6.3 Diagnosing leaf water status and leaf hydraulic strategy
Seasonal midday ΨL values were variable across species and stands, but in general, Q. alba experienced a lower overall midday ΨL and a broader range. Larger species-specific declines in midday ΨL occurred in the more arid forest stands (e.g., NC_E chronosequence and MO), while the smallest occurred in NC 15yo (Fig. 7a). Leaf hydraulic strategy was primarily determined by species (Fndf,ddf = 22.20, p = <0.001), and no influence of age on mean ΨLinterquartile ranges were detected (age and age-species interactions NS). The L. tulipifera and A. saccharum mean ΨL interquartile ranges were indistinguishable, but significantly lower than Q. alba (Fig. 7b); mean ΨL interquartile ranges across stands were 0.709 MPa (SE = 0.040), 0.358 MPa (SE = 0.022), and 0.372 MPa (SE = 0.098) forQ. alba, L. tulipifera, and A. saccharum , respectively. Overall, Q. alba displayed more anisohydric behavior whileL. tulipifera and A. saccharum were more isohydric.
6.4 Relationship between ΨL regulation and vulnerability to hydraulic failure
The most anisohydric species (Q. alba ) in our study possessed xylem that was more vulnerable to embolism than the more isohydric species. This pattern was consistent at both P12 and P50, where Q. alba mean embolism thresholds were significantly greater (p = <0.001 for P12 and P50, respectively) than the more isohydricL. tulipifera and A. saccharum (Fig 8). Mean Ψsafety,P12 values were −1.77 MPa (SE = 0.166), −0.35 MPa (SE = 0.078), and 0.63 MPa (SE = 0.319), and mean Ψsafety,P50 values were −0.147 MPa (SE = 0.196), 1.94 MPa (SE = 0.143), and 2.634 MPa (SE = 0.351) for Q. alba ,L. tulipifera , and A. saccharum , respectively. Moreover, the degree of isohydricity (as defined by seasonal ΨLinterquartile range) was strongly associated with risk of xylem dysfunction such that ~60% of the variability in Ψsafety across stands was explained by a tree’s relative ΨL control (Fig. 9).
Across all stands, Ψsafety values were smallest and often negative for Q. alba , suggesting that Q. albaoperated at lower proportional level of its maximum hydraulic function relative to L. tulipifera and A. saccharum . Likewise, estimated in-situ Kstem values were consistently lower for Q. alba than other co-occurring study species. For example, in the NC_E chronosequence, where Ψsafety were often smallest, estimated in-situKstem was 0.30 kg m-1s-1 MPa-1 (SE = 0.133) for Q. alba , accounting for a 67.1% reduction of Kmax . By comparison, estimated in-situ Kstem forL. tulipifera and A. saccharum in these sites were 1.12 kg m-1 s-1 MPa-1 (SE = 0.148) and 0.62 kg m-1 s-1MPa-1 (SE = 0.04) (accounting for a 18.79% and 7.02% reduction of Kmax , respectively).
7. Discussion
We tested three hypotheses to assess variability and coordination of key plant hydraulic traits across ten deciduous forest stands. We found little support for hypotheses 1 and 2; we observed that variation of embolism vulnerability was better explained by species than intraspecific factors. While we detected some region and age effects, variation in vulnerability was principally determined by the large species effect. Additionally, our results did not confirm hypothesis 3, which predicted stricter ΨL regulation would be associated more vulnerable xylem. Contrary to the prevailing paradigm, we found that anisohydric Q. alba possessed stem tissues more vulnerable to embolism than isohydric L. tulipifera and A. saccharum . Moreover, we found that Q. alba had small or negative Ψsafety, such that its anisohydric behavior likely occurred with a substantial hydraulic cost.
7.1 Why were embolism thresholds invariant with climate and stand age?
We found little variation in hydraulic vulnerability across climate and age. This result, however, must be reconciled with the body of work demonstrating vegetation’s capacity to acclimate xylem to pedo-climatic conditions (Awad et al ., 2010; Durante et al ., 2011; Gea-Izquierdo et al ., 2012). The clearest trends of acclimation are often found in manipulation experiments (Beikircher & Mayr, 2009; Awad et al ., 2010). However, surveys of hydraulic traits across species’ ranges have found these patterns more ambiguous (Martínez-Vilalta et al ., 2009; Wortemann et al ., 2011; Charra-Vaskou et al ., 2012; Lamy et al ., 2014).
The similarity across climate observed here may be evidence that acclimation reflects a broader set of morphological changes to the whole-plant hydraulic architecture, rather than simple adjustments to stem xylem traits (Lamy et al ., 2014). Modifications to other drought-tolerant traits may therefore explain how Q. alba,L. tulipifera , and A. saccharum establish dominance across diverse climate ranges. Such acclimation may include modifications to leaf:sapwood area ratio (Addington et al., 2006; Martínez-Vilatlaet al ., 2009), root:leaf area ratio (Sperry et al ., 2002), fine root turnover (Meier & Leuschner, 2008), or vulnerability of root tissues (Alder et al., 1996; Wolfe et al ., 2016).
The absence of xylem acclimation with age/height may similarly reflect a reliance on morphological changes to alleviate emerging hydraulic constraints as forests mature. As canopies grow in height, greater xylem tension and pathlength resistance restricts hydraulic transport to canopy leaves (McDowellet al ., 2002; Novick et al ., 2009). To cope with these constraints, species representing the extremes of tree height often show strong patterns of increased stem embolism resistance with increased height in the canopy, indicative of acclimation (Burgess et al ., 2006; Ambrose et al ., 2009). Although age effects on embolism thresholds were minimal across stands, age also had little impact on ΨL decline. Thus, age-related constraints may have been mitigated through whole-plant adjustments that reduce damaging plant water potential gradients, rather than increased xylem resistance.