Get 20M+ Full-Text Papers For Less Than $1.50/day. Subscribe now for You or Your Team.

Learn More →

The biodiversity cost of carbon sequestration in tropical savanna

The biodiversity cost of carbon sequestration in tropical savanna SCIENCE ADVANCES RESEARCH ARTICLE FOREST ECOSYSTEMS Copyright © 2017 The Authors, some rights reserved; The biodiversity cost of carbon sequestration in exclusive licensee American Association tropical savanna for the Advancement of Science. No claim to 1 1 2 3,4 Rodolfo C. R. Abreu, William A. Hoffmann, * Heraldo L. Vasconcelos, Natashi A. Pilon, original U.S. Government 5 3 Davi R. Rossatto, Giselda Durigan * Works. Distributed under a Creative Tropical savannas have been increasingly viewed as an opportunity for carbon sequestration through fire suppres- Commons Attribution sion and afforestation, but insufficient attention has been given to the consequences for biodiversity. To evaluate NonCommercial the biodiversity costs of increasing carbon sequestration, we quantified changes in ecosystem carbon stocks and the License 4.0 (CC BY-NC). associated changes in communities of plants and ants resulting from fire suppression in savannas of the Brazilian −1 −1 Cerrado, a global biodiversity hotspot. Fire suppression resulted in increased carbon stocks of 1.2 Mg ha year since 1986 but was associated with acute species loss. In sites fully encroached by forest, plant species richness declined by 27%, and ant richness declined by 35%. Richness of savanna specialists, the species most at risk of local extinction due to forest encroachment, declined by 67% for plants and 86% for ants. This loss highlights the impor- tant role of fire in maintaining biodiversity in tropical savannas, a role that is not reflected in current policies of fire suppression throughout the Brazilian Cerrado. In tropical grasslands and savannas throughout the tropics, carbon mitigation programs that promote forest cover cannot be assumed to provide net benefits for conservation. INTRODUCTION hotspot (17) and putatively the most species-rich savanna region in Carbon mitigation programs that protect and expand tropical forests the world (18, 19). To understand the long-term impact of current fire are widely embraced as a win-win situation for conservation because policies, we quantified changes in biodiversity and ecosystem carbon they presumably enhance species diversity (1, 2). This presumption stocks at three savanna-forest boundaries actively undergoing forest requires closer scrutiny in the seasonally dry tropics, where large areas encroachment due to fire suppression. At each site, we measured soil of natural grasslands and savannas could be converted to forest and vegetation carbon stocks and surveyed plant communities across through afforestation or fire suppression (3–5). In these biomes, it a network of 30 plots ranging from open savanna (mean tree basal 2 −1 cannot be assumed that increased tree cover will generate net benefits area of 3.1 m ha ) to recently formed forest (mean basal area of 2 −1 for biodiversity because savannascontain many endemiclight- 21.5 m ha ). Aboveground-foraging ant communities were also demanding species that can be displaced or eliminated by forest cover surveyed, serving as a representative animal group. We used a 30-year (3, 6, 7); thus, species-rich savannas may undergo a net loss of species series of Landsat satellite images to quantify rates of vegetation change when replaced by forest. For this reason, it is important to rigorously and to hindcast carbon sequestration and the resulting species loss. quantify the impacts on biodiversity of management practices that seek to maximize carbon sequestration by promoting forest cover. Where savanna management is not aimed specifically at car- RESULTS bon sequestration, other factors may lead to the same outcome of The study area has undergone a steady increase in tree cover over the increased tree density and ecosystem carbon stocks. Tree densities past 30 years. Over this time, there was a continuous increase in the have been increasing in many savannas across the tropics (8), likely enhanced vegetation index (EVI; Fig. 1), a satellite-derived metric that because of fire suppression and increasing atmospheric CO (9–12). is strongly and positively correlated with tree basal area, leaf area index In the savannas of the Brazilian Cerrado, governmental regulations (LAI), and ecosystem carbon stocks across our study sites (figs. S1 and severely restrict the use of fire, and fire suppression is actively S2). As inferred from historical EVI, all plots that are now forest were practiced in most reserves and parks (13) because of widespread per- occupied by savanna or grassland in 1986, corroborating observations ception that fire is detrimental to biodiversity. This is compounded by by reserve personnel working at the site during this period. expansion of crops, pasture, and forest plantations at alarming rates In 2015, ecosystem carbon stocks (plants + surface soils) ranged −1 −1 (14, 15), resulting in highly fragmented landscapes. Because such frag- from a mean of 20.8 Mg ha in open savanna sites to 83.5 Mg ha mentation impedes fire spread, natural ignitions alone are unable to in forest. The gradient in carbon stocks can be attributed primarily to a maintain fire frequencies that were once possible (16). 14-fold increase in tree biomass and, to a lesser extent, a 2-fold We assessed the impact of 30 years of fire suppression and forest increase in surface soil stocks (0 to 20 cm) over the vegetation gradient encroachment in savannas in the Brazilian Cerrado, a biodiversity (Fig. 2A). The annual increase in ecosystem carbon storage was −1 −1 1.19 Mg ha year over the past three decades, as estimated by combining the time series of EVI (Fig. 1) with the empirical relation- Department of Plant and Microbial Biology, North Carolina State University, Raleigh, NC 27695–7612, USA. Instituto de Biologia, Universidade Federal de Uberlândia (UFU), ship between EVI and ecosystem carbon (fig. S2). Av. Pará 1720, Uberlândia, Minas Gerais 38405-320, Brazil. Laboratório de Ecologia e The increase in tree biomass over the vegetation gradient was ac- Hidrologia Florestal, Floresta Estadual de Assis, Instituto Florestal, Assis, São Paulo companied by decreases in the richness of both plant and ant species, 19802-970, Brazil. Instituto de Biologia, Universidade Estadual de Campinas (UNICAMP), Caixa Postal 6109, Campinas, São Paulo 13083-865, Brazil. Departamento de Biologia, with particularly large losses in savanna specialists (Fig. 2). The loss of Universidade Estadual Paulista (UNESP), Campus de Jaboticabal, Jaboticabal, São savanna species was most acute when vegetation reached a stem basal Paulo 14884-900, Brazil. 2 −1 area of approximately 15 m ha and an LAI of 2.5, a threshold that *Corresponding author. Email: wahoffma@ncsu.edu (W.A.H.); giselda.durigan@ corresponds roughly to the transition between savanna and forest as gmail.com (G.D.) Abreu et al., Sci. Adv. 2017; 3 : e1701284 30 August 2017 1of7 | SCIENCE ADVANCES RESEARCH ARTICLE Sites that are currently open shrub savanna Sites that are currently tree-dominated savanna Sites that are currently forest 0.6 0.6 0.6 Forest 0.5 0.5 0.5 Dense savanna Open savanna 0.4 0.4 0.4 0.3 0.3 0.3 0.2 0.2 0.2 1985 1990 1995 2000 2005 2010 2015 1985 1990 1995 2000 2005 2010 2015 1985 1990 1995 2000 2005 2010 2015 Year Fig. 1. Historical changes in EVI over 30 years of fire suppression. Historical changes in the EVI over 30 years of fire suppression, as determined by Landsat images. Each study plot is denoted by a different symbol and a separate regression line. indicated by the disappearance of C grasses (20, 21). Forest sites had encroachment, whereas diversity of shrubs and herbaceous plants 69% fewer savanna plant species and 74% fewer savanna ant species, declined markedly (fig. S3). The decline in these groups is particularly compared to dense savanna. This decline in savanna species was par- noteworthy because shrubs and herbaceous plants comprise 75 to 80% tially offset by an increase in the number of forest specialists (Fig. 2) of plant species in the Cerrado region (25) and 58% of species at the thus, when all species were considered, forest sites had 33% fewer plant study sites (table S1). Of these nontree species, 79% are savanna spe- species and 34% fewer ant species than dense savanna. Because species cialists and are therefore particularly sensitive to forest encroachment. richness of each group was strongly correlated with EVI (r =0.65to Herbaceous communities are vastly understudied in the Cerrado, 0.86, P < 0.0001; fig. S2), we used historical changes in EVI (Fig. 1) to compared to trees, but deserve greater attention considering their high estimate changes in species diversity since 1986. In the sites that are diversity and vulnerability to fire suppression. now forest, we estimate that 27% of plant species and 35% of ant spe- In sites encroached by forest, a few savanna plant species still per- cies have been lost during 30 years of fire suppression, whereas the loss sist at low abundances, and these are nearly certain to continue declining of savanna specialists was 67 and 86% for plants and ants, respectively. if fire is not reintroduced into the system. Some species may depend The loss of plant species involved primarily shrubs and herbaceous directly on fire for reproduction (26, 27), but more generally, plant species, whereas the number of tree species increased (fig. S3). species adapted to savanna environments cannot tolerate the low light These concomitant changes in diversity and ecosystem properties availability under a forest canopy (5, 28, 29). These species mostly resulted in a strong trade-off between ecosystem carbon stocks and persist in forest as occasional moribund individuals with few leaves. species richness across the savanna-forest gradients (Fig. 3). Over A few robust savanna trees remain in the forests, but these are rare the ranges of tree densities studied, each 1% increase in ecosystem car- individuals that are sufficiently tall to access ample light. Even these bon resulted in a 0.27% decrease in total plant species richness and a species are unlikely to persist beyond the life span of these individuals 0.43% decrease in ant species richness (Fig. 3), as determined from the because their seedlings are nearly absent from sites encroached by forest. sensitivity coefficient. When only savanna species were considered, In short, we expect continued loss of savanna specialist species in forest- each 1% increase in ecosystem carbon corresponds to decreases of encroached sites, and the results presented here should be considered 0.98 and 1.38% for plants and ants, respectively (Fig. 3). Together, forest an underestimate of the eventual impact of fire suppression. and generalist plant species increased by 0.76% for each 1% increase in Similar responses to forest encroachment are expected for animal carbon, but there was no detectable change for ants. communities, represented here by ants because they are a diverse Stem basal area in 2015 was significantly correlated with properties group that lend themselves to systematic sampling. As for plants, of surface soils (0 to 20 cm). Sites with higher basal area had greater ant communities are strongly dependent on vegetation structure silt, clay, and organic matter; higher cation exchange capacity (CEC); (30, 31); consequently, forest encroachment results in a notable loss higher availabilities of P, K, B, and S (fig. S4); and higher concentra- of ant species adapted to open vegetation (Fig. 2C). These savanna- tions of P, K, B, and S (fig. S4). Higher basal area was associated with dependent species have evolved within most or all major groups of lower pH, sand, and Cu (fig. S4). animals of the Cerrado region, including reptiles (32), amphibians (33), birds (19), and mammals (34). Among these are species of par- ticular conservation concern, such as the maned wolf (35), Pampas DISCUSSION deer (36), and the greater rhea (37). The interrelationships between Previous studies in the Cerrado have suggested that forest encroach- vegetation, fauna, and fire are complex (38), but generally, animal spe- ment into open savanna results in a net increase in plant species rich- cies specialized to savanna and grassland should be considered vulner- ness (6, 22–24), but these studies have been limited to trees and able to fire suppression and forest encroachment. therefore overlook the high diversity of shrubs and herbaceous plants Forest encroachment at this site was made possible by long- in savannas. Here, diversity of tree species increased with forest term fire suppression, although increasing atmospheric CO may Abreu et al., Sci. Adv. 2017; 3 : e1701284 30 August 2017 2of7 Enhanced vegetation index | SCIENCE ADVANCES RESEARCH ARTICLE A Plants All species r = 0.97 Savanna species Trees 120 Soil (0–20 cm) Herbaceous Total 2 r = 0.53 r = 0.98 r = 0.76 r = 0.74 r = 0.66 20 Forest and generalist species Savanna species 0 All species 2 B Ants r = 0.64 50 2 r = 0.69 r = 0.79 r = 0.71 2 20 r = 0.87 Forest and generalist species Savanna species All species r = 0.67 0 20406080 100 −1 Ecosystem carbon stocks (Mg ha ) 2 Fig. 3. Carbon-biodiversity trade-offs across savanna-forest transitions. Carbon- r = 0.03 biodiversity trade-offs across savanna-forest transitions. Relationships between ecosystem carbon stocks and species richness of (A) plants and (B) ants. All relation- r = 0.84 ships are statistically significant (P < 0.0001). Fitted relationships: all plants, y = −0.271 −0.975 −0.426 245.88x ; savanna plants, y = 1953.4x ; all ants, y = 198.68x ; savanna 10 −1.380 ants, y = 2258.8x . and only one was burned more than once in these three decades. 