Although the division in forestry terminology between "minor" and "major" forest products reflects the bias of foresters managing forests for timber, the slow growth rates of forest hardwoods still makes "major forest products" a useful category, particularly where the same species is used for different purposes in different age classes.

As discussed above (page 13), forest use differs greatly from the use of low species diversity, highly productive vegetation types such as Phragmites wetlands or Cymbopogon thatch-grass, where harvesting is seasonal and easy to manage, with obvious clear cutting of suitable stands.

Bwindi Forest is at the opposite end of the scale. Instead of the 1-year rotation applied in reed cutting (Cunningham, 1985), sustainable harvesting of forest for timber aims at a rotation of 100 years (Leggatt and Osmaston, 1961). Tree growth rates are slow, so unlike reeds which regrow from an underground rhizome within a year, the time between final harvesting and replacement of mature trees is seldom less than 50 years, and often as much as 200 years (for oak trees in Europe for example).

On the basis of growth measurements from Afromontane tree species in southern Africa, Stapleton (1955) considered that under natural conditions, growth time to marketable maturity for timber production from Podocarpus latifolius was 230 years, Ocotea bullata 220 years and Olea laurifolia, 200 years.

Logging, particularly when mechanized, affects mature forest, changing forest structure and species composition, resulting from canopy gap formation (Howard, 1991). In addition to differences caused by topography and soil type, logging results in patchy distribution of both species and size class categories of trees and shrubs used for bellows, building poles, beer boats or bean stakes. It also influences the availability of these resources, either increasing the number of young saplings (bean stakes, building poles) or colonizing species in disturbed sites (e.g. Polyscias fulva, Maesopsis eminii), or decreasing stocks of large hardwoods due to over-exploitation and competing uses for timber.

Competing uses

This situation is complicated further through competing uses for young trees or saplings of the same species, particularly as Bwindi Impenetrable Forest is the most heavily pitsawn of all principal forests in Uganda.

To foresters, whose objective is hardwood timber production, saplings of "reserved species" (Table 14, page 41) represent regenerating timber trees. To people from local rural communities they also represent an important source of beer boats (>50 cm dbh), building material (5-15 cm dbh) or bean poles (1.5-5 cm dbh), with high density wood favoured due to greater resistance to borer attack or fungal infection.

For this reason, despite their "reserved status", hardwood trees such as Newtonia buchananii (omutoyo) and Prunus africana (omumba) are, and probably always have been, favoured for beer boats, and Ocotea usambarensis (omwiha) for building poles.

Cutting of bean stakes is selective for size rather than species, but if saplings of canopy species fit into this category, then they are cut (e.g. cutting of Strombosia scheffleri (omuhika) and Ocotea usambarensis (omwiha) in secondary forest during this survey). Not only are these species a useful resource to local people, they also represent the future forest canopy of the next century.

Trees with low density wood or small shrubby species are not generally subject to such competing uses, however. Examples are:

1. Low density timber species used for beer boats but not for timber or building poles (such as large Ficus species such as Ficus sur (omulehe) and F. ovata (ekyitoma));
2. Secondary forest colonizers Polyscias fulva (omungo) and occasionally Musanga leo- errerae (omutunda) used for blacksmiths' bellows but not for other uses due to the soft wood.
3. Bean poles cut from understorey shrubs such as Psychotria schweinfurthii (omutegashali), but not for timber or beer boats, and rarely for building poles.

Wood requirements and supplies

In terms of supply and demand for fuelwood and building poles, the situation in the DTC area is similar to that described by Howard (1991) for the Bwamba and Bajonjo counties in western Uganda, north-east of Bwindi Impenetrable National Park. Population densities are high, and land-holdings in intensively cultivated landscape are similar - 0.2 ha per person in the DTC area (Kanongo, 1991); 0.26 and 0.19 ha per person respectively in Bwamba and Bakonjo counties (Howard, 1991).

In Bwamba, Howard (1991) calculated that the 121,600 people (17,000 households) would require about 151,000 m³ of fuelwood and 4600m³ building poles every year (on the basis of a fuelwood consumption rate of 1.24 m³ per person per yr and a building pole requirement of 0.27 m³ per household per yr or 0.038 m³ per person per yr). The DTC area is occupied by a similar number of people (over 99,000 people, 19,000 households).

This situation has been worsened by the rapid clearing and burning of indigenous forest for agriculture outside Bwindi Impenetrable National Park. In 1954, approximately 120 km² of forest remained in the DTC area within a 15 km radius outside the Bwindi Forest boundary. By 1972 this had been reduced to 42 km² of forest, and by 1983, less than 20 km² remained (Butynski, 1984). Now, apart from forest patches in less densely populated parishes such as Nteko, virtually nothing remains.

Removal of forest outside of Bwindi Impenetrable National Park has reduced supplies not only of fuelwood and building poles, as Howard (1991) points out, but also of natural forest as a source of bean stakes, which although relatively small in diameter are needed in vast quantity, and beer boats, which although required in smaller numbers are large trees, some of which are species already over-exploited by pitsawyers (Newtonia, Prunus) or which are "keystone species" favoured by frugivorous birds and primates (Ficus, Prunus). Through this process, the national park has become a focal point of harvesting pressure.

