Food or fad? Challenges and opportunities for including seaweeds in a Nordic diet

Flavour descriptive analysis – Association test

The sensory panel consisted of 15 judges aged 28 to 58 who were recruited among project partners, i.e. Møreforsking and Klippfiskakademiet (cooking school) employees. Prior to testing, all judges were somewhat familiar with macroalgae as a raw material, but had limited experience with sensory profiling of seaweed.

Test design: Four species of seaweed were presented to the judges, i.e. (i) sugar kelp Saccharina latissima (Linnaeus) C. E. Lane, C. Mayes, Druehl et G. W. Saunders, (ii) winged kelp Alaria esculenta (Linnaeus) Greville, (iii) oarweed Laminaria digitata (Hudson) J. V. Lamouroux, as well as the red alga dulse [Palmaria palmata (Linnaeus) F. Weber et D. Mohr]. Algal material was acquired from a commercial supplier (The Northern Company™, Oslo, Norway) using dried seaweeds from Iceland. The study used material from a single batch for each species to ensure standardized material. Seaweeds were presented as fragments of 2–5 cm², of approximately 1–2 g dry weight, and identified by 3-digit numbers.

The test included two parts separated by a break of approximately 10 min: (i) first, seaweeds were presented in the dried condition (as supplied), and (ii) prior to presentation, seaweeds were rinsed in cold water, covered with a plastic film and then heated in a microwave oven (30 s at 600 W).

Panel members were asked to list key words which they associated with the texture, flavour, and mouth-feel attributes of the different seaweed species (Lyon and Watson 1994, Monteleone and Langstaff 2014). Subsequently, associated terms were collected and grouped by species and treatment method, and the most consistent terms were selected to design a quality control test (Lyon and Watson 1994, Mojet and de Jong 1994, Gawel et al. 2000, Koch et al. 2012).

Quality control test

The panel of judges consisted of seven chefs aged 17 to 65, gathered for a seaweed cooking workshop. All of the subjects worked regionally, i.e. with a focus on Nordic food traditions, and four of them had some prior experience with using seaweeds in their cooking.

Sample presentation: The same four species of seaweeds as in the Association test, i.e. Saccharina latissima, Alaria esculenta, Laminaria digitata and Palmaria palmata, acquired from the same supplier (The Northern Company™) dried and vacuum packaged, were presented to the panel. Laminaria digitata was used as a reference, as it appeared relatively mid-scale in relation to its attributes after the Association test. Samples were presented as for the Association test, i.e. first in dried form and subsequently pre-treated with heat. Heat treatment in this case consisted of (i) rinsing in cold water and (ii) steaming at 100°C for 15 min.

Test performance: Panel subjects were asked to compare 11 taste, flavour and mouth-feel qualities (Glossary, Table 2) of three species of seaweeds to a reference sample (L. digitata) by using a quality control test relative ranking method (ISO:6564, 1985). Hence, for each sample and quality, panellists related its placement to a reference sample of L. digitata on a numerical scale from -3 to +3, with L. digitata representing 0.

Panel judges underwent a training session in order to get familiarized with the material, after which different perceptions of tasks and use of test-forms were discussed and calibrated for consistent test performance across the panel.

The main testing was carried out in two parts with a 15-min break in between. Part one was carried out with dried seaweed samples and part two was carried out with steamed material.

Cooking workshop

The same chefs who constituted the panel for the quality control tests participated in a cooking workshop following the sensory analysis of the four seaweed species. This was carried out at Klippfiskakademiet, the partner cooking school, with suitable medium-scale facilities. Over the course of a day, the chefs first discussed the integration and use of the selected seaweed species in Nordic cuisine and then prepared a selection of 17 dishes where these ideas were implemented. In a creative process, chefs specifically explored how the final food products were affected by (i) preparation method of seaweed raw material, (ii) amounts of raw material used, (iii) application of seaweeds as sea vegetables, relish or flavour enhancers in broth or condiments, and (iv) physical properties and colour of the material.

However, the procedure was both flexible and unregimented, and the above variables were not systematically controlled for.

Resulting dishes were presented to a tasting panel of 17 volunteers who each selected favourite dishes, awarding 3, 2 and 1 points, respectively for the three highest ranking dishes. In addition, the chefs summarised their experiences and conclusions qualitatively.

Consumer test of fish cakes with and without Saccharina latissima

In November 2014, a consumer test was conducted at Ålesund aquarium, comparing a seafood product (fish cakes) which contained dried flakes of Saccharina latissima, to a control fish cake without seaweed ingredient. The tasting panel consisted of visitors to the aquarium volunteering to participate in this research on seaweed as food ingredients.

