Index | Search | Home | Table of Contents

McCain, R. 1993. Goldenberry, passionfruit, & white sapote: Potential fruits for cool subtropical areas. p. 479-486. In: J. Janick and J.E. Simon (eds.), New crops. Wiley, New York.

Goldenberry, Passionfruit, & White Sapote: Potential Fruits for Cool Subtropical Areas

Richard McCain

    1. Botany
    2. Horticulture
    1. Botany
    2. Horticulture
    3. Passifloras
    4. Hybrids
    5. Conclusions
    1. Botany
    2. Horticulture
  5. Table 1

In the West, as in other agricultural regions of the United States, many farmers, especially small growers, have recently been looking at new and unusual crops to diversify their operations. Factors leading to diver-sification include water limitations, pressures from the development sector, overproduction of many traditional local crops, competition (i.e., lower prices) from out-of-state or foreign growers, and increasing labor costs.

The central coast region of California has a variety of soils and climates conducive to the growing of a large assortment of crops. While this region is well known for its production of strawberries, lettuce, and artichokes, its unique conditions have encouraged farmers to experiment with many unusual crops as well.

For example, since it was founded in 1979, the author, a partner in Quail Mountain Herbs, has grown herbs, salad crops, edible flowers, subtropical fruits, and other specialty produce in the Monterey Bay region of the central coast. Since its inception, Quail Mountain has successfully marketed over 100 different crops nationwide. Quail Mountain's growing grounds are located in several different microclimates. Due to this region's unique geography, at least 25 different microclimates have been reported.

The author has conducted research on a wide variety of herbs, flowers, vegetables, and fruits over the years. Most of the research has focused on finding or creating plants with superior taste, appearance or cultural requirements. This is done through selection and hybridizing to achieve desirable traits. The following fruits are among those crops that show the most potential.



First described by Linnaeus in 1753, the cape gooseberry or goldenberry, Physalis peruviana L., Solanaceae, has been cultivated for many decades along the Andes Mountains of South America. Goldenberry, the proposed name (Legge 1974) now being used to avoid any confusion with the common gooseberry (Ribes ssp.), has been spread by explorers and travelers worldwide, but is still considered a backyard fruit in most areas. Small industries are developing around the goldenberry in countries in central and south Africa, Australia, New Zealand, and India but nowhere has it really achieved large commercial success (Morton 1987). The plant's productivity in poor soils, its ease of cultivation, and low requirement for water and fertilizer has made it an attractive potential crop.

The goldenberry is an herbaceous, erect, alternate-branched shrub with pubescent slightly toothed heart-shaped leaves that appear irregularly along the stems (Moriconi et al. 1990). Yellow, pendulous flowers have campanulate corollas with purple to purplish brown spots. The genus, Physalis, with 100 or so species of annual and perennial herbs, is characterized by the fruit being enclosed in a papery husk or calyx. The goldenberry has a particularly delicious fruit with a tangy pineapple-like flavor. Several members of the genus are exploited for their berries. Among them are the ground cherry, P. pruinosa L. and the tomatillo of Mexico, P. ixocarpa Brat.


Dried ripe fruits from selected clones of the previous seasons are fermented in water for up to 5 days. After the seeds are separated from the pulp, they are planted in flats of sterile peat-lite mix. The flats are kept continually moist. Seeds germinate in 8 to 14 days in an unheated greenhouse. The seedlings are field planted when they are 15 to 20 cm tall with at least 1.0 m between each plant. The plants are watered by either drip irrigation or overhead sprinklers, although drip irrigation provides more control of water consumption and weeds. The plants appear to need little or no fertilizer. Fruit production decreased significantly when fertilizer was applied. When fish emulsion or 12-5-10 (NPK) fertilizer were applied at rates approximately 45 lb. N/acre (50 kg/ha), the plants exhibited a great deal of vegetative growth but produced few flowers or fruit. Plants planted in sandy soil without any amendment or fertilizer produced 150 to 300 flowers per plant with a corresponding number of fruit. Plants typically have a sprawling habit similar in size and growth pattern to their relative, the tomato. The plants should be trellised or staked.

