The egestion of combustible matter (i.e., FE) depends on the suscepti- bility of the feed components to digestion and absorption by the fish, and there are few significant interactions between the feed ingredients of diets that might influence their digestibility. Thus, the DE value of an ingredient
is relatively independent of the composition of the diet in which it is fed. In contrast, the loss of combustible matter through the gills, or in the urine, depends upon a variety of factors, such as the composition of the diet (over- all balance of the amino acids and digestible energy content) and other factors (physiological state of the animal, stress, etc.). As a consequence, the ME content of a given ingredient is not independent of the diet compo- sition and conditions of the fish to which it is fed. As mentioned by Cho and Kaushik (1990), ME has significance only as long as it has been measured with respect to an animal’s response to a complete diet under a given set of biological and environmental conditions.
1.8.1. Dietary Factors
The main factors affecting nonfecal energy losses are those that influ- ence the retention of protein by the body and hence govern the loss of nitrogenous end products through the gills or in the urine. One such factor is the balance between digestible protein (available amino acid) energy and nonprotein energy of the diet. This balance is represented by the ratio of digestible protein (DP) to DE of the diet (DP/DE). Numerous studies have shown that an increase in dietary DE by an increase in dietary nonprotein energy led to a decrease in ammonia nitrogen excretion, UE+ZE, and hence to an increase in ME (Kaushik and Oliva-Teles, 1985). Studies with rainbow trout have shown that the regression slopes between nitrogen in- take and nitrogen excretion as well as the basal nitrogen excretion levels are affected by the DP/DE of the diet. At a dietary DP/DE ratio of 18 mg/kJ, the relation between nitrogen excretion (Ne) and nitrogen intake (NI) was 75.1+0.307×NI, and at the higher DP/DE ratio of 23 mg/kJ, the rela- tion was 84.9+0.343 x NI (Kaushik, 1998). With regard to marine species, there is a relative lack of quantitative data on N excretion rates as affected by dietary DP/DE levels. Available data, however, indicate that as with the salmonids, N excretion is reduced with decreasing DP/DE ratios in species such as seabass (Dicentrachus labrax)and seabream (Sparus aurata)(Kaushik, 1998; Lupatschet al.,2000). It can therefore be concluded that, in general, UE+ZE decreases as DP/DE decreases, at least within a certain range of DP/DE.
This decrease in nonfecal N excretion and UE+ZE is due to the utiliza- tion of nonprotein energy sources for meeting energy requirements, result- ing in a reduction in catabolism of a certain proportion of amino acid for energy purposes. This phenomenon is referred to as “protein-(amino acid) sparing.” Protein-sparing by lipids has been shown to occur in a majority of fish species. Protein-sparing by digestible carbohydrates such as glucose and gelatinized starch is more limited and the object of continuing studies.
The amino acid composition of the diet is another factor that has a deter- minant effect on the efficiency of nitrogen utilization and UE+ZE. Feeding amino acids in excess of the requirement will result in catabolism of the amino acid, with associated excretion of ammonia and loss of energy. The total digestible nitrogen retention efficiency rarely exceeds 50% in rainbow trout (60% in Atlantic salmon) fed diets with very low DP : DE ratios (16 g DP/MJ DE) with a good amino acid balance. It is not clear to what ex- tent this significant catabolism of amino acids, despite an ample supply of nonprotein energy, is related to maintenance requirements, imbalances, or inevitable catabolism of amino acids.
It has been observed that fish exhibit persistent postprandial hypergly- cemia, either after being fed an excessive amount of digestible carbohydrates or after experimental administration of glucose (Bergot, 1979; Furuichi, 1988). Excretion of glucose in the urine (Yokote, 1970; Kakuta and Namba, 1989; Furuichi, 1988; Bureau et al., 1998; Deng et al., 2000) as well as through the gills (Hemre and Kahrs, 1997) has been detected in hyper- glycemic fish. Bureau et al. (1998) showed that rainbow trout that had levels of blood glucose exceeding a certain threshold for renal excretion (ca. 5–10 mM) excreted very significant amounts of glucose in their urine and consequently had significantly increased UE+ZE values. The excre- tion of glucose in the urine means that diets containing high levels of di- gestible carbohydrate may have a ME content lower than that calculated only on the basis of nitrogenous waste energy excretion (Bureau et al., 1998).
1.8.2. Other Factors
Feeding level and water temperature do not appear to have any effect on the ME/DE ratio of diets (Kaushik, 1980a; Azevedo et al., 1998;
Rodehutscord and Pfeffer, 1999). Interspecific differences in nitrogen ex- cretion and consequently ME are little studied. Diaset al.(1999) observed significant differences in efficiency of N retention in seabass and rainbow trout fed similar diets. Marine fish species appear to retain a much lower pro- portion of the digestible protein fed to them than do salmonid fish species and therefore have significantly higher UE+ZE values (Kaushik, 1998).
Differences in N retention efficiency are also evident between salmonid fish species. Atlantic salmon appear to retain a greater proportion of the digestible protein than do rainbow trout when these two species are fed similar diets (Azevedo, 1998). Available data do not appear to indicate any significant influence of genetic origin (strain, family, ploidy) on nitrogen excretion per unit N intake (Kaushiket al., 1984; Oliva-Teles and Kaushik, 1988).
1.9