ACCEPTED AUTHOR VERSION OF THE MANUSCRIPT: Effect of adding essential oils of caraway and rosemary on volatile aroma compounds derived from stored vacuum packaged minced turkey meat

The of oils on the Abstract Changes in the odor of meat during its storage are one of the basic indicators affecting its assessment and possible disqualification. The aim of the study was to determine whether the addition of essential oils may affect the composition and concentration of volatile compounds included in the aroma of stored turkey meat. We investigated the effect of adding essential oil (EO) of caraway (0.02% v/w), rosemary (0.02% v/w) and a mixture of the two (0.01% each) on the composition of volatile compound fractions formed during 10-day storage of vacuum-2 packed minced turkey meat. The EOs used were also evaluated for their influence on microbial contamination (total viable count and lactic acid bacteria count), sensory quality (odor and taste) and the level of fat oxidation (acid value, peroxide value and p-anisidine value) and pH in chill-stored samples. In terms of sensory indicators, the greatest beneficial effect of adding oils was noted in the odor of raw meat. Use of the HS-SPME/GC-MS (headspace-solid phase microextraction/gas chromatography-mass spectrometry) method showed that the addition of oils significantly reduced the amounts of benzeneacetaldehyde, 2-octenal and ethyl 2-methyloctanoate compared with the control sample. In addition, in the presence of essential oils of rosemary and caraway, decreases were noted in benzaldehyde and 9-octadecenal, respectively. These changes may potentially affect the reception of the odor. There was a beneficial effect of the oils in reducing the levels of fat oxidation indicators, including peroxide. However, the oils at applied concentrations had no significant effect on the total viable count and LAB count.

packed minced turkey meat. The EOs used were also evaluated for their influence on microbial contamination (total viable count and lactic acid bacteria count), sensory quality (odor and taste) and the level of fat oxidation (acid value, peroxide value and p-anisidine value) and pH in chill-stored samples. In terms of sensory indicators, the greatest beneficial effect of adding oils was noted in the odor of raw meat. Use of the HS-SPME/GC-MS (headspace-solid phase microextraction/gas chromatography-mass spectrometry) method showed that the addition of oils significantly reduced the amounts of benzeneacetaldehyde, 2octenal and ethyl 2-methyloctanoate compared with the control sample. In addition, in the presence of essential oils of rosemary and caraway, decreases were noted in benzaldehyde and 9-octadecenal, respectively. These changes may potentially affect the reception of the odor.
There was a beneficial effect of the oils in reducing the levels of fat oxidation indicators, including peroxide. However, the oils at applied concentrations had no significant effect on the total viable count and LAB count.
Key words: volatile aroma compounds, essential oil, poultry meat, storage, antioxidative effect Turkey meat is popular due to its relatively low price, low fat content and the fact that it is sensorially acceptable. In the world's poultry production, it takes second position after broiler meat (Oz and Yuzer, 2017). Wong and Sampugna (1993) reported that fat content in the retail raw ground turkey samples analyzed by them ranged from 7.2 to 10.8%. In cited research total content of polyunsaturated fatty acids, expressed as weight percentage of total fatty acid methyl esters, ranged from 24.6 to 32.5% and linoleate 18:2n-6 was the major polyunsaturated acid. Total lipid content determined by these authors in light turkey meat, in dark meat, and in skin was 2 g·(100 g) -1 , 5.1 g·(100 g) -1 , and 39.8 g·(100 g) -1 , respectively.
Fatty acids not only have an effect on the nutritive value, sensory quality and stability of fat 3 during meat storage, but are also the main source for the formation of volatile compounds, with a crucial role of unsaturated fatty acids (Van Ba et al., 2013). Glucose and amino acids are also odor-forming substrates (Corral et al., 2013;Nychas et al., 1988). Transformations of substrates take place due to the activity of meat enzymes, microbial-derived ones and as a result of chemical reactions, including oxidation.
The composition of determined volatile fraction is affected by both species of animal and its diet and the method of sample preparation (Ahn et al. 1999;Ahn et al. 2006).
Studies on extending shelf life of minced meat include the use of additives and modifications of packaging methods. Natural substances that can be used in preserving minced meat include essential oils (EOs) of plant origin. Many of them have antimicrobial and antioxidant properties (Bulut et al., 2009;Othman et al., 2016). Studies on the use of these substances concentrate on the extension of meat durability and the reduction of the pathogen proliferation risk (Vasilatos and Savvaidis, 2013;Boskovic et al., 2017;Irkin and Esmer, 2010). The maximum concentrations of EOs, considered as a preserving addition to meat, are limited not only by the cost of these substances, but above all by the sensory acceptability threshold. Effective concentrations are often higher than sensorily acceptable ones (Nazer et al., 2005). Besides concentration, combinations of specific oils and types of meat may also be a barrier to sensory acceptance. In this study, the oils used were selected as acceptable in turkey meat in the preliminary sensory evaluation. One of the effects of adding EOs to meat may be an improvement in its odor during refrigerated storage (Govaris et al., 4 2010;Chouliara et al., 2007;Shahbazi et al., 2018;Skandamis and Nychas, 2001). The effect observed may be due to both masking the spoiled meat odor by the added volatile substances and the direct influence of EO components on the chemical, biochemical and microbial reactions, resulting in the formation of components of the odor of spoilage.
The aim of the study was to determine whether the addition of essential oils of rosemary and caraway affects the composition of the volatile compound fraction of stored turkey meat. Due to the fact that odor is also affected by the activity of microorganisms and fat oxidation, the extent of changes in these quality factors was also assessed.

