Table Of Content

Scopus 19: 83-91,March 1997 Field observations on Prince Ruspoli's Turaco Tauraco ruspolii LucaBorghesio Although the Musophagidae are among the most characteristic birds oftropical Africa, they have receivedverylimitedattentionfromornithologists,andlittleinformationispresentlyavailableontheir biologyinthefield. Theneedforfieldobservationsisparticularlypressingforthosespecieswhosesurvivalisatrisk. One of these is Prince Ruspoli's Turaco Tauraco ruspolii, an endemic of the southern Ethiopian Highlands with averyrestrictedrange, probably notlargerthan 5000km^. The species is currently consideredEndangeredinthelUCNcriteria(Collaretal. 1994),butowingtodifficultyofaccessand securityproblemsinthearea,ithasneverbeenstudiedindetail. InthispaperIreportsomedataonthebiologyofT.ruspolii,collectedduringanexpeditiontosouth Ethiopiainspring 1995 anddiscusstheminthelightofpresentknowledgeofthegenusTauraco. Study area Observations were carried out in the Ethiopian administrative provinces ofBale and Borana, withinthepresentlyknownrange ofT. ruspolii. Elevation inthe arearanged between 1200 and2000m. Owingtoelevation,temperaturesinthestudyareaarenothigh,withannualmeans between 18°and21°C.Rainfallsintwoseparaterainyseasons(inAprilandOctober), andannualrainfall averages 700mm, with somevariation indifferent areas owingto altitude andexposuretorain-carrying winds (Wolde-Mariam 1969). The main habitat is woodland dominated by Acacia abyssinica, A. seyal, A. brevispica,Dichrostachys cinereaandTerminaliabrownii; figs {Ficussycomorus,F. thonningi,F. vasta) are alsocommon. Northwards, woodlands grade into a much wetter forest habitat, where the dominant tree is Podocarpus gracilior. These forests are the northern limit of distributionofT. ruspolii,whichisusuallyreplacedinsidethembytherelatedspecies T. leucotis (Borghesio, inpress). Near the villages of Neghelli (5°20N, 39°35E) and Arero (4°45N, 38°49E) a different vegetation occurs, dominated byJuniperusprocera; and in Arero,junipers formatrueforesthabitat,which,however,onlyextendsoverasmallarea(probablyno more than 25 km^); elsewhere they grow sparser and are mixed with species coming fromthe adjoining woodlands. Methods Observations were carried out between 23 March and 6 June 1995, during daytime, from 6:30 to 18:30. Data were recorded opportunistically after making sure that the birds had notbeen disturbed orinfluenced in the execution oftheirbehaviour. Three main activities were observed: - feeding: individuals observedpecking orswallowing afooditem; -83- 84 LucaBorghesio - resting: individuals perching in a fixed position for at least two minutes, even if engaged inpreeningorothercomfort activities; - calling: individualsvocalizingfromafixedposition. Birdscouldbealoneoringroups. Agroupwasformedbyindividualsobservedon the same tree or on nearby trees, but moving in the same direction in a co-ordinated way orperformingthe same activity. The day was divided into six 2-h periods and the distribution ofthe observations andofactivitieswascheckedwith testsagainstthenullhypothesisthattheydidnot vary over the day, assuming that the number of hours of fieldwork in each period representedtheexpecteddistributionofthedata.Whendatawerenotenoughtopermit statistical analysis, adjacentperiods weremerged. Plant species were determined using Hedberg & Edwards 1989, Noad & Bimie 1989 andBekele-Tesemmaetal. 