Chapter 9: Crop Plants and Exotic Crops

Chapter 9: Crop Plants and Exotic Plants

ANISE
Pimpinella anisum L., family Umbelliferae
Anise is a minor crop cultivated for its seeds or the oil pressed from them which are used in condiments, beverages, medicines, soaps, and perfumery. Probably no more than a few hundred acres are grown in the United States.

Plant:
Anise is a pubescent annual about 2 feet high (fig. 36). It is usually planted in 15- to 30-inch rows, the plants thinned to 6-inch intervals. Planting is usually in the springtime in temperate climates, although Griffiths et al. (1946*) recommended that it be planted in September in Arizona.

Inflorescence:
The small but copious yellowish-white flowers are in large loose umbels and appear in late summer, about 3 months after planting if the seeds were planted in the spring. Pellett (1947*) reported that the blossoms, which are attractive to honey bees, yield a light-colored honey with a mild elusive flavor. Sievers (1948) stated that 400 to 600 pounds of seed per acre was an annual yield, Rosengarten (1969*) mentioned 500 to 800 pounds. Griffiths et al. (1946*) stated that 600 pounds per acre could be expected in Arizona. The influence of insect pollination on seed production was apparently not considered in relation to these yields.

[gfx] FIGURE 36. - Anise plant in full bloom.

Pollination Requirements:
No reference could be found on the relation of pollinating agents to production of anise, although its flower structure and family relationship would indicate that it benefits from, if it is not entirely dependent upon, insect pollination. Hawthorn and Pollard (1954*) support this by stating that insufficient insect pollination frequently results in reduced yields of some crops, including anise. Growers interested in obtaining the highest possible yields of anise should, therefore, give consideration to its insect pollination.

Pollinators:
Although there seems to be little information on the insect pollination of anise, the honey bee could probably pollinate it adequately, considering the flower's structure and its reported attraction to honey bees.

Pollination Recommendations and Practices:
None.

LITERATURE CITED:
SIEVER, A. F.
1948. PRODUCTION OF DRUG AND CONDIMENT PLANTS U.S. Dept. Agr. Farmers' Bul. 1999, 99 pp.

BLACK PEPPER AND WHITE PEPPER¹⁷Piper nigrum L., family Piperaceae
The plant that yields ground pepper is not grown commercially in the United States, but it is an important one worldwide. In terms of usage and value, pepper is the most important of all spices in world trade. The United States imports 35 to 40 million pounds annually. India and Indonesia account for about two-thirds of the world production. Some pepper is produced in Brazil for 10 days but that it is at its peak of receptivity after (Kevorkian 1964). It has been grown experimentally 3 to 4 days. under glass in Maryland (Creech 1955).
__________
¹⁷ See "Pepper, Green," p. 292.

Plant:
Piper nigrum is a strong, somewhat woody, perennial evergreen vine that may climb to 30 feet in its preferred hot, wet, nonseasonal climates. Under cultivation, growth is usually held to 10 to 15 feet. The plant has oval, dark-green leaves, as much as 7 by 4 inches in size, that arise at the nodes. Hardwood posts or trees provide columnar support for the vines that may reach 5 feet in width (Purseglove 1968 *). There are many cultivars.

The fruit, called a corn, is 1/4 to l/5 inch in diameter and is picked just before it ripens. The corns are separated from the stems, then dried in a manner similar to coffee drying. Within 2 to 3 days, the pericarp turns black. If the corns are ground while in the pericarp, the product is black pepper. If the black pericarp is removed before the fruit is ground, the product is white pepper (Blacklock 1954, Gentry 1955).

Inflorescence:
The flowers are borne on the vine, at the node opposite the leaves, in catkins or spikes. A spike may have 50 to 150 rather inconspicuous yellowish green apetalous florets only 1 to 3 mm in diameter. The florets are usually hermaphrodite but may be unisexual, with staminate and pistillate flowers on the same plant or on separate plants. Frequently, the florets are unisexual near the base of the spike and hermaphrodite toward the tip. Flowering begins at the base and continues to the tip over a 7- to 8-day period (Ridley 1912* Gentry 1955)

The hermaphrodite floret is protogynous, the two to three stamens appearing at the base of the ovary only after the star-shaped stigma with its three to five rays has matured (Cobley 1956*). The stigma may be receptive for 10 days with peak receptivity at 3 to 5 days (Purseglove 1968*). The pollen is then released in gelatinous masses to pollinate receptive stigmas of other flowers. The unilocular ovary produces only one seed. The stigmatic rays are coated with long tubular hairy growths with their tips somewhat bulbous. The feltlike surface acts as a medium for trapping the pollen grains (Anadan 1924).

Pollination Requirements:
Because of the protogynous nature of Piper nigrum, self-pollination of the floret is impossible. Cobley (1956*) stated that cross-fertilization was the rule, but he apparently referred to transfer of pollen between flowers on a plant rather than between plants. Martin and Gregory (1962) concluded that self-pollination between flowers on a plant was undoubtedly the rule. Free (1970*) stated that the stigma may be receptive for 10 days but that it is at its peak of receptivity after 3 to 4 days.

Pollinators:
Anadan (1924) and Menon (1949) considered rain as the pollinating agent of Piper nigrum. This was supported by the observation by Anadan (1924) that a vine protected from rain failed to set fruit. Martin and Gregory (1962) stated that wind pollination, with or without rain, was not very effective. They believed that self-pollination was undoubtedly the rule, but they did not explain how the pollen might have been transferred from the anthers of one flower to receptive stigmas of another. Cobley (1956*) attributed the transfer of pollen to wind, rain, and ants. Free (1970*) stated that pollination was the result of gravity possibly aided by rain or wind. Purseglove (1968*) stated that although the pollen was in gelatinous masses, a light rain would break up these masses, then the pollen grains would be dispersed and finally caught in the papillae of the stigma. He concluded that the degree to which insects assist in pollination is not known. Martin and Gregory (1962) stated that no insects, large or small, visited the spikes. Anadan (1924), as previously mentioned, stated that a vine protected from rain failed to set, even with bees. He did not elaborate on the kind or activity of the bees. No other observer mentioned visitation of the flowers by bees. It is not clear, therefore, the degree to which insects pollinate Piper nigrum.

Pollination Recommendations and Practices:
None.

LITERATURE CITED:
ANADAN, N.
1924. OBSERVATIONS ON THE HABITS OF THE PEPPER VINE WITH SPECIAL REFERENCE TO THE REPRODUCTIVE PHASE. Madras Dept. Agr. Yearbook 1924: 49-69.

BLACKLOCK, J. S.
1954. A SHORT STUDY OF PEPPER CULTURE WITH SPECIAL REFERENCE TO SARAWAK. Trop. Agr. [Trinidad] 31: 40-56.

CREECH, J. L.
1955. PROPAGATION OF BLACK PEPPERS. Econ. Bot. 9: 233-242.

GENTRY, H. S.
1955. INTRODUCING BLACK PEPPER INTO AMERICA. Econ. Bot. 9: 256-268.

KEVORKIAN, A. G.
1964. PEPPER, VANILLA, AND OTHER SPICES. U.S. Dept. Agr. Yearbook 1964: 195-200.

MARTIN, F. W., and GREGORY, L. E.
1962. MODE OF POLLINATION AND FACTORS AFFECTING FRUIT SET IN PIPER NIGRUM L., IN PUERTO RICO. Crop Sci. 2: 295--299.

MENON, K. K.
1949. THE SURVEY OF POLLU AND ROOT DISEASES OF PEPPER. Indian Jour. Agr. Sci. 19: 89-136.

BUCKWHEAT ¹⁹Fagopyrum esculentum Moench, family Polygonaceae
Buckwheat, once a highly important crop in the United States, appears to be in the twilight of its day. In 1918, more than a million acres were grown (Quisenberry and Taylor 1939). Over half of that acreage was in Pennsylvania and New York. Twenty years later the total acreage was less than one-half million. By 1954, only 150,000 acres were harvested, and in 1964 when USDA crop production records for buckwheat were discontinued, only 50,000 acres were harvested. Seed production ranged from 500 to 1,700 lb/acre, depending on various cultural factors, not the least of which was completeness of bee pollination (Carmany 1926, Kopel'skievsky l960, Martin and Leonard 1949*). However, Root (1891) reported a phenomenal yield of 3,840 lb/acre in one instance. The limited acreage of buckwheat in the United States is in the Great Lakes region and eastward. In 1970, Russia led all other countries in buckwheat production with more than 4.5 million acres (United Nations Food and Agriculture Organization (FAO) 1971, p. 80). Buckwheat is grown primarily for the seeds, which are ground into flour and used in buckwheat cakes.
__________
¹⁹Tartary buckwheat, Fagopyrun tarticum (L.) Gaertn., is a more slender plant than F. esculentum, with smaller greenish or yellowish flowers and not as aggregated. According to Naghski (1951), extensive plantings of tartary buckwheat have been made in recent years because it is a good source of rutin, a drug used to reuce capillary blood pressure and relieve atomic radiation injury. This species is seldom visited by bees and is self-fertile ( Garber and Quisenberry 1927).

Plant:
Buckwheat is an annual, 2 to 4 feet tall, with a single stem and several branches bearing heart-shaped leaves (fig. 55). The green to red stem turns brown with age. More reddening is evident with poor seed set. The l/5-inch flowers are in clusters mostly at or near the top of the plant (fig. 56). Flowering is indeterminate, and the plants which are usually broadcast are often harvested with some immature seeds and even flowers present. A field in flower is chalky white and has a pronounced aroma that some people consider unpleasant. Flowering in a field may begin 5 to 6 weeks after planting and may continue for 25 to 30 days (fig. 57).

[gfx] FIGURE 55. - buckwheat plant in bloom.
FIGURE 56. - buckwheat flowering branch.
FIGURE 57. - field of buckwheat in full bloom.

Inflorescence:
The buckwheat flower has no petals - the sexual parts, the ovary, three styles and eight stamens being enclosed in the petallike sepals. Four of the anthers bend out but turn their pollen inward. The other four turn their pollen outward (Knuth 1909*, pp. 341 - 342). Some plants have flowers, referred to as the "thrum" type, with short styles and long filaments so the stamens extend above the styles. Other flowers, referred to as "pin" types, have long styles and short filaments so the stigma is above the anthers. Occasionally, the styles and stamens are at the same height. The long stamens and filaments are fully 3 mm; the shorter ones, about 2 mm. Although each plant bears flowers of only one form, the seeds from either form will produce plants having the dimorphic forms in about equal numbers. The three styles lead to a single ovary with one ovule, so a flower can produce only one seed, which is about one-quarter inch long.

The flower, which opens in the morning around 8 a.m., has eight yellow nectaries alternating with the eight filaments at the base of the ovary, bound together by a cushionlike swelling (Knuth 1909*, pp. 341 - 342). The flower (fig. 56) secretes nectar in copious amounts, but only in the morning hours, during which time it is highly attractive to bees (Phillips and Demuth 1922). Toward noon, the flow lessens, and during the afternoon honey bees usually abandon the plants. Pollen is also collected by honey bees from buckwheat.

A colony of honey bees having access to a field of flowering buckwheat may store 10 to 15 pounds of honey per day (Versehora 1962), and collect 90 to 290 pounds of nectar per acre (Free 1970* Martin and Leonard 1949*). The honey produced by buckwheat is dark with a strong flavor that is usually relished only by people who are accustomed to it; however, there is a greater demand for this honey than can be supplied. The honey is used primarily in the baking of foods. During a buckwheat nectar flow, the apiary may have a strong sometimes nauseating aroma which can be detected for some distance (Pellett 1947*). Mel'nichenko (1963) thought that removal of nectar by bees stimulated greater secretion. He stated that secretion ceases after the flower has been fertilized.

Bukhareva (1964) and Leshchev (1962) reported that some trace elements caused an increase in buckwheat nectar secretion and seed yields. This was supported by, Kopel'skievsky (195S, 1960), Leshchev (1962), anc Skrebtsova (1957) who found that the fertilizers calcium, nitrogen, and phosphorus increased the pollination effectiveness of honey bees. Demianowicz and Ruszkowska (1959) found that all the cultivars tested were important sources of pollen, but some were much better sources of nectar than others. With many nectar-producing plants decreasing or disappearing from a beekeeper's area he might encourage buckwheat planting nearby to supplement his bee forage.

Pollination Requirements:
The buckwheat flower is usually unable to self-pollinate. The flower type prevents the pollen from automatically coming in contact with the stigma. Exceptions include the occasional flowers with pistil and stamens of the same length, which usually have a low degree of self-fertility. A recent selection (F. sagittatum Gilib.) has been developed which has stamens and pistil at the same level, with a high degree of self-fertility, but of no direct commercial value (Marshall 1970). Buckwheat pollen is not windblown, therefore insects are necessary for the transfer of the pollen. Davydova (1954) found that, as is customary for dimorphic flowers, the pollen grains on the two types of stamens are different in size, the flowers with longer stamens having larger grains (46 to 67 by 39 to 55 microns, versus 35 to 44 by 29 to 40 microns for grains on the shorter stamens). The analyses by Davydova (1954) of pellets of pollen taken from honey bees working buckwheat, showing that both types of pollen were present, was confirmed by Roz[s]ov and Sc[k]rebtsova (1958). This proved that the bees move freely from one type of flower to another and are thus effective pollinating agents of this crop.

The necessity of insect pollination for commercial seed production of buckwheat has been well established by Garber and Quizenberry (1927) in the United States and numerous workers in Russia, where this crop is grown so extensively (Elagin 1953, Glukhov 1955, Kashkovskii 1958, Mel'nichenko 1962, and Sevcuk 1946). Free (1970*), after reviewing the pollination of buckwheat, pointed out the need for some controlled cage tests on this crop to determine the degree of self-pollination if any occurred and the quantity of seed that might be expected under different pollination conditions.

Pollinators:
Unquestionably, the honey bee is the best pollinator of buckwheat because it is highly attracted to the buckwheat flower and efficiently and effectively transfers the pollen from anthers to stigmas, whether collecting pollen or nectar.

Leighty (1919) stated that many buckwheat growers believed that the weight per bushel of seed was heavier where the crop had been worked heavily by bees. Elagin (1953) showed the following correlation between the 2-year average yield of buckwheat seeds and distance in meters from the apiary. Distance from apiary

Distance from apiary in meters Yield of buckwheat seed in meters in kilograms per hectare ÒNearÓ 850 500 770 1,000 720 1,500 575

The number of colonies in relation to the area of buckwheat was not given by Elagin (1953), although where five colonies per hectare were present, 80.4 percent of the seeds set, but with only one colony per hectare, the set was only 57.8 percent. The transportation of colonies to the buckwheat fields was encouraged because of their value as pollinators.

Glukhov (1955) obtained 1,700 kg buckwheat seed per hectare within 500 m of the apiary, but production dropped to 1,200 kg in the 500- to 1,000-m range, and only 500 kg/ha at 2,000 to 3,000 m from the apiary. In another field, he obtained 2,500 kg/ha of seed adjacent to the apiary, 1,900 kg at 500 m and 1,300 kg/ha 1,000 m from the apiary. Similarly, Kopel'skievsky (1960) obtained 1,470 kg/ha seed adjacent to an apiary, but only 840 kg/ha 2,000 m away.

Pollination Recommendations and Practices:
There are no recommendations in this country in relation to bee populations and buckwheat seed production. In Canada, one colony per acre is recommended (Smith et al. 1971). In Russia, Kashkovskii (1958) stated that there should be enough bees for each flower to receive five or six visits. Mel'nichenko (1962) stated that about two colonies per acre were needed for saturation pollination and highest buckwheat yields; however, when he used about three colonies per acre, he obtained 1,250 to 1,500 lb per acre. Doubtless, the colonies per acre necessary to supply the five to six visits per flower varies with location and conditions.

LITERATURE CITED:
BUKHAREVA, G. A.
1964. [EFFECT OF TRACE ELEMENTS USED IN PRE-SOWING TREATMENT, AND OF FOLIAR FERTILIZERS, ON NECTAR SECRETION IN SOME AGRICULTURAL CROPS.] Trud. Nauch-Issled. Inst. Pchelovodstvo 140-194. [In Russian, English summary.] AA-769/65.

CARMANY, C. E.
1926. BUCKWHEAT IN MICHIGAN. Mich. Agr. Expt. Sta. Spec. Bul. 151, 11 pp.

DAVYDOVA. N. S.
1954. [ANALYSIS OF HONEY BEE POLLEN LOADS FROM BUCKWHEAT.] Uchen. Zap. Kishinev Gos. Univ. 13: 167-173. [In Russian.] AA-104/64.

DEMIANOWICZ, Z., and RUSZKOWSKA, B.
1959. [THE NECTAR FLOW FROM BUCKWHEAT.] Pszczel. Zesz. Nauk. 3(1): 11 - 24. [In Polish.] AA-342/64.

ELAGIN, I.
1953. [INFLUENCE OF POLLINATION BY BEES ON THE YIELD FROM BUCKWHEAT.] Pchelovodstvo 6: 31-33. [In Russian.] AA-117/54.

GARBER, R. J., and QUISENBERRY, K. S.
1927. SELF-FERTILIZATION IN BUCKWHEAT. Jour. Agr. Res. 34: 185 - 190.

GLUKHOV, M. M.
1955. [HONEY PLANTS.] 512 pp. lzd. 6, Perer. i Dop. Moskva, Gos. Izd-vo Selkhoz Lit-ry. [In Russian.]

KASHKOVSKII
1958. [ECONOMIC RESULTS OF POLLINATING BUCKWHEAT CROPS BY HONEYBEES AND BY HAND.] Byull. Nauch-Tekh. Inf., Kemerovo 2-59-61. [In Russian.] AA-390/60.

KOPEL'KIEVSKY, G. V.
1955. [POLLINATION OF BUCKWHEAT BY BEES.] Pchelovodstvo 32: 41 - 48. [In Russian.] AA-55/57.

______ 1960. [BEES AND BUCKWHEAT SEED CROP.] Pchelovodstvo 37(4): 36 - 39. [ In Russian.] AA-683/63.

LEIGHTY, C. E.
1919. BUCKWHEAT. U.S. Dept. Agr. Farmers' Bul. 1062, 24 pp.

LESHCHEV, V.
1952. [MODERN AGRICULTURE INCREASES NECTAR PRODUCTION OF BUCKWHEAT.] Pchelovodstvo 29: 23 - 26. [In Russian.] AA-54l57.

MARSHALL, H. G.
1970. REGISTRATION OF ÔPENNLINE 10Õ BUCKWHEAT. Crop Sci. 10: 726.

MEL'NICHENKO, A. N.
1962. [BIOLOGICAL BASIS FOR INCREASING THE YIELD OF BUCKWHEAT BY DIFFERENT SOWING DATES AND DEGREES OF SATURATION OF BEE POLLINATION. ] Uchen. Zap. Gor'kov. Univ. 55: 5-43. [In Russian.] AA-667/64.

MEL'NICHENKO. A. N.
1963. [BEES THEMSELVES INCREASE THE NECTAR PRODUCTIVITY OF FLOWERS.] Pchelovodstvo 49(9): 32 - 35. [In Russian. ] AA-566/65.

NAGHSKI, J.
1951. NO HONEY FROM TARTARY BUCKWHEAT. Amer. Bee Jour. 91: 513.

PHILLIPS, E. F., and DEMUTH, G. S.
1922. BEEKEEPING IN THE BUCKWHEAT REGION. U.S. Dept. Agr. Farmers' Bul. 1216, 26 pp.

QUISENBERRY. K. S., and TAYLOR. J. W.
1939. GROWING BUCKWHEAT. U.S. Dept. Agr. Farmers' Bul. 1835, 17 pp.

ROOT, A. I.
1891. THE ABC OF BEE CULTURE. 408 pp. A. I. Root Co., Medina, Ohio.

SEVCUK, I.
1946. [PROSPECTS FOR THE DEVELOPMENT OF APICULTURE IN THE UKRAINE.] Soc. Seljs. Hoz. 12: 32-37. Herbage Abs. 18(3): No. 695.

SKREBTSOVA, N. D.
1957. [POLLINATION OF BUCKWHEAT FLOWERS BY BEES.] Pchelovodstvo 34(9): 48 - 50. [In Russian.] AA-135/59.

SMITH, H., PANKIW, P., KREUTZER, G., and others.
1971. HONEY BEE POLLINATION IN MANITOBA. Manitoba Dept. Agr. Pub. 525, 16 pp.

UNITED NATIONS FOOD AND AGRICULTURE ORGANIZATION (FAO)
1971. 1971 yearbook. V. 25, 829 pp.

VERSEHORA, D.
1962. [BUCKWHEAT AND ITS IMPORTANCE FOR APICULTURE.] Apicultura 15(6): 11-14. Bucharest. [In Romanian.] AA-765/65.

CARAMBOLA
Averrhoa carambola L., family Oxalidaceae
Carambola is a crop of no importance in continental United States. Knight (1965) stated that there was one commercial planting in production in Sarasota County, Fla., but otherwise production was limited to dooryard plants of southern Florida.

Plant:
The carambola is a many-branched, frost-susceptible evergreen tree to 30 feet (Bailey 1949*). It is grown for its waxy-yellow, 2- to 5-inch long fruit, which are characterized by four or five sharp ribs. When the fruit is sectioned, the star-shaped pieces are used ornamentally in salads and punch bowls. The juice is rich in vitamin C. The fruit may also be stewed, preserved, or made into jams and jellies. A fruit has from 1 to 15 pendulous seeds in each rib or segment, depending upon the completeness of pollination.

Inflorescence:
The clusters of fragrant whitish to rose-colored flowers are borne in the leaf axils. They are about five-eighths inch across. There are five petals and 10 stamens in at least two whorls alternating long and short, five without anthers. The ovary has four or five cells with two to four ovules per cell (Ochse et al. 1961*, Purseglove 1968*).

Pollination Requirements:
The flowers are self-incompatible, and not wind pollinated; therefore, insects are necessary in the production of fruit (Knight 1965). Honey bees visit the flowers freely. Nand (1971 ) stated that honey bees, flies, and other insects are the chief pollinating agents of this completely cross-pollinated plant.

Pollination Recommendations and Practices:
None.

LITERATURE CITED:
KNIGHT' R. J. JR.
1965. HETEROSTYLY AND POLLINATION IN CARAMBOLA. Fla. State Hort. Soc. Proc. 76: 375 - 378.

NAND, D.
1971. POLLINATION, FRUIT SET, AND DEVELOPMENT IN CARAMBOLA (AVERRHOA CARAMBOLA LINN.). Indian Jour. Hort. 28: 278-284.

CARAWAY
Carum carvi L., family Umbelliferae
Caraway is grown almost exclusively in The Netherlands on about 10,000 acres for its seeds, which are used to season breads, meats, cheeses, and drinks (van Roon and Bleijenberg 1964). A small acreage is grown in the northern and northwest part of the United States (Rosengarten 1969*).

Plant:
The plant is a slender, smooth, erect, annual, or biennial herb, 1 to 3 feet tall, with thick tuberous roots and narrow leaves (fig. 61). It is planted in the spring (fall of the mild-wintered Southwest) in 15- to 30- inch rows, and may produce 800 to 2,000 pounds of seed per acre (Hawthorn and Pollard 1954*, Rosengarten 1 969*).

[gfx]
FIGURE 61. - Caraway leaves and flowering stems.

Inflorescence:
The yellowish-white flower is markedly protandrous. The stamens release pollen during the first 2 days the flower is open, then wither on the third day. The stigma does not become receptive until the sixth to seventh day. The primary umber is usually in the female stage of flowering when the lateral umbels are in the male stage. Self- fertilization in a flower, and usually within an umber, does not occur (van Roon and Bleijenberg 1964, Knuth 1 908*, P. 4 77). Both nectar and pollen are easily available and attractive to flies and hymenopterous insects.

Pollination Requirements:
Because of the protandry, pollen must be transferred from pollen- producing flowers to receptive stigmas. The pollen is not windblown but must be transferred by insects.

Pollinators:
Bees are the primary pollinators of caraway flowers.

Pollination Recommendations and Practices:
Although there are no recommendations on the pollination of caraway, the flower type and the need for pollinating insects would indicate that where maximum commercial production of seed is desired the grower should provide an ample supply of bees to the field.

LITERATURE CITED:
ROON, E. VAN, and BLEIJENBERG, H. J.
1964. BREEDING CARAWAY FOR NON-SHATTERING SEED. Euphytica 13: 281-293.

CARDAMOM
Elettaria cardamomum (L.) Maton, family Zingiberaceae
Most of the cardamoms are produced in southeast Asia and the southern tip of India. The cardamoms of commerce are the seeds, which are used as spices in seasoning and in medicine (Bailey 1949*, Ridley 1912*).

Plant:
The plant, a shrub, forms a clump 7 to 9 feet tall, that is more or less cultivated in the jungle areas. It requires some shade from taller plants. It produces rhizomes, by which it is propagated, although seeds are also planted at the rate of 650 to 1,000 per acre. The plant will flower 2 years after it is planted and will yield for about 15 years. About 5 months after the flower opens, a three-celled pod is harvested. Each cell of this pod produces seven to nine dark-brown aromatic seeds, the cardamoms or cardamons.

Inflorescence:
The slender flowering stems arise 2 to 3 feet from the rootstock or rhizomes, and produce toward the apex numerous florets in two- to three- flowered racemes. The green calyx tube is 1 1/2 inches long, and the pale green 1/2-inch lobes are narrow and spreading. The flowers open singly or in two or more at a time over a long period.

Pollination Requirements:
Ridley (1912*) stated that the flowers require insect pollination. He concluded that the lack of adequate pollination often contributed to reduced crops.

Pollinators:
Ridley (1912*) stated that the flowers are pollinated by insects, probably some species of bees or a fly.

Pollination Recommendations and Practices:
None.

CARDOON
(See "Artichoke")
CHERIMOYA
Annona cherimola Mill., family Annonaceae ²⁰
The cherimoya occurs naturally in the Andean valleys of Ecuador and Peru (Purseglove 1968*), but it has been carried to various other subtropical areas of the world where it has become very popular. It is grown to a limited extent in Hawaii and Florida, with an estimated 50 to 60 acres in California (Schroeder 1948, 1956). Sarasola (1960) stated that about 300 acres are grown in the area of Almunecar, Spain. There are numerous cultivars (Anonymous 1956, Brooks and Olmo 1952). The hybrid of A. cherimola X A. squamosa L., termed "atemoya," has also been cultivated (Ahmed 1936, Thakur and Singh 1965).
__________
²⁰ Other species of Annona that are cultivated or sometimes cultivated include: A. diversifolia Saff., the Ilana; A. montana Macfad., mountain soursop; A. muricata L., soursop or guanababa; A. squamosa L., sugarapple or sweetsop; and A. reticulata L., custard-apple or bullocks-heart. Unlike A. cherimola, none have floral fragrance, but otherwise their pollination requirements may be similar. They are mentioned here because of their popularity in Asia and the tropics and their potential value in ou subtropical areas (R. J. Kinght, Jr., personal commun., 1971).

Plant:
The cherimoya tree may reach 30 feet tall but is usually much smaller, somewhat spreading or scraggly, and semideciduous. It sheds its leaves in the spring just before the flowers appear (Schroeder 1941). It will grow anywhere the avocado will grow. The 'Bays' cv. is the most satisfactory in California (Schroeder 1956). The plants are usually set about 12 feet apart each way in the field (Ahmed 1936).

Cherimoyas are grown for the fruit, 3 to 6 inches in diameter and l/2 to 2 l/2 pounds in weight, which ripens 5 to 8 months after pollination. The thick brown, green, or gray-green skin has the appearance of rough leather. The fruit can be broken apart easily and the delicious white, sweet (18 percent sugar) pulp eaten out of hand with a spoon. It tastes somewhat like banana or pineapple custard. There may be 20 to 80 black or mahogany-colored oval seed l/2 to 3/4-inch in diameter that separate quite easily from among the delicious pulp. The fruit is seldom good more than 7 or 8 days after harvest (Sarasola 1960).

Inflorescence:
The rather primitive but scented cherimoya flower may be solitary, or there may be two or three in a cluster on a short peduncle (fig. 70). There are three light-green, fleshy petals about an inch long. Almost hidden at the base of the petals are the numerous but practically filamentless stamens, surrounding but just below the stigma, the whole androecium resembling the immature strawberry fruit. There is no nectary.

When the petals open, towards midday, the stigma is covered with a viscid material and is receptive to fertilization. Receptivity may last 2 to 6 hours depending upon temperature and humidity (Schroeder 1971). When receptivity ceases, the stigma dries and turns brown. Later in the day, or more frequently the following morning, the stamens shed their pollen (Brooks and Hesse 1953, Schroeder 1941, Watts 1942). If fertilization occurs, the petals drop within about 24 hours and fruit development proceeds. If fertilization is prevented, the entire inflorescence dries and drops within 4 days.

[gfx] FIGURE.- Longitudinal section of cherimoya flower, x 5, with detail showing an additional stamen and pistil, greatly enlarged.

Pollination Requirements:
The maturation of the pistil before pollen is available creates a pollination problem and prevents ample fruit set on at least some cherimoya cultivars (Watts 1942). Schroeder (1941) established that the flowers are self-fertile but usually not capable of self-pollination. When he hand- pollinated flowers, 70 percent set perfect fruit, 17 percent were misshapen, and 13 percent were "runts." In open-pollinated flowers, only 10 percent produced perfect fruit, 39 percent were misshapen, and 51 percent were runts. The hand-pollinated fruits weighed an average of 461 g, whereas the open-pollinated fruits weighed an average of only 261 g. Thakur and Singh ( 1965) also reported 44 to 60 percent set of hand- pollinated flowers as compared to less than 6 percent of open-pollinated flowers. No explanation was given for the fruit set that occurred in the open-pollinated flowers, but Brooks and Olmo (1952) stated that at least the 'Carter' and 'McPherson' cvs. set well without hand pollination.

Thomson (1970) stated that near the ocean the stigma stays receptive longer and selfing is normal. Clark (1925) also reported that heavy crops result without hand pollination although he admitted that he had never seen pollen-bearing insects visit the flowers, and he thought that self-pollination occurred. Krishnamurthi and Madhava Rao (1963) stated that comprehensive studies on pollination of the annonas are needed, with which there seems to be no disagreement.

Pollinators:
In general, cherimoya flowers have been considered incapable of self-pollination and unattractive to pollinating insects. Ahmed (1936) reported that the flowers do not attract bees but stated: "Insects of the lady-bird type such as Coccinella sp. and Scymnus sp. have been observed to visit the flowers either in search for one of the preys such as aphis, or mealy-bugs or feeding on the pollen-grains.... Ants may also be responsible to a smaller extent." He also stated, "Under normal conditions, as in the home-forests of anonas the insect agency is sufficient. But under cultivation, it has been noticed that such agents may be very scarce or absent altogether, thus inducing very low fruit-setting or none." Wester (1910) concluded that nonproductiveness of cherimoyas was due to the scarcity of pollinating insects, but Schroeder (1971) stated that insects visited the flowers upon occasions. Sarasola (1960) doubted that the good fruit set in Spain was the result of special pollinating insect activity but resulted from self-pollination, although he offered no proof for this assumption.

Usually, growers collect pollen by hand from dehiscing anthers, then pollinate stigmas the following day. Ahmed (1936) showed that a man and a little boy working full time daily throughout the 6-week flowering season could pollinate 1 acre. There seems to be no question that hand- pollinated flowers produce more and better fruit than is obtained under natural conditions, but no attempt has been made to influence the supply of pollinators available to the flowers.

Pollination Recommendations and Practices:
Cherimoya growers either collect pollen and hand-pollinate the flowers or leave the plant to chance pollination and the possibility of little or no set of high-quality fruit.

Considering the high cost of hand pollination, the fact that insects visit the flowers only occasionally, and that fruit set occurs in the natural home of cherimoya, other steps should be taken to improve the production and decrease costs. A search might be made for the particular species of insects responsible for the pollination of the plants in their native habitat. Attention might be given to selecting self-fertile cultivars. An immediate step might be to supply "saturation pollination" with honey bees. This has proven feasible on some other crops. Bee visitation should be aufficient to get an ample supply of pollen to all parts of the stigma at the earliest possible moment after it becomes receptive.

LITERATURE CITED:
ANONYMUS
1956. CHECKLIST OF VARIETIES OF CERTAIN SUBTROPICAL FRUITS. Calif. Avocado Soc. Yearbook 40:43 - 44.

AHMED, M. S.
1936. POLLINATION AND SELECTION IN ANNONA SQUAMOSA AND A. CHERIMOYA. Egypt Min. Agr. Tech. and Sci. Serv., Hort. Sect., Bul. 157, 29 pp.

BROOKS, R. M., and HESSE, C. O.
1953. WESTERN FRUIT GARDENING. 287 pp. University ot California Press, Berkeley and Los Angeles.

____ and OLMO, H. P.
1952. REGISTER OF NEW FRUIT AND NUT VARIETIES: 1920-50. 205 pp. University of California Press, Berkeley and Los Angeles.

CLARK, O. I.
1925. CHERIMOYA INVESTIGATIONS AT POINT LOMA HOMESTEAD. Calif. Avocado Assoc. Ann. Rpt. 1924-25, pp. 14-17.

KRISHNAMURTHI, S., and MADHAVA RAO, V. N.
1963. CERTAIN PROBBEMS IN POBBINATION OF FRUIT CROPS. So. Indian Hort. 11(1/2): 1-17.

SARASOLA, L.
1960. CHERIMOYA CULTURE IN SPAIN. Calif. Avocado Soc. Yearbook 44: 47-53.

SCHROEDER, C. A.
1941. HAND POLLINATION EFFECTS IN THE CHERIMOYA (ANNONA CHERIMOBA). Calif. Avocado Soc. Yearbook 1941: 67-70.

____ 1948. REPORT OF THE COMMITTEE ON SUB-TROPICAL FRUIT VARIETIES. Calif. Avocado Soc. Yearbook 1948: 17 - 19.

____ 1956. CHERIMOYAS, SAPOTES AND GUAVAS OF CABIFORNIA. Calif. Avocado Soc. Yearbook 40: 49 - 56.

____ 1971. POLLINATION OF CHERIMOYA. Calif. Avocado Soc. Yearbook 54: 119 - 122.

THAKUR, D. R., and SINGH, R. N.
1965. STUDIES ON POLLEN MORPHOLOGY, POLLINATION AND FRUIT SETIN SOME ANNONAS. Indian Jour. Hort. 22: 10-18.

THOMSON, P. H. THE CHERIMOYA IN CALIFORNIA. Calif. Rare Fruit Growers Yearbook 1970 (2): 20-34. 138

WATTS, J. C.
1942. THE FRUIT OF THE GODS. Calif. Cult. 89(7): 170.

WESTER, P. J.
1910. POLLINATION EXPERIMENTS WITH ANNONAS. Torrey Bot. Club Bul. 37: 529 - 539.

CLOVE
Syzygium aromaticum Merr. and L. M. Perry, family Myrtaceae
The clove tree produces flower buds that when harvested and dried become the cloves of commerce, which are used primarily as food spices. Clove oil, distilled from the plant, is used in perfumes, medicines, artificial vanillin, dentifrices, and other ways.

Purseglove (1968*) stated that the annual worldwide consumption of cloves was as follows: Indonesia, 8,000 metric tons; India, 3,000; Malaya, 2,000; United States, 2,000; Europe and North Africa, 3,000; and other countries, 1,000 metric tons - about 42 million pounds. Rosengarten (1969*) stated that Tanzania produces three-fourths of the world output.

Plant:
The clove tree, although related to the eucalyptus and some other large trees, is relatively small, 12 to 20 feet or, rarely, to 40 feet tall. The stem is often forked with two or three main trunks. The paired leaves are 3 to 5 inches long, 1 to 1 l/2 inches wide, and highly aromatic.

