5. ORNITHOLOGY.

 

By Jens Otto Svendsen, Louis A. Hansen, Jon Fjeldså, Marcel C. Rahner, Louise B. Pedersen, Henrik Kisbye, Erik Edvardsen and Jacob Kiure.

 

 

5.1. Abstract.

 

The Uluguru Mountains (including foothills) were considered the fourth most important area in East Africa for the conservation of rare forest birds by Collar and Stuart (1988). Five globally Threatened (one of them endemic), three globally Near-threatened (one of them endemic) and six other forest bird species of extremely restricted-range are known from the Uluguru forests.

 

Whereas the Uluguru North F.R. had been visited before by various workers and was known to hold all above-mentioned species (except one of the Near-threatened), the Uluguru South F.R. (by far the largest reserve in the Ulugurus) was largely unknown ornithologically except for visits in the Tchenzema and Bunduki areas on the drier western slope and a short visit in the Nyingwa area on the eastern slopes. These visits revealed the presence of two of the Threatened species in the Uluguru South. It was furthermore uncertain whether certain rare species (especially Swynnerton’s Robin Swynnertonia swynnertoni and Dappled Mountain Robin Modulatrix orostruthus) occurring in the Usambaras and the Udzungwas were genuinely absent from the Ulugurus, or had just been overlooked.

 

This chapter presents the results of ornithological fieldwork carried out in the Ulugurus between 1 October and 15 December 1993. Results from a two days visit in November-December 1994 are also included. The main aim was to describe the bird assemblage of the mountain forests of the Uluguru South and North Forest Reserves using standardised and repeatable techniques and to assess the population status of the two Uluguru endemics (Uluguru Bush Shrike Malaconotus alius and Loveridge's Sunbird Nectarinia loveridgei). Standardised fieldwork was carried out along two altitudinal gradients on the eastern slopes of the Uluguru South F.R. (altitudinal range visited: 1450-2640 m) and along one gradient on the eastern slopes of the Uluguru North F.R. (altitudinal range visited: 1100-1950 m). Brief visits were paid to the western slope of the southern section (Tchenzema in the Uluguru South, 1950-2500 m, and Bunduki F.R., mostly plantations at 1220-1540 m), to two localities on the northwestern slopes of the Uluguru North F.R. (Morningside at 1500 m and Kigurunyembe at 650-850 m) and to the lowland Kimboza F.R. (300 m).

 

The altitudinal distribution of all forest species occurring in the Ulugurus is described, supported by quantitative data from the 13 stations shown on Figure 1.4 and Figure 1.5. Ecological data are given where appropriate. Based on the results from our survey in combination with a study of literature we examine the geographical and altitudinal distribution in the Ulugurus of bird species of special conservation interest (Threatened, Near-threatened and other restricted-range) and the value of the Ulugurus for such species. A summary of earlier fieldwork and recommendations for further fieldwork are given. Our study has enabled us to point at the remaining submontane areas on the eastern slopes of the Uluguru North F.R. (largest areas in the Tegetero-Bagilo-Kinole-Lupanga area on the eastern slopes and also some around Morningside in the northwest) as being the most important to focus conservation efforts on.

 

The taxonomy used in this section follows Collar et al. (1994) for Threatened and Near-threatened species. For all other species we follow Dowsett and Dowsett-Lemaire (1993). For species where the name used in Dowsett and Dowsett-Lemaire differs from the most commonly used in East African ornithology (e.g. however not  for Green Coucal instead of Yellow-bill), the latter can be found in Appendix 5.5.

 

For many bird species, it is difficult to determine whether it is a forest or a non-forest species. In this report, we have chosen to regard species, which use other habitat types (woodland, bracken, gardens etc.) to a very high degree as non-forest species, admitting that some of them are characteristic elements of the forests (especially the Kimboza Forest).

5.2. Introduction, part 1: The ornithological significance of the Uluguru forests (also see Table 5.1).

 

In the ICBP/IUCN Red Data Book Threatened birds of Africa and related islands (Collar and Stuart 1985) 97 Afrotropical species (species from the tropical part of mainland Africa) were treated as Threatened. Of these, a high number (63, i.e. 65 %) are forest dependent species (Collar and Stuart 1985).

 

The importance of the Uluguru Mountains for forest birds has long been known. In a review of key forests for the protection of threatened bird species in Africa by Collar and Stuart (1988) the Uluguru Mountains (including foothills) ranked 16th among all forests on the African continent and fourth among all forests in East Africa in terms of conservation value for the protection of Threatened and Near-threatened bird species. The Ulugurus are furthermore an important part of C24, one of the 221 priority areas for global conservation listed by ICBP (1992).

 

Table 5.1 shows the distribution and conservation status of all forest bird species of special conservation importance occurring in Moreau's (1966) Tanganyika-Nyasa Montane Group (Figure 1.1), the geographical limits of which largely fit C24 as illustrated in ICBP (1992). The Ulugurus hold five Threatened forest bird species. In the Tanganyika-Nyasa Montane Group only the Udzungwa and the Usambara Mountains have more. Some of the Threatened species in the Usambaras are restricted to lowland forest, which shows affinity to coastal forests. It is clear that the Eastern Arc is the richest part of the montane group, with no less than 12 Threatened bird species being endemic to this very limited range of isolated forest-capped mountains.

 

Five Threatened and three Near-threatened species occur in the Ulugurus (global conservation status categories with capitals from Collar et al. 1994).

 

Two of these species are endemic to the Ulugurus: the Uluguru Bush Shrike Malaconotus alius (Threatened) which is a distinctive species occurring at low densities, and Loveridge's Sunbird Nectarinia loveridgei (Near-threatened) which is a valid species in the Nectarinia regia superspecies (Hall and Moreau 1970), and very common.

 

The other Threatened and Near-threatened bird species known from the Ulugurus are: Mrs Moreau's Warbler Scepomycter winifredae (Threatened), White-winged Apalis Apalis chariessa (Threatened), Banded Green Sunbird Antreptes rubritorques (Threatened), Tanzanian Mountain Weaver Ploceus nicolli (Threatened), Southern Banded Snake Eagle Circaetus fasciolatus (Near-threatened) and Uluguru Violet-backed Sunbird Anthreptes neglectus (Near-threatened).

 

Six species, which are of restricted-range (estimated breeding range of less than 50,000 km2) but not Threatened or Near-threatened occur in the Ulugurus. These are: Spot-throat Modulatrix stictigula, Red-capped Forest Warbler Orthotomus metopias, Chestnut-headed Apalis Apalis chapini, Sharpe´s Akalat Sheppardia sharpei, White-chested Alethe Alethe fuelleborni and Kendrick’s Starling Poeoptera kenricki.

 

All five Threatened species, the Near-threatened Nectarinia loveridgei and four of the other restricted-range species (Modulatrix stictigula, Orthotomus metopias, Apalis chapini and Sheppardia sharpei) were in the Ulugurus known only from the mountain forests (forests excluding lowland forests and outlying hills) before our survey and our results did not change this. Thus the mountain forests, of which almost the total area is made up by the Uluguru North and South F.R.s, are the most important for the bird species of special conservation importance. Before this survey it was, however, not known which parts of the mountain forests it is most important to protect.

 

The second Near-threatened species (Circaetus fasciolatus) is known only from the foothill forests, of which the best known is Kimboza Forest. The third (Anthreptes neglectus) is most common in the foothill forests but with an occurrence also in the lowest part of the Uluguru North F.R. Two species of restricted range (Alethe fuelleborni and Poeoptera kenricki) are clearly most common in the mountain forests but occur in the foothills at least seasonally. The bird faunas of the foothill forests in many respects show strong similarities to the threatened lowland forests of coastal Tanzania and certain other Eastern Arc ranges. Thus, the foothill forests are also of significant conservation importance.

 

Five subspecies of montane forest birds are endemic to the mountain forests of the Ulugurus (Stuart and Jensen 1985, Stuart et al. 1993). These are: Andropadus nigriceps neumanni (Eastern Mountain Greenbul ssp.), Sheppardia sharpei bangsi (Sharpe's Akalat ssp.), Apalis thoracica uluguru (Bar-throated Apalis spp.), Orthotomus metopias altus (Red-capped Forest Warbler ssp.), and Phylloscopus umbrovirens fugglescouchmani (Brown Woodland Warbler ssp.). Two of these endemic subspecies are of restricted-range species (Table 5.1). A case may exist for regarding Andropadus nigriceps neumanni as a separate species, as it is the most aberrant of all forms in this polytypic species. The number of endemic subspecies is - with our current knowledge - relatively high for the Ulugurus compared to other mountain ranges as is evident from Table 5.2. This could have to do with the isolated eastern position of the mountains.

 

Only a single specimen exists of the geographically isolated Uluguru population of Abyssinian Crimson-wing Cryptospiza salvadorii (normally regarded a mountain forest species, see Appendix 5.5). The nearest population of this species is in the Kilimanjaro area. Further investigation may show that the Uluguru population of C. salvadorii represents an undescribed subspecies.

 

 

5.3. Introduction, part 2: Earlier ornithological survey work in the Uluguru Mountains (also see Appendix 5.1).

 

Friedmann and Stager (1964) gave a review of earlier ornithological fieldwork in the Ulugurus. Also Stuart and Jensen (1985) briefly gave details on earlier fieldwork in their review of the Uluguru avifauna. In Appendix 5.1 we draw the data from these two publications together, adding new details.

