Official websites use .gov
A .gov website belongs to an official government organization in the United States.
Secure .gov websites use HTTPS
A lock () or https:// means you’ve safely connected to the .gov website. Share sensitive information only on official, secure websites.
Issue: The Chesapeake Bay Program partners are striving to improve habitat conditions for recreational fisheries and other native fishes in the Bay and its watershed. While national fish habitat assessments have been conducted, resource managers need more local information to focus restoration and protection efforts in Chesapeake Bay watershed.
Conducting the fish-habitat assessments are challenging due to the size of the Bay and its watershed, which precludes direct surveys of all waters in an efficient and cost-effective manner. Predictive models can fill this gap by providing estimates of condition for these unsurveyed locations. Therefore, the USGS and NOAA are collaborating to improve fish-habitat assessments for inland waters and the estuary, using available data and innovative analytical methods with plans for a joint assessment in a pilot area.
The USGS examined and tested methods needed for a fish-habitat assessment in the nontidal portion of the Chesapeake Bay watershed. The investigators evaluated a large amount of data as part of the study including (1) fish community and species data, and (2) landscape variables that affect fish habitat conditions in nontidal streams and rivers. The investigators determined there was an insufficient number of comparable reference sites to develop a traditional measure of biological conditions, so they developed a novel method that combined community and species-level indicators.
The steps taken to achieve this assessment (Figure 1) included:
1. Watershed-wide fish and landscape predictor data set generation:
Investigators worked with 20 programs and data providers to gather fish community and species-specific sampling data. A total of 31,660 fish sampling events from 01 January 1969 through 13 December 2019 were collated.
Investigators compiled 56 landscape predictors from the USEPA StreamCat dataset and USGS National Water Quality Assessment program that have been previously summarized to the NHDPlusV2.1 and that have been found to strongly affect or potentially affect stream health and fish habitat.
2. Region-specific fish community measure identification:
Measures of the fish community were calculated including metrics describing composition, tolerances, habitat preferences, and functional traits.
The watershed was divided into ecoregions to better reflect gradient of ecological conditions across the watershed (Figure 2). Fish community metrics were selected based on their utility to indicate overall condition within each region. The regions for the Chesapeake watershed included the Coastal Plains (CPL), Northern Appalachians (NAP), Southern Appalachians Northwest (SAPNW), and Southern Appalachians Piedmont (SAPPIED).
3. Key species of interest identification:
The species analyses focused on those native to the watershed and considered sensitive (e.g., Brook Trout and northern hog sucker).
Managers also wanted assessment for important game species, so species like Smallmouth Bass were included.
4. Model development:
Community measures –Models were built to relate community measures with landscape predictors within each region.
Species measures – Separate models were built for each species to relate landscape variables to species occurrence.
5. Model prediction:
The investigators used the optimized region-specific models to predict fish community measures in 66,867 stream and river reaches across the watershed. Community measures were combined into a single index that reflected the degree of community alteration calculated by taking an average of the decile scores from each community measure, with higher deciles indicating less altered habitat conditions.
Species-specific models were used to predict species occurrence within the native range of each key species identified.
Predictions were made for the years 2001, 2006, 2011, and 2016 based on changes in landcover.
Comparisons were made between the community and selected species assessments of condition.
Findings are provided for both fish community and species-specific results.
The predicted community conditions are shown in figure 3. Higher decile values (green shades) represent places with less altered habitat conditions, while lower deciles (brown shades) are areas with more altered habitat conditions. Decile scores have been averaged by ecoregion so direct comparisons across regions are not possible. Based on figure 3, the summary of the spatial conditions for each ecoregion were:
Results are presented in figure 4 for brook trout (BRT), Northern hog sucker (NHS), Smallmouth bass (SMB), and Torrent Sucker (TRS). The areas where species are predicted to be present (light blue shades) represent places with landscape conditions that could support the species, while areas predicted to be absent (dark blue shades) represent places that may not support the populations of the species.
