Raptors in the East African Tropics and Western Indian Ocean Islands: State of Ecological Knowledge and Conservation Status

Home , Accipiter, African cuckoo, Anjouan scops owl, Bateleur, Bearded vulture, Gyps

j. RaptorRes. 32(1):28-39 ¸ 1998 The Raptor ResearchFoundation, Inc.

RAPTORS IN THE EAST AFRICAN TROPICS AND WESTERN INDIAN OCEAN ISLANDS: STATE OF ECOLOGICAL KNOWLEDGE AND CONSERVATION STATUS

MUNIR VIRANI 1 AND RICHARD T. WATSON ThePeregrine Fund, Inc., 566 WestFlying Hawk Lane, Boise,1D 83709 U.S.A.

ABSTRACT.--Fromour reviewof articlespublished on diurnal and nocturnal birds of prey occurringin Africa and the western Indian Ocean islands,we found most of the information on their breeding biology comesfrom subtropicalsouthern Africa. The number of published papers from the eastAfrican tropics declined after 1980 while those from subtropicalsouthern Africa increased.Based on our KnoM- edge Rating Scale (KRS), only 6.3% of breeding raptorsin the eastAfrican tropicsand 13.6% of the raptorsof the Indian Ocean islandscan be consideredWell Known,while the majority,60.8% in mainland east Africa and 72.7% in the Indian Ocean islands, are rated Unknown. Human-caused habitat alteration resultingfrom overgrazingby livestockand impactsof cultivationare the main threatsfacing raptors in the east African tropics, while clearing of foreststhrough slash-and-burnmethods is most important in the Indian Ocean islands.We describeconservation recommendations, list priorityspecies for study,and list areasof ecologicalunderstanding that need to be improved. I•y WORDS: Conservation;east Africa; ecology; western Indian Ocean;islands; priorities; raptors; research.

Aves rapacesen los tropicos del este de Africa yen islasal oeste del Oc•ano Indico: estado del cono- cimiento eco16gicoy de su conservacitn RESUMEN.--Denuestra recopilacitn de articulospublicados sobre aves rapaces diurnas y nocturnasque se encuentran en Africa yen las islasal oeste del Octano Indico, encontramosque la mayoriade la informaci6n sobre aves rapacesresidentes se origina en la regi6n subtropical del sur de Africa. E1 nfimero de publicacionesprovenientes de los tr6picos del este de Africa, declint despuesde 1980, mientras que aquellos del subtrtpico del sur del continente aumentaron. Con base en nuestra escala de valoraci6ndel conocimientosolo el 6.3% de las especiesresidentes en los tr6picosdel estedel Africa y el 13.6% de lasaves rapaces de islasdel Ocaano Indico puedenser consideradas como Bien Conocidas, mientrasque la mayoria60.8% del estede Africa y 72.2% de las islasal oestedel Octano Indico fueron clasificadascomo Desconocidas.La alteracitn de habitatsde origen antrtpico como resultadodel so- brepastoreoy los impactosde la agricultura son las principalesamenazas para las avesrapaces de los tr6picos del este de Africa, mientrasque la deforestaci6na partir de la quema y tala de bosqueses la amenaza mas i•nportante en las islasdel Octano Indico. Describimosrecomendaciones de conservacitn, identificamoslas especiesprioritarias para estudiare identificamoslas areasde entendimientoecoltgico que deben ser mejoradas. [Traducci6n de C•sar M•rquez]

Raptors are an important tool to focus conser- web below them (Thiollay 1992). Becauseof their vation strategies locally, regionally, and globally top positionsin terrestrial and aquatic food webs, (Watson 1991). They can be used as "umbrella spe- raptorscan be used as indicatorsof worldwidepol- cies" because their large home ranges and low lution by pesticides(Newton 1979). Changesin nesting densities necessitatethat any protected ar- raptor distribution or abundance can serve as a eas encompassingviable populations or complete measure of our impact on landscapes,even in re- communities protect sufficient habitat and popu- mote areas (Reichholf 1974). Finally,raptors are lations of most, if not all, other speciesin the food popular and charismatic and several specieshave becomesignificant "flagships" for increasingpub- lic interest and support of conservationprograms • AlsoNational Museumsof Kenya,Nairobi and Leicester (Burnham et al. 1992, Thiollay 1992). University, U.K. This paper combines our collective knowledge

28 MARCH 1998 EAST AFRICAN AND WESTERN INDIAN OCEAN RAPTORS 29 of raptors in the east African tropics and western 1980 material. Pre-1980 titles were obtained from the b•b- Indian Ocean islandsas part of a worldwide review liography of Brown et al. (1982) and Fry et al. (1988). Post-1980 titles were obtained from a keyword search of their conservation and ecology. We chose to from 1980 through the first quarter of 1996 on the BIOS- combine these two regions in one paper to facili- IS (Biological AbstractsInc.) and RRTAC (Raptor Re- tate comparisonof island with continental situa- search and Technical AssistanceCenter, Boise, ID U.S.A ) tions in adjacent geographicareas. The eastAfri- computer databases.We assumedthat the material obtained representeda significantproportion of all articles can tropics (including all countries between the published in peer-reviewedjournals, and that these Tropicsof Cancerand Capricorn and eastof about sourcesaccurately reflect ecologicalknowledge based on 20øE, Zanzibar and Pemba islands) and the western scientific studies. Ad-hocaccounts published in newslet- Indian Ocean islands (comprising Madagascar, ters or magazinescan provide additional knowledgebut Mauritius, Seychelles,Reunion, and Comores) sup- they were not included in this review as they are difficult to find and their significanceto the understanding of a port roughly23% of the world'savifauna accord- species'ecology is difficult to assess. ing to specieslisted in Martin (1987), Langrand Based on a similar review of the statusand knowledge (1990), Perlo (1995), and Zimmerman et al. of Central and South American raptors (Bierregaard (1996). These areas contain a broad spectrum of 1995), we assigned scores to each species from 1 to 5 from our I4a•owledgeRating Scale (KRS) where I = an- habitats, from montane through lowland, dry, ecdotal, speculativeor unstudied; 2 = 1 nest or 1 year moist, and riparian forests,to open woodland, sa- low samplesize studies; 3 = multi-year,local area studies, vanna, scrubland, desert, marsh, mangroves,and <5 pairs; 4 = multi-year, >5 pairs; and 5 = >10 pairs, others. The eastAfrican tropicssupport 82 species >10 years,or regional studies.This wasrepeated for four of diurnal birds of prey and 23 nocturnal owls subject areas, factors affecting distribution and abundance, population trends, breeding biology, and feeding (Britton 1980, Brown et al. 1982, Fry et al. 1988). ecology. Scoresfor each subject were summed for each Three-quartersof theseraptors, or roughly20% of species (min. 4, max. 20), and speciesranked by score. the world's raptor species,breed within this area Specieswere then grouped into four categoriesof knowl- while the rest use the region as a migratory path- edge based on their scores:Unknown (score 4-6), Little Known (score 7-10), Known (score 11-13), and Well way.The Indian Ocean islandssupport 22 endemic Known (score 14 and above). Although these categories raptor species(15 diurnal birds of prey and seven were subjective,they simplified our review by separating nocturnal owls), five other breeding species,and speciesinto roughly common levels of knowledge. To two Palearcticmigrants. Although both theseareas compare between subject areas, we calculated a "study boast a tremendous diversityof birds that attracts index" (first line of Tables 1 and 2) by stunmingthe KRS for all specieswithin each subject, and converting the the attention of bird watchersworldwide, their rap- total to a percentage of the maximum possibleif all spe- tors are still poorly understood in many respects, cies had KRS of 5. Species' total scoreswere calculated ranging from basicbiology to the factorsaffecting by summing KRS for all speciesin all subject areas and distribution and abundance. Conserving raptors, expressedas a percentageof the maximum possibleif all specieshad KRS of 5. The conservation status of each or using them as a tool to achievebroader conser- species was compiled from two sources (Collar et al vation goals,is impossiblewithout a sound under- 1994, Bennun and Njoroge 1996). standingof their ecologicalrequirements. In this paper, we first give an overview of the R•St:LTS •ND D•SCt:SS•ON state of ecologicalknowledge and conservationsta- State of EcologicalKnowledge and Conservation tus of breeding raptors in the east African tropics Status. We found 251 and 295 articles in our pre- and the western Indian Ocean islands, based on 1980 and post-1980literature searches,respective- 546 published papers and recent listsof speciesin ly. During the pre-1980 era, raptor researchorigi- jeopardy. Second, we outline the main conserva- nated almostequally from countriesin the eastAf- tion issues,contrast threats to raptorson the main- rican tropics (36%) and subtropicalsouthern Af- land with those on islands, summarize information rica (39%), with 14% originating from tropical on causesof threats, and cite particular examples west Africa and 12% from north and central Africa. of problems.Lastly, we discusspriorities for further From 1980 through the first quarter of 1996, the studies and focus of conservation efforts. proportion of publications originating from sub-

