Sensory Characteristics and Consumer Acceptance of Mechanically Harvested California Black Ripe Olives S.M. Lee, C. Sirimuangmoon, A. Gomez-Rico, K. Kitsawad, U. Rosa, J. Burns, W.H. Krueger, E. Fichtner, L. Ferguson and J-X. Guinard April 2, 2012
Overview and goal of the project Table olives produced in the US are hand-picked, a labor intensive and expensive practice. Mechanical harvesting is particularly challenging with table olives. The limiting factors are; 1) fruit damage that affects the final quality (our focus); 2) effective fruit removal technology; and 3) developing an economically viable harvester and defining its operating parameters. The goal of this project was to develop economically feasible mechanical harvesting for California black ripe Manzanillo table olives, the most widely used cultivar in California.
Specific aims of the study For commercial use of mechanical harvesting technology, the final processed fruit quality needs to be guaranteed first. Sensory evaluation was carried out with trained panels and consumers to understand: 1) the sensory characteristics of table olives harvested and processed in different ways 2) their relation to consumer preferences
An update on the findings and accomplishments from our collaboration team (years 2007-2011) Up to date Fruit damage has largely been eliminated. Both canopy contact and trunk shaking technologies have been demonstrated to be effective fruit removal technologies. The trunk shakers can be used in high-density orchards. The canopy harvester can be used in the new high-density orchards and in existing orchards if trees are pruned into hedgerows. However, in practice, both harvesting technologies are limited by tree canopy shape and by their respective platforms. Canopy contact and trunk shaking technologies must be improved through engineering, tested on effective platforms and evaluated on properly trained trees.
Outline I. Design of the study - Experimental design and sampling plan II. Methods - Descriptive Analysis Profiling the sensory characteristics of the olives III. Methods - Consumer Test Measuring the acceptability of the olives among consumers IV. Results - Canopy contact head harvester (yr. 08-09 and yr. 10-11) V. Results Trunk shaking harvester (yr. 09-10) VI. Identification of drivers of likes/dislikes for US consumers VII. Conclusions of the research
Canopy contact shaking head harvester, Central Valley, California 2007
Trunk shaking harvester, Central Valley, California
Materials (Experimental design and sampling plan) Row #1 Row #2 Row #3 Row #4 Row #5 Row #6 *yr 08-09 and yr 10-11: canopy contact harvester; 6 field replications *yr 09-10: trunk shaker; 4 field replications. Each row was randomly assigned to either hand harvest or machine harvest Half were treated at Processor A and the other half at Processor B Harvesting method x Processor Half were processed fresh and half after being held in storage tanks ( stored ) x Processing method Commercial olives from each Processor 8 differently treated olives/ row *Olives from different rows (#1-6) which went through the same treatment were pooled together 2 commercial olives *The olives were standard products to be compared against experimental ones
Table 1. Ten olive samples used in the study Sample abbreviation Processor Commercial Harvesting method A_Comm A Commercial - Processing method A_Hand_F - Hand Fresh olives A_Hand_S Stored olives A_Mech_F Machine* Fresh olives A_Mech_S B_Comm B Commercial - Stored olives B_Hand_F - Hand Fresh olives B_Hand_S Stored olives B_Mech_F Machine* Fresh olives B_Mech_S *yr 08-09 and yr 10-11: canopy contact harvester yr 09-10: trunk shaker Stored olives
Methods: Descriptive Analysis Panel: 8 trained panelists for each year (from the University of California, Davis) Training sessions - Developed descriptors (30-35 descriptors, depending on the year) - Selected references - Concept alignment
Qualitative Reference samples Earthy = Soil + olives Ocean-like = Green seaweed + anchovy + olives
Then trying the sample olives to detect the earthy/soil-like characteristic in the sample. Trying the qualitative reference for earthy/soillike in order to understand the concept of it in the black olives.
