Abstract: Numerical taxonomic study of eleven Combretum Loefl. species common in southeastern Nigeria was carried out based on morphological and anatomical characters. PAST software was used to calculate the Euclidean distances and the coefficients were clustered by unweighted pair-group clustering method (UPGCM). From the 107 characters employed in the numerical analysis, the species were separated into two broad clusters (C1 represented by C. zenkeri and C2-the remaining 10 species). The clustering further suggested that there were groups within the genus hence C2 separated into two subgroups with C. glutinosum, a lone sub-cluster and the other nine species, which separated into four other sub-clusters. The similarity coefficient among the Combretum species studied ranged from 4.69 between C. mooreanum and C. bracteatum to 8.72 between C. glutinosum and C. zenkeri. Generally, results obtained favour species arrangement and clustering based on phylogenetic relationship as in C. mooreanum, C. bracteatum and C. platypterum and habitat preference. The unknown species, Combretum sp.1 is more closely related to C. bracteatum than any other species studied. However, C. constrictum, C. dolichopetalum, C. glutinosum, C. hispidum, C. paniculatum, C. platypterum, C. racemosum and C. zenkeri have been confirmed as individual taxa based on previous classifications. Though each Operational Taxonomic Unit (OTU) is confirmed species, we assert that the closeness observed amongst the OTUs imply that introgression as a mechanism of speciation in the wild, is at play in the genus.
INTRODUCTION
Combretaceae is family of plant with 20 genera and 600 species, of tropical and subtropical regions of the world (Gill, 1988). In West Africa, the family Combretaceae is represented by 9 genera with 72 species and the genus Combretum Loefl. is the largest genus with 49 species (Gill, 1988). Keay (1989) reported 25 species of the genus Combretum Loefl. which are mainly straggling shrubs or lianes in Nigeria. Hutchinson and Dalziel (1954) showed that a number of species are indeterminate and occur in the South-Eastern Nigeria. The ethnopharmacological importance which include anti-inflammatory, anthemintic, anti-biharzia (anti-schistosomal), treatment of malaria, pain, dermatitis, diarrhea, Pneumonia, gonorrhea, syphilis, hypertension and cancer have been reported (Coulidiati et al., 2009; Fyhrquist et al., 2002; Baba-Moussa et al., 1999; Simon et al., 2003; Martini et al., 2004; Batawila et al., 2005).
The existing information on these species in Nigeria is strictly based on morphological descriptions by Hutchinson and Dalziel (1954). These information were however, sourced from herbarium specimens. Literature search revealed neither recent description nor taxonomic treatment of Nigerian members of this genus. Hutchinson and Dalziel (1954) reported indeterminate species and noted that much taxonomic gaps needed to be filled with the identified species in tropical West Africa. This underscores the need for a study approach that utilizes different taxonomic lines of evidence to confirm the identity of species in southeastern Nigeria and other parts of tropical West Africa. In the wild forest of southeastern Nigeria were the species occur, there are conflicting identities with morphological signs of introgression. Thus the study is aimed at resolving difficulties with identifying taxonomic species from this section of Nigeria. It seeks to support the great taxonomic work of Hutchinson and Dalziel (1954), which unarguably is the taxonomic Bible of tropical West Africa. The numerical taxonomy approach adopted for this study has been applied in the elucidation of complex genera by Sornsuwan et al. (2006), Khalik et al. (2008), Dogan et al. (2009) and Osman et al. (2011).
MATERIALS AND METHODS
Plant material: Different specimens of Combretum were collected from southeastern part of Nigeria (Fig. 1) between January 2009 and December 2011. By painstaking reference to the Flora of West Tropical Africa (Hutchinson and Dalziel, 1954), these specimens were sorted into ten known taxa and an indeterminate species (Table 1).
| Table 1: | Species name, locality, coordinates and date of collection of Combretum species studied |
Further authentication of the taxonomic identity of the species was achieved by sending representative samples of each identified taxon to Forestry Herbarium Ibadan (FHI). These specimens were compared with species from the study area deposited at FHI. Morphological and epidermal studies were carried out on these species and characters and their character states (Table 2) deduced from the study were used for numerical analysis. Taxa were deposited at the University of Port Harcourt Herbarium (UPH).
Character observation: In this study, the species were treated as the Operational Taxonomic Units (OUT). Eleven OTUs (Table 1) and one hundred and seven (107) characters derived from the morphology (anther colour, flower colour, stem colour, colour of receptacle plus calyx tube, stem hair, habit, number of wings, average leaf length, average leaf breath, average petiole length, average internode length, leaf type, leaf apex, leaf base, leaf shape, leaf surface and leaf arrangement) and anatomy (crystal type, crystal distribution, number of palisade layers, No. of spongy layers, stomata types/complex/occurrence, trichome type, anticlinal cell wall pattern and epidermal cell wall shape) (Table 2) of the species were used in constructing the dendrogram. These characters were selected without prejudice. Ten samples of each authenticated OUT were either qualitatively observed or quantitatively scored to deduce the results summarized in Table 2.
