The negative effects of inbreeding on fitness are serious concerns for populations of endangered species. Reduced fitness has been associated with lower genome-wide heterozygosity and immune gene diversity in the wild; however, it is rare that both types of genetic measures are included in the same study. Thus, it is often unclear whether the variation in fitness is due to the general effects of inbreeding, immunity-related genes or both. Here, we tested whether genome-wide heterozygosity (20 990 SNPs) and diversity at nine immune genes were better predictors of two measures of fitness (immune response and survival) in the endangered Attwater's prairie-chicken (Tympanuchus cupido attwateri). We found that postrelease survival of captive-bred birds was related to alleles of the innate (Toll-like receptors, TLRs) and adaptive (major histocompatibility complex, MHC) immune systems, but not to genome-wide heterozygosity. Likewise, we found that the immune response at the time of release was related to TLR and MHC alleles, and not to genome-wide heterozygosity. Overall, this study demonstrates that immune genes may serve as important genetic markers when monitoring fitness in inbred populations and that in some populations specific functional genes may be better predictors of fitness than genome-wide heterozygosity.

The primary goal of captive breeding programmes for endangered species is to prevent extinction, a component of which includes the preservation of genetic diversity and avoidance of inbreeding. This is typically accomplished by minimizing mean kinship in the population, thereby maintaining equal representation of the genetic founders used to initiate the captive population. If errors in the pedigree do exist, such an approach becomes less effective for minimizing inbreeding depression. In this study, both pedigree- and DNA-based methods were used to assess whether inbreeding depression existed in the captive population of the critically endangered Attwater's Prairie-chicken (Tympanuchus cupido attwateri), a subspecies of prairie grouse that has experienced a significant decline in abundance and concurrent reduction in neutral genetic diversity. When examining the captive population for signs of inbreeding, variation in pedigree-based inbreeding coefficients (f(pedigree)) was less than that obtained from DNA-based methods (f(DNA)). Mortality of chicks and adults in captivity were also positively correlated with parental relatedness (r(DNA)) and f(DNA), respectively, while no correlation was observed with pedigree-based measures when controlling for additional variables such as age, breeding facility, gender and captive/release status. Further, individual homozygosity by loci (HL) and parental rDNA values were positively correlated with adult mortality in captivity and the occurrence of a lethal congenital defect in chicks, respectively, suggesting that inbreeding may be a contributing factor increasing the frequency of this condition among Attwater's Prairie-chickens. This study highlights the importance of using DNA-based methods to better inform management decisions when pedigrees are incomplete or errors may exist due to uncertainty in pairings.

Captive breeding programs often use a pedigree to identify breeding pairs that maintain genetic diversity and limit inbreeding. However, unintentional breeding of closely related individuals can occur when errors exist in the pedigree and may subsequently result in inbreeding depression. In this study, a DNA-based approach was used to identify parentage assignment errors in the captive pedigree of the critically endangered Attwater's prairie-chicken Tympanuchus cupido attwateri, and tested whether clutch survival increased using DNA-relatedness values for assigning breeding pairs instead of the pedigree and mean kinship. Parentage assignment error was observed in each year surveyed ranging from 2.4 to 7.3%. After correcting identified errors prior to the 2013 breeding season, 11 of 38 assigned pedigree-based breeding pairs still possessed DNA-based relatedness coefficients (r DNA ) ≥0.125 suggesting that additional errors remained in the pedigree. Two approaches were used to prevent breeding among close relatives in 2013 and 2014. Assigned pedigree-based breeding pairs in 2013 that possessed r DNA ≥0.125 were reassigned unrelated breeding partners, while all individuals in 2014 were used to identify the minimum overall r DNA for the breeding population without reference to the pedigree other than to verify founder representation. Both years resulted in a significant reduction in mean parental relatedness among chicks (P < 0.001) and a significant increase in the proportion of the clutch surviving to 5 weeks post-hatch (P ≤ 0.006) compared to 2012 when breeding pairs were assigned using only the pedigree. No significant difference in the proportion of the clutch surviving was observed between 2013 and 2014 (P > 0.300). These results have important implications for the captive management of endangered species, and highlight the importance of periodically evaluating for pedigree errors. To what degree pedigree errors limit fitness in other endangered species captive populations deserves further attention.