J. Bio. Env. Sci. 2016 Journal of Biodiversity and Environmental Sciences (JBES) ISSN: 2220-6663 (Print) 2222-3045 (Online) Vol. 8, No. 6, p. 134-150, 2016 http://www.innspub.net OPEN ACCESS RESEARCH PAPER Woody vegetation groups and diversity along the altitudinal gradient in mountain forest: case study of Kahuzi-Biega National Park and its surroundings, RD Congo Gérard Imani1,2*, Louis Zapfack3, John Kalume1, Bernard Riera4, Legrand Cirimwami2, Faustin Boyemba2 1 Département de Biologie, Université Officielle de Bukavu, Bukavu, RD Congo 2 Département de Botanique, Université de Kisangani, Kisangani, RD Congo 3 Département de Biologie et physiologie Végétales, Université de Yaoundé 1, Yaoundé, Cameroun 4 Laboratoire d’écologie générale, Museum National d’Histoire Naturelle, Brunoy, France Article published on June 11, 2016 Key words: Forest type, Woody distribution, Diversity, Altitude, Mountain forest . Abstract This study aims to determine the type of mountain forests in the Albertine Rift, to understand the distribution of woody species DBH ≥ 10 cm and their diversity. The investigations were conducted within 30 plots of 1 hectare located in Kahuzi Biega National Park and surrounding areas in DR Congo into mountain forest only. In total 16 797 individuals belonging to 212 specific taxa, 161 genera and 66 families were asses. Four forests types were identified along an altitudinal gradient using a hierarchical clustering (HCS) coupled to a correspondence analysis (AFC). These types are sub-mountain (1250-1500 m), lower mountain horizon (1500-1800 m), medium mountain horizon (1800-2400 m) and upper mountain horizon (2400-2600 m). Each forest type is characterized by a specific number of indicators species. The Sarcochore is the dissemination mode which dominates in the lower altitudes (1250-1800 m) and the Ballochore within higher altitudes (1800-2600 m). Based on the analyzes of variance (ANOVA), regression and correlation, the results showed that in general woody diversity decreases as altitude increases but exceptionally, only the specific abundance was positively correlated with the elevation. The additional study of the structural variability in forest types distinguished remains important to improve the understanding of functioning of these mountain ecosystems and ensure their better sustainability. *Corresponding Author: Gérard Imani  imanigerard2006@yahoo.fr 134 | Imani et al. J. Bio. Env. Sci. 2016 Introduction (Kahuzi and Biega) in some marshes and areas along Forests assure multiple functions. One of the most the Park (Kabare and Kalehe territories) (Pierlot, currently discussed is their contribution in climate 1966; Fischer, 1993, 1996). change mitigation (Lescuyer and Locatelli, 1999; Locatelli et al., 2008; GIEC, 2014). This function In mountain areas, distribution, abundance and plant refers to the ability of forest ecosystems to be or not species diversity vary with environmental factors, carbon sinks. The amount of aboveground biomass including altitude (Senterre, 2005; Körner, 2007; accumulated in forest ecosystem depends on the Jump et al., 2009) and other ecological factors forest type. Therefore, the understanding forest complementary to composition and diversity However, perfect assessing its contribution is important before understanding 2013). of this relationship remains questionable in the scientific mitigating (Henry et al., 2010; Djomo et al., 2010; community (Sanchez-Gonzalez et Lopez-Mata, 2005). Moonen et al., 2014). This remains the case for The altitudinal limits of vegetation can vary according mountain forests in the Congolese Albertine Rift. to the particularities of each region (Cuello, 2002; Due to a high diversity, the difficult and always Karger et al., 2011), so that understanding plant fragmentary the species composition, abundance and diversity in different morphologies that take plants according to connection with altitudinal gradient, for some different regions, remain rudimentary and elusive (Forsyth, altitudes, climate (Mangambu, change systematic to the altitude knowledge mountain and forests to attract researcher's attention for the past few decades to 1998). understand their role in the dynamic ecosystem functioning linked with changes in florist Because this relationship is not yet well known in this composition, diversity and structure (Plumptre et al., region, the aim of this study is to distinguish the 2007; Guillaumet, 2009; Delnatte, 2010). we note different forest types in the mountain forests in that these forests are rich in plant communities Kahuzi Biega National Park and its surroundings and (Mutke et al., 2011) and the Albertine Rift is known to to show that these types differ most strongly when be one of the hotspot areas of global biodiversity their altitudes are far removed. We will seek to (Myers et al., 2000; Linder, 2001). answer the question about the influence of altitude on species distribution (or grouping of vegetation), their Because of its recent creation (1970), Kahuzi-Biega functional traits and diversity in mountain forest. National Park (KBNP) was not part of those protected areas intensely explored between 1920 and 1960 Materials and methods (Mangambu, Study area 2013) despite its strategic phytogeographical position. The majority of scientific This study was conducted in the Congolese Albertine investigations carried out directly on this site are Rift region, eastern part of the Democratic Republic largely post-colonial. Among the most relevant, are of Congo, specifically in the KBNP and community the inventories of some taxonomic groups, such as forests around the Park, near Bunyakiri locality (Fig. wood-destroying fungi (Balezi, 2013), diversity in 1). KBNP is between 1 ° 36 'and 2 ° 37' South latitude Bamboo community (Amani et al., 2008), and 27 ° 33 'and 28 ° 46' East longitude. There are biogeography and systematic of ferns (Mangambu, two centers of endemism in the area: Guinea-Congo 2013) and some forest dynamics studies (Balezi et al., and afro-mountain. The study was done only in 2010; Masumbuko, 2011). Therefore, no article or mountain formations. They are located between 1250 thesis are dedicated directly to KBNP except in and 2600m altitude. Beyond was the Afro-alpine historical inventories of large exploration in