Development of volatiles and odor-active compounds in Chinese dry sausage at different stages of process and storage

Huimin Zhou, Bing Zhao, Shunliang Zhang, Qianrong Wu, Ning Zhu, Su Li,Xiaoqian Pan, Shouwei Wang, Xiaoling Qiao

China Meat Research Centre, Beijing Academy of Food Sciences, Beijing 100068, China

Keywords:

Chinese dry sausage

Volatiles

Odor-active compounds

Process

Storage

ABSTRACT

The effect of process and storage on the volatiles and odorant pro file of Chinese dry sausage was evaluated,as well as its physicochemical parameters. Microbial esterification and wine (27.54%-43.35%), and lipid oxidation (11.30%-34.92%) played a key role in flavor profile during process and storage. A significant increase of each volatile was detected during process except gradual decrease of volatiles from spices, while a gradual decrease of each volatile was detected during storage except significant increase of volatiles from lipid oxidation and esterification. 15 and 6 odor-active compounds were respectively correlated (P ＜ 0.05) with the process and storage time. Level of heptanal, 1-octen-3-ol, the ethyl of 2-methylbutanoic, 3-methylbutanoic,butanoic, benzoic, hexanoic, heptanoic, octanoic and decanoic acid were best discriminators of process stage,while (E)-2-nonenal, ethyl hexanoate, ethyl heptanoate, and methyl decanoate, were marker compounds of storage time. An objective method was established to evaluate the stages of process and storage for samples.

1. Introduction

Chinese dry sausage is a naturally fermented traditional dry-cured meat products in north China, and widely recognized by worldwide consumers for its special texture, taste, and flavor [1,2]. The overall acceptance of meat products depends to a large extent on their flavor,which is an important quality of meat. Changes in flavor during the process and storage of Chinese dry sausage may be monitored through the volatile pro file [3,4].

Biochemical changes (lipolysis, proteolysis, lipid oxidation,strecker degradation, and carbohydrate fermentation) take place during manufacturing and ripening of dry-fermented sausage and contribute to volatile development [5-7]. Furthermore,proteolysis and lipolysis take place by the synergistic action of endogenous enzymes and microbe enzymes activity [8,9]. These enzyme activities will be influenced by processing parameters(extrinsic and intrinsic parameters) [10-12]. Therefore, processing parameters can be varied in different ways, resulting in a wide species of sausage types with different sensory attributes during the processing. Furthermore, dry-cured sausages generally have long shelf period under retail conditions due to their low water activity, but proteolysis, lipolysis and lipid oxidation still take place during storage, causing gradual alteration of flavor and loss of acceptability, and limiting human consumption due to the formation of oxidative rancidity and off-flavors [3,13,14]. It is crucial to establish the contribution of the process stage and storage time to the volatiles pro file of Chinese dry sausage. It is necessary to establish an objective method to evaluate the different stages of process and storage of Chinese dry sausage.

Adding a small amount of sugar, soy sauce, Chinese-spices and wine into Chinese dry sausage, and combining with the process of curing, oven-drying and air-drying (10 days production cycle), may result in a specific flavor, which makes it different to Cantonese sausage with large amounts of wine and sugar [15]. Meanwhile, airdrying process is a naturally fermented process without starter culture,which makes it different from dry-cured fermented sausages, such as salami. PH and water activity (aw) of final product are respective 5.5-6.0 and 0.70-0.75, which makes it different from semi-dry sausages [1,3]. So its shelf life reaches 6 months to 9 months under retail conditions, which is mainly limited by the flavor deterioration that accompanies oxidation phenomena, since pathogenic or spoilage bacteria have difficulties to flourish in dry-cured sausages during storage [16].

It is still limited about formation and development mechanism of predominant volatiles in Chinese dry sausage during process and storage, and when these volatiles are essential for the odor. Therefore,the objective of this study was to track the change of volatiles in Chinese dry sausage during the process and storage, and also to determine which key odor-active compounds and at which stage of the process and storage was essential for the odor.

