UNIVERSITY OF CALIFORNIA COLLEGE OF AGRICULTURE AGRICULTURAL EXPERIMENT STATION BERKELEY, CALIFORNIA FOUR NEW CITRUS VARIETIES- THE KARA, KINNOW, AND WILKING MANDARINS AND THE TROVITA ORANGE HOWARD B. FROST BULLETIN 597 DECEMBER, 1935 UNIVERSITY OF CALIFORNIA BERKELEY, CALIFORNIA CONTENTS PAGE Introduction 3 The Kara mandarin 7 Tree 7 Leaves 7 Fruit 7 The Kinnow mandarin 10 Tree 10 Leaves 10 Fruit 10 The Wilking mandarin 11 Tree 11 Leaves 12 Fruit 12 The Trovita orange 13 Tree 13 Fruit 13 Strains 13 FOUR NEW CITRUS VARIETIES- THE KARA, KINNOW, AND WILKING MANDARINS AND THE TROVITA ORANGE 1 2 HOWAED B. FROST 3 INTRODUCTION The three mandarins herein described are noteworthy for new and excellent flavors, good size, and a long season, which lasts from January or February to late spring. They are especially good as juice fruits. The Trovita orange may prove to be a useful substitute for the Washington Navel in some of the hotter citrus districts. These varieties originated as seedlings in the breeding work of the Citrus Experiment Station of the University of California, in 1915. In California, varietal standardization in the citrus industry is ex- tremely rigid. For many years planting has been limited almost entirely to the Washington Navel and Valencia oranges, the Eureka and Lisbon lemons, and the Marsh grapefruit — that is, to one variety, or two varie- ties differing more or less in season, for each of the three mainly grown species. Retention of small areas of other species and varieties, however, has depended in part on a special market demand, general or local, for varieties of different flavor and appearance, such as the Dancy tangerine and blood oranges. This special demand might be much increased by the introduction of new varieties combining high quality with other good characteristics. The much more extensive production of the Dancy in Florida is doubtless due very largely to its better size and lower acidity in the Florida climate. Obviously there are important advantages to the growers, packers, and dealers, in this narrow limitation of varieties. Since, also, the devel- opment of new plantings is slow and expensive, and the development of consumer demand for new flavors is laborious and uncertain, the com- mercial establishing of new citrus varieties, especially of markedly different ones, is likely to be a lengthy and difficult process. We may, however, expect the eventual establishment in California of new citrus 1 Received for publication October 12, 1935. 2 Paper No. 335, University of California Citrus Experiment Station and Graduate School of Tropical Agriculture, Riverside, California. 3 Associate Plant Breeder in the Citrus Experiment Station. [3] 4 University of California — Experiment Station varieties, and new strains of old varieties, which will greatly benefit both producer and consumer. This may come about through introduction from other parts of the world, through discovery of bud variations and accidental seedlings, or through special production and trial of seedlings (as with the varieties described below) . The new forms of citrus to be desired may be divided roughly into two classes : (1) forms much like the present major varieties, but in some respects superior from the viewpoint of either producers or consumers, or preferably both; and (2) varieties that are very different from the present varieties, especially ones combining new and excellent flavors with other desirable characteristics. From our present knowledge of citrus breeding, we may expect that new and better varieties of the former kind, especially any that are closely similar to present varieties, will originate mainly by somatic or bud variation. Bud-variation forms are, as a rule, very similar to the parent variety in most respects, and this fact is recognized in the com- mon practice of calling them "strains" of the parent variety. Although bud variation is fairly common in citrus, the origin of valuable new strains seems to be a rare occurrence. Bud variations may be discovered either as "sporting" branches (or trees which have happened to be budded from such branches), or as nucellar seedlings 4 which have origi- nated from cells that have been changed as a result of bud variation. Such a nucellar variant, it is believed, is the Trovita orange, described below. Selfing and narrow crossing have usually given very unpromising results in citrus breeding, presumably because of the presence in the parent varieties of many recessive genes unfavorable to vigor. For ex- ample, weak, unproductive, totally worthless forms have usually resulted when various citrus varieties have been self -pollinated, or when two closely related varieties, such as Ruby and Valencia orange, or Eureka and Lisbon lemon, have been intercrossed. 