M8L U.S. Department of Commerce Seattle, Washington Volume 93 Number 1 January 1995 Fishery Bulletin Contents *Jfc'&**j*. Cceanr 15 Companion articles JAN 9 1995 Woods Hci; Barlow, Jay The abundance of cetaceans in California waters. Part I: Ship surveys in summer and fall of 1991 Forney, Karin A., Jay Barlow, and James V. Carretta The abundance of cetaceans in California waters. Part II: Aerial surveys in winter and spring of 1991 and 1992 The National Marine Fisheries Service (NMFS) does not approve, recommend, or endorse any proprietary product or proprietary material mentioned in this publication. No reference shall be made to NMFS, or to this publication fur- nished by NMFS, in any advertising or sales promotion which would indicate or imply that NMFS approves, recom- mends, or endorses any proprietary product or pro-prietary material men- tioned here-in, or which has as its pur- pose an intent to cause directly or indirectly the advertised product to be used or purchased because of this NMFS publication. Articles 27 Bruce, Barry D. Larval development of King George whiting, Sillaginodes punctata, school whiting, Sillago bassensis, and yellow fin whiting, Sillago schomburgkii (Percoidei: Sillaginidae), from South Australian waters 44 Carls, Mark G., and Charles E. O'Clair Responses of Tanner crabs, Chionoecetes bairdi, exposed to cold air 57 Cortes, Enric Demographic analysis of the Atlantic sharpnose shark, Rhizoprionodon terraenovae. in the Gulf of Mexico 67 Ellis, Denise M., and Edward E. DeMartini Evaluation of a video camera technique for indexing abundances of juvenile pink snapper, Pristipomoides filamentosus. and other Hawaiian insular shelf fishes 78 Finnerty, John R., and Barbara A. Block Evolution of cytochrome b in the Scombroidei (Teleostei): molecular insights into billfish (Istiophoridae and Xiphiidae) relationships Fishery Bulletin 93 1 1), 1995 97 Kornfield, Irv, Austin B. Williams, and Robert S. Steneck Assignment of Homarus capensis (Herbst, 1 792), the Cape lobster of South Africa, to the new genus Homarinus (Decapoda: Nephropidae) 1 03 Milton, David A., Steven A. Short, Michael F. O'Neill, and Stephen J. M. Blaber Ageing of three species of tropical snapper (Lutjanidae) from the Gulf of Carpentaria, Australia, using radiometry and otolith ring counts 1 16 Natanson, Lisa J., John G. Casey, and Nancy E. Kohler Age and growth estimates for the dusky shark, Carcharhinus obscurus, in the western North Atlantic Ocean 1 27 Rickey, Martha H. Maturity, spawning, and seasonal movement of arrowtooth flounder, Atheresthes stomias, off Washington 1 39 Schmid, Jeffrey R. Marine turtle populations on the east-centrai coast of Florida: results of tagging studies at Cape Canaveral, Florida, 1986-1991 Notes 1 52 Benetti, Daniel D., Edwin S. Iversen, and Anthony C. Ostrowski Growth rates of captive dolphin, Coryphaena hippurus, in Hawaii 1 58 Canino, Michael F, and Elaine M. Caldarone Modification and comparison of two fluorometric techniques for determining nucleic acid contents of fish larvae 1 66 Laidig, Thomas E., and Stephen Ralston The potential use of otolith characters in identifying larval rockfish (5eo<3sfe5 spp.) 1 72 Matsuura, Yasunobu, and Roger Hewitt Changes in the spatial patchiness of Pacific mackerel. Scomber japonicus, larvae with increasing age and size 1 79 Riley, Cecilia M., G. Joan Holt, and Connie R. Arnold Growth and morphology of larval and juvenile captive bred yellowtail snapper, Ocyurus chrysurus 1 86 Secor, David H., T. Mark Trice, and Harry T. Hornick Validation of otolith-based ageing and a comparison of otolith and scale-based ageing in mark-recaptured Chesapeake Bay striped bass, Morone saxatilis 191 Szedlmayer, Stephen T, and Jeffrey C. Howe An evaluation of six marking methods for age-0 red drum, Sciaenops ocellatus 1 96 Wiley, David N., Regina A. Asmutis, Thomas D. Pitchford, and Damon P. Gannon Stranding and mortality of humpback whales, Megaptera novaeangliae, in the mid-Atlantic and southeast United States, 1 985-1 992 207 Awards Abstract. — A ship survey was conducted in summer and fall of 1991 to estimate the abundance of cetaceans in California waters be- tween the coast and approximately 555 km (300 nmi) offshore. Line- transect methods were used from a 53-m research vessel. Approxi- mately 10,100 km were searched, and 515 groups of cetaceans were seen. The estimated abundances and coefficients of variation (in pa- rentheses) of the most common small cetaceans are the following: 226,000 (0.28) short-beaked com- mon dolphins, Delphinus delphis; 78,400 (0.35) Dall's porpoises, Pho- coenoides dalli; 19,000 (0.41) striped dolphins, Stenella coeruleo- alba; 12,300 (0.54) Pacific white- sided dolphins, Lagenorhynchus obliquidens; 9,470 (0.68) long- beaked common dolphins, Delphinus capensis; and 9,340 (0.57) northern right whale dolphins, Lissodelphis borealis. The estimated abun- dances (and CV's) of the most com- mon large cetaceans are 2,250 (0.38) blue whales, Balaenoptera musculus; 935 (0.63) fin whales, Balaenoptera physalus; 756 (0.49) sperm whales, Physeter macrocephalus; and 626 (0.41) humpback whales, Megap- tera novaeangliae. Estimates are also made for other species and for higher-level taxa that could not be identified to species. The abundance of cetaceans in California waters. Part I: Ship surveys in summer and fall of 1991 Jay Barlow Southwest Fisheries Science Center National Marine Fisheries Service. NOAA RO. Box 271. La Jolla. California 92038 Manuscript accepted 31 May 1994. Fishery Bulletin 93:1-14 ( 1995). The abundance of cetaceans in Cali- fornia waters is poorly known for the majority of species found there. For small cetaceans, quantitative estimates of abundance with statis- tical confidence limits are available only for common dolphins, Delphi- nus delphis (Dohl et al., 1986) and for harbor porpoise, Phocoena phocoena (Barlow, 1988). For large cetaceans, such estimates are avail- able for gray whales, Eschrichtius robustus (Reilly, 1984; Buckland et al., 1993a); humpback whales, Meg- aptera nov aeangliae (Calambokidis et al., 1990a, 1993 1 ), and blue whales, Balaenoptera musculus. 1 Estimates have been made for some of the other species (Dohl et al. 2,3 ), but these estimates are more than 10 years old, and most lack informa- tion on statistical precision. Many, and perhaps all, cetaceans in California waters are vulnerable to entanglement and death in gillnet fisheries. A program is now in place to estimate the incidental mortality of cetaceans in the Cali- fornia gillnet fisheries (Lennert et al., in press). It is difficult, however, to assess the impact of gillnet mor- tality on cetacean populations with- out knowing population sizes. Co- ordinated ship and aerial surveys were initiated recently to estimate the abundance of all cetacean spe- cies in the region of California gillnet fisheries. To evaluate the ef- fect of seasonality on cetacean abun- dance, surveys were designed to cover both cold-water months (Feb- Apr) and warm-water months ( Jul- Nov). A ship survey was conducted during the warm-water period of 1991; an aerial survey was conducted during the cold-water periods of both 1991 and 1992. Results from the ship survey are reported here; population estimates from the aerial surveys are reported in a companion paper (Forney et al., this issue). Field methods A line-transect survey was con- ducted from 28 July to 5 November 1991 with the 53-m National Ocean- ographic and Atmospheric Admin- 1 Calambokidis, J., G. H. Steiger, and J. R. Evenson. 1993. Photographic identification and abundance estimates of humpback and blue whales off California in 1991-92. Final Contract Rep. 50ABNF100137, sub- mitted to the Southwest Fish. Sci. Cent., P.O. Box 271, La Jolla, CA 92038, 40 p. 2 Dohl, T. P., K. S. Norris, R. C. Guess, J. D. Bryant, and M. W. Honig. 1978. Cetacea of the Southern California Bight. Part II of Summary of marine mammals and sea- bird surveys of the Southern California Bight area, 1975-78. Final Rep. to the Bu- reau of Land Management, 414 p. [NTIS Rep. No. PB81248189.] 3 Dohl, T. P., R. C. Guess, M. L. Duman, and R. C. Helm. 1983. Cetaceans of central and northern California, 1980-83: status, abundance, and distribution. Final report to the Minerals Management Serv., Con- tract No. 14-12-0001-29090, 284 p. 1 Fishery Bulletin 93(1). 1995 istration (NOAA) vessel McArthur to assess the abun- dance of cetaceans in California waters. Primary cruise tracks were drawn for a unifirm survey of the 814,900 km 2 area between the 18-m (10-fathom) isobath and approximately 555 km (300 nmi) offshore (Fig. 1). Primary observation team The basic survey method was that which was devel- oped and used to estimate the abundance of small cetaceans in the eastern tropical Pacific (Holt and Powers, 1982; Holt, 1987; Holt and Sexton, 1989; Wade and Gerrodette, 1993). The primary observa- tion team consisted of three observers who searched from a viewing height of 10 m above the sea surface: two observers searched with 25x pedestal-mounted binoculars; the third observer searched with unaided eye, and (occasionally) 7x binoculars, and also served as data recorder. Observers rotated among these three duty stations every 1/2 hour, and two observer teams alternated work and rest periods every two hours. Sighting effort was maintained from dawn to dusk whenever weather conditions allowed, and searching covered the entire region from directly in front of the vessel to 90 degrees left and right and Figure 1 Transect lines (thin solid lines) completed during the survey. The bold polygon indicates the limit of the main study area. out to the horizon. Data were recorded on a lap-top computer that had direct input from the ship's GPS (Global Positioning System) navigation system. Re- corded data included sighting conditions (sea state, cloud cover, sun position, etc.), observer positions, the beginning and end of effort, and information per- taining to sightings. When a sighting was made, all observers were made aware of the animals' location. The perpendicu- lar distance from the trackline to the center of the group was estimated from the initial bearing and distance. The initial bearing of a cue (a blow, a splash, or a sighting of animals) was measured relative to the bow of the vessel by means of a calibrated collar on the base of the yoke of the 25x binoculars. The initial distance was typically estimated from a cali- brated reticle scale in the oculars of both the 25x and 7x binoculars with the formula derived by Smith ( 1982) and was calibrated by using radar-measured distances to inanimate objects (Barlow and Lee, 1994). If a shore horizon was closer than 11.1 km (6 nmi), distance was estimated by comparison with the radar-measured distance to shore. Occasionally, for very close animals seen only by the third observer, sighting distances and angles were estimated by eye. If a cue turned out to be a cetacean, effort was inter- rupted and the ship was typically diverted towards the animals in order to obtain esti- mates of species composition and group size. The vessel was not typically diverted for ce- taceans that were greater than 5.55 km (3 nmi) perpendicular distance from the trackline. Species identification was made collec- tively by the team, but quantitative estimates of species composition and group size were made independently by each observer. For estimation purposes, a group was defined as a collection of closely associated individuals (typically within several body lengths of each other) that exhibited cohesive behavior. In the field, however, a single distant sighting might prove to be two behaviorally distinct groups upon closer inspection. In such cases, when it was impossible to determine which was the original group sighted, both groups were pooled to estimate group size and spe- cies composition. For mixed-species groups, species composition was recorded as an observer's estimate of the percentage of each species present in the group. The observers recorded species composition and group-size data in confidential personal notebooks, and the data were transcribed at the end of the day into the computer data record by the cruise leader. Barlow. Abundance of cetaceans in California waters: ship surveys Species identification Observers attempted to classify all the species present in a group to the lowest possible taxonomic level (one member of each team was a cetacean iden- tification expert with at least nine months of at-sea survey experience on prior marine mammal surveys). Several higher taxonomic groups were used in cases where species identification was not possible. These higher groups were beaked whales of the genus Mesoplodon; unidentified sei or Bryde's whales; uni- dentified beaked whales (including members of the genera Mesoplodon and Ziphius); unidentified large whales (including members of the species group "large whale" in Table 1 as well as the genera Esch- richtius and Eubalaena); unidentified baleen whales (including members of the genera Balaenoptera, Megaptera, Eschrichtius, and Eubalaena); unidenti- fied small whales (including members of the species groups "small whales" and "large delphinids" in Table 1 ); unidentified delphinoids (including members of the species groups "small delphinids," 'large delphinids," and "cryptic species" in Table 1); and unidentified cetaceans (which could include any of the species listed above or in Table 1). The number of sightings identi- fied to these higher taxonomic levels is relatively small, and these animals were not included in the abundance estimates for individual species. Conditionally independent observer In addition to the primary observation team, a fourth observer was on duty 81% of the time and looked for cetaceans that were missed by the primary team. This conditionally independent observer was sta- tioned immediately next to the other observers, searched with 7x binoculars and unaided eyes, and did not reveal the presence of cetaceans until after they were clearly missed by the primary observation team (i.e. after they had passed abeam of the vessel or were bow-riding). Nine different people served as independent observers during the survey, and all worked irregular schedules that overlapped with both primary teams. Independent observers did not work more than two consecutive hours. When a sighting was made by the independent observer, that person maintained their normal behavior so as to avoid drawing the attention of the primary observer team. Initial bearing and distance were estimated by eye or with the aid of reticles in the ocular of 7x binocu- lars and a hand-held protractor. After a group was clearly missed by the primary team, the independent observer announced the presence of the animals to the data recorder and gave the initial bearing and distance. Typically the vessel was diverted towards the group, and species composition and group size were estimated by the primary observation team. Analytical methods Cetacean abundance was estimated from survey data with line-transect methods (Buckland et al., Table 1 Number of groups of cetaceans which contained members of the indicated species and species groups. The sum of all species in a group may be greater than the total for that group because the latter contains mixed-species groups. Totals do not include off-effort sightings. Species group and No. of species sightings Small delphinids 285 short-beaked common dolphin, Delphi nus delphis 123 long-beaked common dolphin, Delphinus capensis 6 unclassified common dolphin, Delphinus spp. 8 striped dolphin, Stenella coeruleoalba 24 Pacific white-sided dolphin, Lagenorhynchus obliquidens 12 northern right whale dolphin, Lissodelphis borealis 16 unidentified delphinoid 21 Cryptic species 132 harbor porpoise, Phocoena phoeoena 32 Dall's porpoise, Phocoenoides dalli 97 pygmy sperm whale, Kogia breviceps 3 Large delphinids 37 bottlenose dolphin, Tursiops truncatus 16 Risso's dolphin, Grampus griseus 29 killer whale, Orcinus orca 5 Large whales 127 sperm whale, Physeter macrocephalus 13 Baird's beaked whale, Berardius bairdii 1 Bryde's whale, Balaenoptera edeni 1 Bryde's or sei whale, Balaenoptera edeni or B. borealis 2 fin whale, Balaenoptera physalus 22 blue whale, Balaenoptera musculus 49 humpback whale, Megaptera novaeangliae 13 unidentified baleen whale 9 unidentified large whale 22 Small whales 48 unidentified beaked whale 7 mesoplodont beaked whale {Mesoplodon spp.) 5 Cuvier's beaked whale, Ziphius cavirostris 14 minke whale, Balaenoptera acutorostrata 4 unidentified small whale 11 unidentified cetacean 8 Fishery Bulletin 93(1). 1995 1993b). The basic equation for estimating abundance, N, for grouped animals with line transect is given by N- AnSf(O) 2Lg(0) (1) where A = size of the study area; n = number of sightings; S = mean group size; fiO) = sighting probability density at zero per- pendicular distance; L = length of transect line completed; and g(0) = probability of seeing a group directly on the trackline. Ideally, S, /10), and g(0) would be estimated sepa- rately for each species. However, the presence of mixed-species groups and small sample sizes required pooling for the estimation of/10) andg(O). The param- eter /(0) was estimated with the Hazard rate model (Buckland, 1985). This model was fitted by maximum likelihood with ungrouped perpendicular distances. Perpendicular distances were estimated from bearing and radial distance estimates made by observers. Pooling and stratification for estimating f (0) Pooled /T0)'s were estimated for five species groups: "small delphinids," "large delphinids," "small whales," "large whales," and "cryptic species." The five species groups were defined to include all of the species seen on the survey (Table 1) and were based on patterns of species cooccurrence in groups and on similarities in the physical and behavioral attributes that affect sightability from a ship. As an example, bottlenose dolphins, Tursiops truncatus, were never seen in a single-species group but were seen with Risso's dolphins, Grampus griseus, 13 times, with striped dolphins, Stenella coeruleoalba, one time, and with sperm whales, Physeter macrocephalus , three times. Bottlenose dolphins were pooled together with Risso's dolphins because they were seen most fre- quently with that species and because their sighting characteristics are more similar to Risso's dolphins (medium body size, prominent dorsal fin, occasional low puffy blow, small to medium group size) than to the other two species with which they were seen. Because killer whales, Orcinus orca, were never seen with other species but share the same sighting char- acteristics, these were also included in the species group "large delphinids." The other four groups are "small delphinids" which are of small body size (2-3 m) and are found in medium to large groups; "small whales" which are of medium body size (4-10 m), typically show no blow, often surface inconspicuously, and are typically found in small groups; "large whales" which are of large body size (10-30 m), al- most always show a conspicuous blow, and are found in small to medium groups; and "cryptic species" which are small (1.5—4.0 m), show no blow, typically surface inconspicuously, and are found in small groups. The assignment of higher-than-species taxa to species groups is given in Table 1. In estimating /10) for each species group, I explored stratification by two factors that are likely to affect sightability: sea state and group size. To avoid esti- mating more parameters than are justified by the data, I chose the most parsimonious stratification model by minimizing Akaike's Information Criterion (AIC) (Akaike, 1973), defined as 2 multiplied by the number of parameters used to estimate f\ 0) minus 2 multiplied by the sum of the log-likelihoods of the fitted values of/tO). Sea state was subjectively strati- fied into calm (Beaufort 0-2) and rough ( Beaufort 3- 5), based on the obvious degradation in sighting con- ditions that occurs with the presence of whitecaps at Beaufort 3. I stratified by group size by first finding the group size that divided the data into two samples with approximately the same number of sightings in each. If this stratification resulted in a lower AIC, I explored further stratification into three samples of approximately equal size. The above approach to stratification resulted in different strata for each species group. For small delphinids, AIC was minimized by stratifying group size into the categories 1-20, 21-100, and >100. For large delphinids, optimal stratification was with group size categories of 1-20 and >20. For large whales, AIC was minimized by using group size strata of 1-3 and >3. Because "cryptic species" and "small whales" were seldom seen in rough conditions, I estimated abundance for these species by using only data from calm conditions and did not explore strati- fication by sea state. Group size stratification re- sulted in higher AIC values for "cryptic species" and "small whales," so these groups were not stratified by group size. Sea-state stratification was not cho- sen on the basis of AIC values for any species group. In stratification by group size, estimates of den- sity in the various strata are added together to give an overall density. The equation for estimating abun- dance of each species k is therefore given by N t 1 7 = 1 An hk S hk f hk (0) 2Lg jk (0) (2) where A = size of study area; Barlow. Abundance of cetaceans in California waters: ship surveys l j,k J j-k = number of sightings of species k in group size stratum,/'; = mean group size of species k in group size stratum j; f k (0) - sighting probability density at zero per- pendicular distance for group size stra- tum./ of the species group to which spe- cies k belongs; L = length of transect line completed; and g k (0) = probability of detecting a group directly on the trackline for group size stratum j of the species group to which species k belongs. Perpendicular distance truncation Sightings of distant groups add little to the estima- tion of trackline density and can introduce bias. Buckland et al. (1993b) recommend truncating to eliminate at least the most distant 5% of all sightings. In the current study, groups of cetaceans were typi- cally not pursued for species identification and group size estimation if they were farther than 5.5 km (3 nmi) from the trackline. Therefore, by survey design, perpendicular distances must be truncated at no more than 5.5 km. I used a truncation distance of 3.7 km (2 nmi) for "small delphinids," "cryptic spe- cies," "large delphinids," and "small whales," which eliminated 8.8%, 2.4%, 4.6%, and 12.8% of all groups (respectively). A truncation distance of 5.5 km was used for "large whales," which eliminated 10.9% of groups. Group-size estimation The estimation of group size for cetaceans is diffi- cult and can lead to bias in the estimation of abun- dance. To avoid bias, correction factors were devel- oped for individual observers. The estimates of four of the six primary observers on the present survey had been previously calibrated by means of aerial photographic estimates to represent "true" group size. 4 The "best" estimates of two of these four were found to indicate group size with accuracy and did not require any correction factors. The other two re- quired correction factors, and, for one, correction fac- tors varied significantly from one year to the next. A helicopter was not available to make aerial photo- graphic estimates of group size on the present sur- vey, so correction factors for individual observers were estimated indirectly by comparison with the two 4 Gerrodette, T. D., and C. Perrin. 1991. Calibration of shipboard estimates of dolphin school size from aerial photographs. Admin. Rep. LJ-91-36, available from Southwest Fish. Sci. Cent., P.O. 271, La Jolla, CA 92038. 73 p. observers who, in the previous study, did not require correction. Linear regression was used to compare one obser- ver's estimates of group size to another's for the sub- set of groups that were estimated by both. Group sizes were log 10 -transformed to normalize variances. For the two observers who did not require a correction factor in the previous study, 4 the slope of the regres- sion was 1.009 (SE=0.017), indicating that, relative to each other, the observers were still estimating group size consistently. Correction factors for the other four observers were based on the slope and in- tercept of the regression of their "best" estimates against the mean of "best" estimates of the two who did not need calibration. The group size for each species in a group was es- timated as the average of all observers' corrected estimates of the size of the group multiplied by the average of all observers' estimates of the percentage of that species present (if in a mixed-species group). Probability of detecting trackline groups Estimating the probability that a group on the transect line will be seen, g(0), is fraught with diffi- culties (see Buckland et al. [1993b] for a review of previous attempts). In the context of bias from missed groups of marine mammals, it is useful to think in terms of the dichotomy proposed by Marsh and Sinclair ( 1989): bias can result from groups that were available to be seen but were not (perception bias) and from groups that were not available to be seen either because they did not surface or because they surfaced behind a swell (availability bias). I will make a minimum estimate of perception bias based on data collected by the conditionally independent observer and on the approach given in the Appendix. Because the sample of sightings made by independent observ- ers is small (only 37 cetacean groups), f 2 (0) in Equa- tion 7 was estimated for all cetaceans pooled with- out stratification by group size or sea state. Perpen- dicular distance data were fitted with the Hazard rate model to estimate f 2 (0). (Groups are only avail- able to the independent observer if they were missed by the other observers; therefore the distribution of perpendicular distances need not be monotonically decreasing. In this case, however, it was, and a more general model is not likely to have performed better than the Hazard rate model.) The analytical vari- ances of /"j(O) and f 2 (0) (from the information matrix method) were used in estimating the coefficient of variation ofg^O) from Equation 8, and the variances of n 1 and n 2 were estimated by assuming a Poisson distribution. Consideration of availability bias is deferred to the Discussion section. Fishery Bulletin 93(1), 1995 Coefficients of variation and confidence intervals Coefficients of variation (CV) and confidence inter- vals (CI) of the abundance estimates are based on the bootstrap method (Efron, 1977; Buckland et al., 1993b). The sightings associated with consecutive segments of search effort were combined to form a set of subsamples of 139 km (75 nmi) of search effort (corresponding to approximately one day of survey effort). 5 I drew subsamples randomly with replace- ment from this set of effort segments, and a pseudo- population size was estimated by using the same group size stratification as was used for the actual abundance estimates. For each bootstrap sample, the probability of detecting trackline groups, g(0), was estimated as a random number between and 1 drawn from the probability distribution of a bino- mial ratio with a mean and coefficient of variation equal to the estimated values. This process was re- peated 1,000 times, and the CV of the estimated population size was calculated as the standard error of the 1,000 pseudo-population sizes divided by the estimated population size. Bootstrap 95% confidence intervals on the population estimates were based on the 25th and 976th ranked estimates from the boot- strap samples. Log-normal 95% confidence intervals were based on the method given by Buckland et al. (1993b) and used the bootstrap estimate of CV. Results During the survey approximately 10,100 km of searching effort were completed (Fig. 1), and 515 cetacean groups were seen during the sampling ef- fort. Tracklines included 2,386 km in calm sea states (Beaufort 2 or less) and 7,696 km in rough sea states (Beaufort 3-5). During the survey, 18 cetacean spe- cies were identified (as well as at least one species that could only be identified to genus) (Table 1 ). More detailed data summaries for this survey are pre- sented by Hill and Barlow (1992), including the po- sitions and school sizes of all on- and off-effort sightings of cetaceans and pinnipeds, maps showing the distribution of sightings for each species, distri- butions of perpendicular distances for each species, patterns of association in mixed-species groups, sum- maries of searching effort completed under various conditions, and sighting rates of individual observ- ers. The fit of the probability density functions to 5 Barlow, J. 1993. The abundance of cetaceans in California wa- ters estimated from ship surveys in summer/fall 1991. Admin. Rep. LJ-93-09, available from Southwest Fish. Sci. Cent., P.O. Box 271, La Jolla, CA 92038, 39 p. the distributions of perpendicular distances are il- lustrated by Barlow. 5 Group-size estimation Group-size correction parameters, the slopes and intercepts (in parentheses) of log 10 -transformed re- gressions, were 0.922 (0.03), 1.022 (-0.03), 0.886 (0.07), and 0.777 (0.11) for the four observers who required correction. Three of these observers appear to have underestimated group size, in some cases by a large amount (a group of 500 would have been, on average, estimated as 328, 534, 283, and 152 by these four observers, respectively). Probability of detecting trackline groups Independent observers searched a total of 8,190 km. Approximately 7% of groups were detected only by the independent observer; however, all groups that were detected only by the independent observer were groups of less than 20 individuals and accounted for only 0.7% of the individuals that were seen on the survey. Of all groups that had less than 20 animals and were seen while the independent observer was on duty, 347 were seen by the primary observers, and 40 were seen by the independent observer. Abundance estimation With estimated values of/(0) and ^(0) (Table 2), den- sity and abundance were calculated for 19 cetacean species and 9 higher taxonomic categories (Table 3). Common dolphins were the most abundant cetaceans by a large margin. Of the two recently recognized common dolphin species (Heyning and Pen-in, 1994), the short-beaked variety was much more abundant than the long-beaked variety. Blue whales were the most abundant species of large whale. Discussion Distribution The distributions of cetaceans seen during this sur- vey (Figs. 2—6) are in general agreement with the results of other studies in this area (Leatherwood et al., 1982; Dohl et al., 1986; Smith et al., 1986; Barlow, 1988; Forney et al., this issue; Dohl et al. 2,3 ). How- ever, the observed distribution of some species con- tradicted results of previous studies. Striped dolphins were seen rather commonly in mixed groups with short-beaked common dolphins in southern and cen- tral California between 185 and 555 km (100-300 Barlow: Abundance of cetaceans in California waters: ship surveys Table 2 Estimated values of flO) andg(O) for each ( )f the species group stratifications which were chosen jn the basis of Akaike's [nforma- tion Criterion (AICl minimization. Truncation distances for estimating f[ 1 are 5.5 km for large whales and 3.7 km for all other species. Sample sizes include the total number of groups seen by the primary team, n, the number of groups seen by the primary team when an independent observer was on duty, n ,, and the number of groups seen by the independent observers but not by the primary team, n. 2 . NA indicates information that is not available because it could not be estimated. CV is the coefficient of variation. Primary observers Secondary observers Primary observers Number of si ghtings flO) CV f{0) CV CV Main stratum and substrata n "i n 2 km' 1 flO) km- 1 A0) giO) giO) Small delphinids (3.7 km truncation) group size 1-20 67 58 9 1.258 0.249 1.864 0.147 0.770 0.137 group size 21-100 58 51 0.944 0.336 1.864 0.147 1.000 NA group size 101+ 47 44 0.283 0.193 1.864 0.147 1.000 NA Cryptic species (3.7 km truncation) calm seas 102 78 14 1.574 0.199 1.864 0.147 0.787 0.103 Large delphinids (3.7 km truncation) group size 1-20 15 14 1 0.504 0.306 1.864 0.147 0.736 0.391 group size 21+ 17 17 0.352 NA 1.864 0.147 1.000 NA Large whales (5.5 km truncation) group size 1-3 87 81 3 0.696 0.278 1.863 0.146 0.901 0.073 group size 4+ 26 22 0.256 NA 1.863 0.146 1.000 NA Small whales (3.7 km truncation! calm seas 23 19 1 0.614 0.488 1.864 0.147 0.840 0.218 nmi) from shore. Although striped dolphins were known to inhabit this area (Leather- wood et al., 1982), their frequency of occur- rence was much greater than expected. Blue whales were seen primarily in southern Cali- fornia between 92 and 370 km (50-200 nmi) offshore. In previous years, this species was seen commonly in central California between the coast and 92 km (50 nmi) offshore (Calambokidis et al., 1990b). One species was surprising in its absence: short-finned pilot whales, Globicephala macrorhynchus, were previously common in southern California, es- pecially around the Channel Islands in winter (Leatherwood et al., 1982). (Note: one group of pilot whales was seen and photographed by independent researchers between San Fran- cisco and Monterey on 2 November 1991. 6 ) Abundance Abundance estimates from this study are also in general agreement with previous esti- 6 Jones, P. A, and I. D. Szczepaniak. 1992. Report on the seabird and marine mammal censuses conducted for the long-term management strategy (LTMS), August 1990 through November 1991, for the U.S. Environmental Protection Agency, Region IX, San Francisco. July 1992. 42' : C*pe Mendocino 40' \ ® X X 38' X**® «* I S«n Francisco X o ) u ® # x V 36' \ CO 8 » * * j Pcfci!Conc*pCfcin 34' PACIFIC X X CO 9 >-* * * 1. A & X It a x „ x * x x x CP» 8 x x 9* 30' 132" 130' 128' 126' 124' 122" 120' 118" Longitude Figure 2 Locations of on-effort sightings of short-beaked common dolphins (x), long-beaked common dolphins (O), unidentified common dolphins (A), and striped dolphins ( ). Scientific names are given in Table 1. Fishery Bulletin 93(1). 1995 Table 3 Number of groups seen (n ), mean group size (S), density of individuals , abunda rice estimates (N), 95% confidence intervals (CI) on those estimates, and coefficients of variation (CV) for al species and higher taxa that were identified. Density estimates are based on lengths of transect given in the text and estimates of/tO) andg(O) given in Table 2. Mean group size includes only the indicated species and can therefore be less than the minimum of the group size category (which is defined based on the total number of all species present). Scientific names are given in Table 1. Number Mean Animal Pop. Boot strap Log-normal Lower Upper Lower Upper of groups group size density size 95% 95% 95% 95% Species strata n S km 2 N CV CI CI CI CI Small delphinids short-beaked common dolphin 3.248 225,821 0.279 143,026 419,911 132,139 385,918 group size 1-20 25 11.0 0.261 group size 21-100 52 44.7 1.274 group size 101 + 39 267.3 1.713 long-beaked common dolphin 0.136 9,472 0.683 27,029 2,817 31,842 group size 1-20 1 11.8 0.011 group size 21-100 0.0 0.000 group size 101+ 4 190.2 0.125 common dolphin (unclassified) 0.148 10,286 0.815 573 37,007 2,539 41,664 group size 1-20 6 5.4 0.031 group size 21-100 1 15.1 0.008 group size 101 + 1 661.5 0.109 striped dolphin 0.273 19,008 0.412 8,234 45,864 8,755 41,267 group size 1-20 2 7.7 0.015 group size 21-100 5 29.3 0.080 group size 101 + 14 77.6 0.178 Pacific white-sided dolphin 0.177 12,310 0.537 1,888 27,965 4,590 33,010 group size 1-20 7 11.5 0.076 group size 21-100 3 46.2 0.076 group size 101 + 2 75.4 0.025 northern right whale dolphin 0.134 9,342 0.567 2,125 21,488 3,322 26,272 group size 1-20 10 9.9 0.094 group size 21-100 3 9.4 0.015 group size 101 + 2 75.7 0.025 unidentified delphinoid 0.052 3,603 0.462 1,180 6,197 1,521 8,536 group size 1-20 17 3.2 0.052 group size 21-100 0.0 0.000 group size 101+ 0.0 0.000 Cryptic species harbor porpoise' 31 5.0 0.758 52,743 0.682 147,905 15,714 177,026 Dall's porpoise 69 3.3 1.127 78,422 0.354 33,462 150,487 40,026 153,649 pygmy sperm whale 2 1.3 0.013 870 0.796 2,741 220 3,433 Large delphinids bottlenose dolphin 0.022 1,503 0.481 499 3,819 615 3,674 group size 1-20 4 2.8 0.004 group size 21+ 10 8.3 0.017 Risso's dolphin 0.122 8,496 0.415 4,236 21,676 3,890 18,555 group size 1-20 12 8.3 0.039 group size 21+ 16 25.2 0.082 killer whale 0.004 307 1.196 2,340 48 1,947 group size 1-20 3 3.7 0.004 group size 21+ 0.0 0.000 Large whales sperm whale 0.011 756 0.493 211 1,537 303 1,886 group size 1-3 4 1.2 0.002 group size 4+ 9 6.6 0.009 Barlow: Abundance of cetaceans in California waters ship surveys Table 3 (Continued) Number Mean Animal Pop. Boot strap Log-normal Lower Upper Lower Upper of groups group size density size 95% 95% 95% 95% Species strata n S km- 2 N CV CI CI CI CI Baird's beaked whale 0.001 38 1.025 127 7 203 group size 1-3 0.0 0.000 group size 4+ 1 3.7 0.001 Bryde's whale 0.001 61 1.078 242 11 339 group size 1-3 1 1.9 0.001 group size 4+ 0.0 0.000 Bryde's or sei whale 0.001 63 1.093 232 11 355 group size 1-3 2 1.0 0.001 group size 4+ 0.0 0.000 fin whale 0.013 935 0.635 130 2,607 299 2,925 group size 1-3 17 1.4 0.011 group size 4+ 4 4.7 0.003 blue whale 0.033 2,250 0.381 899 4,131 1,093 4,632 group size 1-3 36 1.6 0.026 group size 4+ 13 3.3 0.007 humpback whale 0.009 626 0.411 196 1,133 289 1,359 group size 1-3 7 1.8 0.006 group size 4+ 3 7.3 0.003 unidentified baleen whale 0.003 214 0.631 26 530 69 665 group size 1-3 5 1.2 0.003 group size 4+ 1 2.1 0.001 unidentified large whale 0.009 629 0.470 167 1,306 262 1,508 group size 1-3 15 1.3 0.009 group size 4+ 0.0 0.000 Small whales unidentified beaked whale 3 3.5 0.019 1,322 0.892 4,541 295 5,921 mesoplodont beaked whale 2 1.0 0.004 250 0.834 746 60 1,040 Cuvier's beaked whale 7 1.9 0.023 1,621 0.823 186 5,555 396 6,637 minke whale 4 1.1 0.008 526 0.971 2,244 106 2,596 unidentified small whale 5 1.0 0.009 645 0.767 127 2,061 170 2,446 unidentified cetacean 3 1.7 0.009 620 0.879 2,026 141 2,731 ' More precise estimates for harbor porpoise are recently available in Barlow and Forney (1994). mates (Dohl et al., 1986; Barlow, 1988; Calambokidis et al., 1990a; Dohl et al. 23 ). This is the first cetacean survey in California waters to include the region between 277 and 555 km (150-300 nmi) offshore. The studies of Dohl et al. 2 - 3 included only the inshore 185 km (100 nmi) of the present study area, making di- rect abundance comparisons difficult. The mark-re- capture population estimates of blue and humpback whales by Calambokidis et al. (1990, a and b) were based on individuals sighted near the coast. Further- more, the estimates of Dohl et al. 2 - 3 do not have as- sociated statistical confidence intervals. Hence, ac- curate comparisons with previous studies can be made only for the more coastal species and mean- ingful statistical tests of differences can be made for even fewer species. Direct comparisons with the 1991 and 1992 aerial surveys (Forney et al., this issue) are planned for future publications. The abundance of harbor porpoise estimated for 1984 and 1985 was approximately 9,576 (CV=0.51) (Barlow [1988] his regions 1-4), which is smaller than the present estimate of 52,700 (CV=0.68). This dis- crepancy may be due to the inappropriate design of the present survey for a coastal species such as har- bor porpoise. Humpback whale abundance in central California was estimated as 338 based on aerial surveys from Au- gust to November of 1980-83 (Dohl et al. 3 ); however, this estimate does not include a correction factor for submerged whales. Based on mark-recapture methods, the abundance of humpback whales in 1991 and 1992 was estimated to be 581 (CV=0.03). 1 This estimate is Fishery Bulletin 93(1). 1995 42' x 3"««x \ 1 40' ^i» f Cape Mendocino 38' x X o SrV *%\^ft»*w Latitude X J *] Point Conception 34' A 4 ^ -. Loe ^^V__Ari9eto6 PACIFIC OCEAN V 1 32' \ A 30' *■* 132' 130' 128' 126' 124' 122' 120' 118' Longitude Figure 3 Locations of on-effort sightings of Dall's porpoise (x ), northern right whale dolphins ( ), and Pacific white-sided dolphins (O). Scientific names are given in Table 1. 42 O *1 1 a, ( Cape Mendocino 40' A \ \ ° ) 38' & *-A H V\ \ San Francisco N, Latitude CO a r A\ \ ' *\ 34' | Point Conception O A -oS-X-*"* 01 " PACIFIC a '*-& \ OCEAN * * \ 32' AA 30' ' 132' 130' 128' 126' 124' 122' 120' 118' Longitude Figure 4 Locations of on-effort sightings of killer whales (O), Risso's dolphins (A), and bottlenose dolphins (x). Scientific names are given in Table 1. very close to the present estimate of 626 and is well within its 95% confidence interval. For two species, new estimates of abun- dance appear to be substantially different from previous estimates. For the late 1970's, the combined summer and fall estimate of common dolphin abundance was 57,270 (CV=0.17) (Dohl et al., 1986). Although the methods used were very different and the area surveyed was smaller in that study, es- timates for other small cetaceans are simi- lar in the two studies. A large increase in common dolphin abundance is likely. This could have resulted as an effect of the 1991— 92 El Nino. Although there were no surface temperature manifestations of El Nino in the study area at the time of the survey, it is pos- sible that common dolphins were moving into California waters from farther south as a result of El Nino changes there. Since 1980, a decline has been noted in the abundance of the northern stock of common dolphins south of 30°N (Anganuzzi et al., 1993), and those authors hypothesize that this could have been caused by a general northward move- ment of that stock. This interpretation is con- sistent with the increases noted here, but the magnitude of the decrease in the south (from approximately 500,000 in 1980 to approxi- mately 100,000 in 1991 [Anganuzzi et al. 1993]) is greater than the entire estimated population in California waters. The abundance of blue whales, based on the current line-transect data (2,250), is also much higher than recent estimates made from individual-identification mark-recap- ture techniques (904 based on left-side pho- tographs and 1,112 based on right-side pho- tographs). 1 Although some mark-recapture estimates may be biased low because of geo- graphic heterogeneity in habitat use by indi- vidual whales (Hammond, 1990), the meth- ods used for mark-recapture should have minimized those effects. 1 South of the present study area, the abundance of blue whales was estimated to be 1,415 (CV=0.24) based on line- transect ship surveys in the eastern tropical Pacific from 1986 to 1990 (Wade and Ger- rodette, 1993). The latter study included sightings made along the coast of Baja Califor- nia (which probably belong to the California feeding population) as well as sightings made near the Costa Rica Dome and along the Equa- tor (which are likely to be part of a different population; Reilly and Thayer [1990]). Barlow: Abundance of cetaceans in California waters, ship surveys I I 42 ' ° 1 Of Cape Mendocino 40 38' x \ San Francisco N Latitude 5> Ox |> A 0\ o \ 34' *^\ Point Conception >yO *^ *— -. LO. PACIFIC ©x^x,, ^ , J%. OCEAN #*##* • \ X X ^ 32" 30 x x * o x ^ x 132' 130' 128' 126' 124' 122' 120' 118' Longitude Figure 5 Locations of on-effort sightings of fin whales ( ), humpback whales (O), blue whales (x), and sperm whales ( ). Scientific names are given in Table 1. Probability of detecting trackline groups The probability of detecting a trackline group of ani- mals,^), varied between 0.74 and 1.0 (Table 2 ). The data clearly indicated that small groups are much more likely to be missed than are large groups. This is intuitively obvious and justifies stratifying by group size when estimating ^(0) values. The fraction of trackline harbor porpoise seen in calm seas has been estimated previously to be 0.78 (with five ob- servers on a similar platform in California, Barlow [1988] and 0.70 (with six observers in the Gulf of Maine, Palka [1993]). The higher value ofg(0) esti- mated here for "cryptic species" with only three ob- servers (0.81) may be due to the inclusion of Dall's porpoise which may be easier to see or may simply be an artifact of small sample size. These estimates of the fraction of animals seen include only animals that were available to be seen. Availability bias is likely to be large for species such as beaked whales, which have extremely long dive times, and harbor porpoise and Dall's porpoise, which have shorter dive times but seldom are seen more than 0.5 km from the ship and may therefore remain submerged during the entire time they are within visual range. Correcting for availability bias is more difficult than for perception bias. Attempts that have been made so far have involved detailed modeling of the surfacing behavior of the animal and the searching behavior of the researchers (Doi, 1971, 1974; Barlow et al., 1988; Stern, 1992; Kasamatsu and Joyce 7 ). In addition, there are still problems with estimating perception bias because the methods used here assume that all animals are equally available to be seen if they sur- face. Heterogeneity in sightability (e.g. ani- mals that splash vs. animals that do not) gen- erally will result in an underestimate of the fraction missed. Additional work is needed to obtain complete estimates of the fraction of trackline animals seen for all species. Previous studies of Dall's porpoise have shown that attraction to the vessel is a greater problem for estimating the abun- dance of this species than are missing trackline animals (Turnock and Boucher 8 ). Turnock and Quinn (1991) estimated a cor- rection factor of 0.2378 (CV=0.3391) to ad- just Dall's porpoise abundance estimates for ship surveys (effectively then, g =4.2). That study was based, however, on a design that used only one observer who searched with 7x binoculars and unaided eyes. In the present study, very few Dall's porpoise ap- peared to be attracted to the vessel; of those sighted in calm conditions and used for abundance estimation, only 10% (9 of 88) of the Dall's porpoise groups approached the vessel to "ride the bow wave," and 89% (78 of 88) were exhibiting a "slow roll" sur- facing behavior at the time they were first sighted. Because attraction to the vessel was less than in other studies and because most Dall's porpoise were sighted before showing any apparent reaction to the vessel (perhaps because 25x binoculars were used), the magnitude of bias is probably less than that es- timated by Turnock and Quinn ( 1991). Statistical precision An attempt was made to account for most sources of sampling error in the bootstrap estimates of confi- dence intervals and coefficients of variation. How- ever, several sources of variation could not be easily included. The process of selecting a stratification 7 Kasamatsu, F., and G. G. Joyce. 1991. Abundance of beaked whales in the Antarctic. Int. Whaling Comm. working paper SC/43/012. 8 Turnock, B. J., and G. C. Boucher. 1990. Population abundance of Dall's porpoise, Phocoenoides dalli, in the western North Pacific Ocean. Int. Whaling Comm. working paper SC/42/SM10. 12 Fishery Bulletin 93(1), 1995 40" 38 ■£ 36 ' 34" 30' 132' 120' 118' Longitude Figure 6 Locations of on-effort sightings of beaked whales of the genus Mesoplo- don (...), Cuvier's beaked whales (O), Baird's beaked whales ( ), and unidentified beaked whales (x). Scientific names are given in Table 1. model by minimizing AIC would have been too time consuming to include in the bootstrap procedure; hence, precision estimates are contingent on the cho- sen models being approximately correct. Variability in estimating mean group size was included implic- itly in the Monte Carlo sampling, but it was assumed that the group size estimate for any given group was accurate. Pooling of data to estimate /10) and g(0) introduces a bias (to the extent that individuals dif- fer within a pooled group) which is not accounted for in precision estimates. All of these factors would tend to result in precision being overestimated. Overall, coefficients of variation are likely to be too small and true confidence intervals are probably wider than those reported. Acknowledgments This survey could not have been accomplished with- out the diligent work of many people, including the officers and crew of the RV McArthur. S. Hill served as cruise coordinator. The six primary observers were W. Armstrong, S. Benson, J. Cotton, D. Everhardt, M. Lycan, and R. Mellon. The independent observ- ers included E. Archer, K. Forney, S. Hill, S. Kruse, M. Lowry, V. Philbrick, B. Taylor, and P. Wade (and J. B.). The ship-board data log- ging software was written by J. Cubbage (and J. B.). Observer training was provided by S. Hill, A. Jackson, W. Perryman, and R. Pit- man. Data were edited and archived by A. Jackson and K. Wallace. Sighting distribu- tions were plotted with software written by T Gerrodette. The survey design was im- proved by thoughtful suggestions from T Gerrodette and D. DeMaster. This manu- script was improved by helpful suggestions from S. Buckland, K. Burnham, J. Calam- bokidis, J. Carretta, K. Forney, T Gerrodette, J. Laake, R. Brownell, P. Wade, and two anonymous reviewers. Literature cited Akaike, H. 1973. Information theory and an extension of the maximum likelihood principle. In B. N. Petran and F. Csaaki (eds.), International symposium on information theory, 2nd ed., 1451 p. Akadeemiai Kiadi, Budapest, Hungary. Anganuzzi, A. A., S. T. Buckland, and K. L. Cattanach. 1993. Relative abundance of dolphins associated with tuna in the eastern Pacific Ocean: analysis of 1991 data. Rep. Int. Whaling Comm. 43:459^165. Barlow, J. 1988. Harbor porpoise (Phocoena phocoena ) abundance es- timation in California, Oregon and Washington: I. Ship surveys. Fish. Bull. 86:417-432. Barlow, J., and K. A. Forney. 1994. An assessment of the 1994 status of harbor porpoise in California. U.S. Dep. Commer., NOAA Tech. Memo. NMFS. NOAA-TM-NMFS-SWFSC-205, 17 p. Barlow, J., and T. Lee. 1994. The estimation of perpendicular sighting distance on SWFSC research vessel surveys for cetaceans: 1974 to 1991. U.S. Dep. Commer., NOAA Tech. Memo. NMFS. NOAA- TM-NMFS-SWFSC-207, 46 p. Barlow, J., C. Oliver, T. D. Jackson, and B. L. Taylor. 