Fire and grazing modify grass community response to environmental determinants in savannas: Implications for sustainable use

• We investigated effects of recurring forest fires on herbaceous species composition.
• Remotely sensed data was used to identify areas with different fire frequencies.
• Species heterogeneity and abundance of preferred grass reduced with fire frequency.
• Livestock grazing counteracted negative effect of high fire frequency on vegetation.
• Fire frequency should be managed at low levels to sustain ecosystem processes.

Tropical dry forests and savannas are important repositories of plant diversity and ecosystem services in the tropics. These ecosystems are also used extensively for grazing by livestock, and represent a critical element of the rural economy of many tropical countries. Fire is considered as a part of co-evolution in these ecosystems across the globe. However, in recent decades, there has been a shift in historical fire regime. Fire has become more frequent in these landscapes, and could be further enhanced under climate change. This poses threats to existing biodiversity, ecosystem processes, and rural economy. We asked how variability in fire frequency has influenced diversity and heterogeneity in grass species composition, and richness and abundance of grass species preferred by large herbivores (referred to as grazing acceptability) in a South Indian tropical savanna forest. We assumed that an increase in fire frequency acts as the active constraint and limits an ecosystem from attaining the maximum heterogeneity, and the maximum grazing acceptability (maximum richness and abundance of grass species preferred by herbivores) in its native settings. We used MODIS active fire and burned area products to estimate fire frequency across the landscape. A nested sampling approach was used to collect information on vegetation and soil at different fire frequencies. Quantile regression analyses indicated that diversity and heterogeneity in grass species composition as well as grazing acceptability decreased with increasing fire frequencies. We found that livestock grazing intervened with the observed vegetation patterns; grass species diversity and heterogeneity, and grazing acceptability increased with grazing intensity at lower quantiles. Other measured covariates, rainfall, and soil-fertility, alone were not able to explain the observed vegetation patterns in the landscape. The results show a need to control annual fires but allow and manage intermittent fires in this landscape. A complete suppression of fire is not desirable as fire releases nutrients from burning of deeper-rooted vegetation and thus acts as a periodic nutrient pump. It also played an important role in maintaining the grass cover by reducing shrub cover. Hence, it is important to consider the complex interactions between fires–grazers–soil–vegetation to develop effective management practices. We conclude that fire frequency should be managed at low to intermediate levels (one fire in every 5–9 years, resembling the native settings), and grazing regulated, in order to sustain wild and domestic herbivores, biodiversity, and other key ecosystem processes and services over the long-term.