Abstract: The diversity and distribution of climbing plants were investigated in two 1-ha plots in the KNUST botanic garden. Each plot was divided into sixteen 25x25 m quadrats and all trees and shrubs ≥10 cm dbh were enumerated in each quadrat. All vines on tree species were identified and counted while all lianas ≥1.5 cm were enumerated. The mechanisms of climbing of the climbers were studied. A total of 951 climbing plants belonging to 82 species, 49 genera and 26 families were identified. These consisted of 72 liana and 10 vine species. Seven hundred and sixteen trees belonging to 77 species and 27 families hosted the climbers. Climbing plants made up of about 43% of all the species (except herbs) identified. Lianas alone constituted about 38% of the woody species. The commonest and most dominant species were Strophantus barteri Franch., Smilax species and Griffonia simplicifolia (Vahl ex DC.) Baillon. Most of the climber species showed clumped distribution. Stem twinning was the most predominant climbing mechanism. The number of climbers using the different climbing mechanisms differed significantly (p<0.001). Host species, habitat and their interaction had significant (p<0.001) effects on climber density.
INTRODUCTION
Climbers are plants that germinate on the floor of the forest and grow, at least for part of their life, or when the forest closes up around them, by winding around, leaning on, or anchoring or adhering to other plants (Jongkind and Hawthorne, 2005) to attain great stature (Swaine et al., 2005).
They occur in many plant families with only a few families such as Dioscoreaceae, Cucurbitaceae and Convolvulaceae consisting completely of climbing plants. In Ghana, out of the 125 families of vascular plants in the forest flora, 66 have climbing species (Swaine et al., 2005). The most species rich climber families are Rubiaceae, Leguminosae, Celestraceae and Apocynaceae, with each family containing more than 50 climber species (Gentry, 1991; Schnitzer and Bongers, 2002). Climbing plants also show great diversity in their climbing mechanisms (Bongers et al., 2005; Jongkind and Hawthorne, 2005). These include stem twiners, branch twiners, root/adhesive climbers, tendril climbers, scramblers and hook/thorn climbers.
Climbers occur in all woody ecosystems of the world although a high abundance is considered to be characteristic of tropical and subtropical forests (Bongers et al., 2005). Specifically, in tropical rain forest, they comprise about 25-30% of species diversity (Schnitzer and Bongers, 2002). They occur in between crowns and many grow on several canopies (Caballé, 1998). In most tropical forests many more small climbers and relatively few large ones are present at forest edges and in forest fragments compared to forest interiors and large sized forests (Schnitzer and Bongers, 2002). Tendril climbers are more suitable to gaps and forest edges, where smaller diameter supports are more common, than in forest interiors (Putz and Holbrook, 1991). They are therefore more randomly distributed in young forests during which time small diameter supports are available. Stem and branch twiners on the other hand, are more commonly evenly distributed in later successional forests (Dewalt et al., 2000). The spatial distribution pattern of plants affects future processes, both of plants themselves and of a range of other organisms with which they interact (Parren, 2003). Most climbing plants as a group or at a species level show clump distribution at different levels.
Few studies on climbers such as Kokou et al. (2002) have distinguished three categories of climbing plants namely woody climbing plants (lianas), herbaceous climbing plants (vines) and climbing shrubs. However, most authors (e.g., Caballé, 1998; Muoghalu and Okeesan, 2005) consider the climbing shrubs as lianas and therefore recognize lianas and vines as the two main groups of climbers. Lianas may reach the crowns of trees in the top strata and be exposed to sunlight when they reach maturity but vines are unable to reach anywhere near the canopy of mature forest (Jongkind and Hawthorne, 2005).
Climbers play important ecological roles in the forest ecosystem dynamics and functioning (Nabe-Nielsen, 2001; Bongers et al., 2002). They contribute substantially to canopy closure after tree fall and help stabilize the microclimate underneath (Schnitzer and Bongers, 2002). Lianas in particular add considerably to forest plant diversity and provide valuable habitat and connections among tree canopies that enable arboreal animals to traverse the tree tops (Schnitzer and Carson, 2001). The contribution of lianas to species richness of tropical forests can be impressively as high as 31% (Hall and Swaine, 1981). While lianas contribute substantially to the forest structure, the contribution of vines is only negligible due to the small number of species composing it (Muoghalu and Okeesan, 2005).
