Impact of Four Different Chlorella vulgaris Strains on the Properties of Durum Wheat Semolina Pasta

Abstract

Microalgae are a promising protein source due to their high protein content; high reproductivity; and low carbon, water, and arable land footprints. In this study, the impact of adding 3 and 5% of four Chlorella vulgaris strains, namely Smooth (SCV), Honey (HCV), White (WCV), and New Honey C. vulgaris (NHCV), on the processing, cooking behavior, color, firmness, structure, and sensory properties of durum wheat semolina pasta was investigated. It was hypothesized that (1) changes in physical properties depend on strain and concentration, (2) acceptability varies by strain due to different colors, odors, and flavors, and (3) the absence of fishy odors and flavors is crucial for acceptance rather than color. The results show that high-quality pasta could be produced with all C. vulgaris strains and concentrations. Cooking time and water absorption of all samples decreased but only significantly for the samples with NHCV added. Also, the bite resistance (determined instrumentally and sensorially) increased for almost all samples due to increasing protein and fiber content. A clear concentration dependency could not be found. In terms of sensory acceptance, NHCV performed the best, and an unaltered typical odor was identified to be crucial rather than color or the absence of fishy odor.

1. Introduction

The green microalgae Chlorella vulgaris and the cyanobacterium Arthrospira platensis (Spirulina) have a long history of use and have been cultivated in Europe for many years as food and food additives [1]. Especially in recent years, there is an increasing interest in plant proteins and other alternatives to meat due to the ethical, health, and environmental concerns that go along with meat consumption [2,3,4]. Compared to terrestrial crops, microalgae have many advantages, such as high productivity; a high content of macro- and micro-nutrients; and low carbon, water, and arable land footprints [5]. However, food products containing either whole microalgae or compounds derived from microalgae are a minority on the food market as they are primarily sold as dietary supplements, e.g., capsules, tablets, or powder for ingestion [6]. According to Lafarga [6], several challenges related to production capacities and costs, as well as legislation, must be overcome, but also the intense green color and a fishy taste and aroma represent one of the main problems in the integration of microalgae into food. The acceptance of products containing (micro)algae depends strongly on the traditional diet and is significantly higher in eastern countries [7,8].

The addition of microalgae into recipes for traditional food products, such as bread, is a global trend [6]. However, not much is known about the effects of microalgae supplementation on food properties. Fanari et al. [9] reported that adding 2.5 or 5.0% Spirulina or differently colored C. vulgaris strains (green, yellow, white) to energy bars not only changes the color, flavor, and texture of the products but can offer different colorways as well as different sensory profiles during product development. In the case of a fermented whey sport drink enriched with Spirulina (0.25–0.75%) and a fermented yogurt enriched with Spirulina (0,25–1.0%), only lower concentrations (0.5 and 0.25%, respectively) proved to contribute positively to stability and taste [10,11]. Also, Graça et al. [12] reported that the substitution of wheat flour dough with 1.0–5.0% green C. vulgaris was only beneficial up to a concentration of 3.0%, while at higher concentrations, dough rheology, bread texture, flavor, and bread aging were negatively affected. Green microalgae incorporation into broccoli soup (0.5–2.0% Spirulina sp., Chlorella sp. or Tetraselmis sp.) and into wheat tortillas (0.5–3.0% Nannochloropsis sp. or Tetraselmis sp.) revealed changes in color and higher amounts of bio-accessible polyphenols in the microalgae-containing products [13,14]. Although the sensorial acceptability index was reduced for soups containing microalgae compared to the control, it was still 80% for samples containing 0.5% Spirulina or Tetraselmis sp. [13]. The green color and the fishy taste, therefore, seem to be less disturbing in green foods, which have a strong taste of their own. When adding the green microalgae Nannochloropsis and Tetraselmis to baked crackers and bread, substitution was limited to 2.5% for crackers and 1.0–2.0% for bread containing Nannochloropsis and Tetraselmis, respectively. The sensory evaluation showed the products to be only competitive with the controls while having an improved nutritional value and including a “trendy” ingredient [15]. These findings are supported by recent consumer studies showing that young, physically active, well-educated Italian men who are interested in healthy eating and primarily have a plant-based diet are open to trying new products, such as pasta containing 3% Spirulina [16]. In addition to consumer openness, the label on the packaging, e.g., organic, Nutri-Score, or vegan, also plays a major role in whether consumers see added value in buying microalgae-enriched pasta and would pay more money for it [17].

