562 
Fishery Bulletin 95(3), 1997 
have affected escape rates through the meshes by 
altering the shape of the codend. The percentage of 
fish and shrimp that escaped during trawling, as 
opposed to escaping during haulback, is also un- 
known. Watson et al. (1993) reported that most spe- 
cies escaped through escape openings during trawl 
haulback or when fish were crowded near the open- 
ings. Further diver evaluations are necessary to iden- 
tify methods by which fish and shrimp escape. 
Differences in catch rates of brown and white 
shrimp observed for many of the BRD’s may have 
been due to species-specific behavior or size differ- 
ences (or both). White shrimp swim more actively 
than brown shrimp during the day (Wickham and 
Minkler, 1975). The white shrimp caught by the con- 
trol nets were larger, on average, than the brown 
shrimp, a finding that is typical for Louisiana catches 
(Keithly and Baron-Mounce 2 ). In addition, the white 
shrimp caught in the spring were substantially larger 
than those in the fall. 
These data are derived from a fishery-independent 
study; results from commercial shrimping could dif- 
fer. Had the Eymard BRD been used in a larger trawl 
and without a hoop, the results might have been quite 
different. Many fish and shrimp may 
have been lost during haulback and 
although we had mechanical re- 
trieval, a larger commercial vessel 
may have had faster retrieval. Al- 
though we trawled near shrimp boats 
whenever possible, at times no 
shrimp boats were present in a sam- 
pling area. When this was the case, 
we began trawling in an area where 
shrimp had been caught previously; 
if few or no shrimp were caught, we 
moved to another area. Moving short 
distances (one or two km) could re- 
sult in very different catches. Be- 
cause of time and fuel limitations, 
however, movements of very long dis- 
tances were not feasible. Provided 
that shrimp were being caught, we 
did not relocate if large quantities of 
fishes or crabs were also present. In 
this situation, a shrimper would most 
likely relocate in an attempt to find 
more shrimp or cease shrimping un- 
til conditions in the area become 
more favorable. We found higher ra- 
tios of fish to shrimp when shrimp 
catches were low. Other studies have 
reported that bycatch ratios depend 
on shrimp abundance; when few 
shrimp are present, fishing times are 
longer and result in high catches of 
bycatch species (Adkins 7 ). The 20- 
minute tows used in our study were 
three to six times shorter than those 
typically used in commercial opera- 
tions. Longer tows would have neces- 
sitated decreasing the number of 
7 Adkins, G. 1989. A comprehensive as- 
sessment of bycatch in the Louisiana shrimp 
fishery. Final report to NMFS NA89WC- 
H-MF006. La. Dep. Wildlife Fish., Bourg, 
LA, 75 p. 
2,700 
1,800 
900 
0 
2,700 
1,800 
900 
V/// /. Control 
Device 
Authement - Ledet 
Lake Arthur 
Length class (mm) 
Figure 5 
Length-frequency distribution of shrimp collected by the four devices and cor- 
responding control nets. 