The plants are usually grown from seed, then set about 30 feet apart in the grove.

Inflorescence:
The clove tree inflorescence is a terminal branching cyme of 3 to 20 hermaphrodite florets, the whole about 1 l/2 inches long. Each pale yellow floret consists of a cylindrical thick ovary, one-quarter inch long. Above the ovary are four fleshy ovate sepals, and above these are the four tiny petals, numerous slender white 3/8-inch filaments, and a slender central style. The flower opens early in the morning. The united petals separate from the base as a cap, similar to the grape blossom, which is pushed off by the extending stamens. In a few hours, all the anthers are open, and the stigma is receptive (Wit 1969). There are two flowering seasons a year, July to October and November to January. Few flowers develop into fruit. The fruit, called mother of cloves, contains one seed or rarely two seeds. The ovary and sepals constitute the specific part marketed as cloves (Purseglove 1968*, Ridley 1912*).

Pollination Requirements:
Purseglove (1968*) said that no fertile fruits were obtained from bagged flowers. He concluded that cross-pollination was necessary for seed production. Tidbury (1949) stated that no viable seeds have been produced from selfed flowers, indicating that the flowers require cross- pollination. He also concluded that, since vegetative propagation had never been accomplished, pollination from the breeding standpoint becomes important.

Pollinators:
Purseglove (1968*) stated that the flowers are visited and apparently cross-pollinated by bees. Ridley (1912*) merely stated that the fertilization was by some insect. Tidbury (1949) stated that the flowers are visited by bees.

Pollination Recommendations and Practices:
No attempt is made to utilize insect pollination in the production of clove planting seed. The figures by Purseglove (1968*) on tree spacing, tree yields, and total production of cloves would indicate that about 70,000 acres are involved, and reproduction of plants occurs at the rate of about 1,000 acres per year.

The need for sufficient pollinating insects to produce the small amount of seed required to plant 1,000 acres of cloves is probably not acute. There would be a definite need for insect pollinators if the production were concentrated in certain areas and maximum seed production desired. If such were the case, bees could be concentrated in the planting to perform the required pollination.

No known attempts have been made to use pollinating insects in clove seed production.

LITERATURE CITED:
TIDBURY, G. E.
1949. THE CLOVE TREE. 212 pp. Crosby Lockwood and Sons Ltd., London.

WIT, F. 1969. THE CLOVE TREE. In Ferwerda, F. P., and Wit, F., eds., Outlines of Perennial Crop Breeding in the Tropics, pp. 163-174. H. Veenman and Zonen, N. V. Wageningen, The Netherlands.

COTTON
Gossypium spp., family Malvaceae
Cotton is grown primarily for its lint, although the seed, valued at about one-fifth that of the lint, is also used for planting or is crushed, a food oil recovered, and the residue used as a livestock food.

Cotton was naturally dispersed throughout much of the warmer parts of the world even before 3,000 B.C., when its lint was spun and woven into cloth (Gulati and Turner 1928). Only four of the numerous species of Gossypium are cultivated for their spinnable fibers. These are the two diploid Asiatic species, G. arboreum L. and G. herbaceum L., and the two amphidiploid species, G. barbadense L. and G. hirsutum L. The first two are confined primarily to the Old World. The last two are grown both in the New World and abroad. G. hirsutum, known as upland or short staple cotton, is grown most extensively. G. barbadense, known in the United States as American Pima, or Pima, and sometimes called extra-long staple cotton, is grown in long-season areas, such as our Southwestern States.

The crop is grown from Virginia southward and westward to California, in an area often referred to as the Cotton Belt. The more important cultivars ²⁴ of upland cotton planted in 1971 included 'Deltapine 45', which accounted for 25 percent of the total acreage, 'Stoneville 213', which accounted for 18 percent, and 'Acala SJ-1', which accounted for 10 percent. Other major cultivars and their percentages included: 'Lankart LX 571', 7; 'Coker 201', 4; and 'Paymaster 111', 'Stoneville 7A', and 'Stripper 31', each 3 percent (USDA 1971).

Texas, California, Mississippi, and Arkansas were the leading cotton producing States in 1970. The value of the entire U.S. crop was approximately $1.2 billion. The leading cotton producing countries of the world in 1970 are shown in table 12.

The lint percentage of cotton varies from 30 to 40 percent, more often in the high 30's for upland and the low 30's for Pima.

[gfx] fix table 12
TABLE 12. - Leading cotton producing countries of the world in 1970¹ __________________________________________________________ Millions Pounds Millions of of lint of bales Country acres per acre of lint 2 ______________________________________________________________________________ India.................................................................... 19.0 111 4.4 China (mainland)............................................ 12.5 268 7.0 United States.................................................. 311.2 437 10.3 Russia................................................................. 6.8 761 10.8 Brazil.................................................................. 6.0 183 2.3 Pakistan............................................................ 4.3 271 2.4 United Arab Republic................................... 1.7 664 2.3 Turkey................................................................ 1.3 676 1.8 Sudan.................................................................. 1.3 406 1.1 Mexico................................................................ 1.0 692 1.4 __________________________________________________________ l Source: Anonymous 1972. 2 An average bale weighs 480 pounds. 3 Includes about 77,600 acres of Pima.

Recognizing the variation in cotton due to natural speciation, breeder manipulation, the wide distribution, and the conditions under which it is grown, the following discussion will be largely confined to upland cotton. Pima and, to a much lesser degree, the other two species, will also be mentioned. Because of some lack of agreement on the pollination requirements of cotton, more than the usual amount of space is devoted herein to this crop.
___________
²⁴In cotton, a cultivar, or "variety," is neither a clone, a pure line, nor a primary mixture of pure lines. It is usually a progeny row selection, bulked and mass multiplied, during which time insect pollination may have played a major or insignificant role, depending upon the pollinator population present during the flowering season.

Plant:
The cotton plant is a broad-leaved perennial, 2 to 5 feet tall, that is treated as an annual under much of its growing conditions in the United States. It becomes a perennial if the ground in which it grows does not freeze during the winter. The plants will grow and be productive on a wide variety of soils. It is most productive on fertile soil under hot weather and irrigated conditions if rainfall is deficient. The seeds are usually planted 4 to 8 inches apart in about 3-foot rows after all danger of frost is past in the spring. Flowering on the first of its lower branches begins in about 2 months and may continue on succeeding branches and growth another 2 months at about which time the first ripe fruit (bolls) begin to open and expose the mature lint-covered seed (fig. 87). Most of this raw cotton is currently machine-harvested then transported to a cotton gin where the lint is separated from the seed then pressed into bales.

Flowering and fruiting on the plant follows a spiral course from the innermost bud on the oldest and lowest fruiting branch, and ends on the latest growth toward the tip of the plant.

Pima cotton is usually more robust than upland cotton, with waxy- green leaves and smaller bolls. Fruiting on the plant begins later in the season, which tends to restrict its culture to the area with the longest growing season, such as the extreme southwestern United States.

Only about half as much lint per acre is normally produced on Pima cotton as is produced on upland cotton; however, the grower receives about twice as much for the more desirable lint, so the net profit from the two types of cotton is similar. The lint can be removed from the seed with the same type of gin saws that are used on upland cotton, but the quality of the lint is better if it is removed with a roller gin.

[gfx]
FIGURE 87. - Open bolls of cotton.

Inflorescence:
The 2- to 4-inch-long by 2-inch-broad cotton blossom is subtended by three green leafy bracts, each an inch or more across, and a green calyx that fits snugly around the base of the ovary. The five-petal corolla of upland cotton is cream colored when it opens in the morning shortly after sunup, but turns pink in the afternoon and closes toward nightfall never to reopen (fig. 88). On the second day, the color of the petals is a watermelon-red. The typical corolla of Pima cotton is yellow, with a maroon throat or petal spot, and the color changes little with age. The corolla and stamina column usually fall on the second day.

The staminal column surrounds the 1- to 2-inch-long style leading from the ovary and terminating in the l/4 to l/2-inch-long stigma (fig. 89). The ovary contains 5 to 10 ovules in each of three to five sections, carpers, or locules. The stamina sheath, enclosing most of the style, bears numerous stamens l/4 to 1/2 inch long, each terminating in an anther that normally produces an abundance of viable self-fertile pollen, 45,000 grains per flower (Tsyganov 1953). The grains are large, 81 to 143 microns (Kaziev 1964), and coated with a viscid material that causes them to adhere to each other; therefore, cotton pollen is not transported by wind. Each section of the oval, 1-inch boll that develops from the ovary may produce a "lock," a distinct group of lint-entangled seed. These locks are exposed in the open three- to five-sectioned "burr."

The number of flowers on a cotton plant are determined by numerous factors including the available plant food, water supply, variety, and density of the plant population. Usually, about half of the flowers produce mature bolls (Brown 1938, Buie 1928, Dunlap 1945, McNamara et al. 1940, Sen and Afzal 1937). Flowering reaches its peak at about four flowers per plant per day. Between 225 and 400 bolls are usually required to produce a pound of fruit.

[gfx]
FIGURE 88. - Cotton flower, showing general corolla shape an proximity of anthers to stigma. A, Upland cotton flower with flared, cream-colored corolla, and stigma protruding only slightly above the anthers; B, Pima cotton flower, with tubular shaped yellowish corolla and dark "petal spot" toward base. Stigma extends well above the anthers.

FIGURE 89. -Flower of 'Deltapine' upland cotton, x 1.5. A, Side view, showing one of the 3 subbracteal nectaries; b, bracts slightly spread to show one of the 3 inner bracteal nectaries; C, longitudinal section.

NECTARIES AND NECTAR SECRETION OF COTTON:

Nectar is normally produced in five different areas on the cotton plant, although the reason why the nectar is secreted is not clear. Trelease (1879) made a detailed report on cotton nectar secretion and its possible purposes (although he stated "Glover, Agricultural Report 1855, p. 234, mentions these glands - and their secretion of a sweet substance, which ants, bees, wasps, and plant bugs avail themselves of as food"). Kaziev (1964) stated that Delpino, in 1900, was the first to characterize the floral and foliar nectaries of cotton.

The different areas of nectar secretion are (1) floral, (2) inner or circumbracteal, (3) outer or subbracteal, (4) foliar or leaf, and (5) unipapillate (microscopic) areas on the flower peduncles and young leaf petioles (Mound 1962) (fig. 90). These unipapillate nectaries are rarely visited by pollinating insects, contribute nothing to pollination and little or nothing to the welfare of pollinating agents, and will not be further discussed. Trelease (1879) believed that the floral nectaries were associated with pollination but that extra-floral nectaries were associated with attracting harmful insects away from the delicate flower parts. Kottur (1921 ) believed that the nectar and pollen in the flower invited natural crossing.

In addition to the secretion from the nectaries, there is a saccharine exudation of certain aphids, white flies, and thrips on cotton, known in the Sudan as "asal" (Bedford 1921). When this material is present in abundance, it supports growth of a sooty fungus, causing a detrimental blackening of the cotton leaves. At times, the sticky material becomes mixed with the lint with damaging effects (Hadwich 1961). When honey bees are present in sufficient numbers, they collect this material as food, and by removing it they become beneficial in a sense other than as pollinators.

[gfx]
FIGURE 90. - Nectaries of cotton plant. A, Honey bee collecting nectar from a subbracteal nectary; B, inner bracteal nectaries; C, section of calyx removed to show proboscis (tongue) of bee (while bee is inside flower) reaching for floral nectar droplets; D, underside of leaf, showing location of leaf nectaries.

FLORAL NECTARIES:

Within the flower, the nectar exudes from a ring of papilliform cells at the base of the inner side of the calyx (Tyler 1908). Secretion may begin a few hours to a few days before the flower opens, but, because of its unavailability to pollinators until the flower opens, this possible early secretion is of no consequence. The nectar reaches its maximum accumulation by mid-afternoon, the amount depending upon climatic factors, soil fertility, water, and cultivar involved (Kaziev 1959a, b, 1967), and ceases when the petal color begins to change, an indication, according to Kaziev (1964) and Mel'nichenko (1963) that pollination has occurred.

NECTAR COMPOSITION IN RELATION TO ATTRACTIVENESS TO BEES:

At times, honey bees appear to be noticeably reluctant to visit cotton blossoms, even though much nectar and pollen are present. Wykes (1952) studied the preference of honey bees for solutions of various sugars and found that preferences were shown for solutions of single sugars in the following descending order: Sucrose, glucose, maltose, and fructose, and that mixtures of equal parts of all these sugars was the most attractive combination. Vansell (1944a, b) studied the composition of sugars in orange and cotton floral nectar and found the percentages shown in table 13.

[gfx] fix table 13
TABLE 13. - Percentage of sugars and moisture content of cotton and orange nectars _________________________________________________________ Plant Levulose Dextrose Sucrose Approximate Moisture __________________________________________________________ Cotton: ÔAcalaÕ 14.27 13.06 0.71 70 ÔPimaÕ 10.36 9.25 .35 80 Orange: ÔWashington NavelÕ 6.46 5.42 12.87 75 ÔValenciaÕ 6.08 5.06 12.38 77 __________________________________________________________

Butler et al. (1972) collected nectar from leaf, floral, and extra- floral nectaries of 'Hopicala', 'Deltapine Smooth-leaf', 'Deltapine 16', and 'Pima S-2' near Tucson, Ariz., and analyzed it by gas chromatography for its sugars. They also found low sucrose percentages in nectar from floral and subbracteal nectaries.

Ivanova-Paroiskaya (1950) showed the fructose - glucose - sucrose percentages of floral nectar of G. barbadense cotton, cv. '35-1', to be 39.78-37.50-1.63, and for upland cotton, cv. '36 - 7 M', to be 37.85-35.65- 6.89. Kaziev (1964) showed that the range of the sucrose content of cotton nectars was from 2.3 to 7.6 percent, with the total mono-sugars ranging from 21.2 to 46.9 percent. All samples were taken during mid-season flowering. Whether the percentages change with the season has never been determined.

Numerous observations have shown a relatively low percentage of honey bee visits to flowers of cotton during mid-season and a high percentage toward the beginning and the end of the season. For example in August 1952 (unpublished data), at Tucson, Ariz., eight blossoms of 'Pima S-1' were observed constantly from 8:45 until 11:30 a.m. During that time, they received visits from one honey bee, one Bombus spp., 100 Melissodes spp., and five unidentified pollen-collecting bees. In the same plot, on October 10 between 7 a.m. and noon, three blossoms of 'Pima S-1' were visited by 363 honey bees and seven Melissodes bees. The reason for this extreme difference in the number of bee visitors is unknown.

The volume of nectar in the blossoms of Pima cotton is greater than that in upland cotton, but the sugar concentration is lower. As much as 0.1 ml of nectar has been collected at one time from the former but never more than half this amount from upland. The sugar concentration of upland cotton floral nectar is greater (reaching a maximum of about 69 percent) than that of Pima nectar (a maximum of only 34 percent (fig. 91). The volume of floral nectar of both cottons exceeds the volume of extrafloral nectar, but, as previously stated, the floral nectar is less attractive to honey bees.

Nectar secretion of cotton is strongly influenced by soil fertility, as various tests made in Russia have shown. For example, superphosphate increased nectar secretion by 170 percent and potassium by 130 percent (Monokova and Chebotnikova 1955). Extensive studies by Kaziev (1964) showed that nitrogen had no effect on nectar production, but the greatest increases resulted from application to the soil of cattle manure alone or with complete fertilizers.

[gfx] fix diagram
FIGURE 91. - Average and range, in percent, of sugar (soluble solids) of floral nectar of 6 different cotton cultivars grown at the USDA Cotton Research Center, Phoenix, Ariz., August 1957.

SUGAR CONCENTRATION OF FLORAL NECTAR, PERCENT 7n . Approximate _ _ Plant Levulose Dextrose Sucrose Moisture Cotton: 'Acala' 14.27 13.06 0.71 70 ~ , 'Pima' 10.36 9.25 .35 80 60 _ Orange: _ t 'Washington Navel' 6.46 5.42 12.87 75 'Valencia' 6.08 5.06 12.38 77 50 _ ~ _ _ ~_ _

NECTARIES OUTSIDE THE COROLLA (INVOLUCRAL NECTARIES):

There are usually three nectaries just below the sepals at the union of the three bracts, and three others just below the base of the bracts. These have variously been referred to as calyceal and subbracteal, inner and outer involucral, extrafloral internal and extrafloral external, and bracteal nectaries. They begin to function several days before the flower opens, but the day before opening they secrete nectar in copious amounts and this secretion may then continue from several days to 2 to 3 weeks after flowering. When beekeepers report good honey flows from cotton, the bees are usually working these involucral nectaries more intensively than the other nectaries of cotton. Kaziev (1964) also showed that these involucral nectaries were highly attractive to bees. Some upland cotton plants may not have all of these nectaries.

In Hawaiian cotton (G. tomentosum Nutt. ex Seem.), the nectaries are entirely absent. This characteristic has been transferred to some commercial selections to provide a nectariless cotton (Meyer and Meyer 1961). These selections are studied to determine the effect of nectar deprivation on harmful lepidopterious insects on cotton (Lukefahr and Martin 1964, Lukefahr and Rhyne 1960). The nectariless characteristic has been suggested in connection with proposed production of hybrid cotton, the theory being that if only floral nectar were present the bees would visit the flowers more frequently. The opposite, however, may be true, that is, the bees may be attracted to the field by the extrafloral nectaries and then discover the floral ones. Research in this area is requisite for a successful hybrid cotton program if pollinating insects are used.

FOLIAR OR LEAF NECTARIES:

Nectar is also produced on the underside of the leaf in usually one nectary on the primary vein near the blade and petiole juncture. Occasionally, there are two additional nectaries, one on a vein on either side of the primary vein. The leaf nectary begins to function before the leaf reaches full size and may function for 2 or 3 weeks. Leaf nectar secretion on the plant begins when the first flower is in the early bud, or "square", stage and may continue as long as the plant is producing new leaves (Kaziev 1964). The structure of cotton leaf nectaries was described by Reed (191 7).

Honey bees show preference for all of the extrafloral nectaries over floral nectar but under certain conditions collect nectar from all. Ivanova-Paroiskaya (1956) showed the following relative number of honey bee visits to the nectaries: Floral 32, calycular 219, subbracteal 580, and leaf 389. Many other insect species show a preference for the floral nectary of cotton.

COTTON AS A HONEY PLANT:

In many of the areas where cotton is grown it is considered a major honey plant (Benson 1937, Kuliev 19S8, Minkov 1957, Parks 1921). It does not produce as much honey per acre per day as alfalfa, clover, or many other sources (Butler et al. 1972) but because of its longer flowering period, excellent crops of honey can be obtained. Pima cotton is recognized by beekeepers as a better honey source than upland cotton. Vansell (1944a) calculated that 1 acre of Pima cotton was equal to 30 acres of Acala in the production of nectar. Ivanova-Paroiskaya (1950) reported honey crops of 300 kg/ha (267 lb/acre) for G. barbadense compared to 75 to 90 kg/ha (66 to 80 lb/acre) for upland. Normally, when bees are working cotton blossoms, there is a steady, but not spectacular, storage of the high-quality cotton honey. Unfortunately, highly toxic insecticides, frequently used on cotton during the flowering period, kill many bees and prevent the harvest of a honey crop.

Pollination Requirements:
Cotton is usually referred to as a partially cross-pollinated crop, although many breeders have treated it as a completely self-fertile and self-pollinating crop except for accidental and unwanted cross- pollination caused by pollinating insects. Cross-pollination has been referred to as "natural crossing," and is considered detrimental because of the introduction of off-type plants into the progeny. Its impact on production has not been given much consideration. Breeders know that a cotton blossom isolated by any one of several techniques will usually develop a mature boll with viable seed. Breeders also know that they seldom obtain by this method quite as many seed or as much lint from such bolls as they obtain from open-pollinated ones (Stephens 1956). Because this difference can be altered to the benefit of the grower by the activity of pollinating insects, and because of the association of pollinating insects with natural crossing and resultant undesirable cotton types, the subject of natural crossing is reviewed.

NATURAL CROSSING:

The term "natural crossing" is freely used in cotton literature, but not always precisely defined. The generally accepted meaning appears to be "the amount of cross-pollination effected by insects, as opposed to hand cross-pollination, that can be detected in breeding lines." Fryxell (1957) defined natural crossing as "that which occurs between individuals within a population." He stated that the two phenomena (intra- and inter- population crossing) are related, but the distinction is not always made clear. Simpson (1954a) stated that cross-pollination is not readily detected in cotton unless distinctive marker characters are present in the parental lines. For this reason, the extent of natural crossing in open- pollinated fields has been consistently underestimated by some cotton breeders. Natural crossing is usually associated with insect pollination of plants but seldom with its benefit to the cotton plant.

Various breeders have reported the percentage of natural crossing in their area and commented upon its detrimental effect. Balls (1912) reported 13.3 percent natural crossing of cotton in Egypt, and proposed the isolation of plants under mosquito netting to exclude bees. He noted, however, that some strains "resent this treatment and refuse to hold their bolls," a possible indication that he might have been unknowingly dealing with a degree of self-sterility in the plants, although shading of the plants might also have contributed.

Webber (1903) also reported 5 to 10 percent natural crossing between varieties of upland cotton grown in adjacent rows in the United States and concluded that absolute prevention of crossing would require isolation of cotton by 5 to 10 miles from other cotton. Ricks and Brown (1916) reported 2.8 to 18.5 percent in alternate hills in alternate rows. The type or preponderance of insect pollinators responsible for this cross- pollination was not given.

Cook (1921) stated that natural cross-fertilization in the field is one of the major causes of "running out" of varieties. Later, Cook (1932) stated that maintaining the seed stock of superior strains rather than creation of new ones was the essential breeding problem. To maintain good seed stocks, he stressed complete isolation from other varieties to prevent contamination by natural crossing. Peebles (1942) considered 1 mile as a sufficient isolation distance.

Ware (1927) stated that the amount of natural crossing, providing all other factors were the same, would be in direct proportion to the number of insects capable of carrying pollen. He made observations on the cross-pollination of cotton in two areas - at Scott, Ark., where the cotton acreage was large, and at Fayetteville, Ark., where there was little cotton in the vicinity other than the breeding plots. Honey bee colonies were fewer at Scott than at Fayetteville, the number of pollinating insects visiting the cotton flowers was much smaller, and less than 1 percent hybrids were obtained, compared to 40.9 percent hybrids at Fayetteville. Ware (1927) concluded that there was a close association between the presence of honey bees and other pollinating insects and the amount of cross-pollination obtained. Afzal and Khan (1950) reported 2 percent natural crossing with seven or eight visits per flower by insects daily, principally Apis dorsata Fab., Anthophora confusa Smith, and Elis thoracica Lepeletier.

Stephens and Finkner (1953), considering the beneficial effect of insect cross-pollination in the possible production of hybrid cotton, indicated that cross-pollination in different areas ranged from 5 to 50 percent or more, which they associated with differences in the effective bee population. They concluded that even in the area where the higher percentage of crossing occurred, the number of bees did not keep pace with the number of flowers available, so that the flowers were worked less effectively during the peak. Their proposed solution was the provision of a supplementary source of bees during the flowering period.

Simpson (1954a) made a survey of the natural crossing across the Cotton Belt and stated that the cooperators in the experiment were "in general agreement concerning the following factors that influence the amount of natural crossing: (1) cotton pollen is relatively heavy and wind is not an agent in pollen dispersal; (2) therefore, the amount of natural crossing in cotton is determined by the number of insect pollinators present in relation to the number of cotton flowers; and (3) intercrossing may be affected by the flowering habits of the varieties grown, by the abundance of unlike -pollen, by location of the fields in relation to insect habitats, by flowering periods of other plants attractive to insect pollinators, by distance between unlike varieties, by topography and barrier crops, and by other environmental, climatic and biotic factors."

Further in the same paper, he stated that natural crossing had heretofore been considered a handicap in breeding programs and a hazard to be avoided, but that a beneficial intermingling of unlike genotypes could be obtained by increasing the population of insect pollinators. This was supported earlier by Brown (1927). Conversely, Harland (1943) proposed that a completely self-fertile cotton might be grown, a direction that some breeders seem to seek, but which is unlikely to result in the most productive type.

Simpson (1954a) also stated: "Cotton is a partially cross-pollinated plant, thus some degree of heterozygosity is maintained indefinitely when open pollinated seeds from an original F₁ population is continued on through F₂, and F₃, and subsequent generations. The relative proportions of selfing and outcrossing determine the amount of hybrid vigor retained in later generations." He hinted that this relationship could be utilized advantageously in current breeding programs. Kalyanaraman and Santhanam (1957) also hinted that breeding programs such as the mass pedigree method (Harland 1949a,b) should be modified to allow increased genetic plasticity through open pollination and bulk methods of breeding.

Knight and Rose (1954) proposed that one stage of the cotton breeding program be conducted in an area of "high natural crossing" for improvement of the variety. This proposal has not been adopted in the United States, although it and the other statements would indicate that in current cotton breeding systems the breeders should strive for a high bee population in their seed increase fields. Somewhat similar usage of pollinating insects has been suggested by Simpson (1954a) and Turner and Miravalle (1961). The absolute dependence on pollinating insects for success in the Knight and Rose (1954) method was stressed by Bhat (1955), who concluded that natural crossing could result in deterioration of the variety, "not as a result of cross-pollination as is often fallaciously argued but because of its inadequacy." Al-Jibouri (1960) seemed to agree, for after reporting an average of only 0.47 percent crossing in Iraq, he concluded that this inbreeding could result in deterioration of the variety.

It stands to reason that the grower would want to preserve heterosis through favorable gene combinations, or crossing of proper types, and refrain from allowing crossing with inferior or unwanted types.

Many cotton breeders seem to believe that the natural crossing in an area, once established, does not vary from year to year. Thus, they speak of areas of high natural crossing without considering that the factors responsible for crossing may be considerably altered from season to season or even within the season. Humbert and Mogford (1927) stated that cross-fertilization in cotton will vary from 2 to 20 percent and will not average over 15 percent under normal conditions. The "normal" conditions were not described. Fikry (1931) also indicated that natural crossing was a rather fixed 4 percent. Sappenfield (1963) reported a range of 1.0 to 32.2 percent (average, 13.6 percent) natural crossing in Missouri; Simpson (1950)²⁵, a range of 3.3 to 90 percent in his Cotton Belt survey; while Thomson (1966) reported 1 to 2 percent natural crossing in the Ord Valley of northwest Australia. He noted that there were no honey bees in the valley and that insecticides were applied weekly or more often, which, he stated, undoubtedly suppressed any wild bee activity.

Natural crossing, once considered a hazard to be avoided, now is being presented as a tool to be utilized in the development of superior varieties. Also, with the high degree of efficient transportation of apiaries by beekeepers from one area or crop to another, there is no longer assurance that a location will have the same pollinator population from year to year or even throughout the season. This alteration in pollinator population is further accentuated by applications of toxic pesticides to crops, which may damage, destroy, or cause removal of the majority of the pollinators in a given area.

In the "natural pollination" studies made by cotton breeders, none have indicated the relative number of pollinators or visits responsible for the crossing.
___________
²⁵ SIMPSON, D. M. MEMO TO: COOPERATORS IN NATURAL CROSSING TESTS (RESULTS OF 1949 TESTS). Tenn. Agr. Expt. Sta. and U.S. Cotton Field Sta., Knoxville, 3 pp. 1950. [Mimeographed.]

MOTES IN COTTON AND THEIR SIGNIFICANCE:

Cotton ovules that fail to develop into seeds with well-developed ginnable fibers are termed "motes." The motes in a cotton crop represent a loss in yield. If 15 to 20 percent or more of the ovules fail to produce seeds with ginnable fibers, the potential yield is lowered just that much (Pearson 1949a). Some structures at the base of certain locks of cotton have the appearance of motes, but they are not derived from ovules. They are termed "false motes." The failure to recognize them as false motes can contribute to an error in estimating the potential seeds that fail to mature (Pearson 1949b).

Rea (1928) reported 6.5 percent motes at the apex of the locks of 'Anton' (upland) cotton and 25.3 at the base, with 11.1 and 38.5 percent, respectively, in 'Bolton' cv. He concluded that this difference might be due to incomplete fertilization. Afzal and Trought (1934) also concluded that mote formation may be due to defective pollination. Rea (1929) found that the higher the percentage of motes, the smaller the boll and the lower the yield of cotton. Hughes (1968) studied motes in 'Bar 14/25', and concluded that most of the motes occurred near the base of the lock in ovary positions 1 and 2. Porter (1936) stated, "To the extent that motes represent unfertilized ovules, it is natural that fewer motes would be found near the apex of the lock, as the ovules near the top of the ovary are favored in fertilization, being reached first by the pollen tubes. If the quantity of pollen deposited on the stigmas is scanty, or much of it defective, there would be a correspondingly small chance of the lower ovules being fertilized." Pearson (1949a) also noted that the number of motes increased with the number of ovules, again an indication of inadequate pollination.

Not only does inadequate pollination contribute to reduced yield through failure of ovules to develop, but mote formation could result in another type of loss. It is easily demonstrated that when a ripe lock of cotton is pulled from the burr, the lock is most likely to break at the location of a mote, and when two motes are adjacent to each other, a break is almost certain. If such a break should occur in the mechanical harvesting of cotton, the remaining portion of the lock would likely be missed by the machine and either lost or only salvaged later at reduced quality during complete destruction of the burr.

The preponderance of bolls with motes in commercial fields indicates the loss of yield being experienced. If the plant "makes up" for motes by setting more bolls, harvest is delayed. The quality of late- developing lint is likely to be inferior to earlier lint.

SHEDDING:

The cotton plant frequently sheds half or more of its fruit. Some of this shedding occurs in the bud stage, but the peak occurs about 5 or 6 days after flowering, then tapers off toward the full-grown boll stage. Many factors contribute to shedding, including humidity, temperature, soil-water conditions, genetic factors, diseases, insects, mechanical injury, and inadequate pollination (Beckett and Hubbard 1932, Brown 1938, Eaton and Ergle 1953, and Lloyd 1920). Rainfall during the day the flower is open can damage the pollen and cause shedding although the amount of shedding from this factor is minor. Kearney (1923) doubted that deficient pollination and fertilization were the primary reasons for boll-shedding in Pima cotton at Sacaton, Ariz. However, Kaziev (1964) showed that bee pollination caused a decrease in the shedding of five common cultivars of upland cotton. McGregor et al. (1955) also reported reduced shedding of 'Pima S-1' cotton visited by bees as compared to plants caged to exclude bees. Occasionally, during periods of extremely hot weather, the cotton flower will fail to produce or dehisce viable pollen. When this occurs, the flower sheds unless pollen is brought to it from another flower by insects.

Many factors, therefore, contribute to shedding by cotton, one of which is inadequate pollination.

NEED FOR POLLEN TRANSFER WITHIN OR BETWEEN PLANTS:

When the cotton pollen grains contact the stigma, they germinate tubes that grow down the "conducting tissue" in the center of the style (Arutinnova 1940). When a tube reaches the ovary, the sperm enters and fertilizes an ovule. About 50 ovules must be fertilized if a full complement of seeds is produced; therefore, at least 50 viable pollen grains must contact the stigma. The stigma is normally receptive to the pollen by the time the flower opens or before (Loden et al. 1950), but receptivity drops sharply after about noon (Janki et al. 1968). Because of this limited time span, most hand pollination in cotton breeding work is made during the forenoon.

The majority of the flowers on a cotton plant are largely self- fertile and, to varying degrees, self-pollinating (Ewing 1918, Kearney and Harrison 1932). The method, time, and type of pollination of the stigma influence the degree of self-pollination. Flowers that receive pollen on the whole surface of the stigma yield more seeds per boll than those pollinated at the base of the stigma (Kearney 1926). The base of the stigma affords less favorable condition for pollen germination or growth than the apex (Iyengar 1938, Kearney 1923). Pollen from anthers on the lower part of the stamina column is best (Trushkin 1956). Repeated applications of pollen to the stigma, such as repeated bee visits, are also helpful (Finkner 1954). Therefore, for best pollination of the stigma, repeated applications with an abundance of pollen from the basal anthers on the tip of the stigma insures the highest percentage of fruit set and maximum cotton production.

The pollen-laden anthers that touch the stigma, resulting in self- fertilization, usually contact only the base of the stigma. When this occurs, self-pollination is assured, but maximum fertilization within the ovary of that boll is not usually accomplished. Arutiunova and Gubanov (1950) indicated that "pollen seems to stimulate pollen" so that increased amounts of it on the stigma increased the percentage of germinating pollen tubes, further insuring ovule fertilization.

Rose and Hughes (1955) increased yield of 'Bar 7/8.2' by 11 percent over naturally pollinated flowers by brush pollination of stigmas. The increase resulted from more bolls set. Kohel (1968) also obtained more seeds per lock and per boll from unbagged open flowers, or hand-pollinated open flowers than from bagged or emasculated flowers. Guseinov and Muktarov (1963) showed that cross-pollination within the variety resulted in increased production.

Some pollens are too weak to compete with self-pollen. For example, when McGregor et al. (1955) grew equal numbers of 'A-44' and 'Red Acala' plants in cages with honey bees, only 2.31 percent of the offspring of 'A- 44' were hybrids, whereas 44.0 percent of the 'Red Acala' offspring were hybrids.

As the pollen tube grows down the style, its nucleus moves a few microns ahead of the sperm (Jensen and Fisher 1968). The sperm and contents are discharged into the germ sac of the ovule in 16 to 32 hours (Gore 1932, Kearney 1923, Saakyan 1962). Surplus pollen tubes that penetrate the ovary are eventually assimilated without damage (Linskens 1964), so there is no damage from surplus pollen on the stigma.

According to Arutiunova (1940), the pollen tube begins to form more quickly if the pollen grain is from a genetically different cultivar. For example, tubes in cross-pollinated flowers were visible within 5 to 10 minutes after the pollen was placed on the stigma, but tubes from self pollen did not appear until after 60 to 150 minutes. Also, tubes from pollen placed on the tip of the stigma grew faster than those from pollen placed on the base of the stigma.

Arutinnova (1940) also studied the effect of the number of pollen grains and the kinds of intravarietal pollen on pollen tube growth in both upland and G. barbadense cottons and reported that in the G. barbadense cottons twice as many tubes reach the ovary with cross-fertilization as with selfed flowers. Except for one cultivar, pollen of upland cottons grew better on its own stigma than on others. In many of the selfed flowers, no pollen tubes developed, but in crossed flowers this failure was rare.

This means that more self pollen must be deposited on the stigma than mixed or cross pollen. The sooner the tube reaches the ovary and fertilizes an ovule, the less the likelihood that the fruit will shed. Arutiunova and Gubanov (1950) also concluded that an increased number of pollen grains on the stigma increased the percentage of germinating grains and tube development. Arutiunova and Kanas (1955) concluded that cross-pollination within the cultivar insures the best seed development. In other words, a well cross-fertilized stigma tip is most likely to result in the best fertilization of the ovules in the ovary.

There seems to be constant warfare within the plant between setting and shedding of fruit. Anything that can be done to influence the battle in favor of fruit setting tends to increase production. The earliest possible thorough application of pollen on the tip of the stigma to insure speediest arrival of the tubes in abundance within the ovary would insure maximum set.

BENEFITS DERIVED FROM INSECT POLLINATION OF COTTON:

The benefits derived by cotton from insect pollination have been cited by numerous workers, largely in Russia, where much attention has been given to this subject, but also in Egypt, India, and the United States. Meade (1918) was the first to call attention to the fact that cotton at San Antonio, Tex., should benefit from insect pollination. He applied supplementary pollen by hand to open flowers that may or may not have been visited by pollinating insects and increased the set of 'Durango' cotton flowers by 10.96 percent and 'Acala' flowers by 5.31 percent. His results convinced him that pollination of cotton by honey bees should increase production.