 

Focusing on the localities we visited (see Table 2.1 and Figures 1.4 and 1.5), we can conclude that the eastern slopes of the Uluguru South F.R. were unknown ornithologically except for some collecting activities by Arthur Loveridge around Nyingwa October 1926 (we suppose his Nyange/Nyingwe/Nyingwa localities are south of Lanzi). The Tegetero-Bagiro area in the Uluguru North F.R. had been visited before by various field workers. The Bunduki area had been visited by Stager for some time in 1964 and also by others (but for some of the visits it is not clear whether they visited the Bunduki F.R. or nearby forest in the Uluguru North and South F.R.s). Kimboza, Tchenzema, and Morningside had also been visited before whereas Kigurunyembe was unknown.

 

Most of the visits to the forests before 1965 were to collect specimens for museums, more recent visits were general surveys. The work described in the following is the attempt to make a quantitative description of the bird species composition along altitudinal gradients, primarily using standardised methods, and to describe the distribution of species of special conservation interest.

 

 

5.4. Methods used on this survey (also see Table 5.3).

 

It is difficult to define a single standardised quantitative measure for the composition of the forest bird community. Factors that complicate the matter are:

 

·                               Some species are very noisy, vocalising incessantly while feeding, or singing much of the day. Others give only one or a few song strophes at daybreak and are silent the rest of the day.

·                               Some species skulk in dense understorey all day, other species forage in the mid and upper strata of 40 m tall forest, and a third group move between all strata.

·                               Some species forage alone, some in pairs, some in mono-specific flocks and some frequently in multi-species feeding parties moving over long distances.

·                               Some species are very common while others are extremely rare.

 

No standard manual exists for comprehensive surveys of Afromontane forests. On this survey we used four complementary methods (see 5.4.1-4 below) at each of the three main study sites (Kimhandu, Lanzi, Tegetero). All methods give a relative, not absolute, measure of abundance. Brief surveys were carried out at other localities. Our efforts are detailed in Table 5.3. In Appendix 5.12 we comment on the efficiency of the methods applied and on the completeness of our survey.

 

JF, LAH, JK and JOS were experienced with visual and vocal identification of Eastern Arc montane forest birds from earlier fieldwork. The rest of the ornithology team visited the Eastern Arc for the first time but had trained identification by studying specimens and by listening to tape recordings.

 

 

5.4.1. Mistnetting for two full days (also see Tables 5.6-5.8 and Appendix 5.12).

 

Aim. To obtain a relative measure of the abundance of lower strata species, especially the most skulking ones and to obtain standard biometric data and blood samples of the birds for later analysis of evolutionary relationships of Eastern Arc montane forest species. Mistnetting was furthermore an important part of the training carried out during the survey.

 

Technique. Usually one full day was used to put up the mistnets. The following two days they were opened just before the onset of the first bird activity (usually around 5.30 a.m.) and kept open until about half an hour before darkness (usually to around 6.15 p.m.). They were normally checked every hour but more frequently in the early morning or when the weather conditions were not optimal. The nets were closed during the night to avoid entangling of bats and destruction by nocturnal ground dwelling mammals. No attempt was made to mistnet nocturnal birds since the only night bird heard was the Wood Owl Strix woodfordii.

 

The length of the net series varied from station to station, depending on topography, available manpower and general impression of the bird density. We generally placed the nets where they would be least visible, but attempted to cover all microhabitats such as shrubs, clearings and areas with mainly open forest floor. Clearings were usually covered by placing the nets just inside the forest edge since a fully visible net in a glade may have a low catch rate. In a few cases nets were run in glades if these were a prominent habitat type but their catch rate was comparable to nets run in denser forest at the same stations. The nets were placed in small groups of rows and/or angles. The net chains were placed some distance apart and not parallel in order to avoid "shadow effects". The nets were usually within 150 m from the camp where the birds were processed.

 

All birds caught were bagged, brought to the camp and ringed (enabling us to distinguish between the individuals captured), and most were measured and scored for brood patch and moult status (following recommendations in Baker and Baker 1990). Birds, which were wet or caught late in the evening, were released immediately, however. All mistnetting data (1135 records) have subsequently been computerized into the database of the East African Ringing Scheme and are available on disks. Some 250 of the birds were bled for DNA samples, which can be done without noticeable extra stress or damage.

 

For each individual we noted in which of the nets it was caught. This has enabled us to compensate for non-standardised efforts at some stations where nets were put up specifically to catch e.g. Scepomycter winifredae (including the captures of these nets would lead to an overestimation of the density of that species), or where some of the nets were run for a little more than two days.

 

 

5.4.2. One-hectare plots (also see Table 5.9 and Appendix 5.12).

 

This method, which is adapted from the variable circular-plot method (Reynolds et al. 1980), is almost identical to a procedure developed for South African forest by Koen (1988) and has been used by JF in a wide range of tropical habitats (e.g. Fjeldså 1993, Fjeldså and Rabøl in press). The method, which aims at obtaining a quantitative measure of the abundance of all non skulking forest species, is adapted to the facts that:

 

·                               Territory mapping methods are too time-consuming for comparison of a large number of study sites (stations).

·                               Steepness and dense undergrowth often makes it impossible to quietly follow a continuous route required for a transect.

·                               The detectability of many (but not all) forest birds drops sharply at 30-50 m distance (Reynolds et al. 1980), which makes point counts difficult to use.

 

Aim. To obtain a quantitative measure of the abundance of all non-skulking species.

 

Technique. Study plots of 100 x 100 m extension (visual judgement) were selected within 50 m elevational range from the mistnetting station (in very few cases in steep terrain: 100 m). Within each plot observations were made for exactly ten minutes, during which time an attempt was made to find every bird in the plot by walking quietly through it. Before starting the counting the observer learned its borders and edges by spotting characteristic elements in the surroundings (without walking around to visit the borders). We attempted to select the plots representatively throughout the local habitat spectrum but the ruggedness of the terrain admittedly caused some constraints. In cases where the terrain was too rugged or shrubby to move quietly for inspecting hidden parts, a vantage point with a good overview of the plot was used.

 

Plots were assessed only between 7 and 10.30 a.m. or after 4 p.m. The reason to wait until after 7 o'clock was to avoid the burst in song activity just after dawn. Plot assessment ceased when bird activity was low, and on certain days we had to stop counting as early as e.g. 9 o'clock, in one case even at 7.40 a.m. The activity after 4 p.m. was usually too low to allow for assessment. No plots were assessed in rainy weather, and on the few mornings with sporadic mist, assessments were made only when the whole plot was visible.

 

Most montane forest bird species are heard far more often than seen, and the method requires that the observer has good skills in identifying birds by their songs, calls and warning notes. The majority of the plots were assessed by those of the participants who were familiar with the voices from earlier fieldwork in the Eastern Arc montane forests. Those of the participants who were in the Eastern Arc for the first time did not make plots until they were familiar with the voices. Normally the observer was alone but in a few cases one or two persons joined him to learn the technique. In the latter case only the main observer was allowed to detect the birds - individuals which this person overlooked were omitted in order to ensure comparability.

 

15 to 25 plots were assessed per station. Each one-hectare square was covered only once. In the results section we will restrict ourselves to count each species only once per plot, also in cases were more than one individual was noted. Birds, which evidently were flushed from one plot to the next, were counted only once.

 

At the Kimhandu-5 (2520 m) station the 10 m tall forest was extremely dense, making it impossible to find any plots that could be easily assessed from the forest interior, and the plots were therefore assessed by observations from an adjacent meadow.

 

 

5.4.3. Tape recordings at dawn (also see Table 5.10 and Appendix 5.12).

 

Aim. To augment species lists obtained by mistnetting and plots.

 

Technique. At every ringing station two tape recordings (two mornings), each of 30 minutes duration were made. The recording was started at the onset of the first bird song at dawn. They were made close to the mistnet group furthest away from the camp. The microphone (a Sennheiser MD-21 U [ball characteristics]) was placed about 160-180 cm above ground. The tape recordings were listened through twice after returning to Denmark.

 

 

5.4.4. General field observations (also see Tables 5.4 and 5.5 however do also see Appendices 5.5-5.8 and Appendix 5.12).

 

Aim. To record Threatened species which often occur at such low densities that they are overlooked by the above mentioned methods. Furthermore to record the complete altitudinal range of all species.

 

Technique. At each locality we searched the area as thoroughly as the time allowed, in order to locate rare species. Where possible we attempted to cover the entire elevational gradient. Altitudes were measured with a Thommen altimeter. Special interest was paid to mixed feeding parties where these occurred (they turned out to be mostly restricted to the northern section) in order to find especially the Threatened species Apalis chariessa, Ploceus nicolli and Anthreptes rubritorques (all known to occur in the Ulugurus). Other rare sunbird species which are not known from the Ulugurus but could potentially occur there also sometimes join such parties.

 

Notes on efforts of general field observations. The altitudinal gradient in the Kimhandu area was surveyed from the lowest forest edge (c. 1450 m) to the highest point in the area (Nongwe, c. 2634 m) but only little time was spent in the altitudinal ranges below 1520 m and from 2200 to 2450 m. In the Lanzi area the gradient from the lowest forest edge (c. 1560 m) up to 2220 m was surveyed relatively well (with least effort laid below 1710 m) whereas the gradient from 2220 up to 2500 m was studied only during a very brief visit to the Lukwangule Plateau. In the Tegetero area the lower part of the flat area (from c. 1050 to c. 1250 m) was visited briefly twice and there could be species not recorded in this large and interesting area. The highest point visited in the Tegetero area was a pass at 1960 m altitude near Luhungo but forest can be found up to 2270 m (Mount Nziwane).

 

 

5.5. Results.

 

Table 5.4 lists the altitudinal records of forest species from our survey (all localities). The earlier knowledge on altitudinal distribution of the species is listed too. Records of non-forest species are listed in Appendix 5.6.

