Mapping of the current distributions of forest-cover types across the Great Dismal Swamp National Wildlife Refuge (the swamp) is critical to understanding the success of ongoing hydrologic and other management techniques used to restore the forest communities of the swamp to those present across the swamp in early colonial times. Aerial photographs, orthophotographs, and vector digital data were used to map forest-cover types of the swamp. The forest-cover types were interpreted and mapped using composition, height, and canopy-closure classes derived from this imagery and field verification. The imagery was obtained using a near-infrared sensor (NIR)carried in an airplane flown across the swamp during the mid-to-late spring of 2015. The original dataset was provided in Universal Transverse Mercator meters, Zone 18, NAD 83 projection. The resolution of the imagery was 0.3-meter pixels. This report explains the metadata for the vector digital geodatabase for forest cover as interpreted from the imagery and field verification. One or two species composition codes represent the major forest types in the canopy of each stand. The first code represents the forest type forming at least 50 percent of the canopy. The second code represents the forest type forming 25 to less than 50 percent of the canopy. A single code indicates that only that forest type forms at least 25 percent of the canopy. Where the forest covered less than 25 percent of an area, the area was classified as emergent species. Minimum mapping areas were 5 acres for forested land and 1 acre for emergent species.

Forest surveys were conducted in nine 20 m x 25 m study plots, split into 3 representatives each for three forest types in Great Dismal Swamp, VA and NC, USA, December 2015 - February 2018. Trees, saplings, and shrubs were identified to species and measured for estimates of standing stocks. Standing stock data include: tree diameter at breast height (dbh), height, and condition; sapling dbh; shrub diameter at root collar, and height. In each plot, roughly 10 co-dominant trees were equipped with dendrometer bands and measured annually for growth estimates.

Forest surveys were conducted in nine 20 m x 25 m study plots, split into 3 representatives each for three forest types in Great Dismal Swamp, VA and NC, USA, December 2015 - February 2018. Trees, saplings, and shrubs were identified to species and measured for estimates of standing stocks. Standing stock data include: tree diameter at breast height (dbh), height, and condition; sapling dbh; shrub diameter at root collar, and height. In each plot, roughly 10 co-dominant trees were equipped with dendrometer bands and measured annually for growth estimates.

Plot-level field data were collected in the summer of 2014 to estimate aboveground and belowground biomass in the Great Dismal Swamp National Wildlife Refuge and Dismal Swamp State Park in North Carolina and Virginia. Data were collected at 85 plots. The location of the center of each plot was recorded with a Trimble ProXH global positioning system (GPS) and differentially corrected. Data files included 1: GDS_plots.csv, 2. GDS_FWD.csv, 3. GDS_LWD.csv, 4. GDS_Shrubs.csv, 5. GDS_Trees.csv, and 6. GDS_plot_summaries.csv. The data contained in GDS_plot_summaries.csv were calculated from the GDS_plots.csv, GDS_FWD.csv, GDS_LWD.csv, GDS_Shrubs.csv, GDS_Trees.csv files using the R statistical software environment (R Core Team, 2019) and code in GDS_AGB_Summaries.R. R Core Team, 2019, R: A Language and Environment for Statistical Computing, R Foundation for Statistical Computing, Vienna, Austria, https://www.R-project.org

Plot-level field data were collected in the summer of 2014 to estimate aboveground and belowground biomass in the Great Dismal Swamp National Wildlife Refuge and Dismal Swamp State Park in North Carolina and Virginia. Data were collected at 85 plots. The location of the center of each plot was recorded with a Trimble ProXH global positioning system (GPS) and differentially corrected. Data files included 1: GDS_plots.csv, 2. GDS_FWD.csv, 3. GDS_LWD.csv, 4. GDS_Shrubs.csv, 5. GDS_Trees.csv, and 6. GDS_plot_summaries.csv. The data contained in GDS_plot_summaries.csv were calculated from the GDS_plots.csv, GDS_FWD.csv, GDS_LWD.csv, GDS_Shrubs.csv, GDS_Trees.csv files using the R statistical software environment (R Core Team, 2019) and code in GDS_AGB_Summaries.R. R Core Team, 2019, R: A Language and Environment for Statistical Computing, R Foundation for Statistical Computing, Vienna, Austria, https://www.R-project.org