0 5 10 15 20 25 Fire is not likely to be the only factor responsible for this variation 2 −1 Stem basal area (m ha ) because encroachment was greatest in soils that have greater water Fig. 2. Carbon stocks and species richness due to forest encroachment. Changes retention capacity (high silt and clay and low sand; fig. S4) and in (A) carbon stocks, (B) plant species richness, and (C) ant species richness over higher concentrations of P, K, B, and S (fig. S4). These soil variables gradients of forest encroachment. All variables were significantly correlated with basal were measured in the top 20 cm of the soil profile, where vegetation area (P < 0.0001) except for the number of forest and generalist ants (P =0.42). can have large influences on soil chemistry (39); thus, causality is difficult to ascertain. Nevertheless, the gradient in sand content re- veals preexisting soil differences that likely influenced rates of forest have contributed by accelerating tree establishment and growth (9–12). encroachment under fire suppression. Regardless of the role that However, encroachment was not uniform across the study area, with soils and nutrients may play in controlling the rate of forest expan- substantial variation in both initial density and subsequent rate of sion (40–43), tree encroachment has continued steadily in all savanna increase (Fig. 1). Previous fire history likely contributed to the initial sites (Fig. 1). differences, because fires were common before the 1980s and would If the current trajectory is not interrupted, we expect most or not have burned uniformly across the area. However, during the all high-quality savanna to be lost from the study area within a decade 30-year study period, only four of the most open plots had burned, or two, with obvious negative consequences for biodiversity. Abreu et al., Sci. Adv. 2017; 3 : e1701284 30 August 2017 3of7 −1 Number of ant species Number of plant species Carbon stocks (Mg ha ) Number of species | SCIENCE ADVANCES RESEARCH ARTICLE Reintroduction of fire would reverse the process of tree encroach- conserving savanna landscapes and biota, burning should occur with ment, but to avoid catastrophic loss of biodiversity, it is important sufficient frequency and intensity to prevent tree densities from to restore a natural fire regime well before the canopy becomes approaching the point at which light-demanding species fail to thrive 2 −1 dense enough to substantially impact the herbaceous layer. Once (that is, basal area of approximately 15 m ha ). This should be the diversity of the herbaceous community has been substantially considered the absolute maximum acceptable tree density for avoiding depleted, it is unlikely to recover readily, even after the vegetation is catastrophic loss of savanna species, and even this density may exceed opened. Restoring “old-growth” grasslands such as these is notoriously the point at which prescribed burning can safely and effectively reverse difficult (44), particularly in the Cerrado, where receding forest is tree encroachment. Ideally, consideration should be given to main- typically replaced by aggressive exotic grasses (21, 45) that preclude taining the mosaic of grassland (campo limpo), shrub savanna (campo the recovery of native herbaceous species (45). sujo), and tree savanna (cerrado sensu stricto) that may have histori- cally occurred at a site. Ultimately, fire management in the Cerrado Implications for policy and management should evolve to consider the ecological role of fire, such as stimulating These findings challenge the widespread assumption that promoting flowering in herbaceous plants and influencing exotic species, but this forest cover has universal co-benefits for carbon sequestration and requires observations that can accrue only if land managers and biodiversity (1, 2). Forest encroachment resulted in increased eco- scientists are allowed to use prescribed fire as a management tool. system carbon at the expense of reduced species diversity, an outcome Until this information is available for the Cerrado, managers can use that is likely repeated wherever forest encroaches on the Cerrado and knowledge from long-term fire experiments in Africa and Australia other species-rich savannas and grasslands in the world. Under these (53, 54), as well as traditional knowledge on fire management in conditions, a carbon-centered conservation strategy may be inappropriate the Cerrado (55). Meanwhile, it is urgent to reformulate the current (46, 47), except in excessively burned savannas, where modest reduc- timid policies that favor fire suppression over all other options. tions in fire frequency or intensity benefit both biodiversity and carbon storage (48). The results presented here are particularly alarming in light of the MATERIALS AND METHODS current policy of fire suppression throughout the Cerrado region (13). Experimental design Policies of fire suppression in Brazil are exacerbated by ongoing land- Study area use change and vegetation fragmentation, which restrict fire spread The Santa Barbara Ecological Station (SBES) is located in the munic- and thereby reduce fire frequency (16). In large, unbroken expanses ipality of Águas de Santa Bárbara, São Paulo State, Brazil, with an area of the Cerrado, lightning fires can sustain frequent burning (49), but of 2715 ha (56). It lies within 22°46′33″ and 22°50′33″ Sand 49°10′27″ this becomes increasingly improbable as vegetation is fragmented and 49°15′36″ W, near the southeastern edge of the Cerrado region, at (16). Where lightning fires are not replaced with sufficient human- altitudes between 600 and 680 m above sea level. The climate in the initiated fires, biodiversity losses on protected lands, as observed here, area is Köppen Cwa-type (56), with monthly mean temperatures will be increasingly common throughout the Cerrado. Even greater varying between 16° and 24°C. The annual mean rainfall is between losses should be expected in the core of the Cerrado region, which 1100 and 1300 mm, with dry winters and rainy summers. The soils of supports greater diversity of savanna species than the study area the study area are deep oxisols with high sand content, low nutrient (50), which lies near the southeastern limit of the Cerrado. To date, content, and high saturation of aluminum (57). the consequences of excluding fire are much more evident in the study Data collection area (51) because of its longer history of habitat fragmentation and The study was undertaken at three savanna-forest transitions located protection from fire. This history offers important lessons that are at distances of 2 to 5 km from each other. A network of 30 plots was relevant throughout the Cerrado andother tropical savannaregions. distributed across these transitions to include 12 in recently formed forest, Environmental legislation in Brazil advanced a step in the right 12 in tree-dominated savanna, and 6 in shrub-dominated savanna. direction by legalizing prescribed fires for conservation of protected These vegetation classes correspond to cerradão, cerrado sensu stricto, areas in the Cerrado (Law number 12651, Article 38, in effect since and campo sujo of the Brazilian classification system (25). Each plot May 2012), but to date, controlled fires have been used only in a few was 20 m × 50 m (0.1 ha), within which the stem diameter of all trees reserves whose management plans specifically permit prescribed greater than 5 cm were measured. All trees and shrubs with stem dia- burning. However, burning cerrado vegetation outside formal re- meters between 1 and 5 cm were measured within 10 5-m × 5-m sub- serves remains illegal (13). This situation could change if the expected plots distributed regularly throughout each plot. Tree seedlings and National Fire Policy mandated by law were implemented (Law num- the herbaceous community were sampled within 40 1-m × 1-m sub- ber 12651, Article 40), regulating the use of fire in natural vegetation plots distributed in a grid within each 0.1-ha plot. All individuals throughout the Cerrado ecoregion. Nevertheless, the overall percep- sampled within these plots were identified to species in the field or tion that fire is destructive can be an obstacle even more difficult to collected for subsequent identification. surpass because managers of protected areas hesitate to prescribe fire Plant species were categorized into eight growth forms: (i) grass; in fear of formal punishment or social condemnation (13). (ii) sedge; (iii) forb, that is, nongraminoid plant without a Negative attitude toward fire is difficult to overcome; however, our persistent woody stem; (iv) subshrub, that is, woody plant typically results offer important direct evidence of biodiversity loss due to less than 75 cm tall as adults; (v) shrub; (vi) palm; (vii) treelet; and woody encroachment when fire is eliminated in the Cerrado. The ex- (viii) tree. On the basis of literature (58, 59), species were also pansion of forests and trees in savanna ecosystems may be a global classified according to habitat preference into three groups: (i) savanna phenomenon, attributed to a complex of different factors in each specialists, (ii) forest specialists, and (iii) generalists, that is, species that region (8, 52), but there is no question that fire is an effective tool to are commonly found in both savanna and forest habitats (table S1). prevent it. Our results indicate that, on any Cerrado land managed for Some species of the latter group may be true generalists that persist Abreu et al., Sci. Adv. 2017; 3 : e1701284 30 August 2017 4of7 | SCIENCE ADVANCES RESEARCH ARTICLE indefinitely in either environment, but most are best considered as at intervals of 10 m along its longest axis. The soil samples were transitional species typically found in forest but readily colonize submitted to a commercial laboratory to quantify coarse sand, fine savanna under fire suppression. sand, silt, and clay fractions; extractable Ca, Mg, K, B, Cu, Mn, Fe, The aboveground-foraging ant community was characterized for and Zn; available P; exchangeable Al; pH and organic matter; and each of the 0.1-ha plots placed in forest or tree-dominated savanna; the derived variables: potential acidity (H + Al), CEC, exchangeable the six plots in shrub-dominated savanna were omitted. Five 2.5-m × acidity, and base saturation and aluminum saturation. A subset of 2.5-m grids were established along the borders of each plot, keeping a 16 plots was sampled to determine carbon stocks in litter, duff, and minimum distance of 20 m between any two