High-temperature tolerance in plants is important in a warming world, with extreme heat waves predicted to increase in frequency and duration, potentially leading to lethal heating of leaves. Global patterns of high-temperature tolerance are documented in animals, but generally not in plants, limiting our ability to assess risks associated with climate warming. To assess whether there are global patterns in high-temperature tolerance of leaf metabolism, we quantified T (high temperature where minimal chlorophyll a fluorescence rises rapidly and thus photosystem II is disrupted) and T (temperature where leaf respiration in darkness is maximal, beyond which respiratory function rapidly declines) in upper canopy leaves of 218 plant species spanning seven biomes. Mean site-based T values ranged from 41.5 °C in the Alaskan arctic to 50.8 °C in lowland tropical rainforests of Peruvian Amazon. For T , the equivalent values were 51.0 and 60.6 °C in the Arctic and Amazon, respectively. T and T followed similar biogeographic patterns, increasing linearly (˜8 °C) from polar to equatorial regions. Such increases in high-temperature tolerance are much less than expected based on the 20 °C span in high-temperature extremes across the globe. Moreover, with only modest high-temperature tolerance despite high summer temperature extremes, species in mid-latitude (~20-50°) regions have the narrowest thermal safety margins in upper canopy leaves; these regions are at the greatest risk of damage due to extreme heat-wave events, especially under conditions when leaf temperatures are further elevated by a lack of transpirational cooling. Using predicted heat-wave events for 2050 and accounting for possible thermal acclimation of T and T , we also found that these safety margins could shrink in a warmer world, as rising temperatures are likely to exceed thermal tolerance limits. Thus, increasing numbers of species in many biomes may be at risk as heat-wave events become more severe with climate change.

SummaryWhy do forest productivity and biomass decline with elevation? To address this question, research to date generally has focused on correlative approaches describing changes in woody growth and biomass with elevation.We present a novel, mechanistic approach to this question by quantifying the autotrophic carbon budget in 16 forest plots along a 3300 m elevation transect in Peru.Low growth rates at high elevations appear primarily driven by low gross primary productivity (GPP), with little shift in either carbon use efficiency (CUE) or allocation of net primary productivity (NPP) between wood, fine roots and canopy. The lack of trend in CUE implies that the proportion of photosynthate allocated to autotrophic respiration is not sensitive to temperature. Rather than a gradual linear decline in productivity, there is some limited but nonconclusive evidence of a sharp transition in NPP between submontane and montane forests, which may be caused by cloud immersion effects within the cloud forest zone. Leaf-level photosynthetic parameters do not decline with elevation, implying that nutrient limitation does not restrict photosynthesis at high elevations.Our data demonstrate the potential of whole carbon budget perspectives to provide a deeper understanding of controls on ecosystem functioning and carbon cycling.

Summary We examined whether variations in photosynthetic capacity are linked to variations in the environment and/or associated leaf traits for tropical moist forests (TMFs) in the Andes/western Amazon regions of Peru. We compared photosynthetic capacity (maximal rate of carboxylation of Rubisco (Vcmax), and the maximum rate of electron transport (Jmax)), leaf mass, nitrogen (N) and phosphorus (P) per unit leaf area (Ma, Na and Pa, respectively), and chlorophyll from 210 species at 18 field sites along a 3300‐m elevation gradient. Western blots were used to quantify the abundance of the CO2‐fixing enzyme Rubisco. Area‐ and N‐based rates of photosynthetic capacity at 25°C were higher in upland than lowland TMFs, underpinned by greater investment of N in photosynthesis in high‐elevation trees. Soil [P] and leaf Pa were key explanatory factors for models of area‐based Vcmax and Jmax but did not account for variations in photosynthetic N‐use efficiency. At any given Na and Pa, the fraction of N allocated to photosynthesis was higher in upland than lowland species. For a small subset of lowland TMF trees examined, a substantial fraction of Rubisco was inactive. These results highlight the importance of soil‐ and leaf‐P in defining the photosynthetic capacity of TMFs, with variations in N allocation and Rubisco activation state further influencing photosynthetic rates and N‐use efficiency of these critically important forests.

