Effects of Storage Conditions on Consumer and Chemical Assessments of Raw ‘Nonpareil’ Almonds Over a Two‐Year Period (2024)

Study design

The study consisted of an incomplete factorial design in which almonds were stored under varying conditions over a 24‐month period and tested by chemical means and by consumer sensory evaluation. Almonds were stored in either unlined cardboard cartons (UC) or in sealed polypropylene bags (PPB). Storage temperatures were 15, 25, and 35 °C at RH levels of 50% or 65%. Samples were also held at 4 °C without RH control. This resulted in 13 different conditions as outlined in Table 1.

Figure ​Figure11 depicts a flowchart for the various sample analyses, and how these were triggered at various times during the storage period. Prior to packaging and storage, almonds were evaluated at baseline by a consumer sensory panel (n = 118) as well as by chemical means. The chemical tests included peroxide value (PV), free fatty acid (FFA), conjugated dienoic acid (CD), 2‐thiobarbituric acid‐reactive substances (TBARS), water activity (A_w), and moisture content (MC). Results of these baseline assessments served as a basis for comparison throughout storage.

Every 2 calendar months over a 24‐month period, samples from all treatment groups were assessed for PVs. When predetermined thresholds were reached for a sample set (PV > 2.0 meq active O₂/kg oil), consumer panels evaluated these samples. In addition, 3 experienced sensory analysts evaluated the samples for indicators of quality degradation following ASTM (2011a) guidelines. This was done to ensure that any suspected deterioration in the samples was further evaluated, even if the PVs were below 2.0 meq active O₂/kg oil. This procedure was followed to ensure that consumer sensory evaluations were conducted at appropriate times in the rejection period.

When threshold values were reached, the almonds were evaluated by a consumer screening panel (n = 35 to 40), and the panelists were asked about the acceptability of the sample as well as “If you had purchased this product would you eat it?” (Hough etal., 2003). If at least 75% of the panelists stated that they would eat the sample, the sample continued to be evaluated by a consumer panel at intervals of 2 calendar months until less than 75% of the panelists indicated they would consume the sample. If ≥25% of panelists stated that they would not consume the sample, these samples were then investigated with a larger confirmatory panel (n = 100 to 110) within 24 hours of the screening panel. If both screening and confirmatory panel rejection rate was ≥25%, the sample was deemed “failed,” and the data from the screening and confirmatory panel were pooled together for subsequent statistical analysis. At the points of their respective failures, samples were assessed by PV, FFA, CD, TBARS, A_w, and MC. Following the chemical assessments at the time of failure, samples were no longer submitted for further sensory or chemical evaluations. At 24 months, all remaining stored samples were evaluated, even if the preliminary criteria for indication of quality deterioration had not been met.

Almond characterization, handling, and packaging

The nuts investigated in this study were whole, raw, unsalted ‘Nonpareil’ almonds with brown skin. The almonds were a composite lot harvested from different orchards in California between September and October of 2014 and were graded “Supreme.” The almonds were pasteurized by propylene oxide fumigation prior to packaging. The nuts were shipped to the Dept. of Food Science & Technology, Athens, GA, U.S.A., in 50 lb cartons and the initial MC was determined to be 4.3%.

The raw almonds were repackaged at the Univ. of Georgia for storage in Uline S‐17960 (100 μm, clear polypropylene) bags or Uline S‐15138 corrugated cardboard boxes (Uline, Waukegan, IL, U.S.A.). The polypropylene material had a water vapor transmission rate of 8 g/m²d and oxygen transmission rate of 860 cm³/m²d, and the UC material provided no protection from atmospheric conditions. Each PPB was filled with 300 ± 5 g of raw almonds and then placed in a Model 600 vacuum packaging system (Henkelman B.V., The Netherlands). The filled bags were subject to vacuum for 30 s, and then flushed with food‐grade N₂ prior to heat sealing, thereby forming a pillow pack for each sample. A total of 30 bags were filled per treatment. The initial oxygen level in the PPB was below 0.5%. For samples stored in cardboard boxes, 900 ± 5 g of raw almonds were placed in each box and then covered with a lid. A total of 12 boxes were used for each experimental treatment.

