Drought shortens subtropical understory growing season by advancing leaf senescence.

Subtropical forests, recognized for their intricate vertical canopy stratification, exhibit high resistance to extreme drought. However, the response of leaf phenology to drought in the species-rich understory remains poorly understood. In this study, we constructed a digital camera system, amassing over 360,000 images through a 70% throughfall exclusion experiment, to explore the drought response of understory leaf phenology. The results revealed a significant advancement in understory leaf senescence phenology under drought, with 11.75 and 15.76 days for the start and end of the leaf-falling event, respectively. Pre-season temperature primarily regulated leaf development phenology, whereas soil water dominated the variability in leaf senescence phenology. Under drought conditions, temperature sensitivities for the end of leaf emergence decreased from -13.72 to -11.06 days °C-1, with insignificance observed for the start of leaf emergence. Consequently, drought treatment shortened both the length of the growing season (15.69 days) and the peak growth season (9.80 days) for understory plants. Moreover, this study identified diverse responses among intraspecies and interspecies to drought, particularly during the leaf development phase. These findings underscore the pivotal role of water availability in shaping understory phenology patterns, especially in subtropical forests.

Nitrogen use strategy drives interspecific differences in plant photosynthetic CO2 acclimation.

Rising atmospheric CO2 concentration triggers an emergent phenomenon called plant photosynthetic acclimation to elevated CO2 (PAC). PAC is often characterized by a reduction in leaf photosynthetic capacity (Asat ), which varies dramatically along the continuum of plant phylogeny. However, it remains unclear whether the mechanisms responsible for PAC are also different across plant phylogeny, especially between gymnosperms and angiosperms. Here, by compiling a dataset of 73 species, we found that although leaf Asat increased significantly from gymnosperms to angiosperms, there was no phylogenetic signal in the PAC magnitude along the phylogenetic continuum. Physio-morphologically, leaf nitrogen concentration (Nm ), photosynthetic nitrogen-use efficiency (PNUE), and leaf mass per area (LMA) dominated PAC for 36, 29, and 8 species, respectively. However, there was no apparent difference in PAC mechanisms across major evolutionary clades, with 75% of gymnosperms and 92% of angiosperms regulated by the combination of Nm and PNUE. There was a trade-off between Nm and PNUE in driving PAC across species, and PNUE dominated the long-term changes and inter-specific differences in Asat under elevated CO2 . These findings indicate that nitrogen-use strategy drives the acclimation of leaf photosynthetic capacity to elevated CO2 across terrestrial plant species.

Across-model spread and shrinking in predicting peatland carbon dynamics under global change.

Large across-model spread in simulating land carbon (C) dynamics has been ubiquitously demonstrated in model intercomparison projects (MIPs), and became a major impediment in advancing climate change prediction. Thus, it is imperative to identify underlying sources of the spread. Here, we used a novel matrix approach to analytically pin down the sources of across-model spread in transient peatland C dynamics in response to a factorial combination of two atmospheric CO2 levels and five temperature levels. We developed a matrix-based MIP by converting the C cycle module of eight land models (i.e., TEM, CENTURY4, DALEC2, TECO, FBDC, CASA, CLM4.5 and ORCHIDEE) into eight matrix models. While the model average of ecosystem C storage was comparable to the measurement, the simulation differed largely among models, mainly due to inter-model difference in baseline C residence time. Models generally overestimated net ecosystem production (NEP), with a large spread that was mainly attributed to inter-model difference in environmental scalar. Based on the sources of spreads identified, we sequentially standardized model parameters to shrink simulated ecosystem C storage and NEP to almost none. Models generally captured the observed negative response of NEP to warming, but differed largely in the magnitude of response, due to differences in baseline C residence time and temperature sensitivity of decomposition. While there was a lack of response of NEP to elevated CO2 (eCO2 ) concentrations in the measurements, simulated NEP responded positively to eCO2 concentrations in most models, due to the positive responses of simulated net primary production. Our study used one case study in Minnesota peatland to demonstrate that the sources of across-model spreads in simulating transient C dynamics can be precisely traced to model structures and parameters, regardless of their complexity, given the protocol that all the matrix models were driven by the same gross primary production and environmental variables.

Nitrogen enrichment alters multiple dimensions of grassland functional stability via changing compositional stability.

