Open Collections

Analyses of competition in binary and ternary mixtures involving a crop and three weed species

University of British Columbia

Several models exist for investigating the effects of plant competition within and among species, i.e. intra- and inter-specific competition. The models for interspecific competition are based upon additive and replacement (de Wit) series experiments. Each approach has previously been used almost exclusively to study the effects of binary mixtures, and each has been used to derive various indices of competitiveness among species. Studies were undertaken in 1980 and 1981 to compare and evaluate the different models, and to investigate the relative performance of species in ternary combinations. Monoculture (density), additive and replacement series experiments involving binary mixtures of barnyard grass (Echinochloa crusgalli), redroot pigweed (Amaranthus retroflexus ) and green foxtail (Setaria viridis) were undertaken in both years; monoculture (density) and replacement series involving binary and ternary mixtures of rapeseed (Brassica napus) with pigweed and foxtail were also investigated in both years; in 1981, the rapeseed-pigweed-foxtail experiments also included additive series mixtures, the experimental design for which permitted the investigation of binary replacement series at different total densities. Monoculture experiments showed that yield of all four species was related to the density according to de Wit's spacing formula. Additive series experiments involving barnyard grass or rapeseed with the other two species showed that the yield of the indicator species followed the Dew's relationship between yield and the square root of the density of the associated species. The present studies have shown that the yield of the latter (in the presence of the indicator species) can also be described by the spacing formula. In binary replacement series experiments, de Wit's relative crowding coefficients (k) were calculated. Estimates of yield obtained from the k-values were usually found to agree well with observed yields. Dew's Index of Competition (CI) was calculated from additive series data for each combination of species tested. Relative crowding coefficients (k), Willey and Rao's Competition Ratio (CR) and McGilchrist's Aggressivity (A) were calculated from binary replacement series data. Both k- and CR- values contain components which relate to intra- and inter-specific competition. The actual relative contributions of intraspecific and interspecific competition were calculated by comparing the effect on a given species of adding equal densities of its own kind or of a second species, to the same total density; the ratio of the former to the latter is a new parameter, termed the Interference Ratio (IR), and is related to the relative crowding coefficient. Intercomparisons of the various measures of competitiveness showed that in both years k-values were highly correlated with A and CR, and in 1980 were also correlated with CI. However, there was only a weak correlation between k and IR. In general, CR-, k- and A-values suggested that barnyard grass and rapeseed were the most competitive species. However, IR-values indicated that the greatest competitiveness was exhibited by pigweed against foxtail. Pigweed was much more sensitive to its own kind than to foxtail. Estimates of k-values for untested combinations based upon either the use of de Wit's spacing formula or upon k-values determined for binary mixtures involving each of the untested pair with a common third species were found to be unreliable. In several mixtures, k was found to be density dependent. In ternary mixtures, the effects on the yield of a given species could not be predicted from its behaviour in the presence of either of its competitors in binary combinations. For example, pigweed and foxtail behaved synergistically against high density rapeseed, but tended to act antagonistically at low rapeseed densities. Although foxtail was consistently the weakest competitor in any binary mixture, it had the greatest effect of any species in determining the competitive interaction between the other species. In order to estimate yield losses, e.g. in crop-weed systems, only additive series data are shown to be of general applicability.

Text

2010-04-15

For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use https://open.library.ubc.ca/terms_of_use.

http://hdl.handle.net/2429/23628

Plant Science

University of British Columbia

Graduate