BIOLOGY - Stream Species Quality Index

How healthy are the assemblages of aquatic species in the streams of each watershed?

Why is this important for biology?

Stream Species Quality Health Scores

watershed health scores thumbnail

key for map

Click map to enlarge and explore Watershed Health Assessments.

Species assemblages are a reflection of their environment, landscape features, land use history, and connectivity among habitats. Landscape and aquatic alterations can degrade conditions that support aquatic species. The presence of expected macroinvertebrate and fish species indicates a healthy aquatic system, capable of supporting a diverse community of organisms. A reduced number of species indicates a decline in habitat condition, loss of connection between suitable habitats, or some other change in the aquatic ecosystem. A large number of species tolerant of altered habitat conditions also signals a decline in stream condition.










Input Data


Calculating watershed index


Calculating catchment index


Ranking and scoring

Overview of results


Interpreting the results


Relationship to other health components

Scientific literature support


Confidence in index

Future enhancements


This slideshow requires the latest version of Adobe Flash Player


Calculating Watershed Health Scores

The macroinvertebrate and fish “observed/expected” approach, known as the O/E ratio (Hawkins et al. 2000) was adopted at the major watershed scale. This approach is based on RIVPACS (River Invertebrate Prediction and Classification System) a statistical model used to predict the expected species at a site based on reference site conditions (Wright 1995). The O/E approach is widely used across the world (Armitage et al. 1983; Norris 1995; Wright 1995; Marchant et al. 1997; Moss et al. 1999) and more recently in the western United States (Hawkins et al. 2000). Species used in the calculations were from stream surveys conducted from the Indicators of Biotic Integrity (IBI) stream survey data of the Minnesota Pollution Control Agency (MPCA).

Site counts of observed fish and macroinvertebrate species are compared to expected species counts. The scores typically range from 0 (no expected species present) to 1 (all expected species present). Scores may exceed one when more species are observed than were expected at a site. Each site in each watershed was given an O/E score. If that score is +/- 1 standard deviation (SD) of values observed at reference sites, that site was considered as meeting the reference condition. If a site had score that was greater than +/- 1 SD from the reference condition, the site was considered in an altered condition. The percentage of sites found to be altered (greater than 1 SD from the expected condition) was used to generate the score.

For watersheds with fewer than 3 sample sites, O/E scores from adjacent watershed(s) in the same river network were averaged to create a larger sample size.

A third metric was calculated using the MN DNR Statewide Mussel Survey data from 1999-2010. The percent of species found live at each survey site was calculated by dividing the number of live species by the number of live species and dead shell species. This “percent live mussel species” value was calculated for each survey site. The average value for all survey sites in each watershed was then calculated.

The percent of scores meeting reference conditions for fish and for aquatic macroinvertebrates were averaged with the percent live mussel species into one value for a combined Stream Quality Index score. 


Ranking and Scoring Watersheds

The percentage of sites meeting reference conditions was used to directly rank and score the results.

The values for: Fish ranged from 22 to 100.
Aquatic invertebrates ranged from 22 to 100.
Live mussel species ranged from 13 to 100.
The combined index values ranged from 23 to 100.

The Stream Species Quality values were ranked by generating equal intervals between 0 and 100; where 100 indicates all expected species are found at a site, i.e., the O/E ratio was > or = 1 or the live/total species ratio was:

    0-20 = none to very few expected species present
  20-40 = few expected species present
  40-60 = some expected species present
  60-80 = many expected species present
80-100 = most to all expected species present


Calculating Catchment Health Scores

Three aquatic species metrics were calculated for survey sites and then applied to the catchment scale.  The macroinvertebrate and fish species data from the Index of Biotic Integrity (IBI) stream surveys of the Minnesota Pollution Control Agency (MPCA) were used for two of the metrics.  Each stream survey site is given an overall IBI "score" based on a large number of measured metrics.  The range of possible IBI scores are from 0 to 100.  For each site, there is a threshold value used to establish biological health or impairment. The IBI score and the threshold value were used to create a site score. 

