^{Avifauna species richness and diversity in Mount Elgon National Park, Uganda}

Abstract
 

      Following from previous studies of Mount Elgon avifauna, this study aimed to assess species richness and distribution of
        birds in the Piswa area. It was found that low numbers of forest specialist birds were in the area, although overall the
        proportion of forest-dependent (i.e., forest specialist, forest generalist and forest-visiting birds) was typical. This
        reconciled apparently conflicting evidence from studies by Katende et al. (1989), who found low numbers of forest
        specialists on Elgon, and Dickinson et al. (1996), who found high numbers of forest birds.

        The low numbers of specialist species were thought to be due to the effects of grazing on vegetation, which has been
        found to alter the species composition of plant communities in the Piswa area. It is suggested that grazing
        be reduced in the area, with provision for the needs of local people, and that further research, on a larger scale and
        including ringing of birds, be carried out before further action is taken at Piswa.
 

Birds are often one of the best groups of ‘indicator species’ used when assessing the ecological status of a given area under study. However, in the case of Mount Elgon, most bird surveys have only determined the presence or absence of species, with arbitrary mention of their ecological status (e.g., Dickinson et al., 1996, eds.; Katende et al., 1989). Based on such surveys, a number of recommendations have been made, mainly concerning the fact that low numbers of species characteristic of undisturbed forest have been found (e.g., Katende et al., 1989). However, in the only major survey carried out since the area achieved national park status in 1993, Dickinson et al. (1996, eds.) found a high proportion of forest dependent birds on Mount Elgon, including those which have a restricted range within Uganda, such as Hunter’s cisticola, Cisticola hunteri.

This study, as part of Project Elgon (1997), aimed to clarify some of these aspects of the ecology of the Mount Elgon avifauna, and to use the information gathered in mist-netting and visual observation to make some recommendations about future management of birds and other wildlife in the Park. The Piswa area, alt. 2850m, in the Kapchorwa district at the northern end of the mountain was concentrated on for a study which, although smaller in scale than those mentioned above, aimed to assess the status of the birds in an area which is currently under intensive grazing pressure (Reed and Clokie, submitted) from cattle and other livestock.
 

Mist nets of 12m in length or 2 x 6m in length, with four shelves, were erected at four sites in the Piswa area, and run for 2 weeks each. Although close together (see map), each of the sites were quite markedly different in the relative amounts of plant species they contained. They were:

Site 1: montane bushland, dominated by the shrub Sambucus africana

Site 2: wooded grassland, dominated by the shrubs Clutia abyssinica and Erica arborea, and the tree, Afrocrania volkensii

Site 3: disturbed bushland, dominated by the shrubs Sambucus africana and Helichrysum sp., and the tree, Dombeya sp.

Site 4: transitional shrubland/forest, dominated by a mixture of trees including Dombeya sp., Afrocrania sp., Juniperus sp., Hagenia abyssinica, and Podocarpus sp., as well as large amounts of the shrub, Plectranthus laxiflora

A more complete breakdown of all the plant species and their relative amounts at each site is given in the appendix. It should be noted that it was very difficult to find sites with sufficiently intact understorey to use for mist-netting. This problem, due to intensive grazing, has been reported in previous studies also, e.g., Katende et al. (1989). All sites were at the same altitude of 2850m, to rule out variation in species richness due to altitude. Locations as given by GPS readings are shown on the map in Figure 1.

Effort was made to spend the same amount of time at each net site, allowing for weather conditions. On average, it was not possible to keep the nets open for more than four hours per day. Visual observations were made throughout the area on most days, to monitor those species which could not be caught in mist nets.

Birds caught in nets were measured as follows: weight, body length, tarsus length, wing length, wing aspect ratio, eye colour, beak shape, beak colour, and where possible, sex. Other identifying marks were recorded when a species was first caught. Species were identified using two field guides: Williams and Arlott (1980), and van Perlo (1995).
 