Pasta is an interesting food to be substituted with algae, as consumers are familiar with different colors (yellow, red, green, black). In addition, a fishy taste could be used positively to improve the taste of vegan pasta dishes with fish (sauce) alternatives. Many different microalgae (C. vulgaris, Spirulina, Isochrysis galbana, Diacronema vlkianum, Dunaliella salina, Nannochloropsis sp.) have already been added to pasta products to enhance the nutritional content or antioxidant potential [18,19,20,21,22,23,24,25,26]. However, only a few studies have looked at the changes in the physical properties of pasta with different algae concentrations and their sensory acceptance. Rodríguez De Marco et al. [25] showed that although the content in the ω-3, poly-unsaturated fatty acid eicosapentaenoic acid (EPA) in pasta increased with increasing concentrations of the green microalgae Nannochloropsis (10–40%), the optimum cooking time, water adsorption, and swelling index decreased, indicating a significant impact of the microalgae on starch swelling and gelatinization. Besides textural changes, the sensory evaluation revealed reduced global acceptability at all microalgae concentrations compared to the control [25]. El-Baz et al. [26] showed that the addition of 1–3% green D. salina to pasta slightly improved nutritional parameters but resulted in reduced dough stability and, therefore, strength with increasing algae concentrations. Besides color changes, the swelling and cooking loss of the pasta increased, indicating the destabilization of the structure. However, flavor and overall acceptability scores were quite high for all samples (>8 on a 10-point scale) [26]. For spaghetti supplemented with the golden-brown microalgae I. galbana and D. vlkianum (0.5–2.0%), a significant increase in EPA and docosahexaenoic acid (DHA) was reported, but again, the quality parameters of optimal cooking time, cooking loss, swelling index, and water absorption changed. Sensory evaluation revealed a fishy flavor for samples containing 2.0% algae, while all other attributes (color, odor, texture, global appreciation) decreased with increasing algae concentration, especially for spaghetti supplemented with D. vlkianum [24]. The same authors reported similar findings for spaghetti containing 0.5–2.0% C. vulgaris (green or orange strains) or Spirulina. In particular, the sensory evaluation indicated lower scores for the attributes of odor and flavor when 2% green (fishy) algae was added, while the orange C. vulgaris strain was most preferred by the panelists [19].

All previous studies show that the use of microalgae in food and pasta is limited to concentrations of 2–3% due to strong sensory changes. However, in order to achieve a measurable added value for the marketing of algae pasta through substitution with microalgae [16], significantly higher concentrations would have to be used. The use of microalgae in food should, therefore, be seen as a nutritional enrichment (introduction of small amounts of macro- and micronutrients), but its use as a source of protein is difficult. In this study, the effect of 3 and 5% addition of four strains of C. vulgaris on pasta was tested. The aim was to determine the most suitable strain in terms of physical stability and sensory evaluation and to develop a marketable product. Three working hypotheses were formulated from the literature above: (1) the physical changes of pasta caused by microalgae are dependent on concentration, species, and/or strain; 2) different algae species and/or strains can introduce different colors and flavors, which have different effects on sensory acceptance; and (3) independent of color, the absence of fishy odors and flavors will positively contribute to pasta acceptability. To test the hypotheses, durum wheat semolina was supplemented with 0.0, 3.0, and 5.0% differently colored (green, yellow, and white) C. vulgaris strains, namely Smooth C. vulgaris (SCV, green), Honey C. vulgaris (HCV, yellow), White C. vulgaris (WCV, white), and New Honey C. vulgaris (NHCV, yellow), which was specifically bred for the EU ProFuture project, as previously described by Schüler et al. [27]. The semolina–algae mixtures were used for fusilli and lasagna sheet formulations. The net power consumption of the pasta machine, as well as the final water content of the dried pasta, were determined. Then, physical properties (optimal cooking time, water absorption, cooking loss, color, bite resistance/firmness, microscopic surface structure) and sensory changes (quantitative-descriptive analysis, acceptance, multiple logistic regression, principal component analysis) of the products were evaluated.

2. Materials and Methods

2.1. Material

The four heterotrophically grown strains of C. vulgaris (SCV, LOT No. L201950311; HCV, LOT No. 2021HC032; WCV, LOT No. 2021WC007; NHCV, LOT No. 2022HC006) were kindly provided by Allmicroalgae Natural Products, S.A, Pataias, Portugal. Organic durum wheat semolina (LOT No. S-1712200000-003) was purchased from Antersdorfer Mühle GmbH & Co Vertriebs KG, Simbach/Inn, Germany. The composition of the ingredients is given in

Table 1. Composition of the algae strains according to the certificates of analysis of the specific batches, as well as the approximate composition of the durum wheat semolina according to its specification.

Parameter (g/100 g)	SCV	HCV	WCV	NHCV	DWS
Protein	26.3	30.5	36.1	33.5	~11.0
Fat	7.0	11.5	7.6	10.5	~1.4
Carbohydrates incl. DF	58.1	53.7	48.8	51.3	~72.1
thereof DF	–	12.9	–	13.1	~3.1
Ash	4.0	2.0	4.6	2.1	–
Moisture	4.9	2.4	2.9	2.6	11.9 1
Energy (kcal/100 g)	367.0	414.5	356.0	408.0	~339.0

¹ Specified as <14, analyzed in more detail with OHAUS MB23 infrared dryer (OHAUS Europe GmbH, Nänikon, Switzerland). ~, approximately; DF, dietary fiber; DWS, durum wheat semolina; HCV, Honey C. vulgaris; NHCV, New Honey C. vulgaris; SCV, Smooth C. vulgaris; WCV, White C. vulgaris.

and in the original documents in Appendix A, Figure A1, Figure A2, Figure A3, Figure A4 and Figure A5.

2.2. Pasta Production

2.2.1. Production

Pasta was produced using a Pasta V 300 tabletop pasta machine (Karl-Heinz Häussler GmbH, Heiligkreuztal, Germany) with a volume capacity of approximately 4 kg and a production capacity of up to 12 kg/h. Brass-coated matrices for the production of lasagna sheets (flat matrix no. 99 V, gap set to 1.6 mm) were used for certain physical analyses, and fusilli (matrix no. 49 c) were used for physical analyses and sensorial testing. Two types of pasta, each with a volume of 4 kg, were produced per production day. The production of each type (control, 3% Chlorella additive, or 5% Chlorella additive per strain in the solid) was repeated three times to obtain three temporally independent replicates per pasta variety (sample). The quantities of ingredients (tap water at 30 °C, durum wheat semolina, Chlorella) were selected so that the final moisture content of the pasta dough (bulk + material moisture) was 34.00 ± 0.04%. Due to the different moisture contents of the raw materials, this resulted in different water and solid contents, whereby the percentages of Chlorella (3 and 5%) relate to the total solid content (74.54–74.73%, depending on the strain) (

Table 2. Relative quantities of ingredients used for the pasta formulations in percent and calculated final dough moisture in percent.

