A genetic index for stripe-pattern reduction in the zebra: the quagga project
Transcription
A genetic index for stripe-pattern reduction in the zebra: the quagga project
A genetic index for stripe-pattern reduction in the zebra: the quagga project Rochelle Parsons , Colleen Aldous-Mycock * & Michael R. Perrin 1 1 1 2 School of Biochemistry, Genetics, Microbiology and Plant Pathology, University of KwaZulu-Natal, Private Bag X01, Scottsville, 3209 South Africa 2 School of Biological and Conservation Sciences, University of KwaZulu-Natal, Scottsville, 3209 Received 15 February 2007. Accepted 21 May 2007 The quagga project aims to breed plains zebra that phenotypically resemble the extinct quagga (Equus quagga quagga), by selective breeding to aggregate desirable characteristics, particularly a reduced striping pattern. The purpose of this study was to produce a genetic selection index to improve stripe-pattern reduction, and hence to produce an efficient and objective selective breeding protocol, which will hopefully be of use in future selection experiments. From images of selectively bred zebras, striping ratios for three regions were calculated. Correlations of parent-offspring relationships resulted in narrow-sense heritabilities. Data from two regions (R1 and R3) were used to create an index to improve selection and reduce striping. The index I = 3.1875 (R1) + 4.8134 (R3) and the response to selection using the index, R = 0.6047 i. Key words: Equus quagga, genetic index, quagga, stripe pattern. INTRODUCTION The extinct quagga The quagga (Equus quagga quagga) was a zebra-like equid once common in South Africa but became extinct approximately 100 years ago due to over-hunting or planned extermination by colonists and competition with livestock. The quagga was morphologically divergent in coat colour from all zebras (Leonard et al. 2005), having brown zebra-like stripes only on the anterior half of its body, while the hind quarters were almost solid brown in colour (Fig. 1c). The core of the quagga’s geographical range was the semi-arid, temperate Karoo where it co-occurred with the plains zebra (E. burchelli) in a narrow belt of overlap north of the Orange river, and Hartmann’s mountain zebra (Equus zebra hartmannae) in neighbouring Namibia to the northwest (Hack et al. 2002). In order to resolve the controversy surrounding the taxonomic status of the quagga, immunological and molecular comparative analyses based on mitochondrial DNA from preserved skins were performed. They demonstrated that the quagga is not a distinct species but one of several subspecies of the plains zebra (Higuchi et al. *To whom correspondence should be addressed. E-mail: [email protected] 1984, 1987; Lowenstein & Ryder 1985). The quagga haplotypes investigated are closely related to one another with an average sequence divergence of 0.6% (Leonard et al. 2005). These data support a close affiliation between the quagga and the plains zebra, since the South African plains zebra differs from the quagga by 1.5% and from other plains zebras by 2.4% in a 395 bp mitochondrial region (Leonard et al. 2005). The results were derived from the mtDNA homologies and –8 substitution rates of 10 substitutions/site/year (Oakenfull et al. 2000). This suggests divergence occurred between the quagga and plains zebra in the Pleistocene, during the glacial maximum around 120 000–290 000 years before present (Hewitt 2000). There are several hypotheses that may explain the differences in coat colouration (Fig. 1) (Ruxton 2000). Each infers some fitness benefit of the striped pattern, resulting from differential selection and subsequent evolutionary change in favour of stripes. The hypotheses relate to predator avoidance, social fitness, thermoregulation