NOTES 



COMPARISON OF CATCHES IN 4.3 M 



AND 12.2 M SHRIMP TRAWLS IN 



THE GULF OF MEXICO 



Shrimp trawls used to assess shrimp and fish popula- 

 tions in the southern United States have varied in 

 length, width, and basic design, making comparisons 

 of results among studies difficult. Fishery manage- 

 ment plans by State and Federal agencies emphasize 

 the need for data that can be reliably compared. 

 Techniques and equipment necessary to measure 

 trawl performance so that data collected with dif- 

 ferent trawls can be compared is costly and time 

 consuming (Watson 1976; Loesch et al. 1976; 

 Wathne 1977; Kjelson and Johnson 1978). Recent 

 emphasis has been placed on standardizing gear and 

 sampling methods (Watson and Bane 1985) and 

 determining the effects on catch and mean length 

 of organisms for different tow durations, mesh sizes, 

 trawl widths, and towing vessels (Clark 1963; Chit- 

 tenden and Van Engle 1972; Green and Benefield 

 1982; Matthews 1982; Cody and Fuls 1985). How- 

 ever, sample sizes generally have been small and 

 only selected species have been analyzed. 



The present study evaluates small trawls as popu- 

 lation sampling devices for penaeid shrimp and other 

 organisms in the Gulf of Mexico. The objective of 

 this study was to compare the catch rates and mean 

 lengths of organisms caught with 4.3 m and 12.2 m 

 trawls pulled during day and night. 



Materials and Methods 



The study area was the Gulf of Mexico off Texas 

 between the Colorado River and Port Mansfield in 

 depths from 7 m to 24 m (Fig. 1). Sample sites were 

 established in 1° latitude by 1° longitude grids 

 within the study area. Twenty randomly selected 

 sites were sampled monthly from November 1982- 

 February 1983. Samples were equally and random- 

 ly distributed between day and night. 



At each site two trawls were towed simultaneous- 

 ly for 15 min at approximately 3 kn from the Texas 

 Parks and Wildlife Department (TPWD) RV 

 Western Gulf, a double-rigged 21.9 m steel-hull 

 shrimp trawler. The 4.3 m trawl (small net) was 

 spread by wooden trawl doors 0.4 m high and 0.8 

 m long and the 12.2 m wide trawl (large net) was 

 spread by wooden trawl doors 0.9 m high and 2.1 



m long. Both nets had 5.1 cm stretched mesh web- 

 bing in the body, 4.4 cm mesh in the bag, and were 

 equipped with tickler chains. 



Trawl catches weighing <10 kg were processed 

 by identifying and counting all organisms in the 

 catch. For larger catches a 10 kg subsample was ran- 

 domly selected from the total catch, and the total 

 number for each species was estimated by dividing 

 subsample counts by the proportion of subsample 

 weight to total weight. Total lengths were measured 

 on no more than 50 individuals of each Penaeus 

 shrimp species and no more than 20 individuals of 

 all other species. The arithmetic mean for length 

 data was calculated for each species in each sample. 



The relationship between number caught (or mean 

 length) in the two trawls was tested for linear, mul- 

 tiplicative and exponential models, and log and 

 square root transformations (Sokal and Rohlf 1981). 

 No significant improvement was found over a linear 

 regression with no transformation. Mean length 

 regressions were developed for species with 10 or 

 more pairs of mean length data (>2 measurements) 

 in each size of trawl (Fig. 2). Catch regressions were 

 developed for those species that were present in at 

 least 20 samples in the large net and were repre- 

 sented by at least 5 samples with >20 individuals 

 in the small net. This insured a sufficiently wide 

 distribution to yield meaningful results (Fig. 3). 



Differences (P < 0.01) between day and night 

 regressions for each species were evaluated using 

 analysis of co variance (Snedecor and Cochran 1980). 



Results 



Small trawls can be used to obtain trend data on 

 mean lengths of species caught in offshore waters. 

 Relationships exist between the catch in the 4.3 m 

 trawl vs. the catch in the 12.2 m trawl. No signifi- 

 cant differences were found in the day-night regres- 

 sions of mean length for any species tested. There 

 was no difference in the day-night catch vs. catch 

 relationship for total organisms or Penaeus setiferus 

 but one did exist for Trachypenaeus sp. and Squilla 

 empusa. 



Mean lengths in the two trawls were directly cor- 

 related for all species that met criteria for regres- 

 sion analysis (Table 1). The regressions of the mean 

 length of fish caught in one net vs. the other for day 

 and night were not significantly different for any 



FISHERY BULLETIN: VOL. 84, NO. 4, 1986. 



981 