1988. Harbor porpoise (Phocoena phocoena ) abundance es- timation in California, Oregon and Washington: II. Aerial surveys. Fish. Bull. 86:433-444. Buckland, S. T. 1985. Perpendicular distance models for line transect sampling. Biometrics 41:177-195. Buckland, S. T., J. M. Breiwick, K. L. Cattanach, and J. L. Laake. 1993a. Estimated population size of the California gray whale. Mar. Mamm. Sci. 9(3):235-249. Buckland, S. T., D. R. Anderson, K. P. Burnham, and J. L. Laake. 1993b. Distance sampling: estimating abundance of biologi- cal populations. Chapman and Hall, London, 446 p. Barlow: Abundance of cetaceans in California waters: ship surveys Burnham, K. P., D. R. Anderson, and J. L. Laake. 1980. Estimation of density from line transect sampling of biological populations. Wildl. Monogr. 72, 202 p. Calambokidis, J., J. C. Cubbage, G. H. Steiger, K. C. Balcomb, P. Bloedel. 1990a. Examination of population estimates of humpback whales in the Gulf of the Farallones, California. Rep. Int. Whaling Comm., Special Issue 12:325-333. Calambokidis, J., G. H. Steiger, J. C. Cubbage, K. C. Balcomb, C. Ewald, S. Kruse, R. Wells, and R. Sears. 1990b. Sightings and movements of blue whales off cen- tral California 1986-88 from photo-identification of individuals. Rep. Int. Whaling Comm., Special Issue 12: 343-348. Dohl, T. P., M. L. Bonnell, and R. G. Ford. 1986. Distribution and abundance of common dolphin, Del- phinus delphis, in the Southern California Bight: a quan- titative assessment based on aerial transect data. Fish. Bull. 84:333-343. Doi, T. 1971. Further development of sighting theory on whales. Bull. Tokai Reg. Fish. Res. Lab. 68:1-22. 1974. Further development of whale sighting theory. In W. E. Schevill (ed. ), The whale problem: a status report, p. 359-368. Harvard Univ. Press, Cambridge, MA. Efron, B. 1977. Bootstrap methods: another look at the jack- knife. Ann. Statistics 7:1-26. Forney, K. A., J. Barlow, and J. Carretta. 1995. The abundance of cetaceans in California waters. Part II: Aerial surveys in winter and spring of 1991 and 1992. Fish. Bull. 93:15-26. Hammond, P. S. 1990. Heterogeneity in the Gulf of Maine? Estimating humpback whale population size when capture probabili- ties are not equal. Rep. Int. Whaling Comm., Special Is- sue 12:135-140. Heyning, J. E., and W. F. Perrin. 1994. Evidence for two species of common dolphins (genus Delphinus ) from the eastern North Pacific. Contrib. Nat. Hist. Mus. Los Angeles Co. 442, 35 p. Hill, P. S., and J. Barlow. 1992. Report of a marine mammal survey of the California coast aboard the research vessel McArthur July 28- November 5, 1991. U.S. Dep. Commer, NOAA Tech. Memo. NOAA-TM-NMFS-SWFSC-169, 103 p. Holt, R. S. 1987. Estimating density of dolphin schools in the eastern tropical Pacific Ocean using line transect methods. Fish. Bull. 85:419^134. Holt, R. S., and J. E. Powers. 1982. Abundance estimation of dolphin stocks involved in the eastern tropical Pacific yellowfin tuna fishery deter- mined from aerial and ship surveys to 1979. U.S. Dep. Commer., NOAA Tech. Memo. NOAA-TM-NMFS-SWFC- 23, 95 p. Holt, R. S., and S. N. Sexton. 1989. Monitoring trends in dolphin abundance in the east- ern tropical Pacific using research vessels over a long sam- pling period: analyses of 1986 data, the first year. Fish. Bull. 88:105-111. Leatherwood, S., R. R. Reeves, W. F. Perrin, and W. E. Evans. 1982. Whales, dolphins, and porpoises of the eastern North Pacific and adjacent Arctic waters: a guide to their identification. U.S. Dep. Commer., NOAA Tech. Rep. NMFS Circular 444, 245 p. Lennert, C, S. Kruse, M. Beeson, and J. Barlow. In press. Incidental marine mammal bycatch in California gillnet fisheries. Rep. Int. Whaling Comm., Special Issue. Marsh, H., and D. F. Sinclair. 1989. Correcting for visibility bias in strip transect aerial surveys of aquatic fauna. J. Wildl. Manage. 53:1017-1024. Palka, D. 1993. Estimates of g(0) for harbor porpoise groups found in the Gulf of Maine in August 1991. Ph.D. diss., Univ. California, San Diego. Reilly, S. B. 1984. Assessing gray whale abundance: a review. In M. L. Jones, S. L. Swartz, and J. S. Leatherwood (eds.), The gray whale, Eschriehtius robustus. Acad. Press, 624 p. Reilly, S. B., and V. G. Thayer. 1990. Blue whale {Balaenoptera musculus) distribution in the eastern tropical Pacific. Mar. Mamm. Sci. 6(4):265-277. Smitb, R. C, P. Dustan, D. Au, K. S. Baker, and E. A. Dunlap. 1986. Distribution of cetaceans and sea surface chlorophyll concentrations in the California Current. Mar. Biol. 91:385-402. Smith, T. D. 1982. Testing methods of estimating range and bearing to cetaceans aboard the RVD.S. Jordan. U.S. Dep. Commer., NOAA Tech. Memo. NOAA-TM-NMFS-SWFC-20 [avail, from National Tech. Information Serv., Springfield, VA 22161], 30 p. Stern, S. J. 1992. Surfacing rates and surfacing patterns of minke whales (Balaenoptera aeutorostrata ) off central California, and the probability of a whale surfacing within the visual range. Rep. Int. Whaling Comm. 42:379-386. Turnock, B. J., and T. J. Quinn II. 1991. The effect of responsive movement on abundance es- timation using line transect sampling. Biometrics 47:701- 715. Wade, P. R., and T. Gerrodette. 1993. Estimates of cetacean abundance and distribution in the eastern tropical Pacific. Rep. Int. Whaling Comm. 43: 477-494. 14 Fishery Bulletin 93fl), 1995 Appendix "S =n m f(0)5. (5) To estimate the total fraction of trackline groups missed owing to perception bias requires that the survey be designed with two teams of completely in- dependent observers. To be independent, both teams would have to search simultaneously, not notifying or cueing each other until a group of animals had passed abeam of the vessel and were clearly missed by the other team. This approach was deemed infea- sible because of the need to approach groups to esti- mate group size and species composition. If the ves- sel was not turned until after all groups had passed abeam, a very large percentage of those groups would not be relocated. The probability of relocation would depend on group size and species composition. These factors would add considerably to the difficulty in interpreting such survey data. Instead, the survey was designed to use a single, conditionally independent observer who was aware of sightings made by the primary team, but who did not reveal the presence of a group until that group was clearly missed by the primary team. Data from the conditionally independent observer are used to make an estimate of the probability that the primary survey team detected a trackline group. The expected number of groups, n, seen very close to the transect line, say within distance 8, can be estimated as n a g(x)h(x)dx ' (3) n 5 = g(x)h(x)dx where n m is the total number of groups seen within the truncation distance co, g(x) is the probability of seeing a group that is at perpendicular distance x, and h(x) is the probability that a group will be at perpendicular distance x (usually assumed to be 1.0 for primary observers at all x). As 5 approaches zero distance, the above equation can be reexpressed as n s n a g(0)h{0)8 g{x)h(x)dx (4) which, from the line-transect definition of /TO) (Burnham et al., 1980), can be simplified to The probability of a trackline group being seen by the primary observers can be expressed as £l(0)= '1,S "is + n 2S I g 2 (0) (6) where the subscript 1 refers to sightings made by the primary observers and subscript 2 refers to sightings missed by the primary observers but seen by the independent observer. Combining Equations 5 and 6 and simplifying results in Sl(0): l lca A<0) n lol k(0) + n 2l J 2 (0)/g 2 (0) (7) Because there were three primary observers and only one independent observer, g x (0) should be greater than or equal tog 2 (0). Thus Si(0) fi(0) (9) (10) CV 2 (n 1 J + CV 2 (n 2(O ) + CV 2 {f 1 (0)) + CV 2 {k(0)). (11) Abstract. Two aerial line- transect censuses of cetaceans were conducted along the California coast during March-April 1991 and February-April 1992. The two sur- veys were designed to provide a combined estimate of cetacean abundance for winter and spring (cold-water) conditions; they com- plemented a summer and fall ship survey in 1991. The study area (264,270 km 2 ) extended about 278 km ( 150 nmi ) off the coast of south- ern California, and 185 km (100 nmi) off the coast of central and northern California. A primary team of two observers searched for cetacean species through bubble windows that allowed an unob- structed view to the sides and di- rectly beneath the aircraft. A third, conditionally independent observer searched through a belly window and reported animals that were missed by the primary team. Ap- proximately 7,069 km and 5,973 km were searched in 1991 and 1992, respectively, resulting in 253 sightings of at least 18 cetacean species (some animals could only be identified to higher taxa). Esti- mates of abundance and coeffi- cients of variation (in parentheses) for the most common small ceta- ceans are the following: 306,000 (0.34) common dolphins, Delphinus spp.; 122,000 (0.47) Pacific white- sided dolphins, Lagenorhynchus obliquidens; 32,400 (0.46) Risso's dolphins, Grampus griseus; and 21,300 (0.43) northern right whale dolphins, Lissodelphis borealis. Abundance estimates (and CV's) for the most common whales are the following: 892 (0.99) sperm whales, Physeter macrocephalus; 392 (0.41) beaked whales, genera Meso- plodon and Ziphius; 319 (0.41) humpback whales, Megaptera novae- angliae; and 73 (0.62) minke whales, Balaenoptera acutorostrata. The abundance of cetaceans in California waters. Part II: Aerial surveys in winter and spring of 1991 and 1992 Karin A. Forney Jay Barlow James V. Carretta Southwest Fisheries Science Center National Marine Fisheries Service. NOAA RO. Box 271, La Jolla, California 92038 Manuscript accepted 31 May 1994. Fishery Bulletin 93:15-26 (1995). California coastal waters are a pro- ductive and highly variable oceano- graphic region with a diverse ma- rine fauna. Coastal fisheries, prima- rily gillnet fisheries, cause the inci- dental death of a variety of marine mammal species (Barlow et al., in press). However, the impact of this mortality can only be evaluated if estimates of population size are available for the affected species. In the late 1970's and early 1980's, abundance estimates were obtained based on aerial surveys, 1 - 2 but esti- mates of precision were not ob- tained for most species. Because of the age and uncertainty of these es- timates, the National Marine Fish- eries Service conducted aerial and shipboard surveys during 1991 and 1992. Based on evidence of season- ality in the abundance and distri- bution of some cetaceans (Leather- wood and Walker, 1979; Dohl et al., 1986), separate abundance esti- mates were obtained for winter and summer conditions. Two aerial sur- veys (March-April 1991 and Febru- ary-April 1992) were completed during cold-water conditions, and one ship survey (July-November 1991) was conducted during warm- water conditions (Barlow, this is- sue). The survey periods were cho- sen based on climatic atlases of the California coast which show that, on average, March and April have the coldest, and September and October the warmest sea-surface tempera- tures (U.S. Navy, 1977). Standard line-transect methods (Burnham et al., 1980; Buckland et al., 1993a) were used from both platforms. Pre- liminary abundance estimates were calculated after completion of the first aerial survey in 1991 (Forney and Barlow, 1993), but confidence limits were large. In this paper, we present combined abundance estimates for the 1991 and 1992 aerial surveys. Survey methods The methods used during the 1991- 92 aerial surveys are described in detail by Forney and Barlow (1993) and Carretta and Forney (1993), and only a summary is presented below. The study area (264,270 km 2 ) 1 Dohl, T. P., K. S. Norris, R. C. Guess, J. D. Bryant, and M. W. Honig. 1978. Cetacea of the Southern California Bight. Part II of Summary of marine mammal and sea- bird surveys of the Southern California Bight area, 1975-1978. Final Report to the Bureau of Land Management, 414 p. [NTIS Rep. PB81248189.] 2 Dohl, T. P., R. C. Guess, M. L. Duman, and R. C. Helm. 1983. Cetaceans of central and northern California, 1980-1983: status, abundance and distribution. OCS Study MMS 84-0045. Minerals Management Ser- vice contract No. 14-12-0001-29090, 284 p. 15 16 Fishery Bulletin 93(1). 1995 encompasses California waters out to a distance of 185-278 km (100-150 nmi) from the coast and roughly a depth of 3,000-4,000 m (Fig. 1). It was defined on the basis of the distribution of fisheries that are known to take marine mammals and does not reflect a distributional boundary for any marine mammal population. Surveys were conducted along transect lines forming two nearly uniform, overlap- ping grids (Fig. 1). The resulting overall grid lines were spaced 41^46 km (22-25 nmi) apart. The loca- tion of the transect grid was chosen without refer- ence to specific areas or topographical features. To avoid potential differences in regional coverage, an attempt was made in each year to complete all transects of the first grid, providing coarse coverage of the entire study area, before beginning the second grid. However, in both years, poor weather conditions prevented the completion of both survey grids. In 1991, 85% (5,326 km) of transect grid 1 and 27% (1,739 km) of grid 2 were completed, and in 1992, jjrto 41°- - / / f — 1 Cape Mendocino 40°- 39°- .., V__ / 7\ San Francisco "'-. 38°- \""/~— -— V ^-J? '- \L_/ ~~7V California \ 37°- Latitude CO 1 35°- \7 / 7\Ft Conception 34°- \ / ^ ^r^"^ 7 P\ Los Angeles Pacific ^^\r~^~~L 7 ^^X 33°- Ocean X. 7~^~7^-----^' / -ZL 32°- 31°- Tn° JU | i | | i | | i | ' | i | ' | 127° 126° 125° 124° 123° 122° 121° 120° 119° 118° 11 7° Longitude Figure 1 Study area with two overlapping transects grids. The solid line represents grid 1, the dotted line grid 2. 81% (5,065 km) of transect grid 1 and 14% (890 km) of grid 2 were completed. The relative proportions of survey effort in different sea state and cloud cover conditions were similar for the two years (Table 1). The survey platform was a twin-engine turbo-prop DeHavilland Twin Otter, flown approximately at an altitude of 213 m (700 ft) and an airspeed of 165-185 km/h (90-100 knots). All cetacean and sea turtle sightings were recorded, but because of the high den- sities of pinnipeds near rookeries, these species were recorded only when seen farther than 10 km from land. Two "primary" observers searched through bubble windows on the left and right sides of the air- craft. These windows allowed observers to view to the side and directly beneath the aircraft with at least 10° of overlap between sides. To achieve higher sight- ing efficiency near the transect line, observers searched for cetaceans only out to a declination angle of 12° (1,004 m perpendicular distance). An addi- tional "secondary" observer monitored the trackline area out to 55° declination angles (on both sides) through a round 45-cm ( 18- in) viewing hole in the belly of the air- craft and reported sightings missed by the primary team. A fourth person re- corded all sighting, effort, and environ- mental data. To minimize observer fa- tigue, all observers rotated between these four active positions and one resting position roughly every 30 min- utes. All observers had previous experi- ence in identifying cetacean species from aerial or shipboard platforms, or both. All survey data were recorded on a laptop computer connected to a LORAN or GPS (Global Positioning System) navigational receiver, providing a con- tinuous record of position (updated every few seconds), altitude, air speed, and survey conditions. Environmental conditions, such as Beaufort sea state, percent cloud cover, and glare, were updated whenever changes occurred. Conversation in the aircraft was re- corded on a central cassette recorder as a backup to the computer record. Observers also recorded individual sighting information into personal notebooks. Surveys were conducted only in Beaufort sea states 0-4. Following the methods described in Forney and Barlow ( 1993) and Carretta and Forney (1993), the aircraft circled for each sighting to obtain species iden- tifications and school size estimates Forney et al.: Abundance of cetaceans in California waters: aerial surveys 17 Table 1 Survey effort (in km) stratified by sea state and percent cloud cover. Beaufort sea state Cloud cover and 1 2 Total 1991 0-24 25-49 50-74 75-100 Total 1992 0-24 25-49 50-74 75-100 Total 212 26 45 76 359 406 2 78 913 66 58 129 1,166 933 8 43 251 486 1,235 1,932 96 331 980 3,338 1,349 141 192 758 2,440 1,346 85 241 532 4,403 273 676 1716 2,205 7,069 1,220 113 47 433 1,813 3,908 262 284 1,519 5,973 Both years combined 0-24 25-49 50-74 75-100 Total 618 26 47 154 845 1,846 74 101 380 2,401 3,280 2,566 8,311 238 199 536 523 288 960 1,737 965 3,235 5,778 4,018 13,042 (each observer made a confidential record of best, high, and low estimate into a personal field notebook). Any additional schools sighted while the aircraft was di- verted from the transect were recorded as 'off-effort' sightings. Only sightings made during active searches on predetermined transect lines Con-effort') were in- cluded for abundance estimation. The secondary obser- ver only reported sightings missed by the primary observer team; these secondary sightings were used to estimate the fraction of animals missed on the transect line. Analytical methods Stratification Because we were not able to complete both grids in all regions of the coast, the study area was divided into four a posteriori geographic areas to approximate uni- form coverage within each stratum (Fig. 2). Environ- mental conditions such as sea state and percent cloud cover were recorded throughout the survey, as they have been shown to influence cetacean sighting rates (Holt and Cologne, 1987; Forney et al., 1991). However, be- cause of the small number of sightings made during each combination of environmental conditions, it was not possible to evaluate their effect quantitatively. Because of the difficulty in identifying beaked whales to species level during aerial surveys, only a combined abundance estimate was obtained for this group. In the preliminary analyses of the 1991 aerial survey data, Forney and Barlow ( 1993) assigned other unidentified species based on a 'nearest identified neighbor' ap- proach. In the analyses presented here, unidentified cetacean sightings were treated separately as either 'unidentified dolphin or porpoise,' 'unidentified small whale,' or 'unidentified large whale,' because they rep- resented only a small fraction of the total animals seen. The small number of sightings for each species made it necessary to pool distributions of perpendicu- lar sighting distances for line-transect calculations. Forney and Barlow (1993) created preliminary spe- cies groups based on considerations of school size, body size and behavior, and pooled distributions for groups that were not statistically different from one another. The same procedure was used for this analy- sis, resulting in the same three species/group-size categories: 1 ) small cetacean groups with 1-10 animals; 2) small cetacean groups with more than 10 animals; and 3) medium and large cetaceans (Table 2). Table 2 Estimates of flO) and g(0), and number of sightings (n) for the three species/group-size categories used in the analysis. Small cetaceans Group size n /t0) g(0) 1-10 > 10 99 53 4.70 2.85 0.67 0.85 Species Harbor porpoise, Phocoena phocoena Dall's porpoise, Phocoenoides dalli Pacific white-sided dolphin, Lagenorhynchus obliquidens Risso's dolphin, Grampus griseus Bottlenose dolphin, Tursiops truncatus Common dolphins Delphinus delphis and D. capensis Northern right whale dolphin, Lissodelphis borealis Medium and large cetaceans Group size n f{0) giO) 1-22 57 2.49 0.95 Species Killer whale, Orcinus orca Small beaked whales, Ziphius cavirostris and Mesoplodon spp. Sperm whale, Physeter macrocephalus Right whale, Eubalaena glacialis Gray whale, Eschrichtius robustus Minke whale, Balaenoptera acutorostrata Blue whale, B. musculus Fin whale, B. physalus Humpback whale, Megaptera novaeangliae Fishery Bulletin 93| 1995 42" 39° - -a B 36° 35° 34° 33° 32° 31°- 30° Pacific Ocean 127° 126° 125° 124° 123° 122° 121° 120° 119° 118° 117° 127° 126° 125° 124° 123° 122° 121° 120° 119° 118° 117° Longitude Longitude Figure 2 Completed transects (solid lines) for 1991 and 1992, and a posteriori geographic strata (separated by broken lines) used in the analysis. Area numbers are shown in circles. Abundance estimation Line transect methods (Burnham et al., 1980; Buck- land et al., 1993a) were applied to estimate abun- dances separately for each species in each stratum: N h yy A n ij,k \j,k /}<0) 2L, gj (0) (1) where Ni = l u,k 'ij.k estimated total number of animals of species k in the study area; number of sightings of species k in area i and species/group-size category,/'; average group size of species k in area i and species/group-size category J, calculated as the total number of animals in all groups di- vided by the number of groups sighted; /"■(0) - the probability density function evaluated at zero perpendicular distance for species/group- size category j; giO) = the probability of detecting a group of ani- mals on the transect line for species/group- size category j ; L = the length of transect surveyed in area i (in km); and A- = the size of area i (in km 2 ). Values for/(0) were obtained for each species/group- size category by fitting the distribution of all per- pendicular sighting distances (primary and second- ary; measured in km) to the Hazard rate model with the statistical software program HAZARD (Buckland, 1985). A value for ^(0) was estimated fol- lowing the methods described in Forney and Barlow ( 1993 ), but because of small sample sizes, it was not possible to estimate the variance ing(0). This should result in a downward bias in the variance of the abun- dance estimates, but bias in the abundance estimates themselves will be reduced. The lengths of transect lines flown, L- (and total sizes, A-), for the four areas are 3,715 km (46,300 km 2 ) for area 1; 2,831 km (63,772 km 2 ) for area 2; 4,461 km (120,108 km 2 ) for area 3; and 2,035 km (34,090 km 2 ) for area 4. Variance estimation Variance in estimated abundance was calculated with bootstrap techniques applied to the complete data Forney et al.: Abundance of cetaceans in California waters: aerial surveys 19 set. The data were subdivided by area into effort seg- ments of equal length, and the segments were then drawn randomly with replacement until the total number of kilometers actually surveyed in each area was reached. This process was replicated 1,000 times. Forney and Barlow (1993) demonstrated that the choice of segment lengths between 5 km and 20 km did not influence the resulting estimates of precision. In this analysis we also performed bootstrap simula- tions for 50 km and 100 km segments and again found that segment length did not affect estimates of vari- ance. For the bootstrap analysis, we chose a segment length of 50 km, which roughly reflects the degree of sampling variability for these surveys (i.e. the dimen- sion of actual gaps in the sampling grid in Figure 2). Each of the 1,000 bootstrap replicates was treated and analyzed as a separate survey: sightings were first stratified into the three species/group-size cat- egories given above. Individual values for n and s were calculated, and/10) was estimated with the pro- gram HAZARD. The estimated value of g(0) was treated as a correction factor known without error. The variance, coefficient of variation, and 95% confi- dence intervals were obtained from the distribution of the 1,000 bootstrap abundance estimates with stan- dard formulae. Because the bootstrap method (Buck- land, 1984) of obtaining confidence intervals can re- sult in the lower 95% confidence intervals being smaller than the actual number of animals seen (or even zero) we also calculated log-normal confidence intervals based on the bootstrap coefficient of variation. Results Detailed results of the survey, including sighting in- formation and plots of sighting locations for all spe- cies sighted are presented elsewhere (Carretta and Forney, 1993). Results relevant to the analyses pre- sented in this paper are given below. A total of 253 cetacean sightings were made (Fig. 3): 213 on effort (while actively searching), and an additional 40 off effort (24 while in transit, 8 beyond 12° declination angle, 7 while circling over another group of animals, and 1 by an off-effort observer). Twenty eight on-ef- fort sightings could not be positively identified to the 42° 38°- T3 3 36° -I 35°- 34° 33° 32° 31° 30° Pacific Ocean , 1 , 1 , 1 , 1 1 1 1 1 1 1 1 1 1 1 1 127° 126° 125° 124° 123° 122° 121° 120° 119° 118° 117° Longitude 42" 38° 37° 0) "O 2 36° 33° 32° 31°- 30 Pacific Ocean . 1 . 1 1 1 1 1 1 F ' 1 1 ' 1 ' I ' 127° 126° 125° 124° 123° 122° 121° 120° 119° 118° 117° Longitude Figure 3 Locations of all 253 cetacean sightings made during the 1991 and 1992 surveys. The 213 on-effort sightings (used in the abun- dance estimation) are shown by diamonds, and the 40 off-effort sightings (e.g. made while circling or in transit) are shown with plus signs. 20 Fishery Bulletin 93(1), 1995 species level. Four of these sightings were identified as ziphiid whales, for which a combined abundance estimate was calculated. The remaining 24 sightings were treated separately in the analyses. 300 400 500 600 700 Perpendicular distance 800 900 1000 100 200 300 400 500 600 700 800 900 1000 Perpendicular distance 300 400 500 600 700 Perpendicular distance 800 900 1000 Figure 4 Distribution of perpendicular sighting distances (100-m intervals; solid line) and Hazard model fit (dotted line) for (A) small cetaceans in groups <10, (B) small cetaceans in groups >10, and (C) medium and large cetaceans. The Hazard model provided adequate fits to the perpendicular distance distributions for the three species/group-size categories (Fig. 4). Estimates of f\0) andg(O) are given for each group in Table 1. Al- though the full transect grid was not completed in either year because of poor weather, the resulting estimates of abundance (Table 3) are the most precise that have been produced to date for this area and sea- son. CVs range from 0.24 to 0.49 for small cetaceans and from 0.35 to 1.11 for large cetaceans. Discussion Comparisons with previous abundance estimates Our abundance estimates (Table 3) can be compared directly with estimates based on 1975-83 aerial sur- veys, 12 which are likely to have similar biases. The estimate of 8,460 Dall's porpoise, Phocoenoides dalli, is similar to previous aerial survey estimates of 3,000-4,000 in winter and spring. 1 - 2 The current es- timate of 122,000 Pacific white-sided dolphins, 3 Lagenorhynchus obliquidens, is greater than the com- bined estimates of 26,000 (spring) to 33,500 (winter) for central and northern California 2 and 5,300 ( Jan- Jun) for southern California. 1 Our estimate of 21,300 northern right whale dolphins is less than the com- bined estimates of 29,000 (spring) to 61,500 (winter) for central and northern California 2 and 5,900 ( Jan- Jun) for southern California. 1 The prior studies do not give estimates of statistical precision for any of the above species, but given the CVs of our estimates, the above differences are not likely to be statistically significant. In contrast to the species above, common dolphins, Delphinus spp., appear to be much more abundant at present than during the period 1975—83. The cur- rent winter estimate (306,000; CV=0.34) is more than an order of magnitude larger than the previous value of 15,488 (CV=0.36; Dohl et al., 1986), and the 99% log-normal confidence limits for these two estimates do not overlap. Preliminary comparisons (Barlow, unpubl. data) of 1979 and 1980 ship surveys with the 1991 ship survey (Barlow, this issue) also show a significant increase in common dolphin abundance. Based on these two separate lines of evidence for winter and summer conditions, the abundance of common dolphins in California appears to have in- 3 Although estimates for Pacific white-sided dolphins based on the combined 1991 and 1992 survey data are over twice the preliminary estimate of 46,000 from only the 1991 data (Forney and Barlow, 1993), the new estimate lies well within the 95% confidence limit of the previous value. Forney et al.: Abundance of cetaceans in California waters: aerial surveys 21 Table 3 Number of groups seen, mean group size, density of individuals , and abundance estimates for cetaceans in the entire California study area, and subdivided by geographic stratum (See Fig. 2). Coefficients of variation (CV) and 95% confidence intervals (CI) for the overall abundance estimates are also given. Unid.=unidentified. Bootstrap CI Log-normal CI Animal Population Species and Number of Mean group density size Lower Upper Lower Upper area groups size km" 2 N CV 95% 95% 95% 95% Harbor porpoise' 18 1.2 0.0060 1,599 0.345 664 2,915 829 3,085 Area 1 0.0 0.0000 Area 2 0.0 0.0000 Area 3 10 1.0 0.0079 949 Area 4 8 1.4 0.0191 650 Dall's porpoise 38 3.1 0.0320 8,460 0.240 5,203 13,361 5,320 13,453 Area 1 9 4.0 0.0342 1,582 Area 2 2 4.5 0.0112 716 Area 3 19 2.6 00395 4,744 Area 4 8 3.0 0.0416 1,418 Pacific white-sided dolphin 21 151.6 0.4605 121,693 0.466 35,404 261,524 51,041 290,144 Area 1 5 24.6 0.0573 2,654 Area 2 7 69.4 0.2945 18,779 Area 3 7 237.1 0.6218 74,678 Area 4 2 457.0 0.7505 25,583 Risso's dolphin 19 47.6 0.1225 32,376 0.456 10,255 65,984 13,812 75,891 Area 1 14 28.5 0.2029 9,396 Area 2 1 8.0 0.0100 636 Area 3 4 124.3 0.1860 22,343 Area 4 0.0 0.0000 Bottlenose dolphin 8 17.9 0.0123 3,260 0.487 618 6,783 1,320 8,052 Area 1 7 20.3 0.0684 3,165 Area 2 0.0 0.0000 Area 3 1 1.0 0.0008 95 Area 4 0.0 0.0000 Common dolphins 27 514.9 1.1568 305,694 0.340 124,730 539,319 159.864 584,552 Area 1 22 592.7 5.8769 272,101 Area 2 4 176.0 0.4161 26,535 Area 3 1 157.0 0.0588 7,058 Area 4 0.0 0.0000 Northern right whale dolphin 31 18.9 0.0807 21,332 0.428 9,151 42,629 9,548 47,658 Area 1 18 12.3 0.1378 6,381 Area 2 4 56.5 0.1395 8,895 Area 3 6 11.8 0.0341 4,091 Area 4 3 22.7 0.0577 1,966 Killer whale 2 1.0 0.0002 65 0.689 133 19 220 Area 1 0.0 0.0000 Area 2 1 1.0 0.0005 30 Area 3 1 1.0 0.0003 35 Area 4 0.0 0.0000 Beaked whales 2 8 1.9 0.0015 392 0.408 151 774 182 845 Area 1 0.0 0.0000 Area 2 3 1.0 0.0014 89 Area 3 2 1.5 0.0009 106 Area 4 3 3.0 0.0058 197 Sperm whale 3 10.0 0.0034 892 0.990 2,798 176 4,506 Area 1 0.0 0.0000 Area 2 2 14.5 0.0134 857 22 Fishery Bulletin 93(1). 1995 Table 3 (Continued) Bootstrap CI Log-normal CI AniTTlfll Po*^>'l atinn Species and Number of Mean group .\I1L Lllill A V density size Lower Upper Lower Upper area groups size km" 2 N CV 95% 95% 95% 95% Area 3 1 1.0 0.0003 35 Area 4 0.0 0.0000 Northern right whale 1 1.0 0.0001 16 1.110 59 3 95 Area 1 1 1.0 0.0004 16 Area 2 0.0 0.0000 Area 3 0.0 0.0000 Area 4 0.0 0.0000 Gray whale 3 25 4.2 0.0108 2,844 0.347 1,187 5,270 1,469 5,507 Area 1 12 3.4 0.0145 669 Area 2 0.0 0.0000 Area 3 11 5.3 0.0170 2,043 Area 4 2 3.0 0.0039 132 Minke whale 3 1.0 0.0003 73 0.616 181 24 223 Area 1 1 1.0 0.0004 16 Area 2 0.0 0.0000 Area 3 1 1.0 0.0003 35 Area 4 1 1.0 0.0006 22 Blue whale 1 1.0 0.0001 30 0.990 100 6 149 Area 1 0.0 0.0000 Area 2 1 1.0 0.0005 30 Area 3 0.0 0.0000 Area 4 0.0 0.0000 Fin whale 2 1.5 0.0002 49 1.012 57 9 254 Area 1 2 1.5 0.0011 49 Area 2 0.0 0.0000 Area 3 0.0 0.0000 Area 4 0.0 0.0000 Humpback whale 8 1.6 0.0012 319 0.407 114 622 148 688 Area 1 1 1.0 0.0004 16 Area 2 0.0 0.0000 Area 3 2 1.5 0.0009 106 Area 4 5 1.8 0.0058 197 Unid. large whale 5 1.2 0.0006 160 0.457 40 348 68 376 Area 1 1 2.0 0.0007 33 Area 2 0.0 0.0000 Area 3 3 1.0 0.0009 106 Area 4 1 1.0 0.0006 22 Unid. small whale 3 1.0 0.0003 68 0.676 188 20 226 Area 1 2 1.0 0.0007 33 Area 2 0.0 0.0000 Area 3 1 1.0 0.0003 35 Area 4 0.0 0.0000 Unid. dolphin or porpoise 15 4.4 0.0180 4,766 0.331 2,050 8,368 2,533 8,966 Area 1 2 1.5 0.0028 132 Area 2 5 4.2 0.0223 1,419 Area 3 7 5.7 0.0258 3,096 Area 4 1 2.0 0.0035 118 1 More appropriate estimates for harbor porpoise are recently avai able in Barlow and Forney ( 1994). (See D iscussion section.) 2 This category includes beaked whales of the genus Mesoplodon and Cuvier's beaked whale, Ziphius cavirostris. No Baird's beaked whales, Berardius bairdii , were seen during the surveys. 3 A more accurate estimate of the entire population of California gray whales is presented in Buckland et al., 1993. (See Discus- sion section.) Forney et al. ; Abundance of cetaceans in California waters: aerial surveys 23 creased dramatically since the early 1980's. The causes of this increase are not known, but it is pos- sible that long-term oceanographic changes (Roemmich, 1992; Roemmich and McGowan, 1994) have resulted in a shift in the distribution of com- mon dolphins into this area. This hypothesis is con- sistent with the observed decline in population size of the northern common dolphin south of our study area (Anganuzzi and Buckland, 1994). Similarly, an apparent decrease in abundance was seen in short-finned pilot whales, Globicephala macrorhynchus . This species was commonly seen in the Southern California Bight on surveys during the late 1970's and early 1980's, 1 ' 2 but only one off-effort sighting of four animals was made during our surveys. Our estimate of 304 humpback whales is roughly half the recent estimate obtained from photo-identi- fication studies. 4 This is quite surprising because humpback whales, Megaptera novaeangliae , in the California feeding population are expected to be in waters off Mexico during the winter and spring sea- son. However, it is possible that some animals had already moved north into California at the time of the sightings. Alternatively, the sighted animals may have been part of the southeastern Alaska feeding population that migrates southward to breed in Mexi- can waters in spring (Baker et al., 1986). Previously published estimates for harbor porpoise, Phocoena phocoena (Barlow, 1988; Barlow et al., 1988; Barlow and Forney, 1994) and gray whales, Eschrichtius robustus (Reilly, 1984; Buckland et al., 1993b), are substantially higher than the estimates presented here. This is probably because the defined study area is not appropriate for the range of these animals. Gray whales have a much larger range and migrate through California waters (southward and then northward) from roughly November to May. Our estimate represents that portion of the population which was migrating through California in March and early April. Harbor porpoise are limited to a narrow coastal band, and our transect lines only over- lapped with this region at specific points. More appro- priate abundance estimates for harbor porpoise are pub- lished in Barlow (1988) and in Barlow and Forney (1994). Comparisons with 1991 ship surveys Although a statistical comparison between these winter and spring aerial survey estimates and the 1991 summer and fall ship survey estimates (Barlow, this issue) is precluded at this time because of dif- ferences in the sizes of the two study areas, a few patterns are noteworthy. Despite the differences in seasonal timing and areal coverage, estimates of abundance are very similar for several species. Simi- lar estimates of abundance were obtained for total common dolphins (306,000 vs. 246,000), northern right whale dolphins, Lissodelphis borealis (21,300 vs. 9,340), bottlenose dolphins, Tiirsiops truncatus (3,260 vs. 1,500), and sperm whales, Physeter macrocephalus (892 vs. 756) (aerial vs. ship esti- mates, respectively). More disparate estimates were obtained for Pacific white-sided dolphins ( 122,000 vs. 12,300), Risso's dolphins, Grampus griseus (32,400 vs. 8,500), harbor porpoise (1,600 vs. 52,700), Dall's porpoise (8,460 vs. 78,400), and total beaked whales, Ziphius cavirostris and Mesoplodon spp. (392 vs. 3,230). It may be important to note that all cases in which the ship estimates are substantially larger than the aerial estimates are for species which spend a large fraction of their time diving (harbor porpoise, Dall's porpoise, and beaked whales). Such species could be more easily missed by aerial observers owing to avail- ability bias. In the case of Pacific white-sided dol- phins and Risso's dolphins, the winter and spring aerial estimates may be larger because of a seasonal movement of animals out of Oregon and Washington in winter. 5 Additional analyses, which account for differences in geographic extent of the aerial vs. ship surveys, are planned in the future. Bias There are several sources of potential bias in this study. First, abundance estimates may be biased low because animals are missed by aerial observers (per- ception bias; Marsh and Sinclair, 1989). This is most likely to be a problem with poor observation condi- tions (high sea state or overcast conditions, or both). We have attempted to estimate the magnitude of perception bias in this study through the use of a conditionally independent observer and have cor- rected abundance estimates to reduce this effect. A second source of downward bias, availability bias (Marsh and Sinclair, 1989), is introduced because animals that are submerged when the aircraft passes overhead are not available to be seen. This effect is 4 Calambokidis, J., G. H. Steiger, and J. R. Evenson. 1993. Pho- tographic identification and abundance estimates of humpback and blue whales off California in 1991-92. Final Contract Re- port 50ABNF100137 to Southwest Fish. Sci. Cent., RO. Box 271, La Jolla, CA 92038, 67 p. 5 Green, G. A., J. J. Braeggeman, R. A. Grotefendt, C. E. Bowlby, M. L. Bonnell, and K. C. Balcomb III. 1992. Cetacean distribu- tion and abundance off Oregon and Washington, 1989-1990. Ch. 1 in J. J. Brueggeman (ed.), Oregon and Washington ma- rine mammal and seabird surveys. Minerals Management Ser- vice Contract Report 14-12-0001-30426 prepared for the Pacific OCS (Outer Continental Shelf) Region. 24 Fishery Bulletin 93(1). 1995 expected to be smallest for species which tend to oc- cur in large groups, such as common dolphins, and largest for species which spend relatively little time at the surface, such as porpoise, beaked whales, and sperm whales. Dive studies (Barlow et al., 1988) may provide information on the magnitude of availability bias, but each species requires a separate assessment of the average proportion of time it spends at the sur- face (and hence is 'available'), and adequate estimates are not currently available for most species in Cali- fornia waters. Rough estimates can be made for Dall's porpoise and humpback whales based on prior stud- ies. Dall's porpoise have similar sighting character- istics to those of harbor porpoise (both have a small body size and generally are found in small groups); thus, assuming that dive patterns are similar and applying the correction factor of 3.1 (CV=0.17) for harbor porpoise, 6 one would obtain a corrected esti- mate of approximately 26,200 Dall's porpoise. Based on a very small sample, a correction factor of 2.7 has been estimated for humpback whales. 7 This would yield a corrected abundance estimate of 861 humpback whales. Clearly, given the magnitude of these correc- tion factors, availability bias can be substantial. Potential upward bias in line-transect analysis can result if factors other than distance to the trackline affect the probability of seeing a school. School size has been shown to affect the probability of detection (Drummer, 1985; Holt and Sexton, 1989), and this can lead to an upward bias in the abundance esti- mate (Quinn, 1985; Drummer and McDonald, 1987; Buckland et al., 1993a). To counteract this effect, we have stratified small cetacean sightings by group size and estimated abundances separately for small and large groups of the same species. This is an artificial separation, but it reduces potential biases that are due to large variation in group size within a single species, such as common dolphins or Pacific white- sided dolphins. Within each stratum, correlations of perpendicular sighting distance with group size are weak and not significant at oc=0.05 (r=0.195 for small cetaceans in groups of 1—10 animals; r=0.169 for small cetaceans in groups of greater than 10 animals; and r=0.183 for whales in groups of all sizes). 6 Calambokidis, J., J. R. Evenson, J. C. Cubbage, P. J. Gearin, and S. D. Osmek. 1993. Development of a correction factor for aerial surveys of harbor porpoise. Draft Final Contract Report to the National Marine Mammal Laboratory, NMFS, NOAA, 7600 Sand Point Way NE, BIN C-15700, Seattle, WA 98115. 36 p. 7 Calambokidis, J., G. H. Steiger, J. C. Cubbage, K. C. Balcomb, and P. Bloedel. 1989. Biology of humpback whales in the Gulf of the Farallones. Final report for Contract CX-8000-6-0003 to Gulf of the Farallones National Marine Sanctuary, NOAA, Fort Mason Center, Bldg. 201, San Francisco, CA 94123, 93 p. In summary, we have attempted to correct for per- ception bias by estimating the fraction of animals missed during these surveys and have minimized potential upward bias with a poststratification by school-size range. However, species-specific availabil- ity bias cannot currently be estimated, and overall our abundance estimates are likely to be biased downward. Precision Estimation of variance for line-transect abundance calculations can be difficult. We have attempted to include most of the sources of sampling error in the bootstrap procedure, which reestimates n, s, and/10) (in Eq. 1) for each replicate. Our analysis revealed that the choice of segment length used for the boot- strap did not affect the resulting estimates of preci- sion within the range of appropriate segment lengths for this study (5-100 km; longer segments would not be appropriate because surveys extended only 100-150 km offshore). However, potential heterogeneity due to the pooling of different species and group sizes for esti- mation of /(0) andg(0) was not accounted for in preci- sion estimates. Furthermore, we did not include the variance ing(0) or in the estimation of group size for each school encountered (however, the variance in the estimated mean group size for the survey was included in the bootstrap procedure). Thus, the coefficients of variation for the abundance estimates (Table 3) are likely to be underestimated and the confidence inter- vals are likely to be too narrow. Considerations for future aerial surveys Two species of common dolphins, short-beaked and long-beaked, are recognized in California waters (Rosel, 1992; Dizon et al., 1994; Heyning and Perrin, 1994). Although clear differences in color pattern, size, and beak length exist between these two forms, it is not currently possible to differentiate them dur- ing aerial surveys; therefore the abundance estimate here is a combined estimate. Unless reliable means of identifying the two species from the air are devel- oped, aerial surveys will not be adequate for future assessments requiring separate estimates of short- beaked and long-beaked common dolphins. Similarly, it was difficult to distinguish between the smaller species of beaked whales during our aerial surveys. The estimates presented for the beaked whales as a group are therefore a combined estimate for Ziphius cavirostris and Mesoplodon spp. All unidentified beaked whale sightings could be narrowed down to these two genera. The only other beaked whale species known to occur in this region, Berardius bairdii, can be readily distinguished based Forney et al.: Abundance of cetaceans in California waters: aerial surveys 25 on its size and was not sighted during this survey. It is likely that the categorization of "small beaked whales" will be necessary on future aerial surveys. The survey grid used here was not designed for species which are restricted to a narrow coastal re- gion. Harbor porpoise are found primarily in waters inshore of the 50-fathom (92-m) isobath (Barlow, 1988). Two distinct populations of bottlenose dolphins are found in California; the inshore form is found only within about 1 km of shore (Hansen, 1990; NMFS 8 ). All of the bottlenose dolphins seen during this aerial survey were at least several miles from the main- land; therefore our estimate is assumed to represent the population of offshore animals. Precise estimates of abundance for harbor porpoise and inshore bottle- nose dolphins will require dedicated aerial surveys designed for those species. Work is currently in progress on both of these projects. 8 Acknowledgments Funding for this project was provided by the Office of Protected Species, U.S. Department of Commerce. The aircraft was provided by the NOAA Aircraft Operations Center. We express special thanks to the pilots: T O'Mara, J. Vance, P. Wehling, and M. White. We are grateful to the meteorological staff of the National Weather Service for their valuable assis- tance in planning flights. Observers were D. Ever- hart, S. Kruse, C. LeDuc, R. LeDuc, and M. Lycan (also J. B., J. V. O, and K. A. F). Survey design was improved by comments from D. Ainley, H. Braham, S. Buckland, K. P. Burnham, D. DeMaster, D. Goodman, T Gerrodette, L. Hansen, W Hoggard, D. Palka, and T Polacheck. Data recording software was developed by J. Cubbage (and J. B.). An earlier draft of this manuscript was reviewed by the participants of the Status of California Cetacean Stocks Work- shop in March- April 1993; we thank all for their ef- forts and reviews. The submitted manuscript was reviewed by J. Calambokidis, R. Hobbs, and an anonymous reviewer. Surveys were conducted under Marine Mammal Protection Act permit 748 and per- mits GFNMS-01-92 and CINMS-01-92 from the Na- tional Marine Sanctuary Program. Literature cited Anganuzzi, A., and S. Buckland. 1994. Relative abundance of dolphins associated with tuna in the eastern Pacific Ocean: analysis of 1992 data. Rep. Int. Whaling Comm. 44:361-366. 8 National Marine Fisheries Service, Southwest Fisheries Sci- ence Center, unpublished data. Baker, C. S., L. M. Herman, A. Perry, W. S. Lawton, J. M. Straley, A. A. Wol man. G. D. Kaufman, H. E. Winn, J. D. Hall, J. M. Reinke, and J. Ostman. 1986. Migratory movement and population structure of hump- back whales (Megaptera novaeangliae) in the central and eastern North Pacific. Mar. Ecol. Prog. Ser. 31:105-119. Barlow, J. 1988. Harbor porpoise, Phocoena phocoena, abundance es- timates in California, Oregon and Washington: I. Ship surveys. Fish. Bull. 86:417-432. 1995. The abundance of cetaceans in California waters. Part I: Ship surveys in summer and fall of 1991. Fish. Bull. 93:1-14. Barlow, J., and K. A. Forney. 1994. An assessment of the 1994 status of harbor porpoise in California. U.S. Dep. Commer., NOAA Tech. Memo. NOAA-TM-NMFS-SWFSC-205, 17 p. Barlow, J., C. Oliver, T. D. Jackson, and B. L. Taylor. 1988. Harbor porpoise, Phocoena phocoena, abundance es- timates in California, Oregon and Washington: II. Aerial surveys. Fish. Bull. 86:433-444. Barlow, J., R. W. Baird, J. E. Heyning, K. Wynne, A. M. Manville II, L. F. Lowry, D. Hanan, J. Sease, and V. N. Burkanov. In press. A review of cetacean and pinniped mortality in coastal fisheries along the west coast of the U.S. and Canada and the east coast of the USSR. Rep. Int. Whal- ing Comm. (Special issue.) Buckland, S. T. 1984. Monte Carlo confidence intervals. Biometrics 40:811-817. 1985. Perpendicular distance models for line transect sampling. Biometrics 41:177-195. Buckland, S. T., D. R. Anderson, K. P. Burnham, and J. L. Laake. 1993a. Distance sampling: estimating abundance of biologi- cal populations. Chapman and Hall, New York, 446 p. Buckland, S. T., J. M. Breiwick, K. L. Cattanach, and J. L. Laake. 1993b. Estimated population size of the California gray whale. Mar. Mamm. Sci. 9:235-249. Burnham, K. P., D. R. Anderson, and J. L. Laake. 1980. Estimation of density from line transect sampling of biological populations. Wildl. Monogr. 72, 202 p. Carretta, J. V., and K. A. Forney. 1993. Report of the two aerial surveys for marine mam- mals in California coastal waters utilizing a NOAA DeHavilland Twin Otter aircraft, March 9-April 7, 1991 and February 8-April 8, 1992. U.S. Dep. Commer, NOAA Tech. Memo. NOAA-TM-NMFS-SWFSC-185, 77 p. Dizon, A. E., W. F. Perrin, and P. A. Akin. 1994. Stocks of dolphins (Stenella spp. and Delphinus delphis) in the eastern tropical Pacific: a phylogeographic classification. U.S. Dep. Commer., NOAA Tech. Rep. NMFS 119, 20 p. Dohl, T. P., M. L. Bonnell, and R. G. Ford. 1986. Distribution and abundance of common dolphin, Del- phinus delphis, in the Southern California Bight: a quan- titative assessment based upon aerial transect data. Fish. Bull. 84:333-343. Drummer, T. D. 1985. Size-bias in line transect sampling. Ph.D. diss., Univ. Wyoming, Laramie, 143 p. Drummer, T. D., and L. L. McDonald. 1987. Size-bias in line transect sampling. Biometrics 43:13-21. 26 Fishery Bulletin 93(1), 1995 Forney, K. A., D. A. Hanan, and J. Barlow. 1991. Detecting trends in harbor porpoise abundance from aerial surveys using analysis of covariance. Fish. Bull. 89:367-377. Forney, K. A., and J. Barlow. 1993. Preliminary winter abundance estimates for ceta- ceans along the California coast based on a 1991 aerial survey. Rep. Int. Whaling Comm. 43:407-415. Hansen, L. J. 1990. California coastal bottlenose dolphins. In S. Leath- erwood and R. R. Reeves (eds.), The bottlenose dolphin, p. 403—420. Academic Press, San Diego. Heyning, J. E., and W. F. Perrin. 1994. Evidence for two species of common dolphins (genus Delphinus) from the eastern North Pacific. Contrib. Sci. (Los Angel.) 422. Holt, R. S., and J. Cologne. 1987. Factors affecting line transect estimates of dolphin school density. J. Wildl. Manage. 51:836-843. Holt, R. S., and S. N. Sexton. 1989. Monitoring trends in dolphin abundance in the east- ern tropical Pacific using research vessels over a long sam- pling period: analyses of 1986 data, the first year. Fish. Bull. 88:105-111. Leatherwood, S., and W. A. Walker. 1979. The northern right whale dolphin Lissodelphis bo- realis Peale in the eastern North Pacific. In H. E. Winn and B. L. Olla (eds.), Behavior of marine mammals: cur- rent perspectives in research. Vol. 3: Cetaceans, p. 85- 141. Plenum Press, New York-London. Marsh, H., and D. F. Sinclair. 