Research Article
Composition and Succession of the Woody Flora of South Nandi Forest, Kenya
Department of Integrated Environmental Science, University of Namibia, Namibia
J.M. Mugo
Department of Forestry and Wood Science, Moi University, Kenya
South Nandi Forest in Kenya is a unique Afromontane rain forest being a transitional between the midland-lowland Guineo-congolian forests and Afromontane forests (White, 1983). Afromontane forests are some of the most species rich ecosystems in Africa (Bussmann, 2004). These ecosystems experience pressure from human disturbance (FAO, 2009) as they share the same limited agricultural high potential areas with high human populations (Burgess et al., 2007a; Cordeiro et al., 2007; Fjeldsa and Burgess, 2008). This often results in significant destruction of the forests. Human or otherwise caused disturbance to forest ecosystems sets in motion successional changes. Such changes require documenting and understanding if the future management of the forests is to be in tune with the floristic realities. There are a number of publications (Burgess et al., 2007b; Schmitt et al., 2010; Bussmann, 2001; Aerts et al., 2011), on the species composition of eastern African Afromontane forests. However, factors influencing tree distribution and how the species composition has changed over time are rarely give attention. Overall, many of the Afromontane forests especially the small sized ones, remain unstudied.
Many of the earlier floristic surveys in Kenyan high forests have been timber inventories aiming at rough estimation of available timber for sawmilling (Forest Department, 1967, 1976). These timber inventories gave little attention to population status of individual species. A few studies have been aimed at documentation of biodiversity (Beentje, 1990; KIFCon, 1994). Consequently these forests remain relatively unknown floristically.
Majority of these indigenous forests have faced heavy exploitation and destruction in the past (Forest Department, 1994). In 2005, the country acquired a new forest act. Under this Act the forests have been put under strict protection with the assumption that they will recover to their former state. The Act allows local communities to form Community Forest Associations (CFA) and acquire management rights which can include harvesting of both wood and non-wood products from the forests. This means the forests are likely to face some controlled exploitation in the near future. Before a CFA can get management rights it must prepare a management plan for the forest or section of the forest it intends to manage (Republic of Kenya, 2005). Under such a dispensation, it is important to compile some baseline information for these forests. There is need to document their remaining biodiversity, gauge their conservation and utilization values. This will provide a reference point upon which to assess the impacts of future management.
Since 1963 there have been three inventories carried out in South Nandi Forest: A reconnaissance inventory survey between 1963 and 1967 (Forest Department, 1967); a timber survey in 1976 (Forest Department, 1976) and a species composition survey (KIFCon, 1994). These studies did not give much attention to plants <5 cm in diameter. Without the coverage of the whole spectrum of tree sizes, species dynamics cannot be well understood and hence one cannot be sure of the species re-colonizing the forest. The current study has attempted to verify the findings of these previous inventories and to compile a checklist of erect, woody species found in the forest. Composition and relative abundance of tree species and succession are given attention. Classification of the forest is also discussed. The ultimate goal of the study was to generate baseline data on composition and succession trends as such data will be essential in guiding future management and evaluating impacts of management.
Study area: South Nandi Forest, declared government forest in 1936, is situated in western Kenya near Kapsabet Township. The location is 0°.00'-0°.15'N, 34°.45'E-35°.07'E (Fig. 1) and it covers an area of about 20,000 ha of which 1400 hectares is plantations of exotic species. The rest is indigenous forest (KIFCon, 1994). Altitude ranges from 1700-2000 m. Mean annual rainfall is 1600-2000 mm and annual temperature is 19. (Republic of Kenya, 1982). Fifty seven and 82 species of woody plants excluding lianas were recorded to be growing in the forest in the 1967 and timber inventories, respectively (Forest Department, 1967; KIFCon, 1994). In White (1983) classification the Nandi forests (North and South Nandi) are placed under Afromontane rain forests. In Whites map Nandis location is the Afromontane center of Endemism at the border of the Lake Victoria regional mosaic.
Floristic survey: The floristic survey had two components: (1) a systematic sampling of a forest block measuring 9x5 km located more or less in the middle of the forest (Fig. 1). This was conducted between 1995 and 1996 (2) Unstructured survey covering the rest of the forest, conducted between 2005 and 2007). Multiple-nearest-tree sampling method developed by Hall (1991) was used. According to this method, the numbers of observations are kept constant for all samples.
