legacy of draught cattle breeds of south india: insights into … · 2021. 1. 21. · 2 35 legacy...

1

1 Title : Legacy of draught cattle breeds of South India: Insights into population

2 structure, genetic admixture and maternal origin

3

4 Authors : Vandana Manomohan1,2, Ramasamy Saravanan 2, Rudolf Pichler1, Nagarajan

5 Murali2, Karuppusamy Sivakumar2, Krovvidi Sudhakar3, Raja K

6 Nachiappan4 and Kathiravan Periasamy1,5*

78 Affiliations : 1Animal Production and Health Section, Joint FAO/IAEA Division of

9 Nuclear Techniques in Food and Agriculture, International Atomic Energy

10 Agency, Vienna, Austria

11 2Veterinary College and Research Institute, Namakkal, Tamil Nadu

12 Veterinary and Animal Sciences University, Chennai, India;

13 3NTR College of Veterinary Science, Gannavaram, Sri Venkateswara

14 Veterinary University, Tirupati, Andhra Pradesh, India;

15 4National Bureau of Animal Genetic Resources, Karnal, Haryana, India;

16 5Animal Genetics Resources Branch, Animal Production and Health

17 Division, Food and Agriculture Organization of the United Nations, Rome,

18 Italy.

19

20 Corresponding: Kathiravan Periasamy21 author22

23

24 Address : Animal Production and Health Laboratory, Joint FAO/IAEA Division of

25 Nuclear Techniques in Food and Agriculture, International Atomic Energy

26 Agency, Seibersdorf, Vienna, Austria

27

28

29 Email : [email protected]; [email protected]

30

31

32 Telephone : 00431260027328

33

34

.CC-BY 4.0 International licenseperpetuity. It is made available under apreprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in

The copyright holder for thisthis version posted January 21, 2021. ; https://doi.org/10.1101/2021.01.21.427560doi: bioRxiv preprint

https://doi.org/10.1101/2021.01.21.427560http://creativecommons.org/licenses/by/4.0/

2

35 Legacy of draught cattle breeds of South India: Insights into population structure,

36 genetic admixture and maternal origin

37

38 Vandana Manomohan1,2, Ramasamy Saravanan2, Rudolf Pichler1, Nagarajan Murali2,

39 Karuppusamy Sivakumar2, Krovvidi Sudhakar3, Raja K Nachiappan4

40 and Kathiravan Periasamy1,5*

41 1Animal Production and Health Section, Joint FAO/IAEA Division, International Atomic

42 Energy Agency, Vienna, Austria; 2Veterinary College and Research Institute, Namakkal,

43 Tamil Nadu Veterinary and Animal Sciences University, Chennai, India; 3NTR College of

44 Veterinary Science, Gannavaram, Sri Venkateswara Veterinary University, Tirupati, Andhra

45 Pradesh, India; 4National Bureau of Animal Genetic Resources, Karnal, Haryana, India;

46 5Animal Genetics Resources Branch, Animal Production and Health Division, Food and

47 Agriculture Organization of the United Nations, Rome, Italy.

48 *Corresponding author email: [email protected]; [email protected] 49

50

51 ABSTRACT

52 The present study is the first comprehensive report on diversity, population structure, genetic

53 admixture and mitochondrial DNA variation in South Indian draught type zebu cattle. The

54 diversity of South Indian cattle was moderately higher. A significantly strong negative

55 correlation coefficient of -0.674 (P6.25%) was observed in Punganur, Vechur, Umblachery and Pulikulam cattle breeds. Two

63 major maternal haplogroups, I1 and I2, typical of zebu cattle were observed, with the former

64 being predominant than the later. The pairwise differences among the I2 haplotypes of South




3

65 Indian cattle were relatively higher than West Indian (Indus valley site) zebu cattle. The results

66 indicated the need for additional sampling and comprehensive analysis of mtDNA control

67 region variations to unravel the probable location of origin and domestication of I2 zebu

68 lineage. The present study also revealed major concerns on South Indian zebu cattle (i) risk of

69 endangerment due to small effective population size and high rate of inbreeding (ii) lack of

70 sufficient purebred zebu bulls for breeding and (iii) increasing level of taurine admixture in

71 zebu cattle. Availability of purebred semen for artificial insemination, incorporation of

72 genomic/molecular information to identify purebred animals and increased awareness among

73 farmers will help to maintain breed purity, conserve and improve these important draught cattle

74 germplasms of South India.

75

76 Key words: Zebu, diversity, inbreeding, genetic relationship, taurine introgression,

77 haplogroup, mismatch distribution, domestication

78

79 1. INTRODUCTION

80 Ever since domestication, animals have been used as an important source of draught power for

81 agricultural production. Among them, cattle have been used for a variety of farm operations

82 like ploughing, carting, tilling, sowing, weeding, water lifting, threshing, oil extraction,

83 sugarcane crushing and transport. Over time, with increasing mechanization of agriculture, the

84 use of draught cattle in farm operations reduced significantly and became negligible in

85 advanced economies. However, in developing countries, particularly in small holder

86 production systems and in hilly and mountainous regions, agricultural production still relies on

87 draught animal power (Zhou et al. 2018). For example, 55% of the smallholder farmers in

88 Swaziland rely on draft animal power for land cultivation, and more than 88% of draft animals

89 found on the Swazi National Land are cattle (Kienzle et al. 2013). In India, the significance of

90 draught animal power has reduced greatly in the last few decades with its share in total farm

91 power availability declining from about 78% in 1960-61 to 7% in 2013-2014 (Singh et al. 2014)




4

92 and expected to further reduce to around 5% by 2020 (Natarajan et al. 2016). Over the years,

93 the annual use of draught animals declined from about 1200-1800 hours to 300-500 hours per

94 pair per year. Consequently, the population of draught cattle steadily decreased at a negative

95 annual compound growth rate of about -0.8%, resulting in significant decline of purebred native

96 breeds.

97 India is home to the largest cattle population (13.1 % of world’s cattle population) in the world

98 which constitutes 37.3 % of its total livestock. As per the 19th livestock census, India has 190.9

99 million cattle, out of which 151 million is indigenous cattle. The diversity of Indian cattle is

100 represented by 60 local breeds, eight regional transboundary breeds and seven international

101 transboundary breeds (FAO, 2015). Of these, 41 have been registered by National Bureau of

102 Animal Genetic Resources, the nodal agency for the registration of livestock breeds in India.

103 The Indian Zebu cattle (Bos indicus) breeds have been classified into three major groups: dairy

104 type, draught type and dual-purpose cattle. The dairy type and dual-purpose cattle breeds are

105 predominantly distributed in North and North-Western India, while most indigenous draught

106 type cattle breeds are located in Southern and Eastern India. Apart from rapid mechanization

107 of agriculture in South India, the increasing cost of maintaining draught type cattle forced the

108 farmers to replace these animals with crossbred cattle that fetched better returns from the sale

109 of milk to dairy cooperatives. During late 1980s and early 1990s, the draught cattle breeds

110 started losing their economic relevance and the farmers increasingly resorted to crossbreeding

111 with commercial taurine cattle like Jersey, Holstein-Friesian and Brown Swiss. Such crossbred

112 cows approximately yielded more than double the milk as that of indigenous draught breeds.

113 Combined with farmers’ preference for crossbreds, the relatively easy access to breeding

114 services (through state animal husbandry departments, state cooperatives and private

115 inseminators, etc.) and availability of better infrastructure for artificial insemination and

116 veterinary care, the proportion of crossbred cattle among the breedable females became much




5

117 higher in South Indian states than rest of India. The crossbred cattle constitute 35.9% of total

118 cattle in Andhra Pradesh, 41.9% in Karnataka, 76.1% in Tamil Nadu and 94.6% in Kerala as

119 against the national average of 27.7%. This resulted in significant genetic dilution of draught

120 cattle breeds of South India with varying levels of zebu-taurine admixture.

121 Indiscriminate crossbreeding, lack of stabilization of exotic inheritance and absence of

122 selection among F2 hybrids led to negative impacts like increased disease susceptibility,

123 decreased fertility and drop in production levels among crossbred cattle. During the last two

124 decades, there has been a renewed interest in the conservation of purebred indigenous draught

125 cattle breeds and implementation of selective breeding programs targeting a range of traits.

126 Selective breeding schemes among certain purebred native breeds like Kangayam, Ongole,

127 Deoni and Hallikar were put in place by making available the purebred frozen semen for

128 artificial insemination. Such programs have been constrained by the availability of superior

129 genetic merit bulls and unavailability of sufficient of number of true to the type, purebred bulls.

