UNITED STATES DEPARTMENT OF THE INTERIOR Harold L. Ickes, Secretary FISH AND WILDLIFE SERVICE Ira N. Gabrielson, Director Fishery Bulletin 35 STUDIES ON THE STRIPED BASS (Roccus saxatilis) OF THE ATLANTIC COAST By DANIEL MERRIMAN From FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Volume 50 UNITED STATES GOVERNMENT PRINTING OFFICE WASHINGTON : 1941 For sale by the Superintendent of Documents, Washington. D. C. ------- -- -- Pric* 20 cents ABSTRACT The results of an investigation of the striped bass (Roccus saxatihs) of the Atlantic coast, from April 1, 1936, to June 30, 1938, are discussed and the systematic characters of the species described in detail on the basis of the literature and material afforded by fin-ray, scale, and vertebral counts, and by measurements on more than 350 individuals. Studies on the fluctuations in abundance of this species over long-term periods show that there has been a sharp decline in numbers. Dominant year-classes have at times raised the level of abundance, but the intensity of the fishery is such that their effects have been short lived. The dominant year-class of 1934 was the largest to be produced in the past half century, although the parental stock was probably as low as it has ever been. There is a good correlation between the production of dominant year-classes of striped bass and below-the-mean temperatures during the periods before, of, and immediately after the main spawning season. The striped bass is strictly coastal in its distribution from the Gulf of St. Lawrence to the Gulf of Mexico, is anadromous, and spawns in spring. Sex ratios in northern waters show that males seldom make up more than 10 percent of the population, while in waters farther south the sex ratios are not so disproportionate. Females first mature as they become 4 years old, males as they become 2 years old. This difference in age at maturity may account for the small percentage of males in northern waters, for the time of the spawn- ing season in the South coincides with the time of the spring coastal migration to the North, which is made up mainly of immature females. The age and rate of growth have been studied by scale analysis and the average sizes of the different age groups, and the growth has been calculated to the eleventh year. Striped bass (3,937) have been tagged, and returns have shown that there is a striking migration to the North in spring, and to the South in fall. The population in northern waters in summer remains static. These migrations do not occur until the bass become 2 years old, and have their greatest intensity off the southern New England and Long Island shores. There is little encroachment by the stock in the Middle Atlantic bight on the populations in the North or South. The available evidence from general observation, tagging, and scale analysis points to the conclusion that the dominant 1934 year-class originated chiefly in the latitude of Cheasapeake and Delaware Bays, and that those fish born as far south as North Carolina contribute directly only a relatively small fraction to the population summering in northern waters. Stomach-content analyses show that bass are universal in their choice of food, a large variety of fishes and Crustacea forming the main diet. It is suggested that the increased bulk and availability of Menidia menidia notata in Connecticut waters late in summer and early in fall are responsible for the increase in, or maintenance of the growth rate of striped bass in this region despite the sharp drop in water temperature at this time. The parasites of the species are discussed and several new host records listed. It is suggested that the bilateral cataracts in a high percentage of individuals bass in the Thames River, Connecticut, are the result of a dietary deficiency. The decline in abundance of the striped bass of the Atlantic coast over long-term periods and its causes are discussed from a theoretical point of view, and it is pointed out that the present practice of taking a large proportion of the 2-year-olds annually is apparently not an efficient utilization of the supply. It also is pointed out that both the fishery and the stock would probably benefit from the protection of these fish until 3 years old, at which time the average individual length is 41 cm. (16 inches), measured from tip of lower jaw to fork of tail. STUDIES ON THE STRIPED BASS (Roccus saxatilis) OF THE ATLANTIC COAST 1 By Daniel Merriman, Osborn Zoological Laboratory, Yale University, formerly Temporary Investigator, Fish and Wildlife Service 1 CONTENTS Pago Introduction 1 Acknowledgments 2 Description of the striped bass 2 Size and range of the striped bass 4 Review of the literature on the life history of the striped bass 5 Fluctuations in abundance of the striped bass V Spawning habits and early life history of the striped bass 15 Sex and age at maturity 20 Page Age and rate of growth 22 Migrations.. 33 Origin of the dominant 1934 year-class 46 Food of the striped bass 52 Parasites and abnormalities of the striped bass 55 Discussion 56 Recommendations 62 Summary and conclusions 63 Tables 66 Bibliography 75 INTRODUCTION The following account of the life history and habits of the striped bass (Roccus saxatilis) is the result of an investigation originally sponsored by the Connecticut State Board of Fisheries and Game, and undertaken by the author. The main objectives of this investigation, throughout its entire course, were to obtain information on the life, history and habits of the striped bass, to study the fluctuations in abundance of this species and their causes, and to accumulate material on the effect of the fishery — both commercial and sporting — on the present supply. The striped bass investigation was begun on April 1, 1936, and was concluded on June 30, 1938. Its headquarters have been the Osborn Zoological Laboratory, Yale University, New Haven, Conn., and, during the summer months, the Niantic River, Conn. — an area where this species is more easily available for study than elsewhere in the immediate vicinity. During the first 3 months the work was financed by a group of Connecticut sportsmen. The Connecticut State Board of Fisheries and Game then supported the investigation through December 31, 1937, and also supplied much of the equipment essential to the progress of the work. By that time it had become apparent, as a result of tagging experiments, that the striped bass was a highly migratory species, and that therefore the problem was essentially coastwise in its scope. Clearly the objectives could not be accomplished satisfactorily by studies in one limitod area. The American Wildlife Institute generously contributed a sub- stantial sum in March 1937 when a break in the continuity of the work would have been a severe blow to its progress, and thus made it possible for the investigation to extend its scope to include a large portion of the Atlantic coast. On July 1, 1937, the United States Bureau of Fisheries insured the financial backing of the investiga- tion for a full year from that date, and the State Board of Fisheries and Game appro- priated a sufficient amount for the continuation of the work within Connecticut. ' The Fishery Bulletin of the Fish and Wildlife Service is a continuation of the Bulletin of the Bureau of Fisheries, which ended with vol. 49. The Fish and Wildlife Service was established on June 30, 1940, by consolidation of the Bureau of Fisheries and the Bureau of Biological Survey. 1 277D89— 41- 744L18 2 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE The North Carolina State Department of Conservation and Development also con- tributed to the striped bass investigation in the fall of 1937, and thus made it possible to accumulate valuable information from the Albemarle Sound region in November 1937 and March, April, and May, 1938. The author has published a preliminary account of the results of the striped bass investigation through December 1936 (Merriman, 1937a). A review covering much of the same material has also appeared in the Transactions of the Second North American Wildlife Conference (Merriman, 1937b), and a paper given at the New England Game Conference on February 12, 1938, and the Third North American Wildlife Conference on February 14, 1938, was published later (Merriman, 1938). Several progress reports submitted to the Connecticut State Board of Fisheries and Game have been mimeographed and sent out in limited numbers. This bulletin, therefore, incorporates some previously published material as well as the main accomplishments of the investigation from its inception to its conclusion. ACKNOWLEDGMENTS Since the author was a graduate student in the Department of Zoology at Yale University during the whole course of this investigation, the facilities of the Osborn Zoological Laboratory were always at his disposal. He especially wishes to acknowl- edge the help and advice of Prof. A. E. Parr, Director of the Peabody Museum. He is also indebted to Mr. Marshall B. Bishop of the Peabody Museum for his excellent work in the field in North Carolina in the spring of 1938, to Mr. Donald L. Pitcher of the Bingham Oceanographic Laboratory, and to many members of the Osborn Zoological Laboratory and the Peabody Museum for then- assistance at various times. Furthermore, the investigation owes much of its progress to Mr. Otto J. Scheer, of New York, who made it possible to tag striped bass at Montauk, L. I., N. Y., in the spring and fall of 1937, to Mr. J. D. Chalk, Commissioner of Game and Inland Fisheries in North Carolina, to Mr. David A. Aylward and Mr. Oliver H. P. Rodman of the Massachusetts Fish and Game Association, and to a number of commercial fishermen and sport fishermen's clubs. It is also a pleasure to acknowledge the assistance of Mr. Earl E. Sisson, who was employed by the Connecticut State Board of Fisheries and Game to aid in the seining and tagging of striped bass. And finally, the writer wishes to express his sincere thanks to his wife, who has done most of the recording in the field and has given her support in every possible way. DESCRIPTION OF THE STRIPED BASS During the past few years the striped bass has been called Roccus saxatilis and Roccus lineatus. These two specific names have been used about equally in the liter- ature, and with more or less indiscrimination. Jordan, Evermann, and Clark (1930) say: This species is usually called Roccus lineatus after Sciaena lineata Bloch (Auslandische Fische, VI, 1792, 02); but it cannot be the same. The form, serrae of the preopercle, and the stout spines of the fin, as well as the asserted locality 'Mediterranean' indicate that the species concerned is Dicentrarchus lupus of Europe. The only resemblance to Roccus is found in the striped color; but Bloch says that the stripes on the sides are yellow. A glance at Block's (loc. cit.) illustration substantiates this statement. The name Roccus saxatilis (Walbaum) therefore appears to be the more valid, and lately it has come into more widely accepted usage. Two common names are regularly applied to this species. North of New Jersey "striped bass" is almost universally used, while to the south "rock" or "rockfish" is the generally accepted terminology. Among other names that have been applied in the past, but are seldom if ever heard now, are "green-heads", "squid-hounds" (Goode, 1884), and "missuckeke-kequock" (Jordan, Evermann, and Clark, loc. cit.). The striped bass, Roccus saxatilis, belongs to the family Serranidae, of the order Percomorphi. It has been well described in most of the standard ichthyological ref- erences for both the Atlantic and Pacific coasts (e. g., Hildebrand and Schroeder, STUDIES ON THE STRIPED BASS OF THE ATLANTIC COAST 3 1928; Bigelow and Welsh, 1925; and Walford, 1937), and the following account is based on these works and on the material afforded by fin-ray, scale, and vertebral counts, and measurements on over 350 individuals 15 cm. in length or greater studied during the investigation. The majority of these fish were taken in Connecticut waters. The numbers indicate the extremes of variation, while those in parentheses are the approximate averages. Morphometric description. — Body elongate, moderately compressed; back little arched; greatest depth (at or slightly posterior to origin of spinous dorsal fin) 3.45 to 4.2 (3.7) (young individuals tend to be more slender than old ones), average least depth (at caudal peduncle) 9.6, average depth at anus 3.9— in standard length. Head long and pointed, 2.9 to 3.25 (3.1) in standard length. Dorsal fin rays: IX (VIII in one individual) — I, 10 to 13 (12); fourth and longest dorsal spine 2.2, first and longest dorsal soft ray 2.0 in head. Anal fin rays III, 10 to 12 (11); first and longest soft ray 2.0 in head. Ventral (pelvic) fin rays: I, 5; length of ventrals 1.9 in head. Pectoral fin rays: 15 to 17; length of pectorals 2.0 in head. The two dorsal fins approximately equal in basal length, the first (spinous) being roughly triangular in outline and origi- nating over the posterior half of the pectoral, the second (soft) usually distinctly sep- arate from the first, its soft rays becoming regularly shorter posteriorly. Anal fin of essentially the same shape as second dorsal and slightly smaller; situated below pos- terior two-thirds of second dorsal. Pectorals and ventrals of moderate size; insertion of ventrals slightly behind that of pectorals. Caudal somewhat forked. Scales: 7 to 9 — 57 to 67 — 11 to 15; typically ctenoid (the character "scales on head cycloid" as given by Jordan, 1884, for the genus Roccus, does not hold true in the striped bass) ; extending onto the bases of all the fins except the spinous dorsal. Vertebrae (includ- ing hypural): 24 or 25 (almost invariably 12 + 13 = 25). Gill-rakers on first arch: 8 to 11 + 1 + 12 to 15 (10 + 1 + 14). Eye 3 to 4.9 in head (less in smaller individuals). Mouth large, oblique, maxillary extending nearly to middle of eye (except in smaU individuals) and broad posteriorly (width at tip nearly two-thirds diameter of eye); lower jaw projecting. Teeth small, two parallel patches on base of tongue; also present on jaws, vomer, and palatines. Preopercle margin clearly serrate. Color in lije. — Dark ohve-green to steel-blue or almost black above as a rule, but occasionally light green. Paling on the sides to silver, and white on the belly. Some- times with a bronze luster on the sides. Sides with seven or eight prominent dark stripes, much the same color as the back. Usually the stripes follow scale rows, three or four above the lateral line, one invariably on the lateral line, and three below it. Normally the two above the lateral line, that on the lateral lino, and sometimesthe first below it, are the longest, reaching or coming close to the base of the caudal. None extend onto the head. All except the lowest are above the level of the pectoral fins. The highest stripes and those below the lateral line tend to decrease in length. The stripes are often variously interrupted and broken. Young of less than 6-7 cm. usually without dark longitudinal stripes, and those of 5-S cm. often with dusky vertical cross- bars ranging from 6-10 in number. Vertical fins dusky green to black, ventrals white or dusky, pectorals greenish. Distinguishing characters. — There is little danger of confusing striped bnss above 10 cm. with any other species either on the Atlantic or Pacific coast. Its prominent dark longitudinal stripes, general outline, and fin structure are sufficient to separate it at a glance from other species. The dorsal fins arc usually clearly separate, but sometimes touch. In specimens less than 7 cm. it is often difficult to distinguish striped bass from the white perch (Morone americana), whose dorsal fins are contin- uous — not contiguous, as in the striped bass. The normally separate dorsals of the larger striped bass become an almost useless character here, and the stripes frequently are not present. The general body outlines of the young of these two species are much alike, although the back tends to be somewhat more arched in the white perch. The most valuable differentiating characters are: (1) The second spine of the anal fin, which is almost equal in length to the third spine and more robust in the white perch, and intermediate in length between the first and the third spines and less robust in the striped bass; (2) the relatively thicker and heavier spines in the fins of the white perch; (3) the sharp spines on the margin of the opercle, of which the striped bass has two and the white perch but one; and (4) the soft rays of the anal fin, usually 9 in the white perch and 10-12, normally 11, in the striped bass. 4 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Two fresh-water Serranids bear a superficial resemblance to the striped bass. Morone interrupta, the yellow bass of the Mississippi Valley, also has seven longitudinal dark stripes, but is immediately distinguished by its slight connection of the dorsals, greater depth of tbe body (2.7 in standard length), lesser number of scales in the lateral line (50-54), lack of teeth on tbe base of tongue, and its robust^ spines of the dorsal and anal, as well as the more numerous spines of the first dorsal (X). Lepibema chrysops, the white bass of the Great Lakes region and Mississippi and Ohio Valleys, also has a number of dark longitudinal narrow stripes. Here the dorsals are separate as in the striped bass, but this species differs in having only a single patch of teeth on the base of the tongue, and in having a much deeper body (over one-third of the length) that is more compressed. SIZE AND RANGE OF THE STRIPED BASS The striped bass most commonly taken at present by commercial and sport fisher- men on the Atlantic coast vary in size from less than 1 pound to about 10 pounds in weight. Individuals up to 25-30 pounds, however, are by no means rare, and not infrequently striped bass up to 50-60 pounds are caught, although, judging from old records, these larger fish are not as abundant as they have been in the past. Bass above 60 pounds are now decidedly rare. The largest striped bass taken in recent years was the 65-pounder caught on rod and line in Rhode Island in October 1936 and one weighing 73 pounds was taken on rod and line in Vineyard Sound, Mass., in 1913 (Walford, 1937). Authentic records show that a striped bass weighing 112 pounds was taken at Orleans, Mass., many years ago (Bigelow and Welsh, 1925), and Smith (1907) reports several weighing 125 pounds caught in a seine near Edenton, N. C, in 1891. «*sS .<*? "'~~ 4 --^ Figure 1.— The striped bass (Roccus saxatilis). The striped bass has a range on the Atlantic coast of North America, where it is indigenous, from Florida to the Gulf of St. Lawrence, and is most common from North Carolina to Massachusetts. Jordan and Evermann (1905) state that its southern limit is the Escambia River in western Florida, on the Gulf of Mexico. Jordan (1929), however, states that the striped bass exists as far west as Louisiana. Bean (1884) records the striped bass from the Tangipahoa River, near Osyka, Miss., and this river also flows through Louisiana. Gowanloch (1933) also mentions the striped bass in his "Fishes and fishing in Louisiana." The striped bass was introduced on the Pacific coast where its present center of abundance is the San Francisco Bay region (Scofield, 1931), and the extreme limits of its distribution are Los Angeles County, Calif., and the Columbia River (Walford, loc. cit.). Walford also states: "There is an indigenous population of bass at Coos Bay, Oreg., about 400 miles north of San Francisco." STUDIES ON THE STRIPED BASS OF THE ATLANTIC COAST 5 This fish is strictly coastwise in its distribution, and records of its being taken more than several miles offshore are extremely rare. It is most commonly taken in salt water, but, since it is anadromous, its capture in brackish and even fresh water is a regular occurrence — particularly during the winter and spring months. It has been taken in the Hudson River as far north as Albany, and is caught in large quan- tities in the Roanoke River at Weldon, N. C, each spring. Temperature appears to play no little part in its distribution (see p. 42), yet the striped bass can be taken at the extreme limits of its range throughout the year. REVIEW OF THE LITERATURE ON THE LIFE HISTORY OF THE STRIPED BASS Mention of the striped bass appears early in American literature. This is un- doubtedly because of its great abundance in times past and its coastal distribution — two factors that made it easily available to the early colonists. Capt. John Smith wrote: The Basse is an excellent fish, both fresh & sake . . . They are so large, the head of one will give a good eater a dinner, & for daintinesse of diet they excell the Marybones of Beefe. There are such multitudes that I have seen stopped in the river close adjoining to my house with a sandc at- one tide as many as will loade a ship of 100 tonnes (Jordan and Evermann, 1905). And one of Captain Smith's contemporary divines wrote: There is a Fish called a Basse, a most sweet & wholesome Fish as ever I did eat . . ... the season of their coming was begun when we came first to New England in June and so continued about three months space. Of this Fish our Fishers take many hundreds together, which I have scene lying on the shore to my admiration . . . (Jordan and Evermann, 1905). William Wood in his New England's Prospect (1635) wrote: The Basse is one of the best fishes in the country . . . the way to catch them is with hooke and line: the Fisherman taking a great cod-line, to which he fasteneth a pcece of Lobster, and throwes it into the sea, the fish biting at it he pulls her to him, and knockes her on the head with a sticke. . . . the English at the top of an high water doe crosse the creekes with long seanes or Basse netts, which stop in the fish; and the water ebbing from them they are left on dry ground, sometimes two or three thousand at a set . . . Such references to the striped bass became increasingly common in the eighteenth and nineteenth centuries, all of them dealing with record catches or the abundance of this species, and extolling the virtues of the bass as a game and food fish. Probably the earliest observations of any consequence on any phase of the life history are those by S. G. Worth, who published a series of papers from 1881 to 1912 on the spawning habits and artificial propagation of the striped bass in the Roanoke River, N. C. (See under section on spawning habits and early life history.) Turning to more modern times, mention is made of the striped bass frequently, but in all the literature dealing with the fishes of the Atlantic coast there is scant information on the life history of this species. Such standard and well-recognized references as Bigelow and Welsh (1925) and Ilildcbrand and Schroeder (1928), sum up the available knowl- edge on the striped bass in a few brief pages. In the past few years, however, the need for further information on this species on the Atlantic coast has resulted in several investigations in different localities, apart from the present work. These have given rise to much interesting material and more general knowledge (e. g., see Vladykov and Wallace, 1937), a great deal of which, however, is yet to be published. Reference to some of this work is made in the following pages. In the last quarter of the nineteenth century striped bass were introduced on the Pacific coast, where they prospered beyond all expectations and soon became the object of an intensive and prosperous fishery conducted by both commercial and sport fishermen. This fishery has been of great importance ever since. The story of this introduction of the striped bass to the Pacific coast is particularly interesting (Throck- morton, 1882; Scofield, 1931, etc.). In 1879 and 18S1 a number of yearling bass were seined in New Jersey, taken across the continent in tanks by train, and planted in San Francisco Bav. A total of only 435 striped bass survived the rigors of these 2 trips. Yet by 1889, 10 years after the first plant, they were caught in gill nets and offered for sale, and in 1899 the commercial net catch alone was 1,234,000 pounds. (i FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE In 1915 the greatest catch in the history of the fishery was made, when 1,784,448 pounds of striped bass were delivered to the markets. Since the World War the annual catch has varied between 500,000 and 1,000,000 pounds. The Division of Fish and Game of California has made thorough studies on the life history of the striped bass, as well as the conservation needs of this species. These have been pub- lished in a long series of papers from 1907 to the present, of which the outstanding publication is that by Scofield (1931). But, because the conditions of the fishery on the Pacific coast differed so much from those on the Atlantic coast, much of the I 1 1 T 1 1 22 : 21 . SIKIfbD BASS - ZO LENGTH-WEIGHT RELATIONSHIP _ 19 * 1936 — 1937 I • - 3» IS . f 17 - T 16 . - 15 14 • 13 12 II . 10 " - . - 8 - . 7 - - 6 ' S - ■'/ ; - 4 " 3 . ,J& ' Z • CENTIMETERS- ZO t 30 MO SOt 60 t TO J BO \SO IOO* I/O t 120 I 130 140 INCHES— 10 15 20 25 30 35 40 45 50 LENGTHS Figuke 2.— Length-weight relationship of the striped bass, based on o26 fish. Measurements are to the fork of the tail. information presented by the Division of Fish and Game of California cannot be applied to the striped bass of the Atlantic. On the Pacific coast the main method of capture was by gill net, and it was easy to eliminate the capture of small fish by regulating the mesh size. At the present time commercial fishing for striped bass is prohibited in California. On the Atlantic coast, however, pound-nets, seines, and other methods of capture are used, and striped bass are taken indiscriminately with a great many other species — a situation which would make it highly impractical and most unfair to the commercial fishermen involved if any attempt were made to control the size categories of striped bass taken in these nets by regulating the mesh size. STUDIES ON THE STRIPED BASS OF THE ATLANTIC COAST Length-weight relationship of the striped bass [Length is stated in centimeters, measured to fork in tail; weight is in pounds] Length 20.-. 21... 22... 23... 24. .. 25. ._ 26..- 27--- 28... 29-.- 30.__ 31..- 32..- 33..- 34... 35... 36-.- 37-._ 38-._ 39-.- 40-.- 41___ 42... 43-.. 44... 45..- 46-.- 47-.- 48-.- 49-.. 50.-. 51--- 52-._ 53... 54-.. 55... 56--- V eight 0.25 .25 .25 .25 . 50 .50 .50 .50 .75 .75 .75 .75 1.00 1.00 1.00 1. 25 1.25 1.50 1. 50 1.75 1.75 2.00 2.00 2.25 2.25 2. 50 2.50 2. 75 3.00 3. 25 3.50 3.75 4.00 4. 25 4. 50 4.75 5.00 Length 57-_- 58.-. 59..- 60--. 61-.- 62... 63... 64... 65..- 66--- 67.-- 68. 69. 70. Weight 5.25 5. 50 5.75 6.00 6.25 6.75 7. 00 7.25 7.75 8.00 8.50 9.00 9.25 9.75 71 10.00 72 10.50 73 11.00 74 11.25 75 11.75 76 12.00 77 12.50 78 13.00 79 13.50 80 14.00 81 14.50 82 15.00 83 15.50 84 16.00 85 16.50 86 17.00 87 17.75 88 18.00 89 18.25 90 19.00 91 19.25 92 19. 75 93 20.25 Length Weight 94 21.00 95 21.25 96 22.00 97 22.50 98 23.00 99 23.50 100 24.25 101 25.00 102 25.50 103 26.00 104 26.75 105 27.25 106 28.00 107 28.75 108 29.25 109 30.00 110 30.75 111 31. 50 112 32.25 113 33.00 114 34.00 115 35.00 116 35. 75 117 36. 75 118 37.50 119 38.50 120 39.50 121 40.50 122 41.50 123 42.25 124 43.25 125 44.25 126 45.25 127 46.25 128 47.25 FLUCTUATIONS IN ABUNDANCE OF THE STRIPED BASS Quotations from early settlers point to the enormous abundance of striped bass in those times. Nor is it difficult to find records of unusual catches in the past century. Thus Caulkins (1852) says in a footnote: Four men in one night, (Jan. 5th, 1811), caught near the bridge at the head of the Niantic River with a small seine, 9,900 pounds of bass. They were sent to New York in a smack, and sold for upwards of $300. (New London Gazette.) A quotation from a letter written by a well-known sportsman to the author, dated August 16, 1937, in which he tells of surf-casting for striped bass in the early 1900's at Xlontauk, Long Island, N. Y., reads as follows: As for quantities, almost any time through late summer and into late October, provided one knew the ropes, one could, almost literally, fill a wagon, although 1, myself, seldom continued beyond local give-away — that is, vintil necessity more or less compelled me to become a rod-and-rcel market fisherman, and I fished like one: on one occasion to the tune of just under a ton of fish in a single period of seven days. And even in the last 2 years, when the dominant 1934 year-class of striped bass appeared along the better part of the Atlantic coast, catches reaching extraordinary proportions have been commonplace. As but one example, it is of interest to mention that 90,000 pounds of striped bass were taken by a single trap in 2 weeks in October 1936, at Point Judith, R. I. 8 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Close examination of the available records reveals that the abundance of striped bass on the Atlantic coast has shown tremendous fluctuations over a long period of years. As will be shown below (see p. 13), this is because the striped bass is subject to year-class dominance, a phenomenon which has received increasing attention in the past quarter century, since it has been found to apply to so many different species. Briefly explained, year-class dominance may be said to be the production of such unusually large quantities of any species in a single year that the members of this age- group dominate the population for a considerable period, and are noticeably more abundant than the individuals produced in the preceding and following years. Such dominant year-classes usually make their appearance only at fairly lengthy intervals. Year-class dominance in any species does not, of course, insure the maintenance of the population at a consistently high level. It is also clear that dominant year- classes are often produced by a comparatively small parental stock (see p. 14), and that therefore — at least down to a certain point — their appearance is not correlated with an unusual abundance of mature and spawning fish. There may even be an inverse correlation between these two factors— that is, a large production in any season by a comparatively small population of mature individuals. Such a correlation has been suggested by Bigelow and Welsh (1925) for the mackerel (Scomber scombrus), the "years of great production always falling when fish are both scarce and average very large ..." This phenomenon is probably most common in particularly prolific species that produce a large number of eggs. Such a species is the striped bass, and such a production of a dominant year-class took place in 1934 (see p. 11). In the case of the striped bass a study of the size of the stock over short-term periods may, therefore, be most deceptive. Thus the first manifestation of a large year-class might give the impression of increasing abundance, or, if the study started shortly after an exceptionally productive year, a sharp decline in the population would be apparent under the conditions of the existent intensive fishery. To get a true picture of the trend in abundance, it is therefore essential to study the fluctua- tions over long-term periods. Accurate catch records, which form the most reliable means of studying the rel- ative size of the population in different periods, are unfortunately not available farther back than the latter half of the nineteenth century. Bigelow and Welsh (1925), however, state: "... that a decrease was reported as early as the last half of the eighteenth century." Nor is it surprising that such a decline was noticed so long ago°when it is considered that the striped bass is a strictly coastwise species, and one that is easily available throughout the year. If haddock (Melanogrammus aeglefinus) (Herrington, 1935), halibut (Hippoglossus hippoglossus) (Thompson and Herrington, 1930), and other offshore fishes have become scarcer through the in- tensity of fishing, and this is admitted, it is much more likely that a purely coastal species such as the striped bass, which is far more accessible and therefore unceas- ingly the object of fishermen's attention, should soon have shown a marked decrease in numbers. Also, the availability of the striped bass and the resultant heavy drain on the stock is not the only factor involved. Since this fish is anadromous, there has been every chance for civilization to do irreparable damage to valuable spawn- ing areas. There is abundant evidence to show that such destruction has often occurred (see p. 16). In view of these facts it was not an unreasonable expecta- tion that the supply should soon have diminished, and that in spite of the produc- tion of dominant year-classes the stock could not be maintained at its original high level. Even in the absence of catch records or figures to prove the point, there can be no question but that the numbers of striped bass along the Atlantic coast have de- creased during at least the past 2 centuries. There have undoubtedly been periods when the population showed sudden and pronounced increases, presumably due to the presence of unusually good year-classes. But these peaks have probably been short-lived, and the general trend over long periods has been downwards. Two series of accurate catch records going back to the latter half of the nineteenth century have been made available to the author. Both of these bear out the above contention and substantiate such a hypothesis. The first record is that of the numbers of striped bass taken annually from 1865 to 1907, on rod and line, by the members of STUDIES ON THE STRIPED BASS OF THE ATLANTIC COAST 9 the Cuttyhunk Club at Cuttyhunk, Mass. 2 A graph of this material is shown in figure 3. (For the annual average poundage of the fish caught and the weight of the largest bass in each year, see table 3.) The most striking fact about this curve is its rapid decline from fairly large numbers to extremely low numbers in the 43-year period that it covers. Unfortunately a rod-and-line fishery such as this one cannot be considered a strictly reliable index of abundance — especially since the members of the club confined themselves to fishing for large bass. Moreover, there is no indication of the intensity of fishing, so that the low numbers in the twentieth century might represent the catch of only a few individuals, while the high numbers before 1880 may be the catch of a much larger group. Therefore, the annual fluctuations in this graph are perhaps not real indications of varving abundance, and the rate of decline may be too steep. Nevertheless, it is difficult to imagine from this evidence that a serious depletion did not take place. Even though such a record, lacking as it does information on the effort expended, cannot represent changes in abundance in detail, there can be little doubt that its downward trend indicates the general decline in abundance over the period it covers. RECORD OF S 1 R PED BASS TAKEN BY MEMBERS OF CUTTYHUNK CLUB, CUTTYHUNK IS. MASS IB65- 1907 NUMBCR5 STRIPED BASS TAfCN FBOM IB65 - 9Q1 Fiqure 3.— Record of tho numbers of striped bass taken by tbe members of the Cuttyhunk Club from 1865 to 1907 (see Table 3). Another record of considerable interest and significance is that of the numbers of striped bass taken in pound-net catches from 1884 to 1937 at Fort Pond Bay, Long Island, N. Y. (see fig. 4 and table 4). From 1884 to 1928 these pound-nets were owned by members of the Vail family, who kept accurate records of the numbers of striped bass caught at each haul. 3 They also indicate the number of traps in opera- tion each year. These varied from 6 to 10, and the catches shown in this graph up to 1928 have been weighted to make them equivalent to a fishing intensity of 10 pound- nets throughout. In 1928 the ownership of these nets changed hands, but the author has been able to complete the records up to the present. 4 Unfortunately no record of the number of pound-nets in operation from 1928 to 1937 had been kept, and al- though this number is known to have varied only from 8 to 12, a small error is thus introduced. The magnitude of the catches is such, however, that this part of the graph — indicated by the dotted line — may be properly considered a reasonably accurate continuation of that before 1929. It is of further interest that these pound- nets have occupied essentially the same position each year over the entire period covered by this record. It is impossible to test the validity of this record as a method of sampling the total population, and thus accurately record fluctuations in abundance that occurred. However, it is probable that it gives a fair indication of the decrease in abundance from 1884 to 1935, and that the 1936 and 1937 peaks give a correct picture of the 1 This record was placed at the author's disposal through the courtesy of Mr. Bruce Crane, Dalton, Mass. ' These records were made available by the U. S. Fish and Wildlife Service and the Bingham Oceanographic Foundation. * These records were made available through the oooperatlon of Capt. Daniel D. Parsons, Montauk, Long Island, N. Y., the present owner. ■- 10 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE magnitude of the increased abundance resulting from the 1934 dominant year-class. The peaks at 1894 and 1895, 1906, and 1922 perhaps also represent good year-classes that bolstered the stock temporarily, but there is no adequate means of checking this, since practically no other records covering the same period are available. Striped bass tend to school heavily, and the presence of several schools might easily form the main part of such a peak as the ones shown at 1906 or 1922 in figure 4. Consequently, it may have been that in these years striped bass were not more numerous, but that one or more large schools hit the traps while on migration and gave a false impression of abundance. In another year the reverse situation might have taken place — that is, that the population was unusually high, but that comparatively few bass happened to strike the pound-nets, thus producing a low point on the curve that is not a true indication of abundance. It is, therefore, best not to assume that these fluctuations represent actual changes in the size of the population — at least not until there is further evidence on this score. STRIPED BASS IN POUND NET CATCHES AT FORT POND BAY, LONG ISLAND, N Y 1884-1937 Figure 4.— Numbers of striped bass taken each year in the pound nets at Fort Pond Bay, L. I., N. Y., from 1SS4 to 1937. The fish- ing intensity has been equalized throughout (see Table 4). The peak years mentioned by Bigelow and Welsh (1925) for the catches from Boston to Monomoy, Mass., from 1896 to 1921, show some discrepancy with those in figure 4. In this area 1897 and 1921 were years in which exceptional catches were made. It will be noticed, however, that these years are close to the peaks at 1895 and 1922 shown in figure 4. It may therefore be true that dominant year-classes were present from 1895 to 1897, and in 1921 and 1922, but that they made their presence felt in successive years in somewhat different areas. The peaks at 1936 and 1937, however, are no doubt reasonably accurate indica- tions of the increased abundance in those years. In 1936 the enormous numbers of striped bass that appeared along the Atlantic coast were mainly made up of fish 2 years old, the age at which this species first makes its appearance in the commercial and sport fishermen's catch in Long Island and New England waters. In 1937 a large proportion of the population along the Atlantic coast was composed of 3-year-olds. STUDIES ON THE STRIPED BASS OF THE ATLANTIC COAST 11 The increased abundance in these 2 years was due, therefore, entirely to the 1934 year- class. This group of fish is treated in some detail in the section on age and rate of growth (p. 26), but a glance at figure 5 will sufficiently emphasize the relative abun- dance of the 3-year-olds in 1937. This figure is composed of three length-frequency curves made up from a random sampling of the commercial catch at different localities. Since striped bass 3 years old ranged in size roughly from 35 to 55 cm. (peak at 40 to 45 cm.) during the period these samplings were made, it is evident that the great majority of the catch was made up of 3-year-olds. LENGTH FREQUENCIES OF STRIPED BASS MAKING UP COMMERCIAL CATCHES IN CAPE COD BAY (A), AT NEWP0RT,R.I.(6), AND AT MONTAUK, L. I. (C) , 1937 RANDOM SAMPLING OF STRIPED BASS SEINED IN CAPE COD BAY, AUGUST 2*4, 1937 RANDOM SAMPLING OF STRIPED BASS CAUGHT IN POUND NET AT NEW PORT. R I , OCTOBER 20 t 21, 1937 RANDOM SAMPLING OF STRIPED BASS CAUGHT IN POUND NET AT MONTAUK, L.I., N 1. OCTOBER 25, 26, 27, 1937 Figure 5.— Length-frequency curves made up from random samplings of the commercial catch in different localities in 1937. Data smoothed by threes in all cases (see Table 5 for original measurements). Additional information on the 1934 year-class is seen in the catch records of a haule-seine fisherman at Point Judith, R. I., from 1928 to 1937. 6 (See figs. 6, 7, and 8.) Not only were the numbers and approximate poundage of the fish taken at each haul recorded, but also the date of each haul and the number of hauls annually, thus making it possible to equalize the fishing intensity throughout the entire period. The same areas were fished over this 10-year period. The annual catch in numbers of fish and total poundage are shown in figure 6, and the average weight of the striped bass taken each year is plotted in figure 7. The small proportions of the catch from 1928 to 1935 correspond well with that shown in figure 4, and the tremendous increase • These records were provided through the courtesy of Mr. Chester Whaley, Wakcfleld, R. I. 12 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE in 1936 and 1937 is added evidence on the size of the 1934 year-class. It will be noticed, however, that the decline in the catch in 1937 is not as sharp as that shown in figure 4, probably due to the fact that this seine fishery at Point Judith took a goodly number of 2-year-olds (members of the 1935 year-class) in the spring of 1937. These fish did not make up as large a proportion of the catch at Fort Pond Bay, Long Island, N. Y., during the 1937 season. The records are not sufficiently accurate to permit an exact analysis of the relative numbers of 2- and 3-year-olds in the 1937 I STRIPED BASS BY SEINE OF 1 1 1 \ \ 1928 — 193 7 1 1 \> p 7UNDS of fis * 1 1 1 i ( 1 1 1 1 , c 1 t • \ \ \ 1 / > c s s f \ / / / / <, , $ * i < J i— ' 1928 1929 1910 1931 19)2 93S I9J« 19S5 I9J6 1917 Figure 6.— Annual total catch of striped bass by seine at Point Judith, Rhode Island, 1928-37. Fishing intensity equalized through- out (see Table 6 for original data). catch at Point Judith. The average annual poundage shows, however, that the catch in 1936 was composed mainly of 2-year-olds, and there is a noticeable increase in the average poundage in 1937, due to the dominance of this same 1934 age-group — at that time 3-year-olds. The decline in the average weight of the striped bass making up the annual catches by seine at Point Judith from 1930 to 1936 is quite ST AVERAGE WEIGHT OF THE RIPED BASS MAKING UP THE ANNUAL CATCHES BY SEINE A POINT JUDITH, R.I., 1928-193 T 7 1932 1933 1934 1935 1936 1937 Figure 7.— Average weight of the striped bass making up the annual catches by seine at Point Judith, R. I., 1928-37 (see Table for original data). striking, the drop in this period being from an 8-pound average to a 2-pound average (see fig. 7). European investigators have shown a similar decline in the average annual weight making up the catch following man's intervention on a virgin stock. Thus after the World War, when the North Sea fisheries began to operate again, the larger size-categories were removed first, and in each succeeding year the catch was made up of fish of a smaller average size. In the case of the striped bass, how- ever, the general decline in the average weight from 1930 to 1936 cannot be explained STUDIES ON THE STRIPED BASS OF THE ATLANTIC COAST 13 in the same manner. This is so because although this particular seine fishery at Point Judith was a new one, it was not operating on a virgin stock, for the striped bass is a highly migratory species and is the object of intensive fisheries of different types along the entire Atlantic coast. A more logical explanation is that this down- ward trend in annual average weight over this period was brought about by the de- creasing numbers of large fish that formed the remnant of a dominant year-class produced some years before. That there was a definite decrease in the proportion of large fish making up the catch from 1930 to 1936 is evident from figure 31, in which the percentages of small, medium, and large fish taken in each year are shown. The peak in the annual average weights at 1930 (fig. 7) was caused by the compara- tively great numbers of large fish that made up the catch. Thereafter the composition of the yearly catch showed a decreasing percentage of fish from the larger size-cate- gories (except in 1935). It seems logical, therefore, that a fairly good remnant of a dominant year-class, whose members had attained a large size, existed in 1930, and that in each successive year this remnant became increasingly smaller, thus producing the downward trend in the annual average weight of bass making up the catch in these years. The sharp drop in average weight in 1936 was primarily due to the appearance of the 1934 dominant year-class in the commercial catch. NUMBERS AND SIZES OF STRIPED BASS MAKING UP THE ANNUAL CATCHES BT SEINE AT POINT JUDITH. R I . 1928 - 1937 LEFT COLUMN IN C*CM TtAH IS FOB APRIL • W«I RIGHT COluun IN CAtM VCAR IS FOR INI IH« Figure 8.— Numbers and sizes of striped bass making up the annual eatciics by seine at Point Judith, R. I., 1928-37. The left column in each year is for April and May, and the right column for June to November. The fishing intensity has been equalized throughout. The tremendous numbers of 2-year-olds in this year is well shown in fig. 8. It will also be noticed that there was an exceedingly small percentage of large fish in this year. The increase in annual average weight in 1937 was due to the increase in size of the members of the 1934 dominant year-class— at this time 3-year-olds. If no other dominant year-class comes along for a considerable period of years, it is to be expected that the annual average weight of the striped bass making up the yearly catch will climb steadily to a certain limit, i. e., until the numbers and larger size of the striped bass born in 1934 become insufficient to increase the average weight of the individuals making up the entire catch. If the production of young then con- tinues at a low level, the annual average weight should show a steady decline until the members of another dominant year-class attain sufficient size to start it on an upward trend again. It seems likely that it is the latter part of this cycle that is shown in figures 6 and 7. The question of precisely what caused the appearance of the dominant year-class of 1934 is of especial interest. Judging from the catch records shown in figures 4, 6, 7 and 8, there can be little doubt that this year-class represents the largest production of striped bass on the Atlantic coast in the past half century or more. Yet it is apparent, as has been pointed out, that the parental stock in 1934 was probably as small as it ever as been (see figs. 4, 6, and 8) (the catch in northern waters can be used as an indication of the size of the stock from Massachusetts to Virginia since this species is highly migratorv within these limits). It would seem, therefore, that the production 14 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE of a dominant year-class of striped bass is in no way dependent on the presence of a great number of mature individuals. It is thus necessary to look to other factors for the explanation of this phenomenon. Russell (1932) has pointed out that especially large dominant year-classes were produced in the North Sea in 1904 simultaneously by three different species — herring (Clupea harengus), cod (Gadus morhua 6 ), and haddock (Melanogrammus aeglefinus) . It would seem from this evidence that environ- mental factors apparently play some part in producing these exceptional year-classes. Russell (loc. cit.) has also mentioned the fact that "... there is no necessary con- nection between the number of eggs produced in a particular spawning season and the amount of fry which survives," and it is apparent that environmental factors are most effective in determining the percentage of survival. This is probably especially true in a species with pelagic eggs, a category to which the striped bass essentially belongs (see p. 18). Since the striped bass is anadromous, anything that might affect the rivers in which this species spawns, and the areas in which the eggs hatch and the larvae develop, is worthy of consideration. Unfortunately, the only records that are available are meteorological. Attempts have been made to correlate both tempera- ture and precipitation, since either is capable of seriously influencing the regions where spawning and early development take place, with the prominent peaks shown in the catch records in figure 4. Such a correlation necessarily assumes that the peaks at 1894 and 1895, 1906, and 1922, represent dominant year-classes, and, as has already been mentioned, it is impossible to test the validity of such an assumption. It also takes for granted that these dominant year-classes were produced 2 years before, since striped bass first make their appearance in the commercial catch as 2-year-olds. In the case of the peak at 1936, it is definitely known tbat a dominant year-class was present, and it is further known that the fish that produced this peak were born 2 years before in 1934. Figure 9 shows the deviations from the mean temperature from 1880 to 1935 at Washington, D. C, for February, March, April, and May. Washington DEVIATIONS FROM THE MEAN TEMPERATURES FOR FEB. MARCH. APRIL, AND MAY. IB80-I935. AT WASHINGTON. DC. Figure 9-The deviations from the mean temperature for February, March, April, and May, 1880-1935, at Washington. D. C. The black columns on the base line indicate the years when exceptionally good catches of striped bass were made, and the arrows connect them with the temperatures 2 years before, when in all probability, dominant year-classes were produced. D. C, was chosen because it is in the general latitude of the majority of the important spawning areas for striped bass. The 4 months from February to May were chosen because May is the main spawning season (see below), and because temperatures over this period may well affect the river temperatures as late as May and thereafter. It will be seen from figure 9 that the peak years in the catch record in figure 4 invariably correspond with a below-the-mean temperature 2 years before. It seems likely, there- fore, that dominant year-classes in the striped bass are produced only on a subnormal temperature. On the other hand, a low temperature during the late winter and sprmg months does not necessarily cause the production of a dominant year-class. There are undoubtedly other factors which must concatenate with a subnormal temperature to bring about such a production. It is impossible to state what these factors are, but examination of the precipitation records shows that there is no correlation between rainfall and the dates 2 years before the peaks at 1884 and 1885, 1906, and 1922 shown in figure 4. The inverse correlation between temperature and this catch record, how- ever^ is good. The coefficient of correlation for the entire catch record (1884-1937) and the temperature over this whole period is —.354, which is significant to the 1- percent level. It is thus highly probable that the production of dominant year-classes in the striped bass is quite closely associated with low temperatures. 