Shoreline birds of Robben Island, Western Cape, South Africa L.G. UNDERHILL, P.A. WHITTINGTON & K.M. CALF Avian DemographyUnit, Universityof Cape Town, Rondebosch,7701, SouthAfrica, e-mail: Igu @maths.uct. ac. za Underhill,L.G., Whittington,P.A. & Calf, K.M. 2001. Shorelinebirds of RobbenIsland, Western Cape, South Africa. Wader StudyGroup Bull. 96: 37-39. RobbenIsland, with a perimeterof 10 km, lies nearthe port of CapeTown, SouthAfrica. Sevensurveys of the shorelinebirds have been made. Of Palearcticmigrant wader species, Ruddy Turnstones and Curlew Sand- piperswere the mostabundant; of the non-Palearcticmigrants, the African Black Oystercatcherwas most abundant.Robben Island qualifies as a Ramsarsite. The possibilityis suggestedthat RobbenIsland could be used as a "remote site" in a network of sites to monitor trends in Palearctic waders. INTRODUCTION AND METHODS populationof this species(Barnes 1998). Numbersof two other African species, Spotted Dikkop and Blacksmith RobbenIsland (33ø49'S,18ø22'E) is bestknown as the place Plover, alsoappear to have increasedon RobbenIsland, but of imprisonmentof NelsonMandela. It wasdeclared a World White-frontedPlovers have inexplicablydeclined. HeritageSite in 1999, on accountof its culturalvalue. How- Four counts were made in the austral summer over two ever, RobbenIsland is alsoof biologicalimportance, espe- decades, one in 1980 and three between 1998 and 2001 cially for its seabirdpopulations and has been identified as (Table 1). These show that the speciescomposition of one of SouthAfrica's ImportantBird Areas (Barnes1998, Palearcticmigrant waderswas typical of rocky and mixed Crawford & Dyer 2000). For example, the island supports rocky-sandy shorelinesof the WesternCape (Ryan et al. importantpopulations of two "Vulnerable"seabirds: 12% of 1988). The majority of thesebirds comprised Ruddy Turn- the global populationof the African PenguinSpheniscus stones,Curlew Sandpipers,Whimbrels, Grey Plovers,Ringed demersus and 4% of Bank Cormorants Phalacrocorax Ploversand Sanderlings;other migrant wadersoccurred in neglectus(Barnes 2000). only small numbers. The island has an area of 507 ha, and a perimeterof ap- Numbersof Curlew Sandpipers,Ruddy Turnstonesand proximately 10 km (Figure 1). It lies in Table Bay, 11 km Sanderlings showed decreases(Table 1). However, the from the port of Cape Town, and 7 km from the nearest borealsummer of 1979 wasa particularlygood breeding year mainland(Barnes 1998, Crawford & Dyer 2000). Most of the for thesethree species (Summers & Underhill 1987), andthe shorelineis rockywith variousdegrees of exposure,depend- large countsmade in 1980 might have representedinflated ent on the extent of offshorereefs. South of Murray's Bay populationsof largelyfirst-year birds. Ongoing austral sum- Harbour is a 400-m sectionof sandy shore; this beach is mer surveysare neededto determinethis, and it is unfortu- probablykept in place by the jetty, which allows sandto natethat only a singlesummer historical survey on which to accumulate. base trends is available. Austral winter surveys are needed We know of seven round-island counts of shoreline birds to determine the extent to which the shoreline of the island (Table 1), i.e. speciesthat forage in or closeto the intertidal is usedby "overwintering"birds; most first-year wadersdo zone. Four of the surveyswere made in the australsummer not migrateto the breedinggrounds from non-breedingareas and three in winter. Four were made within the 10-month as far south as South Africa (Summers et al. 1995). It has periodNovember 2000 to August2001, two in summerand beensuggested that youngbirds move from exposedshore- two in winter. lines, such as that on Robben Island, which are subject to largewinter storms,to more shelteredplaces such as Lange- RESULTS AND DISCUSSION baanLagoon (Pringle & Cooper 1977; Underhill 1987). In winter 2001, 18 Ruddy Turnstoneswere observedin May, Of the non-Palearcticmigrant waders, the African Black beforethe passageof a seriesof severecold frontsin June Oystercatcherwas the mostabundant (for scientificnames and July. In the surveymade in August,none was observed of the studyspecies, please refer to Table 1). This speciesis (Table 1). It is therefore possiblethat these birds moved endemic to South Africa and Namibia, and is classified as away from RobbenIsland to find more shelteredhabitat. "Near-threatened" (Barnes 2000). The increase in numbers On the eveningof 31 March 2001, a flock of 136 Sander- of African Black Oystercatchersfrom 40 in 1977 to a median lingswas seen on the shortsection of sandybeach on Robben of 135 in the 1998-2001 surveys(Table 1) is probablyattrib- Island (PAW). Sanderlingsare thoughtto undertakea loop utable to decreased levels of disturbance on the shoreline. migration clockwise around southernAfrica (Underhill This has occurred since the mid-1990s when the Robben 1997). Therefore such short-terminfluxes of this species IslandMuseum took overmanagement of the islandfrom the probablyrepresent birds on passage. Departmentof CorrectionalServices. At currentpopulation Becausethe entire island is part of the Robben Island levels,the islandprobably supports about 2% of the world Museum,the shorelineis likely to remainunaltered. Robben 37 Bulletin 96 December2001 38 Wader Study Group Bulletin !: Murray's Robben:••?