Small tropical forest trees have a greater capacity to adjust carbon metabolism to long-term drought than large canopy trees

The response of small understory trees to long-term drought is vital in determining the future composition, carbon stocks and dynamics of tropical forests. Long-term drought is, however, also likely to expose understory trees to increased light availability driven by drought-induced mortality. Relatively little is known about the potential for understory trees to adjust their physiology to both decreasing water and increasing light availability. We analysed data on maximum photosynthetic capacity (J_max, V_cmax), leaf respiration (R_leaf), leaf mass per area (LMA), leaf thickness and leaf nitrogen and phosphorus concentrations from 66 small trees across 12 common genera at the world's longest running tropical rainfall exclusion experiment and compared responses to those from 61 surviving canopy trees. Small trees increased J_max, V_cmax, R_leaf and LMA (71, 29, 32, 15% respectively) in response to the drought treatment, but leaf thickness and leaf nutrient concentrations did not change. Small trees were significantly more responsive than large canopy trees to the drought treatment, suggesting greater phenotypic plasticity and resilience to prolonged drought, although differences among taxa were observed. Our results highlight that small tropical trees have greater capacity to respond to ecosystem level changes and have the potential to regenerate resilient forests following future droughts.     

Influence of nutrient versus water supply on hydraulic architecture and water balance in Pinus taeda

We investigated the hydraulic consequences of a major decrease in root‐to‐leaf area ratio (A_R:A_L) caused by nutrient amendments to 15‐year‐old Pinus taeda L. stands on sandy soil. In theory, such a reduction in A_R:A_L should compromise the trees’ ability to extract water from drying sand. Under equally high soil moisture, canopy stomatal conductance (G_S) of fertilized trees (F) was 50% that of irrigated/fertilized trees (IF), irrigated trees (I), and untreated control trees (C). As predicted from theory, F trees also decreased their stomatal sensitivity to vapour pressure deficit by 50%. The lower G_S in F was associated with 50% reduction in leaf‐specific hydraulic conductance (K_L) compared with other treatments. The lower K_L in F was in turn a result of a higher leaf area per sapwood area and a lower specific conductivity (conducting efficiency) of the plant and its root xylem. The root xylem of F trees was also 50% more resistant to cavitation than the other treatments. A transport model predicted that the lower A_R:A_L in IF trees resulted in a considerably restricted ability to extract water during drought. However, this deficiency was not exposed because irrigation minimized drought. In contrast, the lower A_R:A_L in F trees caused only a limited restriction in water extraction during drought owing to the more cavitation resistant root xylem in this treatment. In both fertilized treatments, approximate safety margins from predicted hydraulic failure were minimal suggesting increased vulnerability to drought‐induced dieback compared with non‐fertilized trees. However, IF trees are likely to be so affected even under a mild drought if irrigation is withheld.     

Seasonal acclimation of leaf respiration in Eucalyptus saligna trees: impacts of elevated atmospheric CO2 and summer drought

Understanding the impacts of atmospheric [CO₂] and drought on leaf respiration (R) and its response to changes in temperature is critical to improve predictions of plant carbon‐exchange with the atmosphere, especially at higher temperatures. We quantified the effects of [CO₂]‐enrichment (+240 ppm) on seasonal shifts in the diel temperature response of R during a moderate summer drought in Eucalyptus saligna growing in whole‐tree chambers in SE Australia. Seasonal temperature acclimation of R was marked, as illustrated by: (1) a downward shift in daily temperature response curves of R in summer (relative to spring); (2)≈60% lower R measured at 20^oC (R₂₀) in summer compared with spring; and (3) homeostasis over 12 months of R measured at prevailing nighttime temperatures. R₂₀, measured during the day, was on average 30–40% higher under elevated [CO₂] compared with ambient [CO₂] across both watered and droughted trees. Drought reduced R₂₀ by≈30% in both [CO₂] treatments resulting in additive treatment effects. Although [CO₂] had no effect on seasonal acclimation, summer drought exacerbated the seasonal downward shift in temperature response curves of R. Overall, these results highlight the importance of seasonal acclimation of leaf R in trees grown under ambient‐ and elevated [CO₂] as well as under moderate drought. Hence, respiration rates may be overestimated if seasonal changes in temperature and drought are not considered when predicting future rates of forest net CO₂ exchange.     

