2.4. Statistical analysis
The experimental design for each cultivar was WH-4-023 completely randomized design, replicated four times. Statistical analysis was carried out using Analysis of Variance (SAS Institute Inc., 2011) for each cultivar. A general linear mixed model (PROC GLIMMIX) was performed to determine preharvest and postharvest main and interaction effects on tuber quality. Treatment means were separated using the Tukey’s test with 95% confidence limits.
3. Results and discussion
3.1. Preharvest effects on initial tuber quality
There were no significant differences in tuber firmness due to treatment in Season 1, with an average of 21.4 N. In Season 2, tubers were firmest at H1, and Sub-SD irrigated tubers averaged 23.5 N, 5.7 N firmer than those from the other irrigation methods (Fig. 1). However, with subsequent harvests, Sub-SD had the most significant decline in firmness, losing 9.4 N by H2. There were no differences due to irrigation method at H2 or H3.