UNIVERSITY OF CALIFORNIA COLLEGE OF AGRICULTURE AGRICULTURAL EXPERIMENT STATION BERKELEY, CALIFORNIA GRANULATION OF VALENCIA ORANGES E. T. BARTHOLOMEW, WALTON B. SINCLAIR, AND F. M. TURRELL BULLETIN 647 February, 1941 UNIVERSITY OF CALIFORNIA BERKELEY, CALIFORNIA CONTENTS PAGE Introduction 3 Description of granulation and dry juice sac 4 Field experiments 5 Preliminary tests, 1930-1934 6 Trees in cheesecloth tents (1931) 6 Excessive growth of fruits at stem end and granulation 7 The effect of sprays (1931-1934) 8 Tree differences and constancies in the production of granulation 12 Experiments in budding and top-working 16 Granulation and scaly bark 17 Reduction in frequency of application and in amounts of irrigation water .... 18 Soil characteristics and sampling 19 Irrigation treatment 19 Granulation tests 20 Fruit sizes and yields 22 The effect of the rootstock on the production of granulation 24 Granulation and low temperatures 27 The effect of limb girdling on fruit size and granulation 29 Granulation in relation to size of fruit and location on tree 35 Severity of granulation 39 Laboratory experiments 39 Stages of granulation and volume of juice sacs 39 Nitrogen in juice sacs 40 Reducing and total sugars and total soluble solids 41 Titratable acidity and pH 45 Total pectin as calcium pectate 45 Moisture content of juice sacs 46 Dry matter and inorganic constituents 47 Structural and other changes in juice sacs during granulation 50 Enlargement of juice sacs 50 Hardening of juice sacs and thickening of cell walls 51 Juice-sac collapse 52 Lignification of cell wall 52 Gas bubbles 53 Decrease in color 53 Causes of granulation 54 Causes of dry juice sac 57 Control measures 57 In the field 57 In the packing-house 58 Summary 59 Acknowledgments 62 Literature cited 63 GRANULATION OF VALENCIA ORANGES E. T. BAETHOLOMEW, 3 WALTON B. SINCLAIR, 4 and F. M. TUEEELL 5 INTRODUCTION Granulation may occur in the Valencia orange, Citrus sinensis Osbeck, in all areas in California where this fruit is grown. Although it has been recognized in this state for many years, its economic importance was not fully realized until production reached the place where fruits of ques- tionable quality had to give way to those of the best quality. At the present time, about 5 to 15 per cent (the percentage depending on the locality) of the Valencias picked in southern California from July to Oc- tober cannot be shipped because of granulation. In this country, granulation is found not only in California, but also in Arizona, Texas, and Florida. It is probably present in all countries where citrus is grown ; for it has been reported and accurately described from Palestine, Egypt, India, Siam, South Africa, Australia, Brazil, Honduras, and the British West Indies. In California, granulation is confii ed almost entirely to Valencias, but it may occur to a slight extent in navel oranges from young trees in some areas. In other states in this country and in other countries, it may occur in these varieties and in almost any of the other commercial varieties of citrus, such as grape- fruit, tangerines, and tangelos. In general, no two citrus-growing areas use the same term to designate this disorder. For example, in Florida it is called "dry end," which is a good term in that little juice can be reamed from the stem-end half of a badly granulated fruit, but, strictly speaking, is erroneous because the affected juice sacs actually contain more water than the adjacent healthy ones. In Siam it is called "Koa Sam" which means raw or uncooked rice. This term is applicable because in the advanced stages the affected juice sacs look somewhat like grains of rice. In Trinidad, British West Indies, it is referred to as "corkiness," but because of the physiological and histological condition of the affected sacs, this term does not appear so suitable. In California the term "crystallization" has been used for many years and was probably originally suggested by the appearance of the little 1 Received for publication January 6, 1940. - Paper no. 422, University of California Citrus Experiment Station, Eiverside, California. ! Professor of Plant Physiology and Plant Physiologist in the Experiment Station. 4 Assistant Professor of Plant Physiology and Assistant Plant Physiologist in the Experiment Station. 5 Junior Plant Physiologist in the Experiment Station. [3] Erratum in Bulletin 647: The graph labeled fig. 3 (p. 40) should be fig. 5 (p. 51) ? that labeled fig. 4 (p. 49) should be fig. 3 (p. 40), and that labeled fig. 5 (p. 51) should be fig. 4 (p. 49). The legends of the graphs are correctly numbered. 4 University of California — Experiment Station gas pockets which form in the juice sacs in the intermediate stages of development and which, to the unaided eye, may resemble crystals: there are no crystals in or on the affected juice sacs. "Granulation" ap- pears to be a more descriptive term — one now widely used in California (1, 5) ." "Sclerocystosis" 7 has been suggested as a technical term (2) . Unless otherwise stated, the Valencia variety of orange, Citrus sinen- sis Osbeck, is referred to in this paper in all cases where trees or fruits are mentioned. DESCRIPTION OF GRANULATION AND DRY JUICE SAC Granulation is not caused by a fungus, virus, or bacterium, but results from a series of abnormal physiological activities. As a rule, granulation does not appear in the fruit in California to an appreciable extent until late in July, after the middle of the Valencia picking season. In 1932 a special study was made to determine how early in the development of the fruit granulation may be recognized. In a few cases it was found to be plainly evident in large fruits by the time they had reached the eight-to-one standard maturity test. 8 Observations at earlier dates showed that the contents of some of the juice sacs in the stem end of large fruits from young trees had become gelatinized by the latter part of February, a month or more before the fruits normally matured in the locality from which they were obtained. Further obser- vation and study have shown that gelatinization is a process through which all juice sacs must pass before reaching the condition known to the grower as typical granulation. Some of the characteristics of granulation have already been described by Bartholomew, Sinclair, and Raby {2, 3). It is almost impossible to detect the disorder until the fruit has been cut. In the intermediate stages, however, a badly granulated fruit may feel harder, and in the last stage, softer, at the stem end than a healthy fruit. In the latter part of the season, the peel of a granulated fruit may turn a little greener, especially around the stem end, than a healthy fruit. The preceding statements refer to fruit from the regular bloom. Each year a small percentage of off-bloom fruit is also produced. This fruit and the heavily shaded fruit from the regular bloom are usually not only off-color, but more subject to granulation than other fruit from the regu- lar bloom. Because of the color differences, granulation is more easily detected, without cutting, in these fruits than in other fruits. u Italic numbers in parentheses refer to "Literature Cited" at the end of this paper. 7 "Sclero"=hard or tough, "cyst' -sac, and "osis"=state or condition. 8 The ratio of soluble solids to acids (8:1) required in the juice of oranges mar- keted in California. Bul. 647] Granulation of Valencia Oranges 5 Typical granulation begins in the stem end of the pulp and gradually progresses toward the stylar end. 9 A fruit is considered to be badly gran- ulated when from one third to one half of the stem-end pulp has become affected. In some seasons and in some fruits, the extreme stem end of the pulp appears to be healthy, but small masses of slightly to moderately granulated juice sacs may be found near the "core," from one quarter to one third of the way in from either the stem or the stylar end of the fruit, or from both ends. Plate 1 (A and B) shows granulation in the stem end of the pulp of fruits that have been cut longitudinally and transversely. The peel in B of this figure appears to be excessively thick, but this is because the cut was made very near the stem end of the fruit. Plate 1, I, shows the cut- surface of a longitudinal section of a healthy fruit. Plate 1, at C, G, and H, shows a type of breakdown that is often re- ferred to as "granulation." It may accompany granulation (plate 1, H) or appear independently (plate 1, C, 0) . The juice sacs do not at any time become enlarged or hardened as in granulation, but lose their water con- tent and shrink until they become flattened or almost needle-shaped, their form depending upon the