Integrating Multiple Taxa in aBiological Stream Rating System

State of IllinoisRod R. Blagojevich, Governor

Illinois Department of Natural ResourcesOffice of Resource Conservation

Photo CreditsAll photos were taken by the IDNRWatershed Protection Section staff except:

cover - Kevin Cummings - Bean Creek - Salt Forkpage 2 - Kevin Cummings - Hickory Creekpage 6 - Ed Dewalt - Hydroperla fugitans (Plecoptera: Perlodidae) the Springflypage 6 - Kevin Cummings - Threatened mussels at Salt Forkpage 7 - Kevin Cummings - Mussel sampling on the North Fork Vermilion Riverpage 11 - Chris Taylor - Orconectes propinquuspage 15 - Kevin Cummings - Asian clams at Lone Tree Creekpage 21 - Kevin Cummings - Pink Heelsplitter from the Sangamon River

Acknowledgmentshis work would not have been possiblewithout the previous efforts of the

Biological Stream Characterization WorkGroup that instituted a statewide streamrating system in the late 1980s and theIllinois Natural History Survey thatdeveloped the initial Biologically SignificantStreams listing. We would like to thank allthe members of our Biologically SignificantStreams work group for their efforts atenhancing this project. Special thanks toKevin Cummings, Ed DeWalt, MarkJoseph, Christine Mayer, Chris Phillips,Bob Schanzle, Bob Szafoni, Chris Taylor,John Wilker, and the IDNR streamspecialists for providing access to datafrom streams throughout Illinois. Thisreport is based largely on work done byLeslie Bol and Leon Hinz of the IllinoisNatural History Survey, and Ann MarieHoltrop of the IDNR. Their work wasfunded by IDNR through the Illinois StateWildlife Grant Program (T-20-P-001). �

i

T

ii

Prefacepdated stream ratings are provided inthis report under authority of state law

(see 515 ILCS 5-5 and 520 ILCS 5/2.1).This state law provides the IllinoisDepartment of Natural Resources (IDNR)with ownership of the wildlife and aquaticresources residing within the borders ofthe State of Illinois. The IDNR isdesignated as the agency of stategovernment charged with the regulation,protection, and preservation of thosenatural resources. Tools such as thestream ratings provided in this report areused by IDNR as the basis for fieldprogram implementation for resourceprotection. For over twenty years,resource mangers in Illinois have usedstream biological ratings as a vehicle forthe interpretation, assessment, andcommunication of aquatic resource values.The first stream ratings, published in 1989,were based on a five-tiered classificationsystem predicted largely on the type andcondition of the fishery resource. In July2005, the State of Illinois submitted aComprehensiveWildlife Conservation Planto the U. S. Fish and Wildlife Service aspart of a Congressional mandate to beeligible for future federal funding. The plan

was accepted, renamed the IllinoisWildlifeAction Plan, and became the strategicdocument guiding protection andconservation efforts throughout the state.As the name implies, the Illinois WildlifeAction Plan outlines a plan of action toaddress the particular needs of wildlife thatare declining and presents a targetedapproach to habitat enhancement andconservation. The Wildlife Action Planbroadly addresses all types of wildlifeincluding fish, mussels, amphibians, andreptiles. To help establish baselineconditions against which change promotedby the Illinois Wildlife Action Plan could bemeasured and understood, the followingreport describes in detail a stream ratingprocess based on multiple aquatictaxonomic groups. Users desiring accessto the most current ratings andaddit ional location information areencouraged to search http://www.dnr.state.il.us/orc/BioStrmRatings/. The ratings will provide the IllinoisDepartment of Natural Resources with amechanism for identifying high-qualityexamples of all stream communities andwill guide management and restorationactivities throughout the state.�

U

iii

Table of Contents

Acknowledgments i

Preface ii

Table of Contents iii

Introduction 1

General Approach for Diversity and Integrity Ratings 3

Diversity Ratings 5

Background 5

Approach 5

Examples of Diversity Ratings 12

Map of Diversity Ratings 14

Integrity Ratings 15

Background 15

Approach 15

Examples of Integrity Ratings 20

Map of Integrity Ratings 22

Biologically Significant Streams 23

Map of Biologically Significant Streams 25

Conclusions 26

Data Issues 26

Updates and Revisions 27

Literature Cited 30

Appendix A 33

iv

List of TablesTable 1. The number of sites from each dataset used to calculate diversity ratings.Table 2. Number of taxa corresponding to each class in the Macroinvertebrate Index

of Biotic Integrity (Tetra Tech, Inc. 2007).Table 3. Number of species corresponding to the three classes developed for the

Critical Trend Assessment Program’s Ephemeroptera, Plecoptera, andTricoptera data.

Table 4. Class scores for mussel species richness ratings based on expectationsaccording to drainage and stream size.

Table 5. The number of datasets contributing to final diversity ratings.Table 6. Examples of calculating diversity scores.Table 7. The number of sites from each dataset used to calculate integrity ratings.Table 8. Class scores for mussel single sample intactness percentages based on

expectations according to drainage and stream size.Table 9. Class scores for mussel historical intactness percentages based on

expecatations according to drainage and stream size.Table 10. The number of datasets contributing to final integrity ratings.Table 11. Examples of calculating integrity scores.Table 12. The underlying qualifications for designation as a biologically significant

stream (BSS).

List of Figures

Figure 1. Distribution of diversity scores and corresponding letter ratings.Figure 2. Geographic distribution of diversity ratings.Figure 3. Distribution of integrity scores and corresponding letter ratings.Figure 4. Geographic distribution of integrity ratings.Figure 5. Geographic distribution of biologically significant streams.

AppendixAppendix A. List of threatened and endangered species included in stream ratings.

omprehensive statewide biological,chemical, and physical information

associated with streams in Illinois has beenroutinely collected since 1980 through apartnership between the Illinois Departmentof Natural Resources (IDNR) and the IllinoisEnvironmental Protection Agency (IEPA;Bertrand et al. 1996). This partnership wasestablished in order to assess fish andmacroinvertebrate communities, waterquality, and habitat throughout major basinsof Illinois. In 1984, a Biological StreamCharacterization (BSC) Work Group wasconvened to create a mechanism forinterpreting data collected as part of theinteragency Basin Survey Program, and “toprovide managers an overall prospective ofthe state’s stream resources” (Hite andBertrand 1989). The BSC Work Groupdeveloped stream ratings using letter grades“A” through “E”, thereby establishing ameansof communicating the quality of biologicalresources in streams to diverse stakeholders.

At the time the BSCWork Group began, thefish-based Index of Biotic Integrity (IBI) wasrecently developed, and it became thepredominant stream integrity indicator usedfor rating streams (Hite and Bertrand 1989).In recognition of the need to also protectother stream-dependent organisms in thestate, the Illinois Natural History Survey(INHS) developed a list of BiologicallySignificant Streams (BSS) that incorporateddata on mussel communities and rarespecies (endangered, threatened, watch list)of crustaceans, fish, mussels, and aquaticplants in addition to stream segments ratedas “A” by the initial BSC (Page et al. 1992).The goal of the BSS project was to protect100% of the stream-dependent biodiversity,thus a stream with characteristics that metany one of the established criteria couldachieve status as a BSS (Page et al. 1992).

Despite the lack of regular updates, the BSCand BSS processes generated products thatare still used extensively by diversestakeholders including state and federalagencies, local watershed groups,consultants, environmental interest groups,and municipalities.

In 2006, the IDNR initiated an effort tocombine and update the previous streamrating efforts into a single rating. The purposebehind the project was not only to updateoutdated information (i.e., the existing ratingswere based on data at least 15 years old) butto create a rating system that would helpresource mangers determine efficacy inimplementing the aquatic goals of the IllinoisWildlife Action Plan (State of Illinois 2005). Tobe most useful in evaluating and guidingimplementation of the Wildlife Action Plan,IDNR sought a single rating for streamsegments that represented multiple signalsof stream condition. This intent was similarto the “overall prospective” identified by Hiteand Bertrand (1989). Although the mainpurpose behind stream ratings has changedsince the creation of BSC and BSS, severalother objectives for the development and useof ratings remain. These include:

• Facilitate planning and prudent allocationof State resources in IDNR monitoringactivities;

• Inventory and identify the nature, extent,and distribution of Illinois streamresources;

• Establish a common vehicle for theinterpretation, assessment, andcommunication of aquatic resourcevalues;

• Identify stream segments exhibiting ahigh potential for resource managementor restoration activities;

1

IntroductionC

• Focus greater emphasis on theimportance of uncommon aquatic bioticresources and an awareness of wherethese resources exist.