Kearney (1921) was stimulated by Meade's work to compare naturally pollinated Pima flowers with flowers that received supplementary pollination by hand. At Sacaton, Ariz., where the cotton was relatively isolated and insect pollinators were prevalent, he obtained no significant increase in set of flowers or seed produced. However, at Phoenix, in a large cotton-growing area, he obtained only 1,157 seeds per 100 naturally pollinated flowers, but 1,526 seeds per 100 flowers that received supplementary pollination. As a result, he, too, recommended the keeping of bees in cotton fields for their pollination service. Later, Kearney (1923) concluded that Pima cotton production in the Salt River Valley of Arizona could be substantially increased (about 32 percent) if honey bees were kept around the cotton fields.

The rather comprehensive and convincing research by Shishikin (1946, 1952) apparently awakened his countrymen to the value of bees to cotton. His work was followed by that of numerous workers but especially certain leaders in this work: Kaziev (1955, 1956a, b, 1958, 1959a, b, 1960, 1961a, b, 1963,1964, 1967), Kuliev (1958), Minkov (1953a, b), Radoev (1963, 1965), Radoev and Bozhinov (1961), Skrebtsov (1964), and Trushkin (1956,1960a, b).

In Peru, Mercado Mesa (1956) concluded that insect pollination was of no value to 'Tanguis' cotton (G. barbadense). In a more thorough test Dulanto Bartra (1958) showed that 51.7 percent of the flowers shed if they were not visited by the bee Melitoma euglossoides Lepeletier and Serville compared to only 32.2 percent of flowers visited by this bee.

Shishikin (1946) was the first to use the term "saturation pollination" - the uniform distribution of colonies of honey bees among cotton fields. He showed that saturation pollination, at the rate of one- half colony per acre, increased production of cotton 19.5 percent more than areas dependent upon only local pollinators. The increase over cotton grown in cages, excluding all insects, was 43 percent. In his more complete report, Shishikin (1952) stated that 4,130 to 5,000 colonies were used, at the rate of one colony per hectare, and distributed in groups 1 km apart. The tests dealt with "the old variety, '114', and the new variety '1298'." During the tests, the natural crossing caused by the honey bees varied from 26 to 43 percent. He concluded that the wild pollinators were "far from being able to assure cross-pollination of cotton plants."

In a sense, Babadzhanov (1953) duplicated the test by Kearney (1923) and obtained a similar benefit from supplemental pollination. He reported that it increased the boll set of cultivar '108-F' by 30 percent, the raw cotton per boll by 5 to 10 percent, the seed germination from 93 percent in selfed seed to 98 percent in cross-pollinated seed, and decreased the motes by 12.5 percent. Ter-Avanesyan (1952) showed that crossing within the cultivar varied with the cultivar tested: 8.4 percent in 'Sreder', 14.4 in '8582', and 22 percent in '915'.

These tests were supported in theory by Miravalle (1964), who compared the effects of bulked pollen from several plants with pollen from one flower of the same selection. He found that 76 percent of the bolls set, with 34.47 viable seeds per boll, when the flowers were pollinated with bulked pollen, but only 70 percent set, with only 27.07 seeds per boll, when the flowers were pollinated with pollen from one flower of the same plant.

McGregor et al. (1955) studied the effect of bee pollination upon upland and Pima cotton in cages, some of which contained a colony of honey bees (figs. 92, 93). In their test, 'Pima S-1' produced 24.5 percent more cotton in cages with bees than in cages without bees. This increase was caused by the set of more bolls, with more seeds per boll. The presence of bees did not increase total production of upland 'A-33' or 'A- 44', but the crop set earlier. In an area with a short season, this effect would doubtless be reflected in a greater total yield. Also, the cotton was handpicked, with extreme care taken to collect every seed. If machine harvesting had been used, doubtless more cotton would have been collected where there were fewest motes - in the bee cages. The lock usually breaks if motes are prevalent, and the remaining lint remains unharvested in the base of the burr.

In this test and others, McGregor and Todd (1955, 1956) noted that a decided difference developed before the end of day in the appearance of the petals of the upland cotton flowers visited by bees as compared to those not visited. By mid-afternoon, the visited ones began to change color and wilt, and the petals formed a tight cylindrical roll; whereas, in those flowers not visited by bees, the petals stayed white and open until sunset then became limp and adhered to each other like pieces of wet paper (fig. 94).

Mahadevan and Chandy (1959), using cultivars 'M.U.I.' and 'M.C.U.2' in India, obtained 23 to 34 percent and 40 to 53 percent, respectively, more cotton in open plots than in plots caged to exclude bees. They did not have plots caged with pollinating insects. This leaves unexplained the possible effect of the caging on the plant. Sidhu and Singh (1962), also in India, compared production in cages with Apis indica [cerana] and A. florea and in cages without pollinators and obtained an increase of 17.45 to 18.98 percent in favor of the pollinating insects. The increase was attributed to more and larger bolls.

In the United Arab Republic, Wafa and Ibrahim (1960) also obtained 22.4 percent more 'Ashmouni' cotton with honey bee pollination.

Skrebtsov (1964) obtained 33 percent increase in raw cotton by cross-pollination within the variety with honey bees, and showed that the bees improved hybrid vigor.

Radoev and Bozhinov (1961) obtained 10.6 to 24.4 percent greater yield from flowers freely visited by bees than from flowers tied to exclude insects during flowering. There were 0.5 more seeds per boll, fewer motes, and better seed germination. Radoev (1963) stated that freely pollinated plants set 11.04 percent more cotton than isolated plants, with more sound seed and better germination. Radoev (1965) concluded that the honey bee is the most important insect in Russia in the pollination of cotton, even though only 18 percent of the floral visits contribute to pollination.

Minkov (1953b) studied the effect of pollination on Russian cultivars '611-b' and '108-F', which were visited primarily by wild bees. He found that exclusion of pollinators increased the number of motes.

These tests indicate that the value of insect pollination is not limited to any particular area, species, or cultivar.

The material in the reports by the various authors previously mentioned as well as numerous others, leaves little doubt that cotton is benefited by bees in terms of greater lint and seed production, earliness of harvest, fewer motes, better lint, better germination, and improved qualities in the offspring.

Trushkin (1960a) concluded that the use of bees on cotton must be considered not only possible but expedient. Trushkin (1960b) stated, "The time has come to fully exploit [utilize] honey bees for purposes of obtaining high cotton yields and improving seed quality ...."

[gfx]
FIGURE 92. - Pollination studies involving honey bees on caged cotton plants.
FIGURE 93. - Author taking data for pollination studies on tagged Pima cotton flowers.
FIGURE 94. - Effect of insect pollination on cotton flowers. Photograph taken 4:30 p.m. of the day these 2 flowers opened. Flower on left was in a bee cage. Its petals had changed from cream to pink and tightened into a tubular shape. Flower on right was in a no-bee cage. Its petals remained cream-colored and flared until sunset.

HYBRID VIGOR IN COTTON:

The accentuated effect of cross-fertilization is referred to as heterosis or hybrid vigor. It can result from interspecific (between species), intraspecific (within species), or intervarietal crossing. The possible utilization of hybrid vigor in cotton has been of considerable interest to cotton breeders since Mell (1894) showed that some cotton hybrids exceeded their parents in certain characteristics. Although Kottur (1928) concluded that selfing of G. herbaceum plants for 12 generations had no injurious effect, and Harland (1943) proposed the breeding of a cotton immune from natural crossing, Brown (1942) and O'Kelly (1942) independently concluded that inbreeding of upland cotton reduced production and caused fewer flowers and smaller bolls. Simpson and Duncan (1953) also showed that cultivars selfed for 10 years produced 15 percent less cotton than the original plants. As a result, breeders now generally agree that too much inbreeding is detrimental. Instead, they strive for or desire some degree of outcrossing but, so far, have not been able to control it. The subject of hybrid vigor in cotton was thoroughly reviewed by Loden and Richmond (1951).

Hybrid vigor in cotton has been observed in interspecific crosses as well as in crosses between varieties within the species. Fryxell et al. (1958), Hutchinson et al. (1938), Marani (1967), Stroman (1961), and Ware (1931) in particular showed that crosses between G. barbadense and G. hirsutum were much more productive than either parent. Because of the differences in the characteristics of the lint of the two species, it frequently has objectionable qualities in the hybrid. This problem is less likely to arise in intraspecific hybrids where considerable hybrid vigor has also been shown.

Kime and Tilley (1947) made numerous crosses of commercial cultivars of upland cotton and showed an increase in production of the F₁ ranging from 7 to 20 percent. They doubted that hand cross-pollination to produce the hybrid seed was practical but considered production feasible in areas where "a high percentage of crossing normally occurs." Kohel and Richmond (1969) showed that significant heterosis could be obtained in areas of high natural crossing. Patel and Patel (1952) indicated that in India hand-pollination might be practical. Thakar and Sheth (1955) proposed that the government subsidize hybrid cotton seed production. Simpson (1948) obtained increases in yield from crosses of upland cultivars ranging from 5.7 to 44.2 percent. He, like Meade (1918) and Kearney (1923), also recommended that honey bee colonies be placed around the fields, knowing that hybrid seed could only be produced in quantity with pollinating insects (natural crossing) (Simpson 1954b).

Hybrid vigor has been shown within upland cotton by numerous workers (Barnes and Staten 1961, Christidis 1955, Galal et al. 1966, Hawkins et al. 1962, Lee et al. 1967, and Turner 1953a, b ).

Miller and Lee (1964) reported larger bolls with higher lint yields. Muramoto (1958) showed increases in yield, lint percentage, lint index, and seeds per boll. Ter-Avanesyan and Lalaev (1954) reported yield increase, bolls ripening 5 to 6 days earlier, and some resistance to Verticillium wilt. Trushkin and Truskina (1964) also reported that supplementary pollination by bees increased resistance to wilt, fungal root rot, and Xanthomonas maluacearum. Kaziev (1961b) reported earlier germination of hybrid seed and wilt resistance.

Wanjura et al. (1969) showed the importance of early emergence of the seed to plant survival and yield. For plants emerging on the fifth, eight, and 12th days, the survival was 87, 70, and 30 percent, and the relative yield of the plants was 100, 46, and 29 percent respectively.

The value of the hybrids is not in the mere mixing of plant types in the field, such as might occur in certain breeding programs without insect pollination. Richmond and Lewis (1951) showed that from the standpoint of yield nothing was gained by growing a mixture of seed types in a pure stand although such a mixture, they indicated, might supply a mixture of fibers not obtainable from a lone commercial cultivar.

A method tried by Peebles (1956) consisted of planting alternate rows of 'Pima 32' and 'Pima S-1' and the use of saturation pollination by honey bees. He showed the economic feasibility of this method, but it was not accepted by the industry. An attempt was made by a commercial company (DeKalb Agricultural Association, Inc. 1961) to produce hybrid cotton, but their supply of pollinating insects was apparently inadequate. Instead of attempting to increase the local supply, they concluded that hybrid cotton seed must be produced "in marginal cotton growing areas where bee activity is great." Turner (1959) proposed the planting of appropriate male-sterile and normal-functioning flower seed mixtures, then reaping the benefit of the hybrid vigor caused by thorough cross- pollination between the plants by a high population of honey bees.

One method frequently proposed (Christidis and Harrison 1955) for utilizing hybrid vigor involved male-sterile plants. Allison and Fisher (1964), Fisher (1961), Justus and Leinweber (1960), Justus et al. (1963), Meyer and Meyer (1965), Turner (1948), and Weaver and Ashley (1971) have reported the presence of male sterility in cotton. At one time, the creation of male-sterile plants with a chemical "gametocide" (Sodium 2, 3-dichloroisobutyrate) looked promising (Eaton 1957, McGregor 1958, Meyer et al. 1958, and Rohm and Haas 1958); however, subsequent testing failed to establish its reliability. It also created some female sterility; therefore, its use was discontinued.

Stith proposed (1970) the use of cytoplasmic male-sterile stocks and restorer genes in cotton cross-pollinated by bees. Kohel and Richmond (1962) showed that bees should function satisfactorily in the production of cotton on male-sterile plants. Meyer (1969), commenting on the progress made with cytoplasmic male sterility, stated that the basic plant work has been done, but the bee breeders and their bees still have a lot of work to do. She concluded that the most critical problem in the production of hybrid cotton appeared to be in finding some way to get the male-sterile flowers pollinated.

Hybrid vigor in cotton offers possibilities for increasing cotton production to a new plateau, if insects can be used as the cross- pollinators. Tests mentioned herein show this is possible. The problem is to find the proper cotton combiners and the best utilization of pollinating insects.

Pollinators:
There is agreement that cotton is not wind pollinated (Balls 1915), that all pollen transport outside of the flower requires an active vector, and that "bees" are the best pollinators of cotton. Tsyganov (1953) stated that "bees on the cotton flower are not guests but thoroughly adapted symbionts because they feed and rear their young on the products gathered from the flowers."

The bees most frequently mentioned are the bumble bees (Bombus spp.), honey bees (Apis dorsata, A. florea, A. indica [cerana], and, most frequently, A. mellifera), and the solitary groundnesting Melissodes spp. Other hymenoptera sometimes mentioned include Anthophora spp., Elis thoracica Lepeletier, Halictus spp., Megachile spp., Melitoma euglossoides Lepeletier and Serville, and Nomia spp. Numerous species from several other orders of insects sometimes find their way into cotton flowers, but as Simpson and Duncan (1956) stated, pollen distribution is essentially a "put and take" procedure, and unless the insect consistently visits large numbers of cotton flowers it is relatively ineffective as a pollinator.

In the United States, the bumble bee, honey bee, and Melissodes bees are considered most important as pollinators of cotton (Allard 1910, 1911a, b, Butler et al. 1960, Kearney 1923, McGregor 1959, McMillian 1959, Stephens and Finkner 1953, Theis 1953). In Russia, the honey bee is undisputedly considered the most important. In India, Apis spp., Elis thoracica, and Anthophora spp. have been mentioned (Sidhu and Singh 1961, Khan and Afzal 1950). In Egypt, the honey bee is most commonly seen (Wafa and Ibrahim 1957, 1959).

BUMBLE BEES:

Brown (1927) stated: "Large lubberly bumble bees that get pollen all over their bodies and rub against the stigma of every flower they meet are doubtless the best." With this there is no disagreement. Many of the cotton researchers concerned about insect visitation to cotton flowers in the Cotton Belt east of the Brazos River consider the bumble bee most important (Allard 1910, 1911a, Loden and Richmond 1951, Stephens and Finkner 1953, Theis 1953).

The visits to the plant by bumble bees are predominantly within the flower. Because of its size, the bumble bee can scarcely enter the flower without depositing pollen on the stigma and picking up more from the anthers. Because the nest is provisioned with both nectar and pollen, the bumble bee makes numerous collecting trips to the flowers. Also, bumble bees are colonial and under favorable conditions the population within the nest may increase so that numerous individuals from one nest will be foraging simultaneously in a field.

In isolated cotton test plots in North Carolina, it is not unusual to find a bumble bee on every plant and at times in every flower. Under such conditions, their effectiveness as pollinators could not be surpassed. Incidentally, Dulanto Bartra (1958) reported up to three or four Melitoma euglossoides in a single flower near the nesting sites, but they were very scarce farther away. By contrast, in the western half of the Cotton Belt, an entire day might be spent in a large cottonfield without seeing a bumble bee. Here, they are of no importance whatever as pollinators of cotton.

Bumble bees are colonial only through the active season. The nest is abandoned in the fall, the males die, and the females go into hibernation. Each female that survives hibernation establishes a new nest in the spring. Nests are only established if suitable nesting sites can be found. The colony in the nest then faces numerous hazards throughout the season, such as lack of a continuous source of fresh nectar and pollen, diseases, pests, pesticides, and other environmental factors or agricultural practices. For these reasons, bumble bees are not always present in adequate numbers when desired, and their numbers cannot be increased as desired. The culture of bumble bees for the pollination of cotton holds little promise.

MELISSODES BEES:

The Melissodes bee frequently constructs its nest in the soil in the cottonfield. It visits cotton flowers as a preferred host plant and rarely if ever visits extrafloral nectaries. It provisions each cell of its nest with a l/5-inch pellet of pollen and nectar. The female spends the night in her subterranean nest. She rapidly visits blossom after blossom of either upland or Pima cotton and will cross over from the one to the other on a single trip. Under natural conditions, a single bee may make as many as 200 floral visits in a day (Butler et al. 1960).

Melissodes bees are quite seasonal, therefore, they may be plentiful during one part of the flowering season but rare later in the same season. They are adversely affected by pesticides applied during the daytime, but the females may escape damage from nighttime applications. Little is known about the adverse effects on them of insecticides, soil cultivation, irrigation, or crop rotation. No way is known to increase these bees when desired.

HONEY BEES:

In contrast to bumble bees and Melissodes bees, honey bees show a preference for the extra-floral nectaries of cotton and often seem reluctant to enter the cotton flower. When a honey bee enters a cotton flower, it may emerge coated with pollen, then alight on a leaf, and comb much of the pollen off without attempting to pack it in the pollen baskets on the hind legs. However, all of this pollen is not removed, and a familiar sight, where bees are working cotton, is their incoming at the hive entrance coated with cotton pollen. Radoev (1965), in Russia, stated: "The honey bee is the most important insect in the pollination of cotton." As shown in other places herein, its value in the United States would appear to be in proportion to its use and concentration on the cotton.

At times, honey bees collect small amounts of cotton pollen and transport it to the hive (fig. 95). This usually occurs only when no other pollen is available for the bees. Minkov (1956) concluded that honey bees can collect cotton pollen but seldom do so. On the other hand, Kaziev (1956a, 1964) stated that 15 to 25 percent of the bees were collecting pollen when the average colony was storing 2 to 5 pounds of honey per day. The bees were collecting the pollen from 8 a.m. to 4 p.m. Whether he referred to their actually collecting the pollen in their pollen baskets or whether they were merely entering the hive with pollen on their bodies, indicating that they had been inside the cotton flowers, is not clear.

Pollen collection is not always dictated by supply and demand. Honey bee colonies have been observed in Arizona by the author (unpublished data) and Grout (1955) showing every evidence of pollen deficiency, although these colonies were surrounded by hundreds of acres of both species of cotton in flower. Later in the season, the bees in the same location collected cotton pollen freely. No reason could be determined for this strange behavior. At all times, the honey bees which were concentrated at the rate of one colony for each acre of cotton, were collecting both floral and extrafloral nectar.

In 1957 (unpublished data), I counted the honey bees in cotton flowers on each of five farms at Shafter, Calif., and four farms near Mettler Station, about 50 miles to the south. In both areas, the only cotton grown was 'Acala 4-42', and in both areas cultivation was large scale and dependent on irrigation water. Near Shafter, water was plentiful, and much of the land was devoted to cotton and alfalfa hay production. There were few apiaries near Shafter. Near Mettler Station, the water supply was acute; therefore, the growers devoted some of their land to alfalfa seed production, which required less water than cotton. The alfalfa seed fields and cotton fields were interspersed. About 50,000 colonies of honey bees had been transported into the Mettler area to pollinate about 20,000 acres of alfalfa seed. In the Shafter cotton fields, only 13 honey bees were observed in 1,000 cotton flowers (1.3 bees per 100 flowers), but at Mettler Station 158 were counted (15.8 bees per 100 flowers). One unidentified wild bee was seen.

All cotton fields within one-quarter mile of the alfalfa fields had 20 or more honey bees per 100 flowers. This proved that honey bee populations can be built up to provide "saturation pollination" or the "10 bees per 100 flowers" shown by McGregor (1959) to be sufficient to provide thorough coverage of the stigma with pollen.

In an 80-acre 'Pima S-1' cottonfield at mid-season in Arizona, Johansson (1959) dusted fluorescent pigmented particles in a single, newly opened flower. The following day he recorded the percentage of day-old (closed) flowers showing the presence of such particles (brought the previous day) in relation to distance from the treated flower. There were 212 colonies of honey bees within or along the borders of this field (fig. 96), but few other pollinating insects were active where the test was conducted. His results were: 0 to 50 feet (40.5 percent), 50 to 100 feet (14.0 percent), 100 to 150 feet (3.5 percent), and 150 to 200 feet (1.6 percent). This showed that when honey bees are present in sufficient numbers, they can effectively distribute such particles (and pollen grains) from flower to flower.

Kohel and Richmond (1962) concluded that a single insect visit to a cotton flower is not enough for complete pollination. Ter Avanesyan (1959) showed that 600 to 1,000 pollen grains on the stigma is the minimal "norm" quantity of pollen required. Shoemaker (1911) and Minkov (1953b) mentioned the commonly observed characteristic of the honey bee in alighting on the corolla rim, after which it crawls down the petal with its back to the anthers, rarely touching the stigma. However, when there is sufficient honey bee traffic into the cotton flower some of the individuals "get careless" and alight upon or crawl over the stigma, giving it a liberal coating of pollen in the process. The secret of successful pollination of cotton with honey bees seems to be in having sufficient visitation so that the bees are "forced" to visit the flowers, and considerable bee traffic into and out of the flower results.

Pollinating bees are an obstacle to most cotton breeders attempting to develop pure lines. Each cotton blossom from which the breeder desires seed must be enclosed or isolated in some way, otherwise the pollinating insect may dilute the line by bringing pollen to it from another type of cotton plant. This is particularly true if the breeder is using the individual plant selection method where he breeds offspring from a single plant. However, Brown (1942) and Simpson and Duncan (1953) have shown that continual inbreeding causes a decrease in productiveness not fully compensated by gains in other properties. As a result, Harland (1949a) proposed his mass pedigree selection system in which outcrossing within the mass is permitted, and Knight and Rose (1954) proposed a modification in which there is the initial selfing generation, then a selection of progeny rows, the seeds from which are bulked and grown in an isolated area with the highest pollinator population that can be obtained.

Thus, the pollinating insect changes from a "harmful insect in the development of the plant selection to "beneficial" insect in the later stages of the program but there still remains within the minds of many cotton specialists an aura of animosity toward these insects. This has been coupled with the fact that bees are not "necessary" in the production of cotton. The evidence strongly indicates that for the best interest of the grower the pollinating insects should be protected and their presence encouraged in breeding and seed increase programs and in the production of bulk cotton.

Although the use of honey bees for systematic or saturation pollination of cottonfields is practiced to considerable extent in Russia, resulting in increase and/or improved productivity, the American grower has tended to strive for his increased production through the use of pesticides and other agronomic practices and has given little heed to the beneficial insects. Some growers, after observing that areas near apiaries are frequently more productive, feel that bees are of some value. Others, fearing that their pesticide program might damage the bees and result in legal action by the beekeeper, discourage the keeping of bees near the cottonfields. The request for or the rental of bees to pollinate cotton in the United States is extremely rare.

The cotton plant may flower for 2 months or more; however, Buie (1928) showed that the majority of the flowers that set fruit appeared within 3 to 4 weeks.: the bees were concentrated on the cotton for this period with the use of harmful pesticides curtailed, the bees could perform their pollination service and escape pesticide damage.

[gfx]
FIGURE 95. - Honey bee collecting cotton pollen
FIGURE 96.- Honey bee colonies beside cotton field.

Pollination Recommendations and Practices:
Meade (1918), Kearney (1923), and Stephens and Finkner (1953) recommended the keeping of honey bees near cottonfields, but no ratio of bees per flower or colonies per area was indicated. Shishikin (1952), who studied the effect of bees on about 5,000 acres of cotton in Russia, recommended one colony per acre with the colonies grouped about 0.6 mile apart. McGregor and Todd (1955) suggested one colony per acre. Avetisyan (1965, ch. 5, pp. 209-248) suggested 0.5 to 1.0 colony per hectare (one colony per 2.5 to 5. acres). Glushkov and Skrebtsov (1960) stated that with 4.9 and 6.6 colonies per hectare (2.0 to 2.5 colonies per acre) the cotton production was increased 20.9 and 45 percent, respectively, over the control areas.

The colonies-per-acre ratio (from one-fifth of colony to five colonies per acre have been suggested) is doubtless influenced by the acreage involved, competing crops, and colony strength. The ratio of 10 bees per 100 flowers suggested by McGregor (1959) is a more realist ratio than colonies per acre. Quite probably, a low population of honey bees contributes little or nothing to pollination. A method of maintaining a high population should be considered when honey bees are used. Although not an ideal pollinator of cotton, the honey bee is the only pollinator that can be manipulated on cotton.

LITERATURE CITED:
ANONYMOUS.
1972. COTTON - WORLD STATISTICS. Internatl. Cotton Advisory Com. Quart. Bul. 25(9, part 2): 1 - 144, Washington, D.C.

AFZAL, M., and KHAN, H.
1950. NATURAL CROSSING IN COTTON IN WESTERN PUNJAB. Agron. Jour. 42: 14 - 19, 89 - 93, 202 - 205, 236-238.

______and TROUGHT, T.
1934. MOTES IN COTTON. I. PUNJAB-AMERICAN COTTON. Indian Jour. Agr. Sci. 4: 554-575.

AL-JIBOURI, H. A.
1960. NATURAL CROSSING IN COTTON IN IRAQ. Agron. Jour. 52: 53.

ALLARD, H. A.
1910. PRELIMINARY OBSERVATIONS CONCERNING NATURAL CROSSING IN COTTON. Amer. Breeders' Mag.(4): 247 - 261.

______ 1911a. SOME EXPERIMENTAL OBSERVATIONS CONCERNING THE BEHAVIOR OF VARIOUS BEES IN THEIR VISITS TO COTTON BLOSSOMS. Amer. Nat. 45(538): Part I, 607 - 622; Part II, 668 - 685.

______ 1911b. PRELIMINARY OBSERVATIONS CONCERNING NATURAL CROSSING IN COTTON. Amer. Breeders' Assoc. Mag. 6: 156 - 170.

ALLISON, D. C., and FISHER, W. D.
1964. A DOMINANT GENE FOR MALE-STERILITY IN UPLAND COTTON. Crop Sci. 4: 548. ARUTIUNOVA, L. G. 1940. [GERMINATION AND DEVELOPMENT OF COTTON POLLEN IN INTRAVARIETAL CROSS-POLLINATION.] Jarovizacija 1(28): 18-22. [In Russian.]

______and GUBANOV, G. IA.
1950. [IN REGARD TO THE BIOLOGY OF COTTON FERTILIZATION.] Agrobiologiya 6: 94-99. [ln Russian. ] Biol. Abs. 25(4): 31994, 1951.

______and KANAS, M.
1955. [FACTORS WHICH LESSEN THE FORMATION OF MOTES IN COTTON.] Khlopkovodstvo 5(8): 43 - 47. [In Russian.]

AVETISYAN G. A.
1965. [BEE POLLINATION OF AGRICULTURAL CROPS.] His Pchelovodstvo, Moskva, "Kolos," pp. 209-248. [In Russian.]

BABADZHANOV, F.
1953. [ROLE OF INTRAVARIETAL POLLINATION IN INCREASING THE PRODUCTIVITY OF COTTON.] Khlopkovodstvo 3(7): 34 - 39. [In Russian.]

BALLS, W. L.
1912. THE COTTON PLANT IN EGYPT. 202 pp. Macmillan and Co., Ltd., London.

______ 1915. THE DEVELOPMENT AND PROPERTIES OF RAW COTTON. 221 pp. Black Ltd., London.

BARNES, C. E., and STATEN, G.
1961. THE COMBINING ABILITY OF SOME VARIETIES AND STRAINS OF GOSSYPIUM HIRSUTUM. N. Mex. Agr. Expt. Sta. Bul. 457, 33 pp.

BECKETT, R. E., and HUBBARD, J. W.
1932. THE SHEDDING OF 4-LOCK AND 5-LOCK BOLLS IN UPLAND COTTON. U.S. Dept. Agr. Tech. Bul. 277, 16 pp.

BEDFORD, H. W.
1921. THE ASAL OF COTTON AND ITS CAUSES IN THE SUDAN. Gordon Mem. Col. Khartum, Wellcome Res. Lab. Ent. Sect. Bul. 17, 8 pp.

BENTON, R.
1937. IN THE LAND OF COTTON. Bees and Honey 18: 364 - 365, 378 - 379.

BHAT, N. R.
1955. SOME POINTS DESERVING CONSIDERATION IN THE BREEDING AND PURE SEED MULTIPLICATION OF COTTONS IN INDIA. Indian Cotton Crowing Rev. 9(4): 189 - 196.

BROWN, H. B.
1927. VICINISM OR NATURAL CROSSING IN COTTON. Miss. Agr. Expt. Sta. Tech. Bul. 13, 14 pp.

______ 1938. COTTON. 582 pp. McGraw-Hill Book Co., Inc., New York.

______ 1942. RESULTS FROM INBREEDING UPLAND COTTON FOR A 10-YEAR PERIOD Amer. Soc. Agron. Jour. 34: 1084-1089.

BUIE, T. S.
1928. THE FRUITING HABITS OF THE COTTON PLANT. Amer. Soc. Agron. Jour. 20: 193 - 201.

BUTLER, G. D., JR., LOPER, G. M., McGREGOR, S. E., and others.
1972. AMOUNTS AND KINDS OF SUGARS IN THE NECTARS OF COTTON (GOSSYPIUM SPP.) AND THE TIME OF THEIR SECRETION. Agron. Jour. 64: 364 - 368.

______TODD, F. E., McGREGOR, S. E., and WERNER, F. G.
1960. MELISSODES BEES IN ARIZONA COTTON FIELDS. Ariz. Agr. Expt. Sta. Tech Bul. 139, 11 pp.

CHRISTIDIS, B. G.
1955. HYBRID VIGOR EFFECTS WITH COTTON. Jour. Genet. 53(2): 225-231.

______and HARRISON, G. J.
1955. COTTON GROWING PROBLEMS. 633 pp. McGraw-Hill Book Co., Inc., New York and London.

COOK, O. F.
1921. IMPROVEMENTS IN COTTON PRODUCTION. U.S. Dept. Agr. Cir. 200, 12 pp.

______ 1932. COTTON IMPROVEMENT THROUGH TYPE SELECTION. U.S. Dept. Agr. Tech. Bul. 302, 62 pp.

DEKALB AGRICULTURAL ASSOCIATION, INC.
1961. DEKALB COTTON SEED. THE COTTON WITH THE BEE IN ITS BREEDING. DeKalb Agr. Assoe., Inc., Athens, Ga., 11 pp.

DULANTO BARTRA A.
1958. [THE IMPORTANCE OF THE "BUMBLE BEE", MELITOMA EUGLOSSOIDES LEP. & SERV. IN THE POLLINATION OF THE FLOWERS OF THE TANGUIS COTTON PLANT.] Rev. Peruana de Ent. Agr. 1(1): 6-11. [In Spanish.]

DUNLAP, A. A.
1945. FRUITING AND SHEDDING OF COTTON IN RELATION TO LIGHT AND OTHER LIMITING FACTORS. Tex. Agr. Expt. Sta. Bul. 677, 104 pp.

EATON, F. M.
1957. SELECTIVE GAMETOCIDE OPENS WAY TO HYBRID COTTON. Science 126: 1174 - 1175.

______and ERGLE, D. R.
1953. RELATIONSHIP OF SEASONAL TRENDS IN CARBOHYDRATE AND NITROGEN LEVELS AND EFFECTS OF GIRDLING AND SPRAYING WITH SUCROSE AND UREA TO THE NUTRITIONAL INTERPRETATION OF BOLL SHEDDING IN COTTON. Plant Physiol. 28(3): 503-520.

EWING, E. C.
1918. A STUDY OF CERTAIN ENVIRONMENTAL FACTORS AND VARIETAL DIFFERENCES INFLUENCING THE FRUITING OF COTTON. Miss. Agr. Expt. Sta. Tech. Bul. 8, 93 pp.

FIKRY, M. A.
1931. NATURAL CROSSING IN COTTON. Roy. Agr. Soc. Egypt, Tech. Sect. Bul. 18, 23 pp.

FINKNER, M. D.
1954. THE EFFECT OF DUAL POLLINATION IN UPLAND COTTON STOCKS DIFFERING IN GENOTYPE. Agron. Jour. 46(3): 124 - 128.

FISHER, W. D.
1961. MALE-STERILE COTTON FOUND. Prog. Agr. in Ariz. 13(2): 11.

FRYXELL P. A.
1957. MODE OF REPRODUCTION OF HIGHER PLANTS. Bot. Rev. 23: 135 - 233.

______STATEN, G., and PORTER, J. H.
1958. PERFORMANCE OF SOME WIDE CROSSES IN GOSSYPIUM. N. Mex. Agr. Expt. Sta. Bul. 419, 15 pp.

GALAL, H. E., MILLER, P. A., and LEE, J. A.
1966. HETEROSIS IN RELATION TO DEVELOPMENT IN UPLAND COTTON. Crop Sci. 6: 555.

GLUSHKOV, N. M., and SKREBTSOV, M. F.
1960. INCREASING THE COTTON CROP THROUGH SATURATION POLLINATION BY BEES.] PchelovodstVo 37(7): 29 - 30. [ln Russian.]

GORE, U. R.
1932. DEVELOPMENT OF THE FEMALE GAMETOPHYTE AND EMBRYO IN COTTON. Amer. Jour. Bot. 19: 795 - 807.

GROUT, R. A.
1955. HONEY BEES MAKE HYBRID COTTON POSSIBLE. Amer. Bee Jour. 95: 10-11.

GULATI, A. N., and TURNER, A. J.
1928. A NOTE ON THE EARLY HISTORY OF COTTON. Indian Cent. Cotton Com. Tech. Bul. 17, 10 pp.

GUSEINOV, A. M., and MUKHTAROV, Z. M.
1963. [NEW DATA ON THE EFFECTIVENESS OF INTRAVARIETAL HYBRIDIZATION OF COTTON.] Agrobiologiya 4: 490 - 496. [In Russian. ] Biol. Abs. 45(7): 57028, p. 4681. 1964.

HADWICH, F.
1961. [HONEYDEW ON COTTON.] Melliand Textilber. 42(5): 487 - 490. [In German.] AA-603/64.

HARLAND, S.C.
1943. BREEDING A COTTON IMMUNE FROM NATURAL CROSSING. Nature 151: 307.

______ 1949a. METHODS AND RESULTS OF SELECTION EXPERIMENTS WITH PERUVIAN TANGUIS COTTON. Empire Cotton Growing Rev. 26(3): 163-174.

HARLAND S. C.
1949b. METHODS AND RESULTS OF SELECTION EXPERIMENTS WITH PERUVIAN TANGUIS COTTON Empire Cotton Growing Rev. 26(4): 247-255.

HAWKINS, B. S., PEACOCK, H. A., and BALLARD, W. W.
1962. HETEROSIS AND COMBINING ABILITY IN UPLAND COTTON- EFFECT ON YIELD. Crop Sci. 2: 543-546.

HUGHES, L. C.
1968. MOTES IN VARIETIES OF GOSSYPIUM BARBADENSE. Empire Cotton Growing Rev. 45(3): 266-280.

HUMBERT, E. P., and MOGFORD, J. S.
1927. VARIATION IN CERTAIN LINT CHARACTERS IN A COTTON PLANT AND ITS PROGENY. Tex. Agr. Expt. Sta. Bul. 349, 23 pp.

HUTCHINSON, J. B., PANSE, V. G., and GOVANDE, G. K.
1938. STUDIES IN PLANT BREEDING AND TECHNIQUE. IV, THE INHERITANCE OF AGRICULTURAL CHARACTERS IN 3 INTERSTRAIN CROSSES OF COTTON. Indian Jour. Agr. Sci. 8: 757 - 775.

IVANOVA-PAROISKAYA, M. I.
1950. [HONEY-BEARING COTTON IN UZBEKISTAN.] Pchelovodstvo 9: 479 - 483. [In Russian.]

______ 1956. [NECTAR IN COTTON IN UZBEKISTAN AND ITS POLLINATION BY BEES.] In Krishchunas, I. V., and Gubin, A. F., eds. Pollination of Agricultural Plants. Moskva, Gos. Izd-vo Sel'khoz. Lit-ry, pp. 24-44. [In Russian.]

