54077502
Mathematical
Evaluation
of
Kepone Action Levels
January 10, 1977
f
Prepared by
U. S. ENVIRONMENTAL PROTECTiON AGENCY
Office of Pesticide Programs
Office of Special Pesticide Reviews
Washington, O.C. 20460
r.? 5^/
LIBRARY
.; S- ENVIRONMENTAL PROTECTION
«, i.
-------
-------
Mathematical Evaluation
of
Kepone Action Levels
January 10, 1977
By: Todd W. Thorslund, Sc.D.
Gerald K. O'Mara, M.S.
\
*
Prepared by
U.S. ENVIRONMENTAL PROTECTION AGENCY
OfficejB^Pestjcide Programs
Office of Special Pesticide Reviews
Washington, D.C. 20460
LIBRARY
5. EuVlfiONMENTAL PROTECT10-N
M. 1. 0^817
-------
-------
CONTENTS
Lists of Figures ii
List of Tables iii
Bac kg round 1
Introduction 3
Dose Response Relationship. 4
Action Levels Obtained by Considering
Risk to Individuals 7
Action Levels Obtained by Considering
Risk to Total Population 11
Estimated Risk to Population in Area
Contiguous to Chesapenke Bay Under
Current Action Lev els 16
Effect of Kepone Resicues in Blue fish 17
Append ices
Appendix A. Production and Consumption
of Chesapeake Bay Seafood 21
Appendix B. Assumptions Used to Compute
Implied Risk to a Given Population 40
Appendix C. Tumor Risk to U.S. Populat on
Assuming Log-Probit Model and Different
Exposure Patterns 46
-------
-------
FIGURES
NO. ' I'aR e
! Functional Relationship Between
Kepone in Diet and Incidence of
Hepatocellular Carcinnm, 19
2 Relationship Between Kej one in
Diet and Incidenc< of Hepatocellular
Carcinoma in the !ange of Probable
Human Exposure 20
3 Rela t ionsli ip Between Implied
Calculated Risk Rate and Action
Level for Three Dietary Intake
Levels 21
ii
-------
-------
LI T OF TABLES
No. Page
1 Possible Levels of Seafood Consumption
to he Utilized In Dotermining Action Levels 10
2 Potential Action Lex els for Different Lev els
of Risk and Seafood Consumption 10
3 Implicit Calculated Number of Tumors Under
Present Actions Levels 12
4 Resulting Action Levels Under Different
Risk/Benefit Assumptions 14
5 Implied Risk to Individual in Chesapeake
Bay Area ...16
6 Lifetime Production of Tumors Due to
Consumption of BluefLsh 18
111
-------
-------
RACKGRO 'NO
In the spring of 1976, the Office of Pesticide Pro-
grams, EPA evaluated Kepone residues for Chesapeake Ray
Seafood. As a result, action levels were proposed by EPA
for fin fish, shellfish and blue crabs and late implemented
by FDA. Current action levels for Kepone 'neasured in parts
per million (ppm) are: 0.1 ppm in fin fisl; 0.3 ppm in
shellfish; and 0.4 ppm in bluecrabs.
At tie time the action levels were put into effect,
little was known concerning the extent an>l duration of
the Kepone problem for the Chesapeake B a ^ . EPA's position
was that as additional inforiation became available on the
subject a reexamination of r sks and benefits related to
alternative action levels wo "Id be undertaken.
NCI has recently rade available an updated report
concerning chronic can er bioassay studies of rats and mice
fed kepon>'. In additi >n , data has recently been collected
concerning the amount of seafood consumed that is taken from
Chesapeake Bay. This 'dditional informaticn allows a
reassessment of the ac ion levels for kepone at this time.
This paper is intended to outline a genera! methodology
for assessing the risk of a given action level and for deter-
mining action levels based on implied risk concepts. It
uses kepone as a case example. The method developed allows
-------
one to either set an action Jevel by controlling what amounts
to a conservative estimate of the total ri k or a fixed
ben >fit--risk ratio. This upper bound for risk takes into
account the total human exposure and carcinogenic response
levels of animals. The method is also flexible enough to
directly account for benefits of the adulterated product for
the same action level determination.
Several potential problems can be anticipated in using
the rethod: 1) The ~e is a danger that rather tli in interpreting
the upper bound of risk as a carcinogenic index, a literal
interpretation of risk as an actual estimate might be made;
2) The absolute value of the index is hard to put into
context since at present it has not been used in other
situations.
It is also recognized that the method suggested has
numerous scientific shortcomings. Its utility only becomes
apparent when contrasted to the traditional method of
obtaining an action level which is not appropriate for
effects presumed to have no biological threshold.
-------
IN "RODUCTION
In order to assess the carcinogenic risk to man of
low levels of exposure to a < hem ! al , it is necessary
to extrapolate data from animal ecperiments conducted at
high levels of exposure. Virtually all competent scientists
however recognize that this type cf extrapolation is at best
highly speculative and based on many unverifiable assumptions.
Thujs^j the estimate of thr resulting number of tumors produced
i n man by th e c heroic a 1 _ s h oulid^ not be looked a t as a t r u e ,
a_b^sj)ljj_ti; number. Mor_e aj propr lately , it^ should be viewed as^
a health hazardindex thct incorporates bot_h the degree of
care i^nogejiic ac ^iv_i_ty_ o_f_ a chemical and the exposure lev el s
to which man is s_uspec_t_eci to be subjected. Ii light of
this, the absolute value of the estimate is not in itself of
primary importance. Rather, the factor to focus on is the
change of the index due to differences in regulatory stra-
tegies and to a compariscn of the magnitude o' risk index of
various chemicals.
To obtain a value for the health hazard it is necessary
to define the functimal form of the mathematical relation-
ship between consump ion level and tumor incidence. In
addition, methods that should he uniformally followed are
needed to chose the data base to be utilized in fitting the
selected dose response curve and in estimating human exposure.
-------
i)o_SK Ri-spnr si- RKLATTONJHJP
The functional form chosen for predicting effects of
ingested Kepone was the "one-hit" model wh'ch has many
theoretical and practical advantages. Furthermore, since a
risk estimate is highly dependent upon the model chosen,
using the "one-hit" mo lei in this case also has the added
advantage of being internally consistent since it was
previously employed hy the Carcinogen Assessment Group (GAG)
to predict the ri.ck to humans from the continued in term use
of Kepone ant-bait, Firther, for reasons which are amplified
in Appendix C, the "ona.-hit" model is deemed most appropriate
for this purpose. How'ver it is recognized that other models
could be utilized. In fact, the Administrator's Interim
Procedures and Guideliies suggest that several risk extrapola-
tion models should be utilized when assessing a carcinogen
[F.R.,Vol 41 : 21402-21405(1976)\ . To ob.ain some insight on
how the risk estimites would differ under a different model
and exposure assum ) tions , the log-probit model was also
employed. The results .re shown in Appendix C.
At present, there , s no clearly defimd rationale
within the Agency for selecting or eliminating data in
multiple concentrat ion-sex-species-tumor type experiments.