6 The spelling "morhua," instead of "morrhua" as used by most recent authors, is in keeping with Schultz and Welander (1935) . STUDIES ON THE STRIPED BASS OF THE ATLANTIC COAST 15 In conclusion, it may be said that there is every evidence that over a long-term period the abundance of the striped bass of the Atlantic coast has shown a sharp decline. Dominant year-classes have at times temporarily raised the level of abun- dance, but the intensity of the fishery is such that their effects have been short-lived. This is well shown in figure 4, where it will be noticed that the return to a state ap- proaching the normal low abundance usually follows immediately after the appear- ance of a dominant year-class in the commercial catch. In the 1934 year-class, how- ever, the numbers of striped bass reached such enormous proportions that not only did the 2-year-olds of 1936 dominate the fishery, but the 3-year-olds of 1937 also formed the main part of the catch. None the less, the sharp decline in numbers of bass taken in 1937, as compared with those caught in 1936, is clearly evident, and there can be little doubt that the members of this dominant year-class will be reduced within a few years — under the conditions of the present intensive fishery — to a point where they are negligible. The rate of removal of the different age-groups of the striped bass by the fishery is shown in some measure by the percentage of returns of tagged fish. These percentages are shown in tables 17-20, and 22. It is of inter- est that the extreme in percentage of recapture is seen in the case of 303 fish (pre- dominantly 3-year-olds) tagged and released at Montauk, Long Island, N. Y., in late October 1937. Six months later over 30 percent of these tagged fish had been recap- tured. Furthermore, it is not reasonable to expect that the percentage of tag returns gives a sufficiently great valuation of the rate of removal of the fish of different ages, for, among other reasons, no reward was offered for the return of tags, and it is un- doubtedly true that many of the marked fish that were captured were never reported. It is roughly estimated that about 40 percent of the 2-year-olds of 1936 were taken during their first year in the fishery, and that at least 25-30 percent of the remaining 3-year-olds were caught in 1937. This means that a minimum of 50 percent of the 2-year-olds entering the fishery in the spring of 1936 had been removed by the spring of 1938, neglecting the effect of natural mortality. It thus becomes clear why domi- nant year-classes only raise the level of abundance over short periods, and why, in spite of the occasional increases in number, the general trend of the annual catch of striped bass has been downward. Looking to the future, there is no reason to suppose that the increased abundance caused by the 1934 dominant year-class — huge as it was — will produce any lasting effect on the stock. It is more probable that the return to the normally low level of abundance, so characteristic of the years before 19.'5(i, will soon take place, and that only the production of another dominant year-class will raise the population of striped bass to such unusually high numbers. SPAWNING HABITS AND EARLY LIFE HISTORY OF THE STRIPED BASS It is commonly stated in the standard ichthyological references for the Atlantic coast that striped bass are anadromous, spawning in the spring of the year from April through June, the exact time depending on the latitude and temperature (Smith, 1907, and Hildebrand and Schroeder, 1928). Most of the statements on the spawning of this species have been based on a series of papers in which S. G. Worth (1903 to 1912) discussed the problem of artificial propagation and presented many interesting side- lights on the various phases of spawning and early life history from his studies at Weldon, on the Roanoke River, N. C. Although most of the information in Worth's work is fragmentary, his observations are of value because there has been so little work on any part of the Atlantic coast to corroborate and amplify his statements. The work of Coleman and Scofield (1910) and Scofield (1931) on the Pacific coast indicates that striped bass spawn from April through June in the low-lying delta country adjacent to Suisun Bay, Calif., where the water borders between brackish and fresh. The presence of young fry and small striped bass in the brackish waters of large rivers of the Atlantic coast offers proof that this is an anadromous species, and the absence of juvenile and yearling bass along the outer coast indicates that this species does not undertake coastal migrations until they are close to 2 years old. Thus 277589 — 41 2 16 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Mason (1882), Throckmorton (1882), Norny (1882), and Bigelow and Welsh (1925) present interesting accounts of baby bass being taken in various rivers along the coast in the past (Navesink River, N. J.; Wilmington Creek, Del.; Kennebec River, Maine). Hildebrand and Schroeder (1928) record them as being taken in Chesapeake Bay during the summer months, and Dr. Vadim D. Vladykov, while working on the survey of anadromous fishes for the State of Maryland, also took many juvenile striped bass 5-10 cm. in length on the eastern shore of Chesapeake Bay during the summer of 1936. More recently juvenile bass have been taken in the Hudson River by the New York State Conservation Department, and in the Parker River, Mass., by the author (p. 17). There is also some evidence, from the reported cap tine of baby bass, that isolated spawning areas still exist as far north as Nova Scotia. There can be little doubt that striped bass in early times entered and spawned in every river of any size, where the proper conditions existed, along the greater part of the Atlantic coast, and that as cities were built and dams and pollution spoiled one area after another, the number of rivers that were suitable for spawning became fewer and fewer. At the present time there is every indication that by far the greater part of the production of striped bass along the Atlantic coast takes place from New Jersey to North Carolina, and that the addition to the stock from areas to the north is so small as to be almost insignificant and of little consequence. Thus in Connecticut, where there is much evidence— from the statements of old-time fishermen — that striped bass used to spawn, there is now every reason to believe that spawning seldom if ever occurs. During the entire course of this investigation the author has tried innumer- able times in different localities to find juvenile striped bass in Connecticut waters, for since the juveniles are found close to or in areas where the adults are known to spawn, their presence in Connecticut waters would have indicated the probability of spawning occurring nearby. These efforts never met with any success. Most atten- tion was centered on the Niantic and Thames Rivers, especially the latter, because accounts of baby bass having been caught there within the last 50 years are more numerous than for other regions. Areas similar to those where small bass were taken in the Hudson River in the summers of 1936 and 1937, as well as many other likely localities, have been worked with minnow seines and small-meshed trawls that were efficient enough to catch large numbers of young fish of many other species and occa- sionally even adult striped bass. However, the smallest striped bass taken in Con- necticut waters was a small 2-year-old which measured 23 cm. (9 inches). If spawning occurred to any great extent, small fish 3-8 cm. long, comparable to those caught in other areas in the summer, would most certainly have been found. Plankton and bottom hauls taken at weekly intervals in the Niantic River in an area where bass were known to be present from April through November 1936, have failed to reveal the existence of anything that might be construed as evidence that striped bass spawn there. Further than this, not a single ripe fish of this species has been taken by the author in the course of this investigation in Connecticut waters, although many thousands of bass have been handled at all times of year save the winter months. Inquiries among commercial fishermen in New England and Long Island waters show that ripe striped bass have been caught so rarely and at such irregular times in recent years that their presence can be considered nothing more than abnormal. The fact that large fish that showed no signs of even approaching ripeness were commonly taken in the Niantic River during the spring and early summer months, when bass are known to be spawning in other areas, suggests that this species is not necessarily an annual spawner. The impression from the available information is that spawning does not occur in the region investigated, although it is possible that other Con- necticut waters provide proper breeding grounds. Despite the fact that there is no evidence that striped bass spawn in Connecticut waters at the present time, studies in recent years have disclosed two probable spawn- ing areas in other northern waters. In 1936 the New York State Conservation De- partment took large numbers of juvenile striped bass in various localities on the Hudson River from Beacon downstream. A length-frequency curve of these fish is STUDIES ON THE STRIPED BASS OF THE ATLANTIC COAST 17 shown in figure 10. 7 Curran and Ries (1937) in describing the capture of juvenile striped bass in the Hudson River, say: During the survey few adults but many juvenile striped bass were taken throughout the stretch of river from the city of Hudson to New York. Collections of young for the year were taken first on July 20 in Newburgh Bay. At this time they were 2 inches in length and later study of their scales proved that they were 1936 fish. From Newburgh to Yonkers, about 35 miles downstream, they were found in considerable numbers. Gravelly beaches seemed to be the preferred habitat as few were taken over other types of bottom. In night seining over the gravel they were found to be associated with herring and white perch while daytime hauls showed the herring replaced by shad. Nearly every seine haul in which young striped bass were caught brought in white perch as well. The chlorine as chlorides ranged from 10.0-8,560.0 parts per million (water of low salinity) over this stretch of the Hudson River (Biological Survey (1936), 1937). Larger individuals — up to 2 pounds — have been taken in the Hudson asj,far up as Albany. There can be little doubt, therefore, that the Hudson River is a spawning area for striped bass. Their capture by commercial fishermen in April and May in this region, and the not uncommon reports of ripe individuals at this time of year, is added evidence that spawning takes place in the spring in water that is at least brackish and perhaps entirely fresh. On August 4, 1937, the author took three small striped bass in the Parker River, near Newburyport, Mass. These fish were 7.1, 7.6, and 8.5 cm. long, and subsequent LENGTH FREQUENCIES OF JUVENILE STRIPED BASS FROM HUDSON RIVER. NY., JULY 3 TO SEPT 1.1936 V^p^- ■JMmuJ LEN5IM MILLIMETERS Figure 10.— Length-frequency curve of juvenile striped bass from the Hudson River, July 3 lo Sept. 1, 1936. The number of fish making up this curve is 628. The data have been smoothed by threes. The great majority of these fish were taken in late August (see Table 7 for original measurements). examination of their scales showed them to be juveniles. They were taken about 6 miles from the mouth of the river and about 2 miles below the Byfield Woolen Mills, where a dam prevents anadromous fishes from going further upstream. The bottom, on which these fish were seined was mostly mud and sand, with little gravel and a few scattered rocks. The salinity at this point was 10.23 parts per 1,000, and the water temperature at the surface was 25.5° C. and at the bottom 24.8° C. (ebb tide, one-third out). The depth of the river in this area at this time was 8 feet, and the width 40-50 feet. Other fish found in association with these juvenile striped bass were juvenile white perch {Morone americana) , and various Clupeoid species; snapper bluefish (Pomatomus saltatrix) were also included in seine hauls in this region. The Parker River is free from pollution and is strongly tidal all the way to the Byfield Woolen Mills, where a large amount of fresh water empties into it, particularly in the spring. From this point down, the river winds through the Rowley marshes and eventually empties into Plum Island Sound. It has steep sides, and the rise and fall of the tide along the better part of its length is 5-6 feet. The failure to catch more small striped bass in this river, despite several attempts, is probably best explained by the great difficulty of seining in such an area. The steep sides of the banks and the fast tidal current both make it next to impossible to handle a seine efficiently along 1 The entire collection of striped bass made by the members of the Biological Survey in 1936 was placed at theauthor's disposal in February 1938 by Dr. Dayton Stoner, State Zoologist of the New York State Museum at Albany. N. Y. Further than this, Dr. Moore, Chief Aquatic Biologist of the New York Conservation Department, and other members of the staff, gave the author much t n format ion regarding the capture of small bass in the Hudson River, before the results of the Biological Survey of 1936 were published. 18 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE this river. The capture of only three juvenile striped bass, however, is significant, and probably indicates that striped bass spawn in the Parker River. Added evidence that this is a spawning area is seen in the fact that striped bass are known to winter in this river, as is shown by their capture through the ice by bow-net fishermen. It is considered likely that this is an example of an isolated spawning area in northern waters, supported at least in part by a resident population, and possibly added to by migrants from the south in exceptional years. Although this is the northernmost point from which juveniles have been definitely reported in recent years, there can be no doubt that they were commonly taken in the coastal rivers of the Gulf of Maine in old times (Bigelow and Welsh, 1925), and there is good reason to believe that other isolated spawning areas still exist north of Cape Cod. Another area in which juvenile striped bass were taken was in the Delaware River, near Pennsville, N. J. On November 8, 1937, the author was present when the game protectors for the State of New Jersey Board of Fish and Game Commissioners took 104 small striped bass from the intake wells of a large power plant on the Delaware River, where fish of all sorts are regularly trapped against the screens by the strong flow of water, and are removed and liberated in other regions. A length-frequency curve of this material is shown in figure 1 1 . The examination of scales from these fish showed that the bulk of this sampling was composed of yearlings, and that only a few juveniles from about 9.0-12.5 cm. long were present. It is considered probable, there- fore, that the Delaware River region, including some of the smaller streams that enter Delaware Bay, forms another area in which striped bass spawn. LENGTH FREQUENCIES OF STRIPED BASS TAKEN IN DELAWARE RIVER NEAR PENNSVILLE, N J , NOV. 8, 1937 L E N G T H Figure 11.— Length-frequency curve of juvenile and yearling striped bass taken in the Delaware River, near Pennsville, N. J., on Nov. 8, 1U37. The number of fish included in this graph is 104. The data have been smoothed by threes (see Table 9 for original measurements). It has long been known from the observations of Worth (1903 to 1912) at Weldon, N. C, that striped bass spawn in the Roanoke River. The main observations on the eggs and larvae of the striped bass that are recorded in the literature for the Atlantic coast are taken from Worth's papers, and were made during the time that he con- ducted a hatchery at this point. Bigelow and Welsh (1925) sum up the available information as follows: The eggs (about 3.6 mm. in diameter) are semi-buoyant — that is, they sink but are swept up from the bottom by the slightest disturbance of the water — and this is so prolific a fish that a female of only 12 pounds weight has been known to yield 1,280,000 eggs, while a 75-pound fish probably would produce as many as 10,000,000. The eggs hatch in about 74 hours at a temperature of 58°; in 48 hours at 67°. In recent years the hatchery at Weldon has again resumed operations, thus affording an excellent chance for the study of the eggs and larvae of the striped bass. Others have already accumulated detailed information on this subject (Pearson, 1938), and the following material (from data collected in 1937 and 1938) included herewith, is therefore nothing more than a brief account of some of the more interesting highlights of the spawning and early life history of the striped bass. Spawning in the Roanoke River normally occurs in April and May, although occasionally there are a few stragglers that appear as late as June. It is probable that spawning takes place over a good stretch of the river from Weldon down. (Weldon is over 75 miles by river from Albemarle Sound.) At Weldon the river flows about 4 miles an hour, and is approximately 100 yards wide. Water samples taken on March 29, 1937, showed the chlorinity to be less than 5 parts per million (fresh water), the pH 7.7, and the alkalinity 53.1 estimated as milligrams of bicarbonate per liter. STUDIES ON THE STRIPED BASS OF THE ATLANTIC COAST 19 In 1938 the first spawning striped bass were taken at Weldon on April 11, and by May 10 spawning was apparently completed and the fish had left this locality. This was an unusually early and short spawning season, probably due to the abnormally high temperatures during this time. From April 29 to May 11 the water temperature averaged well over 70° F. (21.11° C.) and at one time reached 77° F. (25.0° C). During the spawning season it is a quite common occurrence to see the so-called "rock-fights" described by Worth (1903), and well known to local fishermen on the Roanoke River. These consist of a great number of small males, 1-3 pounds in weight, and apparently only a single female, appearing on the surface and causing a tremendous commotion by splashing about and creating general confusion. Tbe activity is said to be so great that the fish often injure one another quite seriously, and fishermen who catch striped bass when they are "in fight" attest to this fact and to the number of small males, 10-50 as a rule, that take part in such a display with a single female of from 4-50 pounds. Whether or not this is actually part of the spawn- ing act or a form of courtship does not seem to be definitely established, but general opinion favors the former view. There can be little doubt that the spawning fish at Weldon are composed mainly of males, the females probably never making up as much as 10 percent of the population. In May 1938 the examination of 127 individuals taken at Weldon showed but 6 of them to be females, and much the same sex ratio was found to obtain farther down the Roanoke River at Jamesville, N. C, at the same time. There is no reason to doubt the accuracy of Worth's estimates of the number of eggs produced by a single female striped bass. Records kept at the hatchery at Weldon during 1928, 1929, 1931, 1932, 1937, and 1938, show that the number of eggs per female varied from 11,000 to 1,215,000 in a total of 111 individuals examined in this time. The majority of these fish yielded from 100,000 to 700,000 eggs each. Unfortunately the weights of the individual fish on which these counts were made were not taken, but a single female weighing 4% pounds, taken at Weldon on May 4, 1938, produced 265,000 eggs. The eggs of the striped bass average about 1.10-1.35 mm. in diameter when they become fully ripe, and at the time that they are extruded into the water. During the first hour after fertilization the vitelline membrane expands tremendously, thus creating a large perivitelline space. Measurements on a series of 50 eggs that were preserved 1 hour after fertilization in a solution of 7 percent formaldehyde gave an average measurement of 3.63 mm. in diameter, the extremes being 3.24 and 3.95 mm. Eggs similarly preserved at longer time-intervals after fertilization showed the same general measurements. So far as one can judge from preserved specimens, the description given by Bigelow and Welsh (loc. cit.) of the eggs as being semibuoyant fits perfectly. These eggs are undoubtedly swept far downstream by the strong current, and the protection against injury by jarring afforded by the large perivitelline space is probably of no small consequence in the survival of the developing embryos. The speed of development and the time to hatching is of course dependent on tem- perature. At 71°-72° F. (21.7°-22.2° C.) hatching occurs in about 30 hours, while at 58°-60° F. (14.4°-15.6° C.) hatching normally takes place in about 70-74 hours. In view of the fast current in the Roanoke River, and the rate at which the developing eggs are carried downstream, it is reasonable to assume that hatching probably does not take place until they are close to the mouth of the river or even in Albemarle Sound. Figure 12 shows the different stages of development of striped bass eggs and larvae that were reared in the hatchery at Weldon, N. C. These eggs were fertilized artificially and held at a temperature of 70°- 72° F. (21.1°-22.2° C). The photo- graphs of the eggs were taken from above looking down. A side, view would in reality show that the yolk, with the developing embryo and oil globule, lies at the lower pole of the whole egg as it floats normally in the water. The single large oil globule which is imbedded in the surface of the yolk always lies uppermost, and the blastodisc appears on the side of the yolk in an area that is approximately at a 90° angle with the oil globule— not just opposite the oil globule on the lower pole as Wilson (1891) has shown for the sea bass ("Serranus atrarius" — Wilson, loc. cit., now called Cen- tropistes striatus). Hatchine occurred in 30 hours in the lot under observation, and it will be seen in figure 12 (F) that 6% clays later the yolk sac was almost completely absorbed. 20 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE To the author's knowledge, the smallest striped bass that have ever been taken in their natural habitat were seined along the shore of Albemarle Sound from Mackeys to Rea's Beach, N. C, on May 11, 1938. Since the first spawning fish were taken on April 11 in this year at Weldon, it is likely that these individuals were not more than 1 month old. A length-frequency curve of the 85 juveniles taken at this time is shown in figure 14, and it will be seen that they ranged in size from 1.9-3.1 cm., the peak falling at 2.7 cm. The growth of the striped bass from this age on is further discussed in a later section. In general, then, it may be said that all the evidence points to the fact that the striped bass is anadromous, spawning in the spring of the year, the exact time prob- ably depending on temperature and latitude. It is not definitely established, however, how high a salinity the eggs and larvae of bass will tolerate. Considering the wide variation in the type of river in which bass are known to reproduce, it does not seem unlikely that spawning may at times take place successfully in areas where the water is at least strongly brackish and perhaps even strongly saline. Worth (loc. cit.) first noticed that in raising artificially fertilized eggs of striped bass, an apparatus similar to MacDonald jars — in which the eggs are kept in a strong circulation of water — was necessary in order to get a high percentage of normal development. It would seem, therefore, that a fairly strong current is probably essential for the development of the eggs, but that this may be either tidal, such as that in the Parker River, Mass., or mainly fresh water, as in the Roanoke River. Some possible evidence that spawning does not necessarily always take place in waters of extremely low salinity is provided by the irregular and inconstant manifestation of what appear to be distinct spawning marks on the scales of mature striped bass (see p. 24), for it is generally assumed that such marks are only found on fish that enter fresh water. It would be logical to expect that if all striped bass entered fresh water for spawning purposes, spawning marks on the scales would be more common than they actually are. Such spawning marks are, of course, particularly well-known on scales from salmon (Sahno solar), which do not feed to any great extent during their sojourn in fresh water for spawning purposes, and whose scales are probably partially resorbed during this period, thus forming the characteristic spawning mark. It should be pointed out, however, that striped bass undoubtedly do not stop feeding to the same extent or for a similar length of time during spawning. SEX AND AGE AT MATURITY It is impracticable to get large quantities of striped bass for sex determinations and stomach-content analyses anywhere along the Atlantic coast. This is so because this fish is almost universally shipped to market, and frequently even sold to the individual customers, without being cleaned; hence it was not possible to examine the body cavities in large numbers in the wholesale markets. Since there is no valid method of determining sex without inspecting the gonads, the collection of quanti- tative data on this phase of the work was necessarily limited to the study of fish caught on rod and line by sportsmen and cleaned by the author, to a number of small random samplings of bass that were seined during tagging operations, and to a few fish that were examined on different markets as they were being sold. A total of 676 striped bass caught in northern waters (Long Island and New England) from April to November 1936 and 1937 were examined for sex. These fish ranged in size from 25 to over 110 cm., and in age from 2 years old to over 12 years old. Of these 676 fish, only 9.7 percent were males. One hundred and eighty- three of them were 3 years old or more, and only 4.4 percent of these were males. No males above 4 years old have been found hi northern waters. The remaining 493 fish examined were 2-year-olds, 11.8 percent of which were males. Although the number of fish examined for sex is too small to permit any final conclusions, there is little doubt that the number of males in northern waters seldom reaches much over 10 percent of the entire population. And the evidence so far is that the percentage of males is greatest among the 2-year-olds — that age at which this species first under- takes the migration from further south (see p. 44), and appears in large quantities in northern waters; the percentage of males apparently decreases in the age cate- gories above the 2-year-olds. Fish and Wildlife Service. Fishery Bulletin 35 Plate I Figure 12— Six developmental stapes "I strip.! i . . , ui.l larvae raised til t he hatchery at Weldon, N. < ., at a temperature of 70-72° F Hatching occurred at 30 hours. MasnihVtit mils X 8.2 throughout. A. 1 hour alter fertilization. B. I. hours after fertilization. C. 29 hours after fertilization. D. 20 hours after hatching. E. 60 hours after hatching. I 6 days after hatching. Fish and Wildlife Service, Fishery Bulletin 35 Plate 2 Figuke 13 — Sections through immature and mature striped bass ovaries. A. Immature ovary. B. Mature o vary- before the spawning season. C. Mature ovary — approaching full maturity. Magnification throughout -5 to e months X36. STUDIES ON THE STRIPED BASS OF THE ATLANTIC COAST 21 Such a disproportionate number of females to males is of course most unusual, and it seems unlikely that this condition prevails among the total population of the Atlantic coast. The examination of 29 small bass from Delaware Bay in November 1937 showed approximately 45 percent were males. A sample of 126 bass ranging in size from 21 to 42% cm., from Albemarle Sound, N. C, in March and April 1938 was composed of 31.7 percent male fish. There is also evidence that the composition of the spawning populations of striped bass is predominantly male (p. 19). A theoretical explanation of the strikingly low percentage of males in northern waters is included in the section under migrations (p. 44). In studies of the age at maturity, miscroscopic examination of the gonads pre- sented the most plausible method of procedure in northern waters. The fact that ripe 8 individuals were not available in Connecticut precluded the possibility of studying the age groups making up a spawning population. Gonads from 109 female striped bass ranging in size from 32 to 110 cm. were collected at various intervals from April through November 1936 and 1937. Of these, 46 were fixed in Bouin's fluid and slices from the anterior, middle, and posterior region of each one were cleared in toluene. 9 These were sectioned, stained with Delafield's hematoxylin and eosin, and mounted. Samples of up to 50 ova from each of the three regions of the gonads from which slices were taken were then measured by means of an ocular micrometer. It was soon found that samples from the anterior, middle, and posterior parts of each ovaiy contained eggs of the same general sizes, and that there was no significant difference between the ova of these regions, no matter at what stage of development the gonads were. Thereafter only sections from the middle of each ovary were studied. The remaining 63 ovaries from striped bass collected from April through November 1936 and 1937 were preserved in a solution of 10 percent commercial formalin and water. Slices from the middle of each one of these gonads were then macerated mechanically, until the eggs either floated free or could be easily teased from the surrounding epithelium. Samples of up to 50 ova from each ovary were then meas- ured under a dissecting microscope by means of an ocular micrometer. The measure- ments on the eggs from 109 ovaries by these 2 methods gave comparable results throughout. A study of the measurements of the eggs from striped bass of different sizes almost immediately revealed that there were two easily distinguishable types of ovaries. (See fig. 13.) The first type had eggs whose diameters consistently averaged 0.07 mm. There were occasionally eggs as large as 0.18 mm. in diameter, but more com- monly the largest eggs measured 0.11 mm. The second type contained eggs of two definite size categories; there were small eggs of the same size as all those that were seen in the first type of ovary, averaging 0.07 mm. in diameter, and there were large eggs averaging 0.216 mm. in diameter or greater, the extreme size that has been encountered being 0.576 mm. It is a reasonable assumption, especially in view ol Scoficld's (1931) work, that those ovaries containing only small eggs represent im- mature fish, and that those ovaries having eggs of both small and large size come from fish that are mature, in the sense that the large eggs are those that will be pro- duced the following spawning season. A possible criticism of this assumption is that part of the material examined might have been composed of ovaries from fish that had just completed spawning, and that such ovaries might, therefore, contain only eggs of the small size. On the basis of the distinction between mature and immature individuals proposed above, these fish would then be considered immature, a conclu- sion that would be entirely erroneous. There is no evidence, however, that ovaries from fish that had completed spawning immediately before were included in the material. It has already been pointed out that spawning individuals were not found in the waters from which this material was collected, and it is most unlikely that any freshly spawned bass were studied for the purpose of determining the age of ma- turity. Moreover, by far the greater part of the collection of gonads of striped bass of different sizes took place in the summer and fall, by which time spawning is known to be long since past. Another possible criticism of this method of determining the age at maturity of striped bass is that some of the material may have come from fish that were not spawning the following year, for this species is not necessarily an annual 8 The word "ripe" is used throughout to connote flowing milt or eggs. ' Oil of wintergreen and other clearing agents were also used at first, but in general toluene gave the most satisfactory results. 22 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE spawner (see p. 16), and might therefore not have contained eggs of the larger size although the fish were mature. It is considered unlikely, however, that any serious error in the results is introduced by this means. The results from this method of studying the age at maturity indicate that approximately 25 percent of the female striped bass first spawn just as they are becom- ing 4 years old, that about 75 percent are mature as they reach 5 years of age, and that 95 percent have attained maturity by the time they are 6 years old. The average lengths of individuals of these sizes are discussed in the following section (p. 30), and table 10 gives the results of determining the age at maturity of 109 female striped bass of known length by measurements of the diameters of the ova. The examination of spawning individuals in North Carolina in the spring of 1938 gives added evidence on the age at which female striped bass first spawn. Scale samples from 25 fully ripe females of measured length (43 to 78% cm.) were collected in late April and early May. The smallest of these fish was 43 cm. — a bass that was just becoming 4 years old, but was somewhat smaller than the average individual of this age. There were also 5 other individuals from this lot of 25 mature females that were the same age as this smallest fish. Of the remaining 19 fish, 16 were just reaching 5, 6, or 7 years of age, while the other 3 were 8 or 9 years old. During the period when these mature females were encountered, a great many hundreds of smaller females J -4 INCHES i n/« 1 JUVENILE STRIPED i i BASS FROM AL8ERMARLE SOUND, N C. , MAY II, 1938 /wyy ■HL _^§I wSSk, ^V \<^7{«4;'''f"'(«*i< I i i i I l i i i ' 1 Figure 14. — A length-frequency curve of 85 juvenile striped bass taken in Albemarle Sound on May 11, 1938. Data smoothed by threes (see Table 9 for original measurements). from 1 to 3 years old were handled, but none were ever found to be ripe, thus offering further proof that female striped bass do not arrive at maturity until they reach at least 4 years of age. Male striped bass, on the other hand, become mature and first spawn at a much earlier age. A total of 303 ripe males were encountered in late April and early May in the Albemarle Sound region in 1938. The smallest of these was 21.5 cm. long and was just becoming 2 years old, although it was unusually small for a fish of this age. The largest was 51.5 cm. long, and was just becoming 5 years old. Of the 303 ripe males examined, 150 were just becoming 2 years old, and all the remainder, except the largest individual mentioned above, were becoming either 3 or 4 years old. It thus becomes apparent that a large percentage of male striped bass are mature at the time they become 2 years old, and it is probably true that close to 100 percent are mature by the time they become 3 years old. (See Vladykov and Wallace, 1937.) AGE AND RATE OF GROWTH It has been well established in an ever increasing number of species of fish that scales, since they present more or less concentric rings or annuli, may be used for age determinations. It is generally assumed that the formation of a true annulus is caused by the slowing down or almost complete cessation of growth in the winter, resulting in the arrangement of the circuli so that an annulus appears. Actually, in the striped bass, the annulus does not appear in the winter and only becomes evident by April or May. Further than the determination of age, scale analysis has other vitally important applications in studies on the life histories of fishes. It can be used for growth calculations, is often a method for determining the geographical STUDIES ON THE STRIPED BASS OF THE ATLANTIC COAST 23 point of origin of individual fish, and provides a means of studying migrations — e.g., in salmon, Salmo salar (Masterman, 1913), and herring, Clupea harengus (Dahl, 1907) — age at maturity, and the number of times spawning occurs in different individuals. In the case of the striped bass, there had been no previous work on the Atlantic coast to determine the validity of the scale method for age and rate of growth studies, although Scofield (1931) had applied it successfully on striped bass in California. The preliminary examination of scales immediately disclosed the presence of distinct annuli, which were increasingly numerous, the larger the fish from which the scales were taken. Moreover, the number of annuli were normally constant on different scales taken from a single individual. Also the scales taken from 17 fish that were tagged in 1936 and recaptured from May to September of 1937 invariably showed that the formation of an added annulus had taken place in the winter intervening between the dates of release and recapture. In view of this and much other evidence, it seemed that the scale method was definitely applicable to the striped bass. During the course of the investigation scale samples were taken from approxi- mately 7,000 striped bass of measured length. Over 5,000 of these samples have been mounted and studied. It is essential that all scales be taken from the same area on the different fish if they are to be used for growth-rate studies, for the shape and size of scales from different regions of the body vary to a marked extent and thus scale measurements can only be considered comparable if the samples are homologous. FmuRE 16.— Diagrammatic sketch of a striped bass scale to show parts and method of measurement. Hence all scales were taken from the first or second white stripe above the lateral line in the mid-region of the body directly below the gap between the spinous and soft dorsal fins, for it was found that scales from this area were more consistently suitable for study than those from any other place. A single sample generally consisted of 4 or 5 scales. Length measurements of all striped bass were made from the tip of the lower jaw to the fork in the center of the caudal fin, for it became evident in handling live fish which were being tagged that measurements of this type were the easiest to make and the least subject to error. All lengths given in this bulletin are to the fork in the tail, unless otherwise specified. Figure 16 is a graph for the conversion of different types of length measurements. A flat measuring board with vertical head-piece was always used, and measurements were made to the nearest half centimeter. Scale samples were prepared for study by two different methods. The first 600 were mounted on standard 3- by 1-inch slides with %-inch cover-slips, the mounting medium being corn sirup. All the remaining samples were prepared by taking the impressions of the finely sculptured outer surfaces of the scales on transparent cellu- loid. Lea (1918) first showed with herring scales: . . . that all details which are subjected to observation when the scales are used for the pur- pose of age determination and growth calculations, arise from the play of light on the delicately moulded relief forming the outer surface of the scales (Lea and Went, 1936). Lea produced casts, or imprints of the outer surfaces of scales in thin celloidin films and found them ideal for study. Nesbit (1934a) devised an efficient method of pro- 24 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE ducing scale impressions that was fast and at the same time gave accurate results. This method has been applied with complete success to striped bass scales. Trans- parent celluloid, acetate base, was obtained in sheets 20 by 50 inches and 0.050 inch thick. It was cut into pieces 1 by 2)i inches so that over 100 fitted in an ordinary wooden slide-box of 25-slide capacity. The scale-sample numbers were written on each slide with Volger's Opaque Quick-Drying Ink. The surface of a slide was then softened slightly by spreading a thin film of acetone over it with a glass slide, and the scales making up that particular sample were placed outer surface downward on the area that had been moistened with acetone. The slide and scales were next subjected to pressure under a reinforced seal press having a die approximately 1M inches in diameter. The scales were then removed and the impressions of their outer surfaces were left clearly imprinted on the slide. Measurements on 50 scales from striped bass of all sizes were made before they had been subjected to pressure, and then the impressions of these same scales on transparent celluloid were measured; there was no significant difference in the two measurements. Thus it is clear that no stretching takes place in the scale impression method described above. The ad- vantages of this method are threefold: (1) The cast of the outer surface is easier to 9 9 ID ■a 1 N C HES 25 )0 S3 40 1 1 DIAGRAM FOR CONVERSION a/ • - ■0 (A Hi Z IdSO- M 40- 30- V OF LEN 3TH rf E AS URE M ENTS •y /- S V ''■i ft c / /& O* / / *• k. to % / / A /, /* k / Sj Q X /' ~z O LENG) H TL rot K 0/ TAIl - ctnllME TEftS Fioitre 16. — Diagram for the conversion of different types of length measurements. study than the scale itself because the light does not have to penetrate the fibrillar layers of the scale to show the desired marking; it is also better for photographic purposes. (2) The method is much faster. (3) The cost is far less. All scales, or scale impressions that were studied for age determinations, or on which measurements were made, were first examined under a dissecting microscope, a magnification of about 20 times being satisfactory for most purposes. Those that were measured were then placed in a micro-projection apparatus and the necessary measurements were made on the image, which was magnified 13.75 times. The problem of interpreting annuli correctly at all times in scales from striped bass is somewhat complicated by the occasional presence of accessory, or false annuli. Usually, however, these false annuli are different in structure, so that they are quite often easily recognizable. The false annuli are mainly of two types. The first is a broad accessory annulus that is scarlike in its appearance and is frequently seen on scales from larger fish, extremely rarely on those from smaller individuals 2 or 3 years old. This type of mark invariably appears just outside a true annulus or in close con- junction with it. It seems likely that these are spawning marks, since striped bass are anadromous and spawning occurs in the spring near the time of the formation of a true annulus (pp. 20 and 22). The second type of false annulus has much the same appearance as a true annulus, but is distinguishable on close examination by the STUDIES ON THE STRIPED BASS OF THE ATLANTIC COAST 25 character of the circuli that border it. This type occurs most commonly on scales that overlap a regenerated scale. It appears that the process of regeneration in a scale modifies the growth of adjacent scales sufficiently to form false annuli on the latter. This type was observed frequently, particularly on scale samples from tagged fish that had been recaptured and had regenerated scales in the area from which a sample was taken at the time of their original release. Regenerated scales were common in all samples, often forming at least 10 percent of those examined. _ Sometimes entire samples had to be discarded because there were no scales that were not regen- erated. Up to 15 percent of the samples have been rejected on rare occasions because of false annuli, regenerated scales, and other factors which made the age determinations and scale measurements subject to serious errors. Scales from larger striped bass were found to be much more difficult to read for age than those from smaller individuals. Not only did the first annuli become indistinct, but there were likely to be more false annuli so that age determinations were confusing. For this reason growth calculations by the scale-measurement method have been confined to fish less than 5 years old. Particularly on scales from fish over 8 years old it was almost impossible to be sure that the age reading was correct, and on fish of this size or larger it was only feasible to make approximations as to the age of each individual. As a check on age determinations of striped bass of all sizes the growth rings on otoliths have frequently been counted, and it was foimd that on individuals up to 3 years old this method was satisfactory. The opercular and subopercular bones have also been examined for annular markings, which were best seen after these bones had been cleared in a half-and-half mixture of 5 percent glycerine and potassium hydroxide. On the whole such markings were found to be indistinct and irregular, and did not constitute an adequate means of making age determinations. Since the youngest striped bass taken in Connecticut waters during the course of the investigation were 2 years old, age determinations and rate of growth studies on juvenile and yearling fish were necessarily confined to material from elsewhere. The growth of the larvae has already been discussed under spawning habits and early life history (p. 19). The smallest juveniles that have been taken in their natural habitat have also been described, and, as is shown in figure 14, these fish, which were not more than 1 month old at the time they were seined in Albemarle Sound, averaged about 2.7 cm. in length. Figures 10 and 11 show the range in size of juvenile bass from the Hudson River, and of juvenile and yearling bass from Dela- ware Bay. It is apparent that juvenile striped bass in the Hudson averaged 5-7 cm. in length by the middle of the summer (see fig. 10). The juvenile bass taken in Delaware Bay in November 1937 formed only a small part of the curve shown in figure 11, the bulk of this sample being made up of yearling fish. The juveniles at this time, however, were from 9.5-12.5 cm. long. Growth practically ceases in the winter, and when striped bass become 1 year old in the spring they average 11-12 cm. long. Six yearling individuals taken in the Hudson River in July and August, 1936 and 1937, averaged 14.3 cm. (extremes 12.0-15.9 cm.). The yearlings in the Delaware Bay region (see fig. 11) averaged approximately 19 cm. in November 1937. By the time they become 2 years old striped bass are about 20-23 cm. in length, and it is at this age that this species probably first takes any large part in the coastal migrations. It should be mentioned at this tune, however, that even in juvenile and yearling striped bass there is a tremendous variation about the mean in the meas- urements of any age group at any one time, as can be seen from figure 1 1 . The subject is further complicated since the populations under consideration were from different areas where in all probability slightly different growth rates occur. Thus the lengths given for striped bass of different ages throughout can only be rough approximations. Fish 2 years old and older were sufficiently abundant to give ample material for growth-rate studies in Long Island and New England waters, particularly on the members of the dominant 1934 year-class. Figure 17 shows length-frequency curves of all striped bass measured in Connecticut waters from April through October 1936 and 1937. The prominent peaks that characterize these two curves are mainly made up of the 2-year-olds in 1936 and the 2- and 3-year-olds in 1937, and they give some idea of the relative abundance of the members of the 1934 year-class. The measure- ments that make up these graphs come mainly from seined individuals, but they also come from fish that were caught on rod and line and in pound-nets. Although this 26 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE method of sampling the total population cannot be entirely free from error, it is prob- able that these curves represent the relative proportions of the different size- or age- groups to one another fairly accurately for the general region of the Niantic and Thames Rivers, Conn. The tendency of this species to school heavily, particularly among the smaller size-categories, thus making them more available and easier to catch, may have resulted in an over-emphasis on the relative numbers of the members of the 1934 year-class. And the fact that the larger fish tend to lie among the rocks in or near the surf, in places where they cannot be reached by seining, perhaps pro- vides reason to suppose that these larger fish are not proportionately represented in these graphs. On the other hand, evidence from samplings of the striped bass popula- tion from commercial fishermen's nets in northern waters indicates that the 2-year- LENGTH FREQUENCIES OF ALL STRIPED BASS MEASURED rROM APRIL THROUGH OCTOBER. 1936 LENGTH FREQUENCIES OF ALL STRIPED BASS MEASURED IN CONNECTICUT WATERS FROM APRIL THROUGH OCTOBER, 1937 Figure 17.— Length-frequency curves of all the striped bass measured in Connecticut waters from April through October, 1936 and 1937. The data have been smoothed by threes throughout. See text for further discussion. See Table 11. olds in 1936 comprised over 85 percent of the stock available at this time (see fig. 8) and that the members of this year-class continued to dominate the population in 1937 in spite of the fast rate of depletion of fish of this age due to the highly intensive fishery (see figs. 5, 6, 7, and 8). Evidence from other samplings of the stock in north- ern waters in the summer of 1937 shows that the 2-year-olds of 1937 are apparently represented too strongly in the length-frequency curve for this year (see fig. 17). It is difficult to account for the large proportion of 2-year-olds in the lower graph in figure 17, but it is clear that they were not relatively as abundant in 1937 in all north- ern waters (see fig. 5). It seems probable that the Niantic and Thames Rivers, where most of the fish that make up the length-frequencies in figure 17 were taken, are espe- cially favorable for the smaller sized (2-year-old) bass. The growth by months of the 2- and 3-year-olds seined in Connecticut waters from June through October for 1936 and 1937 is shown in figure 18. It will be seen that the 2-year-olds in June 1936 averaged about 29 cm., and that there was a steady progression in the monthly modes through to October 1936 where the 2-year-olds were roughly 37-38 cm. long. The 3-year-olds in 1936 showed much the same type of growth, the modes of the monthly length-frequency curves for this age-group pro- gressing from 40-41 cm. in June to 48-49 cm. by October 1936. The 2-year-olds of STUDIES ON THE STRIPED BASS OF THE ATLANTIC COAST 27 1937 exhibited approximately the same amount of growth (8-9 cm.) from June through October as fish of the same age in 1936, but it wUl be noticed that they consistently averaged at least 2 cm. larger over this entire period. Thus the modes of the length- frequency curves of the 2-year-olds of 1937 moved from 31 cm. in June to 39 cm. in October. However, the 3-year-olds of 1937, although growing the same amount as fish of the same age in 1936 over an equivalent period of time, averaged 2 cm. smaller throughout, the modes moving from approximately 38 cm. in June to 46 cm. in Octo- ber. The comparison of any of the monthly length-frequency curves in 1936 with its counterpart in 1937 clearly shows that the 2-year-olds in 1937 were distinctly larger than those of 1936, while the 3-year-olds of 1937 were definitely smaller than fish of the same age in 1936. The members of the dominant year-class of 1934 (2 years old in 1936 and 3 years old in 1937) therefore appear to have been below average size. GROWTH OF 2- AND 3-YEAR-OLD STRIPED BASS SEINED IN CONNECTICUT WATERS DURING 1936 AND 1937 CMS "j 1CMFS It Fiourk 18.— The growth of the 2- and 3-year-old striped bass seined in Connecticut waters during 1936 aDd 1937. The curves are smoothed in every case by a moving average of threes. The numbers of fish making up each curve have not been equalized except in that for September 1936, where the total number of fish was divided by three. The dotted line in the June 1937, length-frequency curves is a repetition of curve for the 2-year-olds in October 1936, and is included for the purpose of comparing the 2-year-olds of October 1936, with the 3-year-olds of June 1937 (members of the same year -class) (see Table 12 for original measurements). They were consistently smaller than the fish which were born in 1933 or 1935 were at equivalent ages; both the 1933 and 1935 year-classes were few in numbers by com- fmrison to the dominant 1934 year-class. It is quite clear that this lesser average ength of the members of the dominant 1934 year-class developed before the individuals became 2 years old. The smaller sizes of the individuals making up this dominant age-group agree well with Jensen's (1932) studies on plaice (Pleuronectes platessa) in the North Sea, where it was shown that a strong year-class checks the growth of the fish in this age-group. Jensen (loc. cit.) also points out that the principle of the smaller-than-average size of the individuals making up a dominant year-class, at least in plaice, also appears true from Thursby-Pelham's work, where it is shown that the rich year-class of 1922 was distinguished by a small average length. This is explained by Jensen on the basis of increased competition for food among the members of the 28 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE same size category. Other European investigators, however, have not found that the same phenomenon applies in other species of fish in the North Sea. It is possible that environmental factors, such as low temperatures in the spring and early summer of 1934, played some part in the smaller-than-average size of the members of the 1934 dominant year-class of striped bass. It will be noted in figure 18 that the growth rate of the 2- and 3-year-olds m 1936 and 1937 was fairly steady over the period from June through October. In general, the modes of the length-frequency curves for the 2-year-olds progressed about 2 cm. each month. In October 1936, however, the 2-year-olds appear to have shown an increased growth rate, the mode for this curve having progressed 3-4 cm. beyond that for September. In October 1937 the fish of this age did not exhibit a similarly increased growth rate, but the mode for this length-frequency curve progressed about 2 cm . — an amount about comparable to the growth during the summer months. Since the temperature fell sharply in late September and October in both 1936 and 1937 (see fig. 30), the normal expectation would be that the increase in length at this time would have been less than in the summer months, assuming that the food sup- ply remained constant over this entire period. There are a number of possible ex- planations of this apparently higher growth rate in October. There is some chance that errors in sampling were responsible. Thus it is known that the population was starting to change late in October (see Migrations, p. 37), and there is a slight pos- sibility that fish that had summered farther north, where they apparently grow faster despite somewhat lower average temperatures (see fig. 19) were included in the samples at the end of this month. This does not seem likely, however, for the con- sistent recapture of individuals tagged in this area from June through October gives good evidence to the contrary. Another explanation of the apparently greater growth rate in the fall is suggested by the skewness of the length-frequency curve for October 1936. It will be noted in figure 18 that in all curves for the 2-year-olds, except that for October 1936 the peaks come about midway between the two extremes of the range in size, or below that point. In October 1936, however, the peak falls well above the midpoint between the extremes of size, and there is also a tendency toward the same situation in the curve for October 1937. It may be, therefore, that this apparently greater growth rate is possibly the result of "compensatory growth," the name given by Watkin (1927) to the phenomenon of the smaller fish of a single age group making up a deficiency in size between themselves and the larger fish of the same age group in a relatively short period after having lagged behind for some time. The most probable explanation of the increased growth rate in the faU, however, is that the food supply or its availability increased at this time. The analysis of the stomach contents of striped bass is discussed in a later section of this paper, but for the present it is interesting to consider the fact that this species is voracious in its feeding habits and that it preys on small fish, particularly young menhaden {Brevoortia tyrannus) and shiners (Menidia menidia notata) in Connecticut waters. Both of these species spawn in the spring and early summer, and during July the young are still so small and stay so close to shore that they do not form a large part of the diet of the bass. But by late summer, and particularly early fall, they have increased in size to such an extent that they have added enormously to the available food sup- ply. (For information on the growth rate of Menidia, see Food of the striped bass, p. 53, and fig. 36.) The analysis of stomach contents during September showed that striped bass continually gorged themselves on these small fish to the virtual ex- clusion of other tvpes of food. Furthermore, judging from the relative numbers taken in seine hauls in 1936 and 1937, and from the statements of local fishermen, young menhaden were unusually abundant in Connecticut waters in the latter part of 1936. It is likely that these juvenile menhaden were responsible for the greater growth rate of the striped bass in the fall of 1936, and that the increased availability of the food supply in the late summer each year accounts for the maintenance of or increase in the growth rate through October despite the sharp drop in temperature at this time. As wdl be shown subsequently, there is evidence that the growth rate of the striped bass varies considerably in different localities along the coast. It has already been pointed out, however, that there was a great vaiiation about the mean in measure- STUDIES ON THE STRIPED BASS OF THE ATLANTIC COAST 29 ments of fish from any one region at any one time, and that the samples from different areas may have been composed of stocks from widely separated localities which showed different growth rates. Nevertheless, scale analysis (see Origin of the dominant 1934 year-class, pp. 46-52) points to the fact that the striped bass on which studies were made in northern waters in the summer of 1936 and 1937, were mainly of essentially the same origin and with similar growth rates in their first and second years. Figure 19 shows length-frequency curves for 2- and 3-year-old striped bass taken north and south of Cape Cod in 1937. Those taken north of Cape Cod were from Massa- chusetts, and those south of Cape Cod from Connecticut. The striking difference in the striped bass of the same ages from these two areas is at once apparent. The 2-year-olds north of Cape Cod show a peak at approximately 40 cm., while those south of Cape Cod have a peak near 34 cm. The 3-year-olds from the same areas present peaks at 45 and 40 cm., respectively. It is almost certain that all these fish were of southern origin (see Origin of the dominant 1934 year-class, p. 51), and that they first migrated to northern waters as 2-year-olds in the spring (see Migrations, p. 44). It is possible that the difference in size can be accounted for by differential LENGTH FREQUENCY CURVES OF TWO- AND THREE- YEAR-OLD STRIPED BASS TAKEN NORTH AND SOUTH OF CAPE COD, JUNE- SEPTEMBER, 1937 _ 2YE4RSOL0 3 TEURS OLO Fiqcke 19.