c/EPA
             United States
             Environmental Protection
             Agency
             Environmental Monitoring
             and Support Laboratory
             PO Box 15027
             Las Vegas NV 89114
EPA-600 3-79-030
March 1979
             Research and Development
Ecological
Research Series
Development of
a Strategy for
Sampling Tree Rings

-------
                  RESEARCH REPORTING SERIES

Research  reports of  the  Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series.  These nine broad categories
were established to facilitate further development and application of environmental
technology.   Elimination  of  traditional  grouping was consciously planned to foster
technology transfer and a maximim interface in related fields.  The nine sereies are:


       1.   Environmental Health Effects Research
       2.   Environmental Protection Technology
       3.   Ecological Research
       4.   Environmental Monitoring
       5.   Socioeconomic Environmental Studies
       6   Scientific and Technical Assessment Reports (STAR)
       7.   Interagency Energy—Environment Research and Development
       8.   "Special" Reports
       9.   Miscellaneous Reports
This report has been assigned to the  ECOLOGICAL RESEARCH  series.  This series
describes research on the effects of pollution on humans,plant and animal species, and
materials. Problems are assessed for their long-and short-term influences Investiga-
tions include formations,  transport, and  pathway studies to determine the fate of
pollutants and their effects. This work provided the technical basis for setting standards
to minimize undesirable changes in living organisms in the aquatic, terrestrial, and
atmospheric environments.
This document is available  to the public through the National Technical Information
Service, Springfield, Virginia 22161

-------
                                               EPA-600/3-79-030
                                               March  1979
              DEVELOPMENT OF A STRATEGY
               FOR SAMPLING TREE RINGS
                         by
                   Jerry A. Riehl
Northwest Environmental  Technology Laboratories, Inc
          Mercer Island, Washington  98040
              Contract No. CB-7-0771-A
                   Project Officer
                  Gilbert D. Potter
Monitoring Systems Research and Development Division
   Environmental  Monitoring and Support Laboratory
              Las Vegas, Nevada  89114
   ENVIRONMENTAL MONITORING AND SUPPORT LABORATORY
         OFFICE OF RESEARCH AND DEVELOPMENT
        U.S. ENVIRONMENTAL PROTECTION AGENCY
              LAS VEGAS, NEVADA  89114

-------
                                 DISCLAIMER
    This report has been reviewed by the Environmental  Monitoring and Support
Laboratory-Las Vegas, U.S. Environmental Protection Agency, and approved for
publication.  Approval does not signify that the contents necessarily reflect
the views and policies of the U.S. Environmental Protection Agency, nor does
mention of trade names or commercial products constitute endorsement or
recommendation for use.

-------
                                 FOREWORD

     Protection of the environment requires effective regulatory actions which
are based on-sound technical and scientific information,  This information
must include the quantitative description and linking of pollutant sources,
transport mechanisms, interactions, and resulting effects on man and his
environment.  Because of the complexities involved, assessment of specific
pollutants in the environment requires a total systems approach which tran-
scends the media of air, water, and land.  The Environmental Monitoring and
Support Laboratory-Las Vegas contributes to the formation aid enhancement of
a sound monitoring data base for exposure assessment through programs
designed to:

         develop and optimize systems and strategies for monitoring
         pollutants and their impact on the environment

         demonstrate new monitoring systems and technologies by
         applying them to fulfill special monitoring needs of the
         Agency's operating programs

     This report documents the development of a strategy to retrospectively
monitor pollutant levels in air by measuring arsenic concentrations in samples
of wood obtained from annual growth rings of trees in the vicinity of a
smelter.  Such information can provide a basis to assess the relative
risks to the health and well -being of a local  population exposed to environ-
mental hazards and to aid in the improvement of our environmental quality.
The advantage of such a program will  permit us to assess the effectiveness
of past environmental control technologies with those of the future and
establish an environmental baseline with which to compare future control
technologies.

     Federal and local agencies interested in assessing past exposures of
local populations to hazardous or carcinogenic materials released to the
environment will find this study useful.  In addition, the information it
provides may be used to determine the effectiveness of newer control  method-
ologies.  Additional  information about the study may be obtained by contact-
ing the Monitoring Systems Research and Development Division, Environmental
Monitoring and Support Laboratory, Las Vegas,  Nevada.


                                        '
                                      George B. Morgan
                                          Director
                       Environmental Monitoring and Support Laboratory
                                          Las Vegas
                                     ill

-------
                                  ABSTRACT
    A method for determining retrospective pollution levels has been
investigated.  This method relates arsenic concentration in tree rings to
arsenic-in-air concentrations based qualitatively on arsenic emissions from a
nearby smelter, corrected for climatological  and meteorological effects.   To
evaluate the validity of the method, a unique pollution study area was
identified and characterized in detail.  Several select trees were sampled
and the arsenic concentration determined by neutron activation analysis.
These concentrations were compared to certain known phases in the production
history of the smelter, coupled with the expected climatology and meteorology
of the area.  Positive correlations were found thus satisfying the goals  of
the preliminary project.  Major problems encountered were low arsenic
concentrations and an inadequate number of samples.  Recommendations for
future studies are given.
                                      iv

-------
                                  CONTENTS
Foreword	     i i i
Abstract	      i v
Figures	      vi
Tables	     yii
Acknowledgment	    viii

Section

    1.  Introducti on	       1
    2.  Summary	       2
    3.  Recommendations	       3
    4.  Background Information	       4
             Tree ring analysis	       4
             Probl em i denti fi cati on	       5
    5.  The ASARCO Tacoma Smelter Study Area	       7
             Uniqueness of the study area	       7
             Uniqueness of the sampling sites	       7
             Other important considerations	       9
    6.  The Hi story of the Tacoma Smel ter	      10
             Production and emissions	      10
             Control efforts	      12
    7.  The Sampling Site	      13
             Location	      13
             Subject trees	      13
             Soil	      16
    8.  Experimental Systems and Techniques,	      17
             Neutron activation analysis	,	      17
             Elements analyzed	,	      17
             Tree sample collection	,	      18
             Soi 1 sample col 1 ection	      20
    9.  Meteorological Information	      22
             Climatological data	      22
   10.  Experimental Results and Conclusions,..	      26
             Introduction	,	      26
             Experimental results	      27
             Discussion	      33
             Concl us i ons	      37

References	      38

-------
                                   FIGURES
Number                                                             Page

  1      The ASARCO Tacoma smelter study area	     6
  2      Air pollution monitoring network surrounding
              ASARCO smel ter i n Tacoma	    8
  3      Topographic map of southern portion of Maury
              Island with site of study trees identified	   14
  4      Photographs of the study area	   15
  5      Annual precipitation (A) and 10-year running
              mean (M), 1893-1970	   25
  6      Comparison of upper and lower tree rings - tree #5	   28
  7      Yearly variation of arsenic concentration - tree #5	  34
  8      Yearly variation of arsenic concentration - tree #1	  35
                                    VI

-------
                                   TABLES
Number                                                             Page

  1      Total  Production for the Tacoma ASARCO Smelter/
              Refinery, 1971-1975	     11
  2      Description of Trees Used in the Analysis	     16
  3      Description of Tree Ring Samples	     19
  4      Description of the Soil  Samples	     20
  5      Temperature Statistics (C) for Tacoma, Washington,
              for the Period 1930-1960	     23
  6      Precipitation Statistics (cm) for Tacoma, Washington
              for the Period 1930-1960	     24
  7      Arsenic Content, in Study Trees	     30
  8      Arsenic Content in Tree #1, Sample 1-L ;*	,.,..,..     31
  9      Arsenic Content in Trees #1 and #5 During
              Strike Years	     31
 10      Arsenic Content in Needles, Soil  and Cones	     32
                                    VI1

-------
                              ACKNOWLEDGMENT
    This report is the result of the cooperation, assistance, and
participation of many individuals.

    Personnel of the U.S. Environmental Protection Agency's Region X Office
provided considerable guidance and information during the course of this
study.  These include Dr. Robert Courson, Dr. James Everets, and Mr. Robert
Coughlin.  Their efforts were needed and appreciated.

    I am also greatly indebted to many staff members of the Puget Sound Air
Pollution Control Agency for their valuable input, especially in historical
documentation.

