In the mesic tallgrass prairies of the Great Plains, US, two C4 grass species, Andropogon gerardii Vitman and Sorghastrum nutans (L.) Nash often coexist at sufficiently high canopy cover, population densities, and/or occurrence frequencies to qualify (sensu Gray et al., 2021) as codominant species (Brown, 1985; Duralia and Reader, 1993; Freeman, 1998; Hartnett et al., 1996; Smith and Knapp, 2003). As competitive species, these abundances typically result in high community aboveground net primary productivity (ANPP) (Grime, 1998; Pierre et al., 2011; Smith and Knapp, 2003), making them important for grazing (Vinton et al., 1993), carbon sequestration (Grime, 1998; Kemp et al., 1994; Mahaney et al., 2008; Omonode and Vyn, 2006), nutrient cycling (Grime, 1998; Mahaney et al., 2008), and invasion resistance (Smith et al., 2004). Curiously, A. gerardii is reported to be both more competitive (Silletti et al., 2004; Tilman and Wedin, 1991) and more drought tolerant (Hoffman et al., 2018; Silletti and Knapp, 2002; Swemmer et al., 2006) in the regions where they maintain stable codominance (Brown, 1985; Duralia and Reader, 1993; Freeman, 1998; Hartnett et al., 1996; Smith and Knapp, 2003).
Given the competitive advantages in both wet and dry years this attributes to A. gerardii, it is not clear how S. nutans maintains its codominant status. This uncertainty is further shrouded by their many morphological and physiological similarities that make traditional niche-based explanations less plausible. For instance, these species share several important life history traits, including long lives (Gustafson et al., 2005; Keeler, 2004; Lauenroth and Adler, 2008; USDA, 2021a, 2021b), C4 photosynthetic pathways, and rhizomatous cloning (Benson and Hartnett, 2006; Lauenroth and Adler, 2008; McKone et al., 1998; USDA, 2021a, 2021b). They also bear similarities in their functional traits (e.g., leaf dimensions, leaf gas exchange rates, ANPP) (Forrestel et al., 2015, 2014; Nippert et al., 2009) and responses to fire and grazing disturbances (Bowles et al., 2011; Forrestel et al., 2015, 2014; Hadley and Kieckhefer, 1963; Polley et al., 1992; Towne and Kemp, 2003; Weaver, 1931; Weaver and Fitzpatrick, 1932). Both are competitive in low-nitrogen environments (Berg, 1995; Lett and Knapp, 2003; Mulkey et al., 2008; Silletti and Knapp, 2001), and are tall in stature (Knapp and Hulbert, 1986; Weaver, 1931), but are intolerant of shading (Lett and Knapp, 2003; Weaver and Rowland, 1952) and persistent grazing (Damhoureyeh and Hartnett, 2002; Hartnett et al., 1996; Vinton et al., 1993). Taken together, it is not obvious where niche differences substantial enough to prevent competitive exclusion exist.
One alternative explanation for their codominance, the stress gradient hypothesis (SGH), proposes that as the intensity of environmental stress increases, the net interactions between competing species becomes less negative. This shift occurs as some negative interactions are diminished in their effects, and/or some positive interactions are enhanced, resulting in the presence of otherwise deleterious neighbors becoming beneficial for survival, growth, and/or reproduction relative to their absence (Bertness and Callaway, 1994; Brooker and Callaghan, 1998; Callaway and Walker, 1997; Olofsson et al., 1999; Ploughe et al., 2019).
An annual shift from net-negative to net-positive plant interactions has previously been supported as a possible mechanism of codominance stabilization in a subalpine meadow ecosystem characterized by competitive conditions in early growing seasons, and stressfully dry conditions in late seasons (Kikvidze et al., 2006). The mesic tallgrass prairies where A. gerardii and S. nutans are codominant are similarly characterized by frequent late-season reductions in rainfall (Craine et al., 2012; Hayden, 1998; Knight et al., 1994) and soil moisture (Knapp et al., 2002). If S. nutans benefits from the presence of A. gerardii neighbors (relative to intra-specific neighbors) in the drier months of the growing season (e.g., through the more water-conservative behavior in A. gerardii, see: Connor and Hawkes, 2018; Heisler-White et al., 2009; Silletti and Knapp, 2001; Swemmer et al., 2006), the presence of A. gerardii neighbors may provide a reduction in competitive intensity and a temporary increase in the fitness of S. nutans compared to those in monocultures, reducing the probability of competitive exclusion under stressful conditions and promoting more integrated communities.
To test whether stressful conditions induced by late-season drought can shift the net interaction between A. gerardii and S. nutans to one that is more facilitative, we performed a greenhouse experiment using artificial communities with both species present at different densities. We compared three per capita performance metrics (rates of reproduction, tiller survival, and aboveground biomass productivity) across contrasting community structures (monoculture or mixed) and water stress treatments. First, to confirm that water limitation is a form of abiotic stress that reduces per capita performance, we tested the hypothesis that late-season reductions in water availability would reduce performance in both low- and high-density monocultures of both species. Second, to account for the effects of density per se (regardless of neighbor species identity), we compared performances of low-density monocultures to those at high densities in both low- and high-stress conditions. Lastly, in accordance with the SGH, we tested whether interspecific neighbors would alleviate a portion of the negative effects of water limitation relative to monocultures at a given initial total community density, and whether increasing density of interspecific neighbors negates or bolsters that effect. If so, this would indicate that codominance between these species can be aided by shifts toward more positive interactions between them under periodically stressful abiotic conditions.