Search

Savannas contribute to ca. 30 % of the total terrestrial net primary productivity and are responsible for significant carbon storage. Savannas in South America are mostly found within the Cerrado Domain, which is very threatened and presents remarkable carbon pools. Herein, we used a unique dataset of 21 Cerrado sites spanning 144 permanent field plots in Southeastern Brazil to assess the general patterns of above and belowground carbon stocks. We identified the main environmental and tree diversity drivers of aboveground wood carbon and productivity, belowground carbon stocks (roots and soil), carbon ratios (root:shoot and above:below) and total carbon stocks in the Cerrado through a combination of climatic estimates, fire frequency data, field measurements of vegetation, roots, soil carbon, nutrients and texture, and assessment of different components of diversity (species, functional and phylogenetic). Our findings reveal average aboveground, root, and soil carbon stocks of 20.4, 14.24, and 123.13 Mg.ha

Forest production is strongly dependent on nutrient uptake; however, sustainable management of intensively managed plantations requires an improved understanding of this relationship when fertilization occurs frequently across short rotations. Here, we studied temporal nitrogen (N) concentration ([N]) and content (Nc) dynamics under different silvicultural practices (herbicide, fertilization, and planting density) throughout early loblolly pine (Pinus taeda) stand development (5 years). We describe relationships of [N] and Nc of different stand components (foliage, branches, stem, roots, and competing vegetation) with carbon and biomass. Our results demonstrate that [N] of perennial loblolly tissues do not respond to silvicultural practices and progressively decrease through development. While foliar [N] was most responsive to resource availability, it was not consistent across time. Controlling competing vegetation was crucial to promote the use of site resources by the crop tree and increased loblolly Nc by >500%. However, increased N uptake and expedited growth is dependent upon fertilization early in stand development. At age 5, herbicide plus reduced and full fertilization rates exhibited similar aboveground Nc, which was 32% higher than with herbicide only. Increasing planting density resulted in increased above- and belowground loblolly Nc; however, increases in Nc were not proportional with increases in planting density. Net primary productivity and N uptake were linearly related, but age/development strongly controlled N use efficiency. Our study helps to understand complex relationships between N, biomass, and silvicultural practices during early stand development and demonstrates that temporal evaluation of nutrient dynamics is crucial to better understand loblolly pine growth, carbon sequestration potential, and to inform sustainable silvicultural practices across short rotations.

Retaining harvest residues (HR) in the field is considered an option to recycle carbon (C) and nutrients in short-rotation plantations, but the fate of HR in the soil remains unclear. The effects of HR management and nitrogen (N) availability on soil organic matter (SOM) pools and microbial community structure of forest plantations represent a major knowledge gap. To fill this gap, we explored how management scenarios that differed in the amount and type of HR [removal of all residues (-R), residue retention without (+R/-B), or with bark (+R/+B)] and N availability [0 (-N) or 200 kg N ha-1 (+N)] influenced microbial activity and structure and SOM fractions in a recently afforested grassland. Specifically, we measured HR decomposition, carbon dioxide efflux (respired CO2), phospholipid fatty acid (PLFA), and determined changes in particulate (POM) and mineral-associated organic matter (MAOM) over 12 months using differences in δ13C natural abundance between Eucalyptus HR and grassland soils (δ13C ∼-28 ‰ and -13 ‰, respectively). Microbial respiration was stimulated by HR retention (+R). Bark retention reduced HR half-life by ∼70 days, on average, while N had little influence. Bacterial groups (Gram-positive and Gram-negative) were the primary decomposer of eucalypt HR, while Actinobacteria used more of the former soil organic carbon (SOC). +R/+B increased fungal biomarker PLFA concentration and fungal:bacterial ratio, suggesting a key role of Fungi in the fragmentation of woody HR. N influence on microbial community structure and SOM pools was dependent upon HR management. Retaining HR increased SOC concentrations, most significantly in the 0–1 cm soil depth and POM fraction, and when bark was included. +R/+B resulted in higher POM-C concentrations (up to +37%) and more depleted δ13C-POM compared with -R and +R/-B, respectively. Collectively, our results suggest it is feasible to increase POM-C and fungal abundance through HR management practices in the early stages of decomposition, which may potentially contribute to SOC stabilization in the long-term. Yet our findings remain to be tested in long-term studies, we provide quantitative evidence of the potential of a more conservative HR management to contribute to the sustainability of eucalypt plantations.

