COS 185-1 - Land use exerts stronger control over plant diversity than climate or soils in a tropical savanna of India

Background/Question/Methods

Understanding of how global environmental change shapes plant communities is central for conserving biodiversity. Although climate, soils, and land use change are among the most studied factors that influence biodiversity, their interacting effects have not been studied uniformly around the world. For example, compared to tropical forests, there are far fewer studies on global change in tropical savannas, despite their ecological and economic importance. We address this knowledge gap by quantifying the relative and interactive effects of climate, soil, and land use on savanna plant communities in western Maharashtra, India. Old-growth savannas of this 45,000-km² region span broad environmental gradients (e.g., 450 to 1900 mm annual precipitation), and are being converted to tillage agriculture and tree plantations, making these savannas an ideal system for studying interactions among climate and land use. We sampled n = 15 sites in each of four classes (old-growth savanna, active agriculture, agricultural fallows, and tree plantations) stratified across the region’s climatic and soil gradients. To understand the main and interactive effects of climate, soils and land use on plant diversity, we used linear mixed-models. To quantify these same controls over community composition, we used permutational multivariate analysis of variance (PERMANOVA) and ordinations (NMDS).

Results/Conclusions

Among the factors examined, we found that only land use, not climate or soils, explained a significant proportion of variation in herbaceous plant richness. Herbaceous plant richness was highest in old-growth savannas (12 species/m²) and tree plantations (10 species/m²), intermediate in fallows (9 species/m²), and lowest in tillage agriculture (4 species/m²). For community composition, land use explained the greatest variation (R² = 0.22), while climate (R² = 0.06) and soils (R² = 0.03) explained little. All land uses were compositionally distinct from one another. Perennial grasses, a keystone functional group in old-growth savannas, were rare in tillage agriculture, agricultural fallows, and tree plantations. Our findings support a growing body of literature suggesting that disturbance regimes, whether natural (e.g., fire and herbivores) or anthropogenic (e.g., land use change), can be far more important than climate and soils for understanding the maintenance or decline of tropical savanna biodiversity. Furthermore, land uses that disturb soil or increase tree cover eliminate species that are characteristic of old-growth savannas (e.g., perennial grasses), and these species do not recolonize fallows after agriculture is abandoned. We suggest revising conservation policies to focus on avoiding the conversion of Maharashtra’s last remaining old-growth savannas for agriculture and tree plantations.