are largely physiological, but can be long-lasting. Breeding and development Yields of older, established nucellar budlines have often been higher than those of their parent lines. James W. Cameron, Professor of Horticulture But most importantly, many viruses and viruslike and Geneticist, Emeritus, and agents are eliminated at embryo formation. This Robert K. Soost, Professor of Genetics, has led to widespread propagation of nucellar lines Department of Botany and Plant Sciences and has contributed to the improvement of citrus plantings in many countries during the last two decades. hen Howard B. Frost began citrus breeding at the University of California Citrus Ex- W periment Station in 1914, knowledge of Inheritance of nucellar embryony crossing relationships in the genus was limited to One problem associated with citrus breeding- earlier studies in Florida by the U.S. Department as with many tree fruits-is the scarcity of char- of Agriculture. Frost began a wide series of crosses acters controlled by one or a few genes. Most char- among edible types, but the numbers of hybrids acters are inherited quantitatively, involving many first obtained were limited by nucellar embryony. genes. In addition, genes are apparently seldom This is the phenomenon by which somatic cells of homozygous (that is, existing in two identical the nucellus (tissue in the ovule but outside the copies, or alleles) within a clone. Thus, fruits embryo sac) develop into embryos. Since these showing a wide range of size, shape, color, and nucellar embryos develop asexually, with no male flavor may occur in a single progeny from two cells contributing to their formation, they are usu- parents. ally genetically identical with the seed parent. Because of the importance of nucellor embry- Nucellar embryony has been a major obstacle to ony, U.C. researchers began to study its pattern of the systematic production of F1 hybrids and self- inheritance. By 1959, data indicated that its inher- pollinated progenies. Many cultivars produce al- itance was rather simple within common Citrus most entirely nucellar embryos, and there may be species. Crosses between strictly sexual parents several in a single seed. Others form both nucellar regularly yielded hybrids that were themselves and zygotic (sexual) embryos. Studies have identi- sexually reproducing; crosses between sexual and fied additional cultivars that produce only sexual nucellar parents usually produced hybrids of both progeny, and these have been largely used in more types. Two parents that could reproduce by nucel- recent crosses. lar embryony sometimes produced offspring that The occurrence of nucellar embryony enabled were completely sexual. Tetraploid hybrids (hav- Dr. Frost to compare carefully the tree and fruit ing four sets of chromosomes) were also examined, characters of recent nucellar seedling budlines and they showed the same kind of segregation pat- with those of their older parental budlines. The terns. Nucellar embryony thus appears to be dom- nucellar budlines were usually more vigorous, inant over sexual embryony, and very few genes faster growing, more thorny, and slower to fruit may be involved. than parental ones. These juvenile characteristics In crosses between Citrus and its near relative, Poncimcs, the inheritance is somewhat more com- plex. Many F, hybrids were sexually reproducing, even when both parents could reproduce by nucel- lar embryony. Yet backcrosses of these sexual hy- brids to sexually reproducing Citrus in experiment station tests still resulted in progeny that were essentially all sexual, adding to the evidence that strictly sexual plants may be homozygous for this character. Self-incompatibility in citrus Hybridization in citrus can also be complicated by self- and cross-incompatibility, which occurs in a number of cultivars. Self-incompatible forms set Triploid orange (with 3 little or no seed, and often very little fruit, after sets of chromosomes) self-pollination. Climatic factors, however, some- with few seeds (center) times favor fruit set even in the absence of seeds; is from cross of this condition is called parthenocarpy. tetraploid Ruby orange Self-incompatible cultivars may produce many (left)by diploid King fruits and seeds after cross-pollination by particu- tangor (right). lar pollen parents, indicating cross-compatibility. Right: CES-developed Reactions of this sort occur in a number of plant seedless Chandler pum- €amiliesand are considered to be caused by one or melo was grown in isola- more series of self-incompatibility alleles. Studies tion; seedy fruit at left reported by Dr. Soost in 1965 indicated that such received other citrus alleles are present in several citrus taxa, including pollen. 5ome pummelos and their hybrids, certain grape- Fruit hybrids such as Minneola and Orlando, and

CR-4 the Clementine mandarin. In such cases, cultivars from two or three basic histogenic layers, but a that can serve as pollinators are often needed to nucellar embryo stems originally from a single seed produce good crops of fruit. parent cell. The evidence indicates that the pink grapefruits carry genetic factors for color in only The use of polyploidy one or two of their histogenic layers. Thus nucellar Most common citrus is diploid, with two sets of embryos (believed to arise from layer 11) can either chromosomes, but tetraploids often occur spon- lack a color factor or, conversely, can carry it in all taneously among seedling populations. Dr. Frost their tissues. was one of the first to identify and collect these forms, among a series of citrus cultivars. Although Chemical factors as markers not themselves commercially useful, tetraploids Citrus taxonomy has long