C! fff UNIVERSITY OF CALIFORNIA COLLEGE OF AGRICULTURE * V& AGRICULTURAL EXPERIMENT STATION BERKELEY, CALIFORNIA A Survey of Orchard Practices in the Citrus Industry of Southern California ROLAND S. VAILE BULLETIN 374 January, 1924 Beprinted June, 1929 UNIVERSITY OF CALIFORNIA PRINTING OFFICE BERKELEY, CALIFORNIA 1929 CONTENTS PAGE Foreword 3 Introduction 3 Importance and growth of citrus industry 3 Collection of data 4 Factors upon which information was gathered 5 Analysis of the data 6 Climatic zones and their effects on orchard practices and yields 6 Effect of abnormal weather conditions on yields 8 Effect of soils on yields 9 Age of trees, and yields 11 Fertilization and its effect on yields 13 Yields as affected by nitrogen applications 13 The use of bulky organic manures, and yields 16 Effects of cover crops on fertilizer requirements 17 Irrigation and its effect on yields 19 Amount of water applied 19 Frequency of irrigation 21 Plowing and its effect on yields 21 The 'limit of profitable cultivation' 22 Cost of fruit per pound 23 Net returns per acre 25 Net returns on the investment 25 The most profitable use of nitrogen 26 Cost of developing citrus orchards 28 Conclusions 31 Appendix 32 Homogeneity and accuracy of the data 32 Methods used in analysis 33 Classification of the variables 33 Association of the variables 34 Averages based on the law of large numbers 35 Partial correlation coefficients 36 Method of combining groves from three climatic zones 38 Diminished increment of physical returns with increased use of fertilizer... 38 A SURVEY OF ORCHARD PRACTICES IN THE CITRUS INDUSTRY OF SOUTHERN CALIFORNIA 1 EOLAND S. VAILE2 FOREWORD An attempt has been made to present this summary of the results of a survey of citrus orchard practices in southern California in a manner that would appeal to the orchardist. For the benefit of those who are interested in the methods followed, either in the collection or the analysis of data, certain notes have been included in smaller type as an appendix. These notes will also be of interest to any who wish to consider critically the reliability of the conclusions drawn. INTRODUCTION The purpose of this bulletin is to show from actual field records the influence of fertilization, irrigation, plowing, climate, soil, the age of trees, and costs, upon the profitableness of citrus orchards in southern California. Importance and growth of citrus industry. — The commercial citrus industry in California is but little over fifty years old. It has de- veloped rapidly for an agricultural industry. In 1921 California produced 73 per cent of all the oranges and 88 per cent of all the lemons consumed in the United States and Canada. The shipments of oranges alone from California averaged approximately 1,000 cars each week throughout the year. During the past ten years (1913-14 to 1922-23 inclusive) ship- ments of oranges have increased about 25 per cent, while shipments of lemons have more than doubled. Table 1 gives the approximate ship- ments in packed boxes for that period. The average yield of mature orange groves over a period of years is about 150 packed boxes per acre, while that of lemons is about 175 boxes. Wide variations from these figures are occasionally noted for favorable seasons or on individual groves. A few extreme yields of 1 Paper No. 114, University of California, Graduate School of Tropical Agri- culture and Citrus Experiment Station, Eiverside, California. 2 Assistant Professor of Orchard Practices in the Citrus Experiment Station and Graduate School of Tropical Agriculture; resigned October 1, 1923. 4 University of California — Experiment Station 800 to 1,000 packed boxes per acre are on record for both oranges and lemons, but such are not safe figures on which to base a judgment of the industry. The geographic distribution of the citrus plantings has changed in recent years. The early plantings of citrus trees centered around several locations — notably San Gabriel, Orange, Riverside, and Red- lands. However, following the advent of the Navel orange in 1873, there was a rapid increase of acreage in the interior districts. That variety had first been tested at Riverside and its early reputation was that of an interior fruit. Later the plantings of Navels spread so that it is now the predominant orange throughout the Pomona-San Gabriel valley. Plantings of Valencias — the summer orange — gained impetus some years later and are concentrated to a great extent in the regions nearer the coast. Thus the average age of orchards near the coast is less than that in the interior. TABLE 1 Shipments of Packed Boxes of California Oranges and Lemons for the Season 1913-14 Through 1922-23 (Thousands of boxes, i.e., 000 omitted) Year* Oranges Lemons Year* Oranges Lemons 1913-14 17,900 1,180 1918-19 18,100 4,280 1914-15 15,700 2,800 1919-20.: 15,650 3,780 1915-16 15,100 2,900 1920-21 20,200 4,700 1916-17 19,700 3,260 1921-22 12,700| 4,150 1917-18 7,800t 2,560 1922-23 21,000§ 5,100§ * November 1 to October 31. J Crop reduced by cold. t Crop reduced by heat. § Partially estimated. Data brought to date of reprinting can be found in California Agr. Exp. Sta. Buls. 457 (out of print, can be found in libraries), 460, and 463. Collection of data. — For some years the Citrus Experiment Station has endeavored to bring together many of the accumulating experi- ences of growers in this new industry. Beginning in 1914 the writer collected statements from individual growers showing their methods of production, costs, and yields. Much of this material has been of such a heterogeneous nature that it could not be satisfactorily ana- lyzed. However, the figures on cost of developing citrus groves are of such nature that they may be reported. During the winter of 1922-23 a more intensive survey was under - taken. Approximately one thousand growers were interviewed and specific information gathered from them concerning their orchard practices, costs, and yields. The only records that were used in sub- sequent analysis were those that covered a five-year operating period. Bul. 374] Orchard Practices in the Citrus Industry 5 In this work the writer had the hearty cooperation of the County Agent and the Citrus Committee of the Farm Bureaus in each county in which he. worked. (Data were collected in Los Angeles, San Bernardino, Riverside, Orange, and Ventura counties.) The usual method of procedure was for the County Agent to call a meeting of the Farm Bureau Center in each of the. districts to be surveyed. The details of the work were presented at the meeting. Then for several days following, interested growers made appointments for the investi- gator and accompanied him on his survey calls upon the orchardists who furnished the data. Ordinarily a week was spent in each locality and a different grower accompanied the investigator each day. This method of gathering data naturally led to a sort of selection of groves. Only those were taken for which a five-year record was available, which made it practically essential that one owner should have held the property for that length of time. Many growers were eliminated because they had kept no records covering so long a period. All groves under eight years of age were eliminated ; and only a small number were taken under fifteen years. Many times it was necessary to leave out groves because the owner held two or more separate orchards without having separate records of their maintenance. Two things were very evident in the selection of groves. In the first place not over half of the growers keep any record of their orcharding operations except as to cash expenses and receipts. And in the second place an even smaller number attempt to analyze such records as they do keep with a view either to increasing production or to decreasing costs. For many purposes it is more important to know how much nitrogen is applied to trees than to know how much money is spent for fertilizer. The significance of such a physical measure seems not to have caught the attention of the average grower. Possibly the analysis of these survey data will point to a helpful method of studying the orcharding business. Factors upon which information was gathered. — There are many variable factors in so complex an industry as the production of citrus fruits. This study does not include them all. Some of the items con- sidered are the location of groves both in regard to climate and to soil, the variety and age of the tree, and orchard practices as to fertilization and irrigation. Certain other items — notably pruning and pest control— have been considered only incidentally. Whenever severe pruning had obviously affected the yields of a grove or whenever pests had not been commercially controlled, the groves were discarded from the records. A discussion of the classification of the variable factors will be found in the appendix. 6 University of California — Experiment Station Analysis of the data. — In the analysis of any production data there are always two distinct phases involved. The first of these is the relation between changes in any variable on the one hand and physical production on the other. This relationship is always subject to the law of diminishing physical returns, 3 and if the data are complete enough they may be studied in the light of the law. The practice which results in the greatest yield per acre may be called the 'most produc- tive practice. ' The second phase of the analysis is concerned with the effect which changes in any variable have on the profitableness of the enterprise. When viewed from this angle, the data are to be interpreted in accordance with their approach to the ' limit of profitable cultivation. ' The practice which results in the greatest net returns on the invest- ment may be called the 'most profitable practice. ' The most productive practice and the most profitable practice are such totally different concepts that they require separate consider- ation. In this discussion physical returns will be considered first and then some of the more salient factors concerned with the ''limit of profitable cultivation" will be presented. CLIMATIC ZONES AND THEIR EFFECTS ON ORCHARD PRACTICES AND YIELDS The citrus belt of southern California comprises districts which possess markedly different climates. The climate at Santa Ana is very different from that at Riverside, particularly during the summer months. Santa Ana is typically under a coastal influence, while River- side approaches the arid interior. The Pomona Valley lies midway between the two in the amount of coastal influences which it receives. In this study the citrus belt has been divided into three climatic zones as represented by the above illustrations. 