Fish cakes were prepared in two versions, distinguished only by the content of about 1 g dried sugar kelp per 20 g fish cake in version A, and about 1 g of parsley per 20 g fish cake in version B. The primary purpose of replacing seaweed with steamed (i.e. neutral-flavour) parsley was to remove any visual clues to the content of seaweed ingredients which might bias the evaluation.

Questionnaires consisted of the following elements: (i) age and gender information (5 age groups: <20, 20–35, 36–50, 51–65, >65 years), (ii) choice of preference for product A or B, and (iii) separate assessments of taste and appearance of product A and B using a 7-category scale (dislike strongly, dislike moderately, dislike slightly, neither like nor dislike, like slightly, like moderately, like strongly).

Statistical analysis

For sensory testing, data were collected and pre-processed using common spreadsheet software for descriptive statistics (Association test). In the case of quality control testing, we used multivariate principal component analysis (PCA) using Unscrambler™ software (Version × 10.2, CAMO Software, Oslo, Norway) to visualise primary correlations between the selected seaweed species and their sensory attributes.

Factorial two-way ANOVA (seaweed species and preparation type as independent factors) was applied for comparing rankings of individual qualities (STATISTICA™ software, StatSoft, Boston, USA).

Homogeneity of variances was tested by using Cochran’s test, and Tukey’s honest significant difference (HSD) test was performed for post-hoc comparisons of significant ANOVA results. In case of the consumer test, we used Wilcoxon signed-rank tests to compare assessments for the two product types regarding (i) taste and (ii) appearance.

Flavour descriptive analysis – Association test

From the total of attributes generated during the seaweed Association test (22 attributes for dried and 15 for steamed seaweeds, Table 1), a set of 11 quality descriptors was extracted (Table 2) by grouping the attributes used most frequently. In addition to the primary flavour qualities, others such as “odour intensity”, “grassiness” and “liquorice” flavour, as well as several mouth-feel attributes (“crispiness” and “chewiness”) were prominent. To a large extent, variation in mouth-feel attributes reflected the different preparation methods (dried vs. steamed presentation).

With regard to flavour differences and characterisation of the different seaweed species, Laminaria digitata appeared to be placed most neutrally for the majority of quality descriptors. Hence this species was chosen as the 0-level reference for the quality control test, with which the other three species were compared.

Quality control test

Flavour and mouth-feel qualities of four seaweed species were compared using a set of 11 quality descriptors. In each case, three seaweed species were ranked against Laminaria digitata as a neutrally placed reference, and seaweeds were presented both dried and steamed.

Palmaria palmata was most distinct overall with regard to flavour and mouth-feel qualities compared to the other two species tested against L. digitata as a reference (Figure 1A, where P. palmata is placed furthest from the other species). Standardised PCA plots are commonly used in sensory analysis to visualise the relationship among sensory attributes, as well as among individual product samples. The PCA-loadings plot (Figure 1A) displays the positioning of, and association among, the seaweed attributes. In this case, “odour intensity” and “crispiness” were the attributes primarily distinguishing the seaweeds along axis PC-1, explaining 58% of the variation among samples. Although the distribution of attributes in Figure 1A is much more clustered around zero than the distribution of seaweeds in Figure 1A, comparing results from the PCA-scores and PCA-loadings plots (Figure 1A/B) indicates that Saccharina latissima was positively correlated with the attribute “crispiness”, while P. palmata was positively correlated with “odour intensity”. The vertical axis PC-2 explained 29% of the variation of among the seaweeds’ sensory attributes, separating the attributes “sour”, “salty” and “bitter” on the one hand from “sweet” on the other hand.

Figure 1:

Results of quality control test: Principal component analysis (PCA) plots of seaweeds ranked for 11 sensory attributes relative to Laminaria digitata as the 0-level control. Top panel (A), PCA scores; bottom panel (B), PCA loadings. Samples are identified by species (LD – Laminaria digitata, AE – Alaria esculenta, SL – Saccharina latissima, PP – Palmaria palmata), as well as preparation method (S – steamed, vs. D – dried), with n=7 per group.

Saccharina latissima had the best correlation with flavour attributes “salty”, “sour” and “bitter” (Figure 1B), whilst Alaria esculenta appeared most similar to the reference species L. digitata with the attribute “chewiness” having the best correlation for the species.

Whereas PCA plots are useful in visualising overall variation in the samples with regard to sensory attributes, we used ANOVAs to test for differences in sensory profiles among the species assessed. The pre-treatment of seaweeds had no significant overall effect on the flavour qualities and, among mouth-feel qualities, only chewiness was affected by pre-treatment (ANOVA-results, Table 3). Both flavour and mouth-feel qualities varied significantly among species (Table 3). This result, confirmed by and visualised in the PCA-scores plot (Figure 1A) allowed us to carry out further analyses among seaweeds across treatments (steamed/dried), i.e. by pooling the respective treatment groups.

