N and P constrain C in ecosystems under climate change: Role of nutrient redistribution, accumulation, and stoichiometry.

We use the Multiple Element Limitation (MEL) model to examine responses of 12 ecosystems to elevated carbon dioxide (CO2 ), warming, and 20% decreases or increases in precipitation. Ecosystems respond synergistically to elevated CO2 , warming, and decreased precipitation combined because higher water-use efficiency with elevated CO2 and higher fertility with warming compensate for responses to drought. Response to elevated CO2 , warming, and increased precipitation combined is additive. We analyze changes in ecosystem carbon (C) based on four nitrogen (N) and four phosphorus (P) attribution factors: (1) changes in total ecosystem N and P, (2) changes in N and P distribution between vegetation and soil, (3) changes in vegetation C:N and C:P ratios, and (4) changes in soil C:N and C:P ratios. In the combined CO2 and climate change simulations, all ecosystems gain C. The contributions of these four attribution factors to changes in ecosystem C storage varies among ecosystems because of differences in the initial distributions of N and P between vegetation and soil and the openness of the ecosystem N and P cycles. The net transfer of N and P from soil to vegetation dominates the C response of forests. For tundra and grasslands, the C gain is also associated with increased soil C:N and C:P. In ecosystems with symbiotic N fixation, C gains resulted from N accumulation. Because of differences in N versus P cycle openness and the distribution of organic matter between vegetation and soil, changes in the N and P attribution factors do not always parallel one another. Differences among ecosystems in C-nutrient interactions and the amount of woody biomass interact to shape ecosystem C sequestration under simulated global change. We suggest that future studies quantify the openness of the N and P cycles and changes in the distribution of C, N, and P among ecosystem components, which currently limit understanding of nutrient effects on C sequestration and responses to elevated CO2 and climate change.

Tripartite legume-rhizobia-mycorrhizae relationship is influenced by light and soil nitrogen in Neotropical canopy gaps.

Plants and their soil microbial symbionts influence ecosystem productivity and nutrient cycling, but the controls on these symbioses remain poorly understood. This is particularly true for plants in the Fabaceae family (hereafter legumes), which can associate with both arbuscular mycorrhizal fungi (AMF) and nitrogen (N) -fixing bacteria. Here we report results of the first manipulated field experiment to explore the abiotic and biotic controls of this tripartite symbiosis in Neotropical canopy gaps (hereafter gaps). We grew three species of Neotropical N-fixing legume seedlings under different light (gap-full light, gap-shadecloth, and understory) and soil nitrogen (20 g N·m-2 ·yr-1 vs. 0 g N·m-2 ·yr-1 ) conditions across a lowland tropical forest at La Selva Biological Station, Costa Rica. We harvested the seedlings after 4 months of growth in the field and measured percent AMF root colonization (%AMF), nodule and seeding biomass, and seedling aboveground:belowground biomass ratios. Our expectation was that seedlings in gaps would grow larger and, as a result of higher light, invest more carbon in both AMF and N-fixing bacteria. Indeed, seedlings in gaps had higher total biomass, nodule biomass (a proxy for N-fixing bacteria investment) and rates of AMF root colonization, and the three were significantly positively correlated. However, we only found a significant positive effect of light availability on %AMF when seedlings were fertilized with N. Furthermore, when we statistically controlled for treatment, species, and site effects, we found %AMF and seedling biomass had a negative relationship. This was likely driven by the fact that seedlings invested relatively less in AMF as they increased in biomass (lower %AMF per gram of seedling). Taken together, these results challenge the long-held assumption that high light conditions universally increase carbon investment in AMF and demonstrate that this tripartite symbiosis is influenced by soil nutrient and light conditions.

Light fuels while nitrogen suppresses symbiotic nitrogen fixation hotspots in neotropical canopy gap seedlings.