METHODS tropical southernAfrica increasedto 55% and decreased to 25% from countries in the east African We reviewedthe literature in peer-reviewedjournals on 105 speciesin Africa, 79 of which breed in tropical east tropics, while they increasedfrom the western In- Africa, and 22 that are endemic to the Indian Ocean dian Ocean islands (10%) and decreased from •slands. The review was divided into pre-1980 and post- tropical west and north Africa (each 5%). The 30 V•RAN• AND WATSON VOL. 32, NO. 1

Table 1. Knowledgerating scores (KRS) for distributionand abundance,population biology, breeding biology, and feedingecology of 79 raptor speciesbreeding in the eastAfrican tropics. Species are listedin order of increasing KRS.KRS scores: I = anecdotal,speculative or unstudied;2 = singlenest and singleyear low samplesize studies; 3 = multi-yearlocal area studies; 4 = multi-yearand >10 pairs;5 = >10 pairs,>10 yearsor regionalstudies. Conser- vanon statusfor speciesin the eastAfrican region (Bennun and Njoroge 1996): EN = Endangered,NT = Near Threatened,VU = Vulnerable, RR = RegionalResponsibility. Global conservationstatus categories (Collar et al. 1994): EN = Endangered,NT = Near Threatened,VU = Vulnerable.

DIS- Pop- BREED- FEED-

TRIB. ULAT. ING ING AND BIO- BIOL- ECOL- SPECIES ABUND. LOGY OGY OGY TOTAL

PROPORTION OF STUDIES CONSERVATION IN EACH SUBJECTAREA STATUS

COMMON NAME SCIENTIFIC NAME 11% 9% 26% 20% 13% REGIONAL GLOBAL

Buzzard, Mountain Buteooreophilus 1 1 1 1 4 NT Buzzard, Red-necked Buteoauguralis 1 1 1 1 4 Eagle, Congo Serpent Dryotriorchisspectabilis 1 1 1 1 4 Snake-eagle,Southern Circaetusfasciolatus 1 1 1 1 4 NT NT Banded Falcon, Pygmy Polihieraxsemitorquatus 1 1 1 1 4 Hawk-eagle,Ayre's Hieraaetusayresii 1 1 1 1 4 VU Hawk-eagle,Cassin's Spizaetusafricanus 1 1 1 1 4 Hawk, Long-tailed Urotriorchis macrourus 1 1 1 1 4 Kestrel, Fox Falcoalopex 1 1 1 1 4 Kestrel, Grey Falco ardosiaceus 1 I I 1 4 Kite, African Swallow- Elanus ( Chelictinia)riocourii I 1 1 1 4 tailed Kite, Black Milvus migrans 1 1 1 1 4 Owlet, Albertine Glaucidium albertinure 1 1 1 1 4 VU-RR VU Owlet, Chestnut Barred Glaucidiumcapense caste- 1 1 1 1 4 neum Owlet, Etchecopar's Glaucidiumcapense etcheco- 1 1 1 1 4 pari Owlet, Red-chested Glaucidiumtephronotum 1 1 1 1 4 VU Owl, Fraser's Eagle Bubopoensis 1 1 1 1 4 Owl, Congo-bay Phodilusprigoginei 1 1 1 1 4 VU Owl, Pemba Scops- Otuspembaensis 1 1 1 1 4 VU-RR NT Owl, Usambara Bubo vosseleri 1 1 I I 4 VU-RR VU Secretary-bird Sagittariusserpentarius 1 I I I 4 Sparrowhawk,Black Accipitermelanoleucus 1 I I I 4 Buzzard, Augur Buteoaugur 2 I I I 5 Buzzard, Lizard Kaupifalcomonogrammicus 1 I 2 I 5 Snake-eagle,Brown Circaetus cinereus 1 I 2 I 5 NT Eagle, Western Banded Circaetus cinerascens 1 I 2 I 5 VU Hawk, African Cuckoo Aviceda cuculoides 1 I 2 I 5 NT Falcon, Taita Falcofasciinucha 1 1 2 1 5 VU VU Goshawk, Gabar Micronisus( Melierax) gabar 1 I 2 I 5 Goshawk, Little Banded Accipiterbadius I I 2 I 5 (Shikra) Kestrel, Dickinson's Falco dickinsoni 1 I 2 I 5 Owl, AbyssinianLong- Asioabyssinicus 1 I I 2 5 eared Owl, Pearl-spotted Glaucidiumperlatum 1 1 1 2 5 Owl, Pel's Fishing Scotopeliapeli I 1 2 1 5 VU MA•CH 1998 EAST AFRICAN AND WESTERN INDIAN OCEAN RAPTORS 31

Table 1. Continued.

