July 2003 Bird Numbers 12 (1)

Breeding biology of the Southern Grey-headed Sparrow in the Namib-Naukluft Park, Namibia P.A. Bridgeford NamibRand Nature Reserve, PO Box 43, Maltahöhe, Namibia The Southern Grey-headed Sparrow Passer a similar trend in choice of nest site. Only diffusus is widely distributed in southern two were in natural sites, namely a wood- Africa, except in southwestern Namibia and pecker nest and a hole in the broken end of southwestern Western Cape Province of a branch. Southern Grey-headed Sparrows South Africa (Craig 1997). Despite it being used the Zais nest in the asbestos chimney so common, little is known of its breeding for over 20 years. The nests were filled with biology. Maclean (1993) stated that incuba- feathers, but a few pieces of paper and cloth tion and fledging periods are unrecorded, were also found. Indeed, most nests were so while Tarboton (2001) said that incubation is filled with feathers that it was difficult to see not recorded and nestling/fledging period is eggs or chicks. c.19 days. This paper gives some new infor- mation on the breeding biology of this spe- Breeding season cies in Namibia, based on Southern Grey- Breeding seasons are slightly later in Na- headed Sparrows monitored at Zais (24°01'S, mibia than elsewhere and coincide with rela- 16°09'E) in the Namib-Naukluft Park from tively late summer rains. This study, with 17 1986 to 1993. records, showed January (24% of all active The study area, below the escarpment in nests), February (29%), and March (24%) as the pro-Namib, is in a semi-desert and the peak breeding months at Zais (Table 1). savanna transition zone (Giess 1971). How- The summer rainfall, which generally starts ever, as Zais is situated next to the Tsondab in January, is highest in February/March River, the area is mainly Acacia woodland (Table 2). For Namibia as a whole, the peak- dominated by A. tortilis. The majority of the breeding season was also February (28%) breeding records were from nests in the gar- and March (28%), but extended from Janu- den. Where possible, the nests were checked ary through to May (Table 1). It is interest- visually or by inserting a hand into the nest. ing to note that four of the 100 records for They were monitored as regularly as possi- Namibia are from the Hoesch and Nietham- ble, but most observations were interrupted mer expedition in 1933/34. Tarboton (2001) by official duties away from home. Some gave the breeding season as mainly Novem- eggs were removed from easily accessible ber–March throughout southern Africa. nests and measured with vernier callipers. Although many birds were examined Egg size and clutch size while ringing, no sexual dimorphism could Clutches were 3–4 eggs (mean 3.3 eggs) at be determined. Some ringed birds were sexed Zais (Table 3). This corresponds with clutch after they were seen mating. sizes in Maclean (1993) and Tarboton (2001). Five eggs were measured: range Nest sites 18.2–19.4 mm × 14.2–14.4 mm (mean All the nests at Zais were in man-made struc- 18.9 mm × 14.28 mm). This differs only tures, such as an unused asbestos chimney, marginally from the 19.2 mm × 14.3 mm nest boxes, wall cavities, and the chassis of given by Maclean (1993) and Tarboton a wrecked vehicle. Nest record cards from (2001). Eggs were grey/brown and heavily the Ornithological Section, Ministry of En- blotched with dark brown. vironment and Tourism in Windhoek, show

31 Bird Numbers 12 (1) July 2003

Table 1. Breeding season of Southern Grey-headed Sparrows in Namibia.