Since BSC and BSS were developed, thequantity and quality of aquatic data andassessment tools has increased. Forexample, multi-metric indices have beendeveloped for benthic macroinvertebrates(Tetra Tech, Inc. 2007) and mussels (Szafoni2002), and revised for fish (Smogor 2000).Further, the Basin Survey Program, whichassesses fish and macroinvertebratecommunities, has continued. These availableindices and data presented newopportunitiesto create a rating that reflects how differenttaxonomic groups can respond dissimilarly toshared stream conditions because ofdifferences in life-history, mobility, andsensitivities to stressors (Paller 2001).Specifically in this project we used fish,macroinvertebrate, and mussel informationbecause these taxa reflect steam conditionsat different spatial and temporal scales(Diamond and Serveiss 2001, Freund and

Petty 2007, Kilgour and Barton 1999,Lammert and Allan 1999). For instance, dueto their limited mobility, typically shorter lifespans, and association with streamsubstrate, macroinvertebrates may beindicators of local and more recent streamconditions (Freund andPetty 2007), whereasfish may be better indicators of regionalconditions because they have greatermovement capabilities and longer life cycles.Mussels, due to their limited dispersal asadults, may also indicate local conditions, butdue to longer life spans may reflect historicstressors related to specific areas (DiamondandServeiss 2001). By incorporating varioustaxonomic groups and averagingstandardized taxonomic scores, wegenerated an overall rating for streamsegments that is representative of multiplesignals of stream conditions. This reportdescribes an approach that results inassigning up to three designations for astream segment, which are a diversity rating,integrity rating, and identification as abiologically significant stream.�

2

everal purposes of the previous BSC andBSS processes overlapped between the

two initiatives. Both had objectives to identifythe extent of Illinois stream resources, toidentify stream segments of exceptionalquality, and to focus protection efforts towarduncommon resources or biologicallysignificant streams (Bertrand et al. 1996,Page et al.1992). However, the two initiativesdiffered in their overall intent to rate a stream’sbiological diversity (Page et al. 1992) orbiological integrity (Bertrand et al. 1996; Hiteand Bertrand 1989). For the purposes ofimplementing Illinois’ Wildlife Action Plan,IDNR sought a rating system that wouldinclude both diversity and integritymeasures.Although the approach to obtain the diversityand integrity ratings is similar, we have notdirectly combined the two ratings for anoverall rating. Diversity and integrity ratingswere kept separate because it is possible tohave highly intact communities that are notbiologically very diverse. For instance,species richness expectations for small orcold-water streams are expected to be lowcompared with larger or warmer streams.Therefore, it is possible to have a smallstream that would rate high for integrity butlow for diversity. Additionally, keeping the tworatings separate enables stakeholders withdifferent purposes to consider the rating thatis most applicable to their needs. The letterratings of A-E were maintained for both thediversity and integrity ratings as thesedesignations were used in the previous BSCrevision.

Given the change in focus and use for thisproject from previous stream ratings, weconsidered several aspects of the previousrating processes and modified the processaccordingly. Because multiple data sources

are used to generate a rating, there was aneed to standardize data from differentsources in an effort to give equal weight to allcommunities of organisms found in streams ifadequate and comparable sampling hadoccurred. Second, we sought a data drivenand reproducible process that did not includenarrative information (see Hite and Bertrand1989 and Bertrand et al. 1996 for anexplanation of how narrative information wasused previously). Third, we envisioned aproduct that could be easily updated as newinformation became available.

The general approach for obtaining adiversity or integrity rating is a six stepprocess:

1. Select data for inclusion in the rating.2. Convert raw data to a class score.3. Standardize classes into a proportional

score (P score).4. Average the proportional scores within a

given taxonomic group to obtain a singletaxonomic score (T score).

5. Average proportional and/or taxonomicscore for multiple sites on a valleysegment.

6. Determine the final diversity and/orintegrity rating for a valley segment.

We considered all the information thatcontributed to both integrity and diversityratings in order to identify BiologicallySignificant Streams (BSS). Similar to theinitial BSS effort, we incorporated multipledatasets and identified streams based onavailable taxonomic groups rather thanrelying on the fish data as the primary streamintegrity indicator. However, unlike theadditive approach of the original BSS thatidentified all reaches with appropriately high

General Approach for Diversity andIntegrity RatingsS

3

threatened and endangered speciespresence regardless of what other availableinformation may have indicated, the currentprocess uses a holistic approach thatcombines data sources to determine if thebiologically significant stream designation isappropriate.

Fish, mussel, macroinvertebrate, crayfish,and threatened and endangered speciesdata collected by various state agencieswereused for stream ratings. All datasets wereoverlaid on the 1:100,000 – scale, NationalHydrography Dataset (NHD; USGS 2000)that was refined for a previous project(Holtrop and Dolan 2003). Point locations ofdata that were greater than 60m from thenearest digitized stream line were visuallyinspected using an overlay of aerial images todetermine if the point was associated with alarge river or a small stream that was notdigitized. Points that were associated withlarge rivers and undigitized streams wereseparated into a different file and omittedfrom further analysis. Points that did not fallinto either of these categories were furtherinvestigated to determine if therewas an error

with the spatial coordinates. Errors wereremedied where possible, and points thatcould not be corrected and still fell greaterthan 60m from the nearest stream wereomitted.

Point data or sampling sites for the finalratings were summarized according to valleysegment. Valley segments are aggregationsof linearly adjacent, physically similar streamreaches (Seelbach et al. 1997). Physicalcharacteristics used to define valleysegments were related to stream size(drainage area), surficial geology (bedrock,coarse substrates), discharge (flow yield),and gradient. Valley segments wereindependently derived prior to this projectusing a spatially-constrained clusteringmethod based on the cluster affinity searchtechnique (Brenden et al. 2008). Valleysegment numberswere assigned to datasetsthrough a spatial join in ArcMap 9.2.Datasets were then associated with eachother for calculation of the final ratingaccording to valley segment number in aquery performed in Microsoft Office Access2003.�

4

Backgroundiversity simply defined is the number ofdifferent kinds of things (Angermeier and

Karr 1994) or the variety of life and itsprocesses (Hughes and Noss 1992).Although diversity can be representedmathematically using summary indices or asimple species number, we chose to considerit more broadly as the variety of taxa withinseveral important aquatic groups (e.g.,mussels, fish, macroinvertebrates, andcrayfish). In December 2006, projectstakeholders met and discussed theappropriateness of available datasets forinclusion in the diversity analysis. Weconsidered data collected within the pastdecade (1997-2006) that were collected aspart of IDNR, IEPA, or INHS monitoringprograms. We limited data to theseinstitutions to ensure that collection methodswere standardized, repeatable, and will becontinued in the future so that data will beavailable for revisions of these ratings.

ApproachThe general approach for obtaining adiversity rating is a six step process.

Step 1. Select data for inclusion into therating.

We considered only data that were collectedwithin the past decade. However, if a singlesite hadmore than one sample from the pastdecade, we used the samplewith the highestrichness for inclusion in the final ratingcalculation. We used this approach ratherthan taking the most recent sample or anaverage of the samples because the highestrichness represents a conservative estimateof the biological potential for the site and thisapproach accounts for variation that mayoccur with sampling. Additionally, we did notaverage the data frommultiple samples since

the average could represent a condition thathad not been found at the site. The followingdata were used in the final diversity ratings.

Fish – Fish data from community samplestaken as part of cooperative basin surveysand other department monitoring wereprovided by the IDNR. These data werereviewed by regional IDNR stream biologistsfor verification that the samples wererepresentative of community samples withadequate sampling efficiency. The species

richness metric was retrieved from the Indexof Biotic Integrity (IBI; Smogor 2000)summaries andwas used as a component ofthe diversity rating. A total of 731 sites wereused in the diversity score analysis (Table 1).There were fewer sites with fish speciesrichness than fish IBI scores since theindividual metrics scores used to calculatethe fish IBI were not always available.

5

Diversity Ratings

D

Table 1. The number of sites from each dataset usedto calculate diversity ratings.

Potential Data Source Number of Sites

Fish Species Richness 731Macroinvertebrate Taxa Richness 452CTAP EPT Species Richness 179S1S2 EPT Species Richness 104Mussel Species Richness 596Crayfish Species Richness 18Threatened and Endangered Species Richness 413

Total 2493

6

Aquatic Macroinvertebrates – Datafor aquatic macroinvertebrates werecompiled from three different entities.

MacroinvertebrateTaxa RichnessFirst, benthic macroinvertebrate data werecompiled from the IEPA in Springfield. Thesedata were collected following protocolsestablished for use in the Stream ConditionIndex (Tetra Tech, Inc. 2007), but referred toas the Macroinvertebrate Index of BioticIntegrity (MIBI) in this report. The taxarichness metric was retrieved from the MIBI,and a total of 452 sites were used for the finaldiversity score analysis (Table 1).

