Renaud et al.: Shrimp loss by TEDs in US coastal waters 



133 



funnels. There were 86 and 223 data pairs in the At- 

 lantic for Georgia TEDs with and without accelerator 

 funnels, respectively, and 186 pairs for Super Shooter 

 TEDs with funnels. Frequencies of data collection by 

 geographic area and season (winter: December-Feb- 

 ruary, spring: March-May, summer: June-August, 

 fall: September-November) are presented in Table 1. 



Performance of TED-equipped and standard nets Data 

 were collected from 5937 nets during the 2.5 yr study. 

 Frequency of net problems was tabulated by TED type. 

 The most frequent problems included clogging of the 

 net, twisting of trawl doors and cables, and torn web- 

 bing. In the Gulf of Mexico, no problems occurred dur- 

 ing 86%, 87%, 75%, and 87% of the tows for nets 

 equipped with Georgia TEDs with and without fun- 

 nels, Super Shooter TEDs and standard nets, respec- 

 tively (Table 2). In the Atlantic, the values were 96%, 

 90%, 89%, and 95% for the respective gear types 

 (Table 2). A variety of problems, including but not lim- 

 ited to those with trawl doors, cables, bogging-down of 

 nets, etc., were shown to be net-type independent (e.g., 

 TED-equipped nets or standard nets) in the Gulf of 

 Mexico and net-type dependent in the Atlantic (chi- 

 square, P<0.05). 



Testing of paired tows 



Reduction of shrimp CPUE associated with use of 

 TEDs Mixtures of brown and white shrimp were cap- 

 tured in all areas of the Gulf and Atlantic, except for 



the west coast of Florida where pink shrimp were preva- 

 lent. Shrimp species were not separated for analyses. 



There was no significant difference (P<0.05) in net 

 sizes among vessels in this study, so this parameter 

 was excluded from any further analyses. Summaries 

 of shrimp CPUEs by TED type, season, and area are 

 presented in Tables 3-5. Mean shrimp CPUEs for Geor- 



