22 



Fishery Bulletin 88(1), 1990 



important penaeids. However, their sampling frequen- 

 cy was too low to follow the growth of individual 

 cohorts or identify principal periods of emigration from 

 the system. 



Wliile the timing of shrimp emigration from estuaries 

 varies with species and geographic area, a number of 

 factors have been suggested as causal stimuli. These 

 include declines in temperature (Lindner and Ander- 

 son 1956, Pullen and Trent 1969, Chen 1983, Maty- 

 lewich and Mundy 1985), declines in salinity (Roth- 

 lisberg et al. 1985, Staples and Vance 1986, Jayakody 

 and Costa 1988), increased tidal amplitude and currents 

 (Copeland 1965), and an endogenous rhythm coupled 

 with tidal currents (Hughes 1972). 



This study was initiated to provide information about 

 the growth and emigration of P. indicus in St. Lucia 

 and operated concurrently with a plankton sampling 

 program that provided information about the immigi'a- 

 tion of P. indicus postlarvae (Forbes and Benfield 

 1986b). The present paper contains estimates of 

 seasonal differences in growth rates and periods of 

 emigration for P. indicus. 



Methods 

 Beam trawling 



Sampling was conducted in South Lake and the St. 

 Lucia Narrows, a 21 -km channel connecting the lakes 

 with the Indian Ocean; twelve sampling stations were 

 selected (Fig. 1). Samples were collected 17 August 

 1982-6 November 1984 at 4-6 week intervals coinci- 

 dent with spring tides and the plankton sampling pro- 

 gram. Sampling during the first half of 1984 was 

 disrupted when a cyclone struck at the end of January 

 and cut off access to the area. A single sample was 

 obtained in April and routine sampling i-esumed in 

 July 1984. 



A 1-m beam trawl with a frontal area of 0.3 m- was 

 attached to a 12-mm stretched mesh nylon bag with 

 25-mm mesh nylon wings and a nylon 6-mm mesh cod. 

 The bag was fitted with a floating headrope, weighted 

 baserope, and a tickler chain. The trawl was towed 6 m 

 behind an open boat fitted with two outboard engines. 

 At each site, a 10-min trawl was taken against tidal 

 flow (where present) parallel to, and within 10 m, of 

 the shore. Tow distances were not measured but were 

 probably similar because tidal currents were general- 

 ly weak in the vicinity of the sampling stations. Catches 

 were frozen and returned to the laboratory where 

 species, sex, and carapace length were determined. 

 Water temperatures were recorded mid-trawl at each 

 station. 



Figure 1 



.Sampling stations in the St. Lucia Lake system. Pie charts indicate 

 the contribution of each sampling station to the total beam trawl 

 catch of Penaeus indicus. 



Bait fishery 



A bait fishery operates throughout the Narrows (vicin- 

 ity of stations 1-7) and in lower South Lake (vicinity 

 of stations 8-10) and employs 3.7-m gate trawls with 

 a frontal area of 3.45 m- and a 25.4-mm stretched 

 mesh bag. Frozen samples, provided at irregular inter- 

 vals of several days to several months between 17 Sep- 

 tember 1982 and 2 April 1985, were treated in the same 

 manner as the beam trawl samples. 



Data analysis 



The beam trawl length-frequency data from all sites 

 were pooled because sample size differences among 

 sites were frequently large, and the sample sizes from 

 individual sites were often too small to construct mean- 

 ingful length-frequency histograms. A computer pro- 

 gram (MIX) (MacDonald 1986) was used to fit normal 

 curves to the component cohorts in each distribution. 

 Growth rates were estimated by following the mean 

 size of the youngest cohort over consecutive sampling 

 periods. Older cohorts were excluded from the analysis 

 because of the confounding effect of emigration on the 

 cohort mean. The appearance of a new cohort in the 



