Methods for Evaluating Wetland Condition: Wetland ...€¦ · Methods for Evaluating Wetland Condition: Wetland Biological Assessment Case Studies. Office of Water, U.S. Environmental

United States EnvironmentalProtection Agency

Office of WaterWashington, DC 20460

EPA-822-R-03-013May 2003

Methods for evaluating wetland condition

#14 Wetland Biological AssessmentCase Studies



EPA-822-R-03-013May 2003

Methods for evaluating wetland condition

#14 Wetland Biological AssessmentCase Studies

Edited By

Maine Department of Environmental ProtectionThomas J. Danielson

Connecticut Department of Environmental ProtectionDouglas G. Hoskins

Prepared jointly by:

The U.S. Environmental Protection AgencyHealth and Ecological Criteria Division (Office of Science and Technology)

and

Wetlands Division (Office of Wetlands, Oceans, and Watersheds)



EPA-822-R-03-013May 2003

Notice

The material in this document has been subjected to U.S. Environmental Protection Agency (EPA)technical review and has been approved for publication as an EPA document. The informationcontained herein is offered to the reader as a review of the “state of the science” concerning wetlandbioassessment and nutrient enrichment and is not intended to be prescriptive guidance or firm advice.Mention of trade names, products or services does not convey, and should not be interpreted asconveying official EPA approval, endorsement, or recommendation.

Appropriate Citation

U.S. EPA. 2003. Methods for Evaluating Wetland Condition: Wetland Biological AssessmentCase Studies. Office of Water, U.S. Environmental Protection Agency, Washington, DC.EPA-822-R-03-013.

Acknowledgments

EPA acknowledges the contributions of the following people in the writing of this module: The membersof BAWWG—without their hard work, this report would not have possible. Leigh Dunkelbergerprovided invaluable assistance during her internship with the U.S. EPA Wetlands Division.

This entire document can be downloaded from the following U.S. EPA websites:

http://www.epa.gov/ost/standards

http://www.epa.gov/owow/wetlands/bawwg

ii

iii

Contents

Foreword .................................................................................................................. vii

List of “Methods for Evaluating Wetland Condition” Modules ............. ix

Florida: Monitoring Across a Nutrient Gradient in the Everglades ..... 1

Florida: Development of a Biological Approach forAssessing Florida Wetland Integrity ............................................................... 7

Maine: Developing a Statewide Biological Monitoringand Assessment Program for Freshwater Wetlands................................ 14

Maryland: Developing an IBI Assessment for RestoredWetlands in the Mid-Atlantic States ............................................................... 18

Massachusetts: Use of Multimetric Indices to Examine EcologicalIntegrity of Salt Marsh Wetlands in Cape Cod ..........................................22

Massachusetts: Involving Volunteers in Examining theEcological Integrity of Coastal Wetlands in Cape Cod,Massachusetts .......................................................................................................30

Michigan: Bioassessment Procedures and Baseline ReferenceData for Great Lakes Coastal Marshes and Inland ForestedWetlands in Michigan ...........................................................................................32

Minnesota: Developing Wetland Biocriteria ...............................................37

Minnesota: Dakota County Wetland Health Evaluation Project ........44

Minnesota: University of Minnesota’s IBI Development ............................ 47

Montana: Developing Wetland Bioassessment Protocols ToSupport Aquatic Life Beneficial Use-Support Determinations ...............49

North Dakota: Wetland Bioassessment Protocols for MakingAquatic Life Beneficial Use-Support Determinations ................................59

Ohio: Developing IBI Assessment Methods for Water QualityStandards and Regulatory Decisions .............................................................63

Oregon: Simultaneous Development, Calibration, and Testingof Hydrogeomorphic-Based (HGM) Assessment Procedures andBiological Assessment Procedures ...............................................................72

iv

Pennsylvania: Assessing Wetland Condition at a WatershedScale Using Hydrogeomorphic Models and Measures ofBiological Integrity ..............................................................................................75

Vermont: Classification, Biological Characterization, andBiometric Development for Northern White Cedar Swampsand Vernal Pools ..................................................................................................77

Washington: King County Wetland-Breeding AmphibianMonitoring Program .............................................................................................80

Wisconsin: Developing Biological Indexes for Wisconsin’s Palustrine Wetlands ...........................................................................................82

Wisconsin: Refinement and Expansion of the Wisconsin WetlandBiological Index for Assessment of Depressional PalustrineWetlands ..................................................................................................................86

References .............................................................................................................. 91

Glossary...................................................................................................................93

List of Figures

Figure 1: Area of interest within Everglades protection area ............ 1

Figure 2: WCA 2A monitoring sites along the phosphorus gradient .. 2

Figure 3: Change point analyses of Eleocharis frequency ofoccurrence and biomass data along the SFWMDtransects.............................................................................................. 5

Figure 4: Results of change point analyses performed onmedian total percentage of pollution-sensitive(literature determined) periphyton taxa .................................... 5

Figure 5: Results of change point analyses onmedian Florida index values ............................................................ 6

Figure 6: Florida’s wetland regions and field site locations ............... 9

Figure 7: Relationships between wetland ecological conditionand land use .......................................................................................27

Figure 8: Two sets of multimetric goals for evaluatingrestoration ........................................................................................28

v

Figure 9: Locations of smaller depressional wetlandssampled in 1995 ................................................................................39

Figure 10: Locations of the 44 large depressional wetlandssampled in 1999 ................................................................................39

Figure 11: Montana ecoregions and pilot watersheds ............................ 51

Figure 12: Ecoregions and sampling locations by wetland class .......52

Figure 13: Diatom data related to abiotic factors usingDetrended Canonical Correspondence Analysis.................54

Figure 14: Clusters of wetlands based on diatom data ..........................54

Figure 15: Macroinvertebrate index scores for wetlandsin several classes ...........................................................................57

Figure 16: Level IV ecoregions of North Dakota ....................................... 61

Figure 17: Ecological regions of Ohio and Indiana ..................................64

List of Plates

Plate 1: Cypress wetland .............................................................................. 12

Plate 2: Macroinvertebrate sampling ....................................................... 13

Plate 3: Maine’s stovepipe sampler and sieve bucket ............................ 16

Plate 4: Picking macroinvertebrates from multihabitat sample ....... 16

Plate 5: Cape Cod, MA, salt marsh plant survey ....................................29

Plate 6: Cape Cod, MA, salt marsh nekton: Watch your fingers!(blue crab, Callinectes sapidus) .................................................29

Plate 7: Impaired wetland ..............................................................................66

Plate 8: Least-disturbed wetland ...............................................................66

vi

List of Tables

Table 1. Scoring Criteria for 10 Invertebrate Metrics for IBIfor Large Depressional Wetlands ............................................42

Table 2. Scoring Criteria for 10 Vegetation Metrics for LargeDepressional Wetlands .................................................................43

Table 3: Proposed metrics, proposed metric calculations,and score calculations used for developing wetlandmacroinvertebrate indices............................................................56

vii

Foreword

In 1999, the U.S. Environmental Protection Agency (EPA) began work on this series of reports entitledMethods for Evaluating Wetland Condition. The purpose of these reports is to help States andTribes develop methods to evaluate (1) the overall ecological condition of wetlands using biologicalassessments and (2) nutrient enrichment of wetlands, which is one of the primary stressors damagingwetlands in many parts of the country. This information is intended to serve as a starting point for Statesand Tribes to eventually establish biological and nutrient water quality criteria specifically refined forwetland waterbodies.

This purpose was to be accomplished by providing a series of “state of the science” modules concerningwetland bioassessment as well as the nutrient enrichment of wetlands. The individual module formatwas used instead of one large publication to facilitate the addition of other reports as wetland scienceprogresses and wetlands are further incorporated into water quality programs. Also, this modularapproach allows EPA to revise reports without having to reprint them all. A list of the inaugural set of 20modules can be found at the end of this section.

This series of reports is the product of a collaborative effort between EPA’s Health and EcologicalCriteria Division of the Office of Science and Technology (OST) and the Wetlands Division of theOffice of Wetlands, Oceans and Watersheds (OWOW). The reports were initiated with the supportand oversight of Thomas J. Danielson (OWOW), Amanda K. Parker and Susan K. Jackson (OST),and seen to completion by Douglas G. Hoskins (OWOW) and Ifeyinwa F. Davis (OST). EPA reliedheavily on the input, recommendations, and energy of three panels of experts, which unfortunately havetoo many members to list individually:

Biological Assessment of Wetlands Workgroup

New England Biological Assessment of Wetlands Workgroup

Wetlands Nutrient Criteria Workgroup

More information about biological and nutrient criteria is available at the following EPA website:

http://www.epa.gov/ost/standards

More information about wetland biological assessments is available at the following EPA website:

http://www.epa.gov/owow/wetlands/bawwg

ix

List of “Methods for Evaluating WetlandCondition” Modules

1 ................. Introduction to Wetland Biological Assessment

2 ................. Introduction to Wetland Nutrient Assessment

3 ................. The State of Wetland Science

4 ................. Study Design for Monitoring Wetlands

5 ................. Administrative Framework for the Implementation of a

Wetland Bioassessment Program

6 ................. Developing Metrics and Indexes of Biological Integrity

7 ................. Wetlands Classification

8 ................. Volunteers and Wetland Biomonitoring

9 ................. Developing an Invertebrate Index of Biological

Integrity for Wetlands

10............... Using Vegetation to Assess Environmental Conditions

in Wetlands

11 ............... Using Algae to Assess Environmental Conditions in

Wetlands

12 ............... Using amphibians in Bioassessments of Wetlands

13 ............... Biological Assessment Methods for Birds

14 ............... Wetland Bioassessment Case Studies

15 ............... Bioassessment Methods for Fish

16 ............... Vegetation-Based Indicators of Wetland Nutrient

Enrichment

17 ............... Land-Use Characterization for Nutrient and Sediment

Risk Assessment

18............... Biogeochemical Indicators

19 ............... Nutrient Load Estimation

20 .............. Sustainable Nutrient Loading

Module # Module Title

1

Florida: Monitoring Across a NutrientGradient in the Everglades

Contact Russel Frydenborg

Organization Florida Department of Environmental Protection2600 Blair Stone Road, MS 6511Tallahassee, FL 32399-2400

Phone (850) 921-9821

E-mail [email protected]

Purpose

T his project was initiated to monitor biologi-cal assemblages across a nutrient gradient in

the Florida Everglades in support of regulatory ef-forts to define a numeric water quality criterion forphosphorus. The goal is protection of natural popu-lations of aquatic flora and fauna in the EvergladesProtection Area.

Wetland Type

Freshwater marshes

Assemblages

Algae

Macroinvertebrates

Vascular plants

Status

Revising sampling methods and analyzing data

Project Description

The historic Florida Everglades consisted of ap-proximately 4 million acres of shallow sawgrassmarsh, with wet prairies and aquatic sloughs inter-

spersed with tree islands. Today, only 50% of theoriginal Everglades ecosystem remains, primarily asa result of drainage and conversion of large por-tions of the northern and eastern Everglades to ag-ricultural or urban land use. The remaining portionsof the historic Everglades are located in the WaterConservation Areas (WCAs) and Everglades Na-tional Park (ENP) (see Figure 1).

Figure 1: Area of interest withinEverglades protection area.

The Everglades ecosystem evolved under ex-tremely low phosphorus concentrations and is con-sidered an oligotrophic ecosystem. A large bodyof evidence indicates that phosphorus is the pri-mary limiting nutrient throughout the remainingEverglades. Introduction of excess phosphorus tothe Everglades has resulted in ecological changesover large areas of the marsh. The Everglades For-

2

ever Act (EFA; Section 373.4592, Florida Stat-utes), passed by the Florida Legislature in 1994,stated that waters flowing into the part of the rem-nant Everglades known as the Everglades Protec-tion Area (defined as Water Conservation Areas 1,2A, 2B, 3A, 3B, and ENP) contain excessive lev-els of phosphorus and that a reduction in levels ofphosphorus will benefit the ecology of the Ever-glades Protection Area. The EFA requires theFlorida Department of Environmental Protection(FDEP) and the South Florida Water ManagementDistrict (SFWMD) to complete research necessaryto establish a numeric phosphorus criterion for theEverglades Protection Area.

The SFWMD Everglades System Research Di-vision (ESRD) initiated a succession of studies, be-ginning in 1993 and continuing to the present, aspart of the research and monitoring being conductedin the Everglades for phosphorus criterion devel-opment. Biological monitoring for the ESRD stud-ies was initiated in early 1994 in WCA 2A. Datafrom this and other studies are being used by FDEPin the development of a numeric phosphorus crite-rion for the Everglades Protection Area.

Study Design

SFWMD ESRD initially selected 13 sites along 2transects located downstream of canals discharg-ing into WCA 2A and extending down the phos-phorus gradient into the least affected areas of themarsh. Sampling sites ranged from the canal in-flows (discharge structures on the northeasternmargin of WCA 2A) to a site nearly 15 km down-stream from the canal inflows. Three of the 13 mainsites (sites U1-U3) were specifically chosen to rep-resent the least affected area of WCA 2A with re-spect to anthropogenic disturbance. A series of 15additional “intermediate” sites were added to thestudy later to obtain better spatial coverage of thelower portion of the transects. The sites have beenmonitored for water, sediment, and biological qual-ity. Figure 2 shows the WCA 2A monitoring siteslocated along the phosphorus gradient.

Sampling Methods: Algae

Water Bottles: Phytoplankton samples initiallywere collected monthly and later collectedquarterly using water bottles. Samples werepreserved in the field and sent to the FDEPCentral Biology Laboratory for taxonomic iden-tification.

Diatometers: Racks each containing six glassdiatometer slides were deployed quarterly ateach site. It was determined that an 8-weekdeployment was necessary for sufficient per-iphyton growth. Diatometers were collectedand preserved and sent to the FDEP CentralBiology Laboratory for processing and taxo-nomic identification.

Natural Substrate (benthic): Samples ofbenthic periphyton were collected from surficialsediment cores at the main transect sites on sev-eral occasions. Samples were retained bySFWMD ESRD for processing and taxonomicidentification.

Analytical Methods: Algae

Water Bottles: Samples were processed andenumerated by FDEP Central Biology Labo-

Figure 2: WCA 2A monitoringsites along the phosphorus

gradient.

3

ratory staff according to FDEP standard oper-ating procedures (SOPs) (e.g., AB-04 and AB-05; available at http://www8.myflorida.com/labs/sop/index.htm). Analysis from this studyand other studies has indicated that Evergladesphytoplankton are largely periphyton that hassloughed off into the water column. Thus, algaldata analysis was focused on the periphytondata.

Diatometers: Samples were processed andenumerated by FDEP Central Biology Labo-ratory staff according to FDEP SOPs (e.g.,AB-02, AB-02.1, AB-02.2, and AB-03; avail-able at http://www8.myflorida.com/labs/sop/index.htm).

Natural Substrate (benthic): SFWMD pro-cessed and enumerated natural substratesamples.

Sampling Methods:Macroinvertebrates

Dipnet: SFWMD staff conducted quarterlymacroinvertebrate sampling using a standard D-frame dipnet with a 30-mesh bag from Sep-tember 1994 through November 1995. Thesampling method consisted of the collection of20 0.5 m (in length) discrete dipnet sweeps fromrepresentative habitats in the area of each siteon a given sampling date. The 20 dipnet sweepsfor a given site were combined and sent to theFDEP Central Biology Laboratory for process-ing and taxonomic identification.

Quan Net: Beginning in May 1996, SFWMDstaff conducted quarterly macroinvertebratesampling using the Quan Net method. The sam-pling method consisted of the deployment of a1-m2 frame at the site and the collection of netsamples and all vegetation within the area ofthe frame. Frames were deployed in each ofseveral representative habitats, where present,in the vicinity of each site. Samples from eachsite/habitat were kept separate. Representa-

tive habitats were labeled as cattail, sawgrass,or slough, depending on the predominant veg-etation type. The collected material from eachsite/habitat was subsampled, preserved, andsent to the FDEP Central Biology Laboratoryfor processing and taxonomic identification.

Hester-Dendy: SFWMD staff deployedHester-Dendy artificial substrate samplers ateach of the main transect sites on a quarterlybasis. The samplers were deployed for a 1-month period, after which they were collected,preserved, and sent to the FDEP Central Biol-ogy Laboratory for processing and taxonomicidentification.

Analytical Methods:Macroinvertebrates

Dipnet and Quan Net: FDEP Central Biol-ogy Laboratory staff subsampled the dipnetand Quan Net samples from each site and anal-yzed them according to FDEP SOPs(e.g., IZ-02 and IZ-06; available athttp://www8.myflorida.com/labs/sop/index.htm).

Hester Dendy: FDEP Central Biology Labo-ratory staff processed and analyzed the Hester-Dendy samples from each site according toFDEP SOPs (e.g., IZ-03 and IZ-06; availableat http://www8.myflorida.com/labs/sop/index.htm).

Sampling Methods: Macrophytes

Macrophyte Stem Density and Frequency: InApril 1997, SFWMD staff conducted a studyof macrophytes at the WCA 2A transect sites.A 50-m tape was laid out at each transect site.A 1-m square frame was used every 2 m alongthe tape to delineate the sample area for calcu-lation of macrophyte stem densities (stems/m2)and frequencies (# plots where a species wasfound/total # of plots) by species.

4

Macrophyte Harvesting: On the other sideof the 50-m tape used for establishing stem den-sities and frequencies, SFWMD staff harvestedmacrophytes for biomass measurements, usingthe 1-m square frame at five predetermined lo-cations to mark the sample area for harvesting.

Analytical Methods:Macrophytes

Macrophyte Stem Density and Frequency:Stem densities (stems/m2) and frequencies (#plots where a species was found/total # of plots)by species were counted at each site.

Macrophyte Harvesting: SFWMD staff con-ducted biomass analysis of the harvested mac-rophytes for comparison of the relative biom-ass of several species present at each of theWCA 2A transect sites (e.g., Eleocharis,Nymphaea, Typha).

Lessons Learned

Periphyton, macroinvertebrate, and macrophytecommunities in WCA 2A change substantially fromreference conditions at approximately 7 to 8 kmdownstream of canal discharges into WCA 2A (seeFigures 3–5). Data analysis has shown that bio-logical populations at the two stations (E5 and F5)nearest to the three initial reference sites (U1-U3)are very similar in terms of biological communitystructure. This suggests that these areas, despiteslight phosphorus enrichment, still support referencecondition biota. The somewhat higher phosphorusregimes at the next stations (E4 and F4 and be-yond) are associated with greater biological changes.

Experimental field dosing studies (microcosms) havebeen conducted by SFWMD ESRD and show thatthe addition of phosphorus causes changes in per-iphyton assemblages consistent with those observedin the transect study.

The WCA 2A transect periphyton data for eachsite/date have been analyzed using the entire taxo-nomic assemblage encountered and using lists ofpollution-sensitive and -tolerant species based onavailable literature and experimental phosphorusaddition studies (the microcosms) in WCA 2A.Macroinvertebrate data have been analyzed usingthe Florida Index and the macroinvertebrate com-ponent of the Lake Condition Index (LCI), mea-sures of the numbers of pollution-sensitive taxa in asample that are routinely used by FDEP inbioassessments of streams and lakes. The use ofthese methods with the WCA 2A transect data hasdemonstrated a clear signal of biological disturbancealong the nutrient gradient in WCA 2A. FDEP isusing this information as well as information fromother studies conducted in the Florida Evergladesto develop a numeric phosphorus criterion for theEverglades Protection Area.

Additional Comments

The information provided here is based solely onthe transect study by SFWMD ESRD in WCA 2A.Research and monitoring of Florida Evergladeswater, sediment and biological quality is being con-ducted by several research groups in WCA 2A,WCA 1 (Arthur R. Marshall Loxahatchee NationalWildlife Refuge), Everglades National Park (ENP),and WCA 3A, including studies by SFWMD ESRDsimilar to the WCA 2A transect study.

5

Figure 3: Change point analyses of Eleocharis frequency ofoccurrence and biomass data along the SFWMD transects.

Collected April 1997.

Figure 4: Results of change point analyses performed onmedian total percentage of pollution-sensitive

(literature determined) periphyton taxa.

6

Figure 5: Results of change point analyses onmedian Florida index values.

7

Florida: Development of a BiologicalApproach for Assessing Florida

Wetland Integrity

Contact Mark Brown

Organization University of Florida, Center for WetlandsPhelps Lab – PO Box 116350Gainesville, FL 32611-6350

Phone (352) 392-2309


Website http://www.enveng.ufl.edu/homepp/brown/syseco

Purpose

T o develop “bioindicators” of ecosystemhealth for wetlands in Florida. To achieve this

goal, the project team has developed wetlandecoregions using GIS technology, a classificationscheme for Florida wetlands, sampling protocolsfor herbaceous and forested wetlands, and a quan-titative index of the human disturbance gradient.

Wetland Type

Freshwater marshes

Forested wetlands

Assemblages

Algae

Macroinvertebrates

Vascular plants

Status

Developed field protocols, sampled more than 150wetlands statewide, developed candidate plant,macroinvertebrate, and algae metrics for marshes.Currently sampling and developing metrics for for-ested wetlands.

Project Description

The University of Florida Center for Wetlands isinvolved in a multiyear wetland research projectfunded by the Florida Department of Environmen-tal Protection (FDEP) to develop an integrated bio-logical approach for evaluating Florida’s wetlands.The project goal is to develop an assessment ap-proach that recognizes the utility of both biologicaland functional assessments, and is rapid, reproduc-ible, and meaningful.

The Center for Wetlands began development ofthe assessment approach in 1997. Now in its fifthyear, the project, titled “Development of a Biologi-

8

cal Approach for Assessing Wetland Function andIntegrity,” has four main tasks:

Task 1. Review and synthesize all current andrelevant literature.

Task 2. Develop a wetland classification systemfor wetland types in the State of Florida.

Task 3. Develop a GIS-driven methodology forclassifying bioregions within the State of Florida thatidentify climatic, geologic, and geophysical prov-inces that are sensitive to wetland classes.

Task 4. Develop a bioassessment methodology,biocriteria, and metrics for wetlands in the State ofFlorida.

To date, Tasks 1–3 have been completed, andwork continues on Task 4. Nearly complete is aset of metrics for herbaceous depressional wetlands,and data are currently being analyzed to developmetrics for forested depressional wetlands.

The development of the approach has included areview of technical and scientific literature, a wet-land classification and crosswalk, wetlandregionalization of Florida, and two wetland biologi-cal surveys in summer 1998 and summer 1999. Thebiological surveys were designed to test the validity(and necessity) of the proposed wetland regions,to identify the appropriate biological indicator taxa,and to quantify the gradient of human disturbance.

Study Design

The approach has included a review of technicaland scientific literature, development of a wetlandclassification system and crosswalk with other clas-sifications, wetland regionalization of Florida, and4 years of wetland biological surveys in the grow-ing seasons of 1998, 1999, 2000, and 2001 (a fifthseason of field surveys is under way for 2002). Thebiological surveys were designed to test the validity(and necessity) of the proposed wetland regions,to identify the appropriate biological indicator taxa,and to quantify the gradient of human disturbance.

The wetland classification scheme was organizedwith major classes defined from three variables: (1)dominant vegetation (forested, shrub, herbaceous);(2) geomorphic position (stream channel [flood-plain], flat topography, sloped topography, lakefringe, depressional); and (3) primary water source(rainfall, surface water, ground water). Subclassesare discriminated by modifiers (hydroperiod andplant community associations). Eleven wetlandclasses were identified: forested (river swamp,slough/strand/seepage swamp, lake swamp, depres-sion swamp, flatland swamp); shrub dominated(shrub-scrub swamp); and herbaceous (river marsh,wet prairie, lake marsh, depressional marsh, seep-age marsh). An HTML electronic database has beencompleted that crosswalks existing wetland classi-fications to the new simplified classification schemedeveloped for the bioassessment project.

Regionalization of the State was necessary be-cause there is significant variation in climatic andphysical features of the Florida peninsula and it wasbelieved that these regional differences would equateto variations in bioindicator “signals.” Map cover-ages of physical and climatic variables of the Floridalandscape were used to develop regions that haddifferent characteristic driving energies and land-scape structural characteristics.

The map coverages were combined with GIS mapalgebra to create a spatial hydrologic budget equa-tion for the State. The equation modeled the move-ment of water on the landscape during the ecologi-cally sensitive spring growing season. The outputof the model provided a value for a Potential SoilMoisture Index (PSMI). The PSMI was separatedinto four regions based on both the critical depth ofsaturation and on a statistical clustering of the PSMIvalues (see Figure 6). The classified regions weretested for similarities and differences to determine ifbetween-region variation in wetland type and envi-ronmental variables was greater than within-regionvariations.

9

The importance of hydrology in determining wet-land type and location (the premise behind thePSMI) was then tested using a hierarchical classifi-cation technique (TWINSPAN) and ordination(DCA and CCA) with variables of seasonal andannual rainfall, seasonal and annual potential evapo-transpiration, slope, geology, drainage class, andrunoff. TWINSPAN, DCA, and CCA tested therelationships between wetland type and climatic andphysiographic landscape characteristics. Based onthe geostatistical output, hydrology is indeed a ma-jor determinant of wetland type and location, andsupports the use of the spatial hydrologic budgetequation in delineating wetland regions.

Four years of biological surveys of wetlandsthroughout the State were designed to test theregionalization techniques as well as the metric andbioindicator development. The 1998 pilot field re-search involved surveying 24 herbaceous and for-ested depressional wetlands in north and central

Florida. Major taxonomic assemblages were char-acterized and ranked along a gradient of distur-bances. Sites were located within multiple land usessuch as parks, preserves, pastures, farm fields, wellfields, silviculture plots, and urban areas. Impactsthat were assessed included hydrological modifica-tions, nutrient loading, and altered hydroperiod. Thefirst year of sampling resulted in development ofstandardized sampling procedures, design andimplementation of a statewide sampling program,and identification of community attributes and can-didate metrics.

