Climate, landscape and local hydro-topography shape Amazon tree communities

Multiscale JSDM trees and palms in French Guiana

OSF DOI Gitlab project Website

:::

Guillaume SALZET, Sylvain SCHMITT, Thomas GAQUIERE, Marion BOISSEAUX, Giacomo SELLAN, Stéphane GUITET, Gaëlle JAOUEN, Julien ENGEL, Jean-François MOLINO, Daniel SABATIER, Olivier BRUNAUX, Jérôme CHAVE, Géraldine DERROIRE, Christopher BARALOTO, Stephane TRAISSAC, Eric MARCON, and Mélaine AUBRY-KIENTZ, “ Climate, landscape and local hydro-topography shape Amazon tree communities ” (In prep.).

Abstract

  • The assembly of tropical tree communities is driven by environmental gradients operating at multiple spatial scales, from regional climate to local topography. Yet, the understanding of how these multi-scale environmental gradients jointly structure species’ assembly and the resulting community composition is challenging. We address this issue by analysing how regional climate, landscapes and local topography, structured across nested spatial scales, shape the distribution and coexistence of tree and palm species in the French Guiana forest.
  • We compiled 2,307 forest inventories spanning French Guiana, comprising 226,443 individual trees and palms. We used a hierarchical joint species distribution model with environmental predictors at the three nested spatial scales: regional climatic regimes, geomorphological landscape types and a local topographic wetness index. Specifically, we (1) quantified species’ environmental preferences at each scale, (2) tested for phylogenetic signals in these preferences and (3) used variance partitioning to quantify the contributions of each spatial scale to community composition.
  • 99.3% of the species studied exhibited a significant response to at least one environmental predictor, 84,5% to environmental predictors belonging to two scales and 31% to all three, which underline the multi-scale environmental contributions to species niche. Moreover, environmental responses showed significant phylogenetic signal at all scales, indicating climatic and geomorphic niche conservatisms at regional scales and confirmed a pervasive within-genus divergence along local topographic gradients. At the community level, regional climate, landscape and local topography exerted largely complementary effects, with more than half of the explained variance attributable only to landscape effects and only a small fraction shared among scales.
  • The different environmental factors promote spatial variations and thereby a mosaic of partially distinct habitats enhancing regional and local species coexistence. Our results highlight that analyses of Amazonian tree communities, as well as inventory and monitoring networks, should explicitly stratify sampling along climatic, landscape and local topographic gradients to capture the multiscale filters that structure diversity.
  • Synthesis: Our study demonstrates how species environmental preference is spatially structured, and highlights that geomorphological landscapes and local topographic gradients are key components of tree community assembly in French Guiana. Explicitly incorporating these spatial scales into monitoring strategies should improve our capacity to detect forest responses.