- Email: [email protected]
Effect of thawing rate on distribution of an injected salt and phosphate brine in beef
MEAT SCIENCE Meat Science 75 (2007) 480–486 www.elsevier.com/locate/meatsci
Effect of thawing rate on distribution of an injected salt and phosphate brine in beef B. Uttaro *, J.L. Aalhus Agriculture and Agri-Food Canada, Lacombe Research Centre, 6000 C & E Trail, Lacombe, Alta., Canada T4L 1W1 Received 18 January 2006; received in revised form 8 August 2006; accepted 8 August 2006
Abstract Striping or streaking is a problem in moisture enhanced meats which are injected to low levels. Research was undertaken to determine brine distribution paths and effect of pre-injection thawing rate on brine distribution. Beef rib eyes and eye of rounds were vacuum packed and aged 7 days at 3 C, then frozen ( 35 C) for a month, thawed either quickly in water (5 h at 12–17 C) or slowly in air (48 h at 3 C), injected to 108–110% using a conventional brine containing blue food colouring, chilled overnight, then cut along four planes and photographed for image analysis of % blue, % marbling, and subjective evaluation of brine distribution paths. There were no significant differences in % blue due to treatment. Brine distribution in both muscles was better parallel to fibers than perpendicular to them. A subset of aged rib eyes was injected to 125%. Although brine was more extensively distributed than at lower injection levels, basic patterns remained unchanged. Crown Copyright 2006 Published by Elsevier Ltd. All rights reserved. Keywords: Beef; Moisture enhanced; Brine distribution; Value added; Striping; Thaw rate
1. Introduction The problem of striping or streaking in injected products has been noted in the literature repeatedly over the decades (Gooding, McKeith, Carr, Killefer, & Brewer, 2004; Knight & Parsons, 1988; Voyle, Jolley, & Offer, 1986). Striping was first noted in cured products and was overcome with post-injection tumbling. Today, striping is a problem in moisture enhanced meats. These are generally injected at low levels (108–115%), and do not undergo a tumbling treatment because the product is marketed raw and must still appear desirable in the retail case. There have been some suggestions made, and equipment constructed to change the manner in which meat is injected, in order to minimize striping in moisture enhanced products (Freixenet, 1993), yet the problem still exists. Possibilities for the effect of non-nitrite-containing brines on meat colour have been postulated (Swatland, 2004), and movement of *
Corresponding author. Tel.: +1 403 782 8569; fax: +1 403 782 6120. E-mail address: [email protected] (B. Uttaro).
brine from the injection site, along muscle fiber long axes, has recently been reported (Gooding et al., 2004), confirming speculation by Swatland (2004) and in Voyle et al. (1986). Since there is evidently something about the internal structure of the meat which restricts brine flow, a preinjection treatment which disturbs this structure but which minimally changes the external appearance would be useful. Primary brine components are generally sodium chloride and one or more of the sodium phosphates, usually including sodium tripolyphosphate and/or pyrophosphate. The proposed action of salt at a myofibrillar level is that chloride ions bind to thick and thin filaments, creating an area coated with similar charges thus forcing adjacent regions away from one another (electrostatic repulsion) causing the myofilament lattice to expand as much as possible within the constraint of actomyosin cross-bridges (Bendall, 1954; Offer & Trinick, 1983). Most phosphates used in brines have a higher pH than meat so their introduction raises the pH of their environment, contributing to water retention in this manner. Phosphates act also by
0309-1740/$ - see front matter Crown Copyright 2006 Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.meatsci.2006.08.015
B. Uttaro, J.L. Aalhus / Meat Science 75 (2007) 480–486
dissociating the actin/myosin complex starting at the edge of the A-band (Offer & Knight, 1988). The most rapid swelling of myofibrils occurs when they are exposed to 0.8 M salt (Offer & Trinick, 1983). Since commercial brines for moisture enhanced products are often made up to a 0.68–0.86 M solution (4–5% salt), one might expect that brine retention occurs very rapidly at the injection site, and tapers off beyond as the immediate area becomes saturated, allowing excess brine to move further into the meat and become more dilute with distance from the injection site. Salt and phosphate act synergistically, and in the presence of pyrophosphate, maximum myofibrillar swelling occurs at a salt concentration of only 0.4 M (Offer & Trinick, 1983). Freezing and thawing rates have been known to affect internal meat structure to varying degrees through cellular disruption by ice crystal formation (Ambrosiadis, Theodorakakos, Georgakis, & Lekas, 1994; Gonzalez-Sanguinetti, An˜on, & Calvelo, 1985; Judge, Aberle, Forrest, Hedrick, & Merkel, 1989; Mandigo & Osburn, 1996). Judge et al. (1989) stated that during the thawing process, meat fairly rapidly warms up to the freezing point, and then remains there for a relatively long time. This plateau exists because of a heat absorption gradient, along the edge of which there is ample opportunity for recrystallization, producing large ice crystals. Ambrosiadis et al. (1994) noted that slow thawing rates (e.g. in air at 4 C for 28 h) always caused more damage than fast thawing rates (e.g. in water at 16– 18 C for 1.5 h). Study of morphological changes led them to conclude that the damage was caused by the formation of large extracellular ice crystals during the thawing process. Ngapo, Babare, Reynolds, and Mawson (1999) concurred. Boles and Swan (2002) reported greater brine uptake by previously frozen meat than by never-frozen samples. Brine injection using multi-needle automatic stitch-type injection equipment is controlled by three factors: the pressure with which the brine is introduced, the duration of the injection needles in the meat (also known as ‘dwell time’), and the speed of the conveyor belt. Brine pressure must be at least high enough to overcome the pressure of the meat around the needles. An increase in dwell time by decreasing the head speed allows more time for brine to be introduced into the meat. The speed of the conveyor determines how close rows of injection sites are to one another. The final injection level is achieved by coming to a suitable combination of these three factors. Beef requires more pressure or higher dwell time than pork, and muscle requirements within a species vary. At the low injection levels used for moisture enhanced meats, injection equipment may end up operating at low pressures and high head speeds (therefore minimal dwell times). Injection equipment is fitted with a stripper plate over the needles. This is a sprung plastic block which serves the dual purpose of guiding needles as they move into the meat, and then holding the meat down, ‘stripping’ it from the needles as they are retracted from the meat on the upward stroke. Some
481
