We cleaned and standardized a dataset of species occurrences in the 19th century in Spain, containing information on 480 species and 135 genus of plants and animals in over 23,000 localities across the country. The species data was already available and formatted for the 16th and 21st century. In parallel, we extracted the historical data that was to be used to model the species distributions, that is historical bioclimatic variables and historical human density for the 16th and 19th century, as well as a static Digital Elevation Model. We then built an algorithm to run Environmental Niche Models with this data for a sample of 27 mammal species for which the number of occurrences was sufficient and were indeed successful in modelling the historical distribution of those species.
To quantify the impact of the baseline used for the calibration of Environmental Niche Models and the resulting habitat suitability mapping, we then build several models with differing time used for calibration. Here we modelled the habitat suitability of species (1) in the past (19th centuries) calibrated with historical environmental variables and species occurrences; (2) in the past calibrated with present environmental variables and species occurrences; and (3) in the present calibrated with present environmental variables and species occurrences. Albeit preliminary and for a limited number of species, the results show some marked differences in the modelled habitat affinity of species when using historical or contemporary data for calibration, even though those data are "only" a century apart.
In parallel to running the models, we also built a database of historical ecosystem services supply by integrating the historical species lists with available national assessment of traditional knowledge. This database uses the framework of Nature’s Contribution to People (NCPs) to increase its usability beyond the project. In total, we were able to map 18 types of NCPs to 110 and 172 species for which we have occurrence data available in the 16th and 19th century respectively.
Lastly, we reviewed the diversity of historical sources of biodiversity relevant information, from written material to paleo-archives and artwork, and discuss across all sources types the data that can be mobilized from them, focusing on species, ecosystems, and socio-ecological systems. While doing this, we also documented examples of historical data integration as well as emerging tools and technologies that could improve ongoing historical data mobilization efforts. This work also allowed us to design a 7-point strategy to improve the availability and use of historical data for biodiversity research and conservation.