Fish Data Compilation and Climate Change Assessment for Desert Landscape Conservation Cooperative Fishes

Freshwater fishes are globally among the most imperiled major biodiversity groups and they are especially endangered in the North American deserts of the vast binational Desert LCC. Sixty seven native fish species of conservation concern are in the study area, which includes all of the Desert Landscape Conservation Cooperative (DLCC) in both the US and Mexico.

Essentially all species in our study area are understudied and management of them has been greatly impeded by the intrinsic difficulties of working internationally and by relative lack of, or inaccessibility to, basic knowledge about their distributions and conservation status. We propose to mine data from all online and known US-based institutions holding specimen-based occurrence records from our study area. We will normalize and generally improve data quality to provide a comprehensive, high quality resource that brings together in one GIS-accessible database all of the currently very scattered and relatively un-normalized museum-based records.

We will focus our efforts on data for the Rio Grande basin, which will receive more rigorous and thorough normalization, and manual georeferencing with precision estimates, than will data for the remainder of the study area. In the Rio Grande, we will also do basic quality control on taxonomy and georeferencing following published protocols and use the data to produce Species Distribution Models (SDMs) for selected priority, special interest native and invasive fishes.

SDMs will be constructed for present conditions and three projected climate change scenarios to allow us to assess current and projected future status throughout each species’ range, thus filling vast information gaps throughout data-poor areas in Mexico that might prove vital as source or sink habitats. Projected future distributions will identify landscape-level areas of conservation and restoration priority that may not presently be of high priority, but that may become so in the future. The varying projections under varied climate change scenarios will allow for quantitative assessment of uncertainties. Both the raw occurrence data and current and future SDMs will be valuable tools for diverse future work on regional aquatic biodiversity sustainability in the face of climate change.

Project Goals

We will produce data and decision support tools for the conservation, restoration, and management of U.S. priority freshwater fishes in drainages shared by the U.S. and Mexico by compiling and normalizing biodiversity data for all fishes occurring in internationally shared drainages of the DLCC, exclusive of the Colorado and Gila drainages. We will then focus on the Rio Grande drainage where we will model current distributions of selected special interest fishes and project the models into the future under three different climate change scenarios. The results will demonstrate how changing climates will impose directional pressures that will likely tend to shift species distributions.

Anticipated Outcomes

• Biodiversity data in Microsoft Excel (or other requested format) and on publically accessible Fishes of Texas (FoTX) website
• Environmental data to build species distribution models via University of Texas server
• Research protocols and analysis methods via personal request to the investigators
• Raster data format for future environmental variables for the extent of the focal species’ ranges modeled
• Raster and image data format for current and future projections
• 2 peer-reviewed papers
• Presentations to partners

Principal Investigator

Terence Arundel (USGS)

Final Project Report and Data

Related Project

Final Report: Data provision and projected impact of climate change on fish biodiversity within the Desert LCC

The four primary objectives of this project were to:

1. compile a dataset of fish occurrence records for the entirety of the Rio Grande drainage in the US and Mexico
2. improve that dataset by reformatting dates, synonymizing species names to a modern taxonomy, georeferencing localities, and flagging geographic outliers
3. for those species with sufficient data for modeling, create species distribution models (SDMs)
4. use the environmental conditions determined via those models to project the species distributions into the future under two climate scenarios.

To accomplish those objectives, we compiled 495,101 fish occurrence records mined from 122 original sources into a single database. We then, on the basis of text string searches of the original sources' verbatim locality fields, extracted 145,426 records that we judged to have a reasonable likelihood of being from the Rio Grande drainage. For those records we edited taxonomy, reformatted dates, and finally georeferenced 59,156 (41%) records, which proved sufficient for constructing SDMs for 36 species that met a priori quality assurance criteria. We provide basic interpretation of these models and discuss projections of them into several different future climate forecasts. Products include raw model outputs and symbolized maps helpful in interpretation and comparison, as well as raw data sets and recommendations regarding how all of these product might be used in future management and research efforts.