Durham E-Theses. Studies of Himalayan pheasants in Nepal with reference to their conservation. Lelliott, Anthony D.

Transcription

1 Durham E-Theses Studies f Himalayan pheasants in Nepal with reference t their cnservatin Lellitt, Anthny D. Hw t cite: Lellitt, Anthny D. (1981) Studies f Himalayan pheasants in Nepal with reference t their cnservatin, Durham theses, Durham University. Available at Durham E-Theses Online: Use plicy The full-text may be used and/r reprduced, and given t third parties in any frmat r medium, withut prir permissin r charge, fr persnal research r study, educatinal, r nt-fr-prt purpses prvided that: a full bibligraphic reference is made t the riginal surce a link is made t the metadata recrd in Durham E-Theses the full-text is nt changed in any way The full-text must nt be sld in any frmat r medium withut the frmal permissin f the cpyright hlders. Please cnsult the full Durham E-Theses plicy fr further details. Academic Supprt Oce, Durham University, University Oce, Old Elvet, Durham DH1 3HP e-theses.admin@dur.ac.uk Tel:

2 STUDIES OF HIMALAYAN PHEASANTS IN NEPAL WITH REFERENCE TO THEIR NSERVATION by Anthny D. Lellitt, B.Sc. Submitted t the University f Durham fr the degree f Master f Science The cpyright f this thesis rests with the authr. N qutatin frm it shuld be published withut his prir written cnsent and infrmatin derived frm it shuld be acknwledged.

3 » e 0 e Dedicated t the memry f my father, Denis W. Lellitt s»

4 ABSTRACT The Bld Pheasant Ithaginis cruentus, Satyr Tragpan Tragpan Satyra, Kklass Pheasant Pucrasia macrlpha, and Himalayan Mnal Lphphurus impeyanus, were studied fr seven mnths in the Suth Annapurna regin, Nrth f Pkhara, Nepal, in 1979 and Study was cncentrated in the Pipar area, between 3000m and 4000m altitude, where ppulatin densities were estimated by cunting the numbers f calling males (Tragpan and Kklass), and by Direct Cunts f birds (Bld Pheasant and Mnal). Ppulatin densities ranged frm 2.5 t pairs per km fr the first three species, and althugh a slight decline was nted in the 1980 densities, the ppulatins were cnsidered t be quite healthy. An assessment f the habitat preferences f each species was made, which shwed that these verlapped cnsiderably. Diet and feeding behaviur were studied and cmpared with bservatins made by previus wrkers. All species shwed verlapping fd preferences, but these were different in detail. Aspects f the behaviur f each species were studied, including prtective behaviur, daily activities, breeding behaviur, and vcalizatins. The latter were tape recrded, analys ed snagraphically, and mst are described here in detail. The male Tragpan and Kklass were bserved t make dawn challenge calls; the functin f these and the calls f the ther species are discussed. Observatins made n breeding bilgy and territriality were cmpared with thse in the literature and are als discussed. A f i f t h species, the Cheer Pheasant Catreus wallichii, was sught fr in the Athhazar Parbat regin, Nrth-west f Pkhara, Nepal. A

5 small ppulatin was lcated in 1980, and bservatins were made n these fr fur days in May. The Cheer is included in the relevant sectins alng with the ther fur species. A study f human influence n all the species was undertaken, which included the effects f livestck herds, hunting, burning, and frest clearance. In the Suth Anna puma regin, pressure n the pheasants was nt cnsidered t be t great at present, but i t is thught likely t increase in the near future. The status f the study species is cnsidered and recmmendatins fr their cnservatin are made. These include the setting up f a reserve, regulatins fr hunting, frest prductin and pastralism, and recmmendatins fr educatin and research.

6 ACKNOWLEDGEMENTS I am very grateful t His Majesty's Gvernment f Nepal, in particular t the Ministry f Frests and the Natinal Parks and Wildlife Cnservatin Office fr permissin t wrk in Nepal. This prject was carried ut under the auspices f the Wrld Pheasant Assciatin, and funds were made kindly available thrugh Muntain Travel/Tiger Tps (Nepal) Ltd. In particular, I wuld like t thank: Lt. Cl. J. 0. M. Rberts and Mr. K. Sakya, Chairman and Secretary respectively f W.P.A. Nepal, wh bth devted much f their time t the success f the prject; Mr. P. B. Ynzn and Ms. Heather Wright fr their helpful c-peratin in the fieldwrk; Mr. K. C. R. Hwman and Mr. C. D. W. Savage wh helped plan and execute the study frm England; and Dr. P. R. Evans wh kindly supervised the preparatin f this thesis. Bb Gibbns, Patrick Rbinsn, Michael Green and the ther members f the Durham University Himalayan Expeditin prvided the riginal inspiratin t wrk in Nepal, and I am grateful t them fr their cntinued interest and advice. Dr. Peter Slater and Dr. Jhn Culsn bth made available t me the use f Kay Snagraph machines, Mr. Rn Kettle prvided tape, Dr. Jhn Munr prvided medical supplies, Dr. Tusi Butterfield examined pheasant faecal pellets, Mr. Jn Sctt helped with statistical analysis, Mr. Dave Schafield drew Figure 20, Messrs. Alan Eddy and Arthur Chater identified plant specimens, Miss Barbara Pattersn typed the thesis, and Ms. Antnia Salvage translated German literature and prvided help and encuragement.

7 Thanks are als due t : the Baha'i cmmunity in Kathmandu, Rachel and Ansn Cruch, Dr. Geffrey Davisn, Dr. Peter Garsn, Dr. Tny Gastn, Maj. Iain Grahame, Subasing Gurung, Dr. David Jenkins, Dr. Tim Lvel, Richard Margschis, Prank Pppletn, Dr. Dick Ptts, David Pritehard, Matt Ridley, Bill Rigden, Shel Severinghaus, and Cl. Terry White. Finally, thanks must als g t the fllwing W.P.A. members wh prvided infrmatin n their captive pheasants, D. Bayliss, K. Chalmers Watsn, Chessingtn Z,Edinburgh Z, I. Grahame, K. Hwman, D. Jnes, Lilfrd Hall, Marwell Z, W. Presctt and M. Sawyer: and als the field teams f sherpas and prters, especially Ram Kaji Mangar, Rinzing Sherpa, Nema Chttar and Anta Bahadur Gurung.

8 6 V NTENTS Page ABSTRACT i ACKNOWLEDGEMENTS i i i NTENTS DECLARATION v v i i i 1. INTRODUCTION Bld Pheasant Satyr Tragpan Kklass Pheasant Himalayan Mnal Cheer Pheasant Research t Date 6 2. THE PRESENT STUDY Descriptin f the study area and ther areas visited Climate Majr Vegetatin Types METHODS Fieldwrk Labratry Wrk Questinnaire HABITATS AND POPULATION DENSITIES General Habitats in the Suth Annapurna Regin Habitat Preferences f Pheasants Ppulatin Densities f Pheasants Habitats and Occurrence f Pheasants in the Athhazar Parbat Regin 63

9 vi e Page 5. FOOD AND FEEDING BEHAVIOUR Diet 66 # 5-2 Discussin 72 5.) Feeding Behaviur 7^ 6. BEHAVIOUR AND VOCALIZATIONS 80 # 6.1 Bld Pheasant Satyr Tragpan Kklass Pheasant 109 # 6.4 Himalayan Mnal Cheer Pheasant 1^3 7. BREEDING BIOLOGY AND TERRITORIALITY 1^ Bld Pheasant 1^7 7.2 Satyr Tragpan 1^9 7.3 Kklass Pheasant 152» 7-4 Himalayan Mnal Cheer Pheasant HUMAN INFLUENCE ON THE PHEASANTS AND REMMENDATIONS B'OR THEIR NSERVATION Types f Human Influence Status f the Study Species in Captivity Status f the Study Species in the Wild Recmmendatins fr Future Cnservatin 179 APPENDICES ' 1. Descriptins f the five study species f pheasants Direct Encunter Card Details Questinnaire supplied t W.P.A. members. 192 # 4. Snagrams f pheasant calls. 19^ 5. Crrelatin/Regressin charts fr sunrise vs. pheasant calling. 207 # 6. Census pint lcatins at Pipar. 212

10 vii Page 7. Satyr Tragpan call rate details Time f first call f pheasants with respect t weather 21 4 and sunrise. 9. Satyr Tragpan: First call times in relatin t weather cnditins Kklass Pheasant call-type shift Kklass Pheasant: First call times in relatin t weather cnditins Himalayan Mnal: First call times in relatin t weather cnditins. pp-i CLrL1 - l j. Questinnaire used in Human Influence Study Methd f Snare Mechanism Census f captive pheasants. 225 BIBLIOGRAPHY 226» s»» >

11 DECLARATION Nne f the material cntained in this thesis has previusly been submitted fr a degree in this r any ther University. STATEMENT OF PYRIGHT The cpyright f this thesis rests with the authr. N qutatin frm i t shuld be published withut his prir written cnsent, and infrmatin derived frm i t shuld be acknwledged.

12 1. INTRODUCTION This thesis dcuments the results btained during twelve mnths f study f five species f pheasants in Central Nepal in 1979 and Wdland pheasants are difficult t bserve in the wild (Severinghaus 1977)J and I have therefre integrated my smewhat limited findings with all published infrmatin, t build up a mre cmplete picture f the eclgy f the five species. The pheasants are a grup f lsely-related game birds f the Family Phasianidae and Sub-Family Phasianinae. Accrding t Delacur (1977) the criteria fr distinguishing a pheasant are nt well-defined, but the male bird shuld pssess highly clured, rnate plumage and be f 'large size 1. Other related birds lacking these characteristics are placed in varius grups such as Snwccks, American Quails, and Franclins. The five species f Nepalese pheasant t be discussed in this study are placed in the tribe Phasianini which cmprises sixteen genera that are distinctly separate and prbably f plyphyletic rigin (Delacur 1977). Plumage descriptins can be fund in Appendix Bld Pheasant (Plate 1) The Bld Pheasant, genus Ithaginis, Wagler 1832, Isis cl. 1228, cnsists f nly ne species, Ithaginis cruentus, which can be divided int furteen subspecies (Delacur 1977) which range thrugh the Himalayas int the muntains f West-central China, shwing marked clinal gegraphic variatin. Vaurie (1965) hwever, recgnises nly eleven subspecies. 1 - \ R * ' rrv.. The bird ccurring in Nepal is the Himalayan Bld Pheasant, Ithaginis cruentus cruentus which ccurs acrss nrthern Nepal frm apprximately 83 E lngitude t nrthwestern Bhutan, where i t intergrades

13 Plate 1 Male Bld Pheasant n lw branch f tree. Plate 2 Male Satyr Tragpan n branch f Rhddendrn barbatum tree

14 I I 1* «J

15 with I.e. tibetanus. Its lcal names in Nepal are 'Chilme' (Nepali) and 'Chermung' (Sherpa). Habitat: Resident. Occurs in steep pine, birch, dwarf rhddendrn, dense ringal bamb and juniper frest and scrub. Usually fund clse t the snw line between c 36OO and 4300m altitude during the summer and mving seasnally with the snw line dwn t perhaps 2700m in the winter (Ali and Ripley 1969) Delacur (1977) recrds them descending belw 2700m in the cldest part f their range "but never belw 7000 feet" (2135m). 1.2 Satyr Tragpan (Plate 2) The Tragpans, genus Tragpan, Cuvier 1829, cnsist f five separate species and ne subspecies ccurring alng the Himalayan range int central and eastern China. The species encuntered in Nepal is the Satyr Tragpan, Tragpan satyra, which ccurs rughly frm the Alaknanda river, in Gharwal prvince, thrugh Nepal t 'Darrang', Nrth f the Brahmaputra river in Assam (Ali and Ripley 1969). At the extreme western and eastern limits f its range, its relatinship t the neighburing species (T. melancephalus and T. blythi respectively) is unknwn. The lcal name fr this pheasant in Nepal is 'Mnal', which shuld nt be cnfused with the Himalayan Mnal (Impeyan Pheasant, Lphphurus impeyanus). Habitat: Resident. Occurs in Rhddendrn, ak, dedar, and bamb frest, ften n steep inaccessible slpes with thick undergrwth, between 2400 and 4250m, and dwn t c l800m in severe winters (Ali and Ripley 1969) These heights represent the extremes f range fr this species which is usually fund between 2150 and 3500m thrughut mst f the year.

16 4 1.3 Kklass Pheasant (Plate 3) e Like the Bld Pheasant, the Kklass (Genus Pucrasia, Gray l84l), cnsists f nly ne species, Pucrasia macripha, which can be divided 9 int ten (Delacur 1977), r nine (Vaurie 1965) subspecies ranging alng the Himalayan range frm eastern Afghanistan t West-central Nepal, and frm nrth-eastern Tibet t nrthern and eastern China. Their absence frm the eastern Himalayas t western China is difficult t explain, but gegraphic variatin is marked and clinal, especially in the Himalayas. Systematically, Delacur finds them difficult t place, but * believes their similarities t the ther muntain pheasants utweigh their affinity t the ther grups such as Syrmaticus and Phasianus. In Nepal, P.m. nipalensis ccurs frm western Nepal, where i t intergrades 9 with P.m. macripha, eastwards t apprximately 84 E lngitude at least as far as the Madi river, (Rberts 1980), where i t abruptly disappears.» Its lcal name in Nepal is 'Pkras'. Habitat: Resident. Occurs in Nepal between 2100 and 3650m altitude in Rhddendrn, ak, and cniferus frests with heavy scrub and ringal» bamb. In China, P.m. meyeri has been cllected at feet ( m) in Nrthwest Yunnan, whereas P.m. jretiana has been fund at nly 2000 feet (600m) in Anhwei (Vaurie 1965). * 1.4 Himalayan Mnal r Impeyan Pheasant (Plate 4) The Mnal Pheasants, Genus Lphphurus, Temminck 1813, are cmprised > f three species ccurring alng the Himalayan range frm eastern Afghanistan t the muntains f western China. The species fund in Nepal, Lphphurus impeyanus, has the widest range f the three species,» frm Afghanistan thrugh the Himalayas t Bhutan and Arunchal Pradesh; i t has als recently been discvered in Burma (Yin 1970). Its lcal name in Nepal is 'Danfe 1.

17 Plate 3 Male Kklass Pheasant n branch f Rhddendrn barbatum tree. Plate >\ Male Himalayan Mnal digging in a small field in the Everest Natinal Park.

18 1 1 3

19 Habitat: Resident. Occurs in Birch, Rhddendrn, f i r and high ak frest interspersed with pen pastures and steep craggy hillsides with grassy ledges. In Nepal,usually between 2450 and 4600m seasnally with the snwline n mre pen areas than the ther pheasants described. 1.5 Cheer Pheasant (Plate 5) Systematically "the Cheer Pheasant stands alne" (Delacur 1977). The genus Catreus, Cabanis 1851, cnsists f nly ne species and n subspecies. I t is allied t Phasianus but shws features f ther genera including Syrmaticus, Crssptiln and Lphura. The species, Catreus wallichii, ccurs discntinuusly alng the Himalayas frm Durung Galli and Hazara in Nrth-West Pakistan thrugh India t West-central Nepal. Delacur (1977) shws its range extending t the Sikkim brder, but this is incrrect and prbably i t des nt ccur East f the Kali Gandaki valley (Rberts I98O). Its distributin thrughut the length f its range is very lcal and spradic, and the species is included in the I.U.C.N. Red Data Bk (1979) a s a n endangered species. Its lcal name in Nepal is 'Chir'. Habitat: Resident. I t ccurs n very steep craggy hillsides and cliffs brken by narrw grassy and scrubby ledges. Als the adjacent lightly wded ravines and hllws f ak and Berberis. Fund between 1400 and 5000m, usually restricted t abut 2500m in Nepal with slight seasnal migratin. 1.6 Research t Date Apart frm descriptins f sht specimens, the first reprts in the literature cncerning the abve pheasant species were written by British sprtsmen in nineteenth century India. They include authrs such as Hdgsn (1846), P. Wilsn, Capt. Baldwin, A. Hume, and C. Marshall (all in Hume and Marshall 1879)j and Blanfrd (1898). While their reprts

20 Plate 5 Male Cheer Pheasant in captivity. Plate 6 Himalayan Mnal digging area in tussck grassland.

21

22 were unscientific and very subjective, they nevertheless prvide us with imprtant accunts f distributin,abundance, and behaviur, which wuld therwise be lacking frm this perid. The first mre serius study was made by William Beebe, wh wrte the classic 'Mngraph f the Pheasants' ( ) which was largely based n his wn bservatins f pheasants in the wild. This is nt a scientific wrk, but is an excellent treatise which prvides the basis n which mdern research can be carried ut. Since Beebe's time mst bks n pheasants have relied heavily n his wrk, and that f Hume and Marshall (1879) fr "the bulk f their text n Himalayan Pheasants. Such authrs include Baker (1930), Bates and Lwther (1952), Wayre (1969), AH and Ripley (1969), Rberts (1970), Fleming et. al. (1976), and Delacur (1977); althugh the last fur als have integrated their wn bservatins frm the wild. Serius scientific study did nt begin until 1975, when Hwman, Mirza and thers began censusing pheasants in Pakistan (Hwman 1977, Mirza et. al a and b). Since then, a few studies have taken place in Pakistan and India, and have been mainly cncerned with census wrk and re-intrductin f Cheer Pheasants (Severinghaus 1979, Severinghaus et. al. 1979, Gastn 1979, Gastn 1980a, Gastn and Singh 1980). The First Internatinal Pheasant Sympsium in Kathmandu, Nvember 1979 prvided a platfrm fr all current wrkers n Himalayan Pheasants, and its prceedings cntain a number f papers n this subject (Gastn 1980b, Lellitt and Ynzn 1980b, Mirza (1980 a, b and c), Rberts 1980, Sakya 1980, Shrestha and Nepali I98O, Ynzn and Lellitt 1980, Zeliang 1980). Details f avicultural research n the study species can be fund in relevant jurnals such as American Pheasant and Waterfwl Sciety Magazine, Cage and Aviary Birds, Avicultural Magazine, Annual Reprt f

23 the Ornamental Pheasant Trust (nw the Pheasant Trust), and Jurnal the Wrld Pheasant Assciatin.

24 10 r r r QJ t IS) QJ UJ CNI r UJ i O-HsO 6 r O Cxi r v. r 'a. r a 2 \ r- r s CD 1 0 r c c c CM tt-

25 11 2. THE PRESENT STUDY My study f Himalayan Pheasants in Nepal was undertaken as Wrld Pheasant Assciatin Prject Number 103. I t was carried ut in cllabratin with the Natinal Parks and Wildlife Cnservatin Office f His Majesty's Gvernment f Nepal. The prject aims were t study the status and eclgy f highland pheasants in selected areas, t assess the degree f human influence n their habitats and ppulatins, and t make preliminary recmmendatins t HMG regarding an area which culd be cnsidered as a pheasant r 'game' reserve in the future. The prject ran frm April t Nvember and frm March t May I t cncentrated n ne study area, while shrt visits were made t ther areas fr cmparisn. During the curse f the research, certain perids f absence frm fieldwrk were necessary fr visa requirements, the writing f reprts, and ther matters. Fr the duratin f the spring 1979 field seasn, I was jined by Pralad B. Ynzn, M.Sc, Assistant Lecturer in Zlgy at Tribhuvan University, Kathmandu, wh shared the fieldwrk and made a special study f human influence n the pheasants (see Ynzn and Lellitt 1980). In additin, Heather A. Wright, B.Sc., assisted in the fieldwrk in May and June Descriptin f the Study Area and ther areas visited Pipar Pipar is the name f a small hillck at 3325m elevatin n a ridge f the Annapurna Himal, Nrth f Pkhara, Kaski Zne. (See Figures 1 t 5). In this thesis i t refers t the study area arund Pipar h i l l, 28 25'N Latitude; 83 57'E Lngitude. I t is situated n a minr Nrth-Suth riented ridge due Suth f Machhapuchhare (Fish Tail Peak) and separating the Mardi and Seti river valleys draining frm the

26 12 Rc Nir/ G;anqapu>na a 8090 Annapurna l s A 7555 V Annapurna Ut \ -^7219 A Annadurna Suth s f V VVI r 6W3 / \ // Hachha'puchhace «75- Hink ' m- <? I, v» I M,Bhalu Piparx Krchn A Siding Bharbure Dhiprang 2 Lumle Dhumpus\ Ghachk Lachk 1S2Srru \ I Naudanda Suikhet Hyeqja - Figure 2 Regin Nrth f Pkhara shwing study areas(x) Scale 1: POKHARA X'- Phewa Cnturs at 1525 m - Glacier

27 i 13 5?, 53 St 56- PR 1, KALKI MAMRUNtJ : DANDA X Camp ( 'Cam 33*. * US ) U 1463 XBase, 07- pnd '.KORCHON- \ I :»-3684< / 1 I > 06 V J Camp 04 1 f'imu i i 6> 1 I C 52!5* / '-1297/ Km Figure 6 Lcatin f the Study areas in the upper Seti valley- Based n One Inch Survey f India Series

28 14» I 1Q \0> =3 NAMRUNG i I * 08 camp. Ill /camp i -T : *2 t 07 RIPAR A. base camp i 06 5^ Km B i i i i I 0 SCALE 51 Figure 4: Detail f the Pipar area, shwing camp sites, main trails, census pints, and stream systems.

29 4 T Kalki Danda landslide G th ar e a c a ffsw: 10 C 1 ast Gimp Figure l i Dei,ailed map f i.rail at Pipar shwing census pints.

30 Himalayan ridge. The nearest permanent settlements are nearly 2000m belw Pipar in the Seti river valley t the East. There is hwever, ne cattle 'gth' and a number f sheep/gat 'gths' at Pipar. These gths are pen wden huts used by herdsmen fr themselves and their livestck during the summer mnths, when the latter are grazed in the high level pastures r 'kharkas 1. Study was cncentrated n the eastern side f the Pipar ridge in an area f frest, scrub, and pen tussck grassland dissected by small gullies r 'nullahs'. T the Suth and Suth-West f the ridge, slpes are very steep and parts f the frest are almst Impenetrable. The ridge and its upper flanks are cvered with a tugh tussck grass and are grazed by livestck during the mnsn (June t September). In the nrthern part f the area the ridge jins anther, lcally named 'Namrung' and 'Kalki Danda', which runs apprximately East-West. The ttal grund area cvered by the map f Figure 4 is apprximately 16 km Study was carried ut at Pipar frm 2 April t 28 May 1979, 21 September t 24 Octber 1979, and frm 8 April t 3 May Bhalu Bhalu is the name given t the camp (see Figure 2) set up at an altitude f 3150m, 28 26'N Lat., 83 52' E Lng., six km. Nrth f Siding village in the Mardi village. The area is situated n the Eastfacing slpes f a majr ridge separating the Mardi and Mdi river valleys, draining frm the Annapurna range, and cnsists f tussck grassland and adjacent frest. This ridge cnsists f rcky crags and steep nullahs with similar frest t that f Pipar. The alpine meadws are used as kharkas (grazing areas) in the summer mnths. A visit was made t this area between 1 and 9 June 1979.

31 2.1.3 Kumai Kumai is a h i l l at j5260m altitude n the principal ridge West f Pipar running suthwards frm Machhapuchhare. I t lies 3<3km Suthwest f Pipar and apprximately 3-6km. Nrth-West f Mirsa village in the Set! valley (see Figures 2 and 3)- Attentin was cncentrated n the ridge and East-facing slpes arund Krchn (3684m peak), Kumai, and dwn t 2960m. Again, frest is similar in cmpsitin t Pipar and there are numerus gths and kharkas in evidence. A visit was made t this area frm 27 Octber t 9 Nvember Athhazar Parbat Regin This hilly area lies West f the Kali Gandaki river and includes part f the Myangdi river valley in the Dhaulagiri zne f Nepal. The areas visited are situated at apprximately 28 32' N. Lat., 83 22' E. Lng., 70km. Nrth-West f Pkhara. Study was cncentrated n the steep grassy cliffs and adjacent scrubby frest abve Khibang, Dara, and Muri villages at apprximately 2300m altitude (see Figure 6). These areas were visited n trek frm Pkhara between 11 and 25 August 1979, and between 4 and 25 May Climate The suthern slpes f the Annapurna and Dhaulagiri Himal experience sme f the highest rainfall in Nepal fr tw main reasns: a) Heavy spring thunderstrms ccur almst daily as clud builds up n the ridges. b) The h i l l and muntain ranges t the Suth are lw (less than 2500m), s that during the mnsn perid (June t September) rain cluds sweep in frm the Suth-East t precipitate their f u l l lad n the Suth-facing slpes f the main Himalayan ridge. During this seasn, there is almst

32 18 r 7- ra QJ r r ID l_l ra J A i IT3 r t QJ 0 SO r "? -<3 V r \.ra QJ ra re QJ 00 NO r ra. ra / JZ cn j 2? //» QJ / c/ cn r r en II C3 t c! 0* cn r QJ in t t- i Cl QJ i ai ra / cn P cn r CD/ r ad cn r QJ 1 t r QJ ra y m. in in t- QJ SO r rsi rv i _ m ' en r QJ j - ra? U_ O < r aj r l m j

33 19 cntinuus lw clud and drizzling mist at altitudes abve 2100m, limiting the height at which crps can be grwn. Table 1 shws the mean mnthly rainfall at Pkhara (frm Staintn 1972). We made qualitative bservatins f rainfall at Pipar, cmparing i t with that f Pkhara (which is visible frm Pipar). On mre than twenty ccasins precipitatin was nted at Pipar while Pkhara culd be seen clear f any rain. I therefre cnsider that a 30$ increase in the annual Pkhara rainfall wuld be a cnservative estimate f the mean annual rainfall at Pipar, ttalling 4520mm (I78in.). In Spring 1979, precipitatin in the frm f heavy rain, hail, r snw fell n 36 f the 64 days (56$) spent abve 3000m in the field, usually during the afternn r evening. On 25 days (39$), lw clud belw 3350m (fg) was experienced, usually frm late mrning t shrtly befre dusk, but smetimes as early as During this perid, the dawns were usually clear and clud develped during the mrnings, descending as fg n ccasins. Precipitatin wuld ften f a l l during the afternns, but this wuld cease by dusk t give a clear frsty night. In autumn 1979, precipitatin fell n 22 f the 43 days (51$) spent abut 3000m, and lw clud (fg) was experienced n 36 days (84$). The pattern f weather was similar t that f the spring. The persistence f s much lw clud in the autumn field seasn was unexpected, since the pst-mnsn seasn is renwned fr its gd weather. Hwever, i t seems that while the muntains are clear f clud abve abut 4500m, and the valleys are als clear, a belt f clud ften frms at c2800m m resulting in precipitatin and pr visibility between these altitudes. In spring I98O, precipitatin f e l l n 15 f the 26 days (58$) spent abve 3000m, but lw clud was experienced n nly 4 days (15$). Temperature values were nt recrded, but were estimated t vary

34 20 TABLE 1 Rainfall at Pkhara $ Mean mm in January February March April May June July August September Octber 170 O. 1 Nvember December Ttal: 3477mm 136.9in TABLE 2 Cmparisn f Vegetatinal Classificatins f Staintn (1972) and Dbremez et. al. (1971) in the Suth Annapurna/Dhaulagiri regin. Altitude (m) BETULA UTILIS FOREST STAINTON MOIST ALPINE SCRUB J 1RHODODENDRON [FOREST UPPER [ TEMPERATE 1 MIXED BROAD LEAVED FOREST QUERCUS LAMELLOSA FOREST DOBREMEZ AND JEST ALPINE MEADOWS RHODODENDRON- BIRCH-FIR FOREST RHODODENDRON- OAK-TSUGA FOREST (QUERCUS SEMECARPIFOLIA) QUERCUS LAMELLOSA FOREST BAMBOO "FACIES"

35 21 between -3 C (night) and 20 C (midday) during Spring 1979 and Majr Vegetatin Types The vegetatin f this regin f Nepal has been surveyed by Dbremez and Jest (1971)> and Staintn (1972). They used similar frest zning systems (see Table 2), but thse f Staintn best reflect the vegetatin types encuntered in areas t 2.1.3, Suth f the Annapurna Himal: Rhddendrn Frest Distinguished frm 'upper temperate bradleaved m mixed frest' by dense grwths f Rhddendrn arbreum with ther species such as R. barbatum, Acer, and Srbus,scattered amngst them. Frest at Pipar was predminantly f this type at this altitude, with Berberis and Viburnum frming scrub at edge. Betula u t i l i s Frest Often n East- and Nrth-facing slpes and 32O0-4O00m frming a band at the upper frest limit. Cnsists f Betula u t i l i s, R. campanulatum, Berberis, and Viburnum. Mist Alpine Scrub Uppermst frest is B. u t i l i s with R. camp- Treeline 4500m anulatum. Tusscks f grasses and sedges, and arct-alpine species predminate. Ringal bamb (Arundinaria sp.) is prbably the mst imprtant 'shrub' cmpnent in these frests. I t is fund as large stands and scattered thrughut the frest between 2000m and the tree line. The abundance f this genus and the high prprtin f Rhddendrn frest n the suthern slpes f the Annapurna Himal may be due t the very high rainfall in this area. Staintn (1972) suggests that "heavy rainfall n steep rck slpes leaves nly a thin layer f sil which is sufficient fr the shallw rted Rhddendrn (and bamb), but nt fr larger species". Grubb (1971) suggests that the types f trpical frest grwing at certain altitudes can be gverned by the presence f lw clud (fg) causing a higher sil water cntent and a decrease in the mineralizatin f rganic matter. Certainly in Nepal, Rhddendrn frest cver shws psitive crrelatin with the incidence f highest rainfall and clud cver (central and far East Nepal).

36 22 Cnifers (mstly Abies spectabilis) are uncmmn in the study areas, and Staintn (1972) suggests they have been "driven ut" by the high rainfall which is equivalent t the wettest areas f easternmst Nepal. Vegetatin f The vegetatin in the Athhazar Parbat regin, Suth f the Dhaulagiri Himal is, belw 2700m, significantly different t that f the area Suth f the Annapurna Himal. Accrding t Dbremez and Jest (1971) this is due t the lesser rainfall in the Dhaulagiri area, except n the muntain summits. The altitude zne at which Cheer Pheasant were present abve Muri village ( m), lies in the Dry Oak zne (Quereus lanuginsa) f Dbremez which is equivalent t the Quercus lamellsa zne f the Suth Annapurna regin (see Figure 7). The Dry Oak zne includes R. arbreum, Symplcus spp., and Lynia sp. Abve this zne the frest types are similar t thse f the Suth Annapurna regin, but were nt visited in this study.

37 m 3950m Abies-Bet ula- Rhddendrn 3200m Quercus 2700m A s e m e c a r P'^ ^ a Quercus lanuginsa 2100m Alpine Meadws Abies-Befrula- Rhddendrn Quercus semecarpiflia Quercus tamellsa MYANGDI KHOLA Pinus rxburghii ScNima Casfanpsis Ingslhardtia Sbna rbusra F i gu r-> 7 : Vegetatin stages Suth f the Annapuma and Dhaulagirl massifs (frm Dbremez and Jest 1971) ^ Bld Pheasant Himalayan Mrial VEGETATION ZONES OF STAINTON (1972) = i LU X Betula utilis Alpine Scrub J, Kklass Pheasant ^ddendrn 2000= Satyr Tragpan Quercus lametlsa Figure- 8: Altitude ranges f Pheasants,

38 3. METHODS Ji. 1 Fieldwrk Trapping Capture, tagging, and releasing was nt carried ut n the study species fr a number f reasns. First, the ppulatins I was studying were relatively small in size, and the pssibility f mrtality r injury resulting frm the "types f snare used by the lcals (see sectin 8.1.2) was t great. Any ther methds f trapping were cnsidered time-cnsuming and impractical. Secndly, in such a shrt study i t was nt cnsidered that the returns frm retrapping r bservatins f marked individuals wuld be wrthwhile. Thirdly, we aimed t set a 'gd example' t the lcal peple (wh hunt the pheasants fr fd) by bserving the birds, nt snaring them. Finally, in small, f a i r l y sedentary ppulatins, snaring culd disturb the birds and result in even greater avidance f humans. Hwever, t test the efficiency f the snaring methd, tw traplines were set up in Octber 1979 fr fur days, resulting in the capture f ne Satyr Tragpan alive. Als, in April 1980 a prter set up a trap-line and snared ne Bld Pheasant dead. (see sectin 8.1.2) Data Cllectin This was carried ut using a methd based n that f Severinghaus (1977), by Direct Encunter f pheasants. In a l l the areas visited, the authr and ther fieldwrkers (P. Ynzn and H. Wright in 1979; Ram Kaji Mangar and Nema Chtta Sherpa in 1979 and 198O) wuld search the study area and recrd a l l visual and auditry encunters with pheasants n Direct Encunter cards (See Appendix 2) based n thse f

39 2 5 Severinghaus. A daily recrd f weather, rugh areas searched, ntes n ther species, etc. was als kept. In searching fr pheasants I (and the ther bservers) usually made use f the few small trails in the areas (used by herdmen and livestck) but als searched areas where there were n such t r a i l s. The chances f sighting pheasants were greatly increased i f the bserver was silent and incnspicuus. Walking thrugh trail-less frest created substantial nise by the cracking f branches and leaf l i t t e r underft. Despite this, the paucity f trails in the areas made f f - t r a i l walking frequently necessary. A survey f plant species cmpsitin in the study areas was nt carried ut due t lack f time, lack f facilities fr rapid plant identificatin in Nepal, and the fact that a btanist was nt available t jin me in the field Use f Hides A number f hides were cnstructed in the Pipar study area fr lnger bservatins f pheasant species at clse quarters, and fr phtgraphy, recrding and playback. Hides were built f the surrunding materials in the frest (e.g. dead wd, mss) at sites where pheasants had been nted t ccur. The area adjacent t the hide was nt baited. Hides were usually entered befre dawn and ccupied fr up t three hurs after sunrise. A ttal f apprximately 65 hurs was spent in hides Phtgraphy Phtgraphs f pheasants, their habitats, and ther species were taken with a Pentax KM camera, 200mm Vivitar lens, and a 2x Vivitar Telecnverter. In the case f the Bld Pheasant, phtgraphs were taken directly withut the use f a hide by apprach t within 3m f the birds, due t their tame nature. Other species were phtgraphed frm

40 hides at Pipar shrtly after sunrise. Kklass and Satyr Tragpan were decyed t apprach the hides by playback f their calls. Himalayan Mnal were als phtgraphed by stalking while feeding in ptat fields in the Everest Natinal Park, East Nepal. Cheer Pheasants were nt ph tgraph ed Tape Recrding A l l five species were recrded in the field using a Uher 400 Reprt-L recrder and a Grampian DP 6 micrphne, usually munted in a fibreglass 22-inch (508mm) parablic reflectr. Recrded n EMI and Agfa Lng Play tape at 7"? inches per secnd. Playback f certain calls f Kklass, Satyr Tragpan and Bld Pheasant was carried ut and in sme cases resulted in thse species appraching the recrdist Infrmatin frm lcal villagers Infrmatin cncerning the presence r absence f pheasant species was gained frm lcal villagers, especially shikaris (hunters), and mst identificatin was thught t be reliable, since pheasant psters in Nepali had been distributed t villages in the Seti and ther valleys (Savage 1978). Regarding numbers hwever, the villagers' infrmatin was unreliable, and there was a general tendency t t e l l the questiner what he wanted t hear rather than the truth. This was especially prevalent near Muri and Khibang in the Athhazar Parbat. Therefre, unlike Severinghaus (1977)* Reprted Encunters frm the questining f lcals are nt included in the data, except fr a few qualitative bservatins, and in Sectin Census Fr methds see sectin 4.3-!

