In populations or species with low fitness (high genetic load), a new management strategy called genetic rescue has been advocated to help avoid extinction. In this strategy, unrelated individuals from another population are introduced into the population with low fitness in an effort to reduce genetic load. Here we present ten guidelines that can be used to evaluate when genetic rescue is a good management option, the appropriate procedures for genetic rescue planning and management, and the potential negative genetic consequences of genetic rescue. These guidelines are then used to evaluate the genetic rescue aspects of the recovery programs for the Mexican wolf and the Florida panther.

Although inbreeding can reduce individual fitness and contribute to population extinction, gene flow between inbred but unrelated populations may overcome these effects. Among extant Mexican wolves (Canis lupus baileyi), inbreeding had reduced genetic diversity and potentially lowered fitness, and as a result, three unrelated captive wolf lineages were merged beginning in 1995. We examined the effect of inbreeding and the merging of the founding lineages on three fitness traits in the captive population and on litter size in the reintroduced population. We found little evidence of inbreeding depression among captive wolves of the founding lineages, but large fitness increases, genetic rescue, for all traits examined among F1 offspring of the founding lineages. In addition, we observed strong inbreeding depression among wolves descended from F1 wolves. These results suggest a high load of deleterious alleles in the McBride lineage, the largest of the founding lineages. In the wild, reintroduced population, there were large fitness differences between McBride wolves and wolves with ancestry from two or more lineages, again indicating a genetic rescue. The low litter and pack sizes observed in the wild population are consistent with this genetic load, but it appears that there is still potential to establish vigorous wild populations.

Mexican and red wolves were both faced with extinction in the wild until captive populations were established more than two decades ago. These captive populations have been successfully managed genetically to minimize mean kinship and retain genetic variation. Descendants of these animals were subsequently used to start reintroduced populations, which now number about 40-50 Mexican wolves in Arizona and New Mexico and about 100 red wolves in North Carolina. The original captive Mexican wolf population was descended from three founders. Merging this lineage with two other captive lineages, each with two founders, has been successfully carried out in the captive population and is in progress in the reintroduced population. This effort has resulted in increased fitness of cross-lineage wolves, or genetic rescue, in both the captive and reintroduced populations. A number of coyote-red wolf hybrid litters were observed in the late 1990s in the reintroduced red wolf population. Intensive identification and management efforts appear to have resulted in the elimination of this threat. However, population reintroductions of both Mexican and red wolves appear to have reached numbers well below the generally recommended number for recovery and there is no current effort to re-establish other populations.

Hybridization and introgression are significant causes of endangerment in many taxa and are considered the greatest biological threats to the reintroduced population of red wolves (Canis rufus) in North Carolina (U.S.A.). Little is known, however, about these processes in red wolves and coyotes (C. latrans). We used individual-based simulations to examine the process of hybridization and introgression between these species. Under the range of circumstances we considered, red wolves in colonizing and established populations were quickly extirpated, persisted near the carrying capacity, or had intermediate outcomes. Sensitivity analyses suggested that the probabilities of quasi extinction and persistence of red wolves near the carrying capacity were most affected by the strength of two reproductive barriers: red wolf challenges and assortative mating between red wolves and coyotes. Because model parameters for these barriers may be difficult to estimate, we also sought to identify other predictors of red wolf population fate. The proportion of pure red wolves in the population was a strong predictor of the future probabilities of red wolf quasi extinction and persistence. Finally, we examined whether sterilization can be effective in minimizing introgression while allowing the reintroduced red wolfpopulation to grow. Our results suggest sterilization can be an effective short-term strategy to reduce the likelihood of extirpation in colonizing populations of red wolves. Whether red wolf numbers are increased by sterilization depends on the level of sterilization effort and the acting reproductive barriers. Our results provide an outline of the conditions likely required for successful reestablishment and long-term maintenance of populations of wild red wolves in the presence of coyotes. Our modeling approach may prove generally useful in providing insight into situations involving complex species interactions when data are few.

Abstract. We have evaluated a microsatellite measure proposed as an indicator of inbreeding and outbreeding using a captive wolf population with known inbreeding levels and founder sources. The measure, which is based on the difference in the repeat number for microsatellite alleles within an individual, was not more predictive of the known inbreeding coefficient than microsatellite heterozygosity (it was actually less predictive). We also found no support that the measure was predictive of the level of outbreeding. However, we could not determine if the measure was predictive of very low levels of inbreeding due to matings between remote relatives. Overall, it appears that the usefulness of this measure to identify individuals on the inbred-outbred continuum beyond that of heterozygosity and identify biologically important associations with fitness-related traits may be limited. We suggest that the measure be examined theoretically to determine when (and how much) the predictive value of the measure is different from that of heterozygosity for inbreeding or outbreeding levels in a variety of different scenarios.