Anthropogenic nutrient enrichment is known to alter the composition and functioning of plant communities. However, how nutrient enrichment influences multiple dimensions of community- and ecosystem-level stability remains poorly understood. Using data from a nitrogen (N) and phosphorus (P) addition experiment in a temperate semi-arid grassland that experienced a natural drought, we show that N enrichment, not P enrichment, decreased grassland functional and compositional temporal stability, resistance and recovery but increased functional and compositional resilience. Compositional stability and species asynchrony, rather than species diversity, were identified as key determinants of all dimensions of grassland functional stability, except for recovery. Whereas grassland functional recovery was decoupled from compositional recovery, N enrichment altered other dimensions of functional stability primarily through changing their corresponding compositional stability dimensions. Our findings highlight the need to examine ecological stability at the community level for a more mechanistic understanding of ecosystem dynamics in the face of environmental change.

Can evolutionary history predict plant plastic responses to climate change?

Plant plastic responses are critical to the adaptation and survival of species under climate change, but whether they are constrained by evolutionary history (phylogeny) is largely unclear. Plant leaf traits are key in determining plants' performance in different environments, and if these traits and their variation are phylogenetically dependent, predictions could be made to identify species vulnerable to climate change. We compiled data on three leaf traits (photosynthetic rate, specific leaf area, and leaf nitrogen content) and their variation under four environmental change scenarios (warming, drought, elevated CO2 , or nitrogen addition) for 434 species, from 210 manipulation experiments. We found phylogenetic signal in the three traits but not in their variation under the four scenarios. This indicates that closely related species show similar traits but that their plastic responses could not be predicted from species relatedness under environmental change. Meanwhile, phylogeny weakened the slopes but did not change the directions of conventional pairwise trait relationships, suggesting that co-evolved leaf trait pairs have consistent responses under contrasting environmental conditions. Phylogeny can identify lineages rich in species showing similar traits and predict their relationships under climate change, but the degree of plant phenotypic variation does not vary consistently across evolutionary clades.

Soil phosphorus drives plant trait variations in a mature subtropical forest.

Earth system models are implementing soil phosphorus dynamic and plant functional traits to predict functional changes in global forests. However, the linkage between soil phosphorus and plant traits lacks empirical evidence, especially in mature forests. Here, we examined the soil phosphorus constraint on plant functional traits in a mature subtropical forest based on observations of 9943 individuals from 90 species in a 5-ha forest dynamic plot and 405 individuals from 15 species in an adjacent 10-year nutrient-addition experiment. We first confirmed a pervasive phosphorus limitation on subtropical tree growth based on leaf N:P ratios. Then, we found that soil phosphorus dominated multidimensional trait variations in the 5-ha forest dynamic plot. Soil phosphorus content explained 44% and 53% of the variance in the traits defining the main functional space across species and communities, respectively. Lastly, we found much stronger phosphorus effects on most plant functional traits than nitrogen at both species and community levels in the 10-year nutrient-addition experiment. This study provides evidence for the consistent pattern of soil phosphorus constraint on plant trait variations between the species and community levels in a mature evergreen broadleaf forest in the East Asian monsoon region. These findings shed light on the predominant role of soil phosphorus on plant functional trait variations in mature subtropical forests, providing new insights for models to incorporate soil phosphorus constraint in predicting future vegetation dynamics.

Plant photosynthetic overcompensation under nocturnal warming: lack of evidence in subtropical evergreen trees.

BACKGROUND AND AIMS: Increased plant photosynthesis under nocturnal warming is a negative feedback mechanism to overcompensate for night-time carbon loss to mitigate climate warming. This photosynthetic overcompensation effect has been observed in dry deciduous ecosystems but whether it exists in subtropical wet forest trees is unclear. METHODS: Two subtropical evergreen tree species (Schima superba and Castanopsis sclerophylla) were grown in a greenhouse and exposed to ambient and elevated night-time temperature. The occurrence of the photosynthetic overcompensation effect was determined by measuring daytime and night-time leaf gas exchange and non-structural carbohydrate (NSC) concentration. KEY RESULTS: A reduction in leaf photosynthesis for both species and an absence of persistent photosynthetic overcompensation were observed. The photosynthetic overcompensation effect was transient in S. superba due to respiratory acclimation and stomatal limitation. For S. superba, nocturnal warming resulted in insufficient changes in night-time respiration and NSC concentration to stimulate overcompensation and inhibited leaf stomatal conductance by increasing the leaf-to-air vapour pressure deficit. CONCLUSIONS: The results indicate that leaf stomatal conductance is important for the photosynthetic overcompensation effect in different tree species. The photosynthetic overcompensation effect under nocturnal warming may be a transient occurrence rather than a persistent mechanism in subtropical forest ecosystems.

