- Email: [email protected]
Identification of mongoose (Genus: Herpestes) species from hair through band pattern studies using discriminate functional analysis (DFA) and microscopic examination
Science and Justice 49 (2009) 205–209
Contents lists available at ScienceDirect
Science and Justice j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / s c i j u s
Identification of mongoose (Genus: Herpestes) species from hair through band pattern studies using discriminate functional analysis (DFA) and microscopic examination Vivek Sahajpal a, S.P. Goyal a,⁎, R. Raza b, R. Jayapal b a b
Wildlife Forensic laboratory, Wildlife Institute of India, Post Box#18, Dehradoon-248001, India Wildlife Institute of India, Post Box#18, Dehradoon-248001, India
a r t i c l e
i n f o
Article history: Received 17 May 2008 Received in revised form 21 August 2008 Accepted 22 September 2008 Keywords: Cuticle Medulla Cross sections Discriminate functional analysis Forensic
a b s t r a c t India is home to seven species of mongoose (Herpestes sp). Mongooses are being poached primarily for their hair, which is used in the production of painting and shaving brushes. Prior to September 2002, mongooses were listed under Schedule-IV of the Wildlife (Protection) Act 1972 (India). Indiscriminate poaching of the mongoose created an immediate threat to their survival and hence mongooses have now been placed under Schedule-II of the Wildlife (Protection) Act-1972 (India). In order to convict a person under this legislation, species identification of case related samples is necessary. Four species of mongoose i.e. H. edwardsii, H. smithii, H. palustris and H. urva were characterised by performing discriminate functional analysis (DFA) on measurements of their dorsal guard hair banding pattern and by microscopic hair characteristics (Cuticular, medullar and cross section). It was possible to distinguish between the four species studied, based on both these methods. © 2008 Forensic Science Society. Published by Elsevier Ireland Ltd. All rights reserved.
1. Introduction India is one of the 12 identified mega biodiversity nations containing over 8% of the world's biological diversity [1]. Poaching of wild animals and illegal trade of their parts is one of the major threats to conservation planning in India [2]. Trade in wildlife parts and their products have become a massive business with high annual turnover rates. According to some estimates, the illegal trade in wildlife is probably the second largest illegal business in the World after drug trafficking [1,3]. One animal which has suffered from significant poaching in recent years in India is the mongoose (Herpestes sp.), whose hair is used for the production of painting and shaving brushes. Poaching of the mongoose has caused the status of the mongoose to be reclassified from schedule-IV to schedule-II of the Wildlife (Protection) Act-1972 (India) [4]. However, it is evident from the various seizures made by law enforcement agencies that the poaching of these animals is still a regular occurrence. The Wildlife Forensic laboratory of the Wildlife Institute of India, Dehradoon has received numerous cases involving the use of mongoose hair. This has prompted the present study to characterise the species from hair. The value of using microscopic hair characteristics as a tool for the purpose of species identification has been established by previous work [5–7]. However very few forensic studies on hair identification of Indian wild species have been carried [8,9]. ⁎ Corresponding author. Tel.: +91 135 2640111 115x224; fax: +91 135 2640117. E-mail address: [email protected] (S.P. Goyal).
Fig. 1. Bands considered in the study.
1355-0306/$ – see front matter © 2008 Forensic Science Society. Published by Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.scijus.2008.09.002
206
V. Sahajpal et al. / Science and Justice 49 (2009) 205–209
Fig. 2. Guard hair of four species showing the band pattern.
India has seven species of genus Herpestes viz. Herpestes edwardsii, Herpestes smithii, Herpestes urva, Herpestes palustris, Herpestes javanicus, Herpestes vitticolis and Herpestes brachyurus [10]. Table 1 Eigenvalues. Function
Eigenvalue
% of variance
Cumulative %
Canonical correlation
1 2 3
50.373 1.788 .584
95.5 3.4 1.1
95.5 98.9 100.0
.990 .801 .607
First 3 canonical discriminant functions were used in the analysis. Table 2 Classification function coefficients. Species 2n antipenultimate Antipenultimate Penultimate Apical (Constant)
H. edwardsii
H. smithii
H. urva
H. palustris
10.919 8.358 .392 2.895 − 45.796
11.095 7.284 .715 5.126 − 55.161
34.391 29.016 2.915 − 3.938E− 02 − 418.946
13.584 11.580 − .153 3.588 − 76.317
Fisher's linear discriminant functions. Table 3 Classification results. Predicted group membership Species Original
Count H. H. H. H. % H. H. H. H. CrossCount H. validated H. H. H. % H. H. H. H.
edwardsii smithii urva palustris edwardsii smithii urva palustris edwardsii smithii urva palustris edwardsii smithii urva palustris
Total
Fig. 3. Canonical discriminant functions. 94.3% of cases were correctly classified (Jack knife estimate).
