Interannual and interseasonal Soil CO2 efflux and VOC exchange rates in a Mediterranean holm oak forest in response to experimental drought

Climate models predict drier conditions in the next decades in the Mediterranean basin. Given the importance of soil CO2 efflux in the global carbon balance and the important role of soil monoterpene and volatile organic compounds (VOCs) in soil ecology, we aimed to study the effects of the predicted drought on soil CO2, monoterpenes and other VOC exchange rates and their seasonal and interannual variations. We decreased soil water availability in a Mediterranean holm oak forest soil by means of an experimental drought system performed since 1999 to the present. Measurements of soil gas exchange were carried out with IRGA, GC and PTR-MS techniques during two annual campaigns of contrasting precipitation. Soil respiration was twice higher the wet year than the dry year (2.27 +/- 0.26 and 1.05 +/- 0.15, respectively), and varied seasonally from 3.76 +/- 0.85 mu molm(-2) 2 S-1 in spring, to 0.13 +/- 0.01 mu mol m(-2) s(-1) I in summer. These results highlight the strong interannual and interseasonal variation in CO2 efflux in Mediterranean ecosystems. The drought treatment produced a significant soil respiration reduction in drought plots in the wet sampling period. This reduction was even higher in wet springs (43% average reduction). These results show (1) that soil moisture is the main factor driving seasonal and interannual variations in soil respiration and (2) that the response of soil respiration to increased temperature is constrained by soil moisture. The results also show an additional control of soil CO2 efflux by physiology and phenology of trees and animals. Soil monoterpene exchange rates ranged from -0.01 to 0.004nmol m(-2)S(-1), thus the contribution of this Mediterranean holm oak forest soil to the total monoterpenes atmospheric budget seems to be very low. Responses of individual monoterpenes and VOCs to the drought treatment were different depending on the compound. This suggests that the effect of soil moisture reduction in the monoterpenes and VOC exchange rates seems to be dependent on monoterpene and VOC type. In general, soil monoterpene and other VOC exchange rates were not correlated with Soil CO2 CfflUX- In all cases, only a low proportion of variance was explained by the soil moisture changes, since almost all VOCs increased their emission rates in summer 2005, probably due to the effect of high soil temperature. Results indicate thus that physical and biological processes in soil are controlling soil VOC exchange but further research is needed on how these factors interact to produce the observed VOCs exchange responses. (c) 2007 Elsevier Ltd. All rights reserved.