Vegetative Propagation of Lovoa Trichilioides: Effect Provenancef O S , Substrate, Auxind an S Leaf Area

Journal of Tropical Forest Science 13 (1): 116-129 (2001)

VEGETATIVE PROPAGATION OF LOVOA TRICHILIOIDES: EFFECT PROVENANCEF O S , SUBSTRATE, AUXIND AN S LEAF AREA

Z. Tchoundjeu

International Centre for Research in Agroforestry (ICRtVF), BP 2123, Yaounde, Cameroon; e-mail: [email protected]

R. R. B. Leakey

Centre for Ecology and Hydrology (CEH), Bush Estate, Penicuik, Midlothian, EH26 OQB, Scotland, United Kingdom; e-mail: [email protected]

Received October 1998______

TCHOUNDJEU, Z. & LEAKEY, R. R. B. 2001. Vegetative propagation of Lovoa trichilioides: effects of provenance, substrate, auxins and leaf area. Lovoa trichilioides (African walnut, dibetou or bibolo) is an important commercial timber species indigenou Weso st Centra d tan l Africa. However destructiolace e ,seedth th f k o d san n of young seedlings by shoot-borers have hampered large-scale regeneration of this species. As vegetative propagation is an effective means of multiplying selected trees within a tree improvement programme, this study focused on the main factors affecting rootine th g abilit f leafyo y stem cuttings. When testing different rooting substrates, single-node, leafy stem cuttings rooted significantly better in coarse gravel than in a forest topsoil-gravel mixture. The application of indole-3-butyric acid (IBA) had no clear effect on mean rooting perentage, although three out of four treatments with differing concentrations of IBA gave final results that were significantly better than controle th . However, 50u considerege b IB appropriatn n Aa ca e b o dt e concentration to promote their rooting. Rooting ability was also affected by the node position. Cuttings fro e apicamth l nodes rooted significantly better than those from basal nodes of the same stem. Cuttings with large leaf areas (50-200 cm2) rooted better than those with smaller trimmed leaves (60r thos week9 fo %n i e% wits37 versu d h an s0 0 and 25 cm2 respectively). The optimum of 200 cm2 is higher than that for other tropical tree species suc s Triplochitonha scleroxylon Khayad an ivorensis.

Key words: Rootin gster- n cutting s- non-mis t propagato medi- r a

TCHOUNDJEU LEAKEY& ,Z. , R.R.B. 2001. Pembiakan tampang Lovoa trichilioides: kesan provenans, substrat, auksin dan luas kawasan daun. Lovoa trichilioides (walnut Afrika, dibetou atau bibolo) ialah spesies balak komersial asli yang penting bagi Afrika Tengah dan Barat. Bagaimanapun, kekurangan biji benih dan pemusnahan anak- anak benih muda oleh penggerek pucuk telah menghalang pemulihan besar-besaran bagi spesies ini. Memandangkan pembiakan tampang merupakan cara yang berkesan untukmembiakkan pokok yang terpilih dalam rancangan pembaikan pokok, kajian ini

116 Journal of Tropical Forest Science (1): 116-129 (2001) 117

memfokuskankepadafaktor-faktorutamayangmempengaruhikeupayaan pengakaran bagi keratan batang berdaun. Apabila menguji substrat pengakaran yang berbeza, buku tunggal, keratan batang berdaun mengakar dengan lebih baik dengan bererti id dalam kerikil besar berbanding dengan di hutan campuran tanah atas dan kerikil kasar. Penggunaan asid indola-3-butirik (IBA) tidak mempunyai kesan yang jelas terhadap min peratus pengakaran, walaupun tiga daripada empat rawatan dengan kepekatan IBA yang berbeza memberikan keputusan akhir yang lebih baik dengan bererti berbanding dengan kawalan. Bagaimanapun, 50 ug IBA bolehlah dianggap sebagai kepekatan yang sesuai untuk menggalakkan pengakarannya. keupayaan pengakaran juga dipengaruhi oleh kedudukan buku. Keratan daripada buku apeks mengakar dengan lebih baik dengan bererti daripada keratan dari pangkal buku daripada batang yang sama. Keratan dengan luas daun yang besar (50-200 cm2) mengakar lebih baik daripada keratan dengan daun yang dikerat lebih kecil (60% dalam mas mingga9 u berbandin bag% g i37 denga keratan da n0 n masing-masing cm5 2 denga 2)n . da Optimu n0 m sebanya2 adalacm 0 h k20 lebi h tinggi daripada optimum untuk spesies pokok tropika yang lain seperti Triplochiton scleroxylon dan Khaya ivorensis.