2. Materials and methods

2.1 Sausage production and sampling procedures

Chinese dry sausage formulation included pork lean meat(80 g), pork back fat (20 g), water (10 g), light soy sauce (2 g),Chinese liquor (1 g, ethanol content: 56%, V/V, produced from Beijing red star co., LTD), NaCl (1.8 g), sucrose (1 g), glucose(0.5 g), sodium glutamate (0.2 g), sodium ascorbate (0.15 g), Chinese five-spices powder (0.15 g, Jiangsu saimei food technology co. LTD), sodium nitrite (0.003 g). The pork meat was ground through a 10 diameter mincing plate and vacuum mixed together with the remaining ingredients for 10 min. The mixture was cured at 4 °C for 24 h and stuffed into collagen casing (23 mm diameter, Wuzhou Shenguan protein casing co. LTD) and tied the knot a length of 10 cm, followed by oven-drying for 3 h at 50 °C, then transferred into an air-drying chamber with 75%relative humidity at 15 °C (Binder KMF 720E6, Germany) for 10 days. Then samples were vacuum packed with packaging bag of 3 layers material (polyethylene, aluminum, polyester) and stored in chamber with 50% relative humidity at 25 °C (Binder KMF 720E6, Germany) for 270 days. Three batches samples are made and periodically taken for analyses at different stages of process and storage, including the mixing, curing, oven-drying, air-drying 2 days, air-drying 4 days, air-drying 6 days, air-drying 8 days, airdrying 10 days, storage 30 days, storage 60 days, storage 120 days,storage 210 days, storage 270 days.

2.2 Analysis of physicochemical parameters during the process

Moisture content was determined using national food safety standard methods (GB 5009.3-2016). awwas measured using an awmeter LabMaster-aw (Novasina). The pH of the sample was measured using a digital pH-meter (Sartorius PB-10, Germany).

2.3 Analysis of microbial counts during the process

In triplicate, 25 g sample taken from the mixing before stuffing or from the central part of each sausage at the different stages of process was aseptically added into 225 mL sterile solution with 0.9% NaCl. This mixture was homogenized in a SCIENTZ-11 sterile homogenizer (Ningbo xingzhi biotechnology co., LTD) for 2 min at room temperature. For microbial plate counts, serial decimal dilutions in peptone water were prepared and poured on different agar media.Counts of total mesophilic aerobic bacteria, Staphylococci, lactic acid bacteria were respectively determined on plate count agar (PCA)(Oxoid), mannitol salt agar (MSA) and de Man, Rogosa, Sharpe agar(MRS) after incubation at 37 °C for 48 h. Enterobacteriaceae counts were determined on violet red bile glucose agar (VRBGA) incubated with a double layer at 37 °C for 24 h [17].

2.4 Analysis of volatiles and odor-active compounds

Volatiles of the dry sausages were extracted by dynamic headspace sampling (DHS) and analyzed by thermal desorption-gas chromatography-olfactometry-mass spectrometry (TDS-GC-O-MS)according to the method of Zang et al. [18]. Volatiles of sausage samples were trapped in a Tenax TA tubes (17.8 cm × 0.6 cm ×0.4 cm) with Tenax TA (180 mg). The Tenax TA tubes were fitted to thermal desorption (TDS3, Gerstel, Germany) and Cold Inlet System(CIS 6). Analysis of volatiles was carried out using TRACE 1310 gas chromatograph equipped with TSQ 8000 mass spectrometer(Thermo Fisher Scientific CN, USA) and an olfactory detection port(ODP, Sniffer 9000, brechbuhler, Switzerland). The effluents were split 1:1. An ODP olfactory detector port was used for obtaining the aroma profile. The interface temperature of the ODP was set at 200.0 °C. Moist nitrogen was pumped into the sniffing port continuously to prevent the assessors from drying the nasal cavity. Each ODP user aged between 20 and 35 years was trained to recognise odours by using the treated sample and standard aroma compounds. The odor-active compounds perceived by the 3 assessors were recorded as the time for the onset and end,odour characteristics and intensities of the aroma extracts. The description and the time of each aroma-active compound were con firmed consistently by at least 2 assessors.

