Jerdon's Courser, Once Thought to Be Extinct (Ripley 1952,1982, King 1981)

山階鳥研報 (J. Yamashina Inst. Ornith.), 21: 165-174, 1989

Systematics, Biogeography, and Conservation of Jerdon's Courser Rhinoptilus bitorquatus

S. Dillon Ripley* and Bruce M. Beehler*

Abstract A cladistic analysis of Jerdon's Courser (Rhinoptilus bitorquatus) and eight allied taxa supports the validity of the Afro-Asian genus Rhinoptilus and indicates that the sister-species of the Indian relict bitorquatus is the Three-banded Courser (Rhinoptilus cinctus) of eastern Africa. The present distribution of these two sister forms is evidence for a former biotic link between peninsular India and the savanna habitats of eastern Africa. This distributional trend is corroborated by an additional list of forty-three species or sister-species pairs that exhibit this Afro-Indian pattern. We believe that these data support the notion that there once existed an Afro-Indian fauna that inhabited what was probably a continuous belt of savanna from southern Africa to southern India. The recovery plan for the critically-endangered Jerdon's Courser should include at- tempts to develop a captive population of R. cinctus, which could then be used to rear eggs taken from wild populations of bitorquatus. Captive breeding, in concert with local education and efforts to expand protected areas of prime habitat, offers the most promising in- tegrated strategy for the species' recovery.

Introduction

Jerdon's Courser, once thought to be extinct(Ripley 1952,1982, King 1981),was rediscoveredin the hillsof southern Andhra Pradesh in January 1986 (Bhushan 1986a, 1986b). Although itis impossibleto presentlygive an accurateestimate of the distribution and sizeof thisremnant population,the assumption is that the species'numbers are few and probably dissectedinto tiny subpopulations. It is safeto say that Jerdon's Courser, although revivedto the statusof the living,maintains, at best,a tenuous hold on viability,and extinctionfor the speciesmay be only a matter of time. Still,Bhushan's remarkable discoveryhas rekindledinterest in the biologyof the bird,in part to develop a pragmatic recoveryplan for the species,and also because of thisspecies'relictual status on the Indian Subcontinent(Mukherjee 1974, Ripley et al.1988:547). Blyth (1848) described Jerdon's Courser as Macrotarsius bitorquatus,based on specimens procured by T. C. Jerdon in southernIndia. Subsequent revisersargued that Macrotarsiuswas a spellingerror for Macrotarsus of Lacepede (seePeters 1934), a junior synonym of Himantopus and thus not available. Sharpe (1896)recognized Strickland's Rhinoptilus(1850) for the assemblage that included bitorquatus..Blanfbrd (1897) and Baker (1929) followed Sharpe in assigningbitorquatus to the genus Rhinoptilus,as did Peters (1934)and Ripley(1952). More recentlyRipley (1961, 1982, Ali & Ripley 1969) placed all Indian courser

Received 29 May 1989, accepted 21 August 1989 * NHB Room 336 , Smithsonian Institution, Washington D. C. 20560 USA

165 166 S. D. Ripley and B. M. Beehler species in the genus Cursorius, and Snow (1978) and Urban et al. (1986) considered Rhinoptilus not sufficiently distinct from Cursorius for generic status and treated all of the African forms as Cursorius. These varied treatments illustrate the uncertainty regarding relationships among the various courser lineages. In this paper we present the results of a systematic analysis of Jerdon's Courser and its nearest allies, discuss the taxonomic status of the genera Rhinoptilus and Cursorius, and offer some thoughts on the historical biogeography of the group, focusing on Jerdon's Courser. With these data in hand, we then suggest possible strategies for promoting the species' recovery from the brink of extinction.

