PROPOSITION OF A SIMPLE METHOD OF INTERPRETATION OF VARIETAL EXPERIMENTS

One of the jobs carried out by the section of vegetables cultures of the Federal Station of agronomic research of Lausanne is the testing and the examination of new varieties of strawberry plant presented on the market. It is thereby aimed to present every year a varietal assortment recommended nationally as well as regionally or locally.

As a rule, the experiments are made according to the following plan:

1. 1st year: preliminary test (without any repetition) of all the new varieties which have been proposed and choice of a rather wide group of the best, on the base of the results and of a discussion between experts.
2. 2nd year: thorough test of the group which has been chosen after the first year (three repetitions, one planting-date); elimination of the varieties of low yield or presenting a major deficiency (taste, colour, behaviour when transported, etc.).
3. 3rd year: main test. The remaining varieties are tested simultaneously in different regions (as a rule three or four regions: near the Lake Léman, mountain, Plain of Valais and sometimes The Ticino).
4. 4th year: same as 3rd year. Definition of the recommended assortment. During the following years, the chosen varieties are analyzed one by one in order to reach precise cultural conditions for each variety, such as planting-date, distances in the rows…

However, this plan always raises controversy when the assortment is being determined. Indeed, the statistical study generally centred upon the yields, affords us to be more or less certain when isolating varieties or groups of varieties of different yields, but provides us no information relatively to the other criteria, such as average weight of the fruits, percentage of first quality, rapidity of gathering, precocity, etc., whose importance is not to be rejected in the varietal valuation.

Though all the figures necessary to value the other criteria are known, it is very difficult and even impossible to word and to represent by one resulting value the whole of the answers obtained on the spot.