Changes in Selected Quality Indices in Microbially Fermented Commercial Almond and Oat Drinks

Abstract

(1) Background: Interest in plant analogues for food of animal origin is increasing. There are some pro-healthy food ingredients, such as odd-chain, cyclic, and branched fatty acids, that are perceived to be characteristic for food of animal origin or fermented. The purpose of the present study was to determine whether commercial plant drinks can be valuable nutrient mediums for the multiplication of lactic acid bacteria and yeasts. The goal was also to determine their potential for the production of the above-mentioned groups of fatty acids; (2) Methods: Commercial almond and oat beverages were used to produce 16 new variants of fermented beverages using 3 strains of lactic acid bacteria and 5 strains of yeasts. The apparent viscosity, volatile compounds (e-nose), and fatty acids composition (GC-MS) were analyzed; (3) Results: After 48 h of fermentation, acidity increased in both types of drinks. The gelation of proteins in the majority of the almond beverages increased the apparent viscosity. The highest content of minor fatty acids was determined in oat beverages fermented by Lactiplantibacillus plantarum PK 1.1 and Kluyveromyces marxianus KF 0001 and in the almond beverage fermented by Candida lipolytica CLP 0001. Among the used strains, Yarrowia lipolytica YLP 0001 was found to be a major producer of aromas in both beverages.

1. Introduction

The increase in plant analogues of traditional types of food of animal origin is one of the main trends in the food market globally [1]. The plant-based (PB) food sector under study involves mostly meat, fish, and dairy analogues. This food market sector has attracted significant interest from institutions, researchers, economists, and entrepreneurs. The global market for vegetarian and vegan products was worth US$ 51 billion in 2016 and is still expanding [1]. In the European Union, based on the results of the ‘Smart Protein Project’, the value of the market of plant analogues of products of animal origin increased from EUR 2.4 billion in 2018 to EUR 3.6 billion in 2020 [2]. It is worth noting that this development, at least in part, took place in the harsh realities of the global COVID-19 pandemic. This rapid development of PB analogues is a strict consequence of the growing population of vegetarian, vegan, and other consumers avoiding foods of animal origin. For example, in America, the vegan population grew from nearly 4 million in 2014 to 19.6 million in 2017 (almost 5-fold growth) [3]. A recent study conducted by Bryant [4] showed that in the United Kingdom, the estimated number of vegans is ca 1–2% of the adult population, vegetarians ca 2–7%, and pescatarians 3–9%. People change their eating habits for ethical (animal welfare), ecological (care for the environment), and pro-healthy reasons (plant food reduces the risk of numerous diet-related diseases) [5].

Among PB analogues, milk and its by-products are of high consumer interest. From 2018 to 2020, the market value of PB beverages increased in the EU by ca 37% [2]. The majority of PB beverages are produced from seeds/grains such as almond, soy, coconut, rice, oat, various nuts, and other crops (such as hemp) [6]. In Europe, depending on the country, the highest sales were noted for oat-, almond-, and soy-based milk [2]. The intake of these products can be beneficial for the human body. For example, oat consumption, according to the high levels of beta-glucans present, can reduce the risk of type II diabetes and all-cause mortality [7,8]. This reduced risk of type II diabetes may be associated with the ability of oat beta-glucan to reduce the postprandial glycemic response [8]. Similarly, almond consumption lowers the LDL-cholesterol level, coronary heart disease risk, and associated cardiovascular disease expenditures; lowers diastolic blood pressure; enhances cognitive performance; improves heart rate variability under mental stress; slows facial skin aging from exposure to UV radiation; and supports colonic microbiota by increasing the concentrations of health-promoting colonic bioactive compounds [9].

Although PB products were used for allergic breast-fed infants nutrition decades ago [10], the problem of cow’s milk allergy and intolerance among infants and adults in modern societies is constantly worsening [11,12] and this is another important factor in the growing popularity of PB beverages. A study performed in the United States reported the prevalence of milk allergy to be 4.8% in the overall population aged 6 and over and 21.8% in children from 1 to 5 years of age [13]. These aspects, together with the increasing objections to industrial farming and the suffering of animals, have made PB beverages very attractive. However, it has been established that preparations of soya seed and nuts are also an important source of common allergens [14]. In this regard, fermented products are more tolerable/less dangerous for sensitive individuals [15,16].

During fermentation, biochemical modification and the creation of new valuable compounds occur [17,18]. For example, microbials such as Lactobacillus spp., Bifidobacterium spp., Streptococcus spp., Citrobacter spp., Enterobacter spp., Helicobacter spp., Klebsiella spp., Yersinia spp., Escherichia coli, and Yarrowia lipolytica are able to synthesize and accumulate odd-chain fatty acids (OCFAs) [19], branched-chain fatty acids (BCFAs), and cyclic fatty acids (CFAs) [20,21]. For example, the ratio of CFAs to unsaturated fatty acids is the relevant parameter associated with membrane rigidification in Lactococcus lactis TOMSC161 cells and cell resistance to freeze-drying and storage [22]. Apart from the usefulness of these compounds for microflora, they also seem to be important for decreasing/preventing many health problems. For example, higher concentrations of blood-circulating pentadecanoic and heptadecanoic acids (representatives of OCFAs) are associated with lower risks of cardiometabolic diseases, lower mortality, and anti-inflammatory properties. They also counteract anemia, dyslipidemia, and fibrosis in vivo, and pentadecanoic acid is proposed as a potential essential fatty acid [23]. BCFAs have anti-diabetic, anti-inflammatory, and anti-cancerogenic activity [24,25,26]. Similarly, animal studies indicate that sterculic acid (one of the main representatives of CFAs) can prevent metabolic syndrome development and have an anxiolytic-like effect [27,28]. In general, plant materials are the main source of even-chain carbon fatty acids. In this context, even a small accumulation of OCFAs, BCFAs, or CFAs should be profitable for the pro-healthy quality of fermented products. The estimated daily intake of these fatty acids from the typical diet in Western countries is approximately 1.5 g, including approximately 1.0 g odd, 0.5 g branched, and 12 mg of cyclic fatty acids [17].

The present study aimed to determine whether commercial almond and oat drinks, differing in their protein, fat, and carbohydrate contents, are valuable nutrient mediums for the growth and multiplication of three strains of lactic acid bacteria and five strains of food-grade yeasts and their influence on selected quality indices.

2. Materials and Methods

2.1. Plant Material and Reagents

The plant material used in this study consisted of two commercially available beverages made of almond and oat seeds. The oat drink (“Vemondo” brand, Germany) was made of water, oat (15%), sea salt and calcium carbonate, with a final proximate composition of fat, carbohydrates, and protein of 0.5, 9.7, and 0.4 g/100 mL, respectively. The almond drink (“Auchan” brand, France) was composed of water, almonds (7%), cane sugar, natural aroma, and locust bean gum, with a proximate composition of fat, carbohydrates, and protein of 4.1, 4.9, and 1.7 g/100 mL, respectively. Among the carbohydrates, in both drinks, sugars prevailed, reaching values of 4.2 (almond) and 6.1 g/100 mL (oat).

