United States
Environmental Protection
Agency
Office of Acid Deposition,
Environmental Monitoring and
Quality Assurance
Washington DC 20460
EPA 600/8-87 019
April 1987
National
Surface Water
Survey:
National Stream Survey
Phase I - Pilot Survey
Field Operations Report
-------
EPA 600/8-87/019
April 1987
National Surface Water Survey:
National Stream Survey
Phase I -Pilot Survey
Field Operations Report
by
C.M. Knapp, C.L. Mayer, D.V. Peck,
J.R. Baker and G.J. Filbin
A Contribution to the
National Acid Precipitation Assessment Program
U.S. Environmental Protection Agency
Office of Research and Development
Washington. DC 20460
Environmental Research Laboratory — Corvallls. OR 97333
Environmental Monitoring Systems Laboratory - Las Vegas. NV 89119
-------
Notice
The information in this document has been funded by the U.S. Environmental Protection
Agency under contract no. 68-03-3249 to Lockheed Engineering and Management Services Company,
Inc. It has been subject to the Agency's peer and administrative review, and it has been approved
for publication as an EPA document.
The mention of corporate names, trade names, or commercial products in this report is for
illustration purposes only and does not constitute endorsement or recommendation for use.
This document is one volume of a set which fully describes the National Stream Survey - Pilot
and Phase I. The complete document set includes the major data report, quality assurance plan,
analytical methods manual, field operations report, processing laboratory operations report, and
quality assurance report. Similar sets are being produced for each Aquatic Effects Research
Program component project. Colored covers, artwork, and the use of the project name in the
document title serve to identify each companion document set.
The correct citation of this document is:
Knapp, C. M., C. L Mayer, D. V. Peck, J. R. Baker, and G. J. Filbin. 1987. National Stream Survey,
Phase I-Pilot Survey: Field Operations Report. EPA-600/8-87/019. U.S. Environmental
Protection Agency, Las Vegas, Nevada.
-------
Abstract
The National Stream Survey (NSS) is one of the programs within the National Surface Water
Survey of the U.S. Environmental Protection Agency. The proposed research plan for Phase I of
the NSS was evaluated during a pilot survey conducted in the spring and summer of 1985. A base
of operations that included a mobile laboratory was established at Sylva, NC. Selected locations
of 61 streams in the southern Blue Ridge region of the United States were sampled four times
during a 57-day period. This report chronicles the activities required to plan and conduct the field
operations of the NSS pilot survey.
Preparatory activities for the NSS pilot survey are described, including the personnel training
program and site reconnaissance activities. The equipment and protocols (including quality
assurance measures) used to collect water samples and field measurements of pH, conductivity,
and dissolved oxygen are presented. Field laboratory activities are summarized, including a protocol
for preparing a fraction for analysis of organically-complexed monomeric aluminum species. The
fractionation procedure used was feasible, but alternative methodologies should be investigated.
Certain protocols for collecting samples or for conducting field measurements were compared
against possible alternatives. Filtering samples during collection was determined to be unfeasible.
The use of a peristaltic pump to collect samples was found to be more suitable than collecting
discrete grab samples. Measurements of pH could be conducted at streamside without concern
for effects of CO2 degassing. Experiments investigating the holding time of unpreserved water
samples are presented in two appendices to this report. The results of these experiments indicate
that sample holding times could be extended without compromising the accuracy or quality of the
data.
The NSS pilot survey was completed on schedule and demonstrated that a large-scale
synoptic survey of streams was logistically feasible. The NSS also confirmed that the basic
research design, quality assurance plan, and data analysis plan of the NSS would provide the
necessary information to meet the objectives of the NSS. Pertinent cost information and specific
recommendations regarding various aspects of field operations are provided for those planning
similar projects.
This report is submitted in partial fulfillment of contract 68-03-3249 by Lockheed Engineering
and Management Services Company, Inc., under the sponsorship of the U.S. Environmental
Protection Agency. This report covers a period from January 1985 to July 1985 and work was
completed as of September 1986.
in
-------
Contents
Page
Notice ii
Abstract iii
Figures vii
Tables viii
Acknowledgements x
1. Introduction 1
2. Planning and Preparation for Field Operations '. 3
Site Selection 3
Protocol Development 3
Procurement 6
Field Station Site Selection 6
Personnel Recruitment and Training 7
Field Sampler Training Program 7
Field Laboratory Personnel Training Program 7
Field Laboratory Modifications 8
Site Reconnaissance Activities 8
3. Field Station Operations 11
Field Station Organization 11
Field Station Communications 12
4. Field Sampling Operations 14
Sampling Equipment 14
Quality Assurance and Quality Control of Field Measurements 14
Daily Sampling Activities 15
Preparation for Sampling 15
Sampling and Field Measurements 15
pH Measurement 16
In Situ Conductivity Measurement 17
In Situ Dissolved Oxygen Measurement 17
Other Measurements 18
Post-sampling Activities 18
-------
Contents (continued)
5. Field Laboratory Operations 20
Field Laboratory Personnel 20
Daily Field Laboratory Activities 20
Audit Samples 20
Preparation of the Field Laboratory 22
Receipt of Samples and Data Forms from Field Crews 22
Organization of Samples into a Batch 23
Transfer of Samples to Field Laboratory 23
Sample Analysis and Processing 23
Sample and Data Form Shipment 26
6. Evaluation of Equipment and Protocols 28
Field Laboratory Methods Evaluations 28
Field Laboratory Conductivity Measurements . 28
Preparation of Organic Monomeric Aluminum Fraction 28
Field Equipment Evaluations 29
Field Methods Evaluations 29
In-line Filtration at Streamside 29
Sample Collection Method 29
Streamside pH Measurements 30
In Situ pH Measurement Method 30
Sample Holding Time 30
7. Summary of Field Operations 31
8. Observations and Recommendations 34
Field Safety 34
Planning Activities 34
Field Sampling Operations 34
Field Laboratory Operations . 38
References 41
Appendices
A Syringe Sample Holding Time Study 42
B. Cubitainer Holding Time Study 52
VI
-------
Figures
Number Page
1. Regions of the eastern United States selected for sampling
during the National Stream Survey-Pilot Survey 2
2. Map of National Stream Survey-Pilot Survey study area
showing location of stream reaches selected for sampling 5
3. Watershed characteristics form, National Stream Survey-Pilot Survey 10
4. Flowchart of daily field station operations, National Stream
Survey-Pilot Survey 12
5. Daily field sampling activities, National Stream Survey-Pilot Survey 16
6. Field data form, National Stream Survey-Pilot Survey 19
7. Flowchart of daily field laboratory operations,
National Stream Survey-Pilot Survey 21
8. Field laboratory data form, National Stream Survey-Pilot Survey 24
9. Flowchart of field sample processing and analyses conducted
at field laboratory during National Stream Survey-Pilot Survey 25
10. Shipping form, National Stream Survey-Pilot Survey 27
A-1. Changes in pH and dissolved inorganic carbon over time
in a sample initially having no dissolved carbon dioxide 46
A-2. Changes in DIC and pH over a 12-hour period in a sample
initially supersaturated with carbon dioxide 47
A-3. Changes in DIC and pH over a 165-hour period in a sample
initially supersaturated with carbon dioxide 47
B-1. Mean values for pairs of measurements at 4 holding times
for 10 lakes and streams 55
VII
-------
Tables
Number Page
1. Streams Sampled During the National Stream'Survey-Pilot Survey 4
2. Summary of Field Sampler Training Program, National Stream
Survey-Pilot Survey 7
3. Summary of Field, Laboratory Training Program. National Stream
Survey-Pilot Survey 8
4. Sampling Equipment Used During National Stream Survey-Pilot Survey 15
5. Sampling Activity Time Summary for Summer Sampling Period,
National Stream. Survey-Pilot Survey 31
6. Summary of Distance Traveled by Field Sampling Teams,
National Stream Survey-Pilot Survey 32
7. Selected Cost Estimates, National Stream Survey-Pilot Survey 33
8. National Stream Survey-Pilot Survey Problems and
Recommendations: Field Safety 35
9. National Stream Survey-Pilot Survey Problems and
Recommendations: Site Reconnaissance and Access 35
10. National Stream Survey-Pilot Survey Problems and
Recommendations: Field Sampling Activities and Equipment 36
11. National Stream Survey-Pilot Survey Problems and
Recommendations: Field Laboratory Operations 39
A-1. Dissolved Inorganic Carbon and pH Measurements of a
Sample Initially Having No Dissolved CO2 46
A-2. Dissolved Inorganic Carbon and pH Measurements Over Time
on a Sample Initially Supersaturated with Dissolved Carbon Dioxide 46
A-3. Summary Statistics of pH and DIC Changes Over Time for
Solutions of Different Levels of Dissolved Carbon Dioxide
Stored at Different Temperatures 48
A-4. Summary of F-Tests Performed on Variances of Sample and QCCS
Measurements for Solutions of Different Initial Dissolved
Carbon Dioxide Concentrations . 49
VIII
-------
Tables (continued)
A-5. Comparisons of Variances for pH Measurements of Solutions
Having Different Dissolved CO2 Concentrations Versus a Low
Ionic Strength pH 7.00 Quality Control Check Sample 50
A-6. Summary Statistics for Each Time Interval From Samples
Collected From Padden and Bagley Lakes 51
B-1. Locations of Lakes and Streams 53
B-2. Treatment Means and Overall Standard Error of 25 Chemical
Parameters at Four Holding Times for Three Lakes and Seven Streams 62
B-3 Relative Difference Between the Means for Each Parameter for
Holding Time Intervals of 12 to 24, 24 to 48, 48 to 84,
and 12 to 84 hours 63
IX
-------
Acknowledgements
S. L Pierett was responsible for the overall administration of Lockheed-EMSCO logistical
operations. K. Asbury (Lockheed-EMSCO) coordinated the procurement of equipment and supplies.
K. J. Cabbie (Lockheed-EMSCO)' selected the site for the field station and provided technical
assistance during the early days of the field operation. J. Tuschall (Northrop Services, Inc.,
Research Triangle Park, NC) assisted with establishing contact with local cooperators during site
reconnaissance activities. The assistance and support of N. Myers, S. Monroe, and T. Ashe
(Southwest Technical College, Sylva, NC) are gratefully acknowledged.
Comments on earlier drafts of this report from the following individuals are gratefully
acknowledged: K N. Eshleman and P. Kaufmann (Northrop Services, Inc., Corvallis, OR); V. Sheppe,
S. Drouse, and D. Hillman (Lockheed-EMSCO); and R. D. Schonbrod, (U.S. EPA, Las Vegas, NV).
Special thanks are directed to P. Kellar (Radian Corporation, Research Triangle Park, NC), whose
thorough review improved the quality of the manuscript. M. Faber (Lockheed-EMSCO) served as
technical editor. Formal reviewers were P. Gehring (U.S. EPA, Westlake, OH), J. H. Gibson
(Colorado State University), and T. Janecki (Martin Marietta Corporation, Columbia, MD).
F. Garner (Lockheed-EMSCO) provided statistical advice for the syringe holding time
experiments. For Appendix B (holding time study of bulk water samples), E. A. Yfantis, M. J. Miah,
F. C. Garner, and T. J. Permutt assisted with statistical analyses. J. Scanlan helped with computer
programming.
L. Steele and S. Reppke (Computer Sciences Corp.) are gratefully acknowledged for their
excellent support typing the various drafts of this report. B. Sheets, L. Gurzinski, R. Buell, S. Garcia
and D. Hamby (Lockheed-EMSCO) prepared most of the figures presented in this report. Finally,
acknowledgement is due to W. L. Kinney (U.S. EPA, Las Vegas, NV) who served as project officer
for this survey.
-------
Section 1
Introduction
The apparent acidification of surface
waters in areas of the United States has
recently been the subject of intensive debate.
The influence of acidic deposition on surface
water acidification has been the subject of
numerous studies (e.g., see review by
Altshuller and Linthurst, 1984). Attempts to
extrapolate the results from localized studies
to a regional or national scale have not been
successful. Comparisons between localized
studies are compromised by (1) differences in
methodologies, (2) biased selection of study
sites, (3) small or incomplete data bases, or
both, or (4) lack of adequate quality assurance
information.
In an effort to overcome these problems,
the U.S. Environmental Protection Agency (EPA)
has designed and implemented the National
Surface Water Survey (NSWS) as part of the
monitoring and assessment effort of the
National Acid Precipitation Assessment Pro-
gram. The NSWS is a three-phase program to
provide regional-scale assessments of the
present chemical and biological status of lakes
and streams and to select representative sites
for long-term monitoring of aquatic resources.
The NSWS has two major components,
the National Lake Survey (NLS) and the
National Stream Survey (NSS). The NSS will
investigate lotic systems in regions of the
eastern United States considered to be at risk
from acidic deposition. Phase I of the NSS will
be a synoptic survey in some of these regions
to assess the present chemical status of
streams.
In preparation for Phase I of the NSS,
the U.S. EPA and cooperating scientists con-
ducted a pilot survey during the spring and
summer of 1985. The National Stream Survey-
Pilot Survey (NSS-PS) was conducted in the
southern Blue Ridge area of Tennessee, North
Carolina, South Carolina, and Georgia (Figure
1). The study had the following objectives:
1. Evaluate the proposed NSS-Phase I
protocols and equipment that will be
used to collect, process, and analyze
stream water samples.
2. Evaluate the proposed NSS-Phase I
logistics plan, including site access,
safety, and field operations.
3. Evaluate the proposed NSS-Phase I
data quality objectives and quality
assurance plan.
4. Provide suitable data to evaluate the
proposed NSS-Phase I sampling
design and data analysis plan.
5. Train personnel who will be involved
in NSS-Phase I activities.
This report documents and evaluates the
pilot survey logistical planning activities, field
sampling methods, and field laboratory opera-
tions. Reports on two special experiments
related to sample holding times that were
conducted during the pilot survey are appen-
ded to this report. Analytical methods used in
the pilot survey are described by Hillman et al.
(1987). The quality assurance plan is docu-
mented in Drous6 et al. (1986). The research
design, analytical results, data interpretation,
and study conclusions are presented in Messer
et al. (1986).
-------
42° N
38° N
34° N
30° N
26° N
97° W
93° W
89° W 85° W 81° W
77° W
73° W
Figure 1. Regions of the eastern United States selected for sampling during the National Stream Survey-Pilot
Survey (from Messer et al., 1986).
-------
Section 2
Planning and Preparation for Field Operations
The Environmental Monitoring Systems
Laboratory of the EPA in Las Vegas, Nevada,
was responsible for planning and conducting
the pilot survey field operation. Lockheed
Engineering and Management Services Com-
pany, Inc. (Lockheed-EMSCO) provided logis-
tics support and field personnel. Other groups
and agencies also participated in planning and
implementing the pilot survey. These groups
are described further by Messer et al. (1986).
The major planning activities were the
selection and reconnaissance of sampling
sites, procurement of equipment, development
of sampling and analytical protocols, selection
of a location for base of operations, and the
training of field personnel. These activities are
described more fully in this section.
Site Selection
The research design for the pilot survey
used a stream "reach" as the basic statistical
sampling unit (Messer et al., 1986). A reach
was operationally defined as the length of a
stream that lies between corresponding up-
stream and downstream points of confluence
(or "nodes") with other streams, or to the
headwater when no upstream node was
present. For the pilot survey, 54 such reaches
were selected by using procedures described
by Messer et al. (1986). These reaches,
termed "second stage" (Messer et al., 1986) or
"regular" reaches, were selected as a statisti-
cally unbiased sample without consideration of
accessibility or availability of historical water
quality data. In addition to the regular
reaches, seven additional reaches were in-
cluded for sampling. These additional reaches,
termed "special interest" reaches, were study
sites for other monitoring or research pro-
grams; thus considerable data on water qual-
ity were available. All stream reaches sampled
during the pilot survey are identified, along
with their NSS site identification code, in Table
1. The location of each of these reaches is
shown in Figure 2.
Protocol Development
Field protocols for collecting water sam-
ples from streams and for measuring chemical
parameters both in situ and at streamside
were developed in early January 1985. These
protocols were based on the proposed re-
search plan and on discussions with other
researchers. Protocols were adapted from a
recommended methodology to meet the re-
quirements of the NSS. Methods were modi-
fied for specific equipment or instrumentation,
and appropriate quality assurance and quality
control procedures were incorporated into each
protocol. Procedures to collect water from
midchannel by using a portable peristaltic
pump, to measure pH at streamside, and to
determine dissolved oxygen content and con-
ductivity in situ were developed and were
documented in a draft field operations manual
that was used during the pilot survey.
Additional field protocols were developed
to determine stream stage height and cross-
sectional profile and to classify riparian char-
acteristics in the vicinity of the sample site.
These methods were not documented in man-
uals but were incorporated into the sampling
personnel training program. Field protocols
were modified or were otherwise evaluated
during the pilot survey. These tests are sum-
marized in Section 6.
The physical and chemical parameters
measured in water samples collected during
the pilot survey were the same as those
-------
Table 1. Stream* Sampled During the National Stream Survey-Pilot Survey
NSS-PS*
ID
2A07701
2A07702
2AQ7703
2A07801
2A07802
2A07803
2A07805
2A07806
2A07807
2A07808
2A07810
2A07811
2A07812
2A07813
2A07814
2A07815
2A07816
2A07817
2A07818
2A07819
2A07820
2A07821
2A07822
2A07823
2A07824
2A07825
2A07826
2A07827
2A07828
2A07829
2A07830
2A07831
2A07832
2A07833
2A07834
2A07835
2A07881
2A07882
2A07891C
2A07892C
2A07893C
2A07894C
2A07895C
2A07896C
2A08801
2A08802
2A08803
2A08804
2A08805
2A08806
2A08808
2A08809
2AQ8810
2A08811
2A08891C
2A08901
2A09902
2A09903
2A09904
2A09905
2A09906
Stream Name
Sugar Cove Branch of North River
Childers Creek
Halls Creek
Gulf Fork of Big Creek
Puncheon Fork
N.N.T. to Ellejoy*
Cosby Creek
Roaring Fork of Meadow Fork
North Fork of Dillingham Creek
Armstrong Creek
Little River
False Gap Prong
Correll Branch
Little Sandymush Creek
Reems Creek
Curtis Creek
Eagle Creek
Forney Creek
Bunches Creek
Crooked Creek
East Fork Pigeon River
Grassy Creek
Sweetwater Creek
Brush Creek
Middle Prong of Wast Fork
South Fork of Mills River
Henderson Creek
Welch Mill Creek
Whiteoak Creek
Cathe/s Creek
Mud Creek
North Pocolet River
Tusquitee Creek
Allison Creek
Brush Creek
Middle Saluda River
Walnut Creek
Little Branch Creek
Cosby Creek
Twentymile Creek
Jarrett Creek
Shope Fork
Moses Creek
Pinnical Branch
N.N.T.
Dunn Mill Creek
Owenby Creek
Bear Creek
Weaver Creek
N.N.T. to Kintuestia Creek
Whitepath Creek
Tickanetley Creek
Bryant Creek
Hinton Creek
Chester Creek
Little Persimmon Creek
West Fork of Chattooga River
Nottely River
She Creek
Chattahochee Creek
Deep Creek
State
TN
TN
TN
TN
NC
TN
TN
NC
NC
NC
TN
TN
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
SC
NC
NC
TN
NC
NC
NC
NC
NC
GA
GA
GA
GA
GA
GA
GA
GA
GA
GA
GA
GA
GA
GA
GA
GA
GA
County
Monroe
Polk
Polk
Cocke
Madison
Blount
Cocke
Madison
Bancombe
McDowell
Sevier
Sevier
Haywood
Madison
Buncombe
McDowell
Swain
Swain
Swain
McDowell
Haywood
Henderson
Graham
Swain
Haywood
Transylvania
Henderson
Cherokee
Macon
Transylvania
Henderson
Polk
Clay
Macon
Macon
Greenville
Madison
Madison
Cocke
Swain
Macon
Macon
Jackson
Macon
Murray
Fannin
Fannin
Gilmer
Fannin
Union
Gilmer
Gilmer
Lumpkin
Pickens
Fannin
Rabun
Rabun
Union
Rabun
White
Hebersham
Downstream Site
Latitude Longitude
35*19'21°
35'11'30"
35*05'45"
35'53'45"
35'54'30"
35*47'00°
35'48'00"
35*48'30"
35'46'30"
35'48'30"
35*40'15"
35*41'59"
35*40'15"
35*42'10"
35'42'06"
35*38'30"
35*29'54"
35*31'40"
35*33'38"
35*36'30"
35*27'10"
35'27'53"
35*19'39"
35'19'Or
35*22'21"
35*20'45"
35'22'43"
35*11'08"
35'13'56"
35*12'42"
35*15'17"
SS'ttW
35'04'13"
35*07'12"
35*06'50"
35*07'46"
35'48'Sr1
35*27'04"
35*44'58"
35*28'ir
35*09'07"
35*03'48"
35*19'30"
35*03'28"
34*57'42"
34*56'36"
34*58'15"
34*49'28"
34'52'20"
34'53'28"
34°44'15"
34*37'50"
34*36'36"
34*29'09"
34*38'26"
34*54'18"
34*54'21"
34*49'26"
34*50'06"
34*42'32"
34'40'3r
084'06'OT1
084*2913al
064'21'8a'
083*05'15"
082'34'45"
083'48W
083*14'15"
082*53'45"
oaz'zrxr
o&'ozzr
083*40'3(r
083*23'24"
oss'osw
082*45'3811
082'35'12"
082*09'30'
083*45'50'
083'33'aOf'
083*14'48"
082'06'5a'
082*50'5a'
082*17'05"
Q83'4ffQ2'
083*30'5»'
082'56'18"
082*39>50I
082"23'02'
083*53'3r'
oes'sroff1
082'47'09"
OOZ'SOfA"
082*13'15"
083'48'59"
083*29'34"
083*15'28"
082*32'ir
082'44'15"
083'03'39'1
083^20$'
083*52'16"
083'36'29'
oss'aew
083'06'21"
oas'arse11
084'44'12'
084'2ff28"
084'08'45"
084*33'5T'
084*17'56"
084'01'24"
084*26W
084*16'50"
083'59'55"
084*25'16"
084'10'05"
083'30W
083'11'3Z'
083*54'3a>
083'20'39"
083*44'3a'
083*27'19"
National Stream Survey-Pilot Survey site identification code.
