UNIVERSITY OF CALIFORNIA COLLEGE OF AGRICULTURE AGRICULTURAL EXPERIMENT STATION BERKELEY, CALIFORNIA Water Requirements of Cotton on Sandy Loam Soils in Southern San Joaquin Valley S. H. BECKETT AND CARROLL F. DUNSHEE BULLETIN 537 August, 1932 UNIVERSITY OF CALIFORNIA PRINTING OFFICE BERKELEY, CALIFORNIA CONTENTS PAGE Introduction 3 General description of the area 4 Soil types of the area 5 Water supply 6 Climate 6 Definition of terms 7 Location of experimental plots and methods of procedure 8 Size of plots 9 Irrigation and measurement of water 10 Preparation of seed bed and seeding 11 Soil sampling and computation 11 Preliminary observations, season of 1926 12 Conclusions from preliminary observations 13 Field-plot investigations during 1927, 1928, 1929, and 1930 13 Results of soil-moisture observations 15 Seasonal use of water by cotton 26 Growth and yield 28 Effect of irrigation on size of plants 28 Effect of irrigation on flowering 29 Effect of irrigation on shedding and number of bolls 31 Effect of irrigation on length of fiber and lint index 32 Effect of irrigation on yields 33 Use of water and yields of cotton grown in tanks 33 Purpose of experiment 33 Description of tanks and equipment 35 Irrigation treatments 36 Procedure 37 Use of water by cotton grown in tanks 37 Flowering records and yields of cotton grown in tanks 43 Conclusions.... 44 WATER REQUIREMENTS OF COTTON ON SANDY LOAM SOILS IN SOUTHERN SAN JOAQUIN VALLEY 1 S. H. BECKETT2 and CARROLL F. DUNSHEE3 INTRODUCTION Cotton has been produced on a commercial scale in the San Joaquin Valley of California since 1918, when 1,100 acres of cotton was reported. In 1930, 270,000 acres was harvested; and cotton is now one of the major crops in the Valley. Throughout the cotton-producing area of San Joaquin Valley, production depends almost entirely upon irrigation; and, because of wide variation in the soil types upon which the crop is grown, the irrigation problem is complex. Methods and practices of irrigation adapted to one locality may not necessarily apply to another locality where a different soil type predominates. This bulletin presents the results of a five-year study dealing principally with the irrigation requirements of cotton grown on the sandy loam soils of the Delano series, and with the effects of soil- moisture deficiency on growth and yields. The investigations described covered the years 1926 to 1930, inclu- sive, and were conducted at the United States Cotton Field Station at Shafter, California. 4 i Received for publication May 10, 1932. 2 Professor of Irrigation Investigations and Practice and Irrigation Engineer in the Experiment Station. s Assistant Crop Irrigationist in the Experiment Station ; resigned June 30, 1929. 4 During the first three and one-half years, the field work on which this report is based was in charge of Carroll F. Dunshee, who, with the senior author, organized the work and carried the responsibility of planning the experiments and successfully conducting them through that period. In 1929 and 1930 the field work was conducted by L. C. Schultz, Assistant Crop Irrigationist in the Experiment Station. The senior author has been associated with the study in an advisory capacity from the begin- ning. During 1926, 1927, and 1928, valuable advice and assistance in planning and carrying out the field and laboratory work were received from W. B. Camp, then Agronomist and Superintendent of the Cotton Field Station at Shafter. His successors, Fred W. Herbert and J. E. Hite, continued the splendid cooperation during the last two years of the investigations. John W. Carter, Assistant Scientific Aid; W. E. Clark, Junior Agronomist; and Joe F. Schuh, Assistant Scientific Aid, all of the Cotton Field Station, gave material help in the detailed field work of the experiments. University of California — Experiment Station GENERAL DESCRIPTION OF THE AREA Cotton growing in the San Joaquin Valley is confined principally to the six most southerly counties (Kern, Tulare, Kings, Madera, Fresno, and Merced), with a limited area in Stanislaus and San Joaquin counties. Of the 270,000 acres harvested in California in 1930, 88 per cent was in the San Joaquin Valley. However, the cotton acreage in California is less than 1 per cent of the total for the United States. TABLE 1 Area of Cropped Lands and Area of Cotton Harvested in 1930 in Each of the Counties in the Cotton-Producing Areas of California, with the Average Yields per Acre and the Gross Production in Bales District or county Area of cropped land* Area of cotton harvested in 1930| Average yield of cotton fiber per acre in 1930t Gross production in 1930t acres 316,900 201,600 458,300 83,300 236,300 138,000 675,300 acres 63,400 51,300 47,000 26,800 21,600 25,200 2,900 238,200 pounds 484 602 479 450 354 532 217 495 bales (500 pounds) 64 , 090 Kern County 64,585 47,066 25,200 15,962 Kings County 28,012 1,316 2,109,700 246,231 Palo Verde Valley 32,300 16,000 15,900 1,500 286 392 295 9,517 1,229 17,400 10,746 Imperial Valley 424,100 5,100 8,600 5,800 125 375 225 2,239 Bard area 4,541 429,200 14,400 6,780 270,000 468 263,757 * Land classification by State Engineer, 1929. t Report of California Cooperative Crop Reporting Service, 1930. Table 1 shows the areas of cropped land and the areas of cotton harvested in the cotton-producing counties or districts in California in 1930, the average cotton yields per acre, and the gross production in bales. Figure 1 shows the general location of the principal cotton- producing areas in the San Joaquin Valley. Bul. 537] Water Requirements of Cotton on Sandy Soils 5 Soil Types of the Area. — Throughout the cotton-producing areas of the San Joaquin Valley a wide variation in soil type is found, ranging from the light sandy loams and fine sandy loams of the Hanford, Delano, Fresno, and San Joaquin series to the extremely heavy clay adobes of the Stockton and Panoche series. Although Fig. 1. — Location of the principal cotton-producing areas in the San Joaquin Valley, 1930. Each dot represents 1,000 acres. information is not available for an accurate estimate of the cotton acreage on each of the soil types, a general survey of the area, indi- cates that production on the sandier types amounts to at least 75 per cent of the total. In the Bakersfielcl, McFarland, Wasco, and Hanford areas, the Hanford and Delano fine sandy loams and sandy loams 6 University of California — Experiment Station are the most prominent. In the Madera and Fresno areas, the sandy loams and loams of the San Joaquin, Madera, Fresno, and Hanford series are the most common. The heavier soil types upon which cotton is extensively grown are found in the lands bordering the Buena Vista and Tulare lake beds and in the area extending along the west side of the central part of the valley. In the Buena Vista and Tulare lake areas, the heavy soils are principally the loams, clay loams, and clays of the Tulare and Pond series. On the west side of the valley, including the Tran- quillity, Firebaugh, and Dos Palos areas, the heavy types include the Stockton and Fresno loams and clay loams, the Merced clays, and the very heavy Stockton and Panoche clay adobes. Water Supply. — The water supply for the cotton area of the San Joaquin Valley is obtained by gravity and by pumping. The principal gravity supplies come from the Kern, Kaweah, Kings, San Joaquin, and Merced rivers. The principal pumped areas are found bordering Buena Vista and Tulare lakes in Kern and Kings counties; in