The almond samples receiving RH control were stored in Hotpack environmental chambers (Model 434304, SP Industries, Warminster, PA, U.S.A.). Some samples were also stored at 4 °C in a walk‐in cooler, but had no RH control. The temperature and RH of each chamber were monitored throughout the study.

Sensory participants and demographic information

All participants for the sensory panels were over 18 years of age with no peanut or tree‐nut allergies. Ethnicity and sex were not controlled factors. Participants were recruited from faculty, staff, students, and visitors of the Univ. of Georgia. The criterion for inclusion was that the prospective panelist consumed nuts or nut products at least once per month. All participants were required to provide informed consent prior to participation. The demographics differed slightly for each panel, but across all panels, approximately 78% of the panelists were female with 73% between the ages of 18 and 27.

Consumer sensory evaluation

The sensory evaluation protocol followed ASTM methodology (ASTM, 2011b). For all evaluations, sample cups were coded with random 3‐digit codes and served with 3 almonds per sample cup. The panelists were seated in individual sensory booths equipped with white lighting. The samples were presented one at a time in a counterbalanced order of presentation. Water and baby carrots were provided as palate cleansers.

For all consumer evaluations, the panelists evaluated the almond samples for acceptability of odor, texture, and flavor as well as overall acceptability by rating the samples on a 9‐point hedonic scale. The scale ranged from “extremely dislike” (1) to “extremely like” (9). The panelists were also asked to indicate the favorable and unfavorable traits of each sample (Rousset & Martin, 2001), by responding to the question “Please indicate WHAT in particular you liked or disliked about this almond sample (use WORDS not SENTENCES).” Finally, the panelists were asked to respond to “If you had purchased this product, would you eat it? (yes or no)” and a negative response to this question was defined, for our study, as equivalent to “rejection” of the sample (Hough etal., 2003).

Chemical analyses

All chemical assessments were performed in triplicate. PVs were determined for all samples at intervals of 2 calendar months throughout the study according to AOAC Method 965.33 (AOAC, 2012). FFA, CD, TBARS, A_w, and MC were determined for the samples at baseline and for samples following their “failure” (according to consumer assessment). Samples that did not fail during the study were assessed at the conclusion of the 24‐month period.

FFAs were determined according to AOCS Method Ca 5a‐4020 (AOCS, 2009). CDs were determined according to IUPAC Official Method 2.505 (IUPAC, 1992). TBARS were determined according to AOCS Official Method Cd 19–90 (AOCS, 2009). A_w was determined by loading 2 g (±0.1 g) of the ground almond meal into a calibrated Aqua Lab CX‐2 water activity meter (Pullman, WA, U.S.A.). MC (percentage value of moisture mass within total sample mass) was determined for ground almonds by assessment of mass loss following heating in a forced‐air convection oven at 105 °C until constant weight was achieved (AOAC, 2012).

Statistical analysis

All statistical analyses were performed with SAS 9.4 Statistical Analysis Software (SAS Inst. Inc., Cary, NC, U.S.A.). Normal distribution was verified through univariate analysis. Descriptive statistics were determined for assessments and t‐tests were conducted to test significant changes over time when compared to baseline. Analysis of variance (ANOVA) with Student–Newman–Keuls (SNK) was conducted to detect significant differences between samples at the point of rejection.

Multiple regression analyses were conducted to model acceptability of individual sensory attribute (odor, texture, and flavor) as predictors of overall acceptability. Multiple regression models were also developed for the prediction of the time of failure according to storage parameters and these modeling procedures excluded samples that did not fail during the duration of the study. Additional models were made for the prediction of each of the 4 variables representing consumer acceptability (odor, flavor, texture, and overall) according to the assessed chemical measures, utilizing the data assessed at baseline and at sample failure (or study conclusion) for each of the samples.