Introduction

Lovoa trichilioides Harms. (African walnut, dibetou or bibolo) is an important commercial hardwood timber tree of the Meliaceae found throughout the humid forest zone of Cameroon. Its natural distribution extends across the evergreen forest from Sierra Leon Wesn ei t Afric Angolo t a Centran ai l Africa (Figur. e1) Lovoa trichilioides is a large or emergent forest tree, often reaching a height of r moreo m , 5 wit4 diametea h t breasa r t height exceedine bolTh s ei . m g1 sometimes straight and cylindrical to a height of 25 m where the first branches occur. More commonly, however bole ,th crooke s ei heavild dan y forked illustrating hige th h potential therimprovemenr fo s ei thif to s species crowe .Th n tende b o st heavily branched and fairly open, becoming dome-like on emergent trees. There is often abundant local regeneration in the high forest and adjacent low bush. The seedlings can tolerate shade for a long period and, when the overhead canopy is opened by a falling tree, they start to grow and form a long slender stem that terminate smala n si l crown with large leaves. Despit abundance eth seedf eo , many factors hamper regeneration programmes with this species: • Seed collectio s difficuli n t because mast years occu, t y intervala r 4 3- f o s although some flowering trees can be found in most years. • Seed production is spread over a few weeks and the winged seeds can be carried far fro mothee mth r tree. • Insect smald san l mammal seede s th fee sn do reducin numbere gth s than ca t germinate. • Although fresh seed initially ma s y hav a gooe d germination rate, viability decreases rapidly. Consequently, wildings are usually collected from the natural fores foresr tfo t regeneration programmes, although their growt nursere th n hi y may be quite slow. 118 Journal of Tropical Forest Science (1)3 1 : 116-129 (2001)

Like other members of the Meliaceae, L. trichilioides is attacked by the mahogany shoot-borer larvae mote th th , f heo Hypsipyla robusta (FAO 1958) e hatchinTh . g larvae penetrate the young stem tissues, killing the shoot apex and causing considerable damag ofted deate ean nth youn f ho g trees, hence discouraging large-scale planting.

Figure naturaTh . e1 l distributio Lovoaf no trichilioides Wesn i Centra d an t l Africa (shaded area) (from Vivie Faurn& e 1985)

Vegetative propagatio consideres ni d becaus above th f eeo constraint largeo st - scale regeneration fro opportunite mth seed dan seleco yt masd an t s propagate superior tree clonar sfo l forestry (Leakey 1991). Additionally, clonal propagation s beeha n recommende approacn a s da circumveno ht e Hypsipylath t problem (Newto 1994). al l ne , becaus othen ei r member Meliaceaee th f so , genetic resistance has been identified at low frequency in natural populations (Newton et al. 1999). If amenable to vegetative propagation, such resistant material would be appropriate for forest regeneration and agroforestry (Matos et al in press). This paper reviews the main factors affecting the rooting rates of juvenile material of L. trichilioides inon-misna t propagator.

