Evaluating hybrid speciation and swamping in wild carnivores with a decision‐tree approach

Hybridization is an important evolutionary force with a principal role in the origin of new species, known as hybrid speciation. However, ongoing hybridization can create hybrid swamping, in which parental genomes are completely lost. This can become a biodiversity threat if it involves species that have adapted to certain environmental conditions and occur nowhere else. Because conservation scientists commonly have a negative attitude toward hybrids, it is important to improve understanding of the influence of interspecific gene flow on the persistence of species. We reviewed the literature on species hybridization to build a list of all known cases in the order Carnivora. To examine the relative impact, we also noted level of introgression, whether fertile offspring were produced, and whether there was mention of negative or positive evolutionary effects (hybrid speciation and swamping). To evaluate the conservation implications of hybrids, we developed a decision‐making tree with which to determine which actions should be taken to manage hybrid species. We found 53 hybrids involving 68 unique taxa, which is roughly 23% of all carnivore species. They mainly involved monophyletic (83%) and sympatric species (75%). For 2 species, the outcome of the assessment was to eliminate or restrict the hybrids: Ethiopian wolf (Canis simensis) and Scottish wildcat (Felis silvestris silvestris). Both species hybridize with their domestic conspecifics. For all other cases, we suggest hybrids be protected in the same manner as native species. We found no evidence of genomic extinction in Carnivora. To the contrary, some species appear to be of hybrid origin, such as the Asiatic black bear (Ursus thibetanus) and African golden wolf (Canis lupaster). Other positive outcomes of hybridization are novel genetic diversity, adaptation to extreme environments, and increased reproductive fitness. These outcomes are particularly valuable for counterbalancing genetic drift and enabling adaptive introgression in a human‐dominated world.


Species and boundaries
What defines a species is still being debated, and scientists have come up with 26 different species concepts thus far, none of which are comprehensive (Avise & Ball, 1990;Hausdorf, 2011).The most commonly accepted biological species concept describes a species as a group of actually or potentially interbreeding populations that are reproductively isolated from other such groups (Mayr, 1988).This puts a strong emphasis on barriers to gene flow, which are not always easy to determine between allopatric populations.Rather than complete isolation, populations more commonly display an isolation-by-distance pattern across their range (Rieseberg et al., 2004).Furthermore, interspecific gene flow remains common in the wild, even between genetically and morphologically different species (Petit & Excoffier, 2009).Darwin (1859) suggested that speciation can be maintained in the face of gene flow because it results from adaptation to different spatial, temporal, and sexual environments.Many evolutionary biologists now recognize that species boundaries can be semipermeable, with varying degrees of genetic divergence and exchange (Harrison & Larson, 2014;Mallet, 2008;Wu, 2001).
Even though species, here referred to as stable interbreeding populations with their own genetic signature, may hybridize, they are typically maintained in the wild through various mechanisms.First, prezygotic reproduction barriers often evolved as a consequence of selection against maladaptive interspecific matings, such as behavioral differences in mate choice, temporal segregation of mating season, differential communication signals, karyotypic differences, and gametic incompatibility (Mayr, 1963;Ramiro et al., 2015).Second, species divergence is often the result of ecological niche partitioning, which implies that intermediate, hybrid phenotypes have reduced fitness and are often selected against (Gabryś et al., 2021).This type of postzygotic barrier thereby contributes to the maintenance of distinct taxa.As a gray zone in the speciation process, there has been much debate about how to classify species that have not yet reached complete prezygotic or postzygotic reproductive isolation (Crispo et al., 2011).This is of particular interest at species boundaries, where diverged taxa meet and potentially hybridize (Harrison & Larson, 2014).Even though hybridization can be viewed as a natural process, some conservationists argue that it poses a threat to biodiversity (Wolf et al., 2001).