the DRC, vegetation. Indeed, zone exploration shown that the including the two highest mountains of KBNP forests in the transition zone between 1500 m and 135 | Imani et al. J. Bio. Env. Sci. 2016 1900 m altitude in KBNP are disturbed due to human year (Yamagiwa et al., 2005). The rain abundance activities of mining and carbonization (Amsini et al., varies with altitudinal elevation so that there is a high 2008; Brown et Kasisi, 2008), reason why the study abundance in altitudes between 2400-2600 m and was extended to community forests around but fog beyond 2900 m (Fischer, 1993). The overall always in the same Rift region, targeting slightly average disturbed ecosystems. (Yamagiwa et al., 2005). The average relative radiation 421.8 calories/cm3 / month humidity is 86% (Mangambu, 2013). Soil in KBNP In this zone the average temperature is 20.5 °C. The and in community forests is mostly sandy clay or clay rainfall is abundant, in the range of 1750-2000 mm / with an acidic pH. Fig. 1. Study area showing the distributions of different plots. In this region, The vegetation stage occurs with forests by targeting the least disturbed forest. Inside altitude (Fischer, 1996; Mangambu, 2013) : the parcel, all woody with DBH ≥ 10 cm to 1.30 m Lowland rain forest: between 650-1200m were considered in the study. For each individual, the Transition Rain forest, sub-montane: between 1250- scientific or common name (Yumoto et al., 1994), the 1700 m DBH and height were recorded using respectively a Mountain Rain forest: between 1700-2400 m DBH meter and a Laser Ace. The average altitude of Bamboos community: between 2400-2600 m the plot and its coordinates were also taken. The Afro Sub-alpin stage between 2600-3326 m identification of species collected in forest was Our study focuses only on mountain part, between verified by comparison with reference specimens 1250 to 2600 m altitudes. preserved in the herbarium of the Centre for Research in Natural Sciences Lwiro (LWI), INERA Mulungu Data collection (MLGU). Data on plant diversity were collected in 30 permanent plots (1ha each), as proposed by Gentry Data analysis (1988). Each parcel was subdivided in four small plots In this study, different forest types were identified (0.25ha) in order to facilitate inventories. For a using an ascending hierarchical classification (HCS) balanced sampling on different altitude levels, ten coupled with a correspondence analysis (CA) (Lepš plots were placed respectively between 1250-1700 m; and Šmilauer, 1999; Oksanen, 2014) based on a 1700-2150 m and 2150-2600 m in the mountain binary table, presence-absence (Roux and Roux, 1967; 136 | Imani et al. J. Bio. Env. Sci. 2016 Faith et al., 1987; Lenoir, 2016). This approach altitude on these index according forest types. All avoided the Guttman effect observed with the tests were done using the R 3.0.2 software and abundance data (Meddour, 2011), minimizing the considered significant at the 5% level. effect of rare species and give a map of CA with the inertia of ascending hierarchical classification. The Results aggregation model for the ascending hierarchical Change classification is that of Ward because it uses the distribution along altitudinal gradient concept of inertia. The approach IndVal was used to For this study, 16 797 individuals were recorded in 30 determinate characteristic species of each type of plots of 1 ha each. In total, 212 specific taxa belonging forest (De Cáceres, 2013). For each indicators species, to 161 genera and 66 families were identified. About both probabilities of fidelity and occurrence were 99 % of the individuals recorded were identified up to calculed. The first corresponding to the exclusive species level. in floristic composition and species membership to the forest type, while the second indicates the frequency of the species within the The remaining 26 individuals were identified to the forest. In addition, the diaspore and morphological vernacular name but their genus and families remain types of each species were indicated (Habiyaremye, unidentified. The floristic composition is influenced 1995). by altitude level. Four forest types are distinguished related to the altitudinal gradient (Fig. 2). Diversity index (richness, abundance, Simpson 1-D (Eq1), Fisher alpha (Eq 2), Shannon Weaver (Eq 3) Indeed, second axis (var 4.69; p <0.001) separates Evenness (Eq4)) were used to characterize the sub-montane to upper mountain. The first one (var diversity for each plot (Lande, 1996; Magurran, 2004; 2.13; p = 0.03) separates mountain forest into Oksanen, 2014). different groups. Thus, we have mountain type higher horizon (black), mountain type medium horizon Eq(1) S = ∑ fi2 avec S varies between 0 and 1 for minimum and maximum diversity. Ni is a number of individual for one specie and NT a total number of individual for all species. sub mountain type (blue). The Fig. 3 below shows the altitudinal distribution of these forest types. The sub-mountain type (A) starts Eq(2) S= αln (1+N/α) with α = Fisher alpha index, N= total number of individual and S total number of species. Eq(3) H’= ∑ [(Ni/NT) * Log2 (Ni/NT)] (red), mountain type lower horizon (green) and the from 1250 m to 1500 m level; it is followed by the mountain lower horizon (B) between 1500 m and 1800 m altitude. The mountain type medium horizon (C) is between 1850 and 2400 m and finally, the with H’ = Shannon index. Eq(4) EQ = H’/log2(S) ; EQ = Evenness and varies between 0 and 1. mountain type upper horizon (D) between 2400 and 2600 m altitude. Table 1 summarizes all species which characterize each forest type along the elevational gradient. To understand relationship between diversity and altitude, correlation and multiple regressions were used. Furthermore, the analysis of variance one way (ANOVA) was applied on a linear model between altitude and abundance and between altitude and species richness in order to show the influence of In fact, some species are faithful to their own forest type. In this case, their probability of fidelity (A) is 1. Other species can be found in all plots that compose the forest type, and then their occurrence probability (B) is 1. For example: 137 | Imani et al. J. Bio. Env. Sci. 2016 Table 1. Indicator species for different forest types. Num Species Probability A (Fidelity) B (Occurrence) Indival (IndVal.g) Morphologic Indival p > 5% type Dissemination model 28 indicator species of sub mountain type 1250 to 1500 m level 1 Pycnanthus angolensis 1.0000 1.0000 1.000 0.0009 *** A Sarco 2 Trichilia welwichii 0.9577 1.0000 0.979 0.0009 *** Arb Ballo 3 Ficus urceolaris 0.8511 1.0000 0.923 0.0009 *** A Sarco 4 Erythrococa welwitchii 1.0000 0.8333 0.913 0.0009 *** Arb Ballo 5 Sterculia tragacantha 1.0000 0.8333 0.913 0.0009 *** Arb Sarco 6 Tarenna soyauxii 1.0000 0.8333 0.913 0.0009 *** Arb Sarco 7 Pseudospondias microcarpa 0.9568 0.8333 0.893 0.0009 *** A Sarco 8 Monodora myristica 0.9322 0.8333 0.881 0.0009 *** A Sarco 9 Celtis soyauxii 1.0000 0.8333 0.913 0.0029 ** Arb Sarco 10 Erythrina mildbraedii 1.0000 0.6667 0.816 0.0039 ** A Ballo 11 Leptonychia bampsii 1.0000 0.6667 0.816 0.0059 ** A Sarco 12 Phyllanthus muellerianus 1.0000 0.6667 0.816 0.0059 ** Arb Ballo 13 Vepris orophila 1.0000 0.6667 0.816 0.0059 ** Arb Sarco 14 Markhamia lutea 0.9882 0.6667 0.812 0.0049 ** Arb Ballo 15 Teclea nobilis 0.8750 0.6667 0.764 0.0099 ** A Sarco 16 Drypetes spinosodentata 1.0000 0.5000 0.707 0.0079 ** A Sarco 17 Garcinia volkensii 1.0000 0.5000 0.707 0.0090 ** A Sarco 18 Pentaclethra macrophylla 1.0000 0.5000 0.707 0.0079 ** A Ballo 19 Trema orientalis 0.8400 0.6667 0.748 0.010989 * Arb Sarco 20 Diospyros troupinii 1.0000 0.5000 0.707 0.011988 * Arb Sarco 21 Ficus spbunya2 1.0000 0.5000 0.707 0.0139 * A Sarco 22 Rauwolfia vomitoria 1.0000 0.5000 0.707 0.0109 * Arb Sarco 23 Ricidendron heudelotii 1.0000 0.5000 0.707 0.0169 * A Sarco 24 Ficus arcuatone 0.9492 0.5000 0.689 0.0119 * Arb Sarco 25 Milletia ferruginea 0.9474 0.5000 0.688 0.0349 * A Ballo 26 Spathodea campanulata 0.9231 0.5000 0.679 0.0269 * A Ballo 27 Grewia malacocarpoides 0.8800 0.5000 0.663 0.0449 * Arb Sarco 28 Baphiopsis parciflora 0.8929 0.5000 0.668 0.0449 * A Ballo 15 indicator species of mountain type lower horizon 1500 to 1800 m level 1 Lebrunia buchaie 0.8933 1.0000 0.945 0.0009*** A Sarco 2 Drypetes dinklagei 0.8618 1.0000 0.928 0.0009 *** Arb Sarco 3 Grewia mildbraedii 0.7860 1.0000 0.887 0.0009 *** A Sarco 4 Vitex rubro.aurantia 0.7834 1.0000 0.885 0.0009 *** Arb, A Sarco 5 Cleistanthus polystachyus 0.9744 0.8333 0.901 0.00297 ** A Sarco 6 Manilkara multinervis 0.7961 1.0000 0.892 0.0029 ** Arb Sarco 7 Leplaea mayumbensis 0.7803 1.0000 0.883 0.0019 ** A Ballo 8 Carapa grandiflora 0.7545 1.0000 0.869 0.0030 ** A Ballo 9 Fagara gilletii 0.7241 1.0000 0.851 0.0030 ** A Sarco 10 Diospyros sp. 1.0000 0.6667 0.816 0.0050 ** Arb Sarco 11 Garcinia punctanta 1.0000 0.6667 0.816 0.0060 ** A Sarco 12 Monanthotaxis poggei 1.0000 0.6667 0.816 0.0060 ** Arb Sarco 13 Anthocleista grandiflora 0.9897 0.6667 0.812 0.0050 ** A, Arb Sarco 14 Alchornea laxiflora 0.9615 0.6667 0.801 0.0050 ** Arb Ballo 15 Ocotea usambarensis 0.7000 0.8333 0.764 0.0198 * A Sarco 5 indicator species of mountain type medium horizon 1800 to2400 m 1 Maesa lanceolata 0.8708 0.7857 0.827 0.016 * Arb Sarco 2 Lindackeria kivuensis 0.9383 0.6429 0.777 0.014 * Arb Ballo 3 Allophyllus kiwuensis 1.0000 0.5714 0.756 0.026 * A Sarco 4 Bridelia micranta 1.0000 0.5000 0.707 0.030 * A Ballo 5 Dombea goetzenii 1.0000 0.4286 0.655 0.042 * A Ballo 138 | Imani et al. J. Bio. Env. Sci. 2016 8 indicator species of mountain type upper horizon 2400 m to 2600 m 1 Agauria salicifolia 1.0000 1.0000 1.000 0.0009 *** Arb Ballo 2 Hagenia abyssinica 0.9922 1.0000 0.996 0.0009 *** Arb, A Ptéro 3 Nuxia floribunda 0.9837 1.0000 0.992 0.0009 *** A Ballo 4 Rapanea melanophloeus 0.9314 1.0000 0.965 0.0009 *** A Sarco 5 Nuxia congesta 0.8297 0.7500 0.789 0.0229 * A Ballo 6 Ficalhoa aurifolia 1.0000 0.5000 0.707 0.0159 * A Ballo 7 Myrica salicifolia 1.0000 0.5000 0.707 0.0159 * A Sarco 8 Hypericum revoluta 0.9937 0.5000 0.705 0.0179 * Arb Ballo Legend: = Ballo = Ballochories; Sarco = Sarcochories; Ptero = Pterochories; A = Tree; Arb = Shrub. The sub-mountain type 1250 to 1500 m: 57% of Ricidendron heudelotii. Of these species, some are indicator species are faithful to this type: Pycnanthus characteristic of undergrowth and other from tree angolensis, Erythrococa Sterculia strata. In the undergrowth, we note: Erythrococa Monodora welwitchii, Sterculia tragacantha, Tarenna soyauxii, myristica, Celtis soyauxii, Erythrina mildbraedii, Celtis soyauxii, Phyllanthus muellerianus, Vepris Leptonychia bampsii, Phyllanthus muellerianus, orophila, Drypetes spinosodentata. Only 11% of Vepris orophila, Drypetes spinosodentata, Garcinia indicator species are presents in all plots of this forest volkensii, type. tragacantha, Pentaclethra troupinii, welwitchii, Tarenna Ficus soyauxii, macrophylla, Diospyros Rauwolfia vomitoria, sp., Table 2. Altitudinal distribution of some species in mountain forest. Num Species Probability A (Fidelity) B Indival (IndVal.g) Morphologi Disseminatio Num Indival p > 5% c type n model (Occurrence) Indicators Species between 1250 and 1800 m altitude 1 Polyscias kivuensis 1.0000 1.0000 1.000 0.0009 *** A Ballo FS 2 Bakerisideroxylon sp 1.0000 0.9167 0.957 0.0009 *** A Ballo FP 3 Musanga cecropioides 0.9820 0.9167 0.949 0.0009 *** A Sarco FS 4 Dacryodes edulis 1.0000 0.8333 0.913 0.0009 *** A Ballo FS 5 Lovoa trichilioides 1.0000 0.8333 0.913 0.0009 *** A Ballo FP 6 Trilepisium madagascariens 1.0000 0.8333 0.913 0.0009 *** A Sarco FP, FS 7 Piptadeniastrum africanum 0.8537 0.9167 0.885 0.0009 *** A Ballo FP 8 Tetrorchidium didymostemon 1.0000 0.7500 0.866 0.0039 ** A Sarco FS, FP 9 Cynometra alexadrii 1.0000 0.5833 0.764 0.0069 ** A Ballo FP 10 Lindackeria dentata 1.0000 0.5000 0.707 0.0209 * Arb Ballo FS,FP 11 Macaranga monandra 1.0000 0.5000 0.707 0.017982 * A Ballo FS Indicators Species between 1800 and 2400 m altitude 1 Macaranga neomilbraediana 0.9663 1.0000 0.983 0.0009 *** A Ballo FP,FS 2 Polyscias fulvae 1.0000 0.9444 0.972 0.0009 *** A Sarco FS Indicators Species between 1500 and 2400 m altitude 1 Sapium ellipticum 0.9447 0.9000 0.922 0.02597 * A Sarco FS 2 Tabernaemontana johnstonii 0.9681 0.8000 0.880 0.00599 ** A, Arb Sarco FS 3 Diospyros polystemone 1.0000 0.6500 0.806 0.01399 * A Sarco FP 4 Parinari excelsa 0.9879 0.6500 0.801 0.01998 * A Sarco FP 5 Chrysophyllum gorungosanum 0.9835 0.6000 0.768 0.04795 * A Sarco FP 0.8846 0.941 0.003 ** A Sarco FP 0.003 ** A Sarco FP