Number of nests with eggs or chicks

N Nov. Dec. Jan. Feb. Mar. Apr. May

Zais 17 1 0 4 5 4 1 2 Whole of Namibia 100 1 4 9 28 29 19 10 Total 117 2 4 13 33 33 20 12

Incubation To determine who was incubating at night, I put up a net around the entrance be- Only eight records of incubation period were fore first light and only the colour-ringed obtained; these were 11–14 days (mean 12.5 female emerged. This was the only time I did days). Incubation for the congeneric Cape this. More observations are needed to deter- Sparrow P. melanurus and House Sparrow P. mine with certainty which parent incubates at domesticus is 12–14 days (Maclean 1993; night. Of the other three Passer species in Tarboton 2001). southern Africa, Maclean (1993) stated that In this study, it was found from colour- both male and female incubate, and in the ringed birds that the male and female both Cape Sparrow, the female incubates at night. incubate during the day. Possibly because breeding is in the hottest months of the year, Nestling and post-nestling period the one regularly observed nest was fre- The nestling period was 16–19 days (N = 5; quently left unattended by both adults. At mean 17 days). This is identical to that re- 14:00 on 6 April 1886, at the chimney nest, corded for both Cape and House Sparrows six days after commencing incubation, the (Tarboton 2001). Of all the observations of ambient temperature was 30°C, while the breeding Southern Grey-headed Sparrows, temperature in the nest was 32°C. Both only three records of the post-nestling period adults were in and out of the nest during this are considered to be of any value. These were time and on occasions left the nest for peri- 14, 17, and 21 days respectively. Both par- ods of 3, 7, 10 and 11 minutes. Both birds sat ents fed the chicks, but the female fed more on the edge of the nest with beaks open and often. wings slightly spread after emerging from the nest. Summary The breeding biology of the Southern Grey- headed Sparrow was studied in the Namib- Table 2. Total monthly rainfall in millimetres at Naukluft Park, Namibia, and new infor- Zais 1986–1993. mation gathered on various life-history variables. The majority of nests were in man- Nov. Dec. Jan. Feb. Mar. Apr. May made structures. The peak-breeding season was January–March, generally with 3–4 eggs 10.7 25.4 119.1 242.1 142.5 108.1 9.9 per clutch with both parents incubating. The incubation period was 11–14 days and the nestling period 16–19 days. The post- Table 3. Clutch size of Southern Grey-headed nestling period was 14–21 days, with both Sparrow at Zais, Namibia. parents feeding the chicks. No. of eggs per clutch 2 3 4 5 Giess W. 1971. A preliminary vegetation map of No. of records 1 7 5 – S.W.A. Dinteria No. 4 S.W.A. Scientific

32 July 2003 Bird Numbers 12 (1)

Society, Windhoek. Craig A.J.F.K. 1997. Southern Greyheaded Spar- Maclean G.L. 1993. Roberts’ birds of southern row Passer diffusus. In: Harrison J.A., Allan Africa. John Voelker Bird Book Fund, Cape D.G., Underhill L.G., Herremans M., Tree A.J., Town. Parker V. & Brown C.J. (eds). The atlas of Tarboton W. 2001. Nests and eggs of southern southern African birds.Vol.2. BirdLife South African birds. Struik, Cape Town. Africa, Johannesburg. o o o

The 9th International Roseate Tern Workshop, Wexford, Ireland, April 2003 A.J. Tree ADU Research Associate, PO Box 211, Bathurst, 6166. [email protected]

