The impact of Qodume Shirazi seed mucilage‐based edible coating containing lavender essential oil on the quality enhancement and shelf life improvement of fresh ostrich meat: An experimental and modeling study

Abstract Today, food consumers prefer to use the foods that contain natural preservatives. The purpose of this study was to investigate the effect of Qodume Shirazi seed mucilage (QSSM) and lavender essential oil (LO) on the preservation of ostrich meat during cold storage. The chemical compounds of LO were identified through gas chromatography–mass spectrometry (GC/MS). The ostrich meat samples were coated with the mucilage containing the essential oil at concentrations of 0%, 0.5%, 1%, 1.5%, and 2%, v/v. The control and the coated ostrich meat samples were kept at 4°C and analyzed for microbiological (total viable count, psychrotrophic count, Escherichia coli, Staphylococcus aureus, coliforms, and fungi), physicochemical (moisture content, pH, texture, and color parameters), and sensorial (odor, color, and total acceptance) characteristics during 9 days of storage. GC/MS identified 12 compounds in the essential oil, among which linalool was the major one (43.3%). The lightness (L* value) and hardness of all the ostrich meat samples were reduced during the storage. From a microbiological point of view, the cold storage duration for the control and the coated sample without the essential oil was only 3 days, while for coated samples containing 0.5%, 1%, 1.5%, and 2% essential oil, it was 3, 3, 6, and 9 days, respectively. The coated ostrich meat containing 2% LO had an appropriate quality with an expanded shelf life. The results showed that neural network with 10 neurons in the hidden layer had the lowest mean squared error and mean absolute error and the highest correlation coefficient for predicting the quality and microbial properties of the coated meat samples during storage.


