Journal of Heredity Advance Access originally published online on December 23, 2004
Journal of Heredity 2005 96(2):124-131; doi:10.1093/jhered/esi012
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Cytogenetics of Semi-Fertile Triploid and Aneuploid Intergeneric Vine Cacti Hybrids
From the Department of Life Sciences and The Institutes for Applied Research, Ben-Gurion University of the Negev, PO Box 653, Beer-Sheva 84105, Israel (Tel-Zur, Mizrahi); and Levi Eshkol School of Agriculture, The Hebrew University of Jerusalem, PO Box 12, Rehovot 76100, Israel (Abbo)
Address correspondence to Y. Mizrahi at the address above, or e-mail: mizrahi{at}bgu.ac.il
 |
Abstract |
|---|
 Top Abstract Materials and MethodsResultsDiscussionReferences |
|---|
Crosses between the diploid Hylocereus polyrhizus, as the female parent, and the tetraploid Selenicereus megalanthus, as the male parent, yielded triploid and aneuploid hybrids. The fruits of these hybrids combined the attractive appearance of Hylocereus fruits with the delicious taste of S. megalanthus fruits. The aim of this work was to assess the fertility and breeding potential of the triploid and aneuploid hybrids with a view to developing an improved vine cactus crop. Pollen mother cells at metaphase I revealed univalents, bivalents, trivalents, and occasionally quadrivalents. Chromosome distribution at anaphase I revealed different classes of chromosome segregation as well as lagging chromosomes. At metaphase II, parallel and tripolar spindles were observed. The occurrence of triads was frequent, whereas dyads were rarely observed. Pollen stainability varied among the clones studied ranging from 9.8% to 18.6%. The diameters of the stained pollen grains varied widely, probably as a result of the number of chromosomes. Despite the allotriploid origin of our hybrids, functional female and male gametes were produced in considerable proportions, most likely as a result of balanced chromosome segregation. The triploid and aneuploid clones studied yielded viable seeds whose number per fruit was strongly dependent on the pollen donor.
Vine cacti of the genera Hylocereus (Berger) Br. and R. and Selenicereus (Berger) Br. and R., which are native to northern South America, Central America, and Mexico, are currently being grown as new exotic fruit crops (Mizrahi and Nerd 1999). The genus Hylocereus comprises 16 species and the genus Selenicereus, 20 species (Barthlott and Hunt 1993). Hylocereus species are characterized by elongated stems (usually three-winged), branches carrying aerial roots, very large white (rarely red) flowers, and spineless fruits that are usually edible (Britton and Rose 1963). The fruits have foliaceous scales ranging from a few in number to many. Ovaries and flower tubes bear large foliaceous scales but no spines. Selenicereus species are distinguished by ribbed or winged stems, irregularly occurring aerial roots, flowers that are very often large, and large reddish fruits that are covered with clusters of deciduous spines, bristles, and hairs (Britton and Rose 1963). The scales of the ovaries and flower tubes are small, usually with long hairs and bristles in their axils. Unlike other species of Selenicereus, S. megalanthus has a three-winged stem, like that of Hylocereus, and spiny fruits like those of the Selenicereus. It was classified by Britton and Rose (1963) as a separate genus, Mediocactus, thereby implying both an intermediate morphology and an intermediate taxonomic status between the latter two genera.
Of the Hylocereus and Selenicereus species being cultivated (Barbeau 1990; Cacioppo 1990; Mizrahi et al. 1997), three species are being cultivated on commercial scale: H. undatus (in Colombia, Nicaragua, Israel, and Vietnam), S. megalanthus (in Colombia and Israel), and H. polyrhizus (in Israel) (Cacioppo 1990; Mizrahi et al. 1997). The fruits are becoming increasingly popular on European markets, with the volume of fresh vine cactus fruits imported to Europe having increased more than 10-fold over the past 5 years (Yael Kachel, Israel Ministry of Agriculture, personal communication).
