EPA/600/A-97/089
Whittier 1
Development of IBI metrics for lakes in southern New England
Thomas R. Whittier
Dynamac International, Inc.
200 SW 35th St.
Corvallis, OR 97333
whittier@heart.cor.epa.gov
541-754-4455
I. INTRODUCTION
2. IMPEDIMENTS TO DEVELOPING LAKE IBIS
2.1 SCOPE OF ASSESSMENT
2.2 SAMPLING ISSUES
2.3 PUBLIC PERCEPTIONS AND MANAGEMENT OF LAKES
2.4 ECOLOGICAL FACTORS
3. METHODS
3.1 SAMPLE DESIGN & FIELD METHODS
3.2 QUANTITATIVE METHODS
4. RESULTS
4.1 SPECIES RICHNESS METRICS
4.2 TROPHIC COMPOSITION METRICS
5. DISCUSSION
Submitted as a chapter in Assessing the sustainability and biological integrity of water resource quality using
fish assemblage. Thomas P. Simon (ed). CRC Press. Boca Raton, FL.
The research described in this article has been fimded by the U.S. Environmental Protection Agency. This
document has been prepared at the EPA National Health and Environmental Effects Research Laboratory,
Western Ecology Division, in Corvallis, Oregon, through Contract 68-C5-0005 to Dynamac International
Corp. It has been subjected to the Agency's peer and administrative review and approved for publication.
Mention of trade names or commercial products does not constitute endorsement for use.
-------
Whittier 2
INTRODUCTION
Among the principles implicit in the field of ecology are assumptions that: (1) living organisms react to
changes in their ecosystems, (2) we should be able to detect and quantify changes in biological assemblages
due to adverse stress caused by human activity, and (3) these changes should be predictable and generalizable
to broad classes of ecosystems and assemblages. In the early 1980s Karr and his colleagues successfully
applied these principles to fish assemblages in warmwater streams in the Midwest (Karr 1981,1991; Karr et
aJ. 1986; Lyons 1992; OEPA 1988). The resulting Index of Biotic Integrity (DBI) has been repeatedly
modified for other assemblages, regions and ecosystems (e.g., Leonard and Orth 1986; Miller et aJ 1988;
Steedman 1988), with varying degrees of success.
To date there has been only limited effort toward developing IBIs for lakes. Minns et al. (1994) developed an
eight metric IBI for littoral areas in the Great Lakes. Dionne and Karr (1992) and Jennings et al. (1995)
presented preliminary results for a Reservoir Fish Assmeblage Index for the Tennessee Valley, but concluded
that additional work was needed. The only other published research for inland lakes is in Wisconsin by
Jennings et al. (this book). They evaluated metric variability related to sampling methods, and how the
metrics performed as indicators of human induced stress. They were satisfied with the performance of only
four metrics.
In this chapter, I review some of issues that may account for the apparent difficulty in developing inland lake
IBIs. I then present an evaluation of several candidate metrics for lakes in southern New England, and discuss
some the implications of the results and needs for further research.
-------
Whittier 3
IMPEDIMENTS TO DEVELOPING LAKE IBIS
The reasons that development of lake IBIs lags behind that for warrnwater streams may be divided into four
groups: (1) scope of assessment, (2) sampling issues, (3) perceived values of lakes and the management
practices that support those values and (4) ecological factors. The issues discussed below are generalizations
meant to emphasize differences, exceptions to each point can be found. Clearly, these concerns are not the
exclusive domain of lake assessments.
SCOPE OF ASSESSMENT
For streams the conceptual and actual units for biological assessments are not entire streams, but rather
separated points or reaches along the stream length. In reporting results, the points or reaches are plotted or
tabled as individual samples and the assessment of biotic condition usually integrates the individual samples
and changes along the stream length. For lakes, the conceptual units are usually whole lakes, although
sampling is usually at stations or portions of shoreline; and the data are generally combined to represent the
entire lake. Only when lakes are sufficiently large are sampling, analysis and assessment done on subunits
(e.g., coves, bays, arms).
Another facet of this issue is one of "apples and oranges". Biological assessments need to be developed for
classes of reasonably comparable ecosystems (Plafkin et al. 1989). An IBI is based on expeeted-
characteristics of a particular assemblage type, in a particular size and type of waterbody, in a particular
ecoregion or basin. For streams, assemblage expectations are usually defined for ecoregion (or basin),
temperature type (cold or warm), and stream size. The classes of ecoregions and temperature types usually
can encompass a very large number of streams in the same assessment framework.
-------
Whitlier 4
For lakes, classing waterbodies into reasonably comparable groups is more complex. Stream size describes
habitat volume. Habitat volume in lakes is more multi-dimensional. Lake volume defines only the maximum
possible habitat volume. The useable volume may be limited by oxygen depletion (either natural or amplified
by anthropogenic eutrophication) and temperature extremes in summer or oxygen depletion in winter and thus
may vary greatly within and among years. Natural differences in stratification regimes affect expectations for
the fish assemblages, as does basin shape which determines the proportion of the habitat that is littoral versus
pelagic. Watershed position and relative connectedness are also important; a large headwater lake or a large
isolated lake could be expected to have a fish assemblage which differs from a large well connected lake
closer to a mainstem river.
SAMPLING ISSUES
Biological assessments require data collected in a consistent manner, that represent or index the resident
assemblages (Plafldn et al. 1989). For wadeable streams, electrofishing provides appropriate data for IB1
assessments (e.g., OEPA 1988). For lakes no single gear is sufficient to sample in all habitat types nor for all
species. Numerous studies have examined the issue of gear selectivity, sufficiency of effort and sampling
variability in northern lakes (e.g., Weaver et al. 1993). Whenever more than one gear or method is used,
questions arise of how and whether to combine data, and what constitutes a unit of sampling effort Jennings
et al. (this book) propose evaluating and using data from different methods for different metrics, as well as
limiting the data collection to littoral assemblages.
Problems with using multiple gear are compounded further when certain effective methods cannot be applied
in all lakes. Night beach seining is often the only effective method for collecting cyprinids, darters and
sculpins, but many lakes lack beaches clear of obstacles, suitable for seining. Electrofishing is a good single
method for sampling the littoral zone, but losses its effectiveness in low conductivity lakes common in
-------
Whittier 5
northern regions, and in lakes witli very narrow littoral zones. Likewise, gill nets are the only reliable method
for collecting salmonids and other pelagic species, but resource managers of low productivity northern lakes
often restrict or forbid the use of gill'nets.
PUBLIC PERCEPTIONS AND MANAGEMENT OF LAKES
The public, and therefore management agencies, value different things about lakes than they value for
streams. The public often values a park-like aesthetic for lakes, as a place for homes, vacation cabins, and
parks for camping, picnics, and swimming. People like the water; they like to see it, be near it, on it, or in it,
and often they want to use the water for drinking. Except for drinking water, lakes are more amenable to
these things than are streams. Activities in, on, or near the water lead people to want water clear of obstacles
(vegetation, snags, etc), for boating, fishing and swimming. And of course, people want good fishing,
regardless of whether the desired species is native or even suited for "their" lake.
Lakes are often managed to enhance the values listed above, usually to the detriment of biotic integrity.