the vicinity of Arvin, Shafter, McFarland, Wasco, and Delano in Kern County; in the Tulare and Tipton areas of Tulare County; and above the gravity canals on the west side of the valley in Fresno and Merced counties. Numerous pumping plants are also found in the areas generally covered by the gravity supplies. In practically all of the main pumped areas, the underground water level has been decidedly lowered from year to year. Climate. — The climate is characterized by dry, hot summers and limited rainfall, practically all occurring between harvest and plant- ing. The lower and flatter parts of the valley are most frequently visited by frosts, the usual frost period extending through, the months of December, January, and February. The highest and lowest temperatures recorded in the cotton area of the San Joaquin Valley are 118° and 13° Fahrenheit. Temperatures of 105° to 110° F are not unusual. Table 2 gives a comparison of temperatures and rainfall, and table 3 shows the general frost data at different points in the valley, both being taken from the annual summaries of the United States Weather Bureau. Bul. 537] Water Requirements of Cotton on Sandy Soils TABLE 2 Annual Temperatures and Precipitation at Various Centers of the Cotton-Producing- Areas of the San Joaquin Valley* Length of record Temperature 1 Seasonal rainfall Locality Maximum Minimum Mean Maximum Minimum Average Bakersfield Tulare years 33 40 41 50 44 23 39 °F 118 114 115 116 111 113 110 °F 13 18 17 16 15 14 20 °F 64.6 62.4 63 62.2 63.8 61.6 inches 9 30 13.70 19.45 22 08 19.04 12.51 14.41 inches 2.77 3.07 4.96 3.20 3.58 5.57 3.24 inches 5 71 8 12 9 82 Merced 11.17 10.55 8 77 Los Banos 8.52 * U. S. Dept. Agr. Summary of climatological data for the United States. Section 14. — Central California. TABLE 3 Dates of Killing Frosts and Length of Growing Season at Different Localities in the San Joaquin Valley* Locality Length of record Average date of last killing frost in spring Average date of first killing frost in autumn Average length of growing season Latest date of killing frost in spring Earliest date of killing frost in autumn Bakersfield Tulare years 21 12 26 25 14 24 16 February 22 March 13 February 14 March 1 February 1 March 7 January 16 November 25 November 18 November 19 November 20 December 5 November 13 December 6 days 276 250 288 264 307 251 324 April 1 April 11 April 13 April 29 March 1 May 26 February 18 October 21 October 20 Fresno October 31 Merced October 27 Modesto November 11 October 19 * U. S. Dept. Agr. California. Summary of climatological data for the United States. Section 14.— Central DEFINITION OF TERMS In presenting the results of these investigations, reference is fre- quently made to a number of terms used in irrigation practice. The most common of these are here denned : 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. Apparent specific gravity: the oven-dry weight of a volume of undisturbed soil divided by the weight of an equal volume of water. Moisture equivalent : the amount of water retained in a soil against a centrifugal force 1,000 times gravity. It is expressed as a per- 8 University of California — Experiment Station centage of the oven-dry weight of the soil, and is a laboratory 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 unrestricted 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: 5 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. Temporary wilting percentage: the moisture percentage in the soil at which the leaves of the plant lose their turgidity for short periods of time, but revive without the application of water. Efficiency of irrigation: the percentage of the water applied to the land that is accounted for in soil-moisture increase in the soil mass occupied by the principal rooting system of the plant. Redaction of moisture percentages to acre-inches of water: in reducing moisture percentages to equivalent losses in acre-inches per acre, the formula D = is used, where P represents the moisture percentage, v the apparent specific gravity of the soil (often referred to as the volume weight), d the depth of soil in inches, and D the equivalent depth of water in acre-inches per acre. LOCATION OF EXPERIMENTAL PLOTS AND METHODS OF PROCEDURE In starting the investigation reported herein, it was decided that more information of value could be obtained by an intensive study of the effects of variable irrigation treatments given to plots of limited areas than by less complete studies in commercial fields. The advantages of the smaller field plots are that they may be selected for uniformity in depth and type of soil; that they may be more evenly leveled ; that water may be more evenly applied to them than to larger areas ; and, finally, that they permit a more accurate record of seasonal soil-moisture conditions and fluctuations. The work was centered at the United States Cotton Field Station at Shafter in Kern County because of the splendid facilities offered 5 Veihmeyer, F. J., and A. H. Hendrickson. Essentials of irrigation and culti- vation of orchards. California Agr. Ext. C'ir. 50:1-24. 1930. Bul. 537] Water Requirements of Cotton on Sandy Soils i) there. This station, located about 20 miles northwest of Bakersfield, is in one of the principal cotton-producing' areas of the southern San Joaquin Valley. The soil where the experimental plots were located is classed as a Delano sandy loam, typical of much of the plains area northwest and southeast of Bakersfield. Cotton is one of the principal crops produced in the area occupied by this soil type. Delano sandy loam is grayish-brown and gritty, contains some mica, and is fairly uniform in texture to a depth of 5 feet. Below 5 feet is a layer of coarse sand of varying thickness. This type, though low in organic matter, is very fertile and easily cultivated, and takes water readily. Table 4 gives the principal characteristics of this soil as found at the Cotton Field Station. TABLE 4 Characteristics of Delano Sandy Loams at United States Cotton Field Station on Which Plot Experiments Were Conducted Depth of soil Apparent specific gravity Moisture equivalent Observed field capacity Permanent wilting percentage Water-holding capacity between permanent wilting percentage and field capacity 1.42 1.50 1.53 1.55 1.51 1.47 1.50 per cent 10.11 8.78 10.36 10.52 9.08 5.85 10 10 per cent 11.00 per cent 3.50 acre-inches per acre-foot of soil Third foot Fifth foot Average for first 5 feet 1.35 Irrigation water is obtained by pumping from two wells, one 12 inches and the other 14 inches in diameter. The pumps discharge into an earthen reservoir of about 1 acre-foot capacity. Distribution from this reservoir is made through a concrete-pipe system equipped with necessary gates and valves for constant-flow regulation. Throughout the five years of the experiment, irrigation water was obtainable on demand and in any desired quantity. Size of Plots.