Reagents: Helium for GC (99.999% purity) was purchased from Eurogaz-Bombi (Olsztyn, Poland). Analytical-grade solvents and reagents such as hexane, methanol, chloroform, sodium and potassium bases, and sulphuric acid were purchased from Chempur (Piekary Śląskie, Poland). Standard heptadecanoic acid was purchased from Merck (Darmstadt, Germany).

2.2. Experimental Part

2.2.1. Fermentation

Both PB beverages were subjected to fermentation with the use of 8 microorganisms:

• Lactobacillus delbrueckii ssp. bulgaricus 27/23.
• Lactiplantibacillus plantarum ATCC 8014.
• Lactiplantibacillus plantarum PK 1.1.
• Candida antarctica CAN 0001.
• Torula casei TCS 0001.
• Yarrowia lipolytica YLP 0001.
• Kluyveromyces marxianus KF 0001.
• Candida lipolytica (Yarrowia lipolytica) CLP 0001.

The strain numbers correspond to which strain each sample was inoculated with. Control samples were samples of PB beverages not subjected to the fermentation process.

The used strains came from the strain collection of the Department of Industrial and Food Microbiology, Faculty of Food Sciences of the University of Warmia and Mazury in Olsztyn (Poland). In total, 150 mL of each beverage was inoculated with 3 mL of an active culture of the strains (2% by volume) grown in optimal media (MRS broth sticks or yeast liquid wort). The microbial cell density was ca 2.5 × 10⁵ for bacteria and ca 1.0 × 10⁴ for yeasts. Inoculations were made in 3 replications. Fermentation was conducted for 48 h at a temperature of 37 °C for L. bulgaricus or 30 °C for the other strains.

The cell density of LABs before and after fermentation was determined by the MRS-agar surface method (Merck, Darmstadt, Germany). Incubation was anaerobic and carried out at 30 °C for 72 h. The number of yeasts was determined by the YGC-agar surface method (Merck, Darmstadt, Germany). Incubation was carried out aerobically at 25 °C for 96 h. The number of lactic acid bacteria and the number of yeasts were also determined in the control samples (drinks without the addition of test strains). The limit of detection for the used plate count methods was below 10 CFU per mL.

Immediately after fermentation, the titratable acidity, pH, and apparent viscosity were determined. The remaining sample variant was frozen for further volatiles screening by E-nose identification or lyophilized for fatty acids determination.

2.2.2. Measurements of PB Beverage Acidification

The acidity of the obtained PB beverages was assessed by pH measurement and titration assays. pH was measured with the use of an HI 9125 HANNA Instruments (Woonsocket, RI, USA) device, which was standardized with pH 4 and 7 buffers before use. The titratable acidity was measured according to the method described by Matela et al. [29]. Briefly, 10 g of fermented product was weighed and dissolved in 30 mL of deionized water. Each sample, after the addition of a few drops of indicator (alcoholic phenolphthalein solution), was titrated with the use of 0.1 M NaOH solution. The titratable acidity (TA) was calculated on lactic acid using equation as follows:

$$TA=\frac{V_{NaOH}·c_{NaOH}·meq}{m}·100 \left[\%\right]$$

where:

V_NaOH—volume of used NaOH for titration [mL];

c_NaOH—molar concentration of NaOH [mol/L];

meq—milliequivalent factor for lactic acid (1 mL 0.1 M NaOH correspond to 0.0901 g of lactic acid);

m—weighed mass of sample [g].

2.2.3. Measurement of PB Beverage Viscosity

The apparent viscosity was determined in at least triplicates for the control and fermented beverages. Measurements were performed at 25 °C in a rotational DV-II+ Pro Extra viscometer (Brookfield, Middleboro, MA, USA) with a concentric cylinder using a spindle controlled by Rheocalc V3 software from the same manufacturer. About 70 mL of sample was placed in the stationary cup. The apparent viscosity was obtained at a 200 rpm spindle speed and expressed in mPa·s.

2.2.4. Measurement of the Fatty Acids Contents in the PB Beverages

For analysis of the fatty acids composition, 100 mL of each PB beverage was frozen and subjected to lyophilization in an Alpha 2–4 LSC BASIC freeze-dryer (Martin Christ, Osterode am Harz, Germany). In the next stage, lyophilizate was subjected to lipid extraction with the use of a chloroform-methanol (2:1, v/v) mixture. After solvent addition, each sample was homogenized by T25 Ultra Turrax (IKA, Werke, Germany). Homogenates were subjected to ultrasonic extraction with the use of a VCX750 extractor (Sonics & Materials Inc., Newtown, CT, USA) with the following conditions: frequency of 20 kHz, amplitude of 40 µm, and time of 10 min. After sonication, samples were centrifuged, and the solid residue was subjected to 2-fold extraction with the above-mentioned mixture by shaking and centrifugation. The combined supernatants were transferred to the separation funnels through a paper filter. Next, 0.58% sodium chloride solution was added, and the mixture was left for 18 h in the dark. After, the bottom fraction (chloroform and lipids) was separated. The upper fraction was washed 3-fold with chloroform. The chloroform fractions were combined, dried with anhydrous sodium sulphate, and the solvent was evaporated with an R-210 rotary vacuum evaporator (Büchi, Flawil, Switzerland) at 45 °C.

For methylation, 50 mg of the extracted lipids was weighed. Methyl esters were prepared by the incubation of the samples with 2 mL of a chloroform-methanol-sulphuric acid (100:100:1, v/v/v) mixture and internal standard (heptadecanoic acid) solution. After incubation, zinc powder was added (more than 20 mg per sample—in extent for sulphuric acid neutralization) and solvents were evaporated under a nitrogen stream. Methyl esters were dissolved in hexane and analyzed by the GC–MS technique with the use of a GC–MS QP2010 PLUS system (Shimadzu, Kyoto, Japan) equipped with a BPX70 (25 m × 0.22 mm × 0.25 μm) capillary column (SGE Analytical Science, Victoria, Australia). Helium was used as a carrier gas at a flow rate of 1.3 mL/min. The injector temperature was set at 230 °C. The column temperature was programmed as follows: a subsequent increase from 150 to 180 °C at a rate of 10 °C/min, to 185 °C at a rate of 1.5 °C/min, to 250 °C at a rate of 30 °C/min, and then a 10 min hold. The temperature of the GC–MS interface and ion source were set at 270 and 240 °C, respectively. The ionization energy was set at 70 eV. The range of the recorded total ion current (TIC) was set at 50–500 m/z. Fatty acid methyl esters were identified by comparing their retention times to available standard methyl esters and their mass spectra with mass spectral libraries (NIST14 library, Shimadzu, Kyoto, Japan).

