This Technical Memorandum is one of a series of
publications designed to assist watershed projects,
particularly those addressing nonpoint sources of
pollution. Many of the lessons learned from the
Clean Water Act Section 319 National Nonpoint
Source Monitoring Program are incorporated in these
publications.
United States
Environmental Protection
Agency
Technical Memorandum #4
October 2015
Applying Benthic Macroinvertebrate
James B. Stribling and Steven A. Dressing. 2015. Technical
Memorandum #4: Applying Benthic Macroinvertebrate
Multimetric Indexes to Stream Condition Assessments,
October 2015. Developed for U.S. Environmental Protection
Agency by Tetra Tech, Inc., Fairfax, VA, 14 p.
Available online at https://www.epa.gov/polluted-runoff-
nonpoint-source-pollution/watershed-approach-technical-
resources.
Multimetric Indexes to Stream
Condition Assessments
Introduction
The primary objective of the Federal Water Pollution Control Amendments of 1972—commonly
known as the Clean Water Act (CWA)—"is to restore and maintain the chemical, physical, and biolog-
ical integrity of the Nation's waters" (Title 33 of the United States Code section 1251). To strengthen
the scientific foundation of the CWA, "biological integrity" has been defined as "the ability (of a water
body) to support and maintain a balanced, integrated, biological community having a species compo-
sition, diversity, and functional organization comparable to that of natural habitats in the region" (Karr
and Dudley 1981; Schneider 1992). The capacity for aquatic organisms to survive and reproduce in
nature is controlled by both basic biological and physiological processes of the organisms and char-
acteristics of their immediate environment. A water body with chemical and physical characteristics
that are close to those found in a naturally occurring habitat can be considered to represent chem-
ical and physical integrity, and therefore potentially supportive of a healthy biological condition.
Waterbodies may be physically or chemically degraded by a variety of forces. Environmental distur-
bances are caused either by natural extremes (e.g., hurricanes, volcanic eruptions, earthquakes) or
by human activities (e.g., waste discharge, cropland erosion, stormwater runoff), and both types can
lead to changes in the natural environment. Many human activities lead to direct introduction of
chemical pollution, changes in physical habitat, or alteration of hydrologic processes. These changes
may diminish the availability of food and habitat area—including refuge from natural extremes
such as stream flow or temperature—and can alter reproductive behavior. Such factors are termed
"stressors." For purposes of this technical memorandum, a stressor is defined as any agent that limits
the biological capacity for survival and reproduction. Biological responses to stressor exposure
are varied, but can include direct mortality (e.g., from acute toxicity), longer term chronic effects,
reduced reproductive success, increased incidence of disease and/or predation, and elimination of a
local population from loss of habitat.
Degraded biological conditions result from the buildup of stressors, which results from environ-
mental degradation. As stewards, we must recognize and effectively document environmental
degradation and accept the responsibility to take appropriate actions to address it and its causes.
Using aquatic organisms as indicators of changes to chemical and physical components of the
environment can enhance our ability to identify the level of degradation and the actions to
take (Figure 1). Rigorous and defensible measurement techniques do not exist for all pollutants,
combinations of pollutants, or other potential detrimental factors. But, because aquatic organ-
isms are directly exposed to their immediate surroundings, biological characteristics will reflect
i

-------
Technical Memorandum #4 | Applying Benthic Macroinvertebrate Multimetric Indexes to Stream Condition Assessments
October 2015
Stressor sources
produce
environmental suitability, including factors that are unknown, unmea-
sured, and often, unmeasurable. In addition, the community of aquatic
biota naturally and cumulatively integrates the effects of multiple and
complex stressors over time, whether exposure is continuous, intermit-
tent, or episodic. For those pollutants, or stressors, that are less than

continuous, use of biological indicators sidesteps the need to sample
specifically when the pollutants are present.
Stressors
Figure 1. Relationship between Stressors
and Response Indicators.
Response indicators

