FLOWERING AND FRUITING PATIERNS IN MILICIA EXCELSA AND MILICIA REGIA WELW.

Risper Nyagoy Nyong'o,' J.R. cobbinah and J. Appiah-Kwarteng

Forestry Research Institute of Ghana, UST P.O. Box 63, Kumasi, Ghana

ABSTRACT -:._The reproductive biology of Milicia excelsa and M. regia was studied Tne major distinguishing traits were crown shape, bark texture and leaf traite such as shape, dimension, colour, number of lateral nerves and arrangement on branchelets. Male and female trees are distinguished by the size of crown and tree trunk, forking characteristics, and distribution and .abundance of )wers in the crown. Male flowers in general are longer and slender than Lenale flowers.

It takes approximately 5-6 weeks from time of fertilization to fruit maturation. Premature abscission of fruits is not uncommon. Whiles fruiting may not occur every year the results suggests that in a good fruiting year odum tree can be quite prolific. The observational studies reported here have implications in breeding programme for odum.

Keywords - Milicia spp., reproductive biology, flowering patterns, fruiting patterns

INTRODUCTION

Odum (Milicia excelsa and Milicia lata. The attack leads to formation regia) is one of the most important of galls by the leaves and buds. timber species found in Ghana. It's The insect attack is followed by trade name is Iroko. Milicia was dieback, resulting in retarded until recently known as growth and eventually death of the Chlorophora. It is used for plant (Taylor, 1960; Wagner et ai, construction work, joinery, cabinet 1991)• -making and native food mortars (Taylor,1960). Milicia spp. are not Odum suffers a high level of normally distinguished by the exploitation. Alder (1989) esti• timber trade, as the timbers are mated that if the current rate of very similar (Hawthorne, 1990). exploitation of odum continues, the species will cease to be of Odum is found throughout the high commercial importance by the end of forest zone, in the derived• the 20th century. The high rate of woodland, and in the riverine / exploitation coupled with poor riparian forest of savannah wood• natural regeneration makes the land which is in the Antiaris• futuJ:e of odum bleak. A lot more Chlorophora Association. It prefers work, therefore, needs to be done well drained soils and cannot stand to save this valuable species from impeded drainage (Taylor, 1960; possible commercial extinction. Hall & Swaine,1981). The natural range of odum stretches from the Lack of information on the island of Zanzibar off the coast of reproductive biology of odum is one eastern Africa to the Gambia in the major problem faced in the attempt west (White,1966). Efforts to to genetically improve odum. The establish odum plrtntations have reproductive biology of a species proved futile, mo. _:I because of controls the length of its breeding the severe attack f young shoots, cycle. It is not known exactly when especially on the leaves and buds odum starts flowering but it is by an adult psyl .d fly, Phytolyma speculated that it is after 25 to

'present address - Moi University, P.O.Box 3900, Kenya

Ghana Journal of Forestry Vol.' '994 19 30 years of age (Britwum, pers. necessary. The cq.tkins were "com. ). Odum' s breeding system is enclosed in paper bags. Four allogamous, thus it has hetero• buckets were placed in the labo• genous populations of heterozygous ratory, two with a fan next to the plants, giving the tree breeder bags, two without a fan. The other more genetic material to ·two buckets were placed in a screen manipulate. A species breeding house. system dictates the breeding procedure that can be used and the Seed production extent to which recombination .can be manipulated. Knowledge of Nine trees were selected in the flowering age and breeding systems 1994 season for the study of seed ar~ especially important before production at the Afram Headwaters seed orchard establishment. This Forest reserve. The trees were paper is based on a study of selected on the basis of one or flowering and fruiting patterns in more of the following odum. characteristics:

STUDY AREA AND METHODOLOGY The performance of progeny in a progeny trial. Location of specimen Trees Tree form. The study was carried out mainly on 35 odum trees growing in the wild Forking characteristics. at Afram Headwaters Forest Reserve which is about 30km north of Kumasi Heights were measured from ground town in the dry semi-deciduous level to the point of forking or to forest zone. Also utilised in the the point of the lowest branch. study were nine 20 year ol~ trees Diameter was measured at breast growing at Mesewam research nursery height (DBH), whereas form was about 15km South of Kumasi in the classified as either straight or moist semi-deciduous forest zone. crooked. The following grading procedure was used to quantify Flowering and Fruiting fruiting:

