Understanding and Adapting to Climate Change: A Focus on Snow Specialists (hosted by TWS)

The northern hemisphere is experiencing drastic reductions in snow cover duration as a result of a warming climate. Many plant and animal species inhabiting seasonally snow-covered environments have key adaptations to deep persistent snow cover. These same adaptations can make them vulnerable to warming and many are forecasted to undergo dramatic range contraction to higher latitudes and elevations as a result of reduced snow cover. For example, in animals that exhibit seasonal changes in pelage or plumage color, this molt can become mismatched with the background leading to increased predation rates, sufficient to decrease population growth and extirpate peripheral populations. For animals that rely on snow for thermoregulation (e.g., grouse) or subnivean behaviors like foraging (e.g., marten) changing snow amounts and condition can similarly affect population vital rates. Further, many snow specialists are integral to northern food webs and range contractions and population declines of these species can lead to cascading effects across trophic levels. This symposium will address the current state of knowledge and projected impacts of snow cover reductions and variability on these snow specialist species. Presentations will address management and research concerns, advances in landscape snow cover predictions, and identify synergies and challenges to research and management. Information from this symposium will target a wide audience including foresters, wildlife and conservation biologists, conservationists, climatologists and policymakers.

1:10PM Satellite Remote Sensing of Snow Season Dynamics
  David Gudex-Cross, Volker Radeloff, Benjamin Zuckerberg, Jonathan Pauli, Daniel Fink, Madeleine A. Rubenstein
Accurate mapping of snow cover dynamics over time and space is critical to informing models of climate, hydrology, biodiversity and species distributions, and ecosystem services. However, seasonal snow dynamics remain poorly quantified and understood compared to other climate variables (e.g., temperature and rainfall). While meteorological stations provide important in situ measurements of snow cover, satellite observations have the advantage of offering continuous spatial coverage at global scales. Here, using the powerful cloud-computing platform of Google Earth Engine, we analyzed imagery obtained from the Landsat and MODIS sensors to create new dynamic snow indices (DSIs) that describe annual measures of snow season length, extent, and variability. Initial results show the DSIs capture major latitudinal and longitudinal gradients across the continent – as well as those at a finer scale (e.g., the ‘tension zone’ in Wisconsin) – with a high amount of interannual variability in snow cover and extent, particularly at lower latitudes. Next steps include: 1) comparing the results obtained from the Landsat (30m pixels, 16-day returns) and MODIS (500m pixels, daily returns) sensors, 2) validation of the DSIs with ground-truth data, and 3) testing the ability of each DSI to predict winter species distributions for mammals that rely on the subnivium.
1:30PM Chasing Snow: The Behavioral Consequences of Inconsistent Snow on a Snow Specialist
  Shawn Cleveland, H. Brian Underwood
Climate change is forecasted to alter the timing and duration of snow cover in northern latitudes. For snow dependent species, the cost of inconsistent or absent snow can have serious behavioral and fitness consequences. Predation has been shown to cause drastic reductions in survival, but a less explored area of climate induced stress is that of the predation risk associated with snow variability. Color morph species may result in being mismatched to their surroundings (white hair corresponding to a white, snow covered background vs white hair on a brown back ground, matched vs mismatched respectively), negating the benefit of camouflage. This mismatch can have significant behavioral consequences which may exacerbate fitness consequences. We used snowshoe hares as a model species to investigate the effects of mismatched induced increases in predation risk on behavior and fitness. Our results demonstrate a four-fold reduction in home range size in variable snow years and a corresponding decline in survival. Furthermore, these results suggest that there may be heightened fitness consequences for hares in hardwood vs conifer forests given that hardwoods forest have an additional level of tree phenology (leaf loss resulting in reduced hiding cover) that conifer forest dwelling hares are not subject to. This trifecta of phenology (coat color phenology, snow phenology, and tree phenology) may further explain the range contraction rates that are occurring eastern forest.
1:50PM Consistent Responses to Snow and Forest Cover between Disjunct, Peripheral Populations of Snowshoe Hares
  Sean Sultaire, John Humphreys, Gary Roloff
Southern range boundaries of species may signal thresholds of environmental tolerances useful for predicting responses to climate change. Hence, quantifying environment drivers of occupancy within southern peripheral populations is the subject of considerable conservation research. Spatial variation in environmental tolerances across edge distributions may better signal adaptive potential than summation of environmental tolerances across an entire range boundary. We quantified variation in environmental correlates linked to snowshoe hare occupancy (Lepus americanus) for the southern range boundary in the Great Lakes Region, USA. We surveyed for snowshoe hares in three study areas, central Wisconsin, Upper Peninsula of Michigan, and Lower Peninsula of Michigan, and related hare occurrence to land cover and climate variables. Central Wisconsin and the Upper Peninsula share a broad geographic border whereas the Lower Peninsula has been isolated since the end of the last ice age. We found that relationships between snowshoe hare occurrence and environmental variables exhibited only slight variation across study areas. Snowshoe hare occupancy showed a positive relationship with snow cover duration and forest cover, and a negative relationship with maximum summer temperature. Validation statistics indicated better performance for models specified to allow for region specific variation than a model that combined occurrence data across the study areas. Given the covariates we included in our modeling, our results suggest that environmental tolerances across the southern range boundary of snowshoe hares in the Great Lakes Region are similar, implying that hares have limited adaptive capacity to respond to climate change.
