Aridity, soil and biome stability influence plant ecoregions in the Atlantic Forest, a biodiversity hotspot in South America

Abstract

Our aims were to quantify and map the plant ecoregions of the Atlantic Forest, a biodiversity hotspot that covers ca 150 million ha in eastern South America. We used a data set on the distribution of 4378 shrub and tree species across 711 localities. Plant ecoregions were identified using analyses of species turnover for both species occurrences and relative abundances. We interpolated NMDS axes of compositional variation over the entire the Atlantic Forest extent, and then classified the compositional dissimilarity according to the number of biogeographical ecoregions previously identified through K-means analyses. We assessed the ability of environmental, historical vegetation stability and the current human footprint to explain the occurrence of the identified ecoregions through multinomial logistic regression models. We identified 21 spatially cohesive occurrence and 14 abundance ecoregions. Aridity, soil and historical biome stability were retained in the best model explaining both occurrence and abundance ecoregions. Broad compositional zones were identified through UPGMA cluster analysis of ecoregions, and formed north and south compositional blocks. Our work confirms the existence of a broad north-south divide within the Atlantic Forest, previously suggested based on climatic and amphibious data. Differences between the occurrence and abundance maps suggest the location of transition zones to neighbouring domains and endemism centres. Due to the aggregate nature of our analyses, site-level disturbance degree was not considered, implying that human impacts could be broader then we could detect. There was limited overlap between our results and previous Atlantic Forest regionalization efforts, indicating that multi-taxa, physiognomic and environmental regionalization schemes based on expert opinion or vegetation maps are poor proxies for compositional ecoregions.

Woody plant subregions of the Amazon forest

“…In contrast to that has been found for a variety of tropical vegetation types and floristic assemblages (e.g. Cantidio & Souza, 2019;Castro-Insua et al, 2018;Linder et al, 2012;Marques et al, 2020;Saiter et al, 2016;Tuomisto et al, 2019), we found rather smooth borders between subregions. This is attributable to the absence of extensive mountain ranges or steep climatic gradients in most of the Amazon extension, as well as to the historical stability of the tropical forest biome in the basin (Costa et al, 2017), which may have allowed for extensive species dispersal.…”

“…We also investigated the relationships of the identified subregions with environmental, historical and human correlates. Our results contribute to the global effort of mapping terrestrial subregions (Cantidio & Souza, 2019;Dinerstein et al, 2017;Moura et al, 2016;Olson et al, 2001;Saiter et al, 2016;Silva & Souza, 2018a) and for conservation planning in the Amazon domain.…”

“…Therefore, there is a clear need to use of a richness-free dissimilarity index so that our bioregionalization do not reflect richness gradients and sampling biases instead of the patterns we are aiming to capture (Castro-Insua, Gómez-Rodríguez, & Baselga, 2018). The use of indices not affected by richness gradients like the Simpson index is now a clear recommendation (Castro-Insua et al, 2018) and an emerging consensus, and has been extensively used in biogeographical studies (Cantidio & Souza, 2019;Castro-Insua et al, 2018;Dapporto, Fattorini, Vodǎ, Dincǎ, & Vila, 2014;Kreft & Jetz, 2010;Linder et al, 2012;Miranda et al, 2018;Moura et al, 2016;Silva & Souza, 2018a). Despite the robustness of this dissimilarity index to richness differences, small samples yielding small richness values were almost certainly undersampled in a biodiversity-rich domain like the Amazon and could bias the analyses by being nested within more species-rich assemblages, producing artificially small dissimilarity values.…”

“…This means that the observed proportion of shared species between any pair of samples may be smaller than the true proportion of species shared between sites from which the samples came from, producing observed dissimilarity values lower than the real ones. However, it is important to emphasize that sampling variation is inevitable in geographical-scale studies that rely on heterogeneous knowledge assembled along decades or centuries (Banda-R et al, 2016;Cantidio & Souza, 2019;Miranda et al, 2018;Moura et al, 2016;Neves, Dexter, Pennington, Bueno, & Oliveira Filho, 2015;Neves et al, 2017;Oliveira-Filho & Fontes, 2000); and if we regarded it as impeditive to further studies, many opportunities to improve knowledge would be lost. Insufficient sampling is likely to be a permanent bias in all species-rich community and macroecological studies, since intensive sampling of local diversity (e.g.…”

“…Despite its global importance, and in spite of recent advances in the understanding of the distribution of its tree species and phylogenetic relatedness (Emilio et al, 2010;Oliveira & Nelson, 2001;Slik et al, 2018;ter Steege et al, 2000ter Steege et al, , 2013, a regionalization of the Amazon tree flora is still lacking. Clear and data-driven delimitation of subregions is a growing scientific enterprise (Cantidio & Souza, 2019;Moura, Argôlo, & Costa, 2016;Saiter et al, 2016;Silva & Souza, 2018a;Smith et al, 2018), with consequences for macroecological and biogeographical studies (Fine, 2015), to the practical identification of metacommunities (Leibold et al, 2004;Smith et al, 2018), to the understanding of animal distribution and evolution (Rueda, Rodríguez, & Hawkins, 2010), besides being a requisite for conservation planning and management (Dinerstein et al, 2017;Olson et al, 2001). Here we refer to subregions as land units containing distinct natural species assemblages delimited by their contemporary distribution (Olson et al, 2001).…”