THE confidence with which scientific investigators,familiar with the arid regions, have attacked the problems of dry-farming rests largelyon the known relationship of the water requirements of plants to the natural precipitationof rain and snow. It is a most elementary fact of plant physiology that no plantcan live and grow unless it has at its disposal a sufficient amount of water.

The water used by plants is almost entirely takenfrom the soil by the minute root-hairs radiating from the roots. The water thus takeninto the plants is passed upward through the stem to the leaves, where it is finallyevaporated. There is, therefore, a more or less constant stream of water passingthrough the plant from the roots to the leaves.

By various methods it is possible to measurethe water thus taken from the soil. While this process of taking water from the soilis going on within the plant, a certain amount of soil-moisture is also lost by directevaporation from the soil surface. In dry-farm sections, soil-moisture is lost onlyby these two methods; for wherever the rainfall is sufficient to cause drainage fromdeep soils, humid conditions prevail.

Water for one pound dry matter

Many experiments have been conducted to determinethe amount of water used in the production of one pound of dry plant substance. Generally,the method of the experiments has been to grow plants in large pots containing weighedquantities of soil. As needed, weighed amounts of water were added to the pots. Todetermine the loss of water, the pots were weighed at regular intervals of threedays to one week. At harvest time, the weight of dry matter was carefully determinedfor each pot. Since the water lost by the pots was also known, the pounds of waterused for the production of every pound of dry matter were readily calculated.

The first reliable experiments of the kind wereundertaken under humid conditions in Germany and other European countries. From themass of results, some have been selected and presented in the following table. Thework was done by the famous German investigators, Wollny, Hellriegel, and Sorauer,in the early eighties of the last century. In every case, the numbers in the tablerepresent the number of pounds of water used for the production of one pound of ripeneddry substance:

Pounds Of Water For One Pound Of Dry Matter
  Wollny Hellreigel Sorauer
Wheat   338 459
Oats 665 376 569
Barley   310 431
Rye 774 353 236
Corn 233    
Buckwheat 646 363  
Peas 416 273  
Horsebeans   282  
Red clover   310  
Sunflowers 490    
Millet 447    

It is clear from the above results, obtainedin Germany, that the amount of water required to produce a pound of dry matter isnot the same for all plants, nor is it the same under all conditions for the sameplant. In fact, as will be shown in a later chapter, the water requirements of anycrop depend upon numerous factors, more or less controllable. The range of the aboveGerman results is from 233 to 774 pounds, with an average of about 419 pounds ofwater for each pound of dry matter produced.

During the late eighties and early nineties,King conducted experiments similar to the earlier German experiments, to determinethe water requirements of crops under Wisconsin conditions. A summary of the resultsof these extensive and carefully conducted experiments is as follows:--

Oats 385
Barley 464
Corn 271
Peas 477
Clover 576
Potatoes 385

The figures in the above table, averagingabout 446 pounds, indicate that very nearly the same quantity of water is requiredfor the production of crops in Wisconsin as in Germany. The Wisconsin results tendto be somewhat higher than those obtained in Europe, but the difference is small.

It is a settled principle of science, as willbe more fully discussed later, that the amount of water evaporated from the soiland transpired by plant leaves increases materially with an increase in the averagetemperature during the growing season, and is much higher under a clear sky and indistricts where the atmosphere is dry. Wherever dry-farming is likely to be practiced,a moderately high temperature, a cloudless sky, and a dry atmosphere are the prevailingconditions. It appeared probable therefore, that in arid countries the amount ofwater required for the production of one pound of dry matter would be higher thanin the humid regions of Germany and Wisconsin. To secure information on this subject,Widtsoe and Merrill undertook, in 1900, a series of experiments in Utah, which wereconducted upon the plan of the earlier experimenters. An average statement of theresults of six years' experimentation is given in the subjoined table, showing thenumber of pounds of water required for one pound of dry matter on fertile soils:--