Indicators Species between 1250 and 2400 m altitude 1 Strombosia scheffleri 1.0000 Indicators Species between 1500 and 2600 m altitude 1 Syzygium guineense 0.9957 0.8750 0.933 Legend: = Ballo = Ballochories; Sarco = Sarcochories; Ptero = Pterochories; A = Tree; Arb = Shrub; FP= primary forest; FS= Secondary forest 139 | Imani et al. J. Bio. Env. Sci. 2016 The analysis of diaspores dissemination shows a Monanthotaxis poggei. The species of undergrowth dominance of Sarcochories species (68%), which have are Drypetes dinklagei, Manilkara multinervis, fleshy diaspores that can be transported over long Diospyros sp., Alchornea laxiflora. In fact, 53% of distances, especially by birds. Others species ejected indicator species are presents in all plots of the forest themselves their diaspores (Ballochores). type. The mountain type lower horizon, 1500 to 1800 m: Depending 20% of indicator species are faithful to this forest indicators species are Sarcochores and 20% type: Diospyros sp., Garcinia punctanta, on dissemination model, 80% of Ballochores. Table 3. Diversity measures for each inventory plot with its mean altitude. Altitude Number Family 1260 29 1300 33 1350 24 1385 25 1420 27 1527 32 1570 30 1615 35 1680 26 1702 28 1750 30 1803 27 1840 17 1934 27 1980 26 2020 22 2096 16 2100 11 2145 14 2150 20 2170 29 2240 18 2290 13 2326 17 2340 25 2370 17 2400 9 2435 7 2500 15 2590 13 Moyenne 22 Ecart-type 7,7 Number Genus 49 59 40 40 46 61 54 59 43 47 49 46 21 37 40 31 22 12 15 28 46 26 17 23 30 21 10 7 18 16 34 15,9 Number Species 50 60 46 45 51 68 58 63 50 50 53 48 22 37 42 31 23 12 15 31 49 26 17 23 31 21 10 8 19 18 36 17,5 Abundance Simpson Ficher’s Alpha Shannon index H Evenness Pièlou 410 416 429 442 465 468 404 521 555 517 543 527 322 1018 391 327 534 340 582 740 466 602 361 763 609 595 652 301 1206 1318 561 243,2 0,85 0,95 0,94 0,93 0,93 0,96 0,95 0,96 0,91 0,94 0,94 0,94 0,91 0,91 0,94 0,94 0,87 0,81 0,69 0,88 0,96 0,89 0,84 0,86 0,85 0,72 0,61 0,72 0,81 0,89 0,9 0,1 14,93 19,24 13,06 12,53 14,61 21,87 18,56 18,75 12,96 13,66 14,53 12,84 5,35 7,53 11,93 8,41 4,89 2,42 2,81 6,54 13,81 5,53 3,7 4,51 6,62 4,24 1,68 1,51 3,2 2,95 9,5 6,0 2,81 3,42 3,19 3,05 3,11 3,55 3,49 3,61 2,96 3,13 3,29 3,2 2,67 2,76 3,17 3,08 2,41 1,93 1,51 2,6 3,51 2,49 2,17 2,27 2,4 1,84 1,21 1,54 1,97 2,4 2,7 0,7 0,72 0,84 0,83 0,8 0,79 0,84 0,86 0,87 0,76 0,8 0,83 0,83 0,86 0,76 0,85 0,9 0,77 0,78 0,56 0,76 0,9 0,77 0,77 0,72 0,71 0,6 0,52 0,74 0,67 0,83 0,8 0,1 The mountain type medium horizon, 1800 to 2400 m: is noted that 37% of characteristics species are 60% of species is faithful to this forest type: faithful to this forest type: Agauria salicifolia, Allophyllus kiwuensis, Bridelia micrantha, Dombea Ficalhoa aurifolia and Myrica salicifolia. In the goetzenii. In the undergrowth strata we note: Maesa undergrowth, we note the presence of Hypericum lanceolata and Lindackeria kivuensis. Concerning to revoluta and rarely Agauria salicifolia. the occurrence probability no characteristic species is present in all plots. Furthermore, 60% of species The probability of occurrence shows that 30% of ejected theirs diaspores themselves. species are present in all sites of the forest type. Analysis of dissemination mode highlights the The mountain type upper horizon, 2400 to 2600 m: it dominance of species which ejected themselves theirs 140 | Imani et al. J. Bio. Env. Sci. 2016 diaspores (Ballochores: 63%), followed by fleshy some species may have high plasticity in their diaspores (Sarcochores: 25%) and species with wing- altitudinal distribution and the others being confined like appendages (Pterochores). only to a certain altitude level. The table below summarizes this situation: In the mountain forest of KBNP and its surroundings, Table 4. Axis contribution to the principal component analysis. Axis Eigenvalue % of variance % cumulative variance Comp1 6,77 75,26 75,26 Comp2 1,04 11,55 80,80 Comp3 0,87 9,66 96,46 Comp4 0,21 2,35 98,81 Change in floristic diversity along the altitudinal 001) Shannon (r - 0.72; p value <0, 001) and Pielou (r gradient - 0.44; p value 0.01) have a negative linear Table 3 summarizes the change in diversity along relationship with altitude elevation. altitude. We inventoried 561 ± 243 individuals, 36 ± 17 specific taxa, 34 genera and 22 ± 16 ± 8 families The diversity decreases with the increasing of across a plot. altitude. The number of individuals increases as the altitude (Fig. 4). The floristic diversity has been studied and based on the abundance, species richness, number of genera, Indeed, the abundance is positively associated with family and through diversity indices like Simpson, elevation, although there is considerable variability. Shannon Fischer Alpha and Pielou evenness. Multiple regression reduced to the significant variables (abundance and richness) confirms this The correlation analysis reveals that Simpson (r - variance (F = 10.43, Df = 29, R² = 0.799, adjusted R² 0.59; p value <0, 001), Fischer (r - 0.83; p value <0, = 0.72, p-value <0.001, Table 5). Table 5. Relationship between altitude, richness and abundance. Explicative variables Coefficient Er-T coefficient Test pvalue Constant 2285,85 142,14 16,08 <0,0001*** Richness -16,80 2,27 -7,39 <0 ,0001*** Abundance 0,46 0,16 2,84 0,008** When considering forest types, the number of The most diversified families according to the individuals per hectare is variable (Df = 3; F = 3, 48; p importance of individual number of individuals and 0,029) between plots. The mountain type upper species are Euphorbiaceae (20, 24%), Meliaceae horizon has more individuals per hectare than other (6,16%), Moraceae (5,89%), Monimiaceae (5,86%), types (Fig. 5). Myrsinaceae (4,97%), Fabaceae (4,72%), Rubiaceae (4,17%) and Sapindaceae (4,1%). Species richness decreases with the increasing of altitude. The sub-mountain types have more species Discussion per hectare than other types (Fig. 6) (Df = 3, F = Forest types along the altitude gradient 28.08, p <0.001). Ecological significance of species combination 141 | Imani et al. J. Bio. Env. Sci. 2016 (regrouping) along an altitudinal gradient Andresen, 1998) or Asia (Lee et al., 2014). The theory Altitude is one of the important factors that of determine the combination of plant communities in disappearance and appearance of species