3 ' ' It is well known to practical men that a niggardly application of labor and capital to a piece of land in the cultivation of any crop is little better than wasted, because it will produce so little in proportion to itself ; whereas a more generous application will yield a crop not only larger, but larger in proportion to the amount of labor and capital employed. Up to this point the land is said to yield increas- ing returns to the labor and capital employed in its cultivation. But if the amount of these factors used in cultivating a given piece of land is still further increased, a point will eventually be reached where the product will no longer increase as fast as these factors are increased. Beyond this point the land is said to yield diminishing returns to the labor and capital employed. Though larger applications of labor and capital may continue to produce larger crops, the crops will not be so large in proportion to the labor and capital. >f T. N. Carver, The Distribution of Wealth, page 56. The application of the law of diminishing returns to the present data is discussed further on page 40. Bul. 374] Orchard Practices in the Citrus Industry 7 The division of the southern California citrus belt into three climatic zones is somewhat arbitrary. The main criterion used in the division was the mean maximum temperature during the summer. On this basis the coastal zone includes most of Orange County and the Whittier and Pasadena sections of Los Angeles County; the intermediate zone includes the San Gabriel-Pomona valley; the interior zone includes the San Fernando, Fillmore, Redlands, High- lands, and Riverside districts. The Government Weather Bureau reports show consistent differ- ences in mean monthly maximum temperatures during the summer between these three zones. The month of August has been used as an index of these differences, and all the stations within each zone have been averaged over a five-year period. The figures in table 2 result. TABLE 2 Average Mean-Monthly-Maximum August Temperatures, for Each Climatic Zone, 1917-1921 Coastal 87° F Intermediate 91° F Interior 94° F Tulare County (not included in this survey). 98° F The average annual rainfall for seven stations in each zone is given in table 3. These averages are based on the records of the United States Weather Bureau and cover from 25 to 40 years in individual cases. TABLE 3 Average Annual Kainfall in Climatic Zones Coastal 15.6 inches Intermediate 19.6 inches Interior 16.9 inches Other weather factors of recognized importance to crops are relative humidity and wind. Complete data on these points are lacking, but the effect of the three factors of temperature, relative humidity, and wind may be estimated by the relative rate of evapora- tion of water from the surface of porous porcelain spheres. This rate in the interior district is approximately double that of coastal district. The range of soils is roughly similar in the three districts, although there is a slightly lower proportion of heavy soil in the intermediate 8 University of California — Experiment Station zone. The classification of soils that was finally adopted for this study includes loams and clay loams as 'heavy'; fine sandy loams as 'medium'; sandy loams and gravelly sandy loams as 'light.' 4 Certain data concerning the average yields and practices in the three zones are given in table 4. It is probable that part of the differences in. yield among these different districts is due to climatic conditions. It is equally probable that part of the difference is due to difference in soil, age of trees, and orchard practices. TABLE 4 Kelation Between Climatic Zones, Average Orchard Practices, and Yields Age of trees (years) Yield per acre Fertilization per acre Irrigation per acre Zone Oranges (pounds) Lemons (pounds) Nitrogen applied (pounds) Manure applied (tons) Water applied . (acre inches) Fre- quency • (days) Coastal Intermediate Interior 21 25 27 19,500 18,500 15,400 24,700 24,000 18,900 123 160 122 9.6 7.8 6.3 20.2 24.9 28.6 33 25 30 Lemons produce more fruit per acre than oranges in each of the districts. The excess is less in the interior than elsewhere, indicating that lemon production is at a relative disadvantage as compared with orange production in that zone. The excess of lemon production per acre over orange production for the several zones is : Coastal 40 per cent, or 5200 pounds Intermediate 42 per cent, or 5500 pounds Interior 23 per cent, or 3500 pounds Effect of abnormal weather conditions on yields. — The records given in table 1 show the widespread effect of unfavorable weather conditions in two separate years. In June, 1917, there occurred a few unusually hot days that resulted in a heavy drop of young fruit, while in February, 1922, at least 50 per cent of the orange crop was frozen on the trees. Table 5 presents the fluctuations in shipments from one of the oldest citrus districts in the state. The bearing acreage in this district continued to increase until 1911 or 1912. Since that time there has been a slight reduction in total acreage. 4 For a further discussion of citrus orchard soils, see: Vaile, R. S. Fertilizer experiments with citrus trees. California Agr. Exp. Sta. Bui. 345:508-511. 1922. (Out of print.) Bul. 374] Orchard Practices in the Citrus Industry TABLE 5 Shipment of Citrus Fruits from One of the Oldest Citrus Districts for the Seasons 1890-91 to 1921-22 Inclusive (Thousands of packed boxes, i.e., 000 omitted) Year Boxes Year Boxes 1890-91 415 400 715 530* 800 585 715 1,340 715* 1,310 1,580 1,090* 1,410 1,710 1,730 1,520* 1906-07 1,630 1891-92 1907-08 1,480 1892-93 1908-09 1,740 1893-94 1909-10 l,120t 1894-95 1910-11 1,640 1895-96 1911-12 830* 1896-97 1912-13 250* 1897-98 1913-14 1,060 1898-99 . . 1914-15. .: 910 1899-00 1915-16 900 1900-01 . 1916-17 1,050 1901-02 1917-18 440t 1902-03 1918-19 1,050 1903-04 1919-20 1,080 1904-05 1920-21 1,480 1905-06 1,250* Cold affected yields. t Heat affected yields. EFFECT OF SOILS ON YIELDS Soils may be grouped in several different ways. In this study the classifications of the U. S. Bureau of Soils have been followed. When divided according to geological origin and formation, the citrus soils mainly fall into four series, namely, Hanford, Yolo, Ramona, and Placentia. All of these are alluvial soils. The Hanford and Yolo series are relatively recent, while the Ramona and Placentia are much older. Yolo soils are derived from shales and other sedimentary rocks, while the other three series are derived mainly from granitic rocks. Each of these series is represented in all physical grades from sandy loam to clay loam. There is no apparent difference in average yield per acre of citrus on Yolo, Hanford, or Ramona. soils of the same physical grade. Placentia soils, which are often underlain by hardpan, and which absorb irrigation water very slowly, are less productive of citrus fruits than any of the others. The physical make-up of the soils is of some importance in its effect on citrus yields. The groves planted on medium-textured soil (fine sandy loam) gave the highest average yield. About 5 per cent of the groves are planted on very light soil (sand) and these groves produced 30 per cent less than those on medium-textured soil. From 5 per cent to 10 per cent of the groves were planted on very heavy 10 University op California — Experiment Station soil (clay loam) and these produced 10 per cent less than those on medium-textured soil. Table 6 gives the actual average yields for the several types. Figure 1 is a graphic presentation of the yields in the coastal zone. TABLE 6 Average Yields of Oranges on Different Soil Types, in Pounds per Acre Type Coastal Intermediate Interior 14,400 20,100 21,200 19,600 19,000 13,700 19,200 20,100 18,700 17,900 11,400 16,000 16,700 15,500 Clay loam — very heavy 14,900 Yield in Pounds Per Acre. ZZ50O 20,000 /Zsoo /5,00a - /2,50O _ /O.OOO - So// Types s $ s 5 r Fig. 1. — The relation between soil types and yields in the coastal district; illustrating table 6. The differences in yields between middle ranges of soils (light, medium, and heavy) are so slight that they may be almost disregarded. Comparatively few groves have been planted on the extreme soil types where a real reduction of yields is experienced. In the following study of effect of orchard practices on yields, the groves planted on sand and those planted on clay loam have been excluded. Bul. 374] Orchard Practices in the Citrus Industry 11 AGE OF TREES AND YIELDS There appears to be a definite relation between the age of orange trees and their production. A regular and persistent increase in average production occurs until the groves are at least 35 years old. The small number of groves listed that are 40 years old indicates a still further increase, but the number of examples is too small to warrant safe conclusions. Possibly this apparently regular increase is due to the method of grove selection ; certain it is that