| INTRODUC TI ON
Ostrich meat is considered as a healthy alternative to other red meats in Western societies due to its favorable nutritional properties and one of the most popular sources of protein (low cholesterol and intramuscular fat contents and generally high omega-3 polyunsaturated fatty acids (PUFA x3) percentages). In recent years, there is an increasing trend toward consumption of ostrich meat. Therefore, improving the quality and increasing the shelf life of this product are very important both for local consumption and for export. Much research has been done to prevent microbial growth and sensory properties of ostrich meat. Factors such as initial microbial loads, packaging atmosphere, pH, storage temperature (and temperature abuse), and sampling day could affect the quality of ostrich meat (González-Montalvo et al., 2007). For example, the relatively high pH of ostrich meat is one of the factors that can cause rapid microbial spoilage in some packaging conditions. Another factor that may cause ostrich meat spoilage is its the higher polyunsaturated fatty acid content as compared to beef and chicken meats, which makes ostrich meat to be more sensitive to oxidative deterioration (Divani et al., 2017).
Modern packaging technologies can increase the shelf life of a packaged product by preventing or delaying microbial growth. This can be achieved by the manipulation of the meat microenvironment (González-Montalvo et al., 2007).
Due to the abuse of some chemical additives and preservatives and their harmful effects on human health, many countries have regulated food production. For this reason, interest in studying and researching on natural preservatives and replacing chemical preservatives with natural ones has been extensively increased. In addition, the excessive use of chemical preservatives leads to the development of bacterial resistance against antibiotics. Therefore, the study of natural antimicrobials has also received medical attention (Gonçalves Cattelan et al., 2018).
On the other hand, in addition to the growing demands of consumers to achieve a food with the high level of quality, environmental concerns about the negative impact of plastic packaging residues have recently attained more attention. Therefore, food industry experts try to create the new type of edible films and coatings to preserve food quality.
Edible films and coatings that are named and classified based on the compounds used contain different types of polysaccharides, lipids, proteins, or their combinations. The use of edible films/coatings as food packaging has many advantages, including environmental friendliness, economic viability, long shelf life, and good barrier properties for gases and carriers of other food additives such as antimicrobial, vitamins, and antioxidant agents (Hashemi & Mousavi Khaneghah, 2017).
Edible coatings are edible films made on the exterior surfaces of a food or material (Jooyandeh, 2011). In fact, edible coatings form a thin layer on fresh foods such as fruits, vegetables, and meat, thereby maintaining postharvest quality and increasing crop shelf life. The use of food coatings is a relatively simple, environmentally friendly, and inexpensive technology to reduce the spoilage of fruits and vegetables, and therefore has been relatively more accepted by consumers as compared to other storage methods such as the use of chemical preservatives or radiation (Saha et al., 2017).
Chemical composition and applications of water-soluble gums, also called hydrocolloids, used in foods have been well documented.
Most hydrocolloids are polysaccharides (gum Arabic, guar gum, carboxymethylcellulose, carrageenan, starch, and pectin) or proteins (such as gelatin). Hydrocolloids are widely used in the food industry in food systems for a variety of purposes, for example, as gelling agents, stabilizers, texture modifiers, and thickeners (Koochaki et al., 2008). Furthermore, numerous studies have shown that plant polysaccharides comprise biological features such as free radical scavenging and antivirus capabilities (Jooyandeh et al., 2018).
Alyssum is a genus of about 100-170 species of flowering plants in the family Brassicaceae, native to Egypt, Iran, Iraq, Arabia, and Pakistan. The seeds of this plant contain large amounts of mucilage.
This mucilage is used in traditional medicine, especially in Iran as a traditional medicine (Nafchi et al., 2017).
Qodume Shirazi (Alyssum homolocarpum) seed mucilage (QSSM) has a great potential to be used as a new source of biodegradable film due to its suitable thickening/gelling action. The majority of QSSM is carbohydrate (85.33%) with small amount of uronic acid (5.63%).
QSSM behaves like a typical polyelectrolyte because of the presence of carboxyl and hydroxyl groups. Mucilage extracted from Qodume Shirazi seed can be used as thickening, fat replacer, and stabilizer agent (Monjazeb Marvdashti et al., 2019).
The essential oils (Eos) in plants are recognized as safe (GRAS) and are among the most important natural products extracted from a variety of plants. Because of high antimicrobial and antioxidant properties, EOs are frequently used in the food industry as a flavoring, antioxidant, and antibacterial agent ( Gharibzahedi & Mohammadnabi, 2017).
There are about 8,000 herbal species in Iran, of which 2,300 are aromatic and medicinal herbs and are used in traditional medicine.
Lavender is one of the herbs used in traditional Iranian medicine.
Lavender is one of the most widely used medicinal plants in Iran.
Lavender has been used in traditional medicine until now. This plant is from the genus Lavandula and from the mint family. This plant has 39 species (Palá-Paúl et al., 2004). In Iran, there are two perennial species, Lavandula sublepidota and Lavandula stricta, which the former is an exclusive Iranian species.
Various studies have shown that when antimicrobial agents such as herbal extracts are added to films or edible coatings, they are slowly released to the food surface. So, they stay longer on food. In addition, oxidation can be effectively reduced by selecting suitable coatings that have low oxygen permeability (Sayyad et al., 2016).
In recent years, successful studies have been conducted on the application of various types of polysaccharide edible coatings containing different essential oils in extending the shelf life of different types of meat. Among such coatings, Lallemantia iberica mucilage containing Cuminum cyminum essential oil , Salvia macrosiphon mucilage in combination with Myristica fragrans (Kiarsi et al., 2020), the mixture of Lepidium sativum and Heracleum lasiopetalum (Barzegar et al., 2020), Lallemantia royleana mucilage blended with Allium hirtifolium (Alizadeh Behbahani & Imani Fooladi, 2018a), and Plantago major mucilage combined with Anethum graveolens essential oil (Alizadeh Behbahani et al., 2017) could be pointed out.
Therefore, the objectives of the present research were to produce bioactive edible coating based on QSSM loaded with different concentrations of lavender essential oil (LO) and evaluate the antioxidant and the antimicrobial activities of these substances in vitro and their effect on the quality and shelf life of ostrich meat during cold storage at 4°C.

