White-winged Flufftail Sarothrura ayresi


Justification of Red List Category
This elusive species is listed as Critically Endangered as its population is now thought to be very small, and is believed to be undergoing a very rapid continuing decline in extent, area, and habitat quality, owing to the high rate of loss and degradation of its preferred habitat, seasonal marshland. Populations in both Ethiopia and South Africa are now thought to be extremely small (although there may be some migration between the two areas).

Population justification
The population in South Africa was previously estimated to be 235 birds, with at least a further 210-215 pairs in Ethiopia (A. Shimelis in litt. 1998), i.e. probably 700 mature individuals in total. The species can only be reliably surveyed in the breeding season in Ethiopia, but surveys risk disturbing nesting and thus have been avoided in the past (Y. D. Adebe in litt. 2013), and there is no clear evidence to suggest it doesn't breed in southern Africa (C. Symes in litt. 2016). Nevertheless, the lack of recent records and tiny area of remaining habitat suggest that the total population is now almost certainly fewer than 250 mature individuals, with both the Ethiopian and South African populations numbering fewer than 50 mature individuals each (Y. Adebe, M. Ewnetu, G. Gebreselassie and H. Smit-Robinson in litt. 2013). Surveys may be hampered by the fact that its call is essentially unknown, and thus individuals or areas containing this species may have been completely missed (C. Symes in litt. 2016).

Trend justification
This species's population is suspected to be decreasing very rapidly in line with levels of disturbance, habitat loss and degradation in Ethiopia and South Africa (Atkinson et al. 1996a, Taylor and van Perlo 1998, P.B. Taylor in litt. 1999, De Smidt and Evans 2003, Taylor and Grundling 2003, M. Drummond in litt. 2005, Y. Adebe, M. Ewnetu, G. Gebreselassie and H. Smit-Robinson in litt. 2013).

Distribution and population

The species occurs in Ethiopia (three known sites in the central highlands, with only one known for breeding - the only known breeding area for this species [H. Smit-Robinson in litt. 2016]) (Taylor and van Perlo 1998, Taylor 1998, 1999), Zimbabwe (one record in 1988 [Hustler and Irwin 1995], two records in the 1970s [Taylor and van Perlo 1998], and a possible breeding record in the 1950s [Taylor and van Perlo 1998, Taylor 1999]), and South Africa (ten sites in the Eastern Cape, KwaZulu-Natal and Mpumalanga [De Smidt 2003], and it can occur at low, coastal altitudes [Davies et al. 2015]). Claimed records from Zambia and Rwanda are unproven (Taylor and van Perlo 1998, F. Dowsett-Lemaire and R. J. Dowsett in litt. 1999, P. Leonard in litt. 1999). In South Africa, the total population was estimated to be 235 birds by Taylor and van Perlo (1998), but the area of occupancy has since been estimated at 3.92 km2 in South Africa (72 ha suitable habitat at Middelpunt and 320 ha at Wakkerstroom, the only two recently reliable sites; H. Smit-Robinson in litt. 2013), and due to low confidence in past estimates and continued threats to the species and its habitat over the past 10 years, the regional population in South Africa is thought to be fewer than 50 birds (H. Smit-Robinson in litt. 2013, Evans et al. 2015).

In the Ethiopian highlands, 10-15 pairs bred at Sululta in the late 1990s (Atkinson et al. 1996a, Anon. 1997c, J. S. Ash in litt. 1999) and c.200 pairs were discovered at a new breeding site (Berga floodplain) in 1997 (Anon. 1997c, A. Shimelis in litt. 1998, M. Wondafrash in litt. 2007). In 2005, a small breeding population was discovered at Bilacha in Ethiopia (M. Wondafrash in litt. 2007), with three adults recorded initially in July, followed by the location of three eggs in August and 19 nests in September (Anon. 2006). However, surveys at Bilacha and Weserbi since 2007 have found no evidence of breeding at either site. Bilacha in particular is highly overgrazed, with one bird seen in 2010 the only recent record, and Weserbi is thought to support a maximum of 1-2 pairs (Y. D. Abebe in litt. 2013). At Berga, the average number of birds flushed annually since 2007 was just four, and two days of intensive searching in August 2013 located only 12 individuals in suitable habitat now limited to c.300 ha (Y. D. Abebe and G. Gebreselassie in litt. 2013); 11 more were flushed (and 7 ringed) in August 2014. Although there are many wetlands in the Ethiopian highlands similar in altitude and vegetation to Berga, almost all have become unsuitable owing to overgrazing, and monitoring of wetlands in the vicinity of Berga and Weserbi has failed to produce any records (Y. D. Abebe and G. Gebreselassie in litt. 2013). The Area of Occupancy is estimated at just 5.5 km2 at the three Ethiopian breeding sites in 2013 (based on 150 ha of suitable habitat at Weserbi, 100 ha at Bilacha, and 300 ha at Berga; H. Smit-Robinson in litt. 2013), and searches of apparently similar sites nearby have failed to find the species (Y. D. Abebe and G. Gebreselassie in litt. 2013).