The identification of the isolated volatile compounds was achieved through retention index (RI) and libraries (NIST 08). The retention index of volatiles was determined based on the retention times of n-alkanes (C8-C20) [19]. Quantitative results, expressed as μg/kg,were obtained by comparison with the area of internal standard. The contribution of each odorant was evaluated by OAV, which was measured as the ratio of the concentration of each compound to its detection threshold in water [20]. Odorants with OAVs ≥ 1 in the sample were considered to be odor-active compounds and were listed in Table 3.

2.5 Statistical analysis

Statistical analysis was carried out using the statistical package SPSS 20.0 (SPSS Inc., Chicago, IL). One-way analysis of Variance(ANOVA) was carried out to investigate the effect of process and storage time on volatiles, physicochemical parameters and microbial counts using Duncan’s test. The between-variable correlation was carried out to investigate the correlation of volatiles, physicochemical parameters, microbial counts, process time and storage time using Pearson’s test. Principal component analysis (PCA) and hierarchical cluster analysis (HCA) were carried out to evaluate the influence of process and storage on the odor-active compounds of dry sausage.

3. Results and discussion

3.1 Changes in physicochemical parameters during the process

Values corresponding to basic physicochemical parameters throughout process were shown in Table 1. Moisture content, awand pH values did not show significant differences during the curing and oven-drying, with value around 52.78%-55.25%, 0.94 and 5.99,respectively. However, the moisture content and awsignificantly decreased during air-drying process, reaching to 23.54% and 0.78 at the final product, respectively, which meet the storage requirements.PH values decreased significantly (P ＜ 0.05) in the early period of airdrying, reaching the lowest value of 5.64 at 6 days of air-drying. A significant increase was observed at 10 days of air-drying with value 5.79. The reason was that a large number of lactic acid produced from lactic acid bacteria leaded to a rapid reduction of pH during the early stage of air drying, while decrease of awin sausage limited the reproduction of lactic acid bacteria during later stage of process, and at the same time, some alkaline substances produced from proteolysis resulted in a increase of pH. Chinese dry-cured sausage (aw= 0.85,pH = 5.95) and Chinese smoked-cured sausage (aw= 0.83, pH = 5.73)had similar values while salami had higher awvalue (0.93) and lower pH (5.02) [21].

目前，我國企業(yè)預(yù)算管理理念落后，出現(xiàn)了一定的偏差，不能與現(xiàn)代企業(yè)管理觀念同步，制約了企業(yè)預(yù)算管理水平的提升和進(jìn)步，阻礙了預(yù)算管理對企業(yè)的助推作用，傳統(tǒng)的企業(yè)預(yù)算管理內(nèi)容比較復(fù)雜和混亂，與財務(wù)管理的內(nèi)容相差不多，許多企業(yè)管理者的管理觀念落后，認(rèn)為財務(wù)預(yù)算是企業(yè)的預(yù)算管理，但實際上預(yù)算管理和財務(wù)預(yù)算有著很大的區(qū)別，二者之間是包含關(guān)系。全面預(yù)算管理這不僅僅是財務(wù)預(yù)算。它包括的內(nèi)容更多，與企業(yè)經(jīng)營的各個環(huán)節(jié)都有一定的聯(lián)系，如投資、經(jīng)營、經(jīng)營、市場調(diào)整等。

3.2 Changes in microbial counts during the process

Viable counts of total mesophilic aerobic bacteria on PCA, staphylococci on MSA, lactic acid bacteria on MRS and enterobacteriaceae on VRBGA significantly increased (P ＜ 0.001)during the air-drying except for the day 10 of air-drying (Table 1). In the whole processing, lactic acid bacteria counts were the highest and played a dominant factor, followed by staphylococci. It was observed that evolution of lactic acid bacteria and total mesophilic aerobic bacteria have similar behavior and Pearson correlation coefficient was 0.991 (P ＜ 0.000 1). Similar behaviour was reported by Lorenzo et al. [22] and Domínguez et al. [23] when studying dry sausage samples. Compared to Chinese dry sausage, higher populations of PCA, MSA and MRS at the end of the process were found in drycured sausage from Italy, Germany and Spain [22,24,25]. The cause of this result may be Spanish-style dry sausage with a longer cycle of ripening (50-60 days). In addition, enterobacteriaceae enumeration was performed in order to evaluate the contamination level throughout the process. Then VRBGA counts decreased significantly (P ＜ 0.05) at curing process, mainly due to the addition of salt and wine, or the curing temperature (4 °C). However, the cause of the increasing of enterobacteriaceae counts in the airdrying process might be low acidification insufficient to restrain them. Furthermore, researchers found that the salami had a microbial safety with trace enterobacteriaceae population [1]. On the contrary,a larger enterobacteriaceae population were checked in Chinesestyle sausages, probably due to spontaneous fermentation technology.Domínguez et al. [23] also revealed that the use of starter cultures could increased lactic acid bacteria counts and the total viable counts, which completely reduced the enterobacteriaceae counts compared with spontaneous fermentation batch and improved the hygienic quality.

