Tropical forests are approaching critical temperature thresholds.

The critical temperature beyond which photosynthetic machinery in tropical trees begins to fail averages approximately 46.7 °C (Tcrit)1. However, it remains unclear whether leaf temperatures experienced by tropical vegetation approach this threshold or soon will under climate change. Here we found that pantropical canopy temperatures independently triangulated from individual leaf thermocouples, pyrgeometers and remote sensing (ECOSTRESS) have midday peak temperatures of approximately 34 °C during dry periods, with a long high-temperature tail that can exceed 40 °C. Leaf thermocouple data from multiple sites across the tropics suggest that even within pixels of moderate temperatures, upper canopy leaves exceed Tcrit 0.01% of the time. Furthermore, upper canopy leaf warming experiments (+2, 3 and 4 °C in Brazil, Puerto Rico and Australia, respectively) increased leaf temperatures non-linearly, with peak leaf temperatures exceeding Tcrit 1.3% of the time (11% for more than 43.5 °C, and 0.3% for more than 49.9 °C). Using an empirical model incorporating these dynamics (validated with warming experiment data), we found that tropical forests can withstand up to a 3.9 ± 0.5 °C increase in air temperatures before a potential tipping point in metabolic function, but remaining uncertainty in the plasticity and range of Tcrit in tropical trees and the effect of leaf death on tree death could drastically change this prediction. The 4.0 °C estimate is within the 'worst-case scenario' (representative concentration pathway (RCP) 8.5) of climate change predictions2 for tropical forests and therefore it is still within our power to decide (for example, by not taking the RCP 6.0 or 8.5 route) the fate of these critical realms of carbon, water and biodiversity3,4.

Limits of thermal and hydrological tolerance in a foundation tree species (Populus fremontii) in the desert southwestern United States.

Populus fremontii is among the most dominant, and ecologically important riparian tree species in the western United States and can thrive in hyper-arid riparian corridors. Yet, P. fremontii forests have rapidly declined over the last decade, particularly in places where temperatures sometimes exceed 50°C. We evaluated high temperature tolerance of leaf metabolism, leaf thermoregulation, and leaf hydraulic function in eight P. fremontii populations spanning a 5.3°C mean annual temperature gradient in a well-watered common garden, and at source locations throughout the lower Colorado River Basin. Two major results emerged. First, despite having an exceptionally high Tcrit (the temperature at which Photosystem II is disrupted) relative to other tree taxa, recent heat waves exceeded Tcrit , requiring evaporative leaf cooling to maintain leaf-to-air thermal safety margins. Second, in midsummer, genotypes from the warmest locations maintained lower midday leaf temperatures, a higher midday stomatal conductance, and maintained turgor pressure at lower water potentials than genotypes from more temperate locations. Taken together, results suggest that under well-watered conditions, P. fremontii can regulate leaf temperature below Tcrit along the warm edge of its distribution. Nevertheless, reduced Colorado River flows threaten to lower water tables below levels needed for evaporative cooling during episodic heat waves.

Fine root dynamics across pantropical rainforest ecosystems.

Fine roots constitute a significant component of the net primary productivity (NPP) of forest ecosystems but are much less studied than aboveground NPP. Comparisons across sites and regions are also hampered by inconsistent methodologies, especially in tropical areas. Here, we present a novel dataset of fine root biomass, productivity, residence time, and allocation in tropical old-growth rainforest sites worldwide, measured using consistent methods, and examine how these variables are related to consistently determined soil and climatic characteristics. Our pantropical dataset spans intensive monitoring plots in lowland (wet, semi-deciduous, and deciduous) and montane tropical forests in South America, Africa, and Southeast Asia (n = 47). Large spatial variation in fine root dynamics was observed across montane and lowland forest types. In lowland forests, we found a strong positive linear relationship between fine root productivity and sand content, this relationship was even stronger when we considered the fractional allocation of total NPP to fine roots, demonstrating that understanding allocation adds explanatory power to understanding fine root productivity and total NPP. Fine root residence time was a function of multiple factors: soil sand content, soil pH, and maximum water deficit, with longest residence times in acidic, sandy, and water-stressed soils. In tropical montane forests, on the other hand, a different set of relationships prevailed, highlighting the very different nature of montane and lowland forest biomes. Root productivity was a strong positive linear function of mean annual temperature, root residence time was a strong positive function of soil nitrogen content in montane forests, and lastly decreasing soil P content increased allocation of productivity to fine roots. In contrast to the lowlands, environmental conditions were a better predictor for fine root productivity than for fractional allocation of total NPP to fine roots, suggesting that root productivity is a particularly strong driver of NPP allocation in tropical mountain regions.

