Leaf gas exchange measurements provide an important tool for inferring a plant’s photosynthetic biochemistry. In most cases, the responses of photosynthetic CO assimilation to variable intercellular CO concentrations ( A / C response curves) are used to model the maximum rate of carboxylation by ribulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco, V ) and the rate of electron transport at a given photosynthetically active radiation (PAR; J ). The standard Farquhar-Von Caemmerer-Berry model is typically used with default parameters of Rubisco kinetic values and mesophyll conductance to CO ( g ) derived from tobacco that impairs analytical reliability across species. To study this, here we measured the temperature responses of key in vitro Rubisco catalytic properties and g in cotton ( Gossypium hirsutum cv. Sicot 71) and derived V and J ( J at 2000 µmol m s PAR) from cotton A / C curves incrementally measured at 15°C to 40°C using cotton and tobacco parameters with our new automated fitting R package ‘OptiFitACi’. When applied to cotton, the tobacco parameters produced unrealistic J : V ratio of <1 at 25°C, two- to three-fold higher estimates of V , approximately 50% higher estimates of J and more variable estimates of V and J , compared to model parameterisation with cotton-derived values. We determined that errors arise when using a g of 0.23 mol m s bar or below and Rubisco CO -affinities under ambient O ( K ) outside 461 µbar to 627 µbar to model A / C responses in cotton. We show how the multi- A / C modelling capabilities of ‘OptiFitACi’ serves as a robust, user-friendly extension of ‘plantecophys’ by providing simplified temperature-sensitivity and species-specificity parameterisation capabilities to enable higher accuracy estimates of V and J .

Summary Portable gas exchange analysers provide critical data for understanding plant‐atmosphere carbon and water fluxes, and for parameterising Earth system models that forecast climate change effects and feedbacks. We characterised temperature measurement errors in the Li‐Cor LI‐6400XT and LI‐6800, and estimated downstream errors in derived quantities, including stomatal conductance (gsw) and leaf intercellular CO2 concentration (Ci). The LI‐6400XT exhibited air temperature errors (differences between reported air temperature and air temperature measured near the leaf) up to 7.2°C, leaf temperature errors up to 5.3°C, and relative errors in gsw and Ci that increased as temperatures departed from ambient. This caused errors in leaf‐to‐air temperature relationships, assimilation–temperature curves and CO2 response curves. Temperature dependencies of maximum Rubisco carboxylation rate (Vcmax) and maximum RuBP regeneration rate (Jmax) showed errors of 12% and 35%, respectively. These errors are likely to be idiosyncratic and may differ among machines and environmental conditions. The LI‐6800 exhibited much smaller errors. Earth system model predictions may be erroneous, as much of their parametrisation data were measured on the LI‐6400XT system, depending on the methods used. We make recommendations for minimising errors and correcting data in the LI‐6400XT. We also recommend transitioning to the LI‐6800 for future data collection.

Climate warming will alter photosynthesis and respiration not only via direct temperature effects on leaf biochemistry but also by increasing atmospheric dryness, thereby reducing stomatal conductance and suppressing photosynthesis. Our knowledge on how climate warming affects these processes is mainly derived from seedlings grown under highly controlled conditions. However, little is known regarding temperature responses of trees growing under field settings. We exposed mature tamarack and black spruce trees growing in a peatland ecosystem to whole‐ecosystem warming of up to +9°C above ambient air temperatures in an ongoing long‐term experiment (SPRUCE: Spruce and Peatland Responses Under Changing Environments). Here, we report the responses of leaf gas exchange after the first two years of warming. We show that the two species exhibit divergent stomatal responses to warming and vapor pressure deficit. Warming of up to 9°C increased leaf N in both spruce and tamarack. However, higher leaf N in the warmer plots translate into higher photosynthesis in tamarack but not in spruce, with photosynthesis being more constrained by stomatal limitations in spruce than in tamarack under warm conditions. Surprisingly, dark respiration did not acclimate to warming in spruce, and thermal acclimation of respiration was only seen in tamarack once changes in leaf N were considered. Our results highlight how warming can lead to differing stomatal responses to warming in co‐occurring species, with consequent effects on both vegetation carbon and water dynamics.