effect
Over the past 20-30 years, methods and protocols have been
developed to support characterization of biological condition through
direct field sampling, laboratory and data analysis, and stream and
watershed assessment. Protocols are typically customized to meet
technical needs, with individual monitoring programs synchronizing
capabilities and staff resources within budgetary constraints.
Simultaneously, new quality control (QC) procedures have been
developed to ensure that data and assessments are of known and
acceptable quality (Flotemersch et al. 2006; Stribling 2011).
The organism groups most commonly used in routine biological monitoring and assessment
programs in the United States are benthic macroinvertebrates (BMIs)—aquatic insects, snails,
mollusks, crustaceans, worms, and mites; fish; and/or algae—with indicators most often taking the
form of a multimetric Index of Biological Integrity, or IBI (Karr et al. 1986; Hughes et al. 1998; Barbour
et al. 1999; Hill et al. 2000, 2003). An IBI combines several individual metrics, each of which describes
a different aspect of assemblage structure and function (Barbour et al. 1995). Note that IBIs are occa-
sionally called multimetric indexes (MMIs).The acronym "IBI" is a direct connection of the indicator
to biological integrity of the CWA, whereas "MMI" is a generic term simply meant to communicate,
in part, the numeric structure of the index.
Although fish and algae also are often used for evaluating the biological condition of water bodies,
this technical memorandum focuses solely on BMIs. This group of organisms has continually been
proven for use in routine biological monitoring because field sampling protocols are well-es-
tablished, the level of effort required for field sampling is reasonable (Barbour et al. 1999), and
taxonomic expertise is relatively easily accessible. An IBI has a straightforward numeric structure
that serves to summarize and scale complex biological data. Within boundaries defined by the
sampling and analysis protocol, the BMIs in a stream sample represent the full taxonomic diversity of
the assemblage present at a site. Taxonomic diversity, compositional abundance, and autecological
characteristics (e.g., feeding types, habits, and stressor tolerance values) each conveys information
potentially useful in detecting and understanding degradation and biological responses to it. An
effort is typically made during IBI development to have one or more individual metrics represent as
many different information types, or metric categories, as possible associated with the biological
assemblage, including taxonomic richness, community composition, stressor or pollution tolerance,
functional feeding types, and locomotory habit (Barbour et al. 1999). This diversity of categories
increases the probability that an index can detect the effects of multiple and complex stressors. The
multimetric IBI approach allows numeric consideration of the composite sample content (via the
2

-------
Technical Memorandum #4 | Applying Benthic Macroinvertebrate Multimetric Indexes to Stream Condition Assessments
October 2015
biological index score) and places the sample result on the same scale with other samples from the
same or different sample locations and/or from different water bodies.
The structure of IBIs allows them to be easily disaggregated into individual metrics as well as
individual taxa. Properly calibrated indexes will be made up of metrics that have been tested for
direction of change in the presence of stressors and are based on taxa with which autecological
attributes are associated. It also is possible to evaluate the specific component metrics and taxa that
have the most influence on the overall index score and individual metric values and, thus, the site
assessment. Also, since typically, multiple and complex stressors are present in streams to which the
assemblage is exposed, examining overall index scores and assessments, metric values, and taxa can
often provide evidence pointing to stressor types and intensities.
In a very broad sense, two components make up the
"life cycle" of an IBI: 1) calibration or development,
and 2) application. Driving the need for calibration is
the recognition that one size does not fit all and that
customizing the index makes each indicator more effective
in detecting degradation. Natural variation is observed
in benthic macroinvertebrate assemblages across all
spatial (e.g., regions, landscapes, water body types) and
temporal (e.g., daily, monthly, seasonally, and within and
among years) scales (Figure 2). Calibrating a biological
index defines protocols and procedures that help control
variability in sample data so that differences in index scores
and assessments can be interpreted as indicative of real
environmental change, including detection of degradation
due to stressors from both nonpoint and point sources.
Variability in data resulting from their collection, analysis,
and documentation—known as "systematic error"—is
minimized by using standard operating procedures and
ensuring acceptability for application through use of
routine and consistent QC (Stribling 2011).
Purpose and Audience
The purpose of this technical memorandum is to describe the second component of the IBI life
cycle: its application to assessment of and reporting on aquatic ecological condition of a water
body. The content of the memorandum is presented with the assumption that an index has been
appropriately calibrated for the region and water body type of interest. It describes field sampling
for benthic macroinvertebrates, laboratory processing (i.e., sorting/subsampling and taxonomic
identification), IBI calculation, and site assessment. This technical memorandum should be most
helpful to staff responsible for implementing biological monitoring and assessment programs for
nontidal streams and rivers.
Subsample
Minute
small scale
V/
Sample
Hourly
X/
X/
Reach
Daily
v A
\/
Stream
Weekly
v m
x/
Subwatershed
Monthly

v/
Watershed
Annual
Large scale
x/
Continent
Decadal
Figure 2. Range of Spatial and Temporal Scales Used
in Designing Environmental Monitoring
Programs and Calibrating Biological Indexes
(Stribling 2011).
3