— During 1926, plots 40 feet wide and 300 feet long were used, each containing 9 rows of cotton spaced 48 inches apart. Beginning in 1927, however, the size of all plots was reduced to 16 by 90 feet, and the number of replications increased to nine for each of the major treatments. Four rows 48 inches apart were seeded in each plot, and the cotton was chopped to the same number of plants 10 University of California — Experiment Station in each row, with as nearly as possible an equal spacing of 12 inches between plants. In chopping, special care was taken to leave only normal plants. Soil samples, and all counts, measurements, and yield data were taken each year (including 1926) from the two middle rows of the plots, the outside rows being left as guards. Fig. 2. — A field ditch and rectangular weirs used in delivering and measuring water to cotton plots. Irrigation and Measurement of Water. — Water was delivered to each of the plots either directly from the concrete pipe line or from temporary field ditches. This arrangement permitted uniform application of any desired depth of water to the plots. All water applied to the plots was measured through 1-foot rec- tangular weirs, installed in the field ditches extending along the upper Bul. 537] Water Requirements of Cotton on Sandy Soils 11 end of each tier of plots. Weir measurements of the flow were made at 5-minute intervals during" the irrigation of each plot. Figure 2 shows a field ditch with weirs used in delivering and measuring water to the plots. In each irrigation, including that given before planting, sufficient water was added to bring the soil in each plot to its full field capacity, to a depth of at least 5 feet. Preparation of Seed Bed and Seeding. — Recognized good practices were used in preparing the seed bed and in seeding. Each year selected seed of the Acala variety was used, and a standard width of planting of 48 inches between rows was maintained. All plots were irrigated by flooding; and, while the plants were small, each irriga- tion was followed by a cultivation. As the season advanced, the size of the plants necessitated discontinuance of cultivation. Where the condition of the cotton plant, as indicated by change in color of the foliage or wilting, was used as a basis for determining the time to irrigate, reliance was placed on the judgment of qualified authorities of the Cotton Field Station. Soil Sampling and Computation. — Soil samples were taken on each plot to a depth of 5 feet before and after every irrigation, and a record was made of the quantity of water applied each time. Frequent measurements were also made of the average depth of moisture pene- tration when different depths of water were applied, and of the variation in depth of penetration through the length of the border checks. All soil samples were taken with the improved soil tube. 6 In obtaining the soil samples, the top 3 inches of soil was removed, and the samples were taken from definitely established sampling points at 1-foot intervals, to a depth of 5 feet, Standard methods were used in weighing and drying the soil samples and in computation of mois- ture percentages. Soil samples were taken at the beginning and end of the season as well as before and after each irrigation. From the moisture percentages thus obtained, the amounts of water in acre- inches per acre from each foot of soil were computed by using the Pvd previously discussed formula D = 1 ; the total loss was then reduced to equivalent losses for 30-day periods, The method of cal- culation may perhaps be made clearer by showing the detailed steps in a particular case. To obtain the average rate of loss for treatment 6 Veihmeyer, F. J. An improved soil sampling tube. Soil Science 27:147- 152. 1929. 12 University of California — Experiment Station 1, plot 23, from July 6 to July 14, for example, the average moisture content for each foot of depth on July 6 and July 14 is obtained from table 5, and the percentage loss and equivalent loss in acre-inches per acre calculated as shown in the accompanying table. The average rate of loss per 30 days can then be calculated from the total equiva- lent loss in acre-inches per acre from July 6 to July 14 (8 days) : 2.56X30 Q _ n = 9.60 acre-inches per acre. o These losses were next plotted, and a consumptive use-of-water curve for the season was obtained. The average consumptive use of water for each month was then taken directly from the curve. First foot Second foot Third foot Fourth foot Fifth foot Total Average moisture content, July 6, per 10 2 5.4 4.8 1.50 0.86 9.6 5.8 3.8 1 50 0.68 10 7 7.8 2.9 1.50 0.52 11.9 9.2 2.7 1 50 0.48 10 4 10 3 0.1 1 50 0.02 Average moisture content, July 14, per Loss, July 6 to July 14, per cent Equivalent loss, acre-inches per acre*.... 2.56 * Calculated for each foot by the formula D acre-inch per acre. Pvd . ., -. ., ++u . .,. 4.8x1.50x12 — — ; for the first foot this would be — — 0.86 PRELIMINARY OBSERVATIONS, SEASON OF 1926 The season of 1926 was spent in studying the effects of a limited number of treatments on plots at the Cotton Field Station, and in observing the results of different irrigation practices on commercial fields in the southern San Joaquin Valley. Early in April, four border plots were set aside for these pre- liminary studies, and the following irrigation treatments were outlined : Treatment A (duplicate plots) : To be irrigated when the plants showed by their change from a light green to a dark bluish-green color that they were nearing the wilting stage. This treatment was expected to produce normal growth and yields. Treatment B (single plot) : To be irrigated when the plants showed definite indications of wilt at 4 p.m. Treatment C (single plot) : To be irrigated when the plants showed definite indications of wilt at 9 a.m. This is a more severe treatment than treatment B. Buii. 537] Water Requirements of Cotton on Sandy Soils 13 At different times during" the season, counts were made of the shedding of the squares and of the number of bolls set on twelve representative plants under the 4 p.m. and 9 a.m. wilt treatments. At the end of the season, comparative yield records were obtained from these two plots. Conclusions from Preliminary Observations. — All the growers interviewed, as well as those interested in investigational work in cotton production, believed proper irrigation to be one of the most important factors affecting the yield and the quality of the product. The field observations and plot studies in 1926 led to the following conclusions : 1. Cotton is sensitive to soil-moisture fluctuations; one of the causes of excessive shedding of the squares is unfavorable soil-moisture conditions. 2. The average yield of cotton in the upper San Joaquin Valley was reduced by an amount estimated at from 10 to 15 per cent because of improper irrigation in the latter part of the season. 3. Poor preparation of land for irrigation very often accounts for the ' ' spotted ' ' condition of growth and yields. 4. Where the water is limited, there is a tendency to serve too large an acreage with the supply at hand. The Delano sandy loams at the Shafter Station have a field capa- city of about 11 per cent and a temporary wilting percentage of from 4.0 to 4.5 per cent. Permanent wilting occurred at 3.5 per cent. With the soil at the permanent wilting percentage, 1.35 acre-inches of water per acre was required to bring 1 foot depth of soil to its field capacity. A definite difference was shown in shedding, number of bolls, and yields in the plots irrigated when the plants wilted at 4 p.m. and at 9 a.m. The 4 p.m. -wilt plots had 11.3 per cent less shedding, produced 13.7 per cent more bolls, and yielded 15 per cent more cotton. FIELD-PLOT INVESTIGATIONS DURING 1927, 1928, 1929, AND 1930 After the preliminary observations of 1926, the investigations were limited to field plots and to tanks located at the Cotton Field Station at Shafter. The tank layout and the procedure and results obtained are discussed in a later section of this bulletin. From 1927 to 1930 the method of measuring the amounts of water applied to the plots, the preparation of the seed bed, the selection of seed, and the seeding, were substantially the same as during 1926. 