Poultry meal in pet foods

Poultry protein meal represents a substantial portion of the high quality protein and fat in modern companion animal diets. Poultry protein meal is commonly included at 5 to 40% and can contribute in excess of 85% of the dietary protein and 30% of the dietary fat. Thus, changes to the quality or composition of the protein or fat in poultry protein meal can have profound effects on the nutritional value of the diet. The process of producing poultry protein meal is a daunting task undertaken to stabilize the raw material from microbial deterioration.

Unfortunately, during this process protein and fat quality are affected.

Defining poultry meal

The availability of fresh poultry and rendered poultry products coincided with the commercialization and industrialization of poultry production in the 1940s and 1950s; and feed values for poultry by-product meal were first established in the 1950s (Fuller, 1996). The volume of rendered poultry proteins in 2003 was estimated at 3,073.5 million lbs; and on average the companion animal industry consumes about 23% (Pearl, 2003).

The collective term ‘poultry protein meal’ covers both poultry by-product meal and poultry meal from chicken, turkey, or other poultry origin. The official definitions according to the Association of American Feed Control Officials (AAFCO, 2004) are:

9.10 Poultry By-Product Meal consists of the ground, rendered, clean parts of the carcass of slaughtered poultry, such as necks, feet, undeveloped eggs, and intestines, exclusive of feathers, except in such amounts as might occur unavoidably in good processing practices. The label shall include guarantees for minimum crude protein, minimum crude fat, maximum crude fiber, minimum phosphorous (P), and minimum and maximum calcium (Ca). The calcium (Ca) level shall not exceed the actual level of phosphorous (P) by more than 2.2 times. If the product bears a name descriptive of its kind, the name must correspond thereto.

9.71 Poultry Meal is the dry rendered product from a combination of clean flesh and skin with or without accompanying bone, derived from the parts of whole carcasses of poultry or a combination thereof, exclusive of feathers, heads, feet, and entrails. It shall be suitable for use in animal food. If it bears a name descriptive of its kind, it must correspond thereto.

In short, poultry by-product meal (PBPM) differs from poultry meal (PM) only by the inclusion of heads, feet, and entrails. However, no quality considerations are described in the definitions. A proposal in 1998 to drop the ‘by-product’ designation from poultry products and replace it with the term‘protein’ was rejected by AAFCO. The newest proposal requests that the ‘by-’ be dropped from byproduct.

Proponents of this request are asking for a level playing field with meat meal, fish meal, and lamb meal; while opponents claim that this would be misleading to the consumer. At issue is the name on a label and how it is represented. However, there is little evidence in the literature to support differing names. In a comparison of chicken meal (CM) and chicken by-product meal (CBPM) using a chick protein efficiency ratio (PER) model, no difference was observed in protein quality (Aldrich and Daristotle, 1998). Bednar et al. (2000) fed dogs diets containing either PBPM or PM and reported that ileal digestibility of protein and amino acids was not different between treatments; however, total tract protein digestibility was 6.74% lower for dogs receiving PBPM than for dogs receiving PM. From the perspective of chemical composition, there is a fair amount of overlap between PM and PBPM (Table 1).

Several different grades of rendered poultry products are available. Feed grade is seldom used in pet food because it contains a higher level of ash and lower protein content. Standard pet food grade contains less than 14% ash; and low ash poultry meal and (or) poultry by-product meal contain less than 11% ash. The latter is available in limited quantities at a premium price and typically reserved for low ash cat formulas (Miller, 1996).


Rendering

‘Rendering’ is required to stabilize the mass of poultry co-products that have been removed from the human edible stream. The rendering plant accepts and processes all raw materials received from the animal processing plant and must ‘render’ them stable so as to avoid public health problems. The process transforms raw unused poultry parts into a form that can be easily stored and transported. Rendering, in its simplest description, is a sterilization, dehydration, and resizing process (Miller, 1996). In the US, standard rendering is a ‘high temperature’ process.

This involves extensive heating (approximately 280°F), which drives water and fat from the bone and tissue. The fat is removed by pressing and the remaining ‘cake’ is ground in a hammer mill to a uniform particle size. The fat goes to storage vessels where it can be further processed or sold. Likewise, the ground meal is conveyed to storage silos for cooling and eventual sale, or further processed in an effort to ‘improve’ its chemical composition.