2.2.5. E-Nose Analysis of the Volatile Compounds of the PB Beverages

Frozen samples of the PB beverages were carefully defrosted. Volatile analysis was performed according to the method described in the authors’ previous study [30] with slight modifications. The equipment used in this analysis included the Heracles II E-nose (AlphaMOS, Toulouse, France) equipped with an automatic sampling system and two columns working in parallel mode: a non-polar column MXT-5 and a slightly polar column MXT-1701 (both with a 10 m length and 180 µm diameter, connected with two flame ionization detectors (FIDs)). In total, 1 mL of each PB beverage was placed in a 20 mL headspace vial and sealed with a cap with Teflon septa. Before analysis, vials were incubated at 40 °C for 300 s and shaken at 500 rpm. After incubation, the silicone septa were pierced by the syringe, and 2.5 mL of the headspace was injected to the gas chromatographic system. The separation conditions were as follows: injector temperature of 200 °C, carrier gas (hydrogen) flow rate of 1 mL/min, oven temperature of 55 °C, and FID temperature of 270 °C. Each sample was injected in triplicate. Calibration was performed with the use of an alkane mixture solution (from n-hexane to n-hexadecane), which was used for conversion of the retention time into Kovats retention indices and to identify the volatile compounds using AroChemBase (AlphaMOS, Toulouse, France).

2.2.6. Statistical Analysis

Results were statistically analyzed using Statistica 12.5 software (TIBCO, Palo Alto, CA, USA). The differences between the samples were determined using the analysis of variance (ANOVA) with the Tukey test at the p ≤ 0.05 significance level. The mentioned software was also used for PCA analysis of the results obtained with the E-nose.

3. Results and Discussion

3.1. Oat and Almond Drinks as Mediums for the Growth of the LAB and Yeast Strains Used

The main nutritive component of both plant drinks was carbohydrates (9.7 g/100 mL in oat and 5.0 g/100 mL in almond), with a predominance of sugars (4.2 and 6.1 g/100 mL for almond and oat drinks, respectively). This was a positive premise for fermentation since simple sugars can be easily used as a carbon source. However, the analyzed drinks differed in their fat and protein contents. The contents of both of these components were extremely low in the oat drink (only 0.4 g of protein and 0.5 g of lipids in 100 mL of the oat drink) while the composition of the almond drink (4.1 g and 1.7 g/100 mL for fat and protein contents, respectively) was more similar to non-skimmed cow milk, which usually contains ca 3.6, and 3.2 g/100 mL of fat and protein, respectively [31].

After inoculation by the LAB strains, the initial count was 1.0–2.2 × 10⁴ CFU/mL of the prepared beverages (

Table 1. Lactic acid bacteria (LAB) and yeast counts (CFU/mL) of plant beverages before and after fermentation.

Sample	Oat Beverage				Almond Beverage
	Inoculation Stage		After Fermentation (48 h)		Inoculation Stage		After Fermentation (48 h)
	LAB	Yeast	LAB	Yeast	LAB	Yeast	LAB	Yeast
control	<10	<10	<10	<10	<10	<10	<10	<10
1	1.2·104	<10	1.0·109	<10	1.0·104	<10	3.1·109	<10
2	1.0·104	<10	2.4·109	<10	2.2·104	<10	2.6·109	<10
3	1.6·104	<10	2.5·109	<10	1.8·104	<10	4.8·109	<10
4	<10	2.0·103	<10	4.7·107	<10	3.4·103	<10	4.5·107
5	<10	6.5·103	<10	7.6·107	<10	4.0·103	<10	8.4·107
6	<10	2.0·103	<10	7.0·106	<10	2.5·103	<10	1.4·107
7	<10	9.8·102	<10	1.2·108	<10	1.2·103	<10	7.4·107
8	<10	3.9·103	<10	1.4·106	<10	2.6·103	<10	8.9·105

Control—beverage not fermented; sample numbers correspond with the microorganisms described in the experimental section; the fermentation temperature was 37 °C for sample 1 and 30 °C for samples 2–8.

). After 48 h of fermentation, a significant increase was observed in all LAB strains, resulting in a final concentration of 1.0–4.8 × 10⁹ CFU/mL. Almond beverages were characterized by a slightly higher density of LAB counts. L. plantarum strains have been isolated from many food products, including cheese, meat, fish, fermented vegetables, and fruits, and selected strains are used as probiotics. L. bulgaricus, together with Streptococcus thermophilus, is a typical yoghurt microbiota [32]. In accordance with the requirements of the FAO/WHO Codex Alimentarius standard [33], the number of characteristic microflorae in classical yoghurt must be at least 10⁷ CFU in 1 g of the product throughout the shelf life. The LAB count density in the prepared plant beverages was similar to that found in freshly prepared, fermented milk-based products [34]. In general, in a variety of retail fermented foods, the amount of live LAB ranges between 10⁵ and 10⁹ CFU/g or mL [34]. This shows that both plant matrices were valuable nutrient sources for bacteria growth and multiplication.

After inoculation by the yeast strains, their initial concentration ranged from 9.8 × 10² to 6.5 × 10³ CFU/mL of prepared beverages (Table 1). In the fermented beverages, a significant increase was noted in all strains, reaching a final density that ranged from 8.9 × 10⁵ (almond drink fermented by C. lipolytica) to 1.2 × 10⁸ CFU/mL (an oat drink fermented by K. marxianus). Generally, K. marxianus showed the highest count increase (close to 5 log CFU/mL) while C. lipolytica showed the lowest (close to 2 log CFU/mL). For the other yeast strains, the observed increase was close to 4 log CFU/mL. Kluyveromyces species are components of the yeast content of kefir and kefir grains [35] and of Serro Minas cheese [36]. These strains possess the unique ability (for yeasts) to ferment lactose into lactic acid [37], but K. marxianus also utilizes other sugar substrates (i.e., hemicellulose hydrolysates, xylose, l-arabinose, d-mannose, galactose, maltose, sugar syrup molasses, and cellobiose) [38] and can grow in olive oil as a carbon source [39]. In contrast, the low level of multiplication of C. lipolytica and Y. lipolytica in both used plant beverages is difficult to explain. Y. lipolytica and C. lipolytica are the teleomorph and anamorph names of the same species [40]. These species are widespread in nature and can be found in a number of foods, such as sites contaminated with oil, plants, brie, feta, cheddar, and meat [41]. It was recently shown that when cassava residues with a low lipid content were used as a carbon source, Y. lipolytica QU69 produced 22.3% of protein and 9.4% of lipids in dry biomass, respectively. In this study, at a low lipid content in a nutrition medium (approximately 1%), the strain prioritized cell growth while at a high content of lipids (approximately 20%), the priority was for lipase production, which resulted in a 55% decrease in the lipid content in a fermented biomass of cotton seeds [41].