Functional diversity of photosynthesis during drought in a model tropical rainforest – the contributions of leaf area, photosynthetic electron transport and stomatal conductance to reduction in net ecosystem carbon exchange

The tropical rainforest mesocosm within the Biosphere 2 Laboratory, a model system of some 110 species developed over 12 years under controlled environmental conditions, has been subjected to a series of comparable drought experiments during 2000–2002. In each study, the mesocosm was subjected to a 4–6 week drought, with well‐defined rainfall events before and after the treatment. Ecosystem CO₂ uptake rate (A_eco) declined 32% in response to the drought, with changes occurring within days and being reversible within weeks, even though the deeper soil layers did not become significantly drier and leaf‐level water status of most large trees was not greatly affected. The reduced A_eco during the drought reflected both morphological and physiological responses. It is estimated that the drought‐induced 32% reduction of A_eco has three principal components: (1) leaf fall increased two‐fold whereas leaf expansion growth of some canopy dominants declined to 60%, leading to a 10% decrease in foliage coverage of the canopy. This might be the main reason for the persistent reduction of A_eco after rewatering. (2) The maximum photosynthetic electron transport rate at high light intensities in remaining leaves was reduced to 71% for three of the four species measured, even though no chronic photo‐inhibition occurred. (3) Stomata closed, leading to a reduced ecosystem water conductance to water vapour (33% of pre‐drought values), which not only reduced ecosystem carbon uptake rate, but may also have implications for water and energy budgets of tropical ecosystems. Additionally, individual rainforest trees responded differently, expressing different levels of stress and stress avoiding mechanisms. This functional diversity renders the individual response heterogeneous and has fundamental implications to scale leaf level responses to ecosystem dynamics.     

Long‐term directional changes in upland Quercus forests throughout Oklahoma,USA

Questions: (1) How have the composition and structure of undisturbed upland Quercus forests changed over 50 years across a large region and moisture gradient; (2) What factors are associated with long‐term and broad‐scale changes in these forests? Location: Oklahoma, USA. Methods: We re‐sampled 30 forest stands originally sampled in the 1950s across a large geographical area and compared basal area, tree density, and sapling density between the sampling periods using paired t‐tests, CCA, and DCA. We examined vegetation dynamics in the context of drought indices compiled for the sample period. Results: Total and Quercus stellata basal area and tree density increased, but Q. stellata and Q. marilandica sapling density decreased. Juniperus virginiana and woody species richness increased for all measures. DCA indicated that re‐sampled stands generally changed from Q. stellata–Q. marilandica‐dominated forests to forests with greater woody species richness and more J. virginiana. Q. stellata remained a dominant tree species; otherwise, composition shifted towards mesophytic and invasive woody species. Measurements taken in the 1950s immediately followed a major drought; whereas subsequent decades were significantly moister. Conclusions: Fire exclusion and drought may have played an important role in driving changes towards lower dominance by Quercus, increased importance of mesophytic and invasive species, and greater woody species richness. These phenomena are similar to those found in Quercus‐dominated forests throughout the northern hemisphere.     

Respiration acclimation contributes to high carbon-use efficiency in a seasonally dry pine forest