Protein quality

Fresh meats would be a preferred material with which to construct petfood diets, but this is not always practical for several reasons: 1) expense associated with freezing and chilling, 2) expense involved wit transportation of high amounts of moisture, 3) most extrusion processes will not handle more than 25% fresh meat in a formula, 4) fresh meat reduces production efficiency, and 5) fresh meat diets can be more difficult to stabilize. Therefore, the use of dry meals with concentrated protein is often necessary.

Achieving this dry meal requires rendering; however, the rendering process can have a substantial impact on nutritional quality. Murray et al. (1997) reported that protein and total amino acid digestibility at the ileum in dogs fed a diet containing rendered poultry by-product (meal) was reduced by greater than 10% when compared to a diet containing fresh poultry by-product. However, no differences in total tract protein digestibility were detected. Energy and amino acid digestibility of animal by-product meals can be negatively affected by different rendering processes, high rendering temperatures, extended residence times (Wang, 1997), and high rendering vessel pressures (Shirley and Parsons, 2000).

Poultry protein meals can have a better (Bednar et al., 2000; Yamka, 2003b), equal (Bednar et al., 2000), or poorer (Clapper et al., 2001) protein digestibility than soybean meal. Unlike vegetable proteins, poultry meal is not fraught with some of the anti-nutritional components; however, composition (Locatelli and Hoehler, 2003; Dozier et al., 2003) and performance can be quite variable. As an example, van Kempen et al. (2004) reported that the variability in digestible lysine and methionine in PBPM was 3 times that of soybean meal. Locatelli and Hoehler (2003) reported that protein and amino acid concentrations of 409 PBPM samples from 1999 to 2002 varied widely within years and the mean concentrations changed by several percentage points from one year to the next.

From where does this variability arise? Besides variation in processing conditions, the components making up the raw material mix can also change from day to day, week to week, and season to season. To illustrate this point, one must look at the protein quality of the different raw material components. In a study comparing protein quality of various rendered chicken parts, Aldrich and Daristotle (1998) reported that the PER value of feet and bone/cartilage were 0.87 and 1.22, respectively; whereas heads, viscera, and gizzards/livers/hearts were 2.50, 3.04, and 3.08, respectively. The protein quality of these latter parts was comparable to backs/breastplate and whole birds without feathers (2.88 and 3.43, respectively).

Based on these data, one might conclude that the level of ash (bone residue) would have the greatest impact on nutritional quality of the meal. However, ash level in PBPM (16.3% vs. 7.2%) did not affect ileal digestibility of protein or amino acids or total tract protein digestibility in dogs, amino acid digestibility in cecectomized roosters (Johnson et al., 1998), or protein quality (PER) in chicks (Johnson and Parsons, 1997). Feeding dogs increasing amounts of low ash PM (10.4% to 32.5% of the diet) did not affect protein or amino acid digestibility at the ileum (Yamka et al., 2003a). Thus, increasing consumption of ash from poultry sources does not negatively affect nutrient quality. If not ash, then it is likely that the lower protein quality of bone/cartilage and feet is associated with high levels of connective tissue and a reduction in the ratio of essential to non-essential amino acids. Bone/cartilage can be a component of either PM or PBPM, and feet a component of PBPM; regardless, adding these components to the raw material mix would likely reduce the quality of either meal.

To assure the protein quality of poultry protein meal, the focus of the pet food manufacturer should be on a solid relationship with the supplier/renderer and an understanding of their business processes.

Measurement of amino acids and amino acid digestibility and utilization can be a good source of information for general trends in quality; however, this is too slow, laborious, and costly for day-to-day decisions. Numerous tests have been described to provide rapid estimates. They range from nitrogen solubility tests in various buffers, to in vitro tests with enzymes and acids, to the use of Near Infrared Reflectance (NIR) and Fourier Transform Infrared Reflectance (FTIR) technology (van Kempen et al., 2004). Each provides information supporting purchase and use of consistent poultry protein meal; however, the rapid methods must be validated periodically with animal test data.


