Data for: Herbivores' impacts cascade through the brown food web in a dryland

Food webs can be conceptualized as being powered by energy derived from living and dead vegetation, respectively. Most food web research has focused on "green food webs" which begin with the consumption of living vegetation by herbivores. However, "brown food webs" which stem from the consumption of senescent vegetation by detritivores are also an important channel of energy transfer. In theory, herbivores have the potential to disrupt brown food webs by consuming plant material before it has the opportunity to senesce and become available for detritivores. Here we investigate the effects that grazing by kangaroos whose population had irrupted in the absence of an apex predator has on the brown food web in an arid environment. We compared the cover of living and dead vegetation and the abundances of detritivorous termites and their predators inside one ha herbivore exclosures and nearby control plots. Results show there was more cover of living and dead vegetation inside the exclosures. Similarly, abundances of termites and small vertebrate predators of termites were greater inside the exclosures. Our study provides evidence that consumption of plant material by irruptive herbivores can disrupt the functioning of the brown food web by reducing the flow of energy from plant biomass to termites which in turn translates to reduced abundances of termites and small vertebrates that feed on termites. Our findings have implications for conservation and management because they shed light on a previously unconsidered threat to the functioning of arid ecosystems, disruption of brown food webs by irruptive herbivores.R Studio is required to run the code which comes with the data. Funding provided by: Australian Research CouncilCrossref Funder Registry ID: http://dx.doi.org/10.13039/501100000923Award Number:We compared the cover of living and dead vegetation and the abundances of detritivorous termites and their predators inside one ha herbivore exclosures and nearby control plots. We used R Studio to process and analyse the data