Crabtree et al : Life history of Trachinotus fakatus 



27 



Methods 



Collections 



Permit that we examined were collected from the Florida 

 Keys (/i=308; between 25°40'N. SOnO'W and 24°30'N, 

 82''20'W) during 1995-97 and the Tampa Bay area (» =228; 

 27°40'N, 82°45'W) during 1990-95. Most F^lorida Keys 

 permit were caught with hook-and-line gear (;!=215) or 

 speared («=58) over artificial and natural reefs in the 

 waters off the lower and middle Keys in depths ranging 

 from 10 to 30 m. Other, usually smaller, permit were cap- 

 tured with gill nets (/i = 16), seines (/i = 18). and bottom 

 trawls (n=l) over or near shallow banks adjacent to the 

 Keys. Most of the permit sampled in the Tampa Bay area 

 were small (<400 mm FL) and were captured with seines 

 along sandy beaches; some larger Tampa Bay permit were 

 captured with gill nets (/i=53) or trammel nets (/( = 14). 



Standard length (SL), fork length (FL), and total length 

 (TL) were measured to the nearest millimeter (mm) and 

 weight was measured to the nearest gram. Unless other- 

 wise indicated, all lengths reported in our study are fork 

 lengths. Otoliths (sagittae) were removed, rinsed in water, 

 and stored dry until sectioned; they were later weighed 

 to the nearest 0.01 mg. Gonad weight was recorded to 

 the nearest gram (g), and gonad samples were removed 

 from the fish and preserved in 10'/( buffered formalin; 

 they were later soaked in water for 24 hours and stored in 

 70^7^ ethanol. 



Collections of juvenile permit from sandy beaches off 

 Tampa Bay and the Florida Keys were made with a 21.3 x 

 1.8-m bag seine (6.4-mm mesh in the wings and 3.2-mm 

 mesh in the bag). Seine hauls were made perpendicular 

 to the beach for distances up to 50 m, depending on water 

 depth. Lengths of up to 50 fish from each sample collec- 

 tion were measured to the nearest millimeter. Near Tam- 

 pa Bay, we collected fish at the Gulf of Mexico beaches of 

 Treasure Island (November 1992 -October 1994; 27°46'N, 

 82°46.5'W) and Indian Shores (August 1993-November 

 1994; 27°50'N. 82''50'W). Sampling at each site consisted 

 of five seine hauls every two weeks. Six sandy beaches 

 were sampled monthly in the Florida Keys from July 1994 

 to July 1997: Lower Matecumbe Beach (July 1994-April 

 1996; 24°50.95'N, 80°4415'W), Coco Plum Beach (July 

 1994-April 1996; 24°43.65'N, SFOO.lO'Wi. Clarence P 

 Higgs Beach. Key West (July 1994-July 1997; 24°32.79'N, 

 81°47.26'\V), Bahia Honda State Park (October 1994-May 

 1997; 24°39.81'N, 81°15.44'W). Boca Chica Beach (Oc- 

 tober 1994-April 1996; 24°33.60'N, 81°41.65'W), and 

 Sugarloaf Beach (January 1995-May 1996; 24°36.57'N, 

 81°33.49'W). 



Age and growth 



The left sagitta was usually used for age estimation; how- 

 ever, if the left otolith was broken, lost, or destroyed during 

 processing, the right otolith was substituted. We prepared 

 otoliths for age estimation by embedding them in Spurr, a 

 high-density plastic medium (Secor et al.. 1992). A 1-mm 

 to 2-mm-thick transverse section containing the otolith 



core was cut with a Buehler Isomet low-speed saw with a 

 diamond blade. The section was mounted on a microscope 

 slide with thermoplastic glue (CrystalBond 509 adhesive) 

 and was polished with wet or dry sandpaper (grit sizes 

 ranging from 220-2000) until annuli were visible. Sec- 

 tions were then polished on a Buehler polishing cloth with 

 0.05-gamma alumina powder to remove .scratches. With- 

 out knowledge of fish size or capture date and using a 

 compound microscope equipped with transmitted light, 

 two readers independently counted annuli on each otolith 

 twice. If three of the four readings agreed, then this mode 

 was accepted as the annulus count. If three of the four 

 readings did not agree, each reader again counted annuli 

 independently and without knowledge of previous counts. 

 If three of the resulting six readings agreed, then this 

 mode was accepted as the annulus count. If there were not 

 three readings that agreed, the otolith was excluded from 

 further analysis. In six cases, two sets of three readings 

 that were in agreement occurred. For these six otoliths the 

 two sets of readings differed by only one annulus; there- 

 fore the mean was accepted as the annulus count. 



The percentage of otoliths with an annulus on the edge 

 was then plotted by month so that we could look for a sea- 

 sonal pattern in annulus formation. We did not attempt to 

 measure marginal increments because the margin of per- 

 mit otoliths is highly sculptured and easily broken; how- 

 ever, we did believe that we could discern the presence of 

 an annulus on the otolith's edge. 



The von Bertalanffy (1957) growth equation FL, = L,, 

 (1-e '"'"'"') was fitted to observed age-length data with 

 nonlinear regression procedures. Age was esimated as the 

 annulus count because permit both spawn and form annu- 

 li at about the same time of year. Our estimates of length 

 at age include some seasonal growth that occurred after 

 the formation of the final annulus. Length-weight regres- 

 sions were calculated by linear regression of logju-trans- 

 formed data. 



Sex-specific growth models were compared with an ap- 

 proximate randomization test described by Helser (1996). 

 This test is based on the premise that when the null hy- 

 pothesis of no sex-specific differences in growth is true, a 

 test statistic derived by random assignment of fish to one 

 of two populations will not be different from that observed 

 between sexes. The test statistic is calculated as the re- 

 sidual sums of squares for the sexes-combined von Berta- 

 lanffy growth model minus the residual sums of squares 

 for the two sex-specific models. A probability distribution 

 of the test statistic was generated by a randomization rou- 

 tine with 1000 iterations of the nonlinear models. Only 

 sexed fish were included in the statistical comparison. 



Age validation 



Permit used in the age-validation experiments were cap- 

 tured in waters off the Florida Keys with hook-and-line 

 gear After capture, permit were tagged with dart-type 

 tags and injected with Liquamycin LA-200 (200-mg oxy- 

 tetracycline |OTC|/mL) in the dorsal musculature at a 

 dosage of about 100-mg OTC per kg fish weight. Permit 

 were then held in a 33.5-m-long by 5.5-m-wide by 0.75-m- 