— Length-frequency curves of 2- and 3-year -old striped bass taken north and south of Cape Cod from June through September 1937. Data smoothed by a moving average of threes throughout (see Table 13 for original measurements). migration — that is, that the larger fish of the age-categories concerned migrated far- ther north than the smaller individuals. This is unlikely, however, and the difference in size is probably best explained by differential growth rates in the spring, summer, and early fall in the areas under consideration. The samples from these areas are perhaps poor, in that they are composed of rod-and-line caught fish in order that they might be comparable, for it was impossible to get samplings of the population north of Cape Cod over this entire period by any other method. The differences in size may be slightly exaggerated, owing to the fact that the sampling in the early summer south of Capo Cod was somewhat more intensive than that of the middle and late summer, while the sampling north of Cape Cod was evenly distributed throughout the entire period from June through September 1937. There can be little doubt, however, that in 1937 the 2- and 3-year-old striped bass north of Cape Cod grew much faster than those in Connecticut waters from June through September. The average length attained by striped bass each year from the first to the tenth year has been calculated by two different methods, and is shown in figure 20. It is of some interest that these lengths of striped bass at different ages compare almost exactly with those given by Scofield (1931) and Clark (1938) for striped bass on the Pacific coast. Since bass 2 years old and older were available in Connecticut waters in large numbers, it was possible to calculate the average lengths of the differ- ent age groups simply by making age determinations from the scale samples of fish 30 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE of measured length. This has been done on 2,500 fish, and the results are shown by the solid line in figure 20. The average lengths of striped bass from 1 to 4 years old have been calculated from the scales of 4-year-old bass of measured length (see below). This is indicated in figure 20 by the dot-and-dash line. There is every reason to believe from the available samplings of fish of the ages covered by this part of the graph that the lengths derived by this method are accurate estimates. Further than this, it will be noticed that in the center part of the growth curve in figure 20, where the lengths at different ages calculated by both the above-mentioned methods overlap, there is an almost perfect correspondence in the estimated lengths as derived by the two different procedures. It should be emphasized again, in connection INCHES 5 10 IS 20 25 30 ib P GROWTH STRIPED BASS / / / / /_ * • / CALCULATED FROM MEASUREMENTS ON ft OlFFERENT AGE GROUPS 150 SCALES y f FROM MEASUREMENTS ON 2500 FISH .•' X /■ j s / ,.-■■'" Figure 20. — The growth of the striped bass, as calculated from scales and the average lengths of different age groups. 14 for average lengths of striped bass at the time they become 1 year old, 2 years old, etc., to 9 years old. See Table with figure 20, that the lengths represented on this graph are averages, and that there is a wide variation about the mean in the lengths at any age. This is of course particularly true among the larger sizes, as is indicated by the broken line at the upper end of the growth curve. In general, fish 100 cm. (nearly 40 inches) long average about 25 pounds and are about 11 or 12 years old; those 125 cm. (nearly 50 inches) long weigh approximately 50 pounds and are roughly 20 to 25 years old. The largest striped bass taken in recent years (caught in Rhode Island on rod and line in October 1936) weighed 65 pounds and measured 137 cm. (54 inches); examina- tion of several scales leads the author to believe that this fish was 29, 30, or 31 years old. 10 In calculating the growth of striped bass up to 4 years old by the scale method, the following formula was used: L^C+^iL- ■O L x equals the length of the fish at the end of year "x," \\ the length of the scale in- cluded in the annulus of year "x," V the total length of the scale, L the length of the fish from which the scale is taken, and C the length of the fish when scales first appear. (The use of the factor C has various limitations, see pp. 31-32). The measurements on striped bass scales were made from the focus to the anterior edge of the scale and to the annuli along a line that bisected the angle formed by the junction of the two 10 In connection with the age of striped bass, Bigelow and Welsh (1925) write, in the New York Aquarium lived to an age of about twenty-three years." they are certainly long-lived, for one kept STUDIES ON THE STRIPED BASS Of THE ATLANTIC COAST 31 lateral fields at the focus. (See fig. 15.) Scales from striped bass that were beyond their fifth year were not used, since the annuli were often indistinct and it was there- fore difficult to make precise measurements. Van Oosten (1929), Creaser (1926), and others have pointed out that the validity of the scale method of determining the length of a fish at different years in its life depends on 3 main factors: (1) That the scales remain constant in number and identity throughout the life of the fish; (2) that scale growth is proportional to the growth of the fish; and (3) that the annuli are formed yearly and at the same time of the year. Since it has been proved in many other species that scales do maintain their identity throughout the life of the fish, and because there is no evidence to the contrary in the striped bass, it has been assumed that the first requirement holds true. In testing the relation of scale growth to the growth of the fish, the radii of scales from 153 bass of measured length RELATIONSHIP 1 OF SCALE 1 1 "I - GROWTH. TO GROWTH OF STRIPED BASS ' / / / / / /• y ■J'-' '/• / / / / / / 40 SO FISH IN SO CMS LENGTH Fiouee 21.— The relationship of scale growth to body growth in the striped bass (see Table 15 for original data). from 10.5 to 67 cm. were plotted against the lengths of the fish. (See fig. 21.) It will be noted that there is a good straight-line relationship, and that therefore the scale growth may be considered proportional to the growth of the fish within the limits studied. There is no proof, however, that scale and body growth are pro- portional in the smaller sizes below 11 cm., or in the extreme larger sizes above 67 cm. The formation of annuli has already been discussed, and there can be no doubt that they are formed yearly and at the. same time of year — during the winter. Since all the larval stages of development of the striped bass were not available, it was impossible to determine the factor C (that length at which scales first appear on the fish) by careful examination of preserved material. Bass down to 2.0 cm. were collected in the field, and these all showed prominent scales. Individuals up to 0.5-0.6 cm. (approximately 8 days after fertilization of the eggs and 6 days after hatching) were preserved from the hatchery at Eden ton, N. C, and these did not show any signs of scale formation. It was therefore necessary to estimate at what length scales first appear on striped bass between 0.6 and 2.0 cm. by other means. The material that forms the basis of figure 21 was used for this purpose. A regression equation expressing the body-scale growth relationship of the striped bass was 277589 — 41—3 32 j] FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE obtained by means of the product moments method, and it was found that the line intersected the abscissa at 0.6 cm. This value for the length at which scales first appear seems to be too low in view of the evidence mentioned above, but it has been used for the factor C in the scale formula for lack of any other means of determining it more accurately. There is no evidence, as shown before, that scale growth and body growth in the striped bass are proportional in individuals below 11 cm., and an error in the value of 0.6 cm. for C may thus be introduced, since the method applied above necessarily assumes such a relationship. It is considered likely that scales do not first appear until the bass are about 1.0 cm. long, and that scale growth is not directly ANNUAL INCREMENT IN LENGTH OF STRIPED BASS ~°— — _ CALCULATED FROM GR ANO AVERAGE LENGTHS DATA FROM SCALE AGE GROUPS Figure 22.— The annual increment in the length of the striped bass. The annual increments through the fourth year are calculated from the scales from striped bass of the 1933 year-class caught in northern waters in the summer of 1937. The annual increments in the fifth to eighth years inclusive are calculated from the average lengths of the age groups involved, these lengths being taken from fish caught in northern waters in 1936 and 1937 (see Table 16 for actual figures on annual increment). proportional to body growth until a short time after they have formed. But the error introduced in the calculation of the lengths of striped bass at different ages from the scale formula by this discrepancy in the value for C is negligible, and does not affect the points on the growth curve in figure 20 to a significant extent. It should be men- tioned that the use of a constant, C, although superficially plausible, is not sound theoretically. The scale probably does not begin as a geometric point, but as a plate whose radius may weU approximate the size appropriate for the fish at that time. GROWTH OF TAGGED STRIPED 9ASS AS SHOWN 6 MEASUREMENTS AT TIME OF RELEASE AND SUB- SEQUENT RECAPTURE f,... U, I ... I ,.., I „, I ... I ... | ... I ... I .i. I ... I ... I .... I ,„., I ... I „„ l «, I ... I ... I ,t. I ,„ I ... I ... I ... Uil 1936 1937 1938 Fiouee 23. — The growth of tagged striped bass as shown by measurements at the time of release and subsequent recapture. Thus, in the weakfish (Cynoscion regalis) a negative C would be needed to correct for the negative Lee's phenomenon observed (Nesbit, unpublished material). The annual increment in the length of the striped bass is shown in figure 22. It is apparent that the greatest growth occurs in the third year, that age at which this species first undertakes coastal migrations to any great extent. Thereafter the incre- ment in growth falls off sharply, particularly in the fourth year, and from then on maintains an average of about 6.5-8.0 cm. each year at least up to the eighth year. There is some evidence from the available material that the growth rate decreases still more in the eighth and succeeding years. The growth of tagged individuals that were measured at the times of release and subsequent recapture provides a good means of checking on the calculated growth rate of the striped bass as shown in figure 20. This material is shown in figure 23. STUDIES ON THE STRIPED BASS OF THE ATLANTIC COAST 33 Only measurements which came from reliable sources were included in this graph, and the great majority were on fish that were taken at or near the point of release by the author; hence the growth rates refer mainly to fish in Connecticut waters. The lines connecting any two points in this figure of course only represent the total growth in the period intervening between release and recapture. The growths of these individual tagged fish over different lengths of time and in different seasons of the year check well with the growth rates calculated from other material, and in general substantiate the previously discussed information on the growth of the striped bass. It will be noted that the fastest growths occurred in the small fish (2 years old) in the late summer and early fall of 1936, that the growth rates were slow during the winter of 1936-37 (these measurements were in all probability mainly on individuals that wintered in the north), that the growth rates picked up again in the summer of 1937, and that they slowed down once more during the winter of 1937-38. The normally faster growth rate of the 2-year-olds is also indicated by the relative steep- ness of the lines in the smaller size categories. MIGRATIONS There have been no accounts in the literature of the migrations of the striped bass on the Atlantic coast until the present investigation," with the exception of Pearson's (1933) brief paper which was limited to the movements of bass within Chesapeake Bay. There was, however, much evidence to show that this species makes seasonal movements of considerable magnitude. Thus the examination of catch records of commercial fishermen over a period of years at Montauk, Long Island, N. Y., and Newport and Point Judith, R. I., shows that striped bass are caught in large quantities as a general rule only in the spring and fall of the year. This is shown in figure 24, where the bulk of the pound-net catches at Fort Pond Bay, Long Island, N. Y., from 1884 to 1928, were made either in May or October and November. It is also generally known that the date of capture of striped bass along the coast of the Middle and North Atlantic States by pound-nets and seines in great numbers in the spring is progressively later the farther north these catches are made. Moreover, the reverse is true in the fall; for example, the mam catch at Point Judith, R. I., regularly preceds the time that the fishermen on the south side of Long Island make their biggest hauls. It therefore appeared logical to suppose that striped bass undertake definite coastal migrations to the north and cast in the spring, and to the south and west in the fall. Various tagging experiments to demonstrate the time and extent of these migrations have been carried out during the entire course of the investigation. The results of these taggings are summarized in tables 17, 18, 19, 20, and 22. Two methods of tagging have been earned on. External disc tags have been used the greater part of the time, and internal belly tags have also been tried on juvenile and yearling striped bass. Both of these tags were used at the suggestion of Mr. Robert A. Nesbit, of the United States Bureau of Fisheries. The external disc tag is actually a modification of the Scottish Plaice Label, the main changes consisting of reduced dimensions, the use of celluloid instead of hard rubber, the addition of printing, and the substitution of nickel pins for silver wire as the method of attachment. Sketches illustrating these methods of tagging are shown in figure 25. Scale samples were taken in most cases, and lengths and the dates and localities of release were always recorded on all striped bass that were tagged. The external disc tag proved to be a fairly efficient and practical means of marking striped bass. A single tag of this type consisted of two discs of bright red (DuPont No. 6671) celluloid, each 0.025 inch in thickness and one-half inch in diameter, with a center hole ^2-inch in diameter. Each pair of discs bore the same number in black print across the middle, and the necessary instructions to insure their return were printed in black around the circumference. The discs were made by printing on 0.020-inch opaque celluloid and cementing onto the side bearing the printing a " In California, however , tagging experiments on the striped bass have shown that there were "... no definite migrations, simply a diffusion from the locality In which the bass were tagged" (Clark, 1936). 34 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 0.005-inch transparent celluloid, so that the numbers and legends were covered and protected. The first 1,500 tags bore the words, RETURN TO FISH & GAME, HARTFORD, CONN. In the remaining tags this inscription was changed to, RETURN TAG, etc., etc., since it was found that a certain number of returns were being lost because the original wording was sufficiently misleading so that some individuals thought the whole fish should be sent in and were unwilling to part with their catch. Each tag was attached to the fish by means of a pin. This pin was put through the center hole in one disc and pushed through the flesh of the back between the two dorsal fins — one-fourth to one-half inch below the dorsal contour of the body — in a horizontal plane. The matching disc was then put on that part of the pin that POUND NET CATCHES AT FORT POND BAY, LONG ISLAND, N. Y. BY FIVE -YEAR PERIODS 18 84—1928 I -— T 18 9 4-1898 18 9 9-1903 19 4 - I 9 OS 19 09-1913 '914-1918 19 19-1923 1924 — 1926 APR I MflfTjUNt I JULY~ I AUG I f, E P 1 I OCT Figure 24.— Numbers of striped bass caught in the. pound nets at Fort Pond Bay, L. I., N. Y., from 1884 to 1928, for each 5 days during the fishing season, by 5-year periods. The catches have been weighted to make them equivalent to a fishing intensity of 10 pound-nets throughout (see figure 4, table 4). Note that the catches are made only in the spring and fall of the year It is of interest to note that the size of the spring catches has shown a sharp decline over the period covered by this record, while the size of the fall catches has remained about the same during this time. had come through the flesh on the other side of the body, and the pin was crimped over with a pair of finely pointed pliers in such a way that both discs fitted closely against the back of the fish. The printing on the tags was faced out so that it was immediately evident. It sometimes happened, however, that over periods of more than several months Bryozoans and other forms attached themselves to the tags and obscured the printing and even the color of the discs, so that it was necessary to scrape the entire surface with a sharp knife before the inscription became legible. Mussels (Mytilus edulis) over 1 cm. long have been found on the tags at times, and barnacles (Balanus balanoides) covering the entire disc were by no means uncommon. It became evident from the recapture of tagged individuals that it was best to crimp the pin to such a degree that there was less than one-sixteenth of an inch of free space STUDIES ON THE STRIPED BASS OF THE ATLANTIC COAST 35 between the discs and the sides of the fish. If more space was left to allow for growth, sores were created where the edges of the discs rubbed against the body, and weeds were more likely to catch on the tags and cause added irritation. Moreover, since there have been only a few recaptures of fish marked by this method more than a year after the date of release — the longest recovery of a tag of this type was from a fish that was tagged September 7, 1936, in the Niantic River, Conn., and recovered May 2, 1938, in the Hudson River, off Nyack, N. Y. — there is little point in allowing for much growth. In an attempt to preclude any possibility of chafing, both flat and saucer-shaped discs were used. The flat discs showed far less tendency to cause irritation and to pick up weeds and debris, and were in general more satisfactory, although there is some evidence from recaptures in the summer of 1938 that the saucer-shaped discs stay on longer. Two types of pins were used for attaching : 8653 I) (■•■urn lo Slot* ecvl ot runcnet \ one come, mo< Hord . conn J Figure 25.— Sketches to illustrate the external disc and internal belly lag methods ol marking striped bass. the external tags. Those tried with the first 500 bass were stainless steel insect phis. There was abundant evidence in the early work from the subsequent recapture of fish that still showed a scar in the area where they had been tagged with this type of phi, but had lost the tag, that these pins were not adequate in salt water. Not only did they become brittle and fragile after a short time (no fish marked by means of this pin was recaptured more than 2 months after its release), but their slender shafts showed a distinct tendency to cut through the flesh, thus allowing more room for the movement of the tags and causing sores. All these difncultues were fairly well obvi- ated by the use of heavier noncorrosive nickel pins. The nickel pins were made of No. 20 B. & S. pure nickel wire. The diameter of the head of each pin was not less than 0.080 inch in diameter. The pins were ordered in two lengths, 1% and 1% inches, for use in tagging different sizes of striped bass. These pins never showed any tendency to corrode in salt water. 36 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE The external disc tag method of marking striped bass, however, has two definite disadvantages. These are that the evidence from the recapture of fish tagged by this means shows that the discs do not usually stay on for periods much over 1 year; probably because the pins "migrate" toward the dorsal contour of the fish and are eventually sloughed off, and that it is impractical to tag bass less than 8 inches long with discs and pins of the sizes given above. The internal belly tag devised by Nesbit (1934b) has therefore been used on small striped bass (see fig. 25). Since this type of tag has been used successfully over long-term periods with small weakfish (Cynoscion regalis), herring (Clvpea pallasii), and other species, it seemed logical to expect that it was applicable to juvenile and yearling striped bass. This tag consisted of a piece of bright red celluloid 0.030 inch thick, 1% 6 inches long, and % inch wide, with well- rounded ends. One side of the tag bore the number, and the other side the words RETURN TO STATE BOARD OF FISHERIES AND GAME, HARTFORD, CONN., in black print. The printing was made on 0.020-inch opaque red celluloid, and a 0.005-inch transparent celluloid was cemented to each side so that the numbers and legends were well protected. This type of tag was inserted and carried in the body cavity. A small incision was made in the side of the body wall, % to 1 inch in front of the anus with a scalpel. The tag was then pushed through this incision into the body cavity by means of small forceps, so that it lay parallel to the antero-posterior axis of the fish but well on the side of the body cavity where it did not interfere with or displace any of the viscera. Some 581 juvenile and yearling striped bass have been tagged in this manner, and subsequent recaptures have indicated that this method is both feasible and practical with this species, although the returns to date have been few. The advantages of this method over the external disc tags are that it enables the marking of striped bass down to at least 5 inches, and that it is probably a much better long-time tag — although this latter remains to be definitely proven in this species. The only disadvantage of the internal tag with the striped bass is that this species is practically never dressed until it is sold to the individual customer, and since this fish is commonly shipped great distances to market, the tag is likely not to be found until it is difficult to discover the exact locality and date of capture of the fish that bore it. A total of 3,937 striped bass were marked by means of the external disc and internal belly tags from April 1936 to June 1938. Of this number, 2,573 were tagged in Connecticut and Long Island waters. These were all tagged by the external disc method, and were all 2 years old or more, since there are comparatively few areas in northern waters where juvenile and yearling striped bass are available. Returns from fish tagged in this region reached 544 (21.1 percent of the total) by July 1938 and gave abundant proof of a coastwise northern migration in the spring, a relatively stable population showing no movement of any consequence in the summer, and a southern migration in the fall and early winter. In the period from April through October 1936, 1,397 striped bass were tagged in Connecticut waters, of which 337, or 24.1 percent of the total were returned by July 1, 1938. (See fig. 26 and table 17.) In the spring of 1936 these returns showed that an eastward extension from Connecticut to Rhode Island of what undoubtedly was a mass migration to the north, reaching its peak during May in southern New England waters, definitely took place. During late April and May only a few striped bass were tagged, yet returns from the Thames River, Conn., and Point Judith and Newport, R. I., proved that many of these fish were taking part in what the spring catch records of the seines and pound-nets had suggested was a tremendous mass movement to the north. Fish tagged in the Niautic River, Conn., in May were returned from Point Judith and Newport, a distance of 40 to 50 miles in a straight line, 5 to 7 days after their release. The recapture of tagged fish in the summer and early fall showed that the striped bass population in the Niantic and Thames Rivers remained static. Only minor migrations and movements up to 10 miles from the original point of release were recorded from June to October, and it is significant that during the spring, summer, and early fall, there was not a single recapture of a marked bass to the south or west of the areas in which they were tagged. The stability of the popula- tion through the summer and up to the latter part of October was shown by the con- sistent recapture of tagged fish at or near the localities where they were released. An STUDIES ON THE STRIPED BASS OF THE ATLANTIC COAST 37 extreme example of this is that of a bass that bore tag No. 197, which was seined, tagged, and released in June in the Niantic River. This bass was caught in a trap in Niantic Harbor in July and released, caught on a rod and line in the Niantic River in September by the author and released, and caught and released again while seining for tagging purposes in the Niantic River in early October. Returns from tagged striped bass first indicated that a migration to the south was starting in late October, Figure 26.— Chart of the Atlantic coasfcsbowing the migrations of striped bass as determined by the returns from 1,397 Individuals tagged from April through October 1936 (see table 17). when two fish tagged in the Thames River were recovered in the Niantic. Although these fish had oidy moved about 10 miles, they were the first that had ever been taken to the south or west of the original point or release. Almost immediately thereafter bass that had been tagged in Connecticut waters during the summer began to be caught in large quantities in the pound-nets at Montauk, Long Island, N. Y., 38 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE and in seines and on hook and line on the south side of Long Island. The number of returns from Montauk reached a peak during the first 10 days of November. There- after tags were sent in from bass caught progressively farther south as time went on. No marked fish were caught north and east of the original point of release during the fall and whiter, and it was plainly evident from the examination of commercial fishermen's catch records, as well as from tag returns, that an intensive migration to the south had taken place. Scattered returns of tags throughout the winter and early spring months from New Jersey, Delaware, the entrance to Chesapeake Bay, and North Carolina showed that striped bass may go great distances on their southern migration. In 1937 added tagging experiments were undertaken in Connecticut and Long Island waters to obtain additional information on the northern migration in the spring and the return to tbe south in the fall. A group of 103 striped bass were marked and released at Montauk, Long Island, N. Y., from May 15 to 19, 1937, and 14 of these, 13.6 percent were subsequently recaptured. None of these returns came from points to the south of Montauk, all recaptures being in Long Island Sound, on the New York Figuhe 27 .—Migration routes o( striped bass tagged ami released at Montauk, L.I., N. Y.. May 16-19, 1937. The number of fish tagged was 103, the number of returns 14 (13-6 percent of the total). Note that there were no returns from the south, and con- trast with the results of tagging from the same area in the fall as shown in figure 2S (see table 18) . and Connecticut coasts, or from Ehode Island and Massachusetts (see fig. 27 and table 18). Such results gave added evidence that these bass were being tagged near the end of their northern migration, and that an eastward extension of tins movement was still taking place in May and June. From October 25 to 27, 1937, 303 bass were marked and released at Montauk, from the same nets and in exactly the same place as those that were tagged in the spring. Six months later 95, 31.3 percent, of these fish had been reported. The oidy recaptures to the north of the point of release, until the following spring, occurred almost immediately after tagging took place and were so few in number and so minor in scope that they may be considered insignificant. The longest movement to the north that was recorded in the fall was less than 10 miles. On the other hand, recap- tures to the south and west of the area where the tagged fish were released were so numerous as to make it certain that these fish were taking part in an intensive southern migration at that time of year (see fig. 28 and table 19). Many returns in the fall, winter, and early spring months from the south side of Long Island, New Jersey, Delaware, Chesapeake Bay, and North Carolina as far south as Pamlico Sound, indicated the approximate extent and speed of the migration, and further amplified STUDIES ON THE STRIPED BASS OF THE ATLANTIC COAST 39 the results of 1936. The rate at which striped bass may travel south in the fall is shown by the recapture of several fish tagged at Montauk, 450-500 miles away from the point of release, 35-40 days after the date of tagging — an average of 12 miles per day. This distance was measured in a straight line along the coast, which the fish undoubtedly did not travel. Moreover, there is no proof that the fish left the moment they were tagged or were caught at the other end of their migration as soon as they arrived. It seems likely, therefore, that they averaged far more than 12 miles per day. It is of interest that a considerable number of recaptures in the winter and early spring months were from well up large coastal rivers, where spawning occurs in May, thus indicating that some bass probably winter in or near the spawning areas. It is probable that the majority of the spawning individuals in any year do not move into these areas until the late spring, 12 particularly in southern rivers. A total of 770 striped bass were also tagged from April to October in 1937 in the Niantic and Thames Rivers, Conn., and the returns from these further corroborated the results obtained from other marking experiments in northern waters. (See table 20.) There were an insufficient number of fish tagged in April and May to expect .» • -'■ .. -* \ &-£* \ -1 10 20 APRIL \ 10 20 MAY -s- O O 6URFACC, MIANTIC R A A BOTTOM, NiANTiC R A b. OPEN SEA 10 20 JUNE + 10 20 JULY 10 20 AUO 10 20 SEPT H 50°F 10 20 OCT NOV Figure 30.— Water temperatures in the Niantic River, Conn. The surface and bottom temperatures were taken in an area where striped bass were caught throughout the season. The open sea temperatures were taken at the mouth nf the Niantic tuver, where the water passes through a narrow gut on the incoming tide with such force that the surface and bottom temperatures are the same. The open sea temperatures were taken during the spring and fall migrations of the striped bass. Arrows indicate when the.flrst and last bass of the season were caught. Upper graph is for 1936, lower for 1937. catch records and the examination of scale samples that the migration north in 1936 and 1937 at least reached Maine, and that north of Cape Cod the migrants from further south mingled with resident populations that probably had been isolated for some years past. In the summer of 1937 striped bass were taken in large quantities in Nova Scotia, but it is almost certain that there are self-supporting resident populations in various localities along the Canadian coast, and in the absence of length measurements and scale samples it is impossible to be sure of the origin of these fish. Two alternative possibilities suggest themselves in explanation of tbe presence of striped bass in Nova Scotia; first, that these fish are of northern origin and are completely separate from the i! Part of a letter to tbe author from Mr. John R. Webster, of the U. S. Bureau of Fisheries, dated March 8, 1938. reads, ". . . it now seems almost certain that these flsh passed through the Canal. Mr. Churbuck told me the water around State Pier was loaded with bass and thatjieople fished for them all along the banks of the Canal with great success." 42 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE populations farther south, and second, that they are made up of individuals of mixed origin — that is, that the northern stocks are added to by the migrants from the south. The southernmost return of a striped bass tagged in Connecticut and Long Island waters was from the northern tip of Pamlico Sound, N. C. It is probable that the striped bass of the Southern Atlantic Bight — that part of the coast of United States south of Cape Hatteras — are a completely separate population, that may possibly be added to under rare circumstances by the stock from the Middle Atlantic Bight — Cape Hatteras to Cape Cod — and it seems reasonable to expect that the striped bass population of tbe Gulf of Mexico, which presumably extends as far west as Louisiana is entirely isolated. The Middle Atlantic Bight is undoubtedly the center of abundance for the striped bass over its entire range, and tagging experiments indicate that there is compara- tively little encroachment by this stock on the populations to the north and south. This is well in keeping with the conclusions of Parr (1933), who has shown that the shallow-water fish population of the highly heterothermal Middle Atlantic Bight is bounded on the north by a cold-water barrier in the Cape Cod-Nantucket Shoals region in the summer, and on the south by a warm-water barrier at Cape Hatteras in the winter. Parr (loc. cit.) has pointed out that " . . . in neither locality are such barriers found to be a permanent feature during all seasons." But in the case of the striped bass they exist at those times of year when they are most effective in keeping the bulk of the population of the Middle Atlantic Bight from encroaching on the areas to the north or south. Thus the cold-water barrier at Cape Cod in the summer marks the end of the northern migration in normal years, and the warm-water barrier at Cape Hatteras in the winter may play some part in delimiting the extent of the southern migration, and so at least partially separate the populations north and south of this boundary. The question as to how much temperature influences the migration of the striped bass is one of particular interest. This is a highly eurythermal species, yet tempera- ture variations well within the maximum and minimum limits appear to play some part in determining the time of migration. It seems to be more than coincidence that the times when the first striped bass of the year were taken — in April 1936, 1937, and 1938 — and the times that the last ones of the year were caught — in November 1936 and 1937 — in the Niantic River, Conn., were always when the temperature of the water was approximately the same, 6.0° to 7.5° C. (42.8° to 45.5° F.) (see fig. 30). Moreover, the migration of striped bass on the outer coast of North Carolina in late March and early April 1938 was observed to take place over a period when the water temperatures averaged 7.0° to 8.0° C. (44.6° to 46.4° F.). The migrations north in the spring and the return to the south in the fall do not include all striped bass, for this species is caught consistently through the summer in southern waters and not uncommonly in northern waters in the winter. It is a rela- tively small percentage of the stock that remains north in the winter months. How- ever, those that do stay north are of two types — the individuals that form the resident more or less isolated populations of the north Atlantic, and those that may have had their origin farther south but spend an occasional winter in northern waters. The latter may possibly bolster the northern spawning stocks, but are often composed of individuals that are not spawning in that particular year, for this species is not neces- sarily an annual spawner (see p. 16). Striped bass that do remain in the north through the winter months apparently become dormant and inactive in many cases and actually hibernate to much the same extent that lias been described for the black bass (Micropterus dolomieu) in the northern part of its range by Hubbs and Bailey (1938). Their easy capture through the ice by scoop nets and by gigging testifies to their sluggish state in cold water, and the outward appearance of individuals taken in the winter and extremely early spring often shows that they are in poor condition. Striped bass certainly undergo partial hibernation as far south as New Jersey, the extent of this southern limit undoubtedly being determined by the prevailing tempera- tures. Dormant individuals are most commonly taken in northern waters during the winter in shallow bays and in the brackish waters of estuaries. Thus it appears that although temperatures from 6.5° to 8.0° C. play some part in causing the migrations of this species, their effect is not universal. It may be that the first and last fish of the STUDIES ON THE) STRIPED BASS OF THE ATLANTIC COAST 43 season in such a place as the Niantic River, where striped bass are caught so con- sistently at approximately the same temperature in the spring and fall, are mainly winter residents, but it is also known that migratory individuals are present at the times of the earliest and latest catches. It is of interest to note that during October and November 1936, a time which was characterized by sudden drops in temperature, it was plainly indicated that with each cold snap, and resultant decline in temperature of the water, some of the striped bass in the Niantic River moved out and their place was almost immediately taken by fish that presumably came from farther up the coast. Such changes in the population were definitely observed on at least two occasions, both immediately following sharp drops in temperature. Strong winds and storms in the fall also play a part in causing the fish to undertake their migrations. The maximum temperatures for this species appear to be in the neignborhood of 25°-27° C. (77.0°-80.6° F.), for in New England waters in the latter part of August and early September 1937 when there was a protracted period of exceptionally warm weather (see fig. 30), dead bass in considerable numbers were reported simultaneously in Connecticut and Massachusetts. Such mortality occurred chiefly in shallow- water estuaries where the water temperatures reached especially high levels. A number of dead bass were observed by the author in the Niantic and Thames Rivers at this time, and an examination of them disclosed no parasites or injuries that might possibly have been fatal. The water analyses of the Connecticut State Water Com- mission taken at various intervals in the Thames River near New London , Conn. — an area where many dead bass were found — showed nothing unusual nor the presence of any toxic substances during this period (see table 21). There also was a marked migra- tion of bass that normally spend the entire summer in the Niantic and Thames Rivers out to the cooler coastal waters at the time the water temperatures were so high. This was shown by the recapture of tagged fish outside, and by the almost complete absence of bass in the rivers where they are usually found at this time of year. In view of such facts, the evidence is strong that a temperature of 25°-27° C. (77.0°- 80.6° F.) marks the maximum tolerance limit. This is a water temperature which is seldom exceeded over the entire range of the striped bass. It is of some interest to note that although a considerable number of striped bass weighing from 5 to 25 pounds were marked by external disc tags, there have been no returns from these fish save in the immediate locality at which they were released and within a short time after marking took place. Returns of tagged fish from any other area then the general point of release have been confined to individuals not more than 4 years old. It is difficult to account for this circumstance, and, although it may be that the larger bass did not take such a great part in the migrations as the younger individuals, information as to the size-categories appearing in commercial catches in previous years does not make it seem likely that this is an adequate expla- nation. By the same token, it is improbable that the larger fish migrate in waters farther offshore, thus reducing the chances of their being caught along the coast. It is possible that the larger individuals do not carry the external disc tags as well as the smaller fish, and that the tags are not retained for more than a short while. It is true that the larger the bass the nearer the top of the back the pin bearing the tags must be inserted, because the breadth of the fish makes it impossible for pins only 1% inches long to penetrate to the other side far below the dorsal contour. Other reasons for the lack of returns of the larger tagged fish are, first, the overwhelming abundance of the members of the dominant 1934 year-class, and second, the tendency of the smaller size-categories — 2- and 3-year-olds — to school heavily. This schooling instinct, or schooling "synaprokrisis" (Parr, 1937), tends to make them much more available to commercial fishermen than the larger individuals which are not so strongly inclined to congregate together. The heavy schooling of the smaller fish of definite size-categories was observed countless times in the course of seining for tagging purposes in 1936 and 1937. That these schools tend to travel considerable distances without breaking up is suggested by the recapture in several instances at the same time and in the same area some distance away from the original point of release of two or three fish that had previously been tagged in a single seine haid in the Niantic River. 44 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE The recapture of tagged fish as well as observations on the commercial and sports fisheries for striped bass along the Atlantic coast from Maine to North Carolina gives abundant proof that this species is preeminently coastal in its distribution. But studies of the migrations by tagging experiments give convincing evidence that bass do at times cross open bodies of water of considerable size. Thus the spring migration route north apparently takes striped bass from the tip of Long Island straight across to Connecticut and Rhode Island shores, and in the fall the reverse appears to be true — that bass travel from Rhode Island and Connecticut to Montauk and do not follow all the way around the shore line of Long Island Sound. This is shown by the recap- ture of tagged fish at Montauk shortly after their release in Connecticut waters in the fall, and by the almost complete absence of tag returns at any time from the western half of Long Island Sound. A few fish do round Montauk Point and go west along the north shore of Long Island in the spring (see fig. 27), but the majority go to the north and east. Commercial fishermen of long experience in Rhode Island are convinced that in the fall migration to the south a heavy offshore wind causes the main body of fish to go straight from a point at least as far east as Newport to the tip of Long Island, and that a storm from the south causes the bass to follow down the coast of Rhode Island and part of Connecticut before crossing to Montauk. The evidence from the catch records of pound-nets under different conditions in the fall tends to confirm this view. It also is probable that striped bass often cross the mouths of Delaware and Chesapeake Bays in much the same way that they cross the tip of Long Island Sound. It has been pointed out (see p. 20) that approximately 90 percent of the indi- viduals examined for sex in Long Island and New England waters in 1936 and 1937 were females, and it also appears that there is an increasingly smaller percentage of males in northern waters among the large size-categories. On the other hand, this strikingly abnormal sex ratio does not exist in waters farther south, and the following theoretical explanation of this condition is offered. The spring coastal migration to the north in April and May coincides with the spawning season in the south, and is mainly composed of small immature fish and a relatively small number of individuals that are not spawners in that particular year. Because of the discrepancy in the age at ma- turity of the males and females, the males spawning for the first time at the end of their second year while the females do not become mature at least until the end of their fourth year, many of the males do not take part in the spring migration but stay behind to spawn with the larger females. Thus the migration northward at this time of year is largely made up of immature females 2 and 3 years old. The examination of the size-categories making up the catch in northern waters at different seasons indicates that there is a less intensive migration along the coast in June, which is composed of fish of a much larger average size. In all probability these are mainly females which have completed spawning farther south and have moved up along the coast singly or in small groups. This is demonstrated in figure 31 , where the different sizes of striped bass making up the annual catch of a haul-seine fisherman at Point Judith, R. I., be- fore and after June are shown. It is apparent that the small fish make up the bulk of the catch before June each year, but that thereafter bass of the larger size-categories comprise a far greater part of the catch. In 1936 and 1937 an unusually large per- centage of the total were small fish, due to the dominance of the 1934 year-class. There is no evidence that striped bass younger than 2 years old undertake the coastal migrations discussed above. The complete absence of juvenile and yearling individuals anywhere along the coast, save in or close to areas that have been estab- lished as being places where striped bass spawn, is proof that the coastal migrations do not occur until this species becomes 2 years old. In northern coastal waters, where the author handled many thousands of striped bass, individuals less than 2 years old were only encountered on the rarest of occasions. Two interesting tagging experiments were conducted in North Carolina during March, April, and May, 1938. These were carried on for the purpose of determining to what extent the bass from this region take part in the spring migration to the north, and how much they contribute to the population in northern waters during the spring, summer, and fall. This whole question is discussed in some detail under the section on the origin of the dominant 1934 year-class, where evidence is presented STUDIES ON THE STRIPED BASS OF THE ATLANTIC COAST 45 which supports the conclusion that North Carolina does not contribute directly more than a small percentage to the supply summering in the north. In general the results of these experiments substantiate this view as far as they go. In one of the experi- ments a total of 506 juvenile and small yearlings — fish that were just becoming 1- and 2-year-olds — were tagged internally in the general region of the Sutton Beach haul- seine fishery, between the mouths of the Chowan and Roanoke Rivers in the western end of Albemarle Sound, N. C, with the idea that subsequent recaptures of these fish would demonstrate to what extent bass from this region contribute to the popula- tions farther north. These fish were tagged from April 18 to 28, 1938, and 47 were recaptured in the same area before the fishery closed in May. Several others were taken within a short distance of the point of release in the spring, thus indicating that this method of tagging striped bass is satisfactory, at least for short-time returns. It is hoped that the internal tags will also prove satisfactory for long-time returns, as they have in some other species, so that it will be possible to prove the amount of North Carolina's contribution to northern waters over a period of years. The other tagging experiment in North Carolina during March and April 1938, was conducted partially at the extreme eastern end of Albemarle Sound and mostly on the outer coast in the general region of Kitty Hawk and Nags Head. In this experiment, 600 2-, 3-, and 4-year-old striped bass, of which the great majority were 2-year-olds, were marked with the external disc tags. Of these, 62 were caught in the same general PERCENTAGES OF SMALL. MEDIUM AND LARGE STRIPED BASS MAKING UP THE ANNUAL CATCH BY SEINE AT POINT JUDITH, R.I . 