Forest Ecology and Management 140 (2001) 277±287

Mycorrhizal associations in the rain forest of South Cameroon N.A. Onguenea, T.W. Kuyperb,* aTropenbos Cameroon Programme, P.O. Box 219, Kribi, Cameroon bSubdepartment of Soil Science and Plant Nutrition, Wageningen University, P.O. Box 8005, 6700 EC Wageningen, The Netherlands Received 30 August 1999; received in revised form 19 January 2000; accepted 20 January 2000

Abstract

Mycorrhizal associations of important tree species were investigated within the research area of the Tropenbos Cameroon Programme (TCP), situated on the western portion of the Atlantic Biafrean forest of south Cameroon. Ninety-seven tree species of economic, social, and ecological importance, and three lianas were selected in three sites that differed in altitude, soil clay content, and soil pH. In each site plots were laid out in undisturbed forest. In each plot, seedlings, saplings, juvenile, and mature trees were identi®ed to species level and counted; girth at breast height measured and basal area calculated; root samples were taken and examined for mycorrhizal type and extent of mycorrhizal colonization. All 100 species investigated were mycorrhizal. Tree of 74 species formed exclusively arbuscular mycorrhizas (AM); 23 trees and three Gnetum species formed ectomycorrhizas (ECM). Five of these ECM plants also harbored AM structures. Extent of mycorrhizal root colonization showed large differences for various AM trees; however, colonization of more than half of these trees was less than 25%. Colonization of ECM trees was often higher than 75%. The contribution of ECM trees to basal area varied between 19 and 35%. ECM trees often occurred in small to large clumps. In sustainable forest management plans, existing ECM forest clumps should be given special conservation value. # 2001 Elsevier Science B.V. All rights reserved.

Keywords: Ectomycorrhiza; Arbuscular mycorrhiza; Rain forest; Clumping

1. Introduction economic and social interests of all forest dwellers as well as understanding of processes regulating the The rampant deforestation in the tropics due to functioning of tropical rain forests. While the volumes increasing shifting cultivation and logging practices, of logs extracted and the kinds of non-timber forest urges for the design of appropriate management products are well-known, little information exists on schemes to safeguard remnants of climax tropical rain factors that sustain growth of trees, determine natural forests. Effective strategies and solutions to sustain- forest regeneration, and maintain ¯oristic biodiversity. able forest management require taking into account Mycorrhizas (mutualistic associations between spe- cialized basidio-, asco-, and zygomycetous fungi and roots of most higher plants) constitute the most * Corresponding author. Tel.: ‡31-317-482352; fax: ‡31-317-483766. ef®cient nutrient uptake facilitators, particularly in E-mail address: [email protected] (T.W. Kuyper). nutrient-de®cient soils of tropical regions. As a

0378-1127/01/$ ± see front matter # 2001 Elsevier Science B.V. All rights reserved. PII: S 0378-1127(00)00322-4 278 N.A. Onguene, T.W. Kuyper / Forest Ecology and Management 140 (2001) 277±287 consequence, presence of mycorrhizas results in programme directed towards sustainable management increased plant ®tness and forest tree productivity and use of these forests. (Smith and Read, 1997). In the humid tropics, two major types of mycor- 2. Materials and methods rhizal associations of trees have been reported, viz. ectomycorrhiza (ECM) and arbuscular mycorrhiza 2.1. Research area and sites (AM). In general, AM dominates secondary forests and a large number of primary forests (Janos, 1980, The study was carried out within the TCP research 1996). The ECM occurs either isolated in a mosaic of area, which is situated in the western portion of the AM species or in clumps in undisturbed forest where Atlantic Biafrean forest of south Cameroon (Letouzey, they dominate the canopy (Alexander, 1989a; Newb- 1985), lying within the Congo±Guinea refuge. The ery et al., 1988, 1997). In southeast Asia, the major TCP area covers about 2000 km2 and is situated ECM taxa of rain forests belong to Dipterocarpaceae, between the cities of Lolodorf (38140N, 108440E) in whereas in tropical Africa the major taxa belong to the north, Adjap-Essawo (38020N, 108520E) in the east, Caesalpiniaceae and Uapacaceae (Alexander, 1989a; Akom II (28480N, 108340E) in the south, and Bipindi Fassi and Moser, 1991). Few ECM tree species are (38040N, 108250E) in the west. The climate is humid found in the humid neotropics (Alexander and HoÈg- tropical with two distinct wet seasons (mid-August± berg, 1986; BeÂreau et al., 1997). In tropical areas, mid-November, mid-March±mid-May) and two dry ECM tree species are also encountered in introduced seasons. The rainfall decreases in an easterly direc- pines and eucalypts (Le Tacon et al., 1989). tion, with an annual mean of 2836 mm in Kribi to Within the Biafrean forest of tropical west Africa, 2096 mm in Lolodorf and 1719 mm in Ebolowa. caesalp tree species dominate and these species are Average monthly temperatures vary between 22.9 strongly aggregated (Letouzey, 1968). Clumping is and 27.58C (Olivry, 1986). very conspicuous for ECM caesalps, and species of The landscape gradually shifts from lowlands in the Uapaca (Uapacaceae) usually occur in the same southwest to rolling mountains in the central northern clumps. Newbery et al. (1997, 1998), following an portion to abrupt hills in the northeast. Elevation earlier suggestion by Connell and Lowman (1989), ranges from 50 m a.s.l. near Bipindi to 1057 m a.s.l. hypothesized that the ECM habit and clumping were in the Bingalanda mountain, near Nyangong. The causally related because these ECM caesalps might be substratum consists of Precambrian metamorphic superior competitors for nutrients. However, little is rocks and old volcanic intrusions (Franqueville, known about the distribution of these clumps. In the 1973). In the southwestern lowlands (50±350 m rain forest of Korup National Park in southwest a.s.l.), surface soils are sandy clay loam and moder- Cameroon, 22 caesalp tree species form small clumps ately acid; between 350 and 500 m a.s.l., surface soils (Newbery et al., 1988). Basal area of these ECM trees are highly clayey and strongly acid; above 500 m ranged from 3 to 28% in various transects. In indivi- a.s.l., soils are very highly clayey and very strongly dual 0.32 ha plots where three caesalp species acid (van Gemerden and Hazeu, 1999). (Microberlinia bisulcata, Tetraberlinia bifoliolata, Portions of the area have been recently logged by and T. moreliana) dominated, basal area even ranged the Houthandel Gebroeders Wijma and Zonen B.V. between 45 and 68% (Newbery et al., 1988, 1997). (GWZ) timber company in two concessions. Else- The present study was undertaken to examine the where in the forest, Bantu populations subsist on mycorrhizal status of commercially important tim- shifting cultivation with short fallows of Chromolaena bers, of tree species with signi®cance for local forest odorata (Asteraceae) and long bush fallows, and dwellers, and of ecologically important tree species of on cocoa cash crops, while Bagyeli people live in the tropical rain forest of south Cameroon. The inves- small forest settlements on gathering and hunting. tigation also aimed at determining the distribution and Within the TCP research area, three experimental sites proportion of major types of mycorrhizal associations were chosen in Ebimimbang, Ebom, and Nyangong. in undisturbed forest stands within the research area Location, rainfall data, and soil physico-chemical of the Tropenbos Cameroon Programme (TCP), a characteristics are presented in Table 1. N.A. Onguene, T.W. Kuyper / Forest Ecology and Management 140 (2001) 277±287 279

