Mammalian Genome

, Volume 19, Issue 10, pp 713–723

Morphometrics within dog breeds are highly reproducible and dispute Rensch’s rule

Authors

  • Nathan B. Sutter
    • Cancer Genetics BranchNational Human Genome Research Institute, National Institutes of Health
    • Department of Clinical Sciences, College of Veterinary MedicineCornell University
  • Dana S. Mosher
    • Cancer Genetics BranchNational Human Genome Research Institute, National Institutes of Health
  • Melissa M. Gray
    • Department of Ecology and Evolutionary BiologyUniversity of California
    • Cancer Genetics BranchNational Human Genome Research Institute, National Institutes of Health
Article

DOI: 10.1007/s00335-008-9153-6

Cite this article as:
Sutter, N.B., Mosher, D.S., Gray, M.M. et al. Mamm Genome (2008) 19: 713. doi:10.1007/s00335-008-9153-6

Abstract

Using 27 body measurements, we have identified 13 breed-defining metrics for 109 of 159 domestic dog breeds, most of which are recognized by the American Kennel Club (AKC). The data set included 1,155 dogs at least 1 year old (average 5.4 years), and for 53 breed populations, complete measurement data were collected from at least three males and three females. We demonstrate, first, that AKC breed standards are rigorously adhered to for most domestic breeds with little variation observed within breeds. Second, Rensch’s rule, which describes a scaling among taxa such that sexual dimorphism is greater among larger species if males are the larger sex, with less pronounced differences in male versus female body size in smaller species, is not maintained in domestic dog breeds because the proportional size difference between males and females of small and large breeds is essentially the same. Finally, principal components (PCs) analysis describes both the overall body size (PC1) and the shape (length versus width) of the skeleton (PC2). That the integrity of the data set is sufficiently rich to discern PCs has strong implications for mapping studies, suggesting that individual measurements may not be needed for genetic studies of morphologic traits, particularly in the case of breed-defining traits that are typically under strong selection. Rather, phenotypes derived from data sets such as these, collected at a fraction of the effort and cost, may be used to direct whole-genome association studies aimed at understanding the genetic basis of fixed morphologic phenotypes defining distinct dog breeds.

Supplementary material

335_2008_9153_MOESM1_ESM.doc (2.4 mb)
MOESM1 (DOC 2476 kb)

Copyright information

© Springer Science+Business Media, LLC 2008