and protection from tsetse flies. Cloudsley-Thompson (1984) proposed that differential striping patterns segregated the subspecies while Morris (1990) suggested the stripes serve as a mechanism for intraspecies recognition and herd cohesion (Morris 1990). It has been suggested striping affords South African Journal of Wildlife Research 37(2): 105–116 (October 2007) 106 South African Journal of Wildlife Research Vol. 37, No. 2, October 2007 a b c d Fig. 1. Striping variation between the zebra species and quagga. a, Mountain zebra, Equus zebra; b, Grevy’s zebra E. greveyi; c, Burchell’s zebra, E. burchelli; d, quagga, E. quagga (the London Zoo mare, the only quagga to have been photographed alive, by Frederick York (1870)). protection against attack by tsetse flies and disease transmission. Since the quagga occurred outside the range of the tsetse fly, it has been argued that striping had no selection value in terms of tsetse fly avoidance (Waage 1981). Bennet (1980) proposed that the distinctive coat colour of the quagga appeared relatively quickly (although she regarded the quagga’s closest relative as E. caballus and not E. burchelli ). Extant plains zebras show a geographical gradient in progressive stripe reduction from north to south. This may be correlated with habitat changes, with the most southerly populations being adapted to open country; the quagga represented the extreme limit of the trend (Rau 1978). Individuals (museum specimens and photographs) vary in the degree to which they show ‘classic’ quagga features, particularly the lack of stripes and the shade of brown colouration on the hind quarters. The proposed rapid evolution of coat colour and pattern could perhaps be explained by genetic drift, the disruption of gene flow by geographical isolation and/or an adaptive response to a drier habitat (Leonard et al. 2005). The Quagga Project The key aim of the Quagga Project is to breed a zebra phenotypically identical to the extinct quagga from populations of plains zebras. (The Quagga Project 2006). There is considerable intraspecific variability in the pelts of the preserved quagga museum specimens, and in populations of plains zebra. Some plains zebra exhibit quaggalike characteristics (traits) including shades of brown in their colouring, a reduction in striping and a minor flare in the tail. The project aims to select these traits into individuals which would be quagga-like. In 1987, nine zebras (from a population of 2500 zebras) were selected based on their quagga-like traits and captured at the Etosha National Park to start the breeding programme. Later, other breeding stocks were selected and captured from Etosha and Zululand. This original stock has been kept in eleven localities in the Western Cape. Currently, the quagga project comprises 150 individuals (The Quagga Project 2006), including some of the originally captured breeding stock as well as up to four generations of offspring. The Parsons et al.: Genetic index for stripe-pattern reduction in the zebra 107 Fig. 2. The most quagga-like plains zebra foal from the quagga project (Courtesy of the Quagga Project). success of the project is seen in its third and fourth generation foals in terms of stripe reduction (Fig. 2). However, the brown coat colour has not been as successfully bred into the population. The aim of this investigation was to produce a striping index to facilitate further stripe reduction through selective breeding. It necessitated developing a new method of quantifying striping because existing methods, of visual and manual stripe counting, are time-consuming and may not offer a high degree of reliability. METHODS The specific trait investigated was the ratio of striped to non-striped areas of each zebra. The