Snags and weeds are cleared for boating, swimming, docks, and front yard aesthetics. Retaining walls or
riprap are added for erosion control (Jennings et al. 1996; Christensen et al. 1996). Lakes have been
subjected to slocking and other manipulations of fish assemblages for well over a century. In southern New
England much of the public is not aware that most of the currently widespread game fish are not native (e.g.,
largemouth and smallmouth basses, bluegill, northern pike, black crappie, brown, rainbow and lake trouts).
Management goals are closely linked with who does the managing. Streams tend to be managed by, or under
the authority of state departments of environmental quality, who are concerned with issues defined by the
Clean Water Act and subject to oversight by the U.S. EPA. This is in part due to historic problems from
point-source pollution on streams and rivers. As the obvious water quality problems in streams are corrected,
-------
Whittier 6
many of these agencies are shifting their regulatory emphasis to the biological condition of streams, for which
the IBI is well suited.
Lakes, on the other hand, are managed by various legal entities. In some lake-rich regions where lakes receive
more recreational use than streams, lakes in the public domain are often managed by state fish and game
departments. Their mandates emphasize management for fishability, not biotic integrity. Unfortunately in
some instances, interagency politics has led to lakes and streams becoming the exclusive territory of different
agencies wherein environmental quality agencies often only become involved with lakes over eutrophication
and fish tissue contaminants issues. Finally, unlike streams and rivers, lakes often have what amounts to
owners: private individuals and families, lake and home owners associations, sports clubs, organization
camps, resorts, and water districts. Technically many of these lakes are public, but in reality the owners of
the land surrounding lakes limit access to and activities on "their" lake, and the public agencies often choose
not to press their legal mandate. It is reasonable to assume thai most lake owners do not manage for biotic
integrity.
ECOLOGICAL FACTORS
In general, lakes tend to be new geographic features compared with rivers and streams, and fishes have had
longer to speciate in, and adapt to flowing water. Thus, there tends to be fewer lake-dwelling species. For
example, in southern New England the regional species pool of lake dwellers is considerably smaller than the
already depauperate species pool for lotic systems (Miller et al. 1988; Halliwell et al. This book). There are
only three native sunfish; pumpkinseed is nearly ubiquitous, and the other two are rather uncommon.
Depending on the location, there are one or two each of suckers, darters, and native top carnivores, and three
to five possible lake-dwelling minnows (Whittier et al. I997a). In addition, for more than a century the fish
fauna in New England has been liberally augmented with non-natives. In the Northeast an estimated 64% of
-------
Whittier 7
all lakes have at least one non-native species, with non-native individuals outnumbering the natives in 25% of
lakes (T, Whittier unpub. data). Introductions are especially common in southern New England where nearly
all lakes have non-natives.
Fish species' tolerances to stress (whether natural or anthropogenic) in lake ecosystems often differ from their
tolerances in stream ecosystems. Whittier and Hughes (in review) evaluated 45 species' tolerances to five
anthropogenic stressors in Northeast lakes. They found that eight species usually classified as tolerant or
moderately tolerant of disturbance in streams appear to be intolerant or moderately intolerant of degraded
conditions in lakes. Five species usually classified as intermediately tolerant in streams were very tolerant in
Northeast lakes.
METHODS
SAMPLE DESIGN & FIELD METHODS
The data used here were from the U.S. Environmental Protection Agency's Environmental Monitoring and
Assessment Program (EMAP) Northeast Lakes Pilot (Larsen et al. 1991,1994; Stevens 1994). EMAP
sampled fish assemblages, as well as water chemistry, zooplankton, physical habitat and riparian birds in 179
lakes and reservoirs during the summers of 1992-94 in the Northeast USA (New England, New York, New
Jersey). The lakes were selected, using a systematic random design, from all lakes > 1 ha. For this study, I
used the 50 lakes in southern New England (MA, CT, RI and the southern 1/3 of NH) and five additional
lakes, purposively selected and sampled in 1991 for methods evaluation (Larsen and Christie 1993).
Fish assemblages were sampled with overnight sets of gillnets, trapnets and minnow traps, and by night
seining (Whittier et al. 1997b). The level of effort was determined by lake size. The sampling objective was
-------
Whittier 8
to collect a representative sample of the fish assemblage at each lake. Sites were selected using a stratified
random design. Littoral fish sampling was done at random stations within each macrohabitat class. Pelagic
sample sites were chosen in random directions from the deepest location. The collected fish were identified to
species and counted. As part of the EMAP Quality Assurance procedures (Chaloud and Peck 1994)
specimens of all species were vouchered with the Museum of Comparative Zoology at Harvard University to
confirm identifications and for permanent archival. To develop the metrics evaluated in this study, I
combined data from all gear. Water samples were collected at 1.5 m at the deepest part of the lake. Field
methods are detailed in Baker et al. (1997), which is available from the author. Field and laboratory data
from the Northeast Lakes Pilot may be found on the EMAP's website (http://www.epa.gov/docs/emfjulte/
htmJ/datal/surfwatr/index.html).
QUANTITATIVE METHODS
I chose two measures of anthropogenic stress to evaluate metric performance: total phosphorus (TP) as a
measure of eutrophication stress, and the extent of human activity in the watershed as a measure of
generalized human induced stress. The latter is an indirect measure of stress, but is based on the premise that
increased human activity in the watershed increases the frequency and strength of a multitude of
anthropogenic perturbations to the lake ecosystem. Although acidification is also a stressor in some areas of
the Northeast, only four of the sampled lakes in southern New England had pH<6.0. Preliminary examination
of the fish assemblage data did not reveal any distinct qualitative acidification effects, thus I used pH stress
only to aid interpretation of metric behavior.
Watershed-scale measures of human disturbance were developed from digitized coverages of human
population and road density (Census Bureau 1991; USEPA 1992), and land use/land cover (USGS 1990)
from which proportions of the watershed in urban, industrial/commercial, residential, forested, wetlands and
-------
Whittier 9
agricultural categories were calculated (C. Burch Johnson, unpub. data). To estimate human disturbance in
the watershed, I used the first axis scores of a principal components analysts (PCA) (Gauch 1982, PROC
PRINCOMP; S AS 1985) of the land cover (% forest, % urban, % agricultural), human population density
and road density, as described in Whittier et al. (1997a). The first principal component accounted for 62% of
the variability in these five variables, with % forests loading negatively, and all human activity variables
loading positively (Table 1).
To determine species' native ranges I examined species maps and descriptions in a variety of fisheries texts
(Kendall 1914; Hubbs and Cooper 1936; Hubbs and Lagler 1964; Scott and Grossman 1973; Lee et al. 1980;
Trautman 1981; Becker 1983; Smith 1985; Schmidt 1986; Underhill 1986; Page and Burr 1991), and state
biological survey reports from New York (NYSDC 1927-39) and New Hampshire (NHFGC 1937-39).
Trophic guild and habitat preferences were based on these sources along with the summary tables Ln Halliweil
et al (this book), Karr et al. (1986), Ohio EPA (1989), Lyons (1992), and Minns et al. (1994). Most
tolerance classifications were from Whittier and Hughes (in review); for the remaining uncommon species I
used information contained in all of the above sources to make tolerance classifications.
