Do you think that centralised breeding systems have a significant effect on the genetic diversity of forest?

The answer has to be yes. It applies to all plants and animals kept or grown by humans. If a particular line is seen as "good", it is used for breeding, and that strain becomes very widespread, often removing the genetic variation in the population (whether natural or domesticated).

A very good example is Holstein cattle, the black-and-white cattle usually used for dairy production all over the temperate world. The use of AI from "good" bulls over many years has made the breed extremely uniform genetically. The effective population of these many millions of animals is now under 100! (No wonder Ruth and Pip Archer want to cross-breed...)

However, I don't think you can blame Dutch Elm disease on this, at least not in the UK. The English elm and related species grow naturally by suckering, not usually by seed, so the trees in a district tend(ed) to be a single clone. This makes them highly vulnerable to evolution of the fungus to be able to attack that genetic individual.

This is why wych elm (Ulmus glabra) is usually affected less by the disease. It normally grows from seed, not suckering, so each tree is a genetically distinct individual. The fungus is less likely to be able to attack each individual equally, so many survive or are affected less. Wych elm also tends to grow mixed up with other trees rather than in solid stands, so the beetles must have more trouble finding them.

Another problem with propagation of favoured strains of otherwise wild species is that they may not be as suitable for the situtation as the locally-adapted native ones, and so may actually not do as well in real use. They are strains which have done well in a research station, and which are easy to propagate -- they may be planted in a very different climate and different soil.

Sweetheart with a question like that.. then by gosh you should know the answer..