Assessing ecosystem & population dynamics to determine the primary drivers of Devils Hole Pupfish population trends

Authors: John Umek, Kevin Wilson, Jeff Goldstein, Ambre Chaudoin, Mark Hausner

Year: 2023

Abstract

The critically endangered Devils Hole pupfish (DHP), Cyprinodon diabolis, found only in Devils Hole, was in a steady decline in population size beginning in the mid 1990s. The population reached an all-time low of 35 observable fish in the spring of 2013. Since this all-time low the population has increased to 150-250 observable fish. However, the population remains well below historical levels of 400-600 individuals. Because the reasons for the steady decline and subsequent increase are not fully understood, the multi-agency management team established a Long-Term Ecosystem Monitoring Program (LTEMP) in 2011 to collect ecosystem level information on the ecosystem processes and population dynamics of Devils Hole and the DHP. In 2020, the Desert Research Institute (DRI) began a project to compile, organize, and document these data and make them accessible to researchers and resource managers. This project has two goals: (1) to compile the existing LTEMP data into a digitally ingestible format that can be easily accessed by researchers seeking to analyze the data; and (2) to perform data analysis to examine and quantify relationships between different drivers of habitat quality and metrics of the Devils Hole pupfish population. To accomplish this second goal we examine ecosystem and population dynamics through random forest modeling, a machine learning algorithm. The random forest technique can handle complex, multi-dimensional datasets and capture nonlinear relationships. By analyzing historical data, researchers and policymakers can better understand connections between different drivers of variability in the DHP population and make more informed decisions to minimize negative impacts and develop sustainable management practices. We use the random forest technique to examine the interactions between biotic and abiotic parameters measured in Devils Hole and the observed DHP population. Here we present the results of the analyses, identifying the most significant drivers of variability in the DHP population over the last 15 years.