Whether a single population migrates between Ethiopia and South Africa, or each country hosts its own subpopulation, is still not certain (Taylor and van Perlo 1998, Barnes 2000), although observations from a breeding site in Ethiopia discovered in 2005 show that birds continue to breed into the dry season and may remain in Ethiopia after breeding, rather than migrate (A. Tefera per Anon. 2006), and genetic analyses possibly suggest the two subpopulations may intermingle (Symes et al. in prep). Suggestions that the non-breeding grounds for the Ethiopian population could be within south-west Ethiopia have not been substantiated; seven days of intensive surveys in the Kaffa area in April 2013 found no evidence of the species at seven major wetlands (M. Ewnetu in litt. 2013). Genetic and isotopic analyses are in progress and suggest that the species may be migratory (Smit-Robinson 2014, Dalton et al. submitted, C. Symes in litt. 2016).


Behaviour: The movements of this species are not fully understood. Lack of subspeciation has been interpreted to imply that the birds migrate between the two range areas (del Hoyo et al. 1996, Taylor and van Perlo 1998, Dalton et al. submitted), and this is supported by the fact that breeding has not been observed in South Africa where it is considered by many to be a non-breeding summer visitor (Urban et al. 2005). However the fact that there are overlaps in occurrence has prompted suggestions that strict migration is unlikely (Taylor and van Perlo 1998). One suggestion is that long-distance dispersal occurs when numbers are high (del Hoyo et al. 1996, Taylor and van Perlo 1998), with local movements being predominant at other times. In 2012 a survey was carried out in south-west Ethiopia during the non-breeding season to understand more about the possible movement of birds between Ethiopia and South Africa, however, no birds were found (Drerup 2015). In Ethiopia birds that breed in the central highlands in June-September may move to lower-level habitats during the non-breeding season when the highland areas becomes unsuitable (del Hoyo et al. 1996, Taylor and van Perlo 1998). In South Africa the species is thought to be nomadic, moving in search of its transient habitat (del Hoyo et al. 1996, Anon. 1997c, J. S. Ash in litt. 1999). It is also considered nomadic in Zimbabwe (D. Ewbank in litt. 2014). Birds in Ethiopia are present between June and October (J. S. Ash in litt. 1999), while non-breeding birds in South Africa are present from November to March (P. B. Taylor in litt. 1999), with a few records in May, August and September (Urban et al. 2005). Breeding occurs in July-August (Taylor et al. 2004). Breeding birds occur at a density of 2-4 pairs per hectare (Taylor and van Perlo 1998). Non-breeding birds occur in loose associations (Taylor and van Perlo 1998).

Breeding The species breeds in high-altitude seasonal marshes (between 2,200 and 2,600 m) with dense, rapidly growing vegetation dominated by sedges, grasses and forbs (Taylor 1996, Taylor and van Perlo 1998). It occurs here when vegetation has reached 20-40 cm in height and the ground has not yet become entirely flooded (Taylor et al. 2004). Very soon after hatching, it appears to move its chicks to areas of denser vegetation where the ground is more deeply and continuously flooded (Taylor et al. 2004). The species may have specific microhabitat requirements that have not yet been established. Non-breeding In South Africa it inhabits moist to flooded peat-based habitats (Urban et al. 2005) (mostly at 1,100-1,900 m) where vegetation is dense and dominated by sedges (Carex species) (Urban et al. 2005, Davies et al. 2015), although it is occasionally found in pure stands of Bulrush and reeds (Urban et al. 2005). It forages in mud at the edges of reed beds, in shallow water, in floating mats of aquatic vegetation and occasionally on dry ground (Urban et al. 2005). Of the 10 important sites for the species in South Africa, 9 are within the Eastern Uplands, Great Escarpment Mountains and Highveld peatland ecoregions, emphasising the importance of peat-based habitats (Taylor and Grundling 2003).  Over the next three summer seasons, BirdLife South Africa, Middelpunt Wetland Trust and University of Witwatersrand will use walk in traps and camera traps at the Middelpunt Wetland to investigate in more depth this species's habitat use in South Africa (H. Smit-Robinson in litt. 2016). In 2002, a new site was discovered in northern coastal KwaZulu-Natal following speculation that the species no longer occurred in coastal areas, and further sightings at various locations in Free State and KwaZulu-Natal have been made (Taylor and Grundling 2003, Davies et al. 2015).

