The use of streamflow simulations from the Vflo model and subsequent calculation of streamflow metrics to investigate flow-ecology relationships may be hindered by our inability to accurately model flow variability and extreme flows of the arid Great Plains. The Canadian River and other rivers in the Great Plains tend to have highly variable flows and harsh environmental conditions. The combination of these environmental conditions makes semi-arid and arid regions difficult to represent with a hydrologic model, especially extreme events. In some cases, overestimating flows may be acceptable to water managers (e.g., vulnerability of infrastructures), but could greatly affect estimates of fish species persistence. To address incidences where poor model performance affected metrics derived from Vflo simulations, we suggest three possible options. 1) Restrict flow-ecology relationships to the mainstem of the Canadian River below Lake Meredith, 2) Restrict assessments to streamflow data aggregated at a monthly time step (although typically, this does not match ecological processes well); 3) Focus on streamflow metrics with a high prediction accuracy (e.g., magnitude, timing and duration at some locations). To maximize the number of potential explanatory variables and survey locations available in the Canadian River basin for the development of flow-ecology response models and minimize bias and uncertainty, a combination of these approaches is likely warranted. To move forward on flow-ecology relationships with valid statistical power, the compiled fish data (see processing steps) is best combined with available gage data to improve the development of ecological relationships.

The use of streamflow simulations from the Vflo model and subsequent calculation of streamflow metrics to investigate flow-ecology relationships may be hindered by our inability to accurately model flow variability and extreme flows of the arid Great Plains. The Canadian River and other rivers in the Great Plains tend to have highly variable flows and harsh environmental conditions. The combination of these environmental conditions makes semi-arid and arid regions difficult to represent with a hydrologic model, especially extreme events. In some cases, overestimating flows may be acceptable to water managers (e.g., vulnerability of infrastructures), but could greatly affect estimates of fish species persistence. To address incidences where poor model performance affected metrics derived from Vflo simulations, we suggest three possible options. 1) Restrict flow-ecology relationships to the mainstem of the Canadian River below Lake Meredith, 2) Restrict assessments to streamflow data aggregated at a monthly time step (although typically, this does not match ecological processes well); 3) Focus on streamflow metrics with a high prediction accuracy (e.g., magnitude, timing and duration at some locations). To maximize the number of potential explanatory variables and survey locations available in the Canadian River basin for the development of flow-ecology response models and minimize bias and uncertainty, a combination of these approaches is likely warranted. To move forward on flow-ecology relationships with valid statistical power, the compiled fish data (see processing steps) is best combined with available gage data to improve the development of ecological relationships.

the datasets include: Land surface model (VIC) simulation of historical and future streamflow data. This dataset is associated with the following publication: Reeder, W., F. Gariglio, R.S. Carnie, C. Tang, D. Isaak, C. Qiuwen , Y. Zhongbo, J.A. McKean, and D. Tonina. Some (Fish Might) Like It Hot: Climate and Habitat Quality Variability from Past to Future Climates. SCIENCE OF THE TOTAL ENVIRONMENT. Elsevier BV, AMSTERDAM, NETHERLANDS, 787: 147532, (2021).

Geographic patterns and time trends of water-quality, modeled streamflow, and ecological data were compared along the Canadian River and selected tributaries in northeastern New Mexico to Lake Eufaula in Oklahoma to determine effects of climate change on water quality, streamflows, fish populations and ecological flows in this watershed from 1939 to 2013. Project participants included staff from the Oklahoma Cooperative Fish and Wildlife Research Unit, Vieux and Associates, USGS New Jersey Water Science Center and the USGS Oklahoma Water Science Center. Principal project funding was by the South Central Climate Science Center, with in-kind matching from the project participant organizations.