Materials and methods

The vegetative material used in the various experiments of this study was produced from young seedlings of six provenances collected in natural stands in Cameroon. The provenances originated from areas of Cameroon within the natural range of . trichilioidesL with different climates (Tabl. e1) Journal of Tropical Forest Scier,

Tabl . Some1 e characteristic differenf so t zone origif so Lovoaf no trichilioides provenances used to establish juvenile stockplants of seedling origin

Numbef ro Mean Mean Mean Main Main plants of each Altitude rainfall temperature humidity wet dry Provenance provenance (m) Latitude Longitude (mm) season season OO

Mbalmayo 174 650 3° 05' N 11°00'E 1640 23 77 Sept— Nov . Dec-Feb

Kumba 57 236 N ' 32 ° 4 9° 19' E 2405 29 86 Mar-Oct Nov-Mar §

Makak 34 750 N ' 30 ° 3 11°20'E 1687 23 80 Sept-Oct Dec-Feb o0 Loum 54 150 4° 50' N E ' 05 ° 9 2500 26 83 June-Oct Mar-May

Ottotomo 210 802 3° 36' N 11°19'E 1750 24 80 Sept— Nov Dec-Feb

Edea 129 52 3° 46' N E ' 04 ° 10 2750 26 82 Sept-Oct Dec-Jan 120 Journal of Tropical Forest Science 13 (1): 116-129 (2001)

Production of cuttings from seedlings

Prior to their transfer to the nursery of the Office National de Developpement des Forets (ONADEF), wildings of each provenance were collected from the forest and grown for 9-12 months. The plants were then cut back to a height of 10 cm to minimise the physiological stress of transit and transferred to the experimental site in Mbalmayo. Six hundred stockplants of six provenances of L. trichilioides (100 per provenance) were plante nursere th . Prunin n m di row n 0 yi 1. sg X space 5 0. t da subsequently stimulated sprouting and the production of shoots that formed an appropriate source of future 'juvenile' cutting material. Fifty grams of a compound fertiliser (NPK 20:10:10) were applie eaco dt h plan accelerato t t e their growth (application was repeated after every three months over a year).

Preparatio cuttingf no s

Single-node cuttings were harvested sequentially down the current flush of the shooto tw p s to from managed stockplants, takin e maximugth m cutting length allowe lengte internodee th th y f db ho . Unless; otherwise stated, leave thesn so e cuttings were trimme abouo dt cm0 t5 . l Auxi0.2u 0 %n(1 indole-3-butyric acid, 2 IBA), dissolved in 100% ethanol, was applied to the cutting base with a micrometer syringe. The alcohol was evaporated off using a fan, before the cuttings were set in non-mist propagators describe Leakey db l (1990)a t ye . Cutting length was assessed while setting up the experiment. Rooting ability was assessed once a week from the fourth week through to the twelfth. This involved recording the number of rooted and dead cuttings and the state of flushing of the axillar ycuttingse budth n so . Newly-rooted cutting morr so wite e hon root s longer than 1 cm were potted in black polyethylene bags containing a 50:50 mixture of forest topsoil with rotted sawdust. Unrooted cuttings, or those with roots shorter wer, thacm en1 returne propagatioe th o dt n bed. Potted cuttings were hardened ofnon-misn fi t propagators unde rweek 3 shad2- r sefo before being transferreo dt opee th n nursery.

Effects of different media on rooting ability of L. trichilioides stem cuttings under non-mist propagator (Experiment1)

differeno Tw t rooting media were tested: coarse gravel (CG: 3-1diameterm 0m ) 50:5a d 0an mixtur coarsf eo e grave foresd lan t nopsoil (FT) coarse .Th e graves lwa obtaine sieviny db g river saind. After sieving gravee washeth ,s lwa removo dt e weed seeds, organic matter , cutting0 etc tota30 A . f o cutting3 l( s r plantpe s ) from managed seedling stockplants of 2 provenances (Edea and Kumba) were set under a non-mist propagator using a 2-factorial design with 2 levels of substrate (CG and FT) by 2 provenances (n= 150). There were 25 cuttings in each treatment replicated 3 times (n=75). Journal o1 f Tropical12 Forest Science 13(1):116-129 (2001)