Due to the relatively low native species richness in New England, I anticipated that several commonly used
metrics would not be effective. Thus, I examined a large number of potential metrics (e.g., Miller etal. 1988;
Simon and Lyons 1995). To evaluate candidate metrics I considered the following: How many species
contributed to the metric? What was the statistical distribution of raw scores among lakes? Was there a lake
size effect? How do raw metric scores relate to the two measures of stress? In particular, do the raw metric
values distinguish between the most degraded and the least degraded lakes? What ecological characteristics
do the metrics represent and how do we expect these to change over a range of natural conditions and human-
-------
Whitlier 10
induced stress? I primarily used a graphical-based approach (Fore et al. 1996), examining scatterplots of
raw metric data and residuals from regressions .
RESULTS
In the summers of 1991-94, EMAP sampled 55 lakes and reservoirs in southern New England (SNE; Figure
1). All, except three, lakes were <300 ha (Figure 2). Eighteen of the lakes are in the Northeastern Highlands
ecoregion (NHE), with the remainder in the Northeastern Coastal Zone (NCZ) ecoregion (Omemik 1987).
Ten lakes were eutrophic or hypereutrophic, with the remainder split between mesotrophic and oligotrophic
(Figure 2a-b). Phosphorus values tended to be higher in NHE and in lakes <100ha. There was no association
with lake size for watershed disturbance (Figure 2c-d). Watersheds were less disturbed in NHE. Water level
control structures are common on Northeast lakes (pers. obs.); 10 of SNE sample lakes have "Reservoir" in
their names. Named SNE reservoirs tend to be the larger lakes and have relatively low stressor values.
Forty fish species were collected, 18 of which are native to the region, 20 are introduced and two are native to
some of the lakes (Table 2). Of the 40 species collected, 18 were found in three or fewer lakes. Assemblages
were characterized by centrarchids, yellow perch, chain pickerel, golden shiner and bullheads. Bluegill tended
to dominate in the south, being replaced by yellow perch to the north.
SPECIES RICHNESS METRICS
Nearly all of the original (Karr 1981) and variants (Miller et al. 1988; Simon & Lyon 1995) of the species
richness metrics were problematic in SNE. The total number of species EMAP collected was similar to the
total (44) used by Minns et al. (1994) in the Great Lakes. But slightly more than half of the SNE species are
not native to New England, compared to 25% non-native species collected by Minns et al. Many of the New
-------
Whittier 11
England non-natives were introduced 100+ years ago and are firmJy established (NHFGC 1937-39). Some
authors suggest that these species are now "naturalized" and should be considered as part of the resident
species pool (Halliwell et al, this book). Under that proposal only the two non-North American species
(common carp and brown trout) and populations maintained by stocking (nearly always salmonids in SNE)
could be considered non-native. There'are currently no state run warm-water species stocking programs in
SNE (D. Halliwell pers. comm.) I followed a strict interpretation of native status for this assessment
The number of native species in SNE lakes increased with lake size (^=0.41, p=0.0001; Figure 3a), ranging
from 2 to 9 species (11 in one lake). Two features of these data differed from those in Midwestern streams.
First, there were apparently no fisWess lakes in SNE, while some fairly large streams (watersheds up to 2000
km2) in Ohio were fishless (Whittier and Rankin 1992). Second, the data formed a band rather than a wedge
shape (i.e., variance about the regression line was fairly constant rather than increasing with lake size). To be
a useful metric native species richness residual scores, from the regression with lake size, should generally
decrease with increasing stressor scores (i.e., fewer than "expected" natives in more stressed lakes and more
native species in less stressed lakes). However, there was no apparent pattern in these data (Figure 3b-c).
Jennings et al. (this book) also found no relationship between native species richness and human impact as
measured by a Trophic State Index. Total species richness showed similar patterns and lack of relationship to
stressor measures. Treating "naturalized" (non-stocked) North American species as natives produced a
pattern similar to that for total species richness.
The number of introduced species was a useful metric in Great Lakes littoral areas (Minns et al. 1994). In
SNE all, except three, lakes had between 1 and 6 non-native species (Figure 4a), generally increasing with
lake size (^=0.32 p=0.0001). Residual scores plotted against TP showed a lack of pattern with increased
stress, except that several of the relatively low TP lakes with high residuals (more non-native species than
-------
Whittier 12
expected) were stocked with 1-3 salmonids (Figure 3b). Non-native species residuals showed the expected
pattern with watershed disturbance (Figure 3c); a number of lakes with high residual scores and intermediate
watershed disturbance were stocked. Thus, the number of non-native species adjusted for lake size should be
a useful metric.
Most other species richness and composition metrics used in other IBIs did not appear to be applicable in
SNE. There were only two rarely collected darters (and no sculpins) (Table 2). Likewise, there were only
two suckers, with the creek chubsucker collected only once. Cyprinid species richness was somewhat higher;
however 4 of 5 native minnows were uncommonly collected. Golden shiner was widespread (72% of
sampled SNE lakes) and is tolerant of degraded conditions. Only 8 of the 55 sampled lakes had native
cyprinids other than golden shiner; six of these had only one other minnow. Native minnow species richness
may be a useful metric, but it would be essentially a binary metric. Of the three native sunfish in SNE (Table
2), pumpkinseed was ubiquitous (91 % of sampled lakes). The other two sunfish were uncommon and were
not collected in the same lakes, preferring very different habitats. Therefore a native sunfish metric contained
very little information.
Atlantic salmon and fathead minnow were the only SNE lake fish rated intolerant of anthropogenic stress in
Northeast lakes by Whittier and Hughes (in review). Lowering the threshold to tolerance scores of 11 (of 16)
added 13 uncommon species (11 of these were collected in only 1 or 2 lakes each; Table 2). Only five lakes
had more than one species from the expanded list of intolerants and most had none. There was a very weak
relationship between lake size and intolerant species (^=0.14, p=0,0045). Maximum intolerant species
(uncorrected for lake size) tended to decrease with increased TP, most lakes with intolerant species >0 had
moderate to low watershed disturbance. Thus, this metric may be useful, although most lakes would get the
minimum metric score.
-------
Whittier 13
One of the features of the original sucker species metric was their longevity, providing a "muJtiyear
integrative perspective" (Karr et al. 1986). To address the large-bodied, long-lived component of lake fish
assemblages, I selected five relatively large native species to comprise an analogous metric. There was a
significant, but weak, relationship between large species richness and lake size (^=0.19, p=0.009; Figure 5a).
There was little association between increased TP and fewer large native species (Figure 5b). Both the
regression residuals (Figure 5c) and the raw species counts tended to decrease with increased watershed
disturbance. The four lakes circled in the lower left of Figure 5c are subject to manipulations (e.g.,
reclamation, draining) which are not detected by either stressor measures. This appears to be a useful metric.
A similar analysis using 10 native small species (4 minnows, 2 sunfish, 2 darters, and killifish) showed no
associations with any of the stressor measures.
Perhaps the most frequently replaced metric is % green sunfish (Simon and Lyons 1995). In SNE, % bluegill
would be the most likely candidate. Bluegill is not native and sometimes dominated the assemblages, being
most abundant in CT and RI, becoming less dominant in MA and being replaced by yellow perch as a
dominant in NH. It occurred in only two NHE lakes. In the NCZ ecoregion where it was most widespread, %
bluegill was not associated with higher stressor measures. A variant metric might be % individuals of the
most abundant non-native species (often bluegill). This did not improve the metric performance. Somewhat
better was the proportion of individuals of all non-native species (or conversely proportion of native
individuals). This metric showed no lake size effect, and no relationship with TP (Figure 6a). However, there
was a tendency for the % non-native individuals to increase with watershed stress. Most of the lakes in the
upper left of Figure 6b had small watersheds and were probably more stressed than indicated by the
watershed analysis. This metric is probably useful.
