Tree species on environmental gradients in subtropical forest of northwest Belize, in the Maya Lowlands. (April 2023)

• Record Type: Journal Article
• Authors: Ward, Sheila E; Brokaw, Nicholas; Walling, Stanley; Cortes-Rincon, Marisol
• Is Part Of: Progress in physical geography. Volume 47:Number 2(2023)
• Page Start: 249
• Page End: 269
• Publication Date: 2023-04
• Subjects: subtropical forest -- tree species composition -- tree species distribution -- topography -- soil -- multivariate analysis -- indicator species -- Belize -- Maya Lowlands; Physical geography -- Periodicals; 910.02
• DOI: 10.1177/03091333221143643
• Languages: English
• ISSNs: 0309-1333

Abstract

We studied species composition and spatial distributions of tree species, and the underlying topography and soil, in subtropical forests of northwest Belize, a region in the Maya Lowlands. Our goal was to learn how much the spatial distributions of species vary and are predictable over the landscape. The study was done in old-growth, subtropical moist forest on limestone-derived topography and soil. We identified to species all trees ≥10 cm DBH in 209 400-m 2 plots. For each plot, we characterized topographic setting and analyzed soil nutrients and texture. We recorded 3, 984 individual trees of ∼140 tree species and used the 3, 775 individuals of the 69 species occurring in ≥5 plots in multivariate analyses, including Nonmetric Multidimensional Scaling (NMS). NMS showed that 73% of the variation in species composition per plot was associated with the first three ordination axes. Sixteen out of the 34 quantitative variables we measured were correlated at R 2 > 10% with the axes. Of the categorical variables, Topographic Class was strongly associated with species composition, and USDA Texture Class less so. Of the 69 focal tree species, the abundances of 21 were correlated at R 2 > 10% with one or more axes of the NMS ordination. Importantly, these 21 species accounted for 68% of all individual trees sampled in the 209 plots. Twenty-three species were indicators of particular topographic and soil classes. We conclude that patterns of tree species distribution are strongly and predictably associated with different topographic and soil conditions in this landscape. In the past, the ancient Maya could have used this type of predictable plant–soil relationship to optimize their agriculture. In the future, our results are a basis for predicting local shifts in tree species distributions due to climate change.