Methods

For the systematic analysis we compared Jerdon's Courser to eight other glareolid relatives: seven cursoriines (Pluvianus aegyptius, Rhinoptilus africanus, R. cinctus, R. chalcopterus, Cursorius cursor, C. temminckii, C. coromandelicus, and, as an outgroup taxon, a single glareoline, Stiltia isabella. Study skins of all but Jerdon's Courser were available from the collection of the National Museum of Natural History, Washington, D. C., USA. For Jerdon's Courser, we relied on the detailed descriptions of Blyth (1848), Sharpe (1896), Blanford (1898), and Baker (1929), and the color photograph of Bhushan (1986b). For this species assemblage, we compared nineteen characters re- lated to external morphology and life history habits (Fig. 1, Table 1). For each char-

Fig. 1. Key to Throat Characters 1-7, based on the presence of all of these features in Rhinoptilus cinctus. Refer to Table 1 for details. Conservation of Jerdon's Courser Rhinoptilus bitorquatus 167

Table 1. Character-State Data Used in the Analysis

Table 2. Character-State Coding Matrix for the Analysis 168 S. D. Ripley and B. M. Beehler acter we assigned two or more discrete character states and gave them integer codes (0, 1, 2 etc.) to correspond to the particular phenotypic expression of the character (Table 1). After determining the key characters for analysis, we graded the nine courser species for each character and compiled these results into a character-state matrix (Table 2), which was then analyzed using the PAUP cladistic program, version 2.4 (Swofford 1985), an automated phylogenetic analysis that computes minimum-length trees. Character- state series were considered unordered, as it is difficult to develop a priori primitive- derived sequences for plumage and life history data. Characters were unweighted. Our cladistic analysis employed the "alltrees" exhaustive search, a method in which all possible permutations of branching are tested against the available character-state data. The tree with the fewest number of character "steps" and the highest degree of congruence (reduced homoplasy) is accepted as the most likely phylogeny for the study group. Our biogeographic analysis is influenced by the concept of vicariance biogeography (Rosen 1978, Cracraft 1982) and assumes that phylogenetic differentiation has occurred in tandem with the development of large-scale environmental barriers to gene flow. We attempt to equate isolation and speciation events with known or postulated environmental phenomena. The apparent "pattern" is corroborated by discovery of replicate examples from other taxa that share similar ecological requirements. Finally, we employ the available ecological and systematic information about Jerdon's Courser to develop a possible strategy for engineering the species' recovery from the brink of extinction.

Results

Systematics. The PAUP analysis reviewed more than 135,000 possible branching

Fig. 2. Cladogram of Jerdon's Courser and its relatives. Conservation of Jerdon's Courser Rhinoptilus bitorquatus 169

patterns and from this selected a single tree that most parsimoniously explains speciation in the group that includes Jerdon's Courser (Figure 2). The tree, of thirty-seven steps, exhibits a consistency index of 0.649. The cladogram supports the validity of two polytypic genera in the cursoriine assemblage : Cursorius and Rhinoptilus, and the monotypic genus Pluvianus. Four species (bitorquatus, cinctus, africanus, and chalcopterus) form a well-defined Glade Rhinoptilus, and the species cursor, temminckii, and coromandelicus are allied into the genus Cursorius. Rhinoptilus is defined by unique expression of three characters: throat character 1 (posterior dark band), throat character 4 (mottled middle band), and temporal activity (nocturnal). Thus the peculiar throat-banding (see Fig. 1) and predominantly nocturnal habits are characters unique to the genus, unambiguously distinguishing this Glade from any members of the genus Cursorius. Jerdon's Courser is shown to be most closely allied to the Three-banded Courser (R. cinctus), and the two can be considered sister-species. Uniquely-shared expressions of two characters define this species-pair: leg color and throat character 6 (anterior brown necklace). This alliance is also supported by shared character-states in the following: hallux, bill color, bill morphology, primary stripe, belly pattern, posterior breast band, throat characters 2, 3, 4, 5, 7, crown pattern, dorsal pattern, facial stripe, and wing/tarsus ratio.

Biogeography Thepresence of sister-species populations of Rhinoptilus in eastern Africa and southern India (Fig. 3), respectively, supports the notion that the immediate ancestor to these two at some period ranged from Africa to India. We postulate that this hy- pothetical ancestor's broadly-distributed population was then broken by the develop-

Fig. 3. Map of Ranges of Rhinoptilus cinctus and R. bitorquatus (historical range) . 170 S. D. Ripley and B. M. Beehler ment of an intervening environmental barrier (or barriers) that led to the effective isolation of the two terminal populations and to their subsequent genetic differentiation. Present-day barriers of water and desert associated with the Arabian Peninsula are likely to have played a role in the faunal break (Ripley 1953, 1954).

