Gregory T. Wann Nathan D. Van Schmidt Jessica E. Shyvers Bryan C. Tarbox Megan M. McLachlan Michael S. O'Donnell Anthony J. Titolo Peter S. Coates David R. Edmunds Julie A. Heinrichs Adrian P. Monroe Cameron L. Aldridge 20221102 1st raster digital data https://doi.org/10.5066/P95YAUPH Gregory T. Wann Nathan D. Van Schmidt Jessica E. Shyvers Bryan C. Tarbox Megan M. McLachlan Michael S. O'Donnell Anthony J. Titolo Peter S. Coates David R. Edmunds Julie A. Heinrichs Adrian P. Monroe Cameron L. Aldridge 2022 publication Global Ecology and Conservation Elsevier This dataset contains two predictive lek (breeding site) persistence raster layers covering the U.S. greater sage-grouse distribution. In the United States, locations where males display and breed with females (i.e., leks) are often monitored annually by state wildlife agencies, providing valuable information on the persistence of birds in the surrounding areas. A U.S. range-wide lek database was recently compiled for greater sage-Grouse (O’Donnell et al. 2021), providing a standardized source of information to build statistical models to evaluate environmental characteristics associated with lek persistence. The compiled lek database classified a subset of leks as being either active (leks currently used for breeding activities) or inactive (leks no longer used for breeding activities) based on count data collected over a 20-year monitoring period. We fit the outcome of a lek being active or inactive as a function of environmental predictors characterizing surrounding conditions in a logistic regression model. Covariates included sagebrush cover, pinyon-juniper cover, topography, precipitation, point and line disturbance densities, and landscape configuration metrics. We included the Bureau of Land Management habitat assessment areas (termed mid-scales) as regional random effects in the form of random intercepts and random slopes (for a subset of covariates). The final model included 13 covariates. We predicted conditional probabilities of lek persistence across the U.S. occupied range using the covariate layers and regional mid-scales, which we make available here as a 30-meter resolution continuous raster dataset. The predictions were conditional because they were specific to each mid-scale factor level (i.e., pixel predictions were influenced by the regional mid-scale polygon they fell within via the associated mid-scale intercept and random slope deviations). We applied sensitivity thresholds (capturing percentage of leks correctly classified as active) to the continuous probability layer to bin persistence probabilities into high, medium, low, and marginal areas of persistence, which we make available here as a 30-m categorical raster dataset. These datasets were created to identify areas within the greater sage-grouse U.S. range where breeding sites (i.e., leks) are predicted to persist based on current environmental conditions. The datasets could be used as a general assessment tool for habitats within various defined areas (e.g., by averaging pixels of the conditional predicted lek persistence probabilities), or to identify areas where active leks are most likely to be located (using the habitat categories defined by model sensitivities). 2019 observed Complete As needed U.S. greater sage-grouse distribution -119.5437 -103.4683 49.9095 35.9740 ISO 19115 Topic Category biota None greater sage-grouse lek persistence USGS Metadata Identifier USGS:62d75444d34e3b97e58cad4b Common geographic areas Colorado Montana North Dakota South Dakota Washington California Nevada Utah Wyoming Idaho Oregon United States None greater sage-grouse Kingdom Animalia animals Subkingdom Bilateria Infrakingdom Deuterostomia Phylum Chordata chordates Subphylum Vertebrata vertebrates Infraphylum Gnathostomata Superclass Tetrapoda Class Aves Birds Order Galliformes Fowls Gallinaceous Birds Family Phasianidae Partridges Turkeys Grouse Pheasants Quail Subfamily Tetraoninae Grouse Genus Centrocercus Sage Grouse Species Centrocercus urophasianus Greater Sage Grouse Sage Grouse Greater Sage-Grouse TSN: 175855 None. Please see 'Distribution Info' for details. The authors of these data encourage data users to contact them regarding its intended use. This will help users understand limitations and interpretation of the data, and assumptions of the model used to create the data. Gregory Wann US Geological Survey Ecologist physical