1989. Correcting for visibility bias in strip transect aerial surveys of aquatic fauna. J. Wildl. Manage. 53:1017-1024. Quinn, T. J., II. 1985. Line transect estimators for schooling popula- tions. Fish. Res. 3:183-199. Reilly, S. B. 1984. Assessing gray whale abundance: a review. In M. L. Jones, S. L. Swartz, and J. S. Leatherwood (eds.), The gray whale, Eschrichtius robustus, p. 203-223. Academic Press, Orlando. Roenunich, D. 1992. Ocean warming and sea level rise along the south- west U.S. coast. Science 257:373-375. Roemmich, D. H., and J. A. McGowan. 1994. A long term decrease in zooplankton off California. Paper presented at the 1994 Ocean Sciences Meeting in San Diego, California, February 21-25, 1994. Abstract 0111-6 published in Supplement to EOS, Transactions, American Geophysical Union Vol. 75, No. 3, January 18, 1994. Rosel, P. E. 1992. Genetic population structure and systematic relation- ships of some small cetaceans inferred from mitochondrial DNA sequence variation. Ph.D. diss., Univ. California, San Diego, 191 p. U.S. Navy. 1977. U.S. Navy marine climatic atlas of the world. Vol. II: North Pacific Ocean. NAVAIR 50-1C-529. U.S. Govern- ment Printing Office, Washington, D.C. 20402. Abstract. The larval develop- ment of Sillaginodes punctata, Sillago bassensis, and Sillago schomburgkii is described based on both field-collected and laboratory- reared material. Larvae of the three species can be separated based on a combination of pigment and meristic characters, including extent and appearance of dorsal midline pigment, lateral pigment on the tail, presence or absence of pigment above the notochord tip, myomere number, extent and tim- ing of gut coiling, and size at flex- ion. The most useful meristic char- acter across the range of specimens was number of myomeres. Sillagi- nodes punctata with 42-45 myo- meres are easily distinguished from Sillago schomburgkii with 36-38, and from S. bassensis with 32-35. The timing of gut coiling and its subsequent effect on anus position differed both among the three species examined here and from that previously reported for sillaginid larvae in general. Timing of gut coiling and extent of anus migration are not useful characters for the identification of temperate Australian sillaginids at the fam- ily level but are useful on a specific level. Possible implications of the development of the gut to diet are discussed. Based on the presence of larvae, all three species spawn in South Australian waters. No larvae of a fourth sillaginid species, S. flin- dersi, were found during the study. South Australia is the western dis- tributional limit for S. flindersi and it does not appear to spawn in the area. Larval development of King George whiting, Sillaginodes punctata, school whiting, Sillago bassensis, and yellow fin whiting, Sillago schomburgkii (Percoidei: Sillaginidae), from South Australian waters Barry D. Bruce South Australian Department of Fisheries GPO Box 1625. Adelaide. South Australia 5001 Present address. CSIRO Division of Fisheries, GPO Box 1 538. Hobart. Tasmania. Australia 700 1 Manuscript accepted 15 June 1994. Fishery Bulletin 93:27-43 (1995). The perciform family Sillaginidae (whiting and sand smelts) consists of three genera, three subgenera, and thirty-one species of small to moderately sized fishes found pri- marily in shallow coastal waters of the Indo-Pacific (McKay, 1992). Sillaginids are highly valued food fishes in many tropical and temper- ate waters. The Sillaginidae are re- lated to the Percidae, Sciaenidae, and, to a lesser extent, the Haemu- lidae (McKay, 1985) although their sister group is yet to be determined (McKay, 1992). The most speciose of the three sillaginid genera (Sillago) includes twenty-nine species. The remaining two genera, Sillaginodes and Sillaginopsis, are monotypic. The taxonomy of the family is approach- ing stability; only a few species re- main undescribed (McKay, 1992). Two genera and thirteen species of sillaginids are found in Austra- lian waters. Four species inhabit the waters off South Australia: the King George or spotted whiting, Sillaginodes punctata; yellow fin whiting, Sillago schomburgkii; western school whiting, Sillago bassensis; and eastern school whit- ing, Sillago flindersii. The latter two species were, until recently, con- sidered subspecies of S. bassensis (McKay, 1992). All four species are widely distributed in southern Aus- tralia and form the basis for impor- tant commercial fisheries across their range (McKay, 1985; Kailola et al., 1993; May and Maxwell 1 ). The adult and juvenile biology of each of the four species has previ- ously been documented by several authors (Scott, 1954; Gilmour, 1969; Lennanton, 1969; Robertson, 1977; Weng, 1983, 1986; Burchmore et al., 1988; Jones 2 ; Jones et al. 3 ), but very little is known of their early life his- tory and neither the eggs nor the larvae of any of the four species have previously been described. In 1986, the South Australian Department of Fisheries began an ichthyoplankton program to inves- 1 May, J. L., and J. G. H. Maxwell. 1986. Field guide to trawl fish from temperate waters of Australia. CSIRO Division of Fisheries Res., Hobart, Tasmania, 492 p. 2 Jones, G. K. 1979. Biological investigations on the marine scale fishery in South Aus- tralia. South Australian Dep. Agric. and Fisheries Rep., 72 p. 3 Jones, G. K., D. A. Hall, K. L. Hill, and A. J. Stamford. 1989. The South Australian marine scale fishery: stock assessment, economics, management. South Australian Dep. Fisheries Green Paper, 186 p. 27 28 Fishery Bulletin 93(1), 1995 tigate the larval ecology of commercially important fishes of South Australian waters. An important pre- requisite of any such program is the ability to make an accurate identification of larvae to species. This paper details the development of Sillaginodes punctata, Sillago schomburgkii, and S. bassensis lar- vae collected during this study. Materials and methods Specimens were obtained from plankton and beach seine samples collected between March 1986 and March 1991 aboard the research vessel MRV Ngerin in coastal waters and at various inshore nursery ar- eas off South Australia. Details of sampling locations and procedures are described in Bruce (1989). Briefly, larvae were obtained from stepped oblique tows with 70-cm-diameter bongo nets fitted with 500-micron mesh. Postsettlement (refer to definition below) lar- vae and juveniles were captured with a fine mesh beach seine (7 m x 1.8 m, 2-mm mesh) as well as by dipnetting and diving. The field-collected series of Sillaginodes punctata was supplemented with lar- vae reared in the laboratory at West Beach, Adelaide. All field-collected specimens used for description were fixed in a 10% formalin-seawater solution buff- ered with sodium tetraborate (borax) and were later transferred to a 5% solution buffered with sodium B-glycerophosphate (0.5 g per 1,000 mL). Reared lar- vae were fixed immediately in the 5% solution. Reared S. punctata were used for illustration when possible because of their superior condition. Some pigment differences were apparent between reared and field-collected larvae largely as a result of ex- pansion or contraction of melanophores. Melano- phores of field-collected larvae were generally less expanded than reared specimens. Reared larvae were typically greater in length than similarly developed field-collected material owing to increased shrink- age in the latter. Similar shrinkage effects have been previously reported for a variety of species (Theilacker, 1980; Hay, 1981; Bruce, 1988). Unless specified, devel- opment at length refers to field-collected material. Representative series of S. punctata and Sillago bassensis are deposited with the I.S.R. Munro Fish Collection, CSIRO, Hobart Tasmania. Too few S. schomburgkii larvae were collected to allow a com- plete analysis and all are currently held in a collec- tion maintained by the author at CSIRO Division of Fisheries, Hobart, Tasmania. Developmental terminology and body measure- ments follow Leis and Trnski (1989). The term "postsettlement" is used to describe newly settled individuals prior to the acquisition of scales and ju- venile colour patterns, after which they are referred to as juveniles. Body length measurements (BL) are measured as notochord length, NL (i.e. from the snout tip to the end of the notochord), in preflexion and flexion larvae, and standard length, SL (i.e. from the snout tip to the posterior margin of the superior hypural elements), in postflexion larvae and juve- niles. Body depth is taken at two points. Body depth at pectoral (BDp) is equivalent to "body depth" as defined by Leis and Trnski ( 1989), that is, as "the vertical distance between body margins (exclusive of fins) through the anterior margin of the pectoral fin base." Body depth at anus (BDa) is defined as the vertical distance between body margins (exclusive of fins and, initially, the gut) through the midpoint of the anal opening. BDa includes the gut only after overlying musculature has developed. Sillaginodes punctata eggs were measured with a Zeiss photomi- croscope III fitted with an ITC 510 video camera and linked to an Apple Macintosh SE computer via an HEI 582A video coordinate digitizer. Egg dimensions are reported to the nearest micron. Larvae were measured to the nearest 0.1 mm with a dissecting microscope fitted with an ocular micrometer. Postsettlement larvae and juveniles were measured to the nearest 0.1 mm with vernier calipers. Meristic counts were made on S. punctata and Sillago bassensis specimens cleared and stained with alcian blue and alizarin red-S following Potthoff (1984). Insufficient specimens of S. schomburgkii were available for clearing and staining and therefore all meristics were taken from unstained material. Descriptions are based primarily on the detailed examination of a representative series of specimens; however, comments on pigment and meristic vari- ability stem from the routine examination of all lar- vae collected. The number of specimens examined in detail, the size range covered, and the museum ref- erence numbers (for lodged material) are provided under each species account. Results Identification Larvae were identified to family level from larval sillaginid characters reported in the literature. Silla- ginid larvae are elongate and have 30-^t4 myomeres (Johnson, 1984; Miskiewicz, 1987; Leis and Trnski, 1989). The gut typically reaches to greater than 55% body length in preflexion larvae. The anus is reported to migrate anteriorly during development (often dur- ing flexion) as a result of coiling of the anterior sec- tion of the gut, thus shortening the preanal length Bruce Larval development of Sillagmodes punctata. Sillago bassensis, and Sillago schomburgku 29 (Leis and Trnski, 1989). Sillaginids have a charac- teristic series of melanophores along the dorsal and ventral midlines (particularly prominent in small larvae) and generally have pigment located on the angle of the lower jaw. Three types of sillaginid larvae were found during this study. Specific identity of two of the types (S. schomburgkii and Sillaginodes punctata) was estab- lished by comparing vertebral counts and fin meristics of postflexion larvae to those of adult and juvenile specimens. Smaller larvae were linked by establishing a developmental series based on the extent and appearance of dorsal midline pigment, lateral pigment on the tail, presence or absence of pigment above the notochord tip, myomere number, extent and timing of gut coiling, and size at flexion. The identity of S. punctata was also confirmed by comparison to reared larvae. Though clearly separating Sillaginodes punctata from Sillago schomburgkii, fin meristics and verte- bral counts overlap in the other two South Austra- lian sillaginid species (S. bassensis and S. flindersi), thus making the specific separation of their larvae difficult. Sillaginid larvae from southern Tasmania (where only S. flindersi are found) and larvae be- lieved to be S. flindersi from New South Wales (NSW) coastal waters were compared to the third sillaginid larval type collected in South Australia in order to ascertain its identity. The NSW and Tasmanian (re- ferred to herein as eastern) specimens were highly similar but differed from the South Australian type with respect to two pigment characters. First, east- ern specimens had a single prominent, elongate mel- anophore located below the level of the pectoral fin base and overlying the cleithrum that was absent in South Australian material (Fig. 1). Second, eastern specimens developed external lateral midline pig- ment on the tail at an earlier size (7.2 mm) than did South Australian material (14.8 mm). Two sillaginids are known from Tasmanian waters: Sillaginodes punctata and Sillago flindersi (Lastetal., 1983). Only S. flindersi is known to spawn in Tasmanian waters. The eastern form was thus identified as S. flindersi and the South Australian specimens as S. bassensis. Insufficient material was available across the full size range to render an adequate description of the lar- val development of S. flindersii, and thus this spe- cies is not treated in further detail here. The most useful meristic character separating the three South Australian larval types was number of myomeres. Sillaginodes punctata with 42-45 myomeres are easily distinguished from Sillago schomburgkii with 36—38 and S. bassensis with 32— 35. Meristic details for these three species and S. flindersi are listed in Table 1. Descriptions King George whiting [Sillaginodes punctata Cuvier 1829), Figure 2 Material examined — 75 specimens, 2.0-30.5 mm BL (CSIRO L587-01, L587-02, L587-03, L587-04, L587-05, L587-06; L588-01; L589-01). Larval development — The pelagic eggs of S. punctata are spherical and have an unsegmented yolk and smooth chorion. Late stage eggs are 839- 935 microns in diameter (mean 880, n=25) and have a single oil droplet 246-263 microns in diameter (mean 255, n=25). Reared larvae hatched at 2.00- 2.15 mm (mean 2.07, re=24) at 16.5-18.7°C. The tim- ing of fertilization was not recorded as spawning oc- curred in brood stock tanks overnight. Estimates for incubation period are 48-60 hours. The temperature of the spawning tank was 16.5°C and fertilized eggs were transferred to a 90-liter tank held at 18.0-18.7°C for subsequent incubation, 24 hours prior to hatching. Newly hatched larvae have a posteriorly located oil droplet and adopt a head-down position in rear- ing containers. Yolk absorption was complete in reared larvae by 3.5 mm (8 days), although the small- Figure 1 Detail of head and trunk pigment of (A) Sillago bassensis and (B) Sillago flindersi. Myomeres have been omitted for clarity. Position of cleithrum is indicated by a dotted line. Arrow indicates characteristic melanophore overlying cleithrum in S. flindersi. 30 Fishery Bulletin 93(1), 1995 Table 1 Selected early life history features useful for identifying sillaginid larvae ( sizes are in mrr ). Size at Size at Completion first lat. first of Size at Size at midline dorsal Size at fin Number of Species gut coiling flexion pigment banding settlement formation 6 myomeres Dorsal fin Anal fin Pectoral fin Sillaginodes punctata' 21.0-24.0 5.7-7.0 8.0 6.5-7.0 15.0-18.0 C,P1,A,D2+D1,P2 42^5 XlI-XIII+I,25-27 11,21-24 13-15 Sillago bassensis' 4.1-7.5 4.8-6.5 14.0 12.0-13.0 12.0-13.0 C,P1,A,D1+D2,P2 32-35 X-XII+1, 16-19 11,18-20 15-16 Sillago schomburgkii' >5.1<10.1 7 4.8-? 2.7 <10.1 12.0-13.0 8 36-38 X-XJI+1, 19-22 11,17-20 15-16 Sillago nliata 2 ' 3 <5.0 4.0-5.6 5.3 6.5 15.5 CAD2,D1,P1,P2 30-34 XJ+1,16-18 11,15-17 15-17 Sillago maculata 2 8 4.6-6.5 3.3 10.6 8 C,PU,D2,D1,P2 33-36 XI-XII+1,19-21 11,19-20 15-17 Sillago sihama 4 5.9 5.9 5.9 9.0 8 8 33-34 XI+1,20-23 11,21-23 15-17 Sillago japonica 5 <7.6 <7.6 7.6 7.6 11.5 8 35-37 XI+1,21-23 11,22-24 15-17 ' This study. 2 Miskiewicz, L987. 3 Munro, 1945 4 Uchida et al. , 1958. 5 Mito, 1966 (as Sillago japonicus ); Kinoshita 1988. 6 Based on all elements present ar d ossified, C = caudal PI = pectoral, P2 = pelvic A = anal, Dl = first dorsal D2 = second dorsal. 7 No specimens between 5.1 and 10.1 mm were available. Coiling of the gut had not commenced in the 5.1-mm specimen but had been completed in the 10.1-mm specimen. * Data not available. est field-collected larvae (2.9 mm) had already com- pleted yolk absorption. Larvae are elongate (BDp=ll-16% BL) and have 42^45 myomeres (17-21 abdominal + 23-27 candal). Body depth at anus increases slightly from 7% to 9% BL during development. Other body proportions re- main relatively constant (Table 2). The gut is ini- tially straight and differentiates into defined fore, mid and hind gut sections by 3.7 mm. The gut exhib- its some convolution but does not coil during the lar- val phase. The midgut becomes rugose by approxi- mately 5.0 mm and remains so, although overlying musculature obscures this feature in postsettlement larvae larger than 21.0 mm. The gut begins to coil in postsettlement larvae of 21.0 -24.0 mm and is com- plete by 26.0 mm. Coiling of the gut proceeds with- out migration of the anus and is achieved by elonga- tion and anterior looping of the midgut. Conse- quently, body proportions do not show a significant change in preanal length which remains at 50—52% BL. The gas bladder is first visible in reared larvae by 3.5 mm (5 days) and is prominent and inflated in 86% of field-collected larvae (random subsample, n=50; all larvae collected at night) and all postsettlement lar- vae collected (all postsettlement larvae collected dur- ing day). The gas bladder has its origin at myomeres 2-5 in preflexion larvae but migrates posteriorly dur- ing development to myomeres 13-18 by 18.7 mm. The snout is initially slightly concave in profile, but after flexion, this gradually changes to straight or slightly convex. The eye is round. The mouth ini- tially reaches to below the eye, but is short of the eye in postflexion larvae. Six to eight small villiform teeth are present on the premaxilla by 5.8 mm. The num- ber of teeth increases to 10-12 by late flexion (6.5- 7.0 mm). There are no head spines. Scales are first present around the gut and lateral midline by approximately 27.5 mm. The development of fins in larval and juvenile S. punctata is summarized in Table 1. Completion of fin development occurs in the following sequence: caudal; pectoral; anal and second dorsal (almost si- multaneously); first dorsal; and pelvic. The rays of the caudal fin are present just prior to flexion in larvae of 5.6 mm. Flexion commences by 5.7-6.0 mm and is usually complete by 7.0 mm. Pec- toral fin buds are present in reared larvae as slight swellings on the body above the anterior margin of the oil droplet by 3.1 mm (2 days post hatch). Incipient rays are first visible by 7.5 mm and commence ossifi- cation by 8.5 mm. A full complement of 13-15 pectoral rays is present by 11.5 mm. Anal and second dorsal fin anlagen appear during flexion (5.8 mm). Distinct bases are present by 7.0 mm, incipient rays by 7.2 mm, and ossification commences by 8.0 mm. The anal and sec- ond dorsal fins complete development by 13.0 mm. The Bruce: Larval development of Sillagmodes punctata, Sillago bassensis. and Sillago schomburgku Figure 2 Development of Sillaginodes punctata. (A) 2.1 mm; (B) 2.9 mm; (C) 3.1 mm; (D) 3.5 mm; (E) 3.6 mm; (F) 4.2 mm; (G) 5.8 mm; (H) 6.5 mm; (I) 8.5 mm; (J) 12.0 mm; (K) 18.7 mm postsettlement; (L) 22.4 mmlpostsettlement: myomeres omitted for clarity). A-I are reared specimens, J-L are field-collected specimens. first dorsal fin anlage is present by 6.2 mm. Distinct bases are present by 7.6—8.6 mm and ossification of spines has commenced by 8.5—8.9 mm. The first dorsal fin completes development by 13.1 mm. Pelvic fins first appear as slight swellings on either side of the gut in 9.2-mm larvae. Well-developed buds are present by 13.0 mm, incipient rays form shortly thereafter. The pelvic fin does not complete development until 20.0-21.5 mm. Larval pigment — The oil droplet is well pigmented with large stellate melanophores from at least 24 32 Fishery Bulletin 93(1), 1995 hours prior to hatching until yolk exhaustion. Newly hatched larvae have melanophores scattered over the body. Melanophores appear on the ventral and ante- rior regions of the yolk sac by 2.8 mm and pigment also appears within the finfold (both dorsal and anal) between myomeres 25—32 in reared larvae (not appar- ent in field-collected larvae — probably owing to finfold damage). Finfold pigment disappears by 3.5 mm. Initially, melanophores are scattered over the snout but they disappear by 3.5 mm. Pigment appears at the angle of the lower jaw and is retained throughout the larval period. Melanophores are typically present on the lower jaw, ventrally on the gular membrane, and internally below the otic capsule. Further pigment does not form on the head until after settlement. The dorsal surface of both the gut and the gas bladder are covered with melanophores during development. A linear series of discrete melanophores is present on the ventral midline of the gut in preflexion and flexion lar- vae. Ventral melanophores disappear from the hindgut by 10.0 mm and this region then remains unpigmented. Concurrently, the remaining 5—8 melanophores be- tween the cleithral symphysis and the hindgut become elongate and are retained in postsettlement larvae. Bruce: Larval development of Sillagmodes punctata. Sillago bassensis, and Sillago schomburgkit 33 Figure 2 (continued) By 4.0 mm, pigment on the dorsal surface of the trunk and tail coalesce to form 11-18 discrete, evenly placed melanophores that extend in a linear series posteriorly from the nape to within about 4 or 5 myomeres from the notochord tip. The dorsal sur- face of the notochord tip has 0-3 melanophores (most commonly 1 or 2) and when present they are useful in separating preflexion Sillaginodes punctata from Sillago bassensis and S. schomburgkii, both of which lack pigment dorsally on the notochord tip. The dor- sal series of melanophores on the trunk and tail gradually disappears by the end of flexion (6.5-7.0 mm), excepting those between myomeres 31-40, which become prominent and may extend laterally over the body surface when expanded. Lateral mid- line pigment develops in this area during late flex- ion and is retained throughout the postflexion stage. Dorsal pigment gradually redevelops in postflexion larvae as a series of discrete bands, each comprising 3 or 4 pairs of stellate melanophores. Postsettlement lar- vae have 4—6 such bands which subsequently increase in number to 8-10 as juvenile pigmentation develops. Ventral pigment on the tail in newly hatched lar- vae is initially scattered but coalesces to form a se- 34 Fishery Bulletin 93(1). 1995 Table 2 Body proportions of larvae of Sillaginodes punctata ( expressed as a percentage of body length), d = damaged; g = gas bladder not visible; — = character not yet formed. Specimens between dotted lines were undergoing flexion. Pre -gas - Vent to Body length Pre-anal Pre-dorsal bladder Head Snout Eye anal fin Body depth Body depth (mm) length fin length length length length diameter length at pectoral at anus 3.1 51.6 37.1 22.5 6.4 9.7 12.9 8.1 3.2 53.1 — 23.4 18.7 3.1 6.2 — 10.9 6.2 3.3 51.5 — 28.8 24.2 6.1 9.1 — 15.1 9.1 3.6 51.3 — 22.2 19.4 4.1 8.3 — 11.1 5.5 3.7 51.3 — 24.3 21.6 4.1 8.1 — 12.2 6.8 4.1 47.5 — 24.4 19.5 3.7 7.3 — 12.2 6.1 4.2 50.0 — 23.8 19.0 2.3 7.1 — 11.9 7.1 4.3 55.8 — 32.5 23.2 4.6 9.3 — 16.2 9.3 4.7 51.1 — 34.0 23.4 4.2 8.5 — 14.9 8.5 5.0 50.0 — 30.0 24.0 5.0 8.0 — 13.0 8.0 5.3 52.8 — 32.1 22.6 3.7 7.5 — 13.2 8.5 5.4 50.0 — 29.6 20.4 1.8 7.4 — 13.0 7.4 5.7 47.3 — 31.6 22.8 5.3 7.0 — 12.3 7.9 5.9 45.8 — 28.8 18.6 5.1 6.8 — 11.9 6.8 6.0 48.3 60.0 31.6 23.3 6.7 6.7 — 13.3 8.3 6.2 50.0 31.4 21.0 4.0 6.4 6.4 12.1 8.1 6.3 47.6 50.8 33.3 23.8 6.3 6.3 4.2 11.1 7.9 6.4 51.6 46.9 32.0 20.3 6.2 6.2 — 12.5 7.8 6.5 47.7 49.2 33.1 23.8 6.1 d 6.9 12.3 7.7 6.6 47.0 — 30.3 22.7 5.3 6.1 6.6 10.6 7.5 6.8 50.0 48.5 30.9 20.6 5.9 6.6 — 12.5 8.1 7.0 51.4 52.3 34.3 21.4 d d — 11.4 10.0 7.2 54.1 50.0 37.5 23.6 5.7 7.6 0.7 11.1 8.3 7.6 52.6 53.9 34.2 22.3 5.3 7.2 2.6 11.8 7.9 8.4 52.3 40.4 40.4 21.4 5.9 7.1 0.0 12.5 10.7 9.3 52.3 30.1 39.7 21.5 6.4 6.4 1.6 10.7 9.1 10.3 52.4 31.1 40.8 21.3 5.8 6.8 0.5 11.1 9.7 12.0 50.0 27.5 39.2 21.7 6.7 5.0 0.6 10.0 9.2 13.1 49.6 27.5 40.4 19.8 5.3 5.3 0.7 9.2 8.4 15.7 50.9 27.4 40.8 19.1 5.7 d 0.0 10.2 11.5 16.1 50.9 26.7 41.0 19.2 5.0 5.6 0.0 9.9 9.9 18.2 51.1 27.5 40.6 20.3 6.0 6.0 0.5 9.9 10.4 18.7 49.2 25.7 38.0 19.8 5.3 5.9 1.1 9.6 9.1 ries of closely spaced melanophores extending to the notochord tip by 3.6 mm. Preflexion larvae have 2—4 melanophores ventrally on the notochord tip. Dur- ing flexion, melanophores between myomeres 23-38 become more prominent (similar to the dorsal series). The ventral series of melanophores on the tail be- comes gradually obscured by overlying musculature (excepting the prominent region between myomeres 32-38) in postflexion larvae. Paired external melano- phores develop ventrally on the tail in larvae greater than 8.5 mm and by settlement stage, approximately one pair per myomere is present. This ventral series forms a regular pattern of expanded and contracted melanophores in postsettlement specimens, match- ing the banding pattern of the dorsal series. School whiting [Sillago bassensis Cuvier, 1 829), Figure 3 Materials examined— 40 specimens, 2.3-17.2 mm BL (CSIRO L586-01— 10 specimens). Larval development — The smallest S. bassensis larva examined was 2.3 mm BL. At this size the mouth and gut are functional, the eyes are pig- mented, a gas bladder is present, and yolk absorp- tion is complete. Larvae are elongate (BDp= 13-20% BL) and have 32-35 myomeres ( 11-15+19-23). Body depth at anus increases slightly from 8-12% BL during develop- ment. Other body proportions remain relatively con- stant (Table 3). The gut forms a convoluted tube in the smallest specimen and is already differentiated Bruce Larval development of Sillaginodes punctata, Sillago bassensis, and Sillago schomburgkn 35 Figure 3 Development of Sillago bassensis. (A) 3.2 mm; (B) 4.4 mm; (C) 4.8 mm; (D) 5.9 mm; (E) 7.2 mm; (F) 11.2 mm; (G) 12.7 mm (postsettlement). into fore, mid, and hindgut regions. The midgut be- comes rugose by 3.0 mm and remains so, although overlying musculature obscures this feature prior to settlement. The gut begins to coil in preflexion lar- vae by 4.1 mm. Coiling proceeds without migration of the anus and is achieved by elongation and ante- rior looping of the midgut (Fig. 4). Consequently, body proportions do not show a significant change in preanal length which remains at 47—48% BL. Coil- ing of the gut is completed in postflexion larvae (7.0- 7.5 mm). The gas bladder has its origin at myomeres 2-8 in preflexion larvae but migrates posteriorly during development to myomeres 5-10 in postflexion larvae. The gas bladder is inflated and prominent in 90% of field-collected larvae (random subsample n=40; all larvae were collected at night). 36 Fishery Bulletin 93|1). 1995 The snout is initially slightly concave in profile, but after flexion, this gradually changes to straight or slightly convex. The eye is round. The mouth ini- tially reaches to below the center of the eye but ex- tends only to the anterior margin of the eye in postflexion larvae. Four to six small villiform teeth are present on the premaxilla by 4.7 mm. The num- ber of teeth increases to 7 or 8 during flexion (4.8- 6.5 mm). Head spination is only weakly developed. A single minute preopercular spine is present by 7.8 mm but is not visible after settlement (12.5 mm). A weak posttemporal ridge is present by 7.2 mm and is retained; however, no posttemporal spines develop. The single opercular spine is first visible by 12.7 mm and is retained in juveniles. Scales develop after settlement and are first vis- ible around the gut and lateral midline of the tail by approximately 16.0 mm. The development of fins in larval and juvenile S. bassensis is summarized in Table 1. Completion of fin development occurs in the following sequence: caudal; pectoral; anal, first dorsal and second dorsal fins (almost simultaneously); and pelvic. The rays of the caudal fin are present just prior to flexion in larvae of 4.4 mm. Flexion commences by 4.8-5.0 mm and is complete by 6.8 mm. Pectoral fin buds are present in the smallest specimen (2.3 mm), incipient rays form during flexion (5.8-6.0 mm), and a full complement of 15 or 16 rays is present by 10.0 mm. Anal-fin and second-dorsal-fin anlagen appear during flexion. Distinct bases are present by 6.5—7.0 mm, incipient rays by 6.8-7.1 mm, and ossification of posterior rays has regularly commenced by 7.2 mm. Ossification of dorsal and anal elements proceeds anteriorly and both fins complete development by 10.0-10.5 mm. The first dorsal fin anlage is present by 7.0 mm. Distinct bases are present by 7.5-8.0 mm and ossification of spines has commenced by 8.0-8.5 mm. Development of the first dorsal fin is complete by 10.0-10.5 mm. Pelvic fin buds are present in 7.0- 7.2 mm larvae below the pectoral fin bases. Develop- ment of the pelvic fin is complete by 12.5 mm. Larval pigment — S. bassensis larvae were the least pigmented of the three sillaginid species examined. Pigment on the head in preflexion larvae is lim- ited to the angle of the lower jaw and internally to Bruce: Larval development of Sillagmodes punctata. Sillago bassensis, and Sillago schomburgkri 37 Table 3 Body proportions of larvae of Sillago bassensis (expressed as a percentage of body len gth). d = i iamaged; g = gas bladder not visible; — = character not yet formed. Specimens between dotted lines were undergoing flexion. Pre-gas- Vent to Body-length Pre-anal Pre-dorsal bladder Head Snout Eye anal fin Body depth Body depth (mm) 2.3 length fin length length length length diameter length at pectoral at anus 56.5 _ g 23.9 4.3 8.6 17.4 6.5 3.1 48.4 — 27.4 24.2 d d — 16.1 8.1 3.4 44.1 — 23.5 19.1 5.9 7.3 — 14.7 5.9 3.7 51.3 — 28.4 d 4.1 8.1 — 16.2 8.1 3.8 47.4 — 25.0 22.4 3.9 9.2 — 17.1 6.6 4.1 51.2 — 29.3 26.8 6.1 8.5 — 18.3 8.5 4.2 52.4 — 29.8 26.2 7.1 8.3 — 19.0 10.7 4.3 46.5 — g 23.2 5.8 7.0 — 13.9 6.9 4.4 52.3 — 34.1 28.4 6.8 9.1 — 20.4 11.4 4.5 50.0 — g 24.4 4.4 7.7 — 15.5 8.9 4.6 44.6 — 28.2 23.9 6.5 8.7 — 14.1 7.6 4.7 44.7 — 29.8 22.3 6.4 7.4 — 12.8 6.4 4.8 54.2 36.4 29.2 6.2 8.3 2.1 18.7 12.5 4.9 44.9 — 29.6 22.4 5.1 8.2 — 13.3 8.2 5.3 47.2 — 33.0 26.4 7.5 7.5 — 15.1 8.5 5.5 50.9 38.1 g 30.9 9.1 9.1 0.0 20.9 12.7 5.7 50.9 — 31.6 29.8 8.8 8.8 0.0 17.5 10.5 5.8 44.8 56.9 56.9 25.9 6.9 7.7 0.6 13.8 7.7 5.9 49.1 55.9 g 27.1 8.5 8.5 2.5 18.6 11.0 6.3 46.0 47.6 33.3 26.2 6.3 7.9 0.8 14. 8.7 7.7 49.3 54.5 32.5 25.3 7.8 7.8 0.0 18.8 13.6 7.9 45.6 53.2 32.9 25.3 6.3 7.6 2.5 13.9 8.9 8.9 51.7 38.2 34.3 28.6 7.8 7.8 0.0 18.5 16.3 9.6 47.9 33.3 37.5 25.0 7.3 8.3 0.0 14.6 12.5 10.1 44.5 30.7 g 22.8 4.9 7.9 0.9 14.8 10.9 10.2 45.1 31.3 g 24.5 5.9 8.3 1.0 14.7 10.8 11.3 46.9 33.6 33.6 28.3 7.1 8.0 0.0 16.8 13.3 Figure 4 Morphology of the gut in Sillago bassensis (ventral view of pigment omitted). (A) 2.9 mm; (B) 4.2 mm; (C) 5.5 mm. 38 Fishery Bulletin 93(1), 1995 below the otic capsule. Melanophores are irregularly present ventrally on the gular membrane. Additional pigment on the head does not develop until after settlement. Melanophores then develop immediately anterior to and above the eye as well as on the snout and lower jaw. Larger specimens quickly develop a cap of melanophores over the mid and hindbrain. Pigment on the dorsal surface of the gut consists of 2-7 approximately evenly spaced melanophores in preflexion larvae. This reduces to 2 or 3 just prior to flexion. In postflexion larvae, internal pigment over the gut is restricted to above the gas bladder. Ventral pigment on the gut consists of a midline se- ries of 8-14 melanophores extending from just ante- rior to the cleithral symphysis to the anus in both preflexion and flexion larvae. One to two additional melanophores are usually present either side of this series below the level of the pectoral fin base (76% of larvae, random subsample «=25), forming a diamond pattern when viewed ventrally (Fig. 5). Preflexion larvae have 10-18 discrete, evenly placed melanophores that extend in a dorsal linear series on the trunk and tail to within 1-3 myomeres of the notochord tip. The dorsal surface of the noto- chord tip remains unpigmented throughout devel- opment. The dorsal series of melanophores gradu- ally disappears during flexion (4.9—6.5 mm). Postflexion larvae have 0-3 melanophores (most com- monly 0) below the bases of the second dorsal fin. Dorsal pigment redevelops after settlement as a se- ries of discrete bands each comprising 3-6 pairs of stellate melanophores. The first of these bands de- velops immediately below the posterior-most second dorsal fin rays, 5 or 6 additional bands subsequently Figure 5 Ventral pigment on the gut: (A) Sillaginodes punctata, 4.7 mm; Sillago schomburgkii, 4.4 mm; (C) Sillago bassensis, 4.1 mm. (B) develop anteriorly, and a single band developes pos- teriorly on the caudal peduncle by 20.0 mm. Lateral midline pigment on the tail does not form until after settlement, although some internal pigment may be present over vertebrae between myomeres 25-30 after 11.0 mm. A single row of 14—19 melanophores is present along the ventral midline of the tail in preflexion lar- vae. This ventral row is gradually obscured by over- lying musculature during flexion. Paired external melanophores subsequently develop ventrally on the tail in postflexion larvae, approximately one pair per myomere. After settlement, this ventral series forms a regular pattern of expanded and contracted mel- anophores producing a similar banding pattern to the dorsal series. One to two (most commonly 2) mel- anophores are present ventrally on the notochord tip in preflexion larvae. These are retained in postflexion larvae and, with additional melanophores, form a band of pigment over the caudal-fin ray bases. Yellow fin whiting [Sillago schomburgkii Peters 1865), Figure 6 Material examined — 16 specimens, 2.7-18.7 mm BL. Larval development — The smallest S. schomburg- kii examined was 2.7 mm. At this size the mouth and gut are functional, the eyes are pigmented, a gas blad- der is present, and yolk absorption is complete. Larvae are elongate (BDp= 14-18% BL) and have 36-38 myomeres (15-17+20-22). Body depth at anus increases from 8 to 16% BL during development. Other body proportions remain relatively constant (Table 4). The gut forms a convoluted tube in the smallest specimen and is already differenti- ated into fore, mid and hindgut regions. The midgut becomes rugose by 4.4 mm and re- mains so, although overlying musculature obscures this feature prior to settlement. The gut has not begun coiling in the largest flex- ion-stage larva available (5.1 mm). Coiling of the midgut has begun in the 10.1-mm larva and is well developed in all postsettlement larvae. Insufficient specimens were available to further document the timing of gut coil- ing. Coiling of the gut proceeds without mi- gration of the anus and is achieved by elon- gation and anterior looping of the midgut. Consequently, body proportions do not show a significant change in preanal length which remains at 51-53% BL. The gas bladder has its origin at myomeres 1-8 in preflexion lar- vae and is inflated and prominent in all larvae collected during night tows. The gas bladder is inconspicuous in larvae caught during the day. Bruce: Larval development of Sillaginodes punctata, Sillago bassensis, and Sillago schomburgkn 39 D Figure 6 Development of Sillago schomburgkii. (A) 2.7 mm; (B) 4.4 mm; (C) 5.0 mm; (D) 10.1 mm; (E) 13.0 mm (postsettlement). 40 Fishery Bulletin 93(1), 1995 The snout is initially slightly concave in profile, but after flexion this gradually changes to straight or slightly convex. The eye is round. The mouth ini- tially reaches below the eye but is short of the eye in postflexion larvae. Four to six small villiform teeth are present on the premaxilla by 4.4 mm. The num- ber of teeth increases from 10 to 12 during flexion (from 4.8 to greater than 5.1 mm). Head spination is only weakly developed. One to two preopercular spines are discernible in postsettlement larvae. A weak posttemporal ridge with 1 or 2 small spines is developed in the 10.1-mm postflexion larva and is present in all postsettlement larvae examined. The single opercular spine is not visible in the 10.1-mm larva but is present in postsettlement larvae and is retained in juveniles. Scales develop after settlement and are first vis- ible around the gut and lateral midline by 17.2 mm. Insufficient numbers of specimens were available to document the full sequence of fin development or the completion of flexion in S. schomburgkii. The rays of the caudal fin are present in flexion larvae of 4.8 mm. Flexion commences by 4.8 mm. Pectoral fin buds are present in the smallest speci- men (2.7 mm) and incipient rays form during flex- ion. Rays of the pectoral fin have commenced ossifi- cation in the 10.1-mm postflexion larva. A full comple- ment of 15 or 16 pectoral fin rays is present in the smallest postsettlement larva (12.7 mm). Anal-fin and second-dorsal-fin anlagen appear during flexion. Full complements (spines and rays) of the anal fin and both dorsal fins are present in the 10.1-mm speci- men. The pelvic fin has commenced development in the 10.1-mm larva and has completed development by 12.7 mm. Larval pigmentation — Pigment on the head in preflexion S. schomburgkii larvae is limited to the angle of the lower jaw and internally to the base of the otic capsule. One or two melanophores are also present ventrally on the gular membrane, increas- ing to three during flexion. Additional melanophores develop on the snout tip, scattered over the lateral surface of the head, and a cap of pigment forms over the mid and hindbrain in postflexion larvae. Pigment on the dorsal surface of the gut and gas blad- der consists of 8—10 approximately evenly spaced melanophores. Scattered internal melanophores gradu- ally spread over the lateral walls of the gut in post- flexion larvae. Ventral pigment on the gut consists of a midline series of 8-14 melanophores extending from just anterior of the cleithral symphysis to the anus (Fig. 5). Preflexion larvae have 15-22 discrete, evenly spaced melanophores that extend in a dorsal linear series from the nape to within 2-5 myomeres of the notochord tip. The number of dorsal melanophores decreases to 13 by 5.0 mm. A series of three discrete dorsal bands consisting of 3-5 paired stellate mel- anophores has replaced this dorsal series in the 10.1 mm postflexion larva. Lateral midline pigment in S. schomburgkii larvae is the most pronounced of all three species examined and is present on the tail in the smallest larva (2.7 mm) as 2 or 3 elongated mel- Table 4 Body proportions of larvae of Sillago schomburgkii (expressed as a percentage of body length), d = damaged; g = gas bladder not visible; — = character not yet formed. Specimens between dotted lines were undergoing flexion. Pre-gas- Vent to Body length Pre-anal Pre-dorsal bladder Head Snout Eye anal fin Body depth Body depth (mm) length fin length length length length diameter length at pectoral at anus 2.7 53.7 24.1 20.4 3.7 9.2 14.8 6.3 3.3 51.5 — 25.7 21.2 5.1 9.1 — 15.1 7.6 3.6 48.6 — g 25.0 6.1 8.3 — 13.9 8.3 3.7 50.0 — g 24.3 5.4 8.1 — 16.2 9.4 3.8 51.3 — 25.0 22.4 2.6 7.9 — 14.5 7.9 3.9 52.3 — 28.2 22.3 6.4 7.7 — 15.4 7.7 4.0 51.2 — g 25.0 7.5 7.5 — 17.5 9.0 4.3 52.3 — g 22.1 5.8 8.1 — 15.1 9.3 4.4 53.4 — 28.4 22.7 5.7 8.4 — 14.8 7.9 4.8 53.1 31.2 25.0 6.2 8.3 15.6 10.4 5.0 53.0 — 34.0 25.0 7.0 9.0 8.0 18.0 11.0 10.1 52.5 34.6 g 27.7 8.9 6.9 2.0 17.8 15.8 13.6 49.3 32.3 g 27.2 7.3 8.1 3.6 16.9 14.7 17.2 53.5 36.0 g 30.2 9.3 8.1 2.9 17.4 14.5 Bruce: Larval development of Sillagmodes punctata, Sillago bassensis, and Sillago schomburgkn 41 anophores in the vicinity of myomeres 24-26. Lat- eral midline pigment spreads both anteriorly and posteriorly as a linear series of elongated myomeres during development. By 10.1 mm, lateral midline pigment consists of 18 stellate and approximately evenly spaced melanophores extending from the pec- toral fin to the caudal peduncle. Internal pigment along the vertebrae is visible in the 10.1-mm post- flexion larva but is most pronounced in post- settlement larvae as clusters of melanophores located over every 2-5 vertebrae. A single row of 16-18 melanophores is present along the ventral midline of the tail in preflexion lar- vae. This ventral row is gradually obscured by over- lying musculature during flexion. Paired external melanophores (approximately one pair per myomere) subsequently develop ventrally on the tail in post- flexion larvae, approximately one per myomere. Two to three (most commonly three) melanophores are present ventrally on the notochord tip in preflexion larvae. These are retained in postflexion larvae and form a band of pigment over the caudal-fin ray bases. Discussion Egg or larval development, or both, have been de- scribed for only four other species of sillaginid lar- vae: Sillago japonica (Kamiya, 1925; Ueno and Fujita, 1954; Ueno et al., 1958; Mito, 1966 — as Sil- lago japonicus; Ikeda and Mito, 1988; Kinoshita, 1988; Oozeki et al., 1992); Sillago sihama (Gopinath, 1946; Uchida et al., 1958; Ikeda and Mito, 1988; Kino- shita, 1988); Sillago maculata (Miskiewicz, 1987; Kinoshita, 1988); and Sillago ciliata (Munro, 1945; Miskiewicz 1987; Tosh 4 ). In addition, Miskiewicz (1987, p. 62) reported a series of unidentified sillaginid larvae which, based on pigment on the lat- eral wall of the gut below the pectoral fin base, were almost certainly Sillago flindersi. Characters useful for the identification of tropical sillaginid larvae at the family level and similarity of sillaginid larvae to those from other families have been considered in detail by Leis and Trnski (1989) and Miskiewicz ( 1987). Although most of the charac- ters discussed by these authors also apply to the tem- perate species considered here, an exception was the timing of gut coiling. Leis and Trnski ( 1989) reported that the gut of tropical sillaginid larvae commenced coiling during notochord flexion and was accompa- nied by the anterior migration of the anus. In the South Australian species, coiling of the gut com- menced prior to flexion in S. bassensis, after settle- ment in Sillaginodes punctata, and had not yet com- menced in the largest flexion larva available for Sillago schomburgkii (although coiling of the gut was present in a 10. 1-mm postflexion larva ). In all cases, coiling of the gut proceeded without migration of the anus and was achieved by anterior looping of the midgut. The implications of these variations are un- clear but suggest that, although useful on a specific level, the timing of gut coiling and migration of the anus are not useful characters for the identification of temperate sillaginids at the family level. The significance of gut coiling may relate to shifts in diet. Robertson (1977) reported a dietary shift in postsettlement Sillaginodes punctatus (-punctata) in Westernport Bay (Victoria) between November and December, a shift from harpacticoid copepods, gammarid amphipods, and mysids to larvae of the ghost prawn Callianassa australiensis, polychaetes, and juvenile crabs. Robertson correlated this dietary shift with increasing body size and mouth gape as well as with the availability of C. australiensis lar- vae. However, from his length-frequency data, this period also corresponds to the size range during which postsettlement S. punctata undergo gut coil- ing. Alternatively, because evacuation rates are be- lieved to decrease after gut coiling (Arthur, 1976, and references within; Young 5 ), perceived changes in diet may be confounded by increased food retention times. Stomach contents were not analyzed during this study; they provide a valuable topic for further re- search. Despite seasonal sampling over five years, larvae of only three of the four sillaginid species with adult distributions extending to South Australia were lo- cated during this study. The lack of Sillago flindersi larvae suggests either that this species does not spawn in South Australian waters, that sampling frequency was too course to detect the presence of larvae of this species, or that S. flindersi larvae be- have differently from other sillaginid species and are less prone to capture (e.g. epibenthic and neustonic). Sillago flindersi larvae are frequently encountered in similar sampling regimes in coastal waters of east- ern Australia 6 and in Tasmanian waters (author's pers. observ.) and thus it seems unlikely that a lack of their larvae in South Australian samples repre- sents an artifact of sampling or that their behavior is fundamentally different from other sillaginid larvae. 4 Tosh, J. R. 1903. Notes on the habits, development etc. of the common food fishes of Moreton Bay. Queensland Marine Dep.: Marine Biologist's Report. 5 Young, J. W. CSIRO Div. Fisheries, GPO Box 1538 Hobart, Tas- mania, Australia 7001. Personal commun., 1993. 6 Miskiewicz, A. G. Sydney Water Board, PO Box A53, Sydney South, NSW, Australia 2000. Personal commun., 1993. 42 Fishery Bulletin 93(1), 1995 South Australian waters represent the western distributional limit of S. flindersi in southern Aus- tralia (McKay, 1992; Gomon et al., 1994; Kailola et al., 1993). Spawning times for S. flindersi vary throughout its range; a summer spawning is recorded for Victorian populations (Hobday and Wankowski 7 ). No data are available on either the reproductive con- dition of S. flindersi or on the presence or absence of juveniles in South Australian waters. However, on the basis of a lack of larvae, I suggest that spawning does not occur in South Australia and that the west- ern limit of S. flindersi comprises fish recruited from eastern populations. Acknowlegments I would like to thank G. K. Jones, S. A. Shepherd, A Miskiewicz, J. M. Leis, and A. J. Butler for their help- ful comments on the manuscript. Thanks are also due to A. R. Knight, R. Hudson, R. Moehring, and D. A. Short for their help in collecting and processing plankton samples. Literature cited Arthur, D. K. 1976. Food and feeding of larvae of three fishes occurring in the California Current, Sardinops sagax, Engraulis mordax, and Trachurus symmetricus . Fish. Bull. 74:517-529. Bruce, B. D. 1988. Larval development of blue grenadier, Macruronus novaezelandiae (Hector), in Tasmanian waters. Fish. Bull. 86:119-128. 1989. Studying larval fish ecology: an aid to predict future catches. SAFISH 13(4):4-9. Burchmore, J. J., D. A. Pollard, M. J. Middleton, J. D. Bell, and B. C. Pease. 1988. Biology of four species of whiting (Pisces: Sillaginidae) in Botany Bay, New South Wales. Aust. J. Mar. Fresh- water Res. 39:709-727. Gilmour, A. J. 1969. The ecology of King George whiting Sillaginodes punctatus (Cuvier and Valenciennes) in Westernport Bay, Victoria. Ph.D. thesis, Monash Univ., Australia. Gopinath, K. 1946. Notes on the larval and post-larval stages of fishes found along the Trivandrum coast. Proc. Indian Natl. Sci. Acad. 12(1):7-21. Gomon M. F, C. J. M. Glover, and R. H. Kuiter. 1994. The fishes of Australia's south coast. State Print, Adelaide, Australia, 992 p. Hay, D. E. 1981. Effects of capture and fixation on gut contents and body size of Pacific herring larvae. Rapp. P.-V. Reun. Cons. Int. Explor. Mer 178:395-400. 7 Hobday, D. K., and J. W. J. Wankowski. 1987. School whiting Sillago bassensis flindersi: reproduction and fecundity in east- ern Bass Strait, Australia. Victorian Dep. Conserv., Forests and Lands, Fisheries Div. Int. Rep. 153, 24 p. Ikeda, T., and S. Mito. 1988. Pelagic fish eggs. In M. Okiyama (ed.), An atlas of the early stage fishes in Japan, p. 999-1083. Tokai Univ. Press, Tokyo. Johnson, G. D. 1984. Percoidei: development and relationships. In H. G. Moser, W. J. Richards, D. M. Cohen, M. F. Fahay, A. W. Kendall Jr. , and S. L. Richardson ( eds. ), Ontogeny and sys- tematics of fishes, p. 464-498. Am. Soc. Ichthyol. Herpetol., Spec. Publ. 1. Jones, G. K. 1980. Research on the biology of the spotted ( King George ) whiting in South Australian waters. SAFIC 4:3-7. Kailola, P. J., M. J. Williams, P. C. Stewart, R. E. Reichelt, A. McNee, and C. Grieve. 1993. Australian fisheries resources. Bureau of Rural Sciences, Fisheries Res. and Develop. Corp., Canberra, Australia, 422 p. Kamiya, N. 1925. Description of pelagic fish eggs and their larvae in Tate- yama Bay. J. Imperial Fish. Inst. 21: 71-85. [In Japanese.] Kinoshita, I. 1988. Sillaginidae. In M. Okiyama (ed.), An atlas of the early stage fishes in Japan, p. 449-452. Tokai Univ. Press, Tokyo. Last, P. R., E. O. G. Scott, and F. Talbot. 1983. Fishes of Tasmania. Tasmanian Fisheries Develop- ment Authority, Hobart, Australia, 563 p. Leis, J. M., and T. Trnski. 1989. The larvae of Indo-Pacific shore fishes. Univ. New South Wales Press, 371 p. Lennanton, R. C. J. 1969. Whiting fishery — Shark Bay. Fishing Industry News Service. W. Aust. Dep. Fish. Wildl. 2(1):4-11. McKay, R. J. 1985. A revision of the fishes of the family Sillaginidae. Mem. Queensl. Mus. 22: 1-73. 1992. FAO species catalogue. Vol. 14: Sillaginid fishes of the world. FAO, Rome., 87 p. Miskiewicz, A. G. 1987. Taxonomy and ecology of fish larvae in Lake Macquarie and New South Wales coastal waters. Ph.D. thesis, Univ. New South Wales. Mito, S. 1966. Fish eggs and larvae. In S. Motoda (ed.), Illustra- tions of the marine plankton of Japan, Vol. 7. Soyosha, Tokyo, 74 p. [In Japanese.] Munro, I. S. R. 1945. Postlarval stages of Australian fishes — No. 1. Mem. Queensl. Mus. 12:136-153. Oozeki, Y., P. Hwang, and R. Hirano. 1992. Larval development of the Japanese whiting, Sillago japonica. Jpn. J. Ichthyol. 39:59-66. Potthoff, T. 1984. Clearing and staining techniques. In H. G. Moser, W. J. Richards, D. M. Cohen, M. F. Fahay, A. W Kendall Jr., and S. L. Richardson (eds.), Ontogeny and systematics of fishes, p. 35-37. Am. Soc. Ichthyol. Herpetol., Spec. Publ. 1. Robertson, A. I. 1 977. Ecology of juvenile King George whiting Sillaginodes punc- tatus (Cuvier and Valenciennes) (Pisces: Perciformes) in West- ern Port, Victoria. Aust. J. Mar. Freshwater Res. 28:35-43. Scott, T. D. 1954. The life history of the spotted whiting, Sillaginodes punctatus (Cuvier and Valenciennes) in South Australia. M.Sc. thesis, Univ. Adelaide, Australia. Bruce: Larval development of Sillaginodes punctata, Sillago bassensis, and Sillago schomburgkn 43 Theilacker, G. H. 1980. Changes in body measurements of larval northern anchovy, Engrauhs mordax, and other fishes due to han- dling and preservation. Fish. Bull. 78:685-692. Uchida, K., S. Imai, S. Mito, S. Fujita, M. Ueno, Y. Shojima, T. Senta, M. Tahaku, and Y. Dotu. 1958. Studies on the eggs, larvae and juveniles of Japanese fishes. Series 1: Second laboratory offish biology. Fish. Dep. Fac. Agric, Kyushu Univ., Fukuoka, Japan. [In Japanese.] Ueno, M., and S. Fujita. 