5 These statements apply, of course, only to the sexually produced seedlings, and not to the accom- panying nucellar seedlings. Although pollination seems to be necessary, 4 Such seedlings arise from extra embryos produced by an asexual or budding process in the nucellus of the ovule (the young seed), and hence are expected, like budded trees, to belong to the same variety as the parent. See : Strasburger, Eduard. Uber Polyembryonie [On polyembryony]. Jenaische Zeit- schrift fur Naturwissenschaft 12:647-670. 1878. Webber, H. J. The economic importance of apogamy in Citrus and Mangifera. American Society for Horticultural Science Proceedings 1931:57-61. 1932. 6 Frost, Howard B. Polyembryonv, heterozygosis, and chimeras in Citrus. Hil- gardia 1:365-402. 1926. Webber, H. J., and J. T. Barrett. Root-stock influence in citrus. Ninth International Horticultural Congress Eeport 1930:358-373. 1931. Bul. 597] Four New Citrus Varieties 5 usually or always, for the formation of nucellar embryos, 6 they inherit from their seed parent alone ; their characteristics are not affected by the source of pollen. On the other hand, hybrids between more distantly related forms, such as distinct species of the genus Citrus (for instance, between tanger- ine and grapefruit), are very largely vigorous. In fact, Swingle 7 has concluded that very wide crossing tends to give the most vigorous hybrid seedlings, as when one parent is a species of true Citrus and the other is trifoliate orange (Poncirus) or kumquat (Fortunella). Further, the great variability of first-generation citrus hybrids pro- vides abundant material for the breeder. The hybrids from one cross usually differ greatly in many characteristics, presumably because of the highly heterozygous (hybrid) nature of most citrus varieties. 8 Rather wide crossing, therefore, seems to be the one promising gen- eral means for producing valuable and decidedly different new varieties of citrus. Although most hybrids are much inferior to the parent varie- ties in some respects, occasionally one is superior in important features and deserves horticultural trial. Especially promising crosses are those between the King mandarin and certain varieties which belong to dis- tinct subdivisions of the same species, Citrus nobilis Lour. 9 These crosses are especially remarkable for pleasant and varied flavors. Three hybrids from such crosses are described below. It must be especially emphasized that the Citrus Experiment Station is introducing these new varieties for preliminary trial only, and that none of them can at present be recommended for commercial planting. It is particularly important to note that up to the present time fruit of these varieties has been observed (with the one partial exception noted below for the Kara, and a very few fruits of the Trovita) only at the Citrus Experiment Station, and their suitability for other climatic regions is unknown. Except as is specially noted below, therefore, the descriptions of fruit are based on Riverside cultures, although trees are growing in other citrus districts of California. 6 Toxopeus, H. J. De polyembryonie van Citrus in haar beteekenis voor de cultuur. [Buitenzorg] Algemeen Proefstation voor den Landbouw. Korte Mededeelingen 8:1-15.1930. 7 Swingle, Walter T. New types of citrus fruits for Florida. Florida State Horti- cultural Society Proceedings 1910:36-42. [1911?] 8 Frost, Howard B. Polyembryony, heterozygosis, and chimeras in Citrus. Hil- gardia 1:365-402. 1926. (See also papers by Webber, by Swingle, and by the Hagedoorns, cited in that paper.) 9 According to W. T. Swingle (Citrus. In: Bailey, L. H. Standard cyclopedia of horticulture vol. 2, p. 784. 1914), the King represents the type of the species; the narrow-leaved mandarins, such as the Willow Leaf, and the tangerines, the botanical variety deliciosa; and the satsumas, the variety unshiu. 