Mature neotropical lowland forests have relatively lower symbiotic nitrogen fixation (SNF) rates compared with secondary forests. Canopy gap formation may create transient SNF hotspots in mature forests that increase overall SNF rates in these ecosystems, as canopy gaps are pervasive across the landscape and increasing in frequency. However, what environmental conditions are driving SNF upregulation in canopy gaps is unknown. In a field experiment to test these potential environmental controls on SNF, we grew 540 neotropical nitrogen-fixing legume seedlings (Pentaclethra macroloba, Zygia longifolia, and Stryphnodendron microstachyum) under manipulated light and soil nitrogen availability in canopy gaps and intact forests at La Selva Biological Station, Costa Rica. Seedling biomass, nodule biomass, and SNF (g N seedling-1 h-1 ) were 4-, 17- and 42-fold higher, respectively, in canopy gaps than in the intact forest. Nitrogen additions decreased SNF, but light had a stronger positive effect. Upregulation of SNF in canopy gaps was driven by increased plant growth and not a disproportionate increased SNF allocation. These data provide evidence that canopy gap SNF hotspots are driven, in part, by light availability, demonstrating a potential driver of SNF spatial heterogeneity. This further suggests that canopy gap dynamics are important for understanding the biogeochemistry of neotropical forests.

A New Framework for Evaluating Estimates of Symbiotic Nitrogen Fixation in Forests.

Symbiotic nitrogen fixation (SNF) makes atmospheric nitrogen biologically available and regulates carbon storage in many terrestrial ecosystems. Despite its global importance, estimates of SNF rates are highly uncertain, particularly in tropical forests where rates are assumed to be high. Here we provide a framework for evaluating the uncertainty of sample-based SNF estimates and discuss its implications for quantifying SNF and thus understanding of forest function. We apply this framework to field data sets from six lowland tropical rainforests (mature and secondary) in Brazil and Costa Rica. We use this data set to estimate parameters influencing SNF estimation error, notably the root nodule abundance and variation in SNF rates among soil cores containing root nodules. We then use simulations to gauge the relationship between sampling effort and SNF estimation accuracy for a combination of parameters. Field data illuminate a highly right-skewed lognormal distribution of SNF rates among soil cores containing root nodules that were rare and spanned five orders of magnitude. Consequently, simulations demonstrated that sample sizes of hundreds to even thousands of soil cores are needed to obtain estimates of SNF that are within, for example, a factor of 2 of the actual rate with 75% probability. This represents sample sizes that are larger than most studies to date. As a result of this previously undescribed uncertainty, we suggest that current estimates of SNF in tropical forests are not sufficiently constrained to elucidate forest stand-level controls of SNF, which hinders our understanding of the impact of SNF on tropical forest ecosystem processes.

Soil Carbon Dynamics in Soybean Cropland and Forests in Mato Grosso, Brazil.

Climate and land use models predict that tropical deforestation and conversion to cropland will produce a large flux of soil carbon (C) to the atmosphere from accelerated decomposition of soil organic matter (SOM). However, the C flux from the deep tropical soils on which most intensive crop agriculture is now expanding remains poorly constrained. To quantify the effect of intensive agriculture on tropical soil C, we compared C stocks, radiocarbon, and stable C isotopes to 2 m depth from forests and soybean cropland created from former pasture in Mato Grosso, Brazil. We hypothesized that soil disturbance, higher soil temperatures (+2°C), and lower OM inputs from soybeans would increase soil C turnover and deplete C stocks relative to nearby forest soils. However, we found reduced C concentrations and stocks only in surface soils (0-10 cm) of soybean cropland compared with forests, and these differences could be explained by soil mixing during plowing. The amount and Δ14C of respired CO2 to 50 cm depth were significantly lower from soybean soils, yet CO2 production at 2 m deep was low in both forest and soybean soils. Mean surface soil δ13C decreased by 0.5‰ between 2009 and 2013 in soybean cropland, suggesting low OM inputs from soybeans. Together these findings suggest the following: (1) soil C is relatively resistant to changes in land use and (2) conversion to cropland caused a small, measurable reduction in the fast-cycling C pool through reduced OM inputs, mobilization of older C from soil mixing, and/or destabilization of SOM in surface soils.