No-name tributary.
Special interest reach.
-------
SPECIAL INTEREST SITES
85° W
84° W
Figure 2. Map of National Stream Survey-Pilot Survey study area showing locat'on of stream reaches selected
for sampling (from Messer et al., 1986). (Numbers are the last four digits of the site Identification code
from Table 1).
-------
measured during Phase I of the Eastern Lake
Survey (Linthurst et al., 1986). Analytical
methodologies for these parameters also
remained the same. One additional parameter,
organic monomeric aluminum, was measured
during the pilot survey. A measurement of
conductivity at the field laboratory was also
implemented late into the pilot survey.
The determination of organic monomeric
aluminum was added to provide a more accu-
rate estimate of toxic forms of dissolved
aluminum. Total dissolved aluminum concen-
trations are not always indicative of the poten-
tial toxicity to fishes. Only certain species,
particularly monomeric forms, are believed to
be toxic (Driscoll et al., 1980; Baker and
Schofield, 1982). In the NSWS, water samples
are analyzed for their total extractable alumi-
num content. This extractable aluminum
fraction contains dissolved monomeric species
of aluminum and thus provides a crude esti-
mate of the potential toxicity. The toxicity of
monomeric aluminum species can be mitigated
by the presence of organic ligands (e.g., dis-
solved organic acids). Organic monomeric
aluminum represents a nontoxic component of
the extractable aluminum fraction. The con-
centration of toxic monomeric aluminum spe-
cies could then be estimated as the difference
between total extractable and organic mono-
meric aluminum concentrations in a water
sample.
There are several different methods for
determining the various forms of dissolved
aluminum in natural waters. The procedure for
preparing an organic monomeric aluminum
fraction developed by Driscoll (1984) was
selected for evaluation during the pilot survey.
This methodology did not require expensive or
complex equipment. It could be carried out in
a small space and could be done in a short
time. Both of these attributes made this
methodology suitable for use in a field labora-
tory. The organic monomeric aluminum frac-
tion was prepared by using the same extrac-
tion process as was used for total extractable
aluminum, and this allowed for a more direct
estimation of inorganic monomeric aluminum
concentrations.
The measurement of conductivity of
samples at the field laboratory would provide
a precise estimate of conductivity within 16
hours of collection. This measurement would
serve as an additional check on overall data
quality and sample stability. A protocol for
this measurement was developed and eval-
uated during the pilot survey.
Procurement
The equipment and other supplies re-
quired to perform the required field and field
laboratory protocols were procured between
December 1984 and February 1985 by following
U.S. government-approved purchasing proce-
dures. Experience had indicated that demands
for certain pieces of equipment or supplies
(e.g., membrane filters) could deplete stock-
piles on a nationwide basis. The field labora-
tory trailer and much of its instrumentation
and equipment were already available because
they had been used during Phase I of the
Eastern Lake Survey, and only consumable
supplies and spare parts had to be obtained.
The trailer is described in Morris et al. (1986),
and the specific equipment and supplies used
in the field laboratory are outlined in Hillman et
al. (1986, 1987 in press).
Field Station Site Selection
The site selected for the pilot survey field
station had to meet numerous requirements.
Primary concerns were a reliable supply of
good quality water delivered at moderate
pressure (50 psi), an adequate calibration
room facility, and on-site storage facilities for
supplies and backup equipment. The' location
had to be served by a courier service that
offered morning and afternoon service for
delivery and pickup and that could deliver
samples overnight to other locations. Efficient
field operations also dictated that the field
station and field laboratory be centrally located
within the study area.
Advance travel throughout the study area
was necessary to select a site for the field
station which would accommodate needs for
housing, a field operations facility, and a
communications center. Each potential field
station was evaluated by utilizing a site
assessment sheet developed for this purpose.
After inspection of four proposed sites,
Southwest Technical College in Sylva, NC, was
selected as the field station for the pilot
6
-------
survey. The field laboratory, calibration room,
and equipment storage facility were located
there. Housing and the field station communi-
cations center were located in Cullowhee, NC,
approximately five miles from the field labora-
tory.
Personnel Recruitment and
Training
Field sampling and field laboratory per-
sonnel were recruited between December 1984
and late January 1985. All pilot survey field
personnel had previous experience as field
samplers or field laboratory analysts from
Phase I of the Eastern Lake Survey. All per-
sons involved with the pilot survey field opera-
tion received extensive training specific to the
survey. The training programs, which were held
in Las Vegas, Nevada, are summarized below.
Additional details are documented in an unpub-
lished memorandum report available from the
Acid Deposition Department, Lockheed-EMSCO,
Las Vegas, Nevada.
Field Sampler Training Program
The field sampler training program
(Table 2) was designed to ensure that all
samplers would use precisely the same proce-
dures to collect stream samples and to collect
and record data. The program also served to
familiarize personnel with the duties and
responsibilities of the site coordinator. It was
planned that all pilot survey field samplers
would assume site coordinator positions for
NSS-Phase I operations in 1986.
The 5-day field sampler training course,
which was held from February 19 to 23, 1985,
covered the basic concepts and protocols of
the pilot survey. The training program provided
a detailed introduction to locating and estab-
lishing a sample site, collecting water sam-
ples, conducting streamside and in situ mea-
surements, recording data, and maintaining
equipment. Safety training included wilderness
survival, map reading, and proper operation of
four-wheel drive vehicles. All samplers were
certified in first aid and cardiopulmonary
resuscitation.
Additional training was conducted at the
field station before actual sampling activities
commenced. This training addressed basic
measurements of hydrological parameters in
streams and included practice with the field
equipment and protocols. During the course of
field operations, each field sampler assumed
the duties of site coordinator for a short time.
Table 2. Summary of Field Sampler Training Program,
National Stream Survey-Pilot Survey
1. Employee Orientation
2. Project Orientation and Overview
3. Instrument Operation and Calibration
a) pH meter
b) conductivity meter
c) dissolved oxygen meter
d) peristaltic pump
4. Overview of Field Sampling Operation
5. Care and Packing of Sampling Equipment
6. Site Reconnaissance
a) map training
b) development of stream dossiers
c) completion of reconnaissance forms
d) completion of watershed characteristics form
7. Staff Gauge Placement
8. Completion of Logbook, Calibration Form, Stream
Data Form
9. Field Safety
a) safety gear
b) survival
c) communications
d) four-wheel drive training
10. Simulated Sampling Operation
Field Laboratory Personnel
Training Program
The training program for field laboratory
personnel (Table 3) was designed to ensure
that personnel would analyze and process
water samples in accordance with approved
pilot survey analytical and quality assurance
protocols. During the 5-day field laboratory
training program held in Las Vegas, field
laboratory personnel received training in all
technical aspects of field laboratory operation.
During field operations each person assumed
-------
ths duties of laboratory supervisor to gain
familiarity with the dutias and responsibilities
of the position.
individual's last exposure to MIBK, because
this compound does not have a long residence
time in the body.
SUffiimiaEV Off FloOd
iii, Ctatlenafl
Tralnteg
Suroey-Plllot
1. Employee Orientation
2. Project Orientation and Overview
3. Review of Field Laboratory Operations
a) sample flow
b) jab organization
c) personnel responsibilities
4. Laboratory Safety
5. Review of Methods
a) filtration and sample preservation
b) dissolved inorganic carbon analysis
c) pH determination
d). turbidity determination .
e) true color determination
f) aluminum extraction
g) organic mortomeric aluminum extraction (ion
exchange column)
8. Review of Batch and Shipping Forms, Logbooks
7. Communications
8. Inventory Control
8. Waste Handling and Disposal
10. Data Tracking/QA Plan
11. Simulation of Daily Operation
12. Medical Surveillance
Laboratory personnel were also trained
in laboratory safety and in the use of safety
equipment. Each person was fitted for a
half-mask respirator and underwent medical
surveillance physicals before the survey. All
laboratory personnel were certified in first aid
and cardiopulmonary resuscitation. Because
a hazardous chemical, methyl isobutyl ketone
(IVIIBK), was used in some pilot survey field
laboratory procedures, it was necessary to
monitor laboratory personnel for the presence
and concentration of MIBK in the bloodstream.
At the end of the survey, all personnel received
blood tests to determine the absence of or
presence and concentration of MIBK. This
testing was conducted within 24 hours of an
To improve field laboratory safety, the
field laboratory trailer to be used for the pilot
survey was modified, and additional equipment
was installed. As a result of experience
gained during the Eastern Lake Survey-Phase
I, a safety audit checklist was developed and
implemented for the pilot survey. In addition,
a portable photoionization detector was tested
in the field laboratory during the pilot survey.
This device monitored the concentration of
WilBK vapors in the laboratory and sounded an
alarm when the concentration of MIBK inside
the trailer exceeded 25 ppm. The use of this
.detector eliminated the requirement that all
field laboratory personnel wear respirators
whenever MIBK was being used.
Sift© R©(g©iminisi5s@siini©© A©ftwofti©s
Prior to the initiation of field activities,
basic information on each stream reach was
obtained by telephone from persons who were
familiar with the areas of specific stream
reaches. Members of the Soil Conservation
Service, U.S. Forest Service, U.S. Geological
Survey. National Park Service, Tennessee Valley
Authority, and the Cherokee Indian Tribe, as
well as personnel from state highway patrols,
county sheriff's departments, and other local
law enforcement agencies were called. These
people provided information on ease of access
to the reach, information on driving and hiking
times, and the names of landowners to be
called for access permission.
The basic information and a USGS
1:24,000 scale map delineating the stream
reach were used to compile a dossier on each
stream reach. Appropriate county and state
road maps and any necessary access permits
were added to the dossier before field
operations were initiated.
Each field sampling team was assigned
to sample 20 or 21 stream reaches during the
pilot survey. Two sampling sites were located
on each reach, one near the upstream node,
the other near the downstream node. Each
team visited all assigned sites before sampling
-------
activities commenced. When possible, a
sampling team was accompanied on this first
visit by a local person who was familiar with
the stream reach. Access permission was
acquired prior to the visit; at the sampling site,
site information contained in the dossier was
verified or corrected. During and after the site
visit, a reconnaissance form was completed
for the sampling site. This form was used to
record access routes (roads or trails), driving
and hiking times, required maps, and the
names of people who had authority to grant
permission to access the sampling site.
A watershed characteristics form (Figure
3) was completed during the site visit. This
form identified important characteristics useful
in interpreting the results of chemical analyses.
Photographs were taken at each sampling site
(upstream view, downstream view, and any
unique views that could assist with site
identification). To assure that photographs
could be readily matched with the proper sites,
a card was photographed containing the
sampling date, stream name, site ID number,
sampling team, and frame numbers of all site
photographs. The card was photographed
before any site photographs were taken. After
the site visit, the completed reconnaissance
form, the watershed characteristics form, and
the photographs were added to the dossier.
During the reconnaissance visit, hydro-
logic staff gauges were installed in secure
locations at each site. Gauges were installed
in an area of quiet water (i.e., no obstructions
and no eddies) at the side of the main
channel. At least 6" of the scale was placed
below the surface, and the gauge was
oriented parallel to the direction of flow. The
initial reading of gauge height was marked at
a secondary location on shore.This secondary
reference allowed qualitative changes to be
observed in the event a gauge was lost during
the survey. The staff gauges were left in
place throughout the survey. The cross-
sectional area of the stream at each site was
determined by measuring the depth at 10 to 20
points spaced equally along a transect across
the stream.
-------
NATIONAL SURFACE WATER SURVEY
WATERSHED CHARACTERISTICS
FORM 7
D D M M M Y Y
DATE
COUNTY
STATE
STREAM ID
STREAM NAME
LATITUDE: t > i i°i i i i i i • i LONGITUDE:,
ELEVATION
PHOTOGRAPHS
FRAME ID AZIMUTH
. , . ,(^J LAP CARD
,_,,_, O --"-„"
Ol ' 1 1 1 — 1
1:250.000 MAP NAME
1:24.000 MAP NAME
VISUAL ESTIMATES:
STRFAM VUIOTH-
STREAM DEPTH:
GAGE HEIGHT: I M 1.1 lit
units --»
— o
o
WATERSHED ACTIVITIES/DISTURBANCES
(Check all mat apply)
units
O Roadways:
O Dwellings:
O Agriculture:
O Industry:
D Logging:
O Mining:
D Quarries:
D Beaver dams:
D Livestock:
n Olhnr
D Paved D Unpaved -
n Bridged D Culvert
D Single unit(s) D Multiple unit(s)
O Cropland O Pasture
O Fenced D Unfenced
Spcc'fy Typ^?: ... .
D Above Site D Below Site
D Cattle D Sheep
n Horses n nthor
BANK VEGETATION WITHIN 100 METERS OF
STREAM BED (Check all that apply)
Type
Deciduous Trees
Coniferous Trees
Shrubs:
Wetland Areas:
Grasses:
Rocky/Bare:
Absent Sparse Moderate Hea»y
D D D D
D D D D
D D D D
D D D D
a a a D
aana
COMMENTS D. SEE REVhHbt SIDE
D Grade Distance from Stream
Distance from Stream:
Distance Irom Stream:
Historic*? from Stream-
Distance from Stream"
Diclanrp from Slrnam
Oisf'lncp frnm Stream-
nistanro frnm Stream-
STREAM SUBSTRATE
(Check nil that apply)
Type Absent Sparse
Boulders: O D
Cobble: D D
Gravel: D D
Sand: D d
Silt: D D
Aufwuchs: D D
DATA QUALIFIERS
(?) - nthprs
— O
— O
o
\^f
— - O
o
- \^r
o
^^
n
\^/
o
Moderate Heavy
D D
D a
a a
a a
a a
a a
FIELD CHEW DATA
flRFW in SAMPI FR 7
SAMPLER 1
rKFn HY
Figure 3. Watershed characteristics form, National Stream Survey-Pilot Survey.
10
-------
Section 3
Field Station Operations
Field operations for the pilot survey were
conducted in spring (March 1 through April 30,
1985) and summer (June 26 through July 17,
1985).
Spring sampling, preceded by 2 weeks of
preliminary site visits, began on March 17,
1985. Each stream reach was sampled bi-
weekly near its downstream node. Each
downstream site was sampled three times.
During the third set of site visits, 23 regular
stream reaches were also sampled near their
upstream nodes. Eighteen reaches that were
sampled during the first 3 days of field opera-
tions were sampled a fourth time at the end of
the spring sampling period because of equip-
ment problems and protocol changes.
During the summer sampling period, each
regular reach was sampled once near its
downstream and once near its upstream node
on the same day. Each special interest reach
was sampled once, near its downstream node.
A generalized flowchart of field sampling
and processing operations is given in Figure 4.
Each team sampled one to three streams per
day, 5 consecutive days a week. Samples and
field data forms were delivered to the field
laboratory within 9 hours of sample collection.
At the field laboratory, samples were analyzed,
and aliquots were prepared and preserved
within 18 hours of collection. Preserved ali-
quots were shipped via overnight courier to a
contract analytical laboratory (New York State
Department of Health Laboratory, Albany, NY)
to ensure they would arrive within 48 hours
after collection.
Field Station Organization
The field station in Sylva, NC, was
staffed by a site coordinator, three two-
member sampling teams, and a laboratory
crew of four persons.
The site coordinator was responsible for
the overall operation of the field station.
Duties of the coordinator included:
1. Functioning as the on-site contact for
the NSS management team, the Las
Vegas communications center, and
Lockheed-EMSCO administrative
personnel.
2. Establishing a local communications
center and setting up the field labo-
ratory trailer.
3. Developing daily sampling itineraries
for sampling teams.
4. Serving as the liaison between field
sampling and field laboratory person-
nel.
5. Receiving audit samples.
6. Organizing batches and assigning
sample numbers.
7. Shipping processed sample aliquots
to the contract analytical laboratory
and shipping data forms to the pilot
survey data management center,
quality assurance personnel, and the
NSWS sample tracking office.
8. Filing a daily sampling and operations
report with the Las Vegas communi-
cations center.
9. Serving as a reserve field sampler or
laboratory analyst.
11
-------
INSTRUMENT CALIBRATIONS
AND
QUALITY CONTROL_CHECKS
PACK EQUIPMENT AND SUPPLIES
SAMPLING TEAMS
DEPART
COORDINATOR FILES DAILY
REPORTS WITH LAS VEGAS
COMMUNICATIONS CENTER
REQUEST AUDIT SAMPLES
FOR NEXT DAY
1
DAILY BRIEFING
NEXT DAY'S
ACTIVITIES
SAMPLE COLLECTION
AND FIELD
MEASUREMENTS
'
SAMPLING TEAMS
RETURN TO
FIELD STATION
|
SAMPLES AND DATA
FORMS TRANSFERED
TO
FIELD LABORATORY
'
FIELD LABORATORY
PREPARATORY
ACTIVITIES
i
SAMPLES ORGANIZED
INTO BATCH
FOR PROCESSING
AUDIT
SAMPLES
SHIPPED
|
RECEIVE
AUDIT SAMPLES
AT
FIELD LABORATORY
SAMPLES ANALYZED,
PROCESSED,
AND PRESERVED
ALIQUOTS AND DATA
FORMS SHIPPED
THE NEXT DAY
Figure 4. Flowchart of dally field station operations, National Stream Survey-Pilot Survey.
10. Maintaining field safety standards
and initiating search and rescue
procedures.
Duties of the field sampling and field
laboratory personnel are described in Sections
4 and 5, respectively.
Field Station Communications
Daily communications was maintained
by the site coordinator with the field sampling
teams, the field laboratory, and the Las Vegas
communications center. Communication with
sampling teams in the field served to track
their progress, to resolve unexpected prob-
lems, and to ensure safety. The field labora-
tory was advised of the expected number of
samples and of the expected arrival time of
field crews. The Las Vegas communications
center was informed daily about sampling
activities, shipment of samples, and supply
needs.
Each day the members of each sampling
team (and all accompanying persons) com-
pleted a sampling itinerary form. The form
included physical descriptions of each indivi-
dual, clothing descriptions, names and tele-
phone numbers of a person to call in the event
of an emergency, a description of the planned
12
-------
sampling schedule, and routes of travel. The
completed form could be used by the coor-
dinator to. facilitate search and rescue opera-
tions; if necessary. Each team was required
to telephone the local communications center
or field laboratory by 4:30 p.m. (spring sam-
pling period) or 6:00 p.m. (summer sampling
period) to communicate their location and their
expected arrival time at the field laboratory.
The coordinator called the Las Vegas
communications center every day to report on
the previous day's activities. A communica-
tions form completed by the coordinator was
transcribed in Las Vegas over the telephone.
Conference calls between the coordinator and
members of the project management team
were held weekly.
13
-------
Section 4
Field Sampling Operations
Daily field sampling activities involved
traveling to sampling sites, collecting samples
and field data at each site, and delivering
samples to the field laboratory. Four-wheel
drive vehicles were used for transportation to
most sampling sites. Sites inaccessible by
vehicle were accessed by hiking, boat, or
horseback. In addition to collecting water
samples, the pH of the stream water was
determined at streamside, and measurements
of dissolved oxygen content and conductivity
were measured in situ. The equipment, quality
assurance measures, and field methodologies
are described in this section.
Sampling Equipment
The equipment used by sampling teams
at each stream site is listed in Table 4. A
Cole-Palmer Masterf lex 7533-30 portable peri-
staltic pump and a rechargeable battery were
used in conjunction with 1/4-inch Tygon tubing
(surgical grade) and a 6-foot bamboo boom to
collect water from midchannel. Samples were
collected in 4-liter Cubitainers, in 60-mL poly-
propylene syringes, and in 500-mL wide-mouth
bottles constructed of amber high-density
linear polyethylene.
A Beckman pHI-21 portable pH meter
equipped with an Orion Ross model 81-56
combination electrode was used to measure
pH. After the spring sampling period, the
model 81-56 electrodes were replaced with
Orion-Ross model 81-04 glass-bodied com-
bination electrodes. For waters of low ionic
strength, the 81-04 electrodes had improved
response characteristics over the model 81-56
electrodes. Orion model 231 portable pH
meters were available as backup instruments.
A Yellow Springs Instruments (YSI)
model 33 S-C-T portable meter was used to
measure conductivity in situ YSI model 54 and
model 57 portable oxygen meters were used to
measure dissolved oxygen in situ. These two
meters are nearly identical, and an equipment
shortage necessitated the use of two models.
Quality Assurance and Quality
Control of Field Measurements
Calibration procedures and quality con-
trol checks for the pilot survey were developed
and documented in a quality assurance plan
(Drous6 et al., 1986). Instruments were cali-
brated daily at the field station. At each site
before and after sample measurements, the
calibration of pH and conductivity meters were
checked by using standards of known concen-
tration prepared daily: The dissolved oxygen
meters were calibrated at each site before
sample measurements.
The: calibration of the instrument was
checked by using an independent method at
the field station before and after each day's
use. All calibration information was recorded
in a field logbook and on the field data forms.
To calibrate the pH meter, the previous
day's calibration information was cancelled,
and the meter was standardized with NBS-
traceable pH 7.00 and pH 4.01 buffer solutions.
The standardization values were checked with
new aliquots of the buffer solutions. If the
measured pH of each buffer solution was not
within 0.02 pH units of the theoretical value,
the instrument was restandardized. Next, a
1 x 10"4 N H2S04 solution that had a theoretical
14
-------
Table 4. Sampling Equipment Used During National Stream Survey-Pilot Survey
I. SAMPLE COLLECTION
Reconnaissance forms
Maps
Field logbook and forms
Pump
Battery and cables
Stream data forms
Watershed characteristics forms
Field sample labels (tie-on/adhesive)
Ziploc bags
Portable coolers (w/gel packs)
Pens/pencils/markers
Deionized, H.O (two 4-liter Cubitainers/field blank)
4-liter Cubitainers* (1/sample)
500-ml bottles* (1/sample)
Tygon pump tubing* (10-20'/sample)
Syringes/valves* (3/sample)
II. pH MEASUREMENT
pH meter/case
Batteries
2 Electrodes (with BNC connectors)
ATC probe
Sample chamber
Ringstand/clamp
250-ml beakers
7.00 buffer (2 bottles)
II. pH MEASUREMENT (cont.)
4.00 buffer (2 bottles)
QCCS solution (2 bottles)
Wash bottle with deipnized H2O
Electrode filling solution
Instruction manual
III. CONDUCTIVITY
Meter/case
Batteries
Probe/storage bottle
Calibration tables
Deionized H2O wash bottle
74 pS/cm QCCS (2 bottles)
Instruction manual
IV. DISSOLVED OXYGEN
Meter/case
Batteries
Probe
Calibration chamber
Calibration tables
Bucket
Calculator
Membrane kit
Filling solution
Instruction manual
Prepackaged in plastic bags for each site.
pH of 4.00 was used as a quality control
check sample (QCCS). The QCCS was mea-
sured with each successfully standardized
meter. If the measured pH of the QCCS was
more than 0.10 unit different from 4.00, a new
QCCS was prepared. If the measured pH of
the new QCCS was not with control limits, the
electrode or meter was replaced.