2150 Centre Avenue Bldg C

Fort Collins CO 80526 USA 9702269309 wanng@usgs.gov This work was funded by the Bureau of Land Management We conducted all modeling in R v. 4.0.2 (R Core Team 2020) using the package lme4 v. 1.1-28 (Bates et al. 2015). All native spatial raster predictor data were processed in ArcMap v. 10.7 and v. 10.6 (Esri, Redlands, CA), and in the case of landscape metric predictor variables, created using the R package landscapemetrics v. 1.2-18 (Hesselbarth et al. 2019). Covariate extractions at lek locations were done using the R package exactextractr v. 0.4.0 (Baston 2020). The spatial prediction raster was made in R using the packages raster v. 3.3-13 (Hijmans 2020) and rgdal v. 1.5-16 (Bivand et al. 2020) and applying the model formula to a spatial predictor raster stack. P. E. GESSLER I. D. MOORE N. J. McKENZIE P. J. RYAN 199507 publication International journal of geographical information systems vol. 9, issue 4 n/a Informa UK Limited ppg. 421-432 https://doi.org/10.1080/02693799508902047 J. MARK SAPPINGTON KATHLEEN M. LONGSHORE DANIEL B. THOMPSON 200707 publication Journal of Wildlife Management vol. 71, issue 5 n/a Wiley ppg. 1419-1426 https://doi.org/10.2193/2005-723 Maximilian H. K. Hesselbarth Marco Sciaini Kimberly A. With Kerstin Wiegand Jakub Nowosad 20190805 publication Ecography vol. 42, issue 10 n/a Wiley ppg. 1648-1657 https://doi.org/10.1111/ecog.04617 R Core Team 2020 publication R Core Team (2020) R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL https://www.R-project.org/ https://www.R-project.org/ Daniel Baston 2020 R package version 0.4.0. publication Baston, D (2020) exactextractr: Fast extraction from raster datasets using polygons. R package version 0.4.0. https://CRAN.R-project.org/package=exactextractr https://CRAN.R-project.org/package=exactextractr Douglas Bates Martin Mächler Ben Bolker Steve Walker 2015 publication Journal of Statistical Software vol. 67, issue 1 n/a Foundation for Open Access Statistic https://doi.org/10.18637/jss.v067.i01 Roger Bivand Tim Keitt Barry Rowlingson 2020 R package version 1.5-16 publication Bivand, R, et al. (2020) rgdal: Bindings for the ‘geospatial’ data abstraction library. R package version 1.5-16. https://CRAN.R-project.org/package=rgdal https://CRAN.R-project.org/package=rgdal Robert J. Hijmans 2020 R package version 3.3-13 publication Hijmans, RJ (2020) raster: Geographic data analysis and modeling. R package version 3.3-13. https://CRAN.R-project.org/package=raster https://CRAN.R-project.org/package=raster Michael S. O'Donnel Drew A. Ignizio 2012 publication n/a US Geological Survey https://doi.org/10.3133/ds691 Michael S. O'Donnell David R. Edmunds Cameron L. Aldridge Julie A. Heinrichs Adrian P. Monroe Peter S. Coates Brian G. Prochazka Steve E. Hanser Lief A. Wiechman Thomas J. Christiansen Avery A. Cook Shawn P. Espinosa Lee J. Foster Kathleen A. Griffin Jesse L. Kolar Katherine S. Miller Ann M. Moser Thomas E. Remington Travis J. Runia Leslie A. Schreiber Michael A. Schroeder San J. Stiver Nyssa I. Whitford Catherine S. Wightman 202107 publication Ecological Informatics vol. 63 n/a Elsevier BV ppg. 101327 https://doi.org/10.1016/j.ecoinf.2021.101327 No formal attribute accuracy tests were conducted. No formal attribute accuracy tests were conducted. Dataset is complete for the information presented in the abstract. Users are advised to read all associated metadata carefully. A formal accuracy assessment of the horizontal positional information in the data set has not been conducted. A formal accuracy assessment of the vertical positional information in the data set has not been conducted. Jeffrey J. Danielson Dean B. Gesch 2011 publication n/a US Geological Survey https://doi.org/10.3133/ofr20111073 Digital and/or Hardcopy 2011 publication date DEM 30 arc-second Covariate layer used for modeling and spatial predictions. U.S. Geological Survey 20180404 raster digital data U.S. Geological Survey. 2018. 