| Materials
Qodume Shirazi seed (Alyssum homolocarpum) mucilage (QSSM) and Lavandula sublepidota L were purchased in Mashhad, Iran. Ostrich meat was purchased from local markets in Ahwaz, Iran. The microbial media, including Sabouraud dextrose agar (SDA), plate count agar (PCA), eosin methylene blue (EMB), violet red bile agar (VRBA), and mannitol salt agar (MSA), were supplied from Merck Co. All other chemicals and reagents used in this study were of analytical grade.

| Extraction of the Lavandula stoechas essential oil and determination of chemical composition
Hydrodistillation method using a Clevenger-type apparatus (power: 335 W; duration: 3 hr) was utilized to extract the essential oil from lavender. Quantification of LO components was performed using a Beifen 3420A gas chromatograph (Agilent Technologies 7890 A) equipped with a split injector, a flame ionization detector, and a DB-5 capillary column (30 m × 0.25 mm, 0.25 μm stationary phase thickness). To identify the LO constituents, the obtained retention profile was compared with that of known samples already analyzed by a gas chromatograph (Agilent 7890A) coupled to a mass spectrometer (Agilent 5975C) with similar column and operating conditions (Alizadeh Behbahani & Imani Fooladi, 2018b;Kiarsi et al., 2020).

| Determination of total phenolic content
In this experiment, 20 µl of LO at a concentration of 10 g/L was mixed with 2 ml of distilled water and 100 µl of folic acid reagent. After 3 min, 300 μl of sodium carbonate (Na 2 Co3) solution was added to them, and the solution was stirred for 2 hr, and finally, the absorbance of the solution was measured at 765 nm by spectrophotometer.
The amount of the total phenolic content was determined as mg of gallic acid equivalent (GAE) by means of gallic acid calibration curve.
Total phenolic content was reported as mg of gallic acid per gram (Alizadeh Behbahani et al., 2019a, 2019b).

| Determination of total flavonoid content
Aluminum chloride colorimetric method was used to determine the number of flavonoids. LO in 0.5 ml (1:10 g/ml) and 1.5 ml methanol, 0.1 ml aluminum chloride (10% methanol), and 0.1 ml potassium acetate with 2.8 ml water distillate was combined. The solution was then incubated at room temperature for 30 min. The absorbance of each reaction compound was measured at 415 nm with a spectrophotometer (Alizadeh Behbahani et al., 2019a;Noshad et al., 2018).
The total flavonoid content was expressed as mg quercetin equivalents (QE).

| DPPH free radical scavenging assay
Scavenging activity was performed according to Kiarsi et al. (2020) method. For this purpose, 25 mg of LO was dissolved in 5 ml of distilled water; then, 0.1 ml of this solution was mixed with 3.9 ml of methanol DPPH solution (0.1 M methanol DPPH solution). The samples were then incubated for 60 min in a dark place at room temperature; the absorbance of the samples was measured by spectrophotometer against pure methanol at 517 nm. Pure methanol was used to zero the device. Free radical scavenging activity of the samples was calculated as percentage of inhibition (RSA) using Equation (1): where A 0 and A 1 are the absorbance of the control and sample, respectively. Ascorbic acid was used as a positive control.

| Chemical analysis of gum and ostrich meat
The moisture, ash, crude protein, and fat contents of the QSSM and ostrich meat were determined according to AOAC (2005).

| Preparation of edible coating and treatment of ostrich meat samples
Two factors QSSM and LO as the central composition of active edible coatings were designed. For this aim, 2 g of the QSSM was mixed with 1 ml of Tween-80 and made up to 100 ml with distilled water and heated and agitated using a magnetic stirrer. The LO was added (1) Inhibition of DDPH (%) = 100 × (A 1 − A 0 )∕A 0 to QSSM solution at levels of 0%, 0.5%, 1%, 1.5%, and 2% v/v. After that, 1 group out of the 5 groups of the samples was coated via being

| Microbiological analysis
Ostrich meat samples were taken for microbial analysis on days 0 (after dipping treatment), 3, 6, and 9 days of refrigerated storage.
Then, 5 g of the samples was removed and mixed with 45 ml of sterile physiological serum. The solutions were thoroughly mixed with vertex for 5 min and diluted to 10 -1 . Then, 1 ml of the solution was removed and mixed with 9 ml of sterile physiological serum and diluted to 10 -2 . Next, subsequent dilutions (10 -1 to 10 -9 ) were prepared in the test tubes. Dilutions were obtained for each sample separately.
Then, 100 µl dilutions were removed and microbial counts were investigated by microbial culture according to the following methods:

