For the grasslands in the Great Plains, the east-west precipitation gradient results in three distinct grassland types: tallgrass prairie from Illinois to Kansas, associated with a region where rainfall amounts exceed evaporation losses, shortgrass steppe in the west limited by rainfall and growing season temperatures, and the mixed-grass prairie in the central potion as a transition between the wetter and drier prairies (Maricle and Adler 2011; Olsen et al. 2013; Maricle et al. 2017). Across all three grassland types, long-term precipitation and temperature is more variable compared to other biomes, often creating yearly and seasonally contrasting growing conditions (Zhang et al. 2010; Knapp et al. 2015; Flanagan et al. 2017). For example, Flanagan et al. (2017) analyzed long-term precipitation and temperature records and found a rise in asynchrony in climate maxima in the Great Plains, which results in a widening disparity between abiotic patterns and plant phenology. Here, our results emphasize the large differences in physiological and anatomical responses that can exist within a widespread C4 grass species (A. gerardii) across multiple years and locations with distinct climate histories (contrasting precipitation and temperature) (Fig. 1). Our data illustrates that patterns of response in this widespread species vary across locations, but perhaps most importantly, that this pattern of variation in response to wet/dry years is not uniform within a single latitude.
A large number of studies have investigated how a dominant species (A. gerardii) responds to changes in precipitation (Knapp 1985; Dietrich and Smith 2016; Hoffman et al. 2018). Such studies have elucidated how local ecotypes are genetically distinct across a precipitation gradient that exists from Kansas and Illinois, using common garden experiments (Mendola et al. 2015; Kramer et al. 2018) or within local landscapes (Avolio and Smith 2013; McAllister et al. 2015). These studies provide an understanding of ecotypic responses under novel climate conditions; but do not present a mechanistic understanding for physiological responses of local populations within and across sites. Common garden experiments often do not include key grassland ecosystem drivers (fire and grazing), which have been repeatedly shown to impact physiological responses, biomass, and local ecosystem function (O’Keefe and Nippert 2017; O’Connor et al. 2020; Connell et al. 2020). Adding to the mechanistic understanding in the existing literature, we investigated how local populations of A. gerardii within their site of origin respond to water availability and grazing in comparison to populations across multiple sites.
Significant differences in leaf level physiology, microanatomy, stoichiometry, and biomass were observed across sites and between years in this study. The long-term climate histories of each location were responsible for shaping functional traits of local populations (Fig. 1), allowing for site-specific leaf-level anatomy and physiology (Fig. 2, 3; Table 1; Table S1, 2) (Hoffman and Smith 2020; Bachle and Nippert 2021). In addition, decreased soil moisture availability reduces carbon assimilation, decreases nutrient uptake, and leads to reduced productivity (Lemoine et al. 2018; Jardine et al. 2021). Our data illustrate similar patterns, at the FHPP and KPBS sites, which received significantly less rainfall in the 2018 growing season than the subsequent year (Fig. 1). The drought conditions at both locations resulted in significantly reduced photosynthetic rates, stomatal conductance, and leaf nitrogen content (Fig. 2; Table 1; Table S1). Increasing water stress decreases stomatal aperture, allowing for reduced water loss. However, long durations of water stress can lead to carbon starvation (Lawson and Matthews 2020; Nunes et al. 2020). Similarly, reductions in XA and increased BA were also observed in 2018 (Fig. 3), reflecting changes in water-use strategies. Previous research showed that increased XA allows for greater water transport, but it also increases the likelihood of cavitation during droughts or when the water column is under high tension (Olson et al. 2020).
Intraspecific trait variability (CV) was statistically different between years, but relatively similar across locations (Fig. 5). The greatest variation was reported for gas exchange measurements (An, gs, E) in 2018, which were ~ 2 times higher than the following year (at both FHPP and KPBS) (Fig. 5A). While high variability may be inherent to the instantaneous nature of gas exchange measurements, the CV of physiological responses in 2019 was similar to all microanatomical traits regardless of function (Fig. 5B, C). This decrease in physiological CV may indicate a baseline physiology and associated physiological plasticity of A. gerardii, when water is less limiting. While mean microanatomical traits varied significantly between 2018 and 2019, there was little change in variability (CV) across years (Fig. 5B, C). In fact, most microanatomical variation resulted from water-specific traits (XA, t/b, BA) (Fig. 5C). The diversity in functional trait responses has been reported to protect individuals and populations from detrimental effects of drought (Mori et al. 2013; Kreyling et al. 2017; Roberts et al. 2019).