along a mountain forests (Körner, 2007; Jump et al., 2009) gradient remains an open debate in ecology (Hardy et but this can vary from one area to another so that al., 2004; Fayolle et al., 2014). Each region having its Senterre (2005) believes that no generalized model particularities (Cuello, 2002; Karger et al., 2011). can exist. Our results show that woody composition However, the more pronounced is the gradient, which changes with elevation in the area of Kahuzi-Biega allows multiple combinations of vegetation, especially National Park and its surroundings, which were seen in mountain areas, most species richness becomes elsewhere in the Tanzanian and Kenyan mountains in important the Rift (Lovett et al., 2006; Medley and Maingi, pronounced (Tassin et al., 2004). Our results showed 2014) or in other areas like Reunion island (Tassin et that the floristic composition in forest types is greatly al., different when their altitudes are far removed. 2004), Amazonian forests (Terborgh and ecological and discontinuity the difference that between allowing groups Fig. 2. Distinction of forest types along the altitudinal gradient. Hierarchical clustering analysis (HCS) combined to a correspondence analysis of (CA). The 53% cumulative percentage on the axes is not that of a classic AFC but the values of the inertia of Ward approach of the hierarchical clustering analysis In the mountains of KBNP and its surroundings, no surroundings, based on woody species. specific forest typology to mountain formations has been made until today. However, comments made by Indeed, these forests begin with a sub-mountain or some authors locate these forests between 1350 and transition forest between 1250 and 1500 m then we 2400 m (Fischer, 1996) or between 1250 and 2600 m have mountain type with three different horizons when considering mixed forest with Bambous along the altitudinal gradient. community after 2400m (Mangambu, 2013). The lower horizon between 1500 and 1800 meters is These authors confirm that the mountain forest in followed by the medium horizon, very wider, 1800 this region begins by a sub-mountain forest, or and 2400 m and finally comes the upper horizon, transition forest between 1200 and 1700m. In other restricted, between 2400 and 2600 m in which it way, for improved forest typology, consider the tree possible to see Bamboos communities in some area of woody species (Senterre, 2005). It is the reasons why the park. However, the species combination in we present in this study our results with a specific Albertine Rift mountain varies according to the zones typology of mountain forests of KBNP and its (Bussmann, 2006). 142 | Imani et al. J. Bio. Env. Sci. 2016 Fig. 3. Communities similarity according to the altitudinal gradient. A (sub-mountain), B (lower mountain horizon), C (upper mountain horizon), D (medium mountain horizon). Altitudinal range of species and trend of the particular area while others are widely spread on individual size in relation to forests type forest land. In our study, out of all 212 species The species trait is the specific factor that could recorded, the sub-mountain type (1250 to 1500 m) enable it to have a high plasticity and colonize large has itself 123 (58% of richness) with 28 indicators areas. So why some species are restricted to a species, 57% are found exclusively on this type. Fig. 4. Analysis of the relationship between diversity indices and altitude using a principal component analysis (PCA). The mountain type lower horizon (1500 to 1800) have (13%), with 8 indicators and 37% specifics to this 100 species (47%), with 15 indicators species which forest type. Despite this numerical decrease in species only 20% characterize exclusively this range. The richness, along altitude range, some species disappear mountain type medium horizon (1800 to 2400 m) has while others appear. For example, between 2400- 91 species (43%) and includes 5 indicators species of 2600 m, there are three new species Agauria which 60% own at this range. Finally, mountain type salicifolia, Myrica salicifolia and Erica arborea that upper horizon (2400 to 2600 m) with only 28 species are not found in the lower forest types; 16 species are 143 | Imani et al. J. Bio. Env. Sci. 2016 found only among 1250-1500 and 3 respectively altitudes areas are populated by endogenous and between 1500-1800 m and 1800-2400 m. Thus, specific species (Tassin et al., 2004) reason why they variability in tree species composition becomes remain fragile and have priority for conservation important when altitudes are becoming increasingly (Plumptre et al., 2009). remote. This confirms the hypotheses that high Fig. 5. Abundance variability according altitude. The dark black bar in the middle shows the average while the clear black bars and the x indicate the lower and upper values.Tranche1= Sub-mountain type, Tranche2= Mountain type lower horizon, Tranche3= Mountain type medium horizon and Tranche4= Mountain type upper horizon. Furthermore, our results showed a strong similarity 80% between 2400-2600. Some study confirms that between plots along same altitudinal range except for local environmental conditions can effect on seed the tranche 1800-2400 m which shows the variability weight of tropical plants(Opler et al., 1980). This in abundance and floristic composition whether is condition, can explain a restricted distribution of primary or secondary forest. In mountain regions, the species that characterize high altitudes. It will, for species composition is influenced by slope position these areas, to wait for some animals to come allow (Senterre, 2005). In the case of KBNP mountain wider dissemination after dark fruits that can be forest, we observe that rainfall on the eastern slope is driven by secondary dispersion (consumption of not the same as the West slope of Kahuzi and Biega fallen fruits, hanging, hunting,...) (Basabose, 2002). A Mounts and it influences the composition of pressure to animal population in this part of forest of vegetation between 1900 and 2600 m of altitude KBNP (Pierlot, 1966; Fischer, 1996). This is confirmed by composition. However, the dissemination mode is not our results in terms of woody species distribution the only factor, Soil and climate sometimes correlated among the range 1800-2400 m. with