-------
Whittier 14
There were nine tolerant species collected (Table 2). Eleven of the 12 smallest lakes had >90% tolerant
individuals (Figure 7a). Pumpkinseed were fairly abundant in many of these ponds. Removing pumpkinseed
from this metric had the greatest effect in smaller lakes. There was a clear association between % tolerants
and TP. With pumpkinseed included, all eutrophic and hypereutrophic lakes (except one) had >80% tolerant
individual (Figure 7b). The pattern was less dramatic with pumpkinseed removed (Figure 7d). For watershed
disturbance, the pattern of increasing proportion of tolerant individuals with increasing stress was stronger
with pumpkinseed removed (Figure 7c & e).
TROPHIC COMPOSITION METRICS
As with the species richness metrics, the trophic composition metrics were problematic in SNE lakes. For
example, chain pickerel and American eel were the only native top carnivores (Table 2). Other natives such
as yellow perch, white perch and brook trout are piscivorous as large adults. However, white perch is native
in only about a third of the SNE lakes and tolerant of stressed conditions. Brook trout is native to SNE
streams, but is nearly always maintained by stocking in lakes. The most widespread top carnivore was the
non-native largemouth bass (in 50 of 55 sampled lakes) which was often found in quite degraded lakes, as
was black crappie. I selected eight species that are strongly piscivorous and not highly tolerant of stressed
conditions (Table 2). I also examined this metric with largemouth bass included. Any species maintained by
stocking was not included for that lake, which effectively eliminated Atlantic salmon and lake trout.
For the restricted species list (excluding largemouth bass), there was a slight tendency for % top carnivores to
increase with lake size. The highest scores occurred in the lakes with TP < 20 and low to moderate watershed
disturbance (Figure 8a-b). The one outlier lake had 19% of individuals as rock bass; removing rock bass
would still leave this lake with one of the highest % top carnivores scores. Including largemouth bass tended
to decrease the resolution of this metric (Figure 8c-d). That is, the peaks of the scatter plots moved to the
-------
Whittier 15
right on the stressor plots, and the range of scores approximately doubled. There is concern that removing a
highly tolerant species changes the metric from a trophic guild metric into a tolerance metric. I believe that
the original intent of the metric is maintained with the reduced list.
A % insectivorous individuals metric also required some adjustments. Only seven uncommonly collected
species meet the strict definitions of insectivory of Halliwell et al. (this book). I selected six additional native
generalist feeder species that tend toward the insectivorous end of the trophic spectrum, and which are not
highly tolerant of degraded conditions (Table 2). With this species list % insectivorous individuals scores
related very well to TP and watershed disturbance (Figure 9b-c). The three outlier lakes in the upper right of
the watershed disturbance plot all had yellow perch as the most abundant species. Also problematic for this
metric was that 11 of 12 smallest lakes had <10% insectivores (Figure 9a). Eight of these small lakes had
pumpkinseed as a dominant species. Adding pumpkinseed to the insectivore species list for small lakes
removed the size effect, but did not improve the relationship with the stressor measures.
There are few true omnivore species (with a substantial portion of their diet including plant material) in the
Northeast (Halliwell et al. this book), but a fairly large number of generalist feeder species (feeding at several
trophic levels rather than primarily from one trophic level [Minns et al. 1994]). I selected 10 species for a
generalist feeder metric (Table 2). The maximum scores for % generalist feeder individuals decreased with
lake sizes greater than about 30 ha (Figure lOa). Most of the small lakes had high scores due generally to
dominance by bluegill and pumpkinseed. The lowest scores (< 30%) were all in relatively low productivity
lakes (TP < 20). The highest scores tended to be in lakes with moderate levels of stress (Figure lOb-c).
-------
Whittier 16
DISCUSSION
Seven of the candidate metrics appear to be successful and should provide a framework for future refinement
of lake metrics and development of an ffil for inland lakes in southern New England. Two are species
richness metrics, non-native species richness adjusted for lake size as a negative metric, and large species
richness (which may not need to be adjusted for lake size) as a positive metric. The % non-native individuals
and % tolerant individuals metrics could be considered in the richness and composition category (Karr et al
1986) or in the indicator species category of Simon and Lyons (1995) as negative metrics. Minns et al.
(1994) placed the former metric in the abundance and condition category. It may be possible to develop an
intolerant species metric, or a small species metric. With judicious selection of species, all three trophic
composition metrics appear to be useful.
The results presented here also illustrate some of the conceptual and ecological issues important for assessing
biotic integrity of inland lakes, and differences between lotic and lentic fish assemblages. Lakes in New
England are naturally species depauperate (Schmidt 1986). In this way they are somewhat analogous to
coldwater streams, where increased total species richness is usually an indication of degradation (Lyons et al.
1996). In New England lakes, the source of additional species is introductions, rather than immigration from
warmwater streams. The relationship between native species richness and introductions is complex for SNE
lakes. Moyle and Light (1996) proposed that most successful invasions occur without loss of native species
and that non-native predators should have greater effect on native assemblages than non-predators. A
substantial proportion of non-native species in SNE are predators. For the Northeast as a whole, Whittier et
al. (1997a) demonstrated an association between increased predator richness (usually from non-natives) and
lower minnow richness. However, preliminary analyses could not demonstrate an overall decrease in native
species richness with increased numbers of non-natives species (T. Whittier unpub. data). Thus, in SNE
-------
Whittier 17
there appears to adequate ecological "space" for the additional non-native species in general, but some
portions of the native assemblage may have been extirpated. It probably cannot be known whether SNE lakes
had more minnow species prior to introduction of additional littoral predators.
A third rule for species invasion (Moyle and Light 1996) was that increased invasion success occurs when
native assemblages are depleted or disrupted. The greater introduced species richness, and higher proportion
of non-native individuals in more stressed lakes support this idea. However, it appears that these stress levels
have not generally been high enough to eliminate native species, because native species richness showed no
association with anthropogenic stress.
Fish species tolerance of, or sensitivity to degraded lake conditions appear to differ from their tolerances to
impaired lotic systems. When I used species tolerance classifications from stream IBIs the proportion of
tolerant individuals was lowest in the most stressed lakes (T. Whittier unpub. data). This led Whittier and
Hughes (in review) to evaluate individual species tolerances to five stressor measures in Northeast lakes, and
assign new tolerance ratings. Applying these revised classes to the SNE assemblage data generally produced
the expected associations with stressor measures, except in the smallest lakes and ponds (g 10 ha). These
results, coupled with the larger species pool for lotic fish compared with lentic fishes, pose some interesting
questions relative to how lake ecosystems stress fish compared stream ecosystems, and whether there are any
truly intolerant warmwater lake species.
