Nutrition of the dog : an Australian overview.

Abstract

117 Nutrition of the dog: an Australian overview N.D. Costa Division of Veterinary and Biomedical Sciences, Murdoch University, Murdoch WA Australia 6150 Summary Forty-five percent of Australian households care for one or more pet dogs. Their nutrition presents the challenge of maintaining liveweight and health over a number of years. Commercial dog foods are formulated on the nutrient requirements specified by the Association of American Feed Control Officials (AAFCO) rather than NRC recommendations. Most commercial formulations for feeding dogs provide complete and balanced nutrition with safety margins of 10% or more for most nutrients rather than their minimum requirements. Allometric equations for determining daily energy requirements for dogs produce quite different outcomes in the amount of food required. Furthermore the balance of protein and fat as % ME has not been optimised. Wheat and meat and bone meal are the most commonly used ingredients in Australia for commercial dry foods for dogs. However, wheat and beef-based meat and bone meal are also frequently associated with adverse reactions and allergies to foods. Homemade diets are less likely to be associated with allergies and adverse reactions. Formulating dog foods with defined ratios of omega-6 and omega-3 polyunsaturated fatty acids of between 5: 1 and 10: 1 has been successful in increasing the relative concentration of some weakly or anti-inflammatory eicosanoids. Modulation of eicosanoid metabolism is proposed as a basis for treatment of some skin, gastrointestinal and renal disorders. Relatively high inclusions of dietary fibre (five percent or more) have improved stool consistency possibly through modified gastrointestinal structure and function. Prebiotics such as fjuctooligosaccharides are now being evaluated to enhance the actions of dietary fibre. This paper predicts that the trend of formulating to optimise health as well as supplying adequate nutrients will be extended in the future development of commercial dog foods. Dogs in Australia: setting the scene The dog is man's oldest companion animal, probably beginning domestication about 12,000 to 14,000 years ago in Eurasia. Since that time many breeds have been developed for specific purposes such as hunting, herding and guarding. However, today, most dogs are kept mainly for companion purposes and almost 2.7 million households in Australia have a pet dog, giving rise to a total population of approximately 3.8 million dogs in this country (BIS Schrapnel 1996). Obviously dog numbers are related to the human populations in the various States ofAustralia, but expenditure per dog is particularly high in South Australia and the Territories. (Table 1). Recent Advances in Animal Nutrition in Australia 7997 University of New England, Annidale NSW 235 I, Australia 118 Costa, N.D Table 2 Dog Populations and Expenditure on Dogs in Developed Countries The number of dogs in Australia per head of population is comparable with the USA, European countries and Japan, and the expenditure on pet food and accessories for dogs by Australians is within the range of expenditures in those other developed countries. (Table 2 Euromonitor survey reported in Petfood Industry 1997). Direct expenditure on the nutrition of dogs in Australia is more than A$6 19 million per annum (BIS Schrapnel 1996). Prepared dog foods accounted for more than 60% of all food fed to dogs in 1992 (Feed Grains Study 1995) and this trend of feeding commercial dog foods rather than home-made diets is expected to grow. Uncle Ben's ofAustralia (UBA) is the largest pet food manufacturer in the Australian market and their main brands of Pedigree PAL (the world's biggest pet food brand), Chum dog food and My Dog held 75.7% of the value of the canned market and 53.9% of the packet market in 1994 (Feed Grains Study 1995). UBA purchased more than 100,000 tonnes of feed grain in 1994, mainly as cereal grams but with small amounts of pulses and soybean meal. Overall demand for feed grain by the pet food area grew at 20% per annum during 1994 and 1995, and demand is expected to remain high (Feed Grains Study 1995). Feeding the dog compared with other domesticated species It is important to remember that dogs have been used as the preferred animal model for nutritional studies for over a hundred years and this has provided some fundamental data on digestion and vitamin metabolism (as reviewed by Carpenter 199 1). However, this paper will present an overview of the nutrition of dogs that are kept as friends and companions. Feeding the pet dog differs from production feeding in a number of important ways. With beef cattle, prime lambs, pigs, and broiler chickens, there is a clear production end-point preferably achieved in the shortest time span possible. For dairy cattle and sheep inAustralia, the aim of feeding is to produce milk or wool respectively. The nutrient requirement tables for these species reflect these production aims. Only the horse is similar to the dog, in that there is a large variation in liveweight between breeds reflected in the construction of NRC nutrient requirement tables for horses. Moreover, the purpose for keeping a horse is frequently for