^ I 5 S I ^ 3 "S3 1 avaa n2 Ll B RAR 1 es^smithsonian institution noixdxixsni nvinoshiiws^ss i uvu a rT '/) —p W CO . ,v . CO - AJ (r\ ^ _ CA ^ />« ^ XvV^V^I 7a __. A:. O XQ/ypcX " o xgft dcJZ " ^^Uixs^X q ^ “ N^xii<^X C ■’institution NOIXnXIXSNI^NVINOSHXIlMS S3 I b Vb a II “'l I B R AR I ES^SMITHSONIAN^INSTITUTION T? X^VQSV^ ^ X^VAS^X m '■ \X^V " \JVAS»2>^ m S3lbVb3n~ LIBRARIES SMITHS0N1AN~INSTITUTI0N NOIXfUIXSNI~NVINOSHXIWS S3lbVbail n , z v> z - <■ n z tn . h z a h z <^v a -h wMZ/S/^ z 1 » o > S ' > ^vj^ASt^x -g ' ^ 2 > ^INSTITUTION °° NOIinilJLSNI NVIN0SH1IWS^S3 I d Vd 3 ll\| B RAR I ES^SMITHSONIAN INSTITUTION = _ CO ^ z \ ^ ^ ^ 5; <0 9 c CO CO O -' V^' ” Q Z _j Z -J Z _ S3ldVdail LIBRARIES SMITHSONIAN INSTITUTION NOIlfUllSNI NVIN0SH11INS S3ldVd3n ’ ; ^S|| I^| I ^ 5 A? f o N^uTls^ r rn INSTITUTION NOIlfUllSNI NVIN0SH1IWS S3ldVd3ll LIBRARIES ^SM ITHSON IAN INSTITUTION -.v CO 2: O X^\ pc^x ” ^|fP^ O __ x^uixs^x Q “‘INSTITUTION NOlinillSm^NVINOSHlIIAIS S3 I dVdan^LIBRARI ES"2 SMITHSON I AN”* INSTITUTION z r~ , Z _ r* z n x^vosv^y ^ x^as^x m ' v^v x^ xjyAsn>^ fq o £ co " £ co ~ f: co S3 I dVda 11 LIBRARIES SMITHSONIAN INSTITUTION NOIlfllllSNi NVIN0SH1IWS S3ldVd8!l o Z \, CO Z * CO z co x f°f?' "3^ | g jj; S pff ' ^ > n^' > x^VAS)ii/ ^ S X^osv^^ > ' S ^INSTITUTION ^NOIlfUllSNI ^NVINOSHimS^S 3 I d Vd 0 Ilf LI B RAR I ES^SMITHSONIAN ^INSTITUTION ° co z to ^ - z ^“~ w II . II Z JvXc- . ... rr™^ -*• yj O liSNI^NVINOSHilWS^SB I avaa n^UI BRAR I ESZSMITHSONIANJINSTITUTION NOliniUSNllwiNOSHJ w tvAs^x m R I ES ^SMITHSONIAN "INSTITUTION WNOIinilXSNI — NVINOSHllWS S3 I BVB 8 IT “LI B RAR I ES SMITHSOi w „ _ z . w z <£ z , g S A S 5 ^ fi /s ’ o S IppNpi ^ S x||2|^ 5 t i , w/ j0?' S Witu&y § P ^ — S >*' S V*!SSs^/ >■ 2 ''.Tjjr > IXSNI NVINOSHilWS^SaiBVaail^UBRARlES^SMITHSONIAN INSTITUTION W NOIlfUllSNI NVINOSH 5 X f" 5 « .„. ? _ "> z - s:mxi c ^ ^ s a * c O X^osh^X ««» O 2 _* Z -J ^ RIES SMITHSONIAN INSTITUTION NOIlfUllSNI NVIN0SH1IIMS S3IBVBan LIBRARIES SMITHSO ^ 2 r- __ z r- z r- 2 ° . «, X^iX^X O •" ., O Z. X<^2^x ° fc „ m ^ jg Ng?msgx m -IXSNI NVINQSHIIINS^SB I H VH 8 ll”*LI B RAR I ES ^SMITHSONIAN ^INSTITUTION^ NOIlfUllSNI NVINOSH — _ _ IT z ~ X^>lllST^> jg ^ > R I ESWSMITHSONIAN_ INSTITUTION NOIlfUllSNI^NVINOSHlIWS^SB I B VB 3 11^ >-l B RAR I ES^SMITHSO & * ._„• *.-2 | ^ ^ 1 1 ^ £0 O py^ _ x^iiiiv^ q USNI^NVINOSHillNS S3 I U VB 0 11 B RAR I ES^SMITHS0N1AN_INSTITUTI0N NOIlfUllSNI NVINOSH _ jy^^y m g " ■ m ^ \w^ R I ES SMITHSONIAN "^INSTITUTION^ NOIlfUllSNI '"NVINOSHllWS^S 3 I MVB 0 11 “*LI B RAR l ES SMITHSO ^ z t co z to z , ^ Z ^ H Wk/^. z o x 8 (E, 5 | A t l'r ' ' | I1SNI_NVIN0SH1IIMS^S3 IBVBan^LIBRARI ES^SMITHSONIAN ^INSTITUTION *° NOlifllliSN I _ NVINOSH ^ v> Ik. ^ /&Sw!&\ w zTS5\5x ^ ^ ^ w k fflrba i /M ~ of the Bombay Natural History Society April 1987 Rs. 90 NOTICE TO CONTRIBUTORS Contributors of scientific articles are requested to assist the editors by observ- ing the following instructions: 1. Papers which have at the same time been offered for publication to other journals or periodicals, or have already been published elsewhere, should not be submitted. 2. The MS. should be typed (double spacing) on one side of a sheet only, and the sheets properly numbered. 3. All scientific names to be printed in italics should be underlined. 4. Trinomials referring to subspecies should only be used where identifica- tion has been authentically established by comparison of specimens actually collected. 5. ’ Photographs for reproduction must be clear and show good contrast. Prints must be of a size not smaller than 8.20 x 5.60 cm (No. 2 Brownie) and on glossy glazed paper. 6. Text-figures, line drawings, and maps should be in Indian ink, preferably on Bristol board. 7. References to literature should be placed at the end of the paper, alpha- betically arranged under author’s name, with the abridged titles of journals or periodicals underlined (italics) and titles of books not underlined (roman type), thus: ® ' Banerji, M. L. (1958): Botanical Exploration in East Nepal. J. Bombay nat. Hist. Soc. 55(2): 243-268. Prater, S. H. (1948): The Book of Indian Animals. Bombay. Titles of papers should not be underlined. 8. Reference to literature in the text should be made by quoting the author’s name and year of publication, thus: (Banerji 1958). 9. Synopsis : Each scientific paper should be accompanied by a concise, clearly written synopsis, normally not exceeding 200 words. 10. Reprints : Authors are supplied 25 reprints of their articles free of charge. In the case of joint authorship, 50 copies will be given gratis to be distributed among the two or more authors. Orders for additional reprints should be in multi- ples of 25 and should be received within two weeks after the author is informed of the acceptance of the manuscript. They will be charged for at cost plus postage and packing. 11. The editors reserve the right, other things being equal, to publish a member’s contribution earlier than a non-member’s. Hombill House, Shaheed Bhagat Singh Road, Bombay 400 023. Editors, Journal of the Bombay Natural History Society. VOLUME 84(1) : APRIL 1987 Date of Publication: 15-10-1987 CONTENTS Page Changes in the soft part coloration of the Indian Reef Heron, Egretta gularis (Bose) related to age and breeding status. By B. M. Parasharya and R. M. Naik .. 1 Survey of the Freshwater Turtles of India Part II: The Genus Kachuga. By Edward O. Moll. (With two colour plates and four text-figures ) 7 The Butterflies of the Nilgiri mountains of Southern India (Lepidoptera: Rhopalocera) . By Torben B. Larsen .. 26 Notes on comparative body size, reproductive effort and areas of manage- ment priority for three species of Kachuga (Reptilia, Chelonia) in the National Chambal Sanctuary. By R. J. Rao and L. A. K. Singh. ( With a plate and three text-figures ) . . 55 The genus Piper Linn, in Karnataka, India. By B. A. Rahiman and M. K. Nair. ( With eight text- figures ) . . 66 Polychaeta of the Pulicat Lake (Tamil Nadu). By S. K. Sunder Raj and P. J. Sanjeeva Raj. (With four plates) . . 84 A Catalogue of the Birds in the Collection of Bombay Natural History Society — 32. By Humayun Abdulali . . 105 Geographical distributional list of Ichthyofauna of the Garhwal Himalaya with some new records. By H. R. Singh, S. P. Badola and A. K. Dobriyal. (With a text- figure) . . 126 Sundarbans honey and the mangrove swamps. By Kalyan Chakrabarti . . 133 Annual reproductive cycle of the male field rat, Rattus rattus brunneusculus (Hodgson) in hilly terrain of Mizoram. By N. S. Chauhan and R. N. Saxena. (With a plate) . . 138 Studies on Anopheles (Cellia) maculatus Theobald, 1901 in Bastar District, India (Diptera: C’ulicidae). By Zakir Husain Husainy. (With five text-figures) 145 On some insect galls on Tectona grandis Linn, from India. By P. Jayaraman. (With seven text- figures) . . 157 Two UNUSUAL CASES OF HOMONYMY IN ORTHOPTERA WITH NEW NAMES FOR SPECIES from India. By M. S. Shishodia and R. K. Varshney . . 167 New Descriptions: A new species of Asynapta Loew (Diptera: Cecidomyiidae: Porricondylinae) from Aurangabad, India. By R. M. Sharma. (With twelve text- figures) . . 170 On a new Cyprinid fish of the Genus Danio Hamilton (Pisces: Cyprinidae) from Manipur, India. By R. P. Barman. (With a text- figure) . . 172 Ophiopinotus pinotus gen. et. nov. (Hymenoptera: Torymidae). By Tasawwer Husain and Prem Prakash Kudesia. (With seven text-figures) . . 175 Taxonomic studies on the Marine Ostracoda from the east coast of India. By C. Annapurna and D. V. Rama Sarma. ( With four plates) . . 177 Mecistocerus monubumensis sp. nov. (Coleoptera: Curculionidae: Cryptorhyn- chinae) from India. By Lehna Singh Arya and H. R. Pajni. (With three text- figures ) .. 181 Contribution to the knowledge of Desmids of India — some new taxa from Karnataka State. By G. R. Hegde. (With five text-figures ) . . 183 Dasineura psorafeae (Diptera: Cecidomyiidae) — a new Gall-Midge, infesting inflorescences of Psoralea corylifolia Linn. By R. M. Sharma. (With thirteen text-figures ) . . 186 Obituary S. P. Shahi (1917-1986) . . 190 Reviews 1. Annotated checklist of the birds of Hong Kong. (H. Abdulali) .. 191 2. Threatened animals of India. (R. Reuben) . . 191 3. The useful plants of India. (M. R. Almeida) . . 192 Miscellaneous Notes: Mammals: 1. Group number and composition of Hanuman Langur (Presbytis entellus ) in Jaipur, India. By Reena Mathur and B. Ram Manohar (p. 193); 2. Occurrence of the Bicoloured Leaf-nosed Bat (Hipposideros fulvus ) in Rajasthan. By S. Bhupathy (p. 199); 3. Sighting of a Rusty Spotted Cat (Felis rubiginosa ). By Digveerendrasinh (p. 200); 4. Note on the sighting of a Caracal (Felis caracal ) at the Sariska National Park. By Divyabhanusinh (p. 201); 5. Occurrence of Large Indian Civet (Viverra zibetha) in Orissa. (With a photograph). By L. N. Acharjyo and S. K. Patnaik (p. 201); 6. A note on the food of the Small Indian Civet (Viverricula indica) at Point Calimere Wildlife Sanctuary, Tamil Nadu. By M. Ayyadurai, V. Natarajan, P. Balasubramanian and S. Alagar Rajan (p. 203); 7. Unusual coloration of Nilgai (Boselaphus tragocamelus) . By M. K. Ranjitsinh (p. 203). Birds: 8. Occurrence of the Barheaded Goose (Anser indicus ) in South India. By M. Krishnan (p. 204); 9. The Osprey (Pandion haliaetus haliaetus ) preying on a gull. By M. C. A. Jackson (p. 205); 10. Occurrence of Greyheaded Lapwing. Vanellus cinereus (Blyth) in Bangalore. By S. Subramanya (p. 205); 11. Addition to the birds of Point Calimere, S. India. By Nitin Jamdar (p. 206); 12. Further additions to the avifauna of Point Calimere, Ruddy Shelduck (Tadorna ferruginea). By R. Sugathan, David S. Melville and S. Alagar Rajan (p. 206); 13. The Whitewinged Black Tern, Chlidonias leucopterus (Temminck) in Saurashtra, Gujarat. By Taej Mundkur (p. 208); 14. Occurrence of Pied Crested Cuckoo (Clamator jacobinus ) in Suru valley, Ladakh. By Nitin Jamdar (p. 208); 15. Sighting of Blacknaped Oriole. By D. P. Banerjee (p. 209); 16. Nest of the Pied Myna Sturnus contra Linnaeus. By Vasant R. Naik (p. 210); 17. Kaliveli Tank and Yedayanthittu Estuary — A little known wetland habitat in Tamil Nadu. By Pieter (p 210); 18. The Redfronted Babbler Stachyris rufifrons and Redheaded Babbler S. ruficeps in northern Thailand. By C. J. O. Harrison (p. 214); 19. The Indian Grey Tit (Parus major ) on an abandoned honey comb. By Debi Goenka and Heta Pandit (p. 218); 20. Host plants used by Baya Weaver bird (Ploceus philippinus Linn.) for nesting in eastern Rajasthan. By Satish Kumar Sharma (p. 218). Reptiles: 21. Distribution of the keeled box turtle Pyxidea mouhotii (Gray). By Indraneil Das (p. 221); Breeding the king cobra (Ophiophagus hannah ) in captivity. (With a photo- graph). By Shekar Dattatri (p. 222); 23. A list of the snakes of the Bhavnagar District, Gujarat State, By Raju Vyas (p. 227). Amphibia: 24. A note on the breeding habits of Jerdon’s Ramanella, Ramanella montana. By A. G. Sekar (p. 231); Fishes: 25. On the synonymy of Danio feegraclei Hora, 1937 from Burma (Pisces: Cyprinidae). ( With a text-figure). By R. P. Barman (p. 232). Insects: 26. Observations on Indian Trabutinini Silvestri and Phenacoccini Sulc (Pseudo- coccinae: Pseudococcidae : Homoptera). By R. K. Avasthi and S. Adam Shafee (p. 235); 27. Studies on the biology of Parnara naso Fabr. (Lepidoptera : Hesperiidae) . By V. N. Rao and K. S. Behera (p. 238); 28. Distribution of Drosophila species inhabiting the tropical rain forests of Sampaje Ghats (Coorg District), Karnataka, South India. By N. Muniyappa and G. Sreerama Reddy (p. 240); 29. Polymorphism in Callosobruchus chin- ensis (Linn.) (Coleoptera: Bruchidae). By H. R. Pajni, Sadhana Sahnan and Ruchira Sharma (p. 245); 30. Occurrence and life history of Cassida circumdata Herbst (Coleoptera: Chrysomelidae) in Keoladeo National Park, Bharatpur, India. ( With four text-figures ). By M. John George and K. Venkataraman (p. 248). Other Invertebrates: 31. A note on the food of the spider Argiope arcuata Simon (Family: Araneidae). By V. Natarajan (p. 253). Botany: 32. New distributional records of Dicots for Punjab State (India). By M. Sharma (p. 254); 33. Caldesia parnassiifolia (Linn.) Perl. (Alismataceae) — A New Record for Western Himalaya. By P. C. Pande and Y. P. S, Pangtey (p. 257); 34. Notes on Cypera- ceae of Maharashtra. By S. M. Almeida and M. R. Almeida (p. 257); 35. Oianthus disciflorus Hook. f. — A little known taxon from Peninsular India. (With a text-figure). By R. R. Venkata Raju and T. Pullaiah (p. 258); 36. On the identity and synonymy of Hedyotis congesta R. Br. (Rubiaceae). By D. B. Deb and Ratna Datta (p. 261); 37. A note on the Brachystelma glabrum Hook. f. (Aselepiadaceae) from South India. By E. Vajravelu (p. 262); 38. Notes on the distribution of rare and little known Carex rostrata Stocks from north-west Himalaya. (With two text-figures) . By R. D. Gaur, K. S. Negi and J. K. Tiwari (p. 263); 39. Two new records of grasses from Andhra Pradesh. (With two text-figures). By N. Yesoda, P. Venkateswara Prasamia and R. R. Venkata Raju and T. Pullaiah (p. 265); 40. Rectification of the position of Anthraxon microphyllus (Trin.) Hochst. var. hindustanicus (Jain & Deshpande) Almeida and Almeida. By P. C. van Welzen (p. 268). © JOURNAL OF THE BOMBAY NATURAL HISTORY SOCIETY 1987 APRIL Vol. 84 No. 1 CHANGES IN THE SOFT PART COLORATION OF THE INDIAN REEF HERON, EGRETTA GULARIS (BOSC) RELATED TO AGE AND BREEDING STATUS1 B. M. Parasharya2 3 and R. M. Naik:' Age dependent changes in the soft part coloration of the Indian Reef Heron, Egretta gutaris (Bose) from the day of hatching to adulthood are described. The nuptial colour changes in the soft parts are correlated with the nesting status. Introduction Coloration of the soft parts of the Indian Reef Heron, Egretta gularis (Bose) changes with the age and breeding status, so that the coloration when viewed with other evidences gives a clue to the bird’s age and breeding status; a knowledge of the soft part colour changes has been useful in some of our field studies on the heron. Moreover, in certain areas where the reef heron is sympatric with the Little Egret, Egretta garzetta, they tend to interbreed, and since a striking difference 1 Accepted August 1985. 2 Present address : B. A. College of Agriculture, Gujarat Agricultural University, Anand 388 110, India. 3 Department of Biosciences, Saurashtra Univer- sity, Rajkot 360 005, India. between the two species is coloration of their soft parts, we feel that a detailed description of the soft part colours at different stages of life in both the species should prove useful in further studies on the inter-relationship between the two herons. Plumage colour variations and sequence of plumage changes in the reef heron are des- cribed earlier (Naik and Parasharya 1983). However, a brief review is necessary here. In the white as well as grey phases the natal down is white in most of the feather tracts, but it is subjected to some colour variation ranging from light grey to grey in the dorsal tract. The pin juvenile feathers start appearing, pushing out the natal down at the age of 6-8 days. The juvenile white plumage ranges from an almost pure white to one heavily dappled with grey. The juvenile grey is grey JOURNAL, BOMBAY NATURAL HIST. SOCIETY, Vol. 84 dorsally and light grey or white ventrally. The full adult plumage is attained at the age of 20-21 months. The adult white plumage is pure white and the adult grey is grey all over except the chin and throat which are white. Both the juvenile and adult grey plumages may, or may not, have wing patch of variable size on one or both the wings. Before the breeding season a drooping nuchal crest of two narrow plumes develops in both the sexes and phases. In addition to the crest, filamentous ornamental feathers deve- lop on the scapular region and long, narrow, lanceolate plumes on the lower parts of neck and upper breast. Nuptial feathers are shed just after the nesting is over, but in some cases it happens in April when the birds are still feeding their young, so that in the begin- ning of the breeding season (February) 90% of the birds may be with nuchal crest but during the second peak of breeding (July- August) only 40% may have it. Materials and Methods The present report is based on our field notes made during the studies on the distri- bution (Naik and Parasharya in press), breed- ing biology (Parasharya and Naik in press', Naik and Parasharya in prep.) and ecology of the reef heron (unpublished) and also on our observations of some captive birds maintained in our University aviary for studying the sequence of plumage changes and polymor- phism (Naik and Parasharya 1983). Observations Colours of the soft parts described below are applicable to both the grey and white phases of the Reef Heron, unless mentioned otherwise. There were always a small number of E. garzetta and/or garzetta-gularis hybrids inter-breeding with E. gularis in the breeding colonies, so that some of the variations in the coloration of soft parts may possibly be attri- buted to the hybridization. The age dependent changes in the plumage and coloration of soft parts are summarized in Table 1. Downy nestlings Skin. The freshly hatched chicks have pink- yellow (flesh coloured) skin all over. The skin appears greyish green at the age of 3 days when the feather follicles deep within the skin start generating the juvenile plumage. Eye. Iris colour of the freshly hatched chick is brown, and it remains so until the natal down is pushed out by the juvenile feathers. Beak. The beak of freshly hatched chick is pinkish yellow with some brown at the tip of both the mandibles and base of the upper mandible. By the time the juvenile feathers start appearing as pin, the brown on the upper mandible has increased in density to be prominent. Though the brown on the lower mandible also tends to spread a little, the lower mandible continues to remain almost yellow throughout the downy plumage stage. Legs. At the age of about 3 days, a slight green tinge appears on the tibia and upper tarsus and the feet start turning light yellow. Gradually the tibia and tarsus become light greyish green and the feet become light yellow, tinged with green. Lore. Lore (facial skin) which is pink- yellow at the time of hatching gradually turns dark brown to grey. Chicks in juvenile plumage and subadults Skin. The skin which appears greyish green during the initial stage of the juvenile plum- age growth, starts turning yellow after the age of 10 days, and it is conspicuously yellow 2 SUMMARY OF AGE-DEPENDENT CHANGES IN THE TYPE OF PLUMAGE AND COLORATION OF SOFT PARTS IN SOFT PART COLORATION OF REEF HERON o £ ?**■> £ O £ o £ o ft > o 3 iS O TD u 3 p 3 3 n OH >> 3 >> 8 a ^ a w £ > o > t- 2 13 & ^ +3 a ft s> g d> 22 -5 , 0) ft ^ ti <-h §t 00 £ 2 O 0) =3 e> O ft eo d> CO c3 o ft G 3 d> b c ft co >» O 5 o '>» d) ft 3 ft tH tH 00 >> ft ft 00 G d> o <*> g /-N £, a £ d) tH 00 ft CO £ o 3 £ 2 g ‘5 >> 3 ft £ o ft 3 ft tH 3 00 >» ft ft ft g £ 1 = J3 " c g 3 ,5 1=1 £ O 5h Oh ft OH d> 00 ft 60 C3 H g P» ft 4> .2 ^3 ft O d> cS d> o ft ft ft W ft ft H 3 o JD c> 6 g 3 3 3 3 3 £ o ?>> >> ;>> >> ft ■fj g ft! CO ft! CO ft CO •§ 3 3 ft ft ft ft ft ft 3 G O tH C .£ C G 6 a ft ft 'ft ft 'ft 'ft II £ o J upper mandible, lm = lower mandible. JOURNAL, BOMBAY NATURAL HIST. SOCIETY, Vol. 84 by about the 16th day. Later, the skin starts turning grey but the yellow tinge may conti- nue to persist for a long period. However, all the birds acquire uniformly grey skin before reaching the age of 6 months. Eye. Slight white tinge appears in the brown iris at the age of 11 days, and the brown progressively becomes lighter. The iris turns cream coloured by the age of 4 weeks and yellow by the age of 6 months. Beak. The upper mandible is largely dark brown or grey and the lower one yellow when the juvenile feathers start appearing. By the age of 20 days, threefourths of the upper mandible becomes yellow, the brown being restricted to the base and tip. An overall appearance of the beak is yellow at the age of 4 weeks but the brown at the base and tip continues to persist and does not disappear completely even at the age of 6 months. Legs and feet. When the pins of juvenile feathers appear, the tibia and tarsus are grey- ish green and the feet light yellow. The tibia and tarsus gradually turn yellowish green, but at the age of 15 days, a tinge of grey appears on the front face of the tarsus. At the end of 4 weeks, the tibia and tarsus are greyish green, and the frontal surface of the tarsus is marked with black bands. At the end of 6 months a slight dull green colour still persists on both the tibia and tarsus of some birds, though they have turned completely black in the others. The feet remain light yellowish green at least up to the age of 6 months, but thereafter they gradually turn light yellow. Lore. The lore remains dark till 4 weeks of age. Thereafter, light greenish yellow colour appears among brown and gradually the yellow tinge increases to make the facial skin yellow with only a little tinge of green by the end of 6 months. Adult Eye. The iris of the adult bird is yellow. Beak. The beak colour of the adult bird is variable. Variations in the beak colour are categorized into two basic types, yellow and black, which occur in the grey as well as white morphs in the non-breeding as well as breeding seasons. It may be mentioned here that the birds feeding in certain types of habitat often get their beak smeared with mud which on drying obscures the true beak colour (Naik et al. unpublished). Yellow : This may range from pure yellow to pinkish, (a) Pure yellow: Both the man- dibles are pure yellow. This is the most fre- quent beak type, (b) Pinkish yellow: Though the basic colour of both the mandibles is yellow, there is a pinkish tinge all over. Though infrequent, the pinkish yellow beak was recorded during the non-breeding season also. Black: Both the mandibles are completely black without any yellow tinge. Lore. The lore of the adult birds is yellow. Occasionally we came across birds that had small patches of blue on the yellow lore and also those with bluish green or blue lore. Legs. In non-breeding adults, the tibia and tarsus are black and the feet are bright yellow. The bright yellow of feet often ex- tends up to the distal end of tarsus. Nuptial colour changes The beak, lore and iris are usually yellow, but they acquire the nuptial colours during the courtship (pre-pairing and pair formation) stage of the nesting cycle. In both the grey and white phases, the beak usually remains yellow, but in some it may acquire a red tinge. If the bird happens to have a black beak, no recognisable change in the beak colour occurs. 