grids. Four pitfall traps herbaceous plant pools. These were sampled in four or five 0.25-m were set in each grid, totaling 20 traps per plot. Pitfall traps consisted subplots adjacent to each plot. In each subplot, the ground layer of a small plastic cup (250 ml, 8.5 cm high, and 7.8 cm in diameter) biomass was collected and divided into grass, sedge, and dicot partially filled with water and detergent. The traps remained in fractions. Litter was divided into leaf and wood fractions. When a dis- operation for 48 hours. All ant specimens collected were sorted to cernable organic horizon (duff) was present, this was also collected in morphospecies and, whenever possible, identified to species using a 0.0625-m portion of the subplot. All these components were available taxonomic keys or through comparison with specimens pre- weighed after drying for at least 5 days at 60°C. Each dried duff sam- viously identified by experts and deposited at the Zoological Collec- ple was thoroughly mixed, and a subsample was weighed before tion of the Federal University of Uberlândia, Brazil. ashing in a muffle furnace to determine the fraction of organic matter. Each ant species or morphospecies was classified as savanna spe- The carbon content of the lost mass was calculated using a conversion −1 cialist, forest specialist, or habitat generalist (table S2). Because the factor of 0.58 kg C kg organic matter. The combined carbon content habitat preference of some ant species could not be determined from of litter, duff, and herbaceous plants was summed after scaling each to −1 existing literature sources, these species were classified on the basis of similar units (Mg ha ). Because this combined pool accounted for a abundance at the site. A species was considered a specialist if it oc- small fraction of the ecosystem carbon pools (mean, 13%; range, 9 to curred exclusively, or nearly so, in either savanna or forest and was 21%) and was strongly correlated to tree basal area (r = 0.70), we used considered a generalist if it occurred at similar frequencies in both regression to estimate the size of this pool for the 14 plots for which this habitats. To eliminate the possibility that this post hoc classification was not measured directly. might introduce spurious relationships when testing for vegetation Total tree biomass was estimated for each plot using allometric impact on species richness within the functional types, we also applied equations developed for trees in a range of Cerrado physiognomies an alternate approach to classifying species using subsets of the data. and including both root and shoot biomass (64) as follows For this, abundance data from two of the savanna-forest boundaries were used to classify species, and then this classification scheme was Savanna plots : lnðMÞ¼  1:6516 þ 0:7643* lnðd hÞð1Þ used to determine the number of savanna, forest, and generalist spe- cies at the third boundary. By using successive subsets of the data in this way, the test for vegetation response was independent of the data Forest plotsðcerrada˜oÞ : lnðMÞ¼  2:8573 þ 0:9556*lnðd hÞð2Þ used to classify the species. Fewer species could be classified as habitat specialists in this manner, but in relative terms, results were stronger than when analyzed in the previous manner. Specifically, forest en- where M is the predicted dry biomass (in kilograms), d is the stem croachment was found to result in an 87% decline in the mean number diameter (in centimeters), and h is the height (in meters). Mass of −1 of savanna ant species per plot (3.0 in forest versus 22.3 in savanna) carbon was calculated using a conversion factor of 0.5 kg C kg but a 3.5-fold increase in the number of forest species (12.3 versus 3.7). biomass (65, 66). Consequently, we used the classification scheme based on the full data Remote sensing set for all subsequent analyses because this allowed a greater number To quantify vegetation change over 30 years of fire suppression, we of species to be classified. used vegetation indices derived from Landsat satellite images and ob- In each of the 30 vegetation plots, the LAI of the overstory was tained from the USGS (United States Geological Survey) portal (https:// espa.cr.usgs.gov/). We first evaluated the performance of four vegeta- measured with hemispherical canopy photographs taken at each of tion indices, normalized difference vegetation index (NDVI), enhanced the 40 subplots used to characterize the ground layer community. To ensure conditions of diffuse light, all photos were taken before vegetation index (EVI), soil-adjusted vegetation index (SAVI), and sunrise, after sunset, or under homogeneous overcast skies. A tripod modified soil-adjusted vegetation index (MSAVI) calculated from sur- was used to position the camera (Canon EF 8-15mm fisheye lens) at a face reflectances derived from Landsat 8. Of these four indices, the EVI height of 1 m, and the top of the camera was oriented relative to the was most strongly correlated with tree basal area (r =0.75) andLAI north. Photos were taken with an underexposure of one f stop (60), (r = 0.89; fig. S2); thus, all further analyses were undertaken with this andthe colorimageswereconverted to blackand whiteusing index. To determine long-term vegetation trends, we used all available Hemisfer 2.12 (61, 62) and using maximum blue contrast (63). The images from 1986 to 2016 that were acquired by Landsat 5–8under images were then analyzed with Hemisfer 2.12 using an automatic cloudless conditions and during the wet season months of November threshold for closed-canopy vegetation and with a supervised manual through May. For each image, EVI was spatially averaged for each study threshold under open canopies. The LAI values were averaged over plot. This resulted in multiple values for each calendar year, which were the 40 subplots to obtain a single value for each 0.1-ha plot. then composited by determining the yearly maximum. Ecosystem carbon and soil properties Statistical analysis Samples of surface soil (0- to 20-cm depth) were collected for each Analyses were performed with the base stats package in R (67). The of the 30 plots. A composite soil sample was made with five sub- lm function was used to fit linear and quadratic models, and the samples from surface soil collected immediately adjacent to the plot package nlstools (68) was used for fitting nonlinear models. Abreu et al., Sci. Adv. 2017; 3 : e1701284 30 August 2017 5of7 | SCIENCE ADVANCES RESEARCH ARTICLE 15. R. Beuchle, R. C. Grecchi, Y. E. Shimabukuro, R. Seliger, H. D. Eva, E. Sano, F. Achard, To characterize the relative change in species richness (r) for a Land cover changes in the Brazilian Cerrado and Caatinga biomes from 1990 to relative change in ecosystem carbon (C), we calculated the coefficient 2010 based on a systematic remote sensing sampling approach. Appl. Geogr. dr C of sensitivity, ; also known as elasticity in economics and demog- dC r 58,116–127 (2015). raphy (69). This was calculated as the slope of the relationship between 16. S. Archibald, A. C. Staver, S. A. Levin, Evolution of human-driven fire regimes in Africa. Proc. Natl. Acad. Sci. U.S.A. 109, 847–852 (2012). ln(r) and ln(C). 17. N. Myers, R. A. Mittermeier, C. G. Mittermeier, G. A. B. da Fonseca, J. Kent, Biodiversity We estimated past changes in carbon stocks and species number by hotspots for conservation priorities. Nature 403, 853–858 (2000). using a space-for-time substitution. We first used nonlinear regression 18. B. P. Murphy, A. N. Andersen, C. L. Parr, The underestimated biodiversity of tropical grassy to obtain predictive relationships for estimating ecosystem carbon biomes. Philos. Trans. R. Soc. B Biol. Sci. 371, 20150319 (2016). stocks and species richness as a function of current EVI (fig. S2). These 19. J. M. Cardoso da Silva, J. M. Bates, Biogeographic patterns and conservation in the South American Cerrado: A tropical savanna hotspot. Bioscience 52, 225–234 (2002). regression equations were then used to estimate carbon and richness 20. W. A. Hoffmann, E. L. Geiger, S. G. Gotsch, D. R. Rossatto, L. C. R. Silva, O. L. Lau, based on EVI values for 1986 and 2015. To avoid potential artifacts M. Haridasan, A. C. Franco, Ecological thresholds at the savanna-forest boundary: How resulting from year-to-year variation in the composite EVI values, plant traits, resources and fire govern the distribution of tropical biomes. Ecol. Lett. 15, we used linear trend lines fitted to each plot (Fig. 1) to interpolate 759–768 (2012). the EVI values for these 2 years. 21. D. V. Silvério, P. M. Brando, J. K. Balch, F. E. Putz, D. C. Nepstad, C. Oliveira-Santos, M. M. C. Bustamante, Testing the Amazon savannization hypothesis: Fire effects on invasion of a neotropical forest by native cerrado and exotic pasture grasses. Philos. Trans. R. Soc. B Biol. Sci. 368, 20120427 (2013). SUPPLEMENTARY MATERIALS 22. I. Roitman, J. M. Felfili, A. V. Rezende, Tree dynamics of a fire-protected cerrado sensu stricto Supplementary material for this article is available at http://advances.sciencemag.org/cgi/ surrounded by forest plantations, over a 13-year period (1991–2004) in Bahia, Brazil. content/full/3/8/e1701284/DC1 Plant Ecol. 197,255–267 (2008). fig. S1. Relationships between the EVI and vegetation structure. 23. B. S. Marimon, B. H. Marimon-Junior, T. R. Feldpausch, C. Oliveira-Santos, H. A. Mews, fig. S2. Relationships used to hindcast historical vegetation change. G. Lopez-Gonzalez, J. Lloyd, D. D. Franczak, E. A. de Oliveira, L. Maracahipes, A. Miguel, fig. S3. Response of plant growth forms to vegetation structure. E. Lenza, O. L. Phillips, Disequilibrium and hyperdynamic tree turnover at the forest– fig. S4. Relationships between tree basal area and characteristics of surface soils (0 to 20 cm) cerrado transition zone in southern Amazonia. Plant Ecol. Divers. 7, 281–292 (2014). in 2015. 24. P. S. Morandi, B. H. Marimon-Junior, E. A. de Oliveira, S. M. Reis, M. B. X. Valadão, table S1. Plant species recorded at SBES, Brazil, in 2015. M. Forsthofer, F. B. Passos, B. S. Marimon, Vegetation sucession in the Cerrado-Amazonian table S2. Ant species sampled at SBES, Brazil, in 2015. forest transition zone of Mato Grasso State, Brazil. Edinburgh J. Bot. 73,83–93 (2016). table S3. Details of the 30 study plots in this study. 25. T. S. Filgueiras, Herbaceous plant communities, in The Cerrados of Brazil, P. Oliveira, R. J. Marquis, Eds. (Columbia Univ. Press, 2002), pp. 121–139. 26. L. M. Coutinho, Fire in the ecology of the Brazilian Cerrado, in Fire in the Tropical Biota, J. Goldammer Ed. (Springer, 1990), pp. 82–105. REFERENCES AND NOTES 27. A. Fidelis, C. Blanco, Does fire induce flowering in Brazilian subtropical grasslands? 1. J. Phelps, E. L. Webb, W. M. Adams, Biodiversity co-benefits of policies to reduce forest- Appl. Veg. Sci. 17, 690–699 (2014). carbon emissions. Nat. Clim. Change 2, 497–503 (2012). 28. R. C. R. de Abreu, G. B. de Assis, S. Frison, A. Aguirre, G. Durigan, Can native vegetation 2. J. J. Gilroy, P. Woodcock, F. A. Edwards, C. Wheeler, B. L. G. Baptiste, C. A. M. Uribe, recover after slash pine cultivation in the Brazilian Savanna? For. Ecol. Manage. 262, T. Haugaasen, D. P. Edwards, Cheap carbon and biodiversity co-benefits from forest 1452–1459 (2011). regeneration in a hotspot of endemism. Nat. Clim. Change 4, 503–507 (2014). 29. J. P. Lemos-Filho, C. F. A. Barros, G. P. M. Dantas, L. G. Dias, R. S. Mendes, Spatial and 3. J. W. Veldman, G. E. Overbeck, D. Negreiros, G. Mahy, S. Le Stradic, G. W. Fernandes, temporal variability of canopy cover and understory light in a Cerrado of Southern Brazil. G. Durigan, E. Buisson, F. E. Putz, W. J. Bond, Where tree planting and forest expansion are Braz. J. Biol. 70,19–24 (2010). bad for biodiversity and ecosystem services. Bioscience 65, 1011–1018 (2015). 30. R. Pacheco, H. L. Vasconcelos, Habitat diversity enhances ant diversity in a naturally 4. A. C. Staver, S. Archibald, S. A. Levin, The global extent and determinants of savanna and heterogeneous Brazilian landscape. Biodivers. Conserv. 21, 797–809 (2012). forest as alternative biome states. Science 334, 230–232 (2011). 