The Great Dismal Swamp (the swamp) is a forested peatland in southeastern Virginia and northeastern North Carolina. Since early colonial times, timber harvesting and drainage through a network of ditches constructed to facilitate the harvesting have altered these ecosystems. The U.S. Fish and Wildlife Service has managed the swamp as the Great Dismal Swamp National Wildlife Refuge since 1974 to restore its forest communities to those present in early colonial times. Part of the approach to forest restoration has been to "restore the hydrology." The report by Speiran and Wurster (2020) describes the hydrology and water quality across the swamp. Part of the data used to describe the hydrology and water quality of the Great Dismal Swamp are not available through other publicly accessible databases. These data are derived from three sources: (1) water-quality data collected at 90 sites throughout the swamp by the U.S. Geological Survey and the U.S. Fish and Wildlife Service as part of a one-time synoptic survey on March 15-31, 2016, (2) water-level, flow, and forest-cover data collected across the swamp by the U. S. Fish and Wildlife Service since 2009, and (3) data collected by the U.S. Army Corps of Engineers since 1940 on levels of Lake Drummond and flow from the Feeder Ditch and the Dismal Swamp Canal at Deep Creek, Va., and South Mills, N.C. The water-quality data were used to help identify and verify sources of water to the swamp. Forest-cover data provide a reference for comparison with forest cover in the early 1970s and the future. The remaining data are hydrologic information within and around the swamp. Reference: Speiran, G.K., and Wurster, F.C., 2020, Hydrology and water quality of the Great Dismal Swamp National Wildlife Refuge, Virginia and North Carolina, and implications for hydrologic-management goals and strategies: U.S. Geological Survey Scientific Investigations Report, 2020-XXXX, xx p.

Forest Inventory Phase II GIS data for the Upper Mississippi River National Wildlife and Fish Refuge-Winona District. Data is based off U.S. Army of Corps Engineers' MVP-ES Forest Inventory Phase II Protocol. Data covers 2016-2020 field seasons. All data have been further cleaned and edited. This is the most up-to-date version. See attached protocol for data field definitions.

Forest Inventory Phase II GIS data for the Upper Mississippi River National Wildlife and Fish Refuge-Winona District. Data is based off U.S. Army of Corps Engineers' MVP-ES Forest Inventory Phase II Protocol. Data covers 2016-2020 field seasons. All data have been further cleaned and edited. This is the most up-to-date version. See attached protocol for data field definitions.

Forest Inventory Phase II GIS data for the Upper Mississippi River National Wildlife and Fish Refuge-Winona District. Data is based off U.S. Army of Corps Engineers' MVP-ES Forest Inventory Phase II Protocol. Data covers 2025 field season. All data have been further cleaned and edited. See attached protocol for data field definitions.

Values for area of sustainable forest habitat for each species were obtained as the predicted occupied proportion of each 900 square meter pixel (i.e., occupancy probability x 900) within all forest patches deemed large enough to harbor a sustainable population of the species. The area required for a sustainable population of each species was derived from credible intervals associated with population trends from historical (1966-2015) BBS data (Sauer and others, 2017). For each silvicolous bird species in the Mississippi Alluvial Valley, we assumed the minimum sustainable population was the number of birds needed to ensure ≤1% probability that the population would be extirpated (i.e., drop below a quasi-extinction threshold) during a 100-year period wherein annual population change was randomly selected from the credible interval associated with each species’ population trend. We used the mean of 500 simulation replicates conducted in R (Version 3.4.4; https://www.r-project.org/) as the presumed minimum sustainable population for each species. We arbitrarily set the quasi-extinction threshold at 25 breeding pairs. Because species with credible intervals associated with their trend estimates that were inclusively positive never declined in population, by default these species had a minimum sustainable population of 25 pairs.