many groves have been abandoned in the past as unprofitable before reaching the age of 35 years. However, the data indicate that well-located groves may be expected to continue to yield satisfactorily at least to 35 or 40 years of age if given reasonable care. The actual yields for the different age groups are given in table 7. TABLE 7 Age of Trees and Production of Oranges Yield in pounds per acre Age Coastal Intermediate Interior 10 13,100 18,700 19,300 20,100 20,500 20,900 9,600 17,700 18,300 19,050 19,450 19,800 8,900 15 14,700 20 15,200 25 15,850 30 16,200 35 16,500 If the 35-year-old group is considered as being in full bearing, the younger groves may be represented by percentages of a full yield as given in table 8. The data are graphically presented in figure 2. TABLE 8 Relative Yields of Orchards by Ages when the 35-Year-Old Groves are Considered Full Bearing Age Relative yield 35 100 30 97 25 94 20 92 15 89 10 60 12 University of California — Experiment Station Re/af/ve Yields. /OO r- 90 80 70 60 SO fO /J ZO 25 30 35 Age of Trees in Years Fig. 2. — Relative yields of trees in various age classes ; average yield of 35-year-old groves = 100. (Illustrating table 8.) The rate of increase in yields is almost identical in each of the three climatic zones except that the 10-year-old groves produce rela- tively more near the coast than elsewhere. The three percentages stated separately for that age are : Relative yields 10-year group. Coastal Intermediate Interior 67 52 58 Lemon trees increase in yields at essentially the same rate as do orange trees. The 10-year-old group has been eliminated in the subsequent studies of the effect of orchard practices on yields. Bul, 374] Orchard Practices in the Citrus Industry 13 FERTILIZATION AND ITS EFFECT ON YIELDS The successful orchardists of the citrus industry have realized for some time that the supply of nitrogen available to the trees is one of the dominant factors of fruit production. The orchard trials at the Citrus Experiment Station (Rubidoux plots) have strongly em- phasized this fact during the past ten or more years. The same general conclusion was reached in the Citrus Experiment Station trials at Arlington and the cooperative trials that have been conducted in various parts of the state. Comparatively little evidence has ex- isted, however, as to the exact amount of nitrogen that should be used. 5 The present study gives a general basis for an answer to this question. In this study the total amount of nitrogen carried in con- centrated commercial fertilizers is considered, together with an esti- mate of the amount carried in all applications of bulky organic manures. Yields as affected by variations in nitrogen applications. — The data gathered in this survey show a regular increase in average yields until about 350 pounds of actual nitrogen are applied per acre. With larger applications there is an apparent decrease in yields. The observed yields with varying fertilizer applications are given as percentages of the average yields in table 9. This table applies equally to the groves from each of the climatic zones. The method of combining the zones is fully explained on page 39 of the appendix. The data are also presented in figure 3. TABLE 9 Nitrogen Applied and Relative Orange Yields Nitrogen in pounds Relative per acre yields 50* 76 100 94 150 106 200 114 250 128 300 135 350 143 400 130 * Given as the center of the class, the limits of which are 26 to 75, etc. 5 Vaile, R. S., Fertilizer experiments with citrus trees. California Agr. Exp. Sta. Bul. 345:508. 1922. (Out of print.) 14 University of California — Experiment Station /oo /so 200 250 300 350 A// frozen Applied /n Pounds fbr Acre Fig. 3. — Eelative yields obtained with varying amounts of nitrogen; average yield of all groves := 100. The circles represent average yields for each fertilizer class, as given in table 9. The curve represents the yield to be expected in response to the law of diminishing returns from additional applications of nitrogen, as given in table 27. These relative yields may be converted back into approximate, real yields by using" the average yield for district and variety as 100. Table 10 gives the yields that should be expected under average con- ditions with the use of the amount of nitrogen indicated. TABLE 10 Nitrogen Applied and Yields Yield in thousands of pounds per acre Nitrogen in pounds per acre Coastal Intermediate Interior Oranges Lemons Oranges Lemons Oranges Lemons 50 100 150 200 250 300 350 400 14.8 18.3 20.7 22.2 25.0 26.3 27.9 25.4 18.7 23.2 26.2 28.2 31.6 33.4 35.3 32.0 13.6 16.5 18.5 20.0 22.4 23.6 25.0 22.8 17.2 21.2 24.0 25.8 28.9 30.5 32.3 29.4 11.7 14 4 16.3 17.6 19.7 20.4 22.0 20.0 14.3 17.8 20.0 21.5 24.2 25.6 27.1 24.6 Note:— These figures are based on an average of fruit shipped. pounds of fruit picked to each packed box of All groves that received as much as 350 pounds of nitrogen per acre annually were badly mottled, although most of them were still producing better than average crops. A considerable amount of Buu 374] Orchard Practices in the Citrus Industry 15 mottle-leaf was observed on the groves receiving 250 to 350 pounds of nitrogen. This bears out the observations made on the Arlington experimental plots in 1915-16, when a heavy application of alfalfa hay, carrying nearly 400 pounds of nitrogen to the acre, was followed by a severe case of mottling which did not disappear for several years. Dr. I. G. McBeth has also reported that extreme mottling is often associated with high nitrate content in the soil. 6 It seems probable, because of this association, that the most productive use of nitrogen for citrus orchards is reached somewhere between 300 and 350 pounds per acre annually. The most profitable use will be discussed later. An interesting trial of the effect of varying amounts of nitrogen on yield has been conducted on a mature orange grove at Ontario by the Chaff ey Junior College of Agriculture. 7 Three sections of orchard have been given 100 pounds, 200 pounds, and 300 pounds of nitro- gen respectively. In each case one-half the nitrogen has come from manure and the other half from a quick-acting concentrate. For several years prior to 1921 the entire grove had been lightly but uniformly fertilized. The yield in 1922-23, after the second appli- cation of the varying amounts, was as follows : Pounds nitrogen Pounds of fruit per acre per acre 100 14,300 200 18,300 300 18,700 Three hundred pounds of nitrogen per acre gave the most pro- ductive use, but obviously 200 pounds represented the most profitable use in this case. Of the growers interviewed, 30 per cent are using only 100 pounds of nitrogen per acre while 75 per cent are using less than 225 pounds. Approximately half of the nitrogen applied has been carried in bulky organic manure. It seems probable that citrus production may be more surely and rapidly increased by increasing the applications of fertilizer than in any other way. To apply 350 pounds of nitrogen, half of which is carried in bulky organic manure, one must use 20 tons of stable manure together with 1500 pounds of dried blood or nitrate of lime per acre, or other materials that will supply equivalent amounts of nitrogen. These are large quantities and will cost not less than $150 per acre for fertilizer alone. Only a few growers use such amounts consistently year after year. s McBeth, I. G., Soil nitrogen and nutrition of qitrus plants. Jour. Agr. Kesearch 9:248-251. 1917. 7 The data of the Chaffey Junior College of Agriculture experiment were fur- nished through the courtesy of Mr. C. A. Booth. 16 University of California — Experiment Station Five tons of alfalfa hay, or eight tons of either Lima bean straw, first-grade chicken manure, or clean sheep manure might be substi- tuted for the twenty tons of stable manure. One thousand pounds of sulfate of ammonia or 3000 pounds of cottonseed meal carry approximately the same amounts of nitrogen as 1500 pounds of blood, The choice between the different materials within the two groups is probably to be made entirely on the basis of relative cost, for the difference in their fertilizing values is apparently proportional to the amount of nitrogen which they carry. In the discussion above, the most productive use of nitrogen has been considered. Later certain of the factors which determine the most profitable use will be presented. The use of bulky organic manures, and yields. — The Citrus Exper- iment Station field trials (Rubidoux plots) have emphasized the fact that concentrated nitrogenous fertilizers used persistently without bulky organic material will not permanently maintain healthy citrus trees under the conditions which prevail at Riverside. Causal field observations the state over bear out this conclusion. Groves are occasionally successfully managed for as long as ten years with almost all the fertilizer applied in concentrated form. The day of reckoning is, however, sure to come. In the writer's observations there are a score or more of groves that at one time were looked upon as 'show places,' but that have since deteriorated rapidly through failure to recognize this principle. The survey data have been tabulated in two ways in order to measure the effect of applications of organic matter on yields. First, all of the groves were grouped into three classes according to the proportion of the total nitrogen used which was carried in bulky organic manures. The class divisions were : (1) Less than 40 per cent of nitrogen carried in bulky manure ; (2) 40 to 60 per cent of nitrogen carried in bulky manure; (3) Over 60 per cent of nitrogen carried in bulky manure. The data show that the largest average yields resulted when the largest proportion of the nitrogen was carried in bulky manure. Call- ing the average yield in this class 100, table 11 may be compiled. TABLE 11 Kelation of Nitrogen in Bulky Manures to Yields Nitrogen in Relative bulky manures yields Over 60 per cent 100 60-40 per cent 93 Under 40 per cent 89 Buu 374] Orchard Practices in the Citrus Industry 17 The use- -of .