United St.ites
Fnvimnmental Projection
Agency
Fn vi run mental Monitor my
rind Support L;ihnr.itorv
P 0 Box 1b077
Las Vegas NV 89114
EPA 600 7 79 163
July 1979
Rebearch and Development
Assessment of
Macroinvertebrate Monitoring
Techniques in an Energy
Development Area
A Test of the Efficiency of
Three Macrobenthic Sampling
Methods in the White River
Interagency
Energy-Environment
Research
and Development
Program Report
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad categories
were established to facilitate further development and application of environmental
technology. Elimination of traditional grouping was consciously planned to foster
technology transfer and a maximum interface in related fields. The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the INTERAGENCY ENERGY—ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from the effort
funded under the 17-agency Federal Energy/Environment Research and Development
Program. These studies relate to EPA'S mission to protect the public health and welfare
from adverse effects of pollutants associated with energy systems. The goal of the Pro-
gram is to assure the rapid development of domestic energy supplies in an environ-
mentally-compatible manner by providing the necessary environmental data and
control technology. Investigations include analyses of the transport of energy-related
pollutants and their health and ecological effects; assessments of, and development of,
control technologies for energy systems; and integrated assessments of a wide range
of energy-related environmental issues.
This document is available to the public through the National Technical Informatio
Service, Springfield, Virginia 22161
-------
EPA-600/7-79-163
July 1979
ASSESSMENT OF MACROINVERTEBRATE MONITORING
TECHNIQUES IN AN ENERGY DEVELOPMENT AREA
A Test of the Efficiency of Three
Macrobenthic Sampling Methods in the White River
by
J. E. Pollard
Biology Department
University of Nevada, Las Vegas
Las Vegas, Nevada 89154
and
W. L. Kinney
Water and Land Quality Branch
Monitoring Operations Division
Environmental Monitoring and Support Laboratory
Las Vegas, Nevada 89114
ENVIRONMENTAL MONITORING AND SUPPORT LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U. S. ENVIRONMENTAL PROTECTION AGENCY
LAS VEGAS, NEVADA 89114
-------
DISCLAIMER
This report has been reviewed by the Environmental Monitoring and Support
Laboratory, U. S. Environmental Protection Agency, and approved for
publication. Mention of trade names or commercial products does not
constitute endorsement or recommendation for use.
ii
-------
FOREWORD
Protection of the environment requires effective regulatory actions that
are based on sound technical and scientific data. This information must
include the quantitative description and linking of pollutant sources,
transport mechanisms, interactions, and resulting effects on man and his
environment. Because of the complexities involved, assessment of specific
pollutants in the environment requires a total systems approach that
transcends the media of air, water, and land. The Environmental Monitoring
and Support Laboratory-Las Vegas contributes to the formation and enhancement
of a sound monitoring data base for exposure assessment through programs
designed to:
• development and optimize systems and strategies for
monitoring pollutants and their impact on the environment
• demonstrate new monitoring systems and technologies by
applying them to fulfill special monitoring needs of '
the Agency's operating programs.
The relative efficiencies of three macroinvertebrate collection methods
are assessed in this report with reference to their use in areas of oil shale
development. Results presented herein can be used as a basis for developing
water quality monitoring programs for streams of the semiarid western
regions. Potential users of the information presented include federal,
state, and local environmental and health agencies, as well as private
organizations engaged in water quality monitoring and assessment. Further
information is available from the Water and Land Quality Branch, Monitoring
Operations Division.
e B'. Morgai
Di rector
Environmental Monitoring and Support Laboratory
Las Vegas, Nevada
-------
SUMMARY
The use of benthic macroiinvertebrates as indicators of water quality has
become standard practice for many water quality monitoring programs. It has
been demonstated that the conventional macroinvertebrate sampling methods
used in the eastern and subhumid regions of the United States do not provide
useful data in many instances in the semi arid western United States. Since
the West holds much of the country's coal and oil shale reserves, it is
imperative that reliable methods for biomonitoring be available before
massive energy development occurs. The purpose of the present study was to
evaluate the efficiencies of three methods of macroinvertebrate sampling in a
prototype river of the semi arid west.
Three methods of macroinvertebrate collection were evaluated for
selectivity, reproducibility, capture-effectiveness, and cost efficiency in
the White River near Meeker, Colorado. Samples were collected with a
standard Surber sampler, with a portable invertebrate box sampler (PIBS), and
using the standardized traveling kick method (STKM). The methods were
evaluated in riffles of the White River directly upstream and downstream from
its confluence with Piceance Creek as well as at a comparable riffle at an
upstream control station.
The STKM collected more animals and taxa per sample with equivalent or
lower variability than the other two methods tested. Uhile the Surber
sampler and the PIBS performed similarly in the vicinity of Piceance Creek,
their performance differed at the upstream control station where the PIBS
collected more animals and taxa per sample than the Surber sampler.
Similarly, while sample processing time did not significiantly differ from
the various methods of collection used at any station in the vicinity of
Piceance Creek, differences did exist at the upstream control station with
kick samples requiring the greatest amount of processing time and Surber
samples requiring the least.
The cost efficiency of various methods was estimated by calculating the
number of animals processed per unit time and the number of hours required to
provide standing crop estimates for each sample within a given level of
precision. Cost-efficiency estimates indicated that the STKM was superior to
the Surber or PIBS methods, particularly at the stations in the vicinity of
Piceance Creek.
Selectivity of sampling methodologies was evident at all stations during
all seasons. In general, the small-area Surber sampler and PIBS collected
the more tightly adherent forms such as Hydroptila sp. most efficiently
-------
while the STKM was more efficient at collecting the more loosely attached
forms such as the swimming mayfly Baetis sp.
It is one of the missions of the U.S. Environmental Protection Agency to
provide guidance in terms of biological monitoring methodology. Results of
this study will serve to improve the efficiency of biological monitoring
programs by identification of appropriate techniques for application in
semi arid regions including energy development areas.
