Multivariate trait syndromes in ‘Chardonnay’ in relation to soil bulk density
The seven traits incorporated into our multivariate analysis covaried along two primary PCA axes, which accounted for 46.4% and 21.5% of variation in ‘Chardonnay’ physiological (A mass,R mass, Φ , LLCP), morphological (leaf area, LMA), and chemical (leaf N) traits (Figure 2). The first PCA axis represented ‘Chardonnay’ leaf trait covariation and trade-offs consistent with an intragenotype LES. Specifically, PCA axis 1 was most strongly and positively related to A mass, leaf N, leaf Φ , and R mass (r =0.644-0.888 and p <0.001 in all cases), all of which traded-off with LMA (r =-0.615, p <0.001; Figure 2, Table S3). Therefore, ‘Chardonnay’ leaves expressing a higher PCA axis 1 score were associated with “resource acquiring” LES trait syndromes, while “resource conserving” LES trait syndromes characterized ‘Chardonnay’ leaves with lower PCA axis 1 scores (Figure 2). The second PCA axis reflected the covariation of traits associated with resource investment in light interception, including leaf area (r =0.431,p <0.001) and LMA (r =0.69,p <0.001), and leaf-level light requirements (LLCP,r =0.767, p <0.001; Figure 2, Table S3).
Planting row identity explained 39.6% of the variation in multivariate leaf trait syndromes in ‘Chardonnay’ (PerMANOVAp <0.001, Table S4), with the rows of lowest bulk density being concentrated and differentiated (in terms of their 95% confidence ellipses) at the resource-acquiring end of PCA 1 axis (Figure 2). Planting rows with the highest mean bulk density (i.e., Row 5) did tend to be concentrated at the resource-conserving end PCA axis 1 (Figure 2). However, differentiation of leaves sampled in planting rows 2-5, where bulk density did increase albeit not significantly, was weaker along PCA axis 1. Soil bulk density, as represented categorically by planting row identity, did not influence ‘Chardonnay’ leaf position along PCA axis 2 with rows showing overlapping 95% confidence ellipses along this axis (Figure 2). Neither individual plant identity, nor its interaction with planting row, influenced the position of ‘Chardonnay’ leaves in multivariate trait space (PerMANOVA p =0.788 and 0.301, respectively, Table S4).