4 SOFT PART COLORATION OF REEF HERON The iris which is normally pale yellow turns to bright yellow and may even acquire a red rim in some birds. The lore turns to bright pink red or crimson. The feet and lower tarsus which are normally yellow turn to pink-red. The upper tarsus and thigh region which is dull grey normally do not change colour, but in some birds even these parts also acquire a red tinge. Pink-red in the feet and lore of the individuals perching alone (unpaired) is brighter than that of paired birds or nest-building birds. The pink colour tends to fade during the nest-building phase, and it almost completely fades out when the first egg is laid. The toes turn to pale orange- yellow and ultimately resume their original colour. The lore first turns to pale pink, then to bluish white and thereafter to pale yellow and ultimately to its original yellow colour. If a pair loses its clutch during the incuba- tion period and if it is to re-build the nest, the nuptial colours return to the soft parts once again. The birds breeding in juvenile plumage also acquire nuptial colour when they start nesting. We recorded the beak colour of 181 white- phase reef herons in different stages of their nesting cycle in breeding colony and the re- cord is summarized in Table 2. These data illustrate some of the points we have made hereinbefore for both the phases, namely the yellow is the most frequent beak colour for the birds in every stage of the nesting cycle, the 3^ellow beak may turn pinkish in some (but not in all) birds during the nesting season and a small proportion of birds have their beak black in colour which remains apparently unchanged through the nesting season. Discussion The age-dependent changes in the colora- tion of soft parts of Indian Reef Heron, des- cribed in this paper and summarized in Table 1 for the convenience of the readers, should be useful in ageing the chicks and distinguish- ing the older from younger juveniles in the field. Table 2 Frequency of different beak colours of the WHITE-PHASE REEF HERONS IN RELATION TO THEIR NESTING STATUS, AS RECORDED IN THE NEW PORT Colony, Bhavnagar Nesting stage Number of birds with designated beak colour pinkish yellow yellow black all com- bined Unpaired 1 34 4 39 Courtship 2 20 4 26 Nest-building 2 22 4 28 With eggs 4 43 5 52 With chicks 1 31 4 36 All stages combined 10 150 21 181 In the present state of our knowledge it is not possible to say as to how much variation in the soft part coloration described here is attributed to hybridization between the Little Egret and reef heron that has been occurring since long in our area (Naik and Parasharya 1985) However, it is certain that the appear- ance of blue lore and black beak (which are characteristic features of the Little Egret) in our population of the reef heron is a result of gene flow from the Little Egret populations. Nuptial colour changes in soft parts during early stages of the nesting cycle are now known to be widespread in the family Ardeidae (Blaker 1969). Ali and Ripley (1968) have mentioned colour changes asso- 5 JOURNAL, BOMBAY NATURAL HIST. SOCIETY, Vol. 84 dated with breeding in many Indian Ardeidae, but not for the Indian Reef Heron. In certain populations of the reef heron in Africa, the beak turns from yellow to black as a part of nuptial colour changes (Hancock and Kush- lan 1984). In our population of the reef heron, however, only noticeable change in the beak colour is for the yellow to turn pink- yellow in some birds. A few birds that we observed having pink-yellow beak in the non- breeding season were probably exceptional; an increased blood flow in the subcutaneous network of blood capillaries, either because of some emotional stress or premature hormonal changes, may have turned the beak pinkish. Recently, Hancock and Kushlan (1984) have merged the reef heron with the Little Egret. These authors consider the Little Egret, Egretta garzetta (Linnaeus) as a polytypic species consisting of six subspecies, namely garzetta — an all white race with a range in the temperate, subtropical and tropical zones of the Old World, nigriceps — also an all- white race occurring on the islands of south- eastern Asia and the southwest Pacific, im- maculata — also an all-white race of northern R EFE Ali, S. & Ripley, S. D. (1968): Handbook of the birds of India and Pakistan. Vol. I. Oxford University Press, Bombay. Blaker, D. (1969) : Behaviour of the Cattle Egret Ardeola ibis. Ostrich 40(3) : 75-129. Hancock, J. A. & Kushlan, J. A. (1984) : The Herons Handbook. Croom Helm, London. Naik, R. M. & Parasharya, B. M. (1983): Sequence of plumage changes and polymorphism, in the Indian Reef Heron, Egretta gularis. Sand- grouse 5: 75-81. — — — (in press ) : and eastern Australia, gularis — a polymor- phic race on the west African coast from Mauritania to Gabon, schistacea — another polymorphic race distributed along the coast- line extending from the west coast of Indian subcontinent to Persian Gulf and the Red Sea, and south along the African coast up to Tanzania, and dimorpha — also a poly- morphic race found on Madagascar and other islands in and around Mozambique Channel. While available information on the soft part coloration of the above mentioned assemblage of herons are illustrated by Han- cock and Kushlan (1984), detailed informa- tion of the changes in soft part coloration dependent on age and nesting status in these birds are urgently needed to understand their evolution. Acknowledgement One of us (B.M.P.) is grateful to the Council of Scientific and Industrial Research, New Delhi, for an award of Senior Research Fellowship. E N C E S Impact of the food availability, natural-habitat des- truction and regional cultural variations of human settlements on the nesting distribution of a coastal bird. Egretta gularis, in western India. J. Bombay nat. Hist. Soc. (in prep.) : Inter-breeding and gene flow within Egretta guTaris and E. garzetta complex, along the Gulf of Kham- bhat in western India. Parasharya, B. M. & Naik, R. M. (in press) : Breeding biology of the Indian Reef Heron, Egretta gularis (Bose). J. Bombay nat. Hist. Soc. 6 SURVEY OF THE FRESHWATER TURTLES OF INDIA PART II: THE GENUS KACHUGA 1 Edward O. Moll2 {With two colour plates & four text-figures) [Continued from Vol 83(3): 552] Subgenus Pangshura Contains four species — smith'd, sylhelensis, tecta and tentoria. The subgenus is diagnosed by a suite of apomorphic or derived charac- teristics summarized in Table 1 and Figures 2-6. Members are small to moderate-sized species (< 30 cm CL) with pronounced sexual dimorphism which inhabit a variety of lentic and lotic habitats through much of India. The name is derived from “panshura>” a Bengalese word for chelonian. Kachuga smithii (Gray 1863) Brown Roofed Terrapin — Plate III, A-C Identification : A small species (to 23 cm CL) identifiable from other Pangshura by a relatively low, vaulted shell (H/CL < 44%) having only a weak, horizontal spine (or none) on the third vertebral scute. Description : For coloration see descriptions of subspecies. Head moderate in size with short, tapering snout (less than length of orbit) projecting beyond lower jaw; skin at back of head divided into large irregular scales; upper jaw serrate, lacking medial notch or projection; alveolar surface broad, decked by a finely-serrate ridge on each side, 1 Accepted December 1986. x Dept, of Zoology, Eastern Illinois University, Charleston, Illinois, U.S.A. converging but not meeting at midline. Lower jaw serrate with single, projecting tooth anteriorly, alveolar surface concave except for a median symphyseal ridge and a serrate ridge along lingual surface meeting symphyseal ridge at midline; coronoid process prominent. Hyoid moderately developed; ossified portions include a single-element body with a shallow, rounded notch posteriorly and a shallow “V” shaped notch anteriorly, a pair of small, rounded elements attached at either side of anterior notch (ceratohyals?), a narrow, elongate, outwardly-bowed pair of first cera- tobranchial horns and a pair of short (2-3 times longer than wide) second ceratobran- chial horns. Shell oval, widest across a plane through seventh marginals; posterior margin of cara- pace slightly serrate; median keel low; raised areas at posterior of scutes not pronounced, obtuse; Vertebrals 1, 3 and 4 usually longer than wide, 2 and 5 usually wider than long; seam contact formula — 1M 4> 6M 8< 10>. Bridge long, exceeding length of either anterior (shorter) or posterior (longer) lobe of plastron; axillary somewhat smaller than inguinal scute; plastron truncate anteriorly; notched posteriorly; plastral formula — Ab> F> P> A> H> G. Distribution : Brown roofed terrapins occur in the Indus, Ganges and Brahmaputra Drain- ages of Pakistan, Nepal, India, and Bangla- 7 JOURNAL, BOMBAY NATURAL HIST. SOCIETY, Vol. 84 desh. Figure 9 map the distribution in India as verified by this survey. Geographic Variation : Two subspecies (one new) are recognized herein. Kaclsuga smithii pallidipes subsp. nov. Pale-footed Roofed Terrapin — PI. Ill, B&C Holotype : Field Museum Natural Flistory 224177, adult male in alcohol; collected in the Gandak River, Bherihari Wildlife Sanctuary, Bettiah (West Champaran) District, Bihar on 3 June 1983; original number 2827 Edward O. Moll. Paratypes : IJSNM 257779, adult female in alcohol; Kamali River, Royal Bardia Wild- life Reserve, 2 km N. Thakurdara, Nepal, 23 April 1985 (obtained by Joseph Mitchell); FMNH 224186, adult female skeleton, Ghagra River, near Kailaspuri at Girija Barage, Bahraich District, Uttar Pradesh, 12 January 1983. Identification'. A subspecies of Kachuga smithii differing from the nominate form by the absence of a plastral pattern and a re- duction of pigment on head, limbs, feet and penis. Description of holotype : An adult male measuring 8.6 CL, 6.2 CW, 7.8 PL, 3.1 FI and weighing 88 g. Coloration in life — cara- pace light grayish olive to brownish olive (older scutes) with pale yellow rim around the periphery; single mid-saggital black stripe with cinnamon-rufous center on Vertebrals 2 and 3. running length of the shell but becom- ing obscure on V5 and 6; plastron straw yellow, immaculate, having no dark pattern; vague dark blotches present on ventral side of marginals. Head brownish olive, lightly mottled with smoke gray dorsally; skin creamy white be- hind eyes; iris pale gray; throat immaculate, colorless; a pair of narrow dark stripes ex- tend anteriorly from eyes converging to meet at light gray snout; mandibles bright spectrum yellow; neck smoke gray dorsally, colorless ventrally, unstriped. Limbs — ground color on leading face of front limb smoke gray above elbow and late- ral half of foreleg; large triangular scales on lateral border of foreleg, narrow bandlike scales on anterior aspect; toes and webbing yellow; posterior aspect of limb, feet and medial half of foreleg colorless. Penis colorless lacking the dark pigment characteristic of most members of the order. Description of Paratypes: USNM 257779, adult female measuring 15.6 CL 11.6 CW 15.5 PL 5.65 FI and weighing .453 kg. Colo- ration of preserved specimen — carapace brownish gray with a black, middorsal stripe; plastron light, largely devoid of pattern but tiny smudges of pigment present near poste- rior lateral margins of scutes; bridge and ventral side of Marginals 3-9 with heavy con- centrations of dark pigment. Head and neck dark grayish brown dorsally with lighter cinnamon brown band running posteriorly from eye over tympanum; throat light, colorless. Limbs — anterior face of forelimb and feet generally grayish brown, becoming lighter on medial aspect; lateral border of limb light edged; hind feet immaculate, devoid of pig- mentation; skin of leading face of hind leg grayish brown. FMNH 224186, female measuring 16.2 CL 11.6 CW 15.5 PL 6.0 H and weighing 0.63 kg. Coloration in life — carapace buff with dark brown middorsal stripe; plastron straw yellow, unpatterned. Head and neck drab dorsally with lighter, cinnamon-colored band running posteriorly from eye over tympanum; tip of snout and iris light grayish blue; mandibles light orange yellow; neck dirty gray above, light cream ventrally. 8 FRESHWATER TURTLES OF INDIA Fig. 9. Distribution map of Kachuga smithii in India (see legend of Fig. 7). Solid circles indicate localities of the nominate race, K. s. smithii. Open circles indicate localities of K. s. pallidipes. The half open circle indicates Allahabad, a possible intergrade site between the two races (see text). 9 JOURNAL, BOMBAY NATURAL HIST. SOCIETY, Vol. 84 Limbs with toes and skin flap at lateral edge of hind foot immaculate, devoid of pigment, otherwise outer leading surface of fore and hind legs gray; trailing surface of legs creamy white. Distribution : The three members of the type series of this subspecies come from north- ern tributaries of the Ganges. See Figure 9 and comments on intergradation under remarks. Natural History : Little is known about the natural history of this form. All three records were associated with riverine environments. The holotype was collected during the day in a hoop trap baited with fish set near a sand bank in the Gandak River. The Nepal speci- men was collected by hand under a pile of sticks and debris in a pool on the floodplain of the Karnali River. The other female was taken by a fisherman in the Ghagra River but details are lacking. Remains of a freshwater prawn were found in the gut of the latter specimen. Remarks : Presently it seems best to regard this taxon as a subspecies of K. smithii rather than as a distinct species for the following reasons. 1) The male from the Gandak River has been selected as the holotype or best re- presentative of this race. The two female paratypes from more westerly drainages show some characteristics of the nominate form (more dark pigment on the head, limbs and ventral part of shell) suggesting intergradation. One of the original types used by Gray (1863) in the species description from the Chenab River in northwestern India (BMNH 63.2.21.87) lacks pigmentation on the pecto- ral scutes and only small amounts are present on the abdominals. This too could represent intergradation. More collections are needed to confirm whether the observed variation has resulted from intergradation rather than some other factor such as sexual dichromatism. 2) A precedent exists for this type of variation in a related species. In what appears to be a case of unusual parallelism, populations of Kachuga tentoria from northern tributaries of the Ganges also exhibit a reduction of dark pigment and have no plastral pattern. More collections are available for this species and intergradation is evident in the more western drainages of the Ganges (see K. tentoria). A peculiar situation exists with a series of specimens from Allahabad in Uttar Pradesh. The Zoological Survey of India collection has five specimens of K. smithii from this locality. Three reportedly have the dark plastral pattern (ZSI 457, 471, 472) and two do not [ZSI 451, 200 (1912)]. Another specimen in the British Museum (BMNH 1908.12.28.2) from this locality has dark blotches only on the anal scutes. Whether these specimens were actually obtained in Allahabad or whether they were obtained elsewhere and shipped to markets there is not known. Neither is there any information on the coloration of the rest of the body in this collection. The entire problem requires additional study. Kachuga smithii smithii (Gray 1863) Brown Roofed Terrapin — Plate III, A Identification'. A subspecies of K. smithii having a plastral pattern of large dark brown to black blotches on each scute narrowly bordered with yellow; sides of head, leading surface of limbs, feet and penis dark pig- mented. Description : Female (FMNH 224143); carapace brownish olive bearing a middorsal dark brown stripe; a small dark triangle decks areolar portion of Pleurals 2 & 3 and vertical dark bars border seams between the more posterior marginal scutes and the posterior edge of Pleural 4; plastron, bridge and ventral side of marginal chiefly dark but narrowly bor- 10 FRESHWATER TURTLES OF INDIA dered with light yellow; head and neck olive dorsally; a tawny blotch present behind eye; vague striping evident on lateral portion of neck; iris pale blue-gray; mandibles deep buff yellow; skin on outer surface of limbs olive with bandlike scales on forelegs appreciably lighter than ground color; vague striping pre- sent on hind legs and rump. Size and Sexual Dimorphism : Minton (1966) reported that eight females and three males from Pakistan ranged from 15.3-22.7 CL and 10.1-10.8 CL respectively. Smith (1931) recorded the largest specimen as 23.0 CL 15.5 CW and 8.5 H. This race was rarely encountered on our survey. Six shells found in garbage dumps near Rajmahal, Bihar rang- ed from 13.2 to 18.3 (mean 15.7) CL. A subadult female obtained from fishermen at Kahalgaon, Bihar measured 14.3 CL 10.3 CW 13.7 PL 5.9 H and weighed 0.315 kg. Males differ from females by being consi- derably smaller and by having a longer tail which is heavier at the base. Minton (1966) states that the tail of males projects free about 10 per cent of the carapace length whereas that of the female is about 5 per cent. Natural History : Brown roofed terrapins are typically associated with rivers and occur in current as well as more lentic habitats such as backwaters. Minton (1966) found them to be a social basking species on the Indus where they undergo a period of quiescence from early December to early March. He reported finding females with eggs in early October; a clutch of seven laid by a captive contained eggs 43 to 45 mm long and 22 to 24 mm wide. Chaudhuri (1912) reported that five to eight eggs are buried in sand nests but gave no season. Ewert (1979) reported the mean size of four hatchlings to be 3.92 CL and 3.67 PL. The species is generally report- ed to be omnivorous with a carnivorous bias (Das 1985, Minton 1966. Smith 1931). Gut contents of the subadult female from Kahal- gaon contained only plant material. Distribution : The brown roofed terrapin has been reported from the Indus, Ganges and Brahmaputra Drainages of Pakistan, India and Bangladesh. The Museum d’Histoire Natu- relle in Geneve, Switzerland has specimens catalogued as K. smithii from Assam but I have not examined these. Minton (1966) found the turtle to be common in the Indus Drainage while Smith (1931) considered it to be much rarer in the Ganges Drainage. Reza Khan (1982) also reported that the species is uncommon in Bangladesh. The species was rarely encountered on our survey. Figure 9 maps localities verified for India. Specimens were collected from the following sites : FMNH 224143 — Ganges River, Kahalgaon, c. 50 km W. Sahibgani. Bhagalpur District, Bihar. EOM 2720-2725 — Ganges River, Rajmahal, Dumka (Santhal Parghana) District, Bihar. Type locality of this race is “North- western India: Punjab; “River Chenab...” Syntypes are BMNH 1947.3.4.69-70. Much of the former Punjab is now part of Pakistan and most of the Chenab River now found in India is in Jammu and Kashmir rather than the state of Punjab. Hence the type locality could well be Pakistan. The dot on figure 9 is the western-most point of the Chenab in India. Other preserved specimens verified in- clude : MCZ 3233 — Ludhiana, Punjab. ZSI 17606 — Firozp-ur, Punjab. Kachuga sylhetensis (Jerdon 1870) Assam Roofed Terrapin Identification'. A small terrapin (20 cm CL) differing from other Pangshura by typi- cally having 13 pairs of marginal scutes and 11 JOURNAL, BOMBAY NATURAL HIST. SOCIETY, Vol. 84 a fifth vertebral scute that is wider in the anterior half than in the posterior. Description : Head medium-sized with slightly projecting snout; upper jaw slightly hooked with serrations along the tomium be- ing fine or absent; head patterned with a pair of narrow yellow stripes running posteriorly from eye to meet at mid-line on back of head forming a chevron; an additional light stripe runs along underside of the mandible curving upward to meet tympanum; neck with light longitudinal stripes. Shell steeply peaked as in K. tecta and tentoria but more serrate posteriorly; oval in shape being widest across plane through sixth marginals; median keel relatively narrow on first through third vertebrals; sharp pointed spines present on the third (largest) and fourth vertebrals; Vertebrals 3 and 4 longer than broad, V2 and 5 broader than long and VI either as wide or somewhat wider than long; scute contact formula: 1M 4> 6> 8M 11 <; coloration of preserved specimens olive- brown with a lighter vertebral keel. Plastron oval with slight notch or no notch between anals; plastral formula F >< Ab > P > H > A > G; inguinal and axillary scutes well developed; pattern a large dark blotch on each plastral scute, bridge, and ventral side of marginals. Size and Sexual Dimorphism : As typical for the genus, males are much smaller than the females. The largest specimen recorded is a female 19.7 cm CL (Jerdon 1870). Measure- ments of a typical male and female are: BMNH 1929.11.21.1 M — 8.5 CL 6.8 CW 8.0 PL 4.7 H BMNH 1947.3.4.22 F — 18.3 CL 14.3 CW 17.8 PL 8.4 H In addition to being smaller, males have longer tails which are thicker at the base than those of females. Natural History : Most specimens have come from hill streams; nothing else seems to be known of the natural history. Distribution : The Assam roofed terrapin has been reported from the Khasi, Garo and Naga Hill areas of Bangladesh and Assam. As these areas were off limits to our survey, no living specimens of this species were seen or collected. I have examined museutn speci- mens from the following localities: BMNH 1947.3.4.22 (type) — Khasi Hills, Sylhet District, Bangladesh. BMNH 1929.11.21.1 — Khasi Hills, Cherrapunji, Meghalaya, India. ZSI 110 — Cachar District, Assam, India. ZSI 3923 — Garo Hills, Assam, India. Kachuga tecta (Gray 1831b) Indian Roofed Terrapin — Plate III, D-F Identification'. A small Pangshura (23 cm CL) with a high vaulted shell (height/ length > 45%) most easily distinguished from its closest relative, K. tentoria, by its pattern and brighter coloration. Plastral pattern of small dark blotches or streaks (1-4) on most scutes; head with a large red to orange crescent- shaped blotch behind eye; neck with bright yellow stripes and limbs bearing bright yellow spots. Smith (1931) used as a key character that the second vertebral is longer than the third in tecta but not in tentoria. Although useful, this character is variable being correct in only 76 percent of the 21 K. tecta and only 63 percent of the 45 K. tentoria examined. Description : Sexes colored similarly (live female, Lucknow, U.P.); carapace raw umber (brown) with somewhat lighter middorsal stripe (can also be red) bordered in black and a narrow yellow border on marginals; plastron buff-yellow with 2-4 small, round to elongate, black markings on scutes other than gulars and/or anals which may have only one; two dark spots on bridge, one each on inguinal and axillary scutes and on underside of each 12 FRESHWATER TURTLES OF INDIA Fig. 10. Distribution of Kachuga sylhetensis (see legend of Fig. 7). 13 JOURNAL, BOMBAY NATURAL HIST. SOCIETY , Vol. 84 marginal; head dark with large orange yellow to spectrum orange crescent-shaped, post- ocular blotch curving upward from below eye and extending obliquely across head to meet and form chevron with blotch from opposite side; iris dark; mandible orange yellow be- coming orange beneath snout; neck dark, patterned with yellow stripes being brightest laterally and ventrally; outer surface of limbs dark, studded with bright orange yellow spots on scales; rump with vertical orange yellow stripes on lighter background. Shell oval, steeply pitched, widest at plane through sixth or seventh marginals, posterior edge slightly serrate; middorsal keel widest on first three vertebrals, a prominent pointed spine present on Vertebral 3; VI widest in anterior half, usually as long or slightly longer than wide; V4 flask-shaped, much longer than wide; V5 widest in posterior half and wider than long; V2 and 3 vary in length-width relationships; scute contact formula — 1> 4M 6M 8M 10M. Plastron truncate anteriorly and notched posteriorly; anterior lobe shorter than posterior, both shorter than bridge; plas- tral formula-Ab >< F > H> A> P> G; axillary and inguinal scutes large, subequal in length. Head moderate; snout shorter than orbit; skin at back of head divided into a series of irregular scales; secondary palate broad with single denticulate ridge; edge of tomium coar- sely serrate and lacking medial notch. Lower mandible ending in prominent medial tooth; alveolar surface concave except for a serrate ridge along lingual border which joins a symphyseal ridge at midline; moderate coro- noid process present. Hyoid moderately deve- loped; ossified portion comprising a single-unit, elongated body with a broad shallow notch posteriorly and a deeper, narrower notch anteriorly; a narrow, curving, elongated pair of first branchial horns and a pair of small kidney-shaped second branchial horns. Size and Sexual Dimorphism ; Females are much larger than males. The largest specimen recorded is 23 cm CL (Smith 1931). Minton (1966) reported two adult females and a male from Pakistan measured 16.4, 17.3 and 8.4 cm CL. Measurements of two females and one male examined in this survey are: Live F — 15.3 CL 11.5 CW 14.7 PL 7.3 H weight 0.51 kg. EOM 2784 (BNHS uncataloged) F — 18.3 CL 14.2 CW 17.2 PL 7.3 H weight 0.96 kg. Live M (mature?) - 6.6 CL 5.4 CW 6.3 PL 3 . 7 H weight 0 . 