Insights Regarding Sensory Evaluation of Bitterness Development in Citrus Juice

Delayed bitterness is a well-known phenomenon in citrus juice and has a negative impact on juice quality.  Bitterness results when the tasteless limonoic acid A-ring lactone (LARL) in juice is converted to the bitter compound limonin after juicing.  Citrus varieties that produce juice that becomes bitter have limited utility for processing.  USDA 1-105-106 is a promising new mandarin hybrid that produces juice with outstanding initial quality; however, the development of bitterness in this hybrid had not yet been quantified.  The objective of this study was to quantify the kinetics of delayed bitterness in USDA 1-105-106, ‘Ambersweet’ (a sibling of USDA 1-105-106), and ‘Navel’, a sweet orange prone to development of bitterness.   Fruits were harvested at weekly intervals from November to December of 2009, except ‘Navel’ which was only harvested in December, and juiced. Juice was divided into 4 one-L bottles and served to a trained panel within one hour of juicing, or after 4 and 24 hours at 10 °C.  One additional bottle was stored at -20 °C and served after 4 weeks.  Aliquots of juice from each harvest and sampling time were taken and subsequently analyzed for titratable acidity (TA), soluble solids (SS), sugars, acids, limonoids (limonin, nomilin and their glucosides) and volatiles.  Ten trained panelists rated the juice for sweetness, sourness and bitterness on a 0-10 scale. Half of the panelists could discern bitterness in USDA 1-105-106 in samples stored for 24 hours as well as those that were frozen, but bitterness was not detected in ‘Ambersweet’.  Variation in bitterness sensitivity is well-described and was observed in this specific group of tasters.  Nevertheless, frozen stored ‘Navel’ juice was rated high in bitterness by all panelists.  ‘Ambersweet’ tended to be perceived as more sour, which correlated with higher TA and lower pH.  USDA 1-105-106 had generally higher nomilin levels, regardless of sampling time. There were no sensory character differences due to harvest date, although the first harvest had the lowest pH and TA in USDA 1-105-106, and highest levels of nomilin and limonin in ‘Ambersweet’.  This study clearly showed sensory differences between sampling times for USDA 1-105-106 and ‘Navel’ juices but not for ‘Ambersweet’.  Our results indicate that sensory evaluation should be performed at least after 24 hours of juicing, and/or after freezing to effectively detect potential problems due to delayed bitterness.