3.3 Volatiles and odor-active compounds

A total of 76 volatiles were identified in Chinese dry sausage during the whole process and storage by DHS-GC-O-MS (Table 2). However, only 29 volatiles (10 esters, 4 aldehydes, 3 alcohols,1 acid, 11 hydrocarbons) were detected during whole process and storage. As the process increased, the total volatiles contents of Chinese dry sausages showed a dynamic change. It increased significantly(P ＜ 0.05) during the first 120 days of storage, up to 4 715.47 μg/kg, then decreased insignificantly (P ＞ 0.05) until the end of the storage. Food matrix components, such as proteins and lipids, are known to interact with flavour compounds, which can influence the retention of aroma compounds. The increase in fat content induces an decrease in flavour release of the high-molecular weight compounds in dry sausage [12].

These findings gave an indication of the physicochemical and metabolic processes that occur during manufacture and storage. To understand the formation and deterioration of flavor in Chinese dry sausages, volatiles were grouped according to their possible origin:lipid oxidation, microorganism esterification and wine, amino acid metabolism, carbohydrate fermentation, spices, and unknown origin and contaminants [4,12,26]. Moreover, their interactions that involve many primary and secondary metabolic pathways should be taken into consideration. As process time increased, significant increase of each volatile was detected excepting for a gradual decrease of volatiles from spices, which proceeded from microbial esterification and wine (29.54%at the mixing and 37.33% at the final product), lipid oxidation (11.30% at the mixing and 24.38% at the final product), carbohydrate fermentation(7.32% at the mixing and 12.00% at the final product), spices (20.29% at the mixing and 10.13% at the final product), and amino acid catabolism(4.61% at the mixing and 7.22% at the final product) (Fig. 1). As storage time increased, gradual decrease of volatiles was detectedexcept significant increase of volatiles from lipid oxidation, microbial esterification and wine, which proceeded from microbial esterification and wine (42.7%), lipid oxidation (34.92%), carbohydrate fermentation (5.73%), spices (5.08%), amino acid catabolism (4.07%)at the end stage of storage (Fig. 1.).

Table 1Evolution of physiochemical parameters and the counts (lg (CFU/g)) of total mesophilic aerobic bacteria (PCA), staphylococci (MSA), lactic acid bacteria (MRS)and enterobacteriaceae (VRBGA) throughout the process of Chinese dry sausage (means ± standard deviations. n=6).

3.3.1 Volatiles from microorganism esterification and wine

Microorganism esterification and wine could be the cause of the generation of the 27 volatiles during process and storage(Table 2), comprising 26 esters and alcohol. The content of esters increased significantly 2-8 times from 4 days of air-drying to 120 days of storage, while alcohol added at the beginning of process remarkably decreased from 405.98 μg/kg at mixing to 26.55 μg/kg at the end of storage. Pearson correlation test showed that ethyl esters significantly negatively correlated with alcohol (r= -0.862,P＜ 0.01)and significantly positively correlated with lactic acid bacteria(r= 0.760,P＜ 0.05) andStaphylococcus(r= 0.817,P＜ 0.05). So attention should be given to the contribution of microorganisms and alcohol. Previous researcher had extracted esterase fromLactobacillus plantarum,Staphylococcus xylosusandStaphylococcus carnosus[9,27]. So study predicted that the ethyl esters derived from esterification reaction of alcohols and carboxylic acid by esterase from microorganism during air-drying stage and storage within 120 days for Chinese dry sausages. Esterification between ethanol and carboxylic acid also happens at 45-50 °C temperature [15]. High amounts of ethyl esters indicated that microorganism esterification and alcohol played a critical role in the formation of characteristic flavor of Chinese dry sausage. Ethyl esters were responsible for the characteristic aroma of Chinese dry sausages and masking off-odor due to low threshold values and strong fruity odors.

