Modelled laying dates for Montagu's Harrier (Circus pygargus) across Spain, 2005–2023

This dataset contains modelled laying dates for Montagu’s Harrier (Circus pygargus) across mainland Spain, derived from a nationwide spatio-temporal analysis of breeding phenology, spatial predictors, topography and spring precipitation. Laying dates were predicted using a Generalized Additive Model fitted with mgcv::bam(), including a two-dimensional smooth of spatial coordinates, a smooth term for altitude, April precipitation and a random-effect smooth for year. The dataset includes annual predictions for each 10 × 10 km UTM grid cell considered across mainland Spain for the period 2005–2023. Predictions are expressed as ISO-8601 week-of-year values. The dataset is not restricted to the current breeding distribution of the species; instead, it provides a continuous prediction surface across mainland grid cells that can be subsequently intersected with species distribution, agricultural or risk layers. These outputs form the laying-date component used to quantify temporal overlap between predicted breeding phenology and cereal harvest timing at the national scale. Data include:– Grid cell ID (10×10 km UTM)– Longitude — centroid longitude of the 10×10 km grid cell, in decimal degrees (WGS84, EPSG:4326).– Latitude — centroid latitude of the 10×10 km grid cell, in decimal degrees (WGS84, EPSG:4326).– Year– Predicted laying week (ISO‑8601 week number) Applications:– Conservation planning for ground‑nesting farmland birds– Spatial prioritization of nest protection programs– Risk mapping and agroecological modelling– Reproducibility of the publication’s analyses Methods summary: Laying dates were modelled using a GAM fitted with penalized splines for spatial coordinates and altitude. April precipitation was retained as the only climatic predictor in the final model, whereas May precipitation was excluded because it did not improve model fit. Extreme laying dates were removed using the interquartile range criterion to focus the analysis on the main laying period and reduce the influence of likely replacement clutches. The modelling sequence started from a full model including April and May precipitation, spatial and altitude smooths, and random-effect smooths for locality and year. May precipitation was removed during model simplification. An auxiliary model including locality was retained to assess locality-level heterogeneity, but the final predictive model excluded the locality random-effect smooth because predictions were required for unsampled 10 × 10 km grid cells. Final predictions were generated for each grid cell and year using the corresponding spatial, topographic and April precipitation values. Laying dates were obtained from published data (Santangeli et al. 2014, Moreno-Rueda et al. 2019), as well as contributions by members of the Iberian Group on Harriers (GIA), different regional conservation programs for the species and TRAGSATEC-MITECO. References - Moreno-Rueda, Gregorio, Paula Lopezosa, y José M. Rivas. 2019. «Breeding Biology of Montagu’s Harrier Circus Pygargus in South-Eastern Spain». Ardeola 66 (1): 3-11. https://doi.org/10.13157/arla.66.1.2019.ra1. - Santangeli, Andrea, Enrico Di Minin, y Beatriz Arroyo. 2014. «Bridging the research implementation gap - Identifying cost-effective protection measures for Montagu’s harrier nests in Spanish farmlands». Biological Conservation 177: 126-33. https://doi.org/10.1016/j.biocon.2014.06.022.