Material
The raw material was turkey thighs. Meat was collected one day after slaughtering.
The raw material was next minced using a meat grinder (Ø 3 mm) and divided into four batches. One was a control (without additives). The remaining three were mixed with EOs, one with EO of caraway (Carum carvi L.) (0.02% v/w), the other with oil of rosemary (Rosmarinus officinalis L.) (0.02% v/w) and the last one with a mixture of the two (0.01% each). The selection of oils and their combinations was based on previous sensory evaluations, during which EOs were added to meat. The maximum concentration was also established on the same basis. 100% natural EOs of caraway and rosemary were purchased from a shop and their compositions were established by the HS-SPME/GC-MS method (Table   3). 0.4 kg portions of the products were vacuum packaged using a packaging machine (Vac-Star 1000, Switzerland) and then stored at 1÷2°C in darkness. Analyses were performed in fresh raw material and after 10 days of storage.

5
The odor and color of raw meat as well as the color, odor and taste of cooked meat were assessed by means of sensory evaluation. Samples, in the form of 40 g meatballs, were cooked in water for 15 min and assessed on the basis of the previously compiled table. A score of 5 points was assigned to the highest quality, 3 points was the acceptability threshold, while a score of 2 points was assigned to the poorest (and unacceptable) quality. Evaluation was conducted by a trained 7-member sensory panel.

Rancidity indicators and pH analysis
The anisidine value (p-AV) was determined using the methodology described in Cd 18-90 (AOCS, 2004), extracting the secondary fat oxidation products with isooctane, and then measuring the absorbance of their reaction product with p-anisidine at λ = 350 nm. The peroxide value (PV) was determined using the Cd 8b-90 method (AOCS, 2004). Fat fraction extracts containing glacial acetic acid and potassium iodine solution were titrated with sodium thiosulfate in the presence of starch. The results of the determinations were expressed in meq of active oxygen per kg of product. The acid value (AV) was determined based on the Cd 3d-63 standard (AOCS, 2004), titrating the samples with NaOH solution in the presence of phenolphthalein as an indicator. This value was expressed as the number of mg of free fatty acids per g of product. pH was measured using HI 9025 (Hanna Instruments) pH meter.

Microbiological analysis
Samples for microbiological analyses were placed in sterile bags with the dilution fluid, homogenized, and then a series of dilutions were performed.

