Genetic diversity is one of the three forms of biodiversity recognized by the World Conservation Union ( IUCN ) as deserving conservation. The need to conserve genetic diversity within populations is based on two arguments: the necessity of genetic diversity for evolution to occur, and the expected relationship between heterozygosity and population fitness. Because loss of genetic diversity is related to inbreeding, and inbreeding reduces reproductive fitness, a correlation is expected between heterozygosity and population fitness. Long‐term effective population size, which determines rates of inbreeding, should also be correlated with fitness. However, other theoretical considerations and empirical observations suggest that the correlation between fitness and heterozygosity may be weak or nonexistent. We used all the data sets we could locate (34 ) to perform a meta‐analysis and resolve the issue. Data sets were included in the study, provided that fitness, or a component of fitness, was measured for three or more populations along with heterozygosity, heritability, and/or population size. The mean weighted correlation between measures of genetic diversity, at the population level, and population fitness was 0.4323. The correlation was highly significant and explained 19% of the variation in fitness. Our study strengthens concerns that the loss of heterozygosity has a deleterious effect on population fitness and supports the IUCN designation of genetic diversity as worthy of conservation.

Understanding the consequences of inbreeding has important implications for a wide variety of topics in population biology. Although it is often stated in the literature that the deleterious effects of inbreeding (inbreeding depression) are expected to be more pronounced under stressful than benign conditions, this issue remains unresolved and controversial. We review the current literature on the relationship between the magnitude of inbreeding depression and environmental stress and calculate haploid lethal equivalents expressed under relatively benign and stressful conditions based on data from 34 studies. Inbreeding depression increases under stress in 76% of cases, although this increase is only significant in 48% of the studies considered. Estimates of lethal equivalents are significantly greater under stressful (mean ¼ 1.45, median ¼ 1.02) than relatively benign (mean ¼ 0.85, median ¼ 0.61) conditions. This amounts to an approximately 69% increase in inbreeding depression in a stressful vs a benign environment. However, we find strong lineage effects to be ubiquitous among studies that examine inbreeding depression in multiple environments, and a prevalence of conditionally expressed deleterious effects within lineages that are uncorrelated across environments. These results have important implications for both evolutionary and conservation biology. Heredity (2005) 95, 235-242.

Inbreeding-environment interactions occur when inbreeding leads to differential fitness loss in different environments

The ability of populations to undergo adaptive evolution depends on the presence of quantitative genetic variation for ecologically important traits. Although molecular measures are widely used as surrogates for quantitative genetic variation, there is controversy about the strength of the relationship between the two. To resolve this issue, we carried out a meta-analysis based on 71 datasets. The mean correlation between molecular and quantitative measures of genetic variation was weak (r = 0.217). Furthermore, there was no significant relationship between the two measures for life-history traits (r = -0.11) or for the quantitative measure generally considered as the best indicator of adaptive potential, heritability (r = -0.08). Consequently, molecular measures of genetic diversity have only a very limited ability to predict quantitative genetic variability. When information about a population's short-term evolutionary potential or estimates of local adaptation and population divergence are required, quantitative genetic variation should be measured directly.

No abstract