equipment is fitted with a series of stripper blocks instead of a single bar or plate to better accommodate the non-uniform shape of the meat. The aims of this research were twofold. Firstly, to determine the brine distribution path, and secondly, to determine if the beef packers’ practice of freezing excess boxed meat for a time may be beneficial to brine distribution when the thawed meat is injected. 2. Materials and methods 2.1. Sample collection and pre-injection treatment Over 2 weeks in the early Fall of 2004, 45 marketweight Angus cross-steers underwent conventional slaughter and chilling at the Lacombe Research Centre, Lacombe, Alta. On each week, the ribeyes (longissimus thoracis, LT) and eye of rounds (semitendinosus, ST) were dissected from the right sides and trimmed of all external fat 24 h after slaughter. All muscles were vacuum packaged (Multivac AG800) and aged at approximately 3 C for 7 days. At this time, one third of the samples were then injected while the remaining two thirds were moved to 35 C for approximately 32 days of frozen storage. Half the frozen samples were then thawed slowly in air at 3 C for 2 days, and the other half quickly in water over approximately 5 h. Water temperature ranged from 12 to 17 C during this period. All samples were then injected. 2.2. Sample injection The injection equipment was an InjectStar ‘‘New Twist’’ BI-72 fitted, for this experiment, with two rows (50% offset) of nine single 4 mm-diameter needles spaced 2.5 cm apart. Conveyor advance was set at the smallest step, which produced an even diamond pattern of equidistant injection sites, with no overlaps or gaps. Preliminary work to determine optimum equipment settings revealed that below a brine pressure setting of 1.5 bar, very small blue dots of dyed brine at needle entry sites were seen, but there was no movement of the blue colour further into the meat, indicating that meat pressure around holes in the needle end exceeded brine pressure, and dictating that a minimum brine pressure of 1.5 bar be used. Pressure was 1.5–2 bar, with a head speed of approximately 60 strokes/ min for LTs and 44 strokes/min for STs. The cooled brine (3 C) contained 4.8% each of sodium chloride and sodium tripolyphosphate, and 200 ppm FDC Blue #1 (Calico Food Ingredients Ltd., Kingston, Ont.) and was mixed in reverse-osmosis water. Previous exploratory lab work (unpublished) showed that the blue dye was an excellent indicator of location in the meat of the salt component of the brine, while it may slightly overstate the presence of exogenous phosphorus. Brine pressure and head speed (dwell time) settings were fine-tuned before the injection of each muscle type, using one or two samples. If the outcome was immediately suitable, these calibration samples
482
B. Uttaro, J.L. Aalhus / Meat Science 75 (2007) 480–486
were kept to increase sample numbers, if not, they were discarded. LT samples were oriented for injection with the dorsal surface upward, and STs with the more rounded surface upward. After injection, one or two rows of needle insertion sites at one end of the muscle were marked with stainless steel T-pins, to guide the ‘‘parallel to needles’’ cut described below. In the first week, non-frozen LTs were injected to approximately 125% of their initial weight using a slightly longer dwell time and marginally higher pressure. In the second week non-frozen LTs were injected to 108– 110%. Target injection level for all other samples was 108–110%. Samples were weighed before and after injection, then stored on covered trays at 3 C. 2.3. Brine distribution evaluation After an overnight storage period, muscles were cut to expose four planes: (1) parallel to needles (cut to expose needle entry paths); (2) perpendicular to needles; (3) parallel to fasciculi; (4) perpendicular to fasciculi (Fig. 1). This required four cuts on the LT, but only two on the ST since planes 1 and 2, and planes 3 and 4 were equivalent to one another due to the parallel structure of this muscle. Images of each surface were digitally captured (2272 · 1704 pixels; 72 dpi) with a Canon PowerShot A80 (4 megapixel) fitted with a polarizing filter adjusted 90 to the orientation of polarizing filters on the two pairs of GE 100 W Reveal lights illuminating the sample from 45. Image J (v 1.32j; available at http://www.rsb.info.nih.gov/ij; developed by Wayne Rasband, National Institutes of Health, Bethesda, MD) was used to select the area of interest, then convert the JPEG image to an RGB stack. A modified thresholding approach was used on the R layer to determine the relative number of dark pixels in the area (for brine distribution), and on the G layer to determine the relative number of light pixels (for marbling). Images were studied to determine characteristics of brine distribution.
2.4. Statistics Statistics were performed using SAS v8 (SAS Institute, Inc., Cary, NC) to generate LSMeans through MIXED models using the Kenward-Roger degrees of freedom. 3. Results and discussion 3.1. Image analysis Table 1 shows injection levels by muscle and treatment, excluding the over-pumped non-frozen LTs from Week 1. There were no significant differences. Table 2 shows comparative brine coverage between nonfrozen LT samples injected to high (125%) and low (110%) levels as determined by image analysis. The higher injection level was created by slightly increasing both the brine pressure and the dwell time of the needles in the meat thus effectively increasing the force with which a larger amount of brine was introduced. As expected, the % blue coverage was significantly higher on every face for the higher injection level than for the lower one. Illustrated here, and in Figs. 2 and 3 is the difference this volume makes on the extent of the brine distribution. At the lower injection level, approximately 25 cm3 was dispersed in approximately 5600 cm3 of meat, while at the higher injection level the brine volume was closer to 63 cm3. Table 3 shows comparative brine coverage and marbling by face, across non-frozen, quickly thawed, and slowly thawed samples of similar injection levels. There were no significant differences in % blue due to treatment, despite some significant underlying marbling differences. There were no interactions between % blue and % marbling. Nei-
Table 1 LSMeans of injection levels in LTs and STs across three thawing regimes Treatment
n
LT (% pump)
SE
n
ST (% pump)
SE
Non-frozen Fast thaw Slow thaw
6 9 12
109.50 108.93 108.98
0.603 0.492 0.426
12 13 12
109.50 110.80 109.99
0.973 0.935 0.973
Table 2 LSMeans of brine coverage (% blue) on each face of non-frozen LTs pumped to either 125% or 110% Injection level n % Pump
5
High
SE
n
y
0.59
6
50.54y 84.57y 76.26y 76.95y
3.33 3.33 3.33 3.33
6 6 6 6
125.13
Low
SE x
0.53
29.42x 50.67x 42.72x 50.77x
3.04 3.04 3.04 3.04
109.50
a
Fig. 1. Side views of (a) LT and (b) ST. Arrows show orientation of injection needles and approximate depth reached, as indicated by brine presence. Dotted lines show cuts made to expose faces for study of brine distribution patterns: (1) parallel to needles; (2) perpendicular to needles; (3) parallel to fasciculi; (4) perpendicular to fasciculi.