41 3.1.8 Faeces Cllectin Fresh pheasant drppings were cllected where encuntered in 1979 and preserved in frmalin slutin. I t was cnsidered that Bld Pheasant and Himalayan Mnal drppings were easily identifiable in the field, the frmer by their small size and cylindrical shape, and the latter by their dull dark green clur and their presence n Mnal digging areas (see Plate 6). Kklass and Satyr Tragpan drppings were d i f f i c u l t t t e l l apart and n distinctin was made in Faeces f a l l fur species were analysed fr parasites by Mrs. J. B. Hwman. In 1980, further cllectins were made f pheasant faeces, and in sme cases where the birds were bserved, distinctin culd be made between Kklass and Tragpan faeces. All samples were baked in the sun fr several days until cmpletely dry, after which they were stred in sealed plastic bags, a methd recmmended by Green (1978). A number f drppings f each species were analysed by Dr. J. Butterfield in January N drppings f Cheer Pheasant were cllected in either year Nest Search Searches fr nests f pheasants were made in bth years. Only ne nest was fund, that f a Tragpan in and this was by the accidental flushing f the hen bird. Intensive searching by many peple may be mre effective, but time-cnsuming. 3-2 Labratry Wrk Faecal Analysis - Dr. J. Butterfield kindly analysed samples f 6 drppings frm each pheasant species by separating the cntents in alchl slutin and examining them thrugh a micrscpe. Snagraphic Analysis f Vcalizatins - Snagraphic analysis f

42 1» recrded calls was made n a Kay Electric 66l-B Snagraph with amplitude cntur display and scale magnifer. All Snagrams except thse f the Kklass Pheasant were made using the narrw (45Hz) f i l t e r.» Unless therwise stated, Snagrams shw frequency (0-8 khz) n the rdinate and time alng the abscissa. On the Kklass snagrams the wide band f i l t e r was used, and relative amplitude is indicated by cnturs 6dB apart, the darkest 3.3 Questinnaire being the ludest. A questinnaire n aspects f pheasant vcalizatins and behaviur (see Appendix 3) was sent t f i f t y - f i v e members in pssessin f any f the study species in August Eleven replies were received, and the results are discussed in the relevant sectins.

43 4. HABITATS AND POPULATION DENSITIES 4.1 General Habitats in the Suth Annapurna Regin WOODLAND a) Belw 3000m Mixed frest with bamb Primary tree species Rhddendrn arbreum and Quercus spp. Secndary species Srbus sp., Acer sp»> Symplcs spp., Litsea spp. and ther unidentified species. Canpy height 15 t 20m. Shrub layer cnsists largely f ringal bamb, Arundinaria spp., which can frm large stands dminating areas f frest, especially n Suth- and East-facing slpes. Grund cver: mstly leaf l i t t e r, with a few lw-light intensity species such as Arisaema spp.: ther unidentified. b) m Mixed Rhddendrn Frest (see Plate 7) Primary tree species Rhddendrn arbreum, Acer sp., R. barbatum. Secndary species Srbus sp., Betula u t i l i s, R. campanulaturn, and ther unidentified species. Canpy height up t 15m. Shrub layer cnsists f ringal bamb scattered amngst frest, nt grwing as large stands; Viburnum sp., Berberis sp., and ther unidentified species. Grund cver: herbaceus species cmmn during and after mnsn seasn, largely unidentified. Leaf and branch l i t t e r prminent. c) m (treeline) Betula Frest (see Plate 7) Primary tree species B. u t i l i s, R. campanulatum, R. barbatum. Secndary species Alnus sp., thers unidentified. Canpy height up t 10m. Shrub layer cnsists f R. campanulaturn, bamb, Viburnum, Berberis spp. and ther unidentified species. Grund cver: herbaceus species dense during and after mnsn seasn, including Primula denticulata, Gentiana spp., Anemne sp., and thers largely unidentified.

44 a»»» Plate 7 Rhddendrn/Betula/Berberls habitat in the census area, Pipar Plate 8 Tussck grassland habitat, Suth side f Pipar H i l l

45 : i! I ;

46 31 FOREST SCRUB 3200m t treeline (see Plate 7) Primary species Berberis asiatica, Viburnum grandiflrum, R. barbatum, R. campanulatum. Shrub height up t 7m. Grund cver: grasses and herbaceus species such as Arisaema g r i f f i t h i. Primula denticulata, Impatiens spp., et. al. Much mss n tree trunks and grund. Deep leaf l i t t e r. TUSSOCK GRASSLAND 3000 t abve 4600m Livestck Pastures and Alpine Meadws (see Plate 8) Occurs n Suth-facing slpes dwn t 3000m where livestck grazing has taken place. Abve the treeline, alpine meadws f grassland predminate t the limit f vegetatin (^900m in places). With the exceptin f the Himalayan Pppy, Mecanpsis paniculata, and Gentiana spp., the grasses and ther plant species in this zne have nt been identified. 4.2 Habitat Preferences f Pheasants Bld Pheasant Fr this species, 39 ut f the 53 Direct Encunters (74$) were made in the "wdland" and "wdland/scrub" categries f the Direct Encunter cards (see Table 4). Primary tree and shrub species were nted, and in every case cnsisted f a cmbinatin f Rhddendrn, Betula, Berberis, and Viburnum frming frest and adjining scrub; in 19 cases (36$) ringal bamb was als in evidence. In the frest, the tree height ranged frm 4 t 12m and the canpy was up t 70$ clsed. Understry vegetatin was usually lacking except fr dense stands r mre scattered areas f ringal bamb in varius stages f grwth frm seedling t maturity depending n psitin. All the ringal bamb (Arundinaria) species have a very slw grwth rate and lng l i f e cycle, taking up t 50 years t mature, flwering nce, and then dying. The

47 » 32» TABLE 3 Summary f Altitudinal Distributin f Pheasants Direct Average Lwest Highest - Encunters Altitude (m) Altitude (m) Altitude (m) Bld Pheasant Spring Autumn Spring OO Satyr Tragpan Spring Autumn Spring Kklass Pheasant Spring Autumn Spring Himalayan Mnal Spring Autumn Spring Cheer Pheasant Spring

48 33 c rh X)? is c rh CQ 10 CD OJ p x: CM O w 0) H U 0 w M (1) P G O P i 'l-i Q \ rh w c O CA O X) (D bo co -P a CD C 0> u 0) <D PH c cfl rh T) O O MA OJ KA -p w 0) PL, CM O T3 a ih T3 O OO >> M 3 EH rh p> c c c a g bo CD c JC U PH EH in >5 c X) < rh rh s. p E rh H m c W

49 34 result is large areas f bamb either dead, r in different stages f maturity. In the frest, grund vegetatin was scarce in spring, and cnsisted largely f ferns and small shrubs with much mss. Leaf and branch l i t t e r was deep with numerus rtting fallen trunks. Snw drifts were present in shaded areas until late May. In the pst mnsn seasn, grund cver was greater with herbaceus species present in additin t ferns and shrubs. Rck expsures f gneiss and schist were numerus, smetimes frming crags. The frest where mst bservatins f Bld Pheasants were made was grwing n East-facing slpes with gradients ranging frm f l a t ridge tps t 50 slpes, but mstly bservatins (19$ f sightings) were made in pen scrub. This cnsisted predminantly f Berberis, Viburnum, and small R. campanulaturn bushes up t 5m in height. There was n canpy, and grund vegetatin cnsisted f grasses, msses, and a few herb species. Mst f the scrub was a result f heavy grazing by livestck during the mnsn mnths, s that grass was shrt and clearings were numerus. Pur encunters were made in scrub with grassland, but n bservatins were made in pure grassland which extended dwn t >000m n the Suthfacing slpes f Pipar, r in the alpine meadws abve the tree line. Altitude: in spring 1979> Bld Pheasant were encuntered between 3120m and 3510m, at an average height f 3260m frm 34 bservatins. In autumn 1979> they were nly encuntered n tw ccasins, but drppings were fund at Kumai and Krchn as lw as 3050m and as high as 3650m. In spring I980 Bld Pheasants were fund between 3200m and 3600m at an average height f 3350m frm 14 bservatins. Data are summarised in Table 3 and Figure 8. Slpe and aspect: bth f these tpgraphic features influence vegetatin types and therefre indirectly influence pheasant distributin. Of the

50 fifty-three Direct Encunters, nine {YJ%) were n ridge tps r flat land, and 4l {17%) were n slpes f less than 40. Of these, eighteen (3^%) were n slpes f between 11 and 20. As regards aspect, Bld Pheasants were almst always fund n slpe with an easterly cmpnent f aspect. Thirty (57$>) ccurred n Eastfacing and nly ne (2%) was fund n slpe with a westerly aspect. Data are summarised in Tables 5 and 6. DISCUSSION The Direct Encunters with Bld Pheasant suggest that i t prefers frest and adjining scrub n East-facing slpes between 3050m and 3600m: and is assciated with the genera Rhddendrn, Betula, Berberis, and Viburnum with stands f Arundinaria. Hwever, a certain amunt f bias is present in the results due t a number f factrs, including: 1. N fieldwrk was carried ut belw 3000m r abve 4100m. 2. N fieldwrk was carried ut during the mnsn seasn in July and August, r during the winter (December t March). 3. A high prprtin (apprximately 70%) f the hurs spent searching fr pheasants was carried ut between 3200m and 3650m. 4. Distinct differences between 'frest', 'frest/scrub', and 'scrub' are d i f f i c u l t t define and are therefre subjective. 5. Field searching time was significantly less in thick tangled frest n steep slpes (generally West-facing) than in mre pen frest scrub r grassland n gentler slpes. 6. Althugh times were nt recrded, mst pheasant searching tk place in the mrning and mid t late afternn, when the birds appeared t be mre active. 7. Birds are mre d i f f i c u l t t lcate in certain habitats than in thers, (e.g. thick frest) and can be warned f the searcher's apprach mre quickly.

51 w u CD P C 3 a) p 0) H Q OJ (0 p EH ^R. >R *R ^ >R i KA OD OD O A 1^ r\ H ft t-- + > a> H P 00 rh 00 p E-i ^R. ^ >A m V< c- c\j OJ 0 ft ft rh rh OJ OJ rh 00 CM K\ (\l (J\ 4" ft h rh rh OJ CXI rh P CI 10 t c0 CD Si CD a. O rh M w <M O 0!^ w a> rh bo a ca p c CD ^ 0) t CM CM a> IH p c C 0 O 0 t p C p CD 0 c CD c 5-i CD c-l.c 0 OH CM O ft LC\ CP rh V c C -p CD 0 U En u 0 0 -p 0 C ca 0) in W r^ Si PH O O < rh EH CP P rh rh! (X, O O CVJ I -ijft I OJ ft A in -3- OJ OA tn 00 CD O OJ O O O O P O OJ ft -3- C0 rh I I I I I OJ ft -=1- al c 0 s c >. co O O O O 0 rh O (~> O r-> ri P O CA) ft JS X E rh c H rh 1 rh rh rh rh tr> O rh CM ft A (0 CD p c -p 0) -P O EH C Q. O ho u 00 m m rh OJ rh OJ LPi ft ft rh p j ft m rh I I I I x B rh I rh rh 1 f 1 I A C feorhojft^-" td

52 TABLE 6 Occurrence f Pheasants n Slpes f different aspect Bld Pheasant 53 Direct Encunters N-NE 0-8 E-SE 30-3 S-SW 2-0 W-NW 0-1 Expsed (n aspect) 9 Satyr Tragpan 60 Direct Encunters N-NE 0-8 E-SE 30-4 S-SW 6-1 W-NW 5-1.Expsed (n aspect) 5 Kklass Pheasant 29 Direct Encunters N-NE 0-5 E-SE 23-0 S-SW 1-0 W-NW 0-0 Expsed (n aspect) 0 Himalayan Mnal 108 Direct Encunters N-NE 3-2 E-SE 13-8 S-SW W-NW 6-0 Expsed (n aspect) l8

53 Having taken these factrs int cnsideratin hwever, the abve statement f Bld Pheasant habitat preference in this area s t i l l remains valid in my pinin. I t cmpares favurably with the species' habitat described in the literature except fr the absence f cnifers, which is thught t be due t the high rainfall (sectin 2.3). Delacur (1977) states that they frequent " f i r and rhddendrn wds and the scrub abve Wherever bamb is fund, they are partial t i t ". Accrding t A l i and Ripley (1969) they ccur in "steep pine frest, dwarf rhddendrn, and dense ringal bamb". Bth these authrs als state that the Bld Pheasant lives at higher elevatins than any ther Himalayan pheasant, and this may well be true in areas where the treeline is at a high altitude such as eastern Nepal and China. Hwever, I never encuntered the species living ut, f the frest, and since the treeline is at apprximately 3650m in the study areas, I d nt believe i t t live abve this height in central Nepal. The Himalayan Mnal hwever, des live abve the treeline and ccurred at cnsistently higher altitudes than the Bld Pheasant in the study areas. This view is supprted by recent bservatins made by Majr I. Grahame (in Delacur 1977). Regarding the apparent preference fr slpes with an easterly facing cmpnent, I cnsider that this is partly due t the greater number f man-hurs spent searching East-facing slpes, and partly due t the abundance f mre suitable habitat n these slpes Satyr Tragpan Of sixty-ne Direct Encunters f this species, fifty-seven (93$) were made in the 'wdland' and 'wdland/scrub' categries f the cards. Of these, f i f t y (82$) were made in pure wdland (see Table 4). Primary species encuntered were R. arbreur, R. barbatum, Betula, and Arundinaria. Other genera such as Quercus ccurred mre cmmnly

54 39 at lwer altitudes. The height f canpy trees in the frest ranged frm 5m t 20m, but averaged 8m t 12m, and the canpy was frm 30$ t 100$ clsed depending n tree height and age. Understry vegetatin was scarce except fr ringal bamb, here cnsidered as a primary species in the frest. Grund vegetatin in the frest was restricted t ferns, msses, and a few herbaceus species (e.g. Araesmia sp.) in spring. Like Bld Pheasant habitat, with which i t has much in cmmn, leaf and branch l i t t e r was thick, and rck expsures numerus. Damp areas and small streams were usually present in Tragpan habitat. Seven bservatins (12%) f Tragpan were made in 'wdland/scrub 1 habitat categry like that described in 4.2.1, while nly three bservatins (5$) were made in pen scrub. N bservatins f Tragpan were made in pen grassland r alpine meadws except the ccurrence f a nest in tussck grass (see sectin 7.2.3). Altitude: The Satyr Tragpan has a much wider and lwer altitudinal range than the Bld Pheasant, but this is nt very clear frm the Direct Encunters. The study area was at the upper end f the Tragpan's range, and i t was encuntered up t 3550m, but at an average height f 3150m. Hwever, the species was n ccasin heard and seen dwn t 2230m when I was n trek t r frm the study areas. Data are summarised in Table 3 and Figure 8. Slpe and Aspect: Of the sixty Direct Encunters, five (8$) were n ridge tps r f l a t land and 44 (74$) were n slpes f less than 40. Of these, sixteen (27$) were n slpes f With regard t aspect, thirty Encunters (50$) were n East-facing slpes, but this data is cnsidered t be very biased (see discussin). Data are summarised in Tables 5 and 6.

55 40 DISCUSSION Since the study areas lie within nly the tp half f the dcumented altitudinal range f the Satyr Tragpan, any results shuld be examined with this in mind when cnsidering the habitat as a whle. The Direct Encunters indicate that i t prefers mixed Rhddendrn frest up t 3500m cnsisting f R. arbreum, R. barbatum, Betula, and Arundinaria. The factrs f pssible bias listed fr Bld Pheasant apply equally well t this species. Althugh the majrity f sightings were made in the eastern part f the study area Nrth f the campsite (see Figure 4), this is entirely due t the greater number f man-hurs spent searching this area f relatively easy terrain. When ther areas such as the suthern and western slpes f the Pipar ridge were surveyed at dawn in an attempt t census by crw-cunt methds (see sectin 4.3*3), up t six calling male Tragpans were heard in Mixed Frest with bamb and Mixed Rhddendrn frest frm 3200m dwnwards ut f hearing range. These areas f frest were nly searched n three ccasins due t their extreme density and steep slpes, but i t is clear that their habitat supprts sizeable numbers f Satyr Tragpans. In additin, this species has been heard frm Thulkhbang (2300m), which lies n the t r a i l frm the Seti valley t Pipar in Quercus lamellsa frest (Staintn 1972) n steep slpes (up t 50 ). The presence f Tragpans n these slpes demnstrates the much wider range f habitat, slpe angle and aspect that the species lives in than is shwn by the Direct Encunters. Since the Satyr Tragpan exhibits a fairly wide altitudinal range (up t 1350m), this means that they live in a wide range f habitat znes frm Quercus lamellsa frest thrugh Rhddendrn frest t Betula frest. The descriptins f habitat in the literature are there-

56 41 fre necessarily vague: "wded muntains f the Central and Eastern Himalayas" (Delacur 1977); "damp evergreen frests and bamb grves" (Fleming, et al 1976). Ali and Ripley (1969) hwever, are mre precise "ak, dedar, and rhddendrn frest in 'khuds' n steep hillsides with scrubby undergrwth and ringal bamb", but they give an unusally high figure f 4250m fr the tp f the Tragpan's range. In cntrast, Delacur (1977) states that i t "nly ccasinally enters the zne f the Bld Pheasant". Hwever, in the study areas these tw species verlapped in altitudinal range by up t 600m (maximum at Kumai in Octber 1979) - see Figure Kklass Pheasant Twenty-nine Direct Encunters were made f this species, and twenty-fur (83$) were made in the 'wdland' and 'wdland/scrub' categries. Of these, thirteen (45$) were in 'wdland/scrub' (see Table 4). Primary species in this categry were R. arbreum, R. barbatum, R. campanulaturn, Berberis, and Viburnum. In eight (28$) f the Direct Encunters ringal bamb was nted. The tree height in this frest and scrub ranged frm 5m t 15m, averaging 8m t 10m, and the canpy was frm fully pen t 75$ clsed. Understry vegetatin cnsisted f ringal bamb in the frest and Berberis bushes in the scrub. Grund vegetatin in the frest, cnsisted f ferns, msses, and grasses with a few herbaceus species, the density depending n canpy cver. Leaf and branch l i t t e r was deep. In the scrub areas, grund vegetatin cnsisted principally f grasses and mss during spring. Rck expsures were cmmn. Five bservatins (17$) f Kklass were in pen scrub/ grassland similar t that described in N Kklass were encuntered in pure pen grassland. Altitude: Frm the Direct Encunters, this species was recrded between 2700m and 3300m, which in this area appear t be the limits f

57 i t s altitudinal range. The study areas were at the upper end f this range. Data are summarised in Table 3 and Figure 8. Slpe and aspect: twenty-tw f the twenty-nine Direct Encunters (76$) were made n slpes between 10 and 30. N bservatins were made n f l a t land r ridge tps and nly ne Encunter was n a slpe f mre than h0. With regard t aspect, a l l bservatins were made n East- r Nrth-East-facing slpes. Data are summarised in Tables 5 and 6. DISCUSSION Direct Encunters with Kklass indicate that i t prefers Rhddendrn Frest and adjacent Berberis/Viburnum scrub with ringal bamb n Eastfacing slpes between 2700m and 3300m. The factrs f pssible bias listed fr Bld Pheasant shuld again be taken int cnsideratin. The abve habitat descriptin, while typical f the study areas, seems atypical fr the species when cmpared t Kklass habitat described in the literature. Severinghaus (1979) prvides a gd review f previus wrk n the Kklass and describes their general habitat as being "mixed hardwd-cnifer frests with gd under-grwth". The Kklass he studied in Pakistan lived predminantly in mixed cnifer frests with dense understry, but nt scrub. Hwever, Gastn (1979) reprted frm Dachigam Sanctuary, Kashmir, that belw 2100m "Kklass were cncentrated in Parratipsis scrub, cniferus frest, particularly where mixed with Indigfera scrub and mixed bradleaved/cniferus frest". Gastn (pers. cmm.) cnsidered that Parratipsis scrub structure bre a clse resemblence t the Berberis scrub in the Nepal study areas. Bth Severinghaus and Gastn were studying Kklass at lwer altitudes ( m and l m respectively) than thse in the Pipar study area. Like the ther tw species, Kklass appear t shw preference fr East-facing slpes frm analysis f the Direct Encunter cards. Hwever,

58 unlike Tragpan, this preference may be mre real than apparent since nly ne Kklass was heard frm Suth- and West-facing slpes f Pipar in 1980, and nly ne was seen there in At Kumai, in Octber 1979» Kklass were heard calling frm bth western and eastern slpes f the ridge. The apparent preference at Pipar fr East-facing slpes is prbably due t the scarcity f Berberis scrub n Suth- and Westfacing slpes. Severinghaus (1979) nted that Kklass ccurred n slpes frm nullah bttms t ridgelines "but preferred the upper slpes and ridgelines". Other wrkers have made different bservatins, 'Pine Martin' (1910) and Bates (1936) fund Kklass mainly in nullah bttms, while Hume and Marshall (1879) and Beebe (1922) f e l t that Kklass preferred upper slpes. A l l my wn Kklass bservatins were made n upper h i l l slpes, but factrs such as habitat availability, seasnal mvements, and disturbance are likely t affect the birds' distributin n a hillside Himalayan Mnal One hundred and eight Direct Encunters were made f this species, f which seventy-ne (66%) were made in the 'grassland' categry. 'Grassland/scrub' and 'grassland/wdland' accunted fr a further twenty-tw bservatins (19$), while nine (8$) were made in pure wdland. (See Table 4). Primary species n the grassland were largely unidentified but cnsisted principally f a tussck grass. Herb species included Mecanpsis paniculata and Primula spp. Species cmpsitin presumably varied with altitude but n details are knwn. Scattered Berberis and Viburnum bushes and Rhddendrn trees r shrubs were ften present. Adjacent 'wdland' and 'scrub' habitats are as described previusly.

59 44 Altitude: this was the highest living species f pheasant studied, ccurring at an average altitude f 3300m t 3400m, and ranging as high as 4000m and as lw as 2800m. Data are summarised in Table 3 and Figure 8. Slpe and aspect: f the 108 Direct Encunters, eighteen (17$) were made n ridge tps, while frty-three (40$) were made n slpes f 20 t 40 ; seventeen (16$) were bserved n slpes greater than 40. There was an apparent preference (41$) fr slpes f a Suth-facing aspect, which is discussed belw. Data are summarised in Tables 5 and 6. DISCUSSION Frm a l l the bservatins made in this study, Himalayan Mnal appears t favur a habitat f pen grassland adjining frest and/r scrub at and abve the tree-line. While the bird clearly des nt spend a l l its time n the pen hillside, i t is ften very easily visible in such terrain, hence the number f sightings n grassland rather than in the frest. In additin t direct bservatins n grassland, the birds' feeding habits, which invlve digging fr rts and frming numerus digging areas, present further evidence f the species' preference fr this habitat. But as with the ther species described, there is a certain degree f bias in the methds f fieldwrk, and the factrs listed in Discussin f Bld Pheasant shuld be cnsidered. In additin t the habitats described abve, I have als seen the species in mixed Abies frest, pen hillsides, and cultivated fields in the Khumbu valley, Everest Natinal Park, East Nepal. Delacur (1977) states that "The birds frequent pen frests, particularly thse f ak, birch, and rhddendrn, with pen glades and grassy slpes, where they feed". Ali and Ripley (1969) refer t "high ak, rhddendrn, and dedar frest interspersed with pen glades and

60 sheep pastures, and precipitus hillsides". Gastn (1979) at Dachigam in Kashmir bserved the species in "mixed bradleaved/cniferus frest, pen Kail Pine frest" and "amng rcky utcrps in grassland" frm m. Perhaps even mre than the Bld Pheasant, this species is said t mve up and dwn with the snwline. Certainly at Pipar in September/ Octber 1979, when the snwline was at abut 5200m the Mnal was nt seen belw apprximately 4300m until mid Octber, when the birds presumably began t mve dwn frm the high alpine meadws and pastures. Ali and Ripley (1969) recrd the Mnal ascending as high as 5000m in summer. Frm the Direct Encunter Recrds, the Mnal shwed a preference fr the Suth- and Suth-West-facing slpes f the Pipar and Kalki Danda ridges. The bulk f the fieldwrk was cncentrated n these ridges, which cnsisted f frest and scrub n the nrthern and eastern sides, with pen grassland and precipitus slpes n the suthern and western sides. The Mnal's preference fr the latter habitat largely explains their a f f i n i t y fr Suth- and West-facing aspects General Discussin A summary f the ccurrence f a l l fur species in brad habitat categries is shwn in Table 4. Cmparing the apparent habitat preferences f the study species, i t appears that while there is a cnsiderable degree f verlap between Satyr Tragpan, Bld and Kklass Pheasants, the Tragpan was mainly encuntered in pure wdland (82$ f Encunters), the Kklass was nearly equally split between wdland/scrub (45$) and wdland (38$), and the Bld Pheasant was fund mainly in wdland (51$), and less in wdland/scrub (23$). The primary tree and shrub genera f Rhddendrn and Berberis were cmmn

61 t a l l three species, in additin Betula and Viburnum were cmmn t tw species. The Himalayan Mnal stands apart frm the ther three species in that i t prefers grassland (66$) and was rarely fund in pure wdland (8$). Since three and smetimes fur species ccur in a superficially similar habitat, i t is prbable that each has its wn niche which is determined by i t s fd and feeding habits. This is discussed in sectin 5.2 Figure 8 shws the cmplete altitude ranges f the fur study species in the Suth Annapurna area. While the Himalayan Mnal pssesses a brad altitudinal range ( m) in the upper part f the scale, and the Saytr Tragpan shws a brad range ( m) in the lwer part f the scale, the Bld Pheasant ( m) and Kklass Pheasant ( m) exhibit mre restricted ranges within the ranges f the tw frmer species. In additin, frm m a l l fur species verlap during the spring seasn (April t June). I t seems therefre, that the habitats at this elevatin,which lie within the Rhddendrn frest and lwermst Betula u t i l i s frest znes f Staintn (1972), and within my brad categry f Mixed Rhddendrn frest, must be particularly rich in ptential fd at this seasn. The areas f frest and alpine meadws abve this zne are mstly s t i l l under snw until late May, s that fd is nt available. The tw species fund in the uppermst habitats, Bld Pheasant and Himalayan Mnal, are therefre fund just belw the snwline in spring. Belw the zne f verlap, fd is readily available at this time, but while the Tragpan and Kklass d ccur dwn t their lwest limits, they are als fund in the rhddendrn frest and scrub up t 3300m. Inter-specific cmpetitin fr fd is discussed in sectin 5-2.

62 Ppulatin Densities f Pheasants Census Methds Techniques fr censusing pheasants and ther game birds have been develped in Nrth America and Eurpe t a high degree f sphisticatin. A number f techniques are in general use, and, when this study was planned, a review f the pssible methds available was undertaken. Severe lgistical cnstraints became apparent when planning t census pheasants in the Nepalese Himalayas, as cmpared with surveys in Nrth America. These include the very steep tpgraphy, absence f rads, scarcity f trails, the small number f persnnel available, and the lack f human settlements abve 1800m. Techniques reviewed fr pssible use in the study areas included the fllwing: Lincln Index - Census is by nting the rati f marked t unmarked adults. Large scale capture f adults thrughut the study area is required. Birds can be tagged with back-tags, ringed, and partly dyed if necessary. Surveys are then carried ut: the first immediately after tagging, and subsequent nes at fixed intervals. The prblems with this type f census are immediately apparent: it is very time cnsuming and wuld be difficult t execute ver muntain frest terrain, tagged birds wuld be very difficult t see in frest, and the rati f tagged t untagged birds needs t be 1:5 r higher (Bergerud and Mercer 1966). Direct Cunt (a) - The area is systematically censused using line transects and dgs t lcate cveys/individual birds. This methd again relies n the tagging f birds and s presents the same prblems as the

63 Lincln Index in additin t laying a number f straight lines thrugh frest n steep slpes. Direct Cunt (b) - One r mre transects are made thrugh a study area attempting t cver a l l vegetatin types. The bserver ntes a l l birds disturbed within a fixed distance frm the line. This methd was feasible in Nepal fr nn-vcalizing r prvcalizing species, but analysis f the results needs care (Gastn 1980a). King Strip Census - Invlves walking alng a number f systematically spaced line transects recrding the number f birds and the distances at which they flush frm the bserver. Prblems presented are similar t thse stated abve. Aerial Census - Transect lines are flwn lw ver the study area. This methd wuld be d i f f i c u l t in muntainus terrain, althugh i t has been successfully carried ut t survey mammals and birds in Russia (Kvalev, 1979). Expense and lack f aircraft ruled ut this methd in Nepal. Call Cunt Census - Observers are spaced alng paths r rads. They listen and nte dwn the directin and calling times f birds. Individual results are cmpared t give an index f the ppulatin. This methd appeared t be the mst viable in Nepal, and i t had already been used with success by Hwman (1977), Mirza (1978), and Severinghaus (1979) in muntainus regins. Other methds include bserving birds r their tracks at water hles, nting tracks in snw, and nting the abundance f drppings n trails r rsts. Nne f these methds were used in Nepal.

64 49 Gastn (1980a),in his recent review f Himalayan Pheasant censusing techniques distinguished between Abslute methds and 'Cmparative' r 'index' methds. Due t the prblems encuntered in Nepal stated abve, I t was decided that nly Index methds wuld be suitable fr census wrk during the study, s the Call Cunt and Direct Cunt (b) methds were carried ut. CALL UNT CBISUS METHOD In the Suth Annapurna study areas, tw f the fur pheasant species, the Tragpan and Kklass males, give characteristic calls at dawn and smetimes during the day. In additin bth sexes f Mnal, and the Cheer Pheasant (further West), call mrning and evening. A cunt f the calling males in a knwn area gives an estimate f the number f males in that area, and i f the rati f males t females is knwn, an estimate f the breeding ppulatin can be btained. Alternatively, i f bth sexes are calling, a direct estimate f ppulatin can be determined. The methd at Pipar was similar t that used by Hwman (1977); fur census pints were marked alng the t r a i l which passes thrugh frest and scrub frm the main campsite nrthwards t Kalki Danda (see Figures 4 and 5) I t was cnsidered in 1979 that fur was the maximum number f pints needed t effectively census the area, and in 1980 the psitin f pint number 4 was mved t a lcatin slightly Nrth f the riginal (see Appendix 6 fr descriptin f census pint lcatins). Census takers walked the t r a i l in daylight and were infrmed at which pint they were t stand; in additin they were played the calls f Kklass and Tragpan s that they fully understd what they were listening fr. On the mrning f each census, each wrker was required t take up psitin at his pre-allcated census pint well befre f i r s t light, t ensure readiness and s that birds were nt disturbed as they were waking. I t was cnsidered that disturbance at this pint culd result

65 50 in nn-calling. Each wrker then recrded the fllwing details: a) Name and census pint number. b) Date, time f start f census, time f finish (pre-arranged). c) Weather cnditins: clud cver, wind speed, precipitatin, temp. d) The directin and apprximate distance frm which each individual bird called, and i f pssible, the time f its f i r s t and last calls. Figure 9- These details were recrded n a diagram similar t that in Immediately after the census was cmpleted the wrkers met t analyse the results in rder t eliminate any uncertainties regarding psitins f birds. A l l the diagrams were cmpared, and by relating the psitin f each census pint alng the t r a i l, the psitin f each bird recrded, and the time i t began/finished calling, a gd estimate f birds heard alng that t r a i l culd be made. In rder t estimate the number f birds in an entire blck f frest, i t is necessary t extraplate the results frm the census alng a single t r a i l, and Gastn (I980a)gives details n hw t d this. At Pipar, due t the tpgraphy f the land I cnsidered that ur censuses cvered the area frm the treeline dwn t apprximately m, which gives an area f rughly 2.3 km f relatively unifrm habitat fr the Satyr Tragpan and Kklass Pheasant. The technique is based n the assumptin that a l l males in a certain area w i l l call every day, which partly depends n the time f year. Kimball (194-9) nted seasnal trends f crwing f the Ring-necked Pheasant (Phasianus clchicus) with a peak in May. Gastn (1980a) prvides a table (here Table 7) shwing the apprximate calling perids f fur western himalayan pheasants. Gates (1966) believed that

66 SPECIES: CENSUSER: DATE: WEATHER; 20/S-/77 -*~,Z*** TIME START: **r TIME FINISH: CENSUS POINT NUMBER; S K K AST K K 0440" K figure Q: Example f a cmpleted census sheet.