The primary cancer dat; utilized in this p, per is confined
to hepatocellular carc'noma in male mice cited ;n NCI's
January 1976 progress report, which is the iame data set
selected by CAG in its kepone ant-bait ana'ysls. CAG chose
this data set because it showed the most pronounced Kepone-
related effect. In a slight departure fror CAG, both
4
-------
concentration levels of Kepone were utilized in the slope
estimate. This gave a more efficient estimate than CAG had
obtained using onl the high concentration.
The functional form of the probability of an animal,
fed x ppm of a chemical in its diet, developing a tumor over
its lifetime using the "one-hit" theory may be expressed
a s
-6x
P(x)=9+(l-0)(l-e ) (1)
where 6 is the probability of developing the tumor due to
causes other than the deliberately-added chemical in its
diet and Bis the unknown "slope" of the "one-hit" model to
be estimated from observed data.
Fitting this model to the data by non-linear weighted
least squares, the result shown in Figure (p.19) is
obtained, where 6= .1634 and $ = .0791. The probabi-
lity of developing a tumor due to Kepone at x ppm in the
diet is then expressed as
* -.0791x
P(x) = l-e (2)
and is also shown in Figure 1. In Figure 2 (p.20) the
relationship at the levels of expected hum,- n exposure is
also illustrated, whict clearly shows that the response at
-------
this level is virtually linear so that the dose response
relationship may be expressed with virtuall no loss of
accuracy as
*
P(x) = .079lx (3)
To use this equation for man the assumption is made that
ppm in the diet is an equivalent dose for all organisms
Thus, the value for X is the Average lifetime pi>m in the
*
diet and P(x) is the probability of developing a cancer
due to Kepone during th i entire life span of the individual.
This relations? ip, even assuming equivalence between life
span exposure of mice and man, w < u 1 d only be strictly valid
for a cohort of men exposed at firth. However, as a limiting
case an additional conservative assumption could be made
that exposure has the same effect over any segment of the
life span. Coupled with the linear dose retpon.se relationship,
this means that we need only consider man years of exposure.
Thus, the probability of getting a tumor in seventy people
exposed for one year is the same as the probability of one
person getting a tumor if exposed for seventy years.
-------
ACTION LEVEL? OBTAINED BY CONSIDERING
RISK TO INDIVIDUALS
An implicit assumption in the establishment of an
action level is tlat t' > e risk for any leve below the action
level is tolerable. This is tru; since tie foodstuff is
allowed to be used as long as its chetiica] content does not
exceed the action level. Since any chemical level has some
potential risk associated with it, tie max imum allowab1e
risk is established by the choice of the action level
Conversely, given a saiety factor to account for inter-
species sensitivity differences and the calculated risk, it
is possible to establi'h a c or r es pond i np, action level. It
should be emphasized that such a determination depends only
upon the esta lished dose response relationship and hypothe-
sized food in ake. It should also be emphasized that the
actual risk is not necessarily equal to the risk associated
with the action level and, in fact, will usually be less
than the maximum due to the fact that ac tual chemical
residues are less than action levels.
EPA has recently developed information to be used
in safety determinations for calculating food intake of
various food stuffs. This information was culled from
several USDA studies and indicates that the average weight
of solid food eaten per day per individual in the U.S.
is 1,930 g. If aj is defined as the action level and
wj the weight in grams eaten per day per Individual of
-------
the jth types of food stuff thei the statistically calculated
risk of developing hepatoe el 1 ular carcinoma in the lifetime
of an individual who eats wj grams of the jth food source
daily is
-5
= ±07' Ixaj wj = 4.098x1 f) x aj wj (4)
i 930
where a j v j is the maximum number of ppm in the diet from
1930
the jth foodstuff.
Although in specialized circumstances one might be
interested in the risk due to a single item, it is more
likely that the major concern would be total exposure
through the entire diet. In this case, the total maximum
calculated risk to exposure of k different food stuffs would
be expressed as
k
P = 4. (98x10 x ^.^ aj w] (5)
From a strictly t ox io 1 og ic al viewpoint the action
level for each foodstuff should be the same since a standard
unit of one foodstuff las the same effect as a standard
unit of another. However, if the nutritional, economic,
and/or recreational value of one foodstuff is greater than
another foodstuff, then it is desirable to directly incorporate
this information in establishing that action level. For each
potential action level an implict risk-benefit ratio exists.
8
-------
To maintain the same r j s k-l> ene f L t ratio for all foodstuffs,
thus insuring an equitable treatment of all sectors of the
fishing industry, the ;ctLon levels should he made proport-
tional to the value per unit volume vj so that
k
V
P = 4.0 ''8x10 xa ^<\ v j wj ( h)
.1=1
where aj = av j . This implies that if a foodstuff has value s
times as great per unit volume as another, a risk s times as
great will be acce 'ted, and fie same risk-benefit ratio will
be kept, where it is assumed that the calculated risk or
risk-benefit ratio has been fixed.
To be specific, consider the case where it is assumed
that all foodstuffs have equal value to the individual.
In this case aj = i, j=l,2,...,k so that
k
-5
P=4.048xlOxa w] (7)
In the c ise of Kepone-"ontaminated seafood, the term
k
\x wj=W is simply the total daily intake of all forms of
3=1
seai'ood from the potentially contaminated ire a in grams.
-------
In j-, eneral, given P,W, and a safety factor ijj , the action
level can be obtained from the relationship
5
a= i^PxlO / 4 . 098W
(8)
The relationship between P and a under three different
assumptions of total seafood intake, W , listed in Table 1, is
shown in Figure 3 (p.22) and Table 2, for the case of no
safety factor, which means that ty =1.
Table 1. Posible Levels of Seafood Consurption to be
Utilized in DC term nine Action Levels
W
Grams/Da ' Ra t i r n a 1e
29 Average U.S. seafood intake, lower bound
490 Average intake from animal sources, practi-
cal upper bound
1,930 Averap, e total food intake, a'so lite upper
bound
Table 2 . Potential Action Levels for Different Levels of
Risk and Seafood Consumption
W
G/Day
29 490 1930
_p
-2
1x10
-3
1x10
-4
1x10
-5
1x10
Act
8.415
.842
.084
.008
ion Levels
.498 .126
.050 .013
.005 .001
I f)
-------
ACTION LEVELS OBTAIN ED _BY CONSIDERING
RISK TO TOTAL POPULATION
If N(x) Is defined as th' total number of individuals
*
exposed at le el x , and P (x), the probability of an induced
tumor at leve x i; defined as in equation (3), then the
total e>pecte number if tumors in the exposed population
producet is
T = ^^ x P (x) N ( x ) = .0791 / ^ x N ( x ) ( 8 )
allx allx
The term / ^ xN(x) is simply the total exposure to the entire
all x
exposed human population and remains the same no matter how the
Kepone is distributed vithin the population since the amount
of f Lsh eaten is fixed. Therefore, under a "one-hit"
hypothesis, exposure patterns do not affect the total tumors
produced within a population.