3.3.2 Volatiles from lipid oxidation

Twenty volatiles formed via lipid oxidation, such as aliphatic aldehydes, alcohols, acids and aliphatic hydrocarbons were detected in the present study (Table 2). In addition, the C6-C10linear chain carboxylic acids in esters were also possible from lipid oxidation or lipolysis.

Aldehydes are probably the most interesting lipid-derived volatiles, which can produce a wide range of odors. The contents of aldehydes changed insignificantly (P＞ 0.05) during the whole process,but increased markedly (P＜ 0.05) from 60 days to 120 days of storage, then kept constant until the end of the display period (Table 2).These results were agreed with those reports by Lorenzo et al.[3]. Octanal, derived from oleic acid autooxidation, was the most abundant aldehydes during whole process, which disagreed with study of Ramírez et al. [28]. Its content increased primarily during the cured and oven-drying period, reaching the maximum (110.3 μg/kg) at 2 days of air-drying, then a rapid decrease was observed during later stage of air-drying and 60-120 days of storage. Pearson correlations showed that octanal content was negatively related with octanoic acid(r= -0.640,P＜ 0.001) and ethyl octanoate (r= -0.595,P＜ 0.001).So the finding indicated that a large amount of octanal was oxidized to form corresponding acid and ester in the later stage of air-drying and storage period. Contrary to octanal, hexanal and nonanal respectively derived from oxidation ofn-6 andn-9 polyunsaturated fatty acid,such as linoleic and arachidonic acids, and their rapid increase during early storage were only observed. The content of heptanal increased continuously during air-drying and storage stage. Heptanal possesses a characteristic fat and rancid smell, and could be considered as a good indicator of the oxidation level in the Chinese dry sausage.(E)-2-nonenal, which imparts a paper and fat odor, was detected until 210 days of storage, and might lead to flavour deterioration.In this study, these aldehydes were both perceived by sniffers and contributed a grass, tallow, and fat odor to the overall flavor of Chinese dry sausage. They exhibit a pleasant flavor at low concentrations, whereas most of their odor are not acceptable at high concentrations. Therefore, development of these aldehydes is directly related to the flavor deterioration of the Chinese dry sausage [18].

Table 2Contents of volatiles at different stages of process and storage in Chinese drysausage.

Table 2 (Continued)

3.3.3 Volatiles from spices

Spices of Chinese dry sausage could be responsible for the generation of the 10 volatiles comprising terpenes, aldehyde and alcohols. These volatiles can also result from animal diet, but they mainly came from the spices, such as zanthoxylum (linalool,limonene and α-terpineol), cinnamon (cinnamaldehyde, benzaldehyde,phellandrene, linaloo, terpinen-4-ol and α-terpineol), ginger (zingiberene,β-phellandrene, α-curcumene and bisabolene), nutmeg (α-pinene,β-pinene and limonene) and clove (β-caryophyllene) [30-32].Except for the D-limonene, eucalyptol and linalool, all volatilesspices decreased (P ＜ 0.05) during process and were not detected at the end of the process and whole storage, so they contributed little to the aroma of sausage. Content of D-limonene was stable during process and storage, which aligned with Lorenzo et al. [12].(E)-cinnamaldehyde, presenting a special cinnamon aroma, was the main ingredient of Chinese-spice cinnamon [33]. Its contents also decreased significantly from 96.23 μg/kg to 6.24 μg/kg during process and was not detected until 60 days of storage, which disagreed with Chen et al. [34]. In addition, cinnamate acid in ethyl cinnamate might derive from oxidation of cinnamaldehyde, because cinnamaldehyde content was positively related with ethyl cinnamate (r = 0.443,P ＜ 0.01). The founding indicated that (E)-cinnamaldehyde was oxidized into corresponding acid, then converted to corresponding ester by microbial enzyme activity [33]. These founding indicated that spices not only had seasoning function, but also had antioxidant activity.

