The following descriptions and metadata indicate the use and purpose of the information supplied in ‘data.csv,’ and ‘code.txt’ accompanying and supporting the density-dependent habitat selection model for wolves described in the manuscript ‘Territorial landscapes: incorporating density-dependence into wolf resource selection study designs’ by O’Neil et al. (RSOS 190282) This codebook.txt file was generated on by ------------------- GENERAL INFORMATION ------------------- 1. Title of Dataset: "Code, data, and metadata document for the manuscript: Territorial landscapes: incorporating density-dependence into wolf resource selection study designs" 2. Author Information: Bump, Joseph K; Beyer, Dean; O'Neil, Shawn Principal Contact Information Name: Joseph K Bump Institution: University of Minnesota Address: Email: bump@umn.edu 3. Date of data collection: Field Data: January 1994 - June 2014 Data Geoprocessing and preparing data for models: June 2014 - May 2017 4. Geographic location of data collection (where was data collected?): Upper Peninsula, Michigan, USA 5. Information about funding sources that supported the collection of the data: Michigan Department of Natural Resources -------------------------- SHARING/ACCESS INFORMATION -------------------------- 1. Licenses/restrictions placed on the data: CC0 1.0 Universal, http://creativecommons.org/publicdomain/zero/1.0/ 2. Links to publications that cite or use the data: (submitted) "Territorial landscapes: incorporating density-dependence into wolf resource selection study designs" 3. Recommended citation for the data: Bump, Joseph K; Beyer, Dean; O'Neil, Shawn. (2019). code, data, and metadata document for the manuscript: Territorial landscapes: incorporating density-dependence into wolf resource selection study designs. Retrieved from the Data Repository for the University of Minnesota, http://hdl.handle.net/11299/201859. --------------------- DATA & FILE OVERVIEW --------------------- 1. File List A. Filename: data.csv Short description: Geoprocessed data from field locations of wolves (12.35Mb, CSV file) B. Filename: code.txt Short description: statistical code (2.003Kb, Text file) 2. Relationship between files: -------------------------- METHODOLOGICAL INFORMATION -------------------------- 1. Description of methods used for collection/generation of data: Individuals were located by aerial tracking of VHF radio collars. Packs were located using visual sightings from aerial tracking and by ground tracking and following wolf tracks from roads and trails during each winter. 2. Methods for processing the data: Delineation of wolf territories was based on 95% kernel density isopleth analysis when 30 or more locations were available. Otherwise, a combination of VHF locations and ground tracking locations were used to delineate the territory. Used (0 vs. 1) column represents an approximate wolf pack location (randomly sampled within each occupied pack territory for each year of study) or random location (randomly sampled from available space primarily outside of all pack territories for each year of study) Covariate data represent the values of underlying landscape attributes, generated from Geographic Information Systems and remotely sensed data sources. These are described in "DATA-SPECIFIC INFORMATION" and within the manuscript. 3. Instrument- or software-specific information needed to interpret the data: Run code.txt on data.csv in RStudio (R version 3.5.2 (2018-12-20)) after installing the r-inla library (http://www.r-inla.org/download) as well as listed dependencies rgraphviz (https://www.bioconductor.org/packages/release/bioc/html/Rgraphviz.html) and graph (https://bioconductor.org/packages/3.8/bioc/html/graph.html). ----------------------------------------- DATA-SPECIFIC INFORMATION FOR: data.csv ----------------------------------------- COLUMN DESCRIPTION Year Represents a wolf’s biological year, corresponding to the time period during which new pups are added to the population until the same time the following year. Used as a random intercept effect within hierarchical generalized linear mixed model structure. Pack Unique wolf pack territory identifier. Wolf packs and territories were established by information from VHF radio telemetry locations combined with intensive ground tracking to delineate territory boundaries. Used as a random intercept effect within hierarchical generalized linear mixed model structure. Used This column represents the binary response variable in a hierarchical model for wolf (Canis lupus) density-dependent habitat selection in the Upper Peninsula of Michigan, USA, 1995–2013. Each year & wolf pack territory combination received 5 locations sampled within the territory boundary, paired with 25 locations randomly sampled within unoccupied and available space. The available space for each pack was constrained to be within 165 km of the geographic center of the defined wolf territory, based on knowledge about wolf dispersal distances in the Upper Great Lakes region. 