Remotely sensed canopy nitrogen correlates with nitrous oxide emissions in a lowland tropical rainforest.

Tropical forests exhibit significant heterogeneity in plant functional and chemical traits that may contribute to spatial patterns of key soil biogeochemical processes, such as carbon storage and greenhouse gas emissions. Although tropical forests are the largest ecosystem source of nitrous oxide (N2 O), drivers of spatial patterns within forests are poorly resolved. Here, we show that local variation in canopy foliar N, mapped by remote-sensing image spectroscopy, correlates with patterns of soil N2 O emission from a lowland tropical rainforest. We identified ten 0.25 ha plots (assemblages of 40-70 individual trees) in which average remotely-sensed canopy N fell above or below the regional mean. The plots were located on a single minimally-dissected terrace (<1 km2 ) where soil type, vegetation structure and climatic conditions were relatively constant. We measured N2 O fluxes monthly for 1 yr and found that high canopy N species assemblages had on average three-fold higher total mean N2 O fluxes than nearby lower canopy N areas. These differences are consistent with strong differences in litter stoichiometry, nitrification rates and soil nitrate concentrations. Canopy N status was also associated with microbial community characteristics: lower canopy N plots had two-fold greater soil fungal to bacterial ratios and a significantly lower abundance of ammonia-oxidizing archaea, although genes associated with denitrification (nirS, nirK, nosZ) showed no relationship with N2 O flux. Overall, landscape emissions from this ecosystem are at the lowest end of the spectrum reported for tropical forests, consist with multiple metrics indicating that these highly productive forests retain N tightly and have low plant-available losses. These data point to connections between canopy and soil processes that have largely been overlooked as a driver of denitrification. Defining relationships between remotely-sensed plant traits and soil processes offers the chance to map these processes at large scales, potentially increasing our ability to predict N2 O emissions in heterogeneous landscapes.

Nitrogen cycling during secondary succession in Atlantic Forest of Bahia, Brazil.

Carbon accumulation in tropical secondary forests may be limited in part by nitrogen (N) availability, but changes in N during tropical forest succession have rarely been quantified. We explored N cycle dynamics across a chronosequence of secondary tropical forests in the Mata Atlântica of Bahia, Brazil in order to understand how quickly the N cycle recuperates. We hypothesized that N fixation would decline over the course of succession as N availability and N gaseous losses increased. We measured N fixation, KCl-extractable N, net mineralization and nitrification, resin-strip sorbed N, gaseous N emissions and the soil δ15N in stands that were 20, 35, 50, and > 50 years old. Contrary to our initial hypothesis, we found no significant differences between stand ages in any measured variable. Our findings suggest that secondary forests in this region of the Atlantic forest reached pre-disturbance N cycling dynamics after just 20 years of succession. This result contrasts with previous study in the Amazon, where the N cycle recovered slowly after abandonment from pasture reaching pre-disturbance N cycling levels after ~50 years of succession. Our results suggest the pace of the N cycle, and perhaps tropical secondary forest, recovery, may vary regionally.

Chronosequence predictions are robust in a Neotropical secondary forest, but plots miss the mark.

Tropical secondary forests (TSF) are a global carbon sink of 1.6 Pg C/year. However, TSF carbon uptake is estimated using chronosequence studies that assume differently aged forests can be used to predict change in aboveground biomass density (AGBD) over time. We tested this assumption using two airborne lidar datasets separated by 11.5 years over a Neotropical landscape. Using data from 1998, we predicted canopy height and AGBD within 1.1 and 10.3% of observations in 2009, with higher accuracy for forest height than AGBD and for older TSFs in comparison to younger ones. This result indicates that the space-for-time assumption is robust at the landscape-scale. However, since lidar measurements of secondary tropical forest are rare, we used the 1998 lidar dataset to test how well plot-based studies quantify the mean TSF height and biomass in a landscape. We found that the sample area required to produce estimates of height or AGBD close to the landscape mean is larger than the typical area sampled in secondary forest chronosequence studies. For example, estimating AGBD within 10% of the landscape mean requires more than thirty 0.1 ha plots per age class, and more total area for larger plots. We conclude that under-sampling in ground-based studies may introduce error into estimations of the TSF carbon sink, and that this error can be reduced by more extensive use of lidar measurements.

