Global models and predictions of plant diversity based on advanced machine learning techniques

Abstract

Despite the paramount role of plant diversity for ecosystem functioning, biogeochemical cycles, and human welfare, knowledge of its global distribution is still incomplete, hampering basic research and biodiversity conservation.Here, we used machine learning (random forests, extreme gradient boosting, and neural networks) and conventional statistical methods (generalized linear models and generalized additive models) to test environment-related hypotheses of broad-scale vascular plant diversity gradients and to model and predict species richness and phylogenetic richness worldwide. To this end, we used 830 regional plant inventories including c. 300 000 species and predictors of past and present environmental conditions.Machine learning showed a superior performance, explaining up to 80.9% of species richness and 83.3% of phylogenetic richness, illustrating the great potential of such techniques for disentangling complex and interacting associations between the environment and plant diversity. Current climate and environmental heterogeneity emerged as the primary drivers, while past environmental conditions left only small but detectable imprints on plant diversity.Finally, we combined predictions from multiple modeling techniques (ensemble predictions) to reveal global patterns and centers of plant diversity at multiple resolutions down to 7774 km 2 . Our predictive maps provide accurate estimates of global plant diversity available at grain sizes relevant for conservation and macroecology.

Phylogenetic endemism of the world’s seed plants

“…Generally, larger regions host more endemics as well as wide-ranged species because of their overall higher plant diversity (47, 58). Here, we observed a negative association between region area and PE when species range sizes were measured as the total area of the occupied regions.…”

“…Tropical montane regions are well-known centers of taxonomic and phylogenetic plant diversity (e.g., ref. 47). Due to their complex topography and geologic and climatic histories, they also hold exceptionally narrow-ranged species (48).…”

“…The merged phylogeny included 267,105 out of 273,838 species (97.5 %) in the dataset. Adding phylogenetic uncertainties via polytomies is a common issue when working with large phylogenetic trees (47, 63). However, phylogenetic metrics based on phylogenetic trees with higher numbers of polytomies have been shown to be highly correlated with metrics based on trees without or with fewer polytomies (64).…”

Macroecology of plant diversity across spatial scales

“…Cai et al . (2023), in this issue of New Phytologist (Cai et al ., 2023; pp. 1432–1445), by benefiting from up‐to‐date machine learning techniques, present a framework to bring a new quality to the global species and phylogenetic diversity predictions across a range of regional grain sizes.

‘The high‐resolution maps provided by Cai et al .

…”

“…In addition, the multiple determinants of species richness might interact in complex ways, being also habitat dependent and scale dependent (Keil & Chase, 2019). Cai et al (2023), in this issue of New Phytologist (Cai et al, 2023; pp. 1432-1445, by benefiting from up-to-date machine learning techniques, present a framework to bring a new quality to the global species and phylogenetic diversity predictions across a range of regional grain sizes.…”