Table 1 Location, elevation, rainfall, soil types and characteristics of experimental sites

Locality Ebimimbang Ebom Nyangong

Location 38030N, 108280E38050N, 108410E28580N, 108450E Elevation (m a.s.l.) 100 440 550 Rainfall (mm)a 1707 2019 1780 Soil types Ultisols Ultisols, Oxisols Oxisols Clay (%)b 10±40 40±60 60±80 pH (water) 6.1 4.7 4.3 Carbon (%) 1.69 2.26 2.21 Nitrogen (%) 0.15 0.18 0.19

Available P in H2O(mg/ml soil) 0.01 0.005 0.002 a Annual means of rainfall collected from 1995 to 1998. b Data derived from van Gemerden and Hazeu (1999).

2.2. Selection of plant species Cameroon, for additional checks and others were identi®ed at the herbarium of Wageningen University, For the mycorrhizal screening, 97 tree species were The Netherlands. Nomenclature of trees follows: selected, belonging to several categories of trees. They AubreÂville (1970); Letouzey (1985) and Vivien and include 40 actually commercialised species and 27 Faure (1985); de Wildeman (1936) for Uapacaceae potentially exploitable timber species. Actually com- and Stevels (1990) for Gnetaceae. mercialised timbers were selected on the basis of the percentage of volumes extracted, as recorded at the 2.3. Plot setting and plant enumeration Kribi port during the ®scal year 1994±1995 (Anon., 1995). Potentially exploitable timbers include lesser- Permanent sampling plots were selected in undis- known tree species with commercial potential in the turbed forest stands in each site. Selection criteria of future according to national standards. One timber tree these stands included absence of recent man-made or is completely protected under Cameroonian law. Six- natural disturbance and presence of different size teen trees are socially important species that provide classes of selected tree species. In each site, 2 ha non-timber forest products such as fruits, leaves, bark, (200 mÂ100 m) plots were laid out in the undisturbed medicines, to local forest populations; species with forest. There were two plots per site, 5 km distant from cultural values are also included in this category (van one another. Seedlings, saplings, juvenile, and mature Dijk, 1994). Some of these might also be potentially trees were identi®ed to species level, counted and girth exploitable as timber trees. Thirteen early-succes- at breast height (GBH) determined. Basal area was sional trees that could build up mycorrhizal inoculum calculated from GBH. during forest regeneration are included as ecologically Seedlings were plants of a small size with 3±5 true important trees. Three local Gnetum species (Gneta- leaves and <1 m height. Saplings were trees of inter- ceae) were added to the list owing to their particular mediate size between seedlings and juvenile reprodu- growth form (liana) and economic position as a food cing trees with GBH between 10 and 49 cm. Juveniles crop. were young reproducing trees different from mature Identi®cation of tree species was done with the help trees with GBH between 50 and 99 cm. Mature trees of local assistants trained earlier by logging compa- had a GBH greater than 99 cm. nies to recognise native timber species. Bark, leaves, ¯owers, and fruits were used and these characters were 2.4. Root sampling and assessment of mycorrhizal compared with photographs in identi®cation manuals colonization (Letouzey and Mouranche, 1952; Letouzey, 1983; Thirakul, 1983). For some species herbarium material Fine roots of saplings, juvenile, and mature trees was taken to the national herbarium in YaoundeÂ, were collected after tracing larger roots from the stem 280 N.A. Onguene, T.W. Kuyper / Forest Ecology and Management 140 (2001) 277±287 collar of target trees. For seedlings, the whole plant tree were considered AM if the following internal and was uprooted with the surrounding soil. Collection of external structures were observed from samples: vesi- root samples was complicated by the abundance of cles, arbuscules, hyphal coils, intra- and intercellular intermingling roots of different species near the soil hyphae without septa, auxiliary bodies on soil hyphae, surface and by the dif®culty to ®nd the tree collar distributive and absorptive hyphae, with or without where primary roots are inserted. Triplicate root sam- external spores, attached to the roots. ples of each selected tree species were taken per age Estimates of ECM and AM colonization were class. Root sampling was conducted during the period categorized in ®ve classes of mycorrhizal root colo- of July±August 1996, January and July 1997. nization (MRCI): Class 1, 1±5%; Class 2, 6±25%; Ectomycorrhiza roots were easily recognisable as Class 3, 26±50%; Class 4, 51±75% and Class 5, 76± such, by the presence of obvious ECM root surface 100% (Kormanik and McGraw, 1982). features (swollen root tips) using a 10Â hand lens (Sight Savers and Bausch and Lomb, Rochester, NY). The reliability of scoring of ECM roots was con®rmed 3. Results in the lab by observing the Hartig net in transverse sections of ®ne roots. ECM roots were placed in a petri 3.1. Mycorrhizal status of selected plant species and dish containing tap water, gently rinsed to remove soil extent of mycorrhizal colonization and adhering organic particles, and then observed under a dissecting microscope at 6.4±40Â magni®ca- All 100 taxa, belonging to 77 genera and 29 tion. Fractional colonization (100% times the number families, were mycorrhizal (Table 2). Seventy-four of ECM root tips divided by the total number of root tree species formed only AM. Twenty-three trees tips) was estimated from a random subsample (con- and three Gnetum species formed ECM. Dual mycor- taining at least 100 root tips). rhizal associations were observed in root samples of Non ECM root samples and portions of ECM root Afzelia, Anthonotha, Uapaca, and Gnetum. AM struc- samples were ®rst rinsed three times with tap water to tures were consistently observed in roots of Afzelia.In remove the alcohol. After rinsing they were cleared the other dual mycorrhizal species, only few AM with 10% KOH for 24 h or over the weekend (Phillips structures occurred on root portions devoid of ECM and Hayman, 1970). Heavily pigmented root samples features. were bleached after immersion for 60 min in alkaline Internal AM structures were not observed in root

