Influence of seasonal vegetation dynamics on hydrological connectivity in tropical drylands

Seasonally dry forests in tropical regions show over 300% inter-annual biomass variability that directly affects the runoff and erosion dynamics. However, biomass fluctuation is mostly overlooked in hydrosedimentological analysis, including in connectivity analysis. The aim of this paper is to understand how the dryland vegetation seasonality in Brazilian drylands affects the potential runoff and sediment connectivity using the Index of Connectivity (stream and outlet targets). Two main analytical steps were used to identify the influence of dry forest biomass fluctuation on connectivity: Creation of vegetation scenarios based on the relationship between rainfall patterns and NDVI fluctuations (Landsat images); Identification of the effect of the vegetation scenarios on Index of Connectivity. The method was applied to a 90 km(2) watershed in NE Brazil, creating a daily vegetation classification using five vegetation scenarios related to rainfall parameters, with average NDVI values from 0.18 during very dry scenarios (<20 mm of antecedent rainfall) to 0.62 in very wet scenario (>500 mm of antecedent rainfall). The primary connectivity behaviour is controlled by a continuous connectivity decrease, reaching 32%, related to increase of humidity and vegetation biomass. At the same time, due to rainfall irregularity, high magnitude rainfall events can occur even during very dry scenarios, when the watershed shows very high potential connectivity. It indicates that connectivity in runoff-dominated regions is temporally variable due to the highly seasonal vegetation and variable incidence of intense rainstorms.