67 nn-crwing was minimal in Ring-necked Pheasants in Wiscnsin, hwever, data fr himalayan pheasants is nt dcumented. K. C. R. Hwman (pers. cmm.) stated that in captive Kklass, autumn calling was predminently by yung males, and that captive first-year Tragpan and Mnal males d nt. call in spring, althugh ther aviculturalists have nt nted this. The age f sexual maturity in himalayan species is nt recrded, but it is pssible that nn-breeding males are silent in Spring. TABLE 7 Seasnal timing f dawn calls in 5 pheasant species (Adapted frm Gastn 1980a) J P M A M J J A S O N D Kklass Pheasant Himalayan Mnal Western Tragpan Satyr Tragpan Cheer Pheasant A number f pssible variables have t be standardised i f this census technique is t be successful: 1. Time f day; calling in Tragpan and Kklass is usually restricted t a shrt perid just befre sunrise. Figure 10 shws the calling rates f chrusing Kklass males n 16 April, 24 April, and 3 May* Censusing was carried ut n these species at this time since individual birds call nly spradically ( i f at all) during the rest f the day, and give n reliable ppulatin index, see sectin 6.J.4. Cheer Pheasants may pssibly be censused at dusk, since they are knwn t call equally well at dusk as at dawn (Hume and Marshall 1879* Beebe 1937, Gastn 1980a). Hwever, when encuntered in Nepal, they never called at dusk, and nly a little at dawn (see sectin 6.5-2).

68 ~ V ""jr * r~ % APRIL 1980 ;unrse S i a 2? 20 UJ f =i 6 LJ 4 24 APRIL 1980 lb SUNRISE L UNRISE TIME(HOURS) Figur. 10: Recrds f Kklass Pheasant calling Intensity fr three Individual mrnings, expressed in crws per 3 minute perid.

69 2. Census pints: these were pltted n a large scale map in additin t being marked n the grund by a cairn r a blaze. In successive censuses, the same pints shuld be used. 3. Weather: a number f wrkers (Kimball 19^-9, Gates 1966) stress the need fr similarity f weather n days that censuses are carried ut. The tw mst imprtant variables are wind velcity and clud cver, s ideally the mrning shuld be calm and clear; censuses shuld never be cnducted in windspeeds f greater than 13 k.p.h. In mst censuses carried ut at Pipar, the sky was less than ^ cvered and the wind n i l. 4. Time f year: fr cmparisn f a ppulatin frm year t year, censuses shuld be carried ut as clse t the date f previus census as pssible. This is in rder t standardise such variables as altitudinal mvement and breeding seasns. Regarding the ther tw species in the Pipar area, the Bld Pheasant and Himalayan Mnal, censusing was mre d i f f i c u l t. While Mnals d call at dawn, and t a lesser extent at dusk, i t appears t be irregular, females call as well as males, and since first-year males are similar in appearance t females, an idea f numbers is difficult t assess. Gastn (1980a)states that in the Mnal,dawn calling "appears t bear l i t t l e relatinship t the numbers f birds present", while P. J. Garsn (pers. cmm.) cnsiders that calling is initiated nly by a stimulus such as a predatr. The Bld Pheasant n the ther hand, has n dawn calling perid and cannt be censused by this methd. Direct Cunt (b) appears t be the best methd fr the Mnal, but i t can nly be carried ut shrtly after dawn r befre dusk since at ther times f day (unless the weather is very vercast, raining r fggy)j the birds w i l l nt flush. This methd is nt very effective fr

70 the Bld Pheasant hwever, since the species des nt flush, and large cveys in winter cver a very wide area, while pairs in spring are usually parchial. The latter fact enables them t be mapped ver several visits t an area, especially i f individual males can be recgnised (fr example by the amunt f 'bld streak' feathers n their breast r the number f spurs n their legs). Ppulatin estimates are given nly fr the Pipar study area (except in the case f Himalayan Mnal) since the time spent at the ther study areas was insufficient fr ppulatin surveys t be cnducted Bld Pheasant This species was rughly censused by mapping the psitin f each pair ver the areas f suitable habitat visited up t IS times. In April and May 1979, in an area f 2.4 km^ (240ha) f pssible habitat, an estimated ppulatin f between six and nine pairs were present. 2 This gives an average density f 2.5 t 3.8 pairs per km f suitable habitat. In April 1980, between six and eight pairs were again believed t be present in the same area, thugh the pairs were nt frequenting exactly the same lcal areas as in As far as is knwn, this is the f i r s t attempt t census Bld Pheasants and s n cmparisns with ther areas can be made. In my pinin, these figures represent a reasnably healthy ppulatin f Bld Pheasant and i t shuld be brne in mind that the study areas lie at the western edge f the range f the species, where its numbers may be thinning ut. The species was als bserved by the authr in the Everest Natinal Park in March 1980 where ne cvey f furteen birds (apprximately seven males and seven females), and anther cvey f six birds (five males and ne female) were seen. Hwever, these bservatins were made

71 prir t cvey breakup fr breeding, and n estimates f ppulatin density can be made Satyr Tragpan This species was censused by the call cunt methd in bth 1979 and I980. On 21 and 23 May, 1979 dawn cunts f calling males were cnducted alng tw kilmetres f the t r a i l running Nrth frm the main campsite at Pipar.: n bth ccasins a ttal f ten birds were heard calling (see Table 8). The area censused has been measured and estimated t be 2 2 apprximately 2.3 km, giving a minimum density f 4.4 pairs per km f suitable habitat, assuming mngamy. In 1980, ne Tragpan census was carried ut n 1 May alng the same t r a i l, but with nly tw cunters (due t lack f manpwer). A ttal f seven calling males were heard (See Table 8), giving a minimum density f 3-0 pairs per km 2. TABLE 8 Censuses f Satyr Tragpans at Pipar in 1979 and I98O Date Number cunted frm Ttal After each census pint Crrectin Weather 21 May, Sky clear, n wind 23 May, Sky vercast, n wind 1 May, Sky clear, n wind The weather n 21 May, 1979 and 1 May, 1980 was similar, althugh n frst was recrded n 21 May. On 23 May, 1979 the sky was vercast, yet the same number f Tragpan were heard as n 21 May. The apparent decrease in ppulatin frm 1979 t 1980 shuld nt be taken at face value. One pssible reasn fr the discrepancy is that the censuses were unavidably carried ut three weeks apart. The species shws seasnal altitudinal mvement, and by 1 May, 1980 less male Tragpans may have reached the height f the census area than had by 21 May 1979«In additin, f u l l t e r r i t r i a l calls f the male Tragpan appear t

72 surrunding frests. reach a peak arund mid-may, s that fewer birds may be calling n 1 May than by 21 May. These pints having been made, it was f e l t that less Tragpans were present at Pipar in April 1980 cmpared with April 1979 (1.2 Direct Encunters cmpared with 19 respectively). In additin, ne freshly predated male Tragpan was fund n 15 April I t is f e l t that the census fr 1980 des reflect a slight decrease in the Tragpan ppulatin at Pipar and further censuses are required in future years in this and ther areas t cmpare ppulatin fluctuatins. Numbers f calling Tragpan males were als nted elsewhere arund Pipar utside the census area, and n every ccasin in suitable habitat belw 3300m frm tw t six birds were heard, indicating mderate abundance (areas nt measured). In additin i t shuld be nted that the census area included frest frm apprximately 3000m t the tree line - the uppermst part f the altitudinal range f the species. Althugh n data n Satyr Tragpan ppulatins have been cllected elsewhere, Z.B. Mirza (1978) has carried ut sme preliminary censuses f Western Tragpan (Tragpan melancephalus) in Pakistan in In Salkhala Game Reserve in May 1977 Mirza recrded twelve calling males, and in additin tw males and fur females were flushed, giving a ttal 2 f furteen pairs (assuming mngamy) in 31 km f frest, r 0.45 pairs 2 per km. At Kuttan Game Reserve in May 1977, nine males were heard calling, seven males and fur females were flushed - a ttal f up t 2 2 sixteen pairs in 26 km f frest, r 0.62 pairs per km. In Machyara in August 1977, a ttal f seven males and six females were flushed in an area f 8 km, althugh a ttal f 52 km was surveyed. These give 2 a maximum f 0.9 pairs per km. The Western Tragpan is listed as endangered in the Red Data Bk (IUCN 1979), s its ppulatin densities wuld be expected t be lw. Nevertheless, the figures fr Satyr Tragpan frm Pipar are encuraging, especially since the numbers f birds in the census area appear t reflect their abundance in the

73 Kklass Pheasant Like the Satyr Tragpan, this species was censused by the call cunt methd in 1979 and On 20 and 22 May, 1979 dawn cunts f calling males were carried ut. alng apprximately tw km. f the t r a i l running Nrth frm Pipar campsite. On bth ccasins, eleven males were heard calling (see Table 9) in the census area f apprximately km, giving a minimum density f 4.8 pairs per km f suitable habitat, assuming mngamy. On 6 Octber and 11 Octber, 1979 similar cunts ttalled 6 and 8 calling males respectively, indicating densities f 2.6 and 3-5 pairs per km. In 1980, Kklass cunts were carried ut n 23 April and 1 May by tw cunters alng the same t r a i l. On bth ccasins a ttal f six calling males were heard (see Table 9), 2 giving a minimum density f 2.6 pairs per km. On 20 May, 1979 the weather was lw clud t apprximately 3300m and a temperature f c C, while n the ther three ccasins i t was clear and calm with a slight frst. TABLE 9 Censuses f Kklass Pheasants at Pipar in 1979 and I980 ^. Number cunted frm Ttal After,, Date,, r-, Weather each census pint Crrectin May, Sky vercast. Lw clud t 3500m. N wind. 22 May, Sky mstly clear. Wind NW frce 1. 6 Oct Sky mstly clear. 11 Oct Sky vercast, n wind. Rain during night. 23 April, Sky clear. N wind. 1 May, Sky clear. Wind frce 1.

74 As with the Tragpan censuses, the apparent differences in ppulatin size between the tw years shuld nt be interpreted t l i t e r a l l y. Factrs such as the difference in date between the tw years, and the smaller number f census takers shuld be taken int cnsideratin. I t shuld als be nted that the census area was at a cmparatively high elevatin in the species' range, and i t can be assumed that the figures given are underestimates. Hwever, i t was nted that there appeared t be fewer calling Kklass present at Pipar thrughut the I98O field seasn than had been heard in Annual censuses f the area wuld be f great value in cmparing future ppulatin fluctuatins. Unlike the Satyr Tragpan, Kklass were nt heard calling frm ther areas arund Pipar utside the census area. On the suthern, western, and nrthern slpes f frest where Tragpans were heard calling, nly a single Kklass was heard n tw ccasins. While i t is unlikely that the species was entirely absent frm these slpes, i t may have ccurred nly at lw densities. A mderate amunt f recent data has been cllected n ppulatin densities f Kklass Pheasant in Pakistan by Hwman (1977), Mirza (1978), and Severinghaus (1979). The latter authr gives a gd summary f ppulatin densities which I reprduce here (Table 10) having cnverted int metric units. The range in densities varies cnsiderably frm t 10.4 pairs per km, while Severinghaus himself estimated a density 2 f 22.5 pairs per km. He cnsiders that these ranges cannt be explained in terms f habitat differences r disturbance, but instead by "the different methds used t estimate the land areas n which the densities were based". I believe this explanatin t be the crrect ne, but while the Kklass densities determined by Hwman, Mirza, and 2 myself f a l l within the range f 1.5 t 10.5 pairs per km, Severinghaus' estimate appears t be highly anmalus. His census area

75 60 CA CA rh 1/) CA fa 00t t t-- h-0- c P <n 00 OO 00 c-- u >> CA OO OH -p OA CA CA t-- t-- t- ^ t- CA t>- OA CA rh CA ON CT' H OAr-H r-h rh CA CA OA &0 0) PH S-, rh H H rh C rh rh rh r-l r-h rh c c c H c a> O (0 <0 <0 P-, c c c.c x: -P CD CD CD N E N N N CD NIS) N -p -p VH S IH PH Pn > U ft u UJ ID a- ID H O O O -H -H H Q) H -H -rl P* P. ~ CD > H PH CD TJ 10 CD H -P -H W C CD «C O rl -P -P CM E g X E S E S s: c s s s O U a LOOO -=t OAOO OA OJ UA rh Cvj t>- OJ UA lia 0J 0J OA UA-3" r-hrh O -=} LT\ -=r rh >> & n T> 3 -P P P -p t w w w w w w rl rl H <H -H. EH c v mvj v -=3" CM KA CM CM 3 D, O Pn C CM E X in in in K\VO rh ' OJ rh m rh O rh VO CM r<a CM UA VO VO VO rh CM CM CM KA KA KA KA KA OJ OJ OJ CM OJ bo a H rh rh O <(H UA OJ LT\ KA CU O r-t r-i KA VO O -=t KA UA MO CM KA rh t-< r-< rh VO 00 VO VO OA w OT t-- 0\ rh rh t CD -P X! CM ho O C a H PH w -P <D P2 > CD rh rh O PH >j <+H -p H rh P-, \ mining- «««w\-^ -=t -^r m W W W KA OA KA KA KA CUrH OJ OJ OJ OJ i i i UA t^- CA w \ KA OJ CD I -P =t rh >> CD rh CM ho >> CD 1 1 rh c H rh rh T) a a rh (0 > (0 C H P^ u H rh >J rh > bo ho fao w -P H O O O E a) W co. rh C CD CD rh s c H c rh rh e> CD <0 «<D C3 t~ C OA in c KA KA KA CM CM OJ KA (0 P-. S P E bo ^B cb cs ra LT\MJ w 0\rH I CM KA PH -p 41 tn CD OA CM VO CM O CM U as Q. H An CA O OA O O H -st in KA rh CM

76 size was "visually estimated t be 120 acres by Mr. Malek Ali Mhammed, an experienced frester", but frm Severinghaus' densities I cnsider this may have been grssly inaccurate. Rugh estimates f Kklass ppulatins made by Gastn (1979) in Dachigam Sanctuary, Kashmir 2 give density f apprximately ten pairs per km. A cmparisn f the Pipar ppulatin densities with thse frm elsewhere shws that they l i e in the mid range f mst estimates Himalayan Mnal As nted abve, this species was d i f f i c u l t t census, but n three ccasins in 1979> Direct Cunts f birds were made. On 10 May, 1979 a ttal f ten birds (fur males, fur females/ immatures, tw unknwn) were seen and heard alng a 800m stretch f Kalki Danda ridge between 3500 and 3680m elevatin. The ttal area f suitable habitat alng such a stretch was estimated t be 800 x m = 0.28 km, giving a density f apprximately 37 birds per km (see Table 11). On 22 May, 1979, a ttal f eleven birds (fur males, seven females/immatures) were bserved n the Suth-facing slpe f Namrung, West f camp I I I (see Figure 3). They were seen in 2 an area f apprximately 0.7 km, giving a density f 16 birds per 2 km. Finally, n 4 Nvember, 1979, ten birds (five males and five females/immatures) were seen n the Suth-facing slpes f Krchn (see Figure 3) The birds were bserved ver an area f apprximately knt giving a density f apprximately 25 birds per square km. The errrs invlved in such estimates are likely t be very large, the main ne being that at the time f day when the surveys were carried ut (early mrning hurs), the birds may have gruped tgether in high numbers ver a lcalised area f hillside fr rsting purpses, resulting in anmalusly high density estimates. Because f the lack f extra manpwer, ther parts f the same hillside

77 TABLE 11 Censuses f Himalayan Mnal in Pakistan and Nepal, and Ppulatin Densities derived therefrm Lcality Date Number f Birds Cunted Area Cunted Density (km^) (bird/km 2 ) Surce Pakistan: Kuttan and Salkhala May Mirza I978 Machyara Aug Mirza 1978 Nepal: Pipar (Kalki Danda)10 May This study Namrung 22 May 79 H This study Krchn 4 Nv This study were nt surveyed t determine whether birds were present r nt. Hwever, i t is f e l t that i f surveys had been carried ut later in the day when the birds had dispersed t feed ver the hillside, they culd nly have been flushed with d i f f i c u l t y and many culd have been missed. I t shuld be nted that because this species may have a plygamus mating system, densities have been expressed in terms f 2 2 birds per km Instead f pairs per km. The nly available data n Himalayan Mnal ppulatin densities are thse f Mirza (1978). Using a cmbinatin f the call cunt methd, beating, and the use f dgs, in May 1977, he recrded a ttal 2 f frty-ne flushed in 41.6 km in the Kuttan and Salkhala areas f 2 Pakistan; a density f apprximately 1 bird per km. In August 1977 at Machyara he recrded a ttal f f i f t y calling birds and eleven 2 birds sighted in an area f 34 km f cmparitively gentle slpes, 2 giving a density f 1.5 t 1.8 birds per km. Mirza als nted the d i f f i c u l t y f censusing these birds. In cmparisn t the densities I have nted at Pipar and Krchn, thse f Mirza appear very lw. But

78 63 while my wn densities are prbably ver-estimates, I believe that Mirza's may be underestimates i f surveys were carried ut later than c hurs n clear days, when my wn bservatins indicate that Mnal feed less n pen slpes and are nly flushed with difficulty (see sectin 5.3.4). 4,4 Habitats and Occurrence f Pheasants in the Athhazar Parbat Regin Pieldwrk was carried ut between 2100m and 2600m abve Khibang, Dara, and Muri villages (see Figure 6). Due t the lack f time available, n systematic data was cllected n the habitats visited, but the area lies within the 'Dry Oak' (Querus lanuginsa) frest divisin f Dbremez and Jest (1971). The habitat was subjected t cnsiderable human disturbance in the frm f crp cultivatin (millet and wheat) up t 2320m altitude n Suth-facing slpes, and 2170m n Nrth-facing slpes. Further disturbance was caused by the very large amunt f wd-cutting fr fuel and t a lesser extent, fr building purpses; the grazing f sheep, gats and cattle; and the annual burning f grassland t stimulate new grass grwth. Immediately abve the limit f cultivatin there was a zne f Berberis scrub with Quereus lanuginsa and Rhddendrn arbreum. This was gradually succeeded by fuller frest f these species and thers such as Symplcs sp. and Lynia sp. in pen t clsed canpy up t 6m in height. Hwever, mst f this frest was heavily grazed by livestck thrughut the year, with the result that understry and grund vegetatin were light, and in sme places cmpletely absent. Much f the cuntry abve the limit f cultivatin cnsisted f steep, grass-cvered and bare c l i f f s with small ledges. These ranged in gradient frm 45 t vertical and in places held small patches f stunted trees and bushes. The grass cvering was a tussck grass and appeared similar t that seen at Pipar and elsewhere further East.

79 4.4.1 Cheer Pheasant Althugh this species was said by lcal villagers t be present in the frest and n c l i f f s abve Khibang, Dara, and Muri villages, i t was nly lcated abve the latter. Here i t was seen and heard frm 14 t 17 May, 1980 in pen scrubby frest and grassy c l i f f s frm apprximately 2200m t 2440m altitude n East- and Nrth-facing slpes abve cultivated fields. Birds were nly seen n fur ccasins, but heard n three mre, always frm the same area f hillside. Abve Dara village the habitat was similar t that abve Muri, and was situated just ver the tp f the h i l l, facing Suth. Sherpa Ram Kaji Mangar had seen and heard Cheer in this area in September 1977, but althugh we searched and listened frm 10 t 13 May, we culd nt lcate any. The area abve Khibang village, n the ppsite side f the Myagdi valley was visited frm 16 t 18 August 1979, but n Cheer were seen r heard, prbably because they d nt call during the O mnsn perid. I t was visited again fr ne evening/mrning n 17/18 May, 1980, but n Cheer were lcated n this ccasin either. In this lcality the frest was similar t the ther areas, and the slpes mstly faced Nrth and West. The abve bservatins generally supprt thse fund in the literature cncerning Cheer Pheasant. Fr instance Beebe (1922) stated that "The Cheer is extremely lcally distributed, and seems t me very capricius in i t s chice f habitatins: n ne side f a river yu meet with plenty in suitable spts; n the ther side yu may search 50 square miles f likely-lking cuntry and never see ne". Delacur (1977) describes its habitat as "wild cuntry, haunting precipitus ravines, brken hillsides and c l i f f s with light and stunted frest, brush and grass." With regard t the ppulatin f Cheer Pheasant in the area abve

80 Muri, n nly ne ccasin did the birds call sufficiently t prvide a ppulatin index. Unlike thse described by Gastn and Singh (1980) and thers, the birds did nt call regularly at dawn and dusk, but at 0815 n 14 May, 1980, fur separate individuals were heard calling acrss an area f apprximately 700m frm c l i f f and frest at apprximately 2230m altitude. Althugh Finn (1915) and Delacur (1977) state that the species ccurs in parties f up t 15 birds, Gastn and Singh (1980) reprt that "in April, practically a l l appeared t be in pairs". An estimate f numbers in the Muri area wuld then appear 2 t be apprximately fur pairs in 0.5 km. This figure shuld nt be extraplated ver a larger area hwever, until mre surveys have been carried ut in the Athhazar Parbat regin since the very lcal nature f i t s ccurrence precludes even an apprximate estimate f the ppulatin density f this species Other Species In the areas visited in this regin, nly ne pheasant species ther than Cheer was encuntered. This was a single Kklass male heard calling at dawn n 11 and 12 May at 2600m in stunted Quercus frest, abve Dara village.

81 5. FOOD AND FEEDING BEHAVIOUR This sectin deals mainly with reprts frm the literature, and, t a lesser extent, my wn bservatins. The accunts in the literature are usually very general, and ften qute reprts f previus authrs withut acknwledgement. Pheasant species appear each t have a f a i r l y wide range f diet ver the whle year, but at a specific seasn i t is prbably much mre restricted. A discussin sectin fllws the reprts n diet and feeding behaviur f individual species. 5.1 Diet Bld Pheasant See Tables 12 and 13. The accunts in the literature indicate that principal fd items taken by this species are msses, buds, yung shts, lichens, seeds, fruits, and insects. Althugh this is a rather general l i s t, i t shuld be nted that rts, tubers and grubs were never recrded as being present in its diet, and that leaves were unimprtant. Beebe (1922) prvided a useful, mre detailed descriptin f fd items taken in April. He bserved Bld Pheasants leaping at l i l y stems and seed cases at the edge f deep snw, kncking ut the cntents and eating them. When the birds had mved elsewhere, examinatin f the seed cases revealed the fllwing cntents: earwigs (Order Dermaptera), ladybird (cccinellid) beetles, dipterus larvae, spider, f l i e s, staphylinid beetles, chrysmelid beetles, carab beetles, weevils, mths, msquit, hmpterus insect, and seeds. Cheng (1963) prvided the fllwing l i s t f items fund in the stmachs f sht Bld Pheasants in China: Fragaria chilensis, wild nins, primrses, Pedicularis resupinata, grasses, Clematis,

82 B 67 TAHLE 12 Fd items f different pheasant species as reprted in the literature SPECIES Fd Type Bld Pheasant Satyr Tragpan Kklass Pheasant Himalayan Mnal Cheer Pheasant Seeds ac f km aj' cefhjk cei jk cejk Grain Fruits Acrns Flwers Leaves Buds/ Yung Shts Msses abc 2 fkl c^hkm f~jm p c'~hijm.3 aj m P f afijml df j\i h cefhjk efj efk cefhjk cf jk ef cefjkm el cei eijk ej cejk gk k Lichens Fungi Rts Tubers/Bulbs jn i efk J f efj km f jlm ejkm gjjm Pine and Juniper Tips ^ b k n Grass Insects Grubs ai 5 f" hkmn df j m k cef jm cf jk e^fjkm cef jk ecfjkm Ntes: a - Hdgsn (in Hume and Marshall 1879) b - Hker (185*0 c - Blanfrd (in Hume and Marshall (1879) d - Hume and Marshall (1879) e - Wilsn (in Hume and Marshall 1879) f - Beebe (1922) g - Whistler (1926) h - Cheng (1963) i - Hellmich (1968) j - All and Ripley (1969) k - Wayre (1969) 1 - Fleming et. al. (1976) m - Delacur (1977) n - Grahame (in Delacur 1977) 0 - This study 1 - January 2 - September j> - Western Tragpan 4 - Autumn 5 - April Underline indicates that fd item is regarded as imprtant by authr.

83 68 N -P U as OJ KA -p 0) w c t p PL, OJ X) <D t H H CM H >» -P i-l d as CD C X) c H d rh CD t d CM CD -p X! 0) d O CD H a t CD rh U a T3 >> x: CD X> CD H c 13 X! CD -P d a CD H -P m CD t -3 a> r-l t t a CD bo 6 > C as t H w a i-3 CD i-3 Q. <M O O M T3 M CD CD t -P X> t d u w CD X, a E P CD d 4J H M PC LA CU KA KA KA KA OJ -P TJ c O CD O CD en t CD t PH a w 5 CD -P H O s en O ^~ t f-r CD CD rh x> a 3 d w t CD H O CD a -p d t CD x: PH a rh pq d a OS u EH M >» -P (0 c OJ P d t CD x: OH t w ih KA KA KA d S d rh E H d a ho u >> -p XI M PH t rh X,Q X! CM

84 69 Thalictrum minus, Aletris japnica, Hemistepta lyrata, crwfts, buttercups, Lnicera kunglana, L. kehneana, etc. Als beetles, larvae, and mlluscs. The stmach cntents f 14 Bld Pheasants sht in Khumbu, East Nepal, (Hellmlch 1968) revealed much mss present in fur, grass in tw, fungus in tw, and unidentified green plant material present in seven. My wn bservatins f feeding Bld Pheasants were quite extensive, and the birds were usually seen pecking at msses, leaf litter, and grass. The results f superficial analysis f faeces cllected in bth autumn 1979 and spring 1980 are shwn in Table 13, and they indicate that at Pipar, mss was the mst imprtant fd item taken, althugh much unidentified vegetable matter and tw insect wing cases were als present. Very fine quartz grains were present in mst f the drppings, and these are prbably used fr grinding the vegetable matter in the stmach t facilitate digestin. These grains were smaller than thse fund in the drppings f the ther pheasant species examined. Undubtedly, there are seasnal changes in the principal fd items taken, and these prbably reflect fd availability at different times f the year Satyr Tragpan See Tables 12 and Ij5. Prm the reprts in the literature, this species prefers t feed n leaves, buds, yung shts, fruits, seeds, insects, and t a lesser extent acrns, flwers, rts and tubers. This l i s t f fd items is similar t that f the Bld Pheasant, but i t differs in that leaves and yung shts appear t be the mst preferred vegetable fd, and that mss was never recrded in their diet. Beebe (1922) recrded the stmach cntents f three sht Satyr Tragpans: tw held many trn leaves and flwers f 'paper laurel', and ne f these had als eaten a number f insects - several small

85 earwigs, black ants, a gd sized cckchafer, spiders, and a small white centipede. The third bird cntained rhddendrn flwer petals and laurel leaves. Fleming et. al. (1976) reprt a Tragpan's crp f u l l f the fern pinnae f Drypteris wallichiana. My wn bservatins f feeding Tragpans shwed marked differences between spring and autumn in the srt f items taken. Almst all the feeding birds sighted in the spring seasns f 1979 and 1980 were feeding n the grund, usually in streams and leaf litter. I t is likely that insects wuld be plentiful at these sites as well as yung shts and seeds. In autumn 1979, eleven f the fifteen birds seen feeding were in trees r bushes, and were bserved feeding n Berberis berries and Symplcs berries. Trees f the Genus Symplcs were very cmmn n the area f ridge belw Kumai h i l l, where we made a camp in Nvember Tragpans (mstly females and/r immatures) were seen feeding in these trees every day frm 5 t 9 Nvember, smetimes nly 15m frm my tent, and they shwed a marked attractin t the Symplcs fruits. The results f superficial faecal analysis carried ut in autumn 1979 (Table 13) indicate that mss, lichen, leaves, grass, and rts were taken by the Tragpan, as well as sme insects. While in spring 1980, buds were the main fd item, with a l i t t l e grass. The quartz fragments present in the drppings ranged up t 2mm in size. Clearly these faecal remains represent nly a small prprtin f the f u l l diet f the Tragpan, and this is likely t vary cnsiderably thrughut the year Kklass Pheasant See Tables 12 and 13. Prm the literature accunts, the Kklass shws fd preferences similar t thse f the Tragpan, including leaves, insects, seeds, fruits, and t a lesser extent acrns, flwers, buds/yung shts, msses, rts and tubers in its diet. Shu Chen-huang (in Cheng 1963) reprted the fllwing stmach cntents f tw Kklass

86 Pheasants sht in China: Selaginella ferns, maize, seeds and fruits f slenaceus plants, seeds and tender leaves f pine, spruce, and thers. The degree t which the Kklass is insectivrus is in dispute: accrding t Wilsn (in Hume and Marshall 1879)> Blanfrd (1898) and Wayre (1969), they are principally vegetarians. Hwever, Beebe (1922) reprted that they "prefer insect fd t all else, and spend much f their time in search f it", and Delacur (1977) cnsiders that they are mre insectivrus than mst pheasants. My bservatins f feeding Kklass were limited t ne sighting. This was f a female scratching in leaf litter adjacent t and within a small stream in April Tragpans and Bld Pheasants had been bserved feeding at this site n ccasins in Simple analysis f Kklass drppings cllected in autumn 1979, (Table 13) revealed leaf, grass, and mss fragments, seeds and quartz fragments. Analysis f drppings frm spring 1980 revealed mainly grasses, with a few large mss leaves. If they had been eating insects, at least sme traces shuld have been visible in the faeces, as in thse f the Tragpan Himalayan Mnal See Tables 12 and 13. The reprts in the literature shw that the fd items taken by the Mnal include rts, tubers/bulbs, insects, grubs, fruits, and t a lesser extent seeds, acrns, leaves, buds/yung shts, and fungi. This list differs frm the ther species, in that rts and tubers/bulbs are the mst imprtant fd items, while leaves seeds and thers are less imprtant. Beebe (1922) reprted that sht Mnals 1 crps were full f tuber material which was very hard. He nted that "the edges f the upper mandible must perfrm an imprtant functin in cutting and splitting vegetable tissues f such firm cnsistency". Hellmich (1968) recrded the fllwing items present in the stmachs f six sht birds in Khumbu, East Nepal: green leaves,

87 shts, grass, buds, and seeds. Ali and Ripley(1969) specify berries f Ctneaster micrphylla being taken by the Mnal. All my bservatins f feeding Mnal were f birds digging fr fd. In a number f cases when the digging areas were examined, the birds were fund t have been pecking at the rts and tubers f herbaceus and wdy plants which included Mecanpsis paniculata (Himalayan Pppy) and Arundinaria sp. On ther ccasins, the pheasants appeared t have been feeding n grass rts, mss, and pssibly insects and grubs in the tpsil. In Octber 1979, a Mnal was seen feeding n mss n a rcky surface. A sample f the mss was cllected, and has been kindly identified by Mr. A. Eddy f the British Museum (Natural Histry) as being Rhacmitrium crispulum mixed with sme Bryum sp. Simple faecal analysis carried ut in autumn 1979 (Table 13) revealed mstly tuber fragments present in the Mnal drppings, with sme seeds, leaf fragments, and quartz grains less than 1mm in diameter. Analysis f drppings cllected in spring 1980 shwed mainly large mss r liverwrt fragments with numerus rhizids Cheer Pheasant See Table 12. The accunts in the literature recrd that the Cheer feeds mainly n rts, tuber/bulbs, grubs, insects, and, t a lesser extent n seeds, fruits and grain. These fd items are similar t the preferences f the Mnal, but since their altitudinal ranges hardly verlap, the tw species are unlikely t be in cmpetitin with each ther. Beebe (1922) reprts that the crp f a bird which was sht while digging revealed eleven wire wrms and six cckchafer grubs. I made n bservatins f feeding birds during this study, and n faeces were fund. 5.2 Discussin In wild unmanaged ppulatins f the Cmmn Pheasant Phasianus

88 73 clchicus, the types f fd items taken thrughut the year vary accrding t the seasnal availability f fd. The general seasnal pattern is f grass, leaves and rts taken during the winter and spring, seeds and insects taken in the summer, and grain, seeds, and fruits taken during the autumn (Cllinge , Hammer et. al. 1958* Lachlan and Gray 1973) Himalayan pheasants are likely t shw similar seasnal changes in diet, and sme evidence fr this was seen in the Tragpan (sectin 5.1.2) and the Mnal (sectin 5.1.4). As has been stated previusly (sectin 4.2.5), all fur pheasant species present in the main areas can ccur in similar habitat in the zne f their altitudinal verlap (3050m t 3300m) during the breeding seasn. Hwever, it is unlikely that they are ever in direct cmpetitin fr fd since their fd preferences, althugh verlapping with each ther, are all slightly different, as fllws: Bld Pheasant: small msses, grasses, seeds, and insects. Satyr Tragpan: leaves,buds/shts, insects, and fruits. Kklass Pheasant: grass, leaves, seeds, insects, and fruits. Himalayan Mnal: rts, tubers/bulbs, large msses, insects, grubs and fruits. The tw species with the mst similar fd preferences wuld seem t be the Kklass and the Tragpan. It is pssible that they are in cmpetitin fr fd in the winter, when it is scarce, but fr the rest f the year there is likely t be an abundant fd supply, sufficient fr bth species. In additin, althugh nly ne sighting f a feeding Kklass was made, it is likely that this species feeds in mre pen areas than des the Tragpan (sectin 5.3.3«). Further West, the Cheer Pheasant verlaps altitudinally with the Kklass and Tragpan, but neither were heard in the area where we lcated the Cheer. Mrever, it is unlikely that they are in cmpetitin

