*2.10. Free Thiol Content*

To determine the loss of thiol (sulfhydryl) groups, the 5,5 -Dithiobis (2-nitrobenzoic acid) (DTNB) method was used as described by Lund et al. [13]. Duplicates (2 g) of each sample were homogenised at 16,000 rpm in 40 mL of 5% (*w*/*v*) sodium dodecyl sulfate (SDS) in 0.1 M Tris buffer (pH 8) using a Polytron PT 10–35 GT homogeniser (Thermo Fisher Scientific, Scoresby, VIC, Australia). The homogenates were incubated at 95 ◦C for one hour in covered test tubes. The samples were cooled and centrifuges for 20 min at 1200× *g* using a Rotina 380R Hettich centrifuge (LabGear, VIC, Australia). The supernatants were filtered using Whatman filter paper no 1 and the protein concentration of filtrates was determined at 280 nm using a standard curve with bovine serum albumin (BSA) (Sigma-Aldrich, Castle Hill, NSW, Australia). The samples were diluted to a protein concentration of 1.5 mg.mL−<sup>1</sup> using the SDS homogenisation buffer. The diluted samples were used to determine thiol group concentration by adding 2 mL of 0.1 M Tris buffer (pH 8) and 0.5 mL DTNB to 0.5 mL of sample. Samples were incubated for 30 min in the dark and absorbance at 412 nm was measured using a Multiskan spectrophotometer (Thermo Fisher Scientific, VIC, Australia). The concentration of thiol groups was analysed against a standard curve of L-cysteine prepared in 5% (*w*/*v*) SDS in 0.1 M Tris buffer (pH 8). Total thiol content was calculated and expressed as nmol·mg−<sup>1</sup> protein.

### *2.11. Statistical Analysis*

Data were analysed using restricted maximum likelihood (REML) with GenStat 16th Edition (VSN International, Hemel Hempstead, UK). For pH before packaging, breed (Composite and Merino), feed type (SPD, SCF and SCM), and muscles (*Semimembranosus* and *Vastus lateralis*) were fitted as fixed effects. Pen and carcass number (all nested within, i.e., pen/carcass number) were fitted as random effects. For all other quality traits, breed (Composite and Merino), feed type (SPD, SCF and SCM), packaging method (HioxMAP, TrigasMAP and VSP) were fitted as fixed effects. Pen and carcass number (all nested within) were fitted as random effects. Separate analyses were conducted for each muscle type (*Semimembranosus* and *Vastus lateralis*). *p* < 0.05 was used as the level for significant differences.

#### **3. Results**

#### *3.1. pH and Colour*

Figure 1 shows that the pH of meat prior to packaging significantly differed between the Composite and Merino sheep (*p* = 0.004) and between the *Semimembranosus* and *Vastus lateralis* muscles (*p* < 0.001). While the pH of *Vastus lateralis* was higher compared to *Semimembranosus* for both breeds, the difference was more substantial in meat from Merino compared to Composite sheep. There was no significant effect of feed (*p* = 0.7) on the pH.

**Figure 1.** The pH of *Semimembranosus* and *Vastus lateralis* from Composite and Merino sheep before retail packaging. Values are predicted means ± standard error of differences (SED). *p* (breed × muscle) is 0.096.

Using the CIE L\*, a\*, and b\*-values, the colour stability of lamb was evaluated (Table 1). Breed had differential effects on the lightness (L\*) of *Semimembranosus* and *Vastus lateralis* muscles. Compared to Composite, Merino had a lower L\* value for *Semimembranosus*, yet a higher L\* value for *Vastus lateralis*. A significant effect of breed on a\*, b\*, and hue were

also observed for *Vastus lateralis*, but not *Semimembranosus*. The finishing feed had no effect on any of the colour parameters. The packaging method had a greater influence on all colour parameters (*p* < 0.001 for all) compared to breed and feed effects. While there were small differences between HioxMAP and TrigasMAP, most colour parameters significantly differed between VSP and HioxMAP or between VSP and TrigasMAP for both muscles. Interestingly, when comparing VSP and HioxMAP, hue differed in *Semimembranosus*, but not *Vastus lateralis*. Together, these results show that the choice of packaging methods had a greater influence on the colour stability of sheep meat, compared to breed and feed, and the extent to which of HioxMAP negatively impacts meat colour was muscle dependent. A visual illustration of *Vastus lateralis* in the three packaging methods is presented in Figure 2.

