Effects of brood size on multiple-paternity rates: a case for ‘paternity share’ as an offspring-based estimate

• Published in: Animal behaviour Vol. 78; no. 2; pp. 563 - 571
• Main Authors: Eccard, Jana A., Wolf, Jochen B.W.
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01-08-2009
• Subjects: binomial probability distribution; Biologi; Biology; clutch size; confidence limits; extrapair paternity; litter size; multiple paternity; multiple-male mating; NATURAL SCIENCES; NATURVETENSKAP; paternity share; multiple-male mating; paternity share; clutch size; confidence limits; multiple paternity; litter size; binomial probability distribution; extrapair paternity
• ISSN: 0003-3472; 1095-8282; 1095-8282
• DOI: 10.1016/j.anbehav.2009.04.008

Molecular paternity analysis has opened a new perspective to our understanding of mate choice and the evolution of mating systems. Genetic paternity data can now be produced with relative ease, but the interpretation of the data can constitute a challenge. Particularly in multiple paternity measures of promiscuous species with small broods, such as most birds and mammals, stochastic effects are easily mistaken for differences in mating behaviour. Two major issues can be isolated. First, the popular measure 'multiple paternity rate' is brood-based and the probability of multiple paternal broods stochastically depends on the brood size. This makes 'multiple paternity rate' a poor measure for comparisons across species with different brood sizes. We conducted a representative review of recent literature on paternity in mammals demonstrating that such problematic comparisons are rather frequent. Second, comparisons among studies are made without confidence intervals of the respective estimates. Using a simple binomial model, we illustrate the stochastic problems inherent in the estimation of paternity rates. We propose an offspring-based, alternative measure, the 'paternity share' to overcome these problems. This measure is formally analogous to the popular proportion of 'extra-pair young' (pooled over broods) in studies of extra-pair paternity in socially monogamous birds, but in the case of multiple paternity cannot directly be taken from the empirical data. We provide methods for calculating 'paternity share' and establishing confidence intervals. 'Paternity share' will make comparisons of multiple paternity in promiscuous mating systems across species or studies more reliable.