data analysis was based on digital images of all the animals in the studbook. The studbook was completed at the beginning of 2006 by the late R.E. a b Rau of the Iziko South African Museum. Information on the parentage of each individual was included where known. For each animal, a left, right and hind view were included in the studbook. Data analysis For each individual, the view (left or right view) with the greatest resolution was chosen for the analysis of striping ratios. To ensure that the left and right view ratios did not differ significantly from each other, a random sample of 15 individuals were chosen (10%) from the population. Their left and right ratios compared as a control. Each image was digitally enhanced using software ArcSoft PhotoStudio™ to enhance the contrasts of the striping patterns. To compile the striping ratios, the ® software package AnalySIS Pro 3.2 (build 689) Soft Imaging System was used. c Fig. 3. The selected regions for data analysis; a, Region 1: shoulder to rear; b, Region 2: mid to rear; c, Region 3: rump. 108 South African Journal of Wildlife Research Vol. 37, No. 2, October 2007 In order to define the ratio of striped to non-striped areas a macro imaging procedure was created, and programmed to apply a repetitive process when a new image was added. The process included setting the region of interest; Region 1: the shoulder to rear; 2, the rear to the mid section; 3, the rump (Fig. 3). After setting the region of interest, area measurement was defined first for the non-striped areas and then for the total area. The non-striped areas were identified by adjusting contrast and colour intensity thresholds to exclude the maximum amount of brown and black stripes. The striped area was calculated as the non-striped area subtracted from the total area. For each individual the ratio of striped to total area was calculated. The 2 data unit for area measurement (pixel ),allowed for any image size to be used and ratios to be calculated. Statistical analysis and index calculations The phenotypic variance for each region was ® calculated using GenStat (v. 9.1). Comparative analysis was performed to check whether any of the data for the three regions were redundant. It showed that Regions 1 and 2 were redundant, leaving Regions 1 and 3 for constructing the index. The mean of each parent’s offspring was calculated (appendices A and B). The regression comparison of (a) dams to their offspring means and (b) sires to their offspring means was performed as a control to test for any significant difference between male and female parents. A regression of combined parents (P) to offspring (O) was used to deduce heritability estimates. Heritability can be defined statistically as the proportion of phenotypic variance attributable to genetic variance or more commonly as the extent to which genetic individual differences contribute to individual differences in phenotype (Falconer & Mackay 1996). The slope of the linear regression (bOP) estimates heritability as: bOP = ½h 2 . (1) The heritabilities for both regions were calculated. The phenotypic correlation (rP) and genetic correlation (rA) values were required for the index to be constructed. The phenotypic correlation is an estimate of the association between visible characteristics while the genetic correlation is the correlation between breeding values. Correlation estimates were calculated using Genstat (version 9.1). rp = cov p(13 ) σ p1σ p3 , (2) where covP(13) is the covariance between Regions 1 and 3 in parents and σP is the standard deviation for parents within areas 1 and 3, respectively. cov OP(13) , (3) rA = cov OP(1)cov OP(3) where covOP(13) is the covariance between parent Region 1 and offspring and offspring Region 3, covOP(1) is