CriticalTrends Assessment Program(CTAP)Second, Ephemeroptera (mayflies),Plecoptera (stoneflies), and Tricoptera(caddis flies; EPT) data that werecollected since 1997 as part of CTAP(http://ctap.inhs.uiuc.edu/index.asp) wereobtained. Although theMIBI contains anEPTrichness metric, the CTAP data were usedbecause these data were collected in thespring of the year prior to the emergence ofmany of these species and also typically onsmaller streams than those included in theIEPA sampling. A total of 179 siteswere usedfor the final diversity score analysis (Table 1).

S1S2 EPTThird, we included information on sensitiveEphemeroptera, Plecoptera, and Tricoptera

data provided by Dr. Ed DeWalt (INHS).These data were included because currentlyno EPT species are listed as endangeredor threatened by the I l l ino isEndangered Species Protection Act(http://dnr.state.il.us/espb/datelist.htm),although some species within these ordershave been identified as critically imperiled(S1) or imperiled (S2) at the state level by anINHS entomologist (DeWalt et al. 2005,Favret and DeWalt 2002). S1S2 refers toconservation status ranks used byNatureServe (http://www.natureserve.org/).A total of 104 sites were used for the finaldiversity score analysis (Table 1).

Mussels – Mussel data were obtainedfrom the INHS mollusk collections database(http://www.inhs.uiuc.edu/cbd/collections/mollusk/molluskintro.html) and IDNR. Recordsassociated with freshwater snails, fingernailclams, zebramussels, and Asian clamswerenot included, as well as any records notassociated with stream habitat. In order toquery data that were representative ofcommunity samples, we restricted our datato a list of collectors’ names obtained fromKevin Cummings, the INHSmalacologist andmussel database manager. A total of 596sites were used for the final diversity scoreanalysis (Table 1).

Crayfish – Native crayfish data wereobtained from the INHS crustacean

c o l l e c t i o n da t a ba se ( h t t p : / /www.inhs.uiuc.edu/cbd/collections/crustacean/crustaceanintro.html). Despitethe lack of systematically collected crayfishdata across the state, we included crayfish ina limited capacity in the final diversity ratingsbecause they are abundant in Illinois streamsand we anticipate that additional collectionswill be available for future updates of streamratings. A total of 18 sites were used for thefinal diversity score analysis (Table 1).

ThreatenedandEndangeredSpecies–Data on threatened and endangered (T&E)fish, mussel, crayfish, amphibian, and plantspecies (see Appendix A for species lists)were extracted from the Biotics Databasemaintained by the IDNR Office of ResourceConservation, Division of Natural Heritage. Atotal of 413 sites withT&E species were usedfor the final diversity score analysis (Table 1).

Step 2. Convert raw data to a class score.

One of the objectives for this project was togive equal weight to all communities oforganisms found in streams if adequate and

comparable sampling had occurred. To dothis, we developed classes for each datasetused in the analysis in an attempt to interpretraw data from different sources and classify itsimilarly. Classes were independentlydeveloped for each dataset using eachsample collection as an independent recordrather than pooling samples from a singlesite. For example, if one site had multiplesamples collected between 1997-2006, theneach sample was treated as an independentrecord for the purpose of creating the classscores. Therefore, richness expectationswere based on the number of species youwould expect to find in a single samplingevent. Once the classes were established,only the sample that had the highest richnessfrom each site was used to calculate the finaldiversity rating.

Fish Species Richness — The fishspecies richness metric was retrieved fromthe Index of Biotic Integrity (IBI; Smogor2000) summaries and was used as acomponent of the diversity rating. We usedthe classes developed for IBI because theyaccounted for variation in fish species

7

richness expectations across different sizedstreams, slope, and region. We maintainedthese classes with a single modification. Inthe IBI, fish richness metric scores rangefrom0-6. Because the “0”does not representa true absence of fish, we added “1” to eachclass thereby resulting in class scores from1-7.

Macroinvertebrate Taxa Richness —The MIBI did not have classes associatedwith individual metrics; however theavailability of least-disturbed samplesprovided the opportunity to define classes formacroinvertebrate taxa richness by using thesame approach that was used to defineclasses for individual metrics within the fishIBI (Smogor 2000). The top class for taxarichness was set at the 75th percentile ofreference sites. Using this approach, taxarichness values forMIBI ranged from0 to 35+andwere placed into seven classes (Table 2).Datawere not further stratified by stream sizeor location because previous analysisdetermined that neither affected taxarichness expectations (TetraTech, Inc.2007).

CTAP EPT Species Richness — Inorder to maintain similarity across datasources, we used the 90th percentile as theboundary for the highest class fordatasets that were not developedwith a reference site approach (i.e.,mussels, CTAP EPT macroinvertebrates,S1S2 macroinvertebrates, crayfish, andthreatened and endangered species). Our

rationale was that by raising the standardfor the top class for these datasets to at leastthe 90th percentile, the highest class wouldbe similarly restrictive as the datasets that didhave reference site data available. Using the90th percentile as the cut for the top class,three classes were created (Table 3).

Mussel Species Richness —Amusselspecies richness of ten species or greaterwas previously used to identify BSS (Page etal. 1992) and was also used as the thresholdfor defining the highest classification for thespecies richness factor in the Illinois MusselClassification Index (Szafoni 2002; MCI).However, we investigated the relationshipamong mussel species richness acrossdifferent sized streams defined by steam link(Shreve 1967) within different drainages andsubsequently adopted new class scoresbased on our analysis. Three classes weredeveloped for mussel species richnessexpectations for each of the major drainagesbased on the percentiles within three streamsize groupings of the tributary streams andthemainstem (Table 4). Class one consistedof samples that were below average richnesswithin the drainage (0-49th percentile), classtwo were above average samples (50-89th),and class three were exceptionally highscoring samples (90th percentile and above(Table 4)).

Bonus Points –The final diversity ratingalso integrates information about taxa that

ClassScore TaxaRichness

7 35+6 31 - 345 25 - 304 19 - 243 13 - 182 7 - 121 0 - 6

8

Table 3. Number of species corresponding to the threeclasses developed for the Critical TrendAssessment Program’s Ephemeroptera,Plecoptera, and Tricoptera data. The speciesfrom the three orders are considered together.

Table 2. Number of taxa corresponding to each classin the Macro-invertebrate Index of BioticIntegrity (Tetra Tech, Inc. 2007).

Class Percentile Number of Species

1 <50th 1 - 8

2 50th - 89th 9 - 18

3 90th+ 19+

were deemed important due to their rarity.The S1S2 EPT, Crayfish, and T&E datasetshad a limited range of data and subsequentlywere used differently in the final ratings thanother fish, macroinvertebrate, and musseldata described previously. The rationale forthis is described in steps 4 and 6 below.Class scores for these three datasets werebased on percentiles, but were adjusted inweight based on how these data were addedto the diversity rating.

Step 3. Standardize classes into aproportional score (P score).

All class scores range from “1” to a greaternumber with the greatest number alwaysrepresenting the highest class. In this step,we divided the assigned class score by thetotal number of classes available to obtain aproportional score (P score), which has amaximum of 1. For example, a site that had26 macroinvertebrate taxa falls in class 5,

which equates to a P score of 5/7 (0.714).Proportional scoreswere used to standardizediffering numbers of classes amongvariables.

Step 4. Average the proportional scoresfor the three differentmacroinvertebrate datasets inorder to obtain a single taxonomicscore (T score).

When multiple datasets (i.e., taxa richnessfrom MIBI, EPT richness from CTAP, andS1S2 EPT species) were available formacroinvertebrates, the average of theproportional scores was used to determinethe taxonomic score (i.e., macroinvertebratetaxonomic score). Creating a taxonomicscore allowed us to include informationderived from separate assessments into acombined signal for macroinvertebrates.However, we averaged all availablemacroinvertebrate information into a

Table 4. Class scores for mussel species richness values based on expectations according to drainage and streamsize. Stream size is defined by link number, which is the number of first order streams based on the1:100,000 National Hydrography Dataset (NHD) upstream of a given stream reach. Link codes refer togroupings of link numbers.

9

Class 1 Class 2 Class 3StreamSize Drainage (<50th percentile) (50th - 90th percentille) (90th percentile +)

Illinois <3 3 - 7 8+Mississippli <2 2 - 5 6+Ohio 1 2 3+Wabash <3 3 - 8 9+

Illinois <5 5 - 11 12+Mississippli <5 5 - 10 11+Ohio <2 2 - 3 4+Wabash <5 2 - 10 11+

Illinois <5 5 - 11 12+Mississippli <7 5 - 11 12+Ohio <2 2 - 5 6+Wabash <6 6 - 13 14+

Illinois <9 9 - 10 11+Mississippli <15 15 - 20 21+Ohio <6 6 - 13 14+Wabash <3 3 - 9 10+

Small(Link code 1)

Medium(Link code 2 - 3)

Large(Link code 4 - 6)

Mainstem(Link code 7)

taxonomic score rather than keeping thedatasets separate and averaging themall intoa final score in order to give equal weight tofish, macroinvertebrates, and mussels in thefinal diversity rating.