Total viable count bacteria count in PCA (incubation at 37°C
for 24 hr).
3. Mold and yeast count in SDA (incubation at 27°C for 72 hr).
Finally, all the plates were examined visually for the characteristics of a colony (shape, size, pigmentation, etc.) associated with each growth medium. Microbial colonies were counted and expressed as log10 CFU/g ostrich meat (Hamedi et al., 2017).

| Physicochemical analysis pH measurement
Minced ostrich meat samples (10 g) were blended with 90 ml of distilled water and homogenized. The pH values of sample were measured using a digital pH meter (Dragon Lab, MX-S) at the 25°C (Amiri et al., 2019).

| Moisture content
The moisture contents (MC) of coated ostrich meat at 25°C were measured by drying the samples in an oven (Elektro Helios, 285 12, SE) at 102°C until constant weight (dry sample weight). Moisture content (%) was calculated using Equation (2): where m i and m d are initial and dried sample weights, respectively (Guerrero et al., 2015).

| Texture
Hardness of samples was measured based on Akwetey and Knipe In this test, compression force by a 10 kg weight was applied to the specimens (2 × 2 × 2 cm) up to 50% of its initial height at a constant velocity of 5 mm/s. The test was performed on days 0, 3, 6, and 9, and the samples were evaluated for tissue hardness parameter.

| Color
The color of the coated ostrich meat samples was measured by a Hunter Lab color meter (Konica Minolta, CR-400, JP) and was reported by the CIE system. The colorimeter was calibrated with a white standard. The values of L*, a*, and b*, respectively, which represent the brightness of L from black (zero) to white (100)

| Sensory evaluation
The 9-point hedonic method was used to evaluate the organoleptic properties of ostrich meat samples. For this purpose, the samples were evaluated in terms of color, aroma, and overall acceptability (1: dislike extremely to 9: like extremely) by 15 semitrained panelists.
Six samples/treatments of ostrich meat were given to each panelist separately in small porcelain dishes. The panelists were not aware of the experimental approach, and the samples were blind-coded.
Samples that scored more than 4 were considered as valid samples Nisar et al., 2019).

| Statistical analysis and neural network modeling-genetic algorithm
All experiments were performed in at least triplicate and were analyzed by ANOVA using SPSS software (ver. 21, IBM Inc.). Duncan's multiple range tests with 95% confidence level were used to compare the means.
An artificial neural network (ANN) consisting of a set of interconnected neurons is able to estimate outputs based on input and data. The designed lattice type was multilayer perceptron (MLP) in which the input layer consisted of 2 neurons (LO concentration and shelf life) and the output layer contained 5 neurons (moisture content, color index, total microbial load, hardness, and pH). Levenberg-Marquardt (LM) training algorithms were used to update the ANN weights. This algorithm is one of the most commonly used algorithms because it makes network training very fast and minimizes the level of error. In fact, this algorithm is designed to increase network learning speed based on the Hessen matrix (Salehi & Razavi, 2012).
One of the problems that may occur when training a neural network is network learning. In this case, the error is acceptable when training the network, but when evaluating, the network error is much more than the error of the training data. There are two ways to avoid overlearning: 1-to stop training; 2-to choose the lowest number of neurons in the hidden layer. The second method was used in this study. In order to train the network, the data were first randomly divided into three parts, with 60% of the data used for training, 20% of the data used for evaluation, and 20% of the data used for network testing. During network training, the training process was interrupted when the error between training and evaluation data increased.
The number of neurons in the hidden layer was optimized by genetic algorithm method. The initial population was assumed to be 100 generations and maximum 5 generations. The probability of fusion and mutation was 0.9 and 0.01, and the number of neurons to optimize was 1 to 20, respectively. In order to predict the performance of the obtained networks, the parameters of mean squared error (MSE), mean absolute error (MAE), and correlation coefficient (r) were used to predict the investigated parameter. In order to investigate the relative importance of each network input in modifying the output factors of the model, the relative sensitivity of the output factors to the network input parameters was also evaluated.