The factory calibration of each conductiv-
ity meter was checked by using KCI solutions.
These solutions had specific conductances of
1413, 147, and 74 /uS/cm at 25 *C. Measure-
ments of the 1413 and 147 /L/S/cm QCCS solu-
tions were required to be within 10 percent of
the theoretical value. Readings of the 74
/uS/cm solution were allowed a range of ±10
pS/cm (13.5 percent). Meters or probes that
did not produce acceptable QCCS measure-
ments were replaced.
The dissolved oxygen meter was cali-
brated by using a chamber containing water-
saturated air (air at 100 percent relative humid-
ity). The calibration was checked in water that
had been saturated with bubbled air. Read-
ings were required to be within 0.5 mg/L O2 of
each other, or the probe was serviced, and the
unit was recalibrated.
Daily Sampling Activities
The daily routine of each sampling team
(Figure 5) consisted of preparing for sampling,.
traveling to sites and collecting samples and
field data, and post-sampling activities.
Preparation for Sampling
At the field station, each instrument was
calibrated and checked. The sampling equip-
ment and supplies necessary for the day's
sampling schedule were obtained. A daily
travel itinerary that listed the intended routes
of travel, the sites to be visited, and a physical
description of each member of the team was
filed with the site coordinator.
Sampling and Field Measurements
Upon arrival at a stream site, the two
samplers set up the equipment, checked the
15
-------
calibration of each instrument, and collected all
required samples and field measurements.
The calibration of the pH meter was checked
by using the pH 4.00 QCCS (see p. 14). The
conductivity meter was checked by using the
74 pS/cm QCCS (see above). At each site, the
dissolved oxygen meter was recalibrated with
water-saturated air. Sample collection, in situ
measurements, and streamside measurements
were conducted simultaneously but in such a
way as to avoid affecting one another.
Figure 5. Dally field sampling activities, National
Stream Survey-Pilot Survey.
Sample Collection-
Water samples were collected mid-
channel in an area of flowing water that was
not classified as a riffle. Water was pumped
to the bank by using the peristaltic pump and
the Tygon tubing that was attached to the
sampling boom with clamps. The end of the
tubing was placed below the surface in mid-
channel. This submerged tubing was located
far enough above the bottom of the stream to
avoid stirring up the sediments and introducing
them into the sample. Water from each
stream was collected sequentially in several
containers. A routine sample consisted of one
4-L aliquot (collected in a 4-L Cubitainer), three
60-mL aliquots collected in syringes (one of
which was used to measure pH at stream-
side), and one 500-mL aliquot for total sus-
pended solids analysis (collected in a wide-
mouth polyethylene bottle). All containers
were rinsed three times with stream water
before they were filled.
Each day, one of the three teams col-
lected a field blank sample. The field blank
sample was collected before any routine
samples. To collect the field blank sample,
deionized water (from two 4-L Cubitainers
filled at the field laboratory) was pumped into
a clean 4-L Cubitainer and into a 500-mL
bottle. Syringes were not filled for the field
blank sample.
Each day, one of the three teams also
collected a field duplicate sample at one site.
A field duplicate sample was collected by
filling a second set of containers with stream
water immediately after the routine sample
was collected.
pH Measurement
At each site, pH was measured by two
methods. One measurement was obtained
from a sample aliquot collected in an open
beaker. This sample was exposed to the
atmosphere during collection and measure-
ment and the measurement was operationally
defined as an open-system determination. A
second pH measurement was obtained from
a sample aliquot collected in a syringe and
measured in a sample chamber. The syringes
were affixed directly to the end of the pump
tubing to form an airtight seal. This sample
was not exposed to the atmosphere during
either collection or measurement, and the
measurement was operationally defined as a
closed-system determination. During the pilot
survey, both methods were modified (see
Section 6 for further discussion). The
16
-------
procedures described below were in use at the
end of the survey.
A 250-mL beaker was rinsed three times
with stream water and then was filled with
stream water by using the peristaltic pump.
The electrode and the automatic temperature-
compensating (ATC) probe were immersed in
the sample, and the sample was swirled for
3 minutes. A fresh aliquot was collected, the
electrode and the probe were immersed in the
sample, and the measurement was taken. The
sample was not swirled during the measure-
ment process. The pH and temperature were
recorded when the pH reading was stable
(i.e., changed less than 0.01 pH unit over a
1-minute period).
For the field duplicate measurement, this
process was repeated by using a third 250-mL
aliquot of sample.
A sample chamber, described in Hillman
et al., (1986) was assembled at streamside.
The ATC probe, immersed in a 250-mL beaker
filled with stream water, was used to monitor
the approximate water temperature inside the
chamber during the pH measurement process.
A 60-mL syringe was rinsed three times with
stream water, was attached to the end of the
pump tubing, and then was filled with stream
water by using the peristaltic pump. The
syringe was affixed to the sample chamber,
and the electrode was inserted loosely into the
chamber. The chamber and electrode were
rinsed twice with 5- to 10-mL aliquots injected
from the syringe. The chamber then was filled
with sample injected from the syringe, and the
electrode was inserted into the chamber to
seal it. The initial pH and temperature read-
ings were noted; a stopwatch was used to
track elapsed time.
The pH, temperature, and elapsed-time
values were recorded again when the pH
reading was stable. A 5-mL aliquot (the
approximate volume of the sealed chamber)
from the syringe was slowly injected into the
chamber, and when the pH reading was
stable, values were recorded again.
A second 5-mL aliquot was slowly in-
jected into the chamber, and the process was
repeated. If the pH value for the second 5-mL
aliquot was within 0.03 pH unit of the first
5-mL aliquot, the pH, temperature, and
elapsed-time values were recorded as final
readings for the sample. If they were not in
agreement, a third 5-mL aliquot was injected
and measured. Additional 5-mL aliquots were
injected and measured until the stable pH
readings of two successive aliquots agreed to
within 0.03 pH unit. The pH, temperature, and
elapsed time values of the last aliquot mea-
sured were recorded as final readings.
For the field duplicate measurement, this
process was repeated by using a second
syringe.
In Situ Conductivity !M@@§ur®m®flt
The conductivity probe was attached to
a bamboo sampling boom and was immersed
in the stream to a depth of at least 10 cm
(mid-depth if the site depth was less than 10
cm) in an area of flowing water. The water
temperature and conductivity readings were
recorded when stable (i.e., the conductivity
changed less than 5 /jS/cm over a 1-minute
period). No field duplicate measurements of
conductivity were made.
In Situ Dissolved Oxygen
The dissolved oxygen meter was cali-
brated by sealing the probe in a moist calibra-
tion chamber constructed of 6 inch by 1 inch
flanged stainless steel plumbing pipe sealed
with rubber stoppers at both ends. This
chamber was immersed in the stream for 15 to
20 minutes to produce an atmosphere of
water-saturated air inside the chamber. The
meter and probe were calibrated by using a
correction factor derived from the temperature
of the chamber and the theoretical value of the
partial pressure of oxygen at the elevation of
the sample site.
The dissolved oxygen probe was
attached to the sampling boom and immersed
to a depth of at least 10 cm in an area of
flowing water. The water temperature and
dissolved oxygen values were recorded when
17
-------
the dissolved oxygen reading was stable (i.e.,
changed less than 0.5 mg/L O over a 1-minute
period). No field duplicate measurements of
dissolved oxygen were made.
Other Measurements
All quality control and final field measure-
ments were recorded in logbooks at stream-
side. Additional data recorded at streamside
included stage height, cloud cover, instrument
problems, QCCS results, date, time, team
member identification, elevation of the site,
and unique conditions which could affect
water quality.
All data and observations were tran-
scribed from field logbooks onto standardized
field data forms (Figure 6). Copies of the
watershed characteristics form (see Figure 3
were also carried on each site visit. Any signi-
ficant change in the characteristics of a site
was noted by completing a new form.
Post-sampling Activities
At each site, quality control checks were
repeated on the pH and conductivity meters
after measurements were made. Water sam-
ples were placed in portable soft insulated
coolers that contained chemical refrigerant
packs to maintain sample temperatures near
4 °C during transport to the vehicle. At the
vehicle, samples were transferred to insulated
coolers that contained six to eight chemical
refrigerant packs. Equipment and supplies
were packed for transport to the next site.
At the completion of sampling for the
day, the teams returned to the field laboratory.
The dissolved oxygen meter was checked by
using an air-saturated water solution (see
p. 15). The conductivity and dissolved oxygen
probes were stored overnight in tap water, and
the pH electrodes were stored in 3 M KCI.
Water samples (Cubitainers, sealed syringes,
and 500-mL bottles) were delivered to the
coordinator to be organized into a batch for
processing and analysis in the field laboratory.
18
-------
NATIONAL SURFACE WATER SURVEY D D M M M Y Y
STREAM DATA DATE -------
FORM 4 TIME — — :— — hr
pH METER ID: i i i i
T/COND. ID: i i i — i
STATE STREAM ID STREAM NAME D' S LVED °* 'D' ' ' U~J
LATITUDE: , , , ,°i i i I i i i_i' LONGI1
1:250.000 MAP NAME:
1:24.000 MAP NAME:
;LOUD COVER: i ii ii — i %
RAIN: D PREV. D NO D LIGHT D HEAVY
GAGE HEIGHT: , ,., ,', , ft f~}
IH RISING D FALLING ^"^
DATA QUALIFIERS
(A) INSTRUMENT UNSTABLE
(D REDONE FIRST READING NOT
ACCEPTABLE
© SLOW STABILIZATION
(N) DOES NOT MEET QCC
(#) OTHFR (pyplain)
COMMENTS:
D NOT SAMPLED. SEE BELOW
METER CAL
QCCS INITI
ROUTINE 0
SAMPLE TE
ROUTINE C
DUPLICATE
SAMPLE TE
DUPLICATE
QCCS FINA
SAMPLE REPLICATE
pH
Y N
IBRATION: D D
QCCS = pH 4.00
AL: i — ii — i.i — ii — ' \_y
PEN: i ii i.i ii <\_)
MP.: i ii i.i i ° G(^_)
LOSED: i — . . i.i ii >\_)
I OPEN: i ii i.i ii i Q_)
MP. i — ii—i.i — i ° c\_)
-. CLOSED: i , , i .1 i i I C_J
CONDUCTIVITY >«S
Cond. QCCS = 74 @ 25 ' C
QCCS INITIAL: , . . . , . Q^)
IN SITU: i — ii — ii — i r")
STREAM TEMP.: i i , i.i i ° C (~^
QCCS FINAL: i — i i — • • — • \_s
DISSOLVED OXYGEN mg / I
QCC = Theoretical — Measured
^INITIAL: tZj i i.i i^
IN SITU: i ii i.i i Q_^
+ X~N
AFINAL: , ,, , ., ,(^J
REASON
(CHTECK)/IPLED: ° INACCESSIBLE D NO ACCESS PERMIT D TOO SHALLOW D
FIELD LAB USE ONLY
TRAII FR in
RATP.H in
SAMPI F in
OTHFR
. FIELD CREW DATA FORM DISTRIBUTION
P.RFW in WHITE COPY - ORNL
PINK COPY — EMSL-LV
SAMPI FR 9
r.HFP.KFn RY
Flgura 6. Field data form, National Stream Survey-Pilot Survey.
19
-------
Section
The field laboratory trailer provided a
facility to process dilute water samples in a
controlled, environment within a short time
after collection. Measurements of physico-
chemical parameters that had short holding
times were conducted in the field laboratory.
These measurements included closed-system
determinations of pH and dissolved inorganic
carbon (DIG), turbidity, and true color.
Each sample of stream water was
prepared into several aliquots at the field
laboratory for later, more detailed analyses at
a contract analytical laboratory. These activi-
ties were conducted in an environment that
minimized the potential for contamination.
The laboratory was staffed by a labora-
tory supervisor and three laboratory analysts.
The laboratory supervisor was responsible for
the daily operation, security, and safety of the
laboratory. The laboratory supervisor also
ensured that samples were analyzed and
processed by approved methods and in accor-
dance with the quality assurance plan. The
supervisor performed pH and dissolved inor-
ganic carbon (DIG) determinations on water
samples and transcribed all analytical data
collected in the laboratory onto a standardized
data form. Additional duties included trouble-
shooting instrument malfunctions, maintaining
equipment, and tracking the inventory of sup-
plies and equipment.
The three laboratory analysts ware jointly
responsible for preparing reagents and stan-
dards for the sampling teams and for prepar-
ing preserved sample aliquots for shipment by
the coordinator to the contract analytical
laboratory. Analyst 1 prepared all aluminum
extractions. Analyst 2 was responsible for the
preparation of aliquots requiring filtration.
Analyst 3 conducted turbidity and true color
determinations, prepared all aliquot labels and
containers, prepared unfiltered aliquots, and
preserved all aliquots. Each person in the
laboratory rotated through all four positions to
become familiar with all aspects of the labora-
tory operations.
The daily activities associated with the
operation of the field laboratory (Figure 7)
began with calibration and other preparation
prior to the arrival of samples from field crews
and concluded with packing samples and data
forms for shipment the following day. Field
laboratory analytical and sample processing
methodologies used in the pilot survey are
presented in Hillman at al. (1987). Quality
assurance and quality control measures used
in the laboratory during the pilot survey are
described in Drouse et al. (1986).
To monitor the performance of the field
laboratory and contract analytical laboratory,
natural water samples collected from lakes
having a well-known chemical composition
(termed audit samples) were prepared by
Radian. Corporation, Austin, Texas. Audit
samples were shipped daily via overnight
courier service to tfta field laboratory. The
coordinator was responsible for receiving the
audit samples and for storing them at 4°C
until they were incorporated into a sample
batch for processing and analysis. Details on
the chemical composition and preparation of
audit samples can be found in Drouse et al.
(1986).
20
-------
SAMPLE TRANSFER
FROM FIELD CREWS
COORDINATOR
CHECK IN AUDIT SAMPLES
AND ORGANIZE BATCH
(FORMS 4 AND 5)
LABORATORY
SUPERVISOR
DIC
DETERMINATIONS
DH
DETERMINATIONS
DATA
TRANSFER
(FORMS)
DATA
.DATA
ANALYST 3
ALIQUOT
PREPARATION
AND
PRESERVATION
TURBIDITY
DETERMINATIONS
TRUE COLOR
DETERMINATIONS
ANALYST 2
SAMPLE
FILTRATION
NEXT DAY
ANALYST 1
ALUMINUM
EXTRACTIONS
(ALIQUOTS 2
AND 8)
PREPARE
ALIQUOTS
FOR SHIPPING
POST PROCESSING ACTIVITIES
— r
I
I
I
I
PREPARE REAGENTS
AND SUPPLIES
FOR FIELD CREWS
LABORATORY
CLEAN UP
COORDINATOR
PACK AND SHIP
ALIQUOTS
(FORM 3)
SHIP
DATA FORMS
Figure 7. Flowchart of dally field laboratory operations, National Stream Survey-Pilot Survey.
21
-------
Two types of audit samples ("field" audit
samples and "laboratory" audit samples) were
handled at the field laboratory. Field audit
samples were sent to the field laboratory in
2-L wide-mouth amber polyethylene bottles.
These audit samples were labeled, analyzed,
and processed in the same manner as stream
water samples. Laboratory audit samples
were prepared and processed at Radian Cor-
poration following the same protocols used in
the field laboratory for natural water samples.
Laboratory audit samples were shipped to the
field laboratory in containers identical to the
containers used to ship aliquots from the field
laboratory to the contract laboratory. Labora-
tory audit samples received by the field labora-
tory were not processed but were relabeled
and were incorporated into a sample batch.
Upon receipt of audit samples, the
coordinator completed sample tracking forms
which were later returned to Radian Corpora-
tion. Each audit sample was incorporated
along with natural water samples into a batch
for that day's processing (see p. 23). The
assigned batch and sample identification
numbers were recorded on the audit sample
labels for each field audit sample and for each
aliquot of a laboratory audit sample. The
audit sample labels were then removed and
were placed in a logbook. The batch and
sample identification numbers were recorded
on the field audit sample container. The
various aliquot containers of a laboratory audit
sample were relabeled with the appropriate
sample aliquot labels. All audit samples were
known to be audit samples at the field labora-
tory. However, they were indistinguishable
from stream water samples when received at
the contract laboratory.
• The field laboratory staff began preparing
for daily operation one to two hours prior to
the arrival of samples from the field. Each
day, prior to any sample processing or analy-
ses, the field laboratory floor was mopped,
and all counter surfaces were cleaned to
minimize the potential of contamination from
dust. All instrumentation in the field laboratory
was left "on" or on "stand-by" at all times while
the field laboratory was operational.
The field laboratory supervisor prepared
calibration standards and QCCS solutions
(Hillman at al., 1987) for DIG analysis. The
carbon analyzer was calibrated, and its opera-
tion checked with these solutions. The pH
meter was standardized with pH 4.01 and 7.00
NBS-traceable buffers. The standardization
was subsequently checked with fresh buffers
and with a freshly prepared 1 x 10"4 N H^O,
QCCS. The field laboratory supervisor col-
lected syringe samples for closed-system
determinations of pH and DIC from each field
audit sample.
One analyst prepared reagents, equip-
ment, and labels for use in the two aluminum
extractions. Reagent dispensers were checked
for accuracy- of delivery. Ion exchange
columns used to prepare the fraction of
organic monomeric aluminum were adjusted to
be within 0.5 pH unit of the expected sample
pH values (Hillman et al., 1987) which ranged
between 4.8 and 8.4 during the pilot survey.
The peristaltic pump used in the procedure
was calibrated to deliver 30 mL/min of sample
through ths column.
The second analyst assembled and
organized all equipment and supplies required
for the sample filtration and prepared a log-
book to check off sample aliquots as they
were prepared and preserved. The nephelo-
meter was calibrated and was checked for
proper operation; the color test kit was
assembled; and the logbook for recording
turbidity and true color data was prepared.
The third analyst prepared all necessary
aliquot bottles and aliquot labels for the sam-
ple batch and also prepared materials neces-
sary for aliquot preparation. Sample aliquot
bottles and labels were prepared prior to the
start of processing activities to minimize the
possibilities of error in filling the bottles. To
minimize ths possibility of preserving an ali-
quot with the wrong acid, labels were color-
coded with respect to the type of preservative
required.
Three types of water samples were
received at the field laboratory from field
sampling crews: routine samples, field dupli-
cate samples, and field blank samples. The
collection of these samples was described in
Section 4.
22
-------
All sample containers and field data
forms collected during each day's sampling
operation were transferred to the coordinator.
The temperature of each cooler was checked
and was recorded on the appropriate field
data forms. All sample containers were inspec-
ted for leakage and possible contamination.
Syringes were checked for the presence of air
bubbles. All comments regarding samples
were recorded on the appropriate field data
forms.
Data forms were inspected for com-
pleteness and legibility by the coordinator.
Sampling personnel were debriefed by the
coordinator to discuss any problems encoun-
tered in the field.
Organization of Samples into a Batch
The. coordinator .organized stream water
samples and audit samples into a batch for
processing and analysis. A field batch was
defined as all samples processed by the field
laboratory on a given day. Field batches were
sequentially assigned unique batch identifi-
cation numbers, beginning with 2001 and
concluding with 2057.
Each sample in the batch (routine, field
duplicate, field blank, and audit) was then
assigned, at random, a unique sample identifi-
cation number. The batch and sample identifi-
cation numbers were recorded on all field
sample container labels (4-L Cubitainers,
syringes, and 500-mL bottles). These numbers
were also recorded on the labels of correspon-
ding sample aliquots prepared from each
Cubitainer sample.
After batch and sample identification
numbers were recorded on field sample labels,
the coordinator entered batch information,
stream identification numbers, and sample
codes from all samples on the field laboratory
data form (Figure 8). The stream identification
number and sample code for each sample
were entered on the field laboratory data form
on the line corresponding to its assigned
sample identification number. In the case of
an audit sample, no stream identification
number was entered. The audit sample code
was entered in the "sample code" column.
During the organization of a batch and
until the batch was processed by the field
laboratory, all samples were held at 4 °C,
either in the field laboratory refrigerator or in a
cooler containing frozen chemical refrigerant
packs. When the assignment of sample identi-
fication numbers was complete, the field
laboratory coordinator informed the field
laboratory supervisor.
While the samples were being organized
into a batch, the field laboratory supervisor
and analysts made preparations to process
and analyze samples.
Transfer of Samples to Field
Laboratory
Once the batch was organized and all
field sample containers were properly labeled,
one syringe from each field sample was
placed in the laboratory refrigerator for DIG
. .analysis. The other syringe from each field
sample was placed on a shelf in the labora-
tory to warm to ambient temperature prior to
pH determination. The field laboratory super-
visor collected two syringes from each field
audit sample and labeled them with batch and
sample identification numbers. One audit
sample syringe was placed in the refrigerator
for DIG analysis, and the other was placed on
the shelf with those syringes used for pH
determinations.
Sample Analysis and Processing
The flow of samples through the field
laboratory is diagrammed in Figure 9. The
500-mL aliquots of sample collected in the field
were delivered to the Environmental Biology
Laboratory at Southwest Technical College in
Sylva, NC. This laboratory performed the
determination of total suspended solids by
using standard methodology (U.S. EPA, 1983;
Hillman et al., 1987).