The multiple regression models used in this study were developed using the “Best Subsets” procedure. In essence, 8 models were evaluated using combinations of the predictors (temperature, RH, or packaging). Of these, the best fit from each of the 1‐term, 2‐term, or 3‐term models was selected based on the R² value. From these models, the one with the greatest adjusted R² value was selected. The adjusted R² is related to the mean square error, and thus penalizes the model for adding additional terms. For all statistical assessments, the level of statistical significance was defined at α = 0.05.

Assessment of raw almonds at baseline

The baseline assessments (day 0) of PVs (<0.01 meq active O₂/kg oil), FFA (0.28 acid value), CD (1.43), TBARS (0.030), A_w (0.29), and MC (3.1%) of the raw almonds suggested the almonds met industry‐standard criteria for freshness of raw almonds (Almond Board of California, 2015). Specifically, the PVs, FFAs, and MCs were all below the recommended maximum values for verification of freshness (5 meq active O₂/kg oil, 3.0 acid value, and 3.5% to 6% MC, respectively). The FFA and PV levels were also below those reported for raw almonds prior to lipid deterioration (Lin etal., 2012; Mexis & Kontominas, 2010). In addition, the A_w of the almonds fell within the 0.25 to 0.35 range at which lipid oxidation is typically lowest (Huang, 2014).

The baseline consumer sensory assessments of odor, texture, flavor, and overall acceptability of the raw almonds are shown in Table 2. Mean texture, flavor, and overall acceptability were above 7 on the 9‐point hedonic scale, while mean acceptability of odor was 5.7. It is important to note that consumers were instructed to evaluate the odor of the sample prior to masticating the almonds; thus, it is probable that fewer volatile compounds were detected than would have been during mastication. In addition, as the almonds were raw, it may be that some consumers were expecting odors more characteristic of roasted almonds. Mean texture, flavor, and overall acceptability at baseline can reasonably be interpreted as indicating the product was free of substantial defects or detriments.

At baseline, roughly 6% of the consumer panelists rejected the raw almond sample. Considering the freshness of the almonds, one might expect that no panelists would reject the samples. However, previous studies have found that very fresh almonds can result in a perception of unacceptable quality by some consumers (Hough etal., 2003). It has been postulated that some consumers may be more accustomed to eating older/aged almonds (such as those found in the bulk bins at grocery stores), therefore making the absence of aged and oxidized flavor notes an unexpected, and perhaps undesirable, characteristic.

At baseline, roughly ½ of the panelists described the almonds as “crunchy,” using this term to describe something they liked about the sample. Most panelists also stated that they liked the overall texture and flavor/taste of the sample, with the most commonly used words to describe the flavor being “nutty” and “sweet.” The odor was not listed as a quality that was liked by many panelists. The most common specific complaint regarding odor was that the sample had a “weak” or “mild” odor. The most common complaint about flavor for the sample at baseline was that the sample was “bland” or “mild,” while the most common textural complaint was that it was “hard” and “dry.” These quality concerns for flavor and odor are consistent with the hypothesis that panelists may be responding to the absence of roasted or positive oxidized flavor notes.

Consumer sensory evaluation of almonds stored in polypropylene bags

Table 2 summarizes the consumer sensory results of all samples at their final assessment and includes statistical comparisons with one another and with baseline. Of the 7 samples stored in PPB, consumer panelists rejected 4 of the samples prior to the conclusion of the study. All samples stored at 65% RH were rejected, with those stored at 35 °C rejected at 6 months, those at 25 °C at 16 months, and those at 15 °C at 24 months. In addition, the sample stored at 35 °C and 50% RH was rejected at 12 months. The 2 samples stored at 35 °C were the first to be rejected from the study, suggesting a substantial influence of temperature on consumer acceptability and rejection. Although all samples were “triggered” for consumer evaluation within 12 months, 3 of the samples did not fail within the timeframe of the study (that is, those at 4 °C, 15 °C/50% RH, and 25 °C/50% RH).