-nitrogen so completely dominates yields that a direct correlation between manure used and yields is not high. However, as a second method of grouping, table 12 indicates that more manure does bring more fruit. The highest average class yield is arbitrarily taken as 100. TABLE 12 Belation of Applications of Manure to Yield Manure applied Relative tons per acre yields 0-4.9 86 5-9.9 92 10-14.9 95 15-19.9 100 20 and over 99 Throughout the study of organic manure, the high-grade materials such as alfalfa hay have been converted into equivalent amounts of ordinary manure by multiplying by the ratio of organic matter which they contain. Effect of cover crops on fertilizer requirements. — During the past ten years the use of winter green-manure crops in citrus groves has materially increased. In the winter of 1914-15 it was estimated that less than 40,000 acres of citrus orchards were planted to green-manure crops of any kind. Nearly half of this acreage was planted to cereal crops. In 1915-16 approximately 67,000 acres were planted to leguminous, green-manure crops, and in 1916-17 approximately 100,000 acres were so planted. Since 1917 there have been fluctuations in planting with an apparent upward trend. In an attempt to measure the immediate effect of winter green- manuring on citrus production, the grove records were first sorted according to the amount of nitrogen applied and then each class was subdivided into cover-cropped and clean-cultivated groups. The aver- age yields were practically identical for the cover-cropped and clean- cultivated groves in each nitrogen class. Apparently the use of cover crops does not diminish the necessity for nitrogen applications. The records were also divided into classes according to the amount of bulky manure used and subdivided into winter-cover-cropped and clean-cultivated groups. In every case the cover-cropped group gave a higher average yield than the clean-cultivated group. The difference was greatest where the smallest amount of manure was used, and became regularly less as the requirements for organic matter were better supplied from outside sources. Table 13 gives the comparative 18 University of California — Experiment Station yields when the clean-cultivated groves in each manure class are counted as 100. The per cent above 100 in the cover-cropped group is the increase to be expected from green-manuring, when definite amounts of manure are used. TABLE 13 Effect of Cover Crop on Average Yield of Groves Treated with Bulky Manure applied, tons per acre Relative yields Manure Without cover crop With cover crop 0- 4.9 5- 9.9 100 100 100 100 119 110 10-14.9 15-19.9 108 106 When the cover-cropped groves are compared with the clean- cultivated groves that received five tons additional manure each year the average yields are practically identical (100:101). It seems prob- able, therefore, that the persistent use of green-manure crops will relieve the growers of the necessity of applying such liberal applica- tions of manure. It does not seem probable, however, that mature groves can be maintained in the best condition without some applica- tions of manure, even when winter cover-crops are grown. Analysis made by W. P. Kelley indicates that, on the Rubidoux plots of the Citrus Experiment Station, a winter-green-manure crop and about ten tons of manure per acre annually have just kept the organic content of the soil equal to that of virgin soil. The plots so treated have produced more fruit than plots which have received thirteen tons of manure without the winter cover-crop. The experience of the Experiment Station on its Rubidoux plots where winter legumes have been successfully grown for sixteen con- secutive seasons indicates that a rotation of varieties is desirable. Common vetch, Melilotus, purple vetch, and horse beans have been used at various times. No single crop has been used for more than three consecutive years, and as a result the tonnage of green manure produced has always been heavy. Such rotation avoids some of the cumulative injury from aphis, mildew, and weeds because of the different habits and susceptibility of the several crops. The number of groves recorded in the survey in which summer cover crops have been grown for a five-year period is small. Such groves received large amounts of irrigation water, and their average Bul. 374] Orchard Practices in the Citrus Industry 19 yields were not so high as those of clean-cultivated groves receiving the same amount of water. In several instances the observation was made that the use of summer cover crops for three or more consecu- tive years resulted in very poor tree conditions. It may be possible to grow summer cover crops in citrus orchards to advantage, but methods of general applicability do not seem to have been worked out. Of the 1,000 groves listed, only two were found on which neither manure nor cover crop had been used for as long as five years. Although there is a general recognition of the necessity of maintaining the organic content of the soil, at least 75 per cent of the orchards could be expected to yield somewhat larger crops if more organic material were used. Very substantial improvement to 25 per cent of the groves should result from increased application of manure. IRRIGATION AND ITS EFFECT ON YIELDS Climate and soils each have an important effect on irrigation practice in citrus groves. The survey data indicate that less water is used in general near the coast than in the interior, and that the inter- vals between irrigations are longer in the cooler coastal districts. Climate, therefore, is one important variable. Different soils require different irrigation, even under the same climatic conditions. The amount of water that can be stored for use in a light soil is much less than in a heavy soil and it would, therefore, naturally be used up sooner. This condition calls for more frequent irrigations on light soils than on heavy ones where the climatic conditions are the same. There is no evidence, on the other hand, to show that a given crop will transpire more water when grown on one soil than when grown on another, unless the growth is more vigorous in one case than in the other. With minor exceptions citrus trees appear to grow equally vigorously on the various grades of soil in southern California, and therefore they should use about the same total amount of water, regardless of soil. The survey data bear out this reasoning. There is no material difference in the amount of water commonly applied on different soils, and the average yields with any given amount of water are approximately the same on the various soils being considered. Amount of water applied. — To compare the effect of increasing the amount of water used, the records from each climatic zone were divided into five groups or classes according as the total amount of irrigation water used was light, usual, moderate, heavy, or very heavy. 20 University of California — Experiment Station Different limits were set for these several classifications because the range of treatments was different in the several climatic zones. The most comon usage was made the central class for each zone and the maximum range of usage was equally divided into two classes that used more and two classes that used less than the usual amount. The class divisions are indicated in table 14. The class designated as * usual' represents the amount most commonly used. TABLE 14 Designation of Classes in Irrigation Acre inches applied per season Class Coastal Intermediate Under 16.5 16 5-214 21.5-26.4 26.5-31.4 31.5 and over Interior Light Under 14 14-17.9 18-21.9 22-25.9 26 and over Under 19 19-24.9 25-30 9 31-36.9 The average yields for orchards in these several classes are given in table 15. TABLE 15 Yields and Amount of Irrigation Water, Thousands of pounds per acre Class Coastal Intermediate Interior - Oranges Lemons Oranges Lemons Oranges Lemons Light 18.9 20 5 201 19 9 19.5 .,23.9 26.0 25.6 25.5 24.7 17,5 18.1 IS'5 2(1.6, 22. o' . . 22.7. 23.4 , 24 26.8 28:o 12.3. , 14. 9 J " ■! il.uT/7" 19.1 ' =17.7 •-'* 15.1 18 s 3- Usual '. .".... 19,0 Heavy 23.4 21.7 V In the coastal zone a little less than the amount of water now commonly used gives the highest yield. There is considerable .evidence of injury from over-irrigation throughout this zone, particularly , on heavy soil. This is in general agreement with the findings of Thomas in his study of citrus-grove irrigation in the vicinity of Whittier. 8 s Thomas, E. E., Studies on the irrigation of citrus groves. California Agr. Exp Sta. Bui. 341:353-370. 