JNITED STATES DEPARTMENT OF AGRICULTURE „ BULLETIN No. 924 4 ififc % jjg? Contribution from the Bureau of Plant Industry > Jf&f^&J-U WM. A. TAYLOR, Chief f&F^mTU Washington, D. C. PROFESSIONAL PAPER January 26, 1921 TEAR-STAIN OF CITRUS FRUITS. By John Kt Winston. Pathologist, Office of Fruit-Disease Investigations. CONTENTS. Page. Descripf ion of the disease 1 Review oftbe literature 2 Spraying experiments 3 Cultural work 5 Histological examinat ion 9 Page. Inoculation experiments 9 Conclusions 11 Summary II Literature cited : 12 DESCRIPTION OF THE DISEASE. Florida citrus fruits are subject to two distinct types of tear streaking, namely, wither-tip tear-stain and melanose tear-streak. The former has been attributed to Colletotricfium gloeosporioides Penz., while the latter, which will not be discussed in this bulletin, is doubtless due to a peculiar distribution of the spores of the causal organism (Phomopsis citri Fawcett) in trickling waters. Wither-tip tear-stain is a smooth, more or less brownish discolora- tion of the surface which occurs typically in fingerlike patterns about one-fourth of an inch in width, extending longitudinally toward the stylar or blossom end of the fruit. (PL I.) These streaks may or may not be confined to one side of the fruit, and frequently they appear to arise in areas russeted by rust mites [Eriophyes oleiveros Ashmead). The streaks are usually few in number, seldom more than six or eight; occasionally two or more merge and form a rather wide discoloration, in which event the injury is generally attributed to rust mites. (PI. II, fig. 1.) Wither-tip tear-stain as it occurs in Florida is one of the minor diseases of citrus fruits, and for that reason it has received little attention from investigators. Its effects are principally observed on the round orange and grapefruit, more noticeably and frequently on the latter and to a less extent upon other economic species of citrus. 15580°— 21— Bull. 924 2 BULLETIN 924, U. S. DEPARTMENT OF AGRICULTURE. The economic importance of this blemish is chiefly due to the financial loss which the grower sustains as a result of the lowered market value of the affected fruit, which is rendered unsightly and unattractive. However, tear-stained fruit seems to possess flavor and keeping qualities essentially equal to fruit which is free from these markings. REVIEW OF THE LITERATURE. A careful survey of the literature reveals the fact that very little has been published on the subject. The cause of tear-stain was investigated and first reported on by Rolfs (7), 1 who is quoted as follows : This peculiar form of russeting manifests itself by streaks running over the fruit. from the side that hangs uppermost to the point which is nearest the ground. The cause of this peculiar form of russeting is that somewhere above the fruit a twig occurs which has been infested by the fungus 2 and which contains the hold-over spores. Whenever sufficient rain or dew occurs to cause a dripping from the disease- infected twig on to the fruit, the disseminating spores are liberated from the twig and earned with the rain or dew over the epidermis of the fruit. Many of the spores are left along in different places, and these produce sufficient irritation to the epi- dermis to cause russeting along in streaks; hence we have the peculiar form of tear staining or tear streaks. Again in 1911 Rolfs (8) discussed the subject as follows: Russeting and tear streaking can nearly always be traced back for their beginning to a small dead spur or sprig. The fungus 2 lives in the dead spur or sprig. Water from rains and moisture following heavy dews collect in drops on these sprigs or spurs and the drops when they fall carry with them numerous fungous spores. These spores come in contact with the epidermis of the fruits and germinate, causing minute lesions on the epidermis, too small for complete infection and the production of anthracnose. This explanation as presented by Rolfs, to the effect that C6l~ letotrichum gloeosporioides is the responsible agent, seems to have been accepted by several investigatqrs at various times, among them Fawcett (2-5), Stevens (9), Stevenson (10), and Matz (6). On the other hand, Earle and Rogers (1) recognized in the Isle of Pines two types of russeting or tear staining said to be easily distinguishable, the one attributed to rust mites and the other thought to be caused by a fungus. In discussing wither-tip they make the following statement: One of the serious troubles attributed to the wither-tip fungus in Florida is that water washing down from infected twigs will cause the serious discoloration of the fruit known as tear streaking. The discoloration is often confused with the work of the rust mite, though it is easily distinguished. It frequently occurs here and can be always traced to some dead twig, but we are still uncertain whether it is due to the secretions of the wither-tip fungus or the Diplodia. • The serial numbers in parentheses refer to '• Literature cited " at the end of this bulletin. i iillitotrichum gloeospori< U'< s ii.eiilionni in the context q FEB 281921 ooGUMH&wrs division <\J° TEAR-STAIN OF CITEUS FRUITS. 