The trophic composition metrics presented a number of interesting challenges, in addition to the guild
membership issues (e.g., lack of objective criteria, dietary plasticity and changes with age) discussed by
Minns et al. (1994). In SNE lakes, a number of species with obvious trophic guild membership react to
degraded conditions in the opposite manner to what the metric should indicate. The clearest examples are the
-------
Whittier 18
piscivores. For most of SNE chain pickerel and American eel are the only native top carnivores. Largemouth
bass is clearly a top carnivore, but is tolerant of degraded lake conditions. Black crappie and white perch are
usually placed in the piscivore guild, but are two of the most tolerant species in SNE lakes (Whittier and
Hughes in review), and among the dominant species in many stressed lakes. Finally, 16 of 40 species
collected by EMAP in SNE lakes could be considered as piscivores (Halliwell et al. this book), producing
very high % piscivorous individuals scores (40 -100% for more than half of the lakes). In selecting species
for this metric, I did not try to choose the most sensitive piscivores, but limited the list to the most
carnivorous species that were also not highly tolerant, regardless of the native status. This produced metric
values in the 0 -10% range and the expected association between the metric and stressor scores. Similar
adjustments were needed for the other two trophic composition metrics.
i
Additional work is needed in a number of areas. First, biomass data may have the potential to improve all of
the trophic guild metrics, and provide additional abundance and condition metrics (Minns et al. 1994).
However EMAP, which was not primarily a fisheries survey, did not collect those data. Second, some
standard unit of sampling effort needs to be established. Multiple gear sampling makes that a complex
problem. A sampling scheme combining electrofishing and gill nets may provide abundance data with lower
variability. Third, additional work on lake classification schemes is needed. Even within a relatively
homogenous area like southern New England there are questions about whether lakes in the Northeast
Coastal Zone ecoregion should be held to be same standard as those in the Northeast Highlands-ecoregion.
Finally, work is needed to develop metric scoring and an overall ffil, and to validate both the metrics and the
index with additional data.
-------
Whittier 19
ACKNOWLEDGMENTS
I thank Bob Hughes, Dave Halliwell and Bob Daniels for numerous insightful discussions of these issues.
Martin Jennings, Melissa Drake, Dave Peck, and Bob Hughes provided useful reviews of an earlier draft.
This research was supported by the U.S. Environmental Protection Agency through Contract Number 68-C5-
0005 to Dynamac International, Inc.
REFERENCES
Baker, J. R., Peck, D. V., and Sutton, D. W. (Eds). 1997. Environmental Monitoring and Assessment
Program-Surface Waters: Field Operations Manual for Lakes. EPA/620/R-97/001. U.S. Environmental
Protection Agency, Washington, DC.
Becker, G. C. 1983. Fishes of Wisconsin. University of Wisconsin Press. Madison, WI.
Census Bureau. 1991. Census of population and housing, 1990: Summary tape file I on CD-ROM (name of
State). Washington, DC.
Chaloud, D. J,, and Peck, D. V. (eds.). 1994. Environmental Monitoring and Assessment Program:
Integrated Quality Assurance Project Plan for the Surface Waters Resource Croup, 1994 Activities. EPA
600/X-91/080 (Revision 2.00). U.S. Environmental Protection Agency, Las Vegas, NY.
-------
Whittier 20
Christensen, D. L., Herwig, B. R., Schindler, D. E., and Carpenter, S. R. 1996. Impacts of lakeshore
residential development on coarse woody debris in north temperate lakes. Ecological Applications. 6:1143-1149.
Dionne, M, and Karr, J. R. 1992. Ecological monitoring of fish assemblages in Tennessee River reservoirs.
in Ecological Indicators, McKenzie, D. H,, Hyatt, D. E., and McDonald, V. J.. Eds., Elsevier Applied
Science. London. 259-281.
Fore, L. S., Karr, J. R., and Wisseman, R. W. 1996. Assessing invertebrate responses to human activities:
evaluating alternative approaches. Journal of the North American Benthological Society. 15:212-231.
Gaueh, H. G. Jr. 1982. Multivariate Analysis in Community Ecology. Cambridge University Press, New
York, NY.
Halliwell, D. B. 1989. A classification of streams in Massachusetts - to be used as a fisheries management
tool. Ph.D. Dissertation. University of Massachusetts. Amherst, MA.
Halliwell, D. B., Langdon, R. W.f Daniels, R. A., Kurtenbach, J. P., and Jacobson, R. A. (this book).
Classification of freshwater fishes of the Northeastern United States for use in the development of indices of
biological integrity, in Assessing the Sustainability and Biological Integrity of Water Resource Quality
Using Fish Assemblages. Simon, T. P. Ed., CRC Press. Boca Raton, FL.
Hubbs, C. L., and Cooper, G. P. 1936. Minnows of Michigan. Bulletin of the Cranbrook Institute of Science.
No. 8. Bloomfield Hills, MI.
-------
Whittier 21
Hubbs, C. L., and Lagler, K. F. 1964. Fishes of the Great Lakes Region. Bulletin of the Cranbrook Institute
of Science. No. 26. Bloomfield Hiiis, MI.
Jennings, M, J., Fore, L. S., and Karr, J. R. 1995. Biological monitoring of fish assemblages in Tennessee
Valley reservoirs. Regulated Rivers: Research and Management. 11:263-274.
Jennings, M. J., Bozek, M., Hatzenbeler, G., and Fago, D. 1996. Fish and habitat, in Shoreline Protection
Study: A Report to the Wisconsin State Legislature. PUBL-RS-921-96. Wisconsin Department of Natural
Resources. Madison, WI. pp 1-48.
Jennings, M. J., & 7 others, (this book). Toward the development of an index of biotic integrity for inland
lakes in Wisconsin, in Assessing the Sustainability and Biological Integrity of Water Resource Quality
Using Fish Assemblages. Simon, T. P. Ed., CRC Press. Boca Raton, FL.
Karr, J. R. 1981. Assessment of biotic integrity using fish communities. Fisheries 6(6):21-27.
Karr, J. R. 1991. Biological integrity: a long-neglected aspect of water resource management. Ecological
Applications. 1:66-84.
Karr, J. R., Fausch, K. D., Angermeier, P. L., Yant, P. R., and Schlosser, I. J. 1986. Assessing Biological
Integrity in Running Waters A Method and Its Rationale. Special Publication No. 5. Illinois Natural History
Survey, Champaign, IL.
-------
Whittier
Larsen, D. P., and Christie, S. J. Eds. 1993. EMAP-Surface Waters 1991 Pilot Report. EPA/620/R-93/0
U.S. Environmental Protection Agency. Corvallis, OR.
Larsen, D. P., Stevens, D. L., Selle, A. R,, and Paulsen, S. G. 1991. Environmental Monitoring and
Assessment Program, EMAP-Surface Waters: A northeast lakes pilot Lake and Reservoir Management.
7:1-11.
Larsen, D. P., Thornton, K. W., Urquhart, N. S., and Paulsen, S. G. 1994. The Role of sample surveys for
monitoring the condition of the Nation's lakes. Environmental Monitoring and Assessment. 32:101-134.
Lee, D. S., Gilbert, C. R., Hocutt, C. H., Jenkins, R. E., McAllister, D. E., and Stauffer, J. R., Jr. 1980. Atlat
of North American Fishes. North Carolina Biological Survey. Raleigh, NC.
Leonard, P. M., and Orth, D. J. 1986. Application and testing of an index of biotic integrity in small,
coolwater streams. Transactions of the American Fisheries Society. 115:401-415.
Lyons, J. 1992. Using the Index of Biotic Integrity (IBI) to Measure Environmental Quality in Warmwater
Streams of Wisconsin. General Technical Report NC-149. U.S. Department of Agriculture, Forest Service,
North Central Experimental Station.