The sample most similar to baseline at the end of the study was that stored in PPB at 15 °C/50% RH. The scores for odor and overall acceptability for this sample were not significantly different from baseline when the study concluded at 24 months, and although the flavor was significantly lower when compared to baseline, it had the highest score of all samples. This indicates that the quality of raw almonds is best maintained when stored in PPB, which provide a barrier to both oxygen and water vapor transmission, and when stored at low temperature and humidity. Although at lower temperature, almonds stored at 4 °C had lower acceptability than those at 15 °C. This is likely due to the uncontrolled RH in the walk‐in cooler, which could reach up to 90% humidity at times during the study.

In general, acceptability scores for odor were not significantly different from baseline values throughout the 24‐month storage period. In contrast, significant differences in texture, flavor, and overall acceptability (against their respective baseline values) were observed for all rejected samples at their points of failure.

Consumer sensory evaluation for almonds stored in unlined cartons

Consumer panelists rejected all samples stored in UC at some point during the 24‐month storage period (Table 2). In order, those at 35 °C/65% RH were rejected at 2 months, those at 35 °C/50% RH at 6 months, those at 4 °C at 6 months, those at 25 °C/65% RH at 12 months, and those at 25 °C/50% RH or 15 °C/50% RH at 16 months. Mean panelist responses for the acceptability attributes for these samples were consistently lower than the scores for their counterparts stored in PPB.

Samples stored in UC at 4 °C exhibited the greatest quality deterioration of all samples. These had significantly lower scores for odor, texture, and overall acceptability than did all other samples. However, the flavor acceptability score for this sample differed significantly only from 2 samples stored in PPB (4 °C and 15 °C/50% RH). No differences between this sample and the remaining samples stored in UC were found.

When comparing acceptability scores for all attributes against those at baseline, significant differences were seen for more than half of the sensory results. This highlights that almonds stored in UC demonstrated a greater reduction in sensory quality than those stored in PPB. Analysis of the strong and weak points stated by panelists for the samples at the point of failure showed that samples stored in UC were more frequently described as having a “cardboard” flavor or odor, especially among individuals who rejected the samples. Although “cardboard” is a term that is associated with lipid oxidation, it is feasible that these almonds developed detrimental cardboard flavor notes via direct interaction with the cardboard boxes.

Effects of storage parameters on consumer assessments

Table 3 shows the multiple regression model for the prediction of time (in months) until sample failure, according to storage parameters. The model has a fairly strong predictive strength (adjusted R² = 78.9%) and depicts quantitatively the effects of the assessed storage parameters on raw almond stability. The model predicts that storage of raw almonds in PPB (rather than UC) would prolong consumer acceptability by 7.21 months. This finding generally agrees with an examination of the results, as for each storage condition, every sample packaged in UC failed prior to those in PPB stored under equivalent conditions. This relative protective effect of PPB is expected, as UC provide minimal protection against moisture or oxygen transfer and may also lead to more rapid quality deterioration due to the migration of flavor‐imparting compounds from the packaging to the product (Guinard & Mazzucchelli, 1996).

The model in Table 3 also shows a negative effect of T and RH on sample stability. García‐Pascual, Mateos, and Salazar (2003) reported that there was a significant increase in shelf life of roasted almonds by decreasing the storage temperature from 36 to 8 °C. The authors suggested that there is a protective effect of temperature reduction when samples are exposed to high humidity. Specifically, our model predicts the time until failure is reduced by 0.589 months for each additional °C of storage T (within the assessed range). The model also predicts time until failure is decreased by 0.418 months with each 1% increase in RH. The effects suggested by the model are compatible with an examination of the results, as storage at 65% RH resulted in reduced/shorter shelf life than storage at 50% RH in both packaging materials. At 25 and 35 °C, this 10% reduction in RH was associated with measured extensions in the shelf life of 4 and 6 months, respectively.

