Search

The survival and performance of urban forests are increasingly challenged by urban drought, consequently compromising the sustainability and functionality of urban vegetation. Plant–water relations largely determine species drought tolerance, yet little is known about the hydraulics of urban forest species. Here, we report the leaf hydraulic and carbon traits that govern plant growth and drought resistance, including vulnerability to embolism, hydraulic conductivity and leaf gas exchange characteristics, as well as morphological traits that are potentially linked with these physiological attributes, with the aim of guiding species selection and management in urban forests. Plant materials were collected from mature shrubs and trees on our university campus in Beijing, representing 10 woody species common to urban forests in north China. We found that the leaf embolism resistance, represented by the water potential inducing 50% loss of hydraulic conductivity (P50), as well as the hydraulic safety margin (HSM) defined by P50 and the water potential threshold at the inception of embolism (P12), varied remarkably across species, but was unrelated to growth form. Likewise, stem and leaf-specific hydraulic conductivity (Kstem and kl) was also highly species-specific. Leaf P50 was positively correlated with hydraulic conductivity. However, neither P50 nor hydraulic conductivity was correlated with leaf gas exchange traits, including maximum photosynthetic rate (Amax) and stomatal conductance (gs). Plant morphological and physiological traits were not related, except for specific leaf area, which showed a negative relationship with HSM. Traits influencing plant–water transport were primarily correlated with the mean annual precipitation of species climatic niche. Overall, current common woody species in urban forest environments differed widely in their drought resistance and did not have the capacity to modify these characteristics in response to a changing climate. Species morphology provides limited information regarding physiological drought resistance. Thus, screening urban forest species based on plant physiology is essential to sustain the ecological services of urban forests.

The viability of forest trees, in response to climate change-associated drought, will depend on their capacity to survive through genetic adaptation and phenotypic plasticity in drought tolerance traits. Genotypes with enhanced plasticity for drought tolerance (adaptive plasticity) will have a greater ability to persist and delay the onset of hydraulic failure. By examining populations from different climate-origins grown under contrasting soil water availability, we tested for genotype (G), environment (E) and genotype-by-environment (G × E) effects on traits that determine the time it takes for saplings to desiccate from stomatal closure to 88% loss of stem hydraulic conductance (time to hydraulic failure, THF). Specifically, we hypothesized that: (i) THF is dependent on a G × E interaction, with longer THF for warm, dry climate populations in response to chronic water deficit treatment compared with cool, wet populations, and (ii) hydraulic and allometric traits explain the observed patterns in THF. Corymbia calophylla saplings from two populations originating from contrasting climates (warm-dry or cool-wet) were grown under well-watered and chronic soil water deficit treatments in large containers. Hydraulic and allometric traits were measured and then saplings were dried-down to critical levels of drought stress to estimate THF. Significant plasticity was detected in the warm-dry population in response to water-deficit, with enhanced drought tolerance compared with the cool-wet population. Projected leaf area and total plant water storage showed treatment variation, and minimum conductance showed significant population differences driving longer THF in trees from warm-dry origins grown in water-limited conditions. Our findings contribute information on intraspecific variation in key drought traits, including hydraulic and allometric determinants of THF. It highlights the need to quantify adaptive capacity in populations of forest trees in climate change-type drought to improve predictions of forest die-back.

Despite a wealth of eco-physiological assessments of plant response to extreme drought, few studies have addressed the interactive effects of global change factors on traits driving mortality. To understand the interaction between hydraulic and carbon metabolic traits influencing tree mortality, which may be independently influenced by atmospheric [CO2] and temperature, we grew Eucalyptus sideroxylon A. Cunn. ex Woolls from seed in a full-factorial [CO2] (280, 400 and 640 μmol mol-1, Cp, Ca and Ce, respectively) and temperature (ambient and ambient +4 °C, Ta and Te, respectively) experiment. Prior to drought, growth across treatment combinations resulted in significant variation in physiological and morphological traits, including photosynthesis (Asat), respiration (Rd), stomatal conductance, carbohydrate storage, biomass and leaf area (LA). Ce increased Asat, LA and leaf carbohydrate concentration compared with Ca, while Cp generated the opposite response; Te reduced Rd. However, upon imposition of drought, Te hastened mortality (9 days sooner compared with Ta), while Ce significantly exacerbated drought stress when combined with Te. Across treatments, earlier time-to-mortality was mainly associated with lower (more negative) leaf water potential (Ψl) during the initial drought phase, along with higher water loss across the first 3 weeks of water limitation. Among many variables, Ψl was more important than carbon status in predicting time-to-mortality across treatments, yet leaf starch was associated with residual variation within treatments. These results highlight the need to carefully consider the integration, interaction and hierarchy of traits contributing to mortality, along with their responses to environmental drivers. Both morphological traits, which influence soil resource extraction, and physiological traits, which affect water-for-carbon exchange to the atmosphere, must be considered to adequately predict plant response to drought. Researchers have struggled with assessing the relative importance of hydraulic and carbon metabolic traits in determining mortality, yet an integrated trait, time-dependent framework provides considerable insight into the risk of death from drought for trees.

Short-term acclimation and long-term adaptation represent two ways in which forest trees can respond to changes in temperature. Yet, the relative contribution of thermal acclimation and adaptation to tree physiological responses to temperature remains poorly understood. Here, we grew two cool-origin and two warm-origin populations of a widespread broad-leaved evergreen tree species (Corymbia calophylla (Lindl.) K.D.Hill & L.A.S.Johnson) from a Mediterranean climate in southwestern Australia under two growth temperatures representative of the cool- and warm-edge of the species distribution. The populations selected from each thermal environment represented both high and low precipitation sites. We measured the short-term temperature response of leaf photosynthesis (A) and dark respiration (R), and attributed observed variation to acclimation, adaptation or the combination of both. We observed limited variation in the temperature optimum (Topt) of A between temperature treatments or among populations, suggesting little plasticity or genetic differentiation in the Topt of A. Yet, other aspects of the temperature response of A and R were dependent upon population and growth temperature. Under cooler growth temperatures, the population from the coolest, wettest environment had the lowest A (at 25 °C) among all four populations, but exhibited the highest A (at 25 °C) under warmer growth temperatures. Populations varied in R (at 20 °C) and the temperature sensitivity of R (i.e., Q10 or activation energy) under cool, but not warm growth temperatures. However, populations showed similar yet lower R (at 20 °C) and no differences in the temperature sensitivity of R under warmer growth temperatures. We conclude that C. calophylla populations from contrasting climates vary in physiological acclimation to temperature, which might influence how this ecologically important tree species and the forests of southwestern Australia respond to climate change.

