396 



Fishery Bulletin 102(2) 



D 



a. 

 O 



25 

 20 



5 - 



Caretta 



O 

 O 



o 



— r- 



— l— 

 15 



— r~ 

 25 



Temperature ( : C) 



Figure 2 



CPUE (number of captured turtles/1000 hooks I by sea surface temperature (°C) in 

 each set, by species. Circles = first trip, triangles = second trip, stars = third trip. In 

 each graph, the dashed vertical line, arbitrarily placed at 16.7 °C, marks a separation 

 between the temperatures in the first trip and those in the second and third trips 

 (except for one set in the first trip). Note that the two graphs have different vertical 

 scales, and that, in the construction of this figure, temperature measurements (but 

 not the CPUEsl have been jittered (see "Materials and methods" section i. 



post hoc test). Although leatherback sea turtles could not 

 be hauled aboard for measurements, on board observa- 

 tions and video recordings indicated that they were sub- 

 adult or adult animals. 



Most of the loggerhead turtles were hooked through 

 their mouths or esophagus, but a small number were 

 hooked through their flippers or were found to be simply 

 entangled in the lines. Loggerhead sea turtles taken 

 aboard had their hooks removed, sometimes in a care- 

 less way that caused severe injury, and they were then 

 returned to the sea. Leatherback sea turtles were found 

 entangled in the lines or hooked either through the flippers 

 or carapace or through the mouth. Because no leatherback 

 sea turtle was hauled aboard, we could not tell if any were 

 hooked in the esophagus. 



Discussion 



Achaval et al. (2000 ) reported data obtained from nine trips 

 aboard two different longline vessels operating within the 

 Uruguayan EEZ and in international waters in the South 

 Atlantic in different seasons of the year, and employing 

 different longline methods. Those authors reported that 

 28 loggerhead and 28 leatherback sea turtles were cap- 

 tured in 86 sets with 75.033 hooks in zones I and II, that 

 correspond approximately to the fishing area covered in 

 this study, yielding a CPUE of 0.37/1000 hooks for both 

 loggerhead and leatherback sea turtles. For loggerhead 

 sea turtles, there was a significant difference between our 

 CPUE (Table 1) and that of Achaval et al. (chi-square test, 

 X 2 =226.4, P<0.001 ); whereas for leatherback sea turtles no 

 significant difference was found i chi-square test, x 2 =1.97, 

 P=0.161). 



Although the variations in CPUE observed in our study 

 could be explained by differences in temperatures (Fig. 2), 

 other physical, spatial, or temporal factors (or a combina- 



tion of these factors) could be involved. The trips were car- 

 ried out at different times of the year (Table 1); the third 

 trip was more to the south and closer to the coast, and the 

 first trip had sets more to the east (Fig. 1). 



Our estimates of sea turtle mortality at capture may 

 be lower than the actual mortality rates from longlines 

 because our estimates do not consider postrelease deaths 

 derived from 1) wounds caused by hooks removed from 

 turtles on board, 2) embedded hooks and lines, and 3) 

 stress caused by capture itself. Other researchers have 

 also recognized that, because of factors such as these, 

 there is great uncertainty in the estimates of mortality 

 levels for sea turtles captured in longline gear (Balazs and 

 Pooley, 1994; Eckert, 1994). 



Captured loggerhead sea turtles were smaller (Table 2 1 

 than loggerhead sea turtles nesting in Brazil (minimum 

 CCL = 83.0 cm, average CCL= 103.0 cm, nesting season 

 1982-83 through nesting season 1999-2000; Projeto 

 TAMAR 3 ) and in several places in the North Atlantic and 

 the Caribbean (minimum CCL=75.4 cm, average CCL in 

 the range of 94.0-105.1 cm; Dodd, 1988). However, logger- 

 head sea turtles nesting in Cape Verde, in the northeast- 

 ern Atlantic, are smaller than those nesting in those other 

 places: minimum CCL = 68.0 cm, average CCL=82.9 cm. 

 data from 1998 (Cejudo et al., 20001. There is an overlap 

 between the observed CCL range and that of adult Cape 

 Verde loggerhead sea turtles (seven loggerhead turtles 

 out of 54 observed, or 13.0'7 f , had a CCL equal or greater 

 than the minimum Cape Verde CCL), but the average 

 CCL of the captured loggerhead sea turtles (Table 2 i was 

 well below that of loggerhead sea turtles nesting in Cape 

 Verde. We estimate that the captured loggerhead sea 

 turtles were generally juveniles, although a small number 

 of them could have been adult turtles. However, size is 



3 Projeto TAMAR. 2000. Unpubl. data 



Salvador, BA 40210-970, Brazil. 



Caixa Postal 2219. 