1954. On the development of the egg of Sillago sihama (Ferskal). Jpn. J. Ichthyol. 3:118-120. Ueno, M., T. Senta, and S. Fujita. 1958. Sillago sihama (Ferskal). In Kyushu University eds., Second laboratory of fisheries biology: studies on the eggs, larvae and juveniles of Japanese fishes. Shyuk- ousha, Fukuoka. [In Japanese.] Weng, H. T. 1983. Identification, habitats and seasonal occurrence of juvenile whiting (Sillaginidae) in Moreton Bay, Queens- land. J. Fish Biol. 23:195-200. 1986. Temporal distribution of whiting (Sillaginidae) in Moreton Bay, Queensland. J. Fish Biol. 29:755-764. Abstract. Female and sublegal-size male Tanner crabs, Chionoecetes bairdi, are often caught incidentally in the males- only fishery for this species. Effects of low air temperature during the winter fishery on juvenile and fe- male adult crabs and on the devel- oping eggs brooded by the females were simulated in the laboratory by exposing crabs to cold air (-20 to +5°C) up to 32 minutes; controls were not exposed. Exposure was expressed as degree-hours (°h), the product of temperature (°C) and time (hours). Severe exposure caused death: median lethal expo- sure stabilized at -3.3 + 0.8°h for juveniles and —4.3 ± 0.5°h for adults after 16 days. Exposure also re- duced vigor (measured by righting ability), caused pereiopod auto- tomy, and depressed adult feeding rates and juvenile growth. Expo- sures causing one-half the crabs to cease righting were —1.2 ± 0.3°h for juveniles and -2.1 ± 0.3°h for adults (measured immediately af- ter exposure). Mean pereiopod au- totomy ranged up to 44% for juve- niles exposed to -2°h, and up to 10% for adults exposed to -10.6°h. Ecdysis of juveniles was not af- fected, but exposed juveniles fre- quently shed additional pereiopods with the molt. Prompt return of incidentally caught Tanner crabs to the sea when temperatures are be- low freezing should reduce adverse effects of cold aerial exposure. Responses of Tanner crabs, Chionoecetes bairdi, exposed to cold air Mark G. Carls Charles E. O'Clair Auke Bay Laboratory, Alaska Fisheries Science Center National Marine Fisheries Service, NOAA 1 1305 Glacier Highway Juneau, Alaska 99801-8626 Manuscript accepted 23 May 1994. Fishery Bulletin 93:44-56 (1995). Tanner crabs, Chionoecetes bairdi Rathbun, 1893, are the target of a large commercial pot fishery and are an important commercial spe- cies in Alaskan waters (Otto, 1989). Landings of C. bairdi rose to a peak of 57,923 metric tons (t) in 1978, then declined to 5,390 t in 1987; landings increased to 23,507 t in 1990. 1 Current Alaska fishing regula- tions require release of small (<139- mm carapace width) male and all female C. bairdi. Commercial fish- ery openings in recent years have generally ranged from November through April, 2 when minimum daily air temperatures can drop to -21°C. 3 The amount of time inciden- tally captured crabs remain on deck varies, ranging from a few minutes during pot fishing to hours in some trawling operations (Stevens, 1990). Exposure to cold air during fishing operations may be detrimental to individual crabs (Carls and O'Clair, 1990), exposed egg clutches, and possibly — with sufficient fishing pressure — to the population. Regu- lations also require that Tanner crabs caught incidentally by multi- species trawling operations in the eastern Bering Sea be returned to the sea, but these regulations may be ineffective because of poor survival (22 + 3.6% for C. bairdi) of the culled crabs (Stevens, 1990). Here we report the responses of juvenile and adult female Tanner crabs and their offspring exposed to cold air. Our objectives were to determine the effects (immediate 1 Kruse, G. Alaska Dep. Fish and Game, Div. Commer. Fish., Juneau, AK 99802. Pers. commun., July 1992. 2 ADF&G (Alaska Department of Fish and Game). 1989a. Report to the Alaska Board of Fish- eries. Southeast Alaska and Yakutat (Re- gion 1) 1988/89 shellfish fisheries. Regional Information Rep. No. 1J89-01. ADF&G, Div. Commercial Fisheries, Juneau, AK. 1989b. Westward region shellfish report to the Alaska Board of Fisheries. ADF&G Regional Information Rep. No. 4K89-3. ADF&G, Div. Commercial Fisheries, Westward Regional Office, 211 Mission Rd., Kodiak, AK 99615, 325 p. 1989c. Prince William Sound management area shellfish report to the Alaska Board of Fisheries. ADF&G Regional Informa- tion Rep. No. 2C89-03. ADF&G, Div. Commercial Fisheries, Central Region, 333 Raspberry Rd., Anchorage, AK 99581, 55 p. 1989d. Cook Inlet area shellfish manage- ment report to the Alaska Board of Fish- eries, 1988-89. Regional Information Rep. No. 2H89-03. ADF&G, Div. Com- mercial Fisheries, 333 Raspberry Rd., Anchorage, AK 99581, 75 p. 1989e. Synopsis of the Montague Strait ex- perimental harvest area 1985-1988. ADF&G Regional Information Rep. No. 2C89-04. ADF&G, Div. Commercial Fish- eries, Central Region, 333 Raspberry Rd., Anchorage, AK 99581, 21 p. 1989f. Report to the Board of Fisheries Norton Sound red king crab fishery (sum- mer fishery only). ADF&G Regional In- formation Rep. No. 3N89-05. ADF&G, Div. Commercial Fisheries, Central Re- gion, Juneau, AK, 14 p. 3 NOAA. 1987. Local climatological data, monthly and annual summaries with com- parative data. U.S. Dep. Commer., Na- tional Climatic Data Center, Asheville, NC 28801. 44 Carls and O'Clair: Responses of Chionoecetes bairdi to cold air 45 and long-term) of exposure to cold air on 1) survival; 2) sublethal responses, including righting response, limb autotomy, feeding rate, ecdysis (juveniles), and growth; and 3) reproductive responses including egg survival, zoeal production, zoeal viability, and subse- quent egg extrusion and viability of the extruded clutch. Methods Experimental crabs were collected with crab pots. Juvenile crabs (both sexes) were collected in Auke Bay, Alaska (lat. 58°21'N, long. 134°41'W) on 14 and 19 January 1988. Ovigerous females were captured near Eagle River (lat. 58°31'N, long. 134°48'W) and Lena Point (lat. 58°24'N, long. 134°47'W) in Favorite Channel, Alaska, on 11 February 1988. In the laboratory, carapace length (distance from the posterior margin of the right ocular orbit to the midpoint of the posterior margin of the cara- pace) was measured to the nearest millime- ter. Carapace width was subsequently esti- mated by regressing carapace widths and lengths of Tanner crabs measured at a later date. 4 Live weight was measured to the near- est 0.1 g. Juvenile crab weights ranged from 26 to 229 g (* = 109 ±14 g), and carapace lengths ranged from 35 to 64 mm (3c=49 ±2.3 mm) (Fig. 1). Estimated juvenile cara- pace widths (for both sexes) ranged from 46 to 74 mm (width=-0.237 + 1.318 x length, r 2 =0.994, rc=145). The immature condition of males was determined solely by body size. Adult female crab weights ranged from 182 to 553 g ( x = 329 ± 8 g), and carapace lengths ranged from 65 to 96 mm (x=80 ±1.0 mm) (Fig. 1). Estimated female carapace widths ranged from 85 to 124 mm (width=1.746 x 1.274 x length, r 2 =0.995, n=70). Crabs were maintained in 500-L tanks at ambient seawater temperatures (6.0-6.9°C for juveniles, 5.3-6.0°C for adults) until ex- posure to test air temperatures; after expo- sure they were returned to the same tanks for 32-35 days of observation (4.7-6.7°C for juveniles, 4.7-5. 2°C for adults). A subset of 40 female crabs was retained for an addi- tional three months of observation. Crabs were exposed in a modified chest freezer divided by a vertical baffle into two compartments of unequal size (Carls and O'Clair, 1990). Infrared heat lamps were placed in the smaller compartment for temperature control. To ensure uniform temperatures, a small fan (in the center bottom of the baffle) drew air from the exposure chamber into the small chamber at 45 ± 5 cm/sec. Return air circulated over the baffle into the exposure chamber. Temperatures were measured with a thermistor located in the exposure area near the fan and were regulated manually by switching the heat lamps on or off. Temperatures were con- trolled to ±0. 1°C after the chamber had cooled to the desired temperature. Crabs were exposed to cold air on the plywood bottom of the exposure chamber. Juvenile crabs were randomly placed in six groups with 10 crabs per group and were exposed to cold air on 21 and 25 January (about one week after capture). Exposure temperatures ranged from -5.0 to -20.0°C; exposure durations were 0, 12, 16, or 24 minutes to yield 0, -1.0, -1.5, -2.1, -4.0, and -8.0°h exposures (Table 1). The lengths (F 5 54 =0.06, P>0.99) and 25 r ~n ~\ 20 T 200 260 320 380 Wet weight (grams) Ik _ r _r 500 40- 30- 10- Adult females Juveniles dcuxi 30 17/ Tilm rl i i 50 60 70 Carapace length (mm) n Figure 1 Chionoecetes bairdi length and weight frequencies. Adult female frequencies may not be directly comparable to juvenile frequencies because they were taken from a different locality. 4 r^O.99. Stone, R. NMFS Auke Bay Lab., Juneau, AK 99801-8626. Unpubl. data, May 1992. 46 Fishery Bulletin 93(1), 1995 Table 1 Temperature and duration of exposure of Chionoecetes bairdi to cold air. The number of crabs exposed (n) is also indicated. Controls were not exposed to air. SE = standard error. Air temperature Exposure (Celcius) time mean SE (minutes) Degree-hours n Juveniles — — 0.00 10 -5.0 0.02 12 -0.99 10 -7.5 0.05 12 -1.50 10 -10.2 0.24 12 -2.05 10 -15.0 0.03 16 -4.00 10 -20.0 0.07 24 -8.02 10 Adults 5.1 0.12 8 0.683 7 5.0 0.01 32 2.672 7 — — 0.000 31 -3.2 0.21 4 -0.211 8 -3.1 0.04 8 -0.411 8 -3.1 0.06 16 -0.813 8 -3.0 0.03 32 -1.621 8 -8.2 0.19 4 -0.544 8 -8.1 0.11 8 -1.075 8 -8.1 0.06 16 -2.149 8 -8.1 0.03 32 -4.299 7 -13.1 0.18 4 -0.875 8 -12.9 0.08 8 -1.720 8 -13.0 0.03 16 -3.472 7 -13.0 0.02 32 -6.933 8 -20.3 0.34 4 -1.353 8 -20.1 0.15 8 -2.676 8 -18.4 0.08 16 -4.899 8 -19.9 0.04 32 -10.597 8 weights (F 554 =0.02, P>0.99) of the crabs did not dif- fer significantly between treatments. Change in ju- venile crab body weight was estimated from initial and final measurements (32 d). Female crabs were randomly placed in 20 groups (including controls) in a complete 4 (temperature) by 5 (length of exposure) design, with 7 to 8 crabs per group. Treatment temperatures ranged from -3.1 to -20.3°C and exposure duration ranged from (controls) to 32 minutes (Table 1). Two additional groups were tested at 5°C for 8 and 32 minutes (Table 1). The crabs did not differ significantly in length (F 21U9 =1.13, P=0.324) or weight (P. 21 149 =1.36, P=0.149) between treatments. Exposure took place 16 and 17 February (about six days after capture). Observation continued through 22 June. Mortality and limb autotomy were monitored daily. Crabs were judged dead when scaphognathite move- ment stopped. Generally, dead crabs were rechecked the following day before they were removed from test tanks. The number of legs missing on each crab was counted and autotomized legs were removed from the tanks. Righting response (the time it took a crab to right itself when placed on its back underwater), which we considered to be a measure of vigor, was timed to the nearest 0.01 second immediately after aerial ex- posure and 1, 2, 4, 8, 16, 24, and 32 days thereafter. Crabs that could not right themselves after 2 min- utes were recorded as "not righting" and were placed upright in the tank. A subset of 40 female crabs randomly selected from the entire exposure range was used for reproductive observations. The crabs were isolated 32 days after exposure in covered 70-L tanks that overflowed into 19-L buckets containing conical 363-u mesh nets de- signed to trap zoeae. Flow rates were approximately 1.5 L/minute; 95% turnover time was 2.3 hours and water temperatures ranged from 5.2 to 5.9°C during this period (23 March-11 May). Feeding rates were measured before and after the zoeal hatch while the 40 ovigerous females were in- dividually isolated. Mussels, Mytilus trossulus, were fed ad libitum to crabs during each feeding period. Live mussels were cut in half and drained tissue- side down on paper towels for five minutes, weighed, then placed in the tanks. Twenty-four hours later the remaining food was removed, drained, and weighed as before. At each feeding, four food portions were placed as controls in tanks without crabs. Con- sumption was corrected for the mean weight changes in the control portions. Feeding observations were repeated every 1 to 3 days, from 41 to 60 and from 85 to 98 days after exposure. Zoeae were collected daily, rinsed from the nets, concentrated in a known volume, and subsampled with a 5- or 10-mL Hensen-Stemple pipette (Carls and O'Clair, 1990). Subsamples, which contained a minimum of 200 zoeae, were preserved in 5% forma- lin and counted later; the occasional large subsample was divided with a Folsom plankton splitter before being counted. After zoeal hatching, all debris from each tank bottom was preserved to determine the number of dead eggs and zoeae. Responses of the crabs were related to aerial ex- posure, expressed as the product of air temperature (°C) and length of time in air (hours), i.e. degree-hours (°h). In a similar experiment, Carls and O'Clair ( 1990) demonstrated the usefulness of this technique for interpreting responses to aerial exposure in adult king crabs, Paralithodes camtschaticus. Because the responses of the Tanner crabs to exposure (in °h) were similar in form to those of the king crabs over identi- Carls and O'Clair: Responses of Chionoecetes bairdi to cold air 47 cal treatment ranges (0-32 minutes, -20 to +5°C; see Results section) and could be described by the same types of simple linear or nonlinear models, we used the same technique here. Regression techniques and logit analysis were used to relate response variables to exposure (Berkson, 1957; BMDP, 1983). We compared median lethal re- sponses with log-likelihood ratio tests (Fujioka, 1986). Multiple regression was used to test for dif- ferences in the slopes of regression lines and to ad- just for covariates (Kleinbaum and Kupper, 1983). The relation of selected response variables to one another was tested with parametric correlation. Af- ter one-way analysis of variance, comparisons of treatment means were made with Tukey's or Dunnett's a posteriori multiple comparison tests and judged significantly different if P<0.05. Proportional data were arcsine transformed. Reported error ranges are ±95% confidence limits. Results Mortality Below -1 to -3 degree hours, exposure to cold air killed crabs. Almost all mortality occurred 1-2 days after exposure; in groups where more than half the crabs died, mortality always reached 50% within 2 days. Mortality was inversely related to exposure and increased rapidly below -l°h for juveniles and below -3°h for adults (logistic regressions [large P- values indicate good fits], P Juvenile =0.959, P adul =0.882; Fig. 2). Nearly all deaths occurred within 8 days af- ter exposure; no crabs died after day 16. For juve- niles, calculated median lethal exposures rose from -7.7 ± 3.4°h 1 day after exposure to -3.3 ± 0.8°h 16 days after exposure, and for adults from -7.2 ± 1.6°h to -4.3 ± 0.5°h over the same time period (Table 2). Righting response The speed with which crabs righted themselves when placed on their backs was inversely related to expo- sure (Fig. 3A). The response was most clearly de- scribed by the percentage of crabs not righting within two minutes (logistic regressions, P=0.799 [n=6] for juveniles; P=0.978 [ra=22] for adults; Fig. 3B). Per- centages of crabs not righting increased sharply be- low — 1.0°h for juveniles and below -2.2°h for adults, and crabs ceased righting entirely after exposure to <-4.0°h for juveniles and <-6.9°h for adults (Fig. 3B). Median exposures causing one-half the crabs to cease righting (EC50) were -1.2 ± 0.3°h for juveniles and -2.1 ± 0.3°h for adults, measured immediately after exposure; values declined to —1.6 ± 0.3°h for juve- niles and -3.8 ± 0.5°h for adults measured 32 days after exposure (Table 3). The percentage of crabs unable to right themselves immediately after expo- sure was significantly correlated with cumulative mortality (P Juvenile =0.003, r' venile = 0.91, n=6; P . „<0.001,r^7,=0.67,/i=22) and, therefore, could adult ' adult > serve as a predictor of death. Righting times tended to improve (decrease) dur- ing the first eight days after exposure, but this re- covery was generally not statistically significant. ^k *"^ N juveniles mortality o o \ \ \o \ Percent o \ ° \ ° \ 20- \ X ^ D adults \. □ x "s$2| 0- 1 ' 1 ' 1 ' 1 1 ' 1 -10 -8 -6 -4 -2 Degree- hours Figure 2 Cumulative percent mortality (P) of juvenile and adult female Chionoecetes bairdi, observed 32 days after emersion, as a function of exposure (°h): P /llMnifcj =100 / (1 + e< 3 74 + x 14x ° h) ), -P adu „=100 / (1 + e <6 42 + 1.48x°h)) 48 Fishery Bulletin 93(1), 1995 Righting times of juvenile crabs from all exposures tended to decrease over time (Fig. 4). The righting times of adult crabs exposed to <-2.2°h generally showed little evidence of recovery. Median exposures causing one-half the crabs to cease righting also gen- erally declined, but 95% confidence bars overlapped. Pereiopod autotomy Exposure to cold air caused pereiopod autotomy. Ju- venile pereiopod losses increased from to -2°h but declined towards the most severe exposure (-8°h), possibly because early mortality precluded autotomy (Fig. 5A). Juvenile crabs often dropped legs or chela during aerial exposure, but losses also continued af- ter exposure (Fig. 5B), often during ecdysis (Fig. 6). Adult crabs autotomized fewer pereiopods than ju- veniles; as with juveniles, loss was most frequent immediately after exposure. Loss of pereiopods in adults was directly related to severity of exposure (P<0.001, r 2 =0.85, n=19; Fig. 5A). Autotomy was cor- related with mortality in adult crabs (P<0.001, r 2 =0.81, n=19) but not in juveniles (P=0.621, r 2 =0.07, n=6). Autotomy was also correlated with the percent- age of adult crabs not righting as measured immedi- ately after exposure (P<0.001, r 2 =0.81, n=22). Ecdysis Juvenile crabs began molting 22 January and con- tinued through 21 February. Molt timing was not correlated with exposure (r 2 =0.09). Juveniles exposed 100- 80- — W-ii- \ juveniles \ B A 60 -I o\ o fa I 40- i i i \ 20- adults \ X I I V \ o- odcrofflso o o I 1 1 1 1 1 1 ' 1 ' 1 1 1 ' -10 8 Degree-hours Figure 3 Righting times of juvenile and adult female Chionoecetes bairdi capable of righting (A), and percentage not righting (B) observed 32 days after emersion as functions of exposure (°h). Percentages not righting were 100/(1 + e' 6235 + > 651 * ° h >) for adults and 100/(1 -h> (4701 * 2858 * ° h) ) for juveniles. Error bars are ±1 standard error. The ability of the single surviving -4.9°h adult crab improved over time; its righting time 32 days after exposure was similar to that of controls. Carls and O'Clair Responses of Chionoecetes bairdi to cold air 49 Table 2 Degree-hours caus ng death (LC) in Chionoecetes bairdi exposed to cold air, estimated with logit analysis. The LC number indicates the percentage of crabs affected e.g. LC50 is the median lethal degree-hours. The error term (CI) is the estimated 95% confidence interval. Day LC10 LC30 LC50 LC70 LC90 CI Juveniles 1 -0.9 -5.1 -7.7 -10.4 -14.6 3.4 2 -1.0 -4.6 -6.8 -9.0 -12.6 2.6 4 -1.3 -3.9 -5.5 -7.1 -9.7 1.7 8 -1.3 -2.9 -3.9 -4.8 -6.4 1.1 16 -1.3 -2.5 -3.3 -4.0 -5.2 0.8 32 -1.3 -2.5 -3.3 -4.0 -5.2 0.8 Adults 1 -3.1 -5.6 -7.2 -8.8 -11.3 1.6 2 -2.8 -5.2 -6.7 -8.2 -10.6 1.5 4 -2.6 -4.8 -6.2 -7.5 -9.7 1.3 8 -3.0 -4.0 -4.6 -5.1 -6.1 0.6 16 -3.2 -3.9 -4.3 -4.8 -5.5 0.5 32 -2.8 -3.8 -4.3 -4.9 -5.8 0.6 to cold air frequently lost pereiopods during ecdysis; losses increased from to — 4°h. The only crab ex- posed to -8°h that attempted to molt lost no limbs, but died during ecdysis (Fig. 6). Feeding rates Feeding rates of adult female Tanner crabs were sig- nificantly depressed by exposure to cold air (P ANOVA <0.001). In general, adult females exposed to <— 2.7°h (62% of the median lethal exposure) ate sig- nificantly less than did controls (Tukey test). Feed- ing rates measured shortly before zoeal hatching (41 to 60 days after exposure) were significantly less (P<0.05) for all crabs than feeding rates after zoeal hatching (85-98 days after exposure), but the slopes (feeding rate/exposure) before and after hatching did not differ (multivariate regression, P>0.50; Fig. 7). The frequency of feeding also increased significantly after zoeal hatching (P<0.001) and was significantly related to aerial exposure before and after larval hatching (P /mrar <0.001). The most severely treated crabs (-4.9°h) did not eat before zoeal release but ate 57% of the time after release. Weight change Change in weight of juvenile crabs was reduced by exposure to cold air. Wet weights of juvenile crabs that did not molt declined with increasing exposure severity (P=0.002, r 2 =0.42, n=20; Fig. 8). The weight Table 3 Effective degree-hours causing cessation of i -ighting (EC) in Chionoecetes bairdi exposed to cole air, estimated with logit analysis. The EC number indicates the percentage of crabs affected; EC50 is the median effective degree-hours. The error term (CI) is the estimated 95% confidence interval. Day EC10 EC30 EC50 EC70 EC90 CI Juveniles -0.2 -0.8 -1.2 -1.6 -2.2 0.3 1 -0.7 -1.2 -1.5 -1.7 -2.2 0.3 2 -0.6 -1.1 -1.5 -1.8 -2.3 0.3 4 -0.5 -1.1 -1.4 -1.7 -2.2 0.3 8 -1.1 -1.4 -1.7 -1.9 -2.3 0.3 16 -0.9 -1.3 -1.6 -1.9 -2.4 0.3 24 -0.8 -1.4 -1.7 -2.1 -2.6 0.3 32 -0.9 -1.3 -1.6 -1.9 -2.4 0.3 Adults -1.3 -1.8 -2.1 -2.5 -3.0 0.3 1 -2.0 -2.7 -3.1 -3.5 -4.1 0.4 2 -1.8 -2.4 -2.8 -3.2 -3.8 0.4 4 -2.3 -2.8 -3.1 -3.4 -3.9 0.4 8 -2.1 -2.7 -3.0 -3.4 -4.0 0.4 16 -2.2 -2.7 -3.1 -3.5 -4.0 0.4 24 -2.2 -2.9 -3.4 -3.8 -4.6 0.5 32 -2.4 -3.3 -3.8 -4.3 -5.1 0.5 increment of juvenile crabs that molted also declined with decreasing exposure (Fig. 8). This trend was not significant until an outlier at -2.0°h was removed (P=0.021, r 2 =0.56, n=9). Pereiopod autotomy prob- ably influenced these weight outcomes. The weight changes of juvenile crabs that did not molt were cor- related with righting response measured immedi- ately after exposure (P=0.018, r 2 =0.88, «=5, Y=a + bx 3 ). Changes in weight of adult crabs were not cor- related with exposure (P>0.07, r 2 =0.08, rc=44). Reproduction Exposure of ovigerous female crabs to cold air gen- erally did not affect the eggs or subsequently released zoeae unless the female died; all eggs died if the fe- male died. Timing of initial zoeal release (20 April ± 1 day), duration of release (11+1 day), and median release date (26 April ± 1 day) did not vary with ex- posure (r 2 =0.04, n=44; Fig. 9). Zoeae placed in sepa- rate containers for two days were not significantly affected by exposure prior to hatching (P=0.425, r 2 =0.02, «=43), and 87 ±3% continued swimming through the test period. Larval mortality, measured as the percentage of zoeae that sank to tank bottoms and died (0.4 ±0.2%), did not vary with exposure (^=0.03, n=44). Zoeal mortality (2 ±2%) during swim- ming tests was not correlated with exposure (r 2 =0.09). 50 Fishery Bulletin 93(1), 1995 The percentage of eggs that hatched may have been slightly affected by exposure, but our results were inconclusive (P„„ M „„„_ „„ =0.036, but P, . , regression:arcsin lack of ^•cO.OlO, and ^=0.10). The percentage of eggs hatch- ing in the -5.3°h treatment differed significantly from the control (Dunnett's test), but the difference was minor (99.1% versus 99.8% hatching). Egg extrusion may have been influenced by expo- sure, but the data were inconsistent. Elapsed time between larval hatching and subsequent egg extru- sion tended to be prolonged by exposure, but the re- sponse was variable (P Hnea =0.005, P lack O ^,=0.719, r 2 =0.19, rc=41). Egg extrusion generally occurred two days (median) after zoeal release but ranged from to 18 days; only crabs in the two most severe treat- ments (<—4.3°h) exceeded nine days. The date of ex- Juveniles Adults 60 50- 40- 30- 20 10 60 control Ji BDQ Q - T ^ I ' I ' I ' I - control I ' I I ' ' I ' 1 i ' i ' r 50 40 j§ 30- 20 10-1 60 -1.5 °h I ' I I ' I ' I I ' I ' I ' I ' I I ' I -2.0 °h 24 32 Days after exposure Figure 4 Righting times of juvenile and adult Chionoecetes bairdi as a function of time in days after emersion. Error bars are ±1 standard error. trusion (4 May ±7 days) may also have been changed by exposure, but again the statistical results were incon- elusive (P^ ar =0.011, P*^ ^=0.700, r^O.16, n=41). Exposure did not affect the percentage of Tanner crabs extruding eggs (93%, P linea =0.730, r^O.03, ra=6). Discussion Extreme exposure to cold air was lethal to Tanner crabs. It is also possible that thermal shock caused when the crabs were returned to water following expo- sure was also damaging. Following sublethal exposure, crabs exhibited a slowed righting response, autotomy of pereiopods, depressed feeding rates (adults), and weight loss or reduced weight gain (juveniles). Temperature and duration of treat- ment were both critical factors in de- termining how aerial exposure affected Tanner crabs. In a similar experiment with king crabs, the response of crabs to exposure was clearly observed when exposure was defined as the product of temperature and the length of exposure time (Carls and O'Clair, 1990). The use of this composite variable worked well with the current data set. However, our approach may not be generally appli- cable (for discussion see Carls and O'Clair, [1990]). Design of this experiment precluded independent analysis of temperature and time factors. However, either fac- tor may be predicted as a function of the other. For example, at -10°C, 10% of juvenile crabs may be killed by an 8- minute exposure, and 50% may be killed by a 20-minute exposure. Similarly, a 10-minute exposure would impair right- ing in 50% of juvenile crab at -7°C. Pre- dicted times and temperatures were calculated from degree-hours causing death (LC) or from effective degree-hours causing cessation of righting (EC) esti- mates (Tables 2 and 3); temperature multiplied by time (units are Celcius and hours) matched the LC or EC esti- mates. Predictions of adult and juvenile Tanner crab response are summarized in Figure 10 and Appendices. In our ex- ample (Fig. 10), short-term effects are predicted by ability to right immedi- ately after exposure; impaired crabs may be subject to increased predation at this time. Long-term effects are pre- -1.6 "h -2.2 "h Carls and O'Clair: Responses of Chionoecetes bairdi to cold air 51 &U- : A 40 -_ ,11 " Juveniles / 30 ~ / 20 ~ i 1 t 1 1 1 1 1 1 1 1 1 { ii 10 -j < i t i Adults \ 0- 1 ' 1 "" 1 i i B juveniles only -10 -8 -6 -4 Degree-hours -2 -2.0 °h -4.0 °h -1.5 °h -10 °h I ' I 24 Days after exposure 50 40 30 ■20 = -10 control l -1 — T 32 Figure 5 Percent of total pereiopod loss by juvenile and adult female Tanner crabs as a function of exposure I A) and as a function of time for juveniles (B). Error bars are ±1 standard error. dieted by mortality after exposure-induced death ceased. Mortality of adult Tanner crabs was significantly greater below -3°h and vigor was reduced below -2°h compared with control crabs. Exposures that are this severe probably occur infrequently on the fishing grounds except during winter in the north- ern Gulf of Alaska and the Bering Sea. Data are lacking on the time incidentally captured crabs remain on deck before being released, but dura- tion probably varies widely. Larger vessels employ- ing assembly-line techniques may process crabs more rapidly than do smaller vessels. Poor handling of culls combined with prolonged exposure may fur- ther reduce survival of incidentally caught crabs. Crabs captured incidentally during trawling are probably stressed more than those caught in pots. Stevens ( 1990) reported trawl tows ranging up to 6.4 hours and retention times of Tanner crabs up to 17 hours; the median lethal holding time for Tanner crabs was 8.3 hours. Net type influenced survival, and injuries were present in a greater pro- portion of dead than of live crabs (Stevens, 1990). 100- -e -4 Degree- hours Figure 6 Limb loss by juvenile Chionoecetes bairdi at ecdysis as a func- tion of exposure: percent loss=-8.856 + 4,756.960 e"° 629 * (0h * 10) . Error bars are ±1 standard error. Numbers molting in each group were 4, 4, 2, 6, 1, and 1 for controls through -8°h, re- spectively. 52 Fishery Bulletin 93(1). 1995 -2- T" -5 -4 ~ i ' r~ -3 -2 Degree hours Figure 7 Feeding rates (milligram food/gram crab weight/day) of adult female Chionoecetes bairdi before (F =4.556 + 1.071 pre x °h) and after (F pos ,=12.161 + 1.255 x °h) zoeal hatching as a functions of exposure (°h). Error bars indicate 95% CI. 100- 80- 60- S 40- 20 0- -20- did not molt -3 -2 Degree- hours Figure 8 Changes in wet weight of juvenile Chionoecetes bairdi as a function of exposure. Weights of crabs that molted versus those that did not were treated separately. Error bars are ±1 standard error. Mortality and injury due to aerial exposure have been reported for other commercially harvested de- capod crustaceans. For example, king crabs were af- fected by exposure to cold air, but were less sensitive than Tanner crabs (Carls and O'Clair, 1990). The western rock lobster, Panulirus cygnus, was signifi- cantly affected by >15 minutes exposure to warm air (27-35°C); recapture rates were lower than for un- exposed controls, and the probability of mortality due to predation rose (Brown and Caputi, 1983). We do not know what physiological mechanism(s) caused the abnormal events during ecdysis that of- ten resulted in death. O'Brien et al. (1986) induced apolysis (the separation of integumentary tissues from the exoskeleton during proecdysis) in several species of brachyurans by packing crabs in ice. Apolysis occurred within one hour in most cases and was not caused by death (O'Brien et al., 1986). O'Brien et al. (1986) did not observe ecdysis in their experimental crabs; therefore, the effect of apolysis on the timing, duration, and success of ecdysis in crabs is not known. In the present experiment, al- though mortality occurred during ecdysis in juvenile crabs, the timing of ecdysis was not affected. Evaporative water loss during exposure probably did not contribute significantly to the effects we ob- served. The fact that warmer exposures, such as the 32-minute exposure at +5°C, caused little or no ef- fect supports this supposition. Similar observations were made for king crab (Carls and O'Clair, 1990). In a study by Taylor and Whiteley (1989), the lob- ster Homarus gammarus vulgaris, which rarely comes in contact with air in its natural environment, was exposed to air at 15°C for up to 14 hours. Water loss, inferred from the constancy of most hemolymph ion concentrations, was minimal (Taylor and Whiteley, 1989). Oxygen delivered to H. gammarus tissues was substantially reduced, and C0 2 accumu- lated, but levels returned rapidly to normal after a 14-hour exposure. Lactate levels increased, but el- evation of bicarbonate ions increased the buffering capacity of the hemolymph. Because exposures did not exceed 32 minutes, it is unlikely that reduced oxygen directly caused Tanner crab mortality in our experiment. However, at low air temperatures, gills may have been damaged by frost, thus impairing respiratory gas, metabolite, and ion exchange after the crabs were returned to the water. The ability of crabs to right themselves proved to be a sensitive measure of crab viability. Righting response data collected immediately after exposure correlated strongly with less-immediate responses such as mor- tality and growth. Pereiopod loss also impaired righting. Pereiopod autotomy in adult crabs was a function of exposure. Mortality may have influenced the au- totomy response curve for juvenile crabs: during se- vere exposure, crabs apparently died before autotomy could take place. Aerial exposure reduced weight gain in juvenile crabs and caused weight loss in juveniles that did not molt. However, wet weights of the adult crabs (all anecdysial) did not vary with exposure. This ab- Carls and O'Clair: Responses of Chionoecetes bairdi to cold air 53 control 80- O 60- °to 40- \ O 20- cy* ° a* o - g^Nd 0- O 09B0 i i i i l i i 100- 80- 60- 40- 20- 100- 80- 60- 40- 20- o o o o -3.5 "h CK %°v o \ i i i I | i , , , 1 . i i I i i -I — I — I — r- -2.1 °h o - earoao BO T — I — I — I — I — I — I — I — I — I — I — I — I — I — I — I — T 100 ~r 110 r 120 T 130 100 1 I ' ' 110 T" 1 " 120 Julian day -4.3 "h -4.9 °h 130 Figure 9 Zoeal production (number of zoeae per gram female Chionoecetes bairdi) as a function of time. Curves were fit with smoothing techniques (4253HI (Velleman and Hoaglin, 1981). Units are in °h. sence of weight changes in the adults is puzzling because feeding rates were significantly depressed by exposure. Growth of adult western rock lobsters was reduced by exposure (Brown and Caputi, 1985). Body size, shape, and volume may be important factors in predicting crab response to cold-air expo- sure. Results of the present experiment support this hypothesis: smaller crabs (juveniles) were more sen- sitive to exposure than were larger crabs (adults). Additionally, adult Tanner crabs were more sensi- tive to exposure than were larger king crabs (Carls and O'Clair, 1990), but unknown interspecific fac- tors may have influenced this difference. An experi- ment involving a broad size range of conspecific in- dividuals is needed to test whether sensitivity to ex- posure is size-dependent in crabs. Surprisingly, aerial exposure did not measurably af- fect the developing larvae of exposed females unless the female died. Surviving crabs produced normal zoeae. Moreover, the timing of larval release, larval swimming ability, and viability were not affected by exposure. Longer-term larval responses, such as sur- vival past the first molt and zoeal growth, were not examined. Exposure may have reduced hatching suc- cess (by <1%) of the Tanner crab larvae and possibly may have affected the timing of egg extrusion, but these responses did not vary strongly. Schlieder (1980) re- ported a 13% reduction in hatching success in the stone crab, Menippe mercenaria, compared with controls when the crabs were exposed to air at 27-33°C for two hours. Hatching success was reduced further by a five- hour exposure and by autospasy (Schlieder, 1980). 54 Fishery Bulletin 93(1), 1995 In summary, although environmental conditions as severe as those tested are uncommon on the fish- ing grounds during fishing operations (except in the central and northern Bering Sea), low-temperature aerial exposure during fishing operations can ad- versely affect incidentally captured crabs. Exposure to cold air reduced crab vigor and feeding rates, caused limb autotomy, and killed the crabs in severe situations. Progeny died if exposure to cold air killed females brooding them, otherwise larvae were not measurably affected. Prompt return of incidentally caught Tanner crabs to the sea, especially during ex- tremely cold weather, should reduce adverse effects of exposure to cold air. Adults, Death ////I 50- II 40- / 30 ~. JJIJ 20 ~ yyy/J 90^^^/ 10- =*5?io-^'^ 'I""" 1 "! -30 -20 -10 -30 -20 -10 Temperature (Celcius) Temperature (Celcius) Figure 1 Predicted time in minutes required to cause death or impair righting of juvenile and ovigerous female Chionoeeetes bairdi following expo- sure to cold air. Mortality predictions are based on cumulative mor- tality through day 16; no deaths were observed in the ensuing 16 days. Righting predictions are based on responses immediately after exposure; there was a tendancy for righting times to improve after exposure, but improvements were generally not statistically significant. Acknowledgments We thank Tyrus Brouillette for his technical assistance during this experiment and the reviewers who im- proved this manuscript with their helpful suggestions. Literature cited Berkson, J. 1957. Tables for the maximum likelihood estimate of the logistic function. Biometrics 13:28-34. BMDP. 1983. Statistical software [1983 printing with additions], W. E. Dixon (ed.). Univ. Calif. Press, Berkeley, 735 p. Brown, R. S., and N. Caputi. 1983. Factors affecting the recapture of undersize western rock lobster Panulirus cygnus George returned by fish- ermen to the sea. Fish. Res. (Amst.) 2:103-128. 1985. Factors affecting the growth of undersize western rock lobster, Panulirus cygnus George, returned by fishermen to the sea. Fish. Bull. 83:567-574. Carls, M. G., and C. E. O'Clair. 1990. Influence of cold air exposures on ovigerous red king crabs (Paralithodes camtschatica) and Tanner crabs (Chionoeeetes bairdi) and their offspring. In Proc. int. symp. king and Tanner crabs, Nov. 1989, Anchorage, Alaska, p. 329- 343. Alaska Sea Grant College Program, Univ. Alaska, Fairbanks, AK 99775-5040. Fujioka, J. T. 1986 Log-likelihood ratio tests for comparing dose-response data fitted to the logistic function. U.S. Dep. Commer., NOAA Tech. Memo. NMFS F/NWC-96, 25 p. Kleinbaum, D. ('•., and L. L. Kupper. 1983. Applied regression analysis and other multivariable methods. Duxbury Press, Bos- ton, MA, 556 p. O'Brien, J. J., D. L. Mykles, and D. M. Skinner. 1986. Cold-induced apolysis in anecdysial brach- yurans. Biol. Bull. 171:450-460. Otto, Robert S. 1989. An overview of eastern Bering Sea king and Tanner crab fisheries. In Proc. int. symp. king and Tanner crabs, Nov. 1989, Anchorage, Alaska, p. 9-26. Alaska Sea Grant College Program, Univ. Alaska, Fairbanks AK 99775- 5040. Schlieder, R. A. 1980. Effects of desiccation and autospasy on egg hatching success in stone crab, Menippe mercenaria. Fish. Bull. 77:695-700. Stevens, B. G. 1990. Survival of king and Tanner crabs cap- tured by commercial sole trawls. Fish. Bull. 88:731-744. Taylor, E. W., and N. M. Whiteley. 1989. Oxygen transport and acid-base balance in the haemolymph of the lobster, Homarus gammarus, during aerial exposure and resubmersion. J. Exper. Biol. 144:417-436. Wilt-man, P. F., and D. C. Hoaglin. 1981. Applications, basics, and computing of exploratory data analysis. Duxbury Press, Boston, MA, 354 p. Carls and O'Clair: Responses of Chionoecetes bairdi to cold air 55 Appendix Table 1 Predicted time in minutes required to cause death of the listed percentage of adult Chionoecetes bairdi at indicated temperatures (°C). Calculations are based lethal responses (LC10, LC30, . . . LC90) estimated on day 16. Temperature -1.0 10% 189 30% 50% 233 260 70% 90% 288 332 Temperature 10% 30% 50% 70% 90% -16.0 12 15 16 18 21 -2.0 95 116 130 144 166 -17.0 11.1 13.7 15.3 16.9 19.5 -3.0 63 78 87 96 111 -18.0 10.5 12.9 14.5 16.0 18.4 -4.0 47 58 65 72 83 -19.0 9.9 12.3 13.7 15.2 17.5 -5.0 38 47 52 58 66 -20.0 9.5 11.6 13.0 14.4 16.6 -6.0 32 39 43 48 55 -21.0 9.0 11.1 12.4 13.7 15.8 -7.0 27 33 37 41 47 -22.0 8.6 10.6 11.8 13.1 15.1 -8.0 24 29 33 36 41 -23.0 8.2 10.1 11.3 12.5 14.4 -9.0 21 26 29 32 37 -24.0 7.9 9.7 10.9 12.0 13.8 -10.0 19 23 26 29 33 -25.0 7.6 9.3 10.4 11.5 13.3 -11.0 17 21 24 26 30 -26.0 7.3 9.0 10.0 11.1 12.8 -12.0 16 19 22 24 28 -27.0 7.0 8.6 9.6 10.7 12.3 -13.0 15 18 20 22 26 -28.0 6.8 8.3 9.3 10.3 11.9 -14.0 14 17 19 21 24 -29.0 6.5 8.0 9.0 9.9 11.4 -15.0 13 16 17 19 22 -30.0 6.3 7.8 8.7 9.6 11.1 Appendix Table 2 Predicted time in minutes required to cause death of the listed percentage of juvenile Chionoecetes baird i at indicated tempera- tures (°C). Calculations are based lethal responses (LC10, LC30 , . . . LC90) estimated on day 16. Temperature 10% 30% 50% 70% 90% Temperature 10% 30% 50% 70% 90% -1.0 81 152 196 241 311 -16.0 5.1 9.5 12.3 15.0 19.5 -2.0 41 76 98 120 156 -17.0 4.8 8.9 11.5 14.2 18.3 -3.0 27 51 65 80 104 -18.0 4.5 8.4 10.9 13.4 17.3 -4.0 20 38 49 60 78 -19.0 4.3 8.0 10.3 12.7 16.4 -5.0 16 30 39 48 62 -20.0 4.1 7.6 9.8 12.0 15.6 -6.0 14 25 33 40 52 -21.0 3.9 7.2 9.3 11.5 14.8 -7.0 12 22 28 34 44 -22.0 3.7 6.9 8.9 10.9 14.2 -8.0 10 19 25 30 39 -23.0 3.5 6.6 8.5 10.5 13.5 -9.0 9 17 22 27 35 -24.0 3.4 6.3 8.2 10.0 13.0 -10.0 8 15 20 24 31 -25.0 3.2 6.1 7.8 9.6 12.5 -11.0 7.4 13.8 17.8 21.9 28.3 -26.0 3.1 5.8 7.5 9.3 12.0 -12.0 6.8 12.7 16.4 20.1 26.0 -27.0 3.0 5.6 7.3 8.9 11.5 -13.0 6.2 11.7 15.1 18.5 24.0 -28.0 2.9 5.4 7.0 8.6 11.1 -14.0 5.8 10.8 14.0 17.2 22.2 -29.0 2.8 5.2 6.8 8.3 10.7 -15.0 5.4 10.1 13.1 16.0 20.8 -30.0 2.7 5.1 6.5 8.0 10.4 56 Fishery Bulletin 93(1). 1995 Appendix Table 3 Predicted time in minutes required to impair righting response of the listed percentage of adult Chionoecetes bairdi at indicated temperatures (°C). Calculations are based on righting responses (EC 10, EC30, . . . EC90) estimated immediately after exposure. Temperature 10% 30% 50% 70% 90% Temperature 10% 30% 50% 70% 90% -1.0 79 109 129 148 179 -16.0 4.9 6.8 8.0 9.3 11.2 -2.0 39 55 64 74 89 -17.0 4.6 6.4 7.6 8.7 10.5 -3.0 26 36 43 49 60 -18.0 4.4 6.1 7.2 8.2 9.9 -1.0 20 27 32 37 45 -19.0 4.1 5.8 6.8 7.8 9.4 -5.0 16 22 26 30 36 -20.0 3.9 5.5 6.4 7.4 8.9 -6.0 13 18 21 25 30 -21.0 3.7 5.2 6.1 7.0 8.5 -7.0 11 16 18 21 26 -22.0 3.6 5.0 5.9 6.7 8.1 -8.0 10 14 16 19 22 -23.0 3.4 4.8 5.6 6.4 7.8 -9.0 9 12 14 16 20 -24.0 3.3 4.6 5.4 6.2 7.5 -10.0 8 11 13 15 18 -25.0 3.1 4.4 5.1 5.9 7.2 -11.0 7.1 9.9 11.7 13.5 16.3 -26.0 3.0 4.2 5.0 5.7 6.9 -12.0 6.5 9.1 10.7 12.3 14.9 -27.0 2.9 4.0 4.8 5.5 6.6 -13.0 6.0 8.4 9.9 11.4 13.8 -28.0 2.8 3.9 4.6 5.3 6.4 -14.0 5.6 7.8 9.2 10.6 12.8 -29.0 2.7 3.8 4.4 5.1 6.2 -15.0 5.2 7.3 8.6 9.9 11.9 -30.0 2.6 3.6 4.3 4.9 6.0 Appendix Table 4 Predicted time in minutes required to impair righting response of the listed percentage of juvenile Chionoecetes bairdi at indi- cated temperatures (°C). Calculations are based on righting responses (EC 10, EC30, . . EC90) estimated immediately after exposure. Temperature 10% 30% 50% 70% 90% Temperature 10% 30% 50% 70% 90% -1.0 13 49 71 93 129 -16.0 0.83 3.05 4.45 5.84 8.07 -2.0 7 24 36 47 65 -17.0 0.78 2.87 4.19 5.50 7.59 -3.0 4 16 24 31 43 -18.0 0.74 2.71 3.95 5.19 7.17 -4.0 3.3 12.2 17.8 23.4 32.3 -19.0 0.70 2.57 3.75 4.92 6.79 -5.0 2.7 9.8 14.2 18.7 25.8 -20.0 0.66 2.44 3.56 4.67 6.45 -6.0 2.2 8.1 11.9 15.6 21.5 -21.0 0.63 2.33 3.39 4.45 6.15 -7.0 1.9 7.0 10.2 13.4 18.4 -22.0 0.60 2.22 3.23 4.25 5.87 -8.0 1.7 6.1 8.9 11.7 16.1 -23.0 0.58 2.12 3.09 4.06 5.61 -9.0 1.5 5.4 7.9 10.4 14.3 -24.0 0.55 2.04 2.97 3.90 5.38 -10.0 1.3 4.9 7.1 9.3 12.9 -25.0 0.53 1.95 2.85 3.74 5.16 -11.0 1.2 4.4 6.5 8.5 11.7 -26.0 0.51 1.88 2.74 3.60 4.96 -12.0 1.1 4.1 5.9 7.8 10.8 -27.0 0.49 1.81 2.64 3.46 4.78 -13.0 1.0 3.8 5.5 7.2 9.9 -28.0 0.47 1.74 2.54 3.34 4.61 -14.0 0.9 3.5 5.1 6.7 9.2 -29.0 0.46 1.68 2.45 3.22 4.45 -15.0 0.88 3.26 4.74 6.23 8.60 -30.0 0.44 1.63 2.37 3.12 4.30 Abstract. The Atlantic sharpnose shark, Rhizoprionodon terraenovae, is a small coastal spe- cies caught in recreational fisher- ies and as bycatch in the shrimp trawl and longline fisheries in the Gulf of Mexico. Demographic analyses incorporating the best available information on validated age and growth, age at maturity (t mat ), maximum age (t max ), repro- ductive habits, and age-specific natural mortality and fecundity were performed. An initial set of three life history tables based on input parameters t ma =4, t max =10, constant age 1+ survivorship (S=0.657), and varying first year survivorship (S o =0.432, scenario 1; S o =0.512, scenario 2; S o =0.657, sce- nario 3 or best case scenario) yielded net reproductive rates (i? ) ranging from 0.844 to 1.284, a gen- eration length (G) of 5.8 years, and instantaneous rates of population change (r) ranging from -0.029 to 0.044. Further simulations were performed to test the sensitivity of the computed demographic param- eter values to modifications in vari- ous input biological parameter val- ues (scenarios 4 through 14). Over- all, manipulations of biological pa- rameters m x , t maV and t max caused large variations in demographic parameters r, < x2 , and R , while G remained relatively stable. All the demographic parameters proved more sensitive to changes in S than to changes in S . The initial set of analyses (scenarios 1 through 3) was then rerun with the estimated mean fishing mortality from 1986 to 1989 (F=0.428) added to natu- ral mortality. Age 6+ sharks can enter the fishery under the best case scenario only to allow the population to replace itself. Ages at first capture (A rep ) with F=0.428 that would allow full population replacement were also calculated for scenarios 4 through 14. This study indicates that management of R. terraenovae under the Federal Management Plan (FMP) for sharks of the Atlantic Ocean is based on unrealistic biological characteristics for this species. Demographic analysis of the Atlantic sharpnose shark, Rhizoprionodon terraenovae, in the Gulf of Mexico Enric Cortes Center for Shark Research Mote Marine Laboratory 1600 Thompson Parkway, Sarasota. Florida 34236 Manuscript accepted 31 May 1994. Fishery Bulletin 93:57-66 (1995). The Atlantic sharpnose shark, Rhi- zoprionodon terraenovae, is an abundant coastal carcharhinid spe- cies found in shelf waters of the western North Atlantic and Gulf of Mexico (Compagno, 1984), reported to reach a maximum size of approxi- mately 110 cm total length (Com- pagno, 1984 1. Although not targeted by any U.S. commercial fisheries, in the Gulf of Mexico it is caught in recreational fisheries and discarded as bycatch in the shrimp trawl fish- ery (NMFS, 1993) and shark and reef fish longline fisheries (person, obs. ). However, age at first entry in the various fisheries is unknown. R. terraenovae is grouped under the "small coastal" species category in the Federal Management Plan (FMP) for sharks of the Atlantic Ocean, which determined that this species group was not overfished, based on a stock assessment resulting in an estimate of finite rate of population increase (e r ) of 1.91. Thus, no quotas or size limits exist for this species despite its importance in several fisheries. Biological and life history charac- teristics of R. terraenovae in the Gulf of Mexico are now well docu- mented (Parsons, 1983, 1985; Branstetter, 1986, 1987). However, this information has not yet been applied to analyses of the popula- tion dynamics of this species, nor have the results of such analyses been published. Furthermore, as is the case with most shark species, sufficient information necessary for stock assessment is lacking (Hoff, 1990). Because long-term records of catch and effort or the age composi- tion by species are not available, traditional surplus production mod- els or more elaborate age-structured methods of stock assessment have seldom been used for sharks. Ow- ing to the paucity of fisheries data, several investigators have used de- mographic analysis to gain insight into the population dynamics and exploitation rates of shark re- sources. This type of analysis has been utilized to construct life his- tory tables or Leslie matrices (Caughley, 1977; Krebs, 1985), which are summaries of age-specific mortality and fertility rates operat- ing on a population with the as- sumption of a stable age distribu- tion. This technique allows estima- tion of parameters important to the dynamics of any given population. Thus, Hoenig and Gruber (1990), Cailliet (1992), and Cailliet et al. ( 1992) produced estimates of popula- tion dynamics by applying a demo- graphic analysis of the lemon shark, Negaprion brevirostris, the leopard shark, Triakis semifasciata, and the angel shark, Squat ina californica, respectively. Hoff ( 1990), in addition, estimated maximum sustainable yield for the sandbar shark, Car- charhinus plumbeus, in a modified stock production model incorporat- ing life history information. 57 58 Fishery Bulletin 93(1), 1995 This study was prompted by the existence of vali- dated age and growth data (Branstetter and McEachran, 1986) and reproductive information (Parsons, 1983) on R. terraenovae, which provided several important parameters needed for construc- tion of a life history table, i.e. lifespan, fecundity, and age at maturity. The purpose of this study was 1) to produce, using the life history table approach and the best biologi- cal information available, reliable estimates of de- mographic parameters fori?, terraenovae in the Gulf of Mexico, 2) to assess the sensitivity of the computed demographic parameters of the population to a vari- ety of biological (input) parameter manipulations and harvest scenarios, and 3) to compare the resultant rates of population increase with that calculated for the "small coastal" shark group in the FMP for sharks of the Atlantic Ocean and to evaluate the biological basis of the stock assessment on which present man- agement measures are based. The instantaneous natural mortality rate (M) was calculated from Hoenig's (1983) equation relating maximum age to total mortality rate, derived from data pertaining to unexploited or lightly exploited stocks. A value of 0.42 for Z (instantaneous total mortality rate) was derived from the regression equa- tion ln(Z) = 1.46 - 1.01 in(t max ), where t max is longev- ity in years. Assuming that maximum age was de- termined from a time when there was no fishing di- rected at this species, Z can be approximated to M. The proportion of survivors at the start of each age interval (x) was l x = N (e~ Mx ), where 7V o is the num- ber of individuals at time 0. Demographic parameters were computed follow- ing methodology by Krebs ( 1985 ) and included R o (net reproductive rate per generation), G (generation length in years), and r (intrinsic rate of population change). All the values of r reported in this study were refined by using the Euler equation (Wilson and Bossert, 1971; Krebs, 1985): l r m r = 1. Materials and methods Life history tables incorporating the best biological information available on R. terraenovae in the Gulf of Mexico were constructed. Maximum age (lifespan or longevity; t max ) has been estimated to be 8 to 10 years (Branstetter, 1987), and age at maturity (t .) for females has been estimated at 4 years (Bran- stetter, 1987), and from 2.4 to 3.9 years (Parsons, 1985). For this study, it was assumed that all females reproduced after reaching maturity. Parsons (1983) reported that parturition was annual, gestation pe- riod was 10 to 11 months, and sex ratios at birth were 1:1. He also found a significant relationship between total length of gravid females and number of offspring produced. Fecundity at size was calcu- lated from the regression equation Y = -8.4109 + 0.1396X (r=0.50,P<0.001,n=78; Parsons 1 ), where X is female total length and Y is number of offspring. Female length at age was obtained from the von Bertalanffy growth function for both sexes combined derived by Branstetter ( 1987): L t =Lj 1-e -*«-*>>), where #=0.359, L x =108, and f o =-0.985. Number of offspring was further divided by two, because the natality func- tion (m x ) at age represents the number of female off- spring per female parent and sex ratios at birth are 1:1 and parturition is annual (Parsons, 1983). Reports of unusually large litter sizes in tropical populations of R. terraenovae were also incorporated in some of the analyses that follow by doubling fecundity at age (m ). *=o The finite or annual rate of change (e r ) was then calculated from the refined values of r. In addition, the theoretical population doubling or halving time in years « x2 ) assuming a stable age distribution was com- puted as (In 2)/r or (In 0.5)/r, respectively (Krebs, 1985). The initial set of analyses, consisting of three differ- ent scenarios, was run by using the most reliable input biological parameters: t =10, t mat =A, m r =baseline age- specific natality, and S=0.657. In scenario 1, first year natural mortality was arbitrarily doubled (M=0.42 x 2=0.84) or S o =0.432. In scenario 2, a value of S o =0.512 was obtained from the Leslie matrix algorithm by as- suming an equilibrium (or stationary) population (Vaughan and Saila, 1976). Thus, the following equa- tion was solved for S o after assuming r=0: i-i i=i ("•*!