Sheep are one of the important high quality human protein resources, sheep meat production has been steadily increasing globally along with the world purchase power parity in GDP per capita during the past 20 years, according to the Food and Agriculture Organization of United Nation (FAO) and the World Bank (WB) data set. To some extent, higher quantity of mutton consumption reflects a higher standard of living. This trend is pronounced in China as the consumption level of urban residents has continued growing based on the National Bureau of Statistics of China. The production of mutton in China has increased rapidly while the pork decreased gradually, beef increased slightly, especially after 2014. Moreover, consumers preferred the native lamb or mutton produced on pasture due to the perception that meat could be different and healthier when reared in extensive grazing on a variety of grasses, herbs and browse compared to products obtained from an indoor system [1]. However, what needs to be compromised is that overgrazing has been a severe problem worldwide, especially in the arid and semi-arid regions. To ensure the sustainable development of mutton sheep industry, China began to import forage products in large quantities after 2010, and the import scale has reached 550 million US dollars in 2019 (FAO). In addition, fattening lambs in feedlots with a concentrate-included diet after nursing their mother on pasture, i.e. intensive pasture-based fattening system is an alternative choice when there is limited grassland carry capacity and insufficient forage resources, in which lambs could be intensively fed after weaning with the advantage of accelerating commercialization and producing well-finished carcasses [2].

Hulunbuir, as part of the Eurasian steppe, with a total area of about 997.3×106 hm², is an important animal husbandry region in China, where the Chinese indigenous sheep breed, Hulunbuir sheep (HBS), is raised as the main farm animal by both family farms and private-purpose production. Apart from be able to adapt the local harsh natural environment, the most outstanding characteristic of the HBS is incomparable meat quality with natural aroma with no unpleasant flavor. The problem faced currently in Hulunbuir mutton sheep industry, is the double restrictions of overgrazing and insufficient forage, which calls for profound changes in practices to adopt the indoor system with the pasture grazing benefit [3]. Although fattening lambs with concentrate supplement is known to increase carcass yield, fatness and growth performance, fattening lambs on pasture was recognized as significantly affecting the sensory properties of the cooked meat [4], and leading to greater deposition of conjugated linoleic acid (CLA) which is an essential nutrient for health compared to grain-finishing lambs [5,6].

Recent studies showed that besides growth rate and car cass characters the production system might also affect the meat quality [7]. On the issue, there is inadequate information on the effect feeding regimens have on sheep meat quality especially in flavour, and little is known about the effect of combining of high quality pasture with concentrates on HBS. Whether the concentrate-based finishing would injure the meat flavor of HBS is acknowledged as an important issue, which would influence the acceptability of the sheep meat products [8,9].

The aim of this study was to comprehensibly compare sheep meat quality between the pasture finishing and concentrate-included fattening after pasture weaned for HBS. Specifically, focus on changes in slaughter traits, physical measurements of meat quality, fatty acid profile and the content of compounds responsible for flavor. The results are imperative to formulating improved sheep meat production in Hulunbuir.

Although the sheep meat production systems varied from diversified pasture to feedlot by grain feeding among production countries, feeding with grain based rations in confined barn has been a trend for finishing lambs, especially for temperate-cultivated pastures, in which intensive grazing or grazing plus supplementation were usually adopted. This system is already very common in dairy cow and beef cattle, but relatively less in mutton production because of variety of conditions in sheep, like different digestive physiology and difference in forage quality preferences [14]. The diets of feeding strategies designed with verification were not recommended for sheep nutrition and were not abundant as those for dairy and beef cattle [15]. It was necessary to formulate suitable sheep meat production system conditioned to local climate and grass management.

