And management

Need for effective conservation and sustainable use

Forests are important as a habitat for plants and animals, for their role in the maintenance of cycles of rainfall and biogeochemical substances like carbon and nitrogen, and as a source of all kinds of forest products (e.g. timber, fuelwood, bushmeat, medicines, lianas for construction) satisfying the needs of human beings. Forests are commonly used for recreation and religious ceremonies (e.g. sacred forests). In short, tropical forests are able to fulfil a lot of these functions for society. Tropical forests are also well-known to be generally species-rich, although there is a great variation in this aspect. Recent studies have localised the areas in the world where biological diversity is high, but which are at the same time threatened by destruction (Mittermeier et al. 1999, Myers et al. 2000). The rainforests of West Africa rate among the 25 most important "biodiversity hotspots" identified.

Because of their isolated position the Upper Guinean forests harbour a large number of endemic animal (Brooks et al. 2001) and plant species (Myers et al. 2000, Wieringa & Poorter chapter 6). At the same time these forests are disappearing rapidly (Chatelain et al. chapter 2). A recent overview of deforestation rates in selected areas all over the world (Achard et al. 2002) shows that the overall rate of forest disappearance is declining. This is an average value, however, and in several West African countries, notably

Côte d'Ivoire, the deforestation rates are still extremely high.

For an effective policy for conservation and sustainable use, information is needed on several aspects. We need to know where forests are still present, what their conservation status is, and what their actual threats are. We need to know the distribution of rare and endemic species in the region, and the regions in which they are concentrated. We also need to know the main uses of these forests and how they can be managed sustainably. A conservation strategy may be based on a large variety both of species (plants, animals) as well as on habitat diversity (Margules & Pressey 2000). Conservation strategies often focus on rarer species, because these are generally more likely to go extinct without specific management. Unusual habitats often support unusual species, so conservations tend to focus on these as well.

For a good conservation strategy, apart from these aspects also information is needed on the maintenance of

Figure 8.1 Satellite image of forest cover of the Classified Forest of Mopri, Côte d'Ivoire, in 2000. The central square represents the area that has been conserved in a conservation project. The other areas are in various stages of plantation development and agriculture (figure provided by Chatelain, Hao, Gautier & Spichiger).

Figure 8.1 Satellite image of forest cover of the Classified Forest of Mopri, Côte d'Ivoire, in 2000. The central square represents the area that has been conserved in a conservation project. The other areas are in various stages of plantation development and agriculture (figure provided by Chatelain, Hao, Gautier & Spichiger).

ecosystem processes (maintenance of soil fertility, hydrological functions), the maintenance of plant-animal interactions (pollination, dispersal), and of minimum viable population size of selected species. We need to know how animals and plants react to climate change to be able to predict changes in the vegetation as a result of predicted large scale climatic changes. Finally, we need to know how species react to forest fragmentation that results from logging and agricultural activities.

The realisation of conservation goals requires strategies for managing whole landscapes looking at both production and protection areas. Forest Reserves and National Parks are often the cornerstones of national protection strategies, but safeguarding their protection and conditions alone generally is not enough for effective conservation. This is mainly due to the fact that in most countries reserves are biased samples of biodiversity, mostly confined to remote places and to areas that are unsuitable for commercial activities (Margules & Pressey 2000). In Ghana, however, forest reserves cover all forest types (Hawthorne 1996).

How can botanical and ecological information be used for the conservation and management of the remaining forest resources? In this book we present mainly information on the botanical part of the forests. We describe the forests of the area and classify them into a regional system. We focus on rare and endemic species, their characteristics and their distribution. We address the question of where large numbers of these species are found, Further analyses of the data collected may lead to identification of areas where species are concentrated that deliver products that society needs, like timber, rattans, medicinal plants, and dyes. For a large number of species we have analysed the environmental conditions of the locations where these species are found and we have been able to predict where these environmental conditions are optimal for their occurrence. This could lead to strategies of management directed at use of selected species (e.g. in plantations, homegarden farms and/or in sacred forests).

Deforestation and land use changes

The extent of forests is diminishing drastically. At the scale of West Africa we estimate that between 20% and 50% of the forest cover which existed at the turn of the 19th century remains (Chatelain et a/. chapter 2). The current area with a forest cover is estimated to be between 463,240 km2 (SOFO 1997) and 415,940 km2 (FAO 2000). These overall estimates are necessarily crude, mostly due to the different methods used in the calculations. For tropical Africa, the variation among researchers in the changes in forest cover between 1980 and 2000 is even larger than the average change observed during that period (Matthews 2001). Fairhead and Leach (1998) have shown that the total forested area at the beginning of the century most probably was over-estimated, which has contributed to an incorrect view of deforestation levels. However, no one can realistically deny that over the last 50 years the extent of forest in our region has declined dramatically.

We need an update of information on the forested areas, and on the state of conservation and delimitation of the protected areas. All countries in West Africa show a lack of recent detailed cartographic information, but this is stronger for the countries that have seen war for years such as Liberia and Sierra Leone. Priority zones should be mapped at a higher resolution to be able to follow their development over time in detail.

As in almost all countries around the globe, the moist semi-deciduous forest is the most disturbed one (Chatelain et a/. chapter 2), mainly because these forests harbour a large number of exploitable species and also because the soils of these areas are of the most productive ones in terms of agriculture. These forests thus need more careful sustainable management.

Large areas of forest are still found in southwest Côte d'Ivoire and in southeast Liberia. Smaller, but numerous tracts occur in west Ghana. Taï National Park contains a large part of the remaining forest block at the Liberian-Ivorian border. This famous park is currently actively managed but the effectiveness in the long term is doubtful, as poaching is extreme (Caspary et al. 2001) and cocoa plantations are found in some areas. The political situation over the last past years is not very stable, which is a serious drawback for forest conservation (Richards 2001). Taï National Park needs stronger control of mainly illegal activities, in close cooperation with people living in the area. Alternative land uses need to be stimulated, such as planting trees for wood and for all kind of products in arable fields and in abandoned areas. Planting of Tieghemella heckelii is a good example in this respect (cf. Bonnehin 1998).