31. C. L. Parr, E. F. Gray, W. J. Bond, Cascading biodiversity and functional consequences of a 5. C. L. Parr, C. E. R. Lehmann, W. J. Bond, W. A. Hoffmann, A. N. Andersen, Tropical grassy global change-induced biome switch. Divers. Distrib. 18, 493–503 (2012). biomes: Misunderstood, neglected, and under threat. Trends Ecol. Evol. 29,205–213 (2014). 32. C. Nogueira, G. R. Colli, M. Martins, Local richness and distribution of the lizard fauna in 6. A. F. A. Pellegrini, J. B. Socolar, P. R. Elsen, X. Giam, Trade-offs between savanna natural habitat mosaics of the Brazilian Cerrado. Austral Ecol. 34,83–96 (2009). woody plant diversity and carbon storage in the Brazilian Cerrado. Glob. Chang. Biol. 22, 33. C. A. Brasileiro, R. J. Sawaya, M. C. Kiefer, M. Martins, Amphibians of an open cerrado 3373–3382 (2016). fragment in southeastern Brazil. Biota Neotrop. 5,1–17 (2005). 7. R. C. R. de Abreu, G. Durigan, Changes in the plant community of a Brazilian grassland 34. M. A. Mares, K. A. Ernest, D. D. Gettinger, Small mammal community structure and savannah after 22 years of invasion by Pinus elliottii Engelm. Plant Ecol. Divers. 4, composition in the Cerrado Province of Central Brazil. J. Trop. Ecol. 2, 289–300 (1986). 269–278 (2011). 35. A. T. de A. Jácomo, L. Silveira, J. A. F. Diniz-Filho, Niche separation between the maned 8. N. Stevens, C. E. R. Lehmann, B. P. Murphy, G. Durigan, Savanna woody encroachment is wolf (Chrysocyon brachyurus), the crab-eating fox (Dusicyon thous) and the hoary fox widespread across three continents. Glob. Chang. Biol. 23, 235–244 (2016). (Dusicyon vetulus) in central Brazil. J. Zool. 262,99–106 (2004). 9. I. Oliveras, Y. Malhi, Many shades of green: The dynamic tropical forest–savannah 36. J. M. B. Duarte, A. Vogliotti, E. dos Santos Zanetti, M. L. de Oliveira, L. M. Tiepolo, transition zones. Philos. Trans. R. Soc. B Biol. Sci. 371, 20150308 (2016). L. F. Rodrigues, L. B. de Almeida, F. G. Braga, Avaliação do Risco de Extinção 10. G. R. Moncrieff, S. Scheiter, W. J. Bond, S. I. Higgins, Increasing atmospheric CO overrides do Veado-campeiro Ozotoceros bezoarticus Linnaeus, 1758, no Brasil. Biodiversidade Bras. the historical legacy of multiple stable biome states in Africa. New Phytol. 201, 2,20–32 (2012). 908–915 (2014). 37. C. S. de Azevedo, J. B. Ferraz, H. P. Tinoco, R. J. Young, M. Rodrigues, Time–activity 11. R. Buitenwerf, W. J. Bond, N. Stevens, W. S. W. Trollope, Increased tree densities in South budget of greater rheas (Rhea americana, Aves) on a human-disturbed area: The role of African savannas: >50 years of data suggests CO as a driver. Glob. Chang. Biol. 18, habitat, time of the day, season and group size. Acta Ethol. 13, 109–117 (2010). 675–684 (2012). 38. D. M. J. S. Bowman, G. L. W. Perry, S. I. Higgins, C. N. Johnson, S. D. Fuhlendorf, 12. A. C. Franco, D. R. Rossatto, L. de Carvalho Ramos Silva, C. da Silva Ferreira, Cerrado B. P. Murphy, Pyrodiversity is the coupling of biodiversity and fire regimes in food webs. vegetation and global change: The role of functional types, resource availability and Philos. Trans. R. Soc. London B Biol. Sci. 371, 20150169 (2016). disturbance in regulating plant community responses to rising CO levels and climate 39. E. G. Jobbágy, R. B. Jackson, The vertical distribution of soil organic carbon and its relation warming. Theor. Exp. Plant Physiol. 26,19–38 (2014). to climate and vegetation. Ecol. Appl. 10, 423–436 (2000). 13. G. Durigan, J. A. Ratter, The need for a consistent fire policy for Cerrado conservation. 40. W. J. Bond, Do nutrient-poor soils inhibit development of forests? A nutrient stock J. Appl. Ecol. 53,11–15 (2016). analysis. Plant Soil. 334,47–60 (2010). 14. B. B. N. Strassburg, T. Brooks, R. Feltran-Barbieri, A. Iribarrem, R. Crouzeilles, R. Loyola, 41. L. C. R. Silva, W. A. Hoffmann, D. R. Rossatto, M. Haridasan, A. C. Franco, W. R. Horwath, A. E. Latawiec, F. J. B. Oliveira Filho, C. A. M. Scaramuzza, F. R. Scarano, B. Soares Filho, Can savannas become forests? A coupled analysis of nutrient stocks and fire A. Balmford, Moment of truth for the Cerrado hotspot. Nat. Ecol. Evol. 1,1–3 (2017). thresholds in central Brazil. Plant Soil. 373, 829–842 (2013). Abreu et al., Sci. Adv. 2017; 3 : e1701284 30 August 2017 6of7 | SCIENCE ADVANCES RESEARCH ARTICLE 42. J. Lloyd, J. Lloyd, T. F. Domingues, F. Schrodt, F. Y. Ishida, T. R. Feldpausch, G. Saiz, 60. C. Macfarlane, Y. Ryu, G. N. Ogden, O. Sonnentag, Digital canopy photography: Exposed C. A. Quesada, M. Schwarz, T. Torello-Raventos, M. Gilpin, B. Marimon, B. H. Marinon Jr., and in the raw. Agric. For. Meteorol. 197, 244–253 (2014). J. A. Ratter, J. Grace, G. B. Naradoto, E. M. Veenendaal, L. Arroyo, D. Villarroel, 61. A. Thimonier, I. Sedivy, P. Schleppi, Estimating leaf area index in different types of T. Killeen, M. Steininger, O. L. Phillips, Edaphic, structural and physiological contrasts mature forest stands in Switzerland: A comparison of methods. Eur. J. For. Res. 129, across Amazon Basin forest–savanna ecotones suggest a role for potassium as a key 543–562 (2010). modulator of tropical woody vegetation structure and function. Biogeosciences 12, 62. P. Schleppi, M. Conedera, I. Sedivy, A. Thimonier, Correcting non-linearity and slope 6529–6571 (2015). effects in the estimation of the leaf area index of forests from hemispherical photographs. 43. A. F. A. Pellegrini, Nutrient limitation in tropical savannas across multiple scales and Agric. For. Meteorol. 144,236–242 (2007). mechanisms. Ecology 97, 313–324 (2016). 63. M. Nobis, U. Hunziker, Automatic thresholding for hemispherical canopy-photographs 44. W. J. Bond, Ancient grasslands at risk. Science 351, 120–122 (2016). based on edge detection. Agric. For. Meteorol. 128, 243–250 (2005). 45. J. W. Veldman, F. E. Putz, Grass-dominated vegetation, not species-diverse natural 64. E. D. S. Pinheiro, Análises ecológicas e sensoriamento remoto aplicados à estimativa de savanna, replaces degraded tropical forests on the southern edge of the Amazon Basin. fitomassa de cerrado na Estação Ecológica de Assis, SP. Ph.D. thesis, Universidade de São Biol. Conserv. 144, 1419–1429 (2011). Paulo (2008). 46. T. A. Gardner, N. D. Burgess, N. Aguilar-Amuchastegui, J. Barlow, E. Berenguer, 65. D. A. Clark, S. Brown, D. W. Kicklighter, J. Q. Chambers, J. R. Thomlinson, J. Ni, Measuring T. Clements, F. Danielsen, J. Ferreira, W. Foden, V. Kapos, S. M. Khan, A. C. Lees, L. Parry, net primary production in forests: Concepts and field methods. Ecol. Appl. 11, 356–370 R. M. Roman-Cuesta, C. B. Schmitt, N. Strange, I. Theilade, I. C. G. Vieira, A framework (2001). for integrating biodiversity concerns into national REDD+ programmes. Biol. Conserv. 66. Y. Malhi, T. R. Baker, O. L. Phillips, S. Almeida, E. Alvarez, L. Arroyo, J. Chave, C. I. Czimczik, 154,61–71 (2012). A. Di Fiore, N. Higuchi, T. J. Killeen, S. G. Laurance, W. F. Laurance, S. L. Lewis, 47. J. Phelps, D. A. Friess, E. L. Webb, Win–win REDD+ approaches belie carbon–biodiversity L. M. M. Montoya, A. Monteagudo, D. A. Neill, P. N. Vargas, S. Patino, N. C. A. Pitman, trade-offs. Biol. Conserv. 154,53–60 (2012). C. A. Quesada, R. Salomao, J. N. M. Silva, A. T. Lezama, R. V. Martinez, J. Terborgh, B. Vinceti, 48. J. Russell-Smith, C. P. Yates, A. C. Edwards, P. J. Whitehead, B. P. Murphy, M. J. Lawes, J. Lloyd, The above-ground coarse wood productivity of 104 Neotropical forest plots. Deriving multiple benefits from carbon market-based savanna fire management: Glob. Chang. Biol. 10, 563–591 (2004). An Australian example. PLOS ONE 10, e0143426 (2015). 67. R Core Team, R: A Language and Environment for Statistical Computing (2016). 49. M. B. Ramos-Neto, V. R. Pivello, Lightning fires in a Brazilian savanna national park: 68. F. Baty, C. Ritz, S. Charles, M. Brutsche, J.-P. Flandrois, M.-L. Delignette-Muller, Rethinking management strategies. Environ. Manage. 26, 675–684 (2000). A toolbox for nonlinear regression in R : The package nlstools. J. Stat. Softw. 66, 1773 50. J. A. Ratter, S. Bridgewater, J. F. Ribeiro, Biodiversity patterns of the woody vegetation (2015). of the Brazilian Cerrado, in Neotropical Savannas and Seasonally Dry Forests: 69. H. Caswell, Matrix Population Models (Sinauer Associates, 2001), 722 pp. Plant Diversity, T. Pennington, G. P. Lewis, J. A. Ratter, Eds. (CRC Press, 2006), pp. 31–66. 51. G. Durigan, J. A. Ratter, Successional changes in cerrado and cerrado/forest ecotonal Acknowledgments: We thank M. S. Suganuma, M. G. B. Cava, G. B. de Assis, É. R. Silveira, vegetation in western São Paulo State, Brazil, 1962–2000. Edinburgh J. Bot. 63, 119–130 (2006). F. M. de Souza, J. Maravalhas, and E. Koch for assistance with data collection and the 52. C. E. R. Lehmann, T. M. Anderson, M. Sankaran, S. I. Higgins, S. Archibald, W. A. Hoffmann, employees of the Instituto Florestal for logistical assistance. This work was performed under N. P. Hanan, R. J. Williams, R. J. Fensham, J. Felfili, L. B. Hutley, J. Ratnam, J. San Jose, COTEC Technical-Scientific Commission of the Forestry Institute research license 26108- R. Montes, D. Franklin, J. Russell-Smith, C. M. Ryan, G. Durigan, P. Hiernaux, R. Haidar, 008.476/2014. Funding: This material is based on work supported by the NSF under D. M. J. S. Bowman, W. J. Bond, Savanna vegetation-fire-climate relationships differ among grant number DEB1354943 to W.A.H. H.L.V. was funded by CNPq (National Council for continents. Science 343,548–552 (2014). Scientific and Technological Development) grant #457407/2012-3, N.A.P. was funded by 53. N. van Wilgen, B. W. Govender, H. C. Biggs, The contribution of fire research to fire FAPESP (São Paulo Research Foundation) grant #2016/17888-2, D.R.R. was funded by management: A critical review of a long-term experiment in the Kruger National Park, CNPq grant #301589/2015-1, and G.D. was funded by CNPq grant #303179/2016-3. Author South Africa. Int. J. Wildl. Fire 16, 519–530 (2007). contributions: W.A.H. and G.D. designed the study. R.C.R.A., W.A.H., N.A.P., and G.D. 54. A. Andersen, G. D. Cook, R. J. Williams, Fire in Tropical Savannas: The Kapalga Experiment contributed to the vegetation fieldwork, analysis, and manuscript writing. H.L.V. performed (Spriner-Verlag, 2003). the work related to ant biodiversity and analysis and contributed to writing. W.A.H. performed 55. V. R. Pivello, The use of fire in the Cerrado and Amazonian Rainforests of Brazil: Past and the remote sensing analysis. D.R.R. contributed to manuscript writing and project logistics. present. Fire Ecol. 7,24–39 (2011). Competing interests: The authors declare that they have no competing interests. Data 56. J. A. Meira Neto, F. R. Martins, G. E. Valente, Composição florística e espectro biológico and materials availability: All data needed to evaluate the conclusions in the paper na Estação Ecológica de Santa Bárbara, Estado de São Paulo, Brasil. Rev. Árvore 31, are present in the paper and/or the Supplementary Materials. Additional data related to this 907–922 (2007). paper may be requested from the authors. 57. A. C. G. Melo, G. Durigan, Plano de Manejo da Estação Ecológica de Santa Bárbara (SEMA, 2010). 58. G. Durigan, M. F. Siqueira, G. A. D. C. Franco, W. A. Contieri, A flora arbustivo‑arbórea Submitted 21 April 2017 do Médio Paranapanema: Base para a restauração dos ecossistemas naturais, in Pesquisas Accepted 6 August 2017 em Conservação e Recuperação Ambiental no Oeste Paulista: Resultados da Cooperação Published 30 August 2017 Brasil/Japão, O. Vilas Bôas, G. Durigan, Eds. (Páginas & Letras, ed. 1, 2004), pp. 199–239. 10.1126/sciadv.1701284 59. R. C. Mendonça, J. M. Felfili, B. M. T. Walter, M. C. Silva Júnior, A. V. Rezende, T. S. Filgueiras, P. E. Nogueira, C. W. Fagg, in Flora Vascular do Bioma Cerrado Checklist com Citation: R. C. R. Abreu, W. A. Hoffmann, H. L. Vasconcelos, N. A. Pilon, D. R. Rossatto, 12.356 Espécies, S. M. Sano, S. P. Almeida, J. F. Ribeiro, Eds. (Embrapa Cerrados, 2008), G. Durigan, The biodiversity cost of carbon sequestration in tropical savanna. Sci. Adv. 3, pp. 423–1279. e1701284 (2017). Abreu et al., Sci. Adv. 2017; 3 : e1701284 30 August 2017 7of7 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Science Advances Pubmed Central

