Modelling interactions in grass-clover mixtures

The study described in this thesis focuses on a quantitative understanding of the complex interactions in binary mixtures of perennial ryegrass (Lolium perenne L.) and white clover (Trifolium repens L.) under cutting. The first part of the study describes the dynamics of growth, production and the structural characteristics of contrasting grass and clover cultivars under field conditions. This basic information is used in the second part to quantify light absorption, C0 2 assimilation, radiation use efficiency (RUE) and light competition of the species using a modelling approach.Both species showed a seasonal pattern in growth of the dry matter (DM) and leaf area index (LAI) during the season so that the grass-dominated swards during spring shifted to clover dominance in summer. Without N fertilisation (-N), this seasonality was mainly controlled by the weather conditions. However, in fertilised mixtures (+N) grass was the dominant component of the mixture during the whole season and clover growth was always limited by light. The competitive ability and persistence of clover were determined by the structural characteristics of the cultivars. In the -N swards, both large and small-leaved clover cultivars had a higher proportion of their leaf area at the top canopy layers than their companion grass. In the +N mixtures, this was observed only in largeleaved clover, whereas the small-leaved cultivar was strongly overtopped. The difference between cultivars; was mainly due to the pattern of allocation of the DM into the supporting tissues.Experimental results showed two sources of vertical heterogeneity within the mixed grass-clover canopies: different patterns of LAI distribution and leaf dispersion. The validity of the canopy light partitioning model was considerably improved by introducing these sources of variation into the model. Using this model the RUE of species was calculated for different regrowth periods. Grass and clover had a different RUE in mixture and monoculture, but overall RUE was higher in grass, particularly in spring. The variation in the DM yield of grass under different treatments was due to changes in RUE and absorbed radiation. However, in clover these differences were mainly due to the amount of absorbed radiation. Quantification of light competition showed that in spring grass was the strongest competitor. In summer, the competitive ability of clover was related to N level and clover cultivar. Without N, both large and small-leaved clover were better competitors than grass. However, in the +N swards only the large-leaved clover had the same competitive ability as its companion grass. In both species a leaf N profile developed during regrowth, parallel to the light profile within the canopy. The effect of the observed compared to a uniform leaf N profile on canopy C02 assimilation of species was low, but it was different between mixtures and monocultures.It was concluded that the effect of canopy structure on productivity of species was more important than their assimilatory characteristics. The persistence of white clover under cutting may be improved by choosing cultivars with a higher competitive ability, based on canopy structure.