Currently Liberia, together with several international NGOs is trying to preserve forests in the country, like Sapo National Park. Progress is slow however, not in the least due to the political situation in the country. We suggest that the international community increases political pressure for effective preservation of these last remaining vast areas of forest in the region.

Small forests (2 to 100 ha) represent 10 to 30% of the forested area in Côte d'Ivoire. The detailed local analyses of Chatelain et al. (chapter 2) show that the rural areas of Côte d'Ivoire have almost 76% of the forest cover, while 23% is in classified state forests. At present, these small forests are ignored on maps and in the national statistics because they are very fragmented. Updated forest cover maps should have a resolution sufficiently high (30 X 30 m) to allow the resolution of the land-use mosaics and tree plantations, which currently pose a large problem for a realistic estimate of the forest cover updates. Ivorian villagers who invest in forest conservation have no property rights according to the constitution and to the manual of forestry practice. They can do nothing against the continued extraction of large amounts of wood by logging operators who penetrate into community plantations and forests with the authorisation of a logging permit. Most forest policies are focused on forests managed by the State (like Forest Reserves). New policies are needed that focus on small forests in a fragmented landscape despite difficulties for their identification and (long-term) management. To be able to tackle problems that exert pressure on such forests a small-scale local approach is needed, in close concert and cooperation with the local population. For example, the socio-economic infrastructure of village development of more intensive systems of agriculture, and a re-assessment of subsistence cultivation may all need support or help to keep abreast of the changing environment. For a long time, subsistence cultivation was devalued in favour of the system of export-oriented cultivation, with, sometimes high, negative impacts on the food security of farmers. Also, corridors may be established to link existing patches of forest (including "sacred" forests) into larger scale forest networks. A relatively small amount of land area then is

Box 8.1 Forest corridors and networks for elephants

Forest elephants Loxodonta africana cyclotis have declined greatly in the moist forests of Upper Guinea and dwindled down to a total number estimated at not more than c. 3000 individuals in 1998. Elephants influence forests at two main levels: as opportunistic frugivores, by directly effecting the dispersal and regeneration of certain species; and by trampling, debarking and otherwise disturbing the forest. They tend to make gaps and keep forest open and declining elephant populations are followed by altered successional dynamics in some forests (e.g. Laws et al. 1975, Sheil 1996). They are dispersers of a large number of woody forest species, giving rise to concerns that without elephants the plant diversity of Upper Guinea forest plant communities will not be maintained. Although several studies (e.g Chapman et al. 1992, Lieberman et al. 1987) have pinpointed plant species that are absolutely elephant-dependent for dispersal/germination, two recent studies cannot find hard evidence for that (Hawthorne & Parren 2000, Chifu Nchanji & Plumptre 2003).

Nowadays, these elephants live mostly in scattered forest patches in a matrix of cultivated land, except for those in extensive forests in eastern Liberia (Parren & de Graaf, 1995; Appleton, 1997). The elephants of the Côte d'Ivoire-Ghana border area, the total population of which numbered 685-855 in the late 1980s, are contained by the savannas of the Dahomey Gap to the east and by the so-called V-Baoulé, in Côte d'Ivoire to the west. Movement of elephants between forest fragments, often not exceeding 10,000 ha, in the Côte d'Ivoire-Ghana border area would require forest corridors between nearby fragments. Corridors are linear vegetations connecting at least two isolated habitat fragments that were once connected (Saunders & Hobbs 1991). "Shelterbelts", strips of forests often not more than 1.5 km wide and up to 20 km long (Foggie 1949), established in the mid-1930s in Ghana, provide examples of potential corridor design.

Corridors should be designed to attract forest elephants into them, and important characteristics to take into consideration are the presence of food plants, availability of water, and the size and structure of the corridors. Forest elephants feed opportunistically on certain fleshy fruits when these are available, and rely upon less nutritious foods during periods of fruit scarcity. Favourite forest tree species are Parinari excelsa, Balanites wilsoniana, Duboscia viridiflora, Panda oleosa, Sacoglottis gabonensis and Tieghemella heckelii (Sachtler 1968, Martin 1991, Hawthorne & Parren 2000). In the dry season the water provided by the fruits may be insufficient and elephants come out of the forest in search of water. Artificial waterholes could be created in forest elephant reserves to ensure they stay within the reserve boundaries, as has already been successfully done inside FC Bossematie (Waitkuwait 1992). Corridors on riverbanks, such as the Bia river corridor, would have the advantage of providing readily available drinking water. Corridors along streams and rivers would probably be the most useful and successful as such landscape features have a disproportionately large influence on landscape functions (e.g. population, community, and ecosystem processes) beyond what their relative area on the landscape would suggest (cf. Puth & Wilson, 2001).

needed to be able to improve the existing forest "infrastructure" and thus to improve forest quality and forest functions. For the border area between Côte d'Ivoire and Ghana comparable forest networks (with the elephant as a flagship species, see Box 8.1) have been proposed recently (Parren et al. 2002). Such measures then should lead to a lower pressure on the remaining old-growth forests in the area: planted corridors can be used by the local population for some of their needs (e.g firewood, construction poles). This relief of pressure on old-growth forest may have positive effects on animal and plant populations.

The ideal of actually regaining forest area in corridors may seem a long way off, but in the shorter term, and less contentiously, we suggest that the State first strives to establish a widespread policy of conservation based also on community management, as is widely promoted in Ghana, with varying degrees of success (see also Fairhead & Leach 1998). Villagers need forest to fulfil many of their needs (e.g. wood, medicines, religion). Only in a long-term conservation scheme in which advantages for the local population are clear for the short and long term, villagers will (co-) protect forests. And, before even this stage can be expected to make progress, villagers who conserve the forests need to get the necessary rights to accomplish this conservation action. We therefore recommend managers and politicians in the immediate future focusing on these short-term goals and, when the benefits are obvious to all, the medium and longer term goals outlined above might also be realisable.

Fire as a management tool

The local population can have a strong direct impact on the extent of forest (Gautier & Spichiger chapter 3) especially in the drier areas. There is a general agreement that savannas of the transition zone are not in equilibrium with climate and are a direct consequence of annual fires, which are, however, more likely in periodically drier periods. The climatic climax would be dry semi-deciduous forest and further north (from c. 9°N) dry forests and woodlands. Most Guinean savannas are considered to be of secondary origin. Protection of fire, cattle ranching with exclosures and shifting cultivation with longer fallow periods would lead to forest expansion, and give opportunities for area expansion to the threatened dry forest type.