Climbers form an essential part of the diet of many animals in times of scarcity of flowers and fruits (Putz and Windsor, 1987). This is critical for the survival of tress as many of these animals are essential for dispersal of trees seeds (Guariguata and Pinard, 1998), including the majority of commercially interesting species (Jansen and Zuidema, 2001). In spite of the numerous roles climbers play in ecosystems, little attention has been given to them; they are scantly treated in literature (Bongers et al., 2005). Almost all work on forest plant communities have over relied heavily on trees (Turner et al., 1996) probably due to the commercial value of many trees among other reasons (Bongers et al., 2005). For instance, unlike trees and shrubs which have received much attention in the KNUST Botanic Garden (Anning et al. unpublished data) climbers have not been much studied. Only a few studies on climber species have been carried out in Ghana (e.g., Hall and Swaine, 1981; Swaine et al., 2005). Furthermore, most of these studies are not exclusively on climbers as they involve general botanical surveys, leading to scanty information on climbers. Accordingly, climbing plant diversity and distribution in Ghana is still poorly known. The main objective of the present study was to determine the diversity and distribution of climbing plants in the KNUST Botanic Garden as a way of contributing to the understanding of the general floristic composition and structure of the garden.
MATERIALS AND METHODS
Study area: The study took place in the Kwame Nkrumah University
of Science and Technology (KNUST) Botanic Garden, Kumasi, Ghana (Fig.
1) (latitude 6°35 N-6°40 N and longitude 1°30 W-1°35
W). The 12.9 ha garden comprises of a semi-deciduous rain forest. It was
established in 1960 and since then has played very useful roles in education,
research and recreation.
| Fig. 1: | Map of a section of the Kwame Nkrumah University of Science and
Technology (KNUST) showing the Botanic garden with the study sites
(disturbed and undisturbed) |
some measures have been instituted to prevent people from encroaching on the garden. These include fencing of the whole garden and the provision of security men who guard the garden. In spite of this, some people still manage to sneak in and conduct illegal activities such as felling of trees for firewood, cutting of bamboo and hunting for animals in some parts of the garden. These areas together with the part of the garden that has been developed for recreational activities have thus experienced minimal form of human disturbance. The rest of the garden which forms the major part has remained undisturbed for many years.
The garden is endowed with diverse species including tropical palms, timber species and medicinal plant species. The vegetation is made up of about 68% native species and 32% exotic species (Anning et al., unpublished data). Detailed description of the floristic composition and structure of the KNUST Botanic garden is found in Anning et al. (unpublished data), but with scanty information on climbers.
The semi-deciduous vegetation receives relatively high amount of annual rainfall, approximately 731 mm per year. The average annual temperature ranges from 21.55 to 32.12°C. The average annual humidity in the area is 599.2% (The KNUST meteorological department, 2006).
Sampling: Quantitative inventory of climbers was carried out between August 2006 and January 2007, in two 1 ha plots; one in the disturbed area and the other in the undisturbed area. Each plot was further divided into sixteen 25x25 m subplots (quadrats). The plots were demarcated with the aid of a field compass and flagged pegs. Within each subplot all tree species ≥10 cm diameter at breast height (dbh) were identified. All vines on a tree species were identified and counted, while all lianas ≥1.5 cm dbh were enumerated. The diameter at breast height of trees and lianas was measured at 1.3 m from their rooting base. Each identified tree, vine, or liana was tagged to prevent double enumeration. Shrubs (≥10 cm dbh) were also identified in the survey so as to be able to determine the relative composition of lianas among the woody species. All stems that were rooted at the same place were counted as an individual (Nabe-Nielsen, 2001). All shoots connected by a single runner were regarded as an individual. The frequency of the climbers was quantified by recording the presence or absence of the species in the quadrats (Green et al., 1995) and finding the average. The climbing mechanisms of the climber species were determined based on observations on the field and with reference to checklist by Jongkind (2005). Identification of plants was carried out with the help of plant taxonomists and with reference to relevant local manuals and Floras (Hawthorne, 1990; Arbonnier, 2004; Poorter et al., 2004). Voucher specimens were deposited at the KNUST and Forestry Department herbaria, Kumasi. The taxonomy followed Hutchinson and Dalziel (1994).