sport, often for companionship, and now only spasmodically for work. However, the horse has the advantage being a herbivore and can be sustained at maturity on 100% roughage diets. The dog, on the other hand is predominantly a carnivore, with a relatively high protein requirement (minimum 18%) at maturity. There is a number ofnutritional aims and difficulties in feeding the pet dog. Firstly, the aim of feeding the pet dog is to sustain a relatively constant liveweight over a period of eight to fourteen years! Secondly most pet dogs are fed individually by one or two people, so the second aim of feeding is to reinforce a bond of companionship. As an offshoot of this bonding, the owner knows and often understands a great deal about the feeding habits and preferences of their pet dog. Finally the pet dog is kept until old age or disease makes it no longer desirable for both parties, so that feeding the older or geriatric dog is a problem that companion animal nutritionist must face. Different breeds of dogs age at different rates, and in general, larger breeds such as the Great Dane take longer to mature fully (24 months vs 12 months for Beagles), and yet they have a shorter life span than small to medium size breeds. Thus the nutrition of the companion animal draws as much information from other long-lived and well-researched species such as the human, as fi=om the nutrition of production animals. Specification of nutrient requirements of dogs Most commercial dog foods aim to comply with the concept of being `complete and balanced'. A complete food may be defined as a nutritionally adequate feed for animals other than man, compounded by a specific formula to be fed as the sole ration and capable of Nutrition of the dog: an Australian ovetview 119 maintaining life and/or promoting production without any additional substances being consumed except for water. The more balanced means having all known nutrients in the proper amount and concentration as specified by recognised authorities in the field for given set of physiological states for that species. Formulating a `complete and balanced' dog food requires at least three sets of information: 1 a set of nutrient requirement tables; 2 the chemical composition of the ingredients to be used; and 3 an appreciation of the bioavailability of the nutrients present in the dietary ingredients (Morris and Rogers 1994). Two major American organisations have published tables of nutrient requirements for dogs: the NRC Nutrient Requirements of Dogs 1985, and the Association of American Feed Control Officials (AAFCO) who publish an annual volume including nutrient requirement tables for dogs. The NRC (1985) tables specify the minimum nutrient requirements of growing and adult dogs. In the introduction it is stated that `caution is advised in the use of these requirement tables without the demonstration of nutrient availability, because in some cases requirements have been established on the basis of studies in which nutrients were supplied by highly purified ingredients where digestibility and availability were not compromised by the interaction of dietary constituents and effects of processing'. Since practical diets of both a commercial and `homemade' nature are not f?ee of these interactions and effects, and more importantly, no clear guidelines of bioavailability were published by the NRC, commercial manufacturers were either unable, or chose not, to use NRC 1985 as the basis for their claims of nutritional adequacy. In 199 1, the NRC requested that their recommendations not be used to substantiate claims of nutritional adequacy of manufactured feeds for dogs. As a consequence, the NRC nutrient requirement tables are now not widely used in the Australian petfood industry, with the major exception of the National Registration Authority (NRA) who use the NRC as the authority in determining registration of supplements. More importantly, the issue of bioavailability raised by the 1985 NRC Committee has still not really been addressed in current thinking about nutrient requirement tables. Over the last four to five years, AAFCO have become the pre-eminent authority for specifying nutrient requirements of dogs, not only in the USA but in Australia, Japan and Europe also. Historically, AAFCO's primary function in the USA was to provide model laws, animal feed regulations and ingredient definitions, which the various States could use in their own feed laws and regulations for marketing of petfood across State borders in the USA. It is only since 1990 that AAFCO established a Canine Nutrition Expert Subcommittee to advise on nutrient requirement tables and testing procedures, its first report being released as the AAFCO Dog Food Nutrient Profiles in 1992 and updated each year. These tables do not report the published basis for their recommendations, as does the NRC, but they do include maximum as well as minimum amounts for nutrients that can be associated with toxicities (Table 3). The other major points of difference are the higher AAFCO specifications for iron, copper, zinc, iodine and vitamin E for adult dogs at maintenance. An essential part of the AAFCO (1993) protocol was the introduction of a feeding trial using the petfood as the major means of substantiating