altitude, can also influence the distribution of could influence partially on floristic species (Toledo et al., 2012). In other way, distribution of individuals in a forest area depends, in partly, on dissemination model of Nonetheless, some species have a large distribution species concerned (Habiyaremye, 1995; Kumba et al., throughout the mountain forest; they are present in 2013). Between 1800-2400 levels, dissemination is all forest types in mature as secondary forest. For provided by the plant itself (Ballochores) to 60% and every elevation levels, the presence and abundance of 144 | Imani et al. J. Bio. Env. Sci. 2016 pioneer species neomilbraediana, such as Trilepisium Macaranga madagascariens, opportunistic species are related to natural disturbances and they have relatively low abundance Tetrorchidium didymostemon, Lindackeria dentata, in vary according to several parameters including abundance will also be a function of altitudinal disturbance intensity and in some cases being good gradients. indicators of secondary formations. the undisturbed natural formations. Their These Fig. 6. Richness variability along altitude gradient. The dark black bar in the middle shows the average while the clear black bars and the x indicates the lower and upper values.Tranche1= Sub-mountain type, Tranche2= Mountain type lower horizon, Tranche3= Mountain type medium horizon and Tranche4= Mountain type upper horizon. Woody diversity decreases with increasing altitude 2006; Delnatte, 2010). Only the abundance increases Mountain forests in the Albertine rift is diversified with elevation. Other researches shown that there is (Plumptre et al., 2007). We inventoried 561 ± 243 not a linear variation between altitude and diversity individuals, 36 ± 17 species with DBH ≥ 10 cm per (Lovett et al., 2006; Ren et al., 2006). The authors hectare. In the same forest types in Tanzania Lovett et justify al., (2006) inventoried 700 to 1,000 stems per biogeographic theory of mountain suggesting that hectare for the same diameter but Cizungu (2015) isolated nature of mountain forests prevents the recorded 839 ± 154 individuals for DBH ≥ 6cm at frequent migration of species and the low balance Nyungwe in Rwanda and Masumbuko et al., (2012) provides limited support species (Lieberman et al., observed 568 individuals DBH ≥ 5 cm in the high 1996; Givnish, 1999). We observed the same situation altitude of the KBNP. The average is the same but the in this study; there are more species in lowland than variability observed can be explained by different upper altitude but against more individuals in methodological approaches. altitude. The altitude affects the trees growth; having this loss of diversity by the island fewer large trees for the highest areas could allow the The results showed, as in many cases, a decline of release of space and therefore the proliferation of diversity (Simpson, Fischer alpha, Pielou evenness, individuals at the expense of species immigration richness) with increase in altitude (Bruun et al., limitation (Vazquez and Givnish, 1998; Givnish, 145 | Imani et al. J. Bio. Env. Sci. 2016 1999). In the context of KBNP, search on some sarcochorie is predominant in the lower altitudes taxonomic groups confirm the altitude effect on (1250-1800) while in the higher elevations is the diversity. The richness of Rubiaceae decreases with ballochorie which becomes increasingly predominant altitude (Mwanga Mwanga et al., 2014); that of wood- for species. destroying fungi increases to around 2300 m before decreasing to 2600 m (Balezi, 2013). In the same Woody diversity generally decreases with elevation. case, Mangambu et al., (2013) noted the decrease in While species richness decreases with increasing the diversity of ferns with increase of the altitude. altitude, the abundance is positively correlated with This author demonstrated that factors such as this one increasing. This is due to the small size of temperature, slope, substrates and altitude, it is the trees in high altitude, freeing up space for the benefit latter which greatly influences the wealth of ferns. of other individuals on plot scale. To improve a better understanding of the ecosystem functioning in this Conclusion mountain Species distribution and forest typology of mountain management system. It is important to study the forest in the Congolese Albertine Rift precisely in the variability of the vegetation structure in connection Kahuzi Biega National Park and its surroundings was with this forest typology. forest and make a more effective investigated in this work using a hierarchical cluster associated with a correspondence analysis. Acknowledgement The present work has been financially supported by Woody diversity was assessed using the diversity the FCCC project (Appui à l'UNIKIS: "Forêts et indices. The objective of this study was to distinguish Changement Climatique au Congo") funded by the the different forest types in the mountain forests in European Commission, implemented by CIFOR in Kahuzi Biega National Park and its surroundings and partnership with University of Kisangani (UNIKIS) to show that these types differ more strongly when and RSD for logistical support. The authors also their altitudes are remote. In total 16 797 individuals thank all those who assisted in the collection and Dbh ≥ 10 cm, were surveyed in 30 ha. These processing of data. individuals are grouped into 212 specific taxa, 161 genera and 66 families. References Amani C, Nyakabwa M, Lejoly J. 2008. Analyse Four forest types were identified along the altitudinal floristique de la forêt de Bambous (Sinarundinaria gradient. These types are Sub-mountain (1250- alpina) du Parc National de Kahuzi-Biega. Annales 1500m), lower mountain horizon (1500-1800m), des Sciences et Sciences Appliquées, Université medium Officielle de Bukavu 1, 13-26. mountain horizon (1800-2400m) and mountain top horizon (2400-2600m). Each forest group is characterized by a specific number of Amsini F, Ilambu O, Liengola I, Kujirakwinja indicator species. D, Hart J, Grossmann F, Plumptre AJ. 2008. The Impact of civil war on the Kahuzi-Biega National The lower elevation (1250-1800m) has more indicator Park: Results of surveys between 2000-2008. Report, species than the upper one due to mountain effect Wildlife