POPULATION SIZE AND STRUCTURE AND HABITAT RELATIONS OF THE BARASINGHA (CERVUS Q, DUVAUCELI) IN SUKLA PHANTA WILDLIFE RESERVE, NEPAL By Cari Dietrich Schaaf A DISSERTATION Submitted to Michigan State University in partial fulfiiiment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Fisheries and Wiidiife 1978 ABSTRACT POPULATION SIZE AND STRUCTURE AND HABITAT RELATIONS OF THE BARASINGHA (CERVUS D. DUVAUCELI) IN SUKLA PHANTA WILDLIFE RESERVE, NEPAL By Carl Dietrich Schaaf Population organization and habitat relations of the bara- singha (gervu§_gL duvauceli) were studied between April 1974 and May 1976 in Sukla Phanta Wildlife Reserve, Nepal. Total population counts in March when barasingha were most concentrated on southern grasslands in the reserve yielded 805 ani- mals in 1975 and 908 in 1976. Aerial surveys confirmed barasingha distribution, and showed that they frequented Indian territory bordering the reserve before human settlement began there in 1975. Fecal-pellet counts in March 1976 gave estimates of 1,295 bara- singha, 273 hog deer (Axis porcinus) and 341 chital (Axis axis). For barasingha, this was an estimate of maximum population size because they were concentrated on the grasslands. Hog deer and chital numbers represented only the grassland populations, since both species were more widespread in March. The population comprised 16.4-32.7% stags, 35.2-48.6% hinds, 7.1-16.1% yearlings and 9.20-7.7% fawns. An apparent 12.8% increase in minimum population occurred in 1975-76. The large proportion of Carl Dietrich Schaaf hinds indicated that potential population increase was greater than observed. Comparison with Kanha National Park barasingha data suggested that mortality and not poor breeding success caused the low increase. The population before 1968 and l972, when flash floods drowned many animals reportedly was larger than that observed in 1975-76. Habitat thus may be available to support more bara- singha before carrying capacity is reached, and further population increases can be expected. Habitat surveys during the year indicated that barasingha remained mostly on dry grasslands while forests and savannas were avoided or used to an intermediate degree. Hog deer showed greater preference for seasonally-wet grassland, while chital mostly fre— quented forests and savannas. Fecal-pellet distributions indicated that lowland grasslands and savannas were frequented similarly by the three deer during the premonsoon season, when environmental conditions and young forage grasses caused habitat preferences to be less defined. The grasses Imperata cylindrica, Narenga porphyrocoma, Phragmites karka, Saccharum bengalense and Saccharum spontaneum were eaten by barasingha. The relative abundance of these species on the southern grasslands, and availability of water there during the dry season, influenced habitat selection. “Increaser” plant species found on village grazing grounds were absent on southern reserve grasslands, despite local heavy grazing by livestock. A high water table promoted grass growth, Carl Dietrich Schaaf helping to maintain carrying capacity even during the dry season, while flooding, water-logged soils and fire maintained grassland habitats. Barasingha avoided livestock on southeastern reserve grass- lands during the premonsoon season. As livestock grazing declined there annually, drinking water and fresh grass attracted barasingha to the areas vacated. Elimination of livestock from these grass- lands will free additional habitat for deer. Aside from the annual congregation of the largest herds in March, barasingha in Sukla Phanta moved little in response to seasonal change. Some 32 km2 of lowland grassland, savanna and marsh supported the p0pulation year-round. Future increases in barasingha numbers in less-preferred habitats could indicate dis- turbance on the southern grasslands, or that carrying capacity there had been exceeded. Recommendations for barasingha conservation include con- tinued annual population counts and investigation of additional specified research questions. Disturbance in the form of grass- cutting and gathering in the reserve should be eliminated. Re-establishment or introduction of barasingha into other reserves should be considered. An international reserve to include Sukla 2 Phanta and 10 km of adjacent Indian territory is proposed. ACKNOWLEDGMENTS Many persons helped to plan and execute this study, and their names and contributions are too numerous to list all of them here, though I am grateful to each one. I would like to express my gratitude especially to the following individuals: the members of my Guidance Committee, Drs. George A. Petrides (Chairman), Leslie W. Gysel, Rollin H. Baker, and Victor J. Rudolph; to Mr. and Mrs. Kirti M. Tamang, Col. (ret.) and Mrs. Hikmat B. Bisht and their family, Mr. Arjan Singh, Mr. and Mrs. Balram Singh, Dr. Robert Fleming, Sr., Dr. and Mrs. Robert Fleming, Jr., Mr. Clifford Rice, Mr. Will Weber, Dr. Irving Wyeth, Dr. Collin W. Holloway, Dr. Charles E. Cress, Dr. Stanley Zarnoch, Dr. Gary A. Dawson, Mr. John Blower, Mr. and Mrs. Frank Poppleton, and Maj. (ret.) B. Chhetri. I also thank the numerous officers and staff of the Office of National Parks and Wildlife Conservation, and of the Forest Ministry, His Majesty's Government of Nepal, for obtaining permission for me to do the study, and for their subsequent cooperation and assistance. I am grateful, too, for the use of a United Nations aircraft and the services of pilot G. Fuerer in making aerial deer counts. For generous financial support of the study, I am grateful to the World Wildlife Fund (U. S. and International Appeals), the New York Zoo- logical Society, the Zoological Society of San Diego, the ii San Francisco Zoological Society, and the Fauna Preservation Society. Special thanks go to my mother, Mrs. M. Schaaf, who pro- vided extra personal financial aid when it was most needed, and to my mother-in-law, Mrs. J. Le Blanc, who gave us a home in which to live and work once the field study was completed. Perhaps most important of all, I owe a debt of gratitude to my wife Cherie, who assisted me ably in the field and afterwards, and whose unfailing patience and encouragement were always a source of inspiration. iii TABLE OF CONTENTS LIST OF TABLES LIST OF FIGURES . INTRODUCTION . Extant Wild Barasingha Populations . . The Northern and Southern Subspecies of Barasingha Barasingha in Nepal . Previous Studies . Objectives of the Present Study THE STUDY AREA Location History Climate . Topography and Soils Fauna . THE HABITAT Vegetation Types Sal Forest. Sal Savanna Mixed Deciduous Forest. Khair-Sissoo Forest. Lowland Grasslands Influence of Fire in the Habitat Availability of Water . POPULATION SIZE . Methods . Aerial Surveys Ground Counts . . . Fecal-Pellet Counts . iv Page vii ix Results . . Aerial Surveys and Ground Counts Fecal- Pellet Counts . Discussion REPRODUCTION . Timing of the Rut and Distribution of Rutting Stags . Methods . Results . Discussion . . Age of Hinds at First Breeding. Fawning . . . . POPULATION SEX AND AGE STRUCTURE Methods Results Discussion . . Realized and Potential Population Increase . FACTORS AFFECTING SURVIVAL Poaching Predation . Diseases and EctOparasites Floods . . HABITAT PREFERENCES Methods Deer Counts . . Fecal- Pellet Counts . Results . Deer Counts . . Fecal- Pellet Counts . Discussion FACTORS INFLUENCING HABITAT SELECTION Feeding Habits Methods . Results . Discussion . . Effect of Livestock Grazing on Habitat Use by Barasingha . Effects of Grazing and Fire on the Grasslands of the Southern Phantas . . . Page Page CONCLUSIONS . . . . . . . . . . . . . . . . 81 RECOMMENDATIONS . . . . . . . . . . . . . . 84 SUMMARY . . . . . . . . . . . . . . . . . 88 APPENDIX A Mammals . . . . . . . . . . . . . . . 95 B Ectoparasites and Biting Flies . . . . . . . 98 C Grasses and Sedges . . . . . . . . . . . 101 LITERATURE CITED . . . . . . . . . . . . . . 105 vi Table 10. 11. LIST OF TABLES Barasingha numbers estimated from aerial surveys made in Sukla Phanta Reserve, Nepal, May 1974 and March 1975 . . . . . . . . . Population estimates of three deer species derived from fecal-pellet counts on Sukla and Seta Khera Phantas, February-April l976, Sukla Phanta Reserve, Nepal . . . . . . . . . . . . . . . Morning and evening rutting calls counted at three listening post, compared with morning counts made simultaneously in Sukla Phanta proper, November 1974, Sukla Phanta Reserve, Nepal . Numbers (and percentages) by sex and age of bara- singha tallied in Sukla Phanta Reserve, Nepal, and Kakraha Block, Dudhwa National Park, India . Occurrence of three deer species in six habitats, March 1975, Sukla Phanta Reserve, Nepal . Occurrence of three deer species in six habitats, Why-June 1975, Sukla Phanta Reserve, Nepal . Occurrence of three deer species in six habitats, Aug.-Sept., 1975, Sukla Phanta Reserve, Nepal . Occurrence of three deer species in six habitats, Dec. 1975, Sukla Phanta Reserve, Nepal Occurrence of three deer Species in six habitats, March 1976, Sukla Phanta Reserve, Nepal . Summary, seasonal occurrence of three deer species as observed year-round in six habitats, Sukla Phanta Reserve, Nepal Results of Kruskal-Wallis comparisons to detect sig- nificant differences in fecal pellet numbers from three deer species in dry grassland, seasonally-wet grassland, and lowland savanna, February-April l976, Sukla Phanta Reserve, Nepal . . . . . vii Page 30 33 41 47 57 58 59 60 61 62 64 Table 12. 