Spatial and environmental drivers of macrophyte diversity and community composition in temperate and tropical calcareous rivers

• We analysed the aquatic vegetation of 1354 tropical and temperate river sites.
• MEM, multiple regression and pRDA were used to analyse the datasets.
• Spatial and environmental variables were both significant driving factors.
• Species richness (S) in tropical rivers was only driven by spatial factors.
• In temperate rivers S was driven by both spatial and environmental variables.

The hypothesis was examined that sources of variation in macrophyte species richness (alpha-diversity: S) and community composition (“species-set”), attributable to spatial and environmental, variables, may differ in importance between tropical and temperate calcareous rivers (>10 mg CaCO3 L−1). To test this hypothesis geographic, environmental, and aquatic vegetation data was acquired for 1151 sites on calcareous rivers within the British Isles, supporting 106 macrophyte species (mean S: 3.1 species per sample), and 203 sites from Zambian calcareous rivers, supporting 255 macrophyte species (mean S: 8.3 species per sample). The data were analysed using an eigenfunction spatial analysis procedure, Moran’s Eigenvector Maps (MEM), to assess spatial variation of species richness and community composition at large regional scale (>105 km2: British Isles and Zambia); and at medium catchment scale (104–105 km2: British Isles only). Variation-partitioning was undertaken using multiple regression for species richness data, and partial redundancy analysis (pRDA) for community data. For the British Isles, spatial and environmental variables both significantly contributed to explaining variation in both species richness and community composition. In addition, a substantial amount of the variation in community composition, for the British Isles as a whole and for some RBUs, was accounted for by spatially-structured environmental variables. In Zambia, species richness was explained only by pure spatial variables, but environmental and spatially-structured environmental variables also explained a significant part of the variation for community composition. At medium-scale, in the British Isles, species richness was explained by spatial variables, and only for four of the six RBUs.