Memories, memories, memories . . . came gion of the mitochondrial genome in an at- flooding back when I returned to my home tempt to identify different populations (re- country, Ireland, after a break of more than sults for South Africa pending complemen- 40 years. Blackbirds, Robins, Dunnocks and tary material from the Atlantic becoming Wrens singing in the hedgerows, flights of available for comparison). Greenland White-fronted Geese passing For many years, estimates of world totals over, honking excitedly as orderly skeins of Roseate Tern were given at around 30 000 departed for their breeding grounds; daffo- pairs, but recent extensive coverage of the dils and primroses flowering everywhere, multitudes of islands around the Australian horse chestnuts and sycamores bursting into coastline has come up with some staggering leaf and draught Guinness in the pubs. And figures. A minimum estimate of 70 000 pairs believe it or not, no rain! Putting aside remi- there alone puts Australia into worldwide niscences, it was also great to meet people I perspective and helps to make that continent had known only through correspondence or the tern capital of the world. Thus, the world the literature. So why Ireland for this work- population is now estimated at about 85– shop? Well, the southeast of the country is 90 000 pairs. Although we have a mere 240 now home to the vast majority of Western pairs of Roseates in South Africa, it must be Europe’s Roseate Terns, hence the venue for remembered that these probably represent an the workshop. isolated population of unknown origin and Why are these Roseate Tern workshops therefore remain classified as Endangered. convened? In certain parts of the world, the The workshop was convened by Steve Roseate Tern is classified as Threatened and Newton and Oscar Merne, both from Eire. because of the deteriorating situation in the Oscar is a contemporary of mine, each north Atlantic, researchers from North known to the other, but we had never met. America, Western Europe and the Azores The gift of the gab, common to so many came together to discuss their problems and Irishmen, he certainly has and his knowledge find solutions. From this has grown the Ro- of local history made the region come to life seate Tern workshop. This year, fellow re- in ways other than ornithological. Most searchers from elsewhere in the world were speakers were from America, where detailed invited to attend and contribute. There were research has been carried out on the east- only two southern hemisphere delegates, coast metapopulation for many years now. Anna Lashko from Australia and me. Anna This group was led by Ian Nisbet, the doyen has been working on sequencing the ND6 re- of tern research in the north Atlantic and

33 Bird Numbers 12 (1) July 2003 again a name known to me since my teens, ous monitoring of returning breeding adults prior to his move to the States and mine to each year gives a good idea of individual Africa. survival, strength of mate bond (the stronger The greatest emphasis was placed on the bond the better the chance of rearing management techniques used to re-establish young), improving breeding success with populations that had drifted to a critically gathering years and experience, as well as low level in the Atlantic, and the resultant average longevity of adult birds. Although impact of these techniques. Greatest benefits the oldest known living bird was still going have accrued to the Common Tern, followed strong at 26 years, few appear to reach 20 by the Arctic Tern, but results for the Rose- years. Although birds may visit colonies ate Tern have been very variable and produc- from one year of age, first attempted breed- tivity is low at many sites, particularly sites ing occurs only from years three to five and classified as marginal. Basic techniques in- is rarely successful at first attempt. Special clude: Roseate Tern rings are now in use and these can be read in the field with the use of tele- q creating suitable habitat on islands previ- scopes (much in the way we have been doing ously known to harbour this bird; with Swift Terns recently). These rings will q culling or scaring the burgeoning gull be put to use in South Africa from this com- populations that had literally pushed the ing breeding season. terns off many islands as well as predat- Atlantic populations of Roseates have ing eggs and young; been well studied but little is known of popu- q establishment of a variety of nest boxes, lations in the Indian Ocean, although the the most effective of which appears to be Australians have recently embarked on a unsightly old car tyres sunk into the large-scale ringing and colour-flagging substrate at an angle; project. The number of populations making q and the use of tape recordings to attract use of Australia is not known, but there passing birds. appear to be four separate breed- remove the rings ing populations, each with its own All methods have proved effec- for ornaments or time schedule, as well as birds tive to various extents at different trophies from Japan and Taiwan wintering sites, but the European population in the northeast. To this may be (at just over 900 pairs) has another factor to added likely visitors from the seas around In- contend with. These birds migrate to winter donesia, the Phillipines and further east in quarters in West Africa, especially coastal the Pacific. At this stage, these birds can be Ghana, where young lads catch them for separated only in the hand, based on the food or to remove the rings for ornaments or stage of moult. Thus the peak numbers of trophies. Expeditions have visited the area to birds visiting that continent could well attempt to remedy this situation, and at the number over 200 000. A lot of birds for a same time to establish a local working group species considered internationally Endan- to monitor these happenings. To date they gered only a decade ago. seem to have had a modicum of success. The only other population that is being Other studies looked at the problems of studied, other than our own, is that of Aride population dynamics, including productivity, in the Seychelles. Jaime Ramos, from Portu- survival and longevity. Productivity is fairly gal, has been studying these birds annually easy to monitor at certain accessible colo- since the early 1990s, with earlier studies nies, but once birds leave those colonies, back to 1984, and has collected and pub- survival/mortality factors are difficult to de- lished much data on nest-site selection, for- termine and it is here that recovery data from aging habitats, diet, chick-feeding rates, ringed birds becomes so important. Continu- chick growth and survival, and influence of

