Regional invertebrate cross- and within-taxon surrogacy are scale and taxon dependent

Abstract

This study measured within- and cross-taxon congruence in the diversity of epigeal invertebrates (spiders, beetles and millipedes) and woody vegetation sampled at small spatial grains (0.25 ha) across a large area (30 000 km²) within the Vhembe Biosphere Reserve (VBR) in South Africa. Beetle, spider, millipede and woody plant diversity was recorded at 160 point localities in 20 sites stratified across the dominant vegetation types of the VBR. Surrogacy relationships were explored using multiple linear regression (species richness) and Mantel tests (composition), while complementarity was analysed using the species accumulation index (SAI). Very little (< 10%) of the variation in invertebrate species richness was explained by woody vegetation richness alone, but the relationship improved when vegetation type was added to the regression, especially for beetles. Woody vegetation assemblages showed a positive but weak congruence with beetle assemblages, especially Tenebrionidae. Woody vegetation assemblage showed poor congruence for spider communities in general, with the exception of Lycosidae. Although cross-taxonomic congruence was observed between woody vegetation and invertebrate taxa, the relations based on SAI were stronger than expected for Coleoptera (Carabidae, Scarabaeidae and Tenebrionidae), positive but weak for spiders (all families), and weak for millipedes which had several localised endemics. Tests of higher taxonomic categories as surrogates were shown to have much greater potential than cross-taxon surrogacy. Genera in particular are excellent surrogates for species. Tribal- or generic-level determinations can be a cost-efficient approach for regional conservation planning exercises which aim to represent smaller scale variations in invertebrate diversity.

Keywords:

• surrogacy
• woody vegetation
• invertebrates
• savanna
• Vhembe Biosphere Reserve

ACKNOWLEDGEMENTS

We would like to thank all landowners, provincial conservation authorities and community leaders on whose properties we sampled. We would also like to thank Ms Ruth Muller and Dr Adrian Davis for identification of beetle specimens, as well as the Research and Publications Committee (University of Venda) for financial support.

DISCLOSURE STATEMENT

No potential conflict of interest was reported by the author(s).

ORCID

Colin S. Schoeman http://orcid.org/0000-0001-8585-5857

Norbert Hahn http://orcid.org/0000-0002-1186-7713

Michelle Hamer http://orcid.org/0000-0002-8650-890X

Stefan H. Foord http://orcid.org/0000-0002-9195-2562