Volatile compounds analysis
Commercially available, 100% natural EOs and meat-derived volatile compounds were analyzed by means of the headspace solid phase microextraction and gas chromatography-mass spectrometry (SPME-GC-MS) method. For the determination of volatiles, 2 mL of saturated saline with an internal standard solution (5 mg·L -1 4-methyl-2pentanol and 0.05 mg·L -1 ethyl nonanoate, Sigma-Aldrich) and a 10 µL sample of EO or a 0.5 g sample of turkey meat was placed in a 10 mL vial. The SPME device (Supelco Inc., Bellefonte, PA, USA) coated with PDMS (100 μm) fiber was first conditioned (250°C for 1 h) and next inserted into the headspace under stirring (300 rpm) for 30 min at 60°C.
Subsequently, the SPME device was introduced into the injector port of the Agilent benzaldehyde; Sigma-Aldrich) was based on the comparison of retention times and peak surface area of sample and standard chromatograms. Other detected components were determined semi-quantitatively (g·kg -1 ) by measuring the relative peak area of each identified compound, in relation to that of the internal standard.

Statistical analysis
All analyses were carried out three times.

Sensory changes
After the storage period, the color of both raw and cooked meat remained unchanged and achieved the highest scores (5.0). The taste of cooked meat changed slightly in the case of the control; however, after the storage period it still scored 4.6, which is a good result. The samples with EOs were evaluated better and scored higher grades. The odor of cooked meat was also very stable and for the control sample at the end of the storage period still scored 4.5 and for samples with EOs the scores were higher. In contrast, the odor of raw meat changed significantly (Table 1). After 10 days of storage only the samples with added rosemary essential oil scored on average 3.0 points. The samples containing both essential oils or caraway essential oil scored slightly lower than the acceptability threshold (3.0), but significantly higher than the control probe.

Changes in fat rancidity indicators, meat pH and microbiological quality
The values of fat quality indicators are given in Table 2. The acid value allows the assessment of the extent of hydrolytic changes occurring in the fat fraction. The EOs added individually to meat had no significant effect on acid value, whereas a small positive effect was observed in the case of their mixture, which is consistent with changes in the pH value ( Table 2). The effect of adding EOs was more visible with regard to indicators of oxidative changes. The peroxide value determines the amount of peroxides and hydroperoxides. The addition of rosemary oil, both in the concentration of 0.01% and 0.02%, allowed the initial value of this parameter to be maintained until the end of the storage period. In contrast, in the samples containing only essential oil of caraway, there was an increase in PV, but it was significantly lower than in the control sample. The anisidine value determines the amount of aldehydes resulting from the decomposition of hydroperoxides, mainly 2-alkanals reacting with p-anisidine. As for changes in this value, the addition of both EOs had a positive effect on fat stability; the effect was stronger in meat with added essential oil of caraway. The pH was slightly reduced in all stored samples of minced meat. The smallest change in the pH value compared to fresh meat was noted in the samples with a mixture of EOs. The essential oils used had no inhibitory effect on TVC and LAB count (Table 1). Both TVC and LAB exceeded the threshold of 7 log cfu/g.

Composition of EOs
The composition of dominant volatile compounds in EOs used in minced turkey meat, established by the SPME-GC-MS method, is presented in Table 3. The monoterpene carvone (over 60% of the relative essential oil quantity), followed by trans-dihydrocarvone, Dlimonene, carvacrol and p-cymenene were the main components of caraway EO. EO of 9 rosemary had a larger amount of ingredients with significantly smaller individual percentages.
In this EO, eucalyptol, camphor, β-caryophyllene and borneol were found in the largest relative quantities.

Changes in concentration of the meat-derived volatile compounds
Changes in concentration of the meat-derived volatile compounds are shown in Benzeneacetaldehyde was found only in control samples and those with addition of caraway essentail oil, both after storage. On the other hand, the addition of EO of caraway contributed to the reduction of the amounts of 9-octadecenal, 2-octenal and ethyl 2-methyloctanoate. In turkey samples with the addition of EOs, a lower content of hexathiane and cyclic octaatomic sulfur was also found.