Lyons, J., Wang, L., and Simonson, T. D. 1996. Development and validation of an index of biotic integrity
for coldwater streams in Wisconsin. North American Journal of Fisheries Management. 16:241-256.
-------
Miller, D. L., & 13 others. 1988. Regional application of an index of biotic integrity for
management Fisheries. 13 (5): 12-20.
Minns, C. K., Cairns, V. W., Randall, R. G., and Moore, J. E. 1994. An index of biotic ir
assemblages in littoral zone of Great Lakes* areas of concern. Canadian Journal of Fish
Sciences. 51:1804-1822.
Moyle, P. B., and Light, T. 1996. Biological invasions of fresh water: empirical rules and
Biological Conservation. 78:149-161.
New Hampshire Fish & Game Commission (NHFGC). 1937-1938. Biological Surveys - '.
Concord, NH.
New York State Department of Conservation (NYSDC). 1927-1939. Biological Surveys -
Albany, NY.
Ohio Environmental Protection Agency (OEPA). 1988. Biological Criteria for the protectic
life: Volume I. The role of biological data in water quality assessment. OEPA Division of \
Monitoring and Assessment Columbus, OH.
Ohio Environment Protection Agency (OEPA), 1989. Biological Criteria for the Protection c
Volume HI. Standardized Biological Field Sampling and Laboratory Methods for Assessing
Macroinvertebrate Communities. OEPA Division of Water Quality Planning and Assessment
OH.
-------
Wh
Omernik, J. M. 1987. Ecoregions of the conterminous United States. Annals of the Association of
Geographers. 77: 118-125.
Page, B, M. and Burr, L. M. 1991. A Field Guide to Freshwater Fishes, Houghton Mifflin. Boston
Plafkin, J. L., Barbour, M. T., Porter, K. D., Gross, S. K., and Hughes, R. M. 1989. Rapid bioassest
protocols for use in streams and rivers: benthic macroinvertebrates and fish. U.S. Envkonmental
Protection Agency. EPA/444/4-89-001, Washington, DC.
SAS Institute. 1985. SAS User's Guide: Statistics, Version 6. Gary, NC.
Schmidt, R. E. 1986. Zoogeography of the Northern Appalachians. In. The Zoogeography of North
American Freshwater Fishes. Hocutt, C. H., and Wiley, E. O. Eds. John Wiley & Sons. New York, pj
159.,
Scott, W. B., and Grossman, E. J. 1973. Freshwater Fishes of Canada. Bulletin of the Fisheries Rese;
Board of Canada, 184.
Simon, T. P., and Lyons, J. 1995. Application of the Index of Biotic Integrity to Evaluate Water-Resour
Integrity in Freshwater Ecosystems, in Biological Assessment and Criteria: Tools for Water Resource
Planning and Decision Making. Davis, W.S. and Simon, T. P. Eds., Lewis Publishers. Boca Raton, FL.
245-262.
-------
Smith, L. C. 1985. The Inland Fishes of New York State. Department of Environi
Albany, NY.
Steedman, R. J. 1988. Modification and assessment of an index of biotic integrity ti
in southern Ontario, Canadian Journal of Fisheries and Aquatic Science 45:492-;
Stevens, D. L., Jr. 1994. Implementation of a national monitoring program. Journal
Management. 42; 1 -29.
Trautman, M. B. 1981. The Fishes of Ohio. Ohio State University Press. Columbus,
Underhill, J. C. 1986. The Fish Fauna of the Laurentian Great Lakes, the St. Lawrence
Newfoundland and Labrador, in The Zoogeography of North American Freshwater F
and Wiley, E. O. Eds., John Wiley & Sons. New York. pp. 105-136.
U.S. Environmental Protection Agency (USEPA), Office of Information Resources Maj
Block/Precinct Level Coverage User Documentation. Washington, DC.
U.S. Geographic Survey (USGS) 1990. Land Use and Land Cover Digital Data from I
l:100,000-scale Maps. Data Users Guide 4, Reston, VA.
Weaver, M. J., Magnuson, J. J., and Clayton, M. D. 1993. Analyses for differentiating Ii
assemblages with catch data from multiple sampling gears. Transactions of the America,
122:1111-1119.
-------
Whittier 26
Whittier, T. R., and Hughes, R. M. (In review). Evaluation offish species tolerances to environmental
stressors in Northeast USA lakes.
Whittier, T. R., and Rankin, E. T. 1992. Regional patterns in three biological indicators of stream condition
in Ohio, in Ecological Indicators. McKenzie, D. H., Hyatt, D. E., and McDonald, V. J. Eds., Elsevier
Applied Science. London, pp 975-995
Whittier, T. R., Halliwell, D. B., and Paulsen, S. G, 1997a. Cyprinid distributions in Northeast USA lakes:
evidence of regional-scale minnow biodiversity losses. Canadian Journal of Fisheries and Aquatic Sciences
54:1593-1607.
Whittier, T. R., Vaux, P., Merritt, G. D., and Yeardley, R. B., Jr. 1997b. Fish Sampling, in Environmental
Monitoring and Assessment ProgramSurface Waters: Field Operations Manual for Lakes. Baker, J. R,,
Peck, D. V., and Sutton, D. W. Eds., EPA/620/R-97/001. U.S. Environmental Protection Agency,
Washington, DC.
-------
Wr
Table 1. Principal components analysis of watershed-level human influence measures: three Ian
variables (% forest, % urban, % agricultural), human population density(#/km2), and road densi
variables except % forest were Iogi0(x+l) transformed. The first Principal Component (PCA-T
generalized human influence.
Principal
Component
PC-1
PC-2
% Forests
% Urban
% Agricultural
Population Density
Road density
Eigenvalue
3.09
0.88
Eigenvectors
PCA-1
-0.42
0.49
0.32
0.52
0.46
Proportion of Ct
Variance
0.62
0.17
PCA-2
0.34
0.31
-0.81
0.21
0.30
-------
Whittier 28
Table 2. Fish species native status, tolerance ranks and trophic guild as used in this
study, collected at 55 southern New England lakes by EMAP during summers 1991-94.
Tolerance rating scores from Whitlier & Hughes (in review) range 4-16, I=Intolerant,
MI=Moderately Intolerant, M=Moderate, MT=Moderately Tolerant, T=ToleranL For
Trophic Guild: TC=Top Carnivore, GF=GeneraIist Feeder, IN=Insectivore.
