Pedigree pitfalls and solutions
One common assumption of pedigrees is that the founding individuals
(hereafter, founders) are equally unrelated. As many threatened
populations have experienced significant declines before populations are
pedigreed, founders from these populations may include more variance in
kinship amongst one another than individuals randomly sampled from
non-threatened populations. For instance, a study of
microsatellite-based relatedness in kākāpō (Strigops habroptilus )
demonstrated that full-sibling and half-sibling relationships were
represented amongst founders (Bergner, Jamieson, & Robertson, 2014),
which has been confirmed with thousands of genome-wide SNPs (Department
of Conservation, unpublished data ). Research in Tasmanian devil
has further shown that inferring estimates of relatedness amongst
founders using molecular genetic and geographic data reveals
significantly higher inbreeding coefficients in the years following
establishment, compared to pedigrees that treat founders as equally
unrelated (Hogg et al., 2019). While these effects are ameliorated with
higher pedigree depth from founding individuals (Balloux et al., 2004),
this assumption can be perpetuated when individuals of unknown ancestry
(e.g., supplemented individuals, or individuals with missing
information) are incorporated into the pedigree in later generations. In
addition to these issues, pedigrees are also susceptible to human
transcription errors, which can compound in pedigrees over time
(Hammerly, de la Cerda, Bailey, & Johnson, 2016).
Missing information is often a concern for wild pedigrees where
parentage is difficult to ascertain. For example, accurate parentage can
be challenging due to the time and cost required to monitor breeding
individuals in the wild, coupled with the breeding behaviour of the
organism (e.g., polygamous species like flock breeding birds,
herd-breeding mammals, and open-pollinated plants; Ashley, 2010; Ivy,
Putnam, Navarro, Gurr, & Ryder, 2016; Wildt et al., 2019). This
challenge is exacerbated when individuals lose their identifiers (e.g.,
radio collars, tags, or leg bands; Milligan, Davis, & Altizer, 2003) or
when socially monogamous individuals participate in extra-pair parentage
(e.g., Overbeek et al., 2020). Even in captivity, parentage amongst
polygynous breeders is difficult to track. Individuals with unknown
parentage are either excluded from the potential breeding population due
to kinship uncertainty, are assigned a multiple parentage average
kinship value (i.e., MULTs; Lacy, 2012), or are assigned hypothetical
parentage which reconstructs an assumed pedigree (Ballou et al., 2010).
Building MULTs into a pedigree has been shown to retain greater genetic
diversity while still maintaining inbreeding avoidance as opposed to
removing individuals of unknown parentage, as shown in Arabian oryx
(Oryx leucoryx ; Putnam & Ivy, 2014). However, for species that
experience large differences in reproductive success between
individuals, the MULT approach has been shown to both over- and
under-estimate individual genetic contributions (e.g., Tasmanian devil;
Farquharson, Hogg, & Grueber, 2019). To highlight the severity of such
challenges in group breeding species, over 50% of ungulate species
managed under the auspices of the Association of Zoos and Aquariums
Species Survival Plan® (SSP) have less than half of
their pedigree known; in fact, only 15% have completely known pedigrees
(R.M.Gooley and E.K.Latch, unpublished data ). This is a common
challenge encountered with group-housed captive populations, which can
lead to management challenges with respect to kinship calculations,
breeding recommendations, and genetic diversity retention (Hauser,
Galla, Steeves, & Latch, Preprint ). Resolving unknown parentage
and generating accurate kinship values would lead to more effective
conservation management of intensively managed wild and captive
populations.
To address pedigree challenges, empirical estimates of relatedness from
DNA and geographic data can be used to complete and complement
pedigrees. For relatively diverse wild populations, microsatellite-based
approaches can be informative (McLennan et al., 2018), especially for
inferring close relatives. For example, a recent study from Moran et
al., (2021) highlighted the benefit of using microsatellite markers to
verify parentage in critically endangered California condor
(Gymnogyps californianus ). While microsatellites maintain use,
high density single nucleotide polymorphisms (i.e., SNPs) generated
through high throughput sequencing approaches (i.e., HTS) often provide
better resolution for estimating identity-by-descent, even when
populations are inbred or relationships are more distant (Taylor 2015;
Flanagan et al., 2019; Galla et al., 2020). Empirical data have been
used to enhance pedigrees for use in conservation management, including
estimating founder relationships (Hogg et al., 2019) and reconstructing
pedigrees (Gooley, Hogg, Belov, & Grueber, 2017; McLennan et al., 2018)
in Tasmanian devils and guiding metapopulation management of pedigreed
and non-pedigreed populations of sable antelope (Hippotragus
niger ) and dama gazelle (Nanger dama ; R.M.Gooley Personal
comm. ). Molecular data can also be used to validate uncertain and
semi-wild pedigrees. For example, microsatellite markers have recently
been used to validate pedigrees for intensively-monitored and critically
endangered wild populations of kakī/black stilt (Himantopus
novaezelandiae ; Overbeek et al., 2020) and black robin (Petroica
traversi; Forsdick, Cubrinovska, Massaro, & Hale, 2021).
Fellow conservation practitioners and researchers have queried whether
time and resources should be dedicated towards building pedigrees from
scratch and maintaining them, considering their caveats and the
abundance of HTS data available to produce estimates of relatedness that
are comparable to pedigrees (Speed & Balding, 2015). In this
perspective, we argue that pedigrees remain an invaluable tool in the
conservation genomics era, providing an affordable way to estimate
relatedness and inbreeding (Nietlisbach et al., 2017, but see also
Kardos, Luikart, & Allendorf, 2015), enhance genomic inquiry, and
capture important metadata that genomic data alone cannot (Fig. 1).
Based on our collective experience, we assert that gathering the
behavioural and ecological data that underlie pedigrees not only
advances our understanding of inter-individual relationships, but also
develops strong ties between conservation geneticists, practitioners,
and beyond (see below). In the conservation genomics era, we posit that
the greatest conservation genomic advances will happen when we invest in
both pedigree and genomic data.