Fat quality

Fat in dog and cat diets is used to support the energy needs of the animal, meet its essential fatty acid requirements, aid absorption of fat soluble vitamins, impart flavor, aroma, and texture to the product, and enhance product appearance. Poultry meal contains about 15% fat, the portion that is left after the extraction process. In one of the few reports in the literature for PBPM, the predominant fatty acids were oleic (18:1n-9), palmitic (16:0), and linoleic (18:2n- 6) at 41, 21.7 and 20%, respectively (Table 2; Kirkland and Fuller, 1971).

This agrees fairly well with the poultry fat data of Pesti et al. (2002) and the USDA National Nutrient Database (USDA-ARS, 2003). Whether there is a real difference between the fatty acid profiles of poultry protein meal and poultry fat has not been reported; but one might speculate that the structural lipids remaining in poultry protein meal would be slightly different than those present in bulk poultry fat. Differences between chicken and turkey fat appear to be small. Chicken fat is comprised of 65.6% mono- and polyunsaturated fatty acids with 37.3% as oleic and 19.5% as linoleic compared to turkey fat with 66% as monoand poly-unsaturated fatty acids and 35.9% as oleic and 21.2% as linoleic (USDA-ARS, 2003). It has been suggested that the proportion of linoleic acid has been increasing as more unsaturated vegetable fats are incorporated into poultry diets. This has implications for the nutritive value and the stability of the fat; however, no data were found in the literature to support this claim.

The high level of linoleic acid present in poultry protein meal fat complements the nutrient requirements for dogs and cats (1% and 0.5% of diet DM, respectively). In addition, poultry fat is well accepted by both dogs and cats; and its flavor is preferred over a number of other fat sources.

However, the fat present in poultry protein meal is susceptible to oxidative rancidity. Rancidity is the irreversible oxidation of fat initiated by catalysts such as light, transition metals (iron and copper), heat, and free radicals (molecules possessing an unpaired electron; Figure 1). Once the reaction is initiated, it becomes autocatalytic and proceeds unabated until the reactants are completely exhausted. The point in time at which the rate of the reaction becomes autocatalytic corresponds to the induction point (IP) and from this point forward, the rate of the reaction proceeds rapidly (Figure 2). This rapid deterioration is classically termed propagation and concludes, as described, once the reactants are exhausted.

While heat, pressure, mechanical grinding, and mixing are necessary components of the rendering process, unfortunately they contribute to the course of fat oxidation. During rendering, the inherent animal cellular defense mechanisms are disrupted or destroyed. Lipids are intimately mixed with metabolic enzymes, transition metals, and water. Transition metals, such as iron, are key catalysts to initiate the oxidation process. Interestingly, chicken and turkey lose a greater proportion of heme iron during cooking than do the red meats (Lombardi-Boccia et al., 2002).

These factors all combine to accelerate oxidation of the fat in the poultry protein meal.

Monitoring fat quality depends upon gathering and utilizing data from several different methods. The elapsed time from raw material to finished product can have an impact on the production of free fatty acids (FFA). Free fatty acids are produced as a result of lipase enzyme activity on the triglycerides. The cleavage of fatty acid(s) from the glycerol backbone produces non-esterified or ‘free’ fatty acids. These FFAs are more susceptible to oxidation. The heat process of rendering denatures the lipase enzymes, thus monitoring FFA levels provides an indication of pre-rendering raw material handling. The initial peroxide value (iPV) is commonly measured.

However, its relevance is valid only as an initial postrendering indication of quality. This is because peroxides break down to secondary oxidation products. The secondary oxidation products include aldehydes, ketones, epoxides, etc. It is these secondary oxidation products that result in rancidity and offaromas.

Standard methods for their quantification include p-anisidine values, GC-headspace analysis, and reaction with thiobarbituric acid (TBARS). It is these secondary oxidation products and other reactive oxidation species that can be detrimental. As an example, consumption of oxidized poultry fat has been reported to decrease growth, lower hematocrit counts, decrease half-life of intestinal cells, and affect IgA and lymphocyte proliferation in chicks and pigs (Dibner et al., 1996). Feeding diets with moderate levels of aldehydes was also shown to retard puppy growth and suppress immune function (Turek et al., 2003). In many cases, rancidity in the diet can be overcome with high levels of supplemental vitamin E. However, this is an expensive solution and one that neglects the root problem.