The prevailing environmental conditions may be favourable for afforestation, even when the effect of fire is taken into account (Aubréville 1966). This has been demonstrated by comparing two sets of aerial photographs separated by 25 years in Lamto reserve, Côte d'Ivoire. Despite the occurrence of annual fires, the forest surface is increasing both by direct progression of existing forests and by the appearance of forest islets within the densely wooded types of savannas found on hilltops (Gautier 1989). Hilltops may need continued protection as safe sites for establishment of forest species, and sources of recolonisation of the forest. However, the hilltops on the Accra plains were reserved in the 1930s and still exist. Meanwhile fire continues to rage annually across the lowlands and nothing ever redevelops there. Swaine et al. (1992) demonstrated that forest that recovered in fire protected areas included dry semi-deciduous forest species that had no local seed source, presumably because the majority of them are well dispersed by bats and wind.

In Côte d'Ivoire the timing and frequency of fire as a management tool has shown promising results. Gautier and Spichiger (chapter 3) describe experiments that show that annual early burning (beginning of the dry season) has promoted a shrub-dominated savanna, while annual retarded burning (mid to late dry season) has promoted a savanna with a very low shrub cover because of a greater fire intensity (drier and maybe also more fuel) and the fact that it destroys tree leaves and shoots that establish long before the rainy season. Meanwhile an integral protection from fire has led to a genuine forest structure restoration, even in the vicinity of the southern dry outlier forests in the Dahomey gap (Swaine et al. 1992). These results imply that fire can be used as a management tool to create specific desired vegetation types, and that fire protection in particular is crucial to forest conservation in the semi-deciduous forests.

In northeast Ghana, another fire protection experiment has led to a clear decrease of grass cover and a clear increase in tree number and stature, but in contrast to the above, there has been no establishment of Guineo-Congolian forest species. This is explained by the much more northern position of the plot, north of the forest-savanna transition zone, in a region where dense forest species are very rare and thus forest seeds unavailable. This indicates that successful restoration of Guineao-Congolian forest is only possible if seed sources are available, although many pioneers can apparently arrive from many kilometres or tens of kilometres (Swaine et al. 1992). The environmental matrix is important: gallery forests may be useful in some savanna areas as corridors for Guineo-Congolian species.

Fires can also occasionally sweep under the canopy of the drier types of rainforest (for Ghana, see Swaine 1992). These fires are generally low in intensity and are essentially fuelled by litter (ground fires, Phillips 1974). They extend to the canopy only during exceptionally dry conditions (Swaine 1992). Like logged forests, burnt forests are more prone to renewed burning (Hawthorne 1991). This can be dangerous and large fires therefore have to be strictly under control: prevention is better than healing.

Impact of shifting cultivation on forest extension

There has been considerable debate on the impact of shifting cultivation on the distribution of forest and savanna. It was generally assumed that shifting cultivation would lead to an extension of savannas by destroying woody vegetation (Aubréville 1966). In South America slash and burn agriculture, fire and deforestation have lead to savannisation. In some places like the western extension of the V-Baoulé (see Kouamé et al. chapter 5), there is general agreement that the Pennisetum purpureum savannas are derived from cultivation of forest areas (Adjanohoun 1964).

Field studies in Côte d'Ivoire and Togo have even revealed that shifting cultivation at the forest edge promotes afforestation (Spichiger & Blanc-Pamard 1973); field preparation reduces herbaceous domination in the fallow, and thus limits fuel availability. Fairhead & Leach (1995; 1998) have demonstrated that (especially in Guinea) many human activities promote afforestation (farming system, deliberate planting of forest species in fallows, pasture, fire-breaks). Based on aerial photographs, they show that in some areas this has lead to a 40% increase in forest cover during the last 50 years. Although this might be the case for some areas, the results of Achard et al. (2002), however, show that overall deforestation is far larger than afforestation.

Increase in population pressure as well as transforming farming systems into cash crop production has completely altered the intensity of human impact on vegetation during the last decades. The duration of the fallow in the farming system is of paramount importance. If the fallow period is reduced from c. 10 years to c. 5 years, as is currently happening in many places in the forest-savanna transition zone, then afforestation processes can no longer take place. We therefore suggest to aim at a fallow period of at least 10 years. We have to realise, however, that such young secondary forests will not help much in long-term biodiversity conservation, and as such it will never be a substitute for old-growth forest. It does help, however, in terms of soil fertility maintenance (increasing agricultural production when it is next farmed), or it provides fuelwood for burning (lowering the pressure on old-growth forests), and it may increase common wild animals that can be used as bushmeat, as an alternative for the rarer wildlife (again lowering pressure on old-growth forests).

Conservation: hotspots and other areas

As part of systematic conservation planning a general compilation of data on the biodiversity of a planning region is needed, and conservation goals have to be identified for that region. Based on that, existing conservation areas have to be reviewed and additional conservation areas might be selected with the various stakeholders. Sometimes, more detailed analyses of biodiversity data in the planning region are needed for that. The data and results presented in this book should support this part of the systematic conservation planning process.

Biodiversity hotspots: consequences for natural reserves

Upper Guinea is one of the 25 most important hotspots in the world (areas where biodiversity is very high and threats in terms of deforestation are large). The Upper Guinean forests are estimated to contain 2800 forest plant species of which 650 are endemic, and c. 400 are considered to be rare (chapter 11). These forests also harbour 1320 vertebrate species of which 270 are endemic (Myers et al. 2000). Within Upper Guinea and based on a selected list of endemic and rare plant species for Upper Guinean forests, Wieringa and Poorter (chapter 6) distinguish three areas of high diversity. A belt of rare and endemic species is found about 50-100 km inland, from Sierra Leone, through Liberia to southwest Côte d'Ivoire. The climax of this belt is in Liberia. Additional high diversity areas are found in areas around Abidjan in Côte d'Ivoire and Ankasa in Ghana. A second belt is found in the mountain areas of Mt Nimba/Mt Ziama in Guinea, and around Man in Côte d'Ivoire, and the Atewa Range in Ghana. These areas need protection and attention.