In the second and third field seasons, 77 herba-ceous, depressional wetlands were surveyed in 3regions (south, central, and north). Approximatelyhalf of the wetlands were impacted (agricultural set-ting) and half were reference locations. Many ofthe sites were paired sites (impacted and referenceat close proximity).

Figure 6: Florida’s wetland regions and field site locations.

10

In the fourth season, 72 forested, depressionalwetlands were surveyed in four regions (18 sites ineach of four regions: north, central, and southernpeninsular as well as the panhandle; see Figure 6).Approximately ? of the wetlands were impacted (?in agricultural settings and ? in urban settings) and ?were reference locations. Many of the sites werepaired sites (impacted and reference at close prox-imity).

Quantifying gradients of human disturbanceA Land Development Intensity (LDI) index is be-

ing used to quantify gradients of human disturbancefor wetlands throughout the State. The LDI indexis calculated using land use/land cover characteris-tics, from aerial photographs, of lands within a 100-m buffer surrounding the wetland. Land uses in thearea surrounding a wetland are first characterizedand then an intensity factor is assigned to each landuse type. The LDI algorithm multiplies the percentarea of each land use/land cover in the surrounding100-m buffer by intensity coefficients. The inten-sity coefficients are scaled from 1 to 10 and repre-sent intensity of environmental manipulation as mea-sured by energy use per unit area per unit time. TheLDI index is such that lower LDI values are indica-tive of a lower disturbance level.


MaterialsThree 100-mL collection jars

1-L collection jar

Bottomless collection cup, knife, large pipette,bag, brush/scraper

M3 preservative

1 L of deionized water (for dry sites)

Methods1. Methods vary depending on whether there is

water present at the site. We are experimen-tally collecting dry benthic algae samples when

wetlands are dry. At this time, there is no infor-mation as to the usefulness of this sampling tech-nique. Methods outlined below are for wetsites.

2. Samples are taken depending on substrate andseparated as benthic, epiphyton, metaphyton,and phytoplankton.

3. For epiphyton, the 10 aliquots are dividedequally among herbaceous and woody debris:

a. For herbaceous debris, plant stems are cutfrom the soil to the water surface, andplaced in a zip-lock bag with some wet-land water. The plant matter is shaken andmassaged thoroughly to dislodge the al-gae. Using a large pipette, 10 mL of sampleare extracted into collection jar.

b. For woody debris, a brush is used to dis-lodge algae. If the debris is small enough,it is placed in a bag, similar to the herba-ceous methods above. If the debris islarger, a bottomless collection cup is usedto confine the sample, and it is brushedwhile under water.

4. For benthic algae, a bottomless collection cupis used to isolate a spot on the sediment. Thena large pipette is used to gently stir the surface(top 1 cm) of the sediment, and extract a 10-mL aliquot that is placed in a collection jar.Sampling is repeated at different locations untila total of 100 mL is collected.

5. For metaphyton, a thumbnail size portion of thealgal mat is collected from 10 different loca-tions throughout the wetland.

6. For phytoplankton, a total 100 mL of surfacewater is collected at 10 locations throughoutthe wetland.

7. All samples are preserved with M3 using 5 mLper 100 mL of sample.

8. Samples are sent for later laboratory identifica-tion.

11


MaterialsField Physical/Chemical Characterization DataSheet

Habitat Assessment Sheet

D-Frame dipnet with no. 30 mesh

4-L sample jars

Buffered formalin

Methods1. The wetland is visually examined by walking

throughout the wetland, paying close attentionto its physical and habitat characteristics, and aField Physical/Chemical Characterization DataSheet and Habitat Assessment Sheet is com-pleted. The percent coverage of substrate typeand each habitat type is recorded.

2. A total of 20 sweeps with a D-frame dipnet aredivided between the habitat types based on theirpercentage of total wetland area. Discrete 0.5-m sweeps are performed with the dipnet. Theavailable substrates are sampled as determinedby the above procedures in the followingmanner:

a. For areas without flow, an area of sub-strate that is one dipnet width wide andapproximately 0.5-m long is disturbed bysweeping the net over the area three timesto ensure the capture of organisms.

b. For heavily vegetated areas, the net isjabbed into the base of the vegetation, dig-ging down to (but not into) the substrate,and dislodging organisms using a 0.5-msweeping motion with the net.

c. Sand, muck, mud, and silt (nonmajor habi-tats) are sampled by taking 0.5-m sweepswith the net while digging into the bottomapproximately 1 cm.

3. The number of sweeps for each habitat is re-corded on the Field Physical/Chemical Char-acterization Data Sheet.

4. The collected samples are reduced in volumeafter each discrete sample by dislodging organ-isms from larger debris (but retaining inverte-brates in the net or sieve) and discarding thedebris. The finer debris and organisms are savedin sample jars.

5. Samples are preserved with buffered formalinand shipped to the FDEP Laboratory for iden-tification.

Sampling Methods: Plants

MaterialsField data sheets

Compass

100 m of tape

FDEP’s Florida Wetland Plant IdentificationManual (Tobe et al. 1998)

Methods1. Using a compass (or the GPS unit) four line-

transects are located from four cardinal pointdirections (N, S, E, W) that run parallel to theslope of the wetland, beginning at the uplandedge (0 m) and extending into the center of thewetland.

2. At the beginning of each transect, the approxi-mate wetland/upland edge is located using acombination of wetland plants and hydric soils.

3. Preferably, all four transects are set at one time,with N, S and E, W transects meeting perpen-dicularly in the center, and dividing wetland intofour equal sections. Each transect is startedwith the 0-m point at the wetland–upland edgeand increasing in distance toward the wetlandinterior. A minimum of four data sheets is neededper site.

12

4. Two team members walk along each transectand record species presence of all plants within0.5 m on either side of each 5-m interval (sam-pling area = 5 m2).

5. Plant species names are recorded on the datasheets with full genus and species if known.

6. If unknown, the species is given a unique codeidentifying transect location and number of un-known encountered (N-1, N-2, N-3, E-1, E-2, etc.). Voucher specimens for all unknownspecies are collected, making sure to includeplant inflorescence and roots. Each specimenis tagged with properly labeled masking tape,and put in collection bag then pressed for laterlaboratory identification.

7. Plant nomenclature follows FDEP’s FloridaWetland Plant Identification Manual (Tobe etal. 1998).

Sampling Methods:Water Quality and Soils

Water quality samples are collected, preserved,and immediately sent to the chemistry laboratoryfor analysis. A composite soil sample is taken fromeach vegetation zone and later analyzed for nutri-ents, organic matter content, and physical proper-ties.

Analytical Methods: Plants

Field data are entered into an MS Access data-base for analysis. After entry, each data sheet ischecked by a second technician.

Plate 1: Cypress wetland. Photo: Elizabeth Spurrier

13

Plate 2: Macroinvertebrate sampling. Photo: Kelly Chinners Reiss

14

Maine: Developing a Statewide BiologicalMonitoring and Assessment Program

for Freshwater Wetlands

Contact Jeanne DiFranco

Organization Maine Department of Environmental Protection312 Canco RoadPortland, ME 04103

Phone (207) 822-6424


Website http://www.state.me.us/dep/blwq/monitoring.htm#programs

Purpose

Examine regional differences in wetlandmacroinvertebrate and algal assemblages

Test and refine candidate biological metrics

Diagnose stressors and identify risks to wet-lands from human activities

Develop impairment thresholds and biocriteriafor use in State water quality standards

Wetland Type

Freshwater marshes

Assemblages

Macroinvertebrates

Algae

Status

Expanding to Statewide program on a 5-year ro-tating basin schedule, testing and refining metrics,analyzing data, and developing impairment thresh-olds.

Project Description

In 1998, the Maine Department of EnvironmentalProtection (DEP) began development of a biologi-cal monitoring and assessment program for fresh-water wetlands. Between 1998 and 2000, DEPconducted a pilot study in the Casco Bay water-shed to develop wetland biomonitoring protocolsand identify candidate metrics related to wetlandcondition. During 2001 and 2002, DEP expandedmonitoring to the Saco, Piscataqua, and KennebecRiver watersheds using the methods developed inthe pilot study. DEP uses aquatic macro-inverte-brates as the primary taxonomic assemblage for thisprogram. Algae and diatoms are also collected aspart of a collaborative pilot project undertaken byDr. R. Jan Stevenson of Michigan State Universityto develop algal indicators of wetland integrity.Based on the results of Dr. Stevenson’s work, MaineDEP will evaluate the usefulness of algae to its wet-land biomonitoring program. Assessment of algaehas not been formally integrated into the Maine pro-gram at this time, however.

15

Study Design

The Maine wetland biomonitoring initiative hasbeen incorporated into DEP’s existing BiologicalMonitoring Program, and will be extended to as-sess additional major watersheds Statewide. Wet-land monitoring is currently coordinated with theState’s river and stream program using the follow-ing 5-year rotating basin schedule:

Kennebec, midcoast watersheds 2002

Androscoggin watershed 2003

St. John, Presumpscot watersheds 2004

Saco, southern coastal watersheds 2005

Penobscot, downeast watersheds 2006

Considerations for site selection include hydro-logic regime, geographic distribution of sites, land-scape position, human disturbance gradient, man-agement significance, and accessibility. All wetlandssampled are semipermanently or permanently inun-dated, and range from minimally disturbed poten-tial reference sites to poor-quality wetlands. As of2002, DEP has conducted wetland biomonitoringat 88 different sites encompassing 115 samplingevents. Some sites have been sampled repeatedlyover multiple years.


Macroinvertebrates are currently sampled duringJune and early July. Three different approaches havebeen tested to develop both qualitative and quanti-tative methods. In addition, water samples are ana-lyzed for a suite of physical and chemical param-eters to help in wetland characterization, and to iden-tify potential sources of human impact. Habitat in-formation, dominant plant species, and a scoring ofhuman disturbances are recorded in the field, alongwith measurements of water temperature, pH, dis-solved oxygen, and conductivity.

Multihabitat samplingA qualitative, multihabitat sampling approach was

tested, with the goal of developing a screening levelassessment tool. A standard D-frame net was usedto sample all inundated microhabitats at each site,including emergent vegetation, aquatic macrophytebeds, pools, and channels. Samples were “picked”or sorted from detritus in the field. One to severalorganisms representing each different taxon foundwere placed into a vial of alcohol until no differenttaxa were observed.

Stovepipe samplerMaine DEP designed its own stovepipe sampler

for quantitative samples using a 5-gallon bucket withthe bottom removed. In this method, the samplerwas used to enclose three replicate plots to restrictthe movement of organisms. The stovepipe sam-pler was pressed into the wetland substrate, andthe contents of the sampler were then agitated.Vegetation and surface sediment were placed intoa sieve bucket. The sampler was then swept 10times with a small hand net. Large pieces of veg-etation were washed and discarded; however, finerplant material and detritus were retained. Sampleswere preserved for later sorting and taxonomicanalysis in the laboratory.

Dipnet measured sweepA standard D-frame net is currently used to ob-

tain a semiquantitative sample. A sample is col-lected by submersing the net and sweeping throughthe water column for a distance of 1 meter. The netis bumped against the bottom substrate three times(at the beginning, middle, and end of the sweep) todislodge and collect organisms from the sediment.All material collected is placed in a sieve bucket.Large pieces of vegetation are washed and dis-carded; however, finer plant material and detritusare retained. Three replicate samples are collectedin areas of emergent vegetation. Samples are pre-served for later sorting and taxonomic analysis inthe laboratory.

16


Analyses performed to date reveal significant re-lationships between a number of candidate inverte-brate metrics and watershed development. Manyinvertebrate metrics tested also appear to respondto changes in water quality typically associated withurban nonpoint source pollution, including elevated

conductivity and concentrations of anions, cations,and nutrients. As new data are collected, candi-date metrics will be tested and refined, and regionaldifferences and ecological linkages among wetlandsand other waters will be examined. DEP is alsodeveloping impairment thresholds for wetlands.This is a necessary first step to enable the State touse biological monitoring data in wetland manage-ment decisions and development of wetland-specific water quality standards (designated usesand biological criteria), and to assess wetland con-dition and attainment status.


Quantitative and qualitative algae samples weregathered from the same wetland sites as used formacroinvertebrate sampling. Four algae sampletypes were tested to determine which produced thebest indicators. Samples were collected from thewater column, plant stems, and sediments. Samplesfrom multiple sites within each wetland werecomposited into one sample from each habitat. Inaddition, a multihabitat sample was collected fromeach site. Samples were examined microscopicallyto determine species numbers and relative abun-dances of different species in samples. Chlorophylla was quantified from a separate water columnsample as an indicator of algal biomass.

For sampling, garden shears were used to clipplant stems below the water line. A turkey basterwas used to collect qualitative sediment samples;however, this method was revised in 2002 to ob-tain a more quantitative sample. Sediment is cur-rently collected over a known surface area using apetri dish pressed into the substrate and retrievedwith a spatula. Three replicates are collected andcomposited into a single sediment sample. A longhandled dipper is used to collect water samples.For the multihabitat sample, a dose from each single-habitat sample was combined into one container.

Plate 3: Maine’s stovepipesampler and sieve bucket.

Photo: Maine DEP

Plate 4: Pickingmacroinvertebrates from

multihabitat sample.Photo: Maine DEP

17

Analytical Methods: Algae

Dr. Jan Stevenson of Michigan State University isusing three disturbance indicators: the land use in-dicator developed by Maine DEP, trophic statusindicators (total nitrogen, total phosphorus, andchlorophyll a), and hydrologic and sewage chemi-cals (Ca, Na, Cl). Dr. Stevenson is comparing asuite of algal indicators to determine which types ofindicators respond to the three disturbance indica-tors. The algae indicators include biological integ-rity measures such as genus-species richness,Shannon diversity, and number of taxa in genera.Dr. Stevenson is also using European algalautoecology information to determine environmen-tal characteristics for the taxa. This information willgive an autoecological index that shows a relation-ship to variables such as moisture, organic N, lowoxygen, pH, salt, and nutrients. Although algae wascollected from all sites sampled between 1998 and2002, only samples from 1998 and 1999 have beenprocessed to date because of funding limitations.

Lessons Learned

Incorporating a wetland monitoring componentinto Maine’s existing biomonitoring program hasbeen an efficient way to pool limited resourcesand build on Maine’s successful river and streambiomonitoring experience.

To implement a comprehensive biomonitoringprogram for wetlands, Maine DEP needs tobuild the capacity to assess multiple biologicalassemblages, including algae and vascularplants. This capability will improve the State’sability to evaluate wetland impacts from stres-sors such as nutrient enrichment and hydrologicchanges, and will allow for the assessment ofless frequently inundated wetland types whereaquatic invertebrates are not naturally abundant.Current funding and staff levels prohibit expan-sion of the program at this time.

18

Maryland: Developing an IBI Assessment forRestored Wetlands in the Mid-Atlantic States

Contact Peter Lowe

Organization USGS BRDPatuxent Wildlife Research Center11510 American Holly DriveLaurel, MD 20708-4017

Phone (301) 497-5705


Website http://www.pwrc.usgs.gov/wli/

Purpose

Develop sampling methods for different assem-blages

Develop a yardstick of biological metrics toassess the progress and condition of recon-structed wetlands in Maryland, Delaware, andVirginia

Compare the suitability of different assemblages(plants, macroinvertebrates, amphibians) forassessing wetland condition

Evaluate the sources and magnitude of variancein data collected for biological metrics

Evaluate seasonal and annual biological fluc-tuations for the wetland sites

Wetland Type

Delmarva Bays (depressional wetlands withemergent, scrub/shrub, or forested vegetation)

Assemblages

Amphibians

Macroinvertebrates

Vascular plants

Project History

The project is a joint effort among the USGS,Patuxent Wildlife Research Center, the USDA Natu-ral Resources Conservation Service Wetland Sci-ence Institute, and EPA. Project developmentstarted in 1995 in response to a lack of informationabout the success of wetland mitigation projects,especially those associated with wetland restora-tion on farmlands under set-aside programs. From1996 to 1998, fieldwork focused on a set of re-stored and existing wetlands. In 1999, a secondset of wetlands was studied to evaluate the robust-ness of the metrics developed from the first set. Allfieldwork has now been completed and metricdevelopment is under way.

19

Study Design

Because one of the objectives was to assess thesources and magnitude of variance in the differentmeasurements, we intentionally oversampled oursites compared with what would be done under typi-cal regional or statewide assessments. For 3 years,work was done on a single set of 30 wetlands, in-cluding 22 restored and 8 natural wetlands orDelmarva bays. All wetlands in the database weredepressional, semipermanent, or seasonal.Macroinvertebrates were sampled one to four timeseach year, depending on hydrology; macrophytesduring spring and late summer; and amphibians al-most continuously during the breeding season.

Sampling Methods and Analysis:Macrophytes

Data on plant species composition, abundance,and dominance were collected through the use ofline transects. Pairs of 50 m transects were placedon opposite sides of the wetland at four differenthydrological levels as determined at “full pool” lev-els. The hydrological levels were buffer, less than25 cm, 25 to 45 cm, and greater than 45 cm. Inwetlands larger than 5 acres, the number of transectpairs was doubled. Each transect was divided intofive runs of 5 m each, alternating with 5 m ofunsampled transect. Along each run, sampling oc-curred at specific points (as in point intercept meth-ods) at 1-m intervals. Thus, each run of 5 m con-tained five sample points. Each transect consistedof five runs of five sampling points each. Each hy-drological depth was sampled by 2 transects, bring-ing the total number of sample points per hydro-logical zone to 50. There were:

(4 hydrological zones) H (2 transects each zone)H (5 runs per transect) H (5 points per run)= (200 points per wetland)

This was done in spring and fall for each wetland.At each point, the species name and number of in-dividuals were recorded.

In addition to point data, incidental species alongthe line were recorded to include the species notintercepted at any points along the line. These spe-cies were included in species richness and attributesderived from presence or absence data, but not indominance calculations.

Sampling Methods and Analysis:Macroinvertebrates

Aquatic invertebrates were sampled at approxi-mately 6-week intervals beginning in late May andcontinuing until October and were conducted inassociation with the sampling of water quality,aquatic plants, and hydrological and wetland dimen-sions. Invertebrate samples were collected alongtransects following compass coordinates originat-ing from markers placed in the deepest part of ev-ery wetland before the sampling seasons began.Transect coordinates were randomly selected foreach wetland and sampling time. The method ofcompass points was adapted to each wetland’smorphology.

Samples were collected from three depth rangesalong the transects to determine invertebrate rela-tive abundance, diversity, and relative biomass ineach wetland. As long as water depths wereadequate, samples were collected from along thetransects at the following water depths: less than15.0 cm, 15.1 to 45.0 cm, and greater than45.1 cm.

Samples were collected using a modified Gerkingbox sampler, which is a sheet aluminum box with asliding screen door (1-mm mesh) at the bottom.The sampler has the advantage of allowing simulta-neous collection of benthic, pelagic, neutonic, andplant-associated invertebrates. The sampler waslowered to the floor of the sampling area with thescreen door open. The vegetation in the samplerwas then cut at the mud-water interface and putinto prelabeled plastic bags. Then the screen doorwas slowly closed as the sediments just in front of

20

the advancing screen were stirred into the watercolumn. After the screen was closed, soil materialswere sieved through the screen by shaking the box.All invertebrates caught on the screen were thenplaced in prelabeled plastic bags. The bags of veg-etation and invertebrates were placed on wet ice assoon as possible and stored in a refrigerator untilthe samples could be picked. Each individual waskeyed to the lowest taxonomic level feasible, typi-cally species or genus. Data on species composi-tion and abundance were used to generate metricsin a manner similar to that for plants. An approvedquality assurance/quality control process was fol-lowed, which included independent validation of20% of our samples.

Sampling Methods and Analysis:Amphibians

Each site was sampled for amphibian larvae onceevery 4 weeks. Sampled areas consisted of theperimeter of the open-water portion of the wetlandand lightly vegetated areas that allowed a seine topass. We used a 6H8-m nylon mesh (1/16") seineto sample each wetland by wading out 3 to 5 maway from the shoreline and then moving in towardsthe shoreline in one continuous sweep. Samplingwas time-constrained to 2 hours. If new specieswere caught during the last two sweeps, the sam-pling period was extended until no additional newspecies were found. Amphibian larvae were iden-tified using published keys and by temporarily hous-ing tadpoles until they metamorphosed and couldbe identified.

Drift fences were also used to supplement seiningdata and obtain information on adults andmetamorphs. We initially considered surroundingeach wetland with a drift fence but realized the im-practicality of that idea. Therefore, each wetlandwas provided with 50-cm-tall and 15-m-long driftfences. If possible, the fences were placed alongdrains, travel corridors, and other likely points ofamphibian use in order to maximize capture of indi-

viduals entering and exiting the wetlands. Five-gal-lon plastic buckets were buried in pairs at the endsand midpoint on the inside and outside of the fenceto capture adults entering the ponds and juvenilesleaving the ponds. Wet sponges, rocks, and veg-etation were placed in the buckets to prevent des-iccation and provide some cover and refugia to cap-tured individuals. All amphibians were identified,sexed, and returned to the inside of the fence at thewetland from which they were captured. Juvenilesand metamorphs departing the wetlands were iden-tified, counted, examined for malformations, andreleased on the outside of the drift fence.

General Analytical Methods

A fundamental component of this study is to de-vise a gradient of physical factors (e.g., land use indrainage area, management techniques, landscapefeatures, method of restoration) that affect wetlandhealth. From there, data on the frequency of occur-rence and relative abundance of species, guilds, ortrophic classes will be used to develop attributesfor an index of biological integrity (IBI) for each ofthe assemblages. An attribute will be considered avalid metric if it relates either positively or nega-tively to the physical gradient. We will then com-pare the IBIs developed and determine if they areconsistent in ranking wetlands. Once acceptableIBIs have been developed on the initial wetlandbase, we will apply them to the second set of wet-lands to validate the model. In addition, the vari-ance within our sampling methodology will be as-sessed.

Lessons Learned

OverallMany of the wetlands in our bases were only afew years old when we started and may nothave had sufficient time for ecological and an-thropogenic factors to separate them along aphysical gradient.

21

As a result, development of a reliable and eco-logically meaningful gradient has been one ofthe most difficult parts of this project.

It would be very instructive to revisit these wet-lands after 10 or 15 years and see how theyhave changed.

MacrophytesThere can be considerable differences betweenmid- and late summer in the ability to easilyrecord and identify plants. This is especially truefor graminoids, which are primarily identifiedby fruiting body characteristics. In addition,many legumes, composites, and warm-seasongrasses are present in late summer but not ap-parent in spring.

The inclusion of incidental species added ap-preciably to the number of species identified ina particular wetland. We are evaluating whetherthis inclusion has an effect on the resultingmetrics.

Deep-water areas (greater than 45 cm) havemuch lower species richness than shallowerzones and do not need to be sampled at thesame level of intensity at the same site.

Permanent transects are preferred if data col-lection can continue over several years. Thiswill allow for the annual and seasonal changesthat occur over time through shifts in hydrol-ogy.

MacroinvertebratesPicking, sorting, and identifying aquatic insectswas one of the most laborious aspects of thestudy. Investigators wishing to use invertebratesin their bioassessments should allocate sufficientresources to accomplish the task.

Invertebrate species presence, and especiallyabundance, are seasonally quite variable. Juneto early July before the drying of mid- to latesummer begins seems to be the best months forfinding the greatest diversity and abundance ofmacroinvertebrates.

Amphibians

The reduced species richness of amphibians,compared to macrophytes and macro-invertebrates, may limit the number and typesof metrics that can be developed from thisassemblage.

Adequate sampling for amphibians requiresmore trips and techniques than other assem-blages. This is due to their mobility, multiple lifehistory strategies, and variable breeding peri-ods among species. Sampling for one life stageonly is probably not as effective as sampling foradults and tadpoles in determining amphibianusage of a wetland.

22

Massachusetts: Use of Multimetric Indices toExamine Ecological Integrity of Salt Marsh

Wetlands in Cape Cod

Contact Bruce K. Carlisle

Organization Massachusetts Coastal Zone Management251 Causeway StreetBoston, MA 02114

Phone (617) 626-1200


Website http://www.state.ma.us/czm/wastart.htm

Purpose

T he primary focus of the Cape Cod SaltMarsh Assessment Project is to advance and

improve the salt marsh assessment approach de-veloped by the Massachusetts Coastal Zone Man-agement (MA CZM) team through the applicationand review of the existing protocol in two differentassessment investigations. The first investigation,conducted in the 1999 field season (May to Octo-ber index period), examined salt marsh indicatorsfrom six sites on the Cape Cod Bay coast; thesesites had varying types and intensities of human landuse or disturbance. The second investigation is along-term comparison of indicators from three pairsof salt marshes, each pair having a marsh area withrestricted tidal hydrology and a corresponding areawith normal tidal hydrology. The intent of this workis to document differences in indicators before andafter tidal restoration actions. Through the imple-mentation of these two investigations additionalobjectives will be realized. The collection and com-pilation of data on the condition of relatively undis-turbed sites is of critical importance to the evalua-tion and determination of impaired sites. Thisproject will serve to expand the salt marsh refer-ence site database. Another important aspect ofthis project will be to further examine the suite of

existing metrics and attributes used for biologicalcomparison and to explore new metrics, based onthe project data and literature/information base. Thelong-term tide restriction study will provide insightas to the utility of this assessment approach as ameasure of determining salt marsh restorationprogress and trajectory.

