Could Phosphorus and Calcium Be The Missing Links in Unlocking Radically Low Protein Diets for Commercial Broiler Production?

Stephen Jay Gould famously introduced the term ‘non-overlapping magisteria’ in a Natural History article in March 1997, to describe the separation between scientific and religious lines of enquiry (Gould, 1997). From a broiler nutrition perspective, it would be accurate to represent digestible amino acid and metabolisable energy research, and digestible P and Ca research, using the same vocabulary. Despite these distinct research domains, considerable value has been created in both spheres. For example, before the advent of synthetic amino acids, broiler diets were formulated to contain up to 700 g/kg soybean meal and 350 g/kg CP in order to meet the birds’ requirement for methionine (Pesti, 2009). This contrasts with contemporary broiler diets that include several ‘unbound’ or ‘free’ amino acids and can satisfy the animals amino acid requirement with diet CP concentrations as low as 160-210 g/kg, depending on the age of the chick. Similarly, the commercial introduction of exogenous phytase in 1991 and more recent work on phytase dose optimisation and digestible Ca formulation systems has led to substantial decreases in the use of inorganic phosphate to the point where many broiler grower and finisher diets are entirely denuded of inorganic P sources (Moss et al., 2018; Walk et al., 2021). Despite these ‘localised’ successes, there are very few published papers that explore the potential interaction between dietary CP and diet Ca and P concentrations and these magisteria remain largely non-overlapping. It is the purpose of this short review article to describe some of the relevant literature in this space, offer mechanistic insights that may be of importance and suggest opportunities for further research that may increase the precision of nutrient delivery to the bird without compromising live production metrics.

As far as the author is aware, the first studies that were designed to explore the potential for an interaction between diet CP and P (and by association, Ca) were conducted using rodent models. Hammoud et al. (2017) fed rats a low CP diet (100g/kg CP relative to a standard diet of 200g/kg CP) and titrated P from 0.15g/kg to 3.0g/kg. The authors observed significant increases in weight gain, feed intake, energy efficiency and plasma glucose concentration. In addition, plasma urea nitrogen (N) was reduced from around 6.5 mM/l to 4 mM/l as diet P was increased. Rats that received the low CP diet with the highest concentration of P returned growth performance that was equivalent to those fed the standard CP diet. These observations were confirmed more recently in a rodent model when the addition of digestible lysine and/or P to a low protein diet (100g/kg) resulted in synergistic effects on growth rate and energy efficiency and P alone had a positive effect on protein metabolism and significantly reduced plasma urea N (Ragi et al., 2019).

The role of P in determining digestible amino acid requirements and animal response to dietary CP is not fully established. However, there are three potentially valuable lines of enquiry. First, protein intake has a calciuretic effect (Margen et al., 1974; Zemel, 1988) whereby N and Ca compete for resorption at the renal level. A high protein intake results in excess urinary Ca losses and vice versa. Thus, as broiler diet CP is reduced it is possible that Ca retention is increased, a disequilibrium that may be partially mitigated via supplementation of diets with additional P. This possibility was recently supported by Dao et al. (2022a) where a reduced crude protein diet fed to Ross 308 broiler chickens generated an increase in serum Ca but had no effect on serum P. Importantly, this was associated with a numerical decrease in d28 tibia ash in the chicks that received the low protein diet. Second, low CP broiler diets are compositionally distinct from diets formulated at standard CP concentrations. For example, low CP broiler diets typically have a higher concentration of cereals and lower concentrations of protein meals such as soybean meal or canola. It may be relevant that the concentration of phytate-bound P in cereals is lower than in protein meals (Eeckhout and De Paepe, 1994) and the per se availability of phytate-bound P in protein meals may be higher than for cereals (Weremko et al., 1997). The influence of diet CP per se on P digestibility is likely to be dependent on a number of additional factors such as dietary cation concentration, strategies used in feed formulation to achieve the reduction in crude protein and phytase dosing. Indeed, Dao et al. (2022b) observed an increase in ileal P digestibility in broilers fed a low protein diet, which contradicts earlier observations, so further work to explore the role of low protein diets on mineral digestibility is warranted. Finally, protein synthesis requires appreciable quantities of P for manufacture of ATP (Shariatmadari and Forbes, 1993) and it is possible that additional dietary P may reduce protein catabolism and promote protein accretion via provision of P for ATP synthesis. In short, a low CP diet may simultaneously provide lower concentrations of available phytate-P, promote Ca retention and increase the demand for ATP to drive protein synthesis. Increasing digestible P supply in low CP diets may be an effective strategy to mitigate these influences. This could be done by elevating phytase dosing, addition of inorganic P or reductions in total dietary Ca. Optimal strategies require further research.

In attempt to extend the principals described above to commercial broiler production a study was conducted to specifically explore the interaction between diet CP and P (Cowieson et al., 2020). Ross 308 male broiler chickens were offered diets with low, medium or standard CP concentrations and either low, standard or high available P (for context the grower diets were formulated at 215, 195 or 175g/kg CP and offered at either 4.8, 4.3 or 3.8 g/kg available P). All amino acids were balanced using non-protein bound amino acids and energy, macro- and micro-minerals and dietary electrolyte balanced were equivalent across all dietary treatments. Response of broiler chicks to supplemental available P in terms of body weight corrected feed conversion ratio (FCRc) was more pronounced in the low CP diet than in the standard CP diet, resulting in a significant interaction. Specifically, increasing available P from 3.8 to 4.8g/kg in the diet with standard CP concentration had no effect on FCRc over the full experimental duration (d8-35) whereas the same increase in available P in the diet with the lowest CP concentration resulted in a decrease in FCRc of 7 points. Interestingly, increasing available P concentration resulted in a decrease in plasma uric acid concentration and this was particularly marked at the moderate level of dietary CP resulting in an interaction between diet CP and available P. This confirmed previous research in rodents that dietary P may influence post-absorptive N metabolism and may influence deamination and ammonia detoxification processes. The potential for available P to influence N retention and ammonia management is an area of interest for future study.

The role of P (or more accurately digestible or available P) in amino acid metabolism, nitrogen cycling, ammonia detoxification and nitrogen emissions is not clear. The possibly exists, however, that reducing dietary CP may influence the requirement of the bird for both P and Ca. This may occur via direct mechanisms such as the importance of P in ATP synthesis and protein accretion and the competition between Ca and nitrogen in renal tubules and also indirectly via subtle changes in the soluble phytate concentration of low CP diets. As the global poultry industry moves toward increasingly low CP diets, and in concert with important ongoing work to optimize the deployment of non-protein bound amino acids, some attention to the dietary supply of Ca and P is warranted. Given the complexity of these interactions and the constraints in multi-factor experimental designs a mechanistic modelling approach may be justified.