The first yellow, bell-shaped flowers appear 4 to 5 weeks after transplanting during the onset of warm spring days in April and continue flowering through November unless damaged by an early frost. The plants are pollinated by wind and local insects including bees. While pollination is not a problem, inconsistency in fruit size is a problem.

Goldenberry plants typically are heavy fruit producers. Fruit production begins in August and continues until the plants are killed by frost (usually early December in this area). Yields of 150 to 300 fruits per plant are not unusual. The fruits are ripe when they turn yellow-gold. Unripe fruits are green. The fruits are 1.25 to 2.5 cm in diameter and are encased in a papery, tan husk. When fully ripe, the fruits and husk will naturally dehisce or fall when given a good shake. Harvesting can be accomplished by allowing the fruit to fall on fabric or plastic placed under the plants. Collection is either done by hand picking, or by gathering up the plastic and pouring the fruit into containers. Vacuum harvesting the fruit should be explored. Hand collection is preferable if the fruit is to be sold on the fresh market, to avoid bruising. The fruit is quite durable when left in the husk. Goldenberries are generally sold with the husk left on as many chefs use the husk for decorative purposes. After harvest, the ripe fruit may last several months without refrigeration, if kept dry. They also may be picked partially green and allowed to ripen, but these fruit never become as sweet as vine-ripened fruit.

Fruits seen on the market vary in taste and size. The fruits grown at Quail Mountain have a sweet, tangy taste while some cultivars are mealy and tasteless. There is great genetic variability. The fruit is eaten fresh or cooked. The fruit makes excellent pies and jellies and is very high in pectin. The fresh fruit may be served with husk pulled back for fondue. Goldenberry sauce is a nice accompaniment to a meat dish. While not well known by the retail consumer, the fruit has a strong following among chefs and the market is likely to grow for good quality, vine-ripened fruit.

The plant has been primarily pest and disease free at Quail Mountain. Botyritis mold has been found on ripe fruit. This may be caused by the frequent fog in this area.



The Passifloraceae contains nearly 600 species in 12 genera, four of which are found in the New World. Of the 500 species of the genus Passiflora, only 10 to 12 have been exploited for their fruits (Menzel 1990; Howell 1989). Most commercial production around the world is based on cultivars of P. edulis Sims, its yellow form P.e.f. flavicarpa or their hybrids, which are species of the subgenus Granadilla. There are more than 60 other Passiflora species with edible fruits, many that should excite consumer interest.

The subgenus Tacsonia contains over 40 species with edible fruit (Escobar 1980). Originating from the cool, middle to high altitude regions of the Andes, most of the Tacsonia are suited to the ocean-influenced microclimates of Western North America. The arid inter-Andean valleys are veritable biogeographic islands, each with many endemic species that are isolated from other such valleys by wet tropical forests below and cold Andean tundras above, a situation favoring speciation (Iltis 1988). Although there is much genetic variation in the Tacsonias, a few natural hybridizations do occur (Killip 1938; Escobar 1980), and members of Tacsonia hybridize readily under cultivation. The feasibility of cross-breeding Passifloras to improve their fruits has been demonstrated by Rupert-Torres and Martin (1974) and Escobar (1980).

At Quail Mountain, the author started collecting Passifloras in 1979 and began hybridizing them in 1984. Hybridizing for fruit production has focused primarily on the subgenus Tacsonia. At the present time there are over 100 species of Passifloras in this collection, 15 of which are Tacsonias. Through numerous generations of hybrids and a continuous process of selecting superior plants from hundreds of hybrids, several very promising cultivars have been developed.