Eight separate aliquots were prepared
from each Cubitainer sample (see Hillman et
al., 1987). A 125-mL aliquot of filtered water
from each Cubitainer was collected in an acid-
washed glass beaker for aluminum extractions
(for total extractable and organic monomeric
aluminum). Details of the preparation of the
organic monomeric aluminum fraction are
presented in Section 6.
23
-------
DATE RECEIVED
BV DATA MOT. -
NATIONAL SURFACE WATER SURVEY
BATCH / QC FIELD DATA
FORM 5
LAB TO WHICH
BATCH ID DATCHCFNT
NO. SAM
IN BATC
STATIC!"
SAMPLE
ID
01
M
03
04
05
06
07-
08
09
10
11
1?
13
14
IS
16
17
10'
19
20
21
22'
•f.t
24
2f.
?r.
21
2B
29
3(1
nun
PI-ES DATE SHIPPED
H-
1 ID CHEW ID
STREAM
ID
SAMPLE
CODE
TD
DIG Him LI
OCCS LIMITS
IJCL 2?
LCL 1 B
VALIir OC.CS
LAIlO(IAII)IIV|ill
OCCS LIMITS
IICL 4 1
LCL 3
FIELD LAB
SUPERVIS
PLED
o.
ORATORY
nn
uinniniiY
-------
-FIELD SITES
QUALITY ASSURANCE SAMPLES
, AUDIT SAMPLE .
I PREPARATION LABORATORY '
ROUTINE
SAMPLES
SYRINGES
4-L CONTAINER
500-mL BOTTLE
FIELD BLANK
(DEIONIZED WATER)
4-L CONTAINER
500-mL BOTTLE
TRANSF
AT
:RRED TO
BORATORY
4'C
FIELD DUPLICATION
SYRINGES
4-L CONTAINER
500-mL BOTTLE
AUDIT SAMPLES
FIELD LABORATORY
2-L I EIGHT
CONTAINER I PRESERVED
| ALIQUOTS
FIELD LABORATORY —
TRANSFERRED TO
SOUTHWEST TECHNICAL
COLLEGE FOR ANALYSIS
•
'
TOTAL
SUSPENDED
SOLIDS
'
DATA TRANSCRIBED
TO DATA FORM
T f
^ 500-mL BOTTLES SAMPLES ORGANIZED
SYRINGES INT° BATCH 4-L CO
1 • t V i
DISSOLVED CLOSED-SYSTEM TURBIDITY TRUE
INORGANIC pH MEASUREMENT MEASUREMENT MEASU
1
DATA TRANSCRIBED
TO DATA FORM
NEXT DAY
ALIQUOT PREPARATION
FILTRATION
PRESERVATION
ALUMINUM EXTRACTION
HOLD AT 4°C
SEND COPIES TO
DATA MANAGEMENT CENTER AND
QUALITY ASSURANCE PERSONNEL
NEXT
DAY
|
PREPARE SHIPPING FORM (S)
SHIP ALIQUOTS
TO CONTRACT LABORATORY
Figure 9. Flowchart of field sample processing and analyses conducted at field laboratory during National Stream
Survey-Pilot Survey.
The remaining prepared aliquots were
preserved with ultrapure acid if required
(Hillman et al., 1987). All aliquots were refrig-
erated at approximately 4 *C. After aliquoting,
samples in Cubitainers were allowed to warm
to ambient temperature (18 to 25 *C) before
turbidity and true color determinations were
conducted.
While the samples were being aliquoted
and preserved, the field laboratory supervisor
analyzed the refrigerated syringe samples for
DIG concentration. When these analyses were
completed, the field laboratory supervisor
performed pH determinations by using syringes
that had warmed to ambient temperature.
One routine sample in each batch was
selected at random as the "trailer duplicate."
Two aliquots of this sample from each syringe
were analyzed for DIG and pH. Two subsam-
ples of the trailer duplicate sample from the
Cubitainer were analyzed for turbidity and true
color.
When sample processing operations
were completed, preserved aliquots were
prepared for shipping. Refrigerated aliquots
were checked after 1 to 2 hours to ensure that
container caps were tight. The cap of each
aliquot bottle was taped to the bottle using
electrician's tape wrapped clockwise around
the seal. Each bottle was placed in a plastic
25
-------
bag that was sealed with a twist tie. A set of
six aliquots from each sample (not including
the two aliquots for analysis of extractable
aluminum fractions) was placed in a 1-gallon
Ziploc bag. All aliquots were refrigerated at
4 *C. The aliquots for total and nonexchange-
able extractable aluminum analyses (contained
in 10-mL centrifuge tubes) were taped and
then were bagged separately. The centrifuge
tubes were then stored at approximately 4 C
in an insulated cooler with frozen chemical
refrigerant packs.
When all analyses were completed, data
for DIG, pH, turbidity, and true color were
transcribed from laboratory logbooks onto the
field laboratory data form (Figure 8). All of the
glassware and the work area in the laboratory
were cleaned and organized before the staff
left the laboratory each night. A checklist was
used to complete a safety inspection prior to
departure.
Sample and Data Form
Shipment
The morning after a batch was pro-
cessed, the coordinator, with the assistance
of an analyst, packed aliquot bottles and
centrifuge tubes into containers for shipment
to the contract .analytical laboratory. The
bagged sets of aliquot bottles and centrifuge
tubes were placed in 30-quart insulated ship-
ping copiers which were lined with frozen
chemical refrigerant packs to maintain the
temperature near 4- *C. during, shipment.. The"
centrifuge tubes were taped to the inside of
the cooler. All aliquots from a single sample
were shipped in the same cooler; each cooler
held five to six sets of aliquots. Two frozen
chemical refrigerant packs were placed on top
of the samples, and styrofoam packing mate-
rial was packed in the cooler to prevent ali-
quots from shifting during shipping.
A four-part shipping form for each cooler
was prepared. The shipping form (Figure 10)
identified the aliquots in the cooler. Two copies
of this form were included in the cooler. One
copy was sent from the field laboratory to the
pilot survey sample management office (Viar
and Company, Alexandria, VA), and one copy
was retained in the field laboratory. One copy
of the form received at the contract laboratory
was sent to the sample management office to
confirm the receipt of the aliquots. Aliquots
were shipped from the field laboratory to the
contract laboratory via overnight courier ser-
vice, Monday through Friday.
The coordinator also prepared copies of
field data forms, field laboratory data forms,
and shipping forms for delivery to the data
base manager at Oak Ridge National Labora-
tory (ORNL) and to quality assurance person-
nel at EMSL-LV. The coordinator also tele-
phoned the Las Vegas communications center
and provided a report on the day's sampling
activities (including number and identification
codes of streams visited, information on
sample shipment, requests for supplies, prob-
lems encountered, and subsequent corrective
actions).
26
-------
NATIONAL SURFACE WATER SURVEY
SAMPLE MANAGEMENT OFFICE
P.O. BOX 818 . NSWS
ALEXANDRIA, VA 22314 FORM 3
SHIPPING
FROM
(STATION ID):
SAMPLE
ID
01
02
03
04
OS
06
07
08
09
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
20
29
30
1
TO
(LAB|:
BATCH
ID
DATE PROCESSED
ALIOUOTS SHIPPED
(FOR STATION USE ONLY)
2
3
4
5
6
7
8
orrciucn BY
IF INCOMPLETE IMMEDIATELY NOTIFY:
SAMPLE MANAGEMENT OFFICE
(703) 557-2490
PACSC nr
DATE SHIPPED DATE RECEIVED
AIR-BILL NO.
SPLITS
SAMPLE CONDITION UPON LAB RECEIPT
(FOR LAB USE ONLY)
QUALIFIERS:
•J: ALIQUOT SHIPPED
M: ALIQUOT MISSING DUE TO DESTROYED SAMPLE
WHITE - FIELD COPY P1NK-LABCOPY YELLOW — SMO COPY GOLD — LAO COPY FOR RETURN TO SMO
Figure 10. Shipping form, National Stream Survey-Pilot Survey.
27
-------
Section 6
Evaluation of Equipment and Protocols
The methods manual for the National
Stream Survey (Hillman et al., 1987), describes
all analytical methods used during the pilot
survey in the contract analytical laboratory and
the field laboratory. The contract analytical
laboratory equipment and methods and the
majority of the field mobile laboratory equip-
ment and methods had been used successfully
during Phase I of Eastern Lake Survey in 1984
(Linthurst et al., 1986).
A major objective of the pilot survey was
to evaluate methods and equipment for collec-
ting and processing stream water samples.
Methods evaluated included all field proce-
dures and two new field laboratory procedures
that were not used in previous NSWS studies.
All field equipment and the equipment required
for the two new laboratory methods were also
tested. In addition, two holding time studies
were conducted. Results of these studies
were used to assess the chemical stability of
water samples over time. The following sec-
tions describe the evaluations conducted
during the pilot survey and present recom-
mendations based upon evaluation results.
The holding time studies are described in other
documents that are appended to this report.
Field Laboratory Methods
Evaluations
Field Laboratory Conductivity
Measurements
The portable conductivity meter used
during the pilot survey does not have a high
resolution at the low conductivities (<20
fjiS/cm) observed at many stream sites. During
the final days of the pilot survey, a YSI Model
32 laboratory conductivity meter and 3310
• probe were used to evaluate the feasibility of
conducting conductivity measurements on
stream water samples in the field laboratory.
Because the number of samples measured by
using this procedure was small, the evaluation
results must be interpreted cautiously. The
few measurements that were made, however,
indicated that accurate measurements of
conductivity could be obtained in the field
laboratory without adversely affecting other
aspects of laboratory operations. Measuring
conductivity at the field laboratory would
provide an accurate measurement of conduc-
tivity on samples within 12 hours of their
collection.
Preparation of Organic Monomeric
Aluminum Fraction
The preparation of this fraction involved
passing a filtered aliquot of sample through a
cation-exchange column. After the ion-
exchange process, the aliquot was subjected
to the aluminum extraction protocol used in
NSWS projects (Hillman et al., 1987) to prepare
a fraction for analysis of total extractable
aluminum.
The ion exchange resin (Amberlite 125)
was conditioned before use to bring the pH of
the resin column within 0.5 pH unit of the
expected sample pH. Columns were condi-
tioned by adjusting a 1 x 10~5 N NaCI solution
to the desired pH with HCI or NaOH. This
adjusted solution was passed through the
resin column, and the pH was measured on an
aliquot of the collected eluant. This process
was repeated until the desired pH of the resin
column was achieved.
A 125-mL aliquot of sample was filtered
into a 250-mL Pyrex beaker that had been
washed with 5 percent HN03 and had been
28
-------
rinsed with deionized water. The sample in the
beaker was pumped through the ion-exchange
column at a rate of 30 mL/min. The first 30 ml_
of sample collected from the column was
discarded. The next 20 ml was collected, and
its pH was measured. Three 5- to 10-mL
portions of the sample were used to rinse a
50-mL polycarbonate centrifuge tube. The next
25.0 ml_ of sample was measured as accu-
rately as possible (±0.1 mL) into the centrifuge
tube. The column was then flushed with the
adjusted NaCI solution, an aliquot of the NaCI
solution was collected, and its pH was mea-
sured. This measurement ensured that the
column was conditioned properly for the next
sample. The aluminum extraction procedure
was then performed on the aliquot of sample.
Several aspects of the methodology merit
attention by future workers. First, adjusting the
pH of the NaCI solution was time-consuming,
and the adjusted solution was not stable over
time. Allowing the solution to equilibrate with
the atmosphere overnight before adjusting the
pH sometimes improved the stability of the
solution pH. During the pilot survey, three or
four different columns were prepared daily to
cover the range of pH values. Second, a
standard solution having a known concentra-
tion of organic monomeric aluminum was not
available for use as an audit sample to check
on the accuracy of the preparation procedure.
Third, the methodology as used in the field
laboratory was feasible; however, the addi-
tional equipment required crowded the work
area in the laminar flow hood. It is suggested
that other fractionation methods be investi-
gated for possible implementation in future
surveys.
Field Equipment Evaluations
The initial selection of field equipment
for routine use in the pilot survey was based
on the prior experience of field personnel and
on recommendations from researchers. Under
field conditions, the performance of this equip-
ment was compared to the performance of
several alternative instruments. Several pH,
conductivity, and dissolved oxygen meters
were evaluated on the basis of comparability
in measurements, comparability with labora-
tory measurements, ease of use, portability,
and overall durability. These evaluations
confirmed that the equipment used during the
survey was satisfactory in meeting the needs
of the NSS. Results of these comparisons are
available as an unpublished memorandum
report from the Acid Deposition Department,
Lockheed-EMSCO, Las Vegas, Nevada.
Field Methods Evaluations
Field protocols developed for the pilot
survey were based on analytical methods
described in the NSS methods manual and
upon recommendations of researchers and
instrument manufacturers. Potential modifi-
cations of these methods were evaluated
during field sampling operations. Some modifi-
cations were adopted immediately (e.g., pH).
Other modifications (e.g., sample holding
times) were evaluated experimentally for use in
subsequent NSS or NSWS programs.
In-line Filtration at Streamside
The filtration of samples during the
collection process was tested before the start
of pilot survey field operations. Filtration in
the field, if proven feasible, would minimize the
potential deterioration of samples before their
delivery to the field laboratory. A cartridge
filtration unit that used disposable membrane
filters (47 mm dia., 0.45-/;m pore size Gelman
Met rice I) was fitted into the Tygon pump
tubing. The filtrator unit was tested on both
the suction and the discharge sides of the
peristaltic pump.
The evaluation was conducted over a
3-day period during which samples were col-
lected from 16 streams. Difficulties in perform-
ing the operation were identified as (1) a high
potential for sample contamination during filter
replacement or as a result of filter rupture, (2)
the increased time required at streamside to
conduct filtration operations, (3) an increased
load on the pump which required that the
battery be recharged after only one site, and
(4) an increased requirement for rinse water
and other supplies. The consensus was that
filtration of samples in the field using the
equipment selected was not practical.
Sample Collection Method
During normal sampling operations,
syringe samples were pumped from the
29
-------
stream by using a portable peristaltic pump
and Tygon tubing. This method was evaluated
against the alternative of collecting samples
directly from the stream into the sample con-
tainer. The parameters of concern in syringe
samples were pH and DIG concentration. Both
of these parameters can be altered by direct
contact with the atmosphere. A two-way
analysis of variance indicated that there was
no significant difference (or = 0.05) between
the collection methods on either pH or DIG
measured at the field laboratory. After subjec-
tive evaluation of the methods, pumping was
identified as the preferred method because
samples can be collected in a consistent
manner from streams of all sizes.
Two methods of stream side pH measure-
ment were used during the pilot survey. The
closed-system method was designed to mea-
sure the pH of a sample that had not been
allowed to come into contact with .the atmos-
phere. The open-system method, which does
allow atmospheric contact, is a more conven-
tional approach. Both methods were used at
each stream throughout the entire study.
An experiment was conducted to (1) eval-
uate the comparability of these two methods
and (2) to compare the pH measurements of
samples collected by using a pump with the
pH measurement of samples collected directly
from a stream. Three replicate samples of
each treatment combination (method x col-
lection device) were measured at each of three
streams (12 samples per stream). The mea-
surements were compared by using a two-way
analysis of variance. No significant differences
were detected (a = 0.05); however, the
open-system method required much less
equipment and was simpler to.conduct.
Severa| devices were constructed and
were tested in an effort to measure pH in situ.
The two most promising of these were (1) a
set of nested beakers with the pH electrode
and ATC probe suspended in the inner bsaker
and (2) a short section of PVC pipe with the
electrode and ATC probe inserted through two
openings near the center of the pipe. Mea-
surements made with these devices were
compared with both field pH measurements.
A two-way analysis of variance revealed no
significant differences among methods (a =
0.05).
The results of this test indicate that the
streamside methods used in the pilot survey
yielded pH values that were comparable to in
situ measurements. Streamside measure-
ments proved to be more efficient because-
they require less equipment, are simpler to
perform, and do not require that the fieid
sampler enter the water.
i. Tom®
. Current sample handling procedures used
in all NSWS programs involve processing and
preserving raw samples at the field laboratory
within a specified holding time (i.e., within 12
to 15 hours of sample collection). Future
NSWS projects may require longer intervals
between sample collection and subsequent
stabilization and preservation. Two separate
studies were conducted to evaluate the effect
of holding time on sample chemical composi-
tion. One study involved samples collected
and stored in sealed syringes. The other
involved samples collected and stored in 4-L
Gubitainers.
The details and results of the two
studies are presented in other documents that
are appended to this report. The result of
these studies indicated that syringes could bs
held for up to 7 days under conditions similar
to those during the pilot survey without a
significant change in either pH or dissolved
inorganic carbon. Cubitainers could be held
under conditions similar to those encountered
during tha pilot survey for as long as 84 hours
before processing and preservation without a
significant change in any of the chemical
parameters measured.
30
-------
Section 7
Summary of Field Operations
The pilot survey was completed on
schedule on July 17, 1985. The down stream
node of each of the 61 stream reaches was
sampled four times during the survey. Up-
stream nodes of all reaches were sampled at
least once (during the summer sampling
period). Each of the three teams visited
approximately 20 streams during a 2-week
(10-day) sampling period. Fifty-seven days of
operation (not including days off in the field)
were required to complete the sampling.
A total of 759 samples (routine, field
duplicate, field blank, and audit samples) were
processed by the field laboratory in Sylva, NC,
and were shipped (as 6,072 different aliquots)
to the contract analytical laboratory. Only one
shipment was delayed: the courier service
did not deliver samples to the analytical
laboratory on a Saturday.
The preliminary development of site
dossiers and the reconnaissance of each
sampling site were essential to the successful
completion of the pilot survey. Problems
associated with locating and gaining access to
sampling sites were discovered and were
resolved before actual sampling commenced.
A great deal of time was saved, and the
possibility of not obtaining samples from
designated streams was minimized. The use
of local cooperators expedited the reconnais-
sance process.
During spring sampling, which was
conducted primarily at downstream sites,
teams were in the field for an average of 8.8
hours per day. The average time consumed in
a full day of sampling activities during the
summer sampling period was 14 hours be-
cause upstream and downstream sites were
sampled at all streams. The time spent on
each sampling activity performed during the
summer survey is given in Table 5.
Table 5. Sampling Activity Tim* Summary for Summer
Sampling Period, National Stream Survey-
Pilot Survey
Activity
Time Summary (hours)
Range Average
Calibration and preparation
First stream, upstream site
Drive to access point
Hike to site
Sample
Hike from site
First stream, downstream site
Drive to access point
Hike to site
Sample
Hike from site
Second stream, upstream site
Drive to access point
Hike to site
Sample
Hike from site
Second stream, downstream site
Drive to access point
Hike to site
Sample
Hike from site
Return to Field Station
Final calibration and cleanup
0.5 - 3.0
0.5 - 4.0
0.0 - 4.0
0.5 - 1.5
0.0 - 4.0
0.25 - 1.0
0.0 - 4.0
0.5 - 1.5
0.0 - 4.0
0.5 - 3.0
0.0 - 0.5
0.5 - 1.5
0.0 - 0.5
0.25-
0.0 -
0.5 -
0.0 -
1.0
0.5
1.5
0.5
0.0 - 4.0
0.5 - 1.5
Total
1.5
2.0
0.25
0.75
0.25
0.5
0.25
0.75
0.25
1.5
0.25
0.75
0.25
0.5
0.25
0.75
0.25
2.0
-LP_
14.00
Approximately 45,000 vehicle miles were
traveled by the three field teams during field
sampling activities. Additional miles by foot,
boating, or horse back were required to
access some sampling sites. The breakdown
31
-------
of the total miles required by each field team
to complete sampling activities is presented in
Table 6.
Table 6. Summary of Distance Traveled by Field
Sampling Teama, National Stream Survey-
Pilot Survey
Mode of Travel
Automobile
Boat
Horse
Foot
Daily Miles Traveled
oer Team
Range Average
60 - 400 270
0-4 0.1
0 - 12 0.1
0 - 16 1.0
Total
Miles
Traveled
~ 45.000
16
12
160
On occasions when long hikes were
required to gain access to sampling points,
one or two people assisted the field sampling
team. The primary function of these additional
personnel was to carry some of the sampling
gear and samples. They were especially
useful during the summer session when
upstream and downstream samples were
taken on the same trip.
The methods evaluations, equipment
evaluations, and experiments conducted during
the pilot survey provided information that will
be used to modify the proposed research plan
for NSS-Phase I operations. The pilot survey
demonstrated that field operations could be
conducted in such a manner that samples and
field data from streams over a large geo-
graphic area could be collected and analyzed
consistently and quickly.
Selected cost estimates for the pilot
survey are presented in Table 7. These costs
reflect equipment and other supplies necessary
to collect and process samples. These esti-
mates are provided as an aid to planning
studies of similar scope and purpose. Addi-
tional costs related to personnel support and
to operating a field station are not included.
The cost to equip a team to conduct field
sampling activities (visit 20 streams four times
each) was between $2,500 and $3,000. This
estimate includes purchasing the dissolved
oxygen meters which for the pilot survey were
borrowed rather than purchased. A set of
sample containers (including Cubitainers,
syringes, and aliquot bottles) cost about $20
for each routine, field blank, or field duplicate
sample collected.
The equipment and instrumentation in
the field laboratory cost approximately $40,000
(Morris et al., 1986). This estimate includes
expendable supplies used in the laboratory
during the pilot survey (approximately $5,000).
Because the field laboratory and its equipment
had been purchased for Phase I of the East-
ern Lake Survey, this cost was not part of the
pilot survey budget. A field laboratory costs
about $20,000 to construct (Morris et al., 1986).
Containers for preparing eight preserved
aliquots from each sample cost about $13 per
sample. All containers used during the survey
were subjected to a rigorous cleaning proce-
dure (Hillman et al., 1987) that cost about $30
per sample (including field and laboratory
containers). Preserved aliquots were shipped
via overnight courier service to the contract
analytical laboratory at a cost of about $100
per container (6 to 8 sets of aliquots). Over-
night service was necessary to assure that
aliquots would arrive at the analytical labora-
tory within 48 hours of collection to meet
sample holding times.