2:10PM Reenacting the Range Contraction of Snowshoe Hare Via an Experimental Translocation
  Evan Wilson, Jonathan Pauli
Climate change is disrupting biotic interactions and community assemblages globally. Seasonally snow-covered environments featuring winter-adapted species have proven to be important models for understanding the mechanisms through which climate change affects biotic communities. Snowshoe hares (Lepus americanus), a species of central importance in northern forests, have experienced local declines and distributional range contractions along their southern boundary. These declines have been linked to climate change via the mechanism of camouflage mismatch, an inability to match seasonal coat color to declining snow cover duration. However, this relationship has not been experimentally tested, nor has the role of landscape heterogeneity been evaluated. To identify the mechanisms driving range shifts, we performed an experimental translocation of snowshoe hares along their historic range boundary in Wisconsin. Our approach allowed us to directly observe how altered climatic conditions and landscape features interplay to drive population declines and local extirpation. During the winter of 2017, we translocated, radio-collared, released and monitored 99 snowshoe hares for survival pre-, during and post-mismatch. We used camera traps to monitor the spatial extent of dispersal and habitat associations of hares before, during and after translocation, and multi-species occupancy models to estimate colonization and extinction. Hare survival was strongly associated with camouflage mismatch, resulting in high mortality from predation, though this was buffered by the availability of large patches of young aspen (Populus spp.) or alder (Alnus spp.). Colonization of hares on the landscape was strongly associated with distance from source population and patch size of aspen-alder, while extinction was influenced by interspersion of habitat types and patch size of aspen-alder. As climate change alters ecological communities, identifying mechanisms that drive range-shifts in climate-sensitive species and those landscape features that can be manipulated to buffer such effects will be critical for management aimed at promoting persistence of winter-adapted communities.
2:50PM Refreshment Break
3:00PM Snow-Mediated Plasticity in Phenology Does Not Prevent Camouflage Mismatch
  Alexander Kumar, Marketa Zimova, L. Scott Mills, Jim Sparks
Global reduction in snow cover duration is one of the strongest outcomes of climate change. This loss of snow has severe negative consequences for seasonally color molting species, such as snowshoe hares (Lepus americanus) that rely on snow for crypsis. However, phenotypic plasticity may provide an avenue for hares to adapt to reduced snow duration. Plastic responses could occur in the color molt phenology or through hare behavior minimizing mismatch or its consequences. We quantified molt phenology of 200 wild snowshoe hares and measured microhabitat choice and local snow cover in Montana. Hares did not use behavioral plasticity to minimize coat color mismatch via background matching; instead they preferred snow free (brown) areas regardless of coat color, likely for thermoregulatory reasons. Furthermore, hares also did not adjust their behaviors to mitigate the costs of mismatch on survival by choosing denser vegetation for resting areas. Importantly, we did find snow-mediated plasticity in both the initiation and the rate of the molt; greater snow cover was associated with whiter hares even though hares selected for non-snowy areas. However, the plasticity in average molt phenology in years of differing snow duration was insufficient to eliminate camouflage mismatch with white hares on brown snowless ground. We conclude that phenotypic plasticity is unlikely to facilitate adaptive rescue to camouflage mismatch under climate change.
3:20PM Adaptive Capacity to Camouflage Mismatch: The Heterogeneous Responses to Climate Change across Color Molting Species
  Marketa Zimova, L. Scott Mills, Alexej Siren, Toni Lyn Morelli, Josh Nowak, Jacob Ivan, Jesse Whittington, Scott Newey, Sean Giery, Anders Angerbjorn, Dick Moberg
Mismatch in seasonal molt for animals that turn white in winter and increasingly confront snowless conditions has emerged as an important example of phenological mismatches directly linked to climate change. In multiple species including snowshoe hares and least weasels camouflage mismatch has severe survival costs and may have negative population consequences in the absence of adaptation under climate change. We explored the potential of multiple color molting species to respond adaptively to camouflage mismatch via phenotypic plasticity and evolution. First, we tested whether mountain hares in Scotland responded adaptively to the long-term declines in snow cover. We repeated historical field studies from the 1950s and found that mountain hares did not shift their phenologies, and subsequently experienced increased camouflage mismatch. The attenuated selection pressure in Scotland is likely the leading cause for the static phenologies. Next, we quantified the intra-specific variation in vulnerability to mismatch across snowshoe hares’ range in North America. Using remote camera-trap and high-resolution climate data, we quantified the environmental drivers of molt phenology, phenotypic plasticity, and the extent of camouflage mismatch in three disjunct populations. We found that the extent of buffering against mismatch via phenotypic plasticity varied across snowshoe hare populations, with most limited buffering in regions characterized by short snow-season duration. Finally, we described molt phenology and camouflage mismatch in polymorphic Arctic fox populations in Sweden where both winter brown and white individuals coexist. Higher mismatch was observed in the winter white color morphs; however we believe that these polymorphic populations have a broad scope for evolutionary rescue. In general, we observed high variation in the adaptive capacity to camouflage mismatch within and among species. As climate change will continue to decrease the duration of snow cover, evolutionary shifts will likely be pivotal for local persistence in color molting species.