The very diverse genus Passifloras is characterized by woody or herbaceous vines furnished with tendrils in the leaf axils. The polymorphic, alternate leaves can be extremely variable in size and shape. Some species produce egg mimics to avoid predation by members of the Lepidoptera family or other pests (Gilbert 1975; K.W. Williams and L.E. Gilbert unpub.). The angular leaf stalks usually contain excrescences called glands along the petioles. The flowers located in the leaf axils are extremely colorful and elaborate in most species. The floral parts consist of a calyx with 5 lobes or sepals and 5 petals. Inside the petals are one to several series or rings of filaments forming the corona, the center of which is filled with nectar. The outer rings compose a perfect target for hummingbirds, bats, bees, moths, and other pollinators.

The main characteristics that set the Tacsonia apart from other subgenera in the family are that the filaments are reduced to nubs or tubercles in the corona and the calyx tube is much longer than the sepals. The male and female parts are raised aloft by the narrow columnar gynophore. At the top of the gynophore is the ovary with 5 stamens below and 3 stigmas above.

Many species of Passifloras are self-infertile, but if pollinated, the ovary swells into a fruit that is botanically classed as a berry. Hand pollination greatly increases fruit size. In over 60 species, fruits, that vary in size from that of a marble to a small melon, are filled with a deliciously sweet or tangy acidic pulp. The juice is widely used in many countries as a flavoring for juices, confections, ice cream, and other products. Demand in the United States has been growing steadily in the last few years for fruit which can be sold fresh or juiced for use in other products.


While the cultivation of the many Passifloras may be similar, the following is derived from the author's experimentation with selected Tacsonia species. Each year, passionfruit vines are selected from a growing area of about 1 ha based on overall vigor, consistent flowering and fruiting, fruit size and flavor, time of ripening, and disease resistance. Cuttings from these selected clones are rooted in a propagation house in the fall and over-wintered. Selections are propagated by dipping cuttings with 5 to 10 nodes from vigorous growing shoots in rooting hormone (IBA/IAA combination at a concentration of 1,500 ppm) and placed into flats containing a fast draining mix of perlite and sand on a bench with bottom heat and intermittent mist. Rooting takes three to five weeks. During the winter, well-rooted cuttings are transplanted into containers with a peat-lite mix and receive organic fertilizer once in late January.

The vines are set out in fields at the beginning of February (3 x 1.8 m spacing) and are drip irrigated weekly. The soil is amended with bone meal, blood meal, and seaweed meal prior to planting. The vines are trellised as they perform poorly when sprawling along the ground. The vines commence flowering in May and continue nearly year-round unless killed by a hard freeze. At Quail Mountain, nearly every flower sets a fruit. Pollination is by bees, hummingbirds, and wind. Hand pollination can increase fruit size. Vines produce 50 to 120 fruits each. With 740 vines/ha, estimates of harvest yield are 16,800 to 22,400 kg/ha.


The following Passifloras have commercial possibilities but require further breeding efforts. Most of the author's selections are Tacsonia or other cool growing Passifloras which have not had widespread commercial use in the United States.

Passiflora ampullacea is one of the only white flowered Tacsonias. This Passiflora has not flowered yet in Northern California, but flowered in San Diego, in 1988. Fruits are reported to be larger and with a thicker rind than P. mollissima (NRC 1989). Hardiness is unknown. It is found in the mountains of Southern Ecuador at elevations of 2,600 to 2,800 m.

P. antioquiensis thrives in shade, and must have large amounts of water. Flowers are brick-red, 15 cm across and hang down on peduncles up to 32 cm long. It has flowered and fruited at Quail Mountain and as far north as San Francisco. The fruit (to 12 cm) is similar in taste to the sweet granadilla, (P. ligularis), which is one of the sweetest passionfruits. Originally from the mountains of Colombia up to 3,000 m, it cannot survive more than a light frost.