The nutritive value of poultry fat is compromised with elevated levels of oxidation; wherein, the essential fatty acids and energy value have been shown to decline over time. The proportion of linoleic acid in unstabilized PBPM fat declined from 20% to 11.8% over a 12-week storage period and corresponded to an elevation in peroxide values (PV; Figure 4; Kirkland and Fuller, 1971). In this study, linoleic acid levels did not change for stabilized PBPM. Likewise, linoleic acid declined in puppy diets and subsequently in serum as the level of dietary aldehydes increased (Figure 3; Turek et al., 2003).

In poultry grease, an elevated level of oxidation was associated with lower dietary metabolizable energy (Pesti et al., 2002). There have been suggestions that the fatty acid concentrations and oxidative conditions of poultry protein meal vary with season. This may be possible as seasonal temperatures, poultry feeding practices, and holiday poultry consumption changes; however, no reports were found in the literature that supports this claim.


Oxidation of fat can be retarded, but not eliminated

The goal of most poultry protein meal suppliers is to delay the onset of oxidation through good management practices, quick product turnover, and the judicious application of antioxidants early in the rendering process. The idea is to slow or delay the IP for as long as possible. Poultry protein meal suppliers and some pet food manufacturers use the antioxidant ethoxyquin for this purpose. Ethoxyquin has been shown to be effective at reducing oxidation (Kirkland and Fuller, 1971), and is approved for use on animal feeds (21CFR573.380). Other synthetic antioxidants such as BHA, BHT, TBHQ, and propyl gallate are used separately or in various combinations with general success. However, synthetic antioxidants have fallen out of favor with purchasers of pet foods.

The newer challenge to stabilizing poultry protein meal is the request for natural preservatives. Over the past decade there has been a substantial increase in the number of products that carry a ‘natural’ claim.

This eliminates use of the synthetic antioxidants mentioned previously. Poultry protein meal can be effectively stabilized with natural antioxidant systems; however, it is much more costly and requires more attention to the details. Most natural antioxidant systems are built on a backbone of mixed tocopherols carried in a vegetable oil with added chelators, spices, and emulsifiers. Because the natural antioxidants are oil miscible and mix poorly with water, there are other physical properties (e.g. viscosity) that make them more difficult to work with. The goal is to match the amount of natural antioxidant needed with the desired shelf-life of the poultry protein meal.

Uniform application as early in the rendering process as possible and thorough mixing so that the antioxidant is intimately associated with the lipid in the meal is preferable. Constant monitoring of application systems and efficacy is required.

Rapid and (or) predictive analytical tests are available to assist in decision making about the relative freshness of poultry protein meal samples and the efficacy of the antioxidant system. Most of the methods rely upon an external factor to accelerate the rate of oxidation, i.e. heat, oxygen, or prooxidants, etc. The most appropriate tests for poultry protein meal are the Schaal oven, oxygen uptake (Oxygen Bomb), and pro-oxidant stress tests. Each of these methods requires the determination of the relationship to real-time storage. Once this relationship is known, they can provide very rapid and reliable information about the stability of poultry protein meal.


Summary

Very few studies have been reported in the literature relating changes in protein and fat quality of poultry protein meal to performance and health of dogs and cats. Because the rendering process by which the meal is produced requires heat and mechanical mixing, the protein and fat quality can be dramatically affected. The protein quality of poultry protein meal can be affected by heat damage and dilution with non-essential amino acids; and the fat quality can be negatively affected by oxidative rancidity. Each result in a lower than expected nutritional value associated with the poultry protein meal, and can have harmful effects on the dog and cat. It is recommended that constant monitoring to assure the quality of the protein and fat be practiced in coordination with the renderer/ supplier. Further, judicious use of antioxidants and close monitoring of results will assure a safe and nutritionally adequate poultry protein meal for the companion animal industry.


References