14 University of California — Experiment Station In each of the four years the following irrigation treatments were given : Treatment 1 : Irrigated when the average moisture content of the top 5 feet of soil reached approximately 7 per cent. This is about halfway between field capacity and the permanent wilting percentage. It was anticipated that if the average moisture content were main- tained above this figure, readily available moisture would be present in the entire soil depth throughout the season, and that at no time during the season would the plants show indications of approaching the wilting stage. Treatment 2: Irrigated when the plants wilted at 4 p.m. Mid- afternoon wilt shows a lack of available moisture in the soil ; a change in color of the foliage from a light green to a dark bluish-green indicates that the plants are approaching wilting. Treatment 3 : Irrigated when the plants wilted at 9 a.m. Early- morning wilt was considered to indicate definite distress in the plant resulting from a lack of soil moisture. Treatment 4: Irrigated as in treatment 1 until midseason (about July 20), after which water was not applied until the plants wilted at 9 a.m. It was anticipated that in this treatment an abundance of soil moisture would be available during the first half of the season, with a definite deficiency during the last half. Treatment 5 : Irrigated during the first half of the season (until about July 20) when the plants wilted at 9 a.m., as in treatment 3, after which it was irrigated as in treatment 1. Treatment 5 thus provided for a deficiency in available moisture during the first half of the season and no deficiency during the second half. Treatment 1a : In this treatment, introduced during the 1930 season, water was added when the plants changed from a light green to a dark bluish-green color. This change in color usually occurs several days in advance of wilting and is the first indication given by the plant of soil-moisture shortage. The total depth of water applied during the season was intermediate between the amounts applied in treatments 1 and 2. Treatment Iaa: This, the second new treatment added in 1930, was planned as an intermediate treatment between 1 and 1a, with the water applied when the moisture content of the soil, to a depth of 5 feet, had reached an average of 5.5 to 6.0 per cent. Figure 3 shows the general arrangement of the plots with the plot number and irrigation treatment. Substantially this same arrange- Bul. 537] Water Requirements of Cotton on Sandy Soils 15 ment was followed during 1927, 1928, 1929, and 1930. This arrange- ment provided for nine replications of treatments 1 and 2, eight of treatment 3, and five of treatments 4 and 5. In 1930 five new plots were added to take care of treatment 1a, and four for treatment Iaa. Waste check 26 Waste check 13 Tank Yard © 87 ® 14 ' ® T / ® 88 ® 15 © 2 ® 89 © 16 © 3 © 30 ® 17 © 4 © 31 ® * 18 © 5 ® 32 , ® * 19 © 6 lt ® 33 © % 20 3 © 7 ® 34 © * 21 ® 8 ® 35 © * 22 ® 9 ® 36 ® * 23 © /O ® 37 ® & 2* © // ® 38 ® * 25 © 12 ~we/r fcweir Fig. 3. — Arrangement of field plots, delivery ditches, and' weirs, 1927, 1928, and 1929. The stars indicate plots in which soil samples were taken. The treat- ments, indicated by the numbers in circles, were as follows: (1) irrigated when moisture content reached about 7 per cent; (2) irrigated when plants wilted at 4 p.m.; (3) irrigated when plants wilted at 9 A.M. ; (4) high moisture content early in season, low last half of season; (5) low moisture content early in season, high last half of season. Results of Soil-Moisture Observations. — Throughout each of the four seasons, soil samples were taken to a depth of 5 feet before and after each irrigation at four fixed locations in duplicate plots under treatments 1, 2, and 3, and in single plots under treatments 4 and 5, as well as from duplicate plots in treatment 1a during 1930. The methods of summarizing the data collected are illustrated in tables 5 and 6. Table 5 summarizes the average moisture content of plot 23 at each sampling, with the dates and quantities of irrigation water applied to it during 1927. In table 6 the moisture percentages obtained from plots 19 and 23 (both given treatment 1) have been reduced to acre-inches of water per acre taken by the growing crop from each foot in depth of soil ; similar data were obtained from each of the plots sampled during the four-year period. Soil-moisture deter- minations were made on approximately 3,000 samples each year. Tables 7 to 10 show the rates of use of water in the intervals between irrigations for each treatment during each of the four years, calculated from the moisture percentages found. 16 University of California — Experiment Station TABLE 5 Average Moisture Content and Dates and Amounts of Irrigation for Treatment 1, Plot 23 ; May 11 to November 1, 1927 Dates of sampling May 11 June 2 June 6 June 21 June 24 July 2 July 6 July 14 July 18 July 25 July 28 August 9 August 12 August 23 August 27 September 13 September 16 September 29 October 3 November 1*. Total Average moisture content of soil First foot 10 10 per cent 8.1 6 2 2 3 6 6 1 10.2 5.4 9.4 5 2 10.9 5.2 9 8 5.4 10 8 5.4 10 7 6.8 11.0 10.7 Second foot per cent 8 7.7 10 8.5 10 8 7.4 9.6 5.8 1 7 4 7 4 3 9.4 5.1 10.1 7.2 10 4 6.7 Third foot per cent 9 5 9 3 11.0 98 11 5 9.7 10.7 7.8 9.7 6.4 9.8 5.5 9.8 5.7 7.5 5.2 10.0 7 3 10.8 7.7 Fourth foot per cent 9.2 10.1 11.3 10 3 11.6 10 11 9 9.2 9 2 8.1 9.2 6.7 9.2 6.6 6.6 4.8 5.8 6.7 10.5 8.1 Fifth foot per cent 9.1 9.2 9 3 9.2 10.4 10.1 10.4 10 3 9.4 9.1 8.7 7.1 7.5 6 2 5.9 4.8 4.3 4.6 8.5 7 1 Dates of irrigation June 2 June 22 July 4 July 15 July 26 August 9 August 25 September 14 September 30 Amount of irrigation water applied acre-inches per acre 4 3 3 4 5 5 5 5 4 38 * 1.35 inches of rain on October 29 and 30. Bul. 537] Water Requirements of Cotton on Sandy Soils 17 TABLE 6 Quantities of Water Used in Intervals Between Irrigations on Treatment 1, Plots 19 and 23; April 10 to November 1, 1927 Interval Num- ber of days Soil-moisture loss, acre-inches per acre First foot Second foot Third foot Fourth foot Fifth foot Total Equivalent loss in 30 days Plot April 10 to May 11 May 11 to June 2 June 6 to June 21 June 24 to July 2 July 6 to July 14 July 18 to July 25 July 28 to August 9 August 12 to August 23 August 27 to September 13 September 16 to September 29 October 3 to November 1 25 25 0.88 0.90 94 0.70 1.06 0.90 0.86 70 0.82 0.11 0.11 0.65 0.61 0.79 0.63 0.77 0.68 0.88 0.48 0.68 0.11 0.09 0.18 0.41 0.50 0.70 0.75 0.66 0.74 0.36 49 09 0.11 31 0.20 0.31 0.20 0.61 56 0.40 0.27 0.61 0.11 14 12 14 0.38 31 0.70 0.23 0.18 47 2 2 2 2 3.89 3.03 3.06 1.81 3.07 54 65 95 4 28 8.46 10.95 10.90 9.70 8.26 5.40 4.18 3.18 Plot 23 May 11 to June 2 June 6 to June 21 June 24 to July 2 July 6 to July 14 July 18 to July 25 July 28 to August 9 August 12 to August 23 August 27 to September 13. September 16 to September October 3 to November 1 34 70 0.81 0.86 0.76 1.03 79 0.97 70 0.85 06 27 61 0.68 0.79 1 03 0.74 0.77 0.52 0.67 04 0.22 0.32 0.52 0.60 0.77 0.74 41 0.48 0.56 43 0.18 0.02 0.29 0.05 48 02 0.20 0.05 0.45 29 0.47 0.23 32 20 25 44 1.39 2.08 2 56 2.40 3 57 2.97 2.67 1.70 2.76 60 2.78 7.80 9.60 10 30 8.92 8.10 4.71 3.92 2 86 18 University of California — Experiment Station table 7 Rates of Use of