The hair characteristics of five mongoose species have been reported [11], however these findings are useful only when a complete strand of hair is available. When examining suspected items such as painting brushes the roots and lower portion of hair are mostly
H. edwardsii H. smithii H. urva H. palustris 72 4 0 0 90 5 0 0 72 4 0 0 90 5 0 0
8 76 0 0 10 95 0 0 8 76 0 0 10 95 0 0
0 0 60 0 0 0 100 0 0 0 54 0 0 0 90 0
0 0 0 80 0 0 0 100 0 0 6 80 0 0 10 100
80 80 60 80 100 100 100 100 20 20 60 80 100 100 100 100
a. In cross validation, each case is classified by the functions derived from all cases other than that case. b. 95.7% of original grouped cases correctly classified. c. 94.3% of cross-validated grouped cases correctly classified (Jack knife estimate).
Table 4 Blind test. a. For banding pattern Species
Percent correctly identified based on combination of hair characteristics
H. edwardsii 90 H. smithii 90 H. urva 100 H. palustris 90 Total 92.5 b. For microscopic hair characteristics H. edwardsii H. smithii H. urva H. palustris Total
100 100 100 100 100
V. Sahajpal et al. / Science and Justice 49 (2009) 205–209
207
Fig. 4. Cuticle patterns of Herpestes edwardsii (1), Herpestes smithii (2), Herpestes urva (3) and Herpestes Palustris (4). Magnification (1000×).
trimmed as only the upper portion of the hair is useful for brush manufacturing. This work addresses this issue by examining the guard hair portion of the mongoose species. Four species of genus Herpestes
viz. H. edwardsii, H. smithii, H. urva, H. palustris have been characterised on the basis of their hair banding pattern and by microscopic examination of their guard hairs. The remaining three
Fig. 5. Medulla patterns of Herpestes edwardsii (1), Herpestes smithii (2), Herpestes urva (3) and Herpestes Palustris (4) (400×).
208
V. Sahajpal et al. / Science and Justice 49 (2009) 205–209
Fig. 6. Hair cross- sections of Herpestes edwardsii (1), Herpestes smithii (2), Herpestes urva (3) and Herpestes Palustris (4). Magnification (400×).
species prevalent in India could not be examined due to lack of reference hair samples. 2. Materials and methods Hair samples of four species of genus Herpestes viz. Indian grey mongoose (H. edwardsii), Ruddy mongoose (H. smithii), Crab eating mongoose (H. urva) and Bengal mongoose (H. palustris) were obtained from the collection of the Bombay Natural History Society (BNHS) in Mumbai, India. The dorsal guard hairs of these four species were characterised using two separate methods. Initially the hairs were visually examined for their banding patterns. The four alternate dark and light bands towards the apex of each hair were examined for each species. Hairs showing this region of variability are, in our experience, invariably found in all the brushes containing mongoose hair. These bands were designated from apex to bottom as apical band, penultimate band, 1st antepenultimate and 2nd antepenultimate band (Fig. 1). Individual hairs of each species showing the band pattern are given in Fig. 2. The lengths of individual bands were measured in millimetres (mm) for twenty guard hairs for three to four individuals of each species. Discriminate functional analysis (DFA) was performed for these band length measurements to characterise the species, using SPSS (version 12.0.0) software. This approach is non destructive and preserved the hair sample for further microscopic examination. Microscopic examination of the hair samples was undertaken using a comparison microscope (Leica DMR). All the indices for microscopic hair characteristics were generated using 40–70 hairs for each species. Cuticular, medullar and, cross-sectional patterns were studied for each species at the region of 2nd antepenultimate band. In order to obtain the cuticular impression on a microscopic glass slide, a thin film of gelatin solution in water was used. Medulla slides were
made by immersing cut hair samples in xylene for 4–5 h and subsequently mounting them in D.P.X. [9]. Scale count index values were determined as described by Kirk and Gamble [12]. Medullary index values were calculated using the following formula: Medullary index = Medulla thickness ðMTÞ = Hair Thickness ðHTÞ Cross-sections were obtained by mounting the hair tufts in wax and the cross-sections were cut manually using a shaving blade [9]. Finally a blind study was conducted for species identification through banding patterns analysis with DFA and microscopic hair characteristics using the techniques described. 3. Results and discussion The results of the discriminate functional analysis (DFA) are given in Table 1 (eigenvalues), Table 2 (classification function coefficients) and Table 3 (Classification results). The canonical discriminant functions for the four species are given in Fig. 3. The eigenvalue refers to a parameter value for which a differential equation has a non-zero solution under given conditions and represent the portion of the