H2O2 solution at room temperature. Alkaline H2O2 samples of the following tree species: Diospyros spp, was prepared by adding 3 ml of NH4OH and 30 ml of Hexalobus crispi¯orus, and Picralima nitida. Their 10% H2O2 to 567 ml of tap water (Kormanik and mycorrhizal status could not be clearly delineated due McGraw, 1982). Roots were thoroughly rinsed in to very dense pigmentation. The roots of these taxa water to remove the H2O2 (Rajapakse and Miller, were violaceous black and only attached non-septate 1992). The roots were thereafter acidi®ed with 1% hyphae and the absence of swollen root tips allowed us HCl for 1 min before staining in a solution of acid to infer their AM status. Plagiosiphon and Cynometra fuchsin for 2±3 days. The acid fuchsin stain was possessed swollen root tips that on super®cial exam- prepared from 875 ml of lactic acid, 63 ml of glycerin, ination might be interpreted as ECM tips. 63 ml of tap water, and 0.15 g of acid fuchsin (Kor- Twenty three tree species had 1±5% roots colonized manik et al., 1980). Stained roots were destained for with AM structures and 42 had less than 25% AM 2±3 days in a lactic acid solution devoid of acid colonization. Only Pterocarpus soyauxii, P. mildbrae- fuchsin before observation under a dissecting micro- dii, Alchornea cordifolia, Anthocleista schweinfurthii, scope at 6.4±40Â magni®cation. Five to six 1 cm long and Trema orientalis had always more than 75% roots stained roots were randomly selected, placed on a colonised with AM. Extent of AM colonization was glass slide, and gently squashed under a cover glass to equally variable between early- and late-successional observe the anatomy of AM fungi under a compound tree species. In more than half of the samples inves- microscope at 100±400Â magni®cation and to esti- tigated internal hyphae only could be observed. mate fractional colonization. Root samples of a host Arbuscules and hyphal coils were about equally N.A. Onguene, T.W. Kuyper / Forest Ecology and Management 140 (2001) 277±287 281

Table 2 Mycorrhizal status and extent of mycorrhizal colonization of selected tree species of economic, social, and ecological importance, including Gnetum speciesa

