1.12 Digestion and Absorption Processes (HdE)
1.12.2. Transformation of Substrates and Retention in Tissues
Growing animals accrete new tissues and some of the energy supplied by the diet is stored as protein, lipid, and glycogen. The theoretical efficiency of transformation or retention of substrates in tissue has been calculated for higher vertebrates (Blaxter, 1989; Flatt, 1992). As the metabolism of fish is very similar to that of higher vertebrates, it may be assumed that these
theoretical costs are also valid for fish. Converting glucose into glycogen costs 5% of the energy of glucose as HiE, whereas converting glucose into lipids entails an increase in HiE equal to about 30% of its GE (Blaxter 1989). Conversion of dietary lipids into body lipids is, in theory, about 96%, therefore 4% of the GE of lipids is dissipated as HiE. Conversion of dietary amino acids into body proteins is, in theory, 86% efficient, entailing a HiE of 14% (Blaxter, 1989). Conversion of amino acids into body lipids is, in theory, only 66% efficient, or 34% of energy is lost as HiE.
Determining the cost of nutrient deposition and interconversion in prac- tice is not an easy task. Many studies have approached the problem in an empirical manner by trying to relate ME to RE (or HiE) and then trying to delineate the various determinants of HiE. The most popular approach is a factorial one that was first proposed by Kielanowski (1965): multiple regression of ME intake as the independent variable, with protein and lipid energy deposition rates as the dependent variables to determine the partial energy costs for protein and lipid deposition (Reeds, 1991). In the classical factorial approach, the energy cost of lipid and protein deposition is simply defined as the ME required to promote a defined increment in body pro- tein or lipid. The partial efficiency of ME utilization for whole-body growth (Kpf), protein deposition(Kp), and lipid deposition(Kf) is the ratio of the net energy retained to the corresponding ME intake components.
ME=HEm+(REp/Kp)+(REf/Kf)
Using this approach, Emmans (1994) concluded that the net energy cost of protein retention is 2.54 kJ/kJ of protein retained (that is, 1.54 kJ of heat expended for each 1 kJ of protein deposited), equivalent to aKpof 39.5%.
According to Emmans (1994), when all related factors are accounted for in the analysis, the energy cost of protein deposition does not appear to differ between species. The calculated energy costs of lipid retention were 1.4 and 1.1 kJ/kJ of lipid deposited (i.e., heat losses of 0.4 and 0.1 kJ for each 1 kJ of lipid deposited) when deposited from nonlipid and lipid, respectively.
These are equivalent to aKf =90% when deposited from lipid and a Kf = 70% when deposited from nonfat substrates.
A few studies have used the factorial approach to estimate the HEm,Kp, and Kf of fish. Results of these studies are summarized in Table 1.5. The estimates ofKpvaried between 0.44 and 0.56, and those ofKfbetween 0.72 and 0.91. These values appear to be similar to what has been observed with mammals and birds.
The factorial approach has been criticized because it is much easier to control ME intake than it is to control protein and lipid depositions (Emmans, 1995). Moreover, when different levels of ME intake are used in an experimental design, within a genotype and (or) body weight there may
be a strong correlation between protein and lipid deposition. Therefore, it may be difficult to estimate parameters accurately using a factorial anal- ysis. A multivariate model has recently been proposed by Van Milgen and Noblet (1999) to circumvent these problems. This model appears to give more accurate and reliable estimates of Kp, Kf, and HEm but has never been applied to fish.
It is worth emphasizing that both the factorial and the multivariate ap- proaches are purely empirical and, to some extent, an oversimplification of reality. Protein and lipid deposition is not merely the deposition of energy but a highly complicated process whose rates and efficiencies are governed by dietary factors (nutrient balance and utilization) and biological factors (genetics, types of tissue made, etc.). Factorial and multivariate approaches are, nevertheless, interesting tools for insight into the cause of possible differences between theoretical and observed protein and lipid deposition costs.