It feeds on seeds and vegetation (De Smidt 2003) as well as insects, spiders, earthworms, small frogs and small fish (Urban et al. 2005). The stomach contents of a deceased chick included coleoptera (Dystiscidae) imagines, Diptera larvae (Tipulidae and Tabanidae), and the remains of small crustaceans. Breeding site Nests found in Ethiopia are described as a ball of woven live sedge, Eleocharis or Cyperus, and other plant stems and vegetation, with clutches of 4-6 eggs (Allan et al. 2006, Taylor et al. 2004). Observations at a nest found in August 1999 resulted in an estimated incubation period of 15-16 days.


Seasonal marshes are threatened by drainage (for cultivation and forestry), flooding by dams, catchment erosion, water abstraction, human disturbance, too-frequent burning, and excessive trampling and grazing by livestock and cutting of marsh vegetation for fodder (Atkinson et al. 1996a, Taylor and van Perlo 1998, P. B. Taylor in litt. 1999). Observations in Ethiopia suggest that it moves its chicks very soon after hatching to areas of denser vegetation and deeper flooding before the vegetation at nest sites has grown enough for cutting by local people (Taylor et al. 2004). Grasses and sedges are cut for the culturally important Ethiopian coffee ceremony (De Smidt 2003). In Ethiopia, a serious problem is the rapid growth in the numbers of livestock at around 2.4% per annum, and the resultant grazing of breeding habitat to a very short sward length, such that now the remaining sites for this species are threatened (M. Drummond in litt. 2005, H. Smit-Robinson in litt. 2016).

Bilacha wetland has already been largely converted into agriculture, settlement and grazing land. Suitable habitat at Weserbi wetland has been reduced to almost zero and housing is spreading close to the remaining small patch of wet grassland (M. Ewnetu in litt. 2013). Even the remaining site at Berga is highly threatened as habitat degradation and destruction continue unchecked despite attempts to improve the conservation status of the area. In addition to the existing overgrazing, trampling and grass cutting, local people are encroaching further down to the wetland, converting grassland for susbsistence agriculture and dividing up small plots of land for grazing. Suitable habitat here has already been reduced from c. 400 ha to only 200 ha (M. Ewnetu in litt. 2013).

The peatlands of South Africa are threatened by cultivation, afforestation, grazing, water abstraction, horticulture, peat fires, draining, headcut and donga erosion, siltation, fences and developments such as roads and dams (Taylor and Grundling 2003). The construction of the Braamhoek pumped storage scheme at Bedford Chatsworth marsh in eastern Free State, South Africa, was feared may caused disturbance and damage to habitat (De Smidt 2003), but it has only affected 4.5% of wetland and that is far from areas where this species has been recorded (M. Drummond in litt. 2016). The primary current threat in South Africa is development pressure from the mining industry (H. Smit-Robinson in litt. 2013); important wetland habitat for the species in Mpumalanga province is threatened by mining activities (Uys 2012), and if this species is shown to be migratory then birds could be affected by similar issues on the migratory route (Evans et al. 2015). Genetic analysis suggests that this species may have a low genetic diversity, and so could be limited in its ability to cope with environmental changes (Dalton et al. submitted).

Conservation actions

Conservation Actions Underway
CMS Appendix I and II. Some South African sites have some legal protection, and at least four sites are protected by the landowners (Barnes 2000). At the largest Ethiopian breeding population, the vegetation is not cut for fodder until October-November (M. Wondafrash in litt. 2007), thus giving the birds time to breed without disturbance (Anon. 1997c). In South Africa, the Middelpunt Wetland Trus was formed in 1994 with the main objective of conserving the species and its habitat (De Smidt 2003, M Drummond in litt. 2005). One of the Ethiopian sites, Berga, is on a state-run dairy farm, and formerly so was Weserbi. The farm at Weserbi has been privatised, but the marsh still remains under the control of the central Dairy Farm Enterprise based in Addis Ababa (M. Wondafrash in litt. 2007). There is a Site Support Group at Berga, formed by the Ethiopian Wildlife and Natural History Society (De Smidt 2003). Support from organisations including the Middelpunt Wetland Trust helped to build a primary school (which is now named after the species) for the Berga community and the villagers acknowledge the value of the species (Drummond 2013), and a document is in preparation to formalise the Berga community project (H. Smit-Robinson in litt. 2016). In 2003, a partnership was formed to mitigate the effects of the Braamhoek (formally changed to Ingula ) pumped storage scheme. Ingula is in the process of being proclaimed as a nature reserve (H. Smit-Robinson in litt. 2016). Verloren Vallei, Seeikoeivlei and Ntsikeni Vlei are formally protected (H. Smit-Robinson in litt. 2016). In June 2003, a national Species Action Planning stakeholder workshop was held in Wakkerstroom, South Africa, to assess the threats facing the species in this country, and concluded with the agreement that a South African White-winged Flufftail Action Group be established (De Smidt 2003), which was formed by the Department of Environmental Affairs in South Africa in 2016 (H. Smit-Robinson in litt. 2016).