DIs- PoP- BREED- FEED-

TRIB. ULAT. INC ING AND BIO- BIOL- ECOL- SPECIES ABUND. LOGY OGY OGY TOTAL

PROPORTION OF STUDIES CONSERVATION IN EACH SUBJECTAREA STATUS

COMMON NAME SCIENTIFIC NAME 11% 9% 26% 20% 13% REGIONAL GLOBAL

Owl, SokokeScops- Otus ireneae 1 2 1 1 5 VU-RR Owl, SpottedEagle- Buboafricanus 1 1 1 2 5 Owl, African Wood- Ciccabawoodfordii 1 1 2 1 5 Sparrowhawk,Little Accipiterminullus 1 1 2 1 5 Vulture, Palm-nut Gypohieraxangolensis 1 1 1 2 5 Vulture, White-headed Trigonocepsoccipitalis 1 1 2 1 5 VU Snake-eagle,Black- Circaetuspectoralis 2 1 2 1 6 chested Hobby, African Falco cuvieri 1 1 2 2 6 Hawk, Bat Macheiramphusalcinus 1 1 2 2 6 NT Owl, Cape Eagle- Bubocapensis 2 1 1 2 6 VU Owl, White-facedScops- Otus leucotis 1 1 2 2 6 Sparrowhawk,Ovambo Accipiterovampensis 1 1 3 1 6 NT Vulture, Egyptian Neophronpercnopterus 2 2 1 1 6 NT Vulture, African White- Gypsafricanus 1 1 2 2 6 NT backed Hawk-eagle, African Hieraaetusspilogaster 1 1 4 l 7 Eagle, Long-crested Lophaetusoccipitalis 2 1 2 2 7 Owl, Barred Glaucidiumcapense 4 1 1 1 7 Vulture, Hooded Necrosyrtesmonachus 1 1 4 1 7 Vulture, Rfippell's Griffon Gypsrueppellii 4 1 1 1 7 NT Falcon, Peregrine Falcoperegrinus 1 1 3 3 8 Goshawk,Dark Chanting- Melierax metabates 1 I 3 3 8 Goshawk,Pale Chanting- Melierax canorus 1 1 3 3 8 Harrier-hawk, African Polyboroidestypus 1 1 3 3 8 Falcon, Lanner Falco biarmicus 1 2 3 3 9 Owl, Giant Eagle Bubo lacteus 4 1 2 2 9 Falcon, Red-necked Falcochicquera 1 3 3 3 10 NT Goshawk, African Accipitertachiro 1 3 3 3 10 Kestrel, Rock (Common) Falco tinnunculus 2 2 3 3 10 Kite, Black-winged Elanus caeruleus 1 1 4 4 10 Owl, Barn Tyroalba affinus 1 1 4 4 10 Owl, African Grass Tyrocapensis 1 1 4 4 10 VU Owl, ScopsEurasian- Otusscops 1 1 4 4 10 Eagle, African Crowned Stephanoaetuscoronatus 3 3 3 3 12 VU Eagle, Tawny Aquila rapax 3 3 3 3 12 Eagle, Wahlberg's Aquila wahlbergi 2 2 4 4 12 Harrier, African Marsh Circus ranivorus 1 3 4 4 12 NT Sparrowhawk,Red- Accipitervufiventris 3 3 3 3 12 NT breasted Eagle, Martial Polemaetus bellicosus 3 3 3 4 13 VU Kestrel, Greater Falcovupicoloides 3 3 4 3 13 NT Owl, African Marsh Asiocapensis 4 1 4 4 13 NT Vulture, Lappet-faced Torgostracheliotus 4 4 3 3 14 NT Bateleur Terathopiusecaudatus 4 4 4 4 16 Vulture, Bearded Gypaetusbarbatus 4 4 4 4 16 EN 32 VIRANI AND WATSON VOL. 32, No. 1

Table 1. Continued.

DIs- PoP- BREED- FEED-

TRIB. ULAT. ING ING AND BIO- BIOL- ECOL- SPECIES ABUND. LOGY OGY OGY TOTAL

PROPORTION OF STUDIES CONSERVATION IN EACH SUBJECTAREA STATUS

COMMON NAME SCIENTIFIC NAME 11% 9% 26% 20% 13% REGIONALGLOBAL

Eagle,African Fish- Haliaeetusvodfer 4 4 5 4 17 Eagle, Black (Verreaux's) Aquila verreauxii 4 4 5 5 18

Table 2. Knowledgerating scores(KRS) for distribution and abundance,population biology,breeding biology and feeding ecologyof 22 endemicraptors of the westernIndian Ocean islands.Species are listedin order of increasing KRS. KRS scores:1 = anecdotal,speculative or unstudied;2 = singlenest and singleyear low samplesize studies; 3 = multi-yearlocal area studies;4 = multi-yearand >10 pairs;5 = >10 pairs, >10 yearsor regional studies.Global conservationstatus (Collar et al. 1994): CR = Critical; EN = Endangered;VU = Vulnerable;NT = Near-threatened

DISTRIB. AND POPULAT. BREEDING FEEDING SPECIES CONSER- ABUND. BIOLOGY BIOLOGY BIOLOGY TOTAL VATION PROPORTIONOF STUDIESIN EACH SUBJECTAREA STATUS