1928-1937 LEFT COLUMN IN EACH TEAR IS FOR APRIL + MAY. RIGHT COLUMN IN EACH YEAR IS FOR JUNE - NOV. I9S2 I93S YEARS Figure 31. — The percentages of small, medium, and large striped bass making up the annual catch by seine before and after June at Point Judith, R. I., from 1928 to 1937. The left-hand column is for April and May, and the right-hand column for June to November in each year. See Figure 8 for the same material graphed in terms of actual numbers instead of percentages. area within a short time after they had been tagged, and 46 were again released. By June 15, 1938, there had been 45 returns from these 600 tagged fish from areas some distance away from the point of release. Despite the fact that these fish were tagged at the time of the spring migration to the north, they did not show an intensive one- way movement such as has been proven to take place, for example, in northern waters by tagging in the fall. Thus 24 of the 45 returns were from Pamlico, Croatan, and Albemarle Sounds, indicating that many of the fish tagged on the outer coast moved south and west, some of them being taken in the extreme western tip of Albemarle Sound. The remaining 21 returns came from areas to the north of the point of release; 9 came from the Virginia Beach region; 8 from well into Chesapeake Bay (mainly from the James River and Rappahannock River sections) ; and 4 from more northern wa- ters — 2 from New Jersey, 1 from Wainscott, Long Island, N. Y., and the other from Point Judith, R. I. Had there been a heavy migration to the north at this time from this area, it seems reasonable to expect that in view of the highly intensive fishery for this species as shown by the percentage of recapture from other tagging experiments, there would have been a far greater number of returns from more northern waters. That this tagging experiment was not conducted at a time that was too late to coin- cide with the bulk of the spring migration to the north seems virtually certain, in view of the fact that tagging was started as soon as the outer-coast fishermen began to catch striped bass and was not concluded until the catches had dwindled so that few bass were being taken. Further evidence along this line appears in tables 22A, 22B, 46 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE and 22C, which show that there were no returns from outside the State of North Caro- lina from the small number of striped bass that were released there in March and April, 1937. It does not appear, therefore, from the preliminary results of this work that the North Carolina stock contributes more than a small percentage directly to the summer population in the north. Rather, it seems that the bulk of the northern migration of the striped bass in the spring, and the corresponding return to the south in the fall, takes place between the Chesapeake Bay area and Cape Cod, and that only a relatively small number of migrants from the north and south of these regions take part in these movements. In this connection the author is grateful to Mr. David H. Wallace, of the Chesa- peake Biological Laboratory of the University of Maryland, for giving him the results of a tagging experiment conducted in conjunction with Dr. Vadim D. Vladykov's investigation of anadromous species for the State of Maryland. Of 483 bass tagged from November 15 to 19, 1937, in the east end of Albemarle Sound, in Croatan Sound, and on the outer coast of North Carolina, most of which were yearling and 2- and 3- year-old fish, only 2 had been recovered from northern waters by June 1, 1938, these coming from New Jersey. This is added evidence that North Carolina contributes only a small amount directly to the population summering in northern waters. It is of interest that 1 of these fish tagged on November 15, 1937, was caught in New Jersey on January 16, 1938, showing that some fish migrate north before the spring months. ORIGIN OF THE DOMINANT 1934 YEAR-CLASS The problem of the geographical point of origin of the dominant 1934 year-class, that age-group which has already been discussed at some length, is of particular interest. There is considerable evidence to support the conclusion that these fish were produced mainly in the Chesapeake Bay region. Thus, in the summer of 1935, when the members of this year-class were 1-year-olds and probably averaged 15-20 cm. (approximately 6-8 inches) in length, an unusually great abundance of striped bass of about this size and presumably of this age was observed and reported from Chesapeake Bay by many competent people. Truitt and Vladykov (1936) also "found that fish ranging from 21 to 25 cm. in standard length" seemed to be the most abundant age- category of striped bass in Chesapeake Bay during the early and midsummer in 1936. These fish were undoubtedly 2-year-olds at that time — members of the dominant 1934 year-class. Vladykov and Waflace (1937) also corroborate this information. On the other hand, diligent inquiry ehcited no reports of yearling bass in 1935 from waters farther north. In the light of these observations it therefore seems logical to suppose that this largo group of fish that were 2-year-olds in the summer of 1936, and first appeared in north Atlantic waters in that year, came hi the majority from the Chesa- peake Bay area and that general latitude. (See below for evidence that the dominant 1934 year-class did not come from farther south, p. 49.) From what is now known of the paucity of the spawning areas in the north, it is most unlikely that those, regions north of the latitude covered by Delaware. Bay contributed more than a small fraction to this dominant year-class — or for that matter, that they ever play more than a small and unimportant role in contributing to the total stock along the Atlantic coast under present conditions. Thus it becomes apparent that the striped bass fishery from New Jersey northward is almost entirely dependent for its existence on the stock of bass produced to the south, and on the migrations from the south to the north in the spring, which do not occur until bass become 2 years old or older. Granting that the major portion of the production of striped bass takes place from the northern part of Delaware Bay south, it is of interest to determine how far south the stock contributes to the supply in northern waters, and to what extent different areas contribute to this supply. It is known that the Chesapeake Bay area is an important spawning center, and the work of V. D. Vladykov and D. H. Wallace (as yet unpublished) on tagging striped bass in connection with the survey of anadromous fishes for the State of Maryland has shown that the migration of bass out of Chesapeake Bay to the north in the spring is not an uncommon occurrence. Thus it seems well established that this general region contributes to the supply in the north and is an important center of production. STUDIES ON THE STRIPED BASS OF THE ATLANTIC COAST 47 The question of how much the areas to the south of Chesapeake Bay contribute to the population in the north, and whether or not the dominant year-class of 1934 was produced simultaneously in Albemarle and Pamlico Sounds as well as in Chesa- peake Bay, is of further interest. The author has found no evidence from talking with commercial fishermen in the Albemarle Sound region in 1937 and 1938 that there was an unusually large quantity of yearling bass in 1935 in these waters, as was the case in Chesapeake Bay. Further than this, tagging experiments in March and April in 1938 on the outer coast of North Carolina and in the eastern end of Albemarle Sound tend to show that the bass from this area do not undertake such an intensive migration to the north in the spring, and that they do not contribute a large amount to the summer population in northern waters. It has been pointed out tbat these tagged fish did not show an intensive one-way migration at this time, but rather a diffusion from the point of release with only a small percentage of the fish making definite movements of considerable distance to the north. This was in spite of the fact that these fish were released at exactly the time they would be expected to under- take the spring migration northward, and was in direct contrast to the one-way mass migration southward as shown by tagging hi the north in the fall (see pp. 36-39 and 44-46). It is clear from this information that the stock in North Carolina waters probably contributes only a relatively small percentage directly to the populations summering in the north. There is further evidence from the results of scale analysis that the main source of supply for the summer populations in northern waters is in the Chesapeake Bay area — or at least that general latitude (which includes Delaware Bay), and not from farther south. Unfortunately vertebral counts are of no value in showing the general point of origin of individual striped bass or for racial analysis, for this is a species with a virtually constant number (25) of vertebrae (see p. 3), and therefore the counts show no variation with latitude such as has been shown to occur in other forms (e. g., Hubbs, 1922). Scale and fin-ray counts may possibly be of some use in this respect, but they have not been used in this study because of the impracticably of making such counts, especially where the material was limited and it was desirable to tag a large proportion of the fish that were taken in northern waters. But whereas scale and fin-ray counts were not feasible in conjunction with tagging work, it was perfectly practicable to take scale samples from live fish. For these reasons, and because the scale method has given such successful results in determining points of origin in other species, scale analysis was used throughout for this purpose. The assumption on which such a method rests in a species that spawns over a considerable latitude is that since there are likely to be different environmental factors over the entire range of spawning, there are also likely to be different growth rates which should be reflected in the scales. The problem is, then, to detect these differ- ences in the scales from fish of different latitudes, and to establish that they arc con- stant and therefore good criteria for determining the points of origin of the individuals from which the samples are taken. The striped bass is known to spawn over a wide latitude, and apparently does not migrate along the coast until it becomes approxi- mately 2 j'ears old. Thus, if there are any differences in the growth rate of this species in various localities along the coast, those that are to be used in determining points of origin must be found within that part of the scale bounded by the second annulus. With this in mind, as well as the fact that scale growth is proportional to body growth (see p. 31), the widths of the first and second growth zones of scales from striped bass of known and unknown origin were measured by the method described in the section on age and rate of growth (see fig. 15). Figure 32 shows the length-frequencies of the widths of the growth zones in millimeters on scales from striped bass taken in different localities along the Atlantic coast in 1937. The top three series of length-frequency curves (those from scales from fish taken at (1) Cape Cod Bay, Mass., (2) Harkness Point, Conn., and (3) Mon- tauk, Long Island, N. Y.) are from members of the 1934 dominant year-class — that group of fish whose origin is of especial interest. The samplings of fish from which these three sets of curves come, were made in the summer and fall of 1937 in northern waters. In the three sets of measurements, the widths of the first and of the second growth zones are strikingly alike throughout — a fact which at least suggests 277589 — 41 i 48 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE that the members of the dominant 1934 year-class that visited northern waters in 1937 were of much the same origin. It should be mentioned that measurements of the first and second growth zones on the scales from 2-year-old bass in Connecticut waters in 1936 (members of the 1934 dominant year-class) also gave length-frequency curves that were exactly comparable to those shown in the top three sets of curves in figure 32. Had they been of different origin — from areas scattered along the entire length of the Atlantic coast — it would be expected that the distribution of the length- frequencies of the widths of the first and second growth zones in these cases would have been much wider and not nearly as constant in the range of measurement as they actually are. 1ST GROWTH ZONE CAPE COD BAY HARKNESS PT.CONN MONTAUK.L I . NY 2ND GROWTH ZONE 3RD GROWTH ZONE 4TH GROWTH ZONE AUO 2-4,1937 l SEPT 6, 1937 OCT 26. 1937 NCOWPtfTEWffSIWL ZONE CURRITUCK i MARCH 24.1937 SOUND. H C I COMPLETE MARGINAL ZONE LENGTH FREQUENCIES OF GROWTH ZONES ON SCALES FROM STRIPEO BASS TAKEN IN DIFFERENT LOCAL- ITIES IN 1937 INCOMPLETE MARGINAL ZONE WIDTH OF GROWTH ZONES IN MMS Figure 32.— The length-frequencies of the growth zones on scales from striped bass taken in different localities in 1937. The meas- urements making up each curve have been smoothed by a moving average of threes throughout. One other point is of interest in the length-frequencies of the growth zones on the scales from these fish taken in northern waters in 1937. This is the comparison of the fourth growth zones (incomplete marginal zones) of the samples from Cape Cod Bay and Harkness Point. It has been pointed out in the section on age and rate of growth that there is much evidence that striped bass north of Cape Cod grew much faster than those south of Cape Cod during the. summer of 1937 (see fig. 19 and p. 29). Since scale growth is proportional to body growth (see fig. 21), this phenomenon should be reflected in the scales, and a glance at the length frequencies of the incomplete marginal zones mentioned above (see fig. 32) shows this to be true. Thus the measure- ments of the fourth growth zones of the scales from fish from Cape Cod Bay present a peak slightly in advance of the similar peak for the Harkness Point sample, despite the fact that the sample from Cape Cod Bay was taken more than 1 month earlier than the one from Harkness Point. This is probably best explained by the faster growth rate of the fish summering north of Cape Cod, for if the growth rates were the same, the peak for the Harkness Point sample would have been far in advance of the one for the Cape Cod sample, since it was taken so much later in the summer. STUDIES ON THE STRIPED BASS OF THE ATLANTIC COAST 49 Turning now to the two middle sets of length-frequencies in figure 32, those from scale measurements from fish taken in northern and southern Chesapeake Bay in February and March 1937, it is apparent that these are also from samples of the dominant 1934 year-class at tbe time its members were just becoming 3 years old, and when the third annulus was in the process of formation on the anterior margin of the scale. Looking at the widths of the first two growth zones, it is immediately apparent that the general distribution of the length frequencies and the peaks of the first growth zones and the second growth zones are similar throughout. Furthermore, they coincide almost exactly with the same growth zones of the scales from fish born in the same year but collected at a later date in northern waters — see the top three sets of curves in figure 32. It cannot be assumed, however, although it may well be true, that these samples from Chesapeake Bay are from fish that were produced in that region and had remained there, since it is known that this species often undertakes coastal migrations after it becomes 2 years old. Thus these fish might have moved into Chesapeake Bay in 1936, and might, therefore, not have had their origin in this region. On this account, it is not possible to assert that the similarity in the widths of the first growth zones and those of the second growth zones in the top five sets of curves in figure 32 is proof that the dominant year-class of 1934 originated in Chesa- peake Bay. These similarities do, however, suggest that this is so. Looking at the bottom set of curves in figure 32, those from scales from fish taken in Currituck Sound, N. C, it is again apparent that the widths of the first growth zones are much the same as those for all the other samples in this figure, although they do tend to be slightly less. The widths of the second growth zones of scales of the fish from this area, however, are strikingly different from any that precede it in figure 32. Whereas the widths of the second growth zones of the scales from fish from northern waters and from Chesapeake Bay in 1937 all range from approxi- mately 0.5 mm. to or slightly over 2.0 mm. (with peaks at 1.0 mm.), the widths of the second growth zones of scales from fish from Currituck Sound range from about 2.0 to 3.6 mm. (with a peak at 2.9 mm.). These second growth zones of the scales from fish from Currituck Sound are labelled incomplete marginal zones in figure 32 because the second annuli, although in the process of formation on the anterior mnrgins of the scales, were still indistinct. Therefore, the measurements of the marginal zones are to all intents and purposes equivalent to what those on the second growth zones would have, been had the second annuli been completely formed. It should not be necessary to point out that if there were any differences from this factor, the widths of the second growth zones would have been even greater. There is no doubt that these completely different, and exceptionally wide second growth zones on the scales from fish from Currituck Sound are characteristic of the bass born in that general region in 1935, for these scales were taken from fish that were slightly less than 2 years old, and therefore had not undertaken any coastal migration. Thus the wide second growth zones on scales from fish born in the genera] Albemarle Sound region in 1935 give promise of being a means of distinguishing iisli from this area from those born farther north. And since these wide growth zones are so different from the other growth zones in figure 32, they provide added evidence that the dominant 1934 year-class arose in the general latitude of Chesapeake Bay. They also tend to show that those bass born in North Carolina do not contribute a large proportion of the population that summers in northern waters. On the other hand, the fish that make up the top five sets of curves in figure 32 were all born in 1934, while those that make up the bottom set of curves (Currituck Sound) were bom in 1935; and it should be pointed out that the comparison of the widths of the second growth zones of scales from fish born in different years may be fallacious. Thus there is no evidence from the single sampling in Currituck Sound in 1937 as to whether the wide second growth zone is truly a regional difference that occurs annu- ally, or whether it was only a characteristic of the 1935 year-class. However, scale measurements from samplings of bass of the same age — 2 years old in the spring of 1937— as those from Currituck Sound but taken in different areas, southern New England and southern Chesapeake Bay, appear in figure 33. (Tbe length-frequency curves of the scale measurements of the sample from Currituck Sound shown at 50 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE the bottom of fig. 32 are also repeated for the sake of comparison at the bottom of fig. 33.) These provide proof that the members of the 1935 year-class that contributed to the population summering in northern waters as 2-year-olds in 1937 came, in the main, from the Chesapeake Bay area. Thus the middle set of curves in figure 33 are measurements of the growth zones of scales from fish that were just becoming 2-year-olds in Chesapeake Bay in 1937. They are, in other words, from bass that had not yet migrated to any great extent, and the curve for the second growth zone may therefore be considered typical for bass that had been born inl935 in Chesapeake Bay. The upper set of curves in figure 33 is from measurements of the growth zones of scales from 2-year-old fish taken from northern waters in the summer of 1937. They are from bass of unknown origin that had migrated north along the coast in the spring. It will be noted immediately that the curve for the second growth zone of the scales from northern fish in the summer of 1937 compares well with the similar curve for the bass of the same year-class known to be of Chesapeake Bay origin. LENGTH FREQUENCIES OF GROWTH 20NES ON SCALES FROM TWO-YEAR-OLO STRIPED BASS IN 1937 £ - OCT, I95T ZONE 2-d G»0*TM IONE GROWTH ZONES Figure 33.— The length-frequencies of the growth zones on scales from 2-year-old striped bass taken in southern New England southern Chesapeake Bay. and Currituck Sound (repeated from Figure 32 for comparative purposes), in 1937. The measure- ments making up each curve have been smoothed by a moving average of threes throughout. However, it does not compare well with the similar curve for bass of the same year- class known to be of North Carolina origin. (See lower set of curves, figs. 32 and 33.) There is somewhat of an overlap between the curves of the widths of the second growth zones on scales from fish of the 1935 year-class of known origin from Chesa- peake Bay and North Carolina, so that scales from fish of the same age-group but of unknown origin that show a second growth zone measuring from about 2.0-3.0 mm. might have been born in either of the above-mentioned areas. It is apparent that the majority of the widths of the second growth zones on the scales from fish taken in northern waters in the summer of 1937 fall below 2.0 mm. Judging from these measurements, it is possible to say that the North Carolina fish (assuming the Cur- rituck Sound sampling to be representative of that area) contributed at an absolute maximum about 20 percent of the 2-year-olds summering in northern waters in 1937. The percentage that North Carolina contributed to the northern population at this time was probably much less. In fact, a comparison of the widths of the second growth zones of the scales from fish of the same year-class from Chesapeake Bay and from northern waters in 1937 (see fig. 33) shows that it is possible that North Carolina did not contribute anything directly to the population of 2-year-olds summering in the north in 1937, and that this population came entirely from the Chesapeake Bay area or north of it. The latter, however, is undoubtedly an extreme view. It is thus apparent that in 1937 North Carolina contributed directly not more than a small fraction of the 2-year-old striped bass summering in northern waters, and that the 2-year-old bass in northern areas in that summer came mainly from the Chesa- peake Bay latitudes and perhaps from the Delaware Bay region. There is, however, a possibility that the fish born in North Carolina contribute indirectly to the popu- STUDIES ON THE STRIPED BASS OF THE ATLANTIC COAST 51 lation summering in northern waters — that is, that they move up into Chesapeake Bay in the spring as 2-year-olds (e. g., see under the last part of the section on migra- tions) and then migrate to northern waters a year or more later. This is added evidence that the dominant 1934 year-class, which first appeared as 2-year-olds in northern waters in 1936, came from the general area of Chesapeake and perhaps Delaware Bays, although evidence of the above type should be obtained for severa successive years before it can be considered conclusive proof of the fact that the contribution to northern waters in the spring and summer comes essentially from the latitudes of Chesapeake and Delaware Bays each year. Measurements of the growth zones of scales from striped bass born in 1936 in the Delaware Bay and Albemarle Sound regions are shown in figure 34. It will be noted that the widths of the second growth zones of the scales from the fish of Dela- ware Bay origin born in 1936 are slightly below those for the growth zones on the scales from the fish of Chesapeake Bay origin born in 1935. (Compare upper set of curves in fig. 34 with middle set of curves in fig. 33.) It is probable that this differ- ence is at least in part due to the fact that the second growth zones on the scales from the Delaware Bay fish were not yet quite complete (the fish were taken on November 8, 1937) because the annuli on scales do not appear until spring, although the growth from November to March is almost negligible. Whether or not there is a constant difference in the widths of the second growth zones of scales from fish of Delaware LENGTH FREQUENCIES OF GROWTH ZONES ON SCALES FROM FROM YEARLING AND T WO" YE AR- OLD? STRIPED BASS IN 1937-1938 OvlUBCft B. I9ST Figure 34.— The length-frequencies of the growth zones on scales from yearling and 2-year-old striped bass taken in Delaware Bay and Albemarle Sound in 1937 and 1938. The measurements making up these curves have been smoothed by threes throughout. and Chesapeake Bay origin remains to be seen from sampling over a period of years. It is probable that this method will not provide a good means of distinguishing between bass born in these two regions, as the environmental differences are appar- ently insufficient to cause any constant difference in growth rate during the second year. The widths of the second growth zones of scales from fish born in 1936 in Albe- marle Sound (see lower set of curves in fig. 34) are interesting because although they are quite great, they are not so distinctively different from the others as those from North Carolina collected in 1937 (see bottom set of curves, figs. 32 and 33). They indicate, in other words, that although a wide second growth zone is apparently a characteristic of North Carolina fish from the general region of Albemarle Sound, this characteristic varies from year to year sufficiently so that it can only be used as a means of distinguishing fish of North Carolina origin from fish of Chesapeake Bay origin when the scales from fair samplings of bass that are just becoming 2 years old in the spring, before any coastal migrations have been undertaken, are available from both areas during any one year. In conclusion it should be emphasized once more that the available evidence from general observation, scale analysis, and tagging experiments, gives every indi- cation that the dominant 1934 j^ear-class originated chiefly in the latitude of Chesa- peake and Delaware Bays; that those fish produced in North Carolina contribute directly only a relatively small fraction to the population summering in northern waters; and that the main body of the northern summer population of striped bass comes from the area bounded on the south by Virginia and on the north by New Jersey. Further proof that Chesapeake Bay in general contributes a large propor- tion of the stock summering in northern waters is seen m figure 35, where the catches 52 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE in New York and Maryland are compared in certain years from 1887 to 1935. (The material for this figure is taken from the U. S. Bureau of Fisheries canvass, and is not an annual comparison because the data are incomplete.) It wdl be noted that the trends of the catches in these two localities over this entire period show a remark- able correspondence — an agreement that could not reasonably be expected to occur unless the supply for both areas came mainly from the same source. In view of the evidence already presented, there can be little doubt that this source is the Chesa- peake Bay area. In figure 35 the Maryland catch has been plotted at one-tenth its actual value throughout, a reduction which brings the annual catch in that State TOTAL CATCH Of STRIPED BASS IN MARYLAND AND NEW YORK FROM 1887 TO 1935 BY ALL T V PES " OF GEAR • NEW YORK / \ o -o MARYLAND (total CATCH oivteeo / \ er 10 thro ghout) ^ i — i i .... i ... 1 — 1 . . . . Figure 35. — Total catch of striped bass in certain years by all types of gear in Maryland and New York from 1887 to 1935 (from U.S. Bureau of Fisheries canvass) . Maryland catch reduced to one-tenth throughout. down to the same proportions as that of New York. Assuming the fishing intensity to be about the same in New York and Maryland, it is therefore reasonable to expect that this means that about one-tenth of each year's production of young in Chesa- peake Bay reach New York. However, since immigrants from Chesapeake Bay are also taken in New Jersey and southern New England (unpublished material of V. D. Vladykov, p. 46), it is probable that somewhat more than one-tenth of the annual production of young leave Chesapeake Bay near the time that they become 2 years old, at the beginning of their third summer, and before they are old enough to be of any great value to the Chesapeake Bay fishery. FOOD OF THE STRIPED BASS The stomach contents of over 550 striped bass ranging in size from 6.5 to 115 cm. have been examined during the course of this investigation. These fish were all taken from April to November 1936 and 1937. Most of them were caught in Connecticut waters, although a few came from the Massachusetts coast and others from Long Island and New Jersey. Of the total number of fish examined, the majority were caught on rod and line; the others were taken by net. Over 75 per- cent of the stomachs studied came from bass that ranged in size from 30 to 50 cm. The rugose lining of the stomach of the striped bass probably indicates a rapid rate of digestion. It is apparently not a steady feeder, but may gorge itself over comparatively short periods of time and then stop feeding until its stomach is com- pletely empty again. Stomach-content analyses of individuals taken in the same seine hauls often showed the food to be in similar states of digestion, thus providing evidence that the members of a single school of striped bass feed simultaneously and then digest their food over essentially the same period of time. Often a high percentage of the bass in one haul would be filled with recently eaten fish such as men- haden {Brevoortia tyrannus) or silversides (Menidia menidia notata). Stomach- content analysis of the bass taken in another haul would reveal partially or well- digested food. At other times most of the fish taken together would be entirely empty. Approximately 52 percent of all the stomachs examined were completely empty. This high percentage may be explained, at least in part, by the fact that a large portion of the total number of stomachs examined were from rod-and-line caught STUDIES ON THE STRIPED BASS OF THE ATLANTIC COAST 53 fish, which are commonly empty because bass are more likely to be taken by anglers at the start of a feeding period when they usually have nothing in their stomachs, and also because bass taken on hook and line are often seen to regurgitate recently swallowed food. Studies of the food of juvenile and yearling striped bass ranging from 3-1 lcm. in standard length, seined on gravelly shoals of the Hudson River at Dennings Point, near Beacon, N. Y., have been made by Townes (1937) in connection with the bio- logical survey of the Lower Hudson Watershed carried out in 1936 by the State of New York Conservation Department. The majority of these fish ranged from 3.0-5.5 cm. in length. It was found that the fresh-water shrimp (Gammarusfasciahis) formed about 60 percent of the food, with chironomid larvae the next most important item. Small fish remains (not identified, save for one eel, Anguilla rostrata), leptocerid larvae, and planktonic Crustacea such as Latona, Cyclops, and Eurytemora, formed a small percentage of the food. Hildebrand and Schroeder (1928) examined the stomach contents of small striped bass from the salt and brackish waters of Chesapeake Bay, and found that ". . . the young had fed on Mysis, Gammarus, annelids, and insects." The stomach-content analysis of small bass has been confined in the present study to 3 juveniles ranging from 6.0-7.5 cm. in standard length taken in the Parker River, Mass., on August 4, 1937, and 30 juvenile and yearling individuals from 11-23 cm. long taken in the Delaware River, near Pennsville, N. J., on November 8, 1937. Those from the Parker River all had their stomachs filled with the shrimp, Crago septemspinosus." Those from the Delaware River were large enough to have become more voracious in their feeding habits, as is evidenced by the fact that 19 of the 30 examined contained the remains of fish of different species; the others were empty. A clupeoid species (probably menhaden, Brevoortia tyrannus) formed the main diet, while white perch, Morone americana, and shiners, Notropis hudsonius amarus, wore also commonly eaten. It is of some interest that one bass 16.5 cm. (6K inches) long contained a 7.5 cm. (2.95 inches) Morone americana, and examination of the stomach of an 18.5 cm. (7.28 inches) bass revealed the presence of a 10 cm. (3.94 inches) Notropis sp. The examination of stomach contents of larger striped bass (above 25 cm.) has confirmed the commonly held view that this species is voracious in its feeding habits, and fairly general in its choice of food. It has also made it clear that bass often feed off the bottom, and blind individuals that were frequently taken in the Thames River, Conn, (see under section on parasites and abnormalities of the striped bass), appeared to manage well by feeding only on bottom-dwelling forms such as those included in the list below. The most common form of food in Connecticut waters is the shiner, or silver- sides (Menidia menidia notata). This is a species which spawns in the spring (Hilde- brand, 1922), and the young of each year stay so close to shore and are of such small size that they do not become available to the striped bass as food until August. At this time they reach 2 cm. in length and often stray farther offshore. The growth rate of juvenile Menidia is shown in figure 30. The length-frequency curves making up this graph are from random samples of the population seined at biweekly intervals from July to September 1937 in the Niantic River, Conn. It is apparent from a glance at the modes of these curves that in 1937 a peak of 2.0 cm. was attained shortly after the middle of August. Stomach-content analysis of striped bass 30-50 cm. long in this area in 1936 and 1937 showed that adult Menidia and the common prawn (Palaemonetes vulgaris) formed the main food from April to August, but that in August and September the bass fed on juvenile Menidia to a large extent. Shortly after this change in diet in 1936 there was a decided increase in the growth rate of the 2-year- old striped bass (see p. 28), which, despite the drop in water temperature (see fig. 30), was greatest in October. The presence of what was apparently an unusually great number of juvenile menhaden (Brevoortia tyrannus) in 1936 may also have played a part in this increased growth rate, for from August on striped bass commonly fed '< Identified by Dr. Charles J. Fish, Director of the Marine Laboratory at Narragansett, Rhode Island State College, Kingston, R.I. 54 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE heavily on this species during this year. However, juvenile menhaden were not as abundant in 1937 in this area, yet the growth rate of striped bass in September and October continued much as it had throughout the summer in spite of the drop in temperature (see fig. 18). It therefore appears that the increased food supply of striped bass resulting from the availability of juvenile Menidia after the middle of August may be correlated with the maintenance or increase of the growth rate in the early fall when the water temperature falls rapidly, and when the normal expectation LENGTH FREQUENCIES BY BI-WEEKLY INTERVALS 25-1 JULY 17, 1937 GROWTH OF JUVENILE MENIDIA MENIDIA NOTATA, JULY-SEPTEMBER, 1937 NIANTIC RIVER, CONN GROWTH RATE OF JUVENILE MENIDIA MENIDIA NOTATA S"- 15 SO £5 50 STANDARO LENGTHS IN MMS Figure 36.— The growth of Menidia menidia notata, from July to September 1937, in the Niantic River, Conn. The length-fre- quencies have been smoothed by a moving average of threes throughout (see Table 23 for original data). would be that the growth rate would slow down. Other possible explanations of this apparently faster growth rate of striped bass in the late summer and early fall, such as faidty sampling and "compensatory growth," have been discussed in the section on the age and rate of growth of striped bass. The following comprise all the forms of food found in the stomachs of the 550 striped bass examined in 1936 and 1937: Common types: Shiners, or silversides (Menidia menidia notata) . Menhaden (Brevoortia tyrannus). Shrimp, or prawns (Palaemonetes vulgaris). Mummichogs, or kUlifish (Fundulus hetero- clitus and majalis). Uncommon types: Sand Launces (Ammodytes americanus) . Herring (Clupea harengus). Squid (Loligo pealei). Sandworms (Nereis virens). 15 Bloodworms (Glyccra dibranchiata). a Rare types: Flounders (Pseudopleuronect.es americanus). Eels (Anguilla rostrata). Tomcod (Microgadus tomcod) — one 20 cm. specimen in a 40-cm. striped bass. Clams (Mya arenaria) — of small size. Crabs (Callinectcs sapidus and Ovalipes ocellatus) — of small size. Snails (Litlorina, sp. ?). Mussels (Mytilus edulis). White perch (Morone americana). Mullet (Mugil cephalus). Shiners (Notropis hudsonius amarus). Blennies (Pholis gunellus). Amphipods. Isopods. >« These 2 marine annelids are generally used for bait, thus pieces of them are often found In bass that were caught on rod and line. However, whole individuals also have been observed in the stomachs of striped bass. STUDIES ON THE STRIPED BASS OF THE ATLANTIC COAST 55 It is apparent from a glance at this list that bass feed on a wide variety of animals, and it is likely that a study of stomach contents in other localities would yield as many more species as are common in the coastal waters inhabited by striped bass. In this connection, the examination of the stomach contents of 101 striped bass (yearling to 3-year-olds from the Albemarle Sound region and Manteo, N. C, in April 1938 yielded the following definitely identified forms, to say nothing of those that were too well digested to be identified: Teleosts. — Striped killifish (Fundulus majalis); sea trout, or spotted squeteague (Cynoscion nebvlosus); silver perch (Bair- diella chrysura) ; croaker (Micropogon undulatus) ; gizzard shad (Dorosoma cepedianum) ; spotted ling, or hake, or codling (Phycis regius); anchovy (Anchoviella mitchilli); eel (Anguilla rostrata) ; white perch (Morone americana) ; glut herring {Pomolobus aestivalis); and minnow, or shiner (Notropis, sp.?). Crustacea 16 . — Three species of shrimp (Peneus brasiliensis, Palaemonetes carolinus, Crago septemspinosus) ; young blue crab (Callinectes sapidus); and isopod (Aegathoa oculata). 17 is PARASITES AND ABNORMALITIES OF THE STRIPED BASS Parasites of the striped bass have been collected whenever they were observed from 1936 to 1938. Two species of nematodes have been found that are endoparasitic on the striped bass. The first, Goezia annulata (syn.: Lecanocephalus annulatus Molin), was found in a single specimen in the stomach mucosa, and has been reported and described by Linton (1901) and MacCallum (1921). The second, Dicheilonema rubrum (syn.: Filaria rubra Linton), has been observed in innumerable striped bass. It was found in the peritoneal cavity, usually in the posterior end in close association with the gonads, but it never appeared to do any serious harm to its host. This species has been reported for the striped bass by Railliet (1918), and is described by Linton (1901). Among the forms that are ectoparasitic on the striped bass are two species of copepods which have been found on various occasions. Caligus rapax, which occurs on many species of marine fish, and described by Wilson (1905 and 1932), is not un- common. Argulus alosae Gould was taken on three striped bass in the Niantic River, Conn., in August and September, 1936, thus constituting a new host record for this species; it was described by Wilson (1903). It is also of interest that in the collection of juvenile bass taken from the western end of Albemarle Sound on May 11, 1938, a high percentage of the fish were parasitized by glochidia. It is supposed that these glochidia attached themselves to the fish in the fresh water at or near the mouth of the Roanoke River, and it is not known whether or not they can complete their normal encystment and development after being carried into the brackish waters of Albemarle Sound. A review of the literature indicates that many other parasites have been reported for the striped bass. The monogenetic trematodes include Lepidotes collinsi (Mueller, 1936), Aristocleidus hastatus (Mueller, loc. cit.), Epibdella melleni (Nigrelli and Breder, 1934), Microcotyle acanthophallus, M. cueides, and M. macroura. _ Digenetic trematodes that have been reported on striped bass are Distoma rufoviride (syn.: D. tenue) (Linton, 1898), D. tornatvm, (Linton, 1901), and D. galactosomum. Two cestodes, Ehynchobothrivm bulbifer and R. speciosum, have been reported by Linton (1901 and 1924), the former as plerocercoids in the intestine (adults in Selachians), the latter in cysts in the viscera. Besides the nematodes already mentioned, an Ascaris sp. has also been reported by Linton (1901). Two acanthocephalans, Echinorhynchus gadi (syn.: E. acus) (Linton, 1901) and Pomphorhynchus laevis (syn.: E. proteus), have been taken from striped bass. Two other copepods besides those found by the author are the Lernaeopodid, Achtheres lacae (Wilson, 1915), and the Ergasilid, Ergasilus labracis (Wilson, 1911 and 1932). In regnrd to the general well-being of the striped bass, there is no evidence that any of the parasites that are associated with it are of any great importance. Dichei- lonema rubrum, which is so commonly found in the peritoneal cavity, shows a tendency >• Identified by Dr. Charles J. Fish, Direr-tor of the Marine Laboratory at Narragansett, Rhode Island State College, Kingston, R I ii The Isopod, A. oailata. is normally found parasitic on squid (Loligo pealei) and young mullet (Mvgil sp.), but since neither of these forms was seen in the stomachs of these bass, it is probable that A. oculata was taken by the bass while it was free-swiming during the breeding season. , , , . _ _ , .. '« The author wishes to express his gratitude to Dr. John S. Rankin, of the Department of Biology at Amherst College, for his assistance in the preparation of the material on the parasites of the striped bass, and for his identifications of the nematodes and copepods. 56 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE to become partially embedded in the mesenteries, but the infection never appears to be serious. Goezia annulata, although comparatively rare, is probably a much more serious pest. MacCallum (1921: 261) says: Its mode of living is calculated to interfere very materially with the function of the stomach, inasmuch as it burrows under the mucous membrane, in fact excavating in some cases quite a space where several worms cohabit. . . . There are often several of these nests in the stomach, each nest may be 30 mm. to 40 mm. across, and as they cause a good deal of swelling and irritation, they may and do in some cases so restrict the cavity of the host's stomach that its food cannot be taken in any quantity sufficient to keep it alive. Thus the worms are a very serious menace to the fish. This species is not common in striped bass, however, and according to reports is quite cosmopolitan in its choice of host, having been recorded from many other species of fish. Trematode infections are probably sufficiently rare in striped bass in their natural habitat to be of small importance. Nigrelli and Breder (1934) have shown that many of the Serranid fishes have developed a. resistance to Epibdella melleni, while Jahn and Kuhn (1932) noted that "... the possibility of the development of immunity seems to be more strongly suggested in this family" (Serranidae) . Copepod parasites are also apparently of small consequence to the striped bass. It is worth mention that a surprising number of striped bass were encountered in the Thames and Niantic Rivers, Conn., that had cataracts of the eye. These were found commonly only in the Thames River, where they sometimes reached above 10 percent of the catch by seine. This opacity of the lens was encountered in all degrees from a slightly cloudy to a dead-white condition. It was almost universally bilateral, was rare in 2-year-old bass, and more common in the larger sizes. It was equally common in all months from April to October. A number of dissections under low- power magnification failed to reveal any parasites, such as larval digenetic trematodes, which might reasonably be expected to cause such blindness. Hess (1937) has recently shown that bilateral cataracts are common in trout in New York State, both in hatch- ery and wild stock, and he has proved with rainbow trout (Salmo irideus) ". . . that cataract in these fish is due to an unbalanced diet." He has been able to demonstrate that contagious infection, light, and hereditary factors, are not in any way connected with the production of such cataracts, and that the feeding of trout exclusively on pig spleen caused a high incidence of cataract; while trout fed with beef liver and heart never showed any trace of cataract. It seems likely, therefore, that a dietary deficiency may perhaps account for the high percentage of blind striped bass in the Thames River. It is interesting in this connection that the extraction of carotene by acetone from the liver and fatty tissue of blind and normal bass has tended to show less carotene per gram of tissue in the blind than in the normal individuals, and it is thus possible that a lack of vitamin A is associated with the dietary deficiency causing cataracts. It is also of interest that Schultz (1931) has recorded a case of what gave every appearance of being completely functional hermaphroditism in the striped bass. This fish was taken in Oregon in May, and the eggs in one half of the gonads measured about 1 mm. in diameter, close to the size at the time of spawning (see p. 19), while the male half of the gonads was apparently developing normally. DISCUSSION It has been pointed out that there has been a striking decline in the numbers of striped bass along the Atlantic coast over long-term periods. (See under section on fluctuations in abundance of the striped bass, p. 8, and figs. 3 and 4.) The records show that this decline has been fairly steady from at least as far back as the middle of the nineteenth century, and perhaps before. They also indicate that it has been interrupted only by the occasional appearance of dominant year-classes — groups of striped bass that were produced in such huge amounts in certain years that they caused a marked increase in the numbers caught for short periods (see p. 8, et seq.). It is apparent from the available catch records (see fig. 4), however, that these dominant year-classes did not bolster the stock for more than a few years, and that their effects invariably have been short lived. In other words, the surplus created by them was soon removed, no permanent increase in abundance — and a consequent permanent increase in catch — resulted, and the decline in numbers of striped bass, although tem- porarily interrupted, soon resumed its normal trend. STUDIES ON THE STRIPED BASS OF THE ATLANTIC COAST 57 Of especial importance in this respect is the dominant year-class of 1934, probably the largest production of striped bass in a single year in the past half century, whose members appeared along the Atlantic coast as 2-year-olds in 1936 and were at once subjected to the highly intensive fishery that confronts this migratory species over the greater part of its range. Information gathered in the course of this investigation makes it possible to demonstrate that this dominant year-class was directly responsible for a greatly increased catch, and also to make a rough estimate of the approximate rate at which this surplus was removed. Such an estimate is based on the percentage of tag returns from 2- and 3-year-old striped bass of the dominant 1934 year-class. (See pp. 36-41 and tables 17-20.) It includes all the factors which show that the percentage of tag returns on this age-group was far lower than the actual percentage removed by the fishery from 1936 to 1938. (See pp. 15 and 36.) Using this method, the most reasonable approximations show that about 40 percent of the members of this year-class were removed as 2-year-olds, and that at least 25-30 percent of the remain- ing 3-year-olds were taken by the fishery in 1937 and 1938. If these estimates are correct it means that over 50 percent of the 2-year-olds entering the fishery in the spring of 1936 had been removed by the spring of 1938, neglecting the effect of natural mortality, which is taken up below (see p. 59, et seq.), and which is an important factor in the rate of removal of the members of any population. Even though these estimates are only rough approximations, it is plainly evident that the enormous sur- plus created by the production of the dominant 1934 year-class, resulting in the largest catch of many years in 1936 (see figs. 4 and 6), is rapidly being removed, and that the members of this age-group will soon have been depleted to such an extent that they will no longer bolster the annual catch. Granting, then, that there has been a sharp decline in the numbers of striped bass along the Atlantic coast despite the occasional appearance of dominant year-classes that bolstered the stock temporarily, it is of interest to know what lias caused this decline. Two factors appear to have been responsible — first, the destruction of spawn- ing areas by pollution and dams, and second, overfishing. Let us now consider these two factors in some detail. There can be little doubt that striped bass formerly entered and spawned in nearly every river that was suitable along the better part of the Atlantic coast. As civiliza- tion advanced, dams were built, many of the streams were polluted, and the number of spawning areas that were available became less and less. It has been pointed out under the section on spawning habits and early life history, and elsewhere in this paper, that the majority of the spawning areas for striped bass are now confined to the coastal rivers from New Jersey south. There remain, however, a few isolated localities to the north that are still suitable — probably but a fraction of the areas that were once available. Yet it is clear from the production of the dominant 1934 year-class that there are still a sufficient number of good spawning areas left along the whole Atlantic coast to produce a large supply under the proper conditions. It should not be necessarj- to emphasize the fact that these remaining localities should be carefully protected against anything that might damage them, and other areas should be restored if it is possible. Further investigations on the striped bass should continue the study of spawning areas along the Atlantic coast and determine the necessary requirements for the nor- mal production, fertilization, and development of the eggs and larvae. In the case of some of the isolated spawning areas in northern waters, where the stock appears to have been maintained by a more or less self-supporting and partially resident popu- lation, there is some evidence that intensive winter and spring fisheries on the supply in the spawning localities have practically exhausted the stock. Under normal con- ditions the populations north of Cape Cod are probably not increased to any great extent by migrants from outside — especially from the south. This only occurs under exceptional cases, although it may occur more commonly in the future now that the ("ape Cod canal provides an easy means of access to the north (see p. 41). Thus an intensive fishery in the winter and early spring when the members of such an isolated self-supporting stock are dormant and inactive, and hence more easily available for capture, may come close to entirely depleting a population of this sort. Turning to the other factor, overfishing, which in conjunction with the destruc- tion of spawning areas by dams and pollution has been responsible for the decline in 58 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE abundance of striped bass, tbe problem is to see how overfishing affects the stock. Theoretically this factor may act in two ways — first, by the removal of too high a proportion of undersized and immature fish so that there are too few spawning indi- viduals, and second, by failing to take the members of the available population at the most efficient size. In regard to the removal of too great a number of striped bass before they have been given a single chance to spawn, evidence has already been presented to show that the fishery for the smaller size-categories of bass, 2- and 3-year-olds, is higldy intensive, and that a large percentage of each successive year-class is caught before its members attain maturity. Yet there is no reason to believe that an additional supply of spawning individuals woidd result in an increased production, with the one possible exception noted below. Thus it has been emphasized in the section on fluctuations in abundance of the striped bass that the dominant 1934 year-class was apparently produced by as small a parental stock as there has ever been. This means that in southern waters the production of dominant year-classes is not completely dependent — at least down to a certain limit — on the quantity of spawning individuals. In other words, there appears to be no need for concern over the size of the spawning population in the south as long as it is at least as large as it was in 1934. If such a hypothesis be granted, there can be little good in raising the legal-length limit solely for the purpose of increasing the number of spawning fish — especially since we know that under the conditions of the present fishery the number of striped bass along the Atlantic coast is sufficient to produce a year-class of enormous proportions, such as the one that originated in 1934. There is, however, one way in which an increased number of spawning adults may possibly bolster the supply in northern waters, for this supply has apparently declined in some cases to such an extent that the population has been practically wiped out. It has been shown before that in certain years striped bass from the south migrate north of Cape Cod. Since it has been well established that some of these migratory fish remain in northern waters through the winter, it is a reasonable ex- pectation, if they were mature fish, that they would repopulate some of those areas which formerly supported small populations in northern waters and are still suitable for spawning purposes. Thus the striped bass has been virtually an unknown quantity north of Cape Cod for the past 30 years or more; that is, until the members of the dominant 1934 year-class came north of Cape Cod in huge quantities in 1936 and 1937 and provided a renewed sporting and commercial fishery of considerable size in those waters. It is certainly not unreasonable to predict that if a sufficient number of mature fish repopulate the spawning areas that still remain north of Cape Cod, the stock in northern waters can be replenished and the supply increased and maintained if the fish are given the proper protection. It may therefore be said that measures designed to increase the supply of striped bass along the Atlantic coast by providing a greater number of spawning fish might quite possibly prove ineffective in the more southern waters of the Middle Atlantic Bight, for it is known that there are now a sufficient number of mature individuals to produce huge quantities of fish if the environmental factors are right; witness the dominant 1934 year-class. On the other hand, such measures would probably renew, at least partially, the supply north of Cape Cod where the stocks have been practically exhausted in many instances. The other aspect of overfishing to be considered is whether or not the present fishery along the Atlantic coast takes the available members of the population at the most efficient size, or, whether or not the fishery makes the best possible use of the supply each year. Thompson and Bell (1934), Graham (1935), Thompson (1937), and others, have all discussed the theory of the effect of fishing on various stocks of fish, and have studied the problem of the most efficient utilization of the stock in different species. These papers have laid the foundation for future studies along this line, and it is possible to apply many of the principles set forth in them to the striped bass fishery of the Atlantic coast. Those who are critically interested in this whole subject should refer to the work of these authors. The first problem in connection with the striped bass is to get some measure of the yield from the stock under the existing conditions of the fishery at the present time. Having attained this, it is possible to compare it with the yield from the stock under STUDIES ON THE STRIPED BASS OF THE ATLANTIC COAST 59 different conditions of the fishery and thus determine which is the most advantageous, not only from the point of view of profit to the fisherman, but also in the light of what is known about the life history of this species. In other words, it is desirable to dis- cover at what age (or length) it is most advantageous to start the fishery for striped bass; i.e., whether the fishery gets the most profit out of taking the fish for the first time when they are 2-year-olds (averaging roughly three-quarters of a pound and 12 inches in length) as it does at present, or whether it would benefit by allowing the fish one or two more growing seasons before catching them. In order to find the answers to these questions it is essential that the fishing mortality at different ages — the percentage of fish of each age taken by the fishery — and the natural mortality, be known. This can only be done accurately by careful studies and the collection of detailed statistics on the annual catches of striped bass over long-term periods, although the present work has given some information along these lines. Considering the dominant 1934 year-class, it has been assumed from the percentage of tag returns (see p. 57) that approximately 40 percent of its members were taken by the fishery as 2-year-olds in 1936 and 1937, and that about 25 percent of the 3-year-olds of 1937 and 1938 were also taken by the fishery. It is known from various catch records from Virginia to Rhode Island that only about one- quarter as many 3-year-old striped bass were caught in 1937 as the 2-year-olds that were taken in 1936. This is demonstrated in figure 4, where the catches of a pound- net fisherman at Fort Pond Bay, Long Island, N. Y., were approximately four times as great by number in 1936 as they were in 1937, and where the catch was over 90 percent 2-year-olds in 1936 and 3-year-olds in 1937. Given this information it is possible to estimate the natural mortality in 1936 by the following equation: NM=S 1 -(FM l +S 2 ), wherein NM is the natural mortality in 1936, Si the stock available in 1936, FM t the fishing mortality in 1936, and S 2 the stock available in 1937. Si can be given any arbitrary value, for example, 1,000. If FM X is assumed to be 40 percent of Si (see above), FMi is 400. S 2 is equal to approximately Ay.FM 2 , where FM 2 is the fishing mortality in 1937, for tagging experiments indicate that roughly 25 percent of the 3-year-olds were taken in 1937. FM 2 is known to be % FM U as only one-quarter as many 3-year-olds were taken in 1937 as there were 2-year-olds taken in 1936. Under these conditions FM 2 therefore becomes 100, and in the equation above, where Si was assumed to be 1,000, S 2 becomes 400. Substituting these values in the equation, the natural mortality in 1936 attains a value of 200. Thus of the original 1,000 fish in 1936, 400 were caught as 2-year-olds, and of the remaining 600 fish, 200 were lost through natural mortality. It is therefore apparent that if the estimates on which the figures making up this equation are based are correct, natural mortality accounted for about one-third of the 2-year-olds in 1936 which were not taken by the fishery. It should be pointed out, however, that slight variations in the percentages assigned to FMi and FM 2 , which are only rough approximations, can materially change the value obtained for NM. Taking the figures in the equation above, since they seem to be the best available, it is possible to get some estimate of the yield from the stock under the existing con- ditions of the fishery. Table 1 is a theoretical treatment of 1,000 striped bass of the 1934 year-class to show the rate of removal by the fishery and natural mortality, the numbers and poundage caught, and the market value, when the fish of this age group were caught over a 5-year period from 1936-40 (as 2-, 3-, 4-, 5-, and 6-year-olds). This treatment, in other words, considers the value when the fishery starts catching striped bass for the first time as 2-year-olds, which is exactly what occurred in 1936 along the Atlantic coast. The natural mortality is figured at one-third of the popu- lation, excluding those taken by the fishery. The fishing mortality was estimated to be 40 percent in 1936, 25 percent in 1937, 15 percent in 1938 (when the members of the 1934 year-class were 4-year-olds), 10 percent in 1939 (5-year-olds), and 5 per- cent in 1940 (6-year-olds) — a declining fishing mortality that undoubtedly represents as sharp a decrease in the percentage of fish of any year-class caught each year as could possibly exist, and probably over-estimates the decline in the percentage taken by the fishery as the members of a year-class become older. It will also be noted in table 1 that the price per pound varies with the different size categories under con- 60 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE sideration. Thus the 2-year-olds averaging three-quarters of a pound each are listed as bringing 6.5 cents a pound, the 3-year-olds averaging 2 pounds each as 9.5 cents a pound, and the 4-, 5-, and 6-year-olds as bringing 10 cents a pound throughout. These prices were determined from information collected by the Bureau of Fisheries from an important dealer on the Atlantic coast. The average price per pound for the different size categories was determined by dividing the total dollar volume for each month by the total number of pounds of striped bass purchased each month from March through November 1937. The prices for each of these months were then averaged, giving the average price for the different size categories for the entire period. Since this dealer handled a total of approximately 200,000 pounds during this period, the prices for the different size categories should be accurate estimates. Table 1. — Theoretical treatment of 1,000 striped bass of the 1934 year-class to show the rate of removal by the fishery and natural mortality, the numbers and- poundage caught, and the market value, when the fish were caught over a 5-year period from 1936-40. Note that in this treatment fish ivere caught for the first time when they were 2-year-olds Age Average length Average weight Total weight Average price per lb. Market value Assuming 1,000 bass were available in 1936, of which 400 would be caught in 1936 (fishing mortality, 40 percent); 200 would die in 1936 (natural mortality, 33 percent of those not caught), leaving Years 2 31 cm. (12.2 inches). Pounds 0.75 Pounds 300.0 Cents 6.5 $19. 