Water in Intervals Between Irrigations in Treatments 2, 3, 4, and 5; Season of 1927 Interval Num- ber of days Acre-inches per acre Soil- moisture Equivalent loss in 30 days Interval Num- ber of days Acre-inches per acre Soil- moisture loss Equivalent loss in 30 days Treatment 2 Plot 20 April 10 to May 11 May 11 to June June 9 to July July 13 to July 26 July 28 to Aug. 18 Aug. 20 to Sept. 1 Sept. 3 to Sept. 26 Sept. 29 to Nov. 1 31 57 22 0.49 29 3.63 13 3.59 21 4.92 12 2.24 23 3.35 33 2.46 0.55 0.67 3.75 8.28 7.03 5.60 4.37 2.24 Plot 24 May June July July Aug. 11 to June 2 9 to July 8 11 to July 21 23 to Aug. 1 19 to Sept. 3 Sept. 6 to Sept. 27 Oct. 1 to Nov. 1 0.67 3.46 2.16 3.91 1.65 3.53 2.23 0.91 3.58 6.48 6.18 3.30 5.02 2.16 Treatment 3 Plot 21 Plot 25 April 10 to May 11 31 0.61 0.59 May 11 to June 2 22 0.67 0.91 May 11 to June 2 22 0.77 1 05 June 9 to July 15 36 5.13 4.28 June 9 to July 15 36 4.79 4 00 July 18 to Aug. 6 19 4.08 6.44 July 18 to Aug. 6 19 3.45 5 45 Aug. 9 to Aug. 30 21 4.34 6.20 Aug. 9 to Aug. 27 18 3.42 5.60 Sept. 2 to Sept. 27 25 3.24 3.89 Aug. 30 to Sept. 27 28 3.35 3.59 Oct. 1 to Nov. 1 31 2.93 2.84 Oct. 1 to Nov. 1 31 1.84 1.78 Treatment 4, plot 18 Treatment 5, plot 22 April 10 to May 11 31 0.56 0.54 May 11 to June 2 22 0.97 1.32 May 11 to June 2 22 0.59 0.80 June 6 to June 22 16 2.40 4.50 June 2 to June 21 19 0.87 1.37 June 25 to July 8 13 2.42 5.58 June 25 to July 18 23 5.09 6.64 July 11 to July 20 9 2.84 9.46 July 20 to Aug. 12 23 4.35 5.68 July 23 to Aug. 17 25 5.76 6.90 Aug. 16 to Aug. 27 11 2.55 6.95 Sept. 6 to Sept. 26 20 2.29 3.43 Aug. 30 to Sept. 15 16 2.19 4.10 Sept. 29 to Nov. 1 33 2.33 2 12 Sept. 15 to Oct. 10 25 1.67 2.01 Oct. 14 to Nov. 1 18 1.92 3.20 Bul. 537] Water Requirements op Cotton on Sandy Soils 19 TABLE 8 Rates of Use of Water in Intervals Between Irrigations in Treatments 1, 2, 3, 4, and 5; Season of 1928 Num- ber of days Acre-inches per acre Interval Num- ber of days Acre-inches per acre Interval Soil- moisture loss Equivalent loss in 30 days Soil- moisture loss Equivalent loss in 30 days Treatment 1, plots 22 and 26 Treatment 2, plots 23 and 27 April 19 to June 6 48 1.09 0.68 April 19 to June 6 48 1.22 0.76 June 11 to July 2 21 3.37 4.82 June 6 to July 6 30 2.76 2.76 July 6 to July 14 8 1.69 6.34 July 10 to July 23 13 2.58 5.95 July 19 to July 26 7 1.68 7.21 July 28 to Aug. 9 12 3.36 8.42 July 28 to Aug. 6 9 3 03 10.10 Aug. 13 to Aug. 27 13 2.10 4.85 Aug. 9 to Aug. 17 8 1.98 7.44 Sept. 3 to Nov. 1 59 5.26 2.68 Aug. 20 to Aug. 29 9 1.97 6.57 Sept. 4 to Sept. 11 7 1.28 5.50 Sept. 17 to Oct. 15 28 3.68 3.92 Oct. 18 to Nov. 1 14 1.84 3.92 Treatment 3, plots 24 and 28 Treatment 4, plot 21 April 19 to June 6 48 1.18 0.74 April 19 to June 6 48 1.44 0.90 June 6 to July 15 39 3 40 2.62 June 11 to July 2 21 3.41 4.87 July 21 to Aug. 14 24 5.35 6.69 July 6 to July 14 8 2.23 8.35 Aug. 17 to Sept. 7 21 4.08 5.83 July 19 to Aug. 3 15 3.04 6.08 Sept. 10 to Oct. 16 36 2.85 2.38 Aug. 7 to Aug. 20 13 2.12 4.89 Oct. 19 to Nov. 1 13 95 2 19 Aug. 24 to Sept. 14 21 4.27 6.10 Sept. 20 to Nov. 1 42 2.92 2.08 Treatment 5, plot 26 April 19 to June 6 48 1.09 0.68 June 6 to July 15 39 3.61 2.78 July 21 to July 30 9 2.00 6.67 July 31 to Aug. 11 11 3.60 9.82 Aug. 15 to Aug. 22 7 1.79 7.65 Aug. 25 to Sept. 4 10 2.09 6.27 Sept. 22 to Oct. 13 21 3.52 5.01 Oct. 18 to Nov. 1 14 1.79 3.84 20 University of California — Experiment Station TABLE 9 Rates of Use of Water in' Intervals Between Irrigations in Treatments 1, 2, 3, 4, and 5 ; Season of 1929 Interval Num- ber of days Acre-inches per acre Soil- moisture Equivalent loss in 30 days Interval Num- ber of days Acre-inches per acre Soil- moisture loss Equivalent loss in 30 days Treatment 1, plots 22 and 26 April 10 June 6 June 28 July 19 Aug. 1 Aug. 17 Aug. 31 Sept. 19 to June 5 56 to June 25 19 to July 13 15 to July 27 8 to Aug. 13 12 to Aug. 27 10 to Sept. 14 14 to Nov. 1 43 2 02 0.82 3.06 1.98 4.13 3.78 3.50 4.14 1.08 1.29 6.12 7.42 10.31 11.35 7.50 2.89 Treatment 2, plots 23 and 27 April 10 to June 5 June 6 to July 17 July 20 to Aug. 13 Aug. 17 to Sept. 10 Sept. 13 to Nov. 1 56 2.43 41 2.65 24 4.36 24 3 38 49 4.72 1.30 1.94 5.45 4.24 2.89 Treatment 3, plots 24 and 28 Treatment 4, plot 21 April 10 to June 5 June 6 to July 27 Aug. 2 to Sept. 3 Sept. 7 to Nov. 1 56 2.63 51 3.81 32 4.69 55 5.38 1.41 2.24 4.40 2.93 April 10 to June 5 June 6 to June 24 June 28 to July 13 July 19 to July 27 July 28 to Aug. 12 Aug. 16 to Sept. 3 Sept. 7 to Nov. 1 56 2.02 18 1.22 15 2.49 8 2.39 15 2.79 18 3.29 55 5.92 1.08 2.03 4.98 8.97 5.58 5.48 3.26 Treatment 5, plot 25 April 10 to June 5 June 6 to July 27 Aug. 1 to Aug. 15 Aug. 20 to Sept. 3 Sept. 7 to Sept. 24 Sept. 28 to Nov. 1 Bul. 537] Water Requirements oe Cotton on Sandy Soils 21 TABLE 10 Bates of Use of Water in Intervals Between Irrigations in Treatments 1, 1a, 2, 3, 4, and 5; Season of 1930 Num- Acre-inches per acre Num- Acre-inches per acre Interval ber of Soil- Equivalent Interval ber of Soil- Equivalent days moisture loss in days moisture loss in loss 30 days loss 30 days Treatment 1, plo ts 22 and 2 5 Treatment 1A, plots 20 and 29 April 12 to May 22 40 1.16 0.87 April 12 to May 22 40 1.56 1.17 May 23 to June 21 29 1.44 1.49 May 23 to June 21 29 1.78 1 84 June 24 to July 16 22 3.86 5.26 June 21 to July 16 25 2 25 2 70 July 21 to July 31 10 2.50 7.50 July 21 to Aug. 5 15 3 00 6.00 Aug. 4 to Aug. 15 11 3.14 8.56 Aug. 9 to Aug. 26 17 4.02 7.10 Aug. 18 to Aug. 25 7 2.20 9.42 Aug. 29 to Sept. 22 24 3.78 4 70 Sept. 1 to Sept. 14 14 2.78 5.95 Sept. 26 to Oct. 13 17 2.60 4.59 Sept. 19 to Sept. 30 11 2 04 5.55 Oct. 13 to Nov. 1 19 1.39 2.18 Oct. 1 to Nov. 1 31 1 84 1.78 Treatment 2, plo ts 23 and 2 7 Treatment 3, plo s 24 and 21 i April 12 to May 22 40 1.04 78 April 12 to May 22 40 1.00 0.75 May 23 to July 1 38 2.00 1.57 May 23 to July 1 38 2.30 1.82 July 2 to July 21 19 1.64 2.58 July 2 to July 16 14 1.32 2.82 July 25 to Aug. 14 20 3.46 5.19 July 17 to Aug. 20 34 4 30 3.80 Aug. 18 to Sept. 15 28 4.53 4.85 Aug. 25 to Sept. 9 15 2.25 4 50 Sept. 23 to Oct. 13 20 2.20 3.30 Sept. 12 to Sept. 30 18 2.37 3.95 Oct. 14 to Nov. 1 18 0.96 1.60 Oct. 1 to Oct. 13 12 0.53 1.32 Oct. 14 to Nov. 1 18 0.75 1.25 Treatment 4 plot 21 Treatment 5, plot 25 April 12 to May 22 40 0.88 0.66 April 12 to May 22 40 1.07 0.80 May 23 to June 21 29 1.34 1.39 May 23 to June 11 19 1.22 1.93 June 24 to July 16 22 3.97 5.42 June 12 to July 1 19 0.95 1.50 July 21 to July 31 10 2.50 7.50 July 2 to July 16 14 0.87 1.86 Aug. 1 to Aug. 14 13 2.00 4.61 July 17 to July 30 13 79 1.82 Aug. 18 to Sept. 15 28 5.16 5.52 Aug. 2 to Aug. 15 13 2.41 5.56 Sept. 22 to Oct. 13 21 2.02 2.89 Aug. 18 to Sept. 4 17 3 39 5.98 Oct. 14 to Nov. 1 18 1.22 2.03 Sept. 9 to Sept. 22 13 2.43 5.60 Sept. 26 to Oct. 13 17 2.82 4.98 Oct. 14 to Nov. 1 18 1 30 2.17 22 University of California — Experiment Station Figures 4, 5, and 6 show by means of diagrams the seasonal varia- tion in moisture content of the soil in the plots under treatments 1, 2, and 3 in 1927. Similar diagrams were prepared for each plot from which soil samples were taken during the other three years ; they are not included here because those for 1927 are typical of the others. I" s o JO jr o = = = = ■ i— I* 1 = u B ! ^N s — 1 1 5 1 1 i i I 10 20 30 10 20 April May K) 20 30 10 20 31 D 20 31 10 20 30 10 20 31 June July August September October Fig 1 . 4. — Seasonal variation in moisture content of continuously moist cotton irrigation plot, treatment 1, plot 23, season of 1927. Solid bars at bottom indi- cate irrigations. 