COMMON NAME SCIENTIFIC NAME 18% 10% 16% 6% 15% GLOBAL

Falcon, MadagascarCuck- oo- Avicedamadagascariensis 1 1 I 1 4 Goshawk, Henst's Accipiterhenstii I I I I 4 NT Harrier-hawk, Madagascar Polyboroidesradiatus I I I I 4 Harrier, R•union Circus maillardi 1 1 1 1 4 NT Kestrel, Aidabra Falco newtoni aldabranus 1 1 I 1 4 Kestrel, Madagascar Falco newtoni 1 1 I I 4 Owl, Grand Comoro Scops- Otuspauliani I I I I 4 CR Owl, MalagasyScops- Otus rutilus I I I I 4 Owl, SeychellesScops- Otus insularis I I I 1 4 CR Owl, White-browed Ninox superciliaris I I 1 I 4 Sparrowhawk,Anjouan Accipiterfrancesii pusillus I 1 1 1 4 Sparrowhawk,Frances's Accipiterfrancesii 1 1 1 i 4 Sparrowhawk,Madagascar Accipitermadagascariensis I I I I 4 NT Serpent-eagle,Madagascar Eutriorchis astur 2 1 I I 5 CR Owl, Anjouan Scops Otusrutilus capnodes 2 1 1 I 5 CR Owl, MadagascarLong- eared Asiomadagascariensis I 1 1 2 5 Kestrel, Banded Falco zonii•entris 2 I 2 2 7 Owl, MadagascarRed Tytosoumagnei 2 I 2 2 7 EN Buzzard, Madagascar Buteobrachypterus 3 I 4 4 12 Kestrel, Seychelles Falco araea 4 4 4 4 16 VU Fish-eagle,Madagascar Haliaeetusvociferoides 5 4 4 4 17 CR Kestrel, Mauritius Falcopunctatus 5 5 5 4 19 EN MARCH 1998 EAST AFRIGAN AND WESTERN INDIAN OCEAN RAPTORS 33 post-1980 shift in the proportion of studies from Almost 40% of all the breeding raptorsin Africa eastAfrica to southern Africa may be partly due to have been placed in a recently compiled east Af- the death in 1980 of Leslie Brown who contributed rican list of birds in jeopardy (Table 1, Bennun and almost a quarter of all pre-1980 articles from east Njoroge 1996). This regional list, comprisingspe- Africa. cies found in Kenya, Uganda, Tanzania, Rwanda, Of the tbur subjectareas reviewed, breeding bi- and Burundi, was compiled to give more detailed ology (26%) was best studied in the African rap- resolution of conservationpriorities in these coun- tors, followed in descendingorder by feeding ecol- tries and to set a regional agenda for research, ogy, factors affecting distribution and abundance, monitoring, and conservation.Of the 30 breeding and population biology (Table 1). Indian Ocean raptors listed in jeopardy in the east African re- •sland species were mostly less well-studied than gion, one speciesis listed as Endangered, 14 asVul- continental species,but scored higher in factors nerable, and 15 as Near Threatened (Table 1). In affecting distribution and abundance (18%), fol- contrast,only sevenof the 79 breedingAfrican spe- lowed in descendingorder by breeding biology, cies were listed as globally Threatened (Table 1, population biology,and feeding ecology(Table 2). Collar et al. 1994). This substantial diflbrence is There was little difference in species'total scores because many raptors have continental ranges between continental species (13%, Table 1) and which make their global (but not necessarilyre- Indian Ocean island species(15%, Table 2), both gional) status more secure. There may also be a of which indicate our knowledgeis substantiallyin- regional awarenessthat raptor populations are de- complete. Breeding biologymay be the most often clining becauseonce-common species are not seen studied aspect of raptor ecology on mainland Af- as frequently as they were within memory, but data rica because there are many large species,often are lacking. It may also reflect an inherent prob- inhabiting open landscapes,that build large con- lem of categorizing a species'status based on stan- spicuousnests which are used year after year, thus dard criteria for which data are almost invariably facilitating observationson nesting behavior. The lacking, especiallyon a global scale. relatively small sizes of the Indian Ocean islands Speciesconsidered globally Vulnerable or Near perhaps enable a better and more accuratesurvey Threatened that rated as Unknown in our KRS of distribution and abundance of island raptors. and, therefore, should be targeted for study in- The fact that many Indian Ocean island speciesare cluded Southern Banded Snake-eagle(Circaetusfas- Threatened may also have contributed to an inter- ciolatus),Albertine Owlet ( Glaucidiumalbertinum), est in distribution and abundance studies. Popu- Congo Bay-owl (Phodilusprigoginei), Pemba Scops- lation trends and demographic parameters were owl (Otus pembaensis),Usambara Eagle-owl(Bubo least studied in both regions (Tables 1 and 2) pos- vosselen),and Taita Falcon (Falcofasciinucha) (Ta- sibly being the most difficult kind of studies to ble 1). The Sokoke Scops-owl(Otus ireneae),previ- complete. Monitoring and understanding popula- ously listed as Endangered (Collar and Stuart tion changesis probably the most important kind 1985) but now listed as Vulnerable (Collar et al. of knowledge needed for speciesthat are in jeop- 1994, Bennun and Njoroge 1996), was rated Un- ardy. known, but a recent studyhas increasedour knowl- Of 79 speciesthat breed in the eastAfrican trop- edge of this bird to at least Known in all four sub- ics, five (6.3%) scored a total KRS > 14 and could jects (Virani 1995a). be said to be Well Known, eight (10.1%) scored Of 22 endemic Indian Ocean island species,only from 11-13 and might be considered Known, 18 three (14%) rated as Well Known (Table 2). The (22.8%) were rated Unknown in at least two sub- majority of species (83.6%) were rated Little ject areas, and may be considered Little Known, Known or Unknown (SpeciesTotal --<10,Table 2). while the majority 48 (60.8%) were rated Un- The Endangered Mauritius Kestrel (Falcopuncta- known in all four subject areas (Table 1). Well tus) and the Vulnerable SeychellesKestrel (Falco Known specieswere large and charismatic, or lo- araea) were considered Well Known. The Endan- cally threatened such as the Lappet-facedVulture gered Madagascar Fish-eagle (Haliaeetus vocifero- ( Torgostracheliotus), Bearded Vulture ( Gypaetusbar- ides) was rated Known but recent studies have batus), Bateleur (Terathopiusecaudatus), African raised our knowledge level to Well Known (Watson Fish-eagle(Haliaeetus vocifer), and Black Eagle (Aq- et al. 1996, Rafanomezantsoa 1997, Watson 1997, uila verreauxii) (Table 1). Watsonet al. 1997). Half of the endemic raptors 34 VIRANI AND WATSON VOL. 32, NO. 1 of the Indian Ocean islands were listed at some secutionthrough secondarypoisoning as in Kenya level of global jeopardy (Collar et al. 1994). Of (Thomsett pers. comm.). these,five were listed as CriticallyEndangered, two Threats to raptors on the Indian Ocean islands as Endangered, one as Vulnerable, and three as are similarly related to increasing human popula- Near Threatened (Table 2). tions but they tend to be more extreme becauseof Among the Unknown speciesthat should be tar- the limited area of islands,limited potential for dis- geted for studywere the EndangeredGrand Com- persion of both humans and raptors, and high de- oro Scops-owl(Otus pauliani), SeychellesScops-owl gree of endemism among island raptor species. ( Otusinsularis), Anjouan Scops-owl(Otus capnodes), Threats to raptors depend on the human coloni- and Anjouan Sparrowhawk (Accipiterfrancesiipusil- zation history,the faunal community,and the set lus) (Table 2). Two notable recent rediscoveries, of conservationproblems that are unique to each the Madagascar Serpent-eagle (Eutriorchisastur) island. For these reasons we will describe the con- and MadagascarRed Owl (Tyto soumagnei),have servationissues in some detail for each island sep- changed the species'status from possiblyextinct to arately. extant (Halleax and Goodman 1994, Thorstrom et Madagascar,the largest of the Indian Ocean is- al. 1995); they were listed as CriticallyEndangered lands (587 000 km2), supports22 speciesof diurnal and Endangered,respectively (Collar et al. 1994). and nocturnalbirds of prey.All excepttwo of these Studies on the MadagascarRed Owl in 1994-95 breed on Madagascar.Eleven species,10 of which provided first data on breeding and diet (Thor- are found in forest habitats and one in wetlands strom et al. 1997, Thorstrom and Renfi de Roland and coastal areas, are endemic to the island. Sub- 1997) so that the speciescould be rated as Known fossil remains indicate that two Aquila speciesex- •n two subject areas. isted on Madagascar during the Quaternary Threats to Raptors. The main threatsfacing rap- (Goodman and Rakotozafy 1995) and a Stephan- tors in the eastAfrican tropics can be summarized oaetusspecies has been identified from the Holo- as rapidly changing land-use patterns resulting cene (Goodman 1994), but neither genusnow sur- from an increasein human population size (Sorley viveson Madagascar.The island was relatively re- and Andersen 1994). The precise impacts of cently colonized by humans who are known to changingland useson breeding and migrant rap- have been present since 1890 + 90 B.P. (MacPhee tors are speciesspecific. In general,anthropogenic and Burney 1991). With a current human popu- habitat alteration can influence raptor distribution lation of about 14 million, a population growth and abundance through direct changesin habitat rate of 3.1% per year and per capita income characteristics,such as availabilityof suitable nest- around $230 per year (World Bank 1992), the ma- ing and perching sites (Enderson 1964, Marion and Ryder 1975, Stahlecker1978, Janes 1984), or jor threats to raptors, their habitats, and most oth- •ndirectlythrough changesin prey abundanceand er biota are driven by the subsistenceneeds of the availability (Baker and Brooks 1981). In eastAfrica, human population. The greatest single threat to we believe that four main factorsaffect raptor den- extant endemic raptors is from human induced habitat modification that renders forest or wetland sity and diversity as a result of changes in habitat that accompany human land-use pressures: (1) habitat unsuitable for them. Clearing of eastern overgrazingby domesticlivestock resulting in the rainforest by slash-and-burnagriculture was esti- alteration of original vegetativecover that affects mated to occur at an average rate of 111 000 ha/ prey abundance and distribution (Sorley and An- year from 1950-85 when 3.8 million ha (34% of derson 1994); (2) impact of cultivation that com- original forest cover) remained (Green and Suss- pletely replacesnative vegetationand involvesthe man 1990). Most of the western dry deciduousfor- use of pesticidesand fertilizerspotentially toxic to estshave been replaced by relativelysterile savan- raptors and other vertebrates (Sorley and Ander- na, but deforestation continues in the west when son 1994); (3) human hunting pressure that di- deliberate grasslandfires escapeinto forest areas. minishes prey populations (Sorley and Anderson Lossof wetlandsdue to drainage and modification 1994); and (4) direct human persecution (Keran for production of rice continues and protection of 1981) with resulting changesin behavior (Knight remaining wetlands is unpopular because of the et al. 1989), species abundance and diversity need to put new areasinto production. Only one (Craighead and Mindell 1981), and indirect per- significantwetland existswithin a protected area ]V[ARCH 1998 EAST AFRICAN AND WESTERN INDIAN OCEAN RAPTORS 35