If a population is exposed to k different foodstuffs,
each with a total weight consumed per day of Wj grams, with
an actioi level of a j associated with it, then the implied
total calculated number of tumors that could be expected to
occur is
T = .0791 ^ aj Wj (9)
1930 j=l
1 1
-------
The total weigits eaten per day of the items that have
set prescribed actim levels are shown in Tnble 3. Fish
meal is also considered due to the possibil ty that it
will enter int> the human population via th food chain from
chickens. The data and the basis for exposure estimates
from consuming Chesapeake Ray seafood is given in Appendix
A. To estimate the imp ted c ilculated number of tumors that
could occur for the present action level, Aquation (9) is
applied to the derived weight; and action levels. The
results are sh>wn ii Table 3.
Table 3. implicit Calculated Number of Tumors
Under Present Action Levels
Item
Fin Fish
Crabs
Shellfish
Fish Meal
Total
Weight e ten by
population per day in kg
3.64H
.72
2.32"
1.281
7.985
4
x 10
4
x 10
4
x 10
4
x 10
4
x 10
Present
action lev el
.1
.4
.3
.1
Implied calculated
of tumors produced
per lifetime
149.6
119.2
286.4
52.5
607.7
number
per year
2.137
1.703
4.091
.750
8.680
12
-------
Thus, if all of the commodities taken from the Chesapeake
Bay had residues equalling the current action levels, the
one-hit model Implies that approximately 8.7 hepatocellular
carcinomas per year would be the calculated risk.
In d< ing a risk-benefit analysis, it is considered
preferable to rela:e the risks and benefits as close as
possible to the saie population and if possible to cast both
costs and benefits in terms o health units. Taking this
approach, one measurement would be to relate the number of
people in the U.S. that would be supplied with their ace urns-
tomed intake of food fiom animal sources for a year to each
hepa toe e 11 ul ar carcinor a prod iced.
To obtain the total number of people whose animal food
needs are supported by Chesapeake Bay seafood, the edible
weight in kg/day obtained from fin fish and shellfish is
summed up and added to this is the total weight of edible
chicken obtained from the use of fish meal. This number is
then divii'ed by .41 kgm, the daily animal food intake of an
average Anerican. "he conversion ratio from fish meal to
edible chicken is .385 x .4 = .1>4, where .385 is the
fraction of total chicken weight that is edible and .4 is
the energy conversion ratio from feed intake to weight of
chickens. This number is then divided by the total tumors
produced per year under the assumed action level to obtain
the required benefit/risk index. At present action levels
this amounts to 16,226 Americans supplied their animal food
needs for a year for every hepa toe el lul ar carcinoma produced,
I 3
-------
To set action levels on the basis of a' sumod risks it
is 'lecessary t i specify the total calculate) number of
tumors and the weight of each foodstuff and the relative
value per unit wel| ht of each foodstuff. In Table 4 the
action le\els for everal specified condiions are shown.
Table 4. Resulting Action evels In PPM Under Different Risk/
Benefit Assumptions
Allocation of value
in same mannei as
relative price at
docks j de
Food
Stuff
Fin fish
Crabs
Shellfish
Fish Meal
Price/lb
.54
1.80
.81
.12
Relativ e
v alue
4.50
15.00
6.75
1.00
I/year {
.017
.058
.026
.004
- Numbers
J.68()/year
.151
.502
.226
.033
Same Vn ! no
Weight
of Expected
I/year 8.
.021
.021
.021
.021
' Un i t
Tumors -
)80/year
186
186
186
186
I'resen t
8.680/year
.1
.4
.3
.1
For example, to calculate the action levels for the
case where the rela ive values are equLvnleit to their
dockside price and >nly one tumor per year is allowed,
14
-------
70 tumors per life tirne are equated to the right hand
side of equation (10) ^
T= .Q791a >v vjWj (10)
1.93 j-1
giving the result: 70= .0791a [4.5 x 3.648+15.x.727+6.75x2.329
4
+1.281]xlO 71.93, solving for a, a= .0039-.004 which sets
the action level for fish meal. The other action levels are
obtained by multiplying a by the relative value of the
foodstuff. The levels are shown in thi; third column of
Table 4.
15
-------
ESTIMATED RISK TO POPULATION IN AREA CONTIGUOUS TO
CHESAPEAKE BAY UNDER CURRENT ACTION LEVELS
If the Implied average risk rate to any segment of
society is required it can be computed as
P=
ajfjWj
N
.0791
where Y= 365 x 2.20+ x 1.930; fj is the fraction of
the jth foodstuff that is consumed, by the specified segment
of society; Wj is the yearly total Intake in pounds; and N
is the population size for the considered segment.
Under the assumptions listed in Appendix B, the
following table was created which shows the average implied
risk to an individual in the counties surrounding Chesapeake
Bay in Virginia and Maryland considered separately.
Table 5. Implied Risk to Individuals in Chesapeake Bay Area
Location
Maryland (Bay)
Virginia (Bay)
Population Size
1970 Census
2,052,178
1,004,580
Fraction
consumed
Fin fish
.57
.57
of total
in location
Shellfish
oysters clams
.72 .40
.72 .40
Ci abs Calculated
249.01 =
.53 2,052,178
.53 92.94 = 9
1,004,580
Risk
1.21x1
.25x10
16
-------
EFFECT OF KEFOME RESIDUES IN BLUEFISH
In order to estimate tie actual harm from Kepone in
foodstuffs it would be necessary to devise a comprehensive
sampling plan for both commerical and sport fisheries. At
present, the monitoring efforts undertaken by the states of
Virginia and Maryland an! PDA have been concirned more with
evaluating whether
-------
If it is assume! that the two cited monitoring
A
studies give a sample that is at least roughly repre-
sentative of the population and that present conditions are
stable and will continue into the future, then a crude
estimate of the effects of citing bluefish can he obtained.
In Table 6 the number of t u IT > r s calculated from the weight
data found in Appendix B, using equation (7) ar<- shown
where the observed average residues levels o> .179 for
Virginia and .059 for Maryland are substituted for the
action lev-'ls. The numbers in parentheses ar the weights in
thousands of kg for each catagory ii edible 'eg/day.
Table ii . Lifetime Production of Calculated Tumors due to
Co Tsumption of Bluefish
Met
Spo
h
r
od of Catch
t
Mar
20
(H
Commer ical
Tot
a
I
(
21
yl
. 5
.4
. 7
. 2
. 2
Loc a t
and
4
9)
0
9)
4
ion
Virgi
22.
(3.
6.
( .
28.
n
0
0
8
9
8
ia
'
))
0
3)
1
Tot
42.
7.
50.
a
5
5
0
1
5
0
5
Due to catch volumes, th1 major part of the total risk comes
from the sport fisheries, and even though levels are much
higher in Virginia the total risk is roughly comparable for
both states. The total effect of bluefish is estimated at
the production of 5('.05 calculated tumors pe lifetime or .72 per
year.
*Data sent to Jack Blanchard by Max Kisenberg, Maryland
Department of Health and Mental Hygiene, August 6, 1976.
Appendix dated June 30, 1976 from VJ. J. llargis, Jr., Direc-
tor, Virginia Institute of Marine Science1 Subject: Mounting
Kepone Levels in Chesapeake Bay Fish.