direction of lateral pressures. Isolated juice sacs of this kind are shown in plate 1, E, between healthy (plate 1, D) and granulated juice sacs (plate 1, F) from a fruit that was partially granulated. The condition first becomes evident in the longitudinal cen- ter of the pulp segment (plate 1, G) and in this again is different from granulation. Another characteristic of this disorder in which it is unlike granulation is its prevalence in intermediate- and small-sized fruits. It is often called "blossom-end granulation" because it is usually more pronounced in the stylar end (and in the middle of the fruit) than in the stem end. Because of the specific characteristics of this condition of the pulp in Valencia and other citrus fruits, it seems desirable to distin- guish it from granulation. "Dry juice sac," or simply "dry sac," is sug- gested at this time as being a good, descriptive, nontechnical term, and "xerocystosis," 10 as a technical term. During storage or transit, the ap- parent spread of dry juice sac in the fruit is greater than that of granu- lation. Like granulation, it is a physiological disorder. FIELD EXPERIMENTS In the course of field studies on granulation of Valencia oranges in California, from 1930 to 1938, inclusive, 125,600 fruits were cut and 9 The stylar end of the citrus fruit is the end farthest from the point of attachment, often erroneously called the "blossom" end. The petals are attached at the base of the young fruit ; therefore the blossom end and the stem end refer to the same portion of the fruit. 10 "Xero"=dry, "cyst"=sac, and "osis"=state or condition. 6 University of California — Experiment Station observed, and their condition was recorded. All of these fruits were cut and observed by the same person, so to this extent the personal element as a factor in judging the condition of the cut fruits was eliminated. The fruits came from approximately 1,500 trees in 65 different groves in Riverside, Orange, and Los Angeles counties. In addition, many thou- sand fruits from these and other groves have been used in making gen- eral observations and in biochemical, microehemical, and histological studies. PRELIMINARY TESTS, 1930-1934 The results of some of the field experiments for 1930 to 1934, inclusive, have been previously described {2, 3), and at this time they will be given in summary form only. The last of the 1930 Valencia crop at the Citrus Experiment Station, with the exception of the fruit of 10 thirteen-year-old trees, was picked during the last week in July. Many fruits from this pick were cut, but no granulation could be found. The fruits of the 10 trees that were left were picked and cut on different dates, as follows : on September 16 (3 trees), on October 10 (3 trees), and on November 13 (4 trees). On each of these dates the fruits from five different portions of each tree were segregated for cutting and observation. The average percentages of granulated fruits for each portion of all trees were : inside bottom, 71 ; outside north, 65 ; inside top, 56 ; outside east and west, 54 ; and outside south, 52. These results indicate that granulation is more prevalent in the fruit from some portions of the tree than from others. Other data obtained in this experiment indicate that fruit which is good in the earlier part of the season may become granulated if it is allowed to remain too long on the trees. Trees in Cheesecloth Tents (1931). — The possibility that fogs or clouds, which decrease the amount of available sunshine, might be a factor in causing granulation suggested the following tests. Five four- teen-year-old Valencia trees at the Citrus Experiment Station were en- closed within a single large tent made of thin cheesecloth. The direct sunshine did not strike the foliage, but the tent was well ventilated by means of louver openings. In a near-by grove of six-year-old Valencias, each of 6 trees was enclosed singly in a tent of the same material. Three of the tents were entirely closed, and there was no ventilation except through the cheesecloth. The other 3 tents were especially well venti- lated, for there were no north or south sides, and the tops extended only far enough to prevent the direct rays of the sun from striking the foliage. All tents were installed in May and removed in October. In each case, the trees adjacent to the tents served as controls. Test cuts of the fruits from these trees were made on September 10 and Bul. 647] Granulation of Valencia Oranges 7 October 20 and included a total of 8,993 fruits. Every mature fruit on each tree was cut and observed. The trees were found to have produced average amounts of granulation as follows : 5 trees in large tent and 5 adjacent controls, 17 and 35 per cent, respectively ; 3 trees in small closed tents, 3 in small open tents, and 6 adjacent controls, 57, 57, and 58 per cent, respectively. The shading of the trees within the large tent evidently did not cause an increase in the amount of granulation. In fact, the re- sults indicate that the shading may have caused a reduction in granula- tion. This may be explained, however, on the basis that these trees were naturally low producers of granulation (see "Tree Differences and Con- stancies in the Production of Granulation," p. 12). The tents on the small trees apparently had no effect on the amount of granulation. The results show definitely that, under the conditions of this test, the reduc- tion in the amount of sunshine available for the trees did not cause an increase in the percentage or severity of granulation in the fruit. Excessive Growth of Fruits at Stem End and Granulation. — Early in the study of granulation, the opinion was expressed by some growers that a secondary growth occurred in the stem end of at least some of the Valencia fruits during the summer and that this secondary growth was caused by, or was the cause of, granulation. Information concerning this possibility was sought in 1931 by volumetrically measuring the com- parative increase in size that occurred in the stem and stylar halves of fruits during the period from early summer to early fall. The equator of each fruit was marked by an inked thread fastened to a specially de- vised spring wire. The volume of the stem half and of the stylar half of each fruit was then determined by the amount of water displaced by each half in another specially devised piece of apparatus. Similar volu- metric measurements were made on the same fruits at the time they were to be cut and examined for granulation. Ten of the largest fruits on each of 10 thirteen-year-old trees and on each of 5 six-year-old trees were measured. The results of these tests were not entirely conclusive. The fruits from the thirteen-year-old trees showed a slightly greater increase in the stem halves of the good fruits than in those of the granulated fruits, but the difference is probably not significant. The average volume increase of the stem half was 3 cc greater than that of the stylar half for the good fruits and only 2 cc greater for the granulated. With the six-year-old trees, the differences were greater : the average volume increase of the stem half was 10 cc greater than that of the stylar half for the good fruits and 14 cc for the granulated — a reversal of the results shown by the good and the granulated fruits from the thirteen-year-old trees. The results of these tests at least indicate that there is no definite rela- 8 University of California — Experiment Station tion between granulation and enlargement of the fruit at the stem end. No numerical data have been obtained, but continued observation over the period from 1931 to 1938 has led to the conclusion that the more nearly spherical fruits are no less likely to be granulated than those which are more elongated. As a rule, in the coastal areas where granu- lation is more abundant, the fruits are more spherical than they are in the interior districts where granulation is less abundant. The data were next examined to determine whether there was any more granulation in the fruits that had made the greatest increase in size, as a whole, than in the other fruits. The average volume increase of all the good fruits from the older trees was 7.6 per cent, while that of the granulated fruits was 8.5 per cent. When only 10 of the largest fruits in each of these two groups were considered, the results were just the re- verse : the 10 good fruits showed an increase of 14.9 per cent, while the 10 granulated fruits had increased only 12.5 per cent. The fruits from the young trees did not show this reversal. The average volume increase of all the good fruits from the young trees was 14.9 per cent, while that of the granulated fruits was 19.6 per cent. When only 10 of the largest fruits in each of these two groups were considered, it was found that the 10 good fruits had increased 17.7 per cent, while the 10 granulated fruits had increased 28.3 per cent. The lack of consistent results in these tests on the relation between size increase of fruits and granulation is an example of the inconsisten- cies encountered all through the course of study of granulation. The Effect of Sprays (1931-1934).