«■*) ll S J 7 = 1 1 Parsons, G. Univ. Mississippi, MI 38677. Personal commun.. 1993. where m is fecundity at age, I is the oldest age group in the population ( 10 years), and S ; is survival from age j to y+l. In scenario 3, (referred to as the best case scenario), S was assumed to be equal to survivor- ship in the following years (S o =S=0.657=e-° 42 ). For this best case scenario, the stable age distribution (C^) was calculated according to Krebs ( 1985) and plotted. In a second set of analyses, the input biological or life history parameters (t .,t ,m,S,S) were var- J r mat 1 max 1 x 1 ' o ied to test the sensitivity of the resultant demographic parameters (R , G, r, and t K2 ). These sensitivity analy- ses measured the percentage change of the output de- Cortes: Demographic analysis of Rhizoprionodon terraenovae 59 mographic parameter of interest relative to the best case scenario. In the case of ^ x2 , sensitivity was assessed by calculating a multiplication factor that measures the number of times the population doubling time changes relative to the best case scenario (example: if ^=15.7 in the best case scenario and 4.1 in the altered state, then the multiplication factor [mf]=15. 7/4. 1=3.8, i.e. t^ has been shortened 3.8 times in the altered state). Based on the results from the initial set of analy- ses (scenarios 1 through 3), the existing knowledge of life history traits in R. terraenovae, and the re- sults from the FMP (e r =1.91), input parameter val- ues were manipulated in the direction that would be favorable to population increase and should thus be regarded as optimistic scenarios. The following varia- tions, relative to the best case scenario, were applied: doubling m x (m x =2; scenario 4); reducing t nmt by 1 year (t mat =3; scenario 5) and by 2 years (t mat =2; sce- nario 6); reducing t mat by 1 year and doubling m x (t mat =3, m x =2; scenario 7); reducing t mat by 2 years and doubling m x (t mat -2, m x =2; scenario 8); increas- ing S by 10% (S o =0.723; scenario 9) and by up to 50% (°S o =0.985; scenario 10); increasing S by 10% (S=0.72°3; scenario 11) and by up to 50% (S=0.985; scenario 12); doubling t fuax (t max =20 [note that S and S will also vary, since they depend on t max \; scenario 13); and an extreme manipulation that was under- taken to approximate the FMP value of e r =1.91 (equivalent to an r of 0.647), where t mat was reduced by 1 year, m x was doubled, and S and S o set at 95% (t ,=3, m =2, S=S =0.95; scenario 14). mat ' x ' o ' A third set of simulations was run incorporating the estimated mean instantaneous fishing mortal- ity rate from 1986 to 1989 (F=0A28), as used in the stock assessment of small coastal species (Parrack 2 ) on which the FMP for sharks of the Atlantic Ocean is based, to demonstrate the effect of exploitation and various age-at-first-entry scenarios. Fishing mortal- ity (F) was added to natural mortality (M) in the survivorship function l x = N Q (e~ [M+F]x ), with F initially starting at age 0, then sequentially up to age 9. A rep , the minimum age at which individuals can first enter the fishery and still allow the population to replace it- self (r>0) was calculated by noting the age at which the intrinsic rate of increase (r) becomes zero or positive. These simulations were run first under scenarios 1 through 3, and then under scenarios 4 through 14. Results The initial set of life history tables yielded net re- productive rates per generation (R ), ranging from 0.844 to 1.284, a generation length (G) of 5.8 years, and intrinsic rates of population change (r), ranging from -0.029 to 0.044 (Table 1 ) depending on the value of first year survivorship (S ) used. In scenario 1 (S =0.432), the results indicated that the population would decrease at a rate of 2.9% per year and would halve about every 24 years. Halving times are indi- cated by negative values in the t x2 column. In sce- nario 2 (S =0.512), r is equal to by definition. Un- der the best case scenario (scenario 3; S =0.657), the 2 Parrack, M. L. 1990. A preliminary study of shark exploi- tation during 1986-1989 in the U.S. FCZ. Contrib. MIA- 90-493, NOAA, NMFS, SEFC, Miami, FL 33149, 23 p. Table 1 Simulations of the Gulf of Mexico population of the Atlantic sharpnose shark, Rhizoprionodon terraenovae, under three scenarios that use input parameter values representing the best biological information available. Only natural mortality is included in these analyses. First year survival rates (S ) were obtained as follows: S o =0.432 (scenario 1 ) was obtained by doubling the natural mortality value computed from Hoenig's ( 1983) relationship between mortality rate and maximum age; S o =0.512 (scenario 2) was computed from the Leslie matrix algorithm (see text) assuming an equilibrium population (Vaughan and Saila, 1976). The third line (in italics) represents the best case scenario (scenario 3; S o =S=0.657). Input parameter values' Computed parameter values 2 Scenario 'mat t max m x s s a Ro G r e r **2 1 4 10 \ 3 0.657 0.432 0.844 5.762 -0.029 0.971 -23.9 2 4 10 1 0.657 0.512 1.000 5.762 0. 1.000 — 3 4 10 1 0.657 0.657 1.284 5.762 0.044 1.045 15.7 1 'ma; =a e e at maturity; f mal =maximum age; m .^age-specific natality; S=survivorship after the first year of life; S o =survivorship for the first year of life. 2 fl =net reproductive rate per generation; G=generation length, in years; r=intrinsic rate of population change refined through the Euler equation (see text); e r =finite rate of population change; ^^theoretical doubling (positive values) or halving ( negative values) time in years assuming a stable age distribution. 3 "1" indicates baseline age-specific natality. 60 Fishery Bulletin 93(1), 1995 population increased at 4.5% per year and doubled about every 16 years. The predicted stable age distribution (Cj for the best case scenario (Fig. 1) suggested that about 80% of the population was composed of immature indi- viduals. Because of the lack of data on sizes and ages at first capture in the recreational and commercial 2 3 4 5 6 7 8 Age Class ( years ) Figure 1 Predicted stable age distribution of the Atlantic sharpnose shark, Rhizoprionodon terraenovae , under the best case scenario presented in Table 1, assuming geometric growth (with r=0.044) and constant age-specific mortality and fer- tility rates. fisheries, the actual proportion of the population sub- ject to fishing is unknown. Likewise, no size or age com- position of this population is available from surveys, precluding any comparisons with the theoretical C . Results of the sensitivity analyses indicated that doubling age-specific natality, m x , had a distinct ef- fect (a 286% increase) on the population's rate of in- crease, r (scenario 4), and would allow the popula- tion to double in only 4.1 years or 3.8 times faster than in the best case scenario (Table 2). Generation length, G, remained the same, while net reproductive rate per generation, R o , increased 100% (Table 3). Decreasing age at maturity, t mat , by one year (sce- nario 5) produced a smaller change in r, t x2 , and R o (Tables 2 and 3) than doubling m x , but decreased G by 12% (Table 3). Further decreasing t mat by another year (scenario 6) produced almost the same values of r and t x2 as those obtained in scenario 4 (Table 2), although R o increased only 5% and G decreased by 20%. The combined effect of decreasing t mat and dou- bling m x together (scenarios 7 and 8) produced in- creases in r up to near 700% and t x2 values up to 8 times shorter than in the best case scenario (Table 2). Under scenarios 7 and 8, R o also increased by up to more than 200%, while G decreased by up to 20%. Increasing first year survivorship, S o , by 10% (sce- nario 9) yielded a value of r 39% higher and a value of t x2 1.4 times shorter than in the best case scenario (Table 2), affected R o very little (a 10% increase only), and had no effect on G (Table 3). A further increase Table 2 Simulations of the Gulf of Mexico population of Rhizoprionodon terraenovae to test the sensitivity of computed population rate of increase and doubling time to input biological parameter values. Input val jes were manipulated in scenarios 4 through 14; the best case scenario (BC; top row) is shown in italics to facilitate comparison All other symbols are as defined in Table 1. Scenario Input parameter values Computed parameter values t mal t max m x S So r % change of r 1 «x2 mP BC 4 10 1 0.657 0.657 0.044 15.7 4 4 10 2 3 0.657 0.657 0.170 286 4.1 3.8 5 3 10 1 0.657 0.657 0.111 152 6.2 2.5 6 2 10 1 0.657 0.657 0.168 282 4.1 3.8 7 3 10 2 0.657 0.657 0.265 502 2.6 6.0 8 2 10 2 0.657 0.657 0.356 709 1.9 8.3 9 4 10 1 0.657 0.723 0.061 39 11.4 1.4 10 4 10 1 0.657 0.985 0.117 166 5.9 2.7 11 4 10 1 0.723 0.657 0.123 179 5.6 2.8 12 4 10 1 0.985 0.657 0.378 759 1.8 8.7 13 4 20 1 0.811 0.658 0.228 418 3.0 5.2 14 3 10 2 0.950 0.950 0.634 1,341 1.1 14.3 ; % change of r relative to the best case scenario 2 Multiplication factor indicating the number of times t^ has been shortened relative to the best case scenario 3 "2" indicates baseline age -specific natality values have been doubled. Cortes: Demographic analysis of Rhizopnonodon terraenovae 61 Table 3 Simulations of the Gulf of Mexico population of Rh zoprionodon terraenovae to test the sensitivity of computed net reproductive rate per generation and generat on len gth to input biological parameter values Input values were manipulated in scenarios 4 through 14 the best case scenario (BC top row) is shown in italics to facilitate comparison. All other symbols are as defined in Table 1. Scenario Input parameter values Computed parameter values *mul t max m , S So «o % change of R o ' G % change of G' BC 4 10 1 0.657 0.657 1.28 5.76 — 4 4 10 2 2 0.657 0.657 2.57 100 5.76 5 3 10 1 0.657 0.657 1.72 34 5.06 -12 6 2 10 1 0.657 0.657 1.34 5 4.58 -20 7 3 10 2 0.657 0.657 3.43 168 5.06 -12 8 2 10 2 0.657 0.657 4.08 219 4.58 -20 9 4 10 1 0.657 0.723 1.41 10 5.76 10 4 10 1 0.657 0.985 1.92 50 5.76 11 4 10 1 0.723 0.657 2.05 60 6.06 5 12 4 10 1 0.985 0.657 11.66 809 7.20 25 13 4 20 1 0.811 0.658 4.99 290 8.30 44 14 3 10 2 0.950 0.950 29.6 2,212 6.7 16.3 ' % change of R and G relative to the best case scenario. 2 "2" indicates baseline age-spec fie natality values have been doubled. in S up to 50% (scenario 10) had a more distinct effect on r (166% increase), t x2 (2.7 times shorter), and R (50% increase), but did not affect G (Tables 2 and 3). Increasing age 1+ survivorship (S) by 10% (scenario 11) had a similar effect on all the demo- graphic parameters to increasing S o by 50% (scenario 10; Tables 2 and 3), whereas increasing S by 50% (scenario 12) had a very profound effect on all the demographic parameters, increasing r by 759%, shortening t x2 by almost 9 times (similar to scenario 8), increasing R o by over 800% and lengthening G by 25% (Tables 2 and 3). Doubling longevity (t max ) to 20 years (scenario 13) also markedly affected r (418% increase), t x2 (5 times shorter), and R o (290% increase), and produced the largest value of G (8.3 or a 44% increase) in all sce- narios (Tables 2 and 3). Finally, the extreme manipulations of scenario 14 (reducing t mat to 3 years, doubling m x , increasing S and S o to 95%, with a t max of 10 years) produced a 13- fold increase in r, a value of t x2 more than 14 times shorter, a 22-fold increase in R o and only a 16.3% increase in G (Tables 2 and 3). For all simulations, population doubling time (t x2 ) was lessened and generation length (G) was the de- mographic parameter less sensitive to changes in input biological parameter values. With the estimated mean fishing mortality from 1986 to 1989 (F=0.428) added to natural mortality starting at each age interval from 9 to 0, R Q and r were progressively reduced as F was progressively started closer to age-0 (Table 4). In scenarios 1 (S o =0.432) and 2 (S o =0.512), r was always negative and became increasingly so as simulated fishing started earlier in the life of R. terraenovae. Only by using best case scenario (scenario 3) values could the population be made to replace itself or grow by ma- nipulating age at first capture. When fishing pres- sure was applied between 6 and 5 years of age or about 97 cm total length (TL) the population was able to replace itself. Generation length remained the same under the three scenarios but progressively decreased as fishing mortality included progressively earlier ages. Theoretical halving time also progres- sively shortened as fishing started at younger ages, whereas in the best case scenario doubling time in- creased as age at first capture dropped from 9 to 6 years. The effect of added fishing mortality on survivor- ship can be identified as a progressive decrease in percentage survival as fishing starts progressively earlier in the lifespan of the shark (Fig. 2). Age-spe- cific reproduction also decreases significantly as fish- ing mortality is applied at progressively earlier ages (Fig. 3). When the estimated mean fishing mortality from 1986 to 1989 (F=0.428) was added to natural mor- tality in scenarios 4 through 14 (Table 5), A , the earliest age at which sharks can first be captured to 62 Fishery Bulletin 93(1), 1995 3 4 5 6 7 Age Class ( years ) Figure 2 Survivorship curves for Rhizoprionodon terraenovae un- der the survival conditions presented in Table 1 for the best case scenario (S=S =0.657) and fishing mortality as in Table 4 (F=0.428), starting at three different ages (1, 5, and 8 years). 3 4 5 6 7 Age Class ( years ) Figure 3 Age-specific reproduction for Rhizoprionodon terraenovae under the survival conditions presented in Table 1 for the best case scenario (S=S o =0.657) and fishing mortality as in Table 4 (F=0A28), starting at three different ages (1, 5, and 8 years). Table 4 Simulations of the Gulf of Mexico population of Rhizoprionodon terraenovae under the three same scenarios as in Table 1 but with estimatec mean fishing mortality from 1986 to 1989 (F=0.428 [Parrack 2 ]) added to natural mortality starting at different ages. All symbols are as defined in Table 1. Computations based on the following first year survival rates: scenario 1 (S =0.432); scenario 2 (S = 0.512); and scenario 3 (best case S o =0.657) Age at first capture Demographic parameter 9 8 7 6 5 4 3 2 1 Scenario 1 R 0.83 0.81 0.77 0.71 0.62 0.49 0.32 0.21 0.14 0.09 G 5.70 5.60 5.45 5.27 5.06 4.86 4.86 4.86 4.86 4.86 r -0.03 -0.04 -0.05 -0.07 -0.09 -0.14 -0.23 -0.31 -0.38 -0.46 e r 0.97 0.96 0.95 0.94 0.91 0.87 0.80 0.74 0.68 0.63 'x2 -21.7 -18.2 -14.4 -10.7 -7.4 -4.9 -3.1 -2.3 -1.8 -1.5 Scenario 2 R 0.99 0.96 0.91 0.84 0.73 0.58 0.38 0.25 0.16 0.10 G 5.71 5.60 5.45 5.27 5.06 4.86 4.86 4.86 4.86 4.86 r -0.00 -0.01 -0.02 -0.03 -0.06 -0.11 -0.19 -0.27 -0.35 -0.43 e r 1.00 0.99 0.98 0.97 0.94 0.90 0.82 0.76 0.70 0.65 0). 2 "2" indicates baseline age-specific natality values have been doubled. allow for full population replacement (r >0) given the fishing mortality, became progressively smaller as the value of r increased (see Table 2 for reference). Increasing S by 10% (scenario 9) allowed for an age at first capture of 5 years, while doubling m x (sce- nario 4), reducing t t to 3 years of age (scenario 5), increasing S by 50% (scenario 10), or increasing S by 10% (scenario 11) all had the same effect of allow- ing for an age at first capture of 4 years (90 cm TL) compared with 6 years (99 cm TL) under the best case scenario. Reducing t by 2 years (scenario 6) or increasing £ to 20 years (scenario 13) both al- lowed for an A of 3 years (82 cm TL), while reduc- ing t mat by 1 year and doubling m x (scenario 7) al- lowed an A of 2 years. Under the most extreme manipulations, which included reducing t mat by 2 years and doubling m x (scenario 8), increasing S by 50% (scenario 12), and reducing t mat to 3 years, dou- bling m x , and setting S and S o at 95% (scenario 14), an age at first capture of 1 year (55 cm TL; scenarios 8 and 12) and of years (32 cm TL) could be applied in a given year. Discussion These demographic analyses using the best available information indicate that the Gulf of Mexico popula- tion ofR. terraenovae may be very vulnerable to fish- ing pressure. Results showed that, based on known life history parameters, the population's intrinsic rate of increase was, at best, only r=0.044, equating to a finite rate of e r =1.045, which is much lower than the rate estimated for "small coastal" species in the stock assessment used to develop the FMP for sharks of the Atlantic Ocean (e r =1.91). Furthermore, compa- rable rates to the FMP values were only obtained after extreme manipulations of the input life history parameters, which diverged too widely from observed life history parameters to be realistic. For example, one of the possible scenarios that would yield an es- timate of e r of 1.91 implies that age at maturity has to be decreased from 4 to 3 years, fertility doubled, and survivorship increased by almost 50% relative to the most optimistic initial scenario, i.e. the best case scenario, resulting in estimates of 29.6 for R o , 6.7 for G, and 1.1 for t x2 . This means that, in the absence of fishing, the population would almost double every year. Rhizoprionodon terraenovae is the main species caught in the Texas recreational shark fishery and is also caught by the headboat and other recreational fisheries in the Gulf of Mexico. More importantly, it represents a significant bycatch in the shrimp trawl fishery operating in the Gulf of Mexico and to a lesser extent in the longline reef fish and shark fisheries, and in the gillnet fishery in the same area. The lack of data on the age and size at which individuals of R. 64 Fishery Bulletin 93(1). 1995 terraenovae first enter these fisheries, as well as the relative proportions of each age and size group rep- resented, preclude a more detailed analysis at this time. However, the demographic analysis represent- ing the best case scenario indicated that under the present fishing level R. terraenovae should not enter the fishery until individuals reach about 97 cm TL or almost 6 years of age if the population was managed to just replace itself. There is evidence that smaller ani- mals are being caught in the various fisheries, but the proportions of each age class are unknown. The biological parameters incorporated in sce- narios 1 through 3 represent the best, most reliable information available. Data on age and growth were taken from a tetracycline-validated laboratory study (Branstetter, 1987) which indicated that females mature entering their fifth year of life (age-4) and that maximum age is between 8 and 10 years. In another study (Parsons, 1985), female maturity was estimated at between 2.4 to 3.9 years. However, this study used only males and mean lengths for age classes, which, as pointed out by Branstetter (1987), affected the von Bertalanffy parameters. The possi- bility of earlier female age at maturity and even longer lifespan was incorporated in several of the demographic analyses (scenarios 5, 6, 7, 8, 13, and 14), which evidently yielded more liberal results on which more risk-prone management decisions could be based. The unpublished information on fertility at age was derived from a study on the reproductive biology of R. terraenovae (Parsons, 1983) and relates female to- tal length to number of uterine eggs or embryos for 78 specimens. Parsons ( 1983) also noted that tropical popu- lations of R. terraenovae had been reported to have as many as 12 embryos. This possibility was taken into account by doubling fertility at age in several analyses (scenarios 4, 7, 8, and 14), which again produced more optimistic estimates of population parameters. The age, growth, and reproduction information used in this study was based on animals collected in the northern central and western Gulf of Mexico. The extent to which this information is applicable to the entire population or whether there are different stocks in the Gulf with different age, growth, and reproductive capabilities is not known. For example, I recently examined an 82-cm-TL pregnant female with 3 embryos, measurements which fit nicely the regression equation of Parsons (1983), but which would result in a back-calculated age of 3 years with the von Bertalanffy growth function, although the female could have been older, e.g. age 4, owing to variability in size at age, which is not uncommon in sharks (Kusher et al., 1992, and references therein). The most important and also the most difficult parameter to estimate is natural mortality (M). The value of M used in this study was taken from Hoenig's ( 1983) relationship between longevity and total mor- tality for virgin or lightly exploited stocks. The as- sumption that Z could be approximated to M, or that no fishing mortality occurred during the period for which growth parameters for this species were de- rived, may have been violated. However, the possi- bility of lower natural mortality values was incorpo- rated in several analyses (scenarios 9 through 14). While Hoenig's equation represents a shortcut and obvious simplification of reality, the lack of catch and effort data, or age or size composition for stocks of this species precludes calculation of any other esti- mates of M at this time. Lack or inappropriateness of both fishery and biological data may explain why several other researchers have used the same ap- proach to estimate natural mortality in shark popu- lation studies. Except for age-0 Negaprion brevir- ostris (Manire and Gruber, 1993), no actual age-spe- cific estimates of natural mortality are available for any shark species. The value derived for M (0.42) in this study is equivalent to an annual survivorship of 0.66, which is low when compared with survival estimates for other species of sharks. Values derived from Hoenig's (1983) regression equation include 0.82 for the an- gel shark, Squatina californica, (Cailliet et al., 1992); 0.85 for N. brevirostris (Hoenig and Gruber, 1990); 0.87 for Triakis semifasciata (Smith and Abramson, 1990; Cailliet, 1992); and 0.90 for Carcharhinus plumbeus (Hoff, 1990). Grant et al. (1979) derived a value of 0.90 for the Australian school shark, Galeorhinus australis, using cohort analysis, and Walker ( 1992) used a value of 0.82 in a dynamic pool fishery simulation model of the gummy shark, Mustelus antarcticus, which was also obtained through cohort analysis. The lower survivorship value for R. terraenovae may be due to the smaller size of this species which would make it more sus- ceptible to predation by other sharks, especially at early ages, since pups are born at about only 30 cm TL in coastal waters about 10 m deep (Castro, 1993). The very high estimate of F (0.428) used in this study was derived from a shark stock assessment that is the basis for the recently implemented (26 April 1993) FMP for sharks of the Atlantic Ocean. However, the accuracy of this estimate, based on a 4-year catch-and-effort time series, is uncertain, and the demographic analyses undertaken in this study indicate that R. terraenovae is vulnerable to high removal levels in the early years of life. It is also possible that the age, growth, and repro- ductive data used in this study are only representa- tive of the population at a time when fishing pres- sure was not as high as it is at present. Potential Cortes: Demographic analysis of Rhizopnonodon terraenovae 65 Table 6 Life history parameters for severa species of sharks compared to the best case scenario for Rhizoprionodon terraenovae in the Gulf of Mexico. Species R G r e r 100 m separating adjacent stations, successive video deployments were likely indepen- dent, because the greatest distance offish attraction to the bait was only 48 to 90 m. This estimate was based on average maximum bottom current speeds of 0.1 to 0.2 m/s respectively (Bathen, 1978), a soak time of 10 minutes, and a swimming speed for opakapaka of 0.6 m/s (or approximately 3 body lengths (BL) per second, where one BL=20 cm; Videler, 1993). Depths of all video and longline sets were determined by depth sounders aboard the re- search vessels, and positions were determined by GPS (Global Positioning System) or sighting com- pass, as Loran-C capabilities were unavailable. Longlines were deployed approximately perpen- dicular to depth contours. Bottom longline operations used modified Kali longlines, 6 each with 30 individu- 6 Shiota, P. M. 1987. A comparison of bottom longline and deep- sea handline for sampling bottom fishes in the Hawaiian Ar- chipelago. Honolulu Lab., Southwest Fish. Sci. Cent., Natl. Mar. Fish. Serv., NOAA, Honolulu, HI 96822-2396. Southwest Fish. Cent. Admin. Rep. H-87-5, 18 p. Ellis and DeMartmi. Video camera sampling of Pristipomoides filamentosus abundance 69 21.6 21.5 Figure 1 Area of operations off Kaneohe Bay, Oahu, for video and longline sur- veys for opakapaka, Pristipomoides filamentosus. Stations for video and longline in 1992 are midpoints identified by solid circles. Area of video coverage in 1993 is enclosed within the dotted lines and stations are midpoints identified by hollow circles. ally weighted and buoyed 3-m PVC droppers. Drop- pers were attached along the main line about 18 m apart. A 9.07-kg test, hard monofilament branch leader and a 3.63-kg test, hard monofilament hook leader were used. Each dropper had five leaders with size-12 Izuo circle hooks (AH style), for a total of 150 hooks per longline set. Stripped squid was used as bait. The standard soak time was 30 minutes, and three to four sets were completed each day. Two separate, 8-mm video camera assemblies were used for the video operations. Each video camera was equipped with a No. 1 diopter magnification lens and a wide angle zoom lens with a red filter for underwa- ter correction. Camera focus, sensitivity, and white balance were manually adjusted, but an automatic aperture setting was used. The focus distance for both video cameras was fixed at 2.13 m, and the focal length of the lens was set at 11 mm. Each video cam- era was enclosed in an underwater housing and se- cured in a weighted frame (Fig. 2). A single, 15-cm long bait container was positioned 60 cm in front of the camera lens and mounted on a PVC rod. The bait container held a single (=0.5-kg) mackerel (Scomber sp.) and one whole squid (Loligo sp.) tie-wrapped to the outside, both of which were changed after each deployment. The camera assemblies were manually 70 Fishery Bulletin 93(1). 1995 Figure 2 Baited video camera assembly with bait container positioned 60 cm in front of the camera lens. for each video sequence, three indices of abundance were scored for each spe- cies taped: maximum number (MAXNO); time to first appearance (TFAP); and total duration in sequence (TOTTM). The MAXNO index was de- termined as the peak number of a spe- cies visible at any one time (maximum interval one second) during a deploy- ment. Fork length (FL) to the nearest 0.1 centimeter (cm) was recorded for fish caught on the longline. The FL of opakapaka observed on video was esti- mated and rounded to the nearest 5 cm by comparing fish swimming in the plane of the bait container with the known size of the container. Statistics An average maximum number offish re- corded for data collected in 1992 was cal- culated for each of nine sequences (three video stations) by using a mean weighted by the duration of each occurrence: lowered to the bottom and marked by a buoy; later they were raised to the surface by outboard engine power. Cameras were allowed to rest on the bottom for a standard interval of 10 minutes before retrieval. The duration and number of video camera deploy- ments to be used on the ship cruise were estimated on the basis of three earlier pilot deployments of the video assembly from small craft. These prior tests indicated that about 10 minutes were required to deploy and retrieve the camera assembly. The time to first appearance (TFAP) of opakapaka from the three pilot stations was 227 ± 300 sec (mean ± 1 stan- dard deviation of the data [SD]) after bottom con- tact. A bottom time of 10 minutes was chosen to ac- commodate likely extremes and also to allow 6 deploy- ments per 2-h tape (20 min per deployment x 6 deploy- ments). With two cameras, 12 deployments per day could be made without changing tapes. The maximum number of longline sets was limited to four per day, based on three camera deployments per longline set. Types of data Species presence, total number of individuals per species, and the number of hooks lost were recorded for each longline set. Species presence and duration of squid bait attachment to the bait container (BTM) were recorded for each video sequence. In addition, £*„ N h A w (1) N where X w =the weighted average maximum num- ber of fish, rc=total number of occurrences, X h = maxi- mum number of fish seen in the h th occurrence, N /i =duration (s) of the h th occurrence, and N=YN h = 600 s. Video indexes were calculated as means (of up to 3 deployments) to standardize for multiple deploy- ments per station. Video indices were derived in two logarithmic forms — mean of logs (ML), ]Tln(x,+l) ML = -^ (2) and log of means (LM), LM = In + 1 (3) where ac- = individual datum for a variable (i.e., the value for the variable MAXNO, TFAP, or TOTTM for each deployment at a station) and /i=number of de- ployments per station. The longline index consisted of log-transformed individual set data [In (catch + Ellis and DeMartmi. Video camera sampling of Pristipomoides filamentosus abundance 71 1)]. The number of stations where each species was caught or seen was also tallied for each gear type. For nonzero mean data collected in 1993, the best form of the video index (LM\ see Results section), was calculated as follows: (4) 7 \" LM' = ln x*. .=1 n V /_ where x - individual datum for MAXNO and rc=number of deployments per station. A matrix of Pearson's correlation coefficients was calculated for 1992 data (SAS, 1987) with the log- transformed variables to detect interrelationships among all the video and longline indices. Spearman's rank correlations were also calculated and compared. Multiple linear regression (SAS, 1987) was used to estimate the effect of competition between opakapaka and puffers for longline hooks on the basis of the fol- lowing model: Y = p 1 X 1 + p 2 X 2 +e, (5) where F=ln (opakapaka video MAXNO), Xj=ln (no. hooks lost + no. puffers caught), andX 2 =ln (number of opakapaka caught). The model was run as a for- ward regression without an intercept and with an entry level for significance equal to P<0.10. The pre- cision (repeatability) of video and longline was de- scribed by the coefficient of variation (V, Sokal and Rohlf, 1981; Zar, 1984): V = f-^-|xl I mean , (6) where SD is the standard deviation. Longshore station and relative depth effects for 1993 data were analyzed by using standard para- metric and nonparametric procedures (SAS, 1987). Sample size and power analysis We evaluated video and longline data in a power analysis for the £-test of means. Specifically, we esti- mated the sample sizes required to detect a twofold change in abundance by using either sampling method. Skalski and McKenzie (1982) set a prece- dent for use of the criterion of twofold change in en- vironmental monitoring studies; annual variations much larger than this are typical for marine fishes (Hennemuth et al., 1980; Francis, 1993). The effect size (ES) was calculated as follows: ES = 0.693 SD (7) where 0.693= I ± twofold difference in x I for the natu- ral log ( x ) and SD is the standard deviation. Cohen (Tables 2.3.4 and 2.4.1, 1988) was consulted for the requisite sample sizes. The ES for each gear was evaluated at 13=0.20, power (1-J3)=0.8, and a 2 =0.05. For the 1993 data, ES was also evaluated at a 2 =0.1. Results and discussion Sample composition The mean time to first appearance (TFAP) of opakapaka for 1992 video tapes with opakapaka present (all islands included) was 203 ±165 (SD) sec- onds. The total time (TOTTM) of opakapaka during a deployment averaged 122 ± 133 seconds. The maxi- mum number (MAXNO) of opakapaka appeared on tape at approximately 354 (±153) s, based on the nine video sequences for which the weighted average MAXNO ( X w ) was calculated. These data confirm our initial choice of a 10-min bottom time. In 1992, only windward Oahu data were used for comparisons and statistical analyses, because the opakapaka measures from Maui and Kauai included large percentages (92% and 67%) of "double-zeros" (zero longline catch, zero fish recorded). Catches of P. filamentosus also were greatest for the windward Oahu site; 54 of the 58 juvenile opakapaka were longlined off windward Oahu. Puffers were preva- lent at windward Oahu and at Maui. Both longlined and video-recorded opakapaka were juvenile size (13 to 21 cm FL, and 15 to 25 cm FL, respectively; Kikkawa, 1984; Moffitt and Parrish 4 ). Frequency of occurrence data and total number of species differed between longline catches and video records (Fig. 3). Puffers ranked first in abundance and opakapaka second in both the longline and video data. Video cameras recorded the presence of opakapaka and puffers more often than did the longlines (Fig. 3). Video tapes also recorded a greater diversity of species (Table 1), suggesting greater ac- curacy of the video system. Fish that were not caught by the longline but were seen on video included reef- associated species (e.g. pennant butterflyfish, Heniochus diphreutes, and whitesaddle goatfish, Parupeneus porphyreus, sharks {Carcharhinus sp.), and rays (Dasyatis sp.). Longlines also undersampled the lizardfish, Trachinocephalus myops (Fig. 3), a major component of this deep-water, soft-bottom fish assemblage. 5 No major differences in species compo- sition occurred in video surveys from 1992 and 1993. 1 992 video-longline relations The MAXNO index for opakapaka and puffers was highly correlated with the total duration on film 72 Fishery Bulletin 93(1), 1995 15 O a 0} Ih S-, o o O I G o -t-> CB -♦-> 6 2 10 TORQ = FILA = PORP SHRK DASY TRAC SCOU SERI HENI KASM | video V/A longline Torq-uiganer florealis Pristipomoides filamsntosus = Parupeneus porphyreus = Carcharhinus ap. = Dasyatis sp. Trachinocepfialus myops = unidentified Scombndae Soriola dume-iiii = Hemochus diphreutes = Lutjanus kasmira J I Liu TORQ FILA PORP SHRK DASY TRAC SCOM SERI HENI KASM SPECIES Figure 3 Frequency of occurrence of ten common species on video camera deployments and longline sets conducted during 1992 off windward Oahu (re = 15 stations). Refer to Table 1 for common names. Table 1 Total numbers of fish seen ar d caught at 15 video and longline stations located off windward Oahu during 1992. Total number for a fish taxon for video stations is the sum of the maximum numbers seen on 38 films. Total number for longline stations is the number of fish caught. Species Common names Video Longline Torquigener florealis Bleeker's balloonfish 221 80 Pristipomoides filamentosus Pink snapper (opakapaka) 94 54 Heniochus diphreutes Pennant butterflyfish 25 — Parupeneus porphyreus Whitesaddle goatfish 10 — unidentified Scombridae Tuna or mackeral 6 — Trachinocephalus myops Lizardfish 5 — Carcharhinus sp. Shark 4 — Sphyrna sp. Hammerhead shark 4 — Seriola dumerilii Amberjack 3 — Dasyatis sp. Stingray 3 — Chaetodon miliaris Milletseed butterflyfish 3 — unidentified teleosts Bony fish 2 — Parupeneus pleurostigma Sidespot goatfish 1 — Sufflamen fraenatus Bridle triggerfish 1 — Canthigaster rivulata Maze toby 1 — Parupeneus sp. Goatfish 1 — Lutjanus kasmira Bluestripe snapper •3 Ellis and DeMartini: Video camera sampling of Pristipomoides filamentosus abundance 73 (TOTTM) and time to first appearance (TFAP) of the respective species (Table 2, LM form). The duration of squid bait (BTM index) was significantly corre- lated with the MAXNO index and the other video indices for opakapaka but was more strongly corre- lated with the MAXNO index for puffers (Table 2). Videos indicated that puffers were usually respon- sible for the removal of the squid bait; a direct rela- tionship between puffer numbers and the rate of bait disappearance was evident. Spearman's rank corre- lations mirrored the parametric correlations. After log-transformation, the data pairs were ap- proximately bivariate normal. Among all the video indices, MAXNO was best correlated with InCPUE (LLNO) for opakapaka (Table 2). The ML and LM forms of the MAXNO video index were compared separately with the longline CPUE, and the LM form provided a slight but consistently better Pearson's correlation than did the ML form for both opakapaka and puffers. Therefore, the LM form of the MAXNO index was used for all further parametric compari- sons and analyses. The MAXNO-CPUE relationship was approxi- mately linear (Fig. 4A), and its residual plot showed Table 2 Correlation between log-transformed mean video indices (LM) and log-transformed longline catch per unit of effort (InCPUE) from the 1992 windward Oahu site (n=15 sta- tions). Pearson correlation coefficients (r) are displayed above their respective P- values (Prob> \R\ , H o . Rho=0) for Pristipomoides filamentosus and Torquigener florealis. MAXNO = maximum number seen on tape; TOTTM=total duration of a species an tape; TFAP=time to first appear- ance of a species; BTM=duration of external squid bait; and LLNO=longline CPUE. TOTTM TFAP BTM LLNO Pristipomoides filamentosus MAXNO 0.9665 -0.9143 -0.5748 0.7855 0.0001 0.0001 0.0250 0.0005 TOTTM -0.8500 -0.5681 0.7285 0.0001 0.0271 0.0021 TFAP 0.5729 -0.6467 0.0256 0.0092 BTM -0.2982 0.2803 Torquigener florealis MAXNO 0.9465 -0.5770 -0.6654 0.5365 0.0001 0.0243 0.0068 0.0392 TOTTM -0.6030 -0.5902 0.5932 0.0173 0.0205 0.0198 TFAP 0.5141 -0.1143 0.0499 0.6851 BTM -0.5193 0.0473 neither discernible pattern nor slope (P=1.0, Fig. 4B). If all double-zero data are deleted, the correlation between video MAXNO and longline CPUE loses sig- nificance (r=0.55, P=0.08, n = ll). However, the double-zero data were retained in subsequent analy- ses because there was no a priori reason to believe they did not represent real absences. The observed magnitude of hook loss ( x =32%) in- dicates that longline CPUE is fundamentally inac- curate and biased for sampling this habitat and spe- cies assemblage. Apparently, most hook loss occurred when puffers bit through the leader above the hook. Hook competition is often a problem with longlines when hooked fish begin to saturate available hooks (Rothschild, 1967). Removal of hooks has a similar effect. A multiple linear regression with two descrip- tive variables, a puffer factor (Xj) equal to the num- ber of hooks lost plus puffer catch and opakapaka catch (X 2 ), was run to determine the effect of puffers on the relation between longline CPUE and the video MAXNO index for opakapaka. X : and X 2 were first determined to be uncorrected (r 2 =0.02, P=0.62). The model (Eqn. 5) for the multiple regression was forced through the origin, because neither sampling device could record the presence offish in its absence. The total variation in the opakapaka video index ex- plained by the model was 87% (i? 2 =0.87, P<0.001). Opakapaka longline CPUE explained 83% of the varia- tion (^=0.83, P<0. 001), and the puffer factor explained an additional 4% of the variation (^=0.04, P=0.07). The latter observation suggests that the puffer factor might strongly influence video-longline relations for opakapaka at times of relatively high puffer abundance. Precision for longline and video cameras was sepa- rately examined. For both opakapaka and puffers, cameras had nominally but consistently better pre- cision (V=81% and 48%) than did longline CPUE (V=91% and 71%). 1 993 video statistics The MAXNO video index did not differ between shal- low and deep positions (Student's r=0.27, P=0.79; Kruskal-Wallis x 2 =0.09, P=0.76) in May 1993 (Fig. 5). The mean MAXNO data lack a monotone trend over stations (P=0.5), even though raw MAXNO values were atypically large at several stations 20 - o 15 - O 9 1 O 10 - i 1 ii CI p p o O i I 5 - ? ( I p II • II • O • 8 ( ) c > II • • o 1 » • • 5 10 15 20 Station Figure 5 Scatterplot of maximum number of opakapaka observed (MAXNO) for shallow and deep positions and their mean by stations for data collected during May 1993. Stations are ordered in geographic sequence from farthest southeast (sta. 1) to farthest northwest (sta. 18). Means with component data hidden represent coincident shallow and deep values. 1992) observed that, although time of arrival of the first fish (TFAP) was strongly related to estimated fish densities, the maximum number offish seen at a station (MAXNO) either was unrelated or inversely related to densities. Our observation that MAXNO was highly correlated with an abundance estimate (CPUE ) may at first seem contradictory to these prior findings. However, there are important differences between our methods and those of previous studies: previous deep-sea work operated in unproductive depths >2,000 m and cameras recorded data for at least 11 hours per station, whereas our study was limited to productive depths <100 m for which a rela- tively short soak time ( 10 min) was sufficient. In the deep-sea studies, all bait was open to consumption. The partly internal bait of our system created a res- ervoir of odor that persisted for the soak duration in most cases; puffers removed all bait in only 3 out of 75 deployments. Differences in rates of bait consump- tion between the two deep-sea stations and result- ing variations in bait attractiveness may have con- tributed to the disparity between MAXNO and fish density in the deep-sea studies. The MAXNO and TFAP indices in our study were highly correlated (Table 2). This correlation suggests that the greater the density, the faster the fish ar- rive at the bait. These data agree with the observa- tions of Priede et al. ( 1990), where fish arrived at the camera faster at the station with presumed higher densities. Since the MAXNO and TFAP indi- ces were both significantly correlated with CPUE in our study (Table 2), the MAXNO index was chosen as the best index of abundance because it had the better correlation. A persistent bait source and short soak time may have contributed to this stronger cor- relation. In the future, the use of MAXNO as an in- dex of abundance should be reevaluated separately for each species and application. Conclusions Video cameras provide an accurate tool for sampling juvenile opakapaka, and the video MAXNO variable provides a relatively precise and accurate index of abundance. Based on 1993 data for a series of two camera deployments per station, minima of 17 to 22 pairs of deployments (34-44 sets) per study area would be necessary to detect a twofold change in ju- venile opakapaka numbers (at TRICHIURIDAE SCOMBRIDAE } - SCOMBRIDAE XIPHHDAE ISTIOPHORIDAE Scombrini Gasterochisma Grammatorcynus Scomberomorus + Acanthocybium V Sardini Thunnini Figure 1 Two phylogenetic hypotheses for the scombroid fishes based on morphological evidence. The studies of (A) Johnson (1986) and (B) Collette et al. (1984) are examples of the scombrid- subgroup and the scombrid-sister group hypotheses of billfish (Istiophoridae and Xiphiidae) relationships. Johnson ( 1986) considered billfishes a subgroup of the family Scombridae most closely related to the wahoo, Acanthocybium solandri. Collette et al. ( 1984 ) placed the billfishes as a sister group to the Scombridae. Both hypotheses propose that billfishes and scombrids share a common ancestor to the exclusion of other scombroids. An alternative hypothesis for billfish relationships is that billfishes are not scombroids. This hypothesis has never been depicted explicitly in the form of a cladogram (Gosline, 1968; Nakamura, 1983; Potthoff et al., 1980; Potthoff et al., 1986). al. (1984) study and several additional characters (Fig. 1). Like Collette et al., Johnson proposed that billfishes and scombrids compose a monophyletic group, but he regarded billfishes as a subgroup of the Scombridae. A critical piece of evidence support- ing this hypothesis that billfishes are a derived group within scombrids is the presence of cartilaginous in- terconnections between gill filaments in billfishes and the scombrid Acanthocybium solandri. Based largely on this proposed synapomorphy, Johnson placed Istiophoridae and Xiphiidae as derived scombrids and Acanthocybium as their sister group. This association has been suggested by others (Lutken, 1880; Fraser-Brunner, 1950). However the position of billfishes in Johnson's study was only weakly supported because of homoplasy. For ex- ample, five of the ten character-state transitions that support billfish monophyly on Johnson's cladogram are reversals. We will refer to the Johnson hypoth- esis as the scombrid subgroup hypothesis. Other workers have proposed that billfishes are not scombroids. In 1986, Potthoff et al. published a study of bone development in scombroids in which they discussed scombroid phylogeny They concluded that billfishes are not scombroids because of their lack of resemblance to other scombroids in vertebral number and osteological development. They sug- gested that these characters indicate billfish affini- ties to the percoids. This hypothesis has been sug- gested in previous studies (Potthoff et al., 1980; Nakamura, 1983). We will refer to this hypothesis as the nonscombroid hypothesis. It is evident from the morphological studies that there has been a great deal of homoplasious mor- phological evolution in billfishes. Therefore, it is dif- ficult to reconstruct the evolutionary relationships of this group based on morphology alone. In an at- tempt to derive additional, independent data on scombroid intrarelationships and, in particular, to address the position of billfishes, we compiled a mo- 80 Fishery Bulletin 93(1), 1995 lecular data set that consists of DNA sequences from the mitochondrial gene cytochrome b. This gene codes for a functionally conserved protein that should fa- cilitate sequence alignment over ancient divergences. Additionally, it has been used to examine both in- traspecific genealogy (Finnerty and Block, 1992) and much deeper phylogenetic questions such as the ori- gin of the mammalian orders (Irwin et al., 1991). The initial scombroid radiation probably occurred in the Paleocene epoch (Bannikov, 1985; Carroll, 1988). Therefore, cytochrome b sequence should be phylo- genetically informative about divergences within the suborder. The analysis presented in this paper builds on our earlier molecular study (Block et al., 1993). However, we have improved on the previous study in several ways which allow us to directly test the competing hypotheses of billfish relationships. First, we have obtained sequences from additional outgroups. The inclusion of presumably more distant outgroups per- mits us to address the question of scombroid mono- phyly. This is important because the nonscombroid hypothesis of billfish relationships argues that the Scombroidei is not a monophyletic group. Second, we include sequence information from the scombrid Acanthocybium, a taxon which is integral to the scombrid subgroup hypothesis. Third, we utilize sta- tistical tests to directly compare different hypoth- eses of billfish relationships. Finally, we emphasize character-state changes that accrue relatively slowly in order to minimize the effects of phylogenetic noise. Materials and methods Samples Partial cytochrome 6 sequences (590 base pairs) were obtained from 75 individuals representing 34 spe- cies of perciform fishes: 30 scombroid species and four putative outgroup taxa (Sphyraena, Coryphaena, Mycteroperca, and Morone; Table 1). We included Sphyraena based on the placement by Johnson ( 1986) of this taxon as the most primitive member of the Scombroidei. Several percoid taxa (Coryphaena, Mycteroperca, and Morone) were included because of the suggestion by some authors that billfishes are percoids (Gosline, 1968; Potthoff et al., 1980; Nakamura, 1983; Potthoff et al., 1986). Published cytochrome b sequences from two cypriniform fishes obtained from Genbank were used to root the phylo- genetic analysis (Crossostoma lacustre [Tzeng et al., 1990] and Cyprinus carpio [Chang, 1994]). We veri- fied the outgroup status of the cyprinids by first con- ducting a phylogenetic analysis using published se- quence from the sturgeon Acipenser transmontanus, a holostean, to root a parsimony analysis. We at- tempted but were unable to obtain full length se- quences (590 base pairs) from two fixed and preserved specimens of Scombrolabrax heterolepis possibly be- cause of DNA degradation in these specimens. DNA extraction DNA was obtained from frozen tissue samples of the mitochondria-rich "heater tissue" (found in Istio- phoridae, Xiphiidae, and Gasterochisma melampus; Block, 1986), red muscle, white muscle, or liver. Di- gestion of 0.1-0.6 g tissue was performed in ten vol- umes of extraction buffer containing 100 mM Tris CI (pH 8.0), 10 mM EDTA, 100 mM NaCl, 0.1% SDS, 50 mM DTT, and 0.7 mg/mL proteinase K. Digestion proceeded for 2-4 hours at 41°C. The homogenate was extracted twice with equal volumes of phenol (pH 8.0), once with 1:1 phenol/chloroform, and once with chloroform. The final extract was precipitated with 1/9 volume of 3M sodium acetate (pH 5.2) and 2.5 volumes of 100% ethanol. DNA amplification and sequencing The polymerase chain reaction (PCR) was used to am- plify a 700 base pair region of cytochrome b. A 305 base pair segment (not including primers) was generated by using published oligonucleotide sequences (Kocher et al., 1989). We amplified an overlapping, 425-bp region farther downstream with primers L15079 (5- GAGGCCTCTACTATGGCTCTTACC-3') or L15080 (5- CGAGGCCTTTACTACGGCTCTTACCT-3) and H15497 (5'-GCTAGGGTATAATT GTCTGGGTCGCC- 3). Double stranded amplification was performed in a 100-uL volume containing 50 mM KC1, 10 mM Tris- HC1 (pH 8.3), 1.5-3.0 mM MgCl, 200 ^M of each dNTP, each primer at 1 mM, 1 |j,g of template DNA, and 2 units of Amplitaq DNA polymerase (Perkin- Elmer/Cetus). Most templates were amplified through thirty cycles of PCR [1 minute denaturation (92-95°C), 1 minute annealing (40-50°C), and 3 min- utes extension (72°C)] on an Ericomp thermal cycler. Alternatively, PCR was performed on a DNA Ther- mal Cycler 480 (Perkin-Elmer) with the following temperature cycling regime: 5 cycles of 1 minute de- naturation at 95°C, 1 minute primer annealing at 40°C, 1:30 ramp to 72°C, and one minute extension at 72°C, followed by 25-35 cycles with an annealing temperature of 45°C. An 18-pL aliquot of the double stranded product was run by means of electrophore- sis through a IX TBE 1% agarose gel (Sea Plaque, FMC) at 5 V/cm for 45 minutes. A single stranded template was produced by asymmetric PCR (Gyl- Finnerty and Block: Evolution of cytochrome b in the Scombroldei 81 Table 1 Partial cytochrome b sequences (590 base pairs) were obtained from 34 perciform fishes, including 30 scombroid species. Pub- lished cytochrome b sequences were also obtained from Genbank for two cypriniform fishes. Order and suborder' Family and species Common name n Locales 2 Perciformes:Scombroidei Istiophoridae Istiophorus platypterus sailfish 2 A,P Makaira indica black marlin 2 I Makaira nigricans blue marlin 8 A,P Tetrapturus albidus white marlin 2 A Tetrapturus angustirostris shortbill spearfish 2 P Tetrapturus audax striped marlin 3 P Tetrapturus belone Mediterranean spearfish 2 M Tetrapturus pfluegeri longbill spearfish 1 A Xiphiidae Xiphias gladius broadbill swordfish 6 A,P Scombridae Acanthocybium solandri wahoo 3 A Scomberomorus cavalla king mackerel 1 A Scomberomorus maculata Spanish mackerel 2 A Gasterochisma melampus butterfly mackerel 3 T Auxis thazard frigate mackerel 2 P Euthynnus affinis kawakawa 2 P Euthynnus alletteratus little tunny 2 A Katsuwonus pelamis skipjack tuna 2 P Thunnus alalunga albacore tuna 2 P Thunnus albacares yellowfin tuna 2 P Thunnus maccoyii southern bluefin tuna 2 T Thunnus obesus bigeye tuna 2 P Thunnus thynnus northern bluefin tuna 2 A Sarda chiliensis eastern Pacific bonito 1 P Sarda sarda Atlantic bonito 2 A Scomber scombrus Boston mackerel 2 A Scomber japonicus chub mackerel 2 P Gempylidae Gempylus serpens snake mackerel 2 P Lepidocybium ftavobrunneum escolar 2 P Ruvettus pretiosus oilfish 2 A Trichiuridae Trichiurus lepturus scabbard fish 3 A Perciformes:Percoidei Coryphaenidae Coryphaena equiselis pompano dolphin 2 P Serranidae Mycteroperca interstitialis yellowmouth grouper 1 A Percichthyidae Morone saxatilis striped bass 1 P Perciformes:Sphyraenoidei Sphyraenidae Sphyraena sphyraena Atlantic barracuda 1 A Cypriniformes Balitoridae Crossostoma lacustre hillstream loach Tzeng et al., 1992 Cyprinidae Cyprinus carpio carp Chang et al., 1994 ' Eschmeyer, 1990. 