Table 3. Other Avian Examples of the Afro-Asian Distributional Pattern Conservation of Jerdon's Courser Rhinoptilus bitorquatus 171

This hypothesis of a widespread Afro-Indian form being dissected into eastern and western isolates assumes that an once continuous band of suitable habitat extended from eastern Africa through the Persian Gulf to India. Presumably this environmental corridor would have permitted the expansion of an entire Afro-Indian fauna of savanna or dry-forest forms (given the ecological preferences of Rhinoptilus cinctus and bitorquatus). Is there evidence for an Afro-Indian fauna? By conservative count there are at least forty-three other species-pairs or single- species ranges that approximate the Afro-Indian distribution exhibited by Rhinoptilus (Table 3). Nearly all of these are savanna or woodland forms. In fact, another pair of coursers, Cursorius coromandelicus and C. temminckii exhibit this same distribution. In other instances as in that of the flamingo Phoeniconaias minor, a single species is distributed (patchily) from southern Africa to India. The pattern is most often found in non-passerine taxa or with open country passerines (e. g., larks, pipits, and granivorous finches). It is apparent that the modern salt water barriers of the Red Sea and Persian Gulf were not insurmountable faunal barriers in what must have been a corridor of favorable habitat. That other animal taxa exhibited this pattern is evident from the historical presence of a relictual population of the Cheetah (Acinonyx jubatus) in southern India, the remnant Lion (Panthera leo) population in the Gir Forest of Gujerat, and the presence in India of at least two ungulate species of African affinities (Prater 1980, Uerpmann 1987). That this Afro-Indian fauna is of relatively recent age is clear from the relatedness of the sister-populations inhabiting either side of the Arabian barrier.

Discussion

The closest relative to Jerdon's Courser is shown to be an East African form that is part of a species-radiation largely confined to Sub-Saharan Africa. The importance of an Indian faunal linkage with Sub-Saharan Africa is often overlooked in Indian ornithology, mainly because the significantly more species-rich forest avifauna is so strongly affiliated with that of southeast Asia (Ripley 1959). We believe the African connection to be important in its own environmental context. It is India's dry-habitat avifauna that shows closest affinity to the African biota (Table 3). We believe that major environmental shifts from the west (Africa) as well as from the east (southeast Asia) influenced the evolution and diversification of Indian biota. Wetter periods saw the development of tracts of humid forest linking southern India with Burma and humid south-east Asia. In these forests evolved an Indo-Malayan humid forest fauna. We envision that during other periods expanses of dry forest and savanna expanded all along the rim of the Arabian Sea, from southern Africa to south-central India. The abundant examples of Afro-Indian bird distributions (Table 3) provide evidence that a rich Afro-Indian biota populated this arc of savanna habitat. The vast deserts of Arabia, today a barrier to interchange, were presumably not present during the period when the proposed habitat corridor existed. High water levels in now-dry lakes on the Arabian Peninsula indicate that a more humid climate dominated in the following periods: 24,000-18,000 yr B.P. and 9,000- 172 S. D. Ripley and B. M. Beehler

8,000 yr B.P. (Street & Grove 1980). During the older period India was relatively dessicated, but during the second, India was under a moist climatic regime (Van Campo 1986). The current depth of Bab al Mandab (the strait separating the Red Sea from the Gulf of Aden) and the paleoceanographic data on the sediments of the Red Sea indicate that this salt water barrier was present during the last 150,000 years (Reiss et al. 1980). If the Afro-Indian fauna developed during a moist period, one can assume that salt water barrier(s) were present but were not a significant detriment to expansion of the biota east and west. The salient point to be gotten from the acknowledgment of an "African connection" is that the diversification of the Indian biota has been a complex process, but not necessarily one without order. At some point in the future it will be important to analyze the interplay, in peninsular India, of environmental influences from the discrete biotic colonizations from east and west.