Conservation Society, 31 p. which limits the distribution of species. The altitude between 1800-2400 m presents peculiarities in the Balezi distribution and abundance following specific trait Hymenochaetales dans les forêts de montagne de each species and the location of the plot to the East or l’Est de la République démocratique du Congo : le cas West slope in the Park. The mode of dissemination by du Parc National de Kahuzi-Biega". Thèse de 146 | Imani et al. Z. 2013. Taxonomie et écologie des J. Bio. Env. Sci. 2016 doctorat, Université Catholique de Louvain, Belgique, Cuello N. 2002. Altitudinal changes of forest 223 p. diversity and composition in the Ramal de Guaramacal in the Venezuelan Andes. Ecotropicos 15, Balezi Z, Nyakabwa M, Lejoly J. 2008. Etude 160–176. écologique de la liane envahissante Sericostachys scandens dans la partie de haute altitude du Parc DeCáceres M. 2013. How to use the indicspecies National de Kahuzi – Biega (PNKB) Sud-Kivu, package (ver. 1.7.1). R Project, Centre Tecnologic R.D.Congo. Forestal de Catalunya, 29p. Annales des Sciences et Sciences Appliquées, Université Officielle de Bukavu 1, 1–12. Delnatte C. 2010. Le gradient altitudinal sur les Balezi Z, Shukuru B, Nyakabwa M. 2010. Etude sommets tabulaires de Guyane. Thèse de Doctorat, floristique et structurale du secteur Madiriri dans la Université des Antilles et de la Guyane, France, 292 p. partie de haute altitude du Parc National de KahuziBiega (PNKB), Sud-Kivu, R.D.Congo. Annales des Djomo Sciences et Sciences Appliquées, Université Officielle Gravenhorst G, 2010. Allometric equations for de Bukavu 2, 10-20. biomass estimations in Cameroon and pan moist AN, Ibrahima A, Saborowski J, tropical equations including biomass data from Basabose AK. 2002. Diet composition of chimpanzees inhabiting the montane forest of Kahuzi, Africa. Forest Ecology and Management 260, 1873– 1885. democratic republic of congo. American Journal of Primatology 58, 1–21. Faith DP, Minchin PR, Belbin L. 1987. Compositional dissimilarity as a robust measure of Brown O, Kasisi R. 2008. Renforcement de la ecological distance. Vegetatio 69, 57–68. capacité de l’ICCN à la gestion et la résolution des conflits lies aux ressources naturelles dans le Parc Fayolle A, Swaine MD, Bastin JF, Bourland N, National de Kahuzi-Biega. Report, International Comiskey J, Dauby G, Doucet, JL, Gillet, JF, Institute for Sustainable Development (IISD), 23 p Gourlet-Fleury S, Hardy OJ, Kirunda B, Kouamé FN, Plumptre AJ. 2014. Patterns of tree Bruun H, Moen J, Virtanen R, Grytnes J, species composition across tropical African forests. Oksanem L, Angerbjörn Journal of Biogeography 41, 2320–2331. A. 2006. Effects of altitude and topography on species richness of vascular plants , bryophytes and lichens in alpine Fischer communities. Journal of Vegetation Science 17, 37– BRYOTROP-Expedition to Zaire and Rwanda ( 46. Description of Collecting Sites . The Vegetation of E. 1993. Taxonomic results of the Kahuzi-Biega-National Park / Zaire , Nyungwe Forest Bussmann RW. 2006. Vegetation zonation and and Virunga volcanoes / Rwanda). Tropical Bryology nomenclature of African Mountains - An overview. 8, 13–37. Lyonia 11, 41–66. Fischer E. 1996. Die Vegetation Des Parc National Cizungu L. 2015. Dynamics and loss of reactive De Kahuzi-Biega, Sud-Kivu, Zaire. nitrogen in a central African tropical mountain forest Verlag. Stuttgart, 239 p. Franz Steiner and Eucalyptus plantation. PhD Thesis, Ghent University, Belgique, 176 p. Forsyth T. 1998. Mountain Myths Revisited: Integrating 147 | Imani et al. Natural and Social Environmental J. Bio. Env. Sci. 2016 Mountain Research and Development 18, 107–116. of woody species. Trends in Ecology & Evolution 24, 694–701. Gentry AH. 1988. Change in plant community diversity and floristic composition on environmental Karger DN, Kluge J, Krömer T, Hemp A, and geographical gradients. Annales of the Missouri Lehnert M, Kessler M. 2011. The effect of area on Botanical Garden. 75, 1–34. local and regional elevational patterns of species richness. Journal of Biogeography 38, 1177–1185. GIEC 2014. Changements climatiques 2014: Incidences, adaptation et vulnérabilité – Résumé à Körner C. 2007. The use of “altitude” in ecological l’intention des décideurs. Contribution du Groupe de research. Trends in Ecology & Evolution 22, 569– travail II au cinquième Rapport d’évaluation du 574. Groupe d’experts intergouvernemental sur l’évolution du climat. Organisation météorologique mondiale, Kumba S, Nshimba H, Ndjele L, Cannière C, Genève (Suisse), 34 p. De Visser M, Bogaert J. 2013. Structure spatiale des trois espèces les plus abondantes dans la Réserve Givnish TJ. 1999. On the causes of gradients in Forestière tropical tree diversity. Journal of Ecology 87, 193– Démocratique du Congo. Tropicultura 31, 53–61. de la Yoko, Ubundu, République 210. Lande R. 1996. Statistics and Partitioning of Species Guillaumet J. 2009. La végétation des montagnes Diversity, du Rift. In: Le Rift Est Africain. Editions Institut de Communities. Oikos 76, 5–13. and Similarity among Multiple recherche pour le développement (IRD) ISBN : 9782-7099-1663-9, p 195–204. Lee CB, Chun JH, Ahn HH. 2014. Elevational patterns of plant richness and their drivers on an Habiyaremye MK. 1995. Etude phytoécologique de Asian mountain. Nordic Journal of Botany 32, 347– la dorsale orientale du lac Kivu (Rwanda). Thèse de 357. doctorat, Université Libre de Bruxelles, Belgique, 276 p. Lenoir J. 2016. Correspondances Analyse appliquée à Factorielle des données des de Hardy OJ, Sonké B. 2004. Spatial pattern analysis présence/absence. Notes de travail. Université de of tree species distribution in a tropical rain forest of Picardi, France. 26 p Cameroon: Assessing the role of limited dispersal and niche differentiation. Forest Ecology and Management 197, 191–202. Lepš J, Šmilauer P. 1999. Multivariate Analysis of Ecological Data, Faculty of Biological Sciences, Notes of Course, University of South Bohemia, République Henry M, Besnard A, Asante W., Eshun J, tchèque. 