Summer heat and drought extremes trigger unexpected changes in productivity of a temperate annual/biannual plant community

One consequence of climate warming is an increase in frequency and intensity of heat and drought extremes. The need for realistic experiments examining the effects of extreme events, including their possible interaction, is therefore high. Here, we imposed heat and drought events in a full factorial design, on experimental plant communities consisting of three temperate, annual or biannual herbaceous species. The extremes were simulated in either early or late summer (July or August), as summer is the season where they are expected to have the greatest impact. The 10-day heat waves were simulated with infrared lamps and drought (either 20 or 17 days) was created by withholding water input and removing the water table. During exposure, effects of drought extremes on leaf physiology were predominantly negative in July, and absent in August. Effects of heat extremes were positive, negative or neutral depending on species and period. Heat and drought were additive in July, but interacted in August. The aforementioned negative effects were not reflected in total (aboveground plus belowground) end-of-season community biomass, which, surprisingly, was stimulated in response to drought extremes in both periods. Effects of heat extremes varied but never reduced biomass, likewise to combined heat and drought. This increase in total community biomass originated exclusively from stimulated root growth. The exact mechanism for this unexpected result could not be ascertained, but greater whole-plant nitrogen stocks clearly indicated enhanced nutrient availability. The latter may have arisen from increased net mineralization or from greater root exploration under the influence of mid-season drought. Shifts in species dominance apparently also played a role, as species with pronounced allocation to roots were released from competition. This study demonstrates that community responses to climate extremes are not necessarily negative. The observed responses depended on the soil moisture conditions they induced, which in our experiments varied with the timing of the imposed extremes, even within the same season. (C) 2012 Elsevier B.V. All rights reserved.