Abstract
The domesticated European carp was subjected to a two-way selection for growth rate. Five generations of mall selection for faster growth rate did not yield any response, but subsequent selection between groups (families) resulted in considerable progress while maintaining a large genetic variance. Selection for slow growth rate yielded relatively strong response for the first three generations. Random-bred control lines suffered from strong inbreeding depression and when two lines were crossed, the F1 showed a high degree of heterosis. Selection was performed on pond-raised fish, but growth rate was also tested in cages. A strong pond-cage genetic interaction was found. A theoretical explanation was suggested involving overdominance for fast growth rate and amplification through competition of intra-group but not inter-group variation.