Predictions of warming and drying in the Mediterranean and other regions require quantifying of such effects on ecosystem carbon dynamics and respiration. Long‐term effects can only be obtained from forests in which seasonal drought is a regular feature. We carried out measurements in a semiarid Pinus halepensis (Aleppo pine) forest of aboveground respiration rates of foliage, R_f, and stem, R_t over 3 years. Component respiration combined with ongoing biometric, net CO₂ flux [net ecosystem productivity (NEP)] and soil respiration measurements were scaled to the ecosystem level to estimate gross and net primary productivity (GPP, NPP) and carbon‐use efficiency (CUE=NPP/GPP) using 6 years data. GPP, NPP and NEP were, on average, 880, 350 and 211 g C m^?2 yr^?1, respectively. The above ground respiration made up half of total ecosystem respiration but CUE remained high at 0.4. Large seasonal variations in both R_f and R_t were not consistently correlated with seasonal temperature trends. Seasonal adjustments of respiration were observed in both the normalized rate (R₂₀) and short‐term temperature sensitivity (Q₁₀), resulting in low respiration rates during the hot, dry period. R_f in fully developed needles was highest over winter–spring, and foliage R₂₀ was correlated with photosynthesis over the year. Needle growth occurred over summer, with respiration rates in developing needles higher than the fully developed foliage at most times. R_t showed a distinct seasonal maximum in May irrespective of year, which was not correlated to the winter stem growth, but could be associated with phenological drivers such as carbohydrate re‐mobilization and cambial activity. We show that in a semiarid pine forest photosynthesis and stem growth peak in (wet) winter and leaf growth in (dry) summer, and associated adjustments of component respiration, dominated by those in R₂₀, minimize annual respiratory losses. This is likely a key for maintaining high CUE and ecosystem productivity similar to much wetter sites, and could lead to different predictions of the effect of warming and drying climate on productivity of pine forests than based on short‐term droughts.     

Spruce beetle outbreak was not driven by drought stress: Evidence from a tree‐ring iso‐demographic approach indicates temperatures were more important

Climate change has amplified eruptive bark beetle outbreaks over recent decades, including spruce beetle (Dendroctonus rufipennis). However, for projecting future bark beetle dynamics there is a critical lack of evidence to differentiate how outbreaks have been promoted by direct effects of warmer temperatures on beetle life cycles versus indirect effects of drought on host susceptibility. To diagnose whether drought‐induced host‐weakening was important to beetle attack success we used an iso‐demographic approach in Engelmann spruce (Picea engelmannii) forests that experienced widespread mortality caused by spruce beetle outbreaks in the 1990s, during a prolonged drought across the central and southern Rocky Mountain region. We determined tree death date demography during this outbreak to differentiate early‐ and late‐dying trees in stands distributed across a landscape within this larger regional mortality event. To directly test for a role of drought stress during outbreak initiation we determined whether early‐dying trees had greater sensitivity of tree‐ring carbon isotope discrimination (?¹³C) to drought compared to late‐dying trees. Rather, evidence indicated the abundance and size of host trees may have modified ?¹³C responses to drought. ?¹³C sensitivity to drought did not differ among early‐ versus late‐dying trees, which runs contrary to previously proposed links between spruce beetle outbreaks and drought. Overall, our results provide strong support for the view that irruptive spruce beetle outbreaks across North America have primarily been driven by warming‐amplified beetle life cycles whereas drought‐weakened host defenses appear to have been a distant secondary driver of these major disturbance events.     

Soil respiration is stimulated by elevated CO2 and reduced by summer drought: three years of measurements in a multifactor ecosystem manipulation experiment in a temperate heathland (CLIMAITE)

This study investigated the impact of predicted future climatic and atmospheric conditions on soil respiration (R_S) in a Danish Calluna‐Deschampsia‐heathland. A fully factorial in situ experiment with treatments of elevated atmospheric CO₂ (+130 ppm), raised soil temperature (+0.4 °C) and extended summer drought (5–8% precipitation exclusion) was established in 2005. The average R_S, observed in the control over 3 years of measurements (1.7 μmol CO₂ m^?2 sec^?1), increased 38% under elevated CO₂, irrespective of combination with the drought or temperature treatments. In contrast, extended summer drought decreased R_S by 14%, while elevated soil temperature did not affect R_S overall. A significant interaction between elevated temperature and drought resulted in further reduction of R_S when these treatments were combined. A detailed analysis of short‐term R_S dynamics associated with drought periods showed that R_S was reduced by ~50% and was strongly correlated with soil moisture during these events. Recovery of R_S to pre‐drought levels occurred within 2 weeks of rewetting; however, unexpected drought effects were observed several months after summer drought treatment in 2 of the 3 years, possibly due to reduced plant growth or changes in soil water holding capacity. An empirical model that predicts R_S from soil temperature, soil moisture and plant biomass was developed and accounted for 55% of the observed variability in R_S. The model predicted annual sums of R_S in 2006 and 2007, in the control, were 672 and 719 g C m^?2 y^?1, respectively. For the full treatment combination, i.e. the future climate scenario, the model predicted that soil respiratory C losses would increase by ~21% (140–150 g C m^?2 y^?1). Therefore, in the future climate, stimulation of C storage in plant biomass and litter must be in excess of 21% for this ecosystem to not suffer a reduction in net ecosystem exchange.     

