Soil organic carbon and molecular characterisation of soils and vegetation inputs along a Savannah-rainforest boundary in Central Guyana, South America

Abstract

Amazonian soils have been estimated to contain a globally substantial 66.9 Pg C within 1 m depth. Current uncertainty in model projections for future climate scenarios emphasises the need to better understand soil and vegetation carbon stocks which may become significant sources of CO2 and CH4. Contemporary data of bulk and molecular carbon stocks for full soil profiles and corresponding above ground inputs is needed to understand how these stocks may alter with climate change. The savannah-rainforest boundary is particularly sensitive to alteration in response to these local climatic changes and is thus a focal point of international research. The study site in Central Guyana, which lies within the north eastern Amazon, encompasses pristine and relatively unexplored savannah-rainforest boundary, providing an advantageous location for assessing both soil and vegetation carbon. Soil profiles classified as gleysols (FAO) under rainforest have the greatest soil organic carbon (SOC) stocks of those studied, and are 43% greater than previously published data for tropical regions (up to 1 m depth). Further, estimations of the full soil profile SOC stocks show a 94% increase compared to previous 1 m depth data. Although not inclusive of the whole boundary region, these SOC stocks emphasise the significance of local responses to more extreme weather conditions induced by climate change. Molecular surface SOC characteristics are site specific: likely influenced by local water table depth, mineralogy, vegetation inputs and microbial activity. However, measured environmental variables (pH and water content) show no relationship to molecular characteristics. Gleysols have the most degraded lignin and carbohydrates, indicating high inputs and a faster turnover than the bulk SOC. Drier savannah woodland plinthosols have the greatest amounts of lignin, tannin and carbohydrates, reflecting high inputs. Despite this, this soil has significantly lower SOC stocks than gleysols. If local weather patterns alter towards postulated longer and more intense dry seasons, rainforest die-back may occur. With savannah encroachment, the release of SOC stocks from the swamp forest and forest island gleysols is likely to occur. Phenol-rich soil organic matter may preside in developing areas of savannah woodland, but nevertheless a net decrease in SOC stocks is likely to result. The data collected here can be used to inform management policies and practices to help conserve and monitor the significant stocks of SOC in the swamp forests and forest island on these boundaries.