2 A=Atlantic ocean; P=Pacific Dcean; I=Indian ocean; T= Tasman sea; M=Mediterranean Sea. 82 Fishery Bulletin 93[1), 1995 lensten and Erlich, 1988) carried out in a 100-uL volume containing the same reactants as the initial PCR but using 10 uL of the dissolved gel band and reducing one primer concentration 100-fold. The product was washed by centrifugal dialysis with ster- ile water in Centricon microconcentrators (Amicon) to remove excess dNTP's. Sequencing was performed with the Sequenase kit (United States Biochemical, Cleveland, Ohio) by using the limiting primer from the asymmetric PCR reaction. Data from eight spe- cies were obtained by directly sequencing double- stranded PCR products. The template was purified prior to sequencing (either directly from the PCR reaction mix or following excision of the appropriate band from low-melt agarose) with Magic PCR Preps (Promega). Sequencing was performed with the Sequenase kit according to the specifications of Casanova et al. ( 1991 ). Sequences from Mycteroperca and Morone was obtained after first cloning the PCR products in pGEM t- vector (Promega) according to the manufacturer's instructions. Transformation was carried out by using XL-1 blue cells. Two positive clones were selected for each PCR product. Double- stranded sequencing (Sequenase 2.0) was performed following alkaline denaturation as recommended by the manufacturer. Sequence was obtained from both strands of the amplified fragment for all individuals. Analysis Sequences were aligned by using the Mac Vector pro- gram (IBI Biotechnologies). Maximum parsimony analysis was performed with PAUP 3.1. (Swofford, 1991). Neighbor-joining (Saitou and Nei, 1987) and UPGMA dendograms were constructed with Phylip 3.5 (Felsenstein, 1993). The strength of support for various nodes was assessed by using the bootstrap analysis (Felsenstein, 1985). Specific conditions for each analysis are contained in the figure legends. Competing phylogenetic hypotheses were com- pared by using the "enforce topological constraints" option of PAUP 3.1. This option allowed us to deter- mine the length difference between the most parsi- monious trees that support each hypothesis. The cla- distic permutation test for monophyly and nonmonophyly (Faith, 1991) was then used to ascer- tain whether the more parsimonious hypothesis is significantly better than the competing hypothesis according to the criterion of parsimony. The test was performed as follows. The actual length difference between trees supporting the two opposing hypoth- eses was obtained. Then 99 permuted data sets were constructed from the original data set by randomly shuffling the character states for each character. We then obtained the length difference between trees supporting the two opposing hypotheses for each permuted data set. If the actual length difference was matched or exceeded fewer than 5 times in all 100 data sets (the original data set plus 99 permuted data sets), then the more parsimonious hypothesis was considered to be significantly better than the less parsimonious hypothesis. This corresponds to a to- pology-dependent permutation tail probability, or T- PTP, of less than or equal to 0.05. The effects of character weighting on parsimony analysis were assessed by EOR weighting (Thomas and Beckenbach, 1989; Knight and Mindell, 1993): each type of nucleotide substitution was weighted according to the ratio of its expected number of oc- currences divided by its observed number of occur- rences, or EOR. There are six types of nucleotide substitutions if we disregard the direction of change: A«G, CoT, G<=>T, G<=>C, A»T, and A<=>C. The ob- served number of each substitution type was obtained through pairwise sequence comparisons. Pairwise comparisons were performed between sets of sister species (sister species were identified through an initial unweighted phylogenetic analysis; see Fig. 2). Sister-species comparisons were used for two reasons. First, within a clade, sister species will tend to rep- resent relatively recent speciation events. This recency lessens the chance that multiple substitu- tions have occurred at the same site and that more recent substitutions obscure older ones. Second, all comparisons between pairs of sister species are mu- tually independent. Therefore, if we restrict our com- parisons to sister species, we cannot count the same base substitution twice. We modified the method of Knight and Mindell ( 1993) to derive the expected number of substitutions in each class. This method accounts for differences in the frequencies of the four nucleotides that greatly influence the expected frequency of each substitu- tion type. For instance, if guanine residues are very rare, then substitutions of other nucleotides for gua- nine will also be rare. The L-strand base composi- tion of cytochrome b in scombroid fishes is strongly skewed (Table 2), as it is in other groups examined (for example, Irwin et al., 1991). Cytosines and thy- midines each compose nearly 30% of the total nucle- otide population whereas guanines compose less than 16%. In order to incorporate knowledge of the base composition into our derivation of the expected num- ber of each substitution type, we proceeded as fol- lows. First, the average frequency of each nucleotide (/) was obtained for all species used in the pairwise sequence comparisons. Second, the observed num- ber of each substitution type (S 0(i -j), where i and./ represent two different nucleotides, was obtained by summing the results from all pairwise comparisons of Finnerty and Block: Evolution of cytochrome b in the Scombroidei 83 hliophorus platyplerus Makaira nigricans Tetrapturus albidus Tetrapturus audax Tetrapturus angustirosrns Tetrapturus pfluegeri Tetrapturus belone Makaira indica Xiphias gladius Thunnus alalunga Thunnus albacares Thunnus maccoyii Thunnus thynnus Thunnus obesus Katsuwonus pelamis Euthynnus affinis Euthynnus alletleratus Auxis thazard Sarda chiliensis Sarda sarda Scomberomorus cavalla Scomberomorus maculata Acanthocybium solandri Scomber japon icus Scomber scombrus Gasterochisma melampus Gempylus serpens Ruvettus pretiosus Lci'idticxhmm flavohrunneum Trichiurus lepturus Istiophoridae Xiphiidae Scombridae Gempylidae Tnchiuridae Cor\phaena equiselis Mycteroperca interstitialis Morone saxatilis Sphyraena sphyraena Crossostoma lacustre Cyprinus carpio Figure 2 Phytogeny of the Scombroidei based on an unweighted analysis of 248 phylogenetically informa- tive nucleotide sites. The cladogram depicted is a strict consensus of four equally parsimonious trees identified by using a heuristic search procedure on the program PAUP 3.1. (Swofford, 1991): TBR (tree bissection and reconnection) branch swapping was performed on 10 starting trees generated through random stepwise addition of taxa. Crossostoma and Carpio were specified as the outgroup. Length, consistency index, and retention index are the following: L=1595, CI=0.317, RI=0.539. Circled numbers at nodes indicated the percentage of trials in which a given partition between taxa is supported in 1,000 replications of the bootstrap analysis (Felsenstein, 1985). Only nodes supported in >50% of bootstrap replications are indicated. sister species. The expected number of each of the six substitution types (S^ -|)was then derived as follows: i E[i^j -Ifi fj)( S 0[total])/3- We divide by three because three types of base sub- stitutions are possible for each base, and we are in- terested in obtaining an expectation for one of them. For example, the expected number of A<=>T substitu- tions equals the average frequency of A's (0.23) plus 84 Fishery Bulletin 93(1), 1995 Table 2 Nucleotide substitutions by type determined through where i and./ represent two different nucleotides, were taxa. The expected substitutions for each type, S EUolal] , and f- are the frequency of nucleotides i andy. Average C=0.32. pairwise alignments. The observed substitutions for each type, calculated by summing the results from 8 pairwise comparisons were calculated according to the formula S £(Ma;| =(/)+/')(S , Ma; |)/3, >ase frequencies for the 16 species are as follows: G=0.16, A=0.23, of sister where f T=0.29i TRANSVERSIONS Pairwise comparison Substitution Types TRANSITIONS A»G CoT GoT GoC AoT \»C Total Tetrapturus audax vs. Tetrapturus albidus 1 1 2 Tetrapturus angustirostris vs. Tetrapturus pfluegeri 7 1 8 Makaira nigricans vs. Istiophorus platypterus 1 21 1 23 Euthynnus affinis vs. Euthynnus alletteratus 5 27 3 3 38 Thunnus thynnus vs. Thunnus maccoyii 4 3 1 8 Scomberomorus maculata vs. Scomberomorus cavalla 13 38 1 1 8 11 72 Sarda sarda vs. Sarda chiliensis 6 15 3 2 26 Scomber japonicus vs. Scomber scombrus 25 35 2 5 6 6 79 Total observed substitutions 61 140 3 10 20 22 256 Expected substitutions (see Methods section) 33.28 52.05 38.40 40.96 44.37 46.92 256 Expected/observed ratio (EOR) 0.55 0.37 12.80 4.10 2.22 2.13 the average frequency of T's (0.29) multiplied by the total number of substitutions (256) divided by three, or 44.37 (Table 2). The weights used for each substitution type (Table 2) are the ratios of expected substitutions divided by observed substitutions for that substitution type, rounded to the nearest integer (expected divided by observed ratios, or EOR's). All EOR's less than one were rounded to one. Weights were entered into PAUP 3.1. (Swofford, 1991) in the form of a step matrix. Results Sequence evolution and interfamilial relationships Molecular data sets, such as the cytochrome b se- quences presented in this study, are known to encom- pass subsets of characters that evolve at different rates. Subsets of data that differ in their evolution- ary rates will also differ in their phylogenetic utility. Character state changes that accrue very rapidly should permit resolution of very recent divergences. However, these rapid character state changes can provide false inferences about distant relationships because of homoplasy The likelihood of reversals and independent acquisitions is high if a particular site is evolving rapidly because there are only four pos- sible character states (G, A, T, and C) and only six possible types of character state change (A<=>G, CoT, GoT, GoC, AoT, and AoC). Therefore, in order to make an accurate reconstruction of the earliest branching events in scombroid history, we should emphasize slowly evolving character state changes. In an effort to best utilize the phylogenetic infor- mation from both slowly and rapidly evolving char- acter state changes, our phylogenetic analysis pro- ceeds in several discrete steps. We begin with an unweighted analysis of all informative nucleotide sites. This analysis is strongly influenced by nucle- otide substitutions that accrue rapidly and should be most informative concerning recent speciation events. We then attempt to improve our resolution of more ancient divergences by giving greater weight to less frequent types of nucleotide substitutions. We conclude with a phylogenetic analysis based on the inferred amino acid sequences. The amino acid se- quences evolve very slowly and should provide our most reliable estimates of the earliest splits between lineages. In each instance, the phylogenetic analy- sis is preceded by a discussion of the evolutionary varia- tion in the character subset under consideration. Unweighted nucleotide analysis A 590-base pair fragment of the cytochrome b gene, representing positions 134 through 723 of the hu- man cytochrome b sequence, was aligned across all Finnerty and Block: Evolution of cytochrome b in the Scombroidei 85 1 Istiophorus platypterus 2 Makaira nigricans 3 Makaira indica 4 Tetrapturus albidus 5 Tetrapturus audax 6 Tetrapturus angustirostris 7 Tetrapturus belane 8 Tetrapturus pfluegeri 9 Xiphias gladius 10 Acanthocybiun solandri 11 Sccmberamorus cavalla 12 Sccnteranorus maculata 13 Gasterochisma melampus 14 Auxis thazard 15 Euthynnus affinis 16 Euthynnus alletteratus 17 Katswonus pelamis 18 Thunnus alalunga 19 Thunnus albacares 20 Thunnus maccoyii 21 Thunnus obesus 22 Thunnus thynnus 23 Sarda chiliensis 24 Sarda sarda 25 Scomber japonicus 26 Scomber scombrus 27 Gempylus serpens 28 Lepidccybium flavobrunneum 29 Ruvettus pretiosus 30 Trichiurus lepturus 31 Sphyraena sphyraena 32 Coryphaena equiselis 33 Morane saxatilis 34 Myctoperca interstitialis 20 40 60 80 100 120 TCCTTACACoc rmTLi ' ia xTATGCACTACACCTCAGACATCCc^^ A Y T A R .A..T. .A..T. A. .A. .A. .T..A .T. ..C.T. ..G. .R. .A.T. .. .A.T.G. .T .TG... ..G.C. .C C G. .A. ... AT. A T. .A. .A. -T. .C... .C.T. . .G. ..A.T.G. ..G.. .AGT... .TG. ..A.T... .TG.C. . .G.T. . .G.T. A. .C. A. . . A. X .A. .T. C. .T..T. ..T..A. ,T. . .G. .C T. .T. .T .A .A .A. .T. .C. .C.T. .C..C.T. G.RC T.X. .A. .A. .A. .A. .A. .C.C. .C.T. TG.. .A.T. TG...A.T. .G.A.A.T. .G.C. .G.C. .G.C. .C.T. .C.T. .A.T. .A.T. .TG.C. ..G.C. T. .T. .T. .A..C.TT .A. .C.T. ..T. A. .C.T. ..A. C. .T..C. CC .T..CC CC .A..T..A .A T C C.C. .CC. .CC.( .T X. .T. .A. .T. .A. .T. C C.C. .A. .A. .A. .A. .C.T. .C.T. TG.. .A.T. YG...A.T. A.T. .G.C. .G.C. .G.C. .C.T. .C. C .T. .C G..CC G..CC C.C. C.C. .T..C A.. .A. .A. .T..T. .A. .A ..C.T. .GC.C. .TG...A.T. .TG.T.AGT. . A. A. .C. C .T. .A. .T. CC • R..CC C.C .A. C.C .A. A. .C.T. ..A A. .C.T. . .A A. .C.T. . .A C. .C.T. ..T C.T. .TG. . .A.T G.C. .A C.T A. .T. .C .C.T. . .T C T. ..T C.CT...T. A A A .T. .T. . .C.T. .TG. . .A.T G.C. .A. .C C. .T G T T A. .C CCC. .C .A T T A. .C.T... A A. .C.T. ..A T. .C.T. . .T A. .C.T. . .A A. .TG.T. .T C A A. .C.T. . .A T. .C CC C.C.T..G T..C.T A. .C.T. .A .C -T. .C. .T. .G.C .GAC .GC.C . .G. . .A.T A AC. .A. .C T. .T. .A. -C .A. .C .CC< .CC( .C A. .C... .A. .T. .T..A. .A .C.C .C... TG.T. A.T. .G.7.A.T. .TG.C .TG.C .A. .C. .A..C. .C. • A. .C. C. .T..CC .C.C. A. G. .C.A. .A. X. C. T..CCI CC' .A T.X... -A A.. C.A ..C.T. .QG. .. .G.T.A.T.T. . .CA.A.T... T..T G. .A AC. .T..TGAC. .T...C.T. .T.X. X.A. C A A T..C..A. .T..T .G..A. .T..T. .T.G .T.A. X. X..C CC G..C G .G..A. -AA.C .A. X .A. .A. .A. .T.X. .. ..T.C.T. T. .T C. -T A. -T. . . .T. .A. X A. .T. .A T C. .CCC X. . . AA.T. .T. X. . .AC . .A. .T. .ACA T. X. .T.X T. 240 260 280 CCGCCTTajICGaCT3uT3raCICOCCTGAGGACAAAT TT X CG C X C r . .TCA. T X C . .TCA. T X c T ■3 1 i; Figure 3 Alignment of partial cytochrome b sequence (590 base pairs) across 34 species of perciform fishes and two species of Cypriniformes. Nucleotide position 1 is equivalent to position 134 of the human cytochrome b gene. Intraspecific polymorphism is indicated as follows: R=A/G, Y=C/T, M=A/C, S=C/G, K=G/T, W=A/T, H=A/T/C, D=A/G/T. Ambiguities are indicated by '?' thirty-six species included in the analysis (Fig. 3). No deletions or insertions were detected. Overall, 293 nucleotide positions are variable; 248 were poten- tially phylogenetically informative. As expected for a protein coding sequence, the degree of nucleotide variability differs according to codon position (Table 3). The third position is most variable and the sec- ond position is least variable. Differences in nucle- otide variability at the three codon positions are due to the fact that many third position substitutions are silent, whereas many second position substitutions result in nonconservative amino acid replacements. The differences in substitution rates between codon positions becomes more apparent when we compare 86 Fishery Bulletin 93(1), 1995 1 RT ..T.. c ..T.. ..T.. A. ..T. . ..T.. .A. ..T. . .A. T G..C.C. ..T. . A r T . .c. . .T.. . .C. A A T T 11 G. A T .C r T .T. A A A A ..G ..T.. ..C. A A T T T r. .A CT.C .CG. .. CC 12 A. A T G r .c. A A G r .T. A ..G C A T A A r CT.C. .CG.C. C ..G r. A A r O AT.C. A .CG. .. 17 G. .A. ..A. .T. .C .c. .c. .T. .A. .A. .A. .A. A. -A.CT .CC.C. ..G .T .A. .A. .c. .T. ..c. .A AT.C. .CG.C. .C. 18 G. ..A. .T. .c .T. .c. .A. .A. .A. .T. .T. .A.cr .CC.C. .TO ..G. .T .A. .A. .c. .T. .T. .T. ..c. .A AT.C. .A. .CG.C. .C. 20 G. T A T .c T .c. A r. A A .T. .A.cr ..C.C. .TO T A A r T ..c. .A AT.C. A .CG.C. r 21 G. .T. . .A. .T. .c .T. .c. .T. .A. .A. .A. .T. .A.cr .CC.C. .TG .T. .A. .A. .c. .T. .G.. ..c. .A AT.C. .A. .CG.C. .c. 23 G. r. T .c r .c. r A A r A .A.cr .CC.C. ..G ..G. .T. .A. .c. .T. CT.C. .A. .03... ..T 24 r. T .c c .c. r A A A ■y .A.cr .CC.C. ..G ..G. .T. .A. .c. .T. .Y. .T. .G.. . .c. CT.C. .G. .03... ..T 25 G. ..G. .c. .A. .c .G. .A. .A. .T. .T.cr .AC.C. .TG.G. . ..C. .A. .A. .c. .T. AT.C. .A. .CG.C. .CC 26 G. A r T .c. .A. r A A T .A A ...CA .AC.C. .TG ..T.. ..G. A A c T T .G. T r. ..c. T. AT.C. ..G... rr 27 G. A r. r .c. .TT.A. T A r .T. G ...c. TAC... .TG.G.. C .T A .A c ..T.A. T r ..c. .G AT.C. T .03... TOC.T. ..G ..T.. C A A T CT.C. .C.C. r 29 G. A .c r T .c. A G r G .G.C. TQC.T. . .G.G.. ..T. . ..c. A A c ..c. AT.. . ..G.C. C 30 A. .T.. .A. . . .c. .C. .c. .T. .A. .T. .T. .G. .TA.T. .A.CA .AC.C. .AG . .c. .A. .A. .c. .T. .A. . .CA.T AT... .A. .CG.C. .c ..T. T T A T T .T. T T Y 33 A. T T G A .A. .T r .TT.AT.A. T r .T. A G. .. .CC... .T A T .A. A . .c. .T CT.C. A ..G.C. r 440 460 480 500 520 540 560 580 1 CTGCTATGACTCTAATCCACCTCCTrTTCClOCACGAAACAaG^ 2 C. .A C Ft. .A A. .G G T C .C. .R. .C. .A. .A .GC... .G..G. A. .G..A. .C. .A. .C. .A. .GC... .G. .G. .G..G. .G. .C. .G. .C. 9 TA..CGCA. .CA 10 .A..C. .A. .AA 11 TC A. .AA 12 TG..C. .A. .AA 13 TG..C. .A. .AA 14 .A A 15 .A A. .AA 16 -A A. .AA 17 .A..C AA 18 .A. .C AA 19 .A.. 20 .A.. 21 .A.. 22 .A.. 23 TA. . 24 TA..C..A..AA 25 TG..AGCA..AA 26 TA. .GGCAG.G. 27 .A. .A AA 28 .A AT. A. 29 TG. .A. .A. .AA 30 TA..03CT..A. 31 T. .G.G.C.CC 32 TG..A.CT...T 33 GCC. .C. 34 .A..CT.T. .AC T. .T. TC. .. • A .A. .A. TC.T A. .G..T..G. .C. .C. .C A. .CC.C. .T C C. .T. .C A.TC. .AY TA.T C. .C T A.TA. .AC A.T C G. .T..A. .C.C .C. .C T C. .C A.T... A. .C AAT. . .A. TC.T A A T T. .T T. . .AC. .GT. . .TA C C TC.G. .T. .A. .C. .T. .T. .T C A.TC. .CC.C A C C TC A A G. .C. .A. .T. .T T. . .A A.T. .C.C C. .T. ..C C.C. AAT. .AC C.C.T.T. ..CT. .C.C. AAT. OCT A A C. .A A.TC. GC.G. .T. . .A.T C C .T A. .T T C.C. .C .C. .AAT. T.A TC.T A A C .? A.T. . .AC A C C .C A C.T. .C AAT. T.A TC A T C. .A Y. .Y. .A.T. . .A A.T. .G. .C C .C A C.T. .C. .C. .AATT. A TC.T. T T T. .T. .A.TC. .A A.T C .C.T. .C.TGAT. . . .AA.TC.T T T T T. .A.TC. -A . . .AA . . .AA . ..AA TC.T. . . . TC.T. .T. TC.T. . . . TC.T TC.T. . . . TC.T. . . . TC T T A.. .G..A.. .A. .A.. T. .T. .A.TC. .A .. .A.T. . T. .T. .A.TC .A .. .A.T. . T .T. .A.TC. .A .. .A.T. . .A.TC. .GC . . .A.T. . . .C. -A.T. . .ACG.. . . .A . .C. .A.T.. .CC . -TA.T. . . .C. . .A. .T C. . . .T. .T T A.T. . .CC TA.T C. . X. .C. . .C..G. . T.. T.. A.. .T. -G. . ...C.T C.TGAT... G . ..C.T C.TGAT.. .G ...C.T C.TGAT... G . ..C.T C.TGAT... G . .TC.T. .A. .C. .AATT. .G ..TC.T. .A AATT. .G ...C.T..C T. ..C TA.G. .T C.C .C TT. .C. .T. .A T T... . .T..A..T A..T . . .A .03... CA. CA. . -CCTCA. .CCCACG. .GG.G.CA. . .G.G.CA. . .G.A.CA. . .G.A.CA. . .G.A.CT. . .G.G.CA. ..G.A.CA. . .G.A.CG. . .G.A.CA. . .G.A.CA. .TO.. .CA. .TG...CA. .CG.A.C. T. .A. -CG .03T .CGT.A. . .TG . .G.T.. . .T3 .TO .CG.T. .T 03 CG .CG .03 .03 .03.. . .G.G .TT. . ..T ..T ..T ..T ??? ..T TC T..A. .C T..T..G..C T T C.C A C C. .C T C GAT. . .AG.G C ..GG..??? GC.A A A A. .T A.T. . .CC.C A C A TC A. ,C .T.TA. .A A.C G CC.T G T T A.T. . .AC.C TA C C T. .G T CG. .C AAT. . .C. . .G. . -CA. -T T .C.T. .T C G. .C C.T GCTA. .ACT A. . -T. . .C C T. -T. .T. .T T. . .C.T. .A AATTA.A. .C. .OCT. . .C . .A.T CG.T AA.C A G A. .C ACT3. .ACT A. . .T. . .C .G. .C A. A. A CG. .C TA. .G. .C. .C.C .C. . .C. .TAG. .G. . . -T. -A. .A T G. .T. .C .C .T T CC T. . .T.C OC.C A T. .T TC.ACT. .C . .A. A.T. . .GT.G3. ,CTT. .C. -A. . . CT.A TT.G T C TEA. .OCT T. .T.T. . .C CC T T G T G. -C. -A. . -G.C . .G G T TT T T GTT. . .C CT3. .GC.T T C.C A T. .A. .T T. .TC A A G. .C .CCA. -C .A. . . Figure 3 (continued) the inferred number of substitutions (Table 3). For example, if a nucleotide site is twofold variable, i.e. if two bases occur at that position in an alignment of all species, then at least one base substitution has occurred at that position during the evolutionary his- tory of the species concerned. Likewise, if a position is threefold variable, at least two substitutions have occurred, and so on. By this approximation, the 293 variable positions have experienced at least 521 sub- stitutions, and substitutions at the third position outnumber substitutions at the first and second po- sitions by nearly 4 to 1 and by more than 12 to 1, respectively. Figure 2 presents a phylogeny of the Scombroidei based on an unweighted parsimony analysis of all informative nucleotide sites. In this cladogram, only the relationships among recently diverged taxa are strongly supported. There is support for the mono- phyly of genera within the family Scombridae (Thunnus, Euthynnus, Sarda, Scomber, and Scorn- Finnerty and Block: Evolution of cytochrome b in the Scombroidei 87 Table 3 Variable sites in the cytochrome 6 nucleotide alignment (Fig. 2) according to codon position. Variable sites are fur- ther characterized according to how many nucleotide states are present: 2 states=twofold variable, 3 states=threefold variable, fourstates=fourfold variable. Total sites Variable sites A) twofold variable sites B) threefold variable sites C) fourfold variable sites Minimum inferred substitutions = [(A) + 2(B) + 3(C)] Phylogenetically informative variable sites Codon position 1 2 3 Total 196 197 197 590 72 29 192 293 50 27 69 146 15 2 49 66 7 74 81 101 31 389 521 42 18 188 248 beromorus), and for the monophyly of the family Istiophoridae. Interrelationships within the family Istiophoridae and the genus Thunnus are well resolved. No other nodes are supported by more than fifty per- cent of bootstrap replicates (Felsenstein, 1985). Fur- thermore, there is a substantial polychotomy. Weighted nucleotide analysis The lack of resolution in the unweighted nucleotide analysis is not entirely unexpected. From Table 3, we can surmise that many nucleotide sites have in- curred multiple substitutions and therefore the like- lihood of convergent substitutions or reversals is high. In order to minimize the confounding effects of these homoplasious base substitutions, we have weighted infrequent substitution types more heavily using a modification of the method of Knight and Mindell (1993). If we disregard the direction of char- acter change, we can place all nucleotide substitu- tions into six classes : A<=>G, C<=>T, G<=>T, G<=>C, A<=>T, and A<=>C. Through pairwise sequence comparisons we obtained observed counts for each of these sub- stitution types (Table 2; also see Methods section). We observe a nearly 50-fold difference between the most common (Cc=>T) and the least common (G<=>T) substitution types. Then, from the total number of observed substitutions and the observed frequency of each base, we derived the expected number of oc- currences for each substitution type. The ratios of expected occurrences to observed occurrences for each substitution type (EOR's) were then used to weight the six types of base substitutions. The result of this weight- ing scheme is that substitution types that occur less frequently than expected are weighted more heavily. A phylogeny based on EOR weighting of nucleotide substitutions is presented in Figure 4. It retains all of the strongly supported nodes that appear in the unweighted topology. In addition, the weighted to- pology contains three more basal nodes that are strongly supported by the bootstrap analysis (>50%): the node uniting Gempylidae, Scombridae, and Trichiuridae, the node uniting Xiphiidae and Istiophoridae, and the node uniting Auxis and Euthynnus. This suggests that the character weight- ing scheme has accomplished its goal to some extent: we have retained the phylogenetic signal from rap- idly evolving substitutions while emphasizing the phy- logenetic signal from slowly evolving substitutions. According to the weighted cladogram (Fig. 4), all scombroids fall into two clades. The billfishes com- prise one clade consisting of a monophyletic Istiophoridae and its sister group, Xiphiidae. All other scombroids (Gempylidae, Scombridae, and Trichiuridae) fall into a separate clade. This major split within the suborder Scombroidei is in agree- ment with our previous study (Block et al., 1993). However, in contrast with our previous study, the use of character weighting and the inclusion of more distant outgroups leads to the result that the subor- der Scombroidei is not monophyletic. On the most parsimonious tree, Sphyraena and Coryphaena share a common ancestor with the gempylid-scombrid- trichiurid clade to the exclusion of billfishes, though this node does not receive particularly strong sup- port from the bootstrap analysis. This result indi- cates some support for the hypothesis that billfishes are not scombroids. More importantly, the cladogram excludes the possibility that billfishes and scombrids comprise a monophyletic group within the Scombroidei, as required by the scombrid subgroup and scombrid sister group hypotheses. In summary, the weighted analysis agrees with the nonscombroid hypothesis and conflicts with the scombroid subgroup and scombroid sister group hypotheses. Amino acid analysis Amino acid substitutions occur far less frequently than nucleotide substitutions owing to the strong functional constraints on many regions of the mol- ecule. Cytochrome b is a component of the electron transport chain and spans the inner mitochondrial membrane. The portion of the gene sequenced in this study encodes 195 amino acids corresponding to resi- dues 46 through 240 of the human cytochrome b (Fig. 5). 88 Fishery Bulletin 93|1). 1995 Isliophorus platypterus Makaira nigricans Tetraphtrus albidus Tetrapturus audax Tetrapturus anguslirostris Tetrapturus pfluegeri Tetrapturus belone Makaira indica Xiphias gladius Xiphiidae Thunnus alalunga Tlumnus albacares Thunnus maccoyii Thunnus thynnus Thunnus obesus Katsuwonus pelamis Euthynnus affinis Euthynnus alletteratus Auxis thazard Sarda chiliensis Sarda sarda Scombridae + Gempylidae Scomber japonicus Trichiuridae Scomber scombrus Trichiurus lepturus Scomberomorus cavalla Scomberomorus maculata Acanlhocybium solandri Ruvettus pretiosus Lepidocybium flavobrunneum Gasterochisma melampus Gempylus serpens Sphyraena sphyraena Coryphaena equiselis Morone saxatilis Mycteroperca interstitialis Crossostoma lacustre Cyprinus carpio Figure 4 Phytogeny of the Scombroidei based on a weighted, maximum parsimony analysis of informative nucleotide sites. The six types of nucleotide substitutions are weighted according to the ratio of their expected occurrence to their observed occurrence (see Table 3). Weights used for each sub- stitution type are the following: A«=>G=1, CoT=l, Gt=>T=13, GoC=4, A<=>T=2, and AoC=2. Crossostoma and Carpio were specified as the outgroup. The tree depicted is the single most parsimonious topology identified in a heuristic search: TBR branch swapping was performed on 10 starting trees generated through random stepwise addition of taxa. Tree length is 2,348 steps. PAUP 3.1. was unable to derive consistency and retention indices for the cladogram that incorpo- rated the weighting scheme. Circled numbers at nodes indicate the percentage of trials in which a given partition between taxa is supported in 1,000 replications of the bootstrap analysis (Felsenstein, 1985). This fragment spans four transmembrane domains and includes part of the region implicated as the outer membrane redox reaction center (Howell and Gilbert, 1988; Howell, 1989; Fig. 6). In a comparison of the inferred peptide sequences across the 36 species in- cluded in this study, 134 (69%) of the 195 amino acid residues are invariant, 34 (17%) occur in 2 amino acid states, and 27 (14%) occur in 3 or more states. Fmnerty and Block: Evolution of cytochrome b in the Scombroidei 89 Carpio Crossostcira Myctoperca Morcne Coryphaena Sphyraena Xiphias Istiophoridae (8) Thunnus (5) Euthynnus (4*) Sarda (2) Sccrrbertmorus c. Scorrberonrtrus m. Acanthocybium C^sterochisra Scxrrber japonicus Scarber scarbrus Gerrpylus Lepidocybium Ruvettus Trichiurus Carpio Crossostcira Myctoperca Morone Coryphaena Sphyraena Xiphias Istiophoridae (8) Thunnus {5} Euthynnus (4*) Sarda (2) Scccrbertmorus c. Scarberomorus m. Acanthocyiun Gastercchisma Sconber japonicus Scomber sccnbrus Genpylus Lepidocybium Ruvettus Trichiurus Alignment of in otide sequences gies) and the a sequence withi among the gen< studies are un 1993). 10 20 30 40 50 60 70 80 90 LTuLFIAMHYTSDISTAFSSViraCRI^^ I.. I SP. . -M. ... P.VES. LOO 110 Tyrmr .t .qAVPVMTwr 120 130 140 150 160 170 180 190 X^WlTOa^SVINAXLTOFFAFBFLLPFVIAAOTIEIi^^ .1 AAL.I... .IP AAV.VG. . .IP V. ..TVL.VL. . .1 AAL.IG.. .1 T AVL.V. .S .1 T AVL.M3.. Figure 5 ferred amino acid sequences. The amino acid sequences were inferred from the nucle- presented in Figure 3 by using the translation option of Mac Vector (IBI technolo- nimal mitochondrial genetic code. There was no variation in inferred amino acid i the family Istiophoridae, within the genus Thunnus, within the genus Sarda, nor traAuxis, Euthynnus, and Katsuwonus. Conserved positions identified by previous ierlined in the query sequence, Carpio (Howell and Gilbert, 1988; Esposti et al., This level of variability in amino acid sequence is very similar to that reported in a study of placental mammals, a group whose divergence times are prob- ably comparable to scombroids (Irwin et al., 1991). Much of the variation in scombroid cytochrome b occurs in the transmembrane portion of the molecule and represents substitutions between hydrophobic residues (leucine, isoleucine, and valine). The larg- est stretches of invariant residues (21 and 17) occur in a region implicated as part of the Q o redox reac- tion center (Howell and Gilbert, 1988; Howell, 1989; Fig. 6). All of the functionally constrained sites iden- tified by previous studies are conserved throughout the fishes included in this study (see Fig. 5; Howell and Gilbert, 1988; Esposti et al., 1993). Figure 7 presents a parsimony analysis based on 38 informative amino acid sites. The amino acid se- quences do not provide information about more re- cent speciation events because they evolve very slowly, but they contain important evidence about the relationship of billfishes to other scombroids. The amino acid analysis shares two important similari- ties with the weighted nucleotide analysis: first, Scombridae, Gempylidae, and Trichiuridae comprise a clade, and second, Sphyraena and Coryphaena share a common ancestry with this Scombridae- Gempylidae-Trichiuridae assemblage to the exclusion of the billfishes (Xiphiidae and Istiophoridae). The node uniting Sphyraena with the scombrid-gempylid- trichiurid clade is one of the more strongly supported nodes according to the bootstrap analysis. Therefore, cytochrome b amino acid substitutions support the non- scombroid hypothesis and conflict with the scombrid subgroup and scombrid sister group hypotheses. 90 Fishery Bulletin 93(1). 1995 NHj COOH Figure 6 Variability in amino acid sequence superimposed over a structural model for cyto- chrome b (Howell, 1989). Hypervariable residues, present in three or more amino acid states, are indicated by solid circles. Variable residues, present in two states, are indicated by open circles. The amino acids present at invariant residues are specified on the diagram. Residue 1 of this fragment is equivalent to the forty-sixth residue from the amino terminal of the protein in humans. The strength of the evidence that billfishes are not scombroids can be emphasized by directly examining the amino acid characters that are informative about this issue. Of the 38 informative amino acid sites, no sites unite billfishes and other scombroids to the ex- clusion of other perciforms, whereas eight sites unam- biguously separate billfishes from all other scombroids, i.e. sites where all billfishes possess one character state and all other scombroids possess some other character state (characters 12, 14, 15, 16, 113, 117, 140, and 169; Fig. 5). At all of these sites, billfishes share the same character state as one or more of the percoid fishes. Furthermore, at three of these eight sites (15, 16, and 169), Gempylidae, Scombridae, and Trichiuridae share a common state with Sphyraena to the exclusion of all other species in the study As this character analysis emphasizes, the amino acids are consistent with the hypothesis that billfishes are not scombroids and that Sphyraena is the sister group of a clade consisting of Gempylidae, Scombridae, and Trichiuridae. Intrafamilial relationships Within the family Istiophoridae (Istiophorus, Makaira, and Tetrapturus), cytochrome b nucleotide sequence provides a particularly well resolved and strongly sup- ported phylogenetic signal. This is probably due to the recency of the istiophorid radiation. The maximum se- quence divergence between any two species within this clade is less than five percent. We have performed a more in depth analysis of the interrelationships of istiophorids using the exhaustive search option of PAUP 3.1 (Swofford, 1991). Use of the exhaustive search op- tion guarantees identification of the most parsimoni- ous tree. The topology of this tree is identical to the topology of the istiophorid clade within the more inclu- sive scombroid phylogeny (Fig. 8, cf. Fig. 2). Neighbor- joining and UPGMA analyses produce an identical to- pology. Computer simulations suggest that agreement between these three methods should increase our con- fidence in a phylogenetic hypothesis (Kim, 1993). Finnerty and Block: Evolution of cytochrome b in the Scombroidei 91 / Sarda // Auxis-Euthynnus-Katsuwonus /// Thunnus /y^^ Gasterochisma VI n O ■^^ ifa / Scomberomorus cavalla Scomberomorus maculata 3 a CL V n + /%. (84) <^ /^ \^v O 1 r MK \\ / Acanthocybium T3 y\\\ \^\ Lepidocybium El n ,/ \\\ \^ Gempylus + H T); V=nucleotide transversions (C/T<=>A/G). Within the Istiophoridae transitions outnumber transversions 54 to 6. Neighbor-joining (Saitou and Nei, 1987) and UPGMA den- drograms produced with Phylip 3.5 (Felsenstein, 1993) have the same topology. Distance trees were constructed by using Kimura's (1980) two parameter genetic distance, and by assuming a tran- sition to transversion bias of 9:1. of cytochrome b does not support the monophyly of the genus Makaira. The black marlin, Makaira indica, appears to be the sister group of a clade con- taining all other istiophorids, while the blue marlin, M. nigricans, is sister group of the sailfish, /. platypterus. The most parsimonious tree that contains a monophyletic Makaira is six steps longer than the shortest tree overall (158 ver- sus 152), and on the most parsimonious tree, the M. nigricans-I. platypterus node is strongly supported by bootstrap analysis (85%). Cytochrome b provides good resolution of the relationships of the genera of the tribe Thunnini (Auxis, Euthynnus, Katsuwonus, and Thunnus). According to the nucleotide data the nine Thunnini species sequenced in this study com- prise two clades, one consisting of the genus Thunnus and one containing the other genera: Auxis, Euthynnus, and Katsuwonus. This dis- tinct split in the Thunnini was proposed by Kishinouye in 1923 and is consistent with the mor- phological hypothesis of Collette et al. (1984). Support for the monophyly of the Thunnus clade is particularly robust; however, the relation- ships within the genus cannot be resolved with- out the inclusion of both Thunnus tonggol and Thunnus atlanticus which were not sequenced in this study. The number of substitutions sepa- rating T. thynnus from T. maccoyii (<0.5% se- quence divergence) are small considering their status as separate species. Discussion Interfamilial relationships and the limits of the Scombroidei Throughout this analysis, we have focused on the long-standing controversy over the limits of the Scombroidei and, particularly, whether billfishes are scombroids. Cytochrome b appears to be informative on this issue. In the two phy- logenetic analyses that emphasize the more slowly evolving characters (see Figs. 4 and 7), the most parsimonious tree topology is clearly most consistent with the hypothesis that bill- fish are not scombroids: in each case, one or more nonscombroids share a common ancestry with the scombrid-gempylid-trichiurid clade to the exclusion of billfishes (Table 4). Therefore, according to the criterion of parsimony, the nonscombroid hypothesis is superior to the scombrid subgroup and to the scombrid sister group hypotheses. However, in our opinion, the Finnerty and Block: Evolution of cytochrome b in the Scombroidei 93 most parsimonious trees alone do not constitute suf- ficient evidence to reject these unfavored hypotheses. The question we must ask is the following: How unparsimonious are these hypotheses? In comparing the tree topologies that support each competing hypotheses (Table 4), it is clear that our data refute the notion that billfishes share a com- mon ancestor with the Scombridae to the exclusion of other scombroids (Gregory and Conrad, 1937; Berg, 1940; Fraser-Brunner, 1950; Collette et al., 1984; Johnson, 1986). For example, the shortest trees sup- porting a billnsh-scombrid clade are 13% longer than the minimum-length tree based on inferred amino acid sequence (Table 4). According to the cladistic permutation test for nonmonophyly (Faith, 1991), this length difference constitutes significant evidence against the monophyly of scombrids plus billfishes. The condition that billfishes and scombrids comprise a monophyletic group is a requirement of both the scombrid subgroup and scombrid sister group hypoth- eses. Therefore, according to the cytochrome b data, we reject these two hypotheses. The cytochrome b data clearly support the third hypothesis, that billfishes are not scombroids, though not as strongly as they refute the first two hypoth- eses. According to the inferred amino acid sequences, the shortest tree that supports scombroid monophyly places billfishes as sister group to all other scom- broids and is nearly 3% longer than the most parsi- monious tree overall (145 versus 141 steps). This length difference alone does not constitute signifi- cant evidence against the monophyly of the Scom- broidei according to a cladistic permutation test for nonmonophyly (see Methods section; Faith, 1991). However, as previously mentioned, there are three amino acid characters that unite scombrids, gempylids, and trichiurids with Sphyraena to the exclusion of billfishes (amino acids 15, 16, and 169). There are no characters that unite scombrids, gempylids, and trichiurids with billfishes to the ex- clusion of the putative outgroups. Our study is con- sistent with the hypothesis that billfishes are most closely related to some percoid lineage (Nakamura, 1983; Potthoff et al., 1986). The question of which taxon is most closely related to billfishes remains unanswered. On the basis of this evidence, we sup- port a conservative definition of the Scombroidei, including only the families Scombridae, Gempylidae, Table 4 Comparison of three competing hypotheses of billfish (Istiophoridae and Xiphiidae) relationships based on inferred amino acid sequences from cytochrome 6. 'na' = not applicable. Characteristics of tree which would support hypothesis Hypothesis I: Scombrid subgroup II: Scombrid sister group III: Nonscombroid Scombridae + billfishes are Billfishes are sister group of a a monophyletic group monophyletic Scombridae Acanthocybium is the sister group of billfishes Billfishes do not compose part of a monophyletic group with any other scombroid taxon or taxa Does the most parsimonious tree support the hypothesis? No No Yes If the answer to B is no, how much longer is the shortest tree which does support the hypothesis? 13.5% 13.0% (160 steps vs. 141 steps) (159 steps vs. 141 steps) na Based on the increase in tree length, can we reject the underlying hypothesis with statistical significance? Yes Yes (T-PTP = 0.01) (T-PTP = 0.01) na 1 The topology dependent permutation tail probability ( T-PTP; Faith, 1991) was used to determine the difference. Values of T-PTP<0. 05 were considered significant. See Methods section. significance of the length 94 Fishery Bulletin 93(1), 1995 and Trichiuridae (Cuvier and Valenciennes, 1832; Gosline, 1968; Potthoffet al., 1986). How can these inferences from molecular data be reconciled with the morphological data? We believe that this is an instance where molecular data comple- ment morphological data well. Cytochrome b provides an unambiguous phylogenetic signal that billfishes are genetically distant from other scombroids. In contrast, the existing morphological data does not clearly discriminate between a number of hypoth- eses. The number of character reversals in morpho- logical phylogenies that classify billfishes as scom- broids indicates that there have been many ho- moplastic changes in the billfish lineage. According to the morphological evidence, either billfishes are scombroids and have undergone several reversals to the primitive state, such as their low number of ver- tebrae, or billfishes are not scombroids but have evolved many convergent similarities to scombroids, such as their paired lateral caudal keels. Many of the morphological characters that unite billfishes to other scombroids, particularly to Scom- bridae, may be adaptations for continuous swimming, and are therefore of questionable phylogenetic value. These include hypurostegy, the projection of the cau- dal fin-ray bases anteriorly to cover the hypurals (Col- lette et al., 1984; Johnson, 1986), fusion of the hypurals (Collette et al., 1984; Johnson, 1986), and inter- filamentar gill fusion (Johnson, 1986). Hypurostegy and interfilamentary gill fusion are known to have evolved convergently in nonscombroid taxa (Luvarus imperialis [Leis and Richards, 1984]; and Amia calva [Bevelander, 1934]). The molecular data presented here provide a phylogenetic signal that is indepen- dent of convergent morphological adaptations that might confound phylogenetic analysis. There has been convergent evolution in the molecular characters, but unlike many of the morphological characters men- tioned, this convergent evolution does not appear to be the result of strong selection: most amino acid substi- tutions exchange amino acids with similar size, charge, and degree of polarity. Therefore, when compared with the existing morphological data, the phylogenetic sig- nal in the molecular data is less likely to have been obscured by similar selective pressures acting upon distantly related lineages. Istiophorid phylogeny Historically, there have been numerous disagree- ments over the number of species within the Istiophoridae and their interrelationships (Goode, 1882; Jordan and Evermann, 1926; LaMonte and Marcy, 1941; Nakamura, 1983). This is evidenced by the synonymies for many istiophorids, e.g. the Medi- terranean spearfish, Tetrapturus belone, has also been assigned to Istiophorus (Ben-Tuvia, 1953) and Makaira (Tortonese, 1958). The most thorough treat- ment of billfish systematics to date is a phenetic analysis conducted by Nakamura (1983). Nakamura recognized 11 species of istiophorid billfishes in three genera, including the designation of separate Atlan- tic and Indo-Pacific species for blue marlin (Makaira nigricans and M. mazara) and sailfish (Istiophorus albicans and /. platypterus). The molecular evidence presented here agrees with Nakamura ( 1983) in supporting the monophyly of the genus Tetrapturus, and within this genus, clades con- sisting of audax + albidus and pfluegeri + angus- tirostris + belone. Cytochrome b does not support the recognition of separate Atlantic and Pacific species of blue marlin and sailfish. Previous results (Finnerty and Block, 1992) identified substantial overlap in the cytochrome b haplotypes found among Atlantic and Pacific populations of blue marlin. The sailfish sample in this study includes one Atlantic specimen and one Pacific specimen that differ at only two sites among 590 (0.3%). We infer from the cytochrome b data (Block et al., 1993; and this study) a nonmono- phyletic Makaira and support for a clade consisting of the blue marlin (Makaira nigricans) and the sail- fish (Istiophorus platypterus). Based on the cytochrome b data, istiophorid tax- onomy at the generic level is not concordant with phylogeny. It is premature to suggest taxonomic re- vision of istiophorid genera, but we believe it is im- perative to obtain more molecular data, particularly from nuclear genes, to determine whether the infer- ences presented here can be corroborated. Further- more, we recognize the need for an extensive cladis- tic analysis of istiophorid relationships based on ad- ditional morphological data. Another taxonomic is- sue raised by this study concerns the number of valid Tetrapturus species. An extensive genetic survey of several populations from each species is required to determine the number of evolutionarily independent or reproductively isolated lineages within this genus. Relationships within the genus Thunnus The systematics of the genus Thunnus have been well studied owing to the commercial importance of tu- nas and interest in physiological specializations as- sociated with the evolution of endothermy Collette (1978) suggested a taxonomic subdivision of the ge- nus reflecting a split between tropical species (sub- genus Neothunnus: blackfin tuna, Thunnus atlan- ticus, longtail tuna, Thunnus tonggol, andyellowfin tuna, Thunnus albacares) and species that inhabit cooler waters (subgenus Thunnus: bluefin tuna, Finnerty and Block: Evolution of cytochrome b in the Scombroldei 95 Thunnus thynnus, southern bluefin tuna, Thunnus maccoyii, albacore, Thunnus alalunga, and bigeye tuna, Thunnus obesus). According to this hypothesis, the primitive condition for the genus Thunnus is a tropical distribution, and the cold water tunas com- pose a monophyletic group united by specializations that allowed them to exploit cooler temperate or deep waters. The nucleotide analyses presented in Fig- ures 2 and 4 are not consistent with this hypothesis. The cytochrome b phylogeny groups a tropical spe- cies, the yellowfin tuna, Thunnus albacares, with two species adapted for extremely cold water, the blue- fin tuna and southern bluefin tuna (Thunnus thynnus and Thunnus maccoyii). However, it is premature to draw conclusions about relationships within the ge- nus Thunnus until data are obtained from two tropi- cal species not included in this study, Thunnus atlanticus and Thunnus tonggol. Acknowledgments We thank A. Stewart, J. Kidd, A. Borisy, S. Eng, S. Malik, D. Wang, F. Manu, and V. Master for techni- cal assistance. We are indebted to C. Proctor, P. Grewe, G. DeMetrio, P. Davie, J. Pepperell, K. Dickson, B. Collette, and F. Carey for tissue samples. B. Collette, D. Johnson, J. Graves, and M. Westneat provided valuable assistance on the project and con- structive comments on the manuscript. This research was supported by NSF grant IBN8958225 to B. A. B. and NEH molecular biology training grant IBN8958225 and a Sigma Xi Grant-in-aid of Research to J. R. F. Literature cited Bannikov, A. F. 1985. Fossil scombrids of the USSR. Tr. Paleontol. Inst. Akad. Nauk SSSR 210:59-90. Ben-Tuvia, A. 1953. Mediterranean fishes of Israel. Bull. Sea Fish. Sta. Haifa 8:1-40. Berg, L. S. 1940. Classification of fishes both recent and fossil. Tr. Zool. Inst. Akad. Nauk SSSR 5<2):87-517. [In Russian.] Bevelander, G. 1934. The gills of Amia calva are specialized for respira- tion in an oxygen deficient habitat. Copeia 1934:123-127. Block, B. A. 1986. Structure of the brain and eye heater tissue in mar- lins, sailfish, and spearfishes. J. Morphol. 