1922. (Out of print.) Bul. 374] Orchard Practices in the Citrus Industry 21 In each of the other zones the average yields increase with the use of more water until nearly 50 per cent more than the usual amount is applied. Even in these zones, however, the injurious effects of over- irrigation are noted on certain groves that are very heavily irrigated. Irrigation water in the intermediate and interior zones is so scarce and so expensive that the most profitable use may be reached some- time before the most productive use. This question can only be decided when the cost of developing additional water for a given locality is known. Frequency of irrigation. — The records indicate that longer inter- vals' between irrigations are more desirable near the coast than in the interior. The highest average yields for the groups divided by intervals occurred in the following classes : Coastal, 34-41 days or approximately 5 weeks ; Intermediate, 25-33 days or approximately 4 weeks; Interior, 18-24 days or approximately 3 weeks. The most common irrigation interval in the intermediate district is apparently well adapted to the needs of the trees (25 days). The present common usage in the interior allows too long an interval between irrigations (30 days) for maximum average yields, while near the coast the common interval (33 days) is too short for most productive use. These findings in regard to intervals between irrigation are borne out by soil moisture studies in the several districts. , ( PLOWING AND ITS EFFECTS ON YIELDS : Jitst about half of the groves recorded were plowed each spring during the five-year period. Many of the rest were never plowed, while some were plowed in part of the years only. ** f - There was no difference in average yield between the plowed and the unplowed groves. The occasionally plowed groves had an average yield 15 per cent less than the other groups. It would be difficult to say whether the irregularity of plowing actually caused the de- crease in yields or whether the growers were attempting to correct some other unfavorable condition by modifying their orchard pro- gram. The evidence seems to point to the former — that is, it appears that irregular cuttings of roots by occasional plowing may be ex- pected to temporarily reduce yields, while regular, annual plowing does not haive: such an effect. . ... v:'r 22 University of California — Experiment Station It is of interest that in certain communities no one plows because of an opinion that to do so reduces yields, while in adjoining com- munities such a prejudice is not held and everyone plows. Neither policy has any immediate effect on average yields of a large group of groves, although there are apparently a few authentic cases in which plowing at the wrong time in relation to fruit setting or to soil moisture has resulted in a heavy drop of fruit. It seems, therefore, that in general, good farming practice should dictate the policy to be followed. All the successful schemes of crop rotation include a deeply tilled crop and the use of a legume; plowing and green-manuring are fundamental features of world agriculture. They can both be used — and combined — in citrus culture. The evidence of this survey of growers' experience is that plowing down a winter green-manure crop each spring certainly does no harm. Agricultural history teaches that such a practice is one approach to permanent agriculture. Therefore, wherever possible, winter legumes should be grown and plowed down to a greater depth than the subsequent summer cultivation. It should be remembered, however, that the cover crop may require some extra irrigation water, especially in years of light rainfall. THE ' LIMIT OF PROFITABLE CULTIVATION' The produce of a given piece of land cannot be doubled, trebled, quadrupled, and so on indefinitely, by merely doubling, trebling, and quadrupling the amount of labor and capital expended in its culti- vation. In the data of this survey, the increases in yields following additional nitrogen applications grow constantly less until a point is reached where no increase in crops is experienced. Diminishing physical returns from each additional application of nitrogen are found beginning with the first class into which the data are divided. The 'limit of profitable cultivation' is in general a different thing from either the point of diminishing physical returns or that of most productive use. It represents the most profitable use of the combined factors of production. It is usually reached after diminishing returns set in and before the most productive use is realized. Occasionally it goes outside of these limits. Profits, as ordinarily conceived and as used in this discussion, may be divided into two separate concepts : (1) Returns on capital invested; (2) Payment for assuming the risk of doing business. BUI,. 374] ORCHARD PRACTICES IN THE ClTRUS INDUSTRY 23 No attempt will be made in this discussion to separate these two factors, but both will be included in the term 'profit,' and will be referred to in terms of an interest rate on the investment. Agricultural lands are capitalized in proportion to their earning power. The flexibility of this capitalization with changes in earning power was well illustrated by the rapid rise in the selling prices of corn-belt farms during the war period of high prices for agricultural products, followed by the rapid slump in land values when the prices paid for agricultural products dropped more rapidly than the cost of their production. Ordinarily the demand for agricultural lands is strong enough to keep their capitalization high in proportion to their earning power. Farm management surveys made in various states indicate that on the average, farms are earning only about 3 per cent on the farmers ' estimate of their sales value. Among other data secured from growers in the present survey were their estimates of the present sales value of the groves, as well as their estimate of the actual cost of production of fruit during the past five years. These figures have been compared with the average production in pounds of fruit for the different districts and with the average selling prices of citrus fruits on the trees for the past ten years. The resulting figure shows an average profit of between 3^2 and 4% per cen t for the groves included in the survey. In any consideration of the relative profitableness, either of indi- vidual orchards or of varying cultural practices, at least four factors must be taken into consideration, namely : (1) Cost of production per pound of fruit ; (2) Sales price for the fruit ; (3) The amount of fruit ; (4) The value of the land upon which returns must be paid. The margin between the cost of production and the selling price per pound, multiplied by the yield gives the total net return. This figure divided by the valuation of the orchard gives the per cent of return on this valuation. Each of these factors plays so important a part in the answer to the final question of net profit that it is impossible to consider any one of them without assuming a fixed point or a regular schedule for each of the others. Cost of fruit per pound. — Probably the cost of producing a pound of fruit has most frequently been used as a measure of the comparative merits of two orchard enterprises. This, however, is by no means a sure guide, as is illustrated by the figures in table 16, comparing data for oranges from two southern California districts. 24 University of California — Experiment Station If an interest charge on an estimated valuation of groves in each of the above districts has been included in the cost of production, the costs per box would measure the relative accuracy of such estimated valuations. I| the costs per box, including interest, were equal, the valuations used Would be relatively comparable. If one group showed a higher cost, including interest, than another, its estimated value would be relatively too high. Except when used in this manner, the cost of production per box is important as only one factor in an equation and is not of itself a measure of the merits of the enterprise. TABLE 16 Illustration of the Difference in Valuation per, Acre of Orange Orchards that May Exist Even when Cost of Production per Pound of Fruit is the Same (One acre) Items District No. 1 District No. 2 $155.00 143 1 08 185.00 30.00 1,000 00 $230 00 225 1.07 Returns for fruit on trees Farmer's return on capital Capitalization that will allow 3 per cent return 295.00 65.00 2,165 00 TABLE 17 Net Keturns per Acre for Oranges and Lemons Separately in Each of the Climatic Zones, with Present Cultural Practices Zone Yield, packed boxes Ten-year average price* per packed box Gross returns Total operating costs. t No interest included Net returns Oranges : Coastal Intermediate Interior 244 230 192 $1.50 1.40 1.30 $366.00 322.00 250 00 $230 00 229.00 188 00 $146.00 93.00 62.00 Lemons: Coastal 310 300 236 $1.30 1.30 1 30 $404.00 390.00 307.00 $265 00 274 00 230.00 $139.00 116.00 Interior 77 00 * Valencias have sold for about 45 cents per packed box more than Navels on a ten-year average- On a five-year average this difference has been less than half as much. The different prices assigned to the several zones make allowance for the different proportion of Valencias, on the basis of the ten- year difference in price. t Differences in costs between the zones are due in the main to different fertilizer practices, different pest control needs, and different tax assessments. Bul. 374] Orchard Practices in the Citrus Industry 25 Net returns per acre. — Table 17 reports the data collected in the survey that are necessary to figure the net returns per acre for orange and lemon groves separately in each of the climatic zones. The per-acre yields and operating costs are the averages from the survey reports from each zone. The prices used are the approximate ten-year average prices (1912-1922) paid to the growers for fruit on the trees by the associations affiliated with the California Fruit Growers' Exchange. Net returns per acre are indicative of relative valuations of orchards on an acreage basis. This criterion of value must not be confused, however, with the factor of net returns on the investment. Financially the citrus grower is interested in the relation which exists between net returns per acre and the present exchange value of that acre. Net returns on the investment. — The averages of growers' own estimates of the present exchange value of their groves are widely different for the three climatic zones. In table 18 the returns per acre are converted into terms of returns on a given financial investment. The growers' estimate of valuation is the datum with which the calcu- lations begin. The illustration is based on an investment of $20,000 because that is the approximate average (modal) value of California citrus orchards. TABLE 18 Average Net Keturns on the Investment for Oranges and Lemons Separately in Each of the Climatic Zones, with Present Cultural Practices Growers' estimated valuation, per acre Net returns obtained on estimated value, per cent What to expect from a $20,000 investment at growers' estimated value Zone Acres pur- chased Net returns to capital Hours of labor that owner might perform* Value of owners' labor* at 30 cents per hour Total possible income to labor and capital Oranges : $4,100 2,680 1,540 3 56 3.48 4 02 4.9 7.5 13.0 $711 696 804 640 975 1,690 $194 292 506 $905 988 1,310 Lemons: $3,500 2,600 1,800 4.00 4.45 4.28 5.7 7.7 11.1 $800 890 835 865 1,230 1,780 $260 370 535 $1,060 1,260 1,370 * These columns refer only to man labor and are based on the average experience of 75 growers who furnished diary reports to the writer from 1915 to 1918. The average was 130 hours manual labor per acre annually on oranges and 160 hours on lemons. This does not include picking or hauling of fruit, or fumigating labor. The difference between oranges and lemons is due mainly to different pruning requirements. 