Irruptions in plant and animal populations are not uncommon but the factors underlying irruptions are rarely explored quantitatively. In addition, it has been suggested that these irruptions may be reduced by predators or herbivores, but there is a paucity of controlled experimental evidence. Using data from the Kenya Long-term Exclosure Experiment (KLEE), we show that populations of perennial Hibiscus spp. (primarily H. flavifolius) show multiple short-term irruptions a year after rainy periods, increasing in abundance in some cases by more than an order of magnitude before declining in ensuing months and years. We demonstrate that these irruptions are largely limited to experimental plots from which large mammalian herbivores have been excluded, particularly megaherbivores (elephants and giraffes). This represents a rare controlled replicated experimental demonstration of top-down regulation of irruptions. African elephants and giraffes are often at greater risk of local extinction than other large mammals, and their absence appears to destabilize this African savanna ecosystem, providing additional support for their conservation.

The Kenya Long-term Exclosure Experiment (KLEE) was established in 1995 in a semi-arid savanna rangeland on the Laikipia Plateau to examine the separate and combined effects of livestock, wildlife, and megaherbivores (elephants and giraffes) on their shared environment and on each other. The long-term nature of this experiment also allowed us to measure these effects and related questions of stability and resilience in the context of multiple drought-rainy cycles. Here we outline some of the lessons learned over the last 29 years. In particular, we summarize three ways that KLEE exemplifies the value of long-term studies: 1) identifying experimental effects that take a many years to express themselves, 2) quantifying the effects of different years, especially multiple droughts, and 3) capturing time periods long enough to see the signature of systemic, anthropogenic change in the broader landscape. Across all aspects of a long-term study such as this one, there is a need to incorporate both consistency and flexibility to ensure deeper understanding.

Both termites and large mammalian herbivores (LMH) are savanna ecosystem engineers that have profound impacts on ecosystem structure and function. Both of these savanna engineers modulate many common and shared dietary resources such as woody and herbaceous plant biomass, yet few studies have addressed how they impact one another. In particular, it is unclear how herbivores may influence the abundance of long-lived termite mounds via changes in termite dietary resources such as woody and herbaceous biomass. While it has long been assumed that abundance and areal cover of termite mounds in the landscape remains relatively stable, most data are observational, and few experiments have tested how termite mound patterns may respond to biotic factors such as changes in large herbivore communities. Here, we use a broad tree density gradient and two landscape-scale experimental manipulations—the first a multi-guild large herbivore exclosure experiment and the second a tree removal experiment– to demonstrate that patterns in termite mound abundance and cover are unexpectedly dynamic. Termite mound abundance, but not areal cover not significantly, is positively associated with experimentally controlled presence of cattle, but not wild mesoherbivores (15-1000 kg) or megaherbivores (elephants and giraffes). Herbaceous productivity and tree density, termite dietary resources that significantly affected by different LMH treatments, are both positive predictors of termite mound abundance. Experimental reductions of tree densities are associated with lower abundances of termite mounds. These results reveal a richly interacting web of relationships among multiple savanna ecosystem engineers and suggest that termite mound abundance and areal cover is intimately tied to herbivore-driven resource availability.