the covariance between parent and offspring in Region 1 and covOP(3) is the covariance between parent and offspring in Region 3. Covariance is the measure of how much two traits vary together. The breeding objective in this study was striping reduction. The breeding value is evaluated as a composite of all striping characters evaluated when trying to calculate the score (or index). It was calculated separately for each individual. In order to generate the index, the solutions of the b coefficients in the following equations were used as the coefficients for the trait improvement: b1P11 + b3P13 = a1A11 + a3A13 b1P31 + b3P33 = a1A31 + a3A33 , (4) where Pii is the phenotypic variance for each region and Pij is the covariance between regions. Aii is the additive variance of the regions and Aij is the additive covariance between regions (i,j = 1, 3): Aii = hi2σPi2 (5) Aij = r Ah i h j σ Pi σ Pj . (6) The economic weights may reflect preferences or simply arbitrarily fixed values. Ideally, an economic weight of a single trait should reflect the marginal benefit from a one unit improvement. Economic weights (a) were assigned as the inverse of the phenotypic standard deviation of 1 1 each region i.e. a1 = /σ1 and a3 = /σ3 (Falconer & Mackay 1996). By solving the above equation simultaneously for the b coefficients, the index was formed by 2 substitution. The variance of the index (σI ) was then calculated as: σI = b1 (a1A11 + a3A13) + b3 (a1A31 + a3A33) . 2 (7) The response to selection (R ) based on the index was then be predicted by: R = σ Ii , (8) where i is the intensity of selection (Falconer & Mackay 1996). Parsons et al.: Genetic index for stripe-pattern reduction in the zebra 109 Fig. 4. Regression of offspring on sire and mare values for shoulder to rear striping (Region 1). RESULTS For regression analysis for the calculation of index coefficients, data for 22 dams and 16 sires (with a minimum of two offspring) were used. Since there were no significant differences (P < 0.0001) between the left and right coat patterns, further analyses were done using the flank with the greater resolution. The regression controls for sires (Appendix A) and dams (Appendix B) provided regression coefficients of 0.2306 and 0.2151, respectively for Region 1 (Fig. 4). The effect of gender on coat patterning was insignificant and so regressions were calculated using the combined data set. For Region 1 the mean striping ratio was 0.411, with a minimum ratio of 0.93 and a maximum ratio of 0.630. For Region 3 the mean striping ratio was 0.396 with a minimum ratio of 0.056 and a maximum ratio of 0.668. The regression coefficient was 0.2171 for Region 1 and 0.2705 for region 3 (Fig. 5). The heritability for each region was calculated 2 2 from these values using equation 1; h1 and h3 are 0.4342 and 0.5410, respectively. The phenotypic variance of Region 1 (σ2P1 or P11) is 0.0098 and of Region 3 (σ2P3 or P33) is 0.0134. The phenotypic covariance of Regions 1 and 3 (covP(13) equivalent to P13 = P31) is 0.0064. The additive variance of Region 1 (A11) and Region 3 (A33), derived from Equation 5, are 0.0042 and 0.0072, respectively. The additive correlation (rA) was calculated as 0.3953 (Equation 3) and the additive covariance between the regions (A13 = A31) was 0.0021 (Equation 6). The economic weightings a1 and a3 were estimated at 10.1272 and 8.6457, respectively, and therefore b1 and b3 equate to 3.1875 and 4.8134, respectively. These calculations yield an overall index equation of: I = 3.1875 (R1) + 4.8134 (R3) . The variance of the index (σI2) is 0.6048 (Equation 7), and the response to selection equation is then given by R = 0.6047i. DISCUSSION The results demonstrate a strong correlation between parents and offspring with