Cytological observations have shown that Hylocereus and Selenicereus species are diploid, 2n = 2x = 22, whereas S. megalanthus is tetraploid, 2n = 4x = 44 (Beard 1937; Lichtenzveig et al. 2000; Spencer 1955; Tel-Zur et al. submitted). The tetraploid S. megalanthus crosses readily with the diploid Hylocereus species (Lichtenzveig et al. 2000; Tel-Zur et al. 2004b). On the basis of crossability and the intermediate morphology of S. megalantus, it has been suggested that this species is an allotetraploid that may have arisen by natural intergeneric hybridization between yet unknown Hylocereus and Selenicereus species (Lichtenzveig et al. 2000). Recent cytological and molecular studies have provided additional evidence in support of the proposed allotetraploid origin of S. megalanthus (Tel-Zur et al. 2004a, 2004b).
The large, attractive Hylocereus fruits lack taste, a disadvantage that has limited their marketability, whereas S. megalanthus fruits are very sweet and tasty but are inferior to Hylocereus fruits in size and appearance. An additional disadvantage of the S. megalanthus fruits is their spiny peel. From horticultural and commercial points of view, a combination of the large size, spinelessness, and attractive appearance of Hylocereus sp. fruits with the taste features of S. megalanthus fruits would be ideal.
Intergeneric-interploidy hybridization between Hylocereus species and S. megalanthus yielded triploids, pentaploids, hexaploids, and aneuploid hybrids (Tel-Zur et al. 2003). The cross between the diploid H. polyrhizus, as the female parent, and the tetraploid S. megalanthus, as the male parent, resulted in triploid and aneuploid hybrids that yielded fruits combining the attractive appearance of H. polyrhizus with the taste qualities of S. megalanthus. The fruit weight (200–400 g) of some of the triploid clones contributes to their marketability potential and makes them good candidates for commercial production. However, the spiny peel of all the triploid and aneuploid hybrids, inherited from S. megalanthus, is still a disadvantage that poses a breeding challenge.
It has been found that fruit set and fruit weight of vine cacti is positively correlated with seed number (Weiss et al. 1994). Preliminary observations showed that both triploids and aneuploids are partially fertile and produce viable pollen grains and viable seeds. At present, there is a lack of data on the cytological basis of gamete fertility and the influence of the pollen donor on seed set and fruit weight in these triploid hybrids.
In the present work, the meiotic behavior and viability of the pollen and ovules of our triploid and aneuploid hybrids were investigated, as was the influence of the pollen donor on seed set and fruit weight. In addition, the potential of these hybrids for future breeding programs was analyzed.
|
Materials and Methods |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Plant Material and Growth Conditions
The hybrids used in this study were obtained from a cross between H. polyrhizus (2n = 2x = 22, the female parent) and S. megalanthus (2n = 4x = 44, the male parent). The study was carried out during the flowering seasons of 2000–2002 on 3- to 4-year-old hybrid plants growing in Beer-Sheva and at the Habesor Research Station (Israel, 31°N, 34°E). Plant husbandry details were the same as those previously described by Lichtenzveig et al. (2000). Artificial pollination was performed manually with a brush in the morning on the day of anthesis. During the 2002 flowering season, pollination was controlled by covering each flower with a paper cup before and after pollination.
Cytological Observations of Pollen Mother Cells at Meiosis
Flower buds, 5–6 cm long, were fixed for 24 h in 3:1 ethanol:glacial acetic acid and then stored in 70% ethanol at 4°C. Meiotic chromosomes at metaphase I were observed in pollen mother cells (PMCs) excised from anthers of the fixed flower buds. PMCs were squeezed out of the anthers into a drop of 2% acetocarmine and squashed. The preparations were observed under a Zeiss AxiosKop 2 light microscope and photographed with a Zeiss AxioCam, AxioVision program, version 3.0.6 SP2.
Ovule Number, Pollen Stainability, and Pollen Diameter
Ovules were counted in at least four flower buds picked 1 or 2 days before anthesis. For each bud, the number of ovules was counted in a weighed sample of ovary tissue. The total number of ovules per bud was then calculated as the weight of the ovary x (no. of ovules in sample/weight of sample). Pollen stainability was determined in a minimum of 300 pollen grains per flower collected at anthesis from four or five flowers. Although 2% acetocarmine, Alexander's reagent, and fluorescein diacetate were equally effective stains, acetocarmine was used to stain the pollen, because it can be used for both fresh and stored pollen, whereas the other reagents are suitable only for fresh pollen. The diameters of a minimum of 300 stained pollen grains per flower, in three or four flowers, were measured with the aid of a calibrated micrometer under a Zeiss AxiosKop 2 light microscope.