— -In the summer of 1931, 3 thir- teen-year-old trees at the Citrus Experiment Station were sprayed with lime. In the fall, the fruits of these trees were cut and observed, and very little granulation was found. The lime-spray test was repeated in 1932 on a much larger scale in a fifteen-year-old grove at the Citrus Experiment Station and in a ten- year-old grove near Santa Ana. The latter grove was chosen because in previous years it had produced a great deal of granulation. Thirty trees in each grove were sprayed with a lime whitewash composed of 40 pounds of slaked lime and 1 pound of Kayso spreader in each 100 gallons of water. Twenty of the trees in each grove were sprayed in March. In May, 10 of these trees were resprayed, and 10 more were given a first appli- cation. Three test cuttings of the fruit from these trees and 10 control trees were made during the late summer and early fall. In each test, 50 fruits were chosen at random from both the north and the south half of each tree. This called for the cutting of 24,000 fruits. From two to four transverse cuts were made on each fruit to determine the absence or presence, and the severity, of granulation. Since no significant differ- Bul. 647] Granulation of Valencia Oranges ences resulted from the different methods of spraying, the results of all of them are combined for comparison with the unsprayed trees. The consistency of the results of the random sampling is indicated in figure 1. The granulation percentages for 100 fruits from each of 10 trees were determined, and then at once a second, similar determination was made on other fruits from the same group of 10 trees. The figure shows that there was very close agreement in the two tests. 100 80 D < 60 K O K 5 40 u 20 o a Vi -I i 1 'I 1 ! ,i ■I \A 15 17 16 19 TREE NUMBER 20 21 22 23 Fig. 1. — Comparative results of two successive granulation tests made on the same day. In each test, 100 fruits were picked at random from each of the same 10 trees (nos. 14-23), and percentages of granulated fruits were determined. (The vertically hatched columns represent the results of the first pick.) In order to determine the comparative numbers of fruits of different sizes and the degree of granulation, a field sizing and grading board was devised and constructed (plate 2, A) . To prevent unnecessary complica- tion in compiling the results, the fruits were graded into only five sizes : 100's, 150's, 200's, 288's, and 344's. Each of these was arbitrarily chosen to represent a different group within the range of sizes (80's to 360's) used in the standard pack of the California Fruit Growers' Exchange. The sizing and grading board was so constructed that as a fruit was picked and sized, then cut and inspected, the moving of an indicator on the proper dial recorded its size and also its quality, such as "good," or, with reference to granulation, "slight," "moderate," or "bad." Unfortunately for the experiment, the fruit from the trees in the grove at the Citrus Experiment Station did not have enough granulation in it to warrant making the complete test. The cutting of 3,100 fruits from these trees, at intervals during the latter part of the picking sea- son, showed a maximum of only 3 per cent granulation on any one tree, and in these fruits the granulation was very slight. 10 University of California — Experiment Station Granulation in the grove of ten-year-old trees near Santa Ana was bad and afforded an excellent opportunity for determining the effects of the lime spray. The results of this test are shown in summary form in table 1. These data show that 67 per cent of the fruit from the unsprayed trees and only 34 per cent of that from the sprayed trees was granulated. The data also show that fruit from the sprayed trees was much less se- verely affected than that from unsprayed trees. (Even fruits containing only two or three affected juice sacs were classed as granulated.) TABLE 1 Granulation in Fruits from Unsprayed and Lime-sprayed Trees, Santa Ana, 1932* Treatment Unsprayed Lime-sprayed. Proportion of fruits granulated in various sizes 100's per cent 95 80 150's per cent 84 70 200's per cent 62 27 288's per cent 29 8 344's per cent 10 1 All sizes per cent 67 34 Severity of granulation, all sizes Slight per cent 18 19 Moder- ate Bad per cent 34 7 per cent 15 * Samples consisted of 50 fruits, chosen at random, from both the north and the south half of each tree — 10 trees unsprayed and 30 sprayed. Fruits were segregated into the five sizes in a field sizing board and were cut at once for observation. An attempt was made in 1933 to determine whether the reduction in the amount of granulation in the fruit from the lime-sprayed trees was caused by a chemical or by a physical effect of the lime. The tests were made in the same 2 groves in which the 1932 tests had been made. Ten trees in one row in each grove were sprayed with an aqueous mixture of barium sulfate and cornstarch (75 pounds barium sulfate, 38 pounds uncooked cornstarch, and 1.9 pounds of agar, in 150 gallons of water). Ten other trees in a second row in each grove were sprayed with a mix- ture composed of 6.3 pounds of calcium nitrate and about 1 pound of blood albumin in 150 gallons of water. The amount of calcium nitrate used was calculated to give about 10 ounces to each tree. In a third row in each grove the 10 trees were not sprayed, but 115 to 120 of the largest fruits on each tree were dipped in a lime whitewash of the same concen- tration as had been used as a spray in 1932. In a fourth row in the grove at the Citrus Experiment Station, a similar number of fruits were en- closed in waterproof bags, and the trees were then sprayed with the lime whitewash. The bags were removed as soon as the foliage had dried. In each grove, the fruit on an adjacent block of unsprayed trees served as controls for the spraying tests. Fruits borne adjacently, on the same trees, served as controls for the dipped and bagged fruits. The spray and other applications were made during the first week in May. Bul. 647] Granulation of Valencia Oranges 11 Again, as in 1932, preliminary tests made at monthly intervals showed that there was not enough granulation in the Citrus Experiment Station grove to warrant the completion of the test. While this was dis- couraging so far as the test was concerned, it substantiated the observa- tion that granulation is more prevalent in coastal than in interior districts. In the grove near Santa Ana, two granulation cutting tests were made, one in early July and the other 6 weeks later. In the two tests, 11,000 fruits were cut and examined. The total percentages of granulated fruits from the trees which received the different spray treatments were : sprayed with barium sulfate and cornstarch, 69 ; with calcium nitrate, 72 ; unsprayed (controls) , 87. Neither the barium-starch nor the calcium nitrate sprays had any very great controlling effect on granulation, either as to amount or severity. It should be stated that the barium- starch mixture did not make the trees nearly so white as the lime spray used in 1932. The fruits dipped in limewater showed 63 per cent, and the controls from the same trees, only 59 per cent, granulation. It would appear from these results that the dipping in lime had caused a slight increase in the amount of granulation. Such was probably not the case, because the records show that 50 per cent of the dipped fruits and only 28 per cent of the controls were size 150 or larger. The large fruits are more likely to granulate than the small ones. As already stated, 115 to 120 of the largest fruits on each tree were dipped, so that the fruits which were cut from the same trees for controls were necessarily smaller in size. The trees bearing the dipped fruits and their controls were in a portion of the grove which normally produced a comparatively low percentage of granulation. This probably explains why a lower percentage of these fruits than of those in the spray plots were granulated. In 1934 the lime-spray tests were repeated in the grove near Santa Ana. Plots of 20 trees each were sprayed with 10, 15, and 20 pounds of lime and 1 pound of Kayso spreader to each 100 gallons of water. Twenty adjacent trees — one row of 10 trees on the north