Crossdating and the challenges in tropical dendrochronology: perspectives from a 10-year effort and seven site collections in eastern Amazonia

Tropical dendrochronology presents many challenges. Few tree species develop reliable annual growth rings that can be accurately dated to the calendar year. Accessing primary forests and old-growth trees requires significant labor and investment. Researchers have difficulty finding the trees, resulting in limited sample size, one of the key factors for the development of successful tropical tree-ring chronologies. Skepticism exists regarding whether Douglas's method should be the sole approach for dating tropical trees. However, cross-dating—finding a common growth pattern across a large area—remains the only way to accurately assign calendar years to growth rings. This method is essential for developing centuries-long tree-ring chronologies. Our studies demonstrate that Douglas's method works in tropical dendrochronology. Once trees are correctly dated, other methods can be applied, such as quantitative wood anatomy, isotopes, wood density, and radiocarbon dating. This study describes a nearly 10-year effort to construct a network of tropical tree-ring chronologies in eastern Amazonia. It includes the key challenges that prevented dating trees at some of the tropical sites visited. Samples of 342 trees of the species Cedrela odorata, distributed across seven locations, were collected from living and legally harvested trees in forests of eastern Amazonia. Three tree-ring width chronologies have been successfully dated, including a new tree-ring width chronology from Cedrela, in the Altamira National Forest, dated from 1885 to 2016. Provisional chronologies of tree-rings from Cedrela are presented here: (1) a 190-year record from Inupuku and a (2) 328-year chronology from Mukuru. Both sites are located in the Jari River valley, home to the tallest trees ever discovered in the Amazon basin. A third 113-year record from the Monte Alegre site, located in the Rio Paru State Forest. Our results demonstrate the influence of local physical and topographical soil attributes, in terms of their moisture retention capacity, on the successful development of tree-ring chronologies in some locations. Stand and gap dynamics, as well as sample size, also play an important role in whether trees can be dated or not. Despite these challenges, our efforts show that crossdating is possible in primary tropical forests, and the advantage of having precisely dated trees is the ability to learn about climate variability over the past centuries in the vast and largely unknown Amazonian territory.