Face 1 2 3 4 x,y
5 5 5 5
Significant differences (P 6 0.05) within a row. Face 1, parallel to needles; face 2, perpendicular to needles; face 3, parallel to fasciculi; face 4, perpendicular to fasciculi. a
B. Uttaro, J.L. Aalhus / Meat Science 75 (2007) 480–486
483
Fig. 2. LT injected to 109.4%. Dark areas indicate brine presence. Face 1, upper right: parallel to needles. Face 2, left: perpendicular to needles. Face 3, center right: parallel to fasciculi. Face 4, bottom right: perpendicular to fasciculi.
Fig. 3. LT injected to 125.8%. Dark areas indicate brine presence. Face 1, upper right: parallel to needles. Face 2, left: perpendicular to needles. Face 3, center right: parallel to fasciculi. Face 4, bottom right: perpendicular to fasciculi.
ther a fast thaw nor a slow thaw treatment affected the extent to which brine was distributed in the meat when the injection level was kept constant. To keep the injection level constant, dwell times were very slightly decreased for fast-thaw samples from the setting for non-frozen samples, and further marginally reduced for slow-thaw samples from fast-thaw settings. Since % blue did not change among treatments, this indicates that some part of the fascicular structure was successfully disturbed by the treatment, allowing more brine retention, but that
longitudinal interfascicular loosening and separation that occurs during slow thawing (Ambrosiadis et al., 1994; Ngapo et al., 1999) did not facilitate brine distribution. 3.2. Subjective observations Subjective observations of brine distribution patterns on exposed faces revealed a number of consistent features for each muscle type across all treatments. Fig. 2 shows a typical LT injected to 108–110%, Fig. 3 a LT injected to 125%,
484
B. Uttaro, J.L. Aalhus / Meat Science 75 (2007) 480–486
Table 3 LSMeans of brine coverage (% blue) and marbling on each face of LTs and STs across three thawing regimes Treatment
LT
ST A
Face
Face n
1
2
3
4
SE
n
1&4
2&3
SE
% Blue Non-frozen Fast thaw Slow thaw
6 9 12
29.42 34.01 35.52
50.67 46.86 46.91
42.72 41.63 48.97
50.77 43.72 46.42
3.4 2.9 2.4
12 13 12
28.20 34.05 32.22
54.24 51.38 50.54
3.5 3.4 3.5
% Marbling Non-frozen Fast thaw Slow thaw
6 9 12
8.62 9.21 7.84
10.04 9.36 8.23
8.91 8.94 8.25
1.0 0.8 0.7
12 13 12
4.89 5.00 4.96
a,b A
9.11b 8.29b 5.97a
6.86b 5.46a 4.68a
0.5 0.5 0.5
Significant differences (P 6 0.05) within a column. Face 1, parallel to needles; face 2, perpendicular to needles; face 3, parallel to fasciculi; face 4, perpendicular to fasciculi.
Fig. 4. ST injected to approximately 109%. Dark areas indicate brine presence: (a) slow thaw; (b) fast thaw. Faces 2 and 3, left: perpendicular to needles and parallel to fasciculi. Faces 1 and 4, right: parallel to needles and perpendicular to fasciculi.
Fig. 4a a ST thawed slowly and injected to 108–110%, and Fig. 4b a ST thawed quickly and injected to 108–110%. On Face 1 (parallel to needles) of LTs injected to 108– 110%, brine was found to be deposited in the top 2/3 to 3/4 of the muscle, and in many cases there were blue stripes at the same spacing as needles (Fig. 2). These were visible near the center of the face, indicating poor lateral brine distribution here, while nearer to the edges they became blurred, indicative of better distribution. Brine was consistently injected throughout the needle path as was evidenced by uniform thickness of the blue bands. The bottom edge of this face was often outlined in blue and was due to absorption of brine from the bottom of the tray while samples sat in the cooler overnight. Faces 3 (parallel to fasciculi) and 4 (perpendicular to fasciculi) showed better
distribution of brine parallel to fiber orientation than perpendicular to it. Face 2 showed blue dots, often near the center of the face, indicating minimal brine deposition at individual injection sites. The edges to either side of this area appeared quite blue, often in somewhat blurred fine stripes that followed fiber direction. Face 4 showed areas of blue which appeared to be limited by some level of intramuscular connective tissue since in some places the blue stayed within the easily-discerned muscle bundles whereas in others the blue ran between them, suggesting the perimysium is involved. In several samples this distribution along the length of muscle bundles was particularly well-illustrated through the contrast between Faces 1 and 3 and Faces 1 and 4, for although needle entry depth was clearly only 2/3 of meat depth on Face 1, Face 3 showed dyed
B. Uttaro, J.L. Aalhus / Meat Science 75 (2007) 480–486
brine across the entire cross-section, and Face 4 showed much of the depth of the LT dyed blue. Face 2 (perpendicular to needles) consistently showed that the best brine distribution was along the edges of the LT while the poorest was at the center of the cranial end. Faces 1 and 4 (parallel to needles and perpendicular to fasciculi) of STs injected to 108–110% were both exposed by a single cut due to the fiber orientation of this muscle (Fig. 4). Needle penetration depth was similar to the LT, and lateral brine distribution (i.e. across fasciculi) was seen to be limited. Injection sites in the form of clear blue lines were obvious, particularly near the center of the sample, and a blue lower border indicated brine uptake from the tray during storage. Faces 2 and 3 (perpendicular to needles and parallel to fasciculi) were also exposed by one cut. Best brine distribution was at the edges of the muscle, and poorest, as evidenced by blue dots or even just holes where needles entered but no brine was deposited, was generally in a band just off-center of the longitudinal axis. Presence of brine along muscle fiber length was evident. There is a transverse band or sheet of connective tissue in the ST which is immediately evident on cutting open this muscle lengthwise. In a number of cases, particularly when overall dye distribution was poor, such as away