S1S2 EPT data were added to themacroinvertebrate taxonomic score as bonuspoint data rather than averaged into the taxascore in order to ensure that the presence ofthese sensitive taxa always improved astream rating. The maximum number ofbonus points was awarded to samples withthree or more species as this corresponds tothe 90th percentile for the number of speciesfound per sample. Samples with 1-2 specieswere awarded half the maximum. Thediversity score prior to adding bonuspoints is based on the average of themacroinvertebrate taxonomic score,the fish proportional score and themussel proportional score. Since themacroinvertebrate taxonomic score ispotentially 1/3 of the overall diversityscore, and S1S2 EPT potentially contribute1/3 to the macroinvertebrate taxonomicscore, the S1S2 EPT data potentiallycontribute 1/9th (0.11) of the pre-bonuspoints diversity score. We therefore,assigned 0.11 for samples with 3+ and 0.055for 1-2 species.

Some valley segments had S1S2 EPT dataavailable but lacked other macroinvertebratedata. In these cases we added the bonuspoints after the fish and mussel taxonomicscores had been averaged (Step 5).However, since the data were added at adifferent point in the process, the bonuspoints were divided by three since theywouldcontribute to a third of the diversity score priorto the T&E and Crayfish bonus points beingadded. Therefore, for valley segmentswithout other macroinvertebrate data, 0.037was added when there were 3+ species and0.018 for samples with 1-2 species.

Step 5. Average proportional and/ortaxonomic score for multiple siteson a valley segment.

When multiple sites were associated with aparticular valley segment within a dataset, theaverage of these proportional or taxonomic(for macroinvertebrates) scores was used tocalculate the final diversity score. An averagefrom the different sites was used rather thanconsidering the highest proportional scorefrom the valley segment since conditionswithin the streamsegmentmay vary betweensites and an average for the whole valleysegment was a better representation than thesignal from a single site.

Step 6. Determine the final diversity ratingfor a valley segment.

The final diversity score is based on fivepotential data sources: average of the fishproportional scores available for the valleysegment, average of themussel proportionalscores available for the valley segment, theaverage macroinvertebrate taxonomicscores, as well as crayfish and T&E speciesrichness.

ThreatenedandEndangeredSpecies(T&E)Aquatic T&E data were added to thediversity score after the fish proportionalscores, mussel proportional scores, andmacroinvertebrate taxonomic scores havebeen averaged. BecauseT&E species wereone of five potential values contributing to afinal diversity rating, the 95th percentile ofT&E values (i.e., 2+ species) was awarded0.2 (1/5) bonus points. Sites having oneT&Especies were awarded 0.1 bonus points. Themaximum pointsT&E species could add to afinal diversity score was 0.2, even if morethan one sample for a given valley segmenthad 2+T&E species.

10

CrayfishSimilarly to T&E species, crayfish are addedas bonus points after available fish,macroinvertebrate, and mussel informationhad been averaged. However, bonus pointsfor crayfish were only awarded to samplesthat had three or more species. Three ormore species represented the 95th percentileof available data and resulted in 0.1 bonuspoints.

The final diversity score for a valley segmentwas calculated as:

DiversityScore=average (average fishspeciesrichnessP scores + averagemussel species Pscores + averagemacroinvertebrateT Scores)+ threatened and endangered species bonuspoints + crayfish bonuspoints,wherePscore=proportional score and T score = taxonomicscore.

The cut-offs for the final diversity letter ratingswere determined by visually inspecting thedistribution of the diversity scores (Figure 1).We also attempted to have a similarpercentage of valley segments within eachletter category as the previous BSC projects.A total of 1127 valley segments wereassigned a diversity rating of A-E (Figure 2).This represents 3% of the total 38046 valleysegments that exist for the state of Illinois. Ofthe valley segments that were rated, thepercentagewith the assignment of the ratingsA-E is 13, 22, 38, 25 and 1 respectively.While this procedure has been developed forassigning ratings using multiple datasets,approximately one half of the total valleysegments that were rated had data availablefrom only one dataset (Table 5).

11

Table 5. Number of datasets contributing to finaldiversity ratings.

Figure 1. Distribution of diversity scores and corresponding letter rating. The percentage of valley segmentswith diversity ratings of A-E is 13, 22, 38, 25, and 1 respectively.

Distribution ofDiversity Scores

Datasets TotalValley Segments

1 565

2 370

3 134

4 44

5 11

6 3

Total 1127

12

Examples of Diversity RatingsTo further illustrate the diversity process, wepresent several examples (Table 6). In thefirst example, only one dataset is associatedwith the valley segment. The fish speciesrichness is 15, which corresponds to a classscore of 5. To obtain the proportional score,5 is divided by the total number of classes,which is 7. Since there are no other datasetsto average with the fish species richness, thefinal diversity score is the same as the fishproportional score. A final diversity score of0.714 equates to a letter rating of C.

In the second example, data are availablefrom three taxonomic groups. The fishspecies richness is 22, which equates to a

class score of 6 and a proportional score of0.857. The mussel species richness is 6,which equates to a class score of 2 and aproportional score of 0.667. Themacroinvertebrate taxa richness is 42, whichequates to a class score of 7 and aproportional score of 1. The diversity scoreis determined by averaging these threeproportional scores. The final score of 0.841corresponds to a letter rating of C.

The third example has two sets ofmacroinvertebrate data as well as fish andmussel data. The fish species richness is 10,equating to a class score of 3 and aproportional score of 0.429. The musselspecies richness is 1, equating to a class

Table 6. Examples of calculating diversity scores.

ValleySegment

FishSpeciesRichness

Fishspeciesrichnessclassscore

Fishproportionalscore

Musselspeciesrichness

Musselspeciesrichnessclassscore

Musselproportionalscore

Macroinvertebratetaxarichness

Macroinvertebratetaxarichnessclassscore

Macroinvertebratetaxarichnessproportionalscore

CTAPEPTspeciesrichness

CTAPEPTspeciesrichnessclassscore

CTAPEPTspeciesrichnessproportionalscore

S1S2EPTspecierichness

S1S2EPTspecierichnessbonuspoints

Macroinvertebratetaxonomicscore

Pre-bonuspointsDiversityscore

Crayfishspeciesrichness

Crayfishspeciesrichnessbonuspoints

ThreatenedandEndangeredspeciesrichness

ThreatenedandEndangeredspeciesrichnessbonuspoints

FinalDiversityScore

DiversityRating

Examplewithsingledataset

21679

15

5

0.714(5/7)

0.714

0.714

C

Examplewiththreetaxonomic

groups

39073

22

6

0.857(6/7)

6

2

0.667(2/3)

42

7

1(7/7)

1

0.841

0.841

B

Examplewithtwomacroinvertebrate

datasets

37913

10

3

0.429(3/7)

1

1

0.333(1/3)

31

6

0.857(6/7)

17

2

0.667(2/3)

0.76

0.51

0.51

D

ExamplewithS1S2EPTbonus

points

3557

20

3

1(3/3)

1

0.055

1.055

1.055

1.055

A

Examplewithtwomusselsitesandthreatenedandendangeredspecies

bonuspoints

44269

33

7

1 (7/7)

1 and 13

1 and 3

0.667(averageof0.33and1)

40

7

1 (7/7)

1

0.889

2

0.2

1.089

A

13

score of 1 and a proportional score of 0.333.The macroinvertebrate taxa richness is 31equating to a class score of 6 and aproportional score of 0.857. The CTAP EPTspecies richness is 17 equating to a classscore of 2 and a proportional score of 0.667.Before the diversity score can be calculated,availablemacroinvertebratedataarecombinedinto a taxonomic score. Themacroinvertebratetaxonomicscore isdeterminedbyaveraging themacroinvertebrate taxa richness proportionalscore and the CTAP EPT proportional score.The final diversity score (0.51with a diversityrating of D) is calculated by averaging the fishand mussel proportional scores and themacroinvertebrate taxonomic score.

The fourth example also has two datasetsavailable for macroinvertebrates. However,one of the datasets is S1S2EPT bonus data.The CTAP ETP species richness is 20,which represents a class score of 3 and aproportional score of 1. There is one S1S2EPT species associated with the valleysegment that is awarded 0.055 bonus points.The macroinvertebrate taxonomic score istherefore the CTAP EPT proportional score

plus the S1S2 EPT bonus points. Since noother data are available, the final score isequal to the macroinvertebrate taxonomicscore (1.055 with a diversity rating of A).

