MODELING OF MUSHROOMS (AGARICUS BISPORUS) OSMOTIC DEHYDRATION PROCESS IN SUGAR BEET MOLASSES

Mushrooms (Agaricus bisphorus) were osmotically dehydrated in sugar beet molasses solutions at concentrations of 60%, 70% and 80%d.m., at operating temperatures of 25 °C, 35 °C and 45 °C during 0.5h, 1h, 1.5h, 2h, 3h, and 5 hours. Moisture content, water activity (aw), microbiological parameters (total plate counts, Enterobacteriaceae, total yeasts and moulds count) and content of mineral elements (potassium, magnesium, iron and calcium) were determined in the osmodehydrated mushroom samples. Response surface methodology and analysis of variance were selected to estimate the main effects of the process variables (temperature, time, concentration) on process performance and selected mushroom attributes (microbiological counts, chemical composition and mineral content). Increase in the values of applied osmotic process parameters led to the significant increase in the content of minerals (for example, an increase in K and Ca content by 269.42% and 939.03%, respectively) and a decrease in aw values (from 0.941 to 0.811), decrease in microbiological load and relative protein content (decrease by 33.07%), indicating a possibility for prolonged shelf life and suitability for further processing. The osmodehydrated mushrooms could be considered as ingredients for new functional (semi)products due to improved nutritive profile.


INTRODUCTION
Osmotic dehydration involves the immersion of a substrate (food) in a solution of high concentration, which partially dehydrates the food (Mújica-Paz, Valdez-Fragoso, Lopez-Malo, Palou & Welti-Chanes, 2003;Erle & Schubert, 2001). The driving force for water removal is the concentration gradient that is established on the opposite sides of the cell membrane.
This gradient causes two counter currents: dissolved substances of osmotic solution diffuse into the material being treated whilst water from the material flows into osmotic solution (Qiu, Zhang, Tang, Adhikari & Cao, 2019;Ciurzyńska, Kowalska, Czajkowska & Lenart, 2016;Ahmed, Qazi & Jamal, 2016). Water is removed by capillary forces and diffusion, UDK 543.068+582.28:582.662 Original research paper DOI: 10.5937/ffr47-28436 while counter current leaching and solute uptake are transmitted by diffusion (Shi & Xue, 2009;Rahman & Perera, 2007). Both flows depend on the cellular structure and membrane permeability (Fernandes, Gallão & Rodrigues, 2009;Amami et al. 2007). Products with reduced water content are obtained by osmotic dehydration. They are classified as partially preserved (minimally processed) products of high quality with prolonged shelf life (Silva, Fernandes & Mauro, 2014). Intermediate moisture foods (IMF) are gaining in importance in widespread use because their characteristics are similar to fresh food products but shelf life is extended (Qiu, Zhang, Tang, Adhikari & Cao, 2019). If a fresh appearance of a dehydrated product is required, the effect of dehydration must be at least 30%, which makes such products moderately stable and could be further processed (dried, frozen or treated with additives) (Khan, 2012;Phisut, 2012) in the dairy, bakery or confectionery industries. Choice regarding proper osmotic solution is important factor for successful dehydration, and this decision is mainly influenced by the water loss/solid gain ratio. In the case of osmotic solution with lower concentration -the components penetrate in deeper layers of the treated material (Rodrigues & Mauro, 2004), while higher osmotic solution concentrations lead to faster water loss (Waliszewski, Delgado & Garcia, 2002). The osmotic solution must have a low value of water activity (a w ), an acceptable sensory quality and must achieve food safety requirements. Potential solutions are sucrose, glucose, NaCl, corn syrup, starch concentrates, etc., which are selected according to the dehydrated product properties (Ahmed, Qazi & Jamal, 2016). Alternative osmotic solutions, such as maple syrup, high fructose corn syrup, honey, ethanol, etc., have been attracting more attention due to high dehydration rates, optimal rehydration capacity, and negligible sensory effect on the substrate (Qiu, Zhang, Tang, Adhikari & Cao, 2019).
During osmotic dehydration, simultaneously with the diffusion of water from the tissue into the surrounding solution there is an antidote diffusion of dissolved substances from the osmotic solution into the surface layers of the treated material (Rastogi & Raghavarao, 2004;Chiralt & Fito, 2003). Sugar beet molasses has been recognized as suitable osmotic solution due to its exceptional and complex chemical and nutritional composition. Molasses is an important by-product of sugar beet/sugar cane refining industry, remaining after no more sugar can be extracted by crystallization from the raw crop (Šarić et al., 2016). Cane and beet molasses are viscous polycomponent systems with a high dry matter content (around 80%) and are significant source of many micronutrients (in the dissolved state), especially potassium, calcium, iron, magnesium and the vitamin B complex (Sauvant, Perez & Tran, 2004). From a nutritional point of view, sugar beet molasses significantly enriches the content that is dehydrated in terms of minerals and vitamins (Cvetković et al., 2019;Lončar et al., 2015;Nićetin et al., 2015a;Nićetin et al., 2015b). Some investigations were conducted using sugar beet molasses as osmotic solution for dehydration of plant origin food Nićetin et al., 2017;Koprivica, Pezo, Ćurčić, Lević & Šuput, 2014;Mišljenović, Koprivica, Jevrić & Lević, 2011) and animal origin food Ćurčić, Pezo, Filipović, Nićetin & Knežević, 2015;Filipović et al., 2014).
In previous works, mushrooms were osmotically pretreated/treated with different solutions: sucrose, citric acid, but mostly NaCl solution (González-Pérez et al., 2019;Darvishi, Azadbakht & Noralahi, 2018;Gupta, Bhat, Chauhan, Ahmed & Malik, 2015;Doymaz, 2014), at different process parameters (concentration, time, etc.). The aim of the study presented in this paper was to investigate and mathematically model the effects of varying processing parameters applied during the osmotic dehydration of mushrooms in sugar beet molasses on basic osmodehydration responses (water loss, solid gain, dry matter content), microbiological parameters, mineral and chemical composition of treated mushrooms.