32
-------
Table 7. Selected Cost Estimates', National Stream Survey-Pilot Survey
Cost (dollars)
Field Sampling
Vehicle Rental 1,000 per team
(monthly)
Field Equipment:
pH meter and electrode 600 per team
Conductivity meter and probe 540 per team
Dissolved oxygen meter and probe 900 per team
Staff gauges 40 per team
Peristaltic pumps and batteries 500 per team
Sampling Containers:
4-L Cubitainers 3 per sample
60-mL syringes 15 per sample
500-mL bottles 2 per sample
Field Laboratory and Analytical Support:
Instrumentation, equipment, and supplies 40,000*
Aliquot containers 13 per sample
Bottle cleaning 30 per sample
Shipping cost (overnight courier service) 100 per container
(6-8 samples)
• Based on estimates provided by personnel responsible for procurement or quality assurance.
b Field laboratory and its instrumentation was available from the Eastern Lake Survey. Necessary supplies for the pilot
survey were approximately $5,000.
33
-------
This section summarizes the problems
experienced during the pilot survey and pro-
vides associated recommendations in tabular
form. The information is presented in four
categories: field safety, planning activities,
field sampling operations, and field laboratory
operations.
Table 8 summarizes problems associated
with field safety. The daily itinerary form,
which provided the field station with a record
of the proposed routes of travel and estimated
time in the field, was very useful. Portable
radios would have improved communications
in some areas where telephones were not
convenient. Many field samplers complained
of fatigue from driving the same roads day
after day. A possible solution is to rotate
samplers among teams such that a team
would always contain one member who is
familiar with the assigned sampling sites and
their access.
Preliminary site reconnaissance, as
stated previously, was necessary to facilitate
pilot survey field operations and to ensure that
the objectives of the survey were achieved.
Specific recommendations are listed in Table 9.
In addition, future NSS projects should empha-
size (during training) consistent guidelines for
selecting a sampling location at a given
stream site.
Recommendations concerning field
sampling operations are listed in Table 10.
Recommendations related to data quality
issues are also summarized in the following
sections.
Field samplers should receive detailed
training in basic hydrology, including the use
and placement of staff gauges. Sampling
containers and disposable collection equip-
ment should be prepackaged into kits for each
site to ensure that supplies are not forgotten
and to reduce potential contamination during
sample collection. The feasibility of reusing the
pump tubing should be investigated with
respect to potential contamination. Reusing
tubing after rinsing with deionized water or
sample water before collection would be much
less expensive and would save space and
weight during field excursions.
Protocols for measuring pH in the field
during the pjlot survey were designed to pro-
vide pH measurements that were equivalent in
accuracy and precision to field laboratory
measurements. It should be stressed that
care of the pH meter and of the electrodes in
the field is critical to the success of the proto-
col. The etching procedure recommended by
the manufacturer should be used to clean the
ceramic junction of the electrode, if required.
The method involves removing the filling
solution from the electrode and refilling it with
deionized water; this is followed by immersing
and stirring the electrode in a 50 percent
(weight/volume) solution of sodium hydroxide
for 3 minutes. The electrode is then drained,
is rinsed twice with a 3 N KCI solution, is
rinsed in pH 7.00 buffer for 2 minutes, and is
refilled with 3 N KCI. This procedure improved
the response and stabilization time of the
electrode.
Measurement values of in situ conduc-
tivity may have been affected because, for the
34
-------
Table 8. National Stream Survey-Pilot Survey Probleme and Recommendatlone: Field Safety
Problem
Solutions Employed
Recommendations
Vehicle identification.
Lack of radio communication.
Failure to meet call-in protocol.
Lack of safety protocol specifi-
cally relevant to storm epi-
sodes.
Lack of adequate shelter during
storm events.
Driver's fatigue.
None used.
Daily itinerary filed by all teams.
Teams required to call-in at pre-
determined times.
Emphasized importance of call-
in to team members.
Added 1/2 hour delay from time
of required call-in to initia-
tion of search and rescue.
None used.
Standard supply included:
One space blanket per per-
son.
One raincoat per person.
One pair of sturdy, water
repellent shoes per
person.
Switched driving duties among
team members.
Driver education during training.
Use vehicle identification placard to identify a
vehicle as being used for the NSS in
remote locations.
Investigate cellular or mobile radio units for all
sampling vehicles.
Require call-in according to protocol.
Require daily itinerary.
Emphasize call-in during training.
Set flexible time for call-in for teams traveling
to remote stream site.
Develop safety protocol for teams isolated
during storm events.
Supply the following gear to each person:
Space blanket.
Rain coat.
Rain pants.
Sturdy, water repellent shoes.
Sleeping bag.
Supply the following gear to each team:
Lightweight tent.
Shelter half or rain fly to protect sampling
gear.
Discuss fatigue in driver training course.
Have long-haul truck driver teach methods of
reducing fatigue.
Shorten driving duties.
Provide driver with mechanical drowsiness
detector.
Switch driving duties.
Table 9. National Stream Survey-Pilot Survey Problems and Recommendations: Site Reconnaissance and Acctts
Problem
Solutions Employed
Recommendations
Inadequate maps (1:24.000 and
1:250,000 scale maps were
often 30 years old).
Inadequate sample site selec-
tion criteria.
Site access.
Provided several map sources.
Training addressed generalities
of site selection.
Permission was acquired to gain
access to a site.
If permission could not be
acquired, a new site was
selected.
Provide as many map types as possible,
including:
USGS Topo maps.
State transportation maps.
County road maps.
USFS road maps.
Fishing and hunting maps.
Provide the most recent maps of each type
available.
Use USGS or Fish and Wildlife service person-
nel to train sampling personnel and to
emphasize site selection during training.
Visit several stream sites and identify good
and poor sample locations.
Use local ccoperators at all sites to assist in
gaining access permission.
During training, stress friendliness to land
owners and willingness to communicate
about the need for access.
35
-------
Table 10. National Stream Survey-Pilot Survey Problem* and Recommendation*: Field Sampling Activities and
Equipment
Problem
Solutions Employed
Recommendations
Unclear objectives for stream
stage measurement.
Poor stream gauge placement.
Loss of stream gauges due to
vandalism and high flows.
Excessive battery discharge in
cold weather and during
in-line filtration experiments.
Tripped pump circuit breaker in
cold weather and in-line
filtration experiments.
Faulty or loose battery cable
connections.
Higher than expected consump-
tion of surgical grade Tygon
tubing, syringe, Cubitainers, syr-
inge valves, and 500-mL aliquot
bottles.
Incomplete sample collection
supplies taken into field by
teams.
Special experiments and studies
required large amount of
supplies.
Failure of 5 out of 9 portable pH
meters in field.
Developed protocol for collection
of minimal information.
Attempted to place staff at
neutral location so staff
could be read accurately at
all stage heights.
Established secondary reference
marks on permanent fea-
tures for future re-establish-
ment of staff base height.
Recharged battery daily.
Carried spare batteries.
Batteries were kept as warm as
possible in transit.
Waited for circuit breaker to
cool and reset.
Purged tubing for short period
before taking sample.
Inspect and tightened terminals
periodically.
Removed faulty terminals.
Carried replacement cables.
Inventory was checked weekly,
supplies were ordered well
in advance of need.
Reduced length of tubing used
for blank samples.
Cleaned and reused 500-mL ali-
quot bottles for suspended
solids.
Pre-made kits were assembled
by teams in advance of
demand (allowed advance
warning of short supplies).
Developed daily checklist of sup-
plies.
None.
Returned faulty meters as fail-
ures occurred.
Provided alternative field pH
meter for times when ade-
quate numbers of primary
meters were not available.
Carried meters in carrying cases
or vest to minimize shock.
Clarify objectives of hydrologic data collection.
Design data collection methods to fit
needs.
Emphasize staff placement in training.
Visit several streams with an experienced
hydrologist to locate good placements and
illustrate poor placement.
Establish all sites with a secondary reference
on first visit.
Carry extra, fully charged batteries.
Do not filter samples at streamside.
Use thin-walled pump tubing through pump
head (may require reuse of pump tubing).
Heavier duty pump needed for hard pumping.
Thin-walled Tygon tubing would be useful,
at least through pump head. May require
reuse of a short section of pump tubing.
Do not filter samples at streamside.
Carry replacement cables.
Solder terminals onto wire.
Provide replacement banana plugs.
Estimate number of samples prior to entering
field.
Order sufficient supplies for all samples well
in advance of start-up date for sampling.
Supply 10 percent minimum overage for unex-
pected consumption related to special
samples and contaminated or damaged
containers and supplies.
Prepare kits for at least one week of sampling
in advance of need.
Maintain inventory weekly.
Use daily checklist before departing for field
sites.
Plan special studies well in advance and allow
for advance supply of materials.
Allow 10 percent minimum overage for waste
in all experiments.
Maintain a complement of two functional
meters per team.
Protect meters from shock in transport and
use by carrying them in cases carried near
sampler's chest.
(continued)
36
-------
Table 10. (continued)
Problem
Solutions Employed
Recommendations
Battery failure and lack of
replacement batteries for pH
meters.
Bubbles in coils of pH electrode.
Slow pH electrode response.
pH electrode breakage.
Slow pH electrode response for
stream samples and appar-
ent hysteresis following
QCCS measurement.
Thermal inequality between pH
chamber and ATC bath.
Poor comparability between field
and lab pH measurements.
Stream conductivity below cali-
bration and QCCS range.
Poor conductivity probe position
while calibrating and
sampling.
Back-up lithium batteries were
provided as quickly as they
could be supplied.
Kept one full set of replacement
batteries for all meters.
Tapped electrode to remove
bubbles.
Carried electrodes in vertical
position.
Twirled electrodes by cable prior
to reading at each site.
Replaced epoxy bodied elec-
trodes with glass-bodied
electrode.
Etched glass-bodied electrodes
periodically with NaOH.
Carried electrodes in upright
position in carrying case or
vest pocket.
Back-up electrodes were carried
by each team and also kept
at the field station.
Refined protocol to allow longer
electrode equilibration time
in pH measurement.
Attempted to develop low ionic
strength, circumneutral
QCCS for pH.
Attempted to design chamber
that allowed ATC to fit in
chamber with electrode.
Sheltered ATC bath and pH
chamber from direct sunlight.
Modified field protocol to:
Improve calibration.
Provide stringent criteria for
stability of a reading.
Closely mimic field lab pro-
tocol.
Eliminate dependence on
auto-lock feature of
portable pH meter.
Tested low conductivity (14.7
pS/cm) QCCS in field.
Adjusted cord length to keep
probe off bottom of calibra-
tion bottle, but submerged
in standard while calibrating.
Required probe to be positioned
under surface and in line
with current while sampling.
Supply at least one functional set of backup
batteries with each meter.
Maintain a store of one additional set of
batteries for each meter at a central
location.
Twirl electrode by cable before reading at
each site.
Carry electrode in padded case and upright
position.
Use glass bodied electrode. Etch glass-
bodied electrodes as necessary with
NaOH.
At least 2 back-up electrodes should be avail-
able for each team in the field.
Adopt modified protocol:
Conditioning soak followed by successive
aliquots measured to stability within
and among aliquots.
Continue to use pH 4.01 HjSC^ QCCS.
Conduct open pH measurement only.
Continue to develop chamber design to com-
pensate for temperature change in small
sample.
Utilize field procedures which simulate field
lab protocol (Hillman et al. 1987).
Do not use auto-lock features of pH meters
designed to automatically determine
stability of pH measurements.
Continue to use 74 pS/cm QCCS at each site.
Use a 14.7 fjS/cm QCCS as a low range
QC check at calibration room daily.
Set up probes to facilitate calibration in
250-mL bottles.
Standardize subsurface and in situ position of
probe for sampling.
(continued)
37
-------
Tabl* 10. (continued)
Problem
Solutions Employed
Recommendations
Poor resolution of field meter at
low end of conductivity
scale.
Lack of expertise in maintenance
of dissolved oxygen meters.
Unclear qualitative data
recorded on watershed
characteristics form.
Some pertinent information was
not requested on field data
forms.
Illegible forms, difficult key
punching.
Switched to meter with multiple
range capabilities for sum-
mer sampling.
Replaced membrane every other
week.
Developed expertise with main-
tenance procedures.
Attempted to train all teams to
record data in a consistent
manner. .
None.
Interacted with data manage-
ment personnel frequently to
clarify problems.
Repetitious data required on
various forms.
None.
Lower range of meter used in summer worked
well, should evaluate low range instrument
under field conditions.
Measure conductivity at field laboratory.
Include maintenance procedures in training
program.
Practice probe maintenance several times in
lab environment before attempting it in the
field.
Give numerical guidelines for all qualitative
data: e.g., sparse = 5 individuals visible,
or sparse = 5 percent of total.
Emphasize during training.
Clarify purpose of this
Add time, gauge height, and flow velocity at
arrival and at departure on field data form.
Add percentage canopy cover to watershed
characteristic form.
Add space to identify duplicate, blank, and
other special samples on field data form.
Add box for "special samples taken," "not
sampled,0 and boxes to clarify tracking of
experiments, episode samples, duplicates,
and blanks.
Use standard units for stream width, depth
(don't mix feet and meters).
Use standard qualifier to denote when a
QCCS check failed.
Use leading or training zeros to fill all data
blocks in a field.
Standardize shape of numbers and letters
used on forms.
Add hundredths digit to gauge height entry
blank.
Clarify "±" in dissolved oxygen block.
Replace "Replicate Number" with "Site Visit
Number" on Form 4.
Train field personnel' on data entry considera-
tions.
Train field personnel on proper use of com-
ments and data qualifiers.
Redesign forms to eliminate repetitious infor-
mation.
range of conductivities encountered (generally
<50 juS/cm), the meter was not easily read.
The meter that was most reliable under field
conditions had a poorer resolution at conducti-
vities less than 40 juS/cm. Meters having
greater resolution that were tested were not
suitable for use under field conditions. A
low-conductivity QCCS should be used in the
field (e.g., 5 x 10"4 N KCI solution having a
specific conductance of 14.7 pS/cm at 25 C),
and conductivity should be measured in the
field laboratory to provide an accurate meas-
urement of this parameter within a short time
after sample collection.
Field Laboratory Operations
Problems and recommendations regard-
ing field laboratory operations are summarized
in Table 11. Dust in the laboratory was a
problem at the beginning of the pilot survey,
38
-------
Table 11. National Stream Survey-Pilot Survey Problems and Recommendations: Field Laboratory Operations
Problem
Solutions Employed
Recommendations
Feed water to the laboratory
was of poor quality, affected
performance of the reverse
osmosis system.
Laminar flow hood not func-
tioning efficiently.
Dust in laboratory.
Condensation coming into lab
via A/C units.
Laminar flow hood crowded with
reagents and glassware
used for filtration, aluminum
extractions.
Preparation time for organic
monomeric aluminum aliquot
excessively long.
Lack of audit sample or QCCS
having known concentration
of organic monomeric alumi-
num.
50-mL centrifuge tubes often
leaked during MIBK extrac-
- tion procedure.
Teflon O-ring in MIBK aliquots
dissolved into solution.
Turbidity readings fluctuated
because of presence of
different size particles.
Lack of a QCCS <2 NTU.
Lack of acceptable QCCS for
. total suspended solids.
Testing of pH after preservation
required large numbers of
pipette tips.
Aliquots sometimes preserved
with wrong acid.
Changed carbon prefilters every
other day.
None.
Moved laboratory away from
heavy traffic area.
Units operated during process-
ing only when necessary.
Prepared aliquots of organic
monomeric aluminum first.
Analyst 2 assisted with prepar-
ing aluminum extraction
.extracts after filtration was
completed.
Processed samples through col-
umn in order of decreasing
pH.
Analyst 2 assisted in prepara-
tion and extractions.
None.
Tubes were inspected during
rinsing- and again before
shaking.
Use of sealing O-rings was
First reading on nephetometer
was recorded.
5 NTU standard was diluted 1:3
to make a 1.25 NTU (Labo-
ratory prepared solution
yielded acceptable readings
for 3-4 days.)
Diatomaceous earth suspension
used was not adequate.
Nonheparinized glass capillary
tubes used to test pH.
Labels color-coded as to type of
acid required.
Install sediment filter in front of carbon pre-
filter to remove large paniculate material.
Inspect blower unit for proper wiring and
installation.
Locate the laboratory away from areas of
traffic and dust, or locate the laboratory
indoors in a controlled environment (e.g.,
warehouse).
Investigate use of replaceable filters on units,
or locate laboratory indoors in controlled
environment. Provide outside drain for
condensation. °
Conduct all filtering first, then use entire work
area for aluminum extractions.
Investigate alternative methods for preparing
or analyzing dissolved aluminum fractions.
Establish pH criteria for preparing organic
monomeric aluminum aliquots.
Develop and .test appropriate standard solu-
tion.
Investigate alternative methodologies.
Use only 50-mL Coming tubes with orange
screw-top sealing caps.
Use different type of container or obtain O-ring
of solvent-resistant material.
Standardize protocol for turbidity measure-
ment.
Use fresh standard that is certified.
Acceptable standard must be developed.
Continue to use capillary tubes.
Continue to use colored labels.
39
-------
but the problem was corrected by relocating
the laboratory trailer a short distance from its
original location.
The supply of reagent grade water in the
field laboratory was barely adequate. The
daily requirements of three sampling teams for
reagent grade water plus the daily require-
ments of the laboratory were barely satisfied
with the existing reverse osmosis system that
produced about 4-L per hour. The quality of the
field water to the laboratory was also poor;
this required that carbon pref ilter cartridges be
changed every 48 hours. During the pilot
survey, a fibrous pref ilter cartridge system was
installed in front of the carbon prefilter cart-
ridge to filter the field water before it entered
the reverse osmosis system. This filter ex-
tended the life of the purification cartridges. If
future NSS operations require that a field
laboratory support more than three sampling
teams or if daily sample loads per laboratory
are increased, the present water system would
not be sufficient.
Audit samples or other QC samples
should be developed for the organic mono-
meric aluminum fraction and for total sus-
pended solids analysis if these parameters are
to be measured in future NSS operations.
Alternate methodologies • for fractionating
dissolved aluminum species should be eval-
uated.
40
-------
References
Altshuller, A. P. and R. A. Linthurst (eds.).
1984. The Acidic Deposition Phenomenon
and Its Effects. Critical Assessment
Review Papers. Vol. II: Effects
Sciences. EPA 600/8-83/016 BF. U.S.
Environmental Protection Agency, Wash-
ington, D.C.
Baker, J. P., and C. L. Schofield. 1982. Alumi-
num toxicity for fish in acidic waters.
Water, air, and Soil Poll. 18:289-309.
Driscoll, C. T., J. P. Baker, J. J. Bisogni, and
C. L. Schofield. 1980. Effect of alumi-
num speciation on fish in dilute acidified
waters. Nature 284:161-164.
Driscoll, C. T. 1984. A procedure for the
fractionation of aqueous aluminum in
dilute acidic waters. Int. J. Environ.
Anal. Chem. 16:267-283.
Drouse, S. K., D. C. Hillman, L. W. Creelman,
and S. J. Simon. 1986. National Surface
Water Survey, National Stream Survey
(Pilot, Mid-Atlantic - Phase I, Southeast
Screening, and Mid-Atlantic Episode-Pilot)
Quality Assurance Plan. EPA
600/4-86/044. U.S. Environmental Protec-
tion Agency, Las Vegas, Nevada.
Hillman. D. C., J. F. Potter, and S. J. Simon.
1986. National Surface Water Survey,
Eastern Lake Survey (Phase I-Synoptic
Chemistry) Analytical Methods Manual.
EPA 600/4-86/009. U.S. Environmental
Protection Agency, Las Vegas, Nevada.
Hillman, D. C., S. J. Simon, and S. Pia. 1987.
National Surface Water Survey, National
Stream Survey (Pilot, Mid-Atlantic-Phase
I, Southeast Screening and Mid-Atlantic
Episode-Pilot) Analytical Methods Manual.
EPA 600/4-86/044. U.S. Environmental
Protection Agency, Las Vegas, Nevada.
Linthurst, R. A., D. H. Landers, J. M. Eilers, D.
F. Brakke, W. S. Overton, E. P. Meier, and
R. E. Crowe. 1986. Characteristics of
Lakes in the Eastern United States,
Vol. I: Population Descriptions and
Physico-Chemical Relationships.
EPA-600/4-86/007a. U.S. Environmental
Protection Agency, Washington, D.C. In
press.
Messer, J. J., K. N. Eshleman, S. M. Stam-
baugh, and P. R. Kaufmann. 1986.
National Surface Water Survey: National
Stream Survey-Phase I Pilot Survey. EPA
600/04-86/023. U.S. Environmental Pro-
tection Agency, Washington, D.C.
Morris, F. A., D. V. Peck, M. B. Bonoff, and
K. J. Cabbie. 1986. National Surface
Water Survey, Eastern Lake Survey-
Phase I Field Operations Report. EPA
600/4-86/010. U.S. Environmental Protec-
tion Agency, Las Vegas, Nevada. In
press.
U.S. EPA. 1983 (revised). Methods for Chemi-
cal Analysis of Water and Wastes. EPA
600/4-79/020. U.S. Environmental Protec-
tion Agency, Cincinnati, Ohio.
41
-------
Appendix A
Syringe Sample Holding Time Study
by
Eileen M. Burke and D. C. J. Hillman
Lockheed Engineering and Management Services Company, Inc.
Las Vegas, Nevada
Introduction
In the various programs within the
National Surface Water Survey (NSWS), water
samples are analyzed for pH and dissolved
inorganic carbon (DIG) in such a way that the
sample is not exposed to the atmosphere
during collection or analysis. These "closed-
system" measurements are required because
changes in the dissolved CO2 concentration
can significantly change these parameters in
samples from dilute waters having little or no
buffering capacity. Samples are collected in
syringes, sealed, and analyzed at a field labo-
ratory within 12 to 18 hours of collection.
Proposed changes in the logistical plans for
future NSWS activities (e.g., accessing lakes or
streams from the ground rather than by heli-
copter) will require longer holding times for
syringe samples before closed-system mea-
surements can be made. This study investi-
gated whether longer holding times would
significantly alter the pH or DIG of dilute water
samples stored in sealed syringes.
Three factors were initially considered to
be significant in affecting pH or DIG in a
syringe sample: biological activity, chemical
activity, and permeation of CO2 across the wall
of the syringe. Biological and chemical activity
are assumed to be negligible because of the
holding conditions employed. Samples are
stored in the dark at 4 *C until analysis. Thus
the permeation of CO2 is the important con-
sideration in how long syringe samples will
provide an accurate representation of in situ
conditions of pH and DIG. This study was
composed of four experiments designed to
investigate the stability of syringe samples at
different concentrations of dissolved CO2,
different holding temperatures, and different
levels of biological activity.
