Locating meadow study sitesMeadow centers as recorded in the ‘Copy of sitecords_areaelev from Caruthers thesis.xls’ file delivered by Debinski in November 2012 were matched to polygons as recorded in files ‘teton97map_area.shp’ and ‘gallatin97map_area.shp’ both also delivered by Debinski in November 2012.In cases where the meadow center did not fall within a meadow polygon, if there was a meadow polygon of the same meadow TYPE nearby (judgment was used here), the meadow center was matched with the meadow polygon of same meadow TYPE. In total, 29 of 30 Gallatin meadow sites and 21 of 25 Teton meadow sites were positively located.Identifying meadow pixels for analysisThe native MODIS 250-meter grid was reprojected to match meadow data and added to the GIS project window along with the meadow polygons. For context, aerial photography from ESRI’s basemap streaming services were also added to the ArcMap project. MODIS pixels that were at least half-covered by meadow polygon area were used in further ndvi analysis. Meadows that did not cover at least half of one MODIS pixel were eliminated from the analysis. In total, 17 Gallatin meadow sites (M1= 0; M2= 0; M3= 4; M4= 4; M5= 4; M6=5), covering at least half of 39 MODIS pixels (M1= 0; M2= 0; M3= 12; M4= 4; M5= 6; M6= 17), were used in further analysis and 16 Teton meadow sites (M1=3; M2=1; M3=4; M5=5; M6=3) covering at least half of 1252 MODIS pixels (M1= 105; M2= 1; M3= 25; M4=0 ; M5= 19; M6=1102), were used in further analysis.List of site names that were located, but not used in the NDVI analysis b/c they were too small: Gallatin – Porcupine Exclosure; Twin Cabin Willows; Figure 8; Taylor Fork; Teepee Sage; Daly North; Wapiti (Taylor Fork); Specimen Creek; Bacon Rind M1; Bacon Rind M4, Teepee wet; Daly SouthTeton – Cygnet Pond; Christian Pond; Willow Flats North; Willow Flats South; Sound of MusicMODIS preprocessing methods: MODIS MOD13Q1 representing observations of normalized difference vegetation index (NDVI) from March 2000 through December 2012 were downloaded from the USGS Land Processes Distributed Area Archive Center (LPDAAC) during the spring of 2013. Also downloaded at the same time were grids that described the estimated reliability of NDVI observations and the actual day of the year for each NDVI observation used in maximum compositing routines by the MODIS program. All MODIS data layers were reprojected to match meadow data layers.All negative NDVI values which are thought to correspond to standing water, partial snow-cover or wet bare soil were set to ‘NA’values (Huete, Justice and van Leeuwen 1999)The following steps were used to remove any conifer/evergreen signal from NDVI data and are based on an understanding that each pixel has a different “background”(i.e. no-growth) greenness against which any seasonal change must be compared (Beever et al. 2013; Piekielek and Hansen 2013). These methods also help to eliminate long gaps in data that can allow smoothing algorithms to interpolate beyond the valid range of data (in the case of NDVI from 0 –1):Annual minimum NDVI values that were labeled as high-quality were identified in the 13 year time-series.The bottom first percentileof a distribution of minimum values was used as the “background”value to fill-in missing values when the target was identified as being under snow cover.All NDVI values identified by pixel-reliability grids as being of high or marginal quality werepreserved and snow-covered pixels and dates were filled in with each pixels “background”value.Composite day of year grids were used to identify the actual date from which the 16-day maximum composite NDVI value came.Each pixel’s entire time-series (2000 –2012) was smoothed in a weighted regression framework against time using smoothing splines (Chambers and Hastie 1992). NDVI data of marginal quality and snow-covered background values contributed half the weight to final smoothed values as did high-quality values. The final smoothed values were used to interpolate the time-series to a daily time-step and to record annual NDVI amplitudes. Land surface phenology metrics were calculated as follows:Start of season (SOS) –the first annual day of year when smoothed NDVI crosses half of its annual amplitude (White et al. 2009).End of Season (EOS) –the last day of year when smoothed NDVI crosses half of its annual amplitude.Maximum annual NDVI (MAX) –the highest annual smoothed NDVITiming of annual maximum (DOYMAX) –the smoothed day of year when NDVI reaches its maximum valueEstimated annual productivity (INDVI) –the integrated area under the growing season (SOS to EOS) NDVI curve (Goward et al 1985).

