A cc ep te d A rt ic le This article has been accepted for publication and undergone full peer review but has not been through the copyediting, typesetting, pagination and proofreading process, which may lead to differences between this version and the Version of Record. Please cite this article as doi: 10.1111/gcb.13226 This article is protected by copyright. All rights reserved. Received Date : 05-Oct-2015 Accepted Date : 08-Dec-2015 Article type : Research Review Title: Carbon dynamics of mature and regrowth tropical forests derived from a pantropical database (TropForC-db) Running Head: TropForC database Authors: Kristina J. Anderson-Teixeira1,2* Maria M. H. Wang1 Jennifer C. McGarvey1 David S. LeBauer3 Author Affiliations: 1. Conservation Ecology Center; Smithsonian Conservation Biology Institute; National Zoological Park, Front Royal, VA, USA 2. Center for Tropical Forest Science-Forest Global Earth Observatory; Smithsonian Tropical Research Institute; Panama, Republic of Panama 3. Carl Woese Institute for Genomic Biology, University of Illinois, Urbana, Il, USA A cc ep te d A rt ic le This article is protected by copyright. All rights reserved. *Corresponding Author: phone: 1-540-635-6546 fax:1-540-635-6506 email: teixeirak@si.edu Keywords: tropical forest; regeneration; secondary; intact; carbon cycle; biomass; productivity; net ecosystem exchange Paper type: Review Abstract Tropical forests play a critical role in the global carbon (C) cycle, storing ~45% of terrestrial C and constituting the largest component of the terrestrial C sink. Despite their central importance to the global C cycle, their ecosystem-level C cycles are not as well characterized as those of extra-tropical forests, and knowledge gaps hamper efforts to quantify C budgets across the tropics and to model tropical forest- climate interactions. To advance understanding of C dynamics of pantropical forests, we compiled a new database, the Tropical Forest C database (TropForC-db), which contains data on ground-based measurements of ecosystem-level C stocks and annual fluxes along with disturbance history. This database currently contains 3,568 records from 845 plots in 178 geographically distinct areas, making it the largest and most comprehensive database of its type. Using TropForC-db, we characterized C stocks and fluxes for young, intermediate-aged, and mature forests. Relative to existing C budgets of extra-tropical forests, mature tropical broadleaf evergreen forests had substantially higher gross primary productivity (GPP) and ecosystem respiration (Reco), their autotropic respiration (Ra) consumed a larger proportion (~67%) of GPP, and their woody stem growth (ANPPstem) represented a smaller proportion of net primary productivity (NPP, ~32%) or GPP (~9%). In regrowth stands, aboveground biomass increased rapidly during the first 20 years following stand-clearing disturbance, with slower accumulation following agriculture and in deciduous forests, and continued to accumulate at a slower pace in forests aged 20-100 years. Most other C stocks likewise increased with stand age, while potential to describe age trends in C fluxes was generally data-limited. We expect that TropForC-db will prove useful for model evaluation and for quantifying the contribution of forests to the global C cycle. The database version associated with this publication is archived in Dryad (DOI: 10.5061/dryad.t516f) and a dynamic version is maintained at https://github.com/forc-db. A cc ep te d A rt ic le This article is protected by copyright. All rights reserved. Introduction Tropical forests, including both regrowth and intact forests, play a critical role in the global carbon (C) cycle. They store an estimated 45% of terrestrial C and account for over one third of terrestrial gross primary production (GPP; Bonan, 2008; Beer et al., 2010). Tropical forests also constitute the largest component of the terrestrial C sink. In recent years (early 2000’s), forest regrowth on ~557 Mha of abandoned agricultural land in tropical regions has represented an estimated sink of 1.4-1.7 Pg C yr-1 (Pan et al., 2011; Grace et al., 2014; Lewis et al., 2015)—an amount equal to ~20% of annual fossil fuel emissions or over half of the estimated terrestrial land sink over a similar period (2000-2009; Le Quéré et al., 2013). At the same time, intact tropical forests have on average been increasing in biomass over recent decades (Phillips, 1998; Lewis et al., 2009a; Muller-Landau et al., 2014), sequestering an estimated 0.5-1.0 Pg C yr-1 in the early 2000’s (Pan et al., 2011; Grace et al., 2014; Lewis et al., 2015). Moreover, natural disturbance-recovery cycles result in substantial ecosystem-atmosphere CO2 exchange. For instance, in the Amazon alone, natural disturbances release an estimated 1.3 Pg C yr-1, which is more than compensated for by CO2 sequestration through tree growth (Espírito-Santo et al., 2014). Thus, CO2 exchanges between the tropical forest biome and the atmosphere meaningfully influence atmospheric CO2. In the present era of global change, tropical forests play a central role in determining the rate of increase in atmospheric CO2. Tropical deforestation is of key significance; from 1990 to 2007, CO2 emissions from tropical deforestation were ~3 Pg C yr-1, equivalent to ~40% of global fossil fuel emissions (Pan et al., 2011). Efforts to reduce tropical deforestation (e.g., REDD+; UNFCCC, 2008, 2015), if successful, will contribute substantively to reduction of anthropogenic CO2 emissions (Houghton et al., 2015). At the same time, tropical forests are changing in response to climate change and other global change pressures, and this will alter their CO2 exchange with the atmosphere (e.g., Malhi et al., 2014; Anderson-Teixeira et al., 2015). Although currently C sinks, intact tropical forests could become net C sources if, for example, drought and other disturbances substantially increase tree mortality (e.g., Lewis et al., 2011; Brienen et al., 2015). Climate change is likely to increase the frequency and intensity of some natural disturbances (e.g., storms; droughts; IPCC, 2013; Trenberth et al., 2014), and regional C balances will be strongly influenced by tropical forest regrowth dynamics, which are also likely to be altered by climate change (Anderson-Teixeira et al., 2013). Altered disturbance-recovery dynamics have the potential to have a much stronger influence on regional C balances than metabolically-driven changes (Kurz et al., 2008; Running, 2008; Anderson-Teixeira et al., 2013). The net response of the tropical forest biome to global change pressures will influence the future trajectory of atmospheric CO2, yet remains highly uncertain. Despite the importance of tropical forests to the global C cycle, their C cycles are not as well understood as those of extra-tropical forests, and important gaps in our knowledge of their ecosystem-level C cycles hamper scientific and societal efforts to quantify C budgets across the tropics and to model tropical forest-climate interactions. More data are required to understand C cycling in tropical forest ecosystems, how it compares to C cycling in extra-tropical forests, and how A cc ep te d A rt ic le This article is protected by copyright. All rights reserved. it is influenced by environmental variation (U.S. DOE, 2012; Bustamante et al., 2015; Malhi et al., 2015). Modeling the role of tropical forests in Earth’s changing climate system presents a significant challenge (U.S. DOE, 2012), and regrowth forests in particular remain poorly represented in Earth system models (ESMs; Arora & Boer, 2010; Schwalm et al., 2010). Improved data on how C stocks and fluxes of tropical forests are affected by disturbance history and climate is needed to parameterize, test, and validate ESMs (Friedlingstein et al. 1999; Bonan 2008; Schwalm et al. 2010). Moreover, high uncertainty regarding C stocks and fluxes of tropical forests— particularly regrowth forests—introduces substantial uncertainty into the global forest C balance (Pan et al., 2011). On a practical level, national inventories of C stocks and fluxes are central to international climate change mitigation efforts (e.g., greenhouse gas accounting under the Kyoto Protocol), yet the IPCC guidelines for national greenhouse gas inventories (IPCC, 2006) present estimates of tropical forest C stocks and accumulation rates that are based on a very small subset of available data. Improved data on C stocks and fluxes of tropical forests are therefore key to more accurate quantification of the role of tropical forests the global C cycle. Pantropical data on C stocks and fluxes of tropical forests are critical to understanding the C dynamics of tropical forests, how these compare to the better-characterized C dynamics of extra- tropical forests, and the role of tropical forests in the global C cycle. Relevant data have been collected at many locations throughout the world (Table 1), yet appropriate synthesis has been lacking. Recent global compilations of forest C data are scant on tropical forests (Luyssaert et al., 2007; Liu et al., 2014; Michaletz et al., 2014)—in part because tropical data are relatively less abundant, but also because there has not been a focused effort to identify and incorporate relevant tropical forest data. Moreover, most existing forest C databases have limited information about disturbance history—information that is key to analyzing trajectories of forest recovery. Here, we present a new database on C dynamics of tropical forests, the Tropical Forest C database (TropForC-db) and use it to synthesize knowledge to date about tropical forest C dynamics and identify key uncertainties. TropForC-db synthesizes data on ground-based measurements of C stocks and flows along with site disturbance history for forests throughout the tropics. It focuses on C stocks and annual C fluxes, drawing upon existing data compilations and data from original studies identified using a literature search. We use this database, which is the largest and most comprehensive of its type, to characterize C stocks and annual fluxes for young, intermediate-aged, and mature/ intact forests—comparing the latter to existing budgets for extra- tropical forests—and to examine patterns of C accumulation following stand-clearing disturbance. Materials & Methods Overview of TropForC-db TropForC-db is the tropical component of ForC-db, which is a global C database that is currently under development and maintained at https://github.com/forc-db. Its structure is derived from and compatible with that of BETY-db (www.betydb.org; LeBauer et al., 2010). In brief, the database consists of a series of cross-referenced data tables describing (1) sites, (2) plots and A cc ep te d A rt ic le This article is protected by copyright. All rights reserved. their history, (3) measurements of C cycle variables, (4) variables, (5) disturbance/history event type, (6) plant functional types (PFTs)/ species, (7) methodologies, and (8) allometries (Table 2). Records of plot locations within the database were designed to preserve maximum possible location information given in original publications and to allow grouping of related plots. Specifically, plots were grouped by site and area, where site is the most precisely described location with unique site conditions (e.g., latitude/ longitude, elevation, edaphic conditions), and area groups one or more geographically proximate plots. Each plot was described in terms of its history and dominant vegetation. Plot most commonly refers to a contiguous sampling area; however, when an original study presented only average values for non-contiguous replicate plots, they are treated as a single plot within our database. Depending on the level of site detail given by the original publication, a plot may be treated as a unique site or may share site data with other plots. Geographically proximate sites were grouped into areas, where area was defined as a group of sites where no site is >0.25° latitude or longitude distant from another site in the group. This groups chronosequences—i.e., plots differing in time since a stand-clearing disturbance—within a single area. Thus, “areas” group plots suitable for direct comparison, whereas “sites” link plots to the most precisely described geographic, climatic, and edaphic data. Each plot was described in terms of known history of events affecting the entire plot and relevant to understanding the C cycle. Events recorded included major natural and anthropogenic disturbances (e.g., fires, major storms, harvest, tillage), initiation of forest