6 University of California — Experiment Station The oldest Riverside planting of these varieties consists of two trees of each variety budded on trifoliate-orange stocks; these were set in a crowded trial orchard in 1920 and 1921. In 1930-32, nine to fourteen trees of each variety were planted at normal distances in another or- chard. The mandarins in this young planting were budded on the fol- lowing stocks : sour orange, sweet orange, trifoliate orange, Rough lemon, grapefruit, Cleopatra mandarin (Kara and Kinnow only), and Cunningham citrange (Kara only) ; the Trovita was budded on sweet, sour, and Cleopatra stocks. These trees are too young to furnish reliable evidence on stock adaptation, but so far the tree growth has seemed satis- factory on all stocks used, and considerable fruit has been produced on the first five stocks, as listed above, by the Kara and Wilking, on the first three by the Kinnow, and on the first two by the Trovita. It is probable that these mandarins, like the Dancy tangerine, will succeed on a considerable range of stocks. The Rough lemon has elsewhere been found to be an unsatisfactory stock for varieties of the mandarin group, because of a tendency to produce coarser fruits, with more rag and puffing, than do other ordinary stocks. A little evidence has already been secured that Rough-lemon stock increases the tendency of the Wilking to undesirable firmness of pulp, and it is not to be recommended for that variety. For the Kara, however, it deserves further trial, to- gether with other stocks ; it may prove to be valuable, since it seems to reduce the acidity at the expense merely of a moderate shortening of the season. Most of the trees in the young planting were propagated from bud- wood selected for thornlessness, and the mandarins retain little of the seedling thorniness, although small thorns are still frequent on the Kinnow. The Wilking has been the earliest of these mandarins to come into bearing, and the Kinnow the latest; the Kara so far gives most promise of regular annual bearing, and of maintaining good size with heavy yield. The Trovita orange still retains considerable seedling thorniness, but this will doubtless decrease as the variety becomes older. In the descriptions of the fruit of the mandarins, especial weight is given to the evidence from the young Riverside orchard ; the descriptions of tree and leaf characters are based mainly on these vigorous young trees. The descriptions and illustrations of the fruit are made from fresh specimens, either just picked or stored for a short time. Except as is noted in the legend of figure 1, all the illustrations show fruit from River- side trees; figures 2 to 4, and the Orange County part of figure 1, are from young trees. Bul. 597] Four New Citrus Varieties 7 The composition analyses of fruit were made approximately accord- ing to the methods officially prescribed for maturity testing of oranges in California (the so-called "8-to-l-test") . Samples of 20 fruits each were used, with a Sunkist motor extractor ; the soluble solids in the juice were determined with a hydrometer calibrated for pure solutions of sucrose, and the acid (as anhydrous citric acid) by titration with sodium hydrox- ide and phenolphthalein. The number of trees, years, and determinations represented are small, and the values stated are to be considered merely as rough indications of expectation under Riverside conditions. 10 In order to facilitate adequate trial of the varieties herein described, budwood in very small amounts will be furnished free of charge to citrus nurserymen and growers prepared to use it. The supply of bud- wood is small, but applications will be filled as far as practicable until sufficient distribution has been made to provide for a thorough trial of the adaptability of these varieties to the principal citrus regions of California. 11 THE KARA MANDARIN This is a first-generation hybrid of satsuma (believed to be the variety Owari) as seed parent, pollinated by King mandarin; the name is de- rived from the names of the parents (Kara instead of Kari, to obviate uncertainty about the pronunciation) . At Riverside the Kara is produc- tive ; the fruit is of good size and excellent quality ; the acidity is high in the earlier part of the season and quite sufficient later, so that this variety is especially good for juice ; the rind puffs somewhat. Trees of this variety are growing well in the principal citrus districts of Cali- fornia, but seem more vigorous near the coast than in the hottest in- terior regions. Some fruit was produced near Tustin, in Orange Comity, California, in 1934, and this was decidedly unlike the Riverside fruit described below ; it was flatter and smoother, somewhat smaller, usually entirely without neck, and nearly seedless (fig. 1). Tree. — Top larger than Owari satsuma, vigorous; shape similar to satsuma, globose ; moderately open, often irregular in outline ; branches moderately numer- ous, stout, mainly spreading, the less vigorous ones very often drooping; fruit Avell scattered through tree, mainly shaded. Leaves. — Much resembling satsuma in size and shape, but somewhat smoother and darker green; large (maximum size of blade, about 2% x 6^ inches, or 7 x 16 centi- meters), rather thick and rigid; shape broadly lanceolate; base acute; apex mod- erately acuminate; margin crenulate-serrulate ; usually somewhat concave above; petiole long (usually 17 to 25 millimeters), narrowly to very narrowly winged (more than that of parents). Fruit. — Size (fig. 2) of small or medium orange, averaging 3 to 5 fruits per pound ; diameter 2% to 3 inches (5% to 7% centimeters), height 2% to 2% inches (5% 10 The analyses were made by A. C. Austin, Laboratory Assistant in the Citrus Experiment Station. 