225-1 Spatiotemporal Patterns of Fire, Precipitation, and Vegetation Response across the Wetland-Savanna Mosaic of Brazil’s Pantanal
Session: Critical Zone Science: Intersection of Processes Linked to Geomorphology, Ecology, Fire and Climate (Posters)
Poster Booth No.: 220
Presenting Author:
Erica ScarpittiAuthors:
Scarpitti, Erica Ann1, Goodbred, Steve2, Bennartz, Ralf3, Jorge, Malu4(1) Earth and Environmental Sciences, Vanderbilt University, Nashville, TN, USA, (2) Earth and Environmental Sciences, Vanderbilt University, Nashville, TN, USA, (3) Earth and Environmental Sciences, Vanderbilt University, Nashville, TN, USA, (4) Earth and Environmental Sciences, Vanderbilt University, Nashville, TN, USA,
Abstract:
The Pantanal is the world’s largest freshwater wetland and shaped by a composite of several alluvial fans and lowland river systems. Its low-gradient terrain supports a diverse vegetation mosaic that is controlled by seasonal flooding, drought, and fire activity. The frequency and intensity of these cycles, along with local hydrology, determines how the Pantanal’s diverse ecosystems are distributed across the landscape. Increasing climate extremes and land-use changes in recent decades raise questions about how these pressures impact the maintenance and recovery of flood- and fire-adapted ecosystems across the Pantanal.
We analyzed 23 years (2001–2023) of monthly remote-sensing data to quantify vegetation recovery across five land-cover types in the Pantanal: forest, savanna, grassland, wetland, and pasture. We evaluated monthly NDVI (MOD13A1), fire activity (MCD64A1), and precipitation (CHIRPS) across land-cover boundaries defined by MapBiomas. After deseasonalizing each series, we applied Spearman’s rank and lagged correlations (1–6 months) to assess how rainfall and fire influence vegetation greenness (i.e., NDVI). At the biome scale, precipitation peaks in January and NDVI in March, followed by drought and fire activity that culminate in September. Across the Pantanal, rainfall most strongly enhances NDVI at a one-month lag (ρ = 0.38, p < 0.001), with effects fading by four months. In contrast, fire sharply suppresses NDVI at a one-month lag (ρ = –0.44, p < 0.001), with recovery typically within three months.
Across land-cover types, grasslands record the largest burned area annually and experience the largest post-fire NDVI declines (ρ = –0.47 at lag 1). Savannas show the most prolonged fire effects (ρ = –0.23 at lag 3), while forests exhibit intermediate sensitivity to both drivers. Wetlands sustain elevated fire activity into October and November, well after fires subside in other land-cover types. Notably, in 2020, forests burned as extensively as fire-adapted savannas, while in 2023, wetlands burned as extensively as fire-prone grasslands; both patterns not seen elsewhere in the time series. Our results reveal that vegetation responses to fire and precipitation vary in timing, intensity, and duration across land-cover types, underscoring the need to better understand disturbance–recovery dynamics in complex ecosystem mosaics.
Geological Society of America Abstracts with Program. Vol. 57, No. 6, 2025
doi: 10.1130/abs/2025AM-6871
© Copyright 2025 The Geological Society of America (GSA), all rights reserved.
Spatiotemporal Patterns of Fire, Precipitation, and Vegetation Response across the Wetland-Savanna Mosaic of Brazil’s Pantanal
Category
Topical Sessions
Description
Session Format: Poster
Presentation Date: 10/21/2025
Presentation Room: HBGCC, Hall 1
Poster Booth No.: 220
Author Availability: 3:30–5:30 p.m.
Back to Session