130 With the absence of pedigree records, the level of taurine inheritance in most purebred draught

131 cattle populations is not clearly understood. Further, information on genetic differentiation,

132 population structure and maternal inheritance among South Indian cattle breeds is scanty.

133 Hence, the present study was undertaken with the following objectives: (i) assess biodiversity

134 and population structure among purebred South Indian draught cattle breeds (ii) estimate level

135 of genetic admixture among purebred zebu cattle and (iii) identify mitochondrial DNA based

136 maternal lineages and evaluate phylogeography of South Indian cattle.

137

138 2. MATERIALS AND METHODS

139 2.1. Animal ethics statement

140 Blood samples were collected by jugular venipuncture into EDTA vacutainer tubes from

141 various purebred zebu, commercial taurine and crossbred cattle. Blood sampling was




6

142 performed by local veterinarians in the respective native breed tracts following the standard

143 good animal practice. The consent was obtained from animal owners for the usage of the

144 samples in this study. Therefore, no further license from the “Institutional Committee for Care

145 and Use of Experimental Animals” of the Tamil Nadu Veterinary and Animal Sciences

146 University, Chennai, India and the Joint FAO/IAEA Division, International Atomic Energy

147 Agency, Vienna, Austria was required.

148 2.2. Sample collection

149 A total of 529 blood samples were collected from cattle representing 14 breeds/populations

150 that are located in South/Peninsular Indian states: Ongole (n=51) and Punganur (n=18) from

151 Andhra Pradesh; Hallikar (n=36) and Deoni (n=53) from Karnataka; Vechur (n=26) from

152 Kerala; Alambadi (n=29), Bargur (n=52), Kangayam (n=51), Pulikulam (n=34), Umblachery

153 (n=33), Holstein-Friesian (n=15), Jersey (n=34), Holstein-Friesian crossbred (n=39) and Jersey

154 crossbred (n=58) from Tamil Nadu. A stratified random sampling procedure was followed to

155 collect blood from unrelated cattle with typical phenotypic features and located in randomly

156 selected villages of the native breed tract. With the absence of pedigree records in most

157 instances, unrelatedness was ensured by interviewing the farmers about the breeding history.

158 Genomic DNA was extracted by standard phenol chloroform method (Sambrook and Russell,

159 2001). The DNA samples were subjected to quality control using Nanodrop spectrophotometer

160 and stored at −20ºC until further processing.

161 2.3. Microsatellite genotyping

162 All the DNA samples were subjected to PCR amplification using 27 FAO recommended

163 microsatellite markers (FAO, 2011; Supplementary Table ST1). The forward primers

164 conjugated to one of the three fluorescent dyes (FAM, HEX and ATTO550) were used for

165 genotyping these marker loci. Polymerase chain reaction was performed under following

166 conditions: initial denaturation for 5 min at 950C, followed by 35 cycles of denaturation for 30s




7

167 at 950C, 1 minute at specific annealing temperatures of marker locus, elongation at 720C for 1

168 min with the final extension at 720C for 10 min. The PCR products were then electrophoresed

169 after multiplexing in an automated DNA analyzer ABI3100 (Applied Biosystems, USA) with

170 ROX500 (Applied Biosystems, USA) as internal lane control. All the 27 STR loci were

171 multiplexed in six sets for genotyping as shown in Supplementary Table ST1. Determination

172 of allele size and extraction of genotype data was performed using GENEMAPPER software.

173 2.4. Sequencing mitochondrial DNA control region

174 Primers were custom designed and synthesized to amplify 1207 bp bovine mitochondrial DNA

175 control region using Primer 3 version 4.0 (http://bioinfo.ut.ee/primer3-0.4.0/). The amplified

176 mtDNA control region included nucleotide positions 15167 to 30 of the complete bovine

177 mitochondrial genome sequence available at NCBI GenBank accession number AF492350

178 (Hiendleder et al. 2008). The location and sequence of forward and reverse primers are as

179 follows: BTMTD1-F (Positions 15167-15186; 5’ AGGACAACCAGTCGAACACC 3’) and

180 BTMTD1-R 5’ (Positions 11-30; GTGCCTTGCTTTGGGTTAAG 3’). The amplicon included

181 complete D-loop (control region) flanked by partial cytochrome B, complete t-RNA-Thr and

182 complete t-RNA-Pro coding sequences at 5’ end while partial t-RNA-Phe sequence flanked the

183 3’end. Polymerase chain reaction was performed in a total reaction volume of 40µl with the

184 following cycling conditions: initial denaturation at 95°C for 15 min followed by 30 cycles of

185 95°C for 1 min; 58°C for 1 min; 72°C for 1 min with final extension at 72°C for 10 min.

186 Purified PCR products were sequenced using Big Dye Terminator Cycle Sequencing Kit

187 (Applied Biosystems, U.S.A) on an automated Genetic Analyzer ABI 3100 (Applied

188 Biosystems, U.S.A).

189 2.5. Statistical analysis

190 The presence of null alleles in the dataset was checked using Microchecker version 2.2.3

191 (Oosterhout et al. 2004). The neutrality of microsatellite markers used in the present study was




8

192 evaluated by FST outlier approach as implemented in LOSITAN version 1 for windows (Antao

193 et al. 2008). Basic diversity indices (allelic diversity and heterozygosity) and genetic distances

194 (pairwise Nei’s genetic distance and pairwise inter-individual allele sharing distance) among

195 South Indian cattle were estimated using MICROSATELLITE ANALYZER (MSA) version

196 4.05 (Dieringer and Schlotterer, 2003). A UPGMA tree was constructed based on pairwise

197 Nei’s genetic distances using PHYLIP version 3.5 (Felsenstein, 1993). F statistics for each

198 locus (Weir and Cockerman, 1984) was calculated and tested using the program FSTAT

199 version 2.9.3.2. The exact tests for Hardy-Weinberg Equilibrium to evaluate both

200 heterozygosity deficit and excess at each marker locus in each population (HWE) were

201 performed using GENEPOP software. Analysis of molecular variance (AMOVA) was

202 performed using ARLEQUIN version 3.1 (Excoffier et al. 2005). Multidimensional scaling

203 display of pairwise FST was conducted using SPSS version 13.0. Bayesian clustering analysis

204 was performed to evaluate population structure and genetic admixture among South Indian

205 cattle using STRUCTURE 2.3.4 (Pritchard et al. 2000). The program was run with the

206 assumption of K=2 to K=15 with at least 20 runs for each K. The number of burn in periods

207 and MCMC repeats used for these runs were 200000 and 200000 respectively. The procedure

208 of Evanno et al. (2005) was used to estimate the second order rate of change in LnP(D) and

209 identify the optimal K.

210 Mitochondrial DNA sequence editing, assembly of contigs and multiple sequence alignments

211 were performed using Codon Code Aligner version 3.7.1. Mitochondrial DNA diversity

212 parameters including nucleotide diversity, haplotype diversity, parsimony informative sites and

213 average number of nucleotide differences were calculated using DnaSP, version 4.10 (Rozas,

214 2009). Frequency of different haplotypes, haplotype sharing and pairwise FST were estimated

215 using ARLEQUIN 2.1. MEGA version 6.0 (Tamura et al. 2011) was utilized to identify optimal

216 DNA substitution model and construct maximum-likelihood phylogeny. HKY+I+G was




9

217 identified as the optimal model for the current sequence dataset based on Akaike information

218 criterion (AIC). The following sequences were used as references for different maternal

219 lineages reported in zebu and taurine cattle: I1 (AB085923, AB268563, AB268579,

220 EF417970, EF417971 and L27733), I2 (AB085921, AB268559, AB268570, AB268573,

221 AB268581 and AY378133), T1 (DQ124399, EU177841 and EU177842), T2 (DQ124383,

222 DQ124393, EU177849 and EU177851), T3 (EU177815, EU177816, EU177838 and

223 EU177839), T4 (DQ124372, DQ124375, DQ124377 and DQ124392), and T5 (EU177862,

224 EU177863, EU177864 and EU177865). Additionally, 298 mtDNA control region sequences

225 deposited to GenBank from Bhutan (n=121), Nepal (n=14), China (n=47), North India (n=53),

226 West India (n=45) and Vietnam (n=18) were utilized for comparative analysis. The accession

227 number and details of GenBank sequences used in the study are provided in Supplementary

228 Table ST2. Pairwise FST derived from mtDNA haplotype frequency were utilized to perform

229 principal components analysis using SPSS version 13.0. Mismatch distribution and the tests

230 for demographic expansion were performed using ARLEQUIN 3.1. Three parameters viz. tau,

231 theta 0 and theta 1 were estimated under the sudden expansion model and the tests for goodness

232 of fit were performed based on sum of square deviations (SSD) between the observed and

233 expected mismatch and Harpending’s raggedness index. The South Indian cattle breeds were

234 subjected to tests for selective neutrality through estimation of Tajima’s D and Fu’s FS. Median

235 Joining (MJ) network of South Indian cattle haplotypes was reconstructed using NETWORK

236 4.5.1.2 (Bandelt et al. 1999).