-------
Technical Memorandum #4 | Applying Benthic Macroinvertebrate Multimetric Indexes to Stream Condition Assessments
October 2015
The biological assessment results presented in this technical memorandum do not tell environ-
mental managers what actions to take; rather, they stop with the documentation of the presence/
absence and severity of a problem (in this case, biology was not degraded, and thus, no problem
exists). Additional analysis and interpretation is often necessary to increase confidence when deter-
mining necessary and appropriate actions. Stressor identification and causal analysis (USEPA 2010) is
a systematic approach for determining the stressors and stressor sources most likely to be causing
the problem, and improve support for decisions on the types and locations of stressor control activi-
ties (e.g., best management practices, stream restoration, stormwater management).
Quality Control and the Use of Measurement Quality
Objectives
Reliability of biological assessment results can be impaired by errors introduced throughout the
process, beginning with field sampling and sample sorting and subsampling and ending with metric
and index calculation, and site condition assessment. Use of a series of QC checks—for example,
measurement quality objectives (MQOs) (Stribling et al. 2011)—integrated into the process helps the
assessor to recognize, control, and minimize errors (Table 1). If observed values exceed the MQOs,
it does not automatically indicate the existence of invalid or unacceptable data points. Rather, the
values are targeted for closer scrutiny to determine possible reasons for the exceedance and might
indicate a need for corrective actions.
Table 1. Key Measurement Quality Objectives (MQO) Used for QC Evaluation and Tracking Data Quality
Performance Characteristic
MQO
Field sampling precision
(multimetric index)
CV < 10%, for a sampling event (field season, watershed, or other
strata)
CI90 < 15 index points (on a 100-point scale)
RPD < 15
Field sampling completeness
Completeness > 98%
Sorting/subsampling bias
PSE>90 (for > 90% of externally QC'd sort residues)
Taxonomic precision
Median PTD < 15% for overall sample lot (samples with PTD > 15%
examined for patterns of error)
Median PDE < 5% (samples with PDE > 5% should be further
examined for patterns of error)
Taxonomic completeness
Median PTC > 90% (samples with PTC < 90% should be examined
and taxa not meeting targets should be isolated)
mAbs diff < 5%
Notes:
CI90 = 90% confidence interval; CV = coefficient of variability; mAbs diff = median absolute difference;
PDE = percent difference in enumeration; PSE = percent sorting efficiency; PTC = percent taxonomic completeness;
PTD = percent taxonomic disagreement; QC = quality control; RPD = relative percent difference.
4

-------
Technical Memorandum #4 | Applying Benthic Macroinvertebrate Multimetric Indexes to Stream Condition Assessments
October 2015
IBI Application
Field Sampling
Sites to be sampled are typically preselected and defined in the office using maps and other avail-
able data, regardless of whether they are targeted or probability-based. Assessments at targeted
sites are for the specific locations from which the samples are taken, whereas those from proba-
bility sites can also be aggregated to broader-scale assessments, such as watershed. Reach length
sampled varies among programs, but 100 meters is widely used for wadeable streams (generally,
1st—4th order Strahler). Some programs sample reach lengths that are a multiple of the median
wetted channel width (MWCW), such as 20x or 40x (Flotemersch et al. 2011).
For its national surveys of streams and rivers, the U.S. Environmental Protection Agency defined
a sample reach as 40x the mean wetted width (USEPA 2004) and collected BMI samples along 11
transects evenly distributed throughout the reach. Using the transect approach helps minimize and
control bias by the field crews in selecting the specific habitats sampled in the reach. An alternative
approach—proportional distribution—has the field crews estimate the proportion of different habitat
types in a defined reach (e. g., 100m, and distribute a fixed level of sampling effort in proportion to
their frequency of occurrence throughout the reach) (Barbour et al. 1999; Stribling 2011). Samples
from both approaches result in a biological descriptor (i.e., using BMI) of the full sampling reach.
Carter and Resh (2013) reported that D-frame nets are
commonly used across the United States by biological
monitoring programs for composite sampling of BMI.
Organic and inorganic sample material (e.g., leaf litter,
small woody twigs, silt, and sand) are composited in oi
or more containers (Figure 3), preserved in 95 percent
denatured ethanol, and delivered to laboratories for
processing.
Repeat sampling for a monitoring and assessment
program provides data for both field QC evaluation
and for calculating different measures of field
sampling precision (Stribling 2011; Flotemersch et al.
2006). A rate of 10 percent duplication is relatively
common among programs (Carter and Resh 2013),
and we recommend collecting duplicate samples at
a minimum of three reaches. A duplicate sample in
streams and other flowing waters would be one taken
from an adjacent reach (i.e., one that begins at either
the upstream or downstream extent of the primary
reach). Reaches for which duplicate samples are taken
are randomly selected prior to initiation of the field
effort. Sample labeling should be identical between
primary and duplicate samples, with the exception
Field
ne

o.
£
o
Laboratory	^7
sorting/subsampling using the Caton griddedscreen \
C1
"Clean"
subsample
C2
Sort
residue
C3
Unsorted sample
remains
Figure 3. Benthic macroinvertebrate samples are field-
collected, composited, and preserved in one
or more containers with 95% ethanol. They are
transferred to the laboratory, and the sorting/
subsampling process results in three sample
units: containers CI, C2, and C3.
5