A total of 24 species were caught in the mist nets, with a further 13 being observed only. Of the 24 species caught, 15 of these were also visually observed in the area. Table 1 shows the list of species found in the survey, together with their status as forest generalists, specialists, visitors, migrants, or non-forest species, and the sites at which they were found.

The species caught in mist nets comprised 101 individuals in all. This low number is typical for Elgon, where lower catch rates than other areas have frequently been reported (e.g., Matthews, 1996; in Dickinson et al., 1996, eds.). As Table 1 shows, some were found exclusively at one site, whilst others were found at two, three or all four sites. With the exception of one species, the thick-billed seedeater, Serinus burtoni, all species found at more than one net site were forest generalists or forest visitors. This conforms to the theory that, generalists being more widely distributed, are more likely to be found in disturbed habitats than specialists (Pomeroy and Ssekabiira, 1990).

Only five forest specialists were found, and not all of these were highland forest specialists. The encroachment on habitat lower down the mountain may be responsible for this: non-highland specialists may have moved upwards to find suitable habitat. Given that van Heist (1994) reported 144 species of forest bird on Mount Elgon, i.e., 44% of the total for Uganda, this is a very poor figure indeed. Further to this, Dickinson et al. (1996, eds.), reported that, over the whole mountain, 25% of bird species were forest specialists.

However, in the Piswa area, this study shows that only 13.5% of the species encountered were forest specialists, with the vast majority being generalists or non-forest species. Forest generalists comprised 32.4% of all species found, forest visitors comprised 29.7%, with the remaining 24.3% of species encountered being non-forest species. In other words, forest-dependent species comprised a total of 75.7% of all species encountered. Given the close proximity of the Benet grasslands, however, this is not surprising. Whether these grasslands have always been in the area is not known, but they are certainly kept at an artificial climax level by grazing cattle, goats, and donkeys (van Heist, 1994). Figure 2 shows the percentage of the total number of species caught and/or observed at Piswa in each ecological category. This graph shows starkly the very low levels of forest specialist species, which when compared with the data for the mountain as a whole, does not bode well for the birds of the Piswa area. Following this, it was felt that some indication of how well birds were adapting to each type of habitat, most of which were disturbed, should be found. This was done using the data from species found at more than one net site.

Although in many cases the birds were not sexed during capture, and there are often significant intersexual differences in the weights of a species, it was hoped that, by taking the mean weight of a species at each of the sites it was found, and testing for significance, some measure of how well species were doing in each area could be obtained. Excluded from this were the juvenile and subadult white-starred forest robins, Pogoncichlastellata, which were found at sites 1 and 2. It was, however, interesting to note that no adults of that species were found at sites 1 and 2, nor were any juveniles or subadults found at site 3, where the adults were found. Table 2 shows the results of that analysis.

On the subject of species richness, the most species rich sites were sites 2 and 3, with 16 species each. Some of these occurred at both sites. Site 1 had 7 species captured in nets, while at site 4, only 5 species were captured. No species found at site 4 were exclusive to that site; all occurred at least one other site. Only one of the site 1 species was exclusive to that site, the grey-headed sparrow, Passer griseus, an open habitat species.

On the whole, the Piswa area was species poor, which is clear despite the small scale of this project. Particularly worrying is the low number of forest specialist birds, and the encroachment by non-forest species. The reasons behind this, and possible solutions to the problem, are discussed below.
 

It can be seen from the data presented that the low number of forest specialists, and the low number of species encountered as a whole, was most likely due to the disturbance of the vegetation in the area. Site 4 is a classic example of a disturbed forest/bushland transitional area (Pratt, 1997, pers. comm.), given the large number of different tree species in one area. An undisturbed area would be much more uniform. Since not all species of birds that live normally at lower altitudes could adapt when forced to make a move upwards following encroachment of their habitats lower down, this is not surprising. However, the large proportion of forest generalists and visitors indicate that some encroachment on what was formerly the habitat of specialists has occurred, the vegetation structure now being more suited to such species.

