EPA 842-R-16-001
Appendices to
A Practitioner's Guide to the Biological Condition Gradient:
A Framework to Describe Incremental Change in
Aquatic Ecosystems
February 2016

-------
   &EPA
   Appendices to
   A Practitioner's Guide to the Biological Condition Gradient:
   A Framework to Describe Incremental Change in
   Aquatic Ecosystems
Front cover sources:
1. Vermont Department of Environmental Conservation
2. Christophe Quintin, flickr
3. ©istockphoto.com
4. USEPA
5. Vermont Department of Environmental Conservation
6. ©istockphoto.com
7. USDA NRCS

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient              February 2016
                     Appendices to

 A Practitioner's Guide to the Biological Condition
  Gradient: A Framework to Describe Incremental
            Change in Aquatic Ecosystems
Note: For a complete list of the acronyms in this document, see the "Abbreviation and Acronym" list with
Chapters 1-6 of this document.

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                 February 2016

Contents
Appendix A: Generalized Stress Axis Support Materials
Appendix B: Examples of Development of the Biological Condition Gradient for
Large Rivers, Estuaries and Coral Reefs

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                      February 2016
Appendix A: Generalized Stress Axis Support  Materials

Appendix A includes a series of tables that provide a conceptual Generalized Stress Axis (GSA) construct,
as well as examples of pressure and stressor indicators for key environmental processes and elements;
the stressors that are produced when these processes and elements are altered by human disturbance;
and possible mechanisms of stressor action on the aquatic biota and habitats. As described in Chapter 5,
indicators of pressure and stressors can be applied to quantify the GSA. Scenarios for two climatic
regions of the U.S., humid-temperate and arid, are included. The tables are listed below.

A-l. Conceptual scenarios for stress related changes in the major environmental factors that influence
biological condition. The scenarios describe potential changes in the factors in two regions of the United
States: a humid-temperate and an arid region.

A-2. Potential impacts of climate change at low/medium and high stress levels.

A-3. Potential indicators for stressors associated with altered flow,  material transport, channel structure,
and riparian/watershed structure.

A-4. Examples of some fundamental environmental processes and materials (e.g., flow, material
transport, channel structure, riparian and watershed structure, biological interactions) that can be
altered by human-induced disturbances (i.e.., pressure indicators) and create stressors.

Background
This appendix is not  a comprehensive compendium of indicators, but it has drawn upon existing sources
of information such as the EPA Recovery Potential Screening (RPS) method (Norton et al. 2009) and
online resource tools.1 The RPS website provides step-by-step instructions for evaluating the
recoverability of degraded watersheds based on user-selected and weighted ecological, stressor, and
social indicators. The site contains reference materials on recovery potential indicators, including their
definitions, relevance to restorability, data sources, measurement methods, and relevant points from
the technical  literature. This online resource also includes a master list of indicators,  including indicators
for pressures, alteration in ecosystem processes and elements, and stressors, all of which can be applied
to development of the Biological Condition Gradient (BCG) x-axis, the GSA. Associated with the RPS site
is the Watershed Index  Online (WSIO),  a national watershed indicator library and online comparative
watershed analysis tool that houses the data for hundreds of ecological, stressor, and social indicators
compiled nationally on the HUC12 watershed scale.2 At the WSIO site, state-specific RPS tools for the
lower 48 states containing over 200 WSIO  indicators at the 12-digit hydrologic unit code (HUC12) scale
can also be downloaded3 (HUC12 watersheds average 35 square  miles in area).

Whereas these online resources have derived many of their metrics from commonly used geospatial
data sources such as land cover, transportation, and impaired waters data sets, they have gone into
considerably more detail and variation than basic "% in the watershed" statistics in order to provide
1 More information is available at: http://www.epa.gov/rps. Accessed February 2016.
2 Watershed Index Online is available at http://www.epa.gov/watershed-index-online. Accessed February 2016.
3 See statewide tools at http://www.epa.gov/watershed-index-online/watershed-index-online-wsio-download-
statewide-tools. Accessed February 2016.
                                                                                            A-l

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                     February 2016


greater choices for the most locally relevant or issue-specific metrics to use. For example, the national
data's variations on impervious cover statistics include riparian zone totals, proportion of different
impervious cover densities in each watershed, and percent by proximity of impervious cover to surface
waters versus merely percent in the watershed overall. Summary metrics by watershed also address the
extent of 303(d)-listed (impaired) waters per watershed by major pollutant, occurrence of existing total
maximum daily loads,  and other management-relevant metrics. These social indicators also could
potentially provide a third axis, also useful to implementing the BCG in practice, that provides insights
on social factors that may make improving biological condition more (or less) easy to accomplish.

Additionally, the U.S. Environmental Protection Agency's (EPA's) Causal Analysis/Diagnosis Decision
Information System, or CADDIS, is a website developed to help scientists and engineers in the regions,
states, and tribes conduct causal assessments in aquatic systems.4 It is organized into five volumes:

    •  Volume 1: Stressor Identification provides a step-by-step guide for identifying probable causes of
       impairment in a particular system, based on EPA's Stressor Identification process. Those
       interested in conducting a complete causal assessment, learning about different types of
       evidence, or reviewing a history of causal  assessment theory, should start with this volume.

    •  Volume 2: Sources, Stressors & Responses provides background information on many common
       sources, stressors, and biotic responses in stream ecosystems. Those interested  in viewing
       source- and stressor-specific summary information (e.g., for urbanization, physical habitat,
       nutrients, metals, pH, and other stressors), should start with this volume.

    •  Volume 3: Examples & Applications provides examples illustrating different steps of causal
       assessments. Those interested in reading completed causal assessment case studies, seeing how
       Stressor Identification worksheets are completed, or examining example applications of data
       analysis techniques, should start with this volume.

    •  Volume 4: Data Analysis provides guidance on the use of statistical analysis to support causal
       assessments. Those interested in learning how to use data in a causal assessment, should start
       with this volume.

    •  Volume 5: Causal Databases provides access to literature databases and associated tools for use
       in causal assessments. Those interested in applying literature-based evidence to a causal
       assessment, should start with this volume.

The conceptual diagrams in the EPA Causal Analysis CADDIS Volume 2 can provide a starting point
description of how human activities can lead to stressors and biological effects.
4 For more information, visit: http://www3.epa.gov/caddis/ssr  home.html. Accessed February 2016.
                                                                                            A-2

-------
                    Appendices to A Practitioner's Guide to the Biological Condition Gradient
                                                                                         February 2016
Appendix A-l. Conceptual scenarios for stress related changes in the major environmental factors that influence biological condition (see
Chapter 5, Figure 19). The scenarios describe potential changes in the factors in two regions in the United States: a humid-temperate and an
arid region. The stressor levels are qualitative and used only to describe relative differences in magnitude. Both local and watershed scale
factors are important for determining the condition of streams. Note that under the BCG conceptual model, alterations to the factors of flow
regime, water quality, energy source, and physical habitat structure represent increased stressors or categories of stressors (attributes of the
GSA (BCG x-axis)), whereas alterations to biotic interactions are included in the BCG y-axis attributes (e.g., attribute VI (non-native or
intentionally introduced species)).
 o
 ce.
 ce.
 LLJ
 o.
 D
 X
        Stressor Level
            No/
            Low
     Flow Regime
Within the naturally
occurring range, includes
floods & low flows at
natural  rates, intervals,
and extent; High
connectivity with ground
water maintained
         BCG X-Axis
      Water Quality
Within the naturally
occurring range, or with only
minimal increase in nutrients
& sediments, including flood-
related turbidity & summer
warming (linked to season
and antecedent moisture
conditions); no point sources
of nutrients or toxic
substances; usually cool or
cold & dissolved oxygen (DO)
saturated

Within the naturally
occurring range, typically
rare, and no materials in
amounts toxic to aquatic
biota
   Energy Source       Physical Habitat Structure
Within the naturally
occurring range
expected for a
stream with a
particular channel
width; smaller
systems typically
dominated by
riparian woody
vegetation, unless
naturally
autochthonous
Within the naturally occurring
range for a stream of a particular
size & slope; typically, large
woody debris (LWD) abundant;
coarse substrate; overhanging
vegetation and undercut banks
are present
                                     BCG Y-Axis
                                  Biotic Interactions
Within the naturally occurring
range expected, e.g.,
anadromy and potamodromy;5
beavers common; aquatic
invasive species (AIS) non-
detrimental; anomalies (e.g.,
deformities, erosion, lesions,
and tumors (DELT)) due to
disease or parasitism absent or
infrequent; no or insignificant
historical range changes
' The terms "anadromy" and "potamodromy" relate to long distance river migrants. See glossary.
                                                                                                                                              A-3

-------
                     Appendices to A Practitioner's Guide to the Biological Condition Gradient
                                                                                               February 2016
       Stressor Level
          Medium
O
ce.
 .
LLJ
Q.
o

D
     Flow Regime
Flashy; greater maxima
and minima; increased
drought frequency; some
water withdrawals; low
to moderate amount of
wetland drainage;
damming may reduce
annual floods and
droughts; some
groundwater extraction
close to the floodplain
          BCG X-Axis
      Water Quality
Enriched with nutrients and
ions; turbidity may increase,
moderate diel warming; small
DO sags may occur but these
rarely violate criteria; point
sources of nutrients or toxic
substances minor or if they
exist are treated; fish kills
rare

Suspended and dissolved
materials in amounts rarely
toxic to aquatic biota, but
mercury or persistent organic
contaminants may become of
chronic concern to top
piscivoresdue to
bioaccumulation; sediment
contamination may be
detectable but not causing
effects in benthic biota
   Energy Source       Physical Habitat Structure
Autochthonous
production higher
than expected in
lower order streams;
filamentous algae
may be present
Reduced amounts of LWD in
channel; fines slightly to
moderately more abundant than
expected from stream power;
pool substrate moderately
embedded; reduced extent of
undercut banks, overhanging
vegetation, and habitat
complexity; some loss of pool
volume and pool/riffle
proportions may be altered
                                        BCG Y-Axis
                                    Biotic Interactions
Altered fish age structure from
fishing and stocking may
change predation and
competition with a significant
effect on native populations;
expected beaver populations
diminished; higher than
expected occurrence of DELT
anomalies due to parasitism or
disease; sensitive AIS may
dominate, tolerant AIS may be
present; minor to moderate
historical range alterations;
cosmopolitan species may
extend distributions further
upstream
                                                                                                                                                       A-4

-------
                     Appendices to A Practitioner's Guide to the Biological Condition Gradient
                                                                                                February 2016
       Stressor Level
            High
ce.
<
ce.
LLJ
Q.
D
X
     Flow Regime
Flashy; highly altered low
regimes lead to greater
risk of drought/flood;
mostly or entirely human
controlled runoff in
urban and agricultural
areas; water withdrawals
& impoundments,  if
present, fundamentally
alter the nature of the
ecosystem
          BCG X-Axis
      Water Quality
Highly enriched, turbid,
warm; large diel DO &
temperature changes;
chemical and point sources of
nutrients or toxic substances
inadequately treated or
overwhelmed by untreated
diffuse toxic pollution.  Dams
when present produce
altered thermal regime and
nutrient dynamics

Dissolved, suspended, or
sediment-associated
materials may reach
concentrations that are
chronically or acutely toxic to
biota or can affect growth &
reproduction; high to
extreme sediment
contamination; anomalies
when associated with toxic
impacts are abundant &
serious; fish consumption
advisories serious.
Pharmaceuticals and/or
personal care products
present in effluent at
concentrations high enough
to be routinely detected in
water and/or tissue
   Energy Source       Physical Habitat Structure
Secondary
production sustained
mostly by
autochthonous or
imported fine
particulate organic
matter or dissolved
organic matter;
water may be too
turbid for benthic
filamentous algae to
develop. Strong diel
periodicity in
dissolved oxygen
concentrations,
including night-time
anoxia
Simplified or manmade,
straightened and/or leveed;
wood, undercut banks, &
overhanging vegetation absent
or non-functioning; rubble &
trash common, substrates highly
armored or embedded;
sedimented with sand or silt;
aquatic macrophytes missing or
extremely rare; riparian habitats
reduced or destroyed; dam
impoundments often present
                                        BCG Y-Axis
                                     Biotic Interactions
Dominated by transient fishes
or tolerant AIS; historically
common keystone species
extirpated and once-common
species now threatened,
endangered, or extirpated
from large portions of their
historical ranges due to
changes in predation and
competition relationships
and/or alteration of physical
habitat structure by stocked
fish; beavers transient or
absent
                                                                                                                                                         A-5

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
February 2016
BCG X-Axis BCG Y-Axis
Stressor Level Flow Regime Water Quality Energy Source Physical Habitat Structure Biotic Interactions














0
ce.
LLJ
(J
Q
ce.















No/
Low


















Medium














Within the naturally
occurring range, or only
slightly altered, includes
floods & low flows at
natural rates, intervals,
and extent; floods flashy;
annual scouring flows;
high connectivity with
ground water











Altered, increasingly
flashy; increased drought
frequency; some water
withdrawals and wetland
drainage; flow
alterations mitigated to
some extent by
environmental flow
releases; some
groundwater extraction
close to the floodplain;
dams may be present



Within the naturally
occurring range, with only
minimal increase in nutrients
& sediments, includes flood
turbidity & summer warming;
depending on soils, may be
naturally saline or alkaline;
relative ionic concentration
affected by evaporation
(Griffith 2014); enriched
where beaver present; ash
from 5-20 year fire cycles; no
point sources
Within the naturally
occurring range, typically
rare, but may be natural
sources of arsenic &
selenium; No toxics in
amounts toxic to aquatic
biota
Enriched, warmer & saltier,
turbid at low flows, small DO
sags; point sources if present
with treatment. No fish kills

No acute toxicity is observed,
but chronic toxicity is
possible due to
bioaccumulation. Fish
consumption advisories likely
for sensitive populations.
Low concentrations of
personal care products and
Pharmaceuticals observed in
wastewater receiving waters
Within the naturally
occurring range,
varies with channel
width, typically
dominated by
riparian woody
vegetation in small
unconstrained
channels;
heterotrophic &
autochthonous in
wider systems








Mostly
allochthonous, but
increasingly
autochthonous in
narrow streams;
wide streams
heterotrophic or
autochthonous with
increasing amounts
of filamentous algae





Within the naturally occurring
range, varies with geology,
substrate, flow, size, slope, soil,
latitude, elevation, & orography;
relatively stable riparian
vegetation, LWD in flats














Minor amounts of incision,
widening, or shallowing; reduced
LWD in channel; fines greater
than expected from stream
power; bed coarsening from
upstream dams; pool substrate
increasingly embedded; reduced
aquatic macrophytes, undercut
banks, & overhanging vegetation






Within the naturally occurring
range expected e.g.,
potamodromy; AIS absent or
non-detrimental. Intermittent
and ephemeral invertebrate
species traits are linked to flow
permanence, drought and
flood cycles (Bonada et al.
2007)











Altered fish age structure from
fishing and stocking may
change predation and
competition with a significant
effect on native populations;
AIS more common and
beginning to reduce
competitors & prey;
potamodromy reduced (Death
et al. 2009)





                                                                                                                         A-6

-------
                     Appendices to A Practitioner's Guide to the Biological Condition Gradient
                                                                                               February 2016
       Stressor Level
            High
O
ce.
LLJ
(J
I/)
o
     Flow Regime
Human controlled; large
inter-basin transfers;
ground water
overdrawn; effluent
dominated streams
below cities.
Highly altered
drought/flood regime;
droughts yield more dry
channels; withdrawals &
dams severely alter
nature of the ecosystem
          BCG X-Axis
      Water Quality
Highly enriched, turbid,
warm; large diel DO changes;
effluent dominated; point
sources may be inadequately
treated or overwhelmed by
untreated diffuse toxic
pollution; dams produce
altered thermal regime

Toxics may be present in
chronic or acutely toxic
amounts; bioengineered
chemicals can affect growth
& reproduction; high to
extreme sediment
contamination; fish
consumption advisories
serious. Pharmaceuticals
and/or personal care
products present in effluent,
water, and/or tissue
   Energy Source       Physical Habitat Structure
Mostly
autochthonous or
imported fine
particulate or
dissolved organic
matter; filamentous
algae common if
turbidity allows it
Largely manmade, straightened,
and/or leveed; little or no LWD,
undercut banks, or overhanging
vegetation; highly sedimented
with sand or silt; sand bottom
streams/rivers dominated by silt;
construction rubble & trash
common; aquatic macrophytes
missing or extremely rare;
riparian habitats reduced or
destroyed
                                        BCG Y-Axis
                                    Biotic Interactions
Assemblages dominated by
tolerant species including
tolerant AIS; once-common
species now threatened,
endangered, or extirpated
from large portions of their
historical ranges due to
changes in predation,
competition, and/or behaviors
by stocked fish; potamodromy
rare and erratic
                                                                                                                                                        A-7

-------
                     Appendices to A Practitioner's Guide to the Biological Condition Gradient
                                                                                              February 2016
Appendix A-2. Potential impacts of climate change at low/medium and high stress levels
                                                            ^FT7!
 O
 ce.
 LLJ
 (J
 I/)
 o
        Stressor Level
       Low to Medium
     Flow Regime
Depending on the
region, climate change
likely to increase drought
frequency and duration
(leading to greater
intermittent flows);
greater incidence of
intermittency in
headwaters; increased
frequency and
magnitude of floods due
higher incidence of
intense storms; overall
higher variability in
streamflow
             BCG X-Axis
        Water Quality
Degraded water quality due to
climate change impacts may be
observed as a result of increased
loading due to flood frequency
and duration. Low flows may
result in higher water
temperatures and episodic low
oxygen levels, increasing rates of
some geochemical processes.
Lower nutrient loading and
export rates expected during low
flow

Climate change impacts  may
result from watershed loading
during floods. Low flows may
result in episodic low oxygen
levels, increasing rates of some
geochemical processes
    Energy Source
Climate change impacts
(i.e., extended drought
cycles) may negatively
impact riparian systems
leading to reduced
shading, disruptions in
input of CWD and
particulate organic
matter (POM),
greater autochthonous
production
 Physical Habitat Structure
Climate change is expected
to reduce the extent of
coldwater habitat, reduce
stream shading as
macrophytes, overhanging
vegetation, and riparian
vegetation are reduced.
Increases in flood frequency
and intensity will alter
channel structure except in
confined channels or
bedrock dominated systems.
Increased potential for
increased debris flows in
mountain environments
with increased fire
frequency (Braune et al.
2008; Cannon et al. 2010)
        BCG Y-Axis
    Biotic Interactions
Climate change will result in
range contractions of
coldwater species and
expansion of cool and warm-
water species. Biodiversity
shifts are likely due to shifts
in energy sources,
availability of nutrients, and
changes in food webs at low
flow. Invertebrates may
experience increased
voltinism. Increased
incidence of drought may
accelerate incidence of
intermittency leading to
reduction in secondary
production (Freeman et al.
2007). The increased length
of the growing season will
increase yearly algal and
cyanobacterial biomassand
production
                                                                                                                                                       A-8

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
February 2016
BCG X-Axis BCG Y-Axis
Stressor Level Flow Regime Water Quality Energy Source Physical Habitat Structure Biotic Interactions





O
ce.
2
LLJ
(J
I/)
O

D













O
ce.
2
LLJ
(J
I/)
o
DC








High

















Low to Medium




















Climate change expected
to lead to increased
frequency and duration
of drought in some
regions, leading to
increased incidence of
intermittency; increased
incidence of extreme
storms









Climate change likely to
increase drought
frequency, magnitude of
floods. Winter
precipitation is predicted
to increasingly fall as
rain, increasing
likelihood of rain-on-
snow events, leading to
less infiltration, and
lower snowpacks. Earlier
snow melt and lower

and longer summer
baseflows will shift
seasonal flow patterns.
Fire frequency and
intensity may influence
flow regime, with post-
fire peaks observed
(Benda etal. 2003)
Climate change impacts
expected as a result of increased
drought and flood frequency and
duration likely will exacerbate
anthropogenic disturbances. Low
flows may result in low oxygen
levels and increased water
temperature, increasing rates of
some geochemical processes

In some regions, climate change
impacts expected as a result of
increased drought frequency and
severity. Likely acceleration of
mercury methylation and
increased bioavailability of toxic
metals. Mobilization of toxic
chemicals likely during floods
Climate change impacts may be
observed as a result of increased
drought and flood frequency and
duration. Low flows may result in
low oxygen levels, increasing
rates of some geochemical
processes. Post-fire floods may
contribute additional sediment
to channel; impacts of fire on
riparian vegetation may produce
elevated temperatures (Mahlum
etal. 2011)


Climate change impacts may
result from watershed loading
during floods. Low flows may
result in low oxygen levels,
increasing rates of some
geochemical processes

Climate change impacts
(i.e., extended drought
cycles) may negatively
impact riparian systems
reducing or eliminating
sources of CWD and
POM











Climate change impacts
(i.e., extended drought
cycles) may negatively
impact riparian
systems, disrupting
input and retention of
LWDandPOM.
Increased frequency
and duration of channel
drying will reduce litter
breakdown rates (Corti
etal. 2011; Datryetal.

2011)







Effects of climate change
due to intense storms and
more frequent and intense
drought may exacerbate
effects of anthropogenic
disturbances












Climate change is expected
to reduce the extent of
coldwater habitat, reduce
stream shading as
macrophytes, overhanging
vegetation, and riparian
vegetation are reduced.
Post-fire flood sediment
contributions may alter
riverine habitats











Climate change impacts are
less evident due to
simplified assemblages and
reduced richness (Durance
and Ormerod 2007, 2009).
Increased incidence of
intermittency in headwaters
may further simplify
assemblage structure and
lead to further reductions in
secondary production
(Freeman etal. 2007).
Temperature sensitive taxa
also are vulnerable to
organic pollution (Hamilton
etal. 2010)


Climate change may result in
range expansion and
contractions, with expected
losses in coldwater habitat
and species and expansion
of cool and warm-water
species (Comte et al. 2013).
Desiccation-sensitive taxa
drop out with increased
intermittence (Arscott et al.
2010). Invertebrate density,
richness, and condition

metrics decline with reduced
taxon richness lower in
response to reduced
discharge




                                                                                                                         A-9

-------
                     Appendices to A Practitioner's Guide to the Biological Condition Gradient
                                                                                               February 2016
       Stressor Level
            High
ce.
<
o
ce.
     Flow Regime
Climate change expected
to lead to increased
frequency and duration
of drought, and
increased incidence of
extreme storms
             BCG X-Axis
        Water Quality
Climate change impacts
expected as a result of increased
drought and flood frequency and
duration likely will exacerbate
anthropogenic disturbances. Low
flows may result in low oxygen
levels and increased water
temperature,  increasing rates of
some geochemical processes

In some regions, climate change
impacts expected as a result of
increased drought frequency and
severity. Likely acceleration of
mercury methylation and
increased bioavailability
    Energy Source
Climate change impacts
(i.e., extended drought
cycles) may negatively
impact riparian systems
reducing or eliminating
sources of CWD and
POM
 Physical Habitat Structure
Effects of climate change
due to intense storms and
more frequent and intense
drought likely to exacerbate
effects of anthropogenic
disturbances
        BCG Y-Axis
    Biotic Interactions
Climate change exacerbates
anthropogenic disturbances.
Reduced discharge has less
effect on invertebrate
community structure and
function (Death et al. 2009)
                                                                                                                                                      A-10

-------
                  Appendices to A Practitioner's Guide to the Biological Condition Gradient
February 2016
Appendix A-3. Potential indicators for stressors associated with altered flow, material transport, channel structure, and riparian/watershed
structure. The indicators of stress will differ depending on which environmental factor is being considered (e.g., flow regime, water quality,
energy, physical habitat structure, biotic interactions). Ideally, indicator scores are benchmarked to undisturbed or minimally disturbed
conditions for a given region or channel type, and stressors represent a departure from the natural range of variation. Biotic responses can be
related to these stressor indicators (shaded column).

Major factor

Flow Regime















Water Quality








Potential Stressor Indicators Listed for Different Environmental Process
Flow Alteration

• Frequency of low and
high flow events
• Annual flow variability
• Fleshiness (change in
flood peak and
duration)
• Base flow rate
• Stream power
• Timing of peak flow







• Temperature regime
• DO regime
• Flooding linked to
sediment & nutrient
loads
• Flooding linked to
contaminant loading


Material Transport

• Stream power
• Fleshiness (changes
in flood peak and
duration)
• Particle size
distribution
• LWD dam density
• LWD transport
• Organic matter (OM)
transport
• Erosion rates
• Streambed stability
• Turbidity



• Sediment loading
rates
• Sediment bound
metals,
contaminants
• Nutrient
concentrations


Channel Structure

• Discharge
• Hydraulic storage in
catchment
• Hydraulic storage in
flood plain
• Change in transient
storage capacity
• Pool-riffle structure
• Erosion rates
• Evidence of active
erosion
• Streambed stability
• Nutrient spiraling
rate


• DO regime
• Temperature regime
• No direct indicators






Riparian /
Watershed
Structure
• Peak flow
• Flood-mediated
sediment, OM, and
nutrient deposition
in floodplain











• Sediment/nutrient
loading rates
• Water temperature
• Contaminant loading
from upland or
riparian
• Contaminant bound
sediment from
roads, parking lots
Biological Responses
Biotic Structure/
Function

• Macrophyte density
• Dominance of flow-dependent
taxa
• Periphyton biomass
• Fish species traits associated with
flow regime (e.g., rheophils and
nonguarding lithophils &
lithopelagophils are replaced by
residents, generalists, and
polyphils)
• Invertebrate traits associated
with flow regime (= low crawling
rate, short adult life span,
erosional rheophily, med size at
maturity, cool/cold thermal
preference
• Increased biochemical oxygen
demand
• Increased
nitrification/denitrification
• Algal bloom
• Rates of critical biogeochemical
processes (e.g., mercury
methylation)

                                                                                                                                A-ll

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
February 2016

Major factor

Energy Source






Physical Habitat
Structure






Biotic Condition
& Interactions
(Responses)








Potential Stressor Indicators Listed for Different Environmental Process
Flow Alteration

• OM decomposition
rates
• OM input and
retention



• % fines
• Armoured substrate
• Pool/riffle sequence
• OM and LWD input and
retention
• Flood height
• Width:depth

• % rheophilic fish taxa
• Biotic condition scores
(invertebrates)
• Invertebrate diversity
and functional feeding






Material Transport

• Particulate organic
matter/Dissolved
organic matter
concentrations
• Nutrient
concentrations
compared to natural
• LWD density
• % fines
• % embeddedness
• Bed stability




• Dissolved and
particulate nutrient
uptake
• Primary/secondary
production rates
• Invertebrates in drift





Channel Structure

• LWD retention
• OM retention
• Solute retention




• Pool-riffle sequence
• Large woody debris
volume
• LWD input and
retention
• Bank erosion rates


• Assemblage
structure
• Primary and
secondary
production
• Rheophilic species
• Migrator fish
• Spawning habitat



Riparian /
Watershed
Structure
• OM quantity and
composition
• Primary production
rates
• Metabolism


• Floodplain
connectivity
• LWD storage
• OM storage
• Riparian
fragmentation


• Benthic metabolism
rates
• Food web structure
compared to natural
• Riparian buffer
fragmentation





Biological Responses
Biotic Structure/
Function

• Food web alteration






• Macrophytes/algal mats
• AIS dominance of species that
change structure (e.g.,
macrophytes, carp, zebra/quagga
mussels)
• Amount of overhead cover for
fish
• Noxious weeds
• AIS (e.g., native game fish decline,
hatchery fish increase)
• Native fish/benthos and riparian
vegetation and birds relative to
aliens
• Sensitive specialists compared to
tolerant generalists (birds, fish,
invertebrates, plants)
• Fish disease and anomaly rate
• Benthic metabolism,
decomposition rates
                                                                                                                       A-12

-------
                  Appendices to A Practitioner's Guide to the Biological Condition Gradient
February 2016
Appendix A-4. Examples of some fundamental environmental processes and materials (e.g., flow, material transport, channel structure,
riparian and watershed structure, biological interactions) that can be altered by human-induced disturbances (e.g., pressure indicators) and
create stressors (Figures 20, 21, and 26). Possible mechanisms for stressor production resulting from the altered processes are listed, as well
as examples of management actions to reduce the stress.
Pressure Indicators Mechanism for Stressor Production Management Actions to Reduce Stress
FLOW ALTERATION
ALTERATION OF MATERIALS
TRANSPORTED
% impervious area
Road density
% urban area
Population-density
Storm sewer miles
# diversions per catchment; quantity of water
diverted
# of dams; cumulative volume of water in
reservoirs
Proportion of river length channelized
#dams, # diversions
Point source discharge constituent levels; # point
source discharges per catchment; density of
point source discharges
Length (in km) of intact riparian buffers
% impervious surfaces, population density
Acceleration of water flow; reduction of infiltration
and groundwater recharge; hot pavement and
reduced retention times increases stream
temperatures
Increases impervious surface; increases number of
roads crossings streams and culverts
Reduces groundwater infiltration; increases peak
flows; decreases peak flow duration
Increases impervious surface
Increased potential for stormwater overflows
Decreased base flow; increased intermittency in
headwater streams
Impacts natural flow regime by depressing flood
height & duration; maintains base flow downstream
but may prevent fish migration upstream
Reduced base flow
Flow alteration; increased sediment retention
upstream; sediment pulses when gates opened
Increased discharge of pollutants, including excess
nutrients, toxic materials, and particulates from point
sources; enhanced flow increases erosion
Erosion of surface solutes, sediments, and warmer
water; reduced riparian canopy increases algal
biomass, increases nutrient uptake, enhances benthic
production
Eroded material adds carbon and nutrients that
increase biological activity; increased contaminants
Reduce impervious surface; install pervious
pavements
Implement low-impact development strategies
Plant trees in headwaters; urban forests and parks;
restore urban streams
Restore riparian and floodplain vegetation
Restore natural flow regime; move storm sewers
out of stream beds
Use appropriate culvert type and size
Restore connectivity to floodplain; fish ladders
Install flood retention structures
Restore natural flow regime
Water quality management actions (e.g., permits);
sediment retention basin
Riparian restoration
Low-impact development
                                                                                                                              A-13

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
February 2016
Pressure Indicators Mechanism for Stressor Production Management Actions to Reduce Stress
ALTERATION OF MATERIALS TRANSPORTED
# road crossings per catchment; road density
% row crops
Atmospheric deposition
CAFO size and density
Area (in km2) of tile drains
Area of catchment logged; length of logging
roads
# mines per area; area of valley fill
# acres of irrigated cropland with no BMPs
implemented
# quarries per catchment
Acres of drained wetlands
Length of armored channel
Increased road density and numbers of road crossings
may result in episodic, high volume flow events that
erode stream banks and produce increased pollutant
(e.g., metals, oils) and sediment loading
Nutrient and pesticide applications and irrigation
contributes to contaminated runoff and increased
sedimentation
Increased nitrogen, carbon, and sulfur loading
Manure results in increased nutrient, pathogen,
pesticide, antibiotic, and sediment loading
Enhance discharge especially during storms
Enhanced surface runoff from bare soils
Enhanced windborne fines and surface-derived
sediments, salts and metals, acid drainage
Elimination of natural steams and forest habitats
Increase in surface runoff, sedimentation,
eutrophication, with higher levels of pesticides,
herbicides, solutes leached from soils (e.g., salts,
selenium)
Sediment loading
Storm storage
Increased stream power; reduced erosion potential
Road maintenance; retention ponds; riparian
habitat restoration; road placement
Buffer strips
Substitute cleaner fuels; implement "clean"
combustion technologies; implement non
combustion energy production and use (e.g., solar,
wind)
Composting; appropriate manure application;
retention structures; reduced antibiotic and
pesticide management practices
Two-stage ditch
Buffer strips; use temporary road crossings in
winter; select cutting replacing clear-cutting
Sediment retention basins
Alter mining practices (e.g., mountain top removal
and valley fill practices to traditional underground
mining practices)
Drip irrigation based on soil moisture levels
Buffer strip
Retention basin
Restore natural channel
                                                                                                                       A-14

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
February 2016
Pressure Indicators Mechanism for Stressor Production Management Actions to Reduce Stress
CHANGE IN CHANNEL STRUCTURE
CHANGE TO RIPARIAN AND WATERSHED STRUCTURE
Length (in km) of channelized stream
Stream length stabilized by riprap/concrete
# dams; volume in reservoirs
# diversions; volume of water diverted
Culvert density
Density of road crossings
Presence of valley fills; extent of valley fill
Length of levees per catchment
Length of intact riparian zone
Evidence of snagging of LWD
Evidence of connectivity disruption or artificial
connections established
Fragmentation of riparian zone
Riparian width
% shading
Levees
Tile number-drains/ditches
Length (in km) of streamside roads
Surface area of off-stream ponds or wetlands
Area of valley bottom grazing
Area of aggregate mining
Flow alteration; habitat loss
Hardening of shoreline alters flow and erosion
processes and simplifies habitat
Solute, sediment transport interrupted
Flow regime altered; base flow impacted
Flow disruption
Riparian alteration; sedimentation; flow alteration
Direct engineering activities; elimination of channel;
water quality impacts
Connection to floodplain disrupted
Loss of shading; loss of OM
Habitat loss
Habitat loss; vector for non-native invasive species
(NIS) established
Loss of natural vegetation cover leading to habitat
loss, increased sediment/nutrient input
Reducing, disturbing, or completely removing riparian
cover increases sedimentation and reduces habitat
and other effects
Increased solar insolation leading to greater algal
biomass results in greater daytime photosynthesis and
night-time respiration
Floodplain disconnect from river and prevent
replenishment of flow, nutrients, and sediments
Altered flow regime; increased nutrient input
Sedimentation
Loss of connectivity; loss of flood storage
Manure from cattle adds carbon and nutrients that
increase in-stream biological activity; cattle moving
through streambed results in habitat modification
Increased fine sediment loading
Restore natural channel shape and flow regime
Restore natural streambanks
Remove dams where appropriate
Restore natural flow regime
Install appropriate culvert type/size
Mitigate dust, employ proper drainage tactics
Stop activity; employ retention basins
Reconnect floodplain
Restore natural riparian vegetation
Restore natural wood structures
Restore connectivity or employ structures to
remove connectivity
Restore vegetation cover
Restore natural vegetation type and extent
Plant trees in riparian zone
Restore natural channel form; engineer flow,
nutrients and sediment delivery to mimic natural
regime
Two-stage ditch
Install buffer strip; pave road near stream
Restore connectivity
Fence pasture; bridges for cattle to cross streams;
manage manure
Install sediment basin
                                                                                                                       A-15

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
February 2016
Pressure Indicators Mechanism for Stressor Production Management Actions to Reduce Stress
CHANGE IN BIOLOGICAL
CONDITION AND ACTIVITY
#NIS
$ of baitfish sales
# fishing licenses issued
# of dams and reservoirs
Stocking programs; accidental introduction from
aquaculture facility
Habitat modification or negative biotic interactions by
invasive plants and fish
Over harvest
Prevent fish migration upstream
Education; enhanced facility inspections; programs
to extirpate NIS
Education; boat inspections
Harvest limits; education; reduce number of
licenses
Restore connectivity fish ladders
                                                                                                                       A-16

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                     February 2016
Appendix A References

Arscott, D.B., ST. Larned, M. Scarsbrook, and P. Lambert. 2010. Aquatic invertebrate community
      structure along an intermittence gradient: Selwyn River,  New Zealand. Journal of the North
      American Benthological Society 29(2):530-545.

Benda, L, D. Miller, P. Bigelow, and K. Andras. 2003. Effects of  post-wildfire erosion on channel
      environments, Boise River, Idaho. Forest Ecology and Management 178:105-119.

Bonada, N., M. Rieradevall, and N. Prat. 2007. Macroinvertebrate community structure and biological
      traits related to flow permanence in a Mediterranean river network. Hydrobiologia 589:91-106.

Braune, E., O. Richter, D. Sondgerath, and F. Suhling. 2008. Voltinism flexibility of a riverine dragonfly
      along thermal gradients. Global Change Biology 14:470-482.

Cannon, S.H., J.E. Gartner, M.G. Rupert, J.A. Michael, A.M. Rea,  and C. Parrett. 2010. Predicting the
      probability and volume of post-wildfire debris flows in the intermountain western  United States.
      Geological Society of America Bulletin 122(1/2):127-144. doi: 10.1130/B26459.1.

Comte, L., L. Buisson, M. Daufresne, and G. Grenouillet. 2013. Climate-induced changes in the
      distribution of freshwater fish: Observed and predicted trends. Freshwater Biology 58:625-639.

Corti, R., T. Datry, L. Drummond, and ST. Larned. 2011. Natural variation in immersion and emersion
      affects breakdown and invertebrate colonization of leaf litter in a temporary river. Aquatic
      Sciences 73(4):537-550. doi: 10.1007/s00027-011-0216-5.