Sample preparation
Fresh mushrooms (Agaricus bisporus) were bought at a local greengrocery, after which their basic composition was determined, given in Supplementary material, Tables S1 and S2. Mushrooms were wiped with a damp cloth, cut into 5 mm-slices, after which they were immersed in sugar beet molasses diluted to the concentrations of 60%, 70% and 80% of dry matter (d.m.). Basic composition of used sugar beet molasses is given in Table 1. The sample (mushroom slices) to osmotic solution (molasses) weight ratio was 1:5. The process was performed in laboratory jars at temperatures of 25, 35 and 45 °C under atmospheric pressure, in a constant temperature chamber (KMF 115 l, Binder, Germany). The samples were stirred every 15 minutes. Processing conditions regarding stirring, intensity, duration and frequency were the same for all experimental runs, so the results could be comparable. Process duration was varied between process times of 0.5, 1, 1.5, 2, 3, and 5 hours. After dehydration treatment, the samples were rapidly rinsed with running water, absorbed from excess water and further processed for the required tests.

Calculations
Osmotic dehydration process responses of dry matter content (DMC), water loss (WL) and solid gain (SG) were calculated and presented as mean values and standard deviation of three parallel runs: (1) ( 3) where m d , m i and m f are dry matter, initial and final samples mass (g), respectively; z i and z f are the initial and final water fraction mass (g water / g sample ), respectively; s i and s f are the initial and final total solids fraction mass (g total solids / g sample ), respectively.

Response Surface Methodology
Response Surface Methodology (RSM) and Analysis of Variance (ANOVA) were selected to estimate the main effects of the process variables on osmodehydration process, microbiological, and compositional responses during osmotic dehydration of mushroom samples.
Models were fitted to the response surface generated by the experiment. The model used was function of the variables: The following second order polynomial (SOP) model was fitted to the data. Thirteen models of the following form were developed to relate thirteen responses (Y) to three process variables (X): k=1-13; where β kij are constant regression coefficients.
ANOVA and RSM were performed using StatSoft Statistica, for Windows, ver. 12 program. The model was obtained for each dependent variable (or response) where factors were rejected when their significance level was less than 95%.