89 74 fr fd with a species that principally digs fr rts, tubers, insects, and grubs. Regarding animal fd, P. clchicus is reprted t take up t 20$ (by vlume) in August/September (Lachlan and Gray 1973) > 43% in June/ July (Cllinge ), while Hammer et. al. (1958) recrded insects present in Qff f crps in June. Himalayan Pheasants are all reprted t cnsume animal fd, and I cnfirmed this fr the Bld,Tragpan, and Mnal. I cnsider that while the Mnal may actively dig fr insects and grubs, the ther species eat them nly when they cme acrss them in their search fr vegetarian fd. Exceptins t this may ccur where there is a super-abundance f insect fd. Animal fd prbably frms a small but cnstant prprtin f the items taken by himalayan pheasants, becming mre imprtant in the summer mnths, in a similar way t insects in the diet f the Red Gruse Lagpus lagpus scticus in Nrthern England (Butterfield and Culsn 1975). 5.3 Feeding Behaviur 5.).1 Bld Pheasant The Bld Pheasant was bserved feeding at all times f the day, and the species did nt appear t have any special feeding perids. Hwever, i t will be shwn that the birds are mst mbile between 0700 and 1000 hurs, with significantly less mbility during the afternn (sectin 6.1.3). If feeding is related t 'mbility' i t wuld appear that feeding is greatest during the mrning perid 0700 t 1000 hurs. Hwever, nn-mbile birds culd als be feeding actively. Of 46 ccasins when the nature f a bird's activity culd be determined, 36 encunters were with feeding birds. They fed by pecking at the grund, raising their heads nly a little after each peck, althugh n sme ccasins wary birds assumed an upright, alert

90 psture after each series f abut five pecks. Arbreal feeding was cmmn: birds ften fluttered r jumped up t a lw branch f a tree r bush, and pecked at the mss cvering. This was ften dne when birds were aware f an bserver's presence; perhaps mving t a site abve grund gave them sme degree f security. Bld Pheasants were never bserved scratching at the grund with their feet in the manner f the dmestic chicken (Gallus gallus) r the Cmmn Hill Partridge (Arbrphila trquela). Neither were they ever seen digging in the grund with their b i l l. On ne ccasin (15 May, 1979), a male Bld Pheasant was bserved twice jumping ff the grund vertically, neck utstretched, in rder t catch flying insects. Hwever, n n ther ccasins were birds actually seen eating insects. Drinking was bserved n nly ne ccasin, 19 April, 1980, when the female bird f a pair crssed a small snw patch and pecked at i t abut seven times, tipping its head back slightly t swallw the melted snw Satyr Tragpan The Tragpan was seen feeding at mst times f the day, except between 1000 and 1200 hurs. Beebe (1922) and ther wrkers have reprted that feeding takes place mainly in the early mrning and the late afternn, and my bservatins reprted later, indicate greatest mbility befre 0700, with a smaller peak during the mid afternn (sectin 6.2.3). Of the 29 sightings f feeding birds (all made in frest habitat), 10 (3^) were feeding in damp areas r small streams, 10 (34$) were feeding in dry leaf litter, and 9 (32$) were feeding in bushes r trees. When feeding in streams r leaf litter, birds wuld scratch litter and debris backwards with their feet, in the manner f dmestic chickens, and peck at what was revealed by this actin. The technique was bserved n nly tw ccasins, but drppings, and ther

91 evidence f Tragpans feeding in stream curses and damp areas, were seen n numerus ccasins. When feeding in trees in spring, Tragpans were seen plucking and eating leaves and buds, while in the autumn, arbreal feeding was principally n fruits. Like the Bld Pheasant, this species was never seen digging fr fd with its b i l l. 5.3«3 Kklass Pheasant Due t its very shy nature, the Kklass was seen feeding n nly ne ccasin, althugh i t was disturbed n a number f ccasins when i t was prbably feeding. It is nt reprted t have any specific feeding perids, and my data regarding mbility levels (sectin 6.30) are heavily biased in favur f the early mrning. The nly bird bserved feeding undisturbed was a female watched frm a hide n 12 April, While feeding in a small stream, she acted in a very similar way t a Tragpan, scratching backwards with her feet, and pecking at the revealed litter. Later n, when feeding in dry leaf litter, she was walking very slwly, deliberately, and silently, pecking ccasinally at the grund. I t is likely that Kklass feed mre quietly than either the Bld Pheasant r the Tragpan, since these species were ften lcated while feeding, due t the nise they made in the leaf litter. Unlike the Tragpan, i t is likely that the Kklass feeds mre ften in scrub and at the edges f the frest (Severinghaus 1979). Prbable feeding birds were disturbed frm such lcatins n eight ccasins, and in Gharwal, Beebe (1922) recrded pairs nt uncmmnly ging ut n t pen rcky slpes and scratching deep hles in the turf. Similarly, M. W. Ridley (pers. cmm. 1980) bserved Kklass n cliffs and in a clearing n tw ccasins in Himachal Pradesh, in September Arbreal feeding was nt bserved in my study except fr a female Kklass pecking at mss n a tree branch. This appeared t be a displacement activity rather than feeding (sectin 6.j5. *0

92 5.3.4 Himalayan Mnal The Mnal was bserved feeding at all times f the day, and my bservatins regarding mbility indicate that the species was mbile thrughut the day (see sectin 6.4.3)- Of the 55 bservatins f feeding birds, 42 (76%) were feeding in grassland, 8 (15$) were feeding in wdland, and 5 (9$) were in scrub. Hwever, i t was nticeable that the incidence f birds feeding in pen grassland was related t the weather, as shwn in Table 14. When the sky was clear and the sun was shining, Mnals were bserved feeding in grassland nly befre 0900 hurs, and generally befre 0800 hurs. Hwever, i f the sky was vercast, and especially i f lw clud (fg) was present, Mnal were seen feeding n pen grassland at any time f day. Birds feeding in wdland and scrub were encuntered at any time f the day whatever the weather. I t appears that Mnal feed n pen grassland in the early mrning, ften befre sunlight has reached the slpes. I f the rest f the mrning is clear, i t is likely that they then mve t feed in wdland and scrub, where they are less easily encuntered. I f the day becmes vercast, with lw clud and/r rain (the usual pattern f weather at Pipar, see sectin 2.2), Mnals return t r cntinue t feed n the pen slpes. Qualitative bservatins f this nature have als been made by the Himachal Wildlife Prject team, in Himachal Pradesh (M. W. Ridley pers. cmm. 1980). I cnsider that this behaviur is due t the large size f the species and the male bird's cnspicuus plumage (sectin 6.4.2). In the early mrning, the sun is weak, is at a very lw angle and casts lng shadws; in additin, therrnals are likely t be weak r absent at this time f day, s that large aerial predatrs such as eagles and vultures cannt use them (n large raptrs were ever seen saring befre 0830 hurs). The Mnal therefre, can feed n pen slpes withut being t cnspicuus, and there Is less

93 78 TABLE 14 Incidence f Himalayan Mnal feeding In different habitats in relatin t weather cnditins Grassland Wdland/Scrub Sky Clear 0825** * * ** OOO 3* 0710** O625* *2** 0600* 0848» 0440* ** 0615* * 0545* ** 0815* 3** 0600* ** *2** 0545* ** Sky Overcast 1150+* 0855+* 0800+** 1720* 1620+* 1735+* 1620+* **.1440* ** 0930* ** * ** 0450+* 1220+* ** 1630* 1735** * * 0800+* 1445+* ** ** * ** I830* 2** * ** * 0950+* ** Sky Clear Sky Overcast O * 1240* 0920* 0900* * 0bl5** 1450* 1320* 1300* 1525** 1535* ** Ntes: 1 - Bird in shade 2 - Birds under bush 3 - Rain 4 - Hail * - Male ** - Female Figures indicate time f day (hurs) T7>,-> rt>

94 prbability f attack frm raptrs. If the day cntinues t be clear and sunny, Mnals feeding in the pen are very expsed t view and prne t predatin; therefre they mve int scrub r wded areas t feed in cver in such cnditins. When i t is vercast, fggy, r raining, Mnals are much less easily seen, and therefre they mve ut t feed n pen slpes again. This hypthesis makes a number f assumptins, including the fllwing: (1) Mnals prefer t feed n pen slpes, (2) Bth sexes act similarly despite the female's mre cryptic clratin, and (3) Mnals are mre prne t predatin n pen slpes than they are in the frest. Mnals feed mainly by digging with their pwerful bills, smetimes alne, but ften in pairs r grups. At Pipar, n mre than pairs and tris f birds were ever seen digging tgether, but in the Everest Natinal Park, ten t fifteen birds were seen digging in a single grup f fields. The digging technique invlved pecking hard at the grund in rder t disldge earth and rts, and then scratching the lsened earch backwards with their feet. Birds were seen feeding n tubers in the grund, and als n fd items disldged by the pecking and scratching actin. Mre casual feeding, invlving pecking at mss, herbs, and grass leaves was als bserved n a number f ccasins. Arbreal feeding was never bserved in the Mnal Cheer Pheasant Feeding in the Cheer Pheasant was nt bserved in my study except by Ram Kaji Mangar, a sherpa in my emplyment. He saw a male Cheer feeding in Dry Oak wdland n 14 May, 1980 at 1350 hurs, and a female feeding n grassland n 16 May, 1980 at 0520 hurs, but n details f feeding technique were recrded. Cheer are reprted t feed mainly in the mrning and evening, by digging in a manner similar t the Mnal, and t have favurite digging places (Wilsn in Hume and Marshall 1879, Beebe 1922).

95 6. BEHAVIOUR AND VOCALIZATIONS This sectin deals with daily and seasnal patterns f behaviur f the five principal pheasant species studied. I t cncentrates n the infrmatin I btained frm wild birds in Nepal, and relates this t published bservatins made by ther wrkers n bth wild and captive birds. In additin, sme useful material was btained frm individual members f the Wrld Pheasant Assciatin wh replied t a shrt questinnaire sent t them. A cpy f the questinnaire is included at Appendix 3«6.1 Bld Pheasant Wariness This species was the least wary f all the five pheasants studied. In 53 Direct Encunters the cumulative amunt f time fr which I bserved Bld Pheasants was apprximately 1040 minutes, with an average bservatin time f 20 minutes per Direct Encunter. This alne gives sme indicatin f the cmparative ease f bservatin when cmpared t the ther species (sectins 6.2.1, 6.3.1, and 6.4.1). Mst authrs cnsider i t t be f a cnfiding nature: Beebe (1922) stated that i t is "nt wary unless persecuted"; All and Ripley (1969) asserted that i t is "tame and fearless t the pint f stupidity, the members f a cvey... allwing bird after bird t be killed by the ambushed hunter". Hwever, Beebe's pint "unless persecuted" seems t be imprtant in determining the bird's wariness, since Hker (in Hume and Marshall 1879) described them as "very wild", and i t can be assumed that the birds wuld develp sme degree f cautiusness i f heavily hunted by Man. My wn bservatins suggest that the Bld Pheasant's wariness varies frm individual t individual and depends

96 n the suddenness f apprach t the bird. On a number f ccasins in spring 1979* a cvey f up t three pairs was apprached, phtgraphed, and recrded at a distance f Jm. Other pairs in the area, hwever, shwed cnsiderable alarm n being apprached by a human, and usually retreated. In autumn 1979* cveys were rarely seen, and nne were tame, while in spring 1980 nly ne pair was as apprachable as thse seen in In general, i f birds were startled by sudden apprach they wuld call in alarm and retreat, but i f apprach was gradual, with slw mvements, the birds wuld ften remain feeding, r watch the bserver inquisitively Prtective Behaviur Escape - As described abve, the reactin f the Bld Pheasant t Man was variable, but n clse apprach the bird wuld retreat. In every case, escape was n ft away frm the bserver, and birds were seen flying nly up t r dwn frm a lw tree r bush, frm which they wuld stare at the intruder. In additin, n 17 April, 1979* a male Bld Pheasant was seen flying apprximately 30m ver the camp dwn t the Suth-East ut f sight. These bservatins f grund-escape are cnsistent with thse frm the literature: Hume and Marshall (1879) described the bird's flight as "shrt and feeble"; Ali and Ripley (I969) nted that i t is a "swift and strng runner... lth t take wing", while Delacur (1977) cnfirmed that "they rely n their legs t escape". Beebe (1922) hwever, bserved that when suddenly alarmed, the cvey wuld flush and scatter, but with less abrupt alarm the birds wuld run n ft "necks utstretched, heads and tails held rather high" I t is pssible that his first bservatin was made n pen grund, in which case the birds might flush twards cver. When a pair was escaping n ft thrugh wdland, the nise made by the birds' feet, by disturbing leaf litter, was cnsiderable. In

97 i 32 additin, ne r bth f the birds wuld ften call in alarm. They usually stpped after retreating 10 t 15m away, and cmmenced feeding, uttering call types 1, 2, r 3- Such vcalizatins (see sectin 6.1.4) were f great help in lcating pairs r cveys in dense scrub r frest. Reactins t ther predatrs - Respnse t pssible predatrs ther than Man was nly seen n ne ccasin. On 21 May, 1979 a single male was bserved feeding fr ten minutes. During this perid an adult Lammergeier (Gypaetus barbatus) flew lw (30m) ver the scrub in which the bird was feeding. At the sund f its wings the Pheasant lked upwards and std mtinless withut calling, but did nt cruch dwn; the bird having passed ver, the Pheasant resumed feeding. Beebe (1922) dcumented a similar bservatin in which an eagle flew ver; in this case, a male Bld Pheasant squatted, turned its head sideways, and stared at the sky, remaining mtinless fr three minutes until the bird passed ver. Cryptic Clratin - The female Bld Pheasant shws ideal cryptic clratin fr a frest bird, being an almst unifrm brwn, with fine streaks. On the ther hand, the male shws a degree f bth cryptic and phaneric cluring, fr althugh parts such as the t a i l and breast are brightly clured (streaked red), the verall impressin f grey and green appears cryptic and disruptive when viewed in scrub and wdland. In display, the male puffs ut his feathers t draw attentin t his phaneric cluring,while in Beebe's bservatin f the Bld Pheasant's reactin t the eagle (see abve), he nted that all the bird's red clratin remained hidden while i t was in the cruched psitin. Interspecific Cmmunicatin - Severinghaus (1977) reprted that the alarm call f Steere's Babbler (Licichla steeri) prbably had an effect n Swinhe's Pheasant (Lphura swlnhei) and Mikad Pheasant

98 (Syrmaticus mikad), and may have caused them t escape r becme silent and immbile; althugh he had n direct evidence t supprt this. On 19 April, 1979, I nticed that the alarm calls f Tits (Paridae) caused a pair f Bld Pheasants t becme alert, lking arund warily, but after 5 t 10 secnds they resumed feeding. On 14 May, 1979> I bserved a single male react in a similar manner n hearing the alarm calls f passerines Daily Life Mbility Levels Table 15 shws the number f sightings f Bld Pheasant thrughut the day. These may be partly biased by my wn daily mvements, in tw main ways: (1) Generally, I made lnger and mre frequent attempts at bservatin in the mrning perid than in the evening; and (2) Mrning bservatin attempts were smetimes made n different days frm thse in the evenings. Bth f these srts f bias were made due t better weather cnditins ccurring in the mrnings than in the afternns and evenings, making bservatin bth mre likely and mre pleasant. Fr analysis the day was divided int five equal parts: , OfOO , J.00, , and hurs (see Figure 11). I t was fund that there was a significant difference in the number f p bservatins between these intervals (X = v=4, p=0.1). The pattern f sightings thrughut the day may be a measure f the mbility f the pheasants, since mbile birds are likely t be mre easily seen. I t then appears that mbility is greatest during the mrning with a peak between and 10.00, and is significantly less during the afternn.

99 84 TABLE 15 Sightings f Bld Pheasant in relatin t Time f Day, 1979 and I980 Time f Day Number Sighted Unknwn 2 53 Rsting I never bserved this species rsting, but Beebe (1926) nted that i t rsts cmmunally in firs in winter and in lw rhddendrn bushes and scrub in spring. Grahame (in Delacur 1977) saw Bld Pheasant fly "up t rst at the tp f a 20 ft tree", and he has subsequently reprted (pers. cmm. 1980) the species rsting in lw trees and bushes. Hker (in Hume and Marshall 1879) reprted f this species that "during winter i t appears t burrw under r in hles amngst the snw, fr I

100 Figure 11: Histgram f pattern f sightings f Bld Pheasant thrughu".si'; day. 13C TIME(HOURS) ( a Figure 12: Histgram f pattern f sightings f Satyr Tragpan B p thrughut the day.

101 have snared i t in January in regins thickly cvered with snw, at an altitude f feet" (j5660m). I agree with Grahame that Hker's evidence was incmplete and circumstantial, that the bird's build is nt adapted t brrwing, and that the bservatin is unlikely t be true Vcalizatins Described here are five varieties f call f fairly definite meaning. As far as is knwn,tw (Types 1 and 2) have nt been described befre, and nne have been described in detail by any authrs. 1. 'Sree' call - A high-pitched squeak f shrt duratin, similar t the first syllable f the Type 3 'Squeal' call. While feeding, this was heard called frm fur t nine times per minute n average. 2. 'Trill' call - A high-pitched t r i l l lasting apprximately 0.8 secnds, ften heard in cnjunctin with Type 3 call. The call appeared t be triggered by mild alarm, fr instance i f ne bird f a pair culd nt see its mate, r if apprach by a human was t clse. While feeding, this call was heard frm 0 t 4 times per minute. These calls carried nly a shrt distance (up t 20m) and were made at any time f day by bth sexes. They were used t maintain cntact between a pair f birds, usually while feeding, when they frmed an intermittent lw 'chatter'. Appendix 4.1 shws a snagram f call 2 and shws that the t r i l l is made up f an initial part lasting apprximately 0.4s beginning as a pure tne at 4200 hertz rising t a diffuse tne at 6000hz. The secnd part f the call cnsists f a series f 7 frequency mdulatins ranging in frequency frm l^oohz at their trughs t 6000hz at their peaks. The first mdulatin lasts 0.08s, and the remaining six last 0.24s; the ttal call lasts abut 0.8s.

102 On five ccasins, single males were seen and heard calling these cntact calls althugh n ther birds were in evidence; in three f these cases the birds were calling Types 3* ^> and 5 (see belw), in additin t the cntact calls. These calls appear t be similar in functin t thse f the Bbwhite Quail Clinus virglnianus (Stkes 1967) and the Chukr Partridge Alectris graeca (Stkes I96I), in which the fd cntact call "serves t keep the birds aware f thers in the grup". 3. 'Squeal' call - A piercing muted squeal: 1 Sree cheeu cheeu cheeu' lasting up t 3«2 secnds depending n the number f repetitins f the 'cheeu' syllable (range 2 t 6, n = 14). Appendix 4.2 shws a snagram f this call shwing the initial 'sree' syllable t have a fundamental frequency at abut 700 t 1300 hz, with sets f vertnes at hz and hz. The first 'cheeu' syllable lasts abut 0.4s and successive nes becme mre attenuated, dwn t 0.15s n a f i f t h repetitin. Each 'cheeu' syllable is cmpsed f a frequency mdulatin frm apprximately 1500 hz peaking t 5500 hz and dwn t a series f very rapid mdulatins at abut 3000 hz which give rise t a harsh tne in the call. These syllables als shw a pure tne fundamental at abut 800 hz with a set f harmnics repeated every 800 hz. The interval between syllables ranges frm 0.12 t 0.l8s. This call carried a lng distance, up t 200m n a calm day, and was made by bth sexes at any time f day. Its pssible functins are interesting: i t appeared t serve t bring scattered individual members f a cvey tgether in a similar way t the "Rally call" f the Chukr (Stkes 1961), the "Grup mvement calls" f the Bbwhite (Stkes I967), and the "Scial cntact call" f the Califrnia Quail Lphrtyx califrnicus (Williams 1969). Of the nine ccasins when mre than ne pair f birds was bserved, in eight cases they were making this call while feeding thrugh scrub r wdland; in additin there were a number f ccasins

103 38 n which calling was heard but n birds were seen. All the members f a cvey culd make the call; i f a bird became separated frm the rest i t wuld call, and, upn hearing answering calls frm the ther members, mve t jin them. The call was als used fr cntact between bth the sexes f a single pair - f the 29 ccasins when a single pair was bserved they made this call n 21 ccasins. Stkes (196l and 19&J) suggests fr the Chukr and the Bbwhite that the rally call als serves the functins f repelling intruding males, and spacing ut cveys by keeping neighburing cveys infrmed f a cvey's lcatin. He nted that a female Bbwhite easily recgnised the call f her mate, and when separated frm him wuld respnd nly t his call. He therefre cnsiders that the birds f a single cvey can be attracted by the recgnisable calls f members f their cvey when scattered, but be repelled by calls frm anther cvey. In the Bld Pheasant I fund circumstantial evidence supprting this thery when I recrded a cvey f three pairs and immediately played back their wn 'Squeal' calls t them. Three birds (tw males and ne female) then apprached the recrder with great curisity and tameness, calling ludly. On cessatin f the playback they rejined the cvey. Hwever, when this same recrding was played back n five ccasins t ther pairs and cveys knwn t be in ther areas arund Pipar in 1979 and I980, n birds ever apprached the recrder, but n ne ccasin there was a vcal respnse f the 'Squeal' call frm an unknwn number f birds. Thus, the playback f the call f ne cvey t ther birds may have served t repel them. Stkes (1961) als suggests that the rally call f the Chukr may be sexual in functin, serving t attract females. But in the Bld Pheasant pair frmatin is likely t ccur within the cvey, and I saw n evidence f males attracting females by calling. Hwever, I cnsider that the 'Squeal' call prbably has a third functin in this

104 species, that f 'mild alarm'. Of the 14 ccasins when the Type 4 'Chik' call and alarm behaviur (see belw) was bserved, n 12 f these the 'Squeal' call was given in assciatin with them. In additin, a pair disturbed suddenly while feeding quietly wuld ften make the 'Squeal' call up t five times each, and i f nt apprached mre clsely wuld resume feeding. Williams (1969) suggests that the structure f the scial cntact call f the Califrnia Quail is a cmprmise between the selective pressures f the need f the bird t annunce its lcatin (t ther Quails), but nt t indicate i t t readily t pssible predatrs. He cnsiders that the restricted frequency range f the call cmbined with its extended duratin is the cmprmise: an intermediate between the easy-t-lcate, wide frequency range, shrt duratin call and the difficult-t-lcate, narrw frequency range, extended duratin call. In structure, the 'Squeal' call f the Bld Pheasant shws sme similarities t the Califrnia Quail's "cu ca cw" (assembly) call. The latter lasts fr an average f 0.73s and is repeated 1 t 9 times, whereas the Pheasant's call syllables last frm 0.15 t 0.4s and are repeated 2 t 6 times. Hwever, the Quail's call shws a frequency range f hz and an average f abut 2160 hz, whereas that f the Bld Pheasant ranges frm hz with a mean frequency f apprximately 3350 hz. Althugh the calls are nt identical,these similarities suggest that the cvey call is similar in functin t the assembly call f the Califrnia Quail, and may have evlved in the same way. In the literature, this call type has been described by Blanfrd (in Hume and Marshall 1879) a s "a peculiar lng cry like the squeal f a kite". Beebe (1922) described the species inaccurately as "remarkably silent" and the "cvey call alarm nte" as a "drawn ut 'see-e-e-e-e-lpe 1 snapping ff shrt at the end". In my pinin, Fleming et. al. (1976)

105 90 have prvided the mst accurate phnetic rendering: "a lud grating alarm screech 'kzeeuuk-cheeu-cheeu-chee'". 4. 'Chik' call - A high pitched, strident 'chik' f shrt duratin repeated at intervals f apprximately ne t tw secnds. This call was heard n twelve ccasins uttered by bth sexes, and in the case f the male bird was usually accmpanied by call 5 and alarm behaviur (see belw). 5. 'Buzz' call - A 'buzzing' nise f shrt duratin (1 secnd) made by the male bird nly. I suspect that this nise is nt made by the respiratry tract f the bird, but by its feathers. Grahame (1976) describes the display f the male as accmpanied by "a 'purrh' made by suddenly fanning ut the f a i l feathers"; this may be the same nise that I heard n a number f ccasins. Neither f these calls were recrded. The frmer carried a mderate distance (up t 50m) and bth were heard in spring and autumn. I f apprached suddenly by Man, alarm in bth sexes was shwn by erect crest feathers, upright stance and fanned t a i l, fllwing by strutting 2 arund a small area (abut lm ) with a prnunced frward mvement f the head and neck with each step. In additin the head wuld be bbbed dwn clse t the grund every few secnds. I f clser apprach was made t the birds they wuld usually run r flutter away calling Types J, 4, and 5> but i f the intruder apprached n clser and remained quiet, the alarm wuld wane and they wuld resume feeding. Pew studies have been made f the alarm behaviur f pheasants, but the abve descriptin f the Bld Pheasant shws much in cmmn with the behaviur f the Chukr Partridge and Bbwhite Quail which have been intensively studied. The 'chik' call (4) was described by Hdgsn (in Hume and Marshall 1879)

106 91 as "ship... ship", while Blanfrd (p. cit.) nted a "shrt mnsyllabic nte f alarm" and Beebe (1922) reprted "a sharp sudden series f ntes 'seep! seep! seepl'" Breeding Behaviur Curtship Display This was nt bserved in the wild, but mating was seen n ne ccasin when a pair were feeding tgether n mss-cvered rcks and leaf litter in frest. The female std n a raised rck and made the 'Trill' call t the male.; the frmer then squatted dwn n a rck, the male walked ver and quickly munted her while bth called sftly. The male then cpulated fr 10 t 15 secnds while hlding n t the female's nape feathers with his b i l l, then jumped ff and bth birds cntinued feeding quietly. Beebe (1922) describes the display f the male as fllws: "The cck spreads its tail and wings, drpping the latter and raising the crest, puffs ut the breast feathers, and struts befre the hen, turning arund and arund". He als nted males fighting fiercely. Accrding t Grahame (1976) the display is similar t that f the Kklass Pheasant (which is lateral), but silent except fr the "purrh" nise (prbably call type 5) made by the fanning f the tail feathers. He bserved much display between tw pairs kept in adjacent aviaries between which there was free passage fr the birds. In many cases, chasing f male after male and male after female tk place, but n pugnacity was evident. Mating Systems My wn bservatins n Bld Pheasants were made principally during the breeding seasn. Out f 49 bserved Direct Encunters at this seasn, 35 {11%) were f pairs r pair-multiples, 10 (20$) were f single males, ne was f fur males and ne was f a single female.

107 In additin, n 5 June, 1979* tw pairs were seen with five unfledged chicks and tw apparently 'lse' males. This evidence wuld suggest that mngamy was the nrmal mating system in my study areas since the single and 'lse' males culd have been paired but their mates were incubating. Blanfrd (in Hume and Marshall 1879) and Beebe (1922) nted a 1:1 rati f males t females, r a slight prepnderence f males, in pre-breeding seasn cveys, and stated that the species was mngamus. In Ithaginis cruentus tibetanus hwever, Sheriff (in Ludlw 1944) nticed a rati f tw males t ne female during the breeding seasn n three ccasins and was infrmed by the lcal Tibetans that bth plygamy and plyandry were practised by the species. In February, Grahame (1971) reprted ratis f at least fur males t ne female in cveys f I.caffinis in the Everest regin. Hwever, recent bservatins made by sherpas via Cl. J. Rberts (pers. cmm. 1979)* again n I.e. affinis in the Everest regin, suggest a purely mngamus relatinship in April and May, but with male-nly cveys in late May, when the female birds were prbably incubating. Gregariusness With the exceptin f the Eared Pheasants (Genus Crssptiln), the Bld Pheasant is prbably the mst gregarius f the family Phasianidae. In the mid-nineteenth century Hdgsn (in Hume and Marshall 1879) reprted frm Nepal "flcks f 20 t 30 birds" and als "packs... f as many as 70 t 100 birds". While i t is unlikely that numbers as high as the latter are fund tday, cveys f up t 30 t 40 birds have been bserved in Nepal in recent years (K. Sakya pers. cmm. 1979). In Sikkim Beebe (1922) nted flcks f 15 t 40 birds, while Ali and Ripley (1969) usually fund cveys f 5 t 10 birds, and smetimes mre than 30. In this study, I nly ever saw cveys f up t six birds (thugh n ne ccasin with five yung). Sightings in the autumn field seasn were disappintingly few: ne bservatin f fur

108 males, and the ther f an unknwn number f birds. Extensive searches at Kumai and Krchn revealed lcalities with ten t fifteen fresh drppings suggesting that a cvey (r cveys) were present in the area, althugh nt seen. In mid March 1980 I saw ne cvey f furteen and anther f six birds in the Everest Natinal Park, East Nepal. I t appears that after the breeding seasn and during the winter, cveys f family grups frm and mve arund a lcalised area which includes each pairs' breeding 'hme range'. At the apprach f the breeding seasn in April, the cveys break up int pairs, which smetimes assciate lsely during May and June. Cl. J. Rberts (in Grahame 1976) reprted frm the Everest Natinal Park that "On 29 April... the birds had all paired, but laying had nt started nly (up t 21 May) fund pairs. The cvey instinct had disappeared i t seemed". This cvey and pairing characteristic, alng with ther behaviural and anatmical features, is anther way in which the Bld Pheasant resembles the Partridge family mre clsely than i t des the Pheasants. 6.2 Satyr Tragpan Wariness Tragpans are knwn t be shy and wary, but because I was able t bserve them frm hides, the cumulative amunt f time fr which I watched Satyr Tragpans was 615 minutes, in 48 visual Direct Encunters. All authrs agree regarding the birds' wariness: Hume and Marshall (1879) cnsidered i t "very difficult t bserve, keeping t the thick undergrwth" Ali and Ripley (1969) cnsidered i t like the Western Tragpan (T. melancephalus) t be "very shy and wild", while Delacur (1977) stated i t was "extremely shy and wary". Of the 48 bservatins, in nly 8 (17$) the bird was nt disturbed and did nt flee frm the

109 bserver. Sightings were usually very brief, the mdal bservatin time being less than 1 minute per Direct Encunter. Usually the bird was encuntered withut warning while the bserver was searching thrugh frest, and quickly escaped by running int thick cver r flying int trees r bushes ut f sight. On nly ne ccasin did this nt happen: On 11 April, 1979 I suddenly came acrss a pair f Tragpan feeding quietly during a heavy hail- and thunderstrm; bth birds saw me, cntinued feeding n leaf litter in wdland withut alarm, and I mved ff befre they did. Presumably they were nt warned f my apprach due t the nise f the strm Prtective Behaviur Escape - Of the 39 bservatins f escaping birds, 20 were grund escapes, and 19 were by flushing. Unlike the Bld Pheasant the Tragpan is very arbreal in its habits - f the 50 Direct Encunters in which the bird's psitin was knwn, 20 (40$) were made f birds in trees. If a bird was apprached when in a tree, i t wuld always fly dwn t the grund and run int cver ut f sight; althugh n tw ccasins a bird flew frm tree t tree t escape. Birds disturbed n the grund ran r flushed int cver ut f sight, thugh usually nt int a tree. On initial escape frm an Intruder, the nise made by the bird running, fluttering r flying was very cnsiderable and was ften accmpanied by the 'wah wah' call which was cntinued after the bird was ut f sight. This ften enabled the bird t be re-lcated, but after a secnd disturbance it wuld mve much further away, s that detectin was very difficult. My wn bservatins agree well with thse frm the literature: Finn (1915) described the species as "arbreal", Beebe (1922) stated that i t was "nt uncmmn in trees", while Delacur (1977) recrds that "They are very arbreal, mre s than any ther Pheasants".

110 There are few ntes n the actual methds f escape recrded by ther wrkers, but Ali and Ripley (1969) nted the clsely related Western Tragpan "skulking away thrugh thick undergrwth n the least disturbance r flying up and cncealing itself mst effectively in densely fliaged branches f trees". 'Freezing' Behaviur - 'Freezing' behaviur was bserved n ne ccasin, 16 May, 1979, shrtly after mating had taken place. At 05.58, after feeding amngst leaf litter in a small stream, the female f a pair std mtinless just as the sunlight reached this part f the frest. She std alert and immbile fr 21 minutes, and then resumed feeding as befre. I t is prbable that the male was acting in a similar manner, because he was seen clse t the female as sn as she began mving again, but had nt been bserved while she was 'freezing'. I t was pssible that I was the cause f this behaviur and the female was reacting t my presence, but since I was well cncealed in a natural hide (tree rts) and had been bserving the birds since 05.25, I cnsider this unlikely. Anther pssibility is that the bird was reacting t anther predatr in the area which I culd nt see. Cheke and Cles (1975) bserved similar behaviur in the Pheasant (Phasianus clchicus) in England when three birds were n a tree. It is very pssible that Tragpans might shw 'freezing' behaviur when perched in a tree. If s, many birds culd have been missed in my study areas. Cryptic Clratin - The female Satyr Tragpan shws gd cryptic clratin, being rufus brwn abve bltched with pale markings. It wuld be better camuflaged than the male i f sitting n the nest, and prbably carries ut the incubatin. The male, althugh brightly clured arund the head, neck and breast, is a dull brwn n its

111 wings and lwer back, giving a degree f cryptic cluring when viewed frm abve r behind. In the dull light f the frest, even the red clratin f the neck and breast are much less nticeable. Interspecific Cmmunicatin - I t is prbable that the Tragpan reacts t the alarm calls f mst species by escaping r becming silent. When n 27 May, 1979 a Kklass Pheasant was disturbed frm a tree, and flew dwn t the grund calling in alarm, a nearby male Tragpan, which had been uttering the 'wail' call (see sectin ), ceased as the Kklass alarm sunded and shrtly afterwards flew t the grund t escape Daily Life Mbility Levels Table 16 shws the daily number f sightings f Tragpan. These are prbably biased twards the early mrning by a greater amunt f time spent in hides at this time f day and the factrs listed in 6.I.3. On analysis, dividing the day int 5 equal parts ( , , , , and see Figure 12) i t was fund that there was a very significant difference in the 2 number f sightings between these intervals (X = v=4 p=0.1). Again, i f the number f sightings is indicative f mbility levels, i t wuld appear that this species is mst mbile in the early mrning. Mbility then decreases after 07.00, rises again in mid afternn, and falls ff in the later afternn. Mbility may be related t feeding times (discussed in sectin 5-3.2). Rsting The Tragpan was always encuntered rsting in trees, (12 birds heard, 2 seen), either Rhddendrn barbatum r Betula utilis. The

112 97 TABLE 16 Sightings f Satyr Tragpan In Relatin t Time f Day, 1979 and 1980 Time f Day Number Sighted height abve the grund ranged frm 5m t 12m. When calling at dawn in spring, the males mved frm branch t branch, usually descending the tree, as the light strengthened. After a perid f up t 60 minutes they finally flew t the grund. In April-May 1979> ne male was knwn t rst in the same small grup f rhddendrn trees fr up t 38 cnsecutive nights, while ther males were knwn t rst in the same tree fr 7 r mre cnsecutive nights. All and Ripley (1969) refer t the Tragpan having the habit f rsting in trees, but ther authrs d nt mentin i t - prbably because they assumed i t was likely in such an arbreal species.