Warming reshaped the microbial hierarchical interactions.

Global warming may alter microbially mediated ecosystem functions through reshaping of microbial diversity and modified microbial interactions. Here, we examined the effects of 5-year experimental warming on different microbial hierarchical groups in a coastal nontidal soil ecosystem, including prokaryotes (i.e., bacteria and archaea), fungi, and Cercozoa, which is a widespread phylum of protists. Warming significantly altered the diversity and structure of prokaryotic and fungal communities in soil and additionally decreased the complexity of the prokaryotic network and fragmented the cercozoan network. By using the Inter-Domain Ecological Network approach, the cross-trophic interactions among prokaryotes, fungi, and Cercozoa were further investigated. Under warming, cercozoan-prokaryotic and fungal-prokaryotic bipartite networks were simplified, whereas the cercozoan-fungal network became slightly more complex. Despite simplification of the fungal-prokaryotic network, the strengthened synergistic interactions between saprotrophic fungi and certain prokaryotic groups, such as the Bacteroidetes, retained these phyla within the network under warming. In addition, the interactions within the fungal community were quite stable under warming conditions, which stabilized the interactions between fungi and prokaryotes or protists. Additionally, we found the microbial hierarchical interactions were affected by environmental stress (i.e., salinity and pH) and soil nutrients. Interestingly, the relevant microbial groups could respond to different soil properties under ambient conditions, whereas under warming these two groups tended to respond to similar soil properties, suggesting network hub species responded to certain environmental changes related to warming, and then transferred this response to their partners through trophic interactions. Finally, warming strengthened the network modules' negative association with soil organic matters through some fungal hub species, which might trigger soil carbon loss in this ecosystem. Our study provides new insights into the response and feedback of microbial hierarchical interactions under warming scenario.

Research challenges and opportunities for using big data in global change biology.

Global change biology has been entering a big data era due to the vast increase in availability of both environmental and biological data. Big data refers to large data volume, complex data sets, and multiple data sources. The recent use of such big data is improving our understanding of interactions between biological systems and global environmental changes. In this review, we first explore how big data has been analyzed to identify the general patterns of biological responses to global changes at scales from gene to ecosystem. After that, we investigate how observational networks and space-based big data have facilitated the discovery of emergent mechanisms and phenomena on the regional and global scales. Then, we evaluate the predictions of terrestrial biosphere under global changes by big modeling data. Finally, we introduce some methods to extract knowledge from big data, such as meta-analysis, machine learning, traceability analysis, and data assimilation. The big data has opened new research opportunities, especially for developing new data-driven theories for improving biological predictions in Earth system models, tracing global change impacts across different organismic levels, and constructing cyberinfrastructure tools to accelerate the pace of model-data integrations. These efforts will uncork the bottleneck of using big data to understand biological responses and adaptations to future global changes.

Warming-induced global soil carbon loss attenuated by downward carbon movement.

The fate of soil organic carbon (SOC) under warming is poorly understood, particularly across large extents and in the whole-soil profile. Using a data-model integration approach applied across the globe, we find that downward movement of SOC along the soil profile reduces SOC loss under warming. We predict that global SOC stocks (down to 2 m) will decline by 4% (~80 Pg) on average when SOC reaches the steady state under 2°C warming, assuming no changes in net primary productivity (NPP). To compensate such decline (i.e. maintain current SOC stocks), a 3% increase of NPP is required. Without the downward SOC movement, global SOC declines by 15%, while a 20% increase in NPP is needed to compensate that loss. This vital role of downward SOC movement in controlling whole-soil profile SOC dynamics in response to warming is due to the protection afforded to downward-moving SOC by depth, indicated by much longer residence times of SOC in deeper layers. Additionally, we find that this protection could not be counteracted by promoted decomposition due to the priming of downward-moving new SOC from upper layers on native old SOC in deeper layers. This study provides the first estimation of whole-soil SOC changes under warming and additional NPP required to compensate such changes across the globe, and reveals the vital role of downward movement of SOC in reducing SOC loss under global warming.

Traits mediate drought effects on wood carbon fluxes.