side and another on the south side of the plot — served as controls. The spray was applied on April 11. Because of the lack of granulation in the two previous years, the test was not made in the grove at the Citrus Experiment Station. As in previous years, 50 fruits were picked at random from both the north and the south half of each tree for cutting — a total of 8,000 fruits. This year only one fruit-cutting test was made. The results were not conclusive. Granulated fruits on the trees to which the 10, 15, and 20 pounds of lime were applied averaged 62, 51, and 40 per cent, respec- tively, but the control trees produced only 48 per cent. Probably the 12 University op California — Experiment Station factor of chance operated here as, apparently, in the large tent experi- ment in 1931. While the number of fruits affected was about the same for this year (1934), the degree of severity of granulation in the fruits was less than it had been in the two previous years, not only in this grove, but in all Valencia-growing districts in California. TREE DIFFERENCES AND CONSTANCIES IN THE PRODUCTION OF GRANULATION In the course of these experimental tests, from 1930 to 1938, it was found that some trees produce a larger number of granulated fruits and fruits more severely granulated than adjacent trees in the same grove. The granulation yields for the years 1933 to 1937, on the 4 trees on which this difference was first especially noticed, are recorded in table 2. One TABLE 2 Tree Differences and Constancies in the Production of Granulation, 1933-1937* Proportion of fruits granulated Tree no. 1933 1934 1935 1936 1937 5-year average per cent per cent per cent per cent per cent per cent 15 89 63 75 41 43 62 16 96 78 71 60 59 73 22 32 40 9 16 2 20 25 39 35 6 5 7 18 * Trees located near Santa Ana and not treated. Fruits were selected as for data in table 1, except in 1937, when all fruits on one large limb on each tree were cut (trees had been top-worked and all other limbs removed). hundred fruits per tree were used in making these determinations, ex- cept in 1937, as explained in the following paragraph. The first 2 trees listed in the table show a much higher percentage of fruits affected than the other 2 trees, and records not given in the table show that the severity of granulation in the fruits from these trees was much greater than in fruits from the other 2 trees. In 1935, for trees 15 and 16, and in 1933 and 1937, for trees 22 and 25, the percentages were reversed from their relative positions in other years. It should be stated that the percentages for 1937 were obtained from fruits from only one large limb on each tree. All the other large limbs had been removed from the trees when they were top-worked in the preceding summer (see p. 17) . On the single limbs of trees 15, 16, 22, and 25, there were 73, 96, 135, and 298 fruits, respectively ; all fruits were picked and cut. In order to get a more complete record of possible differences in Bin,. 647 Granulation op Valencia Oranges 13 amounts of granulation produced by different trees in the same grove, fruit tests were made in 27 groves in Orange and Los Angeles counties in 1936. Five trees near each corner of each grove (20 trees in each grove) were selected and tagged. The fruit was allowed to remain on these trees until late in the picking season to give the fruit a good chance to granulate. The trees in the different groves ranged from eight to thirty years of age. When the test cuttings were made, 25 of the largest fruits TABLE 3 Grove Averages and Tree Differences in Granulation Production, 1936* Age of trees Fruits granulated Grove no. Age of trees Fruits granulated Grove no. Grove average Range in individual trees Grove average Range in individual trees 1 years 8 20 25 9 9 20 30 16 25 12 12 12 10 25 per cent 91 35 45 51 38 29 14 37 23 60 44 14 31 18 per cent 76- 96 4- 60 12- 76 16- 72 4- 56 4- 76 0- 28 0- 84 0- 78 12-100 12- 96 0- 48 8- 68 0- 48 15 16 years 22 12 22 19 19 19 19 12 22 ? 15 18 18 per cent 43 35 23 25 25 16 24 5 20 34 27 5 17 per cent 0-100 2 12- 64 3 17 18 19 0- 84 4 0- 56 5 0- 84 6 20 21 0- 56 7 0- 72 8 22 23 24 25 26 27 0- 52 9 0- 48 10 0- 84 11 0- 68 12 0- 28 13 0- 40 14 * Tests included 20 trees in each of 27 groves located in Orange and Los Aneeles counties, a total of 540 trees. Twenty-five of the largest fruits from each tree, a total of 13,500 fruits, were tested. were chosen from each tree, making a total of 13,500 fruits. The largest fruits were chosen because they were the ones that were most likely to be granulated. The results of these 1936 tests are presented in table 3, which shows the minimum and maximum percentages of granulation on trees within a given grove and the average percentage for each grove. Similar fruit cuttings were made in 14 of the 27 groves over a period of three successive years (1936-1938) to obtain further data on tree constancy in the production of granulation. These tests entailed the cutting of 15,000 additional fruits during 1937 and 1938. The data on only 4 of these groves are given (table 4) , but they are typical of results obtained in all of them. About half of the trees were constant over the entire three-year period in the production of either large, moderate, or small numbers of granulated fruits, while many of the trees were con- stant in only two of the three years. 14 University of California — Experiment Station TABLE 4 Tree Tendencies toward Constancy in the Production of Granulation, 1936-1938 Row and tree no. Number of fruits granulated (of 25 fruits cut) Row and tree no. Number of fruits granulated (of 25 fruits cut) 1936 1937 1938 1936 1937 1938 Grove no. 1 Grove no. 5 Row 4: Tree 4 23 22 24 23 25 23 23 20 24 23 23 19 22 22 24 24 24 22 20 23 22 25 20 17 24 25 22 23 23 23 20 13 13 11 19 23 19 24 17 10 24 25 25 20 25 25 25 23 24 24 23 19 22 25 24 24 23 22 15 18 Row 3: Tree 1 14 11 15 7 7 5 6 12 11 7 9 2 1 10 5 24 9 10 13 14 16 11 19 20 17 10 18 12 14 22 9 6 9 6 13 24 18 12 10 15 19 Tree 5 Tree 6 Tree 2 Tree 3 4 15 Tree 7 Tree 8 Row 7: Tree 5 Tree 4 Tree 5 Row 5: Tree 6 14 15 5 Tree 6 Tree 7 7 Tree 7 Tree 8 11 Tree 8 Tree 9 10 Tree 9 Tree 10 8 Row 15: Tree 6 Row 11: Tree 6 10 Tree 7 Tree 8 Tree 9 Tree 7 Tree 8 Tree 9 7 15 4 Tree 10 Tree 10 9 Row 19: Tree 3 Tree 4 Row 14: Treel Tree 2 22 10 Tree 5 Tree 3 11 Tree 6 Tree 7 Tree 4 Tree 5 7 13 Grove no. 3 Grove no. 9 Row 5: Tree 3 11 16 7 11 7 3 8 15 18 19 15 8 12 9 16 * 13 8 12 6 10 3 12 14 25 5 9 15 19 10 17 3 12 17 24 4 5 5 13 8 13 22 14 7 18 10 9 4 14 18 18 5 9 17 7 4 19 Row 3: Tree 34 9 3 10 12 2 3 6 3 7 4 19 5 4 1 6 2 10 8 8 2 3 1 3 13 6 1 7 2 13 2 2 2 5 19 Tree 4 Tree 35 5 Tree 5 Tree 36 5 Tree 6 Tree 37 8 Tree 7 Tree 38 16 Row 4: Tree 14 Row 4: Tree 51 . . . 6 Tree 15 Tree 52... 24 Tree 16 . Tree 53 6 Tree 17 Tree 54 3 Tree 18 Tree 55 4 Row 11: Tree 18 Row 7: Tree 49 3 Tree 19 Tree 51 13 Tree 20 Tree 52 4 Tree 21 Tree 53 2 Tree 22 Tree 54 . 8 Row 10: Tree 4 Row 8: Tree 35 15 Tree 5 Tree 36 19 Tree 7 Tree 37 10 Tree 8 Tree 38 8 Tree 9 Tree 39 5 Data not available; these trees had been picked by mistake by the pickers. Bul. 647 Granulation of Valencia Oranges 15 The fact that certain trees in a given grove produce more granulation than the adjacent trees in the same grove is shown in histogram form in figure 2. The figure also shows that some trees are comparatively con- 80 60 40 20 80 60 40 I- 20 < -J 1 80 < £ 60 H 40 | 20 h 5 so o DC u h _l _J 2 UJ 2 ^ .06 O > .04 --"■-" , , . r , ! GRANULATION 2 STAGES Fig. 4. — Comparative amounts of dry matter in juice sacs in stages (healthy) and 1, 2, and 3 (granulated). Determinations were based on average dry weight per individual juice sac. juice sacs — that is, on a percentage basis — the dry matter was 1.83 per cent higher in the healthy than in the granulated juice sacs. When the dry weight was based on amount of increase per individual juice sac (fig. 4), the differences were reversed and were much more marked, which indicates that there had been an actual growth of the juice sacs during the early stage of granulation. The increase in dry weight in stage 1 over stage was 206 per cent. When equal weights of these particular healthy and granulated juice sacs (stages 1 and 2) were compared, rather than individual sacs, it was found that the excess of dry matter in the granulated sacs over the dry matter in the healthy sacs was similar to that shown in table 19, but a little greater. Small differences were to be expected because the samples of fruit came from different localities. The segregation of the granulated juice sacs into stages 1, 2, and 3 50 University of California — Experiment Station was, of course, arbitrary; therefore the differences in dry weight