The biodiversity cost of carbon sequestration in tropical savanna

Loading next page...
 
/lp/pubmed-central/the-biodiversity-cost-of-carbon-sequestration-in-tropical-savanna-FaNLdSQv5M

References

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

Publisher
Pubmed Central
Copyright
Copyright © 2017 The Authors, some rights reserved; exclusive licensee American Association for the Advancement of Science. No claim to original U.S. Government Works. Distributed under a Creative Commons Attribution NonCommercial License 4.0 (CC BY-NC).
ISSN
2375-2548
eISSN
2375-2548
DOI
10.1126/sciadv.1701284
Publisher site
See Article on Publisher Site

Abstract

SCIENCE ADVANCES RESEARCH ARTICLE FOREST ECOSYSTEMS Copyright © 2017 The Authors, some rights reserved; The biodiversity cost of carbon sequestration in exclusive licensee American Association tropical savanna for the Advancement of Science. No claim to 1 1 2 3,4 Rodolfo C. R. Abreu, William A. Hoffmann, * Heraldo L. Vasconcelos, Natashi A. Pilon, original U.S. Government 5 3 Davi R. Rossatto, Giselda Durigan * Works. Distributed under a Creative Tropical savannas have been increasingly viewed as an opportunity for carbon sequestration through fire suppres- Commons Attribution sion and afforestation, but insufficient attention has been given to the consequences for biodiversity. To evaluate NonCommercial the biodiversity costs of increasing carbon sequestration, we quantified changes in ecosystem carbon stocks and the License 4.0 (CC BY-NC). associated changes in communities of plants and ants resulting from fire suppression in savannas of the Brazilian −1 −1 Cerrado, a global biodiversity hotspot. Fire suppression resulted in increased carbon stocks of 1.2 Mg ha year since 1986 but was associated with acute species loss. In sites fully encroached by forest, plant species richness declined by 27%, and ant richness declined by 35%. Richness of savanna specialists, the species most at risk of local extinction due to forest encroachment, declined by 67% for plants and 86% for ants. This loss highlights the impor- tant role of fire in maintaining biodiversity in tropical savannas, a role that is not reflected in current policies of fire suppression throughout the Brazilian Cerrado. In tropical grasslands and savannas throughout the tropics, carbon mitigation programs that promote forest cover cannot be assumed to provide net benefits for conservation. INTRODUCTION hotspot (17) and putatively the most species-rich savanna region in Carbon mitigation programs that protect and expand tropical forests the world (18, 19). To understand the long-term impact of current fire are widely embraced as a win-win situation for conservation because policies, we quantified changes in biodiversity and ecosystem carbon they presumably enhance species diversity (1, 2). This presumption stocks at three savanna-forest boundaries actively undergoing forest requires closer scrutiny in the seasonally dry tropics, where large areas encroachment due to fire suppression. At each site, we measured soil of natural grasslands and savannas could be converted to forest and vegetation carbon stocks and surveyed plant communities across through afforestation or fire suppression (3–5). In these biomes, it a network of 30 plots ranging from open savanna (mean tree basal 2 −1 cannot be assumed that increased tree cover will generate net benefits area of 3.1 m ha ) to recently formed forest (mean basal area of 2 −1 for biodiversity because savannascontain many endemiclight- 21.5 m ha ). Aboveground-foraging ant communities were also demanding species that can be displaced or eliminated by forest cover surveyed, serving as a representative animal group. We used a 30-year (3, 6, 7); thus, species-rich savannas may undergo a net loss of species series of Landsat satellite images to quantify rates of vegetation change when replaced by forest. For this reason, it is important to rigorously and to hindcast carbon sequestration and the resulting species loss. quantify the impacts on biodiversity of management practices that seek to maximize carbon sequestration by promoting forest cover. Where savanna management is not aimed specifically at car- RESULTS bon sequestration, other factors may lead to the same outcome of The study area has undergone a steady increase in tree cover over the increased tree density and ecosystem carbon stocks. Tree densities past 30 years. Over this time, there was a continuous increase in the have been increasing in many savannas across the tropics (8), likely enhanced vegetation index (EVI; Fig. 1), a satellite-derived metric that because of fire suppression and increasing atmospheric CO (9–12). is strongly and positively correlated with tree basal area, leaf area index In the savannas of the Brazilian Cerrado, governmental regulations (LAI), and ecosystem carbon stocks across our study sites (figs. S1 and severely restrict the use of fire, and fire suppression is actively S2). As inferred from historical EVI, all plots that are now forest were practiced in most reserves and parks (13) because of widespread per- occupied by savanna or grassland in 1986, corroborating observations ception that fire is detrimental to biodiversity. This is compounded by by reserve personnel working at the site during this period. expansion of crops, pasture, and forest plantations at alarming rates In 2015, ecosystem carbon stocks (plants + surface soils) ranged −1 −1 (14, 15), resulting in highly fragmented landscapes. Because such frag- from a mean of 20.8 Mg ha in open savanna sites to 83.5 Mg ha mentation impedes fire spread, natural ignitions alone are unable to in forest. The gradient in carbon stocks can be attributed primarily to a maintain fire frequencies that were once possible (16). 14-fold increase in tree biomass and, to a lesser extent, a 2-fold We assessed the impact of 30 years of fire suppression and forest increase in surface soil stocks (0 to 20 cm) over the vegetation gradient encroachment in savannas in the Brazilian Cerrado, a biodiversity (Fig. 2A). The annual increase in ecosystem carbon storage was −1 −1 1.19 Mg ha year over the past three decades, as estimated by combining the time series of EVI (Fig. 1) with the empirical relation- Department of Plant and Microbial Biology, North Carolina State University, Raleigh, NC 27695–7612, USA. Instituto de Biologia, Universidade Federal de Uberlândia (UFU), ship between EVI and ecosystem carbon (fig. S2). Av. Pará 1720, Uberlândia, Minas Gerais 38405-320, Brazil. Laboratório de Ecologia e The increase in tree biomass over the vegetation gradient was ac- Hidrologia Florestal, Floresta Estadual de Assis, Instituto Florestal, Assis, São Paulo companied by decreases in the richness of both plant and ant species, 19802-970, Brazil. Instituto de Biologia, Universidade Estadual de Campinas (UNICAMP), Caixa Postal 6109, Campinas, São Paulo 13083-865, Brazil. Departamento de Biologia, with particularly large losses in savanna specialists (Fig. 2). The loss of Universidade Estadual Paulista (UNESP), Campus de Jaboticabal, Jaboticabal, São savanna species was most acute when vegetation reached a stem basal Paulo 14884-900, Brazil. 2 −1 area of approximately 15 m ha and an LAI of 2.5, a threshold that *Corresponding author. Email: wahoffma@ncsu.edu (W.A.H.); giselda.durigan@ corresponds roughly to the transition between savanna and forest as gmail.com (G.D.) Abreu et al., Sci. Adv. 2017; 3 : e1701284 30 August 2017 1of7 | SCIENCE ADVANCES RESEARCH ARTICLE Sites that are currently open shrub savanna Sites that are currently tree-dominated savanna Sites that are currently forest 0.6 0.6 0.6 Forest 0.5 0.5 0.5 Dense savanna Open savanna 0.4 0.4 0.4 0.3 0.3 0.3 0.2 0.2 0.2 1985 1990 1995 2000 2005 2010 2015 1985 1990 1995 2000 2005 2010 2015 1985 1990 1995 2000 2005 2010 2015 Year Fig. 1. Historical changes in EVI over 30 years of fire suppression. Historical changes in the EVI over 30 years of fire suppression, as determined by Landsat images. Each study plot is denoted by a different symbol and a separate regression line. indicated by the disappearance of C grasses (20, 21). Forest sites had encroachment, whereas diversity of shrubs and herbaceous plants 69% fewer savanna plant species and 74% fewer savanna ant species, declined markedly (fig. S3). The decline in these groups is particularly compared to dense savanna. This decline in savanna species was par- noteworthy because shrubs and herbaceous plants comprise 75 to 80% tially offset by an increase in the number of forest specialists (Fig. 2) of plant species in the Cerrado region (25) and 58% of species at the thus, when all species were considered, forest sites had 33% fewer plant study sites (table S1). Of these nontree species, 79% are savanna spe- species and 34% fewer ant species than dense savanna. Because species cialists and are therefore particularly sensitive to forest encroachment. richness of each group was strongly correlated with EVI (r =0.65to Herbaceous communities are vastly understudied in the Cerrado, 0.86, P < 0.0001; fig. S2), we used historical changes in EVI (Fig. 1) to compared to trees, but deserve greater attention considering their high estimate changes in species diversity since 1986. In the sites that are diversity and vulnerability to fire suppression. now forest, we estimate that 27% of plant species and 35% of ant spe- In sites encroached by forest, a few savanna plant species still per- cies have been lost during 30 years of fire suppression, whereas the loss sist at low abundances, and these are nearly certain to continue declining of savanna specialists was 67 and 86% for plants and ants, respectively. if fire is not reintroduced into the system. Some species may depend The loss of plant species involved primarily shrubs and herbaceous directly on fire for reproduction (26, 27), but more generally, plant species, whereas the number of tree species increased (fig. S3). species adapted to savanna environments cannot tolerate the low light These concomitant changes in diversity and ecosystem properties availability under a forest canopy (5, 28, 29). These species mostly resulted in a strong trade-off between ecosystem carbon stocks and persist in forest as occasional moribund individuals with few leaves. species richness across the savanna-forest gradients (Fig. 3). Over A few robust savanna trees remain in the forests, but these are rare the ranges of tree densities studied, each 1% increase in ecosystem car- individuals that are sufficiently tall to access ample light. Even these bon resulted in a 0.27% decrease in total plant species richness and a species are unlikely to persist beyond the life span of these individuals 0.43% decrease in ant species richness (Fig. 3), as determined from the because their seedlings are nearly absent from sites encroached by forest. sensitivity coefficient. When only savanna species were considered, In short, we expect continued loss of savanna specialist species in forest- each 1% increase in ecosystem carbon corresponds to decreases of encroached sites, and the results presented here should be considered 0.98 and 1.38% for plants and ants, respectively (Fig. 3). Together, forest an underestimate of the eventual impact of fire suppression. and generalist plant species increased by 0.76% for each 1% increase in Similar responses to forest encroachment are expected for animal carbon, but there was no detectable change for ants. communities, represented here by ants because they are a diverse Stem basal area in 2015 was significantly correlated with properties group that lend themselves to systematic sampling. As for plants, of surface soils (0 to 20 cm). Sites with higher basal area had greater ant communities are strongly dependent on vegetation structure silt, clay, and organic matter; higher cation exchange capacity (CEC); (30, 31); consequently, forest encroachment results in a notable loss higher availabilities of P, K, B, and S (fig. S4); and higher concentra- of ant species adapted to open vegetation (Fig. 2C). These savanna- tions of P, K, B, and S (fig. S4). Higher basal area was associated with dependent species have evolved within most or all major groups of lower pH, sand, and Cu (fig. S4). animals of the Cerrado region, including