Journal of Agricultural Research 
Vol. XXVII, No. * 
114 
DISCUSSION 
An examination of Table III shows less difference in yield of tops and 
grain, and of water requirement based respectively on these, than might 
be expected. It is to be borne in mind, however, that the reduction in 
yield is not in this case complicated by diminution of the soil moisture 
to a dangerous point, as frequently occurs in the field when drought and 
rust are coincident. Furthermore, the prolonged growth period un¬ 
doubtedly resulted in an abnormally high water requirement for all cul¬ 
tures, whether rust-infected or not, and tended to eliminate more marked 
differences that might have been evident in a shorter period of rapid 
growth. 
In connection with their extensive researches on the water requirement 
of crop plants, Briggs and Shantz (3, 4 ,5) summarized the results reported 
by earlier investigators. Thus the water requirement for wheat (tops), 
as reported in literature, ranged from 235 found by Lawes at Rotham- 
stead to 554 found by Leather in India. The results here reported lie 
well within this range. Briggs and Shantz found the average water 
requirement for six varieties of wheat to vary from 400 to 500 in different 
years. The water requirement based on grain production, in the present 
instance, is abnormally high in nearly every case, being comparable to 
that reported originally by Briggs and Shantz for plants growing under 
conditions unfavorable to economical use of water, and more than twice 
the figures given in their later papers {loc. cit.). However, at least one 
culture, IVB of the control series, compares favorably with their results 
in all respects. The following conclusions seem to be warranted by the 
data: 
Infection by either stem rust or leaf rust results in a significant reduc¬ 
tion of yield of tops and of grain that is independent of the soil moisture 
supply, and this reduction is evident with only moderately heavy infec¬ 
tion. Taking the figures for the stem rust and control series at face 
value, it appears that an amount of rust estimated at 30 per cent may 
cause a 20 per cent reduction in yield even though the soil moisture 
content remains at a favorable level. 
Accompanying the reduction in yield, there is practically as great use 
of water as in healthy plants; that is, rusted plants have a higher water 
requirement, based on yield of both tops and grain. This is shown to 
be significant only in the case of stem rust and is of a higher order for 
grain than for total dry matter. The lesser effect on yield of tops, 
resulting from rust infection, is probably due to the fact that the soil 
water supply was at all times sufficient; under field conditions dry weather 
may greatly diminish the yield of the entire plant. 
The effect of rust in diminishing yield and increasing the water require¬ 
ment seems less likely to be due to increased transpiration, resulting 
from rupture of the epidermis at postule formation than to other causes, 
for the total amount of water used by rusted and healthy plants is about 
the same. Leaf rust, which produces but slightly erumpent pustules as 
compared with stem rust, and hence less injury to the epidermis, also 
depresses yield, more particularly that of grain, but also of tops, though 
in each case to a less extent. Chief among the causes of yield reduction 
arising from rust infection are to be reckoned the destruction of chloro- 
plasts in the infected tissue and the drain on the elaborated food reserves 
of the host. 