be- tween stages 1 and 2, and between stages 3 and as shown in the graph (fig. 4) may not be significant. The main points of interest are the great increase in dry matter in stages 1 and 2 and the sharp decrease in stage 3. In a succeeding section (see "Juice-Sac Collapse," p. 52) it will be shown that the granulated sac begins to disintegrate during stage 2, as is shown by the accumulation of gas in certain internal cells. This disin- tegration may account for the loss of dry matter by the time stage 3 has been reached. The amount of ash, on a basis of both fresh and dry weight, was al- ways lower in the healthy than in the granulated juice sacs (table 19). The differences were fairly consistent, ranging, on a fresh-weight basis, from 0.123 to 0.191 per cent, and on a dry-weight basis from 2.68 to 3.77 per cent. On an average, there was 0.16 per cent more ash on a fresh- weight basis and 3.18 per cent on a dry-weight basis in the granulated than in the healthy juice sacs. A study of table 19 will also show that the principal inorganic con- stituents were highest in the ash from the granulated juice sacs. This was true for calcium, magnesium, potassium, and sodium, and in most cases for phosphate. Similar tendencies have been reported by Haas and Klotz (6) in connection with their work on gradients in citrus fruits. The differences for calcium and magnesium are especially large, there being an average of only 46.88 per cent as much calcium and 59.09 per cent as much magnesium in the healthy juice sacs as in the granulated ones. The sodium values for the season of 1935 (samples 1 to 4) were higher than those for 1936 (samples 5 to 10). The fruit from which the samples were taken came from three different localities, and it is prob- able that the trees from which the 1935 samples were taken had access to comparatively large amounts of sodium. The form in which these several elements existed in the two kinds of juice sacs tested will not be discussed here. The significant and interesting thing about them is that they accumulated in excess in the granulated sacs. STRUCTURAL AND OTHER CHANGES IN JUICE SACS DURING GRANULATION Enlargement of Juice Sacs. — That the juice sacs enlarge during the process of granulation is indicated by the fact that a volume equal to 16.4 cubic centimeters (1 cubic inch) required approximately 475 healthy sacs, whereas the same volume required only approximately 213 sacs in stage 1, 146 sacs in stage 2, and 295 sacs in stage 3 of granulation. The comparative sizes of healthy sacs and of those in the three stages of granulation are shown graphically in figure 3 (p. 40). As mentioned be- Bul. 647 J Granulation of Valencia Oranges 51 fore, some of those in the collapsed condition (plate 2, B, stage 3) appear to be larger than they really are because of the position they were in when photographed. They may be broad but are comparatively thin (plate 4,^). Hardening of Juice Sacs and Thickening of Cell Walls. — Evidence that juice sacs harden as well as enlarge as they become granulated is shown by figure 5. In making these determinations, individual juice sacs < 5 .0 3.0 GRANULATION- STAGES Fig. 5. — Comparative pressures (in ounces) necessary to crush healthy (stage 0) juice sacs and those in stages 1, 2, and 3 of granulation to 1 mm in thickness. were tested, and pressure was applied gradually until the sac had been crushed to a thickness of 1 mm. A pressure of approximately 140 grams (5 ounces) was found to be sufficient to crush the average healthy juice sac, while to crush a sac that was in the early or medium stage of granu- lation took from 500 to 3,000 grams (20 to 100 ounces) or more. In stage 1 of granulation, the hardness appears to be due principally to the fact that the juice in the cells of the sac becomes more or less gelatinized. In stage 2 this condition persists in some of the cells, but what is more sig- nificant is that the walls of some of the internal cells may become as much as twenty-five times as thick as they were before the sac began to granu- late. Photomicrographs of cross sections of healthy and granulated juice sacs are shown in plates 2, C, and 3, A, B. Plate 2, C, shows that each healthy juice sac is composed of numberless small, thin-walled cells and is not merely a single cell or a little bag filled with juice, as it is so often, 52 University op California — Experiment Station erroneously, thought to be. Each of the small cells that make up the in- terior of healthy juice sacs remains intact and alive, even after picking, so long as the fruit is not destroyed. The walls of the cells which compose the interior of the healthy juice sac are very thin, and for that reason it is impossible to tease out or iso- late a single cell. When the walls of some of these cells have become thick- ened and hardened, as illustrated in plate 3, B, it is not difficult to isolate the cells with dissecting needles under low magnification. Plate 4, A, is a photomicrograph of two thick-walled cells isolated from the in- terior of a granulated juice sac and many times larger than any of the cells in healthy juice sacs. The large dark rings shown within these cells are internal gas bubbles. The black dots on the surface indicate the be- ginning stages of cell-wall etching, which finally results in the escape of the gas and the disintegration of the cell. Plate 4, B, is a photomicrograph of a cross section through the walls of three adjoining cells near the center of a granulated juice sac. The walls had become very much thickened, but etching had not yet begun. The dark line between the walls of the adjoining cells represents the material which holds the walls together. It contains a comparatively large amount of calcium pectate. Plate 4, C, is a photomicrograph of a longitudinal section through an enlarged, granulated cell whose contents have disappeared and whose wall has become very much etched. Juice-Sac Collapse. — The final stage of granulation (stage 3) is typi- fied by a softening and collapse of the juice sac, caused by the progres- sive disintegration of the cells, which begins near the center of the juice sac. Plate 4, D, shows the beginning of the formation of the central cav- ity. The collapse begins at about this stage and continues, as more of the cells disintegrate, until the sac becomes flattened or more or less angu- lar, its form depending upon the directions of lateral pressure. Plate 4, E, is a photomicrograph of a cross section through a collapsed granu- lated juice sac. Under certain conditions it is not easy to distinguish between frozen and granulated collapsed juice sacs. The collapsed granulated sacs, how- ever, usually have less color than similar frozen sacs, and microscopic examination shows that the granulated sacs have an uneven or rough- ened appearance on the surface, which the frozen sacs do not have. The uneven appearance is caused by the presence of some of the thick-walled, not-yet-disintegrated cells just inside the juice-sac wall. Lignification of Cell Wall. — As has been stated, the early stages of granulation are characterized by the gelatinization of the contents of at least a portion of the internal cells of the juice sacs. Soon after the cell Bul. 647 Granulation of Valencia Oranges 53 contents gelatinize, or during the later stages of gelatinization, changes occur in the walls of these cells. Microchemical and histological studies have shown that these changes culminate in the formation of lignin or ligninlike substances. The walls of such cells (plate 3, B) become hard and woody, more or less like the wood cells in the trunk and limbs of the tree. As granulation progresses, these hard- walled cells and adjacent, thinner- walled cells disintegrate ; an internal cavity is thereby formed in the juice sac (plate 4, D), and lateral pressures cause the sac to col- lapse (plate 4, E). Gas Bubbles. — Gas bubbles appear within the cells at the same time that the walls are thickening and hardening. Examples are shown in TABLE 20 Carotin Concentration in "Healthy" and Granulated Juice Sacs (Expressed in gammas per gram of fresh weight) Year "Healthy"* juice sacs Granulated juice sacs Stage 1 Stage 2 1936 gammas 13.9 18.3 16.1 gammas 10 10.0 gammas 5.5 1937 4.7 5 1 * For explanation of "healthy," see text, p. 41. t Gamma values of averages per mg. are: 0.0161, 0.0100. and 0.0051, respectively. plate 4, A. Cells such as these may be easily isolated from the surrounding smaller cells, and the gas bubbles within them can be seen with the un- aided eye. Lateral pressure on such a cell causes the gas bubble to shift position, but it cannot be forced out of the cell until the wall is ruptured. As the cell walls and the