 

Table 5.5 illustrates altitudinal records in a different way than Table 5.4: Presence/absence of species within +/- 50 m elevational range of the 12 stations at our three main localities.

 

Tables 5.6-5.8, 5.11 and 5.12 show numbers of birds mistnetted at our various main localities.

 

Table 5.9 shows which species were recorded on the plots.

 

Table 5.10 shows which species were recorded by tape recording at dawn.

 

Figure 5.1 ranks the relative abundances of species at our 12 stations as shown by mistnetting for two full days and plot counts.

 

Figure 5.2 shows the number of forest species ringed per station (mistnetting for two full days) towards number of individuals ringed.

 

Figure 5.3 shows the number of forest species recorded per station on plots towards number of plots assessed.

 

Abbreviations used for locality names in Tables 5.4-5.10:

 

Kim/Kimh.:                            Kimhandu

Lan.:                                           Lanzi

Teg.:                                           Tegetero

Tchen./Tchenz.:                   Tchenzema

Kimb.:                                       Kimboza

Kigur.:                                       Kigurunyembe

Bund.:                                       Bunduki

 

Abbreviations used in Sections 5.5.1-4 and Appendix 5.5 to indicate by which methods a species was recorded:

 

M             = Mistnetting.

P               = One-hectare plots.

T              = Tape recordings at dawn.

G              = General field observation.

 

 

5.5.1. Observations of the five Threatened species (also see Appendix 5.3).

 

Below we list observations from our survey. The general geographical distribution of the five Threatened species is described in Appendix 5.3 with special notes on the Ulugurus. The sizes of the Uluguru populations of the five Threatened species is discussed briefly in Section 5.6.1.

 

Mrs Moreau's Warbler Scepomycter winifredae (M, P, T, G). A minimum of 33 different territories were identified at Kimhandu, Lanzi, Tegetero and Tchenzema. No standardized measurements of territory density were obtained. However, because of the powerful song, it could be determined that in certain well suited areas (Kimhandu at c. 1800-2000 m, Tchenzema at 2000-2200 m and Lanzi at 1700-1900 m) there was generally 2-300 m between the territory centres. This figure is a very rough estimate and should not be used for calculations of population size or strict comparisons with other mountain ranges but can serve to give the reader an idea of the density in some places. The densities appeared to vary greatly. Below we give the altitudes of all territories recorded during the survey.

 

At Kimhandu we recorded the species near the Kimhandu-1 (1520 m) camp, where a single territory was encountered (altitude: 1700 m, well north of the Msuluzi river) and in the areas surveyed from the Kimhandu-3 (1940 m) station where a minimum of nine territories were identified and the species was heard frequently (territories at 1700, 1820, 1850, 1880, 1920, 1920, 1960, 1960, 2020 m). At Kimhandu-3 it was furthermore scored on two of the 22 plots assessed, and a pair was mistnetted (both birds simultaneously) in nets put up selectively to obtain a blood sample from the species.

 

At Lanzi the species appeared to be more widespread. A minimum of 15 territories were identified (altitudes: 1660, 1670, 1710, 1710, 1770, 1780, 1860, 1900, 1900, 1920, 1920, 1940, 2000, 2070 and 2160 m). Four birds were mistnetted at the Lanzi-1 (1710 m) station. These four individuals are believed to belong to the two territories listed for 1710 m altitude. Two of the birds are included in the two days standard effort (mistnetted with one hours interval and probably a male and a female; caught in a light gap where we had put up a net to cover this habitat type which was prominent around the camp). The third individual was mistnetted inside forest in a net that had been run for more than two days and the fourth was caught in nets put up selectively (in the other territory near the camp) to obtain extra blood samples from the species. The species was scored on one of the 16 plots that were assessed at Lanzi-1 (1710 m) and on one of the 15 plots that were assessed at Lanzi-3 (2110 m).

 

At Tegetero only three territories were identified (altitudes: 1475, 1550-1600 and 1740 m), and the species appears to occur rather scattered below 1950 m in this area (1950 m was the highest altitude we visited here).

 

East of Tchenzema c. six territories were encountered (c. 2000, 2050, 2150, 2280, 2370 and 2430 m). Two of these (2000 and 2050 m) were in heavily disturbed forest near the lower forest edge. The ones from 2370 and 2430 m were in larger forest glades right below the scarp of the Lukwangule Plateau. The species was recorded on four of the 25 plots assessed at Tchenzema (the territories at 2000, 2050 and 2150 m altitude, all in 10-15 m tall forest).

 

Notes on the ecology. Most of the pairs we recorded were found in large natural forest glades with dense herbs covering the ground. Many of the light gaps were big and spectacular with lots of creepers, herbs and other epiphytes hanging down from tall trees in the edges or from single trees in the middle of the light gaps. Also Williams (1951) reported the species to be found in such habitats. In a few cases we found the species in light gaps with the ground mainly covered by bracken, in some cases the territory was in an area containing a mosaic of twenty-five meter tall forest and small light gaps with up to two meter tall herbaceous vegetation, and in other cases (Tchenzema) the species was found in thick, herbaceous vegetation in heavily disturbed, low forest. Stuart and Jensen (1985) reported the species to be particularly associated with dense undergrowth in small clearings and along valleys. The male and the female sing a duet consisting of loud and very characteristic whistles. Most of the territories mentioned above were identified only by hearing this duet, the species was rarely seen. More types of song as well as the warning call were recorded and will be presented on a tape (see also Svendsen&Hansen 1992.). To the best of our knowledge, no tape recordings have been published yet of this species. The rather few visual observations of the species during our survey were almost exclusively of pairs travelling through the understorey together.

 

White-winged Apalis Apalis chariessa (G). The species was recorded only at Tegetero. A female was recorded at 1260 m altitude in a mixed feeding party, and the following day at 1300 m 2-3 birds, including at least one male, were seen in a mixed feeding party. All birds were seen high up in the canopy.

 

Banded Green Sunbird Anthreptes rubritorques. Not recorded on this survey. This small canopy and edge species was not recorded on our survey which is not surprising, taking into consideration how rare the species appear to be here and how little time there was available for field observations. Very little fieldwork was carried out below 1250 m and forest edge habitats were not searched.

 

Uluguru Bush Shrike Malaconotus alius (P, G). The species was not seen at any of the localities visited in the Uluguru South F.R. and after having learnt its distinct and far-carrying voice at the next locality (Tegetero) we believe that it was not heard at other localities than Tegetero. A more thorough search for rare species in the Kimhandu and Lanzi (and Tchenzema) areas would probably reveal the presence of M. alius, but, if present, it must occur at very low densities.

 

At Tegetero a minimum of four home ranges/territories were identified. These are called A, B, C and D below.

 

Territory A. At 1320 m altitude a single individual was seen under good conditions on three occasions in a certain small forest glade (5, 6 and 7 December) some 500 metres from our Tegetero-1 (1345 m) camp. On the first occasion (5 December) it was in the canopy of trees bordering the gap, from 15 m height in the lower part of the canopy of an edge tree to 25 m height in the top of other trees, generally keeping to the upper part of the canopy. When seen 15 m above the ground it foraged in a cluster of leaves where also a Many-coloured Bush Shrike M. multicolor was seen the following day. On the second occasion (6 December) it was seen 25 to 30 m above the ground in the upper part of the canopy. This bird was seen eating a big winged insect, possibly a cicada. On the third occasion (7 December) it was seen 15 m above the ground in the lower part of the canopy. On 3 December a single individual was heard at a plot (Table 5.9) which must have been at least 400 m from the above-mentioned light gap but we have chosen to regard this individual as a bird of the same home range.

 

Territory B. At 1520 m altitude (a ridge, much closer to our Tegetero-1 [1345 m] camp than to our Tegetero-2 [1535 m camp]) an individual was heard on 4 December.

 

Territory C. At 1500-1535 m near our Tegetero-2 (1535 m) camp. On several occasions an individual was calling from the upper canopy for a long time without moving into view (a habit shared with M. multicolor). Its distinctive and loud call was tape-recorded and will be presented on a tape (Svendsen in prep.). To the best of our knowledge, the species has never been tape recorded before. On several occasions another bird, most likely a mate, answered the call. Although heard on several occasions (often at some distance, however), the species was seen on only a single occasion in this area. The visual observation is of a bird foraging and calling in the canopy of a 30 m tall tree. A Waller's Red-winged Starling Onychognathus walleri rested peacefully near its nest hole five meters from the bush shrike. The bush shrike had remained practically motionless (and therefore invisible) but frequently calling in the canopy of the same tree for at least 10 minutes before appearing in the sunlight. On two occasions where the observer followed the species for hours through the forest (being able to locate it from time to time from its call but unable to spot it), a M. multicolor (also vocally active) appeared to follow approximately the same route, some 25 m from the M. alius. They moved so slowly, however, that it was difficulty to determine whether there was any interrelationship between them. M. alius was scored on one of the plots assessed at the Tegetero-2 station (Table 5.9).

 

Territory D. At c. 1710 m altitude a single individual was seen and heard on a plot (Table 5.9). This is the only observation of the species in this area despite some time was spent there.

 

Tanzanian Mountain Weaver Ploceus nicolli. Not recorded on this survey. P. nicolli was not recorded during our survey. This is not surprising in the light of its elusiveness and the short time available for doing general field observations.

 

 

5.5.2. Observations of the three Near-threatened species (confer Appendix 5.4).

 

The general geographical distribution of these three species is described in Appendix 5.4 with notes on earlier Uluguru records.

 

Southern Banded Snake Eagle Circaetus fasciolatus (G). A single juvenile bird was seen in Kimboza Forest during the two days spent at Kibungu Chini by JF and JK.