Hybrid swamping and speciation
Hybridization can become a conservation threat when humaninduced changes have facilitated secondary contact between unique or rare species adapted to certain climatic conditions (Allendorf et al., 2001).Gene flow between locally adapted taxa, referred to as introgression, can disrupt coadapted gene complexes and beneficial combinations of alleles, including chromosomal rearrangements, genome expansions, and epigenetic changes (Baack & Rieseberg, 2007).Furthermore, hybrids could have reduced fitness relative to the parental taxa, referred to as outbreeding depression (Edmands, 2007).This is particularly relevant for allopatric species, which may lack prezygotic reproduction barriers (Crispo et al., 2011).Furthermore, if hybrids are fertile and exhibit no strong reduction in fitness, they may displace pure lineages of the parental taxa, referred to as hybrid swamping.Ongoing hybridization can create hybrid swarms, in which parental genomes become completely lost, known as genomic extinction (Rhymer & Simberloff, 1996).Hybrid swarms often result from secondary contact with domestic animals or invasive species (Mooney & Cleland, 2001;Pereyra, 2016;Randi, 2008).Although hybridization with domestic animals is normally perceived as a biodiversity threat (e.g., Gabryś et al., 2021;Randi, 2008;Rhymer & Simberloff, 1996), some argue that it is not always a detriment and could provide adaptive advantages (Anderson et al., 2009;Coulson et al., 2011;Pilot et al., 2021;Schweizer et al., 2018).
As opposed to genetic swamping, interspecific hybridization may also lead to beneficial combinations of alleles, thereby offering a selective benefit (Garant et al., 2007;Otis et al., 2017).If hybrid fitness is high, hybrid swarms could even lead to the origin of unique lineages over time (Nolte & Tautz, 2010).When the insertion of new gene variants leads to an increase in fitness, it is referred to as adaptive introgression (Seehausen, 2004).Hybrids may also be able to fill ecological niches that were previously unoccupied, referred to as bounded-hybrid superiority (Moore, 1977).As a result, transgressive segregation could occur, which is the creation of extreme phenotypes in hybrid populations outside the phenotypic limits of both parental taxa (Bell & Travis, 2005), which may enable range expansion.Hybridization could also increase genetic diversity and thus overcome the risks of genetic drift and differentiation in small populations (Beninde et al., 2018).As such, hybridization could be an important evolutionary force, with a principal role in the origin of new species known as hybrid speciation (Barton, 2001;Mallet, 2007).Although there are many examples of hybrid speciation in plants (Hegarty & Hiscock, 2005, and included references), examples in animals are rare (Mavárez & Linares, 2008).

Conservation implications
The relative impact of hybrid speciation and swamping in the wild is still poorly understood, even though it has important implications for species conservation (Hirashiki et al., 2021;Mallot, 2008).It is generally believed that distinct lineages should not be admixed and that hybrids could jeopardize the existence of species (Allendorf et al., 2001).Therefore, they are often not granted legal protection.For instance, the wildcat (Felis silvestris) is highly protected under European legislation, whereas hybrids with domestic cats (Felis catus) are legally allowed to be killed, even though they make up the majority of the wildcat population in some areas (Fredriksen, 2016).The best documented example of species hybridization and its management implications is that of gray wolves (Canis lupus) and coyotes (Canis latrans) in North America (Bohling & Waits, 2015;Fredrickson & Hedrick, 2006;Rutledge et al., 2012).In the Alligator River National Wildlife Refuge, coyotes were captured and sterilized to avoid introgression with eastern wolves (Gese & Terletzky, 2015).Eastern wolves (Canis lupus lycaon) and red wolves (Canis lupus rufus) were considered native species and have received enormous amounts of conservation funding, often at the cost of the gray wolves (Gese & Terletzky, 2015).Whole-genome data show that eastern and red wolves are in fact hybrids of gray wolves and coyotes and should therefore not be protected differently from purebred individuals (VonHoldt et al., 2016).
That hybrids should be conserved is still widely disputed (Chan et al., 2019).A line is normally drawn between natural and human-mediated hybridization (Allendorf et al., 2001;Genovart, 2009), although this distinction is not always straight forward.Many species have recently been expanding their range through natural dispersal, enabled by climate change (Wayne & Shaffer, 2016), increasing the chance that allopatric species meet.For instance, hybridization between golden jackals (Canis aureus) and wolves is not entirely natural because jackals only recently expanded their range into Europe (Trouwborst et al., 2015).Furthermore, high human-induced mortality is known to increase hybridization of these 2 species (Moura et al., 2014).To advise conservationists on how to legally protect hybrid populations, Wayne and Shaffer (2016) proposed a decision-tree framework based on the following 3 questions: is admixture a natural occurrence between 2 native species; are hybrids ecological surrogates for endangered species that provide similar or identical ecosystem functionality and result in similar community interactions; and will restoration of the native habitat for endangered species select for native alleles and decrease the fraction of genomic contributions from the nonendangered species?If the answers to these questions are all yes, they advise legal protection of the hybrid.