054 kg. In addition to size males differ from females by having a longer, thicker tail in which the vent opens beyond the carapacial rim. Hatchlings-. Four hatchlings from eggs laid by a female at Lucknow, Uttar Pradesh averag- ed 3.15 CL 2.8 CW 2.8 PL 1.8 H and 7 grams weight. Young K. tecta are strikingly attractive turtles. The carapace is bright lime green rimm- ed with a yellow orange border. There is a broken middorsal stripe of flame scarlet bor- dered in black and each pleural has a tiny black spot at the posterior dorsal edge where a lateral keel would be. The plastron is chrome orange patterned with small, irregular, black blotches on each scute including the axillary, inguinal and underside of the marginals. The head and neck are dark olive in ground color with the latter decked with bright yellow stripes. A large crescent-shaped flame scarlet blotch begins under the eye, curves upward behind the eye and meets its counter part from the opposite side at the back of the head forming a “V” shaped figure. A small flame scarlet spot marks the posterior of each eye lid. The scales of the limbs as well as the webbing between the toes are bright yellow on a dark olive background. 14 FRESHWATER TURTLES OF INDIA Natural History : Indian roofed terrapins chiefly inhabit lentic habitats (tanks, nullahs and backwaters) in the Ganges and Indus drainages. Slow moving or quiet vegetation- choked waters appear optimal. We observed this species in a weedy backwater of the Ghagra River, a small impoundment pond near Lucknow, U.P. and in a slow moving nullah crammed with aquatic vegetation which flowed into a nearby ox bow lake in the Udaipur Forest Area of Bihar. Khan (1982) reported K. tecta occupying flowing and stag- nant waters in Bangladesh. In the Narmada River where K. tentoria and K. smithii are seemingly absent, we found K. tecta to be moderately common. Like its close relative the Indian tent terra- pin, the Indian roofed terrapin is commonly seen basking on logs or the river bank. Parshad reported that it is herbivorous (Smith 1931). However, we caught one specimen in a hoop trap baited with chicken entrails. A specimen from Lucknow laid a clutch of eight eggs on January 13. The eggs averaged 37 x 21 mm and 10.75 grams. Relative to the female reproductive efforts, calculations are: RCM — 0.21, EMI — 2.1, ELI — 24.2 and EW1 — 14. Distribution : The Indian roofed terrapin definitely occurs in the Indus, Narmada, Ganges and Brahmaputra River Systems of Pakistan, India and Bangladesh. Based on the Indian distribution, it probably occurs in Nepal as well. A series of specimens in the Museum of Comparative Zoology (MCZ 3459, 3460, and 3462) labelled Rangoon, if verified would extend the range much farther East. Figure 11 depicts the distribution in India as verified by the survey: Live F — Kukrail, nr. Lucknow, Lucknow District, India. Live M — Harhi Nala, 15 km W Bettiah. Bettiah Dist,, Bihar. EOM 2658 — 6 km S Katarnia Ghat, nr. Girija Barage, Bahraich District, U.P. EOM 2868 — Bedaulia, Manika, Muzaffarpur Dis- trict, Bihar. EOM 2784 (uncataloged specimen BNHS) — Nar- mada River, Dhavdi Ghat, nr. Punasa, East Nimar District, M.P. In addition the following preserved speci- mens have been examined and verified : BNHS 1290-1291 — Chandola Lake, nr. Ahmedabad, Ahmedabad District, Gujarat. ZSI 17609 — Makhu, Firozpur District, Punjab. ZSI 21672 — Magwall Village, Jammu District, Jammu-Kashmir. Uncatalogued specimens BNHS — Hindon River, Mohen Nagar, nr. Ghaziabad, Meerut District, U.P. ZSI 18015 — Baradighi Tea Estate, Jalpaiguri Dis- trict, West Bengal. ZSI 19236 — Cherrapunji, Meghalaya, India. Remarks : The Indian roofed terrapin is the only Kachuga to be listed on CITES (Appen- dix I) or the Indian Wildlife (Protection) Act (Schedule I). Reasons for listing are obscure; we found the species to be relatively common in the aforementioned sites and it did not appear to be a popular market species. Khan (1982) judged it to be the most common turtle in Bangladesh. Nevertheless because of the turtle’s status, the aforementioned pre- served specimens were either taken as shells or in the case of EOM 2784, a turtle which drowned in a fisherman’s net. All specimens from the survey were left within the country. Kadiuga tentoria (Gray 1834) Indian Tent Terrapin — Plate II Identification : A moderate-sized Pangshura (27.1 cm CL) with a high, vaulted shell (height /length > 45%); differing from K. tecta by having one or two small reddish to brownish spots behind the eye instead of a broad crescentic band and by having a plastron with a single large dark blotch per scute or Jacking in dark markings. Description : Shell oval being widest at a 15 JOURNAL, BOMBAY NATURAL HIST. SOCIETY, Vol. 84 Fig. 11. Distribution of Kachuga tecta in India (see legend of Fig. 7). E F (A) Kachuga tentoria tentoria — Juvenile (8.1 cm CL) from the Godavari River near Manthani, A.P. (B) Kachuga t. tentoria — Ventral view of A. (C) K. t. flaviventer — Male (8.3 cm CL) from the Gandak River, Bettiah District, Bihar. (D) K. t. flaviventer — Ventral view of C. (E) K. t. circumdata — Male (8.5 cm CL) from the Yamuna River nr. Etawah. (F) K. t. circumdata — 1 Male (8.1 cm CL) and female (23.5 cm CL) from the Yamuna River near Etawah. Plate II J. Bombay nat. Hist. Soc. 84(1) Moll: Freshwater Turtles E F (A) Kachuga smithi smithi — Female (14.2 cm„CL) from the Ganges near Kahalgaon, Bihar. (B) K . smithi subsp. nov. — Male (8.6 cm CL) from the Gandak River, Bettiah District, Bihar. (C) K. smithi subsp, nov. — Ventral view of B. (D) K. tecta — Immature female (6.4 cm CL) from near Bettiah, Bihar. (E) K. tecta — Hatchling (3.0 cm CL) from Lucknow, U.P. (F) K. tecta — Ventral view of E. Plate III J. Bombay nat. Hist. Soc. 84( 1) Moll: Freshwater Turtles FRESHWATER TURTLES OF INDIA plane passing through the seventh marginals; a prominent middorsal keel runs the length of the carapace breaking at the end of each vertebral to form a knob or spine, the most pronounced being a sharp upward projecting spine on V3; in adults Vertebrals 3 and 4 typically longer than wide with 5 being wider than long; VI and 2 variable, with 1 often being hour glass or bell shaped (pinched in the middle) in large individuals; seam contact formula — 1> 4> 6M 8M 10M. Plastron truncate anteriorly, notched posteriorly; plas- tral formula — AB > F > P > H > A > G; bridge long exceeding lengths of both the shorter fore lobe and longer hind lobe of plastron; axillary somewhat smaller than in- guinal scute. Cloacal bursae present with pro- nounced villous lining. Head medium-sized with short, pointed, projecting snout; skin at back of head divided to form a series of irregular-shaped scales; upper jaw serrate lacking median notch or prominent projections; alveolar surface broad, bearing single “V” shaped denticulate ridge. Lower jaw similarly serrate with single, pro- minent, projecting tooth at apex; lower alveo- lar surface concave bordered by serrate ridge along lingual surface meeting a short symphy- seal ridge at midline. Hyoid moderately deve- loped with ossified portions including a single- element body having a prominent rounded notch posteriorly and a smaller “V” shaped notch anteriorly, a pair of thin, elongate, out- wardly bowed first ceratobranchial horns and a pair of small, rounded second ceratobran- chial horns. Distribution : The Indian tent terrapin is restricted to drainages of rivers flowing into the Bay of Bengal in India, Nepal and Bangla- desh. Figure 12 depicts the distribution in India as verified by the survey. Remarks: Until recently K. tentoria has been considered a subspecies of K. tecta. However, as both appear to be sympatric over a broad geographic area including Bangla- desh (Khan 1982), Bihar, and Uttar Pradesh (this paper); herein I follow Pritchard (1979) in regarding them as separate species pending additional study. Smith (1931) lists the type locality for K. tentoria as Dhond (Krishna River Drainage), Poona District of Maha- rashtra. See Mertens (1969) for a history of the nomenclature of these two species. - Three subspecies of K. tentoria (one re- surrected) are recognized herein. — Kachuga t. tentoria in the rivers of peninsular India; K. t. circumdata in the western and central drainage of the Ganges and K. t. flaviventer in the eastern Ganges and its northern tribu- taries. Kachuga tentoria tentoria (Gray 1834) Indian Tent Terrapin — Plate II, A+B Identification : A race with a dark plastral pattern, no pleuro-marginal ring and reddish head markings. Description : Sexes colored similarly (FMNH 224163 juv., Godavari River); carapace anti- que brown, unicolor except for hazel to amber stripe along middorsal keel from Vertebrals 1-3; plastron yellow with large dark blotches on each scute, bridge, axillary, inguinal and underside of marginals; ground color of head olive to brownish olive; a poorly defined clay band present behind eye; red markings in- clude a small red postocular spot in clay band, a smaller red mark located at dorsal posterior edge of eye and a thin poorly defined red line in occipital region; iris gray olive; mandibles straw yellow; neck with dull straw yellow stripes on lateral and ventral portions; limbs olive with edges of scutes cream; rump mark- ed with vertical black and cream stripes. Size and Sexual Dimorphism : Males are much smaller than females. Three males from the Mahanadi River measured 8.1, 9.7 and 17 2 JOURNAL , BOMBAY NATURAL HIST. SOCIETY, Vo I. 84 Fig. 12. Distribution of Kaehuga tentoria in India (see legend of Fig. 7). Starred circles indicate localities of the nominate race, K. t. tentoria. Solid circles indicate localities for K. t. circumdata. Open circles indicate localities for K. t. flaviventer. FRESHWATER TURTLES OF INDIA 10.9 cm CL. Another 7.5 cm CL with no sperm in the epididymides was judged sub- adult. Six females from this locality ranged from 19.2 to 23.0 (mean 20) cm CL. Two other females 18.5 and 18.0 cm CL with no enlarged follicles on their ovary and relatively small oviducts were judged subadult. Proportions of a typical male and female of this race follow: BNHS 1329 F — 19.5 CL 14.2 CW 18.9 PL 8.9 H weight 0.86 kg. FMNH 224141 M — 9.7 CL 7.5 CW 9.2 PL 4.6 H weighing 0.105 kg. In addition to size, males differ from females by having a longer (preanal > postanal) and thicker preanal portion of the tail with a vent opening beyond the edge of the carapace. Natural History. Indian tent terrapins occur in small to large rivers of peninsular India. They are frequently seen on logs and rocks basking. Females appear completely herbi- vorous. They never entered hoop traps baited with chicken entrails or fish and other than a small feather, the guts of four individuals from the Mahanadi River contained only leaves and stems of vascular plants. Males and juveniles appear more omnivorous. They did enter bait- ed hoop traps and the stomach of one male examined contained 75 percent vegetation and a fresh water prawn. Ovaries of two females collected on the Mahanadi River in mid- February appeared post-reproductive having few enlarged follicles and several small, old corpora lutea. Distribution : Kachuga t. tentoria ranges from at least the Mahanadi River drainage southward to the Krishna drainage. The turtle was taken at the following localities on the survey: FMNH 224163 — Godavari River, Manthani, Karimnagar Dist., A.P. Live Juvenile — Godavari River, Polavaram, West Godavari Dist., A.P. FMNH 224141, BNHS 1329 — Mahanadi River, Tikarpura, Dhenkanal, Orissa. The following additional records have been verified from preserved collections: ZSI (Type Kachuga t. intermedia Blanford 1870) — Hasdo River, Bilaspur Dist., M.P. ZSI 17775 — Seonath River, Bilaspur, Bilaspur Dist., M.P. ZSI 16767 & 68 — Mahanadi River, Cuttack, Cuttack Dist., Orissa. Kachuga tentoria flaviventer (Gunther 1864) Plain-bellied Tent Terrapin — Plate II, C & D Identification: A small tent terrapin (20.3 cm CL) with reduced pigmentation, an un- patterned plastron and little or no striping on neck and rump. Description : Sexes colored similarly. Male (FMNH 224178) — carapace brownish olive; light middorsal stripe with pale orange wash on Vertebrals 1 and 2, cream on V3, becoming faint on V4 and 5; pleuro-marginal juncture and border of shell also cream; plastron, bridge and underside of marginals cream and unpatterned; head pale, mottled with brownish olive; a near colorless patch washed with pale salmon extending from behind eye back over mastication musculature to meet patch from opposite side; immediately posterior a dark horizontal line marks juncture of head and neck; neck colorless to pale cream; a small irregular splotch of cinnamon rufous occurs in colorless area immediately behind eye; iris light smoke gray; limbs almost colorless ex- cept for a scattering of dark pigment along leading face; webbing and underside of feet creamy. Female (FMNH 224132) colored as above with the following exceptions: carapace light buff or cinnamon ground color; the central stripe being a darker, tawny coloration with a lighter center; plastron unpatterned but dark blotches present on underside of marginals; head smoke gray dorsally, cream laterally and on mandibles; a light cinnamon -brown spot 19 JOURNAL, BOMBAY NATURAL HIST. SOCIETY, Vo 1. 84 behind eye and another at posterior dorsal edge of eye; three additional spots at posterior of head (one medial flanked by two lateral); webbing of feet pale yellow. Size and Sexual Dimorphism : Sexes widely disparate in size. A female and two shells pre- sumed to be female from Kahalgaon measured 16.5, 16.9 and 18.8 CL. Five males from the Gandak River in northwestern Bihar showing well developed secondary sex characters rang- ed from 6.6 to 8.8 (mean 8.0) CL. Measure- ments of two typical specimens are: FMNH 224132 F — 16.5 CL 12.4 CW 16.1 PL 7.9 H weight 0.64 kg. FMNH 224178 M — 8.2 CL 6.3 CW 7.7 PL 4 . 1 H weighing 0 . 09 kg. In addition to size males differ from females by having a longer tail (preanal > postanal portion) which is relatively thicker at the base. Hatchlings’. Vijaya (1982e) provided mean measurements for hatchlings from six clutches of the pale-bellied tent terrapin (see com- ments under geographic variation) laid in nests along the Rapti River near Gorakhpur, U.P. Largest and smallest of these means were: Clutch I — 2.7 CL 1.7 CW 2.3 PL 1.6 H and 6.5 g weight. Clutch V — 3.35 CL 2.7 CW 3.1 PL 1.8 H and 10 g weight. Shell coloration — carapace mottled light and dark olive with a light yellow stripe along the middorsal keel; pleuro-marginal ring light geranium pink at hatching fading in the first few months to a pale olive; plastron buff yellow and unmarked; underside of marginals with a pepper like dusting of dark pigment; head ground color pale olive; two flesh pink spots located posterior to eye and at posterior of eyelid respectively and a narrow, flesh pink bar, broken in the middle, at back of head; vague striping discernible on neck and rump. Twelve hatchlings obtained from three clutches of eggs collected at the Katarniaghat Gharial Sanctuary in U.P. had the following mean dimensions: 3.55 CL 3.23 CW 3.2 PL 2.0 H and 9.0 g weight. Their shells were slightly serrate posteriorly with carapacial spines being very small except for a prominent projection on V3. VI was broadest anteriorly lacking the pinched bell-shape of large adults. None had a plastral pattern but varied as to the presence of dark pigment on the under- side of the marginals. Amounts varied from none to having a dark blotch on each scute. Another variable feature was the amount of red in the pattern. Some had an extensive amount including a geranium pink pleuro- marginal ring, middorsal stripe and head pattern. Concerning the latter at one extreme some individuals had both a transverse pink line across the back of the head, two pink spots at the snout and a postocular, pinkish crescent behind the eye. The crescent (similar to that of K. tecta but less extensive) was formed by an elongation of the postocular spot to meet the elongated one at the rear of the eyelid (this was evident in other indivi- duals in which the merger was incomplete). At the other extreme were individuals with no pink whatsoever; the pink areas of other individuals were colorless. Natural History : Like the other races of this species the pale-bellied tent terrapin ap- pears to be chiefly a river turtle. We found this race in Katarniaghat Gharial Sanctuary, an impoundment of the Ghagra River, but otherwise all were seen or collected in areas of flowing water. Many were observed bask- ing near the bank in a sandy area of the Gandak River in May. Gut contents of one male and one female examined contained only leaves and stems of aquatic vegetation. How- ever, five males were taken in traps baited with chicken entrails and fish suggesting that they may be somewhat omnivorous. Vijaya (1982e) found nests of this turtle 20 FRESHWATER TURTLES OF INDIA in “soft, clayey river bank soil” along the Rapti River near Gorakhpur from 6-8 Dec- ember. Nests varied from 15 to 26 cm in depth and were located 3 to 14 metres from the water. Six nests contained from 4 to 8 (mean 6) eggs per nest. She provided mean egg sizes for each nest ranging from 41 x 29 mm and 1 1 . 7 g to 45 x 27 mm and 18.5 g. Incubation times (time to emergence from substrate?) in artificial nests ranged from 125 to 134 days at nest temperatures varying between 27° and 28°C. On this survey we found three clutches comprising 6, 7, and 10 eggs in nests made by this turtle in sand banks along the Ghagra River in the Katarniaghat Gharial Sanctuary on December 5. Mean size of the eggs was 42 x 25 mm and 15.3 g weight. Mean in- cubation time for 12 hatchlings (to emergence from egg) was 95 days from eggs kept on moist cotton in plastic boxes at ambient tem- peratures ranging from 24° to 33 °C. Distribution : The pale-bellied tent terrapin inhabits the northern tributaries of the Ganges and possibly the Ganges proper from Bihar eastward. So far I have examined no speci- mens from West Bengal or Bangladesh. Speci- mens collected on the survey are from the following localities. FMNH 224142 — Katarniaghat Gharial Sanctuary, Ghagra River, Bahraich Dist., U.P. Hatchlings (Vijaya 1982e) — Rapti River, nr. Gorakhpur, Gorakhpur Dist., U.P. BNHS 1339 & FMNH 224178 — Gandak River, Bherihari Wildlife Sanctuary, Bettiah (West Cham- naran) Dist., Bihar. FMNH 224132 — Confluence of Kosi and Ganges River, Khalgaon. c. 50 km W Sahibgani, Bhagal- pur Dist., Bihar. Remarks : Gunther (1864) recognized this taxon as distinct and described it as a new species Pangs hura flaviv enter. The description was based on a single specimen presumed to be from India but lacking in precise locality data. It had been collected by a Mr. Mc- Clelland who had also sent several other speci- mens of Bengal species. Subsequent authors (e.g. Boulenger 1889, Smith 1931) considered it a variant of Kachuga tecta. Now that more specimens have been found, it is evident that this is a legitimate taxon. However, it appears to be a subspecies of Kachuga tentoria rather than a separate species. Evidence for this comes from the aforementioned hatchlings examined from the Rapti and Ghagra Rivers. Characteristics of these specimens particularly those from the Ghagra River appear to be intergrading with those of K. tentoria circum- data. The presence of circumdata characters (i.e. the pink pleuro-marginal ring and pink head markings) on some but not others of this group is typical of an intergrading popula- tion. The subject requires more study but for now I believe a subspecies designation best fits the evidence. Kachuga tentoria circimndata (Mertens 1969) Pink-ringed Tent Terrapin — Plate II, E+F Identification : A moderate-sized tent terra- pin (to 27.1 CL) having a single large dark blotch on all plastral scutes, a reddish ring at pleuro-marginal juncture of carapace, and reddish head markings. Description : Males appear somewhat darker than females otherwise sexes colored similarly (FMNH 224162 M) — carapace olive-green with a geranium pink ring around pleuro- marginal juncture; middorsal stripe comprised of streaks of geranium pink bordered by black; plastron straw yellow with large dark blotch covering over half of each scute; bridge and underside of marginals black bordered in yellow; inguinal with black spot but not axillary; seams of plastral scutes and margi- nals washed with pink; head olive green with a geranium pink circular spot behind eye and a pair of short, oblique, geranium pink bars at posterior of head; iris smoke gray; mandibles 21 JOURNAL, BOMBAY NATURAL HIST. SOCIETY, Vol. 84 light straw yellow with orange wash becoming olive near snout; neck olive gray with dull, cream colored stripes on sides and venter; rump and base of tail also striped (stripes more pronounced than in other races). A female (BNHS 1340) was similar but differed as follows: Carapace antique brown with a cinnamon-rufous rather than pink pleuro -marginal ring; middorsal stripe vague, almost nonexistent; head markings less red and more cinnamon rufous; two irregular spots (instead of bars) present at back of head. Size and Sexual Dimorphism : Females great- ly exceed males in size. A collection of thirteen females from the Chambal River in the Morena District of M.P. ranged from 18.1 to 27.1 (mean 22.1) cm CL. A sample of 11 males from the Yamuna River in the Etawah District of U.P. ranged from 7.7 to 8.8 (mean 8.3) cm CL. Measurements of a typical male and female are: FMNH 224162 M — 8.4 CL 6.7 CW 7.8 PL 4.6 Hand 0.0825 kg. Live F — 23.3 CL 17.2 CW 22.5 PL 11.2 H and 1.6 kg. In addition to size males differ from females by having a longer tail with a proportionately thicker base. Hatchlings : A single hatchling from an egg obtained at Deogarth Ghat, on the Chambal River measured 3.7 CL 3.5 CW 3.4 PL 2.1 H and weighed 11.5 g. Coloration — carapace smoke gray with vague dark mottling on most scutes with a cinnamon-rufous pleuro-marginal ring; anterior portion of middorsal stripe also cinnamon-rufous bordered with black; plastron straw yellow with large, black blotches cover- ing most of the scutes and light cinnamon- rufous along scute seams; head olive with bright geranium pink markings including a spot at posterior edge of upper eyelid and adjacent skin, a larger postocular spot and a transverse bar across back of head; chin with three additional pink spots along lateral edge of mandible; iris light gray; neck with cream stripes on an olive gray background; limbs gray with cream-colored scutes, webbing and lateral skin flaps; rump with dark and cream stripes. See Moll (1985) for a colored photo- graph. Natural History : The pink-ringed tent ter- rapin is a riverine form which readily basks at any opportunity. We observed hundreds of these turtles basking on logs, islands and along the banks of the Yamuna and Chambal Rivers. In January when the larger Kachuga were scarce and presumably dormant, this species was still basking in some number along the Chambal. There is some indication of habi- tat separation between the sexes and age groups — small males and juveniles were rare- ly seen in the river proper where females were common. However, in a backwater behind a sandbar projecting out from the shore of the Yamuna River, we once collected 11 males, 6 juveniles and 3 females. Three yearlings were also captured in a small (10' wide) weedy tributary of the Chambal hiding in and around vegetation masses. Digestive tracts of one male, one female and two immature females were examined. The male contained a beetle and an equiva- lent amount of aquatic vegetation whereas the female guts were packed with vegetation alone. This supports findings from the other races indicating that females tend to be chiefly herbi- vorous while males are more omnivorous. Nesting was occurring along the sandbanks of the Chambal River when we arrived there on 19 January but all nests found had been destroyed by predators. Jackal tracks and occasionally hyena tracks were associated with these nests. Two eggs, one cracked and one entire remained in one of the open nests. The cracked egg measured 47 x 28 mm while the FRESHWATER TURTLES OF INDIA entire egg was 47 x 27 mm and weighed 19.5 g. Rao and Singh (1985) reported that nesting occurs from October through January in the National Chambal River Gharial Sanctuary. They calculated the turtles’ average clutch size as 6 (3-12) eggs and thought two clutches to be typical. Mean egg size was 48.6 x 27.6 mm and 21 g. Using data from eight of their females, I caculated the mean female repro- ductive effort as: RCM — 0.08, ELI — 22, EWI — 12, and EMI — 1.4. Distribution : The pink-ringed tent terrapin occurs in the upper and central Ganges and such tributaries as the Hindon, Yamuna, Chambal and Gomati. Mertens (1969) named the type locality for this race as Meerut, Meerut District, U.P. He was not confident about the localities of his paratypes which were listed as Calcutta or vicinity of Calcutta. Based on our observations from this survey, it seems likely that Merten’s paratypes were market specimens shipped in from more western states. Because of the heavy market trade in turtles of West Bengal, distribution records from this state are often unreliable. More likely the race of tent turtles which occurs naturally in West Bengal is K. t. flavi- venter or some as yet undescribed popula- tion. Specimens of this race were collected at the following localities on our survey: FMNH 224105 & 224109 — Hindon River, nr. Ghaziabad, Meerut Dist., U.P. BNHS 1335 & FMNH 224162 — Yamuna River, 5 km S Etawah, Etawah Dist., U.P. Refe Acharji, M. N. (1950): Edible chelonians and their products. J. Bombay nat Hist. Soc. 49: 529- 532. (1955): A collection of chelo- nians and snakes from Chota Nagpur, Bihar. Rec. Indian Mus. 53: 383-392. Anderson, J. (1876): On the cloacal bladders and on the peritoneal canals in Chelonia. J. Linn. Soc. London. 