237

238 3. RESULTS

239 3.1. Basic diversity measures and Hardy-Weinberg equilibrium

240 In the present study, a total of 13932 genotypes were generated across 27 microsatellite loci in

241 516 cattle belonging to 10 draught type zebu breeds, 2 commercial taurine breeds and 2




10

242 crossbred populations. The mean basic diversity indices of different South Indian cattle breeds

243 are presented in Table 1. Allelic diversity in terms of mean observed number of alleles varied

244 from 5.93 (Punganur) to 8.07 (Pulikulam) across different zebu cattle breeds, while it was 4.7

245 and 5.81 in the commercial taurine breeds, Jersey and Holstein-Friesian respectively. The

246 observed heterozygosity ranged from 0.598 (Kangayam) to 0.671 (Alambadi and Punganur)

247 while the expected heterozygosity varied between 0.637 (Kangayam) and 0.727 (Punganur)

248 among the South Indian zebu cattle. Similarly, allelic richness was lowest in Kangayam and

249 highest in Pulikulam cattle. Assessment of within breed diversity based on inter-individual

250 allele sharing distance revealed Kangayam with the lowest mean distance among individuals

251 (0.568) while Pulikulam cattle showed the highest mean distance (0.647). The allelic diversity

252 observed among South Indian zebu cattle was relatively low as compared to that of West-

253 Central Indian (Shah et al. 2012), North Indian (Sharma et al. 2015) and East Indian (Sharma

254 et al. 2013) cattle. The primary domestication centre of zebu (Bos indicus) cattle in North-

255 Western part of India near Indus valley could explain the higher diversity of North Indian cattle

256 compared to that of South Indian breeds. Despite the relatively low allelic diversity, the

257 observed and expected heterozygosity of South Indian cattle were moderately higher (>60%)

258 and comparable to that of West-Central, North and East Indian cattle (Shah et al. 2012; Sharma

259 et al. 2013, 2015). As expected, the crossbred cattle (Jersey crossbreds and Holstein-Friesian

260 crossbreds) exhibited a higher level of allelic diversity and heterozygosity than indigenous zebu

261 cattle.

262 The mean estimated FIS values were positive in all the zebu cattle breeds and ranged between

263 0.027 (Hallikar) and 0.108 (Pulikulam). Breeds like Umblachery, Vechur, Ongole and Bargur

264 showed mean FIS greater than 7% indicating significant heterozygosity deficit, possibly due to

265 close breeding and potential inbreeding. The mean FIS estimates observed in most South Indian

266 cattle breeds were higher than those reported for East Indian (Sharma et al. 2013) and most




11

267 North Indian cattle with the exception of Mewati and Gaolao (Sharma et al. 2015). However,

268 the mean FIS estimates reported for West-Central Indian cattle were much higher than that of

269 South Indian cattle (Shah et al. 2012). Similarly, the mean FIS for Vechur and Ongole cattle

270 observed in the present study was significantly lower than reported previously by Radhika et

271 al. (2018) and Sharma et al. (2015) respectively. The test for Hardy-Weinberg equilibrium

272 (HWE) was conducted on a total of 378 breed X locus combinations. Among the 270 breed X

273 locus combinations tested for zebu cattle, about 62 (22.96%) deviated significantly (P

12

292 significant outliers with high FST values, indicating positive selection in force (Supplementary

293 Figure SF1). Of these, HEL5 exhibited significant heterozygosity deficit and consequent

294 deviation from HWE in 12 out of 14 studied populations. Both the loci have been reported to

295 be associated with production and/or reproduction traits in cattle (de Oliveira et al. 2005; Li et

296 al. 2010; Vanessa et al. 2018). Specifically, the homozygous long alleles at HEL5 (allele size

297 159-169) were significantly associated (p=0.022) with long calving interval in Brangus-Ibage

298 cattle, a composite breed with taurine and zebu ancestry (de Oliveira et al. 2005). In the present

299 study, significant difference in the frequency of long (allele size 160-180) and short alleles

300 (allele size 148-158) was observed among the zebu, taurine and crossbred cattle, resulting in

301 higher estimation of FST. Selective processes affect neutral loci when the latter is in linkage

302 disequilibrium with other loci subjected to selection. Considering the possibility of selective

303 forces operating at the two FST outlier loci, HEL5 and HEL13 were removed from subsequent

304 analysis of genetic relationship, admixture and population structure of South Indian cattle.

305 3.2. Genetic differentiation and phylogeny

306 The global F statistics among the South Indian cattle breeds are presented in Supplementary

307 Table ST3. The overall global FIT, FST and FIS among zebu, taurine and crossbred cattle was

308 0.143, 0.101 and 0.047 respectively, while those estimates among zebu cattle alone were 0.109,

309 0.057 and 0.056 respectively. The results indicated 10.1% of the total genetic variation among

310 all the studied cattle were due to between breed differences and it reduced to 5.7% when only

311 zebu cattle breeds were considered. The global FST among all cattle breeds/populations ranged

312 from 0.052 (INRA005) to 0.223 (ETH152), whereas among zebu cattle breeds, it varied

313 between 0.024 (INRA005) and 0.114 (HAUT24). Among the investigated loci, ETH152

314 provided relatively more information on zebu-taurine divide while HAUT24 provided

315 significant information on differences among zebu cattle breeds. The global FST among South

316 Indian zebu cattle observed in the present study was comparable to that of East Indian zebu




13

317 (Sharma et al. 2013) cattle. However, much higher genetic differentiation had been reported

318 among West-Central (Shah et al. 2012) and North Indian (Sharma et al. 2015) cattle breeds.

319 The pairwise FST and Nei’s genetic distance among the draught type zebu, taurine and crossbred

320 cattle are presented in Table 2. The pairwise FST ranged from 0.0234 (Hallikar-Alambadi) to

321 0.107 (Kangayam-Punganur) among the South Indian zebu cattle breeds. Alambadi cattle was

322 observed to be genetically less differentiated from Hallikar (FST=0.0234), Pulikulam

323 (FST=0.0237) and Bargur (FST=0.0259) while Kangayam cattle showed highest differentiation

324 from Punganur (FST=0.107), Vechur (FST=0.1034) and Hallikar (FST=0.0933). Interestingly,

325 significantly high genetic differentiation (FST=0.0919) was observed among the two

326 dwarf/miniature type Vechur and Punganur cattle. Similar findings were observed with respect

327 to the estimates of pairwise Nei’s genetic distance among the South Indian zebu cattle breeds.

328 The radial tree and dendrogram constructed following UPGMA algorithm based on pairwise

329 Nei’s genetic distance is presented in Figure 1. Broadly, the taurine, zebu and their crossbred

330 cattle clustered separately as expected. Among the zebu cattle, a major cluster was formed by

331 Alambadi, Bargur, Umblachery, Pulikulam, Deoni, Ongole and Hallikar, within which three

332 sub-clusters were observed. Alambadi and Bargur formed the first sub-cluster, Umblachery

333 and Pulikulam formed the second sub-cluster while Deoni, Ongole and Hallikar formed the

334 third sub-cluster. Kangayam, Vechur and Punganur cattle breeds did not form part of any of

335 these three clusters and were distinct among themselves.

336 3.3. Genetic relationship and population structure

337 Understanding genetic relationship and detecting the cryptic genetic structure is an important

338 first step to elucidate the demography and ancestral history of cattle populations. The

339 multidimensional scaling (MDS) analysis were used to investigate the relationship among the

340 studied cattle breeds. Multidimensional scaling (MDS) is an ordination technique used to

341 display information contained in a distance matrix and to visualize relationship among




14

342 variables within a dataset. This method helps to reduce the dimension of data and to detect

343 structure among populations in a typical genetic diversity study. The MDS plots derived from

344 pairwise FST values are presented in Figure 2. The observed s-stress value was 0.01202 when

345 all the breeds/populations were considered while it was 0.08001 when the zebu cattle alone

346 were considered. The MDS plots showed a pattern of clustering similar to that of phylogeny.

