These data contain the results from the North American Bat Monitoring Program's (NABat) integrated species distribution model (iSDM) for tricolored bats (Perimyotis subflavus). The provided tabular data include predictions (with uncertainty) for tricolored bat occupancy probabilities (i.e., probability of presence) based on data from the entire summer season (May 1–Aug 31), averaged from 2017-2022, in each NABat grid cell (5km x 5km scale) across the range of the species. Specifically, predictions represent occupancy probabilities in the pre-volancy season in the summer (May 1 – July 15), i.e., the period of time before juveniles can fly and become detectable. Predictions were produced using an analytical pipeline supported by web-based infrastructure, Bayesian hierarchical modeling, and iSDM framework which integrated stationary acoustic, mobile transect acoustic), and live-capture data to model the recent summer distribution of the species while accounting for imperfect detection and species misclassification. A tabular file is included detailing the average occupancy probability predictions (from 2017-2022) of each 5 km x 5 km grid cell in the species range, including means, standard deviations, and the 95% Bayesian credible intervals. These data can be cross-referenced to the NABat CONUS 5km master sample for analytical or visualization purposes.

False positive occupancy analysis predictions with model uncertainty based on summertime data provided to support the three bat species status assessment (SSA) for Myotis lucifigus (MYLU), Myotis septentrionalis (MYSE), and Perimyotis subflavus (PESU). The objectives outlined by the Fish and Wildlife Service’s SSA team were to estimate summertime distributions across the entire species range. Statistical analysis included five types of response data requested from the North American Bat Monitoring Program database (NABat): automatically identified stationary acoustic calls, manually vetted stationary acoustic calls, automatically identified mobile acoustic calls, manually vetted mobile acoustic calls, and capture records. Statistical analysis was for the summertime distribution modeling, data collected between June 1 and Sept 1 during 2010 until 2019 were only included.

This data release contains the North American Bat Monitoring Program (NABat) Master Sampling Grid at the 5 km x 5 km scale with biologically relevant covariates for NABat analyses attributed to each cell of the 5 km x 5 km grid frame for the continental United States. It was created using ArcPro and the 'sf', 'tidyverse', 'dplyr' and 'exactextractr' packages in R to extract covariates from multiple data sources following the 10 km x 10 km attributed grid process as well as adding additional covariates. These covariates include the habitat characteristics such as percent of wetlands, forest, deciduous and coniferous forest, dominant and subdominant oak types, the number of tree and oak species, topographic features such as physiographic diversity, elevation, and the presence of karst terrain features or water feature, climate variables such as mean temperature and precipitation, and subterranean human structures such as the number and length of culverts. This layer provides the predictive covariates used in the integrated species distribution model for tricolored bats (Perimyotis subflavus, see External Related Resources). The attributed grid can also support future modeling efforts and data visualizations.

These data contain supplemental results and model code from the North American Bat Monitoring Program's (NABat) integrated species distribution model (iSDM) for the tricolored bat (Perimyotis subflavus). These data also serve as supplemental results, source data used to create figures, and model code to the companion manuscript: "Integrated distribution modeling resolves asynchrony between bat population impacts and occupancy trends through latent abundance " published in Communications Biology. Predictions were produced using an analytical pipeline supported by web-based infrastructure, Bayesian hierarchical modeling, and multi-scale integrated species distribution modeling (MS-iSDM) framework which integrated stationary acoustic, mobile transect acoustic, and live-capture data to model the recent summer distribution of the species while accounting for imperfect detection and species misclassification. The provided tabular data include predictions (with uncertainty) for tricolored bat summer distributions (relative abundance and occupancy probability) based on data from the entire summer season (May 1–Aug 31), for each from 2012-2022. Predictions represent relative abundances and occupancy probabilities in the pre-volancy season in the summer (May 1 – July 15), i.e., the period of time before juveniles can fly and become detectable. Results are summarized at 4 different spatial scales (Range-wide, state-level, 10 kilometer (km) x 10 km grid-cells, and 5 km x 5 km quadrants). At the grid-cell level, predictions (with uncertainty) are provided for relative abundance each year (2012-2022), and the overall proportional change in relative abundance between 2012-2022. At the quadrant level, predictions (with uncertainty) are provided for occupancy probabilities (i.e., probability of presence) each year (2012-2022), and for the overall proportional change in occupancy probability between 2012-2022. At the state-level, average relative abundance (across all grid cells) and average occupancy probability (across all quadrants) is provided for each state and year. Trend estimates for total proportional change between 2012-2022 are also provided for each state for average relative abundance and average occupancy probability, while additional trend metrics (absolute change) between 2012-2022 are provided for average occupancy probability. At the range-wide scale, average relative abundance (across all grid cells) and average occupancy probability (across all quadrants) is provided for each year, along with the overall trends in both metrics from 2012-2022. Predictions at the grid cell (10km x 10km) and quadrant (5km x 5km) can be cross-referenced to the NABat CONUS 5km master sample and/or NABat CONUS 10km master sample for analytical or visualization purposes (see related products). Model code was provided to document the JAGS model used to produce the results. Parameter estimates from the final model and model comparisons used to make figures in the manuscript are also provided.

