Showing papers by "Andy Purvis published in 2021"

The impacts of biofuel crops on local biodiversity: a global synthesis

Abstract: Concerns about the impacts of climate change have led to increased targets for biofuel in the global energy market. First-generation biofuel crops contain oil, sugar or starch and are usually also grown for food, whereas second-generation biofuel is derived from non-food sources, including lignocellulosic crops, fast-growing trees, crop residues and waste. Biofuel production drives land-use change, a major cause of biodiversity loss, but there is limited knowledge of how different biofuel crops affect local biodiversity. Therefore, a more detailed understanding could inform more environmentally-conscious decisions about where to grow which biofuel crops. We synthesised data from 116 sources where a potential biofuel crop was grown and estimated how two measures of local biodiversity, species richness and total abundance, responded to different crops. Local species richness and abundance were 37% and 49% lower at sites planted with first-generation biofuel crops than in sites with primary vegetation. Soybean, wheat, maize and oil palm had the worst effects; the worst affected regions were Asia and Central and South America; and plant species richness and vertebrate abundance were the worst affected biodiversity measures. Second-generation biofuels had smaller, but still significant, effects: species richness and abundance were 19% and 25%, respectively, lower in such sites than in primary vegetation. Our models suggest that land clearance to cultivate biofuel crops reduces local biodiversity. However, the yield of biofuel from different crops influences the biodiversity impacts per unit of energy generated, and the geographic and taxonomic variation in effects are also relevant for making sustainable land-use decisions.

Fast, scalable, and automated identification of articles for biodiversity and macroecological datasets

Abstract: Aim: Understanding broad‐scale ecological patterns and processes is necessary if we are to mitigate the consequences of anthropogenically driven biodiversity degradation. However, such analyses require large datasets and current data collation methods can be slow, involving extensive human input. Given rapid and ever‐increasing rates of scientific publication, manually identifying data sources among hundreds of thousands of articles is a significant challenge, which can create a bottleneck in the generation of ecological databases. / Innovation: Here, we demonstrate the use of general, text‐classification approaches to identify relevant biodiversity articles. We apply this to two freely available example databases, the Living Planet Database and the database of the PREDICTS (Projecting Responses of Ecological Diversity in Changing Terrestrial Systems) project, both of which underpin important biodiversity indicators. We assess machine‐learning classifiers based on logistic regression (LR) and convolutional neural networks, and identify aspects of the text‐processing workflow that influence classification performance. / Main conclusions: Our best classifiers can distinguish relevant from non‐relevant articles with over 90% accuracy. Using readily available abstracts and titles or abstracts alone produces significantly better results than using titles alone. LR and neural network models performed similarly. Crucially, we show that deploying such models on real‐world search results can significantly increase the rate at which potentially relevant papers are recovered compared to a current manual protocol. Furthermore, our results indicate that, given a modest initial sample of 100 relevant papers, high‐performing classifiers could be generated quickly through iteratively updating the training texts based on targeted literature searches. These findings clearly demonstrate the usefulness of text‐mining methods for constructing and enhancing ecological datasets, and wider application of these techniques has the potential to benefit large‐scale analyses more broadly. We provide source code and examples that can be used to create new classifiers for other datasets.

Annual changes in the Biodiversity Intactness Index in tropical and subtropical forest biomes, 2001–2012

Abstract: Few biodiversity indicators are available that reflect the state of broad-sense biodiversity—rather than of particular taxa—at fine spatial and temporal resolution. One such indicator, the Biodiversity Intactness Index (BII), estimates how the average abundance of the native terrestrial species in a region compares with their abundances in the absence of pronounced human impacts. We produced annual maps of modelled BII at 30-arc-second resolution (roughly 1 km at the equator) across tropical and subtropical forested biomes, by combining annual data on land use, human population density and road networks, and statistical models of how these variables affect overall abundance and compositional similarity of plants, fungi, invertebrates and vertebrates. Across tropical and subtropical biomes, BII fell by an average of 1.9 percentage points between 2001 and 2012, with 81 countries seeing an average reduction and 43 an average increase; the extent of primary forest fell by 3.9% over the same period. We did not find strong relationships between changes in BII and countries’ rates of economic growth over the same period; however, limitations in mapping BII in plantation forests may hinder our ability to identify these relationships. This is the first time temporal change in BII has been estimated across such a large region.