Species Family Category AM ECM MRCIb Antrocaryon klaineanum Anacardi. Ta ‡ 1 Trychoscypha acuminata Anacardi. N ‡ 1 Enantia chlorantha Annon. N ‡ 1 Hexalobus crispiflorus Annon. N ‡ U Alstonia boonei Apocyn. N ‡ 1 Picralima nitida Apocyn. N ‡ U Canarium schweinfurthii Burser. Ta ‡ 4 Afzelia bipindensis Caesalpi. Ta ‡‡ 4 Afzelia pachyloba Caesalpi. Ta ‡‡ 4 Amphimas ferrugineus Caesalpi. Tb ‡ 3 Amphimas pterocarpoides Caesalpi. Tb ‡ 2 Anthonotha fragrans Caesalpi. Tb ‡c ‡ 3 Anthonotha macrophylla Caesalpi. Tb ‡ 4 Berlinia bracteosa Caesalpi. Ta ‡ 5 Berlinia confusa Caesalpi. Ta ‡ 5 Brachystegia cynometroides Caesalpi. Ta ‡ 5 Brachystegia eurycoma Caesalpi. Tb ‡ 5 Brachystegia zenkeri Caesalpi. Tb ‡ 5 Cynometra hankei Caesalpi. Tb 3 Cynometra sanagaensis Caesalpi. Tb ‡ 3 Daniella ogea Caesalpi. Ta ‡ 1 Detarium macrocarpum Caesalpi. Ta ‡ 1 Dialium spp Caesalpi. Tb ‡ 1 Didelotia africana Caesalpi. Tb ‡ 5 Didelotia letouzeyi Caesalpi. Ta ‡ 5 Distemonanthus benthamianus Caesalpi. Ta ‡ 4 Erythrophloeum ivorense Caesalpi. Ta ‡d 4 Gilbertiodendron brachystegioides Caesalpi. Ta ‡ 5 Gilbertiodendrom dewevrei Caesalpi. Ta ‡ 5 Gossweilerodendron balsamiferum Caesalpi. Ta ‡ 2 Guibourtia tessmannii Caesalpi. Ta ‡ 3 Julbernardia seretii Caesalpi. Tb ‡ 5 Monopetalanthus le-testui Caesalpi. Tb ‡ 5 Monopetalanthus microphyllus Caesalpi. Tb ‡ 5 Oxystigma buchholzii Caesalpi. Tb ‡ 3 Oxystigma mannii Caesalpi. Tb ‡ 3 Pachyelasma tessmannii Caesalpi. Tb ‡ 3 Paraberlinia bifoliolata Caesalpi. Tb ‡ 5 Plagiosyphon longitubus Caesalpi. Tb ‡ 3 Plagiosiphon multijugus Caesalpi. Tb ‡ 3 Scorodophloeus zenkeri Caesalpi. N ‡ 1 Tetraberlinia bifoliolata Caesalpi. Ta ‡ 5 Toubaouate brevipaniculata Caesalpi. Tb ‡ 5 Garcinia kola Clusi. N ‡ 1 Garcinia lucida Clusi. N, E ‡ 1 Terminalia superba Combret. Ta ‡ 4 Diospyros spp Eben. Tp ‡ U Alchornea cordifolia Euphorbi. E ‡ 5 Drypetes gossweileri Euphorbi. N ‡ 1 Ricinodendron heudelotii Euphorbi. Ta ‡ 3 Pterocarpus mildbraedii Fab. Tb ‡e 5 Pterocarpus soyauxii Fab. Ta ‡e 5 Gnetum africanum Gnet. N ‡ 5 Gnetum buchholzianum Gnet. N ‡ 5 Gnetum sp. Gnet. N ‡c ‡ 5 282 N.A. Onguene, T.W. Kuyper / Forest Ecology and Management 140 (2001) 277±287

Table 2 (Continued )

