全球变化知识资源中心

Soils around the world are being exposed to weather events which are unprecedented in recent history. To maintain the delivery of soil-related ecosystem services and to promote greater soil resilience it is essential to understand how plant-soil systems respond to these extreme events. In this study we replicated a recent period of extreme rainfall and prolonged spring flooding in a temperate grassland which had no previous history of flooding. Intact soil mesocosms (Eutric Cambisol) 1 kg weight were subjected to a simulated long-term spring flood (15 degrees C, 2 months) and maintained in the light with above ground indigenous vegetation (Loliumperenne L) or dark with and without indigenous vegetation to simulate different flood typologies. In comparison to a no flood control treatment, nutrient cycling, water quality, air quality (greenhouse gas emissions), habitat provision and biological population regulation shifts were evaluated. Flooding induced a rapid release of nutrients into the soil solution and overlying floodwater, resulting in potential nutrient losses up to 15 mg Fe, 16 mg NH4+, 360 mg DOC and 28 mg DON, per mesocosm. The presence of plants increased the rate of nutrient release (especially P), with the effects magnified when light transmission through the floodwater was restricted (1.3 mg P vs 0.2 mg P, per mesocosm). Flooding induced a rapid decline in redox potential and subsequent production of CH4, especially in the darkened treatments (10 and between 11 and 16 times higher than the control, without and with light restrictions, respectively). Upon removal of the floodwater, the accumulated NH4+ was nitrified leading to a shift in greenhouse gas emissions, from CH4 to N2O emissions. N2O was only significantly produced in the mesocosms kept under light restrictions (13 times higher than in other two treatments). Flooding eliminated earthworms, reduced grass production after soil recovery (from 28 g for control mesocosms to 11 g and < 1 g for flooded mesocosms without and with light restrictions, respectively). Soil microbial biomass was also reduced (up to a 22-27% of the total PLFAs) and flooding induced shifts in microbial community structure, particularly a loss of soil fungi. The soil fungi content quickly recovered (4 weeks) when light was not restricted during the flood period, however, no such recovery was seen in the darkened treatments. Overall, we conclude that extreme flood events cause rapid and profound changes in soil function. Both the impact of the flooding and the time to recover is exacerbated when light is restricted (e.g. in sediment laden floodwater). In addition, our results suggest that the presence of flood-resilient plants can mitigate against some of the negative impacts of flooding on soil functioning.