UNIVERSITY OF CALIFORNIA COLLEGE OF AGRICULTURE AGRICULTURAL EXPERIMENT STATION BERKELEY, CALIFORNIA COTTON IRRIGATION INVESTIGATIONS IN SAN JOAQUIN VALLEY, CALIFORNIA 1926 TO 1935 FRANK ADAMS, F. J. VEIHMEYER, and LLOYD N. BROWN BULLETIN 668 August, 1942 UNIVERSITY OF CALIFORNIA BERKELEY, CALIFORNIA CONTENTS PAGE Investigations and experiments at Shafter, 1926 to 1930 3 Soil-moisture record, 1927 to 1930 5 Plant responses to irrigation, 1927 to 1930 5 Summary of principal experiments at Shafter, 1927 to 1930 7 Investigations during 1931 7 Descriptions of the areas 7 Amounts of water applied and soil-moisture conditions 9 Summary of irrigation and soil -moisture conditions, growth, and yield for all areas, 1931 12 Soil cracking 13 Root development 15 Investigations during 1932 and 1933 15 Irrigation treatments, 1932 and 1933 17 Method of soil sampling for moisture determinations, 1932 and 1933 18 Soil-moisture record, 1932 and 1933 18 Use of water, 1932 and 1933 38 Main-stem length growth, 1932 46 Blossom count, 1932 50 Yields of cotton, 1932 and 1933 52 Investigations during 1934 56 Irrigation treatments, 1934 57 Replication and sampling plots, 1934 58 Soil-moisture sampling plan, 1934 58 Final moisture-equivalent and permanent -wilting-percentage determinations, 1934 59 Cultural record, 1934 59 Soil-moisture record, 1934 61 Use of water, 1934 63 Main-stem length growth, 1934 63 Yields of cotton, 1934 65 Weight of plants, 1934 66 Investigations during 1935 66 Irrigation treatments, 1935 67 Replications and sampling plots, 1935 67 Soil-moisture sampling plan, 1935 68 Cultural record, 1935 69 Soil-moisture record, 1935 71 Use of water, 1935 72 Main-stem length growth, 1935 72 Sheds of bolls and squares, 1935 75 Yields of cotton, 1935 76 Weight of plants, 1935 77 Uneven growth of cotton plants, 1935 77 Depth of use of water by cotton plants 79 Cotton classification and seed and lint determinations 81 Discussion of results and summary 85 The soil-moisture record, 1927 to 1935 87 Plant responses t 89 Water applied and water used 92 COTTON IRRIGATION INVESTIGATIONS IN SAN JOAQUIN VALLEY, CALIFORNIA 1926 TO 1935 1 FRANK ADAMS, 2 F. J. VEIHMEYER, 3 and LLOYD N. BROWN 4 Irrigation studies on cotton during the five years 1926 to 1930 were pre- sented in Bulletin 537. 5 These studies were conducted by the California Agricultural Experiment Station in cooperation with the United States Bureau of Plant Industry Division of Cotton and Other Fiber Crops and Diseases, at the United States Cotton Field Station at Shafter, Cali- fornia, on sandy-loam soil. From 1931 to 1935 further investigations and experiments were made at other locations in San Joaquin Valley where soils are of heavier texture than those at Shafter. The investigations and experiments from 1931 to 1935 were conducted cooperatively with the United States Bureau of Agricultural Engineering Division of Irri- gation and with the Bureau of Plant Industry Division of Cotton and Other Fiber Crops and Diseases. 6 All of the investigations and experi- ments were with Acala cotton. INVESTIGATIONS AND EXPERIMENTS AT SHAFTER, 1926 TO 1930 The soils at Shafter are sandy loams of the Delano series. They were found to have an average moisture equivalent in the upper 5 feet of about 10.0 per cent, with an average permanent wilting percentage of about 3.5. The last number is the average soil moisture percentage at which cotton plants growing in this soil could extract moisture without wilting. 1 Received for publication June 11, 1941. 2 Professor of Irrigation and Irrigation Economist in the Experiment Station and on the Giannini Foundation. 3 Professor of Irrigation and Irrigation Engineer in the Experiment Station. 4 Specialist in Agricultural Extension. 5 Beckett, S. H., and Carroll F. Dunshee. Water requirements of cotton on sandy loam soils in southern San Joaquin Valley. California Agr. Exp. Sta. Bui. 537:1-48. 1932. 15 The Bureau of Agricultural Engineering Division of Irrigation, W. W. McLaugh- lin, Chief, contributed substantially to funds, equipment, and personnel ; and the United States Cotton Station, F. W. Herbert and George J. Harrison successively in charge, aided materially through consultation, and during 1935 supplied field headquarters and laboratory facilities at Shafter. The field work was in the main done by Lloyd N. Brown and in part, during 1931, by L. C. Schultz. Dr. L. D. Doneen, assisted by Roy Wray, took over the final field work, beginning October 1, 1935. In the preparation of this report for publication, the authors have had the collab- oration of several members of the University of California College of Agriculture Division of Irrigation, particularly Professor C. V. Givan, who analyzed the data relating to the use of water by the cotton plants in 1932, 1933, 1934, and 1935. Assistance was given in typing and preliminary drafting by the Works Progress Administration under Project 8817 B-13. [3] 4 University of California — Experiment Station Investigations during the first season at Shaf ter were preliminary and were the basis of planning the experiments of the following four years. In the latter there were three principal treatments: treatment 1, in which the average moisture content in each of the upper 5 feet of soil in each plot was to be maintained above a lower limit of 7 per cent, the approximate midpoint between the average moisture equivalent and the average permanent wilting percentage — an objective that was not fully accomplished; treatment 2, in which the cotton plants were to be irri- gated whenever they showed wilt at 4 p.m. but would recover during the night; and treatment 3, in which water would be applied when the plants showed wilt at 9 a.m., that is, made little or no recovery during the night. Wilting of the cotton plants at 4 p.m. and 9 a.m. was a more or less arbitrary standard used by the federal cotton specialists at Shafter to indicate approaching or present moisture deficiency, and it was adopted in the irrigation experiments on their advice. The federal cotton specialists also considered that a change in color of the foliage from a light green to a dark bluish green indicated approaching wilting. Two other treatments — 4 and 5 — were variations of treatment 1. Treatment 4 was irrigated the same as treatment 1 until about July 20, after which water was not applied until the plants wilted at 9 a.m. ; treat- ment 5 was irrigated up to about July 20 whenever the plants wilted at 9 a.m., and thereafter as treatment 1. In 1930 there were two additional treatments : treatment 1a, irrigated when the plants changed from light green to a dark bluish green ; and treatment Iaa, which was to be irri- gated when the moisture in the upper 5 feet of soil had been reduced to an average of 5.5 to 6.0 per cent, although the record is not clear as to whether this objective was accomplished. Supplemental experiments were made in tanks in 1928, 1929, and 1930. In these supplemental experiments, treatments 1, 2, and 3 of the plot experiments were repeated. Since comparable experiments were not conducted from 1931 to 1935, the tank experiments will not be discussed in this bulletin. Although moisture determinations were made from about 3,000 soil samples in each of the last four years of the plot experiments at Shafter, 1927 to 1930, it was deemed sufficient, when publishing the results of the experiments, to present what were considered typical soil-moisture charts, and no complete analysis of the seasonal soil-moisture changes in the various treatments was attempted. But the rates of use of water each year for the different treatments, the average moisture content of the soil at the time of each irrigation, the average use of water by months each year, and the irrigation data were tabulated (tables 6 to 12 of Bul- letin 537). Bul. 668] Cotton Irrigation 5 SOIL-MOISTURE RECORD, 1927 TO 1930 The average soil moisture in the plots in treatment 1 was above 7 per cent, except in the first and second feet. But the moisture before irriga- tion frequently dropped below 7 per cent and apparently to the perma- nent wilting percentage in each of the upper 3 feet. Having received an average of 7 irrigations each year, the soil moisture was above the desired 7 per cent much of the time. While no definite minimum moisture per- centage was sought in the other treatments, since wilting at 9 a.m. or at 4 p.m. or the color of the plants was to determine the times of irrigation, the average soil moisture content at the times of irrigation was within about 1 per cent of the average permanent wilting percentage of 3.5 per cent in the upper 2 feet in treatment 2, to the depth of 5 feet in treat- ment 3, and in the upper 2 feet in treatment 1a. The data for the other treatments were not reported. Summarizing the soil-moisture records for treatments 1, 2, and 3 : The soil moisture in treatments 2 and 3 was definitely lower than in treat- ment 1. In all three, however, it frequently dropped to the permanent wilting percentage before irrigations. The depth of water used by the cotton plants in treatment 1 was about 29 inches, and in treatments 2 and 3, about 21 and it) inches, respectively. In each case these figures include evaporation losses from the soil and also any water that may have been lost by downward percolation. Ac- cording to Bulletin 537, the depths of water applied during the growing season to these three treatments were approximately 33, 19, and 17 inches, respectively. PLANT RESPONSES TO IRRIGATION, 1927 TO 1930 During each year, 1927 to 1930, plant responses to the different irri- gation treatments were measured in terms of height of plants, flowering, shedding, number of bolls, yield, lint index, length of staple, and uni- formity. The nature of plant responses in the different treatments at Shafter, 1927 to 1930, can best be brought out: (1) by comparing the responses in treatment 1, in which soil moisture content was maintained above the permanent wilting percentage except before some of the irrigations, with the responses under treatments 2 and 3, in which the plants were allowed to wilt before irrigating; and (2) by comparing the responses under treatment 1 with those under treatments 4 and 5, which were variations of treatment 1. As between treatment 1 and treatments 2 and 3 taken together, using average values for each treatment, the former is found to have produced 6 University of California — Experiment Station each year higher plants, a larger number of bolls per plant, a higher percentage of 5-lock bolls, heavier seed cotton and cottonseed, and higher yields per plant. With the exception of 1927, treatment 1 produced the larger number of flowers per plant. Differences in percentage of shedding were less consistent. 7 Fig. 1.- — -Location of fields on which cotton irrigation investigations were made in 1931. As between treatment 1 and treatments 4 and 5 taken together, dif- ferences in responses were not always in favor of treatment 1. This was the case for height of plants in 1928, 1929, and 1930, and the number of bolls per plant in 1928 and 1929. In yield, however, treatment 1 was higher each year than either treatments 4 or 5. As between treatments 4 and 5 — the former being allowed to wilt before irrigating, up to about July 20, and the latter, after about July 20 — treatment 5 produced higher plants and more bolls per plant, but did not exceed treatment 4 in number of flowers per plant, percentage of shedding, percentage of 5-lock bolls, or yields. 7 No statistical analysis was made of the differences in plant responses in the experiments of 1927 to 1930, and the data needed for such an analysis are no longer available. BuL - 668 1 Cotton Irrigation 7 SUMMARY OF PRINCIPAL EXPERIMENTS AT SHAFTER, 1927 TO 1930 The more favorable plant responses in treatment 1 as compared with treatments 2 and 3 are clearly the most important result of the five years of investigation and experiment at Shaffer. If yields alone are compared, the least favorable result was produced in treatment 3, in which the cotton plants were allowed to wilt at 9 a.m. before irrigating. Exclusive of preirrigation, an average depth of about 33 inches of water was applied annually to the plots in which the soil moisture was maintained above the permanent wilting percentage, except possibly for brief periods before some of the irrigations. The average depth of water used each year by the cotton plants, including that evaporated from the soil and any lost by seepage downward, was about 30 inches. August was the month of maximum use, averaging about 9 inches. INVESTIGATIONS DURING 1931 During 1931 preliminary field observations were made on six farms on the west side of San Joaquin Valley near Firebaugh and Los Banos (fig. 1). In each case areas of from 5 to 10 acres were selected for deter- mining the use of water and gathering data regarding growth and yield, with limited observations on soil cracking and root development. DESCRIPTION OF THE AREAS Soil classifications in the following descriptions are mainly from the recently completed but unpublished soil-survey maps of the Los Banos area. Four areas were irrigated with well water and two with canal water, the difference in the sources of water being thought to be significant from a water-quality standpoint. As a check on this, water used on each field was later analyzed by the Division of Chemistry of the University of California College of Agriculture at Davis ; the results appear in table 1. Fawcett, Well Supply. — The Fawcett farm, served by well water, is located about 6 miles south of Los Banos, as shown in figure 1. The area under special observation was about 5 acres. The soil is Sorrento clay loam, rather sandy, uniform to a depth of at least 6 feet. The land was only fairly well leveled, so that there was variation in irrigation appli- cations and in plant growth. The season 1931 was the second consecu- tive year in cotton, the land having previously been farmed for many years to grain, unirrigated. McClarren, Well Supply. — The McClarren farm, served by well water, 8 University of California — Experiment Station is south of Los Banos. The area under observation was about 5 acres. The soil is clay loam, rather sandy. The land was not very well leveled, so that the stand was spotty. It had been in grain for many years and was in cotton for the second year. Biancucci, Canal Supply. — The Biancucci farm that is served by canal water is in the Firebaugh subdivision. The area under observation was TABLE 1 Chemical Analysis of Irrigation Water Used on Cotton Fields under Observation in 1931* Constituent Hotchkiss well water ( P H 8.6) Fawcett well water (pH 8.6) Biancucci no. 1 well water t (pH8.1) Biancucci no. 2 well water J (pH 8.1) McClarren well water (pH 8.4) Biancucci canal water (pH 8.2) Hotchki; canal water Sodium and potas- sium^ Na + and K + . Calcium, Ca + + Magnesium, Mg + + . . . . Ferric and ferrous iron, Fe + + andFe + + + .... Aluminum, Al + + + . . . . Chloride, CI" Sulfate, S04 — Bicarbonate, HCO3 - . . Carbonate, CO3" - Nitrate and nitrite, N0 3 -andN0 2 - Phosphorus, P Silica, Si0 2 Total solidsl! Carbon dioxide, CO2 p.p.m. 227 40 28 Tracel 1.8 37 430 252 1.1 Trace 32 963 12 p.p.m. 137 55 45 Trace 1.8 17 355 270 2.2 Trace 101 1,137 9 p.p.m. 790 279 Trace 4.2 1,250 1,148 1.7 Trace 31 4,772 30 p.p.m. 565 83 20 Trace 2.0 789 400 174 Trace 56 2,402 5 p.p.m. 268 97 62 Trace 2.0 433 236 135 0.5 33 405 7 p.p.m. * Analyses reported by the late Dr. C. S. Bisson, University of California. t Used on the Biancucci canal area to supplement the canal water. t Used on the Biancucci well area. § Computed as sodium, which was determined by difference. It was assumed that the amount of potassium present was so small that it could be omitted from the computation. 