Study 03 · University of Porto · 2018

Lima Thesis
University of Porto

Original document

Lima, D. F. C. (2018). Unraveling the evolutionary history of livestock dog breeds. Master's thesis, Universidade do Porto · PDF · 2.5 MB

⬇ Download PDF
Study factsheet

Independent confirmation
from Portugal

AuthorDiogo Filipe Coutinho Lima
SupervisorRaquel Godinho (CIBIO-InBio / Universidade do Porto)
Co-supervisorsIgnacio Doadrio (CSIC) and Pedro Silva (CIBIO-InBio)
InstitutionFaculdade de Ciências, Universidade do Porto (FCUP) — CIBIO-InBio
Year2018
TypeMaster's thesis — Mestrado em Biodiversidade, Genética e Evolução
Main focusEvolutionary history of livestock guarding and herding dogs (Iberia and Eurasia)
Markers (the PLE is here)43 autosomal microsatellites and 6 from the Y chromosome — 294 dogs from 15 Iberian breeds
Markers (guarding dogs only)Illumina CanineHD chip, over 170,000 SNPs — 24 Eurasian guarding breeds
PLE samples13 specimens (Spanish Wolfdog)
AnalysesGenetic diversity, STRUCTURE, haplotype networks, genetic distances (FST)

This study is particularly valuable because it was not designed to study the PLE. Its subject is the evolutionary history of livestock guarding and herding dogs; the Spanish Wolfdog is just one of the fifteen Iberian breeds analysed. That a Portuguese team, using different markers and with no stake in the lobero's cause, places the PLE exactly where the Spanish studies place it is independent cross-validation: the most solid thing science can offer.

The scope should be made precise: the high-density genomic analysis (over 170,000 SNPs) was applied only to the 24 Eurasian guarding breeds. The Perro Lobo Español, as a herding dog, was analysed with the 43 autosomal microsatellites, alongside the other Iberian breeds. The PLE data in this study therefore come from the microsatellites, not from the SNP chip.

Breeds analysed

The Iberian block analysed with microsatellites brings together fifteen guarding and herding breeds, Portuguese and Spanish: Cão de Castro Laboreiro, Serra da Estrela, Cão de Gado Transmontano, Rafeiro do Alentejo, Mastín Español, Mastín del Pirineo, Montaña del Pirineo, Cão da Serra de Aires, Barbado da Terceira, Cão de Fila de São Miguel, Cão de Água Português, Pastor Vasco, Carea Leonés, Can de Palleiro and the Spanish Wolfdog —the Perro Lobo Español—.

One point that has often been misread deserves emphasis: this study does not include the German Shepherd or the Czechoslovakian Wolfdog. Its comparative frame is entirely that of Iberian and Eurasian working dogs. It therefore offers —and claims to offer— no direct comparison of the PLE with those two breeds. That comparison is the one provided by the Doadrio (2014) and Villegas (2017–2019) studies.

Key results

The PLE and the Carea Leonés in the same cluster

This is the work's most striking finding. In the Bayesian STRUCTURE analysis, part of the Carea Leonés specimens do not group with the other dogs of their own breed, but in the same cluster as the Perro Lobo Español. A shared genetic background among working dogs of the north-west of the peninsula, detected by a team that was not looking for it.

Normal diversity, no bottleneck

With 43 microsatellites, the PLE shows an expected heterozygosity of 0.67 (observed 0.63), in the mid-to-high range of the Iberian herding dogs analysed, which runs from 0.63 (Cão da Serra de Aires) to 0.69 (Carea Leonés). These are values typical of an established, diverse population: there is no trace of the genetic impoverishment that betrays recently created breeds.

Iberian model versus Central European model

Modern and reconstructed breeds show lower diversity and closed clusters. Iberian functional populations —including the PLE— show gradients, historical overlaps and high variability. These are different realities that require different frames of analysis.

What this study does not allow us to claim

Honesty about the limits of the evidence is what makes it credible. In this work, the Perro Lobo Español is represented by thirteen specimens: a small sample, enough to place the population on the Iberian genetic map, but not enough to describe its internal structure in detail. The heterozygosity values, moreover, are not directly comparable with those of Doadrio (2014) or Villegas (2019), because the marker panels differ —43 microsatellites here against the 18 of the ISAG panel— and each panel produces its own figures.

What this study does establish, and it is no small thing, is that the PLE behaves as one more Iberian population, related to the working dogs of its territory, with no signs of recent creation. Nothing more, and nothing less.

The most relevant conceptual contribution

The thesis states explicitly that Iberian dog populations cannot be interpreted through the Central European model of a breed. The PLE's isolation is functional and geographical, not aesthetic. Historical gene flow did exist, but within a coherent regional framework.

This scientifically legitimises the PLE as a native population, not as an «artificial breed» nor as a derivative of the German Shepherd. The problem is not that the PLE does not fit the FCI model: it is that this model was not designed for realities such as the PLE's.

The three studies: cross-validation

STUDY 01 · DOADRIO 2014

The PLE separates from the German Shepherd.

STUDY 02 · VILLEGAS 2019

The PLE has high genetic diversity and forms an Iberian cluster of its own.

STUDY 03 · LIMA 2018

The PLE fits within the Iberian continuum and shares a cluster with part of the Carea Leonés.

Three different teams, three different marker panels, no shared interest — and one and the same result. The Perro Lobo Español is an ancient functional population, genetically diverse, distinct from modern lupoid breeds and naturally integrated into the Iberian system of rustic dogs. When studies that were not designed to agree do agree, in science that has a name: convergence of evidence.