Domestication of the bush mango (Irvingia spp.): some exploitable intraspecific variations in west and central Africa
Irvingia gabonensis and Irvingia wombolu, the eating and the cooking types, respectively, of bush mango, have been identified by the International Centre for Research in Agroforestry as priority wild fruit tree species for domestication. With information on farmer trait preference, and in collaboration with national agricultural research systems, rangewide germplasm collections have been made. Germplasm banks have been established for genetic conservation and for early assessments of progeny variation. These trials, together with observations on mature field-grown trees, are revealing substantial intraspecific phenotypic differences.
Morphological variations in vegetative and reproductive characteristics, including fruit quality attributes, have been found. Much of this intraspecific variation in the Irvingia species of west and central Africa provides opportunities for genetic improvement by genotypic selection and vegetative propagation. This could promote the domestication process of this important wild tree as a crop for agroforestry. This genetic improvement needs to go hand in hand with the commercialization of the bush mango's products to provide means for farmers to supplement their other forms of income, and to help develop permanent land-use options, based on improved materials, as an alternative to slash-and-burn agriculture.
The natural forests of west and central Africa are rich in natural resources and have tremendous biodiversity (FAO 1983), particularly in trees that provide food, fuel, fibre, medicines and various other products, including construction and building materials. Irvingia gabonensis (Engl.) Engl. and Irvingia wombolu (Vermoesen) produce edible fruits and seeds (fig. 1). The sustainability of these natural resources has been the concern of various workers (National Research Council 1991), particularly with the continued clearing and selective exploitation of forests (Palmberg 1984). This situation is causing substantial loss in germplasm and also degradation of particular genepools (Gale & Lawrence 1984). The need for conservation, domestication and development of sustainable production systems, such as are possible through agroforestry, has been emphasized if the environment is to recover from this degradation process. The genetic benefits obtained from the domestication of mango (Mangifera indica L.), Citrus spp., breadfruit (Artocarpus altilis (Z) Fosb.) and avocado (Persea americana Miller) in the tropics has shown that large genetic gains are possible and suggests that similar opportunities may await the native wild fruits of west and central Africa.
Figure 1. Irvingia gabonnensis: (a) whole fruit, (b) vertical section and (c) extracted seed covered by hard and fibrous endocarp.
The Irvingia domestication programme arose from the identification of farmer's priorities and trait preferences (Ladipo et al. 1995) using ICRAF's guidelines for priority setting among candidate species (Franzel et al. 1996, Jaenicke et al. 1995). The next step is to study the array of variation available in the species. This is done over the natural range of the species (fig. 2), with particular reference to those traits identified by farmers as being the characteristics appropriate for genetic selection. Consequently, the desirable inherent variation as observable phenotypic characteristics is identified and ready for selection.
Figure 2. Natural distribution of Irvingia gabonensis in West and Central Africa.
The International Centre for Research in Agroforestry (ICRAF) has organized, with its national agricultural research system (NARS) partners, a systematic collection of good quality seeds from Ghana, Nigeria, Cameroon and Gabon, with emphasis on southeast Cameroon and northern Gabon, the centre of diversity of the trees Irvingia gabonensis and Irvingia wombolu (Ladipo et al. 1994, Ladipo et al. 1995).
It is well known that considerable variation exists within and between populations of tropical trees (National Research Council 1991). The patterns and extent of genetic diversity in forest trees are strongly determined by their mating systems and geneflow. However, very little is known about these aspects of Irvingia spp., except that in I. gabonensis the flower is hermaphroditic. Controversy, however, exists about the level of outbreeding. Ujor (1995) has reported low levels of inbreeding, while recent work by Ladipo (unpublished) indicates 100% outbreeding. This indicates that I. gabonensis, like other tropical trees, is highly heterozygous.
An important step in a tree improvement programme, particularly in the case of wild species, is definition of the selection criteria. The type and intensity of selection will depend on the pattern of variability within and between populations (Bawa 1976), and variability is the building block for trait improvement through breeding or genetic selection.
An individual or a population's expression of its genetic composition represents an interaction with its environment. This determines the phenotypes of the individual plants and thus their observable morphological characteristics. Most of these characteristics are influenced by a large number of genes. The influence of a single gene on such polygenic traits may be very small. However, the expression of rare recessive genes may result in a trait that is important for the development of a potentially valuable cultivar.