1993. Results Vocalizations Fourdifferentvocalizations wererecorded. These were: 1. the"main"call ( = song), withprobableterritorialfunction; 2. asoftcontactcall utteredbybirds onthemove; 3. agrowling call, withprobable long-distancecontactfunction; 4. astresscall. The main call clearly resembles that ofother Tauraco species (Dowsett-Lemaire & Dowsett 1988,Fryetal. 1988)initsgeneralfeatures.Itiscomposedofanintroductory cluckingnote,lastingaboutonesecond,andfollowedbyasequenceofabouttenlow- pitched kuk sounds uttered at a speed of 2-3 s^ The song of T. leucotis, although similar,iseasilydistinguishedinthefieldasitisclearlylouderandmorecroakinginits intonation. The song ofT. ruspoliiis not frequent, and it was heard on only eleven occasions during the survey. Singing birds were usually hidden inside the foliage oftrees and difficuhto locate. As is common in theMusophagidae (Candy 1984, Fry etal. 1988, Decoux & Erard 1992), the call ofone bird often stimulatedthe answerofup to two othersnearby. Ononeoccasion, atWadera (5°45N, 39°20E), in aPodocarpusforest whereT.leucotiswasmuchmoreabundantthanT.ruspolii,thesongofthefirstspecies triggeredthatofthe second. Twodifferentcallshadaprobablecontactfunction. Thefirstwasasoftcrrr,about one second in duration, probably corresponding to the soft chirrr-cha reported by Benson (1945). This was regularly uttered while moving among the branches, and probablypermittedindividualsinacoupleoragrouptokeepclosetoeachother; this call,althoughaudibleonly atshortdistances, oftenallowedthedetectionofbirdsthat wouldhaveotherwisegoneunnoticed. Obsen^ationson Tauracoruspolii 85 The second contact call was a coarse, loud growl, 1-2 s in duration, usually repeatedseveraltimesforupto30 sandinachoruswithotherindividualsresponding from nearby. Birds engaged in these vocalizations were usually sitting in hidden positions amongtree foliage. Theprobablecontactfunction is suggestedby the short duration of these choruses, in which all the birds called simultaneously, while responsesdeterminedbythe"main"callwere spreadoversomeminutes. Apparently, there was some variation in these growling calls in different parts ofthe range ofT. ruspolii: AtWadera,theywereindistinguishablefromthecallsutteredbyT. leucotis, andonatleasttwooccasions,inthePodocarpusforestnorthofthevillage,individuals ofthetwospeciesparticipatedinthesamechorus.AtArero,whereT. leucotiswasnot present,thecalls were shorterinduration andless coarse intimbre. Thelastvocalizationwasasuddenshriek,ofveryshortduration,givenasasignof fear, usually when apersonunexpectedly appearednearabird. Diet T. ruspoliiseemstofeedlargelyorperhapscompletelyonfruit. Variousotherspecies ofturacoshavebeenreportedtorelypartly onanimalfood (insects), especially while raising young (Jarvis & Currie 1978, Fry et al. 1988), but T. ruspolii was never observedtodoso,althoughthispossibilityisnotruledout, sincebreedingindividuals werenotobservedduringthe survey. Table 1.Foodplantso/Tauracoruspolii Species Family Fruitdiam Fruitcolour Habitat (cm) Podocarpusgracilior Podocarpaceae 2.0 green Podocarpusforest,forest margins Juniperusprocera Cupressaceae 0.5 green Juniperusforest Ficussycomorus Moraceae 2.0 brownish woodland Ficusvasta Moraceae 1.5 green/yellow Juniperusforest,woodland Ficusthonningi Moraceae 0.8 green JuniperusandPodocarpus forests,forestmargin, woodland Veprisdainelli Rutaceae 2.0 orange Podocarpusforest,forest margin Tecleasp. Rutaceae 0.5 green woodland (prob.simplicifolia) Syzygiumguineense Myrtaceae 1.0 purple riverbanks Cordiaafricana Boraginaceae 1.0 yellowish cultivatedatArero Acokantheraschimperi Apocynaceae 1.5 yellowish