110 p Adu-Bredu S, Valentini R, Bernoux M, SaintAndré L. 2010. Wood density, phytomass variations Lescuyer G, Locatelli B. 1999. Rôles et Valeurs des within and among trees, and allometric equations in a forêts tropicales dans le changement climatique. Bois tropical rainforest of Africa. Forest Ecology and et Forêts des Tropiques 260, 5–18. Management 260, 1375–1388. Lieberman D, Lieberman M, Peralta R, Jump AS, Mátyás C, Peñuelas J. 2009. The Hartshorn G. 1996. Tropical forest structure and altitude-for-latitude disparity in the range retractions composition on a large scale altitudinal gradient in 148 | Imani et al. J. Bio. Env. Sci. 2016 Costa Rica. Journal of Ecology 84, 137–152. Woody climbing plants influence the structure of the mountain forest in the Kahuzi-Biega National Park , Linder HP. 2001. Plant diversity and endemism in DR Congo. Regional environnemental Changes 12, sub-Sahara tropical Africa. Journal of Biogeography 951–959. 28, 169–182. Meddour R. 2011. La méthode phytossociologique Locatelli B, Kanninen M, Brockhaus M, Colfer sigmatiste ou Braun-Blanqueto-Tuxenienne. notes de CJ, Murdiyarso D, Santoso H. 2008. Face à un cours. Université Mouloud Mammeri, Algérie, 40 p avenir . incertain: populations Comment peuvent les s’adapter forêts au et les changement Medley KE, Maingi JK. 2014. Biogeographic climatique. Regards sur la forêt No 5 CIFOR, Bogor, Patterns of Forest Diversity at Mount Kasigau , Indonésie. 86 p Kenya. Journal of East African Natural History 103, 1–24. Lovett JC, Marshall AR, Carr J. 2006. Changes in tropical forest vegetation along an altitudinal Moonen, gradient in the Udzungwa Mountains National Park, Boyemba F, Muys B. 2014. Accurate assessment of Tanzania. African Journal of Ecology 44, 478–490. Congo basin forest carbon stocks requires forest type PCJ, Ballaert SV, Verbist B, specific assessments. Geophysical Research 16, 75-75. Magurran AE. 2004. Measuring biological diversity. Blackwell Publ. Malden, Oxford Victoria, Mutke J, Sommer JH, Kreft H, Kier G, UK. ISBN 0-632-05633-9. 70 p. Barthlott W. 2011. Vascular Plant Diversity in a Changing World : Global Centres and Biome-Specific Mangambu MJ. 2013. Taxonomie, biogéographie et Patterns. Biodiversity Hotspots 1, 83-96. écologie des Ptéridophyes de l’écosystème forestier des montagnes du Parc National de Kahuzi-Biega à Mwanga Mwanga IJ, Wabika DP, Imani G, l'Est de la R.D.Congo. Thèse de Doctorat, Université Balezi d’Anvers/Belgique, 463 p. Rubiaceae dans les forêts de montagnes à l’Est de la Z. 2014. Distribution altitudinale des R.D. Congo. Le cas du Parc National de Kahuzi-Biega Mangambu MJ, Habiyaremye M, Janssen T, et la Forêt communautaire de Rwaga-Nirindja. Van Diggelen R, Robbrecht E, Ntahobavuka Continental Journal Applied Sciences 9, 1–9. HH. 2013. Diversité des Fougères et leurs alliées le long du gradient altitudinal au sein de l'écosystème Myers N, Mittermeier RA, Mittermeier CG, forestier des montagnes du Parc National de Kahuzi- Fonseca GA, Kent J. 2000. Biodiversity hotspots Biega for conservation priorities. Nature 403, 853–858. (RD CONGO). International Journal of Environnemental Studies 70, 259–283. Oksanen J. 2014. Cluster analysis: tutorial with R. Masumbuko N. 2011. Ecologie de sericostachys Working paper. University of Oulu, Finlande.13p. scandens, liane envahissante dans les forêts de montagne du Parc National de Kahuzi-Biega, République Démocratique du Congo. Thèse de Doctorat, Université Libre de Oksanen J. 2015. Vegan : ecological diversity. Working paper. University of Oulu, Finlande.12 p. Bruxelles, Belgique, 192 p. Opler P, Baker H, Frankie G. 1980. Plant reproductive Masumbuko N, Habiyaremye M, Lejoly J. 2012. succession 149 | Imani et al. characteristics in Neotropical during secondary Lowland Forest J. Bio. Env. Sci. 2016 Ecosystemes. Tropical succession 12, 40–46. Sharma N, Raina A. 2012. Composition , structure Pierlot R. 1966. Structure et composition de forest and diversity of tree species along an altitudinal denses d’Afrique Centrale, specialement celles du gradient in Jammu province of north-western Kivu. Académie royale de Bruxelles, Belgique. 373p. Himalayas , Jammu and Kashmir , India. Journal of Biodiversity and Ecological Sciences. 2, 86-93. Plumptre A, Davenport T, Behangana M, Kityo Tassin J, Derroire G, Rivière J. 2004. Gradient R, Eilu G, Ssegawa P, Ewango C, Meirte D, altitudinal de la richesse spécifique et de l’endémicité Kahindo C, Herremans M, Peterhans J, de la flore ligneuse indigène à l'île de La Réunion Pilgrim J, Wilson M, Languy M, Moyer D. (archipel des Mascareignes). Acta Botanica Gallica 2007. The biodiversity of the Albertine Rift. Biological 151, 181-196. Conservation 134, 178–194. Terborgh J, Andresen E. 1998. The composition Plumptre A, Kujirakwinja D, Nampindo S. of Amazonian forests: patterns at local and regional 2009. Conservation of landscapes in the Albertine scales. Journal of Tropical Ecology 14, 645–664. Rift. Working Paper N-36 Word Conservation Society. 7 p. Toledo M, Peña-Claros M, Bongers F, Alarcón A, Balcázar J, Chuviña J, Leaño C, Licona JC, Ren HB, Niu SK, Zhang LY, Ma KP. 2006. Poorter L. 2012. Distribution patterns of tropical Distribution of vascular plant species richness along woody species in response to climatic and edaphic an elevational gradient in the Dongling Mountains, gradients. Journal of Ecology 100, 253–263. Beijing, China. Journal of Integrative Plant Biology 48, 153–160. Vazquez AJG, Givnish TJ. 1998. Altitudinal Gradients in Tropical Forest Composition, Structure, Roux G, Roux M.1967. A propos de quelques and Diversity in the Sierra Altitudinal gradients in méthodes de classification en phytosociologie. Revue tropical forest composition, structure, and diversity in de Statistique Appliquée 2, 59–72. the Sierra de Manantlan. Journal of Ecology 86, 999– 1020. Sanchez-Gonzalez A, Lopez-Mata L. 2005. Plant species richness and diversity along an altitudinal Yamagiwa J, Basabose AK, Kaleme K, Yumoto gradient in Sierra Nevada, Mexico. Diversity and T. 2005. Diet of Grauer’s Gorillas in the Montane Distributions 11, 567–575. Forest of Kahuzi, Democratic Republic of Congo. International Journal of Primatology 26, 1345–1373. Senterre B. 2005. Recherches méthodologiques pour la typologie phytogéographie de des la forêts végétation denses et la Yumoto T, Yamagiwa J, Mwanza N, d'Afrique Maruhashi T. 1994. List of Plant Specles ldentified tropicale. Thèse de Doctorat, Université libre de in Kahuzi-Biega National Park, Zaïre. Tropics 3, 295– Bruxelles, Belgique.116p. 308. 150 | Imani et al.