Understanding of the extent of acclimation of light-saturated net photosynthesis (A ) to temperature (T), and associated underlying mechanisms, remains limited. This is a key knowledge gap given the importance of thermal acclimation for plant functioning, both under current and future higher temperatures, limiting the accuracy and realism of Earth system model (ESM) predictions. Given this, we analysed and modelled T-dependent changes in photosynthetic capacity in 10 wet-forest tree species: six from temperate forests and four from tropical forests. Temperate and tropical species were each acclimated to three daytime growth temperatures (T ): temperate - 15, 20 and 25 °C; tropical - 25, 30 and 35 °C. CO response curves of A were used to model maximal rates of RuBP (ribulose-1,5-bisphosphate) carboxylation (V ) and electron transport (J ) at each treatment's respective T and at a common measurement T (25 °C). SDS-PAGE gels were used to determine abundance of the CO -fixing enzyme, Rubisco. Leaf chlorophyll, nitrogen (N) and mass per unit leaf area (LMA) were also determined. For all species and T , A at current atmospheric CO partial pressure was Rubisco-limited. Across all species, LMA decreased with increasing T . Similarly, area-based rates of V at a measurement T of 25 °C (V ) linearly declined with increasing T , linked to a concomitant decline in total leaf protein per unit leaf area and Rubisco as a percentage of leaf N. The decline in Rubisco constrained V and A for leaves developed at higher T and resulted in poor predictions of photosynthesis by currently widely used models that do not account for T -mediated changes in Rubisco abundance that underpin the thermal acclimation response of photosynthesis in wet-forest tree species. A new model is proposed that accounts for the effect of T -mediated declines in V on A , complementing current photosynthetic thermal acclimation models that do not account for T sensitivity of V .

Mesophyll conductance ( g m ) is known to affect plant photosynthesis. However, g m is rarely explicitly considered in land surface models (LSMs), with the consequence that its role in ecosystem and large‐scale carbon and water fluxes is poorly understood. In particular, the different magnitudes of g m across plant functional types (PFTs) are expected to cause spatially divergent vegetation responses to elevated CO 2 concentrations. Here, an extensive literature compilation of g m across major vegetation types is used to parameterize an empirical model of g m in the LSM JSBACH and to adjust photosynthetic parameters based on simulated A n − C i curves. We demonstrate that an explicit representation of g m changes the response of photosynthesis to environmental factors, which cannot be entirely compensated by adjusting photosynthetic parameters. These altered responses lead to changes in the photosynthetic sensitivity to atmospheric CO 2 concentrations which depend both on the magnitude of g m and the climatic conditions, particularly temperature. We then conducted simulations under ambient and elevated (ambient + 200 μmol/mol) CO 2 concentrations for contrasting ecosystems and for historical and anticipated future climate conditions (representative concentration pathways; RCPs) globally. The g m ‐explicit simulations using the RCP8.5 scenario resulted in significantly higher increases in gross primary productivity (GPP) in high latitudes (+10% to + 25%), intermediate increases in temperate regions (+5% to + 15%), and slightly lower to moderately higher responses in tropical regions (−2% to +5%), which summed up to moderate GPP increases globally. Similar patterns were found for transpiration, but with a lower magnitude. Our results suggest that the effect of an explicit representation of g m is most important for simulated carbon and water fluxes in the boreal zone, where a cold climate coincides with evergreen vegetation.

No abstract

Globally, trees originating from high-rainfall tropical regions typically exhibit lower rates of light-saturated net CO assimilation (A) compared with those from high-rainfall temperate environments, when measured at a common temperature. One factor that has been suggested to contribute towards lower rates of A is lower mesophyll conductance. Using a combination of leaf gas exchange and carbon isotope discrimination measurements, we estimated mesophyll conductance (g ) of several Australian tropical and temperate wet-forest trees, grown in a common environment. Maximum Rubisco carboxylation capacity, V , was obtained from CO response curves. g and the drawdown of CO across the mesophyll were both relatively constant. V estimated on the basis of intercellular CO partial pressure, C , was equivalent to that estimated using chloroplastic CO partial pressure, C , using 'apparent' and 'true' Rubisco Michaelis-Menten constants, respectively Having ruled out g as a possible factor in distorting variations in A between these tropical and temperate trees, attention now needs to be focused on obtaining more detailed information about Rubisco in these species.

No abstract