Effect of fruiting and drought or flooding on carbon balance of apple trees

The response of fruiting or deblossomed trees to water stress such as drought or flooding was investigated in six semi open-top cuvettes each containing one apple (Malus domestica Borkh. cv. Golden Delicious) tree. Xylem water potentials of leaves dropped from -1.2 to -4.1 MPa within 7 d of drought, the effect being enhanced by fruiting. Apple trees without fruits showed smaller reductions in net photosynthetic rate (P _N ) and dark respiration rate (R _D ) after 2 d of drought and hence more positive carbon balances relative to fruiting trees. Flooding for 4 d had a more pronounced effect on P _N than on transpiration, resulting in a reduced water use efficiency (WUE). This reduction in WUE was greater in the non-fruiting trees. Flooding reduced P _N of the whole apple canopies irrespective of fruiting; aple trees without fruits increased R _D resulting in a less positive carbon balance relative to fruiting trees. Fruiting increased the sensitivity to drought of apple trees (R _D and P _N ), but decreased their sensitivity to flooding (R _D and WUE), suggesting different adaptation mechanisms for the two forms of water stress. This revised version was published online in August 2006 with corrections to the Cover Date.     

r‐ and K‐selection in fluctuating populations is determined by the evolutionary trade‐off between two fitness measures: Growth rate and lifetime reproductive success

In a stable environment, evolution maximizes growth rates in populations that are not density regulated and the carrying capacity in the case of density regulation. In a fluctuating environment, evolution maximizes a function of growth rate, carrying capacity and environmental variance, tending to r‐selection and K‐selection under large and small environmental noise, respectively. Here we analyze a model in which birth and death rates depend on density through the same function but with independent strength of density dependence. As a special case, both functions may be linear, corresponding to logistic dynamics. It is shown that evolution maximizes a function of the deterministic growth rate r₀ and the lifetime reproductive success (LRS) R₀, both defined at small densities, as well as the environmental variance. Under large noise this function is dominated by r₀ and average lifetimes are small, whereas R₀ dominates and lifetimes are larger under small noise. Thus, K‐selection is closely linked to selection for large R₀ so that evolution tends to maximize LRS in a stable environment. Consequently, different quantities (r₀ and R₀) tend to be maximized at low and high densities, respectively, favoring density‐dependent changes in the optimal life history.     

Associations between carbon isotope ratios of ecosystem respiration, water availability and canopy conductance