    For allowing me to freely choose trees for sampling on their property,  I
am equally indebted to Mr. and Mrs. Alan Houston and Mr. Lewis Duncan.  Their
interest in this project and enthusiasm to cooperate were heart-warming.

    Finally,  I wish to acknowledge the input of Dr. William Kreiss, a
meteorologist with Physical Dynamics,  Inc., of Seattle.  His general guidance
and encouragement were appreciated and his meteorological expertise made
Section 9 possible.

-------
                                  SECTION 1


                                INTRODUCTION
    The primary difficulty in relating the incidence and causation of human
disease to environmental  factors in epidemiological  studies results from the
lack of good retrospective data.  Because many diseases develop over a period
of years, it is of major importance to determine an  accurate chronology of
those items that lead directly to the initiation of  the disease.   Few
accurate methods exist, especially in the area of pollution-caused diseases.
This project is a first step in the development of such a method.

    The technique of dendrochronology is most often  used to estimate the age
of a tree.  By carefully evaluating the successive layers of xylem growth, or
"tree rings" as they are commonly called, from the pith to the cambium, the
age of a tree can be accurately determined.   This technique is widely known
and used.  Also, local  climatic history can  be reconstructed from  a detailed
analysis of tree rings.  In general, the width of the ring is related to the
precipitation and temperature exposure.  In  addition, patterns of  varying
ring widths are established over a period of years due to exposure to various
weather conditions, and these can be traced  from tree to tree in a given
climatic area.  These patterns have been used to date items ranging from the
age of windfalls to the age of objects constructed from wood.

    Directly related to this, but even less  well known, is the fact that
other information is also stored in the rings of xylem growth. As will be
discussed in this report, some trace chemical elements known to be
nonessential  for tree growth are found in the xylem.   Furthermore, the
concentration of these  nonessential  elements in tree  rings can vary from year
to year, and it has been demonstrated that they can  be related to  a pollution
source many kilometers  away.  Such elements  possibly  enter the tree via the
soil by rainout or fallout.

    The development of  this method for retrospective  determination of
pollutant levels is based on the above facts.  The subject of this study is
the demonstration that  the concentration of  some pollutants in the rings is
proportional  to the pollution to which the trees are  exposed.

-------
                                  SECTION 2


                                   SUMMARY
    The concentration of arsenic in tree rings was found to be proportional
to the annual  emission of arsenic from a copper smelter 8 kilometers away
from the trees.  The study site is on an island, and so transport from the
stack by air has been validated.  Although positive correlations were made,
several anomalies exist that need to be answered.  An apparent lag of one to
two growing seasons in the uptake of arsenic from the soil  was identified in
tree rings taken from lower portions of the tree.  On the other hand, no lag
was noted between production output from the smelter and tree rings taken
from higher elevations in the tree, suggesting a direct uptake from the air
through the needles.  Although the evidence is not conclusive, some species
of trees appear to behave differently from others, but all  species evaluated
showed the lag effect in samples collected nearest the root system.

-------
                                  SECTION 3


                               RECOMMENDATIONS
    The data contained in this report suggest that tree ring analyses  are
valid indicators of past pollution levels.   This presents the possibility of
retrospective evaluation of pollution.   Several  important steps must be
taken, however, before the method can be demonstrated  to be quantitative.

    First, pollutants other than arsenic need to be evaluated for this study
area.  Second, a larger number of sample trees needs to be evaluated.   Third,
other species of trees need to be evaluated in greater detail  than was
possible in this study.  Fourth, duplicate  increment samples should be taken
during the sampling stage, and one of these should be  mounted,  polished and
varnished in order to eliminate or at least minimize ring-counting errors.
Fifth, definite climatological-growth relationships need to be  developed, a
major factor in this type of study.  Finally, higher neutron flux should be
used for the analysis to maximize the pollutant  activity in the rings.

    Whereas this study was conducted  in a short  time and on a limited  budget,
future studies should be extended in  time and budget in order to accurately
assess all of the possible variables  and draw conclusive results.

-------
                                  SECTION 4


                           BACKGROUND INFORMATION
TREE RING ANALYSIS

    The trace element concentrations in the xylem ring of trees are known to
be proportional to the elemental concentration in the soil  (Lepp,  1975).
Early studies  (Sheppard and Funk, 1975) have shown that the rings of trees
whose roots were exposed to contaminated water in an Idaho river downstream
from a mine contained trace-metal concentrations that were directly
proportional to the yearly pollution levels of the river and the yearly
output of the mine.  Tree rings examined in certain Pennsylvania trees
(Pillay, 1975) contained concentrations of trace metals known to be present
in the local atmosphere, and the observed variations in the rings suggested a
relationship proportional to the pollution in the environment.  Heavy metal
studies by Ault et al. (1970) and Ward et al. (1973) have shown that lead
levels in tree rings could be correlated with local traffic density.  On the
other hand, a detailed study by Szopa et al. (1973) showed that lead levels
in oak trees near an abandoned highway continued to register high lead levels
after the abandonment, suggesting that direct yearly correlations may be in
error.  (This  study, however, did not consider the possibility of root uptake
of lead from the soil.)  Finally, unpublished tree-ring data taken by the
author on trees located in polluted and unpolluted environments have clearly
indicated that certain environmental contaminants present in air pollution
are concentrated in tree rings, that the concentrations of these elements
appear to be a function of the type of soil in which the tree is located, and
that the concentration varies annually.

    While all  of these results indicate that a relationship may exist between
the concentration in the rings and the local environment, a detailed analysis
of all relevant variables was needed in order to evaluate the validity of
this method as a tool for retrospectively determining pollution levels.  This
report describes the progress to date concerning the first step in this
verification process—a simple demonstration of the correlation between
tree-ring concentration and the polluted environment of the tree.

-------
PROBLEM IDENTIFICATION

    Several  parameters needed to be determined in order to develop this
method for use as a technique for retrospective determination of pollution.
These included tree-related parameters.  Processes related to the uptake,
transport, and deposition of trace elements in the tree needed to be
understood.   Items such as diffusion of trace elements to adjacent rings,  the
preferential  uptake of certain species from the soil,  the effects of soil
diffusion and leaching, the uptake of the elements directly from the air
through broad leaves and needles, all  needed to be understood.  Equally
important were the pollution source-related parameters.  An accurate
chronology of the source emissions was needed, as well  as the various
transport processes and mechanisms.  In addition, a well-defined
meteorological  history was mandatory.   No one parameter stood out as most
important.  All  were related in one way or another and had a direct  effect on
the accuracy of the method—hence on the demonstration of the method
validity.

    A most important first step in the demonstration of the validity of this
method was the identification and characterization of  an ideal study area.
This unique  site was found to be located in a second growth area in  the Puget
Sound of northwestern Washington.  The area contained  a single major source
of pollution, and the only direct means of pollutant transport was by air
(Fig. 1).  Equally important was the tree-ring concentration/pollution
environment  correlation mentioned above.  This has been verified, at least
qualitatively, and the remainder of this report discusses the results.

-------
                                                 GREATER PUGET
                                                    SOUND
Figure 1.    The ASARCO Tacoma smelter study area

-------
                                  SECTION 5
                    THE ASARCO TACOMA SMELTER STUDY AREA
    Several  important parameters qualified the selected study area as  ideal
for this demonstration project.  These were related to four distinct factors
that substantially reduced the level  of error: a well-defined and
characterized study system; ideal  sampling sites; ideal  chemical  tracer
elements and techniques for the analysis; and an established meteorological
history of the area.