Species
# of Lakes
# of Lakes as Introduced or Tolerance Trophic
as Natives Stocked Rating Guild
American eel
Anguilla rostrata
alewife
Alosa pseudoharengus
gizzard shad
Dorosoma cepedianum
common carp
Cyprinus carpio
common shiner
Luxilus cornutus
golden shiner
Notemigonus crysoleucas
bridle shiner
Notropis bifrenatus
spottail shiner
N, hudsonius
fathead minnow
Pimephales promelas
creek chub
Semotilus atromaculatus
fallfish
S. corporalis
8
37
9/M
10/M
5/T
12/M
5/T
13/MI
14/MI
15/1
12/M
11/M
TC
GF
IN
GF
IN
IN
GF
IN
IN
-------
Table 2. (continued)
# of Lakes
Species
white sucker
Catostomus commersoni
creek chubsucker
Erimyzon oblongus
white catfish
Ameiurus catus
yellow bullhead
A. natalis
brown bullhead
A. nebulosus
channel catfish
Ictalurus punctatus
tadpole madtom
Noturus gyrinus
northern pike
Esox lucius
chain pickerel
E. niger
rainbow smelt
Osmerus mordax
rainbow trout
Oncorhynchus tnykiss
Atlantic salmon
Salmo solar
brown trout
5. trutta
# of Lakes as Introduced or Tolerance Tropl
as Natives Stocked Rating Guil
23 8/MT GF
1 9/M IN
2 »/M
20 6/MT GF
42 5/T GF
2 --/M
1 , -Ml IN
3 10/M TC
34 7/M TC
2 11/M GF
7 --/MI
1 15/1 TC'
3 --/M
-------
Whittier 30
Table 2. (continued)
Species
# of Lakes
# of Lakes as Introduced or Tolerance
as Natives
Stocked
Rating
Trophic
Guild
brook trout
Salvelinus fontinalis
lake trout
S. namaycush
banded killifish
Fundulus diaphanus
mummichog
F. heteroclitus
white perch
Morone americana
rock bass
Ambloplites rupestris
banded sunfish
Enneacanthus obesus
redbreast sunfish
Lepomis auritus
pumpkinseed
L, gibbosus
bluegill
L. macrochirus
smallmouth bass
Micropterus dolomieu
largemouth bass
M. salmoides
black crappie
Pomoxis nigromaculatus
11
50
35
15
50
21
13/MI
14/MI
10/M
-IT
5/T
12/M
9/M
10/M
4/T
4/T
10/M
5/r
TC*
IN
GF
TC
IN
IN
GF
GF
TC
-------
Table 2, (continued)
Species
# of Lakes
# of Lakes as Introduced or Tolerance Trophic
as Natives Stocked Rating Guild
swamp darter
Ethesotoma Jusiforme
tesselated darter
E, olmstedi
yellow perch
Percaflavescens
44
IN
11/M
6/MT
IN
-------
Whittier 32
Figure Captions
Figure 1. Locations of the 55 lakes sampled by the Environmental Monitoring and Assessment Program
(EMAP) in southern New England during the summers of 1991-94.
Figure 2. Relationship between sampled southern New England lake surface areas and two stressor
measures: total phosphorus (ug/L), and watershed disturbance (1st axis PCA scores for % forest, % urban, %
agricultural landuses, road density and human population density), plotted by ecoregions (a & c: solid circles
= Northeast Highlands ecoregion, open circles = Northeast Coastal Zone ecoregion), and lake type (b & d:
solid circles = named Reservoirs, open circles = named Ponds or Lakes).
Figure 3. Native Species Richness, (a) Number of native species by lake size, (b) Residuals of the native
species by lake size regression plotted against total phosphorus, c) Residuals plotted against watershed
disturbance PCA axis 1 (higher axis 1 value = increased human activity in the watershed).
Figure 4. Non-Native Species Richness, (a) Number of non-native species by lake size, (b) Residuals of the
non-native species by lake size regression plotted against total phosphorus, c) Residuals plotted against
watershed disturbance PCA axis 1 (higher axis 1 value = increased human activity in the watershed).
Figure 5. Native Large Species Richness, (a) Number of native large species by lake size, (b) Residuals of
the native large species by lake size regression plotted against Total Phosphorus, c) Residuals plotted againsi
watershed disturbance PCA axis 1 (higher axis 1 value = increased human activity in the watershed).
-------
Figure 6. Percent of individuals of non-native species. (a) % non-native individuals b
% non-native individuals by watershed disturbance PCA axis 1 (higher axis 1 value = i
activity in the watershed).
Figure 7. Percent of individuals of tolerant species (a) % tolerant individuals by lake size
pumpkinseed. (b) % tolerant individuals by total phosphorus, including pumpkinseed. c
individuals by watershed disturbance including pumpkinseed. d) % tolerant individuals I
excluding pumpkinseed. e) % tolerant individuals by watershed disturbance excluding pt
Figure 8. Percent of individuals of top carnivore species, (a) % top carnivore individuals I
excluding largemouth bass, (b) % top carnivore individuals by watershed disturbance excli
bass, c) % top carnivore individuals by total phosphorus, including largemouth bass, (d) 95
individuals by watershed disturbance including largemouth bass.
Figure 9. Percent of individuals of insectivore species. (a) % insectivore individuals by Iak<
insectivore individuals by total phosphorus, c) % insectivore individuals by watershed disti
1 (higher axis 1 value = increased human activity in the watershed).
Figure 10. Percent of individuals of generalist feeder species. (a) % generalist feeder individ
size, (b) % generalist feeder individuals by total phosphorus, c) % generalist feeder individu;
disturbance PCA axis I (higher axis 1 value = increased human activity in the watershed).
-------
-------
Southern New England Lakes
Total Phosphorus
120 -
,-J00 '".
3; 80 :
w :
o 6° :
Q. '.
§ 40 :
r-
o. :
20 :
0 :
1
o a 120 -
100 :
80:
,
60 :
o
o * :
40 :
o o o ;
9 1 0 -
10 100 1000 1
b
0 °
o
0 °
o °
0 o o o
° 0 o *
m S» 0 ** Q W
9
10 100 1000
Lake Area (ha) Lake Area (ha)
Northeastern Highlands * Named Reservoirs
o Northeastern Coastal Zone ° Named Lake/Pond
Watershed Disturbance
4 -
,- 3 -
Watershed PCA-'
O -* N)
i i i
-1 -
-2 -
o o ° C 4 -
o
00
o 3 -
o
0 0
n O f. ~
o° o 0
cP °° °. o o °
0 " ~ 1 "
0 0 °
*
* *
-------
Native Species
CO
-------
Non-Native Species
10 -
8 -
(ft
-------
Native Large Species
6 -
4 -
OT
CD
*** ft
o *
CD
1 -
o -
1
3"
2-
to 1 -
CO
3
"Ti n
. 0 -
CD
Of _i .
-2-
-3 -
a
. *
* * MM m m mm m . *
.. ... ...
... *M .
10 100 1000
Lake Area (ha)
Phosphorus
b
2 -
.
\^f . *
* *
» o -
»»-»* . u
!»** * .
. «
v
1 * (
^* \
. -2-
T 1 1 1 1 ] 1 . 1 1 1 1 1 1 1 1 1 1 1 1 1 1 , 1 1 1 -3
Species Lakes
Brown Bullhead 42
Chain Pickerel 34
\lit !» 4 ^~ *^ C*t t *+ $* &+ &+ ^fc rt
White Sucker 23
American Eel 8
Creek Ghubsucker 1
Watershed Disturbance
c
. ^ 4
* .*
.. .*
* * * .» * »
.* *%» .
* * m *
^~~~ ^-*^- *
^ ^ .
* * .)
/ .
^^^ .
25
50
75
100
125
-2
-1
1
Total Phosphorus (^g/L)
Watershed PCA Axis 1
-------
% Non-Native Individuals
Phosphorus
Watershed Disturbance
100 -
80 :
10
to :
l60-
>
E 40
^ -
20 -_
0 :
a
. %
* * * »
A
«
* * *
* *
.* *
'x .
w
. £**
1 . . * i i i i i i | i i i . j t * i . j i < . i ]
0 25 50 75 100 125
100 T.
: b
80 : .
*
60 : m *
* » *
40 '-_ «% .