P. cumbalensis fruits have long been used in South America for flavoring. Escobar (1987) reports at least seven distinct varieties of P. cumbalensis with variety goudatiana the best tasting. Flowers are rose to purple and fruits are red and banana shaped. It has flowered in San Francisco but not at Quail Mountain. P. cumbalensis can be found in the Andes from western Venezuela to northern Peru at elevations of 1,800 to 4,100 m.

P. manicata flowers and fruits nearly year round at Quail Mountain. The flowers are bright reddish-orange and attract hummingbirds. The fruits are small, (2.5 x 5 cm) with good flavor. Widely distributed in the Andes from 1,500 to 2,500 m in elevation.

P. mollissima is a widespread and variable species found from Venezuela down the central cordillera of Columbia to southeastern Peru and western Bolivia at 2,000 to 3,200 m in altitude. It is grown commercially in many areas and is reputed to be one of the best for juice. P. mollissima flowers from May to the first freeze (December-January). The yellow oblong fruits are approximately 10 to 15 cm long and have overtones of citrus. Some clones can be sweet. The species grows vigorously and shoots develop from the roots after frost injury.

P. mixta is a very diverse species found at higher elevations than P. mollissima (2,500 to 3,600 m) but with a similar range in the Andes. P. mixta has survived -7°C. Flowers are light pink to rosy-peach and bloom in successive flushes throughout the year in a cool coastal climate. The fruits are 2 to 3 cm wide and 8 to 12 cm long and are usually yellow-green when ripe. Fruits have a hint of raspberry flavor.

P. trisecta is found from southern Peru to Bolivia within a very limited range of elevation (2,400 to 2,800 m). It has broad white flowers with a short calyx tube. The anthers and stigmas are all located on one side of the flower. P. trisecta has flowered and fruited at Quail Mountain. The fruits are green and egg-shaped, 3 to 4 x 7 to 8 cm. Plant hardiness is unknown.

P. trifoliata is found in the Andes of central and southeastern Peru at 3500 to over 4000 m. The plant is very pubescent, appears to be hardy, but is susceptible to spider mites. Flowers are pink and similar to P. mixta although miniature in size (2 to 3 x 7 to 8 cm). Fruit is egg-shaped (3 to 4 x 5 cm) with three indentations longitudinally.

P. tripartita may be a variant of P. mollissima, but is limited to Columbia. The flowers are similar in color to P. mollissima but the petals and sepals flare open more. Fruit is similar in size, shape, and color to P. mollissima but tastes tarter.


P. antioquiensis (P. mollissima x P. exoniensis) is an extremely vigorous plant. The flowers are larger than P. antioquiensis and darker pink than P. mollissima. The fruit reported to be delicious (Vanderplank 1991).

P. manicata x P. mollissima flowers all year round on the coast of California. It must be hand pollinated to produce fruit. Fruit is larger than P. manicata, but smaller than P. mollissima.

P. mixta x P. mollissima is a vigorous grower but not as floriferous as P. mixta, nor as self-fertile as P. mollissima.

P. mollissima x P. mixta is fast growing and hardier than either parent with showier, vibrant pink flowers. Fruit is much larger and juicier than either parent with a different, but delicious flavor.

P. tripartita x P. mollissima is a weak grower with bright green leaves. The flowers are similar to the P. tripartita parent. The fruit retains the sour flavor of P. tripartita.

P. trisecta x P. mixta is fast growing, but has not flowered or fruited yet.

(P. manicata x P. mollissima) x P. mollissima is vigorous and hardier than either original parent. The fruit is delicious and as large as P. mollissima but pointed at the end.