growth (e.g., initiation of natural succession, planting), management (e.g., fertilization, thinning), and experimental manipulation (e.g., irrigation). Smaller-scale natural disturbances such as tree fall were not included. We used “stand age” to refer to the age of the oldest cohort of trees within the plot; however, we note that the database does include records from some large, heterogeneous plots containing multiple stands that may vary in age (e.g., CTFS-ForestGEO plots; Anderson-Teixeira et al., 2015). Thus, some mature /intact plots contain, but are not dominated by, stands of younger age. When not reported directly, stand age was estimated based on the year of initiation of forest regrowth. When stand age was reported, we used it to calculate the year of establishment of the oldest trees within a plot, which was recorded as part of the plot’s history. Particularly for older stands, this may differ from the year of initiation of forest regrowth following disturbance. Measurements of ecosystem-level C stocks and annual fluxes were included. Most variables were defined as in Chapin et al. (2006) or Luyssaert et al. (2007). Definitions of the variables presented here are given in Tables 3-4, and complete definitions of all variables and equations relating the variables are included in the data files (Anderson-Teixeira et al. 2016). Measurement records included the sites and plots at which measurements were made, dominant plant functional type or species (in the case of plantations), stand age, measurement methods and dates, reported measurement values and error, and covariates important to interpretation of the measurement (e.g., minimum DBH, allometries used; Table 2). A single plot was commonly linked to measurements of multiple variables, and a single variable could be measured multiple times in the same plot. Complete metadata are given in an associated data publication in Dryad Digital Repository (http://dx.doi.org/10.5061/dryad.t516f; Anderson-Teixeira et al., 2016). A cc ep te d A rt ic le This article is protected by copyright. All rights reserved. Data Sources We used previous data compilations on ecosystem-level C stocks and annual fluxes for regrowth or mature tropical forests to identify original publications relevant to the database (Table 1). We referred to the original publications to check data presented in these original compilations and to obtain additional information (e.g., plot history). In addition, in fall 2013- spring 2014, we used Google Scholar to search the literature for additional studies focused on regrowth forests and those quantifying dead wood. Data compilations and individual studies included in the database are listed in Table 1. The search for relevant data was substantive but not comprehensive; we are aware of relevant data that were not included in the database as of September 2015. Plots were included in TropForC-db if located within the tropics (latitude ≤ 23.5°) and tree- dominated (including savannas with woody vegetation). Peat forests were included. The data compilation effort focused on unmanaged forests, but managed or plantation forests were included when their data were included in a previous compilation. We focused the search around ecosystem-level measurements of biomass, dead wood, and major annual C fluxes (Tables 3-4), also including a number of other relevant variables when reported in studies including data on focal variables (see metadata of Anderson-Teixeira et al. 2016 for complete list). While the database does include soil carbon data, we did not attempt to make a comprehensive compilation of soil carbon data. Marín-Spiotta & Sharma (2013) provides a recent compilation of pan-tropical data on soil carbon in forests of different ages and disturbance histories. Data were obtained from tables or extracted from figures using WebPlotDigitizer v.3.8 (Rohatgi, 2015). All data were converted to standardized units: Mg [dry biomass or C] ha-1 for stocks and Mg [dry biomass or C] ha-1 yr-1 for fluxes. The database preserves measurements in biomass or C as reported by the original study. For analyses presented here, we converted biomass to C using the approximation that biomass is 47% C (IPCC, 2006). Additional geographic and climatic data for all sites were extracted from global databases. Biogeographic zones were delineated using the map of Olson et al. (2001), and FAO ecozone classification was obtained from FAO’s GeoNetwork (http://www.fao.org:80/geonetwork). Climate zone was extracted from the ESRI Köppen-Geiger map (downloaded June 2014 from http://maps3.arcgisonline.com/ArcGIS/rest/services/A-16/Köppen- Geiger_Observed_and_Predicted_Climate_Shifts/MapServer). Analyses For the purpose of the descriptive statistics reported here, we grouped forests into three age classes determined based on benchmarks related to biomass recovery and alignment with existing international standards (IPCC, 2006): young (age ≤20), intermediate-aged (20100 plots included stem aboveground net primary productivity (ANPPstem), total biomass, foliage biomass, total root biomass, and dead wood (including standing dead wood and coarse woody debris; Tables 3-4). There were 26 variables measured in ≥ 25 plots, and 21 variables with fewer records (Anderson-Teixeira et al., 2016). Mature/intact and young forests were better represented in terms of the number of variables measured (36 variables each), while intermediate-aged stands had records for fewer variables (26 variables). The three age classes were approximately equally represented in terms of aboveground biomass records (34% young; 33.5% intermediate-aged, 25.6% mature/intact; remainder age unknown). There was very low representation of C fluxes in young and intermediate-aged stands (17.9% and 6.7% of C flux records, respectively); for instance, the database contains only one set of eddy flux measurements A cc ep te d A rt ic le This article is protected by copyright. All rights reserved. in naturally regenerating tropical forest (a young stand) and only one measurement each of belowground and total NPP in naturally regenerating intermediate-aged tropical forest stands. C stocks and fluxes of mature forests Carbon stocks and fluxes of mature/ intact unmanaged tropical broadleaf evergreen forests are summarized in Figure 3 and are also presented in Tables S4 (stocks) and S7 (fluxes). This ensemble C budget was internally consistent; that is, none of the C fluxes shown in Fig. 3 differed significantly from the sum of its component C flux terms (Fig. 3; analyses not shown). Descriptive statistics for all mature/intact forests (not limited to unmanaged broadleaf evergreen) are presented in the Supplementary Information (Tables S2-7). C cycling in regrowth forests C stocks commonly exhibited significant age trends both across and within age categories (Figs. 3-6). Trends were similar when all forests were included and when analyses were limited to unmanaged broadleaf evergreen forests; here, we present results for the latter; those for all forests are presented in Tables S2-S4. Aboveground, total root, and total biomass all accumulated as forests aged and were highest on average in mature forest stands, with at least marginally significantly differences (all p<0.07) in all pairwise comparisons among young, intermediate and mature age classes (Figs. 3-5; Tables S2-S4). Foliage biomass accumulated with age within the young forest age class: foliage C = 0.9 ± 0.8 + age × 0.3 ± 0.03 (p < 0.001; n = 41 records from 29 plots in 7 areas), but did not vary significantly with age in intermediate-aged forests. Mean foliage biomass of mature forests was significantly higher than those of young forests (4.2 ± 0.4 Mg C ha-1 yr-1; n=18 records from 14 plots in 5 areas), while that of intermediate forests did not significantly differ from either young or mature forests (Tables S2-S4). Fine root biomass differed significantly among age classes (p < 0.05, mixed effects model with age class as the only fixed effect and plot nested within area as the random effect); however, pairwise comparison of plot means between age classes using t-tests were non-significant (Tables S2-S4). Aboveground biomass increased with stand age (Tables S2-S4; S8-S11), and was also significantly affected by regrowth type and FAO ecozone in young stands (Fig. 6a; Tables S8-S11). Specifically, in young stands, aboveground biomass increased with age (p < 0.001), and there was a significant age ✕ regeneration type interaction (p < 0.001), with the steepest slope for plantations and the shallowest slope following cultivation (Fig. 6a; Tables S8, S9). When the analysis was limited to naturally regenerating forests (i.e., plantations excluded), aboveground biomass in young stands was significantly influenced by age (p<0.001) and its interactions with both regeneration type (p = 0.02) and FAO ecozone (p = 0.04; Table S8), where biomass accumulated more rapidly in tropical rainforests and mountain systems than in moist deciduous forests (Table S10). For intermediate-aged stands, aboveground biomass again increased with stand age (p < 0.001), but was not significantly influenced by its interactions with regeneration type (natural regeneration A cc ep te d A rt ic le This article is protected by copyright. All rights reserved. following cultivation or pasture) or FAO ecozone (tropical rainforest or moist deciduous forest; all p > 0.5; Table S8). Dead wood C also varied with stand age (Fig. 6b). It was highly variable in young stands, and there was no significant effect of age within the first 20 years of stand development. In intermediate-aged stands, stocks were lower and less variable, but increased with stand age (dead wood C = 0.6 ± 2.3 + age × 0.1 ± 0.04; p = 0.02). In mature stands, dead wood C was higher— averaging 18 ± 4 Mg C ha-1— and quite variable, ranging from 7 to 46 Mg C ha-1 (n=9 records from 9 plots in 5 areas). Potential to describe age trends in C fluxes was limited by the small number of records of C flux for young and intermediate-aged stands (Figs. 3-5; Tables S1, S5-S7). The database contained only one NEE record for an unmanaged regrowth stand (4 years post-fire). This forest was a stronger C sink (NEE= -4.4 Mg C ha-1 yr-1, where negative sign indicates C sink) than was typical for mature stands (average -1.85 ± 0.69), but not outside the range of observed values (-5.6 to +1.2 Mg C ha-1 yr-1). There were a few significant trends in components of ANPP. First, ANPPstem in unmanaged broadleaf evergreen forests declined across the age classes, with a significant difference between young and mature stands (Tables S5-S7): young stands averaged 3.9 ± 0.5 Mg C ha-1 yr-1, intermediate-aged stands averaged 2.5 ± 0.4 Mg C ha-1 yr-1, and mature stands averaged 2.7 ± 0.1 Mg C ha-1 yr-1 (Figs. 3-5). Second, ANPP increased with age in young unmanaged broadleaf evergreen stands: -1.44 ± 3.79 + age x 1.34 ± 0.33 (p = 0.02). Finally, ANPPlitterfall did not differ among age classes (Tables S5-S7), but did increase with age within the intermediate-age category (Tables S5-S7). There was no detectable age trend in GPP, NPP, foliage ANPP, BNPP, fine root productivity, or soil or ecosystem respiration for unmanaged broadleaf evergreen forests (Tables S5-S7). Discussion Trop ForC-db and the status of tropical forest C data coverage TropForC-db is the largest existing compilation of data on ground-based measurements of C stocks and fluxes in tropical forests. As such, it is valuable for new and more comprehensive analyses of C cycling in mature and regrowth tropical forests, some of which are included here (e.g., Figs 3-6; Tables S2-S7). While the database represents a substantive effort to assemble existing data on tropical forest C stocks and fluxes, it is by no means complete in terms of including all available and relevant data. We encourage future studies that draw upon TropForC-db to seek out additional available data for the variables or forest types of interest and to contribute these data to the database at https://github.com/forc-db. While the > 3,000 records included in TropForC-db represent a substantial body of data on tropical forest C stocks and fluxes, there remain important limitations in terms of data coverage, standardization, and uncertainty. Aboveground biomass is the variable with most records (Tables 3-4), yet coverage within TropForC-db remains sparse for many parts of the tropics—particularly for deciduous forests and parts of Africa, Indomalaya, and Oceania (Fig. 1; Table S1). Moreover, A cc ep te d A rt ic le This article is protected by copyright. All rights