First assessment of bird diversity in the UNESCO Sheka Forest Biosphere Reserve, southwestern Ethiopia: species richness, distribution and potential for avian conservation

The Sheka Zone in southwestern Ethiopia is covered by some of the largest remaining forests in the Eastern Afromontane biodiversity hotspot. Owing to the rich biodiversity and a high degree of endemism, it was declared as a biosphere reserve by UNESCO in 2012 and is considered a Key Biodiversity Area. Detailed knowledge on species diversity and distribution in the reserve is, however, severely limited. From February to April 2016, an assessment of the bird diversity and distribution in the reserve was made for the first time through point count transects, camera-trap recordings and opportunistic observations. In total, 244 bird species were identified, of which 19% was only found within the reserve’s designated protected zones. Our study indicates a remarkable bird species richness across the different habitats in Sheka Forest Biosphere Reserve and can be used as a baseline for future monitoring studies and conservation planning.


INTRODUCTION
Ethiopia is recognized among the 35 most biodiverse regions in the world and its biodiversity is renowned for its high percentage of endemism. The highlands in the southwest, where most of the larger forest tracts remain, are part of the Eastern Afromontane Biodiversity Hotspot (Mittermeier et al. 2011). Natural habitats are highly diverse in this region, where the forest structure changes along a large elevation gradient (Friis 1992), with moist evergreen Afromontane rainforests presenting the dominant vegetation type, but also including bamboo forests and transitional rainforests leading down to semi-deciduous forests at lower altitudes. In most areas, the forests are interspersed with wetlands or, at higher altitude, moorlands. Ethiopian natural forests are however rapidly disappearing (Dessie & Kleman 2007;Reusing 2000), with approximately 11.4% or 12,499,000ha of total forest cover left (FAO 2015).
The Sheka Zone in southwestern Ethiopia still harbors some of the largest remaining Afromontane forests in the country (>100,000ha). Nevertheless, the combination of a rising population, ongoing landallocation to agricultural investors and a lack of landuse planning are increasing pressure on the remaining natural habitats. As such, the deforestation rate within Sheka is one of the highest in Ethiopia, with severe impacts on local economy, culture and environment (Woldemariam & Fetene 2007). Sheka forest is considered a Key Biodiversity Area (Birdlife International 2017) and, in 2012, it was recognized by UNESCO as the Sheka Forest Biosphere Reserve. This recognition has led to the zonation of the area into core-, buffer-and transition zones (Fig. 1). These are, respectively, devoted to long-term protection of intact forests (core zones), participatory forest management and low-intensity production (buffer zones), and sustainable human settlement and agriculture (transition zones) (Gole & Getaneh 2011). The forest furthermore provides an important refuge for the native genetic diversity of wild crop relatives with significant agricultural value. The most prominent example is the indigenous wild coffee Coffea arabica, which can still be found in reasonable densities in the broadleaf forest of Sheka, but also other wild crop relatives that are imperative for local food provisioning such as Enset or 'False Banana' Ensete ventricosum and Ethiopian Cardamom Aframomum corrorima. All combined, there are strong incentives to safeguard this forest also from an economic and agricultural perspective (De Beenhouwer et al. 2013;Aerts et al. 2015).
Biodiversity and conservation research has seen a recent increase in southwestern Ethiopia, mainly in relation to agroforestry (e.g., Hundera et al. 2013;Tadesse et al. 2014); however, biodiversity studies in remote forest regions such as Sheka have been very limited thus far, despite the recognition as a UNESCO biosphere reserve and strong ongoing habitat degradation. Detailed insights in the diversity, distribution and abundance of species in the area are hardly available. Likewise, information on the bird diversity in Sheka forest is very scarce, but limited research done in the broader region (e.g., Woldegeorgis & Wube 2012) provides clear indications for a large bird species diversity in remaining forest tracts of southwestern Ethiopia. Here, we present a first comprehensive inventory of the bird diversity in Sheka Forest Biosphere Reserve, discuss our observations in relation to the delineated management zones, and highlight the biosphere reserve as an understudied yet highly valuable area for bird conservation and continued biodiversity studies.

