| Popis: |
We present an exploratory study examining the use of airborne remote-sensing observations to detect ecological responses to elevated CO2 emissions from active volcanic systems. To evaluate these ecosystem responses, existing spectroscopic, thermal, and lidar data acquired over forest ecosystems on Mammoth Mountain volcano, California, were exploited, along with in situ measurements of persistent volcanic soil CO2 fluxes. The elevated CO2 response was used to statistically model ecosystem structure, composition, and function, evaluated via data products including biomass, plant foliar traits and vegetation indices, and evapotranspiration (ET). Using regression ensemble models, we found that soil CO2 flux was a significant predictor for ecological variables, including canopy greenness (normalized vegetation difference index, NDVI), canopy nitrogen, ET, and biomass. With increasing CO2, we found a decrease in ET and an increase in canopy nitrogen, both consistent with theory, suggesting more water- and nutrient-use-efficient canopies. However, we also observed a decrease in NDVI with increasing CO2 (a mean NDVI of 0.27 at 200 g m−2 d−1 CO2 reduced to a mean NDVI of 0.10 at 800 g m−2 d−1 CO2). This is inconsistent with theory though consistent with increased efficiency of fewer leaves. We found a decrease in above-ground biomass with increasing CO2, also inconsistent with theory, but we did also find a decrease in biomass variance, pointing to a long-term homogenization of structure with elevated CO2. Additionally, the relationships between ecological variables changed with elevated CO2, suggesting a shift in coupling/decoupling among ecosystem structure, composition, and function synergies. For example, ET and biomass were significantly correlated for areas without elevated CO2 flux but decoupled with elevated CO2 flux. This study demonstrates that (a) volcanic systems show great potential as a means to study the properties of ecosystems and their responses to elevated CO2 emissions and (b) these ecosystem responses are measurable using a suite of airborne remotely sensed data. |