Very little information exists on inter- or intraspecific variations in many of the tropical tree species that are candidates for domestication, including the Irvingia species. Variations are reported below for crown shape, leaf colour, phenology and various fruit characteristics observed between populations and individual trees of Irvingia gabonensis in west and central Africa (table 1).
Table 1. Variation in fruit yield in 1995 between
some 5-year-old trees
of I. gabonensis planted in Onne (southeast Nigeria)
|Fruits produced in 1995
The flowering and fruiting process consists of a series of sequential stages, all of which must proceed unhindered for a successful fruit harvest (Sedgley & Griffin 1989). The key steps are floral initiation, flower development, pollination, fruit set, fruit development and seed set. Fruit set may depend on the availability of pollen and the efficiency of pollinators, but seed set that is the development of viable seeds from the ovules of pollinated flowers depends on endogenous factors and their interaction with the environment. All these factors are under genetic control but can be manipulated by management factors. The success of seed set affects the variation in seed viability and thus germination.
Tree-to-tree variation in fruit yield has been observed over 2 years from a plot of I. gabonensis planted at the International Institute for Tropical Agriculture's (IITA), high-rainfall research station at Onne in Nigeria in 1990. For example, trees 160 and 162 produced 6 and 162 fruits in 1994 respectively. In 1995, these two trees produced very similar numbers of fruit and represented the extremes in terms of fruit yield among a small group of trees (table 1). Trees with a good yield of high-quality fruits are obviously potential candidates as future cultivars for agroforestry systems, and hence the selection and mass propagation of these individual trees is a key component of the domestication programme.
Chemical changes during ripening usually involve the conversion of starch to sucrose and reducing sugars. The extent of this conversion affects the sweetness of ripe and mature fruits (Alston 1992) as well as their retention on the tree.
During the Irvingia germplasm collections, it was specified that only mature and ripe fruits, as identified by colour and a slight shake of the mother-tree, should be collected. One hundred mature fruits were collected from each tree (where possible) and sweetness assessed by `tasters'. Four fruits were tasted to determine the quality (sweetness) of the fruits. Fruits were divided into three categories: very sweet, sweet and not sweet. Most were assessed as sweet (table 2), but substantial variation in fruit sweetness between accessions was observed in this exercise.
Table 2. Classification of accessions (by number) for fruit sweetness in I. gabonensis collected in Cameroon (C)
|Very sweet||Sweet||Not sweet|
|C 58||C22, C27||C44|
All other 1994/95 collections
These simple and practical relevant results can be followed up by the more precise and technically sound method used, for example, to assess sweetness in apples (Alston 1992), in which the standard indicator bromocresol green is applied on the cut surface of the fruit to determine the level of sugar content.
Fruit pigmentation is a major attribute in fruit marketing. For example, red apples have greater consumer appeal than green ones (Sedgley & Griffin 1989). The major pigments responsible for fruit colours are chlorophyll, carotenoids and anthocyanins. The chloroplasts in green immature fruits generally lose chorophyll on ripening and increase other pigments (Goodwin & Goad 1970). In many fruits, the carotenoid content varies, depending on cultivar or genotype, conferring variation in orange and yellow colouration in fruits. Increase in vacuolar concentration of anthocyanin as fruits ripen confers red colouration (van Buren 1970).
In the wild, fruit colour is thought to influence the interest of biotic dispersal agents. Irvingia fruits change from green to yellow as fruits ripen. The mature fruit colour has, however, been found to vary from tree to tree (table 3). This immense variability in colour is a resource that can be utilized for colour selection. The resulting consumer appeal should enhance the commercialization process of this wild fruit.
Table 3. Fruit colour, in Irvingia accessions collected in Nigeria (A or B), Cameroon (C) and Ghana (G) in 1994/95, as defined in the Methuen handbook of colour (Korneup & Wanscher 1984)
|Colour type||Colour||Colour code*||Accessions|
|4||brownish (cork) yellow a/b||4-C-4/5||G4|
|5||bright reddish yellow a/b||4-A-7/8||A22
*as defined in the Metheun handbook of colour (Korneup & Wascher 1984)
T = tree
Similar to the case of already domesticated tropical nut trees, the endocarp of Irvingia needs to be cracked open (shelled) to extract the cotyledons (kernels), which are used as thickening agents in soups and stews. In Nigeria, these kernels are called ogbono in Ibo (Okafor 1978) and apon in Yoruba.