The dataset is comprised of historical observations and predictions of winter colony counts at known sites for three bat species (little brown bat, Myotis lucifugus; tricolored bat, Perimyotis subflavus; and big brown bat, Eptesicus fuscus). The dataset consists of two separate but related data files in tabular format (comma-separated values [.csv]). Each data set consists of predicted winter counts derived using winter status and trends modeling methods developed by the North American Bat Monitoring Program (NABat). These two predicted winter count data sets were used to inform NABat summertime status and trends analysis: 1) modeled abundance predictions for all hibernacula for all three species from 2010-2021, and 2) modeled abundance predictions for P. subflavus from 2010-2023 using updated monitoring data. Abundance predictions were derived with a combined modeling approach that applied an exponential linear interpolation model (when there were less than 4 observations per location) and a Bayesian hierarchical model (where there were 4 or more data points per location).

The dataset is comprised of historical observations and predictions of winter colony counts at known sites for three bat species (little brown bat, Myotis lucifugus; tricolored bat, Perimyotis subflavus; and big brown bat, Eptesicus fuscus). The dataset consists of two separate but related data files in tabular format (comma-separated values [.csv]). Each data set consists of predicted winter counts derived using winter status and trends modeling methods developed by the North American Bat Monitoring Program (NABat). These two predicted winter count data sets were used to inform NABat summertime status and trends analysis: 1) modeled abundance predictions for all hibernacula for all three species from 2010-2021, and 2) modeled abundance predictions for P. subflavus from 2010-2023 using updated monitoring data. Abundance predictions were derived with a combined modeling approach that applied an exponential linear interpolation model (when there were less than 4 observations per location) and a Bayesian hierarchical model (where there were 4 or more data points per location).

False positive occupancy analysis predictions with model uncertainty based on summertime data provided to support the three bat species status assessment (SSA) for Myotis lucifigus (MYLU), Myotis septentrionalis (MYSE), and Perimyotis subflavus (PESU). The objectives outlined by the Fish and Wildlife Service’s SSA team were to estimate summertime distributions across the entire species range. Statistical analysis included five types of response data requested from the North American Bat Monitoring Program database (NABat): automatically identified stationary acoustic calls, manually vetted stationary acoustic calls, automatically identified mobile acoustic calls, manually vetted mobile acoustic calls, and capture records. Statistical analysis was for the summertime distribution modeling, data collected between June 1 and Sept 1 during 2010 until 2019 were only included.

False positive occupancy analysis predictions with model uncertainty based on summertime data provided to support the three bat species status assessment (SSA) for Myotis lucifigus (MYLU), Myotis septentrionalis (MYSE), and Perimyotis subflavus (PESU). The objectives outlined by the Fish and Wildlife Service’s SSA team were to estimate summertime distributions across the entire species range. Statistical analysis included five types of response data requested from the North American Bat Monitoring Program database (NABat): automatically identified stationary acoustic calls, manually vetted stationary acoustic calls, automatically identified mobile acoustic calls, manually vetted mobile acoustic calls, and capture records. Statistical analysis was for the summertime distribution modeling, data collected between June 1 and Sept 1 during 2010 until 2019 were only included.

False positive occupancy analysis predictions with model uncertainty based on summertime data provided to support the three bat species status assessment (SSA) for Myotis lucifigus (MYLU), Myotis septentrionalis (MYSE), and Perimyotis subflavus (PESU). The objectives outlined by the Fish and Wildlife Service’s SSA team were to estimate summertime distributions across the entire species range. Statistical analysis included five types of response data requested from the North American Bat Monitoring Program database (NABat): automatically identified stationary acoustic calls, manually vetted stationary acoustic calls, automatically identified mobile acoustic calls, manually vetted mobile acoustic calls, and capture records. Statistical analysis was for the summertime distribution modeling, data collected between June 1 and Sept 1 during 2010 until 2019 were only included.

These data contain the supplementary results corresponding with the journal article: Using mobile acoustic monitoring and false-positive N-mixture models to estimate bat abundance and population trends by Udell et al. (2024) in Ecological Monographs. These results contain the findings from the North American Bat Monitoring Program's (NABat) "Summer Abundance Status and Trends" analyses which used mobile transect acoustic data for three species (tricolored bat, little brown bat, and big brown bat). Data from the entire summer season (May 1–Aug 31) were used in the modeling process. Here, tabular data for each species include predictions (with uncertainty) of relative abundance (and trends over time) in the summer maternity season (May1 - July 16) from 2012-2020. Predictions for status and trends are provided for each species at four different spatial resolutions: 1) across the modeled species ranges, 2) at the state or province level, 3) at the NABat grid cell (10km x 10km scale) level, and 4) for each sampled transect. Predictions were produced using an analytical pipeline supported by web-based infrastructure, Bayesian hierarchical modeling, and 'false-positive N-mixture model' framework which analyzed mobile transect acoustics to model the relative abundance distribution (and trends over time) of each species while accounting for imperfect detection and species misclassification. Tabular files provided include: 1) range-wide average relative abundance predictions by year for each species, 2) range-wide trends in average relative abundance for each species, 3) regional (state/province) average relative abundance by year for each species, 4) regional (state/province) trends in average relative abundance for each species, 5) grid cell-level predictions of relative abundance by year for each species, 6) grid cell-level trends (overall change from 2012-2020) for each species (one file per species), and 7) transect-level estimates of relative abundance by year for each species. Estimates include means, medians, standard deviations, and the 95% Bayesian credible intervals. These data can be cross-referenced to the 'knitted' NABat master sample for CONUS, Canada, and Alaska (NABat_grid_covariates.shp, which is available on ScienceBase).