113 6.2.4 Vcalizatins I have distinguished 4 types f call, but this is by n means a cmplete descriptin f all the calls f this species. 1) 'Wah wah' call Uttered by bth sexes, this call was a lud, rapid, staccat "wah, wah, wah..." heard at any time f the day in spring and autumn. I t was recrded n nly ne ccasin (that f a male n 24 May, 1979)> and the snagram is shwn in Appendix 4.3. The structure f this call is mnsyllabic, with each 'wah' syllable repeated, and having the same harmnic structure f a fundamental at abut 400 hz and sets f vertnes at 800 hz, 1200 hz, and abut 1600 hz, thugh the tp harmnic is slurred dwnwards frm l800-l400 hz. Each 'wah' syllable lasts abut 0.15s, repeated every 0.32s, giving a rate f abut 130 per minute. The bird was als recrded respnding t playback f its wn call (see belw). The structure f the new call was similar, but included anther strng harmnic at apprximately 2100 hz, and nne f the harmnics were slurred dwnwards. The call was als given by females. This was nt heard at clse range but appeared t be identical t that f the male, althugh Beebe (1922) stated that i t differed in the 2 sexes. The functin f the call is uncertain since nly three bservatins were made while the call was being uttered, althugh i t was frequently heard in the distance. Beebe (1922) suggested that the "cck utters i t as a herald t his nuptual display - a high, rather quavering baa! baa! baa! baa! Where this is heard the hen is usually nearby, and, unless smething ccurs t alarm the birds, a display is almst sure t fllw. The hen utters a call cmparable t this when separated frm her nearly grwn yung, the call in this instance being given

114 99 singly and in a slightly higher shriller tne". The nly uninterrupted sighting f a male uttering this call (n 16 May, 1979) ccurred just befre display and mating, which adds sme supprt t Beebe's suggestin. Hwever, I heard the call n a number f ccasins in the autumn field seasn when i t prbably had n cnnectin with display. On 24 May, 1979 the male's call was recrded frm a hide and played back t the bird. I t immediately flew frm a tree t the grund, apprached the hide t within 12m, and resumed calling in respnse t the playback. Fr 90 minutes the bird remained in the area, calling and staring at the hide intermittently. I t flew n t a rhddendrn branch and remained there ccasinally pecking at yung leaves and mss, smetimes while calling simultaneusly. Playback t male birds was als carried ut during 1980, and althugh i t usually resulted in a bird appraching the recrder, i t did nt elicit a vcal respnse n these ccasins. It is pssible that the call may signify antagnism between males, like the alarm call f the Ring-neck Pheasant P.clchicus (Heinz and Gysel 1970). The nly bservatins f the female uttering the call were made when n 16 May, 1979 the hen was answering the male bird, prir t display. This call has been described by ther authrs, but n ther attempts have been made t explain its functin. H. Stevens (in All and Ripley 1969) reprted "a lud pitched 'wak' repeated several times" while Fleming et. al. (1976) nted a "wank-wank". 2) 'Wak Wak' call This call was similar t the 'cntact call' but f lwer amplitude and therefre less audible. I t cnsisted f a "wak... wak was repeated at 1-2 calls per secnd fr up t 2 minutes. It uttered by bth sexes when alarmed n being disturbed by a human, and prbably by ther predatrs. The bird retreated n being disturbed,

115 » 100 and made this call until ut f sight f the intruder; apprach twards the bird wuld ften elicit further calling and retreat. Of the 19 bservatins f flushed birds, 11 (58$) gave the 'wak wak' call as they flew. Beebe (1922) described the flush alarm-call as "a series f lud raucus ntes 'quak! quakl quak! quak! '". Fleming et al (1976) reprted the alarm call as a "ca-rk". Other authrs d nt refer t the alarm, except fr Delacur (1977) wh qutes Beebe. 3) 'Bleat' call A single mnsyllabic, shrt call like the truncated bleat f a sheep r gat, lasting apprximately 1.0 secnd was heard called by the male bird. Due t the pr quality f recrdings f this call, i t has nt been analysed n the snagraph. I t has nt been mentined in the literature, but has been nted by pheasant breeders (pers. cmm). In my experience, the call cmmnly preceded the 'Wail' call (Type 4) and was heard in assciatin with wing whirring. On nly abut 5 ccasins was i t heard withut subsequent calls. On 3 May and again n 26 May, 1979 a male Tragpan was heard in Berberis/Arundlnaria scrub, calling with type 3 whenever I made a lud nise, such as cracking branches. The bird was nt seen, despite careful stalking using the call as a guide, althugh a female was bserved clse by. M. Sawyer (pers. cmm.) nted his captive male Satyr Tragpan t use call-type 3 in assciatin with side display when wing whirring n the grund. His bird never gave the 'Wail' call at any time, t his knwledge. The facts that this call was heard nly in the breeding seasn, was heard in the wild in assciatin with call-type 4 and wing whirring, and has been nted in captivity in assciatin with display, suggest that i t is a breeding call.

116 101 4) 'Wail' call This is a drawn ut mnsyllabic animal-like wailing call made by the male bird nly, and is repeated a number f times t frm the cmplete call. Beginning at lw-pitch, the repetitins becme luder, mre prlnged and crescend twards the end f the call. Rendered phnetically the call is: 'Wah ah... ah ah'. The structure (see Appendix 4.4 and 4.5) f each nte 'ah' is similar and cnsists f a fundamental at abut 500 hz rising slightly t 600 hz by the end f the call. 16 visible sets f vertnes are present frm hz, the secnd, fifth and sixth nes being strngest. Each harmnic rises in frequency frm the beginning t the end f the nte, which lasts apprximately s, the later repetitins being lnger than the earlier nes. Details f recrded call lengths and calling patternscan be fund in Appendix 7 and Figure 13. The length f each call depended n the number f repetitins f the ah nte, and ranged frm secnds (n=25). The number f repetitins in all the birds tape recrded ranged frm 4 t 12 (n=28), with a mde f 6 repetitins. A single individual male's calling pattern was nted n 15-17> 18, and 25 May, 1979* and its number f repetitins varied between 4 and 7 (n=62) with a mde f 5. The pauses between each f u l l call ranged frm 1 t 8 minutes, with a mean f 2 min. 10 secnds. It is pssible that each male may have a calling pattern peculiar t itself and can therefre be recgnised by this, but the number f birds calls studied in detail was insufficient t test this hypthesis. Lcal villagers stated that yung males call with fewer repetitins and with lnger spaces between f u l l calls than d lder birds, but I was unable t cnfirm this. Calling was greatest at dawn, befre sunrise, thugh I 7 v V sme s 7... individuals cntinued t call fr a shrt time after the sun was-tip.

117 a> c r-( c f-! r-, r~. 0 ind eti r-< a. +-> u a > re ii x: cu (0 -P Q. a. M ii II CM H 'O " VI x: u 0) E w ^, P a> x: a> bo,0 a C E ^ d) D x c,- ) CD *-> O 2 c O 4 CM t_ c: 00 rh >> CP C CM O E It a b0. ^ cc CM On A V W i-h cu g,

118 103 Details f the calls f ne individual n 15, 16, 17, 19, and 25 May, 1979 were the nly nes nted: the bird's pattern can be seen in Figure 13- I t shws a cnsiderable variatin in pattern, with the calling perid ranging frm minutes, usually extending up t r just beynd sunrise. I t is pssible that the bird n 19 May was disturbed and hence did nt call after Time f first calling f the Tragpan is given in Appendices 5.1* 5.2, and 8 and Figure 14. Calling began minutes befre sunrise, and shwed psitive crrelatin with the calculated times f sunrise, which were cncrdant with bserved sunrise times. On Figure 14, times f first calling were pltted against date, alng with the calculated sunrise times. The tw appear t shw gd crrelatin, s time f first calling was pltted against time f sunrise n a scatter diagram (Appendices 5.1 and 5-2) and the crrelatin cefficient calculated. This was fund t be 0.93 fr the 1979 data, (v = 48, p = 0.001) and 0.84 (v = 20 p = 0.001) fr Bth cefficients indicate a gd psitive crrelatin between sunrise time and the time f first call, significant at the 0.1$ level. As with the Kklass Pheasant i t is pssible that timing f the first call f the Tragpan is influenced by external factrs such as the weather (see sectin 6.3.4). Times f first calling were divided int 2 categries; 'clear mrnings' (less than half f sky cvered by clud, n rain in night) r 'cludy mrnings' (mre than half f sky cvered, and/r rain in night) and are listed in Appendix 9. On clear mrnings (n=29), the mean time f first calling was 38.2 (S.E. = 1.0) minutes befre sunrise, whereas n cludy mrnings (n=13) the mean was 3^.0 (S.E. = 1.5) minutes befre sunrise. Hwever the difference between these 2 figures is nt significant (v = 40 t = I.36 p = 0.1) suggesting that weather cnditins have n effect n the times f first calling. Further research, with captive birds and artificial daylengths wuld be useful in determining the effects f light intensity n calling behaviur.

119 & Sunrise 00 n 0500 < O ^ APRIL MAY JUNE APRIL SB MAY Figure 'J\: Times f first call f Satyr Tragpan in relatin t sunrise.

120 105 After the early mrning peak, the 'Wail' call was heard irregularly and ccasinally thrughut the rest f the day. Prm qualitative bservatins I nticed that i t was heard mre ften when i t was misty and raining than when clear and sunny. Certain individuals als called at dusk, but nne were recrded. Ali and Ripley (1969) stated that the Western Tragpan made "calls at intervals thrughut the day, mre ften at dusk and daybreak (every 5-10 minutes, smetimes lnger)". Call-type 4 was heard nly in the breeding seasn. I heard it as early as 2 April in 1979 a t Pipar, and birds were s t i l l calling n 6 June, 1979 at Bhalu. N birds were heard giving call-type 4 during the autumn field seasn r during a 2 day visit t the study area n 26/27 Nvember. Gastn (1980a)has heard Western Tragpan calling frm March until June. See Table 7- The exact functin f the wailing call is uncertain. In the Ringnecked Pheasant, Heinz and Gysel (1970) nted that the crw call may have alerted a male t the intrusin f anther male int its territry, but cnsidered that there was n evidence t supprt the hypthesis that crwing inhibits ne male frm entering int anther's defended area. In the Satyr Tragpan the ccurrence f territriality has nt been established (see sectin 7.2.4), but bservatins I made after playback f the 'wah wah' call suggest that the male believed its wn recrded vice was an intruding male. Likewise, playback f call-type 4 n 15 April and 26 April elicited a similar respnse - n bth ccasins the male birds apprached the recrder, but did nt call. Recent wrk n passerines has shwn that calling may be at a peak at dawn in rder t advertise ccupied territries at the time when invasin by territry-less males is greatest (J. R. Krebs pers. cmm.)- And Kacelnik (1979) has suggested that early mrning calling takes place befre i t is light enugh t feed efficiently. Bth f these are

121 pssible reasns fr dawn calling in bth the Tragpan and the Kklass, and the latter thery wuld help t explain why calling is delayed n cludy days, when presumably light intensity levels are lwer. The frequency f calling seems t be cnsiderably reduced in captive birds. Althugh calling has been heard frm April t June (W. Presctt pers. cmm.), i t is generally restricted t a shrt perid each mrning. Wayre (1966) nted that bth Satyr and Temminicks Tragpan (Tragpan temminicki) "nly call nce r twice each day as dawn is breaking." K. Chalmers-Watsn (pers. cmm.). reprted a calling perid f abut half an hur, while W. Presctt's birds called fr nly 30 secnds each mrning, (pers. cmm.). K. Hwman (pers. cmm.) cnsidered that in captive cnditins, the clse prximity f s many ther birds affected the timing f call f all species. He als nted that first-year male Tragpans did nt call. Descriptins f the call f the male Tragpan have been attempted by many authrs, including Jerdn (in Blanfrd 1898), Hume (in Blanfrd 1898), Beebe (1922), Wayre (I966), All and Ripley (1969), Fleming et. al. (1976), and Delacur (1977), but these have all been phnetic and highly subjective Breeding Behaviur Curtship Display Display by the male was bserved n nly ne ccasin, n 16 May, 1979> frm a hide at Pipar. The bird was heard t use call-types 3 and 4, the latter a ttal f 11 times ver a perid f 31 minutes. I t was als heard wing-whirring 3 times befre using call-type 4. At 05.20, tw further wing-whirrs were heard frm the bird n a tree branch, and at a female was seen flying frm a tree t the grund apprximately 35m away. The male appeared nly 15m away with his feathers puffed ut and the female walked twards him. The male then displayed frntally

122 107 twards her with puffed-ut feathers, fanning his t a i l, pening wings slightly and quivering vilently at the same time as lwering his wings and t a i l. At this time he was standing very erect s that the larger f his breast celli were easily visible. The male then ran at the female with his blue fleshy 'hrns' erect and his blue lappet expanded. He munted the female, cpulating fr 5-10 secnds, after which she ran ut frm under him with cnsiderable fluttering and escaped dwnhill ut f sight. The male, with feathers sleeked back n his bdy then walked uphill, als ut f view. At bth birds were seen feeding n leaf litter in a small stream, mving slwly dwn stream abut 10m apart. N further display was bserved. Display has been bserved frequently In captivity, and was first described by Bartiett in Murie (1872). Hume and Marshall (1879) nted three types: 1. The male quivering and "blwn up" shwing his wattle and hrns. 2. The male with erect feathers but n shw f hrns. 3. The male standing erect n a perch and shaking head t expse hrns and wattles. Finn (1915) recrded a frntal and a side display. Full curtship display is prbably lengthy and variable. Delacur (1977) described the fllwing sequence: "The curtship cnsists f several actins representing gradual appraches, varying in intensity, t the final perfrmance: (1) A stately walk arund the female, the wing tward her lwered and partly spread, the shulder n the farther side raised, the bdy being thus flattened, with much f the upper plumage in view. This is the general side display f all Pheasants} (2) A sudden rush with partly spread wings, with r withut the erectin f the hrns r spreading f the wattle; (3) In the last phase, the bird suddenly stps. The plumage f mst f the lwer parts is fluffed ut;

123 108 the half-spread wings mve slwly up and dwn, with wrist edges well ut frm the bdy and the tips pressed inwards and dwnward; the head and neck are shaken spasmdically until the hrns and wattle are spread ut t their utmst. The lappet is spread and retracted with astnishing ease and rapidity. Sn the bird reassumes a nrmal psture, walks ff, picks up fd, nly t start display again in a mment". My wn bservatins f Tragpan display were similar t phases (2) and (3) f Delacur's descriptin, but I never bserved the lateral display (phase 1). Specific calls by the male bird while displaying have nt been dcumented, but M. Sawyer (pers. cmm.) nted call-type 3 being uttered by a captive bird while wing-whirring n the grund, a fact als recrded in the wild (this study). He als nted that side display was "very cmmn", while frntal display was less cmmnly seen. Mating Systems In the 1979 and 1980 breeding seasns, Tragpans were seen either in pairs (10 Direct Encunters, 17$), as single males (22 Direct Encunters, 38$), single females (13 Direct Encunters, 22%), r unknwns. In autumn 1979* birds were seen in small cveys (up t 4 birds) which presumably were family grups. Althugh the species was secretive and the bservatins were f cmparatively shrt duratin, the evidence suggests that i t is mngamus. Since s l i t t l e is knwn abut its breeding behaviur in the wild, its mating system is nt dcumented. Hwever, fr avicultural purpses, they are usually kept in pairs, althugh fertile eggs can be btained frm the rati f 1 male t 2 females.

124 109 Gregarlusness The Satyr Tragpan is generally a slitary bird. Of 58 Direct Encunters, 37 (64$) were f single birds, 13 (22$) f pairs, and the» remainder (16%) were f grups f up t 4 birds, mstly family parties. This slitary behaviur is als recrded by Beebe (1922), Delacur (1977), and All and Ripley (1969). The latter, describing» the general behaviur f the genus, als nted small family parties in the nn-breeding seasn and birds feeding in cmpany with Cheer, Kalij (Lphura leucmelana) and Kklass pheasants in "ut f the way 1 places". 6.3 Kklass Pheasant I Wariness The Kklass was the pheasant bserved least ften in the study 1 areas. A ttal f apprximately 145 minutes bservatin was made in 19 visual Direct Encunters, thugh 45 minutes f this were frm a single Encunter. Mst bservatins (7^#) were f less than five minutes duratin. I fund that the Kklass was by far the mst difficult pheasant t study by bservatin, because f its exceptinal shyness and ability t cnceal itself. Only fur (22$) Kklass sightings were made f undisturbed birds, all the thers were f escaping birds. Observatins frm hides were nt successful (ne sighting), since birds did nt rst in the same grup f trees fr mre than cnsecutive nights, s that prductive placement f hides was three difficult. Except fr Beebe (1922) and Delacur (1977) few authrs refer t the bird's shyness. I t was, hwever, the mst vcal species studied, s that birds culd be easily lcated and censused, i f nt actually seen.

125 Prtective Behaviur Escape Of 19 Direct Encunters f escaping birds, ten were by grund escape, and nine were by flight. Generally, grund escape tk place where cver was dense and i f birds culd hear an intruder apprach frm a distance. Flushing usually tk place when birds were apprached suddenly, r when they were in trees. This species is nt reprted t be as arbreal as the Tragpan, but f 23 Direct Encunters where the bird's psitin was knwn, seven (30$) were made f birds in trees, althugh all f these were made shrtly after dawn. Unlike the Tragpan, the grund escape f the Kklass was usually remarkably silent and was never accmpanied by a call. Flushed birds always called in alarm, which n ccasin was cntinued fr up t five minutes, but usually ceased shrtly after the bird had alighted. The silent grund escape was f cnsiderable advantage t the bird in remaining undetected. On sme ccasins I attempted t 'stalk' Kklass when they were calling at sunrise. The birds wuld usually detect my presence, stp calling, and walk r run away befre they culd be seen. On three ccasins birds were actually glimpsed, but escape was rapid and further detectin was impssible unless by chance. Their escape behaviur was well knwn t sprtsmen in India, since the Kklass was prized as a game pheasant. Hume and Marshall (1879) recrded that i t was difficult t flush and had a rapid flight. Blanfrd (1898) nted that "this bird is swift and difficult t sht". Bates and Lwther (1952) stated "They affrd excellent shting, being very fast n the wing, fr they have the habit f rising well abve the cver and then, turning dwn the slpe, they hurtle dwnhill n halfclsed wings". Ali and Ripley (1969) reprted that the bird "lies clse in cver and difficult t flush withut a dg".

126 I l l 'Freezing' behaviur 'Freezing' behaviur f the Kklass was nt bserved in this study. Hwever, Rberts (1970) reprted that "When startled, they first f all freeze, and if pssible slink away int the undergrwth rather than taking t flight". Similarly, Fleming et. al (.1976) nted that "When apprached i t ften freezes n branch f tree". I t is prbable that 'freezing' behaviur in the Kklass is quite cmmn, but is s effective that i t is rarely bserved. Cryptic clratin Bth sexes f the Kklass shw well-develped disruptive cryptic clratin in the streaked and mttled black, brwn and buff plumage. The male is the brighter f the tw sexes and althugh pssessing disruptive patterning in his plumage, uses this patterning t great effect when displaying the crest, cheek patches and t a i l (see sectin 6.3-5). Interspecific cmmunicatin I t seems likely that Kklass react t the alarm calls f ther species. On 12 May, 1979, while I was recrding a male Kklass calling clse by at dawn, a pair f Red-headed Laughing Thrushes (Garrulax erythrcephalus) nticed my presence and began calling in alarm. The Kklass stpped calling as sn as the thrushes started t call and was heard t fly t the grund and escape by ft int dense cver Daily Life Mbility Levels Table 17 shws the daily frequency f sighting f the Kklass. Since there were nly 29 Direct Encunters with the species, i t Is felt

127 112 that these were t few t adequately assess mbility levels. Hwever, the frequency f sighting was analysed in the same way as thse f Bld Pheasant and Satyr Tragpan (see Figure 15), resulting in a 2 significant difference (X = 49 v = 4 p = 0.1) between early mrning and the rest f the day. Assuming that the number f encunters is related t level f mbility, i t appears that Kklass is very mbile until hurs, and is almst inactive fr the rest f the day. Hwever, I believe that the daily pattern f sightings was heavily biased by: a) The greater amunt f time spent searching fr Kklass in the early mrning, b) By the early mrning crwing f the Kklass making lcatin and bservatin mre likely, and c) Mrning bservatin attempts were smetimes made n different days frm thse in the evening. TABLE 17 Sightings f Kklass Pheasant in relatin t Time f Day, 1979 and 1980 Time f Day O5.OO-O6.OO Number Sighted i5.oo-l6.oo

128 113 20H n TIME (HOURS) ' 'i«ut-'v l'>: Histgram f pattern f sightings f Kklass Pheasant thrughut the day.

129 114 Rsting Kklass were nted t rst always in trees, usually n a branch f a Rhddendrn sp. r Betula utilis frm 4m t 10m abve the grund. Birds were disturbed frm their rsts n six ccasins. At Pipar, in the area Nrth f the main campsite, sme males culd be distinguished by their calling patterns. By nting rughly the psitin frm which the birds were first heard calling each mrning, i t was nticed that any ne bird did nt rst in a particular tree r grup f trees fr mre than three cnsecutive nights. This prbably helps t reduce predatin by night-feeding animals. During the breeding seasn, pairs prbably rst tgether r in adjacent trees until Incubatin begins. On 12 April, 1980, a male was bserved rsting in ne tree, and a female in an adjacent tree apprximately 15m frm the male. When the male called its mrning crw, the female smetimes called immediately afterwards with a sft call (see sectin 6.3.4, call-type 5). This call-and-answer had als been heard in April and September 1979 (then unexplained), and i t is nw assumed that i t was uttered by pairs rsting clse tgether. After the breeding seasn, family grups may rst tgether. This pair-rsting behaviur was als nted by Beebe (1922) Vcalizatins 1) 'Kuk kuk' call This call cnsisted f a harsh, rapid, staccat "kuk-kuk-kuk-kukkuk..." called as the bird flushed n alarm when disturbed by a human. The call was heard n eight ccasins, nce when n human intruders were present, and was never heard uttered by a bird which escaped n ft. On flushing, the call was ften cntinued after the bird had alighted, nce fr up t five minutes. Mst f the descriptins f

130 this call dcumented in the literature are similar t mine: Beebe (1922) described the alarm f the male as a "kuk! kuk! kuk! kuk.' k-kal k-ka! k-ka!". The female's call was nt nted. Delacur (1977) reprted "a lud 'kuk! kuk! kuk! kk! kk! kk! kk! kk! ka! ka! ka!", while Severinghaus (1979) nly nce heard the alarm "a rapid, lud and raucus ka-ka-ka-ka-ka, etc.". 2) 'Aw-cuk' call This call was heard n seven ccasins but was nt tape-recrded. I t cnsisted f a lw-pitched, subdued clucking: 'aw-cuk aw cuk aw-cuk repeated fr up t 15 minutes by bth sexes. I t appeared t be uttered when a bird was cnfrnted by an unfamiliar bject, such as a sitting human. In tw cases, the bird was fllwed in thick scrub, and althugh i t culd hear r partly see the intruder, i t was nt sufficiently alarmed t escape immediately, and gave this anxiety call. On 12 April, I98O a pair f rsting birds was bserved at clse range at dawn. At hurs the male began this call frm its rst, and then crwed regularly until The nearby female, which had been calling sftly (Type 5) t its mate, mved twards me and began t give the 'aw-cuk:' call. She cntinued this fr 15 minutes, acting nervusly and shwing displacement behaviur, such as pecking at the mss n the tree branch. Becming increasingly nervus, she eventually flew ff calling Type 1 at 05.50, while the male flew dwn silently five minutes later. 3) 'Clucking' call f male In their reprt n the display f the Kklass, Harrisn and Wayre (1969) described the frward threat f the male as being accmpanied "by a cntinually repeated, subdued, and rather meldius clucking - a six-syllable call 'Chuk-cher-ra-ka-pa-tcha'." N ther wrkers have recrded such a call.

131 116 3) 'Crw' call f male This call cnsisted f a series f harsh, raucus syllables (usually fur r five), uttered at the same pitch. In my study I taperecrded fur variatins f the call, and these are displayed phnetically in Table 18 and cmpared with the variants that Severinghaus (1979) nted in Pucrasia macrlpha astanea in Pakistan. The calls were analysed by means f a snagraph which shwed that the call types differed in the number f syllables, the duratin f the call, the duratin f each syllable and the space between syllables. In additin i t was nted qualitatively that the amplitude f each syllable als varied. Appendices shw snagrams f call types A-D. TABLE 18 Phnetic Descriptins f male Kklass Pheasant calls recrded in Nepal, and cmparisn with thse recrded in Pakistan (Severinghaus 1979)- a Call Type Call Type Phnetic Descriptin. 1 * (Severinghaus 1979) A B C D ka-wow ka ka KAAA KA ka ka ka - KAAA ka KA ka - KAAA ka D* ka KA ka - KAAA ka kw b ka ka KAAA ka ka 2 ka ka KAAA ka ka 3 ka ka ka - KAAA ka 4 Ntes: a - Capital letters indicate amplitude, Length f 'AAA' indicates length f syllable.hyphenated ntes are called in quick successin. b - last 'ka' given at lwer pitch than previus ne, distinguishing i t frm the previus call.

132 In general, the structure f the crw is multisyllabic, with each syllable having a fundamental frequency f apprximately hz and a harmnic f weak amplitude at apprximately J000 hz. Frequency and amplitude mdulatins give rise t the harsh quality f the call. Table 19 shws the call length and rhythm f syllables fr call Types A-D, and i t can be seen that the differences between Types A and B, and Types C and D merely invlve the additin f extra syllables, with n great difference in the rhythm f the calls. The difference between Types B and C hwever, invlves a change f rhythm in syllables 3 and 4, with syllable 3 being nearly twice as lng in B, and syllable 4 being lnger in C. The mean calling rate ranged frm 1.3 t 2.6 calls per minute. Tw further pssible call types were als nted but were nt cnsidered abve. They cnsisted f Types C and D, but each with an extra syllable given at a lwer pitch and amplitude than the previus nes. Severinghaus (1979) distinguished a similar call (his Type 3)- A pssible explanatin fr these call-types is that they were made by birds inhaling after crwing. Gaunt and Gaunt (1977) bserved in the chicken (Gallus gallus), that t terminate the crw call the bird can reverse the flw f air by inhaling "which may be accmpanied by a distinct, smetimes quite lud, wheezing sund". This descriptin sunds similar t the abnrmal Kklass calls reprted abve, and since i t is likely that the tw species have similar tracheal structure and syringeal mechanisms, this inhalatin activity may prvide the explanatin. The call types hwever, d nt allw individual birds t be recgnised by their vice alne. On any mrning, a given bird wuld usually shift call types frm (fr example) A t B t D. Full details f individual birds' shift f call types can be seen in Appendix 10.

133 O CM O -p bo C CD t-1 r<a VO OA v ON OJ I<A v ON OJ -3- OJ KA VO OJ v UA CD rh -Q rh rh >> UA ON rh rh OJ rh d d C m CD CD rh CQ CD! cd p rh OH 01 rh >> UA I OJ rh r<a =»- OA d d O d a, 55 u ca a, H PH -P CD i-l 1 H 1 1 c!>. i r<a C0A VO KA KA m KA OJ OA m m KA rh OJ OJ OJ KA KA KA KA OJ KA KA KA KA O O O OJ -=)- O VO OA KA KA m O rh rh O O O O d d d d d d d d -P c w CD XI PH w rh,m O S 01.C -P 03! -p ho C <D rh c (D rh i rh rh >> KA KA I OJ OJ i rh -=*- KA 00 OJ in VD OA VO KA KA KA rh rh rh OJ rh rh rh rh rh O O O 00 -=1-00 0J KA VO -3- KA KA rh VD rh OJ rh OJ OJ rh OJ OJ OJ OJ rh OJ O O O r<a VO KA VD OJ OA KA rh KA rh m in OJ OJ OJ 0J OJ OJ OJ OJ OJ OJ OJ OJ OJ OJ O O O OA K- rh m -=t- OA in m in rh rh OJ OJ O H rh rh rh O rh rh rh rh OJ rh O O O O O CD CD rh rh X! i rh rh rh rh >> c ON rh f-, CD H } ^ CD rh 55 rh OJ OJ 00 KA OA KA rh KA KA rh rh rh rh rh rh rh H rh rh rh rh O O O O O O ^-^ ^-^, ^ ' v ^ ' v ' ^ > in VO 00 m H rh 00 m rh UA r- 00 t- VO VD m in VO VD in in VD VD rh rh rh rh rh rh rh rh rh rh rh rh r-h H O O O O rh rh r-h rh rh rh rh rh rh rh rh rh rh O rh a) CD PQ O

134 119 The reasn fr this shift is unknwn. All birds began their mrning calling with between ne and furteen Type A calls, these were fllwed by Type B and/r C, after which Type D usually made up the bulk f the mrning calling, thugh smetimes alternating with Types B and/r C. Frm all the recrded calling perids (n=25)* I have ttalled the number f each type f call in Table 20. TABLE 20 Relative numbers f Kklass call-types recrded. Call Type Number f Calls heard Percentage f ttal A B 32 7 B/C C 43 9 D D* 9 2 Ntes: D* is the Type D call with the extra syllable given at a lwer pitch and amplitude as referred t abve. 1 undetermined: either Type B r Type C. Clearly Type D was the call uttered mst frequently in the study areas. This cntrasts with P.m. castanea in Pakistan (Severinghaus 1979) which used call-type B (his Type 1) mst frequently. He als nted that nly tw birds (ut f seven recrded) clearly shifted call types, and that n birds called with Type A. This cntrasts with the results frm this study in which all birds shifted call types and began with Type A. In study species P.m. nipalensis, i t was as thugh the physical act f crwing needed 'warming up' each mrning: a number f shrt, 2-syllable crws had t be made (Type A) befre the f u l l crw call culd be prduced. This explanatin hwever, is cnsidered unlikely (R. Brackenbury pers. cmm.), s i t is pssible that call-type shifting has a behaviural functin.

135 Further, Severinghaus (1979) suggests the pssibility that vcalizatins may help in the identificatin f different races f the Kklass Pheasant. He nted that the calls recrded in K. Hwman's aviaries in England (prbably P.m. macrlpha) differed cnsiderably frm the calls f P.m. castanea frm Pakistan. Similarly, the calls f P.m. nipalensis differ smewhat frm thse f P.m. castanea recrded by Severinghaus. Hwever, nly by detailed recrding and analysis f calls frm different subspecies and habitats culd this hypthesis be tested. Nevertheless, certain individuals culd be recgnised in the area Nrth f the campsite at Pipar. Sme birds had distinctive calling patterns such as ten Type A repetitins befre mre cmplex call types (mst birds used call Type A nly ne t three times each mrning.). Individual recgnitin in a lnger- study culd be used t lk fr the presence r absence f territries r 'hme ranges' i f the birds' calling psitins were mapped in detail. In my study, hwever, lack f manpwer, cntinuity, and the psitive individual recgnitin f nly tw birds allwed nly limited bservatins f this nature t be made (see sectin 7.3.^). Miller (1978) has shwn that in the Red Jungle Fwl (Gallus gallus) there is an acustic basis fr individual recgnitin, and that since the crw call appears t be territrial and pssibly sexual in functin, there is selective pressure fr the birds t distinguish themselves as individuals n the basis f these calls. In the Kklass, calling was greatest at dawn befre sunrise. Figure 10 shws the frequency f calling per 3-minute perid n three mrnings in April and May On these ccasins, the calling perid lasted frm 30 t 39 minutes (mean 3^ minutes), whereas Kklass recrded in Himachal Pradesh in January 1979 (Gastn 1980a)called fr 21 minutes. I t is pssible that the species has a lnger calling perid during the breeding seasn. The shapes f Figure 10 a-c are all quite different,

136 121 and i t must be assumed that variatin in the pattern f mrning calling is cnsiderable. Figure 10(a) shws a flat-tpped peak f calling which is greatest at 32 minutes and 20 minutes befre sunrise. Figure 10(b) shws a distinct peak earlier in the calling pattern, at 31 minutes befre sunrise, while Figure 10(c) shws a mre prlnged calling perid extending t after sunrise, with a peak at nly 18 minutes befre sunrise. The weather was clear with a slight frst n all three mrnings. Taber (19^9)> in his study f the Ring-necked Pheasant (Phasianus clchicus) nticed that as the seasn advanced, the mrning peak f crwing became gradually later, until by late May i t was nly 10 minutes befre sunrise (in mid March i t had been 40 minutes befre sunrise). Althugh I made nly three bservatins f Kklass crwing, they began t shw a similar trend - the peak f crwing becming later, frm 32 min. befre sunrise n 16 April t l8 minutes befre n 3 May. Time f first calling f the Kklass is given in Appendices 5-3* 5.4 and 8, and Figure 16. Calling began 24 t 48 minutes befre sunrise in 1979* and 29 t 53 minutes befre sunrise in It shwed psitive crrelatin with calculated sunrise times, (Appendices 5.3 and 5-4) which were cncrdant with bserved sunrise times. Severinghaus (1979) suggested that "nset f calling appeared t be related t early mrning light cnditins and the extent t which these were influenced by lcal weather". Aschff (1967) and many ther wrkers have described circadian rhythms in the daily nset f activity in birds, and while pheasants prbably pssess such an internal rhythm (Williams and Stkes 1965), i t appears that the exact timing f the calling is at least mdified by external factrs such as weather. Times f calling befre sunrise were divided (as fr the Tragpan) int tw categries, clear and cludy days, and they are given in Appendix 1L On clear mrnings (n= 20), the mean time f calling was 36.2 minutes (S.E. = 1.1) befre sunrise, whereas n cludy mrnings (n=l4), the mean was 30.4

137 Sunrise 0500 UO 0440 x w CSL APRIL ^10 ti u s > 1 M < I < <rfr-»-i-»v. y 1 W V > > APRIL MAY B'igure lb: Times f first call f Kklass Pheasant in relatin t sunrise.