34 July 2003 Bird Numbers 12 (1) weather conditions on reproductive par- to become pelagic away from the breeding ameters. The breeding population has island and may come to land mainly at night, crashed from approximately 4500 pairs in the or may depart our shores almost completely. 1970s to 1267 pairs in 2002. But how genu- We do not know if our population is an iso- ine this crash may be cannot be fully deter- late or part of a larger metapopulation, such mined as there is a possibility that birds as that of Madagascar with which our birds move around the northwestern Indian Ocean, seem to show the greatest affinities. We have breeding at several sites, and nothing is never been able to do any intensive studies known of their post-breeding ecology. Pro- on productivity, although we have a rough ductivity of tropical Roseates appears to be idea of the number of birds breeding and the lower than that of temperate birds, both in the fact that they seem to be increasing slowly. average number of eggs laid and in the We have little data on survival of the young number of young fledged, thus it is possible of different years as we have little idea of that the overall survival rate may be higher. what proportion of the well-hidden young we There is also a suspicion that tropical birds actually catch and ring each season, nor do may not breed consist- we know if all our off- ently every year. burgeoning gannet population spring return to breed at Comparative studies that is spreading over the terns’ either Bird Island or of foraging behaviour main breeding area Dyer Island. are also of interest as the Management of the distance to feeding Bird Island population grounds may vary enormously between dif- is not a priority as the only problem, other ferent populations. It would appear that the than wet spells damping and chilling eggs species of food fish brought to the chicks and chicks, is the burgeoning gannet popu- varies from year to year and the calorific lation that is spreading over the terns’ main food value of the fish varies from one species breeding area. But it is at the even more to another. Those young fed on rich food fish crowded Dyer Island that management is a in one year will have a better chance of sur- necessity because Cape Cormorants destroy vival to adulthood than those fed on less rich the terns’ vegetation cover for nest-building, foods. exposing the eggs and young to the depreda- Although much has been learnt around tions of numerous Kelp Gulls. Establishment the world, there is a huge field open for fur- of suitable nest-boxes is a priority. ther research into various factors affecting All-in-all, the workshop was very worth- populations of Roseate Terns, both annually while, especially in making contact with and in the long term. In South Africa, Rod other researchers. The hotel accommodation Randall set the scene with his publications in was great, the food good, field trips interest- the 1970s. Norbert Klages and I have added ing (I even saw a lifer – a Snow Goose) and our contribution over the last five seasons, on the last evening we had dinner at the Her- but there is great difficulty and expense in itage Centre where we were entertained with accessing Bird Island during the winter traditional Irish music and dancing. months, and our visitations have been too short and variably infrequent. The Bird Acknowledgements Island colony is difficult to observe, hence My attendance at this meeting was supported we do not have the options of close study as by a National Research Foundation (Sea and at some colonies. Coast 2 Programme) grant to Prof. L.G. We have little idea whether our birds are Underhill, and by the 9th International Rose- resident or migratory, although they do seem ate Tern Workshop. o o o 35 Bird Numbers 12 (1) July 2003