Analysis of data: The diameter at breast height (dbh) measured for the trees and lianas were used to determine the basal area. The basal area of the lianas was used as a component part of the determination of importance value. The Importance Value (IV) of each species was calculated as the sum of the species relative density, relative frequency and relative dominance (Kiruki and Njung`e, 2007).
Analysis of variance (ANOVA) was performed on the data to determine the effects of host species, habitat and the interaction between them on climber density. Furthermore, climber densities among the different categories of climbing mechanisms were compared using analysis of variance. Analysis of variance was also performed to determine significant differences between liana densities in the various diameter classes. The ninth edition of the GENSTAT software was used. All analyses were conducted at a significance level of 5%.
Spatial distribution patterns of the climbers were determined using the Morisita index of dispersion (Morisita, 1959). The diversity of climbers in the study sites were quantified using the Shannon-Wiener index and the Shannon evenness.
RESULTS
Diversity of climbing plants: A total of 951 individuals of climbing plants were identified in the study (Table 1). These belong to 82 species, 49 genera and 26 families. The climber species were made up of 72 liana (woody) species and 10 vine (non-woody) species. Seventy two of the species were identified to the species level, 17 were identified to the genus level and 3 were identified to the family level. The density of climbers was 475.5 individuals per hectare while species richness was 41 per hectare. The Shannon-Wiener index for the disturbed site was 3.6 while the undisturbed site recorded an index value of 3.8. The Shannon-Wiener evenness for the disturbed and undisturbed sites were 0.95 and 0.91, respectively.
Strophantus barteri was the most abundant species making up about
11.78% of all the climber stems and 10.41% of the total basal area of
the lianas. The ten most abundant species accounted for 41% of all the
individual climbers identified in the study (Table 1).
Strophantus barteri also had the highest important value index
followed by Smilax species and Griffonia simplicifolia
(Table 1).
| Table 1: | Diversity and distribution of climbing plants recorded in the KNUST Botanic garden |
| *: Relative frequency, relative density and relative dominance of lianas used in calculating the importance value are not included in the table | |
| Fig. 2: | Contribution of families to climber diversit |
he most diverse families in terms of species richness were Fabaceae (17.1%), Apocynaceae (14.6%), Celestraceae (7.3%), Combretaceae (7.3%), Icacinaceae (6.1%) and Araceae (4.9%), while the most dominant families in terms of number of individuals were Apocynaceae (24.3%), Fabaceae (11.4%), Liliaceae (7.7%), Icacinaceae (7.0%) and Araceae (5.7%) (Fig. 2).
The climber species were hosted on 513 individuals distributed in 77 species, 60 genera and 27 families (Table 2). Hevea brasiliensis supported the highest number of climber species as well as individuals. There were 37 climber species and 286 individuals. Other species of host with high density of climber species were P. microcarpa, C. procera, E. guineensis and F. africana. Host species with the least number of climbers were B. unijugata, D. monbuttensis, C. gigantea, F. exasperata, M. species and X. aethiopica.
There were significant effects of host species (F = 32.20; df = 81; p<0.001), site (F = 195.04; df = 1; p<0.001) and the interaction between host species and site (F = 15.20; df = 81; p<0.001) on the density of climbing plants. Host species alone accounted for more than half (62%) of the variation.
There were eight climbing mechanisms used by the climbing plants (Table
3). More than half (62.5%) of the climber individuals twined around
the host plants whereas a few individuals used leaf tendrils (0.7%) and
hooking (0.3%). Two species, namely C. hispidum and
Smilax species combined twining and thorns. There were significant
differences in the number of climbers using the various climbing mechanisms
(F = 149.25; df = 7; p<0.001).
| Table 2: | Species richness and density of climber species on host trees |
| Table 3: | Frequency of climbing mechanisms used by climbing species |
| Table 4: | Distribution of liana individuals in various diameter classes (≥1.5 cm) |
Distribution of climbing plants: Strophantus barteri, Smilax species and G. simplicifolia were the commonest climber species in all the sites studied (Table 1). S. barteri was present in 94% of all the quadrats studied. The least frequent species were A. multiflora, C. producta, B. breviloba, M. chrysophylla and R. beninensis. Most of the climber species (about 82%) showed clumped dispersion whereas 16% showed uniform dispersion. Only two species (2%) namely Combretum species and L. owariensis exhibited random dispersion.