1 Less than 1 part per 1,000,000. || Computed independently, and hence will not agree with the total of the solids given separately. about 7 acres. The soil is Panoche clay loam, which cracks badly. This was the third year in cotton, which was preceded by five years in alfalfa, one year in wheat, and two years in corn, beginning with the corn. The water distribution was uniform. Owing to failure of the canal supply, irrigations subsequent to the third were from a well (Biancucci no. 1 in table 1 ) . Biancucci, Well Supply. — The Biancucci farm that is served by well water is in the Firebaugh subdivision. The area under observation was about 10 acres. The soil is Panoche clay, heaviest of the six areas studied, and relatively uniform to a depth of about 15 feet; it cracks badly. This was the first year in cotton, which was immediately preceded by two years in wheat. Bul. 668] Cotton Irrigation 9 Hotchkiss, Canal Supply. — The Hotchkiss farm that is served by canal water is in the Firebaugh subdivision. The area under observation was about 5 acres. The soil is Panoche clay soil, fairly uniform; it cracks badly and is underlain, especially at the southern end, by a "weak" hardpan 3 to 4 feet below the surface. This was the third year in cotton, with no previous crop on the land. The land was not very well leveled and irrigation distribution therefore not very even, with the heaviest growth in the wettest portions. Hotchkiss, Well Supply. — The Hotchkiss farm that is served by well water is in the Firebaugh subdivision. The area under observation was about 5 acres. The soil is Panoche silty clay, or clay which on drying feels porous and light, and cracks badly ; it is uniform in texture to at least 6 feet in depth. The land had been cropped for many years with cotton and grain in alternate years. AMOUNTS OF WATER APPLIED AND SOLL-MOISTUBE CONDITIONS All fields were planted and irrigated by the growers, without sugges- tions from our observers, since the object of the investigation this season was to determine conditions and results under usual farm practice. Be- cause of unsatisfactory land preparation on some of the areas, irrigation distribution, which was by furrows, was not always uniform. The result was that some of the sampling stations proved to be unadvantageously located, which rendered the samplings valueless. Use of water was determined by measuring the quantities applied, and by soil-moisture tests to ascertain penetration and distribution of the water and transpiration by the cotton plants. Preirrigations were com- pleted before the investigations were started, the quantities of water applied in these prerrigations not being known. Rectangular weirs with water-stage registers were used to measure the water applied sub- sequent to planting. The amounts of water applied in 1931 were as follows: Depthof Depthof water applied, water applied, Farm and date inches Farm and date inches Fawcett (well water) : McClarren (well water) : June 17-18 2.2 June 19-29 5.4 July 10-12 3.4 July 20-23 6.3 July 29-30 3.3 August 12-13 6.9 August 20-23 3.5 August 25 2.1 Total 12.4 Total 20.7 (Continued on next page) 10 University of California — Experiment Station Depth of Depth of water applied, water applied, Farm and date inches Farm and date inches Biancucci (canal water) : Hotchkiss (canal water) : June 3 5.8 July 6 5.7 July 15 7.0 July 25 1.1 August 13-14 4.5 August 2 0.9 August 27-28 4.3 September 13 5.0 Total 21.6 Total 12.7 Biancucci (well water) : Hotchkiss (well water) : June 20-21 4.0 July 8 7.6 July 29 5.8 July 26 5.7 August 30 3.4 August 25 5.6 Total 13.2 Total 18.9 The amounts given are averages for the portions of the farms under observation. Irrigation "runs" were about 1,300 feet long, which was usual practice for the section. In all cases except the Biancucci well area, the land had been irrigated before seeding, but the depth of water applied in preirrigation is not known. Soil-moisture samplings were concentrated at three stations in each area located at about the centers of the upper, middle, and lower thirds along diagonal lines through the areas. At each sampling, samplings were made of each foot to a depth of 6 feet from 8 holes at each station. They preceded and followed each irrigation and were repeated at inter- vening intervals. Standard practice was followed in taking and drying soil samples and in determining moisture equivalents and permanent wilting percent- ages. 8 These soil constants were determined for three of the fields, the results being presented in table 2. Rainfall for the season at Firebaugh and Los Banos was as follows : * Some of the terms relating to irrigation practice that are used in this bulletin are defined by Beckett and Dunshee as follows: Moisture percentage : the weight of water contained in a volume of soil, divided by the oven- dry weight of the soil and multiplied by 100. It is a comparison between the weight of water in a soil and the oven-dry weight of that soil. Moisture equivalent: the amount of water retained in a soil against a centrifugal force 1,000 times gravity. It is expressed as a percentage of the oven-dry weight of the soil, and is a labora- tory method of estimating the amount of water contained in the wetted portion of a soil. Field capacity: the amount of water held in a soil against the force of gravity under unre- stricted drainage. It is the moisture condition found in the wetted portion of the soil after an irrigation or rain and after removal of the free water by drainage. Permanent wilting percentage : the moisture percentage in the soil at which plants wilt and do not recover unless water is added to the soil. It is expressed in percentage of moisture based on the oven-dry weight of soil. (From: Beckett, S. H., and Carroll F. Dunshee. Water requirements of cotton on sandy loam soils in southern San Joaquin Valley. California Agr. Exp. Sta. Bui. 537:1-48. 1932.) Bul. 668] Cotton Irrigation 11 Year and month Rainfall, in inches, at Firebaugh 1930: November 0.68 December 0.00 1931: January 2.41 February 1.00 March 0.22 April May June July August . . . September 0.22 0.43 0.44 0.00 0.03 0.08 October 0.00 Rainfall, in inches, at Los Banos 0.66 0.00 2.14 0.89 .... 0.73 .... 0.21 . . . . 0.60 .... 0.12 . . . . 0.00 0.03 0.01 0.01 Total, 1931 5.51 5.40 TABLE 2 Moisture Equivalents and Permanent Wilting Percentages for Soils of Cotton Areas under Observation in 1931 Farm and soil depth Moisture equivalent, average of 2 trials Permanent wilting per- centage, aver- age of 10 trials* Available depth of water Fawcett (well water) : 1st foot per cent 20.6 18.1 18.7 21.7 21.3 20.1 33 4 29.5 26.4 33.7 28.6 33.1 30.8 31.3 31.8 30.2 26.7 30.0 per cent 9.9 7.7 7.7 7.4 8.9 9.4 8.5 16 4 inches 2d foot 2 2 3d foot 1 8 4th foot 2 5th foot 2 2 6th foot 2 1 Average Biancucci (canal water) : 1st foot 3 2d foot 15 14 18 16 18 16 2 1 4 6 2 5 2 5 3d foot 2 1 4th foot 2 7 5th foot 2 1 6th foot 2.6 Average Hotchkiss (well water) : 1st foot . 22 18.1 18.4 17.1 14.8 18 1 2d foot 2 3 3d foot 2 3 4th foot 5th foot 2 3 6th foot . . . 2 1 Average * In a few cases permanent wilting percentages are the average of less than 10 trials. 12 University op California — Experiment Station Height and general vegetative development of the cotton plants and shedding and boll production were observed on 5 20-foot rows adja- cent to each of the three sampling stations in each of the six areas. Each plant in these rows was included in the detailed observations. Scars left by sheds of squares and bolls and the bolls remaining on the plants were counted, the latter being separated into those which produced cotton and green bolls that did not open. TABLE 3 Summary of Irrigation and Soil-Moisture Conditions, Growth, and Yields in Srx West-Side Cotton Farms, 1931 Farm Irrigation and soil- moisture conditions Average height of plants* Average vegetative branches per plant* Average bolls per plant maturing cotton* Average yields of lint cotton in pounds per acre* inches number number ■pounds Fawcett -per acre (well water) Poor 38.0 2.8 10 9 477 McClarren (well water) Fair 32 9 1.4 9 3 500t Biancucci (canal water) Fair 38 9 3 2 14.8 421 Biancucci (well water) Good 33 5 2.4 17.6 721 % Hotchkiss I (canal water) Poor 37.5 3.8 13.4 449 Hotchkiss (well water) Poor 28.3 2.9 9.7 586§ * In observation rows, t Estimated. t No second picking on account of rain and flood. Probably lost 250 pounds lint per acre, of which about 50 per cent was bolly cotton. § Second picking not included. SUMMARY OF IRRIGATION AND SOIL-MOISTURE CONDITIONS, GROWTH, AND YIELD FOR ALL AREAS, 1931 Table 3 summarizes the data regarding irrigation, soil moisture, growth, and yields for all of the areas. Irrigation and soil-moisture con- ditions are characterized as "good," "fair," and "poor" on the basis of observations made during the season. In none were these conditions con- sidered excellent. Data on growth are limited to height, nodes per plant, and vegetative branches per plant, the figures given for each area being the average of determinations in the three sets of observation rows. These rows were so situated that they should reasonably reflect conditions at the soil-sampling stations. Yield data are limited to bolls per plant that matured and lint cotton per acre. The table is of value chiefly as showing the variation in results obtained by different growers and in different areas. The only comment that seems justified is that the highest yield Bul. 668] Cotton Irrigation 13 of lint cotton came from the one field on which irrigation and soil con- ditions were classed as "good." The study without question demonstrated the urgent need of good land preparation and good irrigation practice, and of a continuously available soil-moisture supply, in those west-side cotton sections of San Joaquin Valley that were covered in the investi- gations. SOIL CRACKING In 1931 studies were made of soil cracking in some of the observation areas and in several other fields. The purpose was to obtain information Fig. 2. — Extensive cracking of adobe soil in a cotton field in the Biancucci well area. Irrigation furrows were wide and most of the surface was wet when irrigating. The rule in the center of the picture is 1 foot long. regarding depth and direction of cracking, and the extent to which it injured cotton plants by breaking roots. The method used was to trench across 3 cotton rows and to pry off cracked portions of soil along the rows. The Biancucci well area furnished an example of the deep and wide cracking that accompanies use of wide irrigation furrows. The soil on this place is particularly heavy. It was irrigated with a larger head of water than is commonly run in furrows, and at least 75 per cent of the soil surface was wet at each irrigation. In many instances it was pos- sible to look down into the surface cracks from 9 to 12 inches. Figure 2 shows this extensive surface cracking. Figure 3 indicates the soil con- dition to a depth of 3 feet across one of the rows. The major cracks here 14 University of California — Experiment Station Fig. 3. — Subsurface condition where soil cracked as shown in figure 2. The major cracks extended to a depth of about 2 feet, but the principal roots did not appear to be affected. Fig. 4. — Limited depth of penetration of cotton roots in heavy soil. Bul. 668 Cotton Irrigation 15 extended down 2 feet; but the principal roots did not appear to be affected, since they commenced their spread just below the hardened, dried-up layer near the surface. The general observation in this and other fields was that the cracks do not penetrate with much width very far downward, and below about 8 inches or a little more, dwindle to very small crevices. Trenching" in some of the other fields showed less, and in some, more and deeper cracking. Horizontal cracking was found to be generally limited to the upper soil layer, and not important. ROOT DEVELOPMENT Root-development studies during 1931 were limited, but were carried far enough to show the difference be- tween the heavy soils underlain with hardpan or a subsoil of fine texture, and the lighter-textured, deep soils. Figure 4 shows the limited depth of penetration in the former, and figure 5 some of the roots taken from the latter. INVESTIGATIONS DURING 1932 AND 1933 About 2.5 acres under the second lift of Firebaugh Canal, in Fresno County, just across the adjacent road from the Biancucci canal area used for studies in 1931 (see fig. 1), was leased for experiments in 1932 and 1933. The soil is classified as Panoche adobe, one of the heavier soils on which cotton is grown on the west side of San Joaquin Valley. It was found to have a moisture equivalent ranging from about 30 to about 35 per cent and a permanent wilting percentage ranging from about 16 to about 18. The objective in 1932 and 1933, as in our previous cotton irrigation studies, was to determine plant responses to different soil moisture con- tents between the moisture equivalent and the permanent wilting per- Fig. 5. — Some of the cotton experi- mental plants taken from the light-tex- tured, deep soils. The vertical rod in the picture is 12 feet long. The cross stick indicates the position of the soil sur- face. 16 University of California — Experiment Station /93Z -31.5'- ! 73 70 ~S~ 5 74 60_ 3_ £7 I0_ 5£ (2. £ 4 JL 7 f£ 6_ 42_ 2_ 39_ 4_ 36_ 3_ m 9_ 30_ ~T JZ 5_ l± /_ 21 ~6_ I8_ 2_ JS_ Vf uL 3_ 9_ 77_ 6_ l/_ 3 \ Supply ditch l 1933 it C) ii 4^ 5l/dd/(j d/tch Location of sampling points Arrangement of sampling notes Dan6ie tines indicate sampiing ptots 2 = Treatment 56 = P/ot number N Piots 37.5' * 49.5' Sapptj/ canat 72 70 6 73 74 75 5 4 3 70 7/ 72 2 I 72 67 68 69 70 6 5 Mi 65 66 4 3 2 67 62 63 7 72 70 58 59 60 6 5 4 55 56 57 3 2 1 52 53 54 12 70 6 49 50 £1 5 4 3 46 47 48 2 1 72 43 44 45 70 6 5 40 47 42 4 3 2 37 38 39 1 w\ 70 34 33 36 6 5 4 31 32 \33_ 3 2 7 28 29 30 72 to 6 25 26 27 5 4 3 22 23 24 r 2 ry 72 5 19 20 2/ 70 6 5 16 77 78 4 3 2 73 14 15 7 12 \/0 70 71 \I2 6 5 4 7 8 9 3 2 1 4 5 6 \ w 8 7\ 1 1 / 2 3] Fig. 6. — Arrangement of cotton plots in 1932 and 1933. Bul. 668] Cotton Irrigation 17 centage. The work during 1932 demonstrated that plant responses were difficult to detect with conclusiveness in the heavy soil of the experimen- tal tract. Nevertheless, repetition of the experimental program during 1933 appeared justified and this was therefore carried through, giving two full seasons of observation, with the essentials of the programs the same in each year. The experimental area was a strip approximately 115 feet by 1,300 feet cut from a field on the farm of D. M. Biancucci. It was divided into 75 plots 31.5 feet by 49.5 feet, arranged in 25 tiers of 3 plots each. A main irrigation supply ditch extended along the west side of the area, and the separate plots were irrigated through small ditches run across the area at the south end of each tier of plots. In each plot there were 9 rows of cotton approximately 49 feet long, spaced 42 inches apart. Water was measured to the plots over a 90-degree v-notch weir set at the intersection of alternate borders and the supply ditch. Distribution was by furrows. Figure 6 shows for each year the layout, the number- ing of the plots, and the distribution of irrigation treatments. Plots with double ruling were sampled for moisture determinations. IRRIGATION TREATMENTS, 1932 AND 1933 The area chosen was preirrigated and thereafter there were five prin- cipal irrigation treatments (nos. 1, 2, 5, 6, and 10), each replicated 9 times, and several supplemental treatments each year, all as described below. Treatment 1. — The aim in treatment 1 was to irrigate when soil mois- ture in the upper 6 feet of soil had been reduced to about the midpoint between moisture equivalent and permanent wilting percentage. Treatment 2. — It was the intention to irrigate the plots in treatment 2 when the cotton plants showed indication of wilt, but, as no drooping of the leaves was detected, they were not irrigated except on August 31 to September 2, 1933, when irrigation was considered by the observer to be needed to keep the cotton plants alive. Treatment 3. — Plots in treatment 3 were irrigated as treatment 1 until late in August, but not irrigated thereafter. Treatment 4. — The first irrigation in treatment 4 was August 12 in 1932, and August 15 in 1933, followed in each year by 1 later irri- gation. Treatment 5. — Treatment 5 followed usual practice of neighboring growers. The plots were irrigated June 29, July 23, August 18 to 22, and September 16, in 1932; and July 13 to 15, August 11, and September 7, in 1933. Treatment 6. — No irrigation was given in treatment 6. The main ob- 18 University of California — Experiment Station jective was to aid in determining the permanent wilting percentage of the soil in the experimental area and the rates of removal of water by the plants from the successive foot depths of soil. Treatments 7 , 8, and 9. — The plots in treatments 7, 8, and 9 were irri- gated at 1-, 2-, and 3-week intervals, respectively. Irrigations began June 17 in 1932 (2 plots each) and July 13 in 1933 (1 plot each). Treatment 10. — The first irrigation of the plots in treatment 10 was July 22 in 1932 (6 plots), and August 8 to 9 in 1933 (9 plots) ; then 3 later irrigations in 1932, and 2 in 1933. Treatment 11 {1932 only). — The first irrigation in treatment 11 (6 plots) was August 5, after plants had practically ceased growing; then 2 later irrigations. Treatment 12 (1933 only). — The first irrigation in treatment 12 (9 plots) was August 8; then 1 later irrigation. METHOD OF SOIL SAMPLING FOR MOISTURE DETERMINATIONS, 1932 AND 1933 The objectives in soil sampling were : (1) to guide the timing of irri- gations necessary to fulfill the requirements of the treatments, and (2) to furnish seasonal soil-moisture histories for use in interpreting plant responses to irrigation. Samples were taken of each foot to a depth of 6 feet. In treatments 1 to 6, in each year, samplings were made in 2 of the 9 plots in each treatment, and in treatments 7, 8, and 9, in 1 plot in each. There were samplings in 2 plots in each year for treatment 10, in 1 plot in treatment 11 in 1932, and in 2 plots in treatment 12 in 1933. The moisture equivalent and permanent wilting percentage were de- termined for each foot to a depth of 6 feet in each of the 12 sampling- plots in treatments 1 to 6. The averages of these determinations and the average computed available moisture contents at field capacity are given in table 4. The moisture shown at some samplings