3.2. Acidity Changes after the Fermentation of the Used PB Beverages

The acidity values of the control and fermented PB beverages are presented in

Table 2. Acidity and apparent viscosity of the control and fermented (48 h) beverages.

Sample	Oat Beverage									Almond Beverage
	Titratable Acidity (% of Lactic Acid)			pH			Apparent Viscosity (mPa × s)			Titratable Acidity (% of Lactic Acid)			pH			Apparent Viscosity (mPa × s)
control	0.08	±	0.00 a	7.05	±	0.01 h	20.53	±	0.55 e	0.07	±	0.00 a	7.99	±	0.01 i	29.60	±	0.22 b
1	0.76	±	0.00 h	2.80	±	0.01 a	19.10	±	0.64 ab	0.53	±	0.02 e	3.42	±	0.01 d	101.25	±	0.10 h
2	0.53	±	0.00 g	2.82	±	0.01 a	19.33	±	0.55 abc	0.72	±	0.00 f	3.13	±	0.01 a	97.80	±	0.00 g
3	0.47	±	0.00 e	2.90	±	0.01 c	18.44	±	0.43 a	0.73	±	0.00 f	3.23	±	0.01 b	90.63	±	0.60 f
4	0.11	±	0.00 b	4.96	±	0.01 f	19.52	±	0.61 bcd	0.07	±	0.01 a	6.51	±	0.00 h	23.40	±	0.35 a
5	0.17	±	0.01 c	4.51	±	0.01 e	19.72	±	0.52 bcde	0.19	±	0.00 b	5.39	±	0.01 g	60.51	±	0.16 d
6	0.50	±	0.01 f	2.85	±	0.01 b	19.44	±	0.43 abcd	0.79	±	0.01 g	3.28	±	0.00 c	108.84	±	0.26 i
7	0.22	±	0.01 d	4.43	±	0.01 d	19.94	±	0.38 cde	0.22	±	0.00 c	5.31	±	0.01 f	62.43	±	0.21 e
8	0.12	±	0.00 b	5.45	±	0.00 g	20.36	±	0.26 de	0.37	±	0.00 d	4.32	±	0.00 e	48.13	±	0.16 c

Data in the column marked with the same letter are not statistically different; control—beverage not fermented; sample numbers correspond with the microorganisms described in the experimental section.

. The titratable acidity of the control beverages showed a very small concentration of lactic acid equivalent (0.07% and 0.08%) while pH was close to neutral (oat drink with pH 7.05) or slightly basic (almond drink with pH 7.99). After 48 h of fermentation with the LAB strains, the titratable acidity significantly increased in both beverages to values of 0.47–0.76% and 0.53–0.73% of lactic acid equivalent in the oat and almond drink, respectively. The final pH of the produced beverages was below pH 3.5 in all samples, with a higher concentration of hydrogen ions in the oat drinks (the determined pH ranged from 2.80–2.90). In the almond beverages, the pH values were higher and reached values of 3.13–3.42. This indicates a higher concentration of dissociated acids (such as lactic, citric, and succinic) in the fermented oat beverages. For the almond drinks fermented by L. plantarum strains, the low pH was accompanied by significantly higher values of titratable acidity, which also indicates a greater content of undissociated acids, such as free amino acids and free fatty acids.

Acidification to pH below 4.5 is a typical symptom of LAB fermentation and achievement of a final lactic acid bacteria density higher than 8 log CFU/g [42]. L. plantarum is especially responsible for the high acidity of fermented products [43]. Based on LAB bacterium and the type of fermentation (homo- or hetero-), the accumulation of only lactic acid or lactic and acetic acid occurs. L. bulgaricus is an example of an aerobic-to-anaerobic homofermentative bacterium, so lactic acid is the major end product of fermentation produced via the EMP pathway [44,45]. In turn, L. plantarum is a facultative heterofermentative bacterium that produces lactic, acetic, and formic acids under certain conditions and selective substrates via the Dickens–Horecker and Entner Doudoroff metabolic pathways [45]. Both LAB species used in this study can ferment sugars found in plant materials, such as glucose, fructose, and xylose. L. bulgaricus is commonly used in the production of yoghurts and some types of ripening cheeses. It has also been successfully used in the production of fermented plant drinks [42]. It is part of the microflora of many food and feed products, including dairy, meat, fish, and fermented vegetable products. Strains of this species are part of the starter and protective cultures used in the food industry [46]. This species is also a natural microbiota of humans and animals [47].

Changes in the titratable acidity and pH of the fermented-by-yeast beverages were more varied (Table 2). The lowest increases in the titratable acidity (up to 0.22% of lactic acid equivalent) were found for the strains of C. antarctica and T. casei in both beverages, K. marxianus in an almond drink, and C. lipolytica in an oat drink. For other strains, the acidity increase was significantly higher, reaching an upper limit of 0.79% and 0.50% of lactic acid equivalent in the almond and oat beverages fermented by Y. lipolytica, respectively. Moreover, the hydrogen ion concentration in both of these beverages was the highest (oat drink with pH 2.85 and almond drink with pH 3.28). These values were close to the values found in drinks prepared using LAB strains. For other fermented-yeast beverages, the pH values ranged from 4.32 to 6.51.

Yeasts, particularly C. lipolytica and Y. lipolytica, are efficient producers of lipases and proteases [48,49], enabling more advanced and faster hydrolysis of proteins and lipids into amino acids and fatty acids, which together increase the titratable acidity. The activity of lipases is particularly important since the ability to degrade triglycerides is an essential process in ensuring energy homeostasis and the availability of precursors for membrane lipid synthesis [50]. In the case of high acidity (measured as low pH), the production of organic acids that are able to dissociate could also be considered. Recent studies have shown that Y. lipolytica is an efficient producer of citric acid [51,52,53] and succinic acid [54]. The balance of dissociated to undissociated acids (e.g., citric) allows the yeast to survive at a low pH (the amount of undissociated citric acid increases and can permeate the cell wall, altering the internal pH of the microorganisms) [33]. In a highly acidic environment (such as in the beverages fermented in the current study by Y. lipolytica), amino acid molecules also exist in the form of positive NH₃⁺ ions.

Among the analyzed fermented plant drinks, the highest titratable acidity changes were determined for almond beverages. As mentioned earlier, this type of beverage was characterized by the highest fat content (4.1 g/100 mL) and relatively high protein content (1.7 g/100 mL). It appears that amino acids and fatty acids produced in these fermented beverages were not fully utilized at once for microbial growth, and they accumulated in the environment.