reserved. biomass estimates are highly influenced by allometries, which are rarely available on a species- or region-specific basis in the tropics (Chave et al., 2014). Measurements of other C stocks are more sparsely distributed and are also associated with sometimes-high uncertainty and lack of methods standardization; for example, root biomass measurements are labor-intensive and differ in sampling design, root size cutoffs, and sampling depth. Measurements of NPP and components thereof are more sparsely distributed. Moreover, there is variability in terms of how the components of NPP are measured and which are included in estimates of total NPP, ANPP, or BNPP (e.g., see multiple NPP variables and associated definitions in Anderson-Teixeira et al. 2016; Luyssaert et al., 2007). Measuring NPP in forests is challenging; many measurement methods are associated with high uncertainty, and some smaller yet potentially significant components of NPP— e.g., herbivory, volatile organic C (VOC) production—are rarely quantified (Clark et al., 2001a). Finally, measurements of ecosystem-atmosphere CO2 exchange—NEE, GPP, and Reco—are particularly challenging in tropical forests (Kruijt et al., 2004), which are highly underrepresented in terms of eddy flux measurements (Schimel et al., 2015). For C flux variables, coverage is particularly sparse in regrowth forests. Future research aimed at filling some of these gaps will be of great value. C cycling in mature forests TropForC-db allows the most comprehensive analysis to date of C cycling in mature tropical forests, including an ensemble C budget for mature unmanaged broadleaf evergreen forests (Fig. 3). The observed closure of this ensemble C budget is not necessarily to be expected, given that the averages reported here are derived from different sets of plots, that a variety of methodologies are employed (often ignoring smaller component fluxes), and that C flux measurements involve many methodological challenges. In part, the observed closure is attributable to high variability in the data; nevertheless, it suggests that the averages presented here are broadly accurate (or, less parsimoniously, that systematic biases cancel). It is important to bear in mind that this ensemble C budget is unlikely to be completely accurate for any given stand; rather, there is substantial variation around these means. Thus, for best estimates of C stocks or fluxes for a particular forest type or for calculation of C allocation parameters, we recommend recomputation of values based on specifically selected subsets of the database—as opposed to reliance on the means presented here. Our ensemble estimates of C fluxes in mature unmanaged broadleaf evergreen forests (Fig. 3) are generally consistent with previous work characterizing tropical forest C budgets (e.g., Luyssaert et al., 2007a; Malhi, 2012). In terms of C fluxes, the only meaningful differences involve modest differences in GPP and net ecosystem productivity (NEP ≈ -NEE; see Chapin et al., 2006) from the means presented in Luyssaert et al. (2007), where our means are significantly lower than this previous average but are not significantly lower than the 25th percentile of observations. Specifically, our average GPP, 31.6 ± 1.6 Mg C ha-1 yr-1, is lower than Luyssaert et al.’s estimate of 35.6 Mg C ha-1 yr-1, and our average for NEP, 1.9 ± 0.7 Mg C ha-1 yr-1, is lower than Luyssaert et al.’s estimate of 4.1 Mg C ha-1 yr-1. Our revised estimate of average NEP is closer to pantropical averages for C stock changes, being larger but not significantly so. Specifically, average C stock changes for pantropical forests have been estimated at ~0.7 Mg C ha-1 yr-1 for 1990-2007 (Pan et al., 2011), and A cc ep te d A rt ic le This article is protected by copyright. All rights reserved. a pantropical weighted average for changes in C stocks of aboveground biomass alone is 0.34 Mg C ha-1 yr-1 (95% CI: 0.23- 0.45 Mg C ha-1 yr-1; Muller-Landau et al., 2014). If our average NEP is not upwardly biased, this indicates that intact tropical forests are a somewhat stronger C sink than recently estimated (Pan et al., 2011). The ensemble C budget of mature unmanaged broadleaf evergreen forests differs from those of extra-tropical forests in three important ways. First, gross C fluxes into and out of the ecosystem, GPP and Reco, are much larger than in extra-tropical forests, where they rarely exceed 20 Mg C ha-1 yr-1 (Luyssaert et al., 2007). The high productivity is driven primarily by longer growing seasons, as opposed to higher productivity during the growing season (e.g., Hirata et al., 2008; Malhi, 2012). Second, the ratio of NPP to GPP, or carbon use efficiency (CUE), is lower than average values observed for any forest biome globally. While the ratio of average NPP to average GPP is not the same as an average CUE, our NPP/GPP=0.28 is in line with CUE’s observed 0.27-0.46 throughout the tropics (Malhi, 2012). This is on the low end of what has been observed in forests globally (DeLucia et al., 2007; Litton et al., 2007; Luyssaert et al., 2007a; Campioli et al., 2015), but this finding is consistent with the fact that CUE tends to decline with both stand age and the ratio of leaf to total biomass (DeLucia et al., 2007). A number of mechanisms may cause this low CUE, including high respiratory costs associated with high temperatures, greater maintenance costs associated with large tree size, greater C costs of defense, or ‘idling respiration’ of excess C (Chambers et al., 2004; DeLucia et al., 2007; Malhi, 2012). Third, the proportion of NPP allocated to stem productivity (ANPPstem)—as opposed to root or leaf productivity (BNPP and ANPPfoliage, respectively)—is lower in the tropics than in other forest biomes (Luyssaert et al., 2007). Specifically, ANPPstem equaled ~32% of total NPP in our C budget (Fig. 3; slightly higher than the 25% estimates of Malhi et al., 2011 and Malhi, 2012), compared to values ranging from 34% in boreal semiarid evergreen forests to 76% in boreal humid evergreen forests (Luyssaert et al., 2007). This may reflect the fact that the database of Luyssaert et al. (2007) includes some regrowth forests, which tend to allocate more C to ANPPstem (Figs 3-5), or it may indicate greater proportional allocation to leaves and roots in the tropics. Together, lower CUE and lower ANPPstem:NPP mean that ANPPstem of unmanaged broadleaf evergreen forests is similar to that of many extra-tropical forests (Luyssaert et al., 2007; Huston & Wolverton, 2009). Thus, tropical forests metabolize far more C than extra-tropical forests, with most of the difference accounted for by high Ra and greater C allocation to functions other than woody growth. Observed patterns of C allocation within mature unmanaged broadleaf evergreen forests (Fig. 3) have important implications for making inferences about forest productivity from commonly measured variables. The most commonly measured C fluxes in tropical forests are the largest components of ANPP: woody productivity (ANPPstem) and litterfall (ANPPfoliage, ANPPlitterfall; Table 3). In mature unmanaged broadleaf evergreen forests, ANPPstem equals approximately 32% of NPP and 9% of GPP, ANPPlitterfall equals approximately 41% of NPP and 11% of GPP, and total ANPP equals approximately 57% of NPP and 16% of GPP (Fig. 3). Thus, measurements of ANPP and its components capture only a modest proportion of total productivity, and shifting C allocation in response to environmental variability implies that these are not good proxies for total productivity (Doughty et al., 2015; Malhi et al., 2015). We therefore caution against using these to infer A cc ep te d A rt ic le This article is protected by copyright. All rights reserved. responses of NPP or GPP to climatic variation. Nevertheless, understanding the responses of ANPPstem in particular to climatic variation is critical in that this C has a long residence time. C cycling in regrowth forests Aboveground biomass accumulation accounts for the majority of C sequestration by regrowth forests, and understanding the rate at which it accumulates is thereby critical to accurately characterizing the role of tropical regrowth forests in the global C cycle. The slopes of biomass-age mixed effects models are primarily derived from chronosequence data and as such do not necessarily equal to biomass accumulation rates—especially, for example, when the intercepts for young stands deviate substantively from zero. Therefore, these are not directly comparable to values used in global forest C inventories (e.g., Pan et al., 2011; Grace et al., 2014) or IPCC accounting (IPCC, 2006). However, our analyses yield insight into the relative influence of various factors on aboveground biomass accumulation rate. Among many variables examined—including climate variables, vegetation type, ecoregion, and biogeographic zone (analyses not shown)— regeneration type was the most important driver of regeneration rate (Fig. 6a). Specifically, in young forests, biomass accumulated most rapidly in plantations, as has been previously observed (Anderson et al., 2006; Bonner et al., 2013). Conversely, biomass accumulation was slowest following agricultural abandonment, which is consistent with previous work demonstrating that forest regrowth rate declines with increasing frequency and intensity of past agricultural disturbance (Uhl et al., 1988; Fearnside & Guimaraes, 1996; Hughes et al., 1999a; Steininger, 2000; Lawrence, 2005; Lawrence et al., 2010; Bonner et al., 2013; Mesquita et al., 2015). The importance of regeneration type suggests that future tropical forest C inventories could be improved by accounting for disturbance history (Fig. 6a; Bonner et al., 2013), along with seed availability or forest cover in the surrounding landscape (not analyzed here, but see Bonner et al., 2013; Mesquita et al., 2015). Our analysis revealed that biomass accumulation in naturally regenerating young stands was higher in tropical rainforests and tropical mountain systems (this category included many lowland forests in our dataset) than tropical deciduous forests (Table S10), which is in broad agreement with previous studies showing an effect of precipitation on biomass accumulation rate and with IPCC values (Brown & Lugo, 1982; IPCC, 2006; Marín-Spiotta et al., 2008; Poorter et al., 2016). In summary, aboveground biomass increases rapidly with stand age during the first 20 years following stand-clearing disturbance, with slower accumulation in deciduous forests and following agriculture. While biomass accumulation decelerates in intermediate-aged stands, there remains a significant effect of age on aboveground biomass for stands aged 20-100 years (Fig. 6a; Table S3). The ongoing accumulation of biomass in intermediate-aged stands implies that C inventories, which commonly lump regrowth forests >20 years old with mature/ intact forests (IPCC, 2006), could be improved by distinguishing older regrowth forests from mature/intact forests, when feasible. By age 100, regrowth forests on average have similar aboveground biomass to mature/ intact forests (Fig. 6a). However, the time required for biomass to recover to the levels of adjacent undisturbed forests is variable (Martin et al., 2013), and successional processes may continue to drive biomass A cc ep te d A rt ic le This article is protected by copyright. All rights reserved. accumulation even in forests with no known history of stand-clearing disturbance (Chave et al., 2008). Beyond aboveground biomass, which accounts for approximately 60 to 90% of C sequestration in tropical regrowth forests, other C stocks also increase as forests age (Figs. 3-5, Tables S2-S4). Carbon accumulation in leaves, roots, and dead wood follows different temporal patterns than aboveground biomass (see also Martin et al., 2013). Foliage biomass accumulates rapidly in young forests, and does not increase markedly thereafter—a pattern that is common in forests globally (Anderson-Teixeira et al., 2013). Total root biomass effectively tracks aboveground biomass (Tables S2-S4), whereas we detected no age-related trends in fine root biomass. Dead wood accumulation lags behind live biomass accumulation, as has been observed in temperate forest stand development (Harmon, 2009). Following decomposition of legacy dead wood in young stands, dead wood accumulation also contributes to the C sink of regrowth forests, sequestering on average 0.1 Mg C ha-1 yr-1 in intermediate-aged forests, and likely continuing to accumulate beyond stand age 100 (Fig. 6b), as has been observed in temperate forests (Janisch & Harmon, 2002; McGarvey et al., 2014). These findings—combined with eddy covariance measurements indicating that mature unmanaged broadleaf evergreen forests are C sinks (Fig. 3), indicate that tropical forests continue to accumulate C beyond 100 years of age. Carbon accumulation in regrowth forests is fueled by GPP and its allocation among various C flux terms. However, scant data on C fluxes in tropical regrowth forests (Figs. 4-5; Table S1) and the fact that comparisons are being made across forests that vary in many factors other than age limit generalizations about age trends in C fluxes. Our analysis detected no differences across young, intermediate-aged, and mature forests in NEE, respiration (Reco, Rh, Rsoil, etc.), GPP, NPP, BNPP, or most major components of NPP (ANPP, ANPPfoliage, ANPPlitterfall, BNPP, BNPPfine). For many of these variables, existing age trends would not be detectable because of data limitations. However, the lack of detectable age trends in some of these variables is also broadly consistent with observations from higher-latitude forests, where GPP and NPP quickly plateau as forests age, sometimes followed by a modest decline, and where heterotrophic respiration, Rh, appears to be roughly invariant with stand age (Anderson-Teixeira et al., 2013 and refs therein). Contrasting with the lack of an age trend in ANPP, the decline in ANPPstem from young to mature stands indicates decreasing C allocation to stem growth as stands age, eventually declining to an average of only ~6% of GPP in mature/ intact unmanaged broadleaf evergreen forests (Figs. 3-5). Additional measurements of C flux in regrowth tropical forests will be required to clarify how patterns of C flux and allocation change as stands age. Implications & future directions TropForC-db is the largest database on ground-based measurement of ecosystem level C stocks and fluxes in tropical forests, and as such allows us to provide the most comprehensive synthesis to date of C cycling in tropical regrowth and mature forests (Figs 3-6). Our findings refine estimates of average C stocks and fluxes using a more comprehensive pantropical data set and complement findings of previous work. A cc ep te d A rt ic le This article is protected by copyright. All rights reserved. Moving forward, we anticipate that TropForC-db will be of value for various initiatives seeking to understand and manage the role of tropical forests in the global C cycle. Specifically, the data will be useful for synthetic analyses seeking to better understand tropical forest C stocks and flows, how these are shaped by past disturbance, and how they are influenced by environmental factors. It will also be of value for model calibration, benchmarking, and improvement, and will be integrated into the BETYdb/ PEcAn ecosystem modeling system (LeBauer et al., 2010, 2012). Finally, TropForC-db can be used to provide more accurate estimates of C stocks and fluxes for regional to global scale tropical forest C inventories (e.g., (IPCC, 2006; Pan et al., 2011). For instance, estimates of biomass accumulation rates in regrowth forests used by previous studies estimating regional to global scale tropical forest C balances (e.g., Houghton et al., 2000; Achard et al., 2004; Pan et al., 2011) and current IPCC greenhouse gas inventory guidelines (IPCC, 2006) are underlain by only a subset of the data in TropForC-db. We anticipate ongoing development of the database. In addition to archiving of the data associated with this publication (Anderson-Teixeira et al., 2016), we plan to maintain a dynamic instance of the database including forests globally (ForC-db), which can be accessed https://github.com/forc-db. Acknowledgements We thank all authors of the original studies and data compilations included in this database, their funding agencies, and the various networks that support ground-based measurements of C stocks and fluxes in the tropics, without which this database or the type of integrated analyses conducted here would not be possible. We also acknowledge all authors of previous data compilations, particularly Sebastian Luyssaert, for their key role in synthesizing data upon which this database draws. Thanks to Moein Azimi, Amy C. Bennett, Syed Ashraful Alam, and Markku Larjavaara for help assembling data, to Yadvinder Malhi communication regarding data, to Jens Kattge for helpful discussion, and to Erika Gonzalez-Akre, Adam Miller, and two anonymous reviewers for helpful comments on the manuscript. This research was funded by DOE grants DE-SC0008085 and DE- SC0010039 to KAT, Evan DeLucia, and Benjamin Duval and by a Smithsonian Women’s Committee grant to KAT. A cc ep te d A rt ic le This article is protected by copyright. All rights reserved. References Achard F, Eva HD, Mayaux P, Stibig H-J, Belward A (2004) Improved estimates of net carbon emissions from land cover change in the tropics for the 1990s. Global Biogeochemical Cycles, 18, GB2008. Adachi M, Bekku YS, Rashidah W, Okuda T, Koizumi H (2006) Differences in soil respiration between different tropical ecosystems. Applied Soil Ecology, 34, 258–265. del Aguila-Pasquel J, Doughty CE, Metcalfe DB et al. (2014) The seasonal cycle of productivity, metabolism and carbon dynamics in a wet aseasonal forest in north-west Amazonia (Iquitos, Peru). Plant Ecology & Diversity, 7, 71–83. Aide TM, Zimmerman JK, Herrera L, Rosario M, Serrano M (1995) Forest recovery in abandoned tropical pastures in Puerto Rico. Forest Ecology and Management, 77, 77–86. Alves D, Soares JV, Amaral S, Mello E, Almeida S, Da Silva OF, Silveira A (1997) Biomass of primary and secondary vegetation in Rondonia, Western Brazilian Amazon. Global Change Biology, 3, 451–461. Anderson KJ, Allen AP, Gillooly JF, Brown JH (2006) Temperature-dependence of biomass accumulation rates during secondary succession. Ecology Letters, 9, 673–682. Anderson-Teixeira KJ, Miller AD, Mohan JE, Hudiburg TW, Duval BD, DeLucia EH (2013) Altered dynamics of forest recovery under a changing climate. Global Change Biology, 19, 2001– 2021. Anderson-Teixeira KJ, Davies SJ, Bennett AC et al. (2015) CTFS-ForestGEO: a worldwide network monitoring forests in an era of global change. Global Change Biology, 21, 528–549. Anderson-Teixeira K, Wang MMH, McGarvey JC, LeBauer DS (2016) Data from: Carbon dynamics of mature and regrowth tropical forests derived from a pantropical database (TropForC-db). Dryad Digital Repository. http://dx.doi.org/10.5061/dryad.t516f. Aragão LEOC, Malhi Y, Metcalfe DB et al. (2009) Above- and below-ground net primary productivity across ten Amazonian forests on contrasting soils. Biogeosciences Discussions, 6, 2441–2488. Araújo AC (2002) Comparative measurements of carbon dioxide fluxes from two nearby towers in a central Amazonian rainforest: The Manaus LBA site. Journal of Geophysical Research, 107. Araujo-Murakami A, Doughty CE, Metcalfe DB et al. (2014) The productivity, allocation and cycling of carbon in forests at the dry margin of the Amazon forest in Bolivia. Plant Ecology & Diversity, 7, 55–69. Arora VK, Boer GJ (2010) Uncertainties in the 20th century carbon budget associated with land use change. Global Change Biology, 16, 3327–3348. A cc ep te d A rt ic le This article is protected by copyright. All rights reserved. Aththorick TA, Setiadi D, Purwanto Y, Guhardja EDI (2012) Vegetation stands structure and aboveground biomass after the shifting cultivation practices of Karo People in Leuser Ecosystem, North Sumatra. Biodiversitas, 13, 92–97. Baker TR, Phillips OL, Malhi Y et al. (2004) Increasing biomass in Amazonian forest plots. Philosophical Transactions of the Royal Society B: Biological Sciences, 359, 353–365. Baldocchi D (2008) TURNER REVIEW No. 15.“Breathing” of the terrestrial biosphere: lessons learned from a global network of carbon dioxide flux measurement systems. Australian Journal of Botany, 56, 1–26. Bartholomew W V, Meyer J, Laudelout H (1953) Mineral nutrient immobilization under forest and grass fallow in the Yangambi (Belgian Congo) region. Publications de L’Institut National Pour L’Etude Agronomique du Conge Belge Serie Scientifique, 57, 1–27. Becknell JM, Powers JS (2014) Stand age and soils as drivers of plant functional traits and aboveground biomass in secondary tropical dry forest. Canadian Journal of Forest Research, 44, 604–613. Beer C, Reichstein M, Tomelleri E et al. (2010) Terrestrial Gross Carbon Dioxide Uptake: Global Distribution and Covariation with Climate. Science, 329, 834–838. Behera N, Joshi SK, Pati DP (1990) Root contribution to total soil metabolism in a tropical forest soil from Orissa, India. Forest Ecology and Management, 36, 125–134. Beringer J, Hutley LB, Tapper NJ, Cernusak LA (2007) Savanna fires and their impact on net ecosystem productivity in North Australia. Global Change Biology, 13, 990–1004. Bonan GB (2008) Forests and Climate Change: Forcings, Feedbacks, and the Climate Benefits of Forests. Science, 320, 1444–1449. Bongers F, Engelen D, Klinge H (1985) Phytomass structure of natural plant communities on spodosols in southern Venezuela: the Bana woodland. Vegetatio, 63, 13–34. Bonner MTL, Schmidt S, Shoo LP (2013) A meta-analytical global comparison of aboveground biomass accumulation between tropical secondary forests and monoculture plantations. Forest Ecology and Management, 291, 73–86. Brienen RJW, Phillips OL, Feldpausch TR et al. (2015) Long-term decline of the Amazon carbon sink. Nature, 519, 344–348. Broadbent EN, Almeyda Zambrano AM, Asner GP et al. (2014) Integrating Stand and Soil Properties to Understand Foliar Nutrient Dynamics during Forest Succession Following Slash-and- Burn Agriculture in the Bolivian Amazon (ed Bernacchi CJ). PLoS ONE, 9, e86042. Brown S, Lugo A (1982) The storage and production of organic matter in tropical forests and their role in the global carbon cycle. Biotropica, 14, 161–187. A cc ep te d A rt ic le This article is protected by copyright. All rights reserved. Brown IF, Martinelli LA, Thomas WW, Moreira MZ, Cid Ferreira CA, Victoria RA (1995) Uncertainty in the biomass of Amazonian forests: An example from Rondônia, Brazil. Forest Ecology and Management, 75, 175–189. Bustamante MMC, Roitman I, Aide TM et al. (2015) Towards an integrated monitoring framework to assess the effects of tropical forest degradation and recovery on carbon stocks and biodiversity. Global Change Biology, n/a–n/a. Campioli M, Vicca S, Luyssaert S et al. (2015) Biomass production efficiency controlled by management in temperate and boreal ecosystems. Nature Geoscience, 8, 843–846. Carswell FE, Costa AL, Palheta M et al. (2002) Seasonality in CO2 and H2O flux at an eastern Amazonian rain forest. Journal of Geophysical Research D: Atmospheres, 107. Chambers JQ, Santos J dos, Ribeiro RJ, Higuchi N (2001) Tree damage, allometric relationships, and above-ground net primary production in central Amazon forest. Forest Ecology and Management, 152, 73–84. Chambers JQ, Tribuzy ES, Toledo LC et al. (2004) Respiration from a tropical forest ecosystem: Partitioning of sources and low carbon use efficiency. Ecological Applications, 14. Chan N, Takeda S, Suzuki R, Yamamoto S (2013) Establishment of allometric models and estimation of biomass recovery of swidden cultivation fallows in mixed deciduous forests of the Bago Mountains, Myanmar. Forest Ecology and Management, 304, 427–436. Chapin F, Woodwell G, Randerson J et al. (2006) Reconciling Carbon-cycle Concepts, Terminology, and Methods. Ecosystems, 9, 1041–1050. Chave J, Riera B, Dubois M-A (2001) Estimation of biomass in a neotropical forest of French Guiana: spatial and temporal variability. Journal of Tropical Ecology, 17, S0266467401001055. Chave J, Condit R, Muller-Landau HC et al. (2008) Assessing Evidence for a Pervasive Alteration in Tropical Tree Communities. PLoS Biol, 6, e45. Chave J, Réjou-Méchain M, Búrquez A et al. (2014) Improved allometric models to estimate the aboveground biomass of tropical trees. Global Change Biology, n/a–n/a. Chen D (2010) Dynamics and controls of carbon exchange of a tropical montane forest at Jianfengling, China. Chinese Academy of Forestry. Chen Z, Zhang H, Wang B, Zhang Z (1993) Studies on biomass and its allocation of the evergreen broad-leaved forest in Heishiding, Guangdong. Acta Phytoecologica et Geobotanica Sinica, 17, 289–298. [in Chinese]. Chen X, Hutley LB, Eamus D (2003) Carbon balance of a tropical savanna of northern Australia. Oecologia, 137, 405–416. A cc ep te d A rt ic le This article is protected by copyright. All rights reserved. Chen D, Li Y, Liu H et al. (2010) Biomass and carbon dynamics