Fig. 1: | Location of South Nandi Forest and the position of the surveyed forest block |
Table 1: | Size categorization of the plants assessed and number of observations made at each sampling point |
In this case the observations were the number of individual shrubs and trees assessed per size class. Assessment was limited to live, erect, woody plants of sizes shown in (Table 1). Sample points were arranged systematically in a forest block measuring 7.2 km. by 2.4 km laid out near Kobujoi Forest. The block was gridded (600 mx600 m). Sampling was done at the intersections of the grids. A total of 57 sample points were assessed. At each sampling point, the species and diameter at breast height (dbh) of each individual tree of the specified size (Table 1) were recorded. For further documentation of species composition, erect woody plants encountered elsewhere in the forest rather than within the samples were also recorded by species. This was achieved through unstructured sampling, whereby walks were made along existing trails at various points of the forest and species of trees and shrubs along the trails recorded without determining their density.
Average stocking (individuals ha-1) and basal area (m2 ha-1) for trees >2 cm dbh (for 57 samples) were calculated for each species for all size classes combined. A list of all tree and shrub species encountered in the survey was compiled and compared with species lists of the previous surveys (Forest Department, 1967, 1976; KIFCon, 1994).
RESULTS AND DISCUSSION
Species composition and classification of South Nandi: A large number of species are present in South Nandi Forest but with few exceptions occur at low stocking. Stocking (individuals ha-1) and basal area (m2 ha-1) of individual species are shown in (Table 2). A total of 49 species were recorded in the surveyed forest block. Of the 49 species, 45 had also been recorded by either the Forest Department (1967) survey or the KIFCon (1994) survey. Another 37 species (Table 3) were recorded outside the surveyed block, through the unstructured survey. Four species recorded in the current survey; Allophylus ferrugigeus, Chionanthus mildbraedii, Deinbollia kilimandscharica and Psydrax parviflora subsp. rubrocostatum had not been recorded in the previous surveys. Eight species (Table 3) were not recorded in the current survey but had been recorded by the Forest Department (1967) and KIFCon (1994) surveys. The general stocking in South Nandi, where a few species contribute a large proportion of the total stand stocking and majority of the species are only represented by a small number of trees is consistent with stocking in the East Usambara (Afromontane rain forests), Tanzania (Hamilton, 1989); lower attitude, Mt Elgon forest (Hitimana et al., 2004); Kakamega Forest (Odhiambo et al., 2004), afromontane forests at Bonga, Ethiopia (Schmitt et al., 2010), Taita Hills forests in Kenya (Aerts et al., 2011) and not unusual for tropical forests in general (UNESCO, 1978).
For individuals >2 cm dbh, twelve species (Table 4) are of particular interest, combined, they contribute 73.5 % of the total basal area and 72.2 % of the total stocking of the surveyed forest block. The six most abundant species are Tabernaemontana stapfiana, Croton megalocarpus, Drypetes gerrardii, Strombosia scheffleri, Diospyros abyssinica and Casearia battiscombei. These species were reported as the most abundant in the 1976 inventory (Forest Department, 1976) with the exception of Drypetes gerrardii, which was not even recorded then. In the 1976 inventory Strychnos usambarensis was reported as the fourth most abundant species. This species was not recorded in the present study and KIFCon (1994) recorded only one tree in its 3000 records. These inconsistencies suggest that in the 1976 inventory Drypetes gerrardii was misidentified as Strychnos usambarensis (the two species share a local name-Shekoye). Apart from this inconsistency there is no evidence for appreciable change in the relative abundance of species between 1976 and the present.
Comparing the species composition of South Nandi Forest with that of other Afromontane rain forests, the placement of the forest as an Afromontane rain forest (White, 1983) was found appropriate. Twenty four of the fifty five species mentioned by White (1983) as being typical of Afromontane forests are present in South Nandi Forest. The species composition of South Nandi matches well the composition of forests (southern Wellega, Illubabor and western Kefa forests) in neighbouring Ethiopia also considered as Afromontane rain forests. About 50% of the erect woody species reported for these forests (Friis, 1992) are present in South Nandi.
South Nandi differs from other Kenyan Afromontane rain forests in two particular ways. Firstly, Croton megalocarpus which occurs as one of the dominant species, although widespread in the highlands, rarely assumes dominance or co-dominance in other forests (Trapnell and Brunt, 1987). As a dominant, it occurs only in the Nandi forests, Kakamega and in one small forest near Nairobi and another small forest near Meru. Secondly, because of Nandis location, at the interface of Afromontane archipelago-like regional center of endemism with the Lake Victoria regional mosaic (White, 1983), species of Guineo-Congolian origin (notably, Antiaris toxicaria, Harungana madagascariensis, Trilepisium madagascariense and Zanthoxylum gilletii not present or very rare in the other Kenyan Afromontane forests) are present in South Nandi.
