Hydrological cycle of a neotropical forest: carbon and nitrogen inputs and dynamics

Abstract

Forests are essential for maintaining water and nutrients availability, carbon stocks, and the global climate, and their preservation is crucial from a social, economic, and environmental viewpoint. Aspects related to the rainfall partitioning, and the nutrients cycling in forests are a field in constant development. In this context, this thesis aimed to evaluate the dynamics of carbon and nitrogen fluxes via rainfall, in an Atlantic Forest Fragment (Montana Seasonal Semideciduous Forest), in Lavras-MG. And to analyze the main biotic and abiotic factors that influence the stemflow carbon enrichment; and the soil carbon stock behavior up to 1-meter depth in the dry and wet seasons, identifying the factors that affect this storage. Samples were collected between May 2018 and April 2019 of gross rainfall (external) in 3 locations, and throughfall, and stemflow in 10 locations selected inside the forest fragment, considering the most abundant tree species, and the sampling spatial distribution. The physical and chemical variables evaluated included pH, electrical conductivity, total dissolved carbon, nitrate, nitrite, and total Kjeldahl nitrogen. To analyze the carbon concentration and calculate the carbon stock, the litterfall was collected, dried, and weighed monthly and the soil was sample in 7 depths (0-5 cm, 5-10 cm, 10-20 cm, 20-30 cm, 30-40 cm, 40-60 cm, 60-100 cm), in 4 months selected in the period (May, August, December 2018 and April 2019). The carbon concentration was analyzed in the litter and soil samples, corresponding to the selected months. As outcomes, it was identified that rainfall is enriched with carbon and nitrogen when crossing the forest; 75% of the carbon and nitrogen inputs in soil, via net precipitation, was in the wet season; and the same locations had the greatest inputs, regardless of the analyzed period. The stemflow carbon concentration was higher than in the throughfall and gross precipitation, and higher in the dry season than in the wet season. In addition, the carbon enrichment ratios were sensitive to the tree's structural features and to the meteorological conditions, mainly the bark structure, crown area, maximum rainfall intensity, and seasonality. Soil carbon concentration decreases with depth, and in the surface layer (0-10 cm) it was higher in the wet season. The multivariate analysis identified that the variables selected as significant for soil carbon storage were soil temperature, litter carbon content, hydraulic conductivity, and coefficient of variation in the diameter of trees in the plot. The correlation model was able to explain 80% of the carbon stored in the soil. Thus, the results obtained improve our understanding of nutrient deposition, leaching, and absorption processes by canopies and the importance of the tropical forest in the hydrological and nutrient cycle, with emphasis on the carbon cycle, which is closely related to climate change. Such understanding can support better management and conservation strategies for forests and watersheds.