Drought impact on forest carbon dynamics and fluxes in Amazonia.

In 2005 and 2010 the Amazon basin experienced two strong droughts, driven by shifts in the tropical hydrological regime possibly associated with global climate change, as predicted by some global models. Tree mortality increased after the 2005 drought, and regional atmospheric inversion modelling showed basin-wide decreases in CO2 uptake in 2010 compared with 2011 (ref. 5). But the response of tropical forest carbon cycling to these droughts is not fully understood and there has been no detailed multi-site investigation in situ. Here we use several years of data from a network of thirteen 1-ha forest plots spread throughout South America, where each component of net primary production (NPP), autotrophic respiration and heterotrophic respiration is measured separately, to develop a better mechanistic understanding of the impact of the 2010 drought on the Amazon forest. We find that total NPP remained constant throughout the drought. However, towards the end of the drought, autotrophic respiration, especially in roots and stems, declined significantly compared with measurements in 2009 made in the absence of drought, with extended decreases in autotrophic respiration in the three driest plots. In the year after the drought, total NPP remained constant but the allocation of carbon shifted towards canopy NPP and away from fine-root NPP. Both leaf-level and plot-level measurements indicate that severe drought suppresses photosynthesis. Scaling these measurements to the entire Amazon basin with rainfall data, we estimate that drought suppressed Amazon-wide photosynthesis in 2010 by 0.38 petagrams of carbon (0.23-0.53 petagrams of carbon). Overall, we find that during this drought, instead of reducing total NPP, trees prioritized growth by reducing autotrophic respiration that was unrelated to growth. This suggests that trees decrease investment in tissue maintenance and defence, in line with eco-evolutionary theories that trees are competitively disadvantaged in the absence of growth. We propose that weakened maintenance and defence investment may, in turn, cause the increase in post-drought tree mortality observed at our plots.

Greater stem growth, woody allocation, and aboveground biomass in Paleotropical forests than in Neotropical forests.

Forest dynamics and tree species composition vary substantially between Paleotropical and Neotropical forests, but these broad biogeographic regions are treated uniformly in many land models. To assess whether these regional differences translate into variation in productivity and carbon (C) storage, we compiled a database of climate, tree stem growth, litterfall, aboveground net primary production (ANPP), and aboveground biomass across tropical rainforest sites spanning 33 countries throughout Central and South America, Asia, and Australasia, but excluding Africa due to a paucity of available data. Though the sum of litterfall and stem growth (ANPP) did not differ between regions, both stem growth and the ratio of stem growth to litterfall were higher in Paleotropical forests compared to Neotropical forests across the full observed range of ANPP. Greater C allocation to woody growth likely explains the much larger aboveground biomass estimates in Paleotropical forests (~29%, or ~80 Mg DW/ha, greater than in the Neotropics). Climate was similar in Paleo- and Neotropical forests, thus the observed differences in C likely reflect differences in the evolutionary history of species and forest structure and function between regions. Our analysis suggests that Paleotropical forests, which can be dominated by tall-statured Dipterocarpaceae species, may be disproportionate hotspots for aboveground C storage. Land models typically treat these distinct tropical forests with differential structures as a single functional unit, but our findings suggest that this may overlook critical biogeographic variation in C storage potential among regions.

Megafauna and ecosystem function from the Pleistocene to the Anthropocene.