In this fattening study, lambs in the CON group presented higher LW and HCW than ones grazing on pasture in the same periods, followed by a higher DP, indicating that concentrate supplementary fattening in a sheep pen after pasture weaned exhibit a better meat production performance. The result was consistent with the other similar researches about indoor fattening experiments [16] and also in accord with the result of pasture grazing steers, which may be less effective than other diets for beef production because of a decrease in daily gain and low productivity [17]. The difference in growth performance between the two finishing systems suggested that changes in the rearing environment and forage affected the growth performance profoundly. Compared to the condition of the indoor feeding system, the outdoor sheep will spend more time walking due to greater available space and to investigate the new environment, thus consuming additional energy.

Generally, the physical measurements for meat quality ef fected by different production systems related to the levels of gross energy (GE), crude protein (CP) in forage and interaction with the genetic propensity of animals for muscle and fat deposition [4]. The lambs fed in indoor system would produce meat with higher fat content than those with forage based diets at the same LW [18]. In the present study, we observed higher fat content in the carcass of the CON group during the slaughter process because of the higher GE and CP of TMR than the mixed grass in the pasture (Table 1).

According to the perception of consumers, meat color is an important and visual mark used to estimate the quality and freshness of sheep meat at the point of purchase [19]. Unlike lightness and yellowness, redness has a stronger correlation with meat acceptance. When the value of redness is above 14.5, the fresh meat is accepted by above 95% of consumers [20]. In the current study, the mean value both in PAS and CON reached 19.85 and 22.56 respectively, demonstrating the excellent marketing appearance of Hulunbuir lamb. In terms of lightness, the lambs in the CON group presented very significant higher value than that in PAS. It was also in line with the previous studies which had reported higher meat lightness in lambs raised on concentrate-based system than the counterpart in the pasture [21]. Being the color of meat was affected by fat deposition and oxidation process [5], the reasons for a higher lightness of lambs in CON can be attributed to the more fat content within IMF even though the difference did not reach a significant level, and might be induced by less oxidation in metabolism process due to more idling time and less physical activity in housing than outdoors [22]. This inference was also in keeping with the observation of Caneque et al [23], who had noted a higher lightness value in LD of dry lot lambs compared to pasture lambs and also considered it to be due to a different physical activity level.

The ultimate meat pH values (pH ₂₄) of the two groups were between 5.7 to 5.9 which could be considered as the acceptable quality range [24], but the lambs from PAS group had a significantly higher pH value than ones from CON group (p<0.05). Some studies have showed that the ultimate meat pH of lambs fed in extensive conditions is higher than that of lambs subjected to intensive fattening [25]. By contrast, other studies have reported that different feeding systems had a limited or no effect on meat pH value in sheep and goats [15,16]. Apparently, our study was in line with the former observations. Before those inconsistent observations above, a study found that cattle fed low energy rations had lower muscle glycogen contents when compared to one fed high energy rations [26]. A subsequent study reported that high energy rations had a positive effect on glycolysis and post mortem reductions in pH by increasing muscle glycogen reserves and fat thicknesses in sheep and cattle carcasses [27]. Hence, the difference in the value of pH₂₄ to differences in muscle glycogen reserves and energy levels was connected to feeding systems, low muscle glycogen reserve associated with low energy supply [28]. Compared to the energy intake of the CON group, the lambs in the PAS group gained less energy because of lower GE in the ration (Table 1), and muscle glycogen reserves would be depleted more quickly. We speculated that the lower deposition of glycogen reserves in muscle from PAS groups was the reason for high pH value. In addition, WHC may be affected by rations. Our study was also in accord with the previous reports that lambs fattened intensively with concentrate to have higher WHC than ones finished on pasture [29].

The tenderness of meat is another primary consumption character and is affected by the IMF directly. Even though there was no significant difference in IMF between the lambs from CON and PAS group, a higher value was observed in CON lambs, and the meat fed with forage inclusion with concentrate was more tender indicated by the significantly lower value of WBSF. This observation was also consistent with the previous reports [16]. Many studies have reported that the fatness of carcass influences the WBSF and it decreases with the increase of carcass fatness because of the reduction of connective tissue strength [30,31]. Meanwhile, some studies reported that a greater growth rate of lambs may produce an increase in soluble collagen and then result in more tender meat [32]. In the present study, lambs in CON had a higher growth rate than that in PAS and had a higher content of IMF, which may be a feasible explanation for more tender in CON lambs.

