Research

Photosynthetic heat tolerance sets the upper temperature at which plants can assimilate carbon, and therefore an upper bound on the functioning of the biosphere. It is conventionally quantified using threshold temperatures such as the critical temperature for photosystem II function, which characterize the temperature of a heat exposure but not its duration. Because thermal damage accumulates over time, threshold metrics conflate exposure durations that are not equivalent, and they provide no mechanistic basis for testing causal hypotheses or extrapolating to novel temperature and duration combinations.

We recently developed and tested a quantitative theory of photosynthetic heat inactivation grounded in chemical kinetics, in which the temperature and duration of exposure are accounted for explicitly and hypothesized causal mechanisms are testable. Applying the theory to photosystem II temperature responses across 177 species from 157 families, we found that heat tolerance is primarily governed by protein denaturation rather than lipid membrane breakdown, resolving a long-standing ambiguity between two proposed mechanisms. The theory further predicts an upper acclimation limit near 60 °C, set by the structural stability of heat shock proteins, a prediction consistent with global heat tolerance data for 836 species.

Heat tolerance also behaves differently from leaf functional traits. In a common garden that minimizes covariation between taxonomy and climate, carbon economics and energy balance traits showed clear phylogenetic structure, whereas photosynthetic heat tolerance metrics showed none, and varied along an axis orthogonal to traits. Leaf morphology is primarily constrained by evolutionary history, whereas heat tolerance acclimates readily to leaf microclimates. Trait frameworks that treat these on a common footing conflate processes operating on very different timescales.

This work connects to collaborative efforts addressing how heat tolerance should be measured and what it means at larger scales, including how data and theory at one level of organization may not always scale straightforwardly to other levels, how methodological choices influence reported global variation in thermal tolerance, and how photosystem II acclimates on timescales that snapshot measurements cannot resolve.

Metabolic theory posits that the temperature dependence of all biological rates originates in the kinetics of rate-limiting biochemical reactions, and formalizes this using the Arrhenius model and an activation energy. This idea has occupied scientists for more than a century, and while the premise is sound, the mathematical formalization is not, and we think it constrains what the field can learn.

Almost all biological rates exhibit peaked temperature responses. Fitting monotonic models to peaked data yields activation energies that are variable, systematically underestimated, and strongly sensitive to the observed temperature range and the noise structure of the data. Because these parameter estimates propagate directly into projections of biological rate responses to climate change, the consequences are not confined to methodology. We have argued for a shift to peaked models that characterize the full temperature response, and evaluated how two prominent peaked models behave when applied to incomplete and noisy datasets like those scientists actually collect.

Acclimation resolves a second and longer-standing puzzle. Leaf photosynthesis is strongly temperature sensitive, yet growth and productivity measured across climate gradients frequently are not. By integrating relative growth rate theory, metabolic theory, and biochemical photosynthesis theory into a carbon budget model of plant growth, and testing it with concurrent measurements of acute photosynthesis, acclimated photosynthesis, and growth rate in the same individuals, we showed that photosynthetic acclimation mediates how assimilation kinetics scale from leaf to whole plant. Existing models of photosynthetic acclimation failed to reproduce the observed growth responses.

The timescale of acclimation is much shorter than generally assumed. Across 243 assimilation-temperature response curves from 102 species in 96 families grown in a common garden, local environmental conditions were the strongest predictors of response parameters, with the thermal optimum best predicted by mean air temperature on the day of measurement. Response parameters showed no phylogenetic structure and only modest relationships with leaf functional traits or climate of origin. Acclimation on daily timescales is not well represented in current Earth system models.

These temperature responses propagate to ecosystem-scale carbon uptake. Working with collaborators after the 2021 western North American heatwave, we modelled changes in gross primary productivity across the affected region and found that the sign of the response depended on where leaf temperatures fell relative to photosynthetic thermal optima: GPP declined by up to 75% in warm, arid areas where leaves were driven above their optima, and increased by up to 30% in cooler, wetter regions where warming brought leaves closer to them. Thermal optima are thus not only a detail of leaf physiology, they determine whether a heatwave increases or decreases regional carbon uptake.