UNIQUENESS OF THE STUDY AREA

    Much scientific work related to this project, and especially  to the
overall study area, had been done in  the recent past.  Tree-ring  analyses
previously performed by the author had formed the basis  for this
demonstration project.  Detailed studies on the particulate (Nelson and
Roberts, 1975) and gaseous (Washington State, 1976) emissions had been
conducted; dispersion modeling of the gaseous emission had  been accomplished
(Cramer et al., 1976); an extensive air pollution monitoring system was in
use (Fig. 2); the surrounding water and sediments between the source and the
sampling sites on Maury Island (Fig.  2) had been characterized in detail
(Crecelius et al., 1975); and many other pollutant-related  studies, ranging
from arsenic levels in hair and urine (Johnson and Lippman, 1973) to cadmium
levels in vegetable gardens (Heilman  and Ekuan, 1977), had  been conducted.
In general,  considerable information  was available on the selected study
area.
UNIQUENESS OF THE SAMPLING SITES

    Two distinct sampling sites were identified near the City of Dockton  on
Maury Island (see Fig. 2).  Maury Island and Vashon Island  are bridged  with
fill  area and lie west of Seattle and north of Tacoma,  the  two major urban
areas in Puget Sound.  Although close to these cities,  they are virtually
isolated (ferry connections only).  Both islands consist mainly of small
farms and second growth timber, have little or no industrial  sources, and

-------
Figure 2.   Air pollution monitoring  network  surrounding
           ASARCO smelter in Tacoma
                           8

-------
for the most part have no local  anthropogenic stationary pollution sources.
The prevailing winds are northerly and southwesterly, depending on the
season, and both sampling sites lie within the southwesterly plume envelope
of the smelter.  Since Tacoma is bordered on the north by Puget Sound, the
only possible path from the smelter to the sampling sites is by air
transport. The sites lie about 8 kilometers (km) from the smelter, within
known plume touchdown areas.  In Section 7, these sites are described in
detail and the trees that will be studied are illustrated.
OTHER IMPORTANT CONSIDERATIONS

    The experimental  techniques and the meterological  considerations employed
in the study are described in Sections 8 and 9.  They illustate the overall
importance of these parameters in the demonstration project.   In Section 10,
the experimental results are discussed.

-------
                                  SECTION 6


                      THE HISTORY OF THE TACOMA SMELTER
PRODUCTION AND EMISSIONS

    Detailed chronologies of the arsenic production levels and the
corresponding stack emission levels from the American Smelting and Refining
Company (ASARCO) smelter in Tacoma were either non-existent or unavailable
to the public.  However, certain specific facts related to"the stack
emissions were known and hence enabled an evaluation to be made concerning
the validity of tree-ring analysis as a method for retrospective
determination of pollution levels.

    The Tacoma smelter began operation in 1890 as a lead smelter.   In 1902
copper smelting capabilities were added (Cramer et al., 1976;  ASARCO, 1976).
ASARCO bought the smelter in 1905 and used it for both lead and copper
smelting until 1911 when it was converted to a copper-only facility.  The
smelter was used to process both copper ore and copper concentrates into
blister copper.  In 1915, a refinery was added to treat the blister copper
produced at the smelter.

    Since both lead and copper ores contain arsenic in varying amounts, it
follows that arsenic could have been produced at the smelter as early as
1890.  Consequently, a detailed analysis of the tree rings between 1885 and
1915 for arsenic provided valuable information about the validity of
tree-ring analyses, especially concerning diffusion between the rings.  An
evaluation of Section 10 shows this to be true for certain species of trees.

    Theoretically, arsenic emissions from the stack should have been
proportional to the yearly production of arsenic, as is implied above,
corrected of course for the addition of pollution abatement equipment on the
stack.  But ASARCO considered its production information proprietary.
However, certain known facts enabled an evaluation of the validity of the
method.  First, some production data had been released to the U.S.
Environmental Protection Agency (EPA)*  This is shown in Table 1.   Second,
*Coughlin, R. L., private communication, May, 1977
                                     10

-------
although more indirect in nature, it was known that the production levels
were especially low in 1967 and 1968 due to a 9-month  strike  from July 1967
to April 1968.  Because the arsenic in question in  this project  was suspected
of being incorporated into the tree structure during the spring  growth
portion of the yearly cycle, the 1968 (or 1969, depending  on  the lag  between
stack emissions and appearance in the tree) tree rings should have been  low
in arsenic.  A similar situation should have existed in the 1959-60 period
where again a strike of many months duration lowered production  and emissions
levels considerably.  Again, the evidence presented in an  evaluation  of
Section 10 indicates this is indeed true.
                                  TABLE  1.

                   TOTAL PRODUCTION FOR  THE  TACOMA  ASARCO
                   SMELTER REFINERY, 1971-75 (SHORT TONS)


Year
1971
1972
1973
1974
1975
Material
processed
380 x 103
406 x 103
384 x 1()3
350 x 103
333 x 103
Smelter Cu
output
88.2 x 103
99.8 x 103
95.5 x 103
86.6 x 103
72.3 x 103
Refined Cu
output
NA
NA
120.1 x 103
117.4 x 103
119.7 x 103
As203
output
9.8 x 103
13.3 x 103
13.1 x 103
9.7 x 103
NA

Finally, the United States Bureau  of Mines  Annual  Minerals  Yearbook reported
information in 1961, 1962, and 1963 (USBM,  1961;  1962;  1963) that was hoped
could yield a valuable fourth set  of data points  that could have been
utilized in the evaluation.  The 1961 book  reported  a 17% decrease in white
arsenic production (As203—nearly  100% of the  arsenic produced  in the
United States is this species) from the 1960 level;  the 1962 book reported an
increase of 7 percent over 1961; and the 1963  book reports  a 4-percent
decrease from the 1962 production  level. Unfortunately, these  small changes
could not be accounted for because of the variability found in  the data.  A
detailed explanation of this can be found in Section 8.  A  continuing effort
is underway to obtain as much white arsenic production  data as  possible for
incorporation into future phases of the project.
                                     11

-------
CONTROL EFFORTS

    The control of arsenic emissions from the smelter has been both direct
and indirect.  An indirect control resulted from a "shutdown"  policy that
ASARCO followed in the past few years.  When meteorological  conditions forced
gaseous S02 concentrations around the smelter to exceed certain values, the
smelter was automatically shut down.  A detailed search for information on
the "down time" of the smelter was not fruitful, as no records were
available.*  As a direct control, ASARCO started a pilot baghouse in early
1974 in order to determine the best available technology for reducing stack
particulate emissions.  This baghouse was completed near the end of 1976,
according to ASARCO, (ASARCO, 1976) and is still  in operation.  No effect  on
1977 emissions could be determined.  To date, these two control efforts
represent the total control operations that affected arsenic output from the
smelter.
*This control methodology is explained in detail  by Cramer et al.  (1976)  but
no record was available concerning the frequency of these shutdowns..
                                     12

-------
                                  SECTION 7


                              THE SAMPLING SITE
LOCATION

    As mentioned earlier, the two sampling sites used in this project are
located on Maury Island near the town of Dockton, Washington (see Fig.  2),
are in a known smelter plume touchdown area, and lie about 8 km downwind from
the stack.  Site No. 1 (the major site) lies to the southwest of Dockton,
while Site No. 2 (a backup site) lies northwest of the town in the Dockton
King County Park.  Both sites are shown in Figure 3.  Site No. 1 was used for
the initial phase of the project for several reasons:  the trees were of the
desired age; several species of trees were available for incremental  coring;
and the trees were clustered together in a small  area, with each tree having
a similar environmental exposure.  The trees at Site No. 2 were marginal  in
age, of a single species, and more widely scattered—thus, less desirable for
this study.  Permission was granted to the author by both landowners to
collect all of the samples needed.

    Site No. 1 (see Fig. 3) is described legally as:  King County Property,
30-22-03, tax lot 9035, that portion of the north one-half of GL 4 measured
on east line ELY of Manzanita Road, less the county road.  The total  size of
the site is 13.27 acres (53,702 square meters (m2)), and the five subject
trees are located on the site within 100 m of each other.  Figure 4d shows  a
southerly view of the smelter as viewed from the reference point on Figure  3.


SUBJECT TREES

    Previous studies by the author indicated that Pseudotsuga taxi folia
(Douglas Fir) trees provided adequate yearly sample sizes and reasonable
pollutant sensitivities for an arsenic analysis in individual tree rings.
Unfortunatley, finding these trees of the proper age (older than 1890)  and  in
sufficient quantity in close proximity to each other was not possible in the
vicinity of the smelter.  Consequently, Abies grandis (Grand Fir) and Thuja
plicata (Western Red Cedar) were also sampled.  Table 2 describes the trees
selected from study Site No. 1.  This mix was considered acceptable in  that  a
comparison between various types of trees was desired.  Since the increment
borer that was used for tree ring extraction was of limited length and  did
                                     13

-------
                  •   v-Yxiij:^.
                 v^v^M'   \'1'
Manzanfta.
 N

 t
0.5 km   1.0 km
         Figure 3,  Topographic map of southern portion of Maury
                  Island with site of study trees  identified
                               14

-------
(a)  Trunk view of tree #1
   (b)  Trunk  view of tree #5
(.c) Trunk view of tree #3

             Figure 4.  Photographs of the study area
(d)   ASARCO smelter as viewed from
     reference point on Figure 3
                                 15

-------
not penetrate to the core, the age of the larger trees was estimated.  This
and other sampling information will be discussed in detail in Section 8.
Figures 4a, 4b and 4c show pictures of trees  1,  3 and  5.