•ß SandyBay Habour Islan,•• shore r• Langebaan Lagoon Dassen Western Cape Island South Africa Atlantic Robben Bay Ocean Island Cape Town 5O Km Figure 1. Map showinglocation of Robben Island in western Cape. Island is thereforelikely to be valuableas a "remotesite" for RobbenIsland alreadyqualifies as a Wetland of Interna- monitoringpopulation trends of Arctic-breedinglong distant tional Importancein terms of the RamsarConvention be- migrantssuch as Ruddy Turnstones, Curlew Sandpipersand causeit supportsmore than 1% of the global populationof Sanderlings.Such remote sites are localities at which the the African Black Oystercatcher,and of the nominaterace of magnitudeof fluctuationsin breedingsuccess is exaggerated the Swift Tern (Barnes1998). However,the list of bird spe- becausethe shorelineis suboptimalhabitat, occupiedby ciesdefined as "waterbirds"by the RamsarConvention has large numbersof birdsonly in thoseyears in whichoptimal remained static sincethe inception of the Conventionin habitat(such as at LangebaanLagoon) is fully utilized.These 1975. It has not kept pace with the changingdefinition of influxesare likely to be of first-yearbirds. Fluctuation in the wetlands,which hasbeen extended to includeshorelines, and numbersof thesespecies at LangebaanLagoon during the the associatedareas of seaout to a depthof 6 m. To be con- austral summer is small in relation to the variation in breed- sistent with this definition of wetlands, the "waterbird" list ing productivity (Underhill 1987). shouldbe revisedto includespecies that would convention- Other offshoreislands along the southernAfrican coast- ally be classifiedas seabirds.In the caseof RobbenIsland, line couldalso form part of sucha networkof remotemoni- it shouldalso be possibleto motivateits listing as a Ramsar toringsites. There is baselinedata from, for example,Dassen siteon thebasis of itsbreeding colonies of AfricanPenguins Island (Underhill 1992), and several surveyshave subse- and Bank Cormorants.This would be closelyanalogous to quently been made on this island (A.C. Wolfaardt pers. the listingof the shorelineof northernKwaZulu-Natal, South comm.). It would be valuable to do a series of trials over a Africa, for whichthe turtles that come ashore to lay theireggs periodof yearsto establishwhether increases and decreases on thebeaches formed part of the argumentfor listingthe site on theseislands fluctuate synchronously and whether they do (Cowan 1995). indeedreflect fluctuationsin breedingsuccess. Bulletin 96 December 2001 Underhill et aL: Shoreline birds of Robben Island 39 Table 1. Counts of shoreline birds on Robben Island, 1977-2001. The survey covers the shorebirds visible from the perimeter road of the island, and includes the waterbirds within three disused quarries which fill up with rainwater during winter. Date of survey 18 Aug 1977 8 Dec 1980 26 Dec 1998 30 Nov 2000 3 Mar 2001 18 May 2001 12 Aug 2001 Observers G. Arkell, G.D. Underhill, PAW Avian Earthwatch LGU, LGU, C. Marais LGU Demography team KMC KMC H.G. Robertson* Unit Species Little Egret Egretta garzetta 22 29 39 29 26 22 Cattle Egret Bubulcusibis I 13 SacredIbis Threskiornisaethiopicus 57 52 174 129 29 Hadeda Ibis Bosttychiahagedash 2 EgyptianGoose Alopochen aegyptiacus 5 2 3 14 16 South African Shelduck Tadorna cana 2 Yellowbilled Duck Anas undulata 1 5 Moorhen Gallinula chloropus 2 2 2 African Black Oystercatcher Haematopusmoquini 40 89 149 135 143 110 126 RingedPlover Charadriushiaticula 34 17 16 32 White-frontedPlover Ch. marginatus 23 24 2 10 4 2 Kittlitz's Plover Ch. pecuarius 5 11 3 2 2 Grey PloverPluvialis squatarola 17 10 14 3 2 Crowned Plover Vanellus coronams 4 4 4 13 Blacksmith Plover V. armams 16 59 35 42 36 40 TurnstoneArenaria interpres 40 517 235 257 137 18 GreenshankTringa nebularia 1 3 3 1 CommonSandpiper Actitis hypoleucos 2 Curlew SandpiperCalidrisferruginea 339 152 114 151 Knot C. canutus 2 SanderlingC. alba 88 4 3 Whimbrel Numeniusphaeopus 1 38 24 27 14 2 6 SpottedDikkop Burhinuscapensis 19 50 13 42 150 28 Kelp Gull Larus dominicanus 120 139 43 39 336 230 51 Hartlaub's Gull L. hartlaubii 136 305 327 259 187 223 98 Swift Tern Sternabergii 23 95 53 30 300 280 Sandwich Tern S. sandvicensis 232 40 3 Common Tern S. hirundo