Fauna data should be presented based on fauna habitat type

Fauna assemblages vary appreciably among habitat types and there can be a significant loss of important information when survey data from multiple fauna habitats are combined and assessed as a single unit.

The characteristics that define a preferred habitat for one species can differ from a sympatric species. Therefore, minor variations within what we perceive to be a uniform fauna habitat may result in appreciable differences in fauna assemblages at a local-scale. Defining fauna habitats can be a difficult and complex process. For example, habitats might differ in their floristics but are structurally the same (e.g. 2 different dominant eucalypt species), they may have the same floristics but differ structurally (e.g. young and old forest), they may have the same structure and floristics but differ in the level of disturbance, or they may be the same site but surveyed in different seasons or different years (e.g. after drought and after rain).

To further illustrate this point, Thompson and Thompson (2008) surveyed what appeared to be five distinctly different terrestrial habitat types in the Goldfields based on soil and vegetation characteristics with four trapping sites located in each habitat type. Figure 1 shows the relative placement of the trapped fauna assemblages from 20 survey sites based on a principal component analysis (PCA). It might have been expected that given the distinctly different habitat types surveyed, the trapped fauna at each site would have clustered into five groups based on vegetation and soils characteristics. It can be seen from Figure 1 that fauna assemblages in three of the habitats are grouped (sand dune, spinifex sand plain and mulga) but the fauna assemblage in the chenopod shrub land was not clearly separated and there was an affinity between at least one of the mulga spinifex sites and the spinifex sand plain site.

Figure 1. Grouping of trapped fauna for 20 sites in five habitat types

More typically, fauna habitats are not as clearly defined as shown in Figure 1 and there is a graduation or continuum among habitat types. When this is the case the trapped fauna assemblages may not cluster into definitive groups. For example, 59 sites were surveyed in the Hamersley Range in the Pilbara over two seasons (Thompson et al. 2010). The PCA 1 (Figure 2) indicates a continuum in the change of these 59 fauna assemblages based on variations in substrate type from stony over a rocky underlying base at the negative end to a loamy substrate at the positive end. The five sites that have the highest positive values on PCA 2 (i.e. crosses) are from areas that were degraded and disturbed by cattle.

Figure 2. Grouping of fauna assemblages for 59 sites in the Pilbara

Biotic and non-biotic characteristics that we perceived to define fauna habitats may not reflect the characteristics that actually separate different fauna assemblages (i.e. we see the environment differently to the small vertebrate fauna). Therefore, it is suggested that when planning a survey, survey sites are distributed among the various ‘perceived’ fauna habitat types with sufficient sites selected such that if one or two have been incorrectly categorised, then there are sufficient data to adequately represent species richness and relative abundance of the fauna for each habitat type. Ordination should be used to group sites, and this grouping should be the basis of the analysis of the fauna survey data in the report.

Conclusions

Fauna survey data (e.g. species richness, relative abundance, species accumulation curves, etc) should be presented in reports based on fauna habitat types, as there could be a significant loss of important information when the data are combined across a project area that contains multiple habitat types.

Fauna habitats are bested determined based on an appropriate method of ordination of the trapped fauna assemblage.

References

Thompson, G. G. and S. A. Thompson. 2008. Spatial variability in terrestrial fauna surveys; a case study from the goldfields of Western Australia. Journal of the Royal Society of Western Australia 91:219-228.

Thompson, G. G., S. A. Thompson, and G. R. Finlayson. 2010. Spatial and temporal variations in the trapped terrestrial vertebrate fauna of the Hamersley Range, Western Australia. Journal of the Royal Society of Western Australia 93:51-64.