11 Address requests to Director, Citrus Experiment Station, Riverside, California. University of California — Experiment Station Fig. 1.- — Typical fruits of Kara; April, 1934; natural size, side and apical views. Upper two, grown near Tustin, Orange County ; lower two, grown at the Citrus Experiment Station, near Eiverside. Note differences in size, shape, smoothness, and color. Bul. 597] Four New Citrus Varieties to 7 centimeters) ; shape usually slightly to medium oblate (flattened), often near globose, very often somewhat obconic, rather variable ; very often with a small neck, which is variable and somewhat irregular ; often slightly depressed about button, fur- rowing slight; apex considerably flattened, commonly with a small basin; surface Fig. 2. — Typical fruits of Kara, harvested May 28, 1935 (4 months after ripening) ; natural size. moderately grained and often, especially near apex, somewhat bumpy and wrinkled ; surface color deep orange to deep orange-yellow (about midway between orange chrome and orange rufous, E. pi. II, to about midway between cadmium yellow and raw sienna, R. pi. Ill 12 ) ; larger oil glands mainly conspicuous on rind surface ; rind 12 Each reference such as "R. pi. Ill" indicates that the color name or names pre- ceding in the same parenthesis occur in Ridgway's Color Standards, on the plate indicated. Ridgway, Robert. Color standards and color nomenclature. 43 p. 53 col- ored plates. Published by the author, Washington, D. C. 1912. 10 University of California — Experiment Station thin to occasionally medium (usual range, % to Y 5 inch, or 3 to 5 millimeters), peeling easily and cleanly, often puffing after midseason, rather tough ; oil glands very many, size medium ; oil abundant, odor pleasant and rather strong ; axis moderately hollow ; segments 10 to 13, separating easily, soft ; membranes thick near axis, elsewhere mainly thin but tough ; pulp color deep, rich yelloAvish orange (near xanthine orange, E. pi. Ill) ; pulp tender, very juicy; juice vesicles of medium size, rather stout; aroma of juice unique, strong, very pleasant to most persons; sugar high (soluble solids of juice high, about 15 to 17 per cent at full maturity and about 2 per cent less at early maturity) ; acid rather high for mandarin (1% to 2 per cent at early maturity, and above 1 per cent until much overripe); flavor rich and excellent, but usually sour and requiring added sugar during the earlier part of the season; seeds medium in number (averaging about 12 to 20 per fruit), medium to large in size, very plump; cotyledons yellow-tinged to pale greenish yellow; average number of embryos in 50 seeds dissected, 3.2; seedlings from self or cross-pollination nearly all of nucellar origin; season, January or February to April or May. THE KINNOW MANDARIN This is a first-generation hybrid of King as seed parent, pollinated by Willow Leaf (China) mandarin; the name is derived from the names of the parents. Fruit solid even when overripe (see fig. 3), excellent in appearance, and also in flavor until the acidity becomes too low with overripeness ; superior in external appearance to Kara, but rind easily injured and probably unsuited to distant shipment ; size less well main- tained with heavy yield than in the case of Kara. Tree. — Vigorous in all districts; top large for mandarin, somewhat exceeding Kara ; tall, erect, shape ellipsoid columnar to obovoid, very symmetrical ; dense, outline even ; branches very numerous, very many of them long, slender, mostly erect or ascending; fruit largely scattered and very heavily shaded, but also largely borne (sometimes in clusters) at the tips of ascending branches. Leaves. — Very numerous; size medium (maximum, blade about 2% x5 inches, or 5y 2 x 13 centimeters) ; rather thin and pliable (about as King) ; shape intermediate between the parents, broadly lanceolate ; base somewhat rounded, rather obtuse ; apex moderately acuminate ; margin crenulate-serrulate ; petiole of short-medium length (usually 8 to 15 millimeters), practically wingless to very narrowly winged. Fruit. — Size (fig. 3) of small or medium orange, averaging about 3 to 5 fruits per pound; diameter 2% to 3 inches (S 1 /** to 7Y 2 