Patterns of Performance Variation Between Animal Hybrids and their Parents: A Meta-analysis

Hybridization is a widespread phenomenon in animals, and hybrid heterosis/breakdown could be key processes determining the evolutionary dynamics of hybrids. Indeed, hybrids are not consistently disadvantaged compared to the parental lineages, as was historically assumed. Multiple processes could lead to performance differences between parental lineages and their hybrids. Despite many studies evaluated the performance of hybrids, a quantitative synthesis is required to assess the general pattern. Here we used meta-analytic and meta-regression approaches to quantify the fitness differences between parental lineages and their hybrids, and to identify possible processes that could lead to these differences. Specifically, we tested biological and methodological parameters that could determine differences in performance between hybrids and parental lineages. Hybrid performance was extremely variable across studies, being often significantly higher or lower compared to the mean performance of their parents. Nevertheless, the averaged hybrid performance was similar to the fitness of parental lineages, with differences across studies related to how performance was assessed. Genetic divergence between parental lineages, and the approach used to identify hybrids were the parameters most strongly related to variation in hybrid performance. Performance was lower for hybrids between distantly related lineages. Furthermore, study settings and the use of imprecise approaches for hybrid identification (e.g. morphology-based) can bias assessments of performance. Studies performed on wild populations and using genetic approaches for hybrid identification detected more often a decreased hybrid performance, compared to laboratory studies. We highlight the importance of appropriate settings for a realistic understanding of the evolutionary impacts of hybridization.