The ability of plants to maintain an intact water transport system in leaves under drought conditions is intimately linked to survival and can been be seen as adaptive in shaping species climatic limits. Large differences in leaf hydraulic vulnerability to drought are known among species from contrasting climates, yet whether this trait varies among populations within a single species and, furthermore, whether it is altered by changes in growth conditions, remain unclear. We examined intraspecific variation in both leaf water transport capacity (Kleaf) and leaf hydraulic vulnerability to drought (P50leaf) among eight populations of Corymbia calophylla (R. Br.) K.D. Hill & L.A.S. Johnson (Myrtaceae) from both cool and warm climatic regions grown reciprocally under two temperature treatments representing the cool and warm edge of the species distribution. Kleaf did not vary between cool and warm-climate populations, nor was it affected by variable growth temperature. In contrast, population origin and growth temperature independently altered P50leaf. Using data pooled across growth temperatures, cool-climate populations showed significantly higher leaf hydraulic vulnerability (P50leaf = -3.55 ± 0.18 MPa) than warm-climate populations (P50leaf = -3.78 ± 0.08 MPa). Across populations, P50leaf decreased as population home-climate temperature increased, but was unrelated to rainfall and aridity. For populations from both cool and warm climatic regions, P50leaf was lower under the warmer growth conditions. These results provide evidence of trait plasticity in leaf hydraulic vulnerability to drought in response to variable growth temperature. Furthermore, they suggest that climate, and in particular temperature, may be a strong selective force in shaping intraspecific variation in leaf hydraulic vulnerability to drought.

Eucalypts are likely to play a critical role in the response of Australian forests to rising atmospheric CO2 concentration ([CO2]) and temperature. Although eucalypts are frequently phosphorus (P) limited in native soils, few studies have examined the main and interactive effects of P availability, [CO2] and temperature on eucalypt morphology, physiology and anatomy. To address this issue, we grew seedlings of Eucalyptus tereticornis Smith across its P-responsive range (6–500 mg kg−1) for 120 days under two [CO2] (ambient: 400 μmol mol−1 (Ca) and elevated: 640 μmol mol−1 (Ce)) and two temperature (ambient: 24/16 °C (Ta) and elevated: 28/20 °C (Te) day/night) treatments in a sunlit glasshouse. Seedlings were well-watered and supplied with otherwise non-limiting macro- and micro-nutrients. Increasing soil P supply increased growth responses to Ce and Te. At the highest P supplies, Ce increased total dry mass, leaf number and total leaf area by ~50%, and Te increased leaf number by ~40%. By contrast, Ce and Te had limited effects on seedling growth at the lowest P supply. Soil P supply did not consistently modify photosynthetic responses to Ce or Te. Overall, effects of Ce and Te on growth, physiological and anatomical responses of E. tereticornis seedlings were generally neutral or negative at low soil P supply, suggesting that native tree responses to future climates may be relatively small in native low-P soils in Australian forests.

Manipulative experiments have suggested that embolism-induced hydraulic impairment underpins widespread tree mortality during extreme drought, yet in situ evidence is rare. One month after drought-induced leaf and branch dieback was observed in field populations of Eucalyptus piperita Sm. in the Blue Mountains (Australia), we measured the level of native stem embolism and characterized the extent of leaf death in co-occurring dieback and healthy (non-dieback) trees. We found that canopy dieback-affected trees showed significantly higher levels of native embolism (26%) in tertiary order branchlets than healthy trees (11%). Furthermore, there was a significant positive correlation (R2 = 0.51) between the level of leaf death and the level of native embolism recorded in branchlets from dieback-affected trees. This retrospective study suggests that hydraulic failure was the primary mechanism of leaf and branch dieback in response to a natural drought event in the field. It also suggests that post-drought embolism refilling is minimal or absent in this species of eucalypt.

Intraspecific variation in biomass production responses to elevated atmospheric carbon dioxide (eCO2) could influence tree species' ecological and evolutionary responses to climate change. However, the physiological mechanisms underlying genotypic variation in responsiveness to eCO2 remain poorly understood. In this study, we grew 17 Eucalyptus camaldulensis Dehnh. subsp. camaldulensis genotypes (representing provenances from four different climates) under ambient atmospheric CO2 and eCO2. We tested whether genotype leaf-scale photosynthetic and whole-tree carbon (C) allocation responses to eCO2 were predictive of genotype biomass production responses to eCO2. Averaged across genotypes, growth at eCO2 increased in situ leaf net photosynthesis (Anet) (29%) and leaf starch concentrations (37%). Growth at eCO2 reduced the maximum carboxylation capacity of Rubisco (-4%) and leaf nitrogen per unit area (Narea, -6%), but Narea calculated on a total non-structural carbohydrate-free basis was similar between treatments. Growth at eCO2 also increased biomass production and altered C allocation by reducing leaf area ratio (-11%) and stem mass fraction (SMF, -9%), and increasing leaf mass area (18%) and leaf mass fraction (5%). Overall, we found few significant CO2 × provenance or CO2 × genotype (within provenance) interactions. However, genotypes that showed the largest increases in total dry mass at eCO2 had larger increases in root mass fraction (with larger decreases in SMF) and photosynthetic nitrogen-use efficiency (PNUE) with CO2 enrichment. These results indicate that genetic differences in PNUE and carbon sink utilization (in roots) are both important predictors of tree productivity responsiveness to eCO2.