Large herbivores and carnivores (the megafauna) have been in a state of decline and extinction since the Late Pleistocene, both on land and more recently in the oceans. Much has been written on the timing and causes of these declines, but only recently has scientific attention focused on the consequences of these declines for ecosystem function. Here, we review progress in our understanding of how megafauna affect ecosystem physical and trophic structure, species composition, biogeochemistry, and climate, drawing on special features of PNAS and Ecography that have been published as a result of an international workshop on this topic held in Oxford in 2014. Insights emerging from this work have consequences for our understanding of changes in biosphere function since the Late Pleistocene and of the functioning of contemporary ecosystems, as well as offering a rationale and framework for scientifically informed restoration of megafaunal function where possible and appropriate.

Global nutrient transport in a world of giants.

The past was a world of giants, with abundant whales in the sea and large animals roaming the land. However, that world came to an end following massive late-Quaternary megafauna extinctions on land and widespread population reductions in great whale populations over the past few centuries. These losses are likely to have had important consequences for broad-scale nutrient cycling, because recent literature suggests that large animals disproportionately drive nutrient movement. We estimate that the capacity of animals to move nutrients away from concentration patches has decreased to about 8% of the preextinction value on land and about 5% of historic values in oceans. For phosphorus (P), a key nutrient, upward movement in the ocean by marine mammals is about 23% of its former capacity (previously about 340 million kg of P per year). Movements by seabirds and anadromous fish provide important transfer of nutrients from the sea to land, totalling ∼150 million kg of P per year globally in the past, a transfer that has declined to less than 4% of this value as a result of the decimation of seabird colonies and anadromous fish populations. We propose that in the past, marine mammals, seabirds, anadromous fish, and terrestrial animals likely formed an interlinked system recycling nutrients from the ocean depths to the continental interiors, with marine mammals moving nutrients from the deep sea to surface waters, seabirds and anadromous fish moving nutrients from the ocean to land, and large animals moving nutrients away from hotspots into the continental interior.

Impacts of fire on sources of soil CO2 efflux in a dry Amazon rain forest.

Fire at the dry southern margin of the Amazon rainforest could have major consequences for regional soil carbon (C) storage and ecosystem carbon dioxide (CO2 ) emissions, but relatively little information exists about impacts of fire on soil C cycling within this sensitive ecotone. We measured CO2 effluxes from different soil components (ground surface litter, roots, mycorrhizae, soil organic matter) at a large-scale burn experiment designed to simulate a severe but realistic potential future scenario for the region (Fire plot) in Mato Grosso, Brazil, over 1 year, and compared these measurements to replicated data from a nearby, unmodified Control plot. After four burns over 5 years, soil CO2 efflux (Rs ) was ~5.5 t C ha-1  year-1 lower on the Fire plot compared to the Control. Most of the Fire plot Rs reduction was specifically due to lower ground surface litter and root respiration. Mycorrhizal respiration on both plots was around ~20% of Rs . Soil surface temperature appeared to be more important than moisture as a driver of seasonal patterns in Rs at the site. Regular fire events decreased the seasonality of Rs at the study site, due to apparent differences in environmental sensitivities among biotic and abiotic soil components. These findings may contribute toward improved predictions of the amount and temporal pattern of C emissions across the large areas of tropical forest facing increasing fire disturbances associated with climate change and human activities.

Temperature and rainfall interact to control carbon cycling in tropical forests.

Tropical forests dominate global terrestrial carbon (C) exchange, and recent droughts in the Amazon Basin have contributed to short-term declines in terrestrial carbon dioxide uptake and storage. However, the effects of longer-term climate variability on tropical forest carbon dynamics are still not well understood. We synthesised field data from more than 150 tropical forest sites to explore how climate regulates tropical forest aboveground net primary productivity (ANPP) and organic matter decomposition, and combined those data with two existing databases to explore climate - C relationships globally. While previous analyses have focused on the effects of either temperature or rainfall on ANPP, our results highlight the importance of interactions between temperature and rainfall on the C cycle. In cool forests (< 20 °C), high rainfall slowed rates of C cycling, but in warm tropical forests (> 20 °C) it consistently enhanced both ANPP and decomposition. At the global scale, our analysis showed an increase in ANPP with rainfall in relatively warm sites, inconsistent with declines in ANPP with rainfall reported previously. Overall, our results alter our understanding of climate - C cycle relationships, with high precipitation accelerating rates of C exchange with the atmosphere in the most productive biome on earth.

