Shryock, D.F. DeFalco, L.A. Esque, T.C. 20220330 raster digital data Denver, CO U.S. Geological Survey data release https://doi.org/10.5066/P9XQJFEL Daniel F. Shryock Lesley A. DeFalco Todd C. Esque 20220412 publication Ecology and Evolution vol. 12, issue 4 n/a Wiley https://doi.org/10.1002/ece3.8805 Preserving native species diversity is fundamental to ecosystem conservation. Selecting appropriate native species for use in restoration is a critical component of project design and may emphasize species attributes such as life history, functional type, pollinator services, and nutritional value for wildlife. Determining which species are likely to establish and persist in a particular environment is a key consideration. Species distribution models (SDMs) characterize relationships between species occurrences and the physical environment (e.g., climate, soil, topographic relief) and provide a mechanism for assessing which species may successfully propagate at a restoration site. In conjunction with information on species attributes, SDMs facilitate holistic ecosystem restoration by enabling practitioners to identify diverse, resilient assemblages of native species. This project develops SDMs for native species of fundamental ecosystem importance in order to guide restoration of Mojave Desert landscapes. The dataset contained herein provides an SDM for Amsinckia tessellata within its Mojave Desert range based on known occurrences. The purpose of these datasets are to describe the geographic distribution and habitat preferences of native plant species that provide ecosystems services and/or are commonly used in ecological restoration within the Mojave Desert. Each raster dataset represents a probability distribution of habitat suitability, where values range from 0 (very low probability of species occurrence) to 1 (very high probability of species occurrence). Predictions are based on known occurrence records for each species and are subject to model uncertainty. Hence, appropriate scrutiny should be taken when applying predictions to real-world assessments of habitat or other management applications. Users should consult the full metadata provided with each dataset, as well as the Larger Work, for additional details on model algorithms, performance, and uncertainty. 2020 ground condition Complete None planned Mojave Desert -118.9291 -112.6985 37.6714 33.5105 ISO 19115 Topic Category biota None native plant materials development species distribution model USGS Thesaurus native species biogeography habitats maps and atlases USGS Metadata Identifier USGS:61f046b0d34e8b818adc32ac Common geographic areas Mojave California Nevada Utah Arizona Integrated Taxonomic Information System (ITIS) Amsinckia tessellata Kingdom Plantae Subkingdom Viridiplantae Infrakingdom Streptophyta Superdivision Embryophyta Division Tracheophyta Subdivision Spermatophytina Class Magnoliopsida Superorder Asteranae Order Boraginales Family Boraginaceae Genus Amsinckia Species Amsinckia tessellata TSN: 31706 None. Please see 'Distribution Info' for details. The authors of these data request that users direct any questions pertaining to appropriate use or assistance with understanding limitations and interpretation of the data to the individuals/organization listed in the Point of Contact section. Data Manager U.S. Geological Survey, Western Ecological Research Center mailing