Future climate regimes characterized by rising [CO 2 ], rising temperatures and associated droughts may differentially affect tree growth and physiology. However, the interactive effects of these three factors are complex because elevated [CO 2 ] and elevated temperature may generate differential physiological responses during drought. To date, the interactive effects of elevated [CO 2 ] and elevated temperature on drought-induced tree mortality remain poorly understood in gymnosperm species that differ in stomatal regulation strategies. Water relations and carbon dynamics were examined in two species with contrasting stomatal regulation strategies: Pinus radiata D. Don (relatively isohydric gymnosperm; regulating stomata to maintain leaf water potential above critical thresholds) and Callitris rhomboidea R. Br (relatively anisohydric gymnosperm; allowing leaf water potential to decline as the soil dries), to assess response to drought as a function of [CO 2 ] and temperature. Both species were grown in two [CO 2 ] ( C a (ambient, 400 μl l −1 ) and C e (elevated, 640 μl l −1 )) and two temperature ( T a (ambient) and T e (ambient +4 °C)) treatments in a sun-lit glasshouse under well-watered conditions. Drought plants were then exposed to a progressive drought until mortality. Prior to mortality, extensive xylem cavitation occurred in both species, but significant depletion of non-structural carbohydrates was not observed in either species. T e resulted in faster mortality in P. radiata , but it did not modify the time-to-mortality in C. rhomboidea. C e did not delay the time-to-mortality in either species under drought or T e treatments. In summary, elevated temperature (+4 °C) had greater influence than elevated [CO 2 ] (+240 μl l −1 ) on drought responses of the two studied gymnosperm species, while stomatal regulation strategies did not generally affect the relative contributions of hydraulic failure and carbohydrate depletion to mortality under severe drought.

Total daily water use is a key factor influencing the growth of many terrestrial plants, and reflects both day-time and nocturnal water fluxes. However, while nocturnal sap flow (En) and stomatal conductance (gs,n) have been reported across a range of species, ecosystems and microclimatic conditions, the regulation of these fluxes remains poorly understood. Here, we present a framework describing the role of abiotic and biotic factors in regulating En and gs,n highlighting recent developments in this field. Across ecosystems, En and gs,n generally increased with increasing soil water content and vapor pressure deficit, but the interactive effects of these factors and the potential roles of wind speed and other abiotic factors remain unclear. On average, gs,n and En are higher in broad-leaved compared with needle-leaved plants, in C3 compared with C4 plants, and in tropical compared with temperate species. We discuss the impacts of leaf age, elevated [CO2] and refilling of capacitance on night-time water loss, and how nocturnal gs,n may be included in vegetation models. Younger leaves may have higher gs,n than older leaves. Embolism refilling and recharge of capacitance may affect sap flow such that total plant water loss at night may be less than estimated solely from En measurements. Our estimates of gs,n for typical plant functional types, based on the published literature, suggest that nocturnal water loss may be a significant fraction (10-25%) of total daily water loss. Counter-intuitively, elevated [CO2] may increase nocturnal water loss. Assumptions in process-based ecophysiological models and dynamic global vegetation models that gs is zero when solar radiation is zero are likely to be incorrect. Consequently, failure to adequately consider nocturnal water loss may lead to substantial under-estimation of total plant water use and inaccurate estimation of ecosystem level water balance.

Nocturnal water flux has been observed in trees under a variety of environmental conditions and can be a significant contributor to diel canopy water flux. Elevated atmospheric CO(2) (elevated [CO(2)]) can have an important effect on day-time plant water fluxes, but it is not known whether it also affects nocturnal water fluxes. We examined the effects of elevated [CO(2)] on nocturnal water flux of field-grown Eucalyptus saligna trees using sap flux through the tree stem expressed on a sapwood area (J(s)) and leaf area (E(t)) basis. After 19 months growth under well-watered conditions, drought was imposed by withholding water for 5 months in the summer, ending with a rain event that restored soil moisture. Reductions in J(s) and E(t) were observed during the severe drought period in the dry treatment under elevated [CO(2)], but not during moderate- and post-drought periods. Elevated [CO(2)] affected night-time sap flux density which included the stem recharge period, called 'total night flux' (19:00 to 05:00, J(s,r)), but not during the post-recharge period, which primarily consisted of canopy transpiration (23:00 to 05:00, J(s,c)). Elevated [CO(2)] wet (EW) trees exhibited higher J(s,r) than ambient [CO(2)] wet trees (AW) indicating greater water flux in elevated [CO(2)] under well-watered conditions. However, under drought conditions, elevated [CO(2)] dry (ED) trees exhibited significantly lower J(s,r) than ambient [CO(2)] dry trees (AD), indicating less water flux during stem recharge under elevated [CO(2)]. J(s,c) did not differ between ambient and elevated [CO(2)]. Vapour pressure deficit (D) was clearly the major influence on night-time sap flux. D was positively correlated with J(s,r) and had its greatest impact on J(s,r) at high D in ambient [CO(2)]. Our results suggest that elevated [CO(2)] may reduce night-time water flux in E. saligna when soil water content is low and D is high. While elevated [CO(2)] affected J(s,r), it did not affect day-time water flux in wet soil, suggesting that the responses of J(s,r) to environmental factors cannot be directly inferred from day-time patterns. Changes in J(s,r) are likely to influence pre-dawn leaf water potential, and plant responses to water stress. Nocturnal fluxes are clearly important for predicting effects of climate change on forest physiology and hydrology.

Elevated atmospheric [CO(2)] (eC(a)) often decreases stomatal conductance, which may delay the start of drought, as well as alleviate the effect of dry soil on plant water use and carbon uptake. We studied the interaction between drought and eC(a) in a whole-tree chamber experiment with Eucalyptus saligna. Trees were grown for 18 months in their C(a) treatments before a 4-month dry-down. Trees grown in eC(a) were smaller than those grown in ambient C(a) (aC(a)) due to an early growth setback that was maintained throughout the duration of the experiment. Pre-dawn leaf water potentials were not different between C(a) treatments, but were lower in the drought treatment than the irrigated control. Counter to expectations, the drought treatment caused a larger reduction in canopy-average transpiration rates for trees in the eC(a) treatment compared with aC(a). Total tree transpiration over the dry-down was positively correlated with the decrease in soil water storage, measured in the top 1.5 m, over the drying cycle; however, we could not close the water budget especially for the larger trees, suggesting soil water uptake below 1.5 m depth. Using neutron probe soil water measurements, we estimated fractional water uptake to a depth of 4.5 m and found that larger trees were able to extract more water from deep soil layers. These results highlight the interaction between rooting depth and response of tree water use to drought. The responses of tree water use to eC(a) involve interactions between tree size, root distribution and soil moisture availability that may override the expected direct effects of eC(a). It is essential that these interactions be considered when interpreting experimental results.