12: 434-444. BNHS 1 340 — Chambal River, Deogarth, 30 km NE Morena, Morena Dist., M.P. FMNH 224185 — Gomati River, nr. Lucknow, Lucknow Dist., U.P. Acknowledgements Many officials and individuals aided in the surveys from which these observations were made. I am particularly indepted to Shri Samar Singh, Joint Secretary for Wildlife, and his staff for providing introductions and for help- ing me to obtain the necessary permits in the states visited. My colleagues and companions on the survey, Ms J. Vijaya and Satish Bhaskar provided invaluable assistance on all aspects of the study. Mr. J. C. Daniel, Shri P. Kannan and Shri S. Biswas provided consi- derable useful advice based on their extensive knowledge of the Indian chelonian fauna. My thanks to Joseph Mitchell who kindly pro- vided information on the Kachuga smithii specimen used as a paratype in this paper and to the U.S. National Museum for loaning me the specimen. A special thanks to Rom and Zai Whitaker who provided living and labo- ratory space for me at the Madras Crocodile Bank. The project was funded through an Indo- American Fulbright Fellowship and by a grant from the New York Zoological Society. Bom- bay Natural History Society kindly served as my host institution in India. EN CES (1878-79): Anatomical and zoolo- gical researches and zoological results of the Yunnan Expeditions. Reptilia and Amphibia. Calcutta. Boulenger. G. A. (1889): Catalogue of the chelonians, rhvnchocephalians and crocodiles in the British Museum (Natural History). Taylor and Francis, London. Carr, J. L. (1981): Phylogenetic implications of chromosomal variation in the Batagurinae (Testu- 23 JOURNAL, BOMBAY NATURAL HIST. SOCIETY, Vol 84 dines: Emydidae). M. S. Thesis, Texas A&M Uni- versity. Chaudhuri, B. L. (1912): Aquatic tortoises of the middle Ganges and Brahmaputra. Rec. Indian Mus. 7: 212-214. Das, I. (1985) : Indian turtles a field guide. World Wildlife Fund India (Eastern Region), Calcutta. Ewert, M. A. (1979): The embryo and its egg: Development and natural history. Pages 333-413 in M. Harless and H. Morlock eds. Turtles: Perspec- tives and research. John Wiley and Sons, New York. Gaffney, E. S. (1984) : Historical analysis of theories of chelonian relationship. Syst. Zool. 33 : 283-301. Gray, J. E. (1831a): Illustrations of Indian Zoo- logy. Vol. 1, PI. 71. London. (1931b): Synopsis Reptilium or short descriptions of the species of reptiles. Part 1. Cata- phracta, tortoises, crocodiles, and enaliosaurians. Treuttel, Wurz Co., London. (1834) : Characters of several new species of freshwater tortoises (Emys) from India and China. Proc. Zool. Soc. London. 1834 : 53-54. (1855) : Catalogue of shield reptiles in the collection of the British Museum. Part I. Testudinata (tortoises). Taylor and Francis, London. (1862): Notice of two new species of Batagur in the collection of the British Museum. Proc. Zool. Soc. London. 1862: 264 and 265. (1863) : Notice of a new species of Batagur from northwestern India, ibid. 1863: 253. — — - (1869) : Notes on the families and genera of tortoises (Testudinata), and on the charac- ters afforded by the study of their skulls, ibid 1869: 165-225. Groombridge, B., Moll, E. O. & Vijaya, J. (1983): Rediscovery of a rare Indian turtle. Oryx. 17: 130- 134. Gunther, A. (1864): The reptiles of British India. London. Hirayama, R. (1984): Cladistic analysis of bata- gurine turtles (Batagurinae : Emydidae: Testudinoi- dea); A preliminary result. Pages 141-157 in F. de Broin and E. Jimenez-Fuentes, eds. Stvdia Geologica Salmanticensia, vol. especial 1 (Stvdia Palaeochelo- niologica). Editiones Universidad de Salamanca, Salamanca, Spain. Hora, S. L. (1948): The distribution of croco- diles and chelonians in Ceylon, India, Burma and farther East. Nat. Inst. Sci. India. 24(6) : 285-310. Iverson, J. (1985) : Checklist of the turtles of the world with English common names. Soc. Stud. Amphib. Rept. Herpetol. Circ. 14. Jayaram, K. C. (1974): Ecology and distribution of freshwater fishes, Amphibia and reptiles. Pages 517-584 in M. S. Mani, ed. Ecology and biogeo- graphy in India Dr. W. Junk, The Hague. Jerdon, T. C. (1870) : Notes on Indian herpe- tology. Proc. Asiatic Soc. Bengal. 1870: 66-85. Khan, M. A. R. (1982): Chelonians of Bangla- desh and their conservation. J. Bombay nat. Hist. Soc. 79: 110-116. Legler, J. M. (1960) : A simple and inexpensive device for trapping aquatic turtles. Utah Acad. Sci. Proc. 37: 63-66. Loveridge, A. & Williams, E. E. (1957): Revi- sion of the African tortoises and turtles of the suborder Crvptodira. Bull. Mus. of Comp. Zool. 115(6): 163-557. McDowell, S. B. (1964): Partition of the Genus Clemmys and related problems in the taxonomy of the aquatic testudinidae. Proc. zool. Soc. Land. 143(2): 239-279. Mertens, R. ( i 969) : Eine neue Rasse der Dachs- childkrote, Kachuga tecta. Senckenbergiana biol. 50 ( 1/2): 23-30. Minton, S. A. jr. (1966): A contribution to the herpetology of West Pakistan. Bull. American Mus. Nat. Hist. 134: 27-184. Moll, E. O. (1983) : Turtle survey update (Janu- ary-April, 1983). Hamadryad 8(2): 15-17. — — (1984) : Freshwater turtles in India: their status, conservation and management. Hama- dryad 9(3) : 9-17. (1985): Freshwater turtles in India. Sanctuary. 5(1) : 49-59, 66. (in Press a) : India’s freshwater turtle resource with recommendations for manage- ment. Bombay Natural History Society Centenary Symposium. Bombay. (in Press b) : Estuarine turtles of tropica] Asia: Status and management. ,Symposium on endangered marine animals and marine parks (1985). Marine Biological Association of India, Cochin. & Vijaya, J. (1986) : Distri- butional records for some Indian turtles. J. Bombay nat. Hist. Soc. 83 ( 1) : 57-62. Pritchard, P. C. H. (1979): Encyclopedia of turtles. T. F. H. Publ. Inc., Neptune, New Jersey. Rao, R. J. & Singh, L. A. K. (1984) : Ecological relationship among turtles in National Chambal Sanctuary. Interim study report 1. National Chambal Sanctuary, Deori. Morena, M.P. 24 FRESHWATER TURTLES OF INDIA (1985) : Ecological relationship among turtles in National Chambal Sanctuary. Interim study report 2. National Chambal Sanctuary, Deori, Morena, M.P. Sites, J. W. jr, et al. (1984) : Biochemical charac- ters and the reconstruction of turtle phylogenies : Relationships among batagurine genera. Syst. Zool. 33: 137-158. Smith, M. A. (1931) : The fauna of British India, including Ceylon and Burma. Reptilia and Amphibia, vol. 1, Loricata and Testudines. Taylor and Francis, London. Smithe, F. B. (1975) : Naturalist’s color guide. The American Museum of Natural History, New York. Theobald, W. (1868): Catalogue of the reptiles of British Burma, embracing the provices of Pegu, Martaban, and Tenasserim; with descriptions of new or little-known species. /. Linn. Soc., Zool. 1868 : 4-67. Tinkle, D. (1962) : Variation in the shell mor- phology of North American Turtles. I. The cara- pacial seam arrangements. Tulane Stud. Zool. 33 1 - 349. Vijaya, J. (1982a) : The flapshell turtles. Horn- bill, 1982(3 ): 12-14. (1982b) : Rediscovery of a rare turtle from Kerala forest. Hombill, 1982(4 ): 25-26. (1982c): Flapshell nests at Taramani, Madras. Hamadryad 7(1): 12-13. (1982d) : Rediscovery of the forest cane turtle of Kerala, ibid. 7(3) : 2-3. — (1982e): Kachuga tecta hatching at the Snake Park. ibid. 7(3) : 14-15. ( 1 982f ) : Breeding data on Melano- chelys trijuga trijuga and Melanochelys trijuga coro- nata. ibid. 7(3) : 16. (1982g): Pelochelys bibroni from the Gahirmata coast, Orissa, ibid. 7(3) : 17. — (198211) : Focus on conservation of freshwater chelonians in India. Tiger Paper. 9(3) : 11-13. (1982i) : Successful artificial breeding Lissemys punctata granosa. J. Bombay nat. Hist. Soc. 79: 210-211. (1982j) : Rediscovery of the forest cane turtle, Heosemys ( Geoemyda ) silvatica (Repti- lia, Testudinata, Emydidae) from Chalakudy forests in Kerala. J. Bombay nat. Hist. Soc. 79: 676-677. (1983a) : Freshwater turtles in U.P./ Bihar. Hamadryad 5(1): 12-13. (1983b) : Freshwater turtle survey in India 1982-83. ibid. 5(1): 21-23. (1983c) : Auditory cues as possible stimuli for hatching eggs of the flap-shell turtle, Lissemys punctata granosa. ibid. 5(1) : 23. (1983d) : Range extension for the spotted pond turtle. Geoclemys hamiltoni. ibid. 5 (2) : 20. (1983e) : The Travancore tortoise, Geochelone travancorica. ibid. 5(3): 11-13. Vitt, L. & Price, H. (1982) : Ecological and evolutionary determinants of relative clutch mass in lizards. Herpetologica. 38 : 237-255. Whitaker, R. (1982): Bangladesh: A general survey. Hamadryad 7(3) : 4-9. Yadava, M. R. (1980) : Hatching time for the eggs of softshell turtle, Kachuga dhongoka (Gray) at various temperatures. Indian Forester. 1061: 721- 724. Yadava, M. R. & Prasad, B. (1982a): Observa- tions on the breeding biology of Indian tropical pond turtle, Lissemys punctata granosa (Schoepff ) . Indian J. Zoot. 23: 51-56. (1982b) : Morphometric studies on the late limb period stages in embryonic development of the turtle, Kachuga dhongoka (Gray). Indian J. Zoot. 23: 57-62. Zangerl (1969): Zangerl, R. (1969) : The turtle shell. Pages 311- 339. In: C. Gans, A. d’Bellairs and T. Parsons, eds. Biology of the Reptilia, Vol. I, Morphology A. Aca- demic Press, London. 25 THE BUTTERFLIES OF THE NILGIRI MOUNTAINS OF SOUTHERN INDIA (LEPIDOPTERA: RHOPALOCERA ) 1 Torben B. Larsen2 This paper is an account of the three hundred or so butterflies to be found in the Nilgiri Mountains of southern India. Special emphasis is placed on placing the butterflies in perspective in relation to a number of defined ecological zones. Follow- ing a description on the climate, topography and vegetation zones in the Nilgiris is a section on the history of entomological exploration. The major part of the paper is devoted to a summary account of each species known to occur in the Nilgiris. Every attempt has been made to update the nomenclature which is cross-referenced to that of Wynter-Blyth (1957) whose nomenclature actually dates back to the 1930ies. Since this nomenclature is the same for most of peninsular India and since the Nilgiris contain populations of virtually all species known to occur in southern India, the paper should also be useful out of the specific Nilgiri context. The data included will be the basis for a later analysis of the ecological and zoogeographical nature of the Nilgiri butterflies, but some discussed. Introduction The Nilgiri Mountains The Nilgiri Mountains are situated in South India with their centre at 11°25'N and 76°45' E. The name means Blue Mountains and was bestowed upon them by the plains people at least 700 years ago when the Nilgiris were only seen rising in the distance from the steaming and insalubrious jungles that sur- rounded them. The Nilgiris are a well-defined massif that forms the southern limit of the main Western Ghats system that stretches unbroken from Bombay in the north. To the immediate south of the Nilgiris is the Palghat Gap, a stretch of dry lowlands separating the South Sahyadri from the main Ghats system. The distance is not great and on a good day the nearest of the 1 Accepted January 1987. 2 Snoghoj alle 29C, 2770 Kastrup, Denmark. initial non-quantitative conclusions are South Sahyadri ranges, the Annamalais, is clearly visible from the Nilgiri plateau. To the northeast the precipitous Moyar Gorge creates a narrow boundary between the Nilgiris and the Biligiriranga Mountains. The latter can be looked at as the southwestern link, though at best a tenuous one, with the Eastern Ghats system. The Nilgiris may be described as a right- angled triangle with the right angle placed in the northwestern extremity. The western slopes of the main plateau rise abrupt- ly from levels of 100 to 300 m through a steep escarpment to 1800 m or so. At this level the Nilgiri Plateau commences. Though less precipitous the same situation pertains to the southern slopes. The northern slopes rise from the Mysore Plateau from levels of 700-900 m in a less precipitous manner still, and they are hence rather less imposing. However, the Nilgiris most certainly are a sharply defined geographical feature. 26 BUTTERFLIES OF THE NILGIRI MOUNTAINS The plateau of the Nilgiris is hardly a plateau in the strict sense of the word. Valleys drop to 1600 m and peaks rise to more than 2500 m, but most of the area is an undulating landscape confined within the 1800 m contour of the scarp and technically the expression plateau is correct (though slightly disconcert- ing to an author from a country in which the highest point is some 180 m and lovingly known as the Mountain of Heaven). Just to the north of the Nilgiris proper lie the Ouchterlony Valley and the Wynaad with an average elevation of 1200-1300 m, provid- ing level country at an altitude not found elsewhere in the range. They form an integral part of the Nilgiris as well as the link with the Western Ghats proper and will be includ- ed in the systematic part of the paper. When making ones way through the Nilgiris, whether by car or by foot, it is easy to forget how small a geographical feature they actually are. The western slopes extend some 40 km due south, the southern and the northern slopes are both some 60 km. As will be seen there can be few areas of similar size with as much ecological variation within its boun- daries. The area covered by this paper is the Nilgiris as a geographical feature, from the surround- ing lowlands to the highest peaks. Most of this falls into the administrative Nilgiri District (PIN 643 000) of Tamil Nadu State, but much of the lower parts of the western slopes fall into the Mallapuram and Palghat districts of Kerala State, while portions of the southern slopes fall into Coimbatore District of Tamil Nadu. The northern borders of the area covered are clearly defined by the Moyar river, and most of the southern border is defined by the Bhavani river. Elsewhere the limit is simply taken as the foot of the mountains and the immediate surrounding plains. Climate Situated at 11° north the Nilgiris are well into the tropical zone with the result that tem- perature variations during the year are rela- tively modest. Mean monthly averages between the coldest month (December or January) are normally no more than 5°C below that of the warmest summer month (usually May). How- ever, the altitudinal temperature differences are highly significant. The annual mean tem- perature is 28° at the foot of the southern slopes and only about 15° in Ooty at 2200 m. These altitudinal temperature differences have profound ecological effects. Average monthly temperatures for selected localities are given in table 1 below. Rainfall patterns are dominated by the mon- soon regime that affects all of India, but diffe- rent parts of the Nilgiris are affected in different ways. The SW monsoon normally commences in the latter part of May or in early June, continuing till some time in Sept- ember with the occasional pauses. This is India’s life sustaining main monsoon, but the positioning and topography of the Nilgiris are such that the main effects of the SW monsoon are felt on the western and northwestern slopes. The southern and northeastern slopes are partly in a rain-shadow and receive as much of their total rainfall from the retreating NE monsoon as they do from the main mon- soon. The plateau receives rain in an inter- mittent fashion from both the monsoons and from more localised thunder showers at other times of the year. The first four months of the year (January to April, and much of May) are everywhere rather dry, though the southern slopes get some rain even then. Interestingly this dry season is most marked on those parts of the western slopes that otherwise receive more precipitation than any other areas of the Nilgiris. But for the very pronounced drought 27 JOURNAL, BOMBAY NATURAL HIST. SOCIETY, Vol. 84 Table 1 Average MONTHLY TEMPERATURE IN Centigrade for selected Nilgiri localities Month Bhavani S. 290 m Kallar 457 m Silent Valley 914 m Coonoor 1747 m Ooty Bot. 2225 m JAN. 25.3 23.9 17.8 13.7 13.1 FEB. 29.9 25.8 19.5 15.2 13.8 MAR. 29.4 28.3 22.3 17.0 15.2 APR. 31.6 28.9 23.6 18.7 16,4 MAY 31.5 29.0 23.5 19.6 16,8 JUN. 31.0 27.7 20.7 18.8 15.4 JUL. 29.9 26.8 18.8 18.2 14.4 AUG. 30.0 27.0 19.2 18.1 15.2 SEP. 29.9 26.9 19.8 17.6 15.4 OCT. 28.1 26.4 20.1 17.1 15.1 NOV. 26.5 25.0 19.3 15.5 14.1 DEC. 25.0 23.5 18.3 14.1 13.3 ALL YEAR 28.8 26.6 20.2 16.9 14.9 Source: von Lengerke (1977) the wetter parts of the Nilgiris would doubt- less be even more rich in flora and fauna than they already are. I suspect that the exceptional richness of the flora and fauna at Kallar, a locality with a relatively modest 1500 mm of rain a year, is due to the fact that it is ex- ceptionally well distributed in time, avoiding periods of pronounced drought. Some illustra- tive examples of rainfall patterns are given in table 2 below. The volume of rainfall varies by a factor of more than twelve in the Nilgiris. Both low- land and highland areas of the western slopes that are fully exposed to the SW monsoon receive more than 6000 mm a year, and more than 3500 mm is normal in this area. The driest parts of the Moyar Gorge, an area in permanent rain shadow, receive less than 500 mm. Some examples are given in table 3 be- low, compressing data from table 2 above. Ecology Rainfall patterns, temperatures and to a lesser extent topography combine to produce in the small area covered by the Nilgiris a network of very complex ecological conditions with a degree of variation that is not usually found in so small an area. Given the statistics this is not surprising: Elevation ranges from 100 m to well over 2500 m; rainfall from less than 500 mm a year to more than 6000 mm, perhaps in some cases much more; the lowest temperature ever recorded at the foot of Nadgani Ghat is about 18°, while frost is a regular feature at Ooty and elsewhere on the plateau. While broad ecological zones of considera- ble diversity can be identified the net result is not always a very tidy pattern because of topographical quirks, rain-shadows and local peculiarities. To give one example. The 28 Annual rainfall in selected Nilgiri localities (by month and year total in BUTTERFLIES OF THE NILGIRI MOUNTAINS P , a P a a a a rj C d § p § g a a a o < © co ON O' On On 8 $ O' m ON z P < H co T— H 04 04 p- CO VO 04 r— 1 Q O' ON On o 04 X O' w co 04 t-H vo >n o p- Q i— i > co O O' in oo O Z 00 On CO »o P" oo v-H 04 04 H On vn 04 VO O' o O O O On O ON ■vi- i— i o- co 04 CO 04 pj cO VO VO 00 CO ON X P oo 00 On ON o O 04 X vn d On VO 00 >o p- On X |d »n P* o CO co on vo < oi CO 04 On O O' t p- CO O' oo vo O' On o O' P O) in 04 i—i 1—1 H- i 04 00 vo VO CO CO oo 00 z »n ON T— | 04 O' CO p cO P" 04 T-1 l—s CO o 04 On vo p- <3 i— i VO 04 O' VO o- ON S co r-H r— H -A VO 04 oo VO o- X 8 Q. o- 04 VO O' O' o 53 gi o- »o 04 ON CO o- < 04 04 CO m s 05 rT*l o~ O 1—1 On CO O' m a r— H vo z •— i 04 00 X 00 X < T— 1 t 04 1— 1 p- X •—> O oo 3 s 5h i' tn P a 3 *4-1 X P3 a -o p a oi X O a *»H ’Sib a T3 p p d a ’p 5-i o M P in o co E? CO >1 ON a 04 ’id p o Q P p- o p- p O' o 04 04 o oo o in P P 3 >n 04 £ T-^ O P ON p On p a a i a a a a i a a oo cO vn O' oo 00 O' 04 r- o X CO p- 00 o- CO ON CO in X o- vn p- 04 o VO co 04 CO CO in 04 o 04 in CO O' m o p- X oo oo CO X 00 00 00 oo 00 p- CO 04 04 1—1 04 CO 1—1 00 p- O co 00 On ON vn O' CO CO p- co p- X o 04 8 00 04 oo p- p- oO CO X O' p- vn O oo X o X CO CO On 00 o vn in O 8 vn X O' co 04 p- oo m o- CO ON ON 04 On 04 o OO OO O- oO O 1—1 i—< 04 i—i 136 73 48 S 70 p- cO CO CO O' 00 O' CO vn *— < p- o 04 vn O' 04 o CO CO T— 1 T-H P- X O' O' m c4 O- 04 29 Source: von Lengerke (1977). JOURNAL, BOMBAY NATURAL HIST. SOCIETY, Vot. 84 Nadgani site where many of the species re- corded in the systematic part were observed probably had a rainfall of about 6000 mm a year, but just two kilometres further west, where I usually camped, rainfall was down to 3000 mm. On many occasions I was repeated- ly drenched during a day’s collecting only to come back and find that there had been no rain at the camp site. Table 3 Total rainfall in four selected Nilgiri localities AND ITS DISTRIBUTION BETWEEN SW MONSOON MONTHS, NE MONSOON MONTHS AND THE DRY SEASON IN PERCENT Season Western Southern Northern Plateau Silent Valley Kallar Moyar Kotagiri 914 m 457 m 880 m 1982 m SW Monsoon (May-Sep.) 83.5 26.1 46.5 56.2 NE Monsoon (Oct. -Dec.) 13.0 47.9 41.8 35.5 Dry Season (Jan.-Apr.) 3.5 26.0 11.7 8.3 Total 100 100 100 100 Rainfall mm 3968 1478 538 1264 Source: Calculated from table 2. Such differences make the development of a broad ecological framework difficult, though such a framework will be essential for a sub- sequent interpretation of the total Nilgiri butterfly fauna in ecological biogeographical terms. I have however, attempted to make such a framework based on the work of pre- vious authors, leaning heavily on a simplified version of the forest classification by Cham- pion & Seth (1968), as well as on my own observations and conclusions. Ecological framework for the Nilgiris For the purposes of analysing the butterfly fauna I have arrived at a classification frame- work covering twelve main ecological zones. The first, and rather crucial division is one based on altitude, the attendant climatic diffe- rences, and their joint effect on the floristic and faunistic composition. On this basis the Nilgiris may be divided into three major divi- sions as follows: tropical/lowlands 100-1500 m subtropical/middle levels 1500-1900 m MONTANE /plateau 1900 m+ The exact altitudes of change from one divi- sion to the next is subject to local variation and there is often a transition zone. Where the subtropical division merges directly with evergreen tropical forest it may be difficult to define when the transition between the two has actually been effected. The following paragraphs briefly characterise the salient features of each of the eleven zones, which are then again summarised in table 4. TROPICAL DIVISION 1. Rainforest zone. Where rainfall exceeds 3500 mm a year in the lower part of the tropical division, and if topography does not militate against it, the climax vegetation be- comes a fully developed rainforest system approaching that called giant evergreen by Champion & Seth (1968). There is a closed canopy and several storeys with an amazing faunal and floral diversity. Most of the Nadgani Ghat and some areas near Mukkali at the base of the Silent Valley access road belong to this zone. Were it not for the fact that the dry period from January to April was very severe the number of species would be even greater, but some species without diapause mechanisms cannot survive. A significant num- ber of the most interesting South Indian butterflies are effectively limited to this zone. 2. Wet Evergreen zone. This zone is charac- terised by a lower rainfall than the previous 30 BUTTERFLIES OF THE NILGIRI MOUNTAINS one, ranging from 1800 mm a year to 4000 mm at the higher levels of the tropical division. Visually it is less grandiose, the trees lower, and the typical storied structure of true rain- forest has become a confused tangle. It covers the West Coast Tropical Evergreen of Cham- pion & Seth (1968), and in the zone I also include the patches of evergreen found on the southern slopes which they would classify as Semi-Evergreen. From a butterfly point of view it is characterised by a great degree of species diversity, the absence, or near absence, of the true rainforest elements, and the fact that species from the more open formations still do not penetrate. Much of the forests of the western slopes fall into this category. I have not worked any spot in this zone con- sistently, but Kallar forms a transition between it and the mixed deciduous forest (see 4 below). 