ON TAXONOMIC STATUS OF BANDICOTA BENGALENSIS LORDI (WROUGHTON) AND BANDICOTA MAXIMA (pRADHAN ET. AL.) : (SUBFAMILY: MURINAE;

Agrawal (2000) recorded certain remarks on the taxonomic status of Bandicota bengalensis Lordi (Wroughton) and B. maxima (Pradhan et. al.) in the revisionary studies on Indian Murids and Hystricids. While reporting them, he has ~ompletely ignored the results obtained from the biochemical studies of these populations. Though merger of B. bengalensis lordi in B. bengalensis bengalensis has been supported by Agrawal (2000), he has totally neglected the variant populations occurring in the natural population. Probably very few taxonomists are aware of the observations made by Fergusson (1980) that allelic heterozygosity at any genomic site even in a single specimen, irrespective of age and sex, is a concrete proof of occurrence of polymorphism in the gene pool of the naturally occurring population under study.


INTRODUCTION
recorded certain remarks on the taxonomic status of Bandicota bengalensis Lordi (Wroughton) and B. maxima (Pradhan et. al.) in the revisionary studies on Indian Murids and Hystricids.While reporting them, he has ~ompletely ignored the results obtained from the biochemical studies of these populations.Though merger of B. bengalensis lordi in B. bengalensis bengalensis has been supported by Agrawal (2000), he has totally neglected the variant populations occurring in the natural population.Probably very few taxonomists are aware of the observations made by Fergusson (1980) that allelic heterozygosity at any genomic site even in a single specimen, irrespective of age and sex, is a concrete proof of occurrence of polymorphism in the gene pool of the naturally occurring population under study.