190:169-189. Block, B. A., J. R. Finnerty, A. F. R. Stewart, and J. Kidd. 1993. Evolution of endothermy in fish: mapping physiologi- cal traits on a molecular phylogeny. Science 260:210-214. Carroll, R. L. 1988. Vertebrate paleontology and evolution. W.H. Free- man and Co., New York. Casanova, J. L., C. Pannetier, C. Jaulin, and P. Kourilsky. 1991. Optimal conditions for directly sequencing double- stranded PCR products with sequenase. Nucl. Acids Res. 18(13):4028. Chang, Y.-S., Huang, F.-L., and T.-B. Lo. 1994. The complete nucleotide sequence and gene organi- zation of carp (Cyprinus carpio) mitochondrial genome. J. Mol. Evol. 38:138-155. Collette, B. B. 1978. Adaptations and systematics of the mackerels and tunas. In G. D. Sharp and A. E. Dizon (eds.), The physi- ological ecology of tunas. Academic Press, New York P- 7-39. Collette, B. B., T. Potthoff, W. J. Richards, S. Ueyanagi, J. L. Russo, and Y. Nishikawa. 1984. Scombroidei: development and relationships. In H. G. Moser et al. (eds.), Ontogeny and systematics of fishes. Allen Press, Lawrence, KS, p. 591-620. Cuvier, G., and A. Valenciennes. 1832. Histoire naturelle des poissons 8:1-509. Eschmeyer, W. N. 1990. Catalog of the genera of recent fishes. California Academy of Sciences, San Francisco, 697 p. Esposti, M. D., S. De Vries, C. Massimo, A. Ghelli, T. Patarnello, and A. Meyer. 1993. Mitochondrial cytochrome 6: evolution and structure of the protein. Biochimica et Biophysica Acta 1143:243-271. Faith, D. P. 1991. Cladistic permutation tests for monophyly and non- momophyly. Syst. Zool. 40(3):366-375. Felsenstein, J. 1993. PHYLIP (Phylogeny Inference Package), version 3.5. Distributed by the author, Univ. Washington, Seattle, WA. 1985. Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39:783-791. Finnerty, J. R., and B. A. Block. 1992. Direct sequencing of cytochrome 6 reveals highly di- vergent haplotypes in blue marlin, Makaira nigricans. Mol. Mar. Biol. Biotech. 1:206-214. Fraser-Brunner, A. 1950. On the fishes of the family Scombridae. Annu. Mag. Nat. Hist., Ser. 12(3):131-163. Goode, G. B. 1882. The taxonomic relationships and geographical dis- tribution of the members of the swordfish family, Xiphiidae. Proc. U.S. Natl. Mus. 4:415^33. Gosline, W. A. 1968. The suborders of perciform fishes. Proc. U.S. Natl. Mus. 124(3647):l-78. Greenwood, P. H., D. E. Rosen, S. H. Weitzman, and G. S. Meyers. 1966. Phyletic studies of teleostean fishes, with a provi- sional classification of living forms. Bull. Am. Mus. Nat. Hist. 131:339-456. Gregory, W. K. 1933. Fish skulls: a study of the evolution of natural mechanisms. Trans. Am. Philos. Soc. 28(2):75^81. Gregory, W. K., and G. M. Conrad. 1937. The comparative osteology of the swordfish (Xiphias) and the sailfish (Istiophorus). Am. Mus. Novit. 952:1-25. GyUensten, U. B., and H. A. Erlich. 1988. Generation of single-stranded DNA by the polymerase chain reaction and its application to direct sequencing of the HLA-DQA locus. Proc. Natl. Acad. Sci. U.S.A. 85:7652-7656. Howell, N. 1989. Evolutionary conservation of protein regions in the 96 Fishery Bulletin 93(1), 1995 proton-motive cytochrome b and their possible roles in re- dox catalysis. J. Mol. Evol. 39:157-169. Howell, N., and K. Gilbert. 1988. Mutational analysis of the mouse mitochondrial cy- tochrome 6 gene. J. Mol. Biol. 203:607-618. Irwin, D. M., T. D. Kocher, and A. C. Wilson. 1991. Evolution of the cytochrome b gene of mammals. J. Mol. Evol. 32:128-144. Johnson, G. D. 1986. Scombroid phylogeny: an alternative hypothe- sis. Bull. Mar. Sci. 39(1):1-41. Jordan, D. S., and B. W. Evermann. 1926. A review of the giant mackerel-like fishes, tunnies, spearfishes and swordfish. Occas. Pap. Calif. Acad. Sci. 12:1-113. Kim, J. 1993. Improving the accuracy of phylogenetic estimation by combining different methods. Syst. Biol. 42(3):331-341. Kimura, M. 1980. A simple method for estimating evolutionary rates of base substitutions through comparative studies of nucle- otide sequences. J. Mol. Evol. 16:111-120. Kishinouye, K. 1923. Contributions to the study of the so-called scombroid fishes. J. Coll. Agric, Imp. Univ., Tokyo 8:293-475. Knight, A., and D. P. Mindell. 1993. Substitution bias, weighting of DNA sequence evo- lution, and the phylogenetic position of Fea's viper. Syst. Biol. 42(1):18-31. Kocher, T. D., W. K. Thomas, A. Meyer, S. V. Edwards, S. Paabo, F. X. Villablanca, and A. C. Wilson. 1989. Dynamics of mitochondrial DNA evolution in animals: amplification and sequencing with conserved primers. Proc. Natl. Acad. Sci. U.S.A. 86:6196-6200. LaMonte, F. R., and D. E. Marcy. 1941. Swordfish, sailfish, marlin, and spearfish. Ichthyol. Contrib. Int. Game Fish Assoc. l(2):l-24. Leis, J. M., and W. J. Richards. 1984. Acanthuroidei: development and relationships. In H. G. Moser et al. (eds.), Ontogeny and systematics of fishes. Allen Press, Lawrence, KS, p. 547-551. Lutken, C. F. 1880. Spolia Atlantica. Dansk. Vid. Selsk. Skrift. Copen- hagen, Ser. 5, 12:409-613. McDowell, J. R., and J. E. Graves. In review. Genetic analysis of billfish population structure. Nakamura, I. 1983. Systematics of the billfishes (Xiphiidae and Istiophoridae). Publ. Seto Mar. Biol. Lab. 28:255-396. Potthoff, T., W. J. Richards, and S. Ueyanagi. 1980. Development of Scombrolabrax heterolepis (Pisces, Scombrolabracidae) and comments on familial relationships. Bull. Mar. Sci. 30:329-357. Potthoff, T., S. Kelley, and J. C. Javech. 1986. Cartilage and bone development in scombroid fishes. Fish. Bull. 84:647-678. Regan, C. T. 1909. On the anatomy and classification of the scombroid fishes. Annu. Mag. Nat. Hist., Ser. 8, 3:66-75. Saitou, N., and M. Nei. 1987. The neighbor-joining method: a new method for recon- structing phylogenetic trees. Mol. Biol. Evol. 4:406-425. Swofford, D. L. 1991. PAUP: phylogenetic analysis using parsimony, Ver- sion 3.1. Computer program distributed by the Illinois Natural History Survey, Champaign, IL. Thomas, W. K., and A. T. Beckenbach. 1989. Variation in salmonid mitochondrial DNA: evolution- ary constraints and mechanisms of substitution. J. Mol. Evol. 29:233-245. Tortonese, E. 1958. Elenco dei leptocardi, ciclostomi, pesci cartilaginei ed ossei del Mare Mediterraneo. Atti. Sco. Ital. Sci. Nat. 97(4):309-345. Tyler, J. C, G. D. Johnson, I. Nakamura, and B. B. Collette. 1989. Morphology ofLuvarus imperialis (Luvaridae), with a phylogenetic analysis of the Acanthuroidei (Pisces). Smithson. Contrib. Zool. 485:1-78. Tzeng, C.-S., S.-C. Shen, and P. C. Huang. 1990. Mitochondrial DNA identity of Crossostoma (Homalopteridae, Pisces) from two river systems of the same geographical origin. Bull. Inst. Zool. Acad. Sin. 29:11-19. Abstract. — Three species of nephropid lobsters have been rec- ognized in the genus Homarus: the American and European lobsters, H. americanus and H. gammarus of the northwestern and northeast- ern Atlantic, respectively, and the Cape lobster of South Africa, H. capensis, few specimens of which have been studied until recently. Analysis of new specimens allows reconsideration of the systematic status of this species and a subse- quent transfer to a monotypic new genus Homarinus. Far smaller than its northern relatives, with a maximum observed carapace length of 47 mm, the Cape lobster has first chelae adorned with a thick mat of plumose setae and less abundant setae on the carapace, tail fan, and abdominal pleura, whereas these setae are absent in Homarus. Relative length and shape of the carpus on pereopod 1, tooth pattern on cutting edges of first chelae, shape of the linguiform rostrum, large size of oviducal openings, and structure of male pleopods differ from corresponding features in Homarus. Comparative analysis of DNA from the mito- chondrial 16s rRNA gene demon- strated considerable sequence di- vergence of the Cape lobster (9.7%) from its putative congeners. The magnitude of this estimate relative to that between the two North At- lantic species (1.3%) further sug- gests that taxonomic revision is warranted. Assignment of Homarus capensis (Herbst, 1 792), the Cape lobster of South Africa, to the new genus Homarinus (Decapoda: Nephropidae) Irv Kornfield Department of Zoology and Center for Marine Studies University of Maine, Orono. Maine 04469 Austin B. Williams National Marine Fisheries Service Systematics Laboratory National Museum of Natural History. Smithsonian Institution Washington, DC 20560 Robert S. Steneck Department of Oceanography and Ira C. Darling Center University of Maine, Walpole, Maine 04573 Manuscript accepted 25 September 1994. Fishery Bulletin 93:97-102 (1995). Until now, three species of neph- ropid lobsters have been recognized in the genus Homarus Weber, 1795 (see Holthuis, 1991)://. americanus H. Milne-Edwards, 1837, the north- western Atlantic American lobster; H. gammarus (Linnaeus, 1758), the northeastern Atlantic-Mediterra- nean European lobster; andH. cap- ensis (Herbst, 1792), the South Af- rican Cape lobster. All are found in cool or cold temperate waters, and the North Atlantic species range into subarctic waters. The northern H. americanus and H. gammarus are well-known, abundant, and eco- nomically valuable species, but the southern H. capensis has long been problematic because only a few specimens (13 males, 1 female) were known to exist in collections (Barnard, 1950; Wolff, 1978; Hol- thuis, 1991). Gilchrist (1918) had seen only three specimens and re- marked (p. 46) that "it is a very rare species, and is not even known to Cape Fishermen." Kensley (1981) recorded its distribution in the Cape Province as Table Bay to East Lon- don, and recent new collections ex- tend the range to Transkei (Kado et al., 1994). Regardless of its rarity, sufficient specimens of the Cape lobster, liv- ing and preserved, are now avail- able for analysis of its distribution, morphological, and genetic at- tributes, and systematic status. Results of our studies indicate that this species should be removed from Homarus and placed in a genus of its own; this paper provides sup- porting evidence for this action and offers supplementary descriptive information on the species. Homarinus, new genus Figs. 1-4 Type species — Homarus capensis (Herbst, 1792) by present designa- tion and monotypy. Description — Carapace moderately compressed, narrower than deep, sparsely setose, middorsal carina barely evident on gastric region, ob- solescent on thoracic region posterior to deep cervical groove. Rostrum 97 98 Fishery Bulletin 93(1), 1995 Figure 1 Homarinus capensis (Herbst). Living male, carapace length 3.41 cm, photographed in an aquarium in Sea Fisheries Research Institute, Cape Town, South Africa, by Robert Tarr. (a) Left lateral; (6) dorsal. Kornfield et al.: Cape lobster taxonomy 99 a b d i— 2 — / Figure 2 Male pleopods (pi); mesial views of pi 1 (slight lateral folds on tips not shown in these views), and mesial views of appendix masculina on mesial ramus of pi 2: (a and 6) Homarinus capensis, left (USNM 251452); (c and d) Homarus americanus, right (USNM 13952); (e and f)H. gammarus, right (USNM 2085). Scale is 1 mm; bar 1 applies to c through f\ bar 2 applies to a and 6. linguiform in dorsal view, broad at base where mar- gins coalesce with orbits, margins bearing 4-6 small spines and gradually tapering anteriorly to rather abruptly pointed or narrowly rounded tip, reaching distal 1/3 of penultimate article of antennular peduncle, shallow dorsal concavity running its entire length. Telson and uropods with thick fringe of plumose setae on distal margin and with scattered non- plumose long setae dorsally on these appendages and sixth abdominal segment. Telson as wide at base as long, with lateral margins slightly sinuous and subparallel bearing obsolescent spines and rugae, each side ending in fixed posterolateral spine; ter- minal margin beyond spine broadly convex; distal 1/3 of surface bearing obsolescent transverse rugae. Uropods broadly subovate, sparsely setose on dorsal surface; mesial ramus broadest near posterior mar- gin with width about 0.73 length, row of obsolescent lateral marginal spines ending in fixed posterolat- eral spine; lateral ramus with width about 0.72 length, diaresis well behind midlength bearing row of fixed but irregularly worn spines ending in stron- gest spine at posterolateral angle. Chelae of first pereopods with thick coat of long plumose setae on upper surface of palm, overhang- Figure 3 Homarinus capensis (Herbst), tail fan (from figure in H. Milne-Edwards, 1851). ing extensor margin and distributed a distance along fixed finger; similar setae on mesial surface of car- pus and ventral surface of merus. Fingers not gap- ing; those of major chela with crushing teeth (often worn) opposed from near base to about midlength 100 Fishery Bulletin 93(1). 1995 Ha Hg He Ha Hg He Ha Hg He Ha Hg He Ha Hg He Ha Hg He Ha Hg He Ha Hg He ggtcgcaaacttttttgtcgatatgaactctcaaaataaataacgctgtt 5 atccctaaagtaacttaaatttttaatcaacaancaanggatcanttaca 100 . ca.c.a . t. . cacnnnnnnaaatatctctgtattttaaatttaaacagttacnnaaatta g t c....t..a....a..t tatcatcgtcgccccaacgaaataattntagtatataaataatattaaac c ...t ac.c g t.. tttcaactcatctaattatatactaaattattaagctttatagggtctta . . .t ..a...t g.a tcgtccctttaaaatatttaagccttttcacttaaaagtcaaattcaatt c . . tg . a . tttgtgtttgagacagtttgcttcttgtccaaccattcatacaagcctcc 150 200 250 300 350 ac.t.t. aattaagagactaatgactatgctaccttc 380 . g. nn. Figure 4 Partial sequence for the mitochondrial 16s rRNA gene. Sequences for Homarus americanus (Ha), H. gammarus (Hg), and Homarinus capensis (He) have been deposited with GenBank Accession Numbers U11238, U11246, and U11247 respectively. Dots indicate nucleotides identical with Ha; letters indicate nucle- otide substitutions at the homologous sites. Sites marked 'n' have unresolved nucleotides. ing form -inus, resembling. The gen- der is masculine. Homarinus capensis (Herbst, 1 792), new combination Synonymy— Holthuis (1986:243, fig. 1) gave an exhaustive synonymy for Homarus capensis, and a later (1991:59) less inclusive account. These treatments are so recent and readily available that reiteration here would be unnecessarily redun- dant. Succeeding reference to the species follows. Homarus capensis. — Kado, Kittaka, Hayakawa and Pollock, 1994:72, figs. 2, 3, 4. followed by row of intermittent noncrushing moder- ate conical teeth with 4-6 smaller ones in intervals between them; minor chela with latter pattern of noncrushing teeth on cutting edge of each finger; tips of fingers on each chela curved toward each other and crossing. Carpus of major chela elongate; anterior margin with two prominent spines and smaller ones between, palmar condyle subcircular and flattened, with sug- gestion of spines or tubercles on its anteromesial margin; dorsomesial margin strongly tuberculate and partly obscured by setae; shorter dorsolateral mar- gin also tuberculate but less prominently so; strong low spines on mesioventral margin. Merus bearing subdistal anterolateral spine, well-separated sharp tubercles on mesiodorsal margin, and mesioventral row of fairly uniform small tubercles. Minor chela with similar but less developed orna- mentation; merus with acute spines and spiniform tubercles. Etymology — The name Homarinus is derived from French homard, lobster, and the adjectival combin- Material — Cape Province, South Af- rica. USNM 251451. 16, East Lon- don?, R. Melville-Smith, 92-RMS-O, Nov 1992, regurg., dismembered, carapace length (cl) 26.5 mm, short carapace length (scl) 21 mm, abdo- men length (abdl) 33.0 mm. USNM 251452. 1 6 , southwest Dassen Island [33°26'S, 18°05'E], regurgitated from Sebastichthys capensis, badly crushed and partly dismembered, R.S. Steneck, 92-D-2, 1 Dec 1992, cl 32 mm, scl 25.5 mm. USNM 251453. 19, Still Bay [34°23'S, 21°27'E], dismembered, R. Melville- Smith, RMS7, abdl 45 mm. USNM 251454. 1 9 , Still Bay, regurg., R. Melville-Smith, RMS8, 5 mm, abdl 47 mm. Additional specimens reported to us by R. Melville- Smith, Sea Fisheries Institute, Cape Town: 1 6 , North Dassen Island, tide pool, RSS, 92-D-l, 3 Feb 1992; 19 . Port Alfred, RMS 1; 18, Houghham Park, Algoa Bay; 1 8 , Dassen Island, west side, RMS 3; 1 6 , Cape St. Francis, RMS 4; 18, Cintsa Reef, East London, RMS 5; 1 6 , Sunday's River mouth, RMS 6; 2 6 , Cape St. Francis, RMS 9 and 10; 1 9 , Haga Haga, Transkei coast, RMS 11. Description — As for genus with addition of the fol- lowing details. Abdominal pleura well developed, with rounded angles; pleuron of segment 1 small; pleuron of seg- ment 2 broad, overlapping first and third pleura; pleura 3^t-5 with antero ventral angle rounded, pos- terolateral angle subrectangular; pleuron of segment 6 rounded ventrally, posterolateral angle rounded and confluent with anterolateral angle of telson. Kornfield et al.: Cape lobster taxonomy 101 Telson with dorsal setae distributed in 3 longitu- dinal tracts, central and submarginal on either side; central tuft proximally in midline and another near each anterolateral corner; sparse similar setae on abdominal pleura; lateral ramus of uropod with ven- tral submarginal row of setae laterally. Eyes with distal edge of cornea slightly exceeding level of basicerite tip; this tip reaching to midlength of narrowly rounded antennal scale exceeded by its very strong anterolateral spine (rarely doubled) reaching distal edge of penultimate article in anten- nular peduncle; latter falling short of distal margin of terminal article in antennal peduncle. Epistome with median anterior spine closely flanked at either side by shorter rounded spine. Cheliped of pereopod 1 having fixed finger with narrowed extensor margin set off by shallow submar- ginal groove. Palm with compound row of low for- ward pointing spines and tubercles on flexor surface, similar development on extensor edge originating at carpal condyle and running along proximal margin of palm, across its basal end, and distally for a dis- tance along palm. Oviducal opening on coxa of pereopod 3 oval; its axes 1.3 x 1.8 mm on measured female noted below. Pleopod 1 with distal article broader than shaft and hollowed mesially, forming flattened tubular opening when appressed to opposite member, tip ir- regularly rounded. Pleopod 2 with appendix masculina on mesial aspect of endopod bearing tuft of strong setae at apex. Uropods with protopodite bearing 2 strong spines overhanging proximal end of mesial and lateral ra- mus respectively. Variation — There is minor variation in development of spines, tubercles, etc., among the two females and two males examined. According to Stebbing (1900), sides of the rostrum may have 5, 6, or 7 spines on the margin. Density of setae on exoskeletal parts is subject to considerable variation, owing perhaps to recency of molting, age, or abrasion after preservation. Color — Color of a living animal is shown in Figure 1. Published records summarized by Holthuis (1986) indicate that color may depart considerably from that shown here: coral-red to tawny or reddish yellow, which may have resulted from postmortem changes; or, in the fresh state, "of a rather dark olive colour, not dissimilar to that of the Northern lobster" Gilchrist (1918:45). Molecular characterization — Comparative analysis of a portion of the 16s ribosomal RNA gene from mitochondrial DNA (mtDNA) was conducted by using standard protocols (Kocher et al., 1989). Mito- chondrial DNA's purified by CsCl ultracentrifugation (Lansman et al., 1981) were amplified by PCR with the conserved primers 16sar and 16sbr of Palumbi et al. (1991). Following asymmetric amplification (Homarus americanus and H. gammarus) or cycle- sequencing (Homarinus capensis), DNA's were manu- ally sequenced by the dideoxy chain-termination method of Sanger et al. (1977). Aligned sequences are presented in Figure 4. Sequence divergence be- tween taxa was estimated by using the two-param- eter method of Kimura (1980). Sequence divergence between Homarus americanus and H. gammarus was 1.3%, whereas average divergence between these two species and Homarinus capensis was 9.7%. The 16s rRNA gene is one of the most slowly evolving regions of the mtDNA molecule (Xiong and Kocher, 1994); this conservative property makes it particularly use- ful for comparative studies among distantly related taxa. Though there is no formal recognition of equiva- lence between levels of sequence divergence and taxo- nomic rank (Hillis and Moritz, 1990), it is clear that the relative magnitude of divergence can be a useful taxonomic indicator (Avise, 1994). The magnitude of sequence differentiation that we observed between H. capensis and the two North Atlantic taxa strongly suggested the existence of two discrete clades. Mo- lecular divergence reinforced our conclusions from the reexamination of the morphology of these species. Remarks — Morphological differences between Homarinus capensis and the two species of Homarus are clear cut. Perhaps the most obvious differences are that Homarinus capensis has a dense coat of se- tae on the outer surface of the palms and on other articles of the chelipeds (PI), and scattered setae distributed over the carapace, tail fan, sixth abdomi- nal segment, and pleurae of the remaining abdomi- nal segments; Homarus americanus and H. gam- marus are smooth and glabrous. The telson of Ho- marinus has subparallel sides and its exposed sur- face bears many obsolescent transverse rugae (Fig. 3); the telson of Homarus species has sides converg- ing toward the tip, giving a subtriangular shape. First pleopods are more elongate and slender in Homarus species than in Homarinus (Fig. 2). The two species of Homarus attain large size (Wolff, 1978), whereas Homarinus capensis appears to be much smaller at maturity. No ovigerous females of H. capensis have been found, but openings of the oviducts are at least twice the size of those on com- parably sized specimens of the species of Homarus (see Kado et al., 1994). This suggests that there are fewer eggs with accelerated larval development in Homarinus capensis relative to slower larval devel- 102 Fishery Bulletin 93(1). 1995 opment from smaller more numerous eggs in Homarus species (Kado et al., 1994). Acknowledgments Our conclusions converged independently from two viewpoints. A. B. W. and other carcinologists have long understood grounds for generic separation of the Cape lobster from Homarus on the basis of morphol- ogy. I. K. and R. S. S. concluded this on the basis of genetic divergence and were well into their analysis before forces were joined. A. B. W. drafted the sys- tematic section and assembled the jointly produced text. Keiko Hiratsuka Moore rendered drawings of the pleopods. G. C. Steyskal provided advice on the choice of a new generic name. The manuscript was critically reviewed by W. Glanz, R. B. Manning, and T. A. Munroe. We are indebted especially to colleagues at the Sea Fisheries Institute, Cape Town, South Africa, who helped us in this study; Roy Melville- Smith provided materials and information, and Rob- ert Tarr photographed the living specimen of Cape lobster. George M. Branch, University of Cape Town, provided logistic support and aided in specimen ac- quisition. Yan Kit Tarn and Alex Parker provided sequence data. R. S. S. was supported by grants from the South African Foundation of Research Develop- ment, the Visiting Scholar Fund, and the Student Fund for Visiting Scholars of the University of Cape Town. Molecular work was supported by NOAA Sea Grant (NA90AAD-SG499) and NSF (EHR-9108766 and OCE-9203342). Literature cited Avise, J. C. 1994. Molecular markers, natural history and evolu- tion. Chapman and Hall, NY, 511 p. Barnard, K. H. 1950. Descriptive catalogue of South African decapod Crustacea. Ann. South African Mus. 38:1-837. Gilchrist, J. D. F. 1918. The Cape lobster and the Cape crawfish or spiny lobster. Mar. Biol. Rep. South Africa 4:44-53, 2 pis. Herbst, J. F. W. 1792. Versuch einer Naturgeschichte der Krabben und Krebs nebst einer systematischen Beschreibung ihrer verschiedenen Arten. Vol 2:i — viii, 1-225, pis. 22^46. Hillis, D. M., and C. Moritz. 1990. Molecular systematics. Sinauer Associates, Sunderland, MA, 588 p. Holthuis, L. B. 1986. J. C. Fabricius' (1798) species of Astacus, with an account of Homarus capensis (Herbst) and Eutrichocheles modestus (Herbst) (Decapoda Macrura). Crustaceana 50:243-256. 1991. FAO species catalog. Marine lobsters of the world. An annotated and illustrated catalog of species of interest to fisheries known to date. FAO Fisheries Synopsis 125, 13:viii, 1-292. Kado, R., J. Kittaka, Y. Hayakawa, and D. E. Pollock. 1994. Recent discoveries of the "rare" species Homarus capensis ( Herbst, 1 792 ) on the South African coast. Crus- taceana 67:71-75. Kensley, B. 1981. On the zoogeography of Southern African decapod Crustacea, with a distributional checklist of the species. Smithsonian Contrib. Zool. 338:i-iii, 1-64. Kimura, M. 1980. A simple method for estimating evolutionary rate of base substitution through comparative study of nucleotide sequences. J. Mol. Evol. 16:111-120. Kocher, T. D., W. K. Thomas, A. Meyer, S. V. Edwards, S. Paabo, F. X. Villablanca, and A C. Wilson. 1989. Dynamics of mitochondrial DNA evolution in mam- mals: amplification and sequencing with conserved primers. Proc. Nat. Acad. Sci. (USA) 86:6196-6200. Lansman, R. A, R. D. Shade, J. F. Shapiro, and J. C. Avise. 1981. The use of restriction endonucleases to measure DNA sequence relatedness in natural populations. Ill: Tech- niques and potential applications. J. Mol. Evol. 17: 214-226. Linnaeus, C. 1758. Systema naturae per regna tria naturae, secundum classes, ordines, genera, species, cum characteribus, differ- entiis, synonymis, locis, ed. 10, 1, iii + 824 p. Laurentii Salvii, Holmiae, Stockholm. Milne-Edwards, H. 1837. Histoire naturelle des Crustaces, comprenant 1'anatomie, la physiologie et la classification de ces animaux, Vol. 2, 532 p. Librairie Encyclopedique de Roret, Paris. 1851. Observations sur le squelette tegumentaire des Crustaces decapodes, et sur la morphologie de ces animaux. Ann. Sci. Nat., Paris (3, Zool.)16:221-291, pis. 8-11. Palumbi, S., A Martin, S. Romano, W. O. McMillan, L. Stice, and G. Grabowski. 1991. The simple fool's guide to PCR, v. 2.0. Special pub- lication of the Univ. of Hawaii Dep. Zoology and Kewalo Marine Laboratory, 23 p. Sanger, F., S. Nicklen, and A. R. Coulson. 1977. DNA sequencing with chain-terminating inhib- itors. Proc. Nat. Acad. Sci. (USA) 74:5463-5467. Stebbing, T. R. R. 1900. South African Crustacea. Mar. Investig. South Af- rica 1:14-66, pis. 1-4. Weber, F. 1 795. Nomenclator entomologicus secundum entomologiam systematicum ill. Fabricii, adjectis speciebus recens detectis et varietatibus.... C. E. Bohn, Chilonii [Kiel] et Hamburgi, viii + 171 p. Wolff, T. 1978. Maximum size of lobsters (.Homarus) (Decapoda, Nephropidae). Crustaceana 34:1-14, pis. 1-2. Xiong, B., and T. D. Kocher. 1994. Phylogeny of the sibling species of Simulium venustum and S. verecurdum (Diptera: Simuliidae) based on sequences of the mitochondrial 16s rRNA gene. Mol. Phyl. Evol. 3:293-303. Abstract. — A radiometric age- ing method was used to resolve con- flicting results from ageing tropi- cal lutjanids based on annual ring counts in whole and sectioned otoliths. The number of rings de- tected in sectioned otoliths of Lut- janus erythropterus, L. malabar- icus, and L. sebae from unexploited populations in the Gulf of Carpentaria, Australia, were 1.6 to 2.4 times the number found in whole otoliths. To obtain an inde- pendent estimate of age, we mea- sured 210 Pb/ 226 Ra radioactive dise- quilibria of both whole and cored otoliths. As all species had high lev- els of 226 Ra, they could be aged with relative accuracy by this method. Samples of whole otoliths and cores with a similar ring count had simi- lar radiometric ages. In samples whose sectioned and whole-otolith ages differed by more than 4 years, the whole otolith ring count agreed better with the radiometric age (for an uptake activity ratio i?=0.0). This result stands in marked con- trast to the radiometric age valida- tion of section counts for slow-grow- ing, long-lived fish inhabiting tem- perate to subtemperate waters. In this region, all species lived less than 10 years and grew to a maxi- mum size of up to 600 mm SL. They reached a similar length in one year, but L. erythropterus grew faster than the other two species thereafter. The sexes had the same growth rates. Our results were similar to those found for these spe- cies elsewhere and suggest that in tropical fishes, such as lutjanids, rings observed in sectioned otoliths and other hard parts may not be formed annually. Where possible, ages derived from counts in these structures should be verified by independent methods. Ageing of three species of tropical snapper (Lutjanidae) from the Gulf of Carpentaria, Australia, using radiometry and otolith ring counts David A. Milton CSIRO Division of Fisheries Marine Laboratories. RO. Box 1 20 Cleveland. Queensland 4 1 63, Australia Steven A. Short Environmental Radiochemistry Laboratory ANSTO. Private Mail Bag 1 Menai. NSW. 2234. Australia Present address: Kmgett Mitchell and Assoc. RO. Box 33-849. Auckland. New Zealand Michael F. O'Neill Stephen J. M. Blaber CSIRO Division of Fisheries Marine Laboratories. RO. Box 1 20 Cleveland. Queensland 4 1 63. Australia Manuscript accepted 18 August 1994. Fishery Bulletin 93:103-115 (1995). Tropical fishes can be difficult to age because many species do not deposit annual rings in their hard parts (Longhurst and Pauly, 1987). Lutjanids, which are highly valued commercial fishes in the tropical Indo-Pacific region, often have ring patterns in their hard parts that are difficult to interprete (e.g. Davis and West, 1992). The age and growth of many lutjanid species have been well studied and the results of these studies have formed the basis of age-structured stock assessments upon which the management of these fisheries is based (e.g. Sains- bury, 1988). In the western Pacific, Lutjanus malabaricus has been the most widely studied lutjanid, as it is the main catch of trawl and line fisher- ies in northern Australia, adjacent Indonesian waters, and in the South China Sea. The reported maximum age (up to 10 yr ) and growth param- eters differ both between regions (Lai and Lui, 1974, 1979) and within one area (northern Austra- lia: Lai and Lui, 1979; Chen et al., 1984; Edwards, 1985; McPherson and Squire, 1992). These studies estimated age from growth rings in vertebrae (Lai and Lui, 1979; Chen et al., 1984; Edwards, 1985) or in whole otoliths (McPherson and Squire, 1992). The latter method may underestimate the age of longer-lived species because of the difficulty of distinguishing all the growth rings (Casselman, 1974). The timing of formation of annual growth rings in Lutjanus from northern Australia has not been fully verified. Several authors (Lai and Lui, 1974, 1979; Chen et al., 1984; Yeh et al., 1986; Davis and West, 1992) have concluded that the outer ring is probably deposited 103 104 Fishery Bulletin 93(1), 1995 annually, but the season when this ring is formed varies between studies and species. Such ambigu- ities cast doubt on the validity of the conclusions and suggest that differences in the estimated growth rate and age of Lutjanus populations may be related to problems of interpretation rather than to biological differences. A radiometric method has recently been used suc- cessfully to estimate the age of long-lived fishes (Bennett et al., 1982; Fenton et al., 1991). This method uses the known decay rates of isotopes of Radium-226 ( 226 Ra) and Lead-210 ( 210 Pb) in bony parts to estimate the age of fish. It does not rely on operator interpretation to estimate age and there- fore is particularly useful for ageing long-lived spe- cies where the growth rings are often not clearly de- fined (Casselman, 1974). The objectives of the present study were 1) to esti- mate the age and growth of Lutjanus malabaricus, L. erythropterus, and L. sebae from the Gulf of Carpentaria by counting rings in whole and sectioned otoliths; and 2) to use the 210 Pb/ 226 Ra radiometric ageing method to make an independent age estima- tion of the same fish. S ' U «m T AUSTRALIA > • • 10" • • J • • • • • / ~~ef,"E Figure 1 Map of the Gulf of Carpentaria showing the distribution and relative abundance of Lutjanus erythropterus, L. malabaricus, and L. sebae during a systematic survey in November 1990. Materials and methods Sampling Most samples of Lutjanus erythropterus, L. mala- baricus, and L. sebae were collected during a sys- tematic survey of the Gulf of Carpentaria between long. 136° and 142°E in November 1990. Two similar random-sampling surveys were made across the northern Gulf of Carpentaria (north of lat. 14°30'S) in November 1991 and January 1993. Samples of Lutjanus malabaricus were also collected during a survey of eight areas in the Gulf of Carpen-taria by the commercial trawler Clipper Bird in June 1990 (Fig. 1). Details of survey design, trawl gears, and trawl durations are given in Blaber et al.(1994). Commercial-sized Lutjanus malabaricus (1-3 kg) were obtained from fish retained for sale after the June 1990 survey. During the systematic survey in November 1990, all specimens of the three target species of lutjanids were retained for ageing studies. In November 1991 and January 1993, only fish from length classes underrepresented in previous samples were processed. All fish were measured (standard length [SL] in mm), weighed (±1 g), and sexed, and both sagittae were removed, dried, and stored in la- belled bags for future analysis. Radiometry Radioanalysis requires about 1 g of sample material; therefore, fish were pooled to obtain the necessary sample weight. For juveniles, up to four otoliths were required to obtain this weight. Otoliths used for radioanalysis were chosen in two ways. First, for each species, otoliths from juvenile, maturing, and ma- ture fish that had the same sectioned-otolith ages, similar otolith weights, and similar sizes, and came from the same region of the Gulf of Carpentaria were pooled for radioanalysis. Second, otoliths of the same whole-otolith age and from fish of similar size were abraded with a mechanical sander to a central core approximating the weight (see Table 2), length, and shape of the otolith of a fish whose whole-otolith age was 3 (otolith length=11.6 ± 0.7 for L. erythropterus; 12.7 ± 0.4 for L. malabaricus; and 11.4 ± 0.4 for L. sebae). The exception was sample 2490 for which otoliths were ground to a core age of 2 (weight 0.156 ± 0.005 g; otolith length 9.2 ± 0.16 mm; n=92). The otolith nucleus at the center of the cores was located by examining intact otolith morphology and by sec- tioning other samples (Campana et al., 1990). This age is less than the age at sexual maturity for all species. Milton et al.: Ageing of Lutjanus erythropterus. L malabancus. and L sebae 105 The method of radioanalysis of the otoliths is de- tailed in Fenton et al. (1990, 1991). It involves mea- suring the specific activity of 226 Ra: 210 Pb by alpha- spectrometry. Because of the extremely small spe- cific activities measured (0.01-0.1 dprn-g" 1 for Polo- nium-210 [ 210 Po]), cleanliness is of the utmost im- portance in the analytical procedure. Every item of laboratory ware that contacted the otolith solutions and otoliths was chemically decontaminated in al- kaline 0.05M Na 4 EDTA(pH 10.5). The otoliths were washed and rinsed several times in this solution, then washed several times in 0.1M HC1 (<10 s) and fi- nally washed twice in water. Our analyses of 210 Pb, via its short-lived daugh- ter-proxy 210 Po, and 226 Ra were made with high-reso- lution alpha-spectrometers according to the meth- ods of Fenton et al. (1990). The mean 210 Po reagent blank was 0.0071 ± 0.0012 dpm. Recovery of 210 Po was always at least 90% and instrument background counts (for 208 Po and 210 Po) were less than one countd -1 . 226 Ra was analyzed by a direct alpha-spec- trometry method and chemical yield was measured by gamma spectrometry of a Barium-133 ( 133 Ba) tracer (Fenton et al., 1990, 1991). Mean activity of the 226 Ra blanks was 0.0174 ± 0.0026 dpm, which was lower than in previous studies (e.g. Bennett et al., 1982; Fenton et al., 1991) owing to careful con- trol of reagents. Recovery of 226 Ra (as estimated by the recovery of 133 Ba tracer) was greater than 85% for all samples. (l-e-' --). XT\\ l-(l-R) (l-e-) XT \ -X(t-T) (2) where all parameters are the same as in the previ- ous model, except T, which is the estimated age of the otolith core. A linear mass growth model was assumed only up to the age of the core. The initial uptake 210 Pb/ 226 Ra activity ratio was generally as- sumed to be i?=0.0. This is the most conservative value, and so radiometric age estimates derived with this value must overestimate the maximum possible age of the sample. The above equations were solved numerically by a Newton-Raphson iteration method (Fenton et al., 1991). Stable element analysis The levels of lead and barium in otoliths are pre- sumed to act as stable equivalents of 210 Pb and 226 Ra and so can be used to assess the uptake of the radio- active isotopes and to normalize the radiometric data (Fenton and Short, 1992). Therefore, the concentra- tions of stable lead, barium, strontium (Sr), and cal- cium (Ca) in each otolith sample were measured for an aliquot of each dissolved otolith solution used in the radiometric analysis. Each solution was analyzed by inductively coupled plasma mass spectrometry for lead and barium and by inductively coupled plasma atomic emission spectrometry for strontium and calcium. Data analysis The ages of whole otoliths were calculated on the basis of a single constant (linear) growth rate by the equation originally derived by Bennett et al. (1982): A=l-(1-R) 1 It (1) where A=the ratio of the activity of 210 Pb to 226 Ra activity at time t ( 210 Pb/ 226 Ra) t ; i?=ratio of 210 Pb to 226 Ra at the time of deposition [( 210 Pb/ 226 Ra) ]; and >.=decay constant for 210 Pb (0.03114 yr _1 ). Assump- tion of a single linear mass growth rate produces radiometric ages that are greater than those that would result from assumption of an exponential (non- linear) rate, the bias always favoring a higher value (Campana et al., 1993). It should be understood that using this assumption (linear mass growth of the otolith) will produce age estimates that always over- estimate the real age. For otolith cores, ages were calculated from Smith et al.'s (1991) equation: Otolith ageing Pairs of otoliths from each fish were cleaned of ex- cess tissue, dried at 60°C for 24 h, weighed (± 0.1 mg) and measured along the longitudinal axis with dial calipers (± 0.05 mm). One otolith of each pair was embedded in polyester resin and cross-sectioned with a diamond saw (Augustine and Kenchington, 1987). Thin sections (approximately 200 |im) of each otolith were bonded to microscope slides with thermo- plastic cement. Each section was polished on both faces with 800-grit wet-and-dry carborundum paper before being examined with a video-enhanced light microscope attached to a microcomputer with pre- cise distance-measuring software. The rings (pre- sumed annuli) were counted and the distance be- tween them measured along the dorsal axis adjacent to the sulcus, where they were most clearly distin- guishable. In whole otoliths, the rings were counted against a strong background point light source. Counts of rings in all whole and sectioned otoliths were made independently by two readers. When the ring counts differed, the otoliths were reexamined by both readers. If the counts still differed by more 106 Fishery Bulletin 93(1), 1995 than one, the data were discarded (<5% of all otoliths). If counts differed by one, the higher value was chosen (10% of otoliths). The relative frequency of these discrepancies was similar for all species. Data analysis The length-at-age data were fitted to the repara- meterized von Bertalanffy growth curve of Francis (1988). This method has the advantage that the pa- rameters estimated are independent and can be com- pared directly between species and populations. Most previous studies of lutjanid age and growth have fitted the von Bertalanffy growth equation to data on length at age (e.g. Lai and Lui, 1979; Manooch, 1987; Davis and West, 1992). However, the estimated parameters L m , K, and t Q either do not have direct biological meaning (e.g. Knight, 1968; Schnute and Fournier, 1980; Ratkowsky, 1986) or are extrapo- lations from the data (Ratkowsky, 1986). Francis (1988) extended the equation of Schnute and Fournier (1980) to derive a new set of parameters L v L 2 , and L 3 (his L, I , and l w ), which correspond to the length at the lower, middle, and upper limits of any arbitrarily defined age range, such that: L t =L 1+ (L 3 -L 1 )(l-r 2{t ^ /u "^ ) )/a-r 2 ), (3) where r = (L 3 —L 2 )/(L 2 —L 1 ); L t is the mean length of a fish at age t; and L v L 2 and L 3 are the length at the lower, middle, and upper limits of two arbitrary ages and w. By fitting a curve of this form, extrapola- tions beyond the data are avoided, as the three fit- ted parameters are chosen from within the range of the data and hence can be directly compared with the results of previous studies. In this study, we set = 1 ring and w = 6 rings for each species. This equation has the advantage that the age range to be examined can be chosen by the investigator, rather than having to be the largest and smallest age classes found, as required by the Schnute and Fournier ( 1980) equation. These param- eters (Lj, L 2 , and L 3 ) can also be expected to have similar properties to those of Schnute and Fournier (1980) and not to show the high negative correlation between L x and K (Francis, 1988). All parameters were estimated by an iterative least-squares method (SAS NLIN procedure with the Marquardt option; SAS, 1989). Vaughan and Kanciruk (1982) found that this procedure consis- tently showed the least bias in parameter estimates, converged rapidly, and provided more precise esti- mates than did standard linear techniques. A mea- sure of goodness-of-fit was obtained by calculating an r 2 value from the residual and the explained sums of squares derived from the least-squares regression. Relationship of ring counts in whole and sectioned otoliths with radiometric ages The estimated age from ring counts in whole and sectioned otoliths used in the radiometry were com- pared for all species by two methods. First, the rela- tionship between radiometric age and whole and sec- tioned otolith ages of the same fish were plotted. If the slope of the relationship was not significantly different from 1, the results of the two methods were considered to be in close agreement. Second, the two ageing methods were compared with the radiomet- ric ages with a Wilcoxon matched-pairs ranks test (Conover, 1980). The two hypotheses tested were 1) that whole otolith ring counts underestimated true age (radiometric age) or 2) that sectioned otolith ring counts overestimated true age. Results Radiometry Lutjanus erythropterus— The specific activity of 226 Ra and the 210 Pb/ 226 Ra activity ratio differed among the three samples of L. erythropterus (Tables 1 and 2). The activity ratio was highest in the cored sample (0.118 ± 0.031; Table 2). Ring counts in whole otoliths were linearly related to otolith mass (Fig. 2A) and ring count in sectioned otoliths, though the relationship was significantly weaker CP<0.05). Radiometric age estimates were calculated on the basis of a single constant (linear) growth rate for otolith mass, which removes the need to include the mass growth rela- tion in the radiometric age calculation (Eq. 1). Un- der the assumption of a constant growth model, ra- diometric age estimates were most similar to those obtained from the ring counts in whole otoliths (Table 2). The match was best for the cored otolith sample where model assumptions are less stringent (sample 2673). Lutjanus malabaricus — The specific activity of 226 Ra in L. malabaricus otoliths differed among samples and among size classes (Tables 1 and 2). The 210 Pb/ 226 Ra activity ratios ranged from less than 0.027 to 0.212 and varied to a similar degree in cored and whole-otolith samples (Table 2). Otolith weight was linearly related to the number of rings in whole otoliths (Fig. 2B), and this relationship was stron- ger than that for counts from sectioned otoliths Milton etal.: Ageing of Lutjanus erythropterus, L malabaricus, and L sebae 107 Table 1 Elemental composition of otoliths of three species of Lutjanus used in the radiometric analysis. Whole = whole otoliths used; cored = otoliths cored to age 3+ (L. erythropterus ) or 2+(L. malabaricus and L. sebae ). Numbers in parentheses represent repeated analyses of a sa mple in which several otoliths of similar whole and sectioned age had been combined. 226 Radium 210 Lead Lead Barium Pb/Ba Strontium Calcium Sr/Ca Whole/or dpm-g -1 dpm-g" 1 (Pb) (Ba) mass (Sr) (Ca) mass Species Sample cored (± la) (± la) (ppm) (ppm) ratio (ppm) (ppm) ratio L. erythropterus 2065 Whole 0.2277 ± 0.0132 0.0174 ±0.0047 0.08 18.7 0.004 2,800 395,000 0.0071 2066 Whole 0.1623 ± 0.0087 0.0070 ± 0.0035 0.19 11.5 0.016 2,720 398,000 0.0068 2673 Cored 0.1331 ± 0.0087 0.0157 ± 0.0040 0.66 9.2 0.071 — — — L. malabaricus 2062(2) Whole 0.2390 ±0.0111 -0.0025 ± 0.0045 0.27 8.2 0.033 2,915 462,000 0.0063 2063 Whole 0.0728 ± 0.0066 0.0067 ± 0.0042 3.49 6.5 0.537 3,330 470,000 0.0071 2063(2) Whole 0.0582 ± 0.0049 -0.0010 ± 0.0025 0.22 13.4 0.016 1,990 321,000 0.0062 2063(3) Whole 0.0916 ± 0.0053 0.0072 ±0.0017 0.55 4.8 0.114 2,240 399,000 0.0056 2064 Whole 0.2118 ±0.0164 0.0173 ± 0.0024 2.28 5.8 0.390 3,000 395,000 0.0076 2438 Whole 0.1014 + 0.0064 0.0068 ± 0.0036 0.41 5.3 0.080 2,005 426,000 0.0047 2439 Whole 0.2942 ± 0.0141 0.0133 ± 0.0043 0.06 7.5 0.008 2,250 415,000 0.0054 2440 Whole 0.1080 ±0.0068 0.0135 ± 0.0029 0.18 6.2 0.029 2,910 438,000 0.0067 2489 Cored 0.1678 ± 0.0088 0.0356 ± 0.0055 <0.09 6.7 <0.014 2,160 407,000 0.0053 2490 Cored 0.1219 ±0.0078 0.0180 ± 0.0037 <0.09 6.2 <0.014 2,040 416,000 0.0049 L. sebae 2068 Whole 0.1036 ± 0.0064 0.0139 ± 0.0034 3.13 8.7 0.360 2,360 396,000 0.0060 2069 Whole 0.1046 ±0.0058 0.0312 ± 0.0037 1.38 8.1 0.170 2,510 398,000 0.0063 2070 Whole 0.0460 ± 0.0042 0.0100 ±0.0023 0.46 5.0 0.092 — — — 2647 Cored 0.2143 ±0.0114 0.0373 ± 0.0049 0.50 11.5 0.043 — — — 2648 Cored 0.1756 ± 0.0099 0.0458 ± 0.0052 0.66 9.4 0.071 — — — Table 2 Results of radiometric and direct ageing otoliths of Lutjanus malabaricus, L. erythropterus and L sebae from the Gulf of Carpentaria. Radiometric ages were calculated by using a constant growth rate model and by using R = 0.0 (where R = initial 210 Pb: 226 Ra activity ratio at time of deposition). All errors n radiometric age estimates expressed at la level (n = number of otoliths in sample). SE = Standard error. Numbers in parentheses represent repeated analyses of a sample in which several otoliths of similar whole and sectioned age iad been combined. Mean length Mean otolith Whole Sectioned 210p b .226 Ra Radiometric Species Sample n (mm)± SE mass (g) ± SE otolith age otolith age activity ratio age L. erythropterus 2065 3 316 ±2 0.3315 ± 0.0186 3 3 0.076 ±0.021 5.1 ± 1.5 2066 2 364 ±- 0.3875 ± - 3.3 6 0.043 ± 0.022 2.8 ± 1.5 2673 3 368 ±7 0.4750 ± 0.0507 4 9 0.118 ±0.031 5.5 ±1.1 L. malabaricus 2062(2) 2 310 ± - 0.4852 ± - 3 3 <0.027 (95% CD <1.8 (95% CD 2063 2 350 ± - 0.5775 ± - 4 6 0.092 ± 0.058 5.7 +4.3,-4.0 2063(2) 2 348 ±- 0.6170 ±- 4 6 <0.069 (95% CD <4.6 (95% CD 2063(3) 3 346 ±7 0.6118 ±0.0133 4 6 0.079 ± 0.019 0.8 ±0.8 2064 3 443 ± 7 1.5591 ± 0.0493 6.7 14 0.082 ±0.013 5.6 ± 0.9 2438 4 250 ±2 0.2517 ± 0.0063 3 3.5 0.067 ± 0.036 4.5 +2.6,-2.5 2439 1 650 2.1155 9 13 0.045 ± 0.015 3.0 ± 1.0 2440 1 560 2.1111 7 19 0.125 ±0.028 8.8 + 2.2,-2.1 2489 4 455 ± 10 1.6185 ± 0.1349 9 13 0.212 ±0.035 8.7+1.5,-1.4 2490 5 422 ±7 1.1055 ± 0.0501 8 8 0.148 ±0.032 6.1 ± 1.2 L. sebae 2068 4 222 ± 5 0.2429 ± 0.0261 2 3 0.134 ±0.034 9.5 + 2.7,-2.6 2069 2 304 ±- 0.6152 ±- 3.5 7 0.298 ± 0.039 24.2 +4.2,-3.9 2070 2 399 ±- 1.5658 ± - 5.5 15 0.217 ± 0.054 16.4+5.1,-4.6 2647 3 400 ± 15 1.4462 ± 0.0184 5 12.3 0.174 ±0-023 7.1 ± 1.0 2648 2 462 ±- 2.1000 ±- 7 15.5 0.261 ± 0.033 11.2+1.5,-1.4 108 Fishery Bulletin 93(1). 1995 "l-5-i l erythropterus 6 8 10 Number of rings Figure 2 The relationship between otolith weight and the number of rings counted in whole otoliths of (A) L. erythropterus, (B) L. malabaricus, and (C)L. sebae. (P<0.05). Under the assumption of a constant mass growth model, radiometric age estimates were again most similar to those found for whole otolith ring counts. The match was best for samples of cored otoliths (2489, 2490) where assumption of a mass growth model is almost absent (Table 2). Lutjanus sebae — The specific activity of 226 Ra and the 210 Pb/ 226 Ra activity ratio varied less between samples in L. sebae than in the other species (Tables 1 and 2). As with the other species, otolith weight was linearly related to the ring counts of whole otoliths; therefore, a single constant growth rate was assumed in interpre- tation of the radiometric data (Fig. 2C). The radiomet- ric age estimates of intact otolith samples of juveniles (2068, 2069, 2070), based on the assumption of no allogenic 210 Pb uptake in the otoliths (R=0.0), were higher than the ring counts for both sectioned and whole otoliths (Table 2). Samples 2068 and 2069 were prob- ably subject to high rates of allogenic 210 Pb uptake, as indicated by the high stable Pb/Ba mass ratios (Table 1). Radiometric ages of both sets of cored otoliths were most similar to the age estimates based on whole otolith counts. However, both these samples (2647 and 2648) had very low stable Pb/Ba mass ratios (Table 1). Mod- elled radiometric ages of L. sebae samples (both whole and cored) for different values of R indicate that R - 0. 