-------
CONTENTS
Foreword Ill
Summary 1v
Figures and Tables viii
1. Introduction 1
2. Conclusions 4
3. Recommendations 5
4. Study Area 6
5. Materials and Methods 9
Sampling Methods 9
Sampling Design 9
Sampling Processing 11
6. Results 12
Total Count and Number of Taxa Per Sample 12
Selectivity of Sampling Methods 15
Time Required for Sample Processing 18
Cost Efficiency of Sampling Methods 21
7. Discussion 23
8. References 25
vii
-------
FIGURES
Number Page
1 Map of the study area and location of sampling
stations on the White River, Colorado 7
2 The surber sampler(S), PIBS (B), and traveling
kick net (K) 10
TABLES
Number Page
1 Means (X), sample sizes (N), and coefficients of
variation (CV) for total counts and total number of
taxa per sample for all samples collected 13
2 Two-way factorial analysis of variance of total counts
and total number of taxa for fall 1977 samples .... 14
3 Two-way factorial analysis of variance of total counts
and total number of taxa for spring 1978 samples ... 15
4 Analysis of variance and Student-Newman-Keuls (SNK)
stepwise multiple range test of total counts and
total number of taxa at the upstream control station,
spring 1978 16
5 Mean percentages of total counts per sample for all
major species collected in all samples during the
study 17
6 Means (X), sample sizes (N), and coefficients of
variation (CV) for processing time per sample in
minutes. Cost efficiency estimates are also
presented 19
7 Analysis of variance of processing time per sample in
minutes for three sampling designs 20
vm
-------
INTRODUCTION
This report evaluates biological monitoring approaches and techniques
tested in a short segment of the White River that may be impacted by oil
shale development activities. The evaluation is based on testing conducted
in a stream reach that is subject to impact by activities associated with
construction and operation of commercial-scale oil shale industries on two
federally leased oil shale tracts in Colorado's Piceance Basin.
Specifically, the report examines stream sampling methodologies and
procedures for purposes of characterizing macroinvertebrate communities,
detecting and quantifying changes in community composition and structure, and
associating such changes with causative factors to the extent possible.
Aspects of macroinvertebrate sampling evaluated include: sampling techniques
and devices; parameter selection (e.g., standing crop, number of taxa, etc.);
sampling schedules and intensity; and individual sample site selection and
station distribution.
As the focus in this country turns to energy development, efficient
methods for evaluating the environmental consequences of large-scale energy
production become of paramount interest. In the near future, streams of the
semi arid western United States will be impacted by intensive energy
development in the form of oil shale and coal mining and processing
activities. These activities may result in the release of a variety of
pollutants into our waterways including a vast array of potentially toxic
substances. It is extremely critical, therefore, that efficient methods for
monitoring the ecological effects of massive energy development on our
western waterways be available for use by regulatory agencies and industry as
these activities increase in intensity.
Currently, biological monitoring procedures appropriate for application
to waterways of the western energy resource areas are not well developed.
The U.S. Environmental Protection Agency (USEPA) has been strongly criticized
for failing to provide biological monitoring guidelines for the assessment of
water quality and receiving waters as required by the Federal Pollution
Control Act Amendments of 1972 [PL 92-500 (Westman 1977)]. While guidelines
are urgently needed nationwide, it must be recognized that biological
monitoring requirements vary widely from one area to another. The methods
which are to be used for biological monitoring of receiving waters must be
tailored to meet the specific regional monitoring requirements which vary
from one geographical region to another as well as within geographical
regions.
-------
It has become increasingly obvious that biological monitoring strategies
and techniques traditionally utilized in subhumid and humid regions of the
country are not necessarily well suited for application in semiarid regions
of the western United States where much of the energy resource development is
occurring. This is particularly relevant with respect to utilization of
stream bottom-dwelling macroinvertebrates as environmental indicators in
water quality assessment programs for streams which may be affected by oil
shale developmental activities. Conventional sampling techniques and devices
are of limited utility for aquatic macroinvertebrate investigations in
streams that receive most of their flow from snowpack runoff, resulting in
high discharge levels from April through July and low base flow during much
of the remainder of the year. Periodic intense thundershowers, although
typically localized and of short duration, result in episodic high flows
which further complicate conventional biomonitoring approaches. During high
water stages, use of conventional unit-area samplers is impossible. For
example, utilization of the Surber sampler is restricted to water of less
than 18 inches in depth. The inherent patchy distribution of
macroinvertebrates in these streams further restricts the utility of small
area samplers because large numbers of samples are required to provide
adequate representation of communities in a given habitat. The suspended
sediments and bed load carried in these streams, in combination with highly
variable flow patterns, frequently create unstable substrates for maintenance
of macroinvertebrate populations.
Stream-dwelling macroinvertebrates have long been used by aquatic
ecologists for purposes of evaluating water quality and ecological stability.
Macroinvertebrates are ideally suited for this purpose since they are easy to
sample, relatively immobile and consequently unable to avoid unfavorable
conditions, and are sensitive in varying degrees to most types of pollution.
Since most aquatic insects have life cycles of a year or more, any alteration
in community structure caused by pollution will remain in evidence for a
fairly long period (Hilsenoff 1977). For the above reasons, information
regarding the structure of macroinvertebrate communities can provide a vivid
portrait of the recent water quality history of a stream. For example,
perturbations in the macroinvertebrate community structure can reflect the
impacts of single, short-duration pollution events that would be easily
missed by periodic physical/chemical monitoring alone.
Many methods have been developed for collecting macroinvertebrates in
riffle substrates, ranging from the well-known Surber sampler to the
qualitative window screen method. Surber samplers have long been considered
the quantitative standard for the collection of benthic invertebrates (Hynes
1970). It has been pointed out by Kroeger (1972), however, that the Surber
sampler provides, at best, a biased estimate of benthic standing crop and is
selective for various species of the benthic community. Chutter (1972) also
pointed out the ineffectiveness of the Surber sampler in depths over 30 cm.
Use of enclosed box-type samplers eliminates some of the problems
associated with Surber samplers (Hynes 1971), but the early designs, such as
the Hess sampler, were somewhat cumbersome. Recently, Ellis-Rutter
Associates developed a portable invertebrate box sampler (PIBS) which is
-------
relatively easy to use in flowing waters and reportedly performs better than
the Surber sampler for estimating standing crop of riffle organisms
(Ellis-Rutter Associates 1973). Both Surber and box samplers, however, are
difficult to use in rapidly flowing waters and are restricted to use in
shallow riffle areas (Frost, Huni, and Kershaw 1971).
In addition to being restricted to shallow water use, small-area samplers
such as the PIBS and Surber sampler collect samples with high replicate
variability. Average coefficients of variation for Surber samples range from
50 percent (USEPA 1973) to over 80 percent for samples taken from fauna-poor
uniform riffle areas (Hornig and Pollard 1978). It is inevitable that
samples collected using small-area methods will have relatively high
variability due to the inherent patchiness of the macrobenthic fauna.
Methodologies that have the lowest possible sample variability associated
with them and yet still collect a relatively large number of organisms and
taxa per sample are optimal for biological monitoring programs. Kick samples,
collected by kicking up riffle substrates with subsequent entrapment of
debris and organisms in a net held downstream from the investigator, have
been used to: 1) facilitate the sampling process (Frost, Huni, and Kershaw
1971; Hynes 1970), and 2) to increase precision of the samples and expedite
the sampling process (Kinney, Pollard, and Horning 1978; Horning and Pollard
1978). For example, approximately three 30-second traveling kick samples can
be collected and field-processed in the time required to collect and process
a single Surber sample (Kinney, Pollard, and Horning 1978).
