Webs, mandibles and capsules - Is mapped vegetation type a surrogate for beetle and spider assemblages?

Increasingly, the effectiveness of surrogate species as a management tool for reservation of biodiversity has been questioned. It has been established that mammal and bird distributions correspond to vegetation type, but for invertebrate species this is less clear. This assumption was tested by comparing the communities of species of two invertebrate taxa in forest litter, spiders and beetles with pitfall sampling. Sampling occurred in spring summer and autumn in the foothills of Mount Wellington in 2002 and 2003 within 6 different adjacent eucalypt forest types - Eucalyptus regnans forest (WRE), E. obliqua with broadleaf shrubs (WOB), E. obliqua dry forest (DOB), E. tenuiramis forest on sediments (DTE), E. amygdalina forest on mudstone (DAM) and E. pulchella forest (DPU). The total number of beetles collected was 1726, representing 152 species from 28 families. Spiders totalled 1983 representing 204 species from 20 families. A third of these were juveniles and data were analysed separately with and without the juveniles. Forest type was a significant factor affecting distribution of spiders and beetles but was different for different forest types. There was a significantly different spider community in wet WOB while communities in WRE and dry DOB overlapped suggesting change along a continuum from wet to dry forest. Species responsible were vagrant hunters from the families Corrinidae, Gnaphosidae, Lycosidae, Zodariidae, Zoridae and a Micropholcommatidae web builder. Beetles were also significantly different between dry E. tenuiramis (DTO) and E. amygdalina (DAM) and a wet WRE-WOB-DOB continuum was detected. Species responsible for this separation were Isopteron obscurum (Erichson, 1842): Tenebrionidae, Tetrabothrus claviger (Fauvel, 1878): Staphylinidae and some fungivores - Nemadini (Leiodidae), Scaphidium sp.: Staphylinidae, Thalycrodes australe (Germar, 1848): Nitidulae, and Acrotrichis sp.: Ptilidae. A total of 56 soil, topographic, ground cover, microclimate and vegetation variables were measured. Their significance for predicting the distributions of spiders and beetles better than vegetation alone was examined. Statistical analysis revealed environmental gradients along which beetles and spiders were dispersed. Beetles were distributed along a moisture and a ground cover gradient. Spiders were separated along a soil nutrient and a moisture/temperature gradient. These gradients varied among sites in the same forest type as well as among sites in different forest types, and explained some of the site scale variation in assemblages. At a larger geographic scale, sampling at 36 sites grouped into 3 regions in southeastern Tasmania: Hobart, Levendale and Swansea, tested assemblage differences across a span of 157 km. Beta diversity was highest at the scale of 50 km. This is suggested as the maximum distance that should separate patches of the same forest type in order to capture maximum spatial variation in diversity of beetles and spiders across a vegetation based reserve mosaic. The research highlights the complexity of invertebrate interactions with forest type and environmental variables and indicates that simple prescriptions which can inform planning of reserves are not readily obtainable from examination of assemblages as a whole.