Leaf aging of Amazonian canopy trees as revealed by spectral and physiochemical measurements.

Leaf aging is a fundamental driver of changes in leaf traits, thereby regulating ecosystem processes and remotely sensed canopy dynamics. We explore leaf reflectance as a tool to monitor leaf age and develop a spectra-based partial least squares regression (PLSR) model to predict age using data from a phenological study of 1099 leaves from 12 lowland Amazonian canopy trees in southern Peru. Results demonstrated monotonic decreases in leaf water (LWC) and phosphorus (Pmass ) contents and an increase in leaf mass per unit area (LMA) with age across trees; leaf nitrogen (Nmass ) and carbon (Cmass ) contents showed monotonic but tree-specific age responses. We observed large age-related variation in leaf spectra across trees. A spectra-based model was more accurate in predicting leaf age (R2  = 0.86; percent root mean square error (%RMSE) = 33) compared with trait-based models using single (R2  = 0.07-0.73; %RMSE = 7-38) and multiple (R2  = 0.76; %RMSE = 28) predictors. Spectra- and trait-based models established a physiochemical basis for the spectral age model. Vegetation indices (VIs) including the normalized difference vegetation index (NDVI), enhanced vegetation index 2 (EVI2), normalized difference water index (NDWI) and photosynthetic reflectance index (PRI) were all age-dependent. This study highlights the importance of leaf age as a mediator of leaf traits, provides evidence of age-related leaf reflectance changes that have important impacts on VIs used to monitor canopy dynamics and productivity and proposes a new approach to predicting and monitoring leaf age with important implications for remote sensing.

The variation of productivity and its allocation along a tropical elevation gradient: a whole carbon budget perspective.

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

The linkages between photosynthesis, productivity, growth and biomass in lowland Amazonian forests.

Understanding the relationship between photosynthesis, net primary productivity and growth in forest ecosystems is key to understanding how these ecosystems will respond to global anthropogenic change, yet the linkages among these components are rarely explored in detail. We provide the first comprehensive description of the productivity, respiration and carbon allocation of contrasting lowland Amazonian forests spanning gradients in seasonal water deficit and soil fertility. Using the largest data set assembled to date, ten sites in three countries all studied with a standardized methodology, we find that (i) gross primary productivity (GPP) has a simple relationship with seasonal water deficit, but that (ii) site-to-site variations in GPP have little power in explaining site-to-site spatial variations in net primary productivity (NPP) or growth because of concomitant changes in carbon use efficiency (CUE), and conversely, the woody growth rate of a tropical forest is a very poor proxy for its productivity. Moreover, (iii) spatial patterns of biomass are much more driven by patterns of residence times (i.e. tree mortality rates) than by spatial variation in productivity or tree growth. Current theory and models of tropical forest carbon cycling under projected scenarios of global atmospheric change can benefit from advancing beyond a focus on GPP. By improving our understanding of poorly understood processes such as CUE, NPP allocation and biomass turnover times, we can provide more complete and mechanistic approaches to linking climate and tropical forest carbon cycling.

Herbivory makes major contributions to ecosystem carbon and nutrient cycling in tropical forests.

The functional role of herbivores in tropical rainforests remains poorly understood. We quantified the magnitude of, and underlying controls on, carbon, nitrogen and phosphorus cycled by invertebrate herbivory along a 2800 m elevational gradient in the tropical Andes spanning 12°C mean annual temperature. We find, firstly, that leaf area loss is greater at warmer sites with lower foliar phosphorus, and secondly, that the estimated herbivore-mediated flux of foliar nitrogen and phosphorus from plants to soil via leaf area loss is similar to, or greater than, other major sources of these nutrients in tropical forests. Finally, we estimate that herbivores consume a significant portion of plant carbon, potentially causing major shifts in the pattern of plant and soil carbon cycling. We conclude that future shifts in herbivore abundance and activity as a result of environmental change could have major impacts on soil fertility and ecosystem carbon sequestration in tropical forests.

Allocation trade-offs dominate the response of tropical forest growth to seasonal and interannual drought.