1/3rd arc-second digital elevation models (DEMs) - USGS National Map 3DEP downloadable data collection. U.S. Geological Survey, Reston, Virginia, USA. Available at https://www.usgs.gov/core-science-systems/ngp/3dep/about-3dep-products-services. Accessed 4 Apr. 2018. https://www.usgs.gov/core-science-systems/ngp/3dep/about-3dep-products-services https://viewer.nationalmap.gov/basic/?basemap=b1&category=ned,nedsrc&title=3DEP%20View Digital and/or Hardcopy 20180404 publication date DEM 1/3 arc-second The DEM data were downloaded as tiles and merged from the URL linked provided above. Spatial elevation data used to derive terrain indices for compound topographic index (CTI) and vector ruggedness measure (VRM). These indices were used as covariates for modeling and spatial predictions. Michael J. Falkowski Jeffrey S. Evans David E. Naugle Christian A. Hagen Scott A. Carleton Jeremy D. Maestas Azad Henareh Khalyani Aaron J. Poznanovic Andrew J. Lawrence 201701 publication Rangeland Ecology & Management vol. 70, issue 1 n/a Elsevier BV ppg. 15-24 https://doi.org/10.1016/j.rama.2016.08.002 Digital and/or Hardcopy 2017 publication date Pinyon-juniper cover Covariate layer used for modeling and spatial predictions. Original data were available at a 1-m resolution. We resampled to a 10-m resolution and reclassified pixels in the 4-10% cover categories as 1 and all other remaining categories as 0 to estimate proportion of early-stage pinyon-juniper cover. Collin Homer 2019 dataset https://www.sciencebase.gov U.S. Geological Survey https://doi.org/10.5066/p9mjvqsq Digital and/or Hardcopy 2016 ground condition Sagebrush cover Covariate layer used for modeling and spatial predictions. The sagebrush shrubland fractional component layer was used. Matthew Rigge 2019 dataset https://www.sciencebase.gov U.S. Geological Survey https://doi.org/10.5066/p9ltu2qm Digital and/or Hardcopy 1981 2010 observed Precipitation Source dataset used to create the mean (of Bio 12) annual precipitation layer. U.S. Geological Survey and U.S. Department of Agriculture Unknown tabular digital data https://landfire.gov/lf_remap.php Digital and/or Hardcopy 2016 ground condition 2016 LANDFIRE EVT Source dataset used to create sagebrush habitat and landscape metric covariate layers. Prior to producing lek persistence maps we processed spatial predictor data layers (i.e., covariates) for modeling. The lek persistence model included 13 covariates that were extracted from these processed raster datasets. The covariate raster datasets included a combination of publicly available data (these layers were not modified beyond simple operations such as reprojecting or masking prior to covariate extractions), proprietary data (layers provided to us by partners internally that were not publicly available for distribution), and interim data (intermediate layers we created solely for the purpose of this project). All raster datasets were reprojected to Albers Conical Equal Area in the North American Datum 1983 and resampled to a 30-m resolution unless their native attributes matched these specifications. All raster datasets covered the full extent of the U.S. greater sage-grouse distribution. Below, we describe the covariate raster datasets in their native (starting) format. Unless noted, raster processing was conducted in ArcMap v. 10.7. Boundary distance (interim layer): We calculated a Euclidean distance raster that measured the distance (in meters) to the U.S. greater sage-grouse distribution boundary for every pixel. The greater sage-grouse distribution boundary was an internally produced USGS product unavailable to the public. Elevation (publicly available layer): We used a global 30 arc-second (1-km resolution) digital elevation model (GTOPO30) available from USGS to represent elevation in meters (Danielson and Gesch 2011: https://doi.org/10.3133/ofr20111073). All sagebrush cover (publicly available layer): We used the National Land Cover Dataset (NLCD) 2016 Shrubland Fractional Components for the Western U.S. (ver. 2.0, October 2019) sagebrush shrubland fractional component layer to represent continuous sagebrush cover as a percentage of each pixel (Homer 2019: https://doi.org/10.5066/P9LTU2QM). Pinyon-juniper cover (publicly available layer): Conifer cover was represented using a modeled raster dataset representing conifer conditions from 2012-2013 (Falkowski et al. 2017: https://doi.org/10.1016/j.rama.2016.08.002). We modified this layer by reclassifying it into a binary layer, distinguishing conifer cover classes between 