Appearance Texture/ Mouthfeel Smell (Aroma) Taste/ Flavor Methods: Descriptive Analysis Table 2. Descriptors for olives (yr 08-09, canopy contact harvester) Attribute Reference Attribute Reference Painty Correction fluid Sweetness Sucrose solution Briny Black olive brine Saltiness NaCl solution Ocean-like Green seaweed + anchovy* Umami MSG + brine Fermented Sauerkraut Bitterness Caffeine solution Canny Keys, cans Roasted Roasted sunflower seeds Earthy Potting soil* Buttery Melted butter + brine* Sautéed Mushroom Sautéed Mushroom* Ripeness Unripe ---- Ripe Dried Fruit Dried Prune Firmness Floral Chrysanthemum tea Juicy/ Moist release Size Small ---- Large Crumbly/ Crunchy Oval Round ---- Oval Fibrous Surface roughness Smooth ---- Rough Mouth coating Glossy Dull ---- Glossy Briny after-taste Skin brownness Black ---- Brown Lasting flavor Flesh Brownness Black ---- Brown Astringent Flesh greenness Black ---- Green * Mixed with olives
Methods: Descriptive Analysis Individual Evaluation - Rated intensities of sensory attributes for each sample in triplicate - Randomized order of presentation - Line scale - Serving: 3 olives/ sample (i.e. 2 whole + 1 olive sliced in half) @ room temp. - Proper cleansers were provided
Methods: Consumer test Consumer criteria - Age over 18 years - American - Black olives users and likers N=104~109 (depending on the year) Tasting Questions: - Overall liking - Appearance liking - Flavor (Taste & Smell) liking - Texture liking Exit survey - Socio-economics & demographics - Usage
Methods: Consumer test Tasting Procedure - 9-point hedonic scale - Serving: 2 whole olives/ sample @ room temperature - Randomized order of presentation - Proper cleansers were provided
Results First, the canopy contact head harvesters.
Descriptive analysis results (yr. 08-09 using canopy contact harvester) PC 2 (37.72 %) 1.5 1 Texture & Flavor: Firmness, 0.5 Fibrous, Dried Fruit, Bitterness 0-0.5 BComm BMechS BHandS Appearance: Brightness of flesh Brightness of skin (brown) AComm BMechF BHandF Flavor, After-taste, Mouthfeel: Salty, Umami, Buttery, Flavor lasting, Briny aftertaste, Juicy, Mouth-coating AHandF -1 Texture & Flavor: Crumbly, Fermented Ocean-like, -1.5 Sautéed mushroom AHandS AMechS Appearance: Glossy AMechF -2-2 -1.5-1 -0.5 0 0.5 1 1.5 2 2.5 Fig. 1 Sensory map of 10 olives (yr. 08-09 using canopy harvester) PC 1 (42.56 %) The biggest difference was seen between processing methods - fresh processed vs. others (commercial & stored), primarily along PC 1. A processor difference was observed as the next biggest factor, primarily along PC 2 (37.7%).
Descriptive analysis results (yr. 08-09 using canopy contact harvester) (cont d) PC 2 (37.72 %) 1.5 1 Texture & Flavor: Firmness, 0.5 Fibrous, Dried Fruit, Bitterness 0-0.5 BComm BMechS BHandS Appearance: Brightness of flesh Brightness of skin (brown) AComm BMechF BHandF Flavor, After-taste, Mouthfeel: Salty, Umami, Buttery, Flavor lasting, Briny aftertaste, Juicy, Mouth-coating AHandF -1 Texture & Flavor: Crumbly, Fermented Ocean-like, -1.5 Sautéed mushroom AHandS AMechS Appearance: Glossy AMechF -2-2 -1.5-1 -0.5 0 0.5 1 1.5 2 2.5 Fig. 1 Sensory map of 10 olives (yr. 08-09 using canopy harvester) PC 1 (42.56 %) The difference between harvesting methods was minimal!
Descriptive analysis results (yr. 10-11 using canopy contact harvester) PC 2 (21.24 %) 6 Texture: Firm, Crunchy, Chewy 4 Flavor, After-taste/mouthfeel : Bitterness, 2 Flavor lasting, Astringency 0 Flavor, Appearance: Earthy, Paint-like, Rancid, -2 Metallic, Fermented, Briny, Green/grassy Gradation in flesh, Glossy -4 BComm BHandS BMechS AHandS AMechS Flavor: Overall flavor, Floral, Salty, Sautéed mushroom AHandF AMechF BHandF BMechF Appearance, Flavor, Mouthfeel: Size, Green flesh, Inner roughness, Discoloration, Umami, Sweet, Sugary, Buttery, Ocean-like, Mouth-coating -6 AComm -8-9 -7-5 -3-1 1 3 5 7 9 Fig. 2 Sensory map of 10 olives (yr. 10-11 using canopy harvester) PC 1 (54.25 %) The main difference among products was observed between processing method (fresh processed vs. non-fresh [i.e. stored and commercial]), as shown by PC1 (54.3%).