| Chemical compositions of lavender essential oil (LO)
Analysis of essential oil from lavender was performed by GC/ MS device. The results of this analysis showed the presence of 12 compounds in the essential oil, which accounted for a total of 99.41% of the essential oils. The most representative compounds of LO were monoterpene, and among them, the main constituents were linalool (43.3%), linalyl acetate (19.1%), β-myrcene (11.61%), D-limonene (5.43%), and 1,3,6-octatriene (5.31%) (

| Chemical compounds of ostrich meat
The proximate chemical analysis of the ostrich meat was deter-

| Total viable count (TVC)
The TVC of ostrich meat samples coated with QSSM incorporated with LO (0%, 0.5%, 1%, 1.5%, and 2%) during 9 days of storage at refrigerated temperature is shown in Figure 1a. The results revealed that the microbial count of all examined groups inclined to increase gradually with storage period. However, the increase in TVC in the samples coated with QSSM + LO was lower than the control sample. With the increase in essential oil, this slope was also decreased.
The TVC of the control sample on the 3 days was higher than the standard, indicating that the duration of ostrich meat storage in the refrigerator (4°C) was maximum 3 days. The results showed that the coating of ostrich meat increased the meat shelf life. According to these results, the shelf life of coated ostrich meat containing 2% LO was 6 days at 4°C as compared to the control sample. The pH value of ostrich meat is one of the determinant factors of TVC. The pH of the ostrich meat is near to 6 and is therefore heavily influenced by microbial growth (Van et al., 2005). Seydim et al. (2006) reported that addition of rosemary extract and sodium lactate to ostrich meat can reduce TVC. However, the use of a mixture of sodium lactate and rosemary extract provided better antimicrobial results than when they used alone. Therefore, sodium lactate alone or in combination with rosemary extract can be used to reduce microbial growth and increase the shelf life of minced meat when stored in the refrigerator.

| Psychrotrophic count
The results of statistical analysis showed that the treatments and storage time had a statistically significant effect on psychrotrophic bacterial count in ostrich meat (p ≤ .05). Our results are shown in The researchers found that the use of modified atmosphere rather than the use of conventional air conditions reduced the number of bacterial cells during 7-day storage period. Pseudomonas aeruginosa species are highly aerobic and oxygen-dependent, and oxygen deficiency prevents the growth of these bacteria.

| Staphylococcus aureus count
The results of variance of all treatment data showed that the S.

| Escherichia coli count
The results of E. coli count during storage are illustrated in Figure 1d.
The results showed that the initial level of E. coli in all samples was showed that the essential oil of lavender had a significant antimicrobial effect on both bacteria, but the antimicrobial effect of the essential oil on E. coli was less than that on S. aureus because of its cell wall. Alizadeh Behbahani et al. (2017) reported that the use of Plantago major seed mucilage as a novel edible coating incorporated with Anethum graveolens essential oil significantly reduced the number of E. coli and S. aureus during storage (18 days) in the treated samples.

| Coliform count
The results of counting the total number of ostrich meat coliform bacteria maintained at 4°C are given in Figure 1e. The initial bacterial coliform count was 2.04 log CFU/g. The results showed that the number of coliform bacteria increased during the storage time in all treatments, but on day 9 of storage, there was a significant difference between the control sample (7.17 log CFU/g) with the samples containing 1.5% essential oil (3.8 log CFU/g) and the samples containing 2% essential oil (3.34 log CFU/g). The results of research by Seydim et al. (2006) showed that using vacuum packaging containing rosemary essential oil