(Lake Tsiinanainpetsotsa,Langrand and Goodman ging outside the protected area. Monitoring of 1995). sinall island populations is clearly desirable be- Habitat conservationis the single most impor- causeof their susceptibilityto extinction froin ran- tant priority in Madagascar,with greater emphasis dom catastrophicevents as well as human induced on wetlands than has been achieved to date. Man- change. agement of wild populations of certain species,like Four volcanic islands fbrin the Coinoro archi- the MadagascarFish-eagle (Watson 1997), mayalso pelago (Ngazidja or Grand Colnoro, Ndzuani or be needed in the near future, and captive breeding Anjouan, Mwali or Moheli, and Maore or Mayotte). may be required not long after that. In the long Each island supports its own, mainly forest dwell- term, however, sustainableagriculture and agro- ing avifauna (Louette and Stevens1992). Humans forestry to provide local inhabitants with needed arrived perhaps 1000 years ago and deforestation food, fuel and fiber, accompanied by a reduction on lowland coastalareas is complete. Most endeln- of population growth,are among the prerequisites ic speciesremain only in higher altitude forest and for effective conservation. Addressing social and tree-heath on top of Mt. Karthala on Ngazidja and economic needs of local peoplesare needed if con- on Mwali. Little to no forest remains, even at high servation is to succeed. For example, ways to en- altitudes, on Ndzuani and Maore. The only re- hance local traditional rules and taboos are being inaining endemic raptor speciesare the Grand Co- explored to achieveconservation of habitatsin the inores Scops-owl (Karthala Scops-owl) and An- absenceof protected areas (Razandrizanakanirina jouan Scops-owl.The latter was rediscoveredon and Watson 1997). Ndzuani in 1992 having not been recorded with Besides habitat loss, raptors are persecuted, es- certainty since about 1886 (Safford 1993). It is pecially in the west where we know nestlings of found only in a few remaining patches of upland MadagascarFish-eagle have been taken for food or forest (above 800 in) where the population is es- as pets and adults have been killed. Given the low tilnated as probably 100-200 pairs. The speciesis population size and low productivityof this species endangered by forest clearance and capture for (Watsonet al. 1997), the lossof just a fewnestlings food by people. Safford (1993) recolnlnended an and adults each year has a significant impact on investigationof the feasibilityand consequencesof the species'survival (Watson 1997). Much perse- the species'introduction to neighboring island of cution occursout of ignorance about the value and Mwali, as this may be the species' only hope for rarity of speciesso increased public awarenessis survival. The Grand Coinores Scops-owl is con- urgent. fined to forest (virgin through degraded and sinall The Seychellescomprise three large islands,the patches) between 1000-1900 in on Mt. Karthala on largestof which is Mah6 (145 kln9), and numerous Ngazidja (Herrelnans et al. 1991). A 1989 study es- others. It supportstwo endemic raptors, the forest tilnated that there was perhaps more than 1000 dwelling SeychellesKestrel with a world population pairs on the island assuming10 000 ha of suitable of about 420 pairs on Maht, Silhouette and habitat and a territory size of about 5 ha for each through reintroduction on Praslin, as well as some owl pair. Although encouraging, the data are ten- satellite islands (Watson 1989), and Seychelles uous and there remains the long-term threat of Scops-owlwith a population of perhaps80 pairsall habitat loss through forest fraglnentation froin in highland forestson Maht. Threats to the kestrel fires, logging, and clearing for pasture. The spread were loss of suitable habitat on Praslin in the 19th of the introduced Indian Myna (Acridotherestrist•s) and early 20th centuries, and human persecution. as forest is opened is also a threat to endemic spe- Public awareness,enforcement of bird protection cies froin coinpetition and other interspecific in- laws, and banning of private ownershipof firearms teractions. has inhibited illegal shooting (Watson1981), while Various subspeciesof Frances'sSparrowhawk (A. habitat recoveryon Praslinhas allowedthe reintro- francesii) are found on the Mascarene islands. A. duction of the speciesto this island. Much of the francesiifrancesii occurs on Madagascar,but on Co- present day highland forest habitat (secondary inores it is represented by three subspecies,none growth) of the SeychellesScops-owl is incorporated of which are present on Mwali. A. francesiigwiveau& in the Morne SeychelloisNational Park. A better is found on Ngazidja but is not colnlnon (del Hoyo understanding of the species'food and nesting re- et al. 1994), A. francesiipusillus on Ndzuani was quireInents is needed to assessthe effects of log- once colnlnon but is now probably close to extinc- 36 V• ^•D W^TSO• Vote. 32, No. 1 tion due to hunting (Louette and Stevens1992), ritius, but was somewhat constrained by the very and A. francesiibrutus on Maore is common on the rugged topography.A proportionately larger area moist west side (del Hoyo et al. 1994). Rainforest of native forest remains uncut, but nearly all of it protection wherever it occurs is the highest con- is over 500 m and most over 1000 m elevation (Di- servationpriority on the Comores.Remaining for- amond 1987). Currently 35% of the land surface est is modified by firewood collecting, timber ex- is forest and woodland, 26% agricultural,and the traction, understory clearing for vegetablesand remaining is listed as "other." The only raptor crops, and cattle grazing. The creation of a forest known to breed is the R6union Harrier (Circus reserve on Mt. Karthala would protect habitat for maillardi)that is alsofound on Madagascarand Co- four endemic birds (Louette and Stevens 1992) mores. Today, the speciespersists in exotic vege- and similar habitat protection may be feasible on tation. It suffered a severepost-war decline but has Mwali where the human population is still low. increasedsince banning DDT. Introduction of spe- There is little hope for establishinga reserve on cial protection in 1966 to reduce persecution Ndzuani due to population pressure. Control of seemsineffective (Thiollay pers. comm.). There is hunting which may affect raptors, control of ani- compelling evidence of a yet undescribedScops mal introduction, and public education are also Owl in the forests (Bretagnolle and Atti6 1996) high priorities (Louette and Stevens1992), as are that needs investigation. investigationsinto introduction of the Anjouan Conservation and Study Recommendations. Scops-owlto Moheli and protection of specificnest Conservation needs of island fauna and flora are sites on Ndzuani. generally more urgent than continental species, The island of Mauritius (1843 km 2) was covered except where limited distributions on continents with native forest when Europeans first began to mimic the island situation (e.g., SokokeScops-owl, settle on it near the end of the 17th century Virani 1995a, 1995b). Conservationneeds for rap- (Vaughanand Wiehe 1937), but destructionfor tors on the Indian Ocean islands are variable de- lumber, firewood, and agricultural land was mas- pending on the island,but they have the following sive.Habitat loss,hunting, and introduced species priorities in common: conservationof native hab- probably contributed to the demise of the Mauri- itats, prevention of exotic