More than one-third of all the liana individuals were distributed in the lower diameter class of 1.5-3.5 cm while about one-third of the individuals occurred in the 5.7-7.7 cm diameter class (Table 4). Close to one-fourth of the liana individuals occurred in the 3.6-5.6 cm diameter class. The remaining individuals occurred in the 7.8-9.8 cm and >9.8 cm diameter classes. The largest individual had a diameter of 25 cm. There were significant differences (F = 1718.61; df = 4; p<0.001) between the numbers of lianas in all the diameter classes.
DISCUSSION
Diversity of climbing plants: Species richness of climbers recorded in this study was relatively high as compared with some studies conducted in the tropics. Muthuramkumar and Parthasarathy (2001) for instance recorded 75 liana species in a 30 ha plot at Varagaliar, India while Campbell and Newbery (1993) recorded 40 liana species in a 4 ha plot at lowland Dipterocarp forest, Malaysia. Therefore, 82 climber species obtained in this study is fairly high, considering the use of 2 ha plot and the fact that only lianas of ≥1.5 cm dbh were counted. Few studies such as Dewalt et al. (2000), Nabe-Nielsen (2001) and Senbeta et al. (2005) however, have identified higher number of species than the number (82 species) obtained in the study.
Climber species featured prominently in the plant community of the KNUST botanic garden, forming about 43% of all the species identified (excluding herbs) in this study while lianas alone constituted about 38% of the woody species. This is comparable to the results of Kouame (unpublished data) and Kuzee (unpublished data) in which lianas formed 30 and 40% of the woody species, respectively.
Density of lianas (woody climbers) in the KNUST Botanic Garden compares favourably with the results of many works conducted in other tropical forests including a tropical evergreen forest at Varagalaiar, Anamalais, India (Muthuramkumar and Parthasarathy, 2001) and a lowland tropical rainforest in Southern Cameroon (Parren, 2003). Climber density was higher in the undisturbed site than the disturbed site. This is supported by the higher Shannon-Wiener index recorded for the undisturbed site. This finding is inconsistent with some studies in which disturbance of the forest rather favoured climber density (Hegarty and Caballé, 1991; Bongers et al., 2005). The lower climber density in the disturbed site reflects the level of human influence. Part of the disturbed site is made up of H. brasiliensis plantation which has other trees and shrubs sparsely scattered within it. The establishment of the monoculture of H. brasiliensis meant that other trees would necessarily be excluded. This may not only explain the low diversity of host tress in the disturbed portion of the garden but also explain why H. brasiliensis supported the greatest number of climber species and individuals. The lower diversity of host species in the disturbed site might have affected climber density, as host species influence climber abundance. The high abundance of vines and small size lianas in the disturbed site, reflects the high level of disturbance (Swaine et al., 2005) and openness (Kokou and Caballé, 2005) and the negative effects of deforestation on the abundance of large lianas (Gardette, 1998) in the area. This explains why habitat had a significant effect on the density of climbers. Host species also had a significant effect on the abundance of climber species in the KNUST botanic garden. Some species such as H. brasiliensis, P. microcarpa, C. carapa, E. guineensis and F. africana hosted great numbers of climbers while other species such as B. unijugata, C. gigantea and X. aethiopica hosted a few climber individuals. This shows that some tree species are more susceptible to climber infestation than others. This is in agreement with other studies such as Muoghalu and Okeesan (2005) and Putz and Chai (1987).
The level of contribution of a family to the host species diversity did not have much effect on climber abundance. Although Anacardiaceae and Arecaceae did not contribute much to host species diversity, they hosted more climber individuals than Moraceae, Euphorbiaceae, Apocynaceae and Sterculiaceae which were more diverse. Moreover, about 84.4% of the climber individuals hosted by members of the family Rubiaceae occurred on only one species (H. brasiliensis), confirming the observation that climber density on the hosts depends more on the host species rather than on the family (Campbell and McC Newbery, 1993; Muoghalu and Okeesan, 2005).