from the edges of the muscle, the dyed brine was seen to be distributed through the muscle up to this barrier, but not to cross it. As might be expected, the brine distribution in LTs injected to 125% was much better than at the lower injection levels, and may be attributed to a larger volume of brine being introduced under slightly higher pressure (Fig. 3). Of particular interest is that the same basic distribution pattern could be seen here as in LTs injected to 108– 110%: i.e. brine distribution was better parallel to fasciculi than perpendicular to them, and there was an area of low brine uptake at the center of the cranial portion of the cut (Fig. 3 vs. Fig. 2). Regarding brine distribution in relation to marbling and vascular supply, fat generally remained white, indicating no uptake of the water-based brine. In cases where needles pierced vascular tissues, brine entered, but did not seem to continue very far along the vessels, possibly because of insufficient brine pressure. Distinguishing between pre-injection treatments based on brine distribution patterns was not possible (Fig. 4a and b). Strongly characteristic of both the LT and the ST, although to differing degrees, was that the center portions were often all but free of brine. This was likely contributed to by the inevitably uneven pressure exerted by the stripper plate (a single piece, on this equipment) on a meat sample of non-uniform height. A greater intramuscular pressure would build under the high points of the sample where the bar pressed, than under the low points where it did not. Since little or no brine was deposited under the high points, it can be deduced that internal meat pressure in this region would thus appear to be greater than brine pressure.
485
Another contributor to poor brine distribution in this area may be the severely limited opportunity for sideways movement of fibres located in the center of the muscle to accommodate the introduction of brine, in comparison to fibres located at the edges. That thaw treatments successfully disrupted the internal muscle structure is evidenced by the previously discussed adjustments to dwell time that were necessary to keep brine inclusion levels on target. This agrees with work by Boles and Swan (2002) who reported that thawed meat showed greater brine uptake than fresh meat when injector settings were kept constant. However, subjective observations in the present study show that injury caused by thawing regime was not extensive enough to improve brine distribution across fascicles, offering evidence that lateral brine distribution may be restricted in some manner by the connective tissue. 4. Conclusion Extent of brine distribution was similar for fresh, quickly thawed and slowly thawed meat when injected to similar levels. In both the LT and ST, brine distribution was consistently better parallel to muscle fibers than perpendicular to them. In areas where pressure within the muscle could have exceeded brine pressure, such as in the center of the sample, brine was unable to penetrate. Injection of a greater volume of brine resulted in more extensive brine distribution, although distribution patterns were similar to those seen for low injection levels. Acknowledgements Calico Food Ingredients Inc. provided the powdered food dye. Lacombe Meat Lab staff and Jennifer Schatschneider were the technical assistance on this project, and Wayne Robertson enabled image collection through loaning equipment and aiding in set-up. References Ambrosiadis, I., Theodorakakos, N., Georgakis, S., & Lekas, S. (1994). Influence of thawing methods on the quality of frozen meat and the drip loss. Fleischwirtschaft, 74(3), 284–287. Bendall, J. R. (1954). The swelling effect of polyphosphates on lean meat. Journal of the Science of Food and Agriculture, 5, 468–475. Boles, J. A., & Swan, J. E. (2002). Meat and storage effects on processing characteristics of beef roasts. Meat Science, 62, 121–127. Freixenet, L. (1993). Spray injection of meat: influence of the brine pressure in the quality of injected products. Fleischwirtschaft International, 3, 16–20. Gonzalez-Sanguinetti, S., An˜on, M. C., & Calvelo, A. (1985). Effect of thawing rate on the exudate production of frozen beef. Journal of Food Science, 50, 697–700, & 706. Gooding, J. P., McKeith, F. K., Carr, T. D., Killefer, J., & Brewer, M. S. (2004). Characterization of striping in fresh, enhanced pork loins. In Reciprocal meat conference, Lexington, Kentucky, June 2004. Judge, M. D., Aberle, E. D., Forrest, J. C., Hedrick, H. B., & Merkel, R. A. (1989). Principles of meat science. Dubuque, Iowa: Kendall/Hunt Publishing Co.
486
B. Uttaro, J.L. Aalhus / Meat Science 75 (2007) 480–486
Knight, P., & Parsons, N. (1988). Action of NaCl and polyphosphates in meat processing: responses of myofibrils to concentrated salt solutions. Meat Science, 24, 275–300. Mandigo, R. W., & Osburn, W. N. (1996). Cured and processed meats. In L. E. Jeremiah (Ed.), Freezing effects on food quality (pp. 135–182). New York: Marcel Dekker, Inc. Ngapo, T. M., Babare, I. H., Reynolds, J., & Mawson, R. F. (1999). Freezing and thawing rate effects on drip loss from samples of pork. Meat Science, 53, 149–158. Offer, G., & Knight, P. (1988). The structural basis of water-holding in meat. Part 1: General principles and water uptake in meat processing.
In R. Lawrie (Ed.). Developments in meat science (Vol. 4, pp. 63–171). London and New York: Elsevier Applied Science. Offer, G., & Trinick, J. (1983). On the mechanism of water holding in meat: the swelling and shrinking of myofibrils. Meat Science, 8, 245–281. Swatland, H. J. (2004). Fiber-optic spectrophotometry of streaking in pork loins injected with sodium chloride and tripolyphosphate. Canadian Journal of Animal Science, 8, 385–389. Voyle, C. A., Jolley, P. D., & Offer, G. W. (1986). Microscopical observations on the structure of bacon. Food Microstructure, 5, 63–70.