13. B-1. C-1. Frequency of occurrence and estimated average percent cover of 23 common grasses and sedges in 116 plots in Sukla Phanta Reserve, Nepal . . . . . . . Grazing incidence on several grasses and sedges by three deer species in Sukla Phanta Reserve, Nepal, during the premonsoon and monsoon seasons, 1974-76 Mammals noted during barasingha study in Sukla Phanta Reserve, Nepal . . . . . . . . Ectoparasites and biting flies collected in Sukla Phanta Reserve, Nepal . Grasses and sedges collected in Sukla Phanta Reserve, and near Reserve Headquarters, Nepal . viii Page 68 69 96 99 102 Figure LIST OF FIGURES Sukla Phanta Wildlife Reserve, Nepal . Average monthly temperature and rainfall, 1974-75, reserve headquarters, Sukla Phanta Reserve, Nepal Vegetation types, Sukla Phanta Reserve, Nepal . Ground counts of Barasingha during the premonsoon seasons, 1974-76, Sukla Phanta Reserve, Nepal . . (a) Morning and evening hourly rutting call rates in Sukla Phanta proper, 1974- 75, (b) Percentage of total calls heard from various directions in Sukla Phanta proper, 1974-75, Sukla Phanta Reserve, Nepal Species-area curve for grasses and sedges collected in Sukla Phanta Reserve, Nepal, during October- November 1974-75 . . . . Avoidance by barasingha of areas used by livestock and graziers: (a) barasingha numbers on central and western Seta Khera Phanta and on Sukla Phanta com- pared with numbers of eastern Seta Khera Phanta February- May 1975, and (b) livestock numbers on eastern Seta Khera Phanta, February-May. 1975, Sukla Phanta Reserve, Nepal . . ix Page 10 14 16 31 39 69 76 INTRODUCTION Extant Wild Barasingha Populations The barasingha or swamp deer (Cervus duvauceli) is one of the world's endangered large mammals (IUCN Secretariat, 1976). Native to the northern and central portions of the Indian subconti- nent, the species has declined drastically in recent decades due to the loss of habitat and other pressures exerted by an ever- increasing human population (Schaller, 1967; Schaller and Simon, 1970; Holloway, 1973; Singh, 1973; Schaaf and Singh, 1977). The largest populations still extant in the wild are in three reserved areas in India and one in Nepal. In sequence accord- ing to declining estimated sizes of barasingha populations, these are Dudhwa National Park in Uttar Pradesh, Sukla Phanta Wildlife Reserve in southwestern Nepal, Kaziranga National Park in Assam, and Kanha National Park in Madhya Pradesh. The estimated minimum popu- lation of Dudhwa is 1,200 animals (Holloway, 1973; Schaaf and Singh, 1977), while that of Sukla Phanta is approximately 1,000 (present study). Lahan and Sonowal (1973) reported 520 head in Kaziranga, and it is thought that the population there has since increased (K. Patar, personal communication). Panwar (in press) reported an estimated minimum population of 283 animals in Kanha, where restora- tive measures have resulted in an increase from the lowest population size of 55 animals counted by Schaller in 1965. 1 The Northern and Southern Subspecies of Barasingha Barasingha are divided into two subspecies (Ellerman and Morrison-Scott, 1951), Cgrvu§_g, duvauceli Cuvier 1823 and Cervu§_ g; branderi Pocock 1943. The latter occurs in east central India north of the Godavari River, and today is confined for all practical purposes to the Kanha National Park. The northern subspecies is represented by larger populations and occurs mainly in the other reserves, while a few remnant populations of uncertain status sur- vive outside the reserved areas. The former range of this sub- species includes the alluvial plains of the great river systems of north India, the Indus, Ganges and Brahmaputra. The naturalist Dunbar-Brander (1927) first drew attention to the possibility of there being two subspecies. Pocock (1943) dis- tinguished between them on the basis of museum specimens he examined and Dunbar-Brander's descriptions of animals observed in north and central India. It was noted that the northern deer are more nearly true swamp dwellers since they inhabit alluvial plains which are at least seasonally wet, while the southern animals live in the central Indian highlands on dry ground, though in the vicinity of water. Dunbar-Brander reported that the hooves of the latter population are ". . . hard and well knit, (like those) of (animals) accustomed to gallop on hard ground. . . ." The northern deer are said to have splayed, “spongy" hooves, the better to support them in their wet habitat. The northern deer also were judged to be larger and paler in color than their southern cousins. Measurements of ten skulls (seven from north India and three from central India) made by Pocock tend to confirm the size difference, yet it is noteworthy that the largest antlers were carried by specimens from central India, the biggest measuring 41 inches (104 cms.) around the out- side curve (Dollman and Burlace, 1922). General descriptions of the barasingha have been given by Schaller (1967) and Prater (1971). A large stag may weigh 170- 180 kgs. and stand 135 cms. at the shoulder (Prater, 1971). Pub- lished weights of hinds (Crandall, 1964) include those of two captive animals which weighed 305 and 320 lbs. (138 and 145 kgs.). Perhaps the heaviest specimens on record were two stags of 563 and 588 lbs. (256 and 267 kgs.) reported by the Maharaja of Cooch Behar (1908). ' During the warm months (April-September), barasingha have sleek, reddish coats in which pale spots are sometimes evident. These spots are not composed of white hair, but appear rather like watermarks on stationery, as Dunbar-Brander (1927) noted. During the cool season (October-February) the pelage is greyish-brown and shaggy, and longer on the neck. Mature stags are noticeably darker at that time of year than other members of the population. The term "barasingha" means "l2-horned" and refers to the 12-tined antlers typically carried by mature males. However, many stags have a larger or smaller number of tines, the usual range being 10-14. "Barasingha" is also a vernacular name in India for the Kashmir stag (Cervus elaphus hanglu , and this common name thus may be a source of confusion when one reads the older literature on these different species. Barasingha in Nepal Barasingha inhabit the terai, an extension of the north Indian Gangetic Plain, which fringes the southern edge of otherwise- mountainous Nepal. As late as the 19505, barasingha were still wide- spread in the districts of Banke, Bardia, Kailali, and Kanchanpur, or approximately the western third of the terai from about 25 kms. east of Nepalganj to the western border with India, just beyond the Mahakali River (K. M. Tamang, personal communication). In 1957, barasingha were also reported "in large numbers" in the Chitawan Valley of south-central Nepal. They were found in the then-extensive marshy grasslands north of the Rapti River, just outside the area which is now the Chitawan National Park (K. M. Tamang, personal communication). According to Schaller (1967), a few animals survived in that valley as late as 1963. None occurs there today. The grasslands where they formerly lived have been settled and are now under cultivation. At present three barasingha populations are known to survive in Nepal, each isolated from the others by distance and/or agri- cultural development. Dinerstein (1976) reported "no more than 20" 2 animals from the 348 km Karnali-Bardia Wildlife Reserve in Bardia District. Some 110 kms. further west a population of unknown size 2 and status exists in the 37 km Dhaka Shikar (Hunting) Reserve in Kanchanpur District. About 4 kms. west of Dhaka, Sukla Phanta Wildlife Reserve (147 kmz) harbors about 1,000 animals, the largest population in Nepal, and the subject of the present study. Previous Studies Schaller (1967) obtained the first systematically collected data on the biology of the barasingha, working with the southern form in 1963-65 in the Kanha National Park. In addition to his observations on ecology and behavior, he summarized much of what had been written by earlier observers, mainly sportsmen-naturalists from the days of the British Raj, about the distribution, habits, and natural history of both subspecies. Schaller's study provided a basis for further research, and drew attention to the precarious status of the southern barasingha and the continuing decline of the northern deer. In 1969 an IUCN Study Group briefly visited Kanha to assess the proposed construction of a 74-acre protective enclosure in which a small breeding herd of barasingha could be established, free from predation and other disturbance. It was recommended that a qualified biologist oversee the project and further investigate barasingha ecology in the park (Binney et a1., 1971). Martin (1975) succeeded Schaller, studying barasingha ecol- ogy in Kanha during 1971-73. He concluded that the principal causes for the decline of this subspecies were loss of habitat and poach- ing, coupled with the deer's need for seasonal migration to satisfy requirements for food, water, and suitable fawning grounds. During Schaller's study, barasingha were observed primarily on the Kanha Meadow, where they spent half the year, but from which they dispersed early in the monsoon season until the following winter. Martin concluded that their seasonal range extensions took them outside the park where they conflicted with man. He suggested that disturbance during the October fawning season, rather than disease, as suggested by Schaller, could have caused the low propor- tion of fawns observed in 1964-65. An addition of land to the park in 1964 took in much of the northern monsoon-season range of the deer. In 1969 a forest village north of the Kanha Meadow was removed, and livestock grazing was eliminated from nearby clearings. By the early 19705, the former village site and nearby meadows had become important monsoon-season range and fawning grounds for barasingha (Martin, 1975). Securing