347 The taurine and crossbreds clustered separately, while Kangayam, Vechur and Punganur cattle

348 were distinct from rest of the zebu cattle breeds.

349 Bayesian clustering of individuals without prior population information was performed to

350 assess the cryptic genetic structure in the studied cattle breeds. Among K=2 to 15, the dataset

351 was best described with K=2 genetic clusters at which the second order rate of change in

352 LnP(D) was maximum (Supplementary Figure SF2). At K=2, the clustering of individuals

353 followed the zebu-taurine divide (Figure 3). When K=3 was assumed, Zebu cattle breeds got

354 clustered based on Hallikar or Kangayam ancestry. Both Hallikar and Kangayam cattle were

355 assigned to their respective clusters with >90% proportion of membership coefficient. Deoni

356 (80%) and Ongole (69.2%) were predominantly assigned to Kangayam ancestral cluster while

357 Vechur (86.4%), Punganur (78.7%), Bargur (75.7%) and Alambadi (72.8%) cattle were

358 predominantly assigned to Hallikar cluster. Pulikulam and Umblachery cattle showed

359 admixture from both the ancestry. When K=4 was assumed, Deoni and Ongole clustered

360 together independent of Kangayam ancestry while at K=5, Ongole clustered distinctly from

361 that of Deoni. At K=6, Deoni and Bargur formed a separate cluster each, while Punganur and

362 Vechur clustered together. At K=7, the miniature breeds, Pungaur and Vechur cattle clustered

363 distinctly. Significant genetic admixture of zebu ancestry was observed among Alambadi,

364 Pulikulam and Umblachery cattle consistently under the assumption of K=5 to 8.

365 To further understand the degree of phylogeographic structure in the studied cattle breeds,

366 analysis of molecular variance (AMOVA) was performed to assess the distribution of




15

367 microsatellite variation as a function of both breed membership and geographic origin (Table

368 3). When no grouping was assumed, 94.56% of the total variance among zebu cattle was found

369 within breeds while 5.44% of total variance was explained by between breed differences. When

370 the breeds were grouped according to their province of origin (Grouping I – based on

371 state/province), about 0.32% of total variance was due to differences among groups (P>0.05)

372 while 5.21% was due to differences among populations within groups. When the grouping was

373 done based on the classification of zebu cattle by Joshi and Phillips (1953) (Grouping II),

374 among group variance was negligible (0.28%; P>0.05), while most of the between population

375 variance was within groups (5.26%; P0.05) while variance among populations within groups was 4.41%. When the grouping was

378 based on Bayesian genetic structure (Grouping IV; Group1-DEO; Group2-ONG; Group3-

379 BAR, ALA, HAL, PUL, UMB; Group4-VEC; Group5-PNR; Group6-KAN), among group

380 variance was 2.78% and significant (P

16

392 informative sites. The overall nucleotide diversity and average number of nucleotide

393 differences among South Indian cattle was 0.0111and 10.36 respectively. A total of 207 unique

394 haplotypes were observed with overall haplotype diversity of 0.961 (Table 4). The haplotype

395 diversity observed in most South Indian cattle breeds was relatively higher than that of North

396 Indian zebu cattle located along the Gangetic plains (Sharma et al. 2015). Two major maternal

397 haplogroups, I1 and I2, typical of zebu cattle were observed, with the former being predominant

398 occurring 313 times and the later occurring 166 times. However, there were differences in

399 certain breeds like Vechur and Umblachery in which I2 haplogroup was more frequent than I1.

400 Interestingly, six zebu cattle showed taurine lineage, of which one belonged to T1 (Deoni),

401 three belonged to T2 (one each from Hallikar, Bargur and Kangayam) and two belonged to T3

402 (one each from Deoni and Pulikulam). The overall nucleotide diversity within I1 and I2

403 haplogroups were 0.0024 and 0.0029 respectively while the overall average number of

404 nucleotide differences were 2.32 and 2.77 respectively (Table 5). Among the 207 unique

405 haplotypes, 105 belonged to I1 haplogroup, 74 belonged to I2 haplogroup, 3 belonged to T3

406 haplogroup and one each belonged to T1 and T2 halpogroups (Table 4). The overall haplotype

407 diversity within I1 haplogroup and I2 haplogroup was 0.907 and 0.949 respectively. The

408 haplotype diversity within I2 haplogroup was higher than that of I1 in all the South Indian zebu

409 cattle breeds except Alambadi. Among the observed haplotypes, 27 were found to be shared

410 across 2 or more breeds, of which 18 belonged to I1 haplogroup, 8 belonged to I2 haplogroup

411 and one belonged to T3 haplogroup (Supplementary Table ST4). The haplotype

412 HxJxABDLKOPUVHJ_H13 was the most frequent among I1 haplogroup and shared with all

413 the investigated populations except Punganur cattle. Among the I2 haplogroup,

414 HxJxABDLKOPUV_H7 was the most frequent haplotype and shared among 11 populations

415 with the exception of Punganur, Holstein-Friesian and Jersey cattle. A total of 139 singleton

416 haplotypes were observed, of which 67, 50, 18, 3 and one belonging to I1, I2, T3, T2 and T1




17

417 haplogroups respectively. The higher proportion of singleton haplotypes within I2 haplogroup

418 of South Indian cattle was consistent with relatively higher haplotype and nucleotide diversity

419 observed within this lineage.

420 3.5. mtDNA phylogeny and evolutionary relationship

421 Phylogenetic analysis was performed on all the 207 unique haplotypes along with reference

422 sequences of each haplogroup, the results of which is presented in Supplementary Figure SF3.

423 The phylogenetic tree conformed to haplogroup classification with the formation of two major

424 indicine clusters I1 and I2. There were no major sub-clusters except for one minor clade

425 containing four or more haplotypes in each of the I1 and I2 haplogroups. The minor clade

426 within in I1 haplogroup was formed by four singleton haplotypes (K_H142, L_H120, L_H115

427 and V_H181) from Kangayam, Hallikar and Vechur cattle breed. Similarly, the minor clade

428 within I2 haplogroup was formed by three singleton and two shared haplotypes (P_H155,

429 AD_H61, B_H84, L_H129 and OP_H149) from Pulikulam, Alambadi, Deoni, Bargur, Hallikar

430 and Ongole cattle.

431 The frequency of mtDNA control region haplotypes was further utilized to estimate pairwise

432 genetic differentiation among zebu cattle from South India as well as the zebu inhabiting areas

433 closer to Indus valley sites (West Indian zebu) and Gangetic plains (North Indian zebu and

434 Nepalese zebu). Analysis of I1 haplotypes revealed pairwise FST varying between zero and

435 o.293 while I2 haplotypes showed pairwise FST varying between zero and 0.255. The pairwise

436 FST among I1 haplotypes was zero among NIC-NPC, WIC-NPC, ALA-UMB and HAL-UMB

437 breed combinations and less than 1% among ALA-HAL, WIC-HAL, WIC-UMB, NIC-UMB,

438 ALA-WIC breed combinations. The largest FST among I1 haplotypes was observed between

439 Punganur cattle and other zebu breeds (Vechur, Ongole, Deoni, Bargur and Nepalese) followed

440 by Vechur and other breeds (Pulikulam, Ongole and Umbalchery). In contrast, the pairwise FST

441 among I2 haplotypes was zero between Punganur and other zebu breeds (ONG, BAR, NIC,




18

442 UMB, DEO, HAL, WIC, ALA, KAN) and among HAL-DEO, WIC-DEO and PUL-DEO breed

443 combinations. The largest FST among I2 haplotypes was observed between Nepalese zebu and

444 South Indian zebu (ONG, KAN, PUL, BAR, UMB, VEC, DEO and ALA) followed by Vechur

445 and other breeds (ALA, BAR, UMB, ONG, PUL, HAL and DEO).

446 The pairwise FST for each of the haplogroups were further utilized to perform principal

447 components analysis (PCA) and understand the evolutionary relationships among the zebu

448 cattle breeds (Figure 4). Principal components analysis (PCA) performs orthogonal

449 transformation of a set of correlated variables (genetic distance estimations) into a set of new

450 linearly uncorrelated variables called as principal components. The three largest principal

451 components that explained maximum variation in the dataset were used to project sampled

452 breeds into a three-dimensional geometric space. PCA of I1 haplotypes revealed the three

453 largest principal components explaining 93.04% of the total variance in the dataset. The three-

454 dimensional scattergram drawn using the three largest principal components showed the

455 distinctness of Punganur, Pulikulam and Vechur cattle from all the other South Indian zebu

456 cattle breeds. The zebu cattle from North India and Nepal were also found to be distinct from

457 West and South Indian zebu cattle. Similarly, PCA of I2 haplotypes revealed the three largest

458 principal components together contributing to 91.31% of total variance in the dataset. The

459 three-dimensional scattergram showed Vechur and Nepalese cattle distinct from other zebu

460 cattle breeds.