3.3.4 Volatiles from carbohydrate fermentation

Two carbohydrates fermentation-derived volatiles, acetic acid and butanoic acid, were respectively found at the oven-drying and beginning of the process, then their contents increased to maximum at 8 days of air-drying, which was in accordance with development of the pH values and microbial counts. In addition, content of butanoic acid increased only significantly from 30 days to 120 days of storage.Organic acid might be produced by Lactobacillus and Staphylococci and also by lipid oxidation and degradation [35]. The acetic acid in ethyl acetate and butanoic acid in ethyl butanoate might be also from carbohydrate fermentation by microorganisms.

3.3.5 Volatiles from amino acid catabolism

Among amino acid catabolism derived volatiles, 3-methyl-butanal,acetophenone 2-methyl-1-propanol, and phenylethyl alcohol were detected in this work. In addition, branched-chain acids or branchedchain alcohols in esters might derive from oxidation or reduction of the corresponding aldehydes. These branched-chain aldehyde,such as 2-methylpropanal, 2-methylbutanal, 3-methylbutanal were degradation products of valine, isoleucine, leucine, respectively via Strecker degradation or by microbial degradation. The microbial degradation is more possible in this study since the Strecker degradation requires heat or long ripening period and is facilitated by relatively low water activity [34,36].

Surprisingly, contents of acetophenone, 2-methyl-1-propanol, and phenylethyl alcohol significantly decreased from the mixing stage to air-drying stage. In addition, 3-methylbutanal was only detected until the 120 days of storage. But contents of their corresponding ethyl esters, increased significantly at the air-drying stage, especially 6-10 days of air-drying. The founding indicated that branched-chain aldehyde might be easily oxidized into corresponding acid, then immediately converted to corresponding ester by microbial activity. Beck[37] also demonstrated that α-ketoisocaproic acid was biotransformed to 3-methylbutanal which immediately oxidize into 3-methylbutanoic acid.This con firmed that amino acid catabolism also played a role on flavor formation of sausage at the air-drying stage.

Phenylalanine was precursor of phenylethyl alcohol,acetophenone, benzoic acid in ethyl benzoate and phenylacetic acid in ethyl phenylacetate. Benzeneacetaldehyde was reported to provide a floral aroma and derived from the deamination of 2-phenylethylamine.Ethyl benzoate increased significantly from 4 days of air-drying to 120 days of storage. However, ethyl phenylacetate was detected until 120 days of the storage, and Pearson correlations showed that it was negatively related with acetophenone (r = -0.401, P ＜ 0.05) and phenylethyl alcohol (r = -0.392, P ＜ 0.05). The founding indicated that their conversion between them was mediated by enzyme activity during the storage.

3.3.6 Contribution level of odor-active compounds to flavor of Chinese dry sausages during process and storage

Odor-active values (OAVs) in water of the volatiles throughout the process and storage were calculated and 24 odor-active compounds with OAVs greater than one or applying by olfactometry techniques were shown in Table 3. From the beginning of process, 15 odoractive compounds, including esters, aldehydes, alkene showed OAVs above 1, which were mainly from wine and spices added to sausage at the preparation stage. Some of them, ethyl esters OAV increased during the air-drying process and storage, while lipid oxidationderived aldehydes OAV, mainly increased during storage 30-120 days. In the middle of the process, the other 6 compounds showed OAVs higher than 1, such as heptanal, ethyl 2-methylpropanoate,ethyl 2-methylbutanoate, ethyl decanoate, eucalyptol and 1-octen-3-ol.All of them, with the exception of eucalyptol, also increased during the air-drying process or even early stage of storage. Finally, 17 odor-active compounds were the contributor of overall aroma in the final product (drying 10 days). The main 9 compounds that strongly contributed to the overall aroma were those with OAVs ＞10, such as ethyl hexanoate ＞ ethyl butanoate ＞ ethyl 3-methylbutanoate ＞ethyl 2-methylbutanoate ＞ octanal ＞ ethyl 2-methyl propanoate ＞nonanal ＞ d-limonene ＞ ethyl octanoate. They contributed to the smell of fruit, fat, mint and flower. In addition, 8 compounds with OAVs ＞ 1,such as ethyl heptanoate ＞ eucalyptol ＞ linalool ＞ hexanal ＞ethyl decanoate ＞ ethyl benzoate ＞ heptanal ＞ 1-octen-3-ol, could be important contributor for the smell of grass, fruit, flower, rancid and mushroom. Without taking into consideration of the contribution of aroma compounds derived from spices and wine, the odor-activity compounds were in accordance with results from Olivares et al. [38].