-------
Technical Memorandum #4 | Applying Benthic Macroinvertebrate Multimetric Indexes to Stream Condition Assessments
October 2015
that the latter should be clearly labelled as such (e. g., with "DUP" or "QC"). Data quality indicators
for field sampling cannot be calculated until after laboratory processing is completed (see the
Sample Processing and Data Preparation section). Performance characteristics recommended for
quantifying field sampling precision include relative percent difference (RPD), coefficient of variation
(CV), and 90 percent confidence intervals (CI90). For relevant formulas, see Stribling (2011) and/or
Flotemersch et al. (2006).
Sample Processing and Data Preparation
Several aspects of BMI sample processing must be completed prior to calculating performance
measures or indicator values (i.e., metrics and index). First, the field samples are sorted and
subsampled, which serves to separate individual organisms from nontarget material such as leaf
litter and other detritus, bits of woody material, silt, and sand. Simultaneously, a randomly selected
subset of organisms (the subsample) is isolated for taxonomic identification. Names (accepted
nomenclature) are matched to individual specimens in the subsample, resulting in a list of organisms
by name and the number of individuals of each taxon.
Sorting/subsampling. Recommended equipment and supplies for sorting and subsampling in
the laboratory include the Caton gridded screen and accessories (Figure 4), forceps, sample jars or
vials, and 70-80-percent ethanol. Typically, sorting is done solely with the naked eye or using some
combination of an illuminated ring lamp with 3-5x magnification and a binocular dissecting micro-
scope (10-40x magnification). The process involves the following steps:
•	Step 1. Empty material from sample jar onto gridded subsampling tray; fully and evenly
spread material on tray, giving special attention to the corner squares.
•	Step 2. Randomly select four grid squares.
•	Step 3. Lift screen from outer box, offset (as shown in Figure 4).
•	Step 4. Remove material from the four selected grids and place in separate picking tray (use
cookie cutter and scoop; if necessary, use scissors to cut leaf litter, algae, small twigs)
•	Step 5. Pick ali specimens from
the four grids' worth of sample
materia! (use forceps and/or eye
dropper and keep track of rough
count); place specimens in one
or more viais (Figure 3, C1) with
approximately 70-80 percent
ethanol.
•	Step 6. After all specimens are
removed from picking tray, pour
remaining material from picking
tray into a separate container
(Figure 3, C2), clearly labelled as
"Sort residue."
Figure 4. Caton Gridded Screen Photo with All Accessories,
including "Cookie Cutter" Frame (A), Spatula (B),
Scoop (C), 480-Micron Screen (D), and Outer Tray (E).
6

-------
Technical Memorandum #4 | Applying Benthic Macroinvertebrate Multimetric Indexes to Stream Condition Assessments
October 2015
•	Step 7. If rough count is less than target count, randomly select another grid (1), repeat from
step 4; repeat with single grids until target count is exceeded; make sure to pick final grid to
completion.
•	Step 8. Transfer sample material remaining on the gridded screen back into original or other
sample container (Figure 3, C3), labelling it as "Unsorted sample remains."
Various target counts are used for monitoring, consistent within programs, with the majority being
200-300 (Carter and Resh 2013). This range must be consistent within programs, but can range from
100-600 among programs. The subsampling process results in at least three sample containers for
each sample (Figure 3)1. The first container (C1) has the clean subsample of specimens that will be
given to the taxonomist for identification and counting; this is generally one or more glass vials,
typically 7-9 drams (25-33 ml). Note that some labs prefer primary sorters to segregate certain
organism groups into isolated vials—for example, midges (Chironomidae), snails, clams and mussels
(Mollusca), or scud, sowbugs, and crayfish (Crustacea). The second container (C2) holds the sort
residue—that is, the material remaining in the separate picking tray after all organisms have been
removed. This container will most likely be a 1L Nalgene jar with a plastic screw cap or similar vessel.
The sort residue is retained for use in a QC check for missed specimens. The third container (C3)
holds the unsorted sample remains. If there are problems with the specimens removed by the sorter
or some other QC issue, additional sorting from C3 might be necessary. Stribling (2011) provides
more detail on the sorting and subsampling procedure using the Caton screen. QC is performed
by having an independent sort-checker go through the sort residue (C2) to check for missed spec-
imens. This should be done for a randomly selected set of 10 percent or a minimum of three of the
samples. Percent sorting efficiency (PSE) is the performance characteristic used to quantify sorting
bias (Stribling 2011).
Taxonomic identification. Identification of BMI is the process of associating a single name with
each specimen in the subsample using a technical taxonomic key or comparing it to a reference
collection of preserved organisms. Personnel extensively experienced in taxonomic identification
often use sight identification, or sight ID, to recognize the taxon or morphological and anatomical
features of a specimen without consulting the literature. Whether using morphology-based
dichotomous keys, technical diagnostic literature, or sight ID, the most important factor is that the
taxonomist record the result using the nomenclatural standard specified for the project. The end
result of the overall identification process for a sample is a list of taxa and the number of individuals
of each (Table 2). Quality control for the taxonomic identification step of the assessment process is
achieved through confirmation by the QC taxonomist of a randomly selected subset of the samples
that were already identified by the primary taxonomist (Stribling 2011). Performance characteristics
for quantifying taxonomic precision and consistency are percent taxonomic disagreement (PTD),
percent difference in enumeration (PDE), and percent taxonomic completeness (PTC) (Stribling 2011)
(Table 1).
1 Each of C1-3 can actually be multiple containers. Following identifications by the primary taxonomist, CI will
potentially include slide-mounted specimens (midges and worms). In any case, the different sets of sample
containers should be considered sample subunits.
7