3.3. Viscosity of PB Beverages

The results of the apparent viscosity of the control and fermented beverages are presented in Table 2. Control samples were only slightly differentiated in this regard, reaching values of 20.53 (oat drink) and 29.60 mPa·s (almond drink). None of the LAB or yeast strains used for the fermentation of oat drinks significantly changed the viscosity of the obtained beverages (the noted variation was from 18.44 to 20.36 mPa·s). Much larger changes were observed in the fermented almond beverages. In the majority of samples, the apparent viscosity significantly increased, with the top levels found for samples fermented by LAB and Y. lipolytica strains (values from 90.63 to 108.84 mPa·s). Only the sample fermented by C. antarctica was thinned after fermentation (23.40 mPa·s). The apparent viscosity results for the almond beverages were highly correlated with the titratable acidity (r = 0.92; p < 0.05) and pH (r = −0.89; p < 0.05).

Similar results were recently reported by Espirito-Santo et al. [55] during the fermentation of gruels made of rice and soy milk by selected LABs. In this study, the apparent viscosity of rice milk was significantly reduced while an opposite effect was found for fermented soy milk. According to the cited authors, in addition to the composition of fermented plant material, the viscosity of the final product also depends on several factors such as the kinetics of acidification, the amylolytic and proteolytic activities of the fermenting microflora and the exopolysaccharides produced (or not) by them, the degree of protein gelation, and protein–starch interactions. For the beverages prepared in the current study, a key factor was probably the concentration of the protein and pH, which determine the gel creation. For the main storage proteins of oat (avenins, 12S storage globulin) and almond seeds (12S storage globulin), their isoelectric point is close to 6.5 (almond) and 6.0–8.0 (oat) [56,57]. This indicates that in samples with lower pH values than the isoelectric point, proteins were found to be unfolded, with amino acids ready to form hydrogen or dipole bonds with surrounding water molecules (the almond 12S globulin contains ca 20% of charged residues and additionally ca 45% of polar amino acids without charge—the data for the protein is coded as P14812 in the UniProtKB base). For the almond drink fermented by C. antarctica, the final pH was 6.51, which is near the value of the isoelectric point of storage proteins (there was visible protein precipitation in this sample). However, in oat beverages, the protein concentration (0.4 g/100 mL) seems to be too low to form a gel structure.

In general, each interaction between the amino acid and water immobilizes more water molecules for ion–dipole interaction. Qiao et al. [58] concluded that negatively charged amino acids were more beneficial for protein solubility, followed by positively charged amino acids, and, finally, charge-neutral amino acids. This indicates that for stable protein gel formation, there is a need for the calculation of charge-neutral, charge-positive, and charge-negative amino acids in each protein. For almond and oat proteins, almonds are much richer in both charge-positive and charge-negative amino acids, so this characteristic also confirms the possibility of creating a stable gel with enhanced viscosity [56,57].

3.4. Fatty Acid Composition Changes following the Fermentation of PB Beverages

The major fatty acids of the control and fermented PB beverages were characteristic of the raw plant material used (

Table 3. Fatty acids content in the control and fermented (48 h) oat beverages.

FA	Control	1	2	3	4	5	6	7	8
Major fatty acids (g/100 g of lipids)
C16:0	15.3 bcd	13.9 abc	13.8 abc	16.0 cd	16.6 d	13.5 ab	14.9 bcd	17.0 d	12.5 a
C18:0	1.51 a	1.15 a	1.09 a	1.41 a	1.17 a	1.02 a	1.17 a	1.42 a	1.35 a
C18:1	26.6 abc	24.5 ab	23.3 a	31.2 c	26.6 abc	22.6 a	27.0 abc	29.6 bc	30.2 c
C18:2	29.5 abc	27.4 ab	26.5 a	32.7 c	31.2 bc	26.8 a	29.9 abc	33.1 c	26.0 a
C18:3	1.273 a	1.005 a	0.901 a	1.083 a	0.915 a	0.868 a	0.991 a	1.139 a	0.846 a
total content of major FAs	74.3 ab	68.0 a	65.6 a	82.4 b	76.4 ab	64.7 a	73.9 ab	82.2 b	70.9 ab
Minor fatty acids (mg/100 g of lipids)
C14:0	384 c	264 ab	260 ab	300 abc	329 bc	250 ab	286 abc	314 abc	222 a
methyl 9-methyltetradecanoate	nd a	nd a	nd a	nd a	nd a	nd a	nd a	nd a	10.2 b
C15:0	46.6 a	27.6 a	29.1 a	36.8 a	64.1 a	28.8 a	23.6 a	29.1 a	27.6 a
methyl (z)-5-dodecenoate	21.2 bc	19.0 bc	15.5 bc	25.3 c	nd a	13.1 b	17.6 bc	21.0 bc	15.6 bc
C16:1	314 b	208 a	206 a	233 a	256 ab	207 a	212 a	277 ab	266 ab
methyl 18-methylnonadecanoate	101 b	70 b	70 b	95 b	nd a	59 ab	71 b	82 b	72 b
C20:1	693 a	548 a	494 a	681 a	449 a	418 a	562 a	621 a	449 a
C20:2	nd a	nd a	20.5 ab	15.4 ab	nd a	nd a	24.4 b	20.8 ab	18.4 ab
9-(3,3-dimethyloxiran-2-yl)-2,7-dimethylnona-2,6-dien-1-ol	92 a	112 a	146 ab	288 b	170 ab	81 a	156 ab	202 ab	210 ab
C22:0	60.9 a	42.9 a	33.5 a	37.8 a	21.2 a	40.0 a	34.8 a	50.3 a	37.0 a
methyl (11r,12r,13s)-(z)-12,13-epoxy-11-methoxy-9-octadecenoate	205 a	227 a	273 ab	561 b	308 ab	162 a	309 ab	338 ab	299 ab
methyl 8-(3-octyl-2-oxiranyl)octanoate	nd a	60 b	107 d	75 bc	nd a	nd a	98 cd	nd a	nd a
methyl 5,13-docosadienoate	71.2 ab	85.8 b	41.9 ab	66.3 ab	nd a	72.6 ab	73.5 ab	nd a	44.7 ab
methyl 4-[2-[[2-[[2-[(2-pentylcyclopropyl)methyl]cyclopropyl]methyl]cyclopropyl]methyl]cyclopropyl]butanoate	nd a	nd a	215 ab	249 b	nd a	nd a	nd a	296 b	151 ab
methyl 7,11,14-eicosatrienoate	nd a	nd a	202 b	255 b	nd a	213 b	198 b	328 b	197 b
methyl octadec-17-ynoate	nd a	nd a	114 ab	148 ab	nd a	nd a	nd a	249 b	117 ab
methyl (z)-5,11,14,17-eicosatetraenoate	nd a	nd a	72 a	145 ab	nd a	399 c	125 a	281 bc	76 a
methyl 8-[2-((2-[(2-ethylcyclopropyl)methyl]cyclopropyl)methyl)cyclopropyl]octanoate	nd a	nd a	68.3 ab	138.9 bc	nd a	nd a	80.0 ab	234.3 c	82.4 ab
methyl 10,13,16-docosatrienoate	153 bc	86 ab	151 bc	138 bc	nd a	101 ab	125 abc	250 c	142 bc
total content of minor FAs	2143 a	1751 a	2519 abc	3490 bc	1598 a	2044 a	2395 ab	3593 c	2439 ab

nd—not determined; data in the row marked with the same letter are not statistically different; control—beverage not fermented; sample numbers correspond to the microorganisms described in the experimental section.

and

Table 4. Fatty acids content in the control and fermented (48 h) almond beverages.

