干旱区生态环境知识资源中心

We studied the spatial patterns and temporal dynamics of vegetation structural responses to precipitation variation in grassland, transitional, and desertified-shrubland ecosystems in an 800 km(2) region of Northern Chihuahua, USA. Airborne high-fidelity imaging spectroscopy data collected from 1997 to 2001 provided spatially detailed measurements of photosynthetic and senescent canopy cover and bare soil extent. The observations were made following wintertime and summer monsoonal rains, which varied in magnitude by >300% over the study period, allowing an assessment of ecosystem responses to climate variation in the context of desertification.

Desertification caused a persistent increase in both photosynthetic vegetation (PV) and bare soil cover, and a lasting decrease in nonphotosynthetic vegetation (NPV). We did not observe a change in the spatial variability of PV cover, but its temporal variation decreased substantially. In contrast, desertification caused the spatial variability of NPV to increase markedly, while its temporal variation did not change. Both the spatial and temporal variation of exposed bare surfaces decreased with desertification. Desertification appeared to be linked to a shift in seasonal precipitation use by vegetation from mainly summer to winter inputs, resulting in an apparent decoupling of vegetation responses to inter-annual monsoonal variation. Higher winter rainfall led to decreased springtime spatial variability in the PV cover of desertified areas. Higher summer rainfall resulted in decreased PV cover variation in grassland, transition and desertified-shrubland regions. The effects of desertification on NPV dynamics were more than three times greater than on PV or bare soil dynamics. Using remotely sensed PV and NPV as proxies for net primary production (NPP) and litter dynamics, respectively, we estimated that desertification decreases the temporal variability of NPP and increases spatial variation of litter production and loss. Quantitative studies of surface biological materials and ecosystem processes can now be measured with high ’structural’ detail using imaging spectroscopy and shortwave-infrared spectral mixture analysis.