We tested the hypothesis that the stable carbon isotope signature of ecosystem respiration (δ¹³C_R) was regulated by canopy conductance (G_c) using weekly Keeling plots (n=51) from a semiarid old‐growth ponderosa pine (Pinus ponderosa) forest in Oregon, USA. For a comparison of forests in two contrasting climates we also evaluated trends in δ¹³C_R from a wet 20‐year‐old Douglas‐fir (Pseudotsuga menziesii) plantation located near the Pacific Ocean. Intraannual variability in δ¹³C_R was greater than 8.0‰ at both sites, was highest during autumn, winter, and spring when rainfall was abundant, and lowest during summer drought. The δ¹³C_R of the dry pine forest was consistently more positive than the wetter Douglas‐fir forest (mean annual δ¹³C_R: ?25.41‰ vs. ?26.23‰, respectively, P=0.07). At the Douglas‐fir forest, δ¹³C_R–climate relationships were consistent with predictions based on stomatal regulation of carbon isotope discrimination (Δ). Soil water content (SWC) and vapor pressure deficit (vpd) were the most important factors governing δ¹³C_R in this forest throughout the year. In contrast, δ¹³C_R at the pine forest was relatively insensitive to SWC or vpd, and exhibited a smaller drought‐related enrichment (～2‰) than the enrichment observed during drought at the Douglas‐fir forest (～5‰). Groundwater access at the pine forest may buffer canopy–gas exchange from drought. Despite this potential buffering, δ¹³C_R at the pine forest was significantly but weakly related to canopy conductance (G_c), suggesting that δ¹³C_R remains coupled to canopy–gas exchange despite groundwater access. During drought, δ¹³C_R was strongly correlated with soil temperature at both forests. The hypothesis that canopy‐level physiology is a critical regulator of δ¹³C_R was supported; however, belowground respiration may become more important during rain‐free periods.     

Effect of natural atmospheric CO2 fertilization suggested by open-grown white spruce in a dry environment

Evidence of an atmospheric CO₂‐fertilization effect on radial growth rates was uncovered for open‐grown white spruce in a mixed‐grass prairie of southwestern Manitoba, Canada. Consistent upward trends of the residuals from dendroclimatic models indicated a decreased ability for climatic parameters to predict radial growth. Despite that a similar amount (61%) of the total variation in radial growth index was explained by climate for both young and old trees, residuals from young trees for the period of 1955–1999 demonstrated a stronger upward trend (R²=0.551, P<0.0001) than old trees for the period of 1900–1996 (R²=0.020, P=0.167). Similar to young trees, the residuals from the early growth period (1900–1929) of old trees also demonstrated a stronger upward trend (R²=0.480, P<0.0001) than the period of 1900–1996. Likewise, a comparable period (1970–1999) of young trees also demonstrated a stronger upward trend (R²=0.619, P<0.0001) than the early growth period (1900–1929) of old trees. In addition, postdrought growth response was much stronger for young trees (1970–1999) compared with old trees at the same development stage (1900–1929) (P=0.011) or within the same time period (1970–1999) (P=0.014). There was no difference (P=0.221) in drought recovery between the early (1900–1929) period and the late (1970–1999) period within old trees. Together, our results suggest that (1) open‐grown white spruce trees improved their growth with time at the early developmental stage, and (2) at the same developmental stage, a greater growth response occurred in the late period when atmospheric CO₂ concentration, and the rate of atmospheric CO₂ increase were both relatively high. While it is impossible to rule out other factors, these results are consistent with expectations for CO₂‐fertilization effects.     

Measured and modelled leaf and stand‐scale productivity across a soil moisture gradient and a severe drought

Environmental controls on carbon dynamics operate at a range of interacting scales from the leaf to landscape. The key questions of this study addressed the influence of water and nitrogen (N) availability on Pinus palustris (Mill.) physiology and primary productivity across leaf and canopy scales, linking the soil‐plant‐atmosphere (SPA) model to leaf and stand‐scale flux and leaf trait/canopy data. We present previously unreported ecophysiological parameters (e.g. V_cmax and J_max) for P. palustris and the first modelled estimates of its annual gross primary productivity (GPP) across xeric and mesic sites and under extreme drought. Annual mesic site P. palustris GPP was ～23% greater than at the xeric site. However, at the leaf level, xeric trees had higher net photosynthetic rates, and water and light use efficiency. At the canopy scale, GPP was limited by light interception (canopy level), but co‐limited by nitrogen and water at the leaf level. Contrary to expectations, the impacts of an intense growing season drought were greater at the mesic site. Modelling indicated a 10% greater decrease in mesic GPP compared with the xeric site. Xeric P. palustris trees exhibited drought‐tolerant behaviour that contrasted with mesic trees' drought‐avoidance behaviour.     