5 400 bass available in 1937, of which 100 would be caught in 1937 (fishing mortality, 25 percent); 100 would die in 1937 (natural mortality, 33 percent of those not caught), leaving 3 41 cm. (16.1 inches). 2.0 200.0 9.5 19.00 200 bass available in 1938, of which 30 would be caught in 193S (fishing mortality, 15 percent); 57 would die in 1938 (natural mortality, 33 percent of those not caught), leaving 4 50 cm. (19.7 inches.) 3.5 105.0 10.0 10.50 113 bass available in 1939, of which 11 would be caught iu 1939 (fishing mortality, 10 percent); 34 would die in 1939 (natural mortality, 33 percent of those not caught), leaving 5 58 cm. (22.8 inches). 5.5 60.5 10.0 6 05 68 bass available In 1940, of which 3 would be caught iu 1940 (fishing mortality, 5 percent). 6 66 cm. (26.0 inches). 8.0 24.0 10. 2.40 Total 689.5 57.45 In table 1 it will be seen that the total market value derived from 1,000 bass of the 1934 year-class over the 5-year period 1936-40 was $57.45, the total number of individuals caught was 544, and the total weight taken was 689.5 pounds. These figures represent the yield to the fishery when striped bass are caught for the first time as 2-year-olds (12 inches in length). Table 2 gives similar information for the same number of bass of the 1934 year- class when the fishery did not catch them as 2-year-olds in 1936 but took them for the first time as 3-year-olds in 1937, and caught them over the 4-year period 1937-40. It will be noted that the total market value under these conditions was $64.48, the total number of individuals caught was 242, and the total weight taken was 661.5 pounds. Thus, less than half as many individuals were taken when the fishery first caught bass as 3-year-olds, yet the gross profit was substantially more. It is, there- fore, plainly evident that if the figures upon which these calculations are based are reasonably accurate, the fishery is not utilizing the available supply of striped bass in the most efficient manner when it first takes them as 2-year-olds. Since it has been shown that it is apparently more efficient for the striped bass fishery of the Atlantic coast to start taking the fish as 3-year-olds rather than as 2-year- olds, it is of interest to consider what the yield would be if the fishery waited still another year and did not begin to remove the members of the bass population until they became 4-ycar-olds. Treating the same 1,000 fish of the 1934 year-class in the same manner as shown in tables 1 and 2, with the sole difference that the fishery only operates over a 3-year period from 1938-40, the total market value drops to $43.60, and there appears to be an inefficient utilization of the available stock from every point of view. This striking drop in the gross profit under these conditions is STUDIES ON THE STRIPED BASS OF THE ATLANTIC COAST 61 due to the high value estimated for natural mortality each year, for the amount added in total growth by allowing the fish to live until they are 4 years old does not compensate for the numbers lost through natural mortality under these conditions. Table 2. — Theoretical treatment of 1,000 striped bass of the 1984 year-class to show the rate of removal by the fishery and natural mortality, the numbers and poundage caught, and the market value, when the fish were caught over a 4-year period from 1987-40. Note that in this treatment the fish were caught for the first time when they were 3-year-olds Age Average length Average weight Total weight Average price per pound Market value Assuming 1,000 bass were available in 1936, of which 333 would die in 1936 (natural mortality, 33 percent), leaving Years 2 Pounds Pounds Cents 667 bass available in 1937, ot which 167 would be caught in 1937 (fishing mortality, 25 percent); 167 would die in 1937 (natural mortality, 33 percent of those not caught), leaving 3 41 cm. (16.1 inches). 2.0 334.0 9.5 $31.73 333 bas available in 1938, of which 50 would be caught in 1938 (fishing mortality, 15 percent); 94 would die in 1938 (natural mortality, 33 percent of those not caught), leaving 4 60 cm. (19.7 inches). 3.5 175.0 10.0 17.50 189 bass available in 1939, of which 19 would be caught in 1939 (fishing mortality, 10 percent); 67 would die in 1939 (natural mortality, 33 percent of those not caught), leaving 5 68 cm. (22.8 inches). 5.5 104.5 10.0 10.45 113 bass available in 1940, ot which 6 would be caught in 1940 (fishing mortality, 6 percent). 6 66 cm. (26.0 inches). 8.0 48.0 10.0 4.80 Total number of striped bass caught during 1937-40 , 242. 1 otal 661.5 64.48 In tables 1 and 2 it was shown that the total market value of striped bass taken from the available stock of 1,000 fish of tbe 1934 year-class from 1936-40 (bass caught for the first time as 2-year-olds) was $57.45, as compared with $64.4S when this same stock was utilized by taking its members for the first time when they were 3-year-olds over the period from 1937-40. It should be pointed out that the gain from allowing the fish to become 3 years old before being caught has been figured in these examples as the least that can result. In the first place, the fishing mortality on the members of the 1934 year-class was estimated from tagging experiments as 40 percent in 1936 and 25 percent in 1937. It has been arbitrarily placed at 15 percent in 1938, 10 per- cent in 1939, and 5 percent in 1940, because they are considered the lowest values possible. Whether or not this annual decline in the percentage taken is as steep as indicated above and in tables 1 and 2 is extremely questionable. It is obvious that if this decline is less sharp, the gain from allowing the fish to become 3 years old before being caught is relatively greater. Further than this, the natural mortality of the bass of the 1934 year-class is estimated to be 33 percent of the population (neglecting fishing mortality) in 1936, and it has been arbitrardy placed at 33 percent for the years from 1937 to 1940. Actually, it is extremely unlikely that it remains as high as 33 percent over this period, for it is reasonable to assume that as bass become older than 2 years of age they are less likely to be killed through natural causes. It is possible that when bass become much older the death rate increases, but in the examples in tables 1 and 2 that stage is probably not reached. Thus it is likely that the annual natural mortality of 33 percent from 1937 to 1940 is far too high. If this be so, the gain from allowing the fish to become 3 years old before being caught is again relatively greater than is shown by the total market value in the examples given above. It is evident therefore that the gain from catching striped bass for the first time as 3-year-olds is far more than is shown in tables 1 and 2. Nor should it bp necessary to point out that the figures used in the examples in tables 1 and 2 represent only gross values, and that the net values would be far greater. It is also of importance that if the fishery first starts to operate on the striped bass population when its members are 3 years old, a greater proportion of the stock is given a chance to spawn. It has already been shown (see p. 22) that female striped bass first mature at 4 years of age. If the stocks available at this age are compared in tables 1 and 2, it will" be seen that of the 1,000 original fish of the 1934 year-class only 200 were 62 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE left by 1938 when the fishery started taking the fish for the first time as 2 year-olds, while 333 were left by 1938 when the fishery started to operate on 3-year-olds. _ In other words, on the basis of these calculations about 1% times as many female striped bass would be given a chance to spawn if the fishery were to allow the 2-year-olds to remain in the water and first started to catch them as 3-year-olds. It has previously been pointed out that although a conservation measure designed to increase the stock by adding to the number of spawners in the south has no evidence to prove that it is not a fallacious policy, an increase in the number of mature fish in northern waters should repopulate this area to a certain extent and revive the fishery in this region There are, of course, many spawning areas in northern waters that have been ruined by pollution and dams so that they could not be repopulated, but it is widely believed that depletion in northern waters is in part due to insufficient numbers of spawners. Thus Bigelow and Welsh (1925) say: Since striped bass have dwindled as nearly to the vanishing point in the St. John (which still sees a bountiful yearly run of salmon) as in the estuaries of rivers that have been dammed and fouled by manufacturing wastes, the chief blame for its present scarcity can not be laid to obstruction of the rivers; and as this is a very vulnerable fish, easily caught, always close inshore, always in shallow water and with no offshore reservoir to draw on when the local stock of any particular locality is depleted by such wholesale methods of destruction as the early settlers employed — overfishing must be held responsible. Probably one of the reasons why the depletion in northern waters has been so great is that bass which remain north in the winter become dormant and inactive (see p. 42), and hence far more easily available for capture, so that it is not impossible to wipe out an entire population. Under these circumstances there is good reason to believe that an added number of mature fish in northern waters would assist mate- rially in renewing the supply in these areas, and that this supply could be maintained by affording the population adequate protection. _ It should be mentioned at this point that the abundance of striped bass in Cah- fornia, where the present fishery arose as a result of two small original plantings (see p! 5), has been successfully maintained by protecting this species up to the time they become 4 years old, at which time they are about 20 inches in length. Thus Craig (1930) and Clark (1932 and 1933) have studied the fluctuations in abundance of the striped bass in California, and both of these authors came to the conclusion that "the striped bass population coidd support a commercial fishery as well as a sport fishery" — a conclusion to which, however, the California State legislature apparently paid scant attention, since commercial netting was prohibited by law after August 14, 1931. , , . . , In consideration of all the foregoing evidence, even though it is based on assump- tions that need further corroboration by continued investigation of this species, it seems highly advisable to try the experiment of allowing striped bass to become 3 years old before they are caught in large quantities along the Atlantic coast. Both sportsmen and commercial fishermen should benefit by this apparently more efficient utilization of the available stock, the former by having an increased number of large bass to fish for, and the latter by making a definitely higher profit than they do under the present conditions. An addition to the spawning stock in northern waters, where the supply has been depleted to such an extent that an added number of mature individuals is badly needed, shoidd also result from protecting this species up to the time it becomes 3 years old. RECOMMENDATIONS The preceding section has dealt with a theoretical discussion of the striped bass population of the Atlantic coast. The causes for its decline in numbers over long- term periods, its fluctuations, and the effects of different fishing intensities and natural mortality on the stock under the existing conditions have been considered. Also, an attempt has been made, on the basis of the limited information at hand, to determine how the available supply of striped bass can be utilized most efficiently from every point of view. The data tend to show that the way in which the fishery for striped bass along the Atlantic coast can make the best possible use of the available supply is to start taking the fish as 3-year-olds, when they average 41 cm. (16 inches) to the fork of the tail and weigh roughly from 1% to 2 pounds each. There is apparently STUDIES ON THE STRIPED BASS OF THE ATLANTIC COAST 63 more profit when the fishery first starts to take the bass as 3-year-olds than there is when the fishery starts to take the bass as 2-year-olds, because the greatest increment in growth in the entire life of the striped bass takes place during the third year of life — when the fish are 2 years old. This growth in the third year is sufficient to more than compensate for the losses due to natural mortality, and its advantages are missed when the fish are caught for the first time as 2-year-olds. It is therefore recommended, on the basis of existing knowledge and as a practical experiment in conservation, that striped bass on the Atlantic coast less than 16 inches in length be protected. The problem is, then, how striped bass should be protected up to the time they become 3 years old. Unfortunately the commercial fishery is not one which exists for the purpose of catching this species alone; rather, striped bass are taken in associa- tion with many other forms by different types of gear along the whole coast. It is impossible to make any limitation on the size of mesh to be used, since this would affect the capture of other species that do not need to be protected up to as large a size as do striped bass. Further than this, the striped bass is highly migratory and should be protected along the entire length of its range. It is only feasible, on this account, to suggest a universal length limit (or at least a commercial sale limit) for the entire Atlantic coast, and let the individual States determine by appropriate investigation whether additional restrictions on the gear employed in the striped bass fishery, and on the seasons when the fishery shall operate, would be profitable. It is no great hard- ship for commercial fisheries to return undersized bass to the water, and it is to their ultimate advantage to do so — not only from the point of view of the increased return it should bring them, but also in order to eliminate any legitimate objection by anglers to their fishing methods. That the mortality of these undersized bass from being caught in a net and handled before being released would be small under normal condi- tions is abundantly illustrated by the fact that some of the most successful tagging experiments that have been carried on during this investigation have been made on fish that were caught in seines and pound-nets. It is apparent that there is nothing to be lost and much to be gamed by allowing the striped bass of the Atlantic coast one more growing season than they have under existing conditions in the fishery — that is, by allowing them to become 3-year-olds before they are taken in large quantities. However, the gains from such an experi- mental measure will depend directly upon its universal acceptance along the entire Atlantic coast, and on the complete cooperation of those engaged in the fishery. The adoption of measures designed to protect striped bass of less than 16 inches in length should result in greater profit to the commercial fishermen, an increased supply of larger fish for the sportsmen, and a larger number that reach maturity — of which a certain number should spawn in northern waters and possibly replenish stocks which have been badly depleted. It is also apparent that there is need for much more study on the striped bass of the Atlantic coast. This is especially true since the specific recommendations as to the size limit of the striped bass made in this paper are suggested on an experimental basis. It is therefore essential that more detailed and more accurate catch records be made available, and further biological studies be undertaken in order to trace the results of the recommendation if adopted, to make possible a suitable revision of the size limit if the results indicate that modification would be desirable, and to amplify the results of the present investigation. SUMMARY AND CONCLUSIONS (1) The foregoing report is concerned with the results of an investigation of the striped bass (Roccus samtilis) of the Atlantic coast, from April 1, 1936, to Juno 30, 1938. (2) The general morphology and systematic characters of the species are described in detail on the basis of the literature and material afforded by fin-ray, scale, and vertebral counts, and measurements on more than 350 individuals. (3) The striped bass is strictly coastal in its distribution from the Gulf of St. Lawrence to the Gulf of Mexico. Those most commonly taken at present range from less than 1 pound to 10 pounds in weight; but larger individuals are by no means rare. The largest striped bass of which there is authentic record weighed 125 pounds. 277689—41 5 64 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE (4) Studies of the fluctuations in abundance of the species over long-term periods show that there has been a sharp decline in numbers. Dominant year-classes have at times temporarily raised the level of abundance, but the intensity of the fishery is such that their effects have been short-lived. The dominant year-class of 1934 was the largest to be produced in the past half century, although the parental stock at this time was probably as small as it ever has been. Evidence is presented to show that there is a good correlation between the production of dominant year-classes of striped bass and below- the-mean temperatures during the period before and immediately after the main spawning season. (5) The striped bass is anadromous, spawning from April through June, the exact time depending on the latitude and temperature. The majority of spawning takes place from New Jersey south, although there are a few isolated spawning areas in northern waters. The development of the eggs and larvae is pictured, and the size of the juveniles at different times of the year is discussed. (6) Sex determinations of striped bass in Long Island and New England waters show that the number of males in this northern range of the species seldom reaches much over 10 percent of the population; the percentage of males apparently de- creases in the age-categories above the 2-year-olds. In waters farther south the sex ratios are not so disproportionate. Studies of the age at maturity show that ap- proximately 25 percent of the female striped bass first spawn just as they are becom- ing 4 years of age, that about 75 percent are mature as they reach 5 years of age, and that 95 percent have attained maturity by the time they become 6 years old. A large percentage of the male striped bass are mature at the time they become 2 years old, and probably close to 100 percent are mature by the time they become 3 years old. This difference in the age at maturity of male and female striped bass may well account for the small percentage of males in northern waters, for the time of the spawning season in the south coincides with the time of the spring coastal migration to the north, which is made up mainly of immature females. (See under migrations, p. 44.) (7) The age and rate of growth have been studied by scale analysis and by the average sizes of different age groups. The scale method and its applicability to the striped bass is discussed in full. Striped bass are roughly 12 cm. long when they become 1 year old, 24 cm. when they become 2 years old, 38 cm. when they become 3 years old, and 45 cm. when they become 4 years old. Thereafter the annual in- crement in length is about 7-8 cm. up to the tenth year. The growth rate of striped bass in the summer months in 1937 was much greater just north of Cape Cod than it was slightly south of Cape Cod. The growth rate of 2-year-old striped bass in Connecticut waters was approximately the same from June through October 1937, and increased in September and October 1936, despite the drop in water tempera- ture. This maintenance of or increase in the growth rate in the fall was probably due to increased food supply at this time. The growth and availability of juvenile silversides (Menidia menidia notata) are shown to be of direct consequence in this relation. The members of the 1934 dominant year-class averaged 2 cm. smaller than the members of the 1933 and 1935 year-classes, neither of which were large, at similar ages. This difference in size developed before these fish became 2 years old. (8) A total of 3,937 striped bass have been marked by either external disc tags or internal belly tags. Returns from these tagged fish, and the examination of commercial catch records, show that there is a mass migration to the north in the spring and to the south in the fall, and that the population in northern waters is stationary in the sum- mer. These migrations have their greatest intensity along the southern New England and Long Island shores. They take place chiefly between Massachusetts and Virginia, although bass north and south of these areas play some part in the migrations. The Middle Atlantic Bight is undoubtedly the center of abundance for the striped bass over its entire range, and tagging experiments indicate that there is little encroachment by this stock on the populations to the north and south. Temperature undoubtedly plays some part in the migrations, for in Connecticut waters they have been observed to occur on each occasion when the water reached 7°-8° C. The migrations of the striped bass, however, are not universal, for this species is caught through the summer in southern waters and in northern waters in the winter. Those fish that stay north STUDIES ON THE STRIPED BASS OF THE ATLANTIC COAST 65 in the winter often become dormant and inactive. The evidence is strong that the maximum tolerance limit for the species is 25°-26° C, which is about as high a temper- ature as coastal waters ever reach in the North and Middle Atlantic. Coastal migra- tions are not undertaken by bass less than 2 years old. Tagging experiments conducted in North Carolina in the springs of 1937 and 1938 tend to show that bass from this region contribute directly only a small percentage to the population summering in northern waters. (9) The available evidence from general observation and scale analysis points to the conclusion that the dominant 1934 year-class originated chiefly in the latitude of Chesapeake and Delaware Bays, and confirms the results of the tagging experiments in North Carolina in the springs of 1937 and 1938 mentioned above. (10) Stomach-content analyses on over 550 striped bass from northern waters, and on over 100 individuals from the south, show that bass are general in their choice of food — a large variety of fishes and Crustacea forming the most common diet. (11) Various nematodes and copepods have been found parasitic on the striped bass, and a number of trematodes, cestodes, and acanthocephalans have also been listed by other authors. Glochidia were found on small juveniles from the western end of Albemarle Sound. Several of the parasites listed constitute new host records. None of these parasites are of any great consequence to the general well-being of the striped bass population. A high percentage of bass in the Thames River, Conn., were found to have bilateral cataract. It is suggested that this is the result of a dietary deficiency. (12) The decline in abundance of the striped bass of the Atlantic coast over long- term periods and its causes are discussed, and it is pointed out that the present prac- tice of taking such a large proportion of the 2-year-olds annually is apparently not an efficient utilization of the supply, and that both the fishery and the stock should benefit by protecting this species until it is 3 years old, at which time it is approxi- mately 41 cm. (16 inches) long to the fork of the tail and weighs \% to 2 pounds. The adoption of such experimental measures designed to protect striped bass up to the time they become 3 years old should result in a greater profit for the commercial fishermen, an increased supply of larger fish for the sportsmen, and an added number of individuals that reach maturity, some of winch may possibly spawn in northern waters and thus replenish the stocks in theso areas where in many instances the populations have been exhausted. The need for further studies on the striped bass is emphasized in order that the results of the recommendation, if adopted, may be traced, so that suitable revision of the size limit may be made if the results indicate that modifications would be desirable, and in order to amplify the results of the present investigation. Table 3. — Record of striped bass taken by members of Cultyhunk Club, Cuttyhunk, Mass., 1865-1907 Year Number offish Average weight Largest fish Year Number offish Average weight Largest fish 1865 1,174 659 906 942 887 615 804 681 592 600 724 835 321 648 499 403 184 200 154 124 46 3 28 24.60 65 57 48 47 42 39 37 65 50.25 51 51.50 51 49 50.25 44 64 31.75 43 29.50 27.25 1887 1888 235 29 4 154 43 18 39 80 21 25 59 45 21 14 13 2 4 5 7 1 5 12 16.50 22.50 14 11.75 16.50 16.25 9 17.25 14.25 11.25 9 13 18 14 17 10.75 15 16 9.25 19 42 66 41 41.50 24.25 38.50 35. 50 35.25 36.75 27 33.50 23.76 35 54 29 26 15.75 35 40 9.25 23.50 6.75 6.25 6.75 6.50 7.25 8.60 8 6.75 8.25 9.25 7 10.25 8.25 9.75 9 9.25 10.25 8.25 9 9 22 1889 1890 1869 1891 1892 1893 1870 1872 . 1894 1895 1896 1897... 1898 1899 1873 1874 1875 1876 1877 1878 — -- 1900 1901 1902 1903 1904 1905 1906 1907 1879 1880 1881 1882 - 1883 1884 1885 1886 -- Note.— See fig. 3. 277589 — 41- 66 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Table 4. — Number of striped bass taken each year in pound-nets at Fort Pond Bay, Long Island, N. Y., 1884-1987 Date Number of striped bass Number of pound- nets in operation 6 6 6 6 6 6 7 7 7 7 8 8 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 Date Number of striped bass Number of pound- nets in operation 1884 3,630 1,872 4.354 2,688 2,046 915 720 636 455 1,953 3,643 3,689 35 895 708 189 1, 551 1,310 348 1,107 219 64 3,374 926 425 300 496 1911 221 702 378 1,579 236 804 197 1,310 157 463 240 1,976 401 878 389 321 121 184 100 325 500 35 50 100 400 15,600 4,200 9 9 9 10 10 9 8 7 7 7 7 7 7 7 7 7 7 6 7 8-12 8-12 8-12 8-12 8-12 8-12 8-12 12 12 1885 1912 1886 1913 1887 1914 1888 1915... _ 1889 1916 1890 1917 1891 1918 1892 1919... 1893 1920-. 1894-.. 1921. 1895... 1922 _ 1896.. . .. 1923 1897 1924 1898. 1925-. 1899 1926 1900 1927 1901 - 1928 1902. 1929 1903 1930 1904 1931... _ 1905. . 1932 1906 .- 1933 1907 1934 _ 1908. 1935. 1909 1936 1910 1937 Note.— See figs. 4 and 24. Table 5. — Length-frequency distribution of striped bass making up the random samplings of the com- mercial catch in Cape Cod Bay, at Newport, R. I., and at Montauk, Long Island, N. Y., in 1937 Length (cm.) Number of individuals Length (cm.) Number of individuals Cape Cod Bay Newport, E.I. Montauk, Long Island, N. Y. Cape Cod Bay Newport, E.I. Montauk, Long Island, N. Y. 20 22 1 1 1 57 58 4 2 3 2 4 3 2 2 3 2 3 2 2 2 2 1 3 1 28 69 1 32 1 60 34 1 2 6 3 16 22 17 31 21 21 19 19 17 20 12 23 16 15 11 9 7 5 3 61- . 35 1 2 5 9 12 24 21 40 66 61 34 39 26 31 12 18 6 6 4 2 62 . 1 1 36 63 37 2 1 4 8 22 28 29 44 39 44 25 21 17 14 24 21 12 7 10 64 38 65 . 39 66.. 40 67 41 68 42 70 43 71 44 73.... 45 77.... 46 - 47 78 80.... 48 81... 49 84 ... 60 90 .. 51 96.. . 52 102 1 63 108. 2 Total 55 366 378 413 66 1 Note.— See fig. 6 for length-frequency curves smoothed by threes made up from this material. STUDIES ON THE STRIPED BASS OF THE ATLANTIC COAST 67 Table 6. — Total catch of striped bass by seine at Point Judith, Ii. I., 1928-37 Date Num- ber rounds Number of days fishing (equalizing factor) Average weight (pounds) Date Num- ber Pounds Number of days fishing (equalizing factor) Average weight (pounds) 1928 1929 1930 1931 1932 225 1,050 600 775 1.375 1,925 5,700 4.825 5,200 8.800 19 (X4.4) 83 (XI. 0) 70 (XI. 2) 4S (XI. 7) 60 (XI. 4) 8.5 6.4 8.0 6.6 6.4 1933_ 1934. 1935 1936_ 1937 1,513 234 1,2.50 7,500 4,500 9,625 1,300 7,000 IS. 000 12.000 66 (XI. 3) 31 (X2.7) 58 (XI. 4) 49 (XI. 7) 44 (XI. 8) 6.2 6.5 5.6 2.4 2.7 Note.— See figs. 6 and 7. Table 7. -Length-frequency distribution of juvenile striped bass from the Hudson River, July S-Sept. 1, 1936 Length (mm.) Number of indi- viduals at each milli- moter Length (mm.) Number of indi- viduals at each milli- meter Length (mm.) Number of indi- viduals at each milli- meter 18 12 11 10 12 7 6 13 7 11 8 6 8 6 5 1 3 6 8 4 3 Length (mm.) Number of indi- viduals at each milli- meter 20 1 49 16 14 16 23 15 17 19 22 18 17 19 28 17 17 19 13 10 17 10 18 12 70 91 2 1 3 30 60-... 71 92 61 72 93 31 62 73 94 32 63 74 95 4 33 1 1 2 4 2 64 75 _ 96 34 66 76 97 1 35 66 77 98 36 57 78 99 37 58... 79. 100 38 69. _ 80... 101 39 2 2 4 8 7 8 10 15 10 16 60 81 102 1 40 61 82 103 41 62 63 83 104 1 1 42 84... 106 43 64 86 106 44 66 86 107 1 46 - 66 87 . Total.. 46... 67 88.. 828 47 68 .. 89... 48 69... 90 .. Note.— See fig. 10 for length-frequency curve of this material smoothed by threes. Table 8.- — Length-frequency distribution of juvenile Table 9. — Length-frequency distribution of juve- and yearling striped bass taken in the Delaware nile striped bass taken in Albemarle Sound, iX. C, River, near Pennsnlle, N. J., Nov. 8, 1937 on May 11, 1938 Length (cm.) Number of indi- viduals 10 1 4 4 3 5 6 6 13 19 4 11 11 10 4 2 1 11 12 13.. 14.. 15 16... 17 18 19... 20 21... 22 23 24 26 Total 104 Length (mm.) Number of indi- viduals 20 1 1 3 7 10 9 12 21 12 9 21 22 23 24.... 16. 28.... 27 28 29 Total 85 Note.— See fig. 14 for length-frequency curves of this material smoothed by threes. Noti.— See fig. 11 for length-frequency curve of this material smoothed by threes. 68 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Table 10. — Age at maturity of 109 female striped bass of known length 2-year-olds (num- ber of flsh) 3-year-olds (num- ber of fish) 4-year-olds (num- ber of fish) 5-year-olds (num- ber of flsh) 6-year-olds and over (number of fish) Imma- ture Mature Imma- ture Mature Imma- ture Mature Imma- ture Mature Imma- ture Mature 1 2 1 2 2 3 1 2 3 1 2 1 1 36- -- -- 1 2 2 2 3 2 1 1 1 2 3 2 3 6 3 5 7 1 1 1 2 1 1 1 1 6 1 1 1 2 1 1 Total { 22 100% I 14 6 26.6% 9 25 73. 5% 1 14 93.3% 19 100% Note. — Those Individuals were listed as mature if their ova had attained sufficient si7.e to indicate that spawning would occur the following season. See teit (p. 21). Table 11. — Length-frequency distribution of all striped bass measured in Connecticut waters from April through October, 1936 and 1937 Length (cm.) Number of individuals Length (cm.) Number of individuals Length (cm.) Number of individuals 1936 1937 1936 1937 1936 1937 23 3 4 8 16 21 43 61 83 121 138 190 174 198 162 136 81 35 53 36 36 28 16 27 15 25 23 1 2 6 22 50 62 85 127 111 111 118 102 100 81 72 70 67 43 40 30 25 24 20 49 11 13 12 6 7 11 6 7 8 6 7 9 5 2 6 4 5 10 6 7 6 4 4 4 4 3 16 17 9 6 -7 7 8 6 5 2 2 2 2 2 1 2 1 i 75 3 4 3 1 1 3 24 50 76 25 51 77 26 62 78 27 63 _ 79 28 64 80 29 65 81 2 3 1 2 2 1 30 66 82 31 67..- 83 32 58 84 33 69, 86 34 60. - 86 35 61 87 36 62 88 2 37 63 89 38 64 90 1 2 2 39 65 91 40 66 92 41 67. 93 1 42 68 94 43 69 95 1 44 70 96 71 97 1 72 Total.. 47 73 1,933 1,460 48..- 74. Noti.— See fig. 17 for length-frequency curves of this material smoothed by thiees. STUDIES ON THE STRIPED BASS OF THE ATLANTIC COAST 69 Table 12. — Length- frequency distribution of 2- and 3-year-old striped bass seined in Connecticut waters during 1936 and 1937, grouped by months Number of Individuals Length In 2-year-olds ,1938 3-year-olds, 1936 2-year-olds, 1937 3-year-olds, 1937 June July Aug. Sept. Oct. June July Aug. Sept. Oct. June July Aug. Sept. Oct. June July Aug. Sept. Oct. 25 1 6 8 12 17 23 25 17 15 4 3 4 4 1 1 3 26 1 9 15 11 20 12 10 10 9 8 6 2 1 2 1 1 2 4 16 25 22 28 23 9 11 6 7 2 1 6 10 27 23 34 19 26 8 10 6 4 7 11 4 3 1 1 1 7 9 25 44 68 69 80 64 35 16 7 7 8 7 4 4 3 3 2 2 11 14 21 22 21 21 17 9 6 3 3 1 1 2 29 2 4 11 17 32 22 24 24 32 13 11 8 3 2 1 2 30 1 31 l l 2 7 8 14 11 13 7 7 3 "Y 32 6 7 18 17 24 38 24 26 8 8 10 4 3 6 6 6 11 9 15 14 16 12 6 11 2 2 1 2 1 1 6 11 6 12 16 14 6 10 8 6 7 1 4 33 3 5 5 4 6 7 6 7 7 10 7 5 8 7 8 6 2 3 34 35 36 1 2 5 4 12 4 6 2 2 2 2 1 3 1 6 8 2 2 1 2 1 1 1 2 3 4 4 6 10 10 3 5 2 3 4 3 3 2 1 1 37 1 1 38 39 2 2 40 1 1 3 41 4 42 1 1 5 3 6 2 7 6 2 1 1 2 43 -- 2 5 44 11 11 4 47 4 2 1 6 48 7 49 9 50 5 3 62 1 1 2 63 1 3 65 4 56 3 67 4 145 192 39 12 Total 116 21)1 451 34 10 8 156 156 208 75 118 108 102 33 66 88 Note. — See fig. 18 for length-frequency curves smoothed by threes to show growth from June to October each year. Table 13. — Length-frequency distribution of 2- and S-year old striped bass taken north and south of Cape Cod, June to September 1937 Length (cm.) 2-year-olds (number of Individuals) Length (cm.) 3-year-olds (number of individuals) North of Cape Cod South of Cape Cod North of Cape Cod South of Cape Cod 25 30 26 31 27 32 1 6 7 7 12 8 11 14 10 10 6 8 7 3 3 4 3 1 3 28 33 29 1 34 30 4 8 6 13 7 8 8 3 4 2 1 35 31 . 1 3 3 1 2 6 36 37 2 32 33 38 3 4 9 16 7 15 14 28 3 22 11 11 19 4 4 11 8 5 1 5 2 1 34 39 -... 35 40 36 41 37 42. 38 9 9 17 8 9 9 4 6 43 39 44 40 45 41 46 42 1 1 47 43 48 44 49 50 46 61 47 1 52 63.. 48 49 54. 60 65 --. Total 88 66 57 68 - 59 60 Total 205 124 Note.— See fig. 19 for length-frequency curves of this material smoothed by threes. 70 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Table 14. — Average lengths of striped bass at the time they become 1 year old, 2 years old, etc., to 9 years old Age 1 year old - 2 years old 3 years old 4 years old 5 years old Average length Centi- meters 12.5 23.5 36.5 45.0 63.0 Inches 4.92 9.25 14.37 17.72 20.87 Age 6 years old 7 years old 8 years old 9 years old Average length Centi- meters 61.0 68.5 75.0 82.0 Inches 24.02 26.97 29.53 32.28 Note.— See Dg. 20. Table 15. — Original measurements of the radii of scales from 153 striped bass of measured length from 10.5-87 centimeters long Length (em.) Scale radius (mm.) Length (cm.) Scale radius (mm.) 10.5 11.0 11.5 12.5 13.5 14.5 -- 1.22. 1.37, 1.37. 1.52, 1.59. 1.95, 1.59. 1.81. 1.79,1.70,1.86. 1.92, 1.85. 2.02, 2.09. 1.95. 2.09, 2.24, 2.24. 2.24, 2.09. 2.24, 2.39, 2.31, 2.09, 2.24, 2.37, 2.31. 2.24, 2.53, 2.46. 2.35, 2.35. 2.60, 2.39. 2.53. 2.65, 2.67, 2.53. 2 89, 2.74, 2.43. 2.67. 2.67,2.69,2.77,3.10. 3.03, 2.82. 2.89. 2.70, 2.86. 3.14. 3.40. 3.03. 3.62. 3.36. 3.32,3.58. 3.83. 3.99. 3.90, 3.69. 3.62,4.12. 3.62. 4.12,4.12. 4.19,3.83. 4.19,4.34,4.56,4.05. 32.0. 32.5 33.0 34.5 35.0 35.5 36.0 3.76. 4.16,4.56. 4.12. 4.48,4.30,4.19,4.09,4.02. 4.05,4.26,4.48,4.26. 4.38, 4.26, 5.03, 4.84, 4.48. 4.55,4.84,4.34. 4.52, 4.56, 4.30. 6.10,4.78,4.38. 4.67,4.41,4.56. 4.84, 4.84, 4.91, 4.39, 4.70, 6.06. 4.88, 4.42, 5.27. 5.24, 5.24. 5.20,5.24,4.91. 5.35,4.70.4.91. 5.13,5.49,5.28. 5.67. 5.56,6.11. 5.75. 6.43. 6.18. 5.99. 5.71. 6.40,6.00,6.43. 6.40,6.28,5.85. 6.36. 6.57,6.36,6.71. 6.07, 6.14. 6.36. 7.00. 6.79. 6.93. 7.87. 8.73. 9.17. 15.0 15.5 -. 36.5 37.0 16.0.. 16.5 37.5... 17.5. 18.0 38.0 . 39.6 40.0 18.5. 40.5 41.0 42.0 43.0.. 43.5... 45.0 45.6.. 46.0 46.5 19.5 20.0 20.5 21.5 22.0 22.6 23.0 24.0.. 24.6 25.0 26.0 26.5 47.0 47.5 48.0 48.5 50.0 27.0 51.0 _ 27.5 28.0 29.0 — 29.5 62.0 63.5. 64.0 55.0 62.0 63.0 67.0. 30.0... 30.5 31.0 31.5 Note.— See fig. 21 for graph of relationship of scale growth to body growth in the striped bass, plotted from data in this table Table 16. — Annual increment in the length of the striped bass Age First year Second year. Third year.. Fourth year . Fifth year. . . Sixth year... Seventh year Eighth year. Increment Centi- meters 12.6 11.0 13.0 8.5 8.0 8.0 7.6 6.5 Inches 4.92 4.33 5.12 3.35 3.15 3.15 2.95 2.66 Note.— See fig. 22. STUDIES ON THE STRIPED BASS OF THE ATLANTIC COAST 71 Table 17. — Returns from 1,397 striped bass tagged in Connecticut, Apr. S3 to Oct. 27, 1936 Date of return Total number tagged by the end of each month Original point of release Number of returns each month Locality of recapture Total number of returns each month May 1936 121 331 483 792 L217 1,397 1,397 1,397 1,397 1,397 1,397 1, 397 1,397 1,397 1,397 1,397 1,397 1,397 2 2 6 2 17 3 10 1 3 2 70 3 30 2 1 34 10 4 4 59 7 1 1 2 2 1 3 1 1 2 1 1 2 1 1 8 1 1 1 5 I 1 12 3 1 5 1 1 1 1 1 1 do do Point Judith, R. I do. . 12 ...do Niantic River, Conn .do .. 20 July 1936 do ... .do Thames River, Conn 11 do ..- 5 September 1936 73 do .do Nlantic and Thames Rivers, Conn. South shore of Long Island, N. Y.. 77 do Niantic and Thames Rivers, Conn. do Montauk, Long Island, N. Y South shore Long Island, N. Y Bradley Beach, N. J 74 . do . Rehoboth Beach, Del 7 Toms River, N. J ... do Columbia, N. C 5 Toms River. N. J Rehoboth Beach, Del 3 1 April 1937 ...do -. ..do Hudson River, N . Y Oyster Bay, Long Island, N. Y 4 May 1937 do do .. Wye River, Md . ..do ..do .do .. . 18 June 1937 ..do 16 July 1937 do 6 do 2 September 1937 do. . ...do do ... 2 May 1938 Hudson River, N. Y 1 Total recap- tures. Total percent- age recap- tured. 337 24. 1 72 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Table 18. — Returns from 103 striped bass tagged and released at Fort Pond Bay, Montauk, Long Island, N. Y., May 15-19, 1987 Date of return Number of returns each month Locality of recapture Total number of returns each month May 1937 June 1937... Point Judith, R. I 3 1 July 1937 August 1937... 3 October 1937 5 1 1 May 1938 14 13.6 Total percentage recaptured. Table 19. — Returns from 303 striped bass tagged and released at Fort Pond Day, Montauk, L. I. Oct. 25, 26, and 27, 1937 N. Y., Date of return October 1937 November 1937. December 1937.. January 1938.. February 1938.. March 1938.. April 1938- May 1938.. June 1938.. Total recaptures. . . Total percentage recaptured. Number of returns each month Locality of recapture Gardiners Bay, Long Island, N. Y. Montauk, Long Island, N. Y Gardiners Bay, Long Island, N. Y. Montauk, Long Island, N. Y _ South shore of Long Island, N. Y... Monmouth Beach, N. J Barnegat Bay, N. J South shore of Long Island, N. Y. . Mullica River, N. J._ Indian River, Del Rappahannock River, Va.. ._ Great Choptank River, Md Cape Charles, Va Croatan Sound, N. C Stumpy Point, N. C Pamlico Sound, N. C Barnegat Bay, N. J. __ Mullica River, N. J Egg Harbor, N.J Synapuxent Bay, Md South shore of Long Island, N. Y._ Barnegat Bay, N. J.. Great Egg Harbor River, N. J Rappahannock River, Va __ Hudson River, N. J Barnegat Bay, N. J — Great Egg Harbor River, N. J Rappahannock River, Va New Point, Va Kitty Hawk, N. C Great Bay, N. J York River, Va Potomac River, Va Rappahannock River, Va Plymouth, Mass Point Judith, R. I Asbury Park, N. J Oak Bluffs, Mass Chatham, Mass Total number of returns each month 25 35 100 33.0 STUDIES ON THE STRIPED BASS OF THE ATLANTIC COAST 73 Table 20.— Returns from 770 striped bass tagged in Connecticut, Apr. 19-Oct. SO, 1937 Date of return Total number tagged by the end of each month Original point of release Number of returns each month Locality of recapture Total number of returns each month 182 434 614 628 770 770 770 770 770 770 770 770 Niantlc River, Conn 3 1 4 11 9 2 2 2 1 1 2 1 11 1 1 1 1 4 1 1 1 1 3 4 1 1 1 1 1 1 2 2 6 1 1 3 do 4 July 1937 do Thames River, Conn Thames River, Conn do.. - Niantic River, Conn September 1937 do Thames River, Conn do - . do October 1937 Niantic River, Conn do do do .-.. do Thames River, Conn Niantic River, Conn ....do ....do 21 November 1937 Niantic River, Conn Niantic River, Conn. do Thames River, Conn South shore of Long Island, N. Y .do .. . South shore of Long Island, N. Y do December 1937 Niantic River, Conn do 3 January 1938 ...do BroadkiU River, Del March 1938 do Thames River, Conn. Hudson River, N. Y . ..do. 3 April 1938 Niantic River, Conn do do 5 May 1938 Niantic River, Conn. do do Thames River, Conn Niantic River, Conn 3 Juno 1938 do 8 Total recap- S3 Total precen- tage recap- 12 1 Table 21. — Chemical analysis of the water at 2 stations in the Thames River, Conn., in the of 1987* Locality Date PH Dis- solved oiygen, parts per million Chloride, parts per million Sulfate, parts per million Calcium, parts per million Phos- phates, parts per million Off the submarine base, 1 mile abovo New London on the (June 2 Uuly 1 (Sept. 15 (June 2 {July 1 ISept. 15 7.70 7.64 7.59 7.82 7.74 7.09 7.76 6.30 5. 11 8.80 7.10 6.07 13, 350 14,250 15,350 15,100 15,500 16,400 1,834 2,027 2,176 2,133 2,279 2,279 316 364 254 314 346 400 0.30 Off the State pier at New London, on the west side of the .69 .20 .52 1.38 1 These water analyses were supplied by the Connecticut State Water Commission. The samples were taken as catch samples, and therefore in no way represent a complete tidal cycle. The 2 localities listed above are both places where striped bass are com- monly caught, and where a good number of bass were found dead in late August and early September 1937. 74 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Table 22A. — Returns from 52 striped bass tagged Table 22B. — Returns from 17 striped bass tagged and released at extreme west end. of Albemarle and released off Coinjock, Currituck Sound, Sound, N. C, Mar. 26, Apr. 9, and 21, 1937 N. C, Mar. 27, 1937 Date of return Number of returns each month Locality of recapture Total number of returns each month March 1937 6 5 1 1 4 1 1 April 1937 Edenton, N. C . Columbia, N. C - . Pasquotank River, N. C. 12 Total recap- tures. Total per- centage re- captured. Edenton, N. C ... Hertford, N.C.... 7 19 36.5 Date of return Number of returns each month Locality of recapture Total number of returns each month October 1937 November 1937 December 1937 Total re- captures. Total per- centage re- captured. 1 1 1 1 Currituck Sound, N. C. Kitty Hawk, N.C. Currituck Sound, N.C. Currituck Sound, N.C. 1 2 1 4 23.6 Table 22C. — Returns from 8 striped bass lagged and released at Kitty Hau'k, N. C. (outer coast), Apr. 29 and May 10, 1937 Date of return Number of returns each month Locality of recapture Total number of returns each month 1 1 ] 12.5 Table 23. — Original measurements of Menidia menidia notata to show growth of juveniles from July through September 1937 in the Niantic River, Conn. Standard length in millimeters Number of individuals at each length Standard length in millimeters Number of individuals at each length July 17 Aug. 2 Aug. 17 Sept. 2 July 17 Aug. 2 Aug. 17 Sept. 2 6 2 24 6 16 9 3 4 E 6 7 3 6 2 11 10 6 7 2 1 1 1 24 20 16 21 10 3 3 2 6 26 1 7 13 22 23 29 21 14 5 3 2 1 1 2 5 10 13 13 19 22 22 16 16 10 3 8 6 8 3 26 27 9 28. 1 29 2 1 1 1 1 30... 1 2 6 17 16 16 29 20 17 11 12 13 31. 32 i 2 14 33... 3 2 2 1 2 34 1 36 36 1 fi 8 18 27 26 2 19 38 39 1 Total 22 1 5 200 200 200 197 23 Note.— See flg. 36 for length-frequency curves of this material smoothed by threes. LITERATURE CITED Bean, T. H. 1884. On the occurrence of the striped bass in the Lower Mississippi Valley. Proc. U. S. Nat. Mus., vol. 7, pp. 242-244. Washington. Bigelow, H. B. and W. W. Welsh. 1925. Fishes of the Gulf of Maine. Bull. U. S. Bur. Fish., vol. XL, pt. I, 1924 (1925), pp. 201, 251-256. Washington. Caulkins, F. M. 1852. History of New London, Connecticut, p. 610. Published by the author, New London, Conn. Clark, G. H. 1932. The striped bass supply, past and present. Calif. Fish and Game, vol. 18, No. 4, pp. 297-298. Sacramento. 1933. Fluctuations in the abundance of striped bass (Roccus lineatus) in California. Div. Fish and Game Calif., Fish Bull., vol. 39, pp. 1-18. Sacramento. 1936. A second report on striped bass tagging. Calif. Fish and Game, vol. 22, No. 4, pp. 272-283. Sacramento. 1938. Weight and age determination of striped bass. Calif. Fish and Game, vol. 24, No. 2, pp. 176-177. Sacramento. Coleman, G. A. and N. B. Scofield. 1910. Notes on spawning and hatching of striped bass eggs at Bouldin Island Hatchery. 21st Bienn. Rept., Calif. Board of Fish and Game Comm., 1909-1910, p. 109. Sacra- mento. Craig, J. A. 1930. An analysis of the catch statistics of the striped bass (Roccus lineatus) fishery of Cali- fornia. Div. Fish and Game Calif., Fish Bull. No. 24, pp. 1-41. Sacramento. Creaser, C. W. 1926. The structure and growth of the scales of fishes in relation to the interpretation of their life-history, with special reference to the sunfish Eupomotis gibbosvs. Mus. of Zool., Univ. of Mich., Misc. Pub. No. 17, pp. 1-80. Ann Arbor. Curran, H. W. and D. T. Ries. 1937. Fisheries investigations in the Lower Hudson River. Biological Survey (1936) No. XI, pp. 124-128. State of N. Y. Cons. Dept. J. B. Lyon Co. Albany. Dahl, K. 1907. The scales of the herring as a means of determining age, growth and migration. Rept. Norweg. Fish, and Mar. Inv., vol. 2, pt. 2, No. 6. 36 pp. Goode, G. B. and associates. 1884. The fisheries and fishery industries of the United States, vol. 1, Sec. Ill, pp. 425-428. Washington. Gowanloch, J. N. 1933. Fishes and fishing in Louisiana. State of La. Dept. of Cons., Bull. No. 23, pp. 208- 213. New Orleans. Graham, M. 1935. Modern theory of exploiting a fishery, etc. Jour, du Conseil Internat. pour l'Ex- ploration de la Mer, vol. 10, No. 3, pp. 264-274. Herrington, W. C. 1935. Modifications in gear to curtail the destruction of undersized fish in otter trawling. U. S. Bur. Fish., Inv. Rept. No. 24., vol. 1, 48 pp. Washington. Hess, W. N. 1937. Production of nutritional cataract in trout. Proc. Soc. Kxp. Biol, and Med., vol. 37, pp. 306-309. HlLDEBRAND, S. F. 1922. Notes on habits and development of eggs and larvae of the silversides, Menidia menidia and Menidia beryllina. Bull. U. S. Bur. Fish., vol. XXXVIII (Doc. No. 918), pp. 113-120. Washington. and W. C. Schroeder. 1928. Fishes of Chesapeake Bay. Bull. U. S. Bur. Fish., vol. XLIII, pt. 1, 1927, pp. 247-250. Washington. Hubbs, C. L. 1922. Variations in the number of vertebrae and other meristic characters of fishes correlated with the temperature of water during development. Am. Nat., vol. 56, pt. 645, pp. 360-372. and R. M. Bailey. 1938. The small-mouthed bass. Cranbrook Institute of Science, Bull. No. 10, pp. 1-92. 75 76 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Jahn, T. L. and L. R. Kuhn. 1932. The life history of Epibdella melleni MacCallum, 1927, a nionogenetic trematode para- sitic on marine fishes. Biol. Bull., vol. 62, pp. 89-111. Jensen, A. J. C. 1932. The effect of the plaice fishery on the stock of undersized plaice and its influence on the yield of the plaice fishery in the North Sea. Rapp. et Proc. Verb, des Reunions (Cons. Perm. Internat. pour l'Exploration de la Mer), vol. 80. Jordan, D. S. 1884. Manual of the vertebrates of the Northern United States, p. 231. Jansen, McClurg and Co. Chicago. 1929. Manual of the vertebrate animals of the Northeastern United States. 13th Ed., p. 172. World Book Co. Yonkers-on-Hudson, N. Y. and B. W. Evermann. 1905. American food and game fishes, pp. 373-375. Doubleday, Page and Co. New York. -, B. W. Evermann and H. W. Clark. 1930. Check list of the fishes and fishlike vertebrates of North and Middle America north of the northern boundary of Venezuela and Colombia. Rept. U. S. Comm. Fish., 1928, pt. 2, p. 307. Washington. Lea, E. 1918. Report on the age and growth of the herring in Canadian waters. Canadian Fisheries Expedition, 1914-1915. Ottawa. and A. E. Went. 1936. Plastic copies of microscopical reliefs, especially of fish scales. Hvalradets Skrifter, Nr. 13, pp. 1-18. Oslo. Linton, E. 1898. Notes on trematode parasites of fishes. Proc. U. S. Nat. Mus., vol. 20, pp. 507-548. Washington. 