Ecological impacts of two years of extreme drought in a mesic grassland: Rapid recovery despite an extreme response

Background/Question/Methods

Climate extremes, such as drought and heat waves, are expected to increase in their frequency and intensity over the next century.  We examined the response of a mesic tallgrass prairie ecosystem to two years of experimentally imposed seasonal drought and a short-term heat wave, followed by a recovery year. During 2010 and 2011, we reduced precipitation inputs to 33% of ambient rainfall (drought) to compare responses to a well-watered treatment (ambient rainfall plus supplemental irrigation).  The drought treatment resulted in a 50% reduction in soil water content in both years.  Under these contrasting soil moisture regimes we imposed a two-week mid-summer heat wave at four temperature levels, ranging from 0 to +11 degrees C above ambient.  In 2012, all plots received ambient rainfall plus supplemental irrigation to ensure that long-term average precipitation inputs were received.  This design allowed us to examine the individual and combined effects of drought and heat on ecosystem structure (plant community) and function (aboveground net primary production, ANPP).  We used local long-term historical records of both weather (112 years, Manhattan, KS) and ANPP (27 years, Konza Prairie LTER) to determine whether the imposed treatments and ecological responses were statistically extreme.

Results/Conclusions

The experientially imposed seasonal drought and heat waves were well outside the bounds of normal variability and comparable in magnitude to the most severe years of the 1930’s Dust Bowl.  While we measured no significant direct or combined effects of the imposed heat waves at the ecosystem or community levels, there were significant effects of drought.  Total ANPP was significantly decreased in both drought years, and particularly during the second year of the drought, which was below the 5^th percentile of the long-term LTER record for the site. Despite this extreme ecosystem-level response, we observed full recovery in ANPP in the first year following drought. This occurred even with a substantial drought-induced shift in community composition during the post-drought year, caused by a reordering in the rank abundances of the four dominant species.  In summary, two years of extreme drought led to a decoupling of grassland structure and function; during the drought we observed an extreme reduction in productivity with no change in community structure, while after the event, there was a full recovery in production despite a significant change in community structure. Such changes in community structure could have important consequences for stability in ecosystem function over the long-term.