From Gabon to Cameroon and Nigeria, people utilize different techniques to extract Irvingia kernels. They can be extracted from fruits in the fresh state, or fruits can be fermented and the kernels extracted wet. Alternatively, they can be fermented and sun dried before extracting, packaging or marketing. All these methods are difficult and hazardous. The whole operation also takes a lot of farmers' time: it is estimated to involve over 288 hours in 3 months-time that could have been spent on other farm activities (Ladipo 1995).
During the 1995 Irvingia collections in Gabon, a tree (G28) was found in Bibas, in the north of Gabon, whose nuts split open naturally. The farmer who owned the tree well recognized the advantage of these self-splitting fruits (D. Boland, pers. comm). Fruits were collected from it and divided into three groups; one was processed and spread out to dry while the other two groups were sent to Onne in Nigeria and M'Balmayo in Cameroon for germination and the establishment of living genebanks. After 72 hours, 30 out of the 33 seeds had already split open, exposing intact cotyledons. After 70-72 hours, 93% of the seeds from G28 had split open, while none of the seeds from six other accessions in Gabon split at all. This useful trait is an early splitting of the hard endocarp, a process that usually takes place much later during germination.
In the domestication of I. gabonensis, special attention needs to be placed on fruit shape and form. Spherical fruits are the most common, but a range of other shapes has been recorded (fig. 3). For example, in Cameroon, accession C44 showed an unusual rectangular fruit shape, which was later also found to occur in Nigeria and Gabon. Piriform fruits are also found fairly frequently. This character was recorded for all mother trees in the passport data forms of each accession collected by ICRAF.
Figure 3. Variation in fruit shapes of Irvingia gabonensis.
In Onne (southeast Nigeria), fruit malformation or `dimpling' was observed on some trees. Observations over 1 to 3 years has shown that this characteristic is repeated and is thus probably heritable. However, such deformations could be caused by insect or disease attack. Fruits from suspect trees in Onne were therefore examined in situ throughout their developmental stages. Sample fruits were cut open but did not reveal disease or insect attack at any stage. Cell collapse or failure of cells to divide during development could be responsible for this malformation, but this characteristic requires more investigation. In fact, it is important to identify this and other characteristics that can reduce the value or quality of Irvingia so that they can be eliminated early in the improvement programme for this species.
The time at which fruits ripen for harvest is a major determinant of the annual schedule of management practices in orchards. It also affects the profitability of the commercial crop (Sedgley & Griffin 1989).
The first stage in the flowering process is floral induction or evocation (Sedgley & Griffin 1989). It is not known what triggers this process in Irvingia species, but a substantial variation was observed in the number of floral flushes within the population of 182 trees planted in 1990 (table 4), with most trees not flowering at this age or flowering only once a year, but with a few flowering 2-4 times.
Table 4. Variation in the number of floral flushes
of I. gabonensis 4-5 years old at Onne, Nigeria
|No of flushes||1994||1995|
During the 1994 Irvingia germplasm collection in Cameroon, a large diversity in fruiting time was observed (see Fondoun et al. 1994). Maturity of fruits was early in the southwest and littoral regions, while fruiting was late in the central, eastern and southern regions of the humid lowlands of Cameroon (see fig. 4).
Fruit maturity in Irvingia gabonensis can be in July-August (early) and August-September (late). However, some out-of-season fruiting seems to occur. In Onne (southeast Nigeria) trees T53, T126 and T162 fruited in February (4 months before general fruiting).
In the moist tropical forests of Gabon, Equatorial Guinea and southern Cameroon, field exploration confirmed the existence of indigenous knowledge about fruiting times and showed that trees of Irvingia gabonensis can flower and produce fruits twice a year. During the 1994 and 1995 collections, there was an opportunity to ask farmers who have lived all their lives in the forest about other variations. The Pygmies of the Dja Forest (Cameroon) and the Ibos of southeast Nigeria, in particular, have a lot of indigenous knowledge about their local fruit trees. In this area, some trees produce fruits in January-February and again in June-August. This is quite different from most other areas, which produce fruit only once a year, and indeed, sometimes even only every other year, as was reported by farmers for some trees in the littoral region of Cameroon. Ake Assi and Okafor (unpublished) have also observed trees that flower and fruit more than once a year in Ivory Coast and in Nigeria, respectively.
If variations of this type can be captured in domesticated varieties, the commercial viability of the crop would be greatly enhanced. Research is therefore needed to understand the genetic basis of this characteristic, so it can be effectively utilized in Irvingia.
