eposition
Table 1. ‘Mean observed ratios of strontium 90
in milk to strontium 90 in feed and their associated variation
-SA)
Mean
Station
observed
ratio
St. Louis_.-...-__..___-__-Brainerd (overall}___._-___Brainerd farm 2L___.._____Brainerd farm 10L_.._2___Brainerd farm i12H________Brainerd farm 14H_______--
.
.
.
.
.
.
046
039
038
035
040
O44
Confidence
limits (95
percent)
.
.
.
.
.
.
035-.
035-.
028-,
025—-.
O81-—.
O37-.
057
043
048
045
049
051
micromicrocuries of strontium 90 per gram of
calcium (strontium units) exhibited more variability in the estimates. However, the basic
result is that
Level in feed=& (level in milk)
sITY
so that the levels in the barn feed should be proportional to the previous harvest season’s milk
values. As an empirical verification of this last
statement, average milk levels during the barnfed periods in 1958, 1959, and 1960 were compared with average milk levels during the previous harvest season in each of thefive stations,
The results are shown in figure 3. The level in
the winter seems to be somewhere between 0.9
and 1.1 times the level during the previous harvest season (June-September). Errors are
present because of the impossibility of stating
which part of the harvest used in barn feed
comes from which part of the harvest season.
Despite the errors introduced, an approximately
linear relationship
B(t)=kifne
n be dem-
od by the
linn. (4),
lies meas‘ations of
ra numthe milk-
[5]
can be used for the barn level at time ¢ where
Strontium 90 Levels in Milk
h Reports
Vol. 77, No. 12, December 1962
ry small,
sheds ly-
M(t) =arp(t) S(t) tasp(the™ f MD (z) dt
t+as([1—p(t)]Instap(t)e™
[7]
Some estimate can be made of the half-residence time in advance because of the sporadic
nature of the milk fluctuations. That is, halfresidence times of more than a year produced
predicted levels which did not correspond well
with observed values. The sharp plunges
which follow peak milk levels without preliminary drops in the deposition rates also indicate
that strontinum 90 leaves the pasture fairly rapidly. Studies conducted by the Atomic Energy
Commission (7) have also indicated that milk
levels are greatly influenced by short-term deposition rates.
Therefore, trial values for 7',,.=
0.603/A ranging from 144 month to 3 months
were used in the final formulation.
Computations
For values of 71,2 (half-residence time) on _
the order of 1 month, equation 7 can be written
appproximately
Figure 3. Comparison of levels of strontium 90
in milk during the barn-fed period and levels
during the previous harvest season
25 ,
207
a
fy
05. Tt is
ted using
iclated 95
M(t) =a,p(t)L(t)+a,lt—p(t)]B(t) [6]
Combining equations 3, 5, and 6 yields the final
model
My, is the average milk level over the previous
harvest season.
The average levels in milk will depend upon
the relative numbers of cattle on pasture and
in the barnsince the feed levels in the two locations will usually be quite different. The proportion of cows on pasture, p(¢), for the five
original milk stations has been taken from Harris and co-workers andis given in table 2, The
average milk level at time ¢, .1/(¢), will then be
computed
related to the average of the barn and pasture
feed levels weighted by the proportion of cows
in the barn and pasture respectively during that
time.
Level
during
previous
|
1s
harvest
season
Gaues1)
pb
10;
5
. f
a
of
ate
T
5
o
o Cincinnati
oNew York City
* Sait Lake City
* Sacramento
4 St. Louls
a
oO
.°
———"
10
15
r
7
20
25
Level during barn-ted period
1059