3. Wet Agricultural zone. Till a few cen- turies ago the entire western slopes and good parts of the southern slopes were covered with evergreen forests of the two types discussed above. Both are very fragile eco-systen^s that do not take kindly to human interference and where wood has been exploited virtual eco- logical deserts result, clad only in grasses and one or two dominant dicotyledons. Such areas are entomologically useless, but unfortunately they prevail over long stretches. Much of the area has been under agriculture for the past few centuries, with rice, bananas, turmeric, rubber, palm crops, pistachio and various fruits as the main crops. Plantation teak is the worst of all. Any butterflies that remain are found in fringing riverine forest and untidy gardens, but they are not many. A very few skippers thrive on the rice and some common butterflies of the more open formations manage to establish populations, but on the whole wet agricultural lands are disappointing for butter- flies, except that the occasional forest species maintains a surprising foot-hold here and there. 4. Mixed Deciduous Forest zone. Much of the lowland southern slopes and the central parts of the northern slopes consist of dense mixed deciduous forest with considerable floral and faunal diversity and a wide range of micro-habitats in response to climatic and topographical variation. Wherever permanent moisture is available there is an admixture of evergreen plants, occasionally leading to local almost evergreen patches. The rainfall regime is usually in the range of 1000 to 1800 mm. The physical aspect is a low, but very dense forest, at times almost impenetrable. At the upper levels it merges gradually with the sub- tropical evergreen forest on the southern slopes. The mixed deciduous forest is very different from the moist-deciduous forests of the Wynaad and the Mudumalai Wildlife Sanctuary, which are very open with tall teak as the dominant tree. Moist-deciduous forest is a relatively poor butterfly habitat which is weakly represented in the Nilgiris proper and not dealt with separately in this paper, though notes are given at relevant points in the syste- matic part. The whole of the Kotagiri Ghat below Kunjapannai is typical of the mixed deciduous forest zone. 5. Thorn Forest zone. Once rainfall drops to between 500 and 1000 mm the natural climax vegetation becomes some form of savanna-like forest which visually is not dis- similar to parts of East Africa and where, in fact, much of the dominant vegetation has African affinities, not least species of Acacia. This is also the home of one of the Indian antelopes, the Blackbuck. The floral and faunal composition is much less diverse and varied than the mixed deciduous forest. Most of the area between Masinagudi and Bhavanisagar via the Moyar Gorge forms part of this zone, but it is also represented at many places in 31 JOURNAL, BOMBAY NATURAL HIST. SOCIETY, Vol. 84 the foothills of the southern slopes, when rain- fall is not high enough to support mixed deciduous forest. It is also the home of the most typically African of the South Indian butterflies. 6. Dry Agricultural zone. Most of the land at the foot of the forested slopes now consists of non-irrigated farmlands, much of which would have been thorn forest, and in favoured localities mixed deciduous forest. Often rocky outcrops and natural fences are left untended. Such areas contain many of the species charac- teristic of thorn forest, but by and large it is a very poor butterfly habitat. Here and there irrigated agriculture prevails. Except where trees and palms are grown, such areas mainly contain the most common plains species and those that have adapted to man-made environ- ments. At Kallar, with its extensive planta- tions, some forest species, even those of the evergreen forests, survive. Most of these would probably be absent if Kallar were not imme- diately adjacent to undisturbed forest. SUBTROPICAL DIVISION 7. Subtropical Evergreen Forest zone. The exact altitudinal boundaries of the subtropical zone vary from 1200/1500 m to 1800/1900 m depending on local circumstances, and once these elevations are reached the extreme diffe- rences in annual precipitation that create such diversity in the tropical zone have largely disappeared. Consequently the range of vege- tation zones has also narrowed, and the pre- dominant climax vegetation of the western and the southern slopes is subtropical evergreen forest, except where the scarps are so steep that forest cannot cling to them. Where the subtropical forests abut tropical evergreen forests (e.g. at Silent Valley) the transition between the two is both visually and floristi- cally very gradual. Elsewhere the transition between mixed deciduous and subtropical ever- green is abrupt and very noticeable, such as on the Kotagiri Ghat. Climax subtropical ever- green forest is often almost impenetrable. As far as the butterflies are concerned this zone has a small, but conspicuous, special element. Most of my own experience of this zone is from the forests at Glenburn and Kunjapannai. 8. Sub-climax Subtropical Forest zone. In areas of low rainfall on the northern slopes and in disturbed and rocky areas of the south- ern slopes the subtropical evergreen forest cannot develop fully. The result is some im- poverished associations drawing from the evergreen and deciduous zones the most hardy species, interspersed with grasslands. These are very poor habitats with no major influence on the Nilgiri fauna. 9. Subtropical Agricultural zone. Most of the subtropical zone has been converted to plantation agriculture, with coffee at the lower reaches and tea at the upper. On the whole these are poor butterfly habitats, not least because weedicides and pesticides are applied liberally, frequently and consistently. Butter- flies are only found along streams which, at least in the coffee country, often has a belt of fringing, riverine vegetation. Tea country is a virtual ecological desert, only a handful of butterflies managing to survive. The grassland areas are indicators of recent forest destruction and are not natural habitats for any species. MONTANE DIVISION 10. Montane Evergreen Temperate Forest zone. The plateau of the Nilgiris, mainly above 1900 m, has two characteristic types of natural vegetation. The evergreen forests known as sholas, often growing only in shel- tered places, and rolling open grasslands with only a few Rhododendron trees. The sholas are dark, almost impenetrable forests, their edges often shaped by the prevailing winds. Typical sholas develop under rainfall regimes from as 32 BUTTERFLIES OF THE NILGIRI MOUNTAINS low as 1200 mm to as high as 6000 mm+ without displaying much difference in floral composition or physical aspect. The reason for this is that the sholcts in the most heavy rain- fall area have a rapid run-off of excess water. Patches of shola remain over most of the plateau, and extensive forests may still be found in the Kundahs, at Avalanche and in the Mukurti Peak area. One of the prettiest and most accessible is the Longwood Shola near Kotagiri, which has a most attractive bog at its centre. The floral composition is characterised by a large proportion of plants with Palaearctic and Oriental montane affi- nities. 11. Montane Grasslands zone. In between the sholas the normal vegetation consists of rolling grasslands, but this is now in evidence only in the remoter parts towards the western escarpment between the Sispara Pass and Mukurti Peak. Elsewhere they have been con- verted to tea. These grasslands are the home of the famed Nilgiri Tahr. Champion & Seth (1968) are insistent that the grasslands are derived and that the entire plateau was once clad in forest. The presence everywhere of the fire resistant Rhododendron is their main evidence. Ranganathan (1938) takes issue with this (referring to the first edition of Champion), pointing out that wind and frost would preclude shola vegetation in some parts of the plateau. Blaseo (1971) would appear to support the latter viewpoint. I do not have the expertise to settle the matter, but certainly the vegetation of the grasslands is both more complex and more varied than in the recent grasslands of Sumatra and Papua New Guinea which are definitely the result of human inter- vention. Grasslands were extensively present when the first explorers visited the Nilgiris just 150 years ago. It is difficult to see how a tiny handful of Toda pastoralists could, or for that matter should, have indulged in massive deforestation. 12. Montane Agricultural zone. Whether or not the montane grasslands are derived, it is a fact that more than three quarters of the plateau has been heavily modified by human intervention. Beginning in 1832 large areas have been forested with exotic trees, espe- cially of Australian origin ( Acacia dealbata, A . decurrens, A. melanoxylon and E. globu- lus). One of the latter is now at 78 m one of the tallest trees in all of India. Somewhat later a number of temperate conifers were also planted. Virtually any tree from anywhere that was of potential promise has been tried out in the Nilgiris and remnants are some- times found in surprising places. The amount of tree planting may well have exceeded the deforestation in some areas, but diversity was sacrificed for monoculture. In the Ooty area some homesick English gentlemen spent their summers strewing about the seeds of broom and gorse on the downs, with the result that these plants are locally dominant. Since their arrival some 400 years ago the local Badaga have concentrated on the cultivation of vege- tables with potatoes, carrots and cabbages as the main crops, though wild boar is a pere- nnial problem. However, in terms of both area and visual impression it is tea, tea and yet more tea that predominates as a never-ending ecological desert. Summary This concludes the brief overview of the Nilgiri ecology and outlines the twelve ecolo- gical zone which will later be used for an analysis of the ecology and zoogeography of the butterfly fauna. Table 4 below gives a brief summary of the information. As already mentioned a somewhat robust approach is called for when dealing with major ecolo- gical zones in a area where so much varia- tion can be found within a few kilometres. 33 3 JOURNAL, BOMBAY NATURAL HIST. SOCIETY, Vol 84 Only a robust approach can impose some element of order into what sometimes appears to be anarchy; it should not be used to obscure the fact that genuine anomalies exist and that surprises do occur. It is also necessary to highlight once again the fact that much of the area, especially that above 1200 m, has been the subject of very intensive human intervention for the past 150 years. It its good to know, though, that representative portions have been maintained for posterity. The proposed inclusion of the Nilgiris in the World Biosphere Reserves scheme will hope- fully ensure that the environment is gradually improved. Few places contain as much diver- sity within so little space as the Nilgiris. As an ecosystem it demands our respect. Table 4 Summary of the twelve major ecological zones of the Nilgiris Tropical Division (100-1500 m) 1 . rainforest zone Rainfall 3500 mm+; temperature range 22-30°. Tall closed canopy forest with little undergrowth except where bamboo prevails. Hopea, Dipterocar- pus, many Guttiferae, many Anacardiaceae, Sapota- ceae, Meliaceae, etc. Great floral and faunal diver- sity. Strong affinities to Sundaland. Only western slopes and not usually above 1000 m. 2. wet evergreen zone Rainfall 1800-3500 m; temperature range 18-28°. Dense closed forest, lower than previous zone, and often with well-developed undergrowth. Much local variation and transition to moist-deciduous or to semi-evergreen depending on specific circumstances. Most of western slopes and part of southern slopes. 3. wet agricultural zone Rainfall 2000-5000 mm; temperature range 18-30°. Main crops rice, coconut, coffee, cardamom, fruits and plantation trees. Mainly at the foot of the western slopes. Some of the slopes are grasslands of little diversity slowly regenerating to forest. 4. MIXED DECIDUOUS FOREST ZONE Rainfall 1000-1800 mm; temperature range 20-30°. Low dense forest with thick undergrowth though patches of teak and figs may be taller. Typical trees Anogeissus, Boswelfia, Tamarindus, Santalum, Moringa. Some penetration of African derived flora. Mainly southern slopes and parts of Mudumalai, but also in western slopes rain-shadows. 5. thorn forest zone Rainfall 500-1000 mm; temperature range 22-33°. Open savanna forest with Acacia, Zizyphus, Euphor- bia and other African elements as dominants and much admixture of Afrotropical flora. The eastern half of the northern slopes, the eastern third of the southern slopes, and locally on the southern foot- hills where not disturbed by agriculture. 6. dry agricultural zone Rainfall 500-2000 mm; temperature 20-35°. Crops are mainly rain-fed millets and certain pulses. Many tropical weeds and some remnants of thorn forest vegetation. Very dry during dry season, prone to drought. Much of the plains adjacent to the southern and northern Nilgiris where forest has been lost. Subtropical Division (1300-1900 m) 7. SUBTROPICAL EVERGREEN FOREST ZONE Rainfall 1300-4000 mm; temperature range 15-25°. In some respects transitional between tropical and montane evergreen but with some special elements. Olea dioica is typical. Dense dark forests, now much reduced in extent. Most of western and southern slopes where not cut down, and parts of the NW slopes above the Wynaad and Ouchterlony. 8. SUBCLIMAX FOREST FORMATIONS ZONE Rainfall 1000-1600 mm; temperature range 18-30°. Impoverished version of (7) above where rainfall and soil is deficient. Often clumps of isolated trees in grasslands. Here and there on northern and southern slopes but of marginal importance. 9. SUBTROPICAL AGRICULTURAL ZONE Rainfall 1100-4000 mm; temperatures 15-30°. Coffee at lower levels, tea at upper reaches, with vestiges of forest along rivers and where too steep for plan- tation crops. Montane Division (1800 m++) 10. MONTANE EVERGREEN FOREST ZONE Rainfall 1400-6000 mm; temperature range 10-20°. Dense evergreen forest with Oriental montane and many Palaearctic plants. Among trees members of genera such as Ternstroemia, Eugenia, Micheiia, 34 BUTTERFLIES OF THE NILG1R1 MOUNTAINS Gordonia, Rhododendron, etc. Many low montane plants ( Begonia , Lobelia, Impatiens ) and Palaearctic plants ( Fragaria , Viola, Rubus). Dotted throughout the plateau, but much reduced in extent. 11. MONTANE GRASSLANDS ZONE Rainfall 1300-3500 mm; temperature range 10-20°. Open rolling grasslands with a somewhat complex structure, much variation in structure according to soil, exposure, drainage and rainfall. Only fire resis- tant tree is Rhododendron but exotic gorse and broom prevalent in many areas. Now mainly from western escarpment system to Mukurti. 12. MONTANE AGRICULTURAL ZONE Rainfall 1300-3500 mm; temperature range 10-20°. Three types: monoculture of exotic Acacia, Euca- lyptus etc; monoculture of tea; and vegetable gar- dening (carrot, potato, cabbage and others) with some fallow land. Presently covers most of the plateau. History of exploration General The Nilgiri Mountains long lay as an iso- lated and unknown ‘jewel of nature’ protected by a wide band of malarious jungles, guarded by tigers and fierce elephants, largely unknown to the outside world. On the plateau lived the pastoral Todas exploiting the extensive grasslands, with primitive hunters and gatherers in the surrounding forests, but never exceeding the 10.000 mark as far as popu- lation was concerned. Their impact on the environment must have been marginal, but they developed a culture and a matrilineal kinship system that has been the delight of anthropologists. More must have been written per capita of the Todas than of any other people. The Kotas provided supplementary skills in the type of symbiotic relationship that pastoralists often develop with other groups. Lower down the mountain lived primitive hunter-gatherer societies, Irula, Kurumba, Panniya and related tribes. Their numbers were, in total, somewhat larger than that of the Todas, but they lived in harmony with nature and had neither the skill nor the in- clination to encroach in any major way on the natural conditions. Some four hundred years ago the Badaga appeared on the scene. They were Kannada speaking people fleeing some political up- heaval in what is now Karnataka, possibly the repercussions of the Bijapur Muslim con- quest of the Vijayanagar empire. They were the first agriculturalists on the plateau and settled in the villages that still carry their original names, digging their fields, but not maintaining much contact with their area of origin. Though probably more in number than the resident Todas, it seems that a modus vivendi was arrived at so that frictions bet- ween the radically different life styles of the two groups were minimised. Life probably went on much as it had always done, except for the compact Badaga villages and their agriculture. We do not know, it has to be said, because information on the Nilgiris was almost nil and no written tradition exists. Just around 1600 the modem world made its way to the Nilgiris. Rumours of Christian communities in the distant mountains, derived from the visit of the Apostle Thomas in the early days of Christianity, reached the court of the Bishop of Calicut. He sent off a group of Portuguese priests to investigate matters, but his judgment in choice of emissaries was not too sound. They came back with a report that was not ‘so sure and complete as was desirable’, a rather nice turn of phrase. So a few years later, in 1603, heavier guns were brought to bear. The Jesuit Father, Jacomo Ferreira, led an expedition on behalf of the Bishop of Malabar in Calicut. He brought back a fair amount of circumstantial and anec- dotal information, but of one thing he was certain .... there were no Christians in the Blue Mountains, and so the Portuguese lost interest. 35 JOURNAL , BOMBAY NATURAL HIST. SOCIETY, Vol. 84 In 1799, after the defeat of Tipu Sultan at the hands of the East India Company, the Nilgiris were part of the territory ceded to the Company through the treaty of Sriranga- patna, though probably Tipu never held much actual sway north of the Wynaad. The moun- tain was visible to the large British garrison and civil establishment at Coimbatore, but it was only in 1812 that the first representative of the colonial power visited the mountain itself. This was in the form of a somewhat unimaginative tax collector whose only com- ment was that the Nilgiris were not worth any particular efforts on his behalf A more official and less single-minded party set off during the winter of 1818-1819. They reported back with enthusiasm on the climate, vegetation and wildlife of the plateau, in fact describing a paradise on earth. They were perhaps over- enthusiastic, because their reports were flatly disbelieved at Fort St. George in Madras .... or at least as flatly as bureaucratic niceties would allow: It is somewhat difficult to accept the understandable enthusiasm of a group of men notwithstanding .... etc. But truth will not be concealed, and soon the Collector of Coimbatore became the driv- ing force in the establishment of the Nilgiris as a major hill resort. By 1827 there were seventeen European houses in Ooty and five in Kotagiri. Ten years later large scale esta- blishment of plantations was in full swing and the destiny of the Nilgiris was changed fore- ver. The population swelled with an influx of Tamil speaking plains people, needed for the tea and coffee plantations. The main towns of Ooty, Coonoor and Kotagiri became trading centres over and above their recreational function. By 1860 the foundations of what the Nilgiris still are to-day were laid. Entomological exploration Butterflies from India reached the scientific community already at the time when Carl von Linne published his tenth edition of systema naturae (1758), the starting point of zoolo- gical nomenclature. This book contains many Indian species, and many more were describ- ed by Fabricius and Cramer in 1775. Soon after the Nilgiris were opened up they were visited by explorers and naturalists, both pro- fessional and amateur. The first of these was the Austrian nobleman and naturalist. Baron von Hiigel in the late 183Gies. His material was described by Kollar and the Felders. It is, however, striking how relatively late some of the more prominent South Indian endemic butterflies were actually described as evidenc- ed by the list below : Pachliopta pandiyana Moore 1881 Papilio Uomedon Moore 1874 Papilio dravidarum Wood-Mason 1880 Papilio buddha Westwood 1872 Prioneris sita Felder & Felder 1865 Colias nilagiriensis Felder & Felder 1859 Celatoxia albidisca Moore 1884 Parantica nilagiriensis Moore 1877 Idea malabarica Moore 1843 Mycalesis adolphei Guerin-Meneville 1843 Ypthima chenui Guerin-Meneville 1843 Calaenorrhinus ambareesa Moore 1867 Thoressa honor ei de Niceville 1887 Oriens concinna Elwes & Edwards 1897 The first systematic account of the Nilgiri butterflies dates back almost exactly a century when Sir George Hampson [1888 (1889)] made a comprehensive list based on his five years of residence in the Nilgiri Wynaad as a coffee planter. He collected mainly in the northwestern corner, including the Nadgani Ghat, but only in that part which lay in the Madras Presidency, and not the Malabar side, now in Kerala State. His energies were sub- sequently transferred to moths. On his return to the United Kingdom he joined the British 36 BUTTERFLIES OF THE N1LGIRI MOUNTAINS Museum (Natural History) and produced the magnificent series on moths published as part of FAUNA OF BRITISH INDIA. His total output outstrips most other entomologists that have ever lived, yet despite this, the quality and accuracy of his work is legendary. Hampson’s list contained 275 entries and he expected that ‘no more than about twenty species would be added to it’. On current taxonomical view the list actually contains somewhat less than 260 species. In all just about forty species have in fact been added since, but as a first effort it certainly is no mean achievement. It is a great shame that none of his copious field notes were included. They are available in the British Museum (Natural History) but I have not been in a position to avail myself of them. It is hardly surprising that the nomencla- ture adopted by Hampson is often difficult to decipher for a contemporary reader. We are therefore indebted to Yates (1935) for ‘trans- lating’ Hampson’s list to the language used by Evans (1932) which is largely intelligiblle to-day, and for adding quite a few species from smaller contemporary collections. Many of Yates’s additions were species from the lower parts of the Nadgani Ghat. Yates’s list comes to 282 entries, some of which are now considered forms or synonyms. The next list of Nilgiri butterflies is that of Wynter-Blyth (1944, 1946). At the time he was headmaster at the school at Ketti below Coonoor. Most of his collecting was done at Ketti, and on the Ghat between Coonoor and the plains, and not least at Kallar, a magni- ficent place for collecting butterflies to this very day. He did make a number of visits to the Nadgani Ghat, but because of wartime petrol rationing he did not visit the lower parts on the Kerala side. In his day, of course, the forests stretched much more east, almost to Gudalur. His list came to 290 entries, later supple- mented with a further twenty or so. Some of these were again synonyms or forms, and