138 minutes (S.E. 0.9). The difference between these tw figures is significant (v=32, t=2.70 p=0.02) and indicates that cludy weather in sme way delayed the start f calling. I t is prbable that lwer light intensity levels n cludy days were respnsible fr this delay, but further research is needed t cnfirm this with the use f light meters t assess light intensity accurately. My bservatins agree with thse f Hwman (1977) and Mirza (1978) in that the birds call fr apprximately 30 t 40 minutes frm their rst sites, and then descended t the grund t feed. In sme cases, certain individual birds called frm the grund while they mved abut feeding. This accunts fr the pattern in Figure 10(c) in which calling is shwn t cntinue after sunrise. The fact that nly certain individuals crw frm the grund whilst feeding wuld help t explain why Mirza made the same bservatin as myself, whereas Hwman (1977) and Severinghaus (1979) did nt. Scattered crwing was ften heard thrughut the day. Gastn (1979) analysed the rate f calling thrughut the mrning at Dachigam Sanctuary, Kashmir, and nticed that, fllwing the pre-dawn peak, there was a minr burst f activity f abut 8 calls per hur between and (1^ t &=r hurs after sunrise). At Pipar, day-time crwing appeared t be mre spradic and was nt analysed, althugh i t was bserved that the sund f thunder, r an avalanche, wuld ften stimulate birds t crw, but nly nce r twice each. This bservatin has been reprted by ther wrkers (Beebe 1922, Fleming et. al. 1976, M.W. Ridley pers. cmm.). Unlike the wailing call f the Tragpan, the Kklass crw was heard in bth spring and autumn. I t was heard frm 3 April t 6 June, frm 22 September t 9 Nvember, and again n 28 Nvember. These dates reflect the time perids I spent in suitable habitat. Gastn (1980

139 124 recrded that the Kklass culd be heard between September and June, but nt during July and August, the mnsn mnths (see Table 7). I. Grahame (pers. cmm. I98O) has recrded captive birds calling frm mid-september t mid-july, and "mst intensively" frm March t June. K. Hwman (pers. cmm. 1980) bserved that first year males called nly during the autumn, but Grahame did nt find them calling at a different time frm the adults. The functin f the crw-call f the Kklass is uncertain. Williams and Stkes (1965) believed that "calling in pheasants (P.clchicus) has l i t t l e r n relatin t aggressive encunters between pheasants either during r after fighting... Pheasant crwing may be essentially sexual in functin". J. R. Krebs (pers. cmm.) and Kacelnifc (1979) have suggested explanatins fr dawn calling in passerines (see sectin 6.2.4), and these are pssible reasns fr Kklass calling. I fund sme evidence that the crw f the Kklass is prbably territrial in functin: playback f the Kklass crw was carried ut n apprximately twelve ccasins in 1979 and On five ccasins a male bird apprached the tape recrder, and n seeing me became alarmed and retreated. Playback usually evked a respnse frm surrunding Kklass in that they wuld answer the crw. On sme ccasins, playback started after the main dawn crwing had ceased wuld result in a secnd smaller chrus f crwing. These bservatins wuld supprt the suggestin f Heinz and Gysel (1970), that a male is alerted t the presence f anther male int its territry. Leffingwell (1928) and Allen (1956) suggested that crwing in P. clchicus may have attracted females, but n females ever apprached playback crwing in this study. Harrisn and Wayre (1969) reprted that in Kklass display in captivity, the male crwed nce r twice during relaxed mments. Severinghaus (1979) gave a summary f phnetic renderings f the crw call f the Kklass Pheasant as described in the literature. The

140 125 races described were P.m. macrlpha, P.m. castanea, and P.m. nipalensis. The phnetic renderings were very variable, ranging frm a simple "Kkras" (Pultn 1904) t "Ah! craaki craak-craak! crk!" (Beebe 1922). Such descriptins f bird sng and calls are highly subjective, and I cnsider the variatins that Severinghaus tabulated are due almst entirely t persnal interpretatin. Hwever, a survey f all the wild races f Kklass including the Chinese subspecies might shw subspecific differences in their calls; and pssibly these culd be used in delimiting their gegraphical distributin, as Severinghaus suggests, since the species shws marked clinal gegraphic variatin. 5) 'Oww' call f the female On a number f ccasins in spring and autumn 1979> a sft 'reply' call was heard immediately after the crw call f the male. Phnetically, the reply culd be described as a sft 'ww 1 r 'kerwakw'. I t was suspected that the call was being uttered by a female bird rsting r feeding clse t the male. This was prved n 12 April, 1980 when I bserved a pair at dawn at a rst site (see sectins and 6.3.4). The call culd be interpreted as a cntact call, infrming the male that the female was clse by Breeding Behaviur Curtship Display Display was nt bserved during this study. In captivity i t has been described briefly by Delacur (1977): "he (the male) struts abut, puffing ut the bdy feathers and erecting vertically the lng black tufts f the sides f the head, the brwn crest raised between them." Harrisn and Wayre (1969) have described the display f the Kklass in cnsiderable detail which I shall nt repeat here, and cnsidered that the display t the female was principally lateral with sme clucking

141 126 calls (Type 3), but generally silent. In captivity, display utside the breeding seasn has been bserved frm late December nwards (I. Grahame pers. cmm. 1980), and until September (K. Chalmers Watsn pers. cmm. 1980). Mating Systems Out f thirty Direct Encunters with Kklass, fur were f pairs f birds, tw were f single females, and the rest (80$) were f single males. N birds were seen in grups, and althugh cmparatively few bservatins f the species were made, i t appears t be mngamus. This view has als been taken by mst previus wrkers, including Hume and Marshall (1879), Beebe (1922), Bates and Lwther (1952), Ali and Ripley (1969), and Delacur (1977). The first tw wrkers suggested that birds pair fr a number f years. Fr avicultural purpses they are usually kept in ratis f males t females f 1:1 r 1:2. K. Hwman (pers. cmm. 1980) has kept ratis f up t 1:5, but breeding success in these cases is unknwn. Gregariusness In this study, the Kklass was fund t be very slitary, even after the breeding seasn. Only fur bservatins (13$) were made f mre than ne bird, and there were n bservatins f mre than tw birds tgether. Hwever, frm the mre extensive bservatins dcumented in the literature, i t appears that pairs f birds are nt infrequently encuntered. Hume and Marshall (1879) and Blanfrd (1898) referred t single birds and pairs in spring, with family parties remaining tgether thrugh the autumn and winter. Ali and Ripley (1969) nted "thugh nt gregarius, several birds frequently haunt a particular hillside... day after day"; and Delacur (1977) stated that "during the winter, numbers f adults can be flushed within a shrt distance f ne anther, but they d nt live in parties".

142 6.4 Himalayan Mnal Wariness This species was the mst regularly bserved f all the pheasants I studied in Nepal. The cumulative bservatin time was 1100 minutes in 108 Direct Encunters. Hwever 57 sightings (53$) were f escaping birds, which were viewed fr nly a matter f secnds befre they flew ut f sight, while nly 14 (13$) were fr mre than 20 minutes. The fact that birds lived principally n grassland and less in the adjacent frest meant that they were very wary n pen slpes and wuld smetimes flush when a human intruder was a lng distance (300m) away. i f an bserver was able t lcate a Mnal withut being seen Hwever, first, he culd then sit s t i l l and view i t fr a lng perid f time (up t 125 minutes) withut disturbing i t. I t appears that the wariness f the Mnal is predminantly the result f hunting by Man (as i t may be in the ther species). In the Everest Natinal Park (East Nepal) the Mnal is nt trapped r hunted in any way because the inhabitants f the muntains in this regin are the Buddhist sherpas, wh d nt k i l l birds t eat. In additin, in this area there are villages up t 3800m altitude, well within the range f the Mnal. The result is that the species is accustmed t human presence and has l i t t l e fear f Man. In March 1980, the Mnal in the Everest regin tlerated slw apprach by a human t within abut 8m withut alarm, but clser apprach resulted in grund escape r flying. This tameness appears t be peculiar t the Natinal Park regin; elsewhere in Nepal (R. L. Fleming Jr. pers. cmm.) and in the Western Himalayas (A. J. Gastn pers. cmm.) the species is very wary, and flushes n sight f Man. There is l i t t l e reference t the species' wariness in the literature,

143 128 but even in the last century Wilsn (in Hume and Marshall 1879) reprted that "wherever they are rare, they are als sure t be very wild and shy". Bates and Lwther (1952) als refer t them as shy and nt ften seen Prtective Behaviur Escape Mst f the bservatins f escaping birds in the study areas were f birds flying. The nly grund escapes seen ccurred when an intruder was s t i l l distant (mre than 150m) and the bird culd quickly run ver a ridge r int a gully ut f sight. When a bird was disturbed in frest, the escape was always by flushing - the bird flying up thrugh the canpy t alight ut f sight, usually in a tree. Flushing n pen grund always resulted in the bird flying dwnhill, ften fr up t j500m. This frm f escape was always accmpanied by alarm calls in the female Mnal, and usually als in the male. In cntrast, the Mnals bserved in the Everest Natinal Park in March 1980 usually escaped by walking r running, bth n pen slpes and in the frest. Wilsn (in Hume and Marshall 1879) reprted that "in the frest, when alarmed, i t generally rises at nce withut calling r running far n the grund; but n grassy slpes i t will, i f nt hard pressed run r walk slwly away in preference t getting up". He als recrded that when suddenly alarmed i t flushed with much calling. Bates and Lwther (1952) nted that "when disturbed the Mnal hurls itself dwnhill emitting a vlley f shrill whistles and ften flies a cnsiderable distance befre sweeping t rest". Ali and Ripley (1969) gave a similar accunt and als bserved: "when suddenly cme upn in the frest, especially i f accmpanied by small chicks, i t flies up with much cackling int... a tree and freezes".

144 129 'Freezing Behaviur' This was bserved n many ccasins, but usually nly nticed nce the bird began t mve again. Invariably, when a bird was cme upn suddenly in cver, i t wuld mmentarily 'freeze' (r smetimes 'freeze' until i t was apprached clsely), and then flush. On pen slpes, birds wuld ften 'freeze' n first seeing an intruder, then, depending n whether i t apprached r retreated, wuld flush r remain s t i l l. Lengthy 'freezing', such as that bserved in the Satyr Tragpan was never seen in the Mnal. Cryptic Clratin The female Mnal shws gd cryptic clratin, being almst unifrm brwn abve and belw, with a white rump. She carries ut the incubatin f the eggs (Rberts 1970), as wuld be expected, since the male is brightly clured and iridescent. His phaneric cluring is used t full advantage in the curtship display (sectin 6.4.5), but such extic plumage seems incngruus n a bird which spends much f its time n pen slpes where predatin appears t be easy. Hwever, the Mnal appears t have a pattern f feeding behaviur whereby it feeds n pen slpes nly in pr light cnditins (see sectin 50»^)» I t is pssible that the habit f living n pen slpes is a relatively recent develpment in the species' evlutin, and that i t was riginally strictly a frest bird (as wuld be suggested by the male's bright plumage). Its present habitat may be the result f frest clearance and disturbance by livestck grazing and herders in the summer mnths (in the study areas and in India), and by general disturbance by humans (in the Everest regin).

145 130 Interspecifi Cmmunicatin Like the ther pheasants studied, this species is thught t respnd t the alarm calls f ther species. On 11 April, 1979* while a male Mnal was being watched, a Large Scaly-bellied Wdpecker (Picus squamatus) was heard calling a number f times. The Mnal lked up, very alert, each time the call was made. T the human ear, the call f the Wdpecker sunded similar t the plaintive 'whistle' call f the Mnal (see sectin 6.4.4, call Type 2). On 26 April, 1979 I disturbed a male Mnal at clse range, and i t flew ff a crag dwnhill uttering the 'piping' call very ludly. I nticed that, as i t began the alarm, the numerus passerines (including Sunbirds Aethpyga, Leaf Warblers Phyliscpus, and Black-capped Sibias Heterphasia capistrata) feeding in the trees belw, immediately alarm-called briefly in respnse. Once the Mnal had flwn ver (a matter f secnds), the passerines resumed their nrmal feeding chatter. On 13 April, 1980 a pair f Mnal were being watched feeding n pen grassland abve frest. Anther Mnal was heard calling frm the frest whereupn a Serw (Capricrnis sumatraensis, a gat antelpe) began barking in alarm. The pair then stpped feeding and std upright in an alert psture, lking arund, and facing dwnhill twards the Serw. After three minutes bth birds flew ff silently dwnhill Daily Life Mbility Levels Table 21 shws the frequency f sightings f Mnal thrughut the day. Unlike the pattern fund fr the ther species f pheasants studied, this is nt cnsidered t be heavily biased, since a large number f sightings are invlved, and the time spent searching fr

146 Mnals was spread thrughut the day. The results were analysed by recrding the number f bservatins made fr each hur f the day, frm t hurs. On this basis there was n significant difference in the frequency f sightings between any hur f the day (X = v=l~j>, p=0.1). Figure 17 illustrates the results as a histgram, with the day divided int five three-hurly intervals, as fr the ther species. I f the frequency f sighting is indicative f mbility levels, then the results imply that the Mnal is mbile thrughut the day. The histgram hwever, des shw a greater number f sightings in the mrning, signifying a greater mbility during the perid t 10.00, but this is nt statistically significant. Sme mbility appeared t be related t feeding - see sectin TABLE 21 Sightings f Himalayan Mnal in relatin t Time f Day, 1979 and 1980 Time f Day Number i5.oo-l6.oo

147 ( 1900 TIME (HOURS) Figure 17: Histgram f pattern f sightings f Himalayan Mnal thrughut the day.

148 133 Rsting Mnal were heard and seen rsting bth n rcks and in trees. In early April 1979 a bird was flushed frm a Whitebeam (Srbus sp.) befre dawn, and birds were subsequently heard calling frm wdland rsts n a number f ccasins. Mnal were tape-recrded calling at dawn frm rsts n crags at Pipar in early May, at Hhalu in early June, and at Kurnai in late Octber 1979 At Bhalu, n a crag rst at c3350m altitude, apprximately 30 Mnal drppings were fund, indicating lengthy rsting at this sight. Cmmunal rsting was nt bserved, but cragrsting birds were heard calling clse tgether. On 11, 1J and 14 April, 1980, a pair f birds which had been watched feeding fr up t 55 minutes each evening, flew ff tgether at twilight (abut hurs) dwn t the frest, prbably t rst tgether. Wilsn (in Hume and Marshall 1879) referred t the Mnal rsting in large frest trees "but in summer n grund r rcks". Beebe (1926) nted that they rst in trees "cmmunally but alne", and that in the summer and. early autumn the parents and yung rst tgether. Since the birds are knwn t ascend the muntains during the spring and summer as the snw retreats, i t wuld appear necessary that they rst n rcks when abve the treeline, which wuld accunt fr Wilsn's bservatin Vcalizatins I have distinguished three types f call made by the Mnal: 1) 'Piping' call - This is a successin f high pitched piping ntes beginning very rapidly, and becming mre spaced ut twards the end f the call which lasts up t 10s, (See Appendix 4.10). Each nte lasts fr abut 0>l4s and the intervals between ntes begin at 0.10s, lengthening t apprximately 1.0s as the call prgresses. Each nte is

149 134 cmpsed, f an upward and dwnward slurred sund with a fundamental peak at abut 3500 hz, and a set f vertnes at 7000 hz. This call was ften given by alarmed birds especially when flushed. (All flushed females and seven {11%) flushed males). I t was als given by birds n the grund, and in these cases appeared t have a different but unknwn functin. At dawn, the call wuld accmpany the 'crescend' call (Type 3)* althugh the birds appeared nt t be alarmed (see under Type 3) Beebe (1922) accurately described this call as "a rapid successin f s h r i l l, screeching, whistling ntes multiplied terrr-induced mdificatins f the cmmn call nte: weep! weeep.' weeeep] weeeeep!" 2) 'Whistle' call - This is a lud high-pitched whistle, either single r duble tned, similar t the call f the Curlew (Numenius arquata). Appendix 4.11 shws the structure f this call, apprximately 0.5s in length, with a fundamental frequency at abut 2300 hz and vertnes at 4800 hz and 7300 hz. The call was usually heard as an extensin f the 'piping' call - the ultimate nte f the piping was lengthened int a pure tne and repeated at intervals f ne t five secnd t frm the 'whistle' call. The usual sequence f behaviur was that the bird wuld flush uttering the 'piping' call, f l y dwn, then alight while s t i l l calling. The call wuld then develp int the 'whistle' call, which wuld be uttered fr up t five minutes, while n the grund r in a tree. This call was never heard uttered by a bird in flight, and was actually seen t be called nly by females/immature males. Hwever, since adult males were bserved calling Types 1 and 3* i t is assumed that they t made this call, althugh perhaps less frequently than females. The functin f the 'whistle' call is uncertain. Since i t was an extensin f the 'piping' call (which was ften uttered as an alarm

150 135 respnse), i t may have an alarm functin - pssibly in the nature f an 'anxiety call'. Mnals were seen t respnd t the 'whistle' calls f ther Mnals n a number f ccasins by becming alert and uttering the same call. This respnse was als bserved by M. W. Ridley (pers. cmm.) in Himachal Pradesh, Octber Hume and Marshall (1879) made very similar bservatins t mine regarding this call. Beebe (1922) referred t the cmmn call nte as: "a s h r i l l lud whistle, with but l i t t l e cheeriness". Others, such as Hellmich (1968) and A l i and Ripley (1969) have likened the call t that f the Curlew. 3) 'Crescend' call - This cnsisted f a piping whistle f tw t five ntes, each set repeated faster, with increasing frequency and amplitude t a crescend set, fllwed by a slwing and diminuend f sets. Appendix 4.12 shws ijf sets f three ntes: each ne has a fundamental frequency at hz with sets f vertnes at hz and 690O-76OO hz. Ntes last apprximately 0.30s, and intervals between ntes in a set range frm 0.10s t 0.15s. This call was uttered by bth sexes bth in spring and autumn, thugh i t is prbably made thrughut the year, except during the mnsn (see Table 7). The call was given principally at dawn like call-types (4) and (3) f the Tragpan and Kklass. Quantitative measurements f calling frequency in this species were nt btained, but i t appeared t be irregular and spradic until abut hurs. T a lesser extent, calling tk place at twilight, but again n systematic ntes were taken fr this species. M. Sawyer (pers. cmm. I98O) has heard captive Mnal calling a l l year rund, with sme birds calling at night, (n time specified). Prm my bservatins in the wild,calling during the day depended n the weather - if i t was clear and sunny, Mnal wuld rarely call by day except fr calls 1 and 2 if disturbed. Hwever, if

151 136 precipitatin was falling and/r i t was misty, spradic calling culd be heard. The irregular dawn calling f this species was bserved by Beebe (1922): "The early mrning calling is rather inexplicable. I have knwn f 4 birds, ld and yung, rsting at a certain place night after night, and yet, early in the mrning, a l l 4 wuld regularly g thrugh a perid f repeated calling. The calling was less nticeable and f shrter duratin when the dawn was lwering and cludy than when... bright and clear. I t seems prbable that i t can nly be a mere cncmitant f the nervus excitement f awaking and preparing fr anther day". Gastn (1980a)nted that "dawn calling ccurs spradically thrughut mst f the year, but appears t bear l i t t l e relatinship t the numbers f birds present". The times f f i r s t calling f the Mnal in 1979 are listed in Appendices 5-5 and 8, and Figure 18. Calling began minutes befre sunrise and, like the Tragpan and Kklass, shwed psitive crrelatin with calculated sunrise times (r= v=31 p=0.001). Onset f early mrning in relatin t weather was als tested t see i f there was a relatinship between the tw, as in the Kklass. On clear mrnings (n=23), the mean time f f i r s t calling was Jh.K minutes (S.E.=1.4) befre sunrise, while n cludy mrnings (n=5), the mean was 23.5 (S.E. = 3.2) (see Appendix 12). The difference between these tw times hwever, is nt significant (v=26 t=0.84 p=0.1), indicating that the time f f i r s t call was nt related t early mrning weather cnditins. Hwever, the number f cludy mrnings was very small (5), and i t is pssible that this is the reasn fr the nn-significance. The functins f the 'crescend' call are mre d i f f i c u l t t assess than the dawn calls (Types 4 and 3) f the Tragpan and Kklass. I t is pssibly t e r r i t r i a l in functin, but since i t als uttered utside the breeding seasn, when the t e r r i t r i a l instinct is prbably drmant (sectin 7-4.4), this explanatin is nt very likely. As in the

152 en e Sunrise e ^0440 e e 0400' ^ - M I > APRIL nrp , MAY JUNE Figure 18: Times f first call f Himalayan Mnal in relatin t sunrise.

153 138 discussin f the Tragpan and Kklass calling may be related t the fact that fraging cannt begin until the light intensity reaches a certain minimum value (Kacelnik 1979). This bservatin has nly been made in passerines hwever, and at present the functins f Mnal calling remain bscure Breeding Behaviur Curtship Display The male's display was seen n tw ccasins: n 17 April, 1979 at hurs, with n ther Mnal in sight, a single male std and fanned his t a i l. He then began jerking his bdy and raising his wings slwly abve his head while hlding his fanned t a i l erect. The silent display lasted fr 10 t 15 secnds, whereupn the bird walked up the slpe feeding casually and was nt seen t display again. On 21 April, 1980, at hurs, a male was bserved displaying in the same manner t a female which was feeding in tussck grassland. The birds in this case were slightly bscured by vegetatin and it is pssible that cpulatin tk place. Tw further advances t the female resulted in the latter retreating t feed further away frm the male. Hwever, when he flew apprximately 150m acrss the slpe, she jined him and bth cntinued feeding. On a number f ccasins in spring 1979* and n 13 and 14 April, 1980 male birds were bserved making a "display f l i g h t ". The Mnal i n i t i a l l y std n a prminent rck n an pen slpe tussck grassland. I t then suddenly flew ff, gliding hrizntally but slwly dwnwards with i t s wings held high abve its bdy, and with the white patch n its lwer back prminently displayed. While gliding dwn the bird wuld call sftly, uttering a similar sund t a muted Type 1 'piping' call. It wuld then alight n a rck and lk arund. In I98O, a female was bserved feeding n the same slpe n ne ccasin f display flight.

154 139 The bservatins f Finn (1915), G. S. Rdn (in Beebe 1922), Whistler (1926), Bates and Lwther (1952), and Ali and Ripley (1969) were f an essentially frntal display similar t my wn sightings. Delacur (1977) described the lateral and frntal display seen in captivity. "He (the cck) appraches the hen with lng grping steps, describing wide circles abut her, always with the inner wing lwered t such an extent that the feet and legs are cmpletely hidden. The neck is extended, the b i l l held dwnwards clse t the neck, the crest standing erect and vibrating. As he gets mre excited he pecks nervusly at the grund. He then assumes a frntal psitin and the t a i l is raised and spread t its widest extent, with head and beak t the grund, wings spread, the bird bwing rhythmically frward and back. The feathers f the back and wings are nly slightly raised, but the mantle and neck plumage f l u f f ut, encircling the head. It shws t the best advantage the beautiful metallic clurs, the white back and the rufus t a i l ". Bth frntal and side display have been bserved by a number f Wrld Pheasant Assciatin members pssessing captive Mnal. Whistler (1926) and Ali and Ripley (1969) als described the brief display f l i g h t f the male referred t abve, and in September 1980, members f the Himachal Wildlife Prject ( ), reprted bserving display f l i g h t by males (M.W. Ridley pers. cmm.). Ridley als bserved that the Mnal is unique amng pheasants in having a display flight. Mating Systems The main prblem f attempting t define the mating system f the Himalayan Mnal is the apparent similarity f first-year male t female birds. Wilsn, (in Hume and Marshall 1879) reprted that yung males resemble females fr the f i r s t year, but that the plumage has a darker and mre glssy appearance. When changing their plumage during the first mult, they "appear sptted a l l ver with the metallic hues f the adult".

155 140 He als cnsidered that at fur mnths ld, yung males were "easily distinguishable" frm females and this view is supprted by present-day keepers f captive Mnal. Hwever, I fund that in the field I culd nt distinguish between yung males and females, either in the spring r autumn; the birds appeared t be either distinct males r distinct females. Members f the Himachal Wildlife Prject wrking in the Himachal Pradesh, India, als culd nt distinguish between females and immature males (P. J. Garsn pers. cmm.). Prm my wn bservatins f 38 Direct Encunters f mre than ne Mnal tgether, the fllwing is a l i s t f the numbers f birds bserved: TABLE 22 Numbers f Mnal Observed during the Direct Encunters. Mre than Mre than 1 male with 1 male with MuJ-^P- 1 - e Pairs ^ : -, = males Unknwn 2 females 2 males 2 females 3 females 5 dj. females Because I culd nt distinguish yung males, i t is pssible that sme were mistaken fr females. Hwever, since mre cnsrting pairs were seen than any ther sex rati, my encunters suggest that the Mnal des pair in the nrmal sense and is mngamus. This view hwever, is nt generally supprted by reprts frm the literature: Wilsn (in Hume and Marshall 1879) stated that " i t may be questined whether they d pair r nt in places where they are at a l l numerus; i f they d, i t wuld appear that the unin is disslved as sn as the female begins t s i t, fr the male seems t pay n attentin whatever t her whilst sitting, r t the yung brd when hatched, and is seldm fund with them". Wilsn, (p. cit.) tk a similar view, while ther wrkers, such as Beebe (1922) d nt refer t i t. All and Ripley (I969) and Delacur (1977) bth refer t prbable plygamy but reprt n new evidence t

156 141 supprt the idea. In respnse t my questinnaire (sectin 3*3), seven members f the WPA wh pssess Mnal in captivity reprted keeping birds in a sex rati f 1:1 fr the best results. The abve evidence indicates that the Himalayan Mnal is prbably mngamus, but that the male bird takes n part in incubatin r care f the yung, and is in fact segregated frm them fr much f the year (Beebe 1922). In certain regins where the sex rati in the ppulatin is biased s that there is a greater number f females (fr example where males are heavily hunted, such as Pakistan), i t is pssible that plygamy may take place. Only intensive study f this species in the wild w i l l definitely establish the nature f its mating system. Gregariusness The Mnal is prbably best termed as lsely gregarius species. They were usually encuntered singly (70 Direct Encunters, 65$), r in pairs (18 Direct Encunters, 17$) and less ften in greater numbers (see Table 22). Hwever, n a number f ccasins, I encuntered up t 11 birds irregularly spaced acrss a slpe r set f crags at Kalki Danda and Namrung (Pipar), and als at Krchn in Nvember 1979» These birds, while nt cncentrated in a flck, did shw a degree f gregariusness and mst wuld respnd t an alarm call given by ne f their number. These bservatins were a l l made befre hurs, and i t is believed that after this time the birds dispersed t feed ver a wider area, and did nt cngregate again until evening. Hume and Marshall (1879) did nt refer t the sciability f the Mnal, but nted that in autumn and winter the birds are f a i r l y dispersed in frests, and that the females keep tgether mre than the males. Blanfrd (1898) reprted that the Mnal were "nt in flcks r cveys, but singly r in tws r threes". Beebe (1922) stated they "seem t a

157 142 large extent gregarius, but ties between members f a flck are extremely lax". He bserved a grup f furteen males digging tgether in spring, while Fleming et. al. (1976) recrds "a dzen ccks" feeding tgether Aggressive Behaviur Aggressin between males was bserved n ne ccasin, 7 May, 1979* n Kalki Danda ridge between and hurs. One male appeared frm the Suth side f the pen ridge, crssed ver, and walked int the wdland n the Nrth side, jining tw mre males there. The birds mved ut f sight, and Types 1 and 2 calls were heard, accmpanied by much fluttering. When the birds came back int view, I saw that tw males were fighting by flying at each ther and attacking with their b i l l and spurs. They were als flying up int lw branches f the trees intermittently. The third male tk n part in the fighting, but stayed with the ther tw and fed ccasinally. All three birds mved slwly up slpe in the wd t the Nrth-West, with the fighting cntinuing spradically, until they mved ut f sight. Ram Kaji Mangar (a sherpa) had seen similar behaviur between tw males n the ridge tp n the previus mrning. Delacur (1977) recrds aggressive behaviur by Mnals in captivity. I f females were present he fund that the dminant (ldest) male wuld always chase away r k i l l the yunger nes. Hwever, with n females present he fund that several males wuld live in peace fr several years. M. Sawyer (pers. cmm. I98O) made the fllwing bservatins n three pairs f Mnal kept tgether: "Beginning f April, 1 cck becming dminant, chasing his brther abut, but ignring cck that was resident when a l l put tgether. Threat psture when sighting anther cck was t puff up feathers, especially neck, and t run 7-8 paces frward then stpping, t a i l fanned ut. As seasn prgressed, attacks mre ften.

158 143 Split up: dminant cck with 3 hens. Resident cck then chased ther slightly, but mre interested in hens in adjacent pen". 6.5 Cheer Pheasant Since nly 4 Direct Encunters with Cheer Pheasant were made in this study, my wn bservatins are very limited. This sectin is therefre drawn largely frm the available literature Wariness and Prtective Behaviur The Cheer is knwn t be a very secretive and wary species. Despite searching suitable habitat in August 1979 n birds were fund, while in May I98O Cheer were seen nly when they had drawn attentin t themselves by calling r flying. While earlier authrs d nt specifically refer t the wariness f the species, mre recent nes d: Ali and Ripley (1969) refer t i t as "an extreme skulker" while Rberts (l97) relates that "they rely fr survival n a habit f extreme cautin and wariness". A l l fur bservatins f escaping birds were by flushing, which is in cntrast t mst f the recrds frm literature. The Cheer is well knwn fr its habit f squatting dwn, remaining cncealed and mtinless until the last pssible mment befre flushing. This practice f cncealment and great reluctance t flush has been reprted by mst wrkers, including Hume and Marshall (1879), Beebe (1922 and 1931), Ali and Ripley (1969), Rberts (1970) and Fleming et. al. (1976). They als refer t its marked preference fr escaping by ft uphill, rather than flushing. Hwever, when the bird des flush, i t is extremely rapid in i t s descent dwnhill with i t s wings pulled in at the sides. Hwever, my nly gd bservatin f a Cheer was ne f walking acrss a steep grassy slpe, calling, and after 3 minutes flying dwn int scrubby ak frest.

159 144 Cryptic clratin f bth sexes is well develped. They are generally barred buff, rusty black and grey which is ideally suited t the pen grassland and scrub which they frequent. Beebe (1926) bserved that "they can be clsely apprached due t their cmplete trust in their cncealing plumage". Their rsting habits have been reprted cnflictingly. Hume and Marshall (1879) stated that they perch l i t t l e in trees, but went n t say that they usually rst abve grund. Beebe (1922) reprted that they usually rst n the grund "smetimes in trees and bushes", while in 1926 he bserved "never seen t rst in trees", but n the grund. Hwever Ali and Ripley (1969) reprt Cheer rsting in cmpany in patches f ak frest, while Rberts (1970) recrded them rsting in t a l l trees n the frest edge. I cnsider that bth are pssible, but bearing in mind that the birds live in steep rcky areas, they are mst likely t rst n small crags. In May 1980, birds were heard calling frm c l i f f rsts abut 60m abve the h i l l slpe near Muri village Vcalizatins The Cheer has a varied and nisy repertire f clucks, cackles, and crws, but the principal call can be rendered phnetically as: 'Chk-chk-chk-cherweewa-cherweewa', with the 'wee' syllable uttered at a higher pitch than the rest f the call. The call is illustrated in Appendix 4.13, the f i r s t syllable 'chk' having a wide slurred frequency range with the fundamental at abut hz. The f i r s t syllable f the secnd part f the call 'cher' is similar t the 'chk' syllable in harmnic structure, with a fundamental at hz and sets f vertnes at 4500 hz and 6500 hz. The 'wee' syllable cnsists f a frequency mdulatin frm 1500 hz t a peak at 5000 hz and back dwn t 1500 hz. I t appears t have a harmnic at mre

160 145 than 8000 hz. The final 'wa' syllable cnsists f a small frequency mdulatin at abut 2000 hz. The f u l l 'cherweewa' call last apprximately 0.4s, and is repeated at intervals f abut 0.5s. Mst authrs (Hume and Marshall 1879, Blanfrd I898, Finn Beebe 1922, Ali and Ripley 1969, Delacur 1977) reprt the call f the Cheer t be a very varied 'Chir-a-pir, chir-a-pir, chir, chir, chirwa, chirwa". Whistler (1926) gave a mre detailed descriptin: "a series f nisy squeaks and chuckles, which ring ut... with the clamur that ne is accustmed t assciate with a Guinea-fwl. Sme f the ntes resemble thse f that bird; thers are an exaggerated versin f the squeak f the Silver Pheasant, while sme recall the Chukar". In my study, Cheer Pheasant were heard calling nly in the mrnings, as fllws: 14 May intermittently until 08.45, 15 May briefly briefly, lb May intermittently until May t The fact that the species was heard calling nly in the mrnings was incnsistent with mst reprts in the literature which state that it is mst vcal at dusk (including Beebe 1926, Whistler 1926, Gastn and Singh I980). Playback f the nrmal call f the Cheer was carried ut n tw ccasins during the day n 15 and 16 May, but i t elicited n respnse frm the birds. Beebe (1922) als refers t three mre call types nt heard in this study:

161 146 1) Plush alarm: "a series f screeching chuckles", 2) Cntented calls: "a lw sleepy 'waaaak, waak, waak" 1, 3) Suspicin: "a sharp 'tuk tuk tuk'" Breeding Behaviur Display was nt bserved in the wild. Delacur (1977) reprted that "the cck's display is lateral; i t is simpler than that f the Lng-tailed and Cmmn Pheasants, lacking the vibratin f pen wings r erect pstures, and is generally similar t that f the Eared Pheasants" (Crssptiln). The Cheer is reputedly mngamus (Ali and Ripley 1969» Delacur 1977) and they are usually kept in male:female ratis f 1:1 in captivity, r ccasinally 1:2. They are gregarius thrughut mst f the year, and have been seen in cveys f between five and fifteen (Beebe 1922, Ali and Ripley 1969, Delacur 1977), thugh in this study, nly single birds were seen. Tw r mre birds hwever, were heard calling frm the same pints n three ccasins. This pheasant is prbably the least studied f the Himalayan species, and fr this reasn the bservatins made in the literature are bth scanty and ut f date. Since i t is an endangered species, further research int its status, bilgy, eclgy and behaviur is cnsidered t be f great imprtance.

