Effects of forest management on vertebrates: synthesizing two decades of data from hardwood forests in Missouri, USA.

Modern forest management seeks to balance multiple social, economic, and ecological goals. Different management approaches create different types of disturbances in a forest ecosystem and thus also differ in their impacts on plants, animals, and insects. Understanding these impacts is important for conservation of forest ecosystem function, but challenging due to the large spatial and temporal scale over which management occurs. Most past research has focused on relatively small areas, short time scales, and/or a small number of species. To address this, we examined the effects of two common silvicultural systems (even and uneven aged) on abundance and richness of three vertebrate taxa (birds, small mammals, and herpetofauna) over a two-decade period in a temperate hardwood forest in Missouri, USA. The two systems removed a similar amount of biomass overall, but differed in the intensity, number, and configuration of harvests applied. We found that vertebrate population responses varied by taxa, occurred at multiple spatial scales, and were concentrated in the period following the first harvest entry. Birds generally had the largest changes in relative abundance, both positive and negative, following management. Small mammals and reptiles had smaller, but generally positive, responses; amphibians were mixed. Bird species tended to respond in the same way to both silvicultural systems, while small mammals and herpetofauna did not respond consistently. Thus, for birds, the total amount of harvest disturbance across the landscape drives population responses, while for others the size and configuration of individual harvests is likely more important. Synthesizing results across the vertebrate community at large spatial and temporal scales allows managers to better understand trade-offs when making decisions that will affect wildlife in contrasting ways.

Low-level pathogen infection and geographic location correlate with the skin microbiomes of Columbia spotted frogs (Rana luteiventris) in a montane landscape.

The skin microbiome of amphibians can influence host susceptibility towards the fungal pathogen Batrachochytrium dendrobatidis (Bd), while simultaneously having the potential to be altered by Bd. Severe Bd infections are known to alter the amphibian skin microbiome; however, little is known about microbiome interactions in amphibians with low infection intensity. In addition to disease dynamics, environmental factors may influence the microbiome. To test for patterns in bacterial diversity based on pathogen infection and environmental factors, 399 Columbia spotted frogs (Rana luteiventris) were sampled throughout northern Idaho and northeastern Washington across two years. Bd prevalence and intensity were measured in 376 frogs, revealing a prevalence of 69%, but generally low infection intensity (Mean = 127 Bd zoospore equivalents among infected frogs). Skin bacterial communities were characterized in 92 frogs using 16S rRNA gene amplicon sequencing. Our results indicated correlations of decreasing Shannon diversity and evenness as infection intensity increased. Latitude was correlated with bacterial richness and Faith's Phylogenetic Diversity measures, indicating increased diversity in northern locations. Beta diversity (UniFrac) analyses revealed that skin microbiomes were distinct between infected and uninfected frogs, and infection intensity had a significant effect on microbiome composition. Site explained the majority of microbiome variation (weighted UniFrac: 57.5%), suggesting a combination of local habitat conditions explain variation, as only small proportions of variation could be explained by year, month, temperature, elevation, and latitude individually. Bacterial genera with potential for Bd-inhibitory properties were found with differential relative abundance in infected and uninfected frogs, with higher Stenotrophomonas and lower Pseudomonas relative abundance observed in infected frogs. Further study may indicate if Bd inhibition by members of the skin microbiome is an influence behind the low infection intensities observed and whether low Bd infection intensities are capable of altering skin microbiome composition.

Carnivore Contact: A Species Fracture Zone Delineated Amongst Genetically Structured North American Marten Populations (Martes americana and Martes caurina).

North American martens are forest dependent, influenced by human activity, and climate vulnerable. They have long been managed and harvested throughout their range as the American marten (Martes americana). Recent work has expanded evidence for the original description of two species in North America - M. americana and the Pacific Coast marten, M. caurina - but the geographic boundary between these groups has not been described in detail. From 2010 to 2016 we deployed 734 multi-taxa winter bait stations across a 53,474 km2 study area spanning seven mountain ranges within the anticipated contact zone along the border of Canada and the United States. We collected marten hair samples and developed genotypes for 15 polymorphic microsatellite loci for 235 individuals, and 493 base-pair sequences of the mtDNA gene COI for 175 of those individuals. Both nuclear and mitochondrial genetic structure identified a sharp break across the Clark Fork Valley, United States with M. americana and M. caurina occurring north and south of the break, respectively. We estimated global effective population size (N e ) for each mountain range, clinal genetic neighborhood sizes (NS), calculated observed (H o ) and expected (H e ) heterozygosity, fixation index (F ST ), and clinal measures of allelic richness (Ar), H o , and inbreeding coefficient (F IS ). Despite substantial genetic structure, we detected hybridization along the fracture zone with both contemporary (nuclear DNA) and historic (mtDNA) gene flow. Marten populations in our study area are highly structured and the break across the fracture zone being the largest documented in North America (F ST range 0.21-0.34, mean = 0.27). With the exception of the Coeur d'Alene Mountains, marten were well distributed across higher elevation portions of our sampling area. Clinal NS values were variable suggesting substantial heterogeneity in marten density and movement. For both M. americana and M. caurina, elevationaly dependent gene flow and high genetic population structure suggest that connectivity corridors will be important to ensuring long-term population persistence. Our study is an example of how a combination of global and clinal molecular data analyses can provide important information for natural resource management.