Index Metadata
We include macroinvertebrate index specific metadata below to assist practitioners with 1) choosing an appropriate index, and 2) understanding nuances of index precision, condition determination, and applicability of indices to a specific site. If you desire a full set of index metadata for your reporting needs, please cite either the original index documentation provided on the pages below or consult your index metadata provided within your report. General information about interpreting macroinvertebrate index scores can be found in index report metadata.
Index metadata includes:
- Developer and original application of the index
- Taxonomic resolution required to compute the index
- Sampling timeframe required to apply the index
- Map of reference sites used
- Boxplots of environmental gradients for reference sites
- Index precision and benchmarks
- Additional index details
- Method for accounting for natural environmental gradients – It is important to account for natural variability in assemblages prior to considering anthropogenic impacts. Methods for accounting for this variability include categorical ecoregional approaches, random forest, or discriminant function analysis.
- Predictors of reference assemblages or metrics – Each index uses different environmental gradients to account for natural variability in assemblages. We list all the geospatial variables used to predict natural variability among sites in macroinvertebrate assemblages or metrics. Sources of geospatial data use varies from public datasets such as PRISM to state specific layers that are only available from each state.
- # of Macroinvertebrate groups obtained by hierarchical cluster analysis – To develop O/E models, sites are grouped into similar clusters and then the probability of a given site being in a specific cluster is predicted using the environmental predictors above. The number of clusters depends on sample sizes and within vs. among site variability. The number of sites within a cluster should be >10 and influences the precision of probabilities of capture for taxa.
- Fixed count – number of individuals that are subsampled to account for differences in sampling effort among samples
- Minimum count – minimum number of individuals required for each state to accept the sample as a valid sample
- Operational Taxonomic Unit (OTU) applied – All O/E indices require standardize taxonomy across all reference sites to compute taxa specific probabilities of capture, and therefore all O/E indices use OTUs. Some MMI use OTUs but most standardize taxonomy within a sample to ensure accurate number of unique taxa.
- Rare taxa <0.5 probability of capture included – excluding rare taxa generally increases model precision
- Documentation – Original index documentation provided by each state