Table 4 shows statistical comparisons between samples stored in PPB versus those in UC (but otherwise identical conditions) for the acceptability scores at the time of failure/end of study. At the lower storage temperatures (4 and 15 °C), acceptability scores of all attributes differed significantly between the samples stored in PPB versus UC, whereas for those stored at 35 °C, only texture differed significantly. Specifically, texture acceptability was significantly higher for almonds stored in PPB at 35 °C (6.3 ± 1.9 for 50% RH and 6.1 ± 1.9 for 65% RH) than those stored in UC at 35 °C (5.7 ± 1.9 for 50% RH and 5.5 ± 2.2 for 65% RH). This suggests that the chemically mediated attributes of flavor and odor are more susceptible to changes in temperature. Thus, limiting moisture and oxygen transfer through polypropylene packaging may not effectively counter the increase in chemical reactions due to temperature. However, texture attributes are more susceptible to moisture absorption, as water softens the structure and decreases the fracturability associated with firm, crunchy nuts.

Specific hedonic measures as predictors of overall consumer acceptability

Multiple regression analysis was used to identify the contribution of each attribute as a predictor of overall acceptability for each of the sensory panels (Table 5). The table presents results for each time and set of conditions in which a sensory panel was conducted. At baseline, all 3 sensory attributes were found to be predictors of overall acceptability (R² = 0.80), with contributions of flavor, texture, and odor to overall acceptability being 65.7%, 13.1%, and 1.4%, respectively.

For 65 of the 69 models, the flavor was the largest predictor of overall acceptability, ranging from 53% to 89% (71% to 85% for samples at the point of failure). Texture was the largest predictor for the 4 remaining instances (PPB at 4 °C at 12 months, UC at 4 °C at 2 and 6 months, and PPB at 15 °C/50% RH at 20 months); it is worth noting that for each of these, the models were based upon screening panels, and thus had smaller sample sizes than the confirmatory panels. The odor was the weakest predictor of overall acceptability for every model produced, and it was omitted from the model in more than half of the cases due to lack of significance. In no cases did the odor variable contribute more than 5.1% predictive strength to the model.

The importance of individual attributes to overall acceptability was also analyzed using multiple regression models of pooled consumer data (Table 6). Among these are models for 3 divisions of the data—one using consumer data for all panelists and samples, one using consumer data for panelists that rejected the sample, and one using the data from panelists that accepted the sample. Table 6 lists the parameter estimates for the intercepts as well as for each of the individual sensory attributes and indicates the partial R² for each sensory attribute and the cumulative R². Odor, texture, and flavor were all significant factors in predicting overall acceptability. The model created for all sensory data explained the most variability (87.3%), while that created based on rejected samples explained the least (75.1%). In all 3 models, flavor was the largest determinant of overall acceptability and odor was the smallest. For example, considering all data partial R² values were 0.774, 0.093, and 0.0055 for flavor, texture, and odor, respectively. For panelists who rejected the samples, texture and odor were more important, although flavor was still a dominant factor. Thus, the partial R² values were 0.564, 0.168, and 0.0183 for flavor, texture, and odor, respectively.

The observed differences in the models for panelists who accepted versus those who rejected the samples suggest consumers may use different criteria when evaluating “intent to consume.” Specifically, odor and texture appear to be more important determinants for panelists who reject samples than panelists who do not, indicating that undesirable odors and textures played a larger role than desirable ones in predicting consumer rejection. On the other hand, flavor played a greater role in predicting consumer acceptability.

Chemical measures as predictors of consumer assessment

The results of the chemical assessments for the samples at the time of failure (or at 24 months in the absence of failure) are presented in Table 7. As expected, chemical measures of deterioration increased for the samples during storage. An examination of the average values within the 2 packaging types shows the samples in UC, in sum, to have significantly higher A_w at time of failure (or conclusion of study; P = 0.021) than those in PPB. This is reasonable, considering the lack of vapor protection provided by UC. Although the average value of MC is also higher for UC than for PPB, neither for this measure nor any of the other assessed chemical measures were the averages within package type significantly different from one another.