13 10 5 1 1 10 S I I* O 10 £ O "> 1 \- >*3 1* $ 1 1 eg i 10 20 30 April 10 20 31 May 10 20 30 10 20 31 10 20 31 10 20 30 10 20 31 •June July August September October ^ Fig. 5. — Seasonal variation in moisture content of cotton plot irrigated when plants wilted at 4 p.m., treatment 2, plot 20, season of 1927, Solid bars at bottom indicate irrigations. Bul. 537] Water Requirements of Cotton on Sandy Soils 23 u 10 1 1 1 5 10 f° 1 I ^ 10 1 %0 5 1 i i 10 10 5 to s $ a" h 4 <1 1 10 2C April 30 10 20 May 10 20 30 10 20 Uune -July 31 10 20 31 10 20 30 10 20 31 August September October Fig. 6. — Seasonal variation in moisture content of cotton plot irrigated when the plants wilted at 9 a.m., treatment 3, plot 21, season of 1927. Solid bars at bottom indicate irrigations. Aver r7^ Sept 4.5 3.4 3.3 3.4 4.0 P I 2 s \ October 3.1 2.1 2.3 2.1 2 4 k 7 / a V \ Totat 313 £2.0 21.6 24.4 23.1 \ I 5- V? "1 r v> > t 1 * 6 I 1 J] ■s \ >\ \ I t jf A r- > ^ < s b^ - ' / K "3 \ \ \ ^ . I 1 /' \ fe s 5 4 >d °N V \ t r I 4 \ Y \> % } / s k V \\ L vj --, I t / > ^ , * o / / ■» ? ¥ *■* y^ /<■ '/ / . O . s L^ 1 or i it a K 1o z 3 I j jn e i j ui 8 3 1 ) 2 u. ) 9f 5 3 / 3 J si ,i sr Fig. 7.- -Average seasonal use of water in acre-inches per acre per month, irrigation treatments 1, 2, 3, 4, and 5, season of 1927. 24 University of California — Experiment Station Ai/erage use ^/" rvoter bu rion ths acre -inches per acre Month Treatments / 2 3 4- 5 Apr// JO # ,?. 1 fj \ i ' 5 Sept- SO -*,? 36 23 S.8 / i 1 \ \ October 3.9 <*7 2.2 2.0 4.7 7 ,/ \ • Tota/ 29 2 -»■ *-* 1 ] pf Z // a i 3 to J • j i 3 2 e 3 i 3 to 8 V 3 A 10 lug a u 3 5/ 1 3 3 a -A 3 JA Fig. Average seasonal use of water in acre-inches per acre per month, Average a^e of~ water by months acre -inches per acre. J2 Month Treatmenta 1- 7 > ^ 1 ^ 3 4 S / \ April 0.3 0.3 0.4 ad 04 ^ V May /./ 1.4- 1.6 1.3 t.7 10 / June /.9 t.9 2 J 24 2 7 / YJu/u 6 9 4.3 3.2 7.7 4.7 -4 ^ ^' J \August /0.8 S.2 4.5 66 6.S \s / \ \Sepr 6.0 3.6 3.6 4.1 S9 / H \pctober 2.7 2.2 2.6 2 7 4.1 b 7 ' P \,Tota/ 29.7 19.4 18.2 25.1 £5.9 i> / / ' H \ L # ( \ \\ r k i ^ / \ \ -x 5 \ r / '' 2 V k \, ~-V J) » ^ V *\ N > 5 ^ / '□ * s S s \ •5 tl > ^ 1 "N ^ Si \ 1 } ? / / 3 ■s. ;S s N ?! 3 - t / / 7 *lj * J; z < B ^ / r ^. K, y P 3 - U ? * i- / * 4 t J a / 2 y ^y 2 e 3 j 2 3 i 4^ '9 2 6/. Q V 1 St 3 a 3 1 3 8 sr Fig. 9. — Average seasonal use of water in acre-inches per acre per month, irrigation treatments 1, 2, 3, 4, and 5, season of 1929. Bul. 537] Water Requirements of Cotton on Sandy Soils 25 Average use of water by nonfhs. acre -inches per acre. Month Treatments I z 3 -4 S April 0.3 0.3 0.3 0.3 03 Mac/ I.I i.l 1.2 9 1.2 June 2.4 1.7 1.9 2.3 1.6 9 ^ L ) July 6.4 3.5 27 6.7 2.2 1- l' \ Auqusf 68 S.I 4.3 6.1 5.6 e \ September 6 6 4.3 4.1 3.6 S.7 A ^ k \ October 2.2 2.2 1.5 2.3 3.4 7 ^ Y / K> \ Total 27.6 id. 2 15.7 22.4 20.2 / / / Y^ / \ \ (A h-" .V "^ -U ^ -TU J ! [ ] , ^ 1- ^ \ v If t *» * K K ^ ^ \ it / / •<" > .- \ k> / > ' v 3 V. \ 3 1 ( / 7 > s. s ^ \ 1 S * \ \ J / ' •< \ >i > — - ,^ $ 5 \ fc J X- - K ^ ■■ * K s / to m April io as to 20 May June July w to August io to Sept. 10 id October Fig. 10. — Average seasonal use of water in acre-inches per acre per month, irrigation treatments 1, 2, 3, 4, and 5, season of 1930. Averoge use of water by months, ocre-/nches per ocre. Month Treatments. 1 2 3 4 5 Apr,/ 0-2 0.2 0-3 02 03 Mot/ f.Q 10 II 10 1-1 V s June 32 22 22 3-0 23 9 July rr 51 4.1 16 46 / f -& \ August 6-9 5-3 53 5-3 6 7 6 V September 55 39 3.3 35 5-4 4 V / i f ^ x s \ October 30 24 21 2-3 36 1 Vjbtoi 295 20-6 id- 9 £3.4 24.0 6 1 / * K \ — 7. N / s J A' > - i y \ < ^ \ I / i / i - 3 s N ^ \ h 5 J a / ^ L t, L Y i s v f/ <* r ^ * i ' * '- c Apr,/ 10 20 10 20 rioi/ June 10 July 10 20 August 10 20 September October Fig. 11. -Average seasonal use of water in acre-inches per acre per month, irrigation treatments 1, 2, 3, 4, and 5, 1927 to 1930. 26 University of California — Experiment Station In figures 7, 8, 9, and 10, the data shown in tables 6 to 10 have been plotted as seasonal-use-of-water curves from which the average use of water in acre-inches per acre per month for each treatment is obtained. Figure 11, a composite diagram, shows the average use of water during the four years. These diagrams show that under treatment 1, the use of water by the crop reached its peak during the latter part of July in 1927 and 1928 and about the middle of August in 1929 and 1930, and that during these periods of maximum use, the crop was using water from the soil at the rate of 10.5 acre-inches per acre per month. These periods of maximum use of water are of comparatively short duration and occur at the peak of the flowering period, rather than at the period of maximum temperatures. With the exception of treatment 5, under which a heavy vegetative growth was produced late in the season, the decline in the use of water after the peak of the flowering period was very rapid. In treatments 1 to 4, the September use averaged 65 per cent of that in August, while the October use was 57 per cent of that in September. Table 11 shows the average moisture content of the soil in each treatment at the time of irrigation. The results in this table show comparatively small differences in average moisture content of the top 5 feet of soil among plots irrigated when the plants wilted at 9 a.m. or 4 p.m., or when they showed their first indication of lack of moisture by their change in color. The first indication of wilt as shown in treatment 2 occurs when the moisture in the top foot of soil has been reduced to a point slightly above the permanent wilting percentage. At this time the second and third feet average about 1 per cent and the fourth and fifth feet, 2 per cent moisture above the permanent wilting percentage. Early- morning wilt as shown in treatment 3, occurred when the top 2 feet of soil had reached the permanent wilting percentage, with the third, fourth, and fifth feet averaging only 1 per cent above it. According to one year's observations, the change in color of foliage due to soil moisture shortage apparently occurs when the top 2 feet is 1 per cent above the permanent wilting percentage, but while readily available moisture is still present in the lower depths. Seasonal Use of Water by Cotton. — The average use of water by months and the irrigation data for each of the treatments are given in tables 12 and 13, Bul. 537] Water Requirements of Cotton on Sandy Soils 27 TABLE 11 Average Moisture Content of the Soil in Each Treatment at the Time of Irrigation Per cent moisture Season First foot Second foot Third foot Fourth foot Fifth foot Average Treatment 1 1927 5.4 6.0 5.3 5 5.9 6.4 5.9 6 7 7.2 7 1 7.1 8 3 7.9 8.8 8.6 7.8 7.5 8 3 8 1 6 9 1928 7 1-929 7.1 1930 7 5 4 6 7.1 8.4 7.9 7.0 Treatment 2 1927 4.5 3.9 3.8 3.9 4.5 4.4 4.4 4.6 5.2 4.6 4-7 4.9 6 5 3 5.4 5.5 5.9 5 5.6 5 3 5.2 1928 4.6 1929 4.8 1930 4.8 4 4.5 4.8 5.6 5.4 4.8 Treatment 3 1927 3 7 3.4 3.2 3 2 3 6 3.9 3.6 4 3 4.3 4.1 3.8 4.5 4.5 4.4 4.4 4.9 4.2 4.1 5.5 4.9 4.1 1928 4.0 1929 4.1 1930 4.4 3.4 3.8 4.2 4.6 4.7 4 2 Treatment 1A 1930. 4.0 4.7 5.2 5.3 TABLE 12 Average Use of Water by Months, Seasons 1927 to 1930 Average use, in acre-inches per acre Treatment April May June July Aug. Sept. Oct. Total ! 