461 To understand the underlying genetic structure of each of the two maternal lineages (I1 and

462 I2), analysis of molecular variance among zebu mtDNA haplotypes of cattle from South India,

463 Gangetic plains (North India and Nepalese) and Indus valley (West India) sites were conducted

464 (Table 6). When no grouping was assumed among I1 haplotypes, 6.17% of variance (P

19

467 India, Gangetic Plains and Indus valley sites), among group variance was negligible (0.43%)

468 and insignificant (P>0.05). However, when Punganur, Vechur and Pulikulam breeds were

469 grouped separately from other South Indian zebu breeds, among group variance was 2.91%

470 and statistically significant (P0.05). However, when Vechur and Nepalese cattle were grouped

477 together, among group variance increased to 6.04% and was statistically significant (P

20

492 most of the South Indian breeds with the modal value at two pairwise mismatches. Bimodal

493 distribution was observed among I1 haplotypes in Kangayam, Umbalchery and Punganur

494 breeds with modal values at 1 and 3, 1and 5, 2 and 6 mismatches respectively. In case of zebu

495 cattle from Indus valley site (WIC), the modal value was observed at 3 mismatches while the

496 Gangetic plains Zebu cattle, NIC and NPC showed modal values at 3 and 2 mismatches

497 respectively. The modal values of South Indian I1 lineage observed in the present study is

498 contrary to the reports made earlier (Chen et al. 2010). Similarly, with respect to I2 lineage,

499 modal values were observed around two to four mismatches in most of South Indian and

500 Gangetic Plain Zebu cattle (NIC and NPC) while it was observed at one mismatch in Indus

501 valley site zebu cattle (WIC). Unimodal distribution of I2 haplotypes was observed in most

502 South Indian zebu breeds except Deoni, Kangayam and Pulikulam. The mismatch analysis of

503 pairwise differences revealed the parameter ‘tau’ (the coalescent time of expansion) ranging

504 from 1.268 (Ongole) to 8.406 (Punganur) among I1 haplotypes while it ranged from 1.344

505 (Pulikulam) to 4.182 (Kangayam) among I2 haplotypes of South Indian zebu cattle. The

506 estimated sum of squared deviations (SSD) and the raggedness index was significant (P

21

517 of smaller magnitude was observed within the I1 haplogroup indicating a possible sub-lineage

518 of I1. The South Indian zebu cattle were further subjected to tests for selective neutrality, the

519 Tajima’s D and Fu’s FS. The Tajima’s D statistics was based on frequency spectrum of

520 mutations while Fu’s FS was based on haplotype distribution. The results revealed negative

521 Tajima’s D in all the investigated South Indian breeds for haplogroup I1, of which Deoni,

522 Hallikar and Pulikulam were statistically significant (P

22

542 Alambadi, Ongole, Pulikulam) and medium to large (Deoni, Kangayam) size animals; draught

543 characteristics varying from heavy load pulling capability (Hallikar, Umblachery) to speed and

544 endurance in trotting (Kangayam, Bargur); utilities ranging from draught to penning for dung

545 (Pulikulam, Bargur). Joshi and Phillips (1953) classified Zebu cattle of India and Pakistan into

546 six major groups based on their morphology, breed history and phenotypic characteristics. The

547 South Indian zebu cattle investigated in the present study were classified under three major

548 categories: Group II, Short horned, white or light grey cattle with coffin shaped skulls

549 (Ongole); Group III, Ponderously, built Red and White cattle (Deoni); and Group IV, Mysore

550 type, compact sized cattle (Hallikar, Kangayam, Umblachery, Bargur, Pulilkulam and

551 Alambadi). The bantam weight dwarf type cattle breeds, Vechur and Punganur did not fall into

552 any of these three categories.

553 4.1. Population status, effective population size and inbreeding estimates

554 With rapid mechanization of agriculture in South India, the need for rearing draught cattle came

555 down drastically in the last few decades. This led to reduction in effective population size and

556 consequent increase in rate of inbreeding in many of these breeds (Singh and Sharma, 2017).

557 A significantly strong negative Pearson’s correlation coefficient of -0.674 (P

23

567 had higher mean FIS estimates of 0.089 and 0.076 respectively. Although the total population

568 of both these breeds were relatively higher (39050 and 14154 heads of Umblachery and Bargur

569 respectively), only few hundred breedable males (586 and 624 for Umblachery and Bargur

570 respectively) were present in their respective native breed tracts. Lack of availability of

571 sufficient breeding bulls was one of the main reasons for low effective population size, higher

572 estimates of FIS and high rate of inbreeding in these breeds (Singh and Sharma, 2017). Among

573 the South Indian zebu cattle, the lowest FIS estimate was observed in Hallikar cattle

574 (FIS=0.021), whose total population was greater than 1.2 million heads with breedable male

575 and female population of 5524 and 397357 respectively. Interestingly, Punganur cattle with a

576 total population of less than 3000 and effective population size of ~200 showed positive, but

577 relatively lower estimate of FIS (0.041). This could be possibly due to a better breedable male

578 to breedable female ratio observed in this breed (0.0715) as compared to Vechur (0.0020),

579 Pulikulam (0.0084) and Umbalchery (0.0387) cattle (DADF, 2015).

580 Singh and Sharma (2017) assessed the degree of endangerment of Indian zebu cattle breeds

581 based on estimated breed-wise cattle population data (DADF, 2015), effective population size

582 and estimated rate of inbreeding. Using the criteria laid down by Food and Agriculture

583 Organization of the United Nations (FAO, 2013) and the Indian Council of Agricultural

584 Research-National Bureau of Animal Genetic Resources (ICAR-NBAGR, 2016), they

585 classified the status of Vechur as “critical” (breeding males less than five, breeding females

586 494 and high rate of inbreeding), Punganur and Pulikulam as “endangered” (breeding females

587 more than 300 but less than 3000) and Bargur as “vulnerable” (breeding females more than

588 3000 but less than 6000). Thus, the status of four out of ten South Indian zebu cattle breeds

589 investigated in the present study are in the category of “under risk” from conservation

590 standpoint. The level of inbreeding estimates is consistent with their population status and is a

591 cause for concern from conservation standpoint. It is also worth to mention that the population




24

592 status of “Alambadi” breed is currently not known but has been observed to retain moderate

593 levels of heterozygosity.

594 4.2. Origin, breed history and population structure

595 Most South Indian cattle breeds belong to Mysore type cattle (Group IV), of which Hallikar is

596 considered to be the ancient one and widely distributed with significant and stable population

597 size. Ware (1942) and Joshi and Phillips (1953) mentioned the origin of most Mysore type

598 cattle breeds centred around Hallikar cattle. The results of AMOVA and genetic structure

599 analysis obtained in the present study was in conformity with this theory of origin for most, if

600 not all the breeds. Among group variance was significant (P

25

617 colour, typical of being red and white or red with white spots. These cattle are known for their

618 speed and endurance in trotting, but are very fiery, restive and difficult to train (Ware, 1942;

619 Ganapathi et al. 2009; Pundir et al 2009). It also needs to be noted that among these four Mysore

620 type Tamil Nadu breeds (Alambadi, Bargur Umblachery and Pulikulam), significant selection

621 process had gone into the development of Bargur cattle, particularly for their unique

622 morphological characteristics, herd uniformity and adaptation to zero input, hilly environment.

623 This was evidenced by their individual assignment to unique inferred cluster when K=7 to K=8

624 (Figure 3) was assumed under Bayesian clustering analysis.

625 Pulikulam, is a migratory cattle breed, maintained in a zero-input system in southern parts of

626 Tamil Nadu state of South India. These cattle have predominantly white or greyish coat colour

627 and are mostly used for agricultural operations and rural transport. These cattle are well-known

628 for their utilities as a sport animal for Jallikattu (a traditional bull embracing sport) and penning

629 (a practice of manuring by keeping the cattle overnight in open agricultural fields). These cattle

630 resemble Kangayam in physical appearance but are significantly smaller and compact in size.