Table 5 Cuticular characteristics. Species
Scale margin
Distance between scales
Scale pattern
H. H. H. H.
Crenate Crenate Crenate Crenate
Near Near Near Near
Irregular Irregular Irregular Irregular
edwardsii smithii urva palustris
wave wave wave wave
Scale count index (mean ± S.E)
Scale count index (range)
170 ± 1.23 149 ± 1.08 135 ± 1.22 114 ± 1.27
162–180 144–158 128–142 106–125
V. Sahajpal et al. / Science and Justice 49 (2009) 205–209 Table 6 Medulla and cross-section.
209
species. The cross-sectional structures have not been previously reported and their shape and medulla configuration have been found to be characteristic for the studied species (Table 6).
Species
Medulla type
Cross section
Medullary index (mean ± S.E)
Medullary index (range)
H. edwardsii
Wide, cortical intrusion Wide, cortical intrusion Wide, with vacuoles Medium, large vacuoles
Oval to oblong
0.792 ± 0.005
0.76–0.83
4. Conclusion
Oval to oblong
0.718 ± 0.003
0.70–0.74
Oblong
0.647 ± 0.004
0.63–0.68
Oval
0.596 ± 0.004
0.54–0.61
It is possible to accurately identify mongoose species from their hair band patterns using discriminate functional analysis (DFA). This type of examination has the added value of being non destructive and can be used as a preliminary screening method for the identification of the species from guard hairs often used in the production of painting and shaving brushes containing mongoose hair. Gross cuticular pattern is the same for all the four species, however significant species specific characteristics were provided by the scale count index in the region of 2nd antepenultimate band. Medulla patterns and medullary index values in the region of 2nd antepenultimate band provided the most reliable species specific characteristics. Crosssectional structure also provided a good parameter for study in the characterisation of the species. Through a combination of the band pattern and microscopic hair characteristics it is possible to distinguish between the four species studied.
H. smithii H. urva H. palustris
original total variance in the data that is solely attributable to a particular discriminant function or axis. It is evident from the eigenvalues, classification function coefficients, classification results and the canonical discriminant functions that the four species can be characterised from the band patterns with an accuracy of 94.3%. The blind test results (Table 4a) also confirmed this where the overall accuracy was 92.5% which compared well with the results for the DFA for the known samples studied. Using the band patterning with discriminant analysis is a novel method in the examination of Mongoose hairs. The use of band pattern matching suggested in this work may be limited as the study is based on guard hairs from the dorsum only. Hairs from other regions of body such as the tail are also used in brush manufacture and as such may not compare well with the reference data. However, since bulk of pelage is formed by the hairs from the dorsal and lateral hairs (which are very similar) their content in brushes is inevitably higher in comparison to tail hair. The cuticular, medullar and cross section patterns for the four species obtained by microscopic examination of the region of 2nd antepenultimate band are shown in Figs. 4–6 respectively. The cuticular characteristics are summarized in Table 5 and the medulla and cross-sectional structures are summarized in Table 6. Scale count index values were found to be species specific and non-overlapping for the four species studied as indicated by their range (Table 5). Similarly, medullary index values were also characteristic for each species with no inter-species overlap (Table 6). It is evident from the patterns that mongoose species can be characterised with microscopic hair characteristics. The blind study also indicated that the differing species could be clearly discriminated from one another on the basis of microscopic examination, (Table 4b). The scale pattern, scale distance and scale margins observed in the four species are in confirmation with the earlier studies of De et al. [12]. The medulla patterns observed in this study are also similar to previous studies [12]. The medullary index values observed for H. edwardsii are similar to earlier observations [12], however, the medulla index values for the other three species were found to be different (Table 6). Overall, these values were found to be very characteristic for the studied