The Prairie Pothole Region (PPR) in the northern Great Plains contains millions of wetlands that provide habitat for breeding and migrating birds. Although conservation and management largely focuses on protecting habitat for nesting ducks, other wetland-dependent birds also rely on this region. Land managers want to know whether habitat conserved for ducks provides habitat for other species and how these habitats will be affected by climate change. A primary goal of this research has been to assist managers in conserving areas that will provide habitat to a broad suite of species. We considered how climate change is likely to affect land-use patterns and agricultural conversion risk. We then predicted how climate change will affect the density and distribution of wetlands under future climate conditions based on models incorporating land-use, climate models, hydrology, and distribution of wetland basins. Although the density of wetlands with water will most likely decline across the region, the distribution of wetlands probably will not shift spatially because the location of wetland basins is static. Species distribution modeling techniques projected that geographic ranges of nearly 30 species of wetland-dependent birds will decline by an average of 31% (range: 75% decline to 16% increase) as the climate warms. To test whether waterfowl are effective representatives, or surrogates, for other wetland-dependent birds, we used data from citizen science bird surveys and species life history to mathematically demonstrate how closely wetland birds associate with waterfowl. At small scales in space and time (for example, a small wetland complex on an annual basis), many waterfowl and other wetland birds species were segregated. Yet at larger scales in space and time, the scales at which habitat protection decisions are made, many species appeared to co-occur because various microhabitats were represented in the larger dynamic landscapes through 30-yr time periods.

Species distribution models (SDMs) are commonly used to assess potential climate change impacts on biodiversity, but several critical methodological decisions are often made arbitrarily. We compare variability arising from these decisions to the uncertainty in future climate change itself. We also test whether certain choices offer improved skill for extrapolating to a changed climate and whether internal cross-validation skill indicates extrapolative skill. We compared projected vulnerability for 29 wetland-dependent bird species breeding in the climatically dynamic Prairie Pothole Region, USA. For each species we built 1,080 SDMs to represent a unique combination of: future climate, class of climate covariates, collinearity level, and thresholding procedure. We examined the variation in projected vulnerability attributed to each uncertainty source. To assess extrapolation skill under a changed climate, we compared model predictions with observations from historic drought years. Uncertainty in projected vulnerability was substantial, and the largest source was that of future climate change. Large uncertainty was also attributed to climate covariate class with hydrological covariates projecting half the range loss of bioclimatic covariates or other summaries of temperature and precipitation. We found that choices based on performance in cross-validation improved skill in extrapolation. Qualitative rankings were also highly uncertain. Given uncertainty in projected vulnerability and resulting uncertainty in rankings used for conservation prioritization, a number of considerations appear critical for using bioclimatic SDMs to inform climate change mitigation strategies. Our results emphasize explicitly selecting climate summaries that most closely represent processes likely to underlie ecological response to climate change. For example, hydrological covariates projected substantially reduced vulnerability, highlighting the importance of considering whether water availability may be a more proximal driver than precipitation. However, because cross-validation results were correlated with extrapolation results, the use of cross-validation performance metrics to guide modeling choices where knowledge is limited was supported.

This dataset represents the area in the Greater Yellowstone Ecosystem prioritized for different whitebark pine(Pinus albicaulis) management activities, summarized by climate suitability zones. This data was developed for use in a landscape simulation modeling study aimed at evaluating how well alternative management strategies maintain whitebark pine populations under historical climate and future climate conditions. For the study, we developed three spatial management alternatives for whitebark pine in the Greater Yellowstone Ecosystem representing no active management, current management, and climate-informed management. These management alternatives were implemented in the simulaton model FireBGCv2 under historical climate and three future climate change scenarios - the HadGEM-ES, CESM1-CAM5, and CNRM-CM5 Global Circulation Models under the RCP 8.5 emissions scenario. We worked with the Greater Yellowstone Coordinating Committee's (GYCC) Whitebark Pine Subcommittee to develop this spatial representation of their current management strategy. The treatments mapped represent a set of the treatments recommended in the GYCC Whitebark Pine 2011 Strategy document and include planting blister-rust resistant whitebark pine seedlings, competition removal thinning, wildland fire use and prescribed fire, and protection from mountain pine beetles using verbenone and carbaryl. We used historical and future projections of climate suitability based on species distribution models for whitebark pine (Chang et al. 2014) to map zones of core, deteriorating, and future whitebark pine habitat. Core zones were those areas that are currently suitable for whitebark and remain suitable in the future. Deteriorating zones were where the climatic conditions for whitebark pine are expected to decline. Future zones were areas that are projected to become newly suitable for whitebark pine. We then overlaid our climate zones for whitebark pine with similar projections of future climate suitability for all of whitebark pine’s competitors - Engelmann spruce, subalpine fir, lodgepole pine, and Douglas-fir (Piekielek et al. 2015. We discussed the different combinations of climate suitability zones (core, deteriorating, future) and potential future level of competition (low or high) from other species with the GYCC Whitebark Pine Subcommittee to determine which management activities should be prioritized within each management zone. The result is a map of management zones where different activities are prioritized to meet the goal of maintaining whitebark pine populations. This was used to determine which treatments would be implemented spatially during the simulation modeling, dependent upon additional criteria related to simulated stand-level conditions. In this dataset, we used the resulting map of spatially prioritized management activities to summarize the area prioritized for each management activity that fell within Core, Deteriorating, and Future climate suitability zones