Species Family Category AM ECM MRCIb

Sacoglottis gabonensis Humiri. E ‡ 1 Irvingia gabonensis Irvingi. N, E ‡ 2 Anthocleista schweinfurthii Logani. E ‡ 5 Entandrophragma angolense Meli. Ta ‡ 1 Entandrophragma candollei Meli. Ta ‡ 1 Entandrophragma cylindricum Meli. Ta ‡ 1 Entandrophragma utile Meli. Ta ‡ 2 Guarea cedrata Meli. Ta ‡ 2 Khaya ivorensis Meli. Ta ‡ 2 Lovoa trichilioides Meli. Ta ‡ 3 Cylicodiscus gabunensis Mimos N ‡ 2 Pentaclethra macrophylla Mimos. N ‡ 1 Piptadeniastrum africanum Mimos. Ta ‡ 4 Musanga cecropioides Mor. E ‡ 4 Milicia excelsa Mor. Ta ‡ 1 Pycnanthus angolensis Myristic. Ta ‡ 1 Staudtia kamerunensis Myristic. Ta ‡ 1 Lophira alata Ochn. Ta ‡ 4 Coula edulis Olac. N ‡ 2 Ongokea gore Olac. Tb ‡ 1 Strombosis grandifolia Olac. E ‡ 2 Panda oleosa Pand. Tb ‡ 1 Poga oleosa Rhizophor. N ‡ 1 Mitragina ciliata Rubi. Ta ‡ 1 Nauclea diederrichii Rubi. Ta ‡ 2 Pausynistalia johimbe Rubi. N ‡ 3 Fagara heitzii Rut. Ta ‡ 2 Allophylus schweinfurthii Sapind. E ‡ 2 Blighia welwitschii Sapind. E ‡ 2 Deinbollia pycnophylla Sapind. E ‡ 2 Eriocoelum macrocarpum Sapind. E ‡ 2 Aningeria robusta Sapot. Ta ‡ 3 Autranella congolensis Sapot. Tb ‡ 2 Baillonella toxisperma Sapot. Ta ‡ 2 Chrysophyllum sp. Sapot. Tb ‡ 2 Gambeya africana Sapot. Ta ‡ 3 Omphalocarpum procerum Sapot. Ta ‡ 2 Eribroma oblonga Sterculi. Ta ‡ 2 Pterigota macrocarpa Sterculi. Tb. ‡ 3 Uapaca acuminata Uapac. E ‡c ‡ 4 Uapaca guineensis Uapac. N ‡ 4 Uapaca staudtii Uapac. E ‡ 4 Uapaca vanhouttei Uapac. E ‡ 4 Celtis sp. Ulm. Tb ‡ 4 Trema orientalis Ulm. E ‡ 5 a Category Ð Ta: timbers actually exploited by logging companies; Tb: potentially exploitable timbers in the future or used by local populations for firewood or other domestical needs; Tp: totally protected tree species by Cameroon Forest Law; N: tree species providing non- timber forests products such as fiber, fruits, nuts, bark, to local populations; E: early successional tree species; AM: Arbuscular mycorrhiza; ECM: Ectomycorrhiza. b MRCI: mycorrhizal root colonization class; referring to the following classes of mycorrhizal colonization of fine roots Ð U: undefined; 1: 1±5%; 2: 6±25%; 3: 26±50%; 4: 51±75%; 5: 76±100%. c (‡): Few mycorrhizal internal structures. d Presence of nodules only at the seedling and sapling growth stages. e Presence of nodules at all growth stages. N.A. Onguene, T.W. Kuyper / Forest Ecology and Management 140 (2001) 277±287 283 common. Vesicles were about twice as common as Melangue (North of Ebom). In the Nyangong site, arbuscules. Intraradical hyphae were most often average density of the investigated tree species was straight with rami®cations along the root cortex, 368 stems per hectare and basal area 0.64 m2 ha1. sometimes tortuous and thicker. Various kinds of Uapaca species were dominant in terms of stem vesicles were observed, viz. oval, irregularly lobed, numbers (11%) and basal area (26%). ECM trees and rectangular. Both smooth and echinulate auxiliary contributed 26% of stems and 34% of basal area. Five cells were observed. clumps dominated by Uapaca species and two domi- ECM colonization was much higher, and was often nated by caesalps were observed. They occurred more than 75%. Root samples of ECM tree species mostly on hills. Individual Uapaca trees occurred near showed either abundant rhizomorphs or smooth man- riverside forests. tles with different colours. ECM of Uapaca species were mostly smooth. Root samples of Gnetum species 4. Discussion harbored either sulphur yellow, white, brown, or dark brick ECM, covered with abundant rhizomorphs. 4.1. Mycorrhizal status of selected plant species and Roots were not completely covered by a mantle, extent of mycorrhizal colonization and numerous root hairs were also observed. In dual mycorrhizal species colonization was predominantly In our study we did not observe non-mycorrhizal by ECM. trees, con®rming the ubiquity of mycorrhizal associa- tions in the tropical rainforest. Out of 56 species 3.2. Stand composition and occurrence of ECM investigated in Korup, Cameroon, 55 turned out to associations be mycorrhizal and only one species, viz. Warneckea memecyloides (Melastomataceae) was non-mycorrhi- In the Ebimimbang site, average density of the zal (Newbery et al., 1988). In French Guyana BeÂreau investigated tree species was 452 stems per hectare. et al. (1997) noted that all 75 tree species investigated Basal area was 0.73 m2 ha1. In terms of stem num- were mycorrhizal. bers Diospyros spp. were dominant (45%), but in A low amount of AM colonization was observed in terms of basal area ECM caesalps dominated these root samples of tree species of several families. stands. ECM trees contributed 22% of stems and 35% Amount of colonization was often comparable within of basal area, whereas Diospyros spp contributed only several members of the same family, e.g., low in 8% to basal area. In the Ebimimbang area, six clumps Anacardiaceae, Meliaceae, Myristicaceae, and Sapin- dominated by ECM caesalps were noted within a daceae, and high in Ulmaceae. Colonization was also 25 km2 area, in ®ve north±south, 5 km long transects, high in the nitrogen-®xing legumes Erythrophleum distant of 1 km. All of them were situated on ¯at and Pterocarpus. It has been noted by Janos (1996) terrain. In the Ebom site, average density of the that the correlation between early seral status and investigated tree species was higher, viz. 583 stems independence of mycorrhizas is loose, and it is there- per hectare, but basal area lower, viz. 0.52 m2 ha1. fore not surprising that in our study area early- and Staudtia kamerunensis was numerically dominant late-successional trees did not show differences in (34% of stem numbers, with 5% contribution to basal their amount of colonization. Characteristics of plant area), and Pycnanthus angolensis contributed most to roots usually highly correlate with mycorrhizal for- basal area (28%, with 10% of stem numbers). These mation and abundance (Baylis, 1975). Coarse roots two tree species belonging to the Myristicaceae con- with no or few root hairs are often highly mycotrophic tributed 33% of total basal area. ECM trees contrib- while highly branched, ®ne, long roots with numerous uted 7% of stems and 19% of basal area. Around root hairs have often been observed to derive little Ebom, ECM tree species occurred as individual plants nutritional bene®t from mycorrhizas in controlled or in small groups scattered within the mass of AM experiments (Manjunath and Habte, 1991). In our tree species. Two monodominant clumps of Gilber- study, roots of most tree species were rather coarse tiodendron dewevrei were observed, one at the bottom with few root hairs. However, they did not show of a non-¯ooded valley and the other along the river abundant AM colonization. 284 N.A. Onguene, T.W. Kuyper / Forest Ecology and Management 140 (2001) 277±287