Acknowledgements The authors are grateful to the Director and the Dean, Wildlife Institute of India for the support. The authors extend special thanks to Bombay Natural History Society for providing the reference hair samples of the four species of mongoose. References [1] S.C. Dey, Wildlife trade: Global perspective and the Indian scenario, Central Bureau of Investigation, bulletin, vol. 4, 1996, pp. 6–8. [2] F. Hanfee, Wildlife trade: a handbook for enforcement staff, WWF Tiger Conservation Program, vol. 4, 1998, p. 56. [3] S.K. Mukherjee, Some thoughts on wildlife trade, Cheetal 2 (1996) 30–33. [4] Indian Wildlife (Protection) Act-1972. (With amendments) A Hand guide with Case Law and Commentaries. Natraj Publishers, Dehradun, 2005 215. [5] H. Brunner, B. Coman, The Identification of Mammalian Hair, Inkata Press, Victoria, Australia, 1974, p. 196. [6] T.D. Moore, L.E. Spence, E.E. Dugnolle, Identification of the dorsal guard hairs of some mammals of Wyoming, Wyoming Game and Fish Dept, 1974, p. 77. [7] B.J. Teerink, Hair of West-European Mammals, Cambridge University Press, Cambridge, 1991, p. 223. [8] A. Bahuguna, S.K. Mukherjee, Use of SEM to recognize Tibetan antelope (Chiru) hair and blending in wool products, Science and Justice 40 (3) (2000) 177–182. [9] V. Sahajpal, S.P. Goyal, K. Jayapal, K. Yoganand, M. Thakar, Hair characteristics of four Indian bear species, Science & Justice 40 (2008) 8–15. [10] D.E. Wilson, D.M. Reeder (Eds.), Mammal Species of the World, Smithsonian Institution press, Washington, 1993, p. 1206. [11] J.K. De, S. Chakraborty, R. Chakraborty, Identification of dorsal guard hairs of five Indian species of mongoose, Herpestes Illiger (Mammalia: Carnivora), Mammalia 62 (1998) 285–295. [12] P.L. Kirk, L.H. Gamble, Further investigation of the scale count of human hair, Journal of Criminal Law and Criminology 33 (1942) 276–280.
Contents lists available at ScienceDirect
Science and Justice j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / s c i j u s
Identification of mongoose (Genus: Herpestes) species from hair through band pattern studies using discriminate functional analysis (DFA) and microscopic examination Vivek Sahajpal a, S.P. Goyal a,⁎, R. Raza b, R. Jayapal b a b
Wildlife Forensic laboratory, Wildlife Institute of India, Post Box#18, Dehradoon-248001, India Wildlife Institute of India, Post Box#18, Dehradoon-248001, India
a r t i c l e
i n f o
Article history: Received 17 May 2008 Received in revised form 21 August 2008 Accepted 22 September 2008 Keywords: Cuticle Medulla Cross sections Discriminate functional analysis Forensic
a b s t r a c t India is home to seven species of mongoose (Herpestes sp). Mongooses are being poached primarily for their hair, which is used in the production of painting and shaving brushes. Prior to September 2002, mongooses were listed under Schedule-IV of the Wildlife (Protection) Act 1972 (India). Indiscriminate poaching of the mongoose created an immediate threat to their survival and hence mongooses have now been placed under Schedule-II of the Wildlife (Protection) Act-1972 (India). In order to convict a person under this legislation, species identification of case related samples is necessary. Four species of mongoose i.e. H. edwardsii, H. smithii, H. palustris and H. urva were characterised by performing discriminate functional analysis (DFA) on measurements of their dorsal guard hair banding pattern and by microscopic hair characteristics (Cuticular, medullar and cross section). It was possible to distinguish between the four species studied, based on both these methods. © 2008 Forensic Science Society. Published by Elsevier Ireland Ltd. All rights reserved.