	Control	1	2	3	4	5	6	7	8
Major fatty acids (g/100 g of lipids)
C16:0	6.93 c	6.69 c	6.99 c	6.72 c	6.94 c	6.41 bc	5.96 ab	5.72 a	6.88 c
C18:0	2.27 b	2.23 b	2.34 b	2.26 b	2.26 b	2.05 ab	1.96 ab	1.74 a	2.27 b
C18:1	57.3 d	50.1 ab	52.1 bc	49.5 ab	51.6 abc	49.3 ab	48.1 a	50.4 ab	55.2 cd
C18:2	19.0 b	18.2 b	19.1 b	18.1 b	19.0 b	18.2 b	16.5 a	16.7 a	19.1 b
total content of major FAs	85.5 d	77.2 ab	80.5 bcd	76.6 ab	79.8 bc	75.9 ab	72.5 a	74.6 a	83.4 cd
Minor fatty acids (mg/100 g of lipids)
C14:0	33.0 c	27.0 abc	29.1 abc	26.9 abc	28.0 abc	24.4 a	26.4 ab	31.0 bc	30.6 abc
C15:0	24.5 a	22.8 a	21.6 a	22.3 a	22.9 a	21.6 a	20.4 a	22.8 a	22.7 a
methyl (z)-5-dodecenoate	11.8 a	8.0 a	9.6 a	8.8 a	7.2 a	9.2 a	8.7 a	10.4 a	12.1 a
C16:1	588 bc	526 ab	553 abc	536 abc	541 abc	492 a	518 ab	498 a	609 c
C17:1	128 c	119 bc	122 bc	122 bc	118 bc	112 ab	113 ab	103 a	127 c
methyl 8-(2-hexylcyclopropyl)octanoate	62.8 b	59.4 b	48.2 ab	49.1 ab	41.6 ab	29.0 a	37.3 ab	29.7 a	51.4 ab
2-cis,cis-9,12-octadecadienyloxyethanol	41.3 a	49.6 a	46.8 a	46.3 a	48.8 a	41.0 a	41.0 a	36.9 a	46.3 a
C18:3	18.0 a	17.5 a	20.7 a	23.4 a	22.8 a	22.8 a	14.5 a	17.0 a	18.8 a
methyl 18-methylnonadecanoate	51.5 a	57.1 a	71.0 a	68.0 a	64.1 a	59.4 a	52.6 a	48.3 a	57.2 a
C20:1	36.5 abc	34.2 abc	43.9 c	40.8 bc	37.7 abc	31.3 abc	29.9 ab	25.8 a	37.8 abc
9-(3,3-dimethyloxiran-2-yl)-2,7-dimethylnona-2,6-dien-1-ol	376 abc	514 bc	490 bc	463 abc	380 abc	198 a	311 abc	252 ab	531 c
methyl (11r,12r,13s)-(z)-12,13-epoxy-11-methoxy-9-octadecenoate	126 ab	201 b	201 b	197 b	158 ab	85 a	124 ab	93 ab	191 ab
9-octadecene, 1,1-dimethoxy-, (z)-	16.8 a	33.6 ab	41.6 ab	40.1 ab	39.3 ab	26.2 a	46.4 ab	35.9 ab	63.1 b
methyl 8-(3-octyl-2-oxiranyl)octanoate	nd a	nd a	nd a	nd a	nd a	nd a	nd a	76.8 a	nd a
methyl 5,13-docosadienoate	34.3 a	33.7 a	34.9 a	33.4 a	33.8 a	28.9 a	26.7 a	91.6 a	96.3 a
methyl 9-oxooctadecanoate	68 a	97 a	87 a	90 a	116 a	84 a	126 a	90 a	119 a
methyl 4-[2-[[2-[[2-[(2-pentylcyclopropyl)methyl]cyclopropyl]methyl]cyclopropyl]methyl]cyclopropyl]butanoate	163 a	239 a	203 a	210 a	219 a	108 a	193 a	168 a	276 a
methyl 5-[(1s,2s)-2-undecylcyclopropyl]pentanoate	430 a	451 a	364 a	372 a	329 a	146 a	253 a	199 a	454 a
methyl (z)-5,11,14,17-eicosatetraenoate	35.1 bc	42.8 c	36.8 bc	33.9 bc	29.1 bc	12.1 ab	26.8 bc	nd a	34.5 bc
methyl 8-[2-((2-[(2-ethylcyclopropyl)methyl]cyclopropyl)methyl)cyclopropyl]octanoate	40.7 a	46.3 a	37.2 a	38.4 a	34.7 a	16.9 a	24.5 a	51.6 a	48.8 a
methyl 10,13,16-docosatrienoate	23.5 a	48.4 ab	46.6 ab	43.3 ab	62.1 ab	43.6 ab	48.4 ab	41.7 ab	76.0 b
total content of minor FAs	2309 abc	2628 bc	2507 abc	2464 abc	2334 abc	1590 a	2041 abc	1923 ab	2903 c

nd—not determined; data in the row marked with the same letter are not statistically different; control—beverage not fermented; sample numbers correspond to the microorganisms described in the experimental section.

). Both beverages contained mostly 18-carbon chain fatty acids. Almond fat had the highest content of oleic acid, with a value close to 48–57 g/100 g of lipids, followed by 16–19 g of linoleic acid per 100 g of lipid. Oat fat was mainly composed of comparable shares of linoleic and oleic acids, reaching ca. 49–64 g per 100 g of lipids. The palmitic acid content ranged from 13–16 g/100 g in oat lipids and 6–7 g/100 g in almond lipids. Similar shares of these fatty acids in the used plant materials were previously presented by Kouřimská et al. [59] and Zamany et al. [60].

In all fermented almond beverages, the total content of major fatty acids decreased from 85.5 g/100 g in the control sample to the lowest content of 72.5 g/100 g in a sample fermented by Y. lipolytica (the highest decrease was noted for oleic acid). For the oat beverages, the changes were more differentiated, since in two samples (fermented by L. plantarum PK 1.1. and K. marxianus), there was an increase in the major fatty acids (to ca 82 g/100 g) in relation to the control sample (ca 73 g/100 g). The total content of minor fatty acids (with contents below 1 g/100 g of lipids) varied from ca 1600 mg/100 g of lipids in an almond beverage fermented by C. antarctica and oat beverage fermented by T. casei to ca 3600 mg/100 g of lipids in an oat beverage fermented by K. marxianus. The majority of fermented samples had more of these minor fatty acids than both control samples. Among the minor fatty acids, branched, odd-chain, and cyclic fatty acids represented from 25.1% (oat control beverage) to 73.9% (almond beverage fermented by L. bulgaricus). For the fermented oat beverages, all variants were characterized by enhanced shares of these unique fatty acids (increase in the range of 5–1547 mg/100 g of lipids). In contrast, for the fermented almond beverages, the contents of these compounds increased in the samples fermented by LAB strains and by C. lipolytica. For other strains, a decrease was noted in comparison to the content determined in the control sample.

