Estimating drivers and identifying uncertainties in smallmouth bass population dynamics in an invaded river network

Smallmouth bass (Micropterus dolomieu) is an important recreational sportfish and destructive non-native species when introduced into freshwater habitats. There is therefore a need to understand the drivers of, and uncertainties in, smallmouth bass population dynamics for various management objectives. We combined long-term smallmouth bass catch-effort and early life history data from a non-native population in the Green River sub-basin of the upper Colorado River to develop a demographic model that links interannual variability in environmental conditions to recruitment in three river reaches. We used the model to quantify how hydrology, river temperature, and exploitation drive smallmouth bass population dynamics. Early life stages were influenced by timing of hatching and discharge. Dispersal of age-0 fish and density-dependent dynamics were identified as primary sources of uncertainty. Determining the true nature of density-dependent dynamics is important, as the impact of exploitation-based management actions is dependent on the strengths of any density-dependent feedbacks. Our model provides a framework to predict smallmouth bass population responses to future climate conditions, reservoir operations, and exploitation levels.

Methods
We compiled smallmouth bass catch-effort data collected by boat electrofishing during non-native removal efforts in the Yampa and Green rivers in 2005 through 2022 . We used catch data from the Lily Park, Yampa Canyon, and Middle Green River reaches.  Each spring and summer, multiple sampling trips were conducted in each reach. We defined a trip as a full sampling occasion comprised of multiple electrofishing passes from the top of the reach to the bottom. Sampling trips typically lasted 2 – 4 days. On most trips, counts of removed fish were summarized by three size classes (<100 mm, 100-200 mm, and >200 mm). Catch-effort data for each study reach were extracted from annual reports which have sampling details (Lily Park: Project # 125/98c, Yampa Canyon: Project # 110, Middle Green: Project # 123a) and are publicly available on the Upper Colorado River Endangered Fish Recovery Program website (https://coloradoriverrecovery.org/uc/documents/work-plan-documents/annual-reports/).

In addition to catch data for each removal reach, we also used observations of recaptured tagged smallmouth bass to estimate movement rates among our reaches of interest. Smallmouth bass were sometimes tagged with individual Floy tags during early removal trips to estimate abundance each year. The Recovery Program compiled encounter locations for marked and recaptured Floy-tagged fish across the reaches of interested using the Species, Tagging, Research and Monitoring System database (https://streamsystem.org/index.php).

We calculated hydrologic metrics based on conditions in the Green River at the U.S. Geological Survey (hereafter USGS) gage at Jensen (#09261000; USGS 2023a) and in the Yampa River from the USGS gage at Deerlodge Park (#09260050; USGS 2023b; Figure 1). Temperature metrics for the Green River were also obtained from the Jensen gage, while temperature data from the Yampa River were collected at Echo Park by USFWS (Figure 1). The temperature logger at Echo Park was periodically out of service, in those cases we used daily mean temperature from Deerlodge Park to fill in any data gaps. Mean daily discharge and temperature for each river was summarized as metrics used to predict smallmouth bass hatch dates based on equations from Bestgen and Hill (2016) or as other predictors included in the model. Calculation of each flow metric is described below in description of the model.

The data included here as been simplified and formatted to fit the smallmouth bass population dynamics model presented in the associated paper with this data and code release.