of a tropical mountain rain forest in China. Science China Life Sciences, 53, 798–810. Clark DA (2013) NPP Tropical Forest: La Selva, Costa Rica, 1975-1994, R1. Data set. Available on- line [http://daac.ornl.gov] from Oak Ridge National Laboratory Distributed Active Archive Center, Oak Ridge, Tennessee, USA. doi:10.3334/ORNLDAAC/218. Clark D., Clark D. (2000) Landscape-scale variation in forest structure and biomass in a tropical rain forest. Forest Ecology and Management, 137, 185–198. Clark DA, Brown S, Kicklighter DW, Chambers J, Thomlinson JR, Ni J (2001a) Measuring net primary production in forests: concepts and field methods. Ecological Applications, 11, 356–370. Clark DA, Brown S, Kicklighter DW, Chambers JQ, Thomlinson JR, Ni J, Holland EA (2001b) Net primary production in tropical forests: An evaluation and synthesis of existing field data. Ecological Applications, 11, 371–384. Clark DA, Brown S, Kicklighter DW, Chambers JQ, Thomlinson JR, Ni J, Holland EA (2013) NPP Tropical Forest: Consistent Worldwide Site Estimates, 1967-1999, R1. Data set. Available on-line [http://daac.ornl.gov] from the Oak Ridge National Laboratory Distributed Active Archive Center, Oak Ridge, Tennessee, USA. doi:10.3334/ORNLDAAC/616. Cleveland CC, Townsend AR (2006) Nutrient additions to a tropical rain forest drive substantial soil carbon dioxide losses to the atmosphere. Proceedings of the National Academy of Sciences of the United States of America, 103, 10316–10321. da Costa ACL, Metcalfe DB, Doughty CE et al. (2014) Ecosystem respiration and net primary productivity after 8–10 years of experimental through-fall reduction in an eastern Amazon forest. Plant Ecology & Diversity, 7, 7–24. Crow TR (1980) A Rainforest Chronicle: A 30-Year Record of Change in Structure and Composition at El Verde, Puerto Rico. Biotropica, 12, 42–55. Dang C, Wu Z (1992) Studies on the biomass for Castanopsis Echdnocarpa community of monsoon evergreen broad-leaved forest. Journal of Yunnan University (Natural Sciences), 14, 95–107. [in Chinese]. Davidson EA, Savage K, Bolstad P et al. (2002) Belowground carbon allocation in forests estimated from litterfall and IRGA-based soil respiration measurements. Agricultural and Forest Meteorology, 113, 39–51. Davidson EA, Reis de Carvalho CJ, Vieira ICG et al. (2004) Nitrogen and phosphorus limitation of biomass growth in a tropical secondary forest. Ecological Applications, 14, 150–163. DeAngelis DL, Gardner RH, Shugart HH (1981) Productivity of forest ecosystems studied during the IBP: the woodlands data set. In: Dynamic Properties of Forest Ecosystems (ed Reichle DE), pp. 567–672. Cambridge University Press, Cambridge. A cc ep te d A rt ic le This article is protected by copyright. All rights reserved. Delaney M, Brown S, Lugo AE, Torres-Lezama A, Quintero NB (1997) The distribution of organic carbon in major components of forests located in five life zones of Venezuela. Journal of Tropical Ecology, 13, 697. DeLucia EH, Drake J, Thomas RB, Gonzalez-Meler MA (2007) Forest carbon use efficiency: is respiration a constant fraction of gross primary production? Global Change Biology, 13, 1157–1167. Doughty CE, Metcalfe DB, da Costa MC et al. (2014) The production, allocation and cycling of carbon in a forest on fertile terra preta soil in eastern Amazonia compared with a forest on adjacent infertile soil. Plant Ecology & Diversity, 7, 41–53. Doughty CE, Metcalfe DB, Girardin CAJ et al. (2015) Drought impact on forest carbon dynamics and fluxes in Amazonia. Nature, 519, 78–82. Van Do T, Osawa A, Thang NT (2010) Recovery process of a mountain forest after shifting cultivation in Northwestern Vietnam. Forest Ecology and Management, 259, 1650–1659. Edwards PJ, Grubb PJ (1977) Studies of Mineral Cycling in a Montane Rain Forest in New Guinea: I. The Distribution of Organic Matter in the Vegetation and Soil. The Journal of Ecology, 65, 943. Egunjobi JK, Bada SO (1979) Biomass and Nutrient Distribution in Stands of Pinus caribea L. in the Dry Forest Zone of Nigeria. Biotropica, 11, 130. Epron D, Nouvellon Y, Deleporte P et al. (2006) Soil carbon balance in a clonal Eucalyptus plantation in Congo: Effects of logging on carbon inputs and soil CO2 efflux. Global Change Biology, 12, 1021–1031. Espírito-Santo FDB, Gloor M, Keller M et al. (2014) Size and frequency of natural forest disturbances and the Amazon forest carbon balance. Nature Communications, 5. Esser G, Lieth HFH, Scurlock JMO, Olson RJ (1997) Worldwide Estimates and Bibliography of Net Primary Productivity derived from Pre-1982. 132 pp. Ewel JJ (1971) Biomass changes in early tropical forest succession. Turrialba, 21, 110–112. Falge E, Baldocchi D, Tenhunen J et al. (2002) Seasonality of ecosystem respiration and gross primary production as derived from FLUXNET measurements. Agricultural and Forest Meteorology, 113, 53–74. Fang Y, Mo J, Peng S, Li D (2003) Role of forest succession on carbon sequestration of forest ecosystems in lower subtropical China. Acta Ecologica Sinica, 23, 1685–1694. [in Chinese]. Fearnside P, Guimaraes W (1996) Carbon uptake by secondary forests in Brazilian Amazonia. Forest Ecology and Mangement, 80, 35–46. A cc ep te d A rt ic le This article is protected by copyright. All rights reserved. Fehse J, Hofstede R, Aguirre N, Paladines C, Kooijman A, Sevink J (2002) High altitude tropical secondary forests: a competitive carbon sink? Forest Ecology and Management, 163, 9–25. Feldpausch TR, Rondon MA, Fernandes ECM, Riha SJ, Wandelli E (2004) Carbon and nutrient accumulation in secondary forests regenerating on pastures in central Amazonia. Ecological Applications, 14, 164–176. Folster H, de las Salas G, Khanna P (1976) A tropical evergreen forest site with perched water table, Magdalena valley, Columbia Biomass and bioelement inventory of primary and secondary vegetation. Oecologia Plantarum, 11, 297–320. Fonseca W, Rey Benayas JM, Alice FE (2011) Carbon accumulation in the biomass and soil of different aged secondary forests in the humid tropics of Costa Rica. Forest Ecology and Management, 262, 1400–1408. Frangi JL, Lugo AE (1985) Ecosystem Dynamics of a Subtropical Floodplain Forest. Ecological Monographs, 55, 351. Gehring C, Denich M, Kanashiro M, Vlek PLG (1999) Response of secondary vegetation in Eastern Amazonia to relaxed nutrient availability constraints. Biogeochemistry, 45, 223–241. Gehring C, Denich M, Vlek PLG (2005) Resilience of secondary forest regrowth after slash-and-burn agriculture in central Amazonia. Journal of Tropical Ecology, 21, 519–527. Giardina CP, Ryan MG, Binkley D, Fownes JH (2003) Primary production and carbon allocation in relation to nutrient supply in a tropical experimental forest. Global Change Biology, 9, 1438– 1450. Giardina CP, Binkley D, Ryan MG, Fownes JH, Senock RS (2004) Belowground carbon cycling in a humid tropical forest decreases with fertilization. Oecologia, 139, 545–550. Girardin CAJ, Malhi Y, Aragão LEOC et al. (2010) Net primary productivity allocation and cycling of carbon along a tropical forest elevational transect in the Peruvian Andes. Global Change Biology, 16, 3176–3192. Golley FB (1975) Mineral Cycling in a Tropical Moist Forest Ecosystem. University of Georgia Press, 248 pp. Grace J, Mitchard E, Gloor E (2014) Perturbations in the carbon budget of the tropics. Global Change Biology, 20, 3238–3255. Grimm U, Fassbender HW (1999) NPP Tropical Forest: San Eusebio, Venezuela, 1970-1971. Oak Ridge National Laboratory Distributed Active Archive Center, Oak Ridge, Tennessee, U.S.A. Guimaraes W (1993) Liberacao de carbono e mudancas nos estoques dos nutrientes contidos na biomassa aerea e no solo resultante de queimadas de florestas secundarias em areas de pastagens abandonadas, em Altamira, Para. Instituto Nacional de Pesquisas da Amazonia/ Fundacao Universidade do Amazonas, Manaus, 69 pp. A cc ep te d A rt ic le This article is protected by copyright. All rights reserved. Harmand J-M, Njiti CF, Bernhard-Reversat F, Puig H (2004) Aboveground and belowground biomass, productivity and nutrient accumulation in tree improved fallows in the dry tropics of Cameroon. Forest Ecology and Management, 188, 249–265. Harmon ME (2009) Woody Detritus its Contribution to Carbon Dynamics of Old-Growth Forests: the Temporal Context. In: Old-Growth Forests (eds Wirth C, Gleixner G, Heimann M), pp. 159–190. Springer Berlin Heidelberg. Hirano T, Segah H, Harada T, Limin S, JUNE T, HIRATA R, OSAKI M (2007) Carbon dioxide balance of a tropical peat swamp forest in Kalimantan, Indonesia. Global Change Biology, 13, 412–425. Hirano T, Jauhiainen J, Inoue T, Takahashi H (2009) Controls on the Carbon Balance of Tropical Peatlands. Ecosystems, 12, 873–887. Hirata R, Saigusa N, Yamamoto S et al. (2008) Spatial distribution of carbon balance in forest ecosystems across East Asia. Agricultural and Forest Meteorology, 148, 761–775. Hoshizaki K, Niiyama K, Kimura K et al. (2004) Temporal and spatial variation of forest biomass in relation to stand dynamics in a mature, lowland tropical rainforest, Malaysia. Ecological Research, 19, 357–363. Houghton RA, Skole DL, Nobre CA, Hackler JL, Lawrence KT, Chomentowski WH (2000) Annual fluxes or carbon from deforestation and regrowth in the Brazilian Amazon. Nature, 403, 301–304. Houghton RA, Byers B, Nassikas AA (2015) A role for tropical forests in stabilizing atmospheric CO2. Nature Climate Change, 5, 1022–1023. Hozumi K, Yoda K, Kokawa S, Kira T Production ecology of tropical rain forests in southwestern Cambodia : I. Plant biomass. Hughes RF, Kauffman JB, Jaramillo VJ (1999a) Biomass, carbon, and nutrient dynamics of secondary forests in a humid tropical region of Mexico. Ecology, 80, 1892–1907. Hughes RF, Kauffman JB, Jaramillo VJ (1999b) Biomass, carbon, and nutrient dynamics of secondary forests in a humid tropical region of Mexico. Ecology, 80, 1892–1907. Hughes RF, Kauffman JB, Jaramillo VJ (2000) Ecosystem-scale impacts of deforestation and land use in a humid tropical region of Mexico. Ecological Applications, 10, 515–527. Hughes RF, Kauffman JB, Cummings DL (2002) Dynamics of Aboveground and Soil Carbon and Nitrogen Stocks and Cycling of Available Nitrogen along a Land-use Gradient in Rondônia, Brazil. Ecosystems, 5, 244–259. Huston MA, Wolverton S (2009) The global distribution of net primary production: resolving the paradox. Ecological monographs, 79, 343–377. A cc ep te d A rt ic le This article is protected by copyright. All rights reserved. Hutley LB, Leuning R, Beringer J, Cleugh HA (2005) The utility of the eddy covariance techniques as a tool in carbon accounting: tropical savanna as a case study. Australian Journal of Botany, 53, 663. Hutyra LR, Munger JW, Saleska SR et al. (2007) Seasonal controls on the exchange of carbon and water in an Amazonian rain forest. Journal of Geophysical Research, 112, G03008. IPCC (2003) LUCF Sector Good Practice Guidance. In: IPCC Good Practice Guidance for Land Use, Land-Use Change and Forestry (eds Jim Penman, Michael Gytarsky, Taka Hiraishi, Thelma Krug DK, Riitta Pipatti, Leandro Buendia, Kyoko Miwa TN, Wagner KT and F). Institute for Global Environmental Strategies (IGES), Hayama, Japan. IPCC (2006) Agriculture, Forestry, and Other Land Use. In: 2006 IPCC Guidelines for National Greenhouse Gas Inventories (eds Eggleston S, Buendia L, Miwa K, Ngara T, Tanabe K). Institute for Global Environmental Strategies, Hayama, Japan. IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change (eds Stocker TF, Quin D, Plattner G-K, Tignor M, Allen SK, Boschung J, Nauels A, Xia Y, Bex V, Midgley PM). Cambridge University Press, Cambridge, UK and New York, NY, USA, 1535 pp. Ito A, Oikawa T (2002) A simulation model of the carbon cycle in land ecosystems (Sim-CYCLE): a description based on dry-matter production theory and plot-scale validation. Ecological Modelling, 151, 143–176. Janisch JE, Harmon ME (2002) Successional changes in live and dead wood carbon stores: implications for net ecosystem productivity. Tree Physiology, 22, 77–89. Jepsen MR (2006) Above-ground carbon stocks in tropical fallows, Sarawak, Malaysia. Forest Ecology and Management, 225, 287–295. Jordan CF, Cuevas E, Medina E (1999) NPP Tropical Forest: San Carlos de Rio Negro, Venezuela, 1975-1984. Oak Ridge National Laboratory Distributed Active Archive Center, Oak Ridge, Tennessee, U.S.A. Jung M, Henkel K, Herold M, Churkina G (2006) Exploiting synergies of global land cover products for carbon cycle modeling. Remote Sensing of Environment, 101, 534–553. Kadeba O (1991) Above-ground biomass production and nutrient accumulation in an age sequence of Pinus