Juniperusforest,forest margin 86 LucaBorghesio Table 1 liststenplantsonwhichT. ruspoliiwasobservedtofeed,withthehabitats wheretheywerefoundandinformationonthesizeandcolourofthefruit.Fruitsvaried widelybothincolourandindiameter; itisremarkablethatmanyofthemweregreen, thusconfirmingthatthiscolourisbynomeans avoidedbyturacos (Dowsett-Lemaire 1988). Fruits were always swallowed whole, and this possibly sets the upperlimit of theirdiameter. Figs {Ficus spp.) andthe two conifer species {Juniperusprocera andPodocarpus gracilior) probably represented the most important food plants (Fig. 1): these three speciestotalled74percentoffeeding observations (/t = 81) andsingle individuals or groupsuptoeightwereoftenobservedtofeedonthesetrees, sometimesspendingthe wholedayinorclosetothem. Table 1 lists alsoAcokantheraschimperi, the"poison- arrow tree"; feeding by T. persa corythaix (the Knysna Lourie or Turaco) on the relatedA. spectabilis and by T. hartlaubi onA. longiflora has already been reported (Jubb 1965, Fry et al. 1988), and judged extraordinary, based on the supposed poisonousness of these fruits; however, the ripe fruits of A. schimperi are not poisonous, andarehabituallyeatenbyhumans (Bekele-Tesemma 1993),thusthere is noreasontosuspectthatT. ruspoliipossessesanyparticularabilitytotoleratevegetal poisons. Figure 1.Foodplantsusedby Tauracoruspolii(percentageof overallfeedingobservations) T. ruspoliiis abletoexploitnewfoodsources,whentheybecomeavailable; thisis suggested by the birds observed on 2 June 1995 feeding on a large Cordia africana growinginthevicinityofthevillageofArero.Thistreewascultivatedandthespecies is not reported from elsewhere in the area (Haugen 1992), although it grows in the forests situatedabout 100kmnorthwards, where T. ruspoliiis stillfound. The list of Table 1 is probably far from complete, since during the survey individualsofT. ruspoliiwereobservedonvariousotherunidentifiedtreesandshrubs bearingfruits onwhichthey wereprobably feeding. General behaviour and daily activity rhythm A total of 188 T. ruspolii were observed during this study; ofthese, 77 (41 percent) were solitary and 111 (59 per cent) in groups of 2 to 8 individuals (mean 2.8±1.4, Observations on Tauracomspolii 87 ^L_L 10:31 12:31 14:31 16:31 to to to to 12:30 14:30 16:30 18:30 Figure2.Numbers6>/Tauracoruspoliiobservedduringdifferent2-hperiods median 2). Figure 2 shows the variation ofthe mean numberofindividuals met with during the day, calculated as the number ofindividuals observed in each 2-h period dividedbythenumberofhoursoffieldworkinthatperiod.Thenumberofindividuals observedvariedsignificantlyduringtheday(%-= 11.4,«= 188,d.f.=5,F=0.047)and there was a marked decrease in the middle hours: this was mainly related to groups, whichshowedhighlysignificantchanges,whileobservationsofsingleindividualsdid not(forgroups,t= 18.6,«= 111,d.f. =5,P=0.002; forsingles, =5.3,n=11,d.f. =5,F=0.37). Inthefirstpartoftheday,from6:30to 12:30,groupsaveraged3.1±1.6individuals, while in the following hours, from 12:31 to 18:30, the mean group size dropped to 2.3±0.5; thisdifference,however, wasnot significant (Mann-Whitney, U= 118, two- tailed, P - 0.13), since median and modal dimensions ofthe groups did not differ, being 2 in both cases. Changes in mean values were therefore related to regular observationofgroups of4to 8 individuals atthebeginning ofthe day, this causing a significantly largerrange ofvariation inthe size (and aconsequently higherstandard deviation) ofthe morning groups than in those ofthe