Plants often exhibit proportionately larger photosynthetic responses to the transition from glacial to modern [CO(2)] than from modern to future [CO(2)]. Although this pattern may reflect increased nutrient demand with increasing [CO(2)], few studies have examined the role of nutrient supply in regulating responses to the range of [CO(2)] from glacial to future [CO(2)]. In this study, we examined the effects of P supply (0.004-0.5 mM) on photosynthetic responses of Populus deltoides (cottonwood) seedlings to glacial (200 micromol mol(-1)), modern (350 µmol mol(-1)) and future (700 micromol mol(-1)) [CO(2)]. The A(sat) (light-saturated net photosynthetic rates at the growth [CO(2)]) response to future [CO(2)] decreased with decreasing P supply such that there was no response at the lowest P supply. However, P supply did not affect A(sat) responses to an increase from glacial to modern [CO(2)]. Photosynthetic capacity [e.g., final rubisco activity, apparent, maximal Rubisco-limited rate of photosynthesis (V(cmax)), apparent, maximal electron transport-limited rate of photosynthesis (J(max))], stomatal conductance (g(s)) and leaf P generally increased with increasing P supply but decreased with increasing [CO(2)]. Measures of carbohydrate sink capacity (e.g., leaf mass per unit leaf area, leaf starch) increased with both increasing P supply and increasing [CO(2)]. Changes in V(cmax) and g(s) together accounted for 78% of the variation in A(sat) among [CO(2)] and P treatments, suggesting significant biochemical and stomatal controls on photosynthesis. However, A(sat) responses to increasing [CO(2)] did not reflect the changes in the carbohydrate sink capacity. These results have important implications because low P already constrains responses to increasing [CO(2)] in many ecosystems, and our results suggest that the P demand will increasingly affect A(sat) in cottonwood as [CO(2)] continues to increase.

To determine the effect of growth under elevated CO(2) partial pressures (pCO(2)) on photosynthetic electron transport and photoprotective energy dissipation, we examined light-saturated net photosynthetic CO(2) assimilation (A(sat)), the capacity for photosynthetic O(2) evolution, chlorophyll fluorescence emission and the pigment composition of upper-canopy loblolly pine needles in the eighth year of exposure to elevated pCO(2) (20 Pa above ambient) at the free-air CO(2) enrichment facility in the Duke Forest. During the summer growing season, A(sat) was 50% higher in current-year needles and 24% higher in year-old needles in elevated pCO(2) in comparison with needles of the same age cohort in ambient pCO(2). Thus, photosynthetic down-regulation at elevated pCO(2) was observed in the summer in year-old needles. In the winter, A(sat) was not significantly affected by growth pCO(2). Reductions in A(sat), the capacity for photosynthetic O(2) evolution and photosystem II (PSII) efficiency in the light-acclimated and fully-oxidized states were observed in the winter when compared to summer. Growth at elevated pCO(2) had no significant effect on the capacity for photosynthetic O(2) evolution, PSII efficiencies in the light-acclimated and fully-oxidized states, chlorophyll content or the size and conversion state of the xanthophyll cycle, regardless of season or needle age cohort. Therefore, we observed no evidence that photosynthetic electron transport or photoprotective energy dissipation responded to compensate for the effects of elevated pCO(2) on Calvin cycle activity.

Temperature plays a critical role in the regulation of respiration rates and is often used to scale measurements of respiration to the stand-level and calculate annual respiratory fluxes. Previous studies have indicated that failure to consider temperature gradients between sun-exposed stems and branches in the crown and shaded lower stems may result in errors when deriving stand-level estimates of stem CO(2) efflux. We measured vertical gradients in sapwood temperature in a mature lowland podocarp rain forest in New Zealand to: (1) estimate the effects of within-stem temperature variation on the vertical distribution of stem CO(2) efflux; and (2) use these findings to estimate stand-level stem CO(2) efflux for this forest. Large within-stem gradients in sapwood temperature (1.6 +/- 0.1 to 6.0 +/- 0.5 degrees C) were observed. However, these gradients did not significantly influence the stand-level estimate of stem CO(2) efflux in this forest (536 +/- 42 mol CO(2) ha(-1) day(-1)) or the vertical distribution of stem CO(2) efflux, because of the opposing effects of daytime warming and nighttime cooling on CO(2) efflux in the canopy, and the small fraction of the woody biomass in the crowns of forest trees. Our findings suggest that detailed measurements of within-stand temperature gradients are unlikely to greatly improve the accuracy of tree- or stand-level estimates of stem CO(2) efflux.

We sought to quantify changes in tree species composition, forest structure and aboveground forest biomass (AGB) over 76 years (1930-2006) in the deciduous Black Rock Forest in southeastern New York, USA. We used data from periodic forest inventories, published floras and a set of eight long-term plots, along with species-specific allometric equations to estimate AGB and carbon content. Between the early 1930s and 2000, three species were extirpated from the forest (American elm (Ulmus americana L.), paper birch (Betula papyrifera Marsh.) and black spruce (Picea mariana (nigra) (Mill.) BSP)) and seven species invaded the forest (non-natives tree-of-heaven (Ailanthus altissima (Mill.) Swingle) and white poplar (Populus alba L.) and native, generally southerly distributed, southern catalpa (Catalpa bignonioides Walt.), cockspur hawthorn (Crataegus crus-galli L.), red mulberry (Morus rubra L.), eastern cottonwood (Populus deltoides Bartr.) and slippery elm (Ulmus rubra Muhl.)). Forest canopy was dominated by red oak and chestnut oak, but the understory tree community changed substantially from mixed oak-maple to red maple-black birch. Density decreased from an average of 1500 to 735 trees ha(-1), whereas basal area doubled from less than 15 m(2) ha(-1) to almost 30 m(2) ha(-1) by 2000. Forest-wide mean AGB from inventory data increased from about 71 Mg ha(-1) in 1930 to about 145 Mg ha(-1) in 1985, and mean AGB on the long-term plots increased from 75 Mg ha(-1) in 1936 to 218 Mg ha(-1) in 1998. Over 76 years, red oak (Quercus rubra L.) canopy trees stored carbon at about twice the rate of similar-sized canopy trees of other species. However, there has been a significant loss of live tree biomass as a result of canopy tree mortality since 1999. Important constraints on long-term biomass increment have included insect outbreaks and droughts.