In general, LABs utilize various sugars as carbon sources while yeasts can utilize both sugars and lipids. In the case of bacteria, available sugars are converted, for example, into fatty acids that are desirable by the host, which can be utilized to reduce stress under the impact of changes in the culture conditions during fermentation (temperature, pH, pressure, oxygen, nutrients, and ethanol). These specific fatty acids, such as cyclic fatty acids, modulate the bacteria membrane composition to maintain the lipid bilayer in the liquid crystalline lamellar phase and, more precisely, maintain an optimal level of fluidity within the lipid bilayer [22].

For yeasts, variation in the fatty acid contents in the fermented biomass depends on the initial lipid content in the biomass used. This was confirmed, for example, for Y. lipolytica grown on a poor vs. abundant-in-fat substrate [41]. On the medium abundant-in-fat (cotton seeds with 19.6% of lipids) substrate, there was a decrease in the total fat while on poor fat medium (cassava peels with 1.2% of lipids), there was a significant increase in the final fat content. In the current study, only two yeast strains (C. antarctica and C. lipolytica) were able to produce branched, odd-chain, or cyclic fatty acids.

3.5. Volatile Compound Changes after the Fermentation of PB Beverages

The odorous compounds in the control and prepared beverages are presented in

Table 5. Volatile compounds of the control and fermented PB beverages.

Class of Compounds	Identified Compound	Kovats Retention Indices		Almond Beverages									Oat Beverages
		KI MXT-5	KI MXT-17
				C	1	2	3	4	5	6	7	8	C	1	2	3	4	5	6	7	8
alcohols	n-butanol	654	781	+	+		+	+	+	+	+	+			+	+	+	+	+		+
	1-hexanol	859	986		+		+
	2-phenylethanol	1099	1273	+	+	+	+	+			+								+
	1-propanol, 2-methyl-	619	728		+	+	+				+		+	+	+	+	+	+	+
	propylenglycol	741	941		+	+	+				+			+	+	+			+	+
	terpinen-4-ol	1175	1274		+	+	+	+	+	+		+	+	+	+	+	+	+	+	+	+
	2-propanol	489	597	+		+		+	+	+	+		+	+	+	+	+	+		+	+
	(z)-3-hexen-1-ol	854	-							+
aldehydes	acetaldehyde	446	505	+	+	+	+	+	+	+	+	+	+	+		+	+	+	+	+	+
	propanal	466	590		+		+	+				+
	2-methylpropanal	519	621	+	+	+	+		+		+	+		+	+	+		+			+
	p-anisaldehyde	1251	1447	+	+	+	+	+		+	+	+	+	+	+	+	+	+	+	+	+
	2,4-decadienal, (e,e)-	1304	1456	+						+
	trans-2-undecenal	1369	1482					+									+
	alpha-hexyl cinnamaldehyde	1754	-	+	+	+	+	+	+	+	+	+	+	+	+		+	+	+		+
	3-methylbutanal	646	730		+		+				+
	hexanal	790	885		+		+				+	+	+	+	+	+		+		+	+
	octanal	983	1096		+								+						+
	benzaldehyde	958	1085		+		+	+	+	+							+		+
	pentanal	696	775						+	+		+			+			+
	furfural	827	954			+													+
	decanal	1193	1294														+
amines	pyrrole	777	928					+	+	+	+		+				+	+	+
cyclic compounds	myristicin	1509	1677	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+
esters	isopropyl acetate	646	720	+	+	+	+	+	+	+	+	+	+		+	+	+	+	+	+
	ethyl 3-(methylthio)propanoate	1107	1220	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+
	benzyl phenyl acetate	1810	-	+	+	+		+	+		+	+	+	+	+	+	+	+	+	+	+
	benzyl salicylate	1864	-				+				+			+	+
	2-phenylethyl phenyl acetate	1906	-	+		+		+		+	+	+	+	+	+	+	+	+	+	+	+
	isoamyl acetate	871	931		+	+	+	+		+	+	+					+	+			+
	ethyl propanoate	725	775		+	+			+	+	+	+	+	+	+	+	+	+		+
	methyl 2-methylbutanoate	786	835			+	+		+	+		+							+
	ethyl formate	479	-	+		+			+	+	+		+	+	+	+		+	+	+	+
	ethyl acetate	621	674					+			+								+
	ethyl isobutyrate	739	820					+		+				+		+	+		+
	ethyl butyrate	803	856							+									+
	butyl acetate	820	886					+		+									+
	ethyl tetradecanoate	1805	-							+
	benzyl acetate	1174	1293							+	+		+	+	+	+	+	+		+	+
	benzyl benzoate	1738	-													+				+
	pentyl octanoate	1460	-																+
	ethyl hexanoate	989	1061														+
heterocyclic compounds	maltol	1110	1294																+
ketones	2-heptanone	881	987	+
	butane-2,3-dione	559	697		+	+		+	+	+	+	+	+		+			+	+	+	+
	butan-2-one	587	679		+	+		+	+	+	+	+	+		+		+	+	+	+	+
	2,3-pentanedione	702	794					+		+	+	+	+				+		+
	undecan-2-one	1300	1407														+
lactones	delta-undecalactone	1505	1853	+	+		+	+	+	+	+	+	+	+	+	+	+	+	+	+	+
	delta-decalactone	1504	1731	+		+
	4-undecanolide	1612	1836	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+
	dodecan-4-olide	1689	1904	+		+	+			+		+		+		+			+	+
	delta-dodecalactone	1722	1947	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+
	4-octanolide	1251	1498						+		+			+	+	+		+			+
organic acids	propanoic acid	740	884	+		+	+	+	+								+		+
	3-methylbutanoic acid	875	1016	+		+	+	+	+	+	+	+	+	+	+	+	+	+	+	+
	(e)-9-octadecenoic acid	1934	-	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+
	acetic acid	614	776				+		+	+			+	+		+	+
	valerenic acid	1864	-					+					+			+	+		+	+	+
	butanoic acid	804	955							+
	pentanoic acid	887	1095			+			+		+			+	+	+		+		+
phenolic compounds	4-ethylguaiacol	1288	1428		+	+	+	+	+	+	+	+
	anisyl alcohol	1287	1506					+		+		+	+			+	+		+
	vanillin	1446	1646	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+
	anethole	1286	1405					+									+
terpenes	citronellal	1169	1252	+		+
	thymol	1299	1501		+	+	+	+				+	+	+	+	+		+	+	+	+
	gamma-terpinene	1084	1108				+	+		+	+	+
	citronellol	-	1332				+
	1r-(+)-alpha-pinene	952	943				+
	1s-()-a-pinene	-	955					+
	()-.-beta-.-pinene	983	1005			+											+
Number of identified volatile compounds in PB-beverages				25	30	34	34	37	29	39	36	31	30	29	30	32	36	31	39	27	24

+—compound detected; C—control sample (beverage not fermented); sample numbers correspond to the microorganisms described in the experimental section.