Datry, T., R. Corti, C. Claret, and M. Philippe. 2011. Leaf litter decomposition along a gradient of flow
      permanence in a French temporary river: The memory of drying. Aquatic Sciences 73:471-483.
      doi: 10.1007/S00027-011-0193-8.

Death, R.G., Z.S. Dewson, and A.B.W. James. 2009. Is structure  or function a better measure of the
      effects of water abstraction on ecosystem integrity? Freshwater Biology 54(10):2037-2050.

Durance, I., and S.J. Ormerod. 2007. Climate change effects on  upland stream macroinvertebrates over a
      25-year period. Global Change Biology 13:942-957.

Durance, I., and S.J. Ormerod. 2009. Trends in water quality and discharge confound long-term warming
      effects on river macroinvertebrates. Freshwater Biology.  54(2):388-405. doi:10.1111/j.l365-
      2427.2008.02112.x.

Freeman, M.C., C.M. Pringle, and C.R. Jackson. 2007. Hydrologic connectivity and the contribution of
      stream headwaters to ecological integrity at regional scales. Journal of the American Water
      Resources Association 43(1):5-14.

Griffith, M.B. 2014. Natural variation and current reference for specific conductivity and major ions in
      wadeable streams of the conterminous USA. Freshwater Science 33(1):1-17.
                                                                                          A-17

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                      February 2016


Hamilton, AT., J.D. Stamp, and B.G. Bierwagen. 2010. Vulnerability of biological metrics and multimetric
      indices to effects of climate change. Journal of the North American Benthological Society
      29(4):1379-1396.

Mahlum, S.K., LA. Eby, M.K. Young, C.G. Clancy, and M. Jakober. 2011. Effects of wildfire on stream
      temperatures in the Bitterroot River Basin, Montana. International Journal ofWild/and Fire
      20:240-247.

Norton, D.J., J.D. Wickham, T.G. Wade, K. Kunert, J.V. Thomas, and P. Zeph. 2009. A method for
      comparative analysis of recovery potential in impaired waters restoration planning. Environmental
      Management 44:356-368.
                                                                                           A-18

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                     February 2016
Appendix B: Examples of Development of the Biological Condition
Gradient for Large Rivers, Coral Reefs, and Estuaries

This appendix includes examples of work underway on development and application of the conceptual
Biological Condition Gradient (BCG) framework to large rivers, estuaries, and coral reefs. These
examples illustrate how the BCG framework may be refined for different aquatic systems. The case
studies are included here to generate discussion and share information among state water quality
program managers and scientists interested in applying to BCG to water bodies other than streams and
wadeable rivers. The author's name and affiliation are included for each case study. Contact information
is included for the primary author.

The following case examples are included:
Bl. Upper Mississippi River: Development of a Biological Condition Gradient for Fish Assemblages of the
Upper Mississippi River and a "Synthetic" Historical Fish Community (page B-2)

B2. Narragansett Bay: Development of a Biological Condition Gradient for Estuarine Habitat Quality
(page B-25)

B3. Caribbean Coral Reefs: Benchmarking a Biological Condition Gradient for Puerto Rican Coral Reefs
(page B-53)

B4. New England Rivers: Using the Biological Condition Gradient and Fish Index of Biotic Integrity to
Assess Fish Assemblage Condition in Large Rivers (page B-82)
                                                                                       B-l

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                     February 2016


Bl.  Upper Mississippi  River: Development of a Biological Condition
Gradient for Fish Assemblages of the Upper Mississippi  River and a
"Synthetic" Historical Fish Community

Ed Rankin, Midwest Biodiversity Institute, Columbus, Ohio6

Bl.l Background

As with streams and wadeable rivers, the BCG framework may also be applied to non-wadeable rivers to
help assess attainment of aquatic life use (ALL)) goal conditions,  identify high quality waters, set
incremental biological goals for environmental improvements, and track progress in achieving the
improvements. Typically, regional reference conditions are used to empirically derive numeric biological
thresholds to assess ALL) attainment for streams and rivers where sufficient reference sites exist
(Hughes et al. 1986; Stoddard et al. 2006). However, for more complex and larger river systems (e.g.,
large and great rivers), the extent and types of historical alterations of these waters makes the regional
reference condition approach difficult (Angradi et al. 2009a).7

Large and great rivers are frequently modified by dams, levees, flow controls, water diversions, water
withdrawals, and chemical impacts (e.g., effluents, runoff). Conditions that are considered comparable
to undisturbed and minimally disturbed conditions (Stoddard et  al. 2006), do not exist for these systems,
and the Upper Mississippi River (UMR) (Figure Bl-1) has a  long history of such human alteration
(Alexander et al. 2012). Biological multimetric indices used in biological assessments are typically
derived using data from least disturbed reference sites and/or stressor response data. Three multimetric
indices applicable to large rivers include: the Great River Fish Index (GRFIn) (Angradi et al. 2009); the
Fish Assessment Community Index (FACI) (Emery et al. 2007), and the Ohio Continuous Index of
Biological Integrity (Ohio CIBI) (Rankin 2010). The GRFIn is  based on a stressor derived reference
condition, the FACI is based on a method that uses all the data in a continuous scaling approach for
calibration (Blocksom  2003), and the Ohio CIBI is based on a  regional reference site approach (least
disturbed).

This case study explores a synthetic modeling approach (Armitage et al. 2009) for using historic data to
model a quantitative description of BCG level 1 and 2 conditions—incorporating historical ecology  (e.g.,
McClenachan et al. 2015) with commonly used assessment methods. BCG level 1 and 2 conditions  may
currently not be achievable in many large rivers, but knowing the characteristics of a biotic community
that would be supported under these conditions may assist in defining  incremental and sustainable
biological goals for water quality improvements.  The synthetic modeling approachis applied here in
conjunction with the GRFIn, FACI, and Ohio CIBI to examine how historical  data can help define a
trajectory toward restoration of all, or elements of, a historic fish assemblage in a highly modified
riverine system.
6 ERankin@mwbinst.com
7 The definition of great rivers has differed among investigators, but Angradi et al. (2009b) considered great rivers
as those with catchments > 1,000,000 km2, which would include the three mid-continent rivers included in the
Environmental Monitoring and Assessment Program (EMAP) Great Rivers Evaluation (GRE), the Mississippi,
Missouri, and Ohio Rivers.
                                                                                         B-2

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                      February 2016
B1.2  Upper Mississippi River Biological Assessment Initiative

Currently, ALL) assessments are conducted independently by each of the five states bordering the UMR,8
using different methods and assessing against different thresholds. Yoder et al. (2010) summarized the
variety of methods used to assess ALL) attainment by Illinois, Iowa, Minnesota, Missouri, and Wisconsin.
Furthermore, ALL) assessments are based primarily on chemical and physical water quality data
collected at widely separated fixed stations. The Upper Mississippi River Basin Association (UMRBA),
through its Water Quality Task Force (WQTF), sponsored a project in 2009-2011 to develop a Clean
Water Act (CWA) Biological Assessment Implementation  Guidance Document for the interstate UMR.
This document (Yoder et al. 2011) focused on how to integrate UMR-specific biological assessment
approaches into the water quality management programs of the UMR states. It also provided technical
methods on conducting biological assessments of the UMR and guidance on how to integrate these
methods into the water quality management programs of the UMR states. Additionally, a detailed
analysis of existing biological assessment data collected by EPA's Great Rivers Evaluation (GRE) and the
U.S. Army Corps of Engineers' Long-Term Resource Monitoring programs was undertaken (Miltner et al.
2011) to develop biological assessment thresholds for the UMR.9

Historical knowledge about large and great river fish assemblages is essential in the development of
contemporary measures of biological condition and a determination of what thresholds might  be
attainable. Such information can be obtained from  the accounts of pioneering naturalists, settlers, or
from early fisheries accounts about these systems (e.g., Trautman 1981; Steuck et  al. 2010). Native
American middens and  fossils and subfossils can also provide historical evidence of fish species
occurrence and distribution in these  systems (Lyman 2006; Humphries and Winemiller 2009). A
description of the historical changes  that have occurred in the UMR fish assemblage is available in Pitlo
and Rasmussen (2004).  These accounts provide important insights about the great river fish
assemblages that occurred prior to the extensive alterations of the 19th and 20th centuries.

Some large and great river data sets  in the Midwest United States now have at least a 20-year
accumulation offish assemblage data paired with chemical, physical, habitat, and other stressor data.
Gradients of ecological  sensitivity can be extracted  for many fish species from these data sets by
examining probabilities of occurrence along chemical, physical, and biological stressor gradients. The
combination of contemporary data on species distributions along stressor gradients can be combined
with historical accounts of rare, extirpated, or even extinct species to reconstruct "synthetic" historic
fish assemblages (Armitage et al. 2009). The resulting model can then be used to "back cast" assemblage
condition-stressor relationships to better inform goal setting for a river system.
 Iowa, Illinois, Missouri, Minnesota, Wisconsin
9 The UMR in this study extended from the confluence with the Ohio River at Cairo, IL upstream to the upstream most lock and
dam in Minneapolis, MN.
                                                                                            B-3

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
February 2016
                               Upper Mississippi  River Basin

                                                                          Copyright: ©2013 Esn. DeLorme, NAVTEQ
Figure Bl-1. Map of the UMR Basin.

B1.3  Methodology for Developing a Synthetic Fish Assemblage

To extrapolate fish species and abundances during a historical pre-disturbance10 time period in the
UMR, large river data sets from the Midwest were used to estimate (1) the frequency of occurrence of a
species by biological condition range based on existing fish IBIs; and, (2) the relative catch rates
(numbers/km) using boat electrofishing methods for each species. This information was then combined
with the historical fish distribution information of Steuck et al. (2010) that includes the historical and
present occurrence of fish species in the upper impounded reach, lower impounded reach, and
unimpounded reach of the UMR. Known life history information and descriptions of UMR fish
populations from historical records (e.g., Carlander 1954) were used to derive extrapolated catch
frequencies and abundances that likely occurred prior to the major alterations such as creation of
impoundment by navigational dams and modification of the open river reach by levees  and wing dams.
These frequencies and estimates of abundance were then used to create a "pool" of fish to "sample"
using a random selection process. Ten iterations were performed for each of the three UMR reaches,
and the data were used to calculate the FACI, the GRFIn for the impounded and open river reaches of
the UMR, and the Ohio CIBI for beatable rivers (Table Bl-1). The steps in this process are summarized in
Table Bl-2. In addition to the modeling of pre-settlement conditions, researchers also used early fish
  "Pre-disturbance" conditions reflect free-flowing conditions during which early fish distribution patterns were recorded by
pioneering naturalists and early settlers, as well as investigated using evidence in Native American middens.
                                                                                            B-4

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
                                                                                   February 2016
data from large rivers of Ohio that had very poor conditions to model a "very poor" fish assemblage that
might have occurred during the 1960s-1980s prior to CWA point source pollution control mandates. For
the calculation of the GRFIn, researchers used the current and historical data to recalculate metric
ceiling and floor values (95th and 5th percentiles).
Table Bl-1. Fish indices applied in the UMR Basin.
Index
Acronym
FACI
GRFIn
CIBI
Description
Fish Assessment Community Index— A multimetric fish assemblage index
developed from sites sampled during the Regional EMAP Large Rivers project
Great River Fish Index — A multimetric index created by EMAP-GRE, developed
specifically for the UMR and the lower Missouri River
Continuous Index of Biotic Integrity — A continuous scoring form of the Ohio Fish IBI
created to improve the scoring sensitivity of the original Ohio IBI and to provide the
ability to score historical fish assemblages not truncated to current conditions
Citation
Emery etal. (2007)
Angradi etal. (2009a)
Rankin (2010)
Table Bl-2
conditions
:. Steps in the development of a synthetic fish assemblage that approximates pre-settlement
i in the UMR.
   Step
                                       Activity Description
           Compile historical fish assemblage list for reaches of the UMR (i.e., upper impounded, lower impounded, and
           open river reaches).
            Use existing data to determine response in abundance and probability of occurrence of each species at biological
            condition ranges (Very Poor, Poor, Fair, Good, Excellent; for regionally relevant fish IBIs).
            Estimate typical relative abundance in catch when trend is extrapolated to pre-settlement conditions (use trends
            from step 2 along with life history information, historical descriptions of occurrence, abundances recorded
            elsewhere, etc.). Do separately for upper impounded, lower impounded, open river.
            For rare, extirpated, or extinct species, estimate abundance during pre-settlement periods using historical
            descriptions, life history information, abundances recorded elsewhere, etc. Do separately for upper impounded,
            lower impounded, open river.
           Create "population" of > 100,000 fish for "sampling" by multiplying for each species by the probability of
           occurrence x the average estimated abundance x 1000.
            Begin random selection process for "fishing" historical "synthetic" pool of fish—10 iterations for each reach of
            UMR.
            Randomly select among best large rivers in Ohio/Indiana data to define maximum abundance and species
            richness for each iteration; cap richness at randomly selected site +5 and abundance at relative number/km +
            500
            For each iteration, randomly select, without replacement, individuals until species and abundance caps reached.
            For each iteration a "sampled" assemblage is created, which is scored with appropriate GRFIn, FACI, and Ohio
            CIBI.
                                                                                                        B-5

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                     February 2016


Bl.3.1 Inferring Stressor Levels from Species Assemblages
Miltner et al. (2011) employed multivariate and correlative measures using the existing GRFIn and FACI
indices and other measures to identify limiting stressors to the fish assemblages in the UMR (i.e., a "top-
down" approach). For this case study, an alternative approach used information about individual
species' responses to stressors gained from broad-scale studies of species sensitivities. The approach
used information to infer which stressors were most limiting, to understand the limiting nature of
stressors, and to predict species occurrences and distributions. By examining the inferred stressor levels
during historical  periods, one can begin to  understand which stressor or stressors might be limiting rare
species and estimate the feasibility of restoration from current conditions.

In employing this approach, Weighted Stressor Values (WSVs; Meador et al. 2008) were determined for
each species in the fish assemblage databases for boat electrofishing sites for purpose of ranking the
relative tolerance of fish species to different stressors. Tolerance Indicator Values (TIVs), which are the
ordinal ranks of WSVs (1-10) for each species and stressor (Meador and Carlisle 2007) were also derived
to place stressors on the same numerical scale.11 These values were then summed across all sites and
divided by the total abundance at all sites to arrive at a WSV.12 The TIVs were used to infer the stressor
level at a site based on the biological assemblage data that were collected, for example, where there
was incomplete stressor data. Grand mean TIV values were calculated by creating a mean across all
species weighted by the abundance of that species at a site. One goal of this analysis was to estimate
how stressor conditions varied between current and historical time periods.

Bl.3.2 Assumptions
If sampling were to occur in a riverine habitat with conditions close to "as naturally occurs," it is
assumed that sampling would occur along  the main channel border and the samples collected would
include currently rare or extirpated species from the backwater and side channel habitats. The same
assumption has been made by others to conclude that such sampling has been typically representative
of the conditions in the  backwaters and secondary channels (Angradi 2006; Thorp 1992).

B1.4  Extrapolation of Fish Assemblages to Historical Conditions in the Upper
Mississippi River

Researchers described historical condition in the UMR and extrapolated it to approximate  BCG level 1
and 2 conditions in order to determine the potential to restore UMR fish assemblages towards this
condition. The principal concept is illustrated in Figure Bl-2. The dark blue points in the  BCG levels 3-5
range represent the existing conditions in the UMR along a generalized stressor gradient. This stressor
gradient represents the cumulative stressor load that influences the current condition of the UMR. The
green and grey points in the BCG level 1-2 range reflect pre-settlement and immediate post-settlement
conditions in the UMR prior to its alteration for commercial navigation.
11 WSVs are derived for individual stressor variables (e.g., DO, pH, ammonia-nitrogen) as an average (or maximum
for toxicants) weighted by the abundance of a species at each site
12 Calculating TIV scores standardizes WSVs measured on different scales and allows averaging of the TIVs to create
a cumulative grand stressor rank across major stressor categories.
                                                                                            B-6

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                      February 2016
 >
       o
      o
      .S          Restoration
      _§*           Trajectory
       o
      m
          A               Level of Exposure to Stressors
Figure Bl-2. The BCG is used to depict the position of current-day UMR fish assemblage as measured by the
GRFIn (dark blue points) compared to the "as naturally occurs" pre-impoundment historical condition (green
points) that was approximated by the synthetic model. A restoration trajectory is apparent between these
conditions. The BCG can be used to define incremental biological improvements along that trajectory.

Using the template of the BCG, UMR fish species and/or other suggested measures were assigned to
each of the 10 BCG attributes (Rankin and Yoder 2011). The species assignments provide the probability
of capture and average relative abundance of each species with extrapolations to historical conditions.

Bl.4.1  Attribute I. Historically Documented, Sensitive, Long-lived, or Regionally
Endemic Taxa
Attribute I of the BCG is perhaps among the most influential in defining the characteristics of the UMR as
it naturally occurred. This attribute contains information not  only about a species occurrence, but also
age/size distributions of the long-lived species such as paddlefish, sturgeon, and muskellunge. Nearly all
of the species in BCG attribute I are sensitive to pollutants, habitat loss, and other alterations. Some
species, such as American eel, are tolerant of pollutants, but they were included because of their life
history requirements and to reflect ecological connectance. Because of their migratory habits, the
regular occurrence of this species would reflect that the UMR was well connected to the Gulf of Mexico.
For this study, the distinquishing characteristics for species assigned to this attributed include:  rare,
endemic and long lived.

Bl.4.2  Attribute II.  Highly Sensitive Taxa
The UMR species assigned to attribute II reflect a high level of sensitivity to stressors influencing large
Midwest rivers (Table Bl-3). While many of these species still occur in the UMR, many are currently
rarely observed but would likely be more common  in the UMR under reduced levels of stressors (e.g.,
reduced nutrient enrichment, unimpounded habitats, or connected to backwater and side-channel
habitats, Mac et al. 1998; Fremling et al. 1989).  This group especially reflects sensitivity to habitat
degradation and loss of connectivity (Mac et al. 1998).
                                                                                            B-7

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
February 2016
Table Bl-3. Sample list of fish species collected or reported from the UMR, by state for the first six
BCG attributes.
Species
Silver lamprey
American brook
lamprey
Chestnut lamprey
Paddlefish
Lake sturgeon
Shovelnose sturgeon
Pallid sturgeon
Alligator gar
Shortnose gar
Spotted gar
Longnose gar
Bowfin
Goldeye
Mooneye
Skipjack herring
Gizzard shad
Threadfin shad
Alabama shad
Central mudminnow
Grass pickerel
Northern pike
State Endangered,
Threatened, Extirpated, or
Special Concern
Designation
MN




SC









sc






Wl




sc







E

E






IA

T


E

E












T

IL




E

E
E













MO




E

E














BCG Species Attributes
1
Rare,
Endemic,
Long-lived



X
X

X
X













II
Highly
Sensitive

X*



X



X




X


X



III
Intermediate
Sensitive
X

X









X
X


X



X
IV
Intermediate
Tolerance








X

X
X






X*
X

V
Tolerant















X





VI
Non-
native





















E=endangered
T=threatened
SC=special concern
EX=extirpated
X=generally expected to occur; those species with an * associated with nearby smaller tributaries

Bl.4.3 Attribute III. Intermediate Sensitive Taxa
Intermediate sensitive species persist through the initial stages of increasing levels of stress. However,
they are generally at their highest abundances when stress is the lowest. These species are generally
numerically predominant in natural fish assemblages historically.

Bl.4.4 Attribute IV. Intermediate Tolerant Taxa
These are fish species that are not sensitive to moderate levels of most stressors and can become
predominant as more sensitive species (attribute I, II, and III species) are reduced with increasing stress.
Their mere presence suggests little about stressor levels at low to moderate levels of stress; however,
combined with the absence or reduction of attribute l-lll species, they can be indicative of high stressor
levels. Attribute I  fish species in the UMR are particularly sensitive to the loss of habitat, particularly
floodplain and backwater spawning and nursery habitats (Etnier and Starnes 1993).
                                                                                              B-8

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
February 2016
Bl.4.5 Attribute V. Tolerant Taxa
These species are especially tolerant to most stressors and will persist at increasing levels of stressors
above intermediate levels. At the very highest stressor levels, however, most of these species will be
reduced in abundance.

Bl.4.6 Attribute VI. Non-native or Intentionally Introduced Species
These are fish species that have either been introduced (intentionally or otherwise), and some are now
resident in the UMR. Certain of these species (e.g., silver and bighead carp) are potentially more
deleterious than others because of their disruption of the food web (Freedman et al. 2012). Most are
moderately to highly tolerant of chemical and physical stressors.

Bl.4.7 Attribute VII.  Organism Condition
Attribute VII measures the condition of individual organisms. Several commonly used biological metrics
can be used to gauge the condition of this attribute for fish in UMR large rivers (Table Bl-4). Most large
river fish assemblage programs use external anomalies to measure the degree of exposure to pollution.
Data on multiple year classes for species are generally available from size measurements, and a good
distribution of large, older year classes is generally indicative of good conditions. This would also
translate to a high diversity by numbers and weights and high indices that reflects these characteristics,
such as the Modified Index of Well-Being (Mlwb), which is readily available provided that biomass data
are collected. The Mlwb should be used as a complimentary index with a fish IBI (Yoder and Smith 1999).

Table Bl-4. Candidate measures of organism condition for attribute VII in the UMR BCG.
Name
External Anomalies
Multiple Year Classes
High diversity based on numbers and
weight
Description
Incidence of erosions, lesions, tumors, and deformities observed on fish in a sample.
A low incidence reflects sublethal and indirect stressors (e.g., excessive diel dissolved
oxygen (DO) variations); high incidence can reflect toxic conditions.
Populations of all expected year classes should exist for all species in attribute
groups /-///.
Use the Mlwb and its subcomponents based on numbers and weight.
Bl.4.8 Attribute VIII. Ecosystem Function
Great rivers in natural or close to natural conditions (e.g., undisturbed to minimally disturbed
conditions) support complex ecosystem functions that result in high diversity and abundance across the
various trophic guilds (Mac et al. 1998). Structural measures can potentially be used as surrogates to
infer the intactness of ecosystem function (Table Bl-5). Work is underway to explore use of candidate
surrogate measures (Table Bl-5).
                                                                                           B-9

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
February 2016
Table Bl-5. Candidate measures to infer ecosystem functioning in the UMR.
Metric
Invertivores
Top Carnivores
Omnivores
Description
The historical UMR was characterized by large numbers of specialized invertivores
that fed on the high diversity and production of aquatic invertebrates in multiple
habitat types.
The historical UMR supported a high diversity and biomass of top carnivores that fed
on abundant forage fish and other organisms supported by efficient energy cycling
through the system.
Omnivores were not predominant in the mainstem UMR given the abundant insects
and mussel assemblages that occurred in the river. A shift to predominance by
omnivores would reflect an alteration to nutrient inputs and cycling.
Bl.4.9 Attribute IX. Spatial and Temporal Extent of Detrimental Effects
Attribute IX is especially important for temperate floodplain rivers and the UMR in particular. The extent
of direct alterations to the UMR from the navigational impoundments and leveeing and wing dams in
the open river reach has been system-wide and affects the entirety of the interstate UMR. These
modifications have disconnected much of the mainstem from its former backwaters, modified the flow
regime, and altered the original riverine habitats to a more lentic (upper impounded) or channelized
(open river) condition. The modification of the original forested and wetland-dominated landscape by
row cropping has affected flow, habitat, and water quality. These stressors, especially habitat and flow,
are currently limiting to the recovery potential of the UMR fish assemblages. However, this study does
not explore approaches to quantify this attribute.

Bl.4.10 Attribute X. Ecosystem  Connectance
Attribute X relates directly to the ability of fish species in the UMR main channel to move laterally into
and out of adjacent backwaters, sloughs,  and oxbows that were once characteristic of the UMR. The
periodic but regular inundation of the floodplain to which many attribute l-lll species are adapted has
been altered by impoundment or leveeing of the  UMR. Many of the sensitive species that are now rare
or extirpated were associated with these  connected, but off-channel habitats. While fish can move
upstream and downstream, the ease with which this now takes place has been modified by the
navigational dams and the alteration of flows and riverine habitat. As with attribute VIM, work is
underway to explore use of biological information and attribute definitions as surrogate biological
measure for this attribute.

B1.5  Synthetic Assemblage  Results

The basis for deriving a synthetic historical fish assemblage is the observation that the probability of
capture and average abundance of a species is related to the array of stressors present  in a reach  and is
reflected in the  biological indices used in this study (e.g., GRFIn, Ohio CIBI, FACI). Researchers have used
this information to derive probabilities of capture and extrapolated abundances for the historical  period
prior to impoundment and a time period with poor to very poor water quality conditions caused by
untreated wastewater discharges (e.g., 1960s). Figure Bl-3 illustrates changes in the probability of
capture, relative abundance,  and abundance by capture rate from Ohio data for three key riverine
species in the UMR: the blue  sucker, river darter,  and black buffalo. The extrapolation to historical data
used to derive the pool of potential  fish for the historical IBI was developed using the trend of actual
data, the historical reports of occurrences and distribution (Steuck et al. 2010), and life  history
                                                                                          B-10

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                     February 2016


information and other historical sources that describe the general occurrence of these species in large
Midwest rivers prior to the anthropogenic impacts of the past two centuries. Because natural species
distributions vary geographically, the modeling was done separately for each of the three reaches of the
UMR (upper impounded, lower impounded, and open river) as defined by Miltner et al. (2011).

The GRFIn indices, the regional FACI score, and the Ohio CIBI were then calculated  using the synthetic
data for the impounded and open river reaches of the UMR. The Ohio CIBI scoring  ranges were not
limited by existing conditions, but assumed that species richness metrics were greater in the past and
allowed for higher scoring than current existing species richness levels. As expected, the synthetic data
resulted in higher GRFIn, FACI, and CIBI scores than the sampled data (Figure Bl-4). Estimates of
abundance for these metrics were based on abundances observed at the best existing UMR sites and,
for rare species, were based on extrapolations based on species life history knowledge and historical
descriptions of abundances when available.

Bl.5.1 Initial Reconstruction of Environmental Conditions to Match Biological
Condition Gradient Levels 1-2
The synthetic fish assemblages are intended to approximate levels 1-2 of the BCG for the UMR during
pre-settlement periods. Environmental conditions were inferred based on the grand  ranking of TIV
scores for the synthetic and existing data and plotted against the GRFIn and FACI for the impounded and
open river reaches (Figure Bl-5). Despite some overlap  in terms of the extrapolated stressor levels with
some of the recent data, particularly from sections of the unimpounded reaches of the UMR, there is a
substantial degree of separation between the stressor levels approximating the historical  and the
present day assemblages. The synthesized data representing the 1960s are among the lowest FACI and
GRFIn scores and coincide with the higher ranges  of the extrapolated stressor ranking.
                                                                                         B-ll

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
                                           February 2016
                           Actual Data
Extrapolated
   Data
cker
unda
S
        20
nce
h-1
Ln
u
Ab
h-1
O
  01  01
 .3  >
 M  «
                  • Blue Sucker    - -A- - Blue Sucker - prob. of capture
                   number/km     -o _ B,ue Sucker (# » prob)
                                                             1000

                                                             100
                x  -a
                £&
           10   §  »
                Q.  ^;
                M
                (D
          10      20      30       40       50
                              IBI Midpoint
                          Actual Data
                                                   60
Extrapolated
   Data
        10
                              ~ -A- - River Darter - prob. of capture
                                   River Darter (# * prob.)
S
T-l
o
T-l
iver Da
Probabi
O
y x Abundance
obab
                  20
                          30      40       50
                              IBI Midpoint
                                                   60
9
O
Extrapolated
Tn
ZU

01
u 15
C
O nj
"ro "°
1 §
en < 10
^^ fll
u 2;
03 >
s s
Ol c
a 5



Actual Data Data
ii
• Black Buffalo --A-- Black Buffalo - prob. of capture
- number/km ^3 — Black Buffalo - (# * prob.)
-



_





/
^ f
- ^aL,






£>" *
f

0.**"
t
f
	 • 	





^.
^

* *^~m




—f 	


4
; -*"-
^a— —\ ~~
^
m »
• *"




,^***


s
*


,.-A





, 	 •



-i nnn
1UUU
o
"• DO
5 a*
g. S-
inn su oo
1UU o- c
= n?
•J o
^o
o -^
^ 0
10 = g"
i. =r
a> =:
=. -5
ro
1
10 20 30 40 50 60
IBI Midpoint
Figure Bl-3. Plots of relative abundance, probability of capture, and abundance x probability vs. IBI midpoint for
three riverine fish species: blue sucker, river darter, and black buffalo. Actual abundance data and probability of
capture data generated from data on boatable sites from Ohio and Indiana; extrapolated data estimated using
best professional judgment based on trends in actual data, data on historical distributions in the UMR (Steuck et
al. 2010), and life history information and other historical information.
                                                                                                       B-12

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
February 2016
       100
        80
        60  -
        40  -
        20  -
                   Existing vs. "Synthetic" Historical Extrapolation

0














-j-



















9





c
T



o
t
o
Upper Upper Lower Lower Open Open Open
Impound. Impound. Impound. Impound. River River River
Historical Current Historical Recent Historical Current 1960s
                   Existing vs. "Synthetic" Historical Extrapolation
 tc
 ID
            Upper   Upper    Lower    Lower    Open     Open    Open
           Impound. Impound. Impound. Impound.   River     River     River
           Historical  Current  Historical  Current Historical   Current   1960s
Figure Bl-4. Box-and-whisker plots of FACI scores (top) and GRFIn scores (bottom) for historical "synthetically"
derived fish assemblages (blue) and present-day data (orange) for the upper impounded, lower impounded, and
the open river reaches of the UMR. The red shaded box are synthetically derived scores in the open river during
the 1960s prior to CWA mandated point source controls.
                                                                                                       B-13

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
February 2016
      10
 (D
       6 -
       4 -
       2 -



o Recent - Impounded n Historical Open River
• Recent - Open River A 1960s Estimate
• Historical - Impounded
n
n
t

—
---







• 1
• ^
nn '
n
n

••
'


1 •
G °

O
• ° •
• C
•* + •


G
o° o
GO (OsQ
^o°J^P%
A?i
OO o
,°l
o


G
8
fo°o
of °G
>^
0
J
3 G
O n



O

k
i
A A
3
 u
                4.5       5       5.5       6
                         Grand Stressor Rank
                                                   6.5

o Recent - Impounded n Historical Open River
• Recent - Open River A 1960s Estimate
• Historical - Impounded


—
n
r


m.
m
i
.-


^m GE
1 0 0 0 0
• 0
•


"Cr
^fllW
5 0 f 
-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                      February 2016


Bl.5.2 Using the Biological Condition Gradient to Establish Attainable Biological
Thresholds
The derivation of biological thresholds to assess ALL) attainment using an Environmental Monitoring and
Assessment Program (EMAP) data set focused on the statistical assumptions and consequences of using
several different methods and approaches (Miltner et al. 2011). Researchers in this analysis suggest that
such a statistical approach should be linked explicitly to biological inferences and narratives about their
position along the BCG for improved communication with decisionmakers and stakeholders.

The effort to develop a synthetic historical fish assemblage was done to provide a foundation for
evaluating the attainability of various biological thresholds. This approach helps define biological
thresholds that can be interpreted relative to the CWA biological integrity objective. This framework
assists in understanding current conditions along a gradient of stress and reduces the risk of setting
thresholds based on an assumption that the current condition of a waterbody represents its full
ecological potential when it may be significantly degraded (Humphries and Winemiller 2009). Executing
this approach requires that the key measures of biological condition be linked to stressor gradients so
that evaluation of the recovery potential of a large river, or segment within the river, can be performed.
The statistical approaches conducted in the UMR thresholds analysis (Miltner et al. 2011) were designed
to offer an analysis of various ways of developing biological thresholds and then evaluating the
attainability of these thresholds.

One result was readily apparent in this analysis—all three UMR reaches historically had similar levels of
attribute I, II and III species, i.e., the rare, endemic, long lived species;highly sensitive species; and
intermediate sensitive species. This was especially important for the open river reach, because a variant
of the GRFIn was derived and calibrated to current conditions in this UMR reach, which is highly
modified, thus setting expectations at its current level of alteration. The open river reach is the most
highly modified of all three reaches examined, and this is especially reflected by a higher proportion of
tolerant and exotic individuals (Figure Bl-6). The comparable levels of historic condition also suggest
that the impounded GRFIn and stressor gradients could be extended to the open river. Presently, the
open  river has been treated separately from the impounded sections of the UMR in the derivation and
calibration of the GRFIn.

To relate the BCG attributes to the current indices, the number of species in  each attribute were
compared to the GRFIn and FACI indices. Both indices showed a significant correlation with attribute III
(Figure Bl-7 and Figure Bl-8, top). There was, however, no apparent correlation between BCG attributes
I and II and either the GRFIn or the FACI (Figure Bl-7 and Figure Bl-8, middle and bottom). The lack of
correlation with BCG attributes I and II may well be explained by the large difference between historical
and existing conditions in the UMR - and that these more sensitive species appear to have been
eliminated from the UMR. It could be that populations of these species in the main channel samples
might be more related to the losses in connectivity with the side channel  habitats than with the
conditions in the main channel itself.  The presence of the intermediate sensitive species though
provides some promise that restoration of the more sensitive species may be possible if conditions
improve and/or parts of the river are reconnected.
                                                                                           B-15

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
                           February 2016
f Rare, Long- Lived Mean Species BCG
;BCG Attribute 1) Attribute Level
i CTi CO O NJUnWUn-I^Unun
Number o
Species
O NJ -t
Existingvs. "Synthetic" Historical Extrapolation







I
t
	 1
	 1 	
_L





^


o
T
^






H


ZZT


I
t

t....



Historica Recent Historical Recent Historica Recent
Upper Upper Middle Middle Lower Lower
River Segment
Existingvs. "Synthetic" Historical Extrapolation






1




o



1
	
T




o


















O
U'^t • D J-U'fr'lD i-U'fr1
100
£ 80
o
08 «
Percent Tolerant
Individua
T NJ -^ C
D 0 0 0 C
}f Highly Sensitive
BCG Attribute II)
£ G E
oi S
-O 'u
Z «
n
Existingvs. 'Synthetic" Histor cal Extrapolation




— ^
Q
O
O
o

1 	

	
:













1




^^
_._



•
	 	 -

Historica Recent Historical Recent Historica Recent
Upper Upper Middle Middle Lower Lower
River Segment
Existingvs. "Synthetic" Histor cal Extrapolation


1






o


7
	 1
' 	 '
i





o
i
!
	 T 	



1
i



O
           Upper    Upper   Middle   Middle   Lower   Lower

                        River Segment
               Existingvs. "Synthetic" Historical Extrapolation
Historical  Recent   Historical  Recent  Historical   Recent
 Upper   Upper   Middle   Middle   Lower   Lower

              River Segment
of Intermediate Sensitive
es (BCG Attribute III)
£ G £
l_ U
£ & 5
E ">
3
z
0
T

_L


Historica
Upper












^



T
T
•~ ~^
+'



-+-
t=P
Recent Historical Recent Historical Recent
Upper Middle Middle Lower Lower
                        River Segment
Figure Bl-6. Box-and-whisker plots of BCG attributes for recent data (orange) and historical synthetic data (blue)
for the upper impounded (river mile (RM) 523-812), lower impounded (RM 523-196), and open river (RM 196-
0) UMR for mean species BCG attribute (upper left), mean tolerant and exotic species (upper right), number of
rare, long-lived species (I) (middle, left), highly sensitive species (II) (middle, right), and intermediate sensitive
species (III) (bottom, left).
                                                                                                         B-16

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
                                                                  February 2016
                                  16
                            
                            £ 
-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
February 2016
                                                     FACI
Figure Bl-8. Scatter plots of the FACI index for existing data vs. BCG attributes III (top), II (middle), and I
(bottom) from these data. Historical distribution of synthetic data for these attributes for the UMR is
illustrated with a box plot to the right of each box.
                                                                                                B-18

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                     February 2016


Bl.5.3 Using the Biological Condition Gradient and Biological Indices to Derive
Aquatic Life Use Thresholds for the Upper Mississippi River
BCG attributes were used to derive cutoffs for the GRFIn, FAG, and Ohio CIBI based on best professional
judgement using the relationship between the BCG attributes and index scores. Figure Bl-9 shows plots
of the GRFIn, FAG, and Ohio CIBI versus the number of combined BCG attribute I, II and III species at
each UMR site. The BCG I, II, III species were combined to approximate the fish assemblage of the UMR
that may be achieved, in part or in whole, with WQ improvements, mitigation and/or restoration efforts
in the future. At a minimum, this "synthetic" assemblage can be used to inform efforts to improve
conditions and restore a more naturally functioning and connected river system. The synthetic data are
coded with blue squares to distinguish it from the present-day sampling data (green circles), and the
open river reach is coded with solid  orange circles. The highly degraded synthetic results are also
included as red triangles in order to have the full  breadth of the BCG represented. All three of the
indices reflect a positive relationship with the number of BCG attribute  I, II and III species (Figure Bl-9)
based on a locally weighted regression that minimizes the effect of outliers. The breaks in these  curves
illustrate patterns in the relationships  that can be used to support various options for selecting tiered
impairment thresholds. The break in the curves in these relationships with the weighted BCG is aided by
the availability of the synthetic data to complete the curves (Figure Bl-9) and is informative if higher
aquatic life thresholds are desired. The tighter relationship  between the FAG, the Ohio CIBI, and BCG
attribute I, II and III species compared  to the GRFIn is  likely  related to the similar metrics in these indices
and a broader geographic basis for their derivation. The GRFIn is designed to maximize the association
with a derived stressor gradient (Angradi et al. 2009a). The  Ohio CIBI provides a way to separate high
and low performing sites beyond the current range of that index based  on contemporary conditions.
Actual data from the UMR are also lacking for the time period when point source pollution stressors
were the most severe (1950s-1970s),  which presumably resulted in assemblages characteristic of BCG
level 6. The availability of such data, which were synthesized the same way as were the historical
conditions, should enhance change point analyses and make the indices more sensitive to the extremes
of the disturbance gradient (e.g., the Ohio CIBI). Although the CIBI was originally calibrated for smaller
large rivers, the application of the method can better  illuminate change points since the other indices
were calibrated to accommodate estimates of historical assemblage condition.
                                                                                           B-19

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
                                                            February 2016
                                    Level 6   Level 5  Level 4   Level 3   Level 2 Level 1
                          
-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                      February 2016
B1.6  Conclusions
The UMR, like many Midwest rivers, has been subjected to a series of perturbations that have
accelerated greatly with European settlement beginning in the 19th century. Some of the historically
common UMR species are now rare, but most remain present if even in limited numbers and
distribution. Present-day conditions, however, have improved since the zenith of gross pollution from
untreatei
century.
untreated industrial and human wastewater sources during the late 19th and the first half of the 20th
Linking existing UMR biological indices to the BCG levels and individual attributes may help strengthen
the technical basis and ability to communicate the rationale for setting appropriate and attainable
biological thresholds for the UMR. BCG levels 1 and 2 represent undisturbed or minimally disturbed
conditions and can be characterized based on historical data and records when these conditions no
longer exist. BCG levels 3 and 4 represent biological assemblages that have been subject to increasing
levels of stress, but which still include some representative species that would be expected under
undisturbed or minimally disturbed conditions. As an initial first step, linkage between specific BCG
attributes, the GRFIn, FACI, and the Ohio CIBI were examined and BCG level thresholds proposed.

