Interactive effects of natural and human disturbances on vegetation dynamics across landscapes.

Accurate measures of human effects on landscape processes require consideration of both the direct impacts from human activities and the indirect consequences of the interactions between humans and the landscape. This is particularly evident in systems experiencing regular natural disturbances such as in the mountainous areas of southwestern China, where the remaining population of giant pandas (Ailuropoda melanoleuca) is supported. Here the spatiotemporal patterns of human impacts, forests, and bamboo episodic die-offs combine to determine the distribution of panda habitat. To study the complex interactions of humans and landscapes, we developed an integrated spatiotemporally explicit model of household activities, natural vegetation dynamics, and their impacts on panda habitat. Using this model we examined the direct consequences of local fuelwood collection and household creation on areas of critical giant panda habitat and the indirect impacts when coupled with vegetation dynamics. Through simulations, we found that over the next 30 years household impacts would result in the loss of up to 30% of the habitat relied on by pandas during past bamboo die-offs. The accumulation and spatial distribution of household impacts would also have a considerable indirect influence on the spatial distribution of understory bamboo. While human impacts influence both bamboo die-off and regeneration, over 19% of pre-existing low-elevation bamboo habitat may be lost following an episodic die-off depending on the severity of the impacts and timing of the die-offs. Our study showed not only the importance of the spatial distribution of direct household impacts on habitat, but also the far-reaching effects of the indirect interactions between humans and the landscapes they are modifying.