             TABLE 2.  DESCRIPTION OF TREES USED IN THE ANALYSIS
                                                  Age           Diameter
Tree #               Name             Type       (years)        (meters)
  1        Abies  grandis            Grand Fir      est 130          1.3

  2        Pseudotsuga  taxi folia    Douglas Fir    est 100          1.0

  3        Abies  grandis            Grand Fir           85  -        0.9

  4        Thuja  plicata            Western Red         93          0.9
                                    Cedar

  5        Abies  grandis            Grand Fir      est 110          1.0
 SOIL

    The soil  at  the sampling site  is  called  Everett-Alderwood  (EwC), which is
 a  gravelly sandy loam,  usually characterized by a 6 to 15 percent slope
 (USDA,  1973).   Both Everett and Alderwood  soils are somewhat/similar in that
 they  are well-drained,  dark to grey in  color,  and situated on  a moderately
 strong  substratum.   The soil depth is typically 75 centimeters (cm) and, in
 the undeveloped  states, covered with  timber.  The soil samples that were
 collected closely resembled this description.  As will be discussed in
 Section 8, a  20-cm layer of litter, composed primarily of decaying leaves and
 needles, covered the soil.
                                      16

-------
                                  SECTION 8


                     EXPERIMENTAL SYSTEMS AND TECHNIQUES
NEUTRON ACTIVATION ANALYSIS

    The arsenic determinations in the tree rings were made with neutron
activation analysis (NAA) at the University of Washington Nuclear Reactor
Laboratory in Seattle and the Washington State University Nuclear Reactor in
Pullman.  The University of Washington facility has a graphite-moderated
reactor with a thermal neutron flux of about 10*2 neutrons per centimeter
squared per second.  Washington State University has a swimming pool  reactor
with a thermal neutron flux of about 1Q13 neutrons per centimeter squared
per second.  The U of W reactor flux proved to be marginal  for the analysis
of the trace amounts of arsenic found in the small tree-ring samples;
however, the larger WSU reactor was adequate.  The WSU reactor was to  be used
if arsenic sensitivity proved to be a problem, but was shut down during the
early stages of the project for reactor control  system modifications.  The
use of the smaller flux reactor placed constraints on the earlier analyses
and necessitated the use of two yearly rings per sample in most cases instead
of one.  While this limited the anticipated quantitative accuracy and
affected the proposed yearly tree pollutant-smelter production correlations,
good qualitative and semiquantitative conclusions could be drawn.

    High resolution germanium-lithium (Ge-Li) detectors with state-of-the-
art, computer-supported gamma-ray analyzer systems were used to analyze the
gamma-ray spectra.  All quantitative determinations for 7fy\s were performed
internally by computer employing the photopeak analysis method developed  by
Korthoven (1970).


ELEMENTS ANALYZED

    Arsenic was the only element analyzed in this study.  While previous
unpublished work by the author has shown that other pollutant species were
also present in the tree rings taken from Pseudotsuga taxifolia (Douglas  Fir)
in the same general vicinity, arsenic was selected for a variety of reasons.
First, it is not native to the trees in the study area.  Also, it is  emitted
in large amounts from the ASARCO smelter (Crecelius et a!., 1975), probably
of the order of 2 x 10^ kilograms AS203 per year.  Third, and
                                     17

-------
equally important, natural arsenic consists of 100 percent 75/\s which, as
the target material, has  a good-sized thermal neutron capture cross section
of 4.5 barns; the resultant radioactive species, 76As, has an ideal
half-life of 26.4 hours for the analysis  (Lederer et al . , 1967).  Typically,
arsenic can be detected in concentrations as small as 1 microgram per gram
(ug/g) of material with NAA.  The nuclear reactions that characterize this
analytical determination  are,

       75As + n— *76As + capture   7 -ray

                      +  0-ray +  7 -ray (559 keV)
Other higher energy  gamma  rays  are  also  emitted in the 7^As decays, but
their intensities are  less than the 559-keV photopeak.


TREE SAMPLE COLLECTION

    Several tree-ring  core samples  were  collected from each of the subject
trees  (Table 3).  Two  complete  sets of cores were taken from each tree and
one set was retained as a  U.S.  Environmental Protection Agency archive.  In
addition,  several of the cores  collected were either  retained as samples or
used as  "practice cores" to develop ring-counting, ring-separation, drying,
and weighing   techniques in the laboratory.

    Pruning spray was  employed  to keep tree infection to  a minimum, and a
5-millimeter  (mm) diameter by 38-centimeter (cm) long stainless steel
increment  corer was  used to extract the  cores.  The instrument was kept in an
organic  oil when not in use to  eliminate metal  oxidation  and subsequent
sample contamination.   A silicone lubricant was sprayed on the corer during
operation  to  reduce  friction.  Later examination showed that neither of these
materials  introduced contamination in the ring  samples and no evidence of any
elements characteristic of stainless steel was  noted  in the gamma-ray
spectra.

    All  sample cores were  placed in new  glass tubes in the field and sealed
to reduce  subsequent contamination.  The  tubes,  designed by the author, were
washed  in  concentrated hydrochloric acid and  rinsed with  distilled water
prior  to use  in the  field.  The cores were typically  30 cm long, since that
was the  maximum obtainable with the instrument. This allowed an 85- to
 125-year-old  sample  to be  taken, a range which  proved satisfactory for the
 study.
                                      18

-------
                 TABLE 3.  DESCRIPTION OF TREE RING SAMPLES
Sample
Type
Height above
 ground (m)
Location on tree
   (north 0°)
1-L-A Abies grandis
1-L-B
1-L-C
1-M-A
1-M-B
1-U-A
1-U-B
2-L-A Pseudotsuga
2-L-B taxi folia
2-U-A
2-U-B
3-L-A Abies grandis
3-L-B
3-M-A
3-M-B
3-U-A
3-U-B
4-L-A Thuja plicata
4-L-B
4-L-C
4-M-A
4-M-B
4-U-A
4-U-B
5-L-A Abies grandis
5-L-B
5-M-A
5-M-B
5-U-A
5-U-B
1.4
1.4
1.4
4.8
4.8
5.4
5.4
1.4
1.4
3.8
3.8
1.4
1.4
3.7
3.7
4.3
4.3
1.4
1.4
1.1
4.3
4.3
5.6
5.6
1.4
1.4
4.0
4.0
4.9
4.9
90°
85°
95°
90°
85°
90°
85°
180°
355°
180°
355°
270°
275°
270°
275°
270°
265°
100°
90°
270°
90°
85°
85°
90°
270°
275°
270°
265°
270°
265°

L, M, U = lower, middle,  upper sections  of  tree
                                     19

-------
    Since most of the trees were located on a hillside with a western
exposure, the majority of samples was taken from either the east or the west
side of the tree so that the ladder used could be placed on reasonably level
ground.  The one exception was the Douglas Fir which was located on an
incline with a northern exposure.