We assessed the relative limitations to photosynthesis imposed by stomatal and non-stomatal processes in Dacrydium cupressinum Lamb. (Podocarpaceae), which is the dominant species in a native, mixed conifer-broad-leaved rainforest in New Zealand. For comparison, we included three co-occurring broad-leaved tree species (Meterosideros umbellata Cav. (Myrtaceae), Weinmannia racemosa L.f. (Cunoniaceae) and Quintinia acutifolia Kirk (Escalloniaceae)) that differ in phylogeny and in leaf morphology from D. cupressinum. We found that low foliage phosphorus content on an area basis (P(a)) limited light-saturated photosynthesis on an area basis (A(sat)) in Q. acutifolia. Depth in the canopy did not generally affect A(sat) or the relative limitations to A(sat) because of stomatal and non-stomatal constraints, despite reductions in the ratio of foliage mass to area, foliar nitrogen on an area basis (N(a)) and P(a) with depth in the canopy. In the canopy-dominant conifer D. cupressinum, A(sat) was low, consistent with low values of the maximum rate of ribulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco) carboxylation (V(cmax)). In comparison, the A(sat) response of the three broad-leaved tree species was quite variable. Although A(sat) was high in the canopy-dominant M. umbellata, it was low in the sub-canopy trees W. racemosa and Q. acutifolia. Relative stomatal limitation to photosynthesis was more pronounced in W. racemosa (40%) than in the other three species (28-33%). Despite differences in degree, non-stomatal limitation to A(sat) predominated in all tree species.

Within-leaf variations in cell size, mitochondrial numbers and dark respiration rates were compared in the most recently expanded tip, the mid-section and base of needles of Pinus radiata D. Don trees grown for 4 years in open-top chambers at ambient (36 Pa) or elevated (65 Pa) carbon dioxide partial pressure (p(CO2)a). Mitochondrial numbers and respiratory activity varied along the length of the needle, with the highest number of mitochondria per unit cytoplasm and the highest rate of respiration per unit leaf area at the base of the needle. Regardless of the location of the cells (tip, middle or basal sections), needles collected from trees grown in elevated p(CO2)a had nearly twice the number of mitochondria per unit cytoplasm as those grown in ambient p(CO2)a. This stimulation of mitochondrial density by growth at elevated p(CO2)a was greater at the tip of the needle (2.7 times more mitochondria than in needles grown in ambient CO2) than at the base of the needle (1.7 times). The mean size of individual mitochondria was unaffected either by growth at elevated p(CO2)a or by position along the needle. Tree growth at elevated p(CO2)a had a variable effect on respiration per unit leaf area, significantly increasing respiration in the tip of the needles (+25%) and decreasing respiration at the mid-section and base of the needles (-14% and -25%, respectively). Although a simple relationship between respiration per unit leaf area and mitochondrial number per unit cytoplasm was found within each CO2 treatment, the variable effect of growth at elevated p(CO2)a on respiration along the length of the needles indicates that a more complex relationship must determine the association between structure and function in these needles.

Gas exchange, growth, water transport and carbon (C) metabolism diminish during drought according to their respective sensitivities to declining water status. The timing of this sequence of declining physiological functions may determine how water and C relations compromise plant survival. In this paper, we test the hypothesis that the degree of asynchrony between declining C supply (photosynthesis) and C demand (growth and respiration) determines the rate and magnitude of changes in whole-plant non-structural carbohydrates (NSC) during drought. Two complementary experiments using two tree species (Eucalyptus globulus Labill. and Pinus radiata D. Don) with contrasting drought response strategies were performed to (i) assess changes in radial stem growth, transpiration, leaf water potential and gas exchange in response to chronic drought, and (ii) evaluate the concomitant impacts of these drought responses on the temporal patterns of NSC during terminal drought. The three distinct phases of water stress were delineated by thresholds of growth cessation and stomatal closure that defined the 'carbon safety margin' (i.e., the difference between leaf water potential when growth is zero and leaf water potential when net photosynthesis is zero). A wider C safety margin in E. globulus was defined by an earlier cessation of growth relative to photosynthesis that reduced the demand for NSC while maintaining C acquisition. By contrast, the narrower C safety margin in P. radiata was characterized by a synchronous decline in growth and photosynthesis, whereby growth continued under a declining supply of NSC from photosynthesis. The narrower C safety margin in P. radiata was associated with declines in starch concentrations after ∼ 90 days of chronic drought and significant depletion of starch in all organs at mortality. The observed divergence in the sensitivity of drought responses is indicative of a potential trade-off between maintaining hydraulic safety and adequate C availability.

High nocturnal transpiration rates (5-15% of total water loss in terrestrial plants) may be adaptive under limited fertility, by increasing nutrient uptake or transport via transpiration-induced mass flow, but the response of stomata in the dark to environmental variables is poorly understood. Here we tested the impact of soil phosphorous (P) concentration, atmospheric CO2 concentration and air temperature on stomatal conductance (gs) during early and late periods in the night, as well as at midday in naturally, sun-lit glasshouse-grown Eucalyptus tereticornis Sm. seedlings. Soil P was the main driver of nocturnal gs, which was consistently higher in low soil P (37.3-79.9 mmol m(-2) s(-1)) than in high soil P (17.7-49.3 mmol m(-2)(-1)). Elevated temperature had only a marginal (P = 0.07) effect on gs early in the night (gs decreased from 34.7 to 25.8 mmol m(-2) s(-1) with an increase in temperature of 4 °C). The effect of CO2 depended on its interaction with temperature. Stomatal conductance responses to soil P were apparently driven by indirect effects of soil P on plant anatomy, since gs was significantly and negatively correlated with wood density. However, the relationship of gs with environmental factors became weaker late in the night, relative to early in the night, likely due to apparent endogenous processes; gs late in the night was two times larger than gs observed early in the night. Time-dependent controls over nocturnal gs suggest that daytime stomatal models may not apply during the night, and that different types of regulation may occur even within a single night. We conclude that the enhancement of nocturnal gs under low soil P availability is unlikely to be adaptive in our species because of the relatively small amount of transpiration-induced mass flow that can be achieved through rates of nocturnal water loss (3-6% of daytime mass flow).

Elevated [CO2] and temperature may alter the drought responses of tree seedling growth, photosynthesis, respiration and total non-structural carbohydrate (TNC) status depending on drought intensity and duration. Few studies have addressed these important climatic interactions or their consequences. We grew Eucalyptus globulus Labill. seedlings in two [CO2] concentrations (400 and 640 μl l(-1)) and two temperatures (28/17 and 32/21 °C) (day/night) in a sun-lit glasshouse, and grew them in well-watered conditions or exposed them to two drought treatments having undergone different previous water conditions (i.e., rewatered drought and sustained drought). Progressive drought in both drought treatments led to similar limitations in growth, photosynthesis and respiration, but reductions in TNC concentration were not observed. Elevated [CO2] ameliorated the impact of the drought during the moderate drought phase (i.e., Day 63 to Day 79) by increasing photosynthesis and enhancing leaf and whole-plant TNC content. In contrast, elevated temperature exacerbated the impact of the drought during the moderate drought phase by reducing photosynthesis, increasing leaf respiration and decreasing whole-plant TNC content. Extreme drought (i.e., Day 79 to Day 103) eliminated [CO2] and temperature effects on plant growth, photosynthesis and respiration. The combined effects of elevated [CO2] and elevated temperature on moderate drought stressed seedlings were reduced with progressive drought, with no sustained effects on growth despite greater whole-plant TNC content.