speciesintroduction and tian Owl (Scopscommersoni) along with the Dodo control of existingintroduced exotics,and preven- (Raphuscucullatus) and other fauna during the tion of direct exploitation, persecution,and inter- 18th century (Diamond 1987). Today only small, ference. Conservationneeds for raptorsin the east highly degradedvestiges of forest remain on rug- African tropics include protection of habitat es- ged uplands, principally Black River Gorges pecially where species'distributions are limited to (Cheke 1987, Jones and Owadally 1988). Loss of small habitat patches, such as the Sokoke Scops- native forest, use of organochlorine pesticides,and owl, control of environmental contaminants, and predation by introduced speciesundoubtedly con- control of persecution through public education tributed to the decline of the Mauritius Kestrel and awareness.In particular; becauseraptors typi- which, in 1974, wasreduced to just four wild birds cally need very large rangesin which to forage and confined to about 4000 ha of native forest (Cade survive, current national parks may not be suffi- and Jones 1993). Captive breeding and manage- cient to maintain viable raptor populations. It is ment in the wild has returned the population size important, therefore, to develop a "living with to 56-68 breeding pairs and 222-286 individuals wildlife" ethic among people that will allow coex- (end of 1993-94 breeding season)and there is rea- istence of raptors on land shared with people. son to hope that the population may rise to 200- Where conservationorganizations and individual 250 pairs occupyingdegraded or even exotic for- expertise are absentor poorly developed,such as est-likehabitats (Jones et al. 1994). in Madagascar and Comores, we recommend de- R6union is a volcanic island of 2510 km '• that lies velopment of local capacity through training at 800 km east of Madagascar and 200 km west of technician through doctoral level, technicaladvis- Mauritius. Colonized and still run as a Department ing, and provision of material and financial re- by the French, some endemic fauna are known to sources (Watson and Lewis 1994). have become extinct from hunting and the effects For raptors rated as Little Known and at least of introduced species.The history of human de- Threatened (Tables 1 and 2), priority should be struction of nativevegetation parallels that of Mau- given to studiesthat collect the information need- MARCH 1998 EAST AFRICAN AND WESTERN INDIAN OCEAN RAPTORS 37 ed to determine the best course of action to re- 23, Centre for Biodiversity,National Museums of Ke- duce the probability of the species' extinction. nya, Nairobi, Kenya. Such studies include, but are not limited to, the BERKELMAN,J. 1997. Habitat requirementsand fbragmg following: (1) measuring the species'population ecologyof the MadagascarFish-eagle. Ph.D. dissertation, Virginia Polytechnic Institute and State Univ, size and distribution (e.g., SokokeScops-owl [Vir- Blacksburg,VA U.S.A. ani 1995a, 1995b]) to include genetic distribution BIERREGAARD,R.O., JR. 1995. The biology and conser- if subpopulationsexist that are small and disjunct; vation status of Central and South American Falcon- (2) understandingthe species'nesting and forag- ifortnes:a surveyof current knowledge.Bird Conserv ing habitat needs (e.g., Madagascar Fish-eagle Int. 5:325-340. [Berkelman1997] and MadagascarRed Owl); (3) BRITTON, P.L. [ED.]. 1980. Birds of east Africa, their hab- monitoring population trends, productivity,surviv- itat, status and distribution. East Africa Natural H•s- al to maturity and adult survival (e.g., Madagascar tory Society,Nairobi, Kenya. BRETAGNOLLE, V. AND C. ATTIla. 1996. Comments on a Fish-eagle[Watson et al. 1997]); (4) understanding possiblenew speciesof ScopsOwl Otussp. on R6un- breeding and other behaviorsthat may affect the ion. Bull. A.B.C. 3(1):36. species'ability to recover from low population size BROWN, L.H., E.K. URBAN AND K. NEWMAN. 1982. The (e.g.,Madagascar Fish-eagle [Watson et al. 1997]); birds of Africa. Vol. I. Academic Press, London, U K (5) understandingmovements including migration BURNHAM,W.A., D.F. WHITACREAND J.P. JENNY. 1992. and natal dispersal (e.g., MadagascarFish-eagle The Maya Project: use of raptors as tools for conser- [Rafanomezantsoa1997]); and (6) understanding vation and ecological monitoring of biological diver- the effects of human persecutionand other an- sity.Pages 257-264 in B.-U. Meyburgand R.D. Chan- thropogenic effects (e.g., Madagascar Fish-eagle cellor [EDS.], Raptor conservation today. World Work- [Watson 1997]. In addition to study of Little ing Group on Birds of Prey, Berlin, Germany. CADE,T.J. AND CJ. JONES.1993. Progressin restoration Known Threatened species,in a few casesthat are of the Mauritius Kestrel. Conserv. Biol. 7:169-175. Known, such as the MadagascarFish-eagle, Grand CHEKE,A.S. 1987. An ecologicalhistory of the Mascare- ComoresScops-owl, or Anjouan Scops-owl,we en- ne Islands, with particular reference to extinctions dorseurgent protection of key breedingsites, and, and introductions of land vertebrates.Pages 5-89 •n as needed, preliminary use of wild population A.W. Diamond [ED.], Studies of Mascarene Island management to increase the population size and birds. Cambridge Univ. Press,Cambridge, U.K. reduce the risk of extinction (Louette and Stevens COLLAr, NJ. AND S.N. STUART. 1985. Threatened birds 1992, Safford 1993, Watson et al. 1996, Watson of Africa and related islands: the ICBP/IUCN Red Data Book. International Council fbr Bird Preserva- 1997). Weftands and the water-forest ecotone in tion and International Union for Conservation of Na- Madagascarneed focused conservationefforts if ture and Natural Resources,Cambridge, U.K. they are to survive(Watson 1997). Effortsto pro- COLLAR,NJ., MJ. CROSBYAND AJ. STATTERSFIELD.1994. tect significantblocks of remaining rainforest in Birds to Watch 2. The world list of threatened birds Madagascar(e.g., Projet Masoala 1995, which used Birdlife Conserv. Ser. No. 4. BirdLife International, the area estimatedfor a viable population of Mad- Cambridge, U.K. agascarSerpent-eagles to help justify the size of CRAIGttEAD,F.C., JR. AND D.P. MINDELL. 1981. Nesting protected area) are important for the survivalof raptorsin westernWyoming, 1947 and 1975.J. Wzldl. the MadagascarSerpent-eagle, Madagascar Red Manage.45:865-872. Owl, and other forest speciesand should be en- DELHOYO, J., A. ELLIOTAND j. SARC;ATAL[EDS.]. 1994. Handbook of birds of the world. Vol. 2. New World couragedand duplicated wherever possible. vultures to guineafowl. Lynx Edicions, Barcelona, ACKNOWLEDGMENTS Spain. DIAMOND, A.W. [ED.]. 1987. Studies of Mascarene Island We thank Alan Kemp, Leon Bennun and Andrew Jen- birds. Cambridge Univ. Press,Cambridge, U.K. kins for their constructivereview of this paper. ENDERSON,J.H. 1964. A study of the Prairie Falcon in LITERATURE CITED the central Rocky Mountain Region. Auk 81:332-352 FRY, C.H., S. KEITH AND E.K. URBAN [EDS.]. 1988. The BAKER,J.A. ANDRJ. BROOKS.1981. Distributionpatterns birds of Africa. Vol. III. Academic Press, London, U K. of raptorsin relation to densityof meadowvoles. Con- GREEN, G.M. AND R.W. SUSSMAN. 1990. Deforestation h•s- dor 83:42-47. tory of the eastern rainforestsof Madagascarfrom sat- BENNUN, L.A. AND P. NJOROGE. 1996. Birds to watch in ellite images.Science 248:212-215. eastAfrica: a preliminary Red Data list. ResearchRep. GOODMAN,S.M. 1994. Description of a new speciesof 38 VIP,ANI ANt) W^TSON VOL. 32, NO. 1