Nutrient acquisition, soil phosphorus partitioning and competition among trees in a lowland tropical rain forest.

We hypothesized that dinitrogen (N2 )- and non-N2 -fixing tropical trees would have distinct phosphorus (P) acquisition strategies allowing them to exploit different P sources, reducing competition. We measured root phosphatase activity and arbuscular mycorrhizal (AM) colonization among two N2 - and two non-N2 -fixing seedlings, and grew them alone and in competition with different inorganic and organic P forms to assess potential P partitioning. We found an inverse relationship between root phosphatase activity and AM colonization in field-collected seedlings, indicative of a trade-off in P acquisition strategies. This correlated with the predominantly exploited P sources in the seedling experiment: the N2 fixer with high N2 fixation and root phosphatase activity grew best on organic P, whereas the poor N2 fixer and the two non-N2 fixers with high AM colonization grew best on inorganic P. When grown in competition, however, AM colonization, root phosphatase activity and N2 fixation increased in the N2 fixers, allowing them to outcompete the non-N2 fixers regardless of P source. Our results indicate that some tropical trees have the capacity to partition soil P, but this does not eliminate interspecific competition. Rather, enhanced P and N acquisition strategies may increase the competitive ability of N2 fixers relative to non-N2 fixers.

Solute and sediment export from Amazon forest and soybean headwater streams.

Intensive cropland agriculture commonly increases streamwater solute concentrations and export from small watersheds. In recent decades, the lowland tropics have become the world's largest and most important region of cropland expansion. Although the effects of intensive cropland agriculture on streamwater chemistry and watershed export have been widely studied in temperate regions, their effects in tropical regions are poorly understood. We sampled seven headwater streams draining watersheds in forest (n = 3) or soybeans (n = 4) to examine the effects of soybean cropping on stream solute concentrations and watershed export in a region of rapid soybean expansion in the Brazilian state of Mato Grosso. We measured stream flows and concentrations of NO3- , PO43- , SO42- , Cl- , NH4+ , Ca2+ , Mg2+ , Na+ , K+ , Al3+ , Fe3+ , and dissolved organic carbon (DOC) biweekly to monthly to determine solute export. We also measured stormflows and stormflow solute concentrations in a subset of watersheds (two forest, two soybean) during two/three storms, and solutes and δ18 O in groundwater, rainwater, and throughfall to characterize watershed flowpaths. Concentrations of all solutes except K+ varied seasonally in streamwater, but only Fe3+ concentrations differed between land uses. The highest streamwater and rainwater solute concentrations occurred during the peak season of wildfires in Mato Grosso, suggesting that regional changes in atmospheric composition and deposition influence seasonal stream solute concentrations. Despite no concentration differences between forest and soybean land uses, annual export of NH4+ , PO43- , Ca2+ , Fe3+ , Na+ , SO42- , DOC, and TSS were significantly higher from soybean than forest watersheds (5.6-fold mean increase). This increase largely reflected a 4.3-fold increase in water export from soybean watersheds. Despite this increase, total solute export per unit watershed area (i.e., yield) remained low for all watersheds (<1 kg NO3- N·ha-1 ·yr-1 , <2.1 kg NH4+ -N·ha-1 ·yr-1 , <0.2 kg PO43- -P·ha-1 ·yr-1 , <1.5 kg Ca2+ ·ha-1 ·yr-1 ). Responses of both streamflows and solute concentrations to crop agriculture appear to be controlled by high soil hydraulic conductivity, groundwater-dominated hydrologic flowpaths on deep soils, and the absence of nitrogen fertilization. To date, these factors have buffered streams from the large increases in solute concentrations that often accompany intensive croplands in other locations.

