ECOLOGICAL RECONNAISSANCE OF ADULT ODONATA (INSECTA) OF EASTERN INDIA

Eastern India, in the present context, covers parts of Assam, Bihar, Sikkim and West Bengal mainly, Mizoram, in the east-west direction. Physiographically the states belong to Assam plain, middle and lower ganga plain, and eastern Himalaya respectively. Although there is no physical barrier to divide the area into east and west but the east is more humid and west more dry. The rainfall increases as one luoves from west to east. The average rainfall is variable from 200 em to 400 cm. per year.


INTRODUCTION
Eastern India, in the present context, covers parts of Assam, Bihar, Sikkim and West Bengal mainly, Mizoram, in the east-west direction. Physiographically the states belong to Assam plain, middle and lower ganga plain, and eastern Himalaya respectively. Although there is no physical barrier to divide the area into east and west but the east is more humid and west more dry. The rainfall increases as one luoves from west to east. The average rainfall is variable from 200 em to 400 cm. per year.

HISTORICAL RECORDS
There is no consolidated report on the ecology of dragonflies of eastern India, although Mitra (1994) reported several aspects' of habits and habitats of adults and cited the available records on the subject. Besides that Fraser (1933Fraser ( , 1934Fraser ( & 1936, Asahina (1958), Dasgupta (1988Dasgupta ( , 1996, Gupta et. al. (1995), Lahiri (1975), Prasad and Ghosh (1982), Ram and Prasad (1978) reported some events on the subject.

MATERIAL AND METHODS
The present report is based on the observations on odonata in the field, mainly in the nearby localities of Calcutta. During the preparations of the report records available from other localities have been added.
It is interesting to note that dragonflies select their food preferably smaller than them and never eat their own species. Intraspecies fight for food was noticed only once. Mitra (1994) reported a clash between a female and a male Ischnura senegaiensis on February, 1967 Breeding activities: Breeding season, on the basis of observation on tandem and copulations, of the following ten species could be regarded as follows. A griocnemis pygmea breed throughout the year; Pseudagrioll rubriceps and P. nzicrocephalum, Ceriagrion coromandelianum, Ischnura a. aurora, Crocotlzel1lis servilia breed during the period from June to November; Ischllura senegalensis breed from March to December. Depending on the climatic conditions, breeding periods during the twenty four hour cycle are as follows. Pseudagrion rubriceps, P. microcephalum, Ceriagrion coronzandelianum, Ischnura aurora, I. senegalinsis Agriocnelnis pygmaea, Crocothemis servilia breed from the early morning to noon (6.00 to 12.00 hr), while Brachythemis contaminata and Tholymis tillarga breed from late afternoon to early evening (17.00 to 18.00 hr).
All. species fonn pair and remain near water for oviposition. But Orthetrum sabina fonn pair near water and moves away from water edges to a certain distance probably to avoid <:listurbances. On three occasions viz. on September 15, 1982, September 19, 1982and August 15, 1983 one pair of Orthetrum sabina in each day was found copulate over water even when disturbed by one of us (TRM). Moreover, during the period, November 25 to 30, 1987, in the area of the river Kulik, in the district of West Dinajpur of West Bengal, some specimens of the species (0. sabina) in overcrowdc d localities, were found to wait on land for partners. Besides this, the copUlating pairs were not leaving the area if they felt the wheels were weak, which after sometime got detached and the pair immediately formed wheel agaIn.
Territory fonnation during the breeding could be noticed in Tholymis tillarga, Crocothemis servilia Orthetrum sabina. Intra species competetion as well as inter species competetion for territorie~ could also be noticed. The horizontal dimension of territories depend on the density of the male population, area of the aquatic body and some intrinsic factors.
Elnergence o/imago: Mass emergence of any species could not be seen, However, emergence of certain species could be recorded, for example, Orthetrum sabina could be observed to emerge in the early evening during the summer months at Dum Dum Park, near Calcutta city. Tholymis tillarga emerged in the morning of August at the same locality; Brachythemis contaminata emerged in the midmorning of December in the Kachugaon forest of Assam; and Bradinopyga geminata in the early evening of October-November at Hazaribagh (Bihar). Pantalaflavescens and Crocothemis servilia in the early afternoon of April and midmorning of September of eastern Himalaya.
Migratoryflights: Migratory flights could be seen only in Pantalaflavescens. The flights were noted in September and early October. The directions of the swanns were from the west to the east. The density of the swarm varied from thirty individuals to huge swann.
Reaction tOlvards moblie objects : Six species. Pseudagrion microcephalum, Ictionogomphus rap ax, Brachythemis cOlltaminata, Diplacodes trivia lis, Crocothemis servilia and Pantalaflavescens follow mobile objects. All species follow the objects flying parallel to the object but Diplacodes tri\'ialis follow the object by moving before the object and then watch and repeat the same for some distance. The cause for this type of behaviour is not clearly known, although it is presumed to be a predatory behaviour.
[herolo-regulation and behaviour during change oflveather: The means of thermo-regulation is val.able. Sometimes the dr'Jgonflies take shelter under leaves or under any shade during the summer. During rain they do the same. But very recently one of us (TRM) noted that a pair of Neurothemis tullia tullia (one male and one female) were not moving out to any place while in rain.
During the hot summer Brachythemis contaminata, Diplaeodes trivialis NeurQthelnis tullia tullia cover their thorax with wings and the abdomen remain in obelisk position. On the otherhand Orthetrum sabina Bradinopyga genlinata, Lathrecista asiatica cover their thorax with wings but their abdomen remain in parallel position to the ground. Sometimes Pantala flavescens and Tholymis tillarga remain under shade and sometimes the former remain on wings a few metre above the ground. During the cool season (December to February) usually the population become reduced. But during the solar eclipse in October 1995 when the ambient temperature fell to 23°C dragonflies became inactive they were no reacting even when disturbed, although th~y resumed activities with the rise of temperature as soon as the eclipse was over (Mitra 1996).
Death in adults: Several dead and dying dragonflies have been collected and their conditions have been reported by Mitra (1977Mitra ( & 1994. According to the conditions the death of dragonflies may be classified as follows. (a) Natural deaths: One dead male Onychargia atrocyana, without having any external injury, has been collected. The cause of death could not be ascertained and it is presumed that it has died a natural death.
(b) Death due to starvation: A dead male Crocothemis servila and a dead female Hemianax ephippiger were collected f~om the cobwebs. Since there were no spider it is presumed that the specimens died due to starvation.
(c) Death due to attack of ants: Sometimes Camponotus compressus and Solenopsis geminata attack weak and overfed dragonflies and become causes of deaths. For outside India records cf. Kiauta (1971 (e) Death due·to accidents: Sometimes modem vehicles dash or sucked in flying dragonflies in the grille or press the resting insects (Mitra 1994), for records of outside India cf. Kiauta (1965).
(0 Death due to unknolvn cause: Several dead and dying individuals were collected by one of us (TRM). The apparent reasons for those dead and dying individuals could not be ascertained (Mitra 1994).
Species recorded from houses: Specimen, of certain species have been collected from the house in calcutta and suburban of the city. It is interesting to note that no specimen was collected from the forest rest houses. On the otherhand no mass entry of any species has also been recorded? although roosting on the walls or verandah have been recorded. The cause for this type of behaviour is presimed to be the shrinkage of habitat.

SUMMARY
The paper reviews ecology of adult dragonflies in eastern India with special reference to the fauna of West Bengal.