Results from the DNR Statewide Mussel Survey were used to create a third survey site level input.   Four different population measurements were created and then combined to create the mussel survey site quality index.   


Fish and Aquatic Macroinvertebrate health scores: 

Site Quality Value:

The fish and invertebrate site IBI scores can range from 0 to 100.  This value was compared to the “threshold” value (range 35-51) for impairment for each site. 

If IBI score is higher than threshold, find relative score between threshold and 100:

(IBI Score - threshold) / (100 – threshold) /2 + 0.5

If IBI score is lower than threshold, find relative score between 0 and threshold, as a negative number:

(IBI score)* (1/ threshold)/2

These formulas provide scores between 0-100, whereby if IBI score meets or exceeds threshold values are 50-100, and if it meets or is below threshold, values are 0-50 (Thus a score of 50 is exactly meeting expectation)

Catchment value:

Catchments are assigned the mean of the IBI point scores that fall within their boundary.

Empty catchments that are neighbors to one or more catchments that contain point data have a score based on theaverage score of neighboring catchments that are hydrologically connected. (Note: not average score of all points in all neighboring catchments).  These values are shown as "extrapolated" values when viewed on the map and have a cross hatch to identify them. 


Mussel Health Scores:

Four Mussel Metrics were calculated for each survey site.

Catch per Unit Effort

Catch Per Unit Effort = Total Live Mussels per Site/Search Time per Site 

The maximum "catch per unit effort" is limited by number of mussels that a researcher can retrieve in one minute. This maxmimum value has been set to 25. The CPUE values of 0-25 were multiplied by 4 to create 0-100 scaled values for this metric.

CPUE x 4 = Scaled CPUE

Percent Live Species

Percent Live =  Count of species found live at site/Count of total species found at site

Species Recruitment

Species that have individuals in the age category 0-5 were considered to be recruiting young mussels.

Count of species recruiting/ Count of total live species = Percent Recruiting Species

Relative Tolerance

Step 1: The relative abundance of each species in the population at each site was calculated based on number of individuals found during the survey at that location:
    Count of individual mussels of each species/Total live individuals per site= Relative Abundance of each species

Step2:  Each species is given a tolerance rank in the range of 1-5
5    Intolerant (very sensitive)
4    Sensitive
3    Facultative/intermediate tolerance
2    Tolerant
1    Very Tolerant  (least sensitive)

Step 3: Relative tolerance ranking for each species was calculated

Tolerance Rank X Relative Abundance = Relative tolerance per species

Step 4:  Relative tolerance per site = Sum of relative tolerance of species found at that site.

Step 5:  Relative Tolerance values for a site range from 0-5.  In order to create a comparable metric of 0-100, this value was multiplied by 20. 

Relative Tolerance X 20 = Scaled Tolerance

Mussel Site Quality Score
The overall site quality score was created by combining the four metrics into a mean value.

(Scaled Relative Tolerance + Percent Recruitment + Percent Live + Scaled CPUE) / 4 = Overall Site Score


Ranking and scoring catchments

Catchments are used to calculate three different metric scores based on the mean value of sites that fall within their boundary. This generates three different metric scores (fish, invertebrate and mussel) that can be viewed at the catchment scale.

Empty catchments that are neighbors to one or more catchments that contain point data have a score based on the average score of neighboring catchments that are hydrologically connected; upstream or downstream on the same stream network. These values are used to create an additional layer with "extrapolated" values. When viewed on the map, the extrapolated values are identified with a cross hatch pattern.



 Overview of results

Major Watershed Scale:

The combined Stream Species Quality scores for watersheds ranged from 22 to 100; however, most scores indicated many expected species were present across the state, but the watersheds with most expected species were widely distributed across the state. There were interesting contrasts in the pattern among taxonomic groups. The lowest scores for fish were generally in watersheds on the western border of the state. The lowest scores for macroinvertebrates were in watersheds in north central to northwestern Minnesota. Finally, the lowest scores for mussels were in the southern third of the state.

Catchment Scale:

The fish IBI results show low scores scattered statewide, but most concentrated in the Minnesota River basin and the lower watersheds of the Upper Mississippi, upstream of the Twin Cities. The highest scores are found in north central Minnesota and along the eastern edge; the north shore, the St. Croix basin and the southeast blufflands.