subfossileagle from Madagascar:Stephanoaetus (Aves: PERLO,B.V. 1995. Birds of eastern Akica. Harper Col- Falconiformes) from the depositsof Ampasambazim- lins, London, U.K. ba. Proc. Biol. Soc. Wash. 107:421-428. PROJETMASOALA. 1995. Proposition des limites du Parc • AND L.M.A. RAKOTOZAFY. 1995. Evidence for the National Masoala. CARE International Madagascar, existenceof two speciesof Aquilaon Madagascardur- Wildlife Conservation Society,The Peregrine Fund, ing the Quaternary. Geobios28:241-246. Antananarivo, Madagascar. HALLEAX,D. ANDS. GOODMAN.1994. Rediscoveryof the RAFANOMEZANTSOA,S. 1997. Behaviorand nataldispersal Madagascar Red Owl Tyto soumagnei(Grandidier of fledgling MadagascarFish-eagles. Pages 403-412 in 1878). Bird Conserr.Int. 4:305-311. R.T. Watson [ED.], Madagascarwetlands conservation HERREMANS,M., M. LOUETTEANDJ. STEVENS.1991. Con- project. Progress report III, 1995-1996. The Pere- servationstatus and vocal and morphologicaldescrip- grine Fund, Boise, ID U.S.A. tion of the Grand Comoro ScopsOwl Otuspauliani RAZANDRIZANAKANIRINA,D. AND R.T. WATSON. 1997. The Benson 1960. Bird Conserr. Int. 1:12.5-133. processof developing the wetland community-based JANES,S.W. 1984. Influences of territory composition conservationproject: 1997-1998. Pages 33-44 in R.T. and interspecificcompetition on Red-tailedHawk re- Watson [ED.], Madagascarwetlands conservation proj- productivesuccess. Ecolo• 65:862-870. ect. Progressreport III, 1995-1996. The Peregrine JONES,C.G. ANDA.W. OWADALLY.1988. The life histories Fund, Boise, ID U.S.A. and conservationof the Mauritius KestrelFalco punc- REICHHOLF,J. 1974. Artenreichtum,Haufigkeit und Div- tatus (Temminck 1823), Pink Pigeon Columbamayeri ersitat derGreifvogel in einigen Gebieten yon Suda- (Prevost 1843) and Echo Parakeet Psittaculaeques merika.J. Ornithol.115:381-397. (Boddart 1783). Proc.R. Soc.Arts Sci. Mauritius 5:79- SAFFOgD,RJ. 1993. Rediscovery,taxonomy and conser- 130. vation of the Anjouan ScopsOwl Otuscapnodes (Gur- , W. HECK, R.E. LEWIS, Y. MUNGROO, g. SLADEAND hey 1889). Bird Conserr.Int. 3:57-74. T. CADE. 1994. The restoration of the Mauritius Kes- SORLEY,C.S. AND D.E. ANDERSEN.1994. Raptor abundance in southcentral Kenya in relation to land-use trel Falcopunctatus population. Ibis 137:S173-S180. patterns. Aft.J. Ecol.32:30-38. KER_AN,D. 1981. The incidence of man-caused and nat- STAHLECKER,D.W. 1978. Effect of a new transmission ural mortalitiesto raptors. RaptorRes. 15:108-112. line on wintering birds of prey. Condor80:444-446. KNItlit, R.L., D.E. ANDERSON,M.J. BECHARDAND N.V. THIOLLAY,J-M. 1992. A world review of tropical forest MARR.1989. Geographicvariation in nest-defencebe- raptors. Current trends, research objectivesand con- haviour of the Red-tailed Hawk Buteojamaicensis. Ibis servation strategy.Pages 231-239 in B.-U. Meyburg 131:22-26. and R.D. Chancellor [EDS.], Raptor conservationto- nANGRAND,O. 1990. Guide to the birds of Madagascar. day. World Working Group on Birds of Prey, Berlin, Yale Univ. Press, New Haven, CT U.S.A. Germany. --AND S. GOODMAN. 1995. Monitoring Madagas- THORSTROM, R., R.T. WATSON, B. DAMARY, F. TOTO, M. car's ecosystems:a look at the past, present, and fu- BABAAND V. BABA.1995. Repeatedsightings and first ture of its wetlands.Pages 203-214 in T.B. Herman, capture of a live MadagascarSerpent-eagle Eutriorchis S. Bondrup-Nielson, J.H. Willison and N. Munro astur. Bull. B.O.C. 115:40-45. lEDS.], Ecosystemmonitoring and protected areas. THORSTROM,R. AND L. A. REN• DE ROLAND. 1997. First Science and Management of Protected Areas Associ- nest record and nestingbehaviour of the Madagascar ation, Nova Scotia, Canada. Red Owl (Tyt0 soumagnei).Ostrich 68:42-43. LOUETTE,M. ANDJ. STEVENS.1992. Conserving the en- THORSTROM,R., j. HART AND R.T. WAYSON.1997. New demic birds on the ComoroIslands, I: generalconsid- record, ranging behaviour, vocalization and food of erations on survivalprospects. Bird Conserr.Int. 2:61- the Madagascar Red Owl (Tyto soumag•ei),Ibis 139: 80. 477-481. MACPHEE,R.D.E. AND D.A. BURNEY. 1991. Dating of VAUGHAN, R.E. AND P.O. WIEHE. 1937. Studies on the modified femora of extinct dwarf hippopotamusfrom vegetation of Mauritius, I. A preliminary survey of southern Madagascar:implications fbr constraining plant comInunides.J. Ecol.25:289-343. hmnan colonization and vertebrate extinction events. VIRANI,M.Z.A. 1995a. The ecologyof the endangered J. Arch.Sci. 18:695-706. Sokoke Scops Owl Otus ireneae.M.S. thesis, Univ. MARLON,W.R, AND R.A. RYDER.1975. Perch-siteprefer- Leicester, U.K. ence of four diurnal raptors in northeastern Colora- 1995b. Sokoke Scops Owl in Tanzania. Swara: do. Condor 77:350-352. May-June. MARTIN,B.P. 1987. World birds. GuinnessSuperlatives, WATSON,J. 1981. Population ecology,ibod and conser- Enfield, U.K. vation of the SeychellesKestrel (Falcoaraea) on Maht. NEWTON,I. 1979. Population ecologyof raptors. T. & Ph.D. dissertation, Univ. Aberdeen, U.K. A.D. Poyser,London, U.K. 1989. Successful translocation of the endemic MARCH 1998 EAST AFRICAN AND WESTERN INDIAN OCEAN RAPTORS 39