The studu was dare during the o - No fruiting 1992/ 1')~ an-:> ':. ['S .993/199,1 flower.L',/ s~'~',so~.;j. "':'o'"er development and structure, fruit formation and morphology and seed 2 - Sparse fruiting production were among the factors that were studied. 3 - Moderately prolific

Girdling 4 - Prolific

Girdling to induce flowering was 5 - Very prolific tried on the 20 year old trees growing at Mesewam research Fifty strobili were collected from nursery, in December 1992/Janu~ry these trees and the number of seeds 1993. Position and size of girdles per strobilum per tree estimated. were varied, with girdles of 1.28, 2.56, 3.84 and 5.12cm made on primary and secondary branches. RESULTS AND DISCUSSION

Pollen Handling Botany and Phenology of the tree

Several techniques were used to DistLnguishing characteristics for determine the best method of pollen the two species are summarised in extraction. Flower bearing branches table 1. The leaves have lateral were collected and brought to the nerves, the number ranging from 7 laboratory. These were cut and to 18 depending on species (Table placed in buckets containing ~ater, 1). The ~~ves are simD~9 and In all there were s'x ~uckets with a.ltErnate ,..~-t,.i th·~'· 'lC. L~;'!.,Les each bucket 8ontdLn~ng five caducous~ Leaves or ~~ --,':~l~!1gs, branches. Water was cnanged daily ;lnl:L~e thGse of the '1'acur:2c tree and the branch butts were cut when have serrated edges. Of the 35

Ghana Journal of Forestry Vol.1 1994 20 Table 1. Distinguishing characteristics between M. excelsa and M. regia tree~

Characteristic M. excelsa M. regia

Crown smaller, flat top, larger, rounded, lighter green dark green

Bark dark with white dark, reddish mottles,--rough brown and smooth in texture

Leaf -interno~e short (1-4.Scm) long (1.S-7cm)

shape leaf blade narrow wide and rounder and long length:width 1.7:1 1.4:1

petiole long (3-Scm) short (1-3cm)

texture soft and glossy hard and dull

colour light green dark green lateral nerves 11-22 7-11

arrangement alternate alternate clustered

Flowers borne in clusters borne singly

Table 2. Distinguishing characteristics between male and female odum trees

Characteristic Male Female

Tree trunk long and slender thicker

Crown small, narrow, wide, big and more compact, scattered, lighter in colour darker in colour

Forking less frequent almost in all cases

Flowering more profuse and less profuse and o~rqrs throughout occurs on the upper ..:rown,flowers and outer parts of rlier, over a the crown ~onger period and :ore frequently

Ghana Journal of Forestry Vol.1 1994 21 Figure 1. Branchlet of M. excelsa"female treei-

Gh8naJournal of forestry Vol.1 1994 Figure 2. Branchlet of M. regia female tree.

GhanaJournal of Forestry Vol.1 1994 23 Figure ~ ..Branchlet of M. excelsa male tree.

GhanaJournal of Forestry Vol. 1 1994 24 .. Catk.itlll

Figure 4. Branchlet of M. regia male tree.

