Abstract

Dryland vegetation frequently shows self‐organized spatial patterns as mosaic‐like structures of sources (bare areas) and sinks (vegetation patches) of water runoff and sediments with variable interconnection. Good examples are banded landscapes displayed by Mulga in semiarid Australia, where the spatial organization of vegetation optimizes the redistribution and use of water (and other scarce resources) at the landscape scale. Disturbances can disrupt the spatial distribution of vegetation causing a substantial loss of water by increasing landscape hydrological connectivity and consequently, affecting ecosystem function (e.g., decreasing the rainfall‐use efficiency of the landscape). We analyze (i) connectivity trends obtained from coupled analysis of remotely sensed vegetation patterns and terrain elevations in several Mulga landscapes subjected to different levels of disturbance, and (ii) the rainfall‐use efficiency of these landscapes, exploring the relationship between rainfall and remotely sensed Normalized Difference Vegetation Index. Our analyses indicate that small reductions in the fractional cover of vegetation near a particular threshold can cause abrupt changes in ecosystem function, driven by large nonlinear increases in the length of the connected flowpaths. In addition, simulations with simple vegetation‐thinning algorithms show that these nonlinear changes are especially sensitive to the type of disturbance, suggesting that the amount of alterations that an ecosystem can absorb and still remain functional largely depends on disturbance type. In fact, selective thinning of the vegetation patches from their edges can cause a higher impact on the landscape hydrological connectivity than spatially random disturbances. These results highlight surface connectivity patterns as practical indicators for monitoring landscape health.

Full Text
Paper version not known

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call