Figure 4. Variation in crown shape of Irvingia gabonensis (* = Principal types)
Precocity in plants is known to be under some degree of genetic control, with considerable variation in the time taken for trees to reach maturity and start fruiting. I. gabonensis is generally said to commence flowering after 10-15 years, but Ladipo and Anegbe (1995) have reported variation in the commencement of flowering in young I. gabonensis in southeast Nigeria. This has also been indicated by farmers who reported that fruiting could commence much earlier in a precocious tree genotype. Near Ibadan, Adedeji (pers. comm.) reported that two of his three Irvingia trees began to fruit at year 6. Within a 182-tree plot of I. gabonensis planted in 1990 at Onne, 83 trees flowered in 1994; 100 trees subsequently flowered in 1995 (45%), with many of the same trees flowering and fruiting a second time.
Polyembryony is the formation of more than one embryo in an ovule. It can be an advantage, as it is a form of clonal propagation, the nucellar embryos being identical to those of the female parent. Sedgely & Griffin (1989) have reported that polyembryony is under genetic control in some fruit tree crops. In I. gabonensis the level of polyembryony was found to differ between some accessions collected in the wild. Accessions that showed polyembryony in Ibadan also showed high levels of polyembryony in M'Balmayo, Onne, and Ibadan (see table 5), indicating that it is probably an inherent characteristic in these accessions exhibiting this trait. Assessment of six kernel samples of Irvingia, however, revealed that the weight of the two kernels from a polyembryonic fruit were about equivalent to the weight of the single kernel of normal fruit from the same accession. Therefore, from a production point of view, there seems to be little reason to select for this characteristic. However, the trait could be useful for research purposes.
Table 5. Incidence of polyembryony (% of fruits examined) in some Irvingia accessions (A = Nigeria, C = Cameroon)
The form and shape of tree crowns can vary substantially and is probably due to genetic variation in apical dominance and apical control. Ladipo et al. (1991) identified genetic variation in apical dominance as the process determining clonal variation in the production of branches and branching frequency of Triplochiton scleroxylon K. Schum. trees and thus differences in their branching habit and crown form. In fruit trees, horticulturalists further manipulate tree form by pruning and mechanical treatments to promote flowering and improve fruit yields (Quinlan & Tobutt 1990). Branching and form traits are incorporated into breeding and selection programmes (Alston 1976), as they can be of substantial value for orchard development.
For agroforestry, substantial attention needs to be given to the crown attributes of trees, particularly their size and density, as affected by the numbers of branches, branching angle and branch retention, since crown form will affect the extent to which the tree shades interplanted crops.
Domesticated I. gabonensis acceptable for both agroforestry and intensive orchard management may eventually be required. This might involve the need for different crown and bole forms. These potential differences therefore need to be considered as selection criteria for improved cropping efficiency and to meet cultural and commercial needs.
Okafor (1974) described interspecific crown variations and deliminated Irvingia gabonensis into two varieties (var. excelsa and var. gabonensis). Harris (1993) has more recently reclassified these varieties as species I.wombolu and I. gabonensis, respectively. This reclassification was based on fruit and crown form characteristics.
Isolated mature (10-25 years) trees of I. gabonensis have been assessed for their crown characteristics over the last 4 years in Cameroon, Gabon and Nigeria. Considerable variation is apparent (fig. 4). The genetic nature of these differences needs to be confirmed. Pruned and plantation trees were not included because of the possibility that their crown form results from competition in a forest type situation. It is not known, however, whether these assessed trees have been isolated throughout their life.
Colour change in the leaves of tropical trees is not uncommon as the leaves expand and mature (Longman & Jeník 1987).
In Irvingia gabonensis young leaves are usually pale green and sometimes pink. Field and nursery observations have shown that some trees of Irvingia gabonensis have red leaves (Okafor pers. comm., Ujor 1995) even as mature trees. In our collections, accession B40 (Nigeria) showed this characteristic in 14 out of its 30 seedlings. To a lesser extent, this leaf colouration was also observed in the progeny of four other accessions, in the M'Balmayo Field Station of IITA in Cameroon. This phenomenon occurs only in I. gabonensis and not I. wombolu. Red-leaved genotypes might have potential as cultivars for amenity plantings as the characteristic could be captured clonally by vegetative propagation.
Accessions of I. gabonensis collected in Nigeria, Cameroon and Gabon in August 1994 and January 1995 were divided into three lots for establishment in live genebanks within this subregion (table 6). In Nigeria two genebanks were established, one at Ibadan in collaboration with the National Centre for Genetic Resources and Biotechnology (NACGRAB), and one in Onne in collaboration with the International Institute of Tropical Agriculture (IITA). A third genebank was established at M'Balmyo in Cameroon in collaboration with IRA (Institut de Recherche Agronomique).