Numerical analysis: Numerical taxonomic analysis was based on Sneath and Sokal (1973). The hierarchical cluster analysis for qualitative and quantitative characters was performed using binary Euclidean distance; the dendrogram was drawn using average distance linkage (between groups). The similarity coefficient/Euclidean distance were measured using PAST and clustered by Unweighted Pair-group Clustering Method (UPGCM). Data generated from the different morphological and anatomical characters were scored and converted to binary No. 1 or 0 i.e., present or absent and fed into a computer which generated clusters of the OTUs according to their overall similarity/Euclidean distance using PAST software. Dendrogram was constructed, which represented the diagrammatic illustration of the relationship among the species based on their degree of similarity in which the OTUs are linked together at various levels of resemblance.
RESULTS
The characters and character states generated and used for the cluster analysis is presented in Table 2. From the 107 characters employed in the numerical analysis, the dissimilarity coefficients among the Combretum species studied ranged from 4.69% between C. mooreanum and C. bracteatum to 8.72% between C. racemosum and C. zenkeri (Table 3). At 11.54% dissimilarity as shown in Fig. 1, the OTUs separated into two broad clusters. Cluster C1 contains only one species, C. zenkeri while cluster C2 contains 10 species. However at 11.11% dissimilarity, cluster C2 further separated into two sub-clusters (C3 with only C. glutinosum and C4). Sub-cluster 4 gave rise to C5 and C7 each with two species and C6 with five species (C. mooreanum, C. paniculatum, C. bracteatum, C. platypterum and Combretum sp.1) as shown in Fig. 1.
The unknown species Combretum sp.1 is closest to C. bracteatum and together with C. mooreanum (all three in C9) occur in marshy environment. C. mooreanum, C. bracteatum and C. platypterum in C6 are phylogenetically related in Hutchinson and Dalziel (1954) classification scheme.
| Table 2: | Character and character states used for cluster analysis of Combretum species |
| Table 3: | Range of coefficient of dissimilarity matrix among the eleven Combretum species |
| 1: C. bracteatum (Laws.) Engl. and Diels., 2: C. constrictum (Benth.) Laws, 3: C. dolichopetalum Engl. and Diels., 4: C. glutinosum Perr. ex DC., 5: C. hispidum Laws, 6: C. mooreanum Exell., 7: C. paniculatum Vent., 8: C. platypterum (Welw.) Hutch. and Dalz., 9: C. racemosum P. Beauv., 10: C. zenkeri Engl. and Diels. and 11: Combretum sp.1 | |
DISCUSSION
The close morphological similarities of members of the genus Combretum in Nigeria oftentimes lead to confusion in species identification and collection. Existing taxonomic information on the genus in the country is based on morphological characters documented in Hutchinson and Dalziel (1954), which reported some indeterminate species. Reassessment of the phylogenetic relationship of members of this genus is therefore worthwhile. We have begun this by subjecting species within the southeastern region of the country to numerical analysis and phylogenetic interpretation.
Dendrogram based on morphological and anatomical information grouped these eleven southeastern species into two major clusters C1 and C2. Cluster C1 is distinct containing only C. zenkeri. This indicates that C. zenkeri is distinct from all the other southeastern species studied. The remaining ten species clustered in C2. The unknown species Combretum sp.1 is found midway among these other species. In cluster C2, C. glutinosum is distinct among the species in this cluster. C. racemosum and C. dolichopetalum are closely related supporting Hutchinson and Dalziel (1954). The relative closeness between these species is attributed to their similar anatomical features. However, C. hispidum and C. dolichopetalum did not follow the sequence.
| Fig. 1: | Dendrogram for the hierarchical cluster analysis of 11 southeastern Nigeria Combretum species using Euclidean distance among the taxa |
The divergence between these two species is as result of their anatomical differences. Similarly, C. zenkeri and C. glutinonum are the most diverged species in the dedrogram. On the other hand, the unknown species Combretum sp.1 is more closely related to C. bracteatum and C. paniculatum than any other species in the dendrogram. Based on their percentage dissimilarity, Combretum sp.1 is a new unnamed species but more closely related to C. bracteatum.
It is important to reemphasize the fact that current information is lacking on members of the genus Combretum in Nigeria. A closer comparative assessment of the results obtained from this study with existing information in Hutchinson and Dalziel (1954) confirm the phylogenetic relationship between C. mooreanum, C. bracteatum and C. platypterum. In fact, C. bracteatum is phylogenetically next to C. mooreanum in Hutchinson and Dalziel (1954) classification key. These species also have similar habitat preference occurring in marshy environments. C. constrictum, a low shrubby liana also prefers damp habitats. However, C. glutinosum, which has been previously reported as savannah small tree, was collected within the region. So also was C. paniculatum, which reportedly thrive at the extreme of the forest zone into the savannah region of the country. These may not be unconnected with human dispersal of reproductive structures.
The close affinity observed in morphological and anatomical characters employed in this study across species is worthy of note. These observations were highest amongst members in C9 (C. mooreanum, C. bracteatum and Combretum sp.1). We therefore assert that introgression as a mechanism for speciation is at work on members of the genus Combretum in southeastern Nigeria. Further work is ongoing for the authentication of this assertion.