18
-------
PROBABILITY OF DEVELOPING HEPATOCELLULAR CARCINOMA
DURING LIFETIME AT SPECIFIED DIET LEVEL
TJ
TJ
2
m
o
O
Z
m
z
o
m
O
C
3J
m
m C
^ Z
> O
O
O
2 CD
> m
c/> H
m
m
Z
*
m
-o
O
m
z
o
z
o
o
m
Z
O
m
O
n
-------
o
ci
30 m
a 8 § § § 8
O -» NJ W *» Ol
CD
00
O
O O
to -»
8lv
2 A
s
"8
5g
O
U1
m
TJ
O
2 I
I 3
8
o
CD
\
s
\
(D>
CO
\
\
\
m |-
m C u
g=S
23
230 m
^'O i""
m
-a
O
m
O
O
n
o
30
O
m
>
Z
m
X
s
V)
30
m
30
m
I O
T) Z
H|
O ^
if) SCO
IpiS
1cm
';r- m
,> ;z
o J m
;30|O
:2|S
O Z
>1
m!D
%l*
|!§
m rn
O
m
O
n
C
30
m
20
-------
N)
O
O
z
<
m
O
3J
m
O
CO
I
CD
m
O
3}
H
X
m
m
2
o
r-
m
3> C
3J i-« 3)
-< £= m
ir
m 3J
i- co
m ^
m ^
eo D
O
O
m
m
-------
-------
AI PKNDIX A
-------
-------
APPENDIX A
LIST OF TABLhS
Pae e
1 Total C o 'ii m c ; c Lai C a t c h : V < 1 u e a n d
Pounds in L iv e Weight, State of
Maryland, 1^74 and ! 9 7 r> , 26
' Total f.ommereia Catch in Value and Pounds
i n L Lv e We i p, h t , S t a t e o f V L r t', I n ! a , 1 9 7 4
and 1975 , 27
3 Total Catch and Fstimated C o n R u Total Catch and F, s t mated Consumption of
Selected Fish from Chesapeake Bay an<' its
Tributaries for Maryland...... 31
6 Estimates o' Edible Pound s a nd V a 1u e s of
Fin Fish and Shellfish f r o;> Commercial
Fishing in the Ches, p e a k e "> a y and its
Tributaries during 1 9 7 S , 'M
7 Estimate of Edible I't) u n d s ind Value of
Fin F i s <\ and S h e I 1 f : s h f r c n S p n r I !' i s h i n j;
in the C h e s a p e a K e. B a y a n d its 'I r i h u t a r i e s
during 1974 , 35
H Total ' o u n d s of M a n h a d d r-1 n and A ] o w i v e s
Caught in the C h e s a p e a k e R a v by R o >; i on
and Estimated Pound-; of Fishing, 1 9 7 S 37
9 Av e r a g e Prices for Chesapeake Seat > o d
Species, 1975 59
-------
APPENDIX /
PRODUCTION AND CONSUMPTION OF CHESAPEAKE BAY SEAFOOD
As part of the effort to evaluate risks and benefits
for this study, it was necessary to estimate the total pounds
of edible seafood produced by the Chesapeake Bay for a given
period of time. Consumption was estimated frem production
or catch and by conversion of live weights into edible
portions. Consumable seafood products were obtained from
three sources: commercial seafood industry, sport fishing
and indirect c ons amp t iori, whereby some fish are used to
make fishmeal, which is fed in small amounts to broiler or
fryer ch ic kens.
The amount of seafood in pounds from comercial fishing
was listed for 1975 s the most recent year for which annual
data was available* A survey of recreational fishing was
conducted in 1974 by the National Marine Fisheries Service,
and tbls data was used to approximate the amount of fish obtainei
from sport f i s h .1 n g -.
The risk anal/si In this study was conducted for
a "worst" case1* situation. Tho data presented in this
appendix was consistent with worst case assumptions. For
example, it was assumed that all fish caught are consumed,
and no allowance Is given for spoilage.
-------
The Chisaprake Bay Bipports a large commercial seafood
Industry. In 1974 an-1 19 ; 5 the Bay produced approximately
6% of value snd 12Z of poundage of total U.S. commercial
catch. The Bay produces approximately j<'<£ of the nation's
non food fish used in the production of fishmeal. Sport
fishing is an important itdustry. For most species of fish,
total catch from sport fishing exceeds commercial catch.
The data used ti approximate pounds of Chesapeake Bay
seafood in elible form and corresponding values is presented
in the following tabl ?s. A description is provided for
each table.
Tables 1 and 2
These tables lift total pounds (Jive we;ght) and
total value (docksidc or exvessel) for the most important
fin fish and shellfish caught for che states of Maryland and
Virginia in 1974 and 1975. The data include the Chesapeake
Bay and its tributaries and the Atlantic Ocean for «acli
state. Revenues to Vvitermen exceeded $22.4 rillion in
Maryland and $25.4 million in Virginia in 19 5.
-------
Table 1. Total Commercial Catch: Value and
Pounds in Live We Lgh^,_S^ateL Q|T Nary land ,1974 and_1975
Food fish
j>£e£ie_s_
Fin fish
Bluefish
Striped Bass
Grey Trout
Croakers
All other
Total
1974_
Ib
(000)
574.8
3,567.0
411.8
120.4
2,488.6
7,162,6
$
(000)
$46,0
948,0
54.5
18.5
50.4
$1117.4
1975
Ib
(000)
270.9
2,722.2
891.8
639.1
3192.9
7,716.9
$
(000)
$25.0
1,064.2
79.3
52.8
715.9
$1,937.2
Shell fish:
Crabs
Oysters
All other
Total
Non food fish
Grand Total
26,483.0
18,284.0
7,859.8
52,626.8
**
7638.2
67,430.5
5,085.0
12,707.6
2,890.0
$20,682.6
188.6
$22,418.03
25,900.0
15,953.5
6,755.8
48,609.3
7,194.9
63,524.5
5,145.5
12,947.2
2,245.3
$20,338.042
190.3
$22,465.60
* Includes Chesapeake Bay and tributaries and Atlantic
Ocean.
** Includes ale\'ives, gizzard and shad, hickory shad,
menhaden and sharks.
Source: U. 5, T'ppartment of Commerce, National Oceanic and
Atmospheric Association. Current Fisheries Statistics
No, 6952: Maryland Landings, December, 1975, Washington,
D.C. (February, L976).
26
-------
. .'if ic ial f'.atch in Value and Pounds
S r '- i } i i ii h
* *
'!i- 'L ,Y,iLS.^nl£-»_ J 2Z^ an!_JLLL! ._
19; '-
.,663.2 ^4,656,4
' v 7 3 7 . 7 4,844.0
S 2 3 7 4 . 0 i 0 ,1 9 4 . 0
i . :- ,774.9 $19 ,694.4
1975
::j-;;f i;<
F-i i,: i
B !, a c f I
St ; I :,(
C, r e ;, T
" » o a V e
', > ''. '" t
I ^ \ s 1
s
i r ! .
sb
d n.i
r o' \
* s
f. - .