The species distribution between sites can be explained in part by the type of vegetation present in each case. At site 1, the dominant plant was the shrub, Sambucus africana, a species which appears to be ubiquitous in the Piswa area. Near to this was cultivated land and some human dwellings, in the form of the park rangers’ hut. The bushland in this area was frequently disturbed, on a more or less daily basis, by cattle, goats, and donkeys. A combination of the nearness of human habitation, the disturbance from livestock, and the species-poor plant composition, may all contribute to the low number of species found at site 1. All species netted at site 1 were forest generalists and visitors (see Table 1), or open habitat species.

However, site 4 was even poorer in species than site 1. It was set back a little from trails which were frequented by both livestock and humans. Being transitional shrubland-forest, it should support a wider variety of species than were found. The most likely reason for this not being the case is the lack of any dominant tree species in the area. The mixture of trees indicated a disturbed area, and that those species found there would be generalists, or species dependent on the dominant shrub, Plectranthus laxiflora. Again, this theory was reinforced by the fact that the five species caught at site 4 were all forest generalists and visitors (Pomeroy and Ssekabiira, 1990). Indeed, previous studies show more generalists than specialists being caught on Mount Elgon, despite efforts to find the most undisturbed areas of forest (e.g., Dranzoa and Rodrigues, in Katende et al., 1989; Matthews, in Dickinson et al., 1996 (eds.)).

Taking the two most species-rich sites, sites 2 and 3, assessing the vegetation can again help to determine why some species were present and others absent. Site 2, the wooded grassland, lay close to the Benet grasslands, where many cattle graze daily, but on the other side of the site was forest. Having a variety of trees, with the dominant species clearly being Afrocrania volkensii, and two dominant shrubs, Clutia abyssinica and Erica arborea, this site may have been less disturbed than sites 1 and 4. Other factors, relating to the capture efficiency of the mist nets (Jenni, Leuenberger, and Rampazzi, 1996), may also have contributed to differences in species richness between sites. However, there was little doubt that all four sites are disturbed as regards vegetation.

Site 3, the disturbed bushland, provided some of the most interesting species, such as the northern double-collared sunbird, Nectarinia preussi, the rare black-billed weaver, Ploceus melanogaster, and the scaly-throated honeyguide, Indicator variegatus, one of the few non-passerine species encountered. Livestock were frequently found grazing right by the net at this site, and most birds were probably caught passing from one patch of trees behind the site to another patch near camp. The apparent richness of the site may therefore have been an artefact of the placing of the net to maximise capture efficiency (Jenni et al., 1996). The position of the net, understorey cover, sunlight, and wind speed can all affect capture efficiency, and this may be one reason why only 9 individuals were caught at site 4, whilst 33 were caught at site 3, when both sites were disturbed.

Again, for the visual observations, both this study and others (e.g., Katende et al., 1989; Dickinson et al., 1996 (eds.)), found far fewer specialists than generalists. In this study, only one of the visually observed species, Hartlaub’s turaco, Tauraco hartlaubi, was a specialist species. This, perhaps more than the data from the mist-netting, indicated how widespread forest generalists and visitors are on Mount Elgon. While generalists are usually more widely distributed than specialists (Pomeroy and Ssekabiira, 1990), this particular area is alarmingly poor in specialist species. Mist-netting tends to help locate some of the more elusive species, but since the visual observations of conspicuous species also show few specialists, the patterns from the mist-net data are confirmed by such observations.