The BCG and individual attributes can also be also used to provide a narrative backup to the statistically
derived impairment thresholds of Miltner et al. (2011). The distance between the present-day
conditions and the "as naturally occurs" conditions that once existed in the  UMR leaves much room for
restoration, but restoration also requires an awareness about the status of present-day UMR fish
assemblages with respect to the currently available indices such as GRFIn, FACI, and the Ohio CIBI. The
fact that many of the historically common fish species are still present indicates that habitats still exist to
support at least relict populations of BCG attribute l-lll species.

Appendix Bl  References

Alexander, J.S., R.C. Wilson, and W.R. Green. 2012. A Brief History and Summary of the Effects of River
      Engineering and Dams  on the Mississippi River System and Delta. U.S. Geological Survey Circular
      1375. U.S. Geological Survey, Lincoln, NE.

Angradi, T.R., ed. 2006. Environmental Monitoring and Assessment Program, Great River Ecosystems
      Field Operations Manual. EPA/620/R06/002. U.S. Environmental Protection Agency, Office of
      Research and  Development, Washington, DC. http://www.epa.gov/emap/greatriver/fom.html.
      Accessed February 2016.

Angradi, T.R., M.S. Pearson, T.M. Jicha, D.L. Taylor, D.W. Bolgrien, M.A. Moffett, K.A. Blocksom, and B.H.
      Hill. 2009a. Using stressor gradients to determine reference expectations for great river fish
      assemblages. Ecological Indicators 9(4):748-764.

Angradi, T.R., D.W. Bolgrien, T.M. Jicha,  M.S. Pearson, B.H. Hill, D.L. Taylor, E.W. Schweiger, L. Shepard,
      A.R. Batterman, M.F. Moffett, C.M. Elonen, and L.E. Anderson. 2009b. A bioassessment approach
      for mid-continent great rivers:  The Upper Mississippi, Missouri, and Ohio (USA). Environmental
      Monitoring and Assessment 152:425-442.
                                                                                           B-21

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                      February 2016


Armitage, B.J., R. Mueller, and E.T. Rankin. 2009. An Assessment of Threats to the Biological Reference
      Expectations for Great River Fish Assemblages.  Ecological Indicators Condition of the Wabash
      River Aquatic Ecosystem of Indiana. In two parts. Prepared for The Indiana Nature Conservancy,
      Indianapolis, IN.

Blocksom, K.A. 2003. A performance comparison of metric scoring methods for a multimetric index for
      Mid-Atlantic highlands streams. Journal of Environmental Management 31:670-682.

Carlander, H.B. 1954. A History of Fish and Fishing in the Upper Mississippi River. UMRCC Special
      Publication.

Davies, S.P., and S.K. Jackson. 2006. The biological condition gradient: A descriptive model for
      interpreting change in aquatic ecosystems. Ecological Applications 16:1251-1266.

Emery, E., R. Tewes, R. Argo,  and J. Thomas. 2007. Development of a Probability Based Monitoring and
      Assessment Strategy for Select Large  Rivers within U.S. EPA Region 5. Report in fulfillment of EPA
      Grant RM-83169201. ORSANCO, Cincinnati, OH.

Etnier, D.A., and W.C. Starnes. 1993. The Fishes of Tennessee. University of Tennessee Press, Knoxville.

Freedman, J.A., S.E. Butler, and D.H. Wahl. 2012. Impacts of Invasive Asian Carps on Native Food Webs.
      Final project report to Illinois-Indiana Sea Grant. University of Illinois, Sullivan.

Fremling, C.R., J.L. Rasmussen, R.E. Sparks, S.P. Cobb, C.F.  Bryan, and T.O. Claflin. 1989.  Mississippi River
      fisheries: A case history. Pages 309-351 in D.P.  Dodge, editor. Proceedings of the  International
      Large River Symposium. Canadian Special Publication of Fisheries and Aquatic Sciences 106,
      Ottawa, Ontario.

Hughes, R.M.,  D.P. Larsen, and J.M. Omernik. 1986. Regional reference sites: A method  for assessing
      stream potential. Environmental Management  10:629-635.

Humphries, P., and K.O. Winemiller. 2009. Historical impacts on river fauna, shifting baselines, and
      challenges for restoration. BioScience 59(8):674-684.

Kopf, R.K., C.M. Finlayson, P.  Humphries, N.C. Sims, and S. Hladyz. 2015. Anthropocene  baselines:
      Assessing change and managing biodiversity in  human-dominated aquatic ecosystems. BioScience
      65(8):798-811.

Lyman, R.L 2006.  Paleozoology in the service of conservation biology. Evolutionary Anthropology 15:11-
      19.

Mac, M.J., P.A. Opler, C.E. Puckett Haecker, and P.O. Doran. 1998. Status and Trends of  the Nation's
      Biological Resources. 2 vols. U.S. Department of the Interior, U.S. Geological Survey, Reston, VA.

McClenachan, L, A.B. Cooper, M.G. McKenzie,  and J.  A. Drew. 2015. The importance of  surprising results
      and best practices in historical ecology. BioScience 65(9):l-8.

Meador, M.R., and D.M.  Carlisle. 2007. Quantifying tolerance indicator values for common stream fish
      species of the United States. Ecological Indicators 7(2):329-338.
                                                                                           B-22

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                      February 2016
Meador, M.R., D.M. Carlisle, and J.F. Coles. 2008. Use of tolerance values to diagnose water-quality
      stressors to aquatic biota in New England streams. Ecological Indicators 8(5):718-728.

Miltner, R.J., C.O. Yoder, and E.T. Rankin. 2011. Preliminary Analysis of Biological Assessment Thresholds
      for Determining Aquatic Life Use Attainment Status in the Upper Mississippi River Mainstem. MBI
      Technical Report/2011-5-1. Upper Mississippi River Basin Association, St. Paul, MN.
      http://www.midwestbiodiversityinst.org/publications. Accessed February 2016.

Pitlo, J., Jr., and J.L. Rasmussen. 2004. A Compendium of Fishery Information on the Upper Mississippi
      River. 3rd ed. UMRCC Special Publication. Conservation Committee.

Rankin, E.T. 2010. Calibration of the Ohio IBI and ICI Using Continuous Scoring Methods. Report to Ohio
      EPA,  Division of Surface Water. Ohio University, Voinovich School for Leadership and Public
      Affairs, Athens, OH.

Rankin, E.T., and C.O. Yoder. 2011. Improving Water Quality Standards and Assessment Approaches for
      the Upper Mississippi River: UMR Clean Water Act Biological Assessment Implementation
      Guidance: Development of a Biological Condition Gradient for Fish Assemblages of the Upper
      Mississippi River and Development of a "Synthetic" Historical Fish Community. MBI Technical
      Report/2011-5-2. Submitted to UMRBA WQTF. 24 pp. http://www.umrba.org/wq.htm. Accessed
      February 2016.

Rankin, E.T., and C.O. Yoder. 2012. Improving Water Quality Standards and Assessment Approaches for
      the Upper Mississippi River: Development of a Biological Condition Gradient for Fish Assemblages
      of the Upper Mississippi River and Development of a "Synthetic" Historical Fish Community. MBI
      Technical Report/2011-5-2. Upper Mississippi River Basin Association, St. Paul, MN.
      http://www.midwestbiodiversityinst.org/publications. Accessed February 2016.

Steuck, M.J., S. Yess, J. Pitlo, A. Van Vooren, and J. Rasmussen. 2010. Distribution and Relative
      Abundance of Upper Mississippi River Fishes. Upper Mississippi River Conservation Committee,
      Onalaska, Wl.

Stoddard, J.L., D.P. Larsen, C.P. Hawkins, R.K. Johnson, and R.H. Norris. 2006. Setting expectations for
      ecological condition of running waters: The concept of reference condition. Ecological
      Applications 16:1267-1276.

Thorp, J.E. 1992. Linkage between  islands and benthos in the Ohio River, with implications for riverine
      management. Canadian Journal of Fisheries and Aquatic Sciences 49:1873-1882.

Trautman, M. 1981. The Fishes of Ohio. Revised edition. The Ohio State University Press, Columbus.

Yoder, C.O., and M.A. Smith. 1999.  Using Fish Assemblages in a State Biological Assessment and Criteria
      Program: Essential Concepts and Considerations.  In Assessing the Sustainability and Biological
      Integrity of Water Resources  using Fish Communities, T.P. Simon (ed.), pp. 17-56. CRC Press, Boca
      Raton, FL
                                                                                            B-23

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                     February 2016


Yoder, CO., V.L. Gordon, N.B. Kale, and O.K. Hokanson. 2010. Improving Water Quality Standards and
      Assessment Approaches for the Upper Mississippi River: UMR Clean Water Act Biological
      Assessment Implementation Guidance—Background and Scoping Report. Upper Mississippi River
      Basin Association, St. Paul, MN.

Yoder, C.O., R.J. Miltner, V.L. Gordon, E.T. Rankin, N.B. Kale, and O.K. Hokanson. 2011. Improving Water
      Quality Standards and Assessment Approaches for the Upper Mississippi River: UMR Clean Water
      Act Biological Assessment Implementation Guidance. Upper Mississippi River Basin Association, St.
      Paul, MN. http://www.umrba.org/publications.htmtfwq. Accessed February 2016.
                                                                                           B-24

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                     February 2016


B2.  Narragansett Bay: Development of a  Biological Condition
Gradient for Estuarine Habitat Quality

Emily Shumchenia, PhD, E&C Enviroscape, LLCa; Giancarlo Cicchetti, PhD, and Marguerite C. Pelletier,
PhD, EPA Office of Research and Development, National Health and Environmental Effects Research
Laboratory-Atlantic Ecology Division, Narragansett, Rhode Island

B2.1 Background

Estuarine waters are affected by a variety of stressors acting at several scales: localized point sources of
contaminants; widespread or diffuse nonpoint sources of contaminants such as nutrients; and global
impacts such as climate change. Consequently, these valued ecosystems are greatly affected by the
cumulative impacts of multiple stressors. Overtime, this has led to "severe, long-term degradation of
near-shore marine systems worldwide" (Lotze et al. 2006). As such, it is critical to have a way to
interpret biological condition consistently and independently of assessment methods for estuaries and
other coastal systems.

Biological condition integrates the effects to living organisms from exposure to stressors. A biological
assessment is an effective tool in managing cumulative impacts. Many different biological assessment
methods and biological indices to quantify biological condition have been developed and applied by
scientists, local resource managers, states, and federal agencies. Most assessments evaluate changes in
quality or quantity of ecologically or economically valued habitats, communities, or species relative to a
defined reference condition. These assessments, when applied in different estuaries, often evaluate
very different aspects of biology and use different reference conditions, usually for the good reason that
biology itself differs among estuaries.

Few tools or frameworks exist to evaluate and manage the gradual degradation of estuaries and
estuarine functions along with the ecological and social benefits they provide over decades. A method to
interpret biological condition consistently, regardless of location, time, or assessment method, would
allow scientists and water quality managers to compare assessments  of aquatic resource condition more
uniformly and directly,  and communicate more clearly to the public both the current status of aquatic
resources and their potential for restoration (Davies and Jackson 2006; USEPA 2011).

B2.2 Development of an Estuarine Biological Condition Gradient Framework

The estuarine BCG approach was initially proposed and launched at a 2005 workshop hosted by EPA
(Office of Water and Region 1) in Providence, Rhode Island. Concepts were developed further at
workshops in Maine during the winter of 2006 and spring of 2007. The approach was solidified when the
EPA Office of Water, Region 1, and Office of Research and Development co-sponsored a November 2008
workshop in Narragansett, Rhode Island, inviting many national estuarine experts and managers. The
goal of these efforts was to develop and refine a nationally consistent, integrative estuarine BCG
framework to enable meaningful comparisons among metrics and water bodies.
13 emily.shumchenia@gmail.com
                                                                                        B-25

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                      February 2016
The proposed estuarine framework considered structure, function, condition, connectivity, and non-
native species in water bodies at multiple scales, including the species- and habitat-scale (e.g., seagrass
health measures, benthic faunal indices), as well as the whole-estuary scale (e.g., measures of the
estuarine mosaic of living habitats). The proposed estuarine framework drew from existing ideas on
estuarine biological assessment:

    •  The Biological Condition Gradient: A Descriptive Model for Interpreting Change in Aquatic
       Ecosystems (Davies and Jackson 2006)

    •  Estuarine and Coastal Marine Waters: Bioassessment and Biocriteria Technical Guidance (EPA
       2000)

The attributes, potential metrics, and condition levels developed for the estuarine BCG are shown in
Table B2-1. Table B2-1 is intended to assist with organization of metrics, but users should recognize that
applicable indicators can be estuary-specific, in terms of what data are collected by estuary programs
and how often biological monitoring occurs.
                                                                                            B-26

-------
                  Appendices to A Practitioner's Guide to the Biological Condition Gradient
February 2016
Table B2-1. Attributes and potential metrics developed at the 2008 BCG workshop (left two columns) paired with examples of narrative for
BCG levels (right 6 columns).
Attribute



£
X
01
Q.


"S
c
_o
]M
'u>
o
Q.
£
8
•a
£
ra
01
i.
I
3
i.
%



Potential Metrics and
Descriptions
Measures of water body,
community, or habitat
structure and complexity,
also recognizing loss of
habitats or species due to
human activities. Examples
include macroinvertebrate
or fish indices,
phytoplankton or
zooplankton community
measures, epifaunal
measures, biotope
measures,
presence/quantity of
sensitive taxa or biotopes,
measures of seagrass and
macroalgae





Examples of BCG Level Narratives (based on Davies and Jackson 2006 and recommendations of a panel of experts)
1
Community
composition is as
naturally occurs
except for global
extinctions;
patterns of primary
production,
biotope measures,
and communities
with large, long-
lived and sensitive
species are as
naturally occurs









2
Slight changes in
natural
occurrences of
biotopes or
patterns of
primary
production;
minimal changes
in abundances of
sensitive or
tolerant species











3
Evident changes in
biological metrics;
decreases in
sensitive species and
increases in tolerant
species; some
evident changes in
patterns of primary
production;
decreases in
sensitive habitat
area and changes to
biotope measures
are also evident








4
Moderate changes
in biological metrics;
some sensitive,
large, or long-lived
taxa may be
markedly diminished
or absent; increases
in tolerant species;
many evident
changes in patterns
of primary
production; biotope
measures
significantly altered
with replacement of
natural
habitats/biotopes by
tolerant or non-
natu rally occurring
components


5
Sensitive, large
and/or long-lived
taxa are markedly
diminished or
absent, with a
dominance in
abundance of
tolerant taxa;
significant shifts in
species diversity,
size, and densities of
remaining species;
biotope measures
significantly altered;
most sensitive
natural
habitats/biotopes
lost with
replacement by
tolerant or non-
natu rally occurring
components
6
Sensitive, large
and/or long-lived
taxa are absent,
with extremes in
abundance of
tolerant taxa;
extreme shifts in
species diversity
and in size spectra
of remaining
organisms;
extreme alteration
of natural biotope
measures with
complete loss of
sensitive habitats






                                                                                                                               B-27

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
February 2016
AttriKi I+A
Aiiriuuie






c
_g
]M
'-a
c
8












£
o
tj
C
3
Li-






Potential Metrics and
Descriptions
Measures of the condition
("health") of water bodies,
habitats, or species. Also
includes measures of
resiliency. Examples
include harmful algal
blooms, disease
outbreaks, outbreaks of
other harmful taxa,
measures of habitat or
biotope health such as
seagrass condition or
wetland condition, fish
pathology or shellfish bed
condition, measures of
reproductive success
Measures of energy flow,
trophic linkages and
material cycling, including
proxy or snapshot metrics
that correlate to functional
measures. Examples
include photosynthesis:
respiration ratios, benthic:
pelagic production rates,
chlorophyll a
concentrations, benthic
bioturbation, and form/
biomass of primary
production


Examples of BCG Level Narratives (based on Davies and Jackson 2006 and recommendations of a panel of experts)
1
Diseases, harmful
algal blooms, other
outbreaks,
measures of
reproductive
success, and
condition/health
measures are as
naturally occurs







Energy flows,
material cycling,
and other functions
are as naturally
occur, typically
characterized by
complex
interactions and
many complex
trophic links
supporting large,
long-lived
organisms



2
Diseases,
harmful algal
blooms, other
outbreaks,
measures of
reproductive
success, and
condition/health
measures are as
naturally occurs






Energy flows,
material cycling,
and other
functions are as
naturally occur,
typically
characterized by
complex
interactions and
many complex
trophic links
supporting large,
long-lived
organisms


3
Incidences of
diseases, harmful
algal blooms, and
other outbreaks are
infrequent;
reproductive success
and
condition/health
measures are within
the range of
variability for
naturally occurring
characteristics



Virtually all
functions are
maintained through
operationally
redundant system
attributes; minimal
changes to some
indicative functions








4
Incidences of
diseases, harmful
algal blooms, and
other outbreaks may
be slightly higher
than expected;
reproductive success
and
condition/health
measures may be
slightly lower than
expected




Most functions are
largely maintained
through
operationally
redundant system
attributes, though
there is evidence of
loss of complexity,
efficiency, and shifts
in some rates






5
Incidences of
diseases, harmful
algal blooms, and
other outbreaks are
increasingly
common;
reproductive success
and
condition/health
measures are
significantly lower
than expected




Losses of some
ecosystem functions
are apparent,
manifested as
changes in energy
and material flows,
functional rates, or
reduced complexity








6
Diseases, harmful
algal blooms, and
other outbreaks
are common and
serious,
reproduction is
minimal except for
extremely tolerant
groups, and
condition/health
measures are
extremely low




Most functions
show extensive
and persistent
disruption: shifts in
primary
production,
microbial
dominance, highly
simplified trophic
structure, extreme
shifts in energy and
material
processing rates,
extensively
reduced
complexity
                                                                                                                        B-28

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
February 2016
A+t-riKi I+A
AIIilDUIc







>
•M
\>
'+•»
OJ
VJ
c
3











CD
x
£
01
'^
ra

c
O
Z



Potential Metrics and
Descriptions
Metrics of exchange or
migrations of biota
between adjacent water
bodies or habitats; may be
strongly affected by
factors adjacent to or
larger than the immediate
study area. Proxies may be
used as measures,
including habitat
landscape metrics,
anadromousfish data, or
hydrological measures







Metrics of non-native
species, including
intentionally introduced
species. May include
measures of the impact of
introduced and non-native
species. Examples include
estimated numbers of
species or individuals,
biomass measures of
natives and non-natives, or
replacements of native
species
Examples of BCG Level Narratives (based on Davies and Jackson 2006 and recommendations of a panel of experts)
1
System is highly
connected in space
and time*;
exchanges,
migrations, and
recruitment from
adjacent water
bodies or habitats
are as naturally
occurs.
*Note that some
systems are
naturally closed off,
and this is the level
1 state.





Non-native taxa are
absent, or if
minimally present
do not affect native
biota or natural
processes







2
Ecosystem
connectivity is
unimpaired

















Non-native taxa
are present, but
occurrence has a
non-detrimental
effect on native
taxa or natural
processes






3
Slight loss in
connectivity in space
or time, with slight
decreases in
exchanges,
migrations, or
recruitment from
adjacent water
bodies or habitats











Non-native taxa may
be prominent in
some assemblages
(e.g., crustaceans,
algae, bivalves,
fishes); native taxa
may be reduced






4
Some loss in
connectivity with
adjacent water
bodies or habitats,
but alternative
pathways prevent
complete
disconnects or other
failures











Some replacement
of sensitive native
taxa with
functionally diverse
assemblages of non-
natives







5
Significant loss in
ecosystem
connectivity with
adjacent water
bodies or habitats is
evident; alternative
pathways do not
exist for some taxa;
some near-complete
disconnects exist;
significant
reductions in
naturally occurring
biotopes






Some assemblages
(e.g., crustaceans,
algae, bivalves,
fishes, epifauna) are
dominated by
tolerant non-native
taxa






6
Complete loss in
ecosystem
connectivity in at
least one
dimension (either
spatially or
temporally) lowers
reproductive or
recruitment
success or
prevents migration
or exchanges with
adjacent water
bodies or habitats;
disconnects or
other failures are
frequent; most
naturally occurring
biotopes are
eliminated
Non-native taxa
are often dominant
and may be the
only representative
of some
assemblages (e.g.,
crustaceans, algae,
bivalves, fishes,
epifauna)




                                                                                                                        B-29

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                     February 2016


While conceptually relying on basic BCG principles, the estuarine framework creates a flexible approach
that can be applied to different individual estuaries. The framework promotes a system-level
examination of the estuary and modifies attributes from freshwater descriptions to apply at different
scales of assessment. For example, it expands the organism condition attribute to include habitat
condition. It also begins to develop the BCG attributes for ecosystem function and connectance, which
were not quantitatively defined for application to streams. The framework is  designed so that the
estuary can be conceptualized as a functioning system. Thus, the promise of the BCG framework is that
once it is calibrated for a specific estuary, it can be used to assess overall estuarine condition in the past
and present, and it can be used to develop visions for desired future conditions.

To address these challenges, since the 2008 Narragansett workshop, an estuarine BCG work group
composed of scientists from EPA and the State of Rhode Island proposed a series of "action steps" that
guide coastal scientists and managers through the process of developing a BCG (Cicchetti et al. 2016).
Each step delivers a product or set of products of use to managers, but each step can be applied, or not
applied, as best meets the goals of individual programs. The first action steps of the framework do not
involve an actual BCG per se, but apply accepted management decision tools  to lead management
groups through evaluation of environmental problems. The next steps integrate the actual BCG into
solving these problems by developing reference conditions, narratives for BCG levels, preliminary
biological assessments, and  broad goals. In the final stages a rigorous and quantitative BCG is developed
through expert consensus, and it can support development of quantitative biological thresholds,
potentially other regulatory and stressor-response thresholds, a variety of non-regulatory actions, and
monitoring for effectiveness of management actions. This flexible framework allows scientists and
managers to develop these steps using any method or sequence that would best address specific needs.
Stages and action steps of estuarine BCG development are:
Management steps—clarifying needs and directions of work using larger frameworks such as Structured
Decision Making or Drivers-Pressures-State-Impact-Response:
    1. Identify management clients and stakeholders
    2. Collaborate to define management goals, visions, and objectives
    3. Determine the biological attributes, measures, and stressors most relevant to management
       objectives
BCG development steps for non-regulatory management—setting targets, communicating, motivating:
    4. Delineate and classify the water body and watershed of interest
    5. Organize and analyze existing data for the identified measures;  collect new data, if needed
    6. Define a "minimally disturbed" reference condition for the measures
    7. Develop narrative descriptions of the biology expected at each BCG level; assist management
       partners
Additional BCG development steps for regulatory management—determining impairment, setting
thresholds for actions, linking measures to stressors, monitoring for change:
    8. Convert narrative descriptions to quantitative measures and thresholds for BCG levels
    9. Develop a stressor gradient and stressor-response relationships
    10. Organize, interpret, and report results
    11. Develop decision-support, communication, and monitoring tools; assist management partners

In past experience presenting estuarine BCG concepts to national experts, scientists tended to focus
their attention on step 6 above. The scientists' focus on the integrity of the data and approach to honing
the definition of "minimally disturbed" as a  reference condition for estuaries  was extremely valuable for
making progress toward an estuarine pilot.
                                                                                          B-30

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                     February 2016
B2.3  Establishing Reference Conditions in Historically Disturbed Environments

BCG level 1 has been interpreted as "as naturally occurs" conditions in absence of anthropogenic
disturbance. Participants in the two estuarine BCG development workshops (Cicchetti and Pryor 2010;
Cicchetti 2010) found that defining "natural" condition for estuaries was a challenge because very few (if
any) undisturbed, or pristine, sites exist in coastal ecosystems today (Bald et al. 2005;  Muxika et al.
2007). A more practical reference level that might be used in a BCG is a "minimally disturbed" condition
that represents an ecological state "in the absence of significant human disturbance" (Stoddard et al.
2006) and has been  considered as comparable to a BCG level 2 by stream biologists (Davies and Jackson
2006). These types of sites may still  be difficult to locate in a modern estuary. For this  reason, it may be
desirable to use historical data to describe a minimally disturbed reference level. Historical baselines are
not without their complications either, as (1) human impact pre-dates modern science in essentially all
U.S. and European watersheds, and thus quantitative data are limited (Borja et al. 2011); (2) they are
difficult to calibrate  with current ecosystem status; (3) ecosystems were as dynamic in the past as they
are today; and (4) climate change and the degree of anthropogenic influence can render these baselines
unattainable (Samhouri et al. 2011). Careful definition and anchoring of reference conditions is needed
to avoid shifting baselines (Pauly 1995) where societal and scientific perceptions of what is "good" or
"normal" are  based  on the expectations developed during a human lifetime. As a result, the "best"
conditions that remain in an area or region can be misinterpreted as "minimally disturbed" two or three
decades later (Papworth et al. 2009).

Where historical quantitative stressor/response data are available, management and/or restoration
efforts have been quite successful in utilizing a biological condition-type approach. While not explicitly a
"BCG", scientists and managers in the Chesapeake Bay, Buzzards Bay, Tampa Bay estuary, and Puget
Sound have taken similar historical biological assessment approaches that are certainly conducive to
organization in the estuarine BCG framework. The Chesapeake Bay Program (CBP) is a well-known
pioneer in setting restoration targets. CBP set numerous conservation and restoration goals in its 2000
report for ecosystem components ranging from  oysters to nutrients  and sediments, many of which  are
based on historical baselines (CBP 2000).  In Buzzards Bay, the Buzzards Bay Coalition works with
scientists and land use experts to examine the best available current and historical information for
indicators in three categories: pollution, watershed health, and  living resources (BBC 2011). The
Coalition's work has suggested that Buzzards Bay is currently functioning at half its ecological capacity,
therefore affecting the local economy and quality of life (BBC 2011). The Tampa Bay Estuary Program
(TBEP) used historical seagrass cover, light attenuation, and chlorophyll-a data to set restoration targets
for seagrass recovery (Greening and Janicki 2006). Following the establishment of restoration targets,
average gains of seagrass of 142-202 acres per year occurred between 1988 and 1996 (Greening and
Janicki 2006). TBEP also used a historical reference-based approach to set acreage targets for other
intertidal habitats, again resulting in significant gains (Cicchetti and Greening 2011). On the west coast,
the Puget Sound  Partnership has used several types of reference levels, including historical baselines
from the 19th  century, to set restoration targets for eelgrass, wetlands, bald eagles, and resident
southern killer whales (Samhouri et al. 2011). Both TBEP and the Puget Sound Partnership use a type of
indicator "report card" to track changes in indicators with respect to each reference level and a
response gradient (i.e., the grades on a report card).

The first formal case study of the estuarine  BCG (Shumchenia et al. 2015) compiled a biological stressor-
response gradient in time rather than space for Greenwich Bay, a sub-embayment  of Narragansett Bay
(Figure B2-1). In order to implement this concept, a long record of ecological data was needed.
Anecdotal observations of the ecosystem were available as far back as the 1600s for this embayment.
                                                                                           B-31

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
                                                               February 2016
After assembling ecological histories for seagrass extent, benthic communities, and primary
productivity/shellfish, a "minimally disturbed" range of conditions was anchored by observations prior
to 1850. Like the broader Narragansett Bay and many estuaries in the U.S., the relative importance of
environmental stressors changed overtime, but even qualitative descriptions of the biological
indicators' status provided useful information for defining condition levels. This BCG demonstrated that
stressors rarely acted alone and that declines in one biological indicator influenced the declines of
others. For example, in Greenwich Bay the loss of eelgrass was linked to the loss of scallops.
Documenting the timeline of changing stressors helps demonstrate that management actions of the
past may no longer be appropriate or effective for managing the current stressor landscape.

Assembling this pilot example of the estuarine BCG for a  small but data-rich embayment was the first
step toward testing the framework at the estuary- and watershed-scale. To demonstrate the value of
the estuarine BCG framework in  implementing ecosystem-based management beyond local water
quality and habitat management issues, a broader scale application was necessary.
 Levels of biological condition
     for Habitat Structure
 Taxa, indices, and metrics are    "1
 as naturally occurs

 Some decreases in abundance
 of susceptible taxa and/or slight
 increases in tolerant taxa; slight   2
 changes in other metrics
 including patterns of vegetation
 Evident changes in metrics;
 decrease in susceptible taxa
 and/or increase in tolerant taxa;   3
 evident changes in patterns of
 vegetation
 Significant changes in many metrics;
 marked decreases in abundance of
 susceptible laxa (including large and/  A
 or long-lived taxa) and/or evident      •
 increases in tolerant taxa; patterns of
 vegetation significantly altered
 Many susceptible, sensitive, large and/or
 long-lived taxa are absent, with dominance
 in abundance of tolerant taxa; shifts in     C
 species diversity: sizes and densities of    w
 remaining species significantly altered;
 marked changes in patterns of vegetation
      Pre-colonial
      < 1650 AD
 Colonial
1650-1750
 Maritime
1730-1820
 Industrial     Suburban
1800-1945  1945-present
 Susceptible, sensitive, large and/or
 long-lived taxa are mostly absent, with
 extremes in abundance of tolerant taxa;
 marked shifts in species diversity and in
 size spectra of organisms; marked loss
 of natural vegetation may occur
6
                                   Level of exposure
                                to cumulative stressors
                                    Time
Figure B2-1. A chart developed for an estuarine BCG case study in Greenwich Bay, showing a generalized
stressor gradient through time from the 1600s to present (x-axis), biological condition levels (y-axis), and the
responses of three structural attributes: SG = seagrass; BC = benthic community; and PS = primary productivity
and shellfish.
                                                                                                    B-32

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                     February 2016
B2.4  Measuring Overall Estuary Condition

Estuaries and near-coastal systems are influenced not only by stressors and processes within the system,
but also by watershed and oceanic pressures. The overall condition of the estuary derives from the
conditions of all of these internal and external components, their connections, and their combined
functions. Assembling a conceptual ecosystem model incorporating the physical, chemical, and
biological processes that structure the particular system of interest helps assess overall estuary
condition, but it is a complex undertaking.

Several methods and proxies have therefore been developed to evaluate overall condition of the
estuary or coast, all of which could be incorporated into a BCG approach via comparisons to naturally
occurring or minimally disturbed conditions. Robustness of the assessment will improve as spatial  and
temporal coverage increases and as more assemblages and habitats are considered. Ideally, the
biological measures chosen should cover the entire estuarine gradient and incorporate multiple
components of the estuary (e.g., intertidal  and subtidal; primary and secondary production; benthos and
nekton). Approaches that assess overall condition, which can be combined, include (see Cicchetti et al.
2016 for details):
    1.  Use of structural measures including presence of keystone species or other indicator species;
       numbers of species, groups of species, communities, or habitats; or the extent, composition, or
       arrangement of living habitats, or biotopes
    2.  Use of measures of ecosystem function and connectivity, especially those that derive from
       complex interactions in the entire estuary such as energy flows, trophic webs and linkages,
       carbon or nutrient fluxes, production of diverse biomass, nutrient processing,  or resilience to
       changes
    3.  Use of both biological indicators and stressor values, such as in the Greenwich Bay case study
       (Shumchenia et al. 2015). Information from four biological indicators and several attributes was
       synthesized with information from a generalized stressor gradient and additional information on
       specific stressors to describe the current state of the estuary together with the significant  events
       and processes that have shaped it over time

A variety of methods and proxies can be effective, and they are generally selected to best address  the
unique features,  needs, and available data that characterize an individual  water body. For our
Narragansett Bay case study, the relatively large amount of spatial and temporal data  addressing
benthic habitats, as well as documented linkages between habitat quality and water quality,  were  the
impetus for choosing a habitat mosaic approach.

B2.5  Habitat and Biotope Mosaics

Productive estuaries in a natural state are composed of a mosaic of living habitats (Henningsen 2005) or
biotopes, including seagrass beds, oyster reefs, mussel reefs, salt marshes, mangrove forests, clam flats,
and specific soft-bottom benthic communities. The "Habitat Mosaic" approach recognizes that
anthropogenic stress to an estuary leads to destruction of these living habitats and replacement with
other habitats. The method considers the distribution of these living habitats to be a central  part of
estuarine biology. A further assumption is that a mosaic of biotopes that most resembles the mosaic
that would naturally occur in an estuary will improve biological integrity and provide greatest benefit for
the native communities of organisms that have evolved in that setting over millennia.  An assessment
                                                                                           B-33

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
                             February 2016
using this approach compares acreage data from a time period of interest to acreage data from one or
more time periods in the past, and several quantitative measures are available.

A contribution from the TBEP was to apply this biotope mosaic concept to the development of biological
assessment based on estuary-wide changes to quantity (acres) and distributions (relative proportions) of
habitats over time. Ecological priorities for Tampa Bay were to "Restore the Historic Balance" of critical
habitats in percent compositions of biotope mixes relative to an undisturbed  historic benchmark, as well
as to restore total acres of all living habitats, to the extent possible (Figure B2-2). Tampa Bay
stakeholders and the public were invested in the quantity and diversity of valued habitats, and the
concept of "Restore the Historic Balance" resonated with this community. The appealing visual aspects
of this method proved effective at communicating estuarine condition and  developing stakeholder
visions and goals, which led to management actions and  environmental results. This method can be
used together with other approaches as an important component in the management of estuaries,
linking environmental goals to biotope acres and biotope metrics under the BCG framework.
            New Habitat Restoration Goals
            Support  a  Balanced Approach
            The first update of TBEP's Habitat Master Plan in 15 years
            was completed in 2010, recommending expansion of
            two key habitats — low-salinity salt marshes and salt
            barrens — critical to maintaining biodiversity in the bay
            watershed.

            The revised Habitat Master Plan validates the original
            "Restoring The Balance" approach adopted in 1995,
            that called for restoring habitats in relative proportion to
            their historic acreages in 1950.