centimeters), height 2 to 2% inches (5 to 6y 2 centimeters) ; shape medium to slightly oblate, regular and uniform; base somewhat flattened, often very slightly depressed, furrowing slight ; apex much flattened, commonly with a very shallow basin; surface often very shallowly pitted, commonly almost perfectly smooth, very glossy, scarring easily; surface color re- sembling King, deep yellowish orange (about midway between cadmium orange and xanthine orange, to about midway between orange and Mars yellow, E. pi. Ill) ; larger oil glands mainly very conspicuous on rind surface, rind color between them much lighter (mikado orange to light orange-yellow, E. pi. Ill) ; rind thin (averag- ing about Vs inch; usual range 2 to 4 millimeters), peeling fairly easily, practically no puffing, rather tough ; oil glands many, size averaging medium but seeming espe- cially variable; oil abundant, odor pleasant and moderately strong; axis solid to slightly hollow ; segments 9 to 12, separating fairly easily, rather firm ; membranes thick near axis, thin and tender elsewhere; pulp color deep yellowish orange (near xanthine orange to Mars yellow, E. pi. Ill) ; pulp moderately tender, very juicy; juice vesicles of medium size, stout; aroma of juice unique, moderately strong, very pleasant; sugar very high (soluble solids of juice very high, about 15 to Yiy* per cent at full maturity and nearly reaching the maximum early in the season) ; acid moderate (near l 1 /^ per cent at early maturity, and falling below 1 per cent late in the season) ; flavor very rich, excellent, becoming very sweet late in the season; seeds medium in number and size, averaging about 12 to 24 ; cotyledons somewhat yellow- ish; average number of embryos in 50 seeds dissected, 2.7; seedlings from self- pollination nearly all of nucellar origin; season, January or February to April or May. Bul. 597] Four New Citrus Varieties 11 Fig. 3.— Typical fruits of Kinnow, harvested May 28, 1935 (4i/> months after ripening) ; natural size. THE WILKING MANDARIN This is a first-generation hybrid from the same cross as the Kinnow (King x Willow Leaf), but the tree and fruit differ in many ways; the origin of the name is obvious. The fruit is firm, puffing very little, and probably would stand handling well ; the flavor is very good ; with light yield the pulp is sometimes too firm, and with heavy yield the size may be small. Tree. — Top vigorous, hardly as large as Kara; shape near globose, to oblong columnar, fairly symmetrical ; dense to rather open, outline much less even than with Kinnow; branches numerous, many of them long, rather stout, largely erect or 12 University of California — Experiment Station ascending, the less vigorous ones often somewhat drooping; fruit very generally distributed, often much of it exposed to sunshine. Leaves. — Very numerous; size medium, intermediate between parents (maximum, blade about 2 x 4% inches, or 5 x 12 centimeters) ; thickness medium; shape broadly lanceolate to medium lanceolate, intermediate between the parents; base acute; apex Fig. 4.— Typical fruits of Wilking, harvested May 28, 1935 (4 months after ripening) ; natural size. decidedly acuminate; margin slightly crenulate; usually somewhat concave above; petiole of short-medium length (usually 9 to 14 millimeters), practically wingless. Fruit. — Size (fig. 4) of small or medium orange, averaging about 3 to 5 fruits per pound; diameter 2% to 3% inches (5% to 8 centimeters), height 1% to 3 inches (4^ to 7% centimeters) ; shape usually medium oblate, fairly regular and uniform; base somewhat flattened, stem cavity slight, base moderately furrowed; apex flat- tened, basin usually medium; rind surface rather coarsely and shallowly grained to nearly smooth, glossy; surface color deep yellowish orange (Ridgway's colors are as given for Kinnow above) ; larger oil glands mainly very conspicuous on rind Bul. 597] Four New Citrus Varieties 13 surface; rind thin to medium (averaging near % to Ve inch; usual range, 2 to 5 millimeters), adherence rather slight, rather brittle, peeling fairly easily, very little puffing; oil glands very many, size medium; oil very abundant, odor pleasant and rather strong; axis moderately hollow; segments 9 to 13, separating easily, firm; membranes tough to fairly tender; pulp color rich yellowish orange (somewhat brighter than xanthine orange and Mars yellow, E. pi. Ill) ; pulp moderately tender, very juicy, or firm with light crop (occasionally excessively firm, especially on Bough- lemon stock); juice vesicles of medium size, stout; aroma of juice unique, rather strong, pleasant; sugar high (soluble solids of juice high, about 15 to 17 per