Leaf-level gas exchange data support the mechanistic understanding of plant fluxes of carbon and water. These fluxes inform our understanding of ecosystem function, are an important constraint on parameterization of terrestrial biosphere models, are necessary to understand the response of plants to global environmental change, and are integral to efforts to improve crop production. Collection of these data using gas analyzers can be both technically challenging and time consuming, and individual studies generally focus on a small range of species, restricted time periods, or limited geographic regions. The high value of these data is exemplified by the many publications that reuse and synthesize gas exchange data, however the lack of metadata and data reporting conventions make full and efficient use of these data difficult. Here we propose a reporting format for leaf-level gas exchange data and metadata to provide guidance to data contributors on how to store data in repositories to maximize their discoverability, facilitate their efficient reuse, and add value to individual datasets. For data users, the reporting format will better allow data repositories to optimize data search and extraction, and more readily integrate similar data into harmonized synthesis products. The reporting format specifies data table variable naming and unit conventions, as well as metadata characterizing experimental conditions and protocols. For common data types that were the focus of this initial version of the reporting format, i.e., survey measurements, dark respiration, carbon dioxide and light response curves, and parameters derived from those measurements, we took a further step of defining required additional data and metadata that would maximize the potential reuse of those data types. To aid data contributors and the development of data ingest tools by data repositories we provided a translation table comparing the outputs of common gas exchange instruments. Extensive consultation with data collectors, data users, instrument manufacturers, and data scientists was undertaken in order to ensure that the reporting format met community needs. The reporting format presented here is intended to form a foundation for future development that will incorporate additional data types and variables as gas exchange systems and measurement approaches advance in the future. The reporting format is published in the U.S. Department of Energy's ESS-DIVE data repository, with documentation and future development efforts being maintained in a version control system.

Summary Understanding biological temperature responses is crucial to predicting global carbon fluxes. The current approach to modelling temperature responses of photosynthetic capacity in large scale modelling efforts uses a modified Arrhenius equation. We rederived the modified Arrhenius equation from the source publication from 1942 and uncovered a missing term that was dropped by 2002. We compare fitted temperature response parameters between the correct and incorrect derivation of the modified Arrhenius equation. We find that most parameters are minimally affected, though activation energy is impacted quite substantially. We then scaled the impact of these small errors to whole plant carbon balance and found that the impact of the rederivation of the modified Arrhenius equation on modelled daily carbon gain causes a meaningful deviation of c. 18% day−1. This suggests that the error in the derivation of the modified Arrhenius equation has impacted the accuracy of predictions of carbon fluxes at larger scales since > 40% of Earth System Models contain the erroneous derivation. We recommend that the derivation error be corrected in modelling efforts moving forward.

Summary Minimum conductance ( g w,min ) in leaves is important for water relations in land plants. Yet, its regulation is unclear due to measurement constraints. Cuticle conductance to water vapor ( g cw ) was estimated from the difference between calculated and direct measurement of CO 2 concentration in the leaf airspace ( C i ) of amphi-stomatous tobacco and sunflower. We estimated g cw in a series of light and dark experiments, and partitioned g w,min into cuticle and stomatal components. Some leaves were detached to simulate severe drought through desiccation conditions where g w,min is generally determined. Between light and dark experiments each g cw was in close agreement, and successfully corrected the discrepancies of calculations from direct measurements. In the dark, either stomatal or cuticle conductance dominated the g w,min , suggesting either of them can control the minimum water loss. In the detached leaves, g cw could not be estimated likely due to unsaturation in the leaf airspace, and g w,min was progressively underestimated. Besides cuticle, leaf water status is a potential pitfall of the standard gas exchange model. Our technique is useful to study the minimal gas exchange as well as to refine the model.

Understanding biological temperature responses is crucial to predicting global carbon fluxes. The current approach to modeling photosynthetic temperature responses in large scale modeling efforts uses a modified Arrhenius equation. We rederived the modified Arrhenius equation from the source publication and uncovered a missing term that was dropped between 1942 and 2002. We compare fitted temperature response parameters between the new and old derivation of the modified Arrhenius equation. We find that most parameters are minimally affected, though small errors still exist. We then scaled the impact of these small errors to whole plant carbon balance and found that the impact of the rederivation of the Arrhenius on modelled daily carbon gain causes a meaningful deviation of ~1.8%. This suggests that the error in the derivation of the modified Arrhenius equation has impacted predictions of carbon fluxes at larger scales. We argue that it is time to move beyond the modified Arrhenius paradigm since the current implementation is categorically incorrect and to use more thermodynamically-grounded temperature response equations going forward.

