Doellingeria umbellata (Miller) Nees var. umbellata | Aster umbellatus | 1/2 mile north east of Fort Saskatchewan (Mr. Jaeoy's?? Farm) | in field of wheat on river flat, broken last last summer | Natural Areas Data Bank, Alberta Forestry Lands and Wildlife

An experiment to determine the plant characters of wheat that caused differences in yield between certain crop sequences and tillage methods was conducted, holding all other factors constant within season. The plots were planted to the same variety, on the same date, at the same rate (with one exception), and with the same type of drill. The final yield of winter wheat for the years and treatments studied was shown to be closely determined by the number of kernels per unit area. The kernel weight was never found sufficiently low to upset this foremost consideration. These results would not be expected to hold for excessive shriveling or for varieties of varying kernel size. The number of heads per unit area as an individual indicator of yield was exceeded only by the number of kernels per unit area and the number of spikelets per unit area. Variations in number of heads per unit area were much greater than those in the number of spikelets per head and were chiefly responsible for the differences in number of spikelets per unit area. A number of tillers sufficient to produce a large number of heads per unit area is the first requirement for a good yield. The potential number of heads per unit area is indicated from the tillering period but this may be modified by later conditions. Plants per unit area do not usually play an important part in determining number of heads per unit area because they are compensated by the number of heads per plant. The head characters, including spikelets per head, kernels per spikelet, kernels per head, and head length, did not often vary enough to compensate for material differences in head number. In some cases however, the yield level indicated by the number of heads and number of spikelets per unit area was considerably modified by head characters. Conditions seldom remain favorable throughout the entire season. Drought during early growth causes an elimination of weak plants and less vigorous tillers that would otherwise produce heads. The tillers remaining are usually those which will produce heads containing an above-average number of spikelets. A greater number of spikelets per head was observed for treatments and years in which a deficiency of soil moisture was experienced. During the heading stage some yield adjustment in favorable years is made through the number of kernels per spikelet. In comparison with the influence of head numbers on yield such adjustment is very limited. Severe conditions during the heading and flowering period, however, may in some cases cause a reduction in the number of kernels per spikelet large enough to more than compensate for differences in head numbers between treatments. The final adjustment in yield to environmental conditions is made through kernel weight as the plant nears maturity. Significant treatment differences in this character were not apparent, indicating that hazards which reduce kernel weight on one treatment cause similar reductions on other treatments. Seasonal differences in kernel weight were apparent. There appears to be no simple answer to the problem of character response to different methods of cultivation and crop sequence. Detailed plant descriptions at maturity demonstrate that there are many possibilities for plant characters to reflect the moisture and fertility conditions occurring at different stages of plant growth The examination of characters pointed out that continuing adjustments of the wheat plant tend to maintain it in a condition that will enable it to take advantage of the moisture and fertility conditions then existing. For the conditions covered by this study, the number of kernels per unit area, which is the product of the two characters number of heads per unit area and number of kernels per head, was found most practical for estimating yields. When kernel numbers were not considered, the number of heads per unit area and plant height provided the most useful estimate.

A 47,000-acre demonstration project for the control of wind erosion was established by the Soil Conservation Service, U.S. Dept. of Agriculture, in 1934 in the vicinity of Dalhart, Texas. In 1936, a research station, consisting of 912 acres, was established near Dalhart to study the causes of wind erosion and to develop methods for its control. At both of these locations studies were made of the change in elevation of ground surface of certain types of land from year to year. This change in elevation was caused largely by wind action and was found to vary considerably with different types of vegetative cover. About 16,000 10-foot readings were taken on the demonstration project and several thousand other similar readings were taken at the Sand Dune Research Station. These were made during the years 1937 to 1940, inclusive, all of which were years of below normal rainfall. It was found that: 1. The amount of soil blowing from idle or abandoned lands showed a closer correlation to the precipitation that occurred during the previous years than did crop lands. 2. Milo which was harvested with a combine or header type machine suffered much less from erosion by blowing than did milo which was hand headed. 3. Certain crops, such as broomcorn, cane, and sudan, did not lose soil because of the fact that they are usually planted thickly. Crops such as hegari, milo, kafir, and corn lost a considerable amount of soil, apparently because these crops are usually planted rather far apart in order to produce a grain crop. When the spacing of sorghum crops was controlled, differences such as this were not so apparent. 4. Wheat was variable in its ability to control wind erosion, depending entirely upon the kind of stand obtained. 5. The removal of soil from rows which ran parallel to the direction of the prevailing wind was somewhat greater than from rows which were at right angles to the prevailing wind. This difference, however, was slight in comparison to the differences caused by other factors. 6. It is indicated that terraces without vegetation on their surface do not have much effect in preventing wind erosion. When the vegetation on the terrace was heavier than it was in the intervals between terraces, however, there was a definite tendency for terraces to collect most of the soil lost from the intervals between. 7. Weeds were variable in their ability to collect soil and resist erosion. The effect of Russian thistles depended on their growth, small thick stands accumulating soil during most of the blowing season, while large thistles broke away from their roots and left the ground bare. Silver leaf nightshade which is becoming a serious pest in the Dalhart vicinity was generally an exceedingly poor cover crop, often being no better than bare ground. Other weeds had a consistent tendency to accumulate soil during the blowing season. 8. Since winds of high velocity may occur in any month of the year, either removal or accumulation of soil may occur continually throughout the year. Whether a given field loses or accumulates soil, depends upon its vegetative cover at the time of each high wind.