4-10% (1 values) from all other classes (0 values). Sagebrush habitat (publicly available layer): The BLM provided a binary sagebrush classified habitat layer. The layer was created by reclassifying the 2016 LANDFIRE Existing Vegetation Type (EVT) 2.0.0 dataset categories (7064, 7072, 7079, 7080, 7124, 7125, 7126) as sagebrush habitat (1 values), and all other EVT categories as non-sagebrush habitat (0 values). Sagebrush habitat and sagebrush-associated habitat (publicly available layer): The BLM provided a binary layer that distinguished sagebrush habitat and sagebrush-associated habitat from all other categories. The layer was created by reclassifying the 2016 LANDFIRE Existing Vegetation Type (EVT) 2.0.0 datasets categories to capture both sagebrush habitat (7064, 7072, 7079, 7080, 7124, 7125, 7126) and sagebrush-associated habitat (7062, 7065, 7081, 7085, 7086, 7107, 7123, 7127, 7141, 7148, 7250) as 1 values, and all other EVT categories as 0 values. This layer did not appear in the lek persistence model as a stand-alone covariate, but was used for landscape metric covariate calculations (described below). Landscape metrics (interim layer): We created 3 landscape class metric raster layers, including core area index (a core area metric), edge density (an edge metric), and clumpiness index (an aggregation metric), at the class level within circular buffers that represented different landscape scales (ranging from 1- to 6.4-km radii) using the package landscapemetrics in R. We applied the landscape class metrics to a raster layer that was the sum of the sagebrush habitat and sagebrush-associated habitat rasters. Landscape metrics were computationally intensive to calculate and could not be directly calculated for every 30-m pixel within the greater sage-grouse distribution. To overcome this, we produced a sample grid of points where landscape metrics were calculated at every point for each of the 3 metrics, and then we interpolated landscape metric values between points using inverse weighted distance interpolation implemented with the IDW tool in ArcMap to create a continuous raster dataset. A full description of this process is provided in the supplemental materials document (SM1.4) associated with Wann et al. (In review). Vector ruggedness measure (interim layer): The vector ruggedness measure (VRM) was derived from a 1/3 arc-second (10-m resolution) digital elevation model (USGS 3D Elevation Program: https://apps.nationalmap.gov/downloader/) and applying transformations from the methodology of Sappington et al. (2007). Processing occurred in ArcMap v. 10.6. Pixel values represented a decomposition of slope and aspect into 3-dimensonal vectors summarized within a 1000-m moving window, which provided a measure of terrain ruggedness, independent of slope. Compound topographic index (interim layer): The compound topographic index (CTI) was derived from a 1/3 arc-second (10-m resolution) digital elevation model (USGS 3D Elevation Program: https://apps.nationalmap.gov/downloader/) and applying transformations from the methodology of Gessler et al. (1995). Processing occurred in ArcMap v. 10.6. CTI represents a steady-state wetness, a function of slope and upstream contributing area per unit width that is orthogonal to the flow direction (i.e., ratio of catchment area and slope). Flatter areas will have large values (typically not informative), while smaller catchments with steep slopes will reflect low values of the index. We used the D8 algorithm to define our catchment area (flow accumulation) and the following equation: ln〖(((flow accumulation+1)* 〖pixel area〗_meters ))/tan(〖slope〗_radians). CTI is often used as a surrogate for soil moisture and vegetative productivity; however, this index does not work well for larger geographic extents where precipitation differs. Therefore, we defined our flow accumulation by weighting it using 1980 – 2010 total precipitation normal. We first calculated the index for a 3x3 cell. We then applied moving window scales to create data reflecting different functional scales with aforementioned statistics. Point disturbance (proprietary data layer): This dataset represents a 30-meter resolution point density analysis derived from an aggregate of twelve merged feature classes representing various anthropogenic