Passionfruit vines are easy to bring to fruit if their basic cultural needs are met. Mixed-cropping or perennial cropping systems hold the most promise. At Quail Mountain, vines grow between rows of white sapotes in an orchard-like fashion. Between these rows, are rows of goldenberry and Yacon (Polymnia sonchifolia). At the ground level, a variety of herbs flourish. All of these crops seem to be very compatible with minimal disease or pest problems. Herbs in the Lamiaceae and Apiaceae are advantageous as they attract bees and other pollinating insects when they flower. Many Passifloras are self-infertile, so a sister seedling or closely related species must be in the vicinity to benefit pollination. Hand pollination is very labor intensive and expensive. Combinations of compatible species or clones need to be examined further to improve pollination in larger plantings. Fruits of many Tacsonias are soft and may bruise with mechanical harvesting, so clones must be selected for thicker skins. Selection of superior clones with larger sizes and improved fruit flavor will be continued. Processing and storage of fruit pulp needs to be investigated further. Passiflora fruits produce ethylene gas and need to be packed, stored, and shipped in such a way that this is not a hindrance to fruit quality. There is every reason to believe that with more research a greater variety of passionfruit flavors could be made available to the American consumer.



The white sapote, Casimiroa edulis Llave & Lex, Rutaceae, has attracted interest among rare fruit growers and orchardists in California. The genus Casimiroa contains 5 or 6 species (Thompson 1972; Morton 1987). Among these are three little-known shrubs or small trees from Mexico, C. pubescens Ramirez, C. pringlei Engl., and C. watsonii Engl. Another species, C. emarginata Standley & Steyerin, was described in 1944 from a single specimen found in Guatemala (Morton 1987). The C. sapota Oerst, matasano, is very similar to and often confused with C. edulis, along with C. tetrameria Millsp., the wooly-leaved white sapote. Although these last two species sometimes hybridize with C. edulis, their flavor is considered inferior (Thompson 1972). C. edulis, C. sapota, and C. tetrameria are found in central Mexico. Their range is broad, extending down into Central America as far as Costa Rica. The white sapote, C. edulis, remains the preferred fruit of the genus due to its delicious flavor and wide appeal.

In its native habitat, trees are found at altitudes of 750 to 2,700 m. They do not flourish in the hot, tropical lowlands (Morton 1987), but are cultivated around the world in subtropical areas and regions with a mild Mediterranean climate. The trees have been planted in the northern part of South America, the Caribbean region, Spain, Portugal, Southern France, and Italy. They are grown commercially on a small scale in New Zealand, Australia, and South Africa. White sapote have not been successfully grown in the Philippines, but have been cultivated in other islands of the East Indies (Morton 1987). There are small plantings in Florida, Hawaii, and experimental plantings in three different regions in Israel (Nerd et al. 1990). The white sapote has grown well in California since the early 1800s (Schneider 1986). It is thought to be first introduced by Franciscan monks along with figs, olives, and grapes (Thompson 1972). Some cultivars have fruited well as far north as San Francisco (Thompson 1972).

White sapotes are medium to large-sized, fast growing trees with aggressive spreading roots that help them withstand periods of drought. Mature trees can reach 15 to 18 m in height and produce 900 kg (2,000 lb.) of fruit per year. Thompson, (1972) reported a tree of 'Chestnut' produced nearly 2,700 kg (6,000 lb.) of fruit in 1971 in Vista, California. Grafted trees remain smaller and develop a better canopy than seedlings.


The author first planted a few white sapote trees at Quail Mountain in 1983. In 1986, 190 two-year-old grafted trees, representing 18 cultivars, were planted approximately 6 m apart (346/ha). The trees were drip fertigated weekly, for the first two years, and bi-monthly thereafter. After the trees were two years old, yearly top dressings of goat manure were added to the base of each tree and the trees were foliar fed with seaweed solution in the summer and fall. Very light skirt pruning is practiced annually.

The first cultivars to fruit were 'Suebelle' in 1988 followed by 'Lemon Gold' in 1989. The first frost of the decade occurred in the winter of 1988 (29° to 30°F; -1.7° to -1.1°C), followed by another series of frosts in 1989 (lows to 28°F; -2.2°C). These frosts injured the young growing shoots and caused fruit to drop, but there was otherwise no severe damage and the trees recovered rapidly in the spring. In December of 1990, a devastating "freeze of the century" struck California. Low temperatures in the sapote orchard were 20°F (-6.7°C) followed by 21°F (-6.1°C) and then two weeks of 25 to 28°F (-3.9° to -2.2°C) nights. The cultivars most susceptible to freeze damage were killed below the graft, although most survived to come back from the roots (Table 1).