Ingredient	Control	3% SCV	5% SCV	3% HCV	5% HCV	3% WCV	5% WCV	3% NHCV	5% NHCV
DWS	75.00	72.49	70.88	72.39	70.90	72.39	70.90	72.48	70.81
Microalgae	–	2.24	3.73	2.24	3.73	2.24	3.73	2.24	3.73
Tap water	25.00	25.27	25.39	25.37	25.37	25.37	25.37	25.28	25.46
Total	100.00	100.00	100.00	100.00	100.00	100.00	100.00	100.00	100.00
Final dough moisture	34.02	34.03	34.04	34.01	33.99	34.01	33.99	33.96	33.98

DWS, durum wheat semolina; HCV, Honey C. vulgaris; NHCV, New Honey C. vulgaris; SCV, Smooth C. vulgaris; WCV, White C. vulgaris.

).

The ingredients were mixed and kneaded in two stages. First, the solids were fed into the machine. Durum wheat semolina and microalgae were mixed dry at 100 rpm speed for 6 min. During stage 1, 600 g of the tap water (30 °C) was fed into the running mixer over a period of 3 min at a speed of 100 rpm for a total time of 6 min. During stage 2, the remaining tap water (30 °C) was fed into the running mixer over a period of 2 min at a speed of 100 rpm for a total time of 4 min. The subsequent extrusion was carried out in two passes. First, the fusilli matrix was used at an extruder speed of 100 rpm and an automatic cutter (3–4 cm length). Subsequently, the flat matrix was used at an extruder speed of 100 rpm. The lasagna sheets were cut manually (30 cm length, 9 cm width). During production, the power consumption of the pasta machine was tracked every 10 s using a power meter (VOLTCRAFT SEM6000SE, Conrad Electronic SE, Hirschau, Germany). The net power consumption was calculated by subtracting the power consumption of the idle machine with the extruder and cutter switched on. Per sample, 170 values were used for statistical analysis.

2.2.2. Drying

Drying of the pasta was performed using a proofing cabinet (MIWE condo 4.0608, MIWE Michael Wenz GmbH, Arnstein, Germany). The fresh pasta was spread in a single layer on perforated trays (approx. 1 kg per tray, 8 trays in total per production day) and dried for 20 h at 78% humidity and 45 °C temperature to obtain a final water content < 13% according to the guidelines for pasta products in the German Food Codex [28].

2.3. Physical Analyses

2.3.1. Residual Moisture Content after Drying

The residual moisture content of the dry pasta was determined on ground fusilli using an OHAUS MB23 infrared dryer (OHAUS Europe GmbH, Nänikon, Switzerland). Approx. 200 g fusilli was ground in a cutting mill (Robot Coupe R2, Robot Coupe SNC, Vincennes, France) at 1.500 rpm. For each drying trial, 10.00 ± 0.20 g of ground pasta was weighed into the corresponding aluminum sample dish (Ø 80 mm), evenly distributed, and dried at 140 °C on automatic mode up to constant mass. Results were given in percent by the device. Drying was repeated five times per pasta replicate (n = 15 per sample).

2.3.2. Optimal Cooking Time

The cooking time until optimal bite (“al dente”) was determined according to the Cereals and Grain Association (AACC) method AACC 66-50.01 [29] with slight modifications. Two 600 mL glass beakers (DURAN^® beaker, low shape with spout, with graduation, DWK Life Sciences GmbH, Wertheim/Main, Germany), each containing 300 mL distilled water, were heated on heating plates (SLR 01000541, SI Analytics GmbH, Mainz, Germany) until bubbling boiling. Into one beaker, 25.0 ± 2.0 g fusilli was added, and the time measurement was started immediately. The second beaker was used to fill up the evaporated water with boiling water to keep the water level constant at 300 mL during the whole cooking process. It was also ensured that the water boiling resulted in the generation of air bubbles throughout the cooking process and that the pasta did not stick together. From min 2, one fusilli was removed from the boiling water at 30 s intervals, placed between two layers of cling film, and mashed. The optimum cooking point was defined as the cooking time at which the pasta can be mashed completely, i.e., including the core. If the optimal cooking point was exceeded after a 30 s interval, 10 s steps were introduced. Cooking time until “al dente” was determined three times per pasta replicate (n = 9 per sample).

2.3.3. Water Absorption

The method for determining the water absorption was based on the method for determining the optimal cooking time [29]. As described above, 25.0 ± 0.1 g dry fusilli (w₀) was added into a 600 mL glass beaker containing 300 mL bubbling boiling distilled water and cooked until “al dente” while keeping the water level at 300 mL. Meanwhile, we made sure that the pasta did not stick together and that the water kept boiling the whole time. An analytical balance (Sartorius Entris^® II BCA6202-1S, Sartorius AG, Göttingen, Germany) with a 600 mL round vessel and lid (originality BD 002-600 O, Pöppelmann GmbH & Co. KG Kunststoffwerk-Werkzeugbau, Lohne, Germany) was tared. The pasta was poured into a sieve, allowed to evaporate for 30 s, poured into the pre-tared sample vessel, and tightly closed with the lid. The weight of the cooked pasta (w_c) was noted, and the absorbed water (W_A) in percent was calculated as the difference (Equation (1)), as previously described by Fradique et al. [19]. W_A was determined three times per pasta replicate (n = 9 per sample).