-------
CONCLUSIONS
1. In the present study, the STKM was the most efficient method of
macroinvertebrate collection in terms of capture-effectiveness, cost
efficiency, and reproducibility of replicate samples collected from rocky
bottom substrates.
2. The methods tested for macroinvertebrate collection efficiency varied in
performance according to season. The seasonality of macroinvertebrate
communities is therefore a significant factor in biological monitoring design
considerations for rivers of the semiarid west including the White River
adjacent to the Piceance Basin oil shale developments.
3. Processing time for all samples collected downstream from Piceance Creek
was significantly higher than for those samples collected upstream from
Piceance Creek. This held true for all sampling methods because of the high
content of filamentous algae in downstream samples. The STKM, on the other
hand, provided better cost-efficiency estimates than either the Surber
sampler or the PIBS for collections in this area.
4. All methods of collection were selective to some degree with the Surber
sampler and the PIBS collecting tightly adherent forms more efficiently and
the standardized traveling kick method more effectively collecting the
loosely attached forms.
-------
RECOMMENDATIONS
The recommendations contained herein apply to the Colorado portion of the
White River that is potentially subject to impact by oil shale developments.
In addition, these recommendations are designed to aid in the implementation
of biological designs for streams and rivers of the western United States.
1. It is recommended that riffle habitats similar to those in the
vicinity of Piceance Creek (which contain high amounts of filamentous algae)
be sampled by the standardized traveling kick method. This is recommended
because this method provided the largest number of organisms per sample with
the lowest replicate variability, and no increase in time required for sample
processing relative to other methodologies tested.
2. Since the standardized traveling kick method was decidedly superior
or at least as effective as the Surber sampler and PIBS in terms of
collection efficiency, cost efficiency, and ease of sampling, this method is
recommended for use in streams with characteristics similar to those of the
study reach. This method is particularly applicable in fauna-poor stream
reaches with patchy macroinvertebrate distributions.
3. It is recommended that comparisons of stations or sites within a
river system be restricted to stream habitats with similar substrate
composition. Only in this way can valid comparisons between locations be
made since different benthic substrates support different benthic
communities.
4. Because single-season sampling did not provide adequate
representation of the major taxa inhabiting the White River, it is
recommended that sampling be conducted at least during the spring and fall.
Until further investigations are conducted, it is not known whether sampling
in two seasons is sufficient to generally characterize the benthic macrofauna
of western streams, including the White River.
5. The utility of small-area samplers should not be disregarded for use
in relatively productive stream reaches where estimates of standing crop may
be a major concern. Under these circumstances, an enclosed design sampler
such as the PIBS is recommended over the open-frame Surber sampler because of
the higher efficiency of the box-type sampler.
-------
STUDY AREA
The White River drainage area may contain the most significant
undeveloped, energy mineral resource in the Colorado River Basin in the form
of oil shale deposits in the sedimentary rock of the Green River Formation.
The richest and thickest known oil shale deposits occur in Colorado's
Piceance Basin, and the largest shale-bearing area of the Green River
Formation lies in the Uinta Basin of Utah (USDI 1973). Large portions of
both these areas drain to the White River.
The White River originates in the dissected lava plateaus of the western
slope of Colorado's Rocky Mountains and terminates in the semiarid plains of
Utah's Uinta Basin where it joins the Green River within the bounds of
Uintah-Ouray Indian Reservation. Annual precipitation ranges from more than
60 centimeters in the high headwater plateaus to less than 18 centimeters at
lower elevations (USDI 1973). Most of the precipitation falls as snow at
high elevations, and runoff from the snowpack is fairly gradual. Conversely,
summer and fall thundershowers, which account for much of the precipitation
at lower elevations, result in rapid overland runoff of suspended sediment-
and mineral-laden waters.
The headwaters of the White River are in a relatively primitive forested
area that is impacted little by man's activities. Although iron and zinc
concentrations have been reported in excess of criteria standards (Fox 1977),
water quality in the upper reaches is generally good. The river in the
upstream reaches supports good trout populations and a diversified
invertebrate fauna. The river rapidly deteriorates in quality, partially as
a result of nonpoint-source-pollution loading, both natural and man-induced.
The river becomes increasingly turbid downstream, and chemical water quality
similarly undergoes a pronounced change characterized by a great increase in
dissolved solids. The downstream reaches are subject to rapid fluctuations
in flow and sediment transport, both suspended and bedload, resulting in a
rather meager warm-water fishery and a relatively unstable macroinvertebrate
fauna.
The study area of the White River addressed in this report is limited to
the stream reach in the immediate vicinity of the Piceance Creek confluence
and a control area upstream from Meeker, Colorado (Figure 1). Three sampling
stations were selected. One station was designated as the "upstream control"
station (WR 230, Figure 1) and was located in an area southeast of Meeker,
Colorado, near Rio Blanco County Road RB4, upstream from any significant
pollution inputs. The other two stations were located on the White River
directly upstream (WR 140) and downstream (WR 130) from its confluence with
Piceance Creek (Figure 1). These stations were designated as "upstream" and
"downstream" from Piceance Creek. The river in the vicinity of Piceance
Creek receives moderate levels of nonpoint source pollutants in the form of
-------
MOFFAT COUNTY
RIO BLANCO COUNTY
TRAPPERS
10 20 30
KILOMETERS
INDEX MAP
ANB
Figure 1.
Map of the Study Area and Locations of Sampling Stations on
the White River, Colorado.
-------
dissolved solids and suspended sediments. Piceance Creek is a potential
source of increased pollution to the river as a result of the ongoing oil
shale development in the Piceance Basin.
Uniform riffle areas with similar substrate composition, depth, and flow
characteristics were chosen at each station. All riffle substrates sampled
were composed of medium-sized (10- to 20-cm) cobbles underlain by a mixture
of sand and gravel. The riffles at the stations upstream and downstream from
Piceance Creek had considerably larger amounts of patchy filamentous algal
growths (Cladophora sp.) during the fall and larger amounts of algal debris
in the spring than did the upstream control station. In addition, the
station downstream from Piceance Creek was noticeably more silted than either
of the two upstream stations during the spring of 1978. Observations of
riffle substrates at the stations near Piceance Creek suggested that the
Creek might influence the silting characteristics of riffle substrates within
the sphere of its direct influence. No substantiation for this hypothesis
could be found in historical suspended sediment data (Pennak 1974; Everhart
and May 1973) or from determinations of suspended sediments from samples
collected upstream and downstream from Piceance Creek during spring, 1978
[461 mg per liter and 453 mg per liter, respectively (Pollard, unpublished
data)]. It has also been shown that Piceance Creek is a major contributor of
dissolved solids to the White River (Pennak 1974).
