These methods, instrumentation and/or protocols apply to all data in this dataset:Methods and protocols used in the collection of this data package |
---|
Description: |
STAND-LEVEL DATA ACQUISITION
REMOTE SENSING
Google Earth
Variables: Distance to edge, distance to road, distance to water
We used the Quantum Geographic Information System (QGIS) program distance measuring tool to measure distance to edge, water, and road from goshawk nest locations and paired random points (QGIS Development Team, 2020). To measure distance to edge, we visually inspected satellite imagery from the Google Earth Satellite Hybrid layer (Google, 2020) and measured the distance to the closest open habitat from the forested nest or random point. Because flowing water tended to be difficult to detect visually using satellite imagery, while measuring distance to water we overlaid the National Hydrography Dataset (NHD) flowline vector layer (Environmental Protection Agency, 2014) on the Google Earth Satellite Hybrid layer along with goshawk nest locations and random points. We then measured the distance between the nest or random point and the nearest flowing water according to the NHD Flowline vector layer. Many well-defined illegal roads exist in our study area and are regularly used, but these are not captured in the 2020 Motor Vehicle Use Map (MVUM) we used for our distance to road measurements. Therefore, we measured to these illegal roads rather than the MVUM layer if they were closer to the nest or random point. Across our “distance to” measurements, we placed a 1370 m cap on our values, which reflects the male goshawk home range size quantified by Hasselblad and Bechard (2007) in our study area. At the stand level, “distance to” values exceeding 1370 m were replaced with the capped distance of 1370 m.
Digital Elevation Model
Variables: Elevation, slope, TPI, northness, eastness
We acquired Digital Elevation Model (DEM) data at a resolution of 30 m across our study area, from which we calculated elevation, aspect, slope, and Topographic Position Index (TPI; U.S. Geological Survey, 2013; Wilson et al., 2007; Fleming and Hoffer, 1979). We used cosine transformations to convert aspect into northness and eastness variables (Roberts, 1986). We centered and scaled each of our variables by subtracting their mean then dividing by their standard deviation.
COLLECTED BY HAND
Canopy App
Variables: Canopy closure
We measured canopy closure at each of the 5 points using the CanopyApp phone app (University of New Hampshire, 2014), which uses supervised image classification to create a filter over a photo of the canopy to determine percent canopy closure.
Rangefinder (Nikon 6x20 Prostaff 1000 or equivalent) and/or tape measurer
Variables: SDI, trees per hectare, DBH, tree height, crown depth
To characterize the habitat at each nest site and random paired site, we collected a consistent set of data at the center point (nest tree or center of paired site) and at 4 points 20 m from the center in each of the cardinal directions. We recorded measurements that enabled us to calculate tree density, mean tree height, and mean crown depth (Reynolds et al., 1982). We measured tree height and crown depth of the center tree of each site and of the closest tree with a DBH greater than 15 cm at each of the 4 cardinal points (5 measurements per site). We used the Pythagorean Theorem to determine tree height and crown depth based on measurements taken with a laser rangefinder to the bottom of the live crown, tree top, and tree base. These measurements were taken while kneeling at some distance (usually 6-12 m) with a clear view of each of the 3 points of interest. At each of the 5 data collection points per site, we measured DBH of and distance to the closest 7 trees with DBHs greater than 15 cm (center tree and 4 cardinal points; 35 trees total for both DBH and distance measurements per site). Our DBH variable reflects the mean of the 35 trees measured at each site. We calculated Stand Density Index (SDI) and trees per hectare for each point using our distance and DBH measurements (Lilieholm et al., 1994). In a few cases where our measurement points fell within an unusually dense grouping of trees, our calculations for trees per hectare and SDI were flagged as outliers exceeding maximum SDI values identified by Woodall et al. (2005). Because these extremely high SDI values were not representative of the stand and instead of removing these data points, we set a minimum value for the distance to the 7th tree at 4 m, which was the lowest value that eliminated the outliers. We collected the same set of measurements at the nest and random paired sites.
DATASETS
- Google, 2020. Google Earth Pro, version 7.3.3. Alphabet Inc., Mountain View, California, USA.
- USDA Forest Service, 2020. Sawtooth National Forest motor vehicle use map.
- U.S. Environmental Protection Agency, 2014. National Hydrography Dataset Plus Version 2, Extended Unit Runoff Method.
- U.S. Geological Survey, 2013. USGS NED 1 arc-second 2013 1 x 1 degree.
OTHER CITATIONS
- QGIS Development Team, 2020. QGIS Geographic Information System, version 3.16.0.
- Hasselblad, K., Bechard, M., 2007. Male northern goshawk home ranges in the Great Basin of south-central Idaho. J. Raptor Res. 41(2), 150-155. https://doi.org/10.3356/0892-1016(2007)41[150:MNGHRI]2.0.CO;2
- Stekhoven, D.J., Buhlmann, P., 2012. MissForest–non-parametric missing value imputation for mixed-type data. Bioinfomatics 28(1), 112–118. https://doi.org/10.1093/bioinformatics/btr597
- Wilson, M.F.J., O’Connell, B., Brown, C., Guinan, J.C., Grehan, A.J., 2007. Multiscale terrain analysis of multibeam bathymetry data for habitat mapping on the continental slope. Mar. Geod. 30, 3–35. https://doi.org/10.1080/01490410701295962
- Fleming, M.D., Hoffer, R.M., 1979. Machine Processing of Landsat MSS Data and DMA Topographic Data for Forest Cover Type Mapping. LARS Technical Reports. Purdue University, West Lafayette, IN, pp. 377-390. http://docs.lib.purdue.edu/larstech/80
- Roberts, D.W., 1986. Ordination on the basis of fuzzy set theory. Vegetatio 66(3), 123–131. https://www.jstor.org/stable/20037322
- University of New Hampshire, 2014. CanopyApp, version 1.0.3. Durham, New Hampshire, USA.
