Coordinated Experiments across Space in the EDGE Project
EDGE is a multisite study being conducted in six grasslands spanning climate gradients but also varying in their ecosystem attributes (Figure 6.3). The approach of this experiment solves three major problems typically confronting single-site studies, meta-analyses, and data synthesis approaches
FIGURE 6.3
(See color insert.) Map of the Midwest United States showing locations of the six experimental sites of the EDGE project. SEV, Sevilleta National Wildlife Refuge, NM; CPR, Central Plains Experimental range; HPG, High Plains Grassland Research Center; HAR, Hays Agricultural Research Center; KNZ, Konza Prairie Biological Center.
to predict large-scale responses to climate change. First, consistent methodology among sites removes the uncertainty associated with meta-analyses that often compare experiments having different techniques for measuring ecosystem responses or implementing treatments; this allows for better detection of patterns of sensitivity across sites. For example, if a researcher has to compare results from two experiments, both simulating drought by removing 50% of ambient rainfall, but one study did this by removing rainfall from May to September and the other by removing rainfall from April to August, the researcher would not be able to know how much of the difference in responses was due to differential sensitivities of the systems and how much was due to the differences of drought timing. Second, the EDGE study is conducting extensive monitoring and sampling of ecosystem attributes at each site before and during the experiment. Meta-analyses or data syntheses usually get partial information from each single study. This provides mechanistic understanding into why sensitivity differs across ecosystems and tracks changes in ecosystem attributes as well as overall responses. Third, the EDGE experiment is designed to have three pairs of sites (Figure 6.3), with each pair existing along a precipitation gradient ranging from ~240 to 860 mm of mean annual rainfall and a temperature gradient ranging from 7.6°C to 13.3°C of mean annual temperature. Yet each pair represents two different plant community types existing within a particular climate envelope. For example, the two arid sites at SEV; both receive ~250 mm of precipitation on average and are quite close in space. However, one is a Chihuahuan Desert grassland dominated by Bouteloua eriopoda (black grama) while the other is more like the shortgrass steppe dominated by Bouteloua gracilis (blue grama). Comparison of differential responses between these two sites versus responses across the broad climate gradient will provide insight into how plant communities drive sensitivity of ecosystems to drought within the context of climate.
