Leaf Economics traits in relation to crop domestication syndromes
Studies reporting differences in the LES traits relationships in crops vs. wild plants have supported inferences and hypotheses surrounding the unintended consequences of artificial selection (e.g. Milla et al., 2015; Roucou et al., 2018). Perhaps most consistent with hypotheses related to artificial selection is our find that in comparison to wild plants, ‘Chardonnay’ expressed a steeper increase inA mass and R mass per unit increase in leaf N. Specifically, based on our SMA models fits (Table S2), across the range of leaf N values in ‘Chardonnay’ observed here (1.9-2.9%), predicted A mass increased by ~84.3% (from 0.06-0.38 μ mol CO2g-1 s-1) andR mass increases by ~ 71.8% (from 0.008 to 0.028 μ mol CO2 g-1s-1). Comparatively, in wild plants from the GLOPNET dataset this same increase in leaf N from 1.9 to 2.9% corresponds to only a 47.0% predicted increase in A mass (from 0.138 to 0.264 μ mol CO2 g-1s-1) and 42.2% increase inR mass (from 0.028 to 0.048 μ mol CO2 g-1 s-1). Higher photosynthetic rates for a given value or increase in leaf N concentrations have been similarly detected in rice (Xiong & Flexas, 2018), and may reflect conscious or unconscious artificial selection for more rapid growth responses to N availability in crops vs. wild plants. However, this is not universal among crops. Certain crops, namely coffee, show significantly lower increases inA mass with greater leaf N (Martin & Isaac, 2021; Martin et al., 2017), while others including soy expressA mass-leaf N relationships that are statistically indistinguishable from those in wild plants (Hayes et al., 2019). In sum, the growing literature to which we contribute with our study indicates that LES trait relationships are a unique and idiosyncratic feature of crop domestication syndromes.
In this regard, a novel contribution from our work here is the integration of R into studies evaluating intraspecific or intragenotypic LES in crops. Specifically, previous studies evaluating crop trait (co-)variation in comparison to non-domesticated wild plants have not included R in their analyses (Hayes et al., 2019; Martin et al., 2017; Milla, Morente-López, Alonso-Rodrigo, Martín-Robles, & Stuart Chapin III, 2014; Roucou et al., 2018; Xiong & Flexas, 2018), despite this trait representing a key trade-off along the LES (P.B. Reich et al., 1998; I. J. Wright et al., 2004). The LES trait relationships in ‘Chardonnay’ that included R masswere qualitatively unique, in that none of these bivariate datasets and SMA models intersected the global LES defined by wild plants (Figure 2C, D, and E). Instead, at a given value of A mass, LMA, or leaf N, in nearly all of the leaves measured here (i.e., 43 or 45 leaves), ‘Chardonnay’ R mass was consistently lower than average vs. R mass in wild plants. This indicates that domestication has favoured vines that express leaves with a low rate of C loss at a given rate of structural or chemical investment in C assimilation.
These results have two possible explanations: 1) even the lowest bulk density/compaction values at our study site still restrict physiological functioning; and/or 2) lower R mass for a given value of A mass, leaf N, or LMA is a signature of domestication in Vitis vinifera varieties. Since the primary targets of grape domestication are related to yield, quality, growth form, and harvestability (Keller, 2020), our findings point to an unintended consequence of domestication related to plant C economy. Expanding our work across a wider range of ‘Chardonnay’ growing sites (particularly where bulk density is lower) and grape varieties is therefore central in testing either proposed explanation.
One unexpected finding in our analysis here, were patterns of leaf C variation. Although not a primary focus of our analysis, since it is not considered a primary trait forming the LES (I. J. Wright et al., 2004), we found that this trait covaried in an unexpected pattern along the intragenotype LES in ‘Chardonnay’. Specifically, we detected a statistically significant positive relationship between leaf C andA mass, R mass, and leaf N, and a significant negative relationship between leaf C and LMA (Table S5). Furthermore, when incorporated into an additional PCA, leaf C covaried across the first PCA axis positively withA mass, leaf N, and negatively with LMA (Table S6). Therefore in our dataset, leaf C covaries along LES traits such that higher leaf C values reflect a resource acquisitive trait syndrome. This finding is counter to studies of certain other domesticated plants where leaf C is by in large positively related to leaf construction costs, leaf dry matter content, and LMA (Gagliardi et al., 2015; Martin et al., 2017). In ‘Chardonnay’, coordination of leaf C along an intragenotype LES likely owes to the selection for C loading in leaves and plants in the form of sugars and starches (Keller, 2020).