Clearly, the mountain ranges have many endemic and rare plants. Within these area the mist-bound summits with all their epiphytes are the more diverse parts, and include many species that are uncommon throughout the more extensive lowland forests. The steeper hills are important because of the variety in special habitats and the fact that they are relatively untouched. Hill sanctuaries, as established in Ghana (Hawthorne & Abu Juam 1995), may provide a good framework for planning general protective management throughout the region, as issues of soil erosion, water supply conservation, migratory bird refuges are at a peak here, and yet the sites are generally the least suitable for logging, and not ideal for farming either. Many of the higher hill forests, such as those around Man in Côte d'Ivoire and Tano Ofin and Atewa in Ghana are also rich in rare species that flourish in the hilltop mists; some of these forests have escaped relatively undamaged until the last few years, but are currently succumbing rapidly to destruction, and deserve a special conservation review.

The location of the areas with high plant diversity is strongly related to both water availability and, it is assumed, to the presence of ancient forest refugia. Diversity clearly increases with rainfall and decreases with soil fertility and with distance to presumed ancient forest refuge areas (Wieringa & Poorter chapter 6). In such refuge areas plant species, especially the drought-intolerant ones, are thought to have survived the glacial periods in the past. Because of their isolation during the glacial periods speciation may have occurred. After the glaciation period plant species supposedly dispersed from these refuge areas, leading to higher numbers of less dispersible species nearby. Supposed refuge areas thus need special conservation management inasmuch as any area that is rich in more localised species demands it.

We have to realise that many areas in the Upper Guinea forest block are undersampled with regard to herbarium collections (Figure 6.4 and 6.5). Sampling intensity is particularly low in southeast Liberia. As many of these areas are high rainfall areas we expect that their plant diversity is very high. We recommend herbarium collection expeditions to these areas.

Distribution patterns of rare and endemic species

The rare and/or endemic species of Upper Guinean forests show three characteristic distribution patterns (Holmgren et al. chapter 7). From the 270 species analysed 86% showed a continuous distribution (the species is found over the whole area), 10% showed a continental disjunct distribution (the species is found somewhere in Upper Guinea and also in central Africa) and 3% showed an Upper Guinean disjunct distribution (the species is found in two separated areas within Upper Guinea). Most species had a wide distribution, and 21 species had a distribution smaller than 100 km (Table 7.2). These latter species need extra conservation attention, as species with a small range are generally the most vulnerable to extinction.

In total 26 species in this sample (which we emphasise is not a comprehensive set of all very rare species, many of which are still in the process of being described) are considered very rare as they have been collected only once or twice. Eight of these are found in Liberia, especially around Monrovia and Bomi Hills, in southern Greensville and in the mountains. The rare Ivorian species are concentrated near the border with Liberia and around Abidjan.

The occurrence of 88 out of 112 species was correlated with various environmental factors, of which rainfall was the most important (in 77% of the species rainfall had a significant effect), followed by altitude (in 36% of the species a significant effect). Species with disjunct populations tended to be found in undisturbed areas, and were often shade tolerant species. Species with continuous spatial distributions were often found in open disturbed habitats. How continuous, widespread and abundant a species is, depends on how successful a species is in dispersal and in tolerating open habitats, and a wide range of rainfall conditions. Thus, even within the set of endemic or rare forest species of Upper Guinea the most successful species tend to be pioneer type species. Disturbances have obviously played an important role in shaping the composition and distribution of forest species in this area. As for the species with disjunct distributions, specialising on very specific environmental conditions (in most cases wet and undisturbed), most need special conservation effort.

Three out of the five areas where many rare species and very local species occur also show very high species richness, notably Mount Nimba, southwest Côte d'Ivoire and southwest Ghana (Wieringa & Poorter chapter 6). Areas where environmental gradients are steep (i.e. a long gradient over a relatively short distance) have many environments in a relatively small area. As rainfall is the most important environmental factor determining species diversity, and also determines distribution of most species, areas with a steep rainfall gradient are extra interesting from the point of view of conservation. Climate change thus may have a profound impact on the distribution of species (Holmgren et a/. chapter 7). A series of reserves along the gradient, connected by corridors would be very helpful. Due to the gradual changes species can migrate easily over the whole area in periods when this is needed (e.g. during periods of climate change). Two such areas have been identified. One is the gradient in southeast Liberia and southwest Côte d'Ivoire and the other is the area in southwest Ghana (see Figures 1.3 and 4.3). In the Liberian-Côte d'Ivoire zone several areas can be connected, notably between the Cestos and the Sassandra rivers, connecting the wet coastal forests with the semi-deciduous forests, between Grebo and Taï forests, and between Haut Sassandra and Taï-N'Zo and Niégré. In the Côte d'Ivoire-Ghana zone the forests of Ankasa and Cape Three Points would ideally be connected with the northeast using the Tano and Ankobra rivers, and the Mabi-Songan-Tamin-Yaya forests might be connected with the forests in southwest Ghana. Implementation of such corridors would need huge efforts, however.

Some of the driest forests, especially those in Ghana at the eastern limit of our atlas, are species poor, but of extreme high conservation concern, as these forests include the last stands of several extremely rare species, and the forests are of extremely small, and diminishing extent.

Time and again, however, calls have been made to secure the highly endangered Ta/botie//a gentii stands in Ghana, but they are still not fully secured (e.g. Swaine and Hall 1981; Hawthorne 2001). Because these may have slipped through our (selected species) herbarium analysis, efforts should be made to seek other such pockets of dry forest, possibly with endemics as yet undiscovered, elsewhere on the periphery of our region.

The mountain areas contain many rare and endemic species and need special conservation effort. The Fouta Djalon in Guinea, Mount Loma in Sierra Leone, the Putu Range in Liberia, the Ziama-Nimba-Mont Péko mountain range between Liberia, Guinea and Côte d'Ivoire, and the Atewa range in Ghana are crucial in this respect (Wieringa & Poorter chapter 6, Holmgren et a/. chapter 7, see also Conservation International 2001). Also Banco forest, near Abidjan, ranks high in number of rare and/or endemic species. The scattered dry forests of southern Ghana (Southern Marginal and Southeast outlier types) tend to slip through the analytical net with these sorts of surveys, because they are species poor, but the vanishingly small and highly vulnerable communities include a high proportion of globally rare endemic plants, like the very rare Ta/botie//a gentii and Hunteria ghanensis trees - and so are amongst the highest priorities for conservation effort in our region (Hawthorne and Abu Juam 1995). The Ghanaian Southern Dry forests are again the target of protective measures by the Ghana Forest Service, but it remains to be seen how secure such small pockets can be made without a major ex situ conservation initiative.