reference to reduction in stripe pattern. The absence of any significant difference between gender and parent– offspring correlation in coat pattern indicate that paternal and maternal effects are equivalent in this regard. Relatively large variances suggest there is scope for further stripe pattern reduction through continued selective breeding. The heritability estimates for stripe pattern reduction in both regions are particularly high, suggesting that selection 110 South African Journal of Wildlife Research Vol. 37, No. 2, October 2007 Fig. 5. Regression of offspring on single parent values for shoulder to rear striping (Region 1) and rump (Region 3). based on phenotypic variation would be successful, because the phenotype is a good indicator of the animals’ inherent breeding potential. Results show a definite trend towards stripe pattern reduction. The high heritabilities for striping ratio could be due to the relatively small number of animals it was possible to include in this study. Many examples of heritability of other quantitative traits by both natural and artificial selection for coat colour can be found in the literature (Andersson 2001). Mathematical models have also been produced which show that relatively large coat colour pattern changes can occur with only small changes in the model parameters (Murray 2003). The index is a prediction of breeding success in relation to reducing striping, and shows the rump to be the most important region when attempting to achieve the quagga-like phenotype. However, for successful classification of each animal, both characters should contribute to the index to provide a value upon which any further selection strategies are based. The variance of the index should be used to predict the response to selection when the intensity of selection is known. It can also be compared with the response through simple selection to estimate the index’s efficiency. The data presented above will aid selective breeding in the quagga project population. However, pleiotrophy and epistasis may also be in play and have not been considered in this study. For the purpose of producing a simple index it was assumed that the trait is quantitative with additive gene action and perhaps only dominance deviations were present. Important progress in the data analysis procedure was made in this study. The macro produced along ® with the effective use of the AnalySIS Imaging software allows the breeder to obtain a photograph of the animal from either left or right view. When the ‘cleaned’ image is entered it yields the relevant data within minutes. When the data are entered into the prepared Microsoft Excel spreadsheet, the individual’s index value is calculated automatically. This simple and effective method provides accurate striping ratios from the formula spreadsheet without the use of any manual, mathematical or statistical operations. In conclusion, the index will allow the Quagga Project to simplify its selective breeding protocol and to reduce the striping pattern in the study population. It could be appropriate and useful to use such data collection and indexing methods for quantifiable phenotypic traits in other mammals. To further explore stripe pattern reduction, it is necessary to study the interaction and number of genes involved in the process. The results of such Parsons et al.: Genetic index for stripe-pattern reduction in the zebra a study together with the index generated here will aid achieving the project’s objectives. ACKNOWLEDGEMENTS Micheal Knight and the rest of the team of the Quagga Breeding Project are thanked for the data and their support of this research. We thank the technical staff of the Centre for Electron Microscopy of University of KwaZulu-Natal for their help with image analysis. Carl Roux from the University of Pretoria is also thanked for statistical advice. REFERENCES ANDERSSON, L. 2001. Genetic dissection of phenotypic diversity in farm animals. Nat. Rev. Genet. 