Analytical Methodologies and
Reagents
Dissolved Inorganic Carbon
Determinations . *
A Xertex Dohrmann DC-80 total carbon
analyzer equipped with a Horiba DIR-2000
infrared detector was used. The unit was
equipped with a 1-mL sample loop. The analy-
sis involved converting all of the DIG in a
sample aliquot into CO2 with 5 percent H3PO4,
and carrying the CO2 into the infrared detector
with nitrogen gas.
Deionized water meeting ASTM Type I
specifications (ASTM 1984) was used to pre-
pare all reagents and standards. A 1,000 mg/L
C calibration stock standard was prepared by
dissolving 8.825 g of over-dried, primary stan-
dard grade Na2CO3 in deionized water and by
diluting to 1.000 liter. A quality control stock
standard of 1,000 mg/L C was prepared inde-
pendently of the calibration stock solution. A
42
-------
10.00 mg/L C calibration standard was pre-
pared by diluting 5.00 ml_ of the calibration
stock to 500.00 ml_ with deionized water.
Quality control check samples (QCCS)
of 2.00 and 20.00 mg/L C were prepared by
diluting 1.00 mL and 10.00 ml of the quality
control stock solution to 500.00 mL, respec-
tively. Each of these solutions was trans-
ferred to Omnifit borosilicate glass reagent
bottles having three one-way valves. These
bottles were vented to the atmosphere through
tubes of CO2 absorbent (Mallcosorb) and
water absorbent (Aquasorb). The head-space
of each reagent bottle was purged with rea-
gent grade, CO2-free nitrogen gas (N2) for 30
minutes. The DIG standards were maintained
in these bottles under a CO2 free atmosphere
until analysis. The stock solutions were
purged with N2 before preparing any DIG
standards.
An Orion model 611 bench meter
equipped with an Orion Ross glass combi-
nation pH electrode (model 81-02) was used to
determine the pH of all samples. Closed-
system measurements of pH were accom-
plished by using a custom-made sample cham-
ber described in Hillman et al. (1986).
The meter was calibrated each day by
using commercially available NBS-traceable
buffer solutions (pH 7.00 and pH 4.00). A
quality control check sample having a theoreti-
cal pH of 4.00 was prepared by diluting 1.00
mL of a 0.100 N H2SO4 solution to 1.000 L with
deionized water. A pH 7.00 quality control
standard was prepared by diluting 0.27 mL of
a 1,000 mg/L C solution to 1.000 L and by
sparging with 300 ppm CO2 gas in air for 30
minutes.
All samples to be measured were col-
lected and stored in 60-mL polypropylene
syringes (Becton-Dickinson) and were sealed
with Dynatech Mininert syringe valves.
The DIG concentration of a standard or
sample was determined by injecting a 5-mL
aliquot into the sample loop. The instrument
was calibrated with the 10.00 mg/L C stan-
dard. The linearity of the calibration was
checked with the 2.00 and 20.00 mg/L C quality
control check samples and with a blank sam-
ple of deionized water. The calibration was
checked after at least every eight sample
measurements with the quality control check
samples. The sample loop was rinsed
between sample injections with 10 mL of
deionized water.
The pH meter and electrode were cali-
brated and checked initially by using solutions
contained in beakers. The pH 7.00 QCCS was
prepared fresh each day. The pH 4.00 QCCS
was prepared biweekly and was stored refri-
gerated when not in use. All syringe samples
were allowed to equilibrate with room temper-
ature before analysis. The sample chamber
was rinsed with deionized water. A syringe
containing sample was affixed to the chamber.
The chamber was then filled with sample and
was drained twice. The chamber was again
filled with sample, and the electrode was
inserted loosely into the top of the chamber.
An additional 5-mL was injected to eliminate
air bubbles. The chamber was sealed with the
electrode, and an initial pH and temperature
reading was noted. The pH of the aliquot in
the chamber was considered stable when the
reading changed less than 0.02 pH units over
a 1-minute period. After the initial pH reading
had become stable, a 5-mL aliquot was
injected, and the pH was determined. The pH
reading of this aliquot was recorded when
stable. Additional 5-mL aliquots were injected
and measured until the stable pH readings of
two successive aliquots were within 0.03 pH
units of each other. The reading of the last
aliquot to be measured was recorded as the
final pH reading for that syringe sample. The
meter was checked after every four to five
samples with the pH 4.00 quality control check
sample. At least one measurement of the pH
7.00 quality control check sample was made
each day.
Expeorinnieinit t
This experiment tested whether the pH
and DIC of a sample initially devoid of dis-
solved CO2 would change over time upon
exposure to the atmosphere.
A synthetic sample was prepared by
diluting 1.00 mL of a 1,000 mg/L C solution to
1 L. This solution was transferred to a
43
-------
reagent bottle and sparged with N2 for 30 min-
utes. Initial measurements of DIG and pH
(time = 0) were made on this solution by
collecting syringe samples directly from the
reagent bottle. The solution was transferred to
an open beaker and was covered with a watch
glass. Aliquots were collected from this beaker
into syringes, and measurements were made
at 22, 47.4, 119.1, and 168 hours after the initial
measurements.
This experiment tested whether the pH
and DIG of a sample initially supersaturated
with dissolved CO2 would change over time
upon exposure to the atmosphere.
A synthetic sample was prepared by
diluting 1.00 mL of a 1,000 mg/L C solution to
1 L. This solution was transferred to a reagent
bottle and was sparged with 300 ppm CO2 in
air for 30 minutes. Initial measurements of
DIG and pH were made on syringe samples
collected directly from the bottle. The
remaining solution was transferred to a beaker
and was covered with a watch glass.
Syringe samples from the beaker were
collected and analyzed at 24.2, 43.9, 118.3, and
165 hours after initial measurements were
made.
This experiment investigated the temporal
stability of pH and DIG in synthetic samples
representing different levels of initial dissolved
CO2 concentrations. The effect of temperature
on temporal stability of these samples was
also evaluated by holding syringe samples
at two different temperatures: refrigerated
(4-11 °C) and room temperature (21-26 °C).
A solution was prepared by diluting 10.00
mL of a 1,000 mg/L C solution to 10 L. This
solution was sparged with N2 gas for 45
minutes. This solution had no dissolved CO2
initially (pCO2 = 0 atm). Syringe samples were
collected from this solution, and initial pH and
DIG measurements were made (time = 0).
Sixty-four additional syringes were filled with
this sample (32 for pH, 32 for DIG measure-
ments). These were identified as "Level 0" sam-
ples.
The remaining solution was sparged
(with.stirring) with 300 ppm CO2 for 45 min-
utes. This solution initially had a dissolved
CO2 concentration in equilibrium with the
atmosphere (pCO2 = 0.0003 atm). Initial
measurements of DIG and pH were made on
this solution (time = 0). Sixty-four additional
syringes were filled with this sample and were
identified as "Level 1" samples.
Tha remaining solution was sparged
(with stirring) with 3,000 ppm CO2 gas for 45
minutes. This solution was equilibrated with
10 times the atmospheric concentration of CO2
(pCO2 = 0.0030 atm). Initial measurements of
pH and DIG were made (time = 0). Sixty-four
additional syringes were filled with this sample
and were identified as "Level 10" samples.
Finally, the remaining solution was sparged
with 30,020 ppm CO2 gas for 45 minutes. This
solution was equilibrated with 100 times the
atmospheric concentration of CO2 (pCO2
= 0.0300 atm). Initial measurements of DIG
and pH were made (time = 0), and sixty-four
additional syringes were filled with sample.
These were identified as "Level 100" samples.
For each parameter (pH and DIG), there
were 32 syringes from each CO2 concentration
level. Sixteen syringes for each parameter
from each group were held at room tempera-
ture (21-26 C). The remaining sixteen syringes
for each parameter were refrigerated (4-11 °C)
depending on the treatment level.
Four syringes from each leva! and tem-
perature group were analyzed for pH and DIG
at four times after initial measurements were
made at time = 0. The exact times varied with
the treatment level but ranged between 24 and
192 hours after time = 0. Instruments were
recalibrated on each day of analyses. At least
three measurements of QC check samples
were made at each time samples were ana-
lyzed. These quality control solutions were
prepared fresh each day, except for the pH
4.00 QCCS. In many cases, two aliquots from
a syringe were measured for pH and DIG.
This experiment investigated the stability
of pH and DIG in samples of natural lake
water held in sealed syringes. These samples
represented waters similar to those under
-------
study in the NSWS in terms of chemical com-
plexity and biological activity.
Water was collected from Padden Lake
and Bagley Lake, Washington. Samples were
collected from below the surface and were
stored in 4-L Cubitainers with no headspace
and were shipped to the laboratory. The
containers were kept sealed and at 4 °C until
use. For each lake, 52 syringes (26 for pH, 26
for DIG) were filled and sealed. Twelve
syringes from each lake were initially analyzed
for either pH or DIG. Twenty syringes from
each lake were held at 8 C, and twenty
syringes were held at room temperature (21-
26 °C). Two syringes from each lake were
analyzed for either pH or DIG at 13.0, 87.3,
108.9, and 170 hours after initial measurements
were made. Two syringes were labelled for
each sample. Measurements of pH and DIG
were made on each syringe, and the process
provided duplicate measurements for each
parameter.
Data collected for Experiments 1 and 2
were not analyzed statistically. Data from
Experiments 3, 4, and 5 were evaluated for
temporal changes in dissolved CO2 concentra-
tion by using the following approach.
Because samples from each time group
in a given experiment were analyzed on dif-
ferent days, any variation or differences obser-
ved can result from temporal changes in a
chemical parameter over time or from day-to-
day differences in the operation of an instru-
ment. These two sources of variation are
confounded, and comparing measurements of
samples among days by ANOVA or similar
analyses will not distinguish between the two
sources of variation.
A simplified approach was used for the
experiments conducted in this study. In a
given experiment, the means of measurements
made at each time interval were considered as
observations drawn from a population having
a grand mean estimated by X and a variance
about this mean estimated by s2(x). Holding
time in itself is not considered as a classifi-
cation factor. The means from each time
group should be normally distributed according
to the Central Limit Theorem.
The means of the measurements of QC
check samples made at each time were
treated_in a similar manner, having a grand
mean Y and variance about this mean s2(y).
The variance of the QC measurements pro-
vides an estimate of the day-to-day variation in
the operation of the instrument, because they
are standards of known concentration pre-
pared fresh each day with negligible error
which is due to preparation. The variance of
the sample measurements provides an esti-
mate of the variation which is due not only to
instrument differences but also to changes in
the chemical composition of the sample over
time. If the dissolved CO2 concentration of a
sample changes over time, the measurement
of DIG or pH will be farther away from the
grand mean of the population, and the esti-
mate of s2 will be larger.
The two variances (samples and QC)
were compared with an F-test. The hull hypo-
thesis was that s2 (sample) £ s2 (QCCS). This
was a one-tailed test with the alternate hypo-
thesis being s2 (sample) > s2 (QCCS). Rejec-
tion of the null hypothesis (P = 0.05) would
suggest that changes in dissolved CO2 concen-
tration in the samples over time contributed
more variation than day-to-day differences in
the operation of the instrument.
The results from Experiment 1 are pre-
sented in Table A-1 and Figure A-1. The pH of
the sample decreased substantially over the
first 22 hours, while the DIG concentration
nearly doubled during the same period of time.
Between 22 hours and 168.9 hours, the pH
fluctuated up and down between 7.2 and 7.7,
while the DIG concentration slowly increased
throughout the same period.
These results indicate the sample equili-
brated with the atmosphere with respect to
dissolved CO2 within about 24 hours. Fluctua-
tions observed after 24 hours may have
resulted from changes in atmospheric CO2
levels or from day-to-day differences in instru-
ment operation. This experiment demonstrates
the necessity of protecting samples low in
dissolved CO2 concentration from atmospheric
exposure.
45
-------
INITIAL Pco,=0
10.0-
9.0
7.0
3.0
3 2.0
^
o
o
0 1.0
0 20 60 100 140 . 180
TIME AFTER PREPARATION'(HOURS)
0 20 60 100 140 180
TIME AFTER PREPARATION (HOURS)
Figure A-1. Chang** In pH and dissolved Inorganic carbon over tint* In a aample Initially having no dissolved
carbon dioxide.
Table A-1. Dissolved Inorganic Carbon and pH
Measurements of a Sample Initially
Having No Dissolved CO2
Time After Preparation pH
(hours) (n=1)
0
0.33
22.0
47.4
119.1
168.9
9.47
_
7.23
7.44
7.22
7.68
DIG (ma/L)
Mean
1.109
1.113
2.110
2.154
2.167
2.249
SD
0.025
0.001
0.011
0.016
0.002
0.010
n
2
2
2
2
2
4
Experiment 2
The results of Experiment 2 are summa-
rized in Table A-2. One sample was analyzed
during a 12-hour period (Figure A-2), while the
second sample was analyzed during a 165-
hour period (Figure A-3). The results of both
sets of measurements are combined in Table
A-2. The initial measurements of the 165-hour
sample are not presented, but they were very
near those values obtained from the 12-hour
sample. The values for pH increased from
5.66 to 7.13 over the first 12 hours and were
fairly stable after that time. A mean pH of
9.01 was measured on the sample at 43.9
hours. This value is an outlier, but no explana-
tion for the discrepancy could be found, as
measurements of quality control samples at
the same time indicated the meter and elec-
trode were operating properly. The DIC con-
centration of the sample decreased slowly
over the first 2 hours and then decreased
more sharply until 12 hours, after which the
measurements appeared to stabilize.
Table A-2. Dissolved Inorganic Carbon and pH
Measurements Over Time on a Sample
Initially Supersaturated with Dissolved
Carbon Dioxide
Time After
Preparation (hours)
0.0
0.25
1.0
2.1
4.1
5.5
8.6
12.0
24.2
43.9
118.3
165.0
Mean
5.66
5.67
5.70
5.74
5.94
6.04
656
7.13
6.86
9.01
7.18
7.31
PH
SD
0.02
0.01
0.02
0.01
0.01
0.01
0.01
0.01
0.08
0.13
0.06
0.01
DIG (ma/U
n
2
2
2
2
2
2
2
2
2
2
3
2
Mean
12.94
12.41
11.58
10.60
7.80
6.55
3.29
2.29
2.53
2.40
2.17
2.18
SD
0.160
0.020
0.130
0.150
0.052
0.150
0.011
0.011
0.008
0.018
0.016
0.016
n
2
2
2
2
2
2
2
2
2
2
3
2
The results of this experiment demon-
strate that samples high in dissolved CO2
concentration should also be protected from
atmospheric exposure prior to analysis.
Experiment 3
Summary statistics for each of the four
levels of dissolved CO2 concentrations are
presented in Table A-3. For the Level "0"
solution, mean pH decreased 0.1 pH unit over
46
-------
EXPERIMENT TWO-PART ONE
7.2-
6.8-
3.4-
X
a
6.0-
5.6
V*
12.0-
10.0-
~ 8.0-
•^
o>
E
6.0-
4.0-
2.0-
TIME AFTER PREPARATION (HOURS)
4812
TIME AFTER PREPARARTION (HOURS)
Figure A-2. Changes In DIG and pH over • 12-hour period In * •ample Initially supersaturated with carbon
dioxide. (Means with standard deviations are presented).
EXPERIMENT TWO - PART TWO
Outlier
9.0
7.5
6.5
, w
12.5
T
T___— — ' " ^ 12-°'
' 3 ''
1 0 3.0
/ 5
/ 2'5'
2.0-
r \
i
\
\
I \
\
\
Xl~-I^
''
20 60 100 140 180
TIME AFTER PREPARATION (HOURS)
20 60 100 140 180
TIME AFTER PREPARATION (HOURS)
Figure A-3. Changes In DIG and pH over a 165-hour period In a sample Initially supersaturated with carbon dioxide.
(Means with standard deviations are presented).
47
-------
Table A-3. Summery Statistic* of pH and DIG Chang** Over Tim* for Solution* of Different Level* of Dlaaolvad
Carbon Dioxide Stored at Different Temperatures (valuaa arc presented a* maana ± SD, with aampla
alzaa In paronthaeaa.)
Time After
Preparation (hours)
Refrigerated
Room Temperature
PH
DIG (mg/L)
DIC (mg/L)
LEVEL "0" SOLUTION
0.0
24.5
96.8
119.8
191.2
0.0
18.3
91.4
113.8
161.0
0.0
24.0
47:6
144.3
166.3
0.0
22.9
45:7
120.5
215.1
9.46 ±0.07
9.40 ±0.04
9.37 ±0.02
9.32 ±0.09
9.36 ±0.02
7.52 ±0.01
7.48 ±0.06
7.41 ±0.13
7.45 ±0.09
7.52 ±0.08
6.70 ±0.01
6.75 ±0.01
6.74 ±0.02
6.72 ±0.01
6.72 ±0.03
5.67 ±0.01
5.70 ±0.01
5.71 ±0.02
5.70 ±0.01
5.74 ±0.01
(3)
(7)
(6)
(8)
(8)
(3)
(8)
(8)
(8)
(8)
(3)
(7)
(8)
(8)
(8)
1.538
1.535
1.479
1.487
1.514
±0.023
±0.033
±0.015
±0.040
±0.036
(3)
(8)
(7)
LEVEL "1" SOLUTION
2.134 ±0.003
2.091 ±0.056
2.142 ±0.006
2.110 ±0.066
2.078 ±0.037
(3)
(7)
(7)
(8)
(8)
LEVEL "W SOLUTION
2.840 ±0.038
2.919 ±0.020
3.067 ±0.085
2.800 ±0.028
2.857 ±0.068
(3)
(7)
(8)
(8)
(8)
LEVEL "100" SOLUTION
(3) 12.41 ±0.12 (3)
(8) 12.51 ±0.22 (8)
(8) 12.33 ±0.10 (7)
(5) 12.29 ±0.13 (5)
(5) 12.02 ±0.15 • (4)
9.46 ±0.07
9.39 ±0.03
9.22 ±0.02
9.09 ±0.07
8.78 ±0.03
7.52 ±0.01
7.44 ±0.07
7.21 ±0.05
7.15 ±0.04
7.10 ±0.05
6.70 ±0.01
6.73 ±0.03
6.74 ±0.04
6.67 ±0.06
6.68 ±0.03
(3)
(8)
(7)
(8)
(8)
(3)
(8)
(8)
(8)
(8)
(3)
(8)
(8)
(8)
(8)
5.67 ±0.01 (3)
6.71 ±0.01 (8)
5.74 ±0.02 (8)
5.77 ±0.01 (8)
5.87 ±0.01 (7)
1.538 ±0.023
1.533 ±0.029
1.525 ±0.034
1.596 ±0.017
1.757 ±0.023
2.134 ±0.003
2.137 ±0.025
2.283 ±0.024
2.276 ±0.051
2.286 ±0.036
2.840 ±0.038
2.928 ±0.059
3.067 ±0.086
2.860 ±0.093
2.927 ±0.088
12.41 ±0.1£
12.06 ±0.10
11.17 ±0.16
10.41 ±0.05
9.44 ±0.22
(3)
(8)
(8)
(8)
(8)
(3)
(8)
(8)
(8)
(7)
(3)
(8)
(8)
(8)
(8)
(3)
(8)
(8)
(6)
(4)
the course of the experiment in the refrigerated
syringe samples and decreased almost 0.7 pH
units in the samples stored at room tempera-
ture. In the same solution, OIC exhibited a
slight overall decrease (0.03 mg/L) over the
course of the experiment in the refrigerated
samples, and increased by approximately 0.2
mg/L in the samples held at room temperature.
The simultaneous decrease of pH and DIC in
the refrigerated samples is not possible based
on the known relationship between DIC and
pH, and the observed change is a result of
random experimental error.
For the Level "1" solution, pH measure-
ments indicated no net change over the course
of the experiment in the refrigerated samples
and a maximum change of 0.1 unit from time
— 0, which occurred at 91.4 hours. Samples
held at room temperature showed a steady
decrease in pH over time, with a net change of
approximately 0.4 units between time = 0 and
161 hours. The DIC content of this solution
changed very little in xthe refrigerated samples
over the course of the experiment (approxi-
mately 0.05 mg/L net change). In the samples
held at room temperature, the DIC content
showed a net increase of approximately
0.15 mg/L.
Samples of the Level "10" solution that
had been refrigerated exhibited a maximum
change in pH of .0.04 units from initial mea-
surements over the course of the experiment,
and a net change of only 0.02 units. A similar
pattern was observed in samples held at room
temperature; a net decrease of 0.02 units was
observed. The DIC content of refrigerated
samples showed a net gain of 0.017 mg/L,
while samples held at room temperature
showed a net increase of 0.08 mg/L. The
maximum difference from initial measurements
was observed at 47.6 hours for both pH and
DIC.
48
-------
For the Level "100" solution, refrigerated
samples exhibited a steady increase in pH over
time, with- a net change of 0.07 units, while
samples held at room temperature exhibited a
similar pattern of change, with a net change of
0.2 units. The DIG content of refrigerated
samples showed a steady decrease over time,
with a net change of approximately 0.4 mg/L
for refrigerated samples and a net change of
approximately 3 mg/L in samples held at room
temperature.
For each solution, parameter, and tem-
perature group, the mean values for each time
period were averaged, and the variance was
calculated. The same calculations were per-
formed on the mean values for the QC check
samples that had been analyzed along with
the samples. For pH, the pH 4.00 QCCS was
used initially. For DIG, the 2.00 mg/L C QCGS
was used for all solutions except for the Level
"100" solution; the 20.00 mg/L C QCCS was
used for the Level "100" solution. A F-statistic
was calculated from the ratio of the sample
variance to the QCCS variance for each solu-
tion, parameter, and temperature group. The
results of these tests are presented in Table
A-4.
There was a significant effect (p < 0.05)
of holding time on pH measurements for the
Level "0" and Level "1" solutions that had been
stored refrigerated. The other refrigerated
solutions were not affected by holding time to
a significant degree with respect to pH. The
dissolved inorganic carbon content of all
solutions that were stored refrigerated was
not significantly effected by holding time.
