112 



Fishery Bulletin 89(1). 1991 



We interpret our data to mean P. burti spawns pri- 

 marily—or most successfully— in temporally separate, 

 discrete, Winter (late January-mid-May), and Fall 

 (early September-late October) periods. The consis- 

 tently well-separated, bimodal length frequencies— on 

 which we place great emphasis— must reflect a tem- 

 poral separation in spawning activity or success, or 

 growth and mortality, that originates during late fall 

 and early winter. However, the well-developed^onads 

 and Gravid/Ripe fish we observed suggest spawning 

 could occur throughout that period, possibly at a low 

 level. Allen et al. (1986) and Vecchione (1987) also 

 found consistently bimodal size distributions of P. burti 

 in the northcentral Gulf. Allen et al. (1986) considered 

 them separate spawning peaks in a continuous, not 

 temporally separated, spawning. The actual degree of 

 temporal separation in spawning of P. burti may be im- 

 portant to resolve, because it may influence (1) ap- 

 propriate management practices, (2) how many popula- 

 tions and stocks exist in P. burti, concepts which are 

 not necessarily the same, and (3) how speciation occurs 

 in Peprilus. Given properly randomized geographical 

 sampling, age determination by daily otolith incre- 

 ments (Jones 1986) might resolve the question of how 

 intense is late-fall to early-winter spawning and 

 whether or not, and to what degree, spawning is tem- 

 porally separate. 



Although P. burti appear to spawn primarily in two 

 main periods, Winter and Fall, it also appears that in 

 each period there is much variation in cohort spawn- 

 ing periodicity or success, or in recruitment periodicity. 

 We observed Winter cohorts to appear as distinct abun- 

 dant groups in April in one year but not until June in 

 two other years. Similarly, one Fall cohort was distinct- 

 ly bimodal over a several-month time period. A more 

 exact method of age determination than length fre- 

 quencies, however, is needed to more clearly interpret 

 these phenomena. 



Our interpretation of spawning periodicity in P. burti 

 is similar to findings that other Gulf species spawn in 

 discrete Winter-Spring and Late Summer-Fall periods 

 related to current transport, including Cynoscion are- 

 narius (Shlossman and Chittenden 1981), C. nothus 

 (DeVries and Chittenden 1982), Larimus fasciatus 

 (Standard and Chittenden 1984), Menticirrhus ameri- 

 canus (Harding and Chittenden 1987), and Polydac- 

 tylus octonemus (Dentzau and Chittenden 1990). 

 Spawning of P. burti in the northwestern Gulf, and for 

 many of these other fishes, probably is timed to coin- 

 cide with currents (Shlossman and Chittenden 1981) 

 that transport eggs and larvae from spawning areas 

 to nurseries, assuming P. burti has pelagic eggs and 

 larvae like P. alepidotus and P. triacanthus (Martin and 

 Drewry 1978). Spawning, or its absence, coincides with 

 wind-induced, up- or downcoast, alongshore coastal 



currents which drive circulation in the northwestern 

 Gulf with seasonal reinforcement from the Missis- 

 sippi— Atchafalaya discharge (Kelly et al. 1981). Aver- 

 age winds are downcoast (toward Mexico) during 

 August/September-April/May but upcoast (toward 

 Florida) during May/ June-July/ August. Nearshore cur- 

 rents parallel the coast. Upcoast wind stress causes up- 

 coast alongshore currents which (1) are reflected in 

 high inshore salinity off Galveston/Freeport and fall- 

 ing sea levels during early to midsummer (Marmer 

 1954, Kelly et al. 1981), and (2) coincide with the sum- 

 mer period of little spawning we observed in P. burti 

 and which is reported in other species just cited. 

 Downcoast wind stress causes downcoast (toward Mex- 

 ico) alongshore currents, onshore surface Ekman 

 transport, and downwelling which (1) are reflected in 

 rising sea levels February-May and August-October, 

 (2) transport low-salinity water downcoast causing a 

 salinity minimum off Galveston/Freeport during Sep- 

 tember and October, and May and June, and (3) coin- 

 cide with the two major spawning periods we suggest 

 for P. burti and which are reported as major or minor 

 periods in other species just cited. Alongshore currents 

 continue downcoast from late fall to early winter. 

 Seemingly, however, no distinct, abundant groups of 

 P. burti originated then, which may reflect low-level 

 spawning or spawning success. Similarly, little or no 

 spawning occurs then in the other species just cited. 

 Temporal variation in the average meteorological and 

 hydrographic patterns may be the primary reason for 

 the variation we noted in cohort spawning and recruit- 

 ment periodicity between and within years. 



Age determination and growth, 

 maximum size and age, and mortality 



Our findings on age and growth in P. burti are new, 

 because this species has not been aged previously. It 

 would be desirable to corroborate them by analysis 

 of daily otolith increments (Jones 1986). However, 

 that may not prove feasible, because recent studies 

 using scales, opercula, vertebrae, and thin-sectioned 

 otoliths, fail to consistently show clear daily increments 

 or annuli (Allen et al. 1986). Therefore, it appears 

 length frequencies are the only way to age P. burti at 

 present. As in our study, supporting length collections 

 must be frequent in time and over a long duration, 

 because cohort boundaries and age are not clear every 

 month. However, they are quite clear in certain months 

 (for examples, the Winter 79 and 80 groups in May or 

 June 1979 and 1980, the Fall 78 and 79 groups in 

 December 1978 and 1979). From the clear groups, one 

 can work chronologically backward and forward in time 

 and gradually assign age and cohort boundaries with 

 reasonable certainty. This process, however, is not as 