Effect of thawing rate on distribution of an injected salt and phosphate brine in beef B. Uttaro *, J.L. Aalhus Agriculture and Agri-Food Canada, Lacombe Research Centre, 6000 C & E Trail, Lacombe, Alta., Canada T4L 1W1 Received 18 January 2006; received in revised form 8 August 2006; accepted 8 August 2006
Abstract Striping or streaking is a problem in moisture enhanced meats which are injected to low levels. Research was undertaken to determine brine distribution paths and effect of pre-injection thawing rate on brine distribution. Beef rib eyes and eye of rounds were vacuum packed and aged 7 days at 3 C, then frozen ( 35 C) for a month, thawed either quickly in water (5 h at 12–17 C) or slowly in air (48 h at 3 C), injected to 108–110% using a conventional brine containing blue food colouring, chilled overnight, then cut along four planes and photographed for image analysis of % blue, % marbling, and subjective evaluation of brine distribution paths. There were no significant differences in % blue due to treatment. Brine distribution in both muscles was better parallel to fibers than perpendicular to them. A subset of aged rib eyes was injected to 125%. Although brine was more extensively distributed than at lower injection levels, basic patterns remained unchanged. Crown Copyright 2006 Published by Elsevier Ltd. All rights reserved. Keywords: Beef; Moisture enhanced; Brine distribution; Value added; Striping; Thaw rate
1. Introduction The problem of striping or streaking in injected products has been noted in the literature repeatedly over the decades (Gooding, McKeith, Carr, Killefer, & Brewer, 2004; Knight & Parsons, 1988; Voyle, Jolley, & Offer, 1986). Striping was first noted in cured products and was overcome with post-injection tumbling. Today, striping is a problem in moisture enhanced meats. These are generally injected at low levels (108–115%), and do not undergo a tumbling treatment because the product is marketed raw and must still appear desirable in the retail case. There have been some suggestions made, and equipment constructed to change the manner in which meat is injected, in order to minimize striping in moisture enhanced products (Freixenet, 1993), yet the problem still exists. Possibilities for the effect of non-nitrite-containing brines on meat colour have been postulated (Swatland, 2004), and movement of *
Corresponding author. Tel.: +1 403 782 8569; fax: +1 403 782 6120. E-mail address: [email protected] (B. Uttaro).
brine from the injection site, along muscle fiber long axes, has recently been reported (Gooding et al., 2004), confirming speculation by Swatland (2004) and in Voyle et al. (1986). Since there is evidently something about the internal structure of the meat which restricts brine flow, a preinjection treatment which disturbs this structure but which minimally changes the external appearance would be useful. Primary brine components are generally sodium chloride and one or more of the sodium phosphates, usually including sodium tripolyphosphate and/or pyrophosphate. The proposed action of salt at a myofibrillar level is that chloride ions bind to thick and thin filaments, creating an area coated with similar charges thus forcing adjacent regions away from one another (electrostatic repulsion) causing the myofilament lattice to expand as much as possible within the constraint of actomyosin cross-bridges (Bendall, 1954; Offer & Trinick, 1983). Most phosphates used in brines have a higher pH than meat so their introduction raises the pH of their environment, contributing to water retention in this manner. Phosphates act also by
0309-1740/$ - see front matter Crown Copyright 2006 Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.meatsci.2006.08.015
B. Uttaro, J.L. Aalhus / Meat Science 75 (2007) 480–486
dissociating the actin/myosin complex starting at the edge of the A-band (Offer & Knight, 1988). The most rapid swelling of myofibrils occurs when they are exposed to 0.8 M salt (Offer & Trinick, 1983). Since commercial brines for moisture enhanced products are often made up to a 0.68–0.86 M solution (4–5% salt), one might expect that brine retention occurs very rapidly at the injection site, and tapers off beyond as the immediate area becomes saturated, allowing excess brine to move further into the meat and become more dilute with distance from the injection site. Salt and phosphate act synergistically, and in the presence of pyrophosphate, maximum myofibrillar swelling occurs at a salt concentration of only 0.4 M (Offer & Trinick, 1983). Freezing and thawing rates have been known to affect internal meat structure to varying degrees through cellular disruption by ice crystal formation (Ambrosiadis, Theodorakakos, Georgakis, & Lekas, 1994; Gonzalez-Sanguinetti, An˜on, & Calvelo, 1985; Judge, Aberle, Forrest, Hedrick, & Merkel, 1989; Mandigo & Osburn, 1996). Judge et al. (1989) stated that during the thawing process, meat fairly rapidly warms up to the freezing point, and then remains there for a relatively long time. This plateau exists because of a heat absorption gradient, along the edge of which there is ample opportunity for recrystallization, producing large ice crystals. Ambrosiadis et al. (1994) noted that slow thawing rates (e.g. in air at 4 C for 28 h) always caused more damage than fast thawing rates (e.g. in water at 16– 18 C for 1.5 h). Study of morphological changes led them to conclude that the damage was caused by the formation of large extracellular ice crystals during the thawing process. Ngapo, Babare, Reynolds, and Mawson (1999) concurred. Boles and Swan (2002) reported greater brine uptake by previously frozen meat than by never-frozen samples. Brine injection using multi-needle automatic stitch-type injection equipment is controlled by three factors: the pressure with which the brine is introduced, the duration of the injection needles in the meat (also known as ‘dwell time’), and the speed of the conveyor belt. Brine pressure must be at least high enough to overcome the pressure of the meat around the needles. An increase in dwell time by decreasing the head speed allows more time for brine to be introduced into the meat. The speed of the conveyor determines how close rows of injection sites are to one another. The final injection level is achieved by coming to a suitable combination of these three factors. Beef requires more pressure or higher dwell time than pork, and muscle requirements within a species vary. At the low injection levels used for moisture enhanced meats, injection equipment may end up operating at low pressures and high head speeds (therefore minimal dwell times). Injection equipment is fitted with a stripper plate over the needles. This is a sprung plastic block which serves the dual purpose of guiding needles as they move into the meat, and then holding the meat down, ‘stripping’ it from the needles as they are retracted from the meat on the upward stroke. Some
481