Climate change may alter forest composition by differentially affecting the responses of faster- and slower-growing tree species to drought. However, the combined effects of rising atmospheric CO2 concentration ([CO2]) and temperature on drought responses of trees are poorly understood. Here, we examined interactive effects of temperature (ambient, ambient + °C) and [CO2] (290, 400 and 650mu;l l(-1)) on drought responses of Eucalyptus saligna Sm. (faster-growing) and E. sideroxylon A. Cunn. ex Woolls (slower-growing) seedlings. Drought was imposed via a controlled reduction in soil water over 1-2 weeks, re-watering seedlings when leaves visibly wilted. In ambient temperature, the effect of drought on the light-saturated net photosynthetic rate (Asat) in E. saligna decreased as [CO2] increased from pre-industrial to future concentrations, but rising [CO2] did not affect the response in Eucalyptus sideroxylon. In contrast, elevated temperature exacerbated the effect of drought in reducing Asat in both species. The drought response of Asat reflected changes in stomatal conductance (gs) associated with species and treatment differences in (i) utilization of soil moisture and (ii) leaf area ratio (leaf area per unit plant dry mass). Across [CO2] and temperature treatments, E. saligna wilted at higher soil water potentials compared with E. sideroxylon. Photosynthetic recovery from drought was 90% complete 2 days following re-watering across all species and treatments. Our results suggest that E. saligna (faster-growing) seedlings are more susceptible to drought than E. sideroxylon (slower-growing) seedlings. The greater susceptibility to drought of E. saligna reflected faster drawdown of soil moisture, associated with more leaf area and leaf area ratio, and the ability of E. sideroxylon to maintain higher gs at a given soil moisture. Inclusion of a pre-industrial [CO2] treatment allowed us to conclude that susceptibility of these species to short-term drought under past and future climates may be regulated by the same mechanisms. Further, the beneficial effects of rising [CO2] and deleterious effects of elevated temperature on seedling response to drought were generally offsetting, suggesting susceptibility of seedlings of these species to short-term drought in future climates that is similar to pre-industrial and current climate conditions.

To detect seasonal and long-term differences in growth and photosynthesis of loblolly pine (Pinus taeda L.) exposed to elevated CO(2) under ambient conditions of precipitation, light, temperature and nutrient availability, seedlings were planted in soil representative of an early, abandoned agricultural field and maintained for 19 months in the field either in open-top chambers providing one of three atmospheric CO(2) partial pressures (ambient, ambient +15 Pa, and ambient +30 Pa) or in unchambered control plots. An early and positive response to elevated CO(2) substantially increased total plant biomass. Peak differences in relative biomass enhancement occurred after 11 months of CO(2) treatment when biomass of plants grown at +15 and +30 Pa CO(2) was 111 and 233% greater, respectively, than that of plants grown at ambient CO(2). After 19 months, there was no significant difference in biomass between +15 Pa CO(2)-treated plants and ambient CO(2)-treated plants, whereas biomass of +30 Pa CO(2)-treated plants was 111% greater than that of ambient CO(2)-treated plants. Enhanced rates of leaf-level photosynthesis were maintained in plants in the elevated CO(2) treatments throughout the 19-month exposure period despite reductions in both leaf N concentration and ribulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco) activity during the first 11 months of CO(2) exposure. Reductions in Rubisco activity indicated photosynthetic adjustment to elevated CO(2), but Rubisco-mediated control of photosynthesis was small. Seasonal shifts in sink strength affected photosynthetic rates, greatly magnifying the positive effects of elevated CO(2) on photosynthesis during periods of rapid plant growth. Greater carbon assimilation by the whole plant accelerated plant development and thereby stimulated new sinks for carbon through increased plant biomass, secondary branching and new leaf production. We conclude that elevated CO(2) will enhance photosynthesis and biomass accumulation in loblolly pine seedlings under high nutrient conditions; however, reductions over time in the relative biomass response of plants to elevated CO(2) complicate predictions of the eventual magnitude of carbon storage in this species under future CO(2) conditions.

Trees adapted to mesic and xeric habits may differ in a suite of physiological responses that affect leaf-level carbon balance, including the relationship between photosynthesis (A) and respiration at night (R(n)). Understanding the factors that regulate physiological function in mesic and xeric species is critical for predicting changes in growth and distribution under changing climates. In this study, we examined the relationship between A and R(n), and leaf traits that may regulate A and R(n), in six Eucalyptus species native to mesic or xeric ecosystems, during two 24-h cycles in a common garden under high soil moisture. Peak A and R(n) generally were higher in xeric compared with mesic species. Across species, A and R(n) covaried, correlated with leaf mass per area, leaf N per unit area and daytime soluble sugar accumulation. A also covaried with g(s), which accounted for 93% of the variation in A within species. These results suggest that A and R(n) in these six Eucalyptus species were linked through leaf N and carbohydrates. Further, the relationship between A and R(n) across species suggests that differences in this relationship between mesic and xeric Eucalyptus species in their native habitats may be largely driven by environmental factors rather than inter-specific genetic variation.

Forest canopies exchange a large part of the mass and energy between the earth and the atmosphere. The processes that regulate these exchanges have been of interest to scientists from a diverse range of disciplines for a long time. The International Union of Forest Research Organizations (IUFRO) Canopy Processes Working Group provides a forum for these scientists to explore canopy processes at scales ranging from the leaf to the ecosystem. Given the changes in climate that are being experienced in response to rising [CO(2)], there is a need to understand how forest canopy processes respond to altered environments. Globally, native and managed forests represent the largest terrestrial biome and, in wood and soils, the largest terrestrial stores of carbon. Changing climates have significant implications for carbon storage in forests, as well as their water use, species diversity and management. In order to address these issues, the Canopy Processes Working Group held a travelling workshop in south-east Australia during October 2010 to examine the impact of changing climates on forest canopies, highlighting knowledge gaps and developing new research directions.