SOIL SAMPLE COLLECTION

    Soil samples were collected with a 2.5-cm diameter soil corer.  The
length of the corer was about 15 cm, but the instrument was designed with
extension arms which allowed sample cores to be taken up to 60-cm depth.  In
general, about 20 cm of litter overburden was present and most soil samples
were collected between 20-cm and 35-cm depth.  Table 4 describes the soil
samples in detail.  All soil samples were placed in clean, scalable
polyethylene vials in the field to keep contamination to a minimum.  The
typical size of each sample was 2.0-cm diameter by 2.5-cm length .  A
stainless steel spatula was used to cut the samples to length, while in the
corer and to transfer them to the vials in the field.  Due to the dampness of
the soil, most samples remained intact until removed for drying and weighing
in the laboratory.
                  TABLE  4.  DESCRIPTION OF THE SOIL SAMPLES
Sample
Tree
Material
Depth from
surface (cm)
Location from tree
(m)  (north = 0°)
1-S-l
l-S-2
l-S-3
l-S-4
l-S-5
l-S-6
5-S-l
5-S-2
5-S-3
5-S-4
5-S-5
5-S-6
5-S-7
1
1
1
1
1
1
5*
5
5
5
5
5
5
Litter
Soil
Soil
Soil
Soil
Soil
Litter
Litter**
Soil
Soil
Soil
Soil
Soil
10 - 12.5 1.0 60°
20 - 22.5
22.5 - 25
25 - 27.5
27.5 - 30
30 - 32.5
10 - 12.5 0.8 270°
20 - 22.5
22.5 - 25
25 - 27.5
27.5 - 30
30 - 32.5
32.5 - 35

*  This core  was  taken midway  between trees 4 and 5
** Approximately  2.5  cm  of decaying bark  was encountered between
   the litter and soil.
                                      20

-------
    Needle samples were extremely difficult to  obtain  in  that  the  lowest
branches on most trees were over 15 meters (m)  above the  ground.   Tree #1
did, however, have a small  branch at about 7 m  and  a limited number  of Grand
Fir needles were otained.   In order to have a comparison  with  other  needles,
both live needle and cone  samples were collected  from  a small  4-year-old
Grand Fir that lies about  10 m southwest of tree #2 (Douglas Fir)  and about
60 m due east of tree #1.   In addition, dead needles were collected  from the
base of tree #1 and tree #5 while collecting the  soil  samples.  All  samples
were placed in clean polyethylene vials to prevent  contamination.
                                     21

-------
                                  SECTION 9


                         METEOROLOGICAL INFORMATION
    Crucial to accurate chronological dating of past air pollutant emissions
from the ASARCO smelter was a complete record of local  meteorology for the
period of concern.  These data were needed to accurately define dispersion
from the stack to the site, and to establish a normal climatological-growth
relationship for trees at the site.  This relationship was needed for two
reasons: to ensure that trees at the site, which is an area of wet climate
and lack of temperature extremes, do reflect past climatic changes, and to
ensure that the samples taken from the trees reflect normal growth patterns
unaffected by local anthropogenic effects.

    Since complete records of temperature, precipitation, and wind existed
for Tacoma, Seattle, and Vashon Island from 1890 to the present, the author
believed that qualitative relationships could be established for related
tree-ring concentrations and stack emissions. However, many of the
meterological parameters needed to accurately define stack plume dispersion
were not available.  Furthermore, the vast amount of data needed to verify
the climatological-growth relationship for the Puget Sound area did not exist
at the start of the project, thus creating a problem from the outset.

    The analytical sensitivity for arsenic became a problem early in the
analysis; therefore, full-scale dispersion modeling was not required to
interpret the data and a simpler evaluation was made.  The relative freedom
of the study area from localized weather events made it possible to
extrapolate with good reliability the available Tacoma, Seattle, and Vashon
Island climatological data, thus providing the necessary weather/climate
descriptors for the study area.  From this, qualitative estimates of the
yearly tree growth and yearly deposition of arsenic were made.  A comparison
of the growth estimates with the weights of the specific rings was made in
order to evaluate the accuracy of and verify qualitatively the
climatological-growth relationship.


CLIMATOLOGICAL DATA

    Temperature and precipitation data for Tacoma for the period 1930-1960
were taken from Phillips' climatological summary (Phillips, 1960).  Table
                                     22

-------
5 lists the maximum, minimum, and average temperature by months as well as
the record maximum and minimum temperatures for each month of the period
denoting the year in which they occurred.  An evaluation of the annual
results at the bottom of the table shows 1955 to be a year of extremes in
temperature, an important fact that will be discussed later.

    Table 6 lists the mean monthly precipitation for the 1930-1960 period
along with the maximum and minimum monthly averages for the period, and the
monthly 24-hour maximum.  The significant anomaly is the 47.9 cm of
precipitation which occurred in December 1933.  This is also important and
will be discussed later on.  The tabulation of 24-hour maximum precipitation
reinforces the previous statement that 5 cm of rain in 24 hours is a heavy
rainfall for the area.  Figure 5 shows the annual  precipitation for the study
area, along with the 10-year running mean for the years 1893-1970.
         TABLE 5.  TEMPERATURE STATISTICS (°C) FOR TACOMA, WASHINGTON
                   FOR THE PERIOD 1930-1960

Month
Daily
maximum
Daily
minimum
Monthly
average
Record
high
Record
low
Jan.
Feb.
Mar.
Apr.
May
June
July
Aug.
Sept.
Oct.
Nov.
Dec.
Ann.
 7.2
 9.4
11.3
15.0
18
20
23.4
23.0
20.1
15.5
10.6
 8.6
15.2
.3
.5
 1.2
 2.6
 3.7
 5.8
 8.5
10.9
12.6
12.7
10.8
 7.8
 4.3
 3.1
 7.1
 4.4
 5.9
 7.5
10.4
13.4
15.7
18.0
17.8
15.4
11.7
 7.5
 5.8
11.2
19.4 (1935)
22.8 (1938)
22.8 (1934)
30.0 (1934)
32.8 (1936)
36.1 (1955)
35.6 (1958)
35.0 (1936)
31.1 (1944)
27.8 (1931)
20.6 (1949)
18.3 (1934)
36.1 (1955)
-12.2 (1950)
-11.7 (1950)
 -7.8 (1955)
 -4.4 (1955)
 -1.1 (1955)
  2.8 (1955)
  5.6 (1957)
  7.8 (1935)
  2.8 (1948)
 -1.1 (1935)
-13.3 (1955)
-10.0 (1932)
-13.3 (1955)
                                     23

-------
TABLE 6.  PRECIPITATION STATISTICS (cm) FOR TACOMA,
          WASHINGTON, FOR THE PERIOD 1930-1960


Month
Jan.
Feb.
Mar.
Apr.
May
June
July
Aug.
Sept.
Oct.
Nov.
Dec.
Annual
Mean
total
13.6
10.6
9.7
6.0
4.1
3.7
1.9
2.1
4.5
9.7
12.6
15.6
94.1
Monthly
maximum
24.3 (1953)
18.6 (1932)
18.1 (1950)
13.4 (1938)
11.2 (1948)
14.2 (1946)
7.6 (1948)
5.7 (1948)
10.0 (1933)
22.4 (1947)
24.8 (1937)
47.9 (1933)
47.9 (1933)
Monthly
minimum
1.7 (1949)
3.9 (1934)
4.6 (1944)
0.6 (1955)
0.4 (1947)
0.2 (1951)
0.0 (1958)
0.2 (1955)
0.5 (1942)
1.3 (1936)
2.0 (1952)
5.0 (1930)
0.0 (1958)
24- hour
maximum
6.9 (1935)
7.9 (1951)
5.5 (1948)
3.5 (1937)
2.9 (1948)
5.0 (1936)
2.7 (1948)
5.0 (1936)
- 5.4 (1945)
6.1 (1934)
7.0 (1937)
6.3 (1933)
7.9 (1951)
                         24

-------
      1143
      1016
c
o
•r-
-P
O

d)


0.
               1900   1910
1920
I960    1970
                                       YEAR
              Figure 5,   Annual precipitation (A)  and  10-year
                         running mean (M), 1893-1970
                                    25

-------
                                 SECTION 10


                    EXPERIMENTAL RESULTS AND CONCLUSIONS
INTRODUCTION

    As indicated in Section 4, a relationship evidently exists between
pollutant concentrations in tree rings and the pollutants present in the
atmospheric environment of the tree.  Section 5 describes the uniqueness  of
the study area used in this demonstration project: the area surrounding the
Tacoma ASARCO Smelter.  The history of this smelter is given in Section 6.
While a detailed chronology of pollutant emissions could not be compiled  for
the smelter, several specific phases in its history were identified.  These
could possibly indicate the pollution present in the atmospheric environment
of the trees used in this study.  They were:

    -  1890 --  began operation as a smelter,
    -  1911 --  switched to copper smelting,
    -  1930's — depression years,
    -  1959-1960 — shut down during strike,
    -  1965 -- became the only commercial source of arsenic
               production in the United States,
    -  1967-1968 -- shut down during strike,
    -  1971-1974 -- definite records of arsenic production available.

These indicators were used to develop a tree-ring analysis scheme which could
demonstrate that the arsenic concentration in the rings was proportional  to
the ejection of arsenic into the environment.  By evaluating tree rings
before 1890, after  1890, around 1911, during the depression years, during the
strike years, and between 1971 and 1974, evidence could be gathered to
correlate with the  above indicators.