Fruit Set, Seed Set, and Fruit Weight
The hybrid plants, used as the female parents, were back-crossed with H. polyrhizus and S. megalanthus, crossed with the diploid H. undatus, or selfed. Seven to 19 flowers were pollinated for each cross-combination. The fruits that developed were harvested at maturity and weighed, and seed number was determined. Brown- and black-coated seeds with germinability values of 0% to 90–95%, respectively, were obtained. Seed viability was determined as a percent of black-coated seeds that showed a normal appearance.
|
Results |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Meiotic Behavior
Chromosome configurations at metaphase I of two triploids and two aneuploids were comprised of univalents, bivalents, and trivalents (Table 1, Figure 1a). Occasionally, quadrivalents were also observed (clone 12-16). A mean of 4.5 to 7.9 univalents (with a range of 1 to 17) per PMC was found among the different clones. The number of bivalents ranged from 6.7 to 8.8, and the trivalents ranged from 3.0 to 5.5 per PMC. Chromosome disjunction data at anaphase I (Table 2) showed a high percent of balanced (or nearly so) chromosome distribution (e.g., 17:16; 17:17; 18:17). Lagging chromosomes were a common feature of anaphase I (1–4) and were observed in a range of 35% (clone 12-29) to 65% (clone 12-15) of the PMCs analyzed (Table 2, Figure 1b). It seems that some of these lagging chromosomes later form independent nuclei, which may result in polyads (Figure 1c, d).
|
|
|
Normal angular spindle orientation was observed in most PMCs (
87.6%), a value that was in line with the high percentage (69.5 and 83.3% for clones 12-31 and 12-16, respectively) of tetrads observed (Table 3). Parallel (Figure 1e) and tripolar (Figure 1f) spindles were also evident. Both symmetrical (Figure 2a) and asymmetrical tetrads (Figure 2b 1) were found. The frequency of triads (Figure 2b 2) was similar but somewhat higher than that of tripolar spindles (6.0% and 8.5% versus 4.5 and 3.2% for clones 12-31 and 12-16, respectively). Despite the relatively high frequency of parallel spindles (5–9.2%), dyads were very rare: 0.3% in clone 12-16 and 0.25% in clone 12-31 (Figure 2c). Polyad frequency was relatively high, being 8.5–24.0% (Figure 2d). Monads were rarely observed (0.25% in clone 12-31).
|
|
Ovule Number, Pollen Stainability, and Pollen Diameter
The average number of ovules per flower, determined in four clones (Table 4), varied from 3,560 in clone 12-29 to 4,900 in clone 12-31. Pollen stainability lay between 9.8% in clone 12-15 and 18.6% in clone 12-29 (Table 4). Pollen diameter of the stained grains varied widely, that is, between 60 and 160 µm (Table 4), with more than 90% of the stainable pollen grains being 80–120 µm in diameter.
|
Fruit Set, Seed Set, and Fruit Weight
The date and length of the flowering season varied widely among the different clones. The flowering seasons of both H. polyrhizus and H. undatus were long, extending from May and June, respectively, to the end of September. That of S. megalanthus was shorter, extending from mid-September to the end of November. There was thus only a short overlap of the S. megalanthus flowering season with those of the two Hylocereus species. In general, the lengths of the flowering seasons of the hybrids were shorter (only 1 month for clones 12-29 and 12-16) than those of both parent species, with the flowering periods falling almost within the dates of the flowering period of S. megalanthus. Because there was no overlap between the flowering seasons of clone 12-29 and the two diploid species, there was only enough pollen from H. polyrhizus to pollinate two 12-29 flowers and no pollen at all from H. undatus (Table 5).
|
High fruit set was obtained for clones 12-31 and S-75 pollinated with the pollen of the diploid Hylocereus species (H. polyrhizus and H. undatus) and of the tetraploid S. megalanthus. Self-pollination was successful only in clone 12-31, yielding small fruits (67 g). Low fruit set was obtained for clone 12-16 back-crossed with the parental species (Table 5).