The effects of holding time on pH and
DIC were much more pronounced in samples
that had been stored at room temperature.
Only the Level "10" solution pH was not signifi-
cantly affected by holding time. The DIC
content of all but the most supersaturated
solution (Level "100") was not affected by
holding time in samples stored at room tem-
perature. The stronger effects observed in the
samples held at room temperature may have
resulted from the laboratory having an ambient
CO2 level greater than 300 ppm.
The variance associated with measure-
ments of the pH 4.00 QCCS was very small in
all cases (0.0002). This solution has a much
higher ionic strength than the synthetic sam-
ples analyzed in this experiment. The pH of
this QCCS sample was also much lower than
the circumneutral pH values observed in the
synthetic samples. Solutions of lower pH
generate a stronger potential for the meter to
read, and the expected variation about this
potential would be expected to be less than
that expected from a potential generated on a
circumneutral sample. The significant effects
of holding time observed for the pH 4.00 QCCS
may be a result of inherent properties of the
instrumentation rather than a temporal change
in the chemical concentration of the sample.
Table A-4. Summary of F-Te*tt Performed on
Variance* of Sample and QCCS
Measurements for Solution* of Different
Initial Dissolved Carbon Dioxide
Concentration* (number of measurement*
I* In parentheses.)
Solution
s2 (Samples) sz (QCCS) F-Value
REFRIGERATED SAMPLES
pH Measurements
Level "0"
Level "1"
Level "10"
Level "100"
Level "0"
Level "1"
Level "10"
Level "100"
0.0027 (5)
0.0022 (5)
0.0004 (5)
0.0006 (5)
0.0002
0.0002
0.0002
0.0002
DIC Measurements
0.0007 (5) 0.0028
0.0007 (5) 0.0017
0.0108 (5) 0.0064
0.0337 (5) 0.0362
(5) 13.50
(5) 11.00*
(5) 2.00"*
(5) 3.00""
(5) 0.25"8
(5) 0.41"8
(5) 1.69"8
(5) 0.93"8
ROOM TEMPERATURE SAMPLES
pH Measurements
Level "0"
Level "1"
Level "10"
Level "100"
0.0730
0.0343
0.0009
0.0057
(5)
(5)
(5)
(5)
0.0002
0.0002
0.0002
0.0002
(5) 365.0***
(5) 171.6
(5) 4.5™
(5) 28.5
DIC Measurements
Level "0"
Level "1"
Level "10°
Level "100"
0.0095 (5)
0.0064 (5)
0.0079 (5)
1.4686 (5)
0.0028 (5) 3.39n8
0.0017 (5) 3.76n8
0.0064 (5) 1.23"8
0.0362 (5) 40.57**
D" not significant at P = 0.05.
*t significant at p < 0.05.
**^ significant at p < 0.01.
*** significant at p < 0.001.
49
-------
A pH 7.00 QCCS was also prepared each
day. This solution was prepared almost
identically to the synthetic samples, and thus
it more closely mimics the chemical behavior
of the samples, providing a more accurate
representation of the expected analytical
variance. The pH of this solution also more
closely approximates the observed pH of the
sample solutions. The pH 7.00 QCCS was
analyzed only once on most days, and 18 total
measurements were made over the course of
these experiments. A variance based on these
measurements was calculated and compared
against the variances for pH measurements of
samples. The results of these comparisons
are presented in Table A-5. As expected, the
QCCS had a larger variance than the pH 4.00
QCCS. In all cases, there was no significant
effect of holding time on the pH of syringe
samples that could be measured.
Tabloi A-5. Comparisons off Variances ffor.pH
M®ffl8ur®mentQ off' Solutions Having
DlffffotronS DIoooIvod CO2 ConcontratDono
Vorouo a Low Ionic Strength pH 7.00
Quality Control Chock Samplo (numbor off
moaoyramonSo Do In
possible affects of biological activity on sam-
ple stability.
Solution
a2 (Samples) s2 (QCCS) F-Value
REFRIGERATED SAMPLES
Laval "0"
Level "1"
Level "10"
Level "100"
0.0027
0.0022
0.0004
0.0003
(5)
(5)
(5)
(5)
0.2627
0.2827
0.2627
0.2627
(18)
(18)
(18)
(18)
OJOIO"3
oooap
QCOST3
QflOS?0
ROOM TEMPERATURE SAMPLES
Level "0"
Level "1"
Level "10"
Level "100"
0.0730
0.0343
0.0009
0.0057
(5)
(5)
(5)
(5)
0.2627
0.2627
0.2627
0.2827
(18)
(18)
(18)
(18)
oznT*
0.131™
oxraaf3
QO2SP
The results of measurements made at
each time interval on natural water samples
are summarized in Table A-6. The pH of
refrigerated samples from both lakes showed
a gradual decrease over the course of the
experiment. Padden Lake exhibited a larger
net decrease (0.17 pH units) than Bagley Lake
(0.08 units). DIC values for refrigerated sam-
ples slowly increased throughout the experi-
ment; Padden Lake exhibited a larger net
change in DIC (approximately 0.3 mg/L vs. 0.05
mg/L for Bagley Lake).
The variance of sample measurements
for each solution was compared to the vari-
ance of the pH 7.00 QCCS measurements or
of the 2.00 mg/L C QCCS. Refrigerated sam-
ples from both lakes did not exhibit a signifi-
cant effect of holding time for either pH or
DIC. There was no significant effect of
holding time on the pH of samples held at
room temperature, but both lakes exhibited a
significant change in DIC over time. The
results of this experiment suggest that the use
of pH measurement values instead of the
hydrogen ion concentration values may be a
less sensitive indicator of dissolved CO2 con-
centrations than DIC.
Both of these lakes are similar in their
chemical composition to target...waters of
NSWS programs. It appears that natural
water samples can be stored for up to 7 days
in sealed syringes without a significant change
in dissolved CO2 concentration if they are held
in the dark at approximately 4 °C.
na not significant at P = 0.05.
Most of the water samples collected
during NSWS programs are supersaturated
with dissolved CO2 (3-20 x atmospheric)
(Linthurst et al., 1986). The results of this
experiment indicate that samples collected in
syringes can be held for up to 7 days with no
measurable change in dissolved CO2 concen-
tration. It should be noted that the samples
analyzed here were sterile, and natural water
samples having biological activity associated
with them may be much less stable at warmer
temperatures. Experiment 4 evaluated the
The results of these experiments indicate
that water samples not necessarily in equili-
brium with the atmosphere can be stored in
sealed syringes for up to 7 days, if they are
kept refrigerated, without a measurable effect
on either DIC or pH.
The results of these experiments pertain
only to pH and DIC, and no inference should
be made regarding possible changes in other
chemical parameters (e.g., aluminum) in water
50
-------
Table A-6. Summary Statlatlce for Each Tim* Interval From Samplaa Collected From Paddan and BagUy Lakaa
(valuta ar* praaantad aa maana ± SD, with tha numbar of maaauramanta In paranthaaaa.)
Time After
Preparation (hours)
pH Measurements
Refrigerated Room Temperature
PIC Measurements
Refrigerated
Room Temperature
PADDEN LAKE
0.0
13.0
87.3
108.9
159.0
Mean:
Variance:
s2 (QCCS):
F-value:
8.96 10.03
8.96 ±0.09
8.95 ±0.03
8.89 ±0.04
8.79 ±0.04
8.91
0.0054
0.2627
0.021"*
(12)
(4)
(4)
(6)
(6)
(5)
(18)
8.96 ±0.03
8.84 ±0.09
7.85 ±0.17
7.87 ±0.09
7.40 ±0.04
8.18
0.4643
0.2627
1.767""
(12)
(4)
(4)
(5)
(6)
(5)
(18)
4.866 ±0.046
4.930 ±0.090
4.989 ±0.021
4.963 ±0.028
5.115 ±0.045
4.973
0.0085
0.0018
4.722"'
(12)
(4)
(4)
(6)
(5)
(5)
(5)
4.866 ±0.046
4.988 ±0.056
5.470 ±0.124
5.388 ±0.062
5.794 ±0.119
5.301
0.1416
0.0018
78.667*"
(12)
(4)
(4)
(6)
(6)
(5)
(5)
BAGLEY LAKE
0.0
13.0
87.3
108.9
159.0
Mean:
Variance:
s2 (QCCS):
F-value:
6.79 ±0.02
6.80 ±0.03
6.76 ±0.02
6.74 ±0.01
6.71 ±0.03
6.76
0.0014
0.2627
0.005"s
(12)
(4)
(4)
(5)
(6)
(5)
(18)
6.79 ±0.02
6.77 ±0.05
6.46 ±0.05
6.42 ±0.07
6.28 ±0.03
6.54
0.0509
0.2627
0.194"*
(12)
(4)
(4)
(6)
(6)
(5)
(18)
1.154 ±0.021
1.221 ±0.014
1.175 ±0.014
1.179 ±0.026
1.206 ±0.034
1.187
0.0007
0.0020
0.350"*
(12)
(4)
(4)
(6)
(6)
(5)
(5)
1.154 ±0.021
1.269 ±0.021
1.517 ±0.076
1.544 ±0.099
1.677 ±0.045
1.432
0.0459
0.0020
22.95*"*
(12)
(4)
(4)
(6)
(6)
(5)
(5)
"* not significant at P = 0.05.
*** significant at p < 0.001.
held in sealed syringes, as other chemical
parameters may be sensitive to small changes
in pH that were not measured in this study.
Literature Cited
American Society for Testing and Materials.
1984. Annual Book of ASTM Standards.
Vol. II. 01, Standard Specifications for
Reagent Water. D 1193-77 (reapproved
1983). ASTM, Philadelphia, Pennsylvania.
Linthurst, R. A., D. H. Landers, J. M. Eilers,
D. F. Brakke, W. S. Overton, E. P. Meier,
and R. E. Crowe. 1986. Characteristics
of Lakes in the Eastern United States,
Vol. I: Population Descriptions and
Physico-Chemical Relationships.
EPA-600/4-86/007a. U.S. Environmental
Protection Agency, Washington, D.C.
Hillman, D. C. J., J. F. Potter, and S. J. Simon.
1986. National Surface Water Survey,
Eastern Lake Survey-Phase 1 (Synoptic
Chemistry) Analytical Methods Manual.
EPA 600/4-86/009. U.S. Environmental
Protection Agency, Office of Research
and Development, Las Vegas, Nevada.
51
-------
Appendix B
Cubitainer Holding Time Study
by
Martin A. Stapanian1, Allison K. Pollack2, and Bryant C. Hess1
1Acid Deposition Quality Assurance Group
Lockheed Engineering and Management Services Company, Inc.
1050 East Flamingo Road
Las Vegas, Nevada 89119
and
2Systems Applications, Inc.
101 Lucas Valley Road
San Rafael, California 94903
Introduction
The U.S. Environmental Protection Agency
undertook a survey of approximately 3,000
lakes and 300 streams in the United States as
part of the National Surface Water Survey
(NSWS). A primary objective of the NSWS is
to determine the degree to which surface
waters in the United States are at risk as a
result of acid deposition (Linthurst et al., 1986).
A survey of this magnitude has many associ-
ated quality assurance and economic con-
cerns. One concern is that the concentration
of chemicals in samples may change from the
time of collection to the time they are prepared
for analysis (holding time).
In NSWS studies to date, samples were
filtered and preserved at a field laboratory,
which was centrally located in the region
where samples are being collected, and then
were transported by truck or air charter to an
analytical laboratory for chemical analysis
(Linthurst et al., 1986; Messer et al., 1986).
Standard methods of filtration and preserva-
tion assured that minimal change occurred
from the time samples were preserved to the
time they were analyzed at the analytical
laboratory. However, the degree to which
holding time affects the chemistry of water
samples is unknown.
The holding times of samples vary con-
siderably depending on the distance from
sampling site to the field laboratory, method
of transportation, weather conditions, and
terrain. During Phase I of the Western Lake
Survey, conducted in the fall of 1985, for exam-
ple, samples collected by ground crews some-
times required up to 2 days to be transported
to a field laboratory, while samples collected
by helicopter crews required only a few hours.
If holding time affects the concentrations of
chemical parameters in a sample, interpreta-
tion of data from samples with different hold-
ing times becomes difficult. If holding time
has no effect on the concentration of param-
eters in a sample, less expensive methods of
transporting samples can be used, and field
laboratories would be unnecessary.
52
-------
This study investigated changes in any of
25 chemical parameters in samples that were
held 12, 24, 48, and 84 hours before prepara-
tion for analysis. The results may be useful in
designing sampling strategies for future water
quality surveys covering large or remote geo-
graphical areas, i.e., for surveys where the
time between collection and preparation of
samples is variable.
Materials and Methods
Collection and Chemical Analysis of
Samples
Water samples were collected on June
24, 1985, from three lakes in New York, three
streams in Pennsylvania, two streams in
Maryland, and one stream each in South
Carolina and Tennessee. Samples were col-
lected over this wide area to determine if
effects of holding time were due to phe-
nomena of lakes and streams in certain geo-
graphical areas. The sampling locations are
listed in Table B-1.
At each site, two 19-liter (5 U.S. gallons)
Cubitainers were filled with water. The con-
tents were then mixed by rocking the Cubi-
tainers for five minutes. A Tygon tubing
T-connector was attached to transfer water
from both 19-liter Cubitainers simultaneously
into eight 3.7-liter (1 U.S. gallon) Cubitainers.
Each 3.7-liter Cubitainer was rinsed three
times, was filled, and was sealed with no
headspace. Holding times of 12, 24, 48, and
84 hours were randomly assigned to each
Cubitainer. Each lake or stream, therefore,
had one pair of 3.7-liter Cubitainers for each
holding time. Samples were shipped via
charter aircraft to a field laboratory in Sylva,
NC, and arrived within 12 hours of collection
so that samples could be prepared and pre-
served at their assigned holding times. Pre-
served sample aliquots were shipped the
following day to an analytical laboratory by
overnight courier service. Samples were
stored at approximately 4 °C during shipment.
A detailed description of the chemical analyses
and methodologies performed by the analytical
laboratory can be found elsewhere (Hillman et
al., 1986). All samples were analyzed within an
acceptable time after filtration and preserva-
tion (Hillman et al., 1986; Drous6 in prepara-
tion). The methods were the same as those
used during Phase I of the Eastern Lake
Survey (Linthurst et al., 1986), and during the
pilot survey for the National Stream Survey
(Messer et al., 1986).
Statistical Analysis
Standard single-factor repeated mea-
sures analyses of variance (RMANOVA; Winer,
1971) were performed for each parameter.
Specifically, we tested the null hypothesis of
equal mean concentrations of value of a
particular parameter at 12, 24, 48, and 84
hours.
An assumption of the RMANOVA model
is that the measurement error does not vary
as a function of concentration. For Ca2+,
Mg2+, total phosphorus, extractable aluminum,
Na+, Cr conductance, an air-equilibrated
dissolved inorganic carbon, and acid neutraliz-
ing capacity, log transformations were required
Table B-1. Locations of Lakes and Streams
Lake or Stream Name
County
State
Latitude
Longitude
Big Moose Lake
Fly Pond
West Lake
Bear Creek
Cherry Run
Lick Run
Lyons Creek
Morgan Creek
Six Mile Creek
Clear Creek
Herkimer
Herkimer
Herkimer
Lycoming
Clinton
Clinton
Anne Arundel
Kent
Pickens
Anderson
New York
New York
New York
Pennsylvania
Pennsylvania
Pennsylvania
Maryland
Maryland
South Carolina
Tennessee
43*49'45"N
43*44'52"N
43*45'20"N
41*21'31"N
39*59'26"N
41*19'02"N
38*11'00"N
39*15'20"N
34'46'54"N
36*10'00"N
74*51'00"W
74'54'05'W
74'55'13'W
76*50'50"W
77*29'34"W
77'30'52'W-
76'34'OCTW
76*02'30"W
82*50'57"W
84'OOWW
53
-------
to satisfy this assumption. Lyons Creek had
extremely unusual data for total aluminum.
The RMANOVA for total aluminum was par-
formed both with and without data from this
stream.
For each parameter, holding time is of
practical significance when the precision
among the holding times is less than the
precision between standard samples with no
holding time. Forty-one samples obtained
from Big Moose Lake in New York in 1984
were used to estimate analytical precision and
to evaluate laboratory bias and precision
during Phase I of the Eastern Lake Survey
(Best et al., in preparation). These "natural
audit" samples were shown to be chemically
stable during the Survey and therefore pro-
vided a useful estimate of the amount of
variance which was due to the collection and
analysis of samples. The relative standard
deviation (standard deviation + mean) was
used to estimate the precision of the Eastern
Lake Survey natural audit data for each para-
meter (Permutt and Pollack, in preparation).
For pH measurements, the absolute standard
deviations were used. The relative and abso-
lute standard deviation values were corrected
for laboratory bias and trend (Permutt and
Pollack, in preparation). The greater the value
of the relative or standard deviation, the less
is the precision and the greater is the variance
in measuring a parameter.
The precision between two holding times
can be estimated by the relative difference
(difference + mean value) between two holding
times. For pH measurements, precision
between two holding times can be estimated
by the absolute difference between the two
holding times. If the relative (or absolute)
difference between two holding times is less
than the relative (or absolute) standard devia-
tion of the natural audit samples, then the
variance which is due to holding time is less
than the variance which is due to the collection
and analysis of samples. Therefore, statisti-
cally significant effects of holding time become
questionable, or insignificant in the practical
sense, when the relative (or absolute) differ-
ences between holding times are less than the
relative (or absolute) standard deviation of the
natural audit data.
When determining the practical signifi-
cance of holding time, one should also con-
sider whether the effects are limited to a few
lakes or streams or to the majority and
whether the effect is the same for all lakes
and streams sampled. Suppose that the
concentration of a particular parameter
increases at a certain holding time for some
lakes and streams, while the concentration
decreases or does not change at the same
holding time for the remaining lakes and
streams. In such a case, the effects of hold-
ing time may be due to site-specific phe-
nomena. The practical significance of holding
time for that parameter, regardless of the
outcome of the RMANOVA, is questionable.
Plots of the data (Figure B-1) suggest
that for all parameters, holding time did not
have a strong effect over all lakes and
streams. The results of the RMANOVA for
each parameter (Table B-2) indicate that
holding time had a statistically significant
effect (p < 0.05) for 17 parameters. The
results of groups of parameters are discussed
in more detail below.
(Figure B-1, Parts A-G)
Holding time had a statistically signifi-
cant effect on the concentrations of Ca2+, Fe
(total ionic), Na+, and ft/In (total ionic) (p <
0.05, Table B-2). No significant effect of
holding time was observed for Mg2+, K+, and
NH^. The average concentration of Ca2+ and
Fe increased up to 48 hours, then sharply
decreased (Table B-2). However, this was
primarily because of having data from only one
stream for Ca2+ and from only three streams
for Fe (Figure.B-1, Parts A and F). Similarly,
the mean decrease in the concentration of Mn
between 12 and 48 hours is due 'primarily to
data from three streams (Figure B-1, Part E).
Although statistical significance was ob-
served, the practical significance of holding
time on the .concentrations of Ca2+, Na+, Fe,
and Mn is questionable. For Ca2+ and Na+,
the relative difference in concentration between
holding times was always less than the rela-
tive standard deviation of the ELS natural audit
data (Table B-3). The average concentration of
Ca2+ increased up to 48 hours, then sharply
decreased (Table B-2). However, these
54
-------
32
31-
30-
13
12-,
' 4-
3'
2-
10 20 30
40 SO 60
HOURS WITHHELD
70 80 90
10 20 30
40 50 60
HOURS WITHHELD
70 60 90
(0)
10 20 30
40 90 60
HOURS WITHHELD
-Big Mooae Lake
••West Lake
70 60 90
Fly Pond
—- Bear Creek
10 20
30 40 50 60
HOURS WITHHELD
70 80 90
LEGEND
Clear Creek Six Mile Creek Lyon Creek
Figure B-1. Mean values for pairs of measurements at 4 holding times for 10 lakes and streams. (This figure Is 7 pages total. Units are peq/L for
acid-neutralizing capacity and base-neutralizing capacity; j/S/cm for conductance; pH units for pH^Q, pH^c, and pH^; and mg/L for the
remaining parameters.)
-------
0.18-
0.17
0.16
0.15
0.14
O.I 3
0.12
O.I 1
e 0.10
0.08
0.07-
0.06-
0.05-
0.04-
0.03-
0.02-
0.00-
0.14-
0.13
0.12-
0.1 1-
0.10-
0.09-
0.08
*» 0.07
0.06
0.05
0.03
0.02-
0.01
0.00-
_, .;_ ' ... . „.., :. - . •., ... v. te} '.'• °-"°
"~\ ' 0.331
^-. 0.311
x.. __ 0.292
^~~___ X--^ ' ' 0.272-
"^v. "'\ --" . 0.253
^ ^ "'• 0.233-
^ \ '0.214-
^X " ' °-W'
•"~^-'^--~^ o-l5^'
~ ~ T~ — . __ 0. 1 36
"~ ~~ —" " 0. 1 1 7-
0.097-
: -. ''._ ^_ 0.078-
- '"""•' : .' ~ ••— 0.058-
""^-^ 0.039
— . : '-^^T-^'jpT ' -; - -r— n ... 0.000-
0 20 30 ' ' .40 5.6' 60 ' ' ' 70 ' 80 90
HOURS WITHHELD
!<*> 18-1
— ^ ^ — 1 7-
— — — — ! 6"
~~ ~~ -r- -_ 15-
14-
13-
12-
1 1-
y-'--- ,.,::.-:— -^r i 'i-
\ ^•'•"~ . • ' °-
---Y- :--'-' " *-
s-
— ^SV- •-• ~.~.~ ~-^^:.ii^:^-~=^-==-- ' «-
' ' • ~- "-" Q.
0 20 30 40 50 60 "70 ' ' 80 ' 90
HOURS WITHHELD
LEGEND
. . ' (P)
/•.
\
/ X
/' \,
. /' \
/ X-.N
/ N'\
-x / . ~\
•-:^>l.
0 20 30 40 50 60 70 80 90
HOURS WITHHELD
<">
__ •m^..^,^^^=:-~±r-^^.-.-=^=,^^~^,. :
0 20 30 40 50 60 70 80 80
HOURS WITHHELD
J.. ,v
Figure B-1. Continued.