Effects of different concentrations rootingthe trichilioidesL. on of the of lBA stem cuttings (Experiment 2)

Cuttings were collected from five clones (original seedling stockpland an t rooted ramets fro originatin) mit g from Mbalmayo: (MB4, MB5, MB9, MB1d 1an MB50). Five concentrations of IBA were chosen for the experiment: 0,50,100,150 Thirt. ul 0 y1 cuttingn i g u 0 s25 wer d an e use eacdn i h treatmen eacd tan h treatment was replicated three times. In all, 450 cuttings were set under non-mist propagator randomisea n si d block design rootine .Th g substrat 50:5a s e0wa mixturf eo coarse gravel and rotted sawdust.

Effects of different lamina areas on rooting of L. trichilioides stem cuttings (Experiment3)

Five different lamina areas (0, 25, 50, 100 and 200 cm ) were tested in this 2 experiment. Each treatment had 30 cuttings and was replicated 3 times. Thus, 450 cuttings were treated with 50 ug IBA before being inserted in randomised block design unde non-misra t propagator containin g50:5a 0 mixtur coarsf eo e grave rotted an l d sawdust. Statistical analysis

AN OVA or t-test, as appropriate, was done using the Statview 512 (Abacus Concepts, Inc., 24009 Venture Blvd., Calabasas, California, USA) computer package and unless otherwise stated, statistical significance is given at the 5% level (p<0.05). Standard error r percentagefo s f rooteo s r deao d d cuttings were calculated usin proceduree gth f Snedecoo s Cochrand ran e (1989 datr fo ) a with binomial distribution.

Results and discussion

Effects of different rooting media on the rooting of L. trichilioides stem cuttings. (Experiment1)

Rooting started between weeks 3 and 4, and was faster in pure coarse gravel (CG) than in the 50:50 mix with forest topsoil (FT). This difference persisted and at week percentage 7th , cuttingf eo s roote becamG C dn i e significantly greater (p=0.001) than in FT. In addition, there were a greater number of roots per rooted cutting in CG than in the FT. Negative relationships were found between % cuttings rooted numberd an rootf s o rooter spe d cuttin (y=303.14G C n gi - 31.36x; r=0.95T F d )an (y=l 17.53-13.23x 0.93)= ;r . Cutting doublmortalits wa T morr eF o yn i eG thaC n ti in cuttings from two provenances (Table 2). 122 Journal of Tropical Forest Science 13(1):116-129 (2001)

Table 2. Effects of substrate (CG = coarse gravel, FT = forest topsoil) and provenance on the percentages of rooted and dead cuttings of L., trichilioides.

Edea Rumba Overall

Media CG FT CG FT CG FT Rooting % 7815.4 45 ± 6.8 61 ±8.5 50 + 8.3 72 ± 4.6 46 ± 5.0 Mortality % 15 ±4.8 66. ± 7 4 24 ±7.8 47. ± 3 3 0 4. ± 0 2 40 ±5. 2