Wetland Type

Salt marshes

Assemblages

Birds

Fish

Macroinvertebrates

Vascular Plants

Status

The Cape Cod Salt Marsh Assessment Project isapproximately two-thirds complete, with the land-use investigation finished and the longer-term tiderestriction study entering its third year (or index sea-

23

son). The report for the 1999 land-use investiga-tion is being finalized and should be available in earlyfall 2002 from this site or MA CZM’s site (http://www.state.ma.us/czm/wastart.htm). EPA grantfunds will support two additional years of field datacollection, with project completion anticipated inMarch 2004. Field investigations from 2000 through2004 are focused on the assessment of the effectsof tidal hydrology alterations and the restoration ofnormal hydrology on salt marsh ecological integrity.The study involved three salt marsh sites with tidalhydrological restrictions. Of the project sites, one(EOBP) received restoration action in spring 2002,one (EMC) is predicted for fall 2002, and one(MSLP) is still uncertain for restoration action, withunforeseen complexities. In addition, the projectteam expects to sample one additional referencesite each year to expand the regional referencedataset.

Project Description

Massachusetts Coastal Zone Managementlaunched its Wetland Assessment Program with itsfirst effort in wetland bioassessment in 1996-1998at the Waquoit Bay National Estuarine ResearchReserve. A subsequent project was implementedon Massachusetts’ North Shore in 1997-1999.

The EPA Region I sponsored pilot, Cape CodSalt Marsh Assessment Project, began in 1999 andwill continue through 2004. The investigative teamfor this project includes Bruce Carlisle, Tony Wilbur,Jan Peter Smith, and Jay Baker from MA CoastalZone Management, and Anna Hicks, an indepen-dent consultant specializing in aquaticmacroinvertebrates. Partners for this project in-clude the MassBays Program, UMass Extension,the MA Department of Environmental Management(DEM) and its Area of Critical Environmental Con-cern Program, the Waquoit Bay National EstuarineResearch Reserve, the Cape Cod Commission, andthe towns of Barnstable, Eastham, Mashpee, Or-leans, and Sandwich.

The Massachusetts CZM project team will applystandardized sampling and surveying protocol to saltmarsh study sites to gather biological, chemical, andphysical data for the two investigations referred toearlier.

The project team will analyze and express the bio-logical data through a series of existing metrics orattributes, or develop new metrics or attributes asnecessary, based on the project data and literaturebase. Chemical and physical data will be utilized assupporting information sources.

The team will make recommendations for revi-sions, additions, or deletions to its current wetlandsassessment approach.

Study Design

In the 1999 field season, six salt marsh sites withvarying types and degrees of intensity of surround-ing land use were selected. Two sites with minimalhuman land use (conservation land and no tidal hy-drological alteration) were chosen as referencesites, representing the best attainable conditions inthe immediate region.

The four other salt marsh study sites have variedland uses including residential, commercial, andtransportation. The impacts associated with theseland uses include direct stormwater outfalls, largeimpervious areas, septic systems, lawn fertilizer/chemicals, pet waste, automobile emissions/byproducts, and direct habitat alterations.

For the 2000 field season, three salt marshes withtidal hydrological restrictions will be studied. Mea-surements will be made both at the salt marsh af-fected by the tidal restriction (the restricted studysite) and at the salt marsh below the restrictive fea-ture (the reference site). The reference sites ormarshes below the tidal restriction receive normaltidal influence and inundation. In addition, the 2000field season will also include two additional long-

24

term salt marsh reference sites. The continued de-velopment of a robust reference site database iscritical to the continuing evolution and applicationof wetland bioassessment.


At each site, a habitat characterization form willbe completed that summarizes the ambient saltmarsh habitat conditions at the study site. The in-formation collected includes water quality param-eters (temperature, pH, dissolved oxygen, salinity,and color) and habitat descriptors on hydrology,vegetation, substrate, available food sources for in-vertebrates, and degree of human impact.

The sampling protocol will utilize tree devices forapplication in different habitats:

Intertidal zonePlot sampling using wooden frame (18"H18")

Subtidal (permanently flooded) zoneD-net, plot sampling, and auger At each sampling site, three sampling stations will

be selected over a defined linear distance of thecreek channel, stations will be located at 1/3 inter-vals. At each of the three stations a representativecomposite sample of macroinvertebrates will be col-lected as follows:

Intertidal bank zone at low tide: 1 plot sample

Subtidal estuarine zone at low tide: 1 plotsample, 1 D-net sample, and 1 auger sample

Each sample will be placed in a zip lock bag la-beled with site number, site name, date of sampling,sample number, sampling method, and name of sam-pler. The site field sheets also record the relevantsample numbers. Samples will be preserved in 70%isopropenol alcohol and placed in a cooler readyto be transported to the laboratory for sorting andidentification.

Invertebrate samples are taken once in May andagain in August.

Sampling Methods: Vegetation

At each site, the salt marsh wetland vegetationwill be surveyed according to this protocol. Sixtransects will be established based on a stratifiedrandom sampling approach. A defined linear dis-tance of the salt marsh creek channel is established.

The evaluation area will be segmented into threesubunits along equal sections of the creek channel.The first third of this length is subunit #1, the sec-ond third is subunit #2, and the final third is subunit#3. In each of the subunits, two randomly selectedtransects will be laid. The transect locations will bedetermined by a computer random numbersalgorithm.

The transects will run roughly perpendicular fromthe channel to the upland edge, and each transectwill be laid according to a consistent compass bear-ing. Along each transect, 1-m2 quadrats will belocated every 60 feet, starting at the creek edgeand progressing along the entire length of the transectuntil the upland edge. The last quadrat will be lo-cated in the salt marsh fringe community, well withinthe wetland and not on the upland.

Using a standard data sheet, in each quadrat alongeach transect, every plant occurring within that quad-rat will be identified by genus and species. Foreach unique species within the quadrat, the abun-dance of that species will be determined using stan-dardized coverage charts. Investigators will alsodefine the community type that the quadrat falls in:low marsh, high marsh, or fringe. To be as accu-rate as possible, coverage estimates include duff,leaves, bare ground, and open water, collectivelydesignated as “other.” Coverage estimates will beadjusted during the analysis stage to account forthe coverage of this “other” category.

25

Vegetation surveys will be conducted once at eachsite during the peak growing period from mid-Julyto mid-September.

Sampling Methods: Birds

Point counts will be utilized as the primary sam-pling method, using visual and auditory cues. Atleast two expert observers, including the principalinvestigator, will sit quietly from a vantage pointwhere all of the evaluation area can be viewed.Using a standard data sheet, all species and indi-viduals will be counted and recorded by the ob-servers, as they are heard or seen demonstratingany activity in the evaluation area or in a 100-footbuffer area surrounding the evaluation area. Countswill be conducted for a period of 20 minutes, sepa-rated into four 5-minute sample intervals. All indi-viduals will be counted, with a concerted effort notto duplicate individuals. An additional 10 minuteswill be allotted to allow observers to walk slowlyalong the perimeter of the wetland in order to de-tect any species not tallied in the 20-minute count.Several sites may be visited on the same day, withcensus beginning at approximately 6 am and ceas-ing at approximately 8:30 am in order to capturepeak activity. Sites will be sampled in late Augustto capture migrating shorebird usage, as saltmarshhabitats are known to have comparatively fewerbreeding species.

Sampling Methods: Fish

The sampling strategy will be to capture the chan-nel habitat (sub- and intertidal). As detailed above,the evaluation area will be divided into three sub-units along a defined length of creek channel. Sta-tions will be established for each habitat as follows.In the creek channel, fixed stations will be deter-mined by a computer random numbers algorithmproducing a random integer between 0 and 100.The random integer will be the distance of the start-ing point for the seine haul in feet from the startpoint (0’) of each subunit.

Seines will be utilized to sample the creek chan-nel. At three stations, as defined above, the seinewill be dragged through the water column along thecreek bank and substrate for a length of 5 meters.Seines will be carefully withdrawn from the creekand the collected fish will be carefully extracted fromthe seine into a processing bucket. The fish willthen be removed by dipnets and individuals will beidentified and measured. Abundance and total bio-mass by species will be enumerated. All data arerecorded on a standardized field sheet.

One sample run will be conducted each monthfrom April to October.

Analytical Methods

Biological data collected at wetland study sitesare compared to data collected at the wetland ref-erence sites. Multimetric data analysis techniquesare employed to examine attributes and variablesof biological data, and these metrics are combinedinto a quantitative final index. A metric is a param-eter or variable that represents some feature, sta-tus, or attribute of biological assemblage, chemicalstate, or physical condition. In a multimetric ap-proach, several different metrics are chosen in or-der to effectively capture and integrate informationfrom individual, population, guild, community, andecosystem levels and processes. Metrics are se-lected on the basis of literature reviews, historicaldata, and professional knowledge. The quantita-tive output from each metric is then combined toproduce an index. An index is the aggregate ofweighted metric scores that serves to summarizethe biological condition.

Lessons Learned

Through the three pilot projects, the MA CZMproject team has been able to learn from each ap-plication and as a result has made several small,incremental revisions to many components of theprotocols, including adjustments to sampling meth-

26

ods and shifts in the attributes and metrics exam-ined. Each application has also generated resultsthat indicate decreasing biological integrity with in-creasing land-use stressors.

Results from 6 years of wetland assessmentexperience have led the MA CZM team to con-clude that this assessment approach has significantpotential for a number of management applications,including:

Inventory efforts such as identifying and pro-tecting unique and important habitat

Identification of degraded salt marsh sites andpresence of invasive (nonnative) species in or-der to evaluate restoration potential and/or re-port on wetland status (i.e., 305b reports)

Monitoring and evaluating the effectiveness ofrestoration actions

Long-term tracking of salt marsh condition anddocumenting the effects of disturbances (i.e.,stormwater pollution, eutrophication,hydromodification)

Many of these applications have appeal to water-shed-based organizations and agencies as well asvolunteer groups. The techniques of the MA CZMsalt marsh assessment approach have been success-fully taught through a comprehensive volunteer train-ing program on Massachusetts’ North Shore. Atthe time of writing, the Wetland Health AssessmentToolbox program is entering its fourth year, guidingand supporting volunteer groups as they monitorsalt marsh restoration sites in their region (http://www.salemsound.org/wetlands.htm).

Results from this Cape Cod Salt Marsh Assess-ment Project are still being analyzed and writtenup, but some initial details are available: The 1999land-use investigation has confirmed the trend seenin the previous two studies where increasing levelsof land use around a given salt marsh site (wetland)correspond with decreased ecological condition, inthis case indicated by the average of the Plant Com-munity Index and the Invertebrate CommunityIndex.

27

Figure 7: Relationships between wetland ecological condition andland use. The Land Use Index (LUI) score is a measure of human

disturbance. With increasing land use types and intensities, the LUIscore declines. Similarly, the Wetland Ecological Condition

(WEC) is a measure of biological integrity. A lower WEC scoreindicates increasing biological impairment.

28

Figure 8: Two sets of multimetric goals for evaluating restoration.The restored site is on the bottom, its paired reference (directly

below/seaward tide restriction feature) is the middle graph, and aregional reference marsh is displayed at top.

29

Plate 5: Cape Cod, MA, salt marsh plant survey:1 m2 along transect. Photo: B.K. Carlisle

Plate 6: Cape Cod, MA, salt marsh nekton: Watch your fingers!(blue crab, Callinectes sapidus). Photo: B.K. Carlisle

30

Massachusetts: Involving Volunteers inExamining the Ecological Integrity of Coastal

Wetlands in Cape Cod, Massachusetts

Contact Britta Magnuson

Organization Salem Sound 2000201-203 Washington St., Suite 9Salem, MA 01970

Phone (508) 741-7900


Purpose

Training volunteers to do biological assessments

Comparing results of volunteers to trained sci-entists

Wetland Type

Salt marshes

Assemblages

Birds

Fish

Macroinvertebrates

Vascular plants

Status

Completed 3 years of sampling. Analyzing data.

Project Description

“Salem Sound 2000” and “8 Towns and the Bay,”two regional subcommittees of the MassachusettsBays National Estuary Program, participated in asalt marsh monitoring pilot project involving citizenvolunteers in conjunction with UMass CooperativeExtension Service and Massachusetts Coastal ZoneManagement (MCZM).

During the summers of 1999 and 2000, more than40 volunteers participated in training workshops andfield data collection for a variety of parameters:water chemistry, land use index (a habitat assess-ment), aquatic macroinvertebrates, birds, tidal in-fluence, and vegetation. During 1999, professionalscientists did independent assessments, and thevolunteers conducted assessments (using the samesampling protocols as the professionals) with theguidance of trained staff members. Data compari-sons, as well as feedback from volunteer partici-pants, were used to modify and improve trainingprotocols for the 2000 field season. Field data werecollected in 2000 at the same sites, which included

Contact Jan Smith

Organization Mass Bays National EstuaryProgram

251 Causeway Street, Suite 900Boston, MA 02114

Phone (617) 626-1231


Website http://www.state.ma.us/czm/wastart.htm

31

salt marsh sites that were affected by tidal restric-tions and sites impacted by stormwater discharges.The project is expected to be a model in other ar-eas of New England. A volunteer training manual iscurrently being printed and should be available atthe time of this printing. Funding from a privatefoundation has enabled the program to continue fortwo additional years, 2001 and 2002, and it isplanned to add fish as an additional parameter forthe volunteers to measure. The sampling proto-cols are summarized in the MCZM project sum-mary.

Lessons Learned

Volunteers may need ongoing field instruction inorder to ensure quality data collection. The initialcomparison between data collected by volunteersand data collected by professionals indicated gaps,which improved field assistance was able to reduce.Also, asking the volunteers to review and commenton training workshops was helpful in improving pro-gram design. Although there is high turnover in vol-unteers from year to year, we are finding, after 4years into the program, that some volunteers docome back after a year’s break, and that the skillsof the returnees definitely improve after the first year.Offering teacher training credits improves the inter-est of educators in participating in the program, andmany get interesting ideas for future use in their class-rooms.

32

Purpose

Subproject 1: Coastal wetlandsCollect baseline data on water quality and ad-jacent land use, as well as plant, invertebrate,and vertebrate communities from Great Lakescoastal wetland sites experiencing a continuumof disturbance

Continue development of invertebrate and fish-based indices of biological integrity (IBI’s) byplant zone, for Great Lakes coastal wetlands

Continue testing and validation of our IBIs

Subproject 2: Inland, forested depressionalwetlands

Collect baseline data on plant, invertebrate, andvertebrate communities with accompanyingchemical/physical parameters from referenceand impacted forested, depressional wetlandsof southern Michigan

Develop an IBI for forested depressional wet-lands of southern Michigan based on inverte-brates, plants, fish (if present), and birds

Wetland Type

Great Lakes coastal

Inland, forested depressional

Assemblages

Birds

Fish

Macroinvertebrates

Vascular plants

Status

Completed 3 years of sampling. Analyzing data.

Project Description

Subproject 1: Coastal wetlandsWe developed a preliminary IBI for Lake Huron

based on invertebrate data collected from coastalwetlands with funds provided by the Michigan De-

Contact Thomas M. Burton

Organization Michigan State UniversityDepartment of Zoology203 Natural ScienceLast Lansing, MI 48824-1115

Phone (517) 353-4475


Michigan: Bioassessment Procedures andBaseline Reference Data for Great Lakes

Coastal Marshes and Inland ForestedWetlands in Michigan

Contact Donald G. Uzarski

Organization Grand Valley State UniversityAnnis Water Resources InstituteLake Michigan Center740 West Shoreline Dr.Muskegon, MI 49441

Phone (616) 895-3989


33

partment of Environmental Quality, the Nature Con-servancy, U.S. EPA, and the U.S. Geological Sur-vey and published it in the journal Wetlands (Bur-ton et al. 1999). Since publication, we have con-tinued testing the IBI with data from additionalLakes Huron and Michigan fringing wetland sites.We also have continued monitoring a set of initialsites used in IBI development as lake levels de-clined in order to determine whether and how wellthe IBI works as lake levels decline substantiallybelow levels that occurred during IBI development.We have collected extensive data on invertebrates,plant, fish, and bird communities on Lake Huronwetlands since the early 1990’s. Much of our workhas been in collaboration with the Great Lakes Sci-ence Laboratory (BRD-USGS), Michigan NaturalFeatures Inventory (MNFI), and the Ohio Biologi-cal Survey (OBS).

Our preliminary IBI for Lake Huron fringing wet-lands was developed using macroinvertebrate datacollected in 1997 from six wetlands. It was testedusing data collected from 11 Lake Huron wetlands(6 original and 5 additional) in 1998 at lake levelssubstantially lower than in 1997. We continued totest the IBI using data collected from 12 sites (7original and 5 additional) at even lower lake levelsin 1999 and from 5 additional sites at extremelylow water levels in 2000. Even though some plantcommunity zones used in IBI development werenot flooded and could not be sampled in 1999 and2000, the IBI functioned extremely well. In 2001,we sampled similar fringing wetlands of northernLake Michigan along with northern Lake Huron sites.The IBI appears to work for fringing wetlands ofnorthern Lake Michigan as well.

While testing and validating the Lake Huron IBIin 1999, 2000, and 2001, we developed modifica-tions to simplify and improve it. A detailed expla-nation for the modifications was presented at “Wet-lands 2000” in Quebec City in August 2000, andadditional modifications to remove the “preliminary”status were presented in Lake Placid, New York,

in June 2002. The modifications are as follows:

1. The Typha zone should be removed from theIBI.

2. The four diversity and richness metrics shouldbe calculated by plant zone instead of combin-ing data from all plant zones before calculationsare made.

3. Two new metrics should be added to the InnerScirpus Zone:

– Relative abundance Isopoda (%)—Decreases with disturbance

– Relative abundance Amphipoda (%)—Increases with disturbance

4. Use ½ person-hour count to determine num-ber of individuals counted per replicate. Counteither 50, 100, or 150.

A manuscript noting all of the above improvementsis in preparation and will be submitted to the jour-nal Aquatic Ecosystem Health and Managementby September 2002 as part of a special issue oncoastal wetlands.

We are optimistic that our Lake Huron IBI willwork for northern Lake Michigan fringing wetlandsites. These sites include most wetlands along thesouthern shore of the Upper Peninsula of Michiganfrom St. Ignace to the Wisconsin border. Most ofthese wetlands appear to be comparable in plantcommunity composition and structure to the LakeHuron wetlands. Preliminary testing took place in2001 and we expect to be able to recommend theIBI’s use for these types of wetlands.

In 2000, we began development of new fish- andmacroinvertebrate-based IBIs for the drowned rivermouth wetlands of Lake Michigan. We collecteddata from eight sites representing a gradient of an-thropogenic disturbance. Ten sites were sampledin 2001 and several new sites will be added in 2002.We expanded our work on drowned river mouthwetlands in 2001 to Lake Superior wetlands.

34


We started our forested depressional wetlandproject in 1999. Invertebrates were collected threetimes per year from eight forested depressionalwetlands during 1999 and 2000 using dipnets, ac-tivity traps, and black lights. Additional sampleswere collected from a subset of these and someadditional wetlands in 2001. Up to 40 additionalsites will be added in 2002, with emphasis on ob-taining a greater array of impacted sites and ex-tending the sites to isolated depressional wetlandsin the coastal zone of the Great Lakes. Inverte-brates are identified to the lowest taxonomic unitpossible, and preliminary IBI development has be-gun. Accompanying chemical/physical samplestaken from surface water and mini-piezometers havealso been recorded and analyzed for potentialmetrics. Comparisons of the plant and invertebratecommunities have been made using a subset of thesites.

A list of potential metrics, a summary of our verypreliminary analyses, and conclusions that we drewfrom the 2000 data set are as follows:

Significantly higher at reference sites (Mann-Whitney tests):

Isopoda p < 0.001

Amphipoda p = 0.004

Diptera

Culicidae p = 0.004

Ephemeroptera p = 0.015

Trichoptera p = 0.078

Significantly higher at impacted sites (Mann-Whitney tests):

Gastropoda

Lymnaeidae p = 0.004

Planorbidae p < 0.001

Diptera

Chironomidae p = 0.074

Chaoboridae Collected from only onesite

Odonata

Libellulidae p = 0.039

Hemiptera

Lleididae p < 0.001

Coleoptera

Halipidae p < 0.001

Annelida

Hirudinea Collected from only onesite

Summary

Of the measured chemical/physical variables, onlydepth separated impacted and reference sites.Canopy cover was not measured in 1999 and 2000but may have also separated impacted and refer-ence sites. Estimates of canopy cover are currentlybeing obtained for all sites. Eleven (+2 ?) taxashowed potential as metrics, but not during earlyinundation (April). During full inundation (May),the correspondence analysis grouped the wetlandsusing invertebrate community composition into dis-tinct categories.

Surface runoff-influenced impacted sites

Groundwater-influenced reference sites

Precipitation-influenced reference sites

Conclusions

Increased runoff may increase water depth andsubsequently kill trees, opening the canopy.

– Some taxa such as Libellulidae may beresponding to these changes in ambientconditions.

35

Most traditional chemical/physical variables willnot be useful as indicators of anthropogenic dis-turbance in these wetlands. These variables in-clude turbidity, DO, pH, SRP, NH4, conduc-tivity, alkalinity, temperature.

An invertebrate-based IBI for (hardwood) for-ested wetlands of Southern Michigan appearsto be feasible.

Study Design

Subprojects 1 and 2Sites that experience a wide range of anthropo-

genic disturbance, or stressors, are chosen from eachhydrogeomorphic class or subclass of wetland. Theextent of disturbance is determined using surround-ing land use/land cover and limnological data. Ini-tially, correspondence analyses of invertebrate com-munity composition were used to determine if ref-erence sites separated from impacted sites. Whenthey did, individual taxa containing the most inertiaresponsible for the separation were deemed po-tential metrics. Mann-Whitney U tests were thenused to determine if densities of these taxa at refer-ence sites were significantly different from densitiesat impacted sites.

We use medians in place of means in the IBI be-cause medians are more resistant to the overwhelm-ing effects of outliers. Our goal is to typify the wet-land. If an area is sampled that is depleted or con-centrated in the constituents of a metric, the areamay be isolated from anthropogenic disturbance,receiving a dose of disturbance not typical of theentire wetland or vegetation zone, or may containsome “natural” chemical/physical component thatis unique. Regardless of the cause, the area is notrepresentative of the entire wetland. The influenceof these outliers can be dampened by using themedian in place of mean as a measure of centraltendency.

After potential metrics were developed, principalcomponents analysis (PCA) was used to establish

principal components (PCs) based on chemical/physical parameters as well as surrounding (1-kmbuffer) land use/cover data. Pearson correlationswere done between individual metrics and PCs toestablish stressor-ecological response relationships.PCs were then decomposed to explore relativecontributions of individual stressors.

Sampling Methods

Subproject 1: Coastal WetlandsMacroinvertebrate samples. Macroinvertebrate

samples were collected with standard D-frame,0.5-mm mesh dipnets. All major plant communityzones were sampled at each site, including a deepemergent and a shallow, wet meadow zone. If cer-tain depths contained more than one dominant plantcommunity along the shoreline, each plant commu-nity type was sampled.

Dipnet sampling entailed sweeps through the sur-face and middle of the water column and above thesediment surface to ensure that an array of micro-habitats were included in each sample. Dipnets wereemptied into white enamel pans, and 150 inverte-brates were collected by removing all specimensfrom small areas of the pan. Special considerationwas made to ensure that smaller organisms werenot missed, as there is a bias toward larger, moremobile individuals with this technique. Plant detri-tus was left in the pan and sorted through for a fewadditional minutes to ensure that sessile species wereincluded in the sample. If 150 individuals were notobtained after ½ person-hour of field picking, wecollected to the next multiple of 50. The timed countwas used to semiquantify sampling effort so that itcould be used as a metric. Three replicate sampleswere collected in each plant zone to obtain a mea-sure of variance.

Dipnet samples were collected from late Julythrough August. Samples taken from ice-out throughmid-July generally contained less diversity and agreater proportion of early instars of aquatic insects,

36

making identification more difficult. The July–August time period also corresponded to the timewhen plant communities, characteristic of thesewetland systems, achieved maximum annualbiomass.

Fish sampling. Six fyke nets (0.5-in mesh) wereset in each wetland for 1 net-night. Nets were setadjacent to specific plant zones with the leads bi-secting the sampling area. Minnow traps were usedas a secondary method.


Invertebrate sampling. Macroinvertebratesamples were collected with standard D-framedipnets containing a 0.5-mm mesh. Three replicatesamples were collected from three locations in eachwetland: (1) the deepest portion (usually the cen-ter), (2) near the upland, and (3) between thesetwo areas. We attempted to incorporate habitatheterogeneity by sampling as many plant zones aspossible at each location.

Dipnet sampling entailed sweeps through the wa-ter column at the surface, middle of the water col-umn, and above the sediment surface. Dipnets wereemptied into white enamel pans and 150 inverte-brates were collected by focusing on small areas ofthe pan and removing all specimens. If 150 indi-viduals were not obtained after ½ person-hour offield picking, we collected to the next multiple of50. The timed count was used to semiquantify sam-pling effort so that it could potentially be used as ametric. Invertebrate sampling was conducted dur-ing early (April), full (May), and late inundation(June) at each site.

Fish sampling. The temporary pools that domi-nate most depressional forested wetlands are un-likely to contain fish. Thus, fish sampling was only

conducted for permanent pools in depressionalwetlands when they occurred. Small fish traps wereplaced in each of the permanent pools.

Birds surveys. Bird communities were surveyedusing 10-minute, 50-m radius point counts (Ralphet al. 1995) by dual observers to sample the birdcommunities at each count location. The 10-minutecounting period began when the observers reachedthe perimeter of the 50-m radius so that any birdsflushed or silenced by the observer’s approachwere detected (Riffell et al. 1996). The bird com-munities of 30 forested wetlands were sampled in2000 using a total of 6 count visits. A variety ofhabitat measures from each site were obtained asdetailed under plant sampling below, and a manu-script on results is currently being prepared. Nonew sampling has occurred since 2000.