CO2 fluxes from wood decomposition represent an important source of carbon from forest ecosystems to the atmosphere, which are determined by both wood traits and climate influencing the metabolic rates of decomposers. Previous studies have quantified the effects of moisture and temperature on wood decomposition, but these effects were not separated from the potential influence of wood traits. Indeed, it is not well understood how traits and climate interact to influence wood CO2 fluxes. Here, we examined the responses of CO2 fluxes from dead wood with different traits (angiosperm and gymnosperm) to 0%, 35%, and 70% rainfall reduction across seasonal temperature gradients. Our results showed that drought significantly decreased wood CO2 fluxes, but its effects varied with both taxonomical group and drought intensity. Drought-induced reduction in wood CO2 fluxes was larger in angiosperms than gymnosperms for the 35% rainfall reduction treatment, but there was no significant difference between these groups for the 70% reduction treatment. This is because wood nitrogen density and carbon quality were significantly higher in angiosperms than gymnosperms, yielding a higher moisture sensitivity of wood decomposition. These findings were demonstrated by a significant positive interaction effect between wood nitrogen and moisture on CO2 fluxes in a structural equation model. Additionally, we ascertained that a constant temperature sensitivity of CO2 fluxes was independent of wood traits and consistent with previous estimates for extracellular enzyme kinetics. Our results highlight the key role of wood traits in regulating drought responses of wood carbon fluxes. Given that both climate and forest management might extensively modify taxonomic compositions in the future, it is critical for carbon cycle models to account for such interactions between wood traits and climate in driving dynamics of wood decomposition.

Plant evolutionary history mainly explains the variance in biomass responses to climate warming at a global scale.

Evolutionary history shapes the interspecific relatedness and intraspecific variation, which has a profound influence on plant functional traits and productivity. However, it is far from clear how the phylogenetic relatedness among species and intraspecific variation could contribute to the observed variance in plant biomass responses to climate warming. We compiled a dataset with 284 species from warming experiments to explore the relative importance of phylogenetic, intraspecific, experimental and ecological factors to warming effects on plant biomass, using phylogenetic eigenvector regression and variance decomposition. Our results showed that phylogenetic relatedness could account for about half the total variance in biomass responses to warming, which were correlated with leaf economic traits at the family level but not at species level. The intraspecific variation contributed to approximately one-third of the variance, whereas the experimental design and ecological characteristics only explained 7-17%. These results suggest that intrinsic factors (evolutionary history) play more important roles than extrinsic factors (experimental treatment and environment) in determining the responses of plant biomass to warming at the global scale. This highlights the urgent need for land surface models to include evolutionary aspects in predicting ecosystem functions under climate change.

High ecosystem stability of evergreen broadleaf forests under severe droughts.

Global increase in drought occurrences threatens the stability of terrestrial ecosystem functioning. Evergreen broadleaf forests (EBFs) keep leaves throughout the year, and therefore could experience higher drought risks than other biomes. However, the recent temporal variability of global vegetation productivity or land carbon sink is mainly driven by non-evergreen ecosystems, such as semiarid grasslands, croplands, and boreal forests. Thus, we hypothesize that EBFs have higher stability than other biomes under the increasingly extreme droughts. Here we use long-term Standardized Precipitation and Evaporation Index (SPEI) data and satellite-derived Enhanced Vegetation Index (EVI) products to quantify the temporal stability (ratio of mean annual EVI to its SD), resistance (ability to maintain its original levels during droughts), and resilience (rate of EVI recovering to pre-drought levels) at biome and global scales. We identified significantly increasing trends of annual drought severity (SPEI range: -0.08 to -1.80), area (areal fraction range: 2%-19%), and duration (month range: 7.9-9.1) in the EBF biome over 2000-2014. However, EBFs showed the highest resistance of EVI to droughts, but no significant differences in resilience of EVI to droughts were found among biomes (forests, grasslands, savannas, and shrublands). Global resistance and resilience of EVI to droughts were largely affected by temperature and solar radiation. These findings suggest that EBFs have higher stability than other biomes despite the greater drought exposure. Thus, the conservation of EBFs is critical for stabilizing global vegetation productivity and land carbon sink under more-intense climate extremes in the future.

Matrix approach to land carbon cycle modeling: A case study with the Community Land Model.

