Discussion
On an extensive set of 122 herbaceous species covering a broad range of water availability conditions in European temperate grasslands (from dry through mesic to wet grasslands), we documented a coordination between a physiological trait directly linked to water availability (turgor loss point, πtlp) and traits commonly used as proxies of drought tolerance: intrinsic water use efficiency (iWUE), leaf dry matter content (LDMC), specific leaf area (SLA) and maximum plant height (height). By investigating the relationships across species, within major plant functional types (PFTs, forbs and graminoids) and by accounting for plant phylogenetic relatedness, we were able to distinguish the relationships that were strong and more general (πtlp–LDMC and πtlp–iWUE) against those that were weaker (πtlp–SLA) or that could even be misleading at a first glance (πtlp–height).
πtlp is coordinated with intrinsic water-use efficiency
Higher leaf level dehydration tolerance (more negative πtlp) was coordinated with higher intrinsic water use efficiency (less negative δ13C) on all levels considered. This may suggest that though plants adjust their osmotic potential in order to keep stomata open, the mechanism has a limited capacity in temperate plants. This may occur when the intensity of the environmental stress (drought expressed by soil water potential) is not stable but fluctuates deeply below the capacity of the osmotic adjustment. During these episodes, stomata are closed, and the assimilated carbon has less negative δ13C (higher iWUE). The existence of the coordination across our species can be interpreted perhaps as a continuous run (i.e. developing osmotic adjustment) of the drought-exposed plants out of the dehydration stress, which they never quite win (if they won, they would keep stomata open, i.e. lower iWUE), but which is sufficient enough to maintain growth and ensure survival in the more stressful conditions.
πtlp is coordinated with leaf dry matter content
Stronger leaf dehydration tolerance (i.e., more negative πtlp) was tightly coordinated with LDMC, morphological trait commonly used as a proxy of plant’s drought resistance. One explanation is that because low osmotic potential is a response to drought stress (e.g. Májeková et al. 2019), water limitation simultaneously results in smaller cell wall expansion, smaller cells and/or more cell walls, i.e. greater LDMC (Poorter et al. 2009). Our results are in line with previous evidence on coordination between these two leaf-level traits in herbaceous species: on 33 C4 grassland species (Liu & Osborne 2015) and on 19 tallgrass prairie species including C3 and C4 graminoids, forbs and shrubs (Griffin-Nolan et al. 2019). The fact that we found the coordination between πtlp and LDMC to be strong on all levels, together with the previous evidence including also C4 species, suggests that this coordination is biophysically-based and transferable to other systems, plant functional types and plant families.
πtlp is only weakly coordinated with specific leaf area
Our results imply that the mechanistic link between SLA, plant water economics and soil water availability could not be simply assumed on any given scale without further stratification. The coordination between πtlp and SLA appeared only weakly (P = 0.045) in forbs after plants were stratified by their PFTs and then more strongly when their phylogenetic relatedness was accounted for. Our results fit well into the so far ambiguous evidence regarding the coordination between SLA and πtlp. While a weak relationship between more negative πtlp and higher SLA was documented in a big compiled tree dataset (Zhu et al. 2018), no was found on a more local spatial scale for both trees (Maréchaux et al. 2015, 2019) and herbs (Májeková et al. 2019). Moreover, recent evidence found no relationship between SLA and water availability gradients at different spatial and taxonomical scales (for reviews see Bartlettet al. 2012b; Griffin-Nolan et al. 2018). This together suggest that SLA should be considered very carefully as a proxy reflecting the response of plants to water availability.