been controversial, have been important in breeding few-seeded tri- and the origin and relationships of many forms are ploids (with three sets of chromosomes). Fruits uncertain. Intensive studies of a range of chemical with few seeds are desirable for eating out of and physical characters have been directed toward Oroblanco triploid grape hand, but most diploid hybrids are seedy. Triploids solving these problems. Under the leadership of fruit hybrid, developed at can be produced by crossing diploids with tetra- R.W. Scora, U.C. scientists have investigated ter- U.C. in 1980, is nearly ploids, and by 1938 Frost had obtained limited penes of leaf and rind oils, leaf hydrocarbons, pro- seedless and earlier numbers of such hybrids. Recently, large popula- tein and isozyme patterns, oxidase browning, virus ripening than the Marsh tions have been produced and are under study. resistance, and momhological characters. The grapefruit. When a partly sexual tetraploid is used as seed overall results suggest that there are four primary Below: Immature fruit of parent and a diploid as pollen parent, triploids are species, namely, the subtropical Citrus reticulata variegated orange. usually obtained, but the reciprocal cross results in and C. medica, the tropical C. grandis and the yellow rind sectors have few triploids and numerous hybrid tetraploids. recently discovered c. halimii. The latter is some- normal growth; green U.C., Riverside, studies showed that the what intermediate between the genera citrus and sectors are &Dressed. chromosomes occasionally doubled in the diploid Fortunella. Other forms that have often been clas- seed parent, followed by selective survival of the sified as species can, in fact, be identified as mem- resulting tetraploid embryos. An unfavorable bers of complexes involving asexual reproduction. balance between chromosome numbers in the In studies of closely related biotypes, relationships young embryo and endosperm (the nutritive tissue) were found between diploid and tetraploid forms, is responsible for this unusual behavior. However, among identical twin individuals, and among F, such new tetraploid plants, some of which are hybrids of Citrus and Pcm.cirus. strictly sexual, have already been used as addi- U.C. researchers, in cooperation with the Uni- tional seed parents to produce more triploids. versity of Kansas, have recently identified enzyme The triploid embryos that arise from pollination characters that are simply inherited and useful for of tetraploid seed parents by diploids also have a both breeding and taxonomy. The browning reac- certain developmental disadvantage, since they tion in young shoot tissue results from oxidation of sometimes tend to mature earlier and produce phenolic substrates. Presence of the substrate is smaller seeds. They can usually be germinated sat- under single-gene control, which often permits the isfactorily by use of in vitro techniques. separation of nucellar from zygotic seedling plants. Another reaction causes coagulation of shoot Mutations and chimeras tissue in some forms. Coagulation is recessive to Somatic mutations occur frequently in citrus; non-coagulation so that, again, separation of dif- they are often unfavorable, but occasionally highly ferent genetic types is often possible. Analyses of valuable. In Japan, mutations in the satsuma man- several other leaf enzymes, also inherited in darin have resulted in closely related clones differ- single-gene fashion, have been especially reward- ing in characteristics such as color and ripening ing. The presence or absence of as many as eight time. At Riverside, Dr. Frost found that similar of these markers, taken together, can permit early mutations had produced satsuma seedling lines and rapid separation of nucellar from zygotic that varied in ripening time, productivity, and plants in a great many crosses, and can assist in fruit shape. One of these nucellar lines has been determining parentage of some long-established used commercially as an improved cultivar. cultivars. Mutations that arise in single somatic cells are sometimes perpetuated in only a part of the tissues New citrus cultivars of a plant or animal, resulting in a chimera. In Efforts to produce early-maturing citrus fruits of 1942, chimeral citrus trees were identified at moderate acidity have been aided by the use of a Riverside in which certain histogenic, or tissue- clone of pummelo (Citrms grandis) that has essen- producing, cell layers were diploid, whereas others tially no titratable acid. When U.C. plant breeders had become tetraploid. This caused differences in crossed this pummelo with cultivars of medium the height, shape, and leaf texture of these plants. acidity, the resulting hybrids had acid levels aver- We have studied an interesting case of chimer- aging well within the edible range at early mid- ism among pigmented grapefruits, in cooperation season. Compared hybrids with a medium- or high- with the U.S. Department of Agriculture. It is acid pummelo as a parent were usually too sour. known that the Thompson and Foster pink grape- One of the acidless pummelo hybrids, when selfed, fruits arose as limb sports on white grapefruit produced acidless individuals in proportions that trees early in this century. When nucellar seedling suggest single-gene action. These individuals may trees of these cultivars were grown in about 1960, serve as new acidless parents. those of Thompson had no pink in their fruits, whereas those of Foster had deeper color than their parent. Various tissues in many plants arise