Plant sampling. The plant community and otherhabitat variables were described for each 50-m ra-dius bird count area along four habitat sampling ra-dii radiating from the center of each point count sta-tion following procedures detailed in Riffell et al.(1996). The understory habitat was described us-ing a Wien’s pole (Rotenberry and Wiens 1985)following procedures of Riffell et al. (2001), modi-fied as appropriate to adapt them to forestedhabitat.

More traditional plant sampling, of a subset offorested wetlands, was done by Mike Kost andDennis Albert of the Michigan Natural Features In-ventory during 1999 and 2000 using quadrat sam-pling along transects running through each wetland.They have submitted a report on their results to theMichigan Department of Environmental Quality in-cluding recommendations on potential metrics forthese wetlands.

37

Contact Mark Gernes

Organization Minnesota Pollution ControlAgency (MPCA)

Environmental Outcomes Division520 Lafayette RoadSt. Paul, MN 55155

Phone (651) 297-3363


Website http://www.pca.stat.mn.us

Minnesota: Developing Wetland Biocriteria

Contact Judy Helgen

Organization Minnesota Pollution ControlAgency (MPCA)

Environmental Outcomes Division520 Lafayette RoadSt. Paul, MN 55155

Phone (651) 296-7240


Website http://www.pca.state.mn.us

Purpose

Develop the tools to assess wetland conditionby studying vegetation and invertebrates in wet-lands across a range of human disturbance

Develop two separate Indexes of BiologicalIntegrity (IBI) for Minnesota depressional wet-lands

Wetland Type

Depressional wetlands, large and small; ripar-ian wetlands

Assemblages

Macroinvertebrates

Vascular plants

Status

Established indexes of biological integrity (IBIs) forplants and macroinvertebrates. Report completedon IBIs for 44 large depressional wetlands in May2002. In 2002, evaluating data on statistical analy-sis of sampling methods in 9 (invertebrates) and 12(vegetation) large depressions. Beginning June2002, testing IBIs in new ecoregion, comparing IBI

assessments with rapid assessment method and citi-zen monitoring data. Training of many citizen com-munity teams in biological assessment of local wet-lands continued in 2002; a pictorial guide will beproduced in 2002.

Project Descriptions

Minnesota has been developing its wetland bio-logical assessment program since 1992.

1992 – Small depressional reference wetlandstudy

The first project was funded by the State legisla-ture to develop biological reference conditions forsmall depressional wetlands in central Minnesota.Subsequent funding is primarily from EPA. The ini-tial research studied the quantity and quality ofmacroinvertebrates primarily in highest quality, leastdisturbed reference sites and relatively few disturbeddepressional wetlands.

1995 – Expanded small depressional wetlandstudy

A second project funded by EPA was undertakento develop multimetric IBIs for depressional wet-lands in central Minnesota. During the 1995 sam-pling season, MPCA collected data on macro-invertebrates and vegetation for a larger setof depressional wetlands than the 1992 study, rep-

38

resenting a wide range of impairment. MPCA’sreport on the 1995 IBI development was completedin 1999 and will be available on MPCA’s web site(www.pca.state.mn.us).

1999 – Large depressional wetland studyMPCA sampled large depressional wetlands, in-

cluding duplicate sampling at six of the sites, to vali-date metrics developed with smaller depressionalwetlands in 1995. MPCA also developed a scor-ing system for human disturbance (HDS) that in-corporates estimates of disturbances in the bufferarea and in the near-wetland landscape, plus rangesof chemical pollution and alterations within the wet-land. The HDS scores were used as the x-axis tocalibrate the metrics and the IBI scores. Both thevegetation and the invertebrate IBI scores show sig-nificant relations to the human disturbance scoresand to various water and sediment chemical fac-tors. The invertebrate IBI was significantly relatedto HDS, turbidity, phosphorus, chloride, and otherfactors. The vegetation IBI was significantly re-lated to HDS, phosphorus, and chloride in water;and to copper; zinc; and nickel in sediments. Mi-nor changes were made in metrics compared withthose used for the smaller depressions. The finalreport for this project, Indexes of Biological Integ-rity (IBI) for Large Depressional Wetlands in Min-nesota, by M.C. Gernes and J.C. Helgen, was com-pleted in May 2002. In 1999, 27 wetlands inriparian areas of small and medium-sized streamsin the St. Croix River basin were sampled for thevegetation IBI. There data are currently underanalysis.

2001 – Statistical assessment of wetlandmonitoring methods

MPCA sampled nine wetlands in three locationsto test the methods for the invertebrate and vegeta-tion IBIs with statistical procedures. These dataare currently under analysis.

2002 – Citizen trainingIn 2002, MPCA expanded efforts for training citi-

zens in biological assessment of wetlands, training

90 citizens in the invertebrate and vegetation meth-ods (see Minnesota case study entitled “DakotaCounty Wetlands Health Evaluation Project”). Afinal guide for biological monitoring of wetlands bycitizens will be produced by MPCA in 2002. Thisguide includes pictorial keys to wetlands inverte-brates and plants. MPCA also trained school teach-ers in the assessment of ephemeral wetlands in thespring of 2001 and 2002.

In the summer of 2002, MPCA is validating theIBIs for depressional wetlands in 40 wetlands insouthern Minnesota to test regional applicability ofthe methods. In addition, about 10 wetlands moni-tored by citizens will be assessed by MPCA usingthe technical IBIs. The data from the assessmentsby citizens using the modified IBIs and MPCA’stechnical IBIs will be compared. The IBI resultswill be compared with results from a Minnesotarapid assessment method carried out by consult-ants on the same 10 wetlands.

Study Design

In the 1992 project on reference wetlands, 32least disturbed and 3 known disturbed wetlandswere sampled for macroinvertebrates. In the sec-ond research phase in 1995, 27 wetlands weresampled for macroinvertebrates and vegetation.The 27 wetlands represented the full range of hu-man disturbance typical of wetlands in this part ofMinnesota, and were located in the North CentralHardwood Forest ecoregion (Figure 9). The sitesincluded 6 least impaired reference sites and 21 siteshighly impacted from human disturbances such asstormwater (12 sites) and agricultural influences (9sites).

In the 1999 large depressional wetlands project,44 sites were sampled, 6 of which were sampledtwice on the same date. Included were 14 high-quality reference sites, 14 agriculture-influencedsites, and 16 urban wetlands (Figure 10). Thesewetlands were selected to represent the widest rangeof disturbance.

39

In the 2001 statistical assessment of wetlandsmethods project, three sectors of nine wetlandswere sampled for invertebrates and vegetation.

In 2002, approximately 50 large depressionalwetlands representing a gradient of human influenceare being assessed. Approximately 30 wetlands inthe Western Cornbelt Plains and the Northern Gla-ciated Plains ecoregions are being analyzed to de-termine if the IBI developed in the North CentralHardwood Forest (NCHF) ecoregion can be usedin other regions of Minnesota. Five of these siteswill have duplicated sampling. Ten wetlands fromprevious projects in the NCHF ecoregion will beanalyzed. Another 10 wetlands in the NCHFecoregion are being sampled by citizens using theIBIs and by consultants using a rapid assessmentmethod.

In the projects in 1995, 1999, 2001, and 2002,both invertebrate and vegetation sampling weredone. Water chemistry samples were taken in June,

Figure 9: Locations of smaller depressional wetlandssampled in 1995.

Figure 10: Locations of the 44large depressional wetlands

sampled in 1999.

40

and sediments were cored for metals and other fac-tors in later summer.

In addition to invertebrate and vegetation sam-pling, wetlands were sampled for water pH, con-ductance, turbidity, dissolved oxygen, temperature,calcium (hardness), chloride, total suspended sol-ids, total phosphorus, and total nitrogen. Sedimentswere analyzed for 15 heavy metals using ICP meth-ods, as well as for total organic content, texturalclasses, carbonates, chloride, total phosphorus, andtotal nitrogen. Low-altitude aerial photographs weretaken to support the scoring for the human distur-bance gradient.

Sampling Methods: GeneralConsiderations

Stratification of the habitat for each study site wasdone so as to minimize the biological variabilityamong the different strata of the wetland. For thedepressional wetland projects, the following habi-tats were identified: nearshore emergent zone, openwater submergent zone, and floating plant zone.


Sampling was done in the nearshore emergent zoneduring the seasonal index period of June to earlyJuly. This time frame ensured that optimal speciesmaturity and richness were present. In previousfieldwork, MPCA had determined that sampling inMay was too early because some invertebrates aretoo immature for identification. Once collected, in-vertebrate samples were preserved and analyzedin the laboratory. Macroinvertebrates were sampledusing both the dipnet and the activity trap method.Dipnetting captured the greatest richness of inver-tebrates and the activity trap captured the activeswimmers and night-active predators.

DipnetThe D-frame aquatic dipnet with 600-micron mesh

net was used. Two dipnet samples were taken withinthe emergent vegetation zones. A ½O wire screenfixed to a wooden frame was used to keep the veg-etation from the sample. After sweeping the netstrongly through the water column four to five timesand downwards to near the bottom, the contents ofthe dipnet are emptied onto the framed wire screen.The frame is set over a pan containing sieved waterto catch invertebrates as they drop down from thevegetation. For approximately 10 minutes, the veg-etation is gently spread and invertebrates are en-couraged to drop or crawl down to the water in thepan. After separation from the vegetation, the wa-ter is then poured through a 200-micron sieve toconcentrate the sample before preservation. Pres-ervation of the samples is done using 80% ethanol,final concentration. Using a squirt bottle with thealcohol solution, the sample is back-flushed into thesample jar and labeled for later picking and identi-fication.

Activity trapThe bottle trap works as a passive funnel trap that

collects organisms as they swim into the funnel andpass through the neck into the bottle. Made from aclear 2-L round-bottomed plastic beverage bottle,the traps are nearly invisible underwater. The trapsare supported on a 4.5N wooden dowel with a flex-ible half section thin wall PVC pipe that allows rais-ing and lowering the bottle trap on the dowel.

Ten bottle traps were placed in the emergent veg-etation zone and left overnight for two consecutivenights. Placement of the bottle traps was from thenearest shallow shore edge to the inner side of thedeeper emergent vegetation zone in water no greaterthan 1 meter. In the shallowest water, the trapswere placed on the bottom just under the surfaceof the water. Traps were placed horizontally about15-20 cm under the surface. The bottle traps wereback-filled with water leaving no air bubbles insideto reduce the amount of predation within the trap.

41


After the invertebrates were sorted in the lab, theentire sample was identified and counted. The datawere entered into an ACCESS database andscored for 10 metrics, which represent differentmeasures of the invertebrate community. The metricsused for small depressions are in the 1999 reportof the 1995 project. The metrics used for largedepressions are in the 2002 report on the 1999project (Table 1). For the 1995 data, the metricswere validated by plotting them against a ranking ofthe site disturbance based on professional judgmentand against selected chemical variables. For 1999,the metric data were plotted against the human dis-turbance gradient scores (HDS) and some chemi-cal factors to evaluate metric responses. In addi-tion, linear regressions were done on metric dataand IBIs against the measures of human disturbance.

Chemical (e.g., phosphorus, nitrogen, chloride,and heavy metals) and biological data were ana-lyzed for statistically significant relationships to themetrics and IBIs. Of the 10 invertebrate metrics,intolerant taxa, chironomid taxa, and total taxashowed the strongest responses to the estimateddisturbance gradient and water chemistry factors,followed by the Odonata and ETSD metrics. HDSscores, turbidity, and phosphorus and chloride inwater were most significantly related to the inverte-brate IBI; copper in sediments was significant.

Sampling Methods: Vegetation

Vegetation sampling techniques vary greatly for dif-ferent wetland habitats and study designs. MPCAused a releve method for sampling vegetation. Thereleve method was chosen for several reasons. Theprimary reason is that it is easily adapted to widelyvarying habitats and vegetation community struc-ture, which is typical for depressional wetlands. Thisadaptability in sampling methodology is needed for

wetlands that receive significant quantities of waterduring storm events. A second reason that MPCAselected releve sampling over line transect or quadratsampling was that the Minnesota Department ofNatural Resources’ Natural Heritage Program andCounty Biological Survey use releves, andMinnesota’s academic researchers also use a simi-lar releve method for collecting vegetation commu-nity data.

All vegetation sampling was done in July, whichrepresents the typical period of maturity and thebest time in Minnesota for determining communitystructure. In each wetland, a 100-m2 plot was es-tablished in representative sampling locations withinthe emergent and open water submergent zones.After establishing the releve plot, the vegetationcover classes were determined for each plant spe-cies occurring in the plot. Voucher specimens werecollected at least once during the project for eachspecies or taxon encountered. All taxa that couldn’tbe identified reliably to the species level in the fieldwere also collected.

Analytical Methods: Vegetation

Ten vegetation metrics were developed and vali-dated, using methods similar to those for the inver-tebrate IBI. Each promising attribute of the plantcommunity was plotted against a disturbance gra-dient. In the 1995 small depressional wetlandsproject, the disturbance gradient index was devel-oped from professional judgment ratings of severaldisturbance factors including stormwater input, ag-ricultural practices, quality of adjacent buffers, hy-drologic alterations, and historical disturbances. Forthe 1999 large depressional wetlands project, theHDS scores and chemical factors were used. Metricscoring criteria were then developed for the stron-gest responding metrics (Table 2). Metrics werealso plotted against chemical variables to demon-strate their response to traditional water chemistryconcerns.

42

Table 1: Scoring Criteria for 10 Invertebrate Metrics for IBI forLarge Depressional Wetlands

The 10 vegetation metrics showed significant re-sponses to water (chloride and phosphorus) andsediment chemistry (zinc, copper, and nickel). Sen-sitive species were found to be the strongest andmost reliable vegetation metrics.

Sampling Methods:Water and Sediment

Sampling water and sediment chemistry was con-ducted by MPCA staff. Water analysis was doneby the Minnesota Department of Health, and sedi-

ment analysis was done under contract with theUniversity of Minnesota Soils Analytical Labora-tory. See April 1999 report for sampling methodsfor sediment and water chemistry.

Lessons Learned

Vegetation IBIs show great promise for futureapplications in wetlands biological assessment.The wetland plant community is biologically richand sensitive to a variety of human disturbances.Data are acquired in a short time.

43

Table 2: Scoring Criteria for 10 Vegetation Metrics for LargeDepressional Wetlands

Biological metrics respond to many stressorsand to other disturbance factors in the land-scape and in physical aspects of the wetlandsthat can be readily measured.

After applying appropriate stratifications, wefind that wetlands are not chaotic or highly vari-able. They show clear patterns and predict-able responses to human disturbances.

The invertebrate community is sensitive to manydisturbances in wetlands. Responses differ fromstream invertebrates because wetlands inver-tebrates are adapted to daily cycling of dissolvedoxygen in wetlands. What may be a “pollutiontolerant” invertebrate in streams may be a spe-cialist in the wetland habitat.

44

Minnesota: Dakota County Wetland HealthEvaluation Project

Contact Daniel HuffEnvironmental Education Coordinator

Organization Dakota County Environmental Education Program4100 220th St. West, Suite 101Farmington, MN 55024-9539

Phone (651) 480-7734


Website http://www.extension.umn.edu/county/dakota/Environment wetlands/wetld.html

Purpose

Evaluate wetland health using biological datagathered by citizen volunteers using approvedtechniques developed by the Minnesota Pollu-tion Control Agency

Increase biodiversity in wetlands in urban ar-eas by installing Best Management Practices(BMPs)

Conduct a public wetland education effortthrough seminars, workshops, field days, andmedia

Wetland Type

Depressional wetlands

Assemblages

Macroinvertebrates

Vascular plants

Status

Since 1997, the project has operated every sum-mer. In 2002, the program consisted of 15 moni-toring teams representing 19 communities in Da-kota and Hennepin Counties.

Project Description

The Dakota County Wetland Health EvaluationProject (WHEP) uses sampling methods and evalu-ation metrics developed by the Minnesota Pollu-tion Control Agency (MPCA). The project startedin 1997 and was conducted by MPCA and Minne-sota Audubon. A total of 30 wetlands were moni-tored by 5 citizen teams representing the Minne-sota Zoo, Dakota County, and the cities ofBurnsville, Eagan, and Lakeville. MPCA stafftrained volunteers in sampling protocols, qualityassurance, and plant and macroinvertebrate identi-fication. The sampling methods are a scaled-backversion of the Minnesota Pollution Control AgencyWetland Bioassessment Program (see “Minnesota:Developing Wetland Biocriteria” in this module).Each citizen team worked under the direction of alocal teacher or nature center staff. The time com-mitment for volunteers was approximately between40-50 hours per year, including training, field work,and analysis.

45

In 1998, the project expanded to eight citizenteams, and a technical consultant was hired to con-duct MPCA’s full sampling methods to facilitatecomparisons. Results of the comparison indicatethat the volunteer assessments, although not as rig-orous as the professional assessments, provide re-peatable results that are consistent with the moredetailed assessments. However, the volunteers tendto score the high-quality wetlands too low becausethey are unable to identify as many organisms (e.g.,species of Carex) resulting in lower scores for therichness metrics. A detailed description of thisproject is provided on the web site listed under thecontact information (http://www.extension.umn.edu/county/dakota/Environment/wetlands/wetld.html).

In 1999, the funding source changed from EPAgrants to funding from the Minnesota State Legisla-ture through the Minnesota Environment and Natu-ral Resources Trust Fund. The Minnesota StateLegislature approved continued funding for thisproject for the period of 1999 to June 2002. Forthe 1999 season, the project expanded to a total of10 cities. An additional 11 wetlands were assessedfor the first time and 24 previously sampled wet-lands were resampled for trend analysis. The in-vertebrate and vegetation IBI scores were gener-ally consistent for each wetland, although the inver-tebrate score was slightly lower on average thanthe vegetation scores. Each team performed a cross-heck on a wetland monitored by another team. Inaddition, the technical consultant field checked 4 ofthe 35 wetlands sampled by the volunteer teams.The volunteer results were compared to MPCAstandard sampling method results and were consis-tent for most samples.

In 2000, 10 city teams monitored 38 wetlands.Invertebrate scores were generally lower than veg-etation scores, a more significant difference than theprevious year. Below-normal precipitation for theprevious fall, winter, and spring may account forthe poor showing of invertebrate populations. Vol-unteers continued to use MPCA protocols and betrained by the MPCA scientists who developed the

protocols. Eight of the 10 teams performed a cross-check on a wetland monitored by another team.Of these, six of the eight sites showed consistentscores for the invertebrate index and seven of theeight sites showed consistent scores for the vegeta-tion index. The technical consultant field-checked4 of the 38 wetlands sampled by the volunteer teams.In general, the citizen data were consistent withconsultants’ findings.

In 2001, 10 teams representing 11 cities moni-tored 41 wetlands. A wet spring may have contrib-uted to higher invertebrate scores than the 2000monitoring season. Vegetation scores were con-sistent with previous seasons’ scores. Twenty-eightof 39 wetlands sampled for both vegetation andinvertebrates were considered to have consistentscores between the 2 indexes. Seven city cross-checks were performed using the invertebratemetrics. Of these seven, five resulted in similarscores. Six city cross-checks were performed us-ing the vegetation metrics. Of these six, four re-sulted in similar scores. The technical consultantfield-checked 3 of the 41 wetlands sampled by thevolunteer teams. The consultant’s check resultedin identical scores for two of the three volunteerteams and was very similar for the third for the in-vertebrate metrics and the vegetation metrics. Thisshowed higher consistency between the citizen teamsand the professional cross-check than in previousyears. In addition to the 10 Dakota County teams,3 teams within Hennepin County monitored 14wetlands with funding from U.S. EPA and theMinnehaha Creek Watershed District.

In addition to wetland monitoring, a wetlandremediation project was begun in 2001 at one ofthe wetlands previously monitored by the project.Cedar Pond in Eagan, MN, had scored among thelowest of all county wetlands for both vegetationand invertebrates when sampled by volunteers in2000. In cooperation with the City of Eagan, aretaining wall surrounding the pond was removed,the slope was regarded, and three zones of nativewetland vegetation, emergent, wet meadow, and

46

upland buffer, were planted around the wetland.Three rainwater gardens were constructed andplanted to receive and filter runoff from an adjacentcity street. WHEP volunteers participated in plant-ing the rainwater gardens and trees along the up-land buffer.

Complete reports of the 1999, 2000, and 2001monitoring season results can be found on the website given above. Information on the 2001 moni-toring season within Hennepin County can be foundat http://www.hcd.hennepin.mn.us/whep.html.

In 2002, communities sponsoring teams are pro-viding funding for the project. Nine teams repre-senting 11 cities in Dakota County and 6 teams rep-resenting 4 cities and 1 watershed district inHennepin County are participating in the Project.Results from the 2002 monitoring season for bothcounties will be available on the web in February2003. The project is expected to continue.

47

Minnesota: University of Minnesota’sIBI Development

Contact Sue Galatowitsch

Organization University of MinnesotaDepartment of Horticultural Science305 Alderman Hall, 1970 Folwell Ave.St. Paul, MN 55108

Phone (612) 624-3242


Website http://www.hort.agri.umn.edu/mnwet/begin.htm

Purpose

Develop assessment methods to evaluate ecologi-cal condition of wetlands

Wetland Type

Variety

Assemblages

Amphibians

Birds

Fish

Macroinvertebrates

Vascular plants

Status

Analyzing data and writing reports

Project Description

Developing indices of biological integrity (IBIs)for Minnesota was pursued by researchers at theUniversity of Minnesota to enable quality assess-ments of existing and restored wetlands. Eight

series of 15 wetlands (120 sites) were used to de-velop wetland IBIs. Each series covers a majorwetland type in the State and is composed of refer-ence sites (unaltered wetlands in an unimpaired set-ting), sites surrounded by land use typical of theregion, and sites that are highly altered. Plants, birds,fish, invertebrates, and amphibians were surveyedto select the best IBIs for each series.

To identify possible patterns in biological com-munities that may relate to land use differences, eachorganismal data set (except amphibians-few organ-isms encountered during surveys) was explored withTWINSPAN. TWINSPAN organizes data so thatthe most similar sites (as described by their spe-cies) are grouped together as columns on a table,and so that the species with similar habitat affinities(as described by sites where they occur) are groupedtogether as rows. This TWINSPAN table was usedto generate a list of potential indicators for analysiswith land-use data. Species or groups of speciesthat appeared preferential to sites with similar landuse characteristics were deemed to be potential in-dicators. Other common ecological measures, suchas richness (number of species), were routinely in-cluded in the list of potential indicators, as well.Twenty-eight potential indicators identified for thisseries are listed below. Proportional indicators foranimals are calculated as a total of all organisms

48

observed (not a proportion of taxa) unless noted.Absolute abundances for plant species (vegetation)cannot be reliably estimated from cover class data.Importance values, an approximate measure ofabundance, were calculated by summing cover classscores (r=0.1, +=0.5, and classes 1-5).

Potential indicatorsAmphibians: None

Birds (12): Species richness (BSR); total abun-dance of all birds (BABU); number of wetland taxa(BWR); proportion of wetland birds (PWB); abun-dance of brown-headed cowbirds (ABHC); pro-portion of forest-nesting birds (PFB); number oftaxa with large territories (BLTR); proportion ofinsectivorous birds (PBIN); abundance of marshand sedge wrens (CIS); abundance of yellowthroats, swamp sparrows, and LeConte’s sparrows(SSYT); number of open ground nesting species(BOGR); proportion of open ground nesting spe-cies (PBOG).

Fish: None

Invertebrates (4): Total abundance (IABU), taxarichness (ISRI), number of snail species (GASR),proportion of snails (PGAS).

Vegetation (12): Species richness (VSR), inva-sive perennial species importance (IPI), number ofCarex species (CAR), importance of Carex spe-cies (CAI), number of native herbaceous perenni-als (HPNR), importance of native herbaceous pe-rennials (HPNI), number of native perennialgraminoids (GPNR), importance of native peren-nial graminoids (GPNI), number of introduced spe-cies (INR), importance of introduced species (INI),ratio of graminoids to herbaceous species (VGH),ratio of annuals to perennials (VAP).

Relationship of potential indicators to land-use measures

Values for each potential indicator (PI) were cal-culated for each site in the series. PI values werecorrelated with land use data: site alteration score,land use cover at 500 m, 1,000 m, 2,500 m, and5,000 m radii. For the radii data, six correlationswere calculated, one for each land cover category(agriculture, urban, disturbed, forest, grassland, andwetland). For this series a total of 700 relation-ships were tested. Relationships with Pearson cor-relation coefficients greater than 0.53 (p < 0.1) areworthy of further consideration as indicators ofwetland quality. Each of these relationships wereplotted to detect if the high coefficients were basedon outliers. Those with outliers were not consid-ered significant.

Eighteen of the original 28 potential indicators(64%) were found to have a high correlation to landuse. Seven of the bird PIs show a land-use rela-tionship. Birds with large territories (BLTR) aremore common to sites with less agriculture and dis-turbance at the regional scale (2,500-5,000 m).Likewise, regional patterns of urbanization are nega-tively associated with wetland bird richness (2,500m). Wetland bird richness is greater on sites withmore wetlands at this same scale. Overall bird rich-ness is lower with more surrounding urbanization atmost scales. In contrast, most of the seven promis-ing vegetation PIs show stronger land-use relation-ships at local scales. Introduced species (INI, INR)and annuals (VAP) are more common on sites thatare immediately impacted by agriculture and urban-ization (site). Four invertebrate PIs will be furtherconsidered for indicator development. Therichness of snail taxa (GASR) is positively associ-ated with forest and wetland cover from 1,000 to5,000 m.