The mean weights of fruits obtained by uncontrolled pollination during the 2000 and 2001 flowering seasons varied considerably: 118.3 ± 8 g in clone 12-29, 264.8 ± 17 g in clone 12-31, 248.2 ± 23 g in clone S-75, and 92.5 ± 15 g in clone 12-16. The controlled pollinations performed during the flowering season of 2002 yielded fruits of similar weights (Table 5).
The pollen source affected the total number of seeds (viable and nonviable) in clones 12-31 and S-75 in a similar way: pollination with pollen of the diploid Hylocereus species produced a large number of viable seeds (1325–1629), whereas pollination with pollen of the tetraploid S. megalanthus resulted in a drastic decrease of viable seeds numbers (535–683). Despite this low number of viable seeds, fruit weight was similar to that of fruits pollinated with Hylocereus pollen.
The highest percentage of fertilized ovules (yielding both viable and nonviable seeds) was obtained with H. polyrhizus pollen, the values being 7.5% for clone 12-16, 13.6% for clone 12-29, 34.6% for clone 12-31, and 49.5% for clone S-75. For these clones, the percentages of viable seeds were 6%, 12%, 27.5%, and 35.5%, respectively.
|
Discussion |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
The intergeneric triploid and aneuploid hybrids analyzed in this work were found to be partially fertile, having different proportions of viable female and male gametes. The viable gametes may indicate the occurrence of balanced segregation in anaphase I. Lichtenzveig et al. (2000) reported a regular meiotic process in the diploid H. polyrhizus and meiotic irregularities in the tetraploid S. megalanthus. These authors used the unbalanced anaphase I disjunction data of S. megalanthus to explain the reduced pollen viability found in this species. The meiotic chromosome association in metaphase I of our triploid and aneuploid hybrids comprised both rod and ring bivalents, the frequency of the ring bivalents being much lower than that of rod bivalents. This finding probably reflects the intergeneric chromosomal divergence between the parental lines. The univalents and multivalents observed in our triploid and aneuploid hybrids may have contributed to unequal segregation at anaphase I and consequently to a decrease in fertility. Chromosome disjunction at anaphase–telophase I showed several segregation classes. However, it is impossible to determine which segregation class produced viable gametes. It is likely that the presence of at least one homolog from each chromosome is a prerequisite for the maturation of a viable gamete. We assume that most of the products of polyad divisions contain incomplete chromosome complements and are therefore unlikely to mature into functional/stainable pollen grains.
A positive relationship exist between pollen viability and pollen stainability, as was found in the Hylocereus group (>90%) and S. megalanthus (25–40% during early–late season to 70–80% during flowering peaks) by Weiss et al. (1994), Lichtenzveig et al. (2000), and Tel-Zur et al. (2003). Therefore pollen stainability is an informative and reliable test notwithstanding minor difference between pollen viability (in vitro or in vivo) and pollen stainability in these taxa.
The occurrence of abnormal spindle geometry during metaphase II is a frequent cause of unreduced (2n) pollen in dicotyledons (Bretagnolle and Thompson 1995). The angular orientation of the double spindle in the second meiotic division permits the isolation of the four nuclei. Abnormal spindle geometry parallel or tripolar may lead to the formation of triads and dyads (Carputo et al. 1995; Conicella et al. 1996; Lopez-Lavalle and Orjeda 2002; Mok and Peloquin 1975). Our cytological observations revealed the presence of dyads, triads, and polyads, but no correlation was found between the frequency of parallel spindles and the number of dyads observed. In Solanum (Conicella et al. 1996; Watanabe and Peloquin 1993) and in Ipomoea batatas (Lopez-Lavalle and Orjeda 2002), a high correlation was reported between the occurrence of parallel spindles and the occurrence of dyads and hence with 2n pollen. On the other hand, a low correlation was found between the frequency of parallel and tripolar spindles and the frequency of dyads and triads observed in S. commersonii and Phureja-Tuberosom hybrids (Carputo et al. 1995) and in dihaploid roses (El Mokadem et al. 2002). Carputo et al. (1995) proposed that parallel spindles constitute a necessary—but not sufficient—condition for the formation of dyads. Other mechanisms may act at cytokinesis and lead to the formation of dyads. Multiple spindles were reported to be strongly correlated with the presence of polyads in Fucsia, S. tuberosum, and Zea mays (Conicella et al. 1996; Caetano-Pereira and Pagliarini 2001; Tilquin et al. 1984), but no multiple spindles were observed in our triploid hybrids. The micronuclei and polyads observed in somatic Citrus hybrids and Valencia (orange) resulted from the formation of univalents (Del Bosco et al. 1999). Li and Heneen (1999) reported similar findings in intergeneric hybrids between Brassica and Orychophragmus violaceus. Univalents were observed quite frequently at anaphase I in our triploid and aneuploid hybrids. Some of these lagging chromosomes did not migrate toward the two poles at anaphase I or anaphase II and probably contributed to the formation of polyads.