. The aroma of the control PB beverages was relatively gentle and mild (preliminary sensory analysis—data not shown) and typical for the used raw plant material. The aroma was created by 25 (almond) and 30 (oat) tentatively identified volatile compounds, with a prevalence of aldehydes, esters, and lactones in the almond drink (5 compounds in each group) and aldehydes (5 compounds), esters (7 compounds), and organic acids (4 compounds) in the oat drink. After fermentation, the highest increase in the number of identified volatiles was found for both drinks fermented by C. antarctica CAN 0001 (36/37 compounds) and Y. lipolytica YLP 0001 (39 compounds). The aroma profiles of both beverages fermented by yeasts were mostly differentiated, without characteristic compounds for the used strains. In contrast, compounds such as: 1-propanol, 2-methyl; propylenglycol; terpinene-4-ol; isoamyl acetate; 4-ethylquaiacol; and thymol in the almond beverages (they were absent in the control sample, although found in all fermented samples) were characteristic for the LAB-fermented beverages. For the oat drinks, LAB also produced propylenglycol; 2-methylpropanal; 4-octanolide; and pentanoic acid. It is also worth noting that unique compounds were found only in the almond beverages fermented by both strains of L. plantarum (1-hexanol; propanal, 3-methybutanal, hexanal, and benzaldehyde).

A PCA analysis of the variation between the control and fermented beverages based on volatile compounds is shown in Figure 1. This figure confirms the lack of characteristic compounds (absent/present) in the studied drinks, identifying clear variation between the beverages. The drinks that were most differentiated from the others are three almond drinks: the control (the characteristic compound was 2-heptanone), fermented by L. bulgaricus (the characteristic compounds were furfural and beta-pinene), and fermented by T. casei (the characteristic compounds were (z)-3-hexen-1-ol, ethyl butyrate, ethyl tetradecanoate, benzyl acetate, and butanoic acid). In the case of the used microbiota strains, the highest similarity of oat and almonds beverage aromas was found for those fermented by T. casei TCS 0001 (newly created volatiles were pentanal, 4-octanolide, and pentanoic acid) and K. marxianus KF 0001 (created volatiles were propylenglycol and pentanoic acid).

To date, there are still a lack of studies on aromas in plant-based fermented beverages, especially by yeasts. Based on the analogy to dairy products, the aroma of fermented food is mainly dependent on the type of processed food matrix, the microflora species/strains used, and the fermentation conditions. A recent study [61] showed that among the volatiles created by some prominent LAB strains in kefir grain, aldehydes, ketones, alcohols, and acids prevailed. For fermented dairy products produced by LAB strains, acetaldehyde, diacetyl, acetoin, and butanediol are typical aroma compounds as a result of metabolizing citrates [62]. From this catalogue, only acetaldehyde was found in the majority of samples in the current study, but this compound was also found in both control samples. According to Popova and Pinheiro de Carvalho [63], citrate is a compound that plant tissues usually contain in considerable amounts, which can explain the presence of acetaldehyde plant beverages. The presence of this compound is especially valuable since acetaldehyde gives a positive taste and aroma [64]. It is well known that the aroma profile of LAB-fermented food differs between homo- or heterolactic fermentation types [44,65]. It was previously concluded that the aroma in heterofermentation is even more complex with the additional production of acetic acid, CO₂, ethanol, ethyl acetate, hexyl acetate, ethyl hexanoate, isopentyl acetate, and C4-C6 alcohols [66]. Unfortunately, the current study did not confirm this research for the studied plant fermented beverages.

The volatile organic compounds produced by yeasts are more complex. They are products of organic compound transformations, especially via specific pathways such as Ehrlich (alcohols derived from amino acid catabolism), glycolytic (alcohols, acids, and esters derived from glucose catabolism), and β-oxidation (volatile fatty acids, aldehydes, and ketones derived from fatty acids) pathways (more details in Ogunremi et al. [67]). Finally, the composition of volatile compounds produced by yeast is related to the desired nutrition substrate of the species/strain and differs between microflorae. In the current study, the aroma of fermented beverages was created by an increased number of volatile compounds, with a predominance of propanal, n-butanol, 3-methylbutanal, isopropyl acetate; 2-methylpropanal, and 2-methyl-1-propanol. Among the studied microflora strains, Y. lipolytica was found to be a major producer of aromas in both beverages. This could possibly be a result of the ability of this strain to utilize various carbon substrates for growth and multiplication. It is worth mentioning that among the used yeasts, this strain produced the highest content of lactic acid.

4. Conclusions

The current study showed that almond and oat drinks are valuable nutrient sources for bacteria and yeast strain growth and multiplication. LABs multiply with a value close to 5 log, typical for fermented milk products. Among the analyzed yeasts, K. marxianus showed the highest count increase (close to 5 log) while C. lipolytica showed the lowest (close to 2 log). After 48 h of fermentation, sample acidification (measured as the titratable acidity and pH) was the highest in samples fermented by LAB and by Y. lipolytica YLP 0001. This indicates an increase in dissociated acids (lactic, citric, and succinic) and undissociated acids (amino acids and fatty acids). However, the amount of each group of acids depended on the type of fermented drink, with a higher content of dissociated acids in oat beverages and a higher content of undissociated acids in almond beverages. The protein content and composition of amino acids affect the values of the apparent viscosity of fermented beverages. For an oat drink with only 0.4 g of proteins in 100 mL, the protein concentration was too low to create a gel structure of the beverage. In contrast, the apparent viscosity of almond beverages was significantly increased in the majority of variants, with only one exception for C. antarctica (associated with protein precipitation at the isoelectric point for this beverage). In general, the values of apparent viscosity in almond beverages were positively correlated with the titratable acidity and negatively with pH. All fermented oat beverages were characterized by enhanced shares of branched, odd-chain, and cyclic fatty acids, with an increase from 5 to 1547 mg/100 g of lipids. For fermented almond beverages, these fatty acid contents increased only in samples fermented by LABs and by C. lipolytica (up to 459 mg/100 g of lipids). Among all used microorganism strains, Y. lipolytica was found to be a major producer of aromas in both beverages. This may be due to the ability of this strain to utilize various carbon substrates (various sugars and lipids) for growth and multiplication.