Table 2: | List of species recorded during the current study and % representation of each species in the surveyed forest block based on stocking (stems/ha) and basal area (m2 ha-1) for trees >2 cm dbh |
Forest block column: the value inside brackets stands for stems/ha or basal area and the value outside, representation (%) |
Table 3: | Species recorded (+) outside the surveyed block (current study) or documented to be present by the Forest Department (1967) or the KIFCon (1994) Inventory |
Table 4: | Contribution of twelve major species to the total basal area and total stocking of a forest block in South Nandi Forest, Kenya |
These differences, whether reflecting persisting floristic distinctness or a seral phase of Croton dominance, have management implications. A management objective could be maintenance of character if the present forest is shown to contain important germplasm. Alternative option could be to manipulate the forest to encourage change to a more commercially or biologically valuable forest.
Population status of top canopy tree species and succession: Stocking (individuals ha-1) and basal area (m2 ha-1) of trees >2 cm dbh for the top canopy species are shown in Table 5. Casaeria battiscombei, Diospryos abyssinica, Drypetes gerrardii, Strombosia scheffleri and Tabernaemontana stapfiana are fairly well represented in all size classes. Croton megalocarpus and Celtis africana have poor stocking in the smaller tree sizes (≤14 cm dbh), their high contributions to the total basal area are due to high stocking of big trees (>20 cm dbh) for Celtis africana and high stocking of trees >8 cm dbh for Croton megalocarpus.
Table 5: | Stand table for South Nandi, Kenya based on 57 samples, showing stocking (individuals ha-1) and total basal area (m2 ha-1) for individuals >2 cm dbh, for top canopy species |
Olea capensis and Prunus africana have very low stocking for trees >2 cm dbh. Their high contribution to the total basal area is from a few large individuals.
A closer look at the population status of these top canopy species (Table 5), shows some evidence that Croton megalocarpus is losing dominance. The stocking of its small and medium sized trees (>8-20 cm dbh) is low compared with the stocking of its main competitors, Drypetes gerrardii, Tabernaemontana stapfiana and Strombosia scheffleri. The decline is partly explained by the high intolerance of Croton megalocarpus to shade, which restricts it germination to gaps. However, harvesting may be another cause of the decline. The population status of Drypetes gerrardii, Tabernaemontana stapfiana and Strombosia scheffleri suggests present trends will result in these as dominant species in the future (trees >20 cm dbh), partly due to their ability to regenerate under shade. Intensification of the dominance of Tabernaemontana stapfiana and Strombosia scheffleri will be at the expense of more valuable pioneer species such as Croton megalocarpus, Prunus africana and Olea capensis. To contain the extent of dominance of Tabernaemontana stapfiana and Strombosia scheffleri, research on better utilisation of timber from the two species needs to be done to make their harvesting attractive. High stocking and a balanced size-distribution in Casearia battiscombei suggest the species could be managed as the principal plywood source from the forest in future.
Of interest to management for timber are the furniture wood species Aningeria adolfi-friedericii, Fagaropsis angolensis, Olea capensis, Prunus africana and the plywood species Trilepisium madagascariense and Zanthoxylum gilletii. These species have very low stocking for trees >2 cm dbh and contribute very little to the total basal area of the forest. Occurrence is in moderate to high numbers as seedlings and poles (≥50 cm height-8 cm dbh) but at low concentrations in the larger size classes (>8 cm dbh) (Table 5). This may have been caused by a number of factors. Considering their high value for timber and plywood, over-exploitation could explain the low stocking of trees >20 cm dbh. There has been both legal and illegal harvesting of all six of these. Records at Kobujoi Forest Station show that legal harvesting has been in progress for more 30 years. However, over-exploitation would not explain the low stocking of individuals >8-20 cm dbh and it seems few seedlings and poles are recruited to larger size classes. Mortality and time-of-passage patterns and their possible implications for recruitment need to be investigated. Priority in the conservation of these species must be location and protection of the remaining seed sources.
The placement of South Nandi Forest in Afromontane rain forest was found to be appropriate as it composition is similar to that of other Afromontane forests. Top canopy species consist of both pioneer and shade tolerant species. This may be as a result of disturbance mainly logging that has taken place over the last 60 years or more. Drypetes gerrardii, Tabernaemontana stapfiana and Strombosia scheffleri are gaining dominance over previously the dominant species, Croton megalocarpus. Future management should consider utilization of these lesser known species. The better known timber and furniture wood species Aningeria adolfi-friedericii, Fagaropsis angolensis, Olea capensis, Prunus africana and the plywood species Trilepisium madagascariense and Zanthoxylum gilletii are having very low densities and may require special management to enhance their stocking.
The main part of this study was funded by FAO/GEF through the Institutional Support for Protection of East African Biodiversity Project. The unstructured survey was made possible by students of Forestry Department, Moi University, Kenya. We are grateful to Dr. J.B. Hall for guiding part of the study. Assistance given in data collection by Messers Kibiru and Nthuni is also appreciated.