agriculture ͉ carbon ͉ land use change ͉ soybean T he ''arc of deforestation'' along the southern and eastern extent of the Brazilian Amazon is the most active land-use frontier in the world in terms of total forest loss (1) and intensity of fire activity (2). Historically, the dominant pattern of forest conversion has begun with small-scale exploration for timber or subsistence agriculture, followed by consolidation into largescale cattle ranching operations or abandonment to secondary forest (3-5). Recent expansion of large-scale mechanized agriculture at the forest frontier has introduced a potential new pathway for forest loss, generating debate over the contribution of cropland expansion to current deforestation dynamics (5-9). In the nine states of the Brazilian Legal Amazon, mechanized agriculture increased by 36,000 km 2 , † † and deforestation totaled 93,700 km 2 ‡ ‡ during [2001][2002][2003][2004]. Recent gains in the area under cultivation and the productivity of locally adapted crop varieties have made Brazil a leading worldwide producer of grains such as soybeans; the agribusiness sector now accounts for more than one-third of Brazil's gross national product (10).The state of Mato Grosso alone accounted for 87% of the increase in cropland area and 40% of new deforestation during this period. Whether cropland expansion contributes directly to deforestation activity or occurs only through the intensified use of previously deforested areas has important consequences for ecosystem services (11), such as carbon storage, and future deforestation dynamics.Amazon deforestation is Brazil's largest source of CO 2 emissions (12, 13). Carbon fluxes from deforestation are a function of the area of forest loss (14-16) and related forest disturbances, such as fire (17, 18) and logging (17,19), variations in forest biomass across the basin (20), and land use or abandonment after forest clearing (3,21). Land use after forest clearing remains a major source of uncertainty in the calculation of deforestation carbon fluxes because methods to assess deforestation trends in Amazonia have not followed individual clearings over time (4,5,(22)(23)(24)(25)(26)(27)(28). The relative contributions of smallholder agriculture and large-scale cattle ranching to annual forest loss have been inferred from the size of deforestation events (5, 28), but no direct measurements have been available. Rapid growth of large-scale agriculture in Amazonia challenges the historic relationship between land use and clearing size.We determine the fate of large deforestation events (Ͼ25 ha) during [2001][2002][2003][2004] in Mato Grosso State to provide satellitebased evidence for the relative contributions of cropland and pasture to increasing forest loss during this period (Fig. 1). Our approach combines satellite-derived deforestation data, vegetation phenology information from the Moderate Resolution Imaging Spectroradiometer (MODIS; ref. 29), and 2 years of field observations to establish the spatial and temporal patterns of land use after fo...

Tropical carbon emissions are largely derived from direct forest clearing processes. Yet, emissions from drought-induced forest fires are, usually, not included in national-level carbon emission inventories. Here we examine Brazilian Amazon drought impacts on fire incidence and associated forest fire carbon emissions over the period 2003–2015. We show that despite a 76% decline in deforestation rates over the past 13 years, fire incidence increased by 36% during the 2015 drought compared to the preceding 12 years. The 2015 drought had the largest ever ratio of active fire counts to deforestation, with active fires occurring over an area of 799,293 km2. Gross emissions from forest fires (989 ± 504 Tg CO2 year−1) alone are more than half as great as those from old-growth forest deforestation during drought years. We conclude that carbon emission inventories intended for accounting and developing policies need to take account of substantial forest fire emissions not associated to the deforestation process.

Amazonia as a carbon source linked to deforestation and climate change

Summary• Remote sensing data are a key tool to assess large forested areas, where limitations such as accessibility and lack of field measurements are prevalent. Here, we have analysed datasets from moderate resolution imaging spectroradiometer (MODIS) satellite measurements and field data to assess the impacts of the 2005 drought in Amazonia.• We combined vegetation indices (VI) and climatological variables to evaluate the spatiotemporal patterns associated with the 2005 drought, and explore the relationships between remotely-sensed indices and forest inventory data on tree mortality.• There were differences in results based on c4 and c5 MODIS products. C5 VI showed no spatial relationship with rainfall or aerosol optical depth; however, distinct regions responded significantly to the increased radiation in 2005. The increase in the Enhanced VI (EVI) during 2005 showed a significant positive relationship (P < 0.07) with the increase of tree mortality. By contrast, the normalized difference water index (NDWI) exhibited a significant negative relationship (P < 0.09) with tree mortality.• Previous studies have suggested that the increase in EVI during the 2005 drought was associated with a positive response of forest photosynthesis to changes in the radiation income. We discuss the evidence that this increase could be related to structural changes in the canopy.

International audienceIn tropical areas, Dynamic Global Vegetation Models (DGVMs) still have deficiencies in simulating the timing of vegetation phenology. To start addressing this problem, standard Fourier-based methods are applied to aerosol screened monthly remotely sensed phenology time series (Enhanced Vegetation Index, EVI) and two major driving factors of phenology: solar radiation and precipitation (for March 2000 through December 2006 over northern South America). At 1 × 1 km scale using, power (or variance) spectra on good quality aerosol screened time series, annual cycles in EVI are detected across 58.24% of the study area, the strongest (largest amplitude) occurring in the savanna. Terra Firme forest have weak but significant annual cycles in comparison with savannas because of the heterogeneity of vegetation and nonsynchronous phenological events within 1 × 1 km scale pixels. Significant annual cycles for radiation and precipitation account for 86% and 90% of the region, respectively, with different spatial patterns to phenology. Cross-spectral analysis was used to compare separately radiation with phenology/EVI, precipitation with phenology/EVI and radiation with precipitation. Overall the majority of the Terra Firme forest appears to have radiation as the driver of phenology (either radiation is in phase or leading phenology/EVI at the annual scale). These results are in agreement with previous research, although in Acre, central and eastern Peru and northern Bolivia there is a coexistence of 'in phase' precipitation over Terra Firme forest. In contrast in most areas of savanna precipitation appears to be a driver and savanna areas experiencing an inverse (antiphase) relationship between radiation and phenology is consistent with inhibited grassland growth due to soil moisture limitation. The resulting maps provide a better spatial understanding of phenology-driver relationships offering a bench mark to parameterize ecological models