            Under "Restoring The Balance." more than 5.000 acres
            of coastal wetland and upland habitats have been
            restored or enhanced in the Tampa Bay watershed since
            1995.  Some 7,600 acres of seagrasses, the benchmark
            barometer of the bay's health, have been recovered
            since  1982. Additionally, 19 of 28 sites priority land
            acquisition sites have been completely or partially
            purchased, and eight of those have undergone at least
            some restoration.
               CASEINPOINT
                                           dtr Donna Bttllenbtit-h
Mangroves continue to
expand faster than other
tidal wetland habitats, so
more salt marshes and
salt barrens need to be
created to maintain the
historic mosaic of habitats.
and ensure that the bay
continues to support a
diversity of birds, fish and
other creatures. Therefore,
the new goals call for
maintaining the current
mangrove coverage of
15,139 acres, while increasing
the amount of low-salinity
tidal marshes by another
1,918 acres and salt barrens
by another 840 acres to keep
pace. Having such specific
goals helps bay managers
focus restoration efforts on
priority habitats and track
their progress in meeting the
goals.

A Tampa Bay Habitat
Restoration and Protection
Partnership composed of
agencies and organizations
involved in bay restoration
was formed in 2011, to
further improve regional
coordination and
cooperation in identifying
and implementing restoration
and mitigation.
Figure B2-2. TBEP graphic describing "Restore the Balance" (TBEP 2012).
                                                                                                 B-34

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                     February 2016


B2.6  Using Biological Condition Gradient Concepts to Re-Assess Narragansett
Bay Benthic Habitat Quality

This section is from "A re-assessment of Narragansett Bay benthic habitat quality between 1988 and
2008" by Emily J. Shumchenia, Marisa L Guarinello, and John W. King; submitted to Estuaries and
Coasts.14

There are currently efforts to measure whole estuary condition in Narragansett Bay over the past
several decades. The structural measures of benthic biotope extent, composition and arrangement, are
used together with information about two apparently dominant stressors over this time period:
anthropogenic nutrient inputs and warming. Although an explicit estuarine BCG has not yet been
constructed using these data, the framework that is used to present and interpret these data uses BCG
concepts. Furthermore, this scientific analysis is viewed as an important first step toward  constructing a
BCG explicitly. The acceptance of the interpretation of BCG data by the scientific community is an
important pre-cursorto estuarine BCG construction, and ensures it will be credible in presentations to
local management and regulatory agencies.

B2.6.1 Introduction to Narragansett Bay
Narragansett Bay is the second-largest estuary on the east coast of the U.S. (328 km2) and has the most
densely populated watershed, shared between the states of Rhode Island and Massachusetts.
Narragansett Bay is best conceptualized as a bay with several relatively distinct regions and a general
north to south gradient of enrichment  patterns from the heavily populated and narrow Providence River
Reach, to the more ocean-influenced Open Bay consisting of the East and West Passages on either side
of Conanicut Island (Valente et al. 1992; Nixon et al. 2009; Raposa 2009). In addition, a few Shallow
Embayments (e.g., Greenwich Bay and associated coves) form distinct regions of the Bay.  Overall,
Narragansett Bay is known as a phytoplankton-based temperate ecosystem with a mean depth of 8.6 m
and a mean flushing rate of 26 days (Pilson 1985; Nixon et al. 1995). Freshwater input is relatively low
(100 m3 s"1), with the result that the mid-bay is generally well mixed (Nixon et al. 2005). Salinity follows a
down-bay gradient from 20 psu at the  head to 32 psu at the mouth of Narragansett Bay. The annual
temperature varies from about 0 to 24°C. Sediments of  Narragansett Bay are mainly clayey silt and sand-
silt-clay (McMaster 1960).

Narragansett Bay lies near the boundary between the Gulf of Maine/Bay of Fundy (also known as
Acadian) ecoregion to the north and the Virginian ecoregion to the south (Spaulding et  al. 2007). This
boundary is defined by the departure of the Gulf Stream from the coast across the northwestern
Atlantic, with generally more  cold-water species  north of Cape Cod. Recent shifts in marine species
distribution and abundance near this boundary are driven in part by climate change (Oviatt 2004; Collie
et al. 2008; Pinsky et al. 2013). Therefore, Narragansett Bay, which, like most estuaries  is likely
experiencing climate-change-related ecological changes (Nixon et al. 2009), provides an excellent case
study for other estuaries that may experience climate-driven ecological shifts and oligiotrophication in
the future.
14 The text of this section has been altered to include heading, table, and figure numbering consistent with the rest
of the chapter.
                                                                                         B-35

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                      February 2016


62.6.2 Why Benthic Biotopes?
Soft sediment benthic biotopes, i.e., abiotic environments and associated assemblage of species (Connor
et al. 2004; Costello 2009; Davies et al. 2004), are particularly useful for monitoring patterns of organic
enrichment in time and space (Pearson and Rosenberg 1978) because they are effective integrators of
cumulative stressors such as eutrophication and hypoxia (Pearson and Rosenberg 1978; Valente et al.
1992; Germane et al. 2011). The structure of surface sediments and the composition, or successional
stage, of benthic communities are linked to the degree of organic loading to a water body (Rosenberg
2001) and readily indicate recent (weeks to months) water quality conditions (Cicchetti et al. 2006;
Shumchenia and King 2010a). Comprehensive characterizations of benthic biotopes at the whole estuary
scale are rare because of the high level of detail required to populate biotope classifications. Benthic
biotopes that have been defined for a sub-embayment of Narragansett Bay include 'Ampelisca on
shallow mud' and 'Spiochaetopterus on deep coarse sand' (Shumchenia and King 2010b). These coastal
marine biotopes comprise mosaics at the landscape scale (Bostrom et al. 2011) and are ecologically
meaningful units for conservation and management purposes (Salomidi et al. 2012). Biotope mosaics
are interrelated and functionally connected such that a change to one biotope may affect others, as well
as the entire ecosystem (Bostrom et al. 2011). To date, studies to characterize patterns and quantitative
change over time in benthic biotope mosaics have been mostly limited to seagrass, mangrove and
saltmarsh  ecosystems due to the utility of aerial photography in these environments (Bostrom et al.
2011; Pittman et al. 2011; Cicchetti and Greening 2011; Zajac 2008). The composition of a biotope
mosaic and how it changes over time may indicate degradation or recovery of an ecosystem (Dunning et
al. 1992; Wiens  et al.  1993; Pittman et al. 2007), and thus monitoring of biotope mosaics can help assess
the effects of human  alterations and multiple stressors on coastal marine ecosystems (Cicchetti and
Greening 2011). However, benthic biotope characterization for whole-estuary assessments by
traditional sampling methods is incredibly labor-intensive due to the collection, sorting, and
identification of benthic samples.

One method that has increased the efficiency of benthic assessments  is sediment profile imagery (SPI).
SPI  is a rapid reconnaissance technique that delivers clear images of benthic biotopes regardless of
water column turbidity  (Germano et al. 2011). Ideally, SPI images capture an area including the
sediment-water interface and up to 20 cm  below, i.e., the most biologically active zone of the sediment
column.

In 1988, SPI was used in the first comprehensive survey of benthic habitat quality in Narragansett Bay
(Figure B2-3) in  the context of organic enrichment from wastewater treatment facility (WWTF)
discharges (Germano and Rhoads 1988; Valente et al.  1992). The 1988 SPI study provided the first in-situ
snapshot of benthic processes in Narragansett Bay soft sediments. In fact, most researchers were
unaware and "surprised" by the proportion of the bottom that had been exposed to high levels of
organic deposition and low concentrations of dissolved oxygen (DO) (Granger et al. 2000). Many of the
sites identified as having excessive organic enrichment and degraded benthic habitat were in the
Providence River Reach or Shallow  Embayment sub-regions of the Bay; sites near WWTF outfalls, in
coves, or other spatially constricted areas that received effluent (refer to  Figure B2-3; Valente et al.
1992).
                                                                                          B-36

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
February 2016
71°30'0"W 71°20'0'W 71°10'0'W
PROV
"Shallow Embayments"
AC-1
* .AC-2
WC-1
"AC~3 WC-2
GREENWICH BAY WC-3,
*GB-1
EG GC-3
WWTF «GC.2
• COVE
GC-1
•PWR-2
PWR-1 •
ALLEN
• .AH-2
*AH-3
N
A
NARRAC
10km
DENCE . ^k~^-- Providence
*PR-1 River
*PR-2
Fields
Point ~~~-~-^^
*PR-3 Taunton
* PR-4 |V8r
"Providence PR'5 \
River \i
Reach" .», *PR-6
~"^. • PR-/
WARWICK/-) OB-1.
PR-8 «PR-9
*• . «PR-10 OB-2. o
• FALL RIVER
PR-13 . .PR-11 OB-3.\
• • *PR-12 \
• • • nS7 MOUNT
*OB-4 HOPE
^^^B •OB-9 (OB-6 BAY
• , *OB-5
• , Prudence f
OB-10 |s|anc| OB-28
"Open Bay" OB-11 Q.
•c-B-12 W
•OB-17 . •OB-15'OB-13 5
, OB-16 OB-14 lil
OB-18 ^ |
^ ^ *OB-25 §
^ fe *OB-24
ANSETT •„„ •
O OB-19 OB-27
O
NEWPORT
*OB-20 OB"22
*OB-26
"OB-21
RHODE ISLAND SOUND
41°50'0"N
41°40'0"N
41°30'0"N
Figure B2-3. Locations within Narragansett Bay, Rhode Island USA, where sediment profile images were taken in
1988 and 2008. Three sub-regions used in the analyses are highlighted: Providence River Reach stations (PR-);
Open Bay (OB-); Shallow Embayments (inset). Note locations of Fields Point and East Greenwich (EG) Waste
WWTF at labels

62.6.3 Changes in the Narragansett Bay Stress Gradient
Since the 1988 study, there has been a great deal of human intervention and human-mediated change
in Narragansett Bay and its watershed. These human actions translate directly to changes in the stress
gradient experienced by habitats and biota.

Human population in the watershed has increased by about 200,000 since the early 1980s to a total of
about 2 million people (Nixon et al. 2008), increasing both impervious surfaces and WWTF loads.
Between 1980 and 1995, the Field's Point WWTF in Providence (responsible for approximately 55% of
                                                                                           B-37

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                     February 2016
total effluent discharged directly to the Bay) transitioned from being considered by the
U.S. Environmental Protection Agency one of the worst in the country to one of the best (Nixon and
Fulweiler 2012). The Field's Point plant initiated secondary treatment of its sewage in June of 1988, just
months before the benthic habitat quality assessment took place. In the 1990s, seasonal nutrient- and
stratification-driven hypoxia was discovered in upper Narragansett Bay, and has since been monitored
by state and academic programs (Deacutis 2008; Codiga et al. 2009). In 2003, a large fish kill occurred in
Greenwich Bay due to a confluence of unique hydrological conditions and organic loading that caused
severe hypoxia and anoxia (RIDEM 2003). The fish kill resulted in media and political attention, and in
2004, the Rhode Island Department of Environmental Management (RIDEM) issued a statutory mandate
to eleven WWTFs within the upper  Narragansett Bay watershed to reduce summer season nitrogen
discharges to the Bay between 48%-65% with respect to 1995-1996 levels (RIDEM 2005). As of January
2008, the Field's Point WWTF had not yet implemented nitrogen removal procedures and five smaller
WWTFs had implemented biological nitrogen removal upgrades (RIDEM 2008).

Most monitoring efforts since the 2003 fish kill and 2004 nutrient reduction mandate have focused on
DO data to evaluate compliance with  water quality standards (RIDEM 2005) and highlight the summer
recurrence of bottom water hypoxia in upper- and mid-Narragansett Bay (Bergondo et al. 2005;
Deacutis et al. 2006; Melrose et al. 2007; Codiga et al. 2009). Despite recommendations for monitoring
at four "critical boundaries" to detect long-term changes to benthic enrichment of the Bay bottom
(Germano and Rhoads 1988), only limited-term temporal studies of the benthos have been conducted
since (e.g., Cicchetti et al. 2006;  Calabretta and Oviatt 2008; Shumchenia and King 2010a). However,
analyses of the available data do suggest a shift in benthic community composition due to organic
enrichment between the 1950s and 1980s (Frithsen 1990). SPI surveys conducted for purposes other
than benthic habitat quality assessment and/or at different sites and times of year than the 1988 study
suggest a potential decline in benthic  habitat conditions in the Providence River Reach between 1975
and 1988 (Myers and Phelps 1978) and another potential slight decline in the upper Bay between 1988
and 1994 (Diaz 1995).

Exactly 20 years after the 1988 study (i.e., August 2008), and using the same SPI techniques, the same
sites were revisited to reassess benthic biotope status in the context of environmental changes in the
intervening years,  including recent improvements in wastewater treatment aimed at curtailing organic
loading. Using the same image analysis approach (Germano et al. 2011) on both 1988 and 2008 data
sets, the abiotic and biotic features of the surface and near-surface environment were classified into
benthic biotopes to compare biological and physical processes between surveys. To assess benthic
condition throughout the estuary, SPI results were analyzed using a biotope mosaic assessment
approach (Cicchetti and Greening 2011). The spatial distribution, composition and diversity of benthic
biotopes throughout Narragansett Bay in 1988 and 2008 were compared and related to any observed
trends in biotope condition to changes in organic loading. Whether or not oligotrophication of
Narragansett Bay will result from climate changes and anthropogenic nutrient reductions already set in
motion, consistent, comparable, and frequent monitoring of the benthos will be necessary for scientists
and managers to be able to track, investigate, and respond to these stressors.

B2.6.4 Methods
In planning the 2008 data collection and analysis, similarities with the 1988 study were maximized.  The
2008 survey was conducted during the same time of year as the 1988 study (mid-August), over roughly
the same number of consecutive days (four in 2008; five in 1988) and during the same tidal stage (neap).
Planning in this way helped to ensure an improved ability to compare images between years versus
previous SPI comparisons in Narragansett Bay (Germano and Rhoads 1988; Diaz 1995). Despite these
                                                                                         B-38

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                     February 2016


precautions, some limitations need to be considered. These limitations were apparent in both the data
collection and analysis phases of the study and are described more fully below.

Station locations were originally chosen to define Bay-wide trends in benthic habitat quality. The
majority of stations were located in  depths ranging from 5 to 20 m and in unconsolidated sediments.

At each station an Ocean Imaging Systems Inc. (Falmouth, MA) digital sediment profile camera was
deployed for three to five replicate drops. The coordinates of the first camera drop at each station were
compared with the 1988 target location.

Analysis of 2008 Images and Re-analysis of 1988 Images
Original, analog printed black and white images from the 1988 survey were obtained from archives at
the Graduate School of Oceanography, University of Rhode Island, and scanned into digital format at a
resolution of 300 pixels per inch. The 2008 digital images were uploaded to the analysis computer in .jpg
format at 335 pixels per inch. Because of differences in image quality between the two surveys (i.e.,
analog vs. digital, black and white vs. color, scanned vs. raw digital), there were limitations to the types
of quantitative measurements that could be compared between the sets of images. For example, the
measurement of the apparent redox potential discontinuity (aRPD) depth is a key indicator of benthic
habitat quality and a significantly contributing variable of the multi-parameter Organism-Sediment Index
(OSI) that was used to summarize conditions at each station in 1988 (Valente et al.  1992). The depth of
the aRPD is often "over-estimated" in black and white images, and thus comparisons between black and
white and color images would be particularly susceptible to error (Diaz 1995). Therefore, OSI  and aRPD
values between 1988 and 2008 could not be compared. The original 1988 images were re-analyzed using
the same method as the 2008 images. This analysis approach relied on a biotope classification derived
from the abiotic and biotic surface and subsurface features visible in the SPI images.

All SPI photos were imported into Adobe Photoshop CSS and image brightness and contrast were
adjusted manually to increase the detectability of habitat features such as tubes and burrows. For each
station in each survey, each available replicate was examined and the image with the least disturbance
and deepest prism penetration was  selected for analysis. Information on sediment  grain size, surface
features, and subsurface features was recorded following the protocol described in Rhoads and
Germano (1982;  1986).

Sediment Grain Size
Sediment descriptors represented the major modal class for each image. Organic-rich mud, mud, sandy
mud,  sand and gravel were distinguishable classes. Coarse-grained sediments were indicated by shallow
prism penetration and suggested physically dominated habitats. Fine-grained sediments were indicated
by deeper prism  penetration and suggested that the seafloor environment was depositional and less-
frequently physically disturbed. Organic-rich muds were characterized by fine-grained sediments and
very deep prism penetration, little to no visible surface oxidation, and minimal surface disturbance or
roughness.

Surface Features
Surface descriptors included both biogenic and physical features, such as amphipod and worm tubes,
epifauna (e.g., snails, crabs), shells, macroalgae, bacterial mats (e.g., Beggiatoa sp.), feeding
pits/mounds, bedforms, and roughness. The presence of these features was noted  to indicate any
recent disturbance and the degree and nature of biological activity at each station.
                                                                                          B-39

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                     February 2016


Subsurface Features
The presence of burrows, infaunal feeding voids, infauna and gas voids were recorded in each image.
Subsurface features such as burrows, feeding voids, and infauna indicated biologically active
environments, whereas the presence of gas-filled voids at depth indicated high rates of methanogenesis
and anaerobic respiration (Rhoads and Germano 1986) associated with high rates of organic matter
decomposition.

Benthic Biotope Classification
From the abiotic and biotic surface and subsurface descriptors, each image from 1988 and 2008 was
assigned to a benthic habitat/biotope based on a SPI classification scheme previously described for
Narragansett Bay (Diaz 1995). Using guidance from the Coastal and Marine Ecological Classification
Standard (FGDC 2012) to classify an image, the sediment, surface and subsurface descriptors were
summarized  into a short phrase such  as 'Ampelisca spp. beds of low to high density, with other small
tube-building and shallow burrowing  fauna on organic-rich mud.' These descriptions were kept
consistent among images and grouped by dominant biota and/or sediment type into eight biotopes
(Table B2-2).

Table B2-2. Descriptions of the eight biotopes based on observed dominant biota and sediment type to which
sediment profile images were assigned.
Biotope
Ampelisca spp. beds of low to high density, occasionally with other small tube-building and shallow-burrowing
fauna on substrates ranging from organic-rich muds to sand. Beggiatoa spp. or Mulinia lateralis may be present.
Organic-rich muds with various mixes of tolerant species such as Beggiatoa spp., tube-building polychaetes, and
shallow burrowing fauna.
Burrowing fauna on mud with shell hash. Tube-building polychaetes or deep burrowing fauna may be present.
Burrowing fauna on mud. Tube-building polychaetes, larger tube-builders such as Chaetopterus, or deep
burrowing fauna may be present.
Burrowing and tube-building fauna on sandy mud.
Crepidula bed on mud. Mobile crustaceans, gastropods or Beggiatoa spp. may be present.
Very coarse sands with shell hash. Rafting macroalgae or Crepidula beds may be present.
Hard sands with epibenthic sponges, rafting or attached macroalgae, and/or mobile gastropods.
Code
AM
UN.SF
UN.SH
UN.SI
UN.SS
SH.SI
SH.SA
SA
Comparison of 1988 and 2008 Surveys Using Biotope Mosaic Approach
To assess the influence of survey repositioning on interpreted biotope change, the relationship between
biotope change and the 2008 field distance from the 1988 target location was examined. Sites that were
classified as 'changed' were a mean 60.7 m from the target location (n=25) and sites classified as 'no
change' were a mean 56.5 m from the target location (n=15), indicating that repositioning likely had
little to do with change detection.

Once images were assigned to one of eight biotopes, biotope diversity Bay-wide and within each Bay
region (Providence River Reach, Shallow Embayments and Open Bay) was assessed for each survey. The
ratios of biotopes Bay-wide and among Bay regions were measured to assess the composition and
spatial structure of the benthic biotope mosaic. These ratios were calculated using only the subset of
stations that could  be directly compared between years. If a biotope changed between 1988 and 2008,
the type of biotope change was noted.
                                                                                          B-40

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                     February 2016
62.6.5 Results
Fifty-two of the original 56 stations were resampled in 2008. Between the two surveys, 38 total stations
could be directly compared at the biotope level. Change between 1988 and 2008 was examined using a
biotope mosaic approach, where the proportions and spatial arrangement of biotopes Bay-wide and
among three Bay regions was characterized.

The composition and spatial arrangement of benthic biotopes differed markedly between the 1988 and
2008 surveys despite identical levels of benthic biotope diversity. Over half (58%) of the stations visited
changed biotope between 1988 and 2008, with the bulk of that change occurring in the Providence River
Reach and Shallow Embayments. These spatial patterns of benthic biotope change provided information
about where to focus attention for this analysis and suggested focal areas for monitoring future
changes.

B2.6.6 Ampelisca Biotopes Mark "Critical Boundaries"in the Stress Gradient
Biotopes dominated by Ampelisca spp. tubiculous amphipods increased > 5-fold between 1988 and
2008, and expanded into the more urban, anthropogenically-stressed Providence River Reach (Figure
B2-4).

The prominence of ampeliscid amphipods (e.g., Ampelisca spp.) within Narragansett Bay benthic
biotopes dates back at least to an early detailed benthic study of Greenwich Bay (Stickney and Stringer
1957). It has been suggested that ampeliscids are organic  enrichment opportunists (McCall 1977). There
is also debate as to whether ampeliscids serve as indicators of impending hypoxia (Levin et al. 2009) or
of improving conditions (Diaz et al. 2008; Rhoads and Germano 1986). A recent study in Greenwich Bay
(i.e., one of the Shallow Embayments in this study) of benthic response to water quality changes on the
order of weeks to  months showed that ampeliscids colonized quickly and indicated improving water
quality (Shumchenia and  King 2010a). Ampeliscid tube structures have been associated with increased
biogenic activity and oxygen penetration into the sediment (Diaz et al. 2008) and increased hard clam
abundance (MacKenzie et al. 2006), but have also contributed to the exclusion of other tube-dwelling
species (Santos and Simon 1980). In Jamaica Bay, New York, amphipod productivity was so high that it
was likely more than sufficient to support the entire local  winter flounder (Pleuronectes americanus)
population, with Ampelisca abdita making up 88% of the diet of juveniles (Franz and Tanacredi 1992). In
Narragansett Bay, winter flounder juveniles have been most abundant in the Providence River Reach
and Shallow Embayments recently (Meng et al. 2005) so increases in Ampelisca biotopes in these Bay
regions  may have already benefited this important fish species. Ampeliscids do require large quantities
of organic matter to sustain "mat" densities (Franz and Tanacredi 1992; McCall 1977), which signals
eutrophic conditions. It was estimated that 500 g carbon m~2 yr"1 is required to maintain Ampelisca spp.
tube mats in Boston Harbor, approximately 60 miles to the north (Diaz et al. 2008). Assuming this
relationship is relevant to Narragansett Bay, there has likely been enough carbon produced historically
via primary productivity on an annual basis in the Providence River (mean of 559 g C m~2 yr"1), Greenwich
Bay (mean 219-254 g C m~2 yr"1), and the Open Bay north of Prudence Island (mean of 517 g C m~2 yr"1) to
support these communities (Oviatt et al. 2002). Given that primary productivity historically comprised
an estimated 80% of the total organic carbon input to the  Bay (Nixon et al. 1995), it is likely that there
are sufficient amounts of organic matter stored in the sediments on which these ampeliscids thrive.
                                                                                         B-41

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
                                            February 2016
                                                  Open Bay
                                                     (1)
                          1988
                           Shallow
                       Embayments
                              (4)
                 Open Bay
                    (8)
                        Shallow
                        Embayments
                        (3)
                                                                  Providence
                                                                  River Reach
                                                                  P)
                                                                   'Shallow
                                                                   Embayments
                                                                   (2)
                                          'Providence
                                          River Reach
                                          (5)
                                                   Open Bay
                                                      (9),
                           2008
                                                            Shallow
                                                            Embayments
                                                            (6)
                          Providence ^ (2
                          River Reach
                              (2)
                        Providence
                        River Reach
                        (7)
                         Open Bay
                            (1)
/Shallow
 Embayments
 (1)
Figure B2-4. Composition and spatial arrangement of the Narragansett Bay benthic biotope mosaic in 1988 (a)
and 2008 (b) AM = 'Ampelisca spp. beds'; UN.SI = 'burrowing fauna on mud'; UN.SF = 'organic rich muds with
various mixes of tolerant species'.

Dense Ampelisca spp. communities in areas with high organic input and good water quality have been
previously observed within Narragansett Bay and in Boston Harbor (Stickney and Stringer 1957; Diaz et
al. 2008). The cessation of primary sewage discharges to Boston Harbor (Massachusetts, USA) in the
early 1990s appears to have "set the stage" for "widespread increases" in Ampelisca spp. throughout
the harbor. Prior to 1992, organic loading was high but water quality may have been too poor to allow
Ampelisca spp. to thrive (Diaz et al. 2008). In the late 1990s and early 2000s, several years after the
sewage outfall relocation, subsequent declines in Ampelisca spp. tubes were associated with the
reductions in organic loadings to the harbor and the eventual depletion of sediment organic inventories
(i.e., surface sediment total organic carbon) (Diaz et al. 2008). Unlike in Jamaica Bay, reductions in
organic matter and lower numbers of Ampelisca spp. have apparently had a positive effect on fish
species, as the recreational fishing community has made note of significant recent increases in winter
flounder populations in Boston  Harbor (Powers 2015).
                                                                                             B-42

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                      February 2016


It is possible that a pattern similar to the Boston Harbor example is currently occurring in Narragansett
Bay. In 1988 we observed conditions that did not favor widespread Ampelisca biotopes: stations with
high organic loading and surface sediments that indicated poor water quality conditions (Valente et al.
1992). Between 1988 and 2008, conditions theoretically became increasingly favorable for Ampelisca
biotopes: management strategies to reduce organic loadings and improve water quality were initiated.
In 2008, an increase in the proportion of Ampelisca biotopes Bay-wide was observed, and especially in
areas  where organic loading was known to be previously high. Water quality monitoring programs
continue to record hypoxic events/5 but it is possible that hypoxia occurs now over a smaller area, with
less frequency and/or intensity than previous events (the first Bay-wide DO monitoring program did not
begin  until 1999; Prell et al. 2004). Regardless, because continuous monitoring of benthic biotopes did
not occur between  1988 and 2008, it is impossible to determine where Narragansett Bay is "located"
along  the similar trajectory observed in Boston  Harbor. However, if future surveys classify and analyze
the proportions of benthic biotopes Bay-wide, Bay-wide benthic biotope quality could be:
    1.  Staying the same—i.e., Ampelisca biotopes are maintaining position via existing sediment
       organic matter inventories under good  or improving water quality
    2.  Improving—i.e., a decrease in Ampelisca biotopes coupled with increases in 'burrowing fauna on
       mud,, any other benthic biotopes, or even the appearance of new benthic biotopes, as well as a
       benthic  biotope mosaic showing Ampelisca biotopes remaining only in regions with formerly
       high organic loading
    3.  Declining—i.e., a decrease in Ampelisca biotopes coupled with  an increase in 'organic rich muds
       with various mixes of tolerant species' and a benthic biotope mosaic similar to 1988

62.6.7 Recommendations for Future Benthic Biotope Monitoring in Narragansett Bay
In the initial report of the 1988 SPI survey (Germane and Rhoads 1988), the authors propose a benthic
monitoring strategy aimed specifically at tracking the future movement of Bay gradients in organic
enrichment and any associated changes to benthic biotope (habitat) quality. They suggest that future
monitoring should focus on four critical boundaries separating high quality benthic habitats from
organically-enriched areas: the southern edge of the Providence River Reach, the mouth of Greenwich
Bay, the mouth of the Taunton River, and the southwest edge of Prudence Island.  Rather than monitor
the full Bay-wide suite of original stations, the proposed strategy concentrated station locations around
the four critical boundaries,  revisiting 21 original stations and adding 19 new locations to increase the
detectability of gradient movement (Figure B2-5a). The direction of movement of these boundaries in
habitat quality would indicate  expanding zones of enrichment (boundaries move down-Bay, away from
shore) or decreases in enrichment (boundaries  move up-Bay, shoreward).
15 See http://www.dem.ri.gov/bart. Accessed February 2016.
                                                                                          B-43

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
                                                      February 2016
                      GREENWICH BAY
                                           *  *
                    EG
                    WWTF

                        \      OB-8
                        GREEMWCH
                          COVE     QB-9 *
PR-13  PR-12|_«
                                                  PR-11
       *H.OB-7
                                                      OB-6
                                                   OB-14
                        5km
                                              Providence
                                                River
                        5km
Figure B2-5. Proposed benthic biotope monitoring strategy for detecting changes in organic enrichment and
habitat quality from the (a) original 1988 study and (b) modified considering the results of this study. Closed
circles = existing stations; star symbols = proposed new stations. Hatched area = monitoring focal area.
                                                                                                      B-44

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                     February 2016


By re-sampling the full Bay-wide suite of original 1988 stations, the results of this study have shown that
these gradients moved up-Bay and shoreward as of 2008. To account for these changes, an updated
strategy proposes moving the critical boundaries further up-Bay and shoreward, and similarly adds
stations to a number of original locations to improve the detectability of future changes (Figure B2-5b).
In this updated monitoring strategy, 22 original stations would be revisited and 15 new locations would
be added to detect any further movement of the habitat quality gradient up-Bay and shoreward. This
new monitoring strategy focuses on the Providence River Reach, Greenwich Bay, and the upper portion
of the West Passage. A focus on these areas of Narragansett Bay targets biotopes in close proximity to
the major WWTFs discharging directly to the Bay. Due to the rapid nature of SPI data acquisition, this
survey plan could likely be completed in 1-2 days. One monitoring plan option could be to visit these
37 stations annually in August during neap tide, and then visit the full suite of ~50 original stations every
5 years in  order to provide a Bay-wide perspective at a lesser interval. Using the benthic biotope
classification developed for this study, future results can be entered into the time series and interpreted
in the context of the shifting critical boundaries of organic enrichment in the Bay.

B2.6.8 Conclusions
The Ampelisca biotope may be the most important biotope to track in Narragansett Bay. Ampelisca
biotopes appear to follow the "critical boundaries" of organic enrichment, and Ampelisca tubes can exist
in such dense aggregations that they are likely important prey sources for the demersal fish of
Narragansett Bay. Demersal fish species in Narragansett Bay have declined in number over the past 47
years and  especially since 1980, concurrent with increases in water temperature and decreases in
chlorophyll concentrations (Collie et al. 2008). When the critical boundaries of organic enrichment are in
the more shallow, protected (constricted) regions of the Bay as in 2008, robust Ampelisca biotopes may
serve as critical habitats for juvenile demersal fish such as winter flounder. When the critical boundaries
of organic enrichment existed in deeper, less protected waters as in 1988, fewer Ampelisca biotopes
were observed. Therefore, benthic biotope composition and quality in 1988 may have  represented
poorer conditions for the protection and growth of juvenile demersal fish. With future warming and
decreasing anthropogenic nutrient inputs, Ampelisca biotopes should be monitored more frequently as
potential indicators of patterns in organic enrichment and important fish habitat.

Overall, biotope mosaics can be very useful for tracking habitat change if the data are available to
develop such analyses. This case study demonstrated that continuous-coverage habitat data are not
required in order to take a mosaic approach. The proportions of key habitats Bay-wide and the changes
in their spatial arrangements were measurable from point-sample observations. Two observation
periods (i.e., 1988 and 2008) do not enable trends in Bay-wide habitat quality to be gleaned, so a
conceptual model was developed for what future observations of biotope type, proportion, and
arrangement could mean for overall estuarine habitat quality. This type of conceptual model is
essentially a BCG, and could help managers respond quickly according to the latest monitoring results.
Work will continue to better develop these data in a more explicit estuarine BCG format for use as a
communications tool.

B2.7 Status of This Effort
Work continues with the Narragansett Bay Estuary Program (NBEP) to apply the results of this latest
case study, as well as concepts developed in the Greenwich Bay case study, to their next State of the
Watershed Report, due to be released in 2016. These data will support the reporting of a benthic habitat
quality indicator for Narragansett Bay and a proposed program for monitoring future changes in benthic
habitat quality. While not explicitly using the BCG framework, NBEP is considering a habitat  mosaic
                                                                                          B-45

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                     February 2016


approach with two other habitat types to be included—salt marsh and seagrass. The estuarine BCG work
group will contribute a short case example of the BCG approach for these three habitat quality
indicators in the NBEP State of the Watershed Report to introduce the estuarine BCG framework,
highlight similarities between it and the existing NBEP indicator framework, and demonstrate its
potential usefulness for management. A large amount of quantitative data and analyses have been
assembled by the NBEP for their latest State of the Watershed Report. It is possible that the NBEP
community of partners will be interested in further developing BCG concepts, as well as potentially
developing quantitative thresholds for ecosystem metrics to  stressor levels in the near future.

The broader theme of linking benthic habitat quality to common estuarine stressors (e.g., organic
enrichment, climate change) will continue to be explored through partners including the Narragansett
Atlantic Ecology Division16 lab, the Southeast New England Program,17 NBEP, and the Buzzards Bay
Estuary Program. The Narragansett Bay and Buzzards Bay estuaries share similar ecosystem attributes
and have experienced similar stressor histories, but have different supporting data sets, monitoring
programs, and scientific studies. This regional  umbrella allows exploration and testing of the ability of
the estuarine BCG framework to offer common measures of  biological condition among these
ecosystems.

Appendix B2  References

Bald, J., A. Borja,  I.  Muxika, J. Franco, and V. Valencia. 2005. Assessing reference conditions and physico-
      chemical status according to the European Water Framework Directive: A case-study from the
      Basque Country (Northern Spain). Marine Pollution Bulletin 50:1508-22.

BBC. 2011. State  of Buzzards Bay. Buzzards Bay Coalition, New Bedford, MA.
      http://www.savebuzzardsbay.org/document.doc?id=645. Accessed January 2016.

Bergondo, D.L, D.R. Kester, H.E. Stoffel, and W.L. Woods. 2005. Time-series observations during the low
      sub-surface oxygen events in Narragansett Bay during  summer 2001. Marine Chemistry 97:90-
      103.

Borja, A., and D. Dauer. 2008. Assessing the environmental quality status in estuarine and coastal
      systems: Comparing methodologies and indices. Ecological Indicators 8:331-337.

Borja, A., D.M. Dauer, and A. Gremare. 2011. The importance of setting targets and reference conditions
      in assessing marine ecosystem quality. Ecological Indicators 12:1-7.

Bostrom, C., S.J. Pittman, C. Simenstad, and R. Kneib. 2011. Seascape ecology of coastal biogenic
      habitats: Advances, gaps, and challenges. Marine Ecology Progress Series 427:191-217.

Calabretta, C.J., and C.A. Oviatt. 2008. The response of benthic macrofauna to anthropogenic stress in
      Narragansett Bay, Rhode Island: A review of human stressors and assessment of community
      conditions. Marine Pollution Bulletin 56:1680-1695.
16 Narragansett Atlantic Ecology Division is part of EPA Office of Research and Development.
17 The Southeast New England Program is part of EPA Region 1.
                                                                                          B-46

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                      February 2016
CBP. 2000. Chesapeake 2000 Agreement. Chesapeake Bay Program, Annapolis, MD.
      http://www.chesapeakebay.net/documents/cbp  12081.pdf. Accessed January 2016.

Cicchetti, G. 2010. Summary of a Technical Workshop Held October 28 and 29, 2009: Developing a
      Biological Condition Gradient for Narragansett Bay. Report AED-10-088. U.S. Environmental
      Protection Agency, Atlantic Ecology Division.

Cicchetti, G., S. Latimer, S. Rego, G. Nelson, J. Bergen, and L Coiro. 2006. Relationships between near-
      bottom dissolved oxygen and sediment profile camera measures. Journal of Marine Systems
      62:124-141.

Cicchetti, G., and M. Pryor. 2010. Summary of the Estuarine BCG Workgroup November 2008 Workshop,
      and a  Proposed Organizing Framework for Bioassessment of Estuaries. Report AED-10-087. U.S.
      Environmental Protection Agency, Atlantic Ecology Division.

Cicchetti, G., and H. Greening. 2011. Estuarine biotope mosaics and  habitat management goals: An
      application in Tampa Bay, FL, USA. Estuaries and Coasts 34:1278-1292.

Cicchetti, G., M.C. Pelletier,  M. Pryor, S.K. Jackson, P. Bradley, S.P. Davies, C.F. Deacutis, K.J. Rocha, D.L
      Santavy, and E.J. Shumchenia. 2016. A Biological Condition Gradient Framework for Management
      of Estuaries and Coasts. EPA/600/R-15/287. U.S. Environmental Protection Agency, Office of
      Research and Development, Atlantic Ecology Division, Narragansett Rl.

Codiga, D.L, H.E. Stoffel, C.F. Deacutis, S. Kiernan, and C.A. Oviatt. 2009. Narragansett Bay hypoxic event
      characteristics based on fixed-site monitoring network time series: Intermittency, geographic
      distribution, spatial synchronicity, and interannual variability.  Estuaries and Coasts 32:621-641.

Collie, J.S., A.D. Wood, and H.P. Jeffries. 2008. Long-term shifts in the species composition of a coastal
      fish community. Canadian Journal of Fisheries and Aquatic Sciences 65:1352-1365.

