Search

Tropical forests and savanna account for nearly 65% of the total global terrestrial net primary production (NPP); however, there are still large uncertainties in tropical forest NPP because of limited field measurements, especially in the structurally diverse Brazilian savanna (cerrado). To address this uncertainty, we measured patterns of aboveground wood C stocks (Cw) and rates of wood C storage (ΔCw) over a 7‐year period for cerrado forests and woodlands of southern Mato Grosso, Brazil, arrayed across hydrological and soil fertility gradients. We focused on ΔCwbecause it is an important component of NPP, and wood is a stable, long‐term, C storage reservoir. Annual rates of ΔCwwere significantly affected by estimates of P and cation (K and Ca) availability, and analysis of covariance indicated that relationships between ΔCwand nutrient availability were independent of stand hydrology. Both upland and hyperseasonal stands exhibited a decline in ΔCwduring the 2015–16 El Niño event, which was exceptionally warm and dry. A limited analysis of the uncertainty associated with the field measurements ranged from 7% for wood density to 24% for tree density, while the uncertainty associated with derived quantities ranged from 10% for tree height to 41% for Cw. Overall, these results suggest that soil fertility and annual precipitation are important drivers of ΔCwand that warming and drying associated with climate change will cause a decline in aboveground woody C storage for these, and similar, tropical forests and woodlands. Tropical forests and savannas are important reservoirs for global carbon (C) storage, but there are still uncertainties in how much C is stored in these ecosystems, especially for savanna. We measured the amount and rate of C stored in aboveground wood for forests and woodlands of the Brazilian savanna in the southern part of the Amazon Basin. Rates of wood C storage increased as nutrient availability increased, and C storage declined during drought years for both seasonally flooded and non‐flooded forests and woodlands. These results have implications for how climate and land‐use change alter wood C storage in these globally important ecosystems. Cerrado forests and woodlands store 0.16–2.62 MgC • ha−1• y−1in aboveground woodWood C storage increased as P and cation (K and Ca) availability increasedWood C storage declined during the 2015–16 El Niño in seasonally flooded and upland cerrado

Brazilian tropical savanna (referred to as Cerrado) has enormous structural variability ranging from tree- and shrub-free grasslands to tall, dense forests that are found in both well-drained upland and seasonally flooded (hyperseasonal) landforms. These structural and hydrological variations cause variations in aboveground carbon (C) storage, and presumably nitrogen (N) and phosphorus (P) storage and fluxes; however, the effects of these structural and hydrological variations on ecosystem N and P pools and fluxes are less well-known. Thus, our objective was to analyze N and P pools and fluxes for a spectrum of Cerrado physiognomies in both upland and hyperseasonal landforms. N and P pools and fluxes were compiled over an 8-10 year period in structurally different forms of Cerrado, including both upland and hyperseasonal grass dominated (campo sujo), typical wooded (sensu stricto) and dense wooded cerrado, and forested Cerrado stands located in the Cuiaba Basin and Northern Pantanal of southern Mato Grosso, Brazil. We found that aboveground N and P storage was generally higher in forested Cerrado than in non-forest landcover types, regardless of hydrology; however, these patterns were not consistent as wood N and P storage was just as high in typical- and dense-wooded savanna as in forested savannas. Wood typically made up the largest aboveground vegetation N and P reservoir, even for woody-grass savanna (campo sujo) which had a substantially lower tree density than wooded or forested savanna, followed by litter and roots, and then foliage. However, the vast majority of N and P was found in the upper 50 cm soil layer, accounting for 82-98% and 72-98% of the total ecosystem N and P, respectively, and even if only the 0-10 cm soil layer is considered, more than half of the ecosystem N and P was found in soil. Nutrient return via litter production was the largest flux followed by mineralization and root uptake, and nutrient cycling in litter production tended to be higher for forests than for wooded and woody-grass savannas. Savanna woodlands and forests in the Pantanal had higher N and P storage than comparable woodlands and forests in the Cuiaba Basin, presumably because of the shallower and stonier soils of the Cuiaba Basin. However, while the vegetation types studied here had large differences in structure (vegetation height, leaf area index, and density), the forested and wooded savannas had comparable belowground and wood N and P pools.

