Condition-Adjusted Estimator of Reproductive Output in Snakes

A potential problem exists with the use of relative clutch mass (RCM = total clutch mass/female postparturient mass) as an estimate of reproductive output. When examining RCM among individuals or subsamples such as years, sites, age classes etc., one must assume that the mass/length relationship is equal for each individual or subsample. When this assumption is violated, interpretations of reproductive output through RCM may be systematically biased. This paper describes the potential bias in the interpretation of reproductive output through RCM and proposes three alternative estimators for this trait. I argue that the first of these estimators (ovulational residuals) is logically the estimate of relative reproductive output among females or groups, short of monitoring the calories accrued and allocated to reproduction. Ovulational residuals are the residuals from the multiple regression of total clutch mass on both female length and female mass at ovulation, i.e., when investment in reproduction is determined. However, mass at ovulation is not often available, especially for viviparous species because of the gestation period. Therefore, like RCM, the two other proposed alternative estimators (postparturient residuals and RCMI), require postparturient female mass in their calculation, rendering them generally more practical. Correlations between the ovulational residuals and each of the three more practical estimators of relative reproductive output (RCM, postparturient residuals, and RCM1) indicate that RCM1, a linear adjustment of RCM by postparturient condition, most closely approximates the logically best estimate of output (ovulational residuals). Data from a four-year field study of reproduction in Nerodia sipedon at two locations are analyzed using both RCM and RCM1. Additionally, RCM1 values are estimated for four prior studies with snakes, and conclusions resulting from both RCM and RCM1 are compared. Conclusions resulting from RCM1 differ from those resulting from RCM in ways that suggest that RCM1 more accurately represents the reproductive biology of these systems. Additional work is needed, especially with oviparous species, to determine the general validity and applicability of these methods.