rest ofthe day (Moses test of extreme reactions, n^ = 27, n^ = 12, P = 0.001). These large groups were probably associatedwithfeeding activity. Figure3 showsthedistributionofthethreemainactivities withintheday. Feeding hada strongpeak inthemorning anddecreasedinlaterhours (x^= 10.8, d.f. =2,n = 81,P=0.004); feedingindividualscouldbeobservedthroughouttheday, asreported for other species (Jarvis & Currie 1979, Fry et al. 1988), but most ofthe food was certainly taken in the morning, when the turacos often congregated on a single large fruiting tree, which was usually exploited for several days. Vocal activity was at its maximumattheendoftheday,andreachedaminimumaroundmidday (x^=6.4,d.f. 88 LucaBorghesio - 1.6 3 o — SI 1.4 o - a. 1.2 -o C>D - 1.0 w - o 0.8 - gCZO3 0.6 > - 0.4 _C - o 0.2 - c5 0.0 6:30to 10:30 10:31 to 14:30 14:31 to 18:30 Figure3.Diurnalactivitieso/Tauracoruspoliiobservedduringdijferent4-hperiods =2,n=31,P=0.04),whilerestinghadasmallpeak(notstatisticallysignificant)inthe middle hours (x^ = 5.2, d.f. = 2, n = 30, P = 0.07); variations inresting activity were probably underestimated owing to the clearly lower detection rate of birds in the middlepartoftheday (cf. Fig 2). Fewotheractivitieswereobservedduringthesurveyandnostatisticalanalysiswas attemptedonthese, owingtothe small sizeofthe sample. At Wadera, one T. ruspolii was observed on the same tree with a T. leucotis in a Podocarpusforesthabitat; the twobirds hadraisedcrests andwerepossiblyengaged inaggressivebehaviour,butthiscouldnotbeconfirmedwithcertaintysincetheyflew offimmediately aftertheyhadbeen spotted. On a few occasions, usually in the morning between 7:30 and 10:00, the birds engaged in sudden pursuits, with one individual running along branches or flying to nearbytreesandanotherfollowingitclosely;thisbehaviourdidnotseemtobeaform of aggression as it never resulted in physical contact between the individuals performingit; moreover, whenthepursuitended,thebirdsusuallyremainedtogether inthe sametreeoronthesamebranch, showingno signofstress.Themeaningofthis behaviourisnotclear; I alsoobserved itin T. leucotis. Breeding Duringthe survey no nest ofT. ruspolii was found. However, some information was obtained from local people, who usually knew the species quite well, and called it wayuwarointheBorenalanguage,anameusedalsoforT.leucotis.AtSokora(5°37N, 39°18E) and Arero, where T. leucotis was not present, thus avoiding possible confusion, many local people agreed that the nest of T. ruspolii was "cup-shaped, similartothatofapigeon"andcontained"oneortwowhitisheggs".Alltheinformers Observations on Tauracoruspolii 89 saidthatthis nestwas veryrarely found, because it was invariably wellhidden inthe trees.Oneotherinformer,on22May,claimedhavingobservedanoccupiednestofT. ruspoliiaboutaweekbefore,atalocalityabout 15kmnorthofArero;Iwasnotableto visittheplacetoconfirmtherecord. Local movements In contrast to most other species in this family (Moreau 1958, Fry et al. 1988), T. ruspolii seems not to be a completely sedentary species, and some data suggest the existence of small-scale movements. This evidence derives mainly from repeated visitstothe samelocalities. At Sokora, during three visits (25 March, 14-15 and 22-23 April) numerous individuals wereeasily observedinthewoodland surrounding a smallriverandupto about4kmeastandwestofit. Onafourth,latervisit,on27-28 May,nonewasfound away from the river; only five birds were seen, after a lengthy search, all of them feeding on alargeFicus thonningigrowing