Palearctic passerine migrant declines in African wintering grounds in the Anthropocene (1970–1990 and near future): A conservation assessment using publicly available GIS predictors and machine learning

Abstract

The Anthropocene causes many massive and novel impacts, e.g., on migratory birds and their habitats. Many species of migratory birds have been declining on the Palearctic-African flyway in recent decades. To investigate possible impacts on a continental scale, we used 18 predictors extracted from 16 publicly available GIS layers in combination with machine learning methods on the sub-Saharan distributions of 64 passerine migrant species. These bird species were categorized as having experienced a 'Large Decline' (n = 12), a 'Moderate Decline' (n = 6) or 'No Decline' (n = 46) based on European census data from 1970 to 1990. Therefore, we present the first study for these species which uses publically available Open Access GIS-data and a multivariate (n = 18) and multi-species (n = 64) machine learning approach to deduce possible past impacts. We furthermore modelled likely future human population change and climate change impacts. We identified three predictor themes related to the distributions and declines of these migratory birds: (I) locations, represented by African ecosystems, countries, and soil types; (II) human population pressures and land-use intensities, the latter represented by land-use categories, habitat area, and cropland proportion; and (III) climatic predictors. This is the first study to relate migratory bird declines to human population pressures and land-use intensities using this type of analysis. We also identified areas of conservation concern, such as the Sahel region. Our models also predict that the declining trends of migratory birds will continue into the foreseeable future across much of Africa. We then briefly discuss some wider conservation implications in the light of the increasing drivers of biodiversity change associated with the Anthropocene as well as some possible solutions. We argue that only comprehensive systemic change can mitigate the impacts on the migratory birds and their habitats.

Dataset compilation by GRASS GIS for thematic mapping of Antarctica: Topographic surface, ice thickness, subglacial bed elevation and sediment thickness

“…The attempts of combining logic of programming languages and machine learning approaches to cartography resulted in several outputs such as e.g. combination of machine learning classifiers along with GIS techniques (Motta et al 2021), AWK language for processing tables from geodata through conversion and formatting (Lemenkova 2019e), modeling links between environmental, biodiversity and climate change impacts (Walther and Huettmann 2021), integration of GRASS GIS, Python, TeX language, or artificial neural networks for geological engineering modeling (Bragagnolo et al 2020, Lemenkov andLemenkova 2021b), to mention a few of them.…”