An International Species Action Plan was published in 2008 (Sande et al. 2008). A proposal was put forward in 2006 to initiate a captive breeding programme in August that year, based in Pretoria Zoo, and using eggs taken from Berga marsh, Ethiopia (Tarboton and Wondafrash in prep.). The aim would be to study the species's life history and behaviour (Tarboton and Wondafrash in prep.). However, there were concerns that the programme should be carried out in Ethiopia, where it is known to breed, and that releasing birds into its non-breeding range could result in hybridisation with similar species (P.K. Ndang'ang'a in litt. 2006). While the captive breeding programme did not go ahead straight away (M. Wondafrash in litt. 2007), it has started and although releases into the wild are not the main priority, this will occur if the time arrives that makes this necessary (H. Smit-Robinson in litt. 2016). Captive breeding may be the best way forward for this species until its natural breeding grounds receive better protection (D. Allan in litt. 2016). An AEWA Small Grants Fund supported project in Ethiopia ran from July 2012 to September 2014 (Drerup 2015). The project held four awareness-raising workshops, distributed 1,000 copies of a brochure about the species to schools, local government and the local community and produced a billboard poster to increase awareness. The project enabled protection of the species at its breeding site in the Berga wetlands: 3 ha were fenced off and members of the Site Support Group patrolled the area (these activities were to continue post-completion of the project); soil conservation measures were also implemented to reduce erosion (Drerup 2015). DNA analyses suggest that the species may be migratory (Smit-Robinson 2014, Dalton et al. submitted, C. Symes in litt. 2016).

Conservation Actions Proposed
Maintain and restore suitable habitat at breeding areas in Ethiopia through sustainable use under community-based conservation programmes, and it has been proposed that the conservation statue of the Berga Wetland be upgraded (Atkinson et al. 1996a, Taylor and van Perlo 1998, M. Drummond in litt. 2016). Protect additional sites in South Africa, and a long term goal is also to declare Middelpunt Wetland a nature reserve (Barnes 2000, Evans et al. 2015, H. Smit-Robinson in litt. 2016). Continue surveys in Ethiopia and southern Africa to better define its range, population, seasonal movements and habitat requirements (Atkinson et al. 1996a, A. Shimelis in litt. 1998, Barnes 2000). Locate new breeding sites (Taylor et al. 2004, M Drummond in litt. 2005). Rehabilitate degraded wetlands (Taylor et al. 2004). Conduct research to determine the extent of the species's dependency on mire habitat in South Africa (Taylor and Grundling 2003). Ensure integrity and reduce disturbance at eight sites in South Africa (De Smidt 2003). Confirm that the species migrates between Ethiopia and South Africa (De Smidt 2003). Determine and record its principal calls for field studies (De Smidt 2003).


14 cm. Tiny rail. Adult male has chestnut head. Both sexes have black-barred chestnut tail and white wing-patches (very obvious in flight, not visible at rest). Similar spp. No other flufftail Sarothrura has white wing-patches. Female paler below than other female flufftails Voice Was believed to make a soft, double-noted hooting, but it has been proposed that these calls may in fact refer to nocturnal calls of Grey Crowned-crane Balearica regulorum. Hints Best chance of seeing this secretive bird is during the wet season in upland marshes in central South Africa and Ethiopia.


Text account compilers
Benstead, P., Ekstrom, J., Evans, M., Pilgrim, J., Shutes, S., Symes, A., Taylor, J., Ashpole, J, Westrip, J.

Allan, D., Gerrans, C., Robertson, P., Dowsett-Lemaire, F., Symes, C., Kariuki Ndang’ang’a, P., Shimelis, A., Leonard, P., Drummond, M., Taylor, P.B., Ewnetu, M., Tarboton, W., Dowsett, R.J., Nel, O., Lawlor, R. J., Madden, S., Gebreselassie, G., Pretorius, R., Anderson, T., Abebe, Y., Ash, J., Wondafrash, M., Ewbank, D., Smit-Robinson, H.

Recommended citation
BirdLife International (2017) Species factsheet: Sarothrura ayresi. Downloaded from on 20/10/2017. Recommended citation for factsheets for more than one species: BirdLife International (2017) IUCN Red List for birds. Downloaded from on 20/10/2017.