structures. To produce this aggregate point layer a suite of datasets were pre-processed as detailed below and combined in ArcMap 10.6. A Point Density Analysis (ArcMap 10.6.1 environment) was run using the point feature class to calculate a magnitude-per-square kilometer of point features that fall within a 534-meter search radius around each 30-meter cell. The process was limited to an extent covering the range of greater sage-grouse in the western US. A description of the feature classes follows. (1) Power Plants: Acquired September 10th, 2016 via data subscription contract. Copyright McGraw-Hill / S&P Global Platts. The Platts Power Plant geospatial data layer contains point features representing Utility and Non-Utility operated power plants in North America. Data was selected based on attribution using ArcGIS to create a subset including only operational power plants with a locational accuracy of 1 mile or less. (2) Oil, Gas, and Geothermal Wells Currently Active and Active within the past 10 years: A subscription to the IHS Enerdeq Browser (https://ihsmarkit.com/products/oil-gas-tools-enerdeq-browser.html) was used to extract information representing wells active after 9/1/2016 and wells having a last production or last activity date prior to 9/1/2006. These data were exported into a feature class where duplicate API numbers were removed using the Delete Identical function in ArcMap. Spatially duplicate features were removed using the Delete Identical function in ArcMap on the Shape field. Additional data were extracted from the BLM’s Automated Fluid Support System (AFMSS) by Joy Fleet, BLM Program Analyst, on 09/30/2016. Data Contact: AFMSS Program Manager Michael Mulder, mmulder@blm.gov. The extraction included all currently active well and inactive wells which were active within the 10 years prior to 2016. Data includes Federal, Indian, FEE and State wells that are identified in federal or Indian agreements for the following states: CA, CO, ID, MT, ND, NE, NV, OR, SD, UT, WA & WY. These data were provided in the form of an excel spreadsheet and those with Latitude and Longitude information were mapped using the “Add X/Y Data” function in ArcMap. Duplicate API numbers were removed using the Delete Identical function in ArcMap on the API number field. Spatially duplicate features were removed using the Delete Identical function in ArcMap on the Shape field. Duplicate wells between the IHS and AFMSS datasets, as identified by the API number, were removed from the IHS data. Spatially duplicate locations between the IHS and AFMSS well data were removed from the IHS dataset. (3) Vertical Structures : Federal Aviation Administration (FAA) Aeronautical Information Management Terrain and Obstacles Data Team 1305 East-West Highway Silver Spring, MD 20910-3281. https://www.faa.gov/air_traffic/flight_info/aeronav/digital_products/dof/. Accessed on September 29, 2016. This data is a subset of the one file containing digital obstacle data within all the FAA Regions plus some areas of Canada, Mexico, the Caribbean, the Bahamas, and Pacific Areas. Locations of features were in degree, minutes, and seconds format before being converted into decimal degree and mapped using the Add X, Y data function in ArcMap and exported. Spatially duplicate features were removed using the Delete Identical function in ArcMap on the Shape field. (4) Communication Towers: Melvin C. Del Rosario Federal Communications Commission Wireless Telecommunications Bureau Spectrum and Competition Policy Division 445 12th St. SW Portals I Room 6227 Washington, DC 20554 Phone: (202)418-0615 e-mail: melvin.delrosario@fcc.gov. http://wireless2.fcc.gov. Accessed on September 29, 2016 for the following states: AZ, CA, CO, ID, KS, MT, NE, OK, NV, NM, ND, OR, SD, TX, UT, WA, WY. Spatially duplicate features were removed using the Delete Identical function in ArcMap on the Shape field. This data layer is not publicly available owing to part of the data being proprietary in nature. Contact: Anthony J. Titolo, Assessments and Monitoring, Natural Resource Specialist, BLM National Operations Center, Denver Federal Center, Building 50, atitolo@blm.gov, 719-396-4841. Major road line disturbance (proprietary data layer): This dataset represents a 30-meter resolution density analysis raster derived from