- Reynolds, R.T., Meslow, E.C., Wight, H.M., 1982. Nesting habitat of coexisting accipiter in Oregon. J. Wildl. Manag. 46(1), 124–138. https://www.jstor.org/stable/3808415
- Lilieholm, R.J., Long, J.N., Patla, S., 1994. Assessment of goshawk nest area habitat using stand density index. Stud. Avian Biol. 16, 18–23.
- Woodall, C.W., Miles, P.D., Vissage, J.S., 2005. Determining maximum stand density index in mixed species stands for strategic-scale stocking assessments. Forest Ecology And Management. 216, 367–377. https://doi.org/10.1016/j.foreco.2005.05.050
|
| Description: |
FOREST-WIDE DATA ACQUISITION
“Distance to” measurements
Variables: Distance to edge, distance to road, distance to water
For our forest-wide analysis, we quantified distance to water, distance to road, and distance to edge across our study area by rasterizing relevant GIS layers and computing the geographic distance to the nearest road, water, or edge for each cell in the 30 m DEM raster (Hijmans et al., 2021). For distance to water, we used the National Hydrography Dataset (NHD) Flowline vector layer from the 2014 NHD Plus Version 2 dataset and measured the distance to the nearest stream with a mean June flow greater than 0.5 cubic feet second (cfs) based on Enhanced Unit Runoff Method (U.S. Environmental Protection Agency, 2014). We chose the minimum value of 0.5 cfs based on our knowledge of the hydrology of the study area. For distance to road, we used the 2020 Motor Vehicle Use Map (MVUM; USDA Forest Service, 2020). We did not consider illegal roads (unmarked in the 2020 MVUM) in our distance to road measurements for the forest-wide analysis, since it was impractical to identify illegal roads by hand across the whole study area. For distance to edge, we used the 2014 GIS layer from the USDA Forest Service Vegetation Classification, Mapping, and Quantitative Inventory (VCMQ; USDA Forest Service, 2015). We placed a 1370 m cap on our values, which reflects the male goshawk home range size quantified by Hasselblad and Bechard (2007) in our study area. For the forest-wide analysis, we replaced values exceeding 1370 m with imputed values between 0 m and 1370 m, but share them here as NAs (Stekhoven and Buhlmann 2012).
Digital Elevation Model
Variables: Elevation, slope, TPI, northness, eastness
We acquired Digital Elevation Model (DEM) data at a resolution of 30 m across our study area, from which we calculated elevation, aspect, slope, and Topographic Position Index (TPI; U.S. Geological Survey, 2013; Wilson et al., 2007; Fleming and Hoffer, 1979). We used cosine transformations to convert aspect into northness and eastness variables (Roberts, 1986). We centered and scaled each of our variables by subtracting their mean then dividing by their standard deviation.
National Land Cover Database
Variables: Canopy closure
We used the 2016 National Land Cover Dataset to obtain mean canopy closure values (U.S. Geological Survey, 2016).
LANDFIRE
Variables: Tree height, crown depth
We acquired mean canopy height (used as mean tree height) and mean canopy base height from the 2016 LANDFIRE (LF) dataset (U.S. Geological Survey, 2019). We calculated mean crown depth by subtracting the LF canopy base height from the LF canopy height.
DATASETS
- U.S. Environmental Protection Agency, 2014. National Hydrography Dataset Plus Version 2, Extended Unit Runoff Method.
- USDA Forest Service, 2020. Sawtooth National Forest motor vehicle use map.
- USDA Forest Service, 2015. Sawtooth National Forest mid-level existing vegetation classification and mapping. https://www.fs.usda.gov/Internet/FSE_DOCUMENTS/fseprd606637.pdf
- U.S. Geological Survey, 2013. USGS NED 1 arc-second 2013 1 x 1 degree.
- U.S. Geological Survey, 2016. National Land Cover Database Tree Canopy Layer.
- U.S. Geological Survey, 2019. LANDFIRE: LANDFIRE 2016 Remap (LF 2.0.0) Existing Vegetation Type layer.
OTHER CITATIONS
- Hijmans, R.J., Bivand, R., Forner, K., Ooms, J., Pebesma, E., 2021. terra: Spatial Data Analysis.
- Hasselblad, K., Bechard, M., 2007. Male northern goshawk home ranges in the Great Basin of south-central Idaho. J. Raptor Res. 41(2), 150-155. https://doi.org/10.3356/0892-1016(2007)41[150:MNGHRI]2.0.CO;2
- Stekhoven, D.J., Buhlmann, P., 2012. MissForest–non-parametric missing value imputation for mixed-type data. Bioinfomatics 28(1), 112–118. https://doi.org/10.1093/bioinformatics/btr59
- Wilson, M.F.J., O’Connell, B., Brown, C., Guinan, J.C., Grehan, A.J., 2007. Multiscale terrain analysis of multibeam bathymetry data for habitat mapping on the continental slope. Mar. Geod. 30, 3–35. https://doi.org/10.1080/01490410701295962
- Fleming, M.D., Hoffer, R.M., 1979. Machine Processing of Landsat MSS Data and DMA Topographic Data for Forest Cover Type Mapping. LARS Technical Reports. Purdue University, West Lafayette, IN, pp. 377-390. http://docs.lib.purdue.edu/larstech/80
- Roberts, D.W., 1986. Ordination on the basis of fuzzy set theory. Vegetatio 66(3), 123–131. https://www.jstor.org/stable/20037322
|
|
|
|