3:40PM Direct and indirect effects of climate and biotic factors on wildlife communities in the northeastern U.S.
  Alexej Siren
Anthropogenic climate and habitat change are the primary factors attributed to recent distributional shifts and population declines of wildlife species across the globe. Species interactions (e.g., competition, predation) can also influence distribution dynamics but can be correlated with environmental variables. As such, a synergistic understanding of direct and indirect effects is needed to accurately predict population trajectories in the face of climate and habitat change. The northeastern U.S. contains a diversity of wildlife species, including the equatorial edge of several snow specialists (Canada lynx [Lynx canadensis], American marten [Martes americana], and snowshoe hares [Lepus americanus]) and the poleward or altitudinal edges for generalist species (e.g., bobcats [Lynx rufus], fishers [Pekania pennanti]); the former of which are considered vulnerable due to predicted declines in snowpack and boreal forest habitat. The mechanisms that drive the distribution dynamics of these species, including the extent to which climate and biotic interactions influence range edges and shifts, are still unclear. We evaluated the interplay of snowpack variables and biotic factors on a suite of interacting carnivore and herbivore species along the boreal-temperate ecotone in the northeastern U.S. and discuss findings within the context of conservation and management decisions. Notably, we found evidence that competitive interactions between snow specialists and generalists are mediated by snowpack dynamics and that dramatic shifts can occur within a few years across elevational and latitudinal gradients in response to changing snow conditions. We advocate adopting modeling approaches that allow for inclusion of direct and indirect effects to provide a mechanistic understanding of how climate change may potentially influence future wildlife populations.
4:00PM Deep Snow Creates Microrefugia That Influence Habitat Selection and Behavioral Plasticity in a Winter-Adapted Bird
  Amy Shipley, Jennyfer Cruz, Benjamin Zuckerberg
Species in temperate ecosystems employ a variety of strategies to cope with the energetic demands of winter. The ability to select microrefugia that provide suitable protection from the cold is important, especially when key features of the environment are dynamic throughout the winter season. In previous studies, we found that deep snow, which is thermally insulative, mediates the negative effect of cold temperatures on physiological stress in ruffed grouse (Bonasa umbellus). However, it is unknown whether the spatiotemporal characteristics of snow cover influence how grouse actively select habitats through the winter months. We conducted a study of the microclimates associated with 700 grouse roost sites during three winters (2016-2018). We measured snow depth at roost sites, and interpolated snow depth and minimum daily temperature based on data collected from 58 weather stations across our study site. We then used resource selection function analysis to compare environmental conditions and forest cover types between grouse roost sites and random locations. Grouse strongly selected for areas with deep snow at the population level, but there was a high degree of individual variation in selection of snow-covered areas. Temperature did not appear to influence roost site selection. The overall strong preference for areas with deep snow during winter highlights their importance as microrefugia, but the loss of snow cover due to climate change may limit the ability of winter-adapted species to access this dynamic refuge.
4:20PM Seasonal Coat Color As a Model to Foster Adaptive Rescue of Wildlife Confronting Climate Change
  L. Scott Mills
At the individual and species level, wildlife encountering rapid environmental change have only three options: die, move, or adapt. Most scientific and media focus has been on the first two options, implying that animals must either move or perish. The relative lack of focus on the third outcome — local adaptation — is perplexing given the central role that adaptation through natural selection and individual plasticity plays in shaping diversity and persistence of biodiversity. Further, the absence of management-relevant criteria to facilitate local adaptation by wild species is a particularly acute omission in a rapidly changing climate. Polymorphisms in animal color patterns have long been classic models to connect trait variation to processes shaping adaptation in a variable environment. Our research group focuses on a trait found in 21 bird and mammal species globally: the genetically-based seasonal color molt from brown to white as camouflage against snow. Reduced snow duration is a pervasive effect of climate change. Therefore, camouflage mismatch generated by seasonal color molt provides a model to evaluate scope and mechanisms of adaptation in a rapidly changing world. A crucial next step requires generalizing the insights derived from seasonal coat color — and other traits shaped directly by climate — to derive practical guidelines. I will distill key findings and principles to initiate a broader conversation of how adaptive rescue might be added as a new axis for focusing specific conservation efforts to promote climate resilience. Ideally, discussion will begin to frame a roadmap for global policy makers whereby underrepresented hotspots for adaptive rescue could be included with other spatially mappable traits (such as species richness, diversity, and focal species dynamics) in a comprehensive science-based platform for conservation-decision making.
4:40PM Panel Discussion

Organizers: Shawn Cleveland, Gary Roloff, Sean Sultaire
Supported by: The National Wildlife Federation, The Climate Change and Wildlife Working Group, The Snowshoe Hare Working Group

Location: Reno-Sparks CC Date: October 2, 2019 Time: 1:10 pm - 5:00 pm