In the face of inevitable future losses to biodiversity, ranking species by conservation priority seems more than prudent. Setting conservation priorities within species (i.e., at the population level) may be critical as species ranges become fragmented and connectivity declines. However, existing approaches to prioritization (e.g., scoring organisms by their expected genetic contribution) are based on phylogenetic trees, which may be poor representations of differentiation below the species level. In this paper we extend evolutionary isolation indices used in conservation planning from phylogenetic trees to phylogenetic networks. Such networks better represent population differentiation, and our extension allows populations to be ranked in order of their expected contribution to the set. We illustrate the approach using data from two imperiled species: the spotted owl Strix occidentalis in North America and the mountain pygmy-possum Burramys parvus in Australia. Using previously published mitochondrial and microsatellite data, we construct phylogenetic networks and score each population by its relative genetic distinctiveness. In both cases, our phylogenetic networks capture the geographic structure of each species: geographically peripheral populations harbor less-redundant genetic information, increasing their conservation rankings. We note that our approach can be used with all conservation-relevant distances (e.g., those based on whole-genome, ecological, or adaptive variation) and suggest it be added to the assortment of tools available to wildlife managers for allocating effort among threatened populations.

No abstract

No abstract

Climate change is altering fire regimes. As fire regimes change, it is important to understand how mammals respond to these altered post-fire landscapes. Because fires vary in size, severity and landscape context, it is important to know the experimental designs and response variables used to address post-fire responses of mammals. We analysed 48 papers published from 1988 to 2015 that examined responses of small mammals to natural or prescribed fire in North American conifer forests. These papers used different experimental contrasts (e.g. burned vs unburned sites, time series, within-fire heterogeneity). Most studies (89.6%) presented species richness or index-derived abundances of common species as their response variable(s). Many studies did not fully describe the fires being examined; these omissions make it more difficult to interpret and compare results among studies. The limited scope of inference presented by the papers in this review leads us to recommend a minimal set of information that should be presented about each fire studied. We conclude by outlining how different experimental designs and response variables can be used for effective inference. We highlight major pathways forward for examining responses of small forest mammals to the important changes in fire regime that are occurring.

Background Wildfires and forestry activities such as post-fire salvage logging are altering North American forests on a massive scale. Habitat change and fragmentation on forested landscapes may threaten forest specialists, such as Pacific marten (Martes caurina), that require closed, connected, and highly structured habitats. Although marten use burned landscapes, it is unclear how these animals respond to differing burn severities, or how well they tolerate additional landscape change from salvage logging. Methods We used snow tracking and GPS collars to examine marten movements in three large burns in north-central Washington, USA (burned in 2006) and central British Columbia, Canada (burned in 2010 and 2017). We also assessed marten habitat use in relation to areas salvage-logged in the 2010 burn. We evaluated marten path characteristics in relation to post-fire habitat quality, including shifts in behaviour when crossing severely-disturbed habitats. Using GPS locations, we investigated marten home range characteristics and habitat selection in relation to forest cover, burn severity, and salvage logging. Results Marten in the 2006 burn shifted from random to directed movement in areas burned at high severity; in BC, they chose highly straight paths when crossing salvage-blocks and meadows. Collared marten structured their home ranges around forest cover and burn severity, avoiding sparsely-covered habitats and selecting areas burned at low severity. Marten selected areas farther from roads in both Washington and BC, selected areas closer to water in the 2006 burn, and strongly avoided salvage-logged areas of the 2010 burn. Marten home ranges overlapped extensively, including two males tracked concurrently in the 2010 burn. Conclusions Areas burned at low severity provide critical habitat for marten post-fire. Encouragingly, our results indicate that both male and female marten can maintain home ranges in large burns and use a wide range of post-fire conditions. However, salvage-logged areas are not suitable for marten and may represent significant barriers to foraging and dispersal.

No abstract

Wildfires are broad-scale disturbances in North American forests, with impacts that persist for many decades. Further disturbance from post-fire salvage logging is extensively modifying burned landscapes. The removal of habitat structure by fire and salvage logging may affect the persistence of forest-specialist wildlife such as Pacific marten (Martes caurina). However, it is unclear which resources are important to marten on burned landscapes. We used snow tracking and habitat surveys to examine marten habitat selection after three large fires in north-central Washington, USA (10-13 years post-fire), and central British Columbia, Canada (1-2 and 6-9 years post-fire). We developed site-scale habitat models to explain marten foraging and scent-marking post-fire, and assessed further structural changes from salvage logging. Foraging marten chose sites with lower burn severity, greater canopy closure, more vertical structures (trees, snags, saplings, and shrubs), and greater moss/lichen cover than what was generally available. When scent-marking, marten selected structurally-complex sites with abundant deadfall or saplings. Marten moved more quickly when canopy cover was sparse, and rarely used salvage-logged areas. Our results suggest that marten rely on residual habitat structure within large burns, and that secondary disturbance from salvage logging is substantially more harmful to marten than the original fire.