IYENGAR, N. K.
1938. POLLEN TUBE STUDIES IN GOSSYPIUM. Jour. Genet. 37: 69-106.

JANKI, R. P., SINGH, R., PRAKASH, R., and HAQUE, M. F.
1968. FLORAL BIOLOGY OF COTTON (GOSSYPIUM SPP.). Indian Jour. Agr. Sci. 38(3): 528 - 535.

JENSEN W. A., and FISHER, D. B.
1968. COTTON EMBRYOGENESIS: THE ENTRANCE AND DISCHARGE OF THE POLLEN TUBE IN THE EMBRYO SAC. Planta 78(2): 158 - 183.

JOHANSSON, T. S. K.
1959. TRACKING HONEY BEES IN COTTON FIELDS WITH FLUORESCENT PIGMENTS. Jour. Econ. Ent. 52: 572 - 577.

JUSTUS, N., and LEINWEBER, C. L.
1960. A HERITABLE PARTIALLY MALE-STERILE CHARACTER IN COTTON. Jour. Hered. 51: 191-192.

MEYER, J. R., and ROUX, J. B.
1963. A PARTIALLY MALE-STERILE CHARACTER IN UPLAND COTTON. Crop Sci. 3: 428-429.

KALYANARAMAN, S. M., and SANTHANAM, V.
1957. A REVIEW OF RECENT TRENDS IN COTTON BREEDING TECHNIQUES. Indian Cotton Growing Rev. 11(2): 131 - 135.

KAZIEV, T. I.
1955. [ROLE OF BEES IN POLLINATION OF COTTON.] Klopkovodstvo 12: 39 - 40. [ In Russian. ] AA-472/61.

______ 1956a. [THE WORKING OF COTTON BY BEES IN WESTERN AZERBAIDZHAN.] Uchen. Zap. Kirovabad. Ped. Inst. (4): 153-156. [In Russian.] AA-210/61.

______ 1956b. [ACTITY OF HONEYBEES ON COTTON.] Sotsial. Sell Khoz. Azerbaidzhan 8: 36-39. [In Russian.] AA-933/63.

KAZIEV, T. I. 1958. [ADDITIONAL POLLINATION OF COTTON BY BEES.] Pchelovodstvo 35(8): 28 - 34. [In Russian.]

______ 1959a. [INFLUENCE OF FERTILIZATION AND WATERING ON NECTAR SECRETION OF COTTON.] Pchelovodstvo 36(1): 25 - 28. [In Russian.]

______ 1959b. [THE INFLUENCE OF CLIMATIC CONDITIONS ON THE PROCESS OF NECTAR SECRETION IN COTTON.] Pchelovodstvo 36(9): 37 - 39. [In Russian.] AA-342/60.

______ 1960. [THE INFLUENCE OF BEE POLLINATION ON THE QUALITY OF COTTON.] PchelovodstVo 37(7): 33 - 35. [In Russian.] AA-333/62.

______ 1961a. [NECTAR SECRETION IN COTTON AND THE ROLE OF BEES IN INCREASING ITS YIELD.] Report of the Inter-University Scientific Conference, pp. 68 - 69: The Universities' Contribution to Agriculture, Moscow. [In Russian.] AA-935/63.

______ 1961b [INFLUENCE OF BEE POLLINATION ON THE PROGENY OF COTTON VARIETIES.] Uchen. Zap. kirovabad ped. Inst. (8): 100 - 112. [In Azerbaidzhani, Russian summary.] AA-934/63.

______ 1963. THE NECTAR YIELD OF THE COTTON PLANT AND THE TASK OF BEES IN INCREASING ITS FERTILITY. In 19th Internatl. Apic. Cong. Proc., Prague, p. 57.

______ 1964. [PROBLEMS OF NECTAR SECRETION IN COTTON AND THE ROLE OF HONEY BEES IN INCREASING ITS YIELD.] 215 pp. Baku, AzerbaidEhan Gos. Izd. [In Russian.]

______ 1967. [SOME AGROTECHNICAL MEASURES EXPEDITING INCREASED NECTAR PRODUCTIVITY OF COTTON.] In 21st Internatl. Apic. Cong. Proc., College Park, Md., Aug. Pp. 452-454. [In Russian.] KEARNEY T. H. 1921. POLLINATION OF PIMA COTTON IN RELATION TO THE YIELD OF SEED AND FIBER. Jour. Hered. 12(3): 99 - 101.

______ 1923. SELF-FERTILIZATION AND CROSS-FERTILIZATION IN PIMA COTTON. U.S. Dept. Agr. Dept. Bul. 1134, 68 pp.

______ 1926. BAGGING COTTON FLOWERS TO PREVENT ACCIDENTAL CROSS- POLLINATION. Jour. Hered. 17: 273 - 279.

______and HARRISON, G. J. 1932. POLLEN ANTAGONISM IN COTTON. Jour. Agr. Res. 44: 191 - 226.

KHAN, H., and AFZAL, M.
1950. VICINISM IN COTTON. Indian Cotton Growing Rev. 4(4): 227-239.

KIME, P. H., and TILLEY, R. H.
1947. HYBRID VIGOR IN UPLAND COTTON. Amer. Soc. Agron. Jour. 39: 308 - 317.

KNIGHT, R. L., and ROSE, M. F.
1954. A NEW APPROACH TO THE TECHNIQUE OF COTTON SELECTION. Empire Cotton Growing Rev. 31(3): 162-170.

KOHEL, R. J.
1968. EFFECT OF GENOTYPE AND HETEROZYGOSITY OF POLLEN SOURCE AND METHOD OF APPLICATION ON SEED SET AND SEED AND FIBER DEVELOPMENT IN COTTON. Crop Sci. 8: 293-295.

______and RICHMOND, T. R.
1962. AN EVALUATION OF SEED-YIELD POTENTIAL OF COMPLETELY MALE-STERILE COTTON IN AREAS OF HIGH AND LOW NATURAL CROSSING. Agron. Jour. 54: 525 - 528.

______and RICHMOND, T. R.
1969. EVALUATION OF SYNTHETIC VARIETIES OF UPLAND COTTON DEVELOPED UNDER TWO LEVELS OF NATURAL CROSSING. Crop Sci. 9: 647-651.

______ 1921. CROSS-FERTILIZATION AND STERILITY IN COTTON. Agr. Jour. India, part I, 16(1): 52 - 59; part II, 16(4): 406 - 409.

______ 1928. CONTINUED SELF-POLLINATION IN COTTON. Nature 122: 314.

KULIEV, A. M.
1958. THE USE OF BEES TO INCREASE COTTON YIELD. In 17th Internatl. Apic. Cong. Proc. Rome. Pp. 65-66.

LEE. J. A., MILLER, P. A., and RAWLINGS, J. O.
1967. INTERACTION OF COMBINING ABILITY EFFECTS WITH ENVIRONMENT IN DIALLEL CROSSES OF UPLAND COTTON (G. HIRSUTUM L.). Crop Sci. 7: 477-481.

LINSKENS, H. F.
1964. POLLEN PHYSIOLOGY AND FERTILIZATION. 257 pp. Symposium at University of Nijmegen, The Netherlands. Aug. 1963. North Holland Publishing Co., Amsterdam.

LLOYD, F. E.
1920. ENVIRONMENTAL CHANGES AND THEIR EFFECT UPON BOLL- SHEDDING IN COTTON. N.Y. Acad. Sci. Ann. 29: 1-131.

LODEN, H. D., and RICHMOND, T. R.
1951. HYBRID VIGOR IN COTTON - CYTOGENIC ASPECTS AND PRACTICAL APPLICATIONS. Econ. Bot. 5: 387-408.

______LEWIS, C. F., and RICHMOND, T. R.
1950. THE EFFECTS OF TIME AND METHOD OF POLLINATION ON SEED SET IN AMERICAN UPLAND COTTON. Agron. Jour. 42: 560-564.

LUKEFAHR, M. J., and MARTIN, D. F.
1964. THE EFFECT OF VARIOUS LARVAL AND ADULT DIETS ON THE FECUNDITY AND LONGEVITY OF THE BOLLWORM, TOBACCO BUDWORM AND COTTON LEAFWORM. Jour. Econ. Ent. 57: 233 - 235.

______and RHYNE, C.
1960. EFFECTS OF NECTARILESS COTTON ON POPULATIONS OF THREE LEPIDOPTEROUS INSECTS. Jour. Econ. Ent. 53: 242-244.

MAHADEVAN, V., and CHANDY, K. C.
1959. PRELIMINARY STUDIES ON THE INCREASE IN COTTON YIELD DUE TO HONEY BEE POLLINATION. Madras Agr. Jour. 47: 23-26. 188

MARANI, A.
1967. HETEROSIS AND COMBINING ABILITY IN INTRASPECIFIC AND INTERSPECIFIC CROSSES OF COTTON. Crop Sci. 7: 519-522.

MCGREGOR, S. E.
1958. RELATION OF HONEY BEES TO POLLINATION OF MALE-STERILE COTTON. In Eleventh Cotton Impr. Conf. Proc.: 93-94, Houston.

____ 1959. COTTON-FLOWER VISITATION AND PO1LEN DISTRIBUTION BY HONEY BEES. Science 129: 97-98.

____ and TODD, F. E. 1955. COTTON IS BENEFITED BY BEES. Prog. Agr. in Ariz. 7(1): 12.

____ and TODD, F. E. 1956. HONEYBEES AND COTTON PRODUCTION. Gleanings Bee Cult. 84: 649 - 652, 701.

____ RHYNE, C., WORLEY, S., JR., and TODD, F. E.
1955. THE ROLE OF HONEY BEES IN COTTON POLLINATION. Agron. Jour. 47: 23-25.

MCMILLIAN, W. W.
1959. THE POTENTIAL POLLINATORS OF COTTON INMONOKOVA FLORENCE AND DARLINGTON COUNTIES,1955. SOUTH CAROLINA. 56 pp. M.S. Thesis, Clemson Agr. Col.

MCNAMARA, H. C., HOOTON, D. R., and PORTER. D. D.
1940. DIFFERENTIAL GROWTH RATES IN COTTON VARIETIES AND THEIR RESPONSE TO SEASONAL CONDITIONS AT GREENVILLE, TEXAS U.S. Dept. Agr. Tech. Bul. 710, 44 pp

MEADE, R. M.
1918. BEEKEEPING MAY INCREASE THE COTTON CROP. Jour. Hered. 9: 282 - 285.

MELL P. H.
1894. EXPERIMENTS IN CROSSING FOR THE PURPOSE OF IMPROVING THE COTTON FIBER. Ala. Agr. Expt. Sta. Bul. 56, 47 pp.

MEL'NICHENKO, A. N.
1963. [BEES THEMSELVES INCREASE THE NECTAR PRODUCTIVITY OF FLOWERS.] Pchelovodstvo 40(9): 32-35. tIn Russian.] Biol. Abs. 45(7): 62074, p. 5069, 1964.

MERCADO MESA, O.
1956. [AUTOGAMY AND THE EFFECT OF INSECTICIDES ON THE FERTILITY OF TANGUIS COTTON.] Lima Estac. Expt. Agr. de La Molina. lnforme Mens. 30 (346): 30- 36. [In Spanish.]

MEYER, J. R., and MEYER, V. G.
1961. ORIGIN AND INHERITANCE OF NECTARILESS COTTON. Crop Sci. 1: 167-169.

____ Roux, J. B., and THOMAS, R. O.
1958. HYBRID COTTON IS AIM OF RESEARCH PROJECT IN DELTA. Miss. Farm Res. 21(5): 2.

MEYER, V. G.
1969. SOME EFFECTS OF GENES, CYTOPLASM, AND ENVIRONMENT ON MALE STERILITY OF COTTON (GOSSYPIUM). Crop Sci. 9: 237-242.

____ and MEYER, J. R.
1965. CYTOPLASMICALLY CONTROLLED MALE STERILITY IN COTTON. Crop Sci. 5: 444.

MILLER, P. A., and LEE, J. A.
1964. HETEROSIS AND COMBINING ABILITY IN VARIETAL TOP CROSSES OF UPLAND COTTON, G. HIRSUTUM L. Crop Sci. 4: 646-649.

MINKOV, S. G.
1953a. [HONEYBEES AND COTTON.] Pchelovodstvo (8): 41-44. [In Russian. ] AA-210/54.

____ 1953b. [ON VARIETAL ENTOMOPHILOUS CHARACTERISTICS IN COTTON.] KhlopkovodstVo 3(7): 40-44. [In Russian.]

____ 1956. [NECTAR PRODUCTIVITY OF COTTON AND THE ROLE OF BEES IN CROSS-POLLINATION.] Trud. Kazakh. Op. Stants. Pchelovodstva 1: 109-150. [In Russian.] AA-363/68.

____ 1957. NECTAR PRODUCTIVITY OF COTTON IN SOUTH-KAZAKH REGION. Pchelovodstvo 34(12): 35-40. AA-190/59.

MIRAVALLE, R. J.
1964. A NEW BULKED-POLLEN METHOD FOR COTTON CROSS- POLLINATION. Jour. Hered. 55: 276-280.

MONOKOVA, F. N., and CHEBOTNIKOVA, K. M.
1955. [AN INCREASE IN THE NECTAR PRODUCTIVITY OF HONEY PLANTS BY MEANS OF VARIOUS FERTILIZERS.] Pchelovodstvo 8: 44 - 46. [In Russian.] Biol. Abs. 31: 2119, July, 1957.

MOUND, L. A.
1962. EXTRA-FLORAL NECTARIES OF COTTON AND THEIR SECRETIONS. Empire Cotton Growing Rev. 39: 254 - 261.

MURAMOTO, H.
1958. THE GENETIC COMBINING ABILITY OF CERTAIN VARIETIES OF GOSSYPIUM HIRSUTUM AS MEASUR] FOR AGRONOMIC AND SPINNING QUALITIES. 119 pp. Dissertation submitted to University of Arizona as partial fulfillment of requirements for Ph.D. degree, University of Arizona Library, Tucson.

O'KELLY, J. F.
1942. DEGENERATION WITHIN COTTON VARIETIES Agron. Jour. 34: 782 - 796.

PARKS, H. B.
1921. THE COTTON PLANT AS A SOURCE OF NECTAR. Amer. Bee Jour. 61: 391.

PATEL, G. B., and PATEL, C. I.
1952. LONG STAPLE COTTON FROM HYBRID SEED OF CROSS BETWEEN HIRSUTUM AND BARBADENSE SPECIES. Indian Cotton Growing Rev. 6(4): 205-210.

PEARSON, N. L.
1949a. MOTE TYPES IN COTTON AND THEIR OCCURRENC AS RELATED TO VARIETY, ENVIRONMENT, POSITION IN LOCK, LOCK SIZE AND NUMBER OF LOCKS PER BOLL. U.S. Dept. Agr. Tech. Bul. 1000, 37 pp.

____ 1949b. FALSE MOTES IN COTTON, THEIR ORIGIN, DESCRIPTION AND VARIATION IN NUMBER. Jour. Agr. Res. 78: 705 - 717.

PEEBLES, R. H.
1942. PURE SEED PRODUCTION OF EGYPTIAN-TYPE COTTON. U.S. Dept. Agr. Cir. 646, 20 pp.

____ 1956. FIRST ATTEMPT TO PRODUCE HYBRID COTTONSEED. Amer. Bee Jour. 96: 51 - 52, 75.

PORTER, D. D.
1936. POSITIONS OF SEEDS AND MOTES IN LOCKS AND LENGTHS OF COTTON FIBERS FROM BOLLS BORNE AT DIFFERENT POSITIONS ON PLANTS AT GREENVILLE, TEXAS. U.S. Dept. Agr. Tech. Bul. 509, 13 pp.

RADOEV, L.
1963. STUDIES ON BEE POLLINATION AND HONEY PRODUCTIVITY OF COTTON. In 19th Internatl. Apic. Cong. Proc., Liblice, Czechoslovakia. P. 99.

____ 1965. [BEE POLLINATION OF COTTON.] Pchelovodstvo 9: 39 - 41. [In Russian.]

____ and BOZHINOV, M. 1961. [A STUDY ON THE NECTAR SECRETION OF COTTON AND THE ROLE OF BEES IN ITS POLLINATION.] Izv. Kompl. sel. lnst. Chirpan 1: 87 - 108. [In Bulgarian, English summary.]

REA, H. E.
1928. LOCATION OF MOTES IN THE UPLAND COTTON LOCK. Amer. Soc. Agron. Jour. 20: 1064-1068.

____ 1929. VARIETAL AND SEASONAL VARIATION OF MOTES IN UPLAND COTTON. Amer. Soc. Agron. Jour. 21: 481 - 486, 1154-1155.

REED, E. L.
1917. LEAF NECTARIES OF GOSSYPIUM. Bot. Gaz. 63: 229-231.

RICHMOND, T. R., and LEWIS, C. F.
1951. EVALUATION OF VARIETAL MIXTURES OF COTTON. Agron. Jour. 43: 66-70.

RICKS, J. R., and BROWN, H. B.
1916. COTTON EXPERIMENTS. Miss. Agr. Expt. Sta. Bul. 178, 40 pp.

ROHM AND HAAS CO.
1958. FW-450 A SELECTIVE GAMETOCIDE. Information sheet, January. Washington Square, Philadelphia.

ROSE, M. F., and HUGHES, L. C.
1955. PROGRESS REPORT OF THE COTTON BREEDING STATION, SEASON OF 1954-55. Cesira Research Station, Wad Medani, pp. 2-12.

SAAKYAN, T. A.
1962. [AN EMBRYOLOGICAL STUDY OF THE PROCESS OF FERTILIZATION OF COTTON WITH DIFFERENT METHODS OF POLLINATION.] Biol. Nauki (AN Arm. SSR), 15(7): 59 - 65. [In Russian.] Biol. Abs. 45(6): 52655, p. 4273. 1964.

SAPPENFIELD, W. P.
1963. ESTIMATES OF NATURAL CROSSING IN UPLAND COTTON IN SOUTHEAST MISSOURI. Crop Sci. 3: 566.

SEN, K. R. D., and AFZAL, M.
1937. VARIATION IN THE CHARACTERS OF COTTON IN RELATION TO THE POSITION OF BOLLS ON THE PLANT. Indian Jour. Agr. Sci. 7(1): 35 - 47.

SHISHIKIN, E. A.
1946. [HONEY BEES IN THE SERVICE OF COTTON POLLINATION.] PchelovodstVo 23(5-6): 31-32. [In Russian.] Biol. Abs. 21(2): 26151, p. 2551, 1947.

SHISHIKIN, E. A.
1952. [EFFECT OF POLLINATION BY HONEYBEES ON INCREASING THE PRODUCTIVITY OF COTTON.] In Krishchunas, I. V., and Gubin, A. F., eds. [Pollination of agricultural plants.]. Moskva, Gos. Izd-vo Selkhoz Lit-ry, pp. 95 - 103. [In Russian.]

SHOEMAKER, D. N.
1911. NOTES ON VICINISM IN COTTON IN 1908. Amer. Breeders' Assoc. 6: 252 - 254.

SIDHU, A. S., and SINGH, S.
1961. STUDIES ON AGENTS OF CROSS POLLINATION IN COTTON. Indian Cotton Growing Rev. 15: 341-353.

____ 1962. ROLE OF HONEYBEES IN COTTON PRODUCTION. Indian Cotton Growing Rev. 16: 18 - 23.

SIMPSON, D. M.
1948. HYBRID VIGOR FROM NATURAL CROSSING FOR IMPROVING COTTON PRODUCTION. Amer. Soc. Agron. Jour. 40: 970 - 979.

____ 1954a. NATURAL CROSS-POLLINATION IN COTTON. U.S. Dept. Agr. Tech Bul, 1094, 17 pp.

____ 1954b. "RADICAL METHODS" SEEK HYBRID COTTON VIGOR. Tenn. Farm and Home Sci. 9: 4.

____ and DUNCAN, E. N.
1953. EFFECT OF SELECTING WITHIN SELFED LINES ON THE YIELD AND OTHER CHARACTERS OF COTTON. Agron. Jour. 45: 275-279.

____ and DUNCAN, E. N.
1956. COTTON POLLEN DISPERSAL BY INSECTS. Agron. Jour. 48: 305 - 308.

SKREBTSOV, M F.
1964. [THE PROBLEM OF ABUNDANT POLLINATION OF COTTON BY HONEY BEES.l Trud. nauch.-issled Inst. Pchelov. 246-264. [In Russian, English summary.] AA-864/65.

STEPHENS, S. G.
1956. THE COMPOSITION OF AN OPEN POLLINATED SEGREGATING COTTON POPULATION. Amer. Nat. 90(850): 25 - 39.

____ and FINKNER, M. D.
1953. NATURAL CROSSING IN COTTON. Econ. Bot. 7(3): 257 - 269.

STITH, L. S.
1910. A BEE IN HYBRID COTTON PRODUCTION? In The Indispensable Pollinators, Ark. Agr. Ext. Serv. Misc. Pub. 127, pp. 28 - 32.

STROMAN, G. N.
1961. AN APPROACH TO HYBRID COTTON AS SHOWN BY INTRA AND INTERSPECIFIC CROSSES. Crop Sci. 1: 363-366.

TER-AVANESYAN, D. V.
1952. [BIOLOGY OF POLLINATION AND FLOWERING IN COTTON.] Bul. Appl. Bot., Genet. and Plant Breed. 29(2): 149-169. [In Russian.] Plant Breed. Abs. 24: 3155. 1954.

____ 1959. THE EFFECT OF QUANTITY OF POLLEN USED ON THE INHERITANCE OF CHARACTERS. Indian Jour. Genet. and Plant Breed. 19(1): 30-35

____ and LALAEV, G. B. 1954. [INTRAVARIETAL CROSSING OF COTTON PLANTS SOWN AT DIFFERENT TIMES.] Agrobiologiya 6: 57-61.1 In Russian.] Biol. Abs. 31(2): 15023, p. 1481, 1957. 190

THAKAR, B. J., and SHETH, D. S.
1955. COMMERCIAL PRODUCTION OF HYBRID SEED IN BOMBAY STATE. Indian Cotton Growing Rev. 9(3): 120-122.

THEIS, S. A.
1953. AGENTS CONCERNED WITH NATURAL CROSSING OF COTTON IN OKLAHOMA. Amer. Soc. Agron. Jour. 45: 481 - 484.

THOMSON, N. J.
1966. COTTON VARIETY TRIALS IN THE ORD VALLEY, NORTHWESTERN AUSTRALIA. Empire Cotton Growing Rev. 43(1): 18-21.

TRELEASE, W.
1879. NECTAR, WHAT IT IS AND SOME OF ITS USES. In Comstock, J. H., U.S. Dept. Agr. Rpt. on Cotton Insects, pt. 3, pp. 319-343.

TRUSHKIN, A. V.
1956. [WAYS OF IMPROVING THE HEREDITARY CHARACTERISTICS OF COTTON SEED WITH A NEW SYSTEM OF KOLKHOZ SEED RAISING.] Nauchn. Tr. Uzb. Sel'khoz. Inst. 9(1): 75-85. [In Russian.] Biol. Abs. 35(3): 27822, p. 2454, 1960.

____ 1960a. [THE USE OF BEES FOR THE POLLINATION OF COTTON.] Agrobiologiya 5: 787-788. [In Russian.] AA-335l62.

____ 1960b. [BEE POLLINATION INCREASES YIELD OF COTTON.] KhlopkovodstVo 10(12): 33-34. [In Russian.]

____ and TRUSKINA, L. P. 1964. [REPEATED POLLINATION BY BEES INCREASES VIGOR OF COTTON.] Khlopkovodsho 6: 30-31. [In Russian.] Plant Breed. Abs. 36: 2313, p. 308, 1966.

TSYGANOV, S. K.
1953. [REMARKS ON THE POLLINATING ACTIVITY OF HONEY BEES.] Uzbekistan Akad. Nauk Inst. Zool. i. Parasitol.i927 Trudy 1: 91-122. [In Russian.]

TURNER, J. H.
1948. HYBRID VIGOR STUDIES. Ga. Coastal Plain Expt. Sta. Ann. Rpt. Bul. 46, pp. 18-20.

____ 1953a. A STUDY OF HETEROSIS IN UPLAND COTTON. 1. YIELD OF HYBRIDS COMPARED WITH VARIETIES. II. COMBINING ABILITY AND INBREEDING EFFECTS. Agron. Jour. 45: 484 - 486, 487 - 490.

____ 1953b. DIFFERENTIAL RESPONSE OF COTTON VARIETIES TO NATURAL CROSSINGS. Agron. Jour. 45: 246-248.

____ 1959. HYBRID COTTON BREEDING PROGRAM. Calif. Agr. 13(3): 3, 15.

TURNER, J. H., and MIRAVALLE, R. J.
1961. 'HYBRID COTTON' FROM ACALA-4-42 COULD BOOST YIELD AND QUALITY BY 10% ON 60'S. Cotton Digest 33 (31): 18-19.

TYLER, F. J.
1908. THE NECTARIES OF COTTON. In Brown, E., Scofield, C. S., Hedgcock, G. G., an others, U.S. Dept. Agr. Bul. 131, pt. 5, pp. 45 - 56.

UNITED STATES DEPARTMENT OF AGRICULTURE.
1971. COTTON VARIETIES PLANTED 1967-1971. U.S. Dept. Agr. Consumer and Market. Serv.. Cotton Div., Memphis, 32 pp.

VANSELL, G. H.
1944a. COTTON NECTAR IN RELATION TO BEE ACTIVITY AND HONEY PRODUCTION. Jour. Econ. Ent. 37: 528-530.

____ 1944b. SOME WESTERN NECTARS AND THEIR CORRESPONDING HONEYS. Jour. Econ. Ent. 37: 530-532.

WAFA, A. K., and IBRAHIM, S. H.
1957. THE HONEYBEE AS AN IMPORTANT INSECT FOR POLLINATION. Cairo University, Faculty Agr., Bul. 162, 13 pp.

____ and IBRAHIM, S. H.
1959. POLLINATORS OF THE CHIEF SOURCES OF NECTAR AND POLLEN GRAIN PLANTS IN EGYPT. Soc. Ent. Egypt Bul. 43: 133 - 154.

____ and IBRAHIM, S. H.
1960. THE EFFECT OF THE HONEYBEE AS A POLLINATING AGENT ON THE YIELD OF CLOVER AND COTTON. Cairo University, Faculty Agr., Bul. 206, 44 pp.

WANJURA, D. F., HUDSPETH, E. B., JR., and BILBRO, J. D., JE
1969. EMERGENCE TIME, SEED QUALITY, AND PLANTIN DEPTH EFFECTS ON YIELD AND SURVIVAL OF COTTON (G. HIRSUTUM L.). Agron. Jour. 61: 63-65.

WARE, J. O.
1927. COTTON SPACING. Ark. Agr. Expt. Sta. Bul. 220, 80 pp.

____ 1931. COTTON BREEDING, SPACING, AND VARIETY STUDIES. Ark. Agr. Expt. Sta. 43d Ann. Rpt. for fiscal year ending June 30, 1931, pp. 30 - 31.

WEAVER, J. B., JR., and ASHLEY, T.
1971. ANALYSIS OF A DOMINANT GENE FOR MALE STERILITY IN UPLAND COTTON. Crop Sci. 11: 596 - 597.

WEBBER, H. J.
1903. IMPROVEMENT OF COTTON BY SEED SELECTION. U.S. Dept. Agr. Yearbook 1902, pp. 365-386.

WYKES, G. R.
1952. THE PREFERENCE OF HONEY BEES FOR SOLUTIONS OF VARIOUS SUGARS. Jour. Expt. Biol. 29: 511-518.

CROTALARIA
Crotalaria spp., family Leguminosae
The genus Crotalaria contains some 325 species, recognizable by their papilionaceous flowers, smooth leaves, erect growth, roundish pods, and yellow, brownish-yellow, blue, or purple flowers (Bailey 1949*). McKee and Enlow (1931) stated that C. mucronata Desv. [C. striata DC.] was the only one of commercial importance in the United States. Ritchey et al. (1941) tested 11 species for forage, and McKee et al. (1946) stated that four species (C. intermedia Kotschy, C. lanceolata E. Mey., C. mucronata, and C. spectabilis Roth) were extensively grown, primarily in the South (fig. 102). Wheeler and Hill (1957*) listed only two species - C. mucronata and C. spectabilis. Their justification for the growing of crotalaria was that it -

Makes excellent growth on light sandy soil, where it is sown for cover and green manure crops.

Does not harbor nematodes or rootknot.

Is well nodulated with nitrogen-fixing bacteria.

Is an annual (except in nonfreezing areas).

Sets seed in abundance.

The drawbacks to the cultivation of crotalaria are that it -

Contains an alkaloid that is poisonous to livestock or poultry (particularly the showy C. spectabilis).

Has a tendency to harbor certain insects harmful to citrus and pecans.

Has hard seeds that tend to germinate and come up years later and contaminate other crops.

Is treated as a noxious weed in at least one State if the seeds are harvested with corn or soybeans.

Because of these drawbacks and their significance at present, the crop is not being planted and the seed is almost unavailable.

Inflorescence:
Free (1970*), referring to work by Howard et al. (1919), stated that the large showy flowers of C. juncea L. occur in inflorescences, each flower having 10 stamens, five with short filaments and long narrow anthers, and five with long filaments and small round anthers. The long anthers dehisce in the bud, then the filaments of the round anthers elongate and push the pollen to the orifice of the keel. When a heavy insect alights on the wings, the pressure forces the style forward, and a ribbon of pollen is pushed out the orifice and onto the insect's abdomen. When the pressure is released, the style retracts through the mass of pollen, and another ribbon of pollen is extruded on the next insect's visit.

Roberts (1939) stated that crotalaria was a source of some nectar and an abundance of pollen. He did not explain how the bee collects the nectar. Pellett (1947*) stated that a colony of honey bees near Winter Haven, Fla., stored 50 pounds of dark and poor quality honey from C. mucronata and that the bees worked C. spectabilis to some extent. The effects of these visits on the flowers was not mentioned.

Pollination Requirements:
The pollination requirements of crotalaria are not too well understood. Todd (1957*) listed crotalaria as largely self-pollinated. Free (1970*) in discussing C. juncea stated that "When flowers are not visited by insects the continual elongation of the filaments presses the pollen masses onto the stigma so that self-pollination is possible. However, self-fertilization does not occur unless the stigmatic surface is rubbed against an insect's body, and lack of pollinators probably helps to explain why in parts of India few flowers set seed." A study of the effect of pollinating insects on the volume of seed produced by the different species of crotalaria and also on the germination of the seed would be of interest.

Pollinators:
Honey bees seem to be suitable pollinators.

Pollination Recommendations and Practices:
None.

LITERATURE CITED:
HOWARD A., HOWARD, G. L. C., and RAHMAN, K. A.
1919. STUDIES IN THE POLLINATION OF INDIAN CROPS. lndia Dept. Agr. Mem., Bot. Ser., 10: 195 - 220.

MCKEE, R., and ENLOW, C. R.
1931. CROTALARIA, A NEW LEGUME FOR THE SOUTH. U.S. Dept. Agr. Cir. 137, 30 pp.

MCKEE, R., STEPHENS, J. L., and JOHNSON, H. W.
1946. CROTALARIA CULTURE AND UTILIZATION. U.S. Dept. Agr. Farmers' Bul. 1980, 17 pp.

RITCHEY, G. E., MCKEE, R., BECKER, R. B., and others.
1941. CROTALARIA FOR FORAGE. Fla. Agr. Expt. Sta. Bul. 361, 72 pp.

ROBERTS, A. C.
1939. CROTALARIA. Amer. Bee Jour. 79: 84.

FEIJOA
Feijoa sellowiana Berg., family Myrtaceae
Feijoa is a subtropical plant native to South America but grown to a limited extent in Florida and California with occasional plantings of 1 to 2 acres (Schroeder 1949). Feijoa prefers a mild temperature with 30 to 40 inches of rain a year (Bailey 1952). Its product is the delectable fruit that is eaten fresh, out of hand, stewed, or processed into jam or jelly. Clark (1926) reported that 9-year-old plants were producing fruit at the rate of 10,000 lb/acre and had not reached maximum production.

Plant:
The evergreen plant reaches 15 to 18 feet with various shapes. The fruit is round, oval or oblong, 1 to 3 inches long, and dull green with a whitish bloom. The thin skin covers a whitish granular flesh about 1/4- inch thick around a translucent jellylike pulp in which 20 to 30 minute seeds are embedded. The flavor suggests pineapple or strawberry, and the aroma is delightful. The fruit is sometimes called pineapple guava.

The plants are generally spaced 15 to 18 feet apart. Irrigation is necessary for best production in California (Popenoe 1920).

Inflorescence:
The handsome flower, as described by Popenoe (1920), is 11/2 inches across with edible fleshy cupped petals that are white on the outside and purplish within. The long, stiff stamens form a conspicuous crimson tuft in the center, with the one stigma extending above the stamens (fig. 109).

[gfx] FIGURE 109. - Longitudinal section of feijoa flower, x 4.

Pollination Requirements:
Popenoe (1912, 1920) stated that some plants are self-sterile, whereas others are self-fertile, but even the self-fertile ones were not self-pollinating. Ryerson (1914) found in a preliminary test that six different seedling plants were self-sterile. Clark (1926) found one cultivar that apparently did not need cross-pollination, but transfer of pollen within the flower was necessary. Ryerson (1933) again stated that the feijoa tends toward self-sterility.

Pollinators:
Popenoe (1920) and later Ryerson (1933) stated that birds feeding upon the edible petals in their native habitat became dusted with pollen from the tuft of anthers then when they visited another flower they affected cross-pollination. Schroeder (1953) believed that transfer of pollen between plants either by insects or man was required.

Clark (1926) reported that honey bees were the only insects that visited the flowers, and they gathered no nectar, only pollen. Plants screened to keep bees off set only 9 percent of the flowers while those visited by bees set 40 percent of their flowers.

Schroeder (1947) conducted a thorough test on five cultivars under cheesecloth in Los Angeles during 1945 and 1946. He found that open pollinated flowers set about 16 times as well as comparable flowers under cage. He then and later (1953) concluded that the feijoa is pollinated by insects, primarily by bees, and that most cultivars "show markedly improved fruit-set when cross-pollinated. "

Pollination Recommendations and Practices:
The evidence indicates that the plants must be pollinated by bees unless the proper species of birds are available. No recommendations on methods of using bees have been made. Should the acreage increase the proper usage of bees would doubtless be necessary.

LITERATURE CITED:
BAILEY F. L.
1952. CULTURE OF FEIJOA TREES. New Zeal. Jour. Agr. 84: 291 - 293, 295 - 296.

CLARK, O. I.
1926. POLLINATION OF FEIJOAS. Calif. Avocado Assoc. Ann. Rpt. 1925-26: 94-95.

POPENOE, F. W.
1912. FEIJOA SELLOWIANA: ITS HISTORY, CULTURE AND VARIETIES. Pomona Col. Jour. of Bot. 2(1): 217 - 242.

POPENOE, W.
1920. MANUAL OF TROPICAL AND SUBTROPICAL FRUITS. 474 pp. The Macmillan Co., New York.

RYERSON, K. A.
1914. SUMMARY OF PRELIMINARY FEIJOA POLLINATION EXPERIMENTS. Calif. Univ. Jour. Agr. 2: 51-53.

______ 1933. THE FEIJOA. Natl. Hort. Mag. 12: 241 - 245.

SCHROEDER, C. A.
1947. POLLINATION REQUIREMENTS OF THE FEIJOA. Amer. Soc. Hort. Sci. Proc. 49: 161 - 162.