49

Montana: Developing Wetland BioassessmentProtocols To Support Aquatic Life Beneficial

Use-Support Determinations

Contact Randall S. Apfelbeck

Organization Montana Department of Environmental Quality2209 Phoenix AvenueP.O. Box 200901Helena, MT 59620-0901

Phone (406) 444-2709


Website http://www.deq.state.mt.us/ppa/mdm/Wetlands/Lakes&Wetlands_Index.asp

Purpose

Montana’s wetland research objectives are to:

Determine the status and trends in wetland wa-ter quality.

Acquire an understanding of how climate, hy-drologic controls, and geomorphic settings in-fluence wetland biological communities for thedevelopment of successful biocriteria.

Develop biological measurements that could beused in developing biocriteria to define the ex-tent and degree of anthropogenic impacts towetland water quality.

Develop an integrated assessment of wetland,streams and lakes within a watershed.

Report on aquatic health at the watershed levelthrough the development of landscape assess-ment tools.

Develop rapid field assessment protocols toevaluate aquatic ecological conditions by usingindicators and best professional judgement.

Wetland Type

Variety of wetland types

Assemblages

Algae

Macroinvertebrates

Vegetation

Amphibians and aquatic reptiles

Status

Ongoing, revising analytical methods

Project Descriptions

1992Montana Department of Environmental Quality

(DEQ) began developing wetland biological crite-ria. At that time, there was little information con-cerning the status or trends of the water quality ofMontana’s wetlands. Furthermore, Montana’swater quality standards were developed to protectthe beneficial uses (e.g., aquatic life) of lakes; river

50

and stream wetlands were not considered Statewaters when Montana’s water quality standardswere developed. Since 1992, Montana has had anongoing program to develop bioassessment proto-cols and water quality standards that will more ad-equately evaluate and protect the aquatic life thatlive in wetlands.

In 1998, as a result of a TMDL lawsuit, MontanaDEQ shifted focus from development of biologicalcriteria to the development of guidelines for makingbeneficial use decisions that apply to all Montanawaterbodies. Montana’s guidelines for making ben-eficial use decisions can be found at http://nr is .s ta te .mt .us/wis/ tmdlapp/pdf2002/Appendix_A.pdf .

1998In conjunction with Montana DEQ’s research pro-

gram, Montana State University (MSU) designeda study that focused on development of vegetationbiocriteria for western Montana depressional wet-lands (Borth 1998). The focus on vegetationbiocriteria is key in Montana because wetland veg-etation is easier to assess than macroinvertebratesor diatoms for depressional wetlands that are sea-sonally dry. The MSU study sampled vegetationand also macroinvertebrates and diatoms for 24 de-pressional wetlands with similar climate, hydrology,and water chemistry. The research included sam-pling across three levels of human disturbances—minimally impacted, slightly impacted, and moder-ately impacted. The study also involved two an-thropogenic impairments—dryland agriculture andgrazing.

2000Researchers from the University of Montana con-

ducted a study to determine the effects of naturalvariability on the use of macroinvertebrates asbioindicators of disturbance in intermontane depres-sional wetlands in northwestern Montana (Ludden2000). Their study design included collection ofmacroinvertebrate samples and physiochemical datafrom 15 pristine and 6 disturbed intermontane prai-

rie potholes. Their study also included analysis ofmacroinvertebrate samples that were previouslycollected by Borth (1998). The researchers col-lected macroinvertebrate samples from across threeseasons and from three wetland zones. They usedmultivariate detrended correspondence analysis toordinate the raw macroinvertebrate data andphysiochemical variables as secondary matrices toestablish vectored biplots of correlation. Candi-date metrics were analyzed using univariate analy-sis. They determined that no environmental vari-ables were strongly correlated with themacroinvertebrate data, and many of the metricsvaried across wetland zones and across seasons.Nevertheless, 35 candidate metrics were able todiscriminate between minimally and highly disturbedsites. These wetlands were also sampled intensivelyto develop and test a model for assessing depres-sional wetlands using the hydrogeomorphic (HGM)functional assessment approach.

2002Montana DEQ initiated the development of a

comprehensive watershed monitoring and assess-ment program. The comprehensive program willbe developed to determine the causes, effects, andextent of pollution to aquatic resources (includingwetlands) and for developing pollution prevention,reduction, and elimination strategies.

The comprehensive program has three compo-nents: landscape, site-specific, and rapid assess-ment. First, a landscape-level process to evaluateand rank wetlands and watersheds for protectionand restoration is critical to maximize the use of lim-ited financial and management resources. This pro-cess will use existing digital data and evaluate a rangeof landscape impacts using a Geographic Informa-tion System. Second, Montana will continue todevelop site-specific assessment protocols such asbiocriteria (i.e., for assessing amphibians, vegeta-tion, algae, and macroinvertebrate assemblages).This information is important for managers ofMontana’s water resources for making informedwatershed management decisions. Development

51

of site-specific wetland biological, physical, andchemical assessment tools will be used to evaluateaquatic life beneficial use support to determine im-pairment as per section 303(d) of the Clean WaterAct. Third, like most States, Montana has verylimited resources for assessing wetland water qual-ity and aquatic conditions for 305(b) purposes. Forthis reason, Montana will develop an assessmentapproach that is truly rapid and cost effective. Arapid assessment component will be developedbased upon field data collected, tested and refinedduring this study, best professional judgment andinformation gathered from a literature search.

The comprehensive watershed monitoring andassessment program will include a probabilistic studydesign to assess the ecological condition of depres-sional and riverine wetlands and streams. Threepilot subbasins (fourth code HUCs) will be selected.The pilot subbasins will represent Rocky Moun-tain, Intermountain Valleys and Prairie Foothills, andPlains ecoregions (Figure 11).

Within each pilot subbasin, three sixth-code HUCs(fifth-code HUCs for the Middle Milk subbasin)will be randomly selected from each of three dis-turbance strata (high, medium, and low) determinedfrom landscape-scale assessments. For vegetation,

soils, water chemistry, macroinvertebrates, diatoms,and land use, a total of six wetlands (three riverineand three depressional) will be monitored in eachwatershed. A total of 27 watersheds (9 in eachsubbasin) and 162 wetlands (3 riverine and threedepressional in each watershed) will be studied. Forthe determination of the detection/nondetection ofamphibians and aquatic reptiles, all standing waterbodies identified in each watershed will be surveyed.

Note: the remainder of this case study will de-scribe in more detail the study initiated in 1992.

Study Design

The original Montana DEQ study was designedin 1992 and involved sampling 80 wetlandsthroughout Montana during 1993 and 1994. Thebioassessment project included collection ofmacroinvertebrate and diatom samples from wet-lands in all ecoregions of Montana (Apfelbeck2001).

Montana DEQ’s approach to developingbiocriteria involved several study designs aimed atdeveloping tools to help detect human influence onwetland water quality. The original study was de-signed in 1992 and involved sampling 80 wetlands

Figure 11: Montana ecoregions and pilot watersheds.

52

throughout Montana during 1993 and 1994 (Fig-ure 12). The study design included collection ofsamples that represent the wetland’smacroinvertebrate (e.g., aquatic insects) and dia-tom (algae) communities. Sampling methods weredesigned so that 1-2 hours was sufficient for the

data to be collected in the field for each wetland.Samples of each wetland’s water column, sediment,and macroinvertebrate and diatom communitieswere collected. Water-column and sedimentsamples were collected to document impairmentsand for classification purposes.

Wetland Classes: (referred to by number in the above figure)Class 1 Dilute Closed Basins and Headwater Wetlands of the Rocky Mountain EcoregionClass 2 Riparian Wetlands of the Rocky Mountain amd Intermountain Valley EcoregionsClass 3 Groundwater Recharge Closed Basin WetlandsClass 4 Riparian Wetlands of the Plains EcoregionClass 5 Alkaline Closed Basin WetlandsClass 6 Saline WetlandsClass 7 Surface Water Supported Closed Basin WetlandsClass 8 Ephemeral WetlandsClass 9 Open Lake Wetlands of the Plains EcoregionClass 10 Open Lake Wetlands of the Rocky Mountain and Intermountain Valley Ecoregions

Figure 12: Ecoregions and sampling locations by wetland class.

Ecoregions of Montana

53

The sites were classified using ecoregions andhydrogeomorphology, and several of the wetlandclasses were further delineated using water-columnchemistry variables. A representative number ofwetlands from the following Omernik ecoregionswere sampled: Rocky Mountains, IntermountainValleys and Prairie Foothills, Glaciated Plains, andUnglaciated Plains Ecoregions. To reduce season-ality, all wetlands within the same ecoregion weresampled during similar time periods. Wetlands inthe Glaciated Plains Ecoregion were sampled fromearly April through mid-June, wetlands of the Inter-mountain Valleys and Prairie Foothills Ecoregion frommid-June until early August, and wetlands of theRocky Mountain Ecoregion from early July throughSeptember.

The classification framework was developed bysampling the full spectrum of wetland types in Mon-tana. The study was designed such that approxi-mately 75% of the sites were reference and 25%were impaired. This approach was useful becauseit allowed Montana DEQ to determine the refer-ence condition of a wide variety of wetland types.Also, the design provided the opportunity to testthe ability of the biological measurements to detectwater quality impairment.

Anthropogenic impacts such as irrigation or log-ging were included in the study design. If anthro-pogenic activities such as dryland agriculture, irri-gation, feedlots, grazing, silviculture, road construc-tion, hydrologic manipulation, urban runoff, waste-water, mining, and oil and natural gas productionoccurred in the wetland’s watershed, the wetlandwas considered impaired. Wetlands for samplingwere selected on the basis of many variables, in-cluding availability of historical data, special inter-ests by other entities, cooperation by landowners,and accessibility.

In order to classify or document impairment, ahydrogeologist for the Montana Natural HeritageProgram (MNHP) assisted Montana DEQ in de-

veloping a wetland classification system throughsummarizing and interpreting the physical and chemi-cal data. Using topographic maps, field observa-tions, and information gathered from land manage-ment agencies, geomorphic characteristics were in-terpreted and a hydrogeomorphic database devel-oped. Maps for each wetland were completed us-ing a Geographic Information System (GIS). Mapfeatures included hydrologic delineations, land man-agement areas, counties, cities, major transporta-tion corridors, wetland watershed boundaries, andsampling locations. Visit Montana DEQ’s websiteto get more detail on the types of wetland classesand for photographs of each type: http://www.deq.state.mt.us/ppa/mdm/Wetlands/classification.asp.

Sampling Methods: Diatoms

Montana DEQ collected diatoms as compositegrab samples. The algae were identified to the lowesttaxonomic level possible. Samples were collectedusing a 250-mL plastic container and then preservedwith Lugol’s solution. Samples were collected froma location determined to best represent the wet-land. These locations were restricted to areas thatwere easily accessible when wearing hip boots.Sampling was done from April through September.Each site was sampled once.

Analytical Methods: Diatoms

The multivariate approach was used to analyzewetland diatom communities. Multivariate analysisis a statistical approach used by biologists to deter-mine relationships among biota such as diatoms ormacroinvertebrates, and environmental variablessuch as water-column chemistry. The multivariateapproach to investigate relationships between Mon-tana wetland diatom assemblages and environmen-tal variables (mostly water-column chemistry) wasdetrended canonical correspondence analysis(DCCA) and two-way indicator analysis

54

(TWINSPAN). Clusters of diatoms with similarcomposition were graphically displayed usingDCCA. Vectors were displayed and labeled toillustrate the relationship between diatom assem-blages and environmental variables. Longer vec-tors show us a stronger correlation among diatomassemblages and environmental variables (Figure13). Envelopes were used to graphically encloseall reference sites using the wetland class delinea-tions (Figure 14). The Academy of Natural Sci-ences of Philadelphia (ANSP) performed thesubsampling, digestion, and mounting of the dia-toms (see Charles et al. 1996).

Figure 13: Diatom data related to abiotic factors using

Detrended CanonicalCorrespondence Analysis.

Figure 14: Clusters of wetlandsbased on diatom data.

55


Montana DEQ collected macroinvertebrates us-ing a 1-mm mesh D-net in a sweeping motion.Macroinvertebrates were collected from all microen-vironments in a sampling location. These locationswere restricted to areas that were easily accessiblewhen wearing hip boots and that best representedthe wetland. Samples were composited with asso-ciated materials such as vegetation and sedimentand then preserved with ethanol in a 1-L plasticcontainer. An effort was made to collect 300 or-ganisms from each location using a consistentmethod of collection. To ensure preservation,sample bottles were refreshed with ethanol severaldays after collection. Sampling was done once persite from April through September.


The multimetric approach was used to evaluatewetland macroinvertebrate communities. Themultimetric approach incorporates many attributesinto the assessment process and has the ability tointegrate information from the biological communi-ties to provide an overall indication of biologicalcondition or ecological health.

The project contractor assisted Montana DEQ indeveloping wetland macroinvertebrate multimetricindices. The proposed metrics and associated en-vironmental data were evaluated in an attempt todevelop an understanding of ecological relationships,to test each proposed metric’s ability to predictvarious anthropogenic stressors, and to test redun-dancy. Table 3 lists some of the proposed metricsused for macroinvertebrates and Figure 15 showsmacroinvertebrate index values for several classesof wetlands.

The macroinvertebrate samples are subsampledand sorted by contractors using a gridded sorting

pan. A subsampling of 200 organisms is performedfor analysis. All individuals are counted when thereare less than 200 organisms in the sample. Organ-isms are identified to the lowest taxonomic level (atleast genus if possible). The taxonomic level of iden-tification is standardized using the Montana Streamwhen possible. Amphipoda are identified by spe-cies. Common, easily identified midge taxa are iden-tified using a dissecting microscope equipped with25H oculars. Midge taxa that require slide mount-ing are cleaned with warm water solution of 10%KOX, rinsed in distilled water followed by 95%ethanol, and mounted on slides in Euperal. If fewerthan 30 individuals are to be mounted, all individu-als are mounted. If there are more than 30 midgelarvae, at least 10% of each morphotype aremounted. Mounted midges are identified using aZeiss Axiolab phase contrast compound microscopeor equivalent.

The contractor identified the organisms in thewetland samples and standardized the taxonomiclevel of identification based on Montana StreamProtocols. Several taxa were eliminated from con-sideration for metric development, as they weredetermined to be nonbenthic taxa or semiaquaticsurface dwellers and considered uninformative forreflecting water quality. These taxa includedGerridae, Collembola, Dytiscidae, Hydrophilidae,Ostracoda, Anostraca, Copepoda, Cladocera,Notonectidae, and Corixidae.

Other Parameters: WaterChemistry and Sediment

Water-column and sediment samples were col-lected to document impairments and for classifica-tion purposes. Each sample was collected from alocation determined to best represent the wetland.These locations were restricted to areas that wereeasily accessible when wearing hip boots. Fieldchemical measurements, observations, and photo-graphs were recorded at each location.

56

Table 3: Proposed metrics, proposed metric calculations, and scorecalculations used for developing wetland macroinvertebrate indices

Lessons Learned

We found that diatoms and macroinvertebrateswere most useful for evaluating the biologicalintegrity of perennial wetlands with open-waterenvironments that had relatively stable waterlevels and were not excessively alkaline or sa-line.

We concluded that multivariate analysis was auseful tool for developing a wetland classifica-tion system and that hydrogeomorphology andecoregions were practical approaches to clas-sifying wetlands for the development ofbiocriteria.

We determined that both the multimetric andmultivariate techniques were valuable for de-veloping wetland biocriteria.

In most cases, the multimetric and multivariateapproaches that we used to assess themacroinvertebrate and diatom communities bothidentified the same wetlands as impaired.

Two wetland types in the arid west (includingMontana) are difficult to classify. Wetlands suchas potholes are highly complex and difficult toclassify because of both spatial and temporalvariability. For these wetlands, the hydrology,water chemistry, and biology can change dra-

57

Figure 15: Macroinvertebrate index scores for wetlandsin several classes.

58

matically throughout a season or from year toyear as a result of climatic change. For example,the biological community of a wetland oftenchanges through an increase in salinity or a de-crease in water content caused by drought.

Aquatic macroinvertebrate and diatombiocriteria did not appear to be very useful fordetecting impairment in wetlands that usuallylack surface water. Vegetation biocriteria arelikely to be the most appropriate for assessingthe biological conditions of these wetland types.

A more quantitative approach is needed for themeasurement of wetland physical disturbancesif we are to link physical disturbances tochanges in biological communities.

Note: Montana DEQ has developed a set of pro-posed metrics, proposed metrics calculations, andscore calculations used for developing wetlandmacroinvertebrate indices. These are included inTable 3.

59

North Dakota: Wetland BioassessmentProtocols for Making Aquatic Life Beneficial

Use-Support Determinations

Contact Mike Ell

Organization North Dakota Department of HealthDivision of Water Quality1200 Missouri AvenueP.O. Box 5520Bismarck, ND 04333

Phone (701) 328–5214


Purpose

The primary purpose of North Dakota’s wetlandbioassessment program is to develop wetland wa-ter quality standards for North Dakota. This cur-rently involves developing biological communitymetrics and an Index of Biological Integrity (IBI)for temporary and seasonal depressional wetlands.A secondary project goal is to compare the IBIwith hydrogeomorphic (HGM) functional assess-ments.

Wetland Type


Assemblages

Algae

Macroinvertebrates

Vascular plants

Status

Analyzing data, developing IBIs, and expand-ing project

Project Description

North Dakota Department of Health (NDDH)initiated its wetland bioassessment program in 1993as a component of the State’s strategy to developwetland water quality standards. NDDH began fieldsampling in 1995 and is currently analyzing datacollected from 1995 through 1999 to develop as-semblage metrics, IBIs for the assemblages, andultimately biocriteria. The NDDH wetland sam-pling plan includes vascular plants,macroinvertebrates, and algal assemblages.

Development of a North Dakota samplingprotocol for the State’s wetland types is ongoingthrough cooperative work agreements with NorthDakota State University, Departments of Animaland Range Sciences and Zoology.

Study Design

North Dakota’s study design has continued toexpand in scope since the project was initiated. Ini-tial field sampling began in 1995 and 1996 with 13“least disturbed” temporary and seasonal wetlands.In 1997, the number and range of wetlands wasexpanded to 20 temporary and seasonal wetlands.Four of the original 13 “reference” wetlands sampled

60

in 1995 and 1996 were sampled in 1997 along with16 new wetlands. These wetlands captured a largerrange of disturbance categories.

In 1998, the number of wetlands again expanded.An additional 16 temporary and seasonal wetlandswere added, bringing the total number of wetlandssampled to 36 in 1998 and the total number of ref-erence wetlands to 6. The 16 wetlands added in1998 were part of an HGM classification project.These wetlands were added to facilitate a compari-son of the two assessment methods. In addition tothe 16 HGM wetlands, NRCS personnel conductedan HGM assessment for the 20 wetlands selectedby the Department.

During the 1999 sampling season, a second set of30 temporary and seasonal depressional wetlands,not sampled previously, were sampled for vascularplants. Macroinvertebrates and algae were notsampled in 1999.

Since the inception of the project, temporary andseasonal wetlands have been the focus of IBI de-velopment efforts. Further, all wetland study siteshave been within the Northern Glaciated Plains andNorthwestern Glaciated Plains ecoregions (Figure16), a region commonly referred to as the “PrairiePothole Region.”


NDDH currently uses only the “sweep” or jabnet method for macroinvertebrate sampling; how-ever, in the past, activity traps have also been used.Sampling consists of two site visits in June. Duringeach site visit two samples are collected with theemergent vegetative zone.

Sweep netThe method employs a “D” frame net (0.6-mm

mesh size). At each site within the wetland, thesubstrate (i.e., wetland vegetation and benthos) is

“jabbed” for a distance of 1 m. The disturbed areais then swept two more times to ensure a represen-tative sample of macroinvertebrates is collected.Each sample collected within the wetland will beplaced in a shallow pan where any excess debris orwater can be removed. The “cleaned” sample isplaced in a jar a preserved with buffered formalinto a concentration of 10% by volume.


In the laboratory each sample is washed througha 0.6-mm sieve to remove the preservative andcleaned to remove excess debris. The cleanedsample is placed in a shallow white pan divided inquadrants of equal size. Because each sample typi-cally contains a large number of organisms, it is nec-essary to subsample. A subsample is obtained byrandomly selecting a sample quadrant and countingand identifying the organisms in that quadrant. Ad-ditional quadrants are selected at random and or-ganisms counted and identified until a subsample of300 organisms is counted. For samples with lessthan 300 organisms, every quadrant is sampled. Allorganisms are identified to the lowest taxonomiclevel practical.

Sampling Methods: Algae(Phytoplankton)

Phytoplankton samples are collected just belowthe water’s surface in the middle or deepest area ofthe wetland basin. Each wetland is sampled twotimes in June at the same time macroinvertebratesamples are collected. The sample is collected bysubmerging a clean 200-mL sample bottle just be-low the water’s surface and filling it. The sample ispreserved in the field with M3 fixative to a concen-tration of 2% by volume.

61

Figure 16: Level IV ecoregions of North Dakota.

62

Analytical Methods: Algae(Phytoplankton)

Identification and enumeration of phytoplanktonis made using the phase-contrast inverted micro-scope method (Utermohl 1958). Sample analysisis conducted from a sample aliquot ranging from 2to 15 mL, depending on turbidity. Phytoplanktongreater than 20 micrometers in diameter are enu-merated at H125 magnification. (Note: At H125the entire bottom of the chamber is in view.) Fol-lowing enumeration at H125 smaller algae arecounted at H1,250. At least 250 cells of the mostnumerous algae are counted using the strip countmethod at H1,250 magnification (APHA 1989).Cell volumes are estimated for dominant taxa bymeasuring cell dimensions of 50 to 100 individualsusing the closest geometric formulas of Wetzel andLikens (1991) and Tikkanen (1986). For rare taxa,volume estimates are made from fewer than 50 cellmeasurements. Diatoms will be identified afterclearing in 30% hydrogen peroxide and mountingin Hyrax Mounting Medium.

Sampling Methods:Vascular Plants

Working in cooperation with NDSU, the Depart-ment is evaluating three methods to sample the vas-cular plant community. The first method is morequalitative in nature. It involves a simple inventoryof plant species present within each wetland zone,low prairie, wet meadow, and shallow marsh. Theother two methods, termed the point and quadratmethods, are quantitative. The inventory is con-ducted each time the wetland is visited. The pointand quadrat methods are done once each, usuallyin July or August.

In the point method, 200 points are evenly strati-fied in each wetland zone and the nearest plant spe-

cies recorded. The quadrat method involves plac-ing 15 1-m2 quadrats evenly throughout each wet-land zone. Each species is recorded within thequadrat and a Daubenmire cover class is recorded.With both methods, a secondary species list is madefor species encountered, but not sampled.

Note: For specific details on abiotic samplingmethods for nutrients, trace elements, general wa-ter parameters, and sediment chemistry, directly referto North Dakota Department of Health.

Lessons Learned

NDDH has made a great deal of progress withinthe past 5 years. Our experience, however, hasnot been without problems and mistakes. As men-tioned previously, the first 2 years of the programfocused solely on “least impaired” or reference con-dition wetlands. Although beneficial in testing sam-pling methods, the lack of a disturbance gradient inthe study design did allow for the testing of attributesand the selection of metrics. Therefore, beginningin 1997 the NDDH chose wetlands across a dis-turbance gradient, including both reference wetlandsand degraded wetlands.

NDDH has also found it beneficial to stratify wet-lands based on ecoregion and wetland class. Thisminimizes the amount of variation in the biologicalassemblage and allows more sensitivity in the re-sponse of the metrics to the disturbance gradient.Current IBI development efforts are focusing ontemporary and seasonal depressional wetlandswithin the Northern Glaciated Plains and North-western Glaciated Plains ecoregions. In 2000 theNDDH will be cooperating in two projects to de-velop wetland IBIs for semipermanent depressionalwetlands and for floodplain wetlands along the Mis-souri River.

63

Ohio: Developing IBI Assessment Methodsfor Water Quality Standards and

Regulatory Decisions

Purpose

Test and develop biological criteria for wetlandsusing vascular plants, macroinvertebrates, andamphibians as indices of biological integrity(IBIs) for eventual adoption into the State’swater quality standards

Use results from IBIs to calibrate the Ohio RapidAssessment Method for Wetlands to supportregulatory decisionmaking under the State’swetland antidegradation rule, which requires thatwetlands be assigned to one of three catego-ries based on the wetland’s quality and func-tionality.

Wetland Type


Assemblages

Algae

Macroinvertebrates

Vascular plants

Contact John J. Mack

Organization Ohio EPADivision of Surface Water122 South Front StreetP.O. Box 1049Columbus, OH 43216-1049

Phone (614) 644-3076


Status

Refining metrics and developing IBIs

Project Description

The initial objective of this study is to provide thereference data needed to implement the wetlandwater quality standards and wetland antidegradationrule. The pilot metrics developed from this studyshould enable Ohio wetlands to be assigned to oneof the three regulatory categories. Generally, thestudy objectives are as follows:

1. Develop pilot biological metrics that may beused to evaluate the function and ecological in-tegrity of a wetland. These metrics will be basedon the vegetation, macroinvertebrate, and am-phibian data, and will form the basis for wet-land biocriteria.

2. Identify and describe reference wetlands inOhio’s four main ecoregions; Eastern CornbeltPlains, Erie/Ontario Drift and Lake Plain, Hu-ron-Erie Lake Plain, and Western AlleghenyPlateau. These reference wetlands will be usedto develop biocriteria and also as “goals” forwetland mitigation projects.

64

3. Continue to assess whether the Ohio RapidAssessment Method correlates well with themore in-depth measures of wetland quality, andto test and refine breakpoints between the wet-land categories.