following irrigations was more than the determined moisture equivalent, owing either to the presence of some free water at time of sampling, or to variations in the soil ; also, in numerous instances during the season, the soil moisture was reduced substantially below the determined average permanent wilting percentage. The meteorological records at Firebaugh for 1932 and 1933 are given in table 5. SOIL-MOISTURE RECORD, 1932 AND 1933 As previously indicated, the objective in these investigations has been to determine plant responses to different soil moisture contents between the moisture equivalent and the permanent wilting percentage. The results of the soil-moisture determinations have, therefore, been consid- ered in relation to those constants. Bul. 668] Cotton Irrigation 19 TABLE 4 Average Moisture Equivalents, Permanent Wilting Percentages, and Available Moisture at Field Capacity for Soils in 12 Sampling Plots of Experimental Area, 1932 and 1933 Soil depth Moisture equivalent Permanent wilting percentage Available moisture at field capacity Percentage Depth of water* per cent 35 5 31.7 29.9 29.9 33.1 33.8 32.3 per cent 17.7 16.7 16.2 16.3 17.1 17.3 16.9 per cent 17.9 15.0 13.8 13.6 15.9 16.5 15.5 inches 2.9 2d foot 2.4 3d foot 2.2 4th foot 2.2 2.6 6th foot 2.7 15 Computed on basis of average volume weight of 84 pounds per cubic foot of soil. TABLE 5 Meteorological Eecord at Firebaugh Canal Headquarters, June to October, 1932 and 1933 (Observations at 6 :45 a.m. about 3 miles north and 2 miles west of experimental area) Year and month 1932: June July August September October Total or mean 1933: June July August September October Total or mean Mean temperature Maximum Minimum 87 101 97 88 87 92 Precipi- tation inches Evaporation, standard Weather Bureau pan inches 12.74 13.68 11.53 8.23 5.99 52.17 0.41* 14.53 13.36 8.79 3.34f 40.43t Mean relative humidity per cent 53 52 61 67 .4 44§ 6111 73^ 48*< * Record incomplete. t To October 13 only. t No record except for 2 days. § Twenty -three days. || Twenty-nine days. \ Twenty-five days. " Ten days. 20 University of California — Experiment Station The conditions disclosed are shown graphically in figures 7 to 25. The largest increases and losses were in the first foot. The relative flatness of the curves for the fifth and sixth feet of depth in most of the graphs indi- cates that there was relatively small moisture extraction or accretion below the fourth foot. On an earlier page, it was explained that the objective in treatment 1 was to maintain the soil moisture at least above the midpoint between the moisture equivalent and the permanent wilting percentage. In sum- marizing the moisture changes, it has been found helpful to consider the midpoint as well as the moisture equivalent and the permanent wilt- ing percentage for all treatments. Whenever the moisture is within 3 per cent of the midpoint or within 2 per cent of the permanent wilting percentage, it is described as "close" to it, this being considered justified because of the relatively large spread, in the heavy soil of the experi- mental area, between the moisture equivalent and the permanent wilting percentage, and the variations in these values. It is important to bear in mind that all of the treatments were pre- irrigated before seeding, and it is to be noted that the moisture curves in figures 7 to 25 begin at the first sampling after preirrigation, and that the irrigations referred to in the following discussion were all subsequent to the preirrigations. Soil-moisture determinations in field-plot experiments can never be expected to show more than a close approximation to the soil-moisture conditions in the entire plots, because of soil differences even in ap- parently uniform soils. Likewise, average determinations of moisture equivalent and permanent wilting percentage, even when made from a satisfactory number of soil samples, may not apply more than approxi- mately to particular sampling locations. For one or both of these reasons, figures 7 to 25 show some inconsistencies, such as the soil-moisture curve extending substantially above the moisture equivalent or below the permanent wilting percentage. In the following discussion, these incon- sistencies are ignored. It is deemed sufficient to consider the soil-moisture curves, and the portions of the lines representing the average moisture equivalents and permanent wilting percentages, as approximately rather than as precisely indicating the conditions in all of the plots of the vari- ous treatments. In some instances soil-moisture samplings were not suf- ficiently frequent to give assurance that the soil moisture did not reach the permanent wilting percentage earlier than indicated by the curves. Soil Moisture, Treatment 1, 1932 and 1933.— In 1932 (fig. 7), and to a greater extent in 1933 (fig. 8) , the soil moisture in treatment 1 dropped below the midpoint between the moisture equivalent and the permanent wilting percentage in the first foot during portions of the summer Bul. 668] Cotton Irrigation 21 so 40 30 20 10 50 40 30 20 /O SO 40 30 20 fO -> so r« n 1 1 1 „Pbt (4 ! 1 1 / si foot ^ 5^ ^ \ t ? : *%> M.ts- — =d P/ot 45? =^=. ^i j ^ i 1 H 1 i i 2/?d foot — [ u \. — ^ •°--. 3= I ^ 30 fc=i 1 1 1 3rd foot =5 £Q> .-s-o- — -9 •o— — ■"~^» — 40 Jt/7 'foot » m~ *t* — 1 -•— ZZZ 20 10 SO 40 30 20 — - 10 SO 40 30 20 10 — 5 t/i foot — o- — — . — — , — ro. =f •o— - — — C -J— o '"'•-♦ — h — 6 t/? foot — 0- — — -_. ^ "O :T~°- — — -0 -o— — — o /0 20 31 10 20 30 JO 20 3/ /0 20 31 10 20 30 /O 20 31 /0 20 30 Moj/ Jane Jofy Aug. Sept. Oct. Nov. Fig. 7. — Soil-moisture record, treatment 1, 1932. 22 University of California — Experiment Station 1st foot — On. '— K. r- 1 1 Mts: — ~P/o£ J0>"*-3 N ^ | "~~"~°'^ i/ — — z PW.Ps i 1 ^ i — 1 I 50 40 30 20 /O 50 40 30 20 to 30 40 30 20 10 50 40 30 20 =- 2 arf foot ^! p : *=^— - 3rd foot :— . ho ^ ^ i=«a _ ■■— < > — J/ftVoo't — ■^Sl" — » — — to 50 40 30 20 — « to 50 40 30 20 to 1 1 1 StA foot •— ' n-o-^ r ~* =3 — g — -< > 1 1 1 — — ■~ "~ -o-c j=S -o- __» .— .. —— » * — ~ #? 3t tO 20 30 tO 20 31 tO 20 3t tO 20 30 tO 20 31 tO 20 3/ Moy June Juft/ At/g- Sept. Oct. Nov. Fig. 8. — Soil-moisture record, treatment 1, 1933. Bul. 668] Cotton Irrigation 23 months ; it also dropped below the midpoint in the first foot about mid- October. The permanent wilting percentage was reached for only brief periods. 9 In the second foot the moisture curves lie below the midpoint in both years after July, and are close to or reach the permanent wilting percentage after July in 1932, and by August 10 in 1933. During the last two months of the season, the curves for the second foot are so flat as to indicate that the moisture was depleted about to the permanent wilting percentage. In the third foot the curves lie along or above the midpoint throughout the 1932 and until the middle of August in 1933, after which they drop below the midpoint but remain above the perma- nent wilting percentage. For the fourth, fifth, and sixth feet, the slopes of the curves are flat or slightly downward after June in 1932 and throughout the season in 1933, which indicates that there was very little, if any, moisture extraction at those depths. The permanent wilting per- centage was not reached at any time during the season, and irrigations did not greatly change the curves, which indicates that water penetra- tion was limited to the first 2 feet of soil, with the possible exception of the first irrigation. Soil Moisture, Treatment 2, 1932 and 1933.- — The moisture was re- duced to the permanent wilting percentage in the first and second feet of treatment 2 by September 1 in both 1932 and 1933 (figs. 9 and 10). It remained at the permanent wilting percentage in the first and second feet for the remainder of the season in 1932 but was raised above the permanent wilting percentage in 1933 by the irrigation about Septem- ber 1. The curves for the third foot do not reach the indicated permanent wilting percentage in either year, although they closely approach it dur- ing the last month in 1932. They also closely approach the permanent wilting percentage in the fourth foot in October, 1932, if not before. Apparently very little moisture was extracted from the fifth and sixth feet, which indicates limited root penetration to those depths. Soil Moisture, Treatment 3, 1932 and 1933. — In 1932 the soil moisture in treatment 3 was close to or at the permanent wilting percentage after August to a depth of 4 feet (fig. 11). In 1933 (fig. 12) the permanent wilting percentage was reached in the upper 2 feet before August 14 ; after the irrigation on that date, it was not reached again in the upper foot until late in the season, but was soon reached again or closely ap- proached in the second foot. A moisture sample was taken shortly after the irrigation August 14, and then only one more that season, so that the permanent wilting percentage may have been reached earlier than indi- 9 Since the upper portion of the top foot of soil is depleted in moisture by surface evaporation, the average soil moisture in the first foot will presumably be less than in the 6 or 8 inches of depth above the second foot. This probable difference is dis- regarded in the soil-moisture analysis. 24 University of California — Experiment Station 30 40 30 20 10 SO 40 30 20 10 30 40 h^ SO 1 1 1 /si /oot —+ Alt's- — o-« ■—- — __ __ sP/oi /9 *Ptf/PA P/nt sm T-o*. *-- ^ ^ — .— -o 1 1 1 2/?d /boi =zz — •^^ — — " " ~t ■ 1 1 1 3rd /bo i — -- " E ' 1 1 1 4 ifi foot — ? A — 30 40 30 20 to 30 40 30 20 /O 30 40 30 20 -—- 1 1 1 5 ift foos — 1 — — o— to 1 1 J 6 1 /? /bos — *-0- ——. . 8 /O 20 31 /O 20 30 /0 20 3/ /O 20 3/ 20 30 tO 20 3/ to 20 30 May Jt//?e Ju/y Aug. Sept. Oci. Fig. 9. — Soil-moisture record, treatment 2, 1932. Nov. Bul. 668] Cotton Irrigation 25 ! so 40 30 20 /O SO 40 SO 20 10 50 40 30 2/? — 1 1 1 3 re/ foot — o— — V "^ r— — N i — 4 a Voo/ -©- — •*' I — = 20 — - /O SO 40 SO 20 to SO 40 30 20 to 1 1 1 5 £/? foc?£ — — — L .:— «= 6 //? /bos — b '" f"** fO 20 31 tO 20 30 tO 20 3/ tO 20 3t tO 20 30 tO 20 3t tO 20 30 Mat/ June Job/ At/g. Sept. Oct. Nov. Fig. 10. — Soil-moisture record, treatment 2, 1933. 26 University of California — Experiment Station so 40 30 20 /O SO 40 30 30 fO 30 40 ^ 30 \30 I i I 1st foot N^ MF'.i — ssm — ; ^***£fM& *P.h/.t P/t o— — 75 OO ' 1 2 r?d foot — tt — ^ ^ s ti-o-^fc^ " — B = I ! /o 30 40 30 30 10 30 40 30 30 10 30 40 30 30 fO 1 1 1 1 3rd foot o— • ssj ^, ^ ****o 1 1 1 4 1/? foot — — o- — — - -d° — »_ (3-o- 1 5 to 1 foot -cf<> o-o- • — 6 t/j foot -oo y* __ — — ^— — fO 20 31 /0 20 30 /0 20 3f /0 20 31 /0 20 30 /0 20 3/ /O 20 30 May Jane Ja/y Aug. Sept. Oct. No/. Fig. 11. — Soil-moisture record, treatment 3, 1932. Bul. 668] Cotton Irrigation 27 50 40 30 20 10 50 40 30 ZO JO 50 40 30 20 10 50 1 1 ■ 1 /si foot __ '8' \ f** 1 1 M.Ei ,—— — P/ot4 -—3 > s Ni jP/oi 14 1 SI "■■■ 1 1 1 2 at/ /boi . ■***» — _ -<5-X3 <; ^ — ~ So- — ^ 1 I & 40 1 1 1 J rd /boi .^'O ^.^_ *o — — - — s 30 20 fO /i/? foot £m **=: =*-- — — 5U 40 30 20 in 1 1 1 Si/7 fat o— « -6-* "~ 50 40 30 1 1 1 6 1/? fooi 20 in ^*=* — — fO 20 31 /0 20 30 /O 20 31 IO 20 31 /O 20 30 fO 20 31 (O 20 30 May June Ju/y Aug. Sept. Oct. Nov. Fig. 12. — Soil-moisture record, treatment 3, 1933. 28 University of California — Experiment Station cated by the curve. The permanent wilting percentage was reached or closely approached in the third and fourth feet by the end of August in 1932, although it was not reached or closely approached in those depths at any time during 1933. For the fifth and sixth feet, the moisture curves are well above the permanent wilting percentage and relatively flat in both years. Soil Moisture, Treatment 4. — Soil moisture in treatment 4 was de- pleted to, or close to, the permanent wilting percentage before the first irrigation in the upper 2 feet in both 1932 and 1933 (figs. 13 and 14). It was also depleted to, or close to, the permanent wilting percentage in the upper foot before the second irrigation. It was not much above the permanent wilting percentage after the second irrigation in the sec- ond foot in either year. In the fourth foot the permanent wilting per- centage was closely approached from September on in 1933 ; otherwise, the moisture was above or close to the midpoint between the moisture equivalent and the permanent wilting percentage in the lower 3 feet throughout the season. Soil Moisture, Treatment 5. — For purpose of comparison, tracings of the curves for treatment 5 (figs. 15 and 16) were superimposed on the curves for treatment 1, which was irrigated substantially the same as treatment 5. Differences in the curves were so small that the comments given above for treatment 1 apply to treatment 5. Soil Moisture, Treatment 6. — The curves for treatment 6 (figs. 17 and 18) were compared with those for the similar treatment 2 in the manner described for treatments 1 and 5. Here, also, the curves for the two treat- ments lie very close together, the only material differences being due to the irrigation of treatment 2 in 1933 about September 1. Figure 17 indi- cates that very little if any moisture was extracted from the fifth and sixth feet in 1932, although figure 18 shows some extraction from the fifth foot in 1933. Soil Moisture, Treatments 7, 8, and 9. — Treatments 7, 8, and 9, irri- gated at 1-, 2-, and 3-week intervals, respectively, had, of course, high moisture contents throughout the season at all depths (figs. 19, 20, and 21). Difficulties of sampling the wet soil prevented obtaining a satisfac- tory record in 1933. Because of the frequent irrigations, and the heavi- ness of the soil, the soil of the upper foot in treatments 7 and 8 was too wet to warrant attempting to compare the moisture content with the moisture equivalent and the permanent wilting percentage. In the second foot the average was above the midpoint in treatments 7 and 8, and only slightly below it in treatment 9. In the third to sixth feet the curves gen- erally average above the midpoint, treatment 8 being higher than treat- ments 7 and 9. Except for the early samplings in the fourth and fifth feet Bul. 668] Cotton Irrigation 29 i I 50 40 30 BO 10 50 40 30 20 10 50 40 30 20 /O 50 40 1 1 1 /si foot -- P/ot /2 r-*. °^. ^•^ M.t's- — — j ^ % J ^ ^i M IPs nut, w 1 1 •..4 s< — I I 1 2 /?c/ foot o • — — r__^ P — - — ^£;;j =->«- i i i 3 rd foot «« s-c = ? — — /o 50 40 30 20 fO 1 1 1 4 i/? foot PS»: *"~t — — so —* 20 1 1 1 S t/? foot ft x± — — 6U 40 30 6 M fooi / *: -o — =8= - =3- 20 10 — tO 20 3/ fO 20 30 /O 20 3/ /O 20 3/ fO 20 30 fO 20 31 /O 20 30 Mat/ Jt/oe Jt/ft/ At/g. Sept. Oct. Nov. Fig. 13. — Soil-moisture record, treatment 4, 1932. 30 University of California — Experiment Station so 40 30 20 (0 SO 40 30 20 10 50 40 30 rs 1 1 /st foot Plot 33 k. \ Mrs'. — — p/otTr \ \ \ \ ° \p.W.Hs 1 —6 S rf — i i i £/?-r" j—' \ 20 ; to 1 1 1 3 re/ /bo/ - — — SO 40 30 20 /O SO 40 30 20 to 1 1 1 4//t (oot x — o. — • n. — i— 1 1 1 5 £/j Poos' — — — — 1_ so 40 30 I I 1 £//? Zoos' — "*" > 20 tn fO 20 3t tO 20 30 fO 20 3/ JO 20 3/ /O 20 30 /O 20 3/ /O 20 3/ Mat/ Jane July Aug Sept. Oct No/. Fig. 14. — Soil-moisture record, treatment 4, 1933. Bul. Cotton Irrigation 31 so 40 30 20 /O 30 40 30 20 10 SO 40 \ i i i / st foot P/ottt^ | > V V" r- Mfk- o— ~]v — I— J N N v, > v s ^ °-^» — -MP's i Plot 64 i i % 'sS ^--o : — 1 1 1 2 /?- X ; --., fPht Id Mi's- — — P/ot46~'* ^i. X ^ ^J zM IPs >1 ■^ """■"* ^N i 2 — - 2 rtd foot ___ !***« ^ __ "*"*x ^ ^ *---:*-^ r- ! -* — 5 T^ -=8- '— " < > _ — *** is to ^ fO 1 1 1 3 rd foot — — ^ ^F ^ 30 \ 40 % 4 1/? foot 1— .. J^2* Nfc — — 20 /O 30 40 30 20 10 30 40 30 20 /O 1 1 1 5 tn foot *.< — 6— - ::..:$ — — 1 1 1 6 tn foot k-< — 1 "•-— i £--=.-. — o. ■ — — /O 20 31 /0 20 30 /O 20 3/ fO 20 31 /O 20 30 /O 20 31 /O 20 30 May June July Aug Sept. Oct. Nov. Fig. 16. — Soil-moisture record, treatment 5, 1933. Bul. 668] Cotton Irrigation 33 I 50 40 30 20 10 £0 40 30 20 10 50 40 30 20 10 SO 40 30 20 10 50 40 30 20 /O 50 40 30 20 /O 1 1 1 /si foot — c ' Plot Z3~ M.t.sl — -—- r V"" J f P's I Pi 'otS8 ~-o— —■ — fcn i i i / /? i~~* ri- -A - 1 _L _ __ 6 -4 < '\ : t 6 i'T r 7 o 6 "3~ i--« -O- — ... ^= " — • -° — - "" /o i i i 3rd foot r~* 4- —J 1 1 "1 . 1 i — 6 o:^ 1 < 6 . 1 . 6 d 1 cJ 6"r -<=»— i.~~. «m-o Jtf r-O A -ft \ i •> -... ••j"" j/932 M.r.l - — , DM/ D* /933 y *> f N r^ -- -... \r-if ' ■ — — -6 -j— 4i ^t^ -•i ' — -c ^ 1 ! 1 3rd foot . f -*v ■"r'sl ^ ■t-j- L-— . •6 -♦-'-■*. i-f- -*•-.. — - — -— - c 4th foot -^ r*- * kT° i —j. i ^^i i I — — o /o JO 40 30 20 /O 50 40 30 20 to 50 40 30 20 = = 1 1 1 5 M foot — ..... r- ^— , — c f mmmm .K* i s t ■*■ r— - ■•— /o • /S33 ! 1 1 6 1/? foot "~ — ' o /9. 33 — "~ c «- i r-o >•-•— , X r ~" ■**•■ — -O 10 20 31 /0 20 30 10 20 3/ /O 20 31 /O 20 30 fO 20 31 /O 20 30 Moj/ June Ju/y Ai/g. Sept. Oct. Nov. Fig. 20.— Soil -moisture record, treatment 8, 1932 and 1933. Bul. 668] Cotton Irrigation 37 k 1 I 1 /sJ foot j\ V_ ! M.E.S — 1933^ ^ " V 1 M- > ~?93Z>^ 1 /?/ V.KS ""*""" -— - — o 50 40 30 20 /O 50 40 30 20=^^ 40 i i i 2 /?d foot — i< =^ •— o -V, ^ t ;-. V" *«►:::: ~* — -G $ 30 20 fO 1 1 1 3rd foot k L__ F--V ^ — — <3 b 50 40 30 BO (O 50 40 30 20 /O 50 40 30 20 /O 1 1 1 r -T'^ r~- t> r~ ■•«- u. s _ [.,'" ' 1 1 1 *f//£ foot - % *.__. ■L f — ' • / SO % % % 30 20 10 SO 40 30 20 IO SO 40 30 20 to \ i l t^t foot |\ t i i \ ^M49 HEs- i S ! "^x • PWPs * Xj > i / m 54 1 1 ^/ktYwt ^w. m m ui N Sj^^- ^ — " ^^ — ^1F *• 1 1 1 3rd foot ^ 0°^ s *v >*T -o. •»»•. " — < i •O m 1 1 1 4M foot ***■ •^ ^ **'X> ^. ^3 1 1 1 SMfooi t- O •••.^^ 1 1 ! 6 1/? foot ^fe ^^."8 =1 — — « /0 A? J/ / 1 -O— «. D — < — i i i 4 t/? foot __ _ __ _ "^ P^ — ■■ — —*> \ - ! 1 1 3 tb foot — — 0— •*^« t .=-L_ 7 1 1 1 6 tf? foot i ,--•" -^ — 6 -I-. 8T~ 7=* /O 20 3/ 10 20 30 /O 20 31 10 20 3/ tO 20 30 tO 20 31 fO 20 30 Mag June July Aug Sept Oct. Nov. Fig. 23. — Soil-moisture record, treatment 10, 1933. Bul. 668] Cotton Irrigation 41 I so 40 30 20 /O SO 40 30 20 10 SO 40 30 20 10 SO 40 30 20 to SO 40 30 20 /O SO 40 30 20 i i— i — 1 M.E. ; • i «|