Nearly 30 years ago, Whitehead and Jarvis and Whitehead et al. postulated an elegant mechanistic explanation for the observed relationship between tree hydraulic structure and function, hypothesizing that structural adjustments promote physiological homeostasis. To date, this framework has been nearly completely overlooked with regard to varying atmospheric carbon dioxide ([CO(2)]). Here, we evaluated Whitehead's hypothesis of leaf water potential (Ψ(l)) homeostasis in faster-growing (Eucalyptus saligna) and slower-growing (Eucalyptus sideroxylon) tree saplings grown under three [CO(2)] (pre-industrial, current and future) and two temperature (ambient and ambient + 4°C) treatments. We tested for relationships between physiological (stomatal conductance and Ψ(l)) and structural (leaf and sapwood areas (A(l), A(s)), height (h), xylem conductivity (k(s))) plant variables as a function of the [CO(2)] and temperature treatments to assess whether structural variables adjusted to maintain physiological homeostasis. Structural components (A(l), A(s), h) generally increased with [CO(2)] or temperature, while g(s) was negatively correlated with [CO(2)]. Contrary to Whitehead's hypothesis, Ψ(l) did not exhibit homeostasis in either species; elevated temperatures were associated with more negative Ψ(l) in faster-growing E. saligna, and less negative Ψ(l) in slower-growing E. sideroxylon. Moreover, individual structural variables were generally uncorrelated with Ψ(l). However, across both species, the integrated hydraulic property of leaf specific hydraulic conductance (K(l)) was positively correlated with an independent calculation of K(l) determined exclusively from leaf physiological variables. These results suggest that physiological homeostasis may not apply to saplings exposed to global change drivers including [CO(2)] and temperature. Nevertheless, Whitehead et al.'s formulation identified K(l) as a sensitive measure of plant structural-physiological co-variation across species.

Mesophyll conductance, g(m), was estimated from measurements of stomatal conductance to carbon dioxide transfer, g(s), photosynthesis, A, and chlorophyll fluorescence for Year 0 (current-year) and Year 1 (1-year-old) fully sunlit leaves from short (2 m tall, 10-year-old) and tall (15 m tall, 120-year-old) Nothofagus solandrii var. cliffortiodes trees growing in adjacent stands. Rates of photosynthesis at saturating irradiance and ambient CO(2) partial pressure, A(satQ), were 25% lower and maximum rates of carboxylation, V(cmax), were 44% lower in Year 1 leaves compared with Year 0 leaves across both tree sizes. Although g(s) and g(m) were not significantly different between Year 0 and Year 1 leaves and g(s) was not significantly different between tree heights, g(m) was significantly (19%) lower for leaves on tall trees compared with leaves on short trees. Overall, V(cmax) was 60% higher when expressed on the basis of CO(2) partial pressure at the chloroplasts, C(c), compared with V(cmax) on the basis of intercellular CO(2) partial pressure, C(i), but this varied with leaf age and tree size. To interpret the relative stomatal and mesophyll limitations to photosynthesis, we used a model of carbon isotopic composition for whole leaves incorporating g(m) effects to generate a surface of 'operating values' of A over the growing season for all leaf classes. Our analysis showed that A was slightly higher for leaves on short compared with tall trees, but lower g(m) apparently reduced actual A substantially compared with A(satQ). Our findings showed that lower rates of photosynthesis in Year 1 leaves compared with Year 0 leaves were attributable more to increased biochemical limitation to photosynthesis in Year 1 leaves than differences in g(m). However, lower A in leaves on tall trees compared with those on short trees could be attributed in part to lower g(m) and higher stomatal, L(s), and mesophyll, L(m), limitations to photosynthesis, consistent with steeper hydraulic gradients in tall trees.

Eastern dwarf mistletoe (Arceuthobium pusillum Peck) is a hemiparasitic angiosperm that infects white spruce (Picea glauca (Moench) Voss) and red spruce (P. rubens Sarg.) in northeastern North America. The effects of mistletoe infection differ substantially between white and red spruce, with white spruce suffering greater infection-induced mortality. In the present study, we sought to determine the role that species-specific differences in needle-scale responses to parasitism may play in the observed differences in the effect of infection on host tree health. Based on the measurements made, the most apparent effect of parasitism was a reduction in needle size distal to infections. The magnitude of this reduction was greater in white spruce than in red spruce. Eastern dwarf mistletoe was a sink for host photosynthate in red spruce and white spruce; however, there were no adjustments in needle photosynthetic capacities in either host to accommodate the added sink demands of the parasite. Needle total nonstructural carbohydrate concentrations (TNC) were also unaltered by infection. Red spruce needles had higher TNC concentrations despite having lower overall photosynthetic capacities, suggesting that red spruce may be more sink limited and therefore better able to satisfy the added sink demands of parasitic infection. However, if carbon availability limits the growth of the mistletoe, one may expect that the extent of the parasitic infection would be greater in red spruce. Yet in the field, the extent of infection is generally greater in white spruce. Taken together, these results suggest that dwarf mistletoe may not substantially perturb the carbon balance of either host spruce species and that species-specific differences in needle-scale responses to the parasite cannot explain the contrasting effects of infection on white spruce and red spruce.

Photosynthesis of tree seedlings is generally enhanced during short-term exposure to elevated atmospheric CO(2), but longer-term photosynthetic responses are often more variable because they are affected by morphological, biochemical and physiological feedback mechanisms that regulate carbon assimilation to meet sink demand. To examine biochemical and morphological factors that might regulate the long-term photosynthetic response of field-grown trees to elevated CO(2), we grew ponderosa pine (Pinus ponderosa Dougl. ex Laws.) trees in open-top chambers for six years in native soil at ambient CO(2) (35 Pa) and elevated CO(2) (70 Pa) at a site near Placerville, CA. Trees were well watered and exposed to natural light and ambient temperature. At the end of the sixth growing season at elevated CO(2), net photosynthesis was enhanced 53%, despite reductions in photosynthetic capacity. The positive net photosynthetic response to elevated CO(2) reflected greater relative increases in Rubisco sensitivity compared with the decreases resulting from biochemical adjustments. Analyses of net photosynthetic rate versus internal CO(2) partial pressure curves indicated that reductions in photosynthetic capacity in response to elevated CO(2) were the result of significant reductions in maximum photosynthetic rate (20%), Rubisco carboxylation capacity (36%), and electron transport capacity (21%). Decreased photosynthetic capacity was accompanied by reductions in various photosynthetic components, including total chlorophyll (24%), Rubisco protein content (38%), and mass-based leaf nitrogen concentration (14%). Net photosynthesis was unaffected by morphological adjustments because there was no change in leaf mass per unit area at elevated CO(2). An apparent positive response of photosynthetic adjustment in the elevated CO(2) treatment was the redistribution of N within the photosynthetic system to balance Rubisco carboxylation and electron transport capacities. We conclude that trees, without apparent limitations to root growth, may exhibit photosynthetic adjustment responses in the field after long-term exposure to elevated CO(2).