equipment is fitted with a series of stripper blocks instead of a single bar or plate to better accommodate the non-uniform shape of the meat. The aims of this research were twofold. Firstly, to determine the brine distribution path, and secondly, to determine if the beef packers’ practice of freezing excess boxed meat for a time may be beneficial to brine distribution when the thawed meat is injected. 2. Materials and methods 2.1. Sample collection and pre-injection treatment Over 2 weeks in the early Fall of 2004, 45 marketweight Angus cross-steers underwent conventional slaughter and chilling at the Lacombe Research Centre, Lacombe, Alta. On each week, the ribeyes (longissimus thoracis, LT) and eye of rounds (semitendinosus, ST) were dissected from the right sides and trimmed of all external fat 24 h after slaughter. All muscles were vacuum packaged (Multivac AG800) and aged at approximately 3 C for 7 days. At this time, one third of the samples were then injected while the remaining two thirds were moved to 35 C for approximately 32 days of frozen storage. Half the frozen samples were then thawed slowly in air at 3 C for 2 days, and the other half quickly in water over approximately 5 h. Water temperature ranged from 12 to 17 C during this period. All samples were then injected. 2.2. Sample injection The injection equipment was an InjectStar ‘‘New Twist’’ BI-72 fitted, for this experiment, with two rows (50% offset) of nine single 4 mm-diameter needles spaced 2.5 cm apart. Conveyor advance was set at the smallest step, which produced an even diamond pattern of equidistant injection sites, with no overlaps or gaps. Preliminary work to determine optimum equipment settings revealed that below a brine pressure setting of 1.5 bar, very small blue dots of dyed brine at needle entry sites were seen, but there was no movement of the blue colour further into the meat, indicating that meat pressure around holes in the needle end exceeded brine pressure, and dictating that a minimum brine pressure of 1.5 bar be used. Pressure was 1.5–2 bar, with a head speed of approximately 60 strokes/ min for LTs and 44 strokes/min for STs. The cooled brine (3 C) contained 4.8% each of sodium chloride and sodium tripolyphosphate, and 200 ppm FDC Blue #1 (Calico Food Ingredients Ltd., Kingston, Ont.) and was mixed in reverse-osmosis water. Previous exploratory lab work (unpublished) showed that the blue dye was an excellent indicator of location in the meat of the salt component of the brine, while it may slightly overstate the presence of exogenous phosphorus. Brine pressure and head speed (dwell time) settings were fine-tuned before the injection of each muscle type, using one or two samples. If the outcome was immediately suitable, these calibration samples
482
B. Uttaro, J.L. Aalhus / Meat Science 75 (2007) 480–486
were kept to increase sample numbers, if not, they were discarded. LT samples were oriented for injection with the dorsal surface upward, and STs with the more rounded surface upward. After injection, one or two rows of needle insertion sites at one end of the muscle were marked with stainless steel T-pins, to guide the ‘‘parallel to needles’’ cut described below. In the first week, non-frozen LTs were injected to approximately 125% of their initial weight using a slightly longer dwell time and marginally higher pressure. In the second week non-frozen LTs were injected to 108– 110%. Target injection level for all other samples was 108–110%. Samples were weighed before and after injection, then stored on covered trays at 3 C. 2.3. Brine distribution evaluation After an overnight storage period, muscles were cut to expose four planes: (1) parallel to needles (cut to expose needle entry paths); (2) perpendicular to needles; (3) parallel to fasciculi; (4) perpendicular to fasciculi (Fig. 1). This required four cuts on the LT, but only two on the ST since planes 1 and 2, and planes 3 and 4 were equivalent to one another due to the parallel structure of this muscle. Images of each surface were digitally captured (2272 · 1704 pixels; 72 dpi) with a Canon PowerShot A80 (4 megapixel) fitted with a polarizing filter adjusted 90 to the orientation of polarizing filters on the two pairs of GE 100 W Reveal lights illuminating the sample from 45. Image J (v 1.32j; available at http://www.rsb.info.nih.gov/ij; developed by Wayne Rasband, National Institutes of Health, Bethesda, MD) was used to select the area of interest, then convert the JPEG image to an RGB stack. A modified thresholding approach was used on the R layer to determine the relative number of dark pixels in the area (for brine distribution), and on the G layer to determine the relative number of light pixels (for marbling). Images were studied to determine characteristics of brine distribution.
2.4. Statistics Statistics were performed using SAS v8 (SAS Institute, Inc., Cary, NC) to generate LSMeans through MIXED models using the Kenward-Roger degrees of freedom. 3. Results and discussion 3.1. Image analysis Table 1 shows injection levels by muscle and treatment, excluding the over-pumped non-frozen LTs from Week 1. There were no significant differences. Table 2 shows comparative brine coverage between nonfrozen LT samples injected to high (125%) and low (110%) levels as determined by image analysis. The higher injection level was created by slightly increasing both the brine pressure and the dwell time of the needles in the meat thus effectively increasing the force with which a larger amount of brine was introduced. As expected, the % blue coverage was significantly higher on every face for the higher injection level than for the lower one. Illustrated here, and in Figs. 2 and 3 is the difference this volume makes on the extent of the brine distribution. At the lower injection level, approximately 25 cm3 was dispersed in approximately 5600 cm3 of meat, while at the higher injection level the brine volume was closer to 63 cm3. Table 3 shows comparative brine coverage and marbling by face, across non-frozen, quickly thawed, and slowly thawed samples of similar injection levels. There were no significant differences in % blue due to treatment, despite some significant underlying marbling differences. There were no interactions between % blue and % marbling. Nei-
Table 1 LSMeans of injection levels in LTs and STs across three thawing regimes Treatment
n
LT (% pump)
SE
n
ST (% pump)
SE
Non-frozen Fast thaw Slow thaw
6 9 12
109.50 108.93 108.98
0.603 0.492 0.426
12 13 12
109.50 110.80 109.99
0.973 0.935 0.973
Table 2 LSMeans of brine coverage (% blue) on each face of non-frozen LTs pumped to either 125% or 110% Injection level n % Pump
5
High
SE
n
y
0.59
6
50.54y 84.57y 76.26y 76.95y
3.33 3.33 3.33 3.33
6 6 6 6
125.13
Low
SE x
0.53
29.42x 50.67x 42.72x 50.77x
3.04 3.04 3.04 3.04
109.50
a
Fig. 1. Side views of (a) LT and (b) ST. Arrows show orientation of injection needles and approximate depth reached, as indicated by brine presence. Dotted lines show cuts made to expose faces for study of brine distribution patterns: (1) parallel to needles; (2) perpendicular to needles; (3) parallel to fasciculi; (4) perpendicular to fasciculi.