Table 8 shows the predictive models for consumer outputs of odor, flavor, texture, and overall acceptability according to assessed chemical factors. The correlation coefficients suggest decent predictive strengths for these models. Specifically, the model for consumer assessment of odor had the greatest predictive strength (adjusted R² = 71.1%) and the model for flavor assessment had the lowest (adjusted R² = 53.5%). The variable coefficients of the models are all within expectation. It is notable that PV and FFA both associate negatively with the assessment of flavor, odor, and overall acceptability, while the CD and TBARS variables failed to improve these models. This suggests that the chemical assessments of PV and FFA in raw almonds may be of particular utility for the evaluation of raw almond quality throughout storage. The models show that increases in PV were associated with a greater reduction in expected odor assessment (−0.696 reduction for each meq active O₂/kg oil) than for expected assessments of flavor or overall acceptability (−0.174 and −0.143 reductions for each meq active O₂/kg oil, respectively). Conversely, the modeled effects of FFA were of lower magnitude for odor than for flavor or overall acceptability.

The most predictive model for assessed texture quality contained exclusively the factors of A_w and MC. For both variables, the coefficients show negative associations with predicted texture quality, indicating deleterious effects of increases in total and available water in raw almonds. This provides further evidence of the possible usefulness of controlling storage humidity, and for storing raw almonds in packaging that mitigates water vapor transmission.

In addition to the noted observed effect of A_w on textural quality, A_w was also determined to be a useful variable in the prediction of overall acceptability and flavor acceptability. The effect on overall acceptability is unsurprising. The effects on flavor may be secondary. That is, it is known that the rates of lipid oxidation are dependent upon the A_w/MC of nut products. As the A_w rises above 0.2 to 0.3, lipid oxidation can occur more expediently as reactive molecules become more mobile. In addition, there may also be multimodal sensory integration by panelists, with the panelists’ assessments of flavor influenced directly by textural qualities (Forde & Delahunty, 2004).