, ',
r. n
, i
5
" ' 0
, 2
? ,5
5
0) (000)
1 / , 3
64.2
6?. . 'i
i , '
3 5 , l
nl . 3
s 2
' f
4
2
2 . 6
4 ,1
hi
i *
ft ' »
05.
38.
85 .
8
1 1
1
2
7
8
1 b
( 0 0 0 )
3 , -!
1 , <
'* , 0
4 , /
1 i , 4
25 ,8
84.
29.
89.
20.
24.
76.
9
1
8
9
,'
4
$
( 0 (i 0 )
$246.
h 4 ; i ,
554.
51 2.
2,977.
$4,932
2
'.'
1
9
9
35,562.0 5,399.9
', ^ 4 4 . 0 4,399,9
43,008.8 0,433.9
84 114.8 $19,433.9
U n v 1 n <3 P I f i f r,
T o r » 1
r. r , ncl T_^ ' a i ^_
include
v * 1 n c 1 u d .
* '-'' * 1 n c 1 u'i -'
s T? d s h F '
^'« ,576.2
'4,377.8
»8 ,954.0
'2 ,230.2
$921.8
120.5
$1042.3
JAii211 1.
48,106.5 $1158.9
6,825,6 161.6
54,932.1 1320.5
164,923,3 $25,486.2
;> a p e a k i - Bay and tributaries a n ( Atlantic Ocean.
i'd, so It and peeler blue crabs.
i i' i c e s , g zzard shad, hickory ihod, manhaden
'- c- '.- p.» r t jn . n t c f Gommc re e , NdtK-ual Oceanic
and A i id o H p h e r i c itsociation. C u r r c n L Fisheries
S a ?. : - i i c s t J o 6 ' j 3 : Virginia Landings, December
1 n 7 r. Wa s i, L n g t o i , D . C . ' F . b i u a r y 1 9 ? 6 ) .
2 7
-------
Tables 3, 4, and 5
In these tables pounds of fin fish and shellfish are
listed by commercial and sport fishing catches. The estimates
from commercial sources weie obtained by summing up the
respective monthly amounts reported in the Maryland and
Virginia bulletins on current fishery statistics, published
monthly by the National Ma Lne Fisheries Service, National
Oceanic and Atmospheric Association (NOAA), U.S. Department
of Commerce. Data is cate;>orized by body of water. Therefore,
it is possible to lerive t'le amount of catch exclusively for
the Chesapeake Bay and its tributaries. Fin fisli are listed
by total liveweight pounds for each category of fish,
whereas oysters and clams ire listed by pounds of edible
meat. For fin fish and bl'ie crabs it was necessary to
convert live weight of fis'i into edible weights. The
conversion factors (edible weight-live weight) were obtained
from Lee J. Weddig, Executive Director of the National
Fisheries Institute, Inc. The conversion factors are one
third for fin fish and one tenth for blue crabs.
28
-------
TnM-> 1,, '
F | -' \ ;< ( r n **|
K H t I mated Consumption of fie let ted
jMid its Tributaries
1 i !.
(;}
i: i ur t f L
( .J t < (i
"live ve
2,23
2,87
33
i t J 0
1 ,':8
M!
«. , , o
v 9 , n
2,17
3 S
i "' , --, U
3 , V 4
i|
8
1
9
3
7
7
6
0
0
8
9
SI 1'
,1
,4
,9
,3
*6
»o
, 2.
*,'/
9 '
,3
-, o
,4
t
6
1
2
7
9
7
0
3
2
9
0
0
) d
0
0
9
5
d
)
0
0
7
0
0
0
C o n s ujryj t i
b edible w
74
95
1 1
90
C4 '/
y £,
20
2 s 2 3
2 . l> 2
2,17
'Si
> . i
j
o
e
6
7
3
1
9
2
5
]
0
8
3
7
n
IK
,o
,1
,3
,1
, 2
,3
s 4
,0
5 /
V ~l
7 »
4 .
h
5
3
1
2
3
5
0
?
2
9
0
9
t)
3
7
0
5
2
8
0
3
L.
0
00
A- 0
b i 1
-c (i j ) »
" < ', ,
i I T! r i
. i > ,! 1 1 h .
i ' f> ') catch a i e I" r oin j i ) I H s u i v ^ -
( S e - H M b r e , V. , I , ;i u i v (. y i - f K e t i e a
! t c. i 0 u r r «' i, I !' i s h ; i i e s - I a f i s (. j > s
The fits fish classification i n c 1 n d e
c i I p ij >l bass, spot, and 1 1 o ;>
p c r a b (; , o y s t e r s and h a r d s ! i
( wt> a k -
-------
Table 4. Total Catch and Estimated Consumption
Selected Fish from Potomac River, 1975
of
Food fish
Commercial Ci tch
Bluef ish
Croaker
Flounder
Grey Trout
Shad
Striped Bass
All Other Fin fish
Blue Crabs
Oysters
Clams
Catch
(1 b liv e we iRht )
620,786
41 ,658
77,825
181,559
144,930
840,105
1,906,863
2 ,068,417
219,711
Consumption
(Ib edible weight)
206,929
13 ,886
25,942
60,520
48,310
280 ,035
635,521
206 ,842
219,711
Non food fish
A1 e w iv es
Menhaden
16,251,530
3,413,633
* Ex eludes Tributaries of the Potomac River
30
-------
!'>;.!< . ! '»i ', . i - '. .. u-< i 1 ,!!;> u -1 : ' a'' ..in jj t J mi u i
-------
Table 6
This table summarizes commercial food fish data from
Tables 3 to 5 and includes pounds of edible fish by species
for the Chesapeake B; y and its tributaries for Maryland
and Virginia and separately for the Potomac River. For the
entire Chesapeake Bay, approximately 30.75 million Ib of
edible seafood products were produced from commercial
fishing. Almost one-half of this is made up of Maryland
oysters.
32
-------
e
o
u
u.
42
CO
<
14-1
rH
rH
01
42
CO
T3
d
CO
42
CO
H
u.
d
H
tu
M-l
O
CO
01
3
r '
re
j>
o
d
to
CO
T3
d
3
o
CL,
0)
rH
43
T3
[i!
14-1
o
CO
0)
4-J
CO
e
H
4J
CO
W
v£>
OJ
rH
43
CO
H
-K
m
r~.