    In addition to  this, Section 4 also identified several factors that could
adversely affect the interpretation of the experimental results.  These were:

    -  diffusion between rings,
    -  soil leaching and uptake effects,
                                     26

-------
    -  direct uptake from the air by needles, and
    -  meteorological variation.

Other possible adverse effects could be that:

    -  different tree species exhibit differing effects,  and
    -  different trees of the same species have different
       uptake patterns.

The tree rings contained the needed information.  If diffusion  between  rings
was taking place, arsenic should be found in rings formed before 1890.   If
soil leaching and uptake were problems or if direct uptake from the  air was
occurring, one might note differences in arsenic concentrations in the  rings
as a function of height in the tree.  Finally, if species effects were
important, or if different trees of the same species presented  a problem, the
chosen tree mix (three Grand Firs, one Cedar, and one Douglas Fir) should
answer these questions. In order to accomplish this and check for method
validity, the six specific regions of interest based on the history  of  the
smelter were evaluated.
EXPERIMENTAL RESULTS

    A detailed evaluation of the increment cores indicated that  the  preferred
cores were from tree #2,  the Douglas Fir.   They were by far the  largest
individual rings, typically about double the weight of the Grand Fir and
Western Red Cedar.  Here, individual  rings could be analyzed at  the
University of Washington, whereas in all  other trees,  2 years of growth  per
sample were required to have sufficient arsenic activity for counting.
Unfortunately, the center of the tree was  decomposed and only samples dating
back to 1937 could be obtained.   The best  tree was  #5, a Grand Fir.   It  had
the most clearly defined  rings;  its A and  B cores and  its lower, middle  and
upper cores matched reasonably well (see Fig.  6).  Consequently, it  was
selected as first choice.  Tree  #1, a Grand Fir and the largest  tree in  the
group, had extremely narrow rings and had  little similarity in adjacent  rings
as a function of height in the tree.   Since arsenic sensitivity  was  a major
problem and ring counting was almost impossible in  the earlier years, most of
the tree#l samples were  saved as archives for future  evaluation.  One set of
rings was analyzed at Washington State University,with the higher-neutron-
flux reactor, evaluating  individual rings  between 1954 and 1977.  This
allowed a detailed evaluation of both of the strikes.   Similar ring-counting
problems were encountered with tree #4, the Western Red Cedar, and tree #3, a
smaller Grand Fir, especially in the earlier years.  Patterns of similarity
could be noted in certain years  however,  and usually the years following  1940
could be counted with confidence.
                                     27

-------
      1976  =	=   1976
      1972  ;=«---:=   1972

      1967  1—-•---••••=   1967


      i960  !-::::::!   196°
      1Q._  = .,c,r=   1955

      1952

            £  ^-;S   1941
      1941  ^:- ;;,-==   1937

      1937  i::;^;!E   1933
      1933  —;:--' =
             —  .--:=  1926
      1926   — ^'-' =
      1919   —-:::•=   1919
               -v^     1912
      1898  =---_-_-_-.=   1898


      1890  ^: --_-_-_-.=   1890

      1884
           LOWER   UPPER
Figure 6.   Comparison of upper and lower
          tree rings - tree #5
                 28

-------
    It is important to note here that the cores could not be mounted and
polished to aid ring counting in the traditional manner because this would
have introduced contamination in the rings.  In the future, duplicate cores
(approximately 1 cm apart) will  be taken and one sample will be used for ring
counting in the traditional manner and the other will be used for chemical
analysis.

    For this study the cores were placed on clean white paper and patterns or
ring maps (Fig. 6) were constructed for each core.  A sharp-pointed knife was
used to roll the cores back and forth as an aid in ring identification.   As
was noted above, great difficulty was encountered and undoubtedly led to
ring-counting errors, except for tree #5 which was far easier.

    The cores were placed on clean paper and covered with Handi-WrapR so
that they could be cut with the sharp-pointed knife without direct handling.
The Handi-WrapR also kept the small  individual  samples from moving once
they had been cut.  The cuts were made along the lines separating yearly
growth.  The individual samples were then placed on laboratory spot-test
trays that were properly labeled, transferred to the drying ovens, and
weighed to constant weight.  As indicated earlier, several  test cores were
employed in developing the technique.  After weighing, the  ring samples  were
transferred to clean polyethylene vials and irradiated in the reactor.   After
irradiation the samples were transferred to clean unirradiated vials and
counted.

    The experimental  results of the study are shown in Tables 7 through  10.
The concentrations were determined by comparing the sample  activities to the
activity of arsenic standards that were irradiated simultaneously.  The
standards used at the University of Washington (Table 7 data) were A.$203
standards prepared by the author.  The standards used at Washington State
University (Table 8 data) were National  Bureau of Standards Orchard Leaves
(SRM-1571) containing 14±2 yg/g arsenic.  Special  sample holders were used  to
ensure reproducible geometries during both irradiation and  counting at both
laboratories.  Traditional gamma-ray standards, e.g., 137cs, 60co,
65zn, 57co, and ^Ha, were used  to calibrate the gamma spectrometers
each day.  In order to ensure that each sample was exposed  to a similar
neutron flux, a cluster of 11 samples was the maximum possible at the
University of Washington.  At Washington State University,  three levels  were
used and iron-flux monitors were employed to account for flux differences.
This allowed 24 samples to be irradiated at once.   Since the University  of
Washington reactor building was open only 8 hours  per day and all  11 samples
had to be counted, counting times were limited to  a maximum of 40 minutes.
No time restraint was necessary at Washington State University and all 24
samples were counted consecutively for 1-hour per  sample.  The first
irradiations on tree #5 were of 1-hour duration and 1-day cooling to
  Registered trade name
                                     29

-------
                                  TABLE  7.

             ARSENIC  CONTENT  IN  STUDY TREES  (MICROGRAMS As/g Wood)
           Year
                          Samples
5-L
5-M   5-U    4-L   4-M
Year
Sample     Sample
 3-L   Year  2-1
Smelting
begins
Depres-
sion
years
Strike
Stri ke
1886-87
88-89
90-91
90-93
94-95
1910-11
12-13
14-15
16-17
1920-21
22-23
24-25
26-27
28-29
30-31
1932
1933
34-35
1957-58
59-60
61-62
63-64
1965
66-67
68-69
70-71
72-73
74-75
76-77
—
—
Trace
5.10
24.5
22.2
8.88
14.0
12.6
12.0
16.2
8.18
10.2
10.7
6.53
4.62
2.18
18.0
28.7
20.3
11.7
46.9
194
47.2
113
27.7
28.4
—
8.05
7.16
8.75
1.32
3.04
6.77
2.94
13.14
18.5
11.9
57.1
14.3
69.3
36.7
— - 20.9 3.30
7.0 4.82
8.84
6.04 11.6 3.66
4.52 14.8 5.05
6.14
2.87
0.3 5.51 3.27
0.8 6.83 2.57
8.12 0.59
3.13 5.78
8.58
9.67
5.74 22.3 5.13
6.07 43.8 6.44
6.96
15.1
12.5
1886-87
88-89
92-93
94-95
1914-15
16-17
1956-57
58-59
60-61
62-63
64-65
66-67
68-69
70-71
72-73
74-75
76-77
30.7
13.0
12.5
11.0
22.0
10.2
16.9
10.7
10.3
12.4
9.41
12.5
9.83
9.01
5.05
11.9
17.6
1958 9.32
1959 10.6
1960 5.71
1961 3.43
1968 3.73
1969 6.01
1972 4.19
1973 5.48
L, M, U = lower, middle, upper sections of tree.


to reduce 24Na interference); 20-, 30- and sometimes 40-minute counts were
made, depending on the arsenic activity.  The remainder of the University of
Washington irradiations were of 2-hour duration and 2-day cooling.   This
resulted in similar arsenic activity, but decreased the 24Na from about
1-1/2 lifetime decay to 3 lifetimes and hence increased the arsenic
sensitivity with respect to the sodium.  In all cases, sodium was the main
interference.  A similar irradiation-cooling cycle was used at Washington
                                     30