MATERIALS AND METHODS
From Masha (around 7.749 0 N & 35.471 0 E; ca. 2,250m), the largest village in the area and capital of the Sheka Zone, we surveyed a variety of core-, buffer-and transition zones across the three districts, or 'woredas', situated in the biosphere reserve (Masha, Anderacha and Yeki). Our field expeditions took place from 2 February 2016 until 30 April 2016, and nine different 'kebeles' (the smallest administrative division) were visited across the three woredas (Table 1). The humid highlands of southwestern Ethiopia are characterized by a short rainy season from March to April and a long rainy season from June to October, with an average temperature of 18.4°C and yearly precipitation of 1,783mm. Hence, our expeditions were carried out immediately prior to and during the short rainy season. Permission for the field work was granted by the Ethiopian Wildlife Conservation Authority (EWCA) at the national level, the head office of the Sheka Zone and by kebele leaders on the local administrative level. The study was part of a larger expedition to assess the biodiversity in the reserve.
Birds were identified using visual, vocal and camera trap observations. Photo and audio recordings were made to support identification. When recordings of certain species were lacking, only those with double observations were listed. Visual and vocal assessments were done during early morning point transect surveys and on an ad hoc basis in all locations visited throughout the expedition. Because the delineation between core and buffer zones was not always clear in the field, observations herein were collectively assigned to the 'protected zones' within the reserve. Assessments were carried out covering a wide variety of habitats. These were divided into seven categories for a provisional overview of species' habitat occupation throughout the reserve (Table 2). Early morning point transect surveys consisted of six counts of 10 minutes, interspersed by five minutes of walking in a predetermined direction. All birds seen and/or heard within a perimeter of 25 meters around the observer were noted. In this way, a total of 74 point transect surveys were carried out, with one survey always restricted to the same habitat. Species identification was done using the Helm field guide 'Birds   Fig. 1) for a total of approximately 510 camera trap days. Camera trap locations were selected based on their potential to monitor mammal diversity, but additional bird observations were made of crepuscular species and analysed as part of the present study.

RESULTS
We recorded a total of 244 bird species from 55 different families in the Sheka Forest Biosphere Reserve (Table 3; Images 1-4). Of the species recorded in the reserve during our fieldwork, 47 species or 19% were only found in the designated protected zones. All other species were found across a variety of management zones. Of the 155 species identified during the point transect surveys, 16 species were observed across all studied habitats, of which Bradypterus cinnamomeus (89.2% of surveys), Turdus (olivaceus) abyssinicus (83.8%), and Zosterops poliogastrus kaffensis (78.4%) were the most common bird species identified during the surveys. Riverine forest was the most species-rich habitat sampled with on average 24.3 species sampled per survey, while bamboo forest was the least species rich habitat with an average of 13.4 species recorded per survey ( Table 2). Based on our observations, the main habitat types used by each species are noted, except for the species encountered on migration, for which no habitats could be determined (Table 3).
Eleven bird species that were recorded during the surveys are considered threatened based on the IUCN Red List. With the notable exception of the recorded vultures, these observations largely stem from within the designated protected zones of the reserve. Details on these species and their identification are discussed below:

Necrosyrtes monachus (Critically Endangered) (Image 1C)
A smaller brown vulture, identified based on its naked pink head and whitish-grey 'hooded' hindneck and nape. It is common and still very abundant around settlements in the region, but rather rare in a variety of other habitats. Like the other African vultures listed below, it is severely threatened by a combination of factors, including land conversion, active persecution as well as secondary poisoning.

Trigonoceps occipitalis (Critically Endangered)
A rather large blackish vulture with contrasting white belly and large red-colored bill. Singles and pairs are uncommon but found to be widespread in a variety of habitats including settlements.

Gyps africanus (Critically Endangered) (Image 1D)
A typical vulture with a bright brown back, dark brown plumage, dark bill and down feathers on the neck. Seen in groups of up to 20 birds but also often in the presence of other vulture species. Common and widespread in a variety of habitats, most numerous around settlements.

Gyps rueppellii (Critically Endangered)
This vulture is similar to Gyps africanus but easily distinguished based on its yellow bill and scaled appearance in adult birds caused by bright edges on dark feathers. Rather uncommon, but widespread and present in a variety of habitats including settlements.

Torgos tracheliotus (Endangered)
A very large, dark brown vulture with a pink head that shows distinctive skin folds. Pairs and solitary birds were encountered only a limited number of times around smaller settlements and agricultural areas.

Aquila nipalensis (Endangered)
A large eagle with brown upper parts and blackish flight feathers, larger and darker than Aquila rapax. Observed and photographed on only one occasion while on migration northwards on 18 April 2016 above a highland moorland at Gandochi.

Polemaetus bellicosus (Vulnerable)
A very large brown eagle with white body underparts. Only one sighting in a wetland in the protected zone of Shato where a territorial pair was seen and heard in flight on 24 April 2016.

Balearica pavonina (Vulnerable)
An unmistakable crane with predominantly black body plumage and a crown of golden feathers. Limited number of sightings on a single day during a wetland inventory around the town of Masha.

Terathopius ecaudatus (Near Threatened)
Easily recognised in flight from below, based on the black body plumage, black and white wings and chestnut tail. Observed only one time at Shato forest (south) and two times in highland moorland at Gandochi, where a male was photographed on 30 March 2016.

Stephanoaetus coronatus (Near Threatened) (Image 1F)
Large eagle with crest, giving the head a rather triangular appearance. Barred black and white from below, with chestnut underwing coverts. Widespread and common in the core forested areas, very rare elsewhere. Both juveniles and adults were seen and photographed on several occasions throughout the study.