Geographic patterns and time trends of water-quality, modeled streamflow, and ecological data were compared along the Canadian River and selected tributaries in northeastern New Mexico to Lake Eufaula in Oklahoma to determine effects of climate change on water quality, streamflows, fish populations and ecological flows in this watershed from 1939 to 2013. Project participants included staff from the Oklahoma Cooperative Fish and Wildlife Research Unit, Vieux and Associates, USGS New Jersey Water Science Center and the USGS Oklahoma Water Science Center. Principal project funding was by the South Central Climate Science Center, with in-kind matching from the project participant organizations.

The dataset contains long-term and short-term summaries of streamflow alteration and measures of biological condition (fish multi-metric index). Streamflow alteration metrics include the magnitude, duration, frequency, and seasonality of high and low flow streamflow. Biological condition was estimated from the National Rivers and Streams Assessment and National Water Quality Assessment fish sampling programs. Using fish samples, a fish multi-metric index was calculated and categorized into altered versus non-altered fish communities.

The dataset contains long-term and short-term summaries of streamflow alteration and measures of biological condition (fish multi-metric index). Streamflow alteration metrics include the magnitude, duration, frequency, and seasonality of high and low flow streamflow. Biological condition was estimated from the National Rivers and Streams Assessment and National Water Quality Assessment fish sampling programs. Using fish samples, a fish multi-metric index was calculated and categorized into altered versus non-altered fish communities.

The development of a hydrologic foundation, essential for advancing our understanding of flow-ecology relationships, was accomplished using the high-resolution physics-based distributed rainfall-runoff model Vflo. We compared the accuracy and bias associated with flow metrics that were generated using Vflo at both a daily and monthly time step in the Canadian River basin, USA. First, we calibrated and applied bias correction to the Vflo model to simulate streamflow at ungaged catchment locations. Next, flow metrics were calculated using both simulated and observed data from stream gage locations. We found discharge predictions using Vflo were more accurate than using drainage area ratios. General correspondence between predicted discharge and the gage data was apparent; however, flow metrics calculated using the Vflo output did not accurately represent flow variability. This work was part of a multidisciplinary project describing water quality, streamflow and runoff, and ecology of the Canadian River Basin from northeastern New Mexico to Lake Eufaula, Oklahoma. This study was done in cooperation with the South Central Climate Adaptation Science Center.

The development of a hydrologic foundation, essential for advancing our understanding of flow-ecology relationships, was accomplished using the high-resolution physics-based distributed rainfall-runoff model Vflo. We compared the accuracy and bias associated with flow metrics that were generated using Vflo at both a daily and monthly time step in the Canadian River basin, USA. First, we calibrated and applied bias correction to the Vflo model to simulate streamflow at ungaged catchment locations. Next, flow metrics were calculated using both simulated and observed data from stream gage locations. We found discharge predictions using Vflo were more accurate than using drainage area ratios. General correspondence between predicted discharge and the gage data was apparent; however, flow metrics calculated using the Vflo output did not accurately represent flow variability. This work was part of a multidisciplinary project describing water quality, streamflow and runoff, and ecology of the Canadian River Basin from northeastern New Mexico to Lake Eufaula, Oklahoma. This study was done in cooperation with the South Central Climate Adaptation Science Center.

The development of a hydrologic foundation, essential for advancing our understanding of flow-ecology relationships, was accomplished using the high-resolution physics-based distributed rainfall-runoff model Vflo. We compared the accuracy and bias associated with flow metrics that were generated using Vflo at both a daily and monthly time step in the Canadian River basin, USA. First, we calibrated and applied bias correction to the Vflo model to simulate streamflow at ungaged catchment locations. Next, flow metrics were calculated using both simulated and observed data from stream gage locations. We found discharge predictions using Vflo were more accurate than using drainage area ratios. General correspondence between predicted discharge and the gage data was apparent; however, flow metrics calculated using the Vflo output did not accurately represent flow variability. This work was part of a multidisciplinary project describing water quality, streamflow and runoff, and ecology of the Canadian River Basin from northeastern New Mexico to Lake Eufaula, Oklahoma. This study was done in cooperation with the South Central Climate Adaptation Science Center.