Résumé
This dataset compiles (i) observations of Ambrosia artemisiifolia in France between 1970 and 2023 aggregated per EPCI (i.e. a polygon aggregating connex municipalities), (ii) associated environmental predictors and (iii) spatial connectivity between EPCIs.
The complete dataset (processed_data.RData) is ready for use with the dSDM R package, allowing to calibrate Bayesian spatial dynamics models from field observations across space and time.
Observations: There are two types of observations, presence-absence surveys (PA) and presence-only records (PO). Nearly 20 thousand standardized PA surveys were gathered from Biovigilance (Fried et al., 2008) and ENI-HTV (Fried et al., 2022), two programs dedicated to the monitoring of the agricultural flora. Each survey is an exhaustive inventory of plant species in a 2000m² vegetation plot, from which we extracted the presence or absence of Ambrosia artemisiifolia. The PA surveys are distributed across 1243 EPCIs and span 18 years between 2002 and 2019, with 608 presences in total. About 10 thousands PO records of A. artemisiifolia were also gathered from national botanical conservatories and citizen science platforms (Pl@ntNet, iNaturalist, Observation.org). These source datasets monitor a large number of plant species opportunistically. We pulled all valid records from all monitored species and aggregated them per source dataset, day and location, which is interpreted as a visit, i.e. an elementary measure of sampling effort. We included about 5 million visits in total, counted in a table per municipality, year, source dataset and day, provided as detection covariates, along with the number of detections of A. artemisiifolia.
Environmental variables: For each EPCI and year, two climate variables (Era5-Land, see Muñoz 2019) and the percent cover of five simplified types of land cover (CORINE Land Cover) is also provided. These variables were selected as they are known to impact the species survival and reproduction in the literature.
Connectivities: Estimates of connectivity between each pair of municipalities and for three dispersal pathways are also provided: Dispersal by vehicles along highways (derived from the highway network obtained from OpenStreetMap), by watercourses (derived from the river network obtained from BD Topage) and by agricultural machinery (derived from CORINE land cover).
We also provide raw data components, including the spatial polygons of the EPCIs, and the R code to reproduce some of the pre-processing steps.
References:
Fried, G., Norton, L. R., & Reboud, X. (2008). Environmental and management factors determining weed species composition and diversity in France. Agriculture, ecosystems & environment, 128(1-2), 68-76. https://doi.org/10.1016/j.agee.2008.05.003
Fried, G., Le Corre, V., Rakotoson, T., Buchmann, J., Germain, T., Gounon, R., ... & Chauvel, B. (2022). Impact of new management practices on arable and field margin plant communities in sunflower, with an emphasis on the abundance of Ambrosia artemisiifolia (Asteraceae). Weed Research, 62(2), 134-148. https://doi.org/10.1111/wre.12522
Muñoz Sabater, J. (2019): ERA5-Land monthly averaged data from 1950 to present. Copernicus Climate Change Service (C3S) Climate Data Store (CDS). DOI: 10.24381/cds.68d2bb30 (Accessed on 14-04-2024)