' Mf's. j . j — -PW.P.5 1 — ,^j 1 — I I 1 2/?d foot ff *3K — ^■4 ^ ' \ i 30 40 30 20 10 50 40 30 20 10 30 40 30 20 10 50 40 30 ZO 10 50 40 30 20**=-- 3 fd 'foot «= — — 1 — — — 4tf? foot — — i * 10 50 40 30 20 /O 1 1 1 3 tf? foot — 1 — * "*v ^.1 — — i i i 6 tf? foot — —i /O 20 31 /O 20 30 /O 20 3/ 10 20 3/ /O 20 30 /O 20 3/ /O 20 30 May June Ju/y Aug. Sept. Oct- Noi/. Fig. 25. — Soil-moisture record, treatment 12, 1933. Bul. 668] Cotton Irrigation 43 figures 26 and 27. The soil-moisture records for treatments 7, 8, and 9, and for treatment 11 in 1932, were not sufficiently complete to be in- cluded in the analysis. Treatments 3 and 4 were not included because they departed too far from normal practice to be comparable with the other treatments. Loss of waler during mo/iifo, wcftes Treat. P/ot Mat/ . ft/ffe \ /{//(/ Aug. . fept. Oct. / Vor. a fai / S /o a 6 74 45 64 35 49 34 79 63 23 56 9 02 0-4 0-8 0-8 08 09 9 0-7 0.8 t-2 0-3 OJ /./ /■O 70 7.1 /./ 0-9 70 66 56 84 73 35 38 7-6 1-5 73 73 7.1 7.7 9.5 78 76 8 2 3 3 2-9 31 22 3-6 43 30 3-6 49 45 08 70 0-9 0-7 /■2 7-6 25 20 3 3 7-6 0-7 OJ 71 0-5 t 06 i f.O c 70 c 75 t 08 l -05 02 0-2 7 '// 0.3 "55 >36 '07 72 79 7.8 7 8 ^-n4 ^ ^ i 5-35 ^^70-49 *^-~ £2 7-/4 ^70-54 /' ~-~ ^ ^ -" / ■45 jr ^ // V, *? / / // / I 1 7 / / // ''/ V (? 63 '/J /" /s. - : - ---* " f 2-t •■~6-23 ^6-58 '9 J? t %/, ——~ 7 J^ , ^ y ^ I-// X X '/^ \/£ y ^ 1^ 30 28 26 24 22 ft ft "S 20 1 o /fl $ I f6 \s 74 ! I • r ■ ^ i&£*&Ci ^1k v*S^ "ZSte # "^ ~" >W , T *^W^Smm£b*&> l:.w-;i?*' >^,.. Fig. 30. — Cotton plants in treatment 5 on August 5, 1932. They had been irrigated twice, on June 29 and on July 23. Bul. 668] Cotton Irrigation 49 Fig. 31. — Cotton plants in treatment 11 on August 25, 1932. They had received no early irrigations but had received water August 5, 11, and 18. ifr :"dj^ ►1 fi'Tr'Y'':^ Fig. 32. — Cotton plants in treatment 10 on August 25, 1932. They had been irrigated July 22 and August 17 and 20. 50 University op California — Experiment Station other treatments, and made the most total length growth ; and that the curves for treatments 1, 3, and 10 started to flatten earlier than those for treatments 7, 8, 9, and 5. Although of perhaps questionable signifi- cance, 11 the relative positions of the length-growth curves for treatments 1 to 6 are similar to the relative positions of the blossoming curves for the same treatments shown in figure 33. Additional evidence of the effect of the different irrigation treatments on plant growth in 1932 is presented by figures 29 to 32. Figures 29 and 30 show plants in treatments 2 and 5, photographed August 5, 1932. Treatment 2 was unirrigated ; treatment 5 had been irrigated on June 29 and July 23, and the plants were much larger. The plants in treat- ment 2 increased very little in height after the date of the picture. Figures 31 and 32 show plants in treatments 10 and 11, photographed August 25. Treatment 11 had received no early irrigations but had been irrigated August 5, 11, and 18. At the date of the picture, the plants had made no visible response to the irrigations. Treatment 10, on the other hand, had been irrigated earlier, on July 22 ; and also on August 17 and 20 ; and the cotton plants were substantially higher than those in treat- ment 11. BLOSSOM COUNT, 1932 Figure 33 shows the average daily blossom count per plant on approxi- mately 50 plants in each sampling plot in treatments 1 to 6 from July 20 to September 10, 1932. Blossoms had been appearing occasionally about a week before the count began, but blossoming was practically over when the count ended. Taken alone, this chart appears to show some significant relations between irrigation treatments and blossoming. The first blossoming peak occurred simultaneously for all treatments, and prior to the first peak, the number of blossoms appearing daily did not show much variation. After this peak, however, there were differences which may have some meaning. The low group includes unirrigated treatments 2 and 6 and treatment 4, which was not irrigated early in the season. The middle group includes treatment 1, which was irrigated in an attempt to keep the moisture percentage above the midpoint between the moisture equiva- lent and the permanent wilting percentage, but in which the permanent wilting percentage was approached or reached in numerous instances in the upper depths, as shown in figure 8 ; and treatment 3, which was irri- gated only in the first part of the season. The highest curve is for treat- ment 5, which received its first 2 irrigations 6 days before treatment 1. If any conclusion is to be drawn for the blossom curves in this chart, it 11 When analyzing the results of the experiments described herein, differences are considered to be significant with odds of 30 to 1. For these odds, the differences must be about 3.2 times their probable errors. AVERAGE BLOSSOMS PER PLANT PER DAY p r* . (V - O d d d o d o d y rw ^* / / \ ^ *~"*- -4 /T'.'.' A -—*' •"*"" y\ ^.- > ^*< > < \ \ b z /"""" d -• -—■ 'f Y*^ / / z d z ,. -v^ r ' "'• ' -• —- \ \ d Z ♦ ^£ Z / V ""*" *-i ': r-'"Tf /-'- >• ~r~. --*-. , «: ;-•-■ :.:: ZZ / /^ : -<"* r~~ c ^ s. \ v "">* -^ ~~L-*- ,«_• — -->-* ^ >;. — •; 2^-' ^__^ =*S*B ~"— v ^7A ■A <: ^ ' V " ■"" ;/^ ^7 ■ ^ — <\j ro n 'V <*.^ 1 < ^ o. a> too PI AVQ b3d J.NV"ld d3d SWOSSOIQ 3DVU3AV 52 University of California — Experiment Station is that, beginning with the first peak about August 8, the unirrigated treatments 2 and 6, and the treatment which was not irrigated early in the season (no. 4) , reached midsummer with plants too small to maintain a blossoming rate comparable with the rates in the treatments that were irrigated early in the season. Unfortunately, from the standpoint of conclusiveness, the apparent blossom-count relations developed in 1932 were not possible of deter- mination in 1933. A complete chart for 1933 could not be made because observations were omitted for 3 or 4 days in four week-end periods. The data that are available are not very consistent with those of 1932. In 1933, the count for unirrigated treatment 6 was the highest of all treat- ments, and the counts for treatments 1 and 5 ran close together during the period of principal blossoming. At certain times when the count for unirrigated treatment 6 was high, that for unirrigated treatment 2 was low. There were also other discrepancies as compared with the count in 1932. A blossom count for a single season is obviously inconclusive. YIELDS OF COTTON, 1932 AND 1933 Yields for 1932 and 1933 are given in table 8 by plots and treatments. Since there was some variation in the number of plants in the different treatments, the number in each is given. In both 1932 and 1933, the yields in treatments 1 and 5, in which the soil moisture was generally maintained above the permanent wilting percentage, were significantly higher than in treatments 2 and 6, in which, except for treatment 2 in 1933, the soil moisture reached or closely approached the permanent wilting percentage in at least the upper 3 feet by the end of August. Irrigation of treatment 2 about September 1 in 1933, considered by the observer necessary to keep the cotton plants alive, raised the soil moisture content well above the permanent wilting percentage in the upper foot, with some increase in the moisture content in the second foot. This probably accounts for the increased yield in this treatment in 1933 over that in treatment 6, which was not irrigated. In 1932, the yield in treatment 10, in which the soil moisture dropped to the permanent wilting percentage in the upper foot before irrigation and was close to or at the permanent wilting percentage after early August in the second foot, was significantly higher than in treatment 1, but not significantly higher than in treatment 5. The yield in treatment 11, which received its first irrigation August 5, was the same as in treatment 6 and about the same as in treatment 2, neither of which was irrigated. The yield in the "field" was significantly higher than in treatment 1, but not significantly higher than in treatment 5. Bul. 668J Cotton Irrigation 53 TABLE 8 Yield of Seed Cotton by Plots and Treatments, and Average Yields per Plant, 1932 and 1933 1932 1933 Plot no. Plants Yield t Plot no. Plants Yieldt 7 14 21 31 38 45 55 1 62 69 Total Av. per plot. . Av. per plant 15 19 32 39 43 56 63 67 Tot Av. Av. r 9... al per per plot.. . . plant. . 13 20 33 37 44... 57 61 68 Tot Av Av ql per per plot. .. . plant. . number 205 221 190 238 224 227 ...% 255 239 1,799 252 234 204 249 188 226 230 234 244 2,061 260 211 227 185 215 247 222 249 248 2,064 pounds 39.5 39.5 33 51.5 46.0 46.5 25. 5J 48.0 39.5 343.5 19±0.004 18.5 18.0 20.0 22.0 21.0 23 21.5 19.0 23.5 20.7 09±0.009 28.0 36.0 39.0 23.5 49.5 34.5 32 30.5 32.5 305.5 33.9 0.15±0 008 10 20 30 34 44 54 58 1 68 Total Av. per plot.. Av. per plant 5 15 19 29 39 43 53 63 67 Total Av. per plot. . Av. per plant 4 14 24 28 38 48 52 62 72 Total Av. per plot. . Av. per plant number 160 179 179 190 212 211 212 227 % 162 1,505 173 240 152 192 222 203 184 259 1,824 217 183 184 184 218 187 210 1,758 pounds 46.0 59 59.5 46 64.5 54.5 49 5 36. 5X 43.5 422.5 52.8 0.28±0.009 39 35.5 34 43.0 41.0 44 42 32 42.0 352.5 39.2 0.19±0.009 51.0 40 39.0 39.5 43 41.0 36.5 41 45.0 376.0 41.8 0.21±0 009 (For footnotes, see end of table, page 56.) 54 University of California — Experiment Station TABLE 8 (continued) Treat- ment* 1932 1933 Plot no. Av. per plant. Plants number 239 207 224 213 228 236 228 267 183 2,025 222 226 189 198 241 256 244 228 217 2,021 182 203 200 234 217 205 233 243 1,885 136t 252 Yield! Plot no. pounds 26.0 29.5 20 5 31.0 40.5 34 29.5 34 29.0 274.0 30.4 0.14±0.007 50 49.5 40 5 42.5 54.0 50.0 49.0 56 33.5 425.0 47.2 0.21±0.006 26.0 15.5 30.0 22.0 25.0 25.5 18.5 19.0 22.0 203.5 22.6 0.11±0.006 40.01 42.5 9 13 23 33 37 47 57 61 71 Total Av. per plot. . Av. per plant 8 18 22 32 42 46 56 66 70 Total Av. per plot.. Av. per plant 7 17 27 31 41 51 55 65 75t Total Av. per plot.. Av. per plant 0.17 Av. per plant. Plants number 190 190 218 182 238 189 230 251 169 1,857 206 225 217 165 215 159 206 167 190 240 204 198 207 243 183 202J 1,632 Yield f pounds 53 43.0 48 50.0 45.5 48 51 5 44 61.0 444.0 49.3 0.24±0.012 62.0 53.0 46.5 50.5 52.5 42.5 49.5 39.0 39 434.5 48.3 0.24±0.010 29.5 25.0 23.5 28.0 31.0 29.0 27.0 28.0 36. 5t 221 27.6 0.14±0.006 55.0 0.32 Bul. 668] Cotton Irrigation TABLE 8 (continued) 55 Treat- 1932 1933 ment* Plot no. Plai its Yield t Plot no. Plants Yield t [26 28 num 2 2 ber pounc 31 50. C 28 42. ( Is > 2... number 156 pounds 52.5 1 Total 4 59 92. C 46. ( 8 Av. per plant. . 0.2( Av. per plant. . 0.34 [25 : 1 98 45. ( ► 1 158 57.0 30 Total 1 3 96 42. 1 94 87. t 9 < Av. per plot. . . . Av. per plant. . 43 3 2$ Av. per plant. . 0.36 [49 2 68 61. 1 12 209 63.5 50 2 51 58. ( 1 16 210 42 5 51 2 50 50. 1 > 26 135 37.0 52 2 2 11 60. ( 20 58. ( 1 36 214 191 57.5 53 1 40 43.0 54 2 59 53. J » 50 60 204 231 240 185 49.5 55.0 10 , 64 50 5 74 63.5 Total 1,4 59 341.; 56. 1 > Total 1,819 462.0 Av. per plot. . . . Av. per plot. . . . 51.3 Av. per plant. . 0.23=hO 0085 Av. per plant. . 0.25±0.010 U 5 2 61 28.{ 1 11 199 189 184 184 222 200 185 100§ 146 67.5 21 41.5 25 51.5 35 50 45 49 55.5 44.0 59 42.0 12 . 69§ 30. 5§ 73 38.5 Total 1,509 390.5 A v. per plot.. . . 48.8 • A v. per plant. . 0.26±0.010 56 University of California — Experiment Station TABLE 8 (concluded) Treat- 1932 1933 ment* Plot no. Plants Yieldt Plot no. Plants Yieldt ' 9-F 21-F number 225 199 231 224 230 235 1,344 pounds 38.5 37 51.5 48.0 56.0 42.5 273.5 45.6 0.20±0.008 number pounds 33-F Field 45-F 57-F i 69-F Total Av. per plot.. . . Av. per plant. . experi- mental plots * For description of treatments see p. 17. For quantities of water applied, see table 6. t Probable errors are given for the averages per plant. t Left out of totals and average because of excessive verticillium. § Flooded by mistake, September 11. Left out of totals and average. In 1933 the yield in treatment 1 was not significantly higher than in treatment 12, which received only 2 irrigations, beginning August 8. The yield in treatment 5, in 1933, was slightly lower than in treatment 12. In 1932 the yields in treatments 1 and 5 were significantly higher than in treatments 3 and 4, irrigated first on July 8 and on August 12, re- spectively. In 1933, however, the difference between treatments 1 and 3 was significant, but that between treatments 1 and 4 was not. The significance of the differences between the yields in the frequently irrigated plots 7, 8, and 9 could not be determined because there were only 2 plots each in these treatments in 1932 and only 1 plot each in 1933. Irrigating as frequently as every week was extreme practice for the clay soil on which the experiments of 1932 and 1933 were conducted. INVESTIGATIONS DURING 1934 About 2 acres, triangular in form, on the ranch of Robert Stevenson, in Sec. 21, T. 13 S., R. 14 E., were used in the experiments in 1934 (fig. 34) . This tract adjoins the third-lift canal of Firebaugh Canal Company. The soil is classified as Panoche clay loam, grayish in color. Initial field examination indicated that it could be considered fairly uniform to a depth of about 3 feet, becoming stratified below about 3 feet. The gen- eral uniformity was confirmed later by preliminary moisture-equivalent and permanent-wilting-percentage determinations completed July 10. These showed, for the upper 6 feet of soil, average moisture equivalents of from 22.7 to 27.8 per cent and average permanent wilting percent- ages of from 12.0 to 15.1 per cent. For the upper 3 feet of soil, the range Bul. Cotton Irrigation 57 in average moisture equivalent was from 24.6 to 27.8 per cent and that of the average permanent wilting percentages was from 12.6 to 14.3 per cent. More complete moisture-equivalent and permanent-wilting-per- centage determinations were made after the close of the season, on com- posite samples from all of the soil samples taken during the season for moisture determinations. The latter are used in presenting the results of the experiments during 1934. Fig. 34. — Arrangement of cotton plots in 1934. IRRIGATION TREATMENTS, 1934 Three irrigation treatments were outlined and followed as closely as possible, with the preliminary moisture-equivalent and permanent- wilt- ing-percentage determinations as a guide. In treatment 1, the aim was to irrigate when any single soil-moisture determination below a depth of 0.5 foot showed a moisture percentage drop below the moisture equivalent of as much as one third of the differ- ence between the moisture equivalent and the permanent wilting per- centage, that is, as soon as one third of the available soil moisture at field capacity had been used. In treatment 2, the aim was to allow the soil moisture to closely ap- 58 University of California — Experiment Station proach the permanent wilting percentage, but not to reach it in more than one soil sample from any 1-foot depth from anj^ plot. In treatment 3, the aim was to allow the cotton plants to show severe distress in at least two thirds of the plots of the treatment before irri- gating. REPLICATION AND SAMPLING PLOTS, 1934 Each treatment was replicated seventeen times. Soil sampling was confined to 3 double plots for each treatment, and measurements of yields and other plant responses were made only on the 11 plots in each treat- ment in which soil-moisture samples were not taken. This separation of the soil-moisture sampling plots was considered desirable because one of the sampling holes (no. 3) was in each case to be directly adjacent to the taproot of the cotton plant. It was anticipated that this arrange- ment would result in destruction of or serious injury to the plant at each sampling, making the sampling plots unavailable for measuring plant responses. Because of the anticipated destruction of a large num- ber of plants by the repeated samplings during the season, the double (2 diagonally adjacent) plots were set aside for the sampling. The arrangement of the plots and the locations of the sampling plots are shown in figure 34. The insert in the upper lefthand corner of the picture shows the location of the sampling holes. Plots were 35 feet wide and from 37 to 44 feet long. In each plot there were 10 rows of cotton plants, the rows being spaced 42 inches apart. An approach to a Latin- square plot arrangement was followed in 1934, but variations were necessary in that year because of the triangular shape of the tract used for the experiment (fig. 34). SOIL-MOISTURE SAMPLING PLAN, 1934 At each sampling for soil-moisture determinations, soil samples were taken from 6 holes at 3 locations in each plot sampled, and 3 sampling plots were used for each treatment. Sampling was to a depth of 5 feet, and each sample represented 6 inches of soil. Soil-sampling holes at each sampling station were numbered as shown in figure 35. As each plot was sampled, the samples for each 6-inch depth from the 3 holes num- bered alike were combined. There were, therefore, 60 composite samples for each plot sampled, or a total of 540 composite samples at each sam- pling for all plots and treatments. The interval between samplings aver- aged a little less than 9 days. This sampling plan was more intensive than in the previous years of the experiments. Having the extra sampling space provided by the double sampling plots mentioned above eliminated the need for sampling in close proximity to holes previously put down. Bul. 668] Cotton Irrigation 59 FINAL MOISTURE-EQUIVALENT AND PERMANENT-WILTING- PERCENTAGE DETERMINATIONS, 1934 These determinations, previously referred to, were made at Davis. Averages, based on six trials for each 6-inch soil depth, are given in table 9. In the 180 individual permanent-wilting-percentage determina- tions from which the averages were computed, the maximum variation was between 10.66 and 14.74, substantially confirming the preliminary determinations. TABLE 9 Average Moisture Equivalents and Permanent Wilting Percentages, Cotton Experimental Plots, by Depths and Treatments, 1934 Treatment 1 Treatment 2 Treatment 3 Soil depth M.E. P. W. P. M.E. P. W. P. M.E. P. W. P. inches 0- 6 per cent 27.2 27 28.9 28.0 25.5 24 4 23.9 25 30 1 25.8 per cent 12.4 12.4 13.4 13.2 12.4 12.0 11.5 12 13.9 12.3 per cent 26.9 26.2 27 26.9 25.4 24.3 24.0 24.0 27.8 24.3 per cent 12.1 11.7 11.9 12.5 11.7 11 4 11.5 12.4 12.9 11.2 per cent 27.1 27.5 28.8 27.7 26.3 24.0 23.3 26.5 29.3 24.9 per cent 12.2 6-12 12-18 18-24 12.3 13 12.3 24-30. . . 