Index Scores and Condition Statements
We provide numerical index results and narrative condition values based on the applied biological index. Data also include results from our test of model applicability and lists any invasive species present. Specific fields are as follows:
| Column Name | Description |
|---|---|
| SampleID | Unique sample identifier in NAMC database |
| Site Name | Customer provided reach/sample name |
| Model name | Model used in calculations |
| Model type | OE, MMI, other |
| Fixed count | Subsampled total used in modeling |
| O value * | Sampled observed number of taxa |
| E value * | Predicted number of taxa from model |
| Index value, OE or MMI | Final score index score |
| Condition | Text condition statement |
| Model Applicability | Test of model applicability |
| ElevCat | Elevation of catchment from StreamCat used to determine model applicability |
| Precip8110Ws | Mean annual 30 average precipitation from StreamCat used to determine model applicability |
| TMean8110Ws | Mean annual 30 average air temperature from StreamCat used to determine model applicability |
| WsAreaSqKm | Watershed area (km2) from StreamCat used to determine model applicability |
| Invasive species | Presence of aquatic invasive species |
* These fields are absent from MMI reports.
Understanding O/E and MMI Indices
Stream and river bioassessments are based on comparisons of observed biota at sample sites with estimates of the biological potential at each site. Two common tools for conducting bioassessment of lotic systems are observed/expected (O/E) and Multi-Metric (MMI) indices. O/E indices compare the macroinvertebrate taxa observed at sample sites of unknown condition to the assemblages expected to occur in the absence of anthropogenic stressors (Hawkins et al. 2000, Hawkins 2006). In contrast, MMI indices aggregate multiple macroinvertebrate assemblage composition and structure metrics (e.g., total richness, proportion of tolerant individuals, combined richness of mayflies, stoneflies, and caddisflies) to assess biological condition (Stoddard et al. 2008). Metrics that differentiate between reference and degraded conditions are selected, rescaled to standardized scale (e.g., 0 to 100), and aggregated into a single measure of biological condition. Metrics are computed for sample sites of unknown condition and compared to metric values predicted to occur in the absence of anthropogenic activities (Hawkins et al. 2010a, Vander Laan and Hawkins 2014). Both O/E and MMI use empirical models built with data from a network of reference sites to predict conditions of sample sites in the absence of anthropogenic impacts.
O/E scores range from 0 to approximately 1, with a score of zero indicating that sample sites have no taxa in common with expected reference conditions. In contrast, an O/E score of 1 occurs when macroinvertebrate assemblages at sample sites are equal to those of reference conditions. For more detailed information, on how O/E indices are built see the predictive models page and individual index metadata pages. Biological condition of sample sites reported by NAMC is assessed based on benchmarks set by state regulatory agencies. If no state regulatory agency benchmarks exist, the benchmarks are based on the precision of the reference site model used to predict the expected number and type of taxa. Specifically, the standard deviation (SD) of predicted reference site O/E scores, with sample sites scoring less than one SD below the mean of reference sites having ‘minimal departure’ from reference; sites scoring between one SD and two SD having ‘moderate departure’; and sites scoring more than two SD below the mean of reference sites having ‘major departure’ from reference conditions.
MMI scores commonly range from 0 to 100, with lower scores indicating that sample sites significantly deviated from reference sites across all metrics. In contrast, higher MMI scores occur when all macroinvertebrate metrics are equal to or greater than those of reference conditions. MMI scores computed for sample sites are compared to reference site scores, with a percentile approach commonly used to assign different degrees of departure or condition categories (e.g., sample sites falling below the 5th percentile of reference MMI distribution are considered to have major departure). For more detailed information, on how MMIs are built see the predictive models page and individual index metadata pages. Biological condition of sample sites reported by NAMC is assessed based on benchmarks set by state regulatory agencies.
Considerations for Index Interpretation
Sample Sizes
Caution should be exercised when interpreting O/E or MMI scores to ensure sample or laboratory error does not have an undue influence on final scores. For example, taxonomic richness of macroinvertebrate assemblages typically increases asymptotically with the number of individuals in a sample (Vinson and Hawkins 1996). Therefore, indices commonly recommend a minimal number of individuals per sample to minimize this sample artifact (e.g., 200 individuals). Low numbers of individuals can result from sampling and/or laboratory processing errors but can also be a signal of degraded biological condition. For samples with low counts, additional samples should be collected to verify the precision of ‘major’ or ‘moderate’ departures from reference.
Model Applicability
Additionally, care should be taken to ensure the environmental conditions of sample sites are like those of the reference conditions used to developed biological indices. To assist with this, NAMC includes a “ModelApplicability” column which determines whether the sample site's environmental conditions are within the range of experience of the model. A “fail” indicates the model had to extrapolate, rather than interpolate when making predictions. Fails indicate that the stream was an outlier (90th percentile) in environmental space compared to the reference sites used to build the index. O/E or MMI scores and condition ratings should be interpreted cautiously if a site failed the test for range of experience of the model. Specifically, if a site fails, the following environmental gradients should be examined to see which of these contributed to the site being an outlier: watershed area, catchment elevation, 30-year (1981-2010) PRISM annual average precipitation for the watershed, 30-year (1981-2010) PRISM annual average air temperature for the watershed. Then the ecological implications of those environmental gradients for macroinvertebrates should be considered. Large, low, and dry stream end to have fewer macroinvertebrate taxa than smaller, higher, and wetter ones. Macroinvertebrate richness tends to increase with temperature up to a point before it declines in very warm streams. For example, if sites are larger, lower, and drier than reference sites used to develop the index, the model may overpredict the number of macroinvertebrate taxa for the site and lead to poor scores when these sites may be in good condition. On the other hand, reference sites for these larger streams are often lacking because these sites have been most impacted by anthropogenic disturbance. For assistance making these interpretations, contact us.
Nonperennial Streams
Caution should also be used when interpreting macroinvertebrate scores for samples with interrupted flow. Most current indices perform poorly in nonperennial streams. Nonperennial streams frequently have fewer taxa simply due to natural frequent and severe drying events. This may lead to samples falsely flagged as poor condition. If many of your sites were sampled during interrupted flow conditions or if you know many sites in the region to be nonperennial, you should reach out to NAMC for assistance interpreting your data. An alternative index developed using sites across flow permanence gradients may be available and appropriate to use depending on your monitoring questions.
How does NAMC determine Model Applicability?
We developed a nearest-neighbor based approach to identify sites with environmental characteristics that were outside of the environmental space defined by the reference sites. For all indices, we used four environmental attributes that broadly characterize primary environmental gradients in macroinvertebrate assemblages: 1) watershed area, 2) catchment elevation, 3) 30-year (1981-2010) PRISM annual average precipitation for the watershed, 4) 30-year (1981-2010) PRISM annual average air temperature for the watershed. We log transformed precipitation and watershed area to normalized distributions. Then we standardized all four representative variables by scaling between the minimum and maximum values observed in the reference data. We used the standardized variables to calculate Euclidean multivariate distances between each reference site and all other reference sites. We then calculated the average distance of each reference site to the 10 nearest other reference sites and used the 90th percentile of this distribution as a threshold for defining if a new site was out-side of reference site environmental space. To apply this test to new sites, we calculated the average distance of each new site to the 10 nearest reference sites and flagged a new site as an outlier if the average distance exceeded the 90th percentile threshold defined by the distribution of 10-nearest neighbor reference site distances. Boxplot distributions of the 4 environmental gradients above at reference sites used to build the model can then be compared with individual sites environmental gradients to determine why sites were considered outliers. These boxplots are provided on each individual index page.