The presence of arbuscules in roots is generally fungi (Smith et al., 1998). In Africa, reports of AM used to designate plants with functional AM. In this in roots of these latter genera exist for Gnetum afri- investigation arbuscules were rarely observed in most canum (Fassi and Moser, 1991), Uapaca guineensis root samples. Hyphal coils were about as common as (Thoen and Ba, 1989), and U. kirkiana (HoÈgberg, arbuscules. In most samples only intracellular hyphae 1982). In roots of Anthonotha fragrans we once were present. Our observations of the lack or scarcity encountered arbuscules and AM external hyphae with of arbuscules are consistent with those in the tropical echinulate auxiliary cells of a Gigaspora species. wet forests of the neotropics by Janos (1984) and Alexander (1989b) also reported presence of AM BeÂreau et al. (1997). Both the amount of arbuscules fungi in the ECM trees Gilbertiodendron and Pelle- and coils and the amount of internal colonization griniodendron, and Moysersoen and Fitter (1999) might depend on differences in root morphology. reported the presence of AM fungi in Anthonotha, Smith and Smith (1997) reviewed structural diversity Berlinia, Didelotia, Gilbertiodendron, Monopeta- of AM and recognised two types, viz. Arum-type and lanthus, and Tetraberlinia. Paris-type. In the Arum-type, extensive intercellular The frequent occurrence of vesicles indicates that a hyphae and arbuscules develop, while in the Paris- large part of the AM fungi belong to the Glomineae, type these structures are absent and hyphal coils occur and diversity in vesicle shape indicates the presence of commonly. both Glomus (vesicles ellipsoid) and Acaulospora Three families (Caesalpiniaceae, Uapaceae, Gneta- (vesicles more irregularly shaped or rectangular). ceae), 13 genera, and 26 species formed ECM. Thoen Predominance of Glomineae has been reported for (1993) recorded the ECM habit in 18 genera and 50 tropical sites (Gianinazzi-Pearson and Diem, 1982; species of African trees. Most prominent in his list are Sieverding, 1991). Gigasporineae are much rarer, but Caesalpiniaceae (13 genera with 37 species in tribe smooth auxiliary cells of Scutellospora and echinulate Macrolobieae (ˆAmherstieae) and one genus with six auxiliary cells of Gigaspora have both been observed. species (Afzelia) in tribe Detarieae). Paraberlinia and A large number (more than 150) of ECM fungal Touabouate (Macrolobieae) have not been reported species have been observed (T.W. Kuyper and N.A. before as ECM trees. ECM caesalps are not only Onguene, unpublished observations). Important ECM common in the wet tropical forest but also in the fungal genera are Amanita, Russula, Lactarius, and savanna (miombo) of East Africa (HoÈgberg, 1989). various genera of the Cantharellaceae and Boletales. The large number of ECM caesalps in Korup (10 Less important genera include Scleroderma, Cortinar- genera with 22 species) and the TCP area (11 genera ius, and Inocybe. Field observations indicated a low with 19 species) con®rm the importance of the Bia- degree of host-speci®city, as most of these species frean forest for that group. Most of the ECM caesalps have been noted under various Caesalpiniaceae and belong to the ekop group. The name ekop has been Uapacaceae. Scleroderma sinnamariense with its used by various timber prospectors for a variety of bright yellow mycorrhizas and rhizomorphs has only timber species. Letouzey and Mouranche (1952) been encountered with Gnetum (Fassi, 1957) and recognised 10 taxa of ekop with a further 10 unnamed might be considered as host-speci®c. ekops. Of the 10 ekop taxa listed explicitly, nine belong to the ECM tribe Macrolobieae, whereas the 4.2. Stand composition and occurrence of ECM identity of the 10th species was doubtful. associations Among the ®ve dual mycorrhizal plant species, AM structures were present to a different extent. AM Not only the large number of ECM trees (species structures were consistently found in roots of Afzelia richness), but also their contribution to stem numbers species, suggesting that it is truly dual mycorrhizal and basal area (ranging from 19 to 35% in the various (Alexander, 1989b; Moysersoen and Fitter, 1999). In sites) con®rm that the TCP-area, just as Korup, is a Uapaca guineensis and Gnetum species, AM occurred major area for these trees. Basal area of ECM trees incidentally in root portions without ECM, suggesting was even higher in Bityili, a site about 15 km SE of that in conditions of lower ECM inoculum potential Nyangong, reaching 73% (N.A. Onguene, unpub- roots might be susceptible to colonization by AM lished observations). The lower contribution of N.A. Onguene, T.W. Kuyper / Forest Ecology and Management 140 (2001) 277±287 285