The final example illustrates the procedure fordealing with valley segments that may havemore than one sampling site associated withthemand for calculating a final diversity scoreusing threatened and endangered speciesbonus points. The fish species richness is 33equaling a class/metric score of 7 and aproportional score of 1. There are twomussel sites associated with the valleysegment with species richness of 1 and 13.These correspond to class/metric scores of1 and 3 respectively. To determine the finalproportional score for the mussels, theaverage is taken of the two site proportionalscores. The fish and mussel proportionalscores are then averaged before bonuspoints are awarded. Two threatened andendangered species are associated with thevalley segment equating to 0.2 bonus points.Once these are added to the pre-bonus pointdiversity score of 0.889, the final diversityscore is 1.089, which equals an A rating.�

14

Figure 2. Geographic distribution of diversity ratings. Three percent of all valley segments for Illinois have adiversity rating. Access to the diversity data associated with individual streams is available at:http://www.dnr.state.il.us/orc/BioStrmRatings/.

Map of Diversity Ratings

Backgroundiological integrity refers to a system’swholeness (Angermeier and Karr

1994) and the ability of a system tosupport organisms and processescomparable to natural habitat of the region(Hughes and Noss 1992). Indices orassessment measures like the fish andmacroinvertebrate Indexes of BioticIntegrity (Smogor 2000, Tetra Tech, Inc.2007) measure how closely a testcommunity resembles a natural, least-disturbed, or intact community (seeStoddard et al. 2006 for a discussion ofthese terms). Intactness for fish andmacroinvertebrates was determined fromthe indices of biotic integrity in comparisonto least disturbed or reference sites.Intactness for mussels was determined incomparison to historical species richnessexpectations for a site. In December 2006,project stakeholders met and discussedthe appropriateness of available datasetsfor inclusion in the integrity analysis. Weconsidered data collected within the pastdecade (1997-2006) that were collected aspart of IDNR, IEPA, or INHS monitoringprograms. We limited data to theseinstitutions to ensure that collectionmethods were standardized, repeatable,and will be continued in the future so thatdata will be available for revisions of theseratings.

ApproachThe general approach for obtaining anintegrity rating is a six step process.

Step 1.Select data for inclusion into therating.

We considered only data that werecollected within the past decade. However,

if a single site had more than one samplefrom the past decade, we used the samplewith the highest value for inclusion in thefinal rating calculation. We used thisapproach rather than taking the mostrecent sample or an average of thesamples because the highest valuerepresents a conservative estimate of thebiological potential for the site and thisapproach accounts for variation that mayoccur with sampling. Additionally, we didnot average the data from multiplesamples because the average couldrepresent a condition that had not beenfound at the site. The following data wereused in the final integrity ratings.

Fish–Fish data from community samplestaken as part of the cooperative BasinSurvey Program and other departmentmonitoring were provided by the IDNR.These data were reviewed by regionalIDNR stream biologists to verify that thesamples were representative communitysamples with adequate samplingefficiency. Fish Index of Biotic Integrity(IBI) scores from the compiled sampleswere used to calculate integrity ratings. Atotal of 744 sites with calculated Fish Indexof Biotic Integrity (IBI; Smogor 2000)scores were used in the final integrityscore analysis (Table 7).

Aquatic Macroinvertebrates – Benthicmacroinvertebrate data were compiled

15

Integrity Ratings

B

Table 7. The number of sites from each dataset used tocalculate integrity scores.

IntegrityDataset Number of Sites

Fish IBI 744Macroinvertebrate IBI 452Mussel Classification Index 134Mussel Single Sample Intactness 329Mussel Historical Intactness 366

Total 2025

16

from the IEPA in Springfield. These datawere collected following protocolsestablished for use in their StreamCondition Index (Tetra Tech, Inc. 2007),referred to as the Macroinvertebrate Indexof Biotic Integrity (MIBI) in this project. Atotal of 452 sites with total MIBI scoreswere used for the final integrity scoreanalysis (Table 7).

Mussels – Mussel data were obtainedfrom the INHS mollusk collections database(http://www.inhs.uiuc.edu/cbd/collections/mollusk/molluskintro.html) and IDNR.Recordsassociated with freshwater snails,fingernail clams, zebra mussels, and Asianclams were not included, as well as anyrecords not located in streams. In order toquery data that were representative ofcommunity samples, we restricted our datato a list of collectors’ names obtained fromKevin Cummings, the INHS malacologistand mussel database manager. Threevariables were used to determine integrityratings for mussels: mussel communityindex (MCI), single sample intactness, andhistorical intactness.

FreshwaterMussel Classification Index(MCI)Data were obtained from Bob Szafoni(IDNR) for sites where the MCI has beencalculated (Szafoni 2002). The MCI iscomprised of four metrics: speciesrichness, abundance, presence ofintolerant species, and recruitment(Szafoni 2002). Each of these metrics isscored and the scores are then summedto determine an index score. Although theMCI is comprised of multiple metrics likethe fish IBI and MIBI, it differs from thesebecause the response of metrics includedin MCI to human impacts in watershedshas not been considered as part of theMCI development. Because referenceconditions were not used to evaluatemetrics, the resulting MCI scores do notrepresent how far a sampled musselcommunity is from a natural or referencecondition. Rather, they were selected torepresent the characteristics of a healthyfunctioning community. Fundamentally thisis different than the fish andmacroinvertebrate IBIs, however weincluded the MCI in this project with theexpectation that the index will be refined inthe future and the availability of data willincrease. A total of 134 sites were used forthe final integrity score analysis (Table 7).

IntactnessOne metric currently considered forinclusion into the MCI is communityintactness, which is simply defined as theproportion of live species found at site towhat is expected. Initial analysissuggested that the expected valueincreased with the number of samplesavailable for a site. Therefore, wecalculated both single sample andhistorical intactness values to account fordifferent numbers of samples among sites.

17

Both intactness values were calculated fora site using the community sample fromthe past decade with the highest speciesrichness of live mussel species divided bythe total number of species including dead(dead and newly empty shells) and relict(old shells) specimens. For single sampleintactness, the total number of species wasfrom the single sample while for historicalintactness it included all the species foundat the site from all available samples. Ifboth historical and single sampleintactness were calculated for a site, thenhistorical intactness was used in the finalintegrity ratings. A total of 366 historicalintactness sites and 329 non-overlappingsingle sample intactness sites were usedfor the final integrity score analysis (695total mussel sites, Table 7).

Step 2. Convert raw data to a class score.

One of the objectives for this project wasto give equal weight to all communities oforganisms found in streams if adequateand comparable sampling had occurred.To do this, we developed classes for eachdataset used in the analysis in an attemptto interpret raw data from different sourcesand classify it similarly. Classes wereindependently developed for each datasetusing each sample collection as anindependent record rather than poolingsamples from a single site. For example, ifone site had multiple samples collectedbetween 1997-2006, then each samplewas treated as an independent record forthe purpose of creating the class scores.Therefore, integrity and intactnessexpectations were based on the number ofspecies you would expect to find in a singlesampling event. Once the classes wereestablished, only the sample that had thehighest value from each site was used tocalculate the final integrity rating.

Fish Index of Biotic Integrity — Thefish Index of Biotic Integrity (IBI; Smogor2000) scores were used as a componentof the integrity rating. Because the IBIalready had five integrity classesassociated with the index (Smogor 2005),we maintained these classes with littlemodification. In the IBI, the integrityclasses ranged from one (best) to five(worst). We reversed the numbering of theclasses to give the sites with the highest IBIscore a 5 instead of a 1.

Macroinvertebrate Index of BioticIntegrity (MIBI) — The MIBI (Tetra Tech,Inc. 2007) scores, based on seven metrics,were used as a component of the integrityrating. In the MIBI, final scores are placedinto one of four classes, with one being theworst and four being the best. Wemaintained these four classes for thisproject.

MusselsMussel Classification Index (MCI)Szafoni (2002) defined five classes for theMCI ranging from 0-4. We maintainedclasses 1 through 4 for the integrity ratings.Sites with a total score of 0 had no livemussels present and were not included inthe final integrity rating calculations.

IntactnessWe used the 90th percentile as theboundary for the highest class for datasetsthat were not developed with a referencesite approach or did not have classesalready developed for the index. Ourrationale was that by raising the standardfor the top class for intactness the 90thpercentile, the highest class would besimilarly restrictive as the datasets that didhave reference site data available. Wedeveloped classes for historic and singlesample intactness independently. For each,

intactness classes consistedof the 1-10th percentile forclass 1 and the 11-50th, 51-89th and 90th+ percentile forclasses 2, 3, and 4 respectively.Similar to mussel speciesrichness expectations, c l a s seswe re ass i gned according todrainage and stream size(Tables 8 and 9).

Step 3. Standardize classesinto a proportionalscore (P score).

Proportional scores were usedto standardize differing numbersof classes among variables. Allmetric/class scores range from“1” to a greater number with thegreatest number alwaysrepresenting the highest class.In this step, we divided theassigned class score by the totalnumber of classes available toobtain a proportional score (Pscore), which has amaximumof 1.

Step 4. Average theproportional scoreswithin a giventaxonomic group toobtain a singletaxonomic score(T score).