6 The publications cited contain practically all the more important references to wither-tip tear-stain in Florida or near-by sections where the round orange and grapefruit are grown in large quantities, and no presentation of actual experimental data is made therein on which the fungus theory regarding the cause of tear-stain is based. However, it can not be denied that this theory, first stated by Rolfs, is eo plaus- ible that it has been accepted readily without the support of published experimental data. It was not until results very strikingly incon- sistent with this theory appeared in experimental spraying tests that its validity was questioned, and this has led to a thorough investiga- tion of the causation of the type of injury commonly known in Florida as "wither-tip tear-stain." The evidence presented in this bulletin indicates that Colletotrichum gloeosporioides is not responsible for tear staining in Florida. If this fungus ever produces such an effect, it must be extremely rare in that section. SPRAYING EXPERIMENTS. During the season of 1917, in connection with spraying experiments which were conducted in a bearing grapefruit grove for the control of citrus scab, it was observed that tear r.taining was almost entirely absent from those plats that received several applications of lime- sulphur solution, though it was quite prevalent where Bordeaux mixture had been applied during the scab-spraying season. Since this first observation, many similar manifestations of the lack of control of tear-stain by copper sprays have appeared in the experi- mental plats. Such a failure has been proportional to the severity of rust-mite attack. Ordinarily the scab-spraying season in Florida begins in February when the spring growth starts and extends over a period of two or three months. It is followed rather closely by the period of very severe attack by rust mites. The latter are usually most abundant during May and June and are readily controlled by two or three appli- cations of lime-sulphur solution diluted 1 to 66. With this explana- tion the data in Table I can be readily interpreted. In general, the results of this spraying experiment show (1) that tear-stain was controlled on plats 3 and 5, which received applications of lime-sulphur solution shortly before the period of maximum abun- dance of rust mites; (2) that tear-stain was materially reduced on plat 4, which received applications of lime-sulphur solution consid- erably in advance of the period of maximum numbers of rust mites; (3) that tear-stain was not reduced on plats 1 and 2, which received spray applications of copper mixtures. Assuming that tear-stain is caused by a species of Colletotrichum or other fungus, it would be reasonable to expect that plats 1 and 2 would show at least as good control as plat 4. The fact that control of tear-stain was so closely BULLETIN 924, U. S. DEPARTMENT OF AGRICULTURE. associated with rust-mite control is strong evidence that rust mites might be largely concerned in the causation of the condition that is generally recognized as wither-tip tear-stain. Table I. — Effect of various sprays on, the control of tear-stain on grapefruit. Date. 1917 Jan. 29 Mar. 23 Apr. 7 May 2 and 22.. June 1918. February (final count of 1,000 'fruits), 1 car- stained, per cent . I'lat designation and spray application. No. 1, Bordeaux mixture. 3-4-50. 3-4-50. 3-4-50. 3-4-50. Not sprayed. No. 2, Burgundy mixture. 3-3i-50.... 3-3 V50.... 3 :; ( 50... 3-3^-50... Not sprayed. 3S.2 No. 3, Bordeaux mixture. 3-4-50 3-4-50 3-4-50 Lime-sulphur, 1-40. Lime-sulphur, 1-66. 1.3 No. 4, lime- sulphur. 1-40 1-10 1-40 Not sprayed. do... No. 5, lime- sulphur. 1-40.... 1-40 1-40.... 1-40.... 1-66.... No. 6. Not spraved. Do. Do. Do. Lime-sulphur, 1-66. 