**Figure 2.** Visual comparison of *Vastus lateralis* from a Composite sheep finished on standard pelleted diet with grain and hay and packed in (**A**) vacuum skin packaging; (**B**) HioxMAP; or (**C**) TrigasMAP for 10 days at 4 ◦C.

#### *3.2. Water Holding Capacity*

Water holding capacity was measured as purge and cooking losses (Table 2). While the three supply chain factors (breed, feed, and packaging method) appear to influence purge loss to a similar extent, only packaging method had a significant effect (*p* < 0.001) on purge loss. It is worth noting that purge loss of *Semimembranosus* in TrigasMAP (5.7% ± 0.3 SED) was similar to VSP (5.7% ± 0.3 SED) and lower than HioxMAP (6.8% ± 0.3 SED). There was no difference in purge loss of *Vastus lateralis* in HioxMAP and TrigasMAP, indicating differences between the two muscles in their response to water holding capacity. A significant interaction between breed and packaging method was also observed for purge loss (Figure 3). While the purge loss did not differ across the three packaging methods for composite sheep *Semimembranosus* and *Vastus lateralis*, TrigasMAP reduced the purge loss in Merino *Semimembranosus* compared to HioxMAP (Figure 3A). This reduction in purge loss by TrigasMAP was not observed for *Vastus lateralis* (Figure 3B). Packaging method had a significant effect on purge loss of *Semimembranosus* and *Vastus lateralis* from Merino, but not those from Composite sheep.

Merino *Semimembranosus* had a lower cooking loss compared to the same muscle type from Composite sheep (Table 2). Finishing feed did not affect cooking loss for either *Semimembranosus* or *Vastus lateralis*. The difference in cooking loss between the three packaging methods were significant with the lowest cooking loss in TrigasMAP (30.9% ± 0.5 SED), followed by HioxMAP (32.2% ± 0.5 SED) and VSP (35.1% ± 0.5 SED). No significant interactions were observed for cooking loss.


 ± standard error of differences (Coeff ± SED) and level of significance (*p*-values) are presented. For *Semimembranosus* from a Composite lamb, finished on a standard pelleted diet containing grain and cereal hay, and retail displayed in vacuum skin packaging for 10 days. 2 SCF = standard pelleted diet containing 15% camelina forage hay. 3 SCM = standard pelleted diet containing 8% camelina meal (SCM). 4 HioxMAP = high-oxygen modified atmosphere packaging with 80% O2 and 20% CO2; 5 TrigasMAP = trigas modified atmosphere packaging with 50% O2, 30% N2 and 20% CO2; 6 For *Vastus lateralis* from a Composite sheep, fed with standard pelleted diet containing grain and cereal hay, and packaged in vacuum skin packaging.

#### *Foods* **2022** , *11*, 144



measurements

 of sheep

*Semimembranosus*

 (topside)

**Table 2.** Effect of breed, feed and packaging method on water holding capacity and texture

containing 15% camelina forage hay. 3 SCM = standard pelleted diet containing 8% camelina meal (SCM). 4 HioxMAP = high-oxygen modified atmosphere

packaging with 80% O2 and 20% CO2; 5

Composite sheep, fed with standard pelleted diet containing grain and cereal hay, and packaged in vacuum skin packaging.

TrigasMAP = trigas modified atmosphere packaging with 50% O2, 30% N2 and 20% CO2; 6 For *Vastus lateralis* from a

**Figure 3.** Purge loss of (**A**) *Semimembranosus* and (**B**) *Vastus lateralis* from two sheep breeds (Composite or Merino) in three retail packaging methods (VSP = vacuum skin packaging; HioxMAP = high-oxygen modified atmosphere packaging with 80% O2 and 20% CO2; or TrigasMAP = trigas modified atmosphere packaging with 50% O2, 30% N2 and 20% CO2). Values are predicted means ± standard error of differences (SED). *p* (breed × packaging method) values are 0.014 for *Semimembranosus* (**A**) and 0.015 for *Vastus lateralis* (**B**).