In CG, cuttings of Edea provenance rooted better than those from Kumba. In significano n therT s F ewa t difference betweeprovenanceso tw e nth . Although significano n ther s wa e t differenc survivan ei l rates betweeprovenanceo tw e nth s mortalite th , CG cuttingf yo e inth s from Ede highes (TablT awa F Ther. n ri e2) e was, however, no significant difference between the two media for survival rates of cuttings from Kumba. In recent years there have been a number of comparative studies between different rooting medi r tropicaafo l tree species (see e Leaketh . 199 d al t 0yan e references therein). Whilbees ha nt ei clear that specie rooo sd t bette certain ri n media, it is not possible to determine very clear trends. Sawdust was the preferred mediu r Miliciamfo excelsa (Ofor. 1996)al t e i . Irvingia gabonensis (Shiemb. al t oe 1996a), Gnetum africanum (Shiemb . 1996bal t d Ricinodendrone o an ) heudelotii (Shiemb . 1997) al whicf t o e l lowlane ,al h ar d moist forest specie Wesf so t Africa, while the central American species Cordia alliodora (Mesen et al. 1997) and Albizia guachapele (Mesen 1993) rooted bes graveln i t e formeTh . moisa s ri t forest species of the lower slopes of the Cordillera, while the latter is from the dry forests of the Pacific coast. Shorea leprosula, a hillside species of Southeast Asia, did not root particularly well in either sand or coconut fibre (Aminah et al 1995). These different media have differing air:water ratios (Leake . 1990al t ye , Mesen 1993). From this limited evidenc seemt ei s unlikely that species adapte conditiont we o dt s root bette medin ri a wit higha h water holding capacity perhaps i t I . s relevant that tropical tree species var theiyn i r foliar relative water content, wit zony hdr e species having lower values than humid zone species (Newto Jonen& s 1993). However, littl knows ei n abou processee tth s involved likels i t I .y that ther interaction a s ei n betwee watee nth r relation mediue th f scuttingsthoso e d mth an f e o exampler .Fo , different rooting media are known to affect water uptake by cuttings (Grange & Loach 1983) and to have effects on photosynthesis and stomatal conductance of cuttings (Mesen et al. 1997). The relationships between % rooting and numbers of root r rootepe s d cutting suggest tha n thii t s study there were more water resources availabl rootinr efo coars e cuttingn gi th n i e t gravesse l medium.

Effects of different LBA concentrations on the rooting of L. trichilioides stem cuttings (Experiment 2)

Percentage rootin cuttinggn i s treate r cuttind witpe A hg IB 0,50,100 g u 0 20 r ,o wat significantlno s y different. Surprisingly, those treated withr 150upe A gIB cutting were lower than the others, apparently due to higher mortality (26 ± 6.2%). Journal of Tropical Forest Science 13(1):116-129 (2001) 123

However generae th , rootinn o A lfurtheeffec n s laca IB g f f wa ko o t r indicatey db lace f significanth ko t differences between treatment e numbeth n si f rooto r s formed per cutting. There were, however, significant differences between rootin d cuttingan g mortality between nodes within the same shoot. Significant differences were found in the rooting percentages of cuttings from apical and basal nodes (Figure 2), cutting mortality (Figure 3) and the numbers of roots per cutting (4±0.62 vs. 2 ±0.29 for node 1 and node 5 respectively). Cuttings from basal nodes were the most susceptible to mortality and they had the greatest proportion that remained unrooted.

9 8 7 6 5 4 10 11 12 Weeks after taking cuttings

Figur . Effecte2 nodf o s e positio rootine th n no g percentag . trichilioidesL f eo single-node leafy stem cuttings (mean acros treatmentsA sIB ) 124 Journal of Tropical Forest Science 13(1):116-129 (2001)

25 _* 1u & •§ 20

•8 aj -a 15 1 10

N2 N3 N4 N5 Node position

Figure 3. Effects of node position on the percentage mortality of L. trichilioides single-node leafy stem cuttings (mean across IBA treatments)

Ther somes ewa evidenc f clonaeo l variatio rootine th n n i f cutting go s from seedlings of Mbalmayo provenance, with those of four clones rooting better than MB5, the fifth clone (Figure 4). Auxins generally enhance rooting abilit cuttingn yi f mosso t plant species, including tropical trees (Leakey et al 1990, 1994). However, the results from this study show that L. trichilioides seems to be different from many other tropical trees rootint increaseds a , no s gwa , hastene r proliferatedo ranga y db auxif eo n concentrations. However, another experiment with higher auxin concentrations (Tchoundjeu 1989) has suggested that to enhance rooting capacity substantially higher concentration auxif so requiree nar thatd dlargen an ,o r leaved cuttings, beneficiale ar . Journal of Tropical Forest Science 13(1):116-129 (2001) 125