A cross‐biome synthesis of soil respiration and its determinants under simulated precipitation changes

Soil respiration (R_s) is the second‐largest terrestrial carbon (C) flux. Although R_s has been extensively studied across a broad range of biomes, there is surprisingly little consensus on how the spatiotemporal patterns of R_s will be altered in a warming climate with changing precipitation regimes. Here, we present a global synthesis R_s data from studies that have manipulated precipitation in the field by collating studies from 113 increased precipitation treatments, 91 decreased precipitation treatments, and 14 prolonged drought treatments. Our meta‐analysis indicated that when the increased precipitation treatments were normalized to 28% above the ambient level, the soil moisture, R_s, and the temperature sensitivity (Q₁₀) values increased by an average of 17%, 16%, and 6%, respectively, and the soil temperature decreased by ?1.3%. The greatest increases in R_s and Q₁₀ were observed in arid areas, and the stimulation rates decreased with increases in climate humidity. When the decreased precipitation treatments were normalized to 28% below the ambient level, the soil moisture and R_s values decreased by an average of ?14% and ?17%, respectively, and the soil temperature and Q₁₀ values were not altered. The reductions in soil moisture tended to be greater in more humid areas. Prolonged drought without alterations in the amount of precipitation reduced the soil moisture and R_s by ?12% and ?6%, respectively, but did not alter Q₁₀. Overall, our synthesis suggests that soil moisture and R_s tend to be more sensitive to increased precipitation in more arid areas and more responsive to decreased precipitation in more humid areas. The responses of R_s and Q₁₀ were predominantly driven by precipitation‐induced changes in the soil moisture, whereas changes in the soil temperature had limited impacts. Finally, our synthesis of prolonged drought experiments also emphasizes the importance of the timing and frequency of precipitation events on ecosystem C cycles. Given these findings, we urge future studies to focus on manipulating the frequency, intensity, and seasonality of precipitation with an aim to improving our ability to predict and model feedback between R_s and climate change.     

Differences in xylem and leaf hydraulic traits explain differences in drought tolerance among mature Amazon rainforest trees

Considerable uncertainty surrounds the impacts of anthropogenic climate change on the composition and structure of Amazon forests. Building upon results from two large‐scale ecosystem drought experiments in the eastern Brazilian Amazon that observed increases in mortality rates among some tree species but not others, in this study we investigate the physiological traits underpinning these differential demographic responses. Xylem pressure at 50% conductivity (xylem‐P₅₀), leaf turgor loss point (TLP), cellular osmotic potential (π_o), and cellular bulk modulus of elasticity (ε), all traits mechanistically linked to drought tolerance, were measured on upper canopy branches and leaves of mature trees from selected species growing at the two drought experiment sites. Each species was placed a priori into one of four plant functional type (PFT) categories: drought‐tolerant versus drought‐intolerant based on observed mortality rates, and subdivided into early‐ versus late‐successional based on wood density. We tested the hypotheses that the measured traits would be significantly different between the four PFTs and that they would be spatially conserved across the two experimental sites. Xylem‐P₅₀, TLP, and π_o, but not ε, occurred at significantly higher water potentials for the drought‐intolerant PFT compared to the drought‐tolerant PFT; however, there were no significant differences between the early‐ and late‐successional PFTs. These results suggest that these three traits are important for determining drought tolerance, and are largely independent of wood density—a trait commonly associated with successional status. Differences in these physiological traits that occurred between the drought‐tolerant and drought‐intolerant PFTs were conserved between the two research sites, even though they had different soil types and dry‐season lengths. This more detailed understanding of how xylem and leaf hydraulic traits vary between co‐occuring drought‐tolerant and drought‐intolerant tropical tree species promises to facilitate a much‐needed improvement in the representation of plant hydraulics within terrestrial ecosystem and biosphere models, which will enhance our ability to make robust predictions of how future changes in climate will affect tropical forests.     

