Predicting habitat suitability for American woodcock and landscape-level assessment of habitat in West Virginia [electronic resource] /

by Steketee, Ann Klein.

Predicting Habitat Suitability For American Woodcock and Landscapelevel Assessment of Habitat in West Virginia Ann K. Steketee The objectives of this study were to 1) develop models to predict potential habitat suitability for woodcock over large geographic areas in West Virginia (Chapter 4), 2) quantify and describe the local and landscape structural components and spatial patterns of woodcock habitat in the state (Chapter 3), and 3) to determine the current suitability and fate of woodcock habitat that was available in the mid-1970’s (Chapter 5). Differences between 165 woodcock flush points and 165 randomly generated points for proportion of 12 land-cover types and 4 metrics describing landscape composition and pattern were compared within 3 elevation groups and 8 spatial scales. Habitat variables were determined from the Multi-Resolution Land Cover database and FRAGSTATS spatial analysis program. Logistic regression was used to quantify relationships between woodcock use sites and available habitat. Standard reclassification statistics were used to evaluate modeling efficiency and model results were extrapolated to the entire study area. Developed models were generally better at classifying use points than nonuse points. Predicted habitat suitability for the study area ranged from P = 0.0 to 0.96. The two best indicators of woodcock habitat suitability were distance to the nearest wooded wetland and degree of slope. Six habitat variables differed by scale for at least one group of flush points, no variable around random points varied by scale. No habitat variable was important to woodcock at all elevations and most differences occurred in low- and mid-elevations. Using a combination of singing male counts, dog-assisted searches and other means, 5,115 ha of habitat that had been field-checked in the 1970’s were resurveyed. Twenty-five (408 ha) resurveyed sites were classed as definitely not woodcock habitat and 28 sites (173 ha) were classed as unlikely woodcock habitat. Forty-two sites (1040 ha) were classed as possible woodcock habitat; 16 sites (451 ha) were probably woodcock habitat and woodcock were found on 15 sites (3,042 ha). Original classification was a good predictor of status in the 1990's. The majority (83.0 % of sites, 65.6 % of area) of the sites currently classified as unsuitable (definitely not or unlikely) were originally rated as poor or fair habitat. The majority (48.3 % of sites, 93.0 % of area) of the sites currently rated as definite woodcock habitat were originally rated good or exceptional habitat. Reasons for habitat loss included suburban and industrial development, flooding due to dams, conversion to agricultural use (e.g. open fields, pastures, or row crops) and seral advancement of forest vegetation beyond what was suitable for woodcock.