-------
I ' 3-
V)
O
90 I
HOURS WITHHELD
40 50 60
HOURS WITHHELD
a-
NIC CARBON
ot o>
a 4-
O
3
> *-
i-
0
(K) 32-
~— - , |«;
cc
0 10-
1
I 8
n «.
>
O
O
&-.- ^u^^^—^^ ^^-^^^^
0-
0 20 30 40 50 60 70 80 90
HOURS WITHHELD
LEGEND
(L)
-•
— zj-^mi__
.»^.-,-.™™™K-^^»™..,»™
0 20 30 40 50 80 70 80 90
HOURS WITHHELD
V
Flgur* B-1. Ccntlmwd.
-------
32-
S
O
(ED INORGANIC CA
N fc
O CD <
o1 «•
CO
CO
o ••
111
1 4-
S 2-
IT
21-
20-
19-
13-
12-
^
0 1 1-
10 "*
8-
7-
5-
4-
(M) 280-
, — - — 260-
5 160-
H
i.~-~~~ i 140-
g 120-
: : ^ • • • = — 0-
0 20 " 30 40 50 60 70 • 60 90
HOURS WITHHELD
(6) 0.17-
— ~~ - 0. 1 5-
0.14-
; 0.13
0.12
3j 0.09'
< 0.08-
, , ° 0.07.
- --^~- - . 006.
: 0.05
0.04
0.03-
o.oi-
'| 1 | ".' 1 I TTlTpTTrm T 1 J'VI 1 1 I 1 1 I'lf 1 1 1 1 I 1 1 1 1 | I 1 1 | |
0 20 30 • 40 50 60 70 80 80
HOURS WITHHELD
LEQEND
(N)
0 20 30 40 50 60 70 80 90
HOURS WITHHELD
: ! —
0 ZO 30 40 50 60 70 s'o o'fl
HOURS WITHHELD
Figure B-1. Continued.
-------
Ol
(O
0.37J
0.47-
100.13
D
0.12
IO.M
* 0.10-
<
O 0.05
K
0.04-
0.02-
0.01-
0.00-
0.26-
0.25
0.24
I 0.2,
0.22-
0.21
0.20;
0.04-
O 0.03-
0.02-
0.01-
0.00-
^_ CO) 7.00-
~~ -^ _ _ — "" 6.00-
5.00^
4.00H
| 3.00
§ 2.00
< 1.00-
S
"""••—._ "~ 0.75-
~"~-— .. 0.50-
:~.-^- 0.25-
0.00-
0 20 30 40 50 60 70 60 90
HOURS WITHHELD
^^^^^ (8) 0.0680^
^ ~^"^ ~~~~-~^^^ 0.063 1
f^"^^ ^^^^_^ 0.0583
/ ^^^^^ S 0.0534-
/ | 0.0466
/ 3 0.0437
/ 2 0.0369
0.0340-
0.029 1
0.0243
0.0194
: 0.0146
0.0097-
0 20 30 40 50 60 70 80 90
HOURS WITHHELD
LEGEND
(R)
. \
x \
"^^^
X
X
0 20 30 40 50 60 70 80 90
HOURS WITHHELD
,—~~ "x
/ "x.
/ *x.^
/ x^
/ "-X
/
-^^,
0 20 30 40 50 60 70 80 90
HOURS WITHHELD
V
Figure B-1. Continued.
-------
2800-j
2700-j.
> 800
!7oo;
1 400-
N
£ 300
D
T 200-
8
* 100-
0
-100-
150
140
130-
120
I 10
100-
90-
80-
70-
60-
50-
40-
30-
20-
10-
0
(VI
40 50 60
HOURS WITHHELD
40 JO 60
HOURS WITHHELD
§
LEGEND
e C™°" L'°n C'°ek
Figure B-1. .Continued.
-------
(X)
10 20
TT-fT-T
30
I "
> 90
HOURS WITHHELD
TT-p-,
60
' ' I M|
60 90
10 20 JO
40 50 60
HOURS WITHHELD
TO BO 90
0>
(Y)
I
40 90 60
HOURS WITHHELD
70 80 90
LEGEND
Figure B-1. Continued.
-------
Table B-2. Treatment Meana and Overall Standard Error (aa determined by RMANOVA") of 25 Chemical Parametera
at Four Holding Tlmee for Three Lakes and Seven Streama. (Probabllltlea (p) < 0.05 for the repeated
measured analysis of variance Indicate a statistically significant effect of holding time).
Parameter; units
Calcium; mg/L
Magnesium; mg/L
Sodium; mg/L
Potassium; mg/L
Manganese; mg/L
Iron; mg/L
Ammonium; mg/L
Sulfate; mg/L
Nitrate; mg/L
Chloride; mg/L
Dissolved organic carbon; mg/L
Dissolved inorganic carbon; mg/L
Dissolved inorganic carbon, air-equilibrated; mg/L
Conductance; pS/cm
Silica; mg/L
Fluoride; mg/L
Phosphorus, total; mg/L
Aluminum, total' mg/L
Aluminum, total , mg/L
Aluminum, extractable; mg/L
Aluminum, organic monomeric; mg/L
Acid neutralizing capacity; peq/L
Base neutralizing capacity, jieq/L
pH. Initial BNC
pH, initial ANC
pH, air-equilibrated
12
7.012
2.571
1.874
1.238
0.054
0.057
0.039
5.845
1.130
3.276
2.569
4.842
4.493
71.42
7.754
0.070
0.074
0.509
0.113
0.027
0.012
438.00
52.103
6.816
6.742
7.326
-Holding Time (hours)-
24 48
7.185
2.566
1.866
1.274
0.052
0.050
0.036
5.743
2.023
3.269
2.575
4.734
4.476
71.48
7.876
0.071
0.078
0.736
0.145
0.030
0.015
436.75
59.303
6.738
6.703
7.356
7.230
2.592
1.864
1.254
0.040
0.084
0.035
5.735
2.021
3.134
2.457
5.137
4.817
71.80
7.798
0.070
0.066
0.736
0.109
0.033
0.017
436.50
63.053
6.663
6.652
7.299
84
7.019
2.596
1.840
1.243
0.039
0.035
0.031
5.704
2.024
3.071
2.758
5.017
4.858
71.61
7.814
0.071
0.073
0.346
0.132
0.030
0.015
434.38
65.653
6.644
6.635
7.308
SE
0.035
0.013
0.004
0.009
0.003
0.010
0.002
0.022
0.021
0.050
0.049
0.094
0.073
0.125
0.032
0.000
0.003
0.126
0.009
0.001
0.001
2.075
2.518
0.023
0.020
0.015
P
<0.0001"
0.0042"
<0.0001"
0.0643
0.0096
0.0420
0.1871
0.0002
0.0092
0.0483"
0.0002
0.0505
0.0005"
0.0465"
0.7970
0.3770
0.1370
0.0875
0.0592
<0.0001"
0.0342
0.0456"
0.0195
<0.0001
0.0015
0.0514
RMANOVA performed by using log-transfer red values.
Data from Lyons Creek excluded.
changes were due primarily to data from Clear
Creek. The remaining lakes and streams
exhibited stable concentrations. For Ca2+ and
Na+, the mean change in concentration from 12
to 84 hours was 0.1 percent and 1.3 percent,
respectively.
For Mn, the relative differences in con-
centration between samples held 24 and
48 hours and between 12 and 84 hours were 3
percent and 9 percent greater, respectively,
than the relative standard deviation of natural
audit samples (Table B-3). However, the
relative difference in concentration between
samples held 12 and 24 hours and between 48
and 84 hours was much less (19.2 percent and
20.5 percent) than the relative standard devia-
tion of the natural audit data. Morgan Creek,
Lyons Creek, and Six Mile Creek exhibited large
decreases in concentration of Mn from 12 to
84 hours.
For the remaining lakes and streams, the
concentration of Mn was stable from 12 to 84
hours. Similarly, Morgan Creek, Lyons Creek,
and Six Mile Creek exhibited large fluctuations
in the concentration of Fe from 12 to 84 hours.
Relatively large increases in Fe were observed
between 24 and 84 hours for these three
streams, while large decreases were observed
from 48 to 84 hours. The concentration of Fe
increased from 12 to 24 hours for Lyons Creek,
but decreased for Morgan Creek and Six Mile
Creek. For the remaining lakes and streams,
the concentration of Fe was stable from 12 to
84 hours. For Fe, the relative difference
between 48 and 84 hours was 17.3 percent
greater than the relative standard deviation of
the ELS natural audit data. Between 12 and
24, 24 and 48, and 12 and 84 hours, the rela-
tive differences were less than the relative
standard deviation of ELS natural audit data.
The practical significance of the effects of
holding time on Fe and Mn is questionable
because of (1) the lack of an overall trend in
Figure B-1, Parts E and F, (2) the effects of
holding time being mixed and restricted to
62
-------
Table B-3. Relative Difference (Dlffaranca Batwaan Maana of Holding Tlmaa In Quaatlon + Grand Maan of Holding
Tlmaa In Quaatlon) Batwean tha Maana for Each Paramatar for Holding Tlma Intarvala of 12 to 24, 24
to 46, 48 to 84, and 12 to 84 houra.
(hours)
Parameter
Calcium (Ca2+)
Magnesium (Mg2+)
Sodium (Na^)
Potassium (K+)
Manganese (Mn)
Iron (Fa)
Ammonium (NHt)
Sulfate (SOJ)
Nitrate (NO;)
Chloride (CO
Dissolved organic carbon (DOC)
Dissolved inorganic carbon (DIC)
DIC. air equilibrated (DIG..)
Conductance (COND)
Silica (Si02)
Fluoride, total dissolved (F)
Phosphorus, total (P)
Aluminum, total (Al)
Aluminum, total* (Al)
Aluminum, extractable (Alext)
Aluminum, organic monomeric (Alor_); j/g/L
Acid neutralizing capacity (ANC); ifglL
Base neutralizing capacity (BNC); pg/L
pH. initial BNC (pHBNC)c
pH. initial ANC (pHANC)<:
pH. air equilibratedTpH,,,)*
* Absolute standard deviation was used for pH
b Data from Lyons Creek removed.
e Absolute difference.
12-24
•0.022
0.002
0.004
-0.029
0.038
0.131
O.OSO
0.017
0.052
0.002
-0.002
0.023
0.004
-0.001
-0.016
-0.014
-0.053
-0.365
-0.250
-0.105
0.222
0.003
0.129
0078
0.039
0.030
measurements
24-48
•0.006
0.005
0.001
0.016
0.261
0.507
0.028
0.001
0.001
0.042
0.075
-0.081
-0.073
-0.004
0.010
0.014
1.67
0.0
0.280
-0.095
-0.125
0.001
-0.061
0.075
0.051
-0.057
for the natural
48-84
0.030
-0.001
0.007
0.009
0.025
0.823
0.121
0.005
-0.001
0.020
-0.045
0.024
-0.008
0.003
-0.002
-0.014
-0.098
0.898
-0.190
0.095
0.125
0.005
-0.041
0.019
0.017
-0.009
audit samples.
12-84
0.001
0.001
0.013
0.004
0.322
0.478
0.229
0.024
0.050
0.064
-0.073
-0.036
-0.078
-0.003
-0.008
-0.014
0.014
0.381
-0.150
-0.105
0.222
0.008
0.225
0.172
0.107
-0.018
RSOof
Natural
Audit
Data"
0.064
0.019
0.044
0.044
0.230
0.650
0.500
0.072
0.260
0.500
0.088
0.150
0.53
0.150
0.074
0.039
1.28
0.21
0.21
0.320
2.49
0.34
0.03
0.03
0.28
three streams, and (3) the mixed results in
precision in Table B-3.
The An/ons: SC^, NOl, and Cl
(Figure B-1, Parts H-J)
Holding time had a statistically signifi-
cant effect on SO*. NOg, and Cl" (p < 0.05,
Table B-2). The mean concentration of all
three variables decreased over time (Tables
B-2 and B-3). However, the practical signifi-
cance of these decreases is questionable. In
all cases, the relative difference between
holding times was less than the relative stan-
dard deviation of the ELS natural audit sam-
ples (Table B-3). Between 12 and 84 hours,
the mean concentration of SO*, NOg, and Cl"
decreased by 2.4 percent, 5 percent, and 6.4
percent respectively (Table B-3). Figure B-1,
Parts H-J, suggests no strong trend for any of
these parameters with respect to holding time.
Dissolved Inorganic and Organic
Carbon (Figure B-1, Parts K - M)
Holding time had a statistically signifi-
cant effect on the concentration of dissolved
organic carbon (DOC) and air-equilibrated
dissolved inorganic carbon (DIC^), but not on
the concentration of dissolved inorganic car-
bon (DIC) (Table B-2). The practical signifi-
cance of holding time, however, is question-
able. For all of these parameters, the relative
differences between holding times were
always less than the relative standard devia-
tions of ELS natural audit data (Table B-3).
Figure B-1, Parts K and M, suggest no strong
trend for these parameters with respect to
holding time. Between 12 and 84 hours, the
mean concentration of OOC increased by 11.5
percent while the mean concentration of DIC^
increased by 7.8 percent (Table B-3). Only
data from Fly Pond and Lick Run exhibited
63
-------
steady increases in OOC from 12 to 84 hours.
For OlCgq, this increase was noticeable only in
samples from Clear Creek and may have been
due to the introduction of carbon dioxide into
these samples.
(Figure B-1, Part N)
Holding time, had a significant effect on
this parameter (<0.05, Table B-2). The prac-
tical significance of the effect of holding time,
however, is questionable. The relative differ-
ences in mean concentration between holding
times were always less than the relative
standard deviations of ELS natural audit data
(Table B-3). Conductance exhibited no definite
trend over time and was stable for all streams
(Figure B-1, Part N). Samples exhibited a
mean increase of only 0.19 /uS/cm (0.3 percent)
from 12 to 84 hours (Table B-3).
(Figure B-1, Parts O-Q)
Plots of mean concentrations over time
for these three parameters (Figure B-1, Parts
O-Q) reveal no trend. Holding time did not
have a significant effect for these parameters
(p > 0.05, Table B-2). The relative differences
between holding times were always less than
the relative standard deviations of the ELS
natural audit data (Table B-3).
(Figure B-1, Parts R-T)
Holding time had a statistically signifi-
cant effect on the concentrations of total
extractable aluminum (Alext) and organic mono-
meric aluminum (Al ) (Table B-2). When the
data for Lyons Creek were removed, holding
time had a statistically significant effect (p <
0.05) on total aluminum (Altl).
When the data for total aluminum from
Lyons Creek were removed, the relative differ-
ence between 12 and 24 hours and between 24
and 48 hours exceeded the relative standard
deviation of the natural audit data by 4 percent
and 7 percent, respectively (Table B-3). The
relative differences between 48 and 84 and
between 12 and 84 hours were less than this
relative standard deviation by 2 percent and 3
percent, respectively. Figure B-1, Part R, sug-
gests stable concentrations for Altl between 12
and 84 hours. The mean concentration exhib-
ited both slight increases and decreases
(Table B-2), further suggestion that no overall
trend is present. For these reasons, the
practical significance of holding time on Altl is
questionable.
For AleKt, the relative difference between
holding times was always less than the rela-
tive standard deviation of the ELS natural audit
data (Table B-3). Figure B-1, Part S, suggests
no overall trend in the effect of holding time.
Although large fluctuations in the concentration
of Alext occurred for Big Moose Lake, the
remaining sites exhibited stable concentrations
between 12 and 84 hours. For these reasons,
the practical significance of holding time for
Ale)(t is questionable.
Natural audit data were not available for
AI0_, and this made an evaluation of the prac-
tical significance of holding time for this para-
meter more difficult. The mean concentrations
of Alorg were quite low (Table B-2). The con-
centration of AI0(B changed more for Clear
Creek and Big Moose Lake than for any other
lakes or streams during this experiment (Figure
B-1). The concentration of AI0(B for Clear Creek
increased by 0.008 mg/L (22 percent) between
12 and 24 hours and by 0.009 mg/L (19 per-
cent) between 24 and 48 hours. Between 48
and 84 hours, the concentration of
Clear Creek decreased by 0.025 mg/L
percent); this resulted in a net decrease of
0.008 mg/L (27 percent) between 12 and 84
hours. Samples from Big Moose Lake exhi-
bited a gradual increase of 0.011 mg/L (17
percent) between 12 and 84 hours. For the
remaining lakes and streams, the concentra-
tion of Alorg was stable between 12 and 84
hours. These results suggest that the prac-
tical significance of the effects of holding time
on the concentration of AlorQ is questionable.
"Ofs
in
(Figure B-1, Parts U-V)
Molding time had a statistically signifi-
cant effect on BNC, but not on ANC (Table
B-2). The practical significance of holding time
on BNC, however, is questionable. The relative
differences in BNC between holding times
were always much less than the relative
standard deviation of the natural audit data
(Table B-3). The small mean increase in BNC,
although not of practical significance, may
have been due to an introduction of carbon
-------
dioxide into some samples, particularly those
from Morgan Creak and Lyons Creek (Figure
B-1).
(Figure B-1, Parts W-Y)
Holding time had a statistically signifi-
cant effect on pH-alkalinity and pH-acidity, but
not on pH-equilibrated (Table B-2). Between 12
and 84 hours, pH-alkalinity and pH-acidity
decreased by an average of 0.107 and 0.152 pH
units (2.3 percent and 1.8 percent),
respectively. The introduction of carbon dio-
xide into some samples, particularly those
from Morgan and Lyons Creeks, may hava
caused this decrease. The differences
between holding times of 12 and 84 hours for
pH-acidity and pH-alkalinity were greater than
the absolute standard deviations for ELS
natural audit data (Table B-3). Figure B-1, Part
W and Part X, indicates steadily decreasing
levels of pH-acidity and pH-alkalinity from 12 to
84 hours. The difference between holding times
for pH-equilibrated was always less than the
absolute standard deviation of the ELS natural
audit data (Table B-3). Figure B-1, Part Y, indi-
cates stable levels of pH-equilibrated from 12
to 84 hours.
The decrease in pH is expected because
the audit sample has achieved equilibrium with
atmospheric CO2, while actual samples may be
either undersaturated or oversaturated with
respect to CO2. The direction of the shift in
pH will depend upon the proximity of the CO2
concentration of the sample to the atmos-
pheric CO2 partial pressure at the location
where the sample container is opened for
analysis. Samples undersaturated with
respect to the atmospheric partial pressure of
CO2 will decrease in pH because this gas
diffuses into solution. Samples oversaturated
will increase in pH because CO2 degasses out
of the sample. Thus the observed trend in pH
is an overall response of samples to a
decrease due to their CO2 concentrations.
Cubitainers appear to be unsuitable for
maintaining stability of pH (Schock and
Schock, 1982). Burke and Hillman (Appendix
A) observed stable pH readings for up to
7 days in samples held in sealed syringes and
stored at 4 °C.
Although holding time had a statistically
significant effect on 17 parameters, the practi-
cal significance in all cases is questionable.
When statistically significant effects of holding
times were observed for a parameter, at least
one of the following was true: (1) the relative
difference between holding times was less
than the relative standard deviation of the
natural audit data, or (2) graphically, effects of
holding time appear to be restricted to a
minority of the lakes and streams sampled
and are site-specific phenomena. A major
uncertainty in this study is the effect of
holding samples up to 12 hours after collec-
tion. In surveys covering large geographical
areas, it may be logistically impossible to filter
and preserve many samples less than 12 hours
after collection. If samples can be transported
to an analytical laboratory, filtered and pre-
served within 84 hours after collection, the
economic cost of a survey can be decreased
by eliminating field laboratories, by collecting
both samples with ground crews instead of
with helicopter crews or by shipping samples
by standard air instead of overnight courier.
Best, M. D., L W. Creelman, S. K. Drouse, and
D. J. Chaloud. 1987. "National Surface
Water Survey Eastern Lake Survey
(Phase I - Synoptic Chemistry) Quality
Assurance Report." EPA 600/4-86/011.
U.S. Environmental Protection Agency,
Las Vegas, NV. In preparation.
Drouse, S. K. "Evaluation of Quality Assur-
ance/Quality Control Sample Data for the
National Stream Survey." 1986. U.S. En-
vironmental Protection Agency, Las
Vegas, NV. In preparation.
Hillman, D. C., J. F. Potter, and S. J. Simon.
1986. "National Surface Water Survey
Eastern Lakes Survey (Phase I - Synoptic
Chemistry) Analytical Methods Manual."
EPA 600/4-86/009. U.S. Environmental
Protection Agency, Las Vegas, NV.
65
-------
Linthurst, R. A., D. H. Landers, J. M. Eilers, 0.
F. Brakke, W. S. Overton, E. P. Meier, and
R. E. Crowe. 1986. Characteristics of
Lakes in the Eastern United States, Vol.
I: Population Descriptions and Physico-
Chemical Relationships. EPA-600/4-
86/007a. U.S. Environmental Protection
Agency, Washington, D.C.
Messer, J. J., C. W. Ariss, J. R. Baker, S. K.
Drouse, K. N. Eshleman, P. R. Kaufman,
R. A. Linthurst, J. M. Omernik,
W. S. Overton, M. J. Sale, R. D.
Schonbrod, S. M. Stambaugh, and J. R.
Tuschall, Jr. 1986. National Surface
Water Survey: National Stream Survey,
Phase I-Pilot Survey. EPA 600/4-86/026.
U.S. Environmental Protection Agency,
Washington, O.C.
Permutt, T. J. and A. Pollack. 1986. "Analysis
of Quality Assurance Data for the
Eastern Lakes Survey." In Best, M. D., et
al. "National Surface Water Survey
Eastern Lake Survey (Phase I-Synoptic
Chemistry) Quality Assurance Report."
EPA 600/4-86/011. U.S. Environmental
Protection Agency, Las Vegas, NV. In
preparation.
Schock, M. R. and S. C. Schock. 1982. Effect
of Container Type on pH and Alkalinity
Stability. Water Research 16:1455-1464.
Winer, B. J. 1971. "Statistical Principles in
Experimental Design." McGraw-Hill: NY.
trU.S. GOVERNMENT PRINTING OFFICE: 1989 - 648-163/00344
66
------- |