caribaea stands. Forest Ecology and Management, 41, 237–248. Kato T, Tang Y (2008) Spatial variability and major controlling factors of CO 2 sink strength in Asian terrestrial ecosystems: evidence from eddy covariance data. Global Change Biology, 14, 2333–2348. A cc ep te d A rt ic le This article is protected by copyright. All rights reserved. Keller M, Palace M, Hurtt G (2001) Biomass estimation in the Tapajos National Forest, Brazil examination of sampling and allometric uncertainties. Forest Ecology and Management, 154, 371–382. Keller M, Palace M, Asner GP, Pereira R, Silva JNM (2004) Coarse woody debris in undisturbed and logged forests in the eastern Brazilian Amazon. Global Change Biology, 10, 784–795. Kenzo T, Ichie T, Hattori D, Kendawang JJ, Sakurai K, Ninomiya I (2010) Changes in above- and belowground biomass in early successional tropical secondary forests after shifting cultivation in Sarawak, Malaysia. Forest Ecology and Management, 260, 875–882. Kira T (1998) NPP Tropical Forest: Khao Chong, Thailand, 1962-1965. Oak Ridge National Laboratory Distributed Active Archive Center, Oak Ridge, Tennessee, U.S.A. Klinge H, Herrera R (1983) Phytomass structure of natural plant communities on spodosols in southern Venezuela: The tall Amazon Caatinga forest. Vegetatio, 53, 65–84. Klinge H, Rodrigues WA, Brunig E, Fittkau EJ (1975) Biomass and Structure in a Central Amazonian Rain Forest. In: Tropical Ecological Systems: Trends in Terrestrial and Aquatic Research (eds Golley FB, Medina E), pp. 115–122. Springer-Verlag, Berlin, Heidelberg, New York. Kosugi Y, Takanashi S, Ohkubo S et al. (2008) CO2 exchange of a tropical rainforest at Pasoh in Peninsular Malaysia. Agricultural and Forest Meteorology, 148, 439–452. Kraenzel M, Castillo A, Moore T, Potvin C (2003) Carbon storage of harvest-age teak (Tectona grandis) plantations, Panama. Forest Ecology and Management, 173, 213–225. Kruijt B, Elbers JA, von Randow C et al. (2004) The robustness of eddy correlation fluxes for amazon rain forest conditions. Ecological Applications, 14, 101–113. Kunstadter P, Chapman EC, Sabhasri S (1978) Farmers in the forest: economic development and marginal agriculture in Northern Thailand. Farmers in the forest: economic development and marginal agriculture in Northern Thailand. Kurz WA, Stinson G, Rampley GJ, Dymond CC, Neilson ET (2008) Risk of natural disturbances makes future contribution of Canada’s forests to the global carbon cycle highly uncertain. Proceedings of the National Academy of Sciences, 105, 1551–1555. Laclau J-P, Bouillet J-P, Ranger J (2000) Dynamics of biomass and nutrient accumulation in a clonal plantation of Eucalyptus in Congo. Forest Ecology and Management, 128, 181–196. Lawrence D (2005) Biomass accumulation after 10–200 years of shifting cultivation in Bornean rain forest. Ecology, 86, 26–33. Lawrence D, Radel C, Tully K, Schmook B, Schneider L (2010) Untangling a Decline in Tropical Forest Resilience: Constraints on the Sustainability of Shifting Cultivation Across the Globe: Decline in Tropical Forest Biomass Accumulation. Biotropica, 42, 21–30. A cc ep te d A rt ic le This article is protected by copyright. All rights reserved. LeBauer D, Dietze M, Kooper R, Long SP, Mulrooney P, Rhode GS, Wang D (2010) Biofuel Ecophysiological Traits and Yields Database (BETYdb). Energy Biosciences Institute, University of Illinois at Urbana-Champaign. LeBauer DS, Wang D, Richter KT, Davidson CC, Dietze MC (2012) Facilitating feedbacks between field measurements and ecosystem models. Ecological Monographs, 83, 133–154. Letcher SG, Chazdon RL (2009) Rapid Recovery of Biomass, Species Richness, and Species Composition in a Forest Chronosequence in Northeastern Costa Rica. Biotropica, 41, 608– 617. Leuning R, Cleugh HA, Zegelin SJ, Hughes D (2005) Carbon and water fluxes over a temperate Eucalyptus forest and a tropical wet/dry savanna in Australia: measurements and comparison with MODIS remote sensing estimates. Agricultural and Forest Meteorology, 129, 151–173. Lewis SL, Lopez-Gonzalez G, Sonké B et al. (2009a) Increasing carbon storage in intact African tropical forests. Nature, 457, 1003–1006. Lewis SL, Lopez-Gonzalez G, Sonké B et al. (2009b) Increasing carbon storage in intact African tropical forests. Nature, 457, 1003–1006. Lewis SL, Brando PM, Phillips OL, Heijden GMF van der, Nepstad D (2011) The 2010 Amazon Drought. Science, 331, 554–554. Lewis SL, Edwards DP, Galbraith D (2015) Increasing human dominance of tropical forests. Science, 349, 827–832. Li Y, Xu M, Sun OJ, Cui W (2004) Effects of root and litter exclusion on soil CO2 efflux and microbial biomass in wet tropical forests. Soil Biology and Biochemistry, 36, 2111–2114. Litton CM, Raich JW, Ryan MG (2007) Carbon allocation in forest ecosystems. Global Change Biology, 13, 2089–2109. Liu Y, Yu G, Wang Q, Zhang Y (2014) How temperature, precipitation and stand age control the biomass carbon density of global mature forests: Controls on biomass carbon density of global mature forests. Global Ecology and Biogeography, 23, 323–333. Loescher HW, Oberbauer SF, Gholz HL, Clark DB (2003) Environmental controls on net ecosystem- level carbon exchange and productivity in a Central American tropical wet forest. Global Change Biology, 9, 396–412. Lucas RM, Honzák M, Amaral I Do, Curran PJ, Foody GM (1998) Avaliação da composição florística, biomassa e estrutura de floresta tropical em regeneração: a contribuição do sensoriamento remoto. In: Floresta Amazônica: dinâmica, regeneração e manejo (eds Gascon C, Moutinho P), pp. 61–82. Ministério da Ciência e Tecnologia, Instituto Nacional de Pesquisa da Amazônia. A cc ep te d A rt ic le This article is protected by copyright. All rights reserved. Lucas RM, Honzák M, Amaral I Do, Curran PJ, Foody GM (2002) Forest regeneration on abandoned clearances in central Amazonia. International Journal of Remote Sensing, 23, 965–988. Lugo AE (1992) Comparison of Tropical Tree Plantations with Secondary Forests of Similar Age. Ecological Monographs, 62, 1. Lugo AE, Wang D, Herbert Bormann F (1990) A comparative analysis of biomass production in five tropical tree species. Forest Ecology and Management, 31, 153–166. Luo T-X (1996) Patterns of net primary productivity for Chinese major forest types and their mathematical models. Chinese Academy of Sciences, Beijing, China. Luyssaert S, Inglima I, Jung M et al. (2007a) CO2 balance of boreal, temperate, and tropical forests derived from a global database. Global Change Biology, 13, 2509–2537. Luyssaert S, Inglima I, Jung M et al. (2007b) CO 2 balance of boreal, temperate, and tropical forests derived from a global database. Global Change Biology, 13, 2509–2537. Maass M, Martinez-Yrizar A (2001) NPP Tropical Forest: Chamela, Mexico, 1982-1995. Oak Ridge National Laboratory Distributed Active Archive Center, Oak Ridge, Tennessee, U.S.A. Malhi Y (2012) The productivity, metabolism and carbon cycle of tropical forest vegetation. Journal of Ecology, 100, 65–75. Malhi Y, Nobre AD, Grace J, Kruijt B, Pereira MGP, Culf A, Scott S (1998) Carbon dioxide transfer over a Central Amazonian rain forest. Journal of Geophysical Research, 103, 31593. Malhi Y, Baldocchi D, Jarvis PG (1999) The carbon balance of tropical, temperate and boreal forests. Plant, Cell and Environment, 22, 715–740. Malhi Y, Baker TR, Phillips OL et al. (2004) The above-ground coarse wood productivity of 104 Neotropical forest plots. Global Change Biology, 10, 563–591. Malhi Y, Doughty C, Galbraith D (2011) The allocation of ecosystem net primary productivity in tropical forests. Philosophical Transactions of the Royal Society B: Biological Sciences, 366, 3225–3245. Malhi Y, Gardner TA, Goldsmith GR, Silman MR, Zelazowski P (2014) Tropical Forests in the Anthropocene. Annual Review of Environment and Resources, 39, 140906185140006. Malhi Y, Doughty CE, Goldsmith GR et al. (2015) The linkages between photosynthesis, productivity, growth and biomass in lowland Amazonian forests. Global Change Biology, 21, 2283–2295. Marín-Spiotta E, Sharma S (2013) Carbon storage in successional and plantation forest soils: a tropical analysis: Carbon in reforested and plantation soils (ed Ramankutty N). Global Ecology and Biogeography, 22, 105–117. A cc ep te d A rt ic le This article is protected by copyright. All rights reserved. Marín-Spiotta E, Silver WL, Ostertag R (2007) Long-term patterns in tropical reforestation: plant community composition and aboveground biomass accumulation. Ecological Applications, 17, 828–839. Marín-Spiotta E, Cusack D, Ostertag R, Silver W (2008) Trends in above and belowground carbon with forest regrowth after agricultural abandonment in the neotropics. In: Post-Agricultural Succession in the Neotropics (ed Myster RW), pp. 22–72. Springer, New York, NY. Martin PA, Newton AC, Bullock JM (2013) Carbon pools recover more quickly than plant biodiversity in tropical secondary forests. Proceedings of the Royal Society B: Biological Sciences, 280, 20132236–20132236. Mascaro J, Asner GP, Dent DH, DeWalt SJ, Denslow JS (2012) Scale-dependence of aboveground carbon accumulation in secondary forests of Panama: A test of the intermediate peak hypothesis. Forest Ecology and Management, 276, 62–70. McGarvey JC, Thompson JR, Epstein HE, Shugart HH (2014) Carbon storage in old-growth forests of the Mid-Atlantic: toward better understanding the eastern forest carbon sink. Ecology, 96, 311–317. Mesquita R de CG, Massoca PE dos S, Jakovac CC, Bentos TV, Williamson GB (2015) Amazon Rain Forest Succession: Stochasticity or Land-Use Legacy? BioScience, 65, 849–861. Mialet-Serra I, Clement A, Sonderegger N, Roupsard O, Jourdan C, LabouisseE J-P, Dingkuhn M (2005) Assimilate storage in vegetative organs of coconut (Cocos nucifera). Experimental Agriculture, 41, 161–174. Michaletz ST, Cheng D, Kerkhoff AJ, Enquist BJ (2014) Convergence of terrestrial plant production across global climate gradients. Nature. Miller SD, Goulden ML, Menton MC, da Rocha HR, de Freitas HC, Figueira AM e S, Dias de Sousa CA (2004) Biometric and micrometeorological measurements of tropical forest carbon balance. Ecological Applications, 14, 114–126. Montagnini F (2000) Accumulation in above-ground biomass and soil storage of mineral nutrients in pure and mixed plantations in a humid tropical lowland. Forest Ecology and Management, 134, 257–270. Moser G, Leuschner C, Hertel D, Graefe S, Soethe N, Iost S (2011) Elevation effects on the carbon budget of tropical mountain forests (S Ecuador): the role of the belowground compartment. Global Change Biology, 17, 2211–2226. Muller-Landau H, Detto M, Chisholm RA, Hubbell SP, Condit RS (2014) Detecting and projecting changes in forest biomass from plot data. In: Forests and Global Change (eds Coomes DA, Burslem DFRP, Simonson WD), pp. 381–415. Cambridge University Press, Cambridge, United Kingdom. A cc ep te d A rt ic le This article is protected by copyright. All rights reserved. Nakane K (1980) Comparative studies of cycling of soil organic carbon in three primeval moist forests. Japanese Journal of Ecology, 30, 155–172. Navarro MN V, Jourdan C, Sileye T et al. (2008) Fruit development, not GPP, drives seasonal variation in NPP in a tropical palm plantation. Tree physiology, 28, 1661–1674. Nye PH, Greenland DJ (1998) NPP Tropical Forest: Kade, Ghana, 1957-1972. Oak Ridge National Laboratory Distributed Active Archive Center, Oak Ridge, Tennessee, U.S.A. Nygård R, Sawadogo L, Elfving B (2004) Wood-fuel yields in short-rotation coppice growth in the north Sudan savanna in Burkina Faso. Forest Ecology and Management, 189, 77–85. Ogawa H, Yoda K, Ogino K, Kira T (1965) Comparative ecological studies on three main types of forest vegetation in Thailand. II. Plant biomass. In: Nature and Life in Southeast Asia Vol. I (eds Kira T, Iwata K), pp. 49–80. Fauna and Flora Research Society, Kyoto. Olson DM, Dinerstein E, Wikramanayake ED et al. (2001) Terrestrial Ecoregions of the World: A New Map of Life on Earth. BioScience, 51, 933. Onyekwelu JC (2004) Above-ground biomass production and biomass equations for even-aged Gmelina arborea (ROXB) plantations in south-western Nigeria. Biomass and Bioenergy, 26, 39–46. Onyekwelu J (2007) Growth, biomass yield and biomass functions for plantation-grown Nauclea diderrichii (de wild) in the humid tropical rainforest zone of south-western Nigeria. Bioresource Technology, 98, 2679–2687. Orihuela-Belmonte DE, de Jong BHJ, Mendoza-Vega J, Van der Wal J, Paz-Pellat F, Soto-Pinto L, Flamenco-Sandoval A (2013) Carbon stocks and accumulation rates in tropical secondary forests at the scale of