-------
Technical Memorandum #4 | Applying Benthic Macroinvertebrate Multimetric Indexes to Stream Condition Assessments
October 2015
Table 2. Raw Data from One Benthic Macroinvertebrate Sample (Example), including Autecological Designations
Family
Taxon
Count3
STV
FFG
Habit
Enchytraeidae
Enchytraeidae
3
4.9
CG
BU
Lumbricidae
Lumbricidae
2
8.3
CG
BU
Cambaridae
Cambaridae
2
6.3
SV
SP
Asellidae
Lirceus
2
7.3
SV
SP
Elmidae
Ancyronyx
3
2.0
OM
CN
Elmidae
Macronychus
2
2.4
OM
CN
Elmidae
Microcylloepus
1
1.9
CG
BU
Elmidae
Stenelmis
7
4.8
SC
CN
Ptilodactylidae
Anchytarsus
1
4.0
SH
CN
Athericidae
Atherix
1
2.0
PR
SP
Ceratopogonidae
Ceratopogonidae
8
4.7
PR
SP
Chironomidae
Kloosia
1



Chironomidae
Pagastiella
1
0.0
CG
SP
Chironomidae
Polypedilum
13
4.1
SH
Cb
Chironomidae
Tribelos
1
2.9
CG
BU
Chironomidae
Brillia
1
2.9
SH
BU
Chironomidae
Corynoneura
3
3.2
CG
SP
Chironomidae
Cricotopus/Orthocladius
1
5.8
SH
SP
Chironomidae
Parametriocnemus
11
3.1
CG
SP
Chironomidae
Pseudorthocladius
3
1.1
CG
SP
Chironomidae
Rheosmittia
20
7.0
CG
BU
Chironomidae
Ablabesmyia
1
5.0
PR
SP
Chironomidae
NHotanypus
1
3.0
PR
SP
Chironomidae
Thienemannimyia genus gr.
10
6.0
PR
SP
Chironomidae
Micropsectra
1
1.5
CG
CN
Chironomidae
Rheotanytarsus
1
3.3
CF
CN
Chironomidae
Stempellinella
1
1.6
CG
CN
Chironomidae
Tanytarsus
19
3.5
CF
CN
Empididae
Hemerodromia
3
4.2
PR
SP
Tabanidae
Tabanus
1
7.4
PR
SP
Tipulidae
Tipula
2
5.0
SH
BU
Tipulidae
Hexatoma
1
0.0
PR
BU
Heptageniidae
Maccaffertium
4

SC
CN
Neoephemeridae
Neoephemera
1
2.1
CG
SP
Coenagrionidae
Argia
1
6.5
PR
CN
8

-------
Technical Memorandum #4 | Applying Benthic Macroinvertebrate Multimetric Indexes to Stream Condition Assessments
October 2015
Table 2. Raw Data from One Benthic Macroinvertebrate Sample (Example), including Autecological Designations
Family
Taxon
Count"
STV
FFG
Habit
Corduliidae
Neurocordulia
1
5.0
PR
CB
Corduliidae/Libellulidae
Corduliidae/Libellulidae
1
5.0
PR
CB
Chloroperlidae
Perlinella
1
2.0
PR
CN
Leuctridae
Leuctra
7
0.0
SH
SP
Perlidae
Perlesta
25
1.8
PR
CN
Calamoceratidae
Anisocentropus
3
2.0
SH
SP
Hydropsychidae
Cheumatopsyche
6
5.8
CF
CN
Hydropsychidae
Diplectrona
1
2.0
CF
CN
Hydroptilidae
Hydroptila
1
3.8
PI
CN
Leptoceridae
Oecetis
11
2.4
PR
CN
Leptoceridae
Triaenodes
9
0.7
SH
SW
Limnephilidae
Pycnopsyche
1
1.4
SH
SP
Philopotamidae
Chimarra
3
1.2
CF
CN
Polycentropodidae
Cernotina
1
1.2
PR
CN
Polycentropodidae
Neureclipsis
2
2.7
CF
CN
Pisidiidae
Pisidiidae
2
5.4
CF