2: 130–138. BENNET, D.K. 1980. Stripes do not a zebra make, Part I: A cladistic analysis of equus. Syst. Zool. 29: 272–287. CLOUDSLEY-THOMPSON, J.L. 1984. How the zebra got his stripes – new solutions to an old problem. Biologist 31: 226–228. FALCONER, D.S. & MACKAY, T.F. 1996. Introduction to quantitative genetics (4th edn). Longman, New York. HACK, M.A., EAST, R., & RUBENSTEIN, D.I. 2002. Status and action plan for the plains zebra (Equus burchellii). In: P.D. Moehlman (ed.), Equids: zebras, asses and horses. Status survey and conservation action plan, IUCN/SSC Equid Specialist Group, IUCN, Gland, Switzerland. HEWITT, G. 2000. The genetic legacy of the quaternary ice ages. Nature 405: 907–913. HIGUCHI, R., BOWMAN, B., FREIBERGER, M., RYDER, O.A. & WILSON, A.C. 1984. DNA se- 111 quences from the quagga, an extinct member of the horse family. Nature 312: 282–284. HIGUCHI, R.G., WRISCHNIK, L.A., OAKES, E., GEORGE, M., TONG, B. & WILSON, A.C. 1987. Mitochondrial DNA of the extinct quagga: relatedness and extent of post-mortem change. J. Mol. Evol. 25: 283–287. LEONARD, J.A. et al. 2005. A rapid loss of stripes: the evolutionary history of the extinct quagga. Biol. Lett. 1: 291–295. LOWENSTEIN, J.M. & RYDER, O.A. 1985. Immunological systematics of the extinct quagga (Equidae). Experentia 41: 1192–1193. MORRIS, D. 1990. Animal watching: a field guide to animal behaviour. Jonathan Cape, London. MURRAY, J.D. 2003. Mathematical Biology II: Spatial models and biomedical applications. Springer, New York. OAKENFULL, E.A., LIM, H.N. & RYDER, O.A. 2000. A survey of equid mitochondrial DNA: implications for the evolution, genetic diversity and conservation of Equus. Conserv. Genet. 1: 341–355. RAU, R.E. 1978. Additions to the revised list of preserved material of the extinct Cape colony quagga and notes on the relationship and distribution of southern plains zebras. Ann. S. Afr. Mus. 77: 27–45. RUXTON, G.D. 2002. The possible fitness benefits of striped coat colouration for zebra. Mammal Rev. 32: 237–244. ‘The Quagga Project’. Retrieved 15 September 2006 from http://www.quaggaproject.org/ WAAGE, J.K. 1981. How the zebra got its stripes – biting flies as selective agents in the evolution of zebra colouration. J. Entomol. Soc. S. Afr. 44: 351–358. Corresponding Editor: M.I. Cherry Appendix A. Sire parentage and offspring data for body Regions 1 and 3. (Y.O.B. = year of birth.) Sire Dam Offspring Y.O.B. Region 1 Region 3 Alex Melanie Melanie Melanie Howey Charlene Sokkies Melanie Howey Melanie Brenda Charlene Sokkies Melanie Charlene Sokkies Sokkies Shaun Luke Leius Bernard Marilyn Jeanetta Nicola Erina Amanda Niki Vernon Emse Libby Quanashi Simone David 1992 1995 1999 1999 1993 1993 2000 2000 2001 1992 1991 1991 1993 1994 1994 1995 0.476 0.431 0.407 0.493 0.432 0.277 0.476 0.414 0.306 0.459 0.527 0.454 0.599 0.300 0.326 0.459 0.403 0.439 0.530 0.500 0.349 0.252 0.360 0.477 0.452 0.222 0.339 0.288 0.509 0.265 0.245 0.293 112 South African Journal of Wildlife Research Vol. 37, No. 2, October 2007 Appendix A (continued ) Sire Dam Offspring Y.O.B. Region 1 Region 3 Melanie Sokkies Melanie Melanie Melanie Howey Erina Melanie Melanie Dierdre Eric Lois Lexus Manie Hans Kerry Koos Himma Allan Reina Lulu Betty Betty Lulu Betty Lulu Betty Lulu Lulu Simone Monica Lulu Paul Mariette Bad luck Brian Hennie Vossi Monica Dale Chris Mark John Matz George Albert Brenda Tandi Emse Try-Me Tandi Tandi Tandi Tandi Brenda Celeste Try-Me Brenda Tandi Celeste Mike Morne Theresa Denise Ammy Zain Dianne Fransi Estelle Cima