It is a consensus that consumers of sheep meat usually place the most weight on flavor followed by tenderness and juiciness [4], therefore, fatty acid profile plays an important role in the definition of meat quality as well as with differences in organoleptic attributes, like taste and undesired sensory characteristics [33]. Many reports showed that the feeding system could alter the fatty acid profile on beef and lamb [5–7]. Our study has substantiated it again from subcutaneous fat other than from muscle. From the results presented in the study, the most abundant fatty acid was OA (C18:1) (31.63% to 36.91%), then PA (C16:0, 24.61% to 27.54%) and SA (C18:0, 18.84% to 20.65%) in both groups of production systems. The total percentages of the three dominant fatty acids in the CON and PAS group were 78.01% and 82.17% respectively, which were in line with the data of previous research [34–36].

For ruminants, the forage degradation in their rumen is a complex and special process involving multiple activities with microorganisms, where the microbiota synthesize long chain n-3 fatty acids through their enzymatic action from the α-linolenic acid precursor when in a pasture feed system is offered. However, the biohydrogenation pathway affected through shifting trans-vaccenic acid production to C18:1 trans-10 when foraging includes a concentrate or grain finished either in lamb or in beef [37–39]. As shown in Table 1, the pasture has a higher content of C18:3, and this might be the reason that both OA (C18:1) and LA (C18:2) were highly significantly increased in the CON group compared to PAS group in our study. We then speculated there was intensive hydrogenation activity through microbiota in the rumen which was responsible for the formation of higher USFA like CLA while when forage includes a concentrate the result is a significantly higher value of USFA than pasture finished system, and in turn, improving the sheepmeat quality to benefit the human health. It is a plausible explanation that SFA is inclined to decrease in the CON group because the total amount of fatty acid is limited and would favor producing more USFA, resulting in a significantly decreased amount of SFAs like MA (C14:0) and PA (C16:0). Apparently, lambs finished with forage inclusion concentrate had a better fatty acid profile compared to the lambs merely pasture feeding.

The formation of short chain BCFAs was regarded as the main contributor to mutton flavour, which is the species-specific odor and the most important factor for consumers accepting lamb or mutton [4]. A recent study reported that BCFAs were positively correlated with the indexes such as consumer lamb flavor intensity scores, flavor liking, overall flavor impact and odor impact, which means the higher amount of BCFAs the lower acceptability of consumers [9]. The most notable BCFAs were MOA, EOA, and MNA, present in ovine fat and associated with “mutton” flavour [4,13]. Other compounds, such as MP and MI, accumulate in fat reservoirs and were associated with distinct “pastoral” flavor in sheepmeat [40]. Both “mutton” and “pastoral” flavor were found to be undesirable odor recognized by consumers, and higher concentrations of these compounds have been observed when receiving a grain based finishing diet [40]. In the current study, with concentrate inclusion in the ration, the content of MOA in lambs was above 4-fold higher than that of PAS lambs, and EOA was above 2-fold. While MNA could not even be detected in lambs of the PAS group but was observed in lambs of the CON group by a mean value of 20.87 μg/g. Apparently, whether MOA, EOA, or MNA have increased several times in the group of CON, which was in agreement with the observation aforementioned [4]. Compared with the lambs in PAS, lambs in CON resulted in intensified “mutton” flavor.