26 University of California — Experiment Station The most profitable use of nitrogen. — Changes in orchard practices will ordinarily modify both the operating costs per acre and the yields. The complex reaction of such changes on the net returns on the invest- ment can best be illustrated by considering the most profitable use of nitrogen in fertilizing orange groves. The per-acre yields of oranges with various applications of nitro- gen are recorded in table 10. The operating costs for each climatic zone when the average amount of nitrogen is used are given in table 17. These costs may be varied for different nitrogen applications on the assumption that all other operations remain constant. The most profitable use of fertilizer will also depend in part on the price at which the fruit may be sold. In order to make the example complete, the effect of two different selling prices may be considered — one a little below the ten-year average and the other somewhat above it. TABLE 19 The Most Profitable Fertilizer Practice with Two Different Prices of Fruit Oranges Nitrogen in pounds per acre Yield in thousands of pounds per acre Total cost per acre varied by nitrogen used Net returns per acre Net returns on $20,000 investment Zone With price of lHc for fruit on trees With price of 2J^c for fruit on trees With price of lj^c for fruit on trees With price of 2Hc for fruit on trees 350 27.9 $360 $59 $336 $290 $1,645 300 26.3 330 65 327 319 1,605 250 25.0 300 75 325 368 1,590 200 150 22.2 20.7 270 240 72 71 285 277 353 348 1,395 1,355 100 18.3 210 65 247 319 1,210 50 14.8 180 42 190 206 930 350 25.0 $350 $25 $275 $188 $2,060 300 23.6 320 34 270 248 2,010 250 22.4 290 40 270 300 2,010 200 150 20.0 18.5 260 230 46 48 240 232 345 360 1,800 1,740 100 16.5 200 48 212 360 1,580 50 13.6 170 14 170 105 1,275 350 22.0 $340 $-10 $205 $-130 $2,660 300 20.4 310 - 4 210 - 52 2,730 250 19.7 280 14 212 182 2,760 Interior 200 17.6 250 16 190 208 2,470 150 16.3 220 25 187 325 2,430 100 14.4 190 26 170 338 2,210 50 11.7 160 16 132 208 1,730 Bul. 374] Orchard Practices in the Citrus Industry 27 Table 19 presents a summary of the data necessary to determine the most profitable use of nitrogen under the conditions outlined above. Figure 4 presents the same data graphically. A cost of $30 for 50 pounds of nitrogen may seem excessive. It is based on the high prices paid for fertilizers during the five years in- cluded in this study. The assumption is also made that at least one- half of the nitrogen comes from bulky organic manures — generally an expensive source when nitrogen alone is considered. Of course the figure used includes all costs of application. When nitrogen can be purchased for less money the details of the example will be changed. Net Her urn On a *20,0O0 ^ Investment *' Interior Zone ? 250O. r- '2000. - */,5DO. When the Price of fruit is 2^ rT?r Pound 'AOOQ "500 h When the Price of fruit /s /j * Per Pound SO / / — • / / / / / / „ *— * " — _•" - _ ^" - — Intermediate Zone Coastal Zone / /' S"' 1 ——______Coosro/ Zone ■ 1 1 ~- --■ 1 J n term ediore~Z one" ^, 1 1 too /so A// trooen ZOO 250 Pounds r~>er Acre tofigg-vo Fig. L — Net returns with varying fertilizer practices in three climatic zones and with fruit at 2y 2 cents and at 1% cents a pound. The highest point in each curve shows the most profitable practice for the conditions specified. Illustrating table 19. Two deductions, at least, may be drawn from the financial sum- maries submitted. In the first place, heavier fertilization is more profitable when the prices are high than when they are low. If there were assurance of high prices growers generally would be justified in fertilizing almost to the limit of maximum physical production. Certainly present applications could then be profitably doubled. 28 University of California — Experiment Station In the second place, when prices are low, the high-yielding zones (coastal and intermediate) have a higher potential earning power on the capital invested than the interior zone. When prices are high, on the other hand, the advantage is definitely in favor of the interior district. The summary, in table 20, of the net returns for the most profitable fertilizer group brings out this point. TABLE 20 Average Net Keturns on $20,000 Investment, for the Most Profitable Fertilizer Group Fruit V/ic Fruit 2 l Ac Coastal $368.00 $1,645.00 Intermediate 360.00 2,060.00 Interior...- 338.00 2,760.00 The figures presented above may have little permanent value, except as illustrations of a method of analyzing the citrus business. They are based entirely on present land valuations, present operating costs, and past selling prices for fruit. As any of these factors change, the final result in net income will also change. The financially successful citrus grower of the future will be the one who most accu- rately anticipates changes in any of the factors and acts upon such anticipation. Obviously, recommendations as to the 'best' fertilizer practice to follow can be made only when all the factors considered above are known. COST OF DEVELOPING CITRUS ORCHARDS It has been stated above that agricultural land is capitalized on its earning power, and figures have been submitted which indicate that average orange groves are paying net returns of 4 per cent on $3,650 per acre near the coast and on $1,540 in the interior. It is equally true that when sufficient time is allowed for the thorough establishment of an industry, the value of a single unit, such as an orchard, will closely approximate the cost of establishing such a unit. This, of course, does not apply where partial or com- plete monopoly exists. It is clear, however, that the orange industry of California cannot be considered a monopolized industry at present, for there is much more good land in sight than there is good market. Bul. 374] Orchard Practices in the Citrus Industry 29 The past ten years have seen extensive new plantings of oranges along the Gulf coast and in Florida, while California acreage has increased relatively slowly. The failure of orange prices to rise during the war period in keeping with other prices is another indication of the lack of monopolistic character in the industry. TABLE 21 Cost per Acre of Developing Young Orange Orchards in Riverside and San Bernardino Counties, 1914-1919 Land with water, ready to plant $400.00 Trees planted 100.00 Pipe lines (allowing one line across a square ten acres) 25.00 $525.00 Cost per year for first four years: Labor and team work $ 20.00 Fertilizer and cover crop * 6.00 Taxes 5.00 Water charges 14.00 Miscellaneous 7.00 Administration, superintendent, etc 8.00 $60.00 $240.00 Cost per year, 5th-6th years, inclusive: Labor and team work $ 25.00 Fertilizer and cover crop 20.00 Taxes , 7.00 Water 20.00 Pest control 6.00 Miscellaneous 8.00 Administration 9.00 $95.00 $190.00 Cost 7th to 10th years, inclusive: Labor and team work $35.00 Fertilizer 40.00* Taxes 11.00 Water 20.00 Pest control 10.00 Miscellaneous 12.00 Administration 12.00 $140.00 $560.00 Total operating costs for first ten years $1515.00 * 10th year, $20.00 additional fertilizer. 30 University of California — Experiment Station In consequence of this great general situation, the value of orange groves should closely approximate the value of land for other farm crops plus the cost of bringing an orchard into bearing. In the coastal districts prospective citrus plantings must compete with bean lands, which are often priced at $800 to $1,200 per acre for Lima bean production. In the interior the competition is with deciduous fruits, alfalfa, grain hay, and a few other crops, and here land with water is priced at about $400 per acre. During the five years 1914-1919 the writer summarized the cost of developing orange groves on nearly one hundred different projects in the interior districts varying in size from 5 acres to 100 acres. These figures, together with the original land values just cited, have been used to construct table 21, on cost of establishing a young grove. In table 22 the detail figures from table 21 are summarized, together with credits to be expected from sales of fruit, and interest on the total money in the capital account. TABLE 22 Developing Orange Orchards — Interior Zone. Accumulated Cost with Interest Charged and Sales Credited Operating cost Credit for sales Net cost Net credit Interest at 4 per cent Total accumu- lated cost $525 60 60 60 60 95 95 140 140 140 160 $525 60 60 60 60 95 56 62 30 $525.00 $21.00 24.24 27.61 31.11 34.76 39.94 43.68 48.00 51.13 52.53 606.00 690.24 777.85 868.96 998.72 $39 78 110 156 195 1,094.66 1,200.34 1,278.34 $16 35 1,313.47 1,331.00 These figures on the cost of developing orange orchards approach closely