Rougetius rougetii (Near Threatened) (Image 1H)
A brown rail with white undertail coverts. Very common in highland wetlands, also common in Moorlands and along the Gebba river.
Several species that were found in the reserve during the monitoring occur only in a very restricted range in Ethiopia. Although these species are not considered threatened on the IUCN red list nor are endemic, their isolated distribution within Ethiopia and/or occurrence at the extremes of their distribution ranges deserves special conservation attention. Except for Cinnyris chloropygius, these species were exclusively found within the protected zones of the reserve.

Sarothrura elegans
Most common in forest interior, but also more open woodland types. Never seen, but regularly heard at night or early morning during the short rainy season. Sometimes, several males could be heard and were recorded giving a long and low hooting "whoooooooo" lasting approximately three seconds and repeated at intervals of approximately five seconds.

Sarothrura rufa
Adult males have a chestnut red head and chest with otherwise black plumage with narrow but striking white streaks. Adult males were observed two times during an inventory in a wetland around Shato when flushed from about 5m from the observers in a wet grassland.

Bradypterus alfredi
Sound recorded at two different locations in highland bamboo forest were uncommon and localized. A rapid series of short notes all at the same pitch "chit-it chit-it chit-it". (Image 2E) A striking blue kingfisher, distinguished from Halcyon senegalensis based on its blue breast-band and more extensive amount of black on its wings. Rare and secretive, but present around the largest rivers in the area. Often heard in wetland and riverine forest habitat in Shato forest, and one individual was photographed in lowland riverine forest near Tepi.

Anomalospiza imberbis
A small finch-like species. Males were recognised based on their black bill and yellow head and underparts. Only observed once in the largest wetland at the side of the Gebba River. A male was seen on 28 February 2018 for approximately 30 seconds sitting on top of a shrub at 15m from the observers during a wetland survey.

Cinnyris chloropygius
Males showed a metallic green head, breast and upper parts, a red breast-band and olive belly. Uncommon in open riverine woodland in the lowlands of Tepi, where one territorial male was seen actively foraging during a riverine forest survey by several observers.

Cisticola (galactotes) lugubris
A cisticola with black and grey streaked mantle and rufous crown. Extremely common and abundant around wetland, highland moorland and big rivers.

Parophasma galinieri
A distinctive grey bird with whitish forehead and orange-red undertail coverts. Common in highland bamboo forest, where recorded in high densities in forest edges around moorlands, very rare elsewhere.

Oriolus monacha (Image 4B)
A large black-headed oriole with grey wing panel. Very common in a wide variety of highland habitats, not restricted to forests.

Corvus crassirostris
A large raven with white patch on nape and very large bill. Common and abundant around settlements and farmland, rare around wetlands and open forest and not recorded from the core zones.

DISCUSSION
This study highlights the importance of the Sheka Forest Biosphere Reserve for globally threatened and Ethiopian endemic bird species, and in particular raptors and vultures. The results not only emphasize a high species richness in the reserve, but also indicate the importance of the proposed zonation, with increased control and protection, for bird conservation. Our findings are comparable to other recent ornithological studies in Ethiopia (e.g., Engelen et al. 2017;Rodrigues et al. 2018), showing a structurally diverse landscape, mostly in traditional low-intensity use, supporting a diverse range of bird species, and with undisturbed forest habitats and wetlands presenting a central refuge for vulnerable range-restricted and specialist bird species.
Our inventories were nevertheless confined to the short Ethiopian rainy season, as well as a subset of kebeles and transition-, buffer-, and core zones. Additional bird species, including seasonally present migrants, can undoubtedly be recorded during future studies in different periods and subregions. For instance, two endemic red-listed species, expected to occur in the reserve based on distribution maps, were not found (Macronyx flavicollis and Cyanochen cyanoptera).
More extensive surveys, specifically during the long rainy season and in the vast moorlands in the eastern highlands of Anderacha woreda might still indicate their continued presence. In addition, point transect surveys were mainly conducted to map species diversity and distributions across habitats, but were inadequate to accurately characterize the relative abundance of species in the reserve. Thus, our study provides a first indication of bird species richness in Sheka forest, and is a baseline that needs complementary monitoring studies to provide more detailed insights in its species composition, population sizes and dynamics. Our study also showed the added value of camera trapping to record crepuscular species, with Pternistis squamatus and Zoothera piaggiae only being visually observed through camera trap observations.
The large elevation and climatic differences that are present in the reserve add up to a large habitat variation and exceptionally rich bird diversity, underlining the protected status assigned to Sheka forest and the urge to safeguard its habitats from ongoing degradation. Our observations furthermore emphasize the understudied nature of this remote biosphere reserve and the importance of continued biodiversity studies to inform conservation planning. The forests of Sheka provide innumerable services to local communities and many people are directly relying on forest resources for their subsistence. As a result, the forests have been sustained through a long tradition of natural resource management (Woldemariam & Fetene 2007), making Sheka Forest Biosphere Reserve an ideal subject for directed longterm and community-based initiatives to conserve some of the largest remaining Afromontane rainforests.