New insights into the leaf economic spectrum could benefit terrestrial models

Abstract

The role of nonstructural carbohydrates in accurate predictions of ecosystem carbon fluxes is becoming increasingly apparent. Over the long term, NSCs represent a buffer against reductions in photosynthesis that often occur during climate extremes (Dietze et al., 2014). Representing NSC storage is crucial for modelling the environmental controls on plant growth, such as temperature and water availability, which can directly affect cell expansion often before limitations from reduced carbon supply by photosynthesis (Fatichi et al., 2014). Changes in NSC concentrations and the conversion of different forms of carbohydrate (e.g. starch to sugar) are also key indicators of drought tolerance in many species (Signori-Muller et al., 2021). However, representing NSCs in large-scale ecosystem models remains a challenge, and many models still ignore the role of NSCs in determining growth and respiration. Those that do represent NSCs often opt to ignore the spatial variation of NSCs within plants, choosing instead to represent just a single carbohydrate pool (Cho et al., 2022). Doing so avoids the need to model the transport of NSC between organs, which is challenging and can introduce significant uncertainty into the models. By shedding light on the relationship between the comparably well-studied LES and the transport of NSC from leaves to the rest of the plant, Asao et al. present a promising avenue through which these processes can be understood and constrained.

Many terrestrial biosphere models and DGVMs already incorporate aspects of the LES (e.g. Fisher et al., 2015; Harper et al., 2016; Peaucelle et al., 2019), supported by large plant trait databases like TRY (Kattge et al., 2011). The LES differentiates between leaves that prioritize fast resource acquisition and those with a more conservative, ‘slow and steady’ investment strategy. On the ‘fast’ end of the spectrum, leaves invest in increased photosynthetic returns at the expense of structure and tend to have short lifespans. On the ‘slow’ end, leaves invest in structure at the expense of photosynthesis rates and tend to have long lifespans (Fig. 1). These investment strategies often extend beyond the leaf-level; for example, low-wood density and higher background mortality rates are associated with fast-growing pioneer trees in tropical forests. Indeed, in their article, Asao et al. hypothesize that the export of NSCs is prioritized over storage on the ‘fast’ end of the spectrum in order to fuel growth and essential functioning elsewhere.

As pointed out by Asao et al., the resource acquisition/conservation trade-offs are evident in global databases of leaf N and P concentrations (per unit mass), maximum rates of photosynthesis (A_max) and carboxylation (V_cmax), leaf dark respiration (R_dark), and specific leaf area (SLA: leaf area per unit dry mass). Using a global plant trait database, Kattge et al. (2009) calculated relationships between leaf nitrogen (N_a, g m⁻²) content and V_cmax at 25°C (μmol m⁻² s⁻¹) and A_max (μmol m⁻² s⁻¹) for different plant functional types (PFTs). This set a precedent for using plant trait information and the LES in terrestrial biosphere models and DGVMs. For example, parameterization of plant functional types in the Joint UK Land Environment Simulator (JULES) DGVM was expanded to include mass-based leaf N, leaf mass per unit area (LMA: the inverse of SLA), leaf lifespan, and the relationships between N_a and V_cmax,25 from Kattge et al. (2009). This enabled bias reduction of gross primary production (GPP) in many biomes (Harper et al., 2016) and improved simulations of global vegetation and biomass distribution (Harper et al., 2018), although the simulated net primary production (NPP) was often too high (suggesting an underestimation of respiration). Some models have gone beyond the single parameter per PFT approach, enabling a prediction of emergent groupings of plants based on known functional trade-offs, including those in the LES (e.g. Pavlick et al., 2013; Schmitt et al., 2020). Another approach based on the LES is to allow traits to vary or co-vary in response to climate, which enables a dynamic and spatially varying set of relationships between climate, leaf strategies, and resultant biogeochemical patterns (Verheijen et al., 2015). These approaches enable a more physiologically based study of pressing global issues ranging from the resilience of tropical forests to the evolution of the global land carbon sink.

Despite the advances enabled by incorporating plant traits into models, there have been fewer marked improvements in representing plant respiration and carbon residence timescales, especially under changing environmental conditions. The results presented by Asao et al. could allow important improvements in modelling capabilities. For example, most land surface models assume that temperature is the main source of variability in leaf and whole plant respiration. However, significant declines in nocturnal leaf respiration are still seen throughout the night, even when temperature is held constant (Bruhn et al., 2022). Substrate limitation could explain these changes (Jones et al., 2024), as the depletion of leaf carbohydrate stores through the night results in the downregulation of respiration. This hypothesis relies on a tight coupling between substrate concentration and substrate utilization rate. Jones et al. (2024) assumed that the respiration rate depends linearly on the concentration of carbohydrate. Typically, though, the relationship between substrate utilization and substrate concentration is not thought to be linear, and instead is assumed to follow a saturating Michaelis–Menten response where the rate of change of respiration with respect to carbohydrate gets smaller the greater the store of carbohydrate. Accounting for this, we might, therefore, expect the magnitude of the decline in respiration to depend on the initial store of NSC at the start of the night. Asao et al. find that on average leaves appear to hold just enough NSC to support one night of respiration and that the majority of daily carbon gain is exported. However, they also find that the magnitude of this export is correlated with leaf metabolic capacity, with leaves at the ‘fast’ end of the LES typically exporting more of their daily carbon gains. Therefore, it might be reasonable to expect leaves at the ‘fast’ end of the LES to display greater reductions in nocturnal respiration under constant temperature as they start on the ‘steeper’ part of the Michaelis–Menten curve. Interestingly, this is exactly what is seen. Bruhn (2023) found that the magnitude of respiration decline depends strongly on the leaf light conditions (with stronger declines occurring in sun-adapted leaves), which are typically correlated with leaf metabolic capacity.

The impact of short-term respiration dynamics on predictions of respiration and, hence, the global carbon sink is uncertain, but Bruhn et al. (2022) estimated a potential increase of 7–10% in simulated NPP for the present day if the nighttime decline in respiration is accounted for. Improving the representation of plant respiration must be a priority for land surface modelers. Clearly, there are trade-offs associated with the LES that we are just beginning to understand. The work presented here by Asao et al. fills in yet another piece of the jigsaw that will ultimately allow us to predict plant functional dynamics at a global scale.

The good news is that the framework for models to incorporate leaf traits and trade-offs already exists. While we must be careful not to think solely from the perspective of the LES (as Asao et al. demonstrate with their measurements of leaf NSC concentrations – not everything is strongly correlated with LES traits), it does unify much of our understanding of how different functional plant types behave. If we can frame new science in the context of the LES, it should allow this new understanding to be more readily incorporated into models and, hence, inform future predictions of the carbon cycle.

The New Phytologist Foundation remains neutral with regard to jurisdictional claims in maps and in any institutional affiliations.