W_A [%] = (w_c − w₀)/w₀ × 100

(1)

2.3.4. Cooking Losses

Solid loss to cooking was determined on pieces of lasagna sheets by evaporation of the cooking water to constant weight [19]. First, 1000 mL of tap water was added into a 3 L glass beaker (DURAN^® flat flange beaker, DN 150, DWK Life Sciences GmbH, Wertheim/Main, Germany) and heated on an SLR 01000541 heating plate (SI Analytics GmbH, Mainz, Germany) set at level 24 until boiling. The heating plate was set at level 18, 100 g broken lasagna sheets were added, and time measurement was started. After exactly 10 min, the lasagna sheet pieces were fished out of the water with a slotted spoon and discarded. The hot cooking water was transferred into 1 L glass bottles (DURAN^®, GL45, 1 L, Schott AG, Mainz, Germany). After the cooking water had cooled down, it was filled up to exactly 1.000 mL with tap water. The subsequent determination of the dry matter (DM) was carried out in triplicates (n = 9 per sample). First, a weighing bottle (DURAN^® boro 3.3, low form 30 × 50 mm, Lenz Laborglas GmbH & Co. KG, Wertheim am Main, Germany) was filled with approximately 10 g of sea sand (purified by acid and calcined, TH. Geyer GmbH & Co. KG, Renningen, Germany) and a small glass rod, dried in a drying oven at 103 ± 2 °C (UF110, Memmert, Schwabach, Germany), and then cooled in a desiccator. The weight in g was determined (w₁). Subsequently, exactly 5 g of cooking water (w₀) was added and evenly mixed with the aid of the glass rod. After drying to constant mass (at least 8 h) at 103 ± 2 °C, the sample was cooled in the desiccator and weighed again (w₂). DM in percent was calculated using Equation (2).

DM [%] = 100 − (w₀ − (w₂ − w₁))/w₀ × 100

(2)

2.3.5. Color Analysis

The color of dry and cooked (10 min) lasagna sheets was analyzed using a CM-600d colorimeter (Konica Minolta, Tokyo, Japan) under standardized light D65. Per pasta replicate, 10 measurements were taken from the top and 10 from the bottom side of the lasagna sheet (n = 60 per sample). The Euclidian color distance (ΔE) was calculated, as previously described by Van De Walle et al. [30] (Equation (3)):

ΔE = square root ((L*_c − L*_s)² + (a*_c − a*_s)² + (b*_c − b*_s)²)

(3)

where L*_c, a*_c, b*_c and L*_s, a*_s, b*_s are the L*a*b*-values of the control and the sample to compare, respectively.

2.3.6. Bite Resistance/Firmness

To determine the bite resistance/firmness of the pasta, lasagna sheets were prepared as follows. First, 6000 mL of tap water was heated in a 10 L metal pot using an SLR 01000541 heating plate (SI Analytics GmbH, Mainz, Germany) set at level 24 until boiling. The heating plate was set at level 18, 200 g lasagna sheets were added so that they lay crosswise on their side, and time measurement was started. After exactly 10 min, the lasagna sheets were poured into ice water for 30 s. The cool lasagna sheets were spread flat on absorbent paper and turned three times to dry them off. Both sides of the lasagna sheets were cut off so that the final width (w) was 6 cm. Per measurement, one lasagna sheet was placed on a Texture Analyzer TA XT2 (Stable Micro Systems Ltd., Godalming, UK) with a 25 kg load cell. The force (F) in N and the distance (d) in mm were measured using a blunt blade (8 cm width) as a measuring probe. Measurement parameters were set as follows: probe height at start of measurement, 5.0 mm; pre-test speed, 1.0 mm/s; test speed, 0.5 mm/sec; test distance, 96%; trigger value, 0.4 N. The measurement was repeated 10 times per pasta replicate (n = 30 per sample). Bite resistance/firmness (p) in kPa was calculated, taking into account the cut surface in m² (Equation (4)):

p [kPa] = F/(w × d × 1000)

(4)

where w is the width of the lasagna sheets in m and d is the test distance (thickness of the lasagna sheets) in m.

2.3.7. Scanning Electron Microscopy

Scanning electron microscopy (SEM) was used to visualize the surface structure of the pasta. The lasagna sheets were cooked as described for determining firmness/bite resistance, and a certain area was cut out, frozen in liquid nitrogen, and inserted into a cryo-preparation system (K 1250, Emitech SAS, Montigny-le-bretonneux, France) for the sublimation of free water. Microscopic images were taken at different magnifications, as previously described by Witte et al. [31]. After the samples’ surface had been sputtered with gold, the sample was transferred into the microscope (JSM 6460 LV, JEOL, Akishima, Japan) at approx. −180 °C, and an electron beam was generated and accelerated to a voltage of 1–30 kV.

