Limitations of Quantitative Studies 



Careful studies, such as those outlined above, provide 

 valuable information for particular rookeries. But they do not 

 permit generalizations because of the marked differences between 

 beaches. Another problem is that the line between natural and 

 unnatural mortality is often blurred, as few really natural 

 beaches are left. Furthermore, taking measures to reduce 

 particular sources of natural mortality does not necessarily 

 result in corresponding increases in the overall hatch rate, 

 because they may simply provide opportunities for other sources 

 of natural mortality to increase. At present no alternative to a 

 step by step approach is evident. Studies of particular beaches 

 should ideally extend over several seasons to take account of 

 year to year variability. Even then, until quantitative work on 

 attrition in the aquatic phase of the life cycle is undertaken, 

 estimating survivorship to maturity will not be possible by such 

 an approach. In Prick's (1976) study, 2 out of 26 green 

 hatchlings were taken by fish as they swam out to sea off 

 Bermuda; this was within just a few hours and <6.5 km from shore 

 but was by day. More extensive studies of this kind, even if 

 telemetry rather than swimming behind the hatchlings were used, 

 could probably be faulted as looking at predation in a situation 

 into which the experimenter had introduced artificialities, but 

 might nevertheless provide some interesting information. 



Some estimates of survivorship can be made from a 

 combination of information on age (derived with some assumptions 

 from size) and frequency of catching turtles in these different 

 age classes (Frazer 1987b) . These could be useful for comparing 

 survivorship in different places or at different times, but they 

 do not distinguish between natural and unnatural mortality, and 

 they only apply to particular age classes. Thus, for example, an 

 estimated 70% of the juvenile loggerheads off the southeast 

 United States survive each year (Frazer 1987b); much of the 

 mortality is from drowning in shrimping nets. 



Natural Mortality Estimated from Reproductive Output 



Nevertheless, one overriding generality provides some 

 guidance: natural mortality is very high. Moreover some order- 

 of- magnitude estimates can be obtained without studying 

 predation and destruction by abiotic factors directly. In a 

 stable population (if there is such a thing for turtles!) each 

 female must produce 2 successful offspring, assuming a 1:1 sex 

 ratio. According to Fowler (1979) "estimating from past records, 

 each female that nests on the Tortuguero beach lays several 

 thousand eggs, perhaps as many as 10,000 or more, during her 

 reproductive life time." Two successful breeding survivors out 

 of 10,000 give an attrition of 99.98% before breeding. Suppose 

 Fowler's figure for reproductive output is ten times too high. 

 If 1,000 eggs are laid in a lifetime, the mortality is 99.8%. 



256 