-------
State for tree #1 samples.  In all, 108 samples were analyzed  at the
University of Washington and 31 samples at Washington State.   Six of the 108
samples were irradiated twice at the University of Washington  to
experimentally check the difference between 1-hour and 2-hour  irradiations.
Seven of the 108 samples were irradiated at Washington State to
experimentally check the difference between the reactors.   These results are
shown in Table 9.
                    TABLE 8.  ARSENIC CONTENT  IN  TREE #1,
                             SAMPLE  1-L  (ug As/g wood)

Year
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
As
0.393
0.402
0.488
0.541
0.313
0.344
0.441
0.277
0.249
0.499
0.338
Year
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
As
0.432
0.319
0.267
0.335
0.382
0.327
0.430
0.355
0.391
0.571
0.668

                TABLE 9.   ARSENIC CONTENT  IN  TREES #1  and #5
                          DURING STRIKE  YEARS (yg As/g wood)

Year
1965
1966-67
1968-69
1970-71
1972-73
1974-75
1976-77
5-L W.S.U.
0.33
0.44
1.39
1.00
0.65
0.48
0.37
1-L W.S.U.
0.39
0.38
0.30 Strike
0.35
0.39
0.48
1.08
                                     31

-------
    Several additional irradiations were initially conducted on the extra
increment cores to determine if acceptable arsenic sensitivity could be
obtained from single rings.  These studies indicated that at the University
of Washington single rings could be evaluated only for Douglas Firs.  Single
rings could be evaluated for the other species only at Washington State.

    Six different needle samples, one small Grand Fir cone, three soil
samples and three standards, were also irradiated at the University of
Washington reactor.  These results are shown in Table 10.  Two-hour
irradiations and 2-day cooling periods were employed, but counting periods
varied, depending on the activity.  An evaluation of these data shows that
arsenic concentrations are far greater in  all of these than in the rings.
                TABLE  10.  ARSENIC CONTENT  IN NEEDLES, SOILS
                           AND CONES  (yg As/g wood)


Sample
1-N-l
l-N-2
l-N-3
l-N-4
Young Grand Fir
Young Grand Fir
1-N-D
5-S-l
5-S-2
5-S-4

Description
New needle
1-year old
2-year old
2-3 years old
New needle
Cone
Dead needle
Litter
Top 2.5 cm
7.5 cm deep
Weight
(ing)
4.6
5.1
12.2
15.2
8.2
105.9
3.5
19.0
18.1
18.3
Arsenic
content
39,957
7,588
23,820
27,368
10,634
46,260
92,857
27,437
21,833
19,781

     All  samples  were oven-dried to constant  weight.   Since meteorological
 conditions  were  important to  the transport and  rainout-fallout of the
 pollutants, a  comparison between the rainfall in  the  study area and the
 growth (weight)  of the rings  was used as a first  approximation to establish a
 simple growth-climatology correlation.   This in turn  was  used to give an
 estimate of the  expected atmospheric arsenic concentrations.  The ring
 weights and the  expected growth showed  a good correlation.
                                      32

-------
DISCUSSION

    An evaluation of the data in Table 7 indicates that in the case of tree
#5 the method appears to be valid.  For instance, no evidence of arsenic can
be found prior to 1890 in either upper, middle,  or lower section ring
samples.  Also, the arsenic concentration drops  off during the depression
years and there are corresponding decreases following the strike years (see
Fig. 7).  Of special  interest is the fact that in the upper and middle
samples, the arsenic concentration changes show  a direct relation to stack
changes in the same year (see strike year data in Table 7), while in the
lower samples the tree change apparently lags behind the stack change by at
least 1 year.  This is seen to some extent in the Table 7 data for the other
trees as well.  The single-ring data from tree #1 given in Table 8 also show
a drop in arsenic concentration for the 2 years  following the 1959-1960
strike, but the decrease in arsenic for the 1967-1968 strike actually occurs
during the same years (see Fig. 8).  This will be discussed later.

    On the other hand, an evaluation of the data from trees #3 and #4 shows
that arsenic can be found in the rings before 1890.   In both cases the center
of the tree was penetrated during sample extractions.  The importance of this
is uncertain.  In addition, the fact that the cores  of these trees are so
close to 1890 in age may also affect the results.  The most probable answer
is ring-counting errors.  As stated earlier, a definite pattern correlation
could be seen in the tree #5 samples and counting was easily accomplished.
However, these patterns were difficult to establish  and compare in the
majority of the other trees; so counting back to 1890 could be in great
error.

    The decrease in arsenic concentration in tree #5 between 1910 and 1915
suggests the smelter change could have had an effect on the atmospheric
concentration of arsenic.  This can be interpreted to mean that either less
arsenic was present in the atmosphere, or the arsenic was gradually
eliminated from the soil, indicating the soil could  have been saturated.
However, the data of both tree #3 and tree #4, if they are accurate, show
similar concentrations in the 1890's which leads one to believe saturation
effects are small.  On the other hand, the arsenic concentration in tree #5
is greater in 1910-1915 than in 1890-1895 which  indicates a buildup.  Since
no arsenic production records are available from ASARCO and since no records
are available that would indicate how the smelter output should have varied
as it changed from lead to copper smelting, it is unclear what should have
occurred.  It is known that lead arsenate is a common ore of lead, but it is
also known that many copper ores are also high in arsenic.  One thing is
clear: the expected growth during that time period is high for all  of the
years.  Thus, decreasing arsenic in tree #5 between  1910-1915 is probably not
a function of adverse growth, or meteorology.
                                     33

-------
    -o
    o
    o


    en
   30
        20
GO
LU

O


O
o
    a  10
    CO
O  LOWER SAMPLES


V  MIDDLE SAMPLES


D  UPPER SAMPLES
         1885   |    1895


         SMELTING BEGINS
                                       O
                                              0--O
                                                                           a
                                                                        \


                                                                     /  \
                                                                     /    \
                                                                     I    O
                            1910
                           1920
1930
                                                                 i
                                                                 V
                              57.9


                               n
                               11
                           (   I
                           \  I


                            \!
 I   b
 I
I
                                                                                     b
              4-1955  \   1965 /


DEPRESSION YEARS     STRIKE YEARS
 1975
                        Figure 7.  Yearly  variation of arsenic concentration - tree #5

-------
       1.50
       1.25
  o
  o
  CO
  «=c
  cr.
  =3
       1.00
       0.75
  o
  o
  o
  oo
      0.50
      0.25

                      OA

                                         o  /
                                                         o
                                                                '  \  I
             b'
                                                                                     o
                            /
                            /
                           o
50 1955
1960
J
1965
1
1970
1975
                                  STRIKE 1
STRIKE 2
          Figure  8,  Yearly variation of arsenic  concentration  - tree  #1
*U. S. GOVERNMENT PRINTING OFFICE: 1979-684-481
                                             35

-------
    The strike data (1959-1960 and 1967-1968) for trees #2,  #3,  #4,  and  #5
all  indicate the method is apparently valid.   Consistency is noted
throughout.  Again the lower samples in trees #4 and #5 appear to lag  behind
the upper samples by 1 to 2 years.  The more  easily interpreted  yearly ring
data in tree #2 confirms that it is at least  1 year.  The more complete  data
for tree #1 (Fig. 8) show low arsenic concentration for 2 years  following the
early strike, but also show low concentrations during the strike period  for
the later strike.  The explanation for this is not clear.

    The apparent exceedingly high values of arsenic found in tree #5 in  the
periods of 1968-69 and 1972-73 (Table 7) cannot be accounted for in  counting
statistics.  These samples were reirradiated  in the Washington State
University reactor to resolve this anomally and to compare with tree #1, and
the comparative results are shown in Table 9.  The WSU data appear  more
normal for tree #5 and suggest  the U of W data may be incorrect.

    The comparative 2-year data for tree #1 in Table 9 were obtained  by adding
the single-year data reported in Table 8.  An evaluation of-the data in  Table
9 shows that little correlation exists between trees #1 and #5.   The large
increase  in arsenic concentration in tree #1  (Tables 8 and 9) and also in
tree #3  (Table 7) can be explained as an excess of unbound arsenic that  was
present in the sap and may have been deposited during the extensive
oven-drying process.  One would expect the outer sap wood to carry the
majority  of the nutrients (and foreign material).  The lack of correlation
between the U of W and WSU data for tree #5 cannot be explained.