afewmetres fromthe water. About 3 km east ofBobela (4°50N, 38°52E), in a woodland adjoiningJuniperus forest, three T. ruspolii were recorded in a 2-km transect on 28 March. Two months later, on 1 June,thesamelocalitywas searchedforonefullmorning,coveringnoless than 10 km, but no turaco was seen. On the other hand, inside theJuniperus forest, 2kmwestofthevillage,T.ruspoliiwasobservedonbothvisits.AtWadera,inanarea ofbetterrainfallthanBobelaandSokora,T. ruspoliiwasobservedonmanyoccasions between22Marchand26May,inthePodocarpusforestandinthetransitionalhabitat wherewoodlandandforest abutted. Thepossibility that T. ruspolii is not completely sedentary was also confirmedby local informers in the area ofthe river Awata, west ofHare Kelo (5°33N, 39°24E), where the species was saidto be very common during the dry period ofthe year, but becomingmuchrarerduringtherains. These observations suggest that T. ruspolii may regularly perform short-range movements, of probably no more than a few kilometres from the wetter habitats (forests, forest margins, riverine formations) where it appears to be recorded throughoutthe year, tothedrierwoodlands surroundingthem. Discussion Most ofthe information on T. ruspolii in the present papercompares quite well with that available for the other species of the genus. However, some differences are notable. Referringtovocalactivity,T.ruspoliiisarelativelysilentspecies,whilemostother turacosareusuallyquitenoisy(Fryetal. 1988).Thedifferenceisparticularlystriking in comparison with T. leucotis, which is probably its closest relative (Moreau 1958, Hall&Moreau 1962)andissometimessympatricwithT.ruspolii.Itmustberecalled, however,thatthis studyonlycoveredaperiodofunderthreemonths, andthatthereis thus nocertaintythatT. ruspoliiis a silent species duringothertimes ofthe year. 90 LucaBorghesio Thedescriptionofthenestandtheeggsbylocalpeopleisingoodaccordancewith thoseavailableforotherspecies ofturacos (Courtenay-Latimer 1942,Fryetal. 1988, Fotso 1993). The possible nesting record for the end ofMay indicates that breeding couldstartattheendofthelongrains.T. leucotiswasbreedinginthestudyareainthe sameperiod,sincetwonestsofthisspecieswerefound,oneatDawa(5°20N,39°02E) andthe otheratKibreMengist(5°53N, 39°00E). As forits feeding preferences, T. ruspolii obviously depends on the fruits that are most common and easily available in its range. Its diet, however, seems not to be restrictedto afewfoodsources anditis likelythatthe species willproveabletofeed onmostkindsoffruitswith softpulp smallerthan about2.5 cmindiameter. Someparticulars ofthebehaviourofT. ruspolii seemtoberelatedto itspreferred habitats,thatis,relativelydryforestmarginsandwoodlands.Most(59percent)ofthe observationswereofgroupedindividuals,andlargegroupswereespeciallyfrequentin the morning, when most feeding tookplace: this couldbe a consequence ofthe food plants being rarer and more spaced from each other in the woodlands than in the forests,thusobligingbirdsfromarelativelylargeareatogatheroneachfoodplant.On the other hand, individuals of T. leucotis that, in the study area, occupied wetter habitats than T. ruspolii, especially Podocarpus forests, where food sources are commoner, weremostly observedas solitarybirds (Borghesio, unpubl. data). The evidence, although not definitive, for the existence of local movements has never been reported for other turacos, and seems to be an adaptation to seasonal habitats where foodavailability is notconstant. Otherauthors (Ash & Gullick 1989), having re-visited