The southern Brazilian grasslands (Pampas) are a new frontier of eucalypt plantation expansion, but the impacts of this land use change (LUC) on soil organic matter (SOM) are not well documented. In this study, we evaluated changes in the particulate organic matter (POM) and mineral-associated organic matter (MAOM) fractions following the afforestation of a grassland in southern Brazil over an entire eucalypt rotation. Additionally, SOM dynamics was assessed in two soil zones [row (R) and inter-row (IR)] created by soil preparation, which consists on scraping the soil surface from the IR to the R along with subsoiling to build ridges for eucalypt plating (i.e., ridge tillage). The soil was sampled to a depth of 60 cm in the R and IR positions, and the carbon (C) and nitrogen (N) in the SOM fractions were determined. We used natural 13C abundance differences to compare the contributions of the ‘new’ C input (C3, eucalypt stands) to the ‘native’ SOM (C4, grasslands). The conversion of the grassland to eucalypt stands resulted in losses in C and N stocks in both the R and IR after 68 months. Soil management controlled the magnitude and directions of C and N changes. The C loss in the R was ∼1.3-fold higher than that in the IR (20.6 Mg ha−1; p

Production forestry in the southeast US has been partially transitioned to intensively managed short rotations (~10 years), in which multiple silvicultural interventions are performed during forest development. Understanding the responses to silvicultural practices and continued refinement of site-specific recommendations is critical to sustainably maximize forest production. We evaluated the effects of silvicultural practices (herbicide, fertilization, and planting density) on growth, stand homogeneity, and above- and belowground biomass accumulation and partitioning of loblolly pine (Pinus taeda) throughout early stand development (age 5 years) in the southeastern US. Five treatments with eight replications each were tested: no herbicide and no fertilization (C); herbicide only (H); herbicide and half-reduced fertilization rate (R); herbicide and full fertilization (F); and increased stand density (60+ %; 1346 vs. 2152 trees per hectare) with herbicide and full fertilization rate (D). Allometric equations generated from destructive harvests were applied to annual diameter measurements to estimate plot-level biomass and allocation. Herbicide was crucial to promote stand uniformity and increase yield (~600+ % stem biomass compared with C at age 5). Aboveground biomass was similar in R and F treatments, which was ~25% higher than in H at age 5. Increasing planting density along with multiple herbicide and fertilizer applications yielded higher biomass without compromising individual tree size (diameter and height). There was little effect of silviculture practices on allocation patterns. Our results parallel what was found for fertilization with herbicide from a number of loblolly stands under similar conditions and indicate a ~28% volume gain with fertilization during early stand development. Similarly, our results were consistent with other studies implementing similar differences in planting density and suggest a ~26% volume gain through early stand development with an initial 60% increase in planting density. Our study helps to understand complex relationships between production and silvicultural practices during early stand development and demonstrates that silvicultural prescriptions can be optimized to increase sustainability of production forestry.

In Brazil, studies addressing soil carbon (C) stocks under native vegetation are scarce and often limited to land use change comparisons at the plot level, while studies across larger areas are based on legacy data that address multiple land uses, use wet-oxidations methods to determine C concentrations (Cc) and lack proper soil bulk density (ρb) determinations. Together, these factors may result in either underestimations or biased soil C stocks, thereby hampering the development of adequate land use policy. In this study, we aimed to (i) determine the soil C stocks at different depths (0–10, 10–20, 20–40, 40–60 and 60–100 cm) in Cerrado woodland remnants in Minas Gerais (MG), southeastern Brazil, (ii) study the spatial distribution of soil C stocks using regression-kriging (RK), and (iii) provide a description of the soil C stock distribution and its relationship with variables that drive the soil C stock in the MG Cerrado. We designed a large soil inventory, in which soil samples were collected for soil Cc and ρb determination, to compute soil C stocks in the Cerrado biome of MG. Afterwards, we used RK to model the soil C stocks across the MG Cerrado to a depth of 1 m. Our results show that using RK to map C stocks across different soil layers in the MG Cerrado resulted in accurate estimates. The soil C stocks in the MG Cerrado tend to increase in the west-east direction, are positively correlated with altitude and silt + clay content, and are negatively correlated with the mean annual temperature and total annual precipitation. On average, the C stored at the 0–40 cm layer accounted for 56% of the total C stocked in the top 1 m of soil. Approximately 1.03 Pg of C is stored in the MG Cerrado soils at a depth of 1 m. The average soil C stock in the MG Cerrado is approximately 53% larger than the average indicated for the Brazilian Cerrado. Our results highlight the need for more detailed regional scale soil C inventories in the country to refine national C accounting. We believe our study can aid the development of adequate land use policy and could be used as a baseline for land use change studies in the MG Cerrado.