This dataset represents the area in the Greater Yellowstone Ecosystem prioritized for different whitebark pine(Pinus albicaulis) management activities, summarized by land classes. This data was developed for use in a landscape simulation modeling study aimed at evaluating how well alternative management strategies maintain whitebark pine populations under historical climate and future climate conditions. For the study, we developed three spatial management alternatives for whitebark pine in the Greater Yellowstone Ecosystem representing no active management, current management, and climate-informed management. These management alternatives were implemented in the simulaton model FireBGCv2 under historical climate and three future climate change scenarios - the HadGEM-ES, CESM1-CAM5, and CNRM-CM5 Global Circulation Models under the RCP 8.5 emissions scenario. We worked with the Greater Yellowstone Coordinating Committee's (GYCC) Whitebark Pine Subcommittee to develop this spatial representation of their current management strategy. The treatments mapped represent a set of the treatments recommended in the GYCC Whitebark Pine 2011 Strategy document and include planting blister-rust resistant whitebark pine seedlings, competition removal thinning, wildland fire use and prescribed fire, and protection from mountain pine beetles using verbenone and carbaryl. We used historical and future projections of climate suitability based on species distribution models for whitebark pine (Chang et al. 2014) to map zones of core, deteriorating, and future whitebark pine habitat. Core zones were those areas that are currently suitable for whitebark and remain suitable in the future. Deteriorating zones were where the climatic conditions for whitebark pine are expected to decline. Future zones were areas that are projected to become newly suitable for whitebark pine. We then overlaid our climate zones for whitebark pine with similar projections of future climate suitability for all of whitebark pine’s competitors - Engelmann spruce, subalpine fir, lodgepole pine, and Douglas-fir (Piekielek et al. 2015. We discussed the different combinations of climate suitability zones (core, deteriorating, future) and potential future level of competition (low or high) from other species with the GYCC Whitebark Pine Subcommittee to determine which management activities should be prioritized within each management zone. The result is a map of management zones where different activities are prioritized to meet the goal of maintaining whitebark pine populations. This was used to determine which treatments would be implemented spatially during the simulation modeling, dependent upon additional criteria related to simulated stand-level conditions. In this dataset, we used the resulting map of spatially prioritized management activities to summarize the area prioritized for each management activity that fell within different land classifications (mutliple use forests, National Park Service lands, Wilderness lands, and non-federal lands).

Natural resource managers face the need to develop strategies to adapt to projected future climates. Few existing climate adaptation frameworks prescribe where to place management actions to be most effective under anticipated future climate conditions.  We developed an approach to spatially allocate climate adaptation actions and applied the method to whitebark pine (WBP; Pinus albicaulis) in the Greater Yellowstone Ecosystem (GYE).  WBP is expected to be vulnerable to climate-mediated shifts in suitable habitat, pests, pathogens, and fire. We worked with a team of biologists and managers to identify management actions aimed at mitigating climate impacts to WBP. Identified actions were spatially allocated across the GYE under two management strategies: (1) current management and (2) climate-informed management which used projected climate suitability for WBP and competing tree species to place management actions.  The current management strategy reflected current legal, policy and access contraints, such as restricting active management in Wilderness and remote locations, while the climate-informed management strategy was designed to maximize preservation of WBP forests regardless of such constraints. Thus, the climate-informed strategy highlighted how the spatial location of management actions might need to shift to most effectively maintain WBP forests under future climate conditions. The spatial distribution of actions and area treated differed among the current and climate-informed management strategies, with 33-60% more wilderness area prioritized for action under climate-informed management. High priority areas for implementing management actions include the 1-8% of the GYE where current and climate-informed management agreed, since this is where actions are most likely to be successful in the long-term and where current management permits implementation. Areas where climate-informed strategies agreed with one another but not with current management (6-22% of the GYE) are potential locations for experimental testing and monitoring of management actions. Our method for prioritizing locations for climate-adaptation actions is applicable to any species for which information regarding climate vulnerability and climate-mediated risk factors is available.