| Mold and yeast count
The results of the counts of molds and yeasts in ostrich meat stored at 4°C are shown in Figure 1f. The results showed that the initial level of fungi in the samples was 1.08 log CFU/g. Over time, the count in-   the samples decreased at days 3 and 6 and then increased at day 9.
Various studies have shown that microbial activity causes carbon dioxide production; some of the essential oils and extracts in film and edible coatings can reduce or increase the permeability to carbon dioxide (Bifani et al., 2007). It seems that the LO in the QSSM has reduced permeability, resulting in increased carbon dioxide content in the coating, and ultimately increased carbon dioxide, reducing the pH of the samples. On the other hand, increased carbon dioxide can also reduce microbial growth. In the QSSM-coated samples, due to the absence of essential oil, the permeability was higher than that of the QSSM + LO-coated samples, which resulted in lower pH on days 3 and 6 and a higher pH on day 9 as compared to the QSSM + LO samples. Generally, during the storage time, the pH decreased on days 3 and 6 in the coated samples with increasing essential oil con- The moisture content of ostrich meat stored at 4°C is shown in Figure 2b. The results showed that the initial moisture content of the samples was 75%. Over time, the moisture content of all samples decreased, but the rate of moisture content decreased in the coated samples. Also, the results of data analysis of variance showed that all samples had significant differences in the moisture content (p ≤ .05) on day 9. Previous research has shown that water storage capacity has a direct relationship with pH; as pH decreases, water storage capacity decreases. Therefore, one of the reasons for the decrease in the moisture content in the samples is likely to be a decrease in water storage capacity (Botha et al., 2006). On the other hand, Nisar et al. (2019) stated that addition of essential oil to edible coating increased the ability to prevent moisture loss. In fact, water binds to the hydrophilic portions of the coating to exit the coating; the addition of the essential oil to the coating increases the hydrophobic part of the coating matrix, thereby preventing further physical barriers to moisture transfer, ultimately reducing its water transmission.
The results of texture evaluation of the ostrich meat maintained at 4°C are presented in Figure 2c. The results of analysis of variance showed that the hardness of all samples decreased with passing time.
On day 9, the hardness of all samples showed a significant difference (p ≤ .05). In the QSSM + LO samples, the hardness increased on day 3 of the storage and then decreased on days 6 and 9. No significant difference (p ≤ .05) was observed in the QSSM samples on days 1 and 3, but on days 6 and 9, the hardness decreased (Figure 2c).
The results of this study were in line with the findings of Rahnemoon et al. (2018). These researchers investigated the effect of alginate coating containing pomegranate peel extract on shelf life, texture, and color characteristics of chicken breast meat. The researchers attributed the decrease in hardness of the samples to the destruction of the meat tissue by microorganisms. On the other hand, with decreasing water-holding capacity due to lower pH in the QSSM and LO samples, the hardness of the samples increased and then decreased. Lowering the pH reduces water storage capacity.
Therefore, the hardness was increased on day 3. On day 6, due to tissue damage by microorganisms, the hardness decreased despite lower pH. On day 9, due to increases in pH and tissue destruction by microorganisms, the rate of hardness decreased (Lawrie & Ledward, 2006).

| Color measurement
The results of color evaluation of the ostrich meat samples stored at 4°C during 9 days of storage are presented in Figure 3a Changes in the sample containing 2% essential oil were lower than in the other samples during the storage time ( Figure 3c). The results showed that the ΔE factor, which reflects the overall color change of the sample and is dependent on the L*, a*, and b* factors, increased during the storage time ( Figure 3d). Van et al. (2005) reported that increasing pH in ostrich meat darkens the meat color. Vital et al. (2016) investigated the effect of active edible alginate coating with rosemary and oregano essential oils on beef. The results of these researchers are consistent with the present study. These researchers attributed the decrease in factor L* during storage to changes in protein structure due to oxidation that may increase light scattering, and in the samples containing the coating, the reduction was less due to partial inhibition of oxygen exchange. Adding essential oils to the samples due to their antimicrobial properties slowed down the microbial activity, resulting in lower protein degradation rate and lower brightness than the control and QSSM samples. The researchers also reported that in fresh meat, when cutting any kind of meat, the so-called meat color flowers and oxygenated myoglobin pigment produce oxymyoglobin; it is brown in color and reduces the sample red color during storage. In the coated samples up to 6 days of the storage, due to the coating and reduction of oxidation process, the amount of oxymyoglobin was higher and this resulted in redness escalation but it decreased after 6 days of storage.
The essential oil also reduced the rate of oxidation due to its antioxidant substances and the slope of the redness curve decreased with increasing essential oil content. Figure 4 shows the color changes of the control ostrich meat and treated samples after 9 days of storage.