additional habitat was one factor which improved conditions for the deer. In addition, controlled burning of grasses on the Kanha Meadow after the monsoon season increased the forage available during the dry season when chital (Axis_axj§) (Appendix A) and barasingha congregated there near water. Controlled buring also permitted certain perennial grasses to become re-established, since they were not as heavily grazed as when the entire meadow was burned annually, and ungulates concentrated on the subsequent flush of new growth. The practice of baiting tigers into the area had induced disproportionately high predation on the already-small barasingha herd. This practice was eliminated and tigers began to kill more chital. In response to these measures, the barasingha population expanded, and continues to grow and occupy new areas in the park at present (Panwar, in press). Holloway (1973), writing of the northern barasingha in Uttar Pradesh, noted that of 11 localities mentioned by Schaller in 1967 as having known or probable populations, eight could be struck from the list for all practical purposes by 1973. He indicated that an etho-ecological study was an immediate conservation need. This recommendation and subsequent proposals for a world-wide threatened deer conservation program (Cowan and Holloway, 1973, 1974) provided a rationale for the present study, which extended from April 1974 to May 1976 and was the first effort to collect data on the northern subspecies. Objectives of the Present Study As originally conceived, the study would have involved the two largest remaining populations, in Sukla Phanta Wildlife Reserve and the nearby Dudhwa National Park. The logistics involved and formal arrangements required to travel repeatedly between Nepal and India made this an impractical plan. The study was limited to the Sukla Phanta Reserve, with only a few comparative observations made in Dudhwa. The principal objectives were (1) to initiate collection of basic ecological data on the northern subspecies, particularly with respect to population size and structure, (2) to use these and environmental data to assess the current status of the deer and factors affecting its welfare, and (3) to make such recommendations as might be needed to improve conservation of the barasingha through herd and habitat management. THE STUDY AREA Location Sukla Phanta Wildlife Reserve is situated between 28°49' and 28°57' north latitude and 80°07' and 80°15' east longitude in the Kanchanpur District of southwest Nepal. For approximately 6 kms., its southern boundary coincides with the international border between Nepal and India. It is bounded on the west and along part of its southern boundary by a large tributary of the Ganges, the Mahakali River, or Sharda River, as it is called in India (Figure 1). The term "phanta" means a grassy plain, or a grassy opening in the forest. Sukla Phanta is one of the largest grasslands in the area, and gives the reserve its name. History The aboriginal Tharus, agriculturalists and livestock herders (Bista, 1967), were until recently almost the sole inhabitants of the remote western terai, including what is now the reserve (local informants). During the relatively dry and cool winter months, the rulers of Nepal sometimes visited Kanchanpur District to hunt (Smythies, 1942), as the forests and grasslands of this region were noted for an abundance of big game. 10 Figure 1.-—Sukla Phanta Wildlife Reserve, Nepal. 11 MAHEgDRA NAGAR 0 Fun. vllloouih ' -..- m border “ohgkdll 0, —— WW '0, 12 In 1965-66, by decree of the late King Mahendra, some 131 km2 surrounding Sukla Phanta became a Royal Shikar (Hunting) Reserve. Fair-weather tracks were cleared to facilitate the hunt and several villages were removed from the new reserve and the people resettled elsewhere (local informants; Singh, 1966). The eroded dikes of rice paddies, old plow furrows, groves of village mango (Mangifera indica) trees and other signs of former habitation still can be found. Cultivation and livestock grazing have influ- enced the condition of the habitat in and near Sukla Phanta for at least the past half century. Singhpur Village, for example, was founded in 1927 and still exists on the eastern boundary of the reserve (Figure 1). In 1976 Sukla Phanta became a wildlife rather than a shikar reserve, and was added to a newly created system of national parks and nature reserves in Nepal. With a few alterations to its boundaries, it stands today much as demarcated in 1965-66. Addi- tional fair-weather tracks have been cleared, and a permanent bridge built in the early 19705 across the Bawani River provides vehicular access to the southern part of the reserve from headquarters near Signhpur Village during the dry season. A special unit of the Nepal Army was assigned to protect the reserve early in 1976, replacing the forest guards formerly responsible. Administration and develop- ment are the duties of a warden and his staff. 13 Climate No earlier climatological records were available for the study area, but records of temperture and precipitation at reserve headquarters were kept during the study. The highest daytime tem- perature recorded was 48° C on June 14, 1974, and the lowest night- time reading was 2° C, on December 11 and 12, 1974, and January 26, 1975. The largest single rainfall of 128.5 mm occurred on July 12, 1975. The premonsoon hot season or summer lasted from mid-March to mid-June and brought steadily-increasing temperatures with very little rain (Figure 2). Except in some permanent streams and marshes, surface water gradually evaporated. The onset of a steady southeast wind bringing moist air from the Bay of Bengal signaled the beginning of the monsoon (Hagen, 1961; Ramdas, 1974). In Kanchanpur District monsoon rains began on about June 17, 1964, and June 23, 1975. Approximately 94 percent of the precipiation recorded at headquarters fell during the monsoon seasons. In September-October the weather cleared and gradually cooled. Temperatures dropped into the cold season or winter and were lowest in December-January. Occasional light frosts were not unknown, though none occurred during the two winters of the study. Winter rains were infrequent and usually light. After mid-February, cool weather slowly gave way to premonsoon heat. 14 3.552 .m>¢mmu¢ (5.25....— (Av-3m .mzmbcsacn—(m: m>¢mmw¢ .chuzbw r3352: 024 wEfikcESNP hath—‘0! m0<¢m>< .N 550.“. 9.3.20: 0 2 O m < a .4 I < 2 m 6 ms] 10 m? 19. u... 1 1» mNN w M m a m 8? 1cm w. n L H N. a ._._<...z_<¢ .m 2.0. . -mu 1 I m . . Go «in» 2.5 .2... one. . -8 I . . mimh wo<1m i=2 . I ma? . 15.0 min». wogm .x<5 003 o i o 13 came. 2:95;. . . a. . 01.10 m3- :3 15 Topography and Soils The reserve is on a flat, gently undulating plain (elevation ca. 100-150 m above sea level) lying at the base of the Siwalik Hills, and leping gradually south. Soils of the north-Indian plains and Nepalese terai are alluvial, and were laid down by the great north-Indian river systems and their numerous tributaries (Raychaudhury and Rajan, 1971; Mani, 1974). The alluvium was brought down from the Siwaliks and the Himalayan ranges where the large rivers arise, and varied from fine silts and clays to pebbles or larger rocks. This material was deposited as the rivers meandered across the land, and the process continues today. Mani (1974) stated that the older or Pleistocene alluvium called bhangar occupies higher ground than the newer khag§r_which grades into the most recent delta silts. Champion and Seth (1968) noted that sal (Shorea robusta) forests on the plains tend to be confined to the older alluvium. In this study, sal forest in the northeast part of the reserve was seen to be on higher ground than the grassland and mixed deciduous forest to the south. The break in elevation between these two kinds of vegetation was often sharply defined, and ran roughly from southeast to northwest, from the north side of the Singhpur Village fields, along the north bank of the Bawani River and the east side of Sunderi Phanta, and thence north of the Balma-Barkola grasslands (Figure 3). In general, with the exception of seasonal streams and depressions, sal forest soils in the reserve appeared to be sandier and better drained than the 16 Figure 3.--Vegetation types, Sukla Phanta Reserve, Nepal. ‘———‘ c ——_ s g— 521—: 3 fig; :3 ——g—_ 2 1,11 l III 1 "31'1"."- r1111 11 A” k 1 7 I W a ’- —_— 17° 2": N. y \ Ac, \E.>+m!£$l— IS / .l I: I U? 3 ’1 J 5‘ NORTH l m“ 1.5 ME" 1 E w M" M m W 5.00 Paul W “I”. [We M W um. ho .. 