Some further explanation of the vegetation structure in the Piswa area should be given at this point, to help clarify the levels of disturbance involved. The Project Elgon team found in 1996 that the vegetation in the Piswa area, which was cultivated until the evictions of local pepole in 1983 and 1990, showed a decreased species richness compared with non-cultivated areas (Reed and Clokie, 2000). The heavily grazed Benet grasslands have a high diversity of grassland plant species, many of which are resistant to grazing and in nutrient poor soils. Some grassland species persist among the shrubs which then take over. After 15-20 years, the nutrient status of the soil begins to improve. This means increased appearance of shrub and tree species, decreasing amounts of forest edge and pioneer species, and an increase in climax species (Reed and Clokie, 2000). With increased grazing, species composition changed. This is most conspicuous in mature forest plots. Grazing history also played a role in the composition of such areas: some areas ceased to be under very intense grazing pressure in 1983; in other areas, this pressure continued until 1990, following the final evictions. Of the areas that had experienced reduced grazing since 1990 only, those most heavily grazed at present had a higher species richness. This may seem to be advantageous rather than detrimental to the plant community and the animals (including birds) that depend on it. However, grazing means that young trees are not common under the canopy; rather, the grazing resistant grassland species, as explained above, are more likely to be found. This lack of young trees is compounded by dense stands of monocarpic Acanthaceae, such as Mimulopsis alpina. The Project Elgon team concluded previously that ‘cultivated and grazed land begins to regenerate after humans have stopped cultivating and stopped/reduced grazing. The succession of species which follows is affected by previous land use, for an unknown number of years’ (Reed and Clokie, 2000).
 

Such an analysis does not bode well for forest-dependent bird species, as is reflected by the findings of this study. Furthermore, although most of the species encountered were forest generalists and forest visitors, these species are also dependent on trees, if not on particular species of trees. Therefore, it is now possible to clarify the apparently conflicting findings of Katende et al. (1989) and Dickinson et al. (1996, eds.). The data from this study concur with the findings of Katende et al. (1989), in that very few specialists are supported on Mount Elgon, due to the disturbed state of the vegetation. However, the generalist and visitor species found were all still forest-dependent species, many being further subdivided into highland forest species. This concurs with the findings of Dickinson et al. (1996, eds.), in that Elgon has a relatively high proportion of forest-dependent birds.

Based on this conclusion, it is suggested that the levels of grazing in the Piswa area could be reduced to such a level that it is under no more pressure than other areas of the park. Using grazing areas on a rotational basis may be one way to resolve this problem, to take into account the needs of local people as well as those of wildlife. Needless to say, more and larger studies (including ringing of birds) need to be carried out on the effects of grazing on bird species richness before any plan of action can be implemented. What is certain, however, is that the birds of Piswa are existing in a far from ideal situation.
 

Jenni, L., Leuenberger, M., and Rampazzi, F. (1996). Capture efficiency of mist nets with comments on their role in the assessment of passerine habitat use. J. Field Ornithol. 67 (2): 263-274.
Katende, T., Ipulet, P., Rodrigues, R., and Dranzoa, C. (1989). Sustainable development and forest conservation in Uganda. IUCN Technical report No. 1: Birds and Woody Perennials Inventory, Mount Elgon Forest Reserve.
Matthews, R. (1996). In: Dickinson, C., Davenport, T., and Howard, P. (eds.) (1996). Mount Elgon Biodiversity Report (No. 1). Chapter 3: Birds. Uganda Forest Department, Kampala.
Pomeroy, D. and Ssekabiira, D. (1990). An analysis of the distributions of terrestrial birds in Africa. Afr. J. Ecol. 28: 1-13.
Reed, M. S. and Clokie, M .R. J. (submitted). Effects of Grazing and Cultivation on Forest Plant Communities in Mount Elgon National Park, Uganda, African Journal of Ecology 38 (2): 154-162.
Van Heist, M. (1994). Land Unit Map of Mount Elgon National Park. IUCN Technical Report, unpublished.
Van Perlo, B. (1995). Birds of Eastern Africa. Harper Collins, London.
Williams, J. G., and Arlott, N. (1980). Birds of East Africa. Harper Collins, London.

Project co-ordinator: Caroline McParland
Field work team: Caroline McParland and Opus Otuko Joseph Charles
Project advisor: Mr. P. Doyle, Mr. R. Kityo
Report Writer: Caroline McParland