8
-------
MATERIALS AND METHODS
SAMPLING METHODS
Three methods of macroiinvertebrate collection were used in this study
(Figure 2). Samples were collected using: 1) a standard 0.093-m2 (1
square foot) Surber sampler with a 90-cm-long, conical-shaped net; 2) a
Rutter-Ellis portable invertebrate box sampler (PIBS) with a 76-cm-long,
conical-shaped net; and 3) a standardized traveling kick method (STKM) using
a triangular dip net with a mouth opening of 28 cm by 28 cm by 24 cm, and a
76-cm-long conical-shaped net held downstream from the investigator to
collect organisms dislodged by kicking the substrate. All nets were
constructed of Nitex® #571 netting (^30 mesh).
Surber and PIBS samples were collected by the methods outlined in Needham
and Needham (1962) for the collection of Surber samples. Individual sampling
sites for collecting Surber and PIBS samples were chosen where rocks were
visibly piled [selection of rich sites as described by Horning and Pollard
(1978)]. All riffle areas were chosen for uniformity of substrate, depth,
and current velocity.
Standardized traveling kick samples were collected using the triangular
dip net described in the following manner: the collector slowly moved
downstream, vigorously kicking the substrate and holding the net in his
forward path. All kick samples were standardized by holding the net in the
water for 30 seconds and traveling approximately 4 meters downstream.
Markers were placed beside the stream for orientation. An area approximately
3/4 by 4 meters (3 mz) was disturbed during the collection of each
standardized traveling kick sample. All samples were collected from uniform
riffle areas with comparable substrates by the same investigator. Uniform
riffles are those determined by the field investigator to have the least
variability of available sampling sites in terms of substrate composition,
water depth, and stream flow.
SAMPLING DESIGN
Fall. 1977
Surber and traveling kick samples were collected at the stations upstream
and downstream from Piceance Creek during fall 1977 to evaluate the
effectiveness of the two sampling methods. Five replicate samples each were
collected with a Surber sampler and by the STKM from the riffles at these
stations on October 4 and 5, 1977. Each set of five replicate samples
constitued a discrete sample set.
-------
Figure 2. The Surber sampler (S) PIBS (B) and traveling kick net (K).
10
-------
Spring. 1978
Portable invertebrate box and standardized traveling kick samples were
collected at the stations upstream and downstream from Piceance Creek during
spring 1978 using the previously stated design. Surber, PIBS, and traveling
kick samples were collected at the upstream control station during the same
time period. Five replicate samples were collected for each method used at
this station. A restricted portion (about 5 by 10 meters in area) of an
extremely uniform riffle was used for the collection of all samples. Surber
and PIBS sampling sites were selected and then sampled, working in an
upstream direction, before traveling kick samples were collected.
SAMPLE PROCESSING
Samples were initially transferred from the Surber sampler, PIBS, and
traveling kick nets to a bucket with a 12-strand-per-cm (30 mesh) screen on
the bottom to avoid an accidental loss of organisms. Samples were placed in
mason jars and preserved with 100 percent Formalin solution in volumes
approximately equivalent to the amount of organic debris in the sample,
resulting in at least a 5 percent Formalin solution.
In the laboratory, samples were washed clean of Formalin by placing them
in a jar covered with a screen of Nitex® 571 netting, pouring off the
Formalin, and then rinsing the samples thoroughly with water.
Macroinvertebrates and debris were sorted from gravel and sand by placing
the sample in a round-bottom container with water, agitating the sample, and
pouring off the debris and organisms. This process was repeated until no
organisms could be found in the gravel and sand remaining in the container.
Macroinvertebrates were then hand sorted from the debris in a shallow white
pan. All macroinvertebrates in the samples were identified to the species
level, when possible, and enumerated. Dr. Richard Baumann of Brigham Young
University, Provo, Utah, has confirmed the majority of the identifications.
Chironomids and simuliids have been identified only to family; generic
determinations of chironomids will be addressed in a separate report.
11
-------
RESULTS
TOTAL COUNT AND NUMBER OF TAXA PER SAMPLE
The standardized traveling kick method collected the largest and the
Surber sampler collected the smallest number of animals and taxa per sample
at the upstream control station (Table 1). The reproducibility of replicate
samples differed distinctly between sampling methods as evidenced by the
coefficients of variation associated with the means of total counts and total
taxa per sample. In the majority of cases, traveling kick samples had the
lowest coefficients of variation while Surber samples had the highest
coefficients of variation for these parameters. For example, at the upstream
control station, PIBS reproducibility was similar to, although slightly lower
than, kick samples for total count data (23.9 percent CV vs 20.0 percent CV).
Surber samples had the lowest reproducibility for total number of taxa at
this station, with PIBS and traveling kick samples, respectively, showing an
improvement in reproducibility. At the stations in the vicinity of Piceance
Creek, the differences between sampling method reproducibility were much more
striking than at the upstream control station.
The significance of differences between sampling methods in the mean
total count per sample and the mean number of taxa per sample was tested
using analysis of variance (ANOVA) procedures (Sokal and Rohlf 1969). A two-
way factorial analysis of variance (ANOVA) with fixed effects was used to
test the differences among methods of macroinvertebrate collection upstream
and downstream from Piceance Creek. A single classification ANOVA was used
to test the differences among these methods at the upstream control station.
Bartlett's test for homogeneity of variances was applied to determine if any
heteroscedasticity was present in the data (Sokal and Rohlf 1969). No
heteroscedasticity was detected in any of the analyses at the 0.05
probability level. Based upon these results, it was accepted that the
assumptions of normality and homogeneity of variances underlying ANOVA were
not violated and that the tests performed were valid (Sokal and Rohlf 1969).
Significant differences were detected between stations (locations) and
methods of collection upstream and downstream from Piceance Creek during fall
1977 for both total count and number of taxa data (Table 2). On the other
hand, total count and total number of taxa data were not significantly
different (P=0.05) during spring 1978 in terms of the method of sample
collection employed (Table 3). There were significant differences (P=0.05)
between stations, however, for both total counts and number of taxa during
this sampling period.