1901. Parasites of fishes of the Woods Hole region. Bull. U. S. Fish. Comm., vol. XIX, pp. 405-492. Washington. 1924. Notes on cestode parasites of sharks and skates. Proc. U. S. Nat. Mus., vol. 64, Art. 21, pp. 1-114. Washington. MacCallum, G. A. 1921. Studies in helminthology. Pt. 3, Nematodes. Zoopathologica, vol. 1, No. 6, pp. 255- 284. Mason, H. W. 1882. Report of operations on the Navesink River, New Jersey, in 1879, in collecting living striped bass for transportation to California. Rept. U. S. Comm. Fish and Fish., 1879, pp. 663-666. Washington. Masterman, A. T. 1913. Report on investigations upon the salmon. Bd. of Agric. and Fish., Fish. Inv., vol. 1. Merriman, D. 1937a. Notes on the life history of the striped bass (Roccus lineatus). Copeia, No. 1, pp. 15- 36. Ann Arbor. 1937b. Notes on the life history of the striped bass. Trans. Sec. N. Am. Wildlife Conf., pp. 639-648. American Wildlife Institute, Investment Bldg., Washington. 1938. A report of progress on the striped bass investigation along the Atlantic coast. Trans. 3d N. Am. Wildlife Conf., pp. 478-485. American Wildlife Institute, Investment Bldg., Washington. Moore, E., et al. 1937. A biological survey of the Lower Hudson Watershed. Biological Survey (1936), No. XI, pp. 16, 62, 76, 127-128, 225-226, and 260-261. State of N. Y. Cons. Dept. J. B. Lyon Co. Albany. Mueller, J. F. 1936. New gyrodactyloid trematodes from North American fishes. Trans. Am. Fish. Soc, vol. 55, pp. 457-464. Washington. Nesbit, R. A. 1934a. A convenient method for preparing celluloid impressions of fish scales. Jour, du Conseil Internat. pour l'Exploration de la Mer, vol. 9, No. 3, pp. 373-376. 1934b. A new method of marking fish by means of internal tags. Trans. Am. Fish. Soc., vol. 63, pp. 306-307. Washington. Nigrelli, R. F. and C. M. Breder. 1934. The susceptibility and immunity of certain marine fishes to Epibdella melleni, a mono- genetic trematode. Jour. Parasitol., vol. 20, pp. 259-269. Nornt, E. R. 1882. On the propagation of the striped bass. Bull. U. S. Fish Comm., vol. I, 1881, pp. 67-68. Washington. Parr, A. E. . 1933. A geographic-ecological analysis of the seasonal changes in temperature conditions in shallow water along the Atlantic coast of the United States. Bull. Bingham Oceano- graphic Collection, vol. 4, No. 3, pp. 1-90. 1937. On self-recognition and social reaction in relation to biomechanics, with a note on termi- nology. Ecology, vol. 18, No. 2, p. 321-323. Pearson, J. C. 1933. Movements of striped bass in Chesapeake Bay. Maryland Fisheries, No. 22, pp. 15-17. 1938. The life history of the striped bass, or rockfish, Roccus saxatilis (Walbaum). Bull. U. S. Bur. Fish., vol. XLIX, No. 28, pp. 825-851. Washington. STUDIES ON THE STRIPED BASS OF THE ATLANTIC COAST 77 Railliet, A. 1918. Le genre Dieheilonema Diesing, 1861 (Nematoda, Filarioidea) . Bull. Soc. Zool. de France, vol. 43, pp. 104-109. Russell, E. S. 1932. Fishery research: its contribution to ecology. The Journal of Ecology (edited for the British Ecological Society by A. G. Tansley), vol. 20, pp. 128-151. Schultz, L. P. 1931. Hermaphroditism in the striped bass. Copeia, No. 2, p. 64. Ann Arbor. and A. D. Welander. 1935. A review of the cods of the Northeastern Pacific with comparative notes on related species. Copeia, No. 3, pp. 127-139. Ann Arbor. Scofield, E. C. 1931. The striped bass of California (Roccus lineatus). Div. Fish and Game Calif., Fish. Bull. No. 29, pp. 1-82. Sacramento. Smith, H. M. 1907. The fishes of North Carolina. N. Car. Geol. and Econ. Surv., vol. 2, pp. 271-273. Raleigh. Thompson, W. F. 1937. Theory of the effect of fishing on the stock of halibut. Rept. Internat. Fish Comm., No. 12, pp. 1-22. Seattle. and F. H. Bell. 1934. Biological statistics of the Pacific halibut fishery. (2) Effect of changes in intensity upon total yield and yield per unit of gear. Rept. Internat. Fish. Comm., No. 8, pp. 1-49. Seattle, and W. C. Herrington. 1930. Life history of the Pacific halibut. (1) Marking experiments. Rept. Internat. Fish. Comm., No. 2, pp. 1-137. Seattle. Throckmorton, S. R. 1882. The introduction of striped bass into California. Bull. U. S. Fish Comm., vol. I (1881), pp. 61-62. Washington. Townes, H. K., Jr. 1937. Studies on the food organisms of fish. Biological Survey (1936), No. XI, pp. 225-226. State of N. Y. Cons. Dept. J. B. Lyon Co. Albany. Truitt, R. V. and V. D. Vladtkov. 1937. Striped bass investigations in the Chesapeake Bay. Trans. Am. Fish. Soc, vol. 66 (1936), pp. 225-226. Van Oosten, J. 1929. Life history of the lake herring (Leucichthys arledi Le Sueur) of Lake Huron as revealed by its scales, with a critique of the scale method. Bull. U. S. Bur. Fish., vol. XLIV, 1928, pp. 265-428. Washington. Vladtkov, V. D. and D. H. Wallace. 1938. Is the striped bass (Roccus lineatus) of Chesapeake Bay a migratory fish? Trans. Am. Fish. Soc, vol. 67 (1937), pp. 67-86. Walford, Lionel A. 1937. Marine game fishes of the Pacific coast from Alaska to the equator. Univ. of Calif. Press, pp. 93-97. Berkeley. Watkin, E. E. 1927. Investigations on Cardigan Bay herring. Kept. Mar. and Fresh Water Inv., vol. 2, pt. 5, Dept. Zool., Univ. Coll., Wales. Wilson, C. B. 1903. North American parasitic copepods of the family Argulidae, etc. Proc U. S. Nat. Mus., vol. 25, pp. 635-742. Washington. 1905. North American parasitic copepods belonging to the family Caligidae. Proc. U. S. Nat. Mus., vol. 28, pp. 479-672. Washington. 1911. North American parasitic copepods belonging to the familv Ergasilidae Proc. U. S. Nat. Mus., vol. 39, pp. 263-400. Washington. 1915. North American parasitic copepods belonging to the Lernaeopodidae, with a revision of the entire family. Proc. U. S. Nat. Mus., vol. 47, pp. 565-729. Washington. 1932. The copepods of the Woods Hole region, Massachusetts. Bull. U. S. Nat. Mus., vol. 158, pp. 1-635. Washington. Wilson, H. V. 1891. The embrvology of the sea bass (Serranus atrarius). Bull. U. S. Fish Comm., vol. IX, 1889, pp". 209-277. Washington. Wood, W. 1635. New England's Prospect. Tho. Coates for John Bellamie, 83 pp. London. Worth, S. G. 1903. Striped bass hatching in North Carolina. Trans. Am. Fish. Soc, vol. 32 (1902), pp. 98-102. 1904. The recent hatching of striped bass, etc. Trans. Am. Fish. Soc, vol. 33 (1903), pp. 223-230. 1910. Progress in hatching striped bass. Trans. Am. Fish. Soc, vol. 39 (1909), pp. 155-159. 1912. Fresh-water angling grounds for the striped bass. Trans. Am. Fish. Soc. vol. 41 (1911), pp. 115-126. UNITED STATES DEPARTMENT OF THE INTERIOR Harold L. Ickes, Secretary FISH AND WILDLIFE SERVICE Ira N. Gabrielson, Director Fishery Bulletin 36 THE YOUNG OF SOME MARINE FISHES TAKEN IN LOWER CHESAPEAKE BAY, VIRGINIA, WITH SPECIAL REFERENCE TO THE GRAY SEA TROUT Cynoscion regalls (BLOCH) By JOHN C. PEARSON From FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Volume 50 UNITED STATES GOVERNMENT PRINTING OFFICE WASHINGTON : 1941 For tale by the Superintendent of Documents. Washington. D. C. --------._.. Price 10 c ABSTRACT Plankton collections made at the mouth of Chesapeake Bay, Va., yielded specimens of 45 species of marine fishes that were recognized. As a result of these weekly collections during the summer and biweekly collections during the winter, from May to October 1929, from April to December 1930, and during January and March 1931, sufficient data were acquired to provide distributional and descriptive data on 31 of the 45 species recognized. Larval and postlarval stages of the gray sea trout, or weakfish, Cynoscion regalis; the bluefish, Pomatomus saltatrix; the butterfish, Poronotus triacanlhus; the harvestfish, Peprilus alepidolus; and the stargazer, Astroscopus guttatus, are described and illustrated. Collections of juvenile gray sea trout by seine and trawl indicate that this food fish attains an average total length of 16 to 20 cm. at the end of its first year of growth in lower Chesapeake Bay. THE YOUNG OF SOME MARINE FISHES TAKEN IN LOWER CHESAPEAKE BAY, VIRGINIA, WITH SPE- CIAL REFERENCE TO THE GRAY SEA TROUT Cynoscion regalis (BLOCH) J- By John C. Pearson, Aquatic Biologist, Fish and Wildlife Sendee CONTENTS Page Introduction 79 Methods 80 Planktonic fishes 80 Brevoortia tyrannus (Latrobe). Men- haden; Fatback 83 Anchoviella milchilli (Cuvier and Valenciennes). Anchovy 83 Conger conger (Linnaeus). Conger eel 83 Lophopsetla maculata (Mitchell). Windowpane 83 Elropus sp. Etrope 83 Paralichthy s sp. Flounder 84 Ancylopsetta sp. Flounder 84 Achirus fascialus Laeepede. Amer- ican sole; Hogchoker 84 Symphurus plagiusa (Linnaeus). Tonguefish 84 Syngnathus jloridae (Jordan and Gil- bert). Pipefish 84 Syrictes fuscus (Storer). Common pipefish 85 Hippocampus hudsonius DeKay. Sea- horse 85 Menidia menidia (Linnaeus). Silver- side 85 Peprilus alepidolus (Linnaeus). Har- vestfish . 85 Poronolus triacanthus (Peck). Butter- fish 87 Pomatomus saltalrix (Linnaeus). Blue- fish 89 Page Planktonic fishes — Continued. Cenlropristes slriatus (Linnaeus). Sea bass; Blackfish 90 Bairdiella chrysura (Laeepede). Sand perch 91 Micropogon undulalus (Linnaeus). Croaker 91 Menticirrhus americanus (Linnaeus). Kingfish; Whiting 92 Cynoscion tegalis (Bloch and Schnei- der). Gray sea trout; Weak- fish; Squeteague 92 Prionotus sp. Sea robin 97 Tauloga onitis (Linnaeus). Tautog.. 97 M irrogobius lhalassinus (Jordan and Gilbert). Scaled goby. 98 Gobiosoma sp. Naked goby 98 Aslroscopus guilatus Abbot. Star- gazer 98 Hypsoblennius hentz (Le Sueur). Blenny 99 Rissola marginata (De Kay). Cusk eel 99 Gobiesox strumosus Cope. Ovster- fish; Clingfish ..... 99 Sphoeroides maculatus (Bloch and Schneider). Puffer 100 Lophius piscatorius Linnaeus. Goose- fish.. 100 Summary 101 Literature cited 101 INTRODUCTION Our knowledge concerning seasonal and geographic distributions of the planktonic young of most inshore marine fishes of the Atlantic coast is meager. This is especially true of certain common food fishes such as the weakfish, or gray sea trout, Cynoscion regalis, which provides the most valuable inshore fishery along the Middle Atlantic 79 80 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE seaboard. The importance of such information concerning our marine food fishes has been brought out by Bowman (1914), who asked Are the chief spawning places such that when the bulk of the larvae appear from the egg they find themselves in the immediate neighborhood of a locality suitable for development? To what extent do the prevailing physical conditions assist the passive eggs and helpless larvae in securing a suitable habitat for further development? It is of considerable import to the annual success of the American fisheries that there should be an intimate connection between the spawning grounds of a species and the localities suitable for growth. The present paper presents additional distributional and descriptive data on the young of a number of marine fishes regularly occurring in lower Chesapeake Bay. These data should help to increase our knowledge of the spawning season and spawning habitat of these fishes. 1 METHODS The area of Chesapeake Bay included in this study is bounded roughly by Cape Henry and Cape Charles on the east, Lynnhaven Roads on the south, Old Point Comfort on the west, and Back River Light on the north (fig. 1). Plankton collections were made at weekly or biweekly intervals at definite points within this area with a meter ringnet towed by powerboat. All except two of the collecting stations were permanently marked with navigation buoys and nearly all plankton was taken at definite localities over the entire period of collection — extending from May to October 1929, from April to December 1930, and during January and March 1931. The period of each tow was standardized at 15 minutes, the tow usually being with the tide and at as constant a rate of speed as conditions permitted. 2 Col- lections were usually taken from 10:30 a. m. to 2:00 p. m. Both surface and sub- surface tows were frequently made at each station. Subsurface tows were made from 10 to 20 feet below the surface of the water — the depth of water at no station exceeding 30 feet. PLANKTONIC FISHES Over 7,400 young fishes, representing 45 species, were taken in the plankton collec- tions in lower Chesapeake Bay during 1929-30. Of the total number, 7,380 fishes were identified and separated into 31 recognizable species, while 50 fishes were sepa- rated into 14 unknown species. The planktonic young of the sea trout, Cynoscion regalis, constituted over 50 percent of the total number of fish identified; followed in abundance by the young of the common anchovy, Anchoiriella mitchilli; the sea robin, Prionotus sp. ; and the blenny, Hypsoblennius hentz. 3 The numerical seasonal relation- ship of the various species of larval and postlarval fishes in the plankton given by the month and year is presented in table 1. The planktonic fishes, usually in larval or postlarval stages, were secured princi- > Acknowledgment is due the War Department for extended use oflaboratory space at Old Point Comfort, Vs., and to the many fish dealers and fishermen about Hampton Roads for valued information and assistance. Special mention is due Miss Louella E. Cable for the original drawings (figs. 2 to 9, 12 to 21, 24, and 25) in this report. ' The length of the net was approximately 4 meters (13 feet), the upper 1M meters of No. silk bolting cloth (3S meshes to the Inch), the lower 3 meters of No. 2 silk cloth (54 meshes to the inch), and a detachable cap of No. 12 silk cloth (150 meshes to the inch. • Numerically the young of A. mitchilli were far more abundant in the plankton than the young of C. regalis but, owing to the labor involved, only a small proportion of young mitchilli was removed from the plankton, while all the young of C. regalis as well as all other species were removed and identified. YOUNG OF SOME MARINE FISHES IN LOWER CHESAPEAKE BAY, VA. 81 pally from April 1 to November l. 4 The months from May to August yielded the most abundant catches, as well as the largest variety of species. "While certain species, such as the blcnny, Hypsoblennius hentz, and the common pipefish, Syrictes fuscus, were generally found widely distributed in the plankton from early spring until late fall, other species, such as the bluefish, Pomatomus saltatrix, occurred only once. : 1 W Figure 1.— Entrance to Chesapeake Bay, Va. Circled letters indicate plankton-collecting stations. Depth is given in feet. Subsurface collections generally yielded a larger number of fishes than surface tows. Certain species, such as Gobiesox strumosus, however, were taken proportionately more often in surface than in subsurface hauls. Many investigators have found that the surface layers contain few larval fish during the day. Clark (1914), in a study of the larval and postlarval fishes in the vicinity of Plymouth, England, found that night * The term "larval" as used in this paper refers to the growth stages of a fish from the time of hatching to the point where the fln rays appear differentiated and the young fish have considerable power of movement. The term "postlarval" refers to the growth stages following the development of the fin rays to a size where all traces of the larval fln fold have disappeared. The terms "larval" and "postlarval" fill a need for differentiating the more or less helpless young of many marine fishes from the juvenile youngwhich have more or less complete control of their movements. 82 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE hauls yielded a much larger percentage of young forms from the surface layers than did day hauls. Possibly the same condition might have occurred if night collections had been made in Chesapeake Bay (table 2). Table 1.- — Seasonal distribution of young fishes in the plankton, Chesapeake Bay, 1929-80. Nearly all fishes were taken in larval or early postlarval stages Species April 1930 May June 1929 July August September Octo- ber 1930 Novem- ber 1930 1929 1930 1929 1930 1929 1930 1929 1930 34 120 2 1 1 11 34 87 45 44 13 1 19 24 79 1 6 2 1 1 55 2 1 9 8 2,468 1 13 18 6 117 1,540 108 4 87 3 98 268 1 23 2 47 5 13 55 1 99 2 72 i 29 2 26 7 1 33 31 11 5 2 12 118 12 26 38 1 55 2 7 30 34 11 73 23 4 1 11 10 1 4S 1 12 26 9 17 1 17 2 4 32 63 2 53 1 5 4 82 138 75 223 3 2 12 26 8 1 3 1 2 2 15 10 22 3 26 1 4 1 4 1 1 2 5 8 3 3 58 14 3 2 1 Table 2. — The surface and subsurface distribution of planktonic fishes in Chesapeake Bay, expressed as the percentage of hauls in which the various species occurred [108 surface and 140 subsurface hauls were made from May to October, 1929, and 111 surface and 168 subsurface hauls from April to December, 1930, omitting June. No fishes were obtained in 24 surface hauls and 47 subsurface hauls made in January and March. 1931] Species Achirus fasciatus Anchoviella mitchilli Ancyclopsetta sp Astroscopus gultatus Bairdiella chrysura Brevoortia tyrannus Centroprutes slriatus Conger conger Cynoscion regalis Etropus sp Oobicsox strumosus Gobiosoma sp Hippocampus hudsonius- Hypsoblemius bentz Lophius piscatorius Percent of hauls, 1929 Sur- face Sub- sur- face 4(1 Percent of hauls, 1930 Sur- face Sub- sur- face Species Lophopsetta maculata Menidia menidia Menticirrhus americanus. Alicrogobius thalassinus. . Micropogon undulatus Paralichthys sp Peprilus alepidotus Pomatomus saltatrix Poronotus triacanthus Prionotus sp Pissola marginata Symphurus plagiusa Syngnalhus floridae _. Syricte* fuscus _ . . Sphoeroides maculatvs Tautoga onitis. Percent of hauls, 1929 Sur- face Sub- sur- face Percent of hauls, 1930 Sur- Sub- sur- face 22 4 B 2 12 2 1 10 4 2 1 11 8 YOUNG OF SOME MARINE FISHES IN LOWER CHESAPEAKE BAY, VA. 83 BREVOORTIA TYRANNUS (Latrobe). Menhaden; Fatback Distribution. — Young menhaden were taken four times during May 1929 and April 1930 near Old Point Comfort. The scarcity of young indicates that spawning probably occurs outside of the area of collection, although a limited number of men- haden eggs were taken during late summer. The occurrence of these young fish in early spring indicates that some spawning occurs during the winter months, as suggested by Hildebrand and Schroeder (1927). Description. — The young menhaden were from 20 to 24 mm. in length. The young of the species have been described by Kuntz and Radcliffe (1918). ANCHOVIELLA MITCHILLI (Cuvier and Valenciennes). Anchovy Distribution. — Young anchovies were taken from July 6 to Sept. 13, 1929, and from May 16 to Sept. 13, 1930. The larval and postlarval young were the most numerous of all species of fishes in the plankton. The separation of A. mitchilli from its relative, A. epsetus, is difficult if not impossible for young under 5 mm. Conse- quently, numbers of young A. epsetus may be represented in the collections of A. mitchilli. According to the relative abundance of eggs and adults of the two species in lower Chesapeake Bay, however, mitchilli far outnumbers epsetus. Description. — The size range of the young extended from 2.5 to 20.0 mm. The young of A. mitchilli have been described by Kuntz (1914) and the young of A. epsetus by Hildebrand and Cable (1930). CONGER CONGER (Linnaeus). Conger eel Distribution. — A leptocephalus, probably that of C. conger, was taken on Apr. 18, 1930, at Station J. Description. — The larva measured 100 mm. in length and possessed 150 + myomeres. LOPHOPSETTA MACULATA (Mitchell). Windowpane Distribution. — -The young of the windowpane flounder were taken during April and May 1930, at stations nearest the sea. The appearance of young only during April and May suggests an early spring spawning season in the region of Chesapeake Bay. Description. — The young ranged from 2 to 10 mm. in length. They are quite distinctive in appearance. Several stages of the young have been described by Bigelow and Welsh (1925). ETROPUS sp. Etrope Distribution. — Planktonic young of this small flatfish were taken principally in July 1929. Description. — This fish ranged in length from 2.5 to 13 mm. Although the correct generic identification of the young was possible through counts of fin rays of the larger specimens, doubt exists as to the specific identity owing to the probable presence of two species of the genus in the Chesapeake Bay area — namely, E. crossotus and E. microstomus. 84 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE PARALICHTHYS sp. Flounder Distribution. — A fish, perhaps referable to the summer flounder, P. dentatus, was taken on Nov. 28, 1930, at Station B. Description. — The fin rays of this fish, measuring 10 mm. in length, were differ- entiated, but the eye had not completed transition. Pigmentation consisted of three parallel rows of weak chromatophores lying along the dorsal, median, and ventral sides of the body. Each row contained eight distinct chromatophores. The specimen was too badly damaged to permit accurate fin-ray count, although the latter fell within the known range of P. dentatus. ANCYLOPSETTA sp. Flounder Distribution. — Two planktonic young taken on July 12, 1929, at Station B are probably referable to this genus of flatfishes. Description. — The young measured 5 and 6 mm. in length. The most charac- teristic features of the two fish are the pronounced elongation of the first two dorsal rays, the latter reaching nearly a quarter the length of the body, and the elongation of one of the ventral fins into a filament extending to the vent. The other ventral fin is not evident and apparently is undifferentiated. The pigmentation consists of a series of six chromatophores along the upper side of the body ; a single chromatophore along the median line on the posterior part of the body; a thin, black, continuous line along the ventral edge of the body; and many branching chromatophores on the ventral surface of the abdominal cavity. The fishes are symmetrical in shape. ACHIRUS FASCIA TUS (Lacepede.) American sole; Hog choker Distribution.— The planktonic young of this flatfish were taken during July 1929, August 1929-30, and September 1930. Most young were obtained during July 1929 and August 1929-30. This seasonal distribution indicates that the species spawns largely in midsummer. The greatest abundance of young was found about 1 mile off Little Creek, Virginia, near Station G. The latter estuary contains many adult and young fish during the summer months, and may constitute a spawning area. Description. — The length range of planktonic young extended from 1.5 to 4 mm. At 4 mm. the fin rays are clearly differentiated and identification is easily determined. The close resemblance of larval fish at 1.5 mm. to larger sizes permits ready identifica- tion. A strikingly heavy black pigmentation is characteristic of all young Achirus. The latter at 4 mm. in length still retain a symmetrical shape with an eye on each side of the head. The young have been described by Hildebrand and Cable (1938). SYMPHURUS PLAGIUSA (Linnaeus). Tonguefish Distribution. — Several larval tonguefish were secured at Station A on July 9, 1929. Description. — The fish ranged from 5 to 6 mm. in length. The young of this species has been described by Hildebrand and Cable (1930) and is readily identified. SYNGNATHUS FLORIDAE (Jordan and Gilbert). Pipefish Distribution.- — The young of this species were taken during June, August, and September 1929, and during May and July 1930, at many localities. YOUNG OF SOME MARINE FISHES IN LOWER CHESAPEAKE BAY, VA. 85 Description. — The young pipefish ranged in length from 14 to 48 mni. Identi- fication was based on body and tail ring counts. SYRICTES FUSCUS (Storer). Common pipefish Distribution. — The young of this species were taken from May 11 to Sept. 16, 1929, and from May 6 to Nov. 22, 1930. Description. — The length of the young ranged from 9 to 50 mm. Identification was based on body and tail ring counts. HIPPOCAMPUS HUDSONIUS De Kay. Seahorse Distribution.- — The young of the seahorse, Hippocampus hudsonius, were taken in plankton from June 6 to Sept. 13, 1929, and from July 7 to Sept, 12, 1930. Al- though spawning may occur within the bay, the young seahorses were generally taken in masses of floating sea vegetation and probably had drifted in from open sea. Description. — The young fish ranged from 6 to 33 mm., which included the distance from the tip of the snout (head flexed) to the end of the caudal fin. The young of the species has been described by Ryder (1881). MENIDIA MENIDIA (Linnaeus). Silverside Distribution. — The young of the silverside were taken in plankton during May 1929-30. Most young were secured at stations well within the bay. Hildebrand and Schroeder (1928) stated that the largest number of ripe adult M audio, occurred in April and May. Description. — The length range of the young extended from 5 to 9 mm. The various developmental stages have been described by Kuntz and Radclifl'e (1917) for the northern form, M. menidia notata, and by Hildebrand (1922) for the southern, or typical form, M. menidia. PEPRILUS ALEPIDOTUS (Linnaeus). Harvestfish Distribution. — The young of this important food fish were taken in the plankton during July and August, 1929-30, at all stations. The appearance of the young fish accompanied the incursion of large numbers of the coelenterates, Dactylometra and Cyanea. The long tentacles of these stinging "jellyfish" appear to act as a shelter and possibly as a food provider for the young harvestfish, for young fish were frequently observed hovering under the coelenterates. Description. — The lengths of the young fish ranged from 1.5 to 32 mm. The young harvestfish at 1.8 mm. in length has the larval yolksac absent and the larval fin fold entire. The larval gut is elongate, reaching about half the length of the. body. A lateral pigmentation occurs as a scattering of black chromatophores on the body (fig. 2). At 2.5 mm. the young fish possesses the lateral chromatophores in a more pro- nounced and characteristic pattern. One series of pigment cells follows the median line of the body from the pectoral fin to about half way the length of the body, while another, more regular series, lies along the lower side of the body dorsal to the gut. Scattering anastomosed chromatophores are found above the opercle and along the 407898—41 2 86 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE posterior sides of the abdominal cavity. The fin fold remains entire. A reduction in the length of the gut occurs at 2.5 mm. and what appears to be a secondary, or true vent is developed anterior to the gut. Several young at this length showed this peculiar structure, the exact nature of which has not been determined (fig. 3). Figure 2. — Peprilus ahpidotus. From a specimen 1.8 mm. long. The harvestfish at 3.5 mm. is more compressed, the gut has become greatly reduced and only one vent is evident. The location of the chromatophores becomes Figure 3.— Peprilus ahpidotus. From a specimen 2.5 mm. long. more elevated. The fin rays are slightly differentiated, although the fin fold remains entire (fig. 4). The young fish can be easily recognized at 7 mm. for the fin rays are fully differ- entiated. A further deepening of the body takes place and the chromatophores •••,■•.*■••* /*. **»• Figure 4. — Peprilus alepidotus. From a specimen 3.5 mm. long. become more scattered, enlarged, and anastomosed. The pigmentation is confined to the forward part of the body (fig. 5). The young fish becomes still further compressed at 9 mm. The pigmentation is darker and a considerable reduction in the size of each chromatophore occurs (fig. 6). The fish has assumed a characteristic adult shape at a length of 62 mm. The body has become strongly compressed, deep, and oval. The caudal fin has become forked, while the dorsal and anal fins are similar in shape and notably elevated ante- YOUNG OP SOME MAEINE FISHES IN LOWER CHESAPEAKE BAY, VA. 87 riorly. The body ckromatophores have disappeared and their place is taken by a thick peppering of black dots over the sides. The tips of the elevated dorsal and anal fins are heavily pigmented with black (fig. 7) . PORONOTUS TRIACANTHUS (Peck). Butterfish Distribution. — Young butterfish were taken abundantly in plankton from May 25 to Aug. 19, 1929, and from May 28 to Sept. 12, 1930. The young fish, similar to ,, rf* >"■ "*-***+,, ****** fWWM Figure 5. — Peprilus alepidotus. From a specimen 7 mm. long. Peprilus, were generally found in association with the coelenterates, Dactylometra and Cyanea. Butterfish 6 mm. long were secured from May 25 to July 23, indicating a late spring and early summer spawning season. The young were taken at all collecting points. Description. — The young butterfish ranged from 1.8 to 57 mm. in length. On the basis of an extensive series of butterfish from Chesapeake Bay, the writer believes Figure 6.—Peprilus akpidotus. From a specimen 9 mm. long. the fish represented in figures 62, 63, 64, and 65 (Kuntz and Radcliffe 1918) are not the young of the butterfish, Poronotus triacanthus, but most probably the young of a hake, Urophycis. Several fish obtained in Chesapeake Bay in 1929 are herein de- scribed as larval butterfish. Several figures of larger butterfish from Kuntz and Radcliffe (1918) are reproduced to show the gradual transformation to the adult shape. 88 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE The smallest, butterfish taken in the plankton measured 1.8 mm. A fish at this length has lost the yolksac but has the larval fin fold entire. The pectorals are faintly outlined, and a few rays of the caudal are discernable. A series of anastomosed Figure 7. — Peprilus alepidotus. From a specimen 62 mm. long. chromatophores lies along the dorsal region of the abdominal cavity. The ventral edge of the abdominal cavity and the body is sharply bordered with a solid narrow black line (fig. 8). The young possesses a deeper body at 3.7 mm. The fin fold is still entire, although the rays of the caudal are becoming differentiated. The same arrangement Figure 8. — Poroiwtus triacauthtis. From a specimen 1.8 mm. long. of chromatophores exists as in smaller fish, but an additional series of markings are now found along the ventral edge of the body from the gut to the caudal fin. Scat- tered chromatophores may appear at random along the sides, although never abun- dantly or in any definite arrangement as in young Peprilus (fig. 9). Succeeding stages of development have been described by Kuntz and Radcliffe. 6 (figs. 10 and 11). 5 Perlmutter (1939) has also recognized the erroneous descriptions by Kuntz and Radcliffe (1918) and has given figures of young butterfish 2.8 mm. and 3.6 mm. length. YOUNG OF SOME MARINE FISHES IN LOWER CHESAPEAKE BAY, VA. POMATOMUS SALTATRIX (Linnaeus). Bluefish 89 Distribution. — One plankton tow on July 24, 1930 at Station B yielded four specimens of young bluefish. Description. — The young ranged from 4 to 7 mm. in length. The bluefish at 4.3 mm. has the larval fin fold entire, although the dorsal, anal, and caudal fin rays Figure 9—Poronotux triacanthu*. From a specimen 3.7 mm. lone. Fir jure 10.— Poronotus triacanthus. From aspecimeu mm. long. From Kuntz nod KadclilTc (1918). Figure n.—Poronnlnstrinrnnlliiis. From aspecimen 15mm. long. From Kuntz and Radcliffe (1918) are slightly differentiated. The yolksac is absent. Three distinctive series of black dashes occur laterally on I he body; one along the dorsal ridge, another along the median line, and the other along the ventral edge. Other chromatophores occur above the abdominal cavity and on the top of the head. The teeth are well developed and appear quite diagnostic. The writer is unfamiliar with any other local fish in which the teeth are so strongly developed at such an early age (fig. 12). 407898—41 3 90 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE At 7.3 mm. the fish has lost its larval fin fold and the fin rays are clearly differ- entiated. The pigmentation remains essentially the same, but the lateral markings have become more pronounced and the dashes are now joined to form narrow black Figure 12.— Pomatomus saltatrii. From a specimen 4.3 mm. long. bands. The number of chromatophores on the head and on the abdominal cavity also increases (fig. 13). A later stage at 26 mm. no longer possesses the lateral bands but the entire body Figure 13. — Pomatomus saltatrii. From a specimen 7.3 mm. long. is covered with fine black dots. The caudal fin has become forked and the fins, particularly the spinous dorsal, have become further developed (fig. 14). At 72 mm. the young bluefish closely resemble the adult, except that the young fish has a silvery sheen in life and in preservation appears thickly peppered with fine FlGUEE 14.— Pomatomus saltatrii. From a specimen 26 mm. long. dots (fig. 15). Both figures 14 and 15 were furnished to the writer by Samuel F. Hildebrand and Louella E. Cable. The bluefish represented in these illustrations were taken off the coast of North Carolina, near Beaufort. CENTROPRISTES STRIATUS (Linnaeus). Sea bass; Blackfish Distribution. — Larval and early postlarval sea bass were secured during June 1929 and July 1929-30. Most young were taken in July 1929 at Station A. YOUNG OF SOME MARINE FISHES IN LOWER CHESAPEAKE BAY, VA. 91 Description. — The length range of the young extended from 2.5 to 9 mm. Young sea bass remain undescribed but comparison with a series of known sea bass from southern New England waters establishes the identity of the Chesapeake fish. Fin rays may be counted when the young reach 9 mm. in length. A distinctive type of pigmentation along the ventral edge of the body is characteristic of the larvae. BAIRDIELLA CHRYSURA (Lacepede). Sand perch Distribution. — The young of Bairdiella chrysura apparently are hatched largely outside of the area of collection, for only seven larval and postlarval fish were taken in the plankton. The young were secured from June 7 to July 1, 1929, principally at Stations A and B. Young fish ranging from 6 to 28 mm. were commonly taken by trawl on the muddy bottom in Little Creek in July 1930. Description. — The planktonic fish were from 2.5 to 5 mm. in length. Larval and postlarval sand perch are recognized by two vertical bands, the first behind the head Figure 15.—Pomatomus saltatrii. From a specimen 72 mm. long. and the second, less pronounced, about two-thirds the distance from the vent to the tip of the tail. The band nearest the tail is often weak and indistinct. Kuntz (1914) described the eggs and the 3 r oung of the species. MICROPOGON UNDULATUS (Linnaeus). Croaker Distribution. — Notwithstanding a great abundance of juvenile croakers within lower Chesapeake Bay throughout the year, a relatively small number of larval and postlarval fish were taken in the plankton. Young fish were taken on Sept, 13, 1929, and from July 29 to Oct. 17, 1930. Practically all catches were made at stations nearest the sea. An extended spawning period for croakers noted by Hildebrand and Cable (1930) in North Carolina evidently occurs also in the region of Chesapeake Bay. Description. — The young croakers ranged from 1.5 to 15 mm. in length. Larval croakers and larval gray sea trout appeared together in the plankton on several occasions in late July 1930. The two species closely resemble each other when newly hatched. The young croaker at 2 mm. in length, however, possesses a much deeper body than the sea trout at the same size. The croaker usually has a dark, crescent- shaped area above the abdominal cavity, while this marking is usually not as distinct in young sea trout. The pronounced chromatophore at the base of the anal fin, found on all young sea trout, is not especially pronounced on young croakers, although 92 FISHEEY BULLETIN OF THE FISH AND WILDLIFE SERVICE the latter do have a series of ventral chromatophores that greatly resemble compa- rable markings on the sea trout. The ventral chromatophores on the croaker are more numerous, however, and more evenly spaced than on the young sea trout. A percep- tible difference in the shape of the head and snout is also evident in the two species. Larval and young croakers have been described by Welsh and Breder (1923), Pearson (1029), and Hildebrand and Cable (1930). MENTICIRRHUS AMERICANUS (Linnaeus). Kingfish; Whiting Distribution. — The young of Menticirrhus americanus were secured abundantly from June 12 to Sept. 13, 1929, and from July 21 to Sept. 2, 1930. The largest collec- tions were made at Stations A, B, and C. Description. — The length-range of young extended from 1 .5 to 7 mm. Young fish, 3 to 7 mm. long, are characterized by profuse jet-black chromatophores scattered over the entire body. Under 3 mm. pigmentation is restricted to an area along the median line of the body. The jaws at all sizes are tipped with black. Fin-ray counts are possible at 5 mm. The young of M. americanus may be confused with the young of M. saxatilus, a closely related species. However, a comparison with a description of young saxatilus by Welsh and Breder (1923) and of americanus by Hildebrand arid Cable (1934) indicates that the fish from Chesapeake Bay most probably represent the young of americanus. CYNOSCION REGALIS (Bloch and Schneider). Gray sea trout; Weakfish; Sqneteague Distribution. — Over 4,000 young gray sea trout were taken in plankton hauls from May 25 to July 25, 1929. The majority of fish were secured at Stations A, B, C, and D during the latter half of June 1929. In 1930 planktonic sea trout were taken from May 21 to Aug. 1. The seasonal distribution of the young sea trout thus corresponds closely for 2 successive years (table 1 and fig. 23). The young of the gray sea trout were taken in 55 subsurface tows, with an average of 67 fish to a tow, and occurred in 13 surface tows, with an average of 25 fish to a tow. While more subsurface than surface tows were made, a comparison of simul- taneous surface and subsurface hauls at the same station indicates that in most instances the subsurface tow contained far more young fish than the surface tow. The planktonic sea trout decreased in abundance at those stations farther within the bay, compared with localities nearer to the sea. However, protected coves and creeks in the vicinity of Lynnhaven Roads yielded large quantities of young fish (8 mm. and over) just leaving the planktonic existence for a semidemersal life. The young fish were found on the bottom, where they were readily obtainable by trawl and seine. Various creeks from Lynnhaven Roads to the York River also had their complement of young sea trout during early summer, all young probably originating on spawning grounds off the entrance to the bay. Description. — The planktonic sea trout ranged from 1 .5 to 7 mm. in length. At a length of 1.8 mm. they are characterized by a very elongated slender body and by a large eye covering most of the side of the head (fig. 16). The larval fin fold is entire but the pectorals are differentiated, although indistinct. The greatest depth of the body is contained 4.0 to 4.5 times in the length to the end of the notochord. A YOUNG OF SOME MARINE FISHES IN LOWER CHESAPEAKE BAY, VA. 93 series of small black chromatophores is present along the ventral edge of the body extending from the vent to the tail. A chromatophore at midcaudal length, or at the primitive base of the anal, is consistently more pronounced than the rest. Several small chromatophores are found along the ventral edge of the abdomen. No other color markings are evident. The yolksac has been absorbed at 1.8 mm., although Welsh and Breder (1923) found a yolksac present on young of 2.2 mm. length taken in Delaware Bay. 6 The young sea trout at 3 mm. has the body depth proportionately increased. The only color marking is the series of chromatophores along the ventral edge of the Figure 16.— Cynoscion regalia. From a specimen 1.8 mm. long. body. All chromatophores become more pronounced, particularly the one at mid- caudal length. The fin fold remains entire. Minute teeth, usually evident at this length, help to distinguish the young sea trout from some related Sciaenidae such as the sand perch, Bairdiella chrysura, and the croaker, Micropogon undulaivs (fig. 17). Figure 17. — Cynoscion recalls. From a specimen 3 mm. long. The young sea trout at a length of 4.6 nun. has the caudal fin rays evident and shows a slight differentiation of the anal and dorsal fin rays. The fin fold remains entire. The greatest depth of the body is contained 2.7 to 3.0 times in the length to the end of the notochord. The series of ventral chromatophores has largely disap- peared, with the exception of the spot at the base of the anal which appears enlarged and anastomosed. This anal spot is significant for it apparently distinguishes the young of C. regalis from both C. nebulosus and C. nothus. Markings on the abdominal cavity are also pronounced. The mouth is more oblique and the teeth further developed (fig. 18). The young fish is quite readily identified at 8.2 mm. for the anal fin rays are usually distinct, while the soft dorsal rays are almost fully differentiated. The fin fold remains entire to the caudal fin. The greatest depth of the body is now contained about 2.8 times in the standard length. The snout is quite blunt, the lower jaw • All length measurements in this paper are referable to preserved specimens and denote total length. 94 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE projecting but little. The ckromatophore at the base of the anal is extremely pro- nounced, while the markings on the abdominal cavity are somewhat reduced in size and intensity (fig. 19). At 10.5 mm. the young have usually passed out of a planktonic existence and have adopted a semibottom habitat in quiet, muddy coves and creeks. Lateral chroma- Figure 18.— Cynoscion regalia. From a specimen 4.6 mm. long. tophores now prof usely appear, although the spot at the base of the anal still persists. The fin fold has nearly disappeared, while the caudal fin has changed to a symmetri- cally pointed shape (fig. 20). Figure 19.— Cynoscion regalis. From a specimen 8.2 mm. long. At 17 mm. in length the young are characterized by the presence of heavy lateral chromatophores arranged in four indistinct vertical bands or saddles. The chromato- phore at the base of the anal has now disappeared. The amount and intensity of Figure 20.— Cynoscion regalis. From a specimen 10.5 mm. long. pigmentation along the sides of the body seem to depend largely on the type of environ- ment in which the fish is found. Young taken on sandy and light bottom do not have as much pigmentation as fish secured on a muddy, or dark bottom. Tracy (1908), for instance, found several young gray sea trout in sunken canvas bags off Rhode Island YOUNG OF SOME MARINE FISHES IN LOWER CHESAPEAKE BAT, VA. 95 which at 6.5 and 12.5 mm. in length possessed more extensive pigmentation than fish of corresponding sizes taken in Chesapeake Bay. The greatest depth of the body is contained about 3.3 to 3.4 times in the standard length. In both larval and postlarval stages of the gray sea trout the body continues to increase in proportionate depth until at about 17 mm. it commences to decrease. In other words, the body becomes Figure 21.— Cynotelon regalia. From a specimen 17 mm. long. progressively stouter and shorter in proportion to length from the slender, newly hatched fish up to about 17 mm. in length, while after 17 mm. is reached the body tends to become more slender and elongate (fig. 21). Yoimg sea trout over 17 mm. in length are characterized largely by four distinct saddles on the body. Both Eigenmann (1901) and Breder and Welch (1922) have described various stages of the young sea trout (fig. 22). Growth. — Juvenile sea trout were found to grow rapidly during their first summer. Planktonic young ranging from 8 to 10 mm. soon settle to the bottom after entering Figcee 22.— Ci/noncion regalii. From a specimen 32 mm. long. From Welsh and Breder (1923). Chesapeake Bay. Brackish creeks and coves are favorite shelters for the young. Collections of fish at varying intervals during 1929-30 indicate that the young attain an average length of 16 to 20 cm. (6.3 to 7.8 in.) by the end of the first year. A growth diagram of young sea trout collected during their first summer and following spring is shown in figure 23. The length-range of young fish taken during the summer of 1930 is considerably less than for fish secured in 1929. This difference appears largely due to size selec- tion by the type of fishing gear employed. Seines were used exclusively during 1929 and allowed a greater escapement of the smaller fish than occurred in 1930, when fine-meshed trawls were employed. Similarly, year-old fish taken during the spring 96 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE of 1930 by commercial pound nets were larger than fish of the same approximate age taken during June 1930 by experimental trawl. Unfortunately, larger series of young collected at regular intervals at various localities and with all types of gear could not be obtained in order to show the selectivity of the gear and the effect of environment on the size distribution of the young fish. Notwithstanding limitations in the sampling of the juvenile sea trout population, it is believed that the average growth during the first year of life in lower Chesapeake Bay is reliably shown by figure 23. The young sea trout evidently have a length range of at least 10 cm. at the end of the first year of growth. Any clear-cut growth curve must involve large collections of young from diverse localities and by varied types of collecting gear. Eigenmann (1901) stated that juvenile sea trout (squeteague) doubled their length during July and August. This observation appears substantiated for Chesa- INS. CMS 10 25 \ j i r »• ' ; 8 20 :| 1 /' i > r e is , i l '■'; <*-P' Figure 26.— Astroscopus guttatus. Dorsal surface of head; from a specimen 235 mm. long. (1928). From Hildebrand and Schroeder phore pattern and the pigmentation does not extend so far back on the body as in Tautoga. Gobiesox also has a shorter gut and lacks the black-tipped upper jaw most characteristic of young Tautoga. SPHOEROIDES MACULATUS (Bloch and Schneider). Puffer Distribution. — The young of the puffer were taken from June 5 to Aug. 15, 1929, and from May 9 to Sept. 2, 1930. Description. — The lengths of the fish ranged from 1.5 to 4 mm. The early stages of the puffer have been described by Welsh and Breder (1922). LOPHIUS PISCATORIUS Linnaeus. Goosefish Distribution. — The young of this species were taken in small numbers during May 1930 at Stations A, B, and C. Since the adult fish are rarely taken within the bay, spawning probably occurs offshore. Hildebrand and Schroeder (1927) secured newly hatched young on June 10, 1916, in the lower bay. Description. — The young ranged from 3 to 5.5 mm. in length. Bigelow and Welsh (1925) have described the larvae of the species. YOUNG OP SOME MARINE FISHES IN LOWER CHESAPEAKE BAY, VA. 101 SUMMARY 1. The area of study is located at the mouth of Chesapeake Bay and is bounded roughly by Cape Charles and Cape Henry on the east, Lynnhaven Roads on the south. Old Point Comfort on the west, and Back River Light on the north. 2. A series of collecting stations was visited, usually weekly in summer and bi- weekly in winter, to determine the seasonal and geographic distribution and variation of the marine plankton. The present paper deals only with the young fishes taken in this plankton. 3. Forty-five species of fishes were recognized in the plankton. Thirty-one species were identified and 14 remain unidentified. Larval and postlarval stages of the gray sea trout, or weakfish, Cynoscion regalis; the bluefish, Pomatomus saltatrix; the harvestfish, Peprilus alepidotus; the butterfish, Poronotus triacanthus; and the stargazer, Astroscopus guttatus, are described and figured. 4. Collections of juvenile gray sea trout by seine and trawl indicate that this food fish attains an average total length of 16 to 20 cm. (6.3 to 7.8 in.) at the end of its first year of growth in lower Chesapeake Bay. 5. Brief distributional and descriptive records for the planktonic young of 31 species of marine fishes are given. LITERATURE CITED Bkjelow, Henry B., and Welsh, William W. 1925. Fishes of the Gulf of Maine. Bull. U. S. Bur. Fish., vol. 40, pt. 1, 1924 (1925): 567, illus. Clark, R. S. 1914. General report on the larval and post-larval telcosteans in Plymouth waters. Jour. Mar. Biol. Assoc. United Kingdom, 10 (n. s.), 1913-15: 327-394, illus. Eioenmann, Carl H. 1901. Investigations into the history of the young squeteague. Bull. U. S. Fish. Comm.. vol. 21, 1901 (1902): 45-51, illus. Hildebrand, Samuel F. 1922. Notes on habits and development of eggs and larvae of the silversides, Meniilia menidio, and Menidia beryllina. Bull. U. S. Bur. Fish., vol. 38, 1921-22 (1923): 113-120, illus. and Schroeder, William C. 1928. The fishes of Chesapeake Bay. Bull. U. S. Bur. Fish., vol. 43, pt. 1, 1927 (1928): 388, illus. and Cable, Louella E. 1938. Further notes on the development and life history of some teleosts at Beaufort, N. C. Bull. U. S. Bur. Fish., vol. 48, 1938 (1939): 505-642, illus. 1930. Development and life history of fourteen teleostean fishes at Beaufort, N. C. Bull. U. S. Bur. Fish., vol. 46, 1930 (1931): 383-488, illus. 1934. Reproduction and development of Whitings or Kingfishes, Drums, Spot, Croaker, and Weakfishes or Sea Trouts, Family Sciaenidae, of the Atlantic coast of the United States. Bull. U. S. Bur. Fish., vol. 48, 1934: 41-117, illus. Kuntz, Albert. 1914. The embryology and larval development of Bairdiella chrysura and Anchovia milchelli. Bull. U. S. Bur. Fish., vol. 33, 1913 (1915): 3-19, illus. 1916. Notes on the embryology and larval development of five species of teleostean fishes. Bull. U. S. Bur. Fish., vol. 34, 1914 (1916): 409-429, illus. and Radcliffe, Lewis. 1918. Notes on the embryology and larval development of twelve teleostean fishes. Bull. U. S. Bur. Fish., vol. 35, 1915-16 (1918): 87-134, illus. 102 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Pearson, John C. 1929. Natural history and conservation of the redfish and other commercial Sciaenids on the Texas coast." Bull. U. S. Bur. Fish., vol. 44, 1928 (1929): 129-214, illus. 1932. Winter trawl fisher}' off the Virginia and North Carolina coasts. Inv. Rept. No. 10, U. S. Bur. Fish., 1932: 31, illus. Perlmutter, Alfred. 1939. A biological survey of the salt waters of Long Island, 1938. Pt. II, N. Y. Cons. Dept., Sec. I: 11-71, illus. Ryder, John A. 1882. A contribution to the development and morphology of the lophobranchiates (Hippo- campus antiquorum), the sea horse. Bull. U. S. Fish Comm., vol. 1, 1881 (1882): 191-199, illus. Tract, Henry C. 1908. The fishes of Rhode Island. VI. A description of two young specimens of squeteague (Cynoscion regalis), with notes on the rate of their growth. 38th Ann. Rept. Comm. of Inland Fish., State of Rhode Island and Providence Plantations. 1908: 85-91, illus. Welsh, William W., and Breder, C. M., Jr. 1922. A contribution to the life history of the puffer, Spheroides maculalus (Schneider). Zoo- logica, vol. 2, No. 12, 1922: 261-276, illus. 1923. Contributions to life histories of the Sciaenidae of the Eastern United States Coast. Bull. U. S. Bur. Fish., vol. 39, 1923-24 (1924): 141-201, illus. UNITED STATES DEPARTMENT OF THE INTERIOR Harold L. I ekes, Secretary FISH AND WILDLIFE SERVICE Ira N. Gabrielson, Director Fishery Bulletin 37 THE SALMON RUNS OF THE COLUMBIA RIVER IN 1938 By WILLIS H. RICH From FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Volume 50 UNITED STATES GOVERNMENT PRINTING OFFICE WASHINGTON : 1942 For Bale by the Superintendent of Document*. Washington. D. C. ............ Price 10 cents ABSTRACT EXCEPTIONAL DATA are available for the study of the salmon runs of the Columbia River in 1938. Detailed figures on catch were supplied by Oregon and Washington in such form that they could readily be combined with the counts at Bonneville Dam to provide a basis for estimating the escapement. Tables show the catch of each species for each week in each of six zones, and the counts at Bonneville and Rock Island dams. The general course of the run of each species is shown. The numbers of fish bound for the spawning grounds above Rock Island Dam are estimated as follows: Chinook salmon entering Colum- bia River before May 1, 4 percent; during May, 6 percent; June and July, 15 percent; and August to December, 1 percent. Bltieback salmon entering the river during the above periods, 40 percent. Steelhead trout entering the river during June to September, 1 percent; during the rest of the year, 10 percent. Fishing intensities are shown by escapement to catch ratios. Percentages of chinook salmon escapement are less than 15 during May; 17 during June and July; and 33 during the remainder of the year. The June and July runs are now greatly depleted, and an important part of these runs spawns above Rock Island Dam. The blueback salmon escapement is about 20 percent, and of steelhead trout about 33 percent. Weekly and seasonal closed periods are shown to be almost entirely ineffective for increasing the spawning escapement. Exploitation is further increased by the intensive troll fishery conducted from Monterey Bay to southeastern Alaska. Chinook salmon are also subjected to a sport fishery of considerable importance. Main runs of salmon to the Columbia River are practically unprotected and are fished with destructive intensity. II THE SALMON RUNS OF THE COLUMBIA RIVER IN 1938 1 By Willis H. Rich, Professor of Biology, Stanford University and Director of Research, Fish Com- mission of Oregon; in cooperation with the Division of Fishery Biology, Fish and Wildlife Service Page Introduction 103 The Columbia River salmon fishery 104 Data for the runs of 1938 107 Modified tables 113 Nature of the analysis of runs 116 Chinook salmon 119 History of the run of 1938 119 Rate of travel 124 The June-July run 125 CONTENTS Page Chinook salmon — Continued. Intensity of fishing in general 130 Percentage of grilse 132 Blueback salmon 134 Steelhead trout 138 Silver and chum salmon 143 Summary 145 Literature cited 147 INTRODUCTION With the announcement of plans for the construction of the Grand Coulee Dam on the Columbia River in eastern Washington, questions were raised as to the effect that this development would have on the salmon runs and as to the possible means for preserving those salmon populations that had formerly reproduced in the area above the site of the dam. Funds were provided by the United States Bureau of Reclama- tion to the Washington State Department of Fisheries for the purpose of making a preliminary study of possible means for preserving the runs. A report (Washington State Department of Fisheries 1938 2 ) was presented in January 1938, in which the chief recommendation was for an extensive system of artificial propagation. Later the Bureau of Reclamation appointed a board of consultants to review the proposed plan and to make recommendations. In their report (Calkins, Durand, and Rich 1939 3 ) these consultants recommended, substantially, the plan proposed by the Washington Department of Fisheries. In the preparation of this report the writer made an analysis of the available data on the salmon runs of 1938 for the particular purpose of determining the relative importance of those fractions of the runs that would be affected by the construction of the Grand Coulee Dam. Various other facts bearing upon the state of the Columbia River salmon resources and the problems of their conservation were developed during the course of this analysis and it has seemed desirable to amplify the part of the i Contribution No. 7. Department of Research, Fish Commission of Oregon. Report of the preliminary investigations into the possible methods of preserving the Columbia River salmon and steelhead at the Grand Coulee Dam. 121pp. V. S. Bureau of Reclamation, Washington. (Processed.) Report of the board of consultants on the fish problems of the upper Columbia River. 