BUREAU OF ENTOMOLOGY AND PLANT QUARANTINE 
65 
of the results for the decade 1930-39 indicates that the prevalence of 
some races has fluctuated from year to year, that other races have 
decreased in prevalence, some of them almost to the vanishing point, 
that others have shown a general tendency to increase slowly, and that 
one (race 56) increased slowly for a time and then suddenly increased 
very rapidly. Race 56 has ranked first in prevalence for 6 successive 
years, reaching the high of 66 percent in 1938 and dropping to 59 
percent in 1939. Race 56 was isolated from 100 percent of the col- 
lections of rusted bread wheat obtained from North Dakota. If cer- 
tain other races were to predominate in future years as 56 now 
does, the behavior of at least some of these varieties would be entirely 
different from what it now is. 
These studies on population trends, when correlated with the be- 
havior of certain varieties of grain toward rust, indicate clearly the 
necessity for information of this type in explaining the variable 
behavior of commercial varieties and in connection with the breed- 
ing and testing of varieties for resistance to stem rust. For example, 
races 36 and 49 were rather prevalent during the first half of the 
decade but decreased to less than 1 percent in 1939. Similarly, races 
11, 21, and 34, which were rather prevalent during some years of 
the decade, were rather rare in 1938 and 1939. As there is so great 
a difference in the relative susceptibility of wheat varieties to differ- 
ent races, and since the prevalence of races varies so greatly, it is 
becoming increasingly evident that everything possible should be 
done to decrease the number of races. 
Experiments during 1939 confirm preliminary evidence and observa- 
tions that the wheat stem rust fungus comprises an indefinite number 
of biotypes, many of which differ from each other only slightly. 
Further evidence was obtained during 1939 that a great many bio- 
types may result from segregation and recombination on barberries. 
For example, a number of virulent races of rust have been obtained 
as a result of selling a given race on barberry. Furthermore, isola- 
tions from crosses between varieties of stem rust indicate that the 
number of biotypes that can originate as a result of the sexual process 
on the bushes is almost incalculable. There is additional evidence 
that races developed on barberries may become established and be 
disseminated for considerable distances. Such evidence was obtained 
as a result of identifying uncommon races in western North Dakota. 
Barberries are important, not only in connection with the wheat 
stem rust but also in connection with stem rust of oats {Puccini a 
graminis avenae) and rye rust (P. graminis secaMs). This is shown 
by the fact that, of all the identifications of aecial material from 
barberry leaves, wheat stem rust constituted 46 percent, rye stem rust 
34 percent, and oat stem rust 20 percent, in 1939. 
It is now clear that rust which occurs in the southern part of the 
United States is not ordinarily the result of spores blown northward 
from southern Mexico. This conclusion is based on physiologic-race 
surveys and observations. Race 56, which occupied first rank in the 
United Spates for the last 6 years, lias not been found at all in southern 
Mexico. Certain sections of northern Mexico, on the other hand, must 
be considered with the southern part of the United States in the devel- 
opment of rust epidemics. 
267756—40 5 
LIBRARY 
STATE PLANT BOARD 