12
-------
TABLE 1. MEANS (X), SAMPLE SIZES (N), AND COEFFICIENTS OF
VARIATION (CV) FOR TOTAL COUNTS AND TOTAL NUMBER
OF TAXA PER SAMPLE FOR ALL SAMPLES COLLECTED
Sampling Method, location
Total Counts
X CV
Total Number Taxa
X CV
10/4/77
Surber, downstream 5
STKM, downstream 5
Surber, upstream 5
STKM, upstream 5
All Surbers 10
All STKM samples 10
All samples, upstream 10
All samples, downstream 10
4/3/78
PIBS, downstream 5
STKM, downstream 5
PIBS, upstream 5
STKM, upstream 5
All PIBS samples 10
All STKM samples 10
All samples, upstream 10
All samples, downstream 10
4/4/78
Surber, control 5
PIBS, control 5
STKM, control 5
1,031
1,371
336
777
683
1,074
556
1,200
1,190
1,641
59
181
625
911
120
1,416
42.5
36.9
51.5
23.5
48.8
36.1
31.9
40.0
55.6
43.3
54.2
31.4
75.1
55.4
51.0
48.6
654 42.3
1,002 23.9
1,661 20.0
25.6
28.4
17.2
23.0
21.4
25.9
20.1
27.2
23.2
20.8
7.8
9.8
15.5
15.3
8.8
22.0
15.2
22.4
23.8
7.1
8.4
19.8
16.6
12.8
12.1
18.0
7.6
12.7
7.1
21.0
16.8
15.4
10.2
18.7
10.6
23.4
21.8
18.6
In contrast to the stations upstream and downstream from Piceance Creek,
there were significant differences between methods of macroinvertebrate
collection at the upstream control station during spring 1978 for both total
counts and total number of taxa collected (Table 4). Patterns of similarity
between means of total counts and total taxa at the upstream control station
during spring 1978 were tested using Student-Newman-Keuls stepwise multiple
range procedure (Sokal and Rohlf 1969).
13
-------
TABLE 2. TWO-WAY FACTORIAL ANALYSIS OF VARIANCE OF TOTAL COUNTS AND TOTAL
NUMBER OF TAXA FOR FALL 1977 SAMPLES (The locations tested were
upstream and downstream from Piceance Creek and the methods tested
were Surber and traveling kick.)
Source of Variation
Degrees
of
Freedom
Sum of Squares Mean Square
TOTAL COUNTS
Totals 19
Location 1
Methods 1
Interaction 1
Error 16
TOTAL NUMBER OF TAXA
4,940,501.20
2,074,968.20
764,405.00
12,700.00
2,088,427.20
2,074,968.20
764,405.00
12,700.00
130,526.70
15.8969**
5.8563*
0.0973
Totals
Location
Methods
Interaction
Error
* = Significant
** = Significant
*** = Significant
19 500.55
1 252.05
1 101.25
1 8.45
16 138.80
at 0.05 probability
at 0.01 probability
at 0.001 probability
252.02
101.25
8.45
8.68
29.0548***
11.6715**
0.9741
Standardized traveling kick samples contained significantly larger numbers
of animals per sample than Surber or PIBS samples while the Surber and PIBS
samples did not differ significantly (Table 4). High variability of small
area samplers has definite disadvantages as exemplified in this case by the
fact that there were no statistical differences between Surber and PIBS
sample means while there was an actual difference of greater than 300 animals
per sample (Table 4). With one exception, however, the STKM provided a
higher mean number of animals and taxa per sample than either the Surber
sampler or the PIBS (Table 1).
14
-------
TABLE 3. TWO-WAY FACTORIAL ANALYSIS OF VARIANCE OF TOTAL COUNTS AND TOTAL
NUMBER OF TAXA FOR SPRING 1978 SAMPLES (The locations tested
were upstream and downstream of Piceance Creek and the methods
tested were PIBS and traveling kick.)
Degrees
of
Source of Variation Freedom Sum of Squares Mean Square
F
TOTAL COUNTS
Totals 19
Location 1
Methods 1
Interaction 1
Error 16
TOTAL NUMBER OF TAXA
12,737,841.80
8,392,896.80
410,124.80
134,808.20
38,000,012.00
8,392,896.80
410,124.80
134,808.20
237,500.75
35.3384***
1.7268
0.5676
Totals
Location
Methods
Interaction
Error
* = Significant
*** = Significant
19
1
1
1
16
at 0.05 probability
at 0.001 probability
960.80
871.20
0.20
24.20
65.20
871.20
0.20
24.20
4.08
213.7914***
0.0491
5.9387*
SELECTIVITY OF SAMPLING METHODS
Species selectivity by the various sampling methods was not striking at
the upstream control station, although some evidence for sampler selectivity
did exist in the data (Table 5). It appeared that chironomids accounted for
a higher mean percentage of the total number of animals per sample collected
(percent total) with the Surber sampler than with the PIBS or STKM at the
upstream control station. Traveling kick samples had a higher mean percent
total for Hydroptila sp. than Surber and PIBS samples (Table 5). Ephemeral!a
inermis also appeared to be collected most effectively by the standardized
traveling kick method, and least effectively by the Surber sampler.
15
-------
TABLE 4. ANALYSIS OF VARIANCE AND STUDENT-NEWMAN-KEULS STEPWISE MULTIPLE RANGE TEST (SNK) OF TOTAL
COUNTS AND TOTAL NUMBER OF TAXA AT THE UPSTREAM CONTROL STATION, SPRING 1978 (The
methods tested were Surber, PIBS, and standardized traveling kick. Non-significant
(p = 0.05) subsets of group means are indicated by vertical lines; significant (p = 0.05)
differences between group means are indicated by horizontal lines.)
Source of
Variation
TOTAL COUNTS
Total
Methods
Error
TOTAL NUMBER OF
Total
Method
Error
* - Significant
** = Sianifieant
Degrees
of
Freedom
14
2
12
TAXA
14
2
12
at 0.05
at 0.01
Sum of Squares Mean Square
3,592,400.0
2,613,200.0 1,306,600.0
969,197.2 81,599.8
437.7333
212.9333 160.4667
224.8000 18.7333
probability
nrobabilitv
F Method Mean SNK
Surber 654
16.0124** PIBS 1,001
Traveling kick 1,661
Surber 15
5.6833* PIBS 22
Traveling kick 24
-------
TABLE 5. MEAN PERCENTAGES OF TOTAL COUNTS PER SAMPLE FOR ALL MAJOR SPECIES COLLECTED IN ALL SAMPLES
DURING THE STUDY (A major species is one that represented at least 10% of the total count
of any sample set. S = Surber sampler, K = 30-second traveling kick method.)