In California, the trees do well on well drained sandy loam or clay soils. They grow and fruit well on the deep sands of Florida, but may become chlorotic on oolitic limestone (Morton 1987). The young branches are bright green but turn gray and become very strong with age. Trunks of older trees can become buttressed. Leaves are shiny above, glabrous below, and palmately compound with 3 to 5 pointed leaflets.

The small flowers are 5-petaled, creamy white with a greenish tinge, and occur in panicles of 5 to 100 in number. In California, many cultivars bloom in spring, summer, and fall. Blooming time varies among the cultivars which prolongs fruit harvest. Most trees have two successive blooming flushes, separated by several months. The panicles are usually held terminally or in bases of the branch shoots or axils of mature leaves. The flowers sometimes are cauliflorous (Batten 1984). If bees are in the area, pollination is no problem, but many flowers and immature fruits abort naturally. For maximum fruit size, the fruit should be thinned. Fruits ripen gradually about 4 to 5 months after pollination occurs. On most cultivars, the fruit remains green when ripe. The fruit is ripe when the skin yields to slight pressure. Fruits are spherical to slightly oval in shape and are 6 to 11 x 6 to 12 cm in size. Fruits have a cream to yellowish custard-like pulp with a melting flavor of banana, peach, and pear. Each fruit has one to four seeds which resemble those of a large orange or grapefruit and are reportedly fatally toxic if eaten (Morton 1987). The fruit quality is quite good in coastal areas of California (Chandler 1950).

Fruit should be hand harvested as many varieties bruise easily. The fruit may be harvested early, which may be an advantage if the fruit is to be shipped for the fresh market. The fruit is high in ethylene so postharvest handling procedures should avoid prolonged storage. Separation of fruit or wrapping individually may retard ripening. The fruit lasts up to two weeks when ripe, under refrigeration. Fruit should be packaged in a manner that avoids bruising. Many new packaging methods are being used for other crops, such as Asian pears, that may be easily adapted to the sapote. Cultivars with thicker skin are needed. The fruit is liked by most all who try it, so it may "market itself" once it becomes more readily available to the general public.




White Sapote

Table 1. Performance of white sapote cultivars planted as 2-year-old trees in 1986 in central California (Monterey Bay region).

Tree response to 1990 freezez
No. surviving
Cultivar above graft below graft No. dead Fruit Comments Tree type
Chestnut 6 2 2 Large, round, good flavor Commercial cv. in California Vigorous, upright
Denzler 2 4 Hawaiian cv.
Fred 9 1 Hardy Vigorous
Guinn 3 5 2 Performs poorly
Lemon Gold 1 5 Medium, round, good flavor Good producer in Southern California Small to medium
Malibu 1 2 3 Not vigorous
Miller 5 1 Flowers early Strong grower
McDill 7 2 1 Lg., round, yellowish Vigorous
Ortega 1 2 3 Weak grower
Pike 4 2 Lg., pointed, execellent flavor Prolific Small tree
Rainbow 2 7 1 Fast grower Bushy growth
Suebelle 10 Small, green Attracts pests Very bushy
T.S. Suebelle 3 3 Small, yellow Medium vigor Bushy growth
Sunrise 6 Small, green Mostly dead
White 3 3 Poor Mostly dead
Wilson 5 1 Execellent flavor, green Vigorous Bears year-round
Vernon 6 Medium to large Fast grower Rounded
Vista 6 Small, oval, good flavor Alternate bearing
zDecember 1990 freeze; low temperatures of -6.7°C followed by -6.1°C and 2 weeks of -3.9° to -2.2°C nights.

Last update April 24, 1997 aw