162 BREEDING BIOLOGY AND TERRITORIALITY In my prject infrmatin gathered n breeding bilgy and territriality in the pheasant species was limited, due t difficulties f bservatin and nest lcatin. The fllwing shrt accunt is based n my wn field bservatins cmbined with the available accunts frm the literature. 7.1 Bld Pheasant Breeding Seasn The exact timing f the Bld Pheasant's breeding seasn prbably varies cnsiderably thrughut its gegraphical and altitudinal range. With regard t the Himalayan subspecies, Hdgsn (in Hume and Marshall 1879) was infrmed by Nepalese villagers that the breeding seasn lasted frm April t May, and that the yung flew in July. Hker (1854) hwever, saw yung birds in May in East Nepal. E. M. Bailey (in Inglis 1931) bserved a captive pair at Gangtk (Sikkim) in 1924, which laid eggs in late May and incubated fr 29 days. J. Rberts (in Grahame 1976) reprted frm the Everest Natinal Park that pairing f Bld Pheasants was cmpleted by the end f April, and eggs were fund n 13 May. Clearly, altitudinal and climatic factrs play an imprtant rle in the timing f the breeding seasn in different areas. In my study areas, i t was bserved that pairing had taken place by the beginning f April, and mating was bserved n 19 April, I t is likely that the cryptically clured females carry ut the bulk f the incubatin, and, in supprt f this, single males were seen feeding alne frm 27 April nwards in bth years. Yung birds were seen nly in 1979* when tw adult pairs were bserved with tw fledged and tw unfledged yung n

163 148 3 June, and a ttal f five unfledged chicks were seen accmpanied by six adults n 5 June. In the Himalayan regin i t appears that breeding takes place between late March, when cveys break up t frm pairs, and early July, by which time mst yung have fledged Eggs The eggs are "pinkish buff, prfusely speckled and bltched with rich brwn" (Delacur 1977)> but are "extremely variable in clur" and average 48mm x 33mm in size (Grahame 1976). Clutch size is 5 t 12 (Delacur 1977) N eggs were fund during my study Nest Hdgsn (circa 184-6, in Ali and Ripley 1969) reprted that a lse nest f grass and leaves is placed n the grund in grass and bushes. Inglis (1931) recrded the captive Bld Pheasants at Gangtk making a nest as "a mere depressin in the grund with a few dead grass stems will pressed int i t... well cncealed beneath the bush". Accrding t Delacur (1977): "The nest is a hllw in dead leaves r mss, usually near a bulder, a stump, r under a bush". N nests were fund during my study despite extensive searching Territriality The cncept f territriality r fixed hme range in the Bld Pheasant has nt been put frward in the literature. In the Pipar study area, i t was nted that during the spring certain pairs f birds frequented very lcalised areas during April and May. In 1979* tw pairs were individually recgnised (n the basis f the number f leg spurs n the male bird), and f these, ne pair r a single male was bserved in the area belw the main camp site n at least twelve ccasins. Similarly, in 1980, anther pair was seen clse t census

164 149 pint 4 n at least five ccasins. This circumstantial evidence indicates that individual pairs may restrict their activities t a definite area during the breeding seasn, which might be termed their 'hme range'. I t is likely that this area is nt defended, and verlaps with the hme ranges f ther Bld Pheasants. This is indicated by the bservatin f temprary cveys f up t three pairs within areas thught t be the hme range f ne pair; n aggressive behaviur was bserved n these ccasins, althugh sme inter-sex chasing tk place. Althugh n bservatins were made during the mnsn seasn, it is likely that hme ranges break dwn after the breeding seasn. Cveys cnsisting f family grups prbably frm and range ver wide areas t feed, thus accunting fr the relatively few sightings f Bld Pheasants in the 1979 autumn seasn. The cveys prbably persist thrughut the winter, until pairing begins in the fllwing Spring, and hme ranges are again acquired. 7.2 Satyr Tragpan J.2.1 Breeding Seasn Due t its wide altitudinal range (2230m t 3550m), the timing f the breeding seasn in the Tragpan shws cnsiderable altitudinal variatin. I t is reprted t last frm May until June in the wild (Ali and Ripley 1969), while in captivity, males have been bserved displaying in March (Delacur 1977)- Hellmich (1968) sht females n 13 and 16 May, bth f which had well develped varies. In the study areas, the breeding calls f the male (sectin 6.2.4) were heard as early as March, (J. Rberts pers. cmm. 1979)* and cntinued thrughut April, May, and early June. In 1979* mating was bserved n 16 May at 3300m altitude at Pipar, while belw Bhalu, Rinzing Sherpa saw a female Tragpan with fur fledged yung n 31 May at 2640m. At Bhalu camp

165 150 site itselfj we fund a Tragpan's nest at 3l60m which cntained three fresh eggs n 6 June. These three pieces f breeding evidence shw the variability in timing which can exist in the wild. Like the Bld Pheasant, the full breeding seasn prbably runs frm late March, when the breeding calls f the male begin, until early July, when mst yung shuld have fledged Eggs These are "reddish buff, freckled all ver with deeper brick red. Size c65mm x 42mm" (Ali and Ripley 1969). Clutch size 3 t 6; incubatin perid 28 days (Delacur 1977). In my study, nly ne nest was fund: i t cntained tw eggs n 4 June, and three eggs n 6 June, Success f hatching was nt determined because we had t leave the area three days later Nest Members f the Tragpan genus are reprted t nest in trees, ften high up, and t use ld nests f Crws and ther birds in which t depsit their eggs (Delacur 1977) Females apparently carry grass, twigs, and leaves up t the nest and line i t with these items (Beebe 1922, Harrisn 1968 and 1969, Mdy 1976, Delacur 1977) At Bhalu, the nest fund was at 3l60m altitude, but n the grund in lng tussck grass apprximately 10m abve the edge f Berberis scrub adjacent t Rhddendrn frest. The nest was merely an unlined depressin in the grass under a large tussck (abut 0.5m t a l l ), n a slpe f abut 45 f pen grassland with small patches f rck expsure. The nest was lcated by chance, as the female bird flushed silently ff the nest dwn int the frest when tw peple were walking past. On a subsequent visit tw days later, the sitting bird did nt flush until we apprached t within ne metre.

166 Territriality The existence f territriality in the Tragpan has nt been reprted in the literature. At Pipar, certain individual males culd be recgnised by their calling patterns in the breeding seasn (sectin 6.2.4). This enabled these individuals t be lcated at dawn, and i t was nted that ne bird always called frm the same grup f trees, fr up t j8 cnsecutive mrnings. Other birds were nt studied in such detail, but up t five called regularly frm certain lcatins, and n tw birds were ever heard calling frm the same place. In Passerines, sng, display, and active fighting are regarded as mechanisms f territrial defence (Hward 1948), and all f these have been recgnised in the Pheasant Phasianus clchicus (Cramp 1980). Althugh display twards intruders and fighting were nt bserved during my study, the wailing call can be regarded as a challenge call used t advertise the presence f a territrial male (Beebe 1922). Playback f calls (sectin 6.2.4) resulted in male birds appraching the surce f playback, and the respnse f ne male certainly indicated antagnism twards what he regarded as anther male, prbably within his territry. The existence f territriality culd easily be tested in captivity by bserving the behaviur f male birds. In the wild, i t wuld be interesting t use playback and decys within the pstulated territries, and t bserve their effect n the birds. Breeding calls were nt made by Tragpans in the autumn, and i t is likely that any territriality breaks dwn utside the breeding seasn. Further research in this field is necessary.

167 Kklass Pheasant 7-3'l Breeding Seasn The breeding seasn f the Kklass is likely t be similar in timing t that f the Tragpan. Hume and Marshall (1879) reprted the Kklass breeding frm mid-april t mid-june between l830m and 2750m, while Wilsn (in Hume and Marshall 1879) extends the breeding range by 300m either side. Beebe (1922) states that nesting begins in late April at 2l40m and that i t becmes later at higher altitude until terminating in mid-june. These dates are als reprted by Blanfrd (I898), and by Ali and Ripley (1969). The dawn calling f the male Kklass can be heard fr mst f the year (Gastn 1980a), s i t cannt be used as a guide t the time f breeding as i t can in the Tragpan. N direct evidence f breeding in the study areas was btained in either 1979 r 1980, but since birds were present in the areas during the breeding seasn, i t is likely that breeding did take place Eggs These are "glssy, creamy buff with dts and bltches f dark reddish r chclate brwn", averaging 52mm x 37mm in size, and with a clutch size f 4 t 9 (Delacur 1977* P.m. macrlpha) Nest A scrape in the grund rughly lined with sticks, leaves, and grass, cncealed under dense bushes r rcks (Ali and Ripley 1969> P.m. macrlpha).

168 153 7.^.4 Territriality Territriality in the Kklass has nt been reprted in the literature. Like the Tragpan, certain individuals culd be recgnised by their calling patterns every mrning at Pipar. Hwever, birds tended t change their rst sites every few days, and n tw males were every heard calling frm the same rst. The call f the Kklass can be heard mre r less cntinuusly frm September t June (see Table 7), and i t has nt been determined whether there is a nticeable increase in crwing intensity during the breeding seasn. Hwever, the crw call f the male can prbably be regarded as a territrial display call, like that f the Satyr Tragpan. Playback f crw calls clse t a calling male resulted in the bird appraching the surce f playback (sectin 6.3.4) which indicates that he was alerted t the presence f anther male which was prbably within his "territry". This respnse was evked equally well in spring and autumn, suggesting that i f the bird is territrial, the territries may be maintained after the breeding seasn. Like the Tragpan, further research int this aspect f the bird's eclgy shuld be carried ut, bth in captivity and in the wild. 7.4 Himalayan Mnal Breeding Seasn Living at the highest altitudes f the fur pheasant species studied, the Mnal is likely t have a later breeding seasn than the thers. Hume and Marshall (1879) and Blanfrd (1898) recrded the breeding seasn as May t June, while Ali and Ripley (1969) als include April. Bates and Lwther (1952) fund nests n 7 May (n height recrded), 24 May at 3050m, and 26 June at nly 2440m. The latter date seems very late fr such a lw elevatin, and i t may have been a secnd clutch. Hellmich

169 154 (1968) sht an egg-laden female n 7 May at 3500m. In the study areas, frntal display by male birds was bserved n 17 April 1979 and 21 April 1980 at abut 3270m, and flight display was bserved during bth spring field seasns. N nests were fund in either year, but i t is likely that breeding tk place within the study areas in bth years Eggs The eggs are "pale yellwish r reddish buff, freckled and sptted with reddish brwn". Size 64mm x 45mm. Clutch size 4 t 6. (Ali and Ripley 1969) Nest A scrape in the grund under the shelter f a rck r in undergrwth n a steep hillside, hidden by grass r ferns etc. I t may be partly lined with leaves, grass, r mss (Ali and Ripley I969, Bates and Lwther 1952) Territriality The presence r absence f territriality in the Mnal is very difficult t determine. Its dawn call is nt regarded t be a challenge call (as i t is in the Tragpan and Kklass), and althugh i t feeds n pen slpes fr parts f the day, interactins between males were seen nly n tw ccasins. On 6 and 7 May, tw male Mnals were bserved fighting in the early mrnings, with ne ther male lking n (sectin 6.4.6). The reasns fr this were unclear, but i t is pssible that i t was territrial aggressive behaviur. Delacur (1977) has described pugnacity between captive male Mnals (see sectin 6.4.6), which culd als be cnstrued as being territrial in functin. In the wild, i t appears that i f territriality des exist in the Mnal, the male

170 155 prbably establishes a territry sufficient fr all breeding activities. Advertisement culd be bth by dawn calling and aerial display t attract a mate. After the breeding seasn, any territriality may die dwn, allwing birds t cngregate during the autumn and winter, as has been bserved (Hume and Marshall 1879, Beebe 1922, Hellmich 1968, Fleming et. al. 1976).» Further research, using marked birds is needed t determine the ccurrence f territriality in this species Cheer Pheasant Breeding Seasn 1 Accrding t Hume and Marshall (1879), Blanfrd (1898), and All and Ripley (1969), the Cheer breeds between 1200m and 2450m frm April until June. In the Athhazar Parbat regin, I bserved the species briefly during May 1980, but the sightings were t few t determine whether r nt breeding was in prgress Eggs The eggs are dull creamy white t pale grey-buff, sparsely freckled and bltched with light reddish brwn. Clutch size 9 t 14. (All and Ripley 1969)- Egg size 53mm x 39mm. (Baker) Nest A scrape r depressin rughly lined with a few leaves and grass at the ft f a bulder In pen frest, usually well cncealed (All and Ripley I969) Territriality Nthing has been reprted in the literature regarding territriality

171 156 in this species. My wn bservatins were insufficient t prvide any» evidence, and further research n the Cheer is urgently required.

172 8. HUMAN INFLUENCE ON THE PHEASANTS AND REMMENDATIONS FOR THEIR NSERVATION Pralad B. Ynzn, M.Sc, carried ut the principal study f human impact n pheasant habitats fr this prject in May-June 1979* and his findings have been published (Ynzn and Lellitt 1980). He devised a questinnaire, (see Appendix 13), and interviewed a ttal f 38 villagers in the Seti and Mardi valleys n varius aspects f frest prductin and hunting. His results are used in this chapter, alng with further data cllected myself during the perid September t Nvember 1979, and April t May Factrs f Human Influence Gthals Gthals are herdsmen wh travel arund with grazing livestck, bth in the valleys and in high level pastures. In this study, we were cncerned nly with livestck which were being grazed abve 2000m, thus excluding all grazing in and arund villages. At Pipar, the herdsmen perated in small units f ne t five men and cme frm a village r grup f villages in the Seti valley. In 1979, nly ne grup f herdsmen was encuntered at Pipar, and they were questined cncerning the activities f ther grups in the area. Each grup lives in seasnal 'gths' (temprary huts) which are either re-used every year, r else new nes are built. Unlike the stne gths fund in the Langtang regin (DUHE 1977), the nes at Pipar mstly cnsist f simple wden frames made f lng straight branches bund tgether t frm an pen lattice. Principal species used in the frame cnstructin are Rhddendrn sp., Betula utilis, and Arundinaria sp.

173 158 The frames are then enclsed by the additin f wven bamb mats, carried up frm the villages, t frm the rf and walls. Highaltitude gths (ver 4000m) are usually built f rcks, with the additin f wden rf frames and bamb mats. The general distributin f gths in the study area is shwn in Figure 19, and tended t be influenced by the availability f mderately flat grund, which is fund mainly n the ridgelines. The principal pasture areas ranged frm lw-level nes such as Sankhbang (1700m), thrugh intermediate areas such as Pipar (3300m), t high level grazing areas (knwn as 'kharkas') such as thse at the nrthern edge f the study areas at mre than 4000m. Gthals can be divided int tw distinct categries: thse with cattle and water buffal, and thse with sheep and gats; all keep up t three Tibetan Mastiff dgs fr the prtectin f the herds. The frmer grups begin grazing at lw-level pastures in mid May, and mve t mid-altitude nes in late June, as the mnsn appraches. A grup was encuntered at Sankhbang with abut 25 head f cattle n 5 May, 1979 and 3 May, 1980, prir t mving up t intermediate areas such as Pipar. Evidence f the presence f cattle gthals was cmmn at Pipar, Bhalu, and Krchn in the frm f recently-used gths, tracks and faeces f cattle, and grazed slpes and bushes. N such evidence was seen in the areas abve 3500m, and i t is prbable that this height represents the apprximate limit f cattle and buffal grazing in this area. Mediumaltitude pastures such as Pipar are prbably used nly between late June and early t mid September, since n herds were encuntered at this height either in early June, r in late September. Sheep and gat gthals mve away frm the valleys in late May r early June (we encuntered a party at Thulkhbang n 28 May, 1979)» and they then spend mst f the summer at higher altitudes than the cattle

174 159 > OPT NjJJMEROOS- GOTHS I 4 i 1 kalki 334* r 05 i t i v V 4? 07- paid XBase, V \ t r > 0 1 f a ST V i. un \ i Oik OH 1 (Thi OH f ,2 LJ AM Km Scale - Apprximate psitin f gth site Figure 19: Apprximate distributin f knwn gth sites in study areas.

175 160 herders. In September 1979, a grup f three gthals was present at Kalki Danda with abut 48 head f stck. They stated that six gthgrups were present at and beynd the nrthern edge f the area shwn n Figure 19 during the summer. Three grups were frm Ghachk village, tw were frm a village near Lachk, and they themselves were frm Mirsa. Initially seen n Kalki Danda ridge n 22 September, 1979, they mved dwn t a gth area between census pints 3 and 4 (see Figure 5) early in Octber, passed thrugh Pipar campsite n 8 Octber, and spent the rest f the mnth and early Nvember n the suthern and eastern slpes f Pipar. They expected t return t their village in late Nvember, and had dne s by the 27th. Frm questining lcal villagers in Karuwa, i t appears that mst villages emply up t five men frm amng their number t lk after the livestck during the summer. These gthals, wh were implied t be the less intelligent members f the cmmunity, are paid a small amunt f mney and given fd by the wners f livestck in return fr lking after them. The reasn fr the upward migratin f the livestck is principally t avid crp damage in the fields, and presumably because fd is plentiful during the mnsn, even at high altitudes. In many parts f the wrld, the effect f wandering livestck herds n frest and grassland habitats is very serius. In this study, the effects f grazing were assessed purely in a visual manner, and n measurements were taken. It was nted that the frest adjacent t village, extending up t apprximately 1500m had been intensely grazed in certain areas, and shwed a cnsiderable lack f understry grwth. Similarly, in areas immediately adjacent t regularly-used pastures such as Sankhbang, Thulkhbang, and Hile, the lack f grund vegetatin and understry was evident, indicating heavy grazing in these areas t. The effects in mid-range pastures are mre difficult t

176 assess. At Pipar and Kumai, sme f the vegetatin is characterised by pen slpes f tussck grassland, sme f which are the result f frest clearance, while thers may be f natural ccurrence. Sme areas shw a predminance f Berberis scrub, especially the edges f the frest. This shrub is an unpalatable species knwn t flurish in areas f clearance and vergrazing (DUHE 1977), and its presence at Pipar indicates mderately heavy grazing, prbably by cattle. Rhddendrn frest at Kumai was nted t be very denuded f grund vegetatin in the area clse t the gths, but this became prprtinately less prnunced with increased distance frm the gths. In all the study areas, the 'kharka' pastures abve the treeline were nly visited in autumn 1979, when they were clear f snw. Grazing pressure was nt assessed quantitatively, but ersin was negligable, and grass was s t i l l abundant, shwing n signs f vergrazing. The ther main ways in which gthals affect habitats is by disturbance, trapping, and hunting. All f these are lcalised t the area in which a gth-grup perates, and the verall effect n the habitat depends n the density f such grups. It is prbable that disturbance causes birds such as pheasants t mve away frm the lcal area, thereby reducing the likelihd f predatin by trapping r hunting. N new traps were seen in the Pipar r Kumai areas in autumn 1979 after the gthals had left. The principal way in which pheasants are affected by livestck grazing is the reductin f frest grund cver. This is ne f the mst imprtant features f pheasant habitat, since it prvides prtectin and safety fr the birds, which are mstly very shy and wary, and its lss will certainly be detrimental t them. The cmplete clearing f frest t make way fr pasture land is even mre deleterius t the pheasants, except perhaps the Mnal. Hwever, abve 3000m the

177 present levels f clearance and grazing d nt appear t be severelyaffecting the pheasants in the study areas, althugh annual mnitring f livestck grazing and clearance shuld be carried ut. The large areas f almst untuched frest, supprting gd numbers f pheasants, we their cntinued presence t the fact that they are situated n very steep and inaccessible slpes where grazing f cattle is impssible. Hwever, these slpes can be grazed by sheep and gats, and i f their numbers are allwed t increase, such areas may becme threatened Shikaris In this study, the term 'shikari' refers t a man wh, fr part f the year, hunts wild game fr fd and prfit. Each village visited in the Seti and Mardi valleys cntained ne r mre resident prfessinal r amateur shikari. Prm analysis f the results f Ynzn's questinnaire (the raw data f which is depsited in Kathmandu), at least eighteen prfessinal shikaris frm ten nearby villages were active, and they ranged in age frm 25 t 50 years,. Shikar (hunting) parties usually cnsisted f tw t nine men, depending n the ttal number f hunting days, which ranged frm three t twelve. Shikaris hunt mainly by trapping r shting their prey. Pew prfessinal shikaris wn a gun, and mst (80$) are brrwed frm rich villagers wh take a share in the prize. Bth ld-fashined muzzlelading and mre mdern breech-lading guns are used. When ut shting, shikaris prefer t k i l l large game such as Serw, Ghral (Nemrhaedus ghral), r Barking Deer (Muntiacus muntjak), but i f success with these is limited, as is ften the case, then pheasants will be sht. In spring 1979* a 'shikar party' f six men (carrying three guns), spent a ttal f 108 man-hurs hunting withut bserving

178 a single wild mammal. Similarly, frm 10 t 12 Octber, 1979* a shikar party f seven men with fur guns spent three days hunting with n success. Prm the results f Ynzn's questinnaire cncerning pheasants sht, the Kalij Pheasant (ccurring frm 304m t 2450m altitude) accunted fr 70$ f the ttal k i l l, fllwed by the Satyr Tragpan with 27%. Bth f these species range well belw the study area, and i t is thught that shting f pheasants takes place mainly belw 2600m. The technique f flashlighting is used by shikaris t k i l l Kalij Pheasants near t villages. I t invlves finding a Kalij rst, visiting i t by night, and dazzling the rsting birds with a pwerful beam f light. The birds can then be sht singly, as they remain sitting n the branches (Rberts 1980, Severinghaus 1977). Details f trapping are less easy t cme by, but the principal methd is by the ft snare. This invlves a trigger mechanism releasing a 'spring' which tightens a nse arund the bird's leg. The mechanism is illustrated in Figure 20 and described in detail in Appendix 14. A brush fence f small branches abut lm high and up t 50m lng is set ut in a likely spt f frest r grassland. A number f 'gates' are made in the fence, each set with a snare, and they are checked peridically (preferably twice per day) fr trapped birds. Ynzn's questinnaire shwed that in 1978 the trapping ttals f pheasant species were as fllws: Kalij 4-3$, Satyr Tragpan 36$, Himalayan Mnal 21$. The success rate f trapping is nt knwn, but is nt likely t be very high, as Severinghaus (1977) fund in Taiwan. In Octber 1979* I set tw 7-gate trap lines in Rhddendrn frest and Berberis scrub, which in fur days resulted in ne male Tragpan caught alive. In April 1980, a prter at Pipar campsite set a 5-gate trapline fr mre than seven days, but nly succeeded in catching ne male Bld Pheasant.

179 164 TRAP SPRING (RESILIENT SAPLINGr) 1 I CATCH BRUSH WALL PIECE FRAME TRIGGER NOOSE v Figure 20: Diagrammatic sketch f pheasant snare.

180 165 Trapping did nt appear t be carried ut very cmmnly at Pipar, since nly ld (pre-1979) trap fences were seen in bth years, and they were cmpletely absent frm Kumai, Krchn, and Bhalu. I t is thught that trapping takes place mainly at lwer levels, near villages, and als in areas where shikaris cmmnly g ut n hunting expeditins. At present levels i t is nt detrimental t pheasant numbers, but shikaris with sme knwledge f the birds' mvements, and by intensive setting f traplines, culd prbably catch large numbers f birds in lcalised areas Land Burning Lcal villagers carry ut burning f scrubland and grassland in certain areas during February and March each year. I t is dne in rder t prmte new grwth in the tussck grass which is favured by the livestck herds grazed n these areas in the summer mnths. Such burning als helps t prevent the frmatin f scrub and wdland n these areas. Burning was particularly nticed n the suthern slpes f Pipar and Kalki Danda, and n the hillsides abve Dara and Muri villages in the Athhazar Parbat regin. At Pipar, the burnt slpes were dry and blackened with numerus shrt tusscks f stiff grass in early April 1979 and Grwth f new shts began in late April, and by late May all the slpes were becming cvered with new grwth; in September 1979 they were thickly cvered with very varied herbaceus vegetatin and tussck grass. In the Athhazar Parbat regin, burnt slpes had nt begun significant new grwth in mid May I98O, but were well cvered in mid August (1979). Apart frm the effect f preventing the regeneratin f scrub and frest, the practice f burning grassland has tw main effects n the

181 166 vegetatin. First, burning expses bare grund which is subjected t the heavy spring precipitatin which ccurs in the study areas. This has led t a small amunt f ersin at Pipar, which is likely t increase in the future i f the practice f burning is cntinued. Secndly, fires frequently appear t get ut f hand, with the result that adjining edges f frest, and scrub becme burnt. This was evident in a number f places at Pipar, and even mre s West f Dara village in the Athhazar Parbat regin where large numbers f Rhddendrn trees were badly burnt. Cntinued accidental burning f frest adjacent t grassland will slwly increase the latter and reduce the frest cver Ringal Bamb Cutting and Medicinal Plant Cllecting Ringal bamb (Arundinaria sp.) is an imprtant raw material used in the making f numerus husehld items such as carrying-baskets and matting, sme f which are sld in bazaars n a regular basis. Cutting f ringal bamb is carried ut thrughut the year in mst areas, althugh in the Siding area, Mardi valley, the lcal Panchayat Authrity (gvernment) has been practising a seasnal ban n cutting frm mid-may until September. On 6 April, 1980, a ttal f abut 30 bamb cutters was seen passing thrugh Dhiprang village, travelling frm the upper Seti valley t Ghachk village. Each man was carrying up t fifteen bamb stems apprximately 6m in length which makes a ttal day's cut f 450 lengths. The cutting is nt indiscriminate: nly the suitable sized mature stems are chsen, and yung shts are mstly left uncut. The prblem with bamb grwth is that i t is extremely slw, sme himalayan species requiring 40 r mre years t grw, befre flwering nce and dying. Regeneratin is therefre very slw, and i f large areas f bamb are cut befre they have flwered and drpped their seeds, new grwth is nly pssible by vegetative means. At present, cutting is carried ut up t nly apprximately

182 m, but since regeneratin is s slw, lwer stcks might sn becme exhausted, necessitating cutting at higher altitudes. Tragpan and Kklass Pheasant are the species principally affected by present levels f cutting, and the remval f cver is likely t be detrimental t these species' habitats. Cntinuus mnitring f the amunt f bamb crpped in selected areas wuld prvide a useful index fr future management f stcks. Medicinal plant cllecting takes place during the pst-mnsn seasn. In September and Octber 1979* grups f up t five men visited the Pipar area in rder t cllect rts and leaves f certain species t sell in the villages fr medicinal purpses. In sme cases, thugh I did nt bserve i t, pheasant snares are set n the way t the cllectin sites, but the number f pheasants trapped in this manner is likely t be small Fuelwd Cllectin An excellent reprt n the fuelwd requirements f villagers in the Langtang Natinal Park and elsewhere in Nepal, can be fund in DUHE (1977 pp ). Details f fuel cnsumptin by villagers in the Seti and Mardi valleys were nt nted in this study, but the general principles given in the DUHE reprt apply thrughut much f Upland Nepal. The authrs estimated in the Langtang regin, that frm three t ver twenty tnnes (air dried) f firewd is used annually by each husehld in its permanent settlement. They state that cnsumptin per husehld depends n: (i) The distance between the surce f wd and the settlement, (ii) The seasnality in the availability f the fuel surce, ( i i i ) The altitude f the husehld, (iv) The number f peple in the husehld. (v) The wealth f the family.

183 168 Frm bservatins made in private husehlds in the upper Seti valley, I estimate that an 'average' family burns apprximately ne 'lad' f wd (apprximately 40kg. ) per day, r 14.6 tnnes per year. In the upper valley, wd is bth plentiful and n great distance away, s that cllectin is easy. This is nt the case as ne prgresses dwn the valley, and i t is likely that wd cnsumptin is cut dwn as factr (i) increases and factr ( i i i ) decreases. Hwever, accurate estimates f the fuel cnsumptin f the whle valley can be determined nly by studies f this prblem. A general trend in the whle area is fr firewd t be cut clsest t villages, s that the slpes f the lwer valleys (twards Pkhara) have largely been cleared f frest and are terraced fr cultivatin. The upper slpes are less denuded, but since fuelwd is nw scarce n the lwer slpes, the frmer are cming under increasing pressure, s that there is a cnstant 'attack frm belw' alng the lwer frest edges. The pattern is cmplicated by wd cutting by gthals and shikaris. The frmer are crpping fuelwd frm the areas surrunding their gths in the lw and mid-altitude ranges, while the gthals wh use kharkas fr grazing their herds are cutting wd at the treeline. The affect f this crpping depends n: (a) The density f gthals in an area, (b) Hw regularly they visit the same areas, and (c) Whether their herds are grazing abve r belw the treeline. Questining f the gthals in the Pipar area in autumn 1979 revealed that they used abut 1 t 2 'lads' f wd per day where i t was easily available, depending n the altitude. This was because a fire was ften kept burning during the night t ward ff wild animals, and t keep warm. Gthals living abve the treeline wuld manage with less than this, despite the cld, because f the greater distance i t had t be carried.