25.9. a The excessive number of tear-stained fruits on the plat sprayed with Bordeaux mixture may be explained by the fact, quite frequently noted, that rust mites become unusually abundant on trees sprayed with that material. The injury did not resemble Bordeaux russet, but was similar in every respect to the tear staining on plat 6. A part of a grove of grapefruit was sprayed experimentally in June, 1919, using 3-4-50 Bordeaux mixture for plat 1 and 1-G6 lime-sulphur solution for plat 2. Plat 3 was an unsprayed check. This grove had not been sprayed previously that season. At the end of July the fruit in plat 2 was free from tear staining, but numerous tear-stained specimens were observed in plats 1 and 3. The fiuit on these plats ripened and was harvested in February, 1920, with no increase of tear- stain over that observed in July, 1919. This indicates that one appli- cation of weak lime-sulphur solution applied at a time suitable for rust-mite control in that grove was equally effective for the control of tear-stain, while the standard strength of Bordeaux mixture applied at the same time did not reduce this blemish. (PI. II, fig. 2.) Other experimental data obtained during the past four years show without exception similar results. The conclusion of Yothers (11), based on numerous extensive experiments to control rust mites in various parts of Florida, is that bright fruit — i. e., fruit free from rust-mite injury — is invariably free from tear-stain as well. (PI. II, fig. 1.) Growers generally throughout the Florida citrus belt have accom- plished commercial control of tear-stain whenever they have con- trolled rust mites by following the spraying schedules established for these pests. Definite observations in commercial groves show clearly the following facts: (1) Where rust mites are naturally absent tear- stain is not observed; (2) where rust mites occur and are successfully controlled tear-stain is also controlled; (3) where rust mites are Bui. 924, U. S. Dept. of Agriculture. Plate I. Bui. 924, U. S. Dept. of Agriculture. Plate II. Fig. I.— Damage to Grapefruit by Rust Mites. Dark tear stains are found frequently on areas more or less rnsseted by rust mites, especially where' the applications of rust-mite sprays were delayed. For all practical purposes, both tear-stain and russeting are absent from fruit kept free from rust mites. Note the zone of faint russet. / w Fig. 2.— The Work of Rust Mites on Grapefruit Sprayed in June with Bordeaux" Mixture. In the zone of rust-mite activity injury is Confined to the area not covered with spray. Near-by grapefruit trees sprayed on the same date with lime-sulphur solution (1-66) matured fruit entirely free from russeting or tear-stain. TEAR-STAIN OF CITRUS FRUITS. present and are not controlled tear-stain occurs and usually varies proportionately with the severity of the rust-mite attack; (4) rust- mite russet and wither-tip tear-stain develop simultaneously Usually in Florida a small proportion of late citrus blooms occur in June. Fruits from such blooms develop after the normal rust- mite period and almost invariably ripen free from tear staining unless rust mites happen to become abundant during the following winter or early-spring months. CULTURAL WORK. Fungi are admittedly capable of producing tear-streak patterns on the host; for example, Phomopsis citri is known definitely to produce melanose injury in distinct tear streaks on various citrus fruits. Consequently a careful study was made to determine the frequency of the occurrence of Colletotrichum gloeosporioides or other fungi in typical wither-tip tear-stain areas. In preliminary tests made during the fall of 1918 oranges and grapefruit were selected that showed typical wither-tip tear-stain, as well as fruit affected with rust-mite russet which shaded off into tear-stain. Usually 30 cultures were made from each fruit, 10 each from areas selected as follows: (1) Typical rust-mite russet, (2) typical tear-stained streaks or slightly discolored areas on fruits show- ing no streaking, and (3) normal areas not discolored. The fruits were washed 1 minute in 1-1,000 mercuric-chlorid solution and afterwards thoroughly rinsed in sterile water. Bits of tissue were then removed with flamed instruments to sterile Petri dishes and covered with melted corn-meal agar. These cultures were allowed to grow five days at room temperature. The results of this pre- liminary culture test are found in Table II, the percentages being based on the number of the bits of tissue used in each test. Table II. — Culture tests for the isolation of Colletotrichum gloeosporioides ' from normal and affected parts of orange and grapefruit rind. Series. Number of fruits. Bits of tissue cultured. Colletofri- chum gloeospori- oides. Miscel- laneous organ- isms. No growth. 