1. Herbivores form an important link in the transfer of energy within a food web and are strongly influenced by bottom-up trophic cascades. Current hypotheses suggest that herbivore consumption and impact on plants should scale positively with plant resource availability. However, depending on the effect of resources on plant quantity and quality, herbivore impact may vary with different types of resources.2. We test four alternative hypotheses for the relationship between plant biomass, herbivore impact on plant biomass and plant resource gradients, each based on how resources might affect plant abundance and quality to herbivores. We measured plant biomass for four non-consecutive years in a long-term grazing exclosure experiment in the Serengeti National Park that includes seven sites that vary substantially in rainfall and soil and plant nitrogen (N) and phosphorus (P).3. Our data supported the hypothesis that herbivore impact is controlled by plant quality, in this case driven by plant P, as herbivore effects on biomass decreased with higher rainfall but increased with greater plant P, but not N content. To our knowledge, this is the first experimental study to indicate that wild mammalian herbivory is associated with P availability rather than N. 4. Synthesis. Our results suggest that P, in addition to water and N, may play a more important role in driving trophic interactions in terrestrial systems than previously realized. Given the uncertainties in rainfall due to climate change and increasing anthropogenic manipulations of global N and P cycles, our findings emphasize the need to consider multiple resources for understanding how trophic interactions might be influenced by environmental variables.

Herbivores form an important link in the transfer of energy within a food web and are strongly influenced by bottom-up trophic cascades. Current hypotheses suggest that herbivore consumption and impact on plants should scale positively with plant resource availability. However, depending on the effect of resources on plant quantity and quality, herbivore impact may vary with different types of resources. We test four alternative hypotheses for the relationship between plant biomass, herbivore impact on plant biomass, and plant resource gradients, each based on how resources might affect plant abundance and quality to herbivores. We measured plant biomass for four non-consecutive years in a long-term grazing exclosure experiment in the Serengeti National Park that includes seven sites that vary substantially in rainfall and soil and plant nitrogen (N) and phosphorus (P). Our data supported the hypothesis that herbivore impact is controlled by plant quality, in this case driven by plant P, as herbivore effects on biomass decreased with higher rainfall but increased with greater plant P, but not N content. To our knowledge, this is the first experimental study to indicate that wild mammalian herbivory is associated with P availability rather than N. Our results suggest that P, in addition to water and N, may play a more important role in driving trophic interactions in terrestrial systems than previously realized.

Plant mortality is a complex process influenced by both biotic and abiotic factors. In recent decades, widespread mortality events have been attributed to increasing drought severity, which has motivated research to examine the physiological mechanisms of drought-induced mortality, particularly hydraulic failure. Drought-based mortality mechanisms are further influenced by plant interactions with biota such as neighboring plants, insect pests, and microbes. In this review, we highlight some of the most influential papers addressing these biotic interactions and their influence on plant mortality. Plant-plant interactions can be positive (facilitation), neutral, or negative (competition), depending on drought intensity and neighbor identity. For example, standscale mortality likely increases with basal area (an index of competition). However, the diversity of forest stands matters, as more diverse forests suffer less mortality from drought than species-poor forests. Dense forest stands also increase bark beetle attack frequency, which can exacerbate drought stress and mortality, particularly for fast-growing species with lower defense allocation. In some cases, however, drought stress can alleviate biotic attack, depending on feedbacks between plant and pest physiology. Finally, plant interactions with beneficial microorganisms can increase drought tolerance, reduce the likelihood of mortality, and even extend plant distributions into drier habitats. Our review suggests more work is needed in natural herbaceous plant communities as well as dry tropical ecosystems where mortality mechanisms are less understood. Overall, relatively few studies directly link biotic interactions with the physiological mechanisms of mortality. Simultaneous manipulations of biotic interactions and measurements of physiological thresholds (e.g., xylem cavitation) are needed to fully represent biotic interactions in predictive models of plant mortality.

Variation in defense traits likely depends on access to different resources and risk from herbivory. Plant defense theories have predicted both positive and negative associations between defense traits and access to resources, but relatively few studies have explored intraspecific variation in defense traits along multiple resource and mammalian herbivory risk gradients. We assessed relationships between herbivory intensity, multiple resources, and plant defense traits using a widely distributed tropical savanna herb, Solanum incanum. As independent measures of risk from large mammal herbivores are rare, we used a satellite-based vegetation index to predict herbivory intensity at the landscape scale. We found that the satellite-based estimate of herbivory intensity was positively associated with browser abundance and total soil P, but negatively associated with rainfall. Intraspecific defense traits too varied substantially across sites (n=43) but only variation in spine density was associated with herbivory intensity and plant resources, such that spine density was positively associated with both rainfall and soil P, but bimodally associated with herbivory intensity. Taken together, it suggests that defenses maybe favored either where resources for defense are abundant under low but still present risk (i.e, at high rainfall sites) or where resource-expensive plant tissue is at high risk (i.e, at high soil P sites). This hints at the possibility of a shift from a resource-associated (bottom-up) to an herbivory-associated (top-down) control of allocation to defenses along an environmental gradient. Additionally, the independent effect of soil P on a carbon-based defense, spine density, suggests potential for resources that are not components of defenses to also influence allocation to defense traits. Thus, our study provides evidence for the influence of multiple drivers, resources, and herbivory intensity, on anti-herbivore defenses and their shifting relative importance on allocation to defenses along an environmental gradient.

Leaf phenology based classification of woody species into discrete evergreen and deciduous categories is widely used in ecology, but these categories hide important variation in leaf phenological behaviour. Few studies have examined the continuous nature of deciduousness and our understanding of variation in quantitative estimates of leaf shedding behaviour and the causes and consequences of this is limited. In this study we monitored leaf phenology in 75 woody species from a seasonally dry tropical forest to quantify three quantitative measures of deciduousness, namely: maximum canopy loss, duration of deciduousness, and average canopy loss. Based on proposed drought tolerance and drought avoidance strategies of evergreen and deciduous species, respectively, we tested whether the quantitative measures of deciduousness were related to leaf functional traits. Additionally, to understand the functional consequences of variation in deciduousness we examined relationships with the timing of leaf flushing and senescing. We found wide and continuous variation in quantitative measures of deciduousness in these coexisting species. Variation in deciduousness was related to leaf function traits, and the timing of leaf flushing. Along a continuous axis ranging from evergreen to deciduous species, increasing deciduousness was associated with more acquisitive leaf functional traits, with lower leaf mass per area and leaf dry matter content, and greater leaf nitrogen content. These results indicate that the continuous nature of deciduousness is an important component of resource acquisition strategies in woody species from seasonally dry forests.