Successful management of freshwater systems hinges on the ability to effectively identify where to best invest resources toward conservation and restoration. Here the authors developed a flexible approach for non-tidal streams and rivers that considers both a novel composite measure of community biological condition and a species level occurrence analysis with similar data. This approach provides results that can give multiple views of condition and address multiple management needs. When examined at a finer scale (e.g., Fig. 5) results can be used to identify potential problem areas that require attention or areas that are experiencing biological improvement (i.e., lift) possibly via management actions.
Results also allow managers to compare and identify areas where both community and population level measures align (or differ) in predicting reach condition, which could aid in management efforts. For example, a site with a predicted Brook Trout presence but predicted marginal community index habitat score may indicate an area to target restoration to protect this key species, as the community assessment suggests potential habitat degradation which was missed by the species level assessment.
The results of the study have been published in Ecological Indicators (with open access), titled: “Using fish community and population indicators to assess the biological condition of streams and rivers of the Chesapeake Bay watershed, USA.” (https://doi.org/10.1016/j.ecolind.2021.108488)
The project team included Kelly Maloney, Kevin Krause, Matthew Cashman, Wesley Daniel, Benjamin Gressler, Daniel Wieferich, and John Young of the USGS. Kelly Maloney (kmaloney@usgs.gov) and Kevin Krause (kkrause@usgs.gov) are the primary contacts for this project.
Community metrics from inter-agency compilation of inland fish sampling data within the Chesapeake Bay Watershed. https://doi.org/10.5066/P9D6JU4X
Fish community and species distribution predictions for streams and rivers of the Chesapeake Bay Watershed. https://doi.org/10.5066/P9B4BMAG
Map showing change in aggregated ecoregion-specific selected mean metric deciles from 2001 to 2016 (negative indicates more altered) for nontidal stream reaches in the Chesapeake Bay watershed. Insets show two focus areas to better highlight changes.
Map showing change in aggregated ecoregion-specific selected mean metric deciles from 2001 to 2016 (negative indicates more altered) for nontidal stream reaches in the Chesapeake Bay watershed. Insets show two focus areas to better highlight changes.
Maps showing predicted occurrence for the four species of interest in 2016. BRT = Brook Trout, NHS = Northern Hog Sucker, SMB = Smallmouth Bass, TRS = Torrent Sucker. NA infers reach was outside modeling extent for that species.
Maps showing predicted occurrence for the four species of interest in 2016. BRT = Brook Trout, NHS = Northern Hog Sucker, SMB = Smallmouth Bass, TRS = Torrent Sucker. NA infers reach was outside modeling extent for that species.
Map of the Chesapeake Bay watershed with location of community samples identified by habitat size and aggregated ecoregions. Inset show study area relative to the northeastern USA. Aggregated ecoregions include Coastal Plains (CPL), Northern Appalachians (NAP), Southern Appalachians Northwest (SAPNW), and Southern Appalachians Piedmont (SAPPIED).
Map of the Chesapeake Bay watershed with location of community samples identified by habitat size and aggregated ecoregions. Inset show study area relative to the northeastern USA. Aggregated ecoregions include Coastal Plains (CPL), Northern Appalachians (NAP), Southern Appalachians Northwest (SAPNW), and Southern Appalachians Piedmont (SAPPIED).
Community metric index score of stream condition (mean metric decile) for 2001, 2006, 2011, and 2016. Higher values (teal shades) indicate less altered conditions while lower values indicate a higher degree of alteration (brown shades).
Community metric index score of stream condition (mean metric decile) for 2001, 2006, 2011, and 2016. Higher values (teal shades) indicate less altered conditions while lower values indicate a higher degree of alteration (brown shades).
Conceptual diagram illustrating the multi-step approach to assess fish habitat in non-tidal reaches of the Chesapeake Bay watershed.