Andreas Erika Barbara-Anne Medee Megavolt Charlene Mariette Charlene Charlene Charlene Charlene Ziggi Zephyr Allmi Ginny Storm Kwaze Shaun Jeanetta Monica Jeanetta Rene Monica Jeanetta Monica Jeanetta Louis Ryan Lindsay Johan Leslie Caroline Elizabeth Whity 1996 1996 1997 1999 2002 2002 2003 2003 2004 Mean 1993 1994 1989 1991 1992 1992 1993 1993 1995 1996 1997 1997 1998 Mean 1993 1994 1994 1997 1990 1991 1992 1993 1994 1995 1996 1996 1996 1998 Mean 1999 1997 2000 2002 1996 1998 Mean 1997 1999 1999 2003 2000 1998 2001 2002 0.590 0.393 0.624 0.397 0.575 0.444 0.513 0.226 0.476 0.443 0.340 0.324 0.654 0.525 0.389 0.472 0.352 0.464 0.335 0.331 0.196 0.350 0.398 0.395 0.566 0.269 0.448 0.603 0.525 0.496 0.493 0.587 0.601 0.500 0.589 0.458 0.430 0.415 0.499 0.421 0.458 0.514 0.482 0.525 0.482 0.480 0.331 0.406 0.420 0.394 0.425 0.628 0.510 0.423 0.302 0.154 0.363 0.499 0.503 0.373 0.529 0.310 0.355 0.372 0.390 0.266 0.384 0.360 0.512 0.382 0.325 0.348 0.278 0.414 0.280 0.144 0.369 0.342 0.474 0.341 0.420 0.574 0.366 0.308 0.621 0.538 0.338 0.545 0.362 0.216 0.345 0.560 0.429 0.524 0.416 0.508 0.537 0.290 0.397 0.445 0.398 0.521 0.346 0.313 0.289 0.350 0.385 0.404 Parsons et al.: Genetic index for stripe-pattern reduction in the zebra 113 Appendix A (continued ) Sire Dam Offspring Y.O.B. Region 1 Region 3 Susan Nina Fanie Ricky Ricky Ricky Ricky Ricky Ricky Ricky Leon Canya Erica Stephan Teib Deon Ralph Hennie Theresa Theresa Etienne Lal Luke Lulu Mariette Amanda Mariette Lulu Zephyr Mariette Lulu Nicola Zephyr Mariette Amanda Elizabeth Duncan Cedric Gary Tracy Marjean Truida Joy Mientie Nico Butch EricH Robin Henry Ike Marcelle Marcelle Marcelle Marilyn Marcelle Marcelle Marilyn Rene Stelza Marilyn Marcelle Karl Jaunie Rene Stelza Susan Anine Audri Linda Margaret Emilene Marlene Paul Try-Me Try-Me Try-Me Medee Try-Me Erina Anine Klaus Mandy Erna Karen Connie Dolly Eddie Leon Ricky Erica Mathews Amico George Leslie Jeanetta Rebecca Tim 2003 Mean 1998 1996 2000 1997 1999 2001 2002 Mean 1998 2000 Mean 2001 2001 2003 2000 2000 2002 2002 2002 2003 2003 2003 2004 2005 Mean 2002 2003 1998 1999 1999 2001 2001 2001 2003 2004 2004 Mean 2002 1999 2000 2001 2003 2003 2004 Mean 2004 2004 Mean 2003 2003 0.445 0.442 0.307 0.525 0.299 0.331 0.286 0.456 0.482 0.384 0.363 0.402 0.383 0.299 0.492 0.339 0.394 0.361 0.418 0.250 0.357 0.433 0.366 0.496 0.373 0.185 0.366 0.320 0.380 0.380 0.367 0.398 0.093 0.416 0.309 0.367 0.367 0.449 0.350 0.452 0.412 0.400 0.471 0.531 0.621 0.459 0.478 0.497 0.325 0.411 0.457 0.563 0.229 0.359 0.421 0.231 0.245 0.337 0.319 0.411 0.391 0.336 0.316 0.433 0.375 0.203 0.288 0.234 0.378 0.538 0.498 0.215 0.411 0.400 0.459 0.583 0.424 0.056 0.361 0.449 0.307 0.203 0.402 0.289 0.254 0.370 0.292 0.274 0.398 0.347 0.326 0.503 0.381 0.502 0.466 0.533 0.538 0.462 0.483 0.394 0.297 0.345 0.536 0.581 114 South African Journal of Wildlife Research Vol. 37, No. 2, October 2007 Appendix A (continued ) Sire Dam Offspring Y.O.B. Region 1 Region 3 Monica Thor Mike Barbara-Anne Theresa Denise Paddy Fritz Zebbi Ziggi Charlene Charlene Vuyo Lawrence Etienne Tracy Tracy Douw Frank Lindsay Ricky Erika Benni Jacques 2003 Mean 2001 2002 2003 Mean 2003 2004 Mean 2003 2004 Mean 2005 2005 0.540 0.520 0.388 0.454 0.584 0.475 0.286 0.375 0.331 0.330 0.283 0.307 0.379 0.396 0.613 0.577 0.433 0.668 0.515 0.538 0.329 0.328 0.328 0.239 0.381 0.310 0.408 0.382 Mean 0.388 0.395 Appendix B. Dam parentage and offspring data for body Regions 1 and 3. (Y.O.B. = year of birth). Dam Sire Offspring Y.O.B. Region 1 Region 3 Betty Allan Allan Allan Allan Bad luck Brian Vossi Dale Brenda U Alex Albert Albert Albert Joxi Niki Mike Estelle Erika Tandi Albert Albert Albert Albert Albert Albert Barbara-Anne Ammy Zain Dianne Fransi Morne Lulu