Carbon emissions from land-use changes in tropical dry forest systems are poorly understood, although they are likely globally significant. The South American Chaco has recently emerged as a hot spot of agricultural expansion and intensification, as cattle ranching and soybean cultivation expand into forests, and as soybean cultivation replaces grazing lands. Still, our knowledge of the rates and spatial patterns of these land-use changes and how they affected carbon emissions remains partial. We used the Landsat satellite image archive to reconstruct land-use change over the past 30 years and applied a carbon bookkeeping model to quantify how these changes affected carbon budgets. Between 1985 and 2013, more than 142 000 km of the Chaco's forests, equaling 20% of all forest, was replaced by croplands (38.9%) or grazing lands (61.1%). Of those grazing lands that existed in 1985, about 40% were subsequently converted to cropland. These land-use changes resulted in substantial carbon emissions, totaling 824 Tg C between 1985 and 2013, and 46.2 Tg C for 2013 alone. The majority of these emissions came from forest-to-grazing-land conversions (68%), but post-deforestation land-use change triggered an additional 52.6 Tg C. Although tropical dry forests are less carbon-dense than moist tropical forests, carbon emissions from land-use change in the Chaco were similar in magnitude to those from other major tropical deforestation frontiers. Our study thus highlights the urgent need for an improved monitoring of the often overlooked tropical dry forests and savannas, and more broadly speaking the value of the Landsat image archive for quantifying carbon fluxes from land change.

No abstract

No abstract

Glossy privet (Ligustrum lucidum) is a tree native to China that successfully invades forests of central Argentina. To fully understand glossy privet's ecological effects on native forest, it is necessary to accurately map the distribution of glossy privet stands and the changes in biodiversity and forest structure of the invaded areas. The objectives of this paper were (1) to map the distribution of glossy privet stands in an area representative of the Sierras Chicas (Córdoba, Argentina) and (2) compare composition, structure and regeneration between glossy privet invaded stands and native forest stands. Using four Landsat TM images (October 2005, March, May and July 2006) we mapped the distribution of a glossy privet-dominated stand using a support vector machine, a non-parametric classifier. We recorded forest structure variables and tree diversity on 105 field plots. Glossy privet-dominated stands occupied 3,407 ha of the total forested land in the study area (27,758 ha), had an average of 33 glossy privet trees (dbh [ 2.5 cm) per plot and the cover of their shrub and herb strata was substantially reduced compared with native forest. Forest regeneration was dominated by glossy privet in native forest stands adjacent to glossy privet-dominated stands. We conclude that in the Sierras Chicas glossy privet has become a widespread invader, changing the patterns of vertical structure, diversity, and regeneration in native forests.