The aquatic macro invertebrate results show low scores prevalent in the Red River Basin and the mid- Minnesota River Basin, particularly in the Granite Falls watershed. Many scores statewide fall near 50 showing that they are near the expected IBI condition. Slightly higher scores follow the highest fish scores, in the north central and eastern portion of the state.

The mussel results show low scores throughout the southern part of the state from the Missouri River Basin in extreme southwest, the southern tributaries to the Minnesota River basin, as well as the southern streams that flow into Iowa. Pockets of lower scores are also found in the Cannon and Root River watersheds and catchments near the Red River. Higher scores are seen in the Mississippi Headwaters, the north central lakes area and the headwaters of the Red River.


Interpretation of results

Overall, this index indicates that the observed number of stream species is near the number of species expected. However, each species group tells a somewhat different story. Macroinvertebrates and mussels are sensitive to channel modification, siltation, and changes in hydrology. Macroinvertebrates and mussels primarily live on or in the streambed, and they can move to a limited extent. Increased sediment inputs and siltation may extirpate local populations, and low dispersal allows these low numbers to persist. The lowest scores for mussels were found in watersheds with significant agricultural land use in the southern part of the state with some low scores in the Red River Valley. In contrast to mussels, the lowest scores for macroinvertebrates were distributed across the central part of the state. Fish have slightly lower scores across the middle of the state, including portions of the Upper Minnesota, Upper Mississippi, and Upper Red River drainages. The lowest scores for each species group are widely dispersed and when averaged across the state the scores indicate moderate to minor impairment. More investigation is needed to find the drivers for the disparity among species groups.


Relationship to other health components

Water quality

The introduction of contaminants, temperature changes, sediment, nutrients and other inputs into the aquatic environment have a direct impact on the ability of aquatic organisms to survive and thrive. The loss of more sensitive (intolerant) aquatic species is often the direct result of water quality degradation.


Many aquatic species need a variety of habitat types to complete their life cycle. Aquatic systems fragmented by dams, zones of poor water quality, levees, ditches, low flows, and other barriers to movement often eliminate access to essential habitats and result in the loss of aquatic species diversity.


The viability of aquatic species is closely related to changes in the hydrologic flow regime and runoff from the surrounding terrestrial landscape. Many species are sensitive to sediment loads, temperature fluctuations, and the timing and magnitude of discharge. Aquatic species assemblages have been found to be closely linked to hydrologic alteration. Diminished flows were the primary predictors of biological integrity for fish and macroinvertebrates communities at more than 200 sites across the United States and the likelihood of biological impairment was related to the severity of diminished streamflows.


 The geologic setting coupled with land cover alteration influence erosion rates, vulnerability to climate extremes, and groundwater contamination. These changes potentially impact the presence and viability of aquatic species.


Scientific literature support

This index is well supported, resting on a well-developed theoretical foundation, with strong theoretical support in the technical processes of generating the base data and index. There are a broad set of studies connecting changes in ecological, chemical, and physical characteristics of water bodies that are measured with the O/E approach, across the world (Armitage et al. 1983; Wright 1995; Norris 1995; Wright 1995; Marchant et al. 1997; Moss et al. 1999) and more recently used in the western United Sates (Hawkins et al. 2000).

Previous studies have found various relationships between the percent agricultural use and forest harvesting in a watershed and fish assemblages. IBI scores, an alternative assessment approach used by the MPCA, decline with an increase in the percent agricultural land (between 35 and 70%), especially in riparian areas (Roth et al. 1996, Wang et al. 1997, Lyons 2006). Deforestation in riparian areas may also result in a reduction in the number of species or a change in species assemblage to more tolerant species (Jones et al. 1999).

Although fish reflect changes in land use in a watershed, macroinvertebrates may be more sensitive to human-induced changes at local scales. Macroinvertebrates respond to stream conditions relating to discharge and channel characteristics, as well as contaminants (Yagow et al. 2006, Brazner et al. 2007, Riva et al. 2008, Stepenuck et al. 2008). Wang and Kanehl (2003) found a shift from intolerant macroinvertebrate species, such as Ephemeroptera, Plecoptera, and Trichoptera (EPT), to more tolerant species and lower community richness with increasing urbanization.