The wide range of pollen diameters observed may be associated with the different segregation classes observed during anaphase I and may represent different ploidy levels and different viability levels, because it has long been recognized that pollen size correlates with DNA content (Den Nijs and Peloquin 1977; Mendiburu and Peloquin 1976). The diameter of pollen grains of the diploid H. polyrhizus parent was 70–90 µm, whereas that of the tetraploid S. megalanthus parent varied widely, specifically, between 90 and 190 µm (Tel-Zur et al. 2003). The greater part of the pollen grains in S. megalanthus had an intermediate diameter of 120–140 µm and were considered to be the product of regular meiosis, carrying 22 chromosomes (Tel-Zur et al. 2003). The diameter of the pollen grains of our triploid and aneuploids hybrids varied between 60 and 160 µm. The small pollen grains (60–80 µm) were most likely haploid (n = 11), the few large ones (130–160 µm) were probably unreduced gametes, and the abundant medium-size grains (90–120 µm) most likely carried intermediate chromosome numbers. The finding of 7.1% of large pollen grains in clone 12-16 and 6.6% in clone 12-31 accorded well with the estimated 8.8% and 6.3% frequency, respectively, of 2n pollen at the tetrad stage (triads and dyads) in the two clones.
The average number of ovules in the triploid hybrids was large, being similar to that of H. polyrhizus. However, only some of the ovules developed into viable and nonviable seeds. The fertility of female gametes was higher than that of the male gametes in clones 12-31 and S-75. A maximum of 27.5% ovules produced viable seeds in clone 12-31 and 35.4% in clone S-75.
The triploid and aneuploid hybrids were produced as the first step in a breeding program aimed at combining the desirable traits of the two genera. Based on fruit set and fruit weight, the most promising candidates for cultivation are clones 12-31 and S-75. Clone 12-31 is self-compatible, but its small (67 g) fruits are of no economic value. Hence, from a horticultural point of view fruit production with this clone will require cross-pollination. For further breeding, all the progeny described in this article, especially clones 12-31 and S-75, can serve as promising parents. Breeding can be performed by selfing or crossing with either parent to combine desired traits in future hybrids. The intergeneric hybrids show classical hybrid vigor, as evident from their growth and development under extreme environmental conditions of high temperatures, low relative humidity, and long drought periods (unpublished data). However, it should be noted that these have a long juvenile period of 3–5 years, which prolongs the breeding process considerably.
|
Acknowledgments |
|---|
The study was supported in part by the UCLA-BGU Program of Academic Cooperation. The authors thank Mr. Joseph Mouyal (Ben-Gurion University of the Negev) for skillful technical assistance and Ms. Dorot Imber and Ms. Inez Mureinik (Ben-Gurion University of the Negev) for editing the manuscript.
|
Footnotes |
|---|
Corresponding Editor: J. Perry Gustafson
Received March 30, 2004
Accepted July 20, 2004
|
References |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
-
Barbeau G, 1990. La pitahaya rouge, un nouveau fruit exotique. Fruits 45:141–147.
Barthlott W and Hunt DR, 1993. Cactaceae In: The families and genera of vascular plants, vol. 2. (Kubitzki K, ed). New York: Springer-Verlag; 161–197.
Beard CE, 1937. Some chromosome complements in the Cactaceae and a study of meiosis in Echinocereus papillosus. Bot Gaz 99:1–21.[CrossRef]
Bretagnolle F and Thompson JD, 1995. Gametes with the somatic chromosome number: mechanisms of their formation and role in the evolution of autopolyploid plants. New Phytol 129:1–22.[CrossRef]
Britton NL and Rose JN, 1963. The Cactaceae, vol. 2 New York: Dover; 183–212.