SeychellesKestrel Falcoaraea to Praslin. Pages 363- Breeding, growth, development,and managementof 367 in Meyburg B.-U. and R.D. Chancellor, [Los.], the MadagascarFish-eagle (Haliaeetus vodferoides) J Raptorsin the modern world. World Working Group RaptorRes. 30:21-27. on Birds of Prey, Berlin, Germany. --, S. RAZAFINDRAMANANA,R. THORSTROM AND S. RA- WATSON,R.T. 1991. Using birds of prey as an environ- FANOMEZANTSOA.1997. Breeding biology,productivi- mental conservation tool: The Peregrine Fund's ty, siblicide and conservationof the MadagascarFish- World Programme.Environ. Conserv. 18:269-270. eagle. Pages 289-306 in R.T. Watson [ED.], Madagas- 1997. MadagascarFish-eagle conservation. Pages car wetlandsconservation project. Progressreport III, 45-49 in R.T. Watson [ED.], Madagascarwetlands con- 1995-1996. The Peregrine Fund, Boise, ID U.S.A. servation project. Progressreport III, 1995-1996. The WORLDBANK. 1992. World development report 1992: de- Peregrine Fund, Boise,ID U.S.A. velopment and the environment. Oxford Univ. Press, --AND R. LEWIS. 1994. Raptor studies in Madagas- New York, NY U.S.A. car's rainforest. Pages283-290 in B.-U. Meyburg and ZIMMERMAN,D.A., D.A. TURNERAND DJ. PEARSIN.1996. R.D. Chancellor [EDS.], Raptor conservation today. Birds of Kenya and northern Tanzania. Princeton World Working Group on Birds of Prey, Berlin, Ger- Univ. Press,Princeton, NJ U.S.A. many. --, S. THOMSETT, D. O'DANIEL AND R. LEWIS. 1996. Received16 October 1996; accepted 12 November 1997