631 Gunn (1909) opined that these cattle probably had a strain of Mysore cattle blood in them. The

632 present study revealed Pulikulam as a genetically distinct breed from Kangayam but sharing

633 significant ancestry with Hallikar and Deoni cattle (Figure 3; K=4 to K=7). The genetic

634 difference between Pulikulam and Kangayam was also consistent with the reported differences

635 in morphometric parameters that included body length, height at withers and chest girth

636 (Kanakraj and Kathiresan, 2006; Singh et al. 2012).

637 Umblachery is an excellent draught cattle breed reared in the eastern coastal areas of Tamil

638 Nadu state (Rajendran et al. 2008). These are medium sized cattle, selectively bred for short

639 stature and suitability to work in marshy paddy fields. Anecdotal evidence attributed the

640 possible origin and development of Umblachery from Kangayam cattle. Gunn (1909)

641 mentioned that with the exception of appearance of head of these animals due to dehorning,




26

642 these cattle retained the chief characteristics of Kangayam. However, the present study showed

643 significant genetic differentiation among these breeds (FST=0.070) and little or no traces of

644 Kangayam ancestry in Umblachery cattle. Similar to Pulikulam cattle, significant admixture

645 was observed in Umblachery cattle with traces of Hallikar and Deoni ancestry in them. This

646 could be possibly due to trade influx of cattle into native tract of Umbalchery more than a

647 century ago, from Southern Tamil Nadu (Cumbum district) where the now extinct breed called

648 “Kappiliyan” or “Tambiran Madu” of Canarese (Mysore) origin was in abundance (Gunn,

649 1909; Littlewood, 1936).

650 The genetic structure analysis clearly indicated the influence of Hallikar breed in the

651 development of most Mysore type cattle breeds of Tamil Nadu (Alambadi, Bargur,

652 Umblachery and Pulikulam) with the exception of Kangayam. Kangayam is one of the popular

653 draught breeds of South India known for their power (0.8hp per pair of bullocks) and sturdiness

654 (Surendrakumar, 1988). Although, there are reports indicating the existence of Kangayam

655 cattle for nearly three centuries (Reddi, 1957), the origin and development of this breed has

656 been attributed mainly to 33rd Pattagar (community chief) of Palayakottai who selectively bred

657 the local cattle, only with sires available in his herd. The appearance of some of his cows/heifers

658 gave the impression that they had a distinct strain of Ongole blood in them, as the Pattagar was

659 known to possess purebred Ongole cows (Gunn, 1909). But no records were maintained by

660 Pattagar during the early stages of the development of this breed to confirm this. Kangayam

661 cattle conform largely to Mysore type breeds, but with a relatively larger body size, probably

662 indicating the admixture with a heavier breed (Phillips, 1944). However, the results of

663 phylogeny, multidimensional scaling and analysis of molecular variance in the present study

664 clearly showed the genetic uniqueness of Kangayam cattle. The pairwise FST indicated marked

665 differentiation of Kangayam from Hallikar (FST=0.093) and Ongole (FST=0.075) cattle. Further,

666 Kangayam was also significantly differentiated from other Mysore type breeds (Alambadi,




27

667 Bargur, Umblachery and Pulikulam) with pairwise FST ranging from 6.3% to 7.7%. The

668 Bayesian clustering analysis showed Kangayam as a genetically distinct entity from both

669 Ongole and Hallikar cattle. Low levels of admixture observed in Kangayam cattle, when

670 assumed under K=4 to K=8 (Figure 3), was probably due to ancestral zebu inheritance common

671 to all South Indian breeds. The genetic uniqueness of Kangayam cattle might have been the

672 outcome of systematic selective breeding practices of Pattagars initially, and subsequent

673 introduction of herd book registration followed by implementation of various genetic

674 improvement programs since1940s by Government of Madras (Panneerselvam and

675 Kandasamy, 2008).

676 The two bantam cattle breeds, Vechur and Punganur are unique to South India and are probably

677 among the smallest cattle breeds of the world, with the average height of adults ranging from

678 60-100 cm and body weight of 125-200 kg (Nath, 1993; Iype, 1996). Vechur and Punganur

679 cattle have retained moderate levels of diversity despite the small population size and being

680 classified in the critically endangered category. Both these breeds were distinct from rest of the

681 South Indian breeds, not only in terms of morphology but also in terms of milk production

682 efficiency. The cows are considered to be efficient milk producers relative to their size, with

683 peak milk yield scaling up to 3-4 kg/day (Nath, 1993; Iype, 1996). Genetic distance based

684 phylogenetic analysis and multidimensional scaling plot derived from pairwise FST showed

685 Vechur and Punganur to be genetically distinct from other South Indian draught type zebu

686 cattle. Although, Bayesian clustering analysis indicated the influence of Hallikar ancestry in

687 both the breeds (K=3 to K=6), they clustered independently at K=7 and K=8 (Figure 3). It is

688 also worth to mention that a high degree of genetic differentiation was observed among these

689 two dwarf cattle (Vechur and Punganur) breeds (FST=0.092). This is understandable

690 considering the geographical location of their respective native tracts. The isolation of native

691 tract of Vechur cattle (Kottayam district of Kerala state) by rivers, canals and backwaters and




28

692 the distant location of Punganur cattle habitat (Chittoor district of Andhra Pradesh state)

693 coupled with selective breeding practiced by farmers, might have contributed to the genetic

694 differentiation of the two breeds. In addition to restricted gene flow, genetic drift arising out of

695 small population size of both the breeds could have also contributed to their genetic divergence.

696 4.3. Breed purity and level of taurine introgression

697 With the mechanization of agricultural operations and increasing demand for dairy products in

698 the last few decades, the focus of cattle breeding in South India shifted towards improving milk

699 production. Beginning with the Operation Flood program in 1970s, intensive cattle

700 development projects in Southern India were targeted towards genetic improvement of local

701 cattle for milk through crossbreeding with exotic commercial taurine cattle. Although the

702 official breeding policy emphasized the need for selective breeding of recognized zebu cattle

703 breeds, indiscriminate crossbreeding took place in the native tracts of many indigenous breeds.

704 In the absence of pedigree recording under small holder production set up, it is not uncommon

705 that the local cows inseminated with purebred taurine or crossbred semen were subsequently

706 served by locally available sires or vice versa. Often, the choice of semen used for artificial

707 insemination (AI) is left to the inseminators and depends on the availability of purebred zebu,

708 purebred taurine or crossbred semen straws, irrespective of the breed/genotype of the cows

709 (purebred zebu, non-descript local, crossbred cattle, etc.) brought for AI. This resulted in

710 dilution of purebred zebu germplasm with varying levels of taurine admixture in well described

711 zebu cattle populations, thus affecting their breed purity.

712 Bayesian clustering analysis was performed to assess the taurine admixture in South Indian

713 zebu cattle using purebred Jersey and Holstein-Friesian as reference genotypes. Samples from

714 Bos taurus X Bos indicus crossbreds were also genotyped to evaluate the range of taurine

715 inheritance under field conditions. The proportion of membership coefficient obtained for

716 individual animals in the inferred zebu and taurine clusters were utilized to estimate taurine




29

717 admixture and purity of the investigated zebu cattle breeds (Table 8). Ongole had zero

718 individuals with >6.25% taurine admixture while Punganur had the highest proportion (44.4%)

719 of individuals with >6.25% taurine admixture. Among the South Indian cattle, Ongole,

720 Kangayam, Deoni, Alambadi and Bargur breeds showed majority of their individuals (>95%

721 of each of these populations) with less than 6.25% taurine admixture. The relatively low level

722 of taurine admixture in Ongole, Kangayam and Deoni can be attributed to the following factors:

723 (i) better artificial insemination coverage in their respective breed tracts (ii) availability of

724 purebred zebu, true to the type breeding bulls (iii) availability of purebred zebu semen for AI

725 and (iv) better implementation of breeding policy in favour of selective breeding of purebred

726 zebu cattle (Panneerselvam and Kandasamy, 2008; Natarajan et al. 2012; Dongre et al. 2017).

727 In case of Alambadi and Bargur cattle raised in a unique, extensive production system of hilly

728 areas, the access to artificial insemination services is little or absent. The cows in heat are

729 normally served by the bulls that accompany the herd while grazing in the forest areas. For

730 example, in Bargur cattle, the herd composition includes at least 2% of total size as breeding

731 bulls and young male stocks (Ganapathi et al. 2009) in addition to 18-20% bullocks. The

732 farmers select the breeding bulls based on certain phenotypic characteristics including

733 uniformity of coat colour (Pundir et al 2009). The absence of AI facility and natural service

734 being the main breeding practice in the native tract, the introduction of exotic bull semen was

735 limited in Bargur and Alambadi cattle resulting in low levels of taurine admixture in these

736 breeds. With respect to Hallikar cattle, the proportion of individuals with >6.25% taurine

737 admixture was estimated to be 16.7%. This was surprisingly higher, considering the significant

738 and stable population size of this breed (DADF, 2015). Singh et al. (2008) reported the demand

739 for purebred Hallikar semen was higher than the available supply in the native tract and thus

740 the lack of quality breeding bulls might have contributed to increased levels of taurine

741 admixture in this breed.