Descriptive analysis results (yr. 10-11 using canopy contact harvester) (cont d) PC 2 (21.24 %) 6 Texture: Firm, Crunchy, Chewy 4 Flavor, After-taste/mouthfeel : Bitterness, 2 Flavor lasting, Astringency 0 Flavor, Appearance: Earthy, Paint-like, Rancid, -2 Metallic, Fermented, Briny, Green/grassy Gradation in flesh, Glossy -4 BComm BHandS BMechS AHandS AMechS Flavor: Overall flavor, Floral, Salty, Sautéed mushroom AHandF AMechF BHandF BMechF Appearance, Flavor, Mouthfeel: Size, Green flesh, Inner roughness, Discoloration, Umami, Sweet, Sugary, Buttery, Ocean-like, Mouth-coating -6 AComm -8-9 -7-5 -3-1 1 3 5 7 9 Fig. 2 Sensory map of 10 olives (yr. 10-11 using canopy harvester) PC 1 (54.25 %) There was little difference between harvesting methods (Hand vs. Machine).
Descriptive Analysis Results using canopy contact harvester (yrs 08-09 and 10-11) The biggest difference was seen between processing methods - fresh processed vs. others (commercial & stored). The difference between harvesting methods was minimal!
Overall degree of liking: On average (yr. 08-09 using canopy contact harvester) Mean (N=100) liking scores 7 5 3 1 Processing method (Fresh vs. Stored) Commercial vs. others Harvesting method (Hand vs. Machine) Fig. 3 Mean (N=100) overall liking scores (right); and partitioning of sources of variations for overall liking scores (left) (yr. 08-09 using canopy harvester) On average, fresh processed black olives were liked the MOST; and the commercial products were liked the LEAST. The two significant (P<0.05) factors that had driven the differences in the liking scores were 1) commercial vs. non-commercial products and 2) processing methods (fresh vs. stored) Harvesting method was NOT a significant source of variation and there was no difference in liking between hand- and machine-harvested products.
Overall degree of liking: On average (yr. 10-11 using canopy contact harvester) Mean (N=109) liking scores 7 5 3 Processing method (Fresh vs. Stored) 1 Fig. 4 Mean (N=109) overall liking scores (right); and partitioning of sources of variations for overall liking scores (left) (yr. 10-11 using canopy harvester) On average, fresh processed black olives were liked the MOST. There was a significant (P<0.05) difference in consumer acceptance between processing method (fresh vs. stored). There was NO significant difference between mechanically- and hand-harvested olives.
Yr. 08-09 Yr. 10-11 Flavor liking Appearance liking Texture liking Overall liking 0.993 0.436 0.717 Values in bold are significantly different from 0 with a significance level alpha=0.05 Flavor liking Appearance liking Texture liking Overall liking 0.913 0.148 0.714 Values in bold are significantly different from 0 with a significance level alpha=0.05 overall flavor appearance texture overall flavor appearance texture Fig.5 Pearson s correlation coefficients of liking for specific attributes to overall liking (top); and Means for hedonic ratings of the 10 olive samples for overall degree of liking and liking for appearance, flavor and texture (bottom) Flavor liking had the highest correlation with overall degree of liking and appearance liking was the least correlated.
Overall degree of liking: Preference Mapping (yr. 08-09 using canopy contact harvester) F2 (14.55 %) F2 (14.55 %) 1 Most of the consumers are in this area. 8 6 Acomm 4 AMechF 0 2 Bcomm AHandF -1 Few consumers in this area F1 (27.37 %) -1 0 1 0-2 -4-6 BHandS BMechS AMechS AHandS BMechF BHandF F1 (27.37 %) -8-6 -4-2 0 2 4 6 8 Fig. 6 - Internal preference map of 100 individual consumers based on overall degree for liking for 10 olives (yr. 08-09 using canopy harvester) *Each vector (on the left graph) represents individual consumers The 100 consumers were homogeneous in their liking for black olive. The majority preferred fresh processed black olives ; whereas the commercial products were liked the least. Harvesting method was NOT a significant source differentiating one s liking.
