Predicting Annual Fecundity from Nest Success

Abstract

Nest success, the probability that a nest will fledge offspring, is a widely measured parameter in ornithological studies, and is usually estimated by monitoring the fates of nests throughout a breeding season. Because nest success es- timates are commonly reported in the avian literature, a key question in population studies is how to derive annual fecun- dity rates, F (a vital parameter in population biology), from nest success data when breeding individuals are not marked. In this manuscript, we describe a simple, recursive model that generates an estimate of the frequency distribution of an- nual fecundity rates that can be achieved for a species, given precise and unbiased estimates of nest success, the average number of young that fledge per successful nest, and three life history parameters: the maximum number of possible suc- cessful broods per breeding season (J), the maximum number of possible nesting attempts per breeding season (K), and the maximum brood size (B). We illustrate the model for 3 hypothetical species in which the average young that fledge per successful nest is 2 offspring: (1) a single-brooded species in which J =1, K = 3, B = 3; (2) a double-brooded species in which J = 2, K = 3, B = 3, and (3) a triple-brooded species in which J =3, K = 3, B = 3. In general, the frequency distributions of acceptable fecundity solutions for single-, double-, and triple-brooded species are all approximately symmetric, and are defined by Fave (the average acceptable solution), Fmin (the minimum acceptable solution), and Fmax (the maximum acceptable solution). The "breadth" of these distributions, or the difference between Fmin and Fmax, appears to be controlled largely by solutions where an unequal number of young fledge across attempts. However, when examining relationships between annual fecundity and nest success across the full spectrum of nest suc- cess values, we find that, more often than not, non-linear relationships characterized the association between nest success, Fmax, Fmin, and Fave. The only case where nest success predicts annual fecundity in a linear fashion occurred when J = K. Thus, the assumption that nest success is an adequate indicator of annual fecundity should be viewed with caution.