2.4. Sensory Evaluation

After approval by the Ethical Commission for Sensorial tests of ILVO, Quantitative Descriptive Analysis (QDA) sessions of the pasta were performed in a standardized taste lab (ISO 8589:2007 [32], 2007). First, pasta attributes of fusilli with and without 5% SCV were defined in a free choice profiling session with trained panelists, including yellow and green fusilli from a colored mix as references (DELVERDE Eliche Tricolore, bought at a local supermarket). Then, the evaluation of the attributes of typical taste, fishy/off-odor, after/off-taste, saltiness, bite resistance (after chewing 5 times), stickiness, and color was performed on a 10-point linear scale ranging from 0–10 in four independent QDA sessions. The color scales used can be retrieved from the Supplementary Data (Figure S1). In addition, the general acceptance as a newly developed commercial product (yes/no) was surveyed. Each QDA session included (1) an external fusilli control (Barilla Fusilli No. 98, bought at a local supermarket) to be able to better classify the parameters of typical taste (set to 8), fishy/off-odor and after/off-taste (set to 0), saltiness (set to 6), bite resistance (set to 7), and stickiness (set to 2) in the scale; (2) a self-made fusilli containing 3% marine algae Tetraselmis chui to be able to better classify the parameters of fishy/off-odor and after/off-taste (both set to 9) in the scale; and (3) four to five self-made fusilli samples (control or samples with algae added). For each sample, 100 g was cooked in 1 L tap water without salt based on the determined cooking time. For the Barilla control, 100 g was cooked in 1 L tap water containing 15 g NaCl for 11 min. For the T. chui pasta, 100 g was cooked in 1 L tap water without salt for 150 s. All pasta samples were portioned to 15–20 g in small plastic vessels and kept warm at 60 °C (drying cabinet UL 40, Memmert GmbH + Co. KG, Schwabach, Germany) for 10 min before serving. The randomized samples were tasted likewise at 50–55 °C at the same time by 11 panelists.

2.5. Data Analyses

All data analyses [multiple logistic regression analysis (MLRA), principal component analysis (PCA), statistical analyses] were performed using Sigma Plot 15.0 software (Systat Software Inc, San Jose, USA). MLRA was conducted using all sensory attributes as independent variables and acceptance (0 for unacceptable, 1 for acceptable) as the dependent variable. PCA was performed using sensory attributes as variables and sample names as labels. For statistical analyses, One-Way Analysis of Variance (ANOVA), followed by Tukey’s post hoc test (95% confidence interval; p-value = 0.05), was selected to analyze all data. If the normality test (Shapiro–Wilk) and/or equal variance test (Brown–Forsythe) failed, an ANOVA on Ranks followed by Tukey’s post hoc test (95% confidence interval) was performed. This was the case for all data except the sensory parameters of typical odor, fishy/off-odor, and bite resistance. For color analyses, ANOVA on Ranks, followed by Dunn’s post hoc test (95% confidence interval), was used on program default.

3. Results

3.1. Changes in Processing Behavior and Composition

The net power consumption of the pasta machine was analyzed to reveal possible changes in the dough properties (

Table 3. Net power consumption of the pasta machine in W and residual moisture content of dry pasta in percent (means ± SD).

Sample	Net Power Consumption (W)	Moisture Content (%)
Control	190.17 ± 42.84 a	10.24 ± 0.60 a
3% SCV	187.59 ± 42.29 abc	10.80 ± 0.74 ab
5% SCV	176.01 ± 41.64 ac	10.79 ± 0.60 ab
3% HCV	202.85 ± 56.38 abd	9.84 ± 0.60 abc
5% HCV	204.42 ± 56.91 abd	9.81 ± 0.98 abc
3% WCV	204.42 ± 51.83 abd	9.61 ± 1.13 ac
5% WCV	198.37 ± 51.96 abd	10.53 ± 0.94 abc
3% NHCV	206.12 ± 50.04 ad	9.86 ± 0.79 abc
5% NHCV	215.02 ± 56.04 d	9.45 ± 0.90 ac

HCV, Honey C. vulgaris; NHCV, New Honey C. vulgaris; SCV, Smooth C. vulgaris; SD, standard deviation; WCV, White C. vulgaris. Different superscript letters indicate groups with significant difference (confidence interval 95%).

). We assumed that changes in power consumption indicated a change in dough resistance due to changes in viscosity or stickiness caused by altered composition. It should be noted that the SDs are relatively high due to fluctuations in power consumption over the entire production process. Therefore, a significant increase in net power consumption was only observed for the sample 5% NHCV compared to the control and SCV samples, indicating an increase in dough viscosity/stickiness. Differences in the extrusion and drying behavior of the pasta could not be observed. The final moisture content of the dry pasta was between 9.5 and 10.8% for all samples and did not differ significantly from the control. None of the samples, therefore, seemed to have an altered water-holding capacity, and all samples could be regarded as storage-stable at room temperature.

Based on the final moisture content (Table 3) and ingredient composition (Table 1), an increase in the protein and fat content was calculated (0.4–1.2% and 0.2–0.5%, respectively), with the highest protein content in the 5% WCV and 5% NHCV samples and the highest fat content in the 5% HCV sample (Supplementary Data, Table S1). In addition, the carbohydrate content decreased (0.6–1.2%), with the lowest content in the 5% SCV and 5% WCV samples. Dietary fiber (DF) content simultaneously increased with increasing algae concentration and was dependent on the strain (e.g., 0.3 and 0.5% for samples containing 3 and 5% NHCV, respectively).

