Substrate heterogeneity and number of plant species in Everglades savannas (Florida,USA)

Environmental heterogeneity, especially that related to topography, has been proposed to influence numbers of plant species in different sized areas. Despite little variation in elevation, large numbers of vascular plant species occur in some habitats. This study explored possible relationships between number of plant species and substrateheterogeneity in two species-rich habitats, subtropical pine savannas and short-hydroperiod prairies, in the Long Pine Key region of Everglades National Park (Florida, U.S.A.). We examined relationships between numbers of vascular plant species and topographic heterogeneity by measuring numbers of species and elevations in differ-ent sizes of nested plots that spanned five orders of magnitude (0.1 m² to 1000 m² ) and that were located along two transects extending from pine savannas into short-hydroperiod prairies in different areas of Long Pine Key. We also classified substrates and soil depths in 1 m 2 sized submodules within the nested plots. Pine savannas occurred at higher elevations than adjacent short-hydroperiod prairies. Although differences occurred in substrate types and distribution within 1 m 2 plots, numbers of species were not associated with these differences. Vari-ances in elevations were similar in the smallest plots, but increased with area more rapidly in pine savannas than in short-hydroperiod prairies. Plot size explained about 85% of the variation in species numbers, which increased from 2040 per 1 m 2 to 80120 per 1000 m² . An interaction between habitat and scale explained 5% of the variation; more species occurred in shorthydroperiod prairies than pine savannas at scales <10 m² , but the re-verse occurred at scales >10 m². The number of species in pine savannas at scales of 1 m²and 10 m²was positively associated with variation in elevations; no significant relationships were obtained in short-hydroperiod prairies, which lack the fine-scale topographic variation of pine savannas. Our data indicate that substrate het-erogeneity, measured as variation in elevations, is not likely to be involved in the co-occurrence of many species within small areas of these savannas, but may influence numbers of species at larger scales of observation, es-pecially in pine savannas. Why many plant species occur within very small areas in these savannas remains un-answered.