The assumptions that water vapor exchange occurs exclusively through stomata, that the intercellular airspace is fully saturated with water vapor, and that CO 2 gradients are negligible between stomata and the intercellular airspace have enabled significant advancements in photosynthetic gas exchange research for nearly 60 years via calculation of intercellular CO 2 (C i ). However, available evidence suggests that these assumptions may be overused. Here we review the literature surrounding evidence for and against the assumptions made by Moss & Rawlins (1963). We reinterpret data from the literature by propagating different rates of cuticular water loss, CO 2 gradients, and unsaturation through the data. We find that in general, when cuticle conductance is less than 1% of stomatal conductance, the assumption that water vapor exchange occurs exclusively through stomata has a marginal effect on gas exchange calculations, but this is not true when cuticle conductance exceeds 5% of stomatal conductance. Our analyses further suggest that CO 2 and water vapor gradients have stronger impacts at higher stomatal conductance, while cuticle conductance has a greater impact at lower stomatal conductance. Therefore, we recommend directly measuring C i whenever possible, measuring apoplastic water potentials to estimate humidity inside the leaf, and exercising caution when interpreting data under conditions of high temperature and/or low stomatal conductance, and when a species is known to have high cuticular conductance. Highlight Leaf water vapor and CO 2 exchange have been successfully used to model photosynthetic biochemistry. We review critical assumptions in these models and make recommendations about which need to be re-assessed.