Assimilation exceeds respiration sensitivity to drought: A FLUXNET synthesis

The intensification of the hydrological cycle, with an observed and modeled increase in drought incidence and severity, underscores the need to quantify drought effects on carbon cycling and the terrestrial sink. FLUXNET, a global network of eddy covariance towers, provides dense data streams of meteorological data, and through flux partitioning and gap filling algorithms, estimates of net ecosystem productivity (F_NEP), gross ecosystem productivity (P), and ecosystem respiration (R). We analyzed the functional relationship of these three carbon fluxes relative to evaporative fraction (EF), an index of drought and site water status, using monthly data records from 238 micrometeorological tower sites distributed globally across 11 biomes. The analysis was based on relative anomalies of both EF and carbon fluxes and focused on drought episodes by biome and climatic season. Globally P was ≈50% more sensitive to a drought event than R. Network‐wide drought‐induced decreases in carbon flux averaged ?16.6 and ?9.3 g C m^?2 month^?1 for P and R, i.e., drought events induced a net decline in the terrestrial sink. However, in evergreen forests and wetlands drought was coincident with an increase in P or R during parts of the growing season. The most robust relationships between carbon flux and EF occurred during climatic spring for F_NEP and in climatic summer for P and R. Upscaling flux sensitivities to a global map showed that spatial patterns for all three carbon fluxes were linked to the distribution of croplands. Agricultural areas exhibited the highest sensitivity whereas the tropical region had minimal sensitivity to drought. Combining gridded flux sensitivities with their uncertainties and the spatial grid of FLUXNET revealed that a more robust quantification of carbon flux response to drought requires additional towers in all biomes of Africa and Asia as well as in the cropland, shrubland, savannah, and wetland biomes globally.     

The role of interannual,seasonal, and synoptic climate on the carbon isotope ratio of ecosystem respiration at a semiarid woodland

The terrestrial carbon cycle is influenced by environmental variability at scales ranging from diurnal to interannual. Here, we present 5‐years of growing season (day 131–275) observations of the carbon isotope ratio of ecosystem respiration (δ¹³C_R) from a semiarid woodland. This ecosystem has a large necromass component resulting from 97%Pinus edulis mortality in 2002, is dominated by drought‐tolerant Juniperus monosperma trees, and experiences large variability in the timing and intensity of seasonal and synoptic water availability. Mean growing season δ¹³C_R was remarkably invariant (?23.57±0.4‰), with the exception of particularly enriched δ¹³C_R in 2006 following a winter with anomalously low snowfall. δ¹³C_R was strongly coupled to climate during premonsoon periods (～May to June), including fast (≤2 days) responses to changes in crown‐level stomatal conductance (G_c) and vapor pressure deficit (vpd) following rain pulses. In contrast, δ¹³C_R was relatively decoupled from G_c and environmental drivers during monsoon and postmonsoon periods (July–August and September, respectively), exhibiting only infrequent couplings of δ¹³C_R to vpd and soil water content (SWC) with longer lags (～8 days) and variable response slopes (both positive and negative). Notably, δ¹³C_R exhibited consistent dynamics after rainfall events, with depleted δ¹³C_R occurring within 1 h, progressive hourly δ¹³C_R enrichment over the remainder of the night, and net δ¹³C_R depletions over the multiple nights postevent in monsoon and postmonsoon periods. Overall this ecosystem demonstrated strong dependence of δ¹³C_R on precipitation, with an apparent dominance by the autotrophic δ¹³C signal in premonsoon periods when deep soil moisture is abundant and surface soil moisture is low, and weaker coupling during monsoonal periods consistent with increasing heterotrophic dominance when deep soil moisture has declined and surface moisture is variable.     

Satellite detection of cumulative and lagged effects of drought on autumn leaf senescence over the Northern Hemisphere

