Sensory evaluation

Wines from the 2010 and 2011 seasons were analyzed by descriptive analysis and overall difference test, respectively. A descriptive sensory analysis was conducted on the wines of the 2010 season over the course of three weeks during June 2011. A trained panel (n = 15; 9 males and 6 females) recruited from the Prosser community, including members of staff of WSU-IAREC and two commercial wineries located in the Yakima Valley area of WA, was convened. Demographic aspects were recorded at the beginning of the first session and no information about the nature of the study was provided to the panelists in order to reduce bias. Panelists were screened for bitterness sensitivity (sensitivity to 6-n-propylthiouracil (PROP), also known as PROP status) and color blindness [33] and subsequently trained during seven consecutive sessions each lasting between 45 min to 1 hr. Terminology development occurred by consensus of the whole panel. After reviewing the standards, panelists evaluated a series of 10 commercial wines, and discussed the intensity of the different attributes relative to the standards using a 15 cm scale. Reference standards for aroma and color were used for this purpose as described earlier [31]. The overall performance of the panel and individual panelists was evaluated leading to the elimination of three panelists based on preliminary analysis of the data, interaction plots and analysis of outliers. The remaining 12 panelists (df = 11) evaluated the two experimental wines from infected and control vines. To avoid bias due to color, tulip-shaped cobalt black glasses (Libbey, Toledo, OH, USA) were used for aroma and astringency evaluation and clear wine glasses (ISO 3591:1977) used for color evaluations. Panelists assessed the wines in individual booths under white light, at the Sensory Laboratory of WSU-IAREC, Prosser. Each panelist was provided with deionized filtered water (Easy Pure II, Thermo Scientific, Dubuque, IA, USA) and unsalted crackers (Great Value, Bentonville, AR, USA) for palate cleansing in between evaluations. Twenty-five mL aliquots of wine at room temperature (20 ± 1°C) were poured into wineglasses coded with three-digit random numbers and covered to trap volatiles. Wines were presented using a complete randomized design including the three replicates for each wine during four evaluation sessions. Results were collected on ballots and manually decoded with a ruler. All panellists involved in the study signed an informed consent form and the project previously approved by the WSU Institutional Review Board for human subject participation.

A forced-choice triangle test was selected to explore potential differences among the wine produced from symptomatic and non-symptomatic vines during the 2011 season. The overall difference test was selected for the 2011 wines due to initial assessment of the wines showing relatively minor differences between treatments. Previous informal pre-screening of the wines carried out by three experienced wine tasters revealed that wines from the two treatments, and their replicates (2 × 3) were free of off-odors or other taint aromas and thus were suitable for sensory evaluation. Since the main goal of the triangle test is to determine an overall sensory difference between the wines, special emphasis was placed on the control of the type I error (i.e. α-risk). All panelists were recruited from the WSU-IAREC community. Demographic aspects such as age, sex, and red wine frequency consumption were recorded at the beginning of the test. Thirty-three consumers (n = 33; 17 males and 16 females) aged between 21 and 60 years participated in the test. Further demographic information indicated that members of the consumer panel were composed by light to moderate wine consumers with about 79% of them declaring that they drink wine at least once a month and 42% drinking red wine at least once a week (data not shown). Panelists were briefly introduced to the mechanics of the triangle test but no information about the nature of the study was provided in order to minimize bias. Panelists were tested for visual disorders as described above and results of this test indicated that none of the panelists had color deficiencies. Aliquots of 25-mL coded wines, consisting on two treatments (symptomatic and non-symptomatic) and three replicates per treatment, were presented in a complete randomized design at room temperature (20 ± 1°C) as described for the 2010 wine evaluations and panelists were presented with three samples per flight, for a total of two flights. In the first flight, the wines were presented in transparent glasses. In the second flight, a new set of three wines was presented in black glasses. Evaluations were recorded on a ballot designed according to Meilgaard et al. [34]. As in the case of the descriptive analysis, all panelists involved in the study signed an informed consent form previously approved by the Washington State University Institutional Review Board for human subject participation.