^40

MB4 MB5 MB9 MB11 MB50 Clone number

Figur . Clonae4 l variatio rootine th n ni g percentag . trichilioidesL f eo single-node leafy stem cuttings (mean across IBA treatments)

This study observed polarity between cuttings from different nodes in mortality and the lack of rooting. The cause of this mortality is not well known in L. trichilioides. However, in Triplochiton scleroxylon, a West African hardwood, there was some evidence that rooting was dependant on the volume of the cutting's stem portion and hence in the cutting's capacity to store current assimilates prior to the formation of roots and a new sink for them (Leakey & Mohammed 1985, Leakey et al. 1992). In L. trichilioides basal cuttings have been found to have short internodesand large diameters, but to be smaller in volume than the more apical nodes. Thus similarita e b thery yma e wit findinge hth scleroxylon,. T n si although differencee th s betwee speciee nth thein si r growth habicomplicaty ma t e eth relationship . scleroxylonT : fres ha e growth, whil . trichilioideseL grow recurreny sb t flushing. In contrast, basal cuttings of Khaya ivorensis, another member of the 126 Journal of Tropical Forest Science 13(1):116-129 (2001)

Meliaceae, rooted better than apical ones, but long cuttings only rooted better than short cuttings when the cuttings had a large leaf area (Tchoundjeu & Leakey 1996).

Effects of different lamina areas rootingon oftrichilioidesL. stem cuttings (Experiment3)

Leafless cuttings did not root and most (90%) had died by week 9. Cuttings with larger leaf areas (50-200 cm2) rooted significantly better than did cuttings with 25 cm , those with 200 cm rooting best. They also had the greatest mean 2 2 numbe rootf ro r rootespe d cutting (Figurpercentagee Th . e5) deaf so d cuttings decreased with the increasing leaf area, suggesting that cutting survival depended on current photosynthesis and that water stress was not a major problem.

801=

3456789 Weeks after taking cuttings

Figur . Effecte5 f lamino s rootine a areth n ao g percentag auxin-treatedf eo , single-node, leafy stem cutting . trichilioidesL f so Journal o7 f Tropical12 Forest Science 13(1):116-129 (2001)

Leaves of L. trichilioides differ from those of most of the other West African hardwoods that have been propagated vegetatively: T.scleroxylon, T. superba and Nauclea diderrichii (Martin & Quillet 1974, Howland 1975, Koyo 1983, Leakey & Coutts 1989). However, like K. ivorensis, L. trichilioides has compound leaves wit7 opposith3- e pair leafletsf so . Despite this similarity specieo tw e sth , appear requiro t edifferena t leaf are r successfuafo l rooting. Optimum rooting percent- . ivorensisK age f o s under intermittent mist were achieved when leaves were trimmed to 10-50 cm2 (Tchoundjeu & Leakey 1996). In contrast, L. trichilioides* cuttings with larger leaf areas (50-200 cm2) rooted better in the conditions provide non-misa y db t propagator t cleano s ri whethet I . r n intrinsithia s i s c difference betwee speciese nth whether o , r water stres less ss wa marked undee rth latter system reducino s , g transpiration losses. Photosynthetically active larger leaves would produce more assimilates altering the optimum balance between the benefits of photosynthesis and the negative effects of transpiration. Dick and Dewar (1992) have developed a process-based model for rooting in stem cuttings, base principlen do s derived fro studiee . scleroxylon.mth T n so o T validate this model with data from tree species with apparently different responses to standard treatments, more physiological studies are needed. Research in the Meliaceae should seek to explain the apparent interactions between cutting volume, leaf area, cutting origin, assimilate storage, and the effects of stockplant environment, especially ligh nutrientsd tan specien i , s with compound leaved san growt recurreny hb t flushing. This would have important implicatione th r sfo domesticatio f mahoganno related yan d species (Newto . 1994)al t r whicne fo , h vegetative propagatio needes ni tako dt e advantag incidence th w f lo eo a f eo leve naturallf o l y occurring genetic resistanc Hypsipylao et shoot borers (Newton et al. 1999).