83 pp. U. S. Bureau of Reclamation, Denver, Colo. (Processed.) 103 104 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE report that treats of the 1938 run and to present it as a separate publication. For this purpose the data presented in the original report of the board of consultants have been supplemented by data that have become available since the original report was pre- pared. At that time no catch data were available later than the close of the "spring" fishing season on August 25. In this revision the catch data for the "fall" season also have been included. Various omissions and minor changes have been made, and some additional analysis is given. Acknowledgment is due the Bureau of Reclamation and the writer's associates on the board of consultants for permission to use here the material of the original report. Acknowledgment also is due the Washington Department of Fisheries, the Fish Com- mission of Oregon, the United States Army Engineers, and the Bureau of Reclamation for many data used in the original report and in this revision. THE COLUMBIA RIVER SALMON FISHERY Five species of salmon are taken in the commercial fishery on the Columbia River. These are (1) chinook salmon (Oncorhynchus tschawytscha) , (2) silver salmon (0. kisvtch), (3) blueback salmon (0. nerka), (4) chum salmon (0. keta), and (5) steel- head trout (Salmo gnirdnerii) . Fishing is permitted throughout the year except during March and April, and during the period from August 25 to September 10. The open season from May 1 to August 25 is spoken of as the spring season, and that from September 10 to March 1 as the fall season. Comparatively few fish are taken during December, January, and February, however, so that the fall season is practically limited to the period from September 10 to about the end of November. In addition to these seasonal closed periods there is a weekly closed period extending from 6 o'clock Saturday evening until 6 o'clock Sunday evening, effective during the spring open season. Because the estimate of the intensity of the fishery is based on the ratio of the commercial catch to the fish passing Bonneville Dam, it is important to consider the relative extent of spawning which, for each species, takes place above and below this point. Obviously, if a large proportion of the fish of any one species, population, or group of populations spawns below Bonneville Dam, estimates of relative spawning escapement based upon the number of fish passing Bonneville will be in error. Practically all the bluebacks spawn above Bonneville. As is well known, their habit is to spawn only in lakes or the tributaries of lakes in which the young remain for 1 or more years before making the seaward migration, and no lakes typical of those in which bluebacks spawn are to be found in the tributaries of the lower Columbia. The chinooks spawn in nearly all the accessible tributaries of the river, both above and below Bonneville; a fact certain to lead to some error. With one exception, however, this error is probably negligible during the main part of the run because it is chiefly the late fall fish that spawn in tbe lower tributaries. The exception is the considerable run of chinooks that ascends the Willamette River in April and early May. There are, unfortunately, no reliable estimates of the extent of this run, but it forms the basis for an extensive sport fishery in the Willamette River, especially just below the falls at Oregon City. No commercial fishing is permitted in tbe Willamette River itself and the peak of tbe rim is ordinarily past Oregon City by the opening of the season on May 1. Although some of these Willamette River chinooks are un- doubtedly taken in the commercial fishery in the Columbia below the mouth of the Willamette, it does not seem likely that these constitute a large percentage of the total SALMON EUNS OF THE COLUMBIA RIVER IN 1938 105 commercial catch. It is believed, therefore, that error in the estimates of fishing intensity of chinooks, due to spawning in the tributaries that enter the Columbia below Bonneville Dam, is relatively small, even during the first few weeks of the spring open season. After about the middle of May it seems reasonably certain that there is very little error due to this cause until at least the first of August, at which time some fish that will eventually spawn in the smaller tributaries below Bonneville Dam begin to enter the river. In none of these lower tributaries is there a large rim of spawning fish while the count of fish passing Bonneville is at its peak during August and September. These facts indicate clearly that, even during these months, the error in the estimate of fishing intensity based on a comparison of catch with the count at Bonneville will not be serious. As the season advances, however, progressively larger percentages of the fish entering the river are destined to spawn in the lower tributaries. Although the total number of fall fish spawning below Bonneville Dam is probably not large com- pared with the number spawning above the dam, the error will tend to increase, and great dependence cannot be placed on the results of the study of the late fall fish. Steelhead trout spawn generally throughout the accessible tributaries, but ap- parently are more abundant in the upper than in the lower streams. In the case of silver and chum salmon, a very large proportion of the spawning occurs in the tribu- taries below Bonneville Dam, so that the ratio between the count at the dam and the catch gives no reliable indication of the intensity of the fishery. This report deals primarily with the salmon runs of 1938 and it is to be hoped that similar studies, either by this writer or by others, will be made of future runs for which similar data will be available. As a part of the "frame of reference" into which are placed these studies of the runs of individual years, however, it is important to pre- sent something of the earlier history of these runs. This has been done in some detail elsewhere (Craig 1938 4 ; Oregon State Planning Board 1938 5 ; Craig and Hacker 1940; and Rich 1940b) and there is presented here only a graph showing the average annual catch of each species for each 5-year period. The data for this graph have been taken from Craig (1938), and recent numbers of the Pacific Fisherman Year Book. Previous to 1888 there was no segregation of the salmon catch by species, but there can be no doubt that chinooks formed the bulk of the catch. For the first 2 decades during which the pack was segregated the chinooks formed about 80 percent of the total, and it has been assumed that approximately the same percentage existed prior to 1888. No attempt has been mado to estimate the catch of the other species previous to the period 1890-94. The catch in pounds has been estimated from the figures for the canned and mild-cured packs, which include a large part of the total. Further details may be found in the several references given. Figure 1 shows the rapid growth of the industry during the first 2 decades after its inception, a period of 35 or 40 years in which the catch of chinook salmon fluctuated from about 20,000,000 to 30,000,000 pounds and a final period of some 20 years in which there has been a constant decline. In all probability this decline is an indica- tion of true depletion; that is, a reduction in productivity below the point that can be maintained over a long period of time. The picture is complicated by the existence of an extensive oceanic fishery extending from Monterey Bay to southeastern Alaska, which draws heavily upon the supply of Columbia River chinooks (Rich 1941). ' Memorandum regarding fishing in the Columbia River above and below Bonneville Dam. 16 pp., U. S. Bureau of Fish- eries, Washington. (Processed.) « Commercial fishing operations on the Columbia River. 73 pp. Oregon State Planning Board, Portland, Oreg. (Processed.) 106 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE The catch within the river does not, therefore, represent the entire productivity of the runs of this species, but with available data it is not possible to determine with much accuracy what this total productivity actually is. The constant decline of the last 20 years, however, taken in connection with data presented in this report, certainly warrants the conclusion that the chinook runs are seriously depleted. 6 We shall show below that the present exploitation of these depleted ruus is being conducted with an intensity so great that it can only lead to disaster in the not far distant future unless the present trends can be altered. The blueback salmon catch for both of the first 2 periods shown in figure 1 is ap- proximately twice that of the succeeding periods, and there is some reason to think that the abundance of blueback salmon previous to 1890 was at least the equal of that T T CHINOOK BLUEBACK STEEL HE AD SILVER CHUM J L J I I L '80 '85 '90 '95 1900 '05 '10 '15 '20 '25 30 35 40 Figure 1.— Average annual catch by 5-year periods of Chinook salmon, 1866-1938; and of blueback, silver, and chum salmon; and steelhead trout, 1891-1938. which existed during the decade of the 90's. Since 1900, however, there has been little change — the trend is almost horizontal. These facts imply that this species originally was fairly abundant in the Columbia River, but that this early abundance was sharply reduced about 1900, and since that time there has been comparatively little change. This species almost universally spawns in or above lakes and it seems quite possible that the damming of lakes for use as reservoirs without providing adequate fishways, and the unrestricted use of unscreened irrigation ditches, were chiefly responsible for the depletion. In figure 1 considerable fluctuation is shown in the estimated catch of steelbead trout, especially in the early years of the record, but there is little evidence of a marked • Since this report was in page proof an additional study of these data has been made using the methods of the control chart as developed by Shewhart, Demlng, and others, for the control of quality in manufactured products. The results show conclusively that the productivity of the chinook fishery since 1925 has been at a distinctly lower level than was maintained during the period 1876 to 1920. These will be published elsewhere. SALMON RUNS OF THE COLUMBIA KIVER IN 1938 107 trend. It suggests, however, that the slightly reduced averages for the past two 5-year periods may signify some real reduction in abundance. The general trend for both silver and chum salmon is distinctly upward (fig. 1) despite rather wide fluctuations. This doubtless reflects an increased usage of these 2 less desirable species that has come with the reduced abundance of the other species, especially the chinook. DATA FOR THE RUNS OF 1938 In this study of the 1938 salmon runs to the Columbia River, data have been available for the first time in the history of the fishery that have made it possible to evaluate the intensity of the fishery as a whole, the relative intensity at different times and in different parts of the river, and the proportion of the total that is formed by the run to the upper Columbia River (Clarks Fork). These data include the following series: (1) Daily commercial catch in pounds and by species in each of 6 districts corresponding to the 6 counties of the State of Washington that form the northern shore of the Columbia; (2) daily counts, by species, of the salmon passing Bonneville Dam beginning with May 7, and estimates for the period from February 15 to May 6; and (3) daily counts, by species, of the salmon passing Rock Island Dam across the upper Columbia near Wenatchee, Wash., about 100 miles below the site of the Grand Coulee Dam. The latter have been available since the season of 1933. The importance of the data on the Bonneville count and the total daily catches to the proper development of a sound program for the conservation of the salmon of the Columbia River should be emphasized. Without them an intelligent con- sideration of the problems raised by the Grand Coulee Dam would have been im- possible, and they will be of equal importance in the study of any other problems dealing with the maintenance of this valuable resource. For the previous three seasons the Washington Department of Fisheries had collected records of the daily deliveries of each species of salmon in each of the counties of the State bordering on the Columbia River. The Fish Commission of Oregon also had collected data on the daily de- liveries of salmon, but not until 1938 were these presented in such form as to make it possible to combine them with the data from Washington so as to give a record of the total daily deliveries by species and by locality. For no other year are such data available, although figures for 1939 will be in suitable form for study when they are available. Now that a uniform system for presenting the catch data has been started by the two States, it probably will be continued so that in the future data will be available showing the total daily deliveries in each of the six districts. Of equal importance has been the record of counts of fish passing the dams at Bonneville and Rock Island. Since 1933 there have been counts, more or less com- plete, at Rock Island, but the Bonneville Dam was not finally closed to the passage of fish previous to 1938, so that this year marks the beginning of the count at this point. The tremendous value in the conservation program of the count of salmon passing over the Bonneville Dam cannot well be overstated. This count should, by all means, be made a permanent feature and should be in the hands of competent men familiar with the fish and with the techniques of fishery research, and having a primary interest in the fishery problems upon which these data will bear. In presenting these data it has been found expedient to sum them for the smallest practical time interval. The unit of 1 week was selected as the shortest period that would avoid insignificant fluctuations, particularly the disturbing effect of the Sunday closed period. For special purposes the data have also been arranged relative to 108 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE longer time intervals, but these have been selected carefully on the basis of facts apparent from the tabulations made on a weekly interval. The use of relatively short time intervals has been important because of the considerable fluctuation in the commercial value of the salmon, particularly chinooks, during the season. The spring fish, entering the river during the period from April to the early part of August are much more valuable than those running later in the season. Furthermore, the magnitude of the run varies greatly from week to week and some portions of the run are far more seriously depleted than others. The intensity of fishing also varies, and the closed seasons tend to favor certain portions of the run and leave others practically unprotected from intensive exploitation. The commercial and biological importance of the various portions of the run of each species must, therefore, be determined independently, and to do this a relatively short time interval is essential. Because the fishing season begins May 1, the first week in May has been taken as the point of departure, and the weekly intervals, both before and after, are arranged to conform to this. Table 1. — Catch of chinook salmon in the Columbia River, 198S Week ending Outside ' Zone 1 Zone 2 Zone 3 Zone 4 Zone 5 Zone 6 Pounds 2.393 38, 292 8.193 11, 123 11, 309 Pounds Pounds Pounds Pounds Pounds Pounds 212, 139 131, 269 69, 042 36, 655 35, 005 53. 963 81,434 110,052 127, 078 189, 276 154, 680 187, 621 309, 210 658. 106 1, 000, 675 1.121,367 » 960, 365 390, 998 116,015 71,280 43, 170 19, 783 35, 554 48. 300 86. 608 66, 401 133, 845 108, 092 84, 095 123,905 127, 847 482, 395 617. WS 288,497 237.612 105,441 17,411 7,619 539 736 3,392 11,544 11,230 26, 334 34, 388 30. 224 29, 035 16, 158 55, 821 84,876 65,964 38, 294 29, 720 5,097 1,805 20 51 1,683 2,931 4,340 4, 572 5,202 7,414 11,648 7,733 15, ISO 7,755 19.611 41,327 28,531 4,662 1,476 129 30 947 1,113 2,306 4,605 7, 358 5,988 10,111 6,899 9,363 17, 030 29,544 14. 447 May 14 --- 226 38,464 19, 509 May 28 7,395 1,388 9,062 5, 872 6,398 12. 494 5,984 6,580 8,577 21, 454 40. 596 12, S69 9,911 26, 875 18, 896 7, 573 12. 648 2,247 2.037 2,794 502 1.295 4.346 6.955 4,495 4,248 July 2 1,989 July 9 1,217 1.118 July 23 2.872 July 30 - 2.070 6,157 15, 373 40. 676 Aug. 27 37, 532 Sept. 17 188, 933 146,414 30. 414 8.700 14, 481 9,356 2,635 1,477 547 83 65 142, 308 164, 363 30. 701 13. 987 15. 990 9,273 6, 592 2.952 812 167 141 18 96 14 83, 724 130, 829 8,264 4.776 6.303 4,446 2,852 1,305 701 177 30 » 94, 942 39. 201 14, 075 5,385 1,371 1,929 671 1,035 151 142 45. 331 52, 156 » 772, 785 Sept. 24 398, 660 117.743 34 69, 656 23.199 Oct 22 - 28 63 6.022 Oet. 29 -- 5,507 6 1,358 26 11 i Outside may Include some fish caught by troll inside the river and along the coast from Neah Bay to Coos Bay. > The season on the river closed on August 25. ' The fall season opened at noon on September 10 and on that, day 450 pounds were delivered in Zone 4 and 93,83. pounds in Zono 6. Since these catches represented a fishing period of only one-half day, they have been added to the catches of the following week. SALMON RUNS OF THE COLUMBIA RIVER IN 1938 109 Table 2 — -Catch of blueback salmon in the Colum bia River, 1938 Week ending Zone 1 Zone 2 Zone 3 Zone 4 Zone 5 Zone 6 Week ending Zone 1 Zone 2 Zone 3 Zone 4 Zone 5 Zone 6 May 21 Pounds 5 Pounds Pounds Pounds Pounds Pounds July 16 July 23 July 30 Aug. 6 Aug. 13 Pounds 2,310 618 65 7 Pounds 0,771 380 72 8 Pounds 2,408 719 71 Pounds 1,919 177 12 Pounds 4,571 919 71 6 Pounds 29,322 May 28..- 17, 749 4 47 2, 282 20,804 18,326 8,301 3.960 16 8,138 63. 156 59,368 37, 708 4 203 2,872 15,312 27,368 1, 125 905 7.342 13, 534 11,624 468 6,393 7,325 6,430 834 9,040 10, 579 8,117 411 Aug. 20. _. 48 57 July 2 Aug. 271... 15 July 9 ' Season closed on August 25. Table 3. — Catch of steelhead trout in the Columbia River, 1938 Week ending Outside Zone 1 Zone 2 Zone 3 Zone 4 Zone 5 Zone 6 Pounds Pounds 1,833 1,166 722 447 566 S'js 3,448 17,724 20.449 59.666 50, 226 36.944 37.928 22 .679 45,215 32.885 '21,852 Pounds 5,347 4,366 2,465 1,703 627 890 5,070 20, 566 38, 342 94,512 76, 705 56, 598 58. 537 30,246 94,016 72. 520 40, 807 Pounds 1.174 851 243 141 28 39 171 1,247 2.929 10,610 23,060 24,920 11,797 7,362 11.274 15.284 11.1117 Pounds 169 75 41 56 5 7 113 371 1,008 2,142 3,224 4,686 2.225 2. 600 2,908 3.152 3,524 Pounds 449 184 66 74 Pounds 1.298 3.253 May 21 2.212 May 28 1,203 737 10 57 327 503 1,701 3,821 3,409 2.210 3,084 2,204 3.777 3,293 503 255 687 358 38 22 46 72 119 50 100 66 19 12 416 July 2 451 July9 829 July 16 1.029 July 23 6, 204 July 30 3. 524 3,857 Aug. 13 15, 784 Aug. 20 20,863 Aug. 27 Sept. 3 11,568 Sept. 10 Sept. 17 1,826 11,583 6,647 2.258 1. 337 1,730 1,149 1.130 2,240 2,810 2.127 930 2, 757 1,0114 774 427 7.317 30.153 14,878 5.025 3.090 2. 585 2,228 2, 5(13 5.206 5, 497 8, 459 7. ISO 18, 101 8.800 6, 955 9,660 4.332 10, 036 3,741 2,021 1.174 9S2 730 us catches in the Columbia River, 19SS Month Chi- nook Steel- head Blue- back Silver Chum Month Chi- nook Steel- head Blue- back Silver Chum Pounds 59 18, 714 1,170 1,503 8,057 47, 099 Pounds Pounds Pounds Pounds October Pounds 9,753 Pounds 1,186 571 423 Pounds Pounds 4,420 3,164 21 Pounds 1,060 2,725 178 93 311 2,074 5,551 2,219 722 27 166 5 152 218 29 9 Total 86, 355 7,561 2,968 8,146 6,649 Tables 1 to 5 give the aggregate Washington and Oregon catches for 1938, by spe- cies, weeks, and zones. These figures include only those catches that were reported by locality and date. There is a relatively small portion of the total catch that is reported without these important data and these have been excluded from this analysis, although for completeness they are given in table 6. The catch of chinook and silver salmon made in the ocean outside the mouth of the river by troll fishermen was not given in the original report by the board of consultants, but is included here. Occasionally deliveries are reported during the spring season as of Sunday. Since the period from 6 p. m. Saturday to 6 p. m. Sunday is closed to fishing each week during the spring season, such catches have been added to those of the preceding week. Catches made on Saturday are not infrequently held over and delivered on Sunday, and it rarely happens that catches are made after 6 p.m. on Sunday and delivered that same evening. The zones correspond to the Washington counties bordering the river, beginning at the mouth. Zone 1 is that part of the river that is bounded on the north by Pacific County, Zone 2 by Wahkiakum County, Zone 3 by Cowlitz County, Zone 4 by Clark County, Zone 5 by Skamania County, and Zone 6 by Klickitat County. The catch in Zone 5 has, on the advice of both the Washington and Oregon officials, been referred wholly to the area below Bonneville Dam. This zone extends above Bonneville for some distance, but for a part of tbis distance the river is closed to all fishing and the catch in the remaining portion is so small as to be negligible, either when omitted from the record of the catch above or added to the record of that below Bonneville. In this analysis we have necessarily omitted consideration of three elements in the catch which are recognized as important but which cannot, with the data at hand, be evaluated. These are: (1) The catch in the ocean by the troll fishery; (2) the hook- and-line catch by sport fishermen; and (3) the catch made by Indians for their own use, especially at Celilo Falls. The troll fishery is very important, and from southeastern Alaska to the mouth of the Columbia it draws largely upon the supply of Columbia River chinooks — as demon- strated by tagging experiments (Pritchard 1934, Fisheries Service Bulletin, Jan. 3, SALMON RUNS OF THE COLUMBIA RIVER IN 1938 111 1928). Fairly good data are available as to the aggregate troll catch of chinooks and silvers in Alaska, Oregon, and Washington. The percentage of Columbia River fish in this catch, however, undoubtedly varies greatly during the season. There are no satisfactory data on this latter point. Even though we knew the proportions of Columbia River fish in the catch at different times and in different localities, it would be impossible to allocate these to the seasonal runs of the Columbia and thus, eventually, to determine the element in the troll catch derived from the runs to the Columbia River above Rock Island Dam. Likewise, we have no data on the catch of the sport fishery or on that part of the Indian catch that is not sold. All of these elements increase to some unknown extent the economic importance of the salmon runs with which we are here concerned. Table 7. — Estimates and counts of fish passiiig Bonneville, 1938 [The figures up to and including May 7 are estimates based on partial counts only. Differences between the figures given hero and those in the report by Calkins, Durand, and Rich are due to the fact that this table includes the final figures as given by the Army Engineers, in which minor corrections were made of the figures submitted weekly.] Week ending Chi- nook Grilse i Steel- head Blue- back Silver Chum Week ending Chi- nook Grilse ' Steel- bead Blue- back Silver Chum Feb 19 55 158 204 980 1, 207 84 981 7,319 1,927 639 320 138 3, 217 1,022 1,644 164 032 052 520 641 800 4,061 7, 161 6,667 Auc. Aug. 13.-.. Aug. 20 Aug. 27.... 1 1. 3 ... Sept. in . ' 17 ' 24.... Ocl i Oct. 8 Oct. 15 Oct. 22 Oct. 29 Nov. Nov. 1 Nov. 19.... Nov. 26 Deo. 8 1 SI 10 17 Dec. 24 ... Dec. 31 Total- 1.327 4,103 10.112 SO, 693 12,268 2,057 489 161 234 208 47 29 9 5 21 2 2 329 769 1.010 2.166 v. 452 1,581 406 244 99 34 17 40 10 5 1 1 5 4.880 6, 457 6,908 1 7. ' 89 15, ••' 1 13,744 3,935 1,204 857 604 230 253 152 90 60 58 30 43 18 1 13 1,125 621 209 156 78 71 10 1 3 1 1 2 4 68 84 14 339 402 484 1,545 3,359 12,938 5, 097 3, 827 205 1,981 2,932 2, 230 1,240 884 1,855 1,534 1,753 1,357 842 871 53 710 515 334 164 102 204 337 430 115 1,933 239 56 94 240 212 138 18 9 4 1 7 n Mar. 19 Apr. 2 Apr. 9 Apr. 16 Apr. 23 Apr. 30 May 7 May 14 May 21.... May 28 2 68 179 131 572 318 24 153 1,358 5.719 15,441 16, 491 21, 673 7. S35 2,770 945 174 236 225 202 45 46 July 2 July 9 July 16 July 23 July 30.... 13 1 277, 665 36, 757 120,985 75, 040 15,185 2,136 1 Grilse, locally designated as "jack" salmon, are precocious males. These are included in the preceding column headed "Chinooks," the figures in which are, therefore, the totals for this species. In table 7 are given the counts and estimates of the Dumber of salmon and steelhead passing Bonneville Dam during 1938. Actual counting did not begin until May 7, but estimates could be made from partial counts — the so-called "spot" counts — covering the period from the middle of February to and including May 6. These partial counts were made by observers stationed for portions of the day at the several fish ladders. The records consisted of (1) the length of time during which the obser- vations were continued, and (2) the number of fish of each species observed. This is essentially a sampling method, and it is known that the fish do not run uniformly during the entire 2-1 hours, or even during the daylight hours. A fairly good estimate can be made from such records, however, if the hours during which the fish run are determined with care, and if the periods during which the counts are made are suitably distributed. The method adopted here for estimating the total number for the day from the partial counts is to multiply by 12 the average hourly count as determined from the records. This is the method recommended and used by Fred Morton, who 112 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE was actively in charge of the count. This method assumes that the fish are passing over the ladders for 12 hours per day at the same average rate as observed during the period of the count and has been applied to each ladder separately and the sum is the estimated total for the day. For periods during which no count was made a linear interpolation between the preceding and the following days' estimated counts was used. Although not comparable in accuracy to the actual count, these estimates appear to give a reasonable basis for further calculations. A chief source of error in these counts and estimates is undoubtedly the identifi- cation of species as the fish were passing up the ladders. After May 7, when the actual count began, the fish were forced to pass through a small opening in a weir placed across each fish ladder and over a submerged platform painted white. Identification of species under these conditions can be made with some accuracy by careful observers and, in general, reasonable confidence can be placed in the identifications so made. Those made under less favorable conditions must, necessarily, be accepted as the best available. Circumstances may arise in which a particular misidentification is espe- cially likely to occur, in which case it may be recognized and steps taken either to improve the identification or to determine its influence and allow for it in the estimates of the number of fish of the species confused. It is apparent that one such particular case of misidentification might easily arise during the time when the blueback run is at its peak. Grilse, which are approxi- mately the same size as the bluebacks, are among the chinooks and run at the same time, and it has seemed likely that bluebacks might be mistaken for grilse or grilse for bluebacks. An analysis has been made in which the correlation was determined be- tween the percentage of grilse in the total count of chinooks and the number of bluebacks for the 10 weeks of the blueback run — June 11 to August 13. The Pearsonian coefficient of correlation is —0.72. Using the standard procedure the probability of chance occurrence of a coefficient of correlation as high as this is only 0.03, so that the observed negative correlation between the percentage of grilse and the number of bluebacks can be accepted as significant. Furthermore, it seems likely that the relationship between these two variables is curvilinear rather than rectilinear, as assumed by the Pearsonian coefficient, and that a true measure of the correlation would be even higher than that calculated. Our measure is, therefore, conservative. It seems quite likely that this negative correlation can be ascribed to a tendency on the part of the observers to mistake grilse for bluebacks when the blue- backs are numerous. This raises the question as to what other errors there may be in the counts. It is certainly difficult to distinguish species under the conditions of counting unless there is a fairly well marked difference in size, shape, or markings, especially if light condi- tions are not favorable. Observers should not be blamed for making errors under these conditions, but, in view of the evidence of error in identification just given, it would seem proper to investigate carefully to see how extensive these errors may be. The importance of having properly trained and experienced observers is obvious. SALMON RUNS OF THE COLUMBIA RIVER IN 1938 113 Table S.—Co mis of chinook salmon at Rock Island Dam, 1933 to 1938 Week end- ing 1933 1934 1935 1936 1937 1938 Week end- ing 1933 1934 1935 1936 1937 1938 Apr 10 2 11 9 39 87 137 93 47 36 13 11 29 104 126 120 258 Aug. 6 Aug. 13.--. Aug. 20.... Aug. 27 ... Sept. 3 Sept. 10...- Sept. 17.... Sept. 24.... Oct. 1- 257 253 2,000 409 1,154 656 210 70 836 741 3.047 386 133 57 113 67 58 111 350 30 27 3 686 689 2,187 3,342 2,710 1. 104 437 1,077 306 629 123 41 5 848 275 139 65 21 645 42 241 172 102 61 65 371 239 P5 55 9 383 Apr 23 419 14 28 70 78 650 235 1U5 69 94 120 39 77 450 725 196 65 117 509 532 402 282 321 86 59 116 38 90 288 13 84 399 727 254 2!)8 201 vS 91 183 1,245 1,530 2 7 30 63 25 33 19 159 180 148 608 1,791 82 162 May 21 .. 171 May 28 .. 209 515 344 Oct. 8 344 Oct. 15 HI Oct. 22 8 Oct. 29 15 July 10 Nov. 5 July 23 July 30 8 51 Total.. 5,668 7,100 16, 301 6,475 5,132 5,803 T ABLE S . — Counts of blueback salmon at Rock Island Dam, 1933 to 1938 Week end- ing 1933 1934 1935 1936 1937 1938 Week end- ing 1933 1934 1935 1936 1937 1938 IS 3 2 3 6 6 Aue. 20 A.Ug, 27.... Sept. 3 Sept. 10.... Sept. 17.... Sept. 24.... Oct. 1 4,94! 827 125 66 42 26 104 35 30 16 13 4 1 2.172 561 180 23 4 3 1 168 20 14 241 74 128 45 12 6 5 8 266 93 1 4 6 43 35 2 37 2 80 139 871 S.95X 4.530 1,234 677 96 22 93 144 667 561 410 120 5 9 62 1,058 8, 856 2, 203 3,778 9 313 1,865 8,011 4,474 1,217 380 4 7 2,871 6,310 4,077 919 356 61 July 23 July 30.... 1,218 X, 900 16, 868 7,668 1 Total.. Aug. 6 .... Aug. 13 40. 737 2,227 14,011 16, 482 15,069 17.123 Table 10. — Counts o / stcelhead trout at Rock Island Pam, 1933 to 19;!S Week end- ing 1933 1934 1935 1936 1937 1938 Week end- ing 1933 1934 1935 1930 1937 1938 Apr 9 14 7 71 62 3 8 8 29 191 338 146 89 132 S7 9 12 2 9 1 7 14 46 85 200 Aug. 27... Sept. 3 Sept. 10...- Sept. 17 Sept. 24.... Oel l- 149 189 168 173 130 2 2 17 26 25 13 30 51 31 23 16 20 8 4 3 11 5 2 306 336 397 591 411 699 677 215 906 25 5 52 21 68 60 89 109 312 384 265 306 42 25 Apr. 16 49 222 143 243 67 55 395 100 53 28 2" 8 ] 4 12 60 62 57 45 33 Apr 23 56 97 May 7 15 35 135 304 618 70 32 9 11 18 10 27 70 77 74 59 9 55 211 67 15 18 9 26 14 10 12" 36 97 200 90 6 1 4 Oct. 8 126 Oct. 15 39 June 4 Oct. 22. .. 34 Oct ."i 67 . , a ' ' 3 1 1 2 4 1 2 3 4 Nov. IS Nc» . 10 Nov. 20 .. Dec. 3 Dpo. in July 30 Aug. 6 38 131 90 87 Deo '7 Dec 21 Total . . Aug. 13 Aug. 20 1,055 484 ■5,411 1,637 2,214 2,400 1 Includes 20 counted previous to the week ending April 9. Tables S, 9, and 10 give the counts made at Rock Island Dam during the years 1933 to 1938, inclusive. These were all actual counts which are presumably accurate, both as to number and identification. In addition to the records given in these tables, 78 silver salmon were counted at Rock Island Dam late in September and early in October 1938. MODIFIED TABLES The data on the run of 1938 (tables 1 to 10) are presented below in such form as to bring out certain facts bearing upon the biological and economic importance of 114 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE different portions of the salmon runs and upon matters important to their conserva- tion. This section deals with the methods used in forming these modified tables and the reasons for the various modifications that have been introduced. The chief purpose in the original report was to show the contribution that the Rock Island runs make to the commercial catch for different periods and also the intensity with which the run as a whole, and particularly tbe Rock Island component, is being exploited. In that report only the spring runs of chinook and blueback salmon and the steelhead trout were considered. In the present report all of the species of salmon found in commercial quantities in the Columbia River have been included and the data covering the fall season to the end of the year have been considered. Information not available at the time the original report was prepared has, we believe, made possible an improved analysis. Additional facts not pertinent to the original report but bearing on the more general problems of the depletion and conservation of these fishery resources have been introduced. Primarily for the purpose of comparing commercial catch with escapement of fish to the spawning grounds, it has been necessary to convert the catch as given in pounds into numbers of fish. Entirely satisfactory conversion factors (average weights) are not available, so that the estimated numbers as given in the following tables cannot be considered as anything more than reasonable approximations. The terminal digits in the figures as given are not, therefore, to be taken as significant. In the original report the following conversion factors were used in converting the catch, given as poundage landed, into numbers of fish: For chinook salmon 2 systems were used; (1) an average weight throughout the season of 22 pounds, and (2) an average of 15 pounds during May, 20 pounds during June, and 25 pounds during July and August. For bluebacks also 2 systems were used; (1) an average of 3 pounds throughout the season in all zones, and (2) an average of 3 pounds throughout the season below Bonneville (Zones 1 to 5) and 2}i pounds above Bonneville (Zone 6). For steelhead trout an average weight of 10 pounds throughout the season in all zones was assumed. In general these were in accord with accepted figures. In the present report we introduce no change in respect to the figures used for bluebacks and steel- heads, but have considerably modified our treatment of the ckinooks. In another paper (Rich 1940a) the writer has described the seasonal changes in weight of chinook salmon in the commercial catch on the Columbia River during the season of 1939, and the estimated weekly average weights given in that paper have been used in this report to convert poundage to number of fish. The validity of applying the 1939 averages to the 1938 run is perhaps questionable, but appears to us to be by far the most acceptable procedure available. It was shown in the paper just mentioned that a satisfactory empirical graduation of the observed weekly mean weights in 1939 is given by the use of two linear equations. Letting ?/=weekly mean weight, z=the week, with origin at the week of July 9, the data for the first part of the season, up to and including the week ending July 9, are fitted by the equation ?/= 30 +1.782, and those for tbe last part of the season, including again the week of July 9, are fitted by the equation y=Z0— 0.55i. Table 11, gives the estimated weights for each week of the spring season as determined from these equations. For this present report, estimated average weights for the weeks previous to the opening of the fishing season on May 1 and for the fall season have also been determined by the dubious method of extrapolation. We fully recognize the dangers of this procedure but, in the absence of any better objective basis for estimate, believe SALMON RUNS OF THE COLUMBIA RIVER IN 1938 115 it to be justified here. This gives the following estimated weights: For the week ending April 30, 12.20 pounds; April 23, 10.42; September 3, 25.60; September 10, 25.05; September 17, 24.50; September 24, 23.95; October 1, 23.40; October 8, 22.85; October 15, 22.30; October 22, 21.75; October 29, 21.20; and for the week ending November 5, 20.65. After this date so few fish were taken in the fishery that an approximation on the basis of about 20 pounds is adequate for all purposes. Table 11. — Estimated weights of chinook salmon in the commercial catch in Zones 1 and 2 for the spring season of 1939. Figures for the first 3 weeks were extrapolated Week ending Estimated mean weight Week ending Estimated mean weight Week ending Estimated mean weight Week ending Estimated mean weight May 7... May 14 (13.085 (15.76) (17.54) 19.32 21.10 22.88 24.66 26.44 July 2 28.22 30.00 29.45 28.90 28.35 Aug. 6 27.80 July 9 Aug. 13 27.25 Julv 16 Aug. 20 26.70 July 23 Aug. 27 26.15 July 30 In converting poundage of silver and chum salmon to numbers of fish we here adopt an average weight of 10 pounds for both species — the same as that adopted for steelhead trout. This is not in accord with the figures commonly given, viz, 7-9 pounds for silvers and 8-10 pounds for chums. Some years ago, however, the writer measured and weighed several hundred silver and chum salmon taken on the lower Columbia River, and these gave averages for both species that were considerably over 10 pounds — 240 chums averaged 10.3 pounds with a standard deviation of 2.0, and 133 silver salmon averaged 10.9 pounds with a standard deviation of 2.6. This average does not include 16 silver salmon grilse which were in the same collections. The samples came from fish caught in traps and the small grilse are seldom taken by gill nets although, as stated above, this form of gear is of primary importance in the Columbia River fishery. In view of these figures, and the purpose to which the esti- mates are to be put, it seems reasonable to use a conversion factor of 10 pounds for both of these species. 7 Some time is required for the journey of the fish up the river, so that on a given day the fish in the upper river may be expected to represent an entirely different stock from that to be found simultaneously in the lower river, although it is the same stock as was to be found in the lower river during an earlier period. Therefore, in order to aid interpretation of some of the more important data, these have been presented so that as nearly as possible those referring to the same stocks of fish are placed on the same lines in the table. In other words, the several series of data have been so "lagged" that comparable portions are related to the same marginal date — which date is the end of the week in which the fish may reasonably be expected to have entered the river from the ocean. From a careful examination of tables 1 to 5 it appears that a given group of fish that entered the river and were to be found in Zones 1 and 2 in a given week (the week of the marginal date in the table) would be in Zones 3 to 5 the next week, at Bonneville and in Zone 6 during the second week, and at Rock Island the fourth week after their appearance in Zones 1 and 2. In table 12 the dates given in the left-hand margin are those ending the weeks during which the fish were in Zones 1 and 2, the estimated catches made in Zones 3 to 5 ' Since this report went to press a paper by Wilbert Chapman, of the Washington State Department of Fisheries, dealing with the weights of Ash taken in the Columbia River fisheries has appeared. His figures are somewhat different from ours but it is not possible to give a critical discussion of them here. 116 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE were made 1 week later than that indicated by the marginal date, the Bonneville count and the estimated catch above Bonneville were made 2 weeks later than that indicated by the marginal date, and the Rock Island count 4 weeks later. For con- venience we shall refer below to the assumed position of the fish during their upward migration as in Zones 1 and 2 the first week, in Zones 3 to 5 the second week, at Bonne- ville and in Zone 6 the third week, and at Rock Island the fifth week of then- fresh- water migration. The same system was followed in preparing the similar tables for the other species. Thus, reading across any one line, say the line for May 7 in table 12, the first col- umn gives the estimated catch made in Zones 1 and 2 during the week ending May 7, the second column the estimated catch made in Zones 3 to 5 during the week ending May 14, the fourth column the count at Bonneville during the week ending May 21, the fifth column the estimated catch above Bonneville during the week ending May 21, and the seventh column the count at Rock Island during the week ending June 4. Columns 3 and 6 are derived by summing across the rows in the appropriate columns and therefore show totals for the run as a whole — all referred back to the week that the fish were presumably in the extreme lower part of the river and, therefore, approxi- mately to the time that they entered the river. Individual fish undoubtedly vary greatly in respect of their rate of travel up- stream, but the obvious similarity in the trends of all the columns in this table is evidence that, on the average, these assumptions are well founded. NATURE OF THE ANALYSIS OF RUNS From the tables of this structure it is possible, for those species that largely spawn above the site of the Bonneville Dam, to estimate the number of fish of each species that escaped the intensive fishery below Celilo Falls (the upper limit of commercial fishing) in 1938 and were available for reproduction above Bonneville Dam. This is readily done for any desired portion of the season by subtracting the catch above Bonneville from the Bonneville count. Such an estimate of the escapement is subject to error from several causes, of which the following may be mentioned: (1) Error in the count of fish of the different species at Bonneville, (2) error in the catch figures due to the fact that a considerable catch that does uot appear in the record is made by Indians, and to some extent by Whites for home use, and (3) error in converting pounds to number of fish. While these sources of error are present, it is believed that their total effect is relatively small and will not affect the general conclusions that may logically be drawn. Furthermore, in making these estimates no attempt has been made to correct for the spawning that takes place in the tributaries below Bonneville Dam. In the case of the silver and chum salmon such a large percentage of the spawning takes place below Bonneville that a similar analysis has not been made. Also, as mentioned above, there is a considerable part of the faU run of chinooks that spawns below Bonneville so that our study of the fall run is probably Jess reliable than that for the spring season. Since our estimate of the escapement is based primarily upon the count at Bonneville (from which is subtracted only the estimate of the number of fish in the recorded commercial catch above Bonneville) the spawning in the tributaries between Bonneville and the upper end of the commercial fishing district at Celilo Falls will not affect the results. If any considerable portion of the run that is actually derived from the tributaries below Bonneville be ascribed to the SALMON RUNS OF THE COLUMBIA RIVER IN 1938 117 river above Bonneville, this will tend to magnify the importance of the spawning in the river above Bonneville, including that above Rock Island. Undoubtedly a part of the commercial catch of all species except the blueback is composed of fish derived from the tributaries below Bonneville, but it seems probable that this forms a relatively small part of the total catch of chinook salmon, at least until after the peak of the fall run. There is a very large count of chinooks at Bonneville immediately after the beginning of the closed period in August — certain evidence that a large proportion of the fish that are in the river at that time are derived from populations spawning in the higher tributaries. On the whole we feel fairly confident that only a relatively small part of the commercial catch of this important species that is made before the first of October comes from the runs into tributaries below Bonneville. An understanding of the analysis of these runs, particularly in relation to the fish destined to spawn in the upper Colum- bia River above Rock Island Dam, may be aided by the following discussion (see also fig. 2). 8 While this particular treat- ment is related specifically to the run to Rock Island, a similar treatment could be applied to any other tributary runs for which similar data were available. Let us assume: A. That the estimated escapement at Celilo is the total escapement for the total run of the period; and B. That the ratio between the escape- ment at Rock Island Dam and the catch made from the same stocks of fish that furnished this escapement is the same as that between the escapement at Celilo and the total catch. This assumes that there is no appreciable loss between Celilo and Rock Island, and that, for each species, the proportion of Rock Island fish caught is the same as the average for all salmon of the species that are passing through the fishery at the same time. From this it would follow also that the relation between the escapement at Rock Island Dam and the run referable to this escapement will be the same as that between the escapement at Celilo and the total run. Having then determined, for a selected time interval, the total catch, denoted by C, the escapement at Celilo, denoted by Ei, and the count at Rock Island, denoted by E 2 , we are able to determine the following: • This clarifying symbolic treatment was contributed to the original report of the Board ot Consultants by Dr. Durand, who has kindly permitted slightly altered repetition here. 449G68— 42 3 Figure 2.— Diagram of the ultimate subdivisions of the main run of chinook salmon entering the Columbia River, illustrating the various ratios. R denotes total run; C, denotes total catch below Bonneville Dam; BC denotes Bonneville count; Ci denotes catch above Bonnevilt Dam; Ei denotes escapement at upper limit of commercial fishing; Ei denotes escapement at Rock Island Dam; " denotes diversions of unknown amounts at various points In the river. 118 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 1. The fraction of the total run (R) derived from that portion normally spawning above Rock Island. This will be -gr- 2. The fraction of the total catch (C) referable to the Rock Island escapement (Rock Island count). This also will be jgr" 1 W 3. The catch derived from the Rock Island contingent. This will be -gC. This catch in numbers of fish can then be converted into pounds weight on the basis of the assumed average weight per fish. 4. The total run referable to Rock Island. This will be I jfC+E 2 .j Likewise, the ratio of the catch referable to Rock Island to the total run referable to Rock Island. This will be ^C-M ^C+E 2 which reduces directly to C-HC+E,) or to total catch divided by total run, as might be expected. This may also be written, rather neatly, as ^ • That is, the ratio of the catch referable to Rock Island to 1+3 ' C the run referable to Rock Island is the same as the ratio of the total catch to the total run. This again follows from the assumptions A and B. In carrying out the analysis along the lines indicated above, the catch in number of fish and in pounds that may properly be ascribed to fish of the runs to the river above Rock Island has been taken as a measure of what may be termed the absolute importance of the Rock Island factor in the commercial fishery. The percentage of the entire run that, for any period, may be ascribed to these Rock Island fish, may similarly be taken as a measure of the relative importance of the Rock Island factor. These two series serve somewhat different purposes. These values may be determined for any selected portion of the season, and this is important because the Rock Island complement in the total run varies widely from time to time and the ratio of catch to escapement also varies during the fishing season. But for any one period it is possible to determine the ratio of catch to escapement — a ratio that may be applied to the entire run for the period or to fish bound for other tributaries above Bonneville Dam as well as to those destined to tributaries above Rock Island Dam. Given the ratio for any period, the catch ascribable to the upper Columbia may be deter- mined by multiplying the Rock Island count, E 2 , for the corresponding period, by this ratio,Sr> giving ( -^ )E 2 . Or, on the other hand, we may use the fraction of the entire 'El \El/ yp run that may be attributed to the river above Rock Island, -g?> and multiply the total catch C by this fraction to get the number of fish derived from those spawning above Rock Island giving \VjP- Mathematically these two procedures are obvi- ously identical and, where either may be applied, they will give identical results; but the latter procedure, making use of fractions of Rock Island fish in the run, may be applied when necessary to determine the part that the Rock Island fish play in producing the catch in any portion of the river, while the former can only be applied to the catch as a whole. SALMON RUNS OF THE COLUMBIA RIVER IN 1938 119 We will now consider, specifically, certain runs and portions of runs in respect of their importance to the general problems of the preservation of the salmon of the Columbia River, and in particular of those that have derived from the river above Grand Coulee Dam. Although the data have been studied and presented on the basis of time units of 1 week, it is convenient and even more illuminating to consider them also for longer intervals of time which have been selected for various reasons as being of special importance. CHINOOK SALMON HISTORY OF THE RUN OF 1938 On account of the dominating importance of this species in the fishing industry, particular attention has been paid to it. The data are presented in tables 12 to 14 and are shown graphically in fig. 3. The earliest part of the run to the Columbia River above Bonneville does not enter into the commercial fishery — it is past the commercial fishing area before the opening of the season on May 1. The first of the run to contribute to the commercial catch is that which enters the mouth of the river during the week ending April 23. These fish, in general, may be expected to pass Bonneville and to be in Zone 6 during the first week in May — the first week of the spring open season. We have therefore considered as a separate period the weeks up to and including the week ending on April 16. The next period includes the part of the run that provides the peaks in catch and Bonneville count that occur in May. We consider that this period ter- minates with the week ending May 28. The next period includes the succeeding 9 weeks ending on July 30, during which the catch and the Bonneville count were both relatively low, while at corresponding weeks the Rock Island count attained the maximum for the year. In the original report the last period treated covered only the 4 weeks ending August 27 — the last 4 weeks of the spring fishing season. It was impossible to carry the study beyond this because at the time the report was prepared data were not available for the fall season. But, with the data now on hand, it is obvious that the portion of the run beginning with the week ending August 6 and extending to the end of the year should be considered as forming a single unit rather than two or more units. In table 12 it is apparent that the run from the week of August 6 to the end of the year contains the main mode which, for purposes of study, should certainly not be broken up without good reason. Furthermore, table 12 and fig. 3 show that there is a mode in the Rock Island count for this period. In the present report, therefore, we shall take for the final period to be studied the entire remainder of the year after the week ending July 30. The data for these selected periods are given in table 13, which, for comparison, also includes the figures for the last period considered in the original report — July 31- August 27. Table 14 gives some of the more significant comparative figures that may be derived from table 13. 