Three datasets were potentiallyavailable for mussels: MCI score(Szafoni 2002), single sampleintactness, and historicalintactness. If both historical andsingle sample intactness wereavailable for a site, thenhistorical intactness was used in the finalintegrity ratings. When MCI and intactnessscores were both available for mussels,

then the average of the proportional scoreswas used to determine the taxonomicscore (i.e., mussel taxonomic score).Creating a taxonomic score allowed us to

18

Table 8. Class scores for mussel single sample intactness percentagesbased on expectations according to drainage and stream size.Stream size is defined by link number, which is the number offirst order streams based on the 1:100,000 NationalHydrography Dataset (NHD) upstream of a given streamreach. Link codes refer to groupings of link numbers.

Table 9. Class scores for mussel single sample intactness percentagesbased on expectations according to drainage and stream size.Stream size is defined by link number, which is the number offirst order streams based on the 1:100,000 NationalHydrography Dataset (NHD) upstream of a given streamreach. Link codes refer to groupings of link numbers.

Single Sample IntactnessPercentageStreamSize Drainage Class 1 Class 2 Class 3 Class 4

Illinois 1 - 27 28 - 65 66 - 83 84+Mississippli 1 - 19 20 - 50 51 - 83 84+Ohio 1 - 20 21 - 42 43 - 54 55+Wabash 1 - 33 34 - 60 61 - 79 80+

Illinois 1 - 26 27 - 71 72 - 90 91+Mississippli 1 - 35 36 - 71 72 - 88 89+Ohio 1 - 12 13 - 44 45 - 76 77+Wabash 1 - 20 21 - 50 51 - 82 83+

Illinois 1 - 21 22 - 50 51 - 83 84+Mississippli 1 - 32 33 - 64 65 - 77 78+Ohio na na na naWabash 1 - 24 25 - 55 56 - 88 89+

Small(Link code 1)

Medium(Link code 2 - 3)

Large(Link code 4 - 6)

Historical IntactnessPercentageStreamSize Drainage Class 1 Class 2 Class 3 Class 4

Illinois 1 - 22 23 - 50 51 - 79 80+Mississippli na na na naOhio 1 - 15 16 - 27 28 - 59 60+Wabash 1 - 17 18 - 50 51 - 71 72+

Illinois 1 - 20 21 - 62 63 - 79 80+Mississippli 1 - 20 21 - 57 58 - 79 80+Ohio 1 - 14 15 - 31 32 - 53 54+Wabash 1 - 14 15 - 41 42 - 71 72+

Illinois 1 - 11 12 - 44 45 - 69 70+Mississippli 1 - 16 17 - 45 46 - 63 64+Ohio na na na naWabash 1 - 13 14 - 40 41 - 62 63+

Small(Link code 1)

Medium(Link code 2 - 3)

Large(Link code 4 - 6)

include information derived from separateassessments into a combined signal formussels. However, we averaged allavailable mussel information into ataxonomic score in order to give equalweight to fish, macroinvertebrates, andmussels in the final integrity rating.

Step 5. Average proportional and/ortaxonomic score for multiple siteson a valley segment.

When multiple sites were associated with aparticular valley segment for a dataset, theaverage of these proportional or taxonomic(for mussels) scores was used to calculatethe final integrity score. An average fromthe different sites was used rather thanconsidering the highest proportional scorefrom the valley segment since conditionswithin the stream segment may vary andan average for the whole valley segmentwas a better representation than the signalfrom a single site.

Step 6. Determine the final integrity ratingfor a valley segment.

The final integrity score for a valleysegment was calculated as:

Integrity Score = average (average fish IBIP scores + average MIBI P scores +average mussel T scores), where P score= proportional score and T score =taxonomic score

The cut-offs for the final integrity letterratings were determined by visuallyinspecting the distribution of the integrityscores (Figure 3). We also attempted tohave a similar percentage of rated valleysegments within each letter category to theprevious BSC projects. A total of 1019valley segments were assigned an integrityrating of A-E (Figure 4). This represents2.7% of the total valley segments. Thepercentage of valley segments with theassignment of ratings A - E is 9, 31, 45, 10and 5 respectively. While this procedurehas been developed for assigning ratingsusing multiple datasets, approximately onehalf of the total valley segments that wereassigned an integrity score used data fromonly one dataset (Table 10).

19

Figure 3. Distribution of integrity scores and corresponding letter ratings. The percentage of valley segmentswith integrity ratings of A-E is 9, 31, 45, 10, and 5 respectively.

Distribution of Integrity Scores

Examples of Integrity RatingsWe provide several examples to furtherillustrate the integrity rating process (Table11). In the first example only the singledataset of macroinvertebrate IBI isassociated with the valley segment. TheMIBI score is 39.99 which equals a class 2

out of 4; therefore the proportional score is0.5. Since there are no other datasets

available for this valley segment the finalintegrity rating is also 0.5 (Integrity Rating C).

In the second example both the MIBI andfish IBI are available. The fish IBI score is47 corresponding to class 4 and aproportional score of 0.8. The MIBI scoreis 65.39 corresponding to class 3 and aproportional score of 0.75. The average ofthe fish IBI and MIBI proportional scores iscalculated to determine the final integrityscore of 0.775, which equates to anintegrity rating of B.

In the third example, the fish IBI, MIBI, andtwo mussel datasets are available. Thefish IBI score is 55, which is a class 4 scorewith a proportional score of 0.8. The MIBIscore is 78.23 with a class score of 4 anda proportional score of 1. The mussel

20

Table 10. The number of datasets contributing tofinal integrity ratings.

Table 11. Examples of calculating integrity scores.

ValleySegment

FishIBIscore

FishIBIclassscore

FishIBIproportionalscore

MacroinvertebrateIBIscore

MacroinvertebrateIBIclassscore

MacroinvertebrateIBIproportionalscore

MusselClassificationIndexscore

MusselClassificationIndexclassscore

MusselClassificationIndexproportionalscore

Musselsinglesampleintactnesspercentage

Musselsinglesampleintactnessclassscore

Musselsinglesampleintactnessproportionalscore

Musselhistoricalintactnesspercentage

Musselhistoricalintactnessclassscore

Musselhistoricalintactnessproportionalscore

Musseltaxonomicscore

Integrityscore

Integrityrating

Examplewithsingledataset

38663

39.99

2

0.5(2/4)

0.5

C

ExamplebasedonFishandMacroinvertebrate

IBIs

29766

47

4

0.8(4/5)

68.39

3

0.75(3/4)

0.775

B

Examplewithtaverageof

musseldatasets

44269

55

4

0.8 (4/5)

78.23

4

1 (4/4)

16

4

1 (4/4)

29

2 (2/4)

0.5

0.75

0.85

B

Datasets TotalValley Segments

1 515

2 306

3 104

4 80

5 12

Total 1019

21

classification index score is 16 with a classscore of 4 and a proportional score of 1.The single sample intactness percentageis 29, which is a class 2 score and aproportional score of 0.5. The two musselproportional scores are averaged for amussel taxonomic score of 0.75. The final

integrity score is then the average of thefish IBI proportional score, the MIBIproportional score, and the musseltaxonomic score. The final score equals0.85, which is equivalent to an integrityrating of B.�

22

Map of Integrity Ratings

Figure 4. Geographic distribution of integrity ratings. Of the total 38,046 valley segments for thestate, only 2.7% have an integrity rating. Access to the integrity data associated withindividual streams is available at: http://www.dnr.state.il.us/orc/BioStrmRatings/.

iologically Significant Streams (BSS)are defined as streams that have a

high rating or score based on data from atleast two taxonomic groups. This can beachieved by obtaining an A rating either fordiversity or for integrity that is based ondata from two or more taxonomic groups.A second way to achieve this status is fora stream segment to have class scores inthe highest class for at least two differenttaxonomic groups when considering thecombined data from the diversity andintegrity ratings. While these criteria mayseem more rigorous than the previousBSS assessment, we believe this ismerited. By requiring BSS segments tohave either an A rating or high class scoresfrom separate assessments, we assuredthat only the highest rated reaches aregiven biologically significant status. Byconsidering two taxonomic groups, wehave more confidence in the BSSdesignation because at least two signalsare indicating high biological significancewithin the stream.

A total of 1366 valley segments had dataassociated with them. Our primary criteriarequiring a valley segment to contain thehighest class score from two differenttaxonomic groups accounted for 84% of allBSS identifications. However, most valleysegments (56%) that were identified asbiologically significant also received an Arating for Diversity and/or Integrity (Table12).

