238 



Fishery Bulletin 94(2). 1996 



Gulf until 1990 (McDonald, 1990; Hen- 

 wood et al., 1992; Crouse, 1993b; Weber 

 et al. 1 ; Crouse et al. 2 ). Therefore, 1986- 

 89 can be considered pre-TED years, for 

 the most part. In contrast, 1990-93 

 were years during which TED regula- 

 tions were in effect. 



Caillouet et al. (1991) examined the 

 statistical relationship between monthly 

 sea turtle stranding rates (species com- 

 bined) and monthly shrimp fishing in- 

 tensities in the northwestern Gulf of 

 Mexico during 1986-89. They detected 

 significant statistical relationships be- 

 tween stranding rates and fishing in- 

 tensities in some offshore depth inter- 

 vals within to 15 fathoms (fm) (27.4 

 m) in two geographic zones of the north- 

 western Gulf: the upper coast defined 

 as shrimp statistical subareas 17-18, 

 and the lower coast defined as subar- 

 eas 19-21 (Fig. 1; see Kutkuhn, 1962). 



In this paper, we used the approach 

 of Caillouet et al. (1991) to determine 

 whether or not the statistical relation- 

 ships they detected between monthly 

 sea turtle stranding rates and monthly 

 shrimp fishing intensities during 1986- 

 89 continued to exist after TED regula- 

 tions had been promulgated. We tested 

 the null hypothesis that the statistical 

 relationship between monthly sea turtle stranding 

 rates and fishing intensities in two geographic zones 

 of the northwestern Gulf was the same in 1990-93 

 as in 1986-89. The expectation was that regulations 

 requiring use of TED's in shrimp trawls during 1990- 

 93 would reduce the incidental catch of sea turtles 

 and thereby diminish or eliminate the statistical re- 

 lationship between sea turtle stranding rates and 

 fishing intensities. 



Materials and methods 



A data set (subareas 17-21) containing 2,445 sea 

 turtle records was obtained from the Sea Turtle 



Louisiana 

 A 



Mermen xau 

 River 



Gulf of Mexico 



- s 

 Mexico 



Figure 1 



Boundaries of the upper coast (shrimp statistical subareas 17-181 and 

 lower coast (subareas 19-21). and ten depth intervals of the northwestern 

 Gulf of Mexico (see Kutkuhn. 1962; Patella, 1975). 



1 Weber, M.. D. Crouse, R. Irvin. and S. Indicello. 1995. Delay 

 and denial: a political history of sea turtles and shrimp fishing, 

 46 p. Center for Marine Conservation, 1725 DeSales St. NW. 

 No. 500, Washington, DC, 20036. 



2 Crouse, D. T., M. Donnelly, M. J. Bean, A Clark. W. R. Irvin. 

 and C. E. Williams. 1992. The TED experience: claims and 

 reality, 17 p. A report by the Center for Marine Conservation, 

 Environmental Defense Fund, and National Wildlife Federa- 

 tion. Center for Marine Conservation, 1725 DeSales St. NW. 

 No. 500. Washington, DC, 20036 



Stranding and Salvage Network (STSSN) headquar- 

 ters, National Marine Fisheries Service (NMFS) 

 Miami Laboratory, Miami, Florida. It contained 

 records not only of sea turtle strandings but also of 

 turtles caught or entangled in commercial and rec- 

 reational fishing gears. 



Stranded sea turtles, whether live or dead, exhibit 

 no known external or internal signs of capture in 

 shrimp trawls. To our knowledge, only mark-recap- 

 ture studies have provided direct evidence that some 

 tagged sea turtles captured incidentally in shrimp 

 trawls have become stranded (e.g. Manzella et al., 

 1988; Fontaine et al., 1989). When examination of 

 stranded sea turtles provides evidence of other causes 

 of death, then capture in shrimp trawls may be ruled 

 out; however, the remaining strandings still lack 

 definite explanation. Strandings that had no expla- 

 nation were the focus of our analyses. 



Prior to our analyses, we deleted 890 records (419 

 from 1986 to 1989 and 471 from 1990 to 1993) based 

 on codes in data fields named LAT, LONG, NOTE1, 

 NOTE2,... NOTE6, HEADSTART, and TYPEREP 

 (Table 1). The following categories of records were 

 deleted: 