one or two were erroneous. His paper is a very solid one with which I have few quarrels. Wynter-Blyth was later to publish the most recent guide to Indian Butterflies (1957) and the influence of his work in the Nilgiris is clearly visible in his book. Various taxonomic works have since scoop- ed up old unpublished material, and I have collected six or seven species never before recorded from the Nilgiris, so that the pre- sent list contains 299 valid species. As shown at the end of the list of Nilgiri butterflies there are about a score of South Indian butterflies that have not yet been recorded from the Nilgiris. Some of these, probably about a dozen, will one day be found also there. They will be an incentive to future collectors, but more precise ecological data and better etho- logical observations than my own should be the priority, and a much more satisfying pros- pect that the record of a few additional species. Current study I deliberately chose to study the Nilgiri fauna because it was already relatively well known, and because I had some opportunities for comparison stretching back a hundred years. It was clear from the outset that the number of new records would be relatively small, but I was more interested in the gene- ral ecology of the area. After publishing this systematic account of the Nilgiri butterflies it is my plan to convert the data into a more formal ecological-cum-biogeographical analysis over the next few years. Also of interest was an assessment of the extent to which the eco- logical degradation that has taken place in the Nilgiris over the past 100 years had in- fluenced the butterfly fauna. When choosing 37 JOURNAL, BOMBAY NATURAL HIST. SOCIETY, Vol. 84 the Nilgiris I was not unmindful of the fact that most of the southern mountains in India are basically very similar as far as the butterfly fauna is concerned, except that if tropical rainforest is absent, then so are several of the more interesting butterflies. Had I chosen the Ghats below Sultan’s Battery to somewhere in the South Sahyadri I would have had the pleasure of filling out some — largely predic- table — distributional blanks, but it would have been at the cost of data on the ecology of the butterflies concerned. I was in the Nilgiris, with occasional visits to the Annamalais, the Biligiriranga Moun- tains, and Kanara, from 11. iv to 18.x. 1986. During this time I spent roughly half my days in the field visiting all the ecological zones as often as possible. My main areas of work were the Longwood Shola at Kotagiri, the forests around Glenburn, Kotagiri Ghat, Kallar, Nad- gani Ghat and especially the little river at its foot, the area around Masinagudi and the Coonoor Ghat. Many other localities were visited once or twice. It is a pity that I was not able to spend a whole year in the area. The profound dry season from January till April has a very definite effect on the butterflies and it would have been interesting to study this not ieast in the Nadgani Ghat area. Some species are mainly on the wing during dry season in the rainforest zone. However, I do not think this loss of information has introduced serious bias into the systematic part that follows. Acknowledgements I have rarely been so well and generously treated as by my friends and colleagues in the Nilgiris, none of whom I knew before my arrival at base in Kotagiri. The Asokan family, Bunty and Rudi Bosen, the Devadars, senior and junior, Mr. and Mrs. Dharman, Palni and Uma Palaniappan, and Mr. Radcliffe. From many others I received much help and kind- ness. Gordon Thompson provided hospitality, good company and many interesting specimens throughout, and my debts to him are high. Thanks are also due to Nancy Fee for gene- ral encouragement and the necessary impetus to do things that would not otherwise have been done. Augustine and Dominic in Kota- giri did all they could to make my stay at ‘The Nest’ pleasant. I spent some interesting and very pleasant days at the Peechi Forest Research Institute in Kerala, being looked after in the best possible way by Dr. K. S. S. Nair and Dr. G. Matthew. Mr. J. C. Daniel of the Bombay Natural History Society aided with a number of prac- tical issues, as has the Royal Danish Embassy in New Delhi. Philip Ackery and Dick Vane-Wright, Pam Gilbert, Rienk de Jong, Niels Peder Kristen- sen and Brenda Leonard kindly provided a lifeline to library resources and advice over long distances. In Copenhagen Harish Gaonkar kindly undertook the task of screening the manus- cript which resulted in helpful advice on a number of matters and the eradication of some errors that would otherwise have survived. I am indebted, as often before, to the Danish Carlsberg Foundation whose grants have assisted in making this and other pro- jects economically feasible. I have not bracketed the names of authors where the genus has been reassigned since the original description since virtually all Indian butterflies are no longer in their original genus. Systematic part The systematic part that follows gives the basic data on each of the species that I accept 38 BUTTERFLIES OF THE NILG1RI MOUNTAINS as having been genuinely reported from the Nilgiris. The basic sequence is that of Eliot (1978) except that the Lycaenidae have been placed before the Nymphalidae in accordance with modern usage. The subfamilies of the Nymphalidae were in older literature usually accorded family rank, but there is increasing agreement that they should not be. Every effort has been made to refer to the butterflies according to the most modern nomenclature, much of it adopted from Eliot (1978) and a number of recent revisions. It has not been thought the place to discuss nomenclature in any detail, except where ex- cessive confusion might be engendered, or where I have deviated from current practice. Where a name in Wynter-Blyth (1957) differs in genus or species, the name used by him is given in brackets below the current one. Under each species I have endeavoured to give the status and the distribution in the Nilgiris, with whatever other information that might be of special interest. I have also sum- marised in brief terms the global distribution of the Nilgiri species. These are not meant to be definitive, but simply to give an impression of how the South Indian fauna is related to its neighbouring regions. Systematic list of the Nilgiri butterflies Papilionidae Papilioninae troidini 001 . Troides niinos Cramer ( Troides Helena) The southern birdwing is the largest of the Nilgiri butterflies, being endemic to southern India and the Western Ghats. It is replaced by Troides darsius Gray in Sri Lanka, while its closest relative in the Himalayas is Troides aecus Felder & Felder rather than, as often assumed, Troides helena Linne. Haugum & Low (1982-1985) discuss these relationships. The species is not rare and may even be common on the western slopes during the monsoon and immediate post-monsoon months. The main habitat is lowland ever- green forest, but the species is at home also in mixed deciduous forest and the subtropical evergreen. It is a good coloniser of agricultu- ral lands and is often common in coffee plantations. Wandering specimens are not in- frequently met with on the plateau itself though it cannot breed there. The biggest con- centration I ever saw was on some land that had been cleared for coffee a few years ago at Kokode Estate near Sholarmuttam. Hun- dreds were feeding from Lantana growing among the forest trees that had been left as shade trees. The species is active in the early morning, coming down from the tree-tops to feed from Lantana and Mussaenda. By 10.30 they once again ascend and are then very diffi- cult to catch. Females will often be found sitting with the wings open with up to three males assiduously courting them. The adults do not come to water. In captivity, on cold mornings, the butterflies will increase their body temperature by wing-quivering in the manner of many moths, being able to fly actively under conditions where other low- land species are immobilised. The food plants are Aristolochia indica and in the wet zone especially Thottea wallichi, a Troides food plant not mentioned by Haugum & Low (1982-85). Much concern has been express- ed over the conservation status of Bird wing butterflies worldwide, and the European Economic Community has banned their im- port and export even for scientific use. The Southern Birdwing is by no means a threaten- ed species, and in some cases its numbers increase through the agricultural activities of man. 39 JOURNAL, BOMBAY NATURAL HIST. SOCIETY, Vol. 84 002. Pachliopta pandiyana Moore (Tros pandiyana) The malabar rose is an unusual swallow- tail which is endemic to the wetter parts of the Western Ghats system and it is closely related to the Sri Lankan endemic Pachliopta jophon Gray. It appears to be strictly limited to the wettest type of rainforest, where it is some- what local, but sometimes very numerous indeed. I have seen hundreds of fresh speci- mens of both sexes at Lantana on the Nadgani Ghat. Activity began as early as 06.45 when the only other butterflies about were skippers, and it was not unusual to find specimens flying even in heavy rain. It normally flies in dense forest but will visit clearings and roads where flowers grow; at such times the three South Indian Pachliopta may be seen side by side though they are normally ecolo- gically segregated. In the forest the flight of the species is very slow at the lower level of the canopy and despite their great difference of pattern the resemblance to Idea malaharica is great, though I will not go so far as to postulate a co-mimicry relationship. A speci- men escaped from a photo session in Kotagiri and was immediately snapped up by a Red- whiskered Bulbul (Pycnonotus jocosus). By the time the bird had moved out of sight the wings of the butterfly had been crumpled up, but the body not yet damaged. I unfortu- nately did not see whether the butterfly was actually eaten. It is obviously a protected species, but the bulbul would not have known this. As far as is known the only larval food plants are Thottea siliquosa\ according to Jason Weintraub this is characteristic for the group to which this species belongs. 003. Pachliopta aristolochiae aristolochiae Fabricius ( T ros aristolochiae ) The common rose is often referred to as being very common, but this is not normally the case in my experience, though it is wide- spread and obvious. Normally numbers are not all that large, though I did see a huge migration at Kotagiri in autumn of 1957 in- cluding millions of this species and of Pach- liopta hector in roughly equal proportion. It is at home in most types of habitat with the exception of dense, wet forest, but it occurs on the plateau only as a straggler. Contrary to the two other members of the genus the Common Rose will visit water, usually early in the morning well before the Pierids and the Papilio begin. Large quantities of water are ingested and immediately excreted through the anus, so obviously salts or nitrogenous substances are extracted metabolically. The flight is slow and deliberate. Flowers are fre- quently visited and the species often spends the night in communal roosts with P. hector. The species is found practically throughout the Oriental region, being replaced by P. polydorus Linne in New Guinea. One of the female forms of Papilio polytes is an excellent mimic of this butterfly, and in nature it is often necessary to look twice before being certain which of the two species is involved. The female of the Zygaenid moth Histia nilgira Moore is a fine example of co-mimcry. I saw it only once, at Glenburn, and it took some considerable time before I realised that it was not a small P. aristolochiae. The moth is certainly aposematic and exuded a foul smelling yellow substance from the tegulae when handled. At rest it adopted the normal Zygaenid posture and all resemblance to a swallowtail was lost. 004. Pachliopta hector Linne ( Tros hector ) The large and beautiful crimson rose is endemic to Sri Lanka and South India, being found mainly south of the Godavery river and in West Bengal. There are records also from 40 BUTTERFLIES OF THE NILGIRl MOUNTAINS eastern Burma and the Andaman Islands. It is sometimes very common indeed, may migrate by the million, and congregates in small forests and fruit tree groves in commu- nal roosts during winter. One such roost at Mahabalipuram in December 1985 must have contained many tens of thousands and possi- bly more than 100,000. I have never seen as many large butterflies in one place [though of course the wintering sites of the American Monarch Danaus plexippus (Linne) would vastly surpass it in numbers]. P. hector is very fond of flowers, but never comes to water. The main habitat is open dry decidu- ous forest and ill-kept agricultural land at low altitudes, though migrants and vagrants will be found to the highest peaks. I once saw the species literally fall out of the sky at Kallar. On closer investigation it turned out that a supernumerary male had attached itself to a copulating couple, clinging so tenaciously to it that the threesome could be lifted off the ground by holding a wing of any of the speci- mens. The species would repay systematic investigation into diapause and migration. PAPILIONINI 005. Chilasa dytia clytia Linne The common mime, in both sexes, occurs in two forms. The typical form, clytia, is an excellent mimic of species of Euploea, while form dissirnilis is an equally good mimic of the two Tirumala and of Parantica aglea. The slow and deliberate flight of the Danaid models is beautifully copied, but try to miss a specimen with the net — off it goes at a speed which would give even a Charaxid a good run for its money. In nature the mimicry, as it often the case, is much more convinc- ing than in cabinet specimens, and I have rarely been quite certain whether I actually had a Mime in front of me before taking it out of the net. Working on Sri Lankan mate- rial Clarke & Karunaratne (1967) found that /. clytia was the ancestral form, and as in Sri Lanka my own Nilgiri observations indi- cate that there are two specimens of dissi- milis to every specimen of the ancestral form. As in Papilio polytes the relative proportion of the two forms may vary from place to place. Near Karkala in South Kanara Gordon Thompson and I found /. clytia to be more common than dissirnilis. In the Nilgiris the butterfly is quite rare and at best two or three are seen on any one day. Their main habitat seems to be the mixed dry-deciduous forests at low levels and the lowland evergreen forest. It penetrates the lower level of the subtropical evergreen zone and at Kallar may be found in the Arecanut plantations, probably because the larval food plant Cinnamonum is grown there. The adult butterfly will come to bolh flowers and to water where its habit of hover- ing above the flower or puddle will tell it apart from the models. The larva is very con- spicuous (see frontispiece in Woodhouse) while the twig-like pupa is one of the natural masterpieces of camouflage. The species is found throughout the Oriental region, each subspecies carefully tuned to mimetic resem- blance of the local Danaids. It would doubt- less be a fascinating laboratory insect for further genetic research. 006. Papilio demoleus Linne The lime butterfly is common and may be found anywhere in the Nilgiris, though it does not spontaneously enter the densest and the wettest of the lowland evergreen forests. However, numbers fluctuate considerably in time and space in a somewhat unpredictable way. It is a pest on cultivated citrus, but it also feeds readily on any type of wild ruta- ceous plant. I have not found it on cultivated fennel and it is perhaps surprising that none of the South Indian swallowtails has managed to transfer to this plant. In the immediate pre-monsoon period P. demoleus is an avid 41 JOURNAL, BOMBAY NATURAL HIST. SOCIETY, Vol. 84 participant in mudpuddling assemblies, as many as one hundred being found in tight groups, shoulder to shoulder. The distribution of this Oriental species is somewhat puzzling. The core area includes India to Malaya and southern China, with a secondary area in the Australian region. During the past seven hundred years it has colonised Arabia (Lar- sen 1983) where it feeds exclusively on culti- vated Citrus and where it is almost in contact with the African vicariant P. demodocus Esper. It invaded southern Iraq from Iran only since 1957 (Larsen 1977a). In recent years it has established populations also in Sumatra and on the Philippines, almost certainly in res- ponse to human interference with the natural environment. It is greatly to be hoped that the recent invasions will be well chronicled. 007. Papilio liomedon Moore The MALABAR BANDED SWALLOWTAIL is a close relative of the Oriental Papilio demolion Cramer, but is definitely specifically distinct and endemic to the Western Ghats, chiefly in the wettest parts of the lowland evergreen forest zone. It is a rather scarce butterfly which is also somewhat localised. The few Nilgiri records are all from the Nadgani Ghat, and here, too, I saw my own only Nilgiri specimen on flowers (19.0) just on the TN/ Kerala border. In behaviour it is much like the other swallowtails, though according to Gordon Thompson it does not come to water. It is certainly much more common in Kanara than in the Nilgiris, and in late September to early October good series may be collected on the globular red flower heads of Clero- dendron paniculatum in suitable localities. 008. Papilio dravidaram Wood-Mason The malabar raven is another endemic of the Western Ghats system, its closest relative being a species from Assam and Thailand. It is rather more common than P. liomedon, but is not frequently met with. The habitat is chiefly the wettest of the lowland evergreen forests and in the Nilgiris the main locality is Nadgani Ghat, though Gordon Thompson has occasionally taken it at Kallar. The colour pattern is reminiscent of the Euploea but the flight is very rapid and whether a genuine mimetic relationship is involved is perhaps doubtful. The butterfly has the disconcerting habit of suddenly emerging at furious speed from dense jungle through a road or clearing, only to vanish immediately again. It may be hard to distinguish from P. polytes on the wing. Wynter-Blyth (1957) states that it never seems to visit flowers, but that is not the case. I have caught several of my few specimens on Lantana. The species is regularly found at mud patches during the drier months and on hot days. 009. Papilio helenus daksha Moore The red Helen is the third largest butterfly of the Nilgiris and is common in the subtro- pical evergreen forests, relatively common on the plateau, and less common at lower ele- vations. It is a good coloniser and exists well outside of natural forest, and it is one of the few large butterflies to be found in tea planta- tions. It is, however, unable to survive in the drier thorn forests. It is very fond of Lantana blossoms and will sometimes assemble on mud patches, usually in the company of other swallowtails, but not usually with the large agglomerations of Pieridae. The weaving, random flight looks most haphazard, but in fact the butterfly makes rapid progress and is very difficult to catch. It seems to feed on virtually all available rutaceous plants and is easily bred. The global distribution covers almost the entire Oriental region as well as the southern fringes of the Palaearctic in China and Japan. Colour pattern apart, the simi- larity between this species and P. paris tami- lana is strong, and it is interesting that the 42 BUTTERFLIES OF THE NILGIR1 MOUNTAINS South Indian subspecies of both should be the largest of them all. 010. Papilio polytes polytes Linne The common mormon is indeed a common butterfly at lower and middle heights, but it does not ordinarily breed on the plateau, though from time to time it is seen even on the highest peaks. Like others of its kind it does not really penetrate virgin evergreen forest, though it colonises whenever it is dis- turbed by human activity. It comes to flowers readily and sometimes large numbers of males are seen mudpuddling. Such assemblages are usually independent of the Pierid agglomera- tions, occupying somewhat shady spots, often in the company of P. crino and other swallow- tails. The largest assembly that I have per- sonally witnessed consisted of 34 P. polytes, 5 P. crino, 1 P. demoleus, and 2 Graphium doson. Two of the three female forms are ex- cellent mimics of two of the red-bodied swallowtails, the Common and the Crimson Rose respectively, while the third female form is almost like the male. The flight pattern of the mimetic forms is also a wonderful copy of the slow and deliberate flight of the models. During my frequent visits to Kallar I noted all females seen in nature, though in many cases it was necessary to check the body colour to be quite certain whether model or mimic was involved. At Kallar the two mimetic forms predominated, the male-like form be- ing rare and constituting less than 5 percent of the total female population. The two mimetic forms were equally common. The precise data are as follows : hector- mimic 50.4%, aristolochiae mimic 45.7%, and male- like 3.9%. This is in marked contrast to the Delhi population of the same species, where some 60% were mimics of aristolochiae and about 40% male-like. P. hector does not occur in Delhi and the mimetic form is very rare, certainly less than one in a thousand (Larsen 1987). Such differences constitute powerful support for the concept of balanced poly- morphism as discussed by Ford (1975). Frjer (1913) bred the species extensively in Ceylon to work out the genetics and found that the male-like form constituted nearly 50% of the total. These results were puzzling to me till I realised that he was breeding the species at an altitude where the models are scarce indeed, just as in Delhi. The Common Mor- mon is, by any standards, among the world’s most interesting butterflies and an excellent insect for both field and genetic studies. An indispensable starting point for any research is the paper by Clarke & Sheppard (1972) on the genetics of the species. Oil. Papilio polymnestor Cramer The blue mormon is the second largest of the South Indian butterflies, and it is a plea- sure that such a large and handsome butterfly should be common practically everywhere. Only in the driest lowland habitats is it gene- rally absent, though stray specimens will turn up even there. I have seen the occasional specimen at Masinagudi. It is endemic to Sri Lanka and peninsular India, but is closely related to the widely distributed Oriental P. memnon Linne. The latter species has a host of mimetic female forms, some being quite similar to Pachliopta pandiyana, and it is per- haps surprising that the female of the Blue Mormon is similar to the male. Occasional females have bright red spots at the base of the forewing upperside, a feature which enters the mimetic pattern of some memnon forms. This red spot is very much more frequent in Sri Lanka than in the Nilgiris. Possibly the very narrow range and habitat choice of P. pandiyana makes it an unsuitable model. The Blue Mormon is found in all types of habitat, flying with a rapid weaving flight, covering long distances. It visits flowers avidly and comes freely to water. Unlike many swallow- 43 JOURNAL, BOMBAY NATURAL HIST. SOCIETY, Vol. 84 tails it comes to foul substances, otter drop- pings being a great favourite. I once found three males deep inside a cave where a group of otters had been dismembering crabs. The eggs are laid on all the natural rutaceous plants as well as on cultivated citrus. 