the claim of being `complete and balanced'. The test protocols for AAFCO specification are set out each year in their annual publication. The minimum testing protocol for adult maintenance for a dog food requires that it must be the only diet fed to a minimurn of 8 healthy adult dogs (at least one year of age) for a minimum of26 weeks. Daily food consumption may be measured and recorded. Individual body weights must be measured at the beginning, then weekly until the end of the 26th week of the test. Ideally there should be no change in liveweight over the 26 weeks, but under the protocol liveweight cannot vary more than 1.9 times the standard error of the average percentage liveweight change. No animal under test can lose more 15% of liveweight. RBC number, haemoglobin packed cell volume, and serum alkaline phosphatase and albumin must be measured at the beginning and at the 26th week. Twenty-six weeks is a long time for any dog to be fed on only one food, so the feeds that pass these protocols are very palatable. The general health of all dogs undergoing AAFCO tests must be monitored by veterinarians through physical examination and blood tests. In addition to this protocol for maintenance, there are AAFCO protocols specified for pregnancy, lactation and growth. Successful completion of such a protocol can form the basis of a claim for nutritional adequacy of a commercial dog food. In the USA, there are a number of independent companies who can conduct such AAFCO trials but there is none in Australia. To undertake anAAFC0 trial inAustralia is a large commitment of resources and only large companies such as UBA have made an effort to screen their products. The problem of predicting daily energy requirement The energy content of a food for dogs can be calculated using modified Atwater constants in the following equation (AAFCO 1997): Carbohydrate is assumed to be equivalent to the nitrogen free extract and no agreed allowance has been made for any contribution to energy from fibre. 120 Costa, N.D Table 3 AAFCO minimum and maximum nutrient profiles for dog foods on a dry matter basis (1997). Moreover, this equation is based on the assumption that the food is approximately 14.6kJ ME/g DM. If the energy density exceeds 16.7kJ ME /g DM, then the nutrient concentrations in the food must be increased in proportion to 14.6kJ ME /g DM factor. The definitive assessment of energy content is through digestibility trials, but more frequently most energy densities stated on labels for dog foods are predicted via this equation. As more commercial dry dog foods exceed energy densities of 14.6kJ then the balance of nutrients relative to predicted ME become more critical. The statement of `complete and balanced' needs to be supported by AAFCO feeding protocols when used on high energy density feeds. Energy requirements and the balance of nutrients to energy for dogs remains an issue of some contention. One of the major reasons for this difficulty is the unique range of liveweight in dogs, with Chihuahuas weighing approximately one kg, some giant breeds such as Newfoundlands and Great Danes nearly 100 kg, and St I Nutrition of the dog: an Australian oven/iew 121 Bernard dogs 115 kg (Burger 1994). Moreover, different breeds have quite different body composition and hair and coat production leading to differing daily energy needs at the same liveweight (Figure 1). Deriving a single predictive allometric equation for the energy needs of dogs has proved somewhat difficult with most of the debate centring on the power function that should be assigned to the metabolic body weight for dogs (NRC 1985; Heusner 1991; Burger and Johnson 1991; Kienzle and Rainbird 1991; Burger 1994; Case et al. 1995). The power function suggested for resting adult dogs of weight range 5.8 - 48.8 kg (Burger 1994) was: Both of these functions are in general agreement with the work of Heusner (199 1) who suggested a function of 0.67 but they are not in accord with the predictive equation of the NRC 1985 which proposes ME requirement = 100 W OS** kcal/day (4 18 W'.** kJ/day). The Heusner equations have different factors for activity in the general equation for ME requirement; the energy allowance for activity @J/day) = K x w-6' where K= 552 (inactive), 607 (active), and 837 for very active (Case, et al. 1995). All of these equations assume that dogs adjust their food intake to match energy requirements but an increasing incidence of obesity has been reported (NRC 1985; Markwell et al. 1990). The difference inthe amount of food fed daily is quite dramatic when dogs are fed on high rather than low energy diets, 16.7 vs. 14.6 kJ using the various predictive equations (Figure 2). It is preferable that dogs be meal-fed at least twice a day rather than fed ad libitum. Owners who follow advice based on either of the prediction equations, 523 kJ/Wo.75 or 418 kJ/W'.