Wright’s shifting balance theory of evolution in artificial breeding programmes: empirical testing using the model organism Tribolium castaneum [Note 1. Journal Paper Number 15754 of the Purdue University Agricultural ...]

A subdivide/merge breeding system, based on Sewall Wright’s Shifting Balance Theory of Evolution, was modelled using Tribolium castaneum (the red flour beetle). Our objective was to determine the effects of population structure on selection response for increased offspring number, a fitness trait with large nonadditive gene effects. Two different population structures were established: a subdivided population that underwent periodic merging and an undivided control population. The experiment consisted of (1) 150 beetle pairs divided into 10 lines of 15 pairs each, and (2) an undivided control population of 150 beetle pairs. Selection of the subdivided population was performed within lines each generation and in generations 4, 8 and 12, poorer performing lines were eliminated with the remaining lines merged. After merging, selected individuals were randomly paired and re‐subdivided into 10 lines of 15 pairs each. The undivided population was mass selected for increased offspring number each generation using the same selection intensity as the subdivided lines. After 14 generations, selection response in the subdivide/merge system was not significantly different from the control population; however, within each generation, there were lines in the subdivided population that had average larva numbers exceeding those of the undivided population. These results indicate that subdivide/merge systems could be advantageous for shorter‐term breeding objectives. If superior lines are isolated, they can be full‐sib mated or selfed to produce highly inbred lines which can then be crossed – a standard plant breeding practice.