Limits...
Landscape genetics of the nonnative red fox of California

View Article: PubMed Central - PubMed

ABSTRACT

Invasive mammalian carnivores contribute disproportionately to declines in global biodiversity. In California, nonnative red foxes (Vulpes vulpes) have significantly impacted endangered ground‐nesting birds and native canids. These foxes derive primarily from captive‐reared animals associated with the fur‐farming industry. Over the past five decades, the cumulative area occupied by nonnative red fox increased to cover much of central and southern California. We used a landscape‐genetic approach involving mitochondrial DNA (mtDNA) sequences and 13 microsatellites of 402 nonnative red foxes removed in predator control programs to investigate source populations, contemporary connectivity, and metapopulation dynamics. Both markers indicated high population structuring consistent with origins from multiple introductions and low subsequent gene flow. Landscape‐genetic modeling indicated that population connectivity was especially low among coastal sampling sites surrounded by mountainous wildlands but somewhat higher through topographically flat, urban and agricultural landscapes. The genetic composition of populations tended to be stable for multiple generations, indicating a degree of demographic resilience to predator removal programs. However, in two sites where intensive predator control reduced fox abundance, we observed increases in immigration, suggesting potential for recolonization to counter eradication attempts. These findings, along with continued genetic monitoring, can help guide localized management of foxes by identifying points of introductions and routes of spread and evaluating the relative importance of reproduction and immigration in maintaining populations. More generally, the study illustrates the utility of a landscape‐genetic approach for understanding invasion dynamics and metapopulation structure of one of the world's most destructive invasive mammals, the red fox.

No MeSH data available.


Related in: MedlinePlus

Genetic distances among 10 nonnative red fox sampling sites estimated from 13 microsatellite loci (n = 349 foxes), including (A) an unrooted neighbor‐joining tree based on Nei's DA with bootstrap support >65% indicated, and (B) principle coordinates (PCoA), including centroids and standard errors along two axes. Sampling sites were San Joaquin Valley (SJV) north (‐N) and south (‐S), Monterey (Mont), Morro Bay (MB), Santa Barbara (SB), Southern California (SO), Half Moon Bay (HMB), and the San Francisco Bay wetlands (SFB) south (‐S), east (‐E), and west (‐W). Samples were color‐coded for convenience to distinguish sites of the San Francisco Bay area (red), central coast (blue), inland (yellow), and south coast (green).
© Copyright Policy - creativeCommonsBy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4979706&req=5

ece32229-fig-0003: Genetic distances among 10 nonnative red fox sampling sites estimated from 13 microsatellite loci (n = 349 foxes), including (A) an unrooted neighbor‐joining tree based on Nei's DA with bootstrap support >65% indicated, and (B) principle coordinates (PCoA), including centroids and standard errors along two axes. Sampling sites were San Joaquin Valley (SJV) north (‐N) and south (‐S), Monterey (Mont), Morro Bay (MB), Santa Barbara (SB), Southern California (SO), Half Moon Bay (HMB), and the San Francisco Bay wetlands (SFB) south (‐S), east (‐E), and west (‐W). Samples were color‐coded for convenience to distinguish sites of the San Francisco Bay area (red), central coast (blue), inland (yellow), and south coast (green).

Mentions: Observed heterozygosity was lowest in the three San Francisco (SF) Bay Area sampling sites (0.51–0.52) and highest in the northern SJV (0.67), followed closely by Monterey and Half Moon Bay (0.64 each) sites (Table 2). Expected heterozygosity (ranging 0.55–0.71) and allelic richness (ranging 2.2–2.6) were less variable, with most estimates falling within 2 standard errors of others. Five sites showed no statistically significant heterozygote deficiency, whereas five sites exhibited FIS values significantly greater than zero (ranging 0.08–0.22), suggesting the presence of admixture. Genetic effective population size estimates ranged from 2 (95% CI: 1.6–2.5) in Morro Bay to 44.6 (28.9–73.0) in Monterey. All populations showed signatures of demographic bottlenecks, consistent with founder effects (Table 2). The neighbor‐joining tree based on Nei's DA indicated 3 sets of sites that clustered with moderate to high bootstrap support: (1) the SF Bay sampling sites; (2) Half Moon Bay and Monterey; and (3) Santa Barbara (SB) and SO (Fig. 3A). Additionally, within the SF Bay sampling sites, the West SF Bay and South SF Bay clustered together relative to the East SF Bay, consistent with a clockwise stepping‐stone pattern of founding around the Bay. Except for the SF Bay populations, the positioning of populations relative to one another were consistent with their spatial arrangement on the landscape and, although not well supported by bootstrapping, both SJV sampling sites clustered together in the final tree. The PCoA similarly grouped the three SF Bay sites together as relatively distinct from the other sites (Fig. 3B).