the same localities after some years, hypothesized that dramatic numerical decreases couldhave occurred. However, this couldprove incorrect ifthe observations were carried out in different seasons of the year. The only way to correctlyevaluatethepopulationtrendofT.ruspoliiseemstobeofrepeatingcountsin the same localities and atexactly the sameperiod ofthe year, in ordertorule outthe possibility ofmovements thatcouldconcealpopulationchanges. Acknowledgements Manypeoplegavetheirhelpandfriendshipduringthiswork,bothinEthiopiaandinItaly.Theyare listedhereinalphabeticalorder:DrJohnAsh,JohnAtkins,AtoYilmaDellelegn,ProfMauroFasola, AtoAmareKifle,DrYalemtsehayMekonnen,AtoTilayeNigussie,AtoShemseddiNuri,Giampiero Pagani, Prof. Anacleto Sabbadin, Per Ole Syvertsen, Ato Abebe Takwa, Don Turner, Ato Akale Yemane and AtoMelaku Zendu. Theresearch was supported by agrantfrom the Bird Exploration FundandreceivedimportantassistancefromtheEthiopianNaturalHistoryandWildlifeSocietyand theIstitutoItalianodiCulturainAddisAbaba. References Ash,J.S.&Gullick,T.M. 1989.Thepresentsituationregardingtheendemicbreedingbirdsof Ethiopia.Scopus 13: 90-96. Bekele-Tesemma, a., Birnie, A. & Tengnas, B. 1993. Useful trees and shrubsfor Ethiopia. Nairobi: RSCU. Benson,C.W. 1945.NotesonthebirdsofsouthernAbyssinia.Ibis87: 489-509. Observations on Tauracoruspolii 91 BoRGHESio, L. (in press). Observations on the ecology ofT. ruspolii and T. leucotis in south Ethiopia.BulletinoftheBritishOrnithologists' Club. Candy, M. 1984. Habits andbreeding biology ofthe GreatBlueTuraco Corythaeola cristata. JournaloftheEastAfricaNaturalHistorySocietyandNationalMuseum 180: 1-19. Collar, N. J., CrosbyM. J. & STATTERsfiELD, A. J. 1994. Birds to watch2. The WorldListof ThreatenedBirds. Cambridge: BirdLifeInternational. Courtenay-Latimer,M. 1942. KnysnaLourie{Turacuscorythaixcorythaix). Ostrich 13: 1-9. Decoux,J.P.&Erard,C. 1992. TauracomacrorhynchusandT.persainnorth-easternGabon: behaviouralandecologicalaspectsoftheirco-existence.ProceedingsoftheVIIPan-African OrnithologicalCongress: 369-380. Dowsett-Lemaire, F. & Dowsett, R. J. 1988. Vocalisations of the green turacos (Tauraco species)andtheirsystematic status.Tauraco 1: 64-71. Dowsett-Lemaire, F. 1988. Fruitchoice and seed disseminationby birds andmammals inthe evergreenforestsofuplandMalawi.Rev.Ecol. (TerreetVie)43: 251-285. FoTSO,R.C. 1993.BreedingofBannerman'sTuracoTauracobannermanniandBandedWattle- eyePlatysteira laticinctainCameroon.Proceedingsofthe8thPan-African Ornithological Congress:438. Fry,C.H.,Keith,S.,&Urban,E.K. 1988.ThebirdsofAfrica.Vol3.London:AcademicPress. Hall, B. P. & MoreauR. E. 1962. A study ofthe rare birds ofAfrica. Bulletin oftheBritish Museum(Nat.Hist.)Zoology8: 313-378. Haugen,T. 1992.WoodyvegetationofBorana,southEthiopia.EthiopianJournalofScience 15: 117-130. Hedberg, I. & Edwards, S. 1989. Flora of Ethiopia. Vol 3. Addis Ababa: Addis Ababa UniversityandUppsala: UppsalaUniversity. Jarvis,M.J.F. &Currie,M. H. 1979. BreedingofcaptiveKnysnaandPurplecrestedLouries. Ostrich50: 38^4. JuBB, R. A. 1965. Knysna Loerie Turacus corythaix (Wagler) feeding on poisonous plants. Ostrich36: 36-37. MoREAU, R. E. 1958. A contribution to the biology of the Musophagiformes, the so-called plantain-eaters.Ibis 14: 67-112,238-270. NoAD,T. C. &BiRNiE,A. 1989. TreesofKenya.Nairobi:Noad&Brniie. Wolde-Mariam,M. 1969.AnatlasofEthiopia. AddisAbaba: AddisAbabaUniversityPress. LucaBorghesio,C.Re Umberto42,10128Torino,Italy Scopus 19: 83-91, March 1997 Received 30March 1996