The three most abundant species (S. barteri, Smilax species and G. simplicifolia) made up about 22.54% of all the climber stems. The dominance of these species is confirmed by their high important values (32.63, 21.10 and 20.12, respectively), indicating their wider range of ecological adaptation and dispersal modes (Senbeta et al., 2005).
Family dominance was observed among the climbers with the families Fabaceae, Apocynaceae, Celestraceae and Combretaceae being the most important families in terms of species richness. This result is similar to the work of Muoghalu and Okeesan (2005). It is however, not the case with Makana et al. (1998) in which Connaraceae, Dichapetalaceae, Loganiaceae, Euphorbiaceae and Hippocrateaceae were the most dominant families.
Climbers in this study showed great diversity in climbing mechanisms which is higher than many other comparable studies such as Dewalt et al. (2000), Muthuramkumar and Parthasarathy (2000) and Senbeta et al. (2005). Twining around the host predominated among all the climbing mechanisms, with about 63.40% of the species using this mechanism. Similar results have been recorded by several authors (Putz and Chai, 1987; Gentry, 1991; Muthuramkumar and Parthasarathy 2000; Senbeta et al., 2005). There were significant differences in the density of climbers among the different category of climbing mechanisms indicating their importance in the abundance, diversity and distribution of climbers (Nabe-Nielsen, 2001).
Vines formed about 12% of all climber species in the KNUST botanic garden. This confirms the widely held view that vines are much less important than lianas due to the small number of species composing it and their negligible contribution to the forest structure (Muoghalu and Okeesan, 2005). It is however, at variance with the finding of Kokou and Caballé (2005) in which vines were comparatively higher than lianas in forest fragments in Togo.
Distribution of climbing plants: Like many other studies (Parren, 2003), most climbers in the KNUST botanic garden showed clumped distribution. Studies have shown that clonal growth by ramets is quite common by climbers and may account for the clumping patterns of climbers (Caballé, 1994).
Strophantus barteri, Smilax species and G. simplicifolia were the most frequent species in all the plots studied. In the case of G. simplicifolia,earlier study by Hall and Swaine (1981) showed that it was the commonest climbing plant in Ghana, as it was found in all forest types except the upland evergreen forests. G. simplicifolia may also be found in almost any small patch of a forest when it is present (Swaine et al., 2005). These three species were common in both the disturbed and undisturbed sites implying their wider range of ecological adaptation (Senbeta et al., 2005). Their distribution may also suggest that they have efficient modes of seed dispersal, because plant distribution is partly influenced by the sort of seed dispersal. For instance, species like S. barteri with highly plumed seeds may be important for wind dispersal (Gentry, 1991).
The diameter distribution of liana individuals did not show the usual inversed J pattern of some studies (Parren, 2003; Muthuramkumar and Parthasarathy, 2000), in which the number of individuals decreased with increasing diameter. Majority of the species were distributed in the lower diameter classes, indicating that the diameter threshold of 1.5 cm chosen for the study is appropriate, since it included a large number of liana species and individuals. The greater abundance of lianas in small diameter classes may be attributed to their extremely slow stem diameter increment by an inverse resource allocation, as compared to trees.
The study has shown that the semi-deciduous forest of the KNUST botanic garden has high diversity of climbing plants which contribute significantly to the structure and functioning of the plant community of the garden. Diversity of lianas was greater than that of vines. The species showed great diversity in climbing mechanisms. Host species, habitat and the interaction between them as well as human activities were the important factors that determined climber density. Majority of the climbing plants showed clumped distribution. Most of the lianas were distributed in the lower diameter classes.
ACKNOWLEDGMENTS
We are grateful to Mr. Ntim Zakri of the Forestry Department, Kumasi, Ghana and Mr. Degan Amissah of the KNUST botanic garden for their assistance in the taxonomic identification of plants. We are also thankful to the Forestry Department, Kumasi, Ghana for given us access to their herbarium.