Acknowledgements

We wis thano ht Commissione kth Europeaf o n CommunitieK U e XIIth G d s)(D an Overseas Development Administration (now the Department for International Development r fundinfo ) g this research e alsW o. gratefully acknowledge th e assistance of ONADEF (Office National de Developpement des Forets) staff in Cameroon.

References

AMINAH, H. 1995. Vegetative Propagation of Shorea teprosula Miq. by Stem Cuttings, Ph.D. thesis, Universit Edinburghf yo . pp 3 25 . DICK, J. McP. & DEWAR, R. 1992. A mechanistic model of carbohydrate dynamics during adventitious root developmen leafn i t y cuttings. Annals of Botany 70:371-377. FAO. 1958. Shootborer Meliaceae th f so . Unasylva (1)2 1 : 30-31. GRANGE, R. I. & LOACH, K. 1983. The water economy of unrooted cuttings. Journal of Horticultural Science 58:9-17. 8 12 Journal of Tropical Forest Science (1)3 1 : 116-129 (2001)

HOWLAND.P. 1975. Vegetative propagation method Triplochitonr sfo scleroxylon Schum.. K . 110-12Pp 4 in Proceedings of the Symposium on Variation and Breeding of Triplochiton scleroxylon (K Schum). Ibadan, Nigeria. KOYO . 1983P e bouturag. J ,L . servicu ea s plantationede s amelioreee e Terminaliad s superban e Republique Populaire du Congo. Silvicultura 8:331-333. LEAKEY, R. R. B. 1991. Towards a strategy for clonal forestry: some guidelines based on experience with tropical trees . 27-4Pp . Jacksonn 2i . (Ed.E . ,J ) Tree Breeding Improvement.d an Royal Forestry Society of England, Wales and Northern Ireland, Tring, England. COUTTS& . LEAKEY. 1989B P . . dynamice R M ,. Th R , rootinf so Triplochitongn i scleroxylon cuttings: their relatio o leat n f area, node position weighy dr , t accumulation, leaf water potentiad an l carbohydrate composition. Tree Physiology 5:135-146. , DICKLEAKEYB. . . McP,J R . R ,NEWTON & . . 1992C . A , . Stockplant-derived variatio rootinn i g ability: e sourcth f physiologicaeo l youth . 171-17Pp . 8n Massi Production Technology r Geneticallyfo Improved Fast Growing Forest Tree Species. Volume . ! AFOCE L / IUFR O Conference, 14-18 September 1992, Bordeaux, France. , DICKB. LEAKEY . . McPJ ,R . R ,NEWTON .& . 1994C . A , . Captur f genetieo c variatio vegetativy nb e propagation: processes determining success . 72-8 Pp .Newton & . Leakey3n i . B . C . R . ,A R , (Eds.) Tropical Trees: the Potential for Domestication and the Rebuilding of Genetic Resources. HMSO, London, UK. LEAKEY, R. R. B., MESEN,J. F., TCHOUNDJEU, Z., LONGMAN, K. A., DICK.J. McP., NEWTON, A. C., MATIN, A., GRACE,J., MUNROMUTHOKA& . C . . 1990N R , . P , . Low-technology technique vegetative th r sfo e propagatio tropicaf no l trees. Commonwealth Forestry Review (3):247-2579 6 . MOHAMMED& . LEAKEY B 1985. . S R . . R R ,. EffectH , stef so m lengt roon ho t initiatio sequentian i l single- node cuttings of Triplochiton scleroxylon K. Schum. Journal of Horticultural Science 60:431-437. MARTIN QUILLET& . B , . 