To describe competition species must be compared with a standard. By conventional standards, such as grammes dry matter, tiller number and sq. cm leaf area, two species and even one species cannot be compared in the course of the season. Therefore a standard without dimensions was needed. Dividing yield of a species (expressed in any unit) in mixed culture by that in monoculture gave a figure without dimensions, called relative yield.
Conclusions were based on my own pot experiments under controlled conditions and on field experiments of others.
1. The rate at which one species replaces another is independent of the proportion of a species in the mixture.
2. Competitive vigour of a grass species is independent of the other species or varieties associated with it in the mixture. This implies that if there are two experiments in which species A is competing with each of two morphologically identical varieties of species B, the result of competition between the varieties of B can be calculated. Competition in mixtures of two species will predict the competitive trend in more complex mixtures. In all these experiments the monocultures may never be omitted.
3. When a treatment limits growth, in general competition changes at the expense of the more productive species. This is because there will be less competition for light as yield decreases, so benefiting the lower yielding species. Under these conditions the shoot/root ratio of the lower yielding species seems to decrease, another reason why the lower yielding species benefits in competition by suboptimum soil conditions. The above- ground production of a species with a large proportion of roots will respond less sharply to a shortage of minerals or water than a species with a higher shoot/root ratio. Hence, the change in competition in adverse conditions does not depend on the nature of the factor limiting growth.
4. Sowing experiments show that pH considerably affects the development of seedlings, even though pH values within the range 4.2 - 6.8 does not affect competition between fully grown plants. In general the period of establishment and the competition afterwards should be stricktly separated. If this is omitted, completely erroneous conclusions may be drawn about the competitive ability of the species.
5. Monocultures are preferred to mixed cultures for the following reasons.
a. Because the higher yielding species in a mixture is not always the dominant species (Montgomery effect); the apparent advantage is often even a drawback. The risk of the lower yielding species becoming dominant should not be underestimated.
b. The possibility that light is better utilized through the sod structure in a mixed culture can only be important during the period that the leaf canopy of the sod is not yet closed. In absolute terms this advantage is only small, especially if yields per cut are high; under grazing this possibility may be of more importance.
c. A mixture of species with different growth rhythms might yield more. If so the increased yield will only be worthwhile if the species alter their order of dominance within the growing season. But changes in botanical composition are more gradual, so that this advantage of a mixture is lost.
d. When the more productive species fails for some reason, the less productive species in the mixture could temporarily take over the production. But this temporary advantage may change into a persistent drawback, when recovery of the main species is hampered by the other species. On sown permanent grassland, especially under suboptimum conditions, less productive or otherwise undesirable species may in time invade the sward. It remains to be seen whether this is more efficiently prevented by a mixture or by a monoculture.
e. The problem in sowing grassland is the choice of the highest yielding species under the prevailing conditions, rather than the choice of the highest yielding mixture.