| Sensory evaluation
The results of sensory evaluation (color, aroma, and overall acceptability) of ostrich meat samples kept at 4°C are shown in Figure 5a-c.
In this way, ostrich meat samples with a score above 4 can be accepted (Hansen et al., 1995;Nisar et al., 2019). Sensory scores (color) less than 4 were observed for control and QSSM samples on day 9. For the QSSM + LO samples, the retention score by day 9 was greater than 4 ( Figure 5a).
The results of analysis of variance showed that in sensory evaluation of samples (aroma), score less than 4 for control sample was observed on day 3. For the QSSM sample, score less than 4 was obtained on day 6, and for QSSM + LO samples containing 0.5%, 1%, 1.5%, and 2% essential oil up to day 9, a score above 4 was obtained ( Figure 5b). Various studies have shown that Pseudomonas spp.
On the basis of sensory tests, the ostrich meat control, the QSSM, and the QSSM + LO samples had (overall acceptability) score above 4 on days 6, 6, and 9, respectively ( Figure 5c).
Comparison between the results of microbial count and sensory (overall acceptability) of the samples showed that there is a good correlation between the data obtained from these tests. For the sample containing 2% essential oil, the shelf life of microbial tests (total microbial load count) and sensory characteristics (overall acceptability) were appropriate until day 9 of storage. For the control sample, microbial load was adequate up to day 3 and overall acceptability was adequate up to day 6. Fazlara et al. (2017) investigated the effect of using gelatin-Avishan Shirazi (Zataria multiflora Bioss) coating on microbial, chemical, and sensorial characteristics of ostrich fillets in refrigerated conditions. These researchers showed that the control sample had appropriate overall acceptability, microbial load, and physicochemical characteristics up to 3 days. Regarding the gelatin coating without thyme, it was also noted that it retains usability for up to 3 days in terms of microbial load and up to 6 days in terms of physical and sensory changes. For Shiraz thyme treatment alone, they reported acceptability for microbial load up to 6 days and for physical changes up to 9 days, while for gelatin-containing thyme treatment, they found appropriate count for psychrophilic up to 6 days, mesophilic up to 12 days, and sensory and physical changes up to 12 days.

| Neural network modeling-genetic algorithm
The optimal structure of the artificial neural network (ANN) was obtained to model the ostrich meat coating process during trial and error. The best artificial ANN structure and related parameters are shown in Table 2 In order to investigate the effect of input parameters and identify the most effective factor, the sensitivity analysis test was performed on the optimal network. The results showed that the shelf life was more effective in predicting the quality and microbial properties than the essential oil concentration in the coatings (Figure 7).
The results of this study were similar to those of Sagdic et al. (2012);

| CON CLUS ION
The incorporation of LO to a natural hydrocolloid, QSSM, fab- meat samples were reduced during storage. From a microbiological point of view, the cold storage time for control sample and sample coated without essential oil was 3 days, and for 0.5%, 1%, 1.5%, and 2% essential oils, it was 3, 3, 6, and 9 days, respectively. The coating containing 2% LO conferred good quality characteristics to the ostrich meat and expanded its refrigeration shelf life. The results showed that neural network with 10 neurons in hidden layer had the lowest mean squared error (MSE) and mean absolute error (MAE) and the highest correlation coefficient (r) for predicting the quality and microbial properties of coated meat during storage time.

TA B L E 2
The best artificial neural network (ANN) structure for modeling the meat coating process