3.2. Changes in Physical Pasta Properties

According to Fradique et al. [19], cooking time, cooking loss, and swelling index/water absorption are important pasta quality parameters. A change in cooking time indicates changes in the rates of water penetration and starch gelatinization, while a change in cooking loss indicates a change in starch release to the cooking water [19]. When looking at the pasta quality parameters (

Table 4. Optimal cooking time in seconds, water absorption in percent, cooking losses in percent, and bite resistance/firmness in kPa (means ± SD).

Sample	Cooking Time (s)	Water Absorption (%)	Cooking Losses (%)	Bite Resistance/Firmness (kPa)
Control	190.0 ± 15.0 a	105.0 ± 6.5 a	0.3 ± 0.1 a	333.2 ± 58.5 a
3% SCV	180.0 ± 00.0 ab	96.9 ± 1.6 ab	0.4 ± 0.1 a	297.1 ± 45.6 a
5% SCV	183.3 ± 05.0 a	97.0 ± 4.2 ab	0.4 ± 0.1 a	386.3 ± 103.3 ab
3% HCV	180.0 ± 00.0 ab	100.7 ± 3.1 a	0.4 ± 0.0 a	407.0 ± 32.3 bc
5% HCV	181.1 ± 03.3 ab	95.9 ± 3.6 ab	0.4 ± 0.0 a	351.4 ± 33.9 abd
3% WCV	180.0 ± 00.0 ab	97.1 ± 2.9 ab	0.3 ± 0.0 a	437.6 ± 41.7 bc
5% WCV	180.0 ± 00.0 ab	98.0 ± 0.9 a	0.4 ± 0.1 a	296.7 ± 11.9 ad
3% NHCV	160.0 ± 15.0 b	88.2 ± 4.4 b	0.4 ± 0.0 a	456.1 ± 121.9 bc
5% NHCV	170.0 ± 15.0 ab	89.1 ± 3.2 b	0.4 ± 0.0 a	419.8 ± 31.1 bc

HCV, Honey C. vulgaris; NHCV, New Honey C. vulgaris; SCV, Smooth C. vulgaris; SD, standard deviation; WCV, White C. vulgaris. Different superscript letters indicate groups with significant difference (confidence interval 95%).

), the cooking time for all samples with added algae decreased by 10–20 s compared to the control but only significantly for the sample with 3% NHCV.

Only the samples with NHCV also showed a significant reduction in water absorption, while no significant differences were observed in cooking losses, indicating the formation of stable carbohydrate and protein networks. Firmness when cut with a blunt blade, which should reflect the bite resistance of teeth, significantly increased for samples with 3% HCV, 3% WCV, 3% NHCV, and 5% NHCV compared to the control, the 3% SCV sample, and the 5% WCV sample (Table 4). However, due to the large SDs of the 5% SCV and 3% NHCV samples, it cannot be clearly stated whether there is a concentration-related difference in firmness. Further, the addition of 3% SCV and 5% WCV resulted in a slight decrease in firmness compared to the control, but the difference was not significant due to overlapping SDs.

The appearance and the measured color of the dry and cooked lasagna sheets are shown in Figure 1.

The top and bottom surfaces differed in brightness after extrusion and drying (Figure 1a), which is due to air inclusions on the bottom side caused by the nature of the flat matrix (no improvement by changing parameters). However, as it proved difficult to classify sides after cooking (Figure 1b), L*a*b* average values were calculated for all samples without side differentiation (Figure 1c,d). The pictures of the pasta (Figure 1a,b) already show that the samples with SCV became green, and samples with HCV, WCV, and NHCV became more yellowish than the control. This was proven by color measurements of the dry pasta (Figure 1c), showing that all samples except 3% WCV were darker than the control (decreased L* value), the samples with SCV were greener (decreased a* value), the samples with NHCV were redder (increased a* value), and all samples except 3% SCV and 3% WCV were yellower than the control (increased b* values). Similar findings were made for the cooked samples (Figure 1d), but the color differences from the control (ΔE values) decreased after cooking, except for the SCV samples. For both dry and cooked, the color difference to the control was lowest for the WCV samples (3% WCV ΔE = 5.0 and 3.7; 5% WCV ΔE = 9.1 and 6.7), followed by the 3% HCV and the NHCV samples.

To reveal differences in the pasta structure that could explain changes in cooking time, water absorption, and firmness, scanning electron images of 10 min-cooked lasagna sheet cross-sections were taken (Supplementary Data, Figure S2 and Figure 2). Images from the outer edge (top/bottom) towards the center show that the edges were softened to varying degrees. Further inside, a fine network structure is recognizable. The 3% SCV, 3% WCV, and NHCV sample edges were less softened, with 5% SCV the most (Supplementary Data, Figure S2). Approx. 150 µm from the outer edge and at higher magnification, clearly distinguishable structures (semolina kernels) surrounded by water (black areas) are recognizable (Figure 2). A fine protein and carbohydrate network has formed in these structures and partly between the structures. In addition, 2–5 µm large, roundish structures are recognizable in all samples, including the control, which could be spherical starch grains on the one hand (approx. 5 µm) and algae cells (approx. 2 µm, not present in the control) on the other. The amount of unswollen starch granules significantly increases closer to the pasta center independent of the sample (Supplementary Data, Figure S3). The control shows a uniform network with few compact areas/structures (likely DF). A similar network structure is found in the samples with NHCV, but they have a higher proportion of compact and roundish structures. In all other samples, the networks seem to have finer mashes, indicating better swelling and starch gelatinization, as well as higher elasticity. However, larger areas with water inclusions (black holes) are also visible in pasta with added algae, except for the 5% WCV and 5% NHCV samples (Figure 2).