Climate change has substantial influences on autumn leaf senescence, that is, the end of the growing season (EOS). Relative to the impacts of temperature and precipitation on EOS, the influence of drought is not well understood, especially considering that there are apparent cumulative and lagged effects of drought on plant growth. Here, we investigated the cumulative and lagged effects of drought (in terms of the Standardized Precipitation–Evapotranspiration Index, SPEI) on EOS derived from the normalized difference vegetation index (NDVI3g) data over the Northern Hemisphere extra‐tropical ecosystems (>30°N) during 1982–2015. The cumulative effect was determined by the number of antecedent months at which SPEI showed the maximum correlation with EOS (i.e., R_max‐cml) while the lag effect was determined by a month during which the maximum correlation between 1‐month SPEI and EOS occurred (i.e., R_max‐lag). We found cumulative effect of drought on EOS for 27.2% and lagged effect for 46.2% of the vegetated land area. For the dominant time scales where the R_max‐cml and R_max‐lag occurred, we observed 1–4 accumulated months for the cumulative effect and 2–6 lagged months for the lagged effect. At the biome level, drought had stronger impacts on EOS in grasslands, savannas, and shrubs than in forests, which may be related to the different root functional traits among vegetation types. Considering hydrological conditions, the mean values of both R_max‐cml and R_max‐lag decreased along the gradients of annual SPEI and its slope, suggesting stronger cumulative and lagged effects in drier regions as well as in areas with decreasing water availability. Furthermore, the average accumulated and lagged months tended to decline along the annual SPEI gradient but increase with increasing annual SPEI. Our results revealed that drought has strong cumulative and lagged effects on autumn phenology, and considering these effects could provide valuable information on the vegetation response to a changing climate.     

Carbohydrate dynamics and mortality in a piñon‐juniper woodland under three future precipitation scenarios

Drought‐induced forest mortality is an increasing global problem with wide‐ranging consequences, yet mortality mechanisms remain poorly understood. Depletion of non‐structural carbohydrate (NSC) stores has been implicated as an important mechanism in drought‐induced mortality, but experimental field tests are rare. We used an ecosystem‐scale precipitation manipulation experiment to evaluate leaf and twig NSC dynamics of two co‐occurring conifers that differ in patterns of stomatal regulation of water loss and recent mortality: the relatively desiccation‐avoiding piñon pine (Pinus edulis) and the relatively desiccation‐tolerant one‐seed juniper (Juniperus monosperma). Piñon pine experienced 72% mortality after 13–25 months of experimental drought and juniper experienced 20% mortality after 32–47 months. Juniper maintained three times more NSC in the foliage than twigs, and converted NSC to glucose and fructose under drought, consistent with osmoregulation requirements to maintain higher stomatal conductance during drought than piñon. Despite these species differences, experimental drought caused decreased leaf starch content in dying trees of both species (P < 0.001). Average dry‐season leaf starch content was also a good predictor of drought‐survival time for both species (R² = 0.93). These results, along with observations of drought‐induced reductions to photosynthesis and growth, support carbon limitation as an important process during mortality of these two conifer species.     

Dynamics of an introduced population of mouflon Ovis aries on the sub‐Antarctic archipelago of Kerguelen

A commonly reported pattern in large herbivores is their propensity to irrupt and crash when colonizing new areas. However, the relative role of density‐dependence, climate, and cohort effects on demographic rates in accounting for the irruptive dynamics of large herbivores remains unclear. Using a 37‐yr time series of abundance in a mouflon Ovis aries population located on Haute Island, a sub‐Antarctic island of Kerguelen, 1) we investigated if irruptive dynamics occurred and 2) we quantified the relative effects of density and climate on mouflon population dynamics. Being released in a new environment, we expected mouflon to show rapid growth and marked over‐compensation. In support of this prediction, we found a two‐phase dynamics, the first phase being characterised by an irruptive pattern best described by the θ‐Caughley model. Parameter estimates were r_m=0.29±0.005(maximum growth rate), K=473±45 (carrying capacity) and S=2903±396 (surplus) mouflon. With a θ=3.18±0.69 our model also supported the hypothesis that density dependence is strongest at high density in large herbivores. The second phase was characterised by an unstable dynamics where growth rate was negatively affected by population abundance and winter precipitation. Climate, however, did not trigger population crashes and our model suggested that lagged density‐dependence and over‐grazing were the probable causes of mouflon irruptive dynamics. We compare our results with those of Soay sheep and discuss the possibility of a reversible alteration of the island carrying capacity after the initial over‐grazing period.