The terrestrial carbon (C) cycle has been commonly represented by a series of C balance equations to track C influxes into and effluxes out of individual pools in earth system models (ESMs). This representation matches our understanding of C cycle processes well but makes it difficult to track model behaviors. It is also computationally expensive, limiting the ability to conduct comprehensive parametric sensitivity analyses. To overcome these challenges, we have developed a matrix approach, which reorganizes the C balance equations in the original ESM into one matrix equation without changing any modeled C cycle processes and mechanisms. We applied the matrix approach to the Community Land Model (CLM4.5) with vertically-resolved biogeochemistry. The matrix equation exactly reproduces litter and soil organic carbon (SOC) dynamics of the standard CLM4.5 across different spatial-temporal scales. The matrix approach enables effective diagnosis of system properties such as C residence time and attribution of global change impacts to relevant processes. We illustrated, for example, the impacts of CO2 fertilization on litter and SOC dynamics can be easily decomposed into the relative contributions from C input, allocation of external C into different C pools, nitrogen regulation, altered soil environmental conditions, and vertical mixing along the soil profile. In addition, the matrix tool can accelerate model spin-up, permit thorough parametric sensitivity tests, enable pool-based data assimilation, and facilitate tracking and benchmarking of model behaviors. Overall, the matrix approach can make a broad range of future modeling activities more efficient and effective.

Differential responses of carbon-degrading enzyme activities to warming: Implications for soil respiration.

Extracellular enzymes catalyze rate-limiting steps in soil organic matter decomposition, and their activities (EEAs) play a key role in determining soil respiration (SR). Both EEAs and SR are highly sensitive to temperature, but their responses to climate warming remain poorly understood. Here, we present a meta-analysis on the response of soil cellulase and ligninase activities and SR to warming, synthesizing data from 56 studies. We found that warming significantly enhanced ligninase activity by 21.4% but had no effect on cellulase activity. Increases in ligninase activity were positively correlated with changes in SR, while no such relationship was found for cellulase. The warming response of ligninase activity was more closely related to the responses of SR than a wide range of environmental and experimental methodological factors. Furthermore, warming effects on ligninase activity increased with experiment duration. These results suggest that soil microorganisms sustain long-term increases in SR with warming by gradually increasing the degradation of the recalcitrant carbon pool.

Joint structural and physiological control on the interannual variation in productivity in a temperate grassland: A data-model comparison.

Given the important contributions of semiarid region to global land carbon cycle, accurate modeling of the interannual variability (IAV) of terrestrial gross primary productivity (GPP) is important but remains challenging. By decomposing GPP into leaf area index (LAI) and photosynthesis per leaf area (i.e., GPP_leaf), we investigated the IAV of GPP and the mechanisms responsible in a temperate grassland of northwestern China. We further assessed six ecosystem models for their capabilities in reproducing the observed IAV of GPP in a temperate grassland from 2004 to 2011 in China. We observed that the responses to LAI and GPP_leaf to soil water significantly contributed to IAV of GPP at the grassland ecosystem. Two of six models with prescribed LAI simulated of the observed IAV of GPP quite well, but still underestimated the variance of GPP_leaf, therefore the variance of GPP. In comparison, simulated pattern by the other four models with prognostic LAI differed significantly from the observed IAV of GPP. Only some models with prognostic LAI can capture the observed sharp decline of GPP in drought years. Further analysis indicated that accurately representing the responses of GPP_leaf and leaf stomatal conductance to soil moisture are critical for the models to reproduce the observed IAV of GPP_leaf. Our framework also identified that the contributions of LAI and GPP_leaf to the observed IAV of GPP were relatively independent. We conclude that our framework of decomposing GPP into LAI and GPP_leaf has a significant potential for facilitating future model intercomparison, benchmarking and optimization should be adopted for future data-model comparisons.

Joint control of terrestrial gross primary productivity by plant phenology and physiology.

Terrestrial gross primary productivity (GPP) varies greatly over time and space. A better understanding of this variability is necessary for more accurate predictions of the future climate-carbon cycle feedback. Recent studies have suggested that variability in GPP is driven by a broad range of biotic and abiotic factors operating mainly through changes in vegetation phenology and physiological processes. However, it is still unclear how plant phenology and physiology can be integrated to explain the spatiotemporal variability of terrestrial GPP. Based on analyses of eddy-covariance and satellite-derived data, we decomposed annual terrestrial GPP into the length of the CO2 uptake period (CUP) and the seasonal maximal capacity of CO2 uptake (GPPmax). The product of CUP and GPPmax explained >90% of the temporal GPP variability in most areas of North America during 2000-2010 and the spatial GPP variation among globally distributed eddy flux tower sites. It also explained GPP response to the European heatwave in 2003 (r(2) = 0.90) and GPP recovery after a fire disturbance in South Dakota (r(2) = 0.88). Additional analysis of the eddy-covariance flux data shows that the interbiome variation in annual GPP is better explained by that in GPPmax than CUP. These findings indicate that terrestrial GPP is jointly controlled by ecosystem-level plant phenology and photosynthetic capacity, and greater understanding of GPPmax and CUP responses to environmental and biological variations will, thus, improve predictions of GPP over time and space.