**, could run the risk of overfeeding their dogs, especially if their dogs are larger breeds of 30 kg or more. On the other hand, manufacturers who use either of those two equations as the basis for their feeding guides on food labels run the risk of looking uncompetitive compared with companies that quite reasonably frame their recommendations on amounts to feed on the basis of 607 kJ/W's6'. This problem of differences in the amount fed is exacerbated when comparing commercial `supermarket' diets of 14.6 kJ ME /g DM and 12% fat content with high energy diets. The fact that dogs live into old age exacerbates the problems of maintaining an appropriate dietary energy supply. Several studies have found that energy requirements decrease by the order of 20% with increasing age in dogs (Fit&e 199 1; Kienzle and Rainbird 199 1; Finke 1994). Moreover, the optimum contribution of protein to ME is a matter of speculation with Kronfeld et al. (1994) proposing that a protein concentration of 30-40% ME is best suited to the active dog. Deciding the best mix of nutrients and ingredients to optimise energy, protein, fat and fibre in the ageing dog remains a matter for investigation. Types of commercial dog food Commercial pet foods can be categorised according to their moisture content. Dry dog foods in Australia usually contain 6-10% moisture and are most oRen extruded products based on cereals and animal proteins derived as by-products of meat processing in abattoirs. Extrusion is essential to gelatinise (and sometimes partially dextranise) the starches from cereal grains, and to kill bacteria in the ingredients. In addition, dry dog foods usually have 8-15% fat to aid palatability, but increasingly high energy `premium' dry dogs are being formulated with fat content of 1822%. The fat content in dry dog foods must be protected from rancidity by the use of antioxidants such as ethoxyqquin, butylated 722 Costa, N.D hydroxytoluene (BHT), butylated hydroxyanisole (BHA), or vitamin E present as the various chemical forms of tocopherols. Semi-moist dog foods usually contain 14-25% moisture in Australia, and again are often extruded from similar ingredients to dry dog foods but with additional protection against mould and spoilage by using sucrose, glucose, propylene glycol, sorbic and propionic acids, and phosphoric acid. Dogs prefer sweet tastes to acid flavours (preferred by cats), and the sweetness of semi-moist dog foods can aid their palatability. Wet dog foods usually in cans, but increasingly now in `knob' forms (the preferred form in the New Zealand market), contain 78-82% moisture. Wet dog foods are made from offal meats, particularly lung which remains intact during processing, muscle meat, and textured vegetable proteins made from soybean with very few cereal ingredients. Guar and carrageenan gums are often used to form a stable gel of these high moisture products. High heat and/or pressure during manufacture kills bacteria and preserves these wet foods. Ingredients commonly used in commercial dog foods Wheat is the most commonly used cereal grain in dry dog foods in Australia with over 200,000 tonnes used annually (Feed Grains Study 1995). On the other hand, corn is the grain of choice in dog foods manufactured in the USA (Case et al. 1995). This is really a reflection of the availability and price of the two grains in their respective countries. The two' grains have some important differences in carbohydrate composition and fibre constituents (Table 4) as well as in the structure of the starch granules (Bach Knudsen 1993). Extrusion gelatinises the starch in both grains, which a necessary prerequisite for availability of the carbohydrate in starch to the dog's digestive system (Case et al. 1995). However, the higher concentrations of soluble and insoluble non-cellulose polysaccharide (NCP) in wheat are associated with lower digestibilities than corn in dogs. In addition, wheat is more frequently associated with food allergies in dogs (Brown et al. 1995). Nevertheless, wheat is unlikely to be supplanted as the most commonly used grain in Australia because of its cost-effectiveness in commercial dog foods. Rice is more completely digested and absorbed in the small intestine of pigs (McDonald 1996), and rice flour and rice grain is now being used more widely in commercial dry dog foods, particularly for puppies. Rice and often potatoes are also used in dry formulations marketed as `hypoallergenic' diets since these two carbohydrate sources are less frequently associated with adverse and allergenic reactions in skin and the gastrointestinal tract (Guilford 1994; Brown et al. 1995). Meat and bone meal is the most commonly used protein supplement in dry dog foods in Australia with over 24,000 tonnes being used annually (AAS 1993). The proximate analysis of the three grades of meat and bone meal inAustralia is presented in Table 5. This type of meal is usually a byproduct of a combination of animal sources rendered for tallow production but remains one of the major source of essential amino acids for dogs. However, meat and bone meals derived predominantly from beef are associated with a high proportion of allergenic and adverse reactions to commercial foods (Guilford 1994; Brown et al. 1995). The protein quality of meat and bone meal is inversely related to the percentage of ash. The ash fraction represents bone (particularly bone chips), and possibly other material such as eartags and intraruminal `bullets' (e.g. Co, Se), all of which make this fraction so variable. In fact, two reports on the use of meat and bone meal (AAS Report 1993; Allan 1994) have both stressed that the ash