Table 6. Accessions of conserved germplasm of Irvingia spp. in west and central Africa (Nigeria and Cameroon) in 1995
|Accessions conserved (no.)|
|Site||Bank 1||Bank 2||NARS and CG|
The National Academy of Sciences (1975) and others have suggested the need to promote minor crops, while Leakey & Newton (1994) have emphasized the need to exploit their genetic diversity and to domesticate them for agriculture, agroforestry and forestry. ICRAF sees domestication as an opportunity to make agroforestry systems more attractive to farmers and to alleviate poverty, while at the same time promoting tree planting with indigenous trees, diversifying land-use systems and developing dynamic agroecosystems. The combination of domestication strategies with an improved policy environment and techniques to alleviate soil fertility depletion are seen by ICRAF as an opportunity to develop better land-use practices in sub-Saharan Africa (Sanchez & Leakey in press). This is urgent as tropical forests are still being destroyed and their genetic resources continue to be eroded. This paper has enumerated substantial phenotypic variations in Irvingia species, which could be exploited to ensure their successful genetic improvement and domestication. Vegetative propagation (see Ladipo 1995) and selection provide quick and reliable means to capture this genetic variation and to domesticate tree species (Leakey & Jaenicke 1995). This approach would enhance the efforts already carried out by farmers, selecting for crops that fit their specific ecosystems and to provide their needs.
Further study is required of the genetic variability in Irvingia species and to confirm their origin and values. We suggest further work on the following issues:
· Continue to study Irvingia breeding systems
· Continue to collect and conserve more accessions, including the utilization of in-situ techniques
· Continue to identify variation in phenotypic characteristics and initiate corresponding chemical and analytical studies to determine descriptors of fruit quality
· Undertake molecular genetic studies within and between populations to elucidate patterns of genetic variation
· Conduct cytological studies to provide information on chromosome numbers, etc.
· Continue to develop propagation techniques, especially for mature tissues (see Ladipo in prep)
· Develop field trials on station and on farm to test performance of putative cultivars
· Investigate management and genotype interactions to develop production systems
The germplasm and knowledge already assembled indicates that domesticated varieties offer the potential to enhance the fruit quality, including nutrition, environment and economy of the region.
Alston F.H. 1976. Dwarfing and lethal genes in apple progenies. Euphytica 25:505-514.
Alston F.H. 1992. Flavour improvement in apples and pears through plant breeding. Phytoparasitica 20:33-41.
Bawa K.S. 1976. Breeding of tropical hardwoods: an evaluation of underlying bases, current status and future prospects. p 43-59. In: Tropical trees: variation, breeding and conservation, ed. J. Burley & B.T. Styles. Academic Press, London.
FAO. 1983. Food and fruit breeding forest species. In: Examples from East Africa. FAO Forestry Paper No. 44/1. p 172.
Fondoun J.M., Kengue J. and Ndung'u J. 1994. Irvingia germplasm collection in Cameroon. Unpublished report. ICRAF, Nairobi, Kenya.
Franzel S., Jaenicke H. & Janssen W. 1996. Choosing the right trees: setting priorities for multi-purpose tree improvement. ISNAR Research Report No. 8. ISNAR, The Hague. 87 p.
Gale J.S. & Lawrence M.J. 1984. The decay of variability. p 77-85. In: Crop genetic resources: conservation and evaluation, ed. J.H.W. Holden & J.T. Williams. IBPGR, Rome, and John Allen and Unwin, London.
Goodwin T.W. & Goad L.J. 1970. Carotenoids and trite + penoids. p 305-368, In: The biochemistry of fruits and their products, vol. 1, ed. A.C. Hulme. Academic Press, New York.
Harris D.J. 1993. A taxonomic revision and ethnobotanical survey of the Irvingiaceae in Africa. D.Phil. thesis. University of Oxford. 274 p.
Jaenicke H., Franzel S. & Boland D.J. 1995. Towards a method to set priorities amongst species for tree improvement research: a case study from West Africa. Journal of Tropical Forest Science 7:490-506.
Korneup A. & Wanscher J.H. 1984. Methven handbook of colour, 3rd ed., rev. by Don Pavey. Methven. p 16-136.
Ladipo D.O. 1995. Management and genetic improvement of MPTS in humid lowlands of West and Central Africa. p 79. In: ICRAF 1995 annual report. ICRAF, Nairobi, Kenya.