012, Papilio paris tamilana Moore The large and brightly coloured South Indian subspecies of the paris peacock is the largest of all, and it qualifies among the finest butterflies anywhere in the world. It is found in all types of evergreen forest from near sea level till at least 2400 m, but I suspect its main habitat is the subtropical evergreen forest. It is found locally also in the moist- deciduous forests of the Wynaad and Mudu- malai. Wandering specimens may be en- countered in open country, even in the centre of Ooty, but this really is exceptional. Normally it does not stray from dense forest. The species is not really rare, though never numerous, but it may be very difficult to catch. Early in the morning it may be caught at flowers, and on hot days it visits damp patches, sometimes settling with the wings held flat against the substrate, which is some- what unusual in the family. Unusual, too, is the fact that females are often found in this way. At Sholayar in the Annamalai Moun- tains I found the species hilltopping along a ridge in primary forest and saw two copulat- ing pairs. I can subscribe to the view of old observers that P. paris has fixed patrol routes in its jungle habitats .... if you miss a speci- men it is likely to reappear an hour or so later from the same direction. They are fond of flowers, and when half a dozen or so are feeding on the same Lantana patch, the sight is a highlight of all that is enjoyable in natural history. The species is found in South India (but not Sri Lanka), in Orissa, and then from Kumaon east to most of the Oriental region. It is curiously absent from Malaysia. 013. Papilio crino Fabricius The common banded peacock is endemic to peninsular India and Sri Lanka. I have seen no phylogenetic reconstruction of the group, but it would appear that the species is fairly isolated. In South India it is more or less restricted to lowland mixed deciduous forest, with only the slightest of colonising ability. I have, however, in December 1983 seen it in the centre of Mysore, from where it was not recorded in lists from early this century (Watson 1890). The closest South Indian relative of P. crino is P. buddha which is essentially limited to the wettest parts of the evergreen zone with the result that the two species are hardly ever sympatric. P. crino is somewhat migratory so it may on occasion be found far from its natural habitats. The species is not very common in the Nilgiris. except at Kallar, where it may abound about the time of the onset of the monsoon. A few years ago several thousands were caught during one season on behalf of Japa- nese commercial collectors, but the popula- tion does not seem to have suffered. The species is very fond of mudpuddling and I have seen more than a dozen assembled, though none was a female. Both sexes come to Lantana, though visits to flowers are infre- quent at the time when mudpuddling is at its maximum. The courtship display, with the male hovering below the female, is very simi- lar to that of P. polytes. The larval food plant is known to be Chloroxylon, the Satinwood tree, but it seems that no-one has found the larva recently, and I was wholly unsuccessful in this respect. 014. Papilio buddha Westwood The buddha peacock is arguably the finest butterfly in South India. The general pattern is not unlike that of P. crino and on photo- graphs they may look alike. In real life the Buddha is so much brighter and more 44 BUTTERFLIES OF THE NILG1RI MOUNTAINS beautiful than the other gloss swallowtails, and it is a sight of rare beauty. The species is limited to wet lowland evergreen forest but is not usually found where rainfall is much less than 4000 mm a year, and it seems unable to survive in severely disturbed forest. On the Kanara Ghats, however, where paddy fields are interspersed with primary forest, the Buddha Peacock emerges from the forest to feed on Lantana and on Clerodendron pani- culatum. The latter plant is also a great favourite of the Purplebacked Sunbird, which attacks the Buddha in the belief that the butterfly is a supernormal rival, since the green colour of Buddha matches the green crown and the wing shoulders of the bird (Larsen 1987c). Normally the butterfly flies in dense forest at canopy level with a furious flight and is almost impossible to net except when visiting flowers. There are no records from damp patches, though P. crino is an avid mudpuddler. In the Nilgiris it is limited to ihe western slopes and though not uncommon is difficult to collect. It is more common in Kanara. The chief season is just after the end of the SW monsoon in late September and early October, but individuals may be met with at any time. Bell et ad. have data on diapause in Kanara. It should be mentioned that Gordon Thompson once caught a male at Glenburn, an illustration of the potential dispersal power of such a strong butterfly, I have given a more detailed account on the ecology and habits of this species elsewhere (Larsen 1987). LE PTOCIRCI N I 015. Graphmm sarpedon teredos! Felder & Felder (Zetides sarpedon) The common bluebottle is a genuinely common butterfly in the Nilgiris from the lowest levels to the highest peaks, and it is the one species able to colonise agricultural lands. However, it will not survive perma- nently in the driest tracts. It is an avid visitor to both flowers and damp patches, but at lower levels it is often outnumbered by G. doson. By disposition it is a very nervous insect and is not easy to catch. More than most swallowtails it is attracted to natural baits such as rotting grasshoppers and cicadas. I have found the larva on Cinnamonum, but it feeds on numerous other plants as well. The world-wide distribution covers practi- cally all of the Oriental region. 016. GrapMum doson eSeius Fruhstorfer ( Zetides doson ) The common jay is essentially a butterfly of the evergreen forest zone at both tropical and subtropical levels, but it is occasionally found away from forest proper. At low levels in evergreen forest it is sometimes more nume- rous than is G. sarpedon. Large numbers may be seen mudpuddling just before the onset of the SW monsoon, and when the monsoon changes in Sep. /Oct. My childhood records indicate that we did occasionally see it as high as 2000 m, but this is exceptional. The world-wide distribution covers Sri Lanka and southern India, suitable spots of the Eastern Ghats., and then most of the Oriental region. 017. Graphium agamemnon sneakies Felder & Felder ( Zetides agamemnon) The tailed jay with its apple green mark- ings is among the more attractive Nilgiri but- terflies, but on the whole it is less common than the previous two members of the genus. Its relative scarcity is a bit surprising since it is common in cities such as Mysore, Banga- lore and Madras. It is absent from the very driest tracts and not normally resident on the plateau, though it may breed there during summer. While it is an avid visitor to flowers 45 JOURNAL, BOMBAY NATURAL HIST. SOCIETY, Vol. 84 it is much less of a compulsive mudpuddler than the two preceding species. The global distribution covers the entire Oriental region. 018. Pathysa nomius nomius Esper The spot swordtail may be very common during the dry season at the foot of the Kotagiri Ghat and at Kallar, but it is rare elsewhere and at any other time. It seems to be a species of dry mixed deci- duous forest, sharing this habitat with Papilio crino. Just as the latter species has hardly any overlap with Papilio buddha, so F. nomius hardly overlaps with the rainforest species F. anti pat hes. Hampson caught only a few specimens on the northern slopes, and on the Nadgani Ghat I have only caught one, though F. antipathes may be common indeed. Normally the species flies high and fast, but in spring (March to early June) it is an avid visitor to damp patches. It is a migrant and specimens are seen on the plateau from time to time, but it does not breed there. The global distribution covers virtually the entire Orien- tal region. I agree with Eliot (1978) that it is legitimate to maintain the generic name Pathysa for the Swordtails, rather than to subsume them under Graphium. 019. Pathysa antipathes akibiades Fabricius The five bar swordtail is a dramatic in- sect that is limited to the wettest rainforests of southern India and Sri Lanka. Generally it is considered to be scarce, but in the right spots it may actually be very common indeed. The flight is fast and furious, and when on the wing the butterfly looks more like a fast Pierid than like a swallowtail. During Febru- ary/March, the main season, it is an invete- rate mudpuddler. When disturbed on a mud patch specimens will often perch on vegeta- tion close by. It is reputed to visit flowers but I have never personally witnessed this. The species is found in Sri Lanka, S. India, then again from Nepal east to most of the Oriental region. The South Indian subspecies has usually been referred to as ssp. naira Moore. However, Fabricius described ssp. alcibiades after specimens from ‘Tranquebar’, then a Danish colony and examination of the Fabri- cian types in Copenhagen by Harish Gaonkar clearly show them to be of South Indian origin (though Tranquebar cannot be the correct locality). PlERIDAE PlERINAE 020. Delias eucharis Drury The common jezebel is one of the most striking Indian butterflies and it is generally not rare, occasionally being locally abundant. It may be found practically everywhere in the Nilgiris, and it seems that the absence of frost and the presence of the larval food plant, Loranthus are the only ecological require- ments. The adult butterfly is very fond of Lantana flowers, but unlike the montane mem- bers of the genus in Papua New Guinea it is not attracted to water. The species is endemic to India and Sri Lanka and unlike other mem- bers of the genus which are forest dwellers, D. eucharis is common even in major towns and cities. This trait is shared with D. hyparete Linne which replaces it from Burma eastwards. When handled the species feigns death and it is almost certainly aposematic. In Prioneris sita it has a beautiful mimic. 021. Leptosia nina nina Fabricius The psyche is the only Oriental represen- tative of an African genus with half a score or so of very similar species. Eliot (1978) suggests that the Oriental species might be conspecific with the African L. alcesta but this seems unlikely in view of significant differ- ences in haploid chromosome numbers (n=19 in nina and n=12 in alcesta). The species is very common all year at Kallar but otherwise 46 BUTTERFLIES OF THE NILGIRl MOUNTAINS it appears to be remarkably scarce in the Nilgiris, though I have found it as high as 1900 m near Kotagiri in the 1950ies. It seems to avoid the wettest of the evergreen forests and I have not come across it at Nadgani. The flight is weak and fluttering, probably the most feeble of any South Indian butterfly. Flowers are visited, with Tridax as a firm favourite, while water is only occasionally attractive. I have once seen an African member of the genus actually alight on a pool of water. The species is distributed throughout the Oriental region. 022. Prioneris sita Felder & Felder The painted sawtooth is not rare on the western slopes of the Nilgiris and on occa- sion it may be very common on the Nadgani Ghat during the dry season when large num- bers are to be found on damp patches in the company of Graphium doson and Pathysa antipathes. It is distinctly uncommon during the rainy season and is only very rarely found outside of the wettest evergreen forest zone. I have taken a male (23. v) and a female (4.vi) at Kallar, but this is most exceptional. Both sexes are excellent mimics of Delias eucharis , a butterfly with a much wider range. The male, it is true, sometimes flies much faster than the model, in the manner of Hebomoia glaucippe, but at other times, such as when circling round a damp patch, the flight is quite like that of the Jezebel model. The species is endemic to Sri Lanka and the Western Ghats system. 023. Artogeia canidia canis Evans ( Pieris canidia ) The Indian cabbage white is a Palaearctic butterfly that is very common on the plateau above 1800 m, flying in a series of broods throughout the year. The original habitat must have been the edges of sholas, but it is now especially common in agricultural areas where it is a moderately serious pest of cab- bages. In northern India the species is migra- tory and visits the plains in winter (Larsen 1986a), but I have seen no evidence of migra- tion in the Nilgiris. The Bangalore record that so puzzled Wynter-Blyth (1957) probably came as a pupa with agricultural produce. The butterfly spends most of its time flying around in search of flowers and will occa- sionally visit damp patches on hot, dry days. It is found on the higher South Indian mountains but is absent from Sri Lanka. In all probability the total area above 1900 m on that island was too small to support a viable population. Outside of South India it is found from the Himalaya east to Japan, just penetrating the tropics in suitable loca- lities. 024. Cepora nerissa phryne Fabricius ( Huphina nerissa) The common gull is chiefly a butterfly of the drier lowland habitats, though it may be found in clearings at middle heights. Mixed deciduous forest, ill-kept agricultural land, and thorn forest are the main haunts of this gene- rally common butterfly. Due to strong migra- tory tendencies it is also met with on the plateau from time to time but I cannot agree with Wynter-Blyth (1946, 1957) that it is resident much above 1400 m. In spring large numbers may be seen mudpuddling with other Pierids. In behaviour it is more or less a tropi- cal replacement for the Small Cabbage White ( Artogeia rapae) and it is much less of a quarrelsome species than the two lxias. The distribution covers practically the entire Orien- tal region. 025. Cepora nadina remba Moore {Huphina nadina) The lesser gull is limited to the wetter parts of lowland evergreen forest and is never seen in open country. Hampson noted that 47 JOURNAL, BOMBAY NATURAL HIST. SOCIETY, Vol. 84 it was sometimes common, but generally it is uncommon. Usually only single specimens are met with. It spends most of its time flying in the forest where it is almost impossible to collect, but one or two can usually be found on a good mudpuddling patch and occasion- ally on flowers. Its behaviour and general aspect in nature contrast so strongly with those of C. nerissa that it is difficult to accept them as congeneric. Apart from in Sri Lanka and South India the species is found from Nepal east to Taiwan and Sumatra. 026. Anaphaeis aurota Fabricius The caper white is linked to the drier tropical habitats, but owing to a great migra- tory potential it may occasionally be found in numbers practically anywhere. Vast swarms are sometimes encountered and, like many migrants, numbers fluctuate considerably in a most unpredictable way. In the Nilgiris it is most consistently common in the thorn forests around Masinagudi. Under good conditions this butterfly breeds faster and more profuse- ly than practically any other and it may com- pletely strip all available food plants for miles around. The species is Palaeotropical, being found on the Indian subcontinent, Arabia and all over Africa. It occasionally invades the Mediterranean area (Larsen 1986b). 027. Appias indra shiva Swinhoe The plain puffin is a rather scarce butterfly that seems to be centred on the subtropical evergreen forest zone, though it may be found both below and above this level. There are few Nilgiri records. I took one at Tamizagham in Ooty (ii . 1984) and it seems to have a headquarters of sorts on the Coonoor Ghat between Wenlock Bridge and Benhope where Wynter-Blyth caught most of his material and where J have seen it on several occasions. I have seen it once only at Kallar (31. viii) and Gordon Thompson has a few taken at water on the Nadgani Ghat where I never saw it. No clear pattern emerges. It comes freely to flowers. In Sri Lanka the species is considered even more of a rarity than is the case in South India. Otherwise it ranges from Nepal east to most of the Oriental region where it is often common. 028. Appias libythea libythea Fabricius The name striped albatross is something of a misnomer for the South Indian popula- tion where the males are practically imma- culate except for some dark apical shading. The underside of the hindwing is chalky white lacking the cream overlay of the other- wise similar A. albina. A certain distinction between these two species lies in the cell of the forewings. A. libythea has the end cell vein forming a 90° angle to the costal edge, while in other Appias the angle is only 30°. Hamp- son considered it to be a rare species while Wynter-Blyth failed to find it. I have taken it on the Kotagiri Ghat, at Glenburn, Kallar, Ronningtown and even on Nadgani Ghat. It appears to be most unpredictable, possibly because it is migratory, but its headquarters are probably the mixed deciduous forests, though it might be without a permanent head- quarters. It is certainly most erratic in the Delhi area (Larsen 1987b). The range covers Sri Lanka and India, east to the Philippines and Malaysia. Numbers have increased signi- ficantly in Malaysia this century (Eliot 1978) probably because the butterfly does well in secondary vegetation in the wetter tropics. 029. Appias lyncida Inti fascia Moore The chocolate albatross is easily recog- nised by the deep yellow colour of the hind- wing underside and the broad chocolate borders. These are not obscured by the con- siderable level of individual and seasonal variation. I find it a most enigmatic butterfly in the Nilgiris, difficult to interpret in ecolo- 48 BUTTERFLIES OF THE NILG1RI MOUNTAINS gical and distributional terms. Hampson re- cords it simply in the words ‘1000 to 3000 ft’. Wynter-Blyth took only three at Kallar, com- menting that it was reputed to be common on the Kotagiri Ghat, where I never saw it. On 1 . v and 2 . v I saw single males flying towards the SW in Kotagiri town in the manner of normal migrants. I have seen small numbers on about half of my visits to Nadgani Ghat, where it sometimes came to water. Four were collected in mixed decidu- ous forest at Ronningtown. This is a pattern’ singularly lacking in consistency. The world- wide distribution covers Sri Lanka, the Western Ghats, and then from Nepal east to most of the Oriental region. An apparently isolated population exists in Orissa. 030. Appias albina darada Felder & Felder The common albatross is indeed the most common of the Nilgiri Appias, but like other members of the genus it is somewhat unpre- dictable. I would not, however, concur with Wynter-Blyth’s statement that it is ‘most abundant everywhere’. I have taken specimens in most of the localities visited from time to time, though this species will not be found in the drier lowland forests where A. libythea may be met with. The headquarters appears to be in the drier parts of the lowland ever- green forest where it merges with the mixed deciduous. It may be caught at flowers but more usually at damp patches. In the large migration that took place in late May and early June of 1986 (Larsen 1987b) this species contributed less than 0 . 5% of the total (some 20.000 individuals). In the migra- tions of my childhood (Larsen 1978a) there were millions. It ranges from Sri Lanka and the Western Ghats, through suitable places in peninsular India to Sikkim, and from there east to practically the entire Oriental region, New Guinea and NE Australia. 031. Appias wardii Moore The lesser albatross is inappropriately named since males are usually larger and more dramatically marked with black than those of the other species of the genus. The taxonomy of this and related species is diffi- cult. There is something to be said for uniting the large number of often disjunct and distinc- tive taxa under the ‘ paulina umbrella, but the South Indian form is so different from the Sri Lankan that specific status seems the best solution till a full revision of the species- group has taken place. In the wet season form the South Indian taxon is easily identified since it has the black markings normally characteristic of female A. albina . The dry season forms are very similar to male A. albina, though vestiges of the black apical markings sometimes remain. Hampson con- fused the genus so thoroughly that whether he obtained this species or not is uncertain. Wynter-Blyth did not mention it in his main Nilgiri paper. I have caught it on several occasions at Glenburn, once near Naduvattam, and on a few occasions at Nadgani. It is decidedly uncommon. The range is similar to that of A. albina, but the species is much less common, confined to dense evergreen forest, and populations are often disjunct. 032. Colotis amafa amata Fabricius ( Colotis Calais) The South Indian subspecies of the small salmon arab has traditionally been known as ssp. modesta Butler, but the Fabrician type is from South India and has been checked by Harish Gaonkar. At best the name modesta is applicable to the Sri Lankan population, but that hardly merits a name of its own. From North India and west through Arabia and the African Sahel it flies in the very different ssp. Calais Cramer. The Small Salmon Arab is chiefly found in the thorn forest formations and surrounding agricultural land, though I 49 4 JOURNAL, BOMBAY NATURAL HIST. SOCIETY, Vol. 84 have seen specimens also in the mixed deci- duous forest of the Kotagiri Ghat. The South and Central Indian subspecies is very different from the North Indian and African one, and its main food plant seems to be Azima rather than the more normal Salvador a. There might be a case for considering the two taxa speci- fically distinct, and certainly the presence of the two subspecies in India would appear to be due to a multiple invasion from Africa. In the Nilgiris area the species is quite localised but usually common where found, often in the company of the other Colotis species. The distribution covers all of tropical Africa, much of Arabia, the drier parts of the Indian sub- continent. and NE Sri Lanka. 033. Colotis etrida Boisduval The small orange tip is a common butter- fly in the drier lowland habitats, but it pushes further up the mountain and further into the wetter zones than do the other members of genus. There is a considerable degree of individual and seasonal variation, and the species is often wholly absent during the wettest months of the year. Occasional speci- mens are met with on the plateau indicating some capacity for dispersal. The butterfly comes avidly to flowers, especially Tridax, but neither this nor other members of the genus come to water, despite being found in very dry habitats. The genus is African, but the species is endemic to Sri Lanka and the Indian peninsula. 034. Colotis eucharis eucharis Fabricius The plain orange tip is a fairly local butterfly, but it is usually not rare where its food plant, Cadaba indica grows. This is nor- mally on rocky ground in the thorn forest of the foothills, or along the hedges growing at the edge of fields. It is almost invariably found in the company of Colotis danae with which it shares the larval food plant. Usually most of their time is spent flying about stands of Cadaba in search of partners, but they will feed from flowers. On exceptionally hot days the warmest hours are spent in the shade of dense trees, usually right down amongst the roots. The species is an intruder from the Afrotropical region, where it is found through- out the tropical zone. In India it is limited to the Deccan and southern India, being found in NE Sri Lanka as well, but not in Sind and Saurashtra where the Afrotropical species usually occur. This is a classical Sudano-Deccanian distribution pattern. 035. Colotis danae danae Fabricius In South India crimson tip butterfly is quite parallel in distribution and habits to Colotis eucharis but it is rather more common. If only one of the species is present it is almost invaria- bly C. danae. The nominate subspecies is found on Sri Lanka and in peninsular India as far north as Madhya Pradesh. A different sub- species inhabits the area from Saurashtra to Baluchistan. Usually known as ssp. dulcis Butler, I find it impossible to separate it from the Arabian ssp. eupompe Klug. Different subspecies are found throughout dry, tropical Africa. 