We would like to stress the fact that, even though the options are limited for small communities like the Ghanaian Southern dry forests, large stretches of forests are always desirable because small patches are much more vulnerable, especially under current levels of disturbance and climate change.

Tab/e 8.1 Distribution of forest condition, and p/ant diversity, for areas under protection (national forest parks, forest reserves, c/assified forests) for Côte d'Ivoire and Ghana. The forest condition is c/assified into six c/asses: c/ass 1 (exce//ent condition), c/ass 2 (good), c/ass 3 (s/ight/y degraded), c/ass 4 (most/y degraded), c/ass 5 (very poor), and c/ass 6 (no significant forest /eft).

Data for Ghana from Hawthorne and Abu Juam (1995), data for Côte d'Ivoire from H. Dao and C. Chate/ain, amended by V. Be/igné.

The p/ant diversity (Sso va/ues, the number of p/ant species for 50 random co//ection) increases from c/ass 1 to c/ass 7

(c/ass 1: 0-15 species; c/ass 2: 15-30 species; c/ass 3: 30-35; c/ass 4: 35-37.5; c/ass 5: 37.5-40; c/ass 6: 40-42.5; c/ass 7: 42.5-45.0).

The forest types are DS = Dry Semi-deciduous; MS = Moist Semi-deciduous; ME = Moist Evergreen, WE = Wet Evergreen.

Va/ues are percentages oftota/ area under protection.

A. Forest condition

1 (good)





6 (bad)








Côte d'Ivoire

25.6 (inc PN)






B. Plant Diversity

1 (low)






7 (high)









Côte d'Ivoire








C. Forest type















Côte d'Ivoire







Comp/ementarity in se/ection of areas for conservation

In selecting areas for biodiversity conservation the principle of complementarity is very important, apart from high numbers of rare and endemic species, high species richness, uniqueness of the habitat, minimum forest cover and possibilities for connections in networks. Complementarity analysis seeks to include as many species as possible within a network of protected areas, thus sites are selected that add as many species as possible to the protected pool. For Côte d'Ivoire Kouamé et a/. (chapter 5) show that many species are very forest specific (out of the 2126 species in the thirteen forests studied, 624 species were found only in a single one of these forests, see also Figure 5.2) and thus it would seem that a large number of specific forests need to be protected to be able to conserve a large number of species. However, it is acknowledged by all proponents of complementarity analysis that it is much diminished in relevance where 100% complete species lists are not available — a fact that is sadly the norm for forests in our area. Where almost complete checklists for large forest reserves have been compiled from special surveys, it is seen that a significant proportion of the flora of our entire area occurs in limited regions. Between about 24 and 33% of the Upper Guinea forest flora is found in Bia South forest in Ghana, for instance, an "average" type of forest (Hawthorne et a/. 1998). In Ankasa and Draw Forest Block alone about 1200 species are found, about half the forest flora for the entire region. For these reasons, complementarity probably has less to offer than prioritisation by rarity, but we can still adhere to general principles — to try and protect as wide a variety of forest types as possible. This is a principle that was extended by Hawthorne and Abu Juam (1995) in a bid to protect the "hidden" infraspecific biodiversity in terms of genetic diversity within overexploited and widespread species (see also Box 8.2).

With respect to our hotspots of plant diversity we suggest that as much as possible of the hotspot areas will be protected, but also that the number of total protected species will be considered. Kouamé et a/. (chapter 5) distinguish four forest types with local endemic species. Consequently, forest management and protection in Côte d'Ivoire must be oriented to have at least a forest reserve/national park in each of these forest types. In three of these, forest reserves have been established already: Taï NP for western evergreen forest, Banco NP for coastal and Marahoué NP for semi-deciduous forests. The creation of biological reserves in the classified forests of Yaya-Mabi-Songan-Tamin in eastern evergreen forest is the first step towards protection of this type of forest.

For this chapter, we have calculated the plant diversity of the forests in Ghana and in Côte-d'Ivoire in several Upper Guinea-wide classes (Table 8.1). In Ghana the majority of the forest reserves are found in areas with intermediate species diversity. Only 3.4% of the reserve areas are found in high diversity areas (classes 6 and 7). The main reason for this is that high diversity areas and very wet areas form a small proportion in Ghana's forest, and these are the areas with the highest diversity (note here that diversity values are determined on a regional scale, see chapter 6). As almost 100% of all larger extant forest blocks in Ghana are within forest reserves, a bigger issue is how precisely to manage it, not what to gazette. For protected forests in Côte d'Ivoire diversity is slightly higher, especially class 5 is better represented. Overall the variety in diversity is high (see also the maps in chapter 6 and Appendix 4).

Because extremely wet sites only occur in Liberia and Sierra Leone, these hyperwet areas need special protection. The Ghanaian Southern Dry forests, forest that has already been emphasised as a conservation priority, are currently the focus of protective measures by the Ghana Forest Service.

Forest reserves and protected areas

A regiona/ c/assification of forests: consequences for conservation

For the Upper Guinea area several forest classifications have been made in the past (Table 4.5). Most of these are at the scale of one country. In Chapter 4 Bongers et a/. have produced a regional classification based on over 200 sites in the Upper Guinea region. This is a very useful addition to the already existing classifications. In general the classes are comparable but in some cases differences exist. This is for a large part due to the longer environmental gradient existing in the whole area, compared to the one in a single country. For instance, the wet forests in Ghana are not "very wet" when seen in a regional perspective. Also, our classification (and thus also the resulting map (see Appendix 2) is based on a selection of species only, which makes it easy for decision makers to put "their" forest in an international context. In Chapter 4 Bongers et a/. have clearly shown where the new classification is different from the older ones. It is clear that some forests, like the hyperwet forests of Liberia and Sierra Leone do not occur in the other countries. As such forests do not occur in Ghana and Côte d'Ivoire, Liberia and Sierra Leone have a special regional responsibility with regard to specific protection of hyperwet forests. Other forest types, such as the southern marginal outlier forests occur only in Ghana. These forests are extremely dry and have a high proportion of endemic species (and also of endemic individual plants) and thus a high conservation value, despite low species diversity (Hawthorne 1996). Conservation of such forest thus is of a wider importance than for the particular countries alone. The same applies to the upland mountain forests that occur sporadically in Ghana, but mainly in the triangle border of Côte d'Ivoire, Liberia and Guinea.