Water transport and water loss are inseparable from leaf temperature and carbon gain. Transpiration cools leaves, stomatal closure warms them, and hydraulic network traits determine the supply that makes either possible.

Leaves continue to lose water after stomata close, at a rate described by the leaf minimum conductance to water vapor. This flux is small relative to stomatal conductance under favorable conditions, but during hot drought it becomes consequential for plant mortality and leaf-, plant-, and landscape-scale water balance. Studies partitioning minimum conductance into cuticular and stomatal components have reached different conclusions. Measuring minimum and cuticular conductance across 20 to 50 °C in eleven broadleaf species, we found that the dominant pathway shifts from stomatal toward cuticular transpiration as temperature rises, which offers a parsimonious explanation for the disagreement in the literature. Trait relationships with conductance strengthened at higher temperatures, and cuticular conductance was large enough to bias estimates of photosynthetic capacity in species with low stomatal conductance.

Hydraulic network traits reflect two competing selective pressures: minimizing hydraulic resistance, which favours wide conduits, and preventing conduit wall collapse under tension, which does not. Existing theory addresses the first but not the second, and scaling in roots has been largely unexamined. We developed new theory for vertical variation in conduit thickness-to-span ratios and tested it against nearly 600,000 xylem conduits spanning above and belowground organs in five conifer species, using a new bootstrapping approach to minimize sampling bias. Conduits widened with distance from the leaf tip, with an exponent numerically close to several theoretical predictions, and widened from fine to coarse roots, mirroring the aboveground pattern. Thickness-to-span ratios increased from base to tip, and nearly all conduits exceeded a predicted critical collapse limit with a safety factor greater than two, indicating that xylem walls are generally overbuilt relative to what collapse resistance alone requires. Tip-to-base widening also exhibited curvature in log-log space rather than the strict power law most models predict, with the departure concentrated in the uppermost meter, suggesting that fixed-exponent scaling oversimplifies conduit widening near the leaf tip.

With collaborators, we have also argued that plant water use theory has been constructed too narrowly around water as a cost of carbon gain, and that hypotheses invoking thermal stress avoidance, reproduction, and competition for water remain largely untested.

Leaf temperatures exhibit a range of responses to variation in air temperature. Under poikilothermy, leaf temperature tracks air temperature closely. Under limited homeothermy, leaf temperature varies less than air temperature, so that leaves are warmer than air in cool conditions and cooler than air in warm conditions. Under megathermy, leaf temperatures increase more than air for a given air temperature change. All three responses are predicted from leaf energy budget theory and have been observed empirically. Their relative prevalence within the complex microclimates of real plant canopies remains a key open question, and one with direct consequences for how trait-climate relationships, vegetation dynamics, and ecosystem carbon budgets are modeled.

Leaf energy budgets provide the mechanistic basis for addressing this. Radiative, convective, and evaporative fluxes are governed by leaf traits including size, thickness, absorptance, and stomatal conductance, so the same traits that determine carbon economics also determine leaf temperature. We have developed theory that unifies these two perspectives, deriving thermoregulatory behaviour from selection on leaf thermal and photosynthetic traits to maximize lifetime carbon gain. The theory was consistent with global data for leaf temperatures spanning an air temperature range of approximately 60 °C. Related work derives how the slope of the leaf versus air temperature relationship should vary with traits and microenvironment.

Predicting these responses from the traits ecologists routinely measure has proven difficult. Across a 1,100 m elevational gradient spanning desert, montane, and alpine communities, energy balance traits were only weakly related to environmental gradients and poorly predicted by common functional traits. Our (in)ability to measure leaf temperatures compounds the challenge. Because leaf temperature varies strongly with canopy position, measurements taken at the canopy top, which is what most thermal imaging and remote sensing captures, do not accurately quantify canopy-scale thermoregulation.

Our work on heat and mass transfer in plant canopies extends back two decades, to wind tunnel studies of forced convection in conifer branches and buds.