1m 18 heavier southern-grassland soils. However, areas of sandy soils occurred also in the southern grasslands, notably in Sukla and Karaiya Phantas. £2222. Several large mannals have disappeared from the reserve in the past 5-20 years, according to local informants. These included blackbuck (Antilope cervicapra), four—horned antelope (Tetracerus guadricornis), dhole or wold dog (Cuon alpinus), wild (Canis lupus), and striped hyena (Hyaena hyaena). Wild water-buffalo (Bubalus bubalis) also were present but apparently disappeared more than 20 years ago. The western terai also lies within the former range of the Indian rhinoceros (Rhinoceros unicornis) (Gee, 1964), but since it is not remembered by local villagers, this species probably has not occurred in or near the reserve for many years. Apart from mammals (Appendix A) and birds (to be reported in a separate paper), the various species of other groups of animals were not recorded; yet an abundant and varied fauna was observed in the reserve. THE HABITAT Vegetation Types Observations on vegetative physiognomy and species composi- tion made during the study were the basis for distinguishing eight vegetation types. Sal forest and savanna were found on the upland portion (80.5 ka) of the reserve, while the remaining six types were on the lowlands to the south and west (Figure 3). Aerial photos (1:12,000, made during a USAID Forest Resources Survey in 1964) provided information on the extent of large blocks of vegeta- tion which could not be surveyed on the ground. Sal Forest This type occupied the largest continuous area in the reserve, ca. 70.4 kmz. Unlike the mixed deciduous forest to the south, sal forest was more extensive than associated grasslands, and thus included them. Sal forest generally is characterized by the dominance of Shorea robusta, which is highly gregarious and may form almost pure stands (Champion and Seth, 1968). Other common sal forest trees in the reserve included ngerstroemia parviflora, Emblica officinalis, Terminalia belerica, I, algtg, Syzygium cerasoideum and Wrightia tomentosa. Some stands had closed canopies and little ground cover other than fallen leaves, while elsewhere trees were more widely 19 20 separated with an understory of grasses, including Eulaliopsis binata, Narenga_porphyrocoma, Themeda arundinacea, Saccharum bengalense, S, spontaneum,0esmostachya bipinnata and Eulalia spp. Sal Savanna According to Whyte (1974), savannas are generally understood to be grasslands containing some woody vegetation. For purposes of this study, grassland in which the tree canopy cover was about 10— 30% (ocular estimate), and/or including low—growing woody vegetation such as coppicing trees or shrubs, was classified as savanna. Sal savannas comprised roughly 10.1 km2 of the uplands. A small part was seasonally wet, principally where seasonal streams flowed. Tree and grass species on the savannas were similar to those in the sal forest. Sal trees at various stages of development in virtually all the drier savannas suggested that they were part of a seral con- tinuum between the mature Shorea robusta forest climax (Champion and Seth, 1968) and more open grassland resulting from disturbance of the forest, mainly by fire and grazing. Many of the large upland phantas contained extensive stands of coppicing sal no taller than the grasses, and subject to annual burning which presumably pre- vented advanced growth. Mixed Deciduous Forest Mixed deciduous forest occupied about 20.5 km2 on the 2 66.8 km lowlands. Typically, mixed deciduous forest does not 21 contain sal as a dominant species (Champion and Seth, 1968), and sal was entirely absent from this forest in the reserve. Ground cover was variable, consisting only of fallen leaves where the canOpy was dense. Elsewhere broad-leaved plants such as Ageratum sp. or ferns predominated, and grass was found infrequently. Common tree and shrub species were Trewia nudiflora, Syzygium cuminii, Celtis australis, Ficus spp., Bombax ceiba, Albizzia sp., Cedrela toona, Holarrhena antidysenterica, Mallotus philippinensis and Murraya koenigii. The trailing, spiny palm Calamus tenuis was common in wet places. Khair-Sissoo Forest Khair (Acacia catechu) and sissoo (Dalbergia sissoo) are pioneer species which thrive on unstable riverine sites (Champion and Seth, 1968). Sissoo formed almost pure stands on the banks and gravel bars of the Mahakali River. Understory grasses included Narenga porphyrocoma, Apluda mutica and Phragmites karka, among Species identified. While sissoo was limited mainly to the banks of the Mahakali, khair was scattered about the grasslands on moist sites, and occasional specimens occurred in the mixed forest. Khair- sissoo forest occupied only about 4.7 km2 in the southwest corner of the reserve. Lowland Grasslands Most of the reserve west and south of the Bawani River and south of the forest bordering the north edge of Sukla Phanta proper 22 was grassland (Figure 3). It was divided into the four types described below. Grasslands outside the principal southern phantas were not known well enough to classify them similarly. Lowland grasslands covered about 38.5 kmz. Dry Grassland.--Dry grassland was only seasonally well- watered. Soils in some places were sandy, facilitating percolation of water and drying of surface layers. Central Sukla Phanta proper, west Seta Khera, part of east Seta Khera bordered on the north by marsh, and parts of Karaiya Phanta met these criteria. Grass heights varied, depending on species and enviornmental conditions. Dominant grasses were Narenga porphyrocoma, Imperata cylindrica, Saccharum bengalense, S, spontaneum and Desmostachya bipinnata. Dry grassland occupied an estimated 9.9 kmz. Seasonally-Wet Grassland.--This type lay at lower elevations where it was flooded by the Bawani River and monsoon rains, and where topography and soils favored retention of surface water. Some areas remained marshy for up to 6 months after the rains ended. Grasses reached heights of 150 cms. or more. Dominant species included Narenga porphyrocoma, Imperata cylindrica, Saccharum spontaneum, Vetiveria zizanioides, and in eastern Seta Khera, Sclerostachya_fusca. Sedges of the genera Cyperus and Scleria also were found. Northeast Sukla, Bameria and most of Seta Khera and Sunderi Phanta supported wet grasslands. About 8.8 km2 in or adjacent to the main southern phantas were covered by this type. 23 Lowland Savanna.--This savanna extended over an estimated 9.5 km2 on the southern phantas, including portions of Sukla and Seta Khera and most of Karaiya. Savanna occupied dry and seasonally- wet sites, but was found primarily on the former. The main tree species represented were Butea monosperma, Acacia catechu, Dal- bergia sisso (in Karaiya Phanta), Ficus religiosa, Bombax ceiba, Cedrela toona and Sterculia villosa. Grasses were the same species mentioned above. Marsh.--Marsh was characterized by the presence of surface water the year round, or by soils which were water-logged and muddy all year. Dense Phragmites karka, Saccharum spontaneum and Sclerostacbya fusca comprised the dominant vegetation. An estimated 3.1 km2 were covered by marsh east of Sukla Phanta and north of eastern Seta Khera. Additional smaller areas were on the northeast edge of Karaiya Phanta, in south-central Sunderi, and in the Khair- sissoo forest. Influence of Fire in the Habitat Fires annually burned over much of the reserve. Probably all were man-made, and many were started by grass-cutters who entered the reserve each winter to collect thatching material. Some were started by passers-by, or by livestock graziers wishing to hasten new growth. The earliest fires were recorded in Karaiya Phanta on November 29, 1974, and December 4, 1975. By the end of December each year most of the south-central grasslands had burned, 24 although burning continued in peripheral areas and on the uplands until well into the premonsoon season. The first extensive fires in the sal forest and savannas were recorded on March 12, 1975, and about April 15, 1976. On occasion, fires started in India also may have swept across the southern border into the reserve. Within a few days after a burn, the first new grass appeared and soon attracted wild ungulates. By clearing senescent vegetation and stimulating new growth, fire encouraged the annual congregation of barasingha in large herds on the southern phantas. Chital and hog deer (Axis porcinus) also were attracted to the new grass. Burn- ing thus influenced the seasonal distribution of animals, in addi- tion to maintaining the grasslands by preventing forest regeneration. Availability of Water Water was abundant everywhere during and after the monsoon. Some of it remained in temporary ponds until late in the following dry season. Seven ponds were found in southern Sukla and adjacent Seta Khera Phanta. All but two were dry by May each year, and even these disappeared by the end of that month. After May, drinking water was to be found only in the Bawani and Mahakali Rivers, and in the marshes bordering the main phantas. The Chaudhar River dried during the last month of the hot season. Being located peripher- ally, this stream probably was of less importance to wildlife, in particular to barasingha, even when flowing. The Mahakali River, too, may have been a less important source of drinking water since it also is on the periphery of the reserve. Only once, in May, 1974, 25 tracks of two barasingha were found on the Mahakali where it adjoins Karaiya Phanta, and the same day a lone hind was spotted leaving the river bed for the grassland. Dew-falls were recorded in the dry season as late as mid- April, and these also provided water for grazing animals. Addi- tional moisture was present in the green forage, especially on the seasonally wet grasslands, where grass was kept fresh by the rela- tively high water table (Singh, 1966). In the sal forest and savannas, two ponds about 1.5 kms. northeast of Sukla Phanta also contained water into May. Permanent water was found on the periphery of the sal forest and savannas in Sunderi Phanta, the Bawani River and Rani Tal, a small lake on the east side of the reserve (Figure 1). Since grass fires occurred later in the season in the sal forest region, and because the water table was not as high as in the lowlands, less green forage was available before the onset of the monsoon rains than on the southern grasslands. Furthermore, leaf-drop by sal trees in March and the simultaneous growth of new leaves may have resulted in draw-down of the water table and enhanced drought conditions on the uplands, in a manner similar to that reported by Martin (1975) from Kanha National Park. POPULATION SIZE Methods Barasingha congregated annually to the greatest extent on the recently burned grasslands of Sukla and Seta Khera Phantas between February and mid-April. They then afforded the best Oppor- tunity to attempt total population counts. Both aerial surveys and ground counts were made. An additional estimate of population size was obtained from fecal-pellet counts. Aerial Surveys Aerial surveys were conducted when the United Nations plane based in Kathmandu became available, and for this reason, the first survey on May 2-3, 1974, was somewhat after the season of maximum deer concentration, while the second on March 26-27, 1975, took place at the appropriate time of year. Flights using the Swiss Pilatus Porter single-engine STOL aircraft were made on the after- noon of the first day after 1500 hrs., and the following morning before 1000 hrs. during each two-day period. Flights proceeded from the southeast corner of the reserve on a series of east-west, west- east transects. The courses flown also enabled views of about 10 km2 of Indian territory adjoining the southern border of the reserve and lying between the Mahakali River, where it turns briefly south before swinging back to Nepal, and the international boundary. 