0.2 0.2 3 0.2 03 1.0 1.0 1.1 1Q 1.1 3.2 2.2 2.2 3.0 2.3 7.7 5.1 4.1 7.6 4.6 8.9 5.8 5.3 5.8 6.7 5.5 3.9 3.8 3.5 5.4 3.0 2.4 2.1 2.3 3.6 29 5 2 20 6 3 18 9 4 23 4 5 24 28 University of California — Experiment Station As would be expected, treatment 1, with the moisture supply above the permanent wilting percentage throughout the season, showed the greatest use of water, averaging 29.5 acre-inches per acre for the four years. This figure was 43 per cent greater than the use under treat- ment 2, 55 per cent greater than under treatment 3, 28 per cent greater than under treatment 4, and 23 per cent greater than under treatment 5. Table 13 shows that an average of 72.8 per cent of the water applied during the growing season was accounted for in soil-moistur^e increase in the top 5 feet of soil and that 83.5 per cent of the total amount of water applied during the season was used through transpira- tion in producing the crop. TABLE 13 Summary of Irrigation Data; Averages for Seasons 1927 to 1930 Number of irrigations Amount of water, in acre-inches per acre Per cent Treatment Total applied Total applied during growing season Accounted for in soil- moisture increase* Used by the crop Efficiency of irrigation* Water applied which was used by the plants 1 7 4 3 5 5 38.6 24.8 22.6 28.0 26.2 32 6 18.8 16.6 22 26.2 22 6 14.2 12.5 15.5 19.0 29 5 20.6 18 9 23.1 24.0 69.4 75.5 75.4 70 5 72.6 76.5 2 83.0 3 83.7 4 .. 82.5 5... 91.7 72.8 83.5 * For the amounts of water applied during the growing season. Growth and Yield. — During each of the four years, the growth (as measured by the height of plants, flowering characteristics, percent- age of shedding, number of bolls, yield, lint index, length of staple) under the different treatments followed so nearly the same trend that publishing of the detailed results for each year is not considered necessary. Effect of Irrigation on Size of Plants. — The average height of plants produced each year under the different treatments is shown in table 14. This table shows that as an average over the four-year period, the same height of plant was obtained when the plants were allowed to wilt at 9 a.m. early in the season, but with moisture available contin- uously after August 1 (treatment 5), as was obtained when available moisture was present throughout the season (treatment 1). During the fall of the year the plants under treatment 5 were making the Bul. 537] Water Requirements of Cotton on Sandy Soils 29 growth ordinarily made in midsummer. This growth is reflected in the later flowering and greater use of water during September and October. Continuously wilting the plants through the season affected the size of the plant, those under treatment 3 being the smallest and those under treatment 2 next to the smallest each year. TABLE 14 Summary and Averages of Height of Plants, Seasons 1927 to 1930 Height of plants, in inches Treatment 1927 1928 1929 1930 Average 1 65 49.5 42.7 46.5 55.0 53.0 45.0 40 48 53.0 38.2 32.0 31.8 38 45.0 35.6 30.8 27.9 32.5 39.5 48.0 2 39.3 3 35.6 4 41 2 5 .... 48.1* * The height of these plants was due to growth during the fall of the year, whereas normally it is made during midsummer. On the light soil types such as the Delano sandy loams, although a high percentage of moisture may be maintained throughout the season, terminal growth ceases and the plant matures normally as the season advances. Effect of Irrigation on Flowering. — Tables 15 and 16 give the results of the observations made on the dates on which the first flowers appeared, the dates of most intensive flowering and of opening of the first bolls, and on the flower counts each year under the different treatments. The results given in table 15 show that the irrigation treatments as applied had no effect on the date of beginning of flowering. When water was withheld during the first half of the season (treatment 5), the period of most intensive flowering was hastened from 1 to 2 weeks, in three out of the four years, the date of opening of the first bolls was advanced about 1 week in each of the years of record, and the flowering period was extended into the late summer, with a resulting heavy crop of immature green bolls. Considering the number of flowers per plant produced under the different treatments, table 16 shows in every year but one (1927) the greatest number of flowers was produced under treatment 1 and the least number under treatment 3. A full moisture supply during the first half of the season (treatments 1 and 3) also produced the great- est number of flowers per plant during the 5-day period of most intensive flowering. 30 University of California — Experiment Station © O o o o> 5 ^ r/j 01 V, OS I— 1 EH H GO W h- 1 o pq En © DO W < S C/J © DQ ftf O S pq (S pH © m J « Ph ft Mh © DO w H ■^ ft ft © o to co CO -* 05 (M OS S bb bb bb bb 13 3 3 3 3 ««J «! < K < o a o DO ,H t- co co -<*< CM o OS bb bj bj bb bb 3 3 3 3 3 2 «! w bb bb bb bb 3 9 3 3 s j>> >> >J 3 3 13 3 3 1-5 ^ l-S »-s M _fl ■* O £ bb 3 o <* ia >o C3 OS CM os CI ci <5 o> os © o o o _> Ol Ol CO (M "S bb >> >> >, >> 3 3 "< 3 3 3 3 ""3 hi t-3 1-9 rt _,_ 5 o O US o US co O] r -1 CO -- 1 00 m ■o © >o o (M , >> >> ^? >. 0) 0) 3 3 3 3 3 Q >-i ►-s 1-5 1-3 >-5 © us o US o CI CO Ol CO ^ us e US i US CM > i? >. _>> >» 3 3 3 p 3 >-5 >-s ►-s t-s 1-5 OO C Ol OS 3 a 3 3 3 3 3 3 3 3 >-i »"3 •-5 ►■a 1-5 o OS CI ^ c OS CO GO os 9 is OS 0> 3 3 >J o> 3 3 3 3 3 3 3 o >-5 1-5 t-a •-3 •-3 so (3 O o O 3 00 o> 01 o o> OS 3 3 3 G 3 3 3 3 3 3 1-3 >-5 1-5 i-s 1-5 Q to CM to >o to O t~ CI Ol Ol Cu O OS 3 3 3 3 3 3 3 3 3 3 >-s *-8 1-3 •-S 1-5 ^ s *" III a « ^ CM 8 CO UO o co CD CO Tt In cd© r w co CO o as CO o CM «3 CM CO CM a a 0) bfl c3 "# CO co o CO CO o OS CM T3 9 o 1 < o l~ CO CO CO a fl © CO OS o co b- © CO CO r^ r- oo CO t*- CD os «o ■*< >o lO O J- o 22 r cD a> OS CM OS o I~» o o CO '>>» 00 oc CO o co CO lO H3 CD -T 3 *-3 CO OO OS CO o f~ CO CO CO e CO CO "1 oo oa 00 ~ ~ CM OS CO CO t-» © lO OS OO © OO CO «5 CO CO 0) o o o o © o © os OS 00 -3 > CO 00 o 00 a CO CM CM 3 ■«lj o a a © oo OS o OS o cm o o 8 o cm" 2 5# s CO o CD ,_ CD O OS OS o o o o © CO o 35 o 1 OO OS **< «3 o£ CO CN r£ _I ^ 03 a 2 T3 T3 00 CM OS o o e © © 8 OO CO o to t- OS "3 cd " H "*< CO N of cm" "S a CO CM OS o o o © © CO *o oo o CO CM OS OO CM* KO 1* co CO CO o o CM OS «>. CO 00 a CO OS o o >o _ OO »o CO CM CO CM c o» oJ2 OS o 1H OS r~ t^ w a CM OS »o t-- iO »o ^ -a %* >o co CM CO CM 4) 0) S ft "o OO o o OO o CO CM OS CM o CO OS CO T3 t^ CO <* T* '3 £ - CO a s OS ED CD hi. ed c c rt IB cr! c s a © a © OO 1 o CD a a > c c X) o '1 o c J2 C3 fl ctJ -O WOfiO Bul. 537] Water Requirements of Cotton on Sandy Soils 35 Description of Tanks and Equipment. — The twelve tanks used in duplicating- the field trials were made of heavy galvanized iron, 27 inches in diameter and with an inside depth of 64 inches. The sides of the tanks extended several inches below the bottoms, which were fitted with a drainage faucet of small diameter. When filled to within 4 inches of the top, each tank contained 20 cubic feet of soil, the sur- face area being" comparable with the area occupied by a single plant in the field. Fig. 12. — Perforated pipe used in tank experiments for applying irrigation water 6 inches below the soil surface. The tanks were installed in a trench 54 feet long, 64 inches deep, and 30 inches wide, the natural soil being removed in 6-inch layers and, after air-drying", being replaced in the tanks as nearly as possible in its original condition, both as to compactness and depth. A wire screen and 4 inches of fine gravel were placed in the bottom of each tank to facilitate drainage of excess water from the soil. The same weight of air-dry soil was placed in each tank, and sufficient water was added to bring' the soil to its field capacity. In order to reduce evaporation losses from the soil surface and to aid in uniform water distribution, a network of i/p-inch perforated pipe (fig - . 