4. Begin to assess the sensitivity of different meth-ods in evaluating the relationship between wet-land quality and the degree of disturbance

The State of Ohio has well-developed biologicalcriteria (or biocriteria) for streams, such as the In-vertebrate Community Index (macroinvertebrates),the Index of Biological Integrity (fish), and the Modi-

fied Index of Well Being (fish) (Ohio EPA 1988a,band 1989). These indices are codified in OhioAdministrative Code Chapter 3745-1. Until re-cently, however, surface waters of the State thatare jurisdictional wetlands were only genericallyprotected under Ohio’s water quality standards.

On May 1, 1998, the State of Ohio adopted wet-land water quality standards and a wetlandantidegradation rule. These wetland quality stan-dards developed narrative criteria for wetlands andcreated the “wetland designated use.” All jurisdic-tional wetlands are assigned the “wetland designated

Figure 17: Ecological regions of Ohio and Indiana(from U.S. EPA 1995).

Ecoregions** of Ohio and Indiana and Neighboring States

65

use.” The State of Ohio did not attempt at thistime to identify multiple wetland functions as wet-land uses, because of the lack of data to developquantitative water quality criteria for wetlands.However, the development of such biocriteria is theultimate goal and the primary thrust of this project.

The key part of Ohio’s current regulatory pro-gram for wetlands is found in the wetlandantidegradation rule. The wetland antidegradationrule categorizes wetlands based on their functions,sensitivity to disturbance, rarity, and irreplaceabil-ity, and scales the strictness of avoidance, minimi-zation, and mitigation to a wetland’s category.Three categories were established: Category 1wetlands with minimal wetland function and/or in-tegrity; Category 2 wetlands with moderate wet-land function and/or integrity; and Category 3 wet-lands with superior wetland function and/or integ-rity. In order to implement the wetland standardsand antidegradation policy, wetlands must be as-sessed on their relative quality. Ohio EPA has de-veloped a draft Ohio Rapid Assessment Method.The Ohio Rapid Assessment Method has provedto be a fast, easy-to-use procedure for distinguish-ing between wetlands of differing quality. How-ever, it does not and was not intended to substitutefor direct, quantitative measures of wetland func-tion (i.e., biocriteria).

Ohio began development of sampling methodolo-gies and began sampling reference wetlands forbiocriteria development in 1996. To date, Ohiohas sampled 56 wetlands located primarily in theEastern Cornbelt Plains Ecoregion located in cen-tral and western Ohio. These wetlands have in-cluded depressional emergent, forested, and scrub-shrub wetlands; floodplain wetlands; fens; kettlelakes; and seep wetlands. The wetlands being stud-ied span the range of condition from “impacted”(i.e., those that have sustained a relatively high levelof disturbance) to “least-impaired” (i.e., the bestquality sites available).

On the basis of results to date (see Fennessy etal. 1998a,b; Mack et al. unpublished data), Ohio’sresearch supports the use of vascular plants,macroinvertebrates and/or amphibians as biologi-cal metrics in wetlands, and also the continued useand development of the Ohio Rapid AssessmentMethod as a rapid assessment tool.

This work has been funded since 1996 by severalEPA Region 5 Wetland Program DevelopmentGrants.

Study Design

Fifty-seven wetlands were sampled during the1996, 1997, 1998, and 1999 field seasons. Thefirst 2 years of data laid the groundwork for stan-dardizing sampling methodologies, classifying wet-lands, identifying potential attributes, and develop-ing metrics using vascular plants, amphibians, andmacroinvertebrates.

In 1996, Ohio EPA monitored a series of riparianforested across a gradient of disturbance (i.e., leastimpacted to impaired) (Fennessy et al. 1998b).Estimates of the relative level of disturbance weremade on a scale of 1 (most disturbed) to 10 (leastdisturbed) based on visual evidence of disturbances,review of aerial photographs of the wetland and thesurrounding area, and interviews with staff from theNatural Resource Conservation Service and/or thelandowner. In 1996 and 1997, Ohio EPA moni-tored 21 forested and emergent depressional wet-lands. Relative disturbance was evaluated using atiered flow chart to assign a relative disturbancescore and also with the score from the Ohio RapidAssessment Method (Fennessy et al. 1998a, Fig-ure 2.2).

Ohio EPA found a good correlation between thescores of the Ohio Rapid Assessment Method scoreand level of disturbance a wetland site has experi-enced. Higher ORAM scores correlate well with

66

lower levels of disturbance based on our model, asdo lower ORAM scores with disturbed sites. In1999, the ORAM score of the site was used as ameasure of the level of disturbance. So far, thisappears to be a highly effective “x-axis” disturbancegradient for the development of IBIs for wetlandplants.

Reference wetlands are sites or data sets fromsites that typify a class of wetlands within a rela-tively homogeneous physiographic region. Refer-ence sites should include wetlands that have beendegraded or disturbed. Site selection in this studyis made using an ecoregional approach and to re-flect a gradient of disturbance (i.e., least-impactedto impaired).

Reference sites are selected such that relativelysimilar proportions of low-, medium-, and high-dis-turbance sites are sampled. To date, almost all thewetlands studied by Ohio EPA have been locatedin the Eastern Cornbelt Plains (ECBP) ecoregion.For the years 2000 and 2001 field seasons, OhioEPA will be studying reference sites in the ErieOntario Lake Plain (EOLP) ecoregion of northeast-ern Ohio.

Sampling Methods and Analysis:Macroinvertebrates and

Amphibians

Below is a detailed description of the samplingmethodology and analysis process used by OhioEPA for macroinvertebrates and amphibians. Alsoincluded are the lessons learned for sampling theseassemblages.

Amphibian and macroinvertebrate taxa were se-lected as potential indicators of wetland condition.On the recommendation of many field profession-als, experienced in amphibian monitoring, funneltraps were used. The funnel traps also proved tobe extremely effective at sampling wetlandmacroinvertebrate communities. Therefore, thesame sampling protocols are used for both amphib-ians and macroinvertebrates in Ohio EPA’s study.

Funnel trapThe funnel traps used in this project have cylin-

ders constructed of aluminum window screen andfunnel ends made from fiberglass window screen.The funnel traps are similar in design to commer-cially available minnow traps. The cylinders are 18inches long and 8 inches in diameter. The two fun-nel ends are attached to the cylinders and begin 8inches in diameter and taper inward 5 inches to a1¾-inch opening. A string handle that runs fromend to end is attached to the two seams where thecylinder and funnels ends join.

Plate 7: Impaired wetland.Photo: Mick Micacchion

Plate 8: Least-disturbed wetland.Photo: Mick Micacchion

67

To date, Ohio sampling data have come from de-pressional wetland systems in the Eastern Corn BeltPlains (ECBP) ecoregion. The Eastern Corn BeltPlains ecoregion spans most of the western half ofOhio and accounts for about 40% of Ohio’s landmass. Within the ECBP, depressional study wet-lands that demonstrated a gradient of disturbancelevels, from least impacted to greatly disturbed, wereselected.

Selected wetlands are sampled for amphibians andmacroinvertebrates three times during the year(early, middle, and late spring). An early springsample is collected in the period from late Februaryto late March. This early spring sample run allowssampling of adult Ambystomatid salamanders andearly spring macroinvertebrates (e.g., fairy shrimp)that are only present in wetlands for a limited time.Adult salamanders enter wetlands to breed in earlyspring following the first few warm, rainy nights oflate winter and early spring. The actual timing oftheir arrival in Ohio is highly weather dependent andvaries greatly by year and location. A secondsample is collected during the month of April in or-der to collect adult frogs entering wetlands to breed,as well as amphibian larvae already present, and tosample for macroinvertebrates. Mosquito larvae,an important prey item for many predators, areabundant in April in Ohio wetlands. The final sam-pling is conducted between May 15 and June 15.Salamander larvae and frog tadpoles are collected,in addition to the resident macroinvertebrates. Thislast sampling run occurs late enough in the breedingcycle to allow collection of larvae from all breedingamphibians. However, it is still early enough thateven in drought years temporary wetlands will nothave dried up.

Generally, 10 funnel traps are installed at eachwetland. Prior to installing the first funnel trap, theperimeter of the area where standing water ispresent is measured using a hip chain or by pacing.The total perimeter length is then divided by 10,and 10 funnel traps are installed uniformly aroundthe perimeter of the wetland at intervals of 10% of

the total perimeter distance. Each funnel trap loca-tion is permanently marked with flagging for usethroughout the sampling season. The funnel trapsare installed on the bottom at a location deep enoughto submerge the trap. The traps are left in the wet-land 24 hours to ensure unbiased sampling for ani-mals with diurnal and nocturnal activity patterns. Thedesign of the traps allows them to collect any am-phibians and macroinvertebrates that swim or crawlinto the funnel openings.

Upon retrieval, the traps are emptied by invertingone of the funnels and dumping and shaking thecontents into a white sorting pan. Organisms thatcan be readily identified in the field are counted andrecorded in the field notebook and released. Theremaining organisms are transferred to a 1-L plas-tic bottle and preserved with 70% ethanol. Thecontents of each trap are kept in separate, clearlymarked bottles for individual analysis in the labora-tory. If large numbers of amphibians are kept foridentification in the lab, the samples are transferredto formalin for long-term storage. All organismscollected are identified in the lab using appropriatekeys and the results are recorded.

DipnetQualitative collections are made concurrently with

funnel trapping at each wetland once during each ofthe three sampling periods. Qualitative samplinginvolves the collection of amphibians andmacroinvertebrates from all available natural wet-land habitat features. This is achieved by using tri-angular ring frame 30-mesh dipnets and manualpicking of substrates with field forceps. The goal isto compile a comprehensive species/taxa list ofmacroinvertebrates and amphibians at the site. Aminimum of 30 minutes is spent collecting the quali-tative sample. Sampling continues until the field crewdetermines that further sampling effort will not pro-duce new taxa. At least one specimen of all taxacollected during the qualitative sampling is preservedin a jar of ethanol for positive identification in thelaboratory.

68

Hester-Dendy artificial substrate samplerFive Hester-Dendy (HD) artificial substrate sam-

plers were tied to the top of a concrete block andplaced in wetlands where they remained submergedfor 6 weeks. The samplers were collected and pre-served in formalin. All the macroinvertebrates colo-nizing the samplers were counted and identified.

Analytical Methods:Macroinvertebrates and

Amphibians

Macroinvertebrates were identified to genus orspecies. Amphibians were identified to species ex-cept for some small salamander larvae identified togenus. Each funnel trap collection was analyzedindividually so that location-specific information wasnot lost by pooling all samples from a site.

The number of individuals collected in the trapswas divided by the number of hours trapped to givea relative abundance consisting of number of indi-viduals per trap hour. Results from the differentwetland study sites were examined for faunal dif-ferences in distribution and abundance. Analysis ofthe data for potential biological indicators of humandisturbance is under way.

Lessons Learned forMacroinvertebrates and

Amphibians

Funnel traps consistently collected an averageof 10 more macroinvertebrate taxa than quali-tative sampling using dipnets. Funnel traps weremuch more effective in sampling amphibians andfish than sampling with dipnets.

Qualitative sampling collected somewhat moreMollusca and Chironomidae taxa than did fun-nel traps.

Funnel traps collected more leech taxa, Hemi-ptera taxa, Coleoptera taxa, Odonata taxa, andCrustacea taxa than did qualitative sampling.

Hester-Dendy artificial substrate samplers wereineffective for sampling most wetlandmacroinvertebrates, except oligochaetes,Chironomidae, and Mollusca.

A 24-hour sampling period for funnel traps ispreferred as it allows for the collection of noc-turnal species that are infrequently collected bydaytime sampling methods.

A single wetland often has several vegetationclasses. Even if only three main classes are identi-fied (forested, scrub-shrub, and emergent), thewetlands included for study can exhibit multiple com-binations. For example, of the 20 wetlands studiedin Fennessy et al. (1998a), 6 combinations of veg-etation classes were found: emergent, emergent/scrub-shrub, forested, forested/emergent, forested/emergent/scrub-shrub, and forested/scrub-shrub.Thus, a sampling method should be flexible enoughto account for horizontal and vertical variation invegetation.

After testing a transect-quadrat method, Ohio EPAhas adapted the method used by the North Caro-lina Vegetation Survey as its standard vegetationsampling method (Peet et al. 1998). This is a flex-ible, multipurpose sampling method that can be usedto sample such diverse communities as grass- andforb-dominated savannahs, dense shrub thickets,forest, and sparsely vegetated rock outcrops, andhas been used at more than 3,000 sites over 10years as part of the North Carolina Vegetation Sur-vey. This method is appropriate for most types ofvegetation, flexible in intensity and time commitment,compatible with other data types from other meth-ods, and provides information on species compo-sition across spatial scales. It also addresses theproblem that processes affecting vegetation com-position differ as spatial scales increase or decreaseand that vegetation typically exhibits strongautocorrelation (Peet et al. 1998). Peet et al. (1998)state, “Our solution to the problems of scale andspatial auto-correlation is to adopt a modular ap-proach to plot layout, wherein all measurements are

69

made in plots comprised of one or more 10x10-mquadrats or “modules” (100 m2 = 1 are = 0.01 hect-are). The module size and shape were chosen toprovide a convenient building block for larger pots,and because a body of data already exists for plotsof some multiple of this size. The square shape isefficient to lay out, ensures the observation is typi-cal for species interactions at that scale of observa-tion, and avoids biases built into methods with dis-tributed quadrats or high perimeter-to-area ratios”(Peet et al. 1998, p. 264). Basically, the methodemploys a set of 10 modules in a 20H50-m layout.Within the site to be surveyed, these 20H50-m gridsare located such that the long axis of the plot isoriented to minimize the environmental heterogene-ity within the plot.

Once the plot is laid out, all species within the plotare identified, an aggregate wood stem count is made,and cover is estimated at the 0.1-hectare scale. Inaddition, four 10H10-m modules are intensivelysampled in a series of nested quadrats. Within these“intensive” modules, species cover class values andwoody stem tallies are recorded for each moduleseparately and for each nested quadrat separately.In effect then, the method proposed by Peet et al.incorporates use of reléves found in the Braun-Blanquet methodology inasmuch as the length,width, orientation, and location of the modules arequalitatively selected by the investigator on the ba-sis of site characteristics; however, within the mod-ules, standard quantitative floristic and forestry in-formation is recorded, such as density, basal area,cover, etc.

Once the location of the plot or plots has beenselected, the primary purpose of the vegetation sur-vey is to obtain a comprehensive list of all vascularplant species growing at a particular wetland at thetime of sampling and to characterize the relativedominance of these species at several levels of scale(basically herbaceous, shrub, small tree, and largetree scales, or at 1 m2, 10 m2, 100 m2, and 0.1 ha(1,000 m2 or 1 are).

All vascular species within the modules are iden-tified to species. Immature plants or plants missingstructures (e.g., fruiting bodies, etc.) that cannot beidentified to species are identified to genus or fam-ily or noted as unknown. Within the intensivelysampled modules, percent cover is recorded foreach species within modules and nested quadrats.Cover classes suggested by Peet et al. (1998) areused as a faster and more repeatable method forassigning cover values: Class 1 (solitary/few), Class2 (0 to 1%), Class 3 (1% to 2%), Class 4 (2%-5%), Class 5 (5%-10%), Class 6 (10%-25%),Class 7 (25%-50%), Class 8 (50%-75%), Class 9(75%-95%), Class 10 (95%-99%). The midpointsof the cover classes are used to calculate percentcover, relative cover, etc.

Woody stem data (trees, shrubs, and woody li-anas reaching breast height or 1.4 m) are collectedas counts of individuals in diameter classes. Peetet al. (1998) suggest the following diameter classes(in cm): 0-1, 1-2.5, 2.5-5, 5-10, 10-15, 15-20,20-25, 25-30, 30-35, and 35-40, with stemsgreater than 40 cm counted individually and mea-sured to the nearest centimeter. Multiple stems aris-ing from a common root system are recorded sepa-rately if they branch below 0.5 m from the ground.Peet et al. (1998) recommend that the area sur-veyed by stem count be adjusted based on condi-tions at the site, e.g., reduced to 20% of the mod-ules for dense shrubland or increased by 200% foroak savannahs. This is easily implemented by re-ducing the width of the modules for woody speciesonly.

An important part of vegetation surveys is collec-tion, preparation, and depositing of voucher speci-mens in major herbariums in order to document apermanent record of that plant at that location. Al-though staff resources make collecting vouchers ofevery vascular plant infeasible, a voucher specimenof at least 10% of the vascular plant species at anygiven site is collected; however, in every instance inwhich the identity of any species cannot be con-firmed in the field, or where field personnel disagree

70

as to the identity of a species, a voucher specimenis collected for identification in the office. In par-ticular, difficult genuses and families (e.g.,Cyperaceae, Poaceae, Ranunculaceae, Viola, As-ter, Potamogeton), as well as endangered, threat-ened, rare, or otherwise unusual plants, are almostalways collected for confirmation.

Finally, data on standing biomass for emergentwetlands are collected. These data can be used inseveral ways. Biomass production in emergentwetlands dominated by herbaceous vegetation isestimated by harvesting 900-cm2 quadrats in eachwetland. The quadrats are located within the inten-sive modules of each plot. The plants within eachquadrat are cut at the soil surface and placed intopaper bags. In the lab, plants are oven dried at105oC for at least 24 hours and then weighed.

Lessons Learned forVascular Plants

Floristic Quality Assessment IndexesOhio EPA has found that the FQAI score and

subscores of the FQAI, e.g., percent coverage ofplants with coefficients of conservatism of 0, 1,or 2, are very successful attributes and metricsfor detecting disturbance in wetlands. See the fol-lowing references:

Andreas B, Lichvar R. 1995. A Floristic QualityAssessment System for Northern Ohio. WetlandsResearch Program Technical Report WRP-DE-8.U.S. Army Corps of Engineers, Waterways Experi-ment Station.

Herman KD, Masters LA, Penskar MR, ReznicekAA, Wilhelm GS, Brodowicz WW. 1996. Floris-tic Quality Assessment With Wetland Categoriesand Computer Application Programs for the Stateof Michigan. Michigan Department of Natural Re-sources, Wildlife Division, Natural Heritage Pro-gram.

Wilhelm GS, Ladd D. 1988. Natural area as-sessment in the Chicago region. Transactions 53rdNorth American Wildlife and Natural ResourcesConference, pp. 361- 375.

Wilhelm GS, Masters LA. 1995. Floristic Qual-ity Assessment in the Chicago Region and Applica-tion Computer Programs. Lisle, IL: Morton Arbo-retum.

Semiquantitative disturbance/integrity scalesOhio EPA has had good success in developing a

semiquantitative disturbance/biological integrity scalecalled the Ohio Rapid Assessment Method forWetlands v. 5.0. Until more quantitative variablessuch as percent impervious surface are found, thistype of tool is a good candidate for the problematicx-axis in wetland biocriteria development. See also“Plants and Aquatic Invertebrates as Indicators ofWetland Biological Integrity in Waquoit Bay Wa-tershed, Cape Cod,” Carlisle BK, Hicks AL, SmithJP, Garcia SR, Largay BG. Environ Cape Cod1999; 2(2):30-60, where a similar system was usedto rank levels of disturbance.

ClassificationClassification is definitely an iterative process.

Investigators should consider a hydrogeomorphic(HGM) classification scheme if one has been de-veloped for their region of interest, at least as a start-ing point. However, the experience in Ohio sug-gests that grosser classes based on dominant veg-etation (emergent, scrub-shrub, forested, etc.) maywork also. A goal of a cost-effective biocriteriaprogram is to have the fewest classes that providethe most cost-effective feedback. With vegetation,data from Ohio are suggesting that somewhat di-verse wetland types may be “clumpable,” becauseeven though their flora are different at the specieslevel, the quality/responsiveness of their unique florato human disturbance is equivalent. This is also aconcern in States with high degrees of wetland losswhere too few wetlands of a particular HGM classremain to analyze as a separate class.

71

Field and lab methodsAfter experimenting with both transect/quadrat and

releve-style plot methods, Ohio has adopted a plot-based method that allows for a qualitative stratifi-cation of wetland by dominant vegetation commu-nities. This method is flexible and adaptable tounique site conditions, provides dominance data forall species in all strata, provides data that areintercomparable with other common methods, isrelatively easy to learn, and is relatively fast andcost-effective (up to two to three plots can be com-pleted in a day).

Whatever sampling method is adopted, it is es-sential that dominance and density information(cover, basal area of trees, stems per unit area, rela-tive cover, relative density, importance values, etc.)be collected. Many of the most successful attributesOhio has found in developing a vegetation IBI arebased on cover data of the herb and shrub layersand density data of the shrub and tree layers.

Definitely consider using cover classes in generaland a class scheme that works on a doubling prin-ciple to aid in consistent inter-investigator usage (seePeet et al. 1998). Then use the midpoints of theclass for your analysis. This seems to help withconsistent usage and smoothing out the roughnessin cover data.

Finally, it is recommended that initially thesampling method should “overstratify” in both thevertical and horizontal dimensions until it can be de-termined which strata and communities are respond-ing best to human disturbance. Ohio has found thatthe herb and shrub (subcanopy layers) seem torespond the best, although some intermediate treesize classes (e.g. 10- to 25-cm dbh) also appearresponsive.

Overstratifying horizontally may also make senseat the reference development stage; however, ulti-mately the decision whether to split or clump com-munities depends on whether this is necessary todetect the disturbances. “Homogenizing” commu-nity types by placing a releve or transect acrossthem (e.g., aquatic bed to emergent to shrub zone)can be appropriate if splitting does not matter todetect the disturbance. The caveat, of course, isthat you cannot separate the data set later if youdetect something of interest in one of the clumpedcommunities.

Vouchers and QAQCOn the basis of Ohio’s experience, voucher as

much as you can for later confirmation in the laband deposit vouchers in local and regionalherbariums. Definitely collect all Cyperaceae,Poaceae, and Juncaceae, and also consider col-lecting shrubs genuses and families (Salix, Vibur-num, Vaccinium, Rosa, Alder, etc.), Polygonumspp., Aster spp., Viola spp., and Cryptograms.

72

Oregon: Simultaneous Development,Calibration, and Testing of Hydrogeomorphic-

Based (HGM) Assessment Procedures andBiological Assessment Procedures

Purpose

Collect and analyze field data on wetland andriparian plant communities and hydrogeo-morphic (HGM) features to define referencestandard conditions and performance criteriafor four subclasses of wetlands in Oregon, twoin the Willamette Valley in western Oregon andtwo on the Oregon coast

Use that information to develop quantitative butrapid, HGM-based assessment models andprocedures for those subclasses, as well as toidentify new wetland plant community metricsthat have demonstrated relationships with land-scape condition and with internal alterations ofwetlands in this region

Compare results of using HGM models for func-tions with metrics that describe floristic charac-teristics, to verify the results on independent datasets, and demonstrate implementation of themethods for assessing progress of restoredwetlands and for helping recommend compen-sation ratios in a wetland mitigation bank con-text

Contact Dana Field

Organization Oregon Division of State Lands775 Summer St. NESalem, OR 97310

Phone (541) 378-3805, ext. 238


Contact Paul Adamus, Ph.D

Organization Adamus Resource Assessment,Inc.

6028 NW Burgundy DriveCorvallis, OR 97330

Phone (541) 745-7092


Wetland Type

Slope, flats, and estuarine fringe wetlands

Assemblages

Vascular plants

Status

Completed reports for the two Willamette Valleysubclasses, and currently initiating developmentof the estuarine fringe project. Information onpurchasing the reports can be found at:h t tp : / / s t a te lands .ds l . s t a te .o r.us /hgm_guidebook.htm.

Project Description

These projects are being done cooperatively withEPA Region 10 and the Oregon Division of StateLands, which in 1997 identified a need, in the con-text of Section 404 and 401 responsibilities, for amore quantitative assessment method tailored spe-cifically to these wetland subclasses. The proce-

73

dures have been designed to be applicable tononjurisdictional riparian sites as well as to wet-lands. The project began in 1998 with formation ofan interagency oversight committee, intensive re-view of regional literature, and reconnaissance andpreliminary selection of 124 reference sites through-out the Willamette Valley ecoregion. In 1999, thelist of sites was narrowed, workshops of local ex-perts were held to develop function models and fieldprocedures, 38+ volunteers were recruited andtrained to assist with field work, and 2 assessmentteams (“A-teams”) of scientists with help from thevolunteers collected data from the selected sites. Athree-volume final report was published in early2001 (see above). The report for the estuarinefringe wetlands is scheduled for completion in 2004.Under administrative rules recently reissued by theOregon Division of State Lands, applicants for mostState-authorized wetland removal-fill permits arenow required to use the regionalized HGM classifi-cation that was developed by this project, and areencouraged to use the reference-based functionassessment procedures when available for a par-ticular subclass and region of the State.

Study Design

For the Willamette Valley project, plant andhydrogeomorphic data were collected from 109wetland and riparian sites (54 riverine impoundingsubclass, 55 slope/flats subclass). The “riverineimpounding” sites include all wetland and riparianareas within a 2-year river floodplain, e.g., sloughs,oxbows, cut-off channels, beaver impoundments,stream-fed ponds with water control structures.“Slope/Flats” sites include most ash swales, wetprairies, springs, and foothill seeps. From visualinspection alone, most sites in the “slope” HGMclass could not be reliably separated from most sitesin the “flats” HGM class, due to the flat topographyand varied geology of the region, so the two werecombined into a “slope/flats” subclass.