to the grower's average estimate of values of mature groves in the interior districts. Presumably the higher price of land and consequent higher interest charges will produce a similar agreement with the orchard values in the other districts. The orange industry in California has apparently reached a point where rapid increase in plantings is not to be expected because the land is just as valuable for other — and possibly less hazardous — purposes. Bul, 374] Orchard Practices in the Citrus Industry 31 CONCLUSIONS Citrus groves produce more fruit per acre near the coast than in the interior. They do not, on the other hand, return higher net profits on present valuations. The middle range of soils (by texture), which are mainly used for citrus planting in southern California, are about equally well adapted to this purpose. Very sandy soils and heavy clay adobes are, in general, less productive of citrus fruit. Citrus trees may be expected to increase in average yield until at least 35 years of age. There is no evidence that 40-year-old trees produce less fruit because of their age. Nitrogen and bulky organic manures appear to be the only fer- tilizing constituents that need to be supplied for successful citrus production in southern California. About half the nitrogen now applied to citrus groves is carried in bulky organic manures. Average yields increase with increased applications of nitrogen until about 350 pounds per acre are applied. The point of most profitable use depends on several variable factors. Examples are presented illustrating the interplay of these factors. Mottling appar- ently results from excessively heavy applications of nitrogen. Orchards winter-cover-cropped gave higher yields than clean- cultivated orchards which received the same quantities of bulky organic manures. When grown for several years, summer-cover crops seem to result in poor tree condition. The use of cover crops does not seem to lessen the need for nitrogen. Rotation of leguminous cover crops is desirable. Irrigation has an important effect on yields. In general, less water is and should be used near the coast than in the interior. The interval between irrigations should be longer in the cooler coastal districts. The citrus industry seems to have reached a stage of stability, for there is a close agreement between (1) the grower's estimate of the exchange value of groves, (2) the capitalization of the present earn- ing power of groves, and (3) the cost of developing new groves from land purchased at its present valuation for other crops. 32 University of California — Experiment Station APPENDIX Homogeneity and accuracy of the data. — In order to make the individual grove records as nearly comparable as possible, the orchard practices of five specific years (1917-18 to 1921-22) were taken in all cases. For instance, the fertilizer usage for these five years was listed, and the average amount applied each season was found for each individual grove. This average should represent normal prac- tice rather closely. In the case of yields only the three middle years of the five were averaged. The 1917-18 crop was affected so seriously by heat in certain sections that it was discarded, and similarly, the 1921-22 yields were omitted because of the varying degree of frost injury. The yield figures were always checked with the packing-house statements, and the irrigation figures with the irrigation companies' records. Fertilizer records could not be so easily checked, but because of th*e relative importance of this item a written memorandum of amounts applied had usually been kept by the grower. The total number of groves actually used in the calculation was 600. The effect of the number of groves upon the class averages is indicated by tables 23 and 24, summarized at different times during the investigation. TABLE 23 Effect of Nitrogen" on Yields Relative yields Pounds N per acre 1st 50 groves 1st 300 groves Total 600 groves 350 100 100 100 300 102 96 94 250 95 91 89 200 86 83 78 150 82 70 73 100 74 61 66 50 53 57 57 TABLE 24 Effect of Age of Trees on Yields Relative yields Age 1st 50 groves 1st 300 groves Total 600 groves 35 100 100 100 30 99 98 97 25 97 95 94 20 93 92 92 15 83 85 89 10 46 57 60 Bul. 374] Orchard Practices in the Citrus Industry 33 Methods used in analysis. — In the handling of survey data the method usually followed has been to sort the records into classes according to some one variable factor and to determine the average yields or profits for each class. The differ- ences between these averages have been taken as measures of the effect of changes in the variable factor, on the assumption that the law of large numbers has equalized any effects of other factors. Sometimes, when it was obvious that secondary factors were affecting the average of one class to a greater degree than that of another class, the records have been further divided into sub-classes. Frequently, however, this method has reduced the number of records in the sub- classes to a point where the probable error of the average made the differences between class averages of no significance. Confidence in the class averages as ordinarily obtained may be greatly strength- ened if correlation coefficients are determined between the variables that are being compared. When inter-relations exist between several variables, the net agreement between any two of them may be determined by the use of the partial correlation method. This technique has been successfully applied to an analysis of survey data on the fattening of baby beef. 9 Correlation coefficients do not take the place of group averages, for they do not measure the value of one variable which corresponds to a particular value of another variable. What they do measure is the degree to which movements of one variable are accompanied by movements in another variable, and also, whether the movements are in the same or in opposite directions. When correlation coefficients are high, differences in class averages are of significance; when the coefficients are low, differences in averages are probably due to imperfect sampling. In the present study the correlation technique has been utilized extensively. Classification of the variables. — The factors upon which information was gathered fall into two main classes. Certain of the variables, such as amount of fruit produced, fertilizer used, irrigation water applied, and so on, can be. measured quantitatively. The agreement between changes in any of these things can be studied statistically by recognized methods for quantitative variables. Other items, such as climatic environment and soil type, cannot be classified quantitatively and are therefore subject only to those types of analyses that apply to qualitative variables. The list of variables upon which data were obtained and their division into statistical classes is given in table 25. Because of the separation of variables into qualitative and quantitative groups, it became desirable to use the statistical methods both of association and cor- relation in the study of the data. Fortunately for the subsequent analysis a high degree of independence was found to exist among the several variables in each of the groups of the qualitative classifications. That is, for instance, the same proportions of heavily, moderately, and lightly fertilized groves were found in each of the three geographic districts — and similarly on each of the three soil types. 9 Tolley, K. H. The theory of correlation as applied to farnvsurvey data on fattening baby beef. U. S. Dept. Agr. Dept. Bui. 504:1-14. 1917. 34 University of California — Experiment Station TABLE 25 List of Variables Upon- Which Data Were Obtained Quantitative Variables Classes Variable Num- ber Centering at Centers of extreme classes 7 8 5 5 7 10 multiples of 5 years multiples of 50 pounds per acre multiples of 5 tons per acre limits dependent upon climatic district weekly intervals multiples of 20 per cent of the average yield 50 and 400 pounds 4. Average annual application of irrigation water... 5. Interval between irrigations during summer 6. Yields light and very heavy irrigation 2 and 8 weeks 40 per cent and 200 7. Costs. per cent Qualitative Variables Variable Classes Num- ber 3 3 6 6 Designations 1 I 3. 4. Tillage Cover crops number of years plowed during past five number of years grown during past five Association of variables. — Coefficients of association were determined between each of the several qualitative classifications of climate, soil, tillage and cover crop and the quantitative variables of fertilizer and irrigation practices as well. The standard formula for this purpose was used : Q AB .ab—Ab-aB AB . ab + Ab . aB in which the difference between the cross products is divided by the sum of the cross products. In the formula the A 's and the B 's represent the possession of an attribute, such as location in the coastal zone (A) and on heavy soil (B), while the a 's and b 's stand for the lack of that particular attribute. In complete asso- ciation Q = 1 because the second term in both numerator and denominator is 0. In complete disassociation Q = — 1 because the first term in both numerator and denominator is 0. In complete independence Q — because the numerator is 0. Nearly all of the coefficients thus determined lay between — 0.15 ± .06 and -f 0.15 ± .06, showing a high degree of independence. The following exceptions may be noted: Bul. 374] Orchard Practices in the Citrus Industry 35 1. There is a higher proportion of young groves (ten-year class) in the coastal zone than in the rest of the population, for when A = all groves in the coastal zone, and B = all groves in the ten-year class, a = all groves not in the coastal zone, and b = all groves not in the ten-year class, then Q = 0.50 ±: .04, indicating a greater association between the B 's and A '& than between the B 's and a 's. This is of no significance to the subsequent analysis, because all the groves in the ten-year class were eliminated. 