Introduction
C4 photosynthesis has evolved independently in multiple grass lineages (Grass Phylogeny workshop 2012) thus leading to remarkable structural, anatomical and physiological trait diversity (Christin et al. , 2013). Studies suggest that this trait diversity among the C4 species could be attributed to their adaptation to different environmental variables like temperature, fire frequency and precipitation (Edwards & Smith, 2010; Visseret al. , 2012; Zhou et al. , 2018). In general, C4 species mostly occupy the lower latitudes where light availability and temperature likely do not strongly limit photosynthesis and growth (Pearcy & Ehleringer, 1984). Instead, precipitation may be an important factor affecting trait diversity in C4species; particularly, in traits associated with photosynthetic C-gain and transpirational water-loss (Edwards & Still, 2008; Osborne & Sack, 2012; Zhou et al. , 2018). During the adaption to habitats with low water availability, a fundamental challenge for plants will be to maintain photosynthetic C-gain while minimizing transpirational water-loss associated with high evaporative demand. This tradeoff could be achieved partly through coordination of leaf-level photosynthetic and hydraulic traits (Brodribb et al. , 2007; Nardini & Luglio, 2014; de Boer et al. , 2016). However, the extent of variation and coordination among these traits, particularly those associated with internal CO2-diffusion conductance (gm) and leaf hydraulic conductance (Kleaf), has not been well studied in C4 species adapted to habitats with varying water availabilities (Osborne & Sack, 2012; Liu & Osborne, 2015; Taylor et al. , 2018). Although this type of trait variation and coordination has been studied in C3 plants there could be significant differences in C4 plants due to their unique anatomy and physiology (Kocacinar & Sage, 2003; Osborne & Sack, 2012; Ocheltree et al. , 2016; Zhou et al. , 2018).
During adaptation to drier habitats, species can exhibit several leaf-level structural and anatomical traits that can help maximize photosynthetic C-gain at a given water-loss (Wright et al. , 2001; Galmés et al. , 2012; Ivanova et al. , 2018b). For instance, presence of greater stomatal densities and amphistomaty (stomata on both leaf surfaces) in drier habitats, is beneficial as it reduces the role of boundary layer conductance in constraining leaf gas exchange, helps optimize leaf interior water status for CO2 transport by reducing temperature gradients, and helps reduce the effective leaf thickness by decreasing the CO2-diffusion pathlength (Galmés et al. , 2012; Muir, 2018; Drake et al. , 2019; Muir, 2019; Pathare et al. , 2020). Greater stomatal densities and smaller stomatal size in drier habitats may increase theoretical maximum stomatal conductance (gmax) which in turn could help plants maximize C-gain particularly during intermittent periods of water availability (Hetherington & Woodward, 2003; Franks & Beerling, 2009). Furthermore, mesophyll traits like Smes and Sc - the parameters that characterize exchange surfaces for CO2, negatively correlated with water availability in the C3 species of European steppe plant communities and were suggested as indicators of increasingly drought adapted steppe plants (Ivanova et al. , 2018a; Ivanova et al. , 2018b). These structural and anatomical adaptations could help maximize internal CO2-diffusion conductance (gm), at a given stomatal conductance (gsw), thus leading to greater photosynthetic rates (Anet) as well as leaf-level water-use efficiency (WUE) in species adapted to drier habitats (Flexas et al. , 2008; Flexas et al. , 2013; Ivanova et al. , 2018a; Ivanova et al. , 2018b). However, very few studies, mostly based on C3 species, have investigated the leaf-level structural and anatomical traits associated with gm that could be a characteristic of plant adaptation to drier habitats (Ivanova et al. , 2018a; Ivanovaet al. , 2018b). Alternatively, even though C4species can successfully occupy drier and warmer habitats and form grasslands over vast areas globally, there is a little information about leaf-level structural and anatomical traits that influence photosynthetic C-gain and water-use in these species. Specifically, we are unaware of any studies that have investigated the relationship of gm and associated anatomical traits in diverse C4 species from habitats with different water availability. In a previous study (Pathare et al. , 2020) we investigated the structural and anatomical determinants of gm in diverse C4 grasses and found that, leaf thickness, adaxial stomatal densities (SDada), stomatal ratio (SR) and Smes had a positive effect on gm. In the current study, our aim is to determine if the variation in above traits among the C4 species could be related to adaptation to habitats with different water availabilities. Our first hypothesis (H1) is that, C4 grasses adapted to lower MAP will show leaf anatomical traits associated with greater gm in order to maximize photosynthetic C-gain.