reptiles (32), amphibians (33), birds (19), and mammals (34). Among these are species of par- ticular conservation concern, such as the maned wolf (35), Pampas DISCUSSION deer (36), and the greater rhea (37). The interrelationships between Previous studies in the Cerrado have suggested that forest encroach- vegetation, fauna, and fire are complex (38), but generally, animal spe- ment into open savanna results in a net increase in plant species rich- cies specialized to savanna and grassland should be considered vulner- ness (6, 22–24), but these studies have been limited to trees and able to fire suppression and forest encroachment. therefore overlook the high diversity of shrubs and herbaceous plants Forest encroachment at this site was made possible by long- in savannas. Here, diversity of tree species increased with forest term fire suppression, although increasing atmospheric CO may Abreu et al., Sci. Adv. 2017; 3 : e1701284 30 August 2017 2of7 Enhanced vegetation index | SCIENCE ADVANCES RESEARCH ARTICLE A Plants All species r = 0.97 Savanna species Trees 120 Soil (0–20 cm) Herbaceous Total 2 r = 0.53 r = 0.98 r = 0.76 r = 0.74 r = 0.66 20 Forest and generalist species Savanna species 0 All species 2 B Ants r = 0.64 50 2 r = 0.69 r = 0.79 r = 0.71 2 20 r = 0.87 Forest and generalist species Savanna species All species r = 0.67 0 20406080 100 −1 Ecosystem carbon stocks (Mg ha ) 2 Fig. 3. Carbon-biodiversity trade-offs across savanna-forest transitions. Carbon- r = 0.03 biodiversity trade-offs across savanna-forest transitions. Relationships between ecosystem carbon stocks and species richness of (A) plants and (B) ants. All relation- r = 0.84 ships are statistically significant (P < 0.0001). Fitted relationships: all plants, y = −0.271 −0.975 −0.426 245.88x ; savanna plants, y = 1953.4x ; all ants, y = 198.68x ; savanna 10 −1.380 ants, y = 2258.8x . and only one was burned more than once in these three decades. 0 5 10 15 20 25 Fire is not likely to be the only factor responsible for this variation 2 −1 Stem basal area (m ha ) because encroachment was greatest in soils that have greater water Fig. 2. Carbon stocks and species richness due to forest encroachment. Changes retention capacity (high silt and clay and low sand; fig. S4) and in (A) carbon stocks, (B) plant species richness, and (C) ant species richness over higher concentrations of P, K, B, and S (fig. S4). These soil variables gradients of forest encroachment. All variables were significantly correlated with basal were measured in the top 20 cm of the soil profile, where vegetation area (P < 0.0001) except for the number of forest and generalist ants (P =0.42). can have large influences on soil chemistry (39); thus, causality is difficult to ascertain. Nevertheless, the gradient in sand content re- veals preexisting soil differences that likely influenced rates of forest have contributed by accelerating tree establishment and growth (9–12). encroachment under fire suppression. Regardless of the role that However, encroachment was not uniform across the study area, with soils and nutrients may play in controlling the rate of forest expan- substantial variation in both initial density and subsequent rate of sion (40–43), tree encroachment has continued steadily in all savanna increase (Fig. 1). Previous fire history likely contributed to the initial sites (Fig. 1). differences, because fires were common before the 1980s and would If the current trajectory is not interrupted, we expect most or not have burned uniformly across the area. However, during the all high-quality savanna to be lost from the study area within a decade 30-year study period, only four of the most open plots had burned, or two, with obvious negative consequences for biodiversity. Abreu et al., Sci. Adv. 2017; 3 : e1701284 30 August 2017 3of7 −1 Number of ant species Number of plant species Carbon stocks (Mg ha ) Number of species | SCIENCE ADVANCES RESEARCH ARTICLE Reintroduction of fire would reverse the process of tree encroach- conserving savanna landscapes and biota, burning should occur with ment, but to avoid catastrophic loss of biodiversity, it is important sufficient frequency and intensity to prevent tree densities from to restore a natural fire regime well before the canopy becomes approaching the point at which light-demanding species fail to thrive 2 −1 dense enough to substantially impact the herbaceous layer. Once (that is, basal area of approximately 15 m ha ). This should be the diversity of the herbaceous community has been substantially considered the absolute maximum acceptable tree density for avoiding depleted, it is unlikely to recover readily, even after the vegetation is catastrophic loss of savanna species, and even this density may exceed opened. Restoring “old-growth” grasslands such as these is notoriously the point at which prescribed burning can safely and effectively reverse difficult (44), particularly in the Cerrado, where receding forest is tree encroachment. Ideally, consideration should be given to main- typically replaced by aggressive exotic grasses (21, 45) that preclude taining the mosaic of grassland (campo limpo), shrub savanna (campo the recovery of native herbaceous species (45). sujo), and tree savanna (cerrado sensu stricto) that may have histori- cally occurred at a site. Ultimately, fire management in the Cerrado Implications for policy and management should evolve to consider the ecological role of fire, such as stimulating These findings challenge the widespread assumption that promoting flowering in herbaceous plants and influencing exotic species, but this forest cover has universal co-benefits for carbon sequestration and requires observations that can accrue only if land managers and biodiversity (1, 2). Forest encroachment resulted in increased eco- scientists are allowed to use prescribed fire as a management tool. system carbon at the expense of reduced species diversity, an outcome Until this information is available for the Cerrado, managers can use that is likely repeated wherever forest encroaches on the Cerrado and knowledge from long-term fire experiments in Africa and Australia other species-rich savannas and grasslands in the world. Under these (53, 54), as well as traditional knowledge on fire management in conditions, a carbon-centered conservation strategy may be inappropriate the Cerrado (55). Meanwhile, it is urgent to reformulate the current (46, 47), except in excessively burned savannas, where modest reduc- timid policies that favor fire suppression over all other options. tions in fire frequency or intensity benefit both biodiversity and carbon storage (48). The results presented here are particularly alarming in light of the MATERIALS AND METHODS current policy of fire suppression throughout the Cerrado region (13). Experimental design Policies of fire suppression in Brazil are exacerbated by ongoing land- Study area use change and vegetation fragmentation, which restrict fire spread The Santa Barbara Ecological Station (SBES) is located in the munic- and thereby reduce fire frequency (16). In large, unbroken expanses ipality of Águas de Santa Bárbara, São Paulo State, Brazil, with an area of the Cerrado, lightning fires can sustain frequent burning (49), but of 2715 ha (56). It lies within 22°46′33″ and 22°50′33″ Sand 49°10′27″ this becomes increasingly improbable as vegetation is fragmented and 49°15′36″ W, near the southeastern edge of the Cerrado region, at (16). Where lightning fires are not replaced with sufficient human- altitudes between 600 and 680 m above sea level. The climate in the initiated fires, biodiversity losses on protected lands, as observed here, area is Köppen Cwa-type (56), with monthly mean temperatures will be increasingly common throughout the Cerrado. Even greater varying between 16° and 24°C. The annual mean rainfall is between losses should be expected in the core of the Cerrado region, which 1100 and 1300 mm, with dry winters and rainy summers. The soils of supports greater diversity of savanna species than the study area the study area are deep oxisols with high sand content, low nutrient (50), which lies near the southeastern limit of the Cerrado. To date, content, and high saturation of aluminum (57). the consequences of excluding fire are much more evident in the study Data collection area (51) because of its longer history of habitat fragmentation and The study was undertaken at three savanna-forest transitions located protection from fire. This history offers important lessons that are at distances of 2 to 5 km from each other. A network of 30 plots was relevant throughout the Cerrado andother tropical savannaregions. distributed across these transitions to include 12 in recently formed forest, Environmental legislation in Brazil advanced a step in the right 12 in tree-dominated savanna, and 6 in shrub-dominated savanna. direction by legalizing prescribed fires for conservation of protected These vegetation classes correspond to cerradão, cerrado sensu stricto, areas in the Cerrado (Law number 12651, Article 38, in effect since and campo sujo of the Brazilian classification system (25). Each plot May 2012), but to date, controlled fires have been used only in a few was 20 m × 50 m (0.1 ha), within which the stem diameter of all trees reserves whose management plans specifically permit prescribed greater than 5 cm were measured. All trees and shrubs with stem dia- burning. However, burning cerrado vegetation outside formal re- meters between 1 and 5 cm were measured within 10 5-m × 5-m sub- serves remains illegal (13). This situation could change if the expected plots distributed regularly throughout each plot. Tree seedlings and National Fire Policy mandated by law were implemented (Law num- the herbaceous community were sampled within 40 1-m × 1-m sub- ber 12651, Article 40), regulating the use of fire in natural vegetation plots distributed in a grid within each 0.1-ha plot. All individuals throughout the Cerrado ecoregion. Nevertheless, the overall percep- sampled within these plots were identified to species in the field or tion that fire is destructive can be an obstacle even more difficult to collected for subsequent identification. surpass because managers of protected areas hesitate to prescribe fire Plant species were categorized into eight growth forms: (i) grass; in fear of formal punishment or social condemnation (13). (ii) sedge; (iii) forb, that is, nongraminoid plant without a Negative attitude toward fire is difficult to overcome; however, our persistent woody stem; (iv) subshrub, that is, woody plant typically results offer important direct evidence of biodiversity loss due to less than 75 cm tall as adults; (v) shrub; (vi) palm; (vii) treelet; and woody encroachment when fire is eliminated in the Cerrado. The ex- (viii) tree. On the basis of literature (58, 59), species were also pansion of forests and trees in savanna ecosystems may be a global classified according to habitat preference into three groups: (i) savanna phenomenon, attributed to a complex of different factors in each specialists, (ii) forest specialists, and (iii) generalists, that is, species that region (8, 52), but there is no question that fire is an effective tool to are commonly found in both savanna and forest habitats (table S1). prevent it. Our results indicate that, on any Cerrado land managed for Some species of the latter group may be true generalists that persist Abreu et al., Sci. Adv. 2017; 3 : e1701284 30 August 2017 4of7 | SCIENCE ADVANCES RESEARCH ARTICLE indefinitely in either environment, but most are best considered as at intervals of 10 m along its longest axis. The soil samples were transitional species typically found in forest but readily colonize submitted to a commercial laboratory to quantify coarse sand, fine savanna under fire suppression. sand, silt, and clay fractions; extractable Ca, Mg, K, B, Cu, Mn, Fe, The aboveground-foraging ant community was characterized for and Zn; available P; exchangeable Al; pH and organic matter; and each of the 0.1-ha plots placed in forest or tree-dominated savanna; the derived variables: potential acidity (H + Al), CEC, exchangeable the six plots in shrub-dominated savanna were omitted. Five 2.5-m × acidity, and base saturation and