12) connected with a 2-inch standpipe was placed at a depth of 6 inches below the soil surface in each tank. All water was applied 36 University of California — Experiment Station through these subsurface systems. During 1928 a layer of heavy tar paper was placed in the soil mulch 2 inches below the surface to pre- vent possible loss by evaporation. In 1929 and 1930 this paper was replaced with wall board cut to fit the circumference of the tank and placed on the soil surface. The six water-table tanks, 27 inches in diameter and 60 inches deep, were perforated in the bottom and for about 1 foot up the sides. They were then jacketed to within 18 inches of the top by a larger tank which served as a reservoir for maintaining the water level in the inner tank. A %-inch perforated pipe placed in the center of each tank was used in obtaining readings of underground water levels. Fig. 13. — Tank yard, showing method of covering tanks, derrick used in weighing tanks, atmometers, and hedge of cotton grown around trench. Losses were obtained by weighing with a 3,000-pound capacity suspension scale hung from a movable derrick centered over the trench in which the tanks were installed. Figure 13 is a general view of the tank yard, showing tanks, derrick, and lifting device. Irrigation Treatments. — Treatments 1, 2, and 3 as applied to the field plots were duplicated in the tanks as follows : Treatment 1 (four tanks) : Irrigation water added when the soil moisture content in each tank was reduced to 7 per cent (about half- way between the field capacity of 11 per cent and the permanent wilting percentage of 3.5). Bul. 537] Water Requirements of Cotton on Sandy Soils 37 Treatment 2 (four tanks) : Irrigation water added when the plants wilted at 4 p.m. Treatment 3 (four tanks) : Irrigation water added when the plants wilted at 9 a.m. Procedure. — At the time the tanks were filled, a number of soil samples were taken from each foot depth of soil ; and from these the dry weight of soil in each tank was obtained. Occasional soil samples taken through each season substantiated the correctness of these deter- minations. After the first irrigation, in which more than enough water was added to bring the soil to field capacity, it was found that the sandy loam scil overlying the layer of gravel would not drain, the soil imme- diately above the gravel remaining in a saturated condition in all tanks until the excess water was removed by the roots of the cotton plants penetrating it. Similar observations have been made in the field, where fine sand or sandy loam soils have been found to be saturated immediately above a layer of coarse gravel. After the first irrigation, care was taken not to add more than the quantity of water needed to bring the soil to its field capacity. Planting in the tanks was done at the same time as planting in the field plots. Twelve seeds were placed in each tank ; and, at the time the field plots were chopped, only the most thrifty plant was left in each tank. At this time the tar-paper or wall-board covering pre- viously mentioned was put in place. The space between the stalk of the plant and the edge of the hole through which it grew was snugly packed with cotton. Systematic weighing was started at the time of thinning and continued through the season. Daily flower counts were made on all plants through the flowering period, and at the end of the season the weight of seed cotton produced from each plant was obtained. Use of Water by Cotton Grown in Tanks. — The daily use of water and the equivalent use in 30 days by cotton plants grown under treat- ments 1, 2, and 3 during the season of 1928 are given in table 21. The corresponding seasonal use of water is shown in figures 14, 15, and 16, and table 22 gives the total seasonal use of water under the different treatments for each of the three years. 38 University of California — Experiment Station £ is o M O W H £ <«1 M P-l fc 00 Q CJ H cs H 1— 1 O rn O o S* fe m o C/J . CN eo ^3S:§ s »- >r CM CO t^ oc ■* O C5 1C U5 O W 00 C »c © CN ^ 0O N 1C IT. "* co ■>* oc: CM CM tfl 00 e3 O a CO (-i CD rt«w « COS f. oc ■>*< ~H t> t^ b- t>- t>- l> t^ t^ CM 00 CN CO ■* 3 T3 £ i * >> « CO 03 m c? ■« s- 3~ r « ^ ic CM O C e CM CN O O c O C S B OC o O l-H CC o- O) O) O) c -H O CN ■* a- CCco co *"" oc as io rf CM CM O C (O ifl l> CO <— H ~.s 50 H. rt ^ t-h r-l rt ^« 3 . CC CM OO O CM O C <* «*t >a lO CN CO CO CM — J "§ r^ t^ cn 00 CN CO 3 T3 y-l £ DO 05, -=: k g§ w tfi O O — « H CO * u: i-i 00 cr. ©5 CC *3+»0 "T »» w oc « n k c M O ^ - O) N cc CO CN cr Ceo H N « CN CM CM r-H 1- HJSfl k & bfi ^ tfJV a> CO t? eo o c C3i O -— a O O O cr c e C M ^ * t^ oc a CD 2^* ^ s 8, CD 00 CO CC co ^ 6 oe CO 10 •* CM y- a 1 0) (-. cS O 1 « oc o o «■ IO CO *< CN CM i—l (h 0) C O «3 if. oc •^< -H l> t^ «>. t» l> t^ t^ (M 00 CN CO 3 -O 1-1 £ i-i 00 U U 1 "el P 8 C CN s c 1 c 2 > 3 i- CN 1 > 5 >-: o « i > "3 1-: r i > bj c 1 I 5 Of 3 3 bi 3 < e oc CN "cc 3 - CN a c 3 >-: > "5 CN > 1 CO > b <1 5 j b 3 b 3 << 9 b 3 c C Bul. 537] Water Requirements of Cotton on Sandy Soils 39 TABLE 22 Total Water Used in - Acre-Inches Per Acre by Cotton Plants Growing in Tanks Under Treatments 1, 2, and 3; Seasons op 1928, 1929, and 1930 Tank No. Total use of water, acre-inches per acre Treatment 1928 1929 1930 9 52.2 44.37 25.65 12 51.7 45 39 24.50 1 15 51.8 49.56 22.82 18 44.60 22 41 Average 51.90 45.98 23 85 8 38.0 28.82 22.06 2 11 14 36.2 37.6 34.58 33.73 23.41 18.97 17 38.6 34.12 19.29 Average 37.60 32.81 21.33 7 35.6 30.52 20.62 3 10 13 32.9 30.6 22.80 33.61 21.43 18.78 16 Average 28.8 31.90 30.67 29.40 19.29 20 03 Average use of roofer t>u 7TO/?tAi \ /k?re - /ficAes per ocre Montn Treatment / 3 3 , \ /\pnl / \ / \ Maty 0-9 OS 0-3 / / \ / \ June 4-3 3-9 3-3 ■«, \ \ ? Juty /7-9 /?■/ /O 5 5 1 U August /a- 9 /£3 ZO-3 | / September 3-3 73 6 / October 3.3 3.0 2-6 § / \ Totat 54-e 33-3 33.7 st, J F >, \ g/J / * £ \ B J | \ \ &> ? C t 4 \\ \ I / / L '£r \ \ E 1 r l« 7 \ \L V I S /J J 1/ j i t 3 \ * i , \ * i I "W l i r S ( \ >i / \ ^ S \ / \ £ 1 fis i 1 '« t n i * 30 /O 30 3/ tO 30 30 /O 30 3/ /O 30 3/ /O 30 30 /O 30 3/ Mat/ June ^ u ^/ August September October Figure 14. — Average seasonal use of water in acre-inches per acre per month by cotton plants grown in tanks under treatments 1, 2, and 3, season of 1928. 40 University of California — Experiment Station Av eroqe USt of rvoter da months Acre-/nct)es per acre 26 Montn Treatment t 2 3 Apr// Maa 0.3 0-3 0-4 June 2.9 32 3-5 *S Ju/u /3-3 tOO 7-9 r \ August /7.4 //■4 to 4 ' V V September 7-6 5-4 50 / \ October 2-3 2-2 2 1 P \ Total 4-4-3 32.5 29-3 J5 t ' V i 'V-- / ^ r t* '" 3 4^ K \ V /■ G fe K J d m k \ 1 i ^ 1 \ ( i \ 5 / K "' V I -Y y /> x ; "~S- V: ."V ""3, 30 tO 20 3/ /O 20 30 /O 20 31 /O 20 3/ /O 20 30 /O 20 3/ May June Ju/y August September October Fig. 15. — Average seasonal use of water in acre-inches per acre per month by- cotton plants grown in tanks under treatments 1, 2, and 3, season of 1929. A veroae use < if water A/ months. Acre /nches per acre. 25 Montb Treatment. I 2 3 Apr// 0-0 0.0 0-0 May 0-4 0.4 O 4 June t 5 t-9 1.7 20 Ju/u 6-3 6 3 63 Auaust 8 3 6 3 6-9 % September 3 7 2-9 24 \ October t-3 / 4 f-5 Totat 22 5 20 2 t9-2 \* \ % X /O A 1 ,P — 4 \ \ fe /,' S V i V jk $, 3 '4 i / *& "■ j \ A N. 6 V t H 6 f » d ir