17 16 19 170 160 190 38.2 21.1 6.2 18.8 19.3 15.7 43.0 59.6 78.1 1 The use of the name C ilU totrich u m glut osporioides in the culture data of this publication merely follows established custom and is without prejudice as to the proper designation of the organism. Any Colletot- richum showing characteristic growth on corn-meal agar was recorded as this species. It thus appears that Colletotrichum gloeosporioides is recovered in rather low percentage from tear-streaked areas, that it is present to some extent in normal as well as badly russeted areas, and that the frequency of occurrence varies directly with the degree of injury. BULLETIN 924, U. S. DEPARTMENT OF AGRICULTURE. '£ »> o S V-o c =o e r - 8 gs O ri s.§ „fSS C ~ °0 O 'J o lO — :■: <-~ t- oc i- 'O C 'co ^H lO 00 CM CO 00 Xi te XI o" 03 ■^ 1-5 coo CMCNr-Jxo CO t~ •C*j io o c ■* CO ■* 'E bi) uo CO 00 CO ■c t^ lO wo 1^ K ■*cc COCo'uo'iO o 10 CM u 03 lI3 C CO CO -^ CO CO t* ■* ■c i- £ rt CC cj 03 CD ■c ^ 5 ■* 'OCO i-l o ^ U5 ■o U0 CB CO .5 n pf c 1-1 "- 1 - 1 1 ci ^5 *o lO "Eh ON 00 00 1,5 s 3S '-. CC t^ d "" CMIMCM CM o lO «c t^. •o r- Ci CO CO CO QJ 1 CO ■* co ^ CO " o lO t^ C, 00 t^ CO co r^ cci t^ c-I io i-l o -r «# T-H o o >o '- rt rt re <. i ■- -i W^J CO CO '" H d Cirt r- co -^* o >-o COOl O ■CM 0T' CO r^ IT. CO rHCCCNC CJ :•< z »o •O) t^ CO CO T. £ J3 ri so f-C4 'O 03 CM rH io oo «od tfo t^en CO • o CM ^ u- CC 1-1 ""* rH H CM CM : rt CM C0 y ~ t si uo t^ O'J r~»o 1^ Ct ^1 — -f CO O o c- u- o c rH O) CM ^H CN C CO i" -1 co C" - .s Eh d t- POCNO 00 o „ CO o~ C 00 'J s C-l o o- *T t^ -d ■o .3 +3 5 c ~ ~ -r c. o oc oc CM -3 o c ° ci a ^ oi CN CO CN o C-4 si m £ "C o O ? ^ 00 CO »-H --^CO CO CO CO CO TJ4 -rjt I- uo c/ CN CO d _ o • no c H(OK NN COO 00 o o> o c i -r 4 c t COCN CM CO CO (B *5 '~ H *~^ 0^ rH 3 ■c CC »o < 03 CO ? i to tr. 1 - r j o c i cc oc t^ w CX) OS o c ^ :■: . i iH 1-1 *-r 1-- : . . i 1-1 " CM CM CJ oil CM I - t - CC o C-i -X: ^ CO N r^O _ CO 00 N W -r CO -r cc cc CC -^ CM CO ■* CO 5 1 - : "t 3 -^ cu M ' a a -- - a c £ ■fci.B o5 mite russi ing tears' ssetcd are mite russc r stains Dther side n one side one side o '& a;— i.ti i «s«.ssa c/5 i 7 03 ■3 -- Et fruit . -stained tear sta red tear iins ight. -mite ru ly russet ') rust-m ) rust-m CO — lite russ sible ble: -mite ru E-§ E-KS SfcPiQ ^s-r^s^Sfcoss ^ t" ; ? E a. :£ o 'ouco oo ooo ooc o o a a, o o a C'ooooooo c •~ 1.T3 pJ'CO CCCO r O r O f O r O r C'C •CO a QSCG CC CO rs "C ■C'O "CC X £H ■ 03 03 03 03 ■ o3 - I t-i o :o to tX) i (=1 42 : : 'A ■ '. .} II 11 i +^ ■j. f- *^ co u p '3 '3 '3 •a '3 '3 3 •3s'§ '2 1 "c * "C II "S CO CM CO C-l > c-i-r CM "*" — ■* CM C-J Cs -t o °§ .g < - O -JOJO"' 1 . -< C C ^ 5 S P4J« +J -U -U +^ +- +J +^. +J PP4J +i +J ^~ +. '" X ti c 3 OOO O OOOC OO O OCO O o c s r" W H- ] 1- - H - r- £ H ~ o H^ ^ ~- ^ I- — h- TEAR-STAIN OF CITRUS FRUITS. 7 In the fall of 1919 a more extensive test was conducted. Fifteen lots of fruit were involved. Five of these lots were selected by the writer, and the remaining ten lots were selected in various parts of Florida by persons specially chosen for their competence to select typical rust-mite injury and typical wither-tip tear-stain. Each of these lots was sorted into several groups of one to four fruits each according to the variety of fruit, the particular type of effect, and the intensity of it. For a comparative study, cultures were made from fruits in the same lot that were free from blemishes or from unblemished areas on the russeted or tear-stained fruits. One hundred bits of tissue, approximately 1 square millimeter in surface area, were cultured from each test area of each fruit, using: 10 Petri dishes, each with 10 bits of tissue. The results when reduced to a percentage basis, as in Table III, also represent the average numbers of occurrence per fruit. Corn-meal agar was used as a culture medium and the plates were held six days at room temperature. Counts were made of the common saprophytic type of Cladosporium as well as of Colletotrichum colonies. Bacteria and fungi other than these were reckoned as miscellaneous. Two parallel series were made, one for undisinfected tissue and one from similar areas on the same fruits washed with a disinfectant. Bichlorid of mercury solution (1 to 1,000) was used for 1 minute with subsequent rinsing on all disinfected lots except K, L, and M; on these three lots undi- luted fresh hydrogen peroxid was used without rinsing. Table III gives the results, with fractions omitted for the higher percentages. Table III shows that Colletotrichum s attacking < itrus trees in Florida. L T . S. Dept. Agr., Farmers' Bui. 933, 3.8 p., 24 fig. O LIBRARY OF CONGRESS 002 810 981 7 ADDITIONAL COPIES "F THIS PUBLICATION MAT BE PROCURED FROM THE SUPERINTENDENT OF DOCUMENTS GOVERNMENT PRINTING OFFICE WASHINGTON, D. C. 5 CENTS PER COPY