Roswell Crash Site Photographed 1947
Figure 8.2 A large Ceiba pentandra tree is cut in a forest concession at N'Zo (Côte d'Ivoire).

In this book eight broad forest types are characterised. That is not an overview of all the different forest types in the region, only a characterisation of the forest types and forest sites used in this book, at the level of Upper Guinea. The special forest types in Ghana have been mentioned already. Also, swamp forests and mangrove forests, and various kinds of riverine forest and sacred groves in the savanna area are not taken into account. With respect to the mountain forests, all are put in one category here.

We have to realise, however, that at a more local and more detailed scale many more differences can and will be found. For a proper conservation of all these forest types, in-depth analyses are necessary. These can better be done at a more local level. The studies in Ghana (for instance Hall & Swaine 1981, Hawthorne & Abu Juam 1995) are starting points for those studies. Also for Côte d'Ivoire such studies are being undertaken, and for Liberia new studies are being performed currently (Conservation International, Flora & Fauna International).

Status of the areas under protection

In Upper Guinea various kinds of protected areas can be distinguished: national forest parks, forest reserves and classified forests. For the countries of Liberia, Côte d'Ivoire and Ghana we have mapped these forests (Appendix 3). In spite of the fact that some natural areas are protected by law, this legal protection is no guarantee for factual effective protection. The conditions of these forests are quite variable, due to legal and illegal logging (Figure 8.2) for timber but also to clear land for small scale as well as large scale agriculture. In Côte d'Ivoire, for instance, only 11 areas are formally protected as "national parks" or "natural reserves", with in total 1,874,800 ha: Azagny, Banco, Comoé, Haut-Bandama, Iles Ehotilé, Marahoué, Mt Péko, Mt Sangbé, the complex Taï & N'Zo and Mt Nimba (Bakarr et al. 2001). Only Comoé and Azagny National Parks were considered to have a reasonable level of administrative and management support, enough to provide real protection for the parks. All others are considered to have shortage of effective protection. For Taï National Park & N'Zo Reserve effective protection is doubtful, despite numerous international agencies maintaining research projects in the area. Taï is very important as it probably contains the only spots of primary forest for the country, and still contains many of the species that are disappearing from other areas. But currently also Taï (and especially the formerly called buffer zone) is threatened by habitat degradation, poaching, logging, farming, illegal gold mining, and other human activities.

For Ghana, the forest condition was determined by Hawthorne and Abu Juam (1995) into six categories: class 1 (excellent condition), class 2 (good), class 3 (slightly degraded), 4 (mostly degraded), class 5 (very poor), and class 6 (no significant forest left). From the total areas of forest reserves 26.4% was in bad condition (their classes 5 and 6), 16.8% in good condition (their classes 1 and 2) and the rest was intermediate. For Côte d'Ivoire this was done by Hy Dao, Chatelain and amended by Beligné. For the classified forests of the forest zone 6.2% was in bad condition, 36.9% in good condition and the rest in intermediate (Table 8.1). The high percentage in good condition is mainly due to the large area of Taï National Park. For Liberia comparable information is scarce and scattered, and we are not able to give a comparable and reliable overview. For some of the forests information is collected by Conservation International and Flora & Fauna International. D. Byler (personal communication) and M.P.E. Parren (personal communication) considered in very good shape (class 1) Sapo National Park, in good shape (class 2) Grebo National Forest (but degrading -M.P.E. Parren), and parts of Krahn Bassa National Forest, and as slightly degraded (class 3) Gio and Gbi National Forest (but degrading - M.P.E. Parren), the largest part of Krahn Bassa, and Lorma National Forest. The Lamco concession area was in degraded condition (class 4).

Chatelain et al. (chapter 2) show that in the large majority of classified forests in Côte d'Ivoire 40-50% of these forests have been replaced by planted trees and crops. The numerous classified forests in the region of Abidjan, for instance, are not well managed and protected. Their importance for wood production and as a biological reservoir is doubtful, because in many cases they are forest only in name (77% of their area is not forested). In Côte d'Ivoire protection effectively maintains reasonably good quality forests only in national parks (see Table 8.1). To avoid this and to restrict agriculture to areas outside the forests, farmers need to be encouraged to invest in their plantations (e.g. rejuvenation) and to strictly forbid agricultural practice in classified forests. The existing agricultural lands in the surroundings could be more effectively used, and many fallow lands are practically unused. Additionally, conflicts between modern and traditional land rights as well as the principle of land ownership acquired by clearing, are both obstacles in a

Box 8.2 Integrating forest management with academic analysis of endemism and rarity

There are several problems with linking global patterns of rarity to management practice, including the following three examples:

1. There is a continuum of scale of analysis of pattern and of management implementation of rarity (Hawthorne 1996). Endemism and presence/abundance issues are all dependent on scale. Patterns of concentration of more or less (globally) rare species are sharply demarcated at a local scale, but vary subtly across any landscape, from swamps to ridges and from logged to unlogged forest.

2. There are considerable difficulties with obtaining precise information on all species, even if one can see they are globally rare. Although a large subset of species has been reviewed here, they are not a complete set of all high priority species, and even this work has taken many hundreds of person-hours to compile.

3. For many proposed methods of linking the global to the local, obtaining adequate information about the local community is likely to be a problem; forest inventory of all except presence-absence in small areas is highly expensive. Complementarity analysis for instance requires more or less complete species lists for forests to be prioritised.