1 = used wolf location, randomly sampled within each occupied pack territory for each year of study 0 = random location, randomly sampled from available space outside of each pack territory for each year of study X X-Coordinate in Michigan GeoRef Projected Coordinate System; values are censored (set to NA) due to sensitive nature of locations of endangered species within the study area. Coordinate information is not necessary for running the model. Y Y-Coordinate in Michigan GeoRef Projected Coordinate System; values are censored (set to NA) due to sensitive nature of locations of endangered species within the study area. Coordinate information is not necessary for running the model. InclProb Represents an underlying inverse probability of occupancy surface. In other words, indicating the probability that the random point was sampled from unoccupied space, given uncertainty in the occupied distribution. ELEV Covariate data representing local elevation (m). Underlying values were extracted to used and random locations using GIS. Detailed description of the data sources are available in the Supplemental Electronic Material accompanying the article. SLOPE Covariate data representing degree slope. Underlying values were extracted to used and random locations using GIS. Detailed description of the data sources are available in the Supplemental Electronic Material accompanying the article. TRASP Covariate data representing transformed aspect. Underlying values were extracted to used and random locations using GIS. Detailed description of the data sources are available in the Supplemental Electronic Material accompanying the article. ROUGH Covariate data representing topographic ruggedness. Underlying values were extracted to used and random locations using GIS. Detailed description of the data sources are available in the Supplemental Electronic Material accompanying the article. IMPERV Covariate data representing proportion of impervious developed surface. Underlying values were extracted to used and random locations using GIS. Detailed description of the data sources are available in the Supplemental Electronic Material accompanying the article. HWY Covariate data representing distance to major road or highway. Underlying values were extracted to used and random locations using GIS. Detailed description of the data sources are available in the Supplemental Electronic Material accompanying the article. STREAM Covariate data representing local density of streams (km / km2). Underlying values were extracted to used and random locations using GIS. Detailed description of the data sources are available in the Supplemental Electronic Material accompanying the article. PLAND Covariate data representing proportion of public and protected land. Underlying values were extracted to used and random locations using GIS. Detailed description of the data sources are available in the Supplemental Electronic Material accompanying the article. DDWC Covariate data representing distance to deer wintering complex habitat patches. Deer wintering complexes represent winter availability of the primary prey source for wolves due to deep snowpack in this study area during winter. Underlying values were extracted to used and random locations using GIS. Detailed description of the data sources are available in the Supplemental Electronic Material accompanying the article. SNOW Covariate data representing annual average snow depth (cm) at each 30 m cell, specific to each each winter. Underlying values were extracted to used and random locations using GIS. Detailed description of the data sources are available in the Supplemental Electronic Material accompanying the article. SNOW_LTA Covariate data representing long-term average snow depth (cm) at each 30 m cell, averaged across all study years. Underlying values were extracted to used and random locations using GIS. Detailed description of the data sources are available in the Supplemental Electronic Material accompanying the article. ROAD Covariate data representing density of minor roads (local roads and trails; km / km2). Underlying values were extracted to used and random locations using GIS. Detailed description of the data sources are available in the Supplemental Electronic Material accompanying the article. EDGE Covariate data representing density of edge between forested and open land cover types. Underlying values were extracted to used and random locations using GIS. Detailed description of the data sources are available in the Supplemental Electronic Material accompanying the article. AG Covariate data representing the proportion of agricultural field and pasture. Underlying values were extracted to used and random locations using GIS. Detailed description of the data sources are available in the Supplemental Electronic Material accompanying the article. OPEN Covariate data representing the proportion of open, non-forested land cover types. Underlying values were extracted to used and random locations using GIS. Detailed description of the data sources are available in the Supplemental Electronic Material accompanying the article. WDENS Covariate data representing regional wolf density. Wolf density was used to model habitat functional responses (e.g., density-dependent habitat selection). A spatial surface was developed and described in the Methods section of the article, and underlying values were extracted to used and random locations using GIS. Detailed description of the data sources are available in the Methods section of the accompanying article. WATER Covariate data representing the proportion of open water and wetland cover types. Underlying values were extracted to used and random locations using GIS. Detailed description of the data sources are available in the Supplemental Electronic Material accompanying the article. PackYear Combination of 'Pack' and 'Year.' Used in post-hoc analysis. p_occ Value representing an estimate of the proportion of the surrounding study area (<165 km radial buffer) occupied by other wolf territories, specific to the current pack. This variable is not currently implemented in the model but could be used as a substitute for time.