Stream segments identified as biologicallysignificant are unique resources in thestate and we believe that the biologicalcommunities present must be protected atthe stream reach, as well as upstream of

the reach. It is well documented in thescientific literature that the physical andchemical properties of water at a streamsite reflect upstream influences (Omernicket al. 1981, Smart et al. 1981, Hunsakerand Levine 1995). However, we areunaware of any criteria that can definitivelyidentify the upstream extent of influence onbiota within each stream reach identifiedas biologically significant. Therefore, weused some simple, practical constraints forextrapolating from site-specific informationto upstream stream segments to arrive atthe final segments identified as biologicallysignificant. Stream reaches (i.e., arcsdefined as confluence to confluencereaches) upstream of a valley segmentthat was identified as BSS were also

23

Table 12. The underlying qualifications fordesignation as a biologically significantstream (BSS). All BSS were evaluatedbased on information from at least twodatasets from differing taxonomic groups.For streams rated A for diversity orintegrity, at least two datasets fromdifferent taxonomic groups had tocontribute to the final rating. For streamsthat had the highest class score, the twodifferent taxonomic groups could bederived from a combination of both thediversity and integrity datasets.

BiologicallySignificant StreamsB

Rationale Count

2+ highest classes but no A ratings 54Total with A rating 68

Total BSS valley segments 122

Breakdown 2+ highest class ratingsIntegrity A & 2+ highest classes 5Diversity A & Integrity A & 2+ highest classes 11Diversity A & 2+ highest classes 332+ highest classes but no A ratings 54

Total with 2+ highest classes 103

Breakdown A ratings

Diversity A & Integrity A 1Integrity A & 2+ highest classes 5Diversity A 8Integrity A 10Diversity A & Integrity A & 2+ highest classes 11Diversity A & 2+ highest classes 33

Total with A Rating 68

24

identified as biologically significant if ALLof the following criteria applied:

1) The nearest downstream valleysegment has sufficient biologicalinformation to warrant BSS status.

2) The stream reach is part of the BSSand not a tributary connecting to it.

3) The stream reach is not smaller thanthird order in size. Stream order is arelative measure of stream size; largerorders represent larger streams. Usingthird order as a size limit is consistent withthe extent of range for the majority of fish,

mussel, and macroinvertebrate informationused, which predominately was collectedfrom third-order streams and larger.Importantly, not all stream segmentssmaller than third order were denied BSSstatus outright. As per the first criterion,regardless of stream size, if sufficientbiological information was available fromthe valley segment and the informationindicates high integrity or diversity, thesegment was identified for BSS status.

4) The stream reach is free-flowing, i.e.,not obviously part of a lake, reservoir, orlarge river.�

25

Map of Biologically Significant Streams

Figure 5. Geographic distribution of biologically significant streams.Access to the data associated with individual streams is available at:http://www.dnr.state.il.us/orc/BioStrmRatings/.

26

Conclusionshe ratings proposed in this documentincorporate aspects of both previous

BSC and BSS processes. Since thepublication of BSC and BSS, newinitiatives have been implemented tocollect biological information relevantto streams such as the Critical TrendsAssessment Program, MusselClassification Index, and the BenthicMacroinvertebrate Stream ConditionIndex (MIBI in this report). The fish IBI hasalso been revised and the list of threatenedand endangered species has changedsince the original publication of BSS. Withthe additions and changes to these datasources, it was pertinent to reassess thestrengths and weaknesses of the previousstream ratings in the context of supportingimplementation of Illinois’Wildlife Action Plan.The Illinois Wildlife Action Plan identifies abroad array of species in greatest need ofconservation, and therefore it wasappropriate to consider multiple taxonomicgroups in this project. In keeping with theIllinoisWildlife Action Plan’s stream habitatgoal that: “High–quality examples of allriver and stream communities . . . arerestored and managed within all naturaldivisions in which they occur”, the currentstream ratings and identification ofbiologically significant streams provide anew and updated tool to identify and targetsuch areas. By combining multipledatasets from different taxonomic groupsinto a single rating, this project givesratings that are a holistic representation ofstream biological resources. Because weconsidered data in addition to fish, ratingswere applied to an additional 483 valleysegments that lacked fish data.

Data Issues

Other taxonomic groups wereinvestigated but not used because oflimited available data. For example,information on amphibians and reptiles inIllinois were obtained from the INHSamphibian and reptile collection. Of thelisted amphibian and reptile species, theDusky Salamander, is a species found instream habitat (Phillips et al. 1999) and isconsidered an indicator species in smallstreams without fish (Southerland et al.2004). While we included the DuskySalamander in with the T&E species, wedid not include other reptiles andamphibians because we lacked sufficientstatewide information on the distribution ofherpitiles inhabiting streams.

Plant information was also pursuedbecause multiple species were includedpreviously in the Biologically SignificantIllinois Streams (Page et al. 1992)publication. However, of the plant speciesthat are still protected under the IllinoisEndangered Species Protection Act, onlythe heart-leaved plantain (Plantagocordata) is considered an associate ofstream habitat (Herkert and Ebinger 2002).Many of the species included in the originalBSS were aquatic plants associated withpond habitats and therefore were notincluded in our analysis. We consultedState experts, including INHS personnelpreviously involved with BSS (Page et al.1992), to determine if other potentialbotanical datasets were available.However, no additional plant species wereincluded in our ratings since there have notbeen systematic statewide surveys ofplants associated with stream habitat.

T

27

Updates andRevisions

One of the goals of the previous BSCinitiatives was to update stream ratings onan annual basis and to publish the revisedratings every five years. However, theoriginal BSC stream ratings were updatedonly once based on data that werecollected through 1993. Similarly, the BSSproject was based on data collectedthrough 1991 and has not been updatedsince. Therefore, stream designationsidentified in these projects are based ondata that is at least 14 years old. Giventhat these ratings are used by a diversegroup of stakeholders, it was clear that anupdated version was required.

Several reasons may explain why previousstream ratings have changed through thisproject including: a new process evaluating

diversity and integrity data, addition of datapreviously unavailable, revision to the fishIBI and T&E species list, and changes instream condition. Because previousstream ratings may have changed forthese reasons, comparisons of newratings to previous ratings (from Hite andBertrand 1989, Page et al. 1992, Bertrandet al. 1996) are not appropriate. Forexample, a stream reach rated as C in thisreport that was previously B should not beinterpreted automatically as a degradationin stream quality. In addition to a revisedprocess for assigning letter grades,biologically significant streams must nowhave data from two different taxonomicgroups. Therefore, some streamspreviously identified as BSS did notreceive the BSS designation in this effortbecause they lacked sufficient data giventhe change in criteria.

The ratings included in this report canassist in identifying streams that are inneed of restoration or improvedconservation. Given that less than 5% ofthe valley segments in the state have dataassociated with them, this project alsoindicates data gaps and can help prioritizefuture survey efforts. Current fish andmacroinvertebrate indexes are onlyapplicable to wadeable streams, thus welimited ratings to wadeable conditions.Development of assessment tools forheadwaters and larger rivers would allowbroader application of ratings in the future.Systematic surveys of mussels andcrayfishes would support index refinementand broader inclusion of these taxa. Asstatewide surveys increase, the inclusionof other taxa such as herpitiles or aquaticmacrophytes may be possible in futureupdates of the stream ratings.

The final product of diversity and integrityrat ings and biological ly signi f icants t reams, ava i l a b l e a t http://www.dnr.state.il.us/orc/BioStrmRatings/,indicates the data sources thatcontribute to each final rating and includesthe proportional scores for these data. Thisinformation will enable differentstakeholders with varying goals to use theratings and contributing data for theirparticular purposes. For example, if astakeholder wanted to target their efforts atstreams with high mussel species diversitythey would be able to identify thosestreams according to the mussel speciesrichness proportional score contributing tothe final diversity score. Similarly, effortsfocused at streams with a high fish IBIscore could consider the fish IBIproportional score contributing to a finalintegrity score.

28

29

The major data collection programs(collaborative basin surveys, CTAP,Endangered Species Board updates) usedin this project operate on a five yearinterval to assess streams statewide.Therefore, the IDNR intends toupda t e ra t i n g s annua l l y a t http://www.dnr.state.il.us/orc/BioStrmRatings/and publish new ratings, includingdesignating biologically significantstreams, after the completion of eachround of basin surveys. A publishedrevision of ratings should be availableapproximately every 5-6 years. With eachpublished update, a new range of datafrom each of the sources will be selected toencompass the last ten years. For certaindatasets such as the fish andmacroinvertebrate IBIs, the values thatcorrespond to the class scores will not

have to be recalculated since they werealready established. However, for otherdatasets such as the mussel speciesrichness and intactness data, the numberof species that correspond to thepercentiles that were used to determineclass scores will undoubtedly change withthe collection of additional data. For thesedatasets, the values that represent thedifferent class scores should berecalculated using the new data for eachrevision until these values can be moreformally established. In addition, the cut-offs for the letter ratings are based on thedistribution of the final scores. In the futurethese cut-offs could change as new dataare analyzed. Therefore, the final scoresthat correspond to the letter ratings A-Eshould be reevaluated with any update.�

30

Literature CitedAngermeier, P.L. and J.R. Karr. 1994. Biological integrity versus biological diversity as

policy directives. BioScience 44:690-697.