Connor, D.W., J.H. Allen, N. Golding, L.  Kerry, L.M. Lieberknecht, K.O. Northen, and J.B. Reker. 2004. The
      Marine Habitat Classification for Britain and Ireland Version 04.05. In JNCC (2015) The Marine
      Habitat Classification for Britain and Ireland Version 15.03.
      http://incc.defra.gov.uk/MarineHabitatClassification. Accessed January 2016.

Costello, M.J. 2009. Distinguishing marine habitat classification concepts for ecological data
      management. Marine Ecology Progress Series 397:253-268.

Davies, C.E., D. Moss, and M.O. Hill. 2004. EUNIS Habitat Classification Revised 2004. Report to European
      Environment Agency,  European Topic Center on Nature Protection and Biodiversity.

Davies, S.P., and S.K. Jackson. 2006. The Biological Condition Gradient: A descriptive model for
      interpreting change in aquatic ecosystems. Ecological Applications 16:1251-1266.

Deacutis, C.F. 2008. Evidence of Ecological Impacts from Excess Nutrients in Upper Narragansett Bay. In
      Science for Ecosystem-based Management: Narragansett Bay in the 21st Century, eds. A.
      Desbonnet and B.A. Costa-Pierce, pp. 349-382. Springer, New York.
                                                                                           B-47

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                     February 2016


Deacutis, C.F., D.W. Murray, W. Prell, E. Saarman, and L Korhun. 2006. Hypoxia in the upper half of
      Narragansett Bay, Rl, during August 2001 and 2002. Northeastern Naturalist 13:173-198.

Diaz, R.J. 1995. Benthic Habitat Classification of Selected Areas of Narragansett Bay and the Providence
      River, Rhode Island. Prepared for Normandeau Associates, Inc., Bedford, NH by R.J. Diaz and
      Daughters, Ware Neck, VA.

Diaz, R.J., D.C. Rhoads, J.A. Blake, R.K. Kropp, and K.E. Keay. 2008. Long-term trends of benthic habitats
      related to reduction in wastewater discharge to Boston Harbor. Estuaries and Coasts 31:1184-
      1197.

Dunning, J.B., B.J.  Danielson, and H.R. Pulliam. 1992. Ecological processes that affect populations in
      complex landscapes. Oikos 65:169-175.

FGDC. 2012. Coastal and Marine Ecological Classification Standard (CMECS) version IV. FGDC-STD-018-
      2012. Federal Geographic Data Committee.
      https://coast.noaa.gov/digitalcoast/publications/cmecs. Accessed February 2016.

Franz, D.R., and J.T. Tanacredi. 1992. Secondary production of the amphipod Ampelisca abdita Mills and
      its importance in the diet of juvenile winter flounder (Pleuronectes americanus) in Jamaica Bay,
      New York. Estuaries 15:193-203.

Frithsen, J.B. 1990. The Benthic Communities within Narragansett Bay. Technical report #NBP-90-28 of
      the Narragansett Bay Estuary Program.

Germano, J.D., and D.C. Rhoads.  1988. Narragansett Bay Sediment Quality Survey: August 1988.
      Technical report #NBP-89-23 of the Narragansett Bay Estuary Program.

Germano, J.D., D.C. Rhoads,  R.M. Valente, D.A. Carey, and M. Solan. 2011. The use of sediment profile
      imaging (SPI) for environmental impact assessments and monitoring studies: Lessons learned from
      the past four decades. Oceanography and Marine Biology: An Annual Review 49:235-298.

Granger, S.L., M. Brush, B.A. Buckley, M. Traber, M. Richardson, and S.W. Nixon. 2000. An assessment of
      eutrophication in Greenwich Bay. In Restoring Water Quality in Greenwich Bay, ed. M. Schwartz.
      Rhode Island Sea Grant, Narragansett Rl.

Greening, H., and A. Janicki.  2006. Toward reversal of eutrophic conditions in a subtropical estuary:
      Water quality and seagrass response to nitrogen loading reductions in Tampa Bay, Florida, USA.
      Environmental Management 38:163-78.

Henningsen, B. 2005. The maturation and future of habitat restoration programs for the Tampa Bay
      estuarine ecosystem, ed. S.F. Treat. In Proceedings of the Tampa Bay Area Scientific Information
      Symposium, BASIS 4, Tampa Bay Regional Planning Council, St. Petersburg, Florida, October 27-30,
      2003, pp. 165-170.

Levin, L.A., W. Ekau, A.J. Gooday, F. Jorissen, J.J. Middelburg, W. Naqvi, C. Neira, N.N. Rabalais, and J.
      Zhang. 2009. Effects of natural and human-induced hypoxia on coastal benthos. Biogeosciences
      Discussions 6:3563-3654.
                                                                                          B-48

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                     February 2016
Lotze, H.K., H.S. Lenihan, B.J. Bourque, R.H. Bradbury, R.G. Cooke, M.C. Kay, S.M. Kidwell, M.X. Kirby,
     C.H. Petersen, and J.B.C. Jackson. 2006. Depletion, degradation, and recovery potential of
     estuaries and coastal seas. Science 312: 1806-1808.

MacKenzie, C.L, Jr., R.A. Pikanowski, and D.G. McMillen. 2006. Ampelisca amphipod tube mats may
     enhance abundance of northern quahogs Mercenaria mercenaria in muddy sediments. Journal of
     Shellfish Research 25:841-847.

McCall, P.L 1977. Community patterns and adaptive strategies of the infaunal benthos of Long Island
     Sound. Journal of Marine Research 35:221-266.

McMaster, R.L 1960. Sediments of Narragansett Bay system and Rhode Island Sound, Rhode Island.
     Journal of Sedimentary Petrology, 30(2), 249-274.

Melrose, D.C., C.A. Oviatt, and M.S.  Berman.  2007. Hypoxic events in Narragansett Bay, Rhode Island,
     during the summer of 2001. Estuaries and Coasts 30:47-53.

Meng, L, G. Cicchetti, and S. Raciti. 2005. Relationships between juvenile winter flounder and multiple-
     scale habitat variation in Narragansett Bay, Rhode Island. Transactions of the American Fisheries
     Society 134:1509-1519.

Muxika, I., A.  Borja, and J. Bald. 2007. Using historical data, expert judgement and multivariate analysis
     in assessing reference conditions and benthic ecological status, according to the European Water
     Framework Directive. Marine  Pollution Bulletin 55:16-29.

Myers, A.C., and  O.K. Phelps. 1978. Criteria of Benthic Health: A Transect Study of Narragansett Bay,
     Rhode Island. Prepared for the U.S. Environmental Protection Agency, Ecosystem Research
     Laboratory, Narragansett, Rl.

Nixon, S.W., S.L. Granger, and B.L. Nowicki. 1995. An assessment of the annual mass balance of carbon,
     nitrogen, and phosphorus in Narragansett Bay. Biogeochemistry 31:15-61.

Nixon, S.W., B.A. Buckley, S.L. Granger, L.  Harris, A.J. Oczkowski, and L. Cole. 2005. Anthropogenic
     nutrient inputs to Narragansett Bay: A  Twenty-five Year Perspective. Narragansett Bay
     Commission and Rhode Island Sea Grant, 52 pp.

Nixon, S.W., B.A. Buckley, S.L. Granger, LA. Harris, A.J. Oczkowski, R.W. Fulweiler, and L.W. Cole. 2008.
     Nitrogen and Phosphorus Inputs to Narragansett Bay: Past, Present, and Future. In Science for
     Ecosystem-based Management: Narragansett Bay in the 21st Century, eds. A. Desbonnet and B.A.
     Costa-Pierce, pp.  101-175. Springer, New York.

Nixon, S.W., R.W. Fulweiler, B.A. Buckley, S.L. Granger, B.L. Nowicki, and  K.M. Henry.  2009. The impact
     of changing climate on phenology, productivity, and benthic-pelagic coupling in Narragansett Bay.
     Estuarine, Coastal and Shelf Science 82:1-18.

Nixon, S.W., and R.W. Fulweiler. 2012. Ecological footprints and shadows in an urban estuary,
     Narragansett Bay, Rl (USA). Regional Environmental Change 12:381-394.
                                                                                          B-49

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                      February 2016


Oviatt, C.A., A. Keller, and L Reed. 2002. Annual primary production in Narragansett Bay with no bay-
      wide winter-spring phytoplankton bloom. Estuarine, Coastal and Shelf Science 54:1013-1026.

Oviatt, C.A. 2004. The changing ecology of temperate coastal waters during a warming trend. Estuaries
      27:895-904.

Papworth, S.K., J. Rist, L. Coad, and E.J. Milner-Gulland. 2009. Evidence for shifting baseline syndrome in
      conservation. Conservation Letters 2:93-100.

Pauly, D. 1995. Anecdotes and the shifting baseline syndrome of fisheries. Trends in Ecology and
      Evolution 10:430.

Pearson, T.H., and R. Rosenberg. 1978. Macrobenthic succession in relation to organic enrichment and
      pollution of the marine environment. Oceanography and Marine Biology: An Annual Review
      16:229-311.

Pilson, M.E.Q. 1985. On the Residence Time of Water in Narragansett  Bay. Estuaries, 8(l):2.

Pinsky, M.L, B. Worm, M.J. Fogarty, J.L. Sarmiento, and S.A. Levin. 2013. Marine taxa track  local climate
      velocities. Science 341: 1239-1242.

Pittman, S.J., C. Caldow, S.D. Hile, and M.E. Monaco. 2007. Using seascape types to explain  the spatial
      patterns offish in the mangroves of SW Puerto Rico. Marine Ecology Progress Series 348:273-284.

Pittman, S.J., R. Kneib, and C. Simenstad. 2011.  Practicing coastal seascape ecology. Marine Ecology
      Progress Series 427:187-190.

Powers, R. 2015. Boston blackbacks. On The Water New England Edition, pp 50-55.

Prell, W., E. Saarman, D.W. Murray, and C.F. Deacutis. 2004. Summer-Season, Nighttime Surveys of
      Dissolved Oxygen in Upper Narragansett Bay (1999-2003).

Raposa, K.B.  2009. Ecological Geography of Narragansett Bay. In An Ecological Profile of the
      Narragansett Bay Estuarine Research Reserve, eds. K.B. Raposa and M.L.  Schwartz, 77-88. Rhode
      Island Sea Grant, Narragansett Rl.

Rhoads, D.C., and J.D. Germane.  1982. Characterization of organism-sediment relationships using
      sediment profile imaging: An efficient method of Remote Ecological Monitoring of the Seafloor
      (REMOTS System).  Marine Ecology Progress Series 8: 115-128.

Rhoads, D.C., and J.D. Germane.  1986. Interpreting long-term changes in benthic community structure:
      A new protocol. Hydrobiologia 142:291-308.

RIDEM. 2003. The Greenwich Bay Fish Kill -August 2003: Causes, Impacts and Responses. Rhode Island
      Department of Environmental Management, Providence, Rl.

RIDEM. 2005. Plan for Managing Nutrient Loadings to Rhode Island Waters. Rhode Island Department of
      Environmental Management, Providence, Rl.
                                                                                          B-50

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                     February 2016


RIDEM. 2008. Rhode Island Wastewater Treatment Facility Information. Rhode Island Department of
      Environmental Management, Providence, Rl.

Rosenberg, R. 2001. Marine benthic faunal successional stages and related sedimentary activity. Scientia
      Marina 65:107-119.

Salomidi, M., S. Katsanevakis, A. Borja, U. Braeckman, and D. Damalas. 2012. Assessment of goods and
      services, vulnerability, and conservation status of European seabed biotopes: A stepping stone
      towards ecosystem-based marine spatial management. Mediterranean Marine Science 13:49-88.

Samhouri, J.F., P.S. Levin, C.A. James, J. Kershner, and G. Williams. 2011. Using existing scientific capacity
      to set targets for ecosystem-based management: A Puget Sound case study. Marine Policy
      35:508-518.

Santos, S.L, and J.L Simon. 1980. Response of soft-bottom benthos to annual catastrophic disturbance
      in a south Florida estuary. Marine Ecology Progress Series 3:347-355.

Shumchenia, E.J., M.L Guarinello, and J.W. King. Submitted. A re-assessment of Narragansett Bay
      benthic habitat quality between 1988 and 2008. In press. Estuaries and Coasts.

Shumchenia, E.J., and J.W. King. 2010a. Evaluation of sediment profile imagery as a tool for assessing
      water quality in Greenwich Bay,  Rhode Island, USA. Ecological Indicators 10:818-825.

Shumchenia, E.J., and J.W. King. 2010b. Comparison of methods for integrating biological and physical
      data for marine habitat mapping and classification. Continental Shelf Research 30:1717-1729.

Shumchenia, E.J., M.C. Pelletier, G. Cicchetti, S. Davies, C.E. Pesch, C.F. Deacutis, and M. Pryor. 2015. A
      biological condition gradient model for historical assessment of estuarine habitat structure.
      Environmental Management 55:143-58.

Spalding, M.D., H.E. Fox, G.R. Allen, N.  Davidson, Z.A. Ferdana, M. Finlayson, B.S. Halpern, M.A. Jorge, A.
      Lombana, S.A. Lourie, K.D.  Martin, E. McManus, J. Molnar, C.A. Recchia, and J. Robertson. 2007.
      Marine Ecoregions of the World: A Bioregionalization of Coastal and Shelf Areas. BioScience
      57(7):573.

Stickney, A.P., and LD. Stringer. 1957.  A study of the invertebrate fauna of Greenwich Bay, Rhode Island.
      Ecology 38:111-122.

Stoddard J.L, D.P. Larsen, C.P. Hawkins, R.K. Johnson, and R.H. Norris. 2006. Setting expectations for the
      ecological condition of streams:  The concept of reference condition. Ecological Applications
      16:1267-1276.

TBEP. 2012. A Tampa Bay Estuary Program Progress Report 2012. Tampa Bay Estuary Program.
      http://www.tbep.org/pdfs/tbep state of bay  2012 ptr reduced.pdf. Accessed January 2016.

USEPA. 2000. Estuarine and Coastal Marine Waters: Bioassessment and Biocriteria Technical Guidance.
      EPA-822-B-00-024. U.S. Environmental Protection Agency, Washington, DC.
                                                                                          B-51

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                      February 2016


USEPA. 2005. DRAFT: Use of Biological Information to Better Define Designated Aquatic Life Uses: Tiered
     Aquatic Life Uses. EPA-822-R-05-001. U.S. Environmental Protection Agency, Washington, DC.

USEPA. 2011. A Primer on Using Biological Assessments to Support Water Quality Management. EPA
     810-R-11-01. U.S. Environmental Protection Agency, Office of Science and Technology and Office
     of Water, Washington, DC.

Valente, R.M., D.C. Rhoads, J.D. Germano, and V.J. Cabelli. 1992. Mapping of benthic enrichment
     patterns in Narragansett Bay, Rhode Island. Estuaries 15:1-17.

Wiens, J.A., N.C. Stenseth, B. Van Home, and R.M. Ims. 1993. Ecological mechanisms and landscape
     ecology. Oikos 66:369-380.

Zajac, R.N. 2008.  Challenges in marine, soft-sediment benthoscape ecology. Landscape Ecology 23:7-18.
                                                                                          B-52

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                     February 2016


B3.  Caribbean Coral Reefs: Benchmarking a Biological Condition

Gradient for Puerto Rican Coral Reefs

Patricia Bradley18 and Deborah Santavy, PhD., USEPA Office of Research and Development, National
Health and Environmental Effects Research Laboratory-Gulf Ecology Division, Gulf Breeze, Florida

B3.1  Background

More than half of the U.S. population lives in coastal counties—areas that border oceans, bays, and
estuaries (NOAA 2014). In the states of Florida and Hawai'i and the territories of Puerto Rico, U.S. Virgin
Islands (USVI), Guam, American Samoa, and the Commonwealth of the Northern Mariana Islands, nearly
everyone lives within 60 miles of the coast, causing impacts to coral reef ecosystems from development,
fishing pressure, and climate change.

Several federal agencies (e.g., the National Oceanic and Atmospheric Administration (NOAA), the
National Park Service, the U.S. Fish and Wildlife Service, and EPA), together with state and territorial
environmental  and natural resource agencies, are responsible for the management and protection  of
U.S. coral reefs.

The Coral Reef  Conservation Act of 2000 (16 USC § 6401 2000) sets forth the requirement for a national
monitoring program to promote the understanding, conservation, and sustainable use of coral reef
ecosystems.  The President's Ocean Action Plan (The White House 2004) directed EPA to develop
biological criteria and assessment methods for states and territories to evaluate the condition of coral
reefs and surrounding marine water quality (Bradley et al. 2008, Bradley et al. 2010). Currently, EPA
Region 2 is working with Puerto  Rico Environmental Quality Board and USVI Department of Planning and
Natural Resources to revise the territorial water quality standards to be more explicitly protective of
coral reefs and other aquatic habitats, including incorporating ALU language into the water quality
standards and using the BCG to develop narrative and numeric coral reef biological  criteria.
Development of a  coral reef BCG is considered a first step towards coral reef biological criteria. The
model  can be used to assess baseline condition. Use of the BCG in this manner could help avoid
predicaments associated with shifting baselines (Pauly 1995) by tracking biological effects of gradual
regional or global stresses (such as coastal/ocean acidification or rising sea surface temperatures).

 In Puerto Rico  in 2011, EPA assembled a panel of U.S. Caribbean coral reef and fisheries experts
(Bradley et al. 2014). During a series of three workshops and numerous webinars over the following 5
years, the experts  described the aquatic assemblages under natural conditions; identified the
predominant regional stressors; described the BCG, including the theoretical foundation and observed
assemblage response to stressors; developed increasingly quantitative decision rules; and began to
calibrate the conceptual model,  populating it with monitoring data.

This case  study summarizes the status of BCG model development by covering  the following topics:
    1)  Development of a Conceptual BCG Framework for Coral Reefs in the Caribbean (workshop 1)
    2)  Development of Quantitative BCG Models for Two Coral Reef Assemblages  (workshops 2 and 3)
18 Patricia Bradley retired from USEPA in October 2015. Since that time, she has joined Tetra Tech, Inc. She can be
reached at: patbradley@comcast.net.
                                                                                         B-53

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
February 2016
       a.  Fish Expert Breakout Group
       b.  Benthic Macroinvertebrates Breakout Group
       c.  Bringing the Two Assemblage Breakout Groups Together
    3)  Enhancing Monitoring Design and Protocols

B3.2  Development of a Conceptual Model
The first workshop provided proof of concept that the BCG can be adapted for coral reef ecosystems
(Bradley et al. 2014). The panel of experts evaluated and ranked coral reef condition from photographs
and videos collected during EPA's 2010 and 2011 coral reef assessment surveys in shallow waters (< 12
m deep) of southwestern Puerto Rico. The biological assemblages considered were stony corals, fishes,
sponges, gorgonians, and benthic macroinvertebrates. Participants examined the visual media, rated the
condition of various coral reefs and provided rationale for their ratings (Figure B3-1). Descriptions of
characteristics relative to natural and degraded ecological condition were captured during facilitated
discussions.
Figure B3-1. Photos from EPA coral reef sites reflect a range of coral reef conditions, from good (left) to
intermediate quality (middle), to severely degraded (right).

The experts proposed a preliminary narrative BCG with four levels of condition (very good-excellent;
good; fair; and poor), and associated physical and biological attributes (Table B3-1).

Table B3-1. Narrative condition levels and associated BCG attributes (Bradley et al. 2014).
Condition level
Very Good
Excellent
(approximate
BCG Level 1-2)
Attribute descriptions
Physical structure: High rugosity or 3D structure; substantial reef built above bedrock; many
irregular surfaces provide habitat for fish; very clear water; no sediment, floes or films
Corals: High species diversity including rare; large old colonies (Orbicella) with high tissue coverage;
balanced population structure (old and middle-sized colonies, recruits); Acropora thickets present
Gorgonians: Gorgonians present but subdominant to corals
Sponges: Large autotrophic and highly sensitive sponges abundant
Fish: Populations have balanced species abundances, sizes, and trophic interactions
Large vertebrates: Large, long-lived species present and diverse (turtles, eels, sharks)
Other invertebrates: Diadema, lobster, small crustaceans, and polychaetes abundant; some large
sensitive anemone species present
Algae: Crustose coralline algae abundant; turf algae present but cropped and grazed by Diadema
and herbivorous fish; low abundance of fleshy algae
Condition: Low prevalence of disease and tumors; mostly live tissue on colonies
                                                                                           B-54

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
February 2016
Condition level
Good
(approximate
BCG Level 3)







Fair
(approximate
BCG Level 4)







Poor
(approximate
BCG Level 5-6)







Attribute descriptions
Physical structure: Moderate to high rugosity; moderate reef built above bedrock; some irregular
cover for fish habitat; water slightly turbid; low sediment, floes or films on substrate
Corals: Moderate coral diversity; large old colonies (Orbicella) with some tissue loss; varied
population structure (usually old colonies, few middle aged and some recruits); Acropora thickets
may be present; rare species absent
Gorgonians: Gorgonians more abundant than levels 1-2
Sponges: Autotrophic species present but highly sensitive species missing
Fish: Decline of large apex predators (e.g., groupers, snappers) noticeable; small reef fishes more
abundant
Large vertebrates: Large, long-lived species locally extirpated (turtles, eels)
Other invertebrates: Diadema, lobster, small crustaceans, and polychaetes less abundant than levels
1-2; large sensitive anemone species absent
Algae: Crustose coralline algae present but fewer than levels 1-2; turf algae present and longer,
more fleshy algae present than levels 1-2
Condition: Disease and tumor presence slightly above background level; more colonies have
irregular tissue loss
Physical structure: Low rugosity; limited reef built above bedrock; erosion of reef structure obvious;
water turbid; more sediment accumulation, floes and films; Acropora usually gone or present as
rubble for recruitment substrate
Corals: Reduced coral diversity; emergence of tolerant species, few or no living large old colonies
(Orbicella); Acropora thickets gone, large remnants mostly dead with long uncropped turf algae
Gorgonians: Gorgonians more abundant than levels 1-3, replacing sensitive coral and sponge
species
Sponges: Mostly heterotrophic tolerant species and clionids
Fish: Absence of small reef fishes (mostly Damselfish remain)
Large vertebrates: Large, long-lived species locally extirpated (turtles, eels)
Other invertebrates: Diadema absent; Palythoa overgrowing corals; crustaceans, polychaetes and
sensitive anemones conspicuously absent
Algae: Some coralline algae present but no crustose coralline algae; turf is uncropped, covered in
sediment; abundant fleshy algae (e.g., Dictyota) with high diversity
Condition: Higher prevalence of diseased corals, sponges, gorgonians; evidence of high mortality;
usually less tissue than dead portions on colonies
Physical structure: Very low rugosity; no or little reef built above bedrock; no or low relief for fish
habitat; very turbid water; thick sediment film and thick floe covering bottom; no substrate for
recruits
Corals: Absence of colonies, those present are small; only highly tolerant species with little or no live
tissue
Gorgonians: Small and sparse colonies; mostly small sea fans; often diseased
Sponges: Heterotrophic sponges buried deep in sediment; highly tolerant species
Fish: No large fishes; only a few tolerant species remain; lack of multiple trophic levels
Large vertebrates: Usually devoid of vertebrates other than fishes
Other invertebrates: Few or no reef invertebrates; high abundance of sediment dwelling organisms
such as mud-dwelling polychaetes and holothurians
Algae: High cover of fleshy algae (Dictyota); complete absence of crustose coralline algae
Condition: High incidence of disease and low or no tissue coverage on small colonies of corals,
sponges, and gorgonians, if present
                                                                                                    B-55

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                     February 2016
Building on the preliminary BCG framework, EPA worked with the expert panel following the first
workshop to prepare a database with which to develop decision rules and to explore technical
approaches to define the BCG Generalized Stress Axis (GSA). A brief discussion of these two topics
follows.

63.2.1 Coral Reef Database
Puerto Rico's coral reef ecosystems are a complex mosaic of habitats, including mangrove forests,
seagrass beds, and coral reefs (Garcia-Sais et al. 2008), and the ecosystems support 69 shallow-water
(< 40 m) scleractinian species, 260 fish species, 46 shallow-water alcyonarian species, and 500 species of
benthic marine algal flora (Ballantine et al. 2008).

Since no single monitoring program collects all the information needed to develop the BCG, EPA is
assembling coral reef assessment data from past Puerto Rico and USVI studies into a single database.
The initial biological data set included data collected by NOAA and EPA for dominant assemblages (fish,
stony corals, and macroinvertebrates). The data were normalized using the taxonomic naming protocols
of the Integrated Taxonomic Information System. The standardized data are in the original format, a
'crosswalk'  with translations in a standardized format and metadata with geospatial  information
following the ESRI standards.

These data  have been uploaded into the STOrage and RETrieval (STORET)19 Data Warehouse, an EPA
data repository for water quality, biological, and physical data, that is used by state and territorial
environmental agencies for reporting their water quality data under the CWA. Data storage in STORET
makes the data available to the experts and others so they can easily view and interpret the results.

63.2.2 Generalized Stress Axis
EPA and the expert panel discussed the concept of a GSA and focused on three stressors that should be
considered  for coral reefs: (1) land-based sources of pollution, (2) fishing pressure, and (3) global climate
change-associated thermal anomalies.

Land-based Sources of Pollution
EPA began stressor axis work by applying the Landscape Development Intensity index (LDI) to demonstrate
the link between land-based human activity and coral reefs in USVI (Oliver et al. 2011) (see Chapter 5).
The LDI is an integrated measure of the intensity of human activities in a landscape or watershed,
estimated by calculating the input of nonrenewable energy to different land use parcels. The premise that
ecological communities are affected by cumulative human impacts in the surrounding watershed was
shown for wetlands (Brown and Vivas 2005). The LDI index was demonstrated to be an effective
landscape indicator of human impact on St. Croix corals and is being developed for Puerto Rico  (Figure
B3-2).
19 More information about STORET is available at: http://www.epa.gov/storet/. Accessed February 2016.
                                                                                          B-56

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
                            February 2016
                                                     *•*
                                 i km
  0   15  30
                 60
                                120
Legend

  •   NOAA 2014 Coral Survey Sites
LDI
^B I 00 - I 41
     1 42 - 1 96
  ^J 1.97-2.76
^B 2.77-4.33
^•4.34-6.31
Figure B3-2. Puerto Rico 12-digit HUC watersheds and NOAA 2014 National Coral Reef Monitoring Program
(NCRMP) coral stations. Watershed LDI values shown on a green-yellow-red continuum, where green indicates
the lowest human activity and red indicates the highest.

Fishing Pressure
Over-fishing has dramatically altered the composition of biological communities on Caribbean coral
reefs and seagrass beds. Large herbivores and carnivores such as turtles, groupers and sharks that were
once abundant are now ecologically extinct (i.e., populations are so greatly reduced relative to past
levels that the species no longer fulfills its former ecological/functional role). The reduction of these
species has resulted in "trophic level  dysfunction" (Steneck et al. 2004), with food chains now
dominated by small fishes and invertebrates (Hay 1984, 1991; Knowlton et al. 1990; Jackson 1997). EPA
is researching a modification of the approach developed by Ruiz Valentine (2013) to estimate fishing
pressure from commercial, recreational, and artisanal (trap) fishers in Puerto Rico (Figure B3-3).
                                                                                             B-57

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
                                              February 2016
                    Fishing effort
                       (average
                      hours/week)
    Traps per
    fishing zone
  (Shivlani & Koeneke 2010;
    Scharer et al. 2004)
             Recreational
              fishing
           (marinas & density of piersj
                        Buffered:
                     10 nm North coast
                     20 nm East, West, &
                       South coasts
Trap density ranked:
  1:0.02-0.59
  2:0.60-1.84
  3: 1.85-4.56
  Buffered
  (depths):
 1:0-50m
0.5: 51-100m
  1: >100 m
 Buffered
 (depths):
1.5:0-2 km
1:2.1-3 km
 0.5:>3km
                                          Log transformed from
                                               0-lt
                                  Effort (commercial) +• Traps + Recreational

Figure B3-3. Fishing pressure in Puerto Rico (modified from Ruiz Valentine 2013).

Global Climate Change
The projected consequences of global climate change include shifts in ocean temperature, precipitation
patterns, sea level rise, carbonate saturation equilibrium of calcite and aragonite, and other
biogeochemical processes in the oceans. Most mass coral bleaching events have been associated with
increased sea surface temperatures, which led to the long-term degradation of coral reefs worldwide
(Brown and Suharsono 1990; Glynn 1991, 2000; Goreau et al. 1992; Goreau and Hayes 1994; Brown
1997; Goenega et al. 1989; Hoegh-Guldberg 1999; Wilkinson et al. 1999; Goreau et al. 2000; Reaser et
al. 2000; Wilkinson 2000, 2004; Wellington et al. 2001; Hughes et al. 2003; Pandolfi et al. 2003;
Sheppard 2003; Hoegh-Guldberg et al. 2007; Knowlton and Jackson 2008; Mora 2008). Increased carbon
dioxide saturation rates in seawater reduce alkalinity and pH, which many believe will impact the
survival and growth rates of calcium carbonate-secreting organisms (e.g., corals, bivalves, and
calcareous algae) (Orr et al. 2005; Gattuso and Buddemeier 2000; Kleypas et al. 1999; Scavia et al. 2002).
EPA has begun research to include threats from  global climate change in the GSA, by using the NOAA
Coral Reef Watch (CRW) thermal history experimental products to estimate the accumulation of thermal
stress at significant levels to corals (Liu et al. 2012; Figure B3-4).
                                                                                               B-58

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
February 2016
                                                              15 April 2013 SSTAnomaly
                                                                         ^_- High 0 96
Figure B3-4. Sea Surface temperature (SST) anomaly: Example of NOAA's high-resolution (5-km ) thermal
anomaly products. SST Anomaly is the difference between daily SST and corresponding daily climatology. Daily
climatology interpolates monthly mean SSTs, and as such detects cooler or warmer temperatures compared to
long-term averages at specific locations (Source: NOAA CRW, Experimental Products, Thermal History).

B3.3  Development of Biological Condition Gradient  Models for Two Coral Reef
Assemblages

Similar to the freshwater streams efforts, the coral reef experts have been working in two breakout
groups: mobile organisms (mainly fish) and sessile organisms (the benthos).

B3.3.1 Fish Breakout Group
The Fish Breakout Group assigned 128 species (fish observed during EPA's 2010 and 2011 surveys in
Puerto Rico) to attributes.20 The stressor categories that the experts considered most relevant to fish
were land-based sedimentation and fishing pressure. For fishing pressure, the experts considered
whether the species was subject to fishing pressure, the category of fishing (recreational, aquarium,  or
commercial), and whether the species was regulated. The experts did not assign any fish to attribute I,
however they brainstormed fish species from the list of species observed historically in Puerto Rican
coral reefs (n= 260) to be assigned to attribute I.
  See Table B3-5 in the Additional Data section after the references for this case study.
                                                                                          B-59

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
                                          February 2016
The experts used several indicators and metrics to distinguish BCG levels, including taxa richness; total
biomass; sensitive taxa; density of damselfish, piscivores, and other fishes. Box plots showing the
distributions of several metrics across the four assessed BCG levels are shown in Figures B3-5 and B3-6.
Total taxa, richness of sensitive taxa, and total biomass were metrics often used  by the panel (Figure
B3-5). In contrast to taxa richness metrics, the percent density and dominance of damselfish increased
at poorer  BCG ratings (Figure B3-6). While several metrics are strongly associated with the panel's
decisions, they relied on no single metric because most of the boxes show substantial overlap with
adjacent BCG levels.
       26
       24
    (0  22
    X  20
    £  18
    J=  16
      14
    L.  12
    "ro  10

    H  6
       4
       2
     (0
     X
     (0
     «S
    Q_
                                  o-
                                                 5E5
E   soooo
o
in
1
,°   5000
                                      Panel BCG
Figure B3-5. Box plots of fish metrics by panel BCG decision, total taxa, sensitive taxa (attributes II + III),
parrotfish taxa, and total biomass.
                                                                                              B-60

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
                                                                                     February 2016
       CC
     
"£ ~   4°
CD '(/>
p §
            20
                 3456
•~     100

CD
O

CD      8°
                                                         60
                                                    .:=!   40
                                                    n
                                                 w

                                                 CD
                                                 W
                                                         20
                                                              3456
    CD




    ||


    CD
    O

    CD
    CL
       100


       so


       60


       40


       20


        0
                                           Panel BCG

Figure B3-6. Box plots of fish metrics by panel BCG decision: Percent density of sensitive taxa (attributes II + III),
percent biomass of sensitive taxa, and percent density of damselfish.
                                                                                                  B-61

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
February 2016
Quantitative rules were developed using the experts' narrative statements and the box plots to assign
numbers to the narrative rules (Table B3-2).

Table B3-2. Reef Fish BCG Rules

Narrative
Quantitative
BCG Level 3 (n = 15)
Total taxa
Proportion all sensitive taxa
Total biomass
Piscivores
Within-family diversity
(av # spp per family)
Parrotfish
Damselfish
Groupers
Reef habitat rule
Hardbottom habitat rule
Richness moderate to high
Small to moderate proportion of richness
Fish biomass moderate to high
Presence of some snappers and other
piscivores (proportion biomass)
Within-family diversity not responsive
Large body parrotfish present
Damsels do not dominate catch
Groupers present
More stringent
Less stringent
nt_total > 15 (10-20)
nt_att23 > 6 (4-8)
bio_total > 35kg/100m2 (30-40)
pb_SP + pb_LP>0
Not used
nt_parrotfish2> 1(0-2)
pd_damsels < 25% (20-30)
nt_grouper >0
Bests of 7 rules
Bests of 7 rules
BCG Level 4 (n = 17)
Total taxa
Proportion all sensitive taxa
Total biomass
Within-family diversity
(av#spp per family)
Parrotfish
Piscivores
Low to moderate diversity
Some sensitive taxa
Low or higher
Within-group diversity of snappers, grunts,
parrotfish, etc. declined
Parrotfish present
At least one Snapper or piscivore present
nt-total>9(4-14)
nt_att23>3(l-5)
bio_total > 11.5kg/100m2 (7-15)
Not used (see level 3)
Not used
Not used
During the calibration phase, the experts reviewed 11 sites, applying the fish rules (Table B3-2) to assign
a BCG level to each site. Using the confirmation sites, the model correctly predicted nine (82% correct).
There were, however, several issues that arose and could be further investigated. The issues are listed
below with possible approaches for resolution.

Fish Size Distributions
Observations of juvenile and adult fish at a reef site might indicate that a full life cycle is supported at
the site, inferring connectivity at the site for certain species. With observation of a single life stage,
assessors were uncertain about the ability of the reef to support recruitment of juveniles or sustenance
of adults. Therefore, in the BCG rating process, experts requested information about the size
distribution (not just enumeration) of the fish observed. This information need is an example of how the
BCG development  process led to expert discussions and recommendations on monitoring protocols of
coral reef systems  in the future, including cross Federal agency coordination (see Section B3.4). The
experts were familiar with critical sizes that might indicate  single or multiple life stages and could relate
the size and life-stage information to the integrity of the reef fish community.

During the field sampling, size was  recorded in 5-cm intervals for all fish species, but association of
juvenile and adult stages has not yet been completed for this data set. A listing of juvenile and adult size
                                                                                             B-62

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                      February 2016


ranges for fish species might be available in the literature or might be created by the experts based on
professional judgment. Enumeration of juvenile and adults (or size distribution) for future rating
exercises would allow calculation of life-stage metrics for reef fish. The life stage metrics might allow
better discrimination of BCG levels and connectivity. There was also discussion about modifying the
attribute descriptions to reflect whether the fish were long lived—slow growth vs. short-lived—fast
growth.

Correlations among Variables
Relationships among fish metrics, coral metrics, and environmental variables were suggested by the fish
breakout group to increase their understanding of the fish community in response to stressors,
environmental factors, and community interactions. A correlation matrix would be fairly quick to put
together for the existing variables in the database, including relationships that were of particular
interest to the experts (e.g., size of fish vs. size and number of fish; fish metrics and coral metrics).

Calibration to the Full Range of Conditions
Most of the consensus ratings for the sites in the data set were rated level 3 or 4. The distinction
between levels 3 and 4 is often most difficult because these levels have vague narrative differences and
metric distributions often overlap among samples rated 3 or 4. A few individual ratings were at level 2,
and a few consensus ratings were made at 5 and  6. There were conceptual rules developed for level 2
and quantitative rules calibrated for levels 5 and 6. Confirmation ratings were only at levels 3 and 4,
leaving the level 2, 5, and 6 rules un-validated. BCG level 1 was not expected to occur and was not
described conceptually or with model rules.

A full range of biological conditions should be observed and rated to complete and confirm the BCG fish
model. A detailed list of possible data to mine was developed that included databases from Puerto Rico,
Florida, USVI, and the MesoAmerican Reef.