Though we hypothesized a greater gm in C4 grasses adapted to low MAP (H1), one would expect an increase in water cost relative to photosynthetic C-gain, because gm and associated traits have been shown to scale positively with leaf hydraulic conductance (Kleaf) in C3 species (Flexas et al. , 2013; Xiong et al. , 2015; Xiong et al. , 2017; Drake et al. , 2019). Kleaf is an important trait associated with leaf water transport and represents the conductance to flow of water from the leaf petiole through the xylem, then through the bundle sheath and finally through the mesophyll to the site of evaporation (Sack & Holbrook, 2006; Noblin et al. , 2008; Buckley, 2015; Buckley et al. , 2015). Though Kleaf is partitioned between the xylem (Kx) and the outside xylem pathways (Kox), changes to Kox are expected to have the largest effects on Kleaf (Buckley et al. , 2015; Scoffoni et al. , 2017; Xiong & Nadal, 2019). Consequently, leaf-level anatomical traits that influence Kox such as leaf thickness, vein-to-epidermis distance (VED), vein length per unit of leaf area (VLA) and bundle sheath and mesophyll traits are expected to have a significant effect on Kleaf (Griffiths et al. , 2013; Sack et al. , 2013; Buckley et al. , 2015). For instance, greater leaf thickness and VED, if associated with low VLA, may increase the length of post-venous water path thus leading to lower Kleaf. Alternatively, greater VLA in thinner leaves may increase Kleaf by providing additional parallel flow paths through the vein system and decreasing the horizontal pathlength from veins to sites of evaporation (Brodribb et al. , 2007; Sack & Scoffoni, 2013; Buckley et al. , 2015; Drake et al. , 2019). Furthermore, greater bundle sheath (BS) surface area ratio, lower BS cell wall thickness (BSCW) and BS exposed to intercellular air spaces (BSias) and greater Smes may all enhance Kleaf (Buckleyet al. , 2015; Caringella et al. , 2015; Xiong et al. , 2017; Scoffoni et al. , 2018). Investigating the coordination of above traits related to water-use with traits related to C-gain will provide insights into the water cost associated with photosynthetic C-gain in C4 species adapted to habitats with varying water availabilities.
Previous studies on C3 species have shown a strong positive linkage of Kleaf with gsw and hence Anet (Brodribb et al. , 2007; Brodribb & Feild, 2010; Flexas et al. , 2013; Scoffoni et al. , 2016). Additionally, the few studies that address the coordination of Kleaf and gm show that these two traits scale positively with each other in C3 species as they share some structural and anatomical traits that form the mechanistic basis for their coordination independent of gsw (Flexaset al. , 2013; Xiong et al. , 2015; Xiong et al. , 2017) but see (Théroux-Rancourt et al. , 2014; Loucos et al. , 2017; Wang et al. , 2018). For example, Smespositively correlates with both gm and Kleaf (Flexas et al. , 2013; Xiong et al. , 2015; Xiong et al. , 2017) , since greater Smesincreases the number of parallel pathways for CO2-diffusion inside mesophyll cells (Evans et al. , 2009) as well as the evaporating surface area for water thus increasing gm and Kleaf respectively (Sack & Scoffoni, 2013; Xiong et al. , 2017). In summary, the positive correlation of Kleaf with gmimplies a greater water cost associated with greater C-gain, which could be detrimental in drier conditions where using water efficiently will be important for plant growth and fitness. Hence, a safer strategy for plants is to maintain lower Kleaf in drier conditions at the cost of Anet and growth rates (Sinclair et al. , 2008; Nardini & Luglio, 2014; Scoffoni et al. , 2016). However, these generalizations are mostly based on studies of C3 species. It is unclear if these results would apply to C4 grasses that are adapted to relatively drier habitats and may show different coordination between the traits associated with photosynthetic C-gain and transpirational water loss (Kocacinar & Sage, 2003; Ocheltree et al. , 2016; Zhou et al. , 2018). Increased bundle sheath size and vein densities are the anatomical precursors for evolution of C4 species from their C3 ancestors that led to higher Kleaf in the C4 species compared to C3 species. (Osborne & Sack, 2012; Christin et al. , 2013; Griffiths et al. , 2013). At the same time, evolution of carbon concentrating mechanism in C4 species allows maintenance of lower gs and higher leaf water potential. Hence, it has been proposed that once the C4 species evolved, subsequent selection for traits leading to greater Kleaf would be lessened particularly during adaptation to drier habitats and there could be a decoupling between Kleaf and Anet within the C4 lineages (Zhou et al. , 2018). Consequently, in contrast to C3 species, maintaining greater Kleaf in order to achieve higher Anetmay not be necessary in C4 species and Kleaf may be uncoupled from gsw, Anet (Kocacinar & Sage, 2003; Ocheltree et al. , 2016) and potentially gm. However, to our knowledge, there have been no previous studies on the correlation of gm with Kleaf or traits associated with Kleaf in C4 species adapted to habitats with diverse MAP. Building knowledge upon previous evidences, we hypothesized (H2) that C4 species adapted to habitats with low MAP will show traits associated with lower Kleaf that will maximize photosynthetic C-gain at a given water loss.
To test the above hypotheses, we selected 18 C4 grasses that varied significantly in structural and anatomical traits (Pathareet al. , 2020). The grasses were grown under common growth conditions and abundant water and nutrient supply which avoids the influence of environmental conditions on traits and thus helps identify the differences that could be a result of species adaptation to their habitat of evolution or common occurrence (Reich et al. , 2003). We measured important leaf-level structural and anatomical traits associated with photosynthetic C-gain and gm and transpirational water-loss and Kleaf in 18 diverse C4 grasses. There is a significant knowledge gap about how C4-gm variability relates with habitat climate variables like MAP largely because of the lack of techniques to estimate C4-gm in field as well as laboratory conditions. However, the recent developments provide the opportunity to estimate C4-gm under laboratory conditions and thus investigate the relationship of habitat climate variables with gm in diverse C4species. Here, we use a recently developed method, based on modeling of leaf oxygen isotope discrimination during photosynthesis, to estimate gm in 18 diverse C4 grasses (Barbouret al. , 2016; Ubierna et al. , 2017; Ogee et al. , 2018). Kleaf was estimated from anatomical traits like leaf thickness, vein-to-epidermis distance (VED) and vein-to-vein distance (IVD) as recently described by de Boer et al ., (2016) using the semi-empirical model of Brodribb et al ., (2007).