As BCFAs, MOA, EOA, and MNA can be de novo synthesized in adipose tissues [41]. In general, fatty acid synthesis is primarily mediated by the acetyl CoA carboxylase (ACC) and fatty acid synthase (FASN), whereby acetyl-CoA is utilized by FASN to initiate a new acyl chain, is subsequently elongated using malonyl-CoA generated by ACC [42]. Moreover, FASN is able to incorporate methylmalonyl-CoA or ethylmalonyl-CoA units instead of malonyl-CoA to produce methyl or ethyl branches on even-numbered carbon atom in the less efficient condition [43]. Methylmalonyl-CoA and ethylmalonyl-CoA are then produced in the cytosol as side-products by ACC using propionyl-CoA and even butyryl-CoA [44]. For ruminants, the rumen microbes can produce volatile fatty acid (VFA) including propionate and butyrate, which are chiefly absorbed by the epithelium of the rumen and of the omasum [45]. From the data of Table 1, the diet of CON group contained more than 3-fold ether extract (crude fat) than that of PAS, which might be a primary factor for the higher content of BCFAs in the CON group.

Since the amount of MP and MI in the subcutaneous fat of HBS was too low to be detected accurately, their results, therefore, were demonstrated with the relative amount. Although there was some difference between MP and MI in CON and PAS lambs, it didn’t make a lot of sense, as the lambs from two different finish systems had very low levels of MP and MI, which means a very low level of “pastoral” flavor.

Looking through related studies, Frank et al [9] reported that their experimental lambs with HCW from 16.6 to 20.2 kg, being close to the HCW of the present research, had the content of MOA, EOA, MNA, MP, and MI were 220 to 393, 89 to 170, 35.1 to 63.6, 36.0 to 89.2, 32.9 to 118.9 μg/g, respectively [9]. The MOA, EOA and MNA from the lambs of Merino sheep were reported 86, 24, 18 μg/g respectively, and the content of these three compounds in the crossbreed of Merino and Ile de France were 49, 26, and 5.6 μg/g, respectively [46]. Schiller et al. had reported the content and distribution of these 3 compounds among 6 breeds, ranging from 56.9 to103, 13.3 to 19.7, and 15.6 to 46.6 μg/g, respectively [47]. The data above are much higher than the counterparts from the lambs in the CON group, indicating that even if the content of compounds responsible for the unpleasant flavor in the CON group were higher than in the PAS group, they were still lower than most breeds reported. In other words, the odor of HBS would be intensified when concentrate inclusion in feedstuff fed in pens, but it could not damage the excellent characteristics of HBS in flavor fundamentally. Whether these differences in the content of BCFAs could be used as an indicator to distinguish the meat of HBS breed, still needs extensive investigation.

As known, the eating quality of sheep meat covers a wide range of complex traits. In the present study, there were 23 traits involved in detection. Each trait in two production systems has a unique embodiment, how to simplify comprehensive evaluation and select reduced traits as a marked character was in great request. Principal component analysis is therefore an optional consideration in which multivariate statistical analysis can reduce the complexity of the issue. Through this method, the lambs in the study were separated into two parts primarily by the finishing systems they received (Figure 1), it was then possible to understand the commonality and difference clearly regarding the eating quality from different production systems, and Figure 1 supplied the information that sheep meat from the CON group is ruddier, lighter, with a higher content of compounds related with unpleasant flavor and USFAs compared to the meat from PAS group.

In the present study, lambs from two feeding regimens of pasture grazing and concentrate-included feeding in sheep pen were conducted for systematic comparison, especially in flavor character with GC-MS instead of trained sensory panels. The results embodied changes in the rearing environment and forage, which affected the growth performance, meat physical measurements and altered the fatty acid profile profoundly. Taking the fattening system of TMR after weaned in pasture exhibited a higher growth rate in meat productivity, improved eating quality symbolized by redder and brighter in color, tender in taste and better fatty acid profile for human health. Despite the concentrate inclusion feeding system increasing the content of major BCFAs significantly in HBS, it was still lower than most sheep breeds and did not fundamentally deteriorate the excellent characteristic of HBS.

In conclusion, it was evident that the pasture-weaned in tensive fattening system (CON group) offers a good alternative regimen for lamb finishing production, which takes the advantage of both pasture grazing and confined intensive feeding systems and has a wide prospection for extension and application in the HBS meat industry.