Dentzau and Chittenden Population dynamics of Polydactylus octonemus in the Gulf of Mexico 



457 



young towards Texas. The coastal area off south Texas 

 and northern Mexico has converging currents much of 

 the year, water which at least in part heads easterly 

 or northerly along the shelf edge after leaving the coast 

 near 26°N. Whatever their source, pelagic young trans- 

 ported to and entrained in the gyre of the Texas- west- 

 ern Louisiana shelf could be transported to inshore and 

 estuarine nurseries there. 



Bathymetric distribution, recruitment, 

 and movements 



Our finding that P. octonemus is most abundant at 

 depths < 22 m in the northwestern Gulf agrees with 

 Hildebrand (1954) and Chittenden and McEachran 

 (1976), who only captured it in the demersal phase and 

 considered it a member of the white shrimp commu- 

 nity. Maximum depths recorded for the demersal form 

 of this species in the Gulf are 66 m off Louisiana 

 (Springer and Bullis 1956) and 65 m off Texas (Wohl- 

 schlag et al. 1979). 



Our finding that P. octonemus appears late March- 

 April and basically disappears late October-late De- 

 cember agrees with the literature for Texas and Loui- 

 siana including Gunter (1938b, 1945), Perret et al. 

 (1971), and Gallaway and Strawn (1974). Some reports 

 indicate a few fish occur in winter in the white and 

 brown shrimp communities throughout the northern 

 Gulf: (1) in January in Galveston Bay, Texas (Gallaway 

 and Strawn 1974, Sheridan 1983), (2) at 18-22 m in 

 January off Texas (Chittenden and McEachran 1976), 

 (3) in shallow waters in January and February at Sabine 

 Lake and Holly Beach, Louisiana (Perry and Carter 

 1979), and (4) at 40 m in early February off Alabama 

 (Bullis and Thompson 1965). The similar periodicity of 

 reported occurrence suggests similar movements and 

 spawning periodicity throughout the northwestern and 

 north-central Gulf. 



The spring recruitment we found in P. octonemus 

 agrees with data in several studies from the northern 

 Gulf including Gunter (1945) and Miller (1965) in Texas, 

 Gunter (1938b) and Perret et al. (1971) in Louisiana, 

 and Ogren and Brusher (1977) in northwestern Florida. 

 Recruitment appears concurrent in estuaries and the 

 shallow Gulf, because we collected fish 50-105 mm in 

 the Gulf off Texas at < 9 m in April, the same period 

 when Gallaway and Strawn (1974) and Landry (1977) 

 collected fish in Galveston Bay 56-97 mm (40-69 mm 

 SL) and 74-101 mm (53-72 mm SL), respectively. 

 Moreover, demersal phase fish first appeared the same 

 month in both Barataria Bay and the shallow Gulf off 

 Louisiana (Gunter 1938b). 



Our finding that P. octonemus disperse offshore 

 beginning in early summer agrees with the literature. 

 Gunter (1945), Gallaway and Strawn (1974), and Per- 



ret et al. (1971) found abundance declined in estuaries 

 in June- August or September, at about the same time 

 (July-September) we found greatest abundance in the 

 Gulf and size gradients from estuaries to offshore. We 

 found the largest, oldest fish in the deeper areas of the 

 white shrimp community or in the transition commu- 

 nity (Chittenden et al. 1982) as they dispersed offshore. 

 This pattern has also been suggested for other species 

 of the Gulf white shrimp community such as Micropo- 

 gonias undulatiis, Peprilus burii, Cynoscion arenariu^, 

 and Larimus fasciatus (White and Chittenden 1977, 

 Murphy 1981, Shlossman and Chittenden 1981, Stan- 

 dard and Chittenden 1984). 



Age determination and growtln 



Age determination, size at age, and von Bertalanffy 

 parameters have not been previously described for P. 

 octonemus. Our calculations for growth are based on 

 sizes at known time scaled to calculated hatching dates 

 to give age. Regression coefficients and K values are 

 the same regardless of hatching date, however, because 

 curves are fitted to the same time dimension between 

 the first and last collections of a cohort. 



We determined age in months and years by length- 

 frequency analysis because we had almost 4 years of 

 data from cruises so close in time that modes were easi- 

 ly followed, and only one cohort usually occurred at a 

 given time. As in Stenotmnus caprinu^ (Geoghegan and 

 Chittenden 1982) and Larmius fasciatus (Standard and 

 Chittenden 1984), modal-group progression analysis 

 can be a superior method of age determination in P. 

 octonemus because (1) spawning primarily occurs in one 

 discrete period, (2) growth of large and small fish in 

 a cohort appears uniform because variances were 

 generally constant between cruises, (3) length frequen- 

 cies are reasonably normally distributed within cohorts, 

 and (4) age need only be determined for 1 or 2 years, 

 ideal conditions for using length frequencies (Lagler 

 1956, Bagenal and Tesch 1978). 



Although it would be desirable to support our aging 

 findings by the more time-consuming analysis of daily 

 otolith increments (Jones 1986), the growth and other 

 parameters we present provide useful upper or lower 

 boundaries for true values (see comments in Methods). 

 We reiterate that our sizes at age and von Bertalanffy 

 parameters are apparent ones; they are affected to an 

 unknown degree by a combination of recruitment, and, 

 especially as the fish approach age I, emigration, gear 

 avoidance, and change from demersal to pelagic behav- 

 ior. Because of similar problems, Knudsen and Herke 

 (1978) suggested length frequencies should not be used 

 to estimate growth in Micropogonius undulatus. We 

 do not agree, however, because the fundamental prob- 

 lem is not the length-frequency method. Rather, it is 