Spotted Ground Thrush threatened by forest destruction With less than 2500 individuals remaining in that urgent and immediate action be taken to the world, the Spotted Ground Thrush Zoo- conserve the species. Recommended activi- thera guttata is one of Africa’s most endan- ties include raising awareness, research and gered birds and amongst the most sought monitoring, and effective management of after by birders. Some populations of this forest habitats. fast-disappearing thrush migrate between In his opening speech, Mr Tito Mbuvi, African countries. It would be futile to pro- Centre Director of Kenya Forestry Research tect the bird in one country only for it to die Institute (KEFRI), Gede, and also Secretary for lack of suitable habitat in another. Five for the Arabuko-Sokoke Forest Management races of this striking ground-dwelling thrush Team, emphasized the importance of involv- are recognised. They exist in what are now ing local communities and other stakeholders isolated patches of indigenous forest. Two adjacent to Spotted Ground Thrush sites if are migratory coastal races, one in Kenya and conservation actions are to succeed in the Tanzania, and the other in South Africa. The long term. three sedentary races are found in small for- Colin Jackson, Director of A Rocha est fragments in Democratic Republic of Kenya, said that from observations in Congo, Malawi, and Sudan, and recent un- Arabuko-Sokoke, Spotted Ground Thrushes confirmed observations appear to require good suggest also Mozambique. Five races of this striking quality forest for their sur- Under the auspices of ground-dwelling thrush are vival and thus could be the BirdLife Africa Spe- recognised. used as an indicator of a cies Working Group, a healthy forest environ- workshop of international ment. Furthermore, local experts was held at Turtle Bay Beach Club, communities can benefit from birdwatchers Watamu, Kenya from 5–9 May 2003. It was visiting sites such as Arabuko-Sokoke Forest jointly hosted by NatureKenya and National to see the thrush and, in so doing, contribute Museums of Kenya (NMK), facilitated by to schemes like the existing bursary fund, NatureUganda and BirdLife South Africa, Arabuko-Sokoke Schools & Eco-Tourism and co-funded by the Royal Society for Pro- Scheme (www.assets-kenya.org). tection of Birds (RSPB) and the UK Darwin A presentation on the species was given Initiative. The aim of developing an Interna- by Kariuki Ndang’ang’a (Research Scientist tional Conservation Action Plan for this en- at NMK). Doug Harebottle (Avian Demog- igmatic species was achieved, and an Inter- raphy Unit, University of Cape Town), who national Spotted Ground Thrush Working has extensively studied the species in South Group was set up to coordinate the imple- Africa, said that much more information is mentation of the Plan. needed for effective conservation of the spe- The 16 NGO and Government partici- cies. pants in the workshop were drawn from The workshop concluded that, with active BirdLife International partners in South support for the Species Action Plan from the Africa, Malawi, DRC, Kenya and Tanzania. governments and conservation bodies in The workshop identified appropriate stake- countries where the bird occurs, there is hope holders, activities and methods to be in- for the continued survival of this elusive and cluded in the Action Plan and recommended threatened bird. For more information contact Kariuki Ndang’ang’a, NatureKenya, PO Box 44486, 00100 GPO, Nairobi, Kenya. Tel: +254-(0)20-3749986/57; Fax: +254-(0)20-3741049; email: [email protected] or [email protected]

36 July 2003 Bird Numbers 12 (1)

White-billed penguins Phil Whittington Dept. of Zoology, University of Port Elizabeth, Box 1600, Port Elizabeth, 6000