community, landscape and forest type. Agriculture, Ecosystems & Environment, 171, 72–84. Overman JPM, Witte HJL, Saldarriaga JG (1994) Evaluation of regression models for above-ground biomass determination in Amazon rainforest. Journal of Tropical Ecology, 10, 207. Palace M, Keller M, Asner GP, Silva JNM, Passos C (2007) Necromass in undisturbed and logged forests in the Brazilian Amazon. Forest Ecology and Management, 238, 309–318. Pan Y, Birdsey RA, Fang J et al. (2011) A Large and Persistent Carbon Sink in the World’s Forests. Science, 333, 988–993. Parrotta JA (1999) Productivity, nutrient cycling, and succession in single- and mixed-species plantations of Casuarina equisetifolia, Eucalyptus robusta, and Leucaena leucocephala in Puerto Rico. Forest Ecology and Management, 124, 45–77. Peng S, Zhang Z (1994) Biomass, productivity and efficiency of radiation utilization of a typical vegetation in Dinghushan. Science In China (Series B) / Scientia Sinica Chimica, 24, 497–502. [in Chinese]. A cc ep te d A rt ic le This article is protected by copyright. All rights reserved. Phillips OL (1998) Changes in the Carbon Balance of Tropical Forests: Evidence from Long-Term Plots. Science, 282, 439–442. Pinard MA, Putz FE (1996) Retaining Forest Biomass by Reducing Logging Damage. Biotropica, 28, 278–295. Poorter L, Bongers F, Aide TM et al. (2016) Biomass resilience of Neotropical secondary forests. Nature, in press. Proctor J (2013) NPP Tropical Forest: Gunung Mulu, Malaysia, 1977-1978, R1. Data set. Available on-line [http://www.daac.ornl.gov] from Oak Ridge National Laboratory Distributed Active Archive Center, Oak Ridge, Tennessee, U.S.A. doi:10.3334/ORNLDAAC/474. Le Quéré C, Andres RJ, Boden T et al. (2013) The global carbon budget 1959–2011. Earth System Science Data, 5, 165–185. Ramachandran A, Byrappa Gowdu Viswanathan M (2009) A new species of Gymnema (Asclepiadaceae) from the Kollihills in Peninsular India. Adansonia, 31, 407–411. R Core Team (2013) R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL http://www.R-project.org/. Read L, Lawrence D (2003) Recovery of biomass following shifting cultivation in dry tropical forests of the Yucatan. Ecological Applications, 13, 85–97. Rice AH, Pyle EH, Saleska SR et al. (2004) Carbon balance and vegetation dynamics in an old-growth Amazonian forest. Ecological Applications, 14, 55–71. Rocha W, Metcalfe DB, Doughty CE et al. (2014) Ecosystem productivity and carbon cycling in intact and annually burnt forest at the dry southern limit of the Amazon rainforest (Mato Grosso, Brazil). Plant Ecology & Diversity, 7, 25–40. Roupsard O, Bonnefond J-M, Irvine M et al. (2006) Partitioning energy and evapo-transpiration above and below a tropical palm canopy. Agricultural and Forest Meteorology, 139, 252– 268. Running SW (2008) Ecosystem Disturbance, Carbon, and Climate. Science, 321, 652–653. Ryan MG, Binkley D, Fownes JH, Giardina CP, Senock RS (2004) An experimental test of the causes of forest growth decline with stand age. Ecological Monographs, 74, 393–414. Saldarriaga JG, West DC, Tharp ML, Uhl C (1988) Long-Term Chronosequence of Forest Succession in the Upper Rio Negro of Colombia and Venezuela. The Journal of Ecology, 76, 938. Saleska SR (2003) Carbon in Amazon Forests: Unexpected Seasonal Fluxes and Disturbance- Induced Losses. Science, 302, 1554–1557. A cc ep te d A rt ic le This article is protected by copyright. All rights reserved. Salomão R de P (1994) Estimativas de biomassa e avaliação do estoque de carbono da vegetação de florestas primárias e secundárias de diversas idades (capoeiras) na Amazônia Oriental, município de Peixe-boi, Pará. Universidade Federal do Pará / Museu Paraense Emilio Goeldi, Belém, 53 pp. Santos AJB, Silva GTDA, Miranda HS, Miranda AC, Lloyd J (2003) Effects of fire on surface carbon, energy and water vapour fluxes over campo sujo savanna in central Brazil. Functional Ecology, 17, 711–719. Schimel D, Pavlick R, Fisher JB et al. (2015) Observing terrestrial ecosystems and the carbon cycle from space. Global Change Biology, 21, 1762–1776. Schöngart J, Wittmann F (2010) Biomass and Net Primary Production of Central Amazonian Floodplain Forests. , pp. 347–388. Schuur EAG (2005) NPP Tropical Forest: Maui, Hawaii, U.S.A., 1996-1997. Oak Ridge National Laboratory Distributed Active Archive Center, Oak Ridge, Tennessee, U.S.A. Schwalm CR, Williams CA, Schaefer K et al. (2010) A model-data intercomparison of CO2 exchange across North America: Results from the North American Carbon Program site synthesis. Journal of Geophysical Research, 115. Scott G (1977) The importance of old-field succession biomass increments to shifting cultivation. lains- Rocky Mountain Geographical Journal, 6, 318–327. Sierra CA, del Valle JI, Orrego SA et al. (2007a) Total carbon stocks in a tropical forest landscape of the Porce region, Colombia. Forest Ecology and Management, 243, 299–309. Sierra CA, Harmon ME, Moreno FH, Orrego SA, del Valle JI. (2007b) Spatial and temporal variability of net ecosystem production in a tropical forest: testing the hypothesis of a significant carbon sink. Global Change Biology, 13, 838–853. Sierra CA, del Valle JI, Restrepo HI (2012) Total carbon accumulation in a tropical forest landscape. Carbon Balance and Management, 7, 12. Silver WL, Kueppers LM, Lugo AE, Ostertag R, Matzek V (2004) Carbon sequestration and plant community dynamics following reforestation of tropical pasture. Ecological Applications, 14, 1115–1127. Snedaker SC (1970) Ecological Studies on Tropical Moist Forest Succession in Eastern Lowland Guatemala. University of Florida, Gainesville, 260 pp. Sorrensen CL (2000) Linking smallholder land use and fire activity: examining biomass burning in the Brazilian Lower Amazon. Forest Ecology and Management, 128, 11–25. Stape J., Binkley D, Ryan MG (2004) Eucalyptus production and the supply, use and efficiency of use of water, light and nitrogen across a geographic gradient in Brazil. Forest Ecology and Management, 193, 17–31. A cc ep te d A rt ic le This article is protected by copyright. All rights reserved. Steininger MK (2000) Secondary forest structure and biomass following short and extended land- use in central and southern Amazonia. Journal of Tropical Ecology, 16, 689–708. Stocker GC (2013) NPP Tropical Forest: Atherton, Australia, 1974-1985, R1. Data set. Available on- line [http://daac.ornl.gov] from Oak Ridge National Laboratory Distributed Active Archive Center, Oak Ridge, Tennessee, USA. doi:10.3334/ORNLDAAC/577. Swamy SL, Kushwaha SK, Puri S (2004) Tree growth, biomass, allometry and nutrient distribution in Gmelina arborea stands grown in red lateritic soils of Central India. Biomass and Bioenergy, 26, 305–317. Szott LT, Palm CA, Davey CB (1994) Biomass and litter accumulation under managed and natural tropical fallows. Forest Ecology and Management, 67, 177–190. Tanner EVJ (1980) Studies on the Biomass and Productivity in a Series of Montane Rain Forests in Jamaica. The Journal of Ecology, 68, 573–588. Terakunpisut J, Gajaseni N, Ruankawe N (2007) Carbon sequestration potential in aboveground biomass of Thong Pha Phum National Forest, Thailand. Applied Ecology and Environmental Research, 5, 93–102. Trenberth KE, Dai A, van der Schrier G, Jones PD, Barichivich J, Briffa KR, Sheffield J (2014) Global warming and changes in drought. Nature Climate Change, 4, 17–22. Trumbore S, Da Costa ES, Nepstad DC et al. (2006) Dynamics of fine root carbon in Amazonian tropical ecosystems and the contribution of roots to soil respiration. Global Change Biology, 12, 217–229. Uhl C (1987) Factors Controlling Succession Following Slash-and-Burn Agriculture in Amazonia. The Journal of Ecology, 75, 377. Uhl C, Jordan CF (1984) Succession and Nutrient Dynamics Following Forest Cutting and Burning in Amazonia. Ecology, 65, 1476. Uhl C, Buschbacher R, Serrao EAS (1988) Abandoned Pastures in Eastern Amazonia. I. Patterns of Plant Succession. The Journal of Ecology, 76, 663. UNFCCC (2008) Reducing emissions from deforestation in developing countries: approaches to stimulate action. UNFCCC (2015) Adoption of the Paris Agreement. FCCC/CP/2015/L.9, 31. U.S. DOE (2012) Research Priorities for Tropical Ecosystems Under Climate Change Workshop Report, DOE/SC-0153. Vargas R, Allen MF, Allen EB (2008) Biomass and carbon accumulation in a fire chronosequence of a seasonally dry tropical forest. Global Change Biology, 14, 109–124. A cc ep te d A rt ic le This article is protected by copyright. All rights reserved. Veenendaal EM, Kolle O, Lloyd J (2004) Seasonal variation in energy fluxes and carbon dioxide exchange for a broad-leaved semi-arid savanna (Mopane woodland) in Southern Africa. Global Change Biology, 10, 318–328. Vieira S, Trumbore S, Camargo PB, Selhorst D, Chambers JQ, Higuchi N, Martinelli LA (2005) Slow growth rates of Amazonian trees: Consequences for carbon cycling. Proceedings of the National Academy of Sciences, 102, 18502–18507. Vourlitis GL, Priante Filho N, Hayashi MMS, Nogueira J de S, Raiter F, Hoegel W, Campelo, Jr JH (2004) Effects of meteorological variations on the CO2 exchange of a Brazilian transitional tropical forest. Ecological Applications, 14, 89–100. Weaver PL (2001) NPP Tropical Forest: Cinnamon Bay, U.S. Virgin Islands, 1982-1993, R1. Data set. Available on-line [http://www.daac.ornl.gov] from Oak Ridge National Laboratory Distributed Active Archive Center, Oak Ridge, Tennessee, U.S.A. doi:10.3334/ORNLDAAC/473. Webb EL, Fa’aumu S (1999) Diversity and structure of tropical rain forest of Tutuila, American Samoa: effects of site age and substrate. Plant Ecology, 144, 257–274. Wen, D., Wei, P., Kong, G., Zhang, Q., Huang Z (1997) Biomass study of the community of Castanopsis chinensis, Cryptocarya concinna, Schima superb in a southern China Reserve. Acta Ecologica Sinica, 17, 497–504. [in Chinese]. Wen, D., Wei, P., Zhang, Q., Kong G (1999) s on biomass of the lower subtropical evergreen broad- leaved forests in a MAB Reserve of south China. Acta Phytoecologica Sinica, 23, 11–21. [in Chinese]. Williams-Linera G (1983) Biomass and Nutrient Content in Two Successional Stages of Tropical Wet Forest in Uxpanapa, Mexico. Biotropica, 15, 275. Yamakura T, Hagihara A, Sukardjo S, Ogawa H (1986a) Aboveground biomass of tropical rain forest stands in Indonesian Borneo. Vegetatio, 68, 71–82. Yamakura T, Hagihara A, Sukardjo S, Ogawa H (1986b) Tree size in a mature dipterocarp forest stand in Sebulu, East Kalimantan, Indonesia. Southeast Asian Studies, 23, 452–478. Yan J, Wang Y, Zhou G, Zhang D (2006) Estimates of soil respiration and net primary production of three forests at different succession stages in South China. Global Change Biology, 12, 810– 821. Yi, W., Zhang, Z., Ding, M., Wang B (2000) Biomass and efficiency of radiation untilization in Erythrophleum fordii community. Acta Ecologica Sinica, 20, 397–403. [in Chinese]. Yu G-R, Zhu X-J, Fu Y-L et al. (2013) Spatial patterns and climate drivers of carbon fluxes in terrestrial ecosystems of China. Global Change Biology, 19, 798–810. A cc ep te d A rt ic le This article is protected by copyright. All rights reserved. Yu G, Chen Z, Piao S et al. (2014) High carbon dioxide uptake by subtropical forest ecosystems in the East Asian monsoon region. Proceedings of the National Academy of Sciences, 111, 4910– 4915. Zarin DJ, Ducey MJ, Tucker JM, Salas WA (2001) Potential Biomass Accumulation in Amazonian Regrowth Forests. Ecosystems, 4, 658–668. Zhang Z, Ding M (1996) Biomass and efficiency of radiation utilization in monsoon evergreen broadleaved forest in Dinghushan Biosphere Reserve. Acta Ecologica Sinica, 16, 525–534. [in Chinese]. Zhang Y, Tan Z, Song Q, Yu G, Sun X (2010) Respiration controls the unexpected seasonal pattern of carbon flux in an Asian tropical rain forest. Atmospheric Environment, 44, 3886–3893. Supporting Information Additional Supporting Information may be found in the online version of this article: Table S1. Classification of plots and areas included in the TropForC-db database. Table S2. Descriptive statistics on C stocks for young (<20) tropical forest stands. Table S3. Descriptive statistics on C stocks for intermediate-aged (20100)/ intact tropical forest stands. Table S5. Descriptive statistics on C fluxes for young (<20) tropical forest stands. Table S6. Descriptive statistics on C fluxes for intermediate-aged (20100)/ intact tropical forest stands. Table S8. Linear mixed-effects models for aboveground biomass in young and intermediate- aged stands. Table S9. Parameter estimates for linear mixed-effects model for aboveground biomass in young