10 best matches the ring count of whole otoliths (Table 3). Lead.'Barium ratios The stable lead:barium ratios of all samples were plotted against radiometric age assuming an initial activity ratio/? = 0.0 (Fig. 3). Neither L. malabaricus nor L. erythropterus showed an increase in the ratio with increasing age. However, in four of the five L. sebae samples radiometric age increased rapidly with increasing stable lead (Fig. 3). Otolith ageing Lutjanus erythropterus — The growth curves of L. erythropterus based on ring counts in whole otoliths 30- A L sebae • L. malabaricus n L. erythropterus >: 20- 0.3) (Table 4) and lived to a similar age. Lutjanus malabancus — The growth curves express- ing the best fit of length-at-age data from both sec- tioned otoliths and whole otoliths show significant differences (P<0.05) in the estimated growth rates (Fig. 4B). More rings were counted in sectioned otoliths than in whole otoliths from the same fish (Fig. 6B), but were linearly related (whole otolith count=0.64 x (sectioned otolith count) + 0.79; r 2 =0.81, 1,869 =3614.7, P<0.001). Growth parameters of the reparameterized von Bertalanffy equation of male and female L. mala- baricus did not differ except for L 3 ; this parameter was larger in males (P<0.05). Not all fish collected were sexed, but the growth parameters of the com- bined equation differed from that obtained from the subsets that were sexed (Table 4). Lutjanus sebae — Ages based on counts of sectioned and whole otoliths differed significantly in L. sebae over 350 mm SL (P<0.05; Fig. 4C). More rings were detected in the otoliths of these fish when they were sectioned than when examined intact, although the number of rings detected by the two methods were linearly related (Fig. 6C; whole otolith count=0.50 x (sectioned otolith count) + 0.19; r 2 =0.80, F 1 140 =546.0, P<0.0001). The growth parameters of the reparameterized von Bertalanffy equation were similar for both sexes (Table 4). Lutjanus sebae were larger at one year (L x ) Table 4 Growth parameters (SL ± SE) of the reparameterized von Bertalanffy growth equations for Lutjanus malabaricus, L. erythropterus, and L. sebae ( 1-6 rings) from the Gulf of Carpentaria (r 2 = nonlinear i estimate of goodness-of-fit). Species Sex n LjiSE L 2 ±SE L 3 ±SE r 2 L. erythropterus both 172 75.40 ± 11.67 335.21 ± 2.73 457.12 ± 10.01 0.93 females 61 75.03 ± 17.57 346.66 ± 5.07 477.29 ± 15.64 0.95 males 30 86.06 ± 22.0 337.47 ± 4.50 468.66 ± 18.55 0.94 L. malabaricus both 878 78.09 ± 2.99 298.94 + 1.72 424.92 ± 1.37 0.95 females 159 195.37 ± 27.37 329.09 ± 4.31 423.19 ±2.70 0.70 males 73 100.63 ± 15.50 313.62 ± 6.91 442.67 ± 4.54 0.92 L. sebae both 144 122.27 ± 3.22 287.28 ± 2.18 451.501 3.27 0.97 females 14 99.92 ± 17.84 277.43 ± 8.58 443.96 ± 7.54 0.99 males 9 113.50 ± 3.98 284.83 ± 5.30 461.10 ±4.41 0.99 Fishery Bulletin 93(1), 1995 A bUU- L erythropterus 400- \Tu 300- 200- D /W [ ] \ - whole otolith 100- 0- { - sectioned otolith Figure 4 Plot of the mean length-at-age (±) range and the growth curves of the three species oiLutjanus based on ring counts in whole and sectioned otoliths. than were other species (P<0.05). However, at six years of age (L 3 ) they were about the same size as L. erythropterus but were larger than L. malabaricus (P<0.05). Relationship of ring counts in whole and sectioned otoliths with radiometric ages There was a significant linear relationship between both whole and sectioned otolith ring counts and ra- diometric age (Fig. 7; P<0.001 in both cases). The slopes of the lines of best fit differed (/}=1.04 ± 0.11; r 2 =0.84 for whole otolith ring counts and /?=1.83 ± 0.06; r 2 =0.87 for sectioned otolith ring counts). Be- cause the initial activity ratios of the L. sebae samples (R) were obviously greater than 0.0 in at least the whole otolith samples, these were not included in the analyses. There was no significant difference between whole otolith ring counts and radiometric ages for all spe- cies combined (T=53.5; P>0.30, re=15) or for L. malabaricus (T=17.5; P>0.15, rc=10). However, for all species combined we found that the sectioned ring counts were significantly greater than the radiomet- ric age of the same fish (T=6.5; P<0.001, ra=15). The sectioned ring counts of L. malabaricus were also greater than the radiometric ages (T=2; P<0.005, n=10). Discussion This is the first study to use 210 Pb/ 226 Ra activity ra- tios to verify the age of relatively short-lived tropi- cal fishes. Previous studies that have used these ra- tios to estimate age have focussed on species that live to at least 70 years (Bennett et al., 1982; Campana et al., 1990; Fenton et al., 1991). In the Lutjanidae, natural levels of 226 Ra in the otoliths were high, which helped to minimize the variances in the 210 Pb/" 226 Ra activity ratio and hence the errors in the age estimates. Radiometry provided strong evidence that the rings counted in whole otoliths were the best estimate of the true age of the three lutjanids studied. The radiometric methods we used tend to overes- timate age because the assumptions concerning the otolith mass growth model and rate of incorporation of allogenic 210 Pb were conservative. The only con- ceivable mechanism that would lead to underesti- mation of ages radiometrically would be a signifi- cant loss of radon ( 222 Rn) from otoliths during growth (West and Gauldie, in press). Radon is the daughter of 226 Ra and the only gas- eous precursor of 210 Pb in the decay chain. Its mean lifetime is only 4.8 x 10 5 seconds, and its effective (physical) diffusivity in otoliths would be about 0.5 x 10" 12 m 2 -s _1 . Radon diffusion out of otoliths would be further retarded by adsorption to organic matter (Wong et al., 1992). Simple calculations based on the known microstructure of otoliths (Campana and Neilson, 1985) and on the existing data on radon emanation (Morawska and Phillips, 1993) show that significant loss of radon from otoliths is extremely unlikely, as previously suggested from empirical stud- ies (Fenton and Short, 1992). Milton et al.: Ageing of Lutjanus erythropterus, L maiabancus. and L. sebae 1 1 Figure 5 Photographs of an otolith of a 270-mm L. erythropterus showing the discrepancy between the number of rings seen by examining (A) the intact otolith (2 rings) and (B) after sectioning (5 rings). Scale=l cm. Fishery Bulletin 93(1), 1995 Why ages derived from whole and sectioned otoliths were significantly different remains unclear. The differences in ring counts increased with the size of the fish, and the slope of the regression (whole vs. sectioned ring count) was steepest for the fastest- growing species, L. erythropterus. The otoliths were large (up to 30 mm long, and weighing 3 g), so the daily rings during periods of reduced or variable growth of younger fish were still relatively widely spaced. Thus, what would appear as a diffuse, single hyaline zone in a whole otolith examined against reflected light may have appeared in section as a group of hyaline and opaque zones. These problems 14 - A L. erythropterus 12- o 10- ooo o o 8- o o o o 6- ooo ooo 4- oooo ooo 2- ooo o o - * i ' i • i i 2 4 6 8 Sectioned otolith ring count -* -» ro 3 Ol O Ol o 03 IOOO P- IOOOOOO 0) ooooooo ST ooooooo g oooooooo ooooooooo o ooooooooo oo ooooooo o o oo o ooo o 2 4 6 8 10 20- c L. sebae 15- 10- 5- o- lOOOOO oooooo ooooooo oooooo ooooo oo oooo oo oo o 1 1 1 1 . 1 I 1 1 2 4 6 8 10 Whole otolith ring count Figure 6 Plot of the relationship between whole and sec- tioned otolith ring counts of the three species of Lutjanus. in otolith interpretation were most marked in L. erythropterus and led to the greatest discrepancy in ring counts. Studies of the age and growth of Lutjanus mala- baricus from northern Australia and the South China Sea used ring counts in vertebrae (Lai and Liu, 1974, 1979; Edwards, 1985), sectioned otoliths (Chen et al., 1984), and whole otoliths (McPherson and Squire, 1992). Their estimates were similar to the estimates we obtained from whole-otolith ageing, although L. malabaricus from the Great Barrier Reef appear to grow much faster and live at least one year less than those found in other areas (McPherson and Squire, 1992). However, the previous studies and our study provide different estimates of the von Bertalanffy growth parameters L x and K (Table 5). These differ- ences may have major impacts on age-structured fish- ery models (e.g. yield per recruit) that use these pa- rameters to estimate optimal yield. Lutjanus erythropterus and L. sebae from the Gulf of Carpentaria grew at similar rates to those reported from other parts of northern Australia (Ju et al., 1988; McPherson and Squire, 1992) and elsewhere within their range (Druzhinin and Filatova 1980; Yeh et al., 1986; McPherson and Squire, 1992). However, the growth of L. sebae in the Gulf of Carpentaria did not decline as they approached the maximum age ob- served. This may have been caused by an error in the ring count in otoliths of older fish or because the older age classes were not caught in the trawls. The maximum size of L. sebae in Australian waters has been reported to be between 1.0 and 1.4 m (Allen, 12- 10- Slope = 1.04 ±0.11 - y S r 2 = 0.84 / c 8 - D O ° fi- o> 6 c ir 4" o / A / □ • X • A* / • L malabancus whole • S ± • A L erythropterus whole 2" / O L malabancus cores ./ Q L sebae cores / A/- erythropterus core 0" ( ) 2 4 6 8 10 12 Radiometric age (yr) Figure 7 The relationship between radiometric age (yr) and whole otolith ring counts of all samples (except L. sebae whole otolith samples). Milton etal.: Ageing of Lutjanus erythropterus. L. malabancus. and L sebae 13 Table 5 Von Bertalanffy growth parameters of tropical Lutjanus from northern Australia and elsewhere within their range (W=whole otoliths; S=sectioned otoliths; V=vertebrae; U=urohyal; Sc =scales). Species Locality Sex Method K L x Maximum age Reference L. erythropterus Gulf of Carpentaria Both W 0.30 565 6 Present study Great Barrier Reef F w 0.44 500 7 McPherson and Squire (1992) M w 0.41 500 7 McPherson and Squire (1992) Northwest Shelf Both V 0.21 603 7 Juetal. (1988) L. malabaricus Arafura Sea Both V 0.17 707 10 Edwards (1985) Both V 0.12 790 8 Lai and Lui (1979) Gulf of Carpentaria Both w 0.22 592 9 Present study N.W. Australia Both V 0.13 768 8 Lai and Lui (1979) Both V 0.25 715 10 Chen etal. (1984) S. China Sea Both V 0.14 790 11 Lai and Lui (1974) Great Barrier Reef F w 0.23 696 7 McPherson and Squire ( 1992) M w 0.18 820 7 McPherson and Squire ( 1992) Vanuatu Both s 0.31 600 — Brouard and Grandperrin (1984) L. sebae Gulf of Aden Both Sc 0.16 660 11 Druzhinin and Filatova (1980) Gulf of Carpentaria Both w 0.06 1483 9 Present study N.W. Australia Both V 0.13 678 10 Yeh etal. (1986) Great Barrier Reef F w 0.18 851 8 McPherson and Squire (1992) M w 0.15 736 8 McPherson and Squire (1992) L. vittus N.W. Australia F u 0.37 267 7 Davis and West (1992) M u 0.22 346 8 Davis and West (1992) 1985; Grant, 1985) or 16 to 22 kg (Grant, 1985; Allen and Swainston, 1988), which is much greater than we recorded (5 kg). This species may, therefore, live more than 10 years. Indeed, large L. sebae (over 800 mm) from the Great Barrier Reef are known to live on deep coral reefs at depths greater than 60 m 1 ; the deepest part of the Gulf of Carpentaria is only 55 m. This sug- gests that fish may move from this region as they grow. Our radiometric ageing results have several im- portant implications beyond the verification of the age structure of each species. First, they demon- strated that for species that have a high otolith 226 Ra specific activity, 210 Pb/ 226 Ra activity ratios can be used to age fish as young as 3 years with accuracy. Previously these radioisotopes have only been used to age long-lived species (>10 yr; Bennett et al., 1982; Campana et al., 1990; Fenton et al., 1991). Other radioisotope pairs ( 228 Th: 228 Ra) have been used to age short-lived tropical species (Campana et al., 1993), but these are only useful for fish up to 5 years old because of the short half-life of Th-228. Second, for relatively short-lived species, radiomet- ric ageing of whole otoliths and cores using a single- phase linear model of otolith mass growth rate gave similar results. Campana et al. (1990) and Smith et 1 Williams, D. Australian Institute of Marine Science, PMB No. 3, Townsville 4810, Queensland, Australia. Personal commun., 1993. al. (1991) argued that new material accreting to the outer surface of the otolith may not accrete 226 Ra in similar specific activities to the juvenile (t=0). This would invalidate the use of a simple otolith mass growth model to interpret the radiometric data for otoliths of postjuvenile fish. However, even with a single-phase linear mass growth model (a two-phase model would have reduced the age estimates), we were able to verify that the ring counts in whole otoliths were a more accurate measure of the true age than counts from sectioned otoliths (in accord with core radiometric ages). However, we agree with Smith et al. (1991) that otoliths should be cored for radiometric ageing, if possible, which would avoid the use of an otolith mass growth model. The third point that arises from our analyses re- lates to the ratio of allogenic to radiogenic lead in Lutjanus otoliths. We set the uptake activity ratio value at zero (R=0.0) because higher values would have lowered the age estimates (e.g. Smith et al., 1991). However, from the stable lead/barium ratios and the high age estimates of two of the L. sebae samples (2068 and 2069) it appears that, at least for this species, the juveniles may be taking up more allogenic 210 Pb than the adults (Fenton and Short, 1992). There was no systematic increase in the Pb/ Ba mass ratios of L. malabaricus and there is insuf- ficient data for L. erythropterus to be conclusive (Fig. 1 14 Fishery Bulletin 93(1), 1995 3). However, for lutjanids it appears that a Pb/Ba mass ratio <0.2 probably indicates that the assump- tion of a low initial activity ratio (R) is valid, whereas the three samples where the Pb/Ba mass ratio is 0.3- 0.6 indicate that the assumption of low R may be invalid. The Pb/Ba ratios are, therefore, a useful test of the validity of the low R assumption. Finally, this appears to be the first instance where radiometric methods are more consistent with whole- otolith ages rather than sectioned-otolith ages. All previous radiometric studies offish from temperate and subtemperate waters have verified section counts (Bennett et al., 1982; Campana et al., 1990; Fenton et al., 1990, 1991; Smith et al., 1991). The metabolic effects of the annual cycle of inorganic and organic deposition in otoliths may be more pronounced in these environments resulting in clear annuli in otoliths offish from more temperate regions. Conclusions This study has shown that radiometry using 210 Pb/ 226 Ra activity ratios in both whole and cored otoliths can accurately estimate the ages of fish as young as 3 years. Stable leadibarium mass ratios were used to identify samples that may invalidate the assump- tion of constant uptake of allogenic lead (i?=0). For the lutjanids examined, ring counts in sectioned otoliths were shown to overestimate fish ages. Meth- ods such as marginal increment analysis do not verify that the ageing method used is accurate unless the pattern is demonstrated to be consistent for all age classes. This indicates that tropical fish should be aged by two independent methods where possible to help minimize possible ageing errors. Acknowledgments We thank John Salini, David Brewer, and Ted Wassenberg for coordinating otolith collection and Robert Chisari for meticulously performing the ra- diochemical alpha source preparations. Gwen Fenton and Chris O'Brien made constructive comments on an earlier draft of the manuscript. This project was partly funded by the Australian Fishing Industry Research and Development Council (FRDC grants 88/90 and 29/91). Literature cited Allen, G. R. 1985. FAO species catalogue. Vol. 6: Snappers of the world. FAO, Rome, 208 p. Allen, G. R., and R. Swainston. 1988. The marine fishes of north-western Australia. W. A. Museum, Perth, 201 p. Augustine, ()., and T. J. Kenchington. 1987. A low-cost saw for sectioning otoliths. J. Cons. Int. Explor. Mer 43:296-298. Bennett, J. T., G. W. Boehlert, and K. K. Turekian. 1982. Confirmation of longevity in Sebastes diploproa (Pi- sces: Scorpaenidae) from 210 Pb/ 226 Ra measurements in otoliths. Mar. Biol. 71:209-215. Blaber, S. J. M., D. T. Brewer, and A. N. Harris. 1994. The distribution, biomass and community structure of fishes of the Gulf of Carpentaria, Australia. Aust. J. Mar. Freshwater Res. 45:375-396. Brouard, F., and R. Grandperrin. 1984. Les poissons profonds de la pente recifale externe a Vanuatu. Notes et documents d'Oceanographie No. 11. ORSTROM Port- Vila, Vanuatu, 131 p. Campana, S. E., and J. D. Neilson. 1985. Microstructure of fish otoliths. Can. J. Fish. Aquat. Sci. 42:1014-1032. Campana, S. E., H. A. Oxenford, and J. N. Smith. 1993. Radiochemical determination of longevity in flyingfish Hirundichthys affinis using Th-228/Ra- 228. Mar. Ecol. Prog. Ser. 100:211-219. Campana, S. E., K. C. T. Zwanenberg, and J. N. Smith. 1990. 210 Pb/ 226 Ra determination of longevity in redfish. Can. J. Fish. Aquat. Sci. 47:163-165. Casselman, J. M. 1974. Analysis of hard tissues of pike Esox lucius with spe- cial reference to age and growth. In T B. Bagenal (ed. ), The ageing offish, p. 13-27. Unwin Brothers, Ltd., England. Chen, C. Y., S. Y. Yeh, and H. C. Liu. 1984. Age and growth of Lutjanus malabaricus in the north west shelf off Australia. Acta Oceanogr. Taiwan 15:154-164. Conover, W. J. 1980. Practical nonparametric statistics. John Wiley and Sons, New York, 493 p. Davis, T. L. O., and G. J. West. 1992. Growth and mortality of Lutjanus vittus from the north west shelf of Australia. Fish. Bull. 90:395-404. Druzhinin, A. D., and N. A. Filatova. 1980. Some data on Lutjanidae from the Gulf of Aden. J. Ichthyol. 39:8-14. Edwards, R. R. C. 1985. Growth rates of Lutjanidae (snappers) in tropical Australian waters. J. Fish Biol. 26:1-4. Fenton, G. E., D. A. Ritz, and S. A. Short. 1990. 210 Pb/ 226 Ra disequilibria in otoliths of blue grenadier Macruronus novaezelandiae: problems associated with ra- diometric ageing. Aust. J. Mar. Freshwater Res. 41: 467-473. Fenton, G. E., S. A. Short, and D. A. Ritz. 1991. Age determination of orange roughy, Hoplostethus atlanticus (Pisces: Trachichthyidae), using 210 Pb: 226 Ra disequilibria. Mar. Biol. 109:197-202. Fenton, G. E., and S. A. Short. 1992. Fish age validation by radiometric analysis of otoliths. Aust. J. Mar. Freshwater Res. 43:913-922. Francis, R. I. C. C. 1988. Are growth parameters estimated from tagging and age-length data comparable? Can. J. Fish. Aquat. Sci. 45:936-942. Grant, E. M. 1985. Guide to fishes. Dep. Harbours and Marine, Bris- bane, 896 p. Milton et al.: Ageing of Lutjanus erythropterus, L malabancus. and L sebae I 15 Ju, D. R., S. Y. Yeh, and H. C. Liu. 1988. Age and growth of Lutjanus altifrontalis in the waters off northwest Australia. Acta Oceanogr. Taiwan 20: 1-12. Knight, W. 1968. Asymptotic growth: an example of nonsense disguised as mathematics. J. Fish. Res. Board Can. 25:1303-1307. Lai, H. L., and H. C. Liu. 1974. Age determination and growth of Lutjanus sanguineus in the South China Sea. J. Fish. Soc. Taiwan 3:39-57. 1979. Age determination of Lutjanus sanguineus in the Arafura Sea and northwest shelf. Acta Oceanogr. Taiwan 10:160-171. Longhurst, A. R., and D. Pauly. 1987. Ecology of tropical oceans. Acad. Press, London, 407 p. McPherson, G. R., and L. Squire. 1992. Age and growth of three dominant Lutjanus species of the Great Barrier Reef inter-reef fishery. Asian Fish. Sci. 5:25-36. Manooch, C. S. 1987. Age and growth of snappers and groupers. In J. J. Polovina and S. Ralston (eds.), Tropical snappers and grou- pers: biology and fisheries management, p. 329- 373. Westview Press, Boulder. Morawska, L., and C. R. Phillips. 1993. Dependence of the radon emanation coefficient on radium distribution and internal structure of the material. Geochim. Cosmochim. Acta 57:1783-1797. Ratkowsky, D. A. 1986. Statistical properties of alternative parameter- izations of the von Bertalanffy growth curve. Can. J. Fish. Aquat. Sci. 43: 742-747. Sainsbury, K. 1988. The ecological basis of multispecies fisheries, and management of a demersal fishery in tropical Australia. In J. Gulland (ed.), Fish population dynam- ics, p. 349-382. Wiley, Chichester, England. SAS (SAS Institute, Inc). 1989. Non-linear regression. In SAS user's guide: statis- tics, p. 575-606. SAS Inst., Inc., Cary, NC. Schnute, J., and D. Fournier. 1980. A new approach to length-frequency analysis: growth structure. Can. J. Fish. Aquat. Sci. 37:1337-1351. Smith, J. N., R. Nelson, and S. E. Campana. 1991. The use of Pb-210/Ra-226 and Th-228/Ra-228 disequilibria in the ageing of otoliths of marine fish. In P. J. Kershaw and D. S. Woodhead (eds), Radionuclides in the study of marine processes, p. 350-359. Elsevier, London. Vaughan, D. S., and P. Kanciruk. 1982. An empirical comparison of estimation procedures for the von Bertalanffy growth equation. J. Cons. Int. Explor. Mer 40:211-219. West, I. F., and R. W. Gauldie. In press. Determination of fish age using 210 Pb: 226 Ra disequilibrium methods. Can. J. Fish. Aquat. Sci. 51. Wong, C. S., Y. P. Chin, and P. M. Gschwend. 1992. Sorption of radon-222 to natural sediments. Geo- chim. Cosmochim. Acta 56:3923-3932. Yeh, S. Y, C. Y. Chen, and H. C. Liu. 1986. Age and growth of Lutjanus sebae in the waters off northwestern Australia. Acta Oceanogr. Taiwan 16:90-102. Abstract. — Age and growth of the dusky shark, Carcharhinus obscurus, was estimated from bands in the vertebral centra of 122 individuals and from length-fre- quency data from 341 individuals. The von Bertalanffy growth func- tion parameters from the vertebral analysis were considered more ro- bust (L =373, #=0.038, < =-6.28, male; zT=349, #=0.039, * =-7.04, female). Comparison of male and female growth curves generated from vertebral data indicate a sta- tistically significant difference; however, these differences are due primarily to larger sizes attained by adult females. Estimates of age at maturity indicate that dusky sharks follow the typical carchar- hinid pattern of slow growth and late age at maturity. The size at maturity is reported at 231 cm FL and 235 cm FL for males and fe- males, respectively. These lengths correspond to approximately 19 years for males and 21 years for fe- males. The oldest fish aged from ver- tebrae was a 33+ year-old female. Age and growth estimates for the dusky shark, Carcharhinus obscurus, in the western North Atlantic Ocean Lisa J. Natanson John G. Casey Nancy E. Kohler Narragansett Laboratory, Northeast Fisheries Science Center National Marine Fisheries Service, NOAA 28 Tarzwell Drive Narragansett, Rhode Island 02882-1 199 Manuscript accepted 15 July 1994. Fishery Bulletin 93:116-126 (1995). Sharks have become increasingly important in U.S. commercial fish- eries in the western North Atlantic Ocean in recent years. U.S. landings of large coastal sharks, represented primarily by several species in the family Carcharhinidae, increased from 135 to 7,122 metric tons (t) from 1979 to 1989 (Anon., 1993). Musick et al. (1993) reported that annual recreational catches are es- timated to be 35,000 U.S. tons and related annual mortality is over 10,000 U.S. tons (9,074 1). As a group, sharks tend to exhibit slow growth, late age at maturity, and low fecun- dity (Holden, 1973). As a conse- quence of these life history charac- teristics, recruitment in sharks is directly dependent on stock size (Holden, 1973). This direct relation- ship means that elasmobranchs may not be able to recover readily from overexploitation (Holden, 1973). The dusky shark, Carcharhinus obscurus, is part of the species com- plex presently managed under the Secretarial Shark Fisheries Manage- ment Plan (FMP) for the Atlantic Ocean (Anon. 1993). Currently, dusky sharks are harvested in commercial fisheries off the southeastern United States and in the Gulf of Mexico. Rec- reational fishermen off the northeast- ern United States also catch dusky sharks (Casey and Hoey, 1985; Musick et al., 1993). The shark FMP (Anon. 1993) details the need for ac- curate life history information on in- dividual species taken in the shark fishery. Proper management at the species level requires specific infor- mation on age and growth. The dusky shark is a common coastal pelagic species with a world- wide distribution in temperate and tropical waters (Compagno, 1984). In the western North Atlantic, it ranges from as far as Banquereau Bank off Nova Scotia, Canada, to southern Brazil, including the Gulf of Mexico and Caribbean Sea (Hoey, 1983; Compagno, 1984). Tagging studies show dusky shark move- ments from southern New England to Yucatan, Mexico (Casey et al. 1 ; Hoey, 1983). Age and growth studies of large sharks are difficult because many species are highly migratory, mak- ing them available for only short seasonal periods, and different ele- ments of the population segregate spatially by size and sex (Hoenig and Gruber, 1990). In addition, the large size attained by adults makes them difficult to sample. Recent lit- erature has discussed the benefits of growth and longevity estimates attained from tag and recapture 1 Casey, J. G., H. L. Pratt Jr., and C. E. Stillwell. 1980. The shark tagger summary. Newsletter of the Coop. Shark Tagging Program. U.S. Dep. Commer., Northeast Fish. Sci. Cent., Natl. Mar. Fish. Serv., 28 Tarzwell Rd., Narragansett, RI, 02882-1199. 116 Natanson et al.: Age and growth estimates for Carcharhmus obscurus I 17 studies (Casey and Natanson, 1992). These data are not available for the dusky shark nor is validation of vertebral band periodicity. Previous attempts to age the dusky shark were based on limited data and were inconclusive (Lawler, 1976; Hoenig, 1979; Schwartz, 1983). We have attempted to strengthen age esti- mates of C. obscurus by using vertebral band counts together with marginal increment analysis and by us- ing comparisons with length-frequency data. With the von Bertalanffy growth function thus derived, we esti- mate age at maturity and longevity for this species. Materials and methods Data and vertebral samples from dusky sharks were obtained between 1963 and 1993 from research cruises, sport fishing tournaments, and commercial shark fishermen from Cape Cod, Massachusetts, to off the east coast of Florida. Vertebral samples were taken in all months except January, March, and November. Length measurements Total and fork lengths were measured to the nearest centimeter (cm) for each specimen. Fork length (FL) was measured from the tip of the snout to the fork of the tail. Total length (TL) is defined as the distance from the snout to a point on the horizontal axis in- tersecting a perpendicular line extending downward from the tip of the upper caudal lobe to form a right angle (Kohler et al. 2 ). All lengths used are fork lengths unless otherwise noted. FL can be converted to TL by using the regression equation: FL = 0.8352 (TL) -2.2973. [r 2 = 0.99, n = 167] Vertebral samples Vertebral samples were taken from above the bran- chial chamber. Sections of vertebral columns were trimmed of excess tissue and then frozen or preserved in 70% ethanol (Casey et al., 1985). Two vertebrae from each specimen were processed histologically following Casey et al. (1985), with the exception of the use of RDO (DuPage Kinetics) for decalcification. All vertebral sections were cut sagit- tally through the focus to a thickness of 80-100 mi- crons, stained with Harris hematoxylin, and mounted in glycerin jelly (Humason, 1972). Bands in the vertebra were counted from an im- age projected on a Summagraphics MM-1812 digi- tizing tablet (Skomal, 1990). Measurements from the focus to growth bands at points along the internal corpus calcareum were digitized directly into an IBM PC-XT computer. The radius of each centrum was measured from the focus to the distal margin of the intermedialia along the same diagonal as the band measurements. Annual growth marks were defined following Casey et al. (1985) for the sandbar shark, Carcharhinus plumbeus, where the annual mark is defined by a wide translucent zone that traverses the intermedialia and continues into the corpus calcareum as an opaque band. Vertebral sections from 171 dusky sharks were prepared. Bands in the same centrum section were counted at least once by each of four investigators to verify that the band counts were repeatable. Sections were considered unreadable if bands could not be discerned in accordance with the above definition. If two readers considered the section unreadable, the sample was eliminated from the final analysis. Counts were accepted if two or more readers agreed. The individual ring measurements for all readers in agreement were then averaged. If two readers agreed on one count and two on another for the same specimen, the higher count was accepted. Specimens where there was no initial agreement were recounted until two of the investigators reached a consensus or the sections were discarded. The relationship between vertebral radius (VR) and FL was calculated to determine the most appro- priate method for back calculation of the size-at-age data (Ricker, 1969). The FL to VR relationship was linear but did not pass through the origin. There- fore, the Lee method was considered more appropri- ate (Ricker, 1969): I - a + (b x s), where / = the length of fish when the vertebra was obtained; a = the intercept on the length axis; b = the slope of the line; and s = the total vertebral radius. A von Bertalanffy growth function (VBGF) was fit- ted to the data by using the following equation (von Bertalanffy, 1938): L t =L„(l-e- k «-<°>), 2 Kohler, N. E., J. G. Casey, and P. A. Turner. Length-weight re- lationships for 13 Atlantic sharks. Unpubl. manuscr. where L t = predicted length at time t\ L x = mean asymptotic fork length (of the fish); K = a growth rate constant (yr _1 ); and t = the theoretical age at which the fish would have been zero length. 118 Fishery Bulletin 93(1), 1995 Growth in length data were analyzed by using FISHPARM, an IBM PC compatible program (Prager et al., 1987), which implements Marquardt's algo- rithm for nonlinear least squares parameter estima- tion (Marquardt, 1963). Bernard's (1981) multivariate analysis for compar- ing growth curves was employed to test the hypoth- esis that male and female vertebral growth curves were the same. This method also determines which of the von Bertalanffy parameters are the most sta- tistically significant cause of any differences in growth. Marginal increment analysis Validation, the confirmation of the temporal mean- ing of the growth increment (Brothers, 1983), is dif- ficult to attain for large pelagic species and was at- tempted by using marginal increment analysis. The marginal increment ratio (MIR) (Skomal, 1990) was calculated by using the following equation: MIR = (VR - R n )/(R„ - R n ^), where VR = the vertebral radius; R n = the last complete band; and R n _ 1 = the next to last complete band. Mean MIR was plotted against month to locate peri- odic trends in band formation. The MIR relates the edge formation to the width of the previous completed band, which corrects for differences in band width between small and large fish. Length frequency Length-frequency distributions were analyzed by us- ing Shepherd's (1987) model. The sample was sepa- rated by sex and calculations were made at 3-cm in- tervals. Initial values of L m and K, based on biologi- cal parameters obtained from the literature (Springer, 1960; Compagno, 1984) were entered into the pro- gram which was then rerun until the highest score function was attained. The L M and if associated with this score function were used to calculate t Q by using the following equation: t Q = t + (UK) (\n[L x - L t ]/LJ, where t = (birth); L t = mean size at birth; K = the von Bertalanffy growth constant; and L oo = the mean asymptotic fork length. Longevity Estimates of longevity were obtained by using tag and recapture data. Data on eight recaptured dusky sharks at liberty for greater than 10 years were ex- amined. Age at tagging was assigned from the size estimate provided at the time of release. This esti- mated age was based on growth curves derived from vertebrae. The number of years at liberty were then added to estimate age at recapture. Results Vertebral samples Of the 171 processed vertebra, 36 (21.0%) were con- sidered unreadable. Initial agreement by two or more readers was reached on 89 specimens. The remain- ing 50 sections were recounted by two of the investi- gators. A consensus was reached on 37 of those re- counted and the rest were discarded as unreadable. Six were then eliminated for having no information on sex. The remaining 120 (70.2%) consisted of 53 male and 67 female specimens ranging in size from a 73 cm FL neonate to a 296 cm FL adult female. The FL-VR regression showed a linear relationship: FL = 12.82(VR) + 24.99 [n = 114; r 2 = 0.99] . The FL to VR relationship was significantly differ- ent between the sexes for all fish combined (ANCOVA, P<0.05). However, this was due to three large females whose removal from the analysis al- tered the curves and showed the males and females to be statistically indistinguishable (P<0.05) (Fig. 1). We chose to use the combined relationship without those three samples. Back-calculated as compared with empirical length-at-age data show a smaller estimated size for fish of younger ages, when calculated from the ver- tebrae of the older fish, indicating the presence of a slight Lee's phenomenon for both sexes (Table 1). Lee's phenomenon was more pronounced in females and increased with age. The MIR data showed a distinct, periodic trend of increasing increment growth from April through June (female) or July (male); after this peak there was a slight decrease and apparent leveling (Fig. 2). The decrease in incremental growth is not large enough to indicate a double band formation. The graph suggests that an annual winter band is formed between September and April. This band can be vis- ible by February in males; no data were available for females. The time of annulus formation cannot Natanson et al.: Age and growth estimates for Carcharhinus obscurus 300 - 250 g 200 _c ? 150 o _i £ 100 - 50 -■ • ••» Large females not used for regression calculation 10 15 Radius (mm) 20 25 Figure 1 Relationship between vertebra] radius (cm) and fork length (cm) for male and female dusky sharks, Carcharhinus obscurus. be further established owing to a lack of winter samples. January was used as the month of band formation for the assignment of age classes (Casey et al., 1985). Back-calculated length at first band (80.2 cm FL male; 85.8 cm FL female ) corresponded closely to the known size at birth of 85-100 cm TL (Castro, 1983; Compagno, 1984). The first winter band would have formed after approximately six months growth (as- suming January deposition and spring parturition), and the following bands represented annual growth (Branstetter, 1987). The oldest female in the sample was 33+ years and the oldest male, 25+ years. The parameters of the VBGF determined from the back-calculated data were similar to known life his- tory characteristics except that the predicted L^ for males was higher than that for females (Table 2). Those samples that were neonates with no visible birthmark were excluded from the VBGF analysis. Therefore, only 114 samples (47 male and 67 female) were included in the final calculations. The t Q and if values appear simi- lar between the sexes (Table 2). However, the male and female growth curves are significantly different (P<0.05) based on Bernard's ( 1981) multivariate analysis (Table 3). The results indicated that the differences were caused by the t and L x values (in order of significance). The reported size at maturity for the dusky shark is 231 cm FL and 235 cm FL for males and females, respectively. These lengths correspond to 19 years for males and 21 years for females based on the ver- tebral growth curves (Table 4). Length frequency Length observations from a total of 208 female and 133 male dusky sharks were used to calculate von Bertalanffy parameters by using length-frequency analysis. Samples were obtained from 1961 to 1987 for the months May through November. Because of small yearly sample sizes, data for all years were combined by month. A comparison of the VBGF parameters from the length-frequency analysis [LF] with those derived from vertebral analysis (Table 2) shows that the L m and t values from the vertebral analysis for females were lower than those derived from the length-frequency analysis, and that the K value for females was basi- cally the same for both data sets. The length-frequency analysis for males results in a lower L^ than that from the vertebral analysis and in higher t and K values. The VBGF differences in both sexes are not large and both curves indicate late age at maturity (males: 25 yr [LF], 19 yr [vertebral]; females: 16 yr [LF], 21 yr [ver- tebral]) and slow growth (males: if =0.049 [LF], 0.038 [vertebral]; females: #=0.040 [LF], 0.039 [vertebral]) (Fig. 3). The von Bertalanffy parameters for the sexes combined are shown for comparison (Table 2). Longevity Tagging records from NMFS Cooperative Shark Tag- ging Program show that 6,067 dusky sharks were tagged and 131 recaptured between 1962 and 1992. 120 Fishery Bulletin 93(1), 1995 Table 1 Back-calculated and observed size-at-age data for male and female dusky shark Carcharhinus obscurus. Male Ring( age in years) Birth 6 months 1 2 3 4 5 6 7 8 Back-calculated X 80.2 87.3 94.1 103 113.9 124.4 132 140.5 149.7 157.7 163.8 SD 5.6 5.5 6.2 7 9.2 10.6 10.7 11.7 12.7 12.6 10.5 n 47 34 30 28 25 21 20 20 19 17 Observed X 87.7 100.5 123.5 116.7 124.5 138 148 173.5 164 SD 5.1 53. 2.1 5.5 4.9 0.7 7.9 n 13 4 2 3 4 1 1 2 3 Male Ring( age in years) 9 10 11 12 13 14 15 16 17 18 19 Back-calculated X 161.7 179.2 188.6 196.1 202.4 209.8 216.1 220.5 226.5 232.3 237.9 SD 11.6 12.3 10.4 11.8 11.7 13.5 13 11.4 11.7 12.4 12.5 n 14 13 11 10 10 10 10 9 9 8 8 Observed X 172 177.5 177 233 245 260.5 SD 2.1 9.2 n 1 2 1 1 1 2 Male Ring (age in years) 20 21 22 23 24 25 26 27 28 29 30 Back-calculated X 246.4 252.6 254.9 256.6 252.6 255.2 SD 14.5 13.2 15.1 16.1 n 6 6 4 3 1 1 Observed X 256.5 256 263.5 265 SD 3.5 9.2 n 2 1 2 1 1 female Ring (age in years) Birth 6 months 1 2 3 4 5 6 7 8 Back-calculated X 85.8 92.3 99.3 107.4 118.8 128.6 136.2 143.7 150.5 157.8 164.8 SD 5.1 4.5 5 6.3 8.4 9.4 9.5 8.5 8.8 9.4 9.9 n 67 55 52 50 48 45 41 38 35 33 33 Observed X 91.8 104.3 106.5 107 122.3 134.8 138.3 161 154.5 SD 4.5 7.2 6.4 2.8 3.8 16.8 26 12.5 3.5 n 12 3 2 2 3 4 3 3 2 Natanson et al.: Age and growth estimates for Carcharhinus obscurus 121 Table 1 (continued) Female Ring (age in years) 9 10 11 12 13 14 15 16 17 18 19 Back-calculated X 171.5 178.8 186.3 192.9 199.8 206.2 212.4 216.7 222.6 228.4 233.4 SD 10.2 9.6 10.8 10.9 12.4 12.9 13.3 10.5 11.6 12.3 12.6 n 33 30 28 28 28 28 28 27 27 27 27 Observed X 183 190 262 281 SD 15.5 9.9 n 3 2 1 1 Female Ring (age in years) 20 21 22 23 24 25 26 27 28 29 30 Back-calculated X 237.9 242.1 246 149.8 252.9 256.4 258.6 262.9 265.4 266.2 265.1 SD 12.3 12.1 12.5 12.2 12.6 13.5 11.9 13.3 11.8 15.2 12.7 n 26 24 23 22 20 16 15 13 9 5 4 Observed X 253.4 263 284 156.5 161.5 274 266.5 272 270.8 281 262 SD 14.8 2.1 9 16.3 10.5 6.5 n 2 1 1 2 4 1 2 4 4 1 1 Female Ring (age in years 1 31 32 33 34 35 36 37 38 39 40 41 Back-calculated X 267.9 277.1 291.4 SD 15.2 16.6 rc 3 2 1 Observed X 269 269 276 SD rc 1 1 1 Eight of these fish were at liberty from 10.1 to 15.8 years. Estimated ages at tagging were based on the vertebral growth curve and ranged from birth to 27 years. The best example of longevity came from a dusky shark that was tagged at an estimated 27 years (260 cm FL) and was recaptured 12 years later at an estimated age of 39 years (Table 5). Discussion In the present study, vertebral data and length-fre- quency data were independently analyzed to derive estimates of von Bertalanffy growth parameters for the dusky shark. Because of the differences between the methods and their sensitivity to the data used to calculate the VBGF parameters, each method pro- duced slightly different growth curves and, therefore, different estimates of age at maturity and longevity (calculated based on maximum reported size). The length-frequency estimates obtained from the dusky shark data are probably somewhat biased owing to limitations of the data and properties of the length- frequency model (Majkowski et al., 1987; Shepherd 122 Fishery Bulletin 93(1). 1995 et al., 1987; Natanson, 1990). As a slow-growing, long-lived species, the dusky shark may have over- 1.2 MALE 1 OH = 0.8 /\ Z 0.6 < W 04 / V" 02 r^* J FMAMJ JASOND MONTH n= 001 64 11 13 54 00 12 FEMALE or 5 0.8 y\ Z 0.6 < W 04 J 02 J FMAMJ JASOND MONTH n= 001 517 16 22 20 00 Figure 2 Mean vertebral marginal increment ratios (MIR) by month for each of four readers for the dusky shark, Carcharhinus obscurus. Number of samples used to calculate the means for each month are located below the figure. lapping lengths at age which may obscure length modes and bias the estimates of model parameters (Rosenberg and Beddington, 1987; Shepherd et al., 1987). The vertebral method is therefore considered the more robust method and the length-frequency parameters are used for comparison only. Yoccoz (1991) has brought up questions as to the validity of judging biological significance based on statistical tests. He suggests that statistical signifi- cance is not necessarily indicative of biological sig- nificance; this appears to be the case with the dusky shark. The statistically significant differences shown between male and female dusky shark vertebral growth curves may not reflect biological differences. Examination of the length-at-age data suggests that biologically the differences between male and female vertebral curves are small. The age and size at ma- turity differ by only two years and five centimeters for males and females (Table 4). Females are pre- sumed to grow ultimately to a larger size than males. This means that either growth slows in males after maturity or that males do not live as long as females. The vertebral VBGF derived in this study is very similar to the curve attained by Hoenig (1979) for combined sexes for the ages under consideration (birth to 33 years) but is different from data presented by Lawler (1976) and Schwartz (1983). Hoenig's ( 1979) parameter values for the VBGF have a slightly higher L ro and t and lower K than parameters de- rived from vertebral analysis in the present study (Table 2). Lawler (1976), using vertebral analysis to determine the age of female dusky sharks, obtained VBGF-parameter values markedly different from the present study (Table 2). The L x in his study is more than twice as large as the L x reported here and his lvalue suggests a much slower growth rate. These two factors combine to make Lawler's (1976) curve appear as a straight line from birth to 34 years. Table 2 The von Bertal, dusky sharks ai Parameters inffy para id Lawler meters derived in this study compared to those derived in Hoenig's (19791 's (1976) study of female dusky sharks, Carcharhinus obscurus. study of male and female Male Female Combined L„ K t n £» K 200 kg/hr) in spring and fall at depths of 183-365 m and decreased in winter, while above 183 m CPUE was highest (about 50- 150 kg/hr) in July-August but dropped to at or near 0.0 kg/hr in November-March. There was no appar- ent seasonal trend in CPUE at depths >366 m. The proportion of mature females at each macro- scopic maturity stage varied seasonally (Fig. 3). Be- cause spring discards included fish that were the Rickey: Maturity, spawning, and seasonal movement of Atheresthes stomias 131 Market n = 2,167 15 20 25 30 35 40 45 50 55 60 65 70 75 80 Discard n = 185 014 012 0.10 08 0.06 0.04 02 000 15 20 25 30 35 40 45 50 55 60 65 70 75 60 Survey n = 351 15 20 25 30 35 40 45 SO 55 60 65 70 75 Length (cm) Figure 2 Length-frequency distributions of arrowtooth flounder by sample category. Dark bars = males; clear bars = females. same size as fish in winter market samples, discard and market samples were pooled by common month and by keeping years separate. Throughout the year, samples almost always included large spent/resting females that did not show signs of ovarian recrudes- cence. Gravid females first appeared consistently in September 1991. The proportion of developing, gravid, and spent females stayed relatively constant through November. In December the first ripe/run- ning fish and a substantial increase in the propor- tion of spent females were seen. The next available sample was March 1992 when all the mature females were in the spent/resting stage. Developing females reappeared the following May, and their proportion increased into the fall. In 1992, the first gravid and ripe/running fish were seen in November. Length at 50% maturity calculated from survey data was 28.0 cm for males and 36.8 cm for females (Table 5). Estimates of L 50% from pooled market and discard ("commercial") data were lower for males and higher for females than estimates from survey data, although confidence intervals for L 50% overlapped. For females, seasonal estimates of L 5m varied widely. The greatest L 5Q% (>41 cm) was seen before spawn- ing (May-August) and the lowest (<37 cm) during spawning (September-December). Parameters for the logistic function were compared with a likelihood- ratio test (Kimura, 1980). Estimates from commer- cial data were significantly different from survey estimates for both females (x 2 =145.490, P«0.001; Fig. 4) and males (x 2 =79.383, P«0.001). In a com- parison of years, logistic curves fit to September- December 1991 (commercial) and September-Novem- ber 1992 ( survey) data were significantly different (like- lihood-ratio test, x 2 =143.257,P«0.001) although again confidence intervals for L 5QC/c overlapped. Ovarian tissue samples were analyzed histologi- cally from 111 female arrowtooth flounder collected late December 1991 during spawning. Each of the five macroscopic maturity stages was represented and no two macroscopic stages showed the same fre- quency distribution of oocyte types (Fig. 5). Chroma- tin nucleolar, perinucleolar, and atretic oocytes were present in all the sampled ovaries. In ovaries of im- mature fish, none of the oocytes had progressed be- yond the perinucleolar stage. Vitellogenic or yolked oocytes were prevalent in developing and gravid stage ovaries, and hydra ted oocytes were seen frequently in gravid and ripe/running stage ovaries. Oocytes with cortical alveoli were most frequently seen in spent/resting ovaries. Atresia was more prevalent in all the ovarian stages of spent/resting fish than in immature fish (Table 6). The percent occurrence of a atretic oocytes was lowest in ovaries from developing fish and high- est in ovaries from spent/resting fish. Beta atresia was most common in spent/resting ovaries but also occurred in developing, gravid, and immature ova- ries. All the immature and 43.8% of the spent/rest- ing females had perinucleolar stage ovaries. Postovulatory follicles (POF) were present in ovaries from all macroscopic stages except immature. POF were most frequently seen in ripe/running ovaries and were common in gravid and spent/resting ova- ries; whereas 7 of 29 developing females examined for histology had ovaries with POF. Postovulatory follicles were also present in 7 of 21 spent/resting 132 Fishery Bulletin 93(1). 1995 162 79 300 148 165 172 104 41 50 36 0.4 0.2 - 0.0 H — V — V — H 1 — *V — 4 H H — 4 f — H f — V — V — H Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov 1991 rj> <3 1992 [> H Developing £ Gravid Ripe/Running [~ H Spent/Resting Figure 3 Proportions of mature female arrowtooth flounder at each macroscopic maturity stage by month. Number of females listed at top. Table 4 Washington commercial bottom trawl catch rates (kg/hr) of arrowtooth flounder, Atheresthes stomias, by depth interval based on the minimum depth recorded for each tow. n = number of tows. Year Month 1- -182 m 183- -365 m 366-547 m 547+ m n kg/hr n kg/hr n kg/hr n kg/hr 1991 July 1,644 155.7 176 184.0 31 44.6 9 134.5 August 1,391 54.5 179 130.6 21 3.0 5 0.0 September 1,362 14.5 187 157.6 54 27.7 7 0.0 October 860 11.8 274 231.7 56 64.3 4 0.0 November 210 0.0 95 20.3 37 22.7 3 0.0 December 187 0.0 126 8.9 82 35.8 7 0.0 1992 January 221 0.1 120 14.9 82 8.9 18 11.0 February 596 0.0 240 10.7 61 12.2 37 0.4 March 883 0.3 201 19.5 178 39.8 85 8.9 April 462 7.4 155 53.9 110 68.5 31 1.7 May 829 16.3 232 228.2 63 71.0 40 0.0 June 1,131 19.0 174 344.1 38 16.4 90 0.5 July 1,246 53.4 134 50.8 13 1.7 82 1.5 females with perinucleolar-stage ovaries. One 41.0- cm gravid female had no POF and 6.3% of its oocytes were hydrated. All other gravid females had POF. Overall mean oocyte diameters (/im) and standard deviations were as follows: perinucleolar 79.6 ± 34.9 (n=420); cortical alveoli 185.4 ± 23.4 (n=220); vitellogenic 722.9 ± 73.9 (n=290); and hydrated 940.6 ± 206.8 (n=170) (Fig. 6). Chromatin nucleolar stage oocytes were not measured because they never rep- resented the most advanced stage present in an ovary. Mean diameter of perinucleolar oocytes in spent/resting females was about 10 /im greater than that in the immature females (Student's £-test, P<0.003). Some of the variance in hydrated oocyte size Rickey. Maturity, spawning, and seasonal movement of Atheresthes stomias 133 Table 5 Parameter estimates for the logistic model' of proportion mature at length (cm ), length at 50% mature, and 95% confidence intervals, for arrowtooth flounder, Atheresthes stomias from 1991- -92 Washington commercial (pooled market and discard) 1992 Washington survey, and 1972- -75 Oregon data (See Footnote 3 in the text). For females, results are also given for Washi ngton commereial data grouped by months in relation to the spawning season.