Twenty-one odor-active compounds were the contributors to overall aroma of Chinese dry sausage vacuum stored 270 days.The main 13 compounds were those with OAVs ＞ 10, such as ethyl hexanoate ＞ (E)-2-nonenal ＞ ethyl butanoate ＞ octanal ＞ethyl 3-methylbutanoate ＞ ethyl 2-methylbutanoate ＞ ethyl 2-methyl propanoate ＞ nonanal ＞ ethyl octanoate ＞ ethyl decanoate ＞D-limonene ＞ ethyl heptanoate ＞ eucalyptol. In addition, 7 compounds with OAV ＞ 1 could be important aroma contributor, such as 1-octen-3-ol ＞ heptanal ＞ linalool ＞ 3-methylbutanal ＞ hexanal ＞ethyl benzoate ＞ methyl caprate. (E)-2-Nonenal and methyl caprate,new members of odor-active compounds in Chinese dry sausage, were detected after vacuum stored 120 days, exhibiting paper and wine odor, respectively. So they contributed significantly to the formation of the off-flavor after storage and are potential indicators of off flavor.

Table 3Odor activity values (OAVs) in water of odor-active compounds detected in Chinese dry sausages at different stages of process and storage (mean, n = 6).

The OAVs of 15 and 6 odor-active compounds were correlated(P ＜ 0.05) with the process and storage time, respectively. OAVs of heptanal, 1-octen-3-ol, ethyl of 2-methylbutanoic, 3-methylbutanoic and benzoic acid, ethyl butanoate, and ethyl of hexanoic, heptanoic,octanoic and decanoic acid were the best discriminators of process stage in Chinese dry sausage, while OAVs of (E)-2-nonenal,1-octen-3-ol, ethyl hexanoate, ethyl heptanoate and methyl decanoate were marker compounds of storage time.

3.4 The identification of samples through odor-active compounds

Principal component analysis was performed on Chinese dry sausage samples during proces and storage with 24 odor-active compounds. The results showed that the 2 principal components explained 79% and 14% of the variance, respectively (Fig. 2). In the score plot (Fig. 2A), a clear separation between different stage samples could be observed, except for samples S60 and S120, indicating that the 24 odor-active compounds represented the differences between samples of different stages. Samples at the stage of process (P0-P7)were separated from samples at the stage of storage (S30-S270)on PC1, while samples at the early stage of process (P0-P4)and at the later stage of storage (S210-S270) were separated samples at the later stage of process (P5-P7) and at the early stage of storage (S30-S120) on PC2, meaning that PC1 and PC2 were the main factors that contribute to differentiate the samples at different stage of process and storage. From Fig. 2B,O4, O7, O8, O13, O20 and O24 were placed on the negative side of PC1 while the rest 18 variables on the positive side of PC1. It could be observed that samples at the stage of process were highly influenced by O4 (octanal), O7 (E-2-decenal), O8(E-cinnamaldehyde), O13 (3-methylbutyl acetate), O20 (methyl salicylate) and O24 (D-limonene). Samples at the early stage of storage (S30-S120) were characterized by O2 (hexanal), O9(ethyl 2-methylpropanoate), O10 (ethyl butanoate), O11 (ethyl 2-methylbutyrate), O12 (ethyl 3-methylbutyrate), O15 (ethyl heptanoate), O16 (ethyl octanoate), O19 (ethyl benzoate),O21 (eucalyptol) and O23 (linalool), while samples at the later stage of storage (S210-S270) were characterized by O1(3-methylbutanal), O3 (heptanal), O5 (nonanal), O6 (E-2-nonenal), O14 (ethyl hexanoate), O17 (methyl decanoate),O18 (ethyl decanoate) and O22 (1-octen-3-ol). So it indicated that esterification, carbohydrate fermentation, amino acid catabolism and lipid oxidation were the main contribution to the flavor of samples during first 120 days of storage, while lipid oxidation and esterification were the main contributions to the flavor of samples at the 210-270 days of storage.