3.3. Changes in Sensorial Pasta Properties

Sensorial evaluation of the pasta samples in comparison to the control revealed a significant decrease in the typical pasta odor, especially for samples containing SCV or 5% HCV/WCV (

Table 5. Sensorial evaluation of the pasta samples (means ± SD).

Attribute	Scale (0–10)	Control	3% SCV	5% SCV	3% HCV	5% HCV	3% WCV	5% WCV	3% NHCV	5% NHCV
Typical odor	low–intense	6.75 ± 0.49 a	2.41 ± 1.36 b	2.03 ± 0.86 b	4.45 ± 1.19 cd	3.19 ± 1.16 bc	4.42 ± 0.87 cd	3.17 ± 1.33 bc	5.77 ± 0.60 ad	5.21 ± 0.79 d
Fishy/off-odor	low–intense	0.63 ± 0.67 a	5.10 ± 1.51 b	5.20 ± 1.95 b	3.41 ± 1.77 bc	3.80 ± 1.60 bc	3.32 ± 2.03 bc	3.32 ± 1.59 bc	1.86 ± 1.23 ac	1.83 ± 1.19 ac
After/off-taste	absent–very different	0.96 ± 0.97 a	5.21 ± 1.54 b	5.29 ± 1.35 b	4.03 ± 1.77 bc	3.77 ± 1.57 bc	2.75 ± 1.11 abd	2.95 ± 1.93 abd	1.23 ± 0.75 ad	1.79 ± 0.71 acd
Saltiness	low–intense	0.82 ± 1.08 a	3.05 ± 2.12 a	2.47 ± 1.85 a	2.65 ± 1.75 a	2.57 ± 1.92 a	1.65 ± 1.53 a	1.97 ± 1.58 a	1.88 ± 1.21 a	1.57 ± 1.42 a
Bite resistance	soft–firm	2.27 ± 1.21 a	4.21 ± 1.04 b	4.12 ± 1.72 ab	4.39 ± 1.08 b	4.11 ± 1.25 ab	3.64 ± 1.50 ab	2.95 ± 1.46 ab	3.55 ± 1.19 ab	3.88 ± 1.93 ab
Stickiness	low–intense	4.12 ± 3.14 a	4.39 ± 1.55 a	4.14 ± 1.92 a	3.62 ± 1.39 a	3.71 ± 2.08 a	4.34 ± 2.41 a	4.86 ± 1.85 a	3.55 ± 2.15 a	3.66 ± 2.01 a
Color	light–dark	1.27 ± 0.47 a	12.32 ± 0.72 1,bc	13.00 ± 0.00 1,c	4.55 ± 0.52 bd	5.36 ± 1.03 bce	2.41 ± 0.49 ad	3.23 ± 0.61 ade	3.91 ± 0.54 ade	4.91 ± 0.54 bce
	(given scale)	(yellow)	(green)	(green)	(yellow)	(yellow)	(yellow)	(yellow)	(yellow)	(yellow)
Acceptance		91% (10/11)	27% (3/11)	36% (4/11)	55% (6/11)	55% (6/11)	64% (7/11)	27% (3/11)	72% (8/11)	91% (10/11)

¹ Addition of 10 to emphasize the difference because samples were evaluated on a green scale, while all other samples were evaluated on a yellow scale. HCV, Honey C. vulgaris; NHCV, New Honey C. vulgaris; SCV, Smooth C. vulgaris; SD, standard deviation; WCV, White C. vulgaris. Different superscript letters indicate groups with significant difference (confidence interval 95%).

). Except for samples containing NHCV, a significant increase in fishy/off-odor, especially for the green pasta, was noted. After/off-taste significantly increased for samples containing SCV and HCV. Also, WCV samples had a more intense off-taste, but significance was not given due to large SDs. An influence of concentration on the attributes is recognizable but not significant. Further, all samples were perceived to have a higher bite resistance, but a significance to the control was only found for samples containing 3% SCV and 3% HCV. This partly contradicts the instrumental measurements. In terms of color, WCV samples were rated most closely to the control, followed by 3% NHCV. All other attributes did not change significantly, but acceptance decreased for all samples except 5% NHCV. A multiple logistic regression analysis (MLRA, Supplementary Data, Table S2) of acceptance showed that acceptance mainly decreased with decreasing typical odor (coefficient 0.622, p-value 0.006) and slightly with increasing salt taste (coefficient −0.487, p-value 0.014). The PCA, which explains 92% of the variance in the first two principal components, shows a similar result (Figure 3). Typical odor is in the opposite direction to all other attributes, and the control and NHCV samples are closest to it.