Ladipo D.O. In preparation. Vegetative propagation of Irvingia gabonensis from marcotts.
Ladipo D.O. & Anegbe P. 1995. Flowering and fruiting in field grown bush mango (I.gabonensis and I. wombolu) in southeast Nigeria. Paper presented at the 23rd Annual Conference of the Genetics Society of Nigeria, 14-18 May 1995. 9 p.
Ladipo D.O., Adewusi H., Fatokun H.K., Adebisi A. & Obisesan K. 1995. Irvingia wombolu (syn. I. gabonensis var. excelsa Okafor). Germplasm exploration and collection in southern Nigeria. ICRAF report on genetic resources collection activity in HULWA. 13 p.
Ladipo D.O., Britwum S.P.K., Tchoundjeu Z., Oni O. & Leakey R.R.B. 1994. Genetic improvement of west African tree species: past and present. p 239-248. In: Tropical trees: the potential for domestication and rebuilding of forest resources. ed. R.R.B. Leakey & A.C. Newton. HMSO, London.
Ladipo D.O., Leakey R.R.B. & Grace J. 1991. Clonal variation in apical dominance of Triplochiton scleroxylon K. Schum in response to decapitation. Silvae Genetica 40(3/4):135-140.
Leakey R.R.B. & Jaenicke H. 1995. The domestication of indigenous fruit trees: opportunities and challenges for agroforestry. p 15-26. In: Proceedings of 4th International BIO-REFOR Workshop, ed. K. Suzuki, S. Sakurai, K. Ishii and M. Norisada. BIO-REFOR, Tokyo.
Leakey R.R.B. & Newton A.C. 1994. Domestication of `Cinderella' species as the start of a woody-plant revolution. p 3-6. In: Tropical trees: the potential for domestication and rebuilding of forest resources. ed. R.R.B. Leakey & A.C. Newton. HMSO, London.
Longman K.A. & Jeník J. 1987. Tropical forests and their environment, 2nd ed. Longmans, London. 185 p.
National Academy of Sciences. 1975. Underexploited tropical plants with promising economic values, NAS, Washington DC. 188 p.
National Research Council. 1991. Managing global genetic resources: forest trees. National Academy Press, Washington, DC. 228 p.
Okafor J.C. 1974. Varietal delimitation in Irvingia gabonensis (Irvingiaceae). Nigerian Journal of Forestry 4:80-87.
Okafor J.C. 1978. Development of forest tree crops for food supplies to Nigeria. Forest Ecology and Management 17:1-11.
Palmberg C. 1984. Genetic resources of arboreal fuelwood species for the improvement of rural living. In: Crop genetic resources: conservation and evaluation, ed. J.H.W. Holden & J.T. Williams. IBPGR, Rome, and John Allen & Unwin, London.
Quinlan J.D. & Tobutt K.R. 1990. Manipulating fruit tree structure chemically and genetically for improved performance. Horticultural Science 25(1):60-64.
Sanchez P.A. & Leakey R.R.B. In press. Land-use transformation in Africa: three determinants for balancing food security with natural resource utilization. In: Proceedings of European Society of Agronomy Symposium, 7-11 July 1996, Wageningen, The Netherlands. 15 p.
Sedgley M. & Griffin A.R. 1989. Sexual reproduction of tree crops. Academic Press, Sydney, Australia. 378 p.
Ujor G.C. 1995. Reproductive biology of Irvingia gabonensis (O'Rorke) Bail in southern Nigeria: phenology, floral biology and varietal characteristics. PhD thesis, University of Ibadan, Nigeria. 355 p.
van Buren J. 1970. Fruit phenolics. p 269-304. In: The biochemistry of fruits and their products, vol. 1, ed. A.C. Hulme. Academic Press. New York.
Plate 21. A germplasm collector showing fruits of Irvingia gabonensis collected in southwest Nigeria. (photo: D. Ladipo)
Plate 22. Fruits of Irvingia gabonensis (bush mango), one of the priority indigenous tree species for domestication. (photo: D. Ladipo)
Plate 23. Variation in colour and size of mature fruits of Irvingia gabonensis collected in southeast Nigeria. (photo: D. Ladipo)
Plate 24. Kernels (processed and unprocessed) of the seeds of Irvingia gabonensis (bush mango). These are used to thicken so ups and stews and are widely traded in West Africa. (photo: R.R.B. Leakey)