120 FISHEKY BULLETIN OF THE FISH AND WILDLIFE SERVICE Figure 3. — Dominant elements in the 1938 Chinook salmon run, by weeks. SALMON RUNS OF THE COLUMBIA RIVER IN 1938 121 Table 12. — Chinook salmon run in the Columbia River, 193S [ Catch in number of fish estimated from weekly average weights, as determined from the 1939 run. Datfi by corresponding weeks] combined and arranged Week ending Zones 1 and 2 Zones 3 to 5 Total catch below BonneviUe Bonneville estimate and count Catch above Bonneville Total catch Rock Island count Total run Feb 19 4 68 84 14 339 402 484 1,545 3, 359 12, 936 5, 097 3,827 205 1,981 2.932 2.230 1,240 884 1, 855 1,534 1, 753 1.327 4,163 5, 104 10,112 53. 753 80, 693 63. 224 12. 258 2, 057 period is not a natural subdivision of the run and that the count at Bonneville for the period is undoubtedly influenced by the incidence of the dosed season on August 25 have resulted in this and other differences between the data for the month of August and those for the entire period of the fall run. The Rock Island count for the period corresponding to that from July 31 to the end of the year was 1,879, or 1.1 percent of the estimated net escapement. Taking this as the percentage of Rock Island fish in the run as a Whole, the da'tcn that may be attributed to the Rock Island runs is estimated at 3,500 fish, or 91,500 pounds. This is to be compared with an estimate of 1,500 fish of an aggregate weight of lo.iino pounds for the month of August. Table 15 presents the more significant figures bearing On fchfe absolute and relative importance of the Rock Island runs of Chinook salmon. There are given not only the figures obtained through the basis of estimate adopted in this report, but also, for com- parison, those obtained through the two bases used in the original report by Calkins, Durand, and Rich. (The estimates given here for the full season on the bases used in the original report were not, of course, given in that report, which treated the catch only up to August 25.) It is apparent that, in general, the results of all three pro- cedures are of the same order of magnitude so that one may assume with some confi- dence that no gross errors have been introduced. Although we believe that the estimates based on the average weights obtained in 1939 are the most accurate, and should certainly be used for detailed study of parts of the run, it is clear that simpler methods will give approximate results of real value. 124 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Table 15. — Chinook salmon — comparison of certain estimates as made on the following bases: (1) An average weight of 22 pounds throughout the season; (2) average weights of 15 pounds in May, 20 pounds in June, and 25 pounds for the remainder of the year; and (3) average weights for each week as calculated from the trend lines described in the text. The first two were used in the original report by Calkins, Durand, and Rich Basis of estimate Ratio of catch to escapement Percentage of Rock Island fish in total run Catch attributed to Rock Island run— in fish Catch attributed to Rock Island run — in pounds April 17 to May 28 (1). (2). (3). 95, 300 104. 000 109, 000 May 29 to July 30 6.3 5.8 4.9 15.46 15.38 15.20 15, 800 15, 700 12, 300 346,900 379, 000 341,000 July 31 to August 27 (1). (2). (3). 40, 900 41, 000 40, 600 July 31 to December 17 2.3 2.0 1.9 1.16 1.11 1.10 4,400 3,700 3,500 96,600 92, 500 91,500 On the basis of these figures the total catch that may reasonably be attributed to the Rock Island runs is between 500,000 and approximately 600,000 pounds, of which by far the larger proportion was of the valuable spring run. Furthermore, it is of especial importance to note that the Rock Island run forms a particularly large percent- age of the seriously depleted and heavily fished June-July run. RATE OF TRAVEL These data provide additional information relative to the rate of migration up the river. We have given the reasons for thinking that the interval between the time that the fish appear in Zones 1 and 2 and at Bonneville is approximately 2 weeks. The peak of the run that occurs in late August and early September is obviously an important landmark and should, therefore, provide important evidence on this point — evidence that was not available at the time the original report was prepared. From the figures of the numbers of fish caught (estimated on the basis of the trend lines of average weights obtained in 1939) it would seem that the peak of the catch in Zones 1 and 2 came 3 weeks before the peak of the count at Bonneville, instead of 2 weeks (fig. 3). The drop in the catch that occurs between the weeks ending August 20 and 27, however, is due, at least in large part, to the fact that there were only 4 days of fishing in the week ending August 27. The spring fishing season closed on August 25. An estimate may be made of what the catch would have been if the full 6 days of fishing had prevailed, instead of 4 days, by multiplying the estimate already presented by IK- The result is over 71,000 fish; actually a few more than estimated for the week ending August 20. This result indicates strongly that the real peak of abundance in Zones 1 and 2 came not earlier than the week ending August 27 — 2 weeks earlier than the actual peak in the Bonneville count and quite in agreement SALMON RUNS OF THE COLUMBIA RIVER IN 1938 125 with the original assumption. Whether, without the closed period, the peak in the Bonneville count would have come in the week ending September 10 is perhaps some- what doubtful, and no method has occurred to us whereby that can be independently determined. From the total run (table 12) this would seem to be a reasonable infer- ence, but it has been based on the assumption that 2 weeks are required for the journey from the mouth of the river to Bonneville. In passing, it should be emphasized for future use in similar situations that the effect of the closed period has been to so increase the Bonneville count immediately following the beginning of the closed period that it has the effect of shifting the peak of the count upward. This would be true even if the final week of the open period had consisted of 6 days instead of 4 days of fishing. In general, the incidence of a closed period will increase the escapement in the following weeks, but in this case the peak of the run happens to coincide so closely with the beginning of the closed period (probably actually preceding it on the lower river) that the effect is to shift the peak of the escapement upward. Also, in this particular case, the fact that the last week of the open season contained only 4 fishing days had the effect of appar- ently shifting the peak of the catch downward. The combined result was an apparent lag of 3 instead of 2 weeks between the peak of the catch in Zones 1 and 2 and the peak of the count at Bonneville. Similarly, at the beginning of an open period there will be the reverse tendency for the peak of the escapement to be shifted downward and the peak of the catch to be shifted upward. Doubtless the peak of the Bonne- ville count that occurs during the week corresponding to that of April 30 has been so modified. Actually this count was made during the week ending May 14, and the fish passing Bonneville during that week were doubtless partly through Zones 1 and 2 before the fishing season opened on May 1. These rather confusing effects are, of course, due to the complementary relationship existing between the catch and the count at Bonneville. Related to these phenomena is the fact that there uppears to have been some delay in the passage of fish through Zone 6 following the peak of the run and the closed season. This is shown particularly by the fact that during the weeks ending September 10 to October 15 (almost the entire effective fall season) the catch above Bonneville exceeded the Bonneville count. However, we believe that this does not indicate a general lower average rate of travel, but is due, rather, to the combined influence of individual variation in the rate of travel and a constant reduction in the number of fish passing Bonneville. The anomaly, then, of the existence over a number of weeks of a greater catch above Bonneville than count over the dam is closely related to the fact that the peak of the escapement curve is shifted to an earlier date by the incidence of an open season. THE JUNE-JULY RUN As previously mentioned, the June-July run of chinooks is poor compared witli that in May or August, and it is rather generally thought that the populations form- ing this part of the run are the most seriously depleted of any. Some evidence of this was developed at the time the original study was made, but was not included in the original report. It has seemed worth while to pursue the investigation further. As bearing on the extent to which the June-July run has been depleted, we have examined data secured through the cooperation of the Columbia River Packers Asso- ciation. These data are in the form of reports of daily deliveries to this company 449668 — 42 * 126 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE over the period from 1912 to 1937, with the exception of occasional years for which no figures were available. It is unfortunate that similar data are not available for the entire river. During this long period the catch delivered to the association has averaged nearly 25 percent of the total deliveries on the Columbia River, and has ranged quite consistently between 20 and 30 percent. To test the reliability of these data as an index of changes in relative abundance during different periods, the Pear- sonian coefficient of correlation, "r," has been calculated between the total annual deliveries to the company and the total deliveries for the entire fishery as given in the report by the Oregon State Planning Board (1938). Between 1912 and 1937 there were 20 years for which complete records were available, and for these the coefficient of correlation is 0.86. The records appear to show, however, that some change took place about 1934, so that the records for the last 3 or 4 years are not consistent with those for earlier years. We have, therefore, calculated "r" for the 16 years of record between 1912 and 1928. The value is practically 0.9. Both show such a high degree of correlation that reasonable confidence may be placed in the assumption that the deliveries to the Columbia River Packers Association will serve to indicate long-time (secular) changes in relative abundance of chinook salmon in different parts of the season. Table 16. — Monthly totals of deliveries of chinook salmon to the Columbia River Packers Association, 1912-37, in thousands of pounds, for the spring fishing season only Year May June July August Total Year May June July August Total 1912 420 759 859 1, 163 684 717 378 882 854 468 727 624 749 683 1,203 2.194 496 578 643 665 1,194 594 440 973 1,629 1,381 1, 935 2,693 1,811 1,275 1.246 1,436 1,271 1,143 808 1,254 1,628 918 1, 378 1,685 3, 232 2,982 3,489 2,700 3,194 2.394 2, 12S 1,150 4,426 3,741 5,375 7.736 6,223 5,552 5, 756 5,683 6, 514 4,599 4,103 4,000 1924 1925 1926 1927 1928 609 703 169 638 440 296 428 93 329 608 584 992 996 732 704 503 502 (>x 843 616 701 445 1,270 1,100 934 794 702 680 714 456 886 605 676 1,296 1,747 1.6S0 1, 695 1,590 2,781 2,457 2.044 2,783 1,970 2,559 4,167 1913 4,546 1914 3,516 1915 3,830 1916 3,235 1917. 1931 1932 1933... 4,260 1918 4,286 1919 3, 435 1920 ... 1931 4,614 1921 .. 1936... 3,884 1922 1937 . 4,265 1923 . Notb. — The years 1929, 1930, and 1935 are omitted because of incomplete records. From data given in table 16 we have calculated the trends by the method c! averages, and these are shown in figure 4, which has been put on a semilogarithmic grid so that relative changes will be correctly shown and the several trends can be directly compared. It is apparent from this that while a general reduction has taken place, as is shown in each month and also in' the total, the reduction in the July catch has been by far the greatest. From a value of nearly 2,000,000 pounds at the beginning of this period (1912), the line of trend of the July deliveries has dropped to only about 600,000 pounds in 1937. The present deliveries are, there- fore, approximately one-third of what they were during July 25 years ago. At the same time the totals for the entire spring fishing season have dropped from about 6,000,000 pounds to about 3,500,000 pounds. This graph also shows that the deliv- eries during May have been seriously reduced. Curiously enough, the trend of the June deliveries is approximately the same as for the spring season as a whole, although those of May and July show evidence of much more serious depletion. Deliveries in August have not suffered nearly so much as those of the other months of the spring season — perhaps because of increased utilization of these later running fish which arc not of so good a quality as those of May, June, and July. SALMON RUNS OF THE COLUMBIA RIVER IN 1938 127 Before adopting the policy of treating all of the data on the basis of time units of 1 week, the daily records were examined and it soon appeared that there was, espe- cially in June and July 1938, a very definite weekly cycle of abundance as indicated by the catch in Zones 1 and 2. The Sunday closed season, of course, resulted in 1915 1920 1925 1930 1935 Figure 4.— Trends of the total monthly deliveries to the Columbia River Packers Association, 1912-37. practically no catch on that day, but there was a distinct tendency for the catches to be highest early in the week and to drop gradually toward the end of the week. The natural interpretation was that during the Sunday closed period a body of fish entered the river and on Monday (actually beginning Sunday evening) there were available to the fishermen, in addition to those left at closing time on Saturday, all of the fish that had entered the river and that were free of all commercial fishing during an 128 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE entire day; 6 pm. Saturday to 6 pra. Sunday. The effect of this accumulation was to increase the catch during the following day or two, but it wore off until, by the end of the week, little if any effect of the closed period remained. The character of the cycle obviously has been determined by the combined influence of a Sunday closed period and a very intensive fishery which, as shown above, takes approximately 80 percent of the fish entering the river during these 2 months. As an additional line of evidence of a dangerous intensity of fishing we have examin- ed in some detail the daily catches and, for comparison, the daily count at Bonneville Dam for the months of June and July, with attention to the variations in catch and count within weeks; in other words, with respect to the variation that is associated with the day of the week on which the catch or the count was made. The data are presented in table 17, together with certain derived figures. From the figures of catch and count given we have calculated for each week day, excluding Sunday in dealing with catches, (1) the mean of the total deliveries for that day of the week during the 8 weeks under investigation, (2) the mean percentage of the total catch for the week, (3) the mean delivery per gill net, and (4) the mean percentage of the weekly total count at Bonneville. These values are presented in table 18. It is apparent that all three measures relating to the catch show much the same thing; namely, that there is a fairly constant and uniform decrease during the first half of the week, while the catch during the last half is relatively stable and at a much lower level. On the other hand, no such progression is apparent in the count at Bonne- ville. This is as one would expect in view of the fact that has just been demon- strated — that the intensive fishery takes out of the run during the first 3 days of the fishing week a very large part of the fish that have entered the river during the Sunday closed period. Table 17. — Daily catch of chinook salmon in Zones 1 and 2, June 5 to July SO, 1938, and Bonneville count for corresponding runs, June 19 lo August IS, with derived figures showing fluctuations in catch during the week Date Total, all gear Total, gill nets only Number of gill-net deliveries Mean catch per delivery Percentage of weekly total Bonneville count Percentage of weekly total Pounds 605 17,612 16,047 13, 499 15, 867 14, 353 12, 139 Pounds 005 17, 012 16,047 13, 499 15, 867 14, 353 12, 139 7 288 291 244 240 246 235 86 61 55 65 66 58 51 0.7 19.5 17.8 15.0 17.6 16.0 13.5 0.0 15.1 15.8 18.1 15.1 16.8 19.1 0.2 20.8 20.0 16.3 13.8 12.2 16.7 0.2 22.2 20.0 17.2 14.4 12.3 260 283 318 422 283 446 218 191 191 167 159 243 159 130 231 243 128 113 133 36 1 149 79 618 471 261 11.7 12.7 June 7 Tu 14.3 June 8 W 18. 9 June 9 Th ... . 12.7 20.0 9.8 15.4 19, 402 20,542 23,412 19, 502 21, 767 24,681 368 40, 894 39,249 32, 141 27,200 23,979 32,808 389 42, 683 38,411 33, 043 27,712 23, 605 19, 462 20, 519 23,134 19, 025 20,804 21,915 368 35,627 35, 039 26, 481 21, 596 19, 578 23,850 335 336 358 324 318 337 4 416 429 374 332 308 366 58 61 65 59 65 65 92 86 82 71 66 64 65 15.4 June H Tu 13.5 June 15 W 12.8 June 16 Th 19.6 June 17 F ... - - 12.8 10.5 June 19 Su 24.1 25.4 June 21 Tu 13.4 June 22 W --. 11.8 June 23 Th .. 13.9 June 24 F 3.8 June 25 Sa - 7.7 4.2 June 27 M 37. 596 34, 501 31, 638 25, 871 21,452 421 419 391 354 32S 89 82 81 73 66 4.4 June 28 Tu -- 17.3 June 29 W 17.3 June 30 Th 26.4 July 1 F 14.6 1 The ladders were closed this day because of manipulation of the water levels, the count of the following day was divided equally between the two days. In calculating the percentage of the weekly total. SALMON RUNS OF THE COLUMBIA RIVER IN 1938 129 Table 17. — Daily catch of chinook salmon in Zones 1 and 2, June 5 to July SO, 19SS, and Bonneville count for corresponding runs, June 19 to August 13, with derived figures showing fluctuations in catch during the week — Continued Date Total, all gear Total. gill nets only Number of gill-net deliveries Mean catch per delivery Percentage of weekly total Bonneville count Percentage of weekly total July 2 Sa Pounds 26,612 1,413 49.636 54,333 51,809 59,264 54,530 49, 862 3,697 72, 215 52,111 47, 137 36, 822 29,583 22,313 2.591 49,528 46, 097 40. 653 32, 998 40.156 56,263 6.021 87. 161 79. 155 81.890 60. 369 63.156 57,918 Pounds 23.951 374 40,318 44,373 38. 866 42.447 36, 775 34, 826 30 57, 578 39,005 36,259 27, 852 22, 352 15,561 460 39.839 33.295 27,363 22,681 29,917 36,345 285 56.639 59,637 64,596 51.119 53, 332 52, 139 370 11 396 435 419 461 474 496 1 516 492 476 440 394 328 5 460 440 410 364 867 378 4 485 470 645 494 485 530 65 34 102 102 93 92 78 70 30 111 79 76 63 57 48 92 87 76 67 62 82 % 71 117 127 119 103 110 98 13.8 0.4 15.5 16.9 16.2 18.5 17.0 15.6 1.4 27.3 19.3 17.8 13.9 11.2 8.6 1.0 18.4 17.2 15.2 12.3 15.0 21.0 1.4 20.0 18.1 18.8 13.8 14.5 13.3 277 254 231 206 239 231 201 172 213 270 288 209 212 297 264 241 285 189 201 163 160 88 264 269 476 445 1,102 870 747 15.6 16.5 July 4 M 15.1 July 5 Tu 13.4 July 6 W 15.6 July 7 Th 15.1 July 8 F 13.1 July 9 Sa 11.2 Julv 10 Su 12.2 July 11 M 15.4 July 12 Tu 16.4 Julv 13 W 11.9 July 14 Th 12.1 July 15 F 16.9 July 16 Sa 15.1 July 17 Su 18.2 July 18 M 21.5 July 19 Tu 14.2 July 20 W . 15.2 July 21 Th 12.3 July 22 F 12.1 July 23 Sa . 6.6 July 24 Su -. 6.3 July 25 M .-. 6.2 July 26 Tu 11.4 Julv 27 W 10.7 Julv 28 Th 20.4 Julv 29 F 20.4 July 30 Sa 18.0 Table 18. — Variation in certain features of the chinook salmon catch in Zones 1 and 2 and of the Bonneville count during June and July, related to the days of the week Day of the week Sunday Monday Tuesday Wednesday. Thursday... Friday Saturday Thousands mean total catch 47.4 43.2 40.5 35.0 34.9 35.2 Mean percentage of weekly total 19.8 18.1 16.8 14.9 14.4 15.2 Mean delivery per gill net 88.9 83.0 7S I 72.9 72. B 69.8 Mean percentage of weekly total of Bonneville count 13.6 14.5 14.2 14.3 17.4 14.2 11.8 The intensity with which the June-July run is being exploited is shown in still another way by comparing the change in the weekly totals of the catch with the weekly totals of the Bonneville count for the corresponding weeks. These data are given in table 12, where the two series may be readily compared. It is seen that the catch below Bonneville during June and July constantly increased from 2,630 fish in the week of June 4, to 16,363 fish in the week ending July 30. At the same time the number of fish passing the Bonneville Dam remained, except for the last week, below the count for the first week of the period. It is obvious that the effect of an increased run entering the river is not felt at Bonneville — a result, without doubt, of a concurrent increase in the intensity of fishing. It is to be noted that the record of the catch above Bonneville Dam agrees with that of the Bonneville count, and thus supports this interpretation. As a measure of this intensity we may take the total number of 130 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE landings per week derived from the figures given in table 17, and shown in the following statement: Total number of deliveries per week in Zones 1 and 2 during June and July Week ending Deliveries June 11 1,551 June 18 2,008 June 25 2, 229 July 2 2,283 July 9 2,692 July 16 2,647 July 23 2,424 July 30 3,013 It is shown by the preceding statement that the number of deliveries practically doubles during the months of June and July — an increase in fishing effort that could readily account for the fact that the count at Bonneville Dam does not increase, although there is better than a fourfold increase in the number of fish taken in the fishery in Zones 1 and 2. In this connection it has been of interest to determine something of the relation- ship that exists between the abundance of fish as measured by the average poundage per delivery and the number of deliveries. The number of deliveries may be taken as a fair measure of the number of men fishing. We have, therefore, taken these two scries of values from table 17 and calculated the coefficient of correlation. This proved to be +0.75. The interpretation is quite clear that the abundance of fish, as shown by the size of the individual catches, is an important factor in determining the number of fishermen that will fish. INTENSITY OF FISHING IN GENERAL The runs of chinook salmon considerably outweigh in importance and value the runs of all other species in the Columbia River fishery combined. Of the entire run the part that enters the river during spring and early summer, April to July inclusive, is the most valuable on account of the fine quality of the fish. This part of the run, perhaps more than any other, has been adversely affected by the reduction of spawn- ing areas and localities suitable for the rearing of the young fish that has attended the utilization of the water resources in the headwaters, especially for power and irriga- tion. Since the salmon industry began on the Columbia River the chinook has been the mainstay of the fishery and the most relentless exploitation has fallen upon the spring run. It has been shown above that the present intensity of fishing is such that, in 1938, over SO percent of the spring run and between 60 and 70 percent of the main fall run of chinook salmon were taken in the commercial fishery. In this connection it is pertinent to recall that in the regulation of the Alaska salmon fisheries the Federal Government, acting through the Fish and Wildlife Service, has adopted the principle that the escapement should be not less than 50 percent of the entire run. There are sound theoretical grounds for thinking that the maximum sustained yield of the sal- mon fisheries can be maintained with an escapement of this order of magnitude, and the practical results obtained with the Alaska fisheries support this view. It seems SALMON RUNS OP THE COLUMBIA RIVER IN 1938 131 reasonably certain that, at least for the spring run of chinooks on the Columbia, the escapement is well below the level that would provide the maximum sustained yield. Such regulations and restrictions as have been imposed upon the Columbia River salmon fisheries apparently have very little effect insofar as they may act to reduce the intensity of fishing and provide a greater escapement of breeding fish to the spawn- ing grounds. It is to be noted that in the lower river the peaks of both spring and fall runs come within the spring open season so that, insofar as the fishery in the lower river is concerned, the main portions of both runs are exposed to the full force of the exploitation. There is the weekly closed period from C pm. Saturday to 6 pm. Sunday that is in force during the spring fishing season, May 1 to August 25, but it has already been shown that this has little value from the standpoint of conservation; its chief effect being to spread the fishery out over a longer stretch of the river. Again it has been shown that whatever effect the closed season, August 25 to September 10, may have in increasing the escapement through the lower river, it is largely offset by the intensive fishery that exists during September and October above Bonneville Dam. In a larger way this closed season acts much the same as does the weekly closed period, and chiefly tends to distribute the fishery over a wider area without materially increasing the breeding population. The effect of the closed season may be seen by examining table 19, which is a diagram representing the passage of a series of stocks of Table 19. — Effrct of a two-week closed period on the stocks of fish passing up the river at the assumed rate [Lctiers represent stocks of fish] Position of stock "Week Zones 1 and 2 Zones .i in :, Booth and Zone 1 A B C D E F G H 2 A B c O E F G H 3 A 4. B S... C 6.... D 7 E g F 9 G 10 H Note.— Bold-face letters represent closed period. fish through the fishing district at the rate we have assumed to hold. It is obvious from tliis diagram that there is no stock of fish that is wholly protected from exploita- tion by the closed season. For example, stock C is only protected by the closed season from exploitation in Zone 6; stock D is protected in Zones 3 to 6; stock E in Zones 1 to 5; and stock F in Zones 1 and 2 only. But, on the other hand, stock C is open to the very intensive exploitation below Bonnevdle just before the closed season, and stock D to the fishery in Zones 1 and 2 where a very large part of the total catch is made during the week just before the closed season. Stock E, however, is completely protected from the fishery below Bonneville but is exposed immediately after the closed season to the much intensified fishery above Bonneville. Stock F is protected 132 FISHERY BULLETIN OP THE FISH AND WILDLIFE SERVICE from the fishery in Zones 1 and 2 only, and also feels the full force of the intensified fishery above Bonneville, while stock G, entering the river at the end of the closed season, is given no protection at all. The closed season undoubtedly does help to increase the escapement to some degree, but it seems very probable that the heavy, concentrated run that enters the river during August and September is actually less intensively fished than is the spring run. This lowered fishing intensity is perhaps due in part to reduced effort by the fishermen, brought about by the lower price received for the fish, and also to the fact that with constant effort the percentage of fisb caught when the run is light is probably greater than when the run is heavy. The actual catch per unit of effort is, of course, greater with the heavier run, but the efficiency of the total effort, as measured by the ratio of catch to escapement, is probably inversely related to the intensity of the run. Within the last few years the use of fish wheels has been entirely eliminated, and the use of traps greatly curtailed. Ostensibly these restrictions were imposed in the interest of conservation, but they could only be effective insofar as they increased the escapement of fish to the spawning grounds, and correspondingly decreased the commercial catch. It seems rather doubtful that these restrictions have actually had this result, although the available data are inadequate either to prove or disprove the point. It may well be, however, that the elimination of these two forms of gear has only resulted in increasing the catch of other forms, without materially increasing the breeding stock. On the whole it would appear that the chinook salmon runs of the Columbia River are subjected to an exceedingly intensive fishery without any effective protection whatsoever, except such as has been afforded by the elimination of certain forms of gear and by artificial propagation. PERCENTAGE OF GRILSE Along with the larger fish that form the bulk of the chinook salmon run there are always some smaller fish, from 2 to 10 pounds in weight, that are commonly desig- nated as "grilse," or, among the Columbia River fishermen, "jack salmon," or simply "jacks." These are practically all males that have become sexually mature 1 or 2 years younger than the average and have, perforce, joined the spawning migration. It has been shown by Gdbert, Rich, and others that most grdse are in their second and third years, while the larger fish are in their fourth, fifth, or sixth years. In counting the fish past Bonnevdle Dam an effort has been made to record these grilse separately, as shown in table 7, and a study of these records has shown some interest- ing and significant fluctuations in the percentages of these smaller fish (fig. 5). It is apparent from this graph that, except for 2 periods during which the per- centage of grilse is consistently low, the average is about 20 percent. The fluctua- tions that involve only individual weeks may be taken as due to "sampling error," but those that extend over several weeks and show consistent change challenge some other explanation. The 2 periods that show consistently low percentages are those covering the weeks ending June 25 to July 16, and those ending September 10 to September 24. We have already explained the lower percentages of the first period as probably due to confusion of chinook grilse with blueback salmon during the peak of the run of this last species. The second period is that during which the Bonneville count is SALMON RUNS OF THE COLUMBIA RIVER IN 1938 133 greatly increased on account of the closed season from August 25 to September 10. The explanation is obvious. A very large part of the total catch of chinooks in the river below Bonneville is made by means of gill nets, and this type of gear is selec- tive — taking more of the larger fish and permitting most of the smaller ones to pass through. During the closed period this selection is not operating, and both large 35 I 30 - 1 •X. 1 1 o J 1 §25 / I / 1 \ A o / \ /\ / _J / \ / \ / < / \y\ l \ >* a 1 1- \ / \ 1 v/l f\ 1 O 20 - / I / \ / / \ / z / 1 1 \ / 1 \ 1 UJ / \ \ / 1 1 \ / a < tii Ul Ul H -Ot'tober 28, by weeks. the total catch formed by the catches above Bonneville (table 24). For the months of June and July only 4.5 percent of the total catch was token above Bonneville, but during August and September the percentage was 36.2. The relation of this to the net escapement also is shown by the percentages that the catch above Bonneville form of the Bonneville count. For the months of June and July oidy 10.3 percent of the fish counted past Bonneville were later captured hi the fishery above the dam. During the months of August and September, however, 43.9 percent was taken. As with the chinooks, the catch of steelheads above Bonneville during the first few weeks following the closed period of August 25 to September 10 exceeds the Bonneville count. This anomaly has been discussed and there seems to be no reason to doubt that the same factors were operating with the steelheads as with the chinooks. It has seemed possible in the case of the steelheads, however, that this phenomenon SALMON RUNS OF THE COLUMBIA RIVER IN 1938 143 might have been the result of misidentificatiou of this species in the Bonneville count — steelheads being mistaken for the much more numerous chinooks. In order to test the possibility of such misidentification on a large scale in the Bonneville count, a study was made of the ratios of the number of steelheads to chinooks in (1) the catch below Bonneville, (2) the Bonneville count, and (3) the catch above Bonneville for each week over the period beginning June 5 and ending October 29. It is to be expected that such series of ratios would vary over the entire period with the rela- tive numbers of fish of the 2 species, but the general trends of the ratios should be similar in the 3 localities in the absence of disturbing factors — such as misidentifica- tion in the Bonneville count. Figure 8 is a graph of these ratios wherein ordinary arithmetic coordinates are used, since the absolute values are the significant ones. It is apparent from this that the trends are very similar in the 3 localities; which is evidence that the identification at Bonneville was sufficiently accurate and probably was not responsible for the anomalous fact that more fish were recorded in tlie com- mercial catch above Bonneville than were counted over the dam. The data thus graphed are interesting in themselves in addition to their bearing on this particular problem. It is quite obvious that, in numbers of fish, the steel- heads approach the chinooks and, during the June-July period when chinooks are few, greatly exceed them. It is chiefly during the peak of the chinook run in August and September that the ratio is down to about 1 :5 in the catch below Bonneville and the Bonneville count. The parallelism in the 3 trends up to about the middle of September is quite striking and is supporting evidence that, for this part of the run, the assumed rate of travel is satisfactory. SILVER AND CHUM SALMON As mentioned in the introduction, the purposes of the original report by Calkins, Durand, and Rich were such that consideration of the catches of silver and chum salmon was not important. In this revision, however, it is pertinent to include the data available on these 2 species, and to examine these for whatever light they may throw upon the characteristics of the runs. The general features of the runs of silvers and chums are so similar that it is convenient to treat them together. The data for these species are given in modified form in tables 25 and 26. In converting poundage to numbers of fish an average weight of 10 pounds per fish has been used for both species. The same rate of migration up the river has been used as with the other species, although the rate of migration of both Isilvers ,and chums is more doubtful and of far loss significance than in the case of the other species. There is, however, no good evidence that the rate of travel is any different in the case of these 2 species than in the others, although the obvious irregularities in the time at which the main portion of the catches is made in the different zones (tables 4 and 5) lead one to suspect that the rates of travel of these species may be somewhat different. This is a matter that should be investigated, but it is necessary for the present to assume the same rate of travel — which has been done in preparing the modified tables. 144 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Table 25. — Silver salmon run in the Columbia River, 1938 [Catch in number of fish, assuming an average weight of 10 pounds. Data combined and arranged by corresponding weeks] Week ending Catch in Zones 1 and 2 Catch in Zones 3 to 5 Total catch Bonneville count Week ending Catch in Zones 1 and 2 Catch in Zones 3 to 5 Total catch Bonneville count Julv 30 1 2 13 238 1 3 67 1,692 2,636 Oct. 29 Nov. 5 Nov. 12 Nov. 19 Nov. 26 Dec. 3 Dec. 10 Dec. 17 Dec. 24- Dec. 31 Total . 20. 268 23.275 13, 346 6.209 2,275 1,592 5.762 1,069 186 154 4,128 1,979 1,596 831 864 547 204 88 49 24, 396 25,254 14.942 7,040 3,139 2, 139 5,966 1,157 235 154 18 Aug. 6 Aug. 13 Aug. 20 Aug. 27 54 1.454 2,636 115 6,964 4,766 1,933 239 56 94 389 240 212 138 9 4 1 7 Sept. 3 Sept. 10 4 4,559 23, 032 14, 454 10, 578 32, 410 31,914 902 2,470 1,166 2,838 5, 951 5,346 5,116 906 7,029 24,198 17, 292 16. 529 37, 756 37, 030 Sept. 17 Sept. 24 Oct. 1 __ Oct. 8 Oct. 15 Oct. 22 195, 232 34, 329 229, 561 15, 185 Note.— No catch was recorded for Zone 6. Table 26. — Chum salmon run in the Columbia River, 1938 ICatch in number of fish, assuming an average weight of 10 pounds. Data combined and arranged by corresponding weeks] Week ending Catch in Zones 1 and 2 Catch in Zones 3 to 5 Total catch below Bonne- ville Bonne- ville count Catch in Zone 6 Week ending Catch in Zones 1 and 2 Catch in Zones 3 to 5 Total catch below Bonne- ville Bonne- ville count Catch in Zone 6 Sept. 24 3 117 334 4,839 15, 960 22, 370 51, 458 41,601 16, 057 284 293 1.538 2,931 5,361 7,235 5,084 3,112 3 401 627 6.377 18,891 27, 731 58. 693 46,685 19, 169 2 68 179 945 174 236 225 202 26 7 365 28 195 297 Nov. 26 - Dec. 3 Dec. 10. Dec. 17 Dec. 24 Dec. 31. Total 6.074 1.083 2,286 453 44 15 1,210 440 57 23 8 7,284 1,523 2,343 476 52 15 46 13 1 Oct. 1 Oct. 8 Oct. 15 Oct. 22 Oct. 29 Nov. 5 Nov. 12 Nov. 19. _ 162, 694 27,576 190, 270 2,117 892 For both silver and chum salmon it is quite apparent that such a small part of each run goes above Bonneville that the same sort of analysis that was made of the data for the other species would be meaningless for these. Obviously the chief spawning areas are in the tributaries that enter the main river below Bonneville — an inference that is in entire accord with the known facts of the distribution of these species. Not only in the Columbia River but generally throughout their entire range, both silver and chum salmon tend to spawn in the lower tributaries of the larger rivers or in the shorter coastal streams. The same is true of the pink salmon, which do not appear in the Columbia in commercial quantities. Under such circumstances it is not possible even to approximate the number of fish in the entire run because the sum of the fish taken below the dam and those counted past Bonneville do not form a sufficiently large percentage of the whole, and without at least approximate informa- tion as to the total number of fish in the run it is impossible to make the sort of analysis that has been done with the chinooks, bluebacks and steelheads. The silver salmon first appeared in the river about the first of August, but the catch did not amount to much until after the closed period from August 25 to Sep- tember 10. On the other hand, a very large part of the total count past Bonneville was made during the 2 or 3 weeks that were chiefly affected by the closed season. (In table 25, the weeks ending August 20, August 27, and September 3.) Several factors, alone or in combination, may account for these facts. First it appears that a much larger percentage of the earlier fish than of the later ones pass above the dam to spawn in the upper tributaries. Secondly, the intensity of fishing for this species SALMON RUNS OF THE COLUMBIA RIVER IN 1938 145 may be greater after the closed period than before. This may be due in part to a change in the gear used on the lower river after the height of the fall run of chinooks has passed. The silvers, being smaller fish, may be more readily caught with gill nets of smaller mesh than is most effective for the larger chinooks. However this may be, it seems reasonably certain that in 1938 there was a small but fairly well separated run of silver salmon that entered the river late in August. The main part of the run of this species comes from about the middle of Septem- ber to about the middle of November. There is some evidence of separate modes in the run during this time, but it is not conclusive or even very strongly marked. The height of the entire run in the lower river comes close to the middle of October. Chum salmon do not begin to enter the river much before the first of October. From that date on the run gradually increases to a peak that comes about the first week in November. After this the run as gradually decreases to terminate late in December. There is no evidence of significant minor modes. As in the case of the silver salmon, comparatively few of these fish pass Bonneville Dam, although a small catch was recorded from Zone G. It is clear that the majority of the fish of this species spawns in the tributaries below Bonneville Dam. SUMMARY 1. Exceptional data are available for the study of the salmon runs of the Co- lumbia River for 1938. For the first time the catch data for Oregon and Washington were given in similar form so that they could be combined. As a result, the daily catch in pounds of each species in each of 6 zones (corresponding to the parts of the river bounding the 6 contiguous counties of Washington) is available for study. Co- incident with this the Bonneville Dam was closed and fish ladders were constructed, by means of which the fish surmounted the dam. On their way through the ladders the fish were conducted through narrow passages and over white surfaces, and the number of each species was recorded. There have also been available for study the counts of salmon passing through the fish ladders at the Rock Island Dam, on the upper Columbia River near Wenatchee, Wash. 2. By using appropriate conversion factors the catch in pounds has been con- verted into numbers of fish, so as to make these data directly comparable with the counts at Bonneville and Rock Island dams. Tables have been prepared in which are given (1) the weekly catch for each of 3 major areas representing natural groups of wnes, (2) the total catch, (3) the Bonneville count, and (4) the Rock Island count. For each major area the data have been appropriately "lagged" so that, as nearly as possible, those for the same part of the run will lie on the same line as the table is read from left to right. This lag has assumed that fish entering the river and to be found in Zones 1 and 2 one week will be found in Zones 3, 4, and 5 the second week, at Bonneville and in Zone 6 the third week, and at Rock Island the fifth week. These modified tables form the basis for study and analysis. 3. The general course of the run of each species is shown so far as possible by the available data. The chinook salmon enter the river throughout most of the year, but two quite distinct peaks are shown: One near the end of April, the so-called "spring" run, and the other the latter half of August. There is a period of marked scarcity during June and July. The blueback run is of much shorter duration, the main por- tion lasting only 6 or 8 weeks and showing a marked peak toward the end of June. Steelhead trout enter the river throughout the year but the chief run is during the 146 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE months of June to September. There are 5 modes: minor ones about the end of March, the first of May, and the first of December, and major modes early in July and about the middle of August. The run of silver salmon extends from early in August to the end of the year, but centers rather broadly from the middle of September to the middle of October. The chum sahnon run attains a well marked maximum about the first week in November, but extends from about the first of October to about the middle of December. 4. The main parts of the chinook, blueback, and steelhead runs spawn above Bonneville, but silvers and chums spawn chiefly in the tributaries below the dam. 5. There is some evidence of error in the identification of species in the Bonne- ville count. 6. The importance of the runs to the river above Rock Island (largely affected by the dam at Grand Coulee) is shown by the ratio of the Rock Island count to the estimated escapement. Some 4 percent of the very early chinooks passing Bonneville previous to the first of May appear later at Rock Island. Of the May run of this species, about 6 percent apparently went to this portion of the river. Of the June- July run, which is poor and apparently seriously depleted, some 15 percent is attrib- utable to these races. During the remainder of the year only about 1 percent of the estimated escapement appeared here. Approximately 40 percent of the blueback run spawns above Rock Island. In the case of the steelheads, the early and late runs contain 10 percent or more of fish spawning above Rock Island; but during the main portion of the run, June through September, only about 1 percent of these fish go to this portion of the river. 7. The intensity of the fishery for chinooks, bluebacks, and steelheads is measured by the ratio of the commercial catch to the escapement, as calculated from the data given in the modified tables. For the May run of chinooks it is shown that only about 1 fish out of 7 escapes the commercial fishery and is available for the future maintenance of this run. During June and July, a period of great scarcity, only about 1 fish in 6 escapes, and during the remainder of the run, August through Decem- ber, the escapement is considerably better but even at this time about twice as many fish are taken in the commercial fishery as remain to reproduce. These figures do not take into consideration the effect of the intensive oceanic fishery which would materially increase the catch-escapement ratio. In the case of the blueback salmon the ratio of catch to escapement is approximately 4:1, indicating that only about 1 fish out of 5 of this species escapes the fishery. The ratio for the steelheads varies with the season, but for the main part of the run, June to September, it is somewhat greater than 2:1 ; i. e., more than 2 out of 3 steelheads are taken in the fishery. Similar ratios for the silvers and chums cannot be determined because few fish of these species pass Bonneville; consequently no estimate of the net escapement can be made. 8. The weekly closed period, 6 p. m. Saturday to 6 p. m. Sunday, in force during the spring fishing season, May 1 to August 25, is almost entirely ineffective insofar as it may tend to increase the number of breeding fish on the spawning grounds. Its chief effect is to spread the fishing over a longer stretch of the river. This is the result of an intensive fishery conducted over a long area. The closed season from August 25 to September 10 is designed to protect the peak of the chinook run and a portion of the steelhead run, but it acts, in a larger way, much the same as does the weekly closed period in that it chiefly tends to extend the fishing areas. The effect of an increased escapement of fish through the fishing area below Bonneville is almost SALMON RUNS OF THE COLUMBIA RIVER IN 1938 147 entirely offset by tbe very intensive fall fishery that is concentrated in Zone 6, above Bonneville Dam. 9. The closed period of March and April protects from the commercial fishery the run of cbinooks that enters the Willamette River during Aprd and early May, but this run is subjected to an intensive sport fishery below the falls at Oregon City. Unfortunately there are no data on the sport catch or on the Willamette run as a whole. This closed period also protects a small run of cbinooks to the main river, the principal portion of which passes through the commercial fishing area before the season opens on May 1. 10. The main runs of all species of salmon to the Columbia River are practically unprotected from exploitation. If all existing restrictions were removed, it is doubt- ful whether the catch would be materially increased, or, conversely, that the remaining brood stock would be materially decreased. The only present aids to the conservation of those runs are apparently those afforded by artificial propagation, stream improve- ment, and, possibly, the restrictions that apply to the use of traps and wheels. LITERATURE CITED Craig, Joseph A., and Hacker, Robert L. 1940. The history and development of the fisheries of the Columbia River. U. S. Bur. Fish. Bull. 49(32): 133-216. Washington. Pritchard, Andrew L. 1934. Pacific salmon migration: The tagging of the spring salmon in British Columbia in 1929 and 1930. Biological Board of Canada, Bull. 41, 31 pp. Ottawa. Rich, Willis Horton. 1940a. Seasonal variations in weight of Columbia River chinook salmon. Copcia 1: 34-43. New York. 1940b. The future of the Columbia River salmon fisheries. Stanford Ichthyological Bulletin vol. 2, No. 2. Palo Alto. 1941. The present state of the Columbia River salmon resources. Sixth Pacific Science Congress, Proceedings, vol. 3. Berkeley. U. S. Bureau of Fisheries. 1928. Fisheries Service Bulletin No. 152, Jan. 2, 1928. UNITED STATES DEPARTMENT OF THE INTERIOR Harold L. Ickes, Secretary FISH AND WILDLIFE SERVICE Ira N. Gabrielson, Director Fishery Bulletin 38 BIOLOGY OF THE ATLANTIC MACKEREL (Scomber scombms) OF NORTH AMERICA Part I: Early life history, including the growth, drift, and mortality of the egg and larval populations By OSCAR ELTON SETTE From FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Volume 50 ,, **A UNITED STATES GOVERNMENT PRINTING OFFICE WASHINGTON : 1943 Kor sale by the Superintendent of Documents. U. S. Gove/nment Printing Office Wa*hington 25. D. C. - - - - - Price 25 ccnti ABSTRACT This portion of a comprehensive study on the Atlantic mackerel (Scomber scombrus) treats of the early life history from spawning up to about the time the schooling habit develops, with emphasis on the quantitative aspects. Spawning takes place along the Atlantic coast, mostly 10 to 30 miles from shore, from Chesapeake Bay to Newfoundland, with perhaps Yio of the volume between the Chesapeake Capes and Cape Cod; Mo in the southern half of the Gulf of St. Lawrence, and negligible amounts elsewhere. Embryological development at the temperature usually encountered occupies about 1 week. The pelagic eggs are confined to a surface stratum 15-25 meters thick. Hatching at 3 mm. of length, larvae grow to 10 mm. in about 26 days, and to 50 mm. in an additional 40 days, by which length they approximate the typical form for adult mackerel, and assume the schooling habit. In 1932, it is estimated, 64,000 billion eggs were produced south of Cape Cod by a spawning population estimated at 100 million individuals. That year dominant north- easterly winds (which were abnormally strong) drifted one concentration of larvae, originat- ing off northern New Jersey, and another concentration, originating off southern New Jersey, in a southwesterly direction, to localities abreast of Delaware Bay and Chesapeake Capes, respectively. A reversal of dominant winds, consequently of drift, returned both groups to northern New Jersey, by the 9-mm. stage of growth. Mortality during most of the developmental period was 10 to 14 percent per day, but was as high as 30 to 45 percent per day during the 8- to 10-millimeter period when fin develop- ment we s rapid. Survival from spawning of the eggs to the end of the planktonic phase of life (50 mm.) was in the order of 1 to 10 fish per million eggs spawned. This rale of survival is an abnormally low one since the fish from this spawning season were abnormally scarce in the adult populations of subsequent years. The low survival rate is ascribed to the abnormal amount of southerly drift, coupled with a general scarcity of plankton in the spring of 1932. BIOLOGY OF THE ATLANTIC MACKEREL (SCOMBER SCOMBRUS) OF NORTH AMERICA. PART 1: EARLY LIFE HISTORY, INCLUDING GROWTH, DRIFT, AND MORTALITY OF THE EGG AND LARVAL POPULATIONS By Oscar Elton Sette, United Stales Fish and Wildlife Service J- CONTENTS Face Introduction 149 Account of field work 151 Synopsis of results 152 Significance of results 155 Life history 156 Reproductive age 156 Fecundity 156 Spawning grounds and spawning sea- sons 158 Coast of southern New England and Middle Atlantic Statrs___ 158 Gulf of Maine 159 Coast of Nova Scotia 1 (10 Gulf of St. Lawrence 160 Relative importance of the var- ious spawning grounds 161 Numbers of eggs spawned and size of spawning stock 164 l'aKe Life history — Continued. Spawning habits 165 The egg 166 The larva 170 Growth 173 Drift and migration 183 Mortality 191 Appendix 208 Methods of determining size at maturity 208 Methods of collecting eggs and larvae. 209 Enumeration of eggs and larvae 211 Computations of catch per station . . 213 Records of tow netting and catches of 1932 219 Sizes of youngest post-planktonic mackerel 235 Literature cited 230 INTRODUCTION The common mackerel, Scomber scombnis, is found on both sides of the Atlantic Ocean, approximately between the 30th and 50th parallels of north latitude. Although American and European representatives are very much alike in appearance, life history, and habits, their ranges are discontinuous, so that the two populations may be regarded as separate races with no intermigration. Consistent with this view is the observation (Garstang, 1898, p. 284) that the two stocks differ in morphological characters. The American race has from colonial times been caught and marketed in large volume. 1 In the nineteenth century the annual yield occasionally reached 200,000,000 pounds. The present yield is about 60,000,000 to 80,000,000 pounds annually, of which the United States fishery takes about tliree-quarters and the Canadian fishery the remainder (Sette and Needier, 1934, p. 43). 1 The European race, too, is the object of an important commercial fishery, but appears never to have been held as high in esteem or occupied so high a rank among the commercial fishes of Europe as has Its American relative among the fishes of tills side of the Atlantic. Fishery Bulletin 38. Approved for publication May 15, 1939. 149 150 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Among the commercial fishes, the mackerel is remarkable for its spectacular changes in yield. To illustrate this, only a few records need be selected (Sette and Needier, 1934, p. 25). From 116,000,000 pounds in 1834 the United States catch dropped to 23,000,000 pounds in 1840, only to rise again to 137,000,000 pounds in 1848. From its peak of 179,000,000 in 1884, the catch dropped to 30,000,000 in 188(3, o;dy 2 years later. More recently it increased from 13,000,000 pounds in 1922 to 08,000,000 pounds in 1926. For the United States and Canada together the largest catch, 234,000,000 pounds, was landed in 1884, the lowest, 12,600,000 pounds in 1910. Although these fluctuations had profound effects both on the economic welfare of the fis