1. Introduction India is one of the 12 identified mega biodiversity nations containing over 8% of the world's biological diversity [1]. Poaching of wild animals and illegal trade of their parts is one of the major threats to conservation planning in India [2]. Trade in wildlife parts and their products have become a massive business with high annual turnover rates. According to some estimates, the illegal trade in wildlife is probably the second largest illegal business in the World after drug trafficking [1,3]. One animal which has suffered from significant poaching in recent years in India is the mongoose (Herpestes sp.), whose hair is used for the production of painting and shaving brushes. Poaching of the mongoose has caused the status of the mongoose to be reclassified from schedule-IV to schedule-II of the Wildlife (Protection) Act-1972 (India) [4]. However, it is evident from the various seizures made by law enforcement agencies that the poaching of these animals is still a regular occurrence. The Wildlife Forensic laboratory of the Wildlife Institute of India, Dehradoon has received numerous cases involving the use of mongoose hair. This has prompted the present study to characterise the species from hair. The value of using microscopic hair characteristics as a tool for the purpose of species identification has been established by previous work [5–7]. However very few forensic studies on hair identification of Indian wild species have been carried [8,9]. ⁎ Corresponding author. Tel.: +91 135 2640111 115x224; fax: +91 135 2640117. E-mail address: [email protected] (S.P. Goyal).
Fig. 1. Bands considered in the study.
1355-0306/$ – see front matter © 2008 Forensic Science Society. Published by Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.scijus.2008.09.002
206
V. Sahajpal et al. / Science and Justice 49 (2009) 205–209
Fig. 2. Guard hair of four species showing the band pattern.
India has seven species of genus Herpestes viz. Herpestes edwardsii, Herpestes smithii, Herpestes urva, Herpestes palustris, Herpestes javanicus, Herpestes vitticolis and Herpestes brachyurus [10]. Table 1 Eigenvalues. Function
Eigenvalue
% of variance
Cumulative %
Canonical correlation
1 2 3
50.373 1.788 .584
95.5 3.4 1.1
95.5 98.9 100.0
.990 .801 .607
First 3 canonical discriminant functions were used in the analysis. Table 2 Classification function coefficients. Species 2n antipenultimate Antipenultimate Penultimate Apical (Constant)
H. edwardsii
H. smithii
H. urva
H. palustris
10.919 8.358 .392 2.895 − 45.796
11.095 7.284 .715 5.126 − 55.161
34.391 29.016 2.915 − 3.938E− 02 − 418.946
13.584 11.580 − .153 3.588 − 76.317
Fisher's linear discriminant functions. Table 3 Classification results. Predicted group membership Species Original
Count H. H. H. H. % H. H. H. H. CrossCount H. validated H. H. H. % H. H. H. H.
edwardsii smithii urva palustris edwardsii smithii urva palustris edwardsii smithii urva palustris edwardsii smithii urva palustris
Total
Fig. 3. Canonical discriminant functions. 94.3% of cases were correctly classified (Jack knife estimate).
The hair characteristics of five mongoose species have been reported [11], however these findings are useful only when a complete strand of hair is available. When examining suspected items such as painting brushes the roots and lower portion of hair are mostly
H. edwardsii H. smithii H. urva H. palustris 72 4 0 0 90 5 0 0 72 4 0 0 90 5 0 0
8 76 0 0 10 95 0 0 8 76 0 0 10 95 0 0
0 0 60 0 0 0 100 0 0 0 54 0 0 0 90 0
0 0 0 80 0 0 0 100 0 0 6 80 0 0 10 100
80 80 60 80 100 100 100 100 20 20 60 80 100 100 100 100
a. In cross validation, each case is classified by the functions derived from all cases other than that case. b. 95.7% of original grouped cases correctly classified. c. 94.3% of cross-validated grouped cases correctly classified (Jack knife estimate).
Table 4 Blind test. a. For banding pattern Species
Percent correctly identified based on combination of hair characteristics
H. edwardsii 90 H. smithii 90 H. urva 100 H. palustris 90 Total 92.5 b. For microscopic hair characteristics H. edwardsii H. smithii H. urva H. palustris Total
100 100 100 100 100
V. Sahajpal et al. / Science and Justice 49 (2009) 205–209
207
Fig. 4. Cuticle patterns of Herpestes edwardsii (1), Herpestes smithii (2), Herpestes urva (3) and Herpestes Palustris (4). Magnification (1000×).
trimmed as only the upper portion of the hair is useful for brush manufacturing. This work addresses this issue by examining the guard hair portion of the mongoose species. Four species of genus Herpestes
viz. H. edwardsii, H. smithii, H. urva, H. palustris have been characterised on the basis of their hair banding pattern and by microscopic examination of their guard hairs. The remaining three
Fig. 5. Medulla patterns of Herpestes edwardsii (1), Herpestes smithii (2), Herpestes urva (3) and Herpestes Palustris (4) (400×).