Notes:
a The number of individuals in the sample.
BU = burrower; CB = climber; CF = coIlector-fiIterer; CG = collector-gatherer; CN = dinger; FFG = functional feeding group;
OM = omnivore; PI = piercer; PR = predator; SC = scraper; SH = shredder; SP = sprawler; STV = stressortolerance value;
S V = scavenger; SW = swimmer.
Autecological attributes. "Autecology" is the study of the relationship between an individual
taxon (usually a species) and its immediate environmental factors. Attributes related to autecology
are used for calculating some individual metrics and also can contribute to additional interpretation
of disaggregated indexes and metrics. Attributes used for BMI are functional feeding groups (FFG),
habit, and stressor tolerance values (TV). Attributes should be assigned to the taxa list after taxo-
nomic identification and any necessary QC corrective actions. Merritt et al. (2008) is a good source
for FFG and habit attributes, as is Barbour et al. (1999) for TV. Many states, however, have done
additional TV analyses to calibrate more specifically to their region, variable water body types and
conditions, and dominant stressor characteristics. Most current monitoring programs use relational
databases (e.g., Microsoft Office Access* customized to the program structure) that include data
tables with attributes already assigned to individual taxa.
At this point in the process, the data are ready for metric and index calculations and site
assessments.
9

-------
Technical Memorandum #4 | Applying Benthic Macroinvertebrate Multimetric Indexes to Stream Condition Assessments
October 2015
Metric Calculation and IBI Aggregation
Two of the main guidelines used for metric selection during index calibration are to: (1) have at least
one metric in each of the metric categories (Barbour et al. 1995) and (2) test the direction of response
in the presence of stressors. Evaluation and selection of metrics typically involves testing of many
more metrics than end up in the final index. They also are scaled to 100 points, which enables them
to be averaged together as the full, multimetric biological index on each sample. As an example,
calibration performed on data from Mississippi (Bioregion East) resulted in a benthic index made up
of six metrics (Table 3). Only the Hilsenhoff Biotic Index (HBI) requires a formula that is not evident
from the description (Hilsenhoff 1982). It is calculated using the following formula:
Zn.a,
if
where:
n. and a = the number of individuals and the stressor tolerance value of taxon /', respectively.
N = the total number of individuals in the sample.
Table 3. Benthic Macroinvertebrate Metrics, Text Description, Category, and Trend with Increasing Stressor Load
Metric Name
(abbreviation)
Description
Category
Direction of Change
with Increasing
Stressors
Total number of taxa
(TotalTax)
Number of distinct taxa identified in the
subsample
Richness
Decrease
Total number of EPT
taxa (EPTTax)
Number of distinct taxa in the insect orders
Ephemeroptera (mayflies), Plecoptera (stoneflies),
and Trichoptera (caddisflies)
Richness
Decrease
Percent individuals
as Cricotopus/
Orthocladius/
Chironomus of total
Chironomidae
(COC2ChiPct)
Count of individuals in these relatively tolerant
genera as percent total Chironomidae individuals
in the sample
Composition
Increase
Percent individuals
as sensitive EPT
(PSensEPT)
Of all individuals in the sample, the percentage
of individuals in the insect orders Ephemeroptera
(mayflies), Plecoptera (stoneflies), and Trichoptera
(caddisflies), except individuals in the families
Caenidae, Baetidae, Hydropsychidae, and
Hydroptilidae
Composition
Decrease
Number of taxa, as
shredders (ShredTax)
Number of distinct taxa in the sample that are
considered shredders (i.e., they use coarse organic
material—primarily leaf litter—for food
Functional
feeding
group
Decrease
Hilsenhoff Biotic Index
(HBI)
Composite of total relative sensitivity of all
organisms in the sample, calculated as the average
tolerance value of all individuals in the sample
Tolerance
Increase
10

-------
Technical Memorandum #4 | Applying Benthic Macroinvertebrate Multimetric Indexes to Stream Condition Assessments
October 2015
Following is an example metric calculation for Bioregion East using raw sample data that consists
of a list of taxa, number of individuals of each, and each taxon's associated autecological attributes
(i.e., FFG, habit, and TV) (Table 2). This sample result has 51 taxa represented by 209 specimens. The
sample is dominated by midges (Chironomidae; 89 specimens, -43 percent of the whole sample),
but also has a substantial number of caddisflies (Trichoptera, 10 genera in seven families; 38 speci-
mens, -19 percent) and stoneflies (Plecoptera; 33 specimens in three genera, -16 percent).
Individual metric values are calculated using the sample data, but because the metrics are mostly
on different scales, they cannot be directly aggregated. Formulas developed during the index cali-
bration process (not shown in this technical memorandum) allow the individual metric values to be
converted to a 100-point scale (Table 4). The overall benthic IBI score for a single sample at this site is
the mean value of the six metrics, which in this example is 96.3.
Table4.Metric Calculations from Example Sample