Conservation Action Using data in the literature, recent surveys by B. Bhushan (1985, 1986a, 1986b), and results of our systematic treatment, it is possible to develop a plausible strategy for saving Jerdon's Courser. The strategy we support involves four activities: (1) con- tinued survey of the population; (2) increased efforts to protect favored habitat; (3) expanded local education; and (4) captive propagation. At the time of this writing we know of recent sightings of Jerdon's Courser only in the vicinity of the Lankamalai Range of southern Andhra Pradesh. The historical range of the species extends northward across the Godavari River to the Andhra/Madhya Pradesh border and westward to Anantapur. We encourage the Bombay Natural His- tory Society and state wildlife departments to refine and intensify their search methods for the bird. Bhushan discovered that Jerdon's Courser is primarily nocturnal. Thus night searches, though difficult, may be the only way to obtain good counts. Once the survey efforts produce a scattering of records, it will be possible to de- lineate present habitat preferences of the species. These data will be critical for determining which tracts within the historical range of Jerdon's Courser should be pre- served. Ideally, remote sensing photos of courser habitat could be compared with the satellite data for the expanse of southeastern India in order to determine the largest tracts favorable for the species. The largest and least disturbed of these should be given high priority for preservation whether or not they support populations of the bird. Unpopulated tracts can subsequently serve as areas of recolonization by means of release of captive-propagated coursers. The local populace in areas where Jerdon's Courser may occur should be offered bounties for proven reports of live, free-ranging birds. The bounty might offer sufficient incentive for local shikaris to avoid killing a bird that otherwise might go into the stewpot-an economic incentive for forbearance on the part of hunters. All local wildlife officers should know the bird and should be able to evaluate properly incoming local reports from the citizenry. This last requirement is perhaps one of the most difficult to put into effect, because of the difficulty for non-ornithologists to effectively Conservation of Jerdon's Courser Rhinoptilus bitorquatus 173

describe and identify the courser. Captive propagation certainly will require the greatest investment of time and money, but nonetheless may be critical to the future of Jerdon's Courser. We suggest initially developing a model propagation system by using the African sister-species Rhinoptilus cinctus. Once this system is perfected, pairs of cinctus can be used to raise eggs of bitorquatus harvested from the field. Viable clutches can be located by radio- tracking breeding pairs. The propagated captive stock of bitorquatus could be used to produce a stock of young for release into the best tracts of habitat. With this plan, we do not see a need to remove adult, free-ranging bitorquatus from the wild.

Acknowledgments

We thank Bharat Bhushan for enthusiastically sharing with us the recent information on Jerdon's Courser. We thank the Government of India, and the Government of Andhra Pradesh for permission to conduct fieldwork in the Eastern Ghats, home to Jerdon's Courser. Ms. Francine Berkowitz, administrator of the Smithsonian Special Foreign Currency Program, provided financial support of our Indian field studies. We thank the Bird Division, National Museum of Natural History, for loan of specimens from their collection.

References

Au, S. & Ripley, S. D. 1969. Handbook of the birds of India and Pakistan. Vol. 3. Oxford Uni- versity Press/New Delhi. Baker, E. C. S. 1929. The fauna of British India. Birds-Vol. 6. Taylor and Francis/London. Bhushan, B. 1985. Jerdon's or Double-banded Courser Cursorius bitorquatus (Blyth)-preliminary survey. Tech. Rept. No. 9. Bombay Nat. Hist. Soc./Bombay. Bhushan, B. 1986a. Rediscovery of the Jerdon's or Double-banded Courser Cursorius bitorquatus (Blyth). J. Bombay Nat. Hist. Soc. 83: 1-14. Bhushan, B. 1986b. Photographic record of the Jerdon's or Double-banded Courser Cursorius bitorquatus. J. Bombay Nat. Hist. Soc. 86: 159-162. Blanford, W. T. 1898. The Fauna of British India. Birds, Vol 4. Taylor & Francis/London. Blyth, E. 1948. [A new courser from southern India]. J. Asiat. Soc. Bengal 17: 254. Cracraft, J. 1982. Origin and evolution of continental biotas: speciation and historical congruence within the Australian avifauna. Evolution 40: 977-996. King, W. B. 1981. Endangered birds of the world. Smithsonian Institution Press/Washington, D. C. Mukherjee, A. K. 1974. Some examples of recent faunal impoverishment and regression. pp. 330-368 in M. S. Mani (ed.) Ecology and biogeography in India. W. Junk/the Hague, Netherlands. Peters, J. L. 1934. Check-list of birds of the world. Vol. 2. Harvard Univ. Press/Cambridge, Mas- sachusetts. Prater, S. H. 1980. The book of Indian animals. Bombay Nat. Hist. Soc./Bombay. Reiss, Z., Luz, B., Almogi-Labin, A., Halicz, E. & Wolf, M. 1980. Late Quaternary paleooceanography of the Gulf of Aqaba (Elat), Red Sea. Quatern. Res. 14: 294-308. Ripley, S. D. 1952. Vanishing and extinct bird species of India. J. Bombay Nat. Hist. Soc. 50: 902- 906. Ripley, S. D. 1953. Considerations on the origin of the Indian avifauna. Bull. Nat. Inst. Sci. India (New Delhi) 7: 269-275. Ripley, S. D. 1954. Comments on the biogeography of Arabia with particular reference to birds. J. 174 S. D. Ripley and B. M. Beehler

