McEACHRON ET. AL: BYCATCH IN TROTLINE FISHERY 



Figure 3.— Texas coast. 



Table 1 —Number of fishes caught on trotlines in the upper and 

 lower Laguna Madre during February 1985-January 1986. 



were pooled for both bay systems because an in- 

 sufficient number of spotted seatrout were caught 

 for individual bay analyses. Factor levels elimi- 

 nated from length analyses were leaves and win- 

 ter from hardhead catfish, leaves and crab from 

 spotted seatrout, and fall, spring, and summer 

 from black drum. Spring and winter red drum 

 lengths were pooled. Lower Laguna Madre data 

 only were used for red drum, black drum, and 

 hardhead catfish length analyses. Each measured 

 fish length was an observational unit of a trotline 

 set. Sets were a random factor nested within fixed 

 main effect combinations. The nested set effect 

 mean square was used for testing other effects 

 when the set effect was significant. However, the 

 AOV yields approximate F values because un- 

 equal numbers of fish were caught among sets. 



SAS procedures (SAS Institute, Inc. 1980, 1982) 

 were used for all analyses. The significance level 

 for each AOV test was a = 0.01 because the AOV 

 used to examine catch rates of each species had 15 

 potential F tests. This alpha value assured that 



111 



