Predators and rainfall control spatial biogeochemistry in a landscape of fear

Oswald J. Schmitz; PNAS September 29, 2020 117 (39) 24016-24018

Figure: Animal-driven landscape-scale biogeochemical processes are dependent on the interplay among herbivore nutrient demand and release, rainfall, and vulnerability to predation. The ratio of key nutrients such as nitrogen [N] and phosphorus [P] released in herbivore dung varies with herbivore body mass (A). Varying rainfall levels favor different-sized herbivores (B) through a trade-off in forage production and quality. High rainfall produces large quantities of low-quality forage, which favors large herbivores with high forage demands but that can digest its poorer quality. Lower rainfall leads to lesser amounts of higher-quality forage, which favors smaller herbivores. The vulnerability of herbivores to predation decreases with herbivore body mass (C). Together, the amount of rainfall and level of potential predation risk should create a joint gradient in the sizes of herbivores occurring in different landscape locations (D) and create concomitant gradients in N:P ratios released in herbivore dung and taken up in foliar tissue of forage plants (E). These gradients explain contingency in biogeochemical processes across landscapes.

Scientists have long pondered what the world might have been like during the geological epochs immediately predating the late Pleistocene extinction events, when very large, wild mammals occupied almost every ecosystem on Earth (1⇓⇓–4). By piecing together whatever evidence can be gathered from the fossil record, it is speculated that these megasized species (i.e., species with body masses ranging from as small as tens of kilograms to as large as thousands of kilograms) had profound impacts on ecosystems worldwide. It is thought that, through their foraging and trampling, these herbivores determined the relative abundance of different vegetation types (e.g., woody vs. grassland) across landscapes; through selective feeding, they mediated competition among plant species within vegetation types; and through their behavior (habitat selection and migration) and physiology, they spatially redistributed plant-derived nutrients such as nitrogen (N) and phosphorus (P), thereby having an important hand in shaping the spatial patterning of biogeochemical properties across landscapes (4⇓–6). It is further speculated that the nature and degree of impact depended on the interplay between herbivore body size and biophysical aspects of ecosystems such as soil fertility and amounts of rainfall (5). If only there was a way to experimentally test some of these speculations. In PNAS, le Roux et al. (7) report on a study that does just that, but with an interesting twist.

The study takes advantage of the fact that relic assemblages of large, wild mammalian herbivores, resembling the size range found during the Pleistocene, currently exist within protected areas in African savanna ecosystems (5). Such places offer great testbeds to conduct landscape-scale experiments along soil nutrient and rainfall gradients in order to measure the singular and combined effects of different-sized wild herbivores on the vegetation structure and functioning of ecosystems (5, 8). The interesting twist of the le Roux et al. study is that along a rainfall gradient it further resolves the effects of another important factor determining herbivore impact. It is one that is entertained in discussions about how paleoecosystems functioned (3⇓–5) but not fully considered as part of conventional speculations about what drives large herbivore impacts on biogeochemistry. That understudied factor is the role of predators. In particular, how shifts in herbivore behavior induced by the mere risk that they may be captured by predators can shape the patterning of biogeochemical properties across landscapes (9). The study resolves a complex interweaving of factors controlling ecosystem functioning to provide a clear picture of the role that predators and herbivores play in shaping variation in biogeochemical properties across landscapes.

The landscape of Hluhluwe-iMfolozi Park in South Africa, where the study was done, is dotted with numerous grazing lawns interspersed among savanna woodlands. Grazing lawns tend to arise in landscape locations with comparatively higher soil nutrient levels (10). These open landscape locations are characterized by large swaths of highly nutritious, short-statured grass species that are kept in this state by frequent, recurrent herbivore grazing that also keeps less nutritious, tall-statured competitor grass species at bay (10). The level of grazing-lawn productivity is generally related to rainfall levels. As well, their high productivity may be sustained by rapid plant uptake of limiting nutrients that are released from herbivores in their dung and urine (10). This herbivore-driven recycling feedback may enhance productivity by circumventing the slower process of nutrient release arising from soil microbial decomposition and mineralization of senescent plant matter (11). However, the general occurrence of fast recycling—especially of N—among grazing lawns is debated (10).

Still, grazing lawns are nutrient hot spots for herbivores, providing them with N for body protein synthesis and P for bone development (10). The demand for N and P, however, varies with herbivore body mass. Larger herbivores require higher amounts of P per unit body mass than do smaller herbivores owing to the need to develop robust skeletons to support their large mass (12). However, they also require disproportionately less N (13). The challenge faced by all herbivores is that the ratios of N and P in their plant resources often do not closely match their demand (14, 15). Thus, when N:P in resource supply and organismal demand is mismatched, herbivores should respond by releasing the nutrient taken in excess (14, 15). According to these physiological first principles, the nutrient that is released as excess will vary with herbivore body mass. Smaller herbivores, having high N:P demands, should disproportionately release more P (i.e., have dung with low N:P ratios). As P demand increases with herbivore body mass, there should be a concomitant increase in the proportion of N released. Consequently, the ratio of N:P released in herbivore dung should increase with body mass (Fig. 1A).

See more: https://www.pnas.org/content/117/39/24016