Analysis of the osmotic dehydration process responses
The untreated, raw mushrooms were characterized by DMC and a w values of 8.51±0.31% and 0.941±0.04, respectively. Supplementary material, Table S1 displays the results obtained for the four osmotic dehydration process responses (DMC, WL, SG, a w ) that adequately represent the osmotic process. The maximal obtained values were: 58.33±1.11%, 0.7953±0.0158 g/g i.s. , 0.1009±0.0016 g/g i.s. and 0.811± 0.008 for DMC, WL, SG and a w , respectively. The osmodehydrated mushroom sample with the highest levels of DMC, WL and SG, and lowest level of a w , indicating the most intensive mass transfer during the process, was obtained at the highest levels of applied technological parameters (process time of 5 hours, process temperature of 45 °C and molasses concentration of 80%). By analysing the statistically significant differrence between the values of different process time and constant process temperature and molasses concentration, it can be concluded that the effect of the process time had statistically significant influence on all four osmotic dehydration process parameters (for example, comparing significance level of runs no. from 49 to 54, at p<0.05, see in Supplementary material, Table S1). This result harmonises with the findings from other similar studies which reported that the increase in immersion time led to higher moisture loss (Ispir & Toğrul, 2009;Mundada, Hathan & Maske, 2011). Figure 1 shows changes in DMC, WL, SG and a w during 5-h osmotic dehydration.
During osmodehydration, substrate is in contact with osmotic medium, which has high osmotic pressure and mass transfer is established with respect to time. As time is longerosmodehydration effect is higher with respect to the evaluated process parameters (DMC, WL, SG). It was noted that the initial process period is the most important, since mass transfer phenomena are the fastest and have highest impact on further osmodehydration flow. Studies showed that exchange of masses occurred at a fastest rate within the initial 2 h followed by a reduction in drying rate during further processing time (Ramaswamy, 2005).
From the result presented in Supplementary material, Table S1 it can be seen that the temperature statistically significantly affected mass transfer in the process, manifested via statistically significant increase in DMC, WL and SG values and statistically significant decrease of a w values. Findings related to the osmodehydration of mushrooms in molasses are similar to those from other works which reported that temperature affects the rate of osmotic mass transfer (Tortoe, 2010) and that rise in process temperature accelerates water loss, while solid uptake is less affected (Khan, 2012;Tortoe, 2010). As in cases of the two previously analysed technology parameters, the effect of molasses concentration was also statistically significant on all four osmotic dehydration process responses, where the increase in the concentration of molasses solution significantly raised the values of the three responses (DMC, WL and SG), while a w significantly decreased. Other studies reported similar findings (Ispir & Toğrul, 2009;Falade, Igbeka & Ayanwuyi, 2007). The higher concentration of osmotic solution leads to greater osmotic pressure gradients, thereby leading to higher solid gain and water loss throughout the osmotic treatment period (Mundada, Hathan & Maske, 2011).

Effect on microbiological characteristics of the treated mushrooms
The fresh mushroom samples were characterrized by log10 (CFU/g) values of: 5.54±1.33x10 -2 ; 3.36±1.67x10 -2 and 4.64±1.11 x10 -2 for TPC, TYMC and Enterobacteriaceae, respectively, (Supplementary material, Table S2). The highest reduction of present microorganisms and the lowest values of TPC and TYMC (log10 (CFU/g) of 4.25±1.43x10 -2 and 2.99±8.82x10 -2 , respecttively) are obtained at the highest levels of The effect of the process duration had statistically significant influence on the TPC and TYMC log10 (CFU/g) values, where prolonged duration of the osmotic dehydration process led to higher reduction of present microorganisms. An increase in process temperature significantly decreased TPC and TYMC. The effect of molasses concentration exerted also significant effect on both TPC and TYMC in such a way that an increase in the molasses concentration caused a significant decrease in TPC and TYMC. Prolonged duration of osmotic dehydration has led to the longer exposure of selected microorganisms to high osmotic pressure environment of osmotic solutions. Higher values of temperature and molasses concentration has led to the intensified mass transfer of the process, and reduced a w value. As a consequence, the treated samples provided unfavourable conditions for the growth and development of the tested microorganisms. Studies on the effect of osmotic dehydration parameters on pork and chicken meat microbiological profiles showed similar results Filipović et al., 2012).
Regarding microbiological safety, the osmodehydrated mushrooms are not ready-to-eat products so they are not covered by the ordinance of Comission Regulation (EC) 2073/2005. According to the other legal acts, such as the national regulation on sanitary requirements during food processing and trading (Pravilnik o opštim i posebnim uslovima higijene hrane u bilo kojoj fazi proizvodnje, prerade i prometa (2010)) it could be concluded that the detection of low number of Enterobacteriaceae in the mushrooms dehydrated in molasses indicates satisfactory hygiene of the production process.  Figures 3 and 4 depicts the changes in the content of minerals and proximal chemical composition of treated mushrooms during the osmotic dehydration in molasses. Increase in the content of minerals is a direct consequence of solid matter uptake from molasses which is abundant in these components. A decrease in the osmodehydrated mushrooms protein content is a result of nonprotein solid matter uptake (molasses is not a protein source), resulting in a relative decrease of dry matter protein content, although absolute quantities of mushrooms protein contents did not change during the osmotic dehydration process.