12.1 30-36. . . 11.4 36-42. . . 11.1 42-48 . . 13.6 48-54 . . . 13.6 54-60 11.4 CULTURAL RECORD, 1934 The experimental tract was disked, harrowed, and floated on April 23, and then seeded at the rate of about 35 pounds per acre. The seeding was a little heavier than in usual local farm practice in order to insure a good stand, the resulting stand being almost perfect. By May 7 the plants were up and cotyledons were formed. On May 8 the field was cultivated and the borders between plots were plowed up and necessary irrigation ditches made. The cotton plants were thinned to about 14 inches in the rows May 25 and May 26, when they were about 4 inches above the cotyledon leaves. On June 8 stakes were driven to a top eleva- tion even with the cotyledon leaves at the sides of 10 plants in all plots except those used for soil-moisture sampling, measurements of plant heights being made from these stakes weekly until the plants stopped growing late in August. In treatment 1 the first irrigation after seeding was given June 16, and the first soil-moisture samples were taken about 6 days later. Appar- ently the soil moisture had not yet come to equilibrium, for some very wet soil was encountered. At this time the irrigated plants were reported 60 University of California — Experiment Station to be lighter green than those not yet irrigated, and by June 25, the irri- gated plants had the largest leaves. A small amount of verticillium was appearing in the field. The next irrigations of treatment 1 were on July 7 to 8, 20, and 31; the irrigations of treatment 2 were on July 7 to 8 and 28. In general, by the end of July, the plants in treatment 1 appeared to grow faster and were noticeably larger and denser than in treatment 2. The plants in TABLE 10 Meteorological Kecord at Firebaugh Canal Headquarters, July 3 to October 31, 1934 (Observations mostly at 7 a.m. about 4.5 miles north and 2 miles west of experimental area) Mean temperature Precipi- tation Evaporation, standard Month Maximum Minimum Weather Bureau pan 7 100* 99 93 80§ 93 °F 58 1 59 54 46§ 54 inches 0.57 0.57 inches 12.32| 11.48 August September 9.08 October 4.371 Total or mean 37.2511 • Recorded on 13 days, t Recorded on 28 days. t July 3-31. § Recorded on 20 days. H October 1-26. || July 3-October 26. treatment 3, not yet irrigated, had slowed down in growth and showed severe distress by the end of July. The moisture in the upper 3 feet of soil was then at or near the permanent wilting percentage. The first blossoms were noticed in the field on July 2 and a blossom count was started July 14. Treatment 3 began to shed blossoms late in July. The first count of "sheds" was made July 31. When treatment 3 was irrigated on August 2 to 7, the plants had suffered to the point of stunting, but no drooping of the leaves had been detected. By about August 15, the cotton plants in all of the plots had reached their maximum height. Blossoming practically ceased August 10 in treatment 3 and by the end of the month in treatments 1 and 2. The first open bolls appeared in treatment 3 on August 20, and in treatments 1 and 2 on August 28. No new vegetative growth was noticed in the plots after the first few days of September. By the end of the month, the leaf area of the plants was reduced by the falling off of the lower leaves, and the remaining leaves appeared to be mature. A slight amount of thrips damage ap- peared and was most evident in treatment 3. An open-boll count was begun September 6. Bul. 668] Cotton Irrigation 61 The meteorological record for 1934 is given in table 10. The dates and amounts of irrigations were as follows : Dep th of water applied, Treatment and date in inches 1 : June 16 4.5 July 7-8 5.2 July 20 4.5 July 31 5.0 August 17 5.6 August 30-31 6.6 September 19-21 11.4 Total 42.8 2: July 7-8 5.6 July 28-August 7 13.3 August 30-31 6.6 Total 25.5 3 : August 2-7 13.5 The amounts given are averages of all plots in each treatment. SOIL-MOISTURE RECORD, 1934 According to the usual procedure, the soil-moisture record in the sampling plots was kept current through the season as a guide to carry- ing out the prescribed irrigation treatments. In the following resume of the soil-moisture record, however, the permanent wilting percentages referred to are those made at the end of the season. The moisture con- tents at the indicated depths are the averages of all samplings for those depths. The soil-moisture record for the three treatments of the experiment is shown in figure 35. Soil Moisture, Treatment 1, 1934. — In this treatment the objective was to maintain the soil moisture percentage in the upper third between the moisture equivalent and the permanent wilting percentage. The objective was substantially accomplished except in the upper foot. Even in the upper foot the moisture was above the midpoint between the mois- ture equivalent and the permanent wilting percentage except for short periods just before some of the irrigations. In the fourth and fifth feet, the soil moisture was close to the moisture equivalent throughout the season. Soil Moisture, Treatment 2, 1934. — About the only difference between treatment 2 and treatment 1 was a slightly lower drop in the moisture curve in the first and second feet in treatment 2, but as in treatment 1, the permanent wilting percentage was not reached at any sampling. The objective of allowing the soil moisture to closely approach the per- 1 1 1 1 hr si _J y • ^ T 1 > ] , i, r—J- I < _J / , 1 1 t i ■4 i y i / / i / / } ( } } 7 1 i v 4 I' z * -J / / J / } 1 1 1 _L_L J 1: > — . — ■ i 1 ^ ^ ^ & ^ ^ St 51 s m m i _i_ :_ i S- o o •J I J- >_ i — r ^ o 2 ^ s g 'o *^ 1 CO fc«D I4 I M- _ rrr^r- S .._ _Z 1 1 s £> & ^ ^ ^ «\J «\g \ ^ 1 1 1 V ) r ^ '/ -7 / "\ > 1 J 1 \ I I 1 ,/ ■A .:-:: 7 — 1 * t J \ } I i I _J 1 J 7 -V T 7 _ 4 I i 3^ -r n i i , 2 ^ P Z 4. 2 ~\ J t: l-T I * * * ^ ^ ^ ^ S> ^i & ^ «\g ©J V ^» Bul. 668] Cotton Irrigation 63 manent wilting percentage, but not to reach it in more than 1 sample, was not accomplished. Soil Moisture, Treatment 3, 1934. — In this treatment the cotton plants were to be allowed to show severe distress in at least two thirds of the plots before irrigating. Only a single irrigation was given, on August 2 to 7. Before this irrigation the permanent wilting percentage was reached in the first foot and closely approached in the second and third feet, In the fourth and fifth feet it was reduced about to the midpoint between the permanent wilting percentage and the moisture equivalent. The irrigation early in August totaled 13.5 inches in depth. This was sufficient to maintain the soil moisture well above the permanent wilting percentage below the first foot for the remainder of the season, but the approximate flatness of the soil-moisture curve for the upper foot after the first week in September indicates that the soil moisture in that foot was at least approximately at the permanent wilting percentage during the last 5 or 6 weeks of the season. USE OF WATER, 1934 Loss or use of water in 1934 was computed for treatments 2 and 3 and the results are given, by months, in figure 36 12 ; a similar computation for treatment 1 was not made because of insufficiency of soil-moisture sam- plings between irrigations. The maximum monthly use was in July and August in both treatments, that in treatment 2 being about twice that in treatment 3 in those months. The total use of 23.6 inches in depth by treatment 2 about equaled the total depth of irrigations (25.48 inches), not counting preirriga- tions. It is also approximately the same as the use in the "wet" treat- ments in the heavy soil in 1932 and 1933. (See p. 43 and 44.) Because treatment 3 was irrigated only once, surface evaporation included in "loss" or "use" can be assumed to have been less in it than in treatment 2, which was irrigated three times. The plants were defi- nitely and significantly smaller in treatment 3 than in treatment 2, as shown by length-growth measurements and weights of plants (p. 64 and 66), obviously because of differences in irrigation and resulting soil- moisture conditions as brought out in the soil-moisture record. MAIN-STEM LENGTH GROWTH, 1934 Length-growth measurements were started on June 8 on 10 plants in each plot, when the plants were about 5 inches tall. Measurements were made from the top of stakes set level with the cotyledon leaves of the 12 For meaning of "loss" or "use" see discussion of use of water in 1932 and 1933, p. 38. 1 1 1 1 1 1 1 1 1 Loss of water daring months, inches depth ? "rent May Je//?e Ja/j/ Joy. Sept Oct M>y. To to/ — 2 3 08 /5 ,5 35 35 35 03 ?3 45 40 2-3 0.4 236 /50 — 2. / / / / / / / s ~3 / s / I y / . /' '/ /< / _, ^ / ^ ^ 30 23 26 24 22 20 ts /6 14 12 to /O ZO 3/ /O 20 30 /O 20 3/ /O 20 3/ /O 20 30 /O 20 3/ /O 20 30 Mqy Jvr?e Jv/y Aug. Sept. Oct. /Voit. Fig. 36. — Use of water, treatments 2 and 3, 1934. 35 30 25 s 20 I ^ /5 S to / v Treat, '/neat i / ; / / '''' """ .__.- -eatment 2 // S^~ ^Tr eatment 3 // Y J tf 20 30 /0 20 3/ /O 20 3/ /O 20 Jane JU/ * ,934 AU S Sept Fig. 37.- —Main-stem length growth, 1934. Bul. 668] Cotton Irrigation 65 plants measured. None of the plots had been irrigated when the stakes were set. Measurements were made weekly to August 6 and finally on August 22, when main-stem length growth had about ceased. Figure 37 shows the average growth by treatments. Up to the first irrigation of treatment 1 (June 16), all treatments made substantially the same growth, but after that the growth of treatment 1 definitely increased, TABLE 11 Yields of Seed Cotton from Experimetal Plots, 1934* Treatment 1 Treatment 2 Treatment 3 Plot no. Plants Yieldf Plot no. Plants Yieldf Plot no. plants Yieldf number pounds number pounds number pounds 2 116 30 3 118 35 1 110 34 8 115 29 6 114 30 4 121 31 15 132 36 9 100 32 11 129 24 18 128 33 10 129 32 17 127 34 21 134 32 27 133 32 23 132 35 26 121 32 30 121 32 28 122 33 33 126 37 31 130 40 32 128 32 37 126 40 34 122 31 36 123 30 40 124 39 44 128 39 42 117 38 49 124 38 46 126 32 47 123 29 51 121 40 50 119 37 48 122 29 Total . . 1,367 386 Total . . 1,340 372 Total . . 1,354 349 Av. per Av. per Av. per plot . . . 124 35.1 plot . . . 122 33.8 plot . . . 123 31.7 Av. per Av. per Av. per plant . 0.28±0.0064 plant . 0.28±0.0064 plant . 0.26±0.0076 *Trea tment 1 re< ;eived 7 irriga :ions, total ing42.8in ?hes in depth; treatment 2 received 3 irrigations totaling 25.5 inches in depth; treatment 3 received 1 irrigation of 13.5 inches in depth, t Probable errors are given for the averages per plant. as did the growth of treatment 2 after the first irrigation (July 7 to 8). The length growth was not increased in any of the treatments by irri- gation August 2. The growth season ended about the midle of August. The plant heights on final dates of measurement, with probable errors, were: treatment 1, 32.5 ± 0.26; treatment 2, 29.5 ± 0.26 ; treatment 3, 23.6 ± 0.28. Differences between all treatments are significant. YIELDS OF COTTON, 1934 Before gathering the cotton from the 33 plant-response plots used in the experiment, the 2 outside rows in each plot were eliminated, and each end cut off to give net plots 5 rows wide and 30 feet long, except that a few of the net plots were less than 30 feet long, making it necessary in such cases to adjust yields to 30-foot rows. All of the cotton in each 66 University of California — Experiment Station of these reduced plots was picked and the seed cotton weighed. Yields, by plots and treatments, were as given in table 11. There is no significant difference between the yields per plant or per plot in the different treatments. WEIGHT OF PLANTS, 1934 As a further check on plant growth in relation to the different treat- ments, the entire plants, less roots, in each reduced yield plot were weighed. After making the same adjustments for plot lengths that were made for yields of cotton, the average weights per plant were as follows : treatment 1, 0.72 pound ; treatment 2, 0.59 pound ; treatment 3, 0.37 pound. Counting the weight per plant in treatment 1 as 100 per cent, the weight per plant from treatments 2 and 3 came to 82 and 51 per cent, respectively, of that for treatment 1. The number of plants in treatment 2 was 98 per cent, and in treatment 3, 99 per cent of the num- ber in treatment 1. INVESTIGATIONS DURING 1935 In 1935 the investigations were transferred to the Woods Stone ranch, about 2.5 miles west of Shafter, in the N. W. % of N. W. %, Sec. 18, T. 28 S., R. 25 E. In 1932 and 1933 the experiments had been conducted on soil with moisture equivalents ranging from 30 to 35 per cent, and permanent wilting percentages, from 16 to 18. In 1934 they had been on soil with moisture equivalents of about 24 to 30 per cent, and permanent wilting percentages of about 11 to 15. It was now desired to continue with the experiments on soil with moisture equivalents and permanent wilting percentages lower than in 1934, and it was also desired to carry on the work in close proximity to the Government Cotton Station at Shafter. The only tract found which seemed to meet the requirements was on the Woods Stone ranch. Preliminary tests showed a more variable range in moisture equivalents than was desired, but in the absence of a more suitable location the decision was reached to continue the experi- ments on a portion of this ranch. In addition to the preliminary moisture equivalent, test samples were collected March 30 for both moisture-equiv- alent and permaneiit-wilting-percentage determinations. Owing to a necessary shift in sampling plots in treatment 3, an additional set of moisture-equivalent and permanent-wil ting-percentage determinations were made for plots 29, 44, and 52 of that treatment after August 7. Average values for all of the determinations are given in table 12. The soil on which the experiments were conducted was classified in the reconnaissance soil survey as Hanf ord loam, but in a resurvey, made subsequent to the time of the experiment, it was reclassified as Hesperia fine sandy loam. The reconnaissance soil map showed an alkali area from Bul. 668] Cotton Irrigation 67 ^4 to % mile to the south, but the subsequent soil map carries the alkali boundary to or very close to the block selected. The probable effect of alkali on the experiment results is discussed later (p. 77). Headquarters for the investigations during 1935 were at the Govern- ment Cotton Station near Shafter. IRRIGATION TREATMENTS, 1935 As outlined, irrigation treatments were to be the same as in 1934, that is : treatment 1, soil moisture not to be allowed to drop below the upper third of the range between the moisture equivalent and permanent wilt- TABLE 12 Average Moisture Equivalents and Permanent Wilting Percentages for Soil of Cotton Experimental Plots, 1935 Moisture equivalent Permanent wilting percentages Soil depth For the whole area* For plots 29, 44, and 52 (treatment 3)t Treatment 2f Treatment 3t For plots 29, 44, and 52 (treatment 3) § per cent 16.1 16.2 17.0 22.3 21.8 ....1 per cent 16.8 17.1 20.7 23.9 21.1 20.8 per cent 6.84 6.76 7.48 9.18 7.61 ....1 per cent 6.8 6.6 7.8 6.9 6.9 ...1 per cent 7 1 2d foot 7 4 3d foot 4th foot 8.7 9 5 5th foot 9.2 9 3 * From 5 holes distributed over the area. t Average of 2 samples for each foot in depth for each of the 3 plots. % From composite samples from 3 holes for each foot in depth for each of treatments 2 and 3. Values given are averages of 10 trials. § Based on 10 trials for each foot in depth for each of the 3 plots. 