aluminum saturation. A subset of 2.5-m grids were established along the borders of each plot, keeping a 16 plots was sampled to determine carbon stocks in litter, duff, and minimum distance of 20 m between any two grids. Four pitfall traps herbaceous plant pools. These were sampled in four or five 0.25-m were set in each grid, totaling 20 traps per plot. Pitfall traps consisted subplots adjacent to each plot. In each subplot, the ground layer of a small plastic cup (250 ml, 8.5 cm high, and 7.8 cm in diameter) biomass was collected and divided into grass, sedge, and dicot partially filled with water and detergent. The traps remained in fractions. Litter was divided into leaf and wood fractions. When a dis- operation for 48 hours. All ant specimens collected were sorted to cernable organic horizon (duff) was present, this was also collected in morphospecies and, whenever possible, identified to species using a 0.0625-m portion of the subplot. All these components were available taxonomic keys or through comparison with specimens pre- weighed after drying for at least 5 days at 60°C. Each dried duff sam- viously identified by experts and deposited at the Zoological Collec- ple was thoroughly mixed, and a subsample was weighed before tion of the Federal University of Uberlândia, Brazil. ashing in a muffle furnace to determine the fraction of organic matter. Each ant species or morphospecies was classified as savanna spe- The carbon content of the lost mass was calculated using a conversion −1 cialist, forest specialist, or habitat generalist (table S2). Because the factor of 0.58 kg C kg organic matter. The combined carbon content habitat preference of some ant species could not be determined from of litter, duff, and herbaceous plants was summed after scaling each to −1 existing literature sources, these species were classified on the basis of similar units (Mg ha ). Because this combined pool accounted for a abundance at the site. A species was considered a specialist if it oc- small fraction of the ecosystem carbon pools (mean, 13%; range, 9 to curred exclusively, or nearly so, in either savanna or forest and was 21%) and was strongly correlated to tree basal area (r = 0.70), we used considered a generalist if it occurred at similar frequencies in both regression to estimate the size of this pool for the 14 plots for which this habitats. To eliminate the possibility that this post hoc classification was not measured directly. might introduce spurious relationships when testing for vegetation Total tree biomass was estimated for each plot using allometric impact on species richness within the functional types, we also applied equations developed for trees in a range of Cerrado physiognomies an alternate approach to classifying species using subsets of the data. and including both root and shoot biomass (64) as follows For this, abundance data from two of the savanna-forest boundaries were used to classify species, and then this classification scheme was Savanna plots : lnðMÞ¼  1:6516 þ 0:7643* lnðd hÞð1Þ used to determine the number of savanna, forest, and generalist spe- cies at the third boundary. By using successive subsets of the data in this way, the test for vegetation response was independent of the data Forest plotsðcerrada˜oÞ : lnðMÞ¼  2:8573 þ 0:9556*lnðd hÞð2Þ used to classify the species. Fewer species could be classified as habitat specialists in this manner, but in relative terms, results were stronger than when analyzed in the previous manner. Specifically, forest en- where M is the predicted dry biomass (in kilograms), d is the stem croachment was found to result in an 87% decline in the mean number diameter (in centimeters), and h is the height (in meters). Mass of −1 of savanna ant species per plot (3.0 in forest versus 22.3 in savanna) carbon was calculated using a conversion factor of 0.5 kg C kg but a 3.5-fold increase in the number of forest species (12.3 versus 3.7). biomass (65, 66). Consequently, we used the classification scheme based on the full data Remote sensing set for all subsequent analyses because this allowed a greater number To quantify vegetation change over 30 years of fire suppression, we of species to be classified. used vegetation indices derived from Landsat satellite images and ob- In each of the 30 vegetation plots, the LAI of the overstory was tained from the USGS (United States Geological Survey) portal (https:// espa.cr.usgs.gov/). We first evaluated the performance of four vegeta- measured with hemispherical canopy photographs taken at each of tion indices, normalized difference vegetation index (NDVI), enhanced the 40 subplots used to characterize the ground layer community. To ensure conditions of diffuse light, all photos were taken before vegetation index (EVI), soil-adjusted vegetation index (SAVI), and sunrise, after sunset, or under homogeneous overcast skies. A tripod modified soil-adjusted vegetation index (MSAVI) calculated from sur- was used to position the camera (Canon EF 8-15mm fisheye lens) at a face reflectances derived from Landsat 8. Of these four indices, the EVI height of 1 m, and the top of the camera was oriented relative to the was most strongly correlated with tree basal area (r =0.75) andLAI north. Photos were taken with an underexposure of one f stop (60), (r = 0.89; fig. S2); thus, all further analyses were undertaken with this andthe colorimageswereconverted to blackand whiteusing index. To determine long-term vegetation trends, we used all available Hemisfer 2.12 (61, 62) and using maximum blue contrast (63). The images from 1986 to 2016 that were acquired by Landsat 5–8under images were then analyzed with Hemisfer 2.12 using an automatic cloudless conditions and during the wet season months of November threshold for closed-canopy vegetation and with a supervised manual through May. For each image, EVI was spatially averaged for each study threshold under open canopies. The LAI values were averaged over plot. This resulted in multiple values for each calendar year, which were the 40 subplots to obtain a single value for each 0.1-ha plot. then composited by determining the yearly maximum. Ecosystem carbon and soil properties Statistical analysis Samples of surface soil (0- to 20-cm depth) were collected for each Analyses were performed with the base stats package in R (67). The of the 30 plots. A composite soil sample was made with five sub- lm function was used to fit linear and quadratic models, and the samples from surface soil collected immediately adjacent to the plot package nlstools (68) was used for fitting nonlinear models. Abreu et al., Sci. Adv. 2017; 3 : e1701284 30 August 2017 5of7 | SCIENCE ADVANCES RESEARCH ARTICLE 15. R. Beuchle, R. C. Grecchi, Y. E. Shimabukuro, R. Seliger, H. D. Eva, E. Sano, F. Achard, To characterize the relative change in species richness (r) for a Land cover changes in the Brazilian Cerrado and Caatinga biomes from 1990 to relative change in ecosystem carbon (C), we calculated the coefficient 2010 based on a systematic remote sensing sampling approach. Appl. Geogr. dr C of sensitivity, ; also known as elasticity in economics and demog- dC r 58,116–127 (2015). raphy (69). This was calculated as the slope of the relationship between 16. S. Archibald, A. C. Staver, S. A. Levin, Evolution of human-driven fire regimes in Africa. Proc. Natl. Acad. Sci. U.S.A. 109, 847–852 (2012). ln(r) and ln(C). 17. N. Myers, R. A. Mittermeier, C. G. Mittermeier, G. A. B. da Fonseca, J. Kent, Biodiversity We estimated past changes in carbon stocks and species number by hotspots for conservation priorities. Nature 403, 853–858 (2000). using a space-for-time substitution. We first used nonlinear regression 18. B. P. Murphy, A. N. Andersen, C. L. Parr, The underestimated biodiversity of tropical grassy to obtain predictive relationships for estimating ecosystem carbon biomes. Philos. Trans. R. Soc. B Biol. Sci. 371, 20150319 (2016). stocks and species richness as a function of current EVI (fig. S2). These 19. J. M. Cardoso da Silva, J. M. Bates, Biogeographic patterns and conservation in the South American Cerrado: A tropical savanna hotspot. Bioscience 52, 225–234 (2002). regression equations were then used to estimate carbon and richness 20. W. A. Hoffmann, E. L. Geiger, S. G. Gotsch, D. R. Rossatto, L. C. R. Silva, O. L. Lau, based on EVI values for 1986 and 2015. To avoid potential artifacts M. Haridasan, A. C. Franco, Ecological thresholds at the savanna-forest boundary: How resulting from year-to-year variation in the composite EVI values, plant traits, resources and fire govern the distribution of tropical biomes. Ecol. Lett. 15, we used linear trend lines fitted to each plot (Fig. 1) to interpolate 759–768 (2012). the EVI values for these 2 years. 21. D. V. Silvério, P. M. Brando, J. K. Balch, F. E. Putz, D. C. Nepstad, C. Oliveira-Santos, M. M. C. Bustamante, Testing the Amazon savannization hypothesis: Fire effects on invasion of a neotropical forest by native cerrado and exotic pasture grasses. Philos. Trans. R. Soc. B Biol. Sci. 368, 20120427 (2013). SUPPLEMENTARY MATERIALS 22. I. Roitman, J. M. Felfili, A. V. Rezende, Tree dynamics of a fire-protected cerrado sensu stricto Supplementary material for this article is available at http://advances.sciencemag.org/cgi/ surrounded by forest plantations, over a 13-year period (1991–2004) in Bahia, Brazil. content/full/3/8/e1701284/DC1 Plant Ecol. 197,255–267 (2008). fig. S1. Relationships between the EVI and vegetation structure. 23. B. S. Marimon, B. H. Marimon-Junior, T. R. Feldpausch, C. Oliveira-Santos, H. A. Mews, fig. S2. Relationships used to hindcast historical vegetation change. G. Lopez-Gonzalez, J. Lloyd, D. D. Franczak, E. A. de Oliveira, L. Maracahipes, A. Miguel, fig. S3. Response of plant growth forms to vegetation structure. E. Lenza, O. L. Phillips, Disequilibrium and hyperdynamic tree turnover at the forest– fig. S4. Relationships between tree basal area and characteristics of surface soils (0 to 20 cm) cerrado transition zone in southern Amazonia. Plant Ecol. Divers. 7, 281–292 (2014). in 2015. 24. P. S. Morandi, B. H. Marimon-Junior, E. A. de Oliveira, S. M. Reis, M. B. X. Valadão, table S1. Plant species recorded at SBES, Brazil, in 2015. M. Forsthofer, F. B. Passos, B. S. Marimon, Vegetation sucession in the Cerrado-Amazonian table S2. Ant species sampled at SBES, Brazil, in 2015. forest transition zone of Mato Grasso State, Brazil. Edinburgh J. Bot. 73,83–93 (2016). table S3. Details of the 30 study plots in this study. 25. T. S. Filgueiras, Herbaceous plant communities, in The Cerrados of Brazil, P. Oliveira, R. J. Marquis, Eds. (Columbia Univ. Press, 2002), pp. 121–139. 26. L. M. Coutinho, Fire in the ecology of the Brazilian Cerrado, in Fire in the Tropical Biota, J. Goldammer Ed. (Springer, 1990), pp. 82–105. REFERENCES AND NOTES 27. A. Fidelis, C. Blanco, Does fire induce flowering in Brazilian subtropical grasslands? 1. J. Phelps, E. L. Webb, W. M. Adams, Biodiversity co-benefits of policies to reduce forest- Appl. Veg. Sci. 17, 690–699 (2014). carbon emissions. Nat. Clim. Change 2, 497–503 (2012). 28. R. C. R. de Abreu, G. B. de Assis, S. Frison, A. Aguirre, G. Durigan, Can native vegetation 2. J. J. Gilroy, P. Woodcock, F. A. Edwards, C. Wheeler, B. L. G. Baptiste, C. A. M. Uribe, recover after slash pine cultivation in the Brazilian Savanna? For. Ecol. Manage. 262, T. Haugaasen, D. P. Edwards, Cheap carbon and biodiversity co-benefits from forest 1452–1459 (2011). regeneration in a hotspot of endemism. Nat. Clim. Change 4, 503–507 (2014). 29. J. P. Lemos-Filho, C. F. A. Barros, G. P. M. Dantas, L. G. Dias, R. S. Mendes, Spatial and 3. J. W. Veldman, G. E. Overbeck, D. Negreiros, G. Mahy, S. Le Stradic, G. W. Fernandes, temporal variability of canopy cover and understory light in a Cerrado of Southern Brazil. G. Durigan, E. Buisson, F. E. Putz, W. J. Bond, Where tree planting and forest expansion are Braz. J. Biol. 70,19–24 (2010). bad for biodiversity and ecosystem services. Bioscience 65, 1011–1018 (2015). 30. R. Pacheco, H. L. Vasconcelos, Habitat diversity enhances ant diversity in a naturally 4. A. C. Staver, S. Archibald, S. A. Levin, The global extent and determinants of savanna and heterogeneous Brazilian landscape. Biodivers. Conserv. 