30

742 Breeds with highest proportion of individuals with >6.25% taurine admixture included

743 Punganur (44.4%), Vechur (42.3%), Umblachery (33.3%) and Pulikulam (23.5%) cattle.

744 Significant proportion of individuals belonging to these breeds also had >12.5% taurine

745 admixture. Among these, Vechur and Pulikulam have been reported to have severe shortage of

746 breedable males (DADF, 2015; Singh and Sharma, 2017) in their respective native tracts.

747 However, the AI coverage in these areas is high and the semen of purebred, commercial taurine

748 cattle are readily available for insemination in local cows. This might have resulted in

749 indiscriminate crossbreeding of purebred zebu cattle, thereby increasing the level of taurine

750 admixture in them. In case of Punganur cattle, although reasonable number of breedable males

751 are available in the native tract, the level of taurine admixture is relatively higher, possibly due

752 to well established AI programs and readily available exotic taurine semen for insemination.

753 Kannadhasan et al. (2018) utilized community based participatory approach to identify and

754 prioritize the constraints in Umblachery cattle farming: (i) shift in farmers’ preference towards

755 crossbreds for better milk production (ii) lack of interest among farmers in using Umblachery

756 bullocks for agricultural operations (iii) lack of grazing lands (iv) lack of quality purebred

757 breeding bulls and (iv) inadequate efforts of breeders’ association in promoting and improving

758 Umblachery breed. The above factors coupled with high coverage of AI and easy access to

759 taurine/crossbred semen might have contributed to relatively higher levels of taurine admixture

760 in Umblachery cattle.

761 4.4. Maternal lineage diversity and zebu cattle domestication

762 The present study revealed no additional maternal haplogroups in the South Indian cattle other

763 than the typical I1 and I2 zebu lineages reported earlier (Chen et al. 2010). A relatively high

764 average number of nucleotide differences, nucleotide and haplotype diversity was observed in

765 I2 lineage as compared to that of I1lineage. Mismatch distribution analysis of mtDNA control

766 region sequences revealed the modal value at two mismatches for I1 lineage, while it was




31

767 observed at two to four mismatches for I2 lineage in most of the South Indian breeds. The

768 mismatch distribution analysis thus clearly showed high levels of mitochondrial control region

769 diversity in South Indian zebu cattle. In general, extant populations located nearby centres of

770 domestication retain high levels of diversity, while populations located farther away from the

771 domestication centre show reduced diversity. The high levels of mtDNA control region

772 diversity observed in South Indian zebu cattle is interesting. Particularly, the pairwise

773 differences among the I2 haplotypes of South Indian cattle were relatively higher than West

774 Indian (Indus valley site) zebu cattle and comparable to that of North Indian and Nepalese zebu

775 cattle (Gangetic plains zebu) (Chen et al. 2010).

776 Several studies have demonstrated independent domestication and origin of Bos tauus and Bos

777 indicus cattle based on mtDNA variations (Loftus et al. 1994). It is also clearly established that

778 the humpless Bos taurus cattle were domesticated in the Near East during the Neolithic

779 transition about 10 thousand years before present (Troy et al. 2001; Beja-Pereira et al. 2006;

780 Edwards et al. 2007; Kantanen et al. 2009; Bonfiglio et al. 2010). Similarly, the report of Chen

781 et al. (2010) shed much light on the origin and domestication of humped zebu cattle in the

782 Indian sub-continent. They analyzed mitochondrial control region sequences from six different

783 regions of Asia (i) Indus valley region (ii) Gangetic plains (iii) South India (iv) North East

784 India (v) West Asia (vi) East Asia (vii) Central Asia and (viii) South East Asia. Among these,

785 nucleotide diversity was reported to be notably higher in four regions of the Indian sub-

786 continent (Indus Valley, Gangetic Plains, South India and North East India), indicating a vast

787 expanse of area as the probable centre of domestication. Chen et al. (2010) identified the greater

788 Indus valley (West India and present-day Pakistan) as most likely location for the origin and

789 domestication of I1 maternal lineage. They made the above conclusion based on the following

790 evidences: (i) the backbone structure of I1 haplogroup network being formed by cattle

791 populations in the Indus valley region (ii) modal value of mismatch distribution of pairwise




32

792 differences being higher in Indus valley cattle (two mismatches) as compared to those located

793 in Gangetic plains (one mismatch) and South India (one mismatch) and (iii) relatively high I1

794 haplotype diversity observed in Indus valley cattle as compared to cattle in Gangetic plains and

795 South India. However, analysis of I2 haplogroup sequences showed more complex pattern of

796 diversity among cattle located in these regions. Identical modal values of I2 mismatch and

797 nearly identical haplotype diversity were observed in Indus valley and South Indian cattle while

798 the frequency of I2a (an intermediate I1-I2 haplotype) was relatively higher in cattle from

799 Gangetic plains. Hence, Chen et al. (2010) reported the lack of genetic evidence to pin-point

800 a single location/region for the origin of the I2 haplogroup. However, they opined the possible

801 initial domestication of this lineage in the northern part of the Indian sub-continent based on

802 the chronology of available archaeological evidence for domesticated zebu cattle and probable

803 wild aurochs in the Indian subcontinent.

804 The present study based on comprehensive sampling of diverse breeds of South Indian zebu

805 cattle, showed the mismatch modal value of I1 haplogroup to be higher (two mismatch) than

806 reported by Chen et al. (2010). More interestingly, the mismatch distribution analysis of I2

807 haplogroup sequences revealed modal values around two to four mismatches in most South

808 Indian breeds. This was higher than the modal value observed for West Indian (Indus valley)

809 cattle, but comparable to that of North Indian and Nepalese (Gangetic plains) cattle. The

810 intermediate I1-I2 haplotype sequences (I2a) were also observed in certain South Indian cattle

811 breeds like Pulikulam, Deoni and Kangayam. Further, it is worth to mention that the haplotype

812 diversity within South Indian I2 lineage was higher in most South Indian cattle breeds.

813 Principal components analysis of pairwise FST derived from frequency of I2 haplotypes showed

814 the distinctness of Vechur, Kangayam, Punganur and Alambadi cattle, thus indicating the

815 diversity of this lineage within South Indian cattle. The results showed the need for additional

816 sampling of zebu breeds from North, East and North-East Indian cattle to explore and gather




33

817 new evidence on the potential location for the origin and domestication of I2 lineage in the

818 Indian sub-continent.

819

820 Conclusions

821 The present study revealed moderate levels of genetic diversity existing in South Indian zebu

822 cattle. It also established the genetic distinctness of Kangayam, Vechur and Punganur cattle

823 from the rest of the South Indian zebu cattle. The influence of Hallikar breed was observed in

824 the development of most Mysore type cattle breeds of South India with the exception of

825 Kangayam. Analysis of mtDNA control region sequences revealed two major maternal

826 haplogroups, I1 and I2, with the former being predominant than the later. The results indicated

827 the need for additional sampling and comprehensive analysis of mtDNA control region

828 variations to unravel the probable location of origin and domestication of I2 zebu lineage. The

829 present study revealed two major concerns about the status, conservation and improvement of

830 South Indian draught type zebu cattle breeds: (i) Four of the ten investigated zebu cattle breeds

831 are under the risk of endangerment due to small effective population size and high rate of

832 inbreeding (ii) the lack of sufficient purebred zebu bulls for breeding and increasing level of

833 taurine admixture in the native breeds tracts. Recently, the state governments of South India

834 have started establishing conservation units and research centres for improvement of breeds

835 like Alambadi, Bargur, Kangayam and Umblachery. Parental stocks of these breeds are being

836 established by procuring true to the type animals from small holder farmers. However, in the

837 absence of pedigree records, it has been difficult to estimate genetic purity or taurine admixture

838 in those animals. Considering the relatively higher levels of taurine admixture in some of the

839 breeds, it is advisable to perform molecular/genomic evaluation of local cattle to improve breed

840 purity in these units/centres. It also needs to be mentioned that additional efforts are necessary

841 to make the farmers aware of scientific breeding practices and limit the levels of inbreeding




34

842 and taurine admixture in purebred zebu populations. Availability of purebred semen for

843 artificial insemination, incorporation of genomic/molecular information to identify purebred

844 animals and increased awareness among farmers will help to maintain breed purity, conserve

845 and improve these important draught cattle germplasms of South India

846

847 Acknowledgement

848 The present study is part of the Coordinated Research Project D3.10.28 of Joint FAO/IAEA

849 Division of Nuclear Techniques in Food and Agriculture, International Atomic Energy Agency

850 (IAEA), Vienna, Austria. The funding provided by IAEA for internship training of the first

851 author at Animal Production and Health Laboratory, Seibersdorf, Austria is gratefully

852 acknowledged. The authors also thank Dean, Veterinary College and Research Institute,

853 Namakkal, Tamil Nadu, India for providing necessary facilities for the study.

854

855 References

856 Antao T, Lopes A, Lopes RJ, Beja-Pereira A, Luikart G. LOSITAN: a workbench to detect

857 molecular adaptation based on a FST outlier method. BMC Bioinformatics. 2008; 9: 323.