layers between them disintegrate, the gas passes out of the cells, forms gas pockets between them, and finally escapes into the large central cavity, the result of multiple cellular disintegration. Limited analytical tests on the composition of this gas showed, as would be expected, that it was composed largely of carbon dioxide. Decrease in Color. — One of the early visible characteristics of granu- lation is the decrease in color (carotin) in the affected juice sacs. Spec- troscopic and colorimetric determinations have shown that granulated juice sacs contain less and less carotin per gram of fresh weight as the disorder progresses. Table 20 shows the comparative amounts of carotin in "healthy" juice sacs and in those in the early and intermediate stages (stages 1 and 2) of granulation. "Healthy" and granulated sacs were taken from the same fruits. The results show that there was less than two thirds as much carotin per gram of fresh weight in the juice sacs in stage 54 University of California — Experiment Station 1 of granulation as in those that were "healthy," and only one half as much in those in stage 2 as in those in stage 1. Tests were not made on juice sacs in stage 3 of granulation (collapsed sacs), but their visible appearance would lead to the conclusion that they contain only a trace of carotin, if any. CAUSES OF GRANULATION The cause or causes of granulation cannot be definitely stated. The pro- duction of granulation appears to be very closely related to those condi- tions which are responsible for vigorous tree growth and the production of large fruit sizes. In the majority of cases, granulation is much more prevalent in the fruit from young trees than in that from old trees. It may be abundant, however, in fruits produced on old trees that have been heavily pruned. Apparently the fruits from such trees are large and granulated because they have been borne on branches having lux- uriant foliage like that on young trees. If this is true, then factors such as fertilizers, amount of irrigation water, rootstock, and climatic condi- tions may directly or indirectly influence the production of granulation. The fluctuation in the amount and severity of granulation from year to year indicates that climatic conditions may be an influencing factor. The direct relation between granulation and large fruit sizes may be obscured by other factors, however. For example, the largest fruits are usually borne on the south side of the tree, but granulation is most abun- dant on the north side ; one rootstock may cause the production of larger fruits but less granulation than another rootstock that produces slightly smaller fruits ; and in the second year the fruits on girdled limbs may be smaller but contain more granulation than larger fruits on nongir- dled limbs (table 10, p. 35). Tree vigor and fruit size appear to have a definite influence on the amount and severity of granulation, yet the results of the experimental work show that other factors are also important. For example, in some cases a given tree, be it young or old, may constantly produce more granulation than an adjacent tree planted at the same time. This would indicate that some condition inherent in the tree itself may be even more important than cultural practices or soil and climatic conditions. The bud-selection studies as outlined in a previous section of this paper (see "Experiments in Budding and Top-Working," p. 16) throw some light on this phase of the problem. It should be remembered, however, that not all the trees studied showed the constant tendency to produce large or small amounts of granulation from year to year. The fact that Valencias grown in the coastal areas usually have a lower content of total soluble solids than those grown in the interior Bul. 647] Granulation of Valencia Oranges 55 might be thought to be one reason why the former show more granula- tion than the latter (table 7, p. 25) . During the years in which tests were made to determine the possible effect of the rootstock on the produc- tion of granulation, healthy fruits from trees grown on the same root- stocks near the coast and in the interior were analyzed for total soluble solids, acid, and for reducing and total sugars. The results of these analyses show that concentrations of all of these substances were lower, in every case, in the fruits from the coastal area (Tustin) than in those from the interior area (Riverside). Large fruits were found to have a lower concentration of these substances than smaller fruits, and, as a rule, to be more susceptible to granulation. The same holds true for the fruits from the inside and north side of the tree in comparison with those from other parts of the tree. Rough-lemon rootstock was one of the highest producers of granulation in both areas, and in almost every instance over a four-year period, fruits from these trees showed the low- est concentration of total soluble solids, acids, and sugars. Trifoliate orange, however, also one of the highest producers of granulation in both areas, was one of the highest producers of total soluble solids, acids, and sugars. Again, the Brazilian sour orange, which produced the highest average amount of granulation at Tustin and a fairly low amount at Riverside, was also among those which gave high analytical results at both places. Therefore, the analytical results do not entirely substantiate the suggestion that fruits having comparatively low con- centrations of total soluble solids, acid, and sugars are most likely to granulate. The suggestion has been made that the unequal distribution of solu- ble solids, primarily the sugars, in the mature orange fruit may be a factor in the production of granulation. If this were an important factor in granulation, then all mature Valencia oranges, both large and small, should be granulated, because they all have a higher concentration of soluble solids in the stylar end than in the stem end. Likewise, the ma- ture navel orange and the grapefruit have a similar distribution of soluble solids, and yet, in California, these fruits are seldom if ever found to be granulated. Granulation does not usually become visibly evident in the fruit until after the fruit has matured, and the soluble-solids content of the stem end of the fruit, where typical granulation begins, is then noticeably less than that of the stylar end. There are times, however, when a small mass of granulated sacs may be found near the center of the fruit, about a third or a fourth of the way in from the stylar end, where the soluble- solids content is highest, and not in any other portion of the fruit. In such cases, the granulated area is small, and the soluble-solids content 56 University of California — Experiment Station of the stylar end of the fruit remains higher than that of the stem end. Again, the marked difference in the soluble-solids content of the stem and stylar ends does not become evident until the fruit is approaching maturity (1), yet exceptional cases have been found where granulation was visible in large fruits at least two months before they were mature. Such evidence challenges the theory that the differences in soluble solids in the two ends of the fruit are an important factor in the production of granulation. The possible relation of oil sprays to granulation was not experimen- tally determined in the series of tests reported in this paper, because groves were not available. The necessity for applying oil sprays to the groves year after year in the control of red mite made it impracticable to ask growers to offer their groves for such a study. However, in 1928 Higby et al. 2i applied twelve different brands or viscosities of oil sprays, at different times of the year, to twelve sets of 2 trees each in a grove in Orange County. With reference to granulation, the conclusions from these tests were : (1) that the time of application was not an important factor ; (2) that the trees sprayed with heavy and medium oils produced much more granulation than those that were fumigated or than those sprayed with light oils ; and (3) that the effects of oil sprays on the pro- duction of granulation in Orange County were much more marked than was generally reported for the interior districts. While the spray oils that are used today are much lighter than those used in 1928, many growers in the coastal area still feel that oil sprays are responsible for the production of at least a certain amount of granulation. It should be pointed out, however, that in the interior districts the oil sprays, as used today, appear to have little effect on the production of granulation. Re- cently experimental plots have been laid out in two widely separated groves, and the comparative effects of oil sprays and fumigation are being determined, but the data are not ready for publication. The comparatively large number of influencing factors and the ex- ceptions to the experimentally proved trends make it difficult to formu- late any theory concerning the cause of typical granulation that will fit all cases.