linear disturbance features representing highways and major roads. These road data were extracted from licensed ESRI StreetMaps Premium data (2016 - Release #2, Copyright ESRI and its licensors - All rights reserved). Highways and major roads were extracted from the ESRI StreetMaps Premium data using attribute queries (ROAD_CL IN [2, 3, 6]). A Line Density Analysis (ArcMap 10.6.1 environment) was run using the subset of the line features meeting the attribute query criteria to calculate a magnitude-per-square kilometer of linear features that fall within a 534-meter search radius around each 30-meter cell. The process was limited to an extent covering the range of greater sage-grouse in the western US. This data layer is not publicly available owing to the data being proprietary in nature. Contact: Anthony J. Titolo, Assessments and Monitoring, Natural Resource Specialist, BLM National Operations Center, Denver Federal Center, Building 50, atitolo@blm.gov, 719-396-4841. Mean annual precipitation (publicly available layer): The USGS calculated a mean annual precipitation variable using the formula for the Bio 12 statistic (O’Donnell and Ignizio 2012: https://pubs.usgs.gov/ds/691/), which was the summation of annual monthly precipitation (measured in millimeters). Mean annual precipitation was therefore the average of Bio 12 for years 1981-2010 (20-year normal). This product used publicly available 800-m resolution down-scaled PRISM monthly precipitation data (Climate Source, Inc, 2011: http://www.climatesource.com/) to produce the mean annual precipitation layer. DEM 30 arc-second DEM 1/3 arc-second 2016 LANDFIRE EVT Pinyon-juniper cover Sagebrush cover Precipitation 2021 The lek persistence model was fit to lek covariates extracted from the covariate layers produced in Step 1. Briefly, we extracted environmental predictor data at lek locations at 5 circular buffer extents: 1-, 3.2-, 6.4-, 15.0-, and 30-km radii. These extracted values became covariates in a generalized mixed effects logistic regression model relating the binary classification of a lek (as being either active [1] or inactive [0]) to the covariates. For each covariate, a single buffer (scale) was selected that led to the most predictive model fit. Prior to model fitting, all extracted covariates values were z-transformed (i.e., covariate values were centered by subtracting the mean and scaled by dividing by the standard deviation). The lek persistence R model object was used to produce spatial predictions of conditional probabilities of lek persistence across the full extent of the U.S. greater sage-grouse distribution at a 30-m resolution. Predictions were made using the predict function from the raster library in R. The predict function required two arguments: (1) the model object which contained the model formula, and (2) a raster stack with layers corresponding to the predictor variables fit in the model object. Because the covariates fit in the model were summarized at various buffer extents followed by z-transformation, the original raster predictor layers (which represented data on the original untransformed scale) had to be similarly matched. We did this by first calculating moving window rasters that matched the buffer for a given covariate as it appeared in the final model. For example, sagebrush cover was best supported when summarized at a 3.2-km buffer, which was the scale used to represent sagebrush cover in the final model. Therefore, we calculated a moving window raster with a 3.2-km radius from the original sagebrush layer, to match the scale of the covariate as it appeared in the model. We then z-transformed the moving window raster of sagebrush cover using the same mean and standard deviation as was used in the original extracted covariate z-transformation. This process was repeated for all remaining covariates that appeared in the final model. The final model also included a random effect for the Bureau of Land Management (BLM) mid-scale, which were polygons delineating different habitat assessment regions within the U.S. greater sage-grouse distribution. For categorical predictor variables, the raster predict function required a categorical raster. We converted the BLM mid-scale polygon shapefile into a categorical raster, with each pixel value taking on the