21, 797–809 (2012). forest as alternative biome states. Science 334, 230–232 (2011). 31. C. L. Parr, E. F. Gray, W. J. Bond, Cascading biodiversity and functional consequences of a 5. C. L. Parr, C. E. R. Lehmann, W. J. Bond, W. A. Hoffmann, A. N. Andersen, Tropical grassy global change-induced biome switch. Divers. Distrib. 18, 493–503 (2012). biomes: Misunderstood, neglected, and under threat. Trends Ecol. Evol. 29,205–213 (2014). 32. C. Nogueira, G. R. Colli, M. Martins, Local richness and distribution of the lizard fauna in 6. A. F. A. Pellegrini, J. B. Socolar, P. R. Elsen, X. Giam, Trade-offs between savanna natural habitat mosaics of the Brazilian Cerrado. Austral Ecol. 34,83–96 (2009). woody plant diversity and carbon storage in the Brazilian Cerrado. Glob. Chang. Biol. 22, 33. C. A. Brasileiro, R. J. Sawaya, M. C. Kiefer, M. Martins, Amphibians of an open cerrado 3373–3382 (2016). fragment in southeastern Brazil. Biota Neotrop. 5,1–17 (2005). 7. R. C. R. de Abreu, G. Durigan, Changes in the plant community of a Brazilian grassland 34. M. A. Mares, K. A. Ernest, D. D. Gettinger, Small mammal community structure and savannah after 22 years of invasion by Pinus elliottii Engelm. Plant Ecol. Divers. 4, composition in the Cerrado Province of Central Brazil. J. Trop. Ecol. 2, 289–300 (1986). 269–278 (2011). 35. A. T. de A. Jácomo, L. Silveira, J. A. F. Diniz-Filho, Niche separation between the maned 8. N. Stevens, C. E. R. Lehmann, B. P. Murphy, G. Durigan, Savanna woody encroachment is wolf (Chrysocyon brachyurus), the crab-eating fox (Dusicyon thous) and the hoary fox widespread across three continents. Glob. Chang. Biol. 23, 235–244 (2016). (Dusicyon vetulus) in central Brazil. J. Zool. 262,99–106 (2004). 9. I. Oliveras, Y. Malhi, Many shades of green: The dynamic tropical forest–savannah 36. J. M. B. Duarte, A. Vogliotti, E. dos Santos Zanetti, M. L. de Oliveira, L. M. Tiepolo, transition zones. Philos. Trans. R. Soc. B Biol. Sci. 371, 20150308 (2016). L. F. Rodrigues, L. B. de Almeida, F. G. Braga, Avaliação do Risco de Extinção 10. G. R. Moncrieff, S. Scheiter, W. J. Bond, S. I. Higgins, Increasing atmospheric CO overrides do Veado-campeiro Ozotoceros bezoarticus Linnaeus, 1758, no Brasil. Biodiversidade Bras. the historical legacy of multiple stable biome states in Africa. New Phytol. 201, 2,20–32 (2012). 908–915 (2014). 37. C. S. de Azevedo, J. B. Ferraz, H. P. Tinoco, R. J. Young, M. Rodrigues, Time–activity 11. R. Buitenwerf, W. J. Bond, N. Stevens, W. S. W. Trollope, Increased tree densities in South budget of greater rheas (Rhea americana, Aves) on a human-disturbed area: The role of African savannas: >50 years of data suggests CO as a driver. Glob. Chang. Biol. 18, habitat, time of the day, season and group size. Acta Ethol. 13, 109–117 (2010). 675–684 (2012). 38. D. M. J. S. Bowman, G. L. W. Perry, S. I. Higgins, C. N. Johnson, S. D. Fuhlendorf, 12. A. C. Franco, D. R. Rossatto, L. de Carvalho Ramos Silva, C. da Silva Ferreira, Cerrado B. P. Murphy, Pyrodiversity is the coupling of biodiversity and fire regimes in food webs. vegetation and global change: The role of functional types, resource availability and Philos. Trans. R. Soc. London B Biol. Sci. 371, 20150169 (2016). disturbance in regulating plant community responses to rising CO levels and climate 39. E. G. Jobbágy, R. B. Jackson, The vertical distribution of soil organic carbon and its relation warming. Theor. Exp. Plant Physiol. 26,19–38 (2014). to climate and vegetation. Ecol. Appl. 10, 423–436 (2000). 13. G. Durigan, J. A. Ratter, The need for a consistent fire policy for Cerrado conservation. 40. W. J. Bond, Do nutrient-poor soils inhibit development of forests? A nutrient stock J. Appl. Ecol. 53,11–15 (2016). analysis. Plant Soil. 334,47–60 (2010). 14. B. B. N. Strassburg, T. Brooks, R. Feltran-Barbieri, A. Iribarrem, R. Crouzeilles, R. Loyola, 41. L. C. R. Silva, W. A. Hoffmann, D. R. Rossatto, M. Haridasan, A. C. Franco, W. R. Horwath, A. E. Latawiec, F. J. B. Oliveira Filho, C. A. M. Scaramuzza, F. R. Scarano, B. Soares Filho, Can savannas become forests? A coupled analysis of nutrient stocks and fire A. Balmford, Moment of truth for the Cerrado hotspot. Nat. Ecol. Evol. 1,1–3 (2017). thresholds in central Brazil. Plant Soil. 373, 829–842 (2013). Abreu et al., Sci. Adv. 2017; 3 : e1701284 30 August 2017 6of7 | SCIENCE ADVANCES RESEARCH ARTICLE 42. J. Lloyd, J. Lloyd, T. F. Domingues, F. Schrodt, F. Y. Ishida, T. R. Feldpausch, G. Saiz, 60. C. Macfarlane, Y. Ryu, G. N. Ogden, O. Sonnentag, Digital canopy photography: Exposed C. A. Quesada, M. Schwarz, T. Torello-Raventos, M. Gilpin, B. Marimon, B. H. Marinon Jr., and in the raw. Agric. For. Meteorol. 197, 244–253 (2014). J. A. Ratter, J. Grace, G. B. Naradoto, E. M. Veenendaal, L. Arroyo, D. Villarroel, 61. A. Thimonier, I. Sedivy, P. Schleppi, Estimating leaf area index in different types of T. Killeen, M. Steininger, O. L. Phillips, Edaphic, structural and physiological contrasts mature forest stands in Switzerland: A comparison of methods. Eur. J. For. Res. 129, across Amazon Basin forest–savanna ecotones suggest a role for potassium as a key 543–562 (2010). modulator of tropical woody vegetation structure and function. Biogeosciences 12, 62. P. Schleppi, M. Conedera, I. Sedivy, A. Thimonier, Correcting non-linearity and slope 6529–6571 (2015). effects in the estimation of the leaf area index of forests from hemispherical photographs. 43. A. F. A. Pellegrini, Nutrient limitation in tropical savannas across multiple scales and Agric. For. Meteorol. 144,236–242 (2007). mechanisms. Ecology 97, 313–324 (2016). 63. M. Nobis, U. Hunziker, Automatic thresholding for hemispherical canopy-photographs 44. W. J. Bond, Ancient grasslands at risk. Science 351, 120–122 (2016). based on edge detection. Agric. For. Meteorol. 128, 243–250 (2005). 45. J. W. Veldman, F. E. Putz, Grass-dominated vegetation, not species-diverse natural 64. E. D. S. Pinheiro, Análises ecológicas e sensoriamento remoto aplicados à estimativa de savanna, replaces degraded tropical forests on the southern edge of the Amazon Basin. fitomassa de cerrado na Estação Ecológica de Assis, SP. Ph.D. thesis, Universidade de São Biol. Conserv. 144, 1419–1429 (2011). Paulo (2008). 46. T. A. Gardner, N. D. Burgess, N. Aguilar-Amuchastegui, J. Barlow, E. Berenguer, 65. D. A. Clark, S. Brown, D. W. Kicklighter, J. Q. Chambers, J. R. Thomlinson, J. Ni, Measuring T. Clements, F. Danielsen, J. Ferreira, W. Foden, V. Kapos, S. M. Khan, A. C. Lees, L. Parry, net primary production in forests: Concepts and field methods. Ecol. Appl. 11, 356–370 R. M. Roman-Cuesta, C. B. Schmitt, N. Strange, I. Theilade, I. C. G. Vieira, A framework (2001). for integrating biodiversity concerns into national REDD+ programmes. Biol. Conserv. 66. Y. Malhi, T. R. Baker, O. L. Phillips, S. Almeida, E. Alvarez, L. Arroyo, J. Chave, C. I. Czimczik, 154,61–71 (2012). A. Di Fiore, N. Higuchi, T. J. Killeen, S. G. Laurance, W. F. Laurance, S. L. Lewis, 47. J. Phelps, D. A. Friess, E. L. Webb, Win–win REDD+ approaches belie carbon–biodiversity L. M. M. Montoya, A. Monteagudo, D. A. Neill, P. N. Vargas, S. Patino, N. C. A. Pitman, trade-offs. Biol. Conserv. 154,53–60 (2012). C. A. Quesada, R. Salomao, J. N. M. Silva, A. T. Lezama, R. V. Martinez, J. Terborgh, B. Vinceti, 48. J. Russell-Smith, C. P. Yates, A. C. Edwards, P. J. Whitehead, B. P. Murphy, M. J. Lawes, J. Lloyd, The above-ground coarse wood productivity of 104 Neotropical forest plots. Deriving multiple benefits from carbon market-based savanna fire management: Glob. Chang. Biol. 10, 563–591 (2004). An Australian example. PLOS ONE 10, e0143426 (2015). 67. R Core Team, R: A Language and Environment for Statistical Computing (2016). 49. M. B. Ramos-Neto, V. R. Pivello, Lightning fires in a Brazilian savanna national park: 68. F. Baty, C. Ritz, S. Charles, M. Brutsche, J.-P. Flandrois, M.-L. Delignette-Muller, Rethinking management strategies. Environ. Manage. 26, 675–684 (2000). A toolbox for nonlinear regression in R : The package nlstools. J. Stat. Softw. 66, 1773 50. J. A. Ratter, S. Bridgewater, J. F. Ribeiro, Biodiversity patterns of the woody vegetation (2015). of the Brazilian Cerrado, in Neotropical Savannas and Seasonally Dry Forests: 69. H. Caswell, Matrix Population Models (Sinauer Associates, 2001), 722 pp. Plant Diversity, T. Pennington, G. P. Lewis, J. A. Ratter, Eds. (CRC Press, 2006), pp. 31–66. 51. G. Durigan, J. A. Ratter, Successional changes in cerrado and cerrado/forest ecotonal Acknowledgments: We thank M. S. Suganuma, M. G. B. Cava, G. B. de Assis, É. R. Silveira, vegetation in western São Paulo State, Brazil, 1962–2000. Edinburgh J. Bot. 63, 119–130 (2006). F. M. de Souza, J. Maravalhas, and E. Koch for assistance with data collection and the 52. C. E. R. Lehmann, T. M. Anderson, M. Sankaran, S. I. Higgins, S. Archibald, W. A. Hoffmann, employees of the Instituto Florestal for logistical assistance. This work was performed under N. P. Hanan, R. J. Williams, R. J. Fensham, J. Felfili, L. B. Hutley, J. Ratnam, J. San Jose, COTEC Technical-Scientific Commission of the Forestry Institute research license 26108- R. Montes, D. Franklin, J. Russell-Smith, C. M. Ryan, G. Durigan, P. Hiernaux, R. Haidar, 008.476/2014. Funding: This material is based on work supported by the NSF under D. M. J. S. Bowman, W. J. Bond, Savanna vegetation-fire-climate relationships differ among grant number DEB1354943 to W.A.H. H.L.V. was funded by CNPq (National Council for continents. Science 343,548–552 (2014). Scientific and Technological Development) grant #457407/2012-3, N.A.P. was funded by 53. N. van Wilgen, B. W. Govender, H. C. Biggs, The contribution of fire research to fire FAPESP (São Paulo Research Foundation) grant #2016/17888-2, D.R.R. was funded by management: A critical review of a long-term experiment in the Kruger National Park, CNPq grant #301589/2015-1, and G.D. was funded by CNPq grant #303179/2016-3. Author South Africa. Int. J. Wildl. Fire 16, 519–530 (2007). contributions: W.A.H. and G.D. designed the study. R.C.R.A., W.A.H., N.A.P., and G.D. 54. A. Andersen, G. D. Cook, R. J. Williams, Fire in Tropical Savannas: The Kapalga Experiment contributed to the vegetation fieldwork, analysis, and manuscript writing. H.L.V. performed (Spriner-Verlag, 2003). the work related to ant biodiversity and analysis and contributed to writing. W.A.H. performed 55. V. R. Pivello, The use of fire in the Cerrado and Amazonian Rainforests of Brazil: Past and the remote sensing analysis. D.R.R. contributed to manuscript writing and project logistics. present. Fire Ecol. 7,24–39 (2011). Competing interests: The authors declare that they have no competing interests. Data 56. J. A. Meira Neto, F. R. Martins, G. E. Valente, Composição florística e espectro biológico and materials availability: All data needed to evaluate the conclusions in the paper na Estação Ecológica de Santa Bárbara, Estado de São Paulo, Brasil. Rev. Árvore 31, are present in the paper and/or the Supplementary Materials. Additional data related to this 907–922 (2007). paper may be requested from the authors. 57. A. C. G. Melo, G. Durigan, Plano de Manejo da Estação Ecológica de Santa Bárbara (SEMA, 2010). 58. G. Durigan, M. F. Siqueira, G. A. D. C. Franco, W. A. Contieri, A flora arbustivo‑arbórea Submitted 21 April 2017 do Médio Paranapanema: Base para a restauração dos ecossistemas naturais, in Pesquisas Accepted 6 August 2017 em Conservação e Recuperação Ambiental no Oeste Paulista: Resultados da Cooperação Published 30 August 2017 Brasil/Japão, O. Vilas Bôas, G. Durigan, Eds. (Páginas & Letras, ed. 1, 2004), pp. 199–239. 10.1126/sciadv.1701284 59. R. C. Mendonça, J. M. Felfili, B. M. T. Walter, M. C. Silva Júnior, A. V. Rezende, T. S. Filgueiras, P. E. Nogueira, C. W. Fagg, in Flora Vascular do Bioma Cerrado Checklist com Citation: R. C. R. Abreu, W. A. Hoffmann, H. L. Vasconcelos, N. A. Pilon, D. R. Rossatto, 12.356 Espécies, S. M. Sano, S. P. Almeida, J. F. Ribeiro, Eds. (Embrapa Cerrados, 2008), G. Durigan, The biodiversity cost of carbon sequestration in tropical savanna. Sci. Adv. 3, pp. 423–1279. e1701284 (2017). Abreu et al., Sci. Adv. 2017; 3 : e1701284 30 August 2017 7of7

Journal

Science AdvancesPubmed Central

Published: Aug 30, 2017

References