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

-------