Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 11 (1), 491

Dispersion Fields Reveal the Compositional Structure of South American Vertebrate Assemblages


Dispersion Fields Reveal the Compositional Structure of South American Vertebrate Assemblages

Michael K Borregaard et al. Nat Commun.


The causes of continental patterns in species richness continue to spur heated discussion. Hypotheses based on ambient energy have dominated the debate, but are increasingly being challenged by hypotheses that model richness as the overlap of species ranges, ultimately controlled by continental range dynamics of individual species. At the heart of this controversy lies the question of whether species richness of individual grid cells is controlled by local factors, or reflects larger-scale spatial patterns in the turnover of species' ranges. Here, we develop a new approach based on assemblage dispersion fields, formed by overlaying the geographic ranges of all species co-occurring in a grid cell. We created dispersion fields for all tetrapods of South America, and characterized the orientation and shape of dispersion fields as a vector field. The resulting maps demonstrate the existence of macro-structures in the turnover of biotic similarity at continental scale that are congruent among vertebrate classes. These structures underline the importance of continental-scale processes for species richness in individual assemblages.

Conflict of interest statement

The authors declare no competing interests.


Fig. 1
Fig. 1. Species richness and environmental variables in South America mapped in 1° × 1° grid cells.
a Species richness of birds (n = 2869), b mammals (n = 1146), and c amphibians (n = 2265). d Mean annual temperature (°C). e Annual precipitation (mm). f Predicted richness of birds based on a linear model of temperature and precipitation (R2 = 0.63).
Fig. 2
Fig. 2. Asymmetry vectors and assemblage dispersion fields (ADFs) for selected 1° × 1° grid cells in Amazonia.
The focal cell is marked by a black point. Colors indicate the number of species shared with the focal cell. The central region (see “Methods”) of the dispersion field is outlined by a black line. The asymmetry vector extends from the center of the focal cell to the center of gravity (white point) of the central region. ADFs of focal cells are relatively symmetrical near the center of the Amazonian ecoregion (ac) and become increasingly asymmetrical near its periphery (df), for birds (a, d), mammals (b, e), and amphibians (c, f).
Fig. 3
Fig. 3. ADF symmetry diagrams for South American birds, mammals, and amphibians and major vegetation biomes.
Arrows indicate the major axis of asymmetry of the ADFs for individual cells (n = 1689). The degree of asymmetry is indicated by a color scale (green through red colors represent more asymmetrical ADFs). This figure summarizes the vector arrows shown in Fig. 2 for all grid cells on the continent. Histograms show the distribution of vector lengths. The panels show a birds, b mammals, c amphibians, and d vegetation ecoregions. ADF symmetry diagrams using different cutoff values for the central ADF region are shown in Supplementary Figs. 3 and 5.). The map in d was provided by Tiina Särkinen, originally based on the map of the World’s ecoregions developed by the WWF (ref. ; under CC BY 4.0 license:
Fig. 4
Fig. 4. Null model divergence of ADF vector symmetry.
a The deviation of the ADF symmetry vectors of birds from the null model (Supplementary Fig. 7), calculated as the standardized effect size (SES) deviation of vector lengths after 1000 repetitions of the null model. b The relationship between the deviation from the null model and the distance from the ecoregion boundary for the subset of grid cells (n = 342) that are completely encompassed within Amazonia (Supplementary Fig. 9; R2 = 0.52). The least square regression explains 52% of the variation.
Fig. 5
Fig. 5. Predicted assemblage dispersion field (ADF) symmetry diagrams based on predictive models.
The arrows reveal the pattern of biotic similarity. a The predicted ADF diagram from a model where species originate in grid cells according to local energy availability and then spread spatially randomly from there; b the predicted ADF diagram from a model where species richness is constrained to be identical to the empirical and ranges are constrained to be spatially cohesive; c the predicted ADF diagram from a model where each species originate in its original vegetation ecoregion and spreads spatially randomly with the constraint that ecoregion boundaries are crossed with low probability. Vegetation ecoregions are defined by the map shown in Fig. 3d.

Similar articles

See all similar articles


    1. Klopfer PH, MacArthur RH. On the causes of tropical species diversity: niche overlap. Am. Nat. 1961;95:223–226. doi: 10.1086/282179. - DOI
    1. Simpson GG. Species density of North American recent mammals. Syst. Zool. 1964;13:57–73. doi: 10.2307/2411825. - DOI
    1. Pianka ER. Latitudinal gradients in species diversity: a review of concepts. Am. Nat. 1966;100:33–46. doi: 10.1086/282398. - DOI
    1. Currie DJ. Energy and large-scale patterns of animal-and plant-species richness. Am. Nat. 1991;137:27–49. doi: 10.1086/285144. - DOI
    1. Rahbek C, Graves GR. Multiscale assessment of patterns of avian species richness. Proc. Natl Acad. Sci. 2001;98:4534–4539. doi: 10.1073/pnas.071034898. - DOI - PMC - PubMed

Publication types