 

Uluguru Violet-backed Sunbird Anthreptes neglectus (G). A single individual was seen by JF during a two days visit in Kimboza.

 

Loveridge's Sunbird Nectarinia loveridgei (M, P, T, G). Clearly the most abundant bird at every mountain forest locality visited as evidenced from the mistnetting and plot data (Figure 5.1). N. loveridgei made up no less than 271 (28.3 %) of totally 959 birds mistnetted at Kimhandu, Lanzi and Tegetero (Table 5.6). Altitudinal distributions recorded at the main localities were: Kimhandu 1520-2580 m, Lanzi 1685-2475 m, Tegetero 1200-1960 m, Tchenzema 2000-2500 m.

 

At many of the stations where N. loveridgei is the most common species, the mistnetting data express a much bigger difference in abundance between N. loveridgei and other species, than the plot data do. (Figure 5.1). A thorough test of these two methods, including exact methods like territory mapping, has not yet been undertaken for East African forests, and it is therefore difficult to say which of them gives the most correct impression. N. loveridgei is very mobile, always on the move and a rapid flyer, and is therefore very prone to fly into the mistnets. Mistnetting may therefore overestimate the abundance of the species in relation to other species. The plot data (as used in this report) can potentially underestimate the abundance of very common species in relation to other species: only the percentage of plots with the species present and not the abundance on the plots is illustrated for a given station in the histograms.

 

The relatively low numbers at Tegetero-1 (1345 m) and Lanzi-3 (2110 m) need a comment: At Tegetero-1 (1345 m) only 8.4 % of the birds mistnetted were N. loveridgei. This should be seen in the light of the abundance of N. olivacea which is the most abundant sunbird at that station as shown also by the plot data. At Lanzi-3 (2110 m) N. loveridgei occurs at only low density according to the mistnetting data (12.5 % of all birds mistnetted) and further supported by general field observations. It is not clearly supported by the plot data. The vegetation was quite open and dry with some bamboo at Lanzi-3 (2110 m), perhaps with less flowering plants.

 

Notes on ecology. During our survey N. loveridgei was most often seen feeding on nectar or on insects by gleaning twigs and leaves. On a few occasions they were seen flycatching, a behaviour that was also observed on a single occasion by Williams (1951). The birds are very vigorous, flying much around, and appear to be temperamental. Williams (1951) stated that it is one of the most adaptable of the Uluguru forest birds, the species being “often seen about native plantations where trees have been left standing, being especially attracted to an Albizzia-like tree when this is in flower. It is also much in evidence in native pea cultivation, feeding at the flowers of this crop". He states that the species ”remains fairly common both in the remaining forest and outside”. At Ukwama village we saw it outside the forest in small stands of trees and shrubs but it was our clear impression that the species was largely restricted to forest habitat at the time we visited the Ulugurus. It may feed outside the forest edge at other times of the year if attracted to flowering plants. We did not hear the species in wooded areas between Kimboza and the montane forests, as we did with N. olivacea. There are no records from Kimboza Forest of N. loveridgei, and given the enormous population in the montane forests this indicates that the species does not undertake significant vertical migration to the lowland forests. Indications of breeding activity: see Appendix 5.11.

 

Only a single other sunbird species was recorded in forest interior, namely N. olivacea. At Tegetero-1 (1345 m) N. olivacea was more common than N. loveridgei as clearly illustrated by the mistnetting and plot data (Figure 5.1). N. olivacea was mistnetted also at Kimhandu-1 (1520 m) and Tegetero-2 (1535 m) and scored during plot assessments at Tegetero-2 (1535 m) and at Tchenzema (2150 m), but N. loveridgei is by far the most abundant at these stations. It is remarkable that the abundance of N. olivacea has declined so much from the Tegetero-1 (1345 m) station to the Tegetero-2 (1535 m) station. The reason may be that the latter station is much deeper inside the forest and near the steep slope in the western part of the large flat area at Tegetero.

 

5.5.3. Observations of restricted-range species (other than Threatened and Near-threatened).

 

White-chested Alethe Alethe fuelleborni (M, P, T, G): The results of our survey show that this understorey species is common in the Uluguru North and South F.R.s, especially in the submontane and lower montane forest. The mistnetting data provides the best indication of this very shy, rarely seen and not very active singing species which furthermore seldom gives warning calls if a person passes through its territory. For comments on its occurrence in Kimboza Forest, see Appendix 5.8.

 

Sharpe's Akalat Sheppardia sharpei (endemic subsp. bangsi) (M, P, T, G): We recorded this understorey species frequently between 1200 and 2140 m, with most observations being from below 1800 m. The highest densities encountered were clearly those in the Tegetero-1 (1345 m) area (Tables 5.8 and 5.9, Figure 5.1). Also at the Tegetero-2 (1535 m) and Kimhandu-2 (1710 m) stations the species was well represented. It is striking that it was not mistnetted in the Lanzi area and at the two lowest stations in the Kimhandu area. The record from Kigurunyembe is of a single bird heard. There were no earlier published records of this species from the Uluguru South F.R.

 

Chestnut-headed Apalis (Chapin’s Apalis) Apalis chapini (P, T, G): This canopy and mid stratum species is generally common throughout the forests except at the highest altitudes (see plot and tape recording data).

 

Red-capped Forest Warbler (African Tailorbird) Orthotomus metopias (endemic subsp. altus) (M, P, T, G): O. metopias is a common understorey species at all our stations in the Uluguru North and South F.R.s (see mistnetting, plot and tape recording data). According to the plot data it is less common at Kimhandu than at Lanzi and Tegetero but this is not supported by the mistnetting data. O. metopias, which is normally very common in montane and upper montane habitats, was scored on only one of the 25 Tchenzema plots. This is a remarkably low score since the vegetation mosaic at Tchenzema offers a high density of suitable microhabitats (e.g. herbaceous scrub and leaf-rich edges in light-gaps). We heard some O. metopias from small forest patches passed in open country (less than 500 m from "the main forest") at 1500-1700 m.

 

Ripley and Henrich (1966) mention a record of O. metopias at 1000 m from "West Uluguru". Stuart and Jensen (1985) stated this to be unlikely since they were not aware of any forest at this altitude on the western slopes but Figure 1.2 shows that there is still some forest left at this altitude. Ripley and Heinrich are, however, known to have been imprecise with altitudes in the Usambaras in some cases (see Stuart 1983).

 

Spot-throat Modulatrix stictigula (M, P, T, G): This understorey species is very common in the Uluguru North and South F.R.s, occurring along the entire altitudinal gradient. Especially high densities were observed around the Kimhandu-3 (1940 m) camp by general field observations. See also Appendix 5.8.

 

Kendrick’s Starling Poeoptera kenricki (P, G): Heard and seen on only a few occasions during this survey and is probably uncommon in the Ulugurus, being clearly outnumbered by Onychognathus walleri. Stuart and Jensen (1985) stated that there were no earlier records from the southern section but Friedmann (1928) actually mentioned a female collected at Nyange (=Nyingwa according to Stuart and Jensen [1985]) by Loveridge in 1926. See also Appendix 5.8.

 

 

5.5.4. Observations of other forest species.

 

Notes from our survey are given on all other forest species known from the Ulugurus in Appendix 5.5, with some discussion. Our observations of non-forest species are listed in Appendix 5.6 (a few of the species being rather characteristic elements of part of the forests).

 

 

5.5.5. Notes on community structure and species interactions.

 

Since these observations are also of importance for the understanding of conservation matters, notes are given on

·         Observations of mixed feeding parties from the survey. See Appendix 5.7.

·         Observations that can help to understand the phenomenon of seasonal vertical migration. Our own observations are supplemented with earlier from the literature in an attempt to preliminary assess the importance of the lowland Kimboza Forest for species breeding in the mountain forests. See Appendix 5.8.

 

 

5.6. Discussion.

 

 

5.6.1. Population sizes of Threatened and Near-threatened species (confer Figure 1.6).

 

Below we discuss the population sizes of Threatened and Near-threatened species in the Ulugurus, based on the information in Section 5.5.1 and Appendix 5.3 (for Threatened species) plus Section 5.5.2. and Appendix 5.4 (for Near-threatened species). The geographical distribution in the Ulugurus (current knowledge) of bird species of special conservation interest is shown on Figure 1.6.

 

Mrs Moreau's Warbler Scepomycter winifredae. In several papers (e.g. Stuart and Jensen 1985), a rather vague term as e.g. ”fairly common” has been used for this species. This is most understandable since there are no exact density estimates to refer to for this species, which is clearly the most abundant of the Threatened species occurring in the Ulugurus. However, we take the opportunity to stress that B. winifredae occurs at relatively low densities compared to largely all other understorey species found in the Uluguru montane forests. This is well illustrated by our mistnetting and plot data - four of the six individuals mistnetted (probably representing three pairs) were caught in nets put up specifically to catch this species, and it was recorded on only few of our plots. On the other hand, the population in the Ulugurus (and to judge from our survey results perhaps especially the Uluguru South F.R.) is doubtless one of the densest within the range of B. winifredae, and the Ulugurus should certainly be regarded a key area for the survival of this Threatened species. B. winifredae depends completely on forest for its survival but may tolerate considerable disturbance in the forest interior since clearings are its natural habitat. A good estimate of the population size in the Ulugurus cannot be given at the present stage.

