Deciding on an appropriate scale for conservation activities: partitioning alpha and beta ant diversity in North-West Province, South Africa (Formicidae: Hymenoptera)

Abstract

1.	Hierarchical partitioning of diversity is an effective tool in determining at what spatial scales diversity is generated.
2.	This has important implications for managing diversity as we may wish to determine the appropriate scale for conservation activities.
3.	We studied whether ant assemblages are compositionally distinct between different vegetation units and at which hierarchical scale (regional vs local) diversity is being generated. We did this in an attempt to determine at what spatial scales land use management and conservation activities for ants (and therefore invertebrates) should take place.
4.	We sampled ants in 4 vegetation units in North West Province, South Africa, as part of a research expedition for the South African National Survey of Arachnida (SANSA); 92 ant species (Formicidae) were sampled in 25 genera within 6 subfamilies.
5.	Partitioning of diversity showed that alpha diversity was high for points (sampling units) and extremely high for sites, but both were not significantly different from the random model. Although small relative to alpha diversity, beta diversity among points contributed significantly more than expected to overall diversity whereas turnover among vegetation units was significantly lower than expected from the null model.
6.	These results indicate that turnover between points within vegetation units is relatively high; turnover at a regional scale was however very low, supported by the few number of species that were site specific. The lack of any spatial structure within sites might suggest that the studies’ grain was too fine or that ant assemblages in a vegetation unit are homogenous in space.
7.	The high turnover observed in the study between points may be exactly that, and if so may well be attributed to habitat quality and ant dispersal mechanisms – supported by other studies. It may, however, simply be an artefact of incomplete sampling. The low beta diversity observed between vegetation types may be attributed to the fact that ants are extremely good dispersers.
8.	Different colony founding mechanisms could explain why some species contributed more to the turnover between points.
9.	Our results indicate that management activities for conservation of ants are sufficient at scales as fine as 400 m×400 m, yet we caution that small conservation areas should be nested within larger reserves in order to maintain landscape connectedness.

Keywords:

• savanna biome
• ants
• Formicidae
• vegetation units
• alpha and beta diversity
• partitioning
• conservation management

ACKNOWLEDGEMENTS

We would like to thank Hermien Roux, from the North West Province Department of Nature Conservation for help with site selection, logistical support and accommodation while sampling for this study, and Norbert Hahn for help with fieldwork. Financial support came from South African National Survey of Arachnida and University of Venda's Research and Publications Committee.