Growth Analysis of Atrazine-Resistant and Atrazine-Sensitive Biotypes of Chenopodium album

Abstract

Seed from atrazine-sensitive and atrazine-resistant biotypes of lambsquarters ( Chenopodium album L.) was sown in pots containing quartz sand. Plants were grown under a controlled environment in a growth room. On the 15th day after planting, four randomly-selected plants of each biotype were harvested, and both fresh and dry weights of leaves, stems, roots and whole plants, together with leaf area, were determined for each plant. This was repeated at 5-day intervals until the 60th day after sowing. Data from the resulting replicated 2x 10 factorial configuration were analysed using BMDP multiple regression programmes and orthogonal polynomials to produce best fit polynomial expressions of time-toharvest and biotype for the natural logarithm of each response. From these empirical models of plant growth, predictor functions for relative growth rates, leaf area and weight ratios, specific leaf area and unit leaf rate were generated. Examination of computer-generated plots of the various growth indices suggested that time-of-harvest was the most important factor in determining the magnitude of the responses. However, the two biotypes exhibited different growth patterns as indicated by the presence of a significant biotype effect or biotype x time interaction in all cases. The faster-maturing atrazine-sensitive plants tended to have fairly dense leaves with relatively small area whereas the atrazine-resistant plants initially produced small root systems and relatively low density leaves. Moreover, the resistant biotype, despite being lighter at 2 weeks, weighed the same as the sensitive plants by the 60th day due to a consistently higher relative growth rate. The advantages of using balanced or orthogonal configurations together with multiple regression procedures to derive data-based empirical models of plant growth are discussed. Limitations to the routine use of empirical models are also considered.