Fig. 2 PCA results of the 24 odor-active compounds. (A) PCA score plot depicting the distribution of the samples for the first two principal components (PC1 and PC2). P0: mixing; P1: curing; P2: oven-drying; P3:air-drying 2 day; P4: air-drying 4 day; P5: air-drying 6 day; P6:air-drying 8 day; P7: air-drying 10 day; S30: storage 30 day; S60: storage 60 day; S120: storage 120 day; S210: storage 210 day; S270: storage 270 day.(B) PCA loading plot depicting the distribution of the 24 key volatile compounds for the first two principal components (PC1 and PC2).O1-O24: 24 odor-active compounds shown in Table 3.

The 24 odor-active compounds were clustered by twodimensional hierarchical cluster analysis (HCA) to further analyze the differences of Chinese dry sausage during process and storage, which led to 5 clusters (Fig. 3). The first cluster consisted of the 5 samples of P0 -P4. The second cluster consisted of 4 samples of P5 -S30. Samples in P0-P4 had higher relative contents on 20 (methyl salicylate), 7(E-2-decenal), 13 (3-methylbutyl acetate), 8 (E-cinnamaldehyde),24 (D-limonene) and 4 (octanal）from spices, wine and lipid oxidation, whereas samples in P5-S30 had higher relative contents on 9 (ethyl 2-methylpropanoate), 10 (ethyl butanoate), 11(ethyl 2-methylbutyrate) and 12 (ethyl 3-methylbutyrate）from esterification, carbohydrate fermentation and amino acid catabolism,which resulted in the difference between early-stage and later-stage of process. The third cluster consisted of two samples of S60 and S120. The sample of S210 and S270 were separated individually in the dendrogram from the other samples. Compared with S30, samples in S60-S120 had higher relative contents on 19 (ethyl benzoate), 2(hexanal), 3 (heptanal), 1 (3-methylbutanal), 18 (ethyl decanoate), 16(ethyl octanoate), 17 (methyl decanoate) and 5 (nonanal), whereas samples in S210-S270 had higher relative contents on 6 (E-2-nonenal), 22 (1-octen-3-ol), 14 (ethyl hexanoate) from esterification and lipid oxidation, which resulted in the difference between earlystage and later-stage of storage. The results also showed that 14(ethyl hexanoate) of the 24 odor-active compounds was individually clustered as a class, and the rest were clustered as a class. Compared with PCA, a two-dimensional HCA heat map was a more efficient and accurate way to grouping the samples through differences in odoractive compounds.

Fig. 3 Two-dimensional HCA heat map of the 24 odor-active compounds of samples at the different stages of process and storage by using the average clustering method of coupling. 1-24: 24 odor-active compounds shown in Table 3.

4. Conclusions

The majority of volatiles were ethyl esters and aldehydes in Chinese dry sausage at different stages of processing and storage.Microbial esterification and wine (37.33%), and lipid oxidation(24.38%) played a key role in Chinese dry sausage flavor formation.Carbohydrate fermentation (12.00%), spices (10.13%), amino acid catabolism (7.22%) also played an active role in the formation of flavor. A significant increase of each volatiles was detected during process except for a gradual decrease of volatiles from spices. As storage time increased, a gradual decrease of each volatiles percentage was detected except significant increase of volatiles from lipid oxidation and microbial esterification. The OAVs of 15 and 6 odoractive compounds were respectively correlated (P ＜ 0.05) with the process and storage time, which were best discriminators of process and storage time in Chinese dry sausage. Therefore, they offer good possibilities for use as marker compounds of flavor acceptability and deterioration. Compared with PCA, two-dimensional HCA was more efficient to determine the differences in odor-active compounds for sample at the different stages of process and storage.

conflicts of interest

All authors certify that there is no conflict of interests.

Acknowledgments

The authors were very grateful for the financial supports from national key research and development program of China (No.2017YFD0400105).