The number of mussel species was highly variable across the state. Mussels are relatively immobile and live in close contact with both the sediments and the water column, thus along with macroinvertebrates, reflect impacts at a local scale. Mussels may have always been rare in some areas in Minnesota, such as the northeast, thus there may be uncertainty whether mussels are declining, and may be classified as “rare”.

Agriculture, forest harvesting, and urbanization alter the hydrology in a watershed resulting in increased sediment, nutrient and contaminant delivery to streams, and an increase in water flow, which can result in streambank and streambed erosion. The loss of more sensitive (intolerant) aquatic species is often the direct result of water quality degradation (Davies and Jackson 2006).



Confidence in index

Confidence in the metric is reasonably high. The data coverage and quality is high, and the metric is based on a studied set of calculations. Data for this index were collected by the MPCA for over 1,500 sites between 1996 and 2006. The Stream Species Quality Index was calculated for 705 sites where both fish and macroinvertebrates were collected in the same year at the same sites. Thus, there are extensive data available to calculate O/E scores. However, the O/E score only indicates whether an expected species was present, but does not depend on the abundance of species present at a site or whether a species is rare at a site. Rare taxa may have more specialized habitat requirements and exhibit higher sensitivity to environmental disturbances, and serve as early indicators to changes in water quality degradation or of the loss of particular habitat types (Cao et al., 1998; Cao and Williams, 1999).

In addition, Dolph et al. (2011) found that fish and macroinvertebrate assemblages indicated a concordant response in 17 of 25 watersheds. Both fish and macroinvertebrates are included in the index because the US Environmental Protection Agency recommends that two different taxonomic groups be sampled, and Rule 7050 in Minnesota State Statues requires that the MPCA evaluate two taxonomic groups to assess aquatic systems.

The index gains strength because the fish and macroinvertebrate taxa provide information about watershed conditions at a different scale. Fish are mobile and provide an indication of watershed conditions at a broad scale, whereas macroinvertebrates are less mobile and reflect conditions at a reach scale. Taken together, these two taxonomic groups provide a comprehensive overview of conditions in a watershed.

Although fish and macroinvertebrates have been assessed across the state, some watersheds have fewer assessments than others. Most assessments have been conducted in small streams within the 81 watersheds, because small streams represent 80-90% of stream miles within watersheds (Leopold et al. 1964) and most are “wadeable” by biologists. Thus, the condition of the larger rivers is less well represented. These data are collected from small stream reaches (less than 500 m in length) that may reflect conditions in a reach, which may or may not represent the conditions in the entire stream. The MPCA has developed a disturbance index comprised of eight variables designed to quantify the overall degree of human disturbance acting upon a stream site. Streams were designated as reference sites if disturbance scores fell within the upper 25% of scores for their ecoregion. The O/E approach predicts presence or absence of species based on prediction for sites that are classified as least impaired (reference sites). Because levels and types of human disturbance vary across Minnesota, reference sites were chosen to reflect the best obtainable condition in each ecoregion, rather than a single standard that could be applied to the entire state. Thus, a least impacted site does not necessarily represent an unaltered site. For example, a “least impacted” site with the highest score in a watershed could be in an area with extensive agriculture or in a suburban area with suitable habitat for a limited number of organisms. There can be a range in the quality of the reference sites that underlie the scoring, resulting in low expected values, and in turn inflating all scores, which may explain relatively higher values than expected in agricultural regions.


Future enhancements

The process and selection of reference sites could be evaluated to limit sites to those that are least impacted. Additional data to bolster the assessments in watersheds with incomplete sampling would strengthen the assessment of Stream Species Quality.

The watershed level scores will be re-calculated based on the catchment scores.  This will align the results across scales to the same analysis approach and process.  The catchment and watershed level scores will be updated annually based on on-going MPCA IBI data collection and MDNR mussel monitoring.