Digestion of a diet leads to the absorption of amino acids, fatty acids, and sugars, which are the principal metabolic fuels for the body. Catabolism of fats and carbohydrates results in the formation of carbon dioxide and water.
The catabolism of amino acids yields ammonia in addition to carbon dioxide and water. Excretion of nitrogenous waste compounds, of which ammonia amounts to about 85% in most fish species (Kaushik and Cowey, 1991), results in nonfecal energy losses since these compounds contain energy.
Although ammoniotelic, fish excrete small amounts of urea. Urea cycle enzymes have been detected in several species of fish. Purine catabolism is, however, the main source of urea production in fish. Urinary excretion of
Table 1.2
Performance of Rainbow Trout (Initial Weight=7 g/Fish) Fed Practical Dietsa Water temperature
7.5◦C 15◦C
Parameter Diet 1 Diet 2 Diet 1 Diet 2
Lipid sources
Fish oil, herring 16 8 16 8
Beef tallow, fancy, bleachable — 8 — 8
Composition
Digestible protein (DP) (%) 44.0 43.5 44.9 44.4
Digestible energy (DE)(MJ/kg) 19.5 19.9 20.9 20.8
DP/DE (g/MJ) 22.6 21.9 21.5 21.3
Apparent digestibility coefficients (%)
Crude protein 93 93 95 95
Lipid 93 94 98 95b
Energy 83 85 89 89
Performance
Weight gain (g/fish) 13.7 13.1 38.1 39.2
Feed efficiency (gain : feed; as is) 1.32 1.27 1.22 1.15 Retained energy (% digestible intake) 47 47 50 48
aDiets contained fish oil or a fish oil and tallow combination; reared at 7.5 or 15◦C for 12 weeks. From D. P. Bureau, A. M. Harris, and C. Y. Cho. (unpublished data, 1996).
bSignificantly different from diet 1.
other types of combustible materials, such as trimethylamine (TMA) and trimethylamine oxide (TMAO), in certain marine teleosts is also known to occur but has not been quantified under intensive culture conditions (Kaushik and Cowey, 1991). All these nonfecal energy losses, mainly through the gills (branchial energy loss; ZE) and some through the kidneys (urinary energy loss; UE), are unaccounted for by the DE value of a diet, meaning that the DE value of a diet overestimates its actual energy value to the fish. The physiologically available fuel value of the diet to the fish is the metabolizable energy (ME) value, defined as follows:
ME=IE−(FE+UE+ZE)
In the rainbow trout, endogenous (branchial and urinary) nitrogen ex- cretion (UNe+ZNe) rates measured in fish after 3 to 4 days of fasting have been found to vary between 80 and 130 mg N/kg body weight/day (endoge- nous UE+ZE =2.0–3.2 kJ/kg/day), affected most by water temperature
and body weight (Watanabe and Ohta, 1995; Kaushik, 1998). Some re- cent studies with Atlantic salmon suggest that the values might be much lower (Forsberg, 1997). With regard to marine fish, data of Ballestrazziet al.
(1994) and of Dosdatet al. (1996) also show that the UNe rates in European seabass, gilthead seabream, or turbot would be in the range of 100 to 160 mg N/kg/day (endogenous UE+ZE=2.5–4.0 kJ/kg/day), comparable to the values found for rainbow trout (Kaushik, 1998).
Although nonfecal nitrogen losses contribute significantly to the envi- ronmental load in terms of ammonia nitrogen, from the point of view of energy balance, their contribution is small, generally no more than 3% of the ME (Kaushik, 1998).
1.7.1. Measurement
Direct determination of the ME values of fish diets is technically difficult because of the need to measure both branchial and urinary losses released into the aquatic environment in which the fish live. Smith (1971) attempted to overcome these difficulties and developed a procedure which allowed the estimation of the ME values of a number of feedstuffs using rainbow trout 165–530 g in body weight. Before the assays, the fish were anesthetized to allow the insertion of a cannula for urine collection. The fish were then con- fined in a tank with a diaphragm separating the front from the rear portion of the body; they were force-fed the feed as a single daily meal under anes- thetic. The ME values determined by this procedure as a fraction of the DE values ranged from 0.72 to 0.93 (mean=0.87). The procedures employed to separate and collect nitrogen excreted via the gills and kidneys (includ- ing force-feeding) involved considerable handling and were stressful to the fish, which increased the loss of nitrogen (Hunn, 1982) and combustible matter. The increase in nitrogen output, together with the low food intake attained by force-feeding of a single daily meal, might be expected to result in a negative nitrogen balance and a low ratio of ME-to-DE values for many of the feed ingredients studied. This strongly suggests that energy losses via the gill and kidney were greater than would be the case for unrestrained fish feeding normally (Cho and Kaushik, 1990).
Monitoring waste in water in the rearing environment is a commonly used approach. Brett and Zala (1975) determined the diurnal pattern of nitrogen excretion of young sockeye salmon (Oncorhynchus nerka) by allow- ing ammonia to build up during alternate periodical closed–open circuit cycles. Kaushik (1980a,b) was the first to estimate the postprandial excretion rates in a flow-through system in a continuous manner using an autoanalyzer.
This method allows continuous monitoring of ammonia and urea nitrogen excretion under normal physiological conditions even in larval fish (Kaushik
et al., 1982). Under these conditions, however, attention should be paid to the maintenance of a constant flow rate and the precise measurement of low concentrations of ammonia in the outlet water. Application of such a tech- nique has revealed postprandial patterns of ammonia nitrogen excretion to be very similar among phylogenetically different species (Dosdatet al., 1996;
Chakraborty and Chakraborty, 1998).
Urinary cannula or noninvasive measurement of the urine flow rate in conjunction with spot sampling of urine (Curtis and Wood, 1991) is another approach that has been used to estimate the urinary excretion of glucose and UE of fish (Bureauet al., 1998; Denget al.,2000).
Because direct measurement of UE+ZE requires sophisticated and time- consuming techniques, the use of an indirect method to estimate UE+ZE based on nitrogen losses by the fish is considered simpler (Cho and Kaushik, 1985). Since UE+ZE occurs mainly as nitrogenous product losses, the total nonfecal nitrogen loss, branchial and urinary, is estimated by the difference between digested nitrogen and recovered nitrogen as shown in the following expression:
ZN+UN=DN−RN
ZE+UE=(ZN+UN) 24.9 kJ g−1N ME=DE−(ZE+UE)
where ZN is branchial N loss; UN, urinary N loss; DN, digestible N intake;
RN, recovered tissue N; ZE, branchial energy loss; UE, urinary energy loss;
ME, metabolizable energy; and DE, digestible energy.
It has been determined that, in general, ammonia represents at least 85% of the nitrogenous wastes, whereas urea represents less than 15%
(Kaushik and Cowey, 1991). The energy of combustion value of ammonia (82.3% N, by weight) and urea (46.7% N, by weight) is 20.5 kJ/g (24.9 kJ/g N) and 10.5 kJ/g (22.5 kJ/g N), respectively (Bradfield and Llewellyn, 1982).
Because most nitrogen losses are as ammonia, and the difference in the amount of energy loss per gram of nitrogen between ammonia and urea is small, it has been proposed that the loss of 1 g of nitrogen by fish under normal conditions can be equated to an energy loss of 24.9 kJ.
1.8