208
V. Sahajpal et al. / Science and Justice 49 (2009) 205–209
Fig. 6. Hair cross- sections of Herpestes edwardsii (1), Herpestes smithii (2), Herpestes urva (3) and Herpestes Palustris (4). Magnification (400×).
species prevalent in India could not be examined due to lack of reference hair samples. 2. Materials and methods Hair samples of four species of genus Herpestes viz. Indian grey mongoose (H. edwardsii), Ruddy mongoose (H. smithii), Crab eating mongoose (H. urva) and Bengal mongoose (H. palustris) were obtained from the collection of the Bombay Natural History Society (BNHS) in Mumbai, India. The dorsal guard hairs of these four species were characterised using two separate methods. Initially the hairs were visually examined for their banding patterns. The four alternate dark and light bands towards the apex of each hair were examined for each species. Hairs showing this region of variability are, in our experience, invariably found in all the brushes containing mongoose hair. These bands were designated from apex to bottom as apical band, penultimate band, 1st antepenultimate and 2nd antepenultimate band (Fig. 1). Individual hairs of each species showing the band pattern are given in Fig. 2. The lengths of individual bands were measured in millimetres (mm) for twenty guard hairs for three to four individuals of each species. Discriminate functional analysis (DFA) was performed for these band length measurements to characterise the species, using SPSS (version 12.0.0) software. This approach is non destructive and preserved the hair sample for further microscopic examination. Microscopic examination of the hair samples was undertaken using a comparison microscope (Leica DMR). All the indices for microscopic hair characteristics were generated using 40–70 hairs for each species. Cuticular, medullar and, cross-sectional patterns were studied for each species at the region of 2nd antepenultimate band. In order to obtain the cuticular impression on a microscopic glass slide, a thin film of gelatin solution in water was used. Medulla slides were
made by immersing cut hair samples in xylene for 4–5 h and subsequently mounting them in D.P.X. [9]. Scale count index values were determined as described by Kirk and Gamble [12]. Medullary index values were calculated using the following formula: Medullary index = Medulla thickness ðMTÞ = Hair Thickness ðHTÞ Cross-sections were obtained by mounting the hair tufts in wax and the cross-sections were cut manually using a shaving blade [9]. Finally a blind study was conducted for species identification through banding patterns analysis with DFA and microscopic hair characteristics using the techniques described. 3. Results and discussion The results of the discriminate functional analysis (DFA) are given in Table 1 (eigenvalues), Table 2 (classification function coefficients) and Table 3 (Classification results). The canonical discriminant functions for the four species are given in Fig. 3. The eigenvalue refers to a parameter value for which a differential equation has a non-zero solution under given conditions and represent the portion of the
Table 5 Cuticular characteristics. Species
Scale margin
Distance between scales
Scale pattern
H. H. H. H.
Crenate Crenate Crenate Crenate
Near Near Near Near
Irregular Irregular Irregular Irregular
edwardsii smithii urva palustris
wave wave wave wave
Scale count index (mean ± S.E)
Scale count index (range)
170 ± 1.23 149 ± 1.08 135 ± 1.22 114 ± 1.27
162–180 144–158 128–142 106–125
V. Sahajpal et al. / Science and Justice 49 (2009) 205–209 Table 6 Medulla and cross-section.
209
species. The cross-sectional structures have not been previously reported and their shape and medulla configuration have been found to be characteristic for the studied species (Table 6).
Species
Medulla type
Cross section
Medullary index (mean ± S.E)
Medullary index (range)
H. edwardsii
Wide, cortical intrusion Wide, cortical intrusion Wide, with vacuoles Medium, large vacuoles
Oval to oblong
0.792 ± 0.005
0.76–0.83
4. Conclusion
Oval to oblong
0.718 ± 0.003
0.70–0.74
Oblong
0.647 ± 0.004
0.63–0.68
Oval
0.596 ± 0.004
0.54–0.61
It is possible to accurately identify mongoose species from their hair band patterns using discriminate functional analysis (DFA). This type of examination has the added value of being non destructive and can be used as a preliminary screening method for the identification of the species from guard hairs often used in the production of painting and shaving brushes containing mongoose hair. Gross cuticular pattern is the same for all the four species, however significant species specific characteristics were provided by the scale count index in the region of 2nd antepenultimate band. Medulla patterns and medullary index values in the region of 2nd antepenultimate band provided the most reliable species specific characteristics. Crosssectional structure also provided a good parameter for study in the characterisation of the species. Through a combination of the band pattern and microscopic hair characteristics it is possible to distinguish between the four species studied.
