Carbon dynamics in an oldfield ecosystem: Was a multi-factor experiment the best approach for revealing responses to atmospheric and climatic change?

Background/Question/Methods

The recognition that ecosystems are subjected to multiple atmospheric and climatic drivers, which may act differently in combination than separately, gave rise to multi-factor climate change experiments. Climate change experiments, however, cannot create a future reality; they are experimental systems for testing hypotheses and developing process-level understanding of ecosystem response. From this perspective, we look back at the results of a multi-factor experiment in an oldfield ecosystem and consider whether it was an effective approach for determining the response of carbon flux and storage. The Oldfield Community Climate and Atmosphere Manipulation (OCCAM) experiment was initiated with utter disregard for Occam’s Razor. A community comprising seven oldfield species was established within open-top chambers and exposed to factorial combinations of ambient or elevated atmospheric [CO₂], ambient or elevated air temperature, and adequate or reduced precipitation. During the 6-year experiment we measured aboveground plant production, root growth, soil CO₂ efflux, soil carbon content, woody plant establishment, and many other responses. Data were analyzed in relation to main effects, interactions, and the indirect effects of treatments on soil moisture and plant community composition.

Results/Conclusions

Precipitation, warming, and [CO₂] altered soil moisture, and soil moisture was the dominant influence on most aspects of carbon dynamics. Plant community biomass increased in response to elevated CO₂, warming, and adequate precipitation. These effects were additive, with no significant interactions. Fine-root detrital input to soil also increased with adequate precipitation, but there was no effect of elevated CO₂ or warming and no interactions. CO₂ efflux from soil increased with elevated [CO₂] and adequate precipitation; the response to warming depended on season and indirect effects of warming on soil temperature, moisture, and carbon substrate. There were no CO₂ effects or interactions on soil carbon, but carbon cycling in the particulate organic matter pool was slower with reduced precipitation. Changes in plant community composition had a larger impact on ecosystem responses than the direct effect of treatments on individual plants. Based on observations of tree seedling establishment in the oldfield community, climate change effects on woody encroachment could be the largest effect on carbon cycling in this system in the long run. Given the high variability in many of these responses and the paucity of significant interactions, in future experiments increased replication should be given a higher priority than multi-factorial treatments.