### Competing Interest Statement The authors have declared no competing interest.

The Rapid A/Ci Response (RACiR) is a dynamic method of leaf-level gas exchange that allows the calculation of fundamental parameters for photosynthetic capacity in much shorter times than standard steady-state methods. This is accomplished by using CO2 ramps instead of discrete values under steady-state conditions. Here, we present data describing potential pitfalls and provide a list of best practices that are important to follow in order to ensure the data acquired are of high quality. Stinziano et al. (2017) demonstrated that the RACiR generates robust estimates of Vcmax (maximal carboxylation rate) and J (potential rate of electron transport) similar to the standard A/Ci, which can be very useful in phenotyping applications. They also suggested that the RACiR technique may generate new biological insights that are unattainable from slower measurements (Stinziano et al., 2017). Here, we respond to the Letter by Taylor & Long (2019; in this issue of New Phytologist, pp. 621–624), reiterating the points from Stinziano et al. (2017) and emphasizing that applying RACiR methodology outside of the specific conditions and recommendations of Stinziano et al. (2017) requires further testing. Fig. 1(a,b) of Taylor & Long (2019) shows data using a ramp of 200 μmol mol−1 min−1, which Stinziano et al. (2017) stated as being a rate that may compromise the estimate of J. Therefore, it is not surprising to see examples at that speed that generate larger deviations from a standard A/Ci than was in our data, provided in both figures and supplementary files of Stinziano et al. (2017) as well as in Fig. 1(c) of Taylor & Long (2019). Here, we present results demonstrating that higher ramp rates generate larger differences in the apparent compensation point (Fig. 1). These results also demonstrate that CO2 ramping rate can result in apparent assimilation offsets. Such offsets may be due to instrumentation artifacts such as an imperfect empty chamber correction or other effects that are not yet demonstrated. At sufficiently high ramp rates, differences between standard steady-state and dynamic RACiR gas exchange methods would be expected and could reflect real differences in the underlying biology. When CO2 can be changed faster than enzyme activation states, stomatal conductance, mesophyll conductance, or faster than some biochemical pools can respond, then gas exchange data may need reinterpretation, and this new approach may offer opportunities to test model assumptions. However, we note that higher CO2 ramp rates are complicated and we recommend limiting ramp rates to 100 ppm min−1, unless one is explicitly examining the mechanisms behind high ramp rate offsets. Taylor & Long (2019) extended data analyses to parameters that were beyond the scope of Stinziano et al. (2017). Based on potential procedural or biological concerns mentioned earlier, at high ramp rates RACiR may not generate compensation points similar to a standard A/Ci. However, we also point out that standard A/Ci methods used to generate estimates of other parameters presented in Taylor & Long (2019), including gm (mesophyll conductance), Rd (dark respiration), and Γ* (photorespiratory compensation point) are potentially problematic (Pons et al., 2009; Walker & Cousins, 2013; Walker & Ort, 2015; Farquhar & Busch, 2017) and we would not recommend using them. Taylor & Long (2019) also combine data from the 500 to 0 and 300 to 800 μmol mol−1 RACiRs of Stinziano et al. (2017) when performing model fits. While compiling a larger data set for curve fitting is understandable, we do not think it is appropriate here. Notably, the RACiRs in Stinziano et al. (2017) were not conducted in a way to maximize alignment between ramp ranges due to the way the CO2 ramping loops were set up. The order of measurements is known to affect the results as evidenced by the common practice, in a standard A/Ci, of carefully returning to a common mid-point value between low and high CO2 ranges. This is required because of the biological effects occurring during slow measurements, such as enzyme deactivation, in each CO2 range. Therefore, we recommend against combining RACiRs from multiple ranges. When we performed a re-analysis of the Stinziano et al. (2017) data using the Taylor & Long (2019) script, but restricted the fit to the 300–800 μmol mol−1 RACiRs, most of the significant differences presented in Taylor & Long (2019) were no longer significant. However, we do not present the table here because we believe a larger data set is needed for proper statistical analyses. Every parameter estimate has some sort of error associated with it. When relatively small numbers of parameter estimates are compared using a standard statistical test (such as a t-test) this error is essentially ignored, and so we urge caution in declaring true differences between parameter estimates based on relatively limited data. Essentially, a larger data set that includes analyses of all errors would be helpful in order to more fully investigate these issues. Taylor & Long (2019) raise an excellent point: curve fitting approach matters. This issue was thoroughly examined by Gu et al. (2010) and has been repeatedly addressed by multiple groups for well over a decade (e.g. Ethier & Livingston, 2004; Dubois et al., 2007; Sharkey et al., 2007). It is important to recognize that different methods for fitting A/Ci responses can yield different parameter estimates. This makes assigning ‘truth’ more difficult since the method used can affect the outcome. This is especially true for values like gm, Rd, and Γ* such that it is not widely accepted that these are optimally derived from a single standard A/Ci (see methods Pons et al., 2009; Walker & Cousins, 2013; Walker & Ort, 2015; Farquhar & Busch, 2017). The most widely accepted use for the A/Ci is to estimate Vcmax and J (with much less certainty about J) and those values are in ‘reasonably close agreement’ between the standard and RACiR approaches in Taylor & Long (2019) and in Stinziano et al. (2017). While it is logical and valuable to compare standard A/Ci and RACiR approaches for fitting other parameters as Taylor & Long (2019) have done, assigning truth becomes more difficult given the uncertainty of model fitting as well as the methodological points we have raised earlier. Given the importance of careful consideration of ramping conditions and preliminary testing with a species of interest, we are providing a more detailed and explicit procedural guide for new users. RACiR requires characterization of the species in question to determine the best way to setup the CO2 ramps, especially the ramping rate. To date, a ramping rate of 100 μmol mol−1 min−1 has provided the best comparisons with steady-state Vcmax and J. RACiR empty-chamber calibration curves are specific to flow rate, temperature, and ramp rate, and may be sensitive to large changes in water mole fraction during the ramp. As such, we are clarifying key recommendations whenever RACiR is used: Taken together, we think the main conclusions of Stinziano et al. (2017) still hold in that RACiR can generate estimates of Vcmax and J that are substantially similar to estimates derived from the standard approach, and that RACiR is a useful screening tool for rapid phenotyping. The work of Taylor & Long (2019) is helpful in determining where differences in parameter estimates between the two methods may exist, and points to the need for additional research as RACiR is further developed.