Downstream
Piceance
Creek
10/4/77
Taxa
Chironomidae
Oligochaeta
Ephemerel 1 a
i nenrn s
Tricorythodes sp
Baetis sp
Pseudocloeon sp
Rlthrogena
undulata
Isoperla sp
Cheumatopsyche spp
Hydropsyche spp
Hydroptila sp
S
5.2
0.8
0.4
48.7
0.2
4.8
0.0
0.3
12.6
20.3
1.3
K
4.8
0.5
1.3
29.3
1.3
15.4
0.3
1.3
12.7
22.9
0.6
Upstream
Piceance
Creek
10/5/77
S K
7.0
5.3
1.4
14.0
0.6
4.4
20.8
1.0
27.1
12.6
0.0
6.2
1.1
3.4
43.5
0.6
9.4
12.4
1.3
9.2
5.4
0.5
Downstream
Piceance
Creek
4/3/78
PIBS
18.1
29.3
8.1
20.4
9.0
0.1
0.2
1.3
3.4
6.3
0.3
K
23.2
7.5
16.5
31.5
10.5
0.0
0.4
0.9
2.3
3.0
0.2
Upstream
Piceance
Creek
4/3/78
PIBS
19.6
18.5
6.9
1.5
36.5
0.0
2.6
9.6
0.3
1.4
0.0
K
11.4
16.0
12.2
1.3
40.7
0.0
3.0
8.2
0.8
0.4
0.1
Upstream
Control Station
4/4/78
S
58.8
7.2
3.2
0.0
10.3
0.0
0.0
0.1
0.1
0.0
12.3
K
40.5
5.2
16.1
0.1
20.9
0.0
0.0
0.3
0.1
0.1
4.3
PIBS
41.7
8.9
9.0
0.1
8.9
0.0
0.0
0.2
0.1
0.1
14.9
-------
At the station upstream from Piceance Creek in spring of 1978,
chironomids appeared to have been collected slightly more effectively by the
PIBS, while Ephemerella inermis was again favored by the STKM. The STKM and
PIBS methods appeared to be quite selective at the station downstream from
Piceance Creek during spring 1978, with oligochaetes comprising 22 percent
more of the mean percent total in the PIBS than in STKM samples. On closer
inspection, one highly aberrant PIBS sample was discovered with 69.4 percent
oligochaetes in it. Percentage data based on relatively small total counts
(^300) is highly sensitive to the widely variable results obtained from
sampling patchy benthic communities with small-area samplers, and must be
interpreted with care.
Results of sampling during the fall 1977 indicate that the Surber sampler
and the STKM were very selective at stations upstream and downstream from
Piceance Creek (Table 5). At the station upstream from the creek, STKM
samples contained higher percentages of Tricorythodes sp., while Surber
samples contained higher percentages of Rithrogena undulata, Hydropsyche
spp., and Cheumatopsyche spp. This pattern was almost reversed at the
station downstream of the creek where Surber samples contained much higher
percentages of Tricorythodes sp. than did STKM samples. Unlike the spring
data, no anomalies existed in the raw data to account for this pattern.
It was interesting to note that in most cases there were considerable
differences in the species composition of samples collected in the same
season upstream and downstream from Piceance Creek. In a majority of cases
these differences were evident in both methods of collection used. The
percentages of Tricorythodes sp. in samples collected in fall 1977 both
upstream and downstream from Piceance Creek were notable exceptions.
TIME REQUIRED FOR SAMPLE PROCESSING
All methods of macroinvertebrate collection compared at the stations
upstream and downstream from Piceance Creek required very similar amounts of
time to process (sort and count) (Table 6). Surber samples at the upstream
control station, however, required the least amount of time to process, with
PIBS and STKM samples, respectively, requiring an increasing amount of time
for processing.
Differences in the time required for processing macroinvertebrate samples
were tested using the same methods outlined previously (Bartlett's ANOVA, and
SNK). No significant differences in sample processing time (p = 0.05)
existed between individual methods of macroinvertebrate collection at
stations upstream and downstream from Piceance Creek during fall or spring.
Differences did exist among stations (locations) near Piceance Creek during
both spring and fall sampling periods (Table 7). Differences in processing
time between methods were significant (p = 0.05) at the upstream control
station in spring 1978. All sampling methods tested at this station required
a significantly different amount of time (p = 0.05) to process as indicated
by SNK result.
18
-------
TABLE 6. MEANS (X), SAMPLE SIZES (N), AND COEFFICIENTS OF VARIATION
FOR PROCESSING TIME PER SAMPLE IN MINUTES. COST-EFFICIENCY
ESTIMATES ARE ALSO PRESENTED (C/T = The number of animals
processed per hour, n = the estimated number of samples required
for a given level of precision, and CE = n(X/60).)
Processing time Cost-Efficiency
in minutes Estimates
Sampling Methods, locations N X CV C/T n CE
10/4/77
Surber, downstream
30-Second Kick, downstream
Surber, upstream
30-Second Kick, upstream
All Surbers
All 30-Second Kicks
All samples, upstream
All samples, downstream
4/3/78
PIBS, downstream
30-Second Kick, downstream
PIBS, upstream
30-Second Kick, upstream
All PIBS samples
All 30-Second Kicks
All samples, upstream
All samples, downstream
4/4/78
Surber, control
PIBS, control
30-Second Kick, control
5
5
5
5
10
10
10
10
5
5
5
5
10
10
10
10
5
5
5
1,125
1,071
594
677
874
870
646
1,098
809
841
82
120
456
480
101
835
225
420
734
37.1
48.2
47.5
33.3
41.0
45.6
39.7
48.7
25.5
20.5
70.2
31.6
34.1
26.0
48.9
23.1
41.6
44.0
9.8
54
78
36
66
48
72
54
66
90
120
42
90
84
114
72
102
174
144
138
22
17
34
7
--
_ _
__
—
38
23
36
12
--
—
-_
--
22
7
5
417.6
299.7
332.5
76.8
--
__
--
--
513.9
324.0
49.5
24.3
--
--
-_
—
82.7
49.3
60.3
19
-------
ro
O
TABLE 7. ANALYSIS OF VARIANCE OF PROCESSING TIME PER SAMPLE IN MINUTES FOR THREE SAMPLING
DESIGNS (Two-way ANOVA's were used for the tests bracketing Piceance Creek. Single-
classification ANOVA and Student-Newman-Keuls stepwise multiple range test (SNK) were
used at the upstream control station. Significant (P = 0.05) differences between
group means are indicated by horizontal lines. N/A - not applicable.)