26 27 Transects extended between the Mahakali and Chaudhar Rivers, and continued as far as the northern tip of Sunderi Phanta. The same two observers participated in all four flights, one person recording barasingha from each side of the plane. The numbers of animals seen in large groups were estimated or partially counted and then esti- mated. With the straight-and-level flight pattern employed, observ- ers alternately viewed from the north and south sides of the plane, and duplicate sightings of deer were possible. Since sightings were not numerous, however, and the area covered was small (ca. 70 kmz), duplicate counts were eliminated by comparing the data recorded by both observers immediately after flights. When dupli- cates occurred, the larger figure was included in the count, and sightings of small groups by one observer overlooked by the other were added to the total. Ground Counts Ground counts were conducted along a standardized route which gave the greatest coverage to the areas frequented by bara- singha during the period of maximum deer concentration and visi— bility. Counts were made in the morning before 0900 hrs. and/or in the afternoon after 1600 hrs. An observer sitting on top of the cab of a vehicle driven slowly along the route tallied all deer seen, stopping carefully to count large herds with 7 x 35 binoculars or a 20x telescope. The route comprised the East-West Track, beginning where it enters Sukla Phanta proper, up to and including the Seta Khera Loop, and the North-South Track (Figure 1). In 28 addition, two short extensions were made, one to the south from the east end of the East-West Track and approximately 500 m long, the other to the southeast about 300 m just as the Seta Khera Loop turns onto the North-South Track. These extensions gave added cov- erage to east-central Seta Khera and southeast Sukla Phanta proper. Fecal-Pellet Counts Fecal-pellet counts were used to estimate the numbers of barasingha as well as hog deer and chital present on Sukla and Seta Khera Phantas during the 1976 premonsoon season. As noted previ- ously, all three deer species were attracted to new grass growth after winter grass fires, and they frequently shared common grazing grounds at that time of year. In February, 56 50-m2 plots were laid out at 300 m intervals along transects 500 m apart in Sukla Phanta and in western and cen- tral Seta Khera Phanta. All pellets were removed from them, and they were re-examined 60 days later. Lack of rain and little insect activity meant that most pellets dropped during the interim remained jg_situ and undamaged. Pellets were identified by size and shape (Eisenberg, et al., 1970). Those of barasingha tended to be the largest and were barrel-shaped. Chital pellets were narrower and longer relative to barasingha pellets, and hog deer pellets were smaller and more uniformly round than the others. Barasingha frequently defecated while walking, making it difficult to distinguish pellet groups, especially where pellet densities were high. Taber (in press) reported similar behavior 29 for captive chital. For this reason, total numbers of pellets for each species, rather than pellet groups (Bennett et al., 1940; Neff, 1968) were counted when plots were re-examined. Results Aerial Surveys and Ground Counts The May 1974 aerial survey yielded lower totals than the one in March 1975 (Table 1). Using the Spearman rank correlation coeffi- cient (Steel and Torrie, 1960), a high inverse correlation was found between the number of sightings and group size, i.e., the fewer groups seen, the larger the group size (rS = 0.800). A "group" or "herd" consisted of one animal or more. A significant correlation also existed between few groups and large group size during 1975 ground counts (rS = 0.708, p < 0.05) (Figure 4). A similar significant correlation between these variables was not found on ground counts in 1976, though, when rS = 0.346. During ground counts in both years, the highest total num- bers of deer were recorded in March, followed by declines in the latter half of April. In 1975, the abrupt drop from 800 to less than 200 animals between April 11 and 19 may have been caused by visiting royalty on April 13, when a helicopter and a large number of vehicles and personnel were active on Sukla Phanta proper, and could have scattered the herds. A more gradual decline in barasingha numbers was seen in 1976 in the absence of such disturbance (Figure 4). 30 .xec mzow>mea mg» mews» comm ago: some 0: use .xuopmm>wp An cam: »p_>mm; come we; umsm>ou awe» amen we» mocwm cmpmmamg we: mew: mpummcmsp “meaggmsasom moss» weer «mm oomnm m om 0mm sop nwmmw mm mwmfi mom oomumm v cup omm mp IWMMW mm mmm& avm ompn— om cup own op uwwmm me»m& mem e~_-e m. oN_ om“ o_ -mww” N¢»M& :mmm mNPm AEV muowwwmgh .eee .geem Memes seeee« .emmmw edges 2: .5. mnmp coca: new «map xqz e—mawz .m>mew~_ mpcwfia m—v—am :_. mtg: m>d>L3m Fwwme EOLm. UmHQEmpmm men—Ezc MSOCwmmsmmll; m4m<._. 31 "°°°l 1975 v FEBRUARY (1105) (aoo)(eoo) 2 231117." 000' o my C III Ill 0 u 600'" o E g 4004 ~20 2 IL 0 m o 5 E 2004 -10 In 1- : I a 52 2 U) — (9°81 v FEBRUARY 0 MARCH A APRIL 0004 C III Ill 0 IL 600- o E g 400- ~20 2 ll. ° 8 a: E 200- ~10 g 1- 3 6 3 a L , I 415202111 5 52530 DATES WHEN COUNTS WERE MADE FIGURE 4. GROUND COUNTS OF BARASINGHA DURING THE FREMON- SOON SEASONS, 1974-76, SUKLA PHANTA RESERVE, NEPAL. 32 Fecal-Pellet Counts Five of the 56 plots were lost during the 60-day interim because markers were removed by passers-by, or could not be relo- cated in grasses which had grown during that time. The total number of pellets for each species was divided by the average number of pellets per group, calculated from 195 complete pellet groups found (Table 2), to obtain the number of pellet groups represented by the total pellet counts. The results were converted to pellet groups per km2 and divided by 13, the assumed number of groups deposited per day. This figure was used by Eisenberg et al., (1970) to calculate chital numbers in the Wilpattu National Park of Ceylon (now Sri Lanka). The same figure was used for this study because it approximated defecation rates for several other deer species (e.g., McCain, 1948; Van Etten, 1959; Julander et al., 1963; Neff et al., 1965) and, in the absence of more precise data, was believed to be a reasonable estimate for barasingha and hog deer as well. Dividing again by 60, the number of days between clearing and re-examination of the plots, produced estimates of deer per kmz. These figures were multiplied by 12.3 kmz, that portion of Sukla and Seta Khera Phantas represented by the sample plots. Estimated population sizes were 1,295 barasingha, 273 hog deer, and 341 chital (Tab1e 2). Discussion The results obtained from aerial surveys were considered less accurate than ground counts because they were based in part on 33 «.5. m.mp x am mxeu om\dc xev\gmmu\mg:ocm umppma m_\mu mp.Nmm n Louuww cowmgm>coun Ne 83 u 30:. e S x 3 Se N mom._ N~.mmp eo.mpm.m Pm.moo.PNP .. 1. map mom.m~ mpepoh Fem mn.- -.mom._ ¢~.nem._m mp.n¢_ ¢_.m.H mm.~m mm o~¢.w peapcu ma SQ $.83 34%.: 3.5 3; H age mm 83 emwe : mm~.F Fm.mo_ om.mpm.o Fm.nmp.mm m¢.mo~ m~.e.H wn.mo mm mmm.mp msmcwm regmm mmEx Ao\emmmm epcmsa mpxzm .ommp Fweawemu mmmumam meu omega we mmumswumm cowumpaaoauu.m m4m<~ 34 estimates of groups size instead of complete counts. Moreover, estimates made by the two observers differed by as much as 14% for the largest groups. It was observed also that small groups of barasingha (less than 10-15) might not have been distinguished correctly in every instance from hog deer and chital when viewed from the plane, and this could have influenced survey results. These considerations did not apply to ground counts, where time could be taken to study animals with field glasses and to count all individuals. Aerial surveys tended to confirm the assumption derived from observations made on the ground, however, that barasingha were most concentrated on the grasslands of Sukla and Seta Khera Phantas dur- ing the mid-premonsoon season. None were seen from the air in March or May in the upland phantas north of Sukla Phanta pr0per. The relatively small number of animals seen during the May surveys indi- cated that the largest herds had broken up by then, and scattered smaller groups probably were overlooked more easily. During the May 3 survey, 21 animals were seen either on the Indian side of the southwestern border or close to it in Nepal, indicating that at the time, the deer may have used habitat avail- 2 Indian able in India. Human settlement had not begun in the 10 km territory mentioned above, but by March 1975 settlers were present and no deer were seen there. Although the