Niche conservatism—the retention of ecological traits across space and time—is an emerging topic of interest because it can predict responses to global change. The conservation of Grinnellian niche characteristics, like species‐habitat associations, has received widespread attention, but the conservation of Eltonian traits such as consumer–resource interactions remains poorly understood. The inability to quantify Eltonian niches through space and time has historically limited the assessment of Eltonian niche conservatism and the dynamics of foraging across populations. Consequently, the relative influence of endogenous factors like phylogeny versus exogenous features like environmental context has rarely been addressed. We tested Eltonian niche conservatism using a paired design to compare foraging among four populations of American martens Martes americana and Pacific martens Martes caurina, morphologically and ecologically similar sister taxa that are allopatrically distributed throughout western North America. We developed a three‐stage isotopic framework and then quantified dietary niche overlap between the sister species and paired island‐mainland sites to assess the relative influence of endogenous (i.e., species) versus exogenous (i.e., environment) factors on Eltonian niches. First, we calculated pairwise dietary overlap in scaled δ‐space using standard ellipses. We then estimated proportional diets (“p‐space”) for individuals using isotopic mixing models and developed a novel utilization distribution overlap approach to quantify proportional dietary overlap. Lastly, we estimated population‐level proportional diets and quantified the differential use of functional prey groups across sites. We detected no pairwise overlap of dietary niches in δ‐space, and distributions of individual diets in p‐space revealed little overlap in core diets across populations. All pairwise comparisons of individuals revealed significant differences in diet, and population‐level comparisons detected contrasting use of functional prey groups. We developed a multi‐faceted isotopic framework to quantify Eltonian niches and found limited evidence of Eltonian niche conservatism across carnivore populations. Our findings are consistent with the growing recognition of dietary plasticity in consumers and suggest that consumer–resource dynamics are largely driven by exogenous environmental factors like land cover and community composition. These results illustrate the context‐dependent nature of foraging and indicate consumer functionality can be dynamic. A http://onlinelibrary.wiley.com/doi/10.1111/1365-2435.13266/suppinfo is available for this article.

Pacific martens ( Martes caurina ) are often associated with mature forests with complex structure for denning, resting, and efficient hunting. Nonetheless, a small isolated population of the Humboldt subspecies of Pacific martens ( Martes caurina humboldtensis ) occupies a narrow strip of young, coastal forest (< 70 years old) but not inland mature forest in the central Oregon Coast Range. We examined factors contributing to this unexpected distribution of martens by 1) analyzing marten diets using DNA metabarcoding to assess 90 scats, 2) using camera traps to assess differences in the relative abundances of prey, competitors, and predators across a coastal to inland gradient of vegetation types, and 3) quantifying differences in extent of fruit-producing shrubs and vegetation structure within vegetation types. Diets of martens were diverse (12, 10, and 3 species of birds, mammals, and amphibians respectively), and most fall and winter scats contained fruit. Voles, mice, and varied thrushes ( Ixoreus naevius ) were dominant prey items. Voles, mice, and most birds, but not varied thrushes, were more commonly observed in the coastal shrub-dominated forest than in inland forest. The coastal shrub-dominated forest had the highest diversity of vertebrates and potential prey overall. Bobcats ( Lynx rufus ), a key potential predator, were more commonly detected in inland forest. Of potential competitors, western spotted skunks ( Spilogale gracilis ) were more commonly detected in inland forest, with gray foxes ( Urocyon cinereoargenteus ) and raccoons ( Procyon lotor ) detected almost exclusively in coastal forests. Vegetation in coastal forests appears to provide, at least seasonally, more prey and fruit, and more overhead shrub cover compared with inland forest. Remaining plausible hypotheses for the restricted distribution of marten to coastal forests include increased prey, fruit, and overhead cover, and reduced predation risk from bobcats.

Simplifying stand structure to reduce fuel density is a high priority for forest managers; however, affects to Pacific marten (Martes caurina) movement and connectivity are unknown. We evaluated whether thinning forests to reduce fuels influenced movements of Pacific marten. We collected movement paths from 22 martens using global positioning system telemetry to evaluate habitat selection and describe movement patterns. We quantified motion variance, speed, and path sinuosity in 3 stand types that differed in structural complexity (i.e., complex [dense], simple [thinned], and open). We hypothesized marten movement would differ between stand types and predicted that 1) martens would select stand types with increased structural complexity (complex > simple > open); 2) movements would increase in complexity (sinuosity, motion variance) and decrease in speed when martens traveled through stands with increased structural complexity; 3) speeds would increase during summer, indicating increased movement during the breeding season; and 4) males would move more rapidly because of their larger home ranges. Martens traveled 0.5-27.2 km/day and an average (SD) of 1.4 (0.4) km/hour. Martens selected home ranges with fewer openings compared to the study area overall. Within home ranges, martens strongly selected complex stands over simple stands and openings. Speed and movement complexity were most consistent over time and movements were more sinuous and slower in complex stand types compared with openings and simple stands. Movement was erratic and more linear in openings than in both complex and simple stands. In simple stands, movement patterns were intermediate between complex stands and openings. Females generally moved more slowly, sinuously, and less variably compared to males. Martens moved more quickly, less sinuously, and more variably during winter compared to summer. However, martens avoided stands with simplified structure, and the altered patterns of movement we observed in those stands suggested that such treatments may negatively affect the ability of martens to forage without increased risk of predation. Fuel treatments that simplify stand structure negatively affected marten movements and habitat connectivity. Given these risks, and because treating fuels is less justified in high elevation forests, the risks can be minimized by applying treatments below the elevations where martens typically occur. Ó 2016 The Wildlife Society.

This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.

No abstract