While previous research has indicated that microanatomical traits can influence/constrain physiological responses to changes in water availability (Christin et al. 2013; Guha et al. 2018; Edson-Chaves and Graciano-Ribeiro 2018; Wargowsky et al. 2021), few studies have analyzed physiology, stoichiometry, and microanatomy on the same leaf. The importance of this sampling technique allowed us to analyze direct bivariate relationships of both functional trait mean and variability (CV) (Fig. 4; Figure S1, 2). Past research focusing on anatomical and physiological relationships has been mainly constrained to greenhouses or single-year studies (Henry et al. 2012; Bachle and Nippert 2018; Sonawane et al. 2021). These results emphasize how disparate climates across years can result in dissimilar relationships among traits and between traits and climate variables (Fig. 4; Fig. 6; Figure S1, 2). A. gerardii photosynthetic rates correlated positively with increasing leaf nitrogen content (Fig. 4A) when analyzed between years. However, this seemingly tight relationship broke down when analyzing each year and treatment separately (Fig. 4B). Several mean trait relationships in physiological and microanatomical traits displayed opposing trends between 2018 and 2019 (Figure S1, 2), including BSA against gas exchange traits (An, gs, and E). In addition, the timing of precipitation has also been known to impact grassland productivity (Nippert et al. 2006; Craine et al. 2012), which is a result of altered microanatomy and physiology (Fay et al. 2002; Wang et al. 2016; Lemoine et al. 2018). For example, early season rainfall (coinciding with tissue development) allows for the production of larger vessel areas for greater transport potentials, while early season droughts constrain development, which results in smaller vessel areas (Mauseth 1988).
Historically, the Great Plains have provided forage for native mammalian grazers such as Bison bison (bison), and grazing resulted in increased plant diversity and landscape heterogeneity (Knapp et al. 1999; Elson and Hartnett 2017). More recently however, the majority of grazing is accomplished by non-native grazers like cattle. Similar to climate variability and fire, responses to grazing are typically examined at the community or ecosystem levels, while less is understood about the physiological and microanatomical mechanisms responsible for those responses (O’Keefe and Nippert 2017). However, grazing and other forms of herbivory can increase gas exchange rates in order to compensate for the loss of tissue (Pinkard et al. 2011; O’Connor et al. 2020). While this allows for greater carbon assimilation, it requires increased stomatal conductance which inherently leads to greater water loss (Bertolino et al. 2019). During drought conditions, this compensatory response of recently grazed tissues would negatively impact grass physiology, thereby decreasing carbon assimilation and future productivity (Feller 2016; Souther et al. 2020). However, gas exchange rates within grazed locations in this study were nearly identical to the control (Table 1; Fig. 2), even during the dry 2018 growing season. In addition, only three functional traits were impacted by the grazing treatment: MSA, C:N ratios, and biomass production (Table 1). The grazing treatment at KPBS was responsible for most MSA variation, in both 2018 and 2019 (Table S2). In 2018, grazing increased C:N ratios in leaf tissues from FHPP and PRP (Table 1; Table S1). While grazing did impact functional trait variability, it was only observed during the 2018 growing season and only in physiological and water-use microanatomical trait CV (Fig. 5). The lack of treatment response may be due to several factors including: 1) stocking rates at each location may not be conducive to reflect substantive grazing pressure; 2) the experimental design may not have adequately covered/represented each site and subsequent treatment; 3) due to the evolutionary history of A. gerardii in the Great Plains, a heightened grazing intensity may be necessary to induce alternative physiological responses.
These results highlight how trait plasticity can serve as an important tool for understanding the anatomical and physiological mechanisms that facilitate wide distributions of a dominant grass species. This research was completed during the 2018 and 2019 growing seasons which had significantly different water availability among years. Drought conditions in 2018 resulted in decreased gas exchange rates and subsequent biomass production, irrespective of grazing. However, increased water availability in 2019 facilitated high gas exchange rates and the doubling of aboveground biomass. In addition, there was significant variation in microanatomical traits across locations and between sampling years. Together, these results indicate that there are specific leaf construction strategies based on intra-annual climate conditions across the Great Plains. Such leaf construction strategies frame instantaneous physiological responses to climate variability, and also other grassland drivers (i.e., grazing and fire). Results from this study underlie the importance of collecting multiple years of data from native species in natural environments. Our data also emphasizes the need for increased microanatomical research, as we clearly demonstrate site and climate-specific leaf construction strategies are important for understanding and contextualizing physiological responses in a dominant grass species.