absence of deer from this area might be expected when the animals had congregated on the grass- lands to the northeast, it is unlikely that they would have returned 35 to the then—settled Indian territory following the April-May break-up of the large herds. The inverse correlation between herd size and the number of barasingha herds or groups seen, and a lack of sightings except on the south-central grasslands of the reserve, were indications that during the season of maximum deer concentration a high proportion of the total barasingha p0pulation may have aggregated into the largest herds seen. Close agreement among the three maximum ground counts in 1975 also indicated that perhaps few animals remained uncounted. The highest ground counts in 1975 (805) and 1976 (908), therefore, were regarded as the minimum barasingha population each year. With respect to fecal-pellet counts, factors which may affect pellet-group deposition rates have been summarized by Neff (1968) and Nootong (1975). Young deer are known to defecate more often than adults, and green forage like that present on the southern phantas when counts were made also increases defecation rates. The actual rate for each species thus may have been more than 13 groups per day, and population estimates based on the true rates then would have been somewhat smaller than those given in Table 2. Since hog deer and chital were known to be more widespread in the reserve than barasingha during the premonsoon seasons, popu- 1ation estimates for those species apply only to the grasslands sampled. The observed relative abundance of hog deer and chital, 36 however, may indicate their abundance in the reserve as a whole. Judging from the available habitat in which hog deer were seen regu- larly, they probably were the least abundant of the three deer. Chital being the most widespread were likely to be more abundant than barasingha, and therefore, were the most numerous wild ungulates in the reserve. In comparison, estimates of chital density in Wilpattu National Park were 44.4-57.9 per km2 in suitable habitat (Eisenberg et al., 1970; Eisenberg and Lockhart, 1972), approximately twice as many as were present on the southern phantas of the reserve. While an unknown portion of the barasingha population of Sukla Phanta Reserve was not observed during deer counts, it may be assumed that the entire p0pulation present on the southern phantas contributed to the number of fecal pellets counted. The pooulation figure derived from pellet counts (1,295) therefore may be taken as an approximation of maximum barasingha population size in the first quarter of 1976. Of the methods employed to census deer, direct counts made at the appropriate time of year along the ground route described above were relatively quickly done with the least expense, and pro- vided an index to barasingha population size. Ground-count data could be accumulated over several years by reserve personnel and used to monitor changes in barasingha abundance. REPRODUCTION Timing of the Rut and Distribution of Rutting Stags Martin (1975) found that the highest rate of rutting calls or "bugling" by stags among the Kanha deer coincided with the observed peak of mating there in the latter half of January. In Sukla Phanta, calling was used as an indicator of mating since barasingha rarely could be seen in the rank post-monsoon grasses. The largest group recorded in the cold season before the grasses burned was 12 animals. Larger "breeding herds" like those observed in Kanha (Schaller, 1967; Martin, 1975) were not seen to form in Sukla Phanta. Rutting calls were heard primarily after 1600 hrs. and intermittently throughout the night in Sukla Phanta. Calling usually ended in the morning before 0900 hrs. The earliest call of the season was heard on August 31, 1975. That same day the first mature stag in hard antlers was seen. The latest call recorded was on April 18, 1974. Subsequent observations showed that mature stags began to drop antlers in early March, and apart from the instance referred to, calling was not heard in April. Methods To determine the peak mating period and to locate rutting areas, calls and their direction from the observer were recorded in 37 38 the early morning and evening from four "listening posts," located as follows: (1) the guard post at the former Barkola Village site, (2) midway along the side of Sunderi Phanta and 100 m east of that grassland, (3) at a lone Butea monosperma tree 400 m south along and 150 m west of the North-South Track, and (4) in Karaiya Phanta south of (3), at the edge of the forest separating Karaiya and Seta Khera. The Sukla post was monitored 19 times between Septem- ber 22 and December 13 during the 1974-75 rutting seasons. The others each were monitored one morning and one evening in November 1974, except Sunderi, which was visited two mornings and evenings. During one morning at Sunderi and each morning at Barkola and Karaiya, calls were recorded simultaneously by another observer at Sukla. Results Data for 1974-75 from the Sukla post were combined since the rut followed roughly the same schedule each year. Evening calling rates declined from mid-October and remained approximately at the same level until mid-December (Figure 5a). Morning rates peaked in mid-November and were markedly lower at other times. Of 498 calls recorded at Sukla, most came from the southwest and southeast, and the least calling was heard from the north (Figure 5b). Morning calls at two posts were waer than those simultane- ously recorded at Sukla (Table 3). The least activity was noted at Barkola, with nine calls scattered around the post. Morning and 39 Figure 5.--(a) Morning and evening hourly rutting call rates in Sukla Phanta proper, 1974-75 (b) Percentage of total calls heard from various direc- tions in Sukla Phanta proper, 1974-75, Sukla Phanta Reserve, Nepal. 4O E 25 \\ MORNING g \\ —-—— EVENING o . \ E \ '2‘ \ 15 ° \ 23 \ 3 \ ,.—- \ : L—"fi‘ ‘~.v”” \ 5 22 231 21617 20 2911 1211375110221 211 29 61 Isenl l-—OCT.-——| F Nov 1 DEC.r 50. CALENDER DATES 45 4° _ SHADED—NORNINC UNSI'IADED — EVENING PERCENT OF RUTTING CALLS 5 5w w Nw N NE E SE 5!). DIRECTION FROM SUKLA LISTENING POST 41 TABLE 3.--Morning and evening rutting calls counted at three listening post, compared with morning counts made simultaneously in Sukla Phanta proper, November 1974, Sukla Phanta Reserve, Nepal Barkola Sunderi Sunderi Karaiya Totals Three posts, PM 4 12 17 16 49 Three posts, AM 5 3 3 14 25 Sukla, AM 28 40 -- 14 82 evening calls at Karaiya came mainly from Seta Khera to the north- east and northwest. Discussion The reason for the difference in pattern between morning and evening calling rates (Figure 5a) is not known. Possibly the high number of September evening calls represented pre-rut activity as stags challenged one another to establish dominance hierarchies. Behavior of this kind was noted among red deer (Cervus elaphus), where dominant stags were more likely later to establish harems (Lincoln et al., 1970). Although barasingha do not maintain harems, dominant stags have priority in access to estrous hinds (Schaller, 1967; Martin, 1975). Since the gestation period in barasingha is 240-250 days (Kenneth, 1953; Asdell, 1964), the peak of new fawns appearning from late June into August (see beyond) indicated that the high number of rutting calls in Mid-November probably reflected the peak of mating. 42 The frequency and location of calls indicated that rutting activity was greatest in southern Sukla Phanta proper and adjacent Seta Khera. Five permanent wallows used by rutting stags each year of the study were found in southeast Sukla Phanta and north-central Seta Khera. Similar wallows were not found elsewhere, though appar- ently suitable wet, grassy places were present in the lowlands dur- ing the rut. Fidelity to rutting areas was found among the Kanha deer (Schaller, 1967; Martin, 1975), and is known to occur in other species as well (Darling, 1937; Murie, 1951; Ueckermann and Hansen, 1968). Use of certain rutting areas may be traditional among the Sukla Phanta barasingha as well. Each year local people were permitted to cut grass for thatch in the reserve for one month in November-December, during and just after the height of the rut. Since the tall, dense grasses in southern Sukla Phanta and environs were particularly desired, the disturbance caused by large numbers of people and draft animals in the main rutting area may affect pOpulation mating success. Ageof Hinds at First Breeding Schaller (1967) and Martin (1975) found no evidence that yearling hinds participated in the rut in Kanha. Schaller believed that hinds first mate when just over 2 years old and bear fawns at about 3 years of age. Barasingha hinds at Woburn Deer Park in England do not mate until 28-30 months old (0. G. Talbot, personal communication). There was no indication that yearlings in Sukla Phanta took part in the rut and it is likely that most, if not all, 43 of the annual fawn crop was produced by hinds which mated first at 2 years or older. Fawning Owing to the dense vegetation in Sukla Phanta, the wariness of the deer, and the tendency of young fawns to remain hidden, no data could be obtained on the ratio of hinds to new-born fawns. Fawns counted in March (see beyond) had survived the first months of life when mortality was likely to be high. The proportion of the annual crop lost between the June-August fawning season and the following March is unknown. The earliest new fawns were observed on June 28, 1974, shortly after the onset of the monsoon rains. One of four fawns seen on that date was flushed after its mother revealed its hiding place. The other three were with groups of older