25 GhanaJournal of Forestry Vol.1 1994 trees studied, 18 were H. excelsa with male trees starting around the and the rest ,H. regia. Odum is last week of December until early dieocious. Dioecy is quite Har~h. Very old flowers were found widespread among tropical forest under' four mal~ odum trees in trees (Kramer & Kozlowski, 1979). February indicating that some trees Among the 35 trees, 22 were found may start flowering in the early to be female while 13 male. This part of December. Female trees gives an approximate ratio of 2:1. begin flowering later (early However, when the individual January). Considerable variability species are considered there were exists among trees with some trees 12 female and 5 male H. excelsa starting to flower in early March. trees; and 7 female and 10 male H. regia trees. Generally, male flowers bloom and ripen several days before the There are distinct differences in females, therefore, pollen is in flower structure. The male catkins the air by-the time the females are are long while females are shorter receptive. However, long after with many protruding styles female trees have stopped flowering (Hawthorne, 1990). Apart from the (towards the end of the flowering obvious difference in flower season) flowering is still evident structure, other traits have been among a few male trees. It is observed that clearly distinguish typical of dioecious species to males from females. These differ• show profound differences in many eRces are summarised in table 2. reproductive traits (Opler & Brown, 1978) among which is the timing, Flower development duration and frequency of flowering (Kapoor-Vijay & White, 1992). Odum trees do not flower every year. Some of the trees studied In the males, flower,ing starts on that flowered in 1992/1993 did not the upper crown of the tree and flower in 1993/199~ (Table 3). later spreads to the, lower Likewise, some trees thpt flowered branches. More mature flowers are, in the 1993/1994 flowering season therefore, present in the upper failed to flower in the 1992/1994 crown while lower branches have season. This pattern, however, was younger flowers. In general, male observed only on female trees and flower~ on the same tree may be at not male trees. different stages of development. It was also observed that M. excelsa Odum has mixed buds containing both is a later bloomer than M. regia. flowers and leaves. The trees a,re completely leafless before bud Both male and female flowers take break. As soon as the buds open about four weeks to reach maturity. three leaves emerge with the most The flowers reach maturity after' proximal at a more developed stage. the first three leaves are fully On close examination, flowers can formed. The entire tree is leafy be seen immediately after foliage and green at flower maturation. flush but before leaf expansion. There are 3-4 cataphylls (bud Girdling scales) per bud and each has a flower at its base (leaf node). None of the trees flowered through Within the bud are also foliage this induction method. It is leaves which in almost all cases speculated that at 20 years of age, are five in number. Of these, only odum trees may still be too young the two proximal to the cataphylls• to flower. However, success of bear flowers at their bases. In girdling depends on the season, certain cases, flowers are found among other factors. only at the bases of the cataphylls and none at the foliage leafaxils. Flower Morphology' In a few cases, in addition to the former, there may be a flower at The early stages of; floral the axil of the oldest foliage leaf development resembles that of leaf only. In general, therefore, one formation. The flowers lack petals bud may produce 3-5 flowers. and are aggregated into caterpillar The trees 'begin flowering at -shaped catkins with male catkins / different times (non-synchrony)

Ghana Journal of Forestry Vol.1 1994 26 longer than their female counter• they withered away. The flowers parts (Hawthorne, 1990 ). withered too. This failure can be Both male apd female catkins are attributed to the lack of a green in colour with the females controlled environment in the having protruding styles. M. laboratory. excelsa catkins are light green in colour while those of M. regia are Fruit formation and Morphology of a darker shade. _The styles of the female M. excelsa are fine to It takes approximately five to six the touch while those of M. regia weeks from the time of fertiliza• are coarse. Offferences in lengths tion to fruit maturation. When the are apparent between flowers of the fruits are mature they abscise and two species. M. excelsa catkins are are yellowish in colour. The fruits generally longer and the female differ little from the female catkins can attain a length of 8 cm inflorescence from which they while M. regia female catkins grow develop. The fruit softens as it up to about 7 em. (Figures 1 & 2). ripens due to changes· in pectic M. excelsa male flowers attain a compounds and hydrolysis of starch. maximum length of 17 cm compared to Overripe fruits are very soft. 13 cm attained by M. regia male Premature abscission of fruits was flowers (Figs 3 & 4),. Taylor observed in almost all the trees, (1960) estimated the length of male but with differing degrees. This flowers to be 20cm while those of premature abscission may occur females about 2 cm. Male flowers in immediately after fruit set. general are longer and much slender However, compared to some tropical while female flowers are shorter species such as Wawa (Triplochiton and thicker. However, M. regia scleroxylon) it is relatively less female flowers tend to be thicker serious problem in Odum. The fruits than those of M. excelsa. In M. mainly fall underneath mother trees excelsa the catkins are borne in and are dispersed by birds, clusters of 2-3 (as a resulb of the mammals and insects. short internodes) but occur in singles in M. regia (long Seed production internode) . Of the nine trees studied, two Pollen handling trees (Nos. 3 and 8) which did not fruit in 1992/1993 seasons produced No pollen was shed by the flowers lrruits in 1993/1994 season. in the screen house, and trace amounts by the ones in the Similarly two trees (Nos.4 & 28) laboratory. When branches were which produced fruits in 1992 /1993 accidentally enclosed in a warm, season did not produce fruits in moist container, however, pollen 1993/1994, suggesting that not all shed was profuse. It was observed trees produce fruits in every that catkins should be brought in fruiting season. while still green and hard, a few days before maturation. The best Table 4 shows the mean number of stage would be when the flowers seeds extracted' perstrobilum have started shedding pollen at (n=50) per tree, and the estimated their bases. This requires day to number of seeds per kilogram of day monitoring of the flowers in odum fruit. the field. The catkins produce large amounts of pollen, therefore, The results suggest that in a good only a few branches are required to fruiting year an odum tree can be produce pollen for breeding quite prolific. This indicates that purposes. fequndity may not be a major consideration in choosing trees for There was not much success with the the establishment of a, seed female odum bran _s· that were orchard. brought to the ~aboratory. The young female buds developed with time, but at the onset of bud break