Landscape genetics of the nonnative red fox of California
Genetic distances among 10 nonnative red fox sampling sites estimated from 13 microsatellite loci (n = 349 foxes), including (A) an unrooted neighbor‐joining tree based on Nei's DA with bootstrap support >65% indicated, and (B) principle coordinates (PCoA), including centroids and standard errors along two axes. Sampling sites were San Joaquin Valley (SJV) north (‐N) and south (‐S), Monterey (Mont), Morro Bay (MB), Santa Barbara (SB), Southern California (SO), Half Moon Bay (HMB), and the San Francisco Bay wetlands (SFB) south (‐S), east (‐E), and west (‐W). Samples were color‐coded for convenience to distinguish sites of the San Francisco Bay area (red), central coast (blue), inland (yellow), and south coast (green).
© Copyright Policy - creativeCommonsBy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4979706&req=5

ece32229-fig-0003: Genetic distances among 10 nonnative red fox sampling sites estimated from 13 microsatellite loci (n = 349 foxes), including (A) an unrooted neighbor‐joining tree based on Nei's DA with bootstrap support >65% indicated, and (B) principle coordinates (PCoA), including centroids and standard errors along two axes. Sampling sites were San Joaquin Valley (SJV) north (‐N) and south (‐S), Monterey (Mont), Morro Bay (MB), Santa Barbara (SB), Southern California (SO), Half Moon Bay (HMB), and the San Francisco Bay wetlands (SFB) south (‐S), east (‐E), and west (‐W). Samples were color‐coded for convenience to distinguish sites of the San Francisco Bay area (red), central coast (blue), inland (yellow), and south coast (green).
Mentions: Observed heterozygosity was lowest in the three San Francisco (SF) Bay Area sampling sites (0.51–0.52) and highest in the northern SJV (0.67), followed closely by Monterey and Half Moon Bay (0.64 each) sites (Table 2). Expected heterozygosity (ranging 0.55–0.71) and allelic richness (ranging 2.2–2.6) were less variable, with most estimates falling within 2 standard errors of others. Five sites showed no statistically significant heterozygote deficiency, whereas five sites exhibited FIS values significantly greater than zero (ranging 0.08–0.22), suggesting the presence of admixture. Genetic effective population size estimates ranged from 2 (95% CI: 1.6–2.5) in Morro Bay to 44.6 (28.9–73.0) in Monterey. All populations showed signatures of demographic bottlenecks, consistent with founder effects (Table 2). The neighbor‐joining tree based on Nei's DA indicated 3 sets of sites that clustered with moderate to high bootstrap support: (1) the SF Bay sampling sites; (2) Half Moon Bay and Monterey; and (3) Santa Barbara (SB) and SO (Fig. 3A). Additionally, within the SF Bay sampling sites, the West SF Bay and South SF Bay clustered together relative to the East SF Bay, consistent with a clockwise stepping‐stone pattern of founding around the Bay. Except for the SF Bay populations, the positioning of populations relative to one another were consistent with their spatial arrangement on the landscape and, although not well supported by bootstrapping, both SJV sampling sites clustered together in the final tree. The PCoA similarly grouped the three SF Bay sites together as relatively distinct from the other sites (Fig. 3B).

View Article: PubMed Central - PubMed

ABSTRACT

Invasive mammalian carnivores contribute disproportionately to declines in global biodiversity. In California, nonnative red foxes (Vulpes vulpes) have significantly impacted endangered ground‐nesting birds and native canids. These foxes derive primarily from captive‐reared animals associated with the fur‐farming industry. Over the past five decades, the cumulative area occupied by nonnative red fox increased to cover much of central and southern California. We used a landscape‐genetic approach involving mitochondrial DNA (mtDNA) sequences and 13 microsatellites of 402 nonnative red foxes removed in predator control programs to investigate source populations, contemporary connectivity, and metapopulation dynamics. Both markers indicated high population structuring consistent with origins from multiple introductions and low subsequent gene flow. Landscape‐genetic modeling indicated that population connectivity was especially low among coastal sampling sites surrounded by mountainous wildlands but somewhat higher through topographically flat, urban and agricultural landscapes. The genetic composition of populations tended to be stable for multiple generations, indicating a degree of demographic resilience to predator removal programs. However, in two sites where intensive predator control reduced fox abundance, we observed increases in immigration, suggesting potential for recolonization to counter eradication attempts. These findings, along with continued genetic monitoring, can help guide localized management of foxes by identifying points of introductions and routes of spread and evaluating the relative importance of reproduction and immigration in maintaining populations. More generally, the study illustrates the utility of a landscape‐genetic approach for understanding invasion dynamics and metapopulation structure of one of the world's most destructive invasive mammals, the red fox.

No MeSH data available.


Related in: MedlinePlus