Nutrient limitation in tropical secondary forests following different management practices.

Secondary forests now make up more than one-half of all tropical forests, and constraints on their biomass accumulation will influence the strength of the terrestrial carbon (C) sink in the coming decades. However the variance in secondary tropical forest biomass for a given stand age and climate is high and our understanding of why is limited. We constructed a model of terrestrial C, nitrogen (N), and phosphorus (P) cycling to examine the influence of disturbance and management practices on nutrient limitation and biomass recovery in secondary tropical forests. The model predicted that N limited the rate of forest recovery in the first few decades following harvest, but that this limitation switched to P approximately 30-40 yr after abandonment, consistent with field data on N and P cycling from secondary tropical forest chronosequences. Simulated biomass recovery agreed well with field data of biomass accumulation following harvest (R2  = 0.80). Model results showed that if all biomass remained on site following a severe disturbance such as blowdown, regrowth approached pre-disturbance biomass in 80-90 yr, and recovery was faster following smaller disturbances such as selective logging. Field data from regrowth on abandoned pastures were consistent with simulated losses of nutrients in soil organic matter, particularly P. Following any forest disturbance that involved the removal of nutrients (i.e., except blowdown), forest regrowth produced reduced biomass relative to the initial state as a result of nutrient loss through harvest, leaching and/or sequestration by secondary minerals. Differences in nutrient availability accounted for 49-94% of the variance in secondary forest biomass C at a given stand age. Management lessons from this study are the importance of strategies that help retain nutrients on site, recognizing the role of coarse woody debris in immobilization and subsequent release of nutrients, and the potential for nutrient additions to enhance biomass growth and recovery in secondary tropical forests.

Environmental controls on canopy foliar nitrogen distributions in a Neotropical lowland forest.

Distributions of foliar nutrients across forest canopies can give insight into their plant functional diversity and improve our understanding of biogeochemical cycling. We used airborne remote sensing and partial least squares regression to quantify canopy foliar nitrogen (foliar N) across ~164 km2 of wet lowland tropical forest in the Osa Peninsula, Costa Rica. We determined the relative influence of climate and topography on the observed patterns of foliar N using a gradient boosting model technique. At a local scale, where climate and substrate were constant, we explored the influence of slope position on foliar N by quantifying foliar N on remnant terraces, their adjacent slopes, and knife-edged ridges. In addition, we climbed and sampled 540 trees and analyzed foliar N in order to quantify the role of species identity (phylogeny) and environmental factors in predicting foliar N. Observed foliar N heterogeneity reflected environmental factors working at multiple spatial scales. Across the larger landscape, elevation and precipitation had the highest relative influence on predicting foliar N (30% and 24%), followed by soils (15%), site exposure (9%), compound topographic index (8%), substrate (6%), and landscape dissection (6%). Phylogeny explained ~75% of the variation in the field collected foliar N data, suggesting that phylogeny largely underpins the response to the environmental factors. Taken together, these data suggest that a large fraction of the variance in foliar N across the landscape is proximately driven by species composition, though ultimately this is likely a response to abiotic factors such as climate and topography. Future work should focus on the mechanisms and feedbacks involved, and how shifts in climate may translate to changes in forest function.

The phosphorus cost of agricultural intensification in the tropics.