RESULTS AND DISCUSSION
It is now established fact that proteins like hemoglobins, eye lens proteins, keratins, certain enzymes like esterases, etc. are stable proteins and they are often used worldwide as chemotaxonomic parameters for solving taxonomic problems (Dessau"er and Fox, 1964, Selander and Young, 1969, Wright, 1974, Fergusson, 1980and Gavin et. al., 1999).Eye lens proteins and Keratins (hair proteins) are considered to be extremely stable and can be used for resolving taxonomic problems even at the family level (Fergusson, 1980).Even though aware of these results, Pradhan (1979) remarked that merger of B. bengalensis Lordi in B. bengalensis bengalensis could be more practical irrespective of the differences in the Hindfoot measurements and occurrence of Haemoglobin polymorphism in the natural populations.However, he further stated that B. bengalensis lordi could not be merged in B. bengalensis bengalensis by just overlooking the significant differences.But for these prominent differences, he was also in favour of merger, of the two subspecies.
All the morphological measurements reported in Pradhan (1979) and Pradhan et. ale (1989Pradhan et. ale ( , 1993) ) were taken meticulously by following Roonwal and Agrawal (1962).Even if, there is an element of error in taking Hindfoot measurements, as stated by Agrawal (2000), the level of difference can not be to the extent of 8-11 mm (Table 1).Hence, in light of these observations, the authors are, now, confident that two distinctly separate populations exist in the region.They are of the opinion that revisionary taxonomic studies of freshly collected B. bengalensis lordi and B. bengalensis bengalensis populations be carried out with the help of the morphological and osteological studies alongwith the studies based on modern parameters like biochemical, karyological and hair impression pattern studies.The systematic position of these populations can also be resolved by testing these observations on the scales of genetic distances and evolutionary separations (Nei, 1972 andSneath andSokol, 1973).
Taxonomic studies of Bandicota maxima populations were based on these lines only.All the available parameters were tested meticulously on Genetic and Evolutionary scales (Pradhan, 1989).On the basis of data on Genetic I values and UPGMA method of cluster analysis of Nei (1972) and Sneath and Sokol (1973), it became crystal clear that this population is distinctly separate from B. indica and deserves species novum status.In all probabilities and possibilities, B. maxima cannot be synonymised in B. indica.Agrawal (2000) overlooked the results based on the biochemtaxonomical studies reported earlier by Pradhan et. ale (1989Pradhan et. ale ( , 1993)).These studies are being widely used in systematics even today (Peppers and Bradley, 2000).Molar tooth-row 5.9-8.0 6.3-8.5 6.3-7.9 Length 715±0.47.25±0.47.65±1.15 In case of other character like difference in occipitionasal length, which is one of the key characters and not the only character (Pradhan, 1993), it has been observed that occipitonasal length was always equal to or more than and never less than condylobasal length in the juvenile, subadult and adult specimens of B maxima populations studied from different regions (Table 2).
The measurements of these specimens were taken as per Roonwal and Agrawal (1962).
Osteomorphological measurements reported in Pradhan et.al. (1989, 1993) were noted from the freshly collected specimens as well as those registered in Zoological Survey of India and Bombay Natural History Society's collections.Localities of these specimens include Mumbai, Pune, Goa, Kolkata, Pondicherry, some areas of Western Ghats and North India.While noting down the osteological measurements, it was observed that ridges in the flattened occiput region were distinctly pronounced and prominent in B. indica specimens of different sex and age groups.However, in B. maxima populations the occiput region was clearly inflated and swollen and the ridges were not as sharp, prominent and pronounced as those were in B. indica (Table 2).This difference could be noticed even in the juvenile and subadult specimens.
Differences in the hair impression patterns of dorsal guard hairs are distinct (Table 2) and have been used as an additional key character.It is, now, universally accepted fact that basal part of the mammalian guard hair shows definite and uniform hair impression patterns irrespective of sex, age and locality groups (Ryder, 1973 andKoegh, 1985).Similar techniques have been widely used in identification of carnivore species in wildlife studies by Chakraborty et. al. (1996) and De et. ale (1998).Thus the prominent differences seen in the hair impression patterns (Mozaic pattern in B. indica and Chevron in B. maxima) can not be overruled so easily.
Finally, on the basis of consolidation of the results derived from all the parameters (including use of Genetic Identity, UPGMA dendrogram and phylogenetic tree studies), Pradhan et. al. (1989 and1993) came to the ultimate conclusion that there exists a third population under Genus Bandicota, which cannot be merged in B. indica with the help of any circumstantial evidence.Hence, the population has rightly been separated as a new species and named as B. maxima on the basis of possession of the combination of key characters like occipitonasal length equal to or more than condylobasal length, inflated occiput region and mozaic hair impression pattern (Table 2).
We strongly feel and conclude that B. maxima is a valid species which does exist in the natural population and there is no ground for its "tentative" inclusion in B. indica as has been done by Agrawal (2000).Agrawal (2000) in the revisionary studies of Indian Murids merged two subspecies (pallidior and singuri) of Millardia meltada to nominate subspecies meltada.Here he has deviated from the basic definition of subspecies given in Mayr (1969).According to Mayr (1969) the subspecies is "a geographically defined aggregate of local populations which differs taxonomically from other such subdivisions of the species."Ellerman (1961) recognized two subspe, e, ies of Millardia meltada n.amely : Millardia m, e, leada melt ada and Millardia meltada pallidior from India.Latter subspecies differs from the fonner in having more pallid grey dorsum and whitish venter . .He has recognized two geo, graphical areas for these two subspecies.Further he has mentioned that the colour difference of these two subspecies is , an average and not absolute.What Agrawal (2000) found in Rewa, Gwalior , and Darbhanga , collections was nothing but mixed characters of pallidiorand meltada due to mixing (interbreedin, g) of the twoO subspecies in the demarcated fringes of the twoO populations.
In case of Millardia meltada singuri described by MandaI and Ghosh (1981) the specimens of this subspecies were collected from Singur, Dist. Hugli, West Bengal in 1978 and1980.They are more darker than Millardia meltada meltada.Even after 21-23 years of collection there is no sign of foxing in the coloration of these specimens which have been deposited in the National Zoological Collection of Zoological Survey of India in Kolkata.Over and above, these specimens have been collected from an entirely new geographical area.Thus, we find three subspecies in case of Millardia meltada in India viz.: M. m. meltada, M. m. pallidior and M. m. singuri which are present as geographically defined populations, although we came across with some mixing of meltada and pallidior populations living on borders.

Table 1 .
: Measurements (In mm) of some key characters of three Bandicota bengalensis subspecies.

Table 2 ,
. : Distinct differences in some of the key characters 0 " Bandicota indica and B. maxima.