The Avian Demography Unit, Marine and Coastal Management, Robben Island Mu- seum and Earthwatch Institute are currently conducting research on the African Penguin Spheniscus demersus at Robben Island. An important component of this work is the more field-testing of a new design of flipper band surprised (Whittington 2001). The project necessitates when I observed regular monitoring of a number of study another African Penguin with a partially nests, that is carried out by teams of Earth- white bill on Bird Island, Algoa Bay on 3 watch volunteers together with staff of the February 2003. This bird had an entirely participating organisations. white upper mandible but a normal black On 13 July 2002, while carrying out nest lower mandible, including the lower half of monitoring with Earthwatch Team Five, an the greyish vertical band (see sketch). adult penguin was noticed at a nest with two There are several examples of partial, and large, downy chicks. There was something a few of complete albinism in penguins in the different about this bird. After a couple of literature (e.g. Rand 1950; Van Wyk 1995) seconds, I realised that it had an almost com- but I was unable to find any mention of birds pletely white bill. Normally, the bill of the with normal plumage and a leucistic bill. African Penguin is black, with a variable Two other seabird biologists, with over 25 greyish vertical band towards the tip. I as- years of field experience of the African Pen- sumed the occurrence of a white-billed bird guin between them, could not recall having to be a form of albinism. The plumage of the seen a penguin with a white bill. Therefore, bird was, however, perfectly normal (see I assume this to be a rare phenomenon. photograph). During my seven years of fieldwork on Acknowledgements this species, I have seen several examples of Earthwatch Institute and the University of partial albinism in the plumage (see also Port Elizabeth are acknowledged for their BN9[1]: 22), but this was the first occasion support of seabird monitoring work in 2002 on which I had witnessed a bird with any- and 2003, respectively. Robben Island Mu- thing other than a black bill. So I was even seum and in particular, Mario Leshoro, gave logistical support at Robben Island. Marine and Coastal Management kindly allowed the use of their accommodation facilities at Rob- ben Island. Robben Island Museum and Brian Howe provided transport to the islands.

Rand R.W. 1950. Albino penguins. Ostrich 21(1): 33. Van Wyk J.C.P. 1995. Unusually coloured pen- guins at Marion Island, 1993–1994. Marine Ornithology 23(1): 58–60. Whittington P.A. 2001. And the bands played on … Bird Numbers 10(1): 44–45. 37 Bird Numbers 12 (1) July 2003

Firefinches and sunbirds on the move R. Cowgill1, S.B. Davis1 & D. Harebottle2 159 Edmonds Road, Glenwood, 4001 2Avian Demography Unit, UCT, Rondebosch, 7701

Over the past three decades, a number of Another isolated report came from Umboga- KwaZulu-Natal bird species have extended vango south of Durban on 27 August 1991. their ranges, and in doing so have extended Five years later in May 1994, further sight- the types of habitat that they use. Some of ings were made at Newlands East in Durban, these changes can be followed from past and in July 1994 of a pair in the coastal low- records while others have few records. The land forest at North Park. In June 1994, sev- species considered here are Red-billed Fire- eral records also came from Umbogavango finch Lagonosticta senegala and Purple- Nature Reserve, 20 km south of Durban banded Sunbird Nectarinia bifasciata. Of (Cowgill & Davis 1998), also in thick coastal these two species the Red-billed Firefinch lowland forest. These records reflected a sig- has the most available records. nificant influx of birds into the greater Dur- ban area. In May 1995, sightings were re- Red-billed Firefinch corded in Ilanda Wilds Nature Reserve, also Clancey (1964) recorded the Red-billed Fire- a reserve of coastal lowland forest situated finch as common in eastern Zululand with its along the Amanzimtoti River. range extending south to the lower Tugela During 1995, the first inland sightings River. In Natal south of the Tugela, it oc- were made at Shongweni in July, when a curred along the coast to Stanger and near small flock was seen in dry grassland with Durban (April 1962). juveniles being fed by adults. The next The Bird atlas of Natal and Zululand records came from the coastal lowland for- (Cyrus & Robson 1980) has only a July ests of in February record at Stanger and few records in western 1997. In August 1997, an isolated record Zululand. ASAB2 shows an extended distri- from Umdoni Park, Pennington, marked a bution to inland Zululand and an increased farther southward expansion. By May 1998, number of records from the region, reaching there were frequent records from further south to Ballitoville (Ballito). down the south coast at Port Shepstone, Both ASAB2 (Harrison et al. 1997) and Anerley and Umzumbe. By this time, regu- Clancey (1964) described the species’ habi- lar records were being received from both tat as thornveld and riparian thickets but also Amanzimtoti (July 1998) and Umkomaas gardens and chicken runs. Further records (December 1998) where the species occurred from individual observers’ and the Birds in regularly at bird feeders. By July 1999, Red- Reserves Project (BIRP) have revealed a billed Firefinches were well established and southward movement of this species during resident at Umkomaas, numbering 11–13 in dry months. An isolated record of Red-billed a single flock. From August 1999 onward, Firefinch came from Sea Cow Lake, Durban, regular records were received from Um- during the dry year of 1982. The earliest ob- twalume, between Umdoni Park and Um- servations during this time were generally zumbe, indicating that they had become a made at the onset of winter. resident species. In July 1989 a few pairs were observed in During mid-February 2000, a pair of Red- rank vegetation and grass near the old Mel- billed Firefinches bred successfully in Glen- ville sugar mill, 10 km southwest of Stanger. more, Durban, and reared three chicks by