stands, where age and its interaction with regeneration type are fixed effects. Table S10. Parameter estimates for linear mixed-effects model for aboveground biomass in naturally regenerating young stands, where age and its interactions with regeneration type and FAO ecozone are fixed effects. Table S11. Parameter estimates for linear mixed-effects model for aboveground biomass in naturally regenerating intermediate-aged, where age is the fixed effect. A cc ep te d A rt ic le This article is protected by copyright. All rights reserved. Tables Table 1. List of studies with data included in TropForC-db, by region. Region Studies included Global / multi- regional* Clark et al., 2001b, 2013; IPCC, 2003, 2006; Anderson et al., 2006; Litton et al., 2007; Luyssaert et al., 2007; Baldocchi, 2008; Martin et al., 2013; Liu et al., 2014; Yu et al., 2014 Afrotropics Bartholomew et al., 1953; Egunjobi & Bada, 1979; DeAngelis et al., 1981; Kadeba, 1991; Esser et al., 1997; Nye & Greenland, 1998; Laclau et al., 2000; Clark et al., 2001b; Olson et al., 2001; Clark et al., 2013; Nygård et al., 2004; Onyekwelu, 2004, 2007; Veenendaal et al., 2004; Harmand et al., 2004; Epron et al., 2006; Lewis et al., 2009b Australasia Edwards & Grubb, 1977; Chen et al., 2003; Hutley et al., 2005; Leuning et al., 2005; Mialet- Serra et al., 2005; Roupsard et al., 2006; Beringer et al., 2007; Navarro et al., 2008; Clark et al., 2013; Stocker, 2013 Indo-Malaya Hozumi et al.; Ogawa et al., 1965; Kunstadter et al., 1978; Nakane, 1980; Yamakura et al., 1986a, 1986b; Behera et al., 1990; Dang & Wu, 1992; Chen et al., 1993, 2010; Peng & Zhang, 1994; Luo, 1996; Pinard & Putz, 1996; Zhang & Ding, 1996; Esser et al., 1997; Wen, D., Wei, P., Kong, G., Zhang, Q., Huang, 1997; Kira, 1998; Wen, D., Wei, P., Zhang, Q., Kong, 1999; Yi, W., Zhang, Z., Ding, M., Wang, 2000; Clark et al., 2001b, 2013; Ito & Oikawa, 2002; Fang et al., 2003; Hoshizaki et al., 2004; Swamy et al., 2004; Adachi et al., 2006; Jepsen, 2006; Yan et al., 2006; Hirano et al., 2007, 2009; Terakunpisut et al., 2007; Hirata et al., 2008; Kato & Tang, 2008; Kosugi et al., 2008; Ramachandran & Byrappa Gowdu Viswanathan, 2009; Chen, 2010; Kenzo et al., 2010; Van Do et al., 2010; Zhang et al., 2010; Aththorick et al., 2012; Chan et al., 2013; Proctor, 2013; Yu et al., 2013 Neotropics Snedaker, 1970; Ewel, 1971; Golley, 1975; Klinge et al., 1975; Folster et al., 1976; Scott, 1977; Crow, 1980; Tanner, 1980; Klinge & Herrera, 1983; Williams-Linera, 1983; Uhl & Jordan, 1984; Bongers et al., 1985; Frangi & Lugo, 1985; Uhl, 1987; Saldarriaga et al., 1988; Lugo et al., 1990; Lugo, 1992; Guimaraes, 1993; Overman et al., 1994; Salomão, 1994; Szott et al., 1994; Aide et al., 1995; Brown et al., 1995; Alves et al., 1997; Delaney et al., 1997; Lucas et al., 1998, 2002; Malhi et al., 1998, 1999, 2004; Gehring et al., 1999, 2005; Grimm & Fassbender, 1999; Hughes et al., 1999b, 2000, 2002; Jordan et al., 1999; Parrotta, 1999; Clark & Clark, 2000; Montagnini, 2000; Sorrensen, 2000; Steininger, 2000; Chambers et al., 2001, 2004; Chave et al., 2001; Clark et al., 2001b, 2013; Keller et al., 2001, 2004; Maass & Martinez- Yrizar, 2001; Weaver, 2001; Zarin et al., 2001; Araújo, 2002; Carswell et al., 2002; Davidson et al., 2002, 2004; Falge et al., 2002; Fehse et al., 2002; Kraenzel et al., 2003; Loescher et al., 2003; Read & Lawrence, 2003; Saleska, 2003; Santos et al., 2003; Baker et al., 2004; Feldpausch et al., 2004; Li et al., 2004; Miller et al., 2004; Rice et al., 2004; Silver et al., 2004; Stape et al., 2004; Vourlitis et al., 2004; Vieira et al., 2005; Cleveland & Townsend, 2006; Trumbore et al., 2006; Hutyra et al., 2007; Marín-Spiotta et al., 2007; Palace et al., 2007; Sierra et al., 2007a, 2007b, 2012; Terakunpisut et al., 2007; Vargas et al., 2008; Aragão et al., 2009; Letcher & Chazdon, 2009; Girardin et al., 2010; Schöngart & Wittmann, 2010; Van Do et al., 2010; Fonseca et al., 2011; Moser et al., 2011; Mascaro et al., 2012; Clark, 2013; Orihuela-Belmonte et al., 2013; Araujo-Murakami et al., 2014; Becknell & Powers, 2014; Broadbent et al., 2014; da Costa et al., 2014; del Aguila-Pasquel et al., 2014; Doughty et al., 2014; Rocha et al., 2014 Oceania Webb & Fa’aumu, 1999; Clark et al., 2001b, 2013; Giardina et al., 2003, 2004; Ryan et al., 2004; Schuur, 2005 *These data compilations were used to identify original studies relevant to our database. A cc ep te d A rt ic le This article is protected by copyright. All rights reserved. Table 2. Overview of TropForC-db structure and content. Bold indicates variables that link data tables. Table Description Content 1) Sites Geographic, climatic, and edaphic site data Site; city, state, country, geographic coordinates, elevation, mean annual temperature and precipitation, and soil descriptors from original publications or subsequent compilations; area (as grouped here). Additional data acquired or derived from global maps and included in the sites table are biogeographic zone, FAO ecozone, Köppen-Geiger climate zone, and canopy leaf attributes from SYNMAP. 2) Plots & history Known history of each plot or set of replicate plots Site; plot; plot area; date (with its certainty) and level*of known events (distttype) of relevance to the C cycle. When plots have no known disturbance history, the year to which an absence of disturbance is known with confidence is recorded. 3) Measurements Records of ecosystem- level measurements relevant to C cycling Citation; site; plot; dominant vegetation type (PFT/ species)†; stand age at time of measurement; variable name; method_id; measurement dates and their certainty, sample size; measured values and their uncertainty, covariates that are important to interpreting trait data, including allometric_equation 4) Variables Definitions of variables Variable name, units, description, equations, notes, and associated covariates 5) Disturbance Type Definition of disturbance, management or regeneration history event types. Disturbance category, disttype, description, units (when applicable) 6) Plant functional types/ species Definitions of species/ PFT codes PFT (plant functional type), description 7) Methodology Description of methodologies method_id; method citation, variable, notes 8) Allometries Sources and description of allometric equations allometric_equation; citation for equation source; notes A cc ep te d A rt ic le This article is protected by copyright. All rights reserved. Table 3. TropForC-db C flux variables included in Figs 3-5. Shown are variable names and descriptions, the associated variable name(s) in the database, number of records (n), and number of plots (n plots) in the database. All units are Mg C ha-1 yr-1. Complete list of variables is available in Anderson-Teixeira et al. (2016). Variable Description Variable name(s) n n plots Ecosystem C balance- NEE or NEP Annual net ecosystem exchange (- indicates C sink) or net ecosystem production (+ indicates C sink) NEE_annual, NEP_annual 84 29 Production- GPP Annual gross primary production GPP_annual, GPP_C_annual 93 48 NPP Annual net primary production NPP_[1-5](_C) 71 42 ANPP Aboveground NPP ANPP_[1-2](_C) 116 71 ANPPstem Annual stem production; i.e., annual stem aboveground biomass increment ANPP_stem(_C) 201 185 ANPPfoliage Annual foliage production, typically estimated as annual leaf litterfall ANPP_foliage(_C) 73 58 ANPPlitterfall Annual litterfall, including leaves, reproductive structures, and sometimes woody material ANPP_litterfall_[1- 2](_C) 19 16 ANPPfolivory Annual productivity consumed by folivores ANPP_folivory(_C) 112 62 BNPP Total annual belowground NPP BNPP_root(_C) 46 37 BNPPcoarse Annual coarse root production BNPP_coarse root(_C) 44 41 BNPPfine Annual fine root production BNPP_fine root(_C) 42 37 Respiration- Reco Annual ecosystem respiration Reco_annual 66 26 Ra Annual autotrophic respiration R_auto_annual 14 13 Rsoil Annual soil respiration Rsoil_annual 49 33 Rh,soil Annual heterotrophic soil respiration Rsoil_het_annual 2 2 Rh Annual heterotrophic respiration R_het_annual 29 25 Other flux variables 132 123 A cc ep te d A rt ic le This article is protected by copyright. All rights reserved. Table 4. TropForC-db C stock variables included in Figs 3-5. Shown are variable names and descriptions, the associated variable name(s) in the database, number of records (n), and number of plots (n plots) in the database. All units are Mg C ha-1. Complete list of variables is available in Anderson-Teixeira et al. (2016). Variable Description Variable name(s) n n plots Living- Biomass Total live biomass C Biomass_total; C_total 180 105 Aboveground biomass Aboveground live biomass C Biomass_ag; C_ag 105 1 672 Foliage biomass Foliage biomass C Biomass_ag; C_ag 157 115 Root biomass Total root biomass C Biomass_root_total; C_root_total 258 149 Coarse root biomass Coarse root biomass C Biomass_root_coarse ; C_root_coarse 48 30 Fine root biomass Fine root biomass C Biomass_root_fine; C_root_fine 82 53 Nonliving- Dead wood Dead wood, including standing dead wood and coarse woody debris. (C_)Deadwood, Downdeadwood, Standingdeadwood 135 105 Organic layer Organic layer (“forest floor”) C. (C_)Organic layer 106 79 Other stock variables 204 154 A cc ep te d A rt ic le This article is protected by copyright. All rights reserved. Figure captions Figure 1. Geographical distributions of sites included in TropForC-db. Map shows satellite-derived coverage of evergreen and deciduous forest (from SYNMAP; Jung et al., 2006). Inset shows distribution of sites, plots, and records among biogeographic regions (sensu Olson et al., 2001). Figure 2. Histogram of stand age distribution for young and intermediate stands within TropForC- db, broken down by regeneration types, in the order as shown in the legend. The database also contains 380 mature forest plots (old-growth or age>100; not shown here). Figure 3. Diagram of major C stocks and flows in mature (>100 year)/ intact tropical broadleaf evergreen forests. Variables are as described in Tables 3-4, and detailed descriptive statistics are given in Tables S4 and S7. All units are Mg C ha-1 (stocks) or Mg C ha-1 yr-1 (flux). Arrow width is scaled according to magnitude of flux. Dashed arrows indicate fluxes for which no data are included in the database. Figure 4. Diagram of major C stocks and flows in young (<20 year) naturally regenerating tropical broadleaf evergreen forests. For variables with a signficant age trend, regression parameters are given. Detailed descriptive statistics are given in Tables S2 and S5. Other numbers and symbols are as in Fig. 3. Figure 5. Diagram of major C stocks and flows in intermediate-aged (20-100 year) naturally regenerating tropical broadleaf evergreen forests. For variables with a signficant age trend, regression parameters are given. Detailed descriptive statistics are given in Tables S3 and S6. Other numbers and symbols are as in Fig. 3. Figure 6. Age trends in (a) aboveground biomass and (b) dead wood for forest regeneration following stand-clearing disturbances. Maps show the sites from which data were obtained. Forests are grouped by young, intermediate-aged, and mature/intact, showing separate statistical fits for each age class. Regression lines indicate mixed model where age and regeneration type (aboveground biomass in young stands only) are fixed effects and plot nested within area is a random effect (Tables S5-S7, S9, S11). A cc ep te d A rt ic le This article is protected by copyright. All rights reserved. A cc ep te d A rt ic le This article is protected by copyright. All rights reserved. A cc ep te d A rt ic le This article is protected by copyright. All rights reserved. A cc ep te d A rt ic le This article is protected by copyright. All rights reserved. A cc ep te d A rt ic le This article is protected by copyright. All rights reserved. A cc ep te d A rt ic le This article is protected by copyright. All rights reserved. * Level is recorded for some event types. For example, for harvest, a level of 100% indicates clear cut. †For dominant vegetation type, the PFT or species that encompasses all trees in the stand is recorded. PFT categories range in their specificity; for example, “broadleaf evergreen trees” would be applied to a stand dominated entirely by evergreen trees, whereas “broadleaf trees” would be applied to a mixed evergreen/deciduous stand. A single species is recorded only in the case of monoculture plantations. Species and PFT are recorded in the measurements table (as opposed to the sites or treatments & history tables) because species composition changes over time.