F2 (15.37%) F2 (15.37 %) Overall degree of liking: Preference Mapping (yr. 10-11 using canopy contact harvester) 1 15 Cluster 3 liked olives in this area BHandS 5 AComm BMechS 0-1 0 1-5 AMechS BComm AHandS BMechF AHandF BHandF -1 Cluster1 Cluster2 Cluster3 F1 (18.15%) AMechF F1 (18.15 %) -15 Cluster 2 liked -20-10 0 10 20 olives in this area Fig. 7 - Internal preference map of 109 individual consumers based on overall degree for liking for 10 olives (yr. 10-11 using canopy harvester) and indication of group information *Each point (on the left graph) represents an individual consumer Three consumer segments which differ in preference for olives were identified: Cluster 1 (n=14) : Hard to generalize a trend based on this map. Cluster 2 (n=57, majority): Fresh-processed olives Cluster 3 (n=38): Commercial ones and fresh-processed olives from processor B
Consumer test Results using canopy contact harvester (yr 08-09 and 10-11) The majority of American consumers liked olives that were fresh processed. There was no significant difference in overall acceptability between harvesting methods (Hand vs. Machine). Flavor liking had the highest correlation with overall degree of liking and appearance liking was the least correlated.
Results And now to the Trunk shaking harvester.
Descriptive analysis results (yr. 09-10 using trunk shaker) PC 2 (28.92 %) 4 Flavor: Bitterness, Sauteed mushroom 3 2 AHandF BHandF Texture: Firmenss, Crunchy, Fibrous 1 Flavor, Mouthfeel, Appearance & 0 After-taste Umami, Salty Oily, Juicy -1 Glossy, Lasting flavor, Briny AF -2 BHandS BMechS BMechF AMechF BComm AComm -3-4 AHandS AMechS -5-5 -4-3 -2-1 0 1 2 3 4 5 6 7 Fig. 8 Sensory map of 10 olives (yr. 09-10 using trunk shaker) Flavor & Appearance: Ocean-like, Metallic, Earthy, Fermented, Green/Grassy Surface roughness, Brownness PC 1 (39.60 %) The biggest difference was seen between commercial vs. others, primarily along PC 1. The next largest difference was observed between Hand harvested x Fresh processed vs. Store processed in processer A, primarily along PC 2. The difference between harvesting methods was not large, but it was significant as can be seen in the left quadrant.
Mean (N=109) liking scores Overall degree of liking: On average (yr. 09-10 using trunk shaker) 7 5 Harvesting * Processing Processors (A vs. B) Harvesting method (Hand vs. Machine) 3 Processing method (Fresh vs. Stored) 1 Fig. 9 Mean (N=109) overall liking scores (right); and partitioning of sources of variations for overall liking scores (left) (yr. 09-10 using trunk shaker) On average, fresh processed black olives were liked the MOST; the Machine harvested, then processed after storage from processor A product was liked the LEAST. The significant (P<0.05) factors that drove differences in liking were 1) Processing method (fresh vs. stored), 2) Processor (A vs. B), 3) Harvesting method (Machine vs. Hand) and 4) interaction of Harvesting method by Processing method.
Mean (N=109) liking scores Overall degree of liking: On average (yr. 09-10 using trunk shaker) (Cont d) 7 5 Harvesting * Processing Processors (A vs. B) Harvesting method (Hand vs. Machine) 3 Processing method (Fresh vs. Stored) 1 Fig. 9 Mean (N=109) overall liking scores (right); and partitioning of sources of variations for overall liking scores (left) (yr. 09-10 using trunk shaker) Harvesting method was not the largest factor, but there was a noticeable decrease in overall liking of Machine harvested olives compared to Hand harvested ones. This effect was mainly due to the Store processed in Processer A samples. Other treatments showed some decrease, but not a significant one. This finding can be attributed to the over-ripe nature of the fruit harvested that year.
Yr. 09-10 Flavor liking Appearance liking Texture liking Overall liking 0.990 0.916 0.900 Values in bold are significantly different from 0 with a significance level alpha=0.05 overall flavor appearance texture Fig.10 Pearson s correlation coefficients of liking for specific attributes to overall liking (top); and Means for hedonic ratings of the 10 olive samples for overall degree of liking and liking for appearance, flavor and texture (bottom) Flavor liking had the highest correlation with overall degree of liking and appearance liking was the least correlated. For this particular year, appearance and texture likings were also found to be highly correlated to overall liking.