1 No determinations made for sixth foot. Note: All samples taken during the season for soil-moisture determinations were saved for final moisture-equivalent and permanent-wilting-percentage determinations but were lost in a fire before determinations were made. ing percentage; treatment 2, soil moisture to be allowed to closely ap- proach the permanent wilting percentage, but not to reach it in more than 1 sample ; treatment 3, cotton plants to be allowed to show severe distress in at least two thirds of the plots of the treatment before irrigat- ing. These differences in soil moisture were expected to give a basis for obtaining further evidence as to whether the response of cotton plants to soil-moisture conditions is affected by differences in the soil moisture between the moisture equivalent and slightly above the permanent wilting percentage. Water for irrigation was obtained from a near-by well. REPLICATIONS AND SAMPLING PLOTS, 1935 The experimental area was divided into 60 plots, 30 feet by 50 feet. Of these, 20 were used for soil-moisture sampling and 40 were used for measuring plant responses. Among the plant-response plots there were 68 University of California — Experiment Station 14 replications for treatment 1, 13 for treatment 2, and 13 for treatment 3. Arrangement of the plots and locations of those used for moisture sam- pling are shown in figure 38. An attempt was made to distribute the treat- ments on the basis of Latin squares, but because of inadvertent flooding 1 1 1 2 1 3 i 4 1 1 6 2 7 3 8 1 9 3 10 3 II 1 12 1 13 1 14 2 15 3 16 1 17 2 18 1 19 2 20 1 21 1 22 2. 23 1 24 2 25 2 26 1 27 2 28 3 29 1 30 3 31 1 32 1 33 3. 34 1 35 2 36 1 37 2 38 3 1 40 2 41 3 42 2 43 3 44 i 45 2 46 3 47 1 48 1 49 1 50 2 51 3 52 3 53 2 54 1 55 & 3 57 1 58 ft a> 2.= Tr 15= P eatment ot Supply Double line indicate pipe ° s sampling p ots Fig. 38. — Arrangement of cotton plots in 1935. of some of the plots from the adjoining field it was necessary to make shifts in treatment numbers, which resulted in unevenness in treatment distribution. SOIL-MOISTURE SAMPLING PLAN, 1935 The sampling plan was varied slightly from that followed in 1934. In treatments 1 and 3, sampling was by 1-foot, instead of 6-inch, depths. In treatment 2, 6-inch samples were taken until the first irrigation, with Bul. 668] Cotton Irrigation 69 1-foot samples thereafter. Samples were taken in all treatments in 3 holes in each sampling plot, instead of 6, as in 1934. These variations from 1934 were considered justified by the results obtained in 1934 and made it possible for one man to conduct the experiment with very little assistance. In treatments 1 and 2, the 3 sampling holes in each plot corresponded to holes 1, 3, and 6 in the 1934 experiments. (See fig. 34.) Samplings were carried to a depth of 6 feet, instead of 5 feet, as in 1934. CULTURAL RECORD, 1935 The experimental field chosen was part of a larger field planted to cotton. The whole field had been preirrigated early in April. A supple- mental preirrigation was given the experimental tract, beginning on April 17. This wet the soil to a depth of at least 5 or 6 feet. The cotton was planted April 27. As the stand in the experimental plot was not satisfactory, misses were replanted, some a second time. Very few plants of either replanting grew, the difficulty being diagnosed at the time as sore shin. On July 4, it was noted that young cotton plants were "pinching off" at the rate of about 2 a day in the area not yet irrigated. By July a few squares were shedding. The squares appeared to be physiologically defective, rather than to be injured by insects. Late in July shedding of squares was attributed mainly to insect injury. On July 16 it was noted that it was not possible to differentiate between normal shedding and shedding due to insect injury. By August 30 blos- soming was falling off rapidly and by September 13 was about over. The dates and amounts of irrigation were as follows : Depth of water applied, Treatment and date in inches 1 : June 21 4.0 July 9 4.7 July 23 4.2 August 4 4.4 August 19 4.9 September 4 6.7 September 26 4.1 Total .33.0 2 : July 5 5.7 July 26 5.2 August 14 6.7 September 4 6.7 Total 24.3 3 : July 21 and 30 11.7 August 27 8.6 Total 20.3 The amounts given are the averages of all plots in each treatment. <0 ! l 1| ii 1 t 1 i / J } J y C\/ -j> / ^ y 3 fe fj / •£ $ t { \ r> 1 1 1 1 1 d 1 b | T5 ¥ 1j I j J I | I i 1 I r \ - I ,r < j> 1 §s 1 _r 1 7 jj 1 .«. . 2 ,'„ 7 7>- i — , p J { b 1 1 1 — i =f= : %-■ Ll _J -X- ? <■ T X I J:: H Z x- I,*' ffl :^ L ^ ^> 'C) ^ k\ «\J M nJ ^ » *> £ ^ ^ ^ Jf) <\j ^j \! ^ ' «\j <\J ?> <\j /-"'" ,'' - / / Trealme at 3 '// ft / y V J<. S y to eo June /O 20 Ju/j/ 3/ S935 /O ZO so 30 Sept Fig. 41. — Main-stem length growth, 1935. In addition to the measurements on 10 plants, the heights of 25 plants in the center row of each plot, except soil-sampling plots, were measured at the end of the season. Where there were not 25 plants in the center row, the measurements were continued up to the specified number in the adjacent row, beginning at the end of the row. At the same time that these measurements of plant heights were made, the number and lengths of vegetative branches, the number of nodes, and the number of frozen (unopened) bolls was determined for each of the 25 plants. The data are summarized in table 13. Differences in plant heights between treat- ments 1 and 2 and treatments 1 and 3 were significant, but the difference between treatments 2 and 3 was not significant. Because of the unevenness in the growth of the cotton plants, to which reference is made later (p. 77), an additional check on plant heights in 74 University of California — Experiment Station relation to treatments was made by measuring the heights of all plants in alternate (second, fourth, and sixth) rows in all plots, except sampling plots. These measurements were made after harvest. By treatments the average heights were : treatment 1, 55.44 inches ; treatment 2, 51.36 inches; treatment 3, 49.20 inches. These plant heights are in the same order, by treatments, as were those for the 25 plants in each plot shown in table 13, and also those for the 10 plants in each plot on which main- stem length growth was measured approximately weekly through the season. In further analysis of the data relating to plant heights, the experi- mental area was divided into four equal parts separated by lines north TABLE 13 Growth Data, by Treatments, 1935* Treatment Average plant height Vegetative branches per plant, average Average length of vegetative branches Nodes per plant, average Frozen bolls per plant, average 1 inches 56 3 ±0.33 52.3±0.38 51.4±0.41 number 4.60 4.09 4.29 inches 36.33 32.27 33.13 number 21.48 20.33 20.53 number 5.65 2 4.76 3 4.99 * For description of treatments see p. 67. Treatment 1 received 7 irrigations, totaling 33.0 inches in depth; treatment 2 received 4 irrigations, totaling 24.3 inches in depth; treatment 3 received 2 irrigations, totaling 20.3 inches in depth. and south and east and west through the center of the area. The average plant heights and the number of plots were then determined for each treatment in each of the four quarters of the area, this procedure being followed for each of the three sets of plant-height measurement that have been given. Because of the elimination from all consideration of plant responses of the plots in which soil-moisture determinations were made, the number of plots for each treatment in the four quarters of the area was found to be not in every case well balanced, the principal unbalance being for treatments 2 and 3 for all sets of plant heights. This unbalance may be of importance because of the variability of the soil in different parts of the area. A study of the unbalance in treatment 2 indicated that it probably did not materially affect the results. In the case of treatment 3, however, there were 8 plots in the east half of the experimental areas, 4 plots in the northwest quarter, and none in the southwest quarter, for the measurements on all plants and on 25 plants ; and the measurements on 10 plants in treatment 3 were all in the east half of the area. The average plant heights for treatment 3, for the east half and the west half of the experimental area, were : Bul. Cotton Irrigation 75 All plants in alternate rows — east half, 3.64 feet; west half (all in northwest quarter), 5.01 feet. Twenty-five plants in each plot — east half, 3.67 feet; west half (all in northwest quarter), 5.02 feet. Ten plants in each plot — east half, 4.13 feet; west half, no plants measured. The conclusion from the above analysis is that the values for plant heights for treatment 3 in all three sets of measurements probably were affected by causes other than moisture treatments, and that the heights TABLE 14 Average Number of Bolls and Squares per Plant on Ground on Dates of Counting, by Treatments, 1935* Count Treatment 1 Treatment 2 Treatment 3 no. Date Bolls Squares Date Bolls Squares Date Bolls Squares 1 2 3....; 4 5 Aug. 17 Aug. 19 Aug. 28 Sept. 3 Sept. 24 3.95 0.48 7.51 7.82 11.34 31.10 3.70 0.57 2.60 1.67 0.74 9.28 Aug. 6 Aug. 13 Aug. 26 Sept. 3 Sept. 24 1.34 2.56 8.75 13.10 8.56 34.31 0.95 2.58 1.61 2.11 0.21 7.46 Aug. 6 Aug. 16 Aug. 27 Sept. 3 Sept. 24 2.43 8.92 12.84 12.43 7.00 43.62 0.98 5.40 4.24 1.55 0.41 Totals to Sept. 24 . 12.58 * For description of treatments see p. 67. Treatment 1 received 7 irrigations, totaling 33.01 inches in depth; treatment 2 received 4 irrigations, totaling 24.3 inches in depth; treatment 3 received 2 irrigations, totaling 20.3 inches in depth. of plants were affected more by position of the plots than by irrigation treatments. This conclusion is confirmed by the results of a similar analysis of the yields in 1935 in treatment 3. The high average yield in this treat- ment occurred in the northwest quarter of the area and the low average yields in the east half, the same unbalance in number of plots having been present in the case of yields as in the case of measurements of plant heights on all plants and on 25 plants. SHEDS OF BOLLS AND SQUARES, 1935 Beginning in August a count was made of sheds of bolls and squares from the plants in each plot which were measured for main-stem length growth. Shedding started later in treatment 1 than in treatments 2 and 3. The count for all treatments closed September 24. Dates of counts varied with the treatment. The record is presented in table 14. The largest total number of sheds of both bolls and squares was in treatment 3, which received the least water. The smallest number of sheds of bolls was in treatment 1, and of squares was in treatment 2. 76 University of California — Experiment Station YIELDS OF COTTON, 1935 Yields of seed cotton were obtained for a net area 15 feet by 34 feet in each plot, except those in which soil-moisture samples were taken. Reduction in acreage was to eliminate border eifect. Each net plot included 5 rows 34 feet long. TABLE 15 Yields of Seed Cotton by Plots and Treatments, and Average Yields Per Plant, Experimental Plots, 1835* Treatment 1 Treatment 2 Treatment 3 Plot no. Plants Yield Plot no. Plants Yield Plot no. Plants Yield 1 number 92 107 74 96 98 112 102 74 97 99 104 104 66 114 1,339 95.6 pounds 26 33 22 25 26 23 17 23 18 26 18 22 25 29 333 23.8 0.25 ±0.011 6 18 20 25 26 33 36 41 46 51 54 56 59 number 66 112 85 79 89 119 119 102 101 115 99 77 111 1,274 98.0 pounds 13 24 22 25 26 29 28 32 28 28 30 27 29 341 26.2 0.27 ±0.008 5 10 11 16 31 34 42 47 49 53 57 60 number 73 42 64 79 104 119 112 112 85 108 111 111 1,120 93.3 pounds 37 3 24 4 30 19 28 22 25 27 23 30 27 32 26 35 37 48 29 30 24 50 55 25 58 Total.... Av. per plot Av. per plant... Total.... Av. per plot Av. per plant... Total.... Av. per plot .... Av. per plant... 328 27.3 0.29 ±0.026 * For description of treatments see p. 67. Treatment 1 received 7 irrigations, totaling 33.0 inches in depth; treatment 2 received 4 irrigations, totaling 24.3 inches in depth; treatment 3 received 2 irrigations, totaling 20.3 inches in depth. Yields per plot and average yields per plant are given in table 15. The lowest yields per plant were in treatment 1 and the highest in treatment 3. This was a reversal of the results in previous years. But the differences between the yields in the different plots were not large enough to be significant. As shown by the soil -moisture curves in figure 39 (p. 70), there was very little difference in soil moisture in the three treatments, except for the upper 2 feet in treatment 3. Bul. 668] Cotton Irrigation 77 WEIGHT OF PLANTS, 1935 The mature plants, less roots, in the portion of each plot from which the yield of seed cotton was obtained were counted and weighed at the end of the season. The weights by treatments and by plots are given in table 16 in connection with the moisture-equivalent determinations made at the close of the season with a view to ascertaining the reason for unevenness of plant responses. Variation in the number of plants per plot, by treatments, was : treatment 1, 66 to 114 ; treatment 2, 66 to 119 ; treatment 3, 42 to 119. The average weights per plant were : treatment 1, 0.78 pound; treatment 2, 0.63 pound; treatment 3, 0.71 pound. An analysis of these weights showed that, as in the case of yields, the differences are not significant. UNEVEN GROWTH OF COTTON PLANTS, 1935 In a preceding paragraph, reference was made to the fact that pre- liminary moisture-equivalent determinations of the tract selected for the experiments showed a more variable range than was desired. The cultural record (p. 69) brought out the difficulty of obtaining an even stand of cotton in the plots, very few of the replants growing. At the time, this difficulty was attributed to sore shin. As the season progressed, uneven growth of the cotton plants gave further evidence that conditions present were more or less adverse. As the cotton plants approached maturity, unexplainable unevenness of growth, regardless of irrigation treatments, was cleary noticeable over much of the area. The most probable reason for this uneven growth seemed to be soil differences not thus far detected. An attempt was made to check this by more intensive moisture-equivalent determinations. Samples of soil were obtained from the center of the north half and the south half of each plot. In each case, sampling was from the upper 6 feet of soil. Results are presented in table 16. Average weights of the mature plants, as cut from the plots after harvesting of cotton, also are listed in this table. These moisture- equivalent determinations furnish further evidence of the variability of the soil as between the plots, although the averages for these treat- ments are close together. There is no consistent relation between varia- tions in moisture equivalents and variations in the weights of plants in the individual plots. Since the moisture equivalents given in table 16 failed to furnish a satisfactory explanation of the unevenness of growth in the experimental plots, and since the difficulty of obtaining an even stand at the beginning of the season, as well as the obvious unequal response of the cotton plants to the different irrigation treatments, suggested alkali as a possible cause 10 CO X OS M iH fc M DQ H ^ < o -i fin ^J E^ o o o i-3 o o a: ^2 to ,d a "** <= oooooo«o>o OOO |C fr- t- "5 "3, 3 "• 8" ee CO o o 8 "8 5-2 5 QC •*fioco©ooeoio.-i • N N ffl | CN .2 > fe " 0)!D01053H1001N ■ W H OO OS 01 ^ a- o> ft a 3 ^2 V N*0>*»NNlfl • OO ^ -- • 1 c is*" ^OI-OhhhOC . O <-h ^h . cN H Ph 3 1 - +J 6 a 3. c a> o 3 o. Pm o > ++ (- ©-*f & oect>.cooooo500«ot^oo \Z c^ S* P, -p a>OS«N^H»OOSl>..-i • | if oo 03 s- ^HOot^oo^H^oO'-iot^^ ■ r^ as H p* s ' " += C o a a a O o a a Oh ::::::::::::: h «< > ce oooifltoratDHCHiitoo) <— JCOC<3'>*. M cS S rt 1* S "" NNtOOlOMNOSNO'fiO S ** a o £3 3.2 lOCOr-WU5»-l»0'«* a a> a 3 fc i CT ©t^O3OS<-lOfr-OSO*OOc0i-l C as H Ph s 3 o o cS fl a ft o Ph ::::::::::::: -j a ::::::::::::: c eo^osCNt^cicNiot^oooiooo > > CN c^ o- M CO ^ »C »o •c 42? .