3020 State University Drive, Modoc Hall, Suite 4004

Sacramento CA 95819 US 916-278-9485 gs-b-werc_data_management@usgs.gov Funding was provided by the Bureau of Land Management (BLM) Plant Conservation and Restoration Program and Mojave Desert Native Plant Program. R Core Team 2019 3.6.1 application/service Vienna, Austria R Foundation for Statistical Computing https://www.r-project.org/ Attributes were formally assessed via statistical measures of classification accuracy, including the Area under the Receiver Operating Characteristic (AUC) and the True Skill Statistic (TSS). See Larger Work for additional details. Data received review from technicians collecting data and by a project manager. Coordinates were plotted in mapping software to ensure their accuracy, and the raster was visually checked to ensure they covered the full extent. Data are considered complete for the information presented. This raster layer was derived for areas within the Mojave Desert ecoregion. A formal accuracy assessment has not been conducted. Tongli Wang Andreas Hamann Dave Spittlehouse Carlos Carroll 20160608 publication PLOS ONE vol. 11, issue 6 n/a Public Library of Science (PLoS) ppg. e0156720 https://doi.org/10.1371/journal.pone.0156720 Digital and/or Hardcopy 1980 2010 ground condition Climate Climate data were used in statistical models to describe the geographic distribution of native species in the Mojave Desert. U.S. Geological Survey Jesslyn F Brown 2020 raster digital data Sioux Falls, SD U.S. Geological Survey https://doi.org/10.5066/F7PC30G1 http://phenology.cr.usgs.gov http://earthexplorer.usgs.gov Digital and/or Hardcopy 2003 2017 ground condition Phenology Remote sensing phenology data were used in statistical models to describe the geographic distribution of native species in the Mojave Desert. U.S. Geological Survey, National Map 2019 raster digital data Digital and/or Hardcopy 2019 ground condition LiDAR Data from the USGS National Map were used in statistical models to describe the geographic distribution of native species in the Mojave Desert. Consortium of California Herbaria Unpublished material tabular digital data http://ucjeps.berkeley.edu/consortium/ Digital and/or Hardcopy 1950 2020 ground condition CCH Species occurrence data used as input for species distribution models. U.S. National Park Service Unpublished material digital data https://www.nps.gov/im/vmi-products.htm Digital and/or Hardcopy 1998 2014 ground condition NPS Species occurrence data used as input for the species distribution models. We used an ensemble modeling approach to create species distribution models for native species in the Mojave Desert ecoregion. Species occurrence data used as input for the SDM was derived from a variety of public databases including herbaria (Consortium of California Herbaria available at http://ucjeps.berkeley.edu/consortium/; SEInet available at http//:swbiodiversity.org/), vegetation classification studies (National Park Service vegetation inventory products, available at https://www.nps.gov/im/vmi-products.htm), and existing USGS datasets. Species occurrences were mapped and visually assessed to identify potential irregularities in georeferencing prior to model fitting. As covariates in the models, we included spatial layers representing climate, site vegetation characteristics, and topography (see Source Inputs and Larger Work for additional details). Our ensemble modeling approach included three algorithms: generalized additive models (R package mgcv version 1.8-22) (Wood, 2017), random forests (R package randomForest version 4.6-12) (Liaw et al., 2002), and MaxEnt version 3.3.3k (R package dismo version 1.1-4) (Hijmans et al., 2017). For each algorithm, we generated models reflecting all combinations of environmental variables while restricting the total number of terms within any one model to seven to avoid overfitting. Due to the lack of surveyed absence points, we created random selections of pseudo-absences. To account for a pattern of spatial aggregation in the presence points, we first rasterized presences to the modeling resolution (1 square km) and subsequently applied a spatial thinning procedure in which a maximum of three points could be sampled from any 10 square km area. Each model was fit across a series of 50 cross validation runs, with each run consisting of a random sample of pseudo-absences and spatially thinned presence points. For each cross validation run, a 20 percent sample of points was withheld for model evaluation (i.e., 80 percent of points were used for model fitting). Raster surfaces representing candidate models were generated by averaging model predictions across the 50 cross validation runs for each model. Next, ensemble predictions for each algorithm (GAM, Random Forest, and MaxEnt) were generated by taking the weighted average among candidate models from each algorithm based on model performance. Finally, an overall ensemble habitat suitability layer was generated by taking the average of the three individual algorithm ensembles. A layer representing the standard error of the overall ensemble habitat suitability layer was also calculated as the standard deviation in model predictions across all candidate models divided by the square root of the number of candidate models considered. Additional details on the modeling process are available in the Larger Work. References: Hijmans, R. J., Phillips, S., Leathwick, J., and Elith, J. (2017). dismo: Species Distribution Modeling. Retrieved from https://cran.r-project.org/package=dismo Liaw, A., and Wiener, M. (2002). Classification and Regression by randomForest. R News, 2(3), 18-22. Retrieved from http://cran.r-project.org/doc/Rnews/ Wood, S.N. (2017) Generalized Additive Models: An Introduction with R (2nd edition). Chapman and Hall/CRC. 2020 Raster Grid Cell 454 550 1 Universal Transverse Mercator 11 0.9996 -117.0 0.0 500000.0 0.0 row and column 1000.0 1000.0 meters North_American_Datum_1983 GRS 1980 6378137.0 298.2572221010042 AMTE_habitat.tif Raster geospatial data file. Producer defined Value Raster values are a probability distribution of suitable habitat, where values range from 0 (very low probability of species occurrence) to 1 (very high probability of species occurrence). Producer defined 0.0019970783032477 0.998366355896 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 on any other system or for general or scientific purposes, 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. TIF https://doi.org/10.5066/P9XQJFEL None 20220420 Daniel F. Shryrock U.S. Geological Survey, Western Ecological Research Center mailing

3020 State University Drive, Modoc Hall Suite 4004

Sacramento CA 95819 US 916-278-9485 dshryock@usgs.gov FGDC Biological Data Profile of the Content Standard for Digital Geospatial Metadata FGDC-STD-001.1-1999