38 July 2003 Bird Numbers 12 (1) mid-April (M. Hoile pers. comm.). The nest scribed by Clancey (1964) as common on the was built in a potted Plectranthus verti- littoral plains of Zululand with a few strag- cillatus, 2.5 m from a kitchen window on a glers on the coast of Natal, as far south as patio. In May of 2001, another set of three Durban. Cyrus and Robson (1980) recorded chicks was raised but not at the same site. this species as common in Zululand, reach- There was no evidence of breeding during ing south to Mtunzini, but only seasonal 2001, but in April 2003 a pair of adults were around Durban. Skead (1967) reported that accompanied by two feathered young at the this species inhabits riverine bush and occa- feeding table. sionally thorn trees. He recorded the species In 2001, the birds moved south during the as common in northern Zululand, but a rare winter months of June and July to Ramsgate, winter visitor in Durban at the southern edge 20 km south of Port Shepstone. In March of its range. It was not noted as a bird of 2002, a record was submitted from Uvongo coastal forests but as an occasional visitor for River Nature Reserve just north of Ramsgate. nectar. It seems evident that the Red-billed Fire- In Durban, the earliest record of note was finch moves south during dry periods, par- when the species frequented the mangroves ticularly in winter. There appear to have at Beachwood Nature Reserve in July 1978. been three phases of range extension, each The specific breeding date was not recorded, preceded by a small number of sporadic but it nested there during that year. The nest records south of the usual range. Prior to was constructed of fine fibres, covered with 1979, regular observations were made north spider webs and suspended from the end of of the Tugela River (2931AB), and by the a foliaged branch of a White Mangrove Avi- end of the atlas period in 1992, the birds had cennia marina, about 2.5 m above the water. become more common in the Stanger/Ballito Between the Natal Atlas and ASAB, there area (2931AD) but had not been recorded were only a few records. From the late 1980s further south. The species had established until the present, this sunbird was recorded itself in the Durban area (2931CC) by 1994, regularly at Beachwood and Durban Bay- and subsequently in Port Shepstone head. In November 1985, it was reliably re- (3030CB) by 1998 and Ramsgate (3030CD) corded as a summer visitor to Umhlanga by 2001. Ponds, in the rank vegetation around the There has not been a corresponding ex- ponds. In July 1987, it visited the forest mar- pansion of the Steelblue Widowfinch Vidua gins of Stainbank Nature Reserve (Cowgill chalybeata, a brood parasite of the Red- & Davis 1998). During July 1990, Purple- billed Firefinch. Two documented records banded Sunbirds arrived and finally estab- report sightings of this species firstly in lished themselves at both Beachwood Man- Umbogintwini where two males were feed- groves and the . In the ing on a lawn in February, 1999. This is pos- summer of that year, the birds settled in the sibly the most southerly record received so mangrove stands at Bayhead where they far. The second sighting was from SAPPI have remained as a resident species. Stanger in January 2001. Most visitations were during winter months. In June 1993, a few birds visited the Purple-banded Sunbird coastal lowland forest of Na- According to ASAB2, this sunbird has un- ture Reserve where they fed on the nectar of dergone a less dramatic expansion than the Natal Wild Banana Strelitzia nicolai. In the Red-billed Firefinch, but since the Purple- same year, records were regularly received banded Sunbird has diversified its habitat in from Umhlanga Lagoon and Umhlanga Durban, it has become far more numerous Wastewater Treatment Works, all in similar and has also spread southward. forest habitat. When the Common Coral Tree The Purple-banded Sunbird was de- Erythrina lysistemon and Water Berry