Boreal forests are crucial in regulating global vegetation-atmosphere feedbacks, but the impact of climate change on boreal tree carbon fluxes is still unclear. Given the sensitivity of global vegetation models to photosynthetic and respiration parameters, we determined how predictions of net carbon gain (C-gain) respond to variation in these parameters using a stand-level model (MAESTRA). We also modelled how thermal acclimation of photosynthetic and respiratory temperature sensitivity alters predicted net C-gain responses to climate change. We modelled net C-gain of seven common boreal tree species under eight climate scenarios across a latitudinal gradient to capture a range of seasonal temperature conditions. Physiological parameter values were taken from the literature together with different approaches for thermally acclimating photosynthesis and respiration. At high latitudes, net C-gain was stimulated up to 400% by elevated temperatures and CO2 in the autumn but suppressed at the lowest latitudes during midsummer under climate scenarios that included warming. Modelled net C-gain was more sensitive to photosynthetic capacity parameters (V-cmax, J(max), Arrhenius temperature response parameters, and the ratio of J(max) to V-cmax) than stomatal conductance or respiration parameters. The effect of photosynthetic thermal acclimation depended on the temperatures where it was applied: acclimation reduced net C-gain by 10%-15% within the temperature range where the equations were derived but decreased net C-gain by 175% at temperatures outside this range. Thermal acclimation of respiration had small, but positive, impacts on net C-gain. We show that model simulations are highly sensitive to variation in photosynthetic parameters and highlight the need to better understand the mechanisms and drivers underlying this variability (e.g., whether variability is environmentally and/or biologically driven) for further model improvement.

The rapid A-Ci response (RACiR) technique alleviates limitations of measuring photosynthetic capacity by reducing the time needed to determine the maximum carboxylation rate (Vcmax ) and electron transport rate (Jmax ) in leaves. Photosynthetic capacity and its relationships with leaf development are important for understanding ecological and agricultural productivity; however, our current understanding is incomplete. Here, we show that RACiR can be used in previous generation gas exchange systems (i.e., the LI-6400) and apply this method to rapidly investigate developmental gradients of photosynthetic capacity in poplar. We compared RACiR-determined Vcmax and Jmax as well as respiration and stomatal conductance (gs ) across four stages of leaf expansion in Populus deltoides and the poplar hybrid 717-1B4 (Populus tremula × Populus alba). These physiological data were paired with leaf traits including nitrogen concentration, chlorophyll concentrations, and specific leaf area. Several traits displayed developmental trends that differed between the poplar species, demonstrating the utility of RACiR approaches to rapidly generate accurate measures of photosynthetic capacity. By using both new and old machines, we have shown how more investigators will be able to incorporate measurements of important photosynthetic traits in future studies and further our understanding of relationships between development and leaf-level physiology.

Steady-state photosynthetic CO2 responses (A/Ci curves) are used to assess environmental responses of photosynthetic traits and to predict future vegetative carbon uptake through modeling. The recent development of rapid A/Ci curves (RACiRs) permits faster assessment of these traits by continuously changing [CO2 ] around the leaf, and may reveal additional photosynthetic properties beyond what is practical or possible with steady-state methods. Gas exchange necessarily incorporates photosynthesis and (photo)respiration. Each process was expected to respond on different timescales due to differences in metabolite compartmentation, biochemistry and diffusive pathways. We hypothesized that metabolic lags in photorespiration relative to photosynthesis/respiration and CO2 diffusional limitations can be detected by varying the rate of change in [CO2 ] during RACiR assays. We tested these hypotheses through modeling and experiments at ambient and 2% oxygen. Our data show that photorespiratory delays cause offsets in predicted CO2 compensation points that are dependent on the rate of change in [CO2 ]. Diffusional limitations may reduce the rate of change in chloroplastic [CO2 ], causing a reduction in apparent RACiR slopes under high CO2 ramp rates. Multirate RACiRs may prove useful in assessing diffusional limitations to gas exchange and photorespiratory rates.

Accurate values of photosynthetic capacity are needed in Earth System Models to predict gross primary productivity. Seasonal changes in photosynthetic capacity in these models are primarily driven by temperature, but recent work has suggested that photoperiod may be a better predictor of seasonal photosynthetic capacity. Using field-grown kudzu (Pueraria lobata (Willd.) Ohwi), a nitrogen-fixing vine species, we took weekly measurements of photosynthetic capacity, leaf nitrogen, and pigment and photosynthetic protein concentrations and correlated these with temperature, irradiance and photoperiod over the growing season. Photosynthetic capacity was more strongly correlated with photoperiod than with temperature or daily irradiance, while the growing season pattern in photosynthetic capacity was uncoupled from changes in leaf nitrogen, chlorophyll and Rubisco. Daily estimates of the maximum carboxylation rate of Rubisco (Vcmax) based on either photoperiod or temperature were correlated in a non-linear manner, but Vcmax estimates from both approaches that also accounted for diurnal temperature fluctuations were similar, indicating that differences between these models depend on the relevant time step. We advocate for considering photoperiod, and not just temperature, when estimating photosynthetic capacity across the year, particularly as climate change alters temperatures but not photoperiod. We also caution that the use of leaf biochemical traits as proxies for estimating photosynthetic capacity may be unreliable when the underlying relationships between proxy leaf traits and photosynthetic capacity are established outside of a seasonal framework.