Bertrand, W. A., R. L. Hite, and D. M. Day. 1996. Biological stream characterization

(BSC): Biological assessment of Illinois stream quality through 1993. Illinois

Environmental Protection Agency, Bureau of Water, Springfield, IL.

IEPA/BOW/96-058.

Brenden, T.O., L.Wang, P.W. Seelbach, R.D. Clark, Jr., M.J.Wiley, and B.L. Sparks-

Jackson. 2008. A spatially constrained cluster program for river valley segment

delineation from GIS digital river networks. Environmental Modeling and

Software 23: 638-649.

DeWalt, R. E., C. Favret, and D.W.Webb. 2005. Just how imperiled are aquatic

insects? A case study of stoneflies (Plecoptera) in Illinois. Annals of the

Entomological Society of America 98(6): 941-950.

Diamond, J.M. and V. B. Serveiss. 2001. Identifying sources of stress to native aquatic

fauna using a watershed ecological risk assessment framework. Environmental

Science and Technology 35:4711-4718.

Favret, C. and R. E. DeWalt. 2002. Comparing the Ephemeroptera and Plecoptera

Specimen Databases at the Illinois Natural History Survey and Using Them to

Document Changes in the Illinois Fauna. Annals of the Entomological Society of

America 95(1): 35-40.

Fruend, J. G. and J. T. Petty. 2007. Response to fish and macroinvertebrate

bioassessment indices to water chemistry in a minded Appalachian watershed.

Environmental Management 39:707-720.

Herkert, J. R. and J. E. Ebinger. 2002. Endangered and Threatened Species of Illinois:

Status and Distribution. Volume 1: Plants. Illinois Endangered Species

Protection Board, Springfield, Illinois. 161 pp.

Hite, R. L., and W. A. Bertrand. 1989. Biological stream characterization (BSC): A

biological assessment of Illinois stream quality. Special Report #13 of the Illinois

State Water Plan Task Force. IEPA/AC/89-275.

Holtrop, A. M. and C. R. Dolan. 2003. Assessment of streams and watersheds in Illinois:

Development of a stream classification system and fish sampling protocols.

Aquatic Ecology Technical Report 03/15. 28 pp.

Hughes, R.M. and R. F. Noss. 1992. Biological diversity and biological integrity: current

concerns for lakes and streams. Fisheries 17:11-19.

31

Hunsaker, C.T. and D.A. Levine. 1995. Hierarchical approaches to the study of water

quality in rivers. BioScience 45:193-203.

Kilgour, B.W. and D. R. Barton. 1999. Associations between stream fish and benthos

across environmental gradients in southern Ontario. Freshwater Biology

41:553-566.

Lammert, M. and J. D. Allan. 1999. Assessing biotic integrity of streams: Effects of

scale in measuring the influence of land use/cover and habitat structure on fish

and macroinvertebrates. Environmental Management 23:257-270.

Omernick, J. M., Abernathy, A. R., and L. M. Male. 1981. Stream nutrient levels and

proximity of agricultural and forest land to streams: some relationships. Journal

of Soul and Water Conservation 36:227-231.

Page, L. M., K. S. Cummings, C. A. Mayer, S. L. Post, and M. E. Retzer. 1992.

Biologically Significant Illinois Streams: An Evaluation of the Streams of Illinois

based on Aquatic Biodiversity. Center for Biodiversity Technical Report 1992(1).

485 pp.

Paller, M. H. 2001. Comparison of fish and macroinvertebrate bioassessments from

South Carolina coastal plain streams. Aquatic Ecosystem Health and

Management 4:175-186.

Phillips, C. A., R. A. Brandon, and E. O. Moll. 1999. Field Guide to Amphibians and

Reptiles of Illinois. Illinois Natural History Survey Manual 8. 300 pp.

Seelbach, P.W., M. J.Wiley, J. C. Kotanchik, and M. E. Baker. 1997. A Landscape-

based Ecological Classification System for River Valley Segments in Lower

Michigan. Michigan Department of Natural Resources Fisheries Research

Report No. 2036. 51 pp.

Shreve, R. L. 1967. Infinite topologically random channel networks. Journal of Geology

75: 178-186.

Smart, M. M., T.W. Barney, and J. R. Jones. 1981. Watershed impact on stream water

quality: a technique for regional assessment. Journal of Soil and Water

Conservation 36:297-300.

Smogor, R. 2000. Draft manual for calculating Index of Biotic Integrity scores for

streams in Illinois. Prepared for: Illinois Environmental Protection Agency and

Illinois Department of Natural Resources. 23 pp.

Smogor, R. 2005. Draft manual for Interpreting Illinois Fish-IBI Scores. Prepared for:

Illinois Environmental Protection Agency. 26 pp.

Southerland, M. T., R. E. Jung, D. P. Baxter, E. C. Chellman, G. Mercurio and J. H.

Volstad. 2004. Stream salamanders as indicators of stream quality in Maryland,

USA. Applied Herpetology 2:23-46.

32

State of Illinois. 2005. The Illinois Comprehensive Wildlife Conservation Plan and

Strategy – version 1. Prepared by the Illinois Department of Natural resources

and approved by the U.S. Fish and Wildlife Service.

Stoddard, J.L., D.P. Larse, C.P. Hawkins, R.K. Johnson, R.H. Norris. 2006. Setting

expectations for the ecological condition of streams: the concept of reference

condition. Ecological Applications 16(4):1267-1276.

Szafoni, R.E. 2002. Freshwater mussel classification index. In Illinois natural areas

inventory - approved standards and guidelines. pp. 66-71.

Tetra Tech, Inc. 2007. Illinois benthic macroinvertebrate collection method comparison

and stream condition index revision. Prepared for: Illinois Environmental

Protection Agency. 11 pp.

United States Geological Survey. 2000. The National Hydrography Dataset: Concepts

and Contents. Available at:

http://nhd.usgs.gov/chapter1/chp1_data_users_guide.pdf

Appendix A. List of threatened and endangered speciesincluded in stream ratings.

Amphibians

Endangered

Spotted Dusky Salamander (Desmognathus conanti)

Crayfish

Endangered

Indiana Crayfish Orconectes indianensisKentucky Crayfish Orconectes kentuckiensisShrimp Crayfish Orconectes lanciferBigclaw Crayfish Orconectes placidus

Fish

Endangered

Lake Sturgeon Acipenser fulvescensWestern Sand Darter Ammocrypta clarumBluebreast Darter Etheostoma camurumHarlequin Darter Etheostoma histrioCypress Minnow Hybognathus hayiBigeye Chub Hybopsis amblopsPallid Shiner Hybopsis amnisNorthern Brook Lamprey Ichthyomyzon fossorSturgeon Chub Macrhybopsis gelidaGreater Redhorse Moxostoma valenciennesiRiver Chub Nocomis micropogonPugnose Shiner Notropis anogenusBigeye Shiner Notropis boopsBlacknose Shiner Notropis heterolepisTaillight Shiner Notropis maculatusWeed Shiner Notropis texanusNorthern Madtom Noturus stigmosusPallid Sturgeon Scaphirhynchus albus

Threatened

Eastern Sand Darter Ammocrypta pellucidumLongnose Sucker Catostomus catostomusCisco Coregonus artediGravel Chub Erimystax x-punctatus

33

Iowa Darter Etheostoma exileBanded Killifish Fundulus diaphanusStarhead Topminnow Fundulus disparLeast Brook Lamprey Lampetra aepypteraRedspotted Sunfish Lepomis miniatusBantam Sunfish Lepomis symmetricusRiver Redhorse Moxostoma carinatumIroncolor Shiner Notropis chalybaeusBlackchin Shiner Notropis heterodon

Mussels

Endangered

Spectaclecase Cumberlandia monodontaFanshell Cyprogenia stegariaSnuffbox Epioblasma triquetraPink Mucket Lampsilis abruptaWavy-rayed Lampmussel Lampsilis fasciolaHiggins Eye Lampsilis higginsiiOrangefoot Pimpleback Plethobasus cooperianusSheepnose Plethobasus cyphyusClubshell Pleurobema clavaOhio Pigtoe Pleurobema cordatumFat Pocketbook Potamilus capaxKidneyshell Ptychobranchus fasciolarisRabbitsfoot Quadrula cylindricaSalamander Mussel Simpsonaias ambiguaPurple Lilliput Toxolasma lividusRainbow Villosa iris

Threatened

Slippershell Alasmidonta viridisPurple Wartyback Cyclonaias tuberculataButterfly Ellipsaria lineolataElephant-ear Elliptio crassidensSpike Elliptio dilatataEbonyshell Fusconaia ebenaBlack Sandshell Ligumia rectaLittle Spectaclecase Villosa lienosa

Plants

Endangered

Heart-leaved Plantain Plantain cordata

34

Printed by the Authorityof the State of Illinois

XM - 8/08

IOCI0048-09