Parthenogenesis in California Condors: Impact on Genetic Variation

Ryder et al. (2021) recently reported the facultative parthenogenic origin of 2 California condors (Gymnogyps californianus). The 2 condors were descended from different fertile mothers that were housed with fertile males. Paternal contribution from the males was genetically excluded, and both chicks were homozygous (identical in state and presumably identical by descent) for all the markers inherited from their mothers, making their inbreeding coefficient equal to 1.0. Ryder et al. (2021) observed that 2 of 911 condors examined, or a proportion of 0.0022 (0.22%), were the result of parthenogenic reproduction. Given this surprising finding, it is important to examine the expected impact of parthenogenesis on the level of genetic variation in condors.

Intragametophytic selfing in plants, such as in homosporous ferns, also results in homozygosity for the whole genome and, along with the type of parthenogenesis seen in condors, is the most extreme form of inbreeding possible. Theory from intragametophytic selfing (Hedrick 1987a) can be used to examine the potential impact of partial parthenogenic reproduction in the condor population. Using this analogy, given there is a proportion of parthenogenesis (Par) and random mating (1−Par), the expected equilibrium proportion of heterozygosity at a locus is H_e = 2p₁p₂(1−Par) where p₁ and p₂ are the frequencies of the 2 alleles, A₁ and A₂, at a biallelic locus. Given that the observed proportion of parthenogenic reproduction is Par = 0.0022, then H_e/(2 p₁p₂) = 0.9978 = 99.78% or the expected heterozygosity is only slightly reduced from the 100% value when there is no parthenogenesis. In contrast, intragametophytic selfing often has a large impact on genetic variation (Hedrick 1987a), but that is because the intragametophytic selfing rate is often much higher than the rate of random mating. In condors, the opposite is true where, given that Par = 0.0022, the rate of random mating (1−Par = 0.9978) is 454 times the estimated proportion of parthenogenesis.

Furthermore, neither of the pathenogenetically produced male condor chicks successfully reproduced, and they are now deceased. Chick SB260 died at 718 days (1.96 years) before sexual maturity and chick SB517 died at 2915 days (7.98 years), was small, and had scoliosis (see Ryder et al. 2021 for other details). In other words, both these chicks had low fitness and were not reproductive, so that the observed rate of parthenogenesis in successfully reproducing condors was actually 0.0. Although these 2 chicks did not contribute genetically, it is possible that other unknown condor parthenotes could have survived and reproduced in the past.

California condors, like many other rare endangered species, are generally considered to have low genetic variation. A recent study comparing genomic variation in California condors, Andean condors, and the related, but much more common, turkey vultures found the 3 species had 1.26, 0.75, and 1.37 heterozygosity per kilobase, respectively (Robinson et al. 2021). In other words, the California condor had a level of variation nearly as high as that in the turkey vulture and significantly higher than the Andean condor. However, the proportions of the genome in ROH (runs of homozygosity) in California condors, Andean condors, and turkey vultures were >20%, 5.7%, and 4.2%, respectively (Robinson et al. 2021). The much higher level of ROH in California condors is likely due to inbreeding associated with the relatively small effective population size prior to being brought into captivity where close inbreeding has been avoided. This might include inbreeding due to parthenogenesis and subsequent successful reproduction of the condors produced by parthenogenesis. When there is parthenogenesis, whole chromosomes that are transmitted by the mother to her different progeny are identical, potentially resulting in much higher levels of ROH. If the ROH regions are excluded, then California condors have even higher heterozygosity than turkey vultures.

Parthenogenesis can also result in purging of detrimental genetic variation, a potentially positive effect (Hedrick 1994); however, this impact is a function of the proportion of parthenogenesis in the population (Hedrick 1987b) and is expected to be low when the proportion of parthenogenesis is low. For example, assume that genotype A₂A₂ is lethal so that the only viable parthenogenic offspring from an A₁A₂ female are A₁A₁. Therefore, the frequency of the A₂ allele is reduced from ½ in an A₁A₂ female to 0 in her surviving parthenogenic progeny, and the genetic load from lethal alleles is eliminated in the parthenogenic offspring from this individual (Hedrick 2007). Let us assume that the average number of lethal equivalents per individual is approximately 3 (Ralls et al. 1988) and that these are all in the form of lethal alleles. Given that the frequency of parthenogenesis is 0.0022 and that these offspring have no lethal alleles, the average number of lethal equivalents becomes [909(3) + 2(0)]/911 = 2.99. In other words, parthenogenesis in these 2 chicks has a very minimal effect on reducing the number of lethal equivalents in the population.

Facultative parthenogenesis might be advantageous when the population numbers are low, potential mates are rare, or when colonizing new areas, and Ryder et al. (2021) found facultative parthenogenesis in 2.4% of all the female condors in the breeding program. California condors have been through past population bottlenecks, situations potentially favoring parthenogenesis. As discussed above, low levels of parthenogenesis are expected to have minimal impacts on the levels of heterozygosity and detrimental genetic variation, but it might have a substantial impact increasing the level of ROH, as found in California condors.

Acknowledgments

I appreciate comments by Aurora Garcia-Dorado, Marty Kardos, and Bob Lacy.

References