184 169 Defrestatin leads t lss f habitats and therefre f pheasant species, but i t has much wider and mre serius implicatins in terms f ersin, flding, and famine, s that reductin f defrestatin is an abslute necessity. In view f the ever-increasing pressure n the frests in this area and thrughut Nepal, i t is essential that alternative surces f energy be investigated fr use in place f firewd and that re-affrestatin prgrammes be implemented. Status f the Study Species in Captivity Appendix 15 shws the full censuses fr the study species carried ut by the Wrld Pheasant Assciatin in 1976 and The ttals fr each species are listed in Table 23. TABLE 23 Gens uses f the Study Species f Pheasants in Captivity, 1976 and 1979 Species Captive Ppulatin in Percentage increase/decrease Bld Pheasant $ Satyr Tragpan $ Kklass Pheasant $ Himalayan Mnal $ Cheer Pheasant $ The censuses are stated by the c-rdinatr t be underestimates, and therefre shuld nly be used as a rugh guide. Assuming they are accurate t within apprximately 20$, they indicate that three f the species have rughly stable ppulatins in captivity, while the Bld Pheasant has increased by abut 50$, and the Kklass by abut 240$ in three years. I t is encuraging t nte that althugh these tw species are the mst difficult t keep in captivity (Delacur 1977), their numbers are increasing, suggesting that rearing techniques are imprving, Hwever, cmpared with the ther species, the

185 Bld Pheasant numbers are s t i l l very lw. The nly endangered species f Nepal, the Cheer Pheasant shws an apparent slight decline in captive ppulatin, but its numbers are s t i l l relatively high. 8.3 Status f the Study Species in the Wild 8.3.I Bld Pheasant Wrld status f this species is difficult t assess at present due t the fact that nly tw f the eleven subspecies, I.e.cruentus and I.c.tibetanus, ccur in areas frm which infrmatin n their status is readily available. The remaining subspecies ccur in Tibet and China, fr which Cheng (1963) is the nly available reference. He states that i t was "quite numerus" in the areas he visited in 1956, such as the Ch'in Ling and T'ai-pai Shan muntains in Shensi and Szechwan prvinces. Hdgsn (in Hume and Marshall 1879) reprted that prir t 1850, the species was cmmn in Nepal, and that packs f 70 t 100 birds were ften seen. N reprts f the species' status in Nepal were available fr the next century because the cuntry was clsed t freigners until Since then, Ali and Ripley (1969) have described i t as "cmmn and abundant in the high Himalayas in central and eastern Nepal"; while Fleming et. al. (1976) state that i t is "fairly cmmn" (i.e. "in suitable habitat, the chances f seeing them are abut 50$"). In cntrast, Grahame (1971)> frm his limited bservatins in the Everest regin and India believed the species t be "nwhere cmmn". 2 In this study, I fund apprximate densities f 2.5 t 4 pairs per km in suitable habitat. This is relatively sparse, but cmpares favurably with densities f birds such as wdland gruse in Eurpe (Lvell 1979)> which are managed fr hunting, and Kklass Pheasant in Pakistan (see Table 10 and Severinghaus 1979). Hwever, further surveys shuld be cnducted in Nepal and elsewhere befre attempting extraplatin t estimate the wrld ppulatin. In general, i t wuld appear that while

186 171 i t is unlikely t be as cmmn as i t was in Hdgsn's time, the Bld Pheasant is certainly in n danger at present. Future Prspects: This species, having a fairly restricted altitudinal range, at high levels (apprximately JOOOm t 3600m, see Figure 8) will mst prbably receive heavy pressures in the future frm expanding human settlements and activities. In the study areas, pressures will initially be the increased frest clearance and livestck grazing rather than the establishment f permanent settlements. Hwever, in eastern Nepal, and prbably Sikkim, Bhutan, and China, these prblems will be accmpanied by an increase in size and number f villages within the high-level range f the Bld Pheasant. The habitat destructin resulting frm frest clearance culd easily give rise t the frmatin f a number f 'islands' f suitable habitat, each supprting separate ppulatins f Bld Pheasants, and surrunded by altered habitat. Diamnd (1975) has studied such 'islands' and frms a number f cnclusins frm his wrk. These include the fllwing: (i) Larger reserves, and reserves situated clse t ther reserves, can hld mre species than small islated nes, (ii) Different species require different minimum areas t ensure viable ppulatins, ( i i i ) The smaller the reserve, the higher the extinctin rate. Althugh n surveys have been carried ut n Bld Pheasants in ther areas in Nepal r elsewhere, i t is assumed (and hped) that the 'island' phenmenn is s t i l l a lng way ff. T ensure the species' survival, i t nevertheless needs a number f reserves in which its habitat will be prtected frm destructin. I t is knwn t ccur in mderate numbers in the Everest Natinal Park (R.L. Fleming Jr. pers. cmm. 1979) and in

187 172 the Langtang Natinal Park (DIME 1977) Its survival in these tw areas shuld be assured i f paching and habitat deteriratin can be stpped, but the species wuld benefit frm further prtectin, especially in an area cntaining n majr human settlements such as Pipar. Recmmendatins fr the creatin f a reserve in this area are presented in sectin ;.2 Satyr Tragpan The wrld status f this species is als difficult t assess, thugh fr different reasns t the Bld Pheasant. Early authrs such as Hume and Marshall (1879) r Blanfrd (1898) give n indicatin f its abundance, but i t was likely t have been mre cmmn than i t is at present since its habitat was presumably mre widespread thrughut its range. Even recent authrs d nt state whether i t is cmmn r nt, and the reference f Fleming et. al. (1976) t "ccasinal" ("usually nt seen") alludes t the bird's secretiveness rather than its abundance. N surveys have been carried ut n this species, s I can base its status nly n the densities recrded in the study 2 areas. These were apprximately 3 t 4.5 pairs per km f suitable habitat which were rughly similar t the densities f Bld Pheasant, and cnsiderably greater than the densities f the endangered Western Tragpan in Pakistan (less than 1 pair per km - Mirza 1978). After further surveys have been carried ut in ther areas, i t might be pssible t extraplate densities t estimate the size f the wrld ppulatin f the Satyr Tragpan. Future Prspects: The Tragpan has a fairly wide altitudinal range (rughly 2230m t 3550m, see Figure 8), which shws that i t can explit a wider variety f habitat than can species such as the Bld Pheasant. Nevertheless, Its lwer level habitats are likely t cme under very great pressure frm defrestatin in the next few decades.

188 In the Seti and Mardi valleys, adjacent t the study areas, tree cutting fr firewd is being carried ut at a high rate (sectin 8.1.5) and as wded areas near the villages becme exhausted, explitatin f frest within the range f the Tragpan is inevitable. In fact i t has prbably already happened, and the present range f the species in these regins is likely t be smaller than i t was in the past. Unless alternative fuel becmes freely available t the lcal peple, the present high rates f fuelwd crpping are certain t cntinue, and prbably increase. I f this happens, pressure n Tragpan habitat will als increase, and the numbers are likely t decline, althugh birds living at higher altitudes will nt be affected as sn as thse at lwer levels. The 'island' phenmenn (Diamnd 1975) is likely t develp in the same way as i t will fr the Bld Pheasant (sectin 8.3-1) i.e. areas f suitable habitat may becme fragmented and separated by larger areas f altered habitat. Althugh present areas f suitable habitat appear t be in reasnable quantity in the regins I have visited in Nepal, almst nne f these are actually prtected. The Tragpan is unlikely t ccur within the Everest Natinal Park, althugh I fund the remains f ne male utside, which had prbably been killed by peple less than 15km frm the Park bundary. The Durham University Himalayan Expeditin did nt encunter the species within the Langtang Natinal Park, althugh J. L. Fx did s in 1974 (Fx 1974). I have n infrmatin fr the third Natinal Park, Lake Rara. All three f these Natinal Parks cver areas which lie predminently abve the altitude range f the Tragpan, and can therefre nt be regarded as prtecting the species. I t wuld be beneficial t preserve an area such as Pipar in rder t prtect the habitat f the species in the centre f its range. Recmmendatins fr the setting up f a reserve at Pipar can be fund in sectin

189 Kklass Pheasant Five f the nine subspecies f P.maerlpha ccur nly in China. Regarding their status, Cheng (1963) states that "numbers are nt very numerus, and attentin shuld be given t regulated hunting and planned breeding". Fr the ther fur subspecies, which all ccur in the Himalayas, i t is als difficult t btain infrmatin f their status, except fr a few recent surveys. Wilsn (in Hume and Marshall 1879) reprted f P.m.macrlpha that they were fund n nearly all hillsides that were cvered with trees and bushes, which implies a fairly widespread distributin. White (1925) and Davidsn (1896, in Bates and Lwther 1952) reprted frm Kashmir that P.m. biddulphi culd be fund "in sme numbers", and that they were "cmmn". Mre recently, wrkers such as Bates and Lwther themselves (p. cit.) and Ali and Ripley (1969) refer t the presence f the Kklass in Kashmir, India, and Nepal, but give n indicatin f its abundance. Rberts (1970) states that in Pakistan "they are nw rare and encuntered nly with difficulty and after diligent search in areas where even in the last ten years they were plentiful". Table 10 shws a summary f the results f censuses carried ut in Pakistan, India, and Nepal in the last fur years, which shw ranges in 2 ppulatin denisty frm 1.5 t 10.5 pairs per km (excluding Severinghaus' estimate, sectin 4.3.4). The mean density f Kklass in 2 Pakistan is 4.4 pairs per km (n=10), s the results frm the Pipar area f an average f 3.2 pairs per km^ (n=5) cmpare quite favurably. Gastn's 2 figures frm the Dachigam Sanctuary, Kashmir, f 10 pairs per km are based n nly ne survey, but are als encuraging as are thse f Green (1980). Gastn als states (1980b) that Dachigam, Khajiar, and Simla water catchment areas all prbably supprt several hundred pairs f Kklass.

190 175 Future Prspects: In the study areas, the Kklass ccurs within a fairly limited altitudinal range (2700m t 3300m, see Figure 8), within that f the Tragpan, but the species as a whle is reprted t ccur between 600rn and 4880m in China (sectin 1.3) I t is therefre pssible that it may be able t adapt t a wide range f habitats and altitudes, but at present in Nepal, P.m. nipalensis wuld appear t be under similar pressures t the Tragpan. These are principally defrestatin and grazing, and the discussin in sectin applies equally well fr this species. Hwever, unlike the Tragpan, the Kklass is limited t western Nepal, where l i t t l e is knwn f habitat suitability because f its inaccessibility. I t is prbable that suitable habitat is s t i l l present in reasnable quantity in West Nepal, but surveys shuld be carried ut t cnfirm this. Lake Rara Natinal Park is knwn t cntain Kklass Pheasant (Natinal Parks and Wildlife Cnservatin Office staff, Kathmandu, pers. cmm. 1979)> but this is the nly area in which prtectin is reasnably assured. reasnable numbers at Pipar, which Since the Kklass is present in is near the eastern edge f its range, i t wuld be advantageus t prtect this area Himalayan Mnal Early reprts f the abundance f this species indicate that i t was very cmmn last century, but that hunting was reducing its numbers drastically in India. Hume and Marshall (1879) referred t Wilsn sending hme 1000 t 1500 skins per year fr thirty years! Rberts (1980) expresses sme scepticism at these numbers, but i t seems reasnable t assume t h a t the b i r d was v e r y cmmn, at l e a s t up u n t i l the midnineteenth century. Since then, there is l i t t l e infrmatin cncerning numbers f birds. Bates and Lwther (1952) reprt that i t is "nt at all numerus" in Kashmir except in a few lcalities, while Rberts (1970)

191 176 glmily reprts frm Pakistan that " i t has nw becme very rare in all the frests where even ten years ag i t was plentiful". In Nepal i t was never likely t have been hunted with as much intensity as i t was in India and Pakistan (Rberts 1980). Hellmich (1968) cnsiders i t t be widespread in Nepal but nt very abundant, while Fleming et. al. (1976) recrd i t as "cmmn" ("usually seen") in suitable habitat. The census results frm my study are verestimates (see sectin 4.3-5)> but even taking this int accunt, they cmpare very favurably with Mirza's cunts frm Pakistan (see Table 11), which are prbably underestimates. Future Prspects: The Mnal is fund ver mst f the altitudinal range f the Kklass and Bld Pheasants, but als ccurs higher than either f these species, having a range f 2800 t 4000m in the study areas (see Figure 8). This wide range is advantageus t the species in that i f ne part f its habitat is explited, i t may be able t adapt t living nly within the remaining part. Hwever, the Mnal is likely t face similar pressures t bth the Kklass and the Bld Pheasant in the frm f expanding human settlements within its habitat. Its ability t live in areas abve the treeline during the summer may be imprtant in the future i f defrestatin becmes a majr prblem in the Himalayas. Its high altitude range may als help t prevent its habitat becming fragmented int 'islands' f suitable habitat, referred t earlier, but this pssibility cannt be ruled ut. The Mnal enjys the status f being Nepal's mst prtected pheasant species. In additin t being the natinal bird f Nepal, i t ccurs cmmnly in the Everest and Langtang Natinal Parks, and pssibly als ccurs in the Shey Gmpa Wildlife Reserve (Dlp Regin) and Lake Rara Natinal Park. The bird and its habitat are given reasnable prtectin in all these areas, s that its future is fairly secure. Hwever, its presence in a pssible reserve in the Pipar area

192 wuld mean that five f Nepal's six pheasant species wuld be prtected in ne area Cheer Pheasant Prm the accunts in the literature, this endangered species has always had a very lcal distributin thrughut its range, althugh i t was smetimes quite cmmn where present (Hume and Marshall 1879* Blanfrd 1898). Thrughut much f its present range in India and Pakistan, the Cheer is even mre lcally distributed than i t was, and where i t des exist, its numbers are prbably significantly less than they were last century. Rberts (1970) reprts that i t may face cmplete extenctin in Pakistan, and knws f nly tw lcatins, in Hazara and the lwer Neelum Valley (Azad Kashmir) where i t exists. Severinghaus et. al. (1979) and Mirza (1980a) cnsider that i t is nw extinct except frm sme areas f Azad Kashmir. Hwever, since 1977, the Wrld Pheasant Assciatin Pakistan Chapter has been carrying ut experimental re-intrductin f Cheer int the Margala Range Natinal Park, with sme success. N data was frthcming frm India until a survey f the Chail Wildlife Sanctuary, Himachal Pradesh was made by Gastn in April 1979 (Gastn and Singh 1980). He and a number f bservers saw twelve Cheer in the curse f five days, and at least nine pairs were lcated by call cunt censusing. This gave an extraplated ttal f apprximately frty pairs in the reserve, the largest ppulatin currently knwn anywhere. The Red Data Bk (IUCN 1979) states that the species is "endangered in mst parts f its range, and pssible extinct in Pakistan. Becming increasingly rare and lcal in the West Himalayan fthills, where habitat destructin and hunting have caused a serius deteriratin in its status". Very l i t t l e is knwn abut numbers f Cheer Pheasants in Nepal, but they are knwn t ccur in reasnable numbers at Dhrpatan (Fleming et. al

193 1976), and Rberts "fund them nt cmmn, but nt s scarce as t warrant listing as an endangered species" at Dhrpatan in 1976 (Rberts 1980). In this study, I fund fur pairs present near Muri village in an area f apprximately 0.5 km (see sectin 4.4), but n Cheer were lcated in the surrunding area. Bth Dhrpatan and Muri are very clse t the eastern limit f the gegraphical range f this species. Future Prspects: Having carried ut s l i t t l e fieldwrk n the Cheer Pheasant, I find i t difficult t assess its prspects fr the future. I t lives at a lwer altitude (2200 t 2440m) than the ther species studied, and was bserved living in clse prximity t much human activity such as livestck grazing, fuelwd cutting, and arable farming. Hwever, precisely because i t was seen existing with such heavy human influences, I cnsider that i t may have been able t adapt t these, and inhabit areas clse t human settlements. If s, then as lng as man's activities d nt becme t destructive, and that defrestatin can be brught under cntrl, the Cheer may be able t survive. The effect f the 'island' phenmenn (see sectin 8.J>.1) n the Cheer is als difficult t assess. Since i t has always been f lcal ccurrence, i t wuld seem that the 'island' phenmenn has existed in its distributin since the last century, but that its effect has been accelerated as the lss f habitat has increased. The Cheer is nt thught t ccur within any prtected areas in Nepal, althugh the Dhrpatan area may sn be gazetted as a 'hunting blck' t regularise and prtect Blue Sheep (Pseudis nayaur) hunting t the Nrth (J. Rberts pers. cmm. 1980). I cnsider that a very high pririty fr future wrk is t carry ut extensive surveys fr Cheer Pheasant t the East and West f Dhrpatan, t try and prvide sme baseline data n the real status f this species in Nepal, after

194 179 which intensive studies f its eclgy shuld be executed, s that its prtectin can be assured Recmmendatins fr Future Cnservatin One f the fur riginal bjectives f my prject was: "T suggest an area that may be cnsidered as a sanctuary in which Gallifrmes are fully prtected, and further research can be undertaken" (Lellitt 1978). Therefre I l i s t here a number f recmmendatins fr their cnservatin Reserve Establishment I t is suggested that a 'Wildlife Reserve' be set up in the Pipar area. Members f staff f the Natinal Parks and Wildlife Cnservatin Office, His Majesty's Gvernment, Kathmandu shuld survey the area, using the suggestins belw as guidelines fr the establishment f a Reserve as sn as pssible. The ecnmic and lgistical aspects f the creatin f such a Reserve are beynd the limitatins f my study and are nt discussed here. Cnsideratins fr the 'Pipar Wildlife Reserve' 1) The Pipar area supprts mderate densities f fur species f Himalayan pheasants: The Bld Pheasant, Satyr Tragpan, Kklass, and Himalayan Mnal, (see sectin 4.3) and is the nly area knwn where these species ccur within such a narrw altitudinal range, 2) The prtected area needs t be large enugh t maintain stable ppulatins f all fur species, and preferably shuld als include a ppulatin f a f i f t h species, the Kalij Pheasant. Diamnd (1975) states that different species have different prbabilities f persisting n a reserve f a given size. These prbabilities depend n the abundance f the species, the magnitude f its ppulatin

195 l8 fluctuatins, and ther factrs. N wrk has been dne n the ppulatin dynamics f Himalayan Pheasants, s that chsing the minimum size f a self-supprting ppulatin is a matter f judgement. Hwever, the bigger the reserve, the better are the chances f maintaining viable ppulatins. 3) The 'Annapurna Sanctuary' is an area f cnsiderable natural beauty at the headwaters f the Mdi river valley, Nrth-West f Pipar (see Figure 2), and is a ppular turist trek. I t has been prpsed (Sakya 1977) that this shuld be designated as a 'Recreatin Area', which wuld give sme degree f prtectin t the wildlife and frests within its bundaries. Hwever, the Pipar area is unlikely t be included in this 'Recreatin Area', since the Annapurna Sanctuary lies 15 km t the Nrth-West and has well-defined natural bundaries. 4) Fr lgistical purpses, Pipar has the advantage that i t is relatively clse t Pkhara (25km direct, apprximately three days walk), which is the capital f Nepal's Western Develpment Regin. 5) Access t much f the Reserve is very difficult due t steep slpes and unexplred dense frest, which hwever, is beneficial t the existing wildlife. 6) N permanent settlements are included in the prpsed Reserve area except fr Karuwa village (ppulatin apprximately 40 persns), which lies in the extreme Suth-eastern crner. Its presence hwever, des nt pse a threat t habitats and wildlife. Pssible Area fr Pipar Wildlife Reserve (See Figure 21) The area shuld be bunded n the eastern side by the Seti Khla, frm its junctin with the Sadhu Khla (Grid Ref ) t its junctin with the Bhajaundi Khla (546026) at Karuwa village. Frm there, the suthern limit runs nrthwards up the Bhajaundi Khla fr

196 apprximately 1.2km t the junctin with the Chiyali Khla (540038). Prm there i t passes suth-westwards up the Chiyali Khla, taking the suthern f its tw branches at (515034) t the ridge at 2960m (505035). The western limit runs frm this pint nrthwards, fllwing the watershed ver Kurnai (506044) and Krchn (502062), cntinuing t the ridge branch at (494094), where i t dg-legs nrth-eastwards t the peak at 4468m (14,648 feet) (508102). Prm this peak, the nrthern bundary runs suth-eastwards t a peak at 4390m, and frm there eastwards dwn the Sadhu Khla t its junctin with the Seti Khla (574091). This area includes Pipar, Kalki Danda, Krchn and Kumai hills 2 and the Seti river in the East; a ttal area f apprximately 46km. Bundary identificatin in the field is relatively easy since all the bundaries lie alng natural landfrms such as watersheds and streams. The eastern edge f the prpsed reserve lies at a minimum altitude f abut 1300m, which is well within the altitudinal range f the Kalij Pheasant. The species was seen at Sankhbang (1700m) in the frest Nrth f Karuwa village n 7 April, On the basis f the ppulatin densities f the ther pheasant species expressed in sectin 4.3, and the estimated areas f suitable habitat available, Table 24 estimates the numbers f pheasants in this area. TABLE 24 Pssible Suitable Habitat and Pheasant Ppulatin Sizes in prpsed Pipar Wildlife Reserve g e c i Area f 2 Ppulatin Estimate C - suitable habitat (km ) (pairs f birds) Bld Pheasant Satyr Tragpan Kklass Pheasant Himalayan Mnal I7.3 (275 birds) 3 Nte: a - estimate based n density f 16 birds km -, which is prbably an verestimate (sectin 4.3)

197 t 3 5?, Si KALKI DANOA I lama 33*.* 03 X'Camp, i I' u X 08 i 4^ I I 'KORCHQ i-3b8t pnd X Bast J 325', 1 t * I 06 r 1 ) ft 0& 'A wu i v A I ,5 5,5-5v 52 / si Km Figure 21: Map shwing area f prpsed Pipar Wildlife Reserve.

198 Regulatins fr Hunting Under Sectin I I f the Natinal Parks and Wildlife Cnservatin Act 2029, i t is illegal fr any animal (listed in Schedule 2 f the act) t be hunted withut licence. All pheasant species are included in this act, but i t is nt prperly enfrced thrughut Nepal (DUHE 1977* Rberts 1980). The actual extent f illegal hunting is uncertain, but may be cnsiderable in the Seti and Mardi valleys (Ynzn and Lellitt 1980), as elsewhere in Nepal. I t is recmmended that the existing game laws be enfrced by the mre bvius presence f District Frest Officers and ther fficials in the muntain areas. I t is further recmmended that when the Pipar area is established as a Wildlife Reserve, a ttal ban n hunting f all animal and bird species be effected within its bundaries. Hwever, licensed hunters shuld be allwed t cntinue perating utside the reserve, and i t is nt suggested at this stage that hunting be frbidden even during specified seasns Regulatins fr Frest Prductin Puelwd Regulating the amunts f fuelwd cllected and cnsumed presents cnsiderable prblems. Since the lcal peple need wd fr cking and heating, i t is likely that any prpsals t limit wd crpping wuld meet with great ppsitin unless free fuel was prvided in sme ther frm. The cst and lgistics invlved in supplying even a small amunt f fuel (such as kersene) n a regular basis wuld be enrmus. I t is therefre prpsed that a majr educatin scheme be launched in the Seti and Mardi valleys (see sectin 8.4.5) which includes infrming the lcal peple f the need fr frest cnservatin. I t shuld suggest

199 184 that nly minimum fuelwd needs shuld be crpped, and the villagers shuld be tld f the detrimental effects f large-scale frest clearance. At this stage i t is nt prpsed t attempt t regulate the crpping f fuelwd either within the reserve r elsewhere, because f the present lack f alternative fuels, and the difficulty f enfrcing such a regulatin. The Frest Department shuld ensure that cleared areas are given a maximum chance t regenerate. Bamb and Medicinal Plants The crpping f bamb within the prpsed Reserve area shuld be allwed t cntinue, but nly by the inhabitants f Santal, Imu, and Karuwa villages. Bamb frm the Reserve shuld be cut fr the use nly f the peple f these villages, and shuld nt be exprted fr sale elsewhere. Settlements lwer dwn the valley will presumably cntinue t crp bamb elsewhere, but they shuld be encuraged t plant bamb in gulleys and alng stream banks (see Hirsbrunner 1968 fr suggestins). Cllectin f medicinal plants within the prpsed Pipar Reserve area shuld be carried ut nly by lcal villagers, and under the same arrangements as fr bamb cutting. They shuld be crpped nly fr lcal use, and under n circumstances shuld they be sld r exprted elsewhere. A detailed survey f their abundance shuld be carried ut t determine their distributin. Re-affrestatin Re-affrestatin shuld be encuraged in uncultivated cleared areas in the Seti and Mardi valleys in the educatin scheme. Seeds culd be cllected and distributed by the Department f Frests.

200 Regulatins fr Pastralism Althugh nly limited bservatins were made during my study, i t is felt that at present levels, the degree f livestck grazing within the prpsed Pipar Reserve area is nt detrimental t the ttal area f frest. I t is recmmended that villagers frm the Seti and Mardi valleys be allwed t cntinue grazing livestck in their traditinal pasture areas, but that there shuld be n increase in the number f livestck presently using the Pipar area. This necessitates a study f pastralism in the prpsed Reserve area. All gthals shuld be infrmed f the Regulatins fr Hunting, and encuraged t use nly dead wd fr cking and heating. Within the prpsed Pipar Reserve area, the burning f vegetatin shuld be frbidden. I t is suggested that the alleged benefits f pasture burning shuld be investigated (DUHE 1977)> and, i f shwn t be beneficial, they culd be re-intrduced n a cntrlled basis Recmmendatins fr Educatin I f the establishment f a Wildlife Reserve in the Pipar area is t be undertaken, i t is imperative that a substantial educatin prgramme be devised, and implemented in the Seti and Mardi valleys. I f this is nt dne, i t is likely that misunderstandings will arise between the lcal villagers and the authrities wh are setting up the Reserve. This culd lead t the antagnism and ppsitin f the lcal peple, which must be avided at all csts. The educatin prgramme shuld include cverage f the fllwing pints: 1) Awareness f the limited life span f the frests i f present clearance rates are cntinued withut re-affrestatin, and that such clearance will adversely affect future generatins f human inhabitants.

201 186 2) Awareness that increased herd size f livestck can lead t vergrazing, and result in severe deteriratin f pasture areas. j>) Awareness f the need fr cnservatin f rare animals and birds. 4) Understanding that hunting withut a licence is illegal, and that the law is t be enfrced mre strictly in the future, but that licensed hunting is perfectly permissible. 5) Infrmatin cncerning the benefits f limited turist treks t the Reserve area in the future. Sme cnservatin educatin has already been carried ut in the valleys by members f the Wrld Pheasant Assciatin, the staff f Muntain Travel Ltd. (Savage 1978), and participants in my study. I t is prpsed that in future, i t be carried ut by staff f the Natinal Parks and Wildlife Cnservatin Office r Frest Officers wh have been trained in cnservatin principles (see DUKE 1977J P- 114). They shuld travel rund the villages and educate the lcal peple principally by discussing the practical necessities f cnservatin n an infrmal basis. Feedback shuld be btained frm the villagers cncerning their views n cnservatin, and, i f cnsidered imprtant, these shuld be reprted back t the Natinal Parks and Wildlife Cnservatin Office fr cnsideratin. The regulatins and prpsals utlined abve can then be re-assessed Recmmendatins fr Research The fllwing research shuld be carried ut in the prpsed Pipar Wildlife Reserve area, as sn as rssible: (i) Cntinued annual censuses f pheasants and a mre detailed study f the species' eclgy (with special reference t habitat and fd requirements) shuld be undertaken. The status and distributin f 'threatened' and rare mammals

202 187 such as the lepard, serw, and Musk Deer shuld be carried ut. (ii) A full btanical survey f the area, t include prvisin fr permanent transects which can be used t measure quantitative vegetatin changes. Frest cver shuld be mnitred by Phtgraphy frm ppsite slpes, ( i i i ) An assessment f the extent t which fuelwd, timber, and bamb is being crpped frm the area, and the effects f grazing. (iv) Frm studies made elsewhere in Nepal (varius reprts in Natinal Parks and Wildlife Cnservatin Office, DUHE 1977), the effects f turism shuld be assessed. I t shuld be cnsidered whether the intrductin f limited turist facilities n the apprach t Pipar is advisable r nt.

203 188 APPENDIX 1 Descriptins f the five study species f pheasant. 1.1 Bld Pheasant (see Plate 1): The Bld Pheasant is small in cmparisn t mst ther pheasants, and recalls a Partridge in general shape; in fact Beebe (1922) called i t the Bld Partridge. I t exhibits marked sexual dimrphism, the male being grey abve and green belw, streaked with yellw. The upper breast is marked with crimsn (the 'bld stains') as are the wing and t a i l cverts. The frehead and face are black with a crimsn thrat and a red naked rbital patch and cere; b i l l black. The legs are red with up t three spurs. Yung males are similar, but less brilliantly clured. Length c 460mm (Ali and Ripley); wing 19>2l4mm (Baker); b i l l 11mm; Tail l64-178mm; tarsus 66-76mm (Delacur). The female is a bright rufus brwn, finely vermiculated. Frehead and face yellw brwn cnstrasting with grey nape and crest. Bill black, legs red, smetimes with small spurs. Length mm; wing mm; b i l l 9-11mm; tail l40-154mm; tarsus 56-70mm. 1.2 Satyr Tragpan (see Plate 2): The Satyr Tragpan is a large partridge-shaped bird with marked sexual dimrphism. The male is brilliantly clured deep crimsn red. The back is live brwn, the neck and upper breast range red, and the lwer breast deep red; all sptted with black-lined white celli which are larger n the lwer parts. The shulder f the wing is crimsn and the rest is dark brwn, mttled buff. Tail black. Head and crest are black, the latter with a crimsn streak either side. Bill black, lappet blue and scarlet, fleshy hrns blue. Legs

204 189 dull pink. First-year males are brwn with red n upper breast and neck, with a few white celli. Length mm; wing mm; tail mm; b i l l l4-l6mm; tarsus 35-95mm (Delacur). The female is variable, rufus t dull brwn abve, barred and bltched with black and buff. Tail barred rufus brwn. Pale stripes n crwn. Much paler and mttled belw. Bill brwn; legs pale. Length 575mm; wing 215-2j55mm; b i l l 11mm; tarsus 66mm; (Delacur). l.j Kklass Pheasant (see Plate 3): Kklass are medium-sized pheasants with cnsiderable variatin in the plumage and shwing sexual dimrphism. The male is silver grey abve, streaked with black and brwn. The fully-feathered dark green head has a brwn crest 100mm lng, with lng glssy black lateral tufts. There is a prminent white patch n either side f the neck. Rufus brwn belw, lightly streaked with black. Pinted wings and pinted wedge-shaped brwn t a i l. Bill black, legs dark. Length c 6l0mm; wing 2l6-232mm; b i l l 27mm; tarsus 62mm; t a i l 203mm; (Fleming). The female is mttled black and reddish brwn abve, finely streaked with buff. Crwn chesnut buff with shrter crest and n ear tufts. Underparts are pale buff, streaked black, narrwly n breast, bradly n flanks. Thrat white. Bill brwn, legs grey. Length 520mm; wing 212mm; b i l l 29mm; tarsus 55mm; t a i l 166mm; (Fleming). 1.4 Himalayan Mnal (see Plate 4): The Himalayan Mnal is a large dumpy pheasant similar t a Snwcck (Tetragallus), but shwing marked sexual dimrphism. The male is a brilliant iridescent, brnze-green, blue, and purple, with a large white back patch and a brest f lng racquet-shaped green feathers.

205 190 The head is green with the back and sides f the neck an range-cpper clur. Underparts are a velvety-black. The shrt, brad, square tail is range rufus. Orbital patch blue, large b i l l grey, legs dark. Length c 720mm; wing mm; b i l l 50-54nm; tarsus 70-80mm; t a i l mm; crest 75-88mm (Baker). First year male is like the female, but mre mttled black and rufus with sme metallic feathers abve. Female is a variable brwn, mttled and streaked with paler and darker brwn, a shrt head tuft, blue rbit and white thrat. White lwer back and in t a i l. Bill buff, legs pale. Length 635mm; wing 300mm; tarsus 78mm; t a i l 200mm (Delacur). 1.5 Cheer Pheasant (see Plate 5): The Cheer is a lng-tailed, barred brwn pheasant with sme sexual dimrphism. The male has buff and rusty upperparts, clsely barred with black and grey. The head and crest are brwn with a bright red naked rbital patch. Lng pinted buff clured tail barred black. Underparts pale buff grey, barred n flanks. Bill yellwish, legs grey. Length 950-lOOOmm; wing mm; b i l l 25-29mm; tarsus 74-78mm; t a i l mm. Immature is similar t female, but duller. Female is like the male but smaller with mre chestnut belw, shrter crest, and a dull red rbit.

206 APPENDIX 2 DIRECT ENUNTER CARD DETAILS Species: Sex: Immature/yung Time: Date: ID: vis/sund Grid: Weather: Slpe angle: Elev: S.Asp: Wdland/Grassland/Scrub Lcality features: Dm. Sp.: Tree height: U.Density: Tp Canpy: HI: Length Obs.: Escape: Gr/Ps Dist: Up/Dwn Behaviur: Vcal/ther

207 APPENDIX 3 QUESTIONNAIRE SUPPLIED TO W.P.A. MEMBERS VOCALIZATIONS Kklass b a t y r m Tragpan Mnal Calling perid(s) (Time f year) D first year males call at different times f year t adult males? I f s, when? D females call? I f s, when? Daily calling perid: Mrning nly Mrning & Evening Evening nly Apprx. length f daily calling perid(s) I f latter varies with time f year, please give details. FEEDING Preference fr feeding: Mrning All day Evening Mrning & Evening D birds dig fr fd? BREEDING Display by males, between which dates has this been bserved? Mating, apprx. dates bserved. Has display been bserved mating has taken place? after Has further mating between same pair taken place?

208 193 Kklass _ S a t y r Mnal Cheer Tragpan Has agressive behaviur twards birds in adjacent pens been bserved? I f s, what species? Rati f males t females during breeding seasn. Nest site preference (Grund r abve grund) At what age can male birds be distinguished frm females? Brief descriptin f display f male. Any f ther ntes which may be interest. MEMBERS NAME: ADDRESS: TELEPHONE NO:

209 i 194 KM 1 H ' " u CD c c H T) U O <D CM 0) O C t w 0),2 PH xs H a Pi,V1 a LU 00 H PH OH < "1 dx P** ^ Lfl st m JL Csl

210 i!' I ; If PC * - * Li' : ft fl i J i HI i a n ' J r 1 LU i i fl, : n 1 9 '! uj a! 0 i. CM - -r H PH On a 1 > 1 i r vo LTl <f OH ^ n u i r.it I :

211 ! S m m lqt ft 7" ft.).vf' tit* y. <SM HEra %4 X 1 1 lira s?\ '-'JW' Jk.His '!;.?*'.«S S B K S id i I A 'hi '* mm m IKS! 1.1 j UT.i'.'siSV' 3 A,.'.7. v. Hi 4 *re a- p. >3 LJ p 1JJ i a, O '21! EH I F 4 P >3 '.V. '.1 i, >' a fire A. it 1 I 4' 1 Ml r r f > f ; i>- in <f nn M

212 197 IA rh rh 1,Q I cc H ID i.; 00 I i i i ft fir CSS! V2> (US' > Si 0 : -! r I J ( 1 s P LJ LU! ft! i > 1 03 I n /Si as t 1l f,1 a r i I '5 i a-' i i i_n i; v m k 1 I.. > i I i

213 r «a & 1 1.! M St i m i m m r v? N ' ' ' '.V,! f 1 1 Ufa 198 :) 1 >'< 6 1 ; > \l 'a d.'. f r * t 1 t : - * O LJ n It- s V i 1? I: 1 i T J 1 1, i IM I 1 I. :J <!» : J I i

214 c i> tn <f m CM ZH>I

215 *R _ }(*»! <5? #f ; r:;! 200 ITS -Q E S5 H IH CD c fa; c a; X pq H rh O B O m cu cd O -P a CCJ m as a> ra >1 tfl >- < < ID < 50 X H p PH OH c v i_n <f m cvj

216 S A ^? T ^ r J M u -Q E 3 U CD fad 2 H 13 CH O O <D s f-h 0) ft >5 "P I -p c w CD 43 OH W t r-j >5 O g fn bo s O X H Q PH >- < < a; CD < Z 5* c 201 a LU ZH>I

217 as c I! O 1 > 1! z u u sac IB 1 ft i u : U mm * r. 1.0 rtr, ID (-7 P 11 L! - LJ LU i w ft r '.'0 i^- 9* at a, < I \ OO r-. v > in sj- nn (N 7U\J