Degrees
Source of of
Variation Freedom
Sum of Squares
Upstream - Downstream Piceance Creek Area^
Totals
Location
Methods
Interaction
Error
19
1
1
1
16
3,330,963.75
1,023,781.25
101.25
17,111.25
2,289,970.00
Upstream - Downstream Piceance Creek Area^
Total s
Location
Methods
Interaction
Error
Upstream Control
Total
Method
Error
19
1
1
1
16
Station,
14
2
12
3,019,173.75
2,697,451.25
3,001.25
781.25
317,940.00
Spring, 1978
849,123.3333
659,503.3333
189,620.0000
Mean Square
Fall, 1977
1,023,781.25
101.25
17,111.25
143,123.13
Spring, 1978
2,697,451.25
3,001.25
781.25
19,781.25
329,751.6667
15,801.6667
F Method Mean
7.1532*
0.0007 N/A N/A
0.1196
135.7464***
0.1510 N/A N/A
0.0393
Surber 225
20.8682*** PIBS 420
STKM 734
SNK
N/A
N/A
—
-
—
* = Significant at 0.05 probability
*** = Significant at 0.001 probability
-------
COST EFFICIENCY OF SAMPLING METHODS
The cost efficiency of the various sampling methods was estimated in two
different ways. The first cost-efficiency estimate was calculated by
dividing the mean total counts per sample by the mean processing time in
minutes per sample and multiplying by 60, which gives an estimated mean
number of animals processed per hour (C/T). The second type of
cost-efficiency estimate was derived by first calculating an estimation of
the sample size required for a given level of precision by the methods of
Steel and Torrie (I960):
n = t2 CV2.
where n = estimated number of samples required,
t = Student's t value for a given probability level and degrees of
freedom based on the number of replicates (p = 0.05 in the present
study),
CV = coefficient of variation and
L = acceptable percent error of the sample mean from the population
mean (25 percent in the present study).
The sample-size estimate is then multiplied by the mean number of hours
required to process a sample. This cost-efficiency estimate represents the
number of hours required to process sufficient samples to provide a total
count estimate that would be within 25 percent of the population mean for any
sampling method.
Cost-efficiency estimates do not consider the time required for sample
collection since the actual sampling time represented only a minor portion of
the overall effort in terms of total time expended per sample. However,
sampling time is substantially reduced with the STKM since it is possible to
collect approximately three STKM samples in the time required to collect a
single Surber sample (Kinney, Pollard, and Hornig 1978). In addition, the
problems associated with winter sampling are substantially reduced using the
STKM.
In the majority of cases, both estimates of cost efficiency indicated
that the STKM had the best cost efficiency (Table 6). For example, the STKM
required fewer hours to process samples that would be within 25 percent of
the population mean as well as producing a higher number of animals per unit
processing time than the other two methods (Table 6). In addition, the
estimated number of samples required to yield total counts (standing crop)
estimates within 25 percent of the population mean was consistently lower for
STKM samples than for any other method. In fact, considerably more Surber
samples (22) would be required than either PIBS (7) or STKM samples (5) to
achieve the same level of precision in the upper White River. The single
exception to this pattern was represented at the upstream control station
21
-------
during spring 1978. In this case, Surber samples contained a higher number
of animals per unit processing time than the other two methods tested. The
precision of Surber samples, however, was not as good as the other two types
of samples as indicated by the higher C.E. index.
22
-------
DISCUSSION
The present data clearly demonstrated the ineffectiveness of the Surber
sampler and the PIBS in riffles with filamentous algal growths (e.g.
Cladophora sp.). This is undoubtedly a function of the small area covered
for each sample and the inherent patchiness of the stream benthos. On the
other hand, Surber samplers and the PIBS performed reasonably well at the
upstream control station, indicating that in some riffle habitats a
small-area sample would provide a sufficient number of animals and taxa per
sample to provide realistic descriptions of macrobenthic community
composition. The present study clearly demonstrated, however, that the STKM
was superior to or equally effective as the PIBS or Surber sampler for
purposes of biological monitoring. Although the PIBS and Surber samplers
should have provided very similar estimates of species compositions and
density, these data clearly demonstrated that this was not the case. This
was particularly interesting since the method of sampler employment was very
similar and both samplers collected animals from approximately the same
bottom area.
It is obvious from these data that different methods of macroinvertebrate
collection yield different estimates of benthic community composition. In
addition, the PIBS, when compared to the Surber sampler, collected larger
numbers of organisms and taxa per sample although the differences in total
counts may have been a function of the high variability of these samplers.
Sampler selectivity has been reported by other investigators using other
methods of macroinvertebrate collection (Albrecht 1961; Hynes 1961).
Precise estimates of benthic standing crop are not obtained using the Surber
sampler as has been vividly pointed out by Kroeger (1972). It is evident
that none of the sampling methods presently available for macroinvertebrate
collection from rocky-bottom stream substrates provide an unbiased estimate
of benthic standing crop or community composition.
Choosing a methodology that yields the lowest possible replicate
variability and the highest capture efficiency per sampling effort is much
more important to sampling program design than the "quantitative" nature of
the sampling method since no method presently available is unbiased. Although
the STKM would not be considered strictly "quantitative" by most benthic
biologists, samples collected with the STKM can provide a reliable index to
standing crop because of its low replicate variability. For purposes of
determining community changes over time, relatively low variability is of
paramount importance. If estimates of standing crop per unit area are
necessary, an enclosed-design small-area sampler would, in relatively
productive areas, provide adequate data, although a large number of samples
23
-------
may be required to adequately determine absolute values for standing crop.
It is unlikely, however, that a small-area sampler would provide useful data
in the patchy, fauna!-poor reaches of western streams.
All major taxa collected by comparable PIBS or Surber samplers were also
well represented in STKM samples. In most cases the efficiency of the STKM
in terms of the number of animals processed per unit time was also higher
than comparable Surber or PIBS samples. For these reasons, the STKM is
considered the best method for collecting stream macroinvertebrates from
rocky bottom substrates (rubble, coarse gravel, etc.), and is specifically
recommended for use in the White River in proximity to oil shale development
areas.
It should be noted at this time that the data presented in this study
represent an ideal data base from which to develop biological monitoring
designs for the White River in the vicinity of Piceance Creek and in the
upstream reaches of the White River between Buford and Meeker Dome. It is
evident from the estimated number of samples required to obtain standing crop
values within 25 percent of the population mean (Table 7) that precise
estimates of standing crop would require an unrealistic amount of sampling
effort. It was possible, however, to detect differences in standing crop per
sample and total number of taxa per sample between methods and stations when
five replicate samples were collected. In addition, five STKM samples
generally collected 1,000 or more animals, allowing reliable species
percentage data to be generated. Comparisons of total counts, total number
of taxa, and percent species composition of samples collected at stations
bracketing the source of potential water pollution from oil shale development
should provide the information necessary to detect any associated changes in
macrobenthic communities.
Owing to the high natural variability in the macrobenthos of this reach
of the White River, adequate baseline data must be compiled to allow
separation of man-induced changes and natural variability. Since vast
differences in communities were apparent during different seasons, it is
evident that single-season monitoring in rivers similar to the White River is
not sufficient to provide adequate characterization of benthic communities.
24
-------
REFERENCES
Albrecht, M. L. 1961. Ein Vergleich quantitativer Methoden zur Untersuchung
der Makrofauna fliessender Gewasser. Verh. int. Verein. Theor. agnew.
Limnol. 14: 486-490.
Chutter, R. M. 1972. A reappraisal of Needham and Usinger's data on the
variability of a stream fauna when sampled with a Surber sampler.
Limnol. Oceanogr. 17(1): 139-141.