Discussion
The effect of EOs on the results of sensory evaluation was largest in the case of the odor of raw meat. Other authors also observed the beneficial effects of adding EOs on the odor of such products as minced sheep meat, chicken breast meat, minced camel meat and minced beef (Govaris et al., 2010;Chouliara et al., 2007;Shahbazi et al., 2018;Skandamis and Nychas, 2001). Data in Table 4 indicate that the effect obtained may result not only from the masking effect of these additives, but also from the fact that they can directly limit the formation of some compounds. After 10-day storage, in the turkey minced meat without EOs among others, some esters such as ethyl acetate and ethyl decanoate, and carbonyl compounds such as benzeneacetaldehyde and diacetyl appeared. The concentration of some other esters, such as ethyl octanoate, increased. The presence of these components should probably be attributed to the microbial activity (Costello et al., 2013). Another source of these substances is amino acids, from which, depending on their structure (e.g. branched, aromatic or sulfur), are also formed respective alcohols, acids and sulfur volatile compounds (Corral et al., 2013). Another important substrate used by microorganisms is glucose, which is transformed, e.g. into lactic acid, CO2, acetic acid, ethanol, and also 11 diacetyl. The substrates for the latter could be also inosine and ribose (Nychas et al., 1988;Argyri et al. 2015;Pothakos, 2015). Diacetyl, described as giving a butter-like odor, may be produced by lactococci, Leuconostoc spp. and pediococci (Hoover, 2000), as well as Brochothrix thermosphacta (Pin et al., 2002;Illikoud et al., 2019). Moreover, Nieminen et al. (2016) observed that during the storage of raw pork in a modified atmosphere, diacetyl was one of the substances showing the highest correlations with the results of sensory evaluation and bacterial concentration. Hexanal, formed e.g. due to linoleic acid oxidation, was present in the highest concentration of all analyzed carbonyl compounds (Table 4). It is a component of the volatile compounds, which is broadly described in the literature. Its odor is described as fishy, grassy (Gu et al., 2013), or strong, rancid, andunpleasant (MacLeod andCoppock, 1976), and the odor threshold in meat is 0.015 ppm (Montel et al., 1998). According to Shahidi and Pegg (1994), this compound may undergo further oxidation or react with other meat ingredients, which results in limitations of its application as a lipid oxidation indicator.
Studies on the storage of various meat species showed both an increase and decrease in the concentration of this substance after the storage period (Yang et al., 2018;Olivares et al., 2012;Argyri et al., 2015). One of the factors possibly affecting its concentration is the way of feeding animals. Descalzo et al. (2005) observed higher volatile levels of hexanal and several other aldehydes as well as thiobarbituric acid-reacting substances (TBARS) in fresh beef from grain-fed animals than in meat derived from pasture-grazing cattle. Bosse (née Danz) et al.
(2017) reported that inoculation of cured loins with Staphylococcus carnosus reduced the formation of this substance and limited the formation of such fat oxidation products as nonanal, 2-pentanone and nonanone. The presence of EOs in turkey meat did not reduce the content of this constituent. In the case of the experiment carried out in this study, it does not seem to be a useful indicator of fat oxidation or sensory changes regarding aroma. Another component which was detected at a relatively high and similar level in fresh and stored meat 12 samples without EOs added was 2-pentylfuran. This compound could have appeared in meat due to lipid oxidation. It can also be an ingredient released by the heat of cooking, or by mechanical grinding, as took place in the case of the minced turkey meat assessed in the present study. Ercolini et al. (2010) also reported that 2-pentylfuran was produced by 54% of the 65 different strains of Pseudomonas fragi which were isolated by the authors from fresh and spoiled meat. Benzaldehyde and benzeneacetaldehyde are compounds that can be formed as a result of microbial spoilage (Casaburi et al., 2015). Benzeneacetaldehyde was detected by Ferrocino et al. (2013) in vacuum packed beef chops after storage at 1°C. In our studies this component was found only in stored control samples and samples containing caraway EO.
