On the origin of the cyanogenic polymorphism in Trifolium repens L.

Trifolium repensL. andTrifolium nigrescensViv. are two of the approximately six cyanogenic species known in the genus Trifolium. The two species are closely related:T. nigrescensis considered to be one of the diploid ancestors of the amphidiploidT. repens. We studied morphology, meiosis and the cyanogenic system inT. repens(amphidiploid),T. nigrescens(diploid) and their reciprocal hybrids. A comparison of the enzyme linamarase in the species and hybrids shows that there is a general resemblance between their linamarases. Immunological studies indicate that the linamarases must have a somewhat different three‐dimensional structure. These facts are consistent with the view thatT. nigrescens(or an ancestral form of this species) has donated theLigene ofT. repens. The other putative parent,T. occidentalehas probably not donated an activeLigene. The hypothesis of the origin of theLigene does not explain its polymorphism in European populations ofT. repens, asT. nigrescensis monomorphic for cyanogenesis and amphidiploids do not segregate for genes which are homozygous dominant in one of the parents. Segregation forLicould be caused by a gene mutation or a small exchange between homeologous chromosomes. The latter event is more probable. A nigrescens‐repens exchange would give rise to a chromosomal region with reduced homology to both parental chromosomes. The genes in the region of exchange will be tightly linked due to diminished cross‐over frequency. It has been known for years thatLihas effects on the vegetative and reproductive characters inT. repensand we have recently shown that these effects must be the result of genes linked toLi. As the associated characters influence the fitness of the cyanotypes, not only the origin but also the maintenance of the cyanogenic polymorphism is closely related to the evolutionary history ofT. repens .