Research aims
Due to the commonly negative attitude of conservation scientists toward hybrids, it is important to improve understanding of their occurrence in the wild and their influence on the genetic structure of species.We examined how hybrid speciation (the creation of taxa) and hybrid swamping (the loss of taxa) influence species diversity by determining how often hybridization has been observed in wild carnivores; identified instances of hybrid speciation and hybrid swamping; and examined the conservation value of wild hybrids.To evaluate the conservation value of or threat from hybrids and link our findings to potential management implications, we developed a decision-making tree (Figure 1), extended from Wayne and Shaffer (2016).We conducted a literature search to find all known cases of species hybridizations in the order Carnivora.The comparative parameters we considered in this search included range distribution and phylogenetic relationship and limited our review to carnivores so as not to extend the scope beyond what we could realistically

Literature research
To develop a list of hybrid species, we searched Google Scholar and Web of Science for peer-reviewed articles published from 1971 to 2022 that discussed cases of hybridization or interspecific gene flow in wild carnivores.We considered that hybridization can be based on field observations or morphological evidence, whereas interspecific gene flow requires some genetic evidence of introgression.We used the following search strings: "common species name*" OR "scientific family name*" AND "keyword$" OR "keyword$" (*retrieves words with variant zero to many characters, $ retrieves 0 or 1 character), for instance, ("fox*" OR "canidae*") AND ("hybridization$" OR "introgression$").The keywords were hybridization, introgression, interspecific gene flow, transgressive segregation, genomic extinction, hybrid speciation, and hybrid swarm.Some families have multiple common names (for instance, wolves and foxes).We entered these in separate search strings to increase our chances of finding pertinent papers (Appendix S1).In total, we had 29 search strings.Peer-reviewed articles were also drawn from the citations in articles selected from the initial search.We did not list a hybrid twice (i.e., from different locations), but instead included the best scientifically documented case for each species.To keep the results relevant for the decision-making tree, we listed only hybridization events that occurred recently (roughly in the past 100 years) and discarded ancient introgression (that occurred millions of years ago) and introgression with ancestral species (e.g., Lopes et al., 2021).All hybridizations were added to a database and visualized on a world map if exact locations were available.