fraction of meat and bone meal is the single most variable component and the first limitation on the usefulness of meat and bone meal as a replacement for higher quality protein meals. However, the ash fraction does make a significant contribution to the mineral nutrition of dogs. Calcium concentrations can range from 1.2%to 16.4%withameanof 12.3% inmeatmeal with a 50% crude protein specification. The variability in the concentrations of the trace minerals iron, zinc, copper and manganese can also be large. For instance, the range of iron concentrations over 489 samples of Australian meat and bone meals was 2 1.9 to 66080 ppm Table 4 Carbohydrate composition (% DM) of wheat and corn grains (from Bach Knudsen 1993) Nutrition of the dog: an Australian overview 723 Table 5 Average analysis of Australian meat and bone meals of various crude protein contents (AFIC. DM with a mean of 3993 ppm DM for meat and bone meal with a 50% crude protein specification (AFIC 1987). From discussions with pet food manufacturers, the variation in the iron is probably due to metal filings being abraded from machinery during rendering and thus being included in the final product sold to petfood manufacturers. Elemental iron, such as that present in the filings, would not be available for absorption by animals but could constitute a risk as an oxidant of fats in the meal. Magnets can extract this type of impurity and should be part of any high quality rendering process. Interestingly, some heavy metal contaminants such as cadmium and mercury have been found in meat and bone meals. The reason is for this contamination is not clear but one possibility is the inclusion of contaminated offal such as liver and kidney (see Langlands et al. 1988) in the final mix for rendering. The `generic' lines of dry dog food have expanded the use of meat and bone meal in the short term because this section of the petfood market is growing rapidly. There are a number of companies in each State competing in this area of the market. Price is the key factor and both in perception and price, meat and bone meal is the preferred protein meal. Atypical generic line of dry dog food contains a minimum of 20% crude protein and 8% crude fat and the main source of protein is usually meat and bone meal which can account for 4050% of the total crude protein and 25% of the crude fat. Fifty percent or more of the remaining fat is derived from tallow. It is these two components which determine the palatability of generic lines of dog feed. Generic dog foods are always on the borderline of producing a zinc-responsive problem of coat condition because of their high calcium (often greater than 2.0%) and high wheat (often 35% or more) content. Calcium is a competitive inhibitor of zinc absorption, and the phytates in wheat are effective at binding zinc, thereby also decreasing its absorption. Beef meats and offal, particularly lungs are used in wet dog food formulations, and included in some extruded products. Lungs are preferred because they not only hold their shape after pressure cooking, but also they readily absorb colouring compounds. These characteristics ensure that the final product has a `meaty and chunky' look. Beef proteins are implicated in skin and gastrointestinal tract allergies and adverse reactions to a greater extent than any other source of animal protein (Guilford 1994; Brown et al. 1995). This could be because beef is more frequently used compared with meat and meals from other animal sources. On the hand, beef could be more susceptible to Maillard reactions during cooking which has been proposed as one possible reason for the higher incidence of adverse reactions to commercial foods (Guilford 1994). Chicken meat and poultry meal are increasingly used in both dry and wet formulations to increase palatability and protein quality. Poultry meal is higher in protein and less variable in protein quality than meat and bone meal (65-68% versus 52-58% respectively) because it does not contain the coarse ash components common in meat and bone meal. As a consequence, poultry meal can be ground through very fine-mesh screens during processing which to the human eye produces a better looking extruded food. Meals produced solely from lamb and mutton are now being used in commercial dry food formulations and promoted as hypoallergenic because they are less frequently associated with adverse and allergenic reactions (Guilford 1994, Brown et al. 1995). Lamb is often included in homemade diets used in the nutritional management of food allergies in dogs (Brown et al. Soybean meal and textured vegetable protein from soybean are often used in dry and wet formulations respectively. Soybean protein is of high quality but like beef and wheat is frequently associated with food allergies (Guilford 1994; Brown et al. 1995). This could be an index of its use rather than any intrinsic allergenic property of soy protein. Textured soy protein keeps its shape under pressure cooking (like lungs) and is a preferred protein in wet formulations. Tallow is often added to dry food formulations, both within the extruded premix and sprayed onto the product post+xtrusion, to increase energy density and improve palatability of dry dog foods. The colour of poor quality tallows varies with season as chlorophylls are extracted in this fraction during rendering and this `greenness' affects the colour of extruded, dry dog feeds. Vegetable oils such as sunflower and safflower oils are often included in dry and wet formulations to improve the content of linoleic acid (Cl 8:2 06), an essential fatty acid for dogs (Table 3). The proportion of linolenic acid (Cl 8:3 CI) 3) and the total omega-3 fatty acids as well as omega-6 series based on linoleic acid is 1995). 