* 1 L ' 4 Higby, Ralph H., et al. A study of the effect of various oil sprays on the develop- ment and composition of Valencia oranges. Unpublished data, 2 (File 17) : 10-99, California Fruit Growers' Exchange Research Department, Ontario, California. 1928. 25 After this paper went to press, an article by R. J. Benton appeared. Excerpts from this article follow. In California, Dr. E. T. Bartholomew has been working on this problem for a number of years. He has not yet finished his studies, but as far as they have gone his work coincides with innumerable observations made in New South Wales. Dr. Bartholomew states that all rootstocks are prone to produce granulated fruit, but that it appears that less granulation occurs on trees irregularly watered than on trees generously irrigated. That is our opinion, as our view is that granulation results from the vegetative activity of the tree. This condition can be almost regularly seen in our coastal districts on a large scale. . . . While all rootstocks will produce this effect f granulation], it does appear that the most vigorous stocks Bul. 647] Granulation of Valencia Oranges 57 CAUSES OF DRY JUICE SAC Up to the present time, dry juice sac has always been classed by the citrus industry as a phase of granulation. But the condition of the juice sacs and the prevalence of the disorder in fruits of all sizes indicate that it has a different origin than granulation. No experimental work has been done to determine the cause of dry juice sac ; therefore any suggestions in this line have to be based on observation and analogy. In a previous article (4) the suggestion was made that dry juice sac (then classed as granulation) might be caused by temperatures that were not low enough to freeze the fruit and to effect the collapse of one or more segments, yet low enough to cause a physiological disturbance that would show in the fruit when it became overmature. This sugges- tion still appears to have merit, but it may be questioned, in view of the fact that dry juice sac was very abundant in the fruit during the latter part of the 1938 picking season, after the comparatively warm winter of 1937-38. That the fruits have lost much of their vitality and that the affected juice sacs are in a state of senility may be a more plausible sug- gestion. Past experience has shown that if the fruit is picked by the middle of July or by the first of August, very little trouble is experienced from this disorder. The affected juice sacs have largely lost their power to retain moisture, and as a result they shrivel, especially after the fruits have been removed from the tree. CONTROL MEASURES In the Field. — At the present time the only practicable way to avoid granulation is to pick the large fruits before they begin to granulate. Extra picks of this sort may have to be made from one to three or four times in a single season. The large fruits on the north and inside of the trees should receive special attention. Young groves and those growing on the lighter types of soil should be most closely watched, although many exceptions may be found. Careful observations should be made on individual trees and on certain areas within a given grove. Lime spray has been found to reduce greatly the prevalence and severity of granulation in the fruits. But the amount of lime necessary to make the reduction commercially profitable is so large fat least 35 to 40 pounds per 100 gallons of water) that it apparently reduces the vitality influence the earliest granulation. I believe the stock influence to be in the following order — rough lemon, Cleopatra mandarin and sweet orange, with trifoliata last. I feel confident that granulation is not a strain effect, for we have in several orchards valencias propagated from many sources. . . . All show granulation to about the same degree at the time of any examination late in the season. . . . Heavy fertilising alone will not cause granulation, but in conjunction with ample irriga- tions it will encourage the development of this condition in late season oranges, if the trees ai-e in a state to utilise such generous treatment. (Benton, R. J. Granulation. Citrus News 16(9) : 129-30. 1940.) 58 University of California — Experiment Station of the tree. This conclusion is based on the fact that when the trees were subjected to the dry fall and winter winds, those which had been sprayed with lime in the spring lost many more leaves than those which had not been sprayed. It is possible that such a control measure might be used in localities where the dry winds are not such an important factor. Even in sheltered areas, however, the successive applications of lime spray in these amounts over a period of years might prove to be undesirable. So far as is known, no such tests have been made in California or elsewhere. Limited tests have shown that the severity and amount of granulation may be noticeably diminished by reducing the amount of irrigation water applied. The reduction may cause at least a small decrease in yield, though not, apparently, in fruit size, and does not need to be very great in order to obtain a commercially important decrease in granula- tion. In the tests described in this paper, the decrease in yield was not apparent until the fifth year of treatment, while the decrease in amount of granulation was plainly evident after the first year. Although there was eventually a slight decrease in total yield, the yield of fancy and choice fruit was probably greater over the entire period because of the marked decrease in the amount and severity of granulation. The irriga- tion tests which appear to justify these statements were made, however, in only one locality (near Santa Ana) and on only one type of soil (Yolo loam) , so that it is impossible to make a general recommendation for con- trolling granulation by this method. It seems probable that any method of treatment which would prevent the production of large sizes would reduce the severity and amount of granulation in the fruit, because it is usually the thriftiest trees that produce the most granulation. Present evidence indicates rather strongly that the control or lessen- ing of granulation in the future depends upon careful choice of root- stock and that for the starting of trees for new groves and for replanting, buds should be taken from trees that have shown a minimum of granula- tion during the early years of their growth. In the Packing -House. — Evidence to date appears to indicate that typical granulation does not become materially worse after the fruit has been brought to the packing-house. It is true that some of the juice sacs which appeared to be healthy, or nearly so, when the fruit was picked will lose moisture and tend to shrink. For this reason the disorder may appear to spread in the tissues during storage or transit, but after the fruit has been removed from the tree, apparently no new juice sacs be- come affected, nor do those already affected continue to enlarge and harden. Granulation of the juice sacs is a growth phenomenon ; there- fore it is scarcely more probable that healthy juice sacs should become Bul. 647] Granulation of Valencia Oranges 59 granulated after the fruit has been picked than that the fruit itself should grow after it has been removed from the tree. Since, as a rule, the disorder cannot be reliably detected by any sur- face markings, samples have to be cut and different portions of a given lot of fruit discarded or graded accordingly. The fluoroscope as used in the packing-houses at the present time is useful in detecting the severely granulated fruits, but is of little aid when the fruits are only slightly or moderately granulated. The type of trouble which is usually referred to as "blossom-end (stylar-end) granulation," has been described in an earlier portion of this paper (see "Description of Granulation and Dry Juice Sac," p. 4), and "dry juice sac" has been suggested as a suitable name for it. This disorder becomes worse as the fruit is held in the packing-house and can be detected only by cutting the fruit or, in the moderate or severe stages, by the use of the fluoroscope. Little can be done to prevent its spread in the fruit, because the real damage was done before the fruit was picked. Close watch should be kept on all lots of fruit that are to be shipped after the middle