Comparative parameters
For all hybridization events added to our database, we recorded several parameters, when such information was available.First, we noted the suspected cause of the hybridization, which can be natural (e.g., dispersal, male mounting, and species boundary) or human mediated (e.g., climate change, translocation, and invasive species).We also distinguished among allopatric (geographically separated ranges), parapatric (bordering ranges), and sympatric species (overlapping ranges).Theory predicts that allopatric species have an increased risk of hybridization because the strength of prezygotic barriers will be lower, due to the lack of selection against maladaptive interspecific mating (Crispo et al., 2011).We also compared hybridizations between monophyletic (containing all descendants of a particular node in a phylogenetic tree), paraphyletic (containing some, but not all, descendants of a particular node), and polyphyletic taxa (containing the most recent common ancestor).Paraphyletic and polyphyletic species are suspected to have more prezygotic barriers, as a result of karyotype and genome size differences (Stelzer et al., 2011), due to which hybridization is less likely.For each hybridization event, we also noted level of introgression, whether fertile offspring were produced, and whether there was mention of negative or positive evolutionary effects (hybrid swamping, outbreeding vs. hybrid vigor, speciation).When authors imply there has been hybrid speciation or swamping, it does not mean they presented evidence of such; rather, it is more an indication of human perception and attitude toward hybrids.Furthermore, not all hybrids are conclusive because introgression and incomplete lineage sorting are often difficult to tell apart (Wang et al., 2018).Generally, we applied the theory that when hybrids occur only in certain areas, genetic introgression following hybridization can be considered more likely than incomplete lineage sorting (Rapson et al., 2012).

Decision-making tree
To provide conservationists with guidelines on how to act in the case of wildlife hybridizations, we developed a decisionmaking tree for use in analyzing hybrid individually.The decision tree includes 7 statements on whether hybrids are a natural occurrence or derived from anthropogenic influences, involve domesticated or endangered species, cause a reduced fitness in hybrids, and are ecological surrogates and whether ecological restoration could select for native alleles.We used the tree to determine whether carnivore hybrids that occur in the wild should be protected in the same manner as purebred native species, temporarily protected while the frequency of native alleles in the hybrid zone will increase, or actively eliminated or restricted.The approach weights the conservation status of species, the ecological role of admixed individuals, and the relative impacts of hybrid speciation and swamping.

RESULTS
We found 53 hybrids, involving 72 unique taxa or 68 taxa when domestic hybrids were excluded (Figure 2).More hybridizations occurred between monophyletic species (83%) than paraphyletic (9%) or polyphyletic (8%) species.Two paraphyletic and 1 polyphyletic species produced fertile offspring (Appendix S1).In 33 hybrids, evidence of introgression was presented.In hybrid zones where the frequency of hybrids in a population was measured, it averaged 22% (1-85%).The majority of hybrids was between sympatric species (75%), followed by parapatric (21%) and allopatric (4%) species.Most hybrids occurred in the Caniformia suborder (n = 40), of which the Mustelidae (n = 10) and Canidae families (n = 9) were particularly well represented.Species boundaries (n = 27) and male mounting (n = 6) were the only natural causes of hybridization in the wild.For other causes, harvesting (n = 5) and hybridization after domestication (n = 5) were most often mentioned, followed by population declines (n = 3), species introductions (n = 3), habitat loss (n = 1), and climate change (n = 1).Most hybrid events were considered neutral (47%) with regard to their effect on the conservation status of species.In 14 papers, hybridization was considered a positive effect of the admixture event, such as a selective benefit (hybrid vigor) or hybrid speciation, whereas in 14 other papers a negative effect was suspected, such as loss of fitness (outbreeding) or hybrid swamping.
We applied our decision tree to the 53 hybrids and concluded that 46 hybrids should be protected in the same manner as native species (blue in Figure 2), 4 hybrids needed temporary protection (yellow), and 3 hybrids needed active elimination or restriction (red).Thirty-eight hybridizations were a natural occurrence between 2 native, sympatric species (statement 1) (Table 1).Six hybridizations were not between sympatric species but were nonetheless considered natural because they did not derive from recent anthropogenic influences (statement 2).Four hybridizations ended at statement 4, asking whether they include an endangered species or subspecies that occurs nowhere else.For 2 hybridizations, active elimination or restriction was considered advisable as a result: Ethiopian wolf (Canis simensis), which hybridized with domestic dogs, and the Scottish wildcat (Felis silvestris silvestris), which hybridized with domestic cat.For the endemic Cozumel dwarf coati (Nasua nelsoni), which hybridized with the introduced white-nosed coati (Nasua narica), we consider that hybrids did not have reduced fitness (statement 5) and that ecological restoration would select   FIGURE 2 Global carnivore hybrids: (a) locations of known hybrids (pictures from International Union for the Conservation of Nature and iNaturalist libraries; credit for each photo in Appendix S1) (blue, hybrids need to be protected in the same manner as purebred native species; yellow circle, hybrids are temporarily protected while the frequency of native alleles increases; red circles, hybrids should be eliminated or restricted to avoid genetic swamping; numbers within circles described in Table 1) and (b) summary statistics of all carnivore hybridizations included in this study (n = 53).
for native alleles (statement 7).For this species and hybrids between European polecat (Mustela putorius) and domestic ferret (Mustela furo), which form ecological surrogates (statement 6), outcome II was the end point (temporarily protect hybrids while the frequency of native alleles in the hybrid zone will increase).