 

White-winged Apalis Apalis chariessa. A. chariessa has never been recorded in the Uluguru South F.R. It cannot be excluded that it occurs in that section but it seems unlikely that the Uluguru South is anything but a sink habitat: The species generally prefers forest below 1600 m (though recorded up to 2000 m in the Udzungwas on a single occasion, record in Dinesen et al. 1993) but this habitat type is largely missing in the southern section. Furthermore, at least in the Udzungwa Mountains and in the Uluguru North F.R. A. chariessa is seen almost exclusively in mixed feeding parties (e.g. Stuart and Jensen 1985, Stuart et al. 1987, Dinesen et al. 1993, Moyer 1993, this survey, LAH and JOS pers. obs. from the Udzungwa Mountains 1994) and the apparent lack of mixed feeding parties of the drongo party type in the Uluguru South F.R. (see Appendix 5.7) probably further reduces the quality of the Uluguru South for A. chariessa.

 

The observations by Stuart and Jensen and Moyer in the early 1980’s (Appendix 5.3, Stuart and Jensen saw several individuals at 1250-1400 m) indicate that the species is relatively easy to find below 1500 m in the Uluguru North F.R. if a number of feeding parties are studied in the dry season. However, Andersen and his collectors apparently never succeeded in collecting specimens of A. chariessa (they have collected specimens of the other four Threatened species occurring in the Ulugurus). During our survey we recorded the species only twice but we suffered from lack of time for doing general field observations (for a discussion of the completeness of our survey and the methods we used: see Appendix 5.12) and the mobility of feeding parties was low. A good estimate of the population size and density in the Ulugurus cannot be given at the present stage. Significant populations of this species may exist only in the Udzungwas but the observations by Stuart and Jensen indicate that the population in the Uluguru North may be the largest outside the Udzungwas.

 

Banded Green Sunbird Anthreptes rubritorques. Since mid-altitude forest is generally absent today in the southern section, any remaining populations in the Ulugurus are most likely in the Uluguru North F.R. More fieldwork is necessary before we can say anything about the Uluguru population size of this species, which is known from only five specimens. In the Usambaras the species gathers at flowering and fruiting trees inside forest, in edges and even some distance from forest. Further search for the species in the Ulugurus should focus on the lowest part of the Uluguru North F.R., be carried out in the flowering/fruiting season and include visits in the surroundings of the forest.

 

Uluguru Bush Shrike Malaconotus alius. The time consumption of the standardized methods and logistics restricted us from searching as thoroughly for the species as we had hoped. The conclusion to draw from our observations (four territories located at Tegetero and none at the other localities) is that the status of this very vocal species on the eastern slopes of the Uluguru South F.R. is uncertain but that it may exist at very low densities here. It appears to be more common in the Uluguru North F.R., at least in the submontane and lower montane belt. M. alius has been recorded above the submontane belt by earlier workers (e.g. at 2100 m at Tchenzema, see Appendix 5.3) and may use the montane belt frequently, but the fact that this species was not recorded in the Uluguru South F.R. during our survey, whereas four territories were identified in the Uluguru North F.R. between 1320 and 1710 m, suggests that the core habitat is the submontane and lowest montane zone. Practically no forest exists below 1500 m in the south and although the species is known from higher altitudes, the presence of submontane and lower montane forest may be essential for maintaining the population. The Uluguru North F.R. may well prove to be the stronghold of the species.

 

Also most earlier records are from the Uluguru North F.R. but this could be an artefact of the more intensive ornithological activity in this area (the Uluguru North F.R. is easier accessible from Morogoro and has a wider altitudinal span of forest, including the luxuriant submontane areas, thus making it more attractive for ornithologists in lack of sufficient time to visit several localities). The altitudes of Andersen’s specimens are generally 1800 m but that is the case with almost all his specimens which may well have been collected over a wide altitudinal range (see Appendix 5.1).

 

The population size is unlikely to exceed 1000 individuals. Intensive searches of large areas in different kinds of habitat would be necessary to refine this very rough estimate, which is based on few observations. Such a search would also clarify which types of habitat this endemic species prefers.

 

Tanzanian Mountain Weaver Ploceus nicolli. The three Uluguru records are all from the Uluguru North F.R. below 1800 m (possibly below 1600 m).

 

Ploceus nicolli frequently takes part in mixed feeding parties like (and often together with) its congener Dark-backed Weaver Ploceus bicolor. It has been recorded in feeding parties at least at Mazumbai (several observations; in the West Usambaras), in Mwanihana Forest (several observations; eastern slope of the Udzungwas), the Nyumbanitu Mountains (c. four observations, unpublished, JOS 1994; in the Udzungwas) and the Ulugurus (one occasion) (e.g. Stuart and van der Willigen 1978, Collar and Stuart 1985, Stuart and Jensen 1985, Stuart et al. 1987). However, in the Ndundulu Mountains (in the Udzungwas close to the Nyumbanitu Mountains) the species has only been seen singly or in pairs (few observations, max five individuals seen during the fieldwork) and only above the altitudinal range of so-called mixed feeding drongo parties, despite that much effort was allocated into studying feeding parties (Dinesen et al. 1993, LAH and JOS pers. comm.). The altitudinal distribution of the species at the above-mentioned localities is described in Appendix 5.3.

 

The conclusion to draw from the knowledge of the species’ ecology in the Usambaras and the Udzungwas is that it occupies a rather wide altitudinal range in these areas, although occurring at very low densities, and that it is most frequently seen in mixed feeding parties (at least if the observer is in the altitudinal range of these) in the majority of (but apparently not all of) the areas. The species could potentially exist in the Uluguru South F.R. despite of the lack of submontane forest and mixed feeding drongo parties (Appendix 5.7) and probably also occur in montane forest in the Ulugurus. This needs further clarification, whereas it seems safe to assume that the submontane forest is important to the species. With only three observations from the Ulugurus it is too early to say anything about the population size but it is probably uncommon here as in the Udzungwas and the Usambaras - the only other mountain ranges holding the species.

 

Southern Banded Snake Eagle Circaetus fasciolatus. Kimboza Forest is not a key area for C. fasciolatus but is one of many threatened forest patches, which this species depends on throughout its range. The lowland Ruvu F.R. immediately east of Kimboza F.R. is still unknown ornithologically but could potentially hold some breeding pairs. The species could potentially be overlooked (it is hard to see and locate) in the submontane forest parts (e.g. the Kinole area), since in the Usambara Mountains it is common at 900 m in the East Usambaras with a single record from 1450 m in the West Usambaras and one from 1600 m in the Ngurus (Stuart and Turner 1980, Britton 1980, Fuggles-Couchmann 1984a). The Uluguru North F.R. is, however, some 25 km away from the foothill forests, which must be considered the key areas for C. fasciolatus in the Ulugurus. It is still too early to say anything about the population size in the Ulugurus.

 

Uluguru Violet-backed Sunbird Anthreptes neglectus. The species is common in Kimboza F.R. according to Stuart and Jensen (1985). However, a single bird was seen during our survey. Ruvu F.R. is unknown ornithologically but probably also holds a population. The species probably occurs also in other forested areas in the eastern foothills but Stuart and Jensen (1985) give no details about this. It remains to be established whether it uses the woodlands to some extent (known to occur in moist bushed and wooded country at least at Mikindani, Nguhi and Soga in coastal Tanzania [Britton 1980]).

 

A. neglectus was not recorded during our survey at Tegetero. On this background, it seems safe to assume that the species is uncommon above 1250 m on the eastern slopes. Only little time was spent below 1250 m, and the species may be more common between 1000 and 1250 m. The Uluguru South F.R. probably does not offer proper conditions for A. neglectus, since very little forest is left below 1500 m.

 

Loveridge's Sunbird Nectarinia loveridgei. It is difficult to estimate the total population of this common species since no proper density estimates have been obtained yet by territory mapping or total counts inside well-defined smaller areas. Since the species is very abundant throughout these forests, the population must number several million individuals. We have no data on densities to clarify this very rough estimate. The species depends completely on the survival of the forests of the Uluguru North and South F.R.s but is certainly in no danger of extinction. It tolerates some degree of disturbance inside the forest.

 

 

5.6.2. General characteristics of the forest avifaunas at the localities visited.

 

The avian community clearly changes with altitude in the Ulugurus, as is evident from the plot and mistnetting data and the species reviews. This is the case also in other tropical montane forests. Many of the differences between our stations can be explained by the elevational distribution of forest habitat in the areas and the habitat preferences of the forest species. The most important features are mentioned below (5.6.2a-e), further notes are given in Appendix 5.9.

 

 

5.6.2.a. The Uluguru North and South F.R.s in general.

 

Within the Ulugurus the bulk of the species of special conservation interest are completely restricted to the submontane, montane and upper montane forest belts of the Uluguru North and South Forest Reserves (including some small outlying mountain forest patches) with only two Near-threatened species and two other restricted-range species known from the lowland forests.

 

A feature shared between most of our stations in the Uluguru North and South F.R.s is the extremely high abundance of Nectarinia loveridgei.

 

Although some species are mistnetted only at certain altitudes, the set of core species in the mistnetting data does not change much between our 12 mistnetting stations (Table 5.7) and there is no clear difference in number of species mistnetted between our mistnet localities (Kimhandu, Lanzi and Tegetero) (Figure 5.2). The difference between Lanzi-1 and -2 is discussed in Appendix 5.9.c.

 

The plot curves in Figure 5.1 show that many of the species are common and that the tail of low-density species is rather short, compared to what is normally the case in tropical forest bird communities (Fjeldså and Rabøl in press). This is a noteworthy aspect of isolated Eastern Arc avifauna communities, which hold a moderate number of species, and could be seen as a consequence of a high degree of extinction in the past. Those species that are able to survive fill up the void and are generally found at good densities. Most tropical lowland forests, notably the lowlands of South America, hold a much higher number of species, typically with few very common species and a long “tail” of species occurring at only low densities (J. Fjeldså pers. comm.).