ON
i i
0£
d
M 4
l-l
3
T)
CO
O>
H
1-4
CO
3
43
(
^i
H
CO
4J
H
-a
d
to
>i
CO
CO
01
.M
CO
0)
a
CO
w
0)
42
0
(1)
.12
4J
d
H
M)
d
H
42
CO
H
b
-H
to
H
(J
l-i
0)
6
E
o
(J>
ONONCN*& ON oo oo \iO r~- ^j
cy\ O°^ <"-4 CT^ in ON rH O^ ^>i CN
*****
ro * > *
CN CN CN
"O r~-inooNi iinfHooxc^-
d rH^c'or^-^^sDr^, II^N.
to ^D CN in en -^ CN o i I r^ i43 m OO ."30
to N-' m N| oo oc oo
s * -
rH -J-
rH
Cfj
01
H
CJ
0)
CO CX
CO CO
4.J co en
3 03 1.J 43
jr rj o ai to
CO COl-iCUV-i T342rJCO
01 -HCLIT3E-I OIJ-JCJU
H U-l-^d D-O 0)CO
O OICOSr^TJ-H OJ4JE
CU 3OOO)COrJrH3COtO
CL, !-HrJrH!-l424J! irHr^-H
CO CDUti-iOCOCO
rC
ON
vD
»»
rH
rH
O
oo
CO
d -a
0) d
rH 3 CO
H o e
p, Ou to
H rH
3 CJ
cr
0; 13
1^ « CO
0) -H 42 0)
r«t 42 4J CJ
CO 4J d r-l
1 0) 3
CO CU 4J O
4-> d 1 co
42 O 0)
60 d to
f-H ^, O W
0) 43 tO
s X t)
42 43
42 CO rH
CO -H 4-1 CO
H 14-1 42 CJ
14-1 60 -H
4-1 -H 4J
d o 01 co
H & *H
42 rH CO
60 to 4J
. -H 4J CO
E cu o
rJ ? 4J O)
O 42
H-l rH 60 JJ
to d
4-J 4-J »H C2
CO O >s -H
CU 4-J rH
i a co
O 00 -H 4J
4J d -u 42
1 -H rH 60
>-, >-, 3 -H
T3 -H E 0)
co cx 5
0) -H >,
»-i 4-1 43 0)
rH i 1
d 3 -a 43
*r-l Q Q) *H
> «
a >, -H o)
a> 43 i-i
4-J C1J d
tO "O ^O tr^
4J 01
CO p, 0) T3
H u a>
o) i-i cu 4J
U OJ 3 t-J
cfl T3 O
,JD O.
co 01 to 01
01 r-l Ul M
60 CO O
CO 0)
T3 T3 0) M
d d 3 to
3 tO -<
o ,a m
ii, '/I r-l
4-j a; a)
rH Q) rH 4J
r-, rH JD CO
<3 r 1 -H >^
H T3 O
y-i cu
T)
O 4-1 d
4-J O CO
-------
Table 7.
Th' National Marine Fisheries Service, NOAA conducted
a survey of matins recreational fisheries for the Northeast-
ern U,,S, for 1974. Maryland and virginia were also included.
The result'.- of this survey, currently unpublished, were
provided to EPA by the National. Marine Fisheries Service.
Data on sport fish catch was obtained by telephone, mail and
household survey. For each state the following species were
separated individually; bluefish, white perch, striped bass,
spot, and weakfish. The other categories included all other
fin fish, hardshell clams, oysters, and crabs. According to
Ernest Mabrey (NOAA), the senior author of the study, 87.5%
of the total reported weight of catch was from the Chesapeake
Bay. The survey provided percentages as follows:
Species of Fish Percent of Total
Finfish by Weight
Bluefish 47
White Perch 9
Striped Bass 10
S P o t 5
Weakfish 8
All Other 21
Since the percentage breakdown i;f. v, rights for shellfish was
unavailable at t.ho time of thin vr/ I t ing , it was assumed
that crabs made up 80% of shs?1lfJsh, and oysters and clams
each account e d f o r 101 f o; a total of 100%.
3/4
-------
AJ
n
O
ex
C/J
E
0
!j
i4H
_c
CO
H
IM
-H
*H
0)
,n
C/)
c
c -t
CO i-~
i-; r~1
U3
( at
[Xj C*
^
C M
H 3
to *o
<*-( aj
O CU
H
01 IJ
3 TO
rH AJ
CO 3
i-l
a M
C H
cfl
AJ
CO I
-a
c -u
3 C
O CO
PJ
d) CO
t t CO
J3
T3 .S
W CO
OJ
O CO
Cfl
4) CU
AJ ,£^
CO cj
e
°H Oi
AJ jr
CO AJ
Id
Cj
H
r- 60
c
CJ i
>-H ,K
43 W
Cfl -H
t ' ill
1J
»
to a)
4-1 r- (
O ,XO
H -H
-a
u
.
AJ
3
OJ
*~ i
^c
CO -H
~f '0
C W
4
60
U
*H *
^ >
. -1
«!
AJ
O
H
*
3
OJ
2
-a ~»
C T3
CO W
rH
>-T w
».J AJ
CO 3
£ ,H
cc
AJ
O
H
U
01
H
C.
01
a
w
OOOOOOO JOOC5
OOOOOOO-Ocn-3-O
0000000-3. O^^IP-I
>£) .T ,t ,.( ,0 ,;j o 1-- ^
D cy> 1.1 I-H o > t > i ex; rH
en ,-j, ,-( rH D
rH ~* 00 -* f~ <": ft
Cs| fl yr> l~ C-J O4
r. 4 ». ««
t^-»- y* ^ ^H rH ^1
OQOOOOOOOC'O
OOOOOOOcnro-. O
ooooooo-*-^---- ro
£) c^i 0" co r^- co CM oo co co i^- * r-- rH O 00 ir !--
..C rH rH rH CO -*
f ^
O C, O C) O c.
o o o o o o
o o o o o c
,0 to H l»-l
-c. tn [i, to rH
U CO rH
I. CO i, »i k-l flj
,u at co a) o ,c
tn o, -a 4 c. >n -.1; to
-I CD U i A 1 l-J
>A4 a. D c .- < 'A 0,1 o] f i
a) 4-> H 4 i ,b»i co 3 AJ ,c to
S-IMOtflcHAIcOCfltCAJ
r-H ,j; AJ G, CD r--< O -1 > i~- O
PQ rz t/j 1/1 S <; H t.? '" > ' f
o
o
to
-t
I I
rH
t*H
CJ
O
O
o
r j
, H
m
t«
1/1
O
0
CO
Cxi
0
vO
in
rH
*
tO
AJ
O
n
a
c
to
M
o
a
AJ CU
C CJ .C
U C -H
WJ Cfl rH
3 OJ JD
O CX
C
.H 3
M cfl '-'
OJ C
AJ O
« I O
w? iJ 0
i cfl a
1-J «-^4
^H *
03 » C
CO QJ O
U 4J
C» '^ &0
" £ -5
t>0 ^J /"C-
c: to V)
H 03
j^ co 3^
CO , -H CO
cu AJ a
j> CO 0)
i-i "z, a
3
c/^
» C-4
. J xD ;j
u' o rJ
»3 C
> CO Al
01 0 tl)
>... .H -M
J3 tJ CJ
CO CO O
^ T\ VS
AJ CO
cfl
-------
In order to assign a value to recreational benefit
from fishing as distinct from the value of fish for food,
the total expenditures for fishing reported for each state
was proportioned out for the Chesapeake Bay. Then, the
total weights of edible fish by fin fish and shellfish
categories were used in assigning weights for letermining
per pound of edible meat. The values are reported as
follows:
State
Ma ryl and
Virginia
Fish
Category
Fin fish
Shellfish
Fin fish
Shellfish
Weight
Dist r ibut ion
86
14
80
20
Edible Pr iducts
$ per p >und
$2.4 >
$7.8 5
$2.37
8.1t
Total
Expend itur es
$35,230,000
$ 5,933,000
$12,231 ,000
2,976,000
Table 8
Manhaden and alewives are the major ingredients in
the production of fishmeal. The Chesapeake Bay, an important
producer of non food fish, supplies approximately 30% of the
nation's total production. By weight, fishmeal is approxi-
mately 21.005% of the original live weight of fish. This
percent was obtained from the ratio of the contribution of
manhaden in fishmeal to manhaden catch. The ratio was:
.21005 = 199.2 thousand short tons of manhaden in fishmeal
1,896.7 million pounds of raw manhaden
The figures represent total U.S. production and
are averages for 1970-74.