-5 ,2 M" cu 2^ >"S.2 Bul. 668] Cotton Irrigation 79 of the inconsistencies, determinations of total salts were made at the common corner of plots 38, 39, 44, and 45, at the common corner of plots 8, 9, 14, and 15, and in plot 29 : A l-to-5 water extract was used, and the results were computed to the basis of the oven-dry weight of the soil. Samples for these determinations were collected in November, 1940. These samples were from the 0-to-3-f oot and from the 4-to-6-f oot depths, and for each depth ten determinations were made. The average total salts at the corner of plots 38, 39, 44, and 45 was 330 parts per million for the depth of to 3 feet and 294 parts per million for the depth of 4 to 6 feet. For the corner of plots 8, 9, 14, and 15 the averages for the two depths were, respectively, 540 and 693 parts per million. For plot 29 the averages for the 0-to-3-foot and for the 4-to-6-foot depths were 972 and 6,690 parts per million, respectively. Results of these deter- minations confirmed the conclusion that results in 1935 may have been affected by the salt content of the soil. The presence below the surface of layers of sand in some plots may also have had an influence. "While these uneven soil conditions justify some doubt as to the conclusiveness of the investigations in 1935, the results obtained are believed to have some value. DEPTH OF USE OF WATER BY COTTON PLANTS Brief descriptions of root development by cotton plants have previ- ously been presented in reporting on the investigations in 1931 (p. 15). Similar studies were not included in the investigations in 1932 to 1935, but some information on depth of rooting is to be found in the record of water lost in each foot of depth as disclosed by the soil-moisture determi- nations. This is presented in table 17. In this table the water lost is shown for each foot of depth for each treatment in which sufficient soil-moisture determinations were made to furnish a complete seasonal record, the values given being expressed as percentage of the total loss from the full depth sampled. The table indicates that the roots of the cotton plants reached at least to the depths sampled, that is, 6 feet in 1932, 1933, and 1935, and 5 feet in 1934. However, the most important information presented in this table is that, on the average, at least two thirds of the moisture extraction from the heavy soil of the experiments in 1932 and 1933 was from the upper 2 feet of soil, and, in the case of the lighter soils used in 1934 and 1935, approximately three fourths was from the upper 3 feet. In each case, of course, much of the loss from the surface foot was surface evaporation, so that the moisture losses from the first foot are not truly indicative of root distribution in that depth. The loss from the second foot, however, was in each instance much higher than from any lower depth, indicating a larger concentration of roots at that 80 University of California — Experiment Station TABLE 17 Water Lost in Experimental Cotton Plots by Surface Evaporation and Transpiration During Each Season* (Expressed as Percentage of the Total Loss from the Full Depth Sampled) Year and Soil depth treatment no.f 1st foot 2d foot 3d foot 4th foot 5th foot 6th foot 1932: 1 per cent 66.0 35.9 54.3 59.3 56.3 35.0 53.4 51.5 51.8 43.7 53.8 54.9 46.8 27.9 47.2 43.7 46.2 31.0 24.7 27.9 26.8 35.3 31.1 per cent 15.9 23.5 23.3 23.1 19.2 24.3 18.2 21.1 17.4 23.1 18.6 22 22.8 20.9 23.0 21.0 21.1 26.5 25.8 26.2 27.2 27.9 27.5 per cent 6.1 19.9 10 8.3 6.4 20.5 12.7 12 12.0 16.8 12.4 11.4 11.7 19.4 12.6 11.3 13.5 16.2 21.2 18.7 19.3 13.6 16.5 per cent 5.2 17.7 9 6.0 10.7 13.1 8.2 10.0 8.8 7.3 10.2 5.7 10 9 15.4 7.5 9.3 9.4 14.0 16.9 15.5 12.0 13.0 12.5 per cent 5.7 3.9 1.9 4.3 4.4 4.5 5.2 4.3 5.2 5.4 3.3 2.8 3.9 10.0 7.1 7.7 5.7 12.3 11.4 11.9 9.5 4.7 7.1 . per cent 1.1 2 3 -0.9 1.5 4 -1.0 5 3.0 6 10 2.6 2.3 Average. . . . 1933: 1 1.2 4.8 2 3.7 3 1.7 4 3.2 5 3.9 6 6.4 10 2.6 12 7.0 Average 1934 :t 2 4.2 3 Average 1935:{ 2 5.2 3 5.5 Average. . . . 5.4 * Includes all of the soil-moisture record, whether indicating an increase or a decrease. t For quantities of water received by all treatments in each year of the experiments, see table 20. j Too many irrigations in treatment 1 for complete analysis. depth than any lower 1-foot depth. The data of the table are conclusive that the principal root concentrations were in the upper 2 or 3 feet in the heavy soils of the experiments in 1932 and 1933, and in the upper 3 or 4 feet in the lighter soils used in 1934 and 1935. It is the upper 2 or 3 feet in depth in the heavier soils and the upper 3 or 4 feet in the lighter soils that are of most importance in cotton production under San Joaquin Valley conditions. Preirrigations, if they penetrate 5 or 6 feet, should furnish all the soil moisture needed by the cotton plants at depths below the zone of principal root activity. Bul. 668] Cotton Irrigation 81 COTTON CLASSIFICATION AND SEED AND LINT DETERMINATIONS Since there has been some discussion of the quality of cotton grown under irrigation, the assistance of the United States Bureau of Agri- cultural Economics was obtained in classifying samples from the ob- servational and experimental areas each year, 1931 to 1935. The classi- fication in 1931, 1932, and 1933 included the grade, length of staple, uniformity, fiber body, and strength; in 1934 and 1935 it included only the grade and staple length. Before being forwarded to Washington for classification, these same samples were used for determination of weight of seed cotton and weight of seed from 10 bolls, percentage of lint, lint index (weight of lint from 100 seeds), seed index (weight of 100 seeds), and bolls per pound of seed cotton. These latter determinations were made by a member of our own staff under the general direction of the staff at the federal Cotton Station at Shafter. The results of the cotton classification and of the determinations of the other characteristics men- tioned, except for 1931, are presented in tables 18 and 19. The results for 1931 are not included in the tables, because the work in that year was merely observational, without differential irrigation treatments. The samples used for classification and for determination of seed and lint indexes were limited to 10 bolls from each plot and were col- lected from plants considered by the observer to be representative of the plots, rather than at random. Cotton Classification. — An analysis of the differences in the staple lengths given in table 18 shows, for both 4-lock and 5-lock bolls, that, in 1932, the average staple length in treatment 1 was significantly greater than in treatments 2 and 6. In 1933 the average staple length in treat- ment 1 was also significantly greater than in treatments 2 and 6, in the case of 4-lock bolls; but not significantly greater than in treatment 2, and barely significantly greater than in treatment 6, in the case of 5-lock bolls. In 1934 the average staple lengths for both 4- and 5-lock samples were the same for treatments 1 and 2, but the average staple lengths in both treatments 1 and 2 were significantly greater than in treatment 3. In 1935 none of the differences for either the 4- or 5-lock boll samples was significant. In the case of 4-lock boll samples, the average in treatment 1 was significantly greater than in treatment 4 in 1932, and than in treatments 3, 10, and 12, in 1933. In the above summary, differences are considered to be significant in a few instances in which they are from 2.9 to 3.0 times, instead of about 3.2 times, the probable errors of the difference. 82 University of California — Experiment Station TABLE 18 Classification of Cotton from Experimental Irrigation Plots, 1932 to 1935* Year and treat- ment no. Sample Gradej 4-lock bolls 5-lock bolls Staple length t 4-lock bolls 5-lock bolls Character 1932: 1... 2... 3... 4... 5 ... 6... 7... 8... 9... 10... 11... Field 1933: 1... 2... 3 ... 4... 5 ... 6... 7... 8... 9... 10... 12... 1934: 1 2 3 1935: 1... number ISM, 8GM 9GM 2SM, 7GM 2SM, 7GM ISM, 8GM 3SM, 6GM 2GM 2GM 2GM ISM, 5GM 1GM 3SM, 3GM ISM, 8GM 9GM 2SM, 7GM 2SM, 7GM 2SM, 7GM 9GM ISM 1GM ISM 9GM 2SM, 7GM GMLSp GMLSp GMLSp 1GM, 7GMLSp, 5GMEW, lGMSp 1GM, 5GMLSp, 7GMEW 7GMLSp, 5GMEW 9GM 9GM 9GM ISM, 8GM 2SM, 7GM ISM, 8GM 2GM 2GM 2GM 6GM 1GM 6GM 9GM 9GM 9GM ISM, 8GM. 2SM, 7GM. 1SM.8GM. 1GM 1GM ISM 9GM ISM, 8GM. GMLSp GMLSp GMLSp 7GMLSp, 7GMEW. 6GMLSp, 7GMEW. 6GMLSp, 5GMEW, lGMSp.. 1/82 inch 34.56±0 118 33.11±0 135 34 11±0 135 34.00±0.112 34 00±0.116 33.78±0 099 34.00 34.00 34.50 34.33±0.142 1/82 inch 34.17±0.112 34 22±0 099 33.44=b0.118 33.67±0.159 33.89±0.237 34 22±0 150 33. 00=1=0.1 12 36 00 34.00 33 00 33.67±0.112 33.67±0.159 36.09±0.061 36.09±0.143 35.45±0.140 36.64±0.114 36.62±0.122 36.58±0.130 34 33=b0.12l) 33.44±0.163 34 11 ±0 075 34.11±0.075 33 67d=0 112 33.78±0.099 \ 34.50 34.50 34.00 34.50±0 151 33. 83=1=0. 112 33.67±0 159) 33.22±0 099 33.78±0 150 33.56±0 198 34.11±0.075 33.00±0 159 > 35.00 34.00 34.00 33 44±0 118 33.67±0 159 J 35.91±0.061 35.91±0.061 35.54zb0.106 J 36.93±0.048 36.69=1=0. 118 \ 36.67±0.096 All samples graded Extra White and classed as "weak cotton, normal for this kind of cot- ton"; "uniform" in length; "light" in fiber body; and "normal" in strength All samples graded Extra White; one sample in treat- ment 12 classed as "fair" in fiber strength; all oth- ers either as "nor- mal" or "strong ' Grade determined in all cases from composite sam- ples; under "char- acter" all sam- ples classed as "regular" in uni- formity ," nor mal " in strength, and "normal" in body All samples classed as "normal" in character * For descriptions of treatments see p. 17 for 1932 and 1933, p. 57 for 1934, and p. 67 for 1935. In 1932 and 1933, treatments 1 and 5 were "wet" treatments; treatment 10 was a "medium" treatment; treatments 2 and 6 were "dry." In 1934 and 1935, treatment 1 was "wet," treatment 2 received "medium" quantities of water, and treatment 3 received the least water. For quantities of water received by all treatments in each year of the experiments, see table 20. t SM is Strict Middling; GM, Good Middling; LSp, Light Spotted; Sp, Spotted; EW, Extra White. J Probable errors are given where there were enough samples. a JS c e pq a$ 03 JS'S i s o>ir-r^iOcoeoooocoOTt< !>.©-l ■8js 4- CD I 000>OI^O«t^«5©COCN(M 4-° co e 43.3 37.1 40 9 39.4 41.9 36.6 46.2 46.3 44.7 41.9 39.0 ^OON^NNOf^Oie COOS©*-HCOOOCO— icoos-Hco»c«^eO'— > >o 0-*IOCOcOt>.00050»-l CO « •-! k. ^ •<»< OS 00 Oi "0 ■* CD -o OS ■* O ""f >* -<*i OS >0 t^ t^ iti oo 493 s CO O os o CO o o o 4 o CO o 8*3.8 ^ s "* -H -* -H •° -ti 10 -H .o 4 10 -H a ^1 CO 3 y |l| to 00 oo r^ ^ ,_, "oJ2 lO CM U5 o -o CM OS O0 lO cm a> O OO OO CO •* co ,53 s CM O 1-H O co o CM d CO o CO ^H 4-° 3 e CO 4 CO 4 CO 4| CO 4| co 4 CO 4 co 00 o o ■^1 r- 4* rr OS 00 IO uJS 1 CO O o o *-> O C35 ^ b- i-i CO »-l ,So B 4 o 4 © co O CO O CO © CO o 15* .4* Qb ** M - -H ^ -H ^ -H ~ -H - -H ■ H Moi T3 '3 © S*2 t _ "* CM CO CM os »0 o O o CO m s CO rt t n CM -h © r* O i-i CO .-! e -f O ■*• o CO O ■^ o ■* o CO O 4^ S> •"■ -h ^ -H ^ -H - -H ~ -n rt 41 CO CM ■* "5 CM •^ B) CO OO OS OS lO O CO O -H co co -H OO '- — B CO O o o CO O T»< M co ^ t- o ^ CO a;— i m £"0 OS O os o OO o OO o 00 O 00 O A* O ■H 4i ■H ■H ^ ■H ~C«- «> .s°s iiJ H .S.SPS *o CM CO o CO 4* rr co ■>J< CM t- Ui O CO OO OS IO r-. IO OS i-4 £ p o-2 49^ o> o o o CO o OO O CO o OO o CO CM r- o iO o oo d 4*-° ■H ■H 4^ -H ■ti -H 00 C33 t^ >ra « co ** CD £ OO CM •* CM CO 00 ca CO rt CO y-> CO ^ t- -tfl © CM "3 CD J^ o> o o> o a o i>- d t^ o OS o iA^ CO 4 co ^ CO ^ co 4 co 4 co 4 c3 -u c\s v~— u CO CO (M CO 4o o co ^_ CD i^JS i t-- >fl <35 CO lO CO TfC t^ CO IO 1^ OS »H "3 T3 1 49 o 00 © co O 4 o io o •^ o 4 4 42 u^ C3l "° -H 10 -H 10 ^ o 4 « 4 *° -H o o "3 43 bS »o 00 o co o '5 o_S2 s B OO t-- o ■>* 03 CO >0 CO t- CO CO OS o 493 ■*! O »o o 4 d io o •o d CM O 4-° i- "* -H "* -H -* -H "* -H "* -H "* -H CO o -*1 o oo 44 ,n CO o CM en C33 oJ2 s a O oo 00 OS if» t^ »C OJ 00 o O "5 73 CD cd 49 "3 i^ d CM O iO o 03 o t- ^H io 4 ^-° B> OS 4 o ^ OO 4| oo 4 oo 4 oo 4 «"? £ °J 5o bd « t~ 03 oo CM "35 o42 i OO ■**! CM CO CO CO co T-4 t>. *-§ 1*» t^ OS CO CM 49 o o CO ^* CO O 03 O CO t-i 4 d OS ,-( 4^ C3> *■ -H *• -H co 4J ^ -H *- -H CO 4 "2 rt 5 « >h£ CO OS -i CM CO CO OS CM CO T— Bul. 668] Cotton Irrigation 85 Table 18 shows that the differences in average staple length exceeded one thirty-second of an inch in only a single instance. Using as the "wet" treatments and the "dry" treatments, treatments 1 and 2 in 1932, 1 and 6 in 1933, and 1 and 3 in 1934 and 1935, the average magnitudes of the differences was determined. These were found to be : Bolls and treatments , Difference in staple length, in thirty-seconds 5-lock bolls : of an inch Between treatments 1 and 2, and 1 and 6 0.66 Between treatments 1 and 3 21 4-lock bolls: Between treatments 1 and 2, and 1 and 6 83 Between treatments 1 and 3 0.35 Seed and Lint Determinations. — Computations to determine the ex- tent to which differences in seed and lint indexes noted in table 19 are significant gave the following results : In 1934, 4-lock bolls : There is no significant difference between treat- ments 1 and 2 for 4-lock bolls in 1934. The seed index and the lint index are significantly higher in treatment 1 than in treatment 3. Treatment 2 is significantly higher than treatment 3 in weight of seed cotton, weight of seed, seed index, and lint index, and significantly lower than treatment 3 in number of bolls per pound. In 1934, 5-lock bolls : Treatment 1 is significantly higher than treat- ment 2 in weight of seed cotton and the weight of seed, and significantly lower in number of bolls per pound. Treatments 1 and 2 are both signifi- cantly higher than treatment 3 in weight of seed cotton, weight of seed, seed index, and lint index, and significantly lower in number of bolls per pound. In 1935, 4- and 5-lock bolls : Differences in the treatments are signifi- cant in only a few cases for 1935. Treatment 1 is significantly higher than treatment 2, for 4-lock bolls, in lint percentage and lint index, and, for 5-lock bolls, in lint index only. There was no significant difference between treatments 1 and 3 in the case of 4-lock bolls, and only in the weight of seed in the case of 5-lock bolls. Treatment 3 is significantly higher than treatment 2 in lint percentage; otherwise in 1935 the differ- ences between treatments 2 and 3 were not significant for either 4- or 5-lock bolls. DISCUSSION OF RESULTS AND SUMMARY The investigations reported herein were conducted on the west side of San Joaquin Valley during the years 1926 to 1935. The investigations from 1926 to 1930 were at the United States Cotton Field Station near Shafter and consisted of one year of preliminary 86 University of California — Experiment Station studies and four years of plot and tank experiments, in which there were five principal treatments with plots, and three with tanks. Results of these investigations were published in August, 1932, in Bulletin 537 of the University of California Agricultural Experiment Station, and are in part summarized herein. The investigations in 1931 were on six cotton farms near Pirebaugh and Los Banos. There were no differential irrigation treatments in these investigations, the object being to measure the use of water, determine soil-moisture conditions, and ascertain plant responses in selected 5- to 10-acre portions of commercial cotton plantings. During 1932 and 1933 the experiments were on small plots in a large commercial planting on adobe soil under the second lift of Firebaugh Canal, in Fresno County. The soil of the plots has a moisture equivalent of about 30 to 35 per cent, a permanent wilting percentage of about 16 to 18, and a volume weight of about 84 pounds per cubic foot. There were six principal irrigation treatments, each replicated 9 times, and several supplemental treatments. The irrigation treatments included one (treatment 1) in which an effort was made to maintain the soil mois- ture in the upper 6 feet of soil above the midpoint between the moisture equivalent and the permanent wilting percentage, although the perma- nent wilting percentage was, at times, reached ; two (treatments 2 and 6) with no irrigation, except as needed to keep the cotton plants alive ; one (treatment 3) with early but no late irrigations ; one (treatment 4) with late but no early irrigations; one (treatment 5) following usual local commercial field practice; three (treatments 7, 8, and 9) irrigated, respectively, at 1-, 2-, and 3- week intervals; and three (treatments 10, 11, and 12) irrigated first in midsummer. The investigations in 1934 were confined to three treatments, of which treatment 1 was to have the soil moisture maintained as nearly as prac- tical in the upper third of the range between the moisture equivalent and the permanent wilting percentage, being generally comparable to treatment 1 in 1932 and 1933; treatment 2 was to be allowed to closely approach the permanent wilting percentage; and treatment 3 was to be allowed to undergo severe distress before irrigating, this treatment being comparable to treatments 2 and 6 in 1932 and 1933. The experi- ments were on a triangular area adjoining the third lift of Firebaugh Canal, with Panoche clay loam soil having a volume weight of about 78 pounds per cubic foot, the moisture equivalent ranging from about 24 to about 30 per cent, and the permanent wilting percentage ranging from about 11 to about 15. For each of the three treatments there were 11 replications. All soil-moisture determinations were made on plots not used for measuring plant responses. Bul. 668J Cotton Irrigation 87 The experiments in 1935, which were similar to those carried through in 1934, with the same numbered treatments, were conducted on a rec- tangular area, divided into 60 small plots, about 2.5 miles west of Shafter. The soil is Hesperia fine sandy loam. Volume