ECM trees to basal area in Ebom can be explained by and facultatively mycorrhizal plants while replace- the fact that these stands are somewhat younger than ment by obligately mycorrhizal trees will take a longer the stands in Ebimimbang and Nyangong. This can be period of time (Janos, 1996). The more limited ability concluded from the higher number of stems and lower of the ECM symbiosis to establish after large-scale basal area in Ebom and the high representation of disturbance (Janos, 1996) might explain their clumped members of the Myristicaceae, which are generally distribution better than their supposed competitive characteristic for old secondary forests (Thirakul, superiority. Management and conservation practices 1983; Letouzey, 1985). Microberlinia bisulcata, the in rain forests should take these clumps into account, species showing clumping behaviour to the strongest as ECM tree recruitment after large-scale deforesta- degree, is absent from the TCP area, although it occurs tion might be a limiting factor. both in Korup and the Douala-Edea forest (Newbery and Gartlan, 1996). Both Korup and the area around Kribi have been mentioned by Rietkerk et al. (1995) as Acknowledgements glacial (pleistocene) refugia. The similar species com- position between Korup and the TCP area indicates, This work was ®nancially supported by The Nether- however, that neither a very pronounced dry period (as lands Organization for Scienti®c Research (NWO- evidenced in Korup) nor the presence of sandy soils Priority Programme Biodiversity in Disturbed Eco- that are very prone to leaching are necessary for ECM systems). Logistics were provided through the Tro- caesalps to become canopy dominants. penbos Cameroon Programme (TCP). Field assistants The occurrence of mixtures of ECM and AM trees were Jean Baptiste Mva, Eric Semengue, Serge Aba'a in the same stands has raised the question whether and Edouard Nsomoto. Technical support was pro- such forests are or are not in equilibrium (Connell and vided by Veronique Onguene. We acknowledge dis- Lowman, 1989). However, the question of an equili- cussions with Dr. G. Chuyong on the comparability of brium between both functional mycorrhizal types Korup with forests in the TCP area, and comments on might not be resolved because of the variability of earlier versions of the manuscript by Drs. B. Foahom, the climate in the last 30,000 years (Maley and Brenac, L. Brussaard, and X. van der Burgt. 1998). Their data from western Cameroon indicate that caesalps generally have a low frequency, even after a recovery from a climatic deterioration (a References relatively arid climate) that occurred between 2800 and 2000 years BP. However, their data also indicate Alexander, I.J., 1989a. Mycorrhizas in tropical forests. In: Proctor, J. (Ed.), Mineral Nutrients in Tropical Forest and Savanna that both ECM and AM caesalps reacted in a similar Ecosystems. Blackwell Scientific Publications, Oxford, way to climatic changes, which would be consistent pp. 169±188. with observations that both functional types are Alexander, I.J., 1989b. Systematics and ecology of ectomycorrhizal equally ef®cient in extracting phosphorus from these legumes. Mon. Syst. Bot. Missouri Bot. Gdn 29, 607±624. Alexander, I.J., HoÈgberg, P., 1986. Ectomycorrhizas of tropical highly leached poor soils (Moyersoen et al., 1998). angiospermous trees. New Phytol. 102, 541±549. Even if both functional types occupy the same Anon., 1995. Rapport annuel d'activiteÂs 1994±1995, Aout 1995. niche, their potential to recolonize these sites after MinisteÁre de l'Environnement et des ForeÃts (MINEF). Poste major deforestation could be dissimilar. Plant succes- Forestier et de chasse de Kribi-Port. ReÂpublique du Cameroun, sion after deforestation will be determined by the 10 pp. AubreÂville, A., 1970. Flore du Cameroun. LeÂgumineuses-CeÂsalpi- availability of AM and ECM inoculum. AM fungi nõÈoideÂes. MuseÂum National d'Histoire Naturelle. Laboratoire are more capable of surviving land clearing than ECM de PhaneÂrogamie. fungi both because spore viability and longevity is Baylis, G.T.S., 1975. The magnolioid mycorrhiza and mycotrophy higher (having larger spores with substantial energy in root systems derived from it. In: Sanders, F.E., Mosse, B., reserves) and because host tree speci®city is lower Tinker, P.B. (Eds), Endomycorrhizas. Academic Press, New York, pp. 373±389. (Janos, 1996). If spore numbers of AM fungi are high, BeÂreau, M., Gazel, M., Garbaye, J., 1997. Les symbioses AM inoculum might rapidly build up. If spore num- mycorhiziennes des arbres de la foreÃt tropicale humide de bers are low, succession starts with non-mycorrhizal Guyane francËaise. Can. J. Bot. 75, 711±716. 286 N.A. Onguene, T.W. Kuyper / Forest Ecology and Management 140 (2001) 277±287