     It can be seen in Figure 7 that during the depression years (1932-1935)
the arsenic levels were exceedingly low.  A cutback in production in the
early stages of the depression would explain the data if one assumes the
2-year lag.  Since 1930-1932 were exceedingly good growing years in the  Puget
Sound area  (sample weights were above normal  which confirms this), climate
could not account for any variations this large.

    The  scatter  in the data is too high to enable comparisons between  the
measured  values  of arsenic in the rings and direct production data available
from ASARCO in 1971 and 1974.  This variability is one of the limiting
factors  in the analysis.  In general, a plus or minus value of 10 to 15
percent  can be placed on most of the data presented in Table 7.  Typical
40-minute counts yielded an integrated photopeak area of 100 to 500 counts,
which  statistially  indicates about a 10-percent accuracy.  While the
integrated  photopeaks for tree #1 in the WSU data were typically 10 times  as
large, the  excess arsenic found in these years made it impossible to compare
with  the  ASARCO  data.

     Many limiting factors have been identified that need to be resolved  prior
to  validating this method. The limited number of trees evaluated is far  from
the number  that  should be checked.  At least  10 to 20 cores should be
                                      36

-------
evaluated in order to increase accuracy,  and dendrochronologists generally
select no less than 20 cores in a particular area.   On  the other hand,  other
equally important, unscientific factors were identified that  made the method
validation equally difficult.   Since sensitivity is a  problem,  the extraction
of larger bore samples would aid in decreasing  this limitation.
Unfortunately, larger samples  are harder  if not impossible to obtain.   Since
most trees that could be used  in analyses of this type  lie on private
property, permission must be obtained in  order  to sample them.   The author
found this permission hard to  obtain.  After considerable time,  some
individuals could be interested in becoming part of the study—as long  as
small bore samples were to be  taken.  No  one wants  1.5  to 2-cm  cores taken
from his trees, especially when it is known that the larger holes need  to  be
plugged to prevent serious infection.


CONCLUSIONS

    Despite the noted problems, the available evidence  favors this method.
Although the results indicate  that arsenic sensitivity  is far less than that
required to develop a quantitative theory, the  experiments did  show that
changes in smelter production  can be related to changes in tree  pollutant
concentrations.  Further phases of the project  should  eliminate  many problems
and lead to an analytical validation.
                                     37

-------
                                 REFERENCES
1.  ASARCO, Inc. Annual Report, 1976

2.  Ault, W.U., R.G. Senechal , and W.E. Erlebach.  Environmental
    Science and Technology, Vol. 4, 1970. p. 305.

3.  Cramer, H.E. , J.F. Bowers, and H.V. Geary.  Assessment of the
    Air Quality Impact of $$2 Emissions from the ASARCO Tacoma Smelter,
    EPA 910/9-76-028,  Environmental Protection Agency.  July, 1976.

4.  Crecelius, E.A., M.H. Bothner, and R. Carpenter.  Environmental
    Science and Technology, Vol. 2, 1975. p. 325.

5.  Heilman, P.E. and  G.T. Ekuan.  Heavy Metals in Gardens Near the
    ASARCO Smelter, Tacoma, Washington.  EPA-68-01-2989, Environmental
    Protection Agency, April, 1977.

6.  Johnson, D.J. and  L. Lippman.  Environmental Contamination With Lead
    and Arsenic From a Copper Smelter.  Paper 73-AP-37 presented at Pacific
    Northwest  International Section, Air Pollution Control Association
    meeting, Seattle,  Washington,  November 29, 1973.

7.  Lederer, C.M.,  J.M. Hollander, and I. Perlman.  Table of Isotopes, John
    Wiley and  Sons, Inc., 6th Ed., 1967.

8.  Lepp, N.W. Environmental Pollution. Vol. 2, 1975. p. 49.

9.  Phillips,  E.B.  Climatology Summary:  Tacoma, Washington, Office
    of the State Climatologist, U.S. Weather Bureau, Seattle, Washington,
    1960.

10. Nelson, P.A. and J.W. Roberts.  A Comparison of the Efficiency of the
    #1 ESP and the  Pilot Baghouse  in Controlling Particulate Emissions at the
    ASARCO Tacoma Smelter.   Paper  presented at Pacific Northwest Internatinal
    Section, Air Pollution Control Association meeting, Vancouver, B.C.,
    November 19, 1975.

11. Pillay, K.K.S.  Activation Analysis and Dendrochronology for Estimating
    Pollution  Histories, Transactions of American Nuclear Society, Vol. 21,
    Supplement 3, 1975.  p.  22.
                                      38

-------
                           REFERENCES (Continued)
12. Sheppard, J.C. and W.H. Funk.  Environmental Science and Technology,
    Vol. 2, 1975. p. 638.

13. Szopa, P.S., E.A. McGuiness and J.D. Pierce.  Wood Science and
    Technology. Vol. 6, 1973.   p. 72.

14. United States Bureau of Mines - Minerals Yearbook, Vol.  I,
    U.S. Department of Interior, 1961.

15. United States Bureau of Mines - Minerals Yearbook, Vol.  I,
    U.S. Department of Interior, 1962.

16. United States Bureau of Mines - Minerals Yearbook, Vol.  I,
    U.S. Department of Interior, 1963.

17. Ward, N.I., R.R. Brooks and R.D. Reeves.  Environmental  Pollution,
    Vol. 6, 1974. p. 149.

18. Washington State Air Monitoring Data, annual publication of Department
    of Ecology, Air Programs Division,  Olympia, Washington 98504, 1976.
                                     39

-------
                                   TECHNICAL REPORT DATA
                           (Please read Instructions on the reverse before completing)
 REPORT NO.
   EPA-600/3-79-030
2.
                              3. RECIPIENT'S ACCESSION NO.
 flTLE AND SUBTITLE

   DEVELOPMENT  OF  A STRATEGY FOR SAMPLING TREE RINGS
                              5. REPORT DATE
                                  .  March  1979
                                                           6. PERFORMING ORGANIZATION CODE
 AUTHOR(S)
   Jerry A. Riehl
                              8. PERFORMING ORGANIZATION REPORT NO.
 PERFORMING ORGANIZATION NAME AND ADDRESS
   Northwest  Environmental Technology Laboratories,  Inc.
   5709  - 80th S.E,
   Mercer Island, Washington   98040
                              10. PROGRAM ELEMENT NO.

                                    1AA601
                              11. CONTRACT/GRANT NO.

                                     CB-7-0771-A
 2. SPONSORING AGENCY NAME AND ADDRESS
   U.S. Environmental Protection  Agency-Las Vegas, NV
   Office  of Research and Development
   Environmental Monitoring and Support Laboratory
   Las Vegas, NV  89114
                              13. TYPE OF REPORT AND PERIOD COVERED
                              14. SPONSORING AGENCY CODE
                                     EPA/600/07
 5. SUPPLEMENTARY NOTES
   Dr.  Gilbert  D. Potter, Project Officer
16. ABSTRACT
   A method for determining  retrospective pollution  levels  has been investigated.
   This  method relates arsenic  concentration in tree rings  to arsenic-in-air con-
   centrations based qualitatively on arsenic emissions  from a nearby smelter,
   corrected for climatological and meteorological effects.  To evaluate the validity
   of  the method, a unique pollution study area was  identified and characterized in
   detail.   Several select trees were sampled and the arsenic concentration determined
   by  neutron activation  analysis.  These concentrations were compared to certain
   known phases in the production history of the smelter, coupled with the expected
   climatology and meteorology  of the area.  Positive correlations were found thus
   satisfying the goals of the  preliminary project.   Major problems encountered were
   low arsenic concentrations and an inadequate number of samples.  Recommendations
   for future studies are given.
17.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
                                              b.lDENTIFIERS/OPEN ENDED TERMS
                                            c.  COS AT I F;ield/Group
       Dendro chronolo gy
       Arsenic
       Air Pollution
                  Retrospective Monitoring
     06C
     07B,E
     18B
18. DISTRIBUTION STATEMENT

       RELEASE TO  PUBLIC
                 19. SECURITY CLASS (This Report/
                      Unclassified
21. NO. OF PAGES
    48
                                               20. SECURITY CLASS (TMipage)
                                                   Unclassified
                                            22. PRICE
                                                A03
EPA Form 2220-1 (Rev. 4-77)   PREVIOUS EDITION is OBSOLETE

-------