Daytime warming lowers community temporal stability by reducing the abundance of dominant, stable species.

Daytime warming and nighttime warming have the potential to influence plant community structure and ecosystem functions. However, their impacts on ecological stability remain largely unexplored. We conducted an eight-year field experiment to compare the effects of daytime and nighttime warming on the temporal stability of a temperate steppe in northern China. Our results showed that the cover and stability of dominant species, stability of subordinate species, and compensatory dynamics among species strongly influenced community-level stability. However, daytime, but not nighttime, warming significantly reduced community temporal stability mainly through the reduction in the abundance of dominant, stable species. These findings demonstrate the differential effects of daytime and nighttime warming on community stability and emphasize the importance of understanding the changes of dominant species for accurately predicting community dynamics under climate warming.

Gross primary production responses to warming, elevated CO2 , and irrigation: quantifying the drivers of ecosystem physiology in a semiarid grassland.

Determining whether the terrestrial biosphere will be a source or sink of carbon (C) under a future climate of elevated CO2 (eCO2 ) and warming requires accurate quantification of gross primary production (GPP), the largest flux of C in the global C cycle. We evaluated 6 years (2007-2012) of flux-derived GPP data from the Prairie Heating and CO2 Enrichment (PHACE) experiment, situated in a grassland in Wyoming, USA. The GPP data were used to calibrate a light response model whose basic formulation has been successfully used in a variety of ecosystems. The model was extended by modeling maximum photosynthetic rate (Amax ) and light-use efficiency (Q) as functions of soil water, air temperature, vapor pressure deficit, vegetation greenness, and nitrogen at current and antecedent (past) timescales. The model fits the observed GPP well (R2  = 0.79), which was confirmed by other model performance checks that compared different variants of the model (e.g. with and without antecedent effects). Stimulation of cumulative 6-year GPP by warming (29%, P = 0.02) and eCO2 (26%, P = 0.07) was primarily driven by enhanced C uptake during spring (129%, P = 0.001) and fall (124%, P = 0.001), respectively, which was consistent across years. Antecedent air temperature (Tairant ) and vapor pressure deficit (VPDant ) effects on Amax (over the past 3-4 days and 1-3 days, respectively) were the most significant predictors of temporal variability in GPP among most treatments. The importance of VPDant suggests that atmospheric drought is important for predicting GPP under current and future climate; we highlight the need for experimental studies to identify the mechanisms underlying such antecedent effects. Finally, posterior estimates of cumulative GPP under control and eCO2 treatments were tested as a benchmark against 12 terrestrial biosphere models (TBMs). The narrow uncertainties of these data-driven GPP estimates suggest that they could be useful semi-independent data streams for validating TBMs.

Challenging terrestrial biosphere models with data from the long-term multifactor Prairie Heating and CO2 Enrichment experiment.

Multifactor experiments are often advocated as important for advancing terrestrial biosphere models (TBMs), yet to date, such models have only been tested against single-factor experiments. We applied 10 TBMs to the multifactor Prairie Heating and CO2 Enrichment (PHACE) experiment in Wyoming, USA. Our goals were to investigate how multifactor experiments can be used to constrain models and to identify a road map for model improvement. We found models performed poorly in ambient conditions; there was a wide spread in simulated above-ground net primary productivity (range: 31-390 g C m-2  yr-1 ). Comparison with data highlighted model failures particularly with respect to carbon allocation, phenology, and the impact of water stress on phenology. Performance against the observations from single-factors treatments was also relatively poor. In addition, similar responses were predicted for different reasons across models: there were large differences among models in sensitivity to water stress and, among the N cycle models, N availability during the experiment. Models were also unable to capture observed treatment effects on phenology: they overestimated the effect of warming on leaf onset and did not allow CO2 -induced water savings to extend the growing season length. Observed interactive (CO2  × warming) treatment effects were subtle and contingent on water stress, phenology, and species composition. As the models did not correctly represent these processes under ambient and single-factor conditions, little extra information was gained by comparing model predictions against interactive responses. We outline a series of key areas in which this and future experiments could be used to improve model predictions of grassland responses to global change.