of July. The fruit may appear to be reasonably sound when picked. Since the affected sacs have largely lost their power to re- tain moisture, storage at a comparatively high humidity is the only suggestion that can be made for temporarily retarding the spread of the disorder within the fruit. Such fruit should be consumed as soon as possible and should not be sent to eastern markets. SUMMARY Granulation has been reported and accurately described from nine foreign countries, and it is probably present in one or more varieties wherever citrus is grown. The Valencia orange is the only citrus variety in California in which granulation is commercially important. In California, granulation is more prevalent in the coastal than in the interior districts. The fruit on young trees is more subject to granulation than that on old trees. Granulation in California does not usually become commercially im- portant until after the middle of the picking season. Reduction of the amount of light by shading trees did not cause an increase in the amount of granulation in the fruit. Granulation originates in and is usually confined to the stem half of the fruit pulp. In spite of the fact that juice sacs usually become enlarged as they become granulated, there is not a consistent enlargement of the stem end 60 University of California — Experiment Station of granulated fruits ; in most cases the effect of the enlargement appears to be distributed over the entire fruit. Spraying the trees with lime apparently reduced the amount of granu- lation in proportion to the concentration of lime used. Lime spray is not recommended as a control, however, because of the effect on the trees during desiccating winds. There is a strong tendency for certain trees in a given grove to be consistent from year to year in the production of much or little granu- lation. Buds from high and low granulation-producing trees are being grown on both sweet- and sour-orange stocks to determine whether the tenden- cies toward the production of granulation may be transmitted through the bud. Other trees have been cross-top-worked for the same reason. Trees affected with scaly bark do not, as a rule, produce a higher per- centage of granulation than those not affected. Reduction in frequency of application and in amounts of irrigation water resulted in a slight reduction of fruit yield in the last (fifth) year of the treatment, but not in fruit sizes. After the first year of differential treatment, the trees on the wet plots produced much more granulation than those on the dry plots. Analysis showed that there were no appre- ciable differences in total soluble solids, acids, pH, and sugars in fruit samples from the wet and from the dry irrigation plots. Certain rootstocks may augment the amounts of granulation pro- duced, but the tendencies in this direction may not be the same in dif- ferent localities. Low temperatures may augment granulation in fruits that are in- clined to granulate anyway, but apparently not in others. Limb girdling did not increase fruit size or have any effect on amount of granulation in the season after girdling. In the second year, there was a general decrease in the size of fruits on the girdled limbs, but there was more granulation in these fruits than in the larger fruits from the limbs that were not girdled. In general, the larger the fruit, the more likely it is to be granulated. Granulation is more prevalent in fruits from the inside and from the north side of the tree than in fruits from any other portion of the tree. The severity of granulation may range all the way from granulation of only two or three juice sacs to granulation of practically the entire pulp of the fruit. There appears to be no difference in the nitrogen content of healthy and granulated juice sacs. Large fruits have a lower total soluble-solids content than small ones, but small fruits are sometimes as severely granulated as large ones. Bul. 647] Granulation of Valencia Oranges 61 Granulated juice sacs have a higher percentage concentration of mois- ture and inorganic matter and more dry matter per juice sac, but a lower active and total acidity and lower concentration of sugars than healthy juice sacs. Juice sacs in stage 2 of granulation contain approximately twice as much total pectin as those that are in a healthy condition. Juice sacs usually enlarge and become comparatively hard during the earlier stages of granulation and then soften and collapse during the later stages. The enlargement is due to cell enlargement and not to cell multiplication. The hardening is due to gelatinization of the cell con- tents and lignification of the walls of some of the internal cells. The softening and collapse are caused by the progressive disintegration of many of the internal cells. Occasionally juice sacs may become granu- lated without becoming very much enlarged. In the intermediate stages of granulation, gas forms and is tempo- rarily trapped within the hard-walled cells. Later the gas escapes into newly formed intercellular spaces and, finally, into the cavity formed by cellular disintegration. The concentration of coloring matter (carotin) in the juice sacs de- creases as granulation progresses. The actual cause of granulation has not been definitely determined, but some of the possible causes are discussed. Control measures are limited, but various phases of the problem are discussed. The apparent inconsistencies in some of the results and the exceptions to experimentally proved trends indicate that a comparatively large number of factors are concerned in the production of granulation. Extensive laboratory studies have shown that many changes occur in the juice sacs during the process of granulation. As a result of these changes and, no doubt, others that have not been mentioned, it is not surprising that the taste and edibility of the affected fruits have been greatly modified. The preceding summary statements refer to typical granulation and not to "dry juice sac." The latter has been suggested in this paper for the first time as a suitable descriptive term for the physiological disorder which in the past has been called "blossom-end granulation." 62 University of California — Experiment Station ACKNOWLEDGMENTS Appreciative acknowledgments are gladly given to Messrs. J. E. En- derud, E. C. Raby, and B. E. Janes, assistants; to Professor M. R. Hu- berty and his assistants, of the University of California Division of Irrigation, who had charge of the irrigation plots for three years ; to Mr. Floyd D. Young, Senior Meteorologist, United States Weather Bureau, who furnished grove data and temperature records; to Dr. Glenn H. Joseph and Dr. Edwin Bryant, of the California Fruit Growers' Ex- change Research Department, for helpful suggestions on pectin de- termination ; to Mr. J . A. Smiley, of Santa Ana, in whose grove many of the tests were made from 1932 to 1938 ; to the Orange County Fumi- gation Company, who applied sprays, gratis, for three years; to the Irvine Company, of Tustin, the Tustin Hills Citrus Association, and the Murphy Ranch Company of Whittier, which furnished many boxes of fruit from each of several groves, gratis, each year for a period of three years; to the North Whittier Heights and the El Ranchito Citrus asso- ciations, each of which furnished fruit from several groves for one year ; to the many packing-house managers who furnished data and fruit for the low-temperature-granulation studies ; and to all others who have so willingly assisted in furthering the study of this problem. Bul. 647] Granulation of Valencia Oranges 63 LITERATURE CITED 1. Bartholomew, E. T., Walton B. Sinclair, and Byron E. Janes. 1938. Soluble solids in citrus fruits. Science 87:584. 2. Bartholomew, E. T., W. B. Sinclair, and E. C. Kaby. 1934. Granulation (crystallization) of Valencia oranges. California Citro- graph 19(4) : 88-89, 106, 108. 3. Bartholomew, E. T., W. B. Sinclair, and E. C. Baby. 1935. Granulation of Valencia oranges. California Citrograph 21(1) :5, 30. 4. Bartholomew, E. T., Walton B. Sinclair, and Floyd D. Young. 1936. Granulated and frozen Valencia oranges. California Citrograph 21(12) : 458. 5. Fawcett, H. S., and H. A. Lee. 1926. Citrus diseases and their control. 1st ed. 582 p. (See specifically p. 426- 28.) McGraw-Hill Book Company, Inc., New York, N. Y. 6. Haas, A. E. C, and L. J. Klotz. 1935. Physiological gradients in citrus fruits. Hilgardia 9(3) : 179-217. 7. Sutherst, W. F. 1911. Sugar content of oranges. California Cultivator 36(20) :612. 8. Turrell, F. M., and E. T. Bartholomew. 1939. Structural and microchemical changes in granulated orange vesicles. California Citrograph 24(3) : 88, 110. 15m-3, '41(9199)