Reduction of total viable count was not detected in samples with rosemary essential oil (Table   1). However, the absence of benzaldehyde and benzeneacetaldehyde in stored turkey meat with rosemary EO may evidence its influence on microbial metabolism or changes in qualitative composition of TVC. The addition of essential oils also reduced the concentration of hexathiane and cyclic octaatomic sulfur in stored samples. The significant presence of these components was detected by GC-MS-SPME in spoiled pork by Briones et al. (2017), who mentioned that these compounds are associated with the unpleasant odor in the samples of spoiled meat. However, it is difficult to assess the effect of EOs on the sensorially evaluated odor of the examined turkey meat since, for example, it is difficult to evaluate the contribution to the odor of those meat-derived compounds whose decrease in concentration was affected by these essential oils. The final effect of interactions between individual odor components depends on their and their mixture characteristics as well as the concentration of individual components. Different components have various odor thresholds, which can vary depending on the matrix of the food product in which they are detected (Ercolini et al., 2010) For example, based on the odor active value (OAV), Zhu et al. (2018) identified benzaldehyde and benzeneacetaldehyde as two of the components responsible for the chestnut-like aroma 13 commonly attributed to an excellent-quality green tea. These ingredients have also been mentioned, for example, by Gu et al. (2013) as characteristic aroma-active compounds of some edible parts of steamed Chinese mitten crab (Eriocheir sinensis). The authors describe benzaldehyde as bitter and almond-like, while benzeneacetaldehyde was described as rosy and fruity. Benzaldehyde may result from catabolism of the aromatic amino acids and may be formed due to proteolytic activity of either microorganisms or meat's own enzymes (Ercolini et al., 2010). This substance was also detected by Soncin et al. (2007) in the raw breast meat of duck. Calkins and Hodgen (2007) list it among the compounds identified in beef. On the other hand, Freeman et al. (1976) reported that benzaldehyde was one of the components whose presence was associated with aerobic spoilage of chicken breasts and was only detected in the samples stored at 2°C, not in those kept at 10°C. The authors revealed that the temperature of meat storage affected the formation of some substances, e.g. benzaldehyde or 2-butanone in the aerobically stored samples, while others such as ethyl acetate occurred in both samples, stored at 2 and 10°C. The storage temperature of the raw material have an effect on the proliferation and metabolic activity of microorganisms. This study, however, showed that the applied EOs had no significant effect on the total microbial count, which probably mainly consist of lactic acid bacteria (Table 1). At the same time, however, it cannot be excluded that microbiota changed qualitatively. Nevertheless, significant differences were found in the composition of volatile compounds (Table 4). The decrease in concentrations of some volatile compounds that occurred in the stored meat under the presence of essential oils could result from their influence on the composition of the microbiome, which did not significantly differ in total counts, microbial metabolism or chemical interaction, including anti-or prooxidative activity. Farag et al. (1989) also observed the antioxidant properties of both EOs; however, in comparison with essential oil of rosemary, EO of caraway exhibited stronger antioxidant properties, when applied to emulsified linoleic acid. Antioxidant activity 14 of essential oils was reported to be derived partly from the presence of phenolic groups.
Carvone, γ-terpinene and limonene are the components of caraway EO with strong antioxidant activity (Abou El-Soud et al., 2014), while eucalyptol, α-pinene and β-pinene are such components of rosemary EO (Wang et al., 2008). Santos et al. (2017) reported that as one of several essential oils, rosemary EO showed noticeable activity in terms of a bactericidal effect. However, according to the authors the effective concentrations of the examined EOs were too high to be used in food, as the odor was too intense. According to Seidler-Łożykowska et al. (2013) essential oil from caraway has medium antimicrobial properties. The smallest change in the pH value compared to fresh meat noted in the samples with a mixture of EOs also may indicate that despite the fact that essential oils had no effect on the count of microorganisms, their metabolic activity was to some extent limited ( Table 2).
The meat containing a mixture of oils also had the smallest increase in acid value (Table 2).

Conclusions
The use of essential oils in minced meat at sensorially acceptable concentrations (0.02%) had a positive effect on the reduction of fat rancidity as well as on the sensorially evaluated odor of the product.