Agricultural intensification in the tropics is one way to meet rising global food demand in coming decades(1,2). Although this strategy can potentially spare land from conversion to agriculture(3), it relies on large material inputs. Here we quantify one such material cost, the phosphorus fertilizer required to intensify global crop production atop phosphorus-fixing soils and achieve yields similar to productive temperate agriculture. Phosphorus-fixing soils occur mainly in the tropics, and render added phosphorus less available to crops(4,5). We estimate that intensification of the 8-12% of global croplands overlying phosphorus-fixing soils in 2005 would require 1-4 Tg P yr(-1) to overcome phosphorus fixation, equivalent to 8-25% of global inorganic phosphorus fertilizer consumption that year. This imposed phosphorus 'tax' is in addition to phosphorus added to soils and subsequently harvested in crops, and doubles (2-7 Tg P yr(-1)) for scenarios of cropland extent in 2050(6). Our estimates are informed by local-, state- and national-scale investigations in Brazil, where, more than any other tropical country, low-yielding agriculture has been replaced by intensive production. In the 11 major Brazilian agricultural states, the surplus of added inorganic fertilizer phosphorus retained by soils post harvest is strongly correlated with the fraction of cropland overlying phosphorus-fixing soils (r(2) = 0.84, p < 0.001). Our interviews with 49 farmers in the Brazilian state of Mato Grosso, which produces 8% of the world's soybeans mostly on phosphorus-fixing soils, suggest this phosphorus surplus is required even after three decades of high phosphorus inputs. Our findings in Brazil highlight the need for better understanding of long-term soil phosphorus fixation elsewhere in the tropics. Strategies beyond liming, which is currently widespread in Brazil, are needed to reduce phosphorus retention by phosphorus-fixing soils to better manage the Earth's finite phosphate rock supplies and move towards more sustainable agricultural production.

Ecological and evolutionary variation in community nitrogen use traits during tropical dry forest secondary succession.

We assessed the role of ecological and evolutionary processes in driving variation in leaf and litter traits related to nitrogen (N) use among tropical dry forest trees in old-growth and secondary stands in western Mexico. Our expectation was that legumes (Fabaceae), a dominant component of the regional flora, would have consistently high leaf N and therefore structure phylogenetic variation in N-related traits. We also expected ecological selection during succession for differences in nitrogen use strategies, and corresponding shifts in legume abundance. We used phylogenetic analyses to test for trait conservatism in foliar and litter N, C:N, and N resorption. We also evaluated differences in N-related traits between old-growth and secondary forests. We found a weak phylogenetic signal for all traits, partly explained by wide variation within legumes. Across taxa we observed a positive relationship between leaf and litter N, but no shift in resorption strategies along the successional gradient. Despite species turnover, N-resorption, and N-related traits showed little change across succession, suggesting that, at least for these traits, secondary forests rapidly recover ecosystem function. Collectively, our results also suggest that legumes should not be considered a single functional group from a biogeochemical perspective.

Structure and composition of altered riparian forests in an agricultural Amazonian landscape.

Deforestation and fragmentation influence the microclimate, vegetation structure, and composition of remaining patches of tropical forest. In the southern Amazon, at the frontier of cropland expansion, forests are converted and fragmented in a pattern that leaves standing riparian forests whose dimensions are mandated by the Brazilian National Forest Code. These altered riparian forests share many characteristics of well-studied upland forest fragments, but differ because they remain connected to larger areas of forest downstream, and because they may experience wetter soil conditions because reduction of forest cover in the surrounding watershed raises groundwater levels and increases stream runoff. We compared forest regeneration, structure, composition, and diversity in four areas of intact riparian forest and four areas each of narrow, medium, and wide altered riparian forests that have been surrounded by agriculture since the early 1980s. We found that seedling abundance was reduced by as much as 64% and sapling abundance was reduced by as much as 67% in altered compared to intact riparian forests. The most pronounced differences between altered and intact forest occurred near forest edges and within the narrowest sections of altered riparian forests. Woody plant species composition differed and diversity was reduced in altered forests compared to intact riparian forests. However, despite being fragmented for several decades, large woody plant biomass and carbon storage, the number of live or dead large woody plants, mortality rates, and the size distribution of woody plants did not differ significantly between altered and intact riparian forests. Thus, even in these relatively narrow forests with high edge: area ratios, we saw no evidence of the increases in mortality and declines in biomass that have been found in other tropical forest fragment studies. However, because of the changes in both species community and reduced regeneration, it is unclear how long this relative lack of change will be sustained. Additionally, Brazil recently passed a law in their National Forest Code allowing narrower riparian buffers than those studied here in restored areas, which could affect their long-term sustainability.

