Breeding Woodcock Ireland survey data and analysis
Description
This data contains the data and code required to replicate analyses for a two-step species distribution modelling approach to estimate the distribution and abundance of breeding Eurasian woodcock (Scolopax rusticola) across the Republic of Ireland. Quantitative species occurrence data was collected at 238 randomly-selected forest sites by volunteers who counted displaying (roding) male woodcock at dusk in May and June of 2017, 2019 and 2021. Environmental data used consisted of historical bioclimatic variables downloaded from www.worldclim.org, and landcover (forest and urban settlement) derived from datasets provided by Coillte, DAFM Forestry Division, DAERA Forest Service and CORINE landcover. The first step was an ensemble model using biomod2 in R, to estimate the geographical range of breeding woodcock in Ireland using bioclimatic variables at a 1 * 1 kilometre resolution. The second step was a Generalised Additive Model (GAM) using 'mgcv' in R to determine the bioclimatic and landcover drivers behind woodcock abundance, conditional on presence (falling within the geographic range estimated in the first step). The GAM was then projected across 1 km squares with a minimum threshold of forest habitat (0.1 km2) within the estimated range to predict the distribution and abundance across Ireland. Finally, the predicted density from all relevant 1km squares were summed to produce a national population estimate, with standard errors.
Files
Steps to reproduce
A total of 2,783 potential survey areas were generated randomly across forest cover on the island of Ireland using ArcGIS Pro v.3.1.0. Forest cover was based on land cover data provided by Coillte, the Forestry Division of the Irish Governmental Department of Agriculture, Food and the Marine (DAFM), and the Forest Service of the Northern Ireland Governmental Department of Environment, Agriculture and Rural Affairs (DAERA). Potential survey areas were made publicly available online where volunteers could select them and register their intent to survey, along with detailed information about the study species and clear instructions for the survey methodology. Surveys were conducted in 2017, 2019 and 2021. In each year, the survey period ran from 1st May to 30th June. Each site was surveyed between one and, ideally, three times during the survey period. All surveys were conducted from the same observation point by the same surveyor(s) within each survey area. Surveys commenced fifteen minutes before sunset and ran for seventy-five minutes. During this period, each Woodcock seen or heard was registered. Data were submitted through an online application, by email or by post. Insufficient data were received to include Northern Ireland. Verification and cleaning of survey data was undertaken. Nineteen raster datasets representing historical (near-current) bioclimatic variables (see Appendix in Supporting Information, Table S1) from 1970-2000 at 30 arc seconds were obtained from Fick and Hijmans (2017). These were then resampled to ensure a consistent 1 km2 resolution across the study area for integration into the SDM. Forest land cover data for 2020 in vector format was obtained from Coillte, the Forestry Division of the DAFM, and the 2018 Corine Land Cover Inventory (European Environment Agency, 2019). All these sources were then collated to produce a single forest land cover dataset for the region. Blocks of forest land cover were categorised according to type, similar to standard Irish forestry categories (Keane, Mason and Pfeifer, 2018): ‘coniferous’ (comprising ≥75% coniferous trees), ‘broadleaved’ (comprising ≥75% broadleaved trees), and ‘mixed’ (comprising a mix of coniferous and broadleaved trees where neither type exceeds 75%). Forest land cover was further categorised by majority tree age following the method of Keane, Mason and Pfeifer (2018): ‘prethicket’ (less than 8 years), ‘thicket’ (8 to 20 years), and ‘mature’ (more than 20 years, closed canopy). To provide a measure of potential human disturbance, urban settlement was included as a land cover category. Urban land cover data were obtained from the 2018 Corine Land Cover Inventory (European Environment Agency, 2019). To represent significantly built-up urban settlement areas, the following Corine land cover categories were collated: Continuous urban fabric; Discontinuous urban fabric; Industrial or commercial units.
Institutions
- University College Cork BioSciences Institute
- University College Cork Environmental Research Institute
Categories
Funders
- Irish Research Council
- National Association of Regional Game Councils