Intensifying drought eliminates the expected benefits of elevated carbon dioxide for soybean Sharon B. Gray 1† , Orla Dermody 1 , Stephanie P. Klein 1† , Anna M. Locke 1† , Justin M. McGrath 1 , Rachel E. Paul 1 , David M. Rosenthal 1† , Ursula M. Ruiz-Vera 1 , Matthew H. Siebers 1† , Reid Strellner 1 , Elizabeth A. Ainsworth 1,2 , Carl J. Bernacchi 1,2 , Stephen P. Long 1 , Donald R. Ort 1,2 and Andrew D. B. Leakey 1 * Stimulation of C 3 crop yield by rising concentrations of atmospheric carbon dioxide ([CO 2 ]) is widely expected to counteract crop losses that are due to greater drought this century. But these expectations come from sparse field trials that have been biased towards mesic growth conditions. This eight-year study used precipitation manipulation and year- to-year variation in weather conditions at a unique open-air field facility to show that the stimulation of soybean yield by elevated [CO 2 ] diminished to zero as drought intensified. Contrary to the prevalent expectation in the literature, rising [CO 2 ] did not counteract the effect of strong drought on photosynthesis and yield because elevated [CO 2 ] interacted with drought to modify stomatal function and canopy energy balance. This new insight from field experimentation under hot and dry conditions, which will become increasingly prevalent in the coming decades, highlights the likelihood of negative impacts from interacting global change factors on a key global commodity crop in its primary region of production. R ising [CO 2 ] this century is predicted to stimulate the yield of C 3 crops, counteracting the negative impacts of greater drought on future food production 1–3 . The mechanisms most commonly cited to explain greater yield under elevated [CO 2 ] are (1) direct stimulation of photosynthetic CO 2 uptake and, thereby, biomass accumulation and yield; and (2) reduced stomatal conductance ( g s ) driving lower crop water use and conserving soil moisture, which ameliorates yield loss to drought stress when it occurs 4–8 . Consequently, the magnitude of relative yield stimulation by elevated [CO 2 ] is frequently predicted to increase as drought intensifies 8–10 . Soybean (Glycine max Merr.) is the most important oil and protein seed crop globally 11 . Soybean has also been investi- gated widely as a model for understanding the response of C 3 species to global change 12 . However, as with many species, exper- imental testing of CO 2 response in the field has occurred over a limited number of locations and growing seasons, which limits the inference space of the previously published literature to con- ditions with little to no drought stress 13 . In addition, theory predicts that reduced g s at elevated [CO 2 ] might lower canopy water use in crops with short and dense canopies, like soybean, less than in other vegetation types because of weaker coupling to the bulk atmosphere 14,15 . Given that projected crop yields and food security for the latter part of this century are highly sensitive to the magni- tude of CO 2 fertilization effects 1–3 , it is important to address the uncertainty about how soybean responds to elevated [CO 2 ] under the stronger drought stress that is predicted to characterize future growing conditions. Experimental design This study took a two-pronged approach to determine the interaction of drought and elevated [CO 2 ] on the productivity of field-grown soybean. (1) Year-to-year variation in the effect of elevated [CO 2 ] on soil water content and yield was assessed over eight growing seasons in relation to natural variation in drought stress and canopy properties, including leaf area and temperature. A commer- cial soybean cultivar (Pioneer 93B15) was grown at the Soybean Free-Air CO 2 Enrichment (SoyFACE) facility in a replicated exper- iment (N = 4) at ambient [CO 2 ] (376–392 ppm) and elevated [CO 2 ] (550–585 ppm) from 2004–2011. These ranges represent increasing ambient CO 2 over the eight-year experiment and corresponding changes in the target [CO 2 ] for fumigation to maintain consistent treatment. The mean growing season temperatures varied (19.1– 23.2 °C) and total growing season precipitation varied (274–470 mm) (Supplementary Fig. 1). An additional six soybean cultivars (Dwight, HS93-4118, IA3010, LN97-15076, Loda and Pana) were also assessed in this field facility for yield response to elevated [CO 2 ] from 2004–2008. (2) A rainfall exclusion treatment was used to manipulate water availability in combination with CO 2 treatments over three growing seasons from 2009 to 2011 (Fig. 1a). Productivity responses of Pioneer 93B15 in the rainfall exclusion experiment were assessed in relation to soil moisture and rooting dynamics, root-to-shoot signalling and leaf photosyn- thetic gas exchange. SoyFACE is located in the Midwestern United States, where more than 80% of the national and more than 25% of the global soybean crop is grown 16 . This location allows experimental exposure of soybean to climate change treat- ments in a setting that is directly relevant to agricultural production. Over 90% of soybean production in the United States is rain fed 17 . As such, this crop is susceptible to year-to-year variation in drought stress. FACE fumigation was applied using the method of Miglietta et al. 18 , which directly releases CO 2 into the wind stream and does not cause the atmospheric turbulence that has been detected in other experiments where air blowers are used to distri- bute CO 2 -enriched air. Furthermore, the use of FACE technology 1 Department of Plant Biology and Institute for Genomic Biology, University of Illinois at Urbana-Champaign, Champaign, Illinois 61801, USA. 2 United States Department of Agriculture, Agricultural Research Service, Urbana, Illinois 61801, USA. † Present addresses: Department of Plant Biology, University of California, Davis, California 95616, USA (S.B.G.). CSIRO Plant Industry, Urrbrae, South Australia 5064, Australia (M.H.S.). United States Department of Agriculture, Agricultural Research Service, Raleigh, North Carolina 27695, USA (A.M.L.). Department of Environmental and Plant Biology, Ohio University, Athens, Ohio 45701, USA (D.M.R.). Department of Plant Science, Penn State University, State College, Pennsylvania 16802, USA (S.P.K.). *e-mail: leakey@illinois.edu ARTICLES PUBLISHED: 5 SEPTEMBER 2016 | ARTICLE NUMBER: 16132 | DOI: 10.1038/NPLANTS.2016.132 NATURE PLANTS | VOL 2 | SEPTEMBER 2016 | www.nature.com/natureplants 1 © 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.