_______ 1949. THE FEIJOA IN CALIFORNIA. Fruit Varieties and Hort. Digest 4: 99 - 101.

______ 1953. THE POLLINATION OF SOME SUBTROPICAL FRUIT TREES. Lasca Leaves 3: 39-41.

FLAX
Linum usitatissimum L., family Linaceae
Flax is grown in the United States for its seed from which linseed oil is pressed and used primarily in paints. The byproduct, linseed meal, is fed to livestock. More than 2.6 million acres of flax were harvested in the United States in 1969, with an average yield per acre of 750 pounds. There were 1.5 million acres in North Dakota, 652,000 in South Dakota, 388,000 in Minnesota, 100,000 in Texas, 17,000 in Montana, and 3,000 in California. Production per acre within the States varied from 2,240 pounds in California to 700 pounds in North Dakota. The value of the crop was $92.9 million.

In addition to flaxseed, the plant is grown in other countries for its fibrous stalk from which linen is made. Economical U.S. production of fiber flax for linen is not feasible.

Plant:
Flax is a cool-weather crop but grows in warmer climates if it is planted in the fall and harvested at the beginning of the next summer. The plant is a slender annual with small linear leaves and a crown of flowers that later develop into seed pods (fig. 112). The seed flaxes are 15 to 30 inches tall, more branching, and produce more seed than the taller (30 to 48 inches) fiber flaxes.

An average plant produces two to six five-celled pods or capsules, although capable of producing many more. The pod normally produces two seeds in each cell and averages 8.6 seeds per pod (Kozin 1954).

The seeds are planted at the rate of 40 pounds per acre, which provides a high-density plant population necessary for maximum seed production.

[gfx] FIGURE 112. - Seed flax in bloom.

Inflorescence:
Flax flowers are borne on the branch terminals in many-flowered panicles. The petals may vary from white through hues of blue, pink, or lavender (Bailey 1949*). The five petals of the flower unfold at or soon after sunrise, depending upon the temperature, and shed before noon on clear warm days. Flowering continues for several weeks, depending upon soil moisture, but the peak occurs at the end of the first week. New flowers open each day (Knowles et al. 1959).

The five stamens are attached to a fleshy ring at the base of the flower (fig. 113). This ring secretes nectar from five small flat pits on its outer side opposite each stamen (Muller 1883*). The petals are also attached to this ring, alternating with the stamens. They narrow suddenly at the base leaving a round opening between the bases. The anthers of most cultivars are level with the stigmas on the five erect styles, but in some the styles are shorter and in others longer than the stamens (Yermanos and Kostopoulos 1970).

There are four types of flowers in our cultivated flax: the common funnel form, disk shaped with large flat petals, star shaped with narrow in-rolled petals, and tubular flowers.

The flowers are hermaphrodite and slightly protandrous (Eyre and Smith 1916), except for the pollenless male-sterile selections.

Flax provides a small amount of both pollen and nectar for honey bees, the degree of visitation and the material collected apparently depending upon the area and competing floral sources. For example, Scullen and Vansell (1942) considered flax a weak source of both pollen and nectar. Alex (1967) concluded that the bee collected only pollen, and Smirnov (1956) said they collected chiefly nectar. Pellett (1947*) stated the numerous bees are found on flax only where there are large number of colonies in the vicinity.

Smirnov (1956) stated that honey bees visited the blossoms from 7 a.m. to 1 p.m., with most intense visitation between 8 and 11 a.m. He concluded that, because the petals shed so easily soon after they open, particularly when a bee alights upon one, the bee "learns" to alight below the calyx and extend its proboscis between the petal bases from below to obtain the nectar. Such visitation contributes nothing to pollination, the contribution coming only before the bee adapts the new collecting stance.

[gfx] FIGURE 113. - Flax flower, x 3. A, Longitudinal section; B, flower with perianth removed to show nectaries.

Pollination Requirements:
Flax is considered to be normally self-pollinated although some crossing (1 to 6 percent) occurs (Dillman 1938, Dillman and Stoa 1935, Dillman and Goar 1937, Masuo 1958, Robinson 1937), mostly among the large flowered types. Rubis (1970) worked with a male-sterile line having disk-form flowers, and stated that he obtained practically no cross- pollination of the male-sterile lines with lines having tubular flowers; however, good seed set was obtained, indicating heavy cross-pollination, with other lines that had large disk-form flowers.

Self-pollination is influenced by the position of the anthers in relation to the stigmas. If the anthers are above or below the level of the stigmas, there is increased opportunity for cross-pollination (Yermanos and Kostopoulos 1970). In most commercial cultivars, however, they are on the same level. For a brief period after this type of flower opens, there is space between the anthers and the stigma. This permits cross- pollination to occur if foreign pollen is brought to the stigma. Whether or not cross-pollination occurs, the stamens soon bend inward so their pollen touches the stigma and self-pollination results. Flax is another example of plants that offer opportunity for cross-pollination then provide for selfing for survival of the species (Mtiller 1883*).

Several tests have shown that bee pollination improves seed yields in fiber flax. Bezdenezhoykh (1956) in Russia reported that honey bees in cages increased seed production of fiber flax 22.5 percent over plants in cages without bees. Gubin (1945) also studied the effect of bee pollination of fiber flax in Russia and reported that bees increased seed production 22.5 to 38.5 percent. Luttso (1957), also in Russia, reported that bee pollination increased seed production by 29 percent, the number of seeds per capsule by 18 percent, and the weight per seed by 11 percent in comparison to fields without bee pollination. Likewise, Smirnov (1956) showed a 19 percent increase in the number of seeds per capsule, a 22 percent increase in the total weight of seeds, and a 2.2 percent increase in the weight per seed. He also reported that bee-visited plants set up the crop and ceased blooming earlier than plants from which bees were excluded. The reason for the increased size of the seed in addition to seed number was not explained. Usually, when more seeds are produced by a plant, the size of the individual seed decreases.

The influence of bee pollination on oilseed flax is somewhat different. Hassanein (1955) reported that honey bee pollination increased both "quantity and quality" of seeds. However, Pritsch (1965) and Alex (1957) failed to show any benefit in terms of increased seed production, and insect pollination is generally considered unnecessary on U.S.-grown flax.

Shehata and Comstock (1971) discussed the potentials for increased production with hybrid vigor in flax. They obtained an average of 6 percent increase in production with hybrids over the highest yielding cultivar, and they stated that interest in hybrid flax is increasing. For the production of hybrid seed, the pollen must be transferred from the fertile to the male- sterile lines.

Pollinators:
Eyre and Smith (1916) pointed out that flax pollen is produced only in small quantities and is not the windblown type, so they concluded that cross-pollination was entirely by insects.

Dillman (1938) mentioned bumble bees as visitors along with honey bees, and Henry and Chih (1928) mentioned honey bees, a "small bee," and thrips, believing that the last-mentioned insects were important agents in cross-pollination in India. Hassanein (1955) attributed 90 percent of the crossing in flax to honey bees; Smirnov (1954), 26 to 93 percent; and Alles (1961), Bezdenezhnykh (1956), Luttso (1957), and Smirnov (1954) concluded that honey bees were the most important agents. When the honey bee collects pollen, it cannot fail to transfer pollen to the stigmas. This is also true when nectar is collected in the normal way.

Pollination Recommendations and Practices:
Alles (1961) concluded that the number of honey bee visits per flower of fiber flax has a determining effect on flax seed set and weight, but he gave no indication as to how many bees were needed. Kozin (1954) reported a sizeable increase in seeds per boll and seed weight when 40 colonies were placed near a fiber flax field, but he did not indicate the size of the field. Also, he stated that there were 226 bees per hectare (90 bees per acre), which seems to be an extremely low population for the number of flowers per acre. Gubin (1945) indicated that each flower of fiber flax should receive an average of two honey bee visits. This is the most concrete recommendation for the use of bees on flax. Whether it applies equally to fiber and seed flaxes is unknown.

There have been no recommendations for the placement and use of honey bees as pollinators of seed flax in this country. The evidence indicates, however, that if hybrid seed is produced insect pollinators will be needed, of which honey bees seem to be the best. The number of bee colonies that would be necessary in or around such a field to provide adequate pollination is unknown. Because the plant is not overly attractive to bees, the relative number of colonies needed would be large if competing plants were in bloom. The breeders might devote some attention to the relative attractiveness of cultivars with the thought in mind that if hybrid seed production materializes and bees are utilized the incorporation of lines having greater attractiveness could improve the efficiency of hybrid seed production.

LITERATURE CITED:
ALEX, A. H. 1957. POLLINATION OF SOME OILSEED CROPS BY HONEY BEES. Tex. Agr. Expt. Sta. Prog. Rpt. 1960, 5 pp.

ALLES, P. T.

1961. INFLUENCE OF THE MIXTURES OF FLOWER POLLEN, OF MICROELEMENTS AND OF VITAMIN B IN THE PROCESS OF FORMATION OF THE SEEDS OF DIFFERENT VARIETIES OF SUNFLOWER AND FLAX. In 18th Internatl. Apic. Cong. Proc., Madrid, Summaries of the Congress: 5-6. BEZDENEZHNYKH, S. M. 1956. POLLINATION OF FIBER FLAX AND TRAINING OF BEES. ] In Krishchunas, I. V., and Gubin, A. F., eds. Pollination of Agricultural Plants. Moskva, Gos. Izd-vo. Selkhoz Lit-ry, pp. 21-24. [In Russian.] DILLMAN, A. C. 1938. NATURAL CROSSING IN FLAX. Amer. Soc. Agron. Jour. 30: 279 - 286. ______and GOAR, L. G. 1937. FLAXSEED PRODUCTION IN THE FAR WESTERN STATES. U.S. Dept. Agr. Farmers' Bul. 1792, 22 pp. and STOA, T. E. 1935. FLAXSEED PRODUCTION IN THE NORTH CENTRAL STATES. U.S. Dept. Agr. Farmers' Bul. 1747, 18 pp. EYRE, J. V., and SMITH, G. 1916. SOME NOTES ON THE LINACEAE, THE CROSS POLLINATION OF FLAX. Jour. Genet. 5: 189-197. GUBIN, A. F. 1945. CROSS POLLINATION OF FIBRE FLAX. Bee World 26(4): 30, 31. HASSANEIN, M. H. 1955. THE VALUE OF POLLINATING INSECTS TO FLAX SEED PRODUCTION IN EGYPT. Agr. Sci. Ann. [Cairo]: 773 - 784. [In English, Arabic summary.] HENRY, A. W., and CHTH, Tu. 1928. NATURAL CROSSING IN FLAX. Amer. Soc. Agron. Jour. 20: 1183 - 1192 KNOWLES, P. F., ISOM, W. H., and WORKER, G. F. 1959. FLAX PRODUCTION IN IMPERIAL COUNTY. Calif. Agr. Expt. Sta.-Agr. Ext. Sen. Cir.480,27 pp. KOZIN, R. B. 1954. [INFLUENCE OF BEES ON THE SEED YIELD OF FIBER FLAX.] Pchelovodstvo 31 (6): 41-43. [ In Russian. ] AA-204/56. LUTTSO, V. P. 1957. [THE POLLINATION OF FLAX BY HONEY BEES.] Dokl TSKhA 30: 327-331. [In Russian.] AA-276/61. MASUO, Y. 1958. [ON NATURAL CROSSING IN FLAX.] Crop Sci. Soc. Japan Proc. 27: 321-323. [In Japanese.] Abstract in Plant Breed. 29(4): 817-818. Oct. 1959. PRITSCH, G. 1965. [INCREASING THE YIELD OF 0IL PLANTS BY USING HONEY BEES.] Ved. Prace vyzkum. Ustav. Vcelar CSAZV 4: 157-163. [In German.] AA-770/66. ROBINSON, B. B. 1937. NATURAL CROSS-PoLLINATION STUDIES IN FIBER FLAX. Amer. Soc. Agron. Jour. 29(8): 644 - 649. RUBIS, D. D. BREEDING INSECT-POLLINATED CROPS. In The Indispensable Pollinators, Ark. Agr. Ext. Sen. Misc. Pub. 127, pp. 19 - 24. SCULLEN, H. A., and VANSELL, G. H. 1942. NECTAR AND POLLEN PLANTS OF OREGON. Oreg. Agr. Expt. Sta. Bul. 412, 63 pp. SHEHATA A. H., and COMSTOCK, V E. 1971. HETEROSIS AND COMBINING ABILITY ESTIMATES IN F2 FLAX POPULATIONS AS INFLUENCED BY PLANT DENSITY. Crop Sci. 11: 534 - 535. SMIRNOV, V. M. 1954. [CROSS-POLLINATION OF FLAX BY BEES.] Pchelovodstvo (9): 53 - 55. [In Russian.] Abstract in Bee World 37: 124. ______ 1956. [CROSS-POLLINATION OF FIBER FLAX WITH THE AID OF BEES.] In Krishchunas, I. V., and Gubin, A. F., eds. Pollination of Agricultural Plants. MoskVa, Gos. Izd-vo. Selkhoz Lit-ry, pp. 16-21. [In Russian.] YERMANOS, D. M., and KOSTOPOULOS, S. S. 1970. HETEROSTYLY AND INCOMPATIBILITY IN LINUM GRANDIFL0RUM DESF. In The Indispensable Pollinators, Ark. Agr. Ext. Sen. Misc. Pub. 127, pp. 50 - 54.

MAMEY SAPOTE
Calocarpum sapota (Jacq.) Merr., family Sapotaceae
The mamey sapote, mamey, or sapote is native to Central America but is grown primarily as a dooryard planting in mildest temperature areas of California and Florida. (Also see "White Sapote," p. 380.) Campbell (1967) stated that it is easy to grow in southern Florida, where there is a ready market for its fruit. The related species (C. viride Pittier), the green sapote, has fruit 2 l/2 to 3 1/2 inches long, with brownish-green skin, sweet reddish-brown flesh, and a pleasant almondlike flavor (Whitman 1966).

Plant:
The mamey sapote is a large evergreen tree that may grow to 80 feet tall. The leaves are as much as 4 inches wide by 12 inches long. The russet-brown, ovoid fruit is 3 to 6 inches long. The somewhat granular, firm flesh is red to reddish brown and sweet. It may be eaten out of hand or used in preserves or sherbet (Kennard and Winters 1960*). There is usually one large seed. The plant is propagated by seeds. Mowry et al. (1967*) stated that old trees can withstand temperatures as low as 28deg F for several hours, but younger trees are quite susceptible to cold.

Inflorescence:
The l/2-inch, whitish, nearly sessile, bisexual flowers are produced in great numbers, six to 12 together in the axils of fallen leaves on old wood. The flower has five lobes to the corolla, five stamens alternating with five stamodia, a five-celled ovary, and a slender style with the stigma extending beyond the corolla (Bailey 1949*).

Pollination Requirements:
The construction of the flower indicates that insect pollination would aid in the setting of the fruit.

Pollination Recommendations and Practices:
None.

LITERATURE CITED:
CAMPBELL.C.W.
1967. THE MAMEY SAPOTE IN SOUTHERN FLORIDA. Fla State Hort. Soc. Proc. 80: 318-320.

WHITMAN,W. F.
1966. THE GREEN SAPOTE,A NEW FRUIT FOR SOUTH FLA. Fla. State Hort. Soc. Proc. (1965)78: 330-336.

MUSTARD
Brassica spp., family Cruciferae
Three species of mustard are grown commercially for their seed from which an excellent oil and the condiment, table mustard, is produced. They are (1) B. hirta Moench (B. alba (L.) Rabenh.), white or yellow mustard; (2) B. juncea Coss, Indian (oriental and brown types), Chinese, leaf, and trowse mustards and rai; and (3) B. nigra (L.) Koch, black mustard. (Also see "Cole Crops," p. 164, and "Rape," p. 315.)

Mustard is a minor crop in the United States, but in 1941, 124,000 acres were grown in Montana, with small amounts in North Dakota, Oregon, and Washington (Straw 1956). The crop in Canada is expanding with about 200,000 acres in 1970 versus 4 million acres of rape (Downey et al. 1970), and its importance is likely to increase in the United States. Black mustard is not grown in Canada, and the major acreage is devoted to yellow mustard.

Yields of 1,000 to 1,500 pounds of mustard seed per acre are obtained in Canada (Downey et al. 1970), which is more than the U.S. production of 468 to 714 pounds reported by Shaw (1956), the 451 to 524 pounds reported by Robinson (1964), or the 500 to 1,000 pounds of brown or 250 to 1,000 pounds of yellow mustard reported by Martin and Leonard (1949*).

Mustard and rape production is similar in many ways, however, each crop is a serious contaminant of the other. For that reason, the two crops should not be grown on the same farm and preferably not in the same area.

Plant:
Young mustard plants are quite similar to many other Brassica plants; however, differences soon begin to appear. The rosette of broad basal succulent leaves, about 1 foot high, produces the upright flowering stem that reaches 1 1/2 to 10 feet and terminates with the inflorescence. The upper leaves are much smaller than the lower ones and may be deeply lobed or entire and more or less oval. The growing season is similar to that of wheat or slightly shorter. Like rape, mustard is a cool season crop, but is more drought tolerant than rape. The seeds do not shatter, so the seed crop can be harvested (combined) without undue loss. Mustard is usually seeded at the rate of 4 to 10 lb/acre, depending upon the type and cultivar (Downey et al. 1970).

When mustard and rape were compared, Downey et al. (1970) stated "In comparison to 'Echo' rapeseed, yellow mustard is a few days later in maturing, has yellow seed that is about twice the size, and shows more vigor in the seedling stage. Yellow mustard begins flowering at the same time but continues to flower longer. It is similar to 'Echo' in height, is more resistant to shattering, but slightly lower in yield." In general, the mustards are slightly taller and also have thinner leaves and smaller flowers than rape. B. nigra may grow to 10 feet or more in height, with four-sided pods less than 1 inch long. B. hirta and B. juncea grow only to 2 to 4 feet, with B. hirta having pods 3/4 inch to 1 1/2 inches long with strong constrictions between the seeds, whereas pods of B. juncea are longest, 1 1/2 to 2 1/2 inches.

Inflorescence:
The mustard inflorescence is an aggregate of yellow florets at the apex of the raceme, that give a field a deep golden appearance when fully open. The structure of the flower, as given under "Cole Crops," applies equally to the mustard flower. Free (1970*) indicated that the two outer nectaries were somewhat functional but Nieuwhof (1969) stated that they were inactive. Mustard is an excellent source of nectar and pollen for honey bees (Pellett 1947*).

According to Howard et al. (1915), the floret opens between 9 a.m. and noon, and remains open for 3 days. Usually, the stigma projects about 2 mm beyond the petals the afternoon preceding opening of the flower and is immediately receptive. Soon afterwards, however, the corolla begins to grow and reingulfs the stigma. Then the stamens lengthen so that the anthers are level with the stigma, but when the corolla opens, they turn half around. At this period, nectar secretion by the inner nectaries begins. Just before the flower closes, the anthers turn to their former position, and, if any degree of self-fertility exists, selfing can result.

Pollination Requirements:
Muller (1883*) stated that the position of the anthers in relation to the nectaries and stigma makes cross-fertilization likely but by no means inevitable on the visit of pollinating insects. The flower is so constructed that pollen from another flower is likely to be transported to it before its own pollen comes in contact with the stigma. Some of the self-pollen may contact the stigma without the aid of insects, but this contact can be abetted by the bees' visit to the flower. Sampson (1957) showed that compatibility varies with species, cultivar, and even the age of the plant.

Free and Spencer-Booth (1963) found that bees more than doubled seed production of B. alba. In B. juncea, production was increased only 14 percent, an amount that was not statistically significant in their test, but could be of great significance to the grower. Pritsch (1965) also obtained significantly greater yields of white mustard in cages with bees than in cages where bees were excluded. Olsson (1952) obtained a set of 64.7 percent of the flowers, with 2.46 seeds per pod, and 1.75 g per pod with bees excluded, but with bees present these values were increased to 95.3, 4.08, and 2.69, respectively, more than doubling total production. Koutensky (1959) also showed that the seed yield of white mustard was increased 66 percent by honey bee pollination. Howard et al. (1916) indicated that B. juncea was self-fertile but abetted by wind, Downey et al. (1970) stated that oriental and brown mustards (B. juncea) are generally self-pollinated, but yellow mustard (B. hirta) is a cross-pollinated crop. They further stated that wind and bees are both effective in pollination. Free (1970*), citing Akhter (1932) and Olsson (1960), indicated that B. nigra is largely self-sterile.

The above references indicate that yellow mustard is immensely benefited by bee pollination, but the value to oriental or brown mustard is minor, although the actual effect of supplemental pollination has not been too well tested.

Pollinators:
Olsson (1955) found pollen on glass slides exposed 1, 5, 20, and 40 m from fields of rape, turnip rape, and white mustard and he deduced that wind was important in the pollination of these crops. Howard et al. (1916) also believed that wind contributed to pollination. However, mustard is basically an insect-pollinated type of crop, with ample pollen and nectar to attract pollinating insects. Honey bees in particular are attracted to it, and they were shown by Free and Spencer-Booth (1963) to be of great benefit to B. hirta and possibly to B. juncea. The data indicate that repeated visits would be beneficial, thus an ample supply of bees should be present. The number of bees per unit of mustard flowers has not been determined. The flowers are highly attractive to bees for both nectar and pollen so there is no problem in getting visitation if sufficient bees are in the area and the weather permits floral visitation.

Pollination Recommendations and Practices:
No colony recommendations have been made for mustard. Downey et al. (1970) stated, "It has not been found necessary to supply honey bees to produce good seed yields." How maximum production is obtained is not explained, because they indicated that most of the 200,000 acres devoted to mustard seed production is of B. hirta, and the data indicate that production of B. hirta provided with bees is double that where no bees are provided. This indicates that the provision of bee colonies to yellow mustard fields in adequate numbers, probably one to two colonies per acre, should be encouraged.

LITERATURE CITED:
AKHTER, A. R.
1932. STUDIES IN INDIAN BRASSICAE. I. STERILITY AND SELECTIVE POLLEN TUBE GROWTH. Indian Jour. Agr. Sci. 2: 280-292.

DOWNEY, R. K., PAWLOWSKI, S. H., and McANSH, J.
1970. RAPESEED - CANADA'S "CINDERELLA" CROP. Ed 2. Rapeseed Assoc. of Canada Pub. 8, 40 pp.

FREE. J. B., and SPENCER-BOOTH, Y.
1963. THE POLLINATION OF MUSTARD BY HONEYBEES. Jour. Apic. Res. 2: 69 - 70.

HOWARD. A., HOWARD G. L. C., and KAHN. A. R.
1915. STUDIES IN INDIAN OILSEEDS. I. SAFFLOWER AND MUSTARD. Indian Dept. Agr. Mem. Bot. Ser. 7: 237-272.

HOWARD, G. L C., and KHAN, A. R.
1916. STUDIES IN INDIAN OIL SEEDS. I. SAFFLOWER AND MUSTARD. Indian Dept. Agr. Mem. Bot. Ser. 7: 214-272.

KOUTENSKY, J.
1959. [THE POLLINATING EFFECT OF THE HONEY BEE (APIS MELLIFERA L.) ON THE INCREASE IN RAPE AND WHITE MUSTARD YIELDS PER HECTARE.] Ceskoslov. Akad. Zemedel. Ved, Sborn. Rostlinna Vyroba 32(4): 571 582. [In Czech.] AA-441/63.

NIEUWHOF, M.
1969. COLE CROPS. 353 pp. Leonard Hill, London.

OLSSON, G.
1952. [INVESTIGATIONS OF THE DEGREE OF CROSS-POLLINATION IN WHITE MUSTARD AND RAPE.] Sverig. Utsadesfpren. Tidskr. 62(4): 311 - 322. [In Swedish, English summary.]

______ 1955. [WIND POLLINATION OF CRUCIFEROUS OIL PLANTS.] Sverig. Utsadesforen. Tidskr. 65(6): 418-422. [In Swedish, English summary.]

______ 1960. SELF-INCOMPATIBILITY AND OUTCROSSING IN RAPE AND WHITE MUSTARD. Hereditas 46: 241 252.

PRITSCH, G.
1965. [INCREASING THE YIELD OF OIL PLANTS BY USING HONEY BEES.] Ved. Prace Vyzkum. Ustav. Vcelar CSAZV 4: 157-163. [In German.] AA-770/66.

ROBINSON, R. G.
1964. MUSTARD AND RAPE OILSEED CROPS FOR MINNESOTA. Minn. Agr. Ext. Serv. Bul. 311, 12 pp.

SAMPSON, D. R.
1957. THE GENETICS OF SELF- AND CROSS-INCOMPATIBILITY IN BRASSICA OLERACEA. Genetics 42: 252 - 263.

SHAW A. F.
1956. MUSTARD SEED PRODUCTION. Mont. Agr. Ext. Serv. Cir. 261; 6 pp.

NIGER
Guizotia abyssinica (L.f. ) Cass., family Compositae

Niger is grown for its seeds, which yield a yellow, edible semidrying oil with little odor and a pleasant nutlike taste. The oil is used in cooking, oil lamps, soaps, and paints; the pressed cakes are used for livestock feed; and the seeds are fried and eaten or used in chutneys or as a condiment. According to Purseglove (1968*), niger is grown primarily in Ethiopia (100,000 to 200,000 tons of oil produced per year) and India (75,000 tons per year). Chavan (1961) stated that India had 716,000 acres of niger.

Plant:
The plant is a branched annual herb, l/2 to 1 l/2 m tall. The period of growth of the plant to the time of flowering is about 3 months, then another 1 1/2 months are required to ripen the seeds. Pure stands yield 350 to 400 pounds of seed per acre (Purseglove 1968*).

Inflorescence:
The yellow 2- to 3-cm flower heads develop in the leaf axil, two to five in a cluster. Each head contains about eight ray florets and 40 to 60 hermaphrodite disk florets (Free 1970*). Within the disk floret, the anthers are united to form the corolla tube. The style extends through this tube, and the hairy forked stigma is above. The floret opens and liberates its pollen early in the morning, the style emerges about midday, and the stigma lobes separate and curl backward toward evening.

Pollination Requirements:
Howard et al. (1919) found that cross-pollination was common. They reported that the stigma lobes rarely curled back sufficiently to touch their own style, indicating that the plants were self-sterile. This explains why isolated plants set no seed. Although the flowers are hermaphrodite, they are not self-pollinating. Bhambure (1958) confined plants in two cages 1.2 by 1.2 mm, and tagged 40 flower heads in each cage. In one of these cages, bees (Apis cerana) were introduced. In the cage with bees, 40 seeds per head developed. In the one without bees, only 15 seeds per head were harvested. Chavan (1961) obtained similar data.

Pollinators:
The meager data indicate that this important crop is largely dependent upon pollinating insects, and growers who desire maximum bee activity in the field would do well to provide an ample bee supply to each field where seeds are desired.

Pollination Recommendations and Practices:
None.

LITERATURE CITED:
BHAMBURE, C. S.
1958. EFFECT OF HONEY BEE ACTIVITY ON NIGER (GUIZOTIA ABYSSINICA CASS.) SEED PRODUCTION. Indian Bee Jour. 20: 189 - 191.

CHAVAN, V. M.
1961. NIGER AND SAFFLOWER 150 pp. Indian Cent. Oilseeds Com., Hyderabad.

HOWARD, A., HOWARD, G. L. C., and RAHMAN, K. A.
1919. STUDIES IN THE POLLINATION OF INDIAN CROPS. India Dept. Agr. Mem.. Bot. Ser. 10: 195 - 220.

NUTMEG AND MACE
Myristica fragrans Houtt., family Myristicaceae
Nutmeg and mace are produced in the tropical areas of Indonesia and the West Indies. Purseglove (1968*) indicated that annual production of nutmeg amounted to about 170,000 cwt (1 cwt = 112 lb). About 1 pound of mace is obtained for each 10 pounds of nutmeg. This would indicate that between 15,000 and 20,000 cwt (2,240,000 lb) of the volume produced was mace.

Plant:
The nutmeg tree is bushy, 30 to 40 feet tall, resembles an apricot, and the trees are usually spaced about 30 feet apart. It produces a pale orange-yellow fruit about 2 1/2 inches long, that resembles an apricot, but when ripe the l/2 inch thick husk separates into two pieces, disclosing the dark-colored nut, covered with a brilliant scarlet network (aril) known as mace (Nicholls and Holland 1929). When the nuts are harvested, the aril is separated from the nut and sold as mace, and the nut marketed as nutmeg. The seeds must be planted within 3 days after harvest or the viability is lost. The plant will begin fruiting at 5 to 6 years of age, but is at its best by 15 years and will remain at this productive level another 10 to 20 years (Ridley 1912*). A plant may produce 1,800 fruits in a year, yielding 20 pounds of nutmeg and 2 pounds of mace (Nicholls and Holland 1929).

Inflorescence:
The nutmeg tree is dioecious, with male flowers on one tree and female flowers on another. Occasionally, a plant may have a few flowers of the opposite sex, that is, a male tree may have a few female flowers. Occasionally, also, the sex of the plant may change entirely, particularly it may change from all male or staminate flowers to completely female or pistillate flowers (Ridley 1912*).

The bell-shaped pendant, light-yellow flowers are in small cymes on a woody stalk one-half inch in diameter. The 5- to 10-mm male flowers are more globose than the female ones, and have a mass of cylindrical stamens 8 to 12 mm and extending to the flower opening. The slightly larger (10 mm) female flowers, seldom over three in a raceme, are dilated at the base, with a tiny, two-lobed stigma and an ovary that largely fills the corolla. Nectar is produced in both types of flowers at the base of the corolla. The development from flower to ripe fruit requires 6 to 9 months (Flach and Cruickshank 1969). There may be three flowering cycles during the year.

Pollination Requirements:
There seems to be little doubt that cross-pollination is required between trees as there are insufficient flowers of both sexes on any one tree. The pollen must be transported to the numerous pistillate flowers to set the 1,500 to 2,000 nuts expected per year on a mature tree.

Pollinators:
Nutmeg is insect pollinated, but there is lack of agreement as to what insects are responsible. Flack and Cruikshank (1969) stated that "natural pollination is carried out by a moth." Ridley (1912*) stated that he had seen only small bees and small beetles visit the flowers. Nicholls and Holland (1929) stated that pollination is effected only by wind and insects. Purseglove (1968*) said that pollination is probably effected by small insects. It becomes evident that there is insufficient information on the pollination of this crop, but logically its pollination is by insects.

Pollination Recommendations and Practices:
None, yet the evidence indicates that for stable production the grower of nutmeg should arrange for a stable pollinator population on these flowers.

LITERATURE CITED:
FLACH, M., and CRUICKSHANK, A. M.
1969. NUTMEG. In Ferwerda, F. P., and Wit, F., eds. Outlines of Perennial Crop Breeding in the Tropics, pp. 329-338. H. Veenman and Zonen, N. V. Wageningen, The Netherlands.

NICHOLLS, H. A., and HOLLAND, J. H.
1929. A TEXTBOOK OF TROPICAL AGRICULTURE. 639 pp. Macmillan & Co., Ltd., London.

PIMENTO OR ALLSPICE
Pimenta dioica (L.) Merrill, family Myrtaceae
Pimento is a semiwild crop in Jamaica and the nearby islands where most of the world's supply is produced (Chapman 1966). Deliberate planting of pimento in Jamaica is negligible (Chapman and Glasgow 1961).

Plant:
The plant is an aromatic tree to 40 feet tall, with 6-inch oblong leathery leaves that shed twice a year. The dried, unripe fruit is a dark- brown, round berry, one-fourth inch across, known as allspice. Its flavor is considered to be a combination of the flavors of cinnamon, cloves, and nutmeg; hence the name allspice. Oil extracted from the dried berries is a stimulant carminative (Purseglove 1968*). Although the trees superficially appear to be hermaphrodite, some of them actually function as male and others as fruiting female trees. The differences in the two types are recognizable at harvesttime. Chapman (1966) suggested planting or budding male trees as alternate trees in alternate rows to provide pollination for the bearing trees.

Inflorescence:
The inflorescence consists of a cluster of several dozen white flowers, 2 to 6 inches long, each flower having four tiny petals, a single style with one ovary, two ovules, and a cluster of anthers. As the flower opens, the style straightens, and although the stigma is raised above the anthers, the flower appears to be hermaphrodite. However, there are differences in flowers between trees. The barren or male-type tree has many flowers, each of which has 75 to 100 anthers per flower. These flowers produce much pollen. Most of the flowers shed, but one or two per tree may produce one-seeded fruit. The bearing or female-type tree has fewer flowers, and the flowers have fewer anthers. The small amount of pollen produced is nonviable but may serve to lure pollen-coated bees from other trees. A bearing tree may produce 20 pounds of berries per year but yields of 150 pounds for one tree have been recorded. The flowers are attractive to honey bees and some other pollinators (see Chapman 1964*,1966).

Pollination Requirements:
Ward (1961) bagged inflorescences and obtained only 19 berries as compared to more than a thousand obtained from a similar number of flowers that were not bagged, which established that the flower must be cross-pollinated. Chapman and Glasgow ( 1961 ) considered the barrenness physiological.

Pollinators:
Ward (1961) believed that wind was the primary pollinating agent. Chapman (1966) considered the plant to be cross-pollinated by bees.

Pollination Recommendations and Practices:
Chapman (1966) recommended the placement of honey bee colonies in the plantings to transfer the pollen to receptive stigmas. The relative concentration of colonies was not indicated.

LITERATURE CITED:
CHAPMAN G. P.
1966. FLORAL BIOLOGY AND THE FRUITFULNESS OF JAMAICA ALL- SPICE (PIMENTA DIOICA (L.) MERRILL). In 2d Internatl. Symposium on Pollination, London, 1964. Bee World 47(Supp.): 125-130.

______and GLASGOW S. K.
1961. INCIPIENT DIOECY IN PIMENTO. Nature [London] 192: 1205-1206.

WARD, J. F.
1961. PIMENTO. 18 pp. (Hors. Dept. Min. Agr. and Lands) Govt. Printer, Kingston, Jamaica.

PYRETHRUM
Chrysanthemum cinerariifolium (Trevir.) Vis., family Compositae
Pyrethrum is grown for the insecticidal material, pyrethrins, that is found primarily in the flower head. It is grown in numerous countries but Kenya, Tanzania, Ecuador, and Rwanda produced an average of 34.2 million of the 35 million pounds produced per year during 1966-70 (Fowler and Mahan 1972). Pyrethrum has been tested experimentally but has not been grown commercially in the United States (Drain and Shuey 1934, McClintock 1929) although, if mechanical harvesting of the flower heads could be perfected and the hand labor reduced, its growth in certain areas might be feasible.

Plant:
Pyrethrum is a tufted, slender, pubescent perennial 12 to 30 inches high with daisylike flower heads 1 1/2 inches across, on long slender stems. It is adapted to a temperate climate with 45 to 50 inches of rainfall. The seeds are sown in special beds. Four months later the 4 to 5- inch-high plants are transplanted into the field (fig. 163).

In another 4 months, harvest of the just-opening flower heads begins and is repeated every 2 or 3 weeks for several months. After three annual harvests, the plants decrease in productivity and are plowed under and another crop planted. Maximum productivity (800 to 1,000 pounds dried seed heads per acre) may be obtained the second season (Purseglove 1968*).

[gfx] FIGURE 163. - Pyrethrum plants cut with a grain binder and curing in shocks. Experimental production at Glenn Dale, Md.

Inflorescence:
The flower head consists of 18 to 22 white, pistillate ray florets almost an inch long and a tightly packed cluster of 40 to 100 yellow, short, bisexual disk florets (fig. 164).

The flower is not considered highly attractive to honey bees, which seem to collect pollen primarily and only at certain times. The main insect visitors were reported by Kroll (1961) to be adult coleoptera and diptera, and their presence was seasonal. Kroll (1961) and Smith (1958), however, indicated that bees increased production of pyrethrum, so presumably the flowers were visited by these insects.