DISCUSSION
Until recently, hybridizations among wild species were considered rare and a threat to conservation, due to outbreeding depression and the loss of genetically pure taxa through hybrid swamping (Edmands, 2007).Genomic tools have advanced and have illustrated that hybridizations are not only common, but also play a crucial role in speciation (Abbott et al., 2016).In this literature review, we found that recent hybridizations have occurred between 68 wild carnivores, which is roughly 23% of all carnivore species.Many carnivore species remain understudied, which means that some hybrids may be unnoticed.Nonetheless, the percentage is much higher than previous calculations, for which it was estimated that hybridization influences about 10% of animal species (Arnold, 1997).This higher percentage illustrates that hybridization in the wild is common.Even though it is now recognized as a widespread phenomenon, there is still much debate around the consequences for species conservation and management, particularly pertaining to endangered species and results of human-mediated processes (Crispo et al., 2011;Stronen et al., 2012;Wayne & Shaffer, 2016).

Species and boundaries
The nature of hybridizations differs, for instance, whether it involves sympatric or allopatric species.The majority of hybridizations (75%) were between sympatric species.There were only 2 cases of allopatric species hybridization, both of which resulted from reintroductions on islands.Hybridization was common in marine mammals, where it often results from violent interspecific sexual behavior by males, causing substantial introgression throughout the evolutionary history of Otariidae and Phocidae (Berta & Churchill, 2012;Miller et al., 1996).On Macquarie and San Miguel Islands, there are even multiple species hybridizing, making these places somewhat of a melting pot (Lancaster et al., 2006;Miller et al., 1996).Land mammals more often hybridized at the species boundary.For instance, kit fox (Vulpes macrotis) and swift fox (Vulpes velox) admix in a narrow hybrid zone in New Mexico (Mercure et al., 1993), and badgers (Meles meles and Meles leucurus) have a stable hybrid zone running along their species distributions in the Volga-Kama region in Russia (Kinoshita et al., 2019).Two Neotropical cats (Leopardus guttulus and Leopardus geoffroyi) and 2 lynx species (Lynx canadensis and Lynx rufus) also admix at the species boundary, where 2 different environments meet (Koen et al., 2014;Trigo et al., 2014).Habitat-dependent selection might be a main driver of stable hybrid zones, where the habitat differs from types that are normally associated with each of the parental species (Trigo et al., 2014).In some areas, hybrid frequencies reach up to 34-50% (Gaubert et al., 2005;Trigo et al., 2014;Zhigileva et al., 2014), suggesting a selective benefit to the intermediate genotype.
Occasional hybridization does not always lead to genetic exchange, which means the frequency of hybridizations may overestimate the rate of introgression (Harrison & Larson, 2014;Mallet, 2005).Introgression is more likely to occur between recently evolved species, which tend to have less complex prezygotic or postzygotic reproductive barriers (Crispo et al., 2011).Matings between paraphyletic species often produce sterile offspring, which means they pose no risk of hybrid swamping (Mayr, 1988).We found that 83% of the hybridizations were between monophyletic species.We found 2 paraphyletic hybrids that produced fertile offspring, and remarkably, 1 polyphyletic hybrid.The latter occurred between a South American fur seal (Arctocephalus australis) and a South American sea lion (Otaria byronia) in Uruguay (Franco-Trecu et al., 2016).That paraphyletic or polyphyletic species produce fertile offspring can mean 2 things.First, it may indicate colinearity in amino acid sequences, coupled with ancient hybridization (Moritz et al., 1987).Second, introgression between paraphyletic or polyphyletic species could imply that the taxonomic status of the species has not been completely resolved.For instance, smooth-coated otters (Lutrogale perspicillata) hybridize with small-clawed otters (Aonyx cinereus) in Singapore, which is presented as evidence that L. perspicillata should be included in the Amblonyx to avoid paraphyly in the genus (Moretti et al., 2017).Likewise, white-nosed coatis (Nasua narica) hybridize with western mountain coatis (Nasuella olivacea), and it has been suggested that the latter species be placed in the genus Nasua (Ruiz-García et al., 2021).