#### *3.3. WBSF and Texture Profile Analysis*

Breed or finishing feed had no effect on WBSF for either of the two muscles (Table 2). Differences in WBSF between the three packaging methods were only found for *Semimembranosus*, which were tougher in HioxMAP and TrigasMAP, compared to VSP. No significant interactions were observed for WBSF in either muscle types.

The effect of the three supply chain factors on Texture Profile Analysis hardness differed between the two muscles (Table 2). Within the *Semimembranosus* samples, hardness was affected by breed only, with *Semimembranosus* from Merino having a higher hardness value compared to *Semimembranosus* from Composite sheep. The hardness of *Vastus lateralis* was only affected by finishing feed, with SCM having a lower hardness value than that of SPD and SCF. Cohesiveness and chewiness were affected by breed and packaging method in both muscle types. Cohesiveness and chewiness were lower in VSP compared to HioxMAP and TrigasMAP for both muscle type, suggesting a softer texture in a low oxygen packaging environment. No significant interactions were observed for hardness, cohesiveness, and chewiness in either muscle types.

### *3.4. Lipid Oxidation*

Lipid oxidation in meat was assessed using TBARS assay and the levels were expressed as mg MDA/kg of meat. Breed did not affect lipid oxidation in either of the two muscle types (Table 3). However, supplementation of feed with either camelina forage or camelina meal led to a reduction in TBARS values compared to the standard pelleted diet containing cereal hay and grains. Packaging type not only had a significant effect but also to a greater extent (substantially higher coefficients) than feed on TBARS values. TrigasMAP was able to reduce lipid oxidation compared to HioxMAP for *Semimembranosus*, but not *Vastus lateralis.* There was also a significant interaction between finishing feed and packaging method for the *Semimembranosus* samples. Figure 4 shows that the TBARS value of meat in HioxMAP were substantially greater in the control diet (SPD) compared to the two camelina supplemented diets (SCF and SCM), especially for *Semimembranosus*. These results further emphasise the need for sheep meat to be packaged in a lower oxygen environment when sheep feed is not supplemented with antioxidants.


**Table 3.** Effect of breed, feed and packaging method on lipid and protein oxidation measurements of sheep *Semimembranosus* (topside) and *Vastus lateralis* (knuckle).

Coefficients ± standard error of differences (Coeff ± SED) and level of significance (*p*-values) are presented. <sup>1</sup> For *Semimembranosus* from a Composite lamb, finished on a standard pelleted diet containing grain and cereal hay, and retail displayed in vacuum skin packaging for 10 days. <sup>2</sup> SCF = standard pelleted diet containing 15% camelina forage hay. <sup>3</sup> SCM = standard pelleted diet containing 8% camelina meal (SCM). <sup>4</sup> HioxMAP = high-oxygen modified atmosphere packaging with 80% O2 and 20% CO2; <sup>5</sup> TrigasMAP = trigas modified atmosphere packaging with 50% O2, 30% N2 and 20% CO2; <sup>6</sup> For *Vastus lateralis* from a Composite sheep, fed with standard pelleted diet containing grain and cereal hay, and packaged in vacuum skin packaging.

**Figure 4.** Thiobarbituric acid reactive substances (TBARS) values of (**A**) *Semimembranosus* and (**B**) *Vastus lateralis* from sheep finished on three diets (SPD = standard pelleted diet containing grain and cereal hay; SCF = pelleted mixture diet containing 15% camelina forage hay; or SCM = pelleted mixture diet containing 8% camelina meal) and retail displayed in three packaging methods (VSP = vacuum skin packaging; HioxMAP = high-oxygen modified atmosphere packaging with 80% O2 and 20% CO2; or TrigasMAP = trigas modified atmosphere packaging with 50% O2, 30% N2 and 20% CO2). Values are predicted means ± standard error of differences (SED). *p* (feed × packaging method) values are 0.011 for *Semimembranosus* (**A**) and 0.243 for *Vastus lateralis* (**B**).