Assessing nutrient limitation in complex forested ecosystems: alternatives to large-scale fertilization experiments.

Quantifying nutrient limitation of primary productivity is a fundamental task of terrestrial ecosystem ecology, but in a high carbon dioxide environment it is even more critical that we understand potential nutrient constraints on plant growth. Ecologists often manipulate nutrients with fertilizer to assess nutrient limitation, yet for a variety of reasons, nutrient fertilization experiments are either impractical or incapable of resolving ecosystem responses to some global changes. The challenges of conducting large, in situ fertilization experiments are magnified in forests, especially the high-diversity forests common throughout the lowland tropics. A number of methods, including fertilization experiments, could be seen as tools in a toolbox that ecologists may use to attempt to assess nutrient limitation, but there has been no compilation or synthetic discussion of those methods in the literature. Here, we group these methods into one of three categories (indicators of soil nutrient supply, organismal indicators of nutrient limitation, and lab-based experiments and nutrient depletions), and discuss some of the strengths and limitations of each. Next, using a case study, we compare nutrient limitation assessed using these methods to results obtained using large-scale fertilizations across the Hawaiian Archipelago. We then explore the application of these methods in high-diversity tropical forests. In the end, we suggest that, although no single method is likely to predict nutrient limitation in all ecosystems and at all scales, by simultaneously utilizing a number of the methods we describe, investigators may begin to understand nutrient limitation in complex and diverse ecosystems such as tropical forests. In combination, these methods represent our best hope for understanding nutrient constraints on the global carbon cycle, especially in tropical forest ecosystems.

The fate of phosphorus fertilizer in Amazon soya bean fields.

Fertilizer-intensive soya bean agriculture has recently expanded in southeastern Amazonia, and whereas intensive fertilizer use in the temperate zone has led to widespread eutrophication of freshwater ecosystems, the effects in tropical systems are less well understood. We examined the fate of fertilizer phosphorus (P) by comparing P forms and budgets across a chronosequence of soya bean fields (converted to soya beans between 2003 and 2008) and forests on an 800 km(2) soya bean farm in Mato Grosso, Brazil. Soya bean fields were fertilized with 50 kg P ha(-1) yr(-1) (30 kg P ha(-1) yr(-1) above what is removed in crops). We used modified Hedley fractionation to quantify soil P pools and found increases in less-plant-available inorganic pools and decreases in organic pools in agricultural soils compared with forest. Fertilizer P did not move below 20 cm. Measurements of P sorption capacity suggest that while fertilizer inputs quench close to half of the sorption capacity of fast-reacting pools, most added P is bound in more slowly reacting pools. Our data suggest that this agricultural system currently has a low risk of P losses to waterways and that long time-scales are required to reach critical soil thresholds that would allow continued high yields with reduced fertilizer inputs.

Watershed responses to Amazon soya bean cropland expansion and intensification.

The expansion and intensification of soya bean agriculture in southeastern Amazonia can alter watershed hydrology and biogeochemistry by changing the land cover, water balance and nutrient inputs. Several new insights on the responses of watershed hydrology and biogeochemistry to deforestation in Mato Grosso have emerged from recent intensive field campaigns in this region. Because of reduced evapotranspiration, total water export increases threefold to fourfold in soya bean watersheds compared with forest. However, the deep and highly permeable soils on the broad plateaus on which much of the soya bean cultivation has expanded buffer small soya bean watersheds against increased stormflows. Concentrations of nitrate and phosphate do not differ between forest or soya bean watersheds because fixation of phosphorus fertilizer by iron and aluminium oxides and anion exchange of nitrate in deep soils restrict nutrient movement. Despite resistance to biogeochemical change, streams in soya bean watersheds have higher temperatures caused by impoundments and reduction of bordering riparian forest. In larger rivers, increased water flow, current velocities and sediment flux following deforestation can reshape stream morphology, suggesting that cumulative impacts of deforestation in small watersheds will occur at larger scales.