 

While The Eastern Arc Mountains have a relatively poor avifauna compared to many other montane regions in the tropics, they are rich in endemic species. The Uluguru North and South F.R.s are among the key areas in this respect, holding two endemics of their own (Malaconotus alius and Nectarinia loveridgei) and another three Eastern Arc endemics (Ploceus nicolli, Scepomycter winifredae and Anthreptes rubritorques) as well as other Red Data Book species and species of restricted range (see Table 5.1).

 

The border between the avifaunas of the submontane and montane forest and the surrounding agricultural land is generally very abrupt: only a few non-forest species were noted inside the forest in our study areas (see Appendix 5.6), and they were generally not important components of the avian community (e.g. Table 5.5), even at the stations which lay close to the forest edge (Kimhandu-1 and -6 [1520 and 1540 m], Lanzi-1 [1710 m]). However, in the lower part of the forest at Tchenzema (degraded and dry), the Southern Puffback Dryoscopus cubla (actually much of a forest edge species) and Tropical Boubou Laniarius aethiopicus are common (Appendix 5.9.b). In the lowland Kimboza Forest there is a higher proportion of species which are not restricted to forest habitat (Appendix 5.6 and 5.9.c).

 

 

5.6.2.b. Submontane evergreen forest[1].

 

Comparing the Uluguru North and South F.R.s, the most important avifaunistic difference from a conservation aspect is caused by the fact that the northern section holds considerable areas of submontane evergreen forest whereas this habitat type today occupies only very small areas in the southern section (Figure 1.2).

 

Importance for bird species of special conservation interest: Three of the five Threatened species occurring in the Ulugurus are known only from the northern section (Apalis chariessa, Anthreptes rubritorques and Ploceus nicolli). A fourth, the endemic Malaconotus alius, is known mainly from the northern section but has been recorded once or twice in the southern. The altitudinal distributions of these four species in the Ulugurus and elsewhere were reviewed in Section 5.5.1, Appendix 5.3 and Section 5.6.1. The conclusion to draw is that Apalis chariessa and Anthreptes rubritorques depend entirely on the submontane (and possibly lowest montane) zone. Ploceus nicolli has been recorded only three times in the Ulugurus so our knowledge of the species is scanty. However, since at least two of the records are from the submontane or lower montane belt, these zones must be important for the species, perhaps forming its source habitat. Malaconotus alius appears to use the montane zone to some extent but the fact that we did not record it in the Uluguru South F.R. during our survey whereas we recorded four territories in the submontane and lower montane zone in the Uluguru North (where also most earlier records are from) clearly indicates that it depends on the submontane and lowest montane zone. Its occurrence in the montane zone may depend on to which extent the core habitat (submontane and lowest montane forest) remains, as a source pool. Conclusion: the protection of the submontane belt is the most critical of all. This forest type is found almost exclusively in the Uluguru North F.R., especially in the Tegetero-Bagiro-Kinole-Lupanga area on the eastern slopes, with some also near Morningside on the northwestern slopes.

 

Apart from the above-mentioned species the submontane belt of the Uluguru North F.R. furthermore holds small populations of Scepomycter winifredae (Threatened) and Anthrepthes neglectus (Near-threatened; occurs up to 1500, possibly 1800 m) and high densities of the endemic Near-threatened Nectarinia loveridgei (though outnumbered by N. olivacea in the lower part of the belt). Also the full set of restricted-range (other than Threatened and Near-threatened) species occurring in the Ulugurus are found in the submontane belt, one of them (Sheppardia sharpei) being clearly more common here than in the montane belt.

 

Other notes: A characteristic element of the submontane forest is the mixed feeding “drongo parties”. Such parties were a common phenomenon below 1500 m at Tegetero (Appendix 5.7). The Threatened species Apalis chariessa (of which almost all observations in the Eastern Arc are from feeding parties, see Appendix 5.3) is almost completely restricted to such parties in the Ulugurus. The Threatened Ploceus nicolli is known to take part in such parties regularly in the Usambaras and the Udzungwas and probably does so in the Ulugurus too. The principal habitat of the core species of the parties (Phyllastrephus flavostriatus, Apalis melanocephala, Dicrurus ludwigii and Ploceus bicolor) is submontane forest[2] and three of them apparently have populations only in the submontane and lowest montane belt of the Uluguru North F.R.

 

Importance of the presence of a submontane zone below the montane: With the exception of Dicrurus ludwigii, which was seen once at Kimhandu, none of the core species of drongo parties (Dicrurus ludwigii, Phyllastrephus flavostriatus, Ploceus bicolor and Apalis melanocephala) were recorded at the localities studied in the Uluguru South F.R. A likely explanation for the small or non-existing populations of these species in the southern section is that their principal habitat is subtropical forest (apart from lowland forest), a habitat type that has disappeared in the southern section. Thus there is no source habitat left to recruit the montane belt, and the populations are most likely not viable on the long term. They are most likely sink populations, founded by individuals that have spread to the Uluguru South F.R. from source areas in the submontane belt of the Uluguru North or from forested patches in the foothills, e.g. Kimboza. An alternative explanation is that they are relictual populations from earlier times when forest extended further down the slopes, surviving in lobes of forest extending below 1600-1700 m. According to current metapopulation theory, such relictual populations in sink habitat are doomed to vanish. See also Section 5.6.2.b.

 

The bird fauna recorded at Kigurunyembe (altitudinal range visited: 650-850 m) showed strongest parallels to the bird fauna in the lowlands. Our review of the forest reserve descriptions in Lovett and Pócs (1993) revealed that there is an ornithologically unknown patch of submontane forest (800-1100 m) in Mkungwe F.R. This forest patch could prove to contain an interesting avifauna with affinity to the lowland forests as well as the submontane areas of the Uluguru North F.R. Mkungwe is known to be of high conservation value for plants (Lovett and Pócs 1993).

 

Further notes on the submontane belt are given in Appendix 5.9.a.

 

 

5.6.2.c. Montane evergreen forest[3].

 

Whereas the submontane forest zone has largely disappeared in the southern section, the montane evergreen forest remains virtually intact in the southern (as well as in the northern) section. The largest areas of montane forest are clearly found in the Uluguru South F.R. but also the Uluguru North F.R. holds a considerable belt of this habitat type.

 

Importance for bird species of special conservation interest: The montane belt is the most important for the Threatened species Scepomycter winifredae (Section 5.5.1, Section 5.6.1). The Uluguru South F.R. holds a very important population of this species, which we recorded in good densities, at least locally, at Kimhandu, Lanzi and Tchenzema. The montane belt of the Uluguru North F.R. probably also supports an important population. Also for the Near-threatened Nectarinia loveridgei the montane belt is the most important (Figure 5.1, Appendix 5). The montane zone is furthermore very important for those restricted-range species, which are not categorized as Threatened or Near-threatened (Section 5.5.3, Figure 5.1).

 

Malaconotus alius has been recorded once or twice in the montane zone of the Uluguru South F.R. There is one record from the western slopes of the Lukwangule Plateau at 2100 m (Stuart and Jensen 1981) and one from "above Bunduki" (J.G. Williams in 1948), which must be from above 1700 m. There are still no records of this Uluguru endemic from the eastern slopes of the Uluguru South F.R. and, if present here, we believe that it must occur at very low densities, since we neither saw it nor heard its far-carrying call at Kimhandu or Lanzi. Further survey work is necessary to determine its status, especially in those few places which have forest lobes below 1500 m (Figure 1.2). The highest densities of M. alius appears to be in the northern section and the southern section of the Ulugurus with mostly montane forest may be a sink habitat for the species.

 

Further notes on the montane belt are given in Appendix 5.9.b.

5.6.2.d. Upper montane or lower subalpine zone[4].

 

The upper montane or lower subalpine zone at Kimhandu-5 (2520 m) and on the Lukwangule Plateau is characterized by harsh climate with strong day/night temperature differences. The forest is low, dense elfin forest of which bamboos are often a prominent part.

 

Importance for bird species of special conservation interest: One Threatened species, Scepomycter winifredae, was encountered on a single occasion in the upper montane zone at Tchenzema during this survey and appears to use this vegetation belt to some extent. One Near-threatened (Nectarinia loveridgei) and at least four restricted-range species occur in the upper montane zone, some of them at good densities (Figure 5.1.e).

 

Other species: The total number of forest species (21) recorded at Kimhandu-5 (all methods combined, Table 5.5) is considerably lower than the numbers recorded at lower stations at Kimhandu, Lanzi and Tegetero (34-50, see Table 5.5). The low number of species is expressed in the plot data which also show that there is no long tail of low-density species (Figure 5.1.e) and in the tape recording data (Table 5.10). For the interpretation of the plot data and the total number of species recorded it should be born in mind that the vegetation is very impenetrable at Kimhandu-5, implying that most plots were assessed from the forest edge and that general field observations were mostly carried out from the forest edge or from mistnet lanes. Potentially we could therefore have overlooked silent birds occurring at low densities inside dense forest. However, with c. four days of fieldwork in the area we believe that only few species can have been overlooked unless a high number of species occur at extremely low densities up here.

 

The number of species mistnetted at Kimhandu-5 was "normal" (Figure 5.2), contrasting the low number of species recorded by observational methods and tape recordings. The catch rate (54.8 individuals per 2500 MNH) was also "normal" (Table 5.8). The only species mistnetted up here but not at lower stations were the typical high altitude species Phylloscopus umbrovirens and Bradypterus cinnamomeus. Our results indicate that it is mostly the upper and mid strata, species that are absent or occur at so low densities at Kimhandu-5 that we did not record them. The community of understorey birds is relatively similar to understorey communities at lower stations.