Overall degree of liking: Preference Mapping (yr. 09-10 using trunk shaker) F2 (12.80 %) F2 (12.80 %) 1 0.75 Most of the consumers are in this area. 15 10 BMechS 0.5 0.25 0-0.25-0.5 Few consumers -0.75 in this area F1 (30.46 %) -1-1 -0.75-0.5-0.25 0 0.25 0.5 0.75 1 AMechS AHandS BHandS AComm BMechF BComm BHandF AHandF AMechF -15-25 -20-15 -10-5 0 5 10 15 20 Fig. 11 - Internal preference map of 109 individual consumers based on overall degree for liking for 10 olives (yr. 08-09 using canopy harvester) *Each vector (on the left graph) represents individual consumers The 109 consumers were homogeneous in their liking for black olive. The majority preferred fresh processed black olives ; whereas the olives that were Machine harvested, then store-processed were liked the least. Harvesting method and Processing method interaction was found to significant in influencing one s liking. 5 0-5 -10 F1 (30.46 %)
Descriptive analysis and Consumer test Results for the Trunk shaking harvester (yr 09-10) In contrast to the canopy contact head harvester, there were some noticeable differences between hand- and mechanicallyharvested olives both in sensory qualities and consumer acceptance. The trunk shaking technology for over-ripe fruits may need further refinement before its application is satisfactory. However, we were again able to confirm that fresh-processed olives have strong potential in the Californian olive market. This year s study also confirmed that flavor liking had the highest correlation with overall degree of liking.
Identification of drivers of likes/dislikes for Californian consumers
Drivers of liking: yr. 08-09 using canopy contact harvester) Fig. 12 - PLS-Regression of the consumer hedonic ratings onto the sensory attributes from the descriptive analysis (yr. 08-09; canopy contact harvester) Overall liking was highly related to flavor liking; and some extent to texture liking; however, appearance liking was not a significant source in this regression model. - Drivers of flavor likes: Sweet, Buttery, Roasted, Saltiness, Umami, Ripeness, Briny after-taste, Lasting flavor - Drivers of flavor dislikes: Dried fruit, Fermented
Drivers of liking: yr. 09-10 using trunk shaker Fig. 13 - PLS-Regression of the consumer hedonic ratings onto the sensory attributes from the descriptive analysis (yr. 09-10; trunk shaker) Overall liking was highly related to flavor, appearance, and texture liking. - Drivers of flavor likes: Buttery, Sautéed Mushroom - Drivers of flavor dislikes: Fermented fruit, Metallic, Green/Grassy, Earthy - Texture: Balance of firmness, crunchiness, fibrousness & Juicy (firmer preferred) - Driver of appearance dislike: Surface roughness
Drivers of liking: yr. 10-11 using canopy contact harvester) Fig. 14 - PLS-Regression of the consumer hedonic ratings onto the sensory attributes from the descriptive analysis (yr. 10-11; canopy contact harvester) Three clusters had different drivers of likes of dislikes - Cluster 1 (N=14): Not well defined in this regression model - Cluster 2 (N=57): Likes - Buttery, Sweet, Umami, Ocean-like, Sugary, Sautéed mushroom, Floral; Dislikes - Fermented, Green/grassy, Bitter, Earthy/musty, Rancid, Metallic, Briny/Salty - Cluster 3 (N=38) : Disliked firm, chewy, crunchy textures
Identification of driver of likes and dislikes: 3 yr summary (yr 08-09; 09-10; 10-11) A shift in consumer perceptions and preferences was observed. Particularly in the last year, consumer preferences did shift toward a more specific direction. This could be due to their increased familiarity with imported olives. In general, Buttery can be considered an important characteristic that drives consumer preferences; whereas Fermented fruit can be regarded as a characteristic that drives consumers to dislike the product.
Conclusions No significant difference in the sensory profiles nor in consumer acceptance of handand mechanically-harvested olives using the canopy-contact mechanical harvester. Canopy contact head mechanical harvesting can provide a cheaper alternative to traditional hand harvesting of black ripe table olives.
Conclusions The trunk shaking technology for over-ripe fruit may need further refinement before its application yields satisfactory quality. Our research showed strong potential for freshprocessed olives on the California olive market. In general, Buttery has been identified to be an important characteristics that drivers consumer preference; whereas Fermented fruit was found to be an important characteristic that drives consumers to dislike a product.
Acknowledgements US Department of Agriculture California Olive Committee Bell Carter Musco Family Olives Descriptive analysis panelists Consumers Thank you for your attention!