Hybrid speciation and swamping
Naturally occurring hybrid zones can be used to investigate the relative forces of hybrid speciation and swamping (Adavoudi & Pilot, 2022).Hybrid swamping is typically considered disruptive to the genetic integrity of species and local adaptation, whereas hybrid speciation generates genetic diversity and could make species better adapted to local environments (Abbott et al., 2016).In carnivores, 14 articles suggested hybrid swamping, of which 4 involved domestic species and 14 hybrid speciation.For instance, hybrid swamping was thought to occur in marten (Martes) and mink (Mustela) species, for which hybrid zones may act as evolutionary sinks (Colella et al., 2019;Kyle et al., 2003;Lodé et al., 2005).Colella et al. (2019) calculated that genetic homogenization would reduce North American marten's (Martes caurina) mitochondrial diversity by half and eliminate 13 unique nuclear variants.The long-term effect of the loss of these rare variants is unknown.The threat of hybrid swamping was also raised by Kelly et al. (2010), who counted at least 34 possible hybridizations in the Arctic and suggested that these are partly caused by climate change.However, most hybrids were anecdotal, and little evidence was presented with regard to gene introgression or threats to species survival.In fur seals (Arctocephalus spp.), hybrid swamping was thought to have resulted from overharvest (Lancaster et al., 2006).The upland seal (Arctocephalus forsteri snaresensis) is a subspecies that was suspected to have gone extinct due to hybrid swamping (Lento et al., 1997).However, the existence of this species and the subsequent genomic extinction have since been debated (Salis et al., 2016).To our knowledge, there is no known case of genomic extinction in the order Carnivora.In a previous assessment, based on the International Union for the Conservation of Nature (IUCN) Red List database, hybrid swamping seems to have contributed to the extinction of 11 out of 120,369 species (Draper et al., 2021).However, in none of these cases was hybridization the leading cause of extinction.
To the contrary, the positive outcomes of hybridization seem much more common, such as increased genetic diversity, novel adaptive variation, and hybrid speciation (Adavoudi & Pilot, 2022).Gene flow between species has been an important evolutionary force in the bears (Ursidae), causing rapid adaptive radiation (Kumar et al., 2017).For instance, polar bear (Ursus maritimus) and brown bear (Ursus arctos) hybridize commonly on the ABC islands (Cahill et al., 2015), and the Asiatic black bear (Ursus thibetanus) is thought to be of hybrid origin (Zou et al., 2022).Interestingly, although brown bears contain up to 8.8% polar bear ancestry, it does not occur the other way around (Cahill et al., 2015).This unidirectional introgression is fairly common in other species (Karlsson et al., 2014) and could provide interesting insights into adaptation to extreme environments (Cahill et al., 2015).For instance, refugial divergence and introgressive hybridization allowed the stoat (Mustela erminea) to adapt to higher elevations and concurrent climates (Colella et al., 2018).Interspecific gene flow also shaped the evolution of Canis species.It has been suggested that the African golden wolf (Canis lupaster) arose from hybridization between the gray wolf (Canis lupus) and Ethiopian wolf.The red wolf and eastern wolf, once considered unique species, seem to be the result of natural hybridization between wolf and coyote (VonHoldt et al., 2016).In North Carolina, red wolves that hybridize with coyotes seem to prefer admixed mates, and the average number of pups is higher in hybrid litters than purebred red wolf litters (Bohling & Waits, 2015).In sable marten (Martes zibellina), there is a correlation between heterozygosity and sexual maturity rate, further suggesting a selective benefit of hybrids (Kashtanov et al., 2003).