Estimates of net primary (NPP) and ecosystem production (NEP) are needed for tropical savanna, which is structurally diverse but understudied compared to tropical rainforest. We used eddy covariance and inventory methods to estimate carbon (C) fluxes for an upland mixed grassland and a seasonally flooded forest to determine the correspondence between these methods and assess the contribution of C cycling components to the total NPP. Both techniques provided similar estimates of net ecosystem CO2exchange (−3.0 ‒ 2.3 MgC ha−1y−1for eddy covariance vs. −2.0 ‒ 4.3 MgC ha−1y−1for inventory), gross primary production (7.5–16.3 MgC ha−1y−1for eddy covariance vs. 8.7–18.4 MgC ha−1y−1for inventory), and total NPP (0.9–7.5 MgC ha−1y−1for eddy covariance vs. 2.0–9.5 MgC ha−1y−1for inventory). Belowground NPP accounted for 49%–53% of the total NPP for both ecosystems, followed by aboveground litter (26%–27%) and wood (16%–17%) production. Increases in water availability increased the potential for C storage, but the mechanism was different in the savanna types. Compared to other savanna ecosystems, the mixed grassland had a lower productivity and C use efficiency (CUE = NPP/GPP = 0.28), while the Cerrado forest had a high CUE (0.58) and similar C flux rates to other tropical savanna forests. The agreement in the C cycle estimates derived from the eddy covariance and inventory methods increases our confidence in the productivity estimates for these tropical savanna ecosystems. Estimates of net primary and ecosystem production are needed for tropical savannas, which are highly diverse but understudied compared to tropical forests. Several methods have been used to measure primary and ecosystem production; however, these approaches have errors and uncertainties. We used available methods to estimate the productivity of a grass‐dominated savanna in the southern Amazon Basin and a seasonally flooded forest in the northern Pantanal to assess the correspondence in the productivity estimated from these methods and evaluate the contribution of above and belowground processes to the overall ecosystem productivity. Productivity estimates derived from the different techniques were similar. Belowground processes accounted for 49%–53% of the productivity followed by leaf (26%–27%) and wood (16%–17%) production. Increases in water availability increased the productivity for both savanna types, but for different reasons. An increase in soil moisture caused an increase productivity in the mixed grassland but a decline in carbon losses from the seasonally flooded forest. Compared to other savanna ecosystems, the mixed grassland had lower productivity while the forest had high productivity. While our measurements are spatially and temporally limited, the broad agreement in the different methods increases our confidence in the productivity estimates for these tropical savanna ecosystems. Estimates of C cycling from inventory and eddy covariance methods were similar for two floristically different tropical savannas of BrazilBelowground NPP was 49%–53% of the total NPP for both ecosystems, followed by aboveground litter (26%–27%) and wood (16%–17%) productionThe grass dominated savanna had a low C use efficiency (CUE = NPP/GPP = 0.28) while the Cerrado forest had a CUE = 0.58 Estimates of C cycling from inventory and eddy covariance methods were similar for two floristically different tropical savannas of Brazil Belowground NPP was 49%–53% of the total NPP for both ecosystems, followed by aboveground litter (26%–27%) and wood (16%–17%) production The grass dominated savanna had a low C use efficiency (CUE = NPP/GPP = 0.28) while the Cerrado forest had a CUE = 0.58

High atmospheric N inputs to southern Californian shrublands can alter organic matter quantity and quality and influence decomposition and N mineralization rates. We conducted a laboratory experiment to test the hypothesis that N addition would increase microbial respiration and net N mineralization. Intact soil cores and litter from Artemisia californicaLess. were collected from field plots exposed to either ambient or elevated N (50 kg N ha−1yr−1) over a 7‐yr period. Microbial respiration was significantly higher in control soil, and litter amendment, regardless of origin, significantly increased respiration. In contrast, net N mineralization was significantly higher in added N soil, and litter addition significantly depressed net N mineralization. Our results suggest that chronic N deposition causes a decline in litter decomposition and microbial N demand. If indicative of the longer‐term decomposition kinetics, these results suggest that chronic N inputs may increase soil C storage and N availability in these semiarid shrublands.

The Pantanal is a large and diverse wetland that spans three South American countries including parts of Brazil, Paraguay, and Bolivia. An invasive tree, Vochysia divergensPohl (commonly known as Cambara), has been expanding throughout the Pantanal for more than four decades forming monospecific stands. Given the rapid and extensive spread of Cambara, and the potential for woody plant invasion to alter soil organic matter and nutrient stocks, we hypothesized that Cambara invasion would significantly increase soil organic matter and nutrient content. To test this hypothesis we sampled the soil, litter, and vegetation of a monospecific Cambara stand, a grassland (campo) stand in the process of Cambara invasion (transitional), and a campo stand free of Cambara during the dry season in July and August of 2009 and 2010 when stands were free from seasonal flooding. Surface (0–10 cm) soil in sites dominated by Cambara had significantly higher soil organic matter (SOM), P, and cation content, and higher cation exchange capacity (CEC), but soil pH and K+concentration were less coincident with Cambara presence. The variation of soil characteristics was also significantly higher in the transitional site, indicating that the transition from a grassland‐ to a Cambara‐dominated system significantly increased the spatial variability of soil chemistry. These results indicate that Cambara invasion fundamentally alters the C and nutrient storage of Panatanal soils. Future research will investigate if these results are general for Cambara invasion and address mechanisms of how Cambara invasion affects the nutrient cycling and storage of Pantanal soils.

Semiarid chaparral and coastal sage shrublands of southern California have been exposed to high levels of atmospheric N for decades, which has the capacity to increase both N and C storage and cycling in these N‐limited systems. Thus we hypothesize that soil C and N mineralization will be higher in areas that have been exposed to high atmospheric N deposition. This hypothesis was tested in a 50‐wk laboratory incubation experiment where the inorganic N (NH4+ NO3) and CO2production of chaparral and coastal sage soils were repeatedly measured. Soil was incubated in the dark at a constant temperature of 25°C and a soil moisture of 0.25 kg H2O kg−1dry soil (65% water‐filled pore space). Relative differences in N deposition exposure between the study sites were quantified by repeatedly rinsing and collecting the N accumulated on branch surfaces during 1 yr. Temporal trends in cumulative C and N mineralization were best described by single‐pool first‐order and zero‐order models, respectively. Total N mineralization, but not C mineralization, increased linearly with relative N deposition, and NO3accounted for 95% of the total inorganic N accumulated during the 50‐wk incubation. The soil δ15N natural abundance increased with relative N deposition (r= 0.85, P< 0.05) and the soil C/N ratio declined with relative N deposition (r= −0.74, P< 0.05), suggesting that N deposition exposure enhanced N mineralization in part because of increases in the soil organic matter quality (i.e., lower C/N ratio). Furthermore, soil C storage declined as a function of relative N deposition exposure, indicating that high atmospheric N inputs are not likely to stimulate soil C storage in these semiarid ecosystems.