Significant differences among seaweed species for the individual quality attributes are summarised in Figure 2A–D. Palmaria palmata is clearly distinct from the two kelp species regarding mouth-feel and texture attributes in that it is softer, less crispy and more easily dissolved (Figure 2A). Its strong odour corresponds to a large degree with its ranking highest on “umami” flavour (panellists’ feedback, Figure 2A, D). In spite of its name, sugar kelp S. latissima did not appear to be sweeter than the other seaweeds. On the contrary, S. latissima appeared least sweet (Figure 2C) and instead most salty, sour and bitter compared to the other species tested (Figure 2B, C). The two kelps were best distinguished by the attributes “sour”, “salty”, “sweet” and “bitter”, in that A. esculenta appeared more sweet, and less salty, sour and bitter than S. latissima (Figure 2B, C). The flavour qualities, “grassy”, “liquorice” and “sea-association” did not contribute to distinguishing the species tested here (no significant difference among species, ANOVA, Tukey’s HSD-test with p>0.05, Table 3).

Figure 2:

Results of quality control test: Comparative ranking of Alaria esculenta, Saccharina latissima and Palmaria palmata for 11 quality descriptors. For each sample and quality, panellists related its placement to a reference sample of Laminaria digitata on a numerical scale from -3 to +3, with L. digitata representing 0. Average scores of both steamed and dried products (n=14)±SD. Posthoc tests (Tukey HSD) reveal significant differences (p<0.05) among species for the descriptors in each panel as indicated by the letters below the histograms (identical letters indicate no significant differences). For the remaining attributes (“grassiness”, “liquorice” and “sea-assoc.”), no significant differences among species were found.

Cooking workshop

The cooking workshop revealed unexpected possibilities regarding the use of seaweeds in a Nordic cuisine. Seaweeds were included in all types of dishes, including seafood, meat and vegetarian dishes, as well as desserts and even drinks (Table 4). Aside from the wide usability of seaweeds in cooking, one of the primary results from the workshop was that flavour attributes were easily lost when only small amounts of seaweeds were included, i.e. all species used appeared less dominant than previously expected. This led to a gradual increase in the amount of seaweed added to the various dishes throughout the cooking workshop.

Another important aspect was the visual changes of the seaweeds during the course of heat-treatment. Kelps went from olive-brown to bright green at first and then to a less distinct greenish colour. The brown colour of kelps is due to the presence of carotenoid pigments, mainly fucoxanthin, bound to chlorophyll a and c to form light-harvesting complexes (Douady et al. 1994). Whereas heat treatment of brown seaweeds, especially boiling, generally tends to increase phenol content, including fucoxanthin (Rajauria et al. 2010, Amorim-Carrilho et al. 2014), high temperature during cooking may dissociate the fucoxanthin from the light-harvesting complexes and consequently reveal the green colour of chlorophylls.

In contrast, the texture of the kelps used in this test remained comparatively stable, whereas the red seaweed P. palmata easily disintegrated, and lost its colour to the accompanying food ingredients – an effect that was explored further by using P. palmata to colour pasta as well as a vodka drink. The alginate content of brown seaweeds was utilised for altering viscosity, i.e. Saccharina-flakes were added to both desserts (a pancake dough, as well as chocolate ice-cream) so as to increase the dishes’ body. Semi-quantitative evaluation of dishes by a tasting panel revealed that both seafood-dishes, as well as vegetarian food combinations, were perceived to be tasty by the panellists (Table 4). The highest rated dish by far was the chocolate ice-cream with sugar kelp (S. latissima), followed by clipfish (dried and salted cod) broth with A. esculenta, and cod with S. latissima and P. palmata in third place.

In conclusion, seaweeds were easily combined with Nordic dishes and food items, and did not appear dominant in the flavour spectrum. Beyond the actual flavour, they have a large potential to affect texture, appearance and colour, as well as consistency of a variety of dishes.

Consumer test of fish cakes with and without Saccharina latissima

This test revealed clearly that fish cakes in which herbs were substituted by sugar kelp (Saccharina latissima) were equally attractive to consumers with regard to both flavour and appearance [Wilcoxon signed rank test (R): V=646, p=0.73, Figure 3]. Consumer responses to the direct question of preference for one of the two types of fish cake showed that both products were equally popular and that 22% of respondents could detect no difference (Figure 4). Neither age, nor gender affected the responses significantly.

Figure 3:

Frequency distribution of responses from consumers for taste and appearance of fish cakes containing sugar kelp (Saccharina latissima) or no seaweed (control with parsley). n=103 respondents (43 females, 60 males).

Figure 4:

Pie chart showing responses from consumers to question “Which type of fish cake do you prefer?” with fish cakes containing sugar kelp (Saccharina latissima) or no seaweed (control with parsley). n=103 respondents (43 females, 60 males).