animals, indi- cating that they had completed the "hiding phase" common among new- born cervids (Lent, 1974). The estimated age of these three animals was 10-14 days, based on the assumptions that they were among the first-born of that season and that the hiding phase had lasted about a week. Conception of these early fawns then would have occurred in mid-October, roughly a month earlier than the estimated peak of the rut. New fawns were seen 36 times during the 1974 monsoon season, either alone, with their mothers, or in the company of other deer. In 1975, the first new fawn of the season was seen on July 6 in a 44 small group of animals. Thereafter, new fawns were observed 22 times during July and August. Schaller (1967) found no records of twinning in barasingha, and apparently did not observe it among the Kanha deer. Hinds were sometimes seen with two fawns at heel in Sukla Phanta, but this could not be taken as firm evidence of twinning, since one or two fawns sometimes even trailed after stags. POPULATION SEX AND AGE STRUCTURE Sex and age composition counts were made several times between November and May during both years of the study. Figures obtained in March during the season of maximum deer concentration were judged best to approximate actual population structure because the largest counts were obtained, and visibility was better than at other times of year. Methods All counts took place in Sukla Phanta proper, save one in Dudhwa National Park, and were made with binoculars and telescope. The most favorable circumstances for accurate counts occurred when herds were moving slowly, roughly in single file and at right angles to the direction of observation. Individuals then could be distin- guished and classified as adult males, adult females, yearlings, or fawns on the basis of relative body size and the presence of antlers or antler pedicels. Though yearling males carried short, unbranched antlers or "spikes," these were not always visible in the field, and yearlings were tallied only as such, without regard to sex. Since deer could not be closely examined, the classes most difficult to identify were females, yearlings (males as well as females, if spikes were not evident) and well-grown fawns. In April-May 1974, no distinction was made between yearlings and fawns 45 46 of the previous monsoon season, since insufficient skill had been attained at that time in separating the two classes. Results In Sukla Phanta, adult males comprised 16.4-32.7% of animals counted, adult females 35.2-48.6%, yearlings 7.1-16.1%, and fawns 9.2-17.7% (Table 4). The three largest March counts (one in 1975 and two in 1976) included 30-40% of the minimum population in each instance. The smallest March count in 1975 contained an unusually low number of fawns, while the Dudhwa count included a larger propor- tion of immature animals than counts made in Sukla Phanta at the same time of year. Discussion Variations in proportions of deer observed in four sex and age classes were due to differences in visibility and deer distribu- tion at different times of year. As expected, numbers of unclassi- fied deer were greatest in November-December before the grasses were burned. Variation between counts was greatest for adult males, although they usually were easiest to recongize because of their antlers. This variation may be caused by the tendency of stags to wander and to be more widespread during the rut (Schaller, 1967) as well as segregation of adult males from other members of the popula- tion for part of the year. Martin (1975) found that stags left females and young during the premonsoon season, with segregation by sex reaching a peak in May. Segregation was noted in Sukla Phanta 47 maze» mmuspucm .mpeswcm 0205 Lo P we umcwmmu mm: ugw; <0 n xgea Pecowuez nguac .xuopm acesxmxm mnem; mega; mega; use; F 820; _ ecu; P mzowee> ego; F CEO; F msowgm> maowge> .mpeswcm .mpeewcm mpmemce .mpeewce .mpeswcm .mpesvce mpeswce .mpwswce Nn_ mam 5mm no mpm opp mpp sup omugzom mmeEmw o.oo N.mm m.mm m.mm m.mm o.mp m.~e 1. “Fave oo_ \m:3mu Ao.e V Ae._ V Am.“ V A_.e V Am.m_V Am.~ V ee_eem 1. mp e m m_ 1. mp e Imepuca AN.mFV Am.mPV A_.m_V Ae.e_V AS.P_V AN.m V 2R.RFV mm mm mm FF mm FF cm I- maze; Am.mmV Ae.o_V A_.epV Ae.m_V A_.ePV “o.eFV AF.“ V eA~.m.V mmee_ _o ow Nu m Pm m_ m cm 12mm» AP.eNV Ae.mmV Ae.mmV Am.eeV AN.mmV Am.NeV Ae..eV Ae.meV me_e5ee om mmp mp, pm «PP mm me ow up:u< Am.e_V Am.omV AA.~mV Ae.eFV AA.NmV AA.NNV Ae.e~V Am.mNV me_ez mm mmp no PP co. mm mm mm szu< mump cum, osmp mump mnm~ mxmp «nap eumF mpwsa< nose: coco: .umo1.>oz cogmz nose: .Ooo-.>oz xezIFwsa< mvccH .xgma chowumz mzsuzo .0>memm mucosa epxzm cw wwaFeu egmcwmwgen Lo mom can xmm an .xuo—m mgmgxmx use .peamz Ammmepcmugma uceV mgwnE:z--.e m4m

o .mo v apacouwmwcmpm our ooo.? moo moo ooo.o ooo., Fee_eu ooo.. om om ooo._ omo Leeo ooe ooo.P AoF.. oo_.~ ooo.. ooo.. eeoESReeee m_eeoe «ooo._w.oo.w ooo. moo. o P ooo. mo Feoeeo -- ooo. _ o ooo. mp eeoo ooe oeoeoe -- ooo. _N o opo. mo egocemeeee _em -- ooo. o o ooo. AN Peoeeo I- ooo. F o mpo. NP soon we: occo>om -- ooo. om o Poo. Po eeoe_meeeo Pom Amo..w.oo.w ooo. ooo. A o. ooo. .m .eowee omeeoe -- ooo. _ _ mNo. m, eeeo ooe mooooeeeo -- ooo. o2 o Npo. om egocemeeee ooxoz stow“. mfloomm_oom. ooo. oo Pow ooo. eon .eoeeu em_o. w.oo w.omo. mo_. Pm o_ moo. ooN eeoo ooe eeeesem seem. v oo v oop. mew. ope oem Rom. ooo egoceeeeee oee_3oo eoNo. m oo.w.moo. ooo. oop o ooo. Koo .eoeeo N_o. w.oo w.ooo. mop. o. o_ ooo. ooN eeoo ooe oee_mmeeo «Rook. v on v one. woe. mfio ooo ooo. ooo egocemeeee Ago ooo. muoo mfloom. Now. oo oo ooo. ooo _eoeeo oee_mmeeo «took. w.oo w.mmo. Leo. oP mm «mm. oop some ooe ea; eooo. v oe v ooo. oNo. o_o om ooo. mom egocemeeee sppeeomeem o o 15 Come o .cowugooogo o we :owusoooso A.o:V om>gomeo .u0>gmmoo omwwomwm omwmmwmo omuuwoxu "V voam>eom mowumom Heaven: co po>gmu:H :owugoooco a z a z omg< pouch omg< mucmuwmzou mo cowusoaogo poomz .m>gmmwm mucosa opxom .mnmp gogoz .mpoumno; xwm cm mmwumom smut moss» go mucmggouoouu.m m4mo .mo v appcmuw$wcmwm one ooo._ oN oN ooo.F ooo.. Peo_ee ooo.o om oN ooo.P oNo eeoo ooe ooo.o NNo oNo _oo.P ooo.. eeooemeeeo mpooop NeoNN..w oo.w ooo. NN.. N op ooo. no _eeeeo -- ooo. F o NNo. m_ eoeo ooe emeeoe -- ooo. o o NNo. mm egocemeeoe Fem NeooN..w oo.w ooo. ooo. N mo oNo. NN Peewea -- ooo. F o PNo. NP eeeo ooe eeeesem -- ooo. __ o oNo. _o eeoeemeeee Fem NemeN..w_oo.w ooo. Nop. N __ oNo. .m Peoweo emeeoe I.m- I. ooo. _ N oNo. m_ Loeo ooe mooooeoeo ooo. v o v mpo. Noo. o o. o_o. oN eeoepmeeee oex_z eoN_.Hm comm Noo. ooo. NN o moo. ooo Peeeee o_m. N.oe w.oo_. o_m. m. o ooo. NmN eeoo ooe eeeesem eooo. v oe v _oo. Noo. No_ NA ooo. ooo egocemeeee oeeezoo e_oo. mnoo m _oo. NNo. No N ooo. Noe _eomeo loom. w.oo w.moo. NNP. o_ m ooo. ooN eeeo ooe oeepmmeeo eroNo. v oo v ooo. oeo. oN_ ooN ooo. ooo eeoereeee seo ..m- I. moo. o P NNF. mop _eeeeo oee_mmeeo eeooo. w. o w.oNN. ooo. e N, mo_. No eeoo ooe No: eeoom. v oo v ooN. _om. oo oNP Noe. No_ eeoEPReeee spPeeomeem 0 Aug .meo oo .cowp a we cowusooogo A.o:V Igooogo om>2mmno co .om>2omoo cmwwmmnm omwmmwmo cmuumoxm uV umxu>gzm mmwuwom weapon: ~o>gmucH mucmnpwcou compgooogo a z a z omg< Pogo» mw2< we cowugoaoga Poowz .m>2mmmm mucosa opxom .mnmp mcowuaoz .mpoupao; xwm cw mmwomam Emma omega mo mocmggsuooII.m u4mo .ma v xfipcoupmpcmmm oar ooo.. No_ No, ooo.o Nmo Loeweo ooo._ om, mop Foo._ Non eeoo ooe ooo.o oo oo ooo.~ ooo eeoeemeeeo m_eoop 3.. M oo M oNo. 5. mm N NE. N2 3% eooo. w.wo w.ooo. ooo. oN N oop. Po eeeo ooe Someoe eooo. v a v _oo. oNo. NF N oN_. pop egocemeeeo Fem eeNom. Mfloo M oo_. NNN. NN Pm o_P. oo Feeeeo eooo. v oo v ooo. oNo. op o oNF. oo eaoo ooe eeeesem -- . ooo. N_ o oo_. om. eeoeemeeee Fem tooo. M on M ooo. NS. o NS ooo. 3 25.50 :28 Noo. v oo v Noo. o_o. o m Noo. op eeeo ooe mooooeeoo -- ooo. N o oNo. NN eeoewmeeee oexez .I m-.u ooo. oo o moo. ooN Peoeee eoNF..w o.w_ooo. o__. om op ooN. oo eeeo ooe eeeesem ooN. v as v ooo. oop. oN o_ ooN. NNN egocemeeeo oeepzoo loo. M oe M ooo. .No. 3 o ooN. 2N 23.23 eeoNo..w oo.w oPN. ooN. om NNP ooo. No. eeeo ooe oeepmmeeo .eoNo. v on v ooo. oPN. oN oN ooN. ooN egocemeeee Neo -- ooo. o_ o oNo. mo Feoeeo oeepmmoeo .I --.I ooo. op o oo_. oo eeoo ooe No: mop. v oo v ooo. _No. o N Noo. oN eeoeemeeee sp_eeomeom awn £08 on .cowu on Co cowugoaogo A.o:V -Looogo um>gmmno co .cw>gmmno omwwmmnm omnwmmMm uwuuwoxm uV umxm>gom mmwuwom powwow: po>2mch mucmcw$cou cowugooogo owg< quoh oms< Co cowugo o2; Foam: .m>2mmmm mucosa opxom .mmm_ ..powm-.mo< .mpopwoo; xwm cw mmwumom some omega Lo mesmesouoo--.m m4m