Tsjaka Allan Allan Luke Allan Allan Allan Allan Luke Luke Luke Reina Monica Mariette Marjean Hennie Chris Mark George Sebastian Duncan Mientie 1989 1991 1992 1993 Mean 1988 1991 1993 1994 1996 Mean 1996 1990 1991 1992 1993 1994 Mean 1990 1993 1994 2000 1992 1995 1996 1998 1999 2001 2002 Mean 0.654 0.525 0.472 0.464 0.53 0.323 0.459 0.566 0.601 0.458 0.48 0.430 0.525 0.496 0.493 0.587 0.269 0.47 0.340 0.352 0.324 0.361 0.389 0.335 0.331 0.398 0.382 0.299 0.357 0.35 0.384 0.36 0.382 0.348 0.37 0.318 0.222 0.474 0.338 0.216 0.31 0.345 0.366 0.308 0.621 0.538 0.341 0.42 0.294 0.325 0.266 0.538 0.512 0.278 0.414 0.369 0.322 0.203 0.411 0.36 Parsons et al.: Genetic index for stripe-pattern reduction in the zebra 115 Appendix B (continued ) Sire Dam Offspring Y.O.B. Region 1 Region 3 Sokkies Alex Alex Alex Alex Alex Jeanetta Simone Emse David Eric Melanie Alex Alex Alex Alex Alex Alex Alex Alex Alex Alex Alex Alex Libby Nicola Amanda Shaun Luke Leius Deirdre Lois Lexus Manie Koos Himma Charlene Megavolt Alex Megavolt Megavolt Alex Alex Megavolt Megavolt Ziggi Ziggi Ziggi Marilyn Allmi Ginny Vernon Quahashi Storm Kwaze Vuyo Lawrence Marilyn Ike Ike Ike Stelza Audri Emilene Monica Shaun Shaun Shaun Allan George Ryan Leslie Elizabeth Matz Thor Jeanetta Shaun Shaun Shaun Shaun George Louis Lindsay Caroline Whity Tim Try-Me Albert Paul Paul Albert Paul Paul Denise Mandy Erna Andreas Klaus Connie 1993 1994 1991 1995 1996 Mean 1993 2000 2001 1992 1995 1999 1996 1997 1999 2002 2003 2004 Mean 1999 1993 2000 2002 1991 1994 1996 1998 2003 2004 Mean 1999 2001 2004 Mean 1999 2000 2001 1997 2003 Mean 1997 1999 1998 2002 2003 Mean 1997 1999 2000 1996 2002 2003 Mean 0.277 0.326 0.454 0.459 0.393 0.3818 0.599 0.476 0.306 0.476 0.431 0.407 0.590 0.624 0.397 0.575 0.226 0.476 0.47 0.421 0.432 0.514 0.482 0.527 0.3 0.5246 0.482 0.286 0.375 0.43 0.367 0.416 0.367 0.38 0.406 0.425 0.51 0.35 0.54 0.45 0.331 0.42 0.628 0.423 0.563 0.47 0.603 0.412 0.400 0.617 0.452 0.531 0.50 0.252 0.245 0.288 0.293 0.154 0.246 0.509 0.36 0.452 0.403 0.439 0.53 0.302 0.363 0.499 0.503 0.31 0.355 0.42 0.524 0.349 0.508 0.537 0.339 0.265 0.29 0.397 0.329 0.328 0.39 0.402 0.37 0.398 0.39 0.521 0.289 0.385 0.144 0.613 0.39 0.398 0.346 0.35 0.404 0.581 0.42 0.574 0.381 0.502 0.362 0.503 0.533 0.48 116 South African Journal of Wildlife Research Vol. 37, No. 2, October 2007 Appendix B (continued ) Dam Sire Offspring Y.O.B. Region 1 Region 3 Ricky Fanie Fanie Fanie Fanie Fanie Fanie Fanie Leon Lindsay Leon Erica Canya Stephan Teib Deon Ralph Mathews Benni Howey Alex Alex Alex Bernard Erina Hans Mariette Luke Megavolt Luke Luke Luke Cedric Zephyr Tracy Joy EricH Marcelle ? Ike Ike Ike Ike Ike Genis Karl Jaunie Rene Susan Anine Theresa Hennie Hennie Mike Etienne Lal Fritz Zephyr Luke Luke Truida Butch Rene Shaun Ike Shaun Johan Linda Griet Susan Shaun Shaun Nina Rosa Tracy Etienne Etienne Douw Frank Amanda Luke Luke Gary Robin Celeste Albert Albert Cima Medee 1998 2000 1996 1997 1999 2001 2002 2004 2005 Mean 1999 2000 2002 Mean 2001 1997 2000 2002 2003 Mean 1996 2002 2003 1998 1999 2001 Mean 1998 2000 2002 Mean 2002 2003 Mean 2003 2001 2004 Mean 2003 2004 Mean 2003 2004 Mean 2003 2004 Mean 1995 1998 Mean 0.307 0.299 0.525 0.331 0.286 0.456 0.482 0.497 0.379 0.40 0.493 0.414 0.444 0.45 0.492 0.458 0.394 0.25 0.496 0.42 0.318 0.32 0.38 0.38 0.398 0.093 0.31 0.363 0.402 0.454 0.41 0.418 0.366 0.39 0.394 0.309 0.41 0.37 0.445 0.393 0.42 0.33 0.283 0.31 0.339 0.373 0.36 0.5 0.415 0.46 0.421 0.245 0.231 0.337 0.319 0.411 0.391 0.394 0.408 0.35 0.5 0.477 0.373 0.45 0.288 0.416 0.378 0.215 0.583 0.38 0.279 0.449 0.307 0.203 0.289 0.254 0.30 0.316 0.433 0.668 0.47 0.498 0.459 0.48 0.313 0.292 0.277 0.29 0.229 0.389 0.31 0.239 0.381 0.31 0.234 0.424 0.33 0.545 0.56 0.55