PLOT SIZES FOR EXPERIMENTS DESIGNED TO ASSESS CHEMICAL CONTROL AND CULTIVAR RESISTANCE LEVELS TO POTATO LATE BLIGHT
ESG MIZUBUTI1, HS MAYTON1, CD SMART1 M GRASMICK2 and WE FRY1
1Department of Plant Pathology, Cornell University, Ithaca, NY 14853, USA; 2Frito-Lay, Inc., 4295 Tenderfoot Rd., Rhinelander, WI, 54501, USA.
Background and objectives
Each year field experiments are conducted in order to evaluate fungicide efficacy and cultivar resistance against potato late blight caused by Phytophthora infestans. When field space and other resources are limiting, the size of the plots influence the number of treatments and the number of replications. There are trade-offs among size of plots, number of treatments and number of replications. It is expected that larger plots (in regard to smaller plots) will improve accuracy (greater precision and lower variance) and will permit greater magnitude of differences among treatments. Small plots permit a greater number of treatments and/or a greater number of replications (greater chance to discern significant differences). The size of the plots should maximize area usage and accuracy of results. The objective of this research was to estimate the smallest plot that could maintain magnitudes of differences among treatments and a sufficiently low variance.
Materials and Methods
Four experiments each using a different plot size, of two (two plants in a row), five (five plants in a row) , 12 (two rows of six plants) or 64 plant-plots (four rows of 16 plants) were carried out with the same treatments. The treatments were cultivar resistance level, Frito Lay (FL)1533 (suscept.) and FL 1625 (mod. resist.), and dose rate of chlorothalonil, 0, 1/4, 1/2 and 1 (full) of the recommended dose (1.4 ;l/ha). The 2, 5, and 12 plant-plots were sprayed with a handheld sprayer and the 64-plant plots were sprayed with a tractor-mounted boom sprayer (standard procedure). Late blight intensity was rated on nine different days using a severity scale. The area under the disease progress curve (AUDPC) was calculated and all statistical analyses were done on these values. The interactions between plot size and treatments (cultivar and dose) were tested with ANOVA.
Results and Conclusions
There was no significant interaction between plot size and cultivar resistance level (P=0.207). Differences in cultivar resistance levels can be detected using plot sizes as small as 2-plant plots. The effects of fungicide dose varied with plot size (P=0.001). At the full dose of chlorothalonil, the final AUDPC values in the two and five plant-plots were 590 and 570, respectively, whereas at larger plot sizes, 12 and 64 plant-plots, the final AUDPC values were 251 and 339, respectively. A possible explanation would be a higher ratio of external/internal inoculum in the small plots compared with the larger ones, which could contributed to increase the AUDPC . The coefficient of variation (CV) for the 2, 5, 12 and 64 plant-plots experiments were similar: 11.1, 18.1, 15.5, and 13.0%, respectively.
The 12-plant plots had similar values of AUDPC to the 64 plant-plots (standard size), had low CV, good canopy development, allowed uniform inoculation and were easy to spray. Therefore, they were considered as adequate plot size. These are preliminary results and the experiments will be repeated in 1998.
1. Paysour RE, Fry WE, 1983. Phytopathology 73, 1014-20.