Metric3
Valueb
Formula
Scorec
1
TotalTax
51
100*(metric value)/51.5
99
2
EPTTax
15
100*(metric vaiue)/14
100
3
COC2ChiPct
1.1
100*(45-(metric value))/45
97.6
4
PSensEPT
32.4
100*(metric value)/39
83.1
5
ShredTax
9
100*(metric value)/7
100
6
HBI
3.6
100*(8.5-(metric value))/5
98

IBI

Sum of metric scores/6
96.3

Narrative Rating

Non-degraded
Notes:
a Metric abbreviations defined in Table 3.
b Numbers calculated or compiled directly from sample data, list of taxa, and counts of individuals of each,
c Normalized to a 100-point scale using formulas developed during the index calibration process.
Site Assessment
Continuing with our example site for Bioregion East, the IBI impairment threshold is 65.7 on a
100-point scale, meaning that sites with aggregated scores (mean values) falling above it are
considered "non-degraded," or similar to reference conditions. This threshold was defined during
calibration based on the 75th percentile of the reference site distribution. It has a discrimination
efficiency2 of 87.3 percent and a 90-percent confidence interval of ±12.5 points. The biological index
score of 96.3 for this site sample falls well above the threshold, providing good confidence in the
finding that the site is non-degraded. By disaggregating the index, the investigator can see that
individual metrics all have high scores, the lowest being 83 (percent sensitive EPT). In this instance,
2 Sometimes abbreviated as DE, this is an estimate of the accuracy of indexes and metrics characterized during
calibration. It is a statement of their capacity to correctly identify stressor conditions, and is quantified using the
formula: DE = (a/b)x100, where a is the number of a priori stressor sites identified as being below the quantified
biological impairment threshold of the reference distribution (25th percentile, 10th, or other), and b is the total
number of stressor sites.
11

-------
Technical Memorandum #4 | Applying Benthic Macroinvertebrate Multimetric Indexes to Stream Condition Assessments
October 2015
none of the metrics are interpreted as more heavily influencing the narrative of being non-degraded
than any other. A sample with 51 (primarily) genus-level taxa is a solid taxonomic diversity for
macroinvertebrates. More detailed evaluation of the raw sample data, however, including auteco-
logical characteristics of individual taxa and their numerical dominance in the sample, can provide
useful information. For example, the vast majority of individual organisms (75 percent) and taxa
(67 percent) have TV<5, indicating a dominance of stressor-sensitive biota. Caddisflies (Trichoptera)
are generally considered to be stressor-sensitive (i.e., intolerant). The sample has 38 individuals
representing 10 genera in seven caddisfly families. The full range of TV is from 0 to 10 with higher
numbers indicating greater stressor tolerance. All but two of the 10 genera (Cheumatopsyche and
Hydroptila) have TV<3. This level of stressor tolerance (sensitivity) would also be reflected partly by
the metric 'percent individuals as sensitive EPT (PSensEPT)' (Table 2). Although not as diverse in the
sample, stoneflies (Plecoptera) are similarly considered to be sensitive to stressors; the three genera
have TVs of 0.0,1.8, and 2.0. Thus, the assessment summary is that the site is not degraded. It has an
IBI score of 96.3 and the sample is dominated by stressor-sensitive organisms that would most likely
not be present if the location was substantially polluted.
Summary/Recommendations
The information presented here covers the background and procedures for field sampling and
laboratory processing of BMI samples from freshwater streams as well as how to calculate and use
a multimetric index as an indicator of water resource quality. The example biological assessment
presented in this technical memorandum shows the result from a stream site that is non-degraded.
The stressors that may be present have not been substantially detrimental to the survival or repro-
duction of the benthic community. Following these guidelines will provide rigorous and defensible
assessments of ecological condition. The IBI provides a technique for summarizing complex biolog-
ical data into a format that can be scaled and ranked relative to data from other sites and samples
and then translated into an assessment narrative. It is important, however, to use metrics, metric
aggregation techniques, and scoring criteria—especially the degradation decision threshold—that
have been appropriately calibrated for the water body type, site class, and region of the project. Use
of an IBI that has not been so calibrated increases the potential for spurious or misleading results.
Whereas this document takes the process to the point of demonstrating the presence or absence
and severity of a problem, further more detailed analysis of stream biological assessment results
may be needed to identify the stressors and sources causing the degradation. Application of stressor
identification and causal analysis would inform decisions on management actions to be taken to
address identified water quality problems.
12