Conceptual diagram illustrating the multi-step approach to assess fish habitat in non-tidal reaches of the Chesapeake Bay watershed.
Issue: The Chesapeake Bay Program partners are striving to improve habitat conditions for recreational fisheries and other native fishes in the Bay and its watershed. While national fish habitat assessments have been conducted, resource managers need more local information to focus restoration and protection efforts in Chesapeake Bay watershed.
Conducting the fish-habitat assessments are challenging due to the size of the Bay and its watershed, which precludes direct surveys of all waters in an efficient and cost-effective manner. Predictive models can fill this gap by providing estimates of condition for these unsurveyed locations. Therefore, the USGS and NOAA are collaborating to improve fish-habitat assessments for inland waters and the estuary, using available data and innovative analytical methods with plans for a joint assessment in a pilot area.
The USGS examined and tested methods needed for a fish-habitat assessment in the nontidal portion of the Chesapeake Bay watershed. The investigators evaluated a large amount of data as part of the study including (1) fish community and species data, and (2) landscape variables that affect fish habitat conditions in nontidal streams and rivers. The investigators determined there was an insufficient number of comparable reference sites to develop a traditional measure of biological conditions, so they developed a novel method that combined community and species-level indicators.
The steps taken to achieve this assessment (Figure 1) included:
1. Watershed-wide fish and landscape predictor data set generation:
Investigators worked with 20 programs and data providers to gather fish community and species-specific sampling data. A total of 31,660 fish sampling events from 01 January 1969 through 13 December 2019 were collated.
Investigators compiled 56 landscape predictors from the USEPA StreamCat dataset and USGS National Water Quality Assessment program that have been previously summarized to the NHDPlusV2.1 and that have been found to strongly affect or potentially affect stream health and fish habitat.
2. Region-specific fish community measure identification:
Measures of the fish community were calculated including metrics describing composition, tolerances, habitat preferences, and functional traits.
The watershed was divided into ecoregions to better reflect gradient of ecological conditions across the watershed (Figure 2). Fish community metrics were selected based on their utility to indicate overall condition within each region. The regions for the Chesapeake watershed included the Coastal Plains (CPL), Northern Appalachians (NAP), Southern Appalachians Northwest (SAPNW), and Southern Appalachians Piedmont (SAPPIED).
3. Key species of interest identification:
The species analyses focused on those native to the watershed and considered sensitive (e.g., Brook Trout and northern hog sucker).
Managers also wanted assessment for important game species, so species like Smallmouth Bass were included.
4. Model development:
Community measures –Models were built to relate community measures with landscape predictors within each region.
Species measures – Separate models were built for each species to relate landscape variables to species occurrence.
5. Model prediction:
The investigators used the optimized region-specific models to predict fish community measures in 66,867 stream and river reaches across the watershed. Community measures were combined into a single index that reflected the degree of community alteration calculated by taking an average of the decile scores from each community measure, with higher deciles indicating less altered habitat conditions.
Species-specific models were used to predict species occurrence within the native range of each key species identified.
Predictions were made for the years 2001, 2006, 2011, and 2016 based on changes in landcover.
Comparisons were made between the community and selected species assessments of condition.
Findings are provided for both fish community and species-specific results.
The predicted community conditions are shown in figure 3. Higher decile values (green shades) represent places with less altered habitat conditions, while lower deciles (brown shades) are areas with more altered habitat conditions. Decile scores have been averaged by ecoregion so direct comparisons across regions are not possible. Based on figure 3, the summary of the spatial conditions for each ecoregion were:
Results are presented in figure 4 for brook trout (BRT), Northern hog sucker (NHS), Smallmouth bass (SMB), and Torrent Sucker (TRS). The areas where species are predicted to be present (light blue shades) represent places with landscape conditions that could support the species, while areas predicted to be absent (dark blue shades) represent places that may not support the populations of the species.