724 Costa, N.D now seen as important to the properties of dry and wet dog foods. The composition of some commonly used animal fats and vegetable oils is listed in Table 6. Vitamins and trace minerals are often included in commercial dog foods on the assumption that the ingredients themselves contain little or none of these. Allowances are made for those vitamins such thiarnin and folic acid where significant losses occur during processing at high temperature and pressure. Thiamin mononitrate is the preferred form of thiamin for manufacturing because it is less susceptible to loss on cooking. No allowances are made in these inclusions for breed differences in requirements or for sensitivities to particular trace minerals. For instance, copper storage diseases can occur predominantly as an inherited autosomal recessive disorder in Bedlington Terriers and West Highland White Terriers, and in Doberman Pinchers and Cocker Spaniels although the mode of inheritance is not known in these two breeds (Thomburg et al. 1984). Copper accumulation in the liver begins shortly after birth with toxic concentrations of 2000 ppm dry weight of liver (normal range for dogs is 200-400 ppm dry weight) reached at varying ages but usually between 4 and 8 years of age, depending on intake and other interactions. A zinc-responsive disorder often presenting as skin lesions occurs inAlaskan Malamutes, Siberian Huskies, and occasionally Great Danes and Doberman Pinchers (Kunkle 1980, Fadok 1982). Dogs with this affliction require zinc supplementation, as 80 mg Zn per day, throughout their lives (Kunkle 1980). Antioxidants are included in dry and some semimoist dog foods to protect against rancidity of fats. Ethoxyquin, a synthetic antioxidant, is the most effective antioxidant (Coelho 1995). However, some breeders and owners suspect that ethoxyquin is associated with reproductive and other problems in dogs despite the fact that ethoxyquin has been tested as safe (Dzanis 199 1). Over the last few years, dry dog foods have been marketed as containing `natural' vitamin E protection of fats rather than ethoxyquin (Coelho 1995). The comparative effectiveness of antioxidants in protecting against fat rancidity in mixed fat is shown in Table 7. Note the additions of other protective agents such as ascorbic acid to rHeduce the antioxidant, EDTA and citric acid to chelate potential oxidant divalent actions such as iron, and phosphoric acid to maintain low pH. Although synthetic antioxidants such as ethoxyquin are clearly more potent in their action, some purchasers of dry dog food are beginning to show a preference for `natural foods' with high concentrations of fat of higher unsaturation. This presents a challenge to pet food industry to establishing the concentrations of tocopherols needed in proportion to content of fat and its degree of unsaturation to ensure a long (9-12 months) shelf-life under a range of temperatures of storage using the tocopherols. Home-made diets for dogs Most homemade diets are based on meat (Khonke and Hoskins 1994). Carbohydrate is often added in the form of cooked pasta or rice. Vitamins such as vitamin E and A and trace minerals such as calcium and zinc can be deficient in these diets. Nevertheless, complete and balanced diets can be formulated from home ingredients if appropriate combinations of meat, meat and bones, vegetables and pasta, vegetable oils, and vitamin and mineral supplements are used. Some of those most commonly used ingredients are listed in Table 8 along with their energy, protein, fat, Ca and P contents. All-meat diets containing no bone material or providing no access to bones represents the greatest dietary risk from home-made diets for growing or adult dogs due to the severe deficiency of calcium, and phosphorus, as well as a skewed Ca:P ratio of approximately 1:20. Such diets result in nutritional secondary hyperparathyroidism (Hazewinkell989). As a result there are a wide range of supplements available on the market to supply calcium in particular, but also phosphorus to correct these deficiencies and imbalances. The AAFCO minimum level on a dry matter basis for calcium is 0.8% for growth and reproduction Nutrition of the dog: an Australian overview 725 Table 7 Activity of several synthetic and naturally derived antioxidants in mixed fat (Coelho 1995). 726 Costa, N.D and 0.5% for adult maintenance, and a maximum level of 2.5%. Most commercial diets from reputable companies contain 1.1 to 1.8% calcium per kg dry matter. Calcium absorption from the small intestine is predominantly a saturable, carrier-mediated process dependent on 1,25 dihydoxycholecalciferol (Fraser 1995), but in neonates and young animals, there is second non-saturable, diffusion transfer dependent on the concentration of cal