Harvest begins when the florets on a head are about three-fourths open (Hartzell 1943). The pyrethrin content of the flower increases as the flower stage increases: Buds unopened, 0.84 percent; one row of disk flowers open, 1.83 percent; and overblown and ripening, 1.21 percent (Kroll 1964). A similar variation from 0.23 to 1.36 percent pyrethrin was also obtained from different plant sources by Hoyer and Leonard (1936).

[gfx] FIGURE 164. - Pyrethrum flowers in different stages of development.

Pollination Requirements:
The pollination requirements of pyrethrum are not too clear, probably because of differences obtained in tests with different cultivars or under different environmental or ecological conditions. Culbertson (1940) stated that seed formation seemed to be the result of self- fertilization or apomyxis because flower heads bagged and with the anthers removed set seed. Delhaye (1956) stated that pyrethrum is highly self-fertile although a higher set of seed, and seed with higher viability, are obtained when the pollen comes from another clone. Kroll (1961) discussed a test comparing production of plants in cages with bees present, with bees excluded, and open plots. He reported that the analysis of the data was not quite conclusive but gave strong indications that production of pyrethrum is increased by insect pollination, and that fertilized embryos contain more pyrethrin than unfertilized embryos.

Purseglove (1968*) stated, without supporting data, that pyrethrum is self-sterile and must be cross-pollinated to produce viable seeds. He stated that it is insect-pollinated mainly by coleoptera and diptera. Kroll (1961) stated that the percentage of unfertilized and nonviable seeds in the field is very high. This, he concluded, seemed to indicate that the number of insect visitors was never large enough to effect satisfactory fertilization, and, at the same time, it provided a strong argument in favor of the predominance of cross- as opposed to self- fertilization.

The fertile achene was shown by Chandler (1956) to contain 1.05 percent pyrethrin compared to only 0.71 percent of barren achenes, which shows the value of having pollinated flowers for highest pyrethrin production.

Parlevliet and Contant (1970) stated that most clones are highly self-incompatible. Smith (1958) reviewed a test by L. A. Notcutt which showed that the yield of seeds was greatest from cages with bees, least from cages excluding pollinating insects, and intermediate in open plots.

A United Nations (FAO) (1961) report stated that pyrethrum is a cross-fertilized plant that requires insects for cross-pollination, the main pollinators being bees and other hymenoptera.

Brewer (1968) stated that the floret's own pollen cannot reach the receptive surfaces of the style (the stigma) because the styler lobes are closed when they extrude through the anther tube. He concluded that by the time the style becomes receptive, the germination of the floret's own pollen is about past. Delhaye (1956) tested the effect of selfing and crossing on the germination of pyrethrum seed. He found the following: Selfed without bees, 0.0 to 1.0 percent; selfed with bees present, 1.7 to 22.7 percent; crossed without bees, 5.2 to 8.3 percent; and crossed by bees, 17.7 to 27.7 percent. Brewer (1968) concluded:

The flower morphology and the flower morphogenesis of Pyrethrum resemhles closely the classical concept known in the Compositae. The flowering rhythm of the inflorescences encourages crosspollination through:

(1) The individual floret discharges the ripe pollen before it unfolds the receptive surfaces of the style.
(2) When insects visit the inflorescence their path follows the development of the flower, i.e., from the margin to the centre, in order to collect pollen and nectar. Thus they deposit the foreign pollen they carry on fully opened styles.
(3) By sticking together, the pollen mass encourages transport by insects.
(4) Pollen does not germinate on genotypically identical styles. Strong evidence exists that the incompatibility system is sporophytically determined.
(5) The limited life of the pollen after anthesis reduces the chance for own pollen to germinate on styles of the same floret.

Lower germination percentages of the pure seed (P.G.S.) are due to rainfall during the maturing period of the seed.

Pollinators:
The previous references indicate that honey bees are not overly attracted to pyrethrum flowers, as compared to beetles and flies. There has been no attempt to concentrate honey bee colonies near the crop. A test should be conducted to determine the practicality of supplying honey bee colonies to pyrethrum fields for pollination purposes. The use of leafcutter and other wild bees and different species of flies should also be investigated.

Pollination Recommendations and Practices:
There are no recommendations on the use of pollinating insects on pyrethrum, even though the evidence indicates they are beneficial.

LITERATURE CITED:
BREWER, J. G.
l968. FLOWERING AND SEEDSETTING IN PYRETHRUM (CHRYSANTHEMUM CINERARIAEFOLIUM VIS.) A REVIEW. Pyrethrum Post 9(4): 18-21.

CHANDLER, S. E.
1956. BOTANICAL ASPECTS OF PYRETHRUM III. THE NATURAL HISTORY OF THE SECRETORY ORGANS; THE PYRETHRIN CONTENT OF THE FERTILE ACHENES. Pyrethrum Post 4(1): 10-15.

CULBERTSON, R. E.
1940. AN ECOLOGICAL, PATHOLOGICAL AND GENETICAL STUDY OF PYRETHRUM (CHRYSANTHEMUM CINERARIAEFOLIUM VIS.) AS RELATED TO POSSIBLE COMMERCIAL PRODUCTION IN THE UNITED STATES. Diss. submitted to Pa. State Univ., Hort. Dept., as partial furfillment of requirements for Ph.D. degree, 289 pp.

DELHAYE, R. J.
1956. [PRELIMINARY NOTE ON THE FLORAL BIOLOGY AND CONTROLLED FERTILIZATION OF PYRETHRUM, CHRYSANTHEMUM CINERARIAEFOLIUM (TREV.) BOCC.] Bul. Agr. Congo Belge 47: 1327-1343. [In French]

DRAIN, B. D., and SHUEY, G. A.
1934. THE ISOLATION AND PROPAGATION OF HIGH PYRETHRIN STRAINS OF PYRETHRUM. Amer. Soc. Hort. Sci. Proc. 32: 19O-191.

FOWLER, D. L., and MAHAN, J. N.
1972. THE PESTICIDE REVIEW 1971. U.S. Dept. Agr. Stabilization and Conserv. Serv., 56 pp.

HARTZELL, A.
1943. PYRETHRUM CULTURE IN DALMATIA WITH SOME APPLICATIONS TO THE AMERICAS. Jour. Econ. Ent. 36: 320-325.

HOYER, D. G., and LEONARD, M. D.
1936. PYRETHRIN CONTENT OF PYRETHRUM FLOWERS FROM VARIOUS SOURCES. Jour. Econ. Ent. 29: 605-606.

KROLL, U.
1961. THE INFLUENCE OF FERTILIZATION ON THE PRODUCTION OF PYRETHRINS IN THE PYRETHRUM FLOWER. Pyrethrum Post 6(2): 19-21.

____ 1964. PYRETHRUM IN KENYA. Outlook on Agr. 4(4): 177-181.

MCCLINTOCK, J. A.
1929. PYRETHRUM. Tenn. Agr. Expt. Sta. 42d Ann. Rpt.: 40.

PARLEVLIET, J. E., and CONTANT R. B.
1970. SELECTION FOR COMBINING ABILITY IN PYRETHRUM (CHRYSANTHEMUM CINARIAEFOLIUM VIS.). Euphvtica 19: 4-11.

SMITH. F. G.
1958. BEEKEEPING OPERATIONS IN TANGANYIKA, 1949-1957. Bee World 39: 29-36.

UNITED NATIONS FOOD AND AGRICULTURE ORGANIZATION (FAO).
1961. AGRICULTURAL AND HORTICULTURAL SEEDS. FAO Agro-studies 55, 531 pp.

RAPE³²Brassica spp., family Cruciferae
Two species of Brassica are known as rape, a word derived from the Latin word "rapum" meaning turnip. B. napus L. is known in Canada as the Argentine type of rape, and elsewhere as summer rape, winter rape, colza, colza-oil rape, or swede rape. B. campestris L. is known as field mustard, summer turnip rape, Polish rape, toria, and sarson. Sarson is somewhat different from toria for it has both yellow-seeded and brown- seeded cultivars.

Rape is not extensively grown in the United States, but there are about 4 million acres in nearby Canada. About 80 percent of this acreage is planted to B. campestris, 20 percent to B. napus. The oil, pressed from the seed, is used in margarine and shortenings and in salad and cooking oil. The quality of rape oil is equal to or better than soybean oil. Rapeseed meal has found wide acceptance as a food for many classes of livestock. The protein in rape is considered equivalent to that in soybean on a pound-for-pound basis (Downey et al. 1970). Yields reported from Canada range from 1,560 to 2,220 pounds seed per acre.
__________
³² See also, "Cole Crops," and "Mustard."

Plant:
Young rape plants look somewhat like young cabbage plants, with basal leaves 4 to 12 inches or more long and half as broad as long. The flower stalk of B. napus grows to a height of 2 1/2 to 4 feet, whereas B. campestris reaches only 1 l/2 to 3 feet. Rape is a cool-season crop but is susceptible to frost. The seeds of rape are drilled into the soil, just like wheat, either late in the fall (making it a sort of biennial) or early in the spring as an annual.

Inflorescence:
The plant is topped by a mass of golden yellow flowers that in bloom give the field a bright golden appearance.

The flowers are in elongated terminal racemes. There are the four characteristic cruciferous petals and, usually, six stamens, four projecting above the stigma and two shorter than the style. There are four partly concealed nectar glands, two on the inner side of the short stamens, the others between the insertions of each pair of long stamens (Knuth 1908*, p. 96). The latter nectaries become more accessible to bees as the flower matures (Meyerhoff 1958). The fruit is a slender silique or pod 2 to 4 inches long. Knuth (1908*, p. 76) stated that when the flower of B. napus opens, the anthers are still unripe, and those of the four long stamens lie close to the already mature stigma. Before the corolla has fully expanded, the anthers make a half turn and dehisce so the pollen-covered sides are turned outward. The anthers of the short stamens, 2 to 3 1/2 mm below the stigma, remain with their pollen-covered sides toward the style, but they lean outward. When the flower fades, the long stamens recurve, so that automatic self- pollination may occur if the plant is selfcompatible. Flowering extends from 22 to 45 days (Gerard and Cronan 1963, Radchenko 1964).

Rape produces nectar sufficiently to be considered a better honey plant than white or red clover (Hammer 1966). The nectar can be seen glistening in the bottom of the flower all day, and a colony of honey bees may store 15 to 33 pounds of honey per day (Palmer 1959). This, of course, would depend upon the strength of the colony, the number of flowers present, and weather conditions.

Pollination Requirements:
There seems little doubt that B. campestris or Polish rape requires insects for cross-pollination and seed production. Koutensky (1958) showed that production in fields with apiaries beside them was 2,095 kg/ha (1,844 lb/acre), but with apiaries 2.4 km (1.4 miles) distant the production was only 1,275 kg/ha (1,511 lb/acre). Pritsch (1965) studied the pollination of B. campestris in cages with bees compared to cages with bees excluded, although smaller insects had access to both cages, and obtained significantly greater production with bee pollination. Downey and Bolton (1961) reported that the yield of seed in fields stocked with bees was at least 30 percent higher than fields not supplied with bees. Downey et al. (1970) stated that B. campestris is almost completely self-sterile and bees must be provided. White (1970) reported that summer turnip rapes are almost completely cross-pollinated, the "true" rapes about one-third cross-pollinating and twothirds self-pollinating.

In the case of B. napus or Argentine rape, there is some question about the degree of benefit from insect pollinators. Knuth (l 908*, p. 98) reported that "according to some authors the plant is self-sterile," and insect visitation will increase seed production. According to Free and Nuttall (1968), Fujita (1939) reported that B. napus plants caged with bees produced 25 percent more seed than plants caged without bees. Von Rhein (1952) cited other workers who showed that bees caused 17.4 percent more seed per pod and 9.7 percent heavier seeds than were produced on plants not visited by bees. Louveaux and Verge (1952) reported a 50 percent increase in seeds per pod on plants growing near a large apiary as compared to plants caged to exclude bees.

Jenkinson and Jones (1953) reported that although the relationship of anthers to the stigma in individual flowers favors self-pollination, the presence of bees resulted in increased yields, for example, 8.8 seeds per pod with bees present versus 5.3 seeds per pod with bees excluded.

Downey et al. (1970) stated that B. napus is largely self-pollinated, and thus a good uniform set of seed can be obtained without bees. Turnip rape, however, is almost completely self-sterile and requires cross- pollination to set seed. Wind can carry pollen from one plant to another, but insects, particularly bees, are important. Experiments show that when fields of turnip rape (B. campestris) are stocked with bees, earlier and more uniform maturity results. Downey (1964) indicated that when bees were excluded from 'Arlo' cv. of rape only two-thirds as much seed set as when bees were present at flowering. Even so, he indicated that neither native bees nor honey bees were available in sufficient quantities for effective pollination of the large acreages of rape in Canada. Nothing was mentioned about transporting colonies to the fields that had low pollinator populations. The ratio of seeds per flower with bees present was 6.7 compared to 2.8 without bee visitation.

Koutensky (1959) reported that bee pollination increased seed yields of B. napus v. arvensis by 64 percent. Vesely (1962) reported that bee activity increased B. napus v. oleifera seed production by 25 percent and that cross-pollinated plants set the crop of seed and ceased flowering earlier than plants not visited by bees. Pritsch (1965) also reported significantly greater yield of seed in cages with bees than with bees excluded. Free and Nuttall (1968) reported a 13 percent increase in seed yield from cages with bees compared to cages without bees - an amount they did not consider of significance. Downey et al. (1970) indicated that B. napus was 70 percent self- pollinated. Mohammad (1935) stated that in toria and brown-seeded sarson 12 and 20 percent of bagged pods set, whereas 91 percent of yellow- seeded sarson in bags set. He also stated that plants from cross- pollinated seeds were more productive.

Meyerhoff (1954) conducted five tests over 3 years with 'Lembke's' winter rape. He concluded that honey bees increased the number of pods per plant by 53.2 percent, pod length by 6.1 percent, and seeds per pod by 12.6 percent. Zander (1952) also studied 'Lembke's' winter rape in one cage with bees, in one without bees, and an open plot. In mid-May, the plants in the cage with bees had set their seed crop and had ceased flowering, whereas the plants in the cage without were still in full bloom; the stage of the plants in the open was between that of the others. Latif et al. (1960) showed that rape seed production in fields with bees was more than double that in fields where bees (Apis cerana F.) were absent. Olsson (1955) also showed that rape was about one-third cross-pollinated in open fields, whereas white mustard was almost completely self-pollinated. The presence of bees in cages of white mustard doubled the number of seeds per pod and increased the pod set by 50 percent.

Vasil (1964) and Hasler and Maurizio (1949, 1950) have shown that boron, in some unknown way, influences the pollination of B. napus and other plants. More information in this area would be most helpful in understanding the pollination and fruit setting not only in the Brassicas but also in the pollination and fruit setting of other plants.

The above tests showed a benefit from bee pollination ranging from 13 to 64 percent more seeds per pod, and with earlier cessation of flowering. This would indicate that the crop is considerably benefited by insect pollination.

Olsson (1955) gave wind pollination some credit in the setting of rape seed, but most other researchers consider wind as only a minor factor.

Pollinators:
Rahman (1940) studied the pollinators of B. napus in India. He concluded that the dwarf honey bee of India (Apis florea F.), the wild bees (Andrena ilerda Cam. and Halictus sp.), and the fly (Eristalis tenex (L.)) were the most important pollinators.

Free and Nuttall (1968) studied the activity of honey bees on B. napus. They reported that all bees that visited the flowers collected nectar although some collected pollen also. All became covered with pollen, but some removed and discarded it. Those that collected the pollen did so primarily during the morning hours.

Honey bees are the primary pollinators of rape (Belozerova 1960, Nikitina 1950, Radchenko 1964, Vesely 1962). The plant is highly attractive to honey bees, providing both nectar and pollen, and the honey bee is of appropriate size for effective transfer of pollen from anthers to stigma. Hammer (1952) reported as many as 20,000 bees per hectare of rape in fields 31/2 to 4 km from the apiary. Each bee was returning to the hive with 30- to 60-mg loads of nectar, roughly half the weight of a worker bee.

Belozerova (1960) noted that B. napus had 2.326 mg nectar per flower at the beginning of bloom, 1.950 mg during the peak, and 1.350 mg per flower toward the end of blooming. He noted that 96.3, 95.3, and 72.9 percent of the floral visitors at the three different periods were honey bees. Other pollinators in India include Apis florea, A. dorsata, A. cerana, and Andrena ilerda (Kapil et al. 1969).

Pollination Recommendations and Practices:
Hammer (1963, 1966) recommended three colonies per hectare (1.5 colonies per acre); Radchenko (1964), two colonies per hectare (0.8 colony per acre); Downey and Bolton (1961), one colony per acre; White (1970), two colonies per acre; and Vesely (1962) three to four colonies per hectare (1.5 to 2 colonies per acre). Downey et al. (1970) stated that it is not necessary to provide bees to produce good seed yields, which is puzzling when it is remembered that 80 percent of the rape in Canada is B. campestris, which is largely self-sterile. White (1970) said that both summer turnip rapes and true rapes depend on bees for maximum production. The data indicate that a heavy bee population on rape would be beneficial, but no data establish the maximum floral visitation desired. Until more concrete data are available, the one to two strong colonies of honey bees per acre cited above would appear to be a logical usage.

The ideal pollinator population and proper distribution of colonies for most efficient pollination of rape needs to be determined.

LITERATURE CITED:
BELOZEROVA, E. I.
1960. [BEES INCREASE SEED CROP FROM WINTER RAPE.] Pchelovodstvo 37(9): 38-40. [In Russian.] AA-939163.

DOWNEY, R. K
1964. EFFECT OF BEES ON SEED YIELDS OF ARLO RAPESEED. Forage Notes 10: 1.

____ and BOLTON, J. L.

1961. PRODUCTION OF [POLISH AND ARGENTINE] RAPE IN WESTERN CANADA. Canada Dept. Agr. Res. Br. Pub. 1021, 19 pp.

____ PAWLOWSKI, S. H., and MCANSH, J.
1970. RAPESEED - CANADA'S "CINDERALLA", CROP. Ed. 2. Rapeseed Assoc. of Canada Pub. 8, 40 pp.

FREE, J. B., and NUTTALL, P. M.
1968. THE POLLINATION OF OILSEED RAPE (BRASSICA NAPUS) AND THE BEHAVIOUR OF BEES ON THE CROP. Jour. Agr. Sci. (Cambridge) 71: 91-94.

FUJITA, M.
1939. [INFLUENCE OF HONEYBEES ON THE FRUCTIFICATION OF RAPE.] Bul. imp. Zootech. Exp. Stn. Chiba-shi 34: 1. [ In Japanese. ]

GERARD, BROTHER, AND CRONAN, FATHER.
1963. RAPESEED - IOWA GOLD. Amer. Bee Jour. 103: 218-219.

HAMMER, O.
1952. [RAPE GROWING, BEES AND SEED PRODUCTION.] Dansk Landbr. 71: 67-69. [In Danish.] AA-122/54.

____ 1963. [SUMMER RAPE AS A COMPETITOR AFFECTING THE POLLINATION OF CLOVERS.] Dansk Froavl No. 14, 7 pp. [In Danish.] AA-425/68.

____ 1966. SOME PROBLEMS OF COMPETITION BETWEEN SUMMER RAPE AND CLOVER, IN RELATION TO POLLINATION. In 2d Internatl. Symposium on Pollination, London, 1964. Bee World 47, Supp. 1: 99-106.

HASLER, A., and MAURIZIO, A.
1949. [THE ACTION OF BORON ON SEED - SETTING AND NECTAR SECRETION IN RAPE (BRASSICA NAPUS L.).] Phytopath. Zeits. 15(2): 193-207. [In German, English summary.]

____ and MAURIZIO, A.
1950. [INFLUENCE OF VARIOUS NUTRITIVE SUBSTANCES ON THE DEVELOPMENT OF BLOSSOM, THE SECRETION OF NECTAR, AND THE SEED-YIELD, OF NECTAR PLANTS (ESPECIALLY SUMMER RAPE).] Schweiz Landw. Monatshefte 6: 201-211. [In German.]

JENKINSON, J. G., and JONES, G. D. G.
1953. OBSERVATIONS ON THE POLLINATION OF OIL RAPE (BRASSICA NAPUS) AND BROCCOLI (BRASSICA OLERACEA). Bee World 34: 173-177.

KAPIL, R. P., GREWAL, G. S., KUMAR, S., and ATWAL, A. S.
1969. INSECT POLLINATORS OF RAPE SEED AND MUSTARD. (Abstract) 56th Indian Sci. Cong. Proc., pt. 3, p. 509.

KOUTENSKY, J.
1958. [THE RESULTS OF THE POLLINATING WORK OF BEES.] Vcelarstvi 11(5): 72-73. [In Czech.] AA-70/60.

KOUTENSKY, J.
1959. [THE POLLINATING EFFECT OF THE HONEY BEE (APIS MELLIFERA L.) ON THE INCREASE IN RAPE AND WHITE MUSTARD YIELDS PER HECTARE.] Ceskoslov. Akad. Zemedel. Ved, Sborn. Rostlinna Vyroba 32(4): 571-582. [In Czech., English summary.]

LATIF, A., QAYYUM, A., and ABBAS, M.
1960. THE ROLE OF APIS INDICA IN THE POLLINATION OF [OIL SEEDS] "TORIA" AND "SARSON" (BRASSICA CAMPESTRIS VAR. TORIA AND DICHOTOMA). Bee World 41: 283-286.

LOUVEAUX, J. and VERGE, J.
1952. [RESEARCHES ON THE POLLINATION OF WINTER RAPE.] Apiculteur 96 (Sect. Sci.): 15-18. [In French. ] AA-213/53.

MEYERHOFF, G.
1954. [INVESTIGATION ON THE EFFECT OF BEE VISITS ON RAPE.] Arch. f. GeflugelZucht und Kleintierkunde 3(3/4): 259-306. [In German.] AA-99/59.

____ 1958. [BEHAVIOR OF BEES FORAGING ON RAPE.] Leipzig. Bienenztng. 72(6): 164-165. [In German.] AA-89/60.

MOHAMMAD, A.
1935. POLLINATION STUDIES IN TORIA (BRASSICA NAPUS VAR. DICHOTOMA PRAIN), AND SARSON (B. CAMPESTRIS L. VAR. SARSON PRAIN). Indian Jour. Agr. Sci. 5: 125-154.

NIKITINA, A. I.
1950. [HONEYBEES RAISE SEED YIELDS OF TURNIPS AND RUTABAGA.] PchelovodstVo 27(5): 271-274. [In Russian.]

OLSSON, G.
1955. [WIND POLLINATION OF CRUCIFEROUS OIL PLANTS.] Sverig. Utsadesforen. Tidskr. 65(6): 418-422. [In Swedish, English summary.

PALMER, S.
1959. A HONEY PLANT PAR EXCELLENCE. Gleanings Bee Cult. 87: 460-461.

PRITSCH, G.
1965. [INCREASING THE YIELD OF OIL PLANTS BY USING HONEY BEES.] Ved. Prace Vyzkam. Ustav. Vcelar CSAZV 4: 157-163. [In German.] AA-770/66.

RADCHENKO, T. H.
1964. [THE INFLUENCE OF POLLINATION ON THE CROP AND THE QUALITY OF SEED OF WINTER RAPE.] Bdzhil'nitstvo 1: 68-74. [In Ukrainian, Russian summary.] AA- 380/69.

RAHMAN, K. A.
1940. INSECT POLLINATORS OF TORIA (BRASSICA NAPUS LINN., VAR. DICHOTOMA PRAIN) AND SARSON (B. CAMPESTRIS LINN., VAR. SARSON PRAIN) AT LYALLPUR. Indian Jour. Agr. Sci. 10(3): 422-447.

RHEIN, W. VON.
1952. [RESULTS OF TRAINING BEES BY SCENT DURING THE 1952 RAPE FLOW.] Hess. Biene 88(8): 192-194, (9): 218-220. [In German.] AA- 220/53.

VASIL, I. K.
1964. EFFECT OF BORON ON POLLEN GERMINATION AND POLLEN TUBE GROWTH. In H. F. Linskens, ea., Pollen Physiology and Fertilization, pp. 107-119. Symposium held in August 1963 at Nijmegen Univ., The Netherlands.

VESELY V.
1962. [THE ECONOMIC EFFECTIVENESS OF BEE POLLINATION ON WINTER RAPE (BRASSICA NAPUS L., VAR. OLEIFERA METZ.).] Min. Zemedel. Lesn. a Vodniho Hospodar. Ust. Vedtech. Inform. Zemedel. Ekon. 8(9): 659-673. [In Czech., German summary.] AA-866/65.

WHITE, B.
1970. POLLINATION OF COMMERCIAL RAPE SEED CROPS. Australasian Beekeeper 72(4): 99-100.

ZANDER, E.
1952. [RAPE AND BEES.] Z. Bienenforsch. 1(8): 135-140. [In German.] AA-121/54.

SAFFLOWER
Carthamus tinctorius L., family Compositae
Safflower is grown principally in California and Arizona, but has been grown successfully in every State west of the 100th meridian (USDA, 1961, Dennis and Rubis 1966, Shaw and Joppa 1963, Klages 1954, Knowles and Miller 1960). The acreage varies from year to year according to the demand for safflower oil, which is obtained from the crushed seed. The oil is used in paints and in margarine and other human food. In 1963, a peak acreage of around 301,000 acres was grown. Production of seed per acre on irrigated soils has varied from 2,500 to 4,000 lb/acre; on dryland soils, from 500 to 2,500 pounds (Knowles and Miller 1965).

Safflower is frequently planted instead of barley. Although safflower is slightly more costly to produce, the same culture and harvesting equipment can be used on each. When grown in cotton-producing areas, the cotton oil mills process the seed. The residue after the oil is removed is used for livestock feed (Knowles 1955, Halloran and Kneeland 1961). The price has relatively stabilized at $80 to $90 per ton, which also amounts to about $80 to $90 per acre.

Plant:
Safflower, like other such related plants as artichoke, thistles, and star thistles, has spine-tipped leaves that make contact with the plant unpleasant. It is an upright annual, 2 to 6 feet high (fig. 171), with a coarse stem and numerous branches, each of which terminates in a yellow or orange (rarely white to red) flower head (fig. 172) from 1/2 inch to 11/2 inches across (Knowles 1958). It may be planted in rows 18 to 40 inches apart, drilled or broadcast in the field with two to six plants per square foot (Knowles and Miller 1965). The seeds ripen and are harvested 120 to 150 days after planting.

Rubis ³⁵ reported the discovery of a thin-hull mutant that produced seeds with about 10 percent more oil than earlier cultivars. The florets in this selection have delayed anther dehiscence (see "Inflorescence"), which lets the plant serve as a male-sterile line and provides a means for producing hybrid safflower.
__________
³⁵RUBIS, D. D. SAFFLOWER BREEDING AND GENETICS IN ARIZONA. Safflower Conf. Proc., 5pp. University of Arizona, Tucson. 1963. [Mimeographed.]

[gfx] FIGURE 171. - Safflower plant showing branching habit, spiny leaves and flowering heads.
FIGURE 172. - Safflower heads in different stages of developement. A, Head after flowering has ended; B, head in full Flower; C, bud just before first florets appear.

Inflorescence:
There may be 15 to 150 flower heads on a plant, each head enclosed in layers of spine-tipped bracts. The head that terminates the main axis of the plant flowers first, then flowering proceeds downward with those flower heads on the lowest branches opening last. On an individual plant, flowering may extend 10 to 40 days.

There may be 20 to 100 florets in a head (Claassen 1950). Those florets on the outside open first, and opening proceeds centripetally for 3 to 5 days. Within the floret, the style is enclosed by five fused anthers attached at the base by short filaments (fig. 173). The floret begins to elongate by sunrise of the day it opens. Anther dehiscence normally occurs within the fused anther tube shortly after sunrise while the style is elongating. If dehiscence occurs before the style elongates, the stigma pushes through a mass of pollen, becomes coated with pollen, and becomes self-fertilized. If dehiscence occurs after the style elongates so that the stigma passes through the anther tube without becoming pollen coated, self-sterility results. Such flowers must be visited by bees that either bring pollen from other pollen-coated stigmas or transfer pollen from within the tip of the anther tube to the stigma. The thin-hull cultivar has this delayed dehiscence and is therefore functionally male-sterile.

Nectar is secreted at the base of the filaments and is highly attractive to bees, although the quality of honey it produces is poor. ³⁶ The bee collects this nectar at the base of the anther tube from the outside rather than through the tube.

Safflower pollen is also highly attractive to bees and is considered an excellent source by beekeeping standards.
__________
³⁶MCGREGOR, S. E., LEVIN, M. D., and RUBIS, D. D. BEE POLLINATION OF SAFFLOWER. Safflower Conf. Proc., 2 pp. University of Arizona, Tucson. 1963. [Mimeographed.]

[gfx] FIGURE 173. - Longitudinal section of safflower floret. A, Floret, x 6; B, filaments and adjoining area, greatly enlarged.

Pollination Requirements:
Safflower is usually considered to be a self-pollinated crop. Claassen (1950), however, reported cross-pollination ranging from zero to 100 percent, although in most of the plants he used, the detectable crossing ranged from 5 to 40 percent. Knowles (1955) reported that some selections give more vigorous progeny if open-pollinated rather than selfed. The necessity for pollen transfer by insects depends largely upon the growth characteristic of the style. If it elongates and thrusts the stigma through and beyond the anther tube before dehiscence of the pollen, then bee visitation to that stigma is necessary for maximum production (Rubis et al. 1966). If, on the other hand, dehiscence occurs before the style elongates, the stigma usually emerges thoroughly coated with pollen, and self-fertilization can result. This condition is most common in current cultivars; however, the description of the "normal" floret in India by Howard et al. (1915) indicates that delayed dehiscence, such as occurs in the thin-hull selection, may have been much more common in earlier cultivars.

Rubis (1970b) proposed a novel way to use bees to create isolation by overstocking the area with honey bee colonies that would so intensively forage an area that outside pollinators would not enter; therefore, no cross-pollination would occur. This has not been tested in practice.

The few reports on the measured value of pollinating insects to safflower are inconsistent. An extensive review of the literature on safflower by Larson (1962) revealed little on its pollination requirement. Plessers 37 reported that absence of insect pollinators caused a reduction of 47.7 and 36.5 percent in two Indian cultivars and 31.8 percent in the American cv. 'WO-14'. At least some of this reduction might be attributed to cage effect. Eckert (1962) reported lower production from plants caged to exclude bees than from open plots for one cultivar that was "somewhat self-sterile" but no difference between treatments in the "self-fertile" cultivar. Boch (1961) reported that during flowering he obtained twice as much seed from plots to which bees had access, as from plots caged to exclude pollinating insects, but again cage effect might have been a contributing factor. Patil and Chavan (1958) found that both temperature and humidity affected seed setting of bagged flowers. Kursell (1939) was reported by Claassen (1950) to have found extensive self-sterility in different lines, which, if true, would indicate that pollinating insects would have had a beneficial effect. McGregor and Hay (1952) gave a brief nod of approval to the value of pollinating insects. Rubis (1970a) indicated that commercial cultivars are from 75 to 95 percent self-fertile, indicating that their production could be improved with an ample pollinator population.

Pollinators:
Not only honey bees but various other bees and other nectar and pollen-feeding insects visit the blossoms. These may contribute in various degrees to pollination of the flower (Levin et al. 1967, Levin and Butler 1966, Butler et al. 1966, Kadam and Patanker 1942), but, in relative numbers, honey bees are by far the most important. No differences have been observed in attractiveness of different cultivars to honey bees. All seem to be attractive.

Probably, the best test of the value of pollinating insects to safflower was conducted at Tucson, Ariz. (Rubis et al. 1966). In this replicated test with two cultivars, the plots exposed to insect pollinator visitation during the flowering period were compared with plots caged under plastic screen. Pollinating insects were excluded from plants of some cages, whereas a functioning colony of honey bees was added to plants of other cages. Two lines of safflower were used: Line A was a selection from the 'Gila' cv., which normally showed about 5 percent outcrossing; line B was a composite of multiple crosses of the thin-hull selection (Rubis 1962), which normally showed about 80 percent outcrossing. The production of line A was increased about 5 percent by bees, whereas production of line B was approximately doubled.

Dennis and Rubis (1966) concluded that the benefits of honey bees or other pollinating insects to commercial cultivars depended on the amount of self-sterility or crossability in a cultivar. They stated that 'Frio' cv. was lower in self-fertility than 'Gila' cv.; therefore, yield increase from pollinating insect activity on the former would be expected to be greater. They concluded that production of 'Gila', even though it was considered to be self-fertile, could be increased 5 percent or more by honey bee pollination. Knowles and Miller ( 1965) apparently were in agreement for they stated that because safflower is not wind-pollinated, the presence of pollinating insects in abundance was necessary for maximum seed set in types that were deficient in "production" of pollen.

Pollination Recommendations and Practices:
Eckert (1959*) recommended two honey bee colonies per acre of safflower, but few, if any, growers take steps to secure this pollinator population.

Because safflower is an excellent source of nectar and pollen, beekeepers frequently place their colonies near safflower plantings, but not in the density recommended by Eckert (1959*). The data indicate that although the safflower has a high degree of fertility, the grower would profit more than the beekeeper would by having a high population of honey bees visiting his safflower blossoms.

LITERATURE CITED:
BOCH, R.
1961. HONEYBEE ACTIVITY ON SAFFLOWER (CARTHAMUS TINCTORIUS L.) Canad. Jour. Plant Sci. 41: 559-562.

BUTLER, G. D., JR., WERNER, E. G., and LEVIN, M. D.
1966. NATIVE BEES ASSOCIATED WITH SAFFLOWER IN SOUTHCENTRAL ARIZONA. Kans. Ent. Soc. Jour. 39(3): 434 - 436.

CLAASSEN, C. E.
1950. NATURAL AND CONTROLLED CROSSING IN SAFFLOWER, CARTHAMUS TINCTORIUS L. Agron. Jour. 42: 381-384.

DENNIS, R. E., and RUBIS, D. D.
1966. SAFFLOWER PRODUCTION IN ARIZONA. Ariz. Coop. Ext. Serv. and Agr. Expt. Sta. Bul. A-47, 24 pp.

ECKERT J. E
1962. THE RELATION OF HONEY BEES TO SAFFLOWER. Amer. Bee Jour. 102: 349-350.

HALLORAN, H. R., and KNEELAND, J. A.
1961. HIGH PROTEIN SAFFLOWER MEAL FOR CHICKENS. West. Feed and Seed 16(11): 23-24, 70.

HOWARD, A., HOWARD G. L. C., and KHAN, A. R.
1915. STUDIES IN INDIAN OILSEEDS. 1. SAFFLOWER AND MUSTARD. India Dept. Agr. Mem. Bot. Ser. 7: 237-272.

KADAM, B. S., and PATANKAR, V. K.
1942. NATURAL CROSS-POLLINATION IN SAFFLOWER. Indian Jour. Genet. and Plant Breed. 2: 69-70.

KLAGES, K. H. W.
1954. SAFFLOWER PRODUCTION. Idaho Agr. Expt. Sta. Bul. 222, 16 pp.

KNOWLES, P. F.
1955. SAFFLOWER-PRODUCTION, PROCESSING AND UTILIZATION. Econ. Bot. 9: 273-299. KNOWLES, P. F. 1958. SAFFLOWER. Adv. in Agron. 10: 289-323.

______and MILLER M. D.
1960. TIPS ON SAFFLOWER GROWING. Calif. Agr. Expt. Sta. Ext. Serv. Leaflet 126.

______and MILLER, M. D.
1965. SAFFLOWER. Calif. Agr. Expt. Sta. Cir. 532, 50 pp.

KURSELL, C.
1939. [BREEDING WORK ON THE NEW OIL PLANT SAFFLOWER.] Pflanzenbau 15: 463-482. [In German.]