Fish Attributes
If the attributes are to become more useful, the fish need to be characterized first for frequency
(commonness or rarity), and then for sensitivity and type of sensitivity. Commonness could be analyzed
in the existing data set or with similar broad-scale surveys in the region. Sensitivity could be derived
from literature, although toxicological approaches for reef fish are not common. In addition,
connectivity attributes could be considered for each fish  taxon. The experts recommended using the
new NCRMP and other NOAA data to tease out the likelihood of seeing various fish species at a site.
They suggested that the data collected by the Puerto Rico Aqueduct and Sewer Authority for CWA
section 301(h) reporting purposes (fish species, habitat, water quality, sediment, coral species),  could be
used to assess for stressor response.

Fish Observation Protocols
The experts recommended an alternate observation method to collect data on presence offish.
Swimming along a transect causes fish to disperse and hide before they could be observed and
identified. In the transect method, the fish counter does  not look under gorgonians, sponges, etc., and it
is not known how this may skew the  data. An alternative method was suggested (Bohnsack and
Bannerot 1986) that involved sitting (floating) still in one location along the transect for a period long
enough that alarmed fish would return or emerge and be observable in the visible range (about 6.5 m).
The experts recommended that a comparability study of the two methods be conducted concurrently at
the same sites.
                                                                                           B-63

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
February 2016
The fish experts also recommended revising the field method for rugosity since the rugosity measure did
not always reflect what the experts observed in the videos. They felt that the NCRMP topographic
complexity survey protocol would provide more information. NCRMP measures minimum/maximum
depth and maximum vertical relief measurements within the entirety of a 25 m x 4 m transect. NCRMP
also estimates topographic complexity within the transect area.

Defining Attribute X: Ecosystem Connectance
The fish group has initiated a discussion on attribute X. Connectivity between coral reefs, mangroves,
sea grass beds, and lagoons provides a complex and dynamic mosaic that is well documented as a
critical ecosystem attribute (Sale et al. 2008; Christensen et al. 2003; Aguilar-Perera and Appeldoorn
2007; McField and Kramer 2007; Mumby et al. 2004, 2008; Meynecke et al. 2008; Pittman et al. 2011).
Higher densities of juveniles in mangroves and seagrass can be attributed to food availability, structural
complexity, shade, and reduced predation (Beck et al. 2001;  Adams et al. 2006; Dahlgren et al.  2006;
Aguilar-Perera and Appeldoorn 2007) (Figures B3-7 and B3-8). Marine organisms may also make
repeated migrations between habitats on various time scales, especially daily and seasonally (Sale et al.
2010). Ecosystem connectivity is therefore an important attribute to include in a coral reef conceptual
model.
Figure B3-7. Mangroves present (modified from Mumby et al. 2004). Red letter "A" shows juvenile grunts, once
reaching a given size in a seagrass bed, moving to mangroves (B). The mangroves serve as an intermediate
nursery habitat before the fish migrate to patch reefs (C), and fish biomass is significantly enhanced on patch
reefs (C), shallow fore reefs (D), and Orbicella reefs (E). Some fish (F), such as certain species of parrotfish, Scarus
guacamaia, are dependent on mangroves and are not seen where mangroves are absent.
Figure B3-8. Mangroves Absent (Modified from Mumby et al. 2004). If the mangroves are not present, then fish
move directly from the seagrass to the patch reefs, appearing on patch reefs (G) at a smaller size and at lower
density, thus more vulnerable to predation.
                                                                                            B-64

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                      February 2016
The fish experts agreed that marine ecosystems are arrayed in space in response to gradients of:

    •   3D structure
    •   depth
    •   water temperature
    •   salinity
    •   energy (wave regime, tide, currents, eddies)
    •   substrate type

The group felt that high-resolution reef bottom topography (LIDAR or other) was critically needed to
allow for better estimation of connectivity. With high-resolution topography, features related to
connectivity would be recognizable and quantifiable.  High-resolution topography would also indicate
elements of rugosity as well as connectivity,  allowing characterization of broad-scale relief and a
possible basis for classification of reefs. This  project would require coordination among multiple
agencies.

63.3.2 Benthic Macroinvertebrate Breakout Group
The coral experts in the benthic group assigned 46 scleractinian and hydrozoan hard coral species found
in the Western Atlantic to attributes  II-VI based on their sensitivity and tolerance to human induced
stressors. Studies documenting the tolerances of coral species to different anthropogenic stressors are
very limited, so most of the assignments  were based on professional consensus. The experts agreed that
thermal anomalies and  land-based stressors were the most critical threats to corals. They assigned each
species to an attribute level separately, for elevated temperature exposure and sediment exposure.21
The experts did not assign any species to attribute I, and only two species to attribute II. Some taxa were
left unassigned when experts had no opinion.

Coral Reef Habitat Classification
The benthic experts wrestled with the fundamental issue of reef classification. Discussions centered on
defining the expectations of which coral species should be found at each site during attempts to assign
species to a BCG level. Reef traits proposed as important to consider for determining optimum species
composition included habitat classification, geology, sea  level change, sediment exposure, and decadal
temperature anomalies (Hubbard 1997; Hubbard et al. 2009; Costa et al. 2009, 2013; Zitello et al. 2009).
Coral reef environments have distinct horizontal and vertical  zones created by differences in depth,
wave and current energy, temperature, and  light (Zitello et al. 2009). The experts considered several
different reef classification systems in an attempt to incorporate most of the critical reef traits discussed
(Adey and Burke 1976; Darwin 1874; Hubbard et al. 2009). They agreed to use the latest edition of
NOAA's Benthic Habitat reef classification as guidance (Costa et al. 2009, 2013; Zitello et al. 2009), which
classifies reef habitat by a hierarchical structure into reef types, geographic zones, and
geomorphological structures (Table B3-3). The benthic experts agreed to only use the fore reef zone for
assigning reefs to BCG condition levels. The fore reef is defined as the area along the seaward edge of
the reef crest that slopes into deeper water on the barrier or fringing reef type (Costa et al. 2013).
Features associated with a non-emergent reef crest (but still  having a seaward-facing slope that is
significantly greater than the slope of the bank/shelf) are also designated as fore reef. Experts agreed
that fore reefs should be further divided  into two  habitats: coral-based bioherms dominated by Orbicella
species and colonized hard bottoms with gorgonian plains  (Williams et al. 2015). This approach should
21 See Table B3-6 in the Additional Data section after the references for this case study.
                                                                                           B-65

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
                                                                  February 2016
provide a template for application to other well-defined coral reef habitats (e.g., deep fore
reef/escarpment with coral reef coverage) to evaluate in the future.

Table B3-3. Benthic habitat classification scheme used to define discrete habitat classes (Adapted
from Costa et al. 2013)
Reef Type
Reef Geographic Zones
Geomorphological Structures*
Barrier Reef
Shoreline Intertidal
Lagoon
Back Reef
Reef Flat
Reef Crest
Fore Reef
Bank/Shelf
Bank/Shelf Escarpment
Fringing Reef
Shoreline Intertidal
Reef Flat
Reef Crest
Fore Reef
Bank/Shelf
Bank/Shelf Escarpment
Non-Emergent Reef
Crest
Shoreline Intertidal
Bank/Shelf (shallow)
Fore Reef
Bank/Shelf (deep)
Bank/Shelf Escarpment
Coral Reef & Hard bottom (Hard)
Aggregate Reef
Aggregated Patch Reefs
Individual Patch Reef
Pavement
Pavement with Sand Channels
Reef Rubble
Rhodoliths
Rhodoliths with Scattered Coral & Rock
Rock/Boulder
Spur & Groove
Unknown
Unconsolidated Sediment (Soft)
Mud
Sand
Sand with Scattered Coral & Rock
Unknown
Other Delineations
Artificial
Unknown
*Geomorphologic structures are not unique to any reef type or reef geographic zone.

A priori habitat classifications incorporated into sampling designs have important consequences for
assigning sites to BCG condition levels. Natural conditions for different reef types are composed of
different coral species, depending on environmental features such as depth, photosynthetic active
radiation (PAR), water flow, and geology, which influence conditions optimal for different coral species.
Linear reefs (Kendall et  al. 2001) in < 12 m depth  and < 3 miles from shore were targeted in the sampling
design for the data used by the experts in BCG development. Linear reefs are not identified in any of the
reef types or geographic zones normally used, and dead geologic reef structure could not be discerned
from live biologically active coral reefs since the linear reef designation was determined by ship sonar.
Detailed information on the environmental features mentioned above is essential for formulating
expectations about the  benthic assemblages in a  natural state, and it should be captured in future
monitoring efforts.

Benthic Attributes for Coral Reefs
The benthic experts decided that the attributes developed for the Freshwater Streams BCG (Chapter 1,
Table 1), especially relying on attributes I-V, did not apply very well to coral reefs benthic communities.
The experts believed that the tolerances of most of the hard coral species probably varied based on the
individual anthropogenic stressor; when multiple ones were combined the total effects could be additive,
neutral or deleterious but largely unknown. The information and metrics the experts used to evaluate and
rate benthic community included: the amount and quality of reef structure; size and density of massive
reef-building coral species; the amount of live coral tissue on coral skeletons; coral colony density; percent
colony mortality; rugosity; fish species richness; density of gorgonians,  sponges, and Diadema; and density
of "weedy coral species" such as Porites astreoides and Siderastrea siderea. The experts believed two
critical metrics were missing which they required to confidently evaluate the sites: presence and absence
                                                                                               B-66

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
February 2016
of different calcareous and fleshy algae and the percent live coral cover as a planar measurement. They
also stated the presence of coral or gorgonian diseases would be helpful in understanding the health
and resiliency of the coral community. Sampling videos were critical for experts to derive the
information necessary to evaluate benthic condition using the BCG approach. Experts recommended
that more emphasis should be directed toward improving the quality of and protocols for the
underwater videos. The BCG evaluation rules employed more benthic assemblages than just hard corals
or sessile benthic invertebrates. The benthic experts included habitat structure, other benthic
assemblages, and algae.  Narrative rules were developed that reflected the expert judgment on critical
elements and processes  recommended for evaluation of coral community condition (Table B3-4).

Table B3-4. Benthic BCG Narrative Rules. Note: quantitative rules have not yet been developed.
BCG Level 1-2 (minimally disturbed)
Stony corals
Rugosity
Macroinvertebrates
Algae
Sponges
> 45% live cover of coral in fore reef habitat
Minimal recent mortality in large reef-building genera (Orbicella, Pseudodiploria, Colpophyllia, Acropora,
Dendrogyra)
Normal frequency distribution of colony sizes within each species size range to include large, medium, and
small colonies (> 4 cm) and presence of recruits (< 4 cm)
Species composition and diversity composed of sensitive, rare species (Isophyllia, Isophyllastrea,
Mycetophyllia, Eusmilia, Scolymia) present in appropriate habitat type
Very low or just background levels of disease, tissue and skeletal anomalies, and bleaching
Orbicella (fore reef), Acropora (back reef, reef crest) colonies dominant reef structure within respective
zones
High rugosity resulting from large living coral colonies, producing spatial and topographical complexity
Diadema abundant
Reef macroinvertebrates (e.g., Lobsters, crabs) common and abundant
Low levels of invertebrate coral predators (Coralliophylia spp., Hermodice spp.)
Minimal fleshy, filamentous, and cyanobacterial algae present
Crustose coralline algae present, with some turf algae
Phototrophic sponges dominate
Low frequency of Clionid boring sponges
BCG Level 3
Stony corals
Rugosity
Macroinvertebrates
Algae
Sponges
> 25% live cover of coral in appropriate habitat
Higher percentage of tissue loss with signs of recent mortality especially on large reef-building genera
(Orbicella, Pseudodiploria, Colpophyllia, Acropora, Dendrogyra)
Frequency distribution of colony sizes within each species size range starting to become skewed to include
fewer medium and small colonies (> 4 cm) and lower number of recruits than expected (< 4 cm)
Species composition and diversity: sensitive, rare species present in appropriate habitat
Low to moderate levels of disease and bleaching
Orbicella and Acropora colonies still dominant (within respective reef geomorphological zones)
Moderate to high rugosity or reef structure resulting from large living reef-forming and dead coral
colonies, producing spatial complexity (or topographical heterogeneity)
Diadema present
Reef macroinvertebrates (e.g., Lobsters, crabs) present
Minimal presence of fleshy, filamentous, and cyanobacterial algae cover
Crustose coralline and turf algae present
Phototrophic sponges present
Low cover and abundance of Clionid boring sponges
                                                                                           B-67

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
February 2016
BCG Level 4
Stony corals
Rugosity
Macroinvertebrates
Algae
Sponges
> 15% live cover of coral in appropriate habitat
Moderate amount of recent mortality on reef-building genera (Orbicella, Pseudodiploria, Acropora,
Dendrogyra)
Mix of sizes: large colonies may be absent, primarily medium and small colonies; low amount of recruits
Species composition and diversity: sensitive species may be absent (Agaricia, Mycetophyllia, Colpophyllia,
etc.), more tolerant spp present (Montastraea cavernosa, Siderastrea siderea, Porites astreoides; at least
some reef-building corals present but not dominant primarily Orbicella)
Moderate to high levels of disease and potential bleaching on corals and sea fans/branching gorgonians
Usually lower rugosity due to old mostly dead coral structure
Palythoa may be present, but not dominant
Moderate to high amount of fleshy, filamentous, and cyanobacterial algae cover
Moderate cover and abundance of Clionid boring sponges
BCG Level 5
Stony corals
Rugosity
Algae
Macroinvertebrates
Sponges
> 1% live cover of coral in appropriate habitat
High tissue mortality on organisms present. Low amount of live tissue remains on mostly small colonies
Low rugosity comprised of mostly dead and eroded coral structure
Coral cover replaced by fleshy, filamentous, and cyanobacterial algae
Palythoa dominant
Highest presence of Clionid boring sponges
Non-phototrophic sponges dominant
Before proceeding to further develop BCG decision rules for assigning sites to BCG levels, the benthic
experts identified several technical issues to address. A brief explanation of these issues follows,
including summarizing expert comments and recommendations.

Correlations between Benthic Metrics and BCG Condition Levels
Exploration of correlative relationships between coral, macroinvertebrate, rugosity, sponge and
gorgonian metrics with different environmental parameters and BCG condition levels could reveal more
narrative and quantitative rules. Relationships revealed through this exercise need to be evaluated for
biological relevancy and strength of association.

Characterizing Full Range of Conditions
Preferably, data used in the development of decision rules for defining BCG condition levels 1-6 should
range from pristine or natural coral reef condition to severely degraded. Experts agreed that the natural
or BCG level  1 probably does not exist in  Puerto Rico, and if it does it would be in one specific reef
system, Mona Island.22  About 70% of the fore reef sites evaluated for the coral reef BCG calibration were
in fair to poor condition or comparable to levels 4 and 5. There were about 30% of sites assigned to BCG
level 3. No sites were assigned to BCG levels 1, 2, or 6. To anchor the BCG model in natural condition,
the experts discussed what data would represent reference sites. Suggestions were made to use
historical data or records from the 1970s or earlier to develop qualitative and quantitative descriptions
for BCG level 1.
  Mona Island is a protected natural reserve of Puerto Rico located in the western part of the main island within
the Mona Channel.
                                                                                            B-68

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                     February 2016
Additional Metrics
The benthic experts strongly recommended additional metrics be provided to adequately evaluate the
condition of coral reef sites. The experts identified the absence of total percent live cover of coral by
species for each site to be a critical information gap. They found population estimates of density and
percent live tissue useful, but percent of live coral cover has been used over the last four decades and
has been adopted as the standard. Experts indicated that using a metric for percent live coral cover was
more intuitive for them to express their professional judgment. The expert panel also discussed the
value of information on the health of coral and gorgonian colonies (e.g., prevalence of disease,
bleaching, predation), an estimate of coral recruitment, a more robust estimate of rugosity, and
observed mortality classified as whether it occurred recently or years ago. They wanted better
characterization of the benthic environment, advising that it should include additional assemblages and
not just corals, gorgonians, and sponges. They were adamant that algae (sub-classified as crustose
coralline, fleshy, filamentous, or cyanobacteria), zoanthids, seagrass, and the type of bare substrate be
incorporated  quantitatively in any future monitoring programs. These data could lead to metrics more
relevant for developing standardized rules to discern different BCG levels.

Additionally, the benthic experts urged monitoring water quality (temperature, pH, turbidity, chlorophyll
a, nutrients, DO,  etc.), sediment, and other significant environmental conditions at long-term fixed
stations measured at the depth of coral communities, not just at the surface. These measurements
could improve understanding of the responses of coral communities to different anthropogenic
stressors and aid in developing the GSA.

Defining Attribute VII: Organism Condition
Organism condition is very important for maintaining coral  reef integrity because increased levels of
coral diseases and bleaching have been identified as major  responses to anthropogenic disturbance,
contributing to the decline of coral reefs. Stressors impacting corals and reef structure can range from
local, regional, or global in scale (e.g., sea temperature anomalies, nutrient enrichment, sedimentation,
sewage, herbicides, pesticides, and coastal acidification). Many coral reef assessments evaluate
diseases, coral bleaching, amount and condition of tissue, and recent vs. old mortality,  so there are data
available for attribute VII.

In the second workshop, the coral experts recommended developing criteria for BCG attribute VII. Of
particular importance would be: the amount of live coral/gorgonian tissue found on colonies as
compared to the exposed bare skeleton, recent vs. old mortality, condition of tissue, and population
demographics including size frequency distribution  and recruitment. Organism anomalies in the BCG
vary from naturally occurring incidence in levels 1 and 2 to higher-than-expected incidence in  levels 3
and 4. In levels 5 and 6, colony tissue is reduced, the age structure of populations indicates premature
mortality or unsuccessful reproduction, and the incidence of serious anomalies is high. As such,
scleractinian corals' sessile nature makes them good candidates for assessing the impacts of exposure to
anthropogenic stressors. A cadre of methods are amenable to exploring impacts of past exposures such
as using sclerochronology to examine skeletal bands, a method that is analogous to examining tree
rings.

63.3.3 Bringing the Two Expert Assemblage Groups Together
The two expert groups (fish and  benthos) worked independently of each other, to develop individual
BCG models focused on their individual assemblages. During the third workgroup meeting the two sets
of experts met together to share information on their progress and to evaluate fore reef sites together
                                                                                           B-69

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                     February 2016


using both the narrative rules (benthic) (Table B3-4) and quantitative fish rules (Table B3-2). This was an
exercise to attempt to understand how the emerging BCG models and decisions rules for the two
assemblages could be combined.

Discussion outcomes and recommendations include:

    •  Experts noted that there was often not congruence between fish and benthic expert ratings.
       Fish communities are influenced by location (distance from shore), connectivity, physical
       structure of the reef, and fishing pressure. Some species are also strongly influenced by the
       characteristics of the living coral community—for food source and/or shelter. Benthic
       communities are influenced by reef type and history, depth, PAR, water quality, past geology,
       and other factors.

    •  The fish experts were able to predict fish species, size, abundance, and trophic level based on
       the habitat location  and type, or to predict the habitat type based upon the fish species found at
       a site. The benthic experts had only a few condition predictors, which resulted in imprecise BCG
       rankings. The benthic experts used more intuition, gestalt, video evidence,  and extensive
       knowledge of historical conditions to assign sites to BCG levels.

    •  In the multi-assemblage group it became clear to the coral experts that the fish experts had a
       process model upon which they heavily relied to assign sites to BCG levels.  The benthic experts
       expressed their need to develop an analogous benthic model to the one fish experts used. All
       experts agreed that  an analysis of time series data sets for coral and other important benthic
       species in the U.S. Caribbean could be employed to develop this model.

    •  Algae could serve as an early response signal to habitat degradation and would  be an important
       component for determining thresholds or tipping points of BCG levels for coral benthic
       community assessments.

    •  The topographical complexity of the coral reef structure and substrate was considered in
       evaluation of site condition. Most of the experts recommended using more recently developed
       methods for rugosity, which use multiple heights (Dustan et al. 2013), instead of the older linear
       chain approach. Additional observations on the quality and quantity of specific substrate types
       could serve as an indicator for potential recruitment of corals, gorgonians, and other benthic
       organisms. For example, if the substrate is covered with diatoms or scuzzy filamentous algae, it
       is untenable for recruitment.

    •  The experts identified a fate and transport model for land-based stressors in the U.S. Caribbean
       as a critical need, which would require coordination among multiple agencies. Water chemistry
       and physical properties were not generally available in the data set used for site evaluation by
       the experts. However, even if available, those data might have been a poor representation of
       conditions tolerated by the fish community because of their movement and water dynamics.
       Fate and transport models would allow characterization of general stressor conditions at each
       sampling site. Sessile corals could prove to provide an excellent record of past and present
       exposures to different anthropogenic stressors as revealed in their skeletons. Once the model
       was calibrated in Puerto Rico, it could be applied elsewhere.

B3.4  Enhancing  Monitoring Designs and Protocols

The experts identified gaps in available coral reef monitoring and assessment data sets that, if
addressed, would enable more  comprehensive and robust assessment of coral reef condition and
                                                                                          B-70

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                     February 2016


support development of a more complete BCG model. The experts discussed the elements of a
monitoring program that would ensure relevant metrics could be provided at different temporal and
spatial scales for assessing coral reef conditions. They also discussed the need for a sampling protocol
starting with a basic design (including measurements essential to obtain the basic required information),
expanding to multiple tiers with increasing complexity and many more measurements to produce more
detailed metrics. The tiered approach would allow selection of an appropriate assessment methodology
ranging from a simple monitoring protocol for screening purposes to a more comprehensive protocol,
dependent upon available funding, time, study objectives, and available personnel.

The experts recommended that the basic sampling unit, assessment unit, and replication approach be
identified for coral reef surveys for screening or assessing general condition of reefs within a study area.
The experts unanimously recommended that a probabilistic sampling design was appropriate for this
purpose. The benthic group further recommended a stratified design  by reef type, selected randomly
from a sampling frame that excludes areas devoid of living coral reefs. It remains unresolved whether
the most appropriate approach would be to use increased replication of shorter transects (10 m) or
fewer transect replicates of longer  length (25 m); this could be resolved by carefully reviewing the
objectives of the study. The experts also considered that more stringent requirements be developed for
setting up transects, with protocols defining how to select the transect placement and direction  relative
to the shore, as  well as how to anchor lines down so they cannot move between the video
documentation  and the surveyors assessment. The experts believed the observed placement of the
transect in the videos was not always representative of the typical reef habitat found at that site, or the
data they expected to see while assigning a BCG level.

B3.5 Conclusion

Work continues on development of a quantitative decision model for the fore reef zone in Puerto Rico
and will include  application of the approach to other well-defined habitats (e.g., deep fore
reef/escarpment with coral reef coverage). As part of the process, EPA is developing a GSA for Caribbean
coral reef systems that includes land-based sources of pollution, fishing pressure, and global climate
change associated temperature anomalies (see Chapter 5). EPA and U.S. Geological Survey are
collaborating on a framework for attribute VII (organism condition) as it could be applied in a coral reef
BCG, and the fish experts will continue to develop attribute X (ecosystem connectance). It is anticipated
that at least one more workshop and a series of webinars will be needed to complete, test, and calibrate
the quantitative decision model for the fore reef zone. Additionally, descriptions  of BCG levels 1  and 2
will be formulated based on historic data and records, as well as evaluation of additional data sets
provided by expert panel.

Additionally, EPA has developed a database that documents tolerance levels for 39 species of corals and
131 species of fish to different stressors, including thresholds, when known. The database was
populated with  environmental condition information and relevant citations of species'
sensitivity/tolerance to stressors from the Encyclopedia of Life and Web of Science  websites. The cited
literature was organized into a species sensitivity library using EndNote. The plan is to expand the
database to other assemblages (e.g., macroinvertebrates, reptiles, seagrasses, sponges, mangroves,
etc.).
                                                                                          B-71

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                     February 2016
Appendix B3 References

Adams, A.J., C.P. Dahlgren, G.T. Kellison, M.S. Kendall, C.A. Layman, J.A. Ley, I. Nagelkerken, and J.E.
      Serafy. 2006. Nursery function of tropical back-reef systems. Marine Ecology Progress Series
      318:287-301.

Adey, W., and R.B. Burke. 1976. Holocene bioherms (algal ridges and bank barrier reefs) of the eastern
      Caribbean. Geological Society of America Bulletin 87:95-109.

Aguilar-Perera, A., and R.S. Appeldoorn. 2007. Variation in juvenile fish density along the mangrove-
      seagrass-coral reef continuum in SW Puerto Rico. Marine Ecology Progress Series 348:139-148.

Ballantine, D., R. Appeldoorn, P. Yoshioka, E. Weil, R. Armstrong, J. Garcia, E. Otero, F. Pagan, C.
      Sherman, E. Hernandez-Delgado, A. Bruckner, and C. Lilyestrom. 2008. Biology and Ecology of
      Puerto Rican Coral Reefs. In Coral Reefs of the USA, ed. B.M. Riegl and R.E. Dodge, pp. 375-406.
      Springer, New York.

Beck, M.W., K.L  Heck Jr., K.W. Able, D.L Childers, D.B. Eggleston, B.M. Gillanders, B. Halpern,  C.G. Hays,
      K. Hoshino, T.J. Minello, R.J. Orth,  P.F. Sheridan, and M.P. Weinstein. 2001. The identification,
      conservation, and  management of estuarine and marine nurseries for fish and invertebrates.
      BioScience 51:633-641.

Bohnsack, J.A., and S.P. Bannerot. 1986. A Stationary Visual Census Technique for Quantitatively
      Assessing Community Structure of Coral Reef Fishes. NMFS Technical Report 41.  National Oceanic
      and Atmospheric Association, National Marine Fisheries Service. Seattle, WA. 15 pp

Bradley, P., D.L. Santavy, and J. Gerritsen. 2014. Workshop on Biological Integrity of Coral Reefs, August
      21-22, 2012, Caribbean Coral Reef Institute, Isla Magueyes,  La Parguera,  Puerto Rico. EPA/600/R-
      13/350. U.S. Environmental Protection Agency, Office of Research and Development, Atlantic
      Ecology Division, Narragansett, Rl.

Brown, B.E. 1997. Coral bleaching: Causes and consequences. Coral Reefs 16:S129-S138.

Brown, B.E., and Suharsono. 1990. Damage and recovery of coral reefs affected by El Nino  related
      seawater warming in the Thousand Islands, Indonesia. Coral Reefs 8:153.

Brown, M.T., and M.B. Vivas. 2005. Landscape development intensity index. Environmental Monitoring
      and Assessment 101:289-309.

Christensen, J.D., C.F.G. Jeffrey, C. Caldow, M.E. Monaco, M.S. Kendall, and R.S. Appeldoorn. 2003.
      Cross-shelf habitat utilization  patterns of reef fishes in southwestern Puerto Rico. Gulf and
      Caribbean Research 14(2):9-27.

Costa, B.M., M.S. Kendall, K. Edwards, G. Kagesten, and T.A. Battista. 2013. Benthic Habitats of Fish Bay,
      Coral Bay and the St. Thomas  East End Reserve. NOAA Technical Memorandum NOS  NCCOS 175.
      Prepared by the NCCOS Center for Coastal Monitoring and Assessment Biogeography Branch,
      Silver Spring, MD.
                                                                                           B-72

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                     February 2016


Costa, B.M., LJ. Bauer, T.A. Battista, P.W. Mueller, and M.E. Monaco. 2009. Moderate-Depth Benthic
     Habitats of St. John, U.S. Virgin Islands. NOAA Technical Memorandum NOS NCCOS 105. NCCOS
     Center for Coastal Monitoring and Assessment Biogeography Branch, Silver Spring, MD.

Dahlgren, C.P., T. Kellison, A.J. Adams, B.M. Gillanders, M.S. Kendall, J.A. Ley, I. Nagelkerken, and J.E.
     Serafy. 2006. Marine nurseries and effective juvenile habitats: Concepts and applications. Marine
     Ecology Progress Series 312:291-295.

Darwin, C. 1874. The Structure and Distribution of Coral Reefs. 2nd ed. Smith Elder and Company,
     London,  UK.

Dustan, P., O. Doherty, and S. Pardede. 2013. Digital reef rugosity estimates coral reef habitat
     complexity. PlosOne 8(2):e57386.

Garcfa-Sais, J.,  R. Appeldoorn, T. Battista,  L. Bauer, A. Bruckner, C. Caldow,  L. Carrubba, J. Corredor, E.
     Diaz, C. Lilyestrom, G. Garcfa-Moliner, E. Hernandez-Delgado, C. Menza, J.  Morell, A. Pait, J.
     Sabater,  E. Weil, E. Williams, and S. Williams. 2008. The State of Coral Reef Ecosystems of Puerto
     Rico. In The State of Coral Reef Ecosystems of the United States and Pacific Freely Associated
     States: 2008. ed. J.E. Waddell and A.M. Clarke. NOAA Technical Memorandum NOS NCCOS 73.

Gattuso, J.P., and R.W. Buddemeier. 2000. Ocean biogeochemistry: Calcification  and CO2. Nature
     407:311-313.

Glynn, P.W. 1991. Coral reef bleaching in the 1980s and possible connections with global warming.
     Trends in Ecology and Evolution 6:175-179.

Glynn, P. W. 2000. El Nino-Southern Oscillation Mass Mortalities of Reef Corals: A Model of High
     Temperature Marine Extinctions? In Carbonate Platform Systems: Components and Interactions,
     ed. E. Insalaco, P. W. Skelton, and T. J. Palmer, pp. 117-133. The Geological Society of London,
     London,  UK.

Goenega, C., V.P. Vicente, and R.A. Armstrong. 1989. Bleaching induced mortalities in  reef corals from La
     Parguera, Puerto  Rico: A precursor of change in the community structure of coral reefs? Caribbean
     Journal of Science 25:59-65.

Goreau T.J., R.H. Hayes, J.W. Clark, D.J.  Basta, and C.N. Robertson. 1992. Elevated Sea  Surface
     Temperatures Correlate with Caribbean Coral Reef Bleaching. Chapter 9 in A Global Warming
     Forum: Scientific,  Economic, and Legal Overview, ed.  R.A. Geyer, pp. 225-255. CRC Press, Boca
     Raton, FL

Goreau, T.J., and R.M. Hayes. 1994. Coral bleaching and ocean 'hot spots'. Ambio 23:176-180.

Goreau T., T. McClanahan, R. Hayes, and A. Strong. 2000. Conservation of coral reefs after the 1998
     global bleaching event. Conservation Biology 14:5-15.

 Hay, M.E. 1984. Patterns offish and urchin grazing on Caribbean coral reefs: Are previous results
     typical? Ecology 65:446-454.
                                                                                           B-73

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                     February 2016
Hay, M.E. 1991. Fish-Seaweed Interactions on Coral Reefs: Effects of Herbivorous Fishes and Adaptations
      of Their Prey. In The Ecology of Coral Reef Fishes, ed. P.F. Sale, pp. 96-119. Academic Press, New
      York.

Hoegh-Guldberg, O. 1999. Climate change, coral bleaching and the future of the world's coral reefs.
      Marine and Freshwater Research 50:839-866.

Hoegh-Guldberg, O., P.J. Mumby, A.J. Hooten, R.S. Steneck, P. Greenfield, E. Gomez, C.D. Harvell, P.F.
      Sale, A.J. Edwards, K. Caldeira, N. Knowlton, C.M. Eakin, R.  Iglesias-Prieto, N. Muthiga, R.H.
      Bradbury, A. Dubi, and M.E. Hatziolos. 2007. Coral reefs under rapid climate change and ocean
      acidification. Science 318:1737-1742.

Hubbard, O.K. 1997. Reefs as dynamic systems. In: Life and Death of Coral Reefs. Birkeland, C. (Ed.).
      Chapman and Hall Publishing, New York, NY. pp. 43-67.

Hubbard, D., W. Ramirez, D. Cuevas, T. Erickson, and A. Estep. 2009. Holocene reef accretion along the
      north side of Bahia Enriquillo (western Dominican Republic): Unique insights into patterns of reef
      development in response to sea-level rise. Proceedings of the 11th International Coral Reef
      Symposium 1:43-47.

Hughes, T.P., A.H. Baird, D.R. Bellwood, M. Card, S.R. Connolly, and C. Folke. 2003. Climate change,
      human impacts, and the resilience of coral reefs. Science 301(5635):929.

Jackson, J.B.C. 1997. Reefs since Columbus. Coral Reefs 16:S23-S32.

Kendall, M.S., C.R. Kruer, K.R. Buja, J.D. Christensen, M. Finkbeiner, R.A. Warner, and M.E. Monaco.
      2001. Methods Used to Map the Benthic Habitats of Puerto Rico and the U.S. Virgin Islands. NOAA
      Technical Memorandum NOS NCCOS CCMA 152. Silver Spring, MD.

Kleypas, J.A., R.W. Buddemeier, D. Archer, J.P. Gattuso, C. Langdon, and B.N. Opdyke. 1999.
      Geochemical consequences of increased atmospheric carbon dioxide on coral reefs. Science
      284:118-120.

Knowlton, N., J.C. Lang, and B.D. Keller. 1990. Case study of natural population collapse: post-hurricane
      predation on Jamaican staghorn corals. Smithson Contributions to Marine Science 31:1-25.

Knowlton, N., and J.B.C. Jackson. 2008. Shifting baselines, local impacts, and global change on coral
      reefs. PLoS Biology 6(2):e54.

Liu, G., C.M. Eakin, J.L. Rauenzahn, T.R.L. Christensen, S.F. Heron, J. Li, W.  Skirving, A.E. Strong, and T.
      Burgess. 2012. NOAA Coral Reef Watch's Decision Support System for Coral Reef Management. In
      Proceedings of the 12th International Coral Reef Symposium, Cairns, Australia, July 9-13, 2012, pp.

McField, M., and P.R. Kramer. 2007. Healthy Reefs for Healthy People: A Guide to Indicators of Reef
      Health and Social Well-being in the Mesoamerican Reef Region. Healthy Reefs for Healthy People
      Initiative. 208 pp. http://www.gstcouncil.org/images/librarv/Healthy Reef 24 complete-l.pdf.
      Accessed February 2016.
                                                                                           B-74

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                      February 2016


Meynecke, J.O., S.Y. Lee, and N.C. Duke. 2008. Linking spatial metrics and fish catch reveals the
      importance of coastal wetland connectivity to inshore fisheries in Queensland, Australia.
      Biological Conservation 141:981-996.

Mora, C. 2008. A clear human footprint in the coral reefs of the Caribbean. Proceedings of the Royal
      Society of London B. 275:767-773.

Mumby, P.J., A.J. Edwards, J.E. Arias-Gonzalez, K.C. Lindeman, P.G. Blackwell, A. Gall, M.I. Gorczynska,
      A.R. Harborne, C.L.  Pescod, H. Renken, C.C. Wabnitz, and G. Llewellyn. 2004. Mangroves enhance
      the biomass of coral reef fish communities in the Caribbean. Nature 427:533-536.

Mumby, P.J., K. Broad, D.R. Brumbaugh, C.P. Dahlgren, A.R. Harborne, A. Hastings, K.E. Holmes, C.V.
      Kappel, F. Micheli, and J.N. Sanchirico. 2008. Coral reef habitats as surrogates of species,
      ecological functions, and ecosystem services. Conservation Biology 22:941-951.

NOAA. 2014. NOAA's State of the Coast website. National Oceanic and Atmospheric Administration.
      http://stateofthecoast.noaa.gov/.

NOAA. 2016. NOAA Coral Reef Watch, Experimental Products, Thermal History.
      http://coralreefwatch.noaa.gov/satellite/product  exp.php. Accessed February 2016.

Oliver, L.M., J.C. Lehrter, and W.S. Fisher. 2011. Relating landscape development intensity to coral reef
      condition in the watersheds of St. Croix, U.S. Virgin Islands. Marine Ecology Progress Series
      427:293-302.

Orr, J.C., V.J. Fabry, O. Aumont, L. Bopp, S.C. Doney, R.A. Feely, A. Gnanadesikan, N. Gruber, A. Ishida, F.
      Joos, R.M. Key, K. Lindsay, E. Maier-Reimer, R. Matear, P. Monfray, A. Mouchet, G. Raymond, R.G.
      Najjar, G.K.  Plattner, K.B. Rodgers, C.L. Sabine, J.L Sarmiento, R. Schlitzer, R.D. Slater,  I.J.
      Totterdell, M.F. Weirig, Y. Yamanaka, and A. Yool. 2005. Anthropogenic ocean acidification over
      the twenty-first century and its impact on calcifying organisms. Nature 437:681-686.

Pandolfi, J. M., R. H. Bradbury, E. Sala, T. P. Hughes, K. A. Bjorndal, R. G. Cooke, D. McArdle,  L.
      McClenachan, M. J. H.  Newman, G. Paredes, R. R. Warner, and J. B. C. Jackson. 2003. Global
      trajectories of the long-term decline of coral reef ecosystems. Science 301:955-958.