036. Madais fausta fulvia Wallengren ( Colotis fausta) The salmon arab is a pretty and vivacious butterfly of the drier foothills where it may be, on occasion, quite common. It penetrates higher up the mountains and deeper into the evergreen forests than most of the related Colotis species. The flight is rapid and danc- ing and it is often difficult to capture a good series of specimens. As elsewhere in India, the larval food plants are Maerua, though the nominate subspecies from the Middle East feeds on Capparis as well. The species visits flowers freely, not least Tridax, but never comes to damp patches. The subspecies from 50 BUTTERFLIES OF THE N1LGIRI MOUNTAINS peninsular India differs from the others through invariably having white females. In northern India ssp. fulvia has dimorphic females, while in the nominate subspecies all females are salmon. The Indian populations appear to be highly sedentary though the nominate subspecies in Arabia and the Middle East is strongly migratory. The genus is monobasic, very close to Colotis, and the single species is found in Arabia, the Middle East and the Indian subcontinent, just penetrating the East African coastal regions north of Kenya. 037. Ixias marianne Cramer The two Indian Ixias are almost identical in distribution and habits, though in most of India the white orange tip, /. marianne, is slightly less common than /. pyrene. They are both common in the mixed deciduous forest, in thorn forest and in ill-kept agricultural land at low levels. There is little penetration into the subtropical zone, nor into the lowland evergreen forests. In May and June both are among the most prominent species in the mud- puddling assemblages, and both were promi- nent in the large migration that I observed in late May and early June (Larsen 1987b). Normally the butterflies are seen flying about the open scrubland in search of flowers, mates or food plants. /. marianne is endemic to India and Sri Lanka. 038. Ixias pyrene sesia Fabricius The use of subspecific names for the yellow orange tip in India has been the subject of some confusion. I follow Gabriel (1943) in using the name sesia. There is, in any case, so much seasonal and individual variation that the designation of subspecies is difficult. The habits are like those of the preceding species and were the ground colour of the two not different they would be impossible to distinguish in the field. The range stretches from Sri Lanka, throughout India east to Hong Kong and Malaysia. 039. Hebomoia glaucippe australis Butler. The giant orange tip is one of my favou- rite butterflies. It is the largest of the Indian Pierids, and it is a beautiful sight to see it swooping down the mountains along densely clad water courses with the wings held three-fourths open. It is generally common enough in mixed deciduous and open ever- green forest at low levels, sometimes being found also in subtropical evergreen forest. It is not a good coloniser of agricultural land and is rarely found in disturbed areas. It comes to flowers but is then very wary. Often it hovers on the flower in the manner of the Papilionidae, the only Pierid to do so. The species is an avid mudpuddler actively follow- ing river systems on the look-out for good spots which may well be situated some dis- tance from the forest edge. In the afternoon they may be seen moving up-river again. When mudpuddling, together with myriads of other Pierids and some Papilionids, the camouflage pattern of the underside may be seen at its best. Never mind how carefully you study the assemblage, one or two H. glaucippe will be overlooked. The larva is a startlingly effective snake mimic. When handl- ed it rears up its head, revealing a couple of blue eye-spots that are concealed when the larva is at rest. Most likely this is a defence against the Bonnet Macaque Monkey whose instinctive fear of snakes is almost comical in its manifestations (try throwing a little bit of rope out of a car and watch the results). The species is distributed in suitable country throughout the Oriental region. 040. Pareronia Valeria hippia Fabricius ( Parenonia ( sic! ) Valeria) The common wanderer appears to be very rare in the Nilgiris. Hampson mentions it with 51 JOURNAL, BOMBAY NATURAL HIST. SOCIETY, Vol. 84 no detail, Wynter-Blyth and I failed to find it, though there is a genuine specimen from Kallar in the Bombay Natural History So- ciety collections. This scarcity is difficult to understand as it is common enough in forest- ed country around Bangalore and as the forests on the northern slopes seem very suitable for the species. It is worth mentioning that the South Indian population throws up the occa- sional female of the form mimicking Parantica aspasia Fabricius, a Danaid that does not occur in peninsular India. The species is not found in Sri Lanka, but stretches from India deep into the Oriental region. 041 . Pareronia ceylonica Felder & Felder ( Parenonia (sic!) ceylonica) The dark wanderer is endemic to Sri Lanka and southern India where it is more of a wet zone butterfly than is the preceding species. The South Indian population is listed as ssp. pingasa Moore by d’Abrera, but I agree with Talbot that it is not worthwhile separating it from the nominate Sri Lankan. It will be found in the mixed deciduous forests as well as in tropical evergreen at Nadgani. It is normally somewhat scarce but may be quite common at Kallar. The flight is fast and restless through dense vegetation and even when coming to flowers it is wary, so pro- curing a good series is not always easy. I have never seen it at water, though I have seen a picture of one of the Pareronia in a mudpud- dling assemblage. On the wing the female is a most effective mimic of Parantica aglea, very much more so than one would suspect from cabinet specimens. COLIADINAE 042. Catopsilia pomona Fabricius (C. pomona & crocale) The lemon emigrant is a large and power- ful butterfly that is strongly migratory. Almost a million participated in the relatively modest migration that I documented in May /June of 1986 (Larsen 1987b), and many millions were involved in a month or more of intermittent, generally southwards migration during Sept- ember and October. Its main base is the drier lowland formations, but it can be found breeding practically anywhere except on the highest plateau. It is an avid mudpuddler and has a great liking for flowers. There are two chief forms: f. crocale has black antennae and has immaculate undersides; f. pomona has reddish antennae and the underside has dark irroration as well as spots at the end of the cell. Intermediates are found but they are not very common. There is no doubt that they represent forms of one species, but the rela- tive frequency of the two forms varies and their respective functions are quite unknown. The large numbers that settled on damp patches at Kallar in May 1986 were 95% crocale, while during the wet season the two forms were roughly equal in number. Both forms participate in the migrations so we are not faced with the type of phases so well known from migratory locusts. A thorough study of the dimorphism in this species would be most interesting. The range covers the entire Oriental region. New Guinea, parts of Australia to well out in the Pacific. A Malagasy species ( Catopsilia thaurama Rea- kirt) is sometimes, in my view not correctly, linked to C. pomona. 043. Catopsilia pyranthe Linne (C. pyranthe & C. florella) The mottled emigrant is a common butter- fly with both sexes showing a high degree of variation. The florella form with very narrow forewing borders and with well-developed, red-ringed silver spots at the end cell of the underside has traditionally been accorded specific status, especially since it is relatively constant in Africa, while the pyranthe form is fairly constant in eastern Asia. However, 52 BUTTERFLIES OF THE NILG1RI MOUNTAINS in India’s drier tracts both forms and inter- mediates occur with little consistency, except that the florella form is most frequent during the dry season. The species is common in the Nilgiris, sometimes very much so, usually in the py rant he morph. The species is strongly migratory and the progeny of migrants some- times breed even on the plateau, though it is mainly a species of the lowland drier tracts. It is fond of both flowers and damp patches. The distribution covers all of tropical Africa, southern Arabia, the entire Oriental region, New Guinea and parts of Australia. 044. Eurema brigitta rubella Wallace ( Terias libythea) The small yellow is a common butterfly with the widest possible of distributions, both geographically and ecologically. In the Nilgiris it may be found at all levels in most types of habitat, though it tends to avoid the dense evergreen forests. It is, in fact, one of those butterflies that seem to thrive particularly well in areas disturbed by the activities of man. It is not very common and somewhat local in the Nilgiris. Flowers are avidly visited, but the species is rarely seen at damp patches. It has been known to migrate but I have seen no evidence of this in the Nilgiris. E. brigitta was one of the few permanent residents of my compound in Kotagiri, the population usually consisting of four to eights individuals at any given time. One specimen was seen captured in flight and eaten by the Whitespott- ed Fantail Flycatcher ( Rhipidura albicollis). The range covers all of tropical Africa, south- western Arabia, most of the Oriental region and parts of New Guinea and Australia. From Sundaland eastwards it becomes a rather scarce and local grasslands species, possibly because of competition from the many other members of the genus in that area (Holloway 1973 gives an interesting review of the genus). 045. Eurema laeta laeta Boisduval ( Terias laeta) The spotless grass yellow is normally a common species, but on the whole this is not the case in the Nilgiris. Wynter-Blyth was a long time before finding it in numbers near Coonoor and at Gudalur. I have found only modest colonies in one or two places on the Nadgani Ghat and at Masinagudi. Obviously colonies may be found in nearly all types of terrain. There is strong seasonal dimorphism, but 1 have seen only wet season forms in the Nilgiris. The species has a limited distribution compared to some of the other Eurema, be- ing found only in Sri Lanka and from India to Burma, Thailand and Indo-China. 046. Eurema hecabe simulata Moore ( Terias hecabe) The common grass yellow is just that, very common practically everywhere. Since it also has a vast range it is probably among the top ten of the world’s most numerous butterflies. In the Nilgiris it may be found literally anywhere, though it is least common in dense evergreen forest where it is replaced by E. blanda. Indeed, E. hecabe is often parti- cularly common in habitats that have been modified by human activity. The species is a known migrant, but large scale migration has not been observed in the Nilgiris, though some did follow the May 1986 migration (Larsen 1987b). It visits flowers and often comes to damp patches. The range covers the entire old world tropical zone from Africa and Arabia to India and the Oriental region, to New Guinea, Australia, Japan, Fiji and Tonga. 047. Eurema blanda silhetana Wallace ( Terias blanda) The three spot grass yellow is restricted to lowland forests where it is sometimes very common indeed. The flight is often higher 53 JOURNAL, BOMBAY NATURAL HIST. SOCIETY, Vol. 84 above the ground than in the other species of the genus, probably because the larval food plants are tall trees like Albizzia and asso- ciated creepers like Wagatea spicata. Contrary to the other Eurema the present species lays its eggs in large batches and the larvae and the pupae are gregarious throughout, the black pupae being situated so close to each other as to almost touch, sometimes up to fifty at a time. The species may assume pest proportions in Albizzia plantations. E. blanda is found in Sri Lanka, South India’s wetter tracts. Eastern Ghats and then east to practi- cally all the Oriental region to at least Papua New Guinea. 048. Eurema andersonii ormistoni Watkins (Terias andersoni) The one spot grass yelijOW is the odd man out among the South Indian Eurema inasmuch as it is a genuinely rare and local species. In the Nilgiris it seems limited to evergreen tropical forest below 1500 m, which means that it just penetrates the subtropical level. Wynter-Blyth collected a few at Kallar where I have also taken the species. I also have a few from the Glenburn forests as well as one from an evergreen patch on the Kotagiri Ghat. The flight is weak and irresolute, the butterfly coming to both flowers and wet patches. It occurs from Sri Lanka through suitable wet tracts of peninsular India to Malaysia, and it seems to be scarce and local everywhere. 049. Colias nilagiriensis Felder & Felder ( Colias erate ) The nilgiri clouded yellow is quite common on the plateau above 1900 m, flying throughout the year, especially in places with some moisture. The biggest concentration I have seen is in the beautiful bog at Longwood Shola near Kotagiri where hundreds may be seen on a good day, buzzing about just above the surface. The larval food plant here is Parochetus communis , though I suspect other plants are used elsewhere. All South Indian females of this butterfly are white, though in all its closest relatives females occur in both white and yellow forms. While related to the Palaearctic C. erate Esper I can see little reason for not accepting the isolated and morphologically very distinctive South Indian taxon as specifically distinct. As such it is a South Indian endemic species of obvious Palaearctic origin. {to be continued) 54 NOTES ON COMPARATIVE BODY SIZE, REPRODUCTIVE EFFORT AND AREAS OF MANAGE- MENT PRIORITY FOR THREE SPECIES OF KACHUGA (REPTILIA, CHELONIA) IN THE NATIONAL CHAMBAL SANCTUARY1 R. J. Rao2 and L. A. K. Singh3 (With a plate & three text-figures) Record lengths of carapace have been reported from the River Chambal for Kachuga tentoria circumdata (26.5 cm), K. dhongoka (48.0 cm) and K. kachuga (49.0 cm). The ratios, carapace length/breadth and plastron length/breadth were similar in all three species. The ratio carapace length/ shell height was less in K. tentoria as it bears a high dome. Out of 13 areas of nesting identified for manage- ment-priority, 23.0% were shared by all three species and 46.1% by K. tentoria and K. dhongoka. Although biometrical features relating ‘body capacity’ to ‘egg mass occupancy’ in the body indicate similarities between K. tentoria and K. kachuga, only K. tentotia is known to lay more than one clutch per season while others lay only one. The incubation period is 5-8 months for tentoria as winter-laying is there, but is 2 months for the others. The clutch sizes are 4-9 (tentoria), 21-35 (dhongoka) and 11-18 (kachuga). Egg length x breadth cm x weight g are: 4.7 x 2.7 X 21.4 (tentoria), 5.9 x 3.6 x 44.2 (dhongoka) and 7.0 x 4.1 x 57.4 (kachuga). I NTRODUCTION Three species of Kachuga (Emydidae) namely, K. tentoria circumdata, K. dhongoka and K. kachuga occur in the National Cham- bal Sanctuary along the River Chambal of the Gangetic system (Rao and Singh in press). In the following we present data on the body size of these chelonians with preliminary analysis of the observations on their repro- duction. Besides, a list is given of the nesting areas that are of significance to a Manager for the species in the Sanctuary. The National Chambal Sanctuary, created in 1978 extends from Jawaharsagar Dam (Rajasthan) to Kota barrage (Rajasthan) and 1 Accepted September 1984. 2 Camp: National Chambal Sanctuary. Post Box 11, Morena 476 001, Madhya Pradesh. Crocodile Research Centre, Hyderabad 500 264. after a gap of 18 km, from Keshoraipatan (Rajasthan) through Pali (M.P./ Rajasthan) to Pachhnada (Uttar Pradesh). The total length of the river inside the Sanctuary is about 600 km. Materials and Methods Three live and 22 caracases of K. t. circum- data and three caracases each of K. dhongoka and K. kachuga were collected outside water during survey trips made by boat and foot between October 1983 and July 1984. The shell measurements of all specimens were taken as described by McRae et al. (1981) and the data were used to compare the maxi- mum sizes reported in literature for different species (Smith 1933, Pritchard 1979, and Daniel 1983), and determine the similarities in certain biometrical ratios. The carapace 55 JOURNAL, BOMBAY NATURAL HIST. SOCIETY, Vol. 84 length and carapace width (= plastron width) were taken as straightline measurements. Four gravid females, represented by two of tentoria and one from each of the two other species were examined at the ovary (for methods see, Moll and Legler 1971. Rao 1982) to determine the nesting frequency and collect comparative data on clutch and egg sizes in relation to the length of the females. In order to compare the species-wise re- productive effort, the approximate body capa- cities (BV) have been considered against the approximate egg-clutch mass (EM) and egg- clutch occupancy (EV). The following methods were used in calculating BV, EM and EV. PL x PB x SH, BV f , 2 EM = CSxEW, and EV = CSxELxEB In the above, PL and PB are the mean lengths and breadths of plastron, SH, the mean shell height; CS, the mean clutch size (mean number of eggs per clutch); EL and EB, the mean length and breadth of eggs, and EW the mean egg weight (g). All measurements are in centimetres. The incubation period was determined by direct corroboration of nesting activities, period of commencement of embryonic deve- lopment and periods of sightings of hatchlings. Different nesting areas and their manage- ment priorities were determined during the surveys from the extent of nesting activities and predation pressure. Results Size and biometrical ratios. The average shell measurements of dead and live speci- mens of hardshelled turtles collected in the Chambal river are presented in Table 1. The mean carapace lengths were 24.65 cm (Kachuga tentoria circumdata), 44.6 cm. (K. dhongoka ) and 47.6 cm ( K . kachuga). The maximum length of the carapace in the three species were 26.5 cm, 48.0 cm and 49.0 cm, respectively. The plastron was always smaller than the carapace but attached to the latter at such a point in the front that it protruded out a little behind the rear end of the cara- pace. Therefore, the total lengths (Table 1) were larger than the carapace length. The ratios carapace length/width were 1.38, 1.38 and 1 . 32 for K. tentoria circumdata, K. dhongoka and K. kachuga respectively (Table 1). The ratio, plastron length /width were 1.26, 1 . 27 and 1 . 25 respectively and the ratio, carapace length /shell height (body depth) were 2.24, 2.73 and 2.34. For one male K. kachuga the ratios were: carapace length/ width 1.25, plastron length /width 1.14 and carapace length/shell height 2.36. The male (29.5 cm carapace length) was caught by net near Babu Singh ka gher (Table 2) on 5 December 1983. The turtle, being in its breeding colours, had six red longitudinal stripes along the neck, a pair of oblong yellow spots on the throat and the head was brilliantly red on the top and bluish on the sides. Nesting areas, Distribution pattern. Kachuga are very scarce in the river upstream of Pali. Downstream of Pali although pre-nesting acti- vities were seen, no nests were located. The first major nesting site was located at Baroli (Table 2), 57 km downstream of Pali. Out of the total 13 identified nesting sites (Table 2) three (23.0%) were common to all three species, and six (46.1%) were com- mon to K. tentoria and K. dhongoka. There were no regular distance-intervals between any two nesting sites (Figs. 1 and 3). Nesting frequency. Three predated nests, about a week old, discovered on 29 October 1983 were the earliest record of nesting by Kachuga tentoria. Fresh nests (confirmed from fresh turtle tracks) were located on 5 56 NOTES ON THREE SPECIES OF KACHUGA Table 1 Body and egg biometrics of Kachuga tentoria circumdata, K. dhongoka and K. kachuga. for SL. no. 1 through 8 n = 25 (tentoria), 3 ( dhongoka ) and 3 (kachuga). sizes in cm and weight in g (range in parentheses) SI. Aspects K. t. circumdata K. dhongoka K. kachuga No. 1 Carapace 24.65 44.6 47.6 length CL (22.0-26.5) (44.0-48.0) (46.0-49.0) 2 Carapace length in: (a) Smith 1933 23.0 40.0 39.0 (K. tectum/K. t. tentoria) (b) Pritchard 1979 17.7 (7") (K. tentoria) 40.6 (16") 40.6 (16") (c) Daniel 1983 23.0 40.0 39.0 (K. tecta/ K. tentoria) 3 Carapace 17.78 32.3 36.0 width CB (16.5-19.5) (28.0-35.0) (35.0-37.0) 4 Plastron 22.45 41.16 45.0 length PL (21.5-24.0) (40.5-44.0) (43.0-46.0) 5 Total 24.8 44.8 49.3 length TL (22.0-27.5) (40.0-47.5) (47.5-50.5) 6 Shell 11.0 16.3 20.3 height SH (8.0-12.5) (15.5-17.5) (20.0-21.0) 7 Anal 4.0 5.3 6.0 width AB (3. 5-5.0) (5. 0-5. 5) (5. 5-7.0) 8 Anal 3.25 6.6 6.7 notch AN (3. 0-3. 5) (6. 0-7.0) (6. 5-7.0) 9 CL/CB 1.38 1.38 1.32 10 PL/CB 1.26 1.27 1.25 11 CL/SH 2.24 2.73 2.34 12 Clutch 5.95 23.64 15.5 size CS (4-9) (21-35) (11-18) n=20 n=31 n=4 13 Egg 4.755 5.993 7.049 length EL (4. 5-5.0) (5.18-6.62) (6.65-7.54) n=119 n=220 n=62 14 Egg 2.754 3.611 4.161 width EB (2.60^-2.90) (3.20-4.12) (3.76-4.56) n=119 n=220 n=62 15 Egg 12.40 44.21 57.42 weight EW (19.8-23.3) (36,5-57.3) (56.5-58.3) n=119 Mean egg weight Mean egg weight from 22 clutches from 4 clutches 16 BV cu. cm 2195.38 10835.16 16443.0 17 EM sq. cm g 127.33 1045.12 890.01 18 EV sq. cm 77.91 511.58 454.62 19 P(,% 5.79 9.64 5.41 20 P2% 3.5 4.7 2.7 57 JOURNAL, BOMBAY NATURAL HIST. SOCIETY, Vo!. 84 Table 2 Major nesting sites of Kachuga sp. identified in national ciiambal sanctuary, 1983-84. d, K. dhongoka\ k, K. kachuga and t, K. tentoria circumdata. for tentoria, ** first nesting, ***second nest- ing. al, alluvium; bs, bush and os, open sand. SI. No, Location Reference km * River bank (M.P./U.P./ Rajasthan) Site Description Shore Nature of length m nesting ground Species-wise use Dates of nesting data collection 1 Baroli 57 M.P. 500 os/bs t,d,k 1.2.84/3.4.84 *** 2 Rahu 113 M. P./Raj. 20/100 os t 22.11.83/24.11.83** 3 Batesura 123 M.P. 100 OS t 24.11.83** 4 Bharrah 131 M.P. 500 OS t,d 6.3.84/11.3.84/ 25.3.84*** 5 Devgadh (Sarsaini) 165 Raj. 250 OS t 3.11.83** 6 Basai Dang 173 Raj. 500 OS t 30.10.83** 7 Gorkha 205 Raj. 100 OS d 5.4.84 8 Tigri- Rithaura 209 Island (M. P./Raj.) 20 sq. m al d 29.3.84/5.4.84 9 Papripura 214 M.P. 2000 os d 5.4.84 10 Babu Singh ka Gher 226 M.P. (Islet) 500/100 sq. m os t,d,k 4.12.83/26.1.84*** 30. 1 . 84/31 .1.84 11 Pureini 229 Raj. 1000 OS t,d 5.12.83/3.1.84*** 31.3.84 12 Kenjra 296 U.P. 1500 os/bs t,d 31.1.84*** 8.4.84 13 Gyanpura 356 M.P. 200 os t,d 2.3.84*** 13.4.84 * In reference to Palighat (Parbati-Chambal confluence) (Fig. 1). December (10), 3 January (1) and 28 Janu- ary (1). After 28 January no further nesting activities were noted. From all the above nest- ing records, it was presumed that nesting season of Kachuga tentoria circumdata extend- ed from October through January (Fig. 2). A female Kachuga tentoria circumdata (25 cm carapace length) wandering on land at 0730 hrs on 24 November contained six (3+3 right and left) oviductal eggs with a mean weight of 23.3 g and length x width 50±0.564 (50.0-51.5) x 29.08±0.205 (29.0-29.5) mm. The ovaries contained only six fresh corpora lutea (right 4 and left 2). Four ovarian follicles of ovulatory size (25-26 mm) were also present suggesting the pro- duction of at least two clutches per season. On the same day (24 November) at 1600 hrs, a female turtle (carapace length 23.0 cm) was captured on land when it was returning after egg laying. The nest contained 7 eggs with the mean measurements of 46.0 ±0.925 (45.0-47.0) x 28. 0_ 0.534 (27.0-29.0) mm. Mean egg weight was 22. 3 ±0.527 (22.0- 58 NOTES ON THREE SPECIES OF KACHUGA Fig. 1. River Chambal (National Chambal Sanctuary) from Pali (Chambal-Parbati confluence) to Pachhnada showing important nesting areas of Kachuga sp. (1-13, as in Table 2). M.P., Raj. and U.P. are the states of Madhya Pradesh, Rajasthan and Uttar Pradesh through which the river flows. Inset: Position of River Chambal (arrows) in the Gangetic system. JOURNAL, BOMBAY NATURAL HIST. SOCIETY, Vol. 84 60 Fig. 2. Relationship among climatic conditions and breeding activities in small (S) and large (L) Kachuga turtles. N, nesting; H, hatching. Horizontal scale shows months from August (A) 1983 through July (J) 1984. Maximum tempera- ture (°C) shown as dotted line and minimum temperature (°C) as solid line through hollow circles. Hour of sunshine (hr) : dotted line through solid circles, relative humidity (%) : solid line through hollow triangles, and rainfall (mm) : solid line through crosses. NOTES ON THREE SPECIES OF KACHUGA l J J_J L _J 5SSSS58! I r-ssai G\ 6 2? ** <4-1