Within the subclasses, assessment sites were se-lected nonrandomly, in a manner intended to span

gradients of human disturbance, size, and plant com-munity succession. Restored/created sites of vari-ous ages were also included. Nearly all sites wereon public lands, and ranged in size from 0.1 to 233acres. Most sites were nominated by local wetlandexperts. From the collected data, five least-dis-turbed riverine sites were selected as reference stan-dards, as were six least-disturbed slope/flat sites.Use of cluster analysis and other statistical analysismethods verified that sites belonging to the two mainsubclasses were significantly different based on land-scape position and mapped soil characteristics im-portant to ecosystem function.

For defining the disturbance gradient (the “x axis”),information specific to each candidate site was col-lected during the reconnaissance phase. This infor-mation pertained to past management practices,surrounding land use, and recent physical alterations.Physical alterations were noted visually during re-connaissance visits. At each site each type of alter-ation was categorized as (a) absent, (b) physicallyaffects <10% of site, (c) affects >10% of site butnot all, or (d) affects entire site. The categories thatwere used for alterations were:

Flow-impounding – berms, dikes, dams

Flow-impounding – excavations, pits

Flow-redirecting

Water subsidy (e.g., stormwater pipes)

Drainage-inducing (ditches, tile)

Soil compacting (e.g., fill, machinery, cows)

Soil mixing (e.g., plowing)

Soil grading (e.g., flattening)

Vegetation removal (e.g., extreme grazing, log-ging)


At each site, the A-teams identified plants in eachof potentially three hydrologic zones: permanentwater zone, seasonally inundated zone, saturated-

74

only zone. For woody plants, the team walked theentire site and made an overall estimate of relativepercent of the area of each zone occupied by eachshrub species (understory and open) and tree spe-cies. For herbaceous plants, the team assessedrelative cover of each species in 1x1-m quadrats.No more than nine quadrats were used at any site(and fewer at smaller sites and sites with fewer hy-drologic zones). Larger numbers of quadrats werenot used because of time and resource constraints.Within zones, quadrats were located so as to maxi-mize the cumulative number of species found. Thisless formal approach was used because our goalwas to develop metrics based mainly on commu-nity composition rather than quantitative measuresof abundance or cover. Two teams assessed all109 sites in about 50 workdays.


From the collected data, the following plant metricswere compiled for each site:

Number of native herb species, relative to theintensity of sampling (# of plots at each site)(based on analysis using species-area curvesand regression)

Percent of herb species that are native

Percent of the herb species that are “remnant-dependent” (i.e., reputedly most characteristicof unaltered sites)

Percent of the dominant herb species in any plotthat are natives

Percent of the dominant herb species in any plotthat are remnant-dependent

Percent of the true wetland species (facultativeor wetter) that are natives

Percent of shrub species that are natives (whenshrubs present)

Percent of native shrub species that are at leastmoderately dominant

Percent cover of non-native shrubs

Preliminary analysis suggests many of the abovemetrics had a statistically significant association withcategorical observations of partial physical degra-dation of sites (see Study Design, above) and/orwith amount and proximity of surrounding land covercategories (agriculture, urban, natural) that wereassessed visually during fieldwork as well as by aGIS analysis of existing digital imagery. Analysesof the HGM and plant data sets, each containing4,000+ records, were performed on a PC usingExcel, PC-ORD, and NCSS.

Lessons Learned

Random or systematic sampling, whether withina region or within a site, is not always appro-priate for use in identifying good biological in-dicators or developing rapid models for assess-ing wetland condition and function.

Systematic, repeatable, rapid procedures canbe developed for assessing some of the distur-bance gradients. This is a necessary precursorto selecting reference sites that will yield the mostuseful data.

The biological metrics investigated or usedshould be appropriate to the study design andmeasurement protocols.

Data suitable for identifying biological indica-tors of wetland condition sometimes can be col-lected simultaneously with data collected forcalibration of HGM models. This does not nec-essarily require a great deal of additional train-ing or field time.

Shared field experiences are a good forum forsharing wetland knowledge among agencies, andamong agencies and consultants and citizens.Shared field experiences lead to participants feel-ing more vested in the process of developing mod-els and multimetric indexes. This informal “buy-in”can lead to greater willingness of participants to usethe methods that ultimately result.

75

Pennsylvania: Assessing Wetland Condition ata Watershed Scale Using Hydrogeomorphic

Models and Measures of Biological Integrity

Purpose

Develop biological and functional assessmentmethodologies for many wetland types in Penn-sylvania

Wetland Type

Variety of wetland types

Assemblages

Amphibians

Birds

Macroinvertebrates

Vascular plants

Project Description

The Penn State Cooperative Wetlands Center(CWC) has numerous ongoing projects that involvebioassessment methodology for assessing the con-dition of wetlands. These projects include:

Bird community index of biological integrity forthe Mid-Atlantic Highlands

Verified suitability index for the Louisiana wa-terthrush

Amphibian indicator landscape study

Macroinvertebrates indicator pilot study

Otter and beaver interactions in the DelawareWater Gap

Using bioindicators to develop a calibrated in-dex of regional ecological integrity for forestedheadwater systems

In addition to these projects, CWC is conductinga 2-year pilot study to assess the ecological condi-tion of wetlands in the Juniata River watershed ofcentral Pennsylvania. The Juniata River is a majortributary of the Susquehanna River and lies withinthe headwaters region of the Chesapeake Bay eco-system. The objective of the Juniata Wetland Moni-toring Project is to define the ecological health ofwetland resources, thereby providing a scientificcontext for resource managers to plan future pro-tection and restoration activities.

Contact Denice Heller-Wardrop

Organization Penn State UniversityPenn State Cooperative Wetlands

Center301 Forest Resources

LaboratoryUniversity Park, PA 16802

Phone (814) 863-1005


Website http://www.wetlands.cas.psu.edu/

Contact Rob Brooks

Organization Penn State UniversityPenn State Cooperative Wetlands

Center301 Forest Resources

LaboratoryUniversity Park, PA 16802

Phone (814) 863-1596


Website http://www.wetlands.cas.psu.edu/

76

CWC is using HGM functional assessment mod-els that incorporate a floristic quality assessmentindex (FQAI) similar to the one used by Ohio EPA(see Ohio EPA’s case study). They use a three-step process. The first step is to develop and cali-brate HGM functional assessment models that aresensitive to environmental disturbance. The sec-ond step is to randomly sample wetlands within thewatershed that represent a range of HGM classesand disturbance levels. The final step is to apply theHGM models to the data collected in the field todetermine the condition or health of wetlands withinthe Juniata River watershed.

Preliminary HGM functional assessment modelshave been developed by the CWC. These modelsare currently being calibrated using a set of 102reference wetlands that span a variety of HGM sub-classes (mainstem floodplain, headwater floodplain,riparian depression, slope, etc.) and disturbancelevels (severe, moderate, or pristine). Disturbancelevels are assigned based on the surrounding land-scape, the type of buffer present, and potential stres-sors identified at the site. It is essential to charac-terize wetlands across a disturbance gradient todetermine not only the level of function a given wet-land type may achieve, but also the level of func-tioning that is attainable in an impacted landscape.

The second step of the process, data collection,is being undertaken by the CWC and trained in-terns from Penn State University, the University ofPittsburgh at Johnstown, and the Department ofEnvironmental Protection. Two types of data arecollected for each wetland sampled: landscape-level data and site-level data. Landscape-level dataare obtained at the CWC by characterizing landuses within a 1-km radius circle surrounding the

wetland using aerial photographs. This gives an in-dication of the type and magnitude of potential wet-land stressors. For example, activities such as landclearing in the surrounding watershed may stresswetland systems by increasing sediment loads.

In addition to land uses, the buffer type and widthis also determined for each site. It is important tolook at buffers when studying disturbance becausebuffers can ameliorate or magnify the affects of nega-tive landscape activities. A wetland surrounded bya forested buffer may not be as stressed by activi-ties in the landscape as one surrounded by an ur-ban or agricultural buffer.

Site-specific data are collected at each wetlandsite in the field. Site-level responses of wetlands tostressors are many and may involve both abioticand biotic indicators. At each wetland sampled, dataare gathered on microtopography, soils, plants, andanimals. Although not all plant species are highlysensitive to disturbance, the immobility of the plantcommunity, its amenity to remote sensing tech-niques, and easily recognized signs of stress makeit a good indicator of wetland condition. Previousstudies by the CWC have investigated the potentialutility of plant community measures as indicators ofwetland health. Plant community measures that mayprove to be good indicators of disturbance includethe FQAI score, the number or dominance of in-troduced or aggressive native species, and the num-ber of different species of Carex present. The CWCmethod contains a comprehensive plant samplingmethodology that has been used in a variety ofproject types. Three sizes of plots are used torecord various measures of the plant communityincluding percent cover and richness of herbaceousspecies, shrub volume, and tree dbh.

77

Vermont: Classification, BiologicalCharacterization, and Biometric Development

for Northern White Cedar Swampsand Vernal Pools

Contact Doug Burnham

Organization Vermont Department of Environmental ConservationWQD103 S. Main St. – 10NWaterbury, VT 05676

Phone (802) 241-3784 or 244-4520


Purpose

The Vermont Wetlands Bioassessment Project isa collaboration between the Vermont Departmentof Environmental Conservation and the VermontNongame and Natural Heritage Program, with a 3-year time frame. The primary objectives of the firstphase of the Vermont Wetlands Bioassessment Pro-gram are:

Gather chemical, physical, and biological datafrom seasonal pools that will facilitate an eco-logically based classification of minimally dis-turbed (reference) seasonal pools in Vermont

Use both previously and newly collected Non-game and Natural Heritage Program data totry to identify specific biological attributes thatcan serve as indicators of ecological integrity innorthern white cedar swamps.

Wetland Type

Vernal pools

White cedar swamps

Assemblages

Algae

Amphibians

Birds

Macroinvertebrates

Vascular plants

Status

Analyzing data and writing final report. A finalreport is expected in the near future.

Project Description

The Vermont Wetlands Bioassessment project cur-rently focuses on two types of wetlands: (1) vernal(seasonal/ephemeral) pools and (2) Atlantic north-ern white cedar swamps.

Vernal poolsWe sampled a total of 28 vernal pools for breed-

ing amphibians, macroinvertebrates, algae (diatoms),soils, vegetation, and water chemistry in April, May,and June of 1999 and 2000. A variety of otherphysical and riparian observations were also made.

78

Pools were distributed throughout Vermont acrossseven biophysical regions. A range of condition,from “reference” or least disturbed, to highly dis-turbed, was represented in the selection of pools.Upon completion of the field phase of the projectand assessment of relative hydroperiod, it is likelythat 5 of the 28 pools may be less “seasonal” thanoriginally thought. Data are being analyzed for bio-logical signals that may identify pools with more-or-less permanent standing water. Each pool wasvisited twice, the two visits approximately 4 to 6weeks apart, for aquatic macroinvertebrate andwater chemistry sampling. Algae samples were col-lected on the second pool visit. Amphibian surveyseither preceded or coincided with the firstmacroinvertebrate sampling visit. Five of the 28pools were dry on the second sampling visit.

Aquatic macroinvertebrates. Macroinverte-brates were sampled using three different methods:funnel traps to sample the actively swimming inver-tebrates (i.e., beetles, bugs, mosquitoes, crusta-ceans), a D-net to sample benthic invertebrates inthe leaf litter and muck (i.e., snails, bivalves, chi-ronomids, oligochaetes, caddisflies), and a qualita-tive search for any taxa we might have missed withthe previous two methods. Funnel traps were madeof window screen and designed to function like min-now traps. The traps were placed approximately10 m apart and were left in place for approximately24 hours. When the traps were emptied, any am-phibians were returned to the pool and themacroinvertebrates were collected and preserved.The contents of each trap were stored separately.The D-net scoop and qualitative samples were pre-served in the field, and later picked and sorted intotaxonomic orders according to standard protocol.Except for zooplankton, all taxa were identified tothe lowest practicable level. Scoop sampling alongtransects often created more of a disturbance in thepool than was comfortable for those conducting thesampling. The presence of egg masses and/or largeconcentrations of larval amphibians greatly restrictedthe efficiency of the scoop sampling at many pools.Trap sampling was less disruptive, but care had to

be taken to not completely submerge the traps whenadult amphibians were present so that entrappedadults would have air space to breathe. High con-centrations of mosquito larvae tended to affect theefficiency of the traps. Qualitative sampling wasless disruptive (although excessive wading in thepools often created an uncomfortable level of dis-turbance) but provided an inconsistent sampling ef-fort. The majority (70%) of macroinvertebrate taxacollected were represented in trap samples. Twoorders represented a large proportion of all taxacollected: Diptera (35%) and Coleoptera (23%).

Water. Water temperature, pH, and apparentcolor were recorded in the field. The remainder ofthe water sample was preserved and later analyzedfor alkalinity, conductivity, anions, cations, and alu-minum, and again for pH. Field pH values rangedfrom a minimum of 4.41 to a maximum of 7.75,alkalinity values ranged from 0 mg/L to 173.0 mg/L, conductivity readings ranged from 14.9 umhos/cm to 335 umhos/cm, and apparent color rangedfrom 4.5 to 489 over the sampling season. Appar-ent color, alkalinity, and cation concentrations showa general increasing trend within each site over thesampling season, whereas pH and aluminum andanion concentrations exhibited no consistent trend.

Algae. Algae were collected and analyzed fordiatoms from 18 pools sampled in 1999. Algae sam-pling primarily targeted diatoms; however, filamen-tous algae were collected when present. We at-tempted to collect both benthic samples (scrapingalgae from leaves, sticks, and rocks) and plank-tonic diatom samples from each pool. Unfortu-nately, five of the pools were already dry in May,so we were only able to collect planktonic samplesin the remaining 13 pools. Benthic samples werecollected from 17 of the 18 pools. We froze allsamples upon return to the lab, and samples weresent to Dr. Jan Stephenson for identification.

Amphibians. We observed amphibians initiallyduring the amphibian survey, and continued to ob-serve and collect specimens during both rounds of

79

macroinvertebrate sampling. At the beginning ofthe field season, we visually surveyed each pool foregg masses and spermatophores, identified each eggmass, recorded an approximate number of eggs permass, counted and identified breeding adults, anddescribed physical parameters of the habitat. Thetiming of this first visit did not necessarily coincidewith amphibian emergence; however, the funnel trapsused during invertebrate sampling effectively caughtactive amphibians. Breeding adults were typicallycaptured during the first round of sampling, whereastadpoles and larvae were present in the traps onthe second visit. All amphibians caught by the fun-nel traps, including adults, larvae, and egg masses,were identified in the field, counted, and returnedto the pool. We continued to survey egg massesand record physical parameters throughout the fieldseason. All 28 pools sampled showed signs of useby breeding amphibians. Amphibian species com-monly observed included wood frogs (27/28 sites),yellow-spotted salamanders (27/28 sites), Jefferson(hybrid) salamanders (7/28 sites), red-spotted newts(18/28 sites), and green frogs 15/28 sites).

Ongoing and upcoming work. Data are cur-rently being evaluated using a variety of proceduresin order to identify and describe efficient and leastdisruptive sampling methods, and evaluate biologi-cal, chemical, and physical indicators of natural vari-ability and disturbance. Methods of data analysisinclude two-way indicator species analysis(TWINSPAN) and detrended correspondenceanalysis (DCA) as well as standard descriptive andcomparative statistical methods. A detailed site re-port is being developed for each vernal pool site.

Northern white cedar swampsSeven northern white cedar swamps were sur-

veyed in June for breeding birds and vegetation.Two of these cedar swamps were considered to beof reference quality, whereas the other five had somedegree of impairment associated with them. Dis-turbances at the impaired sites included logging,roads, and storm water and agricultural discharge.We visited three of the sites early in summer to

assess the feasibility of sampling aquaticmacroinvertebrates for bioassessment and monitor-ing purposes.

Breeding birds/vegetation. Vegetation and bio-physical data were collected at each site and spe-cies lists are being constructed. Listening stationswere established and a bird census was taken twiceduring the breeding season. The data collected fromthe vegetative, biophysical, and bird sampling willbe compiled and compared to the existing data fromthe statewide inventory of northern white cedarswamps.

Aquatic macroinvertebrates. Sampling foraquatic macroinvertebrates had been scheduled forMay, but because of a busy vernal pool samplingseason, we were unable to visit the cedar swampsuntil June. Unfortunately, a very dry spring mayhave caused the cedar swamps to be drier than usualthis year. We visited two impaired sites and onereference site during the month of June. We wereunable to find any standing or flowing water at ei-ther the reference site or at one of the impaired sites,so we could not effectively sample. However, wefound evidence that suggested the swamps had con-tained water earlier in the season. The second im-paired site contained many braided, slow-flowingchannels and some standing water. We qualitativelysampled three of the channels and a small hollow atthe base of the boulder. The samples were pre-served in the field, picked, and sorted, and will beidentified according to standard protocol. It is pos-sible that aquatic microhabitats in cedar swampsare not available consistently enough to sample foraquatic macroinvertebrates; however, our limitedsampling efforts did not conclusively elucidate thefeasibility of using aquatic macroinvertebrates asbiological indicators in cedar swamps.

Ongoing and upcoming work. Using the datawe have collected, we will work to identify attributesassociated with vegetation and bird assemblagesthat can serve as indicators of ecological integrityand anthropogenic disturbance.

80

Washington: King County Wetland-BreedingAmphibian Monitoring Program

Contact Klaus O. Richter

Organization King County Department of Natural

Resources & ParksWater and Land Resources Division201 S Jackson Street, Suite 600Seattle, WA 98104-3855

Phone (206) 205-5622


Organization Viewpoint Services11040 104th Ave. NEKirkland, WA 98033-4423

Phone 425-822-4016

E-mail korichter @msn.com

Purpose

Provide yearly data regarding wetland conditionand the health of breeding amphibians for environ-mental review, to identify threats to wetlands andamphibians, to evaluate current wetland protectionmeasures, and to improve local-land use decisions.Specifically, to stem amphibian extinctions in urbanlandscapes by first identifying cause-and-effect re-lationships among amphibian distribution, abun-dance, and health and anthropogenic impacts towetlands and uplands, and second, using this infor-mation to provide amphibian conservation recom-mendations.

Wetland Type


Assemblages

Amphibians

Status

Implementing methods

Project Description

The King County Wetland-Breeding AmphibianMonitoring Program was originally designed to pro-vide the county with long-term wetland, amphibian,and landscape information for planning and regula-tory purposes. From 1993 to the present, a mini-mum of 150 volunteers were trained to census am-phibian eggs, juveniles, and adults in 90 freshwaterwetlands of 26 watersheds in King County. Analy-sis of amphibian distribution and health for all wet-lands was completed. A targeted subsample of 21wetlands in 3 rapidly urbanizing watersheds in whichamphibians were declining was additionally moni-tored for wetland hydrology, predators, and wa-tershed condition to determine the causes of popu-lation declines, in hope of stemming declines throughbetter wetland conservation practices.

Today, a skeleton program totally run by volun-teers continues to monitor a few of the initially sur-veyed wetlands without county support. A moreintensive monitoring program targeting amphibiansas bioindicators of wetland condition has been initi-ated with the cooperation and partnership of theUniversity of Washington’s Certificate Program in

81

Wetland Science and Management, in which sev-eral students research and monitor an importantwetland and amphibian issue.

The project’s monitoring goals include thefollowing:

Identify the occurrence of the State-endangeredOregon spotted frog

Determine land uses compatible with wetlandand amphibian conservation objectives

Provide data to help develop and implementregulations for the protection of amphibians andtheir habitats

Identify population distribution status of othercounty declining species

Obtain standardized baseline inventory data onthe distribution, abundance, and health of am-phibians in King County wetlands

Provide information to King County, Washing-ton State Department of Fish and Wildlife,Washington State Department of Ecology, and

Federal resource agencies for developing re-gional wetland and wildlife management pro-grams

Develop an effective public outreach and edu-cation program to train citizens to monitor am-phibians and wetland conditions to foster wet-land stewardship

Educate potential wetland scientists to the di-versity of wetland issues and train them in de-veloping monitoring programs, assessing wet-land condition, and crafting management plansfor better wetland conservation

Investigate local urban amphibian decline andextinction issues

Develop methods for utilizing amphibians asbioindicators of wetland condition by establish-ing cause-and-effect relationships betweenamphibian declines and wetland hydrology,water quality, predators and pathogens, andwatershed land use.

82

Wisconsin: Developing Biological Indexes forWisconsin’s Palustrine Wetlands

Contact Jennifer Hauxwell

Organization Wisconsin Department of Natural ResourcesBureau of Integrated Science Services1350 Femrite DriveMonona, WI 53716

Phone (608) 221-6338


Purpose

Develop a biological index for Wisconsin’spalustrine wetlands

Compare performance of one plant and twomacroinvertebrate multimetric indices

Develop biological integrity rating system forclassifying wetlands

Wetland Type

Palustrine wetlands

Assemblages

Macroinvertebrates

Vascular plants

Status

Findings representing preliminary investigations arepresented in Final Report to EPA Region V, Wet-land Grant #CD985491-01-0 prepared by Wis-consin DNR (January 2000).

Project Description

Field studies for this project were conducted dur-ing the spring and summer of 1998, with laboratoryanalysis and data synthesis completed the followingyear. Funding was provided by a grant from EPARegion 5. The findings formed the basis for a sec-ond EPA-funded grant to refine and further evalu-ate the preliminary indices and expand communi-ties covered to include amphibians, small mammals,diatoms, and zooplankton.

Study Design

We sampled 104 palustrine depression wetlandsdistributed across the major ecoregions of Wisconsinduring early spring and summer of 1998. Studysites included a mixture of least-disturbed referencebasins and degraded or restored wetlands, repre-senting a range in vegetative cover types, waterchemistries, and water duration.


We sampled macroinvertebrates early in the springto minimize influences of immigration-emigration.Three different field collection procedures were

83

evaluated and two laboratory approaches wereused. On all 104 wetlands, we collected two setsof three standard D-frame net sweeps of approxi-mately 1-m length each. Sampling stations (< 60cm deep) were established at equally spaced pointsaround the wetland perimeter that approximatelytrisected the basin (and ensured coverage of themajor plant communities present). The first set ofnet sweeps was concentrated (large coarse materi-als were rinsed, examined, and removed) into a 1-quart container; the second set of net sweeps wasnot field-concentrated, but rather was placed di-rectly into a 1-gallon plastic bag. Both sets werepreserved in ethanol and returned to the laboratoryfor processing. The A third set (used on a subset of17 wetlands) of D-frame net sweep samples wasplaced on a coarse wire screen over a collectionbasin for a period of approximately 10 minutes.Organisms falling (or crawling) through the screenand entering the collection container were collectedand preserved as above. This last set representeda “clean” sample that was much easier to sort andprocess than the standard samples.

Laboratory Methods:Macroinvertebrates

Macroinvertebrates were processed using a two-tiered approach. The first stage consisted of a fixed100-count (sensu Hilsenhoff Biotic Index proce-dures) using a grid-marked tray with 24 cells. Or-ganisms were picked and sorted at a coarse taxo-nomic level, usually order or family only. Followingcompletion of the 100-count sample, we processedthe balance of the sample in its entirety (except forsubsampling dominant taxa). The unconcentrated“bag” samples proved to be too large to process inan economical fashion, so only the complete set of104 “field-concentrated” samples were processed.The 17 “screened” samples generally contained lessthan 100 total organisms and were processed com-pletely.


We used SYSTAT to perform all statistical andgraphical analyses. Percentage data were trans-formed using the arc-sine square-root transforma-tion, and abundance data were either log-trans-formed or power-transformed as applicable. Met-ric development was based on a series of visualcomparisons of community attribute responses tosuspected measures of disturbance using box plotsand jittered dot density plots. Those attributes thatexhibited evidence of separation between referencewetland conditions and wetlands suspected to beimpacted by human disturbance were selected aspotential metrics. Attributes that exhibited incon-sistent or overlapping responses between impactedand reference systems were eliminated from furtherconsideration.


We conducted simplified plant surveys during July1998 using a combination of techniques. This in-cluded a subjective estimate of cover and an ob-jective survey of percent cover and frequency ofoccurrence within six equidistantly spaced 20 by50 cm rectangular quadrats positioned along eachof three transects that trisected the wetland basin(total of 18 quadrats per wetland).

Laboratory Methods: Plants

No biomass or stem counts were performed.Voucher specimens were pressed and identified tospecies when possible, but most plant metrics werebased on a coarser taxonomic level.

84


We developed the plant biological index using thesame procedures described for the twomacroinvertebrate-based indices. Because wewanted to develop a practical tool for managerswith limited botanical expertise, we lumped taxa atvarious taxonomic levels (e.g., family, genus) orstructural groups (e.g., grass-like, emergent) foranalysis. Importance values (average of percentcover and frequency of occurrence) were used asthe attribute of concern for family-genus-specieslevels, and percent cover was used for emergent,submergent, floating-leafed, and open-waterattributes.

Other Data Collected

We also collected associated physical and chemi-cal data on each wetland, including pH, alkalinity,conductivity, color, temperature, clarity, and depth.Riparian cover type within a 100-foot buffer areasurrounding each wetland was subjectively estimatedand recorded, as was shade canopy cover.

Preliminary Findings

Three multimetric indices (two macroinvertebrateand one plant index) were developed. The Wis-consin Wetland Macroinvertebrate Index (WWMI)is a multimetric index based on 15 metrics derivedfrom a total count of organisms in three compositednet sweeps. A total of 69 community attributes wereevaluated. The WWMI is composed of 12 abun-dance metrics, two richness metrics, and one per-centage metric. Abundance metrics include mol-lusks, annelids, fairy shrimp, damselflies, pigmybackswimmers, water boatmen, limnephildcaddisflies, total caddisflies, phantom midges, mos-quitoes, soldier flies, and total invertebrates. Rich-ness metrics are noninsects and total taxa. Thepercent caddisflies is the only percentage metric.Apparent redundancies (e.g., caddisflies) in themetric may or may not be an issue; differences in

taxonomic rate of development in wetlands due tothermal dynamics may require a certain amount ofredundancy to ensure that important taxonomicgroups are accounted for. The WWMI is used torate, rank, or compare wetland biological condi-tion.