Ghana JournaL of Forestry VoL.1 1994 27 Table Performance3.2straight40351v.21.4-21.818.81.924.93.925.53.82.52.228.53.330.615.21.140no3good*5no--noForm 30 good Forkingstraight93/94HeightProgeny(m)92/93DBH(m)M~rphologicalgoodnoyes Characteristics and fruit production of nine trees at No. Tree • Afram Headwaters Forest Reserve. 2823*35122948No data available 3 Fruiting intensity

Table 4. Estimated number of seeds per kilogram among six odum trees.

Tree per(g)strobilum0.20.4Weight0.3100 seeds4107512018817Mean73of no.333,333402,778500,000250,000no./kg403,778Estimatedof seedsof fruit No. Mean3035231283

CONCLUSIONS

Knowledge of reproductive biology controlled crossing is to be used is a prerequisite for any meaning • as a breeding technique. In this ful breeding programme. Tree paper variation has been recognized species vary in size, shape and between sexes, 'species and among location of their flowering buds, individual trees. size and structure of male and female flowers, duration of pollen ?he observational studies reported shed and fruit formation among in this paper gives the base for other trait's. Knowledge of such further work on the reproductive traits are useful especial+y if biology of odum. Continuing

GINnt Journal of Forestry Vol.1 1994 28 research on tree improvement of knowledge of seed production is most species require development of required. For this reason, pollen handling techniques. It is, information such as number of seeds therefore, important that pollen per catkin, number of catkins extraction, storage and testing produced per tree and thus total methods be developed. For breeding number of seeds produced per tree and afforestation purposes, and seed viabIlity is required.

ACKNOWLEDGEMENT - We are grateful to the Government of Ghana and the Internatiq,nal Tropical Timber Organisation (ITTO) through Project PO 75/90 for funding this study.

REFERENCES

Alder, o. (1980" Natural Forest Diversity and Genetic resources. increment, growth, and yield. In The Commonwealth Science Council. Proc. of For.Proj. Sem. 29th-30th Commonwealth Secretariat, London. March, Accra, Ghana. p47-52 248pp.

Ayensu, E.S. & Bentum, A.L.K. Kramer, P.J. & Kozlowski, T.T. (1974) Commercial Timbers of West 1979. Physiology of woody plants. Africa. smithsonian contributors Academic Press, Inc. N.Y. 811pp of Botany No. 14. Smithsonian Institute Press, Washington D.C. Opler, P.A. & Brown K.S. (1978) Sex ratios in tropical forest Hall, J.B. & Swaine, M.D. (1981) trees. Evolution 32:812-821. Geobotany: Distribution and ecology of vascular plants in a Taylor, J.C. (1960) Synecology and tropical rainforest. Forest Silviculture in Ghana. Thomas Vegetation in Ghana. W.Junk Nelson and Sons Ltd. 418pp. Publishers. The Hague. 383pp. Wagner; M.R.; Atuahene, S.K.N. & Hawthorne, W.(1990) Field guide Cobbinah, J.R. (1991) Forest to forest trees of Ghana. Chatham: Entomology in West Tropical Natural Resources Institute for Africa: Forest Insects of Ghana. the Overseas Development Kluwer Academic Publishers, Administration, London. Ghana Dordrecht, 210pp. Forestry Series 1, vit 278pp. White M.G. (1966) The problem of Kapoor-Vijay, P. & White, J. the Phytolyma gall bug in the (1992) Conservation Biology. A establishment of Chlorophora. Training Manual for Biological Inst. Pap.Comm. For. Inst. No.37 Oxford, London. 52pp.

Ghana Journal of Forestry Vol.1. 1994 2.9