• Albert, C. M., Gaziano, J. M., Willett, W. C., & Manson, J. E. (2002). Nut consumption and decreased risk of sudden cardiac death in the Physicians’ Health Study. Archives of Internal Medicine, 162, 1382–1387. [PubMed] [Google Scholar]
• Almond Board of California. (2015). Technical Information Kit‐2015. Retrieved from https://www.almonds.com/sites/default/files/content/attachments/abc_technical_kit_2015.pdf.
• Almond Board of California.(2016). Almond almanac 2016. Retrieved from https://www.almonds.com/sites/default/files/misc/Almanac/2016_almond_almanac.pdf.
• AOAC. (2012). Official methods of analysis of AOAC international (19th ed.). Gaithersburg, MD: AOAC Intl. [Google Scholar]
• AOCS. (2009). Official methods and recommended practices of the American Oil Chemists’ Society. Champaign, IL: AOCS. [Google Scholar]
• ASTM International. (2011a). Lexicon for sensory evaluation: Aroma, flavor, texture, and appearance. West Conshohocken, PA: ASTM Intl. [Google Scholar]
• ASTM International. (2011b). Standard guide for sensory evaluation methods to determine the sensory shelf life of consumer products. West Conshohocken, PA: ASTM Intl. [Google Scholar]
• Berryman, C. E., West, S. G., Fleming, J. A., Bordi, P. L., & Kris‐Etherton, P. M. (2015). Effects of daily almond consumption on cardiometabolic risk and abdominal adiposity in healthy adults with elevated LDL‐cholesterol: A randomized controlled trial. Journal of the American Heart Association, 4, e000993. [PMC free article] [PubMed] [Google Scholar]
• Forde, C. G., & Delahunty, C. M. (2004). Understanding the role cross‐modal sensory interactions play in food acceptability in younger and older consumers. Food Quality and Preference, 15, 715–727. [Google Scholar]
• García‐Pascual, P., Mateos, M., & Salazar, D. M. (2003). Influence of storage conditions on the quality of shelled and roasted almonds. Biosystems Engineering, 84, 201–209. [Google Scholar]
• Guinard, J.‐X., & Mazzucchelli, R. (1996). The sensory perception of texture and mouthfeel. Trends in Food Science & Technology, 7, 213–219. [Google Scholar]
• Harris, L. J., & Ferguson, L. (2013). Improving the safety of almonds and pistachios In Harris L. J. (Ed.), Improving the safety and quality of nuts (pp. 350–376). Cambridge, UK: Woodhead Publishing Ltd. [Google Scholar]
• Hough, G., Langohr, K., Gómez, G., & Curia, A. (2003). Survival analysis applied to sensory shelf life of foods. Journal of Food Science, 68, 359–362. [Google Scholar]
• Huang, G. (2014). Almond shelf life factors. Retrieved from https://www.almonds.com/sites/default/files/content/attachments/2014aq0007_shelf_life_factors.pdf.
• Hudson, B. J. F., & Gordon, M. H. (1999). Evaluation of oxidative rancidity techniques In Allen J. C. & Hamilton R. J. (Eds.), (3rd ed., pp. 53–67). Gaithersburg, MD: Aspen Publishers, Inc. [Google Scholar]
• International Nut and Dried Fruit Foundation. (2010). Nuts and dried fruits – Global statistical review 2004–2009. Rues, Spain.
• IUPAC. (1992). Standard methods for the analysis of oils, fats and derivatives. Oxford, UK: Blackwell Science. [Google Scholar]
• Jaceldo‐Siegl, K., Sabaté, J., Rajaram, S., & Fraser, G. E. (2004). Long‐term almond supplementation without advice on food replacement induces favorable nutrient modifications to the habitual diets of free‐living individuals. British Journal of Nutrition, 92, 533–540. [PubMed] [Google Scholar]
• Kader, A. A. (2013). Impact of nut postharvest handling, de‐shelling, drying and storage on quality. In Harris L. J. (Ed.), Improving the safety and quality of nuts (pp. 22–34). Cambridge, UK: Woodhead Publishing Ltd. [Google Scholar]
• Lin, X., Wu, J., Zhu, R., Chen, P., Huang, G., Li, Y., … Ruan, R. (2012). California almond shelf life: Lipid deterioration during storage. Journal of Food Science, 77, C583–C593. [PubMed] [Google Scholar]
• Martínez‐González, M. A., & Bes‐Rastrollo, M. (2010). Nut consumption, weight gain and obesity: Epidemiological evidence. Nutrition, Metabolism & Cardiovascular Diseases, 21, S40–S45. [PubMed] [Google Scholar]
• Mexis, S. F., & Kontominas, M. G. (2010). Effect of oxygen absorber, nitrogen flushing, packaging material oxygen transmission rate and storage conditions on quality retention of raw whole unpeeled almond kernels (Prunus dulcis). London Weekend Television, 43, 1–11. [Google Scholar]
• Rousset, S., & Martin, J.‐F. (2001). An effective hedonic analysis tool: Weak/strong points. Journal of Sensory Studies, 16, 643–661. [Google Scholar]
• Shahidi, F., & John, J. A. (2013). Oxidative rancidity in nuts In Harris L. J. (Ed.), Improving the safety and quality of nuts (pp. 198–229). Cambridge, UK: Woodhead Publishing Ltd. [Google Scholar]
• USDA. (2014). ERS 2014 Yearbook Tables. Retrieved from https://www.ers.usda.gov/data-products/fruit-and-tree-nut-data/yearbook-tables.aspx.
• Velasc, J., Dobarganes, C., & Márquez‐Ruiz, G. (2010). Oxidative rancidity in foods and food quality In Skibsted L. H., Risbo J., & Andersen M. L. (Eds.), Chemical deterioration and physical instability of food and beverages (pp. 3–31). Cambridge, UK: Woodhead Publishing Ltd. [Google Scholar]