1974G , . Bouturag s arbreede s forestier Congou a s . Bois Foretl Tropiquese s de 155:15-33, 156:36-61, 157:21-40. MATOS; J. C. DES., FERNANCES, E. C. M., SOUSA, SAG., PERIN, R. & ARCOVERDE, M. F. Performance of big- leaf mahogan agroforestrn a yn i y syste westere th mn i n Amazon regio Brazilf no Proceedingsn .I of Big-leaf Mahogany: Ecology Managementd an Confermce. 22-24 October 1996, IITF Juann Sa , , Puerto Rico. (In press). MESEN.J . 1993F . . Vegetative Propagatio Centraf no l American Hardwoods, Ph.D. thesis, Universitf yo Edinburgh. 231 pp. MESEN,J. F., NEWTON, A. C. & LEAKEY, R. R. B. 1997. Vegetative propagation of Cordia alliodora (Ruiz & Pavon) Oken effecte :th IBf so A concentration, propagation mediu cuttind man g origin. Forest Ecology and Management 92:45-54. NEWTON, A. C. &JONES, A. C. 1993. The water status of leafy cuttings of four tropical tree species in mist and non-mist propagation systems. Journal of Horticultural Science 68:653-663. NEWTON, A. C., LEAKEY, R. R. B., POWELL, W., CHALMERS, K., WAUGH, R., TCHOUNDJEU, Z., MATHIAS, P. J., ALDERSON , MESEN,JG. . P , , BAKER F. .RAMNARINE & . P , . 1994S , . Domesticatio f mahoganieso n . Pp. 256-266 in Leakey, R. R. B. & Newton, A.C. (Eds.) Tropical Trees: The Potential for Domestication and the Rebuilding of Genetic Resources. HMSO, London, UK. NEWTON , WATT , LOPEZC. D. . . ,A . ,A CORNELIUS.J F , COREA MESE, & . 1999P. . F A . . NE ,,J . Genetic variation in host susceptibility to attack by the mahogany shoot borer, Hypsipyla grandella (Zeller). Agricultural and Forest Entomology 1:11-18. OFORI, D. A., NEWTON, A. C., LEAKEY, R. R. B. & GRACE,J. 1996. Vegetative propagation of Milicia excelsa by leafy stem cuttings: effects of auxin concentration, leaf area and rooting medium. Forest Ecology and Management 84: 39-48. SHIEMBO, P. N., NEWTON, A. C. & LEAKEY, R. R. B. 1996a. Vegetative propagation of Irvingia gabonensis, a West African fruit tree. Forest Ecology and Management 87:185-192. SHIEMBO, P. N., NEWTON, A. C. & LEAKEY, R. R. B. 1996b. Vegetative propagation of Gnetum africanum Welw. leafa , y vegetabl Wesr efo t Africa, foumal of Horticultural Science 71:149-155. SHIEMBO, P. N., NEWTON, A. C. & LEAKEY, R. R. B. 1997. Vegetative propagation of Ridnodendron heudelotii, a West African fruit tree. Journal of Tropical Forest Science 9:514-525. Journal of Tropical Forest Science 13 (1): 116-129 (2001) 129

SNEDECOR COCHRAN& . . 1989W G . . G , W , . Statistical Methods. editionh 7t e IowTh . a State University Press, Iowa, USA. 503 pp. TCHOUDJEU . 1989Z , . Vegetative Propagatio Tropicae th f no l Hardwoods Khaya ivorensisA. Chevd .an Lovoa trichilioides Harms. Ph.D. thesis, Universit f Edinburghyo . pp 0 27 . TCHOUNDJEU, Z. & LEAKEY, R. R. B. 1996. Vegetative propagation of African mahogany: effects of auxin, node position, leaf area and cutting length. New Forests 11:125-136. VIVIEN FAUR& . J ; E ;J.J. 1985. Arbres des Forets Denses d'Afrique Centrale. Agenc e Cooperatiod n Culturelle et Technique, Paris, France. 565 pp.