FISHERY BULLETIN: VOL. 85, NO. 3 



Although these assumptions may not be satis- 

 fied in all cases, our estimates of probability of 

 capture in a given substation were consistent ex- 

 cept for the two discarded experiments conducted 

 during periods of low loggerhead turtle densities. 

 These findings suggest that some loggerheads en- 

 countering the trawl were able to avoid capture, 

 presumably by moving out of the trawl path. The 

 results also indicate that a consistent percentage 

 of loggerheads were captured by the trawl, facili- 

 tating the estimation of turtle abundance based 

 on number of turtles captured. It should be noted 

 that the probability of capture in a given substa- 

 tion (as presented in our results) is lower than the 

 probability of capture in a given tow. To compute 

 the probability of capture in a single tow, the 

 width of the substation is divided by the width of 

 the trawl and this factor multiplied by the proba- 

 bility of capture in the substation. 



Loggerhead turtle abundance estimates in the 

 Port Canaveral ship channel exhibited large sea- 

 sonal variation (Table 2). The estimated popula- 

 tion levels during the month of February were 

 significantly higher than all other quarterly sur- 

 veys indicating that loggerheads were most abun- 

 dant during winter months. These findings are in 

 agreement with other NMFS surveys in the 

 Canaveral channel from 1978 to 1983 (Table 4) 

 and support the contention of Carr et al. (1980) 

 that loggerhead turtles may hibernate in the Port 

 Canaveral channel in refuge from low water tem- 

 peratures. The fact that the winter of 1981-82 was 

 unusually mild, could account for the lack of an 

 early winter peak in loggerhead turtle abundance 

 observed in previous years. 



Data presented in Table 4, while of limited 



statistical value due to inconsistencies in sam- 

 pling methodologies, are useful for comparisons 

 between this study and other NMFS Canaveral 

 channel surveys. It is worthy of note that mean 

 catch per unit effort (CPUE) by month combining 

 all years was in excess of 10 loggerhead turtles/ 

 hour from November through March with peak 

 concentrations in February and March. Lowest 

 CPUE values and presumably population levels 

 occurred from April through September, which is 

 in agreement with our findings. 



It is evident that loggerhead turtle abundance 

 estimates were highly variable between surveys 

 made in the same quarter (Table 2). We speculate 

 that these fluctuations in population levels were 

 caused by short-term immigration and emigra- 

 tion in response to local weather changes. We 

 have observed daily changes in catch rates which 

 appear to be correlated with passage of weather 

 fronts. 



Distribution of loggerhead turtles within the 

 Port Canaveral ship channel is also of interest. In 

 every survey, stations 9, 10, and 11 exhibited the 

 highest abundances, suggesting that they were 

 preferred turtle habitat. Stations 7, 8, and 12 ex- 

 hibited intermediate population levels and sta- 

 tions 13 and 14 had low turtle abundance levels. 

 Stations 7, 8, 9, and 10 were those where deepest 

 cuts into the seabed have been made by dredging. 

 The bottom was characterized by divers as clay- 

 silt and detritus as opposed to the harder clay- 

 sand bottom outside the channel (Carr et al. 

 1980). 



Interpretation of loggerhead turtle abundance 

 estimates generated from this study is compli- 

 cated by the fact that three different groups of 



Table 4. — Summary of catch per unit effort (CPUE) of loggerhead turtles in the Port Canaveral ship channel (1978-83). Values are in 



turtles per hour standardized to a single 100-ft net. N = number of tows. 



452 