Pauly, D. 1995. Anecdotes and the shifting base-line syndrome of fisheries. Trends in Ecology and
      Evolution 10:430.

Pittman, S.J., G.F.  Renchen, R. Clark, C. Caldow, D. Olsen, and R.L Hill. 2011. Applications of  estuarine
      and  coastal applications in marine spatial planning. Treatise on Estuarine and Coastal Science
      1:163-205.

Reaser, J.K., R. Pomerance, and P.O. Thomas. 2000. Coral bleaching and global climate change: Scientific
      findings and policy recommendations. Conservation Biology 14:1500-1511.

Ruiz Valentine, I. 2013. Development and Testing of a Spatial Tool for the Assessment Anthropogenic
      Threats in Coastal and Marine Habitats in Puerto Rico. Ph.D. thesis, University of Puerto Rico,
      Mayaguez, PR.
                                                                                           B-75

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                     February 2016


Scavia, D., J.C. Field, D.F. Boesch, R.W. Buddemeier, V. Burkett, D.R. Cayan, M. Fogarty, M.A. Harwell,
      R.W. Howarth, C. Mason, D.J. Reed, T.C. Royer, A.M. Sallenger, G. James, and J.G. Titus. 2002.
      Climate change impacts on U.S. coastal and marine ecosystems. Estuaries 25(2):149-164.

Sale, P.F., P. Jacob, and J.P. Kritzer. 2008. Connectivity: What it is, how it is measured, and why it is
      important for management of reef fishes. In: R. Grober-Dunsmore and B.D.  Keller (Eds.).
      Caribbean connectivity: Implications for marine protected area management. Proceedings of a
      Special Symposium, 9-11 November 2006, 59th Annual Meeting of the Gulf and Caribbean
      Fisheries Institute, Belize City, Belize. Marine Sanctuaries Conservation Series ONMS-08-07. Silver
      Spring (MD): U.S. Department of Commerce, National Oceanic and Atmospheric Administration,
      Office  of National Marine Sanctuaries, pp. 16-30.

Sale, P.F., H.  Van Lavieren, M.C. Ablan Lagman, J. Atema, M. Butler, C. Fauvelot, J.D. Hogan, G.P. Jones,
      K.C. Lindeman, C.B. Paris, R. Steneck, and H.L Stewart. 2010. Preserving Reef Connectivity: A
      Handbook for Marine Protected Area Managers. Connectivity Working Group, Coral Reef Targeted
      Research & Capacity Building for Management Program, UNU-INWEH.

Sheppard, C.R.C.  2003. Predicted recurrences of mass mortality in the Indian Ocean. Nature 425:294-
      297.

Steneck, R.S., J. Vavrinec, and A.V.  Leland. 2004. Trophic-level dysfunction in kelp forest ecosystems of
      the Western North Atlantic. Ecosystems 7:523-552.

The White House. 2004. U.S.  Ocean Action Plan: The Bush Administration's Response to the U.S.
      Commission on Ocean  Policy. The White House, Washington, DC.

Wellington, G.M., P.W. Glynn, A.E.  Strong, S.A. Navarrete, E. Wieters, and D. Hubbard D. 2001. Crisis on
      coral reefs linked to climate change. EOS, Transactions of the American Geophysical Union 82:1-6.

Wilkinson, C.R. 2000. World-wide coral reef bleaching and  mortality during 1998: A Global Change
      Warning for the New Millennium? In Seas of the Millennium: An Environmental Evaluation,
      Volume III, Global Issues and Processes, ed. C. Sheppard, pp. 43-57. Elsevier, Amsterdam, The
      Netherlands.

Wilkinson, C. 2004. Status of Coral  Reefs of the World: 2004. 2 volumes. Global Coral Reef Monitoring
      Network & Australian Institute of Marine Science, Townsville, AU.

Wilkinson C., O. Linden, H. Cesar, G. Hodgson, J. Rubens, and A.E. Strong. 1999. Ecological and socio-
      economic impacts of 1998 coral mortality in the Indian Ocean: An ENSO impact and a warning of
      future change? AMBIO 28:188-196.

Williams, S. M. P. J. Mumby, I. Chollett, and J. Cortes. 2015. Importance of differentiating Orbicella reefs
      from gorgonian plains for ecological assessments of Caribbean reefs. Marine Ecology Progress
      Series 530:93-101.

Zitello, A.G.,  L.J. Bauer, T.A. Battista, P.W. Mueller, M.S.  Kendall, and M.E. Monaco. 2009. Shallow-Water
      Benthic Habitats of St. John,  U.S. Virgin Islands. NOAA Technical Memorandum  NOS NCCOS 96.
      Silver Spring, MD.
                                                                                          B-76

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
February 2016
Appendix B3 Additional Information: Species Attribute Assignments Made
during the Biological Condition Gradient Workshop

Table B3-5. Fish Species Attribute Assignments made during the BCG Workshop. Fish observed during
EPA's 2010 and 2011 surveys in Puerto Rico were used for assignments to attributes II-VI. Since no
attribute I species were observed during EPA's 2010 and 2011 surveys, fish experts brainstormed a list
based on species known to be found in Puerto Rico.
Scientific Name
Common Name
Frequency (% of Samples)
/ — Historically documented, sensitive, long-lived, or regionally endemic taxa
Epinephelus itajara
Mycteroperca tigris
Mycteroperca bonaci
Epinephelus striatus
Mycteroperca venenosa
Lutjanus cyanopterus
Scarus coelestinus
Scarus coeruleus
Ginglymostoma cirratum
Carcharhinus perezii
Negaprion brevirostris
Sphyrna mokarran
Galeocerdo cuvier
Aetobatus narinari
Dasyatis americana
Lactophrys triqueter
Acanthostracion polygonia
Lactophrys trigonus
Acanthostracion quadricomis
Caranx bartholomaei
Atlantic Goliath Grouper
Tiger Grouper
Black Grouper
Nassau Grouper
Yellowfin Grouper
Cubera Snapper
Midnight Parrotfish
Blue Parrotfish
Nurse Shark
Caribbean Reef Shark
Lemon Shark
Great Hammerhead Shark
Tiger Shark
Spotted Eagle Ray
Southern Stingray
Smooth Trunkfish
Honeycomb Cowfish
Trunkfish
Scrawled Cowfish
Yellow Jack
Not Observed
Not Observed
Not Observed
Not Observed
Not Observed
Not Observed
Not Observed
Not Observed
Not Observed
Not Observed
Not Observed
Not Observed
Not Observed
Not Observed
Not Observed
Not Observed
Not Observed
Not Observed
Not Observed
Not Observed
II— Highly sensitive taxa
Acanthurus coeruleus
Amblycirrhitus pinos
Anisotremus surinamensis
Aulostomus maculatus
Cantherhines pullus
Chaetodon sedentarius
Chromis cyanea
Chromis multilineata
Clepticus parrae
Elacatinus genie
Elacatinus oceanops
Gramma loreto
Haemulon chrysargyreum
Heteropriacanthus cruentatus
Holacanthus ciliaris
Malacoctenus triangulatus
Melichthys niger
Scarus guacamaia
Scomberomorus regalis
Serranus tigrinus
Blue Tang
Red Spotted Hawkfish
Black Margate
Trumpet Fish
Orange Spotted Filefish
Reef Butterflyfish
Blue Chromis
Brown Chromis
Creole Wrasse
Cleaner Goby
Neon Goby
Fairy Baselet
Small-mouthed Grunt
Glasseye Snapper
Queen Angelfish
Saddled Blenny
Black Durgon
Rainbow Parrotfish
Cero
Harlequin Bass
62%
0.7%
2.9%
8.6%
1.4%
0.7%
1.4%
5.7%
0.7%
2.9%
0.7%
1.4%
2.9%
0.7%
0.7%
4.3%
0.7%
0.7%
1.4%
12.9%
III — Intermediate sensitive taxa
Acanthurus chirurgus
Alphestes afer
Balistes vetula
Doctorfish
Mutton Hamlet
Queen Triggerfish
38.6%
0.7%
7.1%
                                                                                  B-77

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
February 2016
Scientific Name
Bodianus pulchellus
Bo di an us rufus
Cephalopholis cruentata
Chaetodon cop/stratus
Chaetodon striatus
Epinephelus adscensionis
Epinephelus guttatus
Equetus punctatus
Haemulon carbonarium
Haemulon macrostomum
Haemulon sciurus
Halichoeres garnoti
Halichoeres maculipinna
Halichoeres radiatus
Hypoplectrus aberrans
Hypoplectrus chlorurus
Hypoplectrus indigo
Hypoplectrus nigricans
Hypoplectrus puella
Hypoplectrus randallorum
Hypoplectrus unicolor
Kyphosus sectator
Lutjanus apodus
Lutjanusjocu
Lutjanus mahogoni
Microspathodon chrysurus
Mulloidichthys martinicus
Myripristis jacobus
Odontoscion dentex
Pareques acuminatus
Pempheris schomburgkii
Pomacanthus arcuatus
Pomacanthus paru
Scarus taeniopterus
Sparisoma chrysopterum
Sparisoma viride
Sphyraena barracuda
Stegastes planifrons
Stegastes variabilis
Thalassoma bifasciatum
Common Name
Spotfin Hogfish
Spanish Hogfish
Graysby
Foureye Butterflyfish
Banded Butterflyfish
Rock Hind
Red Hind
Spotted Drum
Caesar Grunt
Spanish Grunt
Blue-striped Grunt
Yellowhead Wrasse
Clown Wrasse
Pudding Wife
Yellow/belly Hamlet
Yellowtail Hamlet
Indigo Hamlet
Black Hamlet
Barred Hamlet
Tan Hamlet
Butter Hamlet
Bermuda Sea Chubb
Schoolmaster
Dog Snapper
Mahagony Snapper
Yellowtail Damselfish
Yellow Goatfish
Blackbar Soldierfish
Reef Croaker
Highhat
Glassy Sweeper
Gray Angelfish
French Angelfish
Princess Parrotfish
Redtail Parrotfish
Stoplight Parrotfish
Great Barracuda
Threespot Damselfish
Cocoa Damselfish
Bluehead Wrasse
Frequency (% of Samples)
0.7%
10%
4.3%
54.2%
22.1%
4.3%
5%
1.4%
15%
5%
2.9%
21.4%
33.6%
16.4%
0.7%
7.9%
1.4%
0.7%
8.6%
1.4%
2.9%
1.4%
29.3%
0.7%
5.7%
53.6%
6.4%
11.4%
4.3%
2.1%
2.9%
2.9%
12.1%
7.1%
2.1%
59.3%
1.4%
17.1%
4.3%
86.4%
IV— Intermediate tolerant taxa
Abudefdufsaxatilis
Acanthurus bahianus
Anisotremus virginicus
Canthigaster rostrata
Cephalopholis fulva
Coryphopterus glaucofraenum
Diodon hystrix
Gymnothorax moringa
Haemulon aurolineatum
Haemulon flavolineatum
Haemulon plumierii
Holocentrus adscensionis
Holocentrus rufus
Lachnolaimus maximus
Sergeant Major
Ocean Surgeonfish
Porkfish
Sharpnose Puffer
Coney
Bridled Goby
Porcupine Fish
Spotted Moray Eel
Tomtate
French Grunt
White Grunt
Squirrelfish
LongspineSquirrelfish
Hogfish
21.4%
70%
18.6%
25%
5.7%
5.7%
0.7%
2.1%
3.6%
37.1%
6.4%
25.7%
10.7%
1.4%
                                                                                                    B-78

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
February 2016
Scientific Name
Lutjanus analis
Lutjanus griseus
Lutjanus synagris
Monacanthus
Ocyurus chrysurus
Ophioblennius macclurei
Pseudupeneus maculatus
Sargocentron vexillarium
Scarus iseri
Sparisoma aurofrenatum
Sparisoma rubripinne
Stegastes adustus
Stegastes diencaeus
Stegastes leucostictus
Stegastes partitus
Common Name
Mutton Snapper
Gray Snapper
Lane Snapper
Filefish
Yellowtail Snapper
Red Lipped Blenny
Spotted Goatfish
Dusky Squirrelfish
Striped Parrotfish
Redband Parrotfish
Yellowtail Parrotfish
Dusky Damselfish
Longfin Damselfish
Beaugregory
Bicolor Damselfish
Frequency (% of Samples)
2.1%
0.7%
5%
0.7%
59.3%
18.6%
9.3%
1.4%
73.6%
85.7%
21.4%
42.1%
41.4%
38.6%
66.4%
V — Tolerant taxa

Synodusfoetens
Checkered Puffer
Inshore Lizardfish

1.4%
VI — Non-native or intentionally introduced species
Callogobius clitellus
Pterois
Saddled Goby
Lionfish
0.7%
1.4%
N—Taxa not assigned to an attribute
Calamus calamus
Carangoides ruber
Caranx crysos
Chaenopsis ocellata
Emblemariopsis
Eucinostomus gula
Gerres cinereus
Gobiidae
Gymnothorax
Haemulon melanurum
Haemulon parra
Haemulon striatum
Halichoeres bivittatus
Halichoeres cyanocephalus
Halichoeres poeyi
Holacanthus bermudensis
Malacanthus plumieri
Rypticus saponaceus
Serranus baldwini
Sphyraena borealis
Stephanolepsis hispida
Synodus intermedius
Saucereye Porgy
Bar Jack
Blue Runner
Bluethroat Pickle Blenny
Dark Headed Blenny
Silver Jenny
Yellowfin Mojarra
Gobies
Common Moray Eel
Cottonwick
Sailors Choice
Striped Grunt
Slippery Dick
Yellow/cheek Wrasse
Blackear Wrasse
Blue Angelfish
Sand Tilefish
Greater soapfish
Lantern Bass
Northern Sennett
Planehead Filefish
Sand Diver
1.4%
19.3%
5.7%
4.3%
0.7%
0.7%
1.4%
0.7%
2.1%
0.7%
1.4%
0.7%
54.3%
0.7%
0.7%
0.7%
1.4%
2.1%
2.1%
0.7%
0.7%
2.1%
                                                                                                    B-79

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
February 2016
Table B3-6. Coral species assignments to BCG attributes by professional judgment of coral reef
experts. Sediment tolerance was used as a surrogate for landscape stressors and elevated heat
tolerance as a proxy for climate change stressors. The expected density at single site (distribution
within a site) and frequency of occurrence (distribution among sites) were ranked from low to high.
BCG
Attribute
II
II
III
III
III
III
III
III
III
III
III
IV
IV
IV
IV
IV
IV
IV
IV
IV
IV
IV
IV
IV
IV
IV
IV
IV
IV
IV
V
V
V
V
V
V
V
V
V
V
V
V
V
VI
%
Stations
Present1
5.0
11.4
15.0
2.1
17.9
6.4
7.9
0.7
0.0
4.3
13.6
17.1
0.0
35.0
16.4
0.0
9.3
0.7
0.0
15.7
0.0
4.3
2.9
14.3
52.1
3.6
10.0
26.4
0.0
0.0
0.0
0.0
57.1
64.3
0.0
91.4
7.1
40.0
77.9
3.6
92.9
1.4
31.4
0.0
Scientific Name
Isophyllastrea rigida
Isophyllia sinuosa
Acropora cervicornis
Agaricia lamarcki
Col pophy Ilia natans
Dendrogyra cylindricus
Diploria labyrinthiformis
Eusmilia fastigiata
Helioseris cucullata
Madmcis decactis
Millepora complanata
Acropora palmata
Acropora prolifera
Agaricia agaricites
Agaricia humilis
Cladocora arbuscula
Dichocoenia stokesi
Madracis myriaster
Meandrinajacksoni
Meandrina meandrites
Mussa angulosa
Mycetophyllia aliciae
Mycetophyllia ferox
Orbicella annularis
Orbicellafaveolata
Orbicella j ranks!
Poritesfurcata
Porites porites
Scolymia cubensis
Scolymia lacera
Faviafragum
Manicina areolata
Millepora alcicornis
Montastraea cavernosa
Oculina diffusa
Porites astreoides
Porites divaricata
Pseudodiploria clivosa
Pseudodiploria strigosa
Siderastrea radians
Siderastrea siderea
Solenastrea bournoni
Stephanocoenia intersepta
Tubastrea coccinea
Sediment
Tolerance
^L
^L
3
3
3
34
3
3
3
24
3
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
5
5
5

Heat
Tolerance
^L
^L
3
2
3
3
3
3
3
4
2
3
3
2
2
4
3
3
3
3
2
3
2-3
2
2
2
4-5
4
4
4
4
5
2
4-5
4
5
4
4
4
5
4
4
4

Expected
Density at
Single Site
low
low
low
med - low
med
low
med
low
low
low
hi
med
low
med - hi
med - hi
low
low
low
low
low
low
low
med
med
hi
low
low
med
low
low
med
low
med
med
low
hi
med
hi
hi
med
med
low
med

Expected
Frequency of
Occurrence
low
low
med
med
med
med
hi
low
low
med
low3
med low
low
hi
hi
low
med
low
med
med
low
med
hi
hi
hi
hi
med
hi
low
low
hi
low
hi
hi
low
hi
low
hi
hi
hi
hi
low
low

                                                                                           B-80

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
February 2016
BCG
Attribute
X
X
X
X
X
%
Stations
Present1
21.4
9.3
0.0
0.7
4.3
Scientific Name
/UNKNOWN/
Agariciafragilis
Millepora squarrosa
Mycetophyllia daniana
Mycetophyllia lamarckiana
Sediment
Tolerance





Heat
Tolerance


2


Expected
Density at
Single Site


low
deep
deep
Expected
Frequency of
Occurrence


low


1 Total of 140 stations
2 Only about 50% experts expressed an opinion
3 Limited to shallow depths
                                                                                                        B-81

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                     February 2016


B4.  New England:  Using the Biological Condition Gradient and Fish
Index of Biotic Integrity to Assess Fish Assemblage Condition  in Large
Rivers

Chris Voder, Research Director, Midwest Biodiversity Institute, Columbus, Ohio23

This case study examines the correspondence of BCG levels and attributes with a fish IBI developed for
large, non-wadeable rivers in New England (termed "riverine IBI" hereafter) based on work conducted in
2002 through 2009. The principal objective was the development of a BCG-based fish IBI that could be
used to assess and readily communicate the status of New England rivers. Intended applications include
determining the existing status and quality of individual river reaches and the effectiveness of
management efforts aimed at protecting and restoring native fish assemblages including diadromous
species. Using the BCG to better visualize the "as naturally occurs" riverine fish assemblage proved
essential to developing expectations for BCG levels 1 and 2, and describing the incremental changes
predicted in BCG levels 3 through 6. The riverine IBI that was developed during 2002-2007 in Maine
(Yoder et al. 2008), and further tested and applied throughout New England in 2008-2009 (Yoder et al.
2015), served as the quantitative scale of measurement along the BCG.

The riverine fish fauna of New England has a unique make-up due to its comparative isolation from
drainages to the west and north (Curry 2007). A narrative BCG for fish assemblages for New England was
developed on the basis of expert judgment and historical knowledge of pre-settlement conditions.
Numeric thresholds are proposed based on alignment of BCG attributes with the metrics from the
riverine IBI. The resulting BCG-based IBI model can be used to communicate aquatic life condition (fish)
in New England Rivers and,  based on historic knowledge, describe the fish assemblage expected in an "as
naturally occurs" condition. Thus, a site is assigned to a BCG level based on its IBI score.

The riverine IBI and the attendant BCG were initially developed for the cool-coldwater, moderate-high
gradient riverine ecotype as it is the most common type throughout New England. Many New England
rivers also support several diadromous fish species that comprised a  significant ecological and
commercial aspect of riverine fish assemblages, at least historically. To better address this important
component of the BCG, a supplemental set of IBI metrics were developed to specifically measure the
diadromous component of the fish assemblage. This was done for two purposes: (1) to use the
diadromous metrics (termed "diadromous metrics" hereafter) as a distinct indicator of whether a river is
supporting a fishery on the basis of unobstructed access to freshwater (anadromous) or salt water
(catadromous)  for spawning, and (2) to retain the function of the riverine IBI for assessing rivers where
diadromous species are not expected to occur. Common diadromous fish include sea run Atlantic
salmon (Salmo  salar), Atlantic sturgeon (Acipenser oxyrhinchus), shortnose sturgeon (Acipenser
brevirostrum), American eel (Anguilla  rostrata), sea lamprey (Petromyzon marinus), striped bass
(Morone saxatilis), and three species of Clupeidae known as river herring.24 Historically these species
comprised a significant component of a New England riverine fish assemblage, but their numbers have
been significantly reduced since the mid to late 19th century.  Restoration efforts are currently
widespread and are focused on improving upstream and downstream passage for diadromous fish. The
23 CYoder@MWBinst.com
24 In New England, river herring include alewife (Alosa pseudoharengus), blueback herring (Alosa aestivalis), and
American shad (Alosa sapidissima).
                                                                                        B-82

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                     February 2016


development of a BCG-based riverine IBI provides the opportunity to quantitatively determine how
these restoration efforts affect the whole fish assemblage.

B4.1  Background
A systematic approach to the assessment of fish assemblages in the large, non-wadeable rivers of New
England was initiated in 2002 for the purpose of developing a large rivers fish IBI. Data collection
occurred first in Maine through 2007 and was then extended into the remainder of New England in
2008-2009 as part of a Regional EMAP (REMAP) project. The aggregate effort produced an extensive
and detailed region-wide coverage of riverine fish assemblages and habitat (Figure B4-1). The resulting
database was sufficient to develop and test a riverine IBI. This project also paralleled the early
development and piloting of the BCG by EPA (Davies and Jackson 2006). An expert advisory panel was
formed to provide advice about New England riverine fish assemblages and evaluate the
correspondence between  BCG attributes and riverine IBI metrics. This included using the  BCG to
describe the composition of the "as naturally occurs" riverine fish assemblages for the cool-coldwater,
moderate-high gradient riverine ecotype that prevailed throughout most of New England prior to the
extensive modification of rivers in the 18th and 19th centuries.

B4.2  Riverine Index of Biotic Integrity Development—Summary of Key Tasks

A systematic and tractable methodology for sampling riverine fish in  New England did not exist when
this project was conceived in 2001. As such, methodological issues had to be addressed first and then
followed by the organization of species' autecological information, both of which are essential steps in
the complementary development of a BCG and IBI. These were followed by the more traditional tasks of
selecting and testing candidate IBI metrics, selecting final metrics, calibrating the metrics, and testing
the riverine IBI across a gradient of conditions ranging from  reference to highly impacted. The following
are the major tasks that were accomplished starting with the initial efforts in Maine and then expanding
to rivers throughout New England (referencing the project documents that deal with each):
    1)  Developing an effective and systematically employed sampling methodology (Yoder et al. 2006a)
       with first phase of development occurring for Maine rivers with applicability across New England
       taken into account;
    2)  Establishing a sufficient spatial and temporal database in Maine, then testing throughout New
       England (Yoder et al. 2006b, 2008);
    3)  Describing the autecology of the extant fish fauna to support deriving and testing candidate
       metrics in Maine but application considered across New England (Yoder et al. 2006b, 2008);
    4)  Differentiating major lotic ecotypes in Maine and New England  (Yoder et al. 2006b, 2008);
    5)  Visualizing the expected New England fish assemblages along the BCG (Yoder et al. 2008);
    6)  Establishing reference condition for Maine and  application throughout New England (Yoder et
       al. 2008);
    7)  Deriving a fish IBI for the moderate-high gradient riverine ecotype, first in Maine and then
       applied throughout New England (Yoder et al. 2008); and,
    8)  Testing the BCG and IBI initially developed for Maine with data sets that represent a range of
       conditions and stressors across New England (Yoder et  al. 2015).
                                                                                          B-83

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
 February 2016
           New England Rivers Fish Assemblage Sites
          Data Collected 2002 - 2009
                  150
                                Miles
                                                             Sampling Sites
                                                             Total N = 508
                                                                     lesri
                                                                   MBI
Biodiversity
Institute
Figure B4-1. Locations of fish sampling in the non-wadeable rivers of New England for the Maine rivers and New
England REMAP fish assemblage assessment projects, 2002-2009.
                                                                                    B-84

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                      February 2016


The IBI development process generally followed that described by Hughes et al. (1998), which has
guided many leading examples in North America and elsewhere (see summary by Yoder and Kulik 2003),
and Mebane et al. (2003). The tasks related to IBI derivation and testing and the BCG are summarized in
the order in which they were accomplished.

B4.2.1 Task 1. Development of an Effective Sampling Methodology
A tractable sampling methodology did not exist for riverine fish assemblages in Maine when this project
was first proposed. The lack of an approach was likely due to the status of rivers as "working rivers" that
supported hydroelectric power production and the transport of logs, the latter of which rendered most
rivers as inaccessible for fish sampling (log driving was discontinued in 1975). The approaches developed
for sampling large rivers in the Midwestern U.S. in the 1970s and 1980s were applied and modified
accordingly to suit the prevailing conditions in Maine. A perceived  obstacle was the comparatively low
conductivity of the water, which  threatened to make electrofishing less effective. However, this was
overcome by making adjustments to the equipment during an initial testing phase in 2002. Other
aspects of the methodology were also established at this time (Yoder et al. 2006a).

64.2.2 Task 2. Development of Sufficient Spatial and Temporal Database
The development of a database that is representative of the spatial and temporal aspects of riverine fish
assemblages is an essential part of BCG and IBI development. The sampling conducted throughout
Maine during 2002-2007 provided the database for developing the narrative BCG  and  numeric riverine
IBI. The addition of data from rivers throughout New England in 2008-2009, coupled with the preceding
years of sampling Maine rivers (Figure B4-1), provided a more complete stress gradient for testing the
IBI and the BCG, which is illustrated by the specific assessment examples included  herein.

64.2.3 Task 3. Autecology of Extant New England Riverine Fish Fauna
Describing the autecology of the extant fish fauna is another essential step and includes information that is
used to derive, select, and test candidate IBI metrics and to make BCG attribute assignments. Information
about environmental tolerance, native status, habitat and flow preferences, thermal regime, foraging
habitats, and reproductive habits were compiled25 for 78 fish species known or suspected to occur in the
non-wadeable rivers of New England Proper.26 These classifications were compiled from  a number of
sources about native status, target fish classification, common riverine habitats where each species
occurred, spatial occurrence in the New England region, thermal classification, environmental tolerance,
foraging habits, reproductive habits, and predominant habitat residence. The most recent and
geographically relevant sources, in combination with observations made during nine  years of field studies,
were used to make these assignments. This task fulfilled the breadth and type of information that Karr et al.
(1986) described in the seminal guidance for developing fish IBIs. In addition to several new guilds that have
appeared in contemporary IBIs of the past 10-15 years, a fluvial classification scheme based on the target
fish community method of Bain and Meixler (2000, 2008) and a thermal classification scheme by Hokanson
(1977) were used to better reflect attributes of the BCG for New England riverine fish assemblages.
25 See http://www.midwestbiodiversitvinst.org/reports/31/Maine%20Rivers%20IBI%20Appendix%20Table%20B4-
6%2020160211.pdf. Accessed February 2016.
26 New England Proper includes the rivers that drain directly to the Atlantic Ocean. It excludes the portions of the
Lake Champlain, Hudson-Hoosic, St. Francois, and Lake Memphremagog drainages in Vermont and Massachusetts
(David Halliwell, personal communication).
                                                                                          B-85

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                      February 2016


B4.2.4 Task 4. Differentiating Riverine Ecotypes
It was essential to determine naturally occurring and distinctive strata for classifying New England
Rivers. The classification provides the basis for describing BCG levels 1 and 2 and predicting incremental
changes from these conditions with increasing stress. In addition to commonly applied distinctions such
as cold and warmwater assemblages, there are at least three riverine ecotypes in New England that are
distinguished by baseline habitat characteristics and fish assemblage composition. These ecotypes are:
moderate-high gradient riverine, low gradient riverine, and fresh-brackish water tidal habitats27 (Yoder
et al., 2008). Impounded habitats are viewed as a human-induced modification of moderate-high
gradient riverine habitats, rather than as a distinct and naturally occurring ecotype. However, data from
these modified habitats played an important role in testing the responsiveness of candidate metrics and
the riverine IBI to human-made modifications of natural riverine habitat. Based on observations made in
the field and in analyzing data for this and previous reports (Yoder et al. 2006a,b), there are distinctive
differences in fish assemblage composition between moderate-high gradient riverine habitats and low
gradient riverine habitats. The emphasis of this case study is on the moderate-high gradient riverine
ecotype.

64.2.5 Task 5. Expected Fish Assemblages along the Biological Condition Gradient
Developing an understanding of the natural fish assemblages that historically occurred in the non-
wadeable rivers of New England is critical to determining their current status and potential for
restoration. The BCG concept was employed for this task. Consistent with the BCG conceptual
framework, the "as naturally occurs" fish assemblage represents the assemblage expected in an
undisturbed/minimally disturbed condition for large, non-wadeable rivers in New England and
corresponds with BCG levels 1 and 2. Restoring all New England rivers to such a condition may be
impractical given the economically dependent activities and non-native species introductions that have
substantially altered the fish assemblages in this region for more than two centuries. Nevertheless it is
important to describe this condition because it serves as an essential anchor for the "upper levels" of
the BCG and as an objective reference for assessing current conditions, providing more accurate
understanding of what has been lost and, where possible, what can be restored. Description of the "as
naturally occurs" fish fauna was  based on historical observations by the first European settlers coupled
with expert knowledge about how such assemblages were most likely organized based on current
understanding of species autecology and distribution.28

In developing a BCG model for non-wadeable riverine fish assemblages, general information about the
historical fish assemblages coupled with expert judgment was used in the process. This was
accomplished through an expert advisory workgroup that included scientists from EPA, U.S.  Fish and
Wildlife Service, NOAA, State of Maine water quality and natural resource agencies, the Penobscot
Indian  Nation, and Trout Unlimited. One important outcome of the expert deliberations was the
conclusion that the "as naturally occurs" fish assemblage in the moderate-high gradient riverine ecotype
was largely comprised of native cool-coldwater species which are described as temperate stenotherms
and mesotherms.29 Based on these discussions and using the results of the initial sampling in Maine
27 See http://www.midwestbiodiversitvinst.org/reports/31/Maine%20Rivers%20IBI%20Appendix%20Table%20B4-
6%2020160211.pdf. Accessed February 2016.
28 See http://www.midwestbiodiversityinst.or;
6%2020160211.pdf. Accessed February 2016.
29 See http://www.midwestbiodiversityinst.or;
6%2020160211.pdf. Accessed February 2016.
28 See http://www.midwestbiodiversitvinst.org/reports/31/Maine%20Rivers%20IBI%20Appendix%20Table%20B4-
 °A
29 See http://www.midwestbiodiversityinst.org/reports/31/Maine%20Rivers%20IBI%20Appendix%20Table%20B4-
                                                                                           B-86

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
                                                                             February 2016
(Yoder et al. 2006a,b, 2008), the template for a BCG was developed for the cool-coldwater, moderate-
high gradient riverine ecotype (Figure B4-2). This reflects a comparatively simple, qualitative method of
visualizing what has happened in many instances to the "as naturally occurs" fish assemblage through
time. Observed departures from level 1 attributes and characteristics are the result of historical
modifications to water quality, habitat, flow regime, thermal regime, the native fauna via the
widespread introduction of non-native species, and the loss of connectivity for diadromous species
(Figure B4-2).
   Biological Condition Gradient Conceptual Model: Maine Rivers
       0.
        -
   il
   O  o>
   .2
   m
   O
       re

 O
 C
.2
±i
?
 o
O
       i
       to
              Native inland freshwater & diadromous species (Atlantic salmon, alewife, American
              shad, American eel, brook trout, native cyprinids, white & longnose sucker)

                     Same as tier 1 except:  non-native salmonid species with naturalized
                     populations may co-occur with brook trout.
                               Some native diadromous species are reduced in abundance; shifts
                               towards intermediate tolerances and mesotherms; brook trout are
                               reduced or replaced by non-native naturalized salmonid species.

                                           Some native diadromous species are rare or
                                           absent; moderately tolerant species predominate;
                                           brook trout are absent; non-native mesotherms &
                                           surytherms present; anomalies present.
Native diadromous species are absent or
if present by interventions; some native
cyprinids are absent, replaced by
tolerant and moderately tolerant species;
brook trout are absent; non-native
salmonids are non-reproducing;
non-native eurytherms usually
predominate; anomalies present.
            Native diadromous species rare or absent; tolerant
            species predominate and may become numerous
            (enrichment); species richness reduced in some cases
                                                      (toxic impacts); non-native
                                                      eurytherms predominate;
                                                      anomalies frequent.
           LOW
                       Human  Disturbance Gradient
                                                                    HIGH
Figure B4-2. A BCG model for fish assemblages representative of cool-coldwater, moderate-high gradient
riverine habitats in New England (after Yoder et al. 2008).

While the formal process of developing narrative and numeric decision rules through an expert driven
fuzzy set modeling approach was not used, expert judgment played a role in defining the
correspondence between BCG attributes and IBI metrics and informed deriving IBI thresholds for each
BCG level.  A more detailed description of the parallels and overlap between  BCG attributes and metrics,
and derivation of the riverine IBI and supplemental diadromous IBI (D-IBI) are described below.

B4.2.6 Task 6. Establishing Reference Condition Comparable to Biological Condition
Gradient Levels 1 and 2

The techniques for screening and selecting reference sites has evolved significantly during the past 30
years from a mostly qualitative process first described by Hughes et al. (1986) and used by some of the
                                                                                        B-87

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                     February 2016


pioneering developers of numeric biological criteria (e.g., Ohio EPA 1987; Barbour et al. 1996), to a more
quantitative process (Stoddard et al. 2006) that is now used by EPA and an increasing number of states.
How reference sites are selected and used to develop reference condition are essential components of
how EPA evaluates the level or rigor of biological assessment programs (Yoder and Barbour 2009; USEPA
2013). While the majority of these efforts have focused on wadeable streams, there are now ample
precedents for developing reference condition for large rivers (Hughes and Gammon 1987; Ohio EPA
1987; Lyons et al. 2001; Emery et al. 2003). The prevalence of legacy impacts in most non-wadeable
rivers raises issues about the quality of the reference condition that contemporary sampling data
represent. This is one reason why merging IBI development with the BCG framework is helpful. Ideally,
reference condition is  represented by undisturbed or minimally disturbed conditions, comparable to
BCG levels 1 and 2. However, actually finding a contemporary example of BCG level 1, and sometimes
level 2, has been elusive, especially  in large rivers. BCG level 2 conditions have been observed in other
ecological  regions, though not present everywhere. Given this reality it then becomes important to
understand how the relative states of "best" and "better" occur along the BCG within the domain of the
riverine fish assemblage data across New England, so that the task of reconciling conceptual goals with
societal realities can be dealt with more effectively. Articulating this framework now provides for a more
accurate way of developing attainable thresholds later in the process.

Reference sites were selected  using a combination of position in the landscape (with respect to point
and nonpoint source stressors) and  the intactness of the native fish fauna. The latter included selecting
sites that lacked blackbasses (smallmouth and largemouth bass) and other non-native species based on
knowledge about the negative impacts that these introductions have had on the native fish species that
comprise the sensitive metrics of the riverine IBI (Whittier et al. 2000, 2001; Warner 2005; Yoder et al.
2008). Yoder et al. (2015)  described the native status of the fish species that were either encountered in
the 2002-2009 sampling or reasonably expected to have occurred in recent times.30 The definitions of
Halliwell (2005) were followed in describing the native status of fish species and in deriving candidate
and final IBI metrics relative to native status. Hence, the presence of introduced species was a major
factor in the selection  and/or rejection of reference sites. New England rivers represent a unique
situation in which all of the major river drainages are mostly contained within New England state
boundaries, and all are coastal drainages discharging to Long Island Sound or the Gulf of Maine. As such,
they have  largely been isolated from adjacent drainages such as the St. Lawrence-Great Lakes and
Hudson River drainages since post-glacial times (Curry 2007). This has influenced the character of the
freshwater fish fauna with some species common to these adjacent drainages being historically absent.
Examples are smallmouth and largemouth bass that are not indigenous to any New England river
system, but which were introduced  in the latter part of the 19th century becoming firmly established in
several major river drainages to date (Warner 2005). A few select rivers in northern Maine have not yet
been invaded by blackbasses or other introduced species, and these also tended to represent minimally
impacted conditions in terms of landscape, habitat, thermal, and flow alterations. Hence, these were
selected as the approximation of minimally disturbed reference for the derivation and testing of
candidate  IBI metrics.  In addition to reflecting a minimum of anthropogenic chemical and physical
impacts, they also reflected the absence of non-native species. Reference and a gradient of non-
reference sites were selected to represent the full gradient of stress from BCG levels 1 and 2
(undisturbed/minimally disturbed conditions) to BCG level 6 (severely altered) as follows:

    •  "Minimally disturbed" reference sites lacking non-native species (BCG level 2).
30 See http://www.midwestbiodiversitvinst.org/reports/31/Maine%20Rivers%20IBI%20Appendix%20Table%20B4-
6%2020160211.pdf. Accessed February 2016.
                                                                                           B-88

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
February 2016
    •   Non-reference sites with conductivity > 100 u.S/cm.
    •   The remaining non-reference sites were partitioned by Qualitative Habitat Evaluation Index
       ranges: < 50; 51-75; 76-90; > 90; this imparts a habitat gradient that reflects commonly
       occurring impacts throughout New England.