Face 1 2 3 4 x,y
5 5 5 5
Significant differences (P 6 0.05) within a row. Face 1, parallel to needles; face 2, perpendicular to needles; face 3, parallel to fasciculi; face 4, perpendicular to fasciculi. a
B. Uttaro, J.L. Aalhus / Meat Science 75 (2007) 480–486
483
Fig. 2. LT injected to 109.4%. Dark areas indicate brine presence. Face 1, upper right: parallel to needles. Face 2, left: perpendicular to needles. Face 3, center right: parallel to fasciculi. Face 4, bottom right: perpendicular to fasciculi.
Fig. 3. LT injected to 125.8%. Dark areas indicate brine presence. Face 1, upper right: parallel to needles. Face 2, left: perpendicular to needles. Face 3, center right: parallel to fasciculi. Face 4, bottom right: perpendicular to fasciculi.
ther a fast thaw nor a slow thaw treatment affected the extent to which brine was distributed in the meat when the injection level was kept constant. To keep the injection level constant, dwell times were very slightly decreased for fast-thaw samples from the setting for non-frozen samples, and further marginally reduced for slow-thaw samples from fast-thaw settings. Since % blue did not change among treatments, this indicates that some part of the fascicular structure was successfully disturbed by the treatment, allowing more brine retention, but that
longitudinal interfascicular loosening and separation that occurs during slow thawing (Ambrosiadis et al., 1994; Ngapo et al., 1999) did not facilitate brine distribution. 3.2. Subjective observations Subjective observations of brine distribution patterns on exposed faces revealed a number of consistent features for each muscle type across all treatments. Fig. 2 shows a typical LT injected to 108–110%, Fig. 3 a LT injected to 125%,
484
B. Uttaro, J.L. Aalhus / Meat Science 75 (2007) 480–486
Table 3 LSMeans of brine coverage (% blue) and marbling on each face of LTs and STs across three thawing regimes Treatment
LT
ST A
Face
Face n
1
2
3
4
SE
n
1&4
2&3
SE
% Blue Non-frozen Fast thaw Slow thaw
6 9 12
29.42 34.01 35.52
50.67 46.86 46.91
42.72 41.63 48.97
50.77 43.72 46.42
3.4 2.9 2.4
12 13 12
28.20 34.05 32.22
54.24 51.38 50.54
3.5 3.4 3.5
% Marbling Non-frozen Fast thaw Slow thaw
6 9 12
8.62 9.21 7.84
10.04 9.36 8.23
8.91 8.94 8.25
1.0 0.8 0.7
12 13 12
4.89 5.00 4.96
a,b A
9.11b 8.29b 5.97a
6.86b 5.46a 4.68a
0.5 0.5 0.5
Significant differences (P 6 0.05) within a column. Face 1, parallel to needles; face 2, perpendicular to needles; face 3, parallel to fasciculi; face 4, perpendicular to fasciculi.
Fig. 4. ST injected to approximately 109%. Dark areas indicate brine presence: (a) slow thaw; (b) fast thaw. Faces 2 and 3, left: perpendicular to needles and parallel to fasciculi. Faces 1 and 4, right: parallel to needles and perpendicular to fasciculi.
Fig. 4a a ST thawed slowly and injected to 108–110%, and Fig. 4b a ST thawed quickly and injected to 108–110%. On Face 1 (parallel to needles) of LTs injected to 108– 110%, brine was found to be deposited in the top 2/3 to 3/4 of the muscle, and in many cases there were blue stripes at the same spacing as needles (Fig. 2). These were visible near the center of the face, indicating poor lateral brine distribution here, while nearer to the edges they became blurred, indicative of better distribution. Brine was consistently injected throughout the needle path as was evidenced by uniform thickness of the blue bands. The bottom edge of this face was often outlined in blue and was due to absorption of brine from the bottom of the tray while samples sat in the cooler overnight. Faces 3 (parallel to fasciculi) and 4 (perpendicular to fasciculi) showed better
distribution of brine parallel to fiber orientation than perpendicular to it. Face 2 showed blue dots, often near the center of the face, indicating minimal brine deposition at individual injection sites. The edges to either side of this area appeared quite blue, often in somewhat blurred fine stripes that followed fiber direction. Face 4 showed areas of blue which appeared to be limited by some level of intramuscular connective tissue since in some places the blue stayed within the easily-discerned muscle bundles whereas in others the blue ran between them, suggesting the perimysium is involved. In several samples this distribution along the length of muscle bundles was particularly well-illustrated through the contrast between Faces 1 and 3 and Faces 1 and 4, for although needle entry depth was clearly only 2/3 of meat depth on Face 1, Face 3 showed dyed
B. Uttaro, J.L. Aalhus / Meat Science 75 (2007) 480–486
brine across the entire cross-section, and Face 4 showed much of the depth of the LT dyed blue. Face 2 (perpendicular to needles) consistently showed that the best brine distribution was along the edges of the LT while the poorest was at the center of the cranial end. Faces 1 and 4 (parallel to needles and perpendicular to fasciculi) of STs injected to 108–110% were both exposed by a single cut due to the fiber orientation of this muscle (Fig. 4). Needle penetration depth was similar to the LT, and lateral brine distribution (i.e. across fasciculi) was seen to be limited. Injection sites in the form of clear blue lines were obvious, particularly near the center of the sample, and a blue lower border indicated brine uptake from the tray during storage. Faces 2 and 3 (perpendicular to needles and parallel to fasciculi) were also exposed by one cut. Best brine distribution was at the edges of the muscle, and poorest, as evidenced by blue dots or even just holes where needles entered but no brine was deposited, was generally in a band just off-center of the longitudinal axis. Presence of brine along muscle fiber length was evident. There is a transverse band or sheet of connective tissue in the ST which is immediately evident on cutting open this muscle lengthwise. In a number of cases, particularly when overall dye distribution was poor, such as away