Photosynthetic temperature acclimation could strongly affect coupled vegetation-atmosphere feedbacks in the global carbon cycle, especially as the climate warms. Thermal acclimation of photosynthesis can be modelled as changes in the parameters describing the direct effect of temperature on photosynthetic capacity (i.e., activation energy, Ea ; deactivation energy, Hd ; entropy parameter, ΔS) or the basal value of photosynthetic capacity (i.e., photosynthetic capacity measured at 25°C). However, the impact of acclimating these parameters (individually or in combination) on vegetative carbon gain is relatively unexplored. Here we compare the ability of 66 photosynthetic temperature acclimation scenarios to improve the ability of a spatially explicit canopy carbon flux model, MAESTRA, to predict eddy covariance data from a loblolly pine forest. We show that: (1) incorporating seasonal temperature acclimation of basal photosynthetic capacity improves the model's ability to capture seasonal changes in carbon fluxes and outperforms acclimation of other single factors (i.e., Ea or ΔS alone); (2) multifactor scenarios of photosynthetic temperature acclimation provide minimal (if any) improvement in model performance over single factor acclimation scenarios; (3) acclimation of Ea should be restricted to the temperature ranges of the data from which the equations are derived; and (4) model performance is strongly affected by the Hd parameter. We suggest that a renewed effort be made into understanding whether basal photosynthetic capacity, Ea , Hd and ΔS co-acclimate across broad temperature ranges to determine whether and how multifactor thermal acclimation of photosynthesis occurs.

Phenotyping for photosynthetic gas exchange parameters is limiting our ability to select plants for enhanced photosynthetic carbon gain and to assess plant function in current and future natural environments. This is due, in part, to the time required to generate estimates of the maximum rate of ribulose-1,5-bisphosphate carboxylase oxygenase (Rubisco) carboxylation (Vc,max ) and the maximal rate of electron transport (Jmax ) from the response of photosynthesis (A) to the CO2 concentration inside leaf air spaces (Ci ). To relieve this bottleneck, we developed a method for rapid photosynthetic carbon assimilation CO2 responses [rapid A-Ci response (RACiR)] utilizing non-steady-state measurements of gas exchange. Using high temporal resolution measurements under rapidly changing CO2 concentrations, we show that RACiR techniques can obtain measures of Vc,max and Jmax in ~5 min, and possibly even faster. This is a small fraction of the time required for even the most advanced gas exchange instrumentation. The RACiR technique, owing to its increased throughput, will allow for more rapid screening of crops, mutants and populations of plants in natural environments, bringing gas exchange into the phenomic era.

Climate warming is expected to increase the seasonal duration of photosynthetic carbon fixation and tree growth in high-latitude forests. However, photoperiod, a crucial cue for seasonality, will remain constant, which may constrain tree responses to warming. We investigated the effects of temperature and photoperiod on weekly changes in photosynthetic capacity, leaf biochemistry and growth in seedlings of a boreal evergreen conifer, white spruce [Picea glauca (Moench) Voss]. Warming delayed autumn declines in photosynthetic capacity, extending the period when seedlings had high carbon uptake. While photoperiod was correlated with photosynthetic capacity, short photoperiods did not constrain the maintenance of high photosynthetic capacity under warming. Rubisco concentration dynamics were affected by temperature but not photoperiod, while leaf pigment concentrations were unaffected by treatments. Respiration rates at 25 °C were stimulated by photoperiod, although respiration at the growth temperatures was increased in warming treatments. Seedling growth was stimulated by increased photoperiod and suppressed by warming. We demonstrate that temperature is a stronger control on the seasonal timing of photosynthetic down-regulation than is photoperiod. Thus, while warming can stimulate carbon uptake in boreal conifers, the extra carbon may be directed towards respiration rather than biomass, potentially limiting carbon sequestration under climate change.