Additional information:Use policyThe full-text may be used and/or reproduced, and given to third parties in any format or medium, without prior permission or charge, for personal research or study, educational, or not-for-pro t purposes provided that:• a full bibliographic reference is made to the original source • a link is made to the metadata record in DRO • the full-text is not changed in any way The full-text must not be sold in any format or medium without the formal permission of the copyright holders.Please consult the full DRO policy for further details. To test whether tropical habitat groups across the world can be differentiated by using 25 taxon-free mammalian community structures and to discuss the implications of this 26 analysis for palaeoecological community studies. 27 28
Materials and Methods 29We used mammalian community data for 169 localities, which were assigned a priori to 30 hierarchical Olson (1983) vegetation categories. Species over 500 g were classified into 31 dietary, locomotion, and body mass groups and the resulting group structures were 32 analysed using community structure analyses (NPMANOVA, CAP, SIMPER). 33
34
Results
35The test results show that the mammalian community structures are significantly different 36 between all of Olson's categories. These differences are highest at Olson's major and 37 minor ecosystem levels, and require the least number of variable categories. At the 38 vegetation level, the number of variable categories required to distinguish between them 39 becomes higher. Of the dietary groups, the number of frugivore-granivores, frugivore-40 omnivores, grazers and mixed feeders contribute most to these differences, while the 41 number of arboreal, arboreal-terrestrial and subterranean-terrestrial species are the key 42 locomotor groups. Body mass was not a good discriminator. 43
44
Main conclusions 45As general ecosystem categories are broken down into more precisely defined habitats, it 46 3 requires more detailed knowledge of the species adaptations to distinguish between them. 47Many of Olson's vegetation groups represent a continuum of cover that are, at least at the 48 worldwide comparison, too detailed to differentiate when broad generalities are sought. We 49 suggest using three worldwide tropical major ecosystems in mammalian community 50 structure analyses: "Humid, closed forests", "Seasonal or interrupted forests and 51 grasslands", and "Seasonal, open drylands". Our results also demonstrate that community 52 structures defined by both dietary and locomotor adaptations are powerful discriminators 53 of tropical ecosystems and habitats across the continents we examined, but body mass 54 should be interpreted with caution when the research question pertains to multiple 55 continents.