-------
Technical Memorandum #4 | Applying Benthic Macroinvertebrate Multimetric Indexes to Stream Condition Assessments
October 2015
References/Additional Resources
Barbour, M.T., J.B. Stribling, and J.R. Karr. 1995. The Multimetric Approach for Establishing Biocriteria
and Measuring Biological Condition. Chapter 6 in Biological Assessment and Criteria: Tools for
Water Resource Planning and Decision Making, ed. W.S. Davis and TP. Simon, pp. 63-77. Lewis
Publishers, Boca Raton, FL.
Barbour, M.T., J. Gerritsen, B.D. Snyder, and J.B. Stribling. 1999. Rapid Bioassessment Protocols for
Streams and Wadeable Rivers: Periphyton, Benthic Macroinvertebrates and Fish. Second Edition.
EPA/841 -D-97-002. U.S. Environmental Protection Agency, Office of Water, Washington, DC.
Accessed September 30, 2015.
http://water.epa.aov/scitech/monitorina/rsl/bioassessment/index.cfm.
Carter, J.L., and V.H. Resh. 2013. Analytical Approaches Used in Stream Benthic Macroinvertebrate
Biomonitoring Programs of State Agencies in the United States. Open-File Report 2013-1129. U.S.
Geological Survey, Reston, VA.
Flotemersch, J.E., J.B. Stribling, and M.J. Paul. 2006. Concepts and Approaches for the Bioassessment
of Non-Wadeable Streams and Rivers. EPA/600/R-06/127. U.S. Environmental Protection Agency,
Office of Research and Development, National Exposure Research Laboratory, Cincinnati, OH.
Accessed September 30, 2015. http://www3.epa.aov/reaionl/npdes/merrimackstation/pdfs/ar/
AR-1157. pdf# aa= 1.80358807.1889529594.1444762539.
Flotemersch, J.E., J.B. Stribling, R.M. Hughes, L. Reynolds, M.J. Paul, and C. Wolter. 2011. Reach length
for biological assessment of boatable rivers. River Research and Applications 27(4):520-535. doi:
10.1002/rra.1367.
Hill, B.H., A.T. Herlihy, PR. Kaufmann, R.J. Stevenson, F.H. McCormick, and C.B. Johnson. 2000. Use
of periphyton assemblage data as an index of biotic integrity. Journal of the North American
Benthological Society 19:50-67.
Hill, B.H., A.T. Herlihy, PR. Kaufmann, S.J. Decelles, and M.A. Vander Borgh. 2003. Assessment of
streams of the eastern United States using a periphyton index of biotic integrity. Ecological
Indicators 2:325-338.
Hilsenhoff, W.L. 1982. Using a Biotic Index to Evaluate Water Quality in Streams. Technical Bulletin 132.
Wisconsin Department of Natural Resources, Madison, Wl.
Hughes, R.M., PR. Kaufmann, A.T. Herlihy, T.M. Kincaid, L. Reynolds, and D.P Larsen. 1998. A process
for developing and evaluating indices offish assemblage integrity. Canadian Journal of Fisheries
and Aquatic Sciences 55:1618-1631.
Karr, J.R., and D. Dudley. 1981. Ecological perspective on water quality goals. Environmental
Management 5:55-68.
13

-------
Technical Memorandum #4 | Applying Benthic Macroinvertebrate Multimetric Indexes to Stream Condition Assessments
October 2015
Karr, J.R., K.D. Fausch, P.L. Angermeier, P.R. Yant, and I J. Schlosser. 1986. Assessing Biological Integrity
in Running Waters: A Method and its Rationale. Special Publication 5. Illinois Natural History
Survey, Champaign, IL.
Merritt, R.W., K.W. Cummins, and M.B. Berg, ed. 2008. An Introduction to the Aquatic Insects of North
America. Fourth Edition. Kendall/Hunt Publishing Company, Dubuque, IA.
Schneider, E.D. 1992. Monitoring for Ecological Integrity. Chapter 79 in Ecological Indicators, Vol. 2, ed.
D.H. McKenzie, D.E. Hyatt, and V.J. McDonald, pp. 1403-1419. Elsevier Applied Science, New York.
Stribling, J.B. 2011. Partitioning Error Sources for Quality Control and Comparability Analysis in
Biological Monitoring and Assessment. Chapter 4 in Modern Approaches to Quality Control, ed.
A.B. Eldin, pp. 59-84. INTECH Open Access Publisher, doi: 10.5772/22388.
USEPA (U.S. Environmental Protection Agency). 2004. Wadeable Stream Assessment: Field Operations
Manual. EPA 841-B-04-004. U.S. Environmental Protection Agency, Office of Water and Office of
Research and Development, Washington, DC.
USEPA (U.S. Environmental Protection Agency). 2010. Causal Analysis/Diagnosis Decision Information
System (CADDIS). U.S. Environmental Protection Agency, Office of Research and Development,
Washington, DC. Last updated September 23,2010. Accessed September 30,2015.
h ttp://www. epa. aov/caddis.
14

-------