cent at full maturity and about 2 per cent less at early maturity) ; acid moderate (per- centages about as with Kinnow, but evidently tending to be a little higher) ; flavor rich, very good, very sweet late in the season; seeds medium in number and size, averaging about 10 to 17 per fruit, mostly plump but rather often empty or poorly filled; cotyledons pale greenish yellow to slightly yellow-tinged; rarely more than one embryo (only 51 embryos in 50 seeds dissected) ; seedlings from self-pollination nearly all of sexual origin (true second-generation hybrids) ; season January or Feb- ruary to April or May. THE TROVITA ORANGE This is a good early non-navel orange, which has pollen and a few seeds. Because of its seeds, it seems promising for trial in the hotter citrus dis- tricts where the seedless variety Washington Navel often fails to set good crops. The name (pronounced tro-vee'ta) is Esperanto for found, and was suggested by the accidental finding of the seeds which produced the variety. Three seedlings, doubtless nucellar, all Trovita, were grown from seeds found in one fruit in the main "Washington Navel orange crop of 1914-15 at the Citrus Experiment Station. The parent tree was not located, but the record of the fruit does not mention any peculiarity. Probably a bud variation had occurred in an ordinary tree of navel orange. The seediness of Trovita in isolated plantings is unknown, since so far all trees have been exposed to pollination by other varieties. Very inadequate evidence has been secured on productiveness, although in the Riverside plantings mentioned above it seems at least equal to that of comparable Washington Navel. Tree. — In most respects about like ordinary Washington Navel but showing the greater vegetative vigor which is characteristic of citrus seedlings in general, and therefore somewhat more like nonvariant nucellar seedlings of Washington Navel. Fruit. — Much like Washington Navel in most respects, but the navel structure (extra whorl of carpels) mainly absent or rudimentary, usually with no opening in rind (fig. 5) ; apex of fruit not bulging, but slightly basined as with Valencia; occa- sional twigs or small branches bear navel fruit ; ripening about at the same time as Washington Navel; pulp similar, apparently somewhat more tender and juicy; aroma of juice somewhat unlike Washington Navel, not quite so strong or as good; acid seeming typically lower, and solids-acid ratio higher; rag rather similar to that of Washington Navel, but thicker or tougher near axis, so that the peeled fruit slices somewhat less easily ; in a mixed planting, average number of seeds ranging approxi- mately from 2 to 6 per fruit. Strains. — At the Citrus Experiment Station, and in several cooperative plantings elsewhere, two strains of the Trovita are under trial. These two strains, derived from two of the three original Trovita seedlings, differ in the amount of pollen in the flowers and the number of seeds in the fruits. Only the strain with more pollen and seeds is being distributed in response to requests for budwood, because the other strain seems to be genetically variable, showing much tendency to produce navel-marked and seedless fruits. In fact, studies in progress while this bulletin 14 University of California — Experiment Station is in press show that in the 1935-36 crop 13 on a few trees just coming into bearing (in the second "budding generation" from the seedlings), in several localities, the former strain commonly has a few seeds, while the latter strain is very largely seedless. Even the former strain is probably liable to occasional bud variation to a navel type. Present recommendations, therefore, are that, if budwood is taken Fig. 5. — Fruits of Trovita orange (columns B and D), and of a nucellar seed- ling of Washington Navel which seems to be typical of the latter variety (columns A and C). Harvested May 21, 1929 (overripe). from trees of the Trovita in any of the cooperative plantings, it be taken only from trees listed as belonging to the "strain with more seeds" (trees bearing the parent number "Bll, 12" on their pedigree labels). At least the present trials of the less seedy strain should be continued, however, since its scarcity of seeds is desirable if its productiveness should prove satisfactory. 13 Somewhat favorable preliminary indications have been given by the Trovita near El Centro and in the Coachella Valley, and in the latter region by the man- darins described above. There is an indication that the acidity of the Trovita, but not that of the Kara and the Kinnow, is likely to be definitely lower in the regions mentioned than at Riverside. iim-l,'3(