What determines the seasonal and interannual variation of growth rates in trees in a tropical forest? We explore this question with a novel four-year high-temporal-resolution data set of carbon allocation from two forest plots in the Bolivian Amazon. The forests show strong seasonal variation in tree wood growth rates, which are largely explained by shifts in carbon allocation, and not by shifts in total productivity. At the deeper soil plot, there was a clear seasonal trade-off between wood and canopy NPP, while the shallower soils plot showed a contrasting seasonal trade-off between wood and fine roots. Although a strong 2010 drought reduced photosynthesis, NPP remained constant and increased in the six-month period following the drought, which indicates usage of significant nonstructural carbohydrate stores. Following the drought, carbon allocation increased initially towards the canopy, and then in the following year, allocation increased towards fine-root production. Had we only measured woody growth at these sites and inferred total NPP, we would have misinterpreted both the seasonal and interannual responses. In many tropical forest ecosystems, we propose that changing tree growth rates are more likely to reflect shifts in allocation rather than changes in overall productivity. Only a whole NPP allocation perspective can correctly interpret the relationship between changes in growth and changes in productivity.

The sixth R: Revitalizing the natural phosphorus pump.

Phosphorus (P) is essential for all life on Earth and sustains food production. Yet, the easily accessible deposits of phosphate-rich rock, which underpin the green revolution are becoming rarer. Here we propose a mechanism to help alleviate the problem of "peak phosphorus". In the past, wild animals played a large role in returning P from ocean depths back to the continental interiors. In doing so, they collectively retained and redistributed P within the biosphere, supporting a more fertile planet. However, species extinctions and population reductions have reduced animal-mediated P transport >90% over the past 12,000 years. Recently a 5R strategy was developed to Realign P inputs, Reduce P losses, Recycle P in bio-resources, Recover P in wastes, and Redefine P in food systems. Here, we suggest a sixth R, to Revitalize the Natural Phosphorus Pump (RNPP). Countries are starting to mandate P recycling and we propose a P-trading scheme based on REDD+, where a country could partially achieve its recycling goals by restoring past animal-mediated P pathways. Accrued money from this scheme could be used to restore or conserve wild animal populations, while increasing natural P recycling.

Predicting tropical plant physiology from leaf and canopy spectroscopy.

A broad regional understanding of tropical forest leaf photosynthesis has long been a goal for tropical forest ecologists, but it has remained elusive due to difficult canopy access and high species diversity. Here we develop an empirical model to predict sunlit, light-saturated, tropical leaf photosynthesis using leaf and simulated canopy spectra. To develop this model, we used partial least squares (PLS) analysis on three tropical forest datasets (159 species), two in Hawaii and one at the biosphere 2 laboratory (B2L). For each species, we measured light-saturated photosynthesis (A), light and CO(2) saturated photosynthesis (A(max)), respiration (R), leaf transmittance and reflectance spectra (400-2,500 nm), leaf nitrogen, chlorophyll a and b, carotenoids, and leaf mass per area (LMA). The model best predicted A [r(2) = 0.74, root mean square error (RMSE) = 2.9 µmol m(-2) s(-1))] followed by R (r(2) = 0.48), and A(max) (r(2) = 0.47). We combined leaf reflectance and transmittance with a canopy radiative transfer model to simulate top-of-canopy reflectance and found that canopy spectra are a better predictor of A (RMSE = 2.5 ± 0.07 µmol m(-2) s(-1)) than are leaf spectra. The results indicate the potential for this technique to be used with high-fidelity imaging spectrometers to remotely sense tropical forest canopy photosynthesis.

The megabiota are disproportionately important for biosphere functioning.

A prominent signal of the Anthropocene is the extinction and population reduction of the megabiota-the largest animals and plants on the planet. However, we lack a predictive framework for the sensitivity of megabiota during times of rapid global change and how they impact the functioning of ecosystems and the biosphere. Here, we extend metabolic scaling theory and use global simulation models to demonstrate that (i) megabiota are more prone to extinction due to human land use, hunting, and climate change; (ii) loss of megabiota has a negative impact on ecosystem metabolism and functioning; and (iii) their reduction has and will continue to significantly decrease biosphere functioning. Global simulations show that continued loss of large animals alone could lead to a 44%, 18% and 92% reduction in terrestrial heterotrophic biomass, metabolism, and fertility respectively. Our findings suggest that policies that emphasize the promotion of large trees and animals will have disproportionate impact on biodiversity, ecosystem processes, and climate mitigation.

