Large-scale changes in semi-arid vegetation, such as that found in the Owens Valley, are not considered natural phenomena. Past studies and common sense suggest that natural climate variability should not adversely affect natural vegetation. This theory is based in part on the known drought tolerance of these plant communities. The long-term record of precipitation in the Owens Valley reports several droughts equal to or greater than the six-year drought that occurred in the last decade. Had these droughts led to effects as large as what we saw in the decade, it is unlikely that vegetation conditions recorded by LADWP prior to the drought would have been the same. Therefore, we needed to include human activities in our analyses of potential causes. Based on our fieldwork and a comparison of groundwater levels and vegetation changes, we attributed groundwater pumping by LADWP, which caused a drop in water tables across the valley, to the large decreases in live cover.

The Landsat TM satellite imagery allows us to extend ICWD's field measurements to a regional scale. Therefore, our results broaden the scope to the entire valley, instead of limited point measurements. Our results are as follows: Some meadow and shrub-meadow communities affected by pumping and insufficient recharge showed a greater than 50% decrease in live cover in more than 5%, or 12,500 acres, of the study area. An additional 15%, or 37,500 acres, of the groundwater dependent shrub and meadow communities were affected to a lesser extent by the same drivers. In many of these areas, the imposed stresses have resulted in a shift towards communities dominated by non-native annuals, rather than native perennial grasses and shrubs. Communities which were already dominated by annuals were least likely to be affected by pumping, but showed a strong dependence on annual rainfall. This response was observed over 22%, or 55,000 acres, of the study area. Modest increases in live cover were found in 3%, or 7,500 acres, of the study area, much of this attributable to irrigation practices. Within the limits of our observations, the remaining 55%, or 137,500 acres, of the study area was left largely unchanged despite low annual precipitation, primarily because these communities are either not dependent on groundwater or lie in regions of the valley where there was little change in the depth to groundwater over the 16-year study period.

In this study, we have found that intense land use coupled with natural climate variability can have far reaching effects on a semi-arid ecosystem such as the Owens Valley. Most significantly, these forces can result in a shift toward communities dominated by non-native annuals (weeds). We saw this shift go to completion (100% weeds) in 1%, or 2,500 acres, of the study area, and while our analyses are still underway, we consider this to be a significant change. Over 10%, or 25,000 acres, of the study area showed some increase in weed live cover. Land use history studies have shown that areas that experienced pumping- and drought-induced shifts towards weed-dominated systems during the past 100 years often remained in this unnatural state for several decades. The consequences of this type of non-elastic change are understudied, however, likely will include issues of loss in biodiversity (both plant and animal) and changes in water, carbon, and nitrogen budgets.

The remote sensing study will continue for at least another year. Results of the study may be used to develop a tool to predict on a regional scale the effects of drought and groundwater pumping on Owens Valley vegetation.