Late blight is a polycyclic disease, having several cycles of infection and inoculum production during one growing season. Thus, the infection may be expected to increase proportionally to both the initial amount of inoculum present and the amount of new inoculum produced during the growing season. The amount of inoculum produced is a factor of the host, pathogen, environment and management conditions. When dealing with polycyclic diseases there are several ways to measure resistance, and temporal analysis parameters (disease progress curves) are advised. CIP recommends the area under the disease progress curve (AUDPC) to measure resistance to late blight. The AUDPC is a resistance parameter calculated from the percentages of leaf area affected estimated at different times during the epidemic. The advantage of using AUDPC is that it is simple to calculate, uses multiple evaluations, and does not rely on transformations, which often are based assumptions about the distribution of data points. The AUDPC also has some disadvantages that are discussed below.
1. Experimental design
Evaluation of advanced clones is conducted in randomized complete block design, using a plot size of 5 hills per entry, and 4 replications. It is important to include resistant and susceptible clones (controls) in the experiment. Independent of the number of trials and sites used, clones should be tested over a minimum of two years. Spreader rows should be planted in the borders of each plot to assure uniformity in the inoculum distribution throughout the whole experimental period.
2. Evaluation of disease
To obtain the AUDPC, disease readings are taken on the basis of the percentage of leaf area affected by late blight, at intervals of five to 14 days after infection starts on susceptible clones. Since it is difficult to know exactly when symptoms will start to appear, the first reading is planned for 30 to 40 days after planting, during which time the field should be observed carefully for initiation of disease. Resistance evaluation experiments performed in the field at CIP are conducted under natural inoculum pressure. The test plants are protected from infection with a contact fungicide until well established, usually for the first 30 days after planting. After that date, the application of fungicide is ceased and the plants are exposed to the natural inoculum. Seven days later, evaluation starts and readings are repeated every 5 to 14 days. The frequency of readings depends on the rate of disease increase, which in turn depends on the level of resistance of the materials being tested and the weather. One of the strengths of the AUDPC is that readings do not have to be done on a common schedule in order to compare clones, cultivars or treatments. Readings should continue until the susceptible cultivars reach high levels of disease, near 100 %. It is best to record readings independently (i.e., without knowing the value given at the previous reading) at each date, such as by having someone else write in the field book or by using a cassette recorder.
3. Data analysis
When data have been collected sequentially through the season, the AUDPC is calculated in order to compare cultivars or treatments. The AUDPC is most frequently calculated using the midpoint rule method (Campbell and Madden, 1991). The formula is:
Where yijk is the percent severity observed at time k for the ith treatment in the jth block, tk is the Julian date of the observation, and observations were made on n dates. The variable tk can also be days after planting or emergence, because the absolute value does not matter, but rather the interval between readings.
The AUDPC can be calculated using virtually any statistical analysis program or spreadsheet. Here we give an example using Excel.
Enter the evaluation of the five clones under the corresponding days written in the second line of the table. Next, locate the cursor in cell I4, which corresponds, to the area of clone 1, and calculate the area using the following formula.
Area = ((C4+B4)/2)($C$2-$B$2)((D4+C4)/2)($D$2-$C$2)((E4+D4)/2)*
Press [Enter] and the area (578) will appear in the cell I4. Then copy this cell to the other cells from I4 to I8. Table 2 shows the results.
Table 1. Percentage of foliar area affected by late blight and corresponding AUDPC calculated by a spreadsheet EXCEL.
4. Interpretation of parameters obtained
The AUDPC represents the area under the actual infection curve. It is expressed in %-days and interpreted directly without transformation. The higher the AUDPC the more susceptible the clone or variety. It is often helpful to plot the percentage of leaf area infected versus evaluation date to get a better idea how resistant and susceptible clones or varieties perform in the evaluation experiment (see figure 1).
5. Data recording and summary
A data sheet is included to give an idea of how data may be recorded. Measurements of plant development (flowering, vigor) are generally recorded in addition to disease severity. If the AUDPC is to be calculated the date of each disease severity reading must also be recorded. Data are collected on each experimental unit (each clone or variety within each replication) and the AUDPC is calculated for each experimental unit. Simple statistics (e.g., mean, std.) or analyses (ANOVA, regression) can be calculated using the AUDPC values.
6. Getting around common problems associated with the AUDPC
As mentioned, the AUDPC is robust in the sense that it allows for different intervals between readings. However, the AUDPC can be biased if readings begin too long after disease initiation. Part of the disease progress curve of susceptible materials will be lost. For this reason, it is import to start readings as early as possible.
The AUDPC is also sensitive to lack of uniformity of disease in the field. This can be corrected by blocking in the "correct" direction if one can guess how the disease severity pattern will be in the field.
The AUDPC can be biased if readings are taken too long after susceptible cultivars reach high levels of disease. If this occurs, difference between resistant and susceptible materials may be underestimated. For this reason it is important to stop readings when susceptible materials are severely infected. Alternatively, different AUDPC’s may be calculated, one including all materials and a reduced number of readings, and another including more readings but excluding the most susceptible materials.
The AUDPC is generally not comparable across experiments. In an effort to standardize the AUDPC, researchers often use the relative AUDPC. This is calculated by dividing the AUDPC by the "maximum potential AUDPC". The maximum potential AUDPC is simply the AUDPC a variety or clone would have if it had 100% infection at all readings. Therefore, the maximum potential AUDPC is calculated by multiplying the total number of days between the first and last readings by 100.
Campbell C.L., Madden L. V., 1990. "Introduction to Plant Disease Epidemiology.," John Wiley & Sons, New York City.
Fry, W.E. 1978. Quantification of general resistance of potato cultivars and fungicide effects for integrated control of late blight. Phytopathology 68: 1650-1655