Reintroductions are a common approach for preserving intraspecific biodiversity in fragmented landscapes; however, reintroduced populations are often smaller and more geographically isolated than native populations. Reintroductions may therefore exacerbate the reduction in genetic variability initially caused by population fragmentation due to the small effective population size of the reintroduced populations. Mixing genetically divergent sources is assumed to alleviate this issue by increasing genetic diversity, but the effects on genetic diversity are often not monitored and there are potential negative tradeoffs for mixing genetically distinct sources.

I examined the consequences of mixed-source reintroductions on the ancestral composition, genetic variation and fitness of a small stream fish, the slimy sculpin (Cottus cognatus), from three source populations at nine reintroduction sites in southeast Minnesota. I used microsatellite markers to evaluate allelic richness and heterozygosity in the reintroduced populations relative to computer simulated expectations. I then inferred the fitness of each crosstype in the reintroduced populations by comparing their overall persistence, growth rates, and relative body conditions. Finally, I modeled the response of fitness related variables in the reintroduced populations to variation in habitat using a combination of regression and ordination methods.

Ancestry analysis revealed that one of the three sources had more ancestors than the other two sources at most reintroduction sites. Sculpins in reintroduced populations exhibited higher levels of heterozygosity and allelic richness than the sources, but only slightly higher than the most genetically diverse source population. Simulations of maximum genetic variation indicated only a modest expected increase over the most diverse source. Growth rate, body size, and relative body condition suggest significantly reduced fitness in second generation hybrids. I detected evidence of local adaptation in the source populations based on greater predicted fitness for each source in its respective habitat. This local adaptation is strongly associated with a gradient in winter water temperatures. My study indicates that using more than one source for reintroductions may not substantially enhance genetic diversity. Furthermore, using multiple sources risks disruption of important adaptations and may cause outbreeding depression. Future reintroductions may be improved by evaluating the potential for local adaptation in ongoing reintroduction programs.