WaterSMART: Apalachicola/Chattahoochee/ Flint River (ACF) Basin

Science Center Objects

The Challenge: The DOI WaterSMART (Sustain and Manage America’s Resources for Tomorrow) initiative is developing data and tools to help water managers identify current and future water shortages, for humans and for freshwater ecosystems. Fishes, for example, can decline in diversity and abundance when streamflow becomes too low, for too long.  However, ecologists find that effects of declining streamflow can vary depending on stream characteristics and on traits of local species, confounding predictions of ecological outcomes.  Scientists thus need data on ecological responses to low streamflow in differing physical and biological contexts to better inform water management decisions.

The Challenge: Ecologists generally hypothesize that stream-dwelling fishes and other biota are strongly governed by stream flow regimes.  For example, we expect that fishes will decline in diversity and abundance when stream flows become too low, for too long.  Most of the data to test these hypotheses come from surveys of population status in streams that have differing flow regimes.  Because so many other factors affect stream populations, detecting a clear signal of the effects of streamflow variation have proven difficult. Additionally, because few streams are gaged, methods are needed for estimating stream flow dynamics at locations where biological data are collected.

The Science: We have collected data describing changes in fish assemblages in individual stream sites in relation to stream flow variation through time.  From 2011 through 2017, we conducted seasonal surveys of fishes occupying up to 40 stream sites located in the Blue Ridge, Piedmont and Coastal Plain portions of the ACF basin.  We have applied dynamic occupancy models to these data to estimate effects of low- and high-flow periods on fish species persistence and colonization rates.  In collaboration with USGS hydrologists, we have evaluated the efficacy of using modeled stream flows (as opposed to flows extrapolated from nearest available stream gages) to predict fish responses.  We have also investigated differences in flow effects on fishes in the differing physical settings of upland (mountain and Piedmont) and Coastal Plain streams.

The Future: Our final phase involves applying state-transition models to the fish population data in order to relate stream flows to variation in fish abundances as well as to community turnover.