Cacioppo O, 1990. Pitaya: una de las mejores frutas producidad por Colombia Informativo Agro Economico. Febrero 15–19.
Caetano-Pereira CM and Pagliarini MS, 2001. A new meiotic abnormality in Zea mays: multiple spindles associated with abnormal cytokinesis in both divisions. Genome 44:865–871.[Medline]
Carputo D, Cardi T, Frusciante L, and Peloquin ST, 1995. Male fertility and cytology of triploid hybrids between tetraploid Solanum commersonii (2n = 4x = 48, 2EBN) and Phureja-Tuberosum haploid hybrids (2n = 2x = 24, 2EBN). Euphytica 83:123–129.[CrossRef]
Conicella C, Genualdo G, Errico A, Frusciante L, and Monti LM, 1996. Meiotic restitution mechanisms and 2n pollen formation in a Solanum tuberosum dihaploid and in dihaploid x wild species hybrids. Plant Breed 115:157–161.[CrossRef]
Den Nijs TPM and Peloquin SJ, 1977. Polyploid evolution via 2n gametes. Am Potato J 54:377–386.
Del Bosco SF, Tusa N, and Conicella C, 1999. Microsporogenesis in a Citrus interspecific tetraploid somatic hybrid and its fusion parents. Heredity 83:373–377.[CrossRef][Web of Science][Medline]
El Mokadem H, Crespel L, Meynet J, and Gudin S, 2002. The occurrence of 2n pollen and the origin of sexual polyploids in dihaploid roses (Rosa hybrida L.). Euphytica 125:169–177.[CrossRef]
Li Z, and Heneen WK, 1999. Production and cytogenetics of intergeneric hybrids between the three cultivated Brassica diploids and Orychophragmus violaceus. Theor Appl Genet 99:694–704.[CrossRef]
Lichtenzveig J, Abbo S, Nerd A, Tel-Zur N, and Mizrahi Y, 2000. Cytology and mating systems in the climbing cacti Hylocereus and Selenicereus. Am J Bot 87:1058–1065.
Lopez-Lavalle LAB and Orjeda G, 2002. Occurrence and cytological mechanism of 2n pollen formation in a tetraploid accession of Ipomoea batatas (sweet potato). J Hered 93:185–192.
Mendiburu AO and Peloquin SJ, 1976. Sexual polyploidization and depolyploidization: some terminology and definitions. Theor Appl Genet 48:137–143.
Mizrahi Y and Nerd A, 1999. Climbing and columnar cacti In: New arid land fruit crops. Perspective in new crops and new uses (Janick J, ed). Alexandria: ASHS Press; 358–366.
Mizrahi Y, Nerd A, and Nobel P, 1997. Cacti as crop. Hort Rev 18:291–320.
Mok DW and Peloquin SJ, 1975. Three mechanisms of 2n pollen formation in diploid potatoes. Can J Genet Cytol 17:217–225.[Web of Science]
Spencer JL, 1955. A cytological study of the Cactaceae of Puerto Rico. Bot Gaz 117:33–37.[CrossRef]
Tel-Zur N, Abbo S, Bar-Zvi D, and Mizrahi Y, 2003. Chromosome doubling in vine cacti hybrids. J Hered 94:329–333.
Tel-Zur N, Abbo S, Bar-Zvi D, and Mizrahi Y, 2004a. Clone identification and genetic relationship among vine cacti from the genera Hylocereus and Selenicereus based on RAPD analysis. Sci Hortic 100:279–289.[CrossRef]
Tel-Zur N, Abbo S, Bar-Zvi D, and Mizrahi Y, 2004b. Close genetic relationship among vine cacti of the genera Hylocereus and Selenicereus. Ann Bot 94:527–534.
Tilquin JP, de Brouwer K, and Horvat F, 1984. Unusual cytological patterns in microsporogenesis in a cultivar of Fuchsia. Multiple spindle. Theor Appl Genet 67:413–417.
Watanabe K and Peloquin ST, 1993. Cytological basis of 2n pollen formation in a wide range of 2x, 4x, and 6x taxa from tuber-bearing Solanum species. Genome 36:8–13.
Weiss J, Nerd A, and Mizrahi Y, 1994. Flowering behavior and pollination requirements in climbing cacti with fruit crop potential. HortScience 29:1487–1492.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||