from the edges of the muscle, the dyed brine was seen to be distributed through the muscle up to this barrier, but not to cross it. As might be expected, the brine distribution in LTs injected to 125% was much better than at the lower injection levels, and may be attributed to a larger volume of brine being introduced under slightly higher pressure (Fig. 3). Of particular interest is that the same basic distribution pattern could be seen here as in LTs injected to 108– 110%: i.e. brine distribution was better parallel to fasciculi than perpendicular to them, and there was an area of low brine uptake at the center of the cranial portion of the cut (Fig. 3 vs. Fig. 2). Regarding brine distribution in relation to marbling and vascular supply, fat generally remained white, indicating no uptake of the water-based brine. In cases where needles pierced vascular tissues, brine entered, but did not seem to continue very far along the vessels, possibly because of insufficient brine pressure. Distinguishing between pre-injection treatments based on brine distribution patterns was not possible (Fig. 4a and b). Strongly characteristic of both the LT and the ST, although to differing degrees, was that the center portions were often all but free of brine. This was likely contributed to by the inevitably uneven pressure exerted by the stripper plate (a single piece, on this equipment) on a meat sample of non-uniform height. A greater intramuscular pressure would build under the high points of the sample where the bar pressed, than under the low points where it did not. Since little or no brine was deposited under the high points, it can be deduced that internal meat pressure in this region would thus appear to be greater than brine pressure.
485
Another contributor to poor brine distribution in this area may be the severely limited opportunity for sideways movement of fibres located in the center of the muscle to accommodate the introduction of brine, in comparison to fibres located at the edges. That thaw treatments successfully disrupted the internal muscle structure is evidenced by the previously discussed adjustments to dwell time that were necessary to keep brine inclusion levels on target. This agrees with work by Boles and Swan (2002) who reported that thawed meat showed greater brine uptake than fresh meat when injector settings were kept constant. However, subjective observations in the present study show that injury caused by thawing regime was not extensive enough to improve brine distribution across fascicles, offering evidence that lateral brine distribution may be restricted in some manner by the connective tissue. 4. Conclusion Extent of brine distribution was similar for fresh, quickly thawed and slowly thawed meat when injected to similar levels. In both the LT and ST, brine distribution was consistently better parallel to muscle fibers than perpendicular to them. In areas where pressure within the muscle could have exceeded brine pressure, such as in the center of the sample, brine was unable to penetrate. Injection of a greater volume of brine resulted in more extensive brine distribution, although distribution patterns were similar to those seen for low injection levels. Acknowledgements Calico Food Ingredients Inc. provided the powdered food dye. Lacombe Meat Lab staff and Jennifer Schatschneider were the technical assistance on this project, and Wayne Robertson enabled image collection through loaning equipment and aiding in set-up. References Ambrosiadis, I., Theodorakakos, N., Georgakis, S., & Lekas, S. (1994). Influence of thawing methods on the quality of frozen meat and the drip loss. Fleischwirtschaft, 74(3), 284–287. Bendall, J. R. (1954). The swelling effect of polyphosphates on lean meat. Journal of the Science of Food and Agriculture, 5, 468–475. Boles, J. A., & Swan, J. E. (2002). Meat and storage effects on processing characteristics of beef roasts. Meat Science, 62, 121–127. Freixenet, L. (1993). Spray injection of meat: influence of the brine pressure in the quality of injected products. Fleischwirtschaft International, 3, 16–20. Gonzalez-Sanguinetti, S., An˜on, M. C., & Calvelo, A. (1985). Effect of thawing rate on the exudate production of frozen beef. Journal of Food Science, 50, 697–700, & 706. Gooding, J. P., McKeith, F. K., Carr, T. D., Killefer, J., & Brewer, M. S. (2004). Characterization of striping in fresh, enhanced pork loins. In Reciprocal meat conference, Lexington, Kentucky, June 2004. Judge, M. D., Aberle, E. D., Forrest, J. C., Hedrick, H. B., & Merkel, R. A. (1989). Principles of meat science. Dubuque, Iowa: Kendall/Hunt Publishing Co.
486
B. Uttaro, J.L. Aalhus / Meat Science 75 (2007) 480–486
Knight, P., & Parsons, N. (1988). Action of NaCl and polyphosphates in meat processing: responses of myofibrils to concentrated salt solutions. Meat Science, 24, 275–300. Mandigo, R. W., & Osburn, W. N. (1996). Cured and processed meats. In L. E. Jeremiah (Ed.), Freezing effects on food quality (pp. 135–182). New York: Marcel Dekker, Inc. Ngapo, T. M., Babare, I. H., Reynolds, J., & Mawson, R. F. (1999). Freezing and thawing rate effects on drip loss from samples of pork. Meat Science, 53, 149–158. Offer, G., & Knight, P. (1988). The structural basis of water-holding in meat. Part 1: General principles and water uptake in meat processing.
In R. Lawrie (Ed.). Developments in meat science (Vol. 4, pp. 63–171). London and New York: Elsevier Applied Science. Offer, G., & Trinick, J. (1983). On the mechanism of water holding in meat: the swelling and shrinking of myofibrils. Meat Science, 8, 245–281. Swatland, H. J. (2004). Fiber-optic spectrophotometry of streaking in pork loins injected with sodium chloride and tripolyphosphate. Canadian Journal of Animal Science, 8, 385–389. Voyle, C. A., Jolley, P. D., & Offer, G. W. (1986). Microscopical observations on the structure of bacon. Food Microstructure, 5, 63–70.