H. smithii H. urva H. palustris
original total variance in the data that is solely attributable to a particular discriminant function or axis. It is evident from the eigenvalues, classification function coefficients, classification results and the canonical discriminant functions that the four species can be characterised from the band patterns with an accuracy of 94.3%. The blind test results (Table 4a) also confirmed this where the overall accuracy was 92.5% which compared well with the results for the DFA for the known samples studied. Using the band patterning with discriminant analysis is a novel method in the examination of Mongoose hairs. The use of band pattern matching suggested in this work may be limited as the study is based on guard hairs from the dorsum only. Hairs from other regions of body such as the tail are also used in brush manufacture and as such may not compare well with the reference data. However, since bulk of pelage is formed by the hairs from the dorsal and lateral hairs (which are very similar) their content in brushes is inevitably higher in comparison to tail hair. The cuticular, medullar and cross section patterns for the four species obtained by microscopic examination of the region of 2nd antepenultimate band are shown in Figs. 4–6 respectively. The cuticular characteristics are summarized in Table 5 and the medulla and cross-sectional structures are summarized in Table 6. Scale count index values were found to be species specific and non-overlapping for the four species studied as indicated by their range (Table 5). Similarly, medullary index values were also characteristic for each species with no inter-species overlap (Table 6). It is evident from the patterns that mongoose species can be characterised with microscopic hair characteristics. The blind study also indicated that the differing species could be clearly discriminated from one another on the basis of microscopic examination, (Table 4b). The scale pattern, scale distance and scale margins observed in the four species are in confirmation with the earlier studies of De et al. [12]. The medulla patterns observed in this study are also similar to previous studies [12]. The medullary index values observed for H. edwardsii are similar to earlier observations [12], however, the medulla index values for the other three species were found to be different (Table 6). Overall, these values were found to be very characteristic for the studied
Acknowledgements The authors are grateful to the Director and the Dean, Wildlife Institute of India for the support. The authors extend special thanks to Bombay Natural History Society for providing the reference hair samples of the four species of mongoose. References [1] S.C. Dey, Wildlife trade: Global perspective and the Indian scenario, Central Bureau of Investigation, bulletin, vol. 4, 1996, pp. 6–8. [2] F. Hanfee, Wildlife trade: a handbook for enforcement staff, WWF Tiger Conservation Program, vol. 4, 1998, p. 56. [3] S.K. Mukherjee, Some thoughts on wildlife trade, Cheetal 2 (1996) 30–33. [4] Indian Wildlife (Protection) Act-1972. (With amendments) A Hand guide with Case Law and Commentaries. Natraj Publishers, Dehradun, 2005 215. [5] H. Brunner, B. Coman, The Identification of Mammalian Hair, Inkata Press, Victoria, Australia, 1974, p. 196. [6] T.D. Moore, L.E. Spence, E.E. Dugnolle, Identification of the dorsal guard hairs of some mammals of Wyoming, Wyoming Game and Fish Dept, 1974, p. 77. [7] B.J. Teerink, Hair of West-European Mammals, Cambridge University Press, Cambridge, 1991, p. 223. [8] A. Bahuguna, S.K. Mukherjee, Use of SEM to recognize Tibetan antelope (Chiru) hair and blending in wool products, Science and Justice 40 (3) (2000) 177–182. [9] V. Sahajpal, S.P. Goyal, K. Jayapal, K. Yoganand, M. Thakar, Hair characteristics of four Indian bear species, Science & Justice 40 (2008) 8–15. [10] D.E. Wilson, D.M. Reeder (Eds.), Mammal Species of the World, Smithsonian Institution press, Washington, 1993, p. 1206. [11] J.K. De, S. Chakraborty, R. Chakraborty, Identification of dorsal guard hairs of five Indian species of mongoose, Herpestes Illiger (Mammalia: Carnivora), Mammalia 62 (1998) 285–295. [12] P.L. Kirk, L.H. Gamble, Further investigation of the scale count of human hair, Journal of Criminal Law and Criminology 33 (1942) 276–280.