O'Neill et al. 2010

O’Neill KP, Godwin HW, Jiménez-Esquilín AE, Battigelli JP. 2010. Reducing the dimensionality of soil microinvertebrate community datasets using Indicator Species Analysis: Implications for ecosystem monitoring and soil management. Soil Biology & Biochemistry 42: 145-154.

These authors used a dataset of soil microarthropods to evaluate a method for identifying indicator species for ecosystem monitoring. The method centres on the Indicator Value (IV) of a species, a number that integrates the degree of uniqueness to a place of a species and the abundance of that species within a given habitat. A high IV value indicates both high information content and a high probability of being sampled. The IV is apparently robust to differences in site number and species absolute abundances, and provides a single value for evaluating observed or expected changes in an ecosystem. Indicator species, furthermore, integrate habitat conditions over their lifespans, in contrast to measures of chemical and physical parameters that capture a snapshot of an ecosystem.

The basic evaluation approach here was to identify indicator species along a clear environmental gradient from meadow to forest in West Virginia. The habitat was divided into three zones, with an edge patch between the open meadows and closed-canopy forest. Near-surface soil cores were collected from each zone every month from April 2004 to April 2005 (n = 180), using the top of the mineral soil as the reference depth. Microarthropods were extracted in a modified Macfadyen funnel with a strong and increasing temperature gradient, into 70% ethanol.

Diversity measures, including Simpson’s and Shannon indices, were based on counts of individuals identified to family level (suborder for Acari). Differences between sites were analyzed by 2-way repeated measures ANOVA and Principle Components Analysis, after rare taxa (those that occurred in less than 10% of samples) were removed; rare taxa are extremely unlikely to be identified as indicator species.

Calculating IV for each taxon, regardless of the taxonomic resolution, provides large advantages in labour time and taxonomic expertise, as many microfauna are very difficult to identify to genus or species. These authors state that enumeration of a single sample required more than 1 hour of a trained taxonomist’s time. In studies such as this one, there are further advantages of IV associated with its robustness in the face of many zero measurements (i.e. taxa absent from samples) and the general messiness of these kinds of datasets. However, the ISA approach is intended for 2-stage studies, where an intensive initial survey identifies indicator species (taxa), and later long-term monitoring ignores other species. For studies specificially designed to address biodiversity, such as my own, excluding rare taxa would not be beneficial, and there may be no easy escape from time-consuming morphotaxa sorting.

I have spoken with Dr. Battigelli, the trained taxonomist in this study. He has indicated that while this IV-based approach may not be appropriate for my own work, it nonetheless demonstrates the types of analyses that can be conducted with soil invertebrates identified to middle taxonomic levels. He has assured me I could probably be trained to identify Collembola to Family and Acari to Suborder in a matter of a few days, and he would be interested in futher studies of collected soil invertebrates based on interesting patterns that emerge at these taxonomic levels.