Terrestrial ecosystems often vary dramatically in their responses to drought, but the reasons for this are unclear. With climate change forecasts for more frequent and extensive drought in the future, a more complete understanding of the mechanisms that determine differential ecosystem sensitivity to drought is needed. In 2012, the Central US experienced the fourth largest drought in a century, with a regional-scale 40% reduction in growing season precipitation affecting ecosystems ranging from desert grassland to mesic tallgrass prairie. This provided an opportunity to assess ecosystem sensitivity to a drought of common magnitude in six native grasslands. We tested the prediction that drought sensitivity is inversely related to mean annual precipitation (MAP) by quantifying reductions in aboveground net primary production (ANPP). Long-term ANPP data available for each site (mean length = 16 years) were used as a baseline for calculating reductions in ANPP, and drought sensitivity was estimated as the reduction in ANPP per millimeter reduction in precipitation. Arid grasslands were the most sensitive to drought, but drought responses and sensitivity varied by more than twofold among the six grasslands, despite all sites experiencing 40% reductions in growing season precipitation. Although drought sensitivity generally decreased with increasing MAP as predicted, there was evidence that the identity and traits of the dominant species, as well as plant functional diversity, influenced sensitivity. A more comprehensive understanding of the mechanisms leading to differences in drought sensitivity will require multi-site manipulative experiments designed to assess both biotic and abiotic determinants of ecosystem sensitivity.

Intensification of the global hydrological cycle, ranging from larger individual precipitation events to more extreme multi-year droughts, has the potential to cause widespread alterations in ecosystem structure and function. With evidence that the incidence of extreme precipitation years (defined statistically from historical precipitation records) is increasing, there is a clear need to identify ecosystems that are most vulnerable to these changes and understand why some ecosystems are more sensitive to extremes than others. To date, opportunistic studies of naturally occurring extreme precipitation years, combined with results from a relatively small number of experiments, have provided limited mechanistic understanding of differences in ecosystem sensitivity suggesting that new approaches are needed. Coordinated distributed experiments (CDEs) arrayed across multiple ecosystem types and focused on water can enhance our understanding of differential ecosystem sensitivity to precipitation extremes, but there are many design challenges to overcome (e.g., cost, comparability, standardization). Here we evaluate contemporary experimental approaches for manipulating precipitation under field conditions to inform the design of "Drought-Net", a relatively low cost CDE that simulates extreme precipitation years. A common method for imposing both dry and wet years is to alter each ambient precipitation event. We endorse this approach for imposing extreme precipitation years because it simultaneously alters other precipitation characteristics (i.e., event size) consistent with natural precipitation patterns. However, we do not advocate applying identical treatment levels at all sites -a common approach to standardization in CDEs. This is because precipitation variability varies >5-fold globally resulting in a wide range of ecosystemspecific thresholds for defining extreme precipitation years. For CDEs focused on Accepted ArticleThis article is protected by copyright. All rights reserved. precipitation extremes, treatments should be based on each site's past climatic characteristics. This approach, though not often used by ecologists, allows ecological responses to be directly compared across disparate ecosystems and climates, facilitating process-level understanding of ecosystem sensitivity to precipitation extremes.

Plant traits can be used to predict ecosystem responses to environmental change using a response–effect trait framework. To do this, appropriate traits must be identified that explain a species' influence on ecosystem function (“effect traits”) and the response of those species to environmental change (“response traits”). Response traits are often identified and measured along gradients in plant resources, such as water availability; however, precipitation explains very little variation in most plant traits globally. Given the strong relationship between plant traits and ecosystem functions, such as net primary productivity (NPP), and between NPP and precipitation, the lack of correlation between precipitation and plant traits is surprising. We address this issue through a systematic review of >500 published studies that describe plant trait responses to altered water availability. The overarching goal of this review was to identify potential causes for the weak relationship between commonly measured plant traits and water availability so that we may identify more appropriate “response traits.” We attribute weak trait–precipitation relationships to an improper selection of traits (e.g., nonhydraulic traits) and a lack of trait‐based approaches that adjust for trait variation within communities (only 4% of studies measure community‐weighted traits). We then highlight the mechanistic value of hydraulic traits as more appropriate “response traits” with regard to precipitation, which should be included in future community‐scale trait surveys. Trait‐based ecology has the potential to improve predictions of ecosystem responses to predicted changes in precipitation; however, this predictive power depends heavily on the identification of reliable response and effect traits. To this end, trait surveys could be improved by a selection of traits that reflect physiological functions directly related to water availability with traits weighted by species relative abundance. A http://onlinelibrary.wiley.com/doi/10.1111/1365-2435.13135/suppinfo is available for this article.

During the 1930s Dust Bowl drought in the central United States, species with the C3 photosynthetic pathway expanded throughout C4-dominated grasslands. This widespread increase in C3 grasses during a decade of low rainfall and high temperatures is inconsistent with well-known traits of C3 vs. C4 pathways. Indeed, water use efficiency is generally lower, and photosynthesis is more sensitive to high temperatures in C3 than C4 species, consistent with the predominant distribution of C3 grasslands in cooler environments and at higher latitudes globally. We experimentally imposed extreme drought for 4 y in mixed C3/C4 grasslands in Kansas and Wyoming and, similar to Dust Bowl observations, also documented three- to fivefold increases in C3/C4 biomass ratios. To explain these paradoxical responses, we first analyzed long-term climate records to show that under nominal conditions in the central United States, C4 grasses dominate where precipitation and air temperature are strongly related (warmest months are wettest months). In contrast, C3 grasses flourish where precipitation inputs are less strongly coupled to warm temperatures. We then show that during extreme drought years, precipitation–temperature relationships weaken, and the proportion of precipitation falling during cooler months increases. This shift in precipitation seasonality provides a mechanism for C3 grasses to respond positively to multiyear drought, resolving the Dust Bowl paradox. Grasslands are globally important biomes and increasingly vulnerable to direct effects of climate extremes. Our findings highlight how extreme drought can indirectly alter precipitation seasonality and shift ecosystem phenology, affecting function in ways not predictable from key traits of C3 and C4 species.

No abstract