LARSON, N. G.
1962. SAFFLOWER, 1900-1960. A LIST OF SELECTED REFERENCES. U.S. Dept. Agr. Natl. Agr. Libr. Libr. List 73 author and subject index, 557 references, 31 pp.

LEVIN, M. D., and BUTLER, G. D. JR.
1966. BEES ASSOCIATED WITH SAFFLOWER IN SOUTH CENTRAL ARIZONA. Jour. Econ. Ent. 59: 654-657.

______BUTLER, G. D. JR., and RUBIS D. D.
1967. POLLINATION OF SAFFLOWER BY INSECTS OTHER THAN HONEY BEES. Jour. Econ. Ent. 60: 1481-1482.

McGREGOR, W. G., and HAY, W. D.
1952. SAFFLOWER CANADIAN EXPERIMENTS. Sci. Agr. 32(4): 204-213.

PATIL, J. A., and CHAVAN, V. M.
1958. SELFING METHODS IN SAFFL0WER. Indian Oilseeds Jour. 2: 10-12.

RUBIS, D. D.
1962. RECESSIVE MUTANT "THIN HULL" DESCRIBED. Agron. Abstracts, p. 75.

______LEVIN, M. D., and MCGREGOR, S. E.
1966. EFFECTS OF HONEY BEE ACTIVITY AND CAGES ON ATTRIBUTES OF THIN-HULL AND NORMAL SAFFLOWER LINES. Crop Sci. 6: 11-14.

_______ 1970a. BREEDING INSECT-POLLINATED CROPS. In The Indispensable Pollinators, Ark. Agr. Ext. Serv. Misc. Pub. 127, pp. 19-24.

______ 1970b. BEE-POLLINATION IN THE PRODUCTION OF HYBRID SAFFLOWER. In The indispensable Pollinators, Ark. Agr. Ext. Serv. Misc. Pub. 127, pp. 43-49.

SHAW, A. F., and JOPPA, L.
1963. SAFFLOWER - AN OILSEED CROP. Mont. Agr. Ext. Serv. Cir. 289, 16 pp.

UNITED STATES DEPARTMENT OF AGRICULTURE.
1961. GROWING SAFFL0WER - AN OILSEED CROP. U.S. Dept. Agr. Farmers' Bul. 2133, 16 pp.

SESAME
Sesamum indicum L. family Pedaliaceae
Sesame, sometimes known as benne, is grown for its edible oil pressed from the seed and for the decorticated (hulled) edible seed (Martin and Leonard 1949*).

World production in 1968 was estimated at 640,000 tons of sesame oil. This would indicate that about 10 million acres were devoted worldwide to this crop. In 1955, about 15,000 acres were grown in the United States, mostly in Texas and New Mexico.³⁸ In the United States, sesame is grown in the Southwestern, Southern, and South Central States. Although there was essentially no production in 1971, there is considerable interest in reviving production. Tests have shown that under extremely favorable conditions as much as 2,000 pounds of seed per acre can be produced. Nonshattering cultivars were developed in 1953 (Kinman 1955).
__________
³⁸ KINMAN, M. L. SESAME PRODUCTION. U. S. Dept. Agr., Agr. Res. Ser., Tex. Agr. Expt. Sta. 1956. [Mimeographed.]

Plant:
Sesame is an annual erect herb, 3 to 5 feet tall, which is grown in rows 18 to 42 inches apart with 6 to 10 plants per foot of row (USDA 1958). The 3- to 5-inch-long leaves are opposite, oblong, and, in the older cultivars, smooth and flat. In the nonshattering cultivars the leaves are cupped and have small leaflike outgrowths on their underside. Some cultivars have many branches, whereas others are relatively unbranched. Thousands of cultivars are known, with lifespans ranging from 2 to 6 months. Sesame is killed by frost; however, seed harvest before frost is preferred. Because large-scale cultivation equipment can be used, growers can handle large acreages. A single plant is capable of producing several thousand seeds (Kinman and Martin 1954).

Inflorescence:
The tubular, pendulant, bell-shaped, two-lipped flower is pale rose to white and 3/4 to 1 inch long. The two lobes of the upper lip are shorter than the three lobes of the lower (Bailey 1949*). One flower is produced at the axil of each leaf. The lower flowers usually begin blooming 2 to 3 months after seeding, and blooming continues for some time until the uppermost flowers are open.

Sesame is a source of nectar and some honey for beekeepers primarily because it flowers in midsummer when little else in the area is blooming. It is an excellent source of pollen for bees. It also attracts various other bees and other insects that feed on its pollen or nectar; however, honey bees are the primary visitors (Langham 1944).

Pollination Requirements:
Sesame is usually considered to be a self-pollinated crop (Kinman and Martin 1954) although the amount of cross-pollination that occurs is considerable. Van Rheenen (1968) recorded 5.5 to 9.6 percent crossing but gave no indication as to what pollinators might be responsible. Langham (1944) obtained an average of 4.6 percent (0.50 to 9.58 percent) outcrossing, which he attributed to honey bees. Martinez and Quilantan (1964) observed 0.15 to 9.39 percent crossing, with higher crossing observed during winter when the bee population was higher. Langham (1944) covered plants to exclude insects and obtained relatively as much seed set as on plants exposed to bee visitation. However, Srivastava and Singh (1968) obtained yield increases of 43.66 percent over the best parent when they crossed Meghna with local cultivars and 38.0 percent when they crossed Meghna with wild plants. This indicated that hybrids might be produced that would outyield current cultivars. A crossing method involving bees might prove quite beneficial.

Pollinators:
Honey bees are the primary visitors to sesame flowers. Langham (1944) stated that the bee alights on the protruding lip of the flower and squeezes inside. Later, it emerges coated with pollen and flies to another flower. However, no benefit from such crossing, although established in many other crops that have been considered self-pollinating, has been established for sesame. The high percentage of heterosis shown by Srivastava and Singh (1968) strongly indicates that insect pollination would be beneficial in the production of superior hybrid seed.

The effect of insect visitation on the individual flower has not been studied.

Pollination Recommendations and Practices:
None

LITERATURE CITED:
KINMAN, M. L.
1955. SESAME. Econ. Bot.9(2): 150.

______and MARTIN, J. A
1954. PRESENT STATUS OF SESAME BREEDING IN THE UNITED STATES. Agron. Jour. 46(1): 24-27.

LANGHAM D. G.
1941. NATURAL AND CONTROLLED POLLINATION IN SESAME. Jour. Hered. 35(8): 254-256.

MARTINEZ, H. G., and QUILANTAN, V. L.
1964. [PERCENT OF NATURAL CROSS-FERTILIZATION OF SESAME IN IGUALA, GROS.] Agr. Tecnologia Mex. 11(4): 175-177. [In Spanish.]

RHEENEN, H. A. VAN.
1968. NATURAL CROSS-FERTILIZATION IN SESAME (SESAMUM INDICUM L.). Trop. Agr. [Trinidad] 45(2): 147-153.

SRIVASTAVA D. P., and SINGH, S. N.
1968. HETEROSIS IN SESAME. Indian Bot. Soc. Jour. 47(1/2): 79-88.

UNITED STATES DEPARTMENT OF AGRICULTURE.
1958. SESAME PRODUCTION. U.S. Dept. Agr. Farmers' Bul. 2119,12 pp.

SISAL AND HENEQUEN ³⁹Agave spp., family Agavaceae
Sisal and henequen are long hard fibers used primarily in cordage (ropes, cords, and twine). They are obtained from the 2- to 4-foot-long leaves of agave plants. Sisal, the most important fiber, is obtained from A. sisalana Perr. ex. Engelm. In 1965, it accounted for 779,000 tons or 85 percent of the world supply. Henequen is obtained from A. fourcroydes Lem., and it accounts for practically all of the remaining fibers produced. Sisal is produced in Tanzania, Brazil, Angola, Madagascar, and Haiti. In 1963, almost 1.5 million acres were devoted to sisal production in Africa and more than 300,000 acres in Brazil. Henequen, which is a much weaker fiber than sisal but which has a certain market, is produced primarily in Mexico.
__________
³⁹ Material in this section was largely taken from Lock (1962) and Wienk (1969).

Plant:
The plant has stiff, heavy, persistent leaves, 2 to 4 feet long, 4 to 8 inches wide, and 1 to 4 inches thick that are basal or come from a short stem, 3 to 6 feet long. The flower stalk is a towering spike or panicle, 6 to 15 feet above the rosette of leaves. The plant grows slowly, attaining a height of only 6 inches 9 months after planting and 2 feet at the end of 2 years. It is about full grown at 4 years when its stem is about 8 inches in diameter and the harvesting of the lower leaves begins. An average of 185 leaves may be harvested before leaf growth ceases and the flower stalk or "pole," which resembles a giant asparagus sprout, shoots rapidly upward. From first appearance of the pole through flowering, fruiting, and death of the entire plant covers a span of about 6 months. About 100 plants per acre are maintained for maximum production of fiber.

Inflorescence:

The 11/2- to 21/2-inch, pale-green, funnel-shaped flower is made up of six narrow, united lobes. Six long stamens come from the base of the corolla and surround the ovary has three locules with two series of ovules in each, which develops into a green, fleshy capsule about 2 inches long, turning black at ripening. This capsule may have as many as 300 ovules but usually less than 100 seeds. The fertile seeds are triangular, black, and hard; the unfertilized ovules produce white, papery, nonviable seeds.

Flowering of the floret commences with the extrusion to 2 inches of the six anthers from the apex of the bud 36 to 48 hours before they release pollen and 3 to 4 days before the stigma becomes receptive. The anthers begin to dehisce early in the afternoon, and by next morning all pollen has been released. The style begins to elongate and becomes receptive, but by then the stamens have withered and hang limp. One to 2 days after fertilization, the style withers and ovarian development begins. All flowers on a branch of the panicle do not open at once; therefore, pollen from newly opened flowers can pollinate those that opened earlier. Flowering covers several weeks as it moves from the bottom to the top of the pole.

The flower produces large quantities of nectar and a rather heavy, yellow, strong-smelling pollen, both of which are highly attractive to bees. This pollen is usually completely removed from the anthers by bees before the stigma becomes receptive. The honey produced from agaves is generally of inferior quality with a strong unpleasant aroma, strong taste, and dark color.

Pollination Requirements:
Propagation of the agaves is mainly by bulbils or suckers. The grower prefers this method because it enables him to maintain pure lines. However, where seed production is desired, cross-pollination is necessary. The pollen is released within a flower before the stigma is receptive; therefore, for fertilization to occur, pollen must be transferred to other flowers with receptive stigmas. Because of the large number of ovules in the ovary, numerous pollen grains must be deposited on the relatively small stigma. The heavy pollen is not a wind-carried type, nor would gravity be likely to account for the pollination of the numerous ovules of a flower.

Pollinators:
Bees, and particularly honey bees, are the primary pollinators. For maximum seed production, the grower should consider building up the bee population in the area.

Pollination Recommendations and Practices:
None.

LITERATURE CITED:
LOCK, G. W.
1962. SISAL. 322 pp. Longmans, Green and Co., Ltd., London.

WIENK, J. F.
1969. LONG FIBER AGAVES. Pp. 1-21. In Ferwerda, F. P., and Wit, F., eds., Outlines of Perennial Crop Breeding in the Tropics. H. Veenman and Zonen, N. V. Wageningen, The Netherlands.

SUNFLOWER
Helianthus annuus L., family Compositae
There are two types of annual sunflower, oilseed and nonoil. Oilseed sunflower is a source of high-quality seed oil, which is used in cooking, salads, paints, and as an industrial lubricant. Seed from the nonoil type is used as a bird feed or roasted and marketed as a confectionery product. Until 1972, most of the sunflower acreage in the United States was of the nonoil type.

Minnesota and North Dakota are the major producing States. The crop is being tried in many other States to a limited extent. The oil from cultivars grown in Minnesota contains more of the desirable linoleic acid than the oil from the same cultivars grown in Central or Southern United States.

Other major sunflower producing countries are Russia (11 to 12 million acres), Argentina (3 million acres), and Romania (11/2 million acres). There were 850,000 acres in the United States in 1972 (Robinson 1973).

Plant:
The sunflower is a widely adapted plant. It will grow in the arid Southwest, yet at some stages of growth it will tolerate light frosts. It is native to North America and is the only important annual crop to evolve and be domesticated within the confines of the United states. Little heed was given it, however, until it was transported to Europe and returned, via Canada, as an important oilseed crop. Not until 1947 did oilseed production develop in the United States, in Minnesota. Most of our cultivars originated in Russia, or they have been developed from Russian cultivars. However, Kinman and Earle (1964) showed that some of the best American cultivars outyielded the best Russian cultivars in tests with comparative linoleic acid values.

When handled as a row crop, most growers prefer row widths of 20 to 36 inches with 15,000 to 30,000 plants per acre. The exact plant population most desirable depends upon the type grown, rainfall, temperature, and soil fertility. Average production is about 1,100 pounds seed per acre (Anonymous 1969), although much higher production has been reported by individuals (Killinger 1968, Noetzel⁴³, Trotter and Giran 1970, Weibel etal. 1950).
__________
⁴³ NOETZEL, D. M. 1968 INSECT POLLINATION RESULTS ON SUNFLOWERS. Pp. 108 - 112. Dept. Ent., N. Dak. State Univ., Fargo. 1968. [Mimeographed.]

Inflorescence:
The sunflower stalk has a main or primary flower head and may have one to several secondary heads (fig. 179). All commercial cultivars are composed of almost 100 percent single-headed plants. The main head may have from 1,000 to 4,000 individual florets, depending upon the cultivar and size of the head. A secondary head may have from 500 to 1,500 florets. The outer or ray florets with the showy yellow petals are sterile, having neither stamens nor pistil (fig. 180). The less conspicuous florets making up most of the head are hermaphrodite, but protandrous, and many are self-incompatible. They are normally open 2 or more days. The first day, the anthers release their pollen in the anther tube, which is partly exserted from the corolla. The pollen is collected freely by bees, along with the nectar at the flower base. The second day, the stigma pushes up through any pollen mass remaining, then its two lobes open outward, receptive to pollen but out of reach of its own pollen (Putt 1940).

The opening of all the florets of a single head takes 5 to 10 days. If pollination occurs, the floret withers shortly; otherwise it may "wait" as long as 2 weeks for fertilization. Seed setting on such florets, however, is greatly reduced even with cross-pollination (Avetisyan 1965, pp. 209- 248). A typical head in bloom will have dried florets toward the outside, then a ring of florets with receptive stigmas, a ring of florets shedding pollen, and, finally, unopened florets toward the center. Radaeva (1954) stated that within 3 days after the first flower head begins to open 83 percent of the remaining heads also begin to open.

Both the pollen and the nectar of sunflower are quite attractive to bees throughout the day (Bitkolov 1961, Free 1964). Nectar is secreted at the base of the floret, primarily during the pollen-producing stage of flowering but to some degree while the stigmas are receptive. Extrafloral nectaries in the bracts and on the upper leaves of the plant are sometimes visited by honey bees, particularly in the afternoon (Free 1964).

Sunflowers are considered by beekeepers to be a fair source of pleasant-flavored, yellow-colored honey (Anonymous 1969, Burmistrov 1965). Furgala (1954a) reported that a colony on scales gained 104 pounds in 15 days while on sunflowers, which he considered an indication that the area was underpopulated for adequate sunflower pollination. Baculinschi (1957) calculated the nectar crop at about 20 lb/acre for the entire flowering period. This is roughly equal to nectar production of cantaloupe, as calculated by McGregor and Todd (1952*). Guynn and Jaycox (1973) reported a yield of 80 pounds of honey per colony when 15 colonies were placed at the center of a 45-acre field of sunflowers.

[gfx] FIGURE 179. - Sunflower head, showing the brilliant yellow but sterile ray flowers around the outside, and the fertile florets in different stages of development in the center.
FIGURE 180. - Longitudinal section of sunflower head, x 1/2, with individual florets. A, Ray floret, x 5; B, disk florets in different stages of developement, x 5.

Pollination Requirements:
If pollen from a floret is transferred to the receptive stigma of another flower on the same head and the flowers are self-compatible, seed will set. However, most cultivars are self-incompatible (Free and Simpson 1964), in which case the pollen must come from another plant. When selfing (within the head) occurs, seed setting is usually low (Barrett 1954, Cardon 1922, Kalton 1951), the seeds are undersized, and the oil content and germination are reduced. Selfed seed also sprout more slowly, and production from them is lower than from plants derived from crossed seed. Putt and Heiser (1966) described two types of genetic male sterility in sunflowers.

Putt (1966) observed strong hybrid vigor in sunflowers; 2,291 lb/acre with 33 percent oil in his crosses as compared to 1,156 lb/acre of seed with oil content of 30 percent in his inbreds. Some "hybrid" seed has been produced by use of either a self-incompatible line or one producing little viable pollen, which is pollinated by a line that is a copious producer of viable pollen (Robinson et al. 1961). Genetic male sterility is also used to produce commercial F₁hybrid seed, but full exploitation of the hybrid vigor possible in sunflower will come when cytoplasmic male sterile female parents can be crossed with male parents that will restore fertility to this hybrid (Kinman 1970).

Sources of cytoplasmic male sterility (Leclercq 1970) and fertility restoration (Kinman 1970) have recently been discovered, and preliminary tests indicate that the yields of the F_l hybrids produced by this method are comparable to yields of the best hybrids produced by other means.

When Pritsch (1965) compared seed set in cages with and without bees, his results were inconclusive. However, others have shown a strong beneficial effect from insect pollination. Sanduleac (1960) compared seed production from four pollination treatments: (1) Isolated under gauze; (2) isolated under gauze and hand pollinated four times; (3) open to insect pollination; and (4) open to insect pollination and hand pollinated four times. His production was always greatest on open plots, with no increase derived from the extra hand pollination. His pollinator population, however, was not indicated. Avetisyan (1965) said that for best seed production each floret should be visited 8 to 10 times. Habura (1957) also obtained a higher set of seed from cross-pollinated flowers than from selfed flowers. Pinthus (1959) stated that a higher percentage of selfed seed was produced under lower temperatures at flowering. He showed that although 50 to 60 percent of the seed of bagged flowers developed between October 15 and November 30, only 0.4 to 5.0 percent developed between June 10 and September 25, during the normal period of seed production. Luttso (1956) also compared set of seed on 10 heads exposed to three treatments. His results in grams of seed produced were: Caged without bees, 315 g; exposed to bees, 995 g; and exposed to bees in addition to supplemental pollination of the heads with a soft mitten, 1,000 g. He showed that the labor of 1 man-day per acre was about equal to the bee pollination services. Schelotto and Pereyras (1971) He showed that sunflower seed yield in Argentina was increased five to six times, and the oil content of the seed was increased 25 percent in plots exposed to honey bee colonies as compared to plots isolated from insects during flowering.

Furgala (1970) reviewed the effects of insect pollination on seed production and urged that more research be conducted. The evidence is strong, however, that insect pollination is needed for commercial production of sunflower seed. Posey (1969) compiled 133 pages of literature citations on sunflowers from January 1960 to June 1967. Only 24 of these references concerned sunflower pollination, and only two of the 24 were published in the English language. This indicates the lack of interest in the pollination of this crop outside Russia.

Pollinators:
Honey bees are the primary pollinating agents of sunflowers almost wherever they are grown (Cardon 1922, Cirnu 1960, Fomina 1961, Glukhov 1955, Overseas Food Corporation 1950, Pritsch 1965, Radaeva 1954). If there is a shortage of honey bees in the sunflower fields, a small seed crop is harvested.

In some localities, bumble bees and sundry other wild bees visit sunflowers (Cockerell 1914). Arnason (1966) indicated that in many instances these bees are adequate, but all other researchers have shown that the bulk of the pollination in commercial sunflower production is by honey bees. Radaeva (1954) showed that honey bees are much more effective than wild insects. The evidence indicates that if sufficient flower heads are available for bees to collect honey surplus to the needs of the colony, the ratio of bees per flower is inadequate for maximum seed production.

The exact number of bees needed for maximum pollination has not been determined. Measurements of bee activity have included bees per flower head, bees per unit of row, bees per acre, and colonies per acre. These have been correlated to some degree with seed production. For example, Noetzel⁴⁴ placed one colony per acre at the ends of different sunflower fields in North Dakota. He counted the bees per head and obtained seed production data at different distances from the apiary. He obtained an overall increase of 20 percent due to the bees alone, but beyond 400 feet from the apiaries he got no measurable increase in yield.

Noetzel's production ranged from 1,350 to 4,962 lb/acre within 50 feet of the apiary as compared to the range of from 734 to 2,249 lb/acre at 1,300 feet from the apiary. Robinson et al. (1961) harvested 1,231 to 1,653 lb/acre from their plots in Minnesota but gave no indication of the pollinator population in the plots. Bees were not provided to the plots, but apparently they were abundant in the vicinity. Furgala (1954b) reported that three to five colonies per acre significantly increased yield. Furgala (1954a) reported that a field produced 1,300 lb/acre near the apiary, 900 lb/acre at a distance of 400 feet, and 800 lb/acre at 1,000 feet, whereas the field not supplied with bees produced about 700 lb/acre at all sites.

Alex (1957) obtained only 311 pounds per acre in cages without bees, 602 pounds from cages with honey bees present, and 931 pounds from open plots freely visited by bees, differences that were significant at the 5 percent level of confidence. All were relatively low yields, partly due to drought conditions. Blackman (1951) stated that a shortage of bees can be a limiting factor in maximum seed production. Glukhov (1955) showed the following correlation between honey bee visits and production of seed:

[gfx] fix table:

Number of honey bee kilograms of seed visits per floret from a million florets 1.0 53 1.4 76 3.4 133 6.0 210 10.0 210

This gives support to the recommendation of Avetisyan (1965) that each floret should receive 8 to 10 bee visits.

Kushnir (1960) obtained 1,696 g/8 m² plot at 400 m from the apiary, 1,373 g at 2,000 m, and 266 g from isolated plots. He had shown earlier, however, that weight of seed was not the entire story. For example, he (Kushnir 1957) showed that 100 bee-pollinated seed weighed 9.27 g and had 86.9 percent germination, whereas 100 selfed seed weighed only 2.98 g with only 9.2 percent germination. In shape, the selfed heads were uneven. Later, he (Kushnir 1958) found in another test that the kernels from 100 seeds from bee-pollinated flowers weighed 5.7 g, whereas kernels from 100 selfed seed weighed only 0.5 g.

Kurennoi (1957) likewise found that seed production 400 m from the apiary with 5.1 bees per flower was 1.81 tons /ha (1,465 lb/acre); at 2,000 m, production with 2.9 bees per flower was 1.77 tons/ha (1,384 lb/acre).

Kovalev and Ul'yanichev (1961) studied the effect of bee pollination on 12,000 ha of sunflower. They found that better pollination accounted for an overall 14.5 percent increase in yield of seed.

Cirnu (1960) stated that bee pollination increased production of seed 10 to 30 percent; Cirnu and Sanduleac (1965) reported that use of one colony per hectare on 5,582 ha increased seed production 21 to 27 percent, with bees brought in when 5 percent of the plants were in bloom.

Ponomareva (1958) conducted large-scale experiments with bees during 1950-56 on 66 sunflower farms in various Russian zones. When one colony per hectare was placed by the fields, the bees worked "sufficiently well at pollination and honey-gathering within a radius of 500 m of the apiary." Beyond that distance, seed production dropped rapidly. Overall, the fields that were supplied with bees produced 79 percent more seed than fields not supplied with bees. In 22 districts, where one-half colony per hectare was used, production was 890 kg/ha of seed; whereas, in 19 districts, where one colony per hectare was used, seed production was 1,270 kg/ha - an average increase of more than 42 percent. The number of colonies necessary for maximum production was not determined.

Lecomte (1962) counted 108 bees on 100 sunflower heads in the morning and 100 to 115 per 100 heads of cv. 'B-65-40' in the afternoon, and calculated the bee population in the field at 100,000 per hectare. This would indicate about one bee per head and 100,000 heads per hectare. If these heads contained 1,000 to 4,000 florets each and required 10 days to open, there would be an average of 10 to 40 million new florets per day.

Avetisyan (1965) calculated that only 2 million florets per hectare were available daily, and that each floret should be visited eight times; therefore, 16,000,000 bee visits per hectare per day are required. He further calculated that a foraging bee will visit 1,080 florets per day; therefore, 15,000 foragers could pollinate 1 ha. He reasoned that a colony with 6 kg of honey bees would supply enough foragers to adequately pollinate 2 ha of sunflower, providing the bees have nothing else to work on. This population is far below the 100,000 bees, or one bee per sunflower head, observed by Lecomte, and would seem to be an inadequate forager population for highest seed production.

Avetisyan (1965) based his recommendation on the assumption that there was nothing else for the bees to visit except sunflower, that each forager visits sunflower, and that the number of florets is the same each day. Actually, there are almost always competing plants, the number of florets is greater than the average during the peak of flowering, and many colonies do not contain 6 kg of bees. Allowances should be made for such differences when recommendations are issued. This need for allowances is supported by Benedek et al. (1972) who studied the relation of colony numbers, density of bees on the sunflowers, and seed production. They concluded that seed production is dependent upon density of honey bees on the flowers, but that many factors override the effect of colony concentration around the sunflower field and seed production.

To prevent the sharp decline in production with distance from the apiary, the Russians recommended "converging or saturation pollination" or the distribution of the apiaries in such a way that equal distribution of bees throughout the field is obtained. This is basically the method advocated by Todd and Crawford (1962) of distributing the bees every tenth of a mile in each direction in the field, a method utilized in most alfalfa fields today in Western United States.

Khalifman (1959) stated that heavy honey bee visits to sunflowers not only increased seed set but also limited the damage by the sunflower moth (Homoeosoma nebulella Denis & Schiffermuller), a delayed effect called hysteresis. Martin (1968) considered H. electellum (Hurst) the most serious pest of sunflower in South Carolina. Teetes and Randolph (1970) stated that the period of greatest sunflower moth oviposition was the third to sixth day after the sunflower head began to bloom. This is when the pollinating agents should also be at their peak; however, the use of insecticides highly toxic to honey bees is recommended for this same period. Both control of this harmful moth and the use of insect pollinators are necessary for production of sunflower seed. Unless the programs are so arranged that both are successful, seed production is doomed to failure.
__________
⁴⁴ See footnote 43, p. 345

Pollination Recommendations and Practices:
All research on sunflower pollination indicates that honey bees are the primary pollinating agents, and that colonies should be provided to the field (Barbier and Abid 1966), and that they should be protected from harmful pesticides while they are in the field. The bees should be ready for the pollination task at the onset of flowering. The total flowering period is usually about 20 days, but 83 percent of the heads begin to open within 3 days after the first head opens. Evidence also indicates that the highest bee population and the highest production occur within a few hundred feet of the apiary. If adequate pollination throughout the field is provided, there should be no significant gradient of seed set in relation to apiary location.

The term "saturation pollination," meaning the patterned distribution of groups of colonies, sometimes used on other crops to provide adequate coverage throughout the entire field, is equally applicable and needed in sunflower production. Cirnu (1960) recommended two colonies per hectare, the bees to be moved in at 3 to 5 percent of bloom. Later, Cirnu and Sanduleac (1965) recommended one colony per hectare. Furgala (1954b) recommended one colony per acre, the colonies placed in rows 300 to 400 yards apart. Smith et al. (1971f recommended one-half colony per acre. The evidence is plain that, if the grower wants maximum seed production, he should not skimp on the use of bees.

The number of colonies per acre alone is not too meaningful. Distribution of colonies to give thorough coverage of all blooms is highly important, and strength and other conditions of the colony are equally important. The criterion the grower should use is the bee visits per floret or bees per head throughout his field. The presence of one bee per head throughout the day should provide adequate visitation, but additional research is needed to determine the exact bee population needed for maximum production of sunflower seed.

LITERATURE CITED:
ANONYMOUS.
1969. MINNESOTA SUNFLOWER PRODUCTION 1968. Minn. Beekeeping Mag. Leaflet.

ALEX, A. H.
1957. POLLINATION OF SOME OILSEED CROPS BY HONEY BEES. Tex. Agr. Expt. Sta. Prog. Rpt. 1960, 5 pp.

ARNASON, A. P.
1966. RECENT STUDIES IN CANADA OF CROP POLLINATION BY INSECTS. In 2d Internatl. Symposium on Pollination, London. Bee World 1964: (supp.) 107-124.

AVETISYAN, G. A.
1965. [BEE POLLINATION OF AGRICULTURAL CROPS.] His Pchelovodstvo, pp. 209-248. Moskva, Kolos. [In Russian.]

BACULINSCHI, H.
1957. [NECTAR PRODUCTION ON SUNFLOWERS IN THE STEEPE REGION.] Apicultura 30: 9-10. [ In Romanian. ] AA-305 /60.

BARBIER, E. [C.], and ABID, M.
1966. POLLINATION AND SEED FORMATION IN SUNFLOWER. Al Awamia 21: 63-83.

BARRETT, C. F.
1954. SUNFLOWER POLLINATION. Ent. Soc. Manitoba Proc. 10: 25-28.

BENEDEK, P., MANNINGER, S., and NAGY, B.
1972. THE NUMBER OF COL0NIES AND THE DENSITY OF HONEYBEES IN SUNFLOWER FIELDS IN RELATION TO THE POLLINATION OF THE CROP. Zeitschr. f. Angew. Ent. 71: 385-389.

BITKOLOV, R. S.
1961. [SUNFLOWER AND BEES.] Pchelovodstvo 38(5): 20-21. [In Russian.]

BLACKMAN, M. A.
1951. THE SUNFLOWER. World Crops 3: 51-53.

BURMISTROV, A. N.
1965. THE MELLIFEROUS VALUE OF SOME SUNFLOWER VARIETIES. In 20th Internatl. Apic. Cong. Proc., Bucharest, pp. 320-323.

CARDON, P. V.
1922. SUNFLOWER STUDIES. Amer. Soc. Agron. Jour. 14: 69-72.

CIRNU. I.
1960. [RESULTS OF BEE POLLINATION OF SUNFLOWERS.] Apicultura 33(1): 18-20. [In Romanian.] AA-444l63.

______and SANDULEAC, E.
1965. [THE ECONOMIC EFFICIENCY OF THE SUNFLOWER (HELIANTHUS ANNUUS) POLLINATION WITH THE AID OF THE BEES.] Lucr. Stiint. Stat. Cent. Seri. Apic. 5: 37 - 51. [In Romanian, English summary.]

COCKERELL, T. D. A.
1914. BEES VISITING HELIANTHUS. Canad. Ent. 46: 409-415.

FOMINA, K. Y.
1961. [THE INFLUENCE OF A FIELD-PROTECTIVE FORESTATION ON THE NECTAR PRODUCTIVITY AND SEED YIELD OF SAINFOIN AND THE SUNFLOWER.] In Moskov. Ordena Lenina Sel'skokhoz. Akad. im. KA. Timiryazeva. Dok. TSKhA, 62: 531-536. [In Russian.]

FREE, J. B.
1964. THE BEHAVIOUR OF HONEYBEES ON SUNFLOWERS (HELIANTHUS ANNUUS L.). Jour. Appl. Ecol. 1(1): 19-27.

______and SIMPSON, J.
1964. THE POLLINATION REQUIREMENTS OF SUNFLOWERS (HELIANTHUS ANNUUS L.). Empire Jour. Expt. Agr. 32(128): 340-342.

FURGALA B.
1954a. HONEY BEES INCREASE SEED YIELDS OF CULTIVATED SUNFLOWERS. Gleanings Bee Cult. 82: 532-534.

______ 1954b. THE EFFECT OF THE HONEY BEE, APIS MELLIFERA L., ON THE SEED SET, YIELD AND HYBRIDIZATION OF THE CULTIVATED SUNFLOWER, HELIANTHUS ANNUUS L. Ent. Soc. Manitoba Proc. 10: 28-29.

FURGALA, B .
1970. SUNFLOWER POLLINATIONÑA NEGLECTED RESEARCH PROBLEM AREA. In The Indispensable Pollinators, Ark. Agr. Ext. Serv. Misc. Pub. 127, pp. 37-42.

GLUKHOV, M. M.
1955. [HONEY PLANTS.] 512 pp. Izd. 6, Perer. i Dop. Moskva, Gos. lZd-vo. Selkhoz Lit-ry. [In Russian.]

GUYNN, G., and JAYCOX, E. R.
1973. OBSERVATTONS ON SUNFLOWER POLLINATION IN ILLINOIS. Amer. Bee Jour. 113: 168-169.

HABURA, E. C.
1957. [SELF AND CROSS STERILITY IN SUNFLOWERS.] Ztschr. f. Pflanzenzucht 37: 280-298. [In German.]

KALTON, R. R.
1951. EFFICIENCY OF VARIOUS BAGGING MATERIALS FOR EFFECTING SELF-FERTILIZATION OF SUNFLOWERS. Agron. Jour. 43: 328-331.

KHALIFMAN, I. A.
1959. HETEROSIS IN PLANTS AS THE AFTER-EFFECTS OF POLLINATION BY BEES (HYSTERESIS).1969 Bee World 40: 303-313.

KILLINGER G. B.
1968 NEW AGRONOMIC CROPS FOR FLORIDA. Fla. Agr. Expt. Sta. Sunshine State Agr. Res. Rpt.13(4): 3-5.

KINMAN M. L.
1970. GREETINGS FROM MURRAY L. KINMAN. In 4th Internatl. Sunflower Conf. Proc.:181-183 Memphis.

______and EARLE, F. R.
1964. AGRONOMIC PERFORMANCE AND CHEMICAL COMPOSITION OF THE SEED OF SUNFLOWER HYBRIDS AND INTRODUCED VARIETIES. Crop Sci. 4: 417-420.

KOVALEV A. M., and UL'YANICHEV E. M.
1961. [REGISTRATION EXPERIMENT ON ADDITIONAL SUNFLOWER YIELD FROM BEE POLLINATION.] Pchelovodstvo 7: 7-11. [In Russian.]

KURENNOI, N. M.
1957. [EXPERIMENT ON INCREASING THE EFFECTIVENESS OF SUNFLOWER POLLINATION BY BEES.] Pchelovodstvo 34(9): 42-48. [In Russian.] AA-136/59.

KUSHNIR, L. G.
1957. [ECONOMIC EFFECTIVENESS OF POLLINATION OF SUNFLOWER BY BEES.] Pchelovodstv

______ 1958. [THE BIOLOGICAL EFFECTIVENESS OF SUNFLOWER POLLINATION BY VARIOUS METHODS.] Dokl. TSKhA 36: 81-88 [In Russian.] AA-230/60.

______ 1960. [ECONOMIC ESTIMATION OF SUNFLOWER POLLINATION WITH THE HELP OF BEES AND BY HAND.] Pchelovodstovo 37(1): 22-25. [In Russian.] AA-445/63

LECLERCQ, P.
1970. SUNFLOWER HYBRIDS USING MALE STERILITY. In 4th Internatl. Sunflower Conf. Proc.:123-126. Memphis.

LECOMTE, J.
1962. [OBSERVATIONS ON THE POLLINATION OF THE SUNFLOWER (HELIANTHUS ANNUUS L.).] Ann. de l'Abeille 5(1): 69-73. [In French.] AA-938/63.

LUTTSO, V. P.
1956. [SUNFLOWER POLLINATION BY BEES.] In Krishchunas, I. V., and Gubin, A. F., Pollination of Agricultural Plants. Moskva, Gos. Izd-vo. Selkhoz Lit-ry, pp. 45-52. [In Russian.]

MARTIN, J. A.
1968. SUNFLOWER MAY BE OIL POTENTIAL. S.C. Agr. Expt. Sta. 15(1, 2): 20.

OVERSEAS FOOD CORPORATION.
1950. BEES AND SUNFLOWERS. Rpt. for 1949-50: 93-94, and App. 5, part 2: 105-109. Overseas Food Corp., London.