Despite its recent expansion in eastern US forests, red maple (Acer rubrum L.) generally exhibits a low leaf photosynthetic rate, leaf mass per unit area (LMA) and leaf nitrogen concentration ([N]) relative to co-occurring oaks (Quercus spp.). To evaluate these differences from the perspective of leaf energy investment, we compared leaf construction cost (CC) and leaf maintenance cost (MC) with leaf photosynthetic rate at saturating photon flux density and ambient CO 2 partial pressure (A max ) in red maple and co-occurring red oak (Quercus rubra L.) and chestnut oak (Quercus prinus L.). We also examined relationships among leaf physiological, biochemical and structural characteristics of upper-canopy leaves of these three species at lower (wetter) and upper (drier) elevation sites of a watershed in the Black Rock Forest, Cornwall, NY, USA. Although A max , leaf [N], leaf carbon concentration ([C]) and LMA were significantly less in red maple than in either oak species at both sites, CC per unit leaf area of red maple was 28.2 and 35.4% less than that of red oak at the lower and upper site, respectively, and 38.8 and 32% less than that of chestnut oak at the lower and upper site, respectively. Leaf MC per unit leaf area, which was positively associated with leaf CC (r 2 = 0.95), was also significantly lower in red maple than in either oak species at both sites. When expressed per unit leaf area, A max was positively correlated with both CC (r 2 = 0.65) and MC (r 2 = 0.59). The cost/benefit ratio of CC/A max of red maple was significantly less than that of chestnut oak at the lower site, however, CC/A max did not exhibit any significant interspecific differences at the upper site. Expressed per unit leaf area, CC was correlated positively with LMA (r 2 = 0.90), leaf [N] (r 2 = 0.97), and leaf [C] (r 2 = 0.89), and negatively correlated with leaf molar carbon to nitrogen ratio (r 2 = 0.92). Combined with red maple's general success in many oak-dominated forests, our findings suggest that reduced leaf-level photosynthetic capacity and related leaf characteristics in red maple are partially balanced by lower energy and resource requirements for leaf biomass construction and maintenance, which could enhance the competitive success of this species.

Photosynthesis of tree seedlings is generally enhanced during short-term exposure to elevated atmospheric CO2 partial pressure, but longer-term studies often indicate some degree of photosynthetic adjustment. We present physiological and biochemical evidence to explain observed long-term photosynthetic responses to elevated CO2 partial pressure as influenced by needle age and canopy position. We grew Pinus radiata D. Don. trees in open-top chambers for 5 years in sandy soil at ambient (36 Pa) and elevated (65 Pa) CO2 partial pressures. The trees were well watered and exposed to natural light and ambient temperature. In the fourth year of CO2 exposure (fall 1997), when foliage growth had ceased for the year, photosynthetic down-regulation was observed in 1-year-old needles, but not in current-year needles, suggesting a reduction in carbohydrate sink strength as a result of increasing needle age (Turnbull et al. 1998). In 5-year-old trees (spring 1997), when foliage expansion was occurring, photosynthetic down-regulation was not observed, reflecting significantly large sinks for carbohydrates throughout the tree. Net photosynthesis was stimulated by 79% in trees growing in elevated CO2 partial pressure, but there was no significant effect on photosynthetic capacity or Rubisco activity and concentration. Current-year needles were more responsive to elevated CO2 partial pressure than 1-year-old needles, exhibiting larger relative increases in net photosynthesis to elevated CO2 partial pressure (98 versus 64%). Lower canopy and upper canopy leaves exhibited similar relative responses to growth in elevated CO2 partial pressure. However, needles in the upper canopy exhibited higher net photosynthesis, photosynthetic capacity, and Rubisco activity and concentration than needles in the lower canopy. Given that the ratio of mature to juvenile foliage mass in the canopy will increase as trees mature, we suggest that trees may become less responsive to elevated CO2 partial pressure with increasing age. We conclude that tree response to elevated CO2 partial pressure is based primarily on sink strength and not on the duration of exposure.

We measured responses of leaf respiration to temperature and leaf characteristics in three deciduous tree species (Quercus rubra L., Quercus prinus L. and Acer rubrum L.) at two sites differing in water availability within a single catchment in the Black Rock Forest, New York. The response of respiration to temperature differed significantly among the species. Acer rubrum displayed the smallest increase in respiration with increasing temperature. Corresponding Q(10) values ranged from 1.5 in A. rubrum to 2.1 in Q. prinus. Dark respiration at ambient air temperatures, expressed on a leaf area basis (Rarea), did not differ significantly between species, but it was significantly lower (P < 0.01) in trees at the wetter (lower) site than at the drier (upper) site (Q. rubra: 0.8 versus 1.1 micromol m(-2) s(-1); Q. prinus: 0.95 versus 1.2 micromol m(-2) s(-1)). In contrast, when expressed on a leaf mass basis (R(mass)), respiration rates were significantly higher (P < 0.01) in A. rubrum (12.5-14.6 micromol CO(2) kg(-1) s(-1)) than in Q. rubra (8.6-9.9 micromol CO(2) kg(-1) s(-1)) and Q. prinus (9.2-10.6 micromol CO(2) kg(-1) s(-1)) at both the lower and upper sites. Respiration on a nitrogen basis (R(N)) displayed a similar response to R(mass). The consistency in R(mass) and R(N) between sites indicates a strong coupling between factors influencing respiration and those affecting leaf characteristics. Finally, the relationships between dark respiration and A(max) differed between sites. Trees at the upper site had higher rates of leaf respiration and lower A(max) than trees at the lower site. This shift in the balance of carbon gain and loss clearly limits carbon acquisition by trees at sites of low water availability, particularly in the case of A. rubrum.