STOCHASTIC SIMULATION OF RADIONUCLIDE UPTAKE
809
Table 2— ESTIMATED !“!] ON VEGETATION IN FOUR AREAS IN
SOUTHERN NEVADA AS OF JULY6, 1962*
~
Groom Valley
Vegetation
Stomach contents
Penoyer Valley
Vegetation
Stomach contents
Railroad Valley
Vegetation
Stomach contents
Currant Area
Vegetation
Stomach contents
Variance
Arithmetic mean
of distribution,
pe/g
Size
of
sample
4.075
0.228
17,481
30
3.611
0.272
6,533
27
3.428
0.120
3,695
41
3.488
0.100
3,930
36
3.184
0.017
1,656
29
3.218
0.058
1,980
25.
2.825
2.494
0.110
0.100
890
414
Mean
of logs
31
25
*Estimates are based on decay corrections of bulk samples of vegetation and of stomach contents of jackrabbits (A = 0.126). The antilogarithm
of the mean of the logs is the geometric mean of the distribution and is
not the same as the arithmetic mean.
three times the mean are invariably small, usually less than 2%, An
analysis of actual observations gave similar results. In only 2 of 133
jackrabbits (in 24 samples of 5 to 9 individuals) did the thyroid ‘°4I level
of an individual exceed three times the sample mean,
There is no doubt that, when the model is used in the manner de-
fined, it predicts skewed frequency distributions of '*'I in populations
of consumers, But how well do the predicted distributions agree with
observed frequency distributions? Unfortunately, our 1962 data simply
are not sufficient to make such comparisons. From a given locality
(for a specific time) few animals were taken, and no convincing com-
parisons involving theoretical distributions of 100 or 1000 individuals
and observations of less than 10 can be made,
However, most of the observed distributions suggested asymmetry,
with a few very high values, as compared with the median. For example, two samples are indicated in Table 4, Figure 6 shows these
observations, as well as normal and lognormal distributions based on
the mean and variances of the two samples. The high values observed
in both samples are more probable if the distribution is lognormal
than if it is normal.
DISCUSSION
Although it is difficult to support some of the theoretical predictions
of the model from observations during 1962, other work on frequency