716 



Fishery Bulletin 89(4), 1991 



UPPER COAST 



r = 398 

 P = 0051 

 n = 48 - i 



5- 10 fathoms 



4 

 3 

 2- 



r = 0.384 . 

 P = 00070 



10 - 15 fathoms 



LOWER COAST 



r = 0512 

 P = 0.0002 

 n = 48 



5-10 fathoms 



. V. 



t*V 



r = 0.327 

 P = 0.0232 



n = 48 



10 - 15 fathoms 



IV* 



2 3 



In (E + 



2 3 



In (E + 



Figure 2 



Scatter plots for significant (P<0.05) correlations between transformed 

 monthly sea turtle strandings, ln(S + 1), and transformed monthly shrimp 

 fishing effort, ln(E + 1), on upper coast (shrimp statistical subareas 17-18) 

 and lower coast (subareas 19-21) of the northwestern Gulf of Mexico 

 during 1986-89. 



(Murphy and Hopkins-Murphy 1989, Magnu- 

 son et al. 1990). In fact, sea turtles may con- 

 gregate in shrimping areas to feed on dis- 

 carded by catch (Shoop and Ruckdeschel 1982, 

 Ruckdeschel and Shoop 1988). However, no 

 cause-and-effect relationship between sea 

 turtle strandings and shrimping has been 

 demonstrated to date. 



It is noteworthy that our analyses detected 

 significant correlations despite the wide varia- 

 tion inherent in turtle stranding and fishing 

 effort data. These significant positive correla- 

 tions are circumstantial evidence of a linkage 

 between strandings and shrimping, but do not 

 demonstrate that the strandings were caused 

 by shrimping. Strandings and shrimping occur 

 year-round and both are strongly seasonal, 

 with peaks during warm months. The correla- 

 tions we detected are consistent with earlier 

 findings that incidental capture in shrimp 

 trawls is the major cause of sea turtle mortal- 

 ity associated with human activities, but it is 

 also recognized that other fisheries, dredging, 

 collisions with boats, oil-rig removal with 

 underwater explosives, entrainment in power 

 plants, and directed take contribute to sea 

 turtle mortality at sea (Magnuson et al. 1990). 



Interpretation of statistical relationships 

 between sea turtle strandings and shrimp 

 trawling activity is confounded by the dynam- 

 ics of waterborne transport of stressed, in- 

 jured, or dead turtles to stranding sites. 

 Surface currents, winds, waves, tides, and 

 scavengers, as well as conditions affecting the 

 buoyancy of turtles, can affect their transport 

 toward or away from shore (Murphy and 