39 Bird Numbers 12 (1) July 2003

Syzygium cordatum were in bloom at Bluff billed Firefinch. Instead, the records indicate Nature Reserve in June 1994, these sunbirds a steady southward expansion with dispersal visited the site with a number of males in into non-specific habitat during winter. The eclipse plumage. In July 1995, Virginia Bush question is, what habitat and season is re- Nature Reserve, also a coastal lowland for- quired for it to breed down south? est, was the next site visited by this species. This was followed, in May 1996, by a sight- Other species and further monitoring ing made in Nature Reserve. Other species that require closer examination In July 1997, the species was recorded at of their range extensions include Woolly- Northern Wastewater Treatment Works necked Stork, Scarlet-chested Sunbird, Black where there is no coastal lowland forest. Coucal and White-eared Barbet. It is evident During 1992 to 1995, winter records that BIRP plays a significant part in the con- (June and July ) were received from Umbo- tinuous and regular recording that is required gavango Nature Reserve where a nest was to document changes in distribution ranges. later found in December 1999. The next sig- The personal observations of birders in areas nificant southward movement was in Sep- that lie outside of protected areas, particu- tember 1996 where the species was sighted larly domestic gardens, are also important. at Uvongo (3030CD), and then in Mpenjati An important function of Out-of-Range Re- Nature Reserve in August 1998. corders on bird club committees, is to evalu- Inland movements were recorded from ate and report on all observations in a suit- the first urban observations in Westville able format. The recorder should monitor the (2930DD) in 1992. Individual birds arrived distributions of certain species and draw intermittently in May to feed on a blossom- attention to these, requesting detailed ing Chinese Hat Plant Holmskoldia san- records. These records will provide more guinea. The birds were absent during 1993 accurate details of the distribution and move- but returned in 1994, leaving again in Sep- ment of birds, including as yet unexplained tember of the same year. Visits have been erratic or nomadic movements. regular since then to the present. In June 1999, single birds visited gardens on the Acknowledgements Berea, Durban (2931CC), and later in We wish to thank all those birders for their Cowies Hill in September 2001. In Novem- time and effort in submitting comments, ob- ber 2001, a pair took up residence at Durban servations, BIRP cards, and the ADU for Botanic Gardens. In July 2001, birds were supplying essential data to make this report seen at the Mbumbazi Nature Reserve ap- possible. proximately 15 km inland from Port Shep- stone (3030CB). By July 2002, the bird was Cowgill R. & Davis S.B. 1998. Bird checklists of recorded as a winter visitor at Umtentweni, KwaZulu-Natal reserves Part 1: The greater and at Port Shepstone in October 2002. An Durban area, Mondi BLSA Guide 21. BirdLife unusual record came from Spioenkop Nature South Africa, Johannesburg. Reserve (2829DA) when a dead specimen Clancey P.A. 1964. The birds of Natal and Zulu- land. Oliver and Boyd, Edinburgh. was collected at Spioenkop Lodge in Octo- Cyrus D. & Robson N. 1980. Bird atlas of Natal. ber 2001. Most of the records are of birds University of Natal Press, Pietermaritzburg. moving during winter months. It is still un- Harrison J.A., Allan D.G., Underhill L.G., known where they go during summer breed- Herremans M., Tree A.J., Parker V. & ing months. No recent records of successful Brown C.J. (eds). 1997. The atlas of southern breeding have been received from south of African birds. BirdLife South Africa, Johannes- Durban. burg. It is not evident that the Purple-banded Skead C.J. 1967. The sunbirds of southern Africa. Sunbird is spreading in phases like the Red- South African Bird Book Fund, Cape Town.

40