Animals and the zoogeochemistry of the carbon cycle.

Predicting and managing the global carbon cycle requires scientific understanding of ecosystem processes that control carbon uptake and storage. It is generally assumed that carbon cycling is sufficiently characterized in terms of uptake and exchange between ecosystem plant and soil pools and the atmosphere. We show that animals also play an important role by mediating carbon exchange between ecosystems and the atmosphere, at times turning ecosystem carbon sources into sinks, or vice versa. Animals also move across landscapes, creating a dynamism that shapes landscape-scale variation in carbon exchange and storage. Predicting and measuring carbon cycling under such dynamism is an important scientific challenge. We explain how to link analyses of spatial ecosystem functioning, animal movement, and remote sensing of animal habitats with carbon dynamics across landscapes.

Tropical forest leaves may darken in response to climate change.

Tropical forest leaf albedo (reflectance) greatly impacts how much energy the planet absorbs; however; little is known about how it might be impacted by climate change. Here, we measure leaf traits and leaf albedo at ten 1-ha plots along a 3,200-m elevation gradient in Peru. Leaf mass per area (LMA) decreased with warmer temperatures along the elevation gradient; the distribution of LMA was positively skewed at all sites indicating a shift in LMA towards a warmer climate and future reduced tropical LMA. Reduced LMA was significantly (P < 0.0001) correlated with reduced leaf near-infrared (NIR) albedo; community-weighted mean NIR albedo significantly (P < 0.01) decreased as temperature increased. A potential future 2 °C increase in tropical temperatures could reduce lowland tropical leaf LMA by 6-7 g m-2 (5-6%) and reduce leaf NIR albedo by 0.0015-0.002 units. Reduced NIR albedo means that leaves are darker and absorb more of the Sun's energy. Climate simulations indicate this increased absorbed energy will warm tropical forests more at high CO2 conditions with proportionately more energy going towards heating and less towards evapotranspiration and cloud formation.

ENSO Drives interannual variation of forest woody growth across the tropics.

Meteorological extreme events such as El Niño events are expected to affect tropical forest net primary production (NPP) and woody growth, but there has been no large-scale empirical validation of this expectation. We collected a large high-temporal resolution dataset (for 1-13 years depending upon location) of more than 172 000 stem growth measurements using dendrometer bands from across 14 regions spanning Amazonia, Africa and Borneo in order to test how much month-to-month variation in stand-level woody growth of adult tree stems (NPPstem) can be explained by seasonal variation and interannual meteorological anomalies. A key finding is that woody growth responds differently to meteorological variation between tropical forests with a dry season (where monthly rainfall is less than 100 mm), and aseasonal wet forests lacking a consistent dry season. In seasonal tropical forests, a high degree of variation in woody growth can be predicted from seasonal variation in temperature, vapour pressure deficit, in addition to anomalies of soil water deficit and shortwave radiation. The variation of aseasonal wet forest woody growth is best predicted by the anomalies of vapour pressure deficit, water deficit and shortwave radiation. In total, we predict the total live woody production of the global tropical forest biome to be 2.16 Pg C yr-1, with an interannual range 1.96-2.26 Pg C yr-1 between 1996-2016, and with the sharpest declines during the strong El Niño events of 1997/8 and 2015/6. There is high geographical variation in hotspots of El Niño-associated impacts, with weak impacts in Africa, and strongly negative impacts in parts of Southeast Asia and extensive regions across central and eastern Amazonia. Overall, there is high correlation (r = -0.75) between the annual anomaly of tropical forest woody growth and the annual mean of the El Niño 3.4 index, driven mainly by strong correlations with anomalies of soil water deficit, vapour pressure deficit and shortwave radiation.This article is part of the discussion meeting issue 'The impact of the 2015/2016 El Niño on the terrestrial tropical carbon cycle: patterns, mechanisms and implications'.