Ellis-Rutter Associates. 1973. Advertising publication. P. 0. Box 394,
Douglassville, Pa.
Everhart, W. H. and B.E. May 1973. Effects of chemical variations in aquatic
environments. U.S. Environmental Protection Agency, Office of Research
and Monitoring, Washington, D.C. 117 pp.
Fox, R. L. 1977. Report of baseline water quality investigations on the
White River in western Colorado September-October, 1975 and May-June
1976. EPA-908/2-77-001. S. and A. Division, U.S. Environmental
Protection Agency, Region VIII. 48 pp. + appendices.
Frost, S., A. Huni, and W. E. Kershaw. 1971. Evaluation of a kicking
technique for sampling stream bottom fauna. Can. J. Zool. 49:
167-173.
Hilsenoff, W. L. 1977. Use of Arthropods to evaluate water quality of
streams. Technical Bulletin Mo. 100, Department of Natural Resources,
Madison, Wi. 15 pp.
Hornig, C. E., and J. E. Pollard. 1978. Macroinvertebrate sampling
techniques applicable to streams of semi-arid regions. Environmental
Monitoring Series. EPA-600/4-78-040. U.S. Environmental Protection
Agency, Las Vegas, Nevada. 21 pp.
Hynes, H. B. N. 1961. The invertebrate fauna of a Welsh mountain stream.
Arch. Hydrobiol. 57: 344-388.
Hynes, H. B. N. 1970. The ecology of running waters. University of
Toronto Press, Toronto. 555 pp.
Hynes, H. B. N. 1971. Benthos of flowing waters. In: A manual on
methods for the assessment of secondary productivity in fresh water.
W. T. Edmonson and G. G. Winberg (Eds.). F. A. Davis Company,
Philadelphia, Pa. pp. 66-75.
25
-------
Kinney, W. K., Pollard, J. E., and C. E. Hornig. 1978. Comparison of
macroinvertebrate samplers as they apply to streams of semi-arid
regions. In: Conference proceedings of the 4th joint conference on
sensing of environmental pollutants, New Orleans Hilton, New Orleans
La., Nov. 6-11, 1977. American Chem. Soc. Washington, D.C. pp. 515-518.
Kroeger, L. 1972. Underestimation of standing crop by the Surber sampler.
Limnol. Oceanogr. 17 (3): 475-479.
Needham, J. G., and P. R. Needham. 1962. A guide to the study of
freshwater biology. Holden-Day, Inc., San Francisco. 108 pp.
Pennak, R. W. 1974. Limnological status of streams, summer 1973 -
Piceance Basin, Rio Blanco and Garfield Counties, Colorado. Regional
Oil Shale Study, State of Colorado. Thome Ecological Institute,
Boulder. 50 pp.
Sokal, R. F., and F. J. Rohlf. 1969. Biometry. W. H. Freeman and Co.,
San Francisco. 776 pp.
Steele, R. G. 0., and J. H. Torrie. 1960. Principles and procedures of
statistics with special reference to the biological sciences. McGraw-
Hill, New York. 481 pp.
U.S. Department of the Interior. 1973. Final environmental statement for
the prototype oil shale leasing program (six volumes), Washington, D.C.,
Regions VIII and IX. Vol. 1, Regional impacts of oil shale development.
U.S. Environmental Protection Agency. 1973. Biological field and laboratory
methods for measuring the quality of surface waters and effluents.
Environmental Monitoring Series. EPA-670/4-73-001. U.S. Environmental
Protection Agency. Cincinnati, Ohio. 176 pp.
Westman, W. E. 1977. Problems in implementing U.S. water quality goals.
Amer. Sci. 65: 197-203.
26
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
REPORT NO.
EPA-600/7-79-163
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
ASSESSMENT OF MACRO INVERTEBRATE MONITORING TECHNIQUES
IN AN ENERGY DEVELOPMENT AREA: A test of the efficiency
of three macrobenthic sampling methods in the White Rive
5. REPORT DATE
July 1979
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
J.E. Pollard* and W.L. Kinney
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Environmental Monitoring and Support Laboratory
U.S. Environmental Protection Agency, and
Biology Department, University of Nevada, Las Vegas
10. PROGRAM ELEMENT NO.
1NE625ABZ
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
U.S. Environmental Protection Agency-Las Vegas, NV
Office of Research and Development
Environmental Monitoring and Support Laboratory
Las Vegas, Nevada 89114
13. TYPE OF REPORT AND PERIOD COVERED
Final 1977-1978
14. SPONSORING AGENCY CODE
EPA/600/07
15. SUPPLEMENTARY NOTES
*Biology Department, University of Nevada, Las Vegas,
Las Vegas, Nevada 89154
intiLtiuds uf
16.ABSTRACT inrcB iiininuus UT nidcrulMverUibr'dte cutlectlon were evaluated fur
reproducibility, capture-effectiveness, and cost efficiency in the White River near
Meeker, Colorado. Samples were collected with a standard Surber sampler, with a
portable invertebrate box sampler (PIBS), and using the standardized traveling kick
method (STKM). Methods were evaluated in riffles of the White River directly upstream
and downstream from the confluence of Piceance Creek, as well as at a comparable riffle
at an upstream control station. The traveling kick method collected the largest number
of animals and taxa per sample with equivalent or lower variability than the other two
methods tested. While Surber samplers and the PIRS performed similarly in the vicinity
of Piceance Creek, their performance differed at the upstream control station where the
PIBS collected more animals and taxa per sample than the Surber sampler. Similarly,
while sample processing time did not significantly differ for the various methods of
collection used at any station in the vicinity of Piceance Creek, differences did exist
at the upstream control station with kick samples requiring the greatest amount of
processing time and Surbers requiring the least. The cost efficiency of various
methods was estimated by calculating the number of animals processed per unit time and
the number of hours required to provide standing crop estimates for each sample within
a given precision level. Cost-efficiency estimates indicated that the STKM was
superior to the Surber or PIBS methods, particularly at the station near Piceance
Creek.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b-IDENTIFIERS/OPEN ENDED TERMS
COSATI Field/Group
Benthos
Limnology
Sampling
Water Pollution
Environmental Monitoring
Aquatic Biology
Oil Shale Industry
Semi arid regions
Macroinvertebrates
Standardized traveling-
kick method
Portable invertebrate-
box sampler
Surber Sampler
f.nlnradn
08H
14A.D
19. SECURITY CLASS (ThisReport)
UNCLASSIFIED
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
21. NO. OF PAGES
36
20.
ASS (This page)
8/m
22. PRICE
A03
EPA Form 2220-1 (R«v. 4—77) PREVIOUS EDITION is OBSOLETE
»U.S. GOVERNMENT PRINTING OFFICE: 1979-683 091
------- |