-------
C'-i
-t
'U
0)
O
CO
CO
.-i
ttj
w
-H
O
en
o
iH
W
0)
O
a) j
.-( C
*f3 n)
a c
c o
CO -H
DC
c v i
a.) i*: i
o <
co
C
cfl
5S
«fl
PQ 01
n-i
Q 0; ,£
ij (fl
M CO
13 0) fa
c a
CO 14-4
05 O
0)
r: w
3
O
»H
13
4,1 01
o ,c
t-'l 4J
to
o
tfl O
4J
o ty-
H
H CO H
O C
OJ d OJ
a co
-------
Tab la f)
This table lists prices for each species. The prices
are dollars per pound for b o f" h live weight and edible fish.
Live weight prices a r >.> obtained by dividing total revenues
by state received by watermen by total pounds of catch
for 1975. This procedure is also used for the category,
"all other fin fish." The average live weight price column
is obtained by weighting prices by pounds of fish produced
in each state. Average prices for edible portions is
obtained by dividing live weight prices by the fraction
of edible portion to live weight of fish.
In the analysis, the prices were used to derive relative
values of each seafood product. These relative prices were
used to rank tVie seafood products by value. The study
should not be interpreted as an economic impact analysis.
However, it could serve as a basis for one.
-------
Table 9. Averages Prices for Chesapeake Seafood Species,
1975.
Species
Bluef ish
C roa ker
F 1 ound er
Grey trout
(Weakf ish)
Shad
S t r iped Bass
Ma ryland
.09
.08
.09
.39
Virginia Average
.07 .08
.11 .10
. 284
.14 .12
. 266
.48 .44
Av er ag e Price
.24
.30
.852
.36
.798
1.32
(Rock)
All Other .22 .22 .22 .66
Fin Fish Used
as Food
Weight Price .025
of Alewives
and Menhaden
Fishmeal .119
* Prices are dockside or exvessel equivilent.
Source: tl. S. Department of Commerce, National Oceanic
and Atmospheric Association. Current Economic Analysis
1-26, Industrial Fishery Products: Market Review and
Outlook, NOAA JFPOA 26, Washington, U.C. (February 1976).
39
-------
-------
APPENDIX B
-------
-------
APPENDIX B
LIST OF TABLES
No.
1 Population by Cities and Counties Contiguous
to the Chesapeake Bay in Maryland and
Virginia, 1 9 7 0 ....<...... t ........ ,.......,»
-------
APPENDIX B
ASSUMPTIONS USED TO COMPUTE IMPLIED RISK
TO A GIVEN POPULATION
In the equation,
P= i* aj fj Wj
The following variables are discussed:
aj = action Levels for species j
fj = percent of total production of species j consumed
locally (cities and counties contiguous to the
Chesapeake Bay)
Wj = Total pounds of species j produced in the
Chesapeake Bay annually (based on 1975 data)
Action Lev el s
The a : t i o n levels chosen are the ones that are currently
in effect for shellfish, crabs and oysters and are:
j (Fin fish) =0.1
1
a (Oysters) =0.3
'>
£.
a (Clams) =0.3
1
a (Blue Crabs) = 0.4
l\
42
-------
cropoi I ius M Fit: l> ( HtBunit'sj Loral I;'
f \i ., , .: ,' f 1 i pci .- t is! of :" p e e ios c o n n ! inf! local 1 y are;
i { i: MI i I r-li) 1) /'
1
f nr-f Sjr err,) - 722
>
i , f I aws) 40 2,
! ( H i. u L f i a h ;> ) - 55 %
f. . ,1 V cl o «'>ii'. ioi M < ,- i c r si - ' l.ini'i, -'ii','< c^al-s are
c .>!<<;« u i »:; '' !]!on of thr= UnMrci '". : -i i "" a , -'- ' rroduc-
; ' "' 'ir,'; >i ! j i, i ,..,, aii" ci-( i1 i. ii 1 i^i.ifn '" ; "4 n <\ '' For
''/
. , -i c i ,il . < e;:p,'. ; i-s p i y 1 ;.< :'-. ( i ) i-r cent
',>,.,-(, t li t i" a I i c I >{; p i MC r ii '.. ; '"i ' > ; .1 t- j n r f
': ';'! h ii t t r ii s'f r i. u i r r po i f r.; i ; . i n fish.
"> '"
'" '; ' r ".>';;<(«> v> i:. in,,uli n i total f in ' * ''< - is JMMM ]»t 1 on
>u1 '- I f 1 y ; >- :,-,!,i i pusn.St, u ! ( '. n f i -( i;
! : ' ' ! - - - sfc,e !". i f IK ' Mfh :', i !
ll< ' f <>!;! i |»o p "I ] ;i I Ion : tli,' f" h f
(Tcf> Tf'Mf M, ! -. ubtatn the .!F-nfr;-d prr.p'---' i-- " lii :
t. r o" * ('- T" i ;'. i " 'l , in (urn., f ii i 5, (I»--fis«' - ' :,r.'-i\ to
'"" - ' ; , ', . ' \: > !, ii , Hi'UJtl'.il ,111(1 (.'lllfc'l
f c< i. P ! i i I l h r . ' I i -. i I m i n a , !' I nd ! tip. H ,
i,- l S >i i' i ',> .1 t C ! i i !; i ,< ; "j i .' N a ! i on n i H a - i n c
N a r i o n a I Oo > ;< i> f t mf A i m t, - .j hi- r t r A:, s .ic L a
,n c 111 C! t" C ninms- r r r - ' f- a t i 1 r ) .T ^ !'!.]>.
-------
Table 1. Population by Cities and Counties Contiguous to
the Chesapeake Bay in Maryland and Virginia, 1970.
Maryland
Anne Arundel
Baltimore (City)
Calv ert
Cecil
Dorchester
Kent
Queen Anne
St Marys
Somer set
Talbot
Balt imore
Total
298,042
905,787
20,682
53,291
29 ,405
16,146
18,422
47 ,388
18,924
23 ,682
620.409
2 ,052 ,178
Virg inia
Accorr oc k Co .
North Hampton
Landraster
Midd]esex
M a t h e w s
Glouc ester
York Co.
Isle of Night
Hansemond
Norfolk (City)
Newport News
Hamp ton
Virginia Beach
Chesapea ke
29,044
14 ,442
9,239
9,126
6,295
7 ,168
14,059
33,203
18 ,285
35,166
307 ,951
138 ,177
120 ,779
172 ,106
89,580
Total
44
-------
Tgta_l Pounds Ed 1 b 1 e Seafood
The following list of figures are estimated pounds
of fin fish used as food from both commercial and
sport fishing sour' es by species and location.