For these and other related reasons, and in order to integrate global trends with local management practice and Environmental Impact Assessments (EIA), Hawthorne and Abu Juam (1995), developed and implemented for Ghana a species ("Star") and associated forest ("Genetic Heat Index system") ranking system. This is a fairly crude but equally simple, objective and repeatable method, and has enabled GEF (Global Environment Facility) funded projects to target priority areas (GSBAs) for special management. More detailed examination of these Globally Significant Biodiversity Areas (GSBA) has yielded numerous new species, exemplifying the predictive power of the system. Meanwhile Star and GHI rankings have become part of the common parlance of forestry in Ghana at all scales (tree selection in logged forest; forest EIA — e.g. for the Subri industrial plantations company, seeking low priority parts of Subri FR for Gmelina plantation), and conservation is correspondingly a routine issue in Ghanaian foresters' minds.

lasting land management.

Some of the classified forests in Côte d'Ivoire, however, do have an important role to play with respect to biodiversity: many forest species are confined to one or a few forests only (Kouamé et al. chapter 5). The same accounts for the forest reserves in Ghana (Hawthorne 1996) and probably also in Liberia. In Ghana forest reserves are thought to reasonably well represent the variety of forests available a century ago (Hawthorne 1996), but special conservation measures ("GSBAs", Globally Significant Biodiversity Areas) have recently been instigated by the Ghana Ministry of Lands and Forestry within that portion of the reserved forest with the higher proportion of rare plants, recognising their priority status. Work is underway to demarcate and manage sensitively a modest network of these specially protected GSBAs within the much more extensive and long standing network of forest reserves. The GSBA priorities were selected (using methods described by Hawthorne and Abu Juam (1995) for their high biodiversity conservation priority), with a view to better and sensitive conservation of high priority species, but within a broad concept of ongoing forest management and use. In Liberia effective measures for forest protection are difficult due to the unfavourable political situation over the last decade, but also, construction of national parks with effective measures are currently being planned (e.g. for the Sapo National Park area, D. Byler, J. Suter and E. Waitkuwait personal communications). Much additional work needs to be done in this field before conservation and protection efforts will be effective.

Table 8.1 also shows the percentages of area of protected forests in different forest types. In Ghana the majority of protected forests is in moist semi-deciduous and in moist evergreen forests, with a considerable part also in wet evergreen. In Côte d'Ivoire the wet evergreen forests are well represented with c. 40% in the third type.

Forest corridors and animal conservation

Many plant species depend in their life history on animals, be it for pollination, for dispersal of their seeds or otherwise. For effective conservation in the long term not only conservation of plants is needed but also of the animals that are connected to them. Hunting is widespread in the region and leads to severe declines in animal populations (e.g. Caspary et al. 2001). Many large animals have large home ranges and as a result of fragmentation of forest, the resulting forest patches may be too small to sustain a viable animal population. Thus for effective protection of plant species and their related animals sizeable forest tracts are needed. An alternative is a combination of smaller patches connected to each other with corridors, forming ecological networks. Networks that link forest reserves and classified forests are particularly interesting in this respect. Such corridors are needed as linkages for animals and plants to cross areas to maintain a larger population that might be more viable. Parren et al. (2002), for example, plea for a network

Figure 8.3 Distribution of stock of commercial timber out of the selected species of Chapter 4 (Table 4.1). The number of individuals > 30 cm dbh are shown, together with their interpolation for the whole area. Notice that this is a potential map based on earlier inventory data, and not the actual situation.

0«<t i Lbn? ■ w










1 Vfl-KBO




Figure 8.3 Distribution of stock of commercial timber out of the selected species of Chapter 4 (Table 4.1). The number of individuals > 30 cm dbh are shown, together with their interpolation for the whole area. Notice that this is a potential map based on earlier inventory data, and not the actual situation.

Figure 8.4 Tree bark is sold on the market of Man, Côte d'Ivoire.

directed to forest elephants. They focus on the border areas between Ghana and Côte d'Ivoire (Box 8.1).

Also, some plant species have special value for animals as keystone species. Keystone species provide food for animals when overall food availability is low. Fig trees are known for their overall importance as food source for a large variety of animal species. Such individual species thus need special attention. On the other hand, animals can be extremely important for plants as well, in terms of pollination and seed dispersal for example. The forest elephant is the best-known example (see Box 8.1).

Areas that have high value in terms of plant diversity are not necessarily important for their animals. Logically those areas that have biodiversity values for a large number of plant and animal groups will have a larger chance of being effectively protected. To be able to characterise different areas in Upper Guinea with respect to overall biodiversity, in 1999 Conservation International brought together a large number of specialists. Together they produced a biodiversity hotspot map for terrestrial ecosystems for the whole of Upper Guinea, based on priority areas for mammals, birds, reptiles and amphibians, insects, and plants. For plants the extremely high conservation priority areas were: Liberia Hygrophilous Evergreen Forest, Tai'-Haute Dodo area, Mount Nimba, Gola, Ghana Wet Evergreen Forest. Several of these areas are also hotspots for animals (see maps in Conservation International 2001). The Tai'-Haute Dodo-southeastern Liberia area is important for plants, mammals, birds, reptiles and amphibians. The same accounts for the Gola area in western Liberia, and for the Mount Nimba region. Banco National Park is a hotspot for reptiles and amphibians and insects, but less so for mammals and birds. The Guinea highlands of Sierra Leone and Guinea are of high importance for mammals, the Loma mountains in Sierra Leone stand out in terms of insects. Taking all five groups into account, the overall (regional) conservation priority areas for Upper Guinean forests are found in (from west to east) the Fouta Djalon, the Gola-Lofa-Mano complex, Mount Nimba Range, Tai'-Grebo-Putu Range, the Cestos-Upper Krahn Bassa area, Mount Peko, Abi Lagoon-Cape Three Points area, Ankasa Jema, Krokosua-Sui and finally the Ghana-Togo highlands.