Connell, J.H., Lowman, M., 1989. Low-diversity tropical rain- Letouzey, R., 1985. Notice de la carte phytogeÂographique du forests: some possible mechanisms for their existence. Am. Cameroun au 1:500000. Institut de la carte internationale de la Natl. 134, 88±119. veÂgeÂtation. Toulouse, France. van Dijk, H., 1994. Inventory of the names and functions of non- Letouzey, R., Mouranche, R., 1952. Ekop du Cameroun. Publica- timber forest product species. Intermediate Report, Tropenbos tion 4. Centre Technique Forestier Tropical. Nogent S/Marne, Cameroon Programme (TCP), Kribi, Cameroon, 62 pp. France. Fassi, B., 1957. Ectomycorrhizes chez le Gnetum africanum Welw. Maley, J., Brenac, P., 1998. Vegetation dynamics, palaeoenviron- due aÁ Scleroderma sp. Bull. Soc. Mycol. Fr. 73, 280±286. ments and climatic changes in the forests of western Cameroon Fassi, B., Moser, M., 1991. Mycorrhizae in the natural forests of during the last 28,000 years B.P. Rev. Palaeobot. Palynol. 99, tropical Africa and the Neotropics. In: Fontana, A. (Ed.), 157±187. Funghi, Piante e Suolo. Centro di studio Micologia del Manjunath, A., Habte, M., 1991. Root morphologial characteristics Consiglio nazionalle delle Ricerche, Torino, Italy, pp. 183±202. of host species having distinct mycorrhizal dependency. Can. J. Franqueville, A., 1973. Atlas reÂgional sud-ouest. ReÂpublique Unie Bot. 69, 671±676. du Cameroun. ORSTOM, YaoundeÂ, Cameroun. Moyersoen, B., Alexander, I.J., Fitter, A.H., 1998. Phosphorus van Gemerden, B.S., Hazeu, G.W., 1999. Landscape ecological nutrition of ectomycorrhizal and arbuscular mycorhizal tree survey (1:100 000) of the Bipindi Ð Akom II Lolodorf region seedlings from a lowland tropical rain forest in Korup National in the South West Cameroon. Tropenbos Cameroon Programme Park. Cameroon. J. Trop. Ecol. 14, 47±61. SC-DLO Report. Wageningen, The Netherlands. Moysersoen, B., Fitter, A.H., 1999. Presence of arbuscular Gianinazzi-Pearson, V., Diem, H.G., 1982. Endomycorrhizae in the mycorrhizas in typically ectomycorrhizal host species from tropics. In: Dommergues, Y.R., Diem, H.G. (Eds.), Microbiol- Cameroon and New Zealand. Mycorrhiza 8, 247±253. ogy of Tropical Soils and Plant Productivity. Junk, The Hague, Newbery, D.M., Gartlan, J.S., 1996. A structural analysis of rain pp. 209±251. forest at Korup and Douala-Edea Cameroon. Proc. R. Soc. HoÈgberg, P., 1982. Mycorrhizal associations in some woodlands Edinb. 104B, 177±224. and forest trees and shrubs in Tanzania. New Phytol. 92, 407± Newbery, D.M., Alexander, I.J., Rother, J.A., 1997. Phosphorus 415. dynamics in lowland African rainforest: the influence of HoÈgberg, P., 1989. Root symbioses of trees in savannas. In: Proctor, ectomycorrhizal trees. Ecol. Mon. 67, 367±409. J. (Ed.), Mineral Nutrients in Tropical Forest and Savanna Newbery, D.M., Alexander, I.J., Thomas, D.W., Gartlan, J.S., 1988. Ecosystems. Blackwell Scientific Publications, Oxford, Ectomycorrhizal rain forest legumes and soil phosphorus in pp. 123±136. Korup National Park, Cameroon. New Phytol. 109, 433±450. Janos, D.P., 1980. Mycorrhizae influence tropical succession. Newbery, D.M., Songwe, N.C., Chuyong, G.B., 1998. Phenology Biotropica 12, 56±64. and dynamics of an African rainforest at Korup, Cameroon. In: Janos, D.P., 1984. Methods for vesicular-arbuscular mycorrhiza Newbery, D.M., Prins, H.H.T., Brown, N.D. (Eds.), Dynamics research in the lowland wet tropics. In: Medina, E., Mooney, of Tropical Communities. Blackwell Scientific Publications, H.A., Vasquez-Yanes, C. (Eds.), Physiological Ecology of Oxford, pp. 267±308. Plants of the Wet Tropics. Junk, The Hague, pp. 173±187. Olivry, J.C., 1986. Fleuves et rivieÁres du Cameroun. Collection Janos, D.P., 1996. Mycorrhizas, succession, and the rehabilitation Monographes hydrologiques d'ORSTOM 9, Paris, France. of deforested lands in the humid tropics. In: Frankland, J.C., Phillips, J.M., Hayman, D.S., 1970. Improved procedures for Magan, N., Gadd, G.M. (Eds.), Fungi and Environmental clearing and staining parasitic and vesicular-arbuscular mycor- Change. Cambridge University Press, Cambridge, pp. 129±162. rhizal fungi for rapid assessment of infection. Trans. Br. mycol. Kormanik, P.P., Bryan, W.C., Schultz, R.C., 1980. Procedure and Soc. 55, 158±160. equipment for staining numbers of plant root samples for Rajapakse, S., Miller, J.C., 1992. Methods for studying vesicular- endomycorrhizal assays. Can. J. Microbiol. 26, 536±538. arbuscular mycorrhizal root colonization and related root Kormanik, P.P., McGraw, A.-C., 1982. Quantification of vesicular- physical properties. Meth. Microbiol. 24, 301±317. arbuscular mycorrhizae in plant roots. In: Schenck, N.C. (Ed.), Rietkerk, M., Ketner, P., de Wilde, J.J.F., 1995. Caesalpinioideae Methods and Principles of Mycorrhizal Research. APS Press, and the study of forest refuges in Gabon: preliminary results. St. Paul, Minnesota, pp. 37±45. Adansonia 17, 95±105. Le Tacon, F., Garbaye, J., Ba, A., Beddiar, A.F., Diem, H.G., 1989. Sieverding, E., 1991. Vesicular-Arbuscular Mycorrhiza Manage- L'importance des symbioses racinaires pour les arbres forestier ment in Tropical Agrosystems. GTZ, Eschborn, Germany. en zone tropicale seÁche et en zone humide. In: Proceedings of Smith, F.A., Smith, S.E., 1997. Structural diversity in (vesicular)- the Regional Seminar on Trees in Development in Sub-Saharan arbuscular mycorrhizal symbioses. New Phytol. 137, 373± Africa. IFS, ICRAF House, Nairobi, Kenya, 20±25 February 388. 1989, pp. 302±318. Smith, J.E., Johnson, K.A., Cazares, E., 1998. Vesicular mycor- Letouzey, R., 1968. Etude PhytogeÂographique du Cameroun. P. Le rhizal colonization of seedlings of Pinaceae and Betulaceae Chevalier, Paris. after spore inoculation with Glomus intraradices. Mycorrhiza Letouzey, R., 1983. Manuel de botanique forestieÁre. Afrique 7, 279±285. tropicale. Tome 2A, Centre Technique Forestier Tropical. Smith, S.E., Read, D.J., 1997. Mycorrhizal Symbiosis, 2nd Edition. Nogent S/Marne, France. Academic Press, London. N.A. Onguene, T.W. Kuyper / Forest Ecology and Management 140 (2001) 277±287 287

Stevels, J.M.C., 1990. LeÂgumes traditionnels du Cameroun, une Thirakul, S., 1983. Manuel de dendrologie-Cameroun. Appui eÂtude agro-botanique. WAU Papers 22, 1±90. institutionnel forestier aupreÁs du CENADEFOR. Groupe Thoen, D., 1993. Looking for ectomycorrhizal trees and ectomy- Poulin, ThieÂrault LTeÂe. QueÂbec, Canada. corrhizal fungi in tropical Africa. In: Isaac, S., Frankland, J.C., Vivien, J., Faure, J.J., 1985. Arbres des foreÃts denses d'Afrique Watling, R., Whalley, A.J.S. (Eds.), Aspects of Tropical Myco- centrale. ACCT, MinisteÁre des Relations ExteÂrieures, CoopeÂra- logy. Cambridge University Press, Cambridge, pp. 193±205. tion et DeÂveloppement. Paris, France. Thoen, D., Ba, A.M., 1989. Ectomycorrhizas and putative ectomy- de Wildeman, E., 1936. Contributions aÁ l'eÂtude des espeÁces du corrhizal fungi of Afzelia africana Sm. and Uapaca guineensis genre Uapaca Baill. (EuphorbiaceÂes). MeÂm. Royal Colonial Muell. Arg. in Southern Senegal. New Phytol. 113, 549±559. Belge.