Climate signals and rainfall reconstruction using Cedrela tree-ring chronologies from the Eastern Amazon Basin

Abstract

Understanding the complexity of one of the world's largest convective centers, the Amazon basin, is a challenge given recent reports regarding the increase of extreme moist and dry events in the last decades. In addition to the mega-diversity of flora and fauna, the Amazon has a significant role in atmospheric circulation and in the global energy balance. However, the lack of long-term instrumental hydroclimatic data prevents this trend from being verified. In this sense, climate information derived from tree-ring analyses, a science known as dendroclimatology, has proved to be a good tool for climate reconstruction. This work is part of an international collaboration project between the Federal University of Lavras (UFLA), the University of Arkansas (UARK), the National Institute of Amazonian Research (INPA) and the Argentinean Institute of Nivology, Glaciology and Environmental Sciences (IANIGLA-CONICET ), which aims to apply tree-rings to study the past climate in the Amazon Basin. Wood samples from 519 trees were collected in the eastern portion of the basin thanks to collaboration agreements with legal logging companies. The target species were the Cedros (Cedrela fissilis and Cedrela odorata), which have proven dendroclimatic potential in the Neotropics due to the presence of distinct annual growth rings and mid-high sensitivity to moisture variability. The first chapter presents the first rainfall reconstruction for the equatorial Amazon using a tree-ring chronology built from well-dated Cedrela ring-width time-series from the Paru State Forest, located on the north bank of the Amazon River. The second chapter focuses on the efforts to develop a network of new tree-ring chronologies from different sites in the eastern Amazon with the validation of dating and climate/growth relationships obtained by the first chronology. The data presented here bring new contributions about the hydrological variability in the eastern portion of the basin that predates the existing instrumental data. Interannual and subdecadal frequencies dominate over the entire length of reconstructed data (1786-2016), although extreme episodes prior to anthropogenic changes were identified by growth rings. The resulting chronologies express the influences of the sea surface temperatures of the Atlantic and specially equatorial Pacific on rainfall variability over the eastern Amazon. These results show the potential of a Cedrela network of chronologies to reconstruct water variability in the eastern Amazon, providing a long-term perspective on the occurrence of hydroclimatic extremes and future predictions of what can be expected from this variability in the current scenario of local and global anthropogenic changes.