Forest resources and use

The forests of West Africa have been used extensively for their resources. Assuming that plant composition is indicative for the whole forest ecosystem we can use the forest classification as indicative for the whole system. We thereby assume that a broad cover of vegetation types also implies a broad cover of other forest elements (e.g. animal communities). This means that our overall ecological stratification is based on forest types. By linking for instance forest use to forest types we can predict specific forest uses over a large area. For many areas the main resources were timber. For the forests that have been inventoried for a selection of species (see Bongers et al. chapter 4) we have calculated the total number of commercial trees larger than 30 cm diameter per square km (note that the data used are mostly pre-logging inventory data and thus not indicating the actual but rather the potential timber stock). This stock of commercial species is an indication of the importance of these forests as a timber resource. Figure 8.3 shows that the concentration of timber species was high in Liberia, southwest Côte d'Ivoire (specially Haute-Dodo and Haut-Sassandra) and southwest Ghana. We expected the moist and semi-deciduous forests to be the ones that have most of these trees. However, also some of the wet evergreen forests and hyperwet forests had large stocks of timber species as well (compare Figure 8.3 with Figure 4.2C). This is partly because the moist semi-deciduous forests have been heavily depleted over the last century, much more so than the evergreen forests.

When we compare this potential timber map (Figure 8.3) with the diversity map (Figure 6.7) it is clear that there is no simple link between the potential timber stock and the biodiversity. A general relation exists, however: the lower the timber density is, the higher the diversity. This is also shown for Ghana alone (Hawthorne & Abu Juam 1995). At regional level the negative relation between timber density and species density implies that timber use and biodiversity conservation are, in a broad sense, compatible options, based on spatial segregation. However, some concessionaires only have forests in the evergreen forest areas, and will seek to maximise the use of their concession even if on a national level this is not ideal for a conservation perspective. Comparison of this map with the forest classification shows that on a local level some hyperwet areas have high densities of timber species, but also some dry semi-deciduous or wet evergreen forests.

For local people non-timber forest products are important as well. Bushmeat, for instance, is highly used and its utilisation is currently one of the most important conservation challenges in the tropical moist forest region, both in Upper Guinea as in Central Africa (Bakarr et al. 2001). In West Africa, bushmeat collection has led to drastic decline in many species, and populations of many species are near the brink of extinction. For instance, the primate subspecies Procolobus badius waldroni (Miss Waldron's red colobus) disappeared completely from previously occupied localities in Ghana and Côte d'Ivoire (Bakarr et al. 2001). That hunting indeed is severe in most West African forests is shown by Caspary and co-workers (Caspary et al.2001): during 1996 alone, 35.5 million animals, with an estimated weight of 120,000 tons of carcasses, were consumed in Côte d'Ivoire.

At a smaller scale plants or plant parts are used for construction, artisan products, and for the preparation of medicines (Figure 8.4). For Ghana the non-timber products are listed in Abbiw's (1990) book. For two forests in Côte d'Ivoire, FC Haut Sassandra and FC Scio, Tra Bi (1998) has shown that a total of 304 species were used, of which 65 as sources of food, 73 for construction purposes, 99 for artisanal craftworks, 14 for hunting and 9 for fishing. No less than 182 species were used for medicines. This exemplifies the local importance of many forest plants. Integrated management guidelines are needed to be able to manage the available resources wisely. For bushmeat the blueprint provided by Bakarr et al. (2001) seems a good way ahead.

Long-term conservation and use

The results in this book, and the areas prioritised for conservation, are in line with the recommendations of an international priority-setting workshop in Ghana in 1999. During that workshop a total of 41 regional priorities sites were identified from Guinea to Togo, covering approximately 235,000 km2 of land across six countries (Conservation International 2001). This is equivalent to 23% of the land in those countries. This includes 55,000 km2 of the estimated lowland evergreen forest remaining in Upper Guinea (Sayer et al. 1992b). The priority areas include all existing habitat types, from mountain forest to coastal mangrove, and from moist evergreen to semi-deciduous forest.

For a better and more effective management of Upper Guinean forests, goals directed at conservation need to be combined with goals on short-term and long-term use by local people and state organisations. Needs of local people, and also their best indigenous knowledge of biodiversity and biological management, have to be taken into account. The currently widely advocated approach is to enable local communities to participate in the management of forests.

In most West African countries, education curricula are weak in biodiversity information. Against this background, effective environmental education that combines the best of indigenous and modern information is perhaps the single most important activity in ensuring long-term biodiversity conservation. Regional cooperation in this respect may improve use of collective knowledge regarding proper management.

The challenge of biodiversity conservation is a dynamic one. Although this book focuses on species and forests, this challenge is as much about people. Integrated management plans are necessary, cutting across sectors in

Box 8.3 Conclusions and recommendations

Plant diversity hotspots

• Three prime areas of high plant species diversity are distinguished: a belt 50-100 km inland from Sierra Leone, through Liberia to southwest Côte d'Ivoire; the wet evergreen forests of southwest Ghana, and the area around Abidjan. These are also rich in species of relatively restricted distribution.

• Very rare and very local species, worthy of the forest managers' particular attention, occur specially in five areas: the Liberian coast, around Abidjan, Mount Nimba, southwest Côte d'Ivoire and southwest Ghana.

• The supposed glacial refuge areas (Mount Nimba region, Cape Three Points, Cape Palmas) deserve special conservation attention.

• Hill sanctuaries, as established in Ghana, may provide a good framework for planning general protective management throughout the region.

Forests from a regional perspective

• The regional forest classification presented in this book is a valuable addition to the country-based ones developed earlier. Local forests now can be put in a regional context. This is of high relevance for regional forest conservation and management.

• Hyperwet forest areas in Liberia and Sierra Leone need extra focus because of their regional scarcity.

• Complementarity is a good principle in selection of (additional) sites for forest and species conservation. In Côte d'Ivoire an extra forest reserve (National Park) in the eastern evergreen forests is highly recommended.

• Corridor networks in two areas would be highly valuable to facilitate migration, especially of some animals. One in the southwest of Côte d'Ivoire and border with Liberia, the other in southwest Ghana and border with Côte d'Ivoire.

• Large forest blocks in southeast Liberia are heavily undersampled with regard to herbarium collections. Collection expeditions are highly needed.

Forest condition and preservation

• The condition of the protect

Oplan Termites

Oplan Termites

You Might Start Missing Your Termites After Kickin'em Out. After All, They Have Been Your Roommates For Quite A While. Enraged With How The Termites Have Eaten Up Your Antique Furniture? Can't Wait To Have Them Exterminated Completely From The Face Of The Earth? Fret Not. We Will Tell You How To Get Rid Of Them From Your House At Least. If Not From The Face The Earth.

Get My Free Ebook

Post a comment