Successful management of freshwater systems hinges on the ability to effectively identify where to best invest resources toward conservation and restoration. Here the authors developed a flexible approach for non-tidal streams and rivers that considers both a novel composite measure of community biological condition and a species level occurrence analysis with similar data. This approach provides results that can give multiple views of condition and address multiple management needs. When examined at a finer scale (e.g., Fig. 5) results can be used to identify potential problem areas that require attention or areas that are experiencing biological improvement (i.e., lift) possibly via management actions.
Results also allow managers to compare and identify areas where both community and population level measures align (or differ) in predicting reach condition, which could aid in management efforts. For example, a site with a predicted Brook Trout presence but predicted marginal community index habitat score may indicate an area to target restoration to protect this key species, as the community assessment suggests potential habitat degradation which was missed by the species level assessment.
The results of the study have been published in Ecological Indicators (with open access), titled: “Using fish community and population indicators to assess the biological condition of streams and rivers of the Chesapeake Bay watershed, USA.” (https://doi.org/10.1016/j.ecolind.2021.108488)
The project team included Kelly Maloney, Kevin Krause, Matthew Cashman, Wesley Daniel, Benjamin Gressler, Daniel Wieferich, and John Young of the USGS. Kelly Maloney (kmaloney@usgs.gov) and Kevin Krause (kkrause@usgs.gov) are the primary contacts for this project.
Community metrics from inter-agency compilation of inland fish sampling data within the Chesapeake Bay Watershed. https://doi.org/10.5066/P9D6JU4X
Fish community and species distribution predictions for streams and rivers of the Chesapeake Bay Watershed. https://doi.org/10.5066/P9B4BMAG
Map showing change in aggregated ecoregion-specific selected mean metric deciles from 2001 to 2016 (negative indicates more altered) for nontidal stream reaches in the Chesapeake Bay watershed. Insets show two focus areas to better highlight changes.
Map showing change in aggregated ecoregion-specific selected mean metric deciles from 2001 to 2016 (negative indicates more altered) for nontidal stream reaches in the Chesapeake Bay watershed. Insets show two focus areas to better highlight changes.
Maps showing predicted occurrence for the four species of interest in 2016. BRT = Brook Trout, NHS = Northern Hog Sucker, SMB = Smallmouth Bass, TRS = Torrent Sucker. NA infers reach was outside modeling extent for that species.
Maps showing predicted occurrence for the four species of interest in 2016. BRT = Brook Trout, NHS = Northern Hog Sucker, SMB = Smallmouth Bass, TRS = Torrent Sucker. NA infers reach was outside modeling extent for that species.
Map of the Chesapeake Bay watershed with location of community samples identified by habitat size and aggregated ecoregions. Inset show study area relative to the northeastern USA. Aggregated ecoregions include Coastal Plains (CPL), Northern Appalachians (NAP), Southern Appalachians Northwest (SAPNW), and Southern Appalachians Piedmont (SAPPIED).
Map of the Chesapeake Bay watershed with location of community samples identified by habitat size and aggregated ecoregions. Inset show study area relative to the northeastern USA. Aggregated ecoregions include Coastal Plains (CPL), Northern Appalachians (NAP), Southern Appalachians Northwest (SAPNW), and Southern Appalachians Piedmont (SAPPIED).
Community metric index score of stream condition (mean metric decile) for 2001, 2006, 2011, and 2016. Higher values (teal shades) indicate less altered conditions while lower values indicate a higher degree of alteration (brown shades).
Community metric index score of stream condition (mean metric decile) for 2001, 2006, 2011, and 2016. Higher values (teal shades) indicate less altered conditions while lower values indicate a higher degree of alteration (brown shades).
Conceptual diagram illustrating the multi-step approach to assess fish habitat in non-tidal reaches of the Chesapeake Bay watershed.
Conceptual diagram illustrating the multi-step approach to assess fish habitat in non-tidal reaches of the Chesapeake Bay watershed.