*' * . '
20 : * *
:**.. . .
n - * * *
-2-10 1 2 3 4
Total Phosphorus (n9/L)
Watershed PCA Axis 1
-------
% Tolerant Individuals
1UU
(D
(D 80
to
_to .c
ZJ Q. 60
"O C
lo «'
.2 20 -
O
c
0 -
% ,
. »* a
s
«
s
Species
Largemouth Bass
Pumpkinseed
Brown Bullhead
Golden Shiner
* * . * Bluegill
* Black Crappie
White Perch
* * * Common Carp
/
1 10 100 1000
Lake Area (ha)
Lakes
50
50
42
39
35
21
12
3
Phosphorus Watershed Disturbance
viduals
umpkinseed
O) 00 O
o o o
ig 40;
^ 20 :
0
C
o :
100 n
o> :
| 80 :
_to 2
"3 Q- 60 -
.'> S. '.
"° 0) 40 "
~~ '-6
"1 20:
Q)
0 :
+ . 100 -
* « ** .
* * * i_ «n
h "
: eo-
*jj
I 1 40 -
* t 2°:
n:
/ * * *
* * * *
+
* *
**
.' :"
D 25 50 75 100 125 -2 -1 0 1 2 3
* * 10° 1
* . 80 -
* ** d
. 60 -
^
V 40 -
;*
*# 20 :
.^-
r. . . . i i , . , , i . , , , i o -
,
. ; s
«
«
\. :/-
C
2
4
e
»
*
0 25 50 75 100 125
Total Phosphorus
-2-101234
Watershed PCA Axis 1
-------
Top Carnivores
50 -i
40 -
Iso-
'&
10 -
0 -
70 -
65,
/
40 -
"co
i30-
>
TJ '.
£ 20-
10 -
0 -
50 -
without
Largemouth Bass
40 -
» 30 -
20 -
.* 10-
.i">j*v f
without
Largemouth Bass
*
...
* ** . s ^
* . * ->« %«**«
0 25 50 75 100 125 -2-101234
70 -
^ with 65/:
, Largemouth Bass
40 -
* 30 -
*
20:
*** .
> >'. . 10:
***
i? * *
.*...*. * * o
/ with
Largemouth Bass
*
*
*
*
0
- . V*
./*'.
.* ; ***.*.
"1 i t t y | i i i i p i i r ! j i r r i |r"'i j i i | i ' ' * ' i l ' f * i ' ' ' ' i ' ' * * i * ' ! f i > ' ' i f '
0 25 50 75 100 125 -2-101 234
Total Phosphorus
Watershed PCA Axis 1
-------
Insectivores
13
a
100
80
' 60
i
i
40 -
20 -
0 -
. . a
»»
*
Species
Yellow Perch
Banded Killifish
: * Banded Sunfish
; ' » » Redbreast Sunfish
: * .* * * Fallfish
* »
. Swamp Darter
* * . * * Tesselated Darter
* *
* % ***f . * *
*****
1 10 100 1000
Bridle Shiner
Creek Chub
+ 4 Others
Lake Area (ha)
Lakes
44
11
6
5
5
3
2
2
2
Phosphorus Watershed Disturbance
100 -,
80 -
;
60 :
40 :
.
20 \
0 :
100 -
b
V
* * 80 -
* %
*
60 -
0 '.
****
! 4° :
t
* 20 :
* *
w* *
-------
Generalist Feeders
100
80 -
16
3 60 -
o
!>
"c 40 -
G£
20 -
0 :
^ .
- . Soecies
3
* ^ Pumpkinseed
. Brown Bullhead
* * Golden Shiner
* . Bluegill
f ** . White Sucker
. Yellow Bullhead
* ^
Common Carp
0 9
t . Rainbow Smelt
* * Mummichog
Fathead Minnow
1 10 100 1000
Lake Area (ha)
Lakes
50
42
39
35
23
20
3
2
1
1
Phosphorus Watershed Disturbance
100 n . 10° 1 .
80 :
t/)
§60-
."2 :
"^
"£ 40-
t ^u
20 :
0 :
* . D
80 -
~
w «
!
60 -
&
* 40 :
« ** 20 '
I ^
/« ;
o-
» * ^
* B
*
*
A A
*
*
* *
*
C
*
«
0 25 50 75 100
Total Phosphorus
125 -2-1012-34
Watershed PCA Axis 1
-------
NHEERL-COR-2167A
x ' TECHNICAL REPORT DATA
(Please read instructions on the reverse before co,
1. REPORT NO.
EPA/600/A-97/089
2.
4. TITLE AND SUBTITLE
Development of IBI metrics for lakes in southern New England
7, AUTHOR(S) Thomas R. Whittier
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Dynamac International, Inc.
US EPA NHEERL
200 SW 35th Street
Corvallis, OR 97333
12. SPONSORING AGENCY NAME AND AC
US EPA ENVIRONMENTAL RESEX
200 SW 35th Street
Corvallis, OR 97333
3DRESS
kRCH LABORATORY
5. REPORT DATE
6. PERFORMING ORGANIZATION
CODE
8. PERFORMING ORGANIZATION REPORT
NO.
10. PROGRAM ELEMENT NO.
1 1 . CONTRACT/GRANT NO.
1 3. TYPE OF REPORT AND PERIOD
COVERED
14. SPONSORING AGENCY CODE
EPA/600/02
15. SUPPLEMENTARY NOTES:
1 6, Abstract:
In the early 1980's J.R. Karr and his colleagues developed the Index of Biotic Integrity (IBI) to assess the ecological
condition of fish assemblages in warmwater streams in the Midwest, as a biological tool to further the goals of the
Clean Water Act. The IBI has been repeatedly modified for other assemblages, regions and ecosystems, with varying
degrees of success. However to date, there has been only limited effort toward developing IBIs for inland lakes. The
development of lake IBIs lags behind that for warmwater streams due in part to the increased complexity in the scope
of assessment and sampling issues in lakes. In addition, streams tend to be managed, or under the authority of state
departments of environmental quality which are concerned with issues defined by the Clean Water Act. These
agencies have been shifting their regulatory emphasis to the biological condition of streams. Lakes, on the other hand
tend to be managed by state fish and game departments whose mandates emphasize management for fishability, not
biotic integrity. Finally, a number of ecological factors are less well understood for lake fish assemblages.
This manuscript evaluates the performance of numerous candidate metrics for lake fish assemblages in southern New
England, comparing metric behavior with total phosphorus {as a measure of eutrophication stress) and a generalized
measure of human activity in the watersheds (% of land in urban and agriculture, road and population density). This
manuscript presents solutions to issues of species tolerances to stressors in lakes and to trophic guild membership.
These modifications improved the performance of four candidate metrics (% individuals of highly tolerant species, %
top carnivore individuals, % insectivorous individuals, % generalist feeder individuals). A total of seven metrics appear
to be useful for evaluating biotic integrity of lakes in southern New England,
17.
a. DESCRIPTORS
New England, lakes, metric
development, index of biotic integrity
18, DISTRIBUTION STATEMENT
KEY WORDS AND DOCUMENT ANALYSIS
b. IDENTIFIERS/OPEN ENDED
TERMS
19. SECURITY CLASS (This Report)
20. SECURITY CLASS (This page)
c. COSATI Field/Group
21. NO. OF PAGES: 43
22. PRICE
EPA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION IS OBSOLETE
------- |