Variety X management interactions of red clover (Trifolium pratense L.)

Abstract

Field studies at the Welsh Plant Breeding Station, Aberystwyth were designed to evaluate the responses of diploid and tetraploid representatives from early and late flowering groups of red clover, to various autumn defoliation managements. A cutting trial was established comprising the four varieties under four autumn/winter defoliation managements; undefoliated or defoliated once, twice or three times during this period. During spring/summer all plots experienced the same management which was three silage-type cuts. DM measurements were taken at each defoliation. Plant counts, morphological measurements and chemical and quality analyses were carried out at appropriate times during the trial. Plants were removed from the field to a depth of 400mm at the end of the first winter, planted in soil-filled boxes and placed in a heated light-exclusive chamber. Thus an attempt was made to establish the importance of the level of carbohydrate root reserves in initiating structural regrowth under favourable conditions, but in the absence of photosynthetic products.

The same autumn/winter treatments were imposed on a year-old established sward of an early flowering diploid red clover variety, but defoliation was effected with sheep rather than by mower. Sane comparison was thus obtained between cutting and grazing.

Discriminant and multiple regression analyses were used to identify the factors most sensitive to autumn/winter management for each variety, and to see if predictive models could be established for DM yield from measured components of yield. Detailed measurements were taken at the end of winter in the first year, and before the first and second spring/summer harvests of the same year, to provide the required data. Both techniques of analysis were found to be useful, with high levels of classification obtained for varieties, and accurate predictive models established; e.g. 74% of the variation in DM yield at the first spring/summer harvest was explained from an equation of factors measured at the end of winter for the diploid late flowering variety. This increased to 81% of the variation explained by just two factors in the grazing trial. Overall the most important factors in determining yield variation were root carbohydrate levels and plant number per unit area.

Genotype x environment regression analysis showed that early flowering varieties were detrimentally affected by autumn/winter defoliation to a much greater extent than late flowering varieties. Diploids were more sensitive than tetraploids within each flowering group. At the second spring/summer harvest there was evidence of a compensatory mechanism operating for treatments defoliated two and three times in autumn/winter. This resulted in there being very little difference in TIM between defoliation treatments. This was especially so for the late flowering varieties.

The dark room experiment showed a direct relationship between the concentration of total available carbohydrate of red clover roots and their ability to produce structural growth in the absence of light.

A controlled environment experiment was carried out at Lincoln College using three day/night temperature regimes likely to be experienced in autumn in temperate environments. The same four red clover varieties were grown and the plants subjected to four defoliation regimes during the course of the experiment. This experiment demonstrated that warm temperatures in autumn markedly increased red clover growth, particularly in early flowering varieties. This was, however, associated with a lowering of carbohydrate root reserves which were lowered still further with any defoliation.

These results are interpreted and discussed in terms of present-day management of red clover swards on intensive grassland farms.