Wood resources availability

For beer boats large trees are used (usually >50 cm dbh), and most beer boats are greater than 40 cm in diameter (see Figure 4). Density of large trees suitable for beer boats is low in the multiple-use area, judging from field observation, and a 1 ha plot carried out in an Allanblackia - Syzygium guineense dominated forest site selected as representative of the forest. This had a high number of beer boats per ha, due to the flat terrain and absence of any pitsawing activity there in the past.

Despite this, only 3 (2.6%) of the 114 trees >30 cm dbh (or 4.6% of the trees >50 cm dbh) were suitable for beer boats. With the exception of moist sites with a high density of Ficus trees, pitsawn sites would be expected to have an even lower density of potential beer boats. Growth rates of most tree species favoured for beer boats are unknown for this area. It is likely, however, that Ficus sur and Ficus ovata would be expected to reach the minimum tree diameter (50 cm dbh) suitable for beer boats in 20-30 yr, and Prunus africana or Newtonia buchananii in 40-50 yr, both of which greatly exceed the average life-span of most beer boats (9 yr). With improved road networks and urbanization, marketing of banana beer, and therefore demand for beer boats can be expected to increase.

Building poles come from trees intermediate in size between those used for bean stakes and beer boats. Resource assessments evaluating trees within 20 x 20 m plots for suitability for building purposes on the basis of durability, diameter and straightness, showed that there were fewer building poles per ha (Figure 8) than resource users anticipated from visual assessments made before quantitative work was done. Visual assessments of resource availability made with local people reported in Scott (1992) therefore need to be considered with caution.

Figure 8. Combined number of tree stems in 2 cm size-class intervals for four and three 20 x 20 m plots (combined total of 0.28 ha) in secondary forest in the Ishasha gorge and Ngoto swamp areas respectively, showing the number of stems in the size-class range preferred for building purposes (5-15 cm dbh).

In the seven plots surveyed, only 20.8% (8.3) poles per plot were classed as very good for building poles and 52.7% (21) poles per plot were accepted (see Figure 5, page 35). Although additional plots are required, in the absence of other data this would indicate an average density of 207 very good poles per ha, or a total of 525 useable poles per ha.

Building pole cutting was not widespread, and despite the big demand for poles, high intensity harvesting was limited to a few patches in secondary forest. This is attributed to the widespread cultivation of Eucalyptus and black wattle, and the less labour-intensive harvesting from these cultivated tree species.

Counts in 20 x 20m plots in fields of climbing beans in the Rubuguli and Nteko areas during this survey showed that there were 48,000-52,000 bean stakes per ha of climbing beans. By comparison, two forest plots in Alchornea hirtella (ekizogwa) dominated forest understorey favoured for bean stakes (and the highest density of plots sampled) (Figure 9, page 44), contained 479-630 per 20 x 20m plot, or approximately 12,000-16,000 bean stakes per ha, less than half as many as are required for climbing bean cultivation.

Figure 9. Data from four 20 x 20m plots in secondary forest in the Rushaga area, showing selectivity and high proportion of stems cut for bean stakes in Alchornea hirtella (ekizogwa) dominated understorey (Plots 1 and 2), and low degree of harvesting in less favoured sites (Plots 3 and 4).

Bean stakes last 2-3 seasons, and beans are a major crop in the Rubuguli, Nteko, Rushaga, Nteko and Nyamabale areas. Total demand for bean stakes in the DTC must represent millions of saplings per year.

Two additional points are significant here. First, distribution of Alchornea hirtella dominated stands is patchy, and density of bean poles in the surrounding areas is significantly lower (3000-6000 bean stakes per ha). Second, the Alchornea hirtella patches are already heavily utilized, with 58% (429) of stems cut in the plot with the highest density of bean stakes, and 35% (186) of stems cut in the adjacent, lower density plot .

Cultivation of wood resources

Shortages of fuelwood, building poles and bean stakes are being experienced in the DTC area, and a shortage of large trees for beer boats can be expected in the future. Reasons for wood scarcity, and solutions to the problem are recognized by local people.

Trees (particularly Eucalyptus) are planted in the DTC area, mainly for building poles, but are inadequate to meet either the existing or the future demand (Table 15, page 45). The need for building poles was also the main reason for tree planting in Bwamba, but it was considered to provide only 328 m³ of the total annual demand of the 151,000 m³ of fuelwood, 4600 m³ of building poles and annual increase in demand for timber of 5280 m³. It is likely that a similar situation exists in the DTC project area.

Elephant grass (Pennisetum purpureum) and trees (particularly Eucalyptus) are also planted for bean stakes, while Ficus cuttings are planted for beer boats (recorded in the Ngoto area) and canoes (Lake Bunyonyi). In some areas, even hardwood timber trees have been planted, with Entandrophragma (omuyovi) reaching a dbh of 90 cm within 40 years (Photos 21 and 22, page 45). Such local initiatives need to be recognized and encouraged.

Photo 21. Mr. K. Byarugaba, second generation Entandro-phragma (omuyovi) planter with one of six 1 year old trees at his homestead in the Ngoto area .		Photo 22. Entandrophragma excelsum (omuyovi) planted in 1950 at the same homestead by the late father of Mr. K. Byarugaba.