 

Numerically prominent members of the forest bird assemblage at Kimhandu-5 are: Andropadus nigriceps (the only greenbul recorded above 2200 m at Kimhandu), Cossypha anomala, Modulatrix stictigula, Pogonocichla stellata, Bradypterus mariae, Apalis thoracica, Orthotomus metopias, Phylloscopus umbrovirens, Laniarius fuelleborni, Nectarinia loveridgei and (primarily near the edges) Bradypterus cinnamomeus. The high record of Bradypterus cinnamomeus in the plot data from Kimhandu-5 (Table 5.9, Figure 5.1.e) should be explained by the fact that the plot assessments were carried out from the forest edge - this species was most common on the meadow and in and near the forest edge, occurring at much lower densities inside the forest.

 

 

5.6.2.e. Lowland semi-evergreen forest[5].

 

Importance for bird species of special conservation interest: In the Ulugurus the Kimboza F.R. (probably also Ruvu F.R.) is probably the stronghold for the two Near-threatened species Circaetus fasciolatus and Anthreptes neglectus. One species of restricted range, Alethe fuelleborni, is common in Kimboza at least at certain times of the year. The rest of the assemblage of Threatened, Near-threatened and restricted-range (but not Threatened or Near-threatened) species occurring in the Ulugurus has never been recorded in Kimboza F.R. (except Poeoptera kenricki of which there is only a single observation).

 

Other notes: The following forest species occur in Kimboza F.R. but have never been recorded in the Uluguru North or South F.R.s: Circaetus fasciolatus (Near-threatened), Guttera pucherani, Glaucidium capense, Stactolaema leucotis (abundant), Pogoniulus simplex, Chlorocichla flaviventris, Phyllastrephus terrestris, Phyllastrephus debilis, Erythrocercus holochlorus, Telophorus quadricolor and Lamprotornis corruscus. Species clearly being most common in the foothill forests but with one or a few records from submontane and montane forest areas include Ceuthmochares aereus, Apaloderma narina, Ceratogymna bucinator, Tockus alboterminatus, Phyllastrephus fischeri, Oriolus chlorocephalus, Cossypha natalensis, Neocossyphus rufus, Illadopsis rufipennis, Anthreptes neglectus (Near-threatened) and Macrosphenus kretschmeri. Many of the above-mentioned species are prominent members of the bird communities of the threatened coastal forests further east and of other lowland Eastern Arc forests. Table 5.11 clearly illustrates that the composition of the bird fauna is very different from that of the mountain forests. Further notes on the lowland semi-evergreen forest are given in Appendix 5.9.c.

 

 

5.6.3. The absence of certain species in the Ulugurus (see Appendix 5.10 for details).

 

We had expected to be able to reveal the presence of two additional Threatened understorey species, which occur, disjunct in the Udzungwas and the Usambaras. However, they remain unrecorded and we therefore consider that their distributions are probably truly disjunct. Their absence appears odd since suitable habitats seem to be present in the lower parts of the Uluguru North F.R. and since the Ulugurus climatically seen are believed to have been one of the most stable areas in the Eastern Arc with their high amount of rainfall. The two species in question are:

 

·         Dappled Mountain Robin Arcanator orostruthus.

·         Swynnerton's Robin Swynnertonia swynnertoni.

 

Another species, which occurs disjunct in the Usambaras and the Udzungwas, is:

 

·         Amani Sunbird Anthreptes pallidigaster.

 

 

5.7. Recommendations for future ornithological studies (priorities).

 

Submontane, montane and upper montane areas. It seems safe to assume that the submontane parts of the Uluguru North F.R. are the most important areas to focus conservation efforts on in terms of bird conservation. Our knowledge on the geographical distribution and habitat preferences of some of the Threatened species in the Ulugurus is, however, still fragmentary. Further surveys, focusing entirely on Threatened species, should be undertaken, preferably linked with detailed mapping of the altitudinal distribution of forest. The observers should work without using the time consuming standardised methods: For the simple purpose of finding low density Threatened species we believe that it is more important to search large and diverse areas to obtain a detailed knowledge on core habitats and plasticity of the species. The surveys should include wide altitudinal ranges in the northern as well as the southern section and special attention should be paid to Malaconotus alius. The survey of Threatened species should preferably take place in the cold season for two reasons:

 

·                               Most trees flower in the cold season, and sunbirds, e.g. the Threatened species Anthreptes rubritorques and A. pallidigaster, often gather at flowering trees, making them easier to locate.

·                               Feeding parties are more frequent and mobile in the cold season, making it easier to locate Threatened species like Apalis chariessa and perhaps Ploceus nicolli.

 

A search of large areas of suitable habitat in the Uluguru North F.R. would furthermore enable a definitive determination of whether Arcanator orostruthus, Swynnertonia swynnertoni and Anthreptes pallidigaster (Section 5.6.3) are in reality absent from the Uluguru Mountains.

 

The distributional pattern in the Ulugurus may differ between Threatened bird species and red-listed species in other organism groups. An assessment of habitat preferences of rare taxa in other organism groups should complement the Threatened birds' survey.

 

Lowland forests. The large Ruvu F.R. is largely unknown ornithologically. It is our impression that also the biologically rich Kimboza F.R. is rather incompletely known since it has never been studied for longer periods. No quantitative data exist from Kimboza, apart from sporadic mistnetting data. Future fieldwork in Kimboza and Ruvu F.R.s should be carried out with the methods used in the Uluguru North and South F.R.s during our survey to improve the basis for comparative studies. Enough time should be allocated to determine the abundance of the Near-threatened species Circaetus fasciolatus and Anthreptes neglectus and to record the complete assemblage of species present.

 

Outlying hills. The outlying hills appear to be completely unknown ornithologically. Of these, especially the Mkungwe F.R. with its submontane forest (800-1100 m) could hold an avifauna of high conservation value.

 

Further recommendations are given in Appendix 5.13.

 

5.8. References.

 

Anderson, G.Q.A. and T.D. Evans in prep. A new subspecies of Swynnerton's Robin Swynnertonia swynnertoni from Tanzania. Bulletin of the British Ornithologists' Club.

Baker, N.E. 1984. Recent coastal records of the White-starred Forest Robin Pogonocichla stellata in Tanzania. Scopus 8: 51-52.

Baker, N.E. and L. Baker 1990. Instructions for ringers. Unpublished manual prepared for the WCST/ICBP Coastal Forest Survey. June 1990.

Benson, C.W. 1950. A collection from Chiperoni Mountain, Portuguese East Africa. Bulletin of the British Ornithologists’ Club 70: 51.

Britton, P.L. 1978. The Andersen collection from Tanzania. Scopus 2: 77-85.

Britton, P.L. (ed.) 1980. Birds of East Africa - their habitat, status and distribution. Nairobi: East Africa Natural History Society.

Britton, P.L. 1981. Notes on the Andersen collection and other specimens from Tanzania housed in some West German Museums. Scopus 5: 14-21.

Britton, P.L., H.A. Britton and M.A.C. Coverdale 1980. The avifauna of the Mrima Hill, South Kenya coast. Scopus 4: 73-78.

Brown, L.H., E.K. Urban and K. Newman 1982. The Birds of Africa. Volume I. London: Academic Press.

Cambridge Tanzania Rainforest Project 1994. A biological and human impact survey of the lowland forests, East Usambara Mountains, Tanzania. BirdLife Study Report No. 59. Cambridge, UK: BirdLife International.

Collar, N.J., M.J. Crosby and A.J. Stattersfield 1994. Birds to watch 2: The world list of threatened birds. BirdLife Conservation Series No. 4. Cambridge, U.K.: BirdLife International.

Collar, N.J. and S.N. Stuart 1985. Threatened birds of Africa and related islands: the ICBP/IUCN Red Data Book. Part 1, 3rd edition. Cambridge, U.K: International Council for Bird Preservation and International Union for Conservation of Nature and Natural Resources.

Collar, N.J. and S.N. Stuart 1988. Key forests for threatened birds in Africa. ICBP Monograph No. 3. Cambridge, U.K.: International Council for Bird Preservation.

Dinesen, L., T. Lehmberg, J.O. Svendsen and L.A. Hansen 1993. Range extensions and other notes on some restricted-range forest birds from West Kilombero in the Udzungwa Mountains, Tanzania. Scopus 17: 48-59.

Dinesen, L. T. Lehmberg, J.O. Svendsen, L.A. Hansen and J. Fjeldså 1994. A new genus and species of perdicine bird (Phasianidae, Perdicini) from Tanzania, a relict form with Indo-Malayan affinities. Ibis 136: 2-11.

Dowsett, R.J. and F. Dowsett-Lemaire (eds.) 1993. A contribution to the distribution and taxonomy of Afrotropical and Malagasy birds. Tauraco Research Report No. 5. Liège, Belgium: Tauraco Press.

Dowsett-Lemaire, F. 1989. Ecological and biogeographical aspects of forest bird communities in Malawi. Scopus 13: 1-80.

Evans, T. and G.Q.A. Anderson 1992. A wildlife survey of the East Usambara and Ukaguru Mountains, Tanzania. ICBP Study Report No. 53. Cambridge UK: International Council for Bird Preservation.

Evans, T. and G.Q.A. Anderson 1993a. Notes on Moreau's Sunbird Nectarinia moreaui. Scopus 17: 63-64.

Evans, T. and G.Q.A. Anderson 1993b. Results of an ornithological survey in the Ukaguru and East Usambara Mountains, Tanzania. Scopus 17: 40-47.

Fjeldså, J. 1993. The avifauna of the Polylepis woodlands of the Andean highlands: the efficiency of basing conservation priorities on patterns of endemism. Bird Conservation International 3: 37-55.