Wheat 1048
Corn 589
Peas 1118
Sugar Beets 630

These Utah findings support strongly the doctrinethat the amount of water required for the production of each pound of dry matteris very much larger under arid conditions, as in Utah, than under humid conditions,as in Germany or Wisconsin. It must be observed, however, that in all of these experimentsthe plants were supplied with water in a somewhat wasteful manner; that is, theywere given an abundance of water, and used the largest quantity possible under theprevailing conditions. No attempt of any kind was made to economize water. The results,therefore, represent maximum results and can be safely used as such. Moreover, themethods of dry-farming, involving the storage of water in deep soils and systematiccultivation, were not employed. The experiments, both in Europe and America, ratherrepresent irrigated conditions. There are good reasons for believing that in Germany,Wisconsin, and Utah the amounts above given can be materially reduced by the employmentof proper cultural methods.

The water in the large bottle would be required to produce thegrain in the small bottle.

In view of these findings concerning the waterrequirements of crops, it cannot be far from the truth to say that, under averagecultural conditions, approximately 750 pounds of water are required in an arid districtfor the production of one pound of dry matter. Where the aridity is intense, thisfigure may be somewhat low, and in localities of sub-humid conditions, it will undoubtedlybe too high. As a maximum average, however, for districts interested in dry-farming,it can be used with safety.

Crop-producing power of rainfall

If this conclusion, that not more than 750 poundsof water are required under ordinary dry-farm conditions for the production of onepound of dry matter, be accepted, certain interesting calculations can be made respectingthe possibilities of dry-farming. For example, the production of one bushel of wheatwill require 60 times 750, or 45,000 pounds of water. The wheat kernels, however,cannot be produced without a certain amount of straw, which under conditions of dry-farmingseldom forms quite one half of the weight of the whole plant. Let us say, however,that the weights of straw and kernels are equal. Then, to produce one bushel of wheat,with the corresponding quantity of straw, would require 2 times 45,000, or 90,000pounds of water. This is equal to 45 tons of water for each bushel of wheat. Whilethis is a large figure, yet, in many localities, it is undoubtedly well within thetruth. In comparison with the amounts of water that fall upon the land as rain, itdoes not seem extraordinarily large.

One inch of water over one acre of land weighsapproximately 226,875 pounds. or over 113 tons. If this quantity of water could bestored in the soil and used wholly for plant production, it would produce, at therate of 45 tons of water for each bushel, about 2-1/2 bushels of wheat. With 10 inchesof rainfall, which up to the present seems to be the lower limit of successful dry-farming,there is a maximum possibility of producing 25 bushels of wheat annually.

In the subjoined table, constructed on the basisof the discussion of this chapter, the wheat-producing powers of various degreesof annual precipitation are shown:--

One acre inch of water will produce 2-1/2 bushels of wheat.
Ten acre inches of water will produce 25 bushels of wheat.
Fifteen acre inches of water will produce 37-1/2 bushels of wheat.
Twenty acre inches of water will produce 50 bushels of wheat.

It must be distinctly remembered, however, thatunder no known system of tillage can all the water that falls upon a soil be broughtinto the soil and stored there for plant use. Neither is it possible to treat a soilso that all the stored soil-moisture may be used for plant production. Some moisture,of necessity, will evaporate directly from the soil, and some may be lost in manyother ways. Yet, even under a rainfall of 12 inches, if only one half of the watercan be conserved, which experiments have shown to be very feasible, there is a possibilityof producing 30 bushels of wheat per acre every other year, which insures an excellentinterest on the money and labor invested in the production of the crop.

It is on the grounds outlined in this chapterthat students of the subject believe that ultimately large areas of the 'desert"may be reclaimed by means of dry-farming. The real question before the dry-farmeris not, "Is the rainfall sufficient?" but rather, "Is it possibleso to conserve and use the rainfall as to make it available for the production ofprofitable crops?"