2. The amount of water used in the coastal zone is more nearly uniform than in the other zones, for when A = all groves in the coastal zone, and B = all groves using the ' usual 'io amount of water a = all groves not in the coastal zone, and b = all groves using other than the ' usual ' amount of water, then Q = 0.60 ± .03, indicating that a higher proportion of groves in the coastal zone (A) used the usual amount of water (B) than was the case in the other zones (a). This is equivalent to saying that there is a more pronounced mode of irrigation usage in the coastal district than elsewhere. The range of usage is almost as great in the coastal zone but the non- modal classes are not so well represented. 3. There is a higher proportion of heavily fertilized groves in the intermediate zone than in the other zones, for when A = all groves in the intermediate zone, B = all heavily fertilized groves, a = all groves not in the intermediate zone, b = all groves not heavily fertilized, then Q = 0.40 ± .04 showing that relatively more of the heavily fertilized groves (B) were in the intermediate zone (A) than elsewhere (a). Allowance was partially made for this fact in the figures given in table 10 by using the different observed mean nitrogen applications for the three zones as the datum from which the relative yields were converted into pounds. Averages based on law of large numbers. — Inasmuch as the independence between the several qualitative variables is so high, the law of large numbers should act so that the mean figures for the different classifications are significant of the effect of the qualitative variables. For instance, mean yields for the three climatic zones should be an approximate measure of the effect of climate on pro- duction, and mean yields for the three soil types should be a fair measure of the adaptability of soils to citrus production. The other variables, being independent in their movements, will not affect such means. io For explanation of this term see page 20. 36 University of California— Experiment Station After ascertaining the essential independence of the variables,; the mean yield was determined for each class in the several quantitative arrays, such as ■: age of trees, amount of nitrogen, manure, and water used. These figures -represent the normal effect that may be expected to accompany changes in. age or in orchard practices. Partial correlation coefficients. — In order to study more completely the com- plex relations between the quantitative variables, partial correlation coefficients were determined on six variables, namely: (a) Age of trees; (fr) Annual application of nitrogen; (c) Annual application of manure; (d) Annual application of irrigation water; (e) Interval between irrigations; (/) Yields. The usual formula for partial correlation was employed, in which r ab-cde — ^af-cde • 1"bf-cde r ab-cdcf — , /-,■> \ /-i V (1 — r'af.cde) (1 — r'bf.cde) In this case r a b-cdef represents the intensity of agreement between a and b when their separate agreement with c, d, e, and / is taken into account. A complete list of the net correlation coefficients is given in table 26. The more significant net and partial coefficients are commented on in the following paragraphs. The correlation coefficients between nitrogen used, and yield when age is taken into account, are a measure of the intensity of agreement in the movements of the two variables. These coefficients are all high in relation to their probable errors. The coefficients are: Coastal 0.612±03 Intermediate 0.624±.03 Interior \ 0.690i.03 In the interior and intermediate districts there is a tendency to fertilize more heavily as the groves grow older. This tendency does not appear in the coastal district. The correlation coefficients between nitrogen used and age of trees, when the items of manure and water applied and intervals between irrigations are taken care of, are: Coastal 0. 112±.06 Intermediate 0.510±.03 Interior 0.654±.03 The net coefficients between nitrogen and yields, after all other quantitative factors are taken into account, are: Coastal : 0.372±.04 Intermediate 0.626±.03 Interior 0.543±04 Bul. 374J Orchard Practices in the Citrus Industry 37 The agreement in the coastal zone is slight, in part at least, because some young groves are heavily fertilized while the soil is still so rich that it will not respond. On the whole, these coefficients are high enough to add considerable confidence to the figures obtained as average yields following different nitrogen applications. The coefficients between amount of manure used and amount of nitrogen used indicate that increases or decreases in the use of these two items occur together. This is to be expected inasmuch as half the total nitrogen comes from manure as an average. It appears that when the interval between irrigations is shortened, the total amount of water used in the season is increased. This has resulted, especially in the interior, from a definite attempt on the part of growers to use more water by furnishing occasional extra applications. The coefficients between yields and amount of water used are low. This really bears out the data presented in table 15. The correlation formula used is that of a straight-line fit, while the effect of increased applications of irrigation water is first to increase and later to decrease yields. Naturally such data show little agreement with a straight-line increase. Increased applications of nitrogen are accompanied by increased use of water in the interior and intermediate zones. Apparently certain farmers are working their orchards more intensively in both these ways. On the whole, the correlation coefficients may be said to materially increase the confidence in the significance of the averages presented in the main text of the bulletin. TABLE 26 Net or, Final Correlation Coefficients These coefficients measure the intensity of agreement in the movements of the two variables concerned, when the effect of the four other factors is taken into account. Net coefficients for each zone separately Factors compared Coastal Intermediate Interior -0.070±.05 -0.069±.05 0.234±.05 0.372±.04* 0.052±.05 -0.234±.05 0.019±.06 0.082±05 0.204±05 -0.078±.05 0.285±.04 0.026± 05 0.389±.04* 0.134±.05 -0.124±.04 -0.236±.04 0.204±04 0.273±04 0.626±03* 0.150±04 -0.507±.03* -0.330±04 0.444±03* -0.108±.04 -0.283±04 . 0.760±02* 0.624±.03* 0.638±03* 0.351±04 0.487±03*. — 020±.06 249±.05 0.054±06 0.543±04* 0.172±.05 Frequency of irrigation and amount of water used Frequency of irrigation and amount of manure used... Frequency of irrigation and amount of nitrogen used.... -0.688±.03* 0.106±05 0.937±.01* -0.673±03* Amount of water used and amount of manure used.... Amount of water used and amount of nitrogen used.... 0.436±.04 0.665±03* 0.632±03* Amount of manure used and amount of nitrogen used . 0.523±04* 0.316±05 0.654±.03* Large enough in proportion to probable error to show some agreement in movement. 38 University of California — Experiment Station Method of combining groves from the three climatic zones. — The effect of fertilizer treatments on yields appears to have been very similar in the three climatic zones. The correlation coefficients between yields and nitrogen applied when age of trees is taken into account are : Coastal r = 0.612=fc.03 Intermediate r=0.624±.03 Interior r = 0.690±.03 These coefficients were derived from the formula r ab r ac • Tbc 1'ab-c V (1 — r\ c ) (1 — rVO where a = yield ; b — amount of nitrogen used; c =age of trees. Because of this apparent agreement in effect, the records from the three zones were combined to give the law of large numbers a better chance to operate. This was accomplished by reducing the yield of each grove to a relative of the average yield in its zone. The yields of individual orange and lemon orchards were divided by their respective average yields. Then all of the relatives so found were com- bined into a single series for comparison with fertilizer treatments. For this pur- pose the groves were grouped according to the total amount of nitrogen applied and mean yields were determined for each group. In figuring the amount of nitrogen applied, an estimate was made of the amount carried in any bulky organic manures used, and this was added to the amount carried in commercial fertilizer. Diminished increment of physical returns with increased use of fertilizer. — There is usually a diminished increment of physical return with increased applica- tions of anything which normally causes increased yield. Many series of ferti- lizing and feeding data form the terms of decreasing geometric series, in response to the law of diminishing returns. 11 If the formula for the sum of a decreasing geometric series is applied to the yield data in this survey and the several points corresponding to varying nitrogen applications are determined, there is a very close agreement with the observed facts. The formula referred to is S = — • (1 — R») 1 — R V ' in which S is the sum of n terms, a is the first term and B the ratio of any term to the preceding term. The observed relative yields are shown together with the calculated yield for diminishing returns in table 27. ii Spillman, W. J., Jr. Application of the law of diminishing returns to some fertilizer and feed data. Farm Econ. 5(1) :36-52. 1923. Bul,. 374] Orchard Practices in the Citrus Industry 39 TABLE 27 Diminishing Physical Beturns from Increasing Applications of Nitrogen pounds per acre Relative yields Nitrogen in Observed Theoretical 50 76 76.5 100 94 92.5 150 106 105.5 200 114 117 250 128 127 300 135 135.5 350 143 143 This relationship is illustrated in figure 3. 7m-6,'29