The Anthropocene
That the creation of novel genetic diversity through introgression can gain a selective benefit over parental groups is particularly relevant for species' survival in the Anthropocene (Rutledge et al., 2012).There is no denying that the world is highly transformed and fragmented and that return to the natural state is unrealistic.Instead of hanging onto the genetic integrity of species, it could be considered that introgression, whether natural or not, may allow species to adapt to a humandominated world (Crispo et al., 2011;Stronen et al., 2012).For instance, removing all wolf-coyote hybrids from the landscape could interfere with a driving force behind speciation that provides genetic and phenotypic material necessary for adaptation (Abbott et al., 2013;Stronen & Paquet, 2013).Hybridization can also occur between wild and farmed conspecifics that have escaped from captive breeding facilities, which has happened with foxes (Vulpes spp.) (Noren et al., 2005), minks (Kidd et al., 2009), and polecats (Etherington et al., 2022).Although this has been of great concern to conservationists, no loss in reproductive or adaptive fitness has been observed yet.Even when selective breeding has occurred on farms (e.g., for fur color), it is expected natural habitat and mate choice will select against the non-native alleles (Sacks et al., 2011;Wayne & Shaffer, 2016).Introgression with invasive species may even be beneficial in some cases, for example, when it increases the genetic variability in wild species experiencing a severe bottleneck or inbreeding depression (Feulner et al., 2013;Fulgione et al., 2016).
Introgressive hybridization between domestic animals and their wild relatives can also have extensive effects on phenotypic traits, which are widely perceived as a biodiversity threat (e.g., Gabryś et al., 2021;Randi, 2008;Rhymer & Simberloff, 1996).However, we argue that this very much depends on the circumstances.Wolves and dogs hybridize throughout their range and thereby provide an opportunity to assess the effects of socalled human-induced evolution (Hedrick, 2013).Introgression between wolves and dogs can, in rare cases, provide significant adaptive advantages for both groups by counterbalancing the negative effects of genetic drift and domestication (Pilot et al., 2021).For instance, introgressed domestic dog genes responsible for the black color morph in wolves are thought to offer a heterozygote advantage and selective benefit and to have enriched the genetic legacy of natural wolf populations (Anderson et al., 2009).Additionally, dog-derived immune genes have a positive effect on life span and reproductive success (Coulson et al., 2011;Schweizer et al., 2018).Although domestication does not necessarily lead to genetic erosion, some dog breeds carry a significant amount of deleterious mutation load, which will have the opposite effect (Cruz et al., 2008).Generally, hybridization with domestic animals only becomes a significant threat to biodiversity when excessive genetic swamping threatens to eradicate a species or subspecies that occurs nowhere else (Gómez et al., 2015).This is the situation in the welldocumented case of the Scottish wildcat, in which the majority (85%) of the critically endangered population is introgressed with domestic cat, forming a hybrid swarm (Howard-McCombe et al., 2021).Although purebred species may be lost in the process, it has been argued that hybrid swarms could give rise to unique hybrid lineages (Nolte & Tautz, 2010).