4. Discussion

The main determinant of pasta dough properties, such as strength, extensibility, and stability, is the wheat endosperm proteins, also known as gluten, but additional factors such as starch, non-starch polysaccharides, and non-gluten proteins can play a role [33]. When producing pasta dough, the addition of water makes the proteins rubbery and elastic. They become able to form strands and sheets via intermolecular bonds during mixing, and the formed gluten network helps to trap the starch granules in the pasta and holds the pasta shape during cooking [33,34]. Slight changes in the composition due to the addition of microalgae can, therefore, influence both the dough properties and the quality parameters (cooking time, water absorption, cooking losses) of pasta. This has already been shown when adding C. vulgaris and Spirulina to spaghetti. The authors showed that Spirulina, which stores glycogen, has no impact on the cooking time and cooking losses, while C. vulgaris, which stores starch, did [19]. The compositional results in Supplementary Table S1 show that the samples with a significant change in power consumption (5% NHCV), cooking time, and water absorption (3% and 5% NHCV) have one of the highest protein and DF contents (NHCV protein 0.7–1.2% higher, DF 0.3–0.5% higher), resulting in altered water binding abilities during dough processing and, consequently, altered dough viscosity. Due to semolina substitution by microalgae, the ratio of gluten to non-gluten protein, as well as of starch to non-starch polysaccharides, is supposed to change, too. Unfortunately, no data were available on the starch content of the DWS and microalgae or on the fiber content of SCV and WCV, but freeze-dried SCV (different batch) has been reported to contain approx. 28% starch [30], while the starch content of durum wheat semolina is approx. 70% [35]. Consequently, it can be assumed that the starch content is reduced in all samples containing algae, explaining the reduced cooking time, which is determined by starch swelling and gelatinization [19]. Another explanation would be that the dough network is discontinuous and not strong enough or elastic enough to prevent starch from swelling and gelatinizing before protein coagulation [25]. However, SEM cross-sections of the lasagna sheets (Figure 2) showed the opposite (continuous networks, as well as very fine networks, indicating high elasticity). In addition to the reduced cooking time, the NHCV samples showed reduced water absorption. According to Rodríguez De Marco et al. [25], water absorption is closely dependent on optimal cooking time, which could be one explanation. Another explanation may be that the increased protein and DF contents result in a denser network structure, which would make water absorption more difficult. However, apart from a higher proportion of compact structures (presumably DF) embedded in the protein network, which could explain the higher firmness (Table 4), no difference in the mesh size or thickness of the NHCV network could be detected compared to the control (Figure 2). In addition, all pasta samples can be considered to be of high quality due to the unchanged and very low cooking losses, indicating that all pasta samples have built up a stable protein and carbohydrate network from which no starch passes into the cooking water [19].

While the instrumentally measured color changes also corresponded approximately to the perception of the sensory panelists (WCV was most similar to the control, Table 5), the instrumental firmness measurement (Table 4) did not quite correspond to the perception of the panelists. This could be explained by the fact that the thickness of the lasagna sheets used for the instrumental measurement varied slightly from one production day to the next (extrusion gap had to be set manually). As the cooking time was kept constant at 10 min, the lasagna sheets showed different levels of doneness within a sample (product replicates) and between samples, resulting in not very meaningful results. The sensory evaluation of bite resistance was conducted on fusilli considering the optimal cooking time of each sample. Hence, algae addition seems to increase the bite resistance of all pasta samples, most likely due to higher protein and DF contents, as previously described for pasta with Nannochloropsis sp. addition [25]. MLRA and PCA analyses revealed that neither color, fishy/off-odor or after/off-taste, nor bite resistance had an impact on the acceptance of the pasta samples, which mainly depended on the typical odor. This is an important finding, as the fishy off-odor and taste accompanied by green microalgae were considered the main reasons for dislike in other studies [6,12,15,19,24]. However, it needs to be mentioned that SCV was the only green sample and showed the lowest ratings for typical odor and the highest ratings for fishy/off-odor and after/off-taste. Also, the non-green algae (HCV, WCV, NHCV) performed quite differently in terms of typical odor, fishy/off-odor, and after/off-taste, showing that the specific smell and aroma of each strain has a strong impact on the acceptance of the end product. The NHCV strain bred exclusively for the ProFuture project performed the best, showed unaltered acceptance at 5% concentration, and was only disliked by two more panelists at 3% concentration. This shows that certain characteristics, such as smell and taste, strongly need to be incorporated into breeding and strain selection, as well as in product development. However, the added value of the algae pasta is still low, making marketing more difficult. Indeed, protein and DF contents increased, but the nutritional claims stayed the same as for the control (low in fat, source of fiber, source of protein). To prove the marketability of the 5% NHCV sample, consumer studies are necessary in the future.

5. Conclusions

It was proven that physical pasta properties are dependent on the composition of the C. vulgaris strain used and that the impact of adding C. vulgaris to pasta on processing and quality parameters was low at 3 and 5% concentrations. The protein and DF contents could be slightly increased by adding microalgae to pasta formulations, resulting in increased firmness/bite resistance of the pasta, but they had no impact on the nutritional claims. This means that pasta with 5% Chlorella added may only be advertised as an algae-enriched product, which makes marketing more difficult. It was further proven that different C. vulgaris strains introduced different colors and flavors to the pasta, but sensory acceptance mainly depended on an unaltered typical odor rather than the absence of fishy/off-odors or after/off-taste. This means that the acceptance of microalgae pasta could be significantly increased by an improved form of presentation during sensory tasting (e.g., together with a sauce that masks the off-odor). This also allows the targeted use of the fishy aroma when developing new pasta dishes, such as pasta with vegan fish (sauce), vegan fish lasagna, and soups with garnish to improve the umami taste. However, to achieve an acceptable end product, algae strains should be selected not only based on the strain-specific flavor and odor properties but also on the product/dish to be developed. For example, fishy-tasting and green algae are better suited for dishes that are intended as fish alternatives or are green, e.g., vegan sushi and green soups, while algae with little off-flavor and taste, such as NHCV, could serve a wider range of products, e.g., baked goods, desserts, meat alternatives, and other non-green dishes.