Abstract from Ecosphere: The Prairie Pothole Region, situated in the northern Great Plains, provides important stopover habitat for migratory shorebirds. During spring migration in the U.S. Prairie Potholes, 7.3 million shorebirds refuel in the region's myriad small, freshwater wetlands. Shorebirds use mudflats, shorelines, and ephemeral wetlands that are far more abundant in wet years than dry years. Generally, climate change is expected to bring warmer temperatures, seasonality shifts, more extreme events, and changes to precipitation. The impacts to wetland habitats are uncertain. In the Prairie Potholes, earlier spring onset and warmer temperatures may advance drying of wetlands or, alternately, increased spring precipitation may produce abundant shallow‐water habitats. To look at the availability of habitats for migratory shorebirds under different climate regimes, we compared habitat selection between a historic wet year and a dry year using binomial random‐effects models to describe local and landscape patterns. We found that in the dry year shorebirds were distributed more northerly and among more permanent wetlands, whereas in the wet year shorebirds were distributed more southerly and among more temporary wetlands. However, landscape‐scale variation played a larger role in the dry year. At the local wetland scale, shorebirds selected similarly between years—for shallower wetlands and wetlands in croplands. Overall, while shorebirds were sensitive to local habitat conditions, they exhibited a degree of adaptive capacity to climate change impacts by their ability to shift on the landscape. This indicates an avenue through which management decisions can enhance climate change resilience for these species given an uncertain future—by preserving shallow‐water wetlands in croplands throughout the landscape.

Managing natural resources is fraught with uncertainties around how complex social-ecological systems will respond to management actions and other forces, such as climate. Modeling tools have emerged to help grapple with different aspects of this challenge, but they are often used independently. The purpose of this project is to link two types of commonly-used simulation models (agent-based models and state-and-transition simulation models) and streamline the handling of model inputs and outputs. This innovation will provide researchers with the capability to simulate the interactions of wildlife, vegetation, management actions, and other drivers, and thus answer questions and inform decisions about how best to manage natural resources.

Sagebrush steppe is one of the most widely distributed ecosystems in North America. Found in eleven western states, this important yet fragile ecosystem is dominated by sagebrush, but also contains a diversity of native shrubs, grasses, and flowering plants. It provides critical habitat for wildlife like pronghorn and threatened species such as the greater sage-grouse, and is grazed by livestock on public and private lands. However, this landscape is increasingly threatened by shifts in wildfire patterns, the spread of invasive grasses, and changing climate conditions. While sagebrush is slow to recover after fires, non-native grasses such as cheatgrass thrive in post-fire conditions and the spread of these species can increase the frequency and intensity of wildfires. These changes to the sagebrush ecosystems have implications for big game, threatened wildlife, and ranching. To address this growing concern, resource managers will often try to limit the spread of exotic grasses after fire events by applying herbicides, or will help native species recover through seeding or planting. However, these treatments have mixed results, and poor success is often attributed to droughts, which make it more difficult for seeds and native plants to survive; to the limited amount of time in which these treatments can be applied (usually in the first year after a fire); or because the seeds or plants used aren’t adapted to the environmental conditions of the location where they’re applied. The goal of this project is to improve our understanding of the factors that affect post-fire treatment success. Researchers will use data collected from more than 300 fires over the last 40 years, after which treatments were applied. They will identify the impacts of drought on those treatments, how incorporating information on drought forecasts or extending the period over which treatments are applied could have altered the outcomes, and how managers can better select plant material that will be more adaptable to the conditions of planting locations. Addressing this knowledge gap has been identified as a high priority in the DOI Integrated Rangeland Fire Management Strategy, by the BLM Emergency Stabilization and Rehabilitation Program, and by state management agencies in the West. The results of this project will support adaptive management of sagebrush ecosystems, which will be critical if these ecologically and economically important landscapes are to be maintained into the future. This project was jointly funded by the Southwest, Northwest, and North Central CASCs.