Bomb. Nat. list. Soc. 52: 241-248. Ripley, S. D. 1959. Zoogeographic considerations on the Indian avifauna. J. Bomb. Nat. Hist. Soc. 56: 72-81. Ripley, S. D. 1961. Synopsis of the birds of India and Pakistan. Bombay Nat. Hist. Soc./Bombay. Ripley, S. D. 1982. Synopsis of the birds of India and Pakistan. Second Ed. Bombay Nat. Hist. Soc./ Bombay. Ripley, S. D., Beehler, B. & Krishna Raju, K. S. R. 1988. Birds of the Visakhapatnam Ghats. Part one. J. Bomb. Nat. Hist. Soc. 84: 541-559. Rosen, D. E. 1978. Vicariant patterns and historical explanation in biogeography. Syst. Zool. 27: 159-188. Sharpe, R. B. 1896. Catalogue of the Limicolae in the collection of the British Museum. British Museum (Nat. Hist.)/London. Snow, D. W. (ed.) 1978. An atlas of speciation in African non-passerine birds. British Museum (Nat. Hist.)/London. Street, F. A. & Grove, A. T. 1979. Global maps of lake-level fluctuations since 30,000 yr B. P. Quatern. Res. 12: 83-118. Swofford, D. 1985. Phylogenetic analysis using parsimony. Illinois Nat. Hist. Survey/Champaign, Illinois. Uerpmann, H.-P. 1987. The ancient distribution of ungulate mammals in the Middle East. Dr. Ludwig Reichert Verlag/Wiesbaden. Urban, E. K., Fry, H. K. & Kieth, S. (eds.). 1986. The Birds of Africa. Vol. 2. Academic Press/London. Van Campo, E. 1986. Monsoon fluctuations in two 20,000-yr B. P. Oxygen-isotope/pollen records off southwest India. Quatern Res. 26: 376-388.

クビワスナバシリRhinoptilus bitorquatusの分類,生物地理および保護

インド産のクビワスナバシリJerdon's Courser Rhinoptilus bitorquatusは一時絶滅と考えられていたが(1952～1982),Bhushan(1986)によって再発見された。本種に近縁のアフリカとインドの8種を19の

形質(形態特徴と昼夜行性)についてPAUP分岐分析法(相対形質を0と1または2,3と置く)で比較

したところ,アフリカ産のウロコクビワスナバシリR.cinctusと近い姉妹種であることが証明された。こ

れは生物地理学的に東アフリカからアラビアを介してインドに連なる乾燥帯があったことを示唆し,この

アフリカーインド動物相(Afro-Indian fauna)は他の鳥類約40種や獣類でも立証される(一般にはアジアー

インド系の湿性林が重視されるが)。この希種の保護には姉妹種R.cinctusを人工繁殖させ本種R.bitor- quatusの卵を孵化させる方法をとれば,生息地保護や教育に加えて効果があると思う(文責黒田長久)。

S. Dillon Ripley & Bruce M. Beehler: NUB Room 336, Smithsonian Institution, Washington, DC 20560 U. S. A.