Decision-making tree
We found that most of the hybrids (72%) occurred between native, sympatric species.The only hybridizations considered a threat based on the decision tree were between the Ethiopian wolf and domestic dog and between the Scottish wildcat and domestic cat.The Ethiopian wolf is listed as endangered by the IUCN (197 remaining adults) and has a restricted range in the Ethiopian highlands, where it is threatened by farming and human persecution (Marino & Sillero-Zubiri, 2011).Microsatellite analysis confirm 7 hybrid individuals with domestic dogs, and about 17% of the sampled individuals were phenotypically abnormal (Gottelli et al., 1994).Genome-wide data also show gene flow with African golden wolf (Gopalakrishnan et al., 2018).Considering the small population size of Ethiopian wolves, and presumed loss of fitness, hybridization forms a serious threat to their persistence (Gottelli et al., 1994).For this species, the decision tree led to a recommendation of active elimination or restriction of hybrid species.The same was considered to threaten the Cozumel dwarf coati, which hybridizes with white-nosed coati that were introduced from the mainland (Cuarón et al., 2004).However, introgression does not appear to be common, and ecological restoration will likely decrease the genomic contribution from the non-native alleles (Flores-Manzanero et al., 2022).This is not the case for Scottish wildcats (Howard-McCombe et al., 2021).Because hybrids fill the same ecological niche as Scottish wildcats, the decision tree showed that the conservation policy surrounding this cat should be revised, in a modern rewilding framework (Soulé & Noss, 1998).Hybrids of Scottish wildcat are not legally protected, although original plans to actively cull them have been cancelled and replaced with a trap-neuter-release program, which is less efficient and more expensive, but was chosen because it was supported by the public (Fredriksen, 2016).
Removing or restricting non-native species or hybrids that are firmly established in the wild is not easy and comes with financial burdens and ethical concerns (Stronen & Paquet, 2013).For example, in an intensive effort to save the red wolf from hybrid swamping, the U.S. Fish and Wildlife Service (USFWS) adopted a Red Wolf Adaptive Management Plan, under which coyotes in an area where they naturally occur were killed or sterilized (Gese & Terletzky, 2015).The program did not reduce introgression, and it was concluded that the red wolf population is nonviable on its own (VonHoldt et al., 2016).Nonetheless, the USFWS continues its efforts despite considerable annual financial and social costs.
Attitudes and conservation goals among stakeholders differ, and stakeholders often present conflicting scientific or historical records (Schlaepfer et al., 2011).For instance, some argue that management should be aimed at preserving natural, genetically pure populations (Donfranseco et al., 2019), whereas others believe hybrids should be conserved if they have greater evolutionary fitness (Donfranseco et al., 2019).Some scientists argue that the ecological role of admixed individuals should be valued regardless of their genetic identity, even though a general uncertainty remains as to what can be considered a good ecological surrogate (Donfranseco et al., 2019).Even experts in the field of conservation genetics still struggle to agree on which management actions are required to deal with hybridization in the wild, particularly when they are driven by anthropogenic disturbance, such as wolf-dog hybridization (Hindrikson et al., 2012;Rand, 2008).We conclude that active elimination or restriction of hybrids is hardly ever necessary.
Our results show that interspecific gene flow is a common phenomenon in wild carnivores and that, generally, hybridization should not be viewed as a threat to the genetic integrity of unique taxa.Conservation of wild carnivores should establish a strategy that allows species to increase their genetic variation and adapt to future environmental change, which hybridization could help achieve.Only when hybrids include an endangered or subspecies that occurs nowhere else, do we advise resorting to active elimination or restriction.We believe our decisiontree approach provides conservation managers with simple guidelines on how to proceed in the case of hybridization.

FIGURE 1
FIGURE 1Decision-making tree to determine management action for wild animal hybrid species.

a
More parameters and common names are in Appendix S1. b Statements are described in text and Figure 1.Key: I, protection of hybrid individuals in the same manner as purebred native species; II, temporarily protect hybrid individuals while frequency of native alleles in the hybrid zone increases; III, active elimination or restriction of movement of individuals with non-native ancestry.

TABLE 1
Recommendation for management of wild carnivore hybrids based on a decision-tree approach.a.