πtlp is not coordinated with plant height
The relationship between πtlp and plant height represents a nice example on how a coordination between two traits could be misinterpreted without a further stratification by PFTs or accounting for plant phylogenetic relatedness. When considering all species without any stratification, a weak, but significant, negative relationship appears between πtlp and plant height, suggesting that higher leaf dehydration tolerance (more negative πtlp) is coordinated with taller stature. A potential explanation would offer itself, i.e. that shorter plants growing under taller plants would be sheltered from direct irradiance, and therefore experience less dehydration. However, a closer examination reveals that, in our case, the relationship is driven purely by the pronounced differences between forbs and graminoids in both traits considered and disappears within the PFTs and after accounting for plant phylogenetic relatedness. It seems that the risk of embolism, driving the relationship between plant height and hydraulic traits in woody species (Ryan et al. 2006; Liuet al. 2019), does not play a major role in the shorter, relatively to trees, herbaceous species. Indeed, a height of 1 m generates only –0.01 MPa of gravitational potential, which is negligible when compared with osmotic potential. Rather, plant height in grassland plants seems to be under different and potentially more important selection pressures than water availability, such as competitive ability for light (Keddy & Shipley 1989), thus resulting in little coordination among height and πtlp.
Critical remarks: the use of single traits as proxies of function
One needs to carefully consider that the leaf is a multifunctional organ, and thus bulk leaf traits might not always capture exactly the response to a single abiotic factor in question. For instance, a plant’s response to water availability measured through SLA, a very popular leaf economics spectrum trait, can be simultaneously confounded by light availability, nutrient availability, or herbivory resistance (Walters & Reich 1999; Sack 2004; Poorter et al. 2009; Markesteijn et al. 2011; John et al. 2017). As advocated by Hodgson et al. (2011), one needs to carefully consider how valid the use of such a trait is as a stand-alone proxy of a single function.
Here, we would like to reinforce this view by tentatively proposing three advices. First, we advocate the use of traits known to be directly and mechanistically linked to the factor in question (Griffin-Nolanet al. 2018). In case of water availability, we suggest using πtlp, which is a hydraulic trait feasible to measure on a large number of individuals. Second, if leaf traits are measured that could reflect multiple functions of the leaf, i.e. SLA and/or LDMC, we suggest coupling them with traits known to be mechanistically linked to the factor in question. In case of water availability, this would be πtlp, intrinsic water use efficiency, or any other hydraulic trait. This should be done on at least a subset of species, in order to validate (e.g. by simple correlation) that the morphological traits reflect the function in question.
Third, both approaches can be further reinforced by coupling the quantitative measures with the plant functional types. Indeed, if a relationship (with another trait, or environmental factor) is driven only by the differences in functional types, but does not hold within the types, one might ask whether the trait considered really reflects the function being investigated. For instance, here we showed that stratification by simple plant functional types and/or accounting for species phylogeny might help to identify relationships that are biophysically-based, and therefore potentially better transferable among different systems. Such would be the coordination between πtlp and LDMC, where the strength of the relationship holds true within PFTs, after accounting for phylogenetic relatedness, as well as in other functional groups such as C4 grasses (Liu & Osborne 2015). On the other hand, the need for verifying the relationship is in our case highlighted on the relationship between πtlpand plant vegetative height, which disappeared when further stratifying by PFTs or accounting for phylogenetical relatedness.
Conclusion
Here, we addressed the recent plea for a use of traits that would have a more direct mechanistic link to plant drought resistance and water availability to be considered in community trait ecology and global change ecology (Brodribb 2017; Griffin-Nolan et al. 2018; Volaire 2018). We demonstrated the feasibility of the use of a physiological drought tolerance leaf-level trait, πtlp, that becomes operational via the novel osmometry method for large-scale studies across different plant functional types. In herbaceous plants, πtlp was tightly coordinated with intrinsic water use efficiency and with leaf dry matter content. However, πtlp was not coordinated with plant height, which is the leading dimension of the global spectrum of form and function, and only weakly to SLA, which represents the second most important axis (Díaz et al. 2016). This implies that another important axis related to drought tolerance and water availability has so far been largely omitted in the trait-based ecology. Finally, stratification either by simple plant functional types and/or by accounting for species phylogeny might help to identify relationships that are biophysically-based and under the same selection pressure, and therefore potentially better transferable among different systems.