PINTHUS, M. J.
1959. SEED SET OF SELF-FERTILIZED SUNFLOWER HEADS. Agron. Jour. 51: 626.

PONOMAREVA, E. G.
1958. [RESULTS OF MASS EXPERIMENTS ON THE USE OF BEES AS POLLINATORS OF ENTOMOPHILIC AGRTCULTURAL PLANTS.] Biul. Nauthno-Tekh. Inform (Nauchno-Issled. Inst. Pchelovod.) 3-4: 27-28. [In Russian.]

POSEY, M. H.
1969. SUNFLOWER - A LITERATURE SURVEY, JANUARY l960-JUNE 1967. U.S. Dept. Agr., Natl. Agr. Libr. List 95, 133 pp.

PRITSCH, G.
1965. [INCREASING THE YIELD OF OIL PLANTS BY USING HONEY BEES.] Ved. Prace Vyakum. Ustav. Vcelar CSAZV 4: 157-163. [In German.] AA-770/66.

PUTT, E. E.
1940. OBSERVATIONS ON MORPHOLOGICAL CHARACTER AND FLOWERING PROCESSES IN THE SUNFLOWER (HELIANTHUS ANNUUS L.). Sci. Agr. 21: 167-179.

______ 1966. HETEROSIS, COMBINING ABILITY AND PREDICTED SYNTHETICS FROM A DIALLEL CROSS IN SUNFLOWERS (HELIANTHUS ANNUUS L.). Canad. Jour. Plant Sci. 46: 59-67.

______and HEISER, C. B., JR.
1966. MALE STERILITY AND PARTIAL MALE STERILITY IN SUNFLOWERS. Crop Sci. 6: 165-168.

RADAEVA, E. N.
1954. [BEE POLLINATION INCREASES THE YIELD OF SUNFLOWER SEEDS (HELIANTHUS ANNUUS L.).] Pchelovodstvo (2): 33-38. [In Russian.] AA-156/5

ROBINSON, R. G.
1973. THE SUNFLOWER CROP MINNESOTA. Minn. Agr. Ext. Serv. Bul. 299, rev.,26 pp.

______JOHNSON, F. K., and SOINE, O. C.
1961. THE SUNFLOWER CROP OF MINNESOTA. Minn. Agr. Ext. Serv. Bul. 299, 22 pp.

SANDULEAC, E.
1960. [INSECT POLLINATION OF THE SUNFLOWER.] Lucr. Stiint. Stat. Cent. Seri. Apic. 2: 209-218. [In Romanian, English summary.]

SCHELOTTO, B. and PEREYRAS, N. L.
1971. [AN EVALUATION OF THE ECONOMIC SIGNIFICANC OF POLLINATING SUNFLOWER WITH BEES.] Ciencia y Abejas 1(2): 7-25. [In Spanish, English summary.]

SMITH, H., PANKIW, P., KREUTZER, G., and others.
1971. HONEY BEE POLLINATION IN MANITOBA. Manitoba Dept. Agr. Pub. 525,16 pp.

TEETES, G. L., and RANDOLPH, N. M.
1970. ECOLOGY AND CONTRoL OF THE SUNFLOWER MOTH HOMOEOSOMA ELECTELLEM (HULST) In 4th Internatl. Sunflower Conf. Proc., pp. 184-186. Memphis.

TODD F. E., and CRAWFORD, N. R.
1962. THE RELATION OF LOCATION OF HONEY BEE COLONIES TO ALFALFA SEED SET. In 1st Internatl. Symposium Pollination Proc. Copenhagen, Aug. 1960. Commun. 7, Swedish Seed Growers' Assoc., pp. 78-85.

TROTTER, W., and GIVAN, W.
1970. ECONOMICS OF PRODUCING SUNFLOWERS FOR OIL IN THE UNITED STATES. In 4th Internatl. Sunflower Conf. Proc.: 23-34. Memphis.

WEIBEL, R. O., ROBINSON, R. G., and SOINE, O. C.
1950. ILLINOIS AND MINNESOTA TAKE A LOOK AT SUNFLOWERS. Crops and Soils 3(1): 18-19.

TEA
Camellia sinensis (L.) O. kuntze, family Theaceae
Tea is an evergreen shrub widely cultivated throughout the tropics and subtropics especially in hilly or mountainous regions for its tender leaves that are dried and used for a mildly stimulating beverage. Asia produces more than a billion pounds annually; Africa, more than 200 million pounds; and South America, more than 26 million pounds. We import about 140 million pounds from Africa, 100 million pounds from Asia, and 3 million pounds from South America (Purseglove 1968*). Harler (1969) stated that about 970 pounds per acre were produced in northeastern India.

Tea growing was tried in South Carolina over a century ago on about 300 acres, and, even though it grew well, its production was not economical so it was discontinued (Mitchell 1907).

Plant:
Tea prospers in areas with a moderate temperature, high humidity, and moderate to high rainfall. It is killed by frost. Under cultivation, it is usually kept pruned to a spreading shrub 2 to 5 feet in height, with about 2,000 plants per acre. A plant may live 40 to 100 years, its shoots (the bud and two tender leaves) can be plucked each 7 to 14 days, 4 pounds of which yield 1 pound of dried or "made" tea. Mature plants annually yield about 1,000 pounds of made tea per acre.

The fruit is a thick-walled, brownish-green, three-lobed, and usually three-celled capsule, 3/4 to 1 inch in diameter. Upon maturity, 9 to 12 months after flowering, it splits from the apex to release the 1- to 11/2-cm long seeds.

Tea is planted from seed, the estimated acreage planted annually ranging from 20,000 to 50,000 acres. Planting is at the rate of about 40 pounds of seed per acre. When seed is produced commercially, only 70 to 100 trees per acre are maintained instead of the 2,000 used for production of leaves. Production of 1,000 pounds seed per acre is considered a conservative estimate. This means that annually from 800 to 2,000 acres must be devoted to production of seed.

Inflorescence:
The fragrant flowers, 2.5 to 4.0 cm in length, are axillary, solitary, or in clusters of two to four flowers. They have five to seven white or pink-tinged petals, numerous 1/2-inch-long stamens, with three to five stigmatic lobes of the style about level with the anthers. According to Free (1970*), the flower opens in the afternoon and remains open for 2 days.

Pollination Requirements:
The flowers are pollinated by insects. Tea is virtually self-sterile and almost entirely cross-pollinated (Purseglove 1968*, Wight and Barua 1939, Wu 1967). Kutubidze (1958) reported that supplementary pollination produced more, larger, and heavier capsules, better viability, and a higher grade of seed. Simura and Oosone (1956), in studying the embryology of the tea plant, noted that, as in many other plants, self-pollen grows much more slowly in the style than foreign pollen. Tomo et al. (1956) also concluded that tea was highly self-incompatible, largely due to inhibition of pollen tube growth at the tip of the ovary. Kutubidze (1958) noticed that supplementary pollination of hybrid and commercial strains by mixed pollen of other plants of the same strain increased set of fruit and size of capsule. Bakhtadze (1932) reported that isolated plants had an 85- to 95- percent reduction in seed set. Self-pollination did not help to increase set, and, furthermore, only 34-percent germination resulted from selfed seed, whereas crossed seed had 75-percent germination. A greater percentage of the crossed seeds developed into plants that reached maturity, and these plants were more vigorous than the selfed plants. Harler (1964) stated that only about 2 percent of the tea flowers on a tree produced seed, although by artificial pollination this can be raised to 14 percent. He concluded that, to get even 2 percent, at least nine random trees are needed for cross-pollination. Pollinating agents were not mentioned.

Bakhtadze (1932) stated that bees are the chief pollinating agents of tea, but that there were not enough bees present in his area to effect complete pollination of all the flowers. He made no mention of bringing in pollinators to the crop. Kutubidze (1964) recommended that steps be taken to obtain additional cross-pollination for increasing yield and quality of tea.

Pollination Recommendations and Practices:
None, although the evidence is sufficiently strong to recommend the building up of pollinators in tea seed fields.

LITERATURE CITED:
BAKHTADZE, K.
1932. POLLINATION OF TEA, CAMELIA SINENSIS, IN GEORGIA. Subtropics 2(12): 63-80.

HARBER, C. R.
1964. THE CULTURE AND MARKETING OF TEA. 262 pp. Oxford University Press, London.

______ 1969. UNPRUNED TEA IN UPPER ASSAM. Investors' Guardian 213: 1020-1022.

KUTUBIDZE, V. V.
1958. [INTRAVARIETAL SUPPLEMENTARY POLLINATION OF THE TEA PLANT WITH A POLLEN MIXTURE.] Agrobiologiya 4: 53-56. [In Russian.] Abstract in Plant Breeding Abs. 29(4): No. 4254, p. 835, 1959.

______ 1964. [METHODS OF THE PRODUCTION OF TEA HYBRID SEEDS, THE INCREASE OF ITS YIELD AND THE IMPROVEMENT OF ITS QUALITY.] Makharadze. Vses. Nauch.-Issled. lnst. Chaya. [ Subtropical Crops. ] 4: 39-42. [In Russian, English summary.]

MITCHELL, G. E.
1901. HOME-GROWN TEA. U.S. Dept. Agr. Farmers' Bul. 301, 16 pp.

SIMURA, T., and OOSONE, K.
1956. [STUDIES ON THE FERTILIZATION OF THE TEA PLANT.] Ikushugaku Zasshi-Jap. Jour. Breeding 6: 11-14. [In Japanese, English abstract.]

TOMO N., FUCHINOUE Y., and FUCHINOUE, H.
1956. [STUDIES ON SELF-FERTILIZATION OF THE TEA PLANT.] Chagyo Kenkyu Hokoku-Tea Res. Jour. 7: 14-20 [In Japanese, English Abstract.]

WIGHT, W., and BARUA, P. K.
1939. THE TEA PLANT INDUSTRY, SOME GENERAL PRINCIPLES. Indian Tea Assoc. Tocklai Expt. Sta. Memo. 7, 13 pp.

WU, C. T.
1967. [STUDIES ON THE PERCENTAGES OF FRUIT-SETTING OF SELF- AND CROSS-POLLINATION AND ITS RELATION TO SOME ECONOMIC CHARACTERS OF THE F₁ HYBRIDS IN TEA PLANT Agr. Assoc. China, Jour. 59: 24-39. [In Chinese, English summary.]

TEPHROSIA
Tephrosia vogelii Hook. f., family Leguminosae
Tephrosia is a potential source of rotenone, an important nonresidual insecticide, and also a material useful in killing undesirable fish (Blommaert 1950). It is maintained as a semicultivated plant in dooryards in some primitive areas where it is used for poisoning fish. Seeds are saved and planted, and the plants are tended, yet the plant also occurs in the wild state. It has not been grown commercially, although recent tests (Barnes et al. 1967, Gaskins et al. 1972) indicate that production might be economically feasible if the culture and handling of the crop could be similar to that of some hay crops. Other species of Tephrosia native to the United States have also been tested and show some promise as sources of rotenone (Sievers et al. 1938).

Plant:
Tephrosia is a short-lived, slow-growing, herbaceous, frost- susceptible perennial. Barnes et al. (1 967) suggested that for commercial production of rotenone, which is derived largely from the leaves, the plants should be grown at the rate of 30,000 to 37,000 per hectare. Barnes and Freyre (1969) reported that when plants were spaced 1.0 to 8.9 m apart, the seed yield per plant ranged from 1.1 to 8.9g, with the highest-yielding line producing at the rate of only 70 kg seed/ha.

Many individual plants produce good seed yields, whereas others are poor, indicating that considerable improvement through breeding for seed production is possible. Based on variations among accessions in seed set and others agronomic traits, Martin and Cabanillas (1970) suggested that cross-breeding procedures might result in improved seed production.

Gaskins et al. (1972) stated that seed production is impeded by the flowering requirements of the species, by a naturally occuring system of sterility, and by frequent shortages of pollinating insects.

Inflorescence:
The flower is typically papilionaceous (fig, 183), about an inch across, and purple with white markings or white. The flowers are borne on compact raccemes that bloom over a 3- to 6- week period. There may be 20 to 30 flowers per raceme with up to 200 flowers per plant (Gaskins et al. 1972). Pods usually contain 8 to 16 seeds. The flowers have a faint but definite pleasant aroma and bees visit them freely for both nectar and pollen. Flowering occurs on decreasing day-lengths. If the plant is grown in the United States, it is likely to be killed by frost before flowers appear, therefore the plant requires a tropical home for seed production. The flower may last about 2 days during cool or rainy weather but only 30 hours during dry weather (Martin and Cabanillas 1970)

[gfx] FIGURE 183. - Flowering stems of a tephrosia plant.

Pollination Requirements:
The plant is considered to be self-pollinated. The stigma appears to be receptive to pollination at anthesis. Often it is in contact with dehisced anthers, particularly if the stamens are long. Furthermore, data indicate that self-pollination generally occurs, because recessive white flowered selections grown next to dominant purple flowered ones never produce purple-flowered offspring. Also, flowers bagged before anthesis frequently produce some pods with seed; however, when viewed from an agronomic standpoint the seed set ispoor, and large differences in seed production occur in different locations. Martin and Cabanillas (1970) showed that pollination is a factor by comparing plants in the open, plants caged with bees, and plants caged without bees. The results showed that from 10.8 to 22.8 percent of the pods set in open plots, 17.4 percent set in the cage with bees, but only 0.8 to 3.7 percent of the pods set in cages without bees. There were also fewer seeds per pod in the cage without bees.

Knowing that visiting bees, largely honey bees or carpenter bees, caused scratches on the stigmas, possibly making them more receptive to pollination, some stigmas were intentionally damaged with a needle before pollination. Others were pollinated as gently as possible. The results were significant: a 50 percent increase in pod set and more than 100 percent increase in seed set were obtained from flowers with damaged stigmas.

Pollinators:
Martin and Cabanillas (1970) concluded that "bees appear to have a role in pollinating tephrosia." Honey bees were the most frequent visitors. Some were nectar-seeking bees that visited only "younger" flowers and usually did not touch the stigma. Pollen-seeking bees, on the other hand reportedly visited chiefly older flowers. They forced open the upper suture or short leg of the keel to remove pollen, but the effect of the bee behavior on the flower was not visibly discernable but did not "appear to lead to pollination." How the authors arrived at this conclusion is not clear. They concluded that carpenter bees (Xylocopa brazilianorum L.) were the principal pollinators. Gaskins et al. (1972) concluded that insects facilitate self-pollination but contribute little to cross-pollination. They considered the honey bee too small to trip the flowers, yet they reported that most flowers had scratches on the stigma but were not tripped. A high percentage of these untripped flowers were found to be self-pollinated. Thus, they concluded, bees facilitate self-pollination by changing the relative position of keel, stigma, and pollen, without tripping, preparing the stigma for pollinatin by breaking up the stigmatic surface.

Pollination Recommendations and Practices:
Martin and Cabanillas (1970) recommended that plantings be made close to weedy areas and in abandoned fields where dead trees or legume plantings occur. (They also recommended that research be conducted on methods of increasing natural populations of carpenter bees.) Their discussion of the pollination of this crop is so reminiscent of the early history of alfalfa pollination that one is led to wonder if flooding the field of tephrosia with honey bees might not have the same beneficial effect it has had on alfalfa seed production.

LITERATURE CITED:
BARNES. D. K., and FREYRE, R. H.
1969. SEED PRODUCTION POTENTIAL OF TEPHROSIA VOGELII IN PUERTO RICO. Puerto Rico Univ. Jour. Agr. 53(3): 207-212.

FREYRE, R. H., HIGGINS, J. J., and others.
1967. ROTENOID CONTENT AND GROWTH CHARACTERISTICS OF TEPHROSIA VOGELII AS AFFECTED BY LATITUDE AND WITHIN ROW SPACING. Crop Sci. 7: 93-95.

BLOMMAERT, K. L. J.
1950. THE PLANT TEPHROSIA VOGELII HOOKER, AS A FRESH WATER FISH POISON. Roy. Soc. So. Africa Trans. 32: 247-263.

GASKINS, M. H., WHITE, G. A., MARTIN, F. W., and others.
1972. TEPHROSIA VOGELII: A SOURCE OF ROTENOIDS FOR INSECTICIDAL AND PISCICIDAL USE. U.S. Dept. Agr. Tech. BuI. 1445, 38 pp.

MARTIN, F. W., and CABANILLAS, E.
1970. THE BIOLOGY OF POOR SEED PRODUCTION IN TEPHROSIA VOGELII. U.S. Dept. Agr. Tech. BuI. 1419, 34 pp.

SIEVERS, A. F., RUSSELL, G. A., LOWMAN, M. S., and others.
1938. STUDIES ON THE POSSIBILITIES OF DEVIL'S SHOESTRING (TEPHROSIA VIRGINIANA) AND OTHER NATIVE SPECIES OF TEPHROSIA AS COMMERCIAL SOURCE OF INSECTICIDE. U.S. Dept. Agr. Tech. BuI. 595, 40 pp.

TUNG
Aleurites fordii Hemsl., family Euphorbiaceae
The tung tree has also been known as tung-nut, tung-oil, or china wood-oil tree (Fairchild 1913). The word tung is Chinese for "heart," the general shape of the leaf (Potter and Crane 1957). In 1964, there were about 7.5 million tung trees in the Southern States: Mississippi, 4.4 million; Florida, 1.7 million; Louisiana, 1.3 million; and Alabama, 0.1 million; with a few thousand in Georgia. Production of nuts amounted to 123,300 tons with a value of $7.6 million. Production amounted to only 11,700 tons valued at less than $1 million in 1970, when production estimates in USDA Agricultural Statistics were discontinued.

Tung oil, which is pressed from the nuts of this tree, is used by the protective coating industry in varnishes, enamels, and electrical insulators.

Plant:
Tung is a soft wooded, smooth-barked deciduous tree that may grow to 30 feet. First blooms appear from late February to April before the leaves appear. In bloom, the tree (fig. 187) is highly attractive because of its mass of pink blossoms similar to flowers of catalpa (Catalpa spp.). These are followed by the dense foliage of 3- to 5-inch heart-shaped dark-green leaves. Current orchard recommendations include 100 to 140 trees per acre (Potter and Crane 1957) with suggested spacings ranging from 12 by 30 feet to 15 by 40 feet. Growers allow the trees to start branching 4 to 6 feet above the ground. Tree shape is oval and symmetrical.

[gfx] FIGURE 187. Eleven-year-old tung orchard in full bloom.

Inflorescence:
The colorful and attractive blossoms (fig. 188), which are borne on the ends of the growing shoots of the previous season, vary in type. They may be all staminate, all pistillate, or predominantly one or the other (Dickey and Reuther 1940, McCann 1942). The percentage of pistillate flowers may depend on the vigor of the tree, with more such flowers produced on trees making more vigorous growth (Abbott 1929). The reddish-white flowers occur in panicled cymes or clusters with usually about 60 staminate and one pistillate flower each, with rarely a perfect flower (Newell 1924). Each flower may be an inch or more in length, and the tree is covered with the canopy of blossoms. The pistillate flowers have a three- to five-celled ovary that, when pollinated, produces a top- shaped fruit 2 to 3 inches in diameter, usually bearing five seeds. The blossoms secrete some nectar, and the staminate flowers produce a copious amount of pollen (Pering 1937)., Bees visit the blossoms freely.

[gfx] FIGURE 188. - Closeup of a flowering tung branch.

Pollination Requirements:
Angelo et al. (1942) showed that plants caged to exclude bees set no fruit and that wind or shaking the tree was of no value in fruitsetting, but when a tree was caged with a colony of honey bees a good set was obtained. Others (Hambleton 1950, Pering 1937) also credit honey bees with setting the crop. The tree is not self-sterile. It merely needs the agency to transfer the sticky pollen from the anthers of the staminate flowers to the stigma of the pistillate flower. Brown and Fisher (1941) showed that pollination can occur over several days of the life of the blossom. Webster (1943) concluded that when staminate and pistillate flowers are on separate trees, one staminate tree for 20 pistillate trees was sufficient for satisfactory pollination, provided that some staminate flowers open by the time the pistillate flowers are receptive.

Pollinators:
The pollination of tung trees is dependent upon the honey bee. Other insects visit the blossoms but rarely in sufficient abundance to be of significance. Under orchard conditions, practically every ovule of every pistillate flower is capable of developing a seed. This means that at least one viable pollen grain must land on each of the four or more lobes of the stigma of each flower at the right time to permit fertilization of the ovules. The bee population necessary to accomplish this has not been determined.

Pollination Recommendations and Practices:
No recommendations have been made on the use of bees in the pollination of tung, even though there is no doubt about their need in the production of this crop. The need for transfer of pollen grains from the staminate to all of the lobes of the pistillate flower may not require repeated visits of honey bees, but the grower should not overlook this need. To assure maximum set of nuts, he should arrange for an appropriate number of strong healthy colonies of honey bees so that every pistillate flower is well pollinated. Because of the small acreage now being grown in the United States, the demand for insect pollination is not great.

LITERATURE CITED:
ABBOTT, C. E.
1929. FRUIT BUD DEVELOPMENT IN THE TUNG OIL TREE. Jour. Agr. Res. 38: 679-696.

ANGELO, E., BROWN, R. T., and AMMEN, H. J.
1942. POLLINATION STUDIES WITH TUNG TREES. Amer. Soc. Hort. Sci. Proc. 41: 176-180.

BROWN, R. T., and FISHER, E.
1941. PERIOD OF STIGMA RECEPTIVITY IN FLOWERS OF TUNG TREES. Amer. Soc. Hort. Sci. Proc. 39: 164-166.

DICKEY, R. D., and REUTHER, W.
1940. FLOWERING, FRUITING, YIELD AND GROWTH HABITS OF TUNG TREES. Fla. Agr. Expt. Sta. Bul. 343, 28 pp.

FAIRCHILD, D.
1913. THE CHINESE WOOD-OIL TREE. U.S. Dept. Agr. Burl Plant Indus. BPI Cir. 108, 7 pp.

HAMBLETON, J. I.
1950. ARE HONEY BEES NEEDED FOR THE POLLINATION OF TUNG TREES? Gleanings Bee Cult. 78: 227.

MCCANN, L. P.
1942. DEVELOPMENT OF THE PISTILLATE FLOWER AND STRUCTURE OF THE FRUIT OF TUNG (ALEURITES FORDII). Jour. Agr. Res. 65: 361-378.

NEWELL, W.
1924. PRELIMINARY REPORT ON EXPERIMENTS WITH THE TONG-OIL TREE IN FLORIDA. Fla. Agr. Expt. Sta. Bul. 171, pp. 193-234.

PERING A. H.
1937. TUNG OIL PRODUCTION AND THE BEEKEEPER. Amer. Bee Jour. 77: 526-527.

POTTER, G. F., and CRANE, H. L.
1957. TUNG PRODUCTION. U.S. Dept. Agr. Farmers' Bul. 2031, 35 pp.

WEBSTER, C. C.
1943. A NOTE ON POLLINATION IN BUDDED PLANTATIONS OF TUNG TREES (ALEURITES MONTENA). Nyasaland Agr. Quart. Jour. III(3): 17-19.

VANILLA
Vanilla spp., family Orchidaceae
Three species of Vanilla are of commercial importance: V. planifolia Andrews [ V. fragrans (Salisb.) Ames], V. pompona Schiede, and V. tahitensis J. W. Moore. The first is by far the most important (Childers et al. 1959). Vanilla is cultivated for its pods which, under processing, yield vanilla extract. In 1950, world production was 3 million pounds of extract, of which Madagascar produced more than half and Mexico about a fourth. The United States is the primary consumer of vanilla extract (Childers et al. 1959).

Plant:
Vanilla is a tropical, evergreen, leafy, and somewhat fleshy vine (fig. 189) that may climb to the top of trees, 50 to 75 feet, if unchecked. It has thick, oblong, 6- to 9-inch, dark-green leaves and forms roots opposite the leaves by which it clings to the tree. It is propagated vegetatively, pruned at the tip, and trained onto a trellis. The plants are usually set 6 feet apart in the row, the rows about 9 feet apart (Kanman and Plai 1966). The fruit is a beanlike pod, 4 to 6 inches long by about one-half inch thick (fig. 190). The better pods are 5 inches or more in length. A single pod contains thousands of seeds almost microscopic in size (l/3 to 1/4 mm).

A healthy vine should produce about 100 pods per year, which mature 8 to 9 months after flowering. If too many flowers are pollinated and too many pods set, the vine may be overloaded and will die.

Just before the plant flowers, the grower usually prunes 4 to 6 inches from the vine tip; this stops linear growth and seems to benefit flowering (Childers et al. 1 959).

Irvine and Delfel (1961) dispelled the former belief that plants will not flower unless they are climbing, by showing that inflorescences were produced satisfactorily on horizontal and even descending stems. Of 10 plants studied, one ascending stem had 82 inflorescences, one 60 feet tall and still climbing had only 18, and one descending vine had 29. This proved that maintenance of plants on trellises did not necessarily cause a decrease in yield.

[gfx] FIGURE 189. - Section of vine of vanilla (Vanilla planifolia) growing on tree trunk.
FIGURE 190. - Dried vanilla beans from 4 different cultivars.

Inflorescence:
The small lilylike, greenish-yellow vanilla flowers, 1 l/2 by 2 1/2 inches long (Woebse 1963), develop in axillary racemes (fig. 191). There may be as many as 100 flowers in a raceme but usually there are about 20. Usually, only one flower in a raceme opens in a day, with the entire flowering period of the raceme lasting an average of 24 days. Flowering for an average plant in Puerto Rico begins in January, reaches a peak in March, and ends in June. In the Philippines, flowering extends from March to June, with the largest percentage of the flowers appearing in April (David 1953).

The individual flower has three sepals and three petals, one of the petals being enlarged and modified to form the trumpetlike lip, and a central column comprised of the united stamen and pistil. The anther is at the apex of the column and hangs over the stigma, but a flap or rostellum separates them.

The flower opens in the morning and closes in the afternoon, never to re-open. If it is not pollinated, it will shed the next day. The optimum time for pollination is in midmorning (Childers et al. 1959).

The pollen clings together in a mass and is feebly attractive to certain bees and hummingbirds (Cobley 1956*, DeVarigny 1894). The nectar source is seldom mentioned, although hummingbirds visit flowers primarily, if not exclusively, for nectar. Many species of the family Orchidaceae are noted for their nectar secretion (Darwin 1877*). Correll (1953) stated that the flowers are visited for the "honey" secreted at the base of the lip.

[gfx] FIGURE 191. - Flowers of Vanilla pompona.

Pollination Requirements:
The vanilla flower is self-fertile, but incapable of self-pollination without the aid of an outside agency to either transfer the pollen from the anther to the stigma or to lift the flap or rostellum then press the anther against the stigma. The only time this can be accomplished is during the morning of the one day the flower is open. Unless pollination occurs, the flower drops from the vine the next day. Correll (1953) stated that insect pollinated flowers, being cross-pollinated, produce viable seed, but flowers that are hand pollinated, being self-pollinated, produce only sterile seeds.

Pollinators:
The reference occurs repeatedly in the literature that in its native Mexico the flowers of vanilla are pollinated by small bees of the genus Melipona and also by hummingbirds (Ridley 1912*). Childers and Cibes (1948) noted that this report has not been carefully checked and later Childers et al. (1959) said that there is no experimental proof that they are actually effective pollinators. Mention is made by Childers et al. (1959 p. 477), that "The first effort made toward solving the (pollination) problem was to introduce bees of the genus Melipona from Mexico, but they did not thrive. After this failure a mechanical means of pollination was tried." Then Albius, in 1841, discovered the practical method (Childers et al. 1959) of using a small splinter of wood or a grass stem to lift the rostellum or flap out of the way so that the overhanging anther can be pressed against the stigma to effect self-pollination.

Now, practically all vanilla is produced by hand pollination, which accounts for 40 percent of the total labor cost in vanilla production (Gregory et al. 1967).

No further study seems to have been made on the utilization of Melipona, or other insects, or hummingbirds. No attempt has been made to concentrate pollinating insects for this purpose. It would appear logical that if nectar is secreted, as indicated by Correll (1953), honey bee colonies could be amassed in the area when desired, and the workers could be "forced" to visit the flowers. The relative cost of a high concentration of honey bee colonies as compared to the cost of human labor, would make such exploitation of honey bees highly worthwhile investigating. The reference by DeVarigny (1894) that Cuban bees, whether indigenous or naturalized European bees, were pollinating vanilla in Cuba indicates that bees could be used satisfactorily.

Pollination Recommendations and Practices:
There are no recommendations for the use of bees, bats, birds, or other agencies. The evidence indicates, however, that saturation pollination by honey bees or certain other bees offers possibilities because vanilla in Mexico was probably pollinated by bees at one time to some extent.

LITERATURE CITED:
CHILDERS, N. E., and CIBES, H. R.
1948. VANILLA CULTURE IN PUERTO RICO. U.S. Dept. Agr. Cir. 28, 94 pp.

______CIBES. H. R., and HERNANDEZ-MEDINA, E.
1959. VANILLA - THE ORCHID OF COMMERCE. In Whithner, C. L., The Orchids - A Scientific Survey, pp. 477-508. Ronald Press Co., New York.

CORRELL, D. S.
1953. VANILLA. ITS BOTANY, HISTORY, CULTIVATION AND ECONOMIC IMPORTANCE. Econ. Bot. 7(4): 291-358.

DAVID. P. A.
1953. POLLINATION OF VANILLA FLOWERS BY HAND. Philippine Agr. 37: 301-305.

DEVARIGNY. C.
1894. FERTILIZATION OF THE VANILLA FLOWER BY BEES. Bombay Nat. Hist. Soc. 4: 555-556.

GREGORY. L. E., GASKINS, M. H., and COLBERG, C.
1967. PARTHENOCARPIC POD DEVELOPMENT BY VANILLA PLANIFOLIA ANDREWS INDUCED WITH GROWTH-REGULATING CHEMICALS. Econ. Bot. 21: 351-357.

IRVINE, J. E., and DELFEL, N. E.
1961. FLOWERING BEHAVIOR OF VANILLA. Nature 190(4773): 366.

KANMAN, K., and PILLAI, G. G.
1966. VANILLA, A NEW VENTURE FOR KEROLA FARMERS. Indian Farm. 16(2): 4 - 7.

WOEBSE, A.
1963. POLLINATING THE VANILLA FLOWER. Amer. Orchid Soc. Bul 32(12): 1009-1010.

VERNONIA
Vernonia anthelmintica (L.) Willd., family Compositae
The seeds of vernonia are of current interest because they are the source of an oil containing vernolic acid (Higgins 1968, Krewson et al. 1962). Yield and related results have recently been obtained (Berry et al. 1970, Lessman and Berry 1967, Massey 1969). However, not too much attention was given to the pollination requirements.

Later, through the use of flower color as a genetic marker, Berry and Lessman (1969) noted cross-pollination in open plots. Berry et al. (1970) reported 13 percent outcrossing and noted an abundance of bees. They considered the plant to be essentially self-pollinated.

If this crop is to come into commercial production, more careful and extended studies of its pollination requirement should be made.

LITERATURE CITED:
BERRY. C. D.. and LESSMAN, K. J.
1969. CONTROLLED CROSSING AND INHERITANCE OF FLOWER COLOR IN VERNONIA ANTHELMINTICA Jour. Hered. 60: 75.

BERRY, C. D., LESSMAN, K. J., and WHITE, G. A.
1970. NATURAL CROSS-FERTILIZATION IN VERNONIA ANTHELMINTICA (L . ) WILLD. Crop Sci. 10: 104-105.

HIGGINS J. J.
1968. VERNONIA ANTHELMINTICA: A POTENTIAL SEED OIL SOURCE OF EPOXY ACID. I. PHENOLOGY OF SEED YIELD. Agron Jour. 60: 55-58.

KREWSON, C. F., ARD, J. S., and RIEMENSCHNEIDER, R. W.
1962. VERNONIA ANTHELMINTICA (L. ) WILLD. TRIVERNOLIN, 1, 3- DIVERNOLIN AND VERNOLIC (EPOXYOLEIC) ACID FROM THE SEED OIL. Jour. Amer. Oil Chem. Soc. 39: 334-340.

LESSMAN, K. J., and BERRY, C.
1967. CRAMBE (ABYSSINICA) AND VERNONIA (ANTHELMINTICA) RESEARCH RESULTS AT THE FORAGE FARM IN 1966. Ind. Agr. Expt. Sta. Res. Prog. Rpt. 284,3 pp. \

MASSEY, J. H.
1969. FRUITING PATTERN OF VERNONIA ANTHELMINTICA (L.) WILLD. Agron. Jour. 61: 651-652.

WHITE SAPOTE
Casimiroa edulis Llan. & Lex., family Rutaceae
Hoffman (1970) reported that although white sapote is a tropical plant it is becoming popular in the mild areas of California and Florida. Mowry et al. (1967*) stated that it was growing in scattered locations in these two States. Mortenson and Bullard (1968*) pointed out that this plant is not a true "sapote", which is in the family Sapotaceae, but a relative of the citrus. According to Mowry et al. (1967*), it is more common than any other plant known as sapote. (Also see, "Mamey Sapote," p. 253.)

Plant:
White sapote is a medium to large, erect to spreading, evergreen tree, with leaves 3 by 6 inches in size. It produces ovoid fruit up to 3 inches in diameter, with a greenish skin that becomes yellow at maturity. The creamy or yellowish juicy fruit, rich in vitamin C, has a distinctive sweet aroma. Each fruit has one to five large ovoid seeds imbedded in the flesh. The fruit is eaten out of hand or as a fruit dessert with cream and sugar (Mortensen and Bullard 1968*).

Inflorescence The small (5 mm), whitish-green flowers are produced in great numbers along the branchlets, but few ever produce mature fruit (fig. 196). Abundant nectar is produced from these flowers in southern Mexico, primarily in January and February (Ryerson 1925, Wulfrath and Speck no date). The individual flower has five pale-green, strongly reflexed pews with five short, stout, slightly reflexed stamens. Each stamen arises between two petals. The position of the pews and anthers leaves the globose green ovary exposed. The stigma is sessile, resting directly on the ovary, and leads directly to five ovules, which normally form the one to five seeds in the fruit. Nectar is secreted on the base of the petals next to the ovary.

[gfx] FIGURE 196. - Longitudinal section of 'Neysa' white sapote flower, x 15.

Pollination Requirements:
Mature fruits are sometimes irregular in shape from lack of seed development in one or more carpers, and heavy shed of immature fruit has been reported (Kennard and Winters 1960*). Although the flowers are hermaphrodite, Mustard (1954) showed that there is partial to total pollen sterility within the flower. He concluded that this factor was responsible for failure to set a good crop of fruit. Mortensen and Bullard (1968*) stated that in Florida the 'Dade' cv. must be cross-pollinated because it does not have "normal" pollen. Mustard (1954) reported that partial pollen sterility may be a factor in failure to obtain good sets of fruit. Pollinators The honey bees in visiting the flowers for nectar doubtless serve as pollinators of the plants, particularly if they are sufficiently concentrated in the area of the trees.

Pollination Recommendations and Practices:
None.

LITERATURE CITED:
HOFFMAN, L. E.
1970. THE SAPOTE TREE. Horticulture 48:3.

MUSTARD, M. J.
1954. POLLEN PRODUCTION AND SEED DEVELOPMENT IN THE WHITE SAPOTE. Bot. Gaz. 116:189-192.

RYERSON, H.
1925. THE WHITE SAPOTE OF CALIFORNIA. Calif. Univ. (Berkeley), College of Agr., Div. of Subtrop. Agr. Cir. 3, 2 pp.

WULFRATH, A., and SPECK, J. J.
(n.d.) ENCICLOPEDIA APICOLA. Folleto 28, Ed. 2, Ediciones Mexicanas, Mexico, D.F. 96 pp. [In Spanish.]