Fjeldså, J. and J. Rabøl in press.Variation in avian communities between isolated units of the Eastern Arc montane forests, Tanzania. Le Gerfaut.

Franzmann, N.-E. 1983. A new subspecies of the Usambara Weaver Ploceus nicolli. Bulletin of the British Ornithology Club 103: 49-51.

Friedmann, H. 1927. New birds from Tanganyika Territory. Proceedings of the New England Zoological Club 10: 3-7.

Friedmann, H. 1928. A collection of birds from the Uluguru and Usambara Mountains, Tanganyika Territory. Ibis 4: 74-99.

Friedmann, H. 1929. Two East African Barbets. Proceedings of the New England Zoological Club 11: 35-36.

Friedmann, H. and A. Loveridge 1937. Notes on the ornithology of tropical East Africa. Bulletin of the Museum of Comparative Zoology at Harvard College 81: 1-413.

Friedmann, H. and K.E. Stager 1964. Results of the 1964 Cheney Tanganyikan Expedition. Ornithology. Contributions in Science (Los Angeles) 84: 3-50.

Fry, C.H., S. Keith and E.K. Urban (eds.) 1988. The Birds of Africa. Vol. III. London: Academic Press.

Fuggles-Couchman, N.R. 1939. Notes on some birds of the Eastern Province of Tanganyika Territory. Ibis 3: 76-106.

Fuggles-Couchman, N.R. 1984a. The distribution of, and other notes on, some birds of Tanzania. Scopus 8: 1-17.

Fuggles-Couchman, N.R. 1984b. The distribution of, and other notes on, some birds of Tanzania - Part II. Scopus 8: 73-78 and 81-92.

Fuggles-Couchman, N.R. 1984c. The distribution of, and other notes on, some birds of Tanzania - Part II. Scopus 8: 81-92.

Hall, B.P. and R.E. Moreau 1970. An atlas of speciation in African passerine birds. London: Trustees of the British Museum (Natural History).

ICBP 1992. Putting biodiversity on the map: Priority areas for global conservation. Cambridge, UK: International Council for Bird Preservation.

Jensen, F.P. and S. Brøgger-Jensen 1992. The forest avifauna of the Udzungwa Mountains, Tanzania. Scopus 15: 65-83.

Keith, S., E.K. Urban and C.H. Fry (eds.) 1992. The birds of Africa. Volume IV. London: Academic Press.

Koen, J.H. 1988. A census technique for Afromontane forest bird communities. Suid-Afr. Bosboutydskrif 145: 39-41.

Loveridge, A. 1922. Notes on east African birds (chiefly nesting habits and stomach contents) collected 1915-19. Proceedings of the General Meetings for Scientific Business of the Zoology Society of London for 1922: 837-862.

Loveridge, A. 1933. Reports on the scientific results of an expedition to the southwestern highlands of Tanganyika Territory. I. Introduction and zoögeography. Bulletin of the Museum of Comparative Zoölogy 75: 1-43.

Loveridge, A. 1960. Status of new vertebrates described or collected by Loveridge. Journal of the East Africa Natural History Society 23: 250-280.

Lovett, J.C. and D.C. Moyer in press. Notes on Kawemba Forest Reserve, Iringa District, Tanzania. Bulletin of the East Africa Natural History Society.

Moreau, R.E. 1938. [A new Artisornis]. Bulletin of the British Ornithology Club 58: 139.

Moreau, R.E. 1940. Distributional notes on East African birds. Ibis 82: 454-463.

Moreau, R.E. 1946. The adult of Mrs Moreau’s Warbler. Bulletin of the British Ornithologist’s Club: 66: 44.

Moreau, R.E. 1957. Variation in the western Zosteropidae (Aves). Bulletin of the British Museum of Natural History, Zoology 4: 309-433.

Moreau, R.E. 1966. The bird faunas of Africa and related islands. London: Academic Press.

Moyer, D.C. 1993. A preliminary trial of territory mapping for estimating bird densities in Afromontane forest. Proceedings of the Eighth Pan African Ornithological Congress. Bujumbura, Burundi.

Reichenow, A. 1889. Ueber eine Vogelsammlung aus Ostafrika. Journal für Ornithologie 37: 264-286.

Reichenow, A. 1894. Die Vögel Deutsch-Ost-Afrikas. Berlin: D. Reimer.

Reichenow, A. 1895. Dr. Stuhlmann's neueste Forschung in Ost-Afrika. Ornithologisches Monatsberichte 3: 87.

Reichenow, A. 1900. Neue Forschungen in Deutsch-Ostafrika. Ornithologisches Monatsberichte 8: 38-40.

Reichenow, A. 1900. Neues aus Deutsch-Ostafrika. Ornithologisches Monatsberichte 8: 98-100.

Reichenow, A. 1900-1901. Die Vögel Afrikas. Vol. I. Neudamm: J. Neudamm.

Reichenow, A. 1902-1903. Die Vögel Afrikas. Vol. II. Neudamm: J. Neudamm.

Reichenow, A. 1904-1905. Die Vögel Afrikas. Vol. III. Neudamm: J. Neudamm.

Reynolds, R.T., J.M. Scott and R.A. Nussbaum 1980. A variable circular-plot method for estimating bird numbers. Condor 82: 309-313.

Ripley, S.D. and G. Heinrich 1966. Comments on the avifauna of Tanzania. I. Postilla No. 96, pp. 1-45.

Ripley, S.D. and G. Heinrich 1969. Comments on the avifauna of Tanzania. II. Postilla No. 134, pp. 1-21.

Scharff, N., M. Stolze and F.P. Jensen 1982. The Uluguru Mountains, Tanzania. Report of a study tour 1981. Unpublished report, Copenhagen.

Schuster, L. 1926. Beiträge zur Verbreitung und Biologie der Vögel Deutsch-Ostafrikas. Journal für Ornithologie 74: 138-167, 521-541, 709-742.

Sibley, C.G. and B.L. Monroe Jr. 1990. Distribution and taxonomy of the birds of the world. Newhaven: Yale University Press.

Stuart, S.N. 1981a. A comparison of the avifaunas of seven East African forest islands. African Journal of Ecology 19: 133-152.

Stuart, S.N. 1981b. An explanation for the disjunct distributions of Modulatrix orostruthus and Apalis (or Orthotomus) moreaui. Scopus 5: 1-4.

Stuart, S.N. 1983. Biogeographical and ecological aspects of forest bird communities in eastern Tanzania. Unpublished Ph.D. thesis, Cambridge University, Cambridge U.K.

Stuart, S.N. 1989. The forest bird fauna of the East Usambara Mountains. Pp. 357-361 in Hamilton, A.C. and R. Bensted-Smith (eds.). Forest conservation in the East Usambara Mountains, Tanzania. Gland, Switzerland and Cambridge, U.K.: International Union for the Conservation of Nature and Natural Resources.

Stuart, S.N. and F.P. Jensen 1981. Further range extensions and other notable records of forest birds from Tanzania. Scopus 5: 106-115.

Stuart, S.N. and F.P. Jensen 1985. The avifauna of the Uluguru Mountains, Tanzania. Le Gerfaut 75: 155-197.

Stuart, S.N., F.P. Jensen and S. Brøgger-Jensen 1987. Altitudinal zonation of the avifauna in Mwanihana and Magombera Forests, Eastern Tanzania. Le Gerfaut 77: 165-186.

Stuart, S.N., F.P. Jensen, S. Brøgger-Jensen and R.I. Miller (1993). The zoogeography of the montane forest avifauna of eastern Tanzania. Pp. 203-228 in Lovett, J.C. and S.K. Wasser (eds.) Bi­ogeography and ecology of the rain forests of eastern Africa. Cambridge, U.K.: Cam­bridge.

Stuart, S.N. and D.A. Turner 1980. Some range extensions and other notable records of forest birds from eastern and northea­stern Tanzania. Scopus 4: 36-41.

Stuart, S.N. and T.A. van der Willigen 1978. Report of the Cambridge Ecological Expedition to Tanzania 1978. Unpublished.

Stuart, S.N. and T.A. van der Willigen 1980. Is Moreau's Sunbird Nectarinia moreaui a hybrid species? Scopus 4: 56-58.

Svendsen, J.O. in prep. Additional montane forest bird voices from the Udzungwas and some from the Ulugurus, Tanzania. Unpublished tape with accompanying booklet. To be acquired from Zoological Museum of Copenhagen. Follows the tape by Svendsen and Hansen (1992).

Svendsen, J.O. and L.A. Hansen 1992. Some bird voices from the Udzungwa Mountains, Tanzania. Unpublished tape with accompanying booklet. To be acquired from Zoological Museum of Copenhagen.

Urban, E.K., C.H. Fry and S. Keith (eds.) 1986. The Birds of Africa. Volume II. London: Academic Press.

Williams, J.G. 1950. On the status of Cinnyris mediocris moreaui. Ibis 92: 645-47.

Williams, J.G. 1951. Notes on Scepomycter winifredae and Cinnyris loveridgei. Ibis 93: 469-470.

Go to: Top of page, Contents of Part A, or Next Section on Mammals


[1] Submontane evergreen forest: Between 500 and 1500 m (see Section 3.3).

[2] All four species also have populations in the foothill forests but in the Uluguru North and South F.R.s the core habitat is submontane forest.

[3] Montane evergreen forest: Between 1500 and 2100 m in the Uluguru North F.R. and 1600 to 2400 m in the Uluguru South F.R. (see Section 3.3).

[4] Upper montane or lower subalpine zone: above 2400 m in the southern section and above 2100 m in the northern section, see Section 3.3.

[5] Lowland semievergreen forest: forest between 250 and 500 m, see Section 3.3.