858 Bandelt HJ, Forster P, Rohl A. Median-Joining networks for inferring intraspecific

859 phylogenies. Molecular Biology and Evolution. 1999; 16: 37-48.

860 Beja-Pereira A, Caramelli D, Lalueeza-Fox C, Vernesi C, Ferrand N, Casoli A, et al. The origin

861 of European cattle: evidence from modern and ancient DNA. Proceedings of National

862 Academy of Sciences, USA. 2006; 103: 8113-8118.

863 Bonfiglio S, Achilli A, Olivieri A, Negrini R, Colli L, Liott L., et al. The Enigmatic Origin of

864 Bovine mtDNA Haplogroup R: Sporadic Interbreeding or an Independent Event of Bos

865 primigenius Domestication in Italy? PLoS One. 2010; 5(12): e15760.

866 doi:10.1371/journal.pone.0015760.




35

867 Chen S, Lin B-Z, Baig M, Mtra B, Lopes RJ, Santos AM, et al. Zebu cattle are an exclusive

868 legacy of the South Asia Neolithic. Molecular Biology and Evolution. 2010; 27 (1): 1-

869 6.

870 DADF. Estimated livestock population breed-wise based on breed survey 2013. Published by

871 Department of Animal Husbandry, Dairying and Fisheries, Ministry of Agriculture,

872 Government of India, New Delhi; 2015.

873 de Oliveira JFC, Neves JP, Almeida EA, Steigleder CS, Moraes JCF, Gonçalves PBD, Weimer

874 TA. Association between reproductive traits and four microsatellites in Brangus-Ibagé

875 cattle. Genetics and Molecular Biology. 2005; 28 (1): 54-59.

876 Dieringer D, Schlötterer C. MICROSATELLITE ANALYSER (MSA): a platform indpendent

877 analysis tool for larger microsatellite data sets. Molecular Ecology Resources. 2003; 3

878 (1): 167-169.

879 Dongre VB, Gandhi RS, Salunke VM, Kokate LS, Durge SM, Khandait VN, Patil PV. Present

880 status and future prospects of Deoni Cattle. Indian Journal of Animal Sciences. 2017;

881 87 (7): 800–803.

882 Edwards CJ, Bollongino R, Scheu A, Chamberlain A, Tresset A, Vigne J-, et al. Mitochondrial

883 DNA analysis shows a near eastern Neolithic origin for domestic cattle and no

884 indication of domestication of European aurochs. Proceedings of the Royal Society B.

885 2007; 274:1377-1385.

886 Evanno G, Regnaut S, Goudet J. Detecting the number of clusters of individuals using the

887 software structure: a simulation study. Molecular Ecology. 2005; 14: 2611-2620.

888 Excoffier L, Hofer T, Foll M. Detecting loci under selection in a hierarchically structured

889 population. Heredity. 2009; 103: 285–298.

890 Excoffier L, Laval G, Schneider S. Arlequin ver. 3.0: An integrated software package for

891 population genetics data analysis. Evolutionary Bioinformatics Online. 2005; 1: 47-50.




36

892 FAO. In vivo conservation of animal genetic resources. FAO Animal Production and Health

893 Guidelines. No. 14. Rome; 2013.

894 FAO. Molecular genetic characterization of animal genetic resources. FAO Animal Production

895 and Health Guidelines. No. 9. Rome; 2011.

896 FAO. The second report on the state of the World’s Animal Genetic Resources for Food and

897 Agriculture, edited by Scherf BD and Pilling D. FAO, Rome; 2015.

898 Felsenstein J. PHYLIP: Phylogeny inference package, version 3.5. Department of Genetics,

899 Washington University, Seattle, Washington; 1993.

900 Ganapathi P, Rajendran R, Subramanian A. Distribution and Population Status of Bargur cattle.

901 The Indian Veterinary Journal. 2009; 86: 971–972.

902 Grant WS. Problems and Cautions With Sequence Mismatch Analysis and Bayesian Skyline

903 Plots to Infer Historical Demography, Journal of Heredity. 2015; 106 (4): 333–346.

904 https://doi.org/10.1093/jhered/esv020.

905 Gunn WD. Cattle of Southern India Vol III. Bulletin No. 60. Department of Agriculture,

906 Madras, India; 1909.

907 Hiendleder S, Lewalski H, Janke A. Complete mitochondrial genomes of Bos taurus and Bos

908 indicus provide new insights into intraspecies variation, taxonomy and domestication.

909 Cytogenetic and Genome Research. 2008; 120:150–156. doi: 10.1159/000118756.

910 ICAR-NBAGR. Guidelines for Management of Animal Genetic Resources of India. National

911 Bureau of Animal Genetic Resources (Indian Council of Agricultural Research), Karnal

912 (Haryana), India, 163 p.; 2016.

913 Iype S. The Vechur Cattle of Kerala. Animal Genetic Resources Information. 1996; 18: 59-63.

914 Joshi NR, Phillips RW. Zebu cattle of India and Pakistan. FAO Agriculture Studies No. 19,

915 Food and Agricultural Organization of the United Nations, Rome, Italy. pp. 14-29;

916 1953.




37

917 Kanakraj P, Kathiresan D. Study on the physical and productive characteristics of Pulikulam

918 breed of cattle and Katchaikatty sheep. Final Report of the project submitted to SEVA,

919 Madurai; 2006.

920 Kannadhasan MS, Kathirchelvan M, Rajendran R. Identification and Prioritization of

921 Constraints in Umblachery Breed Cattle Farming Through Participatory Approach.

922 International Journal of Current Microbiology and Applied Sciences. 2018; 7(11):

923 1100-1110. https://doi.org/10.20546/ijcmas.2018.711.128

924 Kantanen J, Edwards CJ, Bradley G, Vinalass H, Thessler S, Ivanova Z, et al. Maternal and

925 paternal genealogy of Eurasian taurine cattle (Bos taurus). Heredity. 2009; 103: 404-

926 415.

927 Kienzle, Ashburner JE, Sims BG. Mechanization for Rural Development: A Review of Patterns

928 and Progress from around the World. In Integrated Crop Management (FAO) No. 20;

929 Food and Agriculture Organization of the United Nations: Rome, Italy; 2013.

930 Li M-H, Iso-Touru T, Laurén H, Kantanen J. A microsatellite-based analysis for the detection

931 of selection on BTA1 and BTA20 in northern Eurasian cattle (Bos taurus) populations.

932 Genetics Selection Evolution. 2010; 42:32. http://www.gsejournal.org/content/42/1/32.

933 Littlewood RW. Livestock of Southern India. Government of Madras, Madras, India. pp. 51-

934 69; 1936.

935 Loftus R, MacHugh D, Bradley D, Sharp P, Cunningham P. Evidence for two independent

936 domestication of cattle. Proceedings of National Academy of Sciences, USA. 1994; 91:

937 2757-2761.

938 Natarajan A, Chander M, Bharathy N. Relevance of draught cattle power and its future

939 prospects in India: A review. Agricultural Reviews, 2016; 37 (1): 49-54.

940 Natarajan A, Senthilvel K, Velusamy K. Productivity performance of Kangayam cattle. Indian

941 Journal of Animal Sciences. 2012; 82 (11): 1440–1441.

.CC-BY 4.0 International licenseperpetuity. It is made available under ap

legacy of draught cattle breeds of south india: insights into … · 2021. 1. 21. · 2 35 legacy...

Documents