o .mo v apucouwCPcmmm one ooo._ oo oo ooo.. Noo Feoeee ooo._ FF NF ooo.o ope eoeo ooe ooo._ mo_ oo_ ooo._ oNo egocemeeee mpeeoe opo. v oo v oo_. Noo. o_ N_ ooN. oNP Feoeee - moo. N _ NAP. NN eooo ooe Amoeoe -- ooo. oN o mop. oop eeoeFmeeee Fem _om. v oo v ooo. ooN. o NF Pop. oop Feeeeo I.-- I. No_. N N om_. oo eeeo ooe eeee>em eom_. v oo v o_o. oNo. NN o .NFN. _PN eeoe_meeee New ANN. v oo v ooo. No_. N o ooo. No Feoeeo Cameo» - ooo. o o ooo. mp eoeo ooe moooooeeo - ooo. o o oNo. oN eeoeemeeee oexwz o_o. w.oo w.ep_. ooN. o N, oop. No, _eoeeo Noo. w.oo w.Poo. mom. _ o oN_. om eoeo ooe eeeesem ooN. v oo v oNo. No_. o_ N_ ”NP. oNF eeoeemeeee oeepzoo Ioop. v oo v ooo. Poo. N_ N ooN. N.N Feomeo I.m- I. No_. o N o_m. No_ eeeo ooe oee_mmeeo eeooN. v o v Noo. ooo. oN NN ooN. ooN egocemeeeo seo - ooo. o o mm_. NNP _eoeeo oeepmeeeo I.m- -. ooN. N o oo_. No eeeo ooe So; some. v e v o_o. oNo. N_ o oo_. Noe egocemeeee »F_eeomeem o 3a .208 on .cowu a $8 cowpgooogo A.o:V -2ooo2o um>gmmno co .om>2mmnc umwwwmwm umwmmwmo umyumaxu uV umxm>2=m mmwumom “maven: Fo>gmucH mucmnwycoo cowugooogo n z a z om2< Pouch omg< wo :owpeooogo Fwnwz ow>memm mucmc—n— mPv—zm .mmmp .Uwa .mvmuwnw; xwm E. wwame meU 025.... $0 wucmLL300CII.w m.._m<._. .Ao_.o v aV ompooNccw mucmgmwmeo .oo A apucou_wwcm_m oar- .Aop.o v oV umpoowocp mucouwo>o .mo v apucouwmwcmwm oar Noo._ ooN ooN oo.N mNm.. NeeNeo ooo., No om ooo.N moo eeeo ooe ooo._ oNN NNN ooo.o ooo._ eeoeNmeeee mNeeoN mmN..w oo.w emN. ooN. mm Ne omN. oNN NeoNeo - ooo. o o NNN. NN gown mo; umwgow - ooo. NN o ooN. om. eeoeNmeeee Fem -Noo..w oo.w ooo. mmo. mN o ooN. mmN Neomee - ooo. m o ooN. oo eeeo ooe oeeesem -- ooo. om o NNN. NNN eeoeNmeeee Nem -- ooo. o o NNo. mN NeeNeo omeeoe - ooo. N o NNo. NN eoeo ooe mooooNoeo - ooo. N o ooo. N. oeoeNmeeee ooxNz Nom. M 2. M mmN. mNm. No oN ooN. mom 3% Nmo..w oo.w oNN. ooN. m, oN mom. mmN eeeo ooe oeeesem eeomN. v oo v Noo. Noo. mo oNN oom. mmo eeoeNmeeeo ooNNzos e-NNm. M ooflM oom. Noe. mm ooN NoN. NNm Neemee I-NoN..w oo.w mNm. ooo. NN NN NNN. omN eeoo ooe oeeNmmeeo eemoo. v oo v oNN. omm. No No NNN. Nmo eeoeNmeeee Neo .I -.I ooo. mm o NNN. ooN NeeNeo oeeNmmeeo mom. v oo v Nmo. ooN. o_ o oo_. mNN eoeo ooe ea; -- ooo. om o ooN. oom eeoeNmeeee mNNNEomeem o 0 “mo .wao a .cow» 8 mo cowpgoooeo A.o:V -2ooogo co>gmmao :o .um>gomoo omwwwowm COWMOMMO ompumoxm "V vmxm>gom mmwumam powwow: Fo>gmucH mocmcwmcou cowugoooeo as 2 a z om2< pouch owg< mo copugoooeo Poooz .0>memm mucosa opxom .mNmF sugoz .muouwno; xmm cm mmmomom some moss» mo mucmggauuoII.m m4mo2s uos sows: Acumoom use wwwumom sumo sowV musONsmoe sopmswm No mmusmsmwwwu saw; umsooeou mo: Amsms53s um>smmso .m> umuumome opou 3mg ms» s_ musmsmwwwu 020 $0 musuwsmoe ms“ .mmmou susm sH .ONON sosu mmmp mo: uwswp smzo_ 0;“ Hosp pposm ow 0203 msowusooosa um>smmso so .um>smmso 0003 mpoevsu os mmsoums umyopsupou on yo: upsou mpo>20usw musmummsou II A V .Aos so Fo>smpsv wusmuNNsou swsuwz mwmp moV umuwo>o Cos umsgmwmgs gmsuwms youwno: u o .Ams v osV umuoumusw museumo>o «moves: u I .Amo A osV umuoumusw musmgm$mgs “mamas: u + +000 AA AA O0 loo vv vv N-V NHV N-V so NoV N-V 0++0 AHV I I .w s: A V o + 0++00 I O I 4. VV VV l0 CO A I O V AA AA COO AAA IOOOO VVV IOO‘I’ +++++ N-V N-Vmo 8V AI o AA 100 vv + + N-V +0 uNm_ sues: swasmooo gnaw-ms< mssuImoz mNm— saga: 2:5 on— sage: NwsEmumo poem-ms< mssuIANz mNsp sugoz Nmmu mo: on_ sues: smssmoma usmm-ms< wssu-ao: mNmF sous: osmsmmogum umwsos Pom osso>om pom amass; msosuwumo umxwz usopzod posmz .0>smmms ousoss o_xsm osso>om usopmmogo usmpmmogu so: xpposomowm .mumuwsos me s_ usoos-somx um>somso mo mmmumom Nmmu mass» 00 moswsgsuuo passwomm .xgussomI-.o— m4m7.772, p < 0.10 species are primarily grazers, food preferences probably influenced habitat selection less then because young grasses were higly pala- table, and a larger variety of forage species was likely to be acceptable than later in the year (Bor, 1960; Heady, 1964; Martin, 1975). Deer thus moved more freely between grassland types than during other seasons. Fecal pellet distributions supported this interpretation, indicating that the three deer species frequented the three grassland types nearly equally until the end of April. Barasingha were the exception, and as indicated by their fecal pellets, occurred more commonly on dry grassland than hog deer or chital. While deer counts showed habitat preferences at the moment they were made, pellet counts reflected deer movements over a 60-day 65 period, including nocturnal movements. These may well have differed from movements observed during the day. Chital, for example, were seen venturing further onto the open grassland of Sukla Phanta proper at night than during the day. Schaller (1967) noted that while barasingha utilize a var- iety of vegetation types, including open forest where grasses are present, historically their greatest abundance was reached in the sandy and marshy riverine grasslands of north India. Of the several vegetation types in the reserve, the southern grasslands are most similar to this environment. In central India, on the other hand, the southern barasingha exist entirely in a sal forest environment. Martin (1975) found that the deer in Kanha National Park infre- quently entered the forest but confined themselves largely to mead- ows associated with this type (cf. sal savanna in Sukla Phanta Reserve). The Kanha deer migrated between seasonal ranges some 7 kms, apart in response to changing availability of water and food. Martin suggested that seasonality is more markedly felt in the central Indian highlands, which lack the alluvial flats of north India, where green forage and water are available even in the dry season. In Sukla Phanta where both lowland grasslands and upland sal forest environments exist, barasingha remained largely in the former, and moved little to satisfy seasonal requirements. FACTORS INFLUENCING HABITAT SELECTION Feeding_Habits Methods Barasingha are primarily grazing animals (Schaller, 1967; Martin, 1975), and an attempt was made to identify food plants and to assess their distribution. Grasses and sedges were collected from 116 lO-m2 plots regularly placed at 300 m intervals along transects 500 m apart. Sixty plots were in Central and western Seta Khera Phanta and in Sukla Phanta proper. The remaining 56 were located in the sal forest and savannas northeast of Sukla Phanta. Plots were examined after the monsoon season in October and early November, 1974-75, when the majority of grasses were in flower or fruit and therefore identifiable. Frequency of occurrence of grass and sedge species was recorded, and an ocular estimate of the area covered by each species in each plot was made. Feeding observations were made during the premonsoon and monsoon seasons in Sukla and Seta Khera Phantas. Because of the wariness of the deer, and due to vegetation density during the mon- soon, few opportunities to see directly what barasingha ate were possible. The presence of two other deer species prevented the identification of grazed or browsed plants as barasingha foods, even when livestock was absent (see beyond). 66 67 Results Fifty-four grasses and sedges were collected in the reserve and at reserve headquarters (Appendix C). Twenty-three species were in the sample plots (Table 12); the remainder were collected outside the areas sampled. Perennial grasses constituted the dominant vege- tation of the grasslands sampled. By using a Species-area curve (Braun-Blanquet, 1932), it was estimated that most of the common species were collected (Figure 6). The lowland (Sukla and Seta Khera Phantas) and upland (sal forest and savanna) environments each yielded 17 species in the areas sampled, with ten common to both. Of all species collected in the reserve, 35 occurred in the lowlands and 23 in the uplands with 12 common to both regions (Appendix C). Sterile (usually unidenti- fiable) grasses were found more frequently in the sal forest and savannas, as were bare ground, forbs, and c0ppicing woody vegetation. Very little nongrass vegetation was found in the lowlands, and bare ground occurred only on the more xeric sandy sites. The grasses Ophiuros exaltatus and Mnesithea 1aevis were not identified as sep- arate species (Table 12) until specialists had examined the collec- tion. Five grasses which barasingha were observed consuming were Imperata cylindrica, Narenga porphyrocoma, Phragmites karka, Saccharum bengalense and Saccharum spontaneum (Table 13). One instance of browsing on Acacia catechu was observed. Barasingha also ate aquatic vegetation, plunging their muzzles into seasonal TABLE 12.--Frequency of occurrence and estimated average percent1 cover of 23 common grasses and sedges in 116 plots in Sukla Phanta Reserve, Nepal Sukla & Seta Khera Sal forest & Savanna Species Frequency Cover Frequency Cover Apluda mutica .08 1-10 -- -- Cymbopogon sp. .33 1-10 .04 1-10 Desmostachya bipinnata .22 20-30 .34 1-10 Dicanthium bladhii .12 1-10 .30 l-lO Eragrostiella bifaria .10 20-30 -- -- Eulalia leschenaultiana -- -- .20 10-20 Eulalia trispicata -- -- .11 1-10 Eulaliopsis binata .02 10-20 .48 10-20 Fimbristylis sp. .18 20-30 -- -- Hemarthria compressa .03 1-10 -- -- Heteropogon contortus -- -- .04 1-10 Imperata cylindrica .40 60-70 .11 30-40 Mnesithea laevis2 2 Ophiuros exaltatus .13 1-10 .36 1-10 Narenga porphyrocoma .63 50-60 .39 1-10 Ophiuros megaphyllus -- -- .02 1-10 Phragmites karka .03 1-10 -- -- Saccharum bengalense .33 30-40 .07 10-20 Saccharum spontaneum .53 20-30 .38 1-10 Scleria alta .03 1-10 -- -- Sorghum nitidum -- -- .11 1-10 Themeda arundinacea -- -- .39 1-10 Vetiveria zizanioides .02 1-10 .02 40-50 Unidentified sterile grass .12 60-70 .55 30-40 Forbs, coppicing woody vegetation .08 1-10 .86 10-20 Bare ground .10 10-20 .64 10-20 IOcular estimate on a scale of 1-10, e.g., l = 1-10% of plot area, 2 = 11-20%, etc. 2Not treated as separate species in the field (see text). .Ckmh mNm. 9.12.2. cwm2m>0212m00b00 02350 .._cwmw¢ (2.25.3 (1.23m 2. Owh0m4400 mwgmm o2< mumm<¢0 20n— m>¢30 em use “mosom New :0 soopmo>wp mo uoNeNm mpswmmoou .uo>somso appesuoe pos ostmsosmos AmVoso>wosos ms» was .uopeoeusw me uoNesm mFowmmoo .mszeem mo msmwm mswzosm mowooomo .oeoou s? ueNeem mowoosm so muo>somoo appesuoe pos msNNesm one .uoNeNo aNsmoNN me: mmesm egos; “somoss mpeswse .oosouw>o Femuseumssoewuo .oosepmsw oso me: peewse oso an msFNeNm 0o usoo uo>somoo smuo> u.oeooeswussse euoEosp oeu—e ewsopom x oemmosoeoo ewespeeEoz x 0.8 $3222: oepesswopo emsoeumoEmoo o.om somosonsmu N goosepsoom Eseesooem oesopemsmn Essesooem m exsex meowEaesss QNF m esooosasosoo emsosez 87 m N o: o7 822:0 3285 N sopzuoeu souoszu X X XXXX XXXX MOXXXXXX a COOmcoz COOMCOEGLQ COOMCOZ zoomCOEwLa COOmcoz zooms—9:250 mowooom uoNeeo mooseemsH uoNese mooseomse euoNeeo moosepmsH Fepwso some mo: esmswmesem .uN-onp .msomeom soomsoe use soomsoeoso one msweou .Feooz .o>eomom epsess epxom sw mowooom soou omega an momuom use mommesm Peeo>om so mosouwosw mseueszI.mp m4m