Effect on the chemical composition and mineral content in the treated mushrooms
The concentrations of minerals and chemical parameters in the mushrooms treated at the applied highest levels of processing parameters ( The effects of individual process parameters on mineral and chemical responses are the same as in the case of basic osmotic dehydration process parameters, since process mass transfer mechanisms are leading to compositional changes of the dehydrating samples. Individual increase of process time, temperature and molasses concentration has led to the statistically significant increase of minerals, sugars and ash, and a significant decrease of protein content. These results are in correlation with previously reported results on mineral and chemical composition of osmodehydrated fish (Lončar et al., 2015), pork meat (Nićetin et al., 2015b) and wild garlic (Šobot et al., 2019).  Increasing the temperature, process duration and molasses concentration solid gain increases that affects the chemical and nutritional composition of the final product. The driving force for osmotic water removal is the concentration gradient established on opposite sides of the cell membrane. During dehydration, two counter current flows take place: water from the material passes into osmotic solution, while the dissolved substances of the osmotic solution diffuse into the material to be treated (Rodrigues and Mauro, 2004). Molasses concentration, regarding rich chemical and nutrient composition, contributes to substrate enrichment during osmotic dehydration. Table 2 shows ANOVA results of the response surface models which were calculated from the osmodehydration process responses (DMC, WL, SG, a w ) at different time, temperature and molasses concentration. ANOVA testing showed that all four responses were statistically significantly influenced by all three technological parameters, where the parameter with the highest influence was duration, followed by temperature and molasses concentration. In all four mathematical models of osmodehydration process responses, all three linear terms (for time, temperature and concentration) were statistically significant. Statistically significant model members are marked with * in Table 2. Quadratic term for time statistically significantly contributed to the model forming of WL, SG and a w , in-dicating that exchange of masses occurred at a faster rate at the initial phase of the process followed by a reduction in drying rate during further processing time (Ramaswamy, 2005).

Analysis of the mathematical models for mass transfer during osmotic dehydration
In the case of quadratic term for temperature, it was statistically significant for DMC, WL and a w .
Interaction of time and concentration was significant in cases of all four responses, while the cross product Time x Temperature was significant in cases of: DMC, SG and a w .
Residual variance was not statistically significant in any of four tested mathematical models, and together with high values od coefficient of correlation (R 2 ) pointed that the applied models adequately represented variation in osmotic process responses within the range of applied technological parameters.

Analysis of the mathematical models for microbiological counts
ANOVA results of the response surface models calculated from the recorded microbiological responses are shown in Table 3. Response for Enterobacteriacea were not modelled, since there was not enough variation between the experimental results. ANOVA testing showed that both modelled responses (TPC and TYMC) were statistically significantly influenced by all three technological parameters, where the most influential parameter was time, followed by temperature and the least influential molasses concentration. The increase in immersion time, pro-cess temperature and osmotic solution concentration lead to higher and accelerated moisture loss until equilibrium level is achieved (Mundada et al., 2011;Khan, 2012;Tortoe, 2010).   Residual variance was not statistically significant in both tested mathematical models, and together with high R 2 values showed that applied models adequately represented variations of microbiological profile responses.

Analysis of the mathematical models for responses related to mineral and chemical composition
Significant mineral enrichment and chemical composition change in the treated mushrooms is a consequence of previously explained mechanisms of solid gain upgrowth, owing to molasses favourable nutritional composition. Tables 4 and 5 shows the ANOVA results of the response surface models calculated for mineral and chemical composition, respecttively.
The presented results show that all seven responses were influenced by all three technological parameters, in the following order of significance: time, temperature, concentration. In cases of all responses, linear terms of all three dehydration parameters were statistically significant. Quadratic term for time, together with cross product of Time x Temperature were also statistically significant in all cases. Residual variance was not statistically significant in any of the tested mathematical models. High R 2 values again indicated on good fit of proposed models to the minerals and chemical responses experimental data.

Mathematical models regression coefficients
Individual coefficients, which can be used for forming thirteen quadratic equations of models for selected responses (mass transfer parameters, microbiological counts, mineral and chemical composition), are shown in Supplemetary material, Tables S5 and 6, along with their statistical significance. Using presented equations and knowing technological parameters of the process (time, temperature and molasses concentration), values of all selected responses at the determined process conditions can be calculated. Data obtained from these models can be used for process and quality control and management of the osmotic dehydration of mushrooms in molasses.

CONCLUSIONS
From the presented results it can be concluded that osmotic dehydration of mushrooms in beet molasses and an increase in its intensity led to the significant increase in mass transfer kinetics, mineral uptake (for example, an increase in K content by 269.42%; and Ca content increase by 939.03%), changes in proximate chemical composition and a decrease in a w values (from 0.941 to 0.811), microbiological load and relative protein content (33.07% decrease). The obtained osmodehydrated mushroom samples were characterized by high levels of DMC, low a w values and favourable microbiological profile, indicating a good base for prolonged shelf life and further processing. Improved nutritive profile of osmodehydrated mushrooms owing mostly to prominent mineral enrichment due to transfer of minerals from molasses, suggest a possibility for production of new and functional products. The proposed mathematical models of changes in selected responses (mass transfer parameters, microbiological counts, mineral and chemical composition) were statistically significant, calculated and observed responses corresponded very well, indicating a satisfactory approximation of responses values within the applied technological parameters.

АCKNOWLEDGEMENTS
This paper is a result of the research within the project TR 31055 (451-03-68/2020-14/ 200134), supported by the Ministry of Education, Science and Technology, Republic of Serbia.