weight of the soil averaged 82 pounds per cubic foot, moisture equivalents ranged from about 16 to about 24 per cent, and permanent wilting percentages from about 7 to about 9. Twenty-one of the 60 plots were used for soil- moisture samplings, the remaining 39 plots being used for measuring plant responses, with 12 to 14 replications for each treatment. Throughout the eight years of the plot experiments attempts were made to determine the effect of the different irrigation treatments on main-stem length growth (height of plants), number and length of vegetative branches, blossoming, shedding of squares and bolls, yields of seed cotton, and quality of cotton produced. In some cases weights of mature plants were determined, and in 1931 there were some observa- tions on soil cracking and root development. A review of the experi- mental procedure followed led to the elimination of some of the data on plant responses. Furthermore, owing to a degree of soil variability in the experimental area used in 1935, the plant responses for that year show some inconsistencies that tend to raise doubt as to the conclusive- ness of the results. Nevertheless, these results are believed to throw some light on the problems of cotton irrigation, and they are included for completeness of the record. THE SOIL-MOISTURE RECORD, 1927 TO 1935 The soil-moisture record is discussed in some detail for each year, with ■accompanying soil-moisture charts for 1932, 1933, 1934, and 1935. The extent to which the general objective of "wet," "medium," and "dry" 13 treatments was reached may be briefly summarized as follows : In 1927 to 1930 : The soil moisture in the "wet" treatment was gener- ally maintained above the permanent wilting percentage. In the other treatments reported, the permanent wilting percentage was reached or generally approached before irrigations in the upper 2 feet in two treat- ments and in the upper 5 feet in one. In 1932 and 1933 : In both years in "wet" treatment 1, the soil moisture dropped somewhat below the midpoint between the moisture equivalent and the permanent wilting percentage in the upper 2 feet in the summer months, and approached or reached the permanent wilting percentage 13 As previously explained, the terms "wet," "medium," and "dry" as used herein are relative only, and are intended merely to differentiate between the unirrigated treatments and the treatments which received the largest and the intermediate quan- tities of water. The soil moisture in the "wet" treatments was not always continuously maintained above the permanent wilting percentage, although that was the objective in the experiments. 88 University of California — Experiment Station after midsummer, especially in the second foot. But from the third foot to the sixth foot, the soil moisture was above or near the midpoint most of the time, and at no time reached the permanent wilting percentage. In "dry" treatments 2 and 6, in both 1932 and 1933, the soil moisture was at or close to the permanent wilting percentage in the upper 2 feet by September and remained there the rest of the season, except for a rise above the permanent wilting percentage in treatment 2 in 1933 due to irrigation in September. The indicated permanent wilting percentage was not reached in either the third or fourth foot, although it was closely approached in October. Treatment 5, irrigated according to local field practice, maintained a soil moisture approximately the same as treat- ment 1, and can be classed with treatment 1 as a "wet" treatment. Treatments 7, 8, and 9, irrigated at 1-, 2-, and 3-week intervals, had high moisture content at all depths throughout the season. With minor excep- tions, the permanent wilting percentage was not approached in any of these three treatments. After July 22 in 1932 and August 8 in 1933, the permanent wilting percentage was reached in treatment 10 in the first foot before each irrigation in 1932, and before the first 2 irrigations in 1933, the soil moisture in each year closing the season at about the permanent wilting percentage. In the second foot the moisture was below the midpoint between the moisture equivalent and the permanent wilting percentage most of the season, and at times approached or reached the permanent wilting percentage, after about midseason. In the fourth foot, one of the plots at times closely approached the permanent wilting percentage. From a soil-moisture standpoint, treatment 10 can be classed as "medium," although it yielded comparably with "wet" treatments 1 and 5. The soil-moisture record for the other treatments in 1932 and 1933 is not considered in this summary because their only purpose was to obtain information on the effect of withholding either early or late irrigations. In 1934: The objective of maintaining the soil moisture in "wet" treatment 1 in the upper third of the range between the moisture equiva- lent and the permanent wilting percentage was satisfactorily accom- plished. In the fourth and fifth feet it was close to the moisture equivalent throughout the season. In "medium" treatment 2 the permanent wilting percentage was not reached in any sampling. The objective of allowing the soil moisture to closely approach the permanent wilting percentage was not accomplished. In "dry" treatment 3 the permanent wilting per- centage was reached in the first foot and closely approached in the second and third feet before the plots were irrigated early in August, at which time they showed distress. This August irrigation, of 13.5 inches in depth, maintained the soil moisture well above the permanent wilting Bul. 668] Cotton Irrigation 89 percentage below the first foot the remainder of the season. The treat- ment was therefore not a "dry" treatment after early in August. In 1935 : At times the soil moisture dropped below the upper third of the range between the moisture equivalent and the permanent wilting percentage in the upper 2 feet in treatment 1, but the permanent wilting percentage was not reached at any time in the treatment. In treatment 2 the soil moisture was but little below that in treatment 1 and the treat- ment can hardly be classed as "medium." In treatment 3 the soil moisture reached the permanent wilting percentage in the upper foot before irrigation at the end of July and August, and also was at the perma- nent wilting percentage in this depth from late in September through October. There was a drop to the permanent wilting percentage in the second foot in mid-October. In July the soil moisture in the third foot was lower than in treatment 2. Otherwise, the record for the third foot to the sixth foot was not substantially different from that for treatment 2. On the basis of the soil-moisture charts, treatment 3 is classed as "medium" rather than "dry," as intended. PLANT RESPONSES Growth and Yields. — A summary of total main-stem length growth (plant heights) , yields of seed cotton, and weights of cotton plants after harvest is presented in table 20. The number and dates of irrigations and total seasonal applications of water, not including preirrigations, and the number of replications in each treatment, are included to assist in interpreting the data. As a further aid in interpretation, treatments that, according to the soil-moisture record, can be considered as "wet," "medium," or "dry" are so designated. The table does not include results of the experiments at Shafter from 1927 to 1930, or the data from the commercial plantings obtained in 1931, since they are not directly comparable with the results obtained in 1932, 1933, 1934, and 1935. Considering only the principal treatments, as indicated in table 20 : Plots in the experiments in 1932 and 1933 which were irrigated to a total seasonal depth of about 16 to about 20 or 21 inches, not including pre- irrigations, with from 3 to 5 applications of water (plots 1, 5, and 10), yielded at a significantly higher rate than the "dry" plots (plots 2 and 6). Length growths (plant heights) in 1932 were also significantly higher in the former than in the latter. In 1932 "medium" treatment 10, which received less water than treatments 1 and 5, and in which the first irrigation was delayed longest, yielded at the highest rate of all the treatments. This treatment was again "medium" in 1933 and the yield was at about the same rate as in treatment 5. Treatment 10 received more water in 1933 than treatment 5, but the first irrigation was about 3 weeks *Q k CO +a OS 63 r-\ O Gd fcH cq pj CO o i— i fn ^ OS co CO o OS i-H PL, ^Ci co d o W ed a w -i PJ En OS h3 I— 1 pq 3 pj < to cc H ►q ■H s fc cc a> M ~ f^ -»J U, ed o ft >H o « Pj to 3 t3 03 C7J +3 •^ 3 c bo 3 3 . ,. « 3 o> 3 i MT3 ^ (-, a> a! <; o c3 aj (B ^ ^ o M to P 2 a o-i oT a.: T3 03 o.^ a.2 OOO OOO iOC200C)0000<=>0 cioooooooooo -*-tO mtoON^W^iO 00 to to (M tO CO ' CO CO CO •»* o • > bfl s i a a . V 4) .coco ^ 10 (CO o3 a> dj|^io 0) aj 0) n. • o S oj* bo a c s s 3 OOOS«1 Ci«« bi^ (9, S ^ 3 • S so ^Jco jjseo * ei >j bb,* co •- ft a ! S3 co CO 3 S^oso> c3 22 Si* oo o >> >> be 3-3-3 S 3 to"ce~i ©» < .3 3 » * M avS'S OJ 73 3 Si- S3 § NWOlCjH CO *-i ©« CO -^ *0 "to t^. 00 OS O (M Tti . to^caco Bul. 668] Cotton Irrigation 91 later in the former than in the latter. In 1933 the yield in "medium" treatment 10 was not at a significantly lower rate than in treatment 1. If preirrigated to a depth of 5 or 6 feet, it seems apparent that, on the heavy soils such as that of the experiments in 1932 and 1933, 3 or 4 irrigations, beginning in June or July, according to the season, will produce as much cotton as more frequent irrigations. While treatments 7 and 8 (not among the principal treatments), which were the most frequently irrigated and which received the most water, yielded at the highest rate in 1933, they did not yield at the highest rate in 1932. Evidence furnished by these treatments, however, is not conclusive because there were but 2 plots each in 1932 and only 1 plot each in 1933. Yields from treatments 3 and 4 (not among the principal treatments) , one of which was irrigated early but not late and the other late but not early, were lower than from treatments 1, 5, and 10 in 1932, but the differences were less consistent in 1933. Treatment 12 (not among the principal treatments) in 1933 yielded between "wet" treatment 1 and "wet" treatment 5. It received only 2 irrigations, totaling 12.5 inches in depth, with the first irrigation on August 8. On the other hand, treatment 11, in 1932, with 3 irrigations totaling 8.6 inches in depth, yielded only as much as "dry" treatment 6, and slightly more than "dry" treatment 2. Thus treatments 11 and 12 furnish some evidence that with heavy soil, well preirrigated, 2 or 3 irrigations applied after July will produce a yield comparable with that obtained with irrigations beginning earlier. A definite conclusion to this effect, however, does not seem justified. As to differences in the three treatments in 1934 and 1935 : There was a clearer differentiation in the number of irrigations and the amounts of water applied than in the soil-moisture records. In main-stem length growth (plant height) and in weight per plant the differences were in the same order in 1934 as the differences in irrigations. Differences in yield from the treatments were not significant in either year. Considering the principal plant responses as a whole, for the years 1932 to 1935, it is clear that allowing the soil moisture to remain at the permanent wilting percentage for extended periods, such as in treat- ments 2 and 6 in 1932 and 1933 (figs. 9, 10, 17, and 18), in the upper 2 or 3 feet of soil will reduce plant heights and yields. Allowing the soil moisture to reach the permanent wilting percentage and remain there for extended periods during the last portion of the season, as in treat- ment 3 in 1932 and 1933 (figs. 11 and 12), will have the same effect. The effect of withholding the first irrigation until August, as in treatments 4, 11, and 12 (figs. 13, 14, 24, and 25) is not conclusively shown. There 92 University of California — Experiment Station was a reduction in yield, as compared to the "wet" treatments, in treat- ments 4 and 11 in 1932, but not in treatments 4 and 12 in 1933. As between the principal "wet" and "medium" treatments, the yields of neither group were consistently higher or lower than those of the other. Blossom Count, 1932. — Counts of blossoms were made each year, but only the 1932 data are considered sufficiently complete to be included. The results for 1932 are shown graphically in figure 33. The first blos- soming peak occurred simultaneously for all treatments, and prior to this first peak the number of blossoms appearing daily did not show much variation. After this peak, the unirrigated or "dry" treatments, and treatment 4, which was irrigated first in August, fall into a low group ; "wet" treatment 1 and treatment 3, which was irrigated only in the first part of the season, fall into a middle group ; and treatment 5, another "wet" treatment, showed the highest blossom count of all treatments. A blossom count for a single year cannot, however, be taken as conclusive. Sheds of Bolls and Squares, 1935. — Sheds of bolls and squares were also counted each year, but only the 1935 record is included. The largest number of sheds of both bolls and squares was in the treatment which received the least water. The smallest number of sheds of bolls was in the treatment which received the most water, and the smallest number of sheds of squares was in treatment 2, for which the soil-moisture record is not much different from that for the treatment which received the most water. Quality of Cotton. — Some data are included regarding the effect of irrigation on the quality of the cotton produced. Certain differences were found which are statistically significant, but, even as between "wet" and "dry" treatments, differences, either in grade or staple length, do not appear to be of sufficient magnitude to be of economic importance. Because of the small number of samples from each treatment that were classified, however, the data presented on the relation of soil moisture to cotton quality are not considered to be conclusive. WATER APPLIED AND WATER USED The quantities of water applied on portions of six commercial cotton farms were measured in 1931. Applications ranged from 12.4 to 21.6 inches in depth, not including preirrigations. These quantities were applied in 3 or 4 irrigations. Quantities of water applied under the various treatments in the experiments from 1932 to 1935 were also measured, and the use or loss of water, including both surface evapora- tion and plant transpiration, was determined from the soil-moisture records. The average quantities of water applied to the principal treat- ments in which an effort, not wholly successful, was made to prevent the Bul. 668] Cotton Irrigation 93 soil moisture from dropping to the permanent wilting percentage (treat- ments 1 and 5 in 1932 and 1933 and treatment 1 in 1934 and 1935) ranged from 15.9 to 20.8 inches in depth on the heavy soils on which the experi- ments were conducted in 1932 and 1933, and from 33.07 to 42.78 inches in depth on the lighter soils in 1934 and 1935, all exclusive of pre- irrigations. The quantities stated were applied in from 3 to 5 irrigations in 1932 and 1933, whereas treatment 1 received 7 irrigations in both 1934 and 1935. Computations of the water used or lost, as determined from the soil- moisture records, were confined to the treatments in which soil-moisture samplings were sufficiently frequent to make the record of use or loss reasonably complete. The data are presented graphically and numer- ically in figures 26, 27, 36, and 40. The conclusion is reached that, on either the heavy soils of the experiments in 1932 and 1933 or the lighter soils of the experiments in 1934 and 1935, cotton plants in San Joaquin Valley will use the equivalent of about 24 inches in depth in producing normal yields of cotton, this use including both surface evaporation and plant transpiration. The experiments do not disclose whether the use would be greater than about 24 inches in depth when the cotton is grown on lands of exceptional fertility that might be expected to produce larger plants than were produced in the experiments described. The quantity of water used is, of course, less than it is necessary to apply, since some losses, in addition to surface evaporation, are in practice unavoidable. However, the data as to water applied, which are sum- marized in table 20, indicate that the application of a depth of 15 to 20 inches of water, for heavy soils, and 20 to 25 inches, on the medium and lighter soils, is ample for cotton in San Joaquin Valley if there has been a preirrigation to a depth of 5 or 6 feet of soil. The depth of water it is necessary to apply to wet the soil to 5 or 6 feet, prior to seeding, will depend in part on the depth to which the soil has been wet by the winter rainfall. If these soils are dry to a depth of 6 feet, the maximum preirri- gation required should not be more than a depth of about 14 inches for the heavier soils or about 11 inches for the medium or lighter soils. lOm-8, '42(9781)