The resulting stress gradient was then used to evaluate the response of the candidate IBI metrics
following the continuous calibration methodology of Mebane et al. (2003). This calibration method was
first used in the Pacific Northwest, which has many similarities to the New England region, including
depauperate cool-coldwater fish fauna impacted by similar stressors (i.e., thermal and flow alterations).
A range of scores for each metric were defined for all 6 BCG levels based on the correspondence
between the metric response and the narrative BCG level descriptions (Figure B4-2, Table B4-1). The
riverine IBI and D-IBI were then derived based on the traditional IBI development approach, summing
the scores for selected metrics into one index value (Yoder et al., 2008)).

Table B4-1. New England riverine IBI metrics with calibrated scoring equations and manual scoring
adjustment criteria  based on their initial application in Maine. Proportional (%) metrics are based on
numbers unless indicated otherwise (for methods used to derive the scoring equations,  see Yoder et
al. 2008).
Metric
Native species richness
Native Cyprinid species
(excluding fallfish)
Adult white & longnose sucker
abundance (biomass)
%NativeSalmonids
%Benthic Insectivores
%Blackbass
%Fluvial Specialist/Dependent
%Macrohabitat Generalists
Temperate stenothermic species
Non-guarding lithophilic species
Non-indigenous species
%DELT anomalies
Scoring Equation
10 * (-0.2462 + (0.0828*numspec2)))
(10 * (0.4457 + (0.0109*allcyp_ff) - (0.00005629 *
(allcypjf2))))
(10 * (0.3667 + (0.008*ws_lns_pb) - (0.000023592 *
(ws_lns_pb2))))
(10 * (0.9537 + (0.00000000039*nat_salm) -
(0.000078892 * (nat_salm2))))
10 * (0.010966*benth_pc_n)
10 - (10 * (-0.09684 + (0.5638*loglO(blackbass))))
(10 * (0.2775 + (0.0073*fluv_pc_n)))
10 - (10 * (0.1017 + (0.0096*macro_gen)))
(10 * (0.7154 + (0.4047*(loglO(steno)))))
(10 * (0.2979 + (0.8975*loglO(lith_ng))))
10 - (10 * (0.1063 + (0.3271*Non-indigenous_sp) -
(0.029*(Non-indigenous_sp2))))
10 - (10 * (0.8965 + (0.1074*loglO(delta))))
Scoring Adjustments
Score = 0
< 3 sp.
Eq"
0
0
0
Eq
0%
> 90%
Osp.
<1
>5
Eq
Score = 10
> 15 sp.
Eq
> 128 kg/km
> 20%
> 91.2%
0
Eq
Eq
>5 sp.
>10
0
0
64.2.7 Task 7. Deriving an Index ofBiotic Integrity for the Cool-Coldwater, Moderate-
High Gradient Riverine Ecotype
The riverine IBI metrics that were initially tested and selected had to take into account two of the
unique characteristics of New England fish fauna. First, there is an inherently low species richness, and,
second, the fauna has been isolated from adjacent regions since post-glacial times (Curry 2007). As such,
New England rivers naturally  lack several species that are common to the same latitudes in adjacent
  No scoring adjustments are necessary; scoring determined by equation (Eq) across entire metric scoring range of 0-10.
                                                                                          B-89

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                     February 2016
drainages. Given that physical factors (e.g., flow regime, habitat, and thermal regime) and biological
factors (e.g., native status) are important in this region, metrics that reflect those characteristics were
derived and selected.

This task included two principal steps: (1) the selection and testing of candidate IBI metrics, and (2) the
derivation of an IBI for the cool-coldwater, moderate-high gradient riverine ecotype. These tasks  have
ample and recent precedence in North America and elsewhere. A growing body of information is now
available for non-wadeable rivers (Yoder and Kulik 2003), including the baseline factors common to New
England (e.g., an appropriate thermal baseline for large rivers, metric testing and selection, and index
development and testing (Hughes and Gammon 1987; Ohio EPA 1987; Lyons et al. 2001; Mebane et al.
2003; Emery et al. 2003)). Developing the metrics for the core riverine IBI involved sequential steps
beginning with identifying candidate metrics, evaluating the responsiveness and relevance of those
metrics along the  BCG, and deriving indices comprised of the  "best"  set of metrics that represent the
ecotype and other strata that are embedded within the process (Yoder et al. 2008).

The riverine IBI metrics and their calibration equations appear in Table B4-1. The  riverine IBI  is scored on
a 0-100 scale making it amenable to scaling along the entirety of the BCG, rather than only a portion of
the BCG as is common to the early IBIs that employed the ordinal calibration and 12-60 scale of Karr et
al. (1986). Furthermore, the selection of riverine IBI metrics was based on emulating attributes of the
BCG (Figure B4-2) that was developed prior to IBI development and calibration.

A supplemental set of four diadromous metrics were developed  in 2011 to better reflect the
diadromous component of the fish assemblage in rivers that have historically supported these species
(Table B4-2). These supplement metrics  are applied only where diadromous fish  have historically been
documented, thus it is not applied where natural barriers have historically prohibited their occurrence.
It is added to the riverine IBI and the resulting index is termed the D-IBI.

Table B4-3 illustrates the match, or correspondence, between the metrics that comprise the riverine IBI
and D-IBI and the  BCG attributes. As discussed above, the first 12 metrics comprise the cool-coldwater,
moderate-high gradient riverine fish  IBI while the four supplemental diadromous metrics are specific to
the diadromous part of the fish assemblage. A complete match between a metric and an attribute
indicates that the species assigned to an IBI metric fits wholly within the definition of the BCG attribute.
Three IBI metrics were complete in their match with a BCG attribute—the percentage offish with
deformities, erosion, lesions, and tumors (DELT) anomalies (attribute VII), the proportion of benthic
insectivores (attribute VIII), and non-guarding lithophils (attribute VIM). A partial match indicates that an
IBI metric includes species that occur in multiple BCG attributes. For example, the fluvial
specialist/dependent IBI  metric includes species that occur in BCG attributes II and III, thus making the
correspondence of that metric for the two attributes partial. The supplemental diadromous metrics are
a surrogate measure of BCG attribute X (ecosystem connectance), with some of the species also
corresponding with attribute I (historically documented, sensitive, long-lived, or regionally endemic taxa.

B4.2.8 Task 8: Assessing New England Large Rivers
The BCG-based riverine IBI was used to assess the condition of large rivers across New England and also
to determine regional scale, reach level, and site-specific stressors (Yoder et al. 2015). Different data
sets were used for four different projects: a regional scale assessment, an intensive  assessment of the
Connecticut River mainstem, a comparison and ranking of major rivers using the riverine IBI and D-IBI.
and a site-specific application. This projects were conducted to explore the utility of using a BCG-based
index for large river biological assessments, with a focus on use of different data sets.
                                                                                           B-90

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                     February 2016


Regional Assessment of New England Large Rivers
The results of a regional scale analysis of condition by BCG level across New England is depicted in
Table B4-4. Two types of data sets were used: a probabilistic regional data set (REMAP32) and a targeted
sites data set. BCG level 2 sites were only identified in western and northern Maine (Figure B4-3; Table
B4-4). These sites also showed a low incidence of stressors, intact habitat, and the absence of non-
native species. The proportion of samples that reflected BCG levels 3 through 6 were not substantially
different between the REMAP probabilistic and targeted results (< 5% difference) (Table B4-3).

The only difference between the two sampling designs was illustrated by the absence of BCG level 2
samples in the REMAP probabilistic data set for New England (Table B4-5). A total of 19 targeted sites in
Table B4-5 had higher IBI scores than the highest scoring REMAP probability site and only 4 of the 27
highest scoring sites were REMAP probabilistic sites. Two of the 27 highest scoring REMAP sites occurred
outside of Maine in the upper Connecticut River in northern New Hampshire. These are the highest
quality sites and rivers in the New England region, and are potential candidates for additional
protections.

Connecticut River Assessment
The Connecticut River mainstem was sampled in 2008 and 2009 from the Third Connecticut Lake in New
Hampshire downstream to the "salt wedge" just upstream from 1-95 in Connecticut. Probabilistic sites
were selected from the 2008-2009 NRSA draw of sites for two levels of coverage with targeted sites
added to fill in "gaps" to complete a longitudinal pollution survey design on the mainstem. Based on the
BCG levels and the corresponding riverine IBI scores, an objective was to compare the estimates of
condition between the two probabilistic sample draws (NRSA base and REMAP) and the intensive
pollution survey design. Though the targeted design detected BCG level 2 conditions that the
probabilistic design did not, the overall proportion of BCG levels for both monitoring designs was
generally comparable (Table B4-4).
32 Regional EMAP, USEPA.
                                                                                          B-91

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
February 2016
Table B4-2. Supplemental diadromous metrics intended to represent the diadromous component of a
riverine fish assemblage in New England (for methods used to derive the scoring equations, see Yoder
et al. 2008)
Metric
Diadromous species richness
Number of American eel
Number of Clupeidae
Number of diadromous fish (all
diadromous species)
Scoring Equation
Score = 0.0318 + 0.227*(Diadromous Species Richness)
Score = 0.0689 + 0.2*(Log Eel Rel. No.) + 0.0616*(Log Eel
Rel. No.)
Score = 0.832*LoglO(Rel. No. Clupeids)A (0.269)
Score = 0.0522 + 0.168*(Log(Diad Rel. No.) +
0.0644A(Log(Diad Rel. No.))
Scoring Adjustments
Score = 0
0
0
0
0
Score = 10
>5 sp.
> 389/km
> 96/km
> 560/km
                                                                                        B-92

-------
                     Appendices to A Practitioner's Guide to the Biological Condition Gradient
February 2016
Table B4-3. Comparison of BCG attributes with riverine fish IBI metrics (metrics 1-12) and four supplemental diadromous metrics (metrics 13-16)
applicable to non-wadeable rivers in New England

BCG Attribute

1. Historically
Documented, Sensitive,
Long-lived, or
Regionally Endemic
Taxa
II. Highly Sensitive Taxa
III. Intermediate Sensitive
Taxa
IV. Intermediate Tolerant
Taxa
V. Tolerant Taxa
VI. Non-native or
Intentionally
Introduced Species
VII. Organism Condition
VIII. Ecosystem Function
IX. Spatial and Temporal
Extent of Detrimental
Effects
X. Ecosystem
Connectance
Riverine Fish IBI Metrics
5"
l/l Q.
Rf
a|
w



4














•



Fluvial Specialis
Dependent
--•





4
4














Temperate
Stenotherms






•















%Native
Salmonids



4


•













4

Adult
White/Longnos
Sucker Biomas
ui n>


4





4










4

Native Cyprinic
C/l






4

4

4










%Macrohabita
Generalists









4

•

4








%Benthic
Insectivores
















•





Non-indigenou
Species









4

4

•






•

sp
c?^
O3
QJ
n
7^-
ff
VI
VI
a>









4



•








Non-guarding
Lithophils
















•





DELT Anomalie















•






Diadromous Metrics
Diadromous
Species



4
















•

No. American E
s.


•
















•

No. Clupeidae




















•

No. Diadromou
Individuals
in

















•

•

   •- complete match between IBI metric and BCG attribute.
   4 - partial match between IBI metric and BCG attribute.
                                                                                                                                     B-93

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
                                       February 2016
Table B4-4. The number and percentage of New England Large River (NELR) REMAP probabilistic and
targeted samples arranged by corresponding BCG level for the riverine IBI

BCG Level
NELR REMAP Probabilistic
Samples
Percent
NELR REMAP Targeted
Samples
Percent
IBI
Level 1: IBI > 95
Level 2: IBI > 80 and IBK 95
Level 3: IBI > 60 and IBI < 80
Level 4: IBI > 40 and IBI < 60
Level 5: IBI > 20 and IBI < 40
Level 6: IBI < 20
Totals
0
0
15
48
78
8
	
149
0
0
10.1
32.2
52.3
5.4
100
0
12
42
127
177
13
371
0
3.2
11.3
34.2
47.7
"
100
^-^
                                                        •   I: >95
                                                        •   II: 80-94.9
                                                        •   III: 60-79.9
                                                        O   IV: 40-59.9
                                                        O   V: 20-39.9
                                                        •   VI: < 20
                                                            -Probabilistic
                                                          A -Targeted
Figure B4-3. Map of non-wadeable fish sampling sites in New England with riverine IBI values colored coded by
BCG levels 1-6.
                                                                                        B-94

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
February 2016
Table B4-5. Riverine IBI and supplemental diadromous metrics and their BCG level equivalents with
the method of estimation (e.g., regression equation, by eye). These were used in the mapping of the
IBI metrics (after Yoder et al. 2015).
IBI Metric
No. of Native Species
No. Temperate
Stenothermic Species
No. of Non-Guarding
Lithophilic Species
%. of Cyprinid Species*
% Native Salmonids
% Benthic Insectivores
% Black Bass
% Fluvial Specialists and
Dependents
% Macrohabitat
Generalists
Adult White, Longnose
Sucker Biomass
Non-Indigenous Species
% DELT Anomalies
Log American Eel
Number/km
Log Diadromous
Number/km
+LogClupeid Number/km
Diadromous Species
Richness
Biological Condition Level
BCG1
>10
>5
>7
>58.2
>4.20
>39.2
-
>96.8
<0.6
>63.4
0
0
>2.5
>2.5
>2.5
5
BCG 2
10
4
6
> 47.3-
58.2
>3.22-
4.20
>30-
39.2
0
> 86.3-
96.8
> 0.6-9.4
> 52.8-
63.5
1
> 0-0.30
> 2.0-2.5
> 2.0-2.5
> 2.0-2.5
4
BCG 3
8-9
3
4-5
> 32.8-
47.3
> 1.91-
3.22
> 17.7-
30.0
> 0-9.2
> 68.7-
86.3
>9.4-
24.2
> 38.7-
63.5
2
> 0.30-
0.50
> 1.5-2.0
> 1.5-2.0
> 1.5-2.0
3
BCG 4
6-7
1-2
3
> 18.3-
32.8
> 0.59-
1.91
>5.3-
17.7
>9.2-
19.3
> 43.9-
68.7
> 24.2-
45.0
> 24.6-
38.7
3
> 0.50-
1.0
> 1.0-1.5
> 1.0-1.5
> 1.0-1.5
2
BCG 5
4-5
0
1-2
>3.8-
18.3
0
<5.3
> 19.3-
29.4
>1.4-
43.9
> 45.0-
80.5
> 10.5-
24.6
4
> 1.0-2.0
> 0.5-1.0
> 0.5-1.0
> 0.5-1.0
1
BCG 6
<4
-
0
<3.8
-
-
>29.4
<1.4
>80.5
<10.5
>5
>2.0
<0.50
<0.50
<0.50
0
Equation/Method for BCG
Cutoff Estimation
Numspec=l. 83+0.018*161
Stenotherms=-l. 94+0.0713*161
Lithophil NG=-1.36+0.0866*IBI
% Cyprinids=-10.6+0.724*IBI
% Nat. Salm. =-2.03+0.0656*161
% Benth. lns.=-19.3+0.616*IBI
% Blackbass=39.5-0.505*IBI
% Fluvial Specialists=
-182+141*log(IBI)
%Macrohab. Gen. =234-
118*log(IBI)
White, LN Sucker=-
3.62+0.705*161
By eye
Threshold by eye
By eye
By eye
By eye
By eye
*excludesfallfish.
                                                                                            B-95

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                      February 2016


Correlations between Index ofBiotic Integrity Metrics and Their Biological Condition Gradient
Equivalents
Spatial patterns in riverine IBI metrics and their corresponding BCG level assignments across New
England were examined as part of the process for indexing the IBI to the BCG. Maps of each IBI metric
were color coded to its equivalent BCG level as depicted in Table B4-5. This allowed for the visualization
of general patterns across New England and also to highlight site- or river reach-specific issues that
might otherwise be obscured by regionally focused analyses and which could warrant more detailed
follow up investigations. Two representative metrics are included in Figure B4-4 and  each exposes
gradients related to the degree of disturbance from the major stressors identified by Yoder et al. (2015).
The non-native species metric, corresponding to BCG attribute VI (presence of non-native taxa), shows
the extent of introductions in central and southern New England and their virtual absence in northern
Maine (Figure B4-4), the latter of which is due primarily to both natural and artificial  barriers to their
ingress. Some non-native introductions such as smallmouth and largemouth bass have had deleterious
effects on native Cyprinids and other indigenous species in both lakes (Whittier et al. 2000, 2001) and
rivers (Yoder et al. 2008). At some locations  more than five non-native species were collected, and these
occurred in river reaches that are most impacted by hydrological alterations and chemical  pollution. As
such, the extent  of non-native species introductions represent one of the major negative influences on
the condition of the native New England riverine fish fauna.

The distribution and occurrence of temperate stenotherms (which are all native species) roughly mirrors
that  of non-native species (Figure B4-4). While this metric represents typically cool-coldwater fish
species, it also represents BCG attributes II and III (sensitive and moderately sensitive taxa) since this
metric is comprised of species that cannot tolerate significant alterations to the natural thermal regime,
habitat, and/or flow regime. This  same pattern generally  held for the other riverine IBI metrics that are
predicted to decrease with increasing stress (e.g., native species richness, % native Salmonids, and
fluvial specialist/dependent species). The pattern exhibited by non-native species generally held for the
metrics predicted to increase with increasing stress such as % blackbasses and macrohabitat generalists
(Yoder etal. 2015).

Comparing New England Mainstem River Reaches along the Biological Condition  Gradient
Sufficient data were available to make a comparative assessment of the status of 36  individual rivers
across New England which was one of the primary objectives of the project (Figure B4-5). The riverine IBI
reflects the status of the resident freshwater  assemblage. The supplemental diadromous metrics highlight
the ability of each individual river and/or river reach to  support a diadromous fish assemblage. As
discussed earlier, the diadromous metrics can be interpreted as an inferred measure of ecosystem
connectance (e.g., free from fish passage barriers) since diadromous fish rely upon free-access to fresh and
marine habitats to complete their natural life cycles. Box-and-whisker plots were used  to display and rank
each river by the 75th percentile value of the summed riverine IBI and diadromous metrics (i.e., the D-IBI)
and by shading the boxes with the corresponding BCG color level based on the median (50th  percentile) D-
IBI value. Ranking each river by the 75th percentile of D-IBI  values reflects the protection and/or restoration
potential of each river  or reach. The rankings were done according to the D-IBI which best incorporates all
of the BCG attributes and the current quality  of each river while revealing when the riverine  IBI and/or D-
IBI exhibit markedly different results. It also shows where the freshwater part of the  assemblage is
positioned along the BCG relative to the status of the diadromous part of the assemblage. The restoration
of the diadromous part of the assemblage will potentially benefit the freshwater fish assemblage in coastal
rivers—indirectly due to restoration of riverine habitat  if dams are removed and directly through the influx
of marine nutrients with diadromous fish (Saunders et al. 2006).
                                                                                            B-96

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
February 2016
                 Non-indigenous
                 Species
                 Present-day
                 Thermal Response
                                                     Number of
                                                      Species
                                                    .   I: 0
                                                    o   II: 1
                                                    •   111:2
                                                    O   IV: 3
                                                    O  V: 4
                                                    •  VI: >5
                                                    Temperate
                                                    Stenothermic
                                                      Species
                                                    • I: >5
                                                        : 4
                                                        : 3
                                                    O  IV: 1-2
                                                    0  V: 0
                                                    •  VI:  --
Figure B4-4. Number of non-indigenous (upper) and temperate Stenothermic species (lower) at New England
large river sites with symbols coded by the IBI metric value that corresponds to BCG levels 1-6.
                                                                             B-97

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
February 2016
                                                New England Large Rivers 2002-9
                   Allagash_Fish R.
                      Aroostook R.
                   W. Br. Penobscot
                Upper Connecticut R.
                   Penobscot Tribs.
                        St. John R.
           Lower Kennebec R. 2002-3
                   E. Br. Penobscot
                 Middle Kennebec R.
                        St. Croix R.
           Lower Kennebec R. 2008-9
                  Kennebec R. Tidal
             Upper Connecticut Tribs.
                  Penobscot R. Tidal
                      Penobscot R.
          Upper Middle Connecticut R.
              Upper Androscoggin R.
                     Upper Saco R.
             Lower Connecticut Tribs.
          Upper Kennebec & Branches
                 Pawcatuk_Wood R.
                       Pawtuxet R.
                     Lower Saco R.
                Lower Connecticut R.
                        Taunton R.
                 Lower Merrimack R.
                    Presumpscot R.
                          Sandy R.
                 Tidal Connecticut R.
                Sebasticook R. 2003
               Quinebaug_Shetucket
              Lower Androscoggin R.
                     Blackstone R.
          Lower Middle Connecticut R.
                 Upper Merrimack R.
                     Housatonic R.
                                 100
                                            Maine Rivers Index of Biotic Integrity (IBI)
Figure B4-5. New England river reach D-IBI and riverine IBI box-and-whisker plots for all sites sampled during
2002-2009 in 36 major riverine segments in New England and ordered by the 75th percentile of the D-IBI. Fill
                                            r.th
color corresponds to the BCG range using the 50  percentile D-IBI or riverine IBI value for each river and reach.
                                                                                                     B-98

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                      February 2016


Assessment of Reach-Level Impacts: Flow Modification
The Connecticut River downstream from the Turners Falls dam is affected by flow diversions to the
Cabot hydropower project. An objective of the 2008-2009 Connecticut River intensive survey was to
assess possible local scale effects from stressors such as habitat and flow modifications. River flows in an
approximate 3.5 mile long reach of the Connecticut River are effectively "by-passed" with most of the
flow being diverted into a canal that provides water to the Cabot hydroelectric generating station. A
minimum flow of 120 cfs is maintained over the Turners Falls dam into the by-pass reach during low
flow periods. The result is a very constricted wetted channel with the much wider physical channel
lacking flows that are comparable to a typical New England moderate-high gradient river of this size. As
a result, the habitat consisted almost entirely of pools with little or no flow velocity in the upper reach
that is represented by the upstream most site (RM 67.9). Four sites were located within and
immediately downstream from the 3.5 mile bypassed reach to assess potential effects of the diversion
of flows (Figure B4-6). The riverine  IBI, D-IBI, and selected IBI metric results for the four sites are shown
in Figure B4-6. The riverine IBI at upstream-most site (RM 67.9) in the bypass reach revealed BCG level 6
(very poor) quality, and the second site (RM 66.9) was BCG level 4 (fair) for the IBI. This site is
downstream from the partial return of flows from the Cabot station feed channel, which was positive for
the fish assemblage. The D-IBI was  one BCG level higher at three of the four sites, indicating a higher
abundance of diadromous species in three samples. The IBI metric results generally reflected BCG levels
5 and 6 with sporadic exceptions. The results for %blackbasses and macrohabitat generalists were
consistent with the high degree of flow alteration and the resulting negative influences of the flow
diversion on habitat  quality in the bypass reach. Simply increasing the minimum flows over the Turners
Falls dam would result in improved IBI scores and an increase in the BCG level. The BCG framework
provides a means to communicate  information to stakeholders to better understand the gains, or losses,
in management decisions.

Assessment of Reach-Level Impacts: Dam Removal
In an effort to document how the fish IBIs responded to the improved habitat and access to diadromous
fish, the lower Sebasticook River has been sampled annually since 2009 as a follow-up to the removal of
the Ft.  Halifax dam at the mouth in Winslow, Maine. A baseline assessment of what was then an
impounded riverine habitat was conducted at three sites upstream from the Ft. Halifax dam in 2003. The
dam was removed in 2008 as part of a FERC relicensing agreement to improve access for river herring to
their historic spawning areas in the Sebasticook River drainage. The Ft. Halifax dam removal was coupled
with improved fish passage at two upstream dams. The results of sampling after the dam removal in
2009, 2010, and 2011 show increases in both the riverine IBI and D-IBI, but particularly so with the latter
(Figure B4-7). The modest improvement in the riverine  IBI reflected  improved riverine habitat for
resident freshwater species, but the capacity for additional improvement is limited by historic
alterations in the flow and thermal regimes and the introduction of  non-native species such as
smallmouth and largemouth bass. Other introduced species such as northern  pike and several sunfish
species have occurred post-dam removal. The D-IBI showed a comparatively larger increase due to
improved access by diadromous species and river herring in particular. These results show not only
improved access, but the success of these species finding and reproducing in their historic spawning
areas as the sampling measures the outmigration of the young-of-year of these species in the late
summer and fall of each year. By including the D-IBI, the results are  a better representation of the BCG
level corresponding to the strong improvement in the diadromous species that comprise attribute X.
                                                                                          B-99

-------
                 Appendices to A Practitioner's Guide to the Biological Condition Gradient
                                                                                           February 2016


  DIBI =
  %Macrohabitat Generalists = 62%
  %Blackbasses = 44%"
  %DELT Anomalies^
                                                                   ,:JCL: ?.. ?.\J\
                                                          Conne.
                                                          IBI = 5
                                                          DIBI = 74
                                                          %MacrohabitafGenera lists = 17%
                                                          %Blackbass
  IBI = 38 (BCG Level 5)
  DIBI= 57
                                                                                                 'pass Reach:
                                                          DIBI = 34
                                                          %Macrohabitat Generalise
%DELT Anomalies =
                                                                        maUes
Figure B4-6. Fish sampling results in and downstream from the Turners Falls bypass reach in the Connecticut River and in the vicinity of the Cabot
hydropower project in 2009 showing riverine IBI and D-IBI scores and selected metric results. Color shading in the cells corresponds to the BCG level for the
IBI, D-IBI, and each IBI metric (see Table B4-5). (Green BCG level 3, Yellow BCG level 4, Orange BCG Level S, Red BCG level 6).
                                                                                                                    B-100

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient
                                                                   February 2016
                                       Sebasticook River
     eg
     o
     +
     e
              100
               80
60
               40
               20
           Ft. Halifax Dam
           Removed 2008
                          O
                                                  -C
                                                         O
          IV
                                                                                  V
                                                                                  VI
               DO
               O
               Q
                                                                                       (D
                                                                                       (D
co
o
o
CN
or
o
o
o
"w
.a

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                     February 2016


would have been expected based on minimal levels of disturbance. Thus far, the resulting IBIs have
utility in detecting the biological impacts from alterations to the flow regime, the thermal regime, and
habitat, each of which is a key focus of contemporary restoration efforts in numerous New England
rivers.

Appendix B4 References
Bain, M.B., and M.S. Meixler. 2008. A target fish community to guide river restoration. River Research
     and Applications 24:453-458.

Bain, M. B., and M. S. Meixler. 2000. Defining a Target Fish Community for Planning and Evaluating
     Enhancement of the Quinebaug River in Massachusetts and Connecticut. Final report by the New
     York Cooperative  Fish and Wildlife Research Unit, Cornell University, Ithaca, NY to the New
     England Interstate Water Pollution Control Commission, Lowell, MA. 51 pp.

Barbour, M.T., J. Gerritsen, G.E. Griffith, R. Frydenborg, E. McCarron, J.S. White, and M.L Bastian. 1996.
     A framework for biological criteria for Florida streams using benthic macroinvertebrates. Journal
     of the North American Benthological Society 15(2):185-211.

Curry, R.A. 2007. Late glacial impacts on dispersal and colonization of Atlantic Canada and Maine by
     freshwater fishes. Quaternary Research 67(2):225-233.

Danielson, T. 2006. Standard Operating Procedure Methods for Sampling Stream and Wetland Algae.
     Maine Department of Environmental Protection, Bureau of Land and Water quality, Division of
     Environmental Assessment, Augusta, ME. DEPLW0634. 18 pp.

Davies, S.P., and S.K. Jackson. 2006. The biological  condition gradient: A descriptive model for
     interpreting change in aquatic ecosystems. Ecological Applications 16(4):1251-1266 + appendices.

Davies, S.P., L. Tsomides, J.L. DiFranco, and D.L. Courtemanch. 1999. Biomonitoring Retrospective:
     Fifteen Year Summary for Maine Rivers and Streams. Maine Department of Environmental
     Protection, Bureau of Land and Water quality, Division of Environmental Assessment, Augusta,
     ME. DEPLW1999-26. 192 pp.

Davies, S.P. and L.  Tsomides. 1997. Methods for Biological Sampling and Analysis of Maine's Inland
     Waters. Maine Department of Environmental Protection, Bureau of Land and Water quality,
     Division of Environmental Assessment, Augusta, ME. 29 pp.

Emery, E.B., T.P. Simon, F.H. McCormick, P.A. Angermeier, J.E. DeShon, CO. Yoder, R.E. Sanders, W.D.
     Pearson, G.D. Hickman, R.J. Reash, and J.A. Thomas. 2003. Development of a Multimetric Index for
     Assessing the Biological Condition of the Ohio River. Transactions of the American Fisheries Society
     132:791-808.

Goldstein, R.M., and T.P. Simon. 1999. Toward a unified definition of guild structure for feeding ecology
     of North American freshwater fishes, pp. 97-122. in T.P. Simon (ed.), Assessing the Sustainability
     and Biological Integrity of Water Resources Using Fish Communities. CRC Press, Boca Raton, FL.

Halliwell, D.B. 2005. Introduced Fish in Maine. Focus on Freshwater Biodiversity, MABP Series. 12 pp.
                                                                                         B-102

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                     February 2016


Halliwell, D.B, R.W. Langdon, R.A. Daniels, J.P. Kurtenbach, and R.A. Jacobson. 1999. Classification of
      freshwater fish species of the northeastern United States for use in the development of indices of
      biological integrity, with regional applications, pp. 301-338. in T.P. Simon (ed.), Assessing the
      Sustainability and Biological Integrity of Water Resources Using Fish Communities. CRC Press, Boca
      Raton, FL

Hartel, K.E., D.B. Halliwell, and A.E. Launer. 2002. Inland Fishes of Massachusetts. Massachusetts
      Audubon Press, Faunal Series No. 3, Lincoln, MA.

Hokanson, K.E.F. 1977. Temperature requirements of some percids and adaptations to the seasonal
      temperature cycle. Journal of the Fisheries Research Board of Canada 34(10):1524-1550.

Hughes, R.M., and J.R. Gammon. 1987. Longitudinal  changes in fish assemblages and water quality in the
      Willamette River, Oregon. Transactions of the American Fisheries Society 116:196-209.

Hughes, R.M., D.P. Larsen, and J.M. Omernik. 1986. Regional reference sites: A method for assessing
      stream pollution. Environmental Management 10:629-635.

Hughes, R.M., P.R. Kaufmann, AT. Herlihy, T.M. Kincaid, L Reynolds, and D.P. Larsen. 1998. A process
      for developing and evaluating indices of fish assemblage integrity. Canadian Journal of Fisheries
      and Aquatic Science 55:1618-1631.

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. Illinois Natural History Survey Special Publication
      5,28pp.

Langdon, R.B. 2001. A preliminary Index of Biotic Integrity for fish assemblages of small coldwater
      streams in Vermont. Northeastern Naturalist 8(2):219-232.

Lyons, J., R.R. Piette, and K.W. Niermeyer. 2001.  Development, validation, and application of a fish-
      based index of biotic integrity for Wisconsin's  large  warmwater rivers. Transactions of the
      American Fisheries Society 130(6):1077-1094.

Mebane, C.A., T.R. Maret, and R.M. Hughes. 2003. Development and testing of an index of biotic
      integrity (IBI) for Columbia River basin and western  Oregon. Transactions of the American
      Fisheries Society 132:239-261.

Nelson, J.S., E.J. Grossman, H. Espinosa-Perez, L.T. Findley, C.R. Gilbert, R.N. Lea, and J.D. Williams. 2004.
      Common and Scientific Names of Fishes from the United States, Canada, and Mexico. Sixth edition,
      American Fisheries Society Special Publication 29. 386 pp.

Ohio Environmental Protection Agency. 1987. Biological Criteria for the Protection  of Aquatic Life:
      Volume II. Users Manual for Biological Field Assessment of Ohio Surface Waters. Division of Water
      Quality Monitoring and Assessment, Surface Water Section, Columbus, Ohio.

Page, L.M., H. Espinoza-Perez, L.T. Findley, C.R. Gilbert, R.N. Lea, N.E. Mandarak, R.L. Mayden, and J.S.
      Nelson. 2013. Common and Scientific Names of Fishes from the United States, Canada,  and
      Mexico, 7th edition. American Fisheries Society, Special Publication 34, Bethesda, Maryland.
                                                                                          B-103

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                     February 2016
Saunders, R., M.A. Hachey, and C.W. Fay. 2006. Maine's diadromousfish community: Past, present, and
      implications for Atlantic salmon recovery. Fisheries 31(ll):537-547.

Stoddard, J.L, D.P. Larsen, C.P. Hawkins, R.K. Johnson, and R.H. Norris. 2006. Setting expectations for
      the ecological condition of streams: The concept of reference condition. Ecological Applications
      16(4):1267-1276

USEPA. 2013. Biological Assessment Program Review: Assessing Level of Technical Rigor to Support
      Water Quality Management. EPA 820-R-13-001. Office of Water, Washington, DC. 144 pp.

Warner, K. 2005. Smallmouth bass introductions in Maine: history and management implications.
      Fisheries 3Q(ll):2Q-26.

Whittier, T.R., D.B. Halliwell, and R.A. Daniels. 2001. Distribution of lake fishes in the northeast-IV.
      Benthic and small water column species.  Northeastern Naturalist 8(4):455-482.

Whittier, T.R., D.B. Halliwell, and R.A. Daniels. 2000. Distribution of lake fishes in the northeast-ll. The
      Cyprinidae (minnows). Northeastern Naturalist 7(2):131-156.

Yoder, C.O., and M.T. Barbour. 2009. Critical technical elements of state bioassessment programs: A
      process to evaluate program rigor and comparability. Environmental Monitoring and Assessment.
      DOI10.1007/S10661-008-0671-1 (accepted for publication).

Yoder, C.O. and B.H. Kulik. 2003. The development and application of multimetric biological assessment
      tools for the assessment of impacts to aquatic assemblages  in large, non-wadeable rivers: A
      review of current science and applications. Canadian Journal of Water Resources 28(2):301-328.

Yoder, C.O., B.H. Kulik, and J.M. Audet. 2006a. The Spatial and Relative Abundance Characteristics of the
      Fish Assemblages in Three Maine Rivers. MBI Technical Report MBI/12-05-1. Grant X-98128601
      report to U.S.  Environmental Protection Agency, Region 1, Boston, MA. 136 pp. + appendices.

Yoder, C.O., B.H. Kulik, B.J. Apell, and J.M. Audet. 2006b. 2005 Maine Rivers  Fish Assemblage
      Assessment: I. Northern Maine Rivers Results; II. Maine Rivers Fish Species Distribution Atlas; III.
      Toward the Development of a Fish Assemblage Index for Maine Rivers. MBI Technical Report 12-
      06-1. Report to U.S. Environmental Protection Agency, Region 1, Boston, MA. 71 pp. + appendices.

Yoder, C.O., R.F. Thoma, LE. Hersha, E.T. Rankin, B.H. Kulik, and B.R. Apell. 2008. Maine Rivers Fish
      Assemblage Assessment: Development of an Index ofBiotic Integrity for Maine Rivers. MBI
      Technical Report 2008-11-2. Report to U.S. EPA, Region 1, Boston, MA. 69 pp.
      http://www.midwestbiodiversityinst.org/publications. Accessed February 2016.

Yoder, C.O., E.T. Rankin, and I.E. Hersha. 2015.  Development of Methods and Designs for the Assessment
      of the Fish Assemblages ofNon-Wadeable Rivers in New England. MBI Technical Report MBI/2015-
      3-3. U.S. EPA Assistance Agreement RM-83379101. U.S. EPA, Office of Research and Development,
      Atlantic Ecology Division, Narragansett, Rl and U.S. EPA, Region 1, Boston, MA. 152 pp.
      http://www.midwestbiodiversityinst.org/publications. Accessed February 2016.
                                                                                         B-104

-------
Appendices to A Practitioner's Guide to the Biological Condition Gradient                      February 2016


Whittier, T.R., and R.M. Hughes. 1998. Evaluation offish species tolerances to environmental stressors
      in lakes in the northeastern United States. North American Journal of Fisheries Management
      18(2):236-252.
                                                                                            B-105

-------