value of the polygon it fell within (i.e., the polygon mid-scale identifier, which was a number). Therefore, when we predicted across the greater sage-grouse range, our predicted probabilities of lek persistence were conditional on the factor levels of the mid-scale, which was treated as a random effect. We masked the predicted conditional probability of lek persistence layer to the U.S. greater sage-grouse distribution polygon, which was buffered by 2-km to ensure all leks used in model fitting fell within the prediction space. DEM 30 arc-second DEM 1/3 arc-second Pinyon-juniper cover Sagebrush cover Precipitation 2016 LANDFIRE EVT 20220330 We binned the predicted conditional probability surface into “habitat quality” categories corresponding to different levels of model sensitivity (true positive classification rate of leks being active at different probability threshold values). For each mid-scale, we calculated probabilities associated with sensitivities at 0–50% (high-quality), 50–75% (medium-quality), 75–95% (low-quality), and 95–100% (marginal). Using this strategy, the high-quality habitat bin included areas with highest estimated probabilities of lek persistence, while the marginal habitat bin included lowest probabilities. Using R and the raster package, we iterated through each BLM mid-scale and conducted binning in the following steps: (1) we clipped and masked the continuous conditional probability surface to the extent of the focal mid-scale; (2) we reclassified probabilities to values of 1, 2, 3, or 4, based on the associated sensitivity thresholds that corresponded to high-quality, medium-quality, low-quality and marginal habitats. We mosaiced the resulting reclassified mid-scale rasters into a single raster, which resulted in the binned habitat raster. 20220330 Raster Grid Cell 43807 51258 1 Albers Conical Equal Area 29.5 45.5 -96.0 23.0 0.0 0.0 row and column 30.0 30.0 meters North_American_Datum_1983 GRS 1980 6378137.0 298.257222101004 lek_persistence_probabilities.tif Raster geospatial data file. Producer Defined Value Unique numeric values contained in each raster cell. Values represent predicted conditional probabilities of lek persistence (continuous values ranging from 0 to 1). Producer Defined 3.7080588599636e-11 0.99872612953186 lek_persistence_bins.tif Raster geospatial data file. Producer Defined OID Internal object identifier. Producer Defined Sequential unique whole numbers that are automatically generated. Value Unique numeric values contained in each raster cell. Producer Defined 1 Probability bins associated with sensitivities at 0-50% (high-quality) Producer defined 2 Probability bins associated with sensitivities at 50-75% (medium-quality) Producer defined 3 Probability bins associated with sensitivities at 75-95% (low-quality) Producer defined 4 Probability bins associated with sensitivities at 95-100% (marginal-quality) Producer defined Count Number of raster cells with this value. Producer Defined 221517708.0 High-quality Producer defined 133569261.0 Medium-quality Producer defined 209477856.0 Low-quality Producer defined 304585900.0 Marginal-quality Producer defined GS ScienceBase U.S. Geological Survey mailing address

Denver Federal Center, Building 810, Mail Stop 302

Denver CO 80225 United States 1-888-275-8747 sciencebase@usgs.gov Unless otherwise stated, all data, metadata and related materials are considered to satisfy the quality standards relative to the purpose for which the data were collected. Although these data and associated metadata have been reviewed for accuracy and completeness and approved for release by the U.S. Geological Survey (USGS), no warranty expressed or implied is made regarding the display or utility of the data for other purposes, nor on all computer systems, nor shall the act of distribution constitute any such warranty. Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the U.S. Government. Digital Data https://doi.org/10.5066/P95YAUPH None 20221102 FORT Metadata Specialist U.S. Geological Survey, Fort Collins Science Center FORT Metadata Specialist mailing address

2150 Centre Avenue Bldg C

Fort Collins CO 80526 US 970-226-9100 970-226-9230 fortdatamanagement@usgs.gov FGDC Biological Data Profile of the Content Standard for Digital Geospatial Metadata FGDC-STD-001.1-1999