458 



Fishery Bulletin 88(3|. 1990 



how best to draw a random sample in time and space 

 to characterize the population. The time dimension 

 has three major aspects of variation (diel, annual, 

 and monthly or seasonal), as does the spatial (along- 

 shelf, across-shelf and estuary, and vertical). No pub- 

 lished Gulf study to our knowledge addressed all six 

 aspects, and few addressed any to the extent we did. 

 Our design directly addressed all three time aspects 

 and the across-shelf spatial dimension. We addressed 

 along-shelf and estuarine variation indirectly by 

 referencing the considerable literature. We were not 

 able to address vertical variation, but apparently no 

 published work has done that. The problem of getting 

 representative data for P. ocioneTOws— indeed, for 

 many other Gulf species— may not be fully solved till 

 all six aspects of sampling are concurrently addressed 

 in one study. 



Maximum size, life span, and mortality 



The largest P. octonemus we collected in the north- 

 western Gulf (229 mm) is very close to that reported 

 from off Virginia (233 and 235 mm ?L, Hildebrand and 

 Schroeder 1928), the only Atlantic coast location with 

 published sizes. Breder (1948) noted this species 

 reaches one foot in length but gave no details to sup- 

 port the figure; it is only 2-3 inches larger than our 

 maximum sizes and might reflect a rough rounding-off 

 by him. The largest P. octonemus we found is slightly 

 larger than maximum sizes in other Gulf studies [off 

 Texas: 213 mm (Gunter 1945), 210 mm (Compton and 

 Bradley 1963), 213 mm (171 mm FL) (Chittenden and 

 McEachran 1976); Off Mississippi: 197 mm (Christmas 

 and Waller 1973); Off Florida: 209 mm ?L (Powell et 

 al. 1972)]. However, these reported maximum lengths 

 are similar to the apparent typical maximum size of 

 202-206 mm we found. Smaller maximum sizes in other 

 Gulf reports such as Gunter (1938b), Ferret et al. 

 (1971), Gallaway and Strawn (1974), and Perry and 

 Carter (1979) may reflect gear selectivity t)r the fact 

 they collected largely in estuaries, surf, and shallow 

 Gulf areas. The studies cited, however, took place over 

 much of the northern Gulf so the demersal stage of P. 

 octonemus appears to show little, if any, east-west 

 spatial variation in maximum sizes. 



Our finding the largest, oldest P. ncto)temus in the 

 deeper areas of the white shrimp community may 

 reflect a broad life-history pattern for species there, 

 many of which show a positive size gradient towards 

 offshore. This has been reported for strongly estuarine 

 dependent species such as Micropogonius undulatus 

 and Cynoscion arenarius (White and Chittenden 1977, 

 Shlossman and Chittenden 1981), and more marine 

 forms such as Peprilus burti and Larimus fasciatus 

 (Murphy 1981. Standard and Chittenden 1984). 



The apparent typical maximum life span of 1 year 

 we found for P. octonemus in the northwestern Gulf 

 agrees with Chittenden and McEachran (1976) and 

 appears reasonable, at least for the demersal stage, 

 because the maximum and typical maximum sizes we 

 observed are similar to, or larger than, all other values 

 reported for the Gulf. The value oi ti = \ year might 

 be a little low because P. octonemus disappears at 9-11 

 months of age to spawn in winter, and because we col- 

 lected one fish 15 months of age in April. That one fish 

 is the only instance in 71 cruises in which a second year 

 class co-occurred with young-of-the-year, so our data 

 clearly indicate that at least the demersal phase of life 

 basically ends at 1 year. Though the literature indicates 

 no east-west spatial variation in maximum size, and age 

 by inference, P. octonemus might survive their first 

 spawning and assume a pelagic behavior. Bullis (1961) 

 observed a dense ball of small Polydactylus sp., which 

 could have been young or adult P. octoyiemus, P. vir- 

 ginicus, or P. oligodon (Fischer 1978), at the surface 

 along the 500-fm curve in August in the western 

 Caribbean. 



Midwater and surface collections are needed to 

 resolve the duration of a pelagic phase in adult P. oc- 

 tonemus; that would clarify the validity of the several 

 life-history parameter values we have termed "ap- 

 parent." In addition, collections with trawls larger than 

 we and other workers have used would clarify if and 

 how gear-avoidance affects parameter values. Despite 

 these possible problems, however, it seems clear that 

 few adult P. octonemus resume a demersal existence 

 in the white or brown shrimp communities after age 

 I. We collected only one fish older than 12 months, and 

 Chittenden and McEachran (1976) caught no fish older 

 than 1 2 months even though they caught 33 specimens 

 on the white shrimp grounds in January. 



The apparent mean time- and cohort-specific total an- 

 nual mortality rates we observed for P. octonemus 

 (97-100%) agree with theoretical estimates of 90-100% 

 (Royce 1972:238) if maximum life span is typically only 

 1-2 years. Though these are also apparent values, they 

 are not that far from the range of values for another 

 well-studied, exploited pelagic species, Brevoortia 

 patronus. In that species, total annual mortality rate 

 is 83-95% depending on age (Nelson and Ahrenholz 

 1986). Moreover, the population dynamics of P. octo- 

 nemus in the northwestern Gulf appear generally 

 similar to those reported for many other species there 

 including Micropogonius undulatus (White and Chit- 

 tenden 1977), Cynoscion arenarius (Shlossman and 

 Chittenden 1981), Cynoscion nothuJi (DeVries and Chit- 

 tenden 1982), Peprilus burti (Murphy 1981), Steno- 

 toniiis caprinus (Geoghegan and Chittenden 1982), 

 Larimus fasciatus (Standard and Chittenden 1984), 

 and Leiostomus xnyithurus (Hata 1985). This similar- 