The cost of lipid deposition determined by the factorial approach ap- pears to be close to the theoretical chemical cost. Results from Emmans (1994) showed that the source of lipid deposition (lipid or nonlipid) af- fected the efficiency value, and this might contribute to differences in Kf
estimates between studies. SinceKf differs depending on the origin of the lipid deposited, the composition of the diet might affect the efficiency of lipid deposition. Dietary intake of pre-formed lipids will lead to a very good efficiency of utilization for lipid deposition, whereas de novo synthesis of lipid from dietary carbohydrates will lead to a slightly lower efficiency. The pathways of lipogenesis in fish are qualitatively similar to those in other ver- tebrates. In salmonids, variation in the dietary lipid level appears to be a more effective modulator of fatty acid synthesis than the level of dietary car- bohydrate (Sargentet al.,1989). Some recent data have confirmed that an increase in dietary lipid levels decreases hepatic lipogenic enzyme activities (Diaset al.,1998) as well asde novofatty acid synthesis (Brauge, 1994). In fish fed high-fat diets, where almost all the lipid deposited is of dietary origin, the cost of lipid deposition is very low.
There are large differences in the energy cost of protein deposition based on theoretical assumptions (86% efficient) and that calculated using the fac- torial approach (40–60% efficient). The reasons for this difference have not been clearly identified. In growing animals, even at nitrogen equilibrium (zero body protein gain), protein is synthesized and energy is consumed for maintaining protein turnover (Milligan and Summers, 1986). According to Reedset al.(1985), the amounts of protein and energy needed at zero nitrogen retention might account for as much as 20% of the maintenance ATP utilization. The ATP cost of Na+/K+ pumping accounts for a similar proportion of ATP utilization at energy equilibrium (Gillet al.,1989). The
rate of protein synthesis greatly exceeds that of protein deposition (Reeds et al.,1981), the efficiency of protein deposition in fish ranging between 40 and 60%. Therefore, changes in protein turnover are a possible explanation for the variable energy cost of protein deposition. On the other hand, when problems associated with structural and kinetic heterogeneity of amino acid pools are involved (Wattet al.,1991) in addition to the difficulty of measur- ing the synthesis of rapidly turning-over proteins (Wheatley et al.,1988), one is led to conclude that current estimates of protein synthesisin vivomay be underestimates, and consequently the energy cost of protein deposition may also be underestimated. No single cause is responsible for the appar- ent higher cost of protein deposition that is predicted in theory. Energy is expended in the biochemical pathways that lead to protein synthesis, and protein degradation and in regulating and integrating the various cellular metabolic activities involved in protein deposition. All these components need to be quantified so reliable Kp values can be obtained from in vivo studies.
It is also evident that the amino acid balance of the diet will contribute to Kp. Catabolism of amino acids due to amino acid excess or imbalances in the diet or a low non protein energy content will result in waste of energy (higher HiE associated with catabolism of amino acids than with that of lipids) and a decrease inKp. Diets with an imbalanced amino acid composition will result in less energy retained as protein and, consequently, result inKpvalues lower than those that would occur if the diet was perfectly balanced.
Cho et al.(1976) showed that an increase in dietary fat levels led to a decrease in HiE. LeGrow and Beamish (1986) confirmed that the increase in oxygen uptake with increasing dietary protein levels was consistent, irre- spective of the dietary fat levels, thus highlighting the importance of dietary DP/DE ratios. In addition, the efficiency of the energy (i.e., net energy) derived from the catabolism of amino acids by the fish is unclear.
It has been suggested that digestible carbohydrates are possibly also signif- icant contributors to HiE (Beamishet al.,1986; Hiltonet al.,1987). Beamish et al.(1986) observed that fish fed a diet with a high glucose content con- sumed more oxygen than fish fed a diet with the same protein level but rich in lipids, suggesting that an increase in digestible carbohydrate intake results in an increase in heat production. The fish fed the diet high in glu- cose had a lower N retention efficiency (N gain/N intake) than the fish fed the diet high in lipids (Beamishet al.,1986). These data suggest that the effect on HiE observed was in fact related more to the variation in dietary lipids and the sparing of protein than to the digestible carbohydrate itself.
However, Helland and Grisdale-Helland (1998) observed that at low intake levels, an increase in digestible starch at the expense of digestible protein resulted in an increase in the oxygen consumption of Atlantic salmon but
no change in the N retention efficiency. Bureau et al. (1998) observed a very poor retention of ME of digestible carbohydrate fed to rainbow trout.
These data suggest that utilization of carbohydrate can be associated with a significant HiE, especially when the DP/DE ratio of the diet is less than 20–21 g /MJ.
1.13