State
Spec ies Ma ryland Vi rg in ia
Fin fish 16,451,311 10,821,616
Shellfish
Crabs 2 627,400 3,221,073
Oysters 15 173,917 2,208,222
Clams 961,857 395,890
Population Estimates
The >opulation chosen consisted of state counties and
cities which are contigious to the Chesapeake Bay. The
population by counties and totals for each sub region for
Maryland and Virginia are listed in Table 1. The populations
are 2,052,178 and 1,004,580 for Maryland and Virginia,
respectively. Since these populations are arbitrarily small,
the resulting probabilities of development of tumors are
somewhat high. However, this does provide a worst case
est imate.
-------
-------
A ' P E N I) I X C
-------
-------
APPENDIX C. FIGURES
N o .
Pae e
1 R i < ]' Hiulpj Present: Action Levels and Assumed
C n is s u m ji t: L i ' i * , , * »...«»».! . ,, » .... e ...........
5 2
2 F o I n : ' i , 1 1 c 1 1 ] , s t > .-I T inn o r ft Under Present Action
L e-f c. I > a n d U n i a t m C o n s urn pt, ion...... ........ ...«.«» 5 3
-------
AFl'EtiDi;, C
RISK TO FOI'HI.ATTmi ^i S IJM ! N (: VH: rKOBIT MODEL AND DIFFERENT
KKPOSUK I1, I'M f JT.RNS
In th? case > i' kepone, the utility of using a two
parameter model to e r, t i m a t o ; i : l> is s ev e r e ] y limited due to
the n a L v r o s < t" L h e o h s. P r v <> d animal data. Tie response 1 ev e 1 s
are extremely close t o> rth^r MI 11: rig any estimate of slope
highly sci i L i Iv e to rardom b i c i .!;; ic a 1 error. Also, only two
exposure > ev e 1 n ex st v n i r h tnikos any evaluation of the
conformity of a two parameter model to the data impossible.
Furthermore, n joint probability distribution of the amounts
of kepone in finfLsh, < tabs, and shellfish taken from
Chesapeake Bay .m i 11 e f such a distribution is not
av a i 1 ab 1 e , a series of B o m o wh n I- arbitrary assumptions
concernirij consumption were nade in order to estimate risk.
It must b recognised that tie risk estimate d e r iv e d from
this model, is sensitive to thcr.c- assumptions. For these
reasons, it" ir-; f o 1 t )-'=.. M > r^'libllity of the "log-probit"
method docs n o f ' a<< r: ; ' . J i- r of r e a 1 i b i 1 i t y as that
possessed by t ! " " : r\ c - i ' ' : ' ' I u a t i o n .
-------
. The expected number of tumors under thi log-probit
h. , , model may > e expressed is:
T = X NSPS
wher e
p - _J ..a
s /2iT -°° fc e ~w
is the probability of a tumor in the sth exposure group due
to an agent and Ns is tie total number of exposed people in
the sth group.
The term Xs = >\ a-W^s/1930 is the ppm in the diet of the
0 = 1 J J
sth exposure group and the other terms are defined below:
m = number of exposure groups
a, B are the intercept and slope parameters of the
log-probit model
k = the number of different food stuf's
aj = action level for jth food stuff
wjs = grams eaten/day of jth food stuff by a
individual in the sth exposure group
In order to evaluate cancer risk using this model
the parameters a; 3 tnust be estimated and an exposure pattern
specified. However, since the Kepone data does not lend
itself to the estimation of a reliable slope for the log-
probit modrl it was necessary to assume a conservative value
for the slope. Following the original suggestion of Mantel-
Bryan, a slope ofg=! was assumed. An intercept value a was
estimated by taking the weighted average of the Intercepts
obtained from estimating the intercept for each dose level
separately in the heptacellular carcinoma male mice data
set. Proceeding in this manner, the log-probit equation
Zp = -.437 + log x
is obtained.
49
-------
Under the assumption that each individual eats fish,
crabs, and shellfish in the same ratio as they are consumed
by the entire population, an implied acceptable risk level
can be obtained given the action levels and the average
grams of seafood consumed per day by the individuals. In
this case the ppm in th3 diet x is simply:
x » a~w/1930
where a" is the weighted average of the action levels of
each seafood type weighted by their total consumption and
is the total seafood consumption in grams/day. In the case
of the present action levels T - .1857 and the log-probit
equation can be written as:
Zp * - 4.454 + loslOw
and is shown is Figure 1. The risk of any average daily w
consumption can then be mathematically or graphically
from Figure 1.
To calculate the total number of tumors produced, it
is necessary to specify a consumption pattern or distribu-
tion of Chesapeake Bay seafood. The difficulty in such a
task is caused by the fact that most indivJduals are not
aware of the origin of the seafood products they consume.
However, under the assumption that each individual eats the
same amount of seafood the number of tumors can be calculated,
50
-------
;t i i,ii-, uni it ion o) v r r a ni s / il a y i h i n i.:r>},>- flint
!J - 7.985 x 10'Ar
> q a, 1 * - ' ' ," IIT b e r of p e o p ] o t li a f: R -i t a 1 1 t h e s r; a f o o d
h ,'i v -p f i i i.»( . r T h r f- n (- n 1 i" 11 m o in 1 ; t x p r r- r, s n ij n n ;
,=, '- -- '-- -. i e"" d I-
i' ' i g n i <> ..' 1 '»r fi 1 i ? ') 1 uef: ot w ''»? rn-i r-
i« li a t t ii e : r ti e. rl i r i i ub t J o n p a ; , . '! i \ . ^ u i :)
if i II ,1 ll '; , '' f) ii i 'i i.l <:!',, The Till.' ll ;i,.i ',1 t' ' U!i ri '
, ', ; t i H> u t I i> n i :i -n n s t i ikolv H(>"ir ^s If, lit -<1
>' 10 and 300 p, rdM,-;/ day g iv ing 'if. a ri mlu
> i A , S 0 0 t n in o r s ,. M must HP r '-in ' 'nh p T or' that
'i -i r i- I) 1 g h 1 v -'> p e ' i; 1 a f I v e «
-------
RISK
P
.5
-4
.3x10
-7
.61x10
Zp
0. w
.159 -1.
.067 .1.5
.023 -2.
.00135 -3.
-4.
5.
APPENDIX C. FIGURE 1 RISK UNDER PRESENT ACTION
LEVELS AND ASSUMED CONSUMPTION
a = .1857
-1. 0. 1. 2. 3. log 10
.1 .316 1. 3.16 10 31,6 100 316 1.000 w gms
-------
'PENDIX C. FIGURE 2 TOTAL CACULATED TUMORS UNDER PRESENT ACTION LEVELS
AND UNIFORM CONSUMPTION
TOTAL
TUMORS
IN 103
7.
1.
0.
a -.1857
T-NP f 2p j_
N = W,P = i J eT
w %/2n -
dt
Iog10w
.1 .316 1 3.16 10 31.6 100 316 1,000
10.000
-------
------- |