The second macroinvertebrate index, termed the100-count macroinvertebrate biotic index (100-count MBI), is based on 10 metrics derived from arandom pick of 100 organisms found in the threecomposited net sweeps. The 100-count MBI iscomposed of 9 percentage metrics and 1 richnessmetric (noninsect taxa). Percentage metrics includepigmy backswimmers, water boatmen, total “bugs,”limnephilid caddisflies, total caddisflies, chironomids,soldier flies, and the sum of Ephemeroptera-Odonata-Trichoptera (EOT) taxa. Noninsects rep-resent the only richness metric in the 100-count MBI.The ninth percentage metric, mollusks, may be onlyuseful in prairie-type wetlands. The 100-count MBIis intended to be applied in the field by experiencedstaff as a means of rapid bioassessment.

The third index is the Wisconsin Wetland PlantBiotic Index (WWPBI) is based on eight (or nine)plant metrics derived from transect data (represent-ing 18 quadrats) and is intended to serve as a supple-mentary index to the WWMBI to rate, rank, andcompare wetland biological condition. Of 24 plantcommunity attributes tested, only one richness(count) metric, one percent metric, and seven im-portance value-based metrics demonstrated con-sistent and predictable response. The single rich-ness metric, total taxa, may require further modifi-cation after reaching some consistency regardingtaxonomic resolution (currently mixed family-genus-species). Importance value-based metrics includedCarex, reed canary grass, cattail, duckweed,bluejoint grass, and “good” plants (the sum of agroup of plants including all Carex, Utricularia,Potamogeton, Calamogrostis, Sagittaria,Polygonum, and Equisetum). An additional impor-tance value-based metric, “pondweeds,” would onlybe applied to wetlands with water duration exceed-

85

ing 7 months per year. The only percentage-basedmetric, floating-leafed plants, would likewise onlybe applied to wetlands with water durations ex-ceeding 7 months.

Lessons Learned

MacroinvertebratesWater duration is an important factor shapingmacroinvertebrate community composition andderived metrics that must be accounted for inmetric scoring.

A coarse level of taxonomic resolution (orderand family) appears to be satisfactory in devel-oping wetland macroinvertebrate metrics.

Issues relating to redundancy among metrics,influences of water chemistry, differences amongecoregions, and seasonal variations need to beaddressed in more detail. Undoubtedly, thesefactors need to be accounted for in establishingrating scores and/or in refining metrics for usein different areas or habitats.

Basic differences exist in macroinvertebratecommunities between wetlands representingwooded kettle depressions and open-prairietype depressions in Wisconsin.

Our greatest difficulty was in selecting and as-signing some measure of “human impact” to thestudy site wetlands. Further research will berequired to quantify the degree of human im-pact in order to refine biological responsemetrics and indices.

The WWMBI was not stable across dates (andnot designed to be); consequently, its use is re-stricted to early spring.

Each macroinvertebrate index has its own setof advantages and disadvantages; further re-finement is required to enable their successfulapplication in the field.

PlantsSome biases were apparent in the WWPBI, asreference kettles scored consistently higher thanprairie wetlands.

WWPBI scores in restored prairie wetlandswere better than in many natural wetlands, sug-gesting that wetland restorations in Wisconsinmay be adequate in terms of “restoring” thevegetative community (not true formacroinvertebrate response).

The WWPBI shows some promise in its per-formance and, because of its taxonomic sim-plicity, it could be applied by nonbotanists.

Followup: Continuing Work

With the assistance of a second EPA wetland grant,we have refined the macroinvertebrate and plantmultimetric indices and explored expanding the com-munity components to include zooplankton (in co-operation with Dr. Stanley Dodson, University ofWisconsin–Madison), diatoms (Paul Garrison, Wis-consin DNR), amphibians, and small mammals.Findings associated with the second grant are re-ported in the second Wisconsin case study below.

86

Wisconsin: Refinement and Expansion of theWisconsin Wetland Biological Index

for Assessment of DepressionalPalustrine Wetlands

Contact Jennifer Hauxwell

Organization Wisconsin Department of Natural ResourcesBureau of Integrated Science Services1350 Femrite DriveMonona, WI 53716

Phone (608)-221-6373


Purpose

Test and refine a Biotic Index for Wisconsin’spalustrine wetlands

Expand list of assemblages to includemacroinvertebrates, zooplankton, diatoms,amphibians, plants, and small mammals

Establish a biological integrity rating system forclassifying wetlands based on the response ofselected biological attributes (metrics) of theabove communities to surrogate measures ofhuman disturbance

Wetland Type

Palustrine wetlands

Assemblages

Macroinvertebrates

Aquatic plants

Zooplankton

Diatoms

Amphibians

Small mammals

Status

Findings presented in Final Report to EPA Re-gion V prepared by Wisconsin DNR dated April2002.

Project Description

This project represents the evaluation and expan-sion phase of an earlier study that resulted in thepreliminary development of a Wisconsin WetlandBiological Index based on plant andmacroinvertebrate metrics (please see final reportto EPA -Wetland Grant #CD985491-01-0). Datafrom this second study were used to refine and fur-ther evaluate the preliminary indices and expandcommunities covered to include zooplankton, dia-toms, amphibians, and small mammals. Field stud-ies for this project were conducted during the springand summer of 2000, with laboratory analysis anddata synthesis completed in 2001. Funding wasprovided by a grant from EPA Region 5.

We sampled 75 palustrine depression wetlands insoutheast Wisconsin during early spring and sum-mer of 2000. Study sites included a mixture of least-disturbed reference basins (17 prairie and 19

87

wooded kettles) and impacted wetlands (18 urbanand 20 agriculture), representing a range in vegeta-tive cover types and water chemistries. A severedrought, which began the previous year, caused mostof the smaller wetlands with short hydroperiods(seasonal and temporary) to dry out, and conse-quently only long-duration wetlands (semiperma-nent and permanent) were sampled.

We sampled macroinvertebrates early in the springto minimize influences of immigration-emigration.On each wetland, we collected three standard D-frame net sweeps of approximately 1-m length each.Sampling stations (less than 60 cm deep) were es-tablished at equally spaced points around the wet-land perimeter that approximately trisected the ba-sin and assured coverage of the major plant com-munities present. The contents of the three netsweeps were concentrated (large coarse materialswere rinsed, examined, and removed) into a 1-quartcontainer, preserved in ethanol, and returned to thelaboratory for processing. Seven field replicateswere collected.

Laboratory Methods:Macroinvertebrates

Each macroinvertebrate sample was processedentirely, using a two-phase approach withsubsampling employed only for extremely abundantorganisms. The first stage involved using a grid-marked tray with 24 cells. Cells were selected ran-domly, and organisms were picked and sorted at acoarse taxonomic level, usually to order or familylevel. The abundance of organisms present in thefirst two or three randomly selected cells was usedto project the abundance of the most common taxain the sample. Taxa whose abundance was esti-mated to exceed 300 in the total sample were thenoverlooked while the balance of the sample wasprocessed (the second phase of processing). Allspecimens were vouchered (preserved in 70% etha-nol) for possible further evaluation.


We conducted simplified plant surveys during July2000 using a combination of techniques. This in-cluded a subjective estimate of cover and an ob-jective survey of percent cover and frequency ofoccurrence within six equidistantly spaced 20x50-cm rectangular quadrats positioned along each ofthree transects that trisected the wetland basin (to-tal of 18 quadrats per wetland).

Laboratory Methods: Plants

No biomass or stem counts were performed.Voucher specimens were pressed and identified tospecies when possible, but most plant metrics arebased on a coarser taxonomic level.

Sampling Methods: Zooplankton

Note: zooplankton studies were conducted by Dr.Stanley Dodson, University of Wisconsin-Madison,Zoology Department, Madison, Wisconsin. E-mail:[email protected].

We collected one zooplankton sample from a cen-tral basin location in each wetland during June 2000using a 5-L plastic bucket. We filtered a knownvolume of water through a No. 10 (60-micron mesh)net to capture zooplankton within. Seven field rep-licates were collected. Samples were preserved in70% ethanol until processed.

Laboratory Methods:Zooplankton

Each sample was scanned at moderate magnifi-cation for species of cladocera, copepods, ostra-cods, and aquatic insects. Slides and dissectionswere made where necessary; for example, to aid inthe identification of copepod species. Total num-ber of male and female Daphnia was counted in

88

each sample. More than 200 slides were prepared,and identification of organisms is in progress. Or-ganisms will be identified to species where possible.

Sampling Methods: Diatoms

Diatom studies were conducted under the direc-tion of Paul Garrison, WDNR, Bureau of IntegratedScience Services, Environmental ContaminantsSection, 1350 Femrite Drive, Monona, WI 53716.E-mail: [email protected]

Using a 1-dram vial as a sample collection de-vice, we collected and composited surficial (upper0-1 cm) sediments from five sites in each wetland.Seven field replicates were collected. Samples werekept on ice or refrigerated until processed.

Laboratory Methods: Diatoms

Each diatom sample will be thoroughly mixed, anda small amount will be placed into a tall beaker.Hydrogen peroxide will be added and the samplewill be allowed to steep for about 5 minutes. Po-tassium dichromate will be added (under a venti-lated hood and handled with safety gloves) to fa-cilitate reduction of organic matter. The sample willbe washed at least four times with deionized waterby centrifuging for 10 minutes. Two portions of thecleaned sample will be dried on separate No. 1cover-slips and mounted with Naphraxâ, and la-beled accordingly. Specimens from both cover slipswill be identified and counted under oil immersionobjective (1,400X) until a total of 250 frustules arecounted. Identification of difficult taxa will be madeusing a scanning electron microscope. Undeter-mined specimens representing a significant portionof a sample will be sent to Dr. Rex Lowe at Bowl-ing Green University, Dr. Jan Stevenson at Louis-ville University, and/or Dr. Gene Stormer at theUniversity of Michigan.

Sampling Methods: Amphibians

Because amphibians are extremely sensitive toweather and temperature, we assessed amphibiancommunities during two separate sampling periods.We conducted standardized frog-toad calling sur-veys (using WDNR protocols) during the first twophenologies (early spring and late spring) betweenthe hours of 8 and 10:30 pm for 10-15 minuteswhen water temperatures were above 50°F duringthe first phenology or above 60°F during the sec-ond phenology. We recorded all calling to permitverification of questionable identifications. In addi-tion to calling, we added to the database any per-sonal observations of amphibians made during anyof the daylight visits and any specimens capturedduring the macroinvertebrate surveys.

Laboratory Methods: Amphibians

No laboratory methods were used.

Sampling Methods:Small Mammals

Small mammal studies were conducted by R.Bautz, WDNR, Bureau of Integrated Science Ser-vices, Ecological Inventory & Monitoring Section,1350 Femrite Drive, Monona, WI 53716. E-mail:[email protected].

We assessed small mammal communities by trap-ping during August-September 2000. On eachwetland, we set 46 baited traps (mixture of 40 mu-seum special grade and 6 larger tomahawk traps)along transects (zigzag scattered routes) in the ri-parian zone (variable dimensions, depending onsetting) for one night. A one-day/night trappingperiod was used to minimize disturbance by rac-coons and other predators. Bait consisted of a mix-ture of peanut butter and rolled oats. Traps werecleaned and rebaited each morning. Specimenswere placed in labeled freezer bags and returned tothe laboratory for identification.

89

Laboratory Methods:Small Mammals

Identifications of rare taxa and species of specialconcern were verified by local experts from theUniversity of Wisconsin.

Other Data Collected

We also collected associated physical and chemi-cal data on each wetland. The water chemistrymeasures included pH, alkalinity, conductivity, color,chloride, calcium, silica, nitrate-nitrite, organic-ni-trogen, and total phosphorus. Chemical analyseswere performed at the Wisconsin State Laboratoryof Hygiene following standard EPA-approved pro-cedures. Physical data collected included waterdepth, apparent color, temperature, and size. Ri-parian cover type within a 100-foot buffer area sur-rounding each wetland was subjectively estimatedand recorded, as well as shade canopy cover.

Analytical Methods: DataAnalysis and Development of

Biological Indices

Data were entered into Excel spreadsheet andAccess databases and examined for entry errors(univariate checks) and anomalies (e.g., bivariateplots) prior to analysis. Various community or spe-cies attributes (e.g., taxa or species richness, diver-sity, presence or absence of selected functional feed-ing guilds, trophic structure, percentages) wereevaluated and scored as potential metrics based ontheir responsiveness to measures of human distur-bance. Community attributes were examined usinga combination of procedures (outlined below) toselect promising metrics for index development.Attributes that exhibited strong positive or negativecorrelation with selected human response variableswere selected as metrics. We compared the sensi-tivity and correspondence among the selected com-munity metrics, and developed a single multimetricindex of ecosystem integrity that related to overallwetland condition.

Because different forms of human disturbanceelicit different responses among the various wet-land biological communities, it is difficult if not im-possible to choose one measure that represents “the”single best measure of human disturbance. For ex-ample, the amphibian community may respond moredirectly to woodland impacts (distance to nearestwoodland, patch size, corridor dimensions, etc.)than to nutrient or pesticide inputs, whereas zoop-lankton and diatom communities may be more re-sponsive to percent row crops in the watershed.Consequently, we evaluated biotic responses to acombination of a priori classes of human disturbanceand a composite chemical index that incorporatednitrogen, phosphorus, and chloride concentrations.The a priori classes represented agricultural andurban impacts as well as subclasses of each, in-cluding three intensity levels of agricultural impactand two forms of urban impact. Biotic responsesoccurring within these classes were measuredagainst the responses in least disturbed kettles andprairie depression wetlands. Statistical proceduresand approaches used in the analysis varied amongthe communities assessed because of inherent dif-ferences in responses and the type of data collected.Canonical correspondence analysis was used toexplore unimodal distributions of the diatoms alongthe various environmental gradients. We usedSYSTAT and a combination of other available sta-tistical software programs to perform statistical andgraphical analyses. Metric development was basedon a series of visual comparisons of community at-tribute responses to measures of disturbance (com-bination of selected water chemistry and land usecharacteristics as discussed above) using box plotsand jittered dot density plots. Attributes that ex-hibited evidence of separation between referenceand impacted wetlands were selected as metrics(or incorporated into the existing list of metrics).Attributes that exhibited inconsistent or overlappingresponses between impacted and reference systemswere eliminated from further consideration.

90

Lessons Learned

Finding willing and cooperative private prop-erty landowners for access to potentially im-pacted wetlands is difficult!

Determining what attribute or attributes (e.g.,land use, chemical contaminant concentration,distance to nearest road) to use as surrogatemeasures of human disturbance is critical to thesuccessful development of reliable multimetricbiological indices.

Drought (or floods) can seriously hamper sam-pling plans and interfere with the best of plans!

The list of suitable metrics among macroinver-tebrate communities of permanent wetlands isdifferent and much shorter than the list of metricssuitable for wetlands with shorter water dura-tion periods.

91

References

Apfelbeck RS. 2001. Development of biocriteriafor wetlands in Montana. In: Batzer D, Rader RB,Wissinger SA, eds. Bioassessment and Manage-ment of North American Freshwater Wetlands. pp.141-166.

APHA 1989 (North Dakota)

Borth C. 1997. Development of vegetation indicesfor water quality and hydroperiod in depressionalwetlands in western Montana. Master’s thesis pro-posal submitted to Montana DEQ. Montana StateUniversity, Bozeman, MT.

Burton TM, Uzarski DG, Gathman JP, Genet JA,Keas BE, Stricker CA. 1999. Development of apreliminary invertebrate index of biotic integrity forLake Huron wetlands. Wetlands 19(4):869-882.

Charles D, Acker F, Roberts NA. 1996. DiatomPeriphyton in Montana Lakes and Wetlands: Ecol-ogy and Potential as Bioassessment Indicators. Areport submitted to the Montana DEQ. PatrickCenter of Environmental Research, EnvironmentalResearch Division, Academy of Natural Sciences,Philadelphia, PA.

Fennessy et al. 1998a, b (Ohio)

Galatowitsch SM, Whited DC, Lehtinen RM,Husveth J, Schik K. 1999. The vegetation of wetmeadows in relation to their land use. EnvironMonitor Assess 60:121-144.

Galatowitsch SM, Whited DC, Tester JR. 1999.Development of community metrics to evaluate re-covery in Minnesota wetlands. J Aquat EcosystemStress Recovery 6:213-234.

Lehtinen RM, Galatowitsch SM, Tester JR. 1999.Consequences of habitat loss and fragmentationfor wetland amphibian assemblages. Wetlands19:1-12.

Ludden VE. 2000. The effects of natural variabil-ity on the use of macroinvertebrates as bioindicatorsof disturbance in intermontane depressional wet-lands in Northwestern Montana, USA. MS The-sis, University of Montana, Missoula MT, p. 163.

Mack et al. unpublished data (Ohio)

Mayer P, Galatowitsch S. 1999. Diatom communi-ties as ecological indicators of recovery in restoredprairie wetlands. Wetlands 19:765-774.

Mensing DM, Galatowitsch SM, Tester JR. 1998.Anthropogenic effects on the biodiversity of ripar-ian wetlands of a northern temperate landscape. JEnviron Manage 53:349-377.

Ohio EPA. 1988a, b, 1989

Peet et al. 1998 (Ohio)

Ralph CJ, Sauer JR, Droege S (tech. eds.). 1995.Monitoring Bird Populations by Point Counts.USDA Forest Service Pacific Southwest ResearchStation, Albany, CA, General Technical Report psw-GTR-149.

Riffell SK, Gutzwiller KJ, Anderson SH. 1996.Does repeated human intrusion cause cumulativedecline in avian richness and abundance? Ecol Appl6:492-505.

Riffell SK, Keas BE, Burton TM. 2001. Area andhabitat relationships of birds in Great Lakes coastalwet meadows. Wetlands 21(4):492-507.

92

Rotenberry JT, Wiens JA. 1980. Habitat struc-ture, patchiness, and avian communities in NorthAmerican steppe vegetation: a multivariate analy-sis. Ecology 61:1228-1250.

Tikkanen. 1986. (North Dakota)

Tobe JD, Burks KC, Cantrell RW, Garland MA,Sweeley ME, Hall DW, Wallace P, Anglin G, NelsoG, Cooper JR, Bickner D, Gilbert K, Aymond N,Greenwood K, Raymond N. 1998. Florida Wet-land Plants: An Identification Manual. Florida De-partment of Environmental Protection. Tallahas-see, FL: Rose Printing Company.

U.S. EPA. 1995 (Ohio)

Utermohl 1958 (North Dakota)

Wetzel and Likens 1991. (North Dakota)

Whited DC, Galatowitsch SM, Tester JR, SchikK, Lehtinen R, Husveth J. 2000. Importance andinfluence of landscape characteristics on patternsof biodiversity in depressional wetlands. Land UseUrb Planning 49:49-65.

93

Glossary

Aquatic life use A type of designated use pertain-ing to the support and maintenance of healthy bio-logical communities.

Assemblage An association of interacting popu-lations of organisms that belong to the same majortaxonomic groups. Examples of assemblages usedfor bioassessments include: algae, amphibians, birds,fish, amphibians, macroinvertebrates (insects, cray-fish, clams, snails, etc.), and vascular plants.

Attribute A measurable component of a biologi-cal system. In the context of bioassessments, at-tributes include the ecological processes or char-acteristics of an individual or assemblage of speciesthat are expected, but not empirically shown, torespond to a gradient of human disturbance.

Benthos The bottom fauna of waterbodies.

Biological assessment (bioassessment) Usingbiomonitoring data of samples of living organismsto evaluate the condition or health of a place (e.g.,a stream, wetland, or woodlot).

Biological integrity “the ability of an aquatic eco-system to support and maintain a balanced, adap-tive community of organisms having a species com-position, diversity, and functional organization com-parable to that of natural habitats within a region”(Karr and Dudley 1981).

Biological monitoring Sampling the biota of aplace (e.g., a stream, a woodlot, or a wetland).

Biota All the plants and animals inhabiting an area.

Composition (structure) The composition of thetaxonomic grouping such as fish, algae, ormacroinvertebrates relating primarily to the kindsand number of organisms in the group.

Community All the groups of organisms living to-gether in the same area, usually interacting or de-pending on each other for existence.

Competition Utilization by different species of lim-ited resources of food or nutrients, refugia, space,ovipositioning sites, or other resources necessaryfor reproduction, growth, and survival.

Criteria A part of water quality standards. Crite-ria are the narrative and numeric definitions condi-tions that must be protected and maintained to sup-port a designated use.

Continuum A gradient of change.

Designated use A part of water quality standards.A designated use is the ecological goal thatpolicymakers set for a waterbody, such as aquaticlife use support, fishing, swimming, or drinking wa-ter.

Disturbance “Any discrete event in time that dis-rupts ecosystems, communities, or population struc-ture and changes resources, substrate availabilityor the physical environment” (Picket and White1985). Examples of natural disturbances are fire,drought, and floods. Human-caused disturbancesare referred to as “human disturbance” and tend tobe more persistent over time, e.g., plowing,clearcutting of forests, conducting urban stormwaterinto wetlands.

Diversity A combination of the number of taxa(see taxa richness) and the relative abundance ofthose taxa. A variety of diversity indexes have beendeveloped to calculate diversity.

Dominance The relative increase in the abundanceof one or more species in relation to the abundanceof other species in samples from a habitat.

Ecological risk assessment An evaluation of thepotential adverse effects that human activities haveon the plants and animals that make up ecosystems.

Ecosystem Any unit that includes all the organ-isms that function together in a given area interact-ing with the physical environment so that a flow of

94

energy leads to clearly defined biotic structure andcycling of materials between living and nonlivingparts (Odum 1983).

Ecoregion A region defined by similarity of cli-mate, landform, soil, potential natural vegetation,hydrology, and other ecologically relevant variables.

Gradient of human disturbance The relativeranking of sample sites within a regional wetlandclass based on degrees of human disturbance (e.g.,pollution, physical alteration of habitats, etc.)

Habitat The sum of the physical, chemical, andbiological environment occupied by individuals of aparticular species, population, or community.

Hydrology The science of dealing with the prop-erties, distribution, and circulation of water both onthe surface and under the earth.

Impact A change in the chemical, physical (includ-ing habitat), or biological quality or condition of awaterbody caused by external forces.

Impairment Adverse changes occurring to an eco-system or habitat. An impaired wetland has somedegree of human disturbance affecting it.

Index of biologic integrity (IBI) An integrativeexpression of the biological condition that is com-posed of multiple metrics. Similar to economic in-dexes used for expressing the condition of theeconomy.

Intolerant taxa Taxa that tend to decrease in wet-lands or other habitats that have higher levels ofhuman disturbances, such as chemical pollution orsiltation.

Macroinvertebrates Animals without backbones(insects, crayfish, clams, snails, etc.) that are caughtwith a 500-800 micron mesh net.Macroinvertebrates do not include zooplankton orostracods, which are generally smaller than 200 mi-crons in size.

Metric An attribute with empirical change in valuealong a gradient of human disturbance.

Minimally impaired site Sample sites within aregional wetland class that exhibit the least degreeof detrimental effect. Such sites help anchor gradi-ents of human disturbance and are commonly re-ferred to as reference sites.

Most-impaired site Sample sites within a regionalwetland class that exhibit the greatest degree ofdetrimental effect. Such sites help anchor gradientsof human disturbance and serve as important refer-ences, although they are not typically referred to asreference sites.

Population A set of organisms belonging to thesame species and occupying a particular area at thesame time.

Reference site (as used with an index of bio-logical integrity) A minimally impaired site that isrepresentative of the expected ecological conditionsand integrity of other sites of the same type andregion.

Stressor Any physical, chemical, or biological en-tity that can induce an adverse response.

Taxa A grouping of organisms given a formal taxo-nomic name such as species, genus, family, etc. Thesingular form is taxon.

Taxa richness The number of distinct species ortaxa that are found in an assemblage, community,or sample.

Tolerance The biological ability of different spe-cies or populations to survive successfully within acertain range of environmental conditions.

Trophic Feeding, thus pertaining to energytransfers.

Wetland(s) (1) Those areas that are inundated orsaturated by surface or groundwater at a frequencyand duration sufficient to support, and that undernormal circumstances do support, a prevalence ofvegetation typically adapted for life in saturated soilconditions [EPA, 40 C.F.R.§ 230.3 (t) / USACE,33 C.F.R. § 328.3 (b)]. (2) Wetlands are lands

95

transitional between terrestrial and aquatic systemswhere the water table is usually at or near the sur-face or the land is covered by shallow water. Forthe purposes of this classification, wetlands musthave one or more of the following three attributes:(a) at least periodically, the land supports predomi-nantly hydrophytes, (b) the substrate is predomi-nantly undrained hydric soil, and (c) the substrate isnonsoil and is saturated with water or covered byshallow water at some time during the growing sea-son of each year (Cowardin et al. 1979). (3) Theterm “wetland” except when such term is part ofthe term “converted wetland,” means land that (a)

has a predominance of hydric soils, (b) is inundatedor saturated by surface or ground water at a fre-quency and duration sufficient to support a preva-lence of hydrophytic vegetation typically adaptedfor life in saturated soil conditions, and (c) undernormal circumstances does support a prevalenceof such vegetation. For purposes of this Act andany other Act, this term shall not include lands inAlaska identified as having a high potential for agri-cultural development which have a predominanceof permafrost soils [Food Security Act, 16 U.S.C.801(a)(16)].

Documents

Methods for Evaluating Wetland Condition: Wetland ...€¦ · Methods for Evaluating Wetland Condition: Wetland Biological Assessment Case Studies. Office of Water, U.S. Environmental