CR-5 Two U.C.-developed hybrids in- two citremons showed certain good characteristics volving the acidless pummelo with lemons, but overall they have not been suffi- have been introduced: the ciently promising. The citrumelo behaved well in Chandler pummelo (1961) and some trials with orange, but showed severe prob- the Oroblanco grapefruit hybrid lems in a Eureka lemon trial. (1980). Both are of good flavor Identifying satisfactory rootstocks for lemons and are usually edible by De- has been especially difficult. Sweet orange is fav- cember in southern California orable but should not be used where gummosis is a plantings. The Oroblanco has problem. Eureka lemons are incompatible with very few seeds, because it is a Troyer, as are some semidense Lisbons. Cleopatra triploid. Studies of seed number mandarin has been perhaps the most useful of the in the Chandler indicate that it mandarin stocks. Several recent trials with Citrus is self-incompatible and, when macrophylla have shown that it gives high yields planted in isolation, can pro- with small tree size, but phloem necrosis can be a duce good yields of seedless limiting factor. fruit. As with other such New crosses between Pmirus and Citrus to Two germinating seeds on right show polyembryony cultivars, this ~h~racteristic obtain rootstocks have been made at Riverside. (production of more than one embryo). Others are varies With climatic conditions Successful rootstock cultivars must be highly monoembryonic. and by years. Fruits with few nucellar in their reproduction, and hybrids with seeds can sometimes be smaller sweet orange have appeared promising in this in size. respect. Some of them show resistance to Phy- The first new citrus cultivars introduced at Riv- tophthwu, tristeza, and nematodes, but their ef- erside (in 1935) were those produced by Dr. Frost. fect on yield and fruit quality is uncertain. One of They were the Kara, Kinnow, and Wilking man- the better ones, unfortunately, has a low yield of darins and the Trovita orange. The Kinnow has at- fruit and seeds. tained considerable importance in Pakistan and in parts of North Africa; it continues to be grown on Trifoliolate leaf of Poncirus a small scale in California. The Wilking, usually Leaves of Poncirus are typically composed of too small for general use, has been a parent of three leaflets, in contrast to the simple (monofoli- several mandarin hybrids with excellent eating olate) leaves of common citrus. It has been of in- quality. In Japan the Trovita, which is an early, terest to examine the inheritance of this trifoliolate few-seeded juice orange, has been used in crosses leaf character. Nearly all of the older, long- with satsuma to produce a new tangor, the established hybrids of Citrus with Pmirus were Kiyomi. known to have high proportions of three-parted leaves. A new U.C. study showed that among 272 Rootstocks and rootstock breeding F1 individuals, all but one had significant pro- Rootstocks are of major importance to citrus cul- portions of multifoliolate leaves, but the percen- ture in most countries, because they help to im- tages varied from 30 percent in some hybrids to part disease resistance, cold hardiness, high yield nearly 100 percent in others. When these hybrids and fruit quality, and other favorable character- were backcrossed to Citrus, most of the resulting istics to their scions. In California, clones of the plants had very few or no multifoliolate leaves, sweet and sour orange were the principal stocks yet some showed proportions up to 90 percent. for many years, until the damaging effects of The conclusion was that the three-leaflet charac- Phytophthoru fungi and the tristeza virus made ter is not completely dominant, but that it is rather them unacceptable in many areas. simply inherited, probably dependent on only a When tristeza began to destroy orchards of few genes. sweet orange on sour orange rootstock, use of a substitute rootstock became imperative. As part of Future prospects comprehensive studies of Citrus and Citrus rela- Advances in breeding and genetics by conven- tives, conducted since 1946 at U.C., Riverside by tional methods will continue to be slow because of W.P. Bitters, many taxa were tested for tristeza quantitative inheritance of most characters, long resistance. The Troyer citrange was outstanding, generation time, and nucellar embryony. However, because sweet orange scions grown on this root- breeding efforts can be directed at more specific stock were tolerant to the disease, and also had goals than in the past because of new information other favorable qualities. Troyer continues to be and plant materials. The range of strictly sexual the most important rootstock for oranges in Cali- parents has been increased, more tetraploids are fornia. A genetically similar cultivar, the Carrizo, available for the production of triploids, and hy- has also been successful for that purpose. brids with improved horticultural characters have One parent of Troyer is Poncirus- the trifoliate been developed. orange- which as a rootstbck also imparts toler- Identification and recovery of sexual seedlings ance to tristeza. Clones of this cultivar have im- should continue to improve with the use of avail- portant rootstock value, but under more limited able biochemical and in vitro techniques. How- conditions than Troyer. Studies were made on ever, progress in in vitro methods of hybridization other old, established hybrids of Poncirus, includ- has been slow, although fusion of somatic proto- ing citremons 1448 and 1449 (hybrids with lemon) plasts (cells with cell walls removed) has been and citrumelo 1452 (a hybrid with grapefruit). The achieved. Research on cell selection has been initi- ated. Continued research should provide a choice of methods for transferring and recombining a wider range of genetic material. CR-6