Nutrient Metabolism

8 downloads 0 Views 194KB Size Report
3 Cod liver oil with added vitamin A (all-trans retinol: 1,000 IU/g) and vitamin D (cholecalcipherol: .... For calcium, the difference in ADC between fish fed D5 and.
Nutrient Metabolism

Dietary Chromic Oxide Does Not Affect the Utilization of Organic Compounds but Can Alter the Utilization of Mineral Salts in Gilthead Sea Bream Sparus aurata1 Felipe Fernandez,*2 Anna G. Miquel,* Roberto Martinez,* Esther Serra,* Jordi Guinea,† Francisco J. Narbaiza,* Anna Caseras** and Isabel V. Baanante** *Departamento de Ecologı´a, †Departamento de Fisiologı´a, Facultad de Biologı´a, Diagonal 645, and **Departamento de Bioquı´mica, Facultad de Farmacia, Diagonal 643, Universidad de Barcelona, 08071 Barcelona, Spain ABSTRACT This study was conducted to determine whether the level of chromic oxide supplemented to diets containing gelatinized starch as the carbohydrate source affects digestibility, body composition, growth performances, and liver enzyme activities in gilthead sea bream, Sparus aurata. Gilthead sea bream fingerlings were fed diets containing gelatinized corn starch as the carbohydrate source and several levels of chromic oxide (0, 5, 10 and 20 g/kg) for 6 wk. No effect of dietary chromium level was detected on carbon, nitrogen, or dry matter digestibility. Calcium and phosphorus digestibility were higher in fish fed the diet supplemented with 5 g/kg chromic oxide than in fish fed the other supplemented diets. Dietary chromium did not affect dry matter, carbon, nitrogen, protein, or lipid concentrations in fish. However, fish fed 5 g/kg chromic oxide generally had higher levels of calcium, phosphorus, and ash than fish fed the other Cr-containing diets. Chromium concentration was significantly higher in fish fed the diets with 0.5 and 1% chromic oxide than in fish fed the control diet. Chromium supplementation of the diets did not affect the specific growth rate, the food efficiency ratio, the protein efficiency ratio, or, protein or nitrogen retention of the fish. Blood glucose and the activity of several liver enzymes involved in carbohydrate metabolism were unaffected by dietary chromic oxide. Alanine aminotransferase was lower in the fish fed the diet with 10 g/kg of chromic oxide than in unsupplemented controls. Our results indicate that chromic oxide can be used as a neutral marker in fish nutrition studies involving organic compounds, but not mineral salts. J. Nutr. 129: 1053–1059, 1999. KEY WORDS:



chromic oxide



digestibility



growth

Chromic oxide is used as a marker in digestibility studies performed in fish and other animals (Bondi 1987, De Silva and Anderson 1995, Jobling 1994, Talbot 1985). However, several papers report doubts about its suitability for fish digestibility studies because the level of chromic oxide in the diet could affect the digestibility values obtained (Bowen 1978, Ringo 1993, Shiau and Liang 1995, Tacon and Rodrigues 1984) as well as growth performance, body composition, and carbohydrate metabolism (Shiau and Chen 1993, Shiau and Liang 1995). Tacon and Rodrigues (1984) used three diets with three external markers (chromic oxide, polyethylene, and acidwashed sand) added simultaneously and in the in the same proportions (0.5, 1, and 2%) to the same basic diet (so the diet with 0.5% chromic oxide also had 0.5% polyethylene and 0.5% acid-washed sand, etc.) and also determined crude fiber as an internal marker. They found no consistency between the results obtained with these four markers, with chromic oxide and polyethylene giving higher digestibilities for the diets with

1 2



carbohydrate metabolism



gilthead sea bream

the higher concentration (2%) of both markers, and crude fiber and acid-insoluble ash giving the same digestibility values for all three diets. They concluded that chromic oxide, at levels ,1%, and crude fiber are reliable external and internal dietary markers for use with rainbow trout (Salmo gairdneri). Shiau and Chen (1993) fed tilapia (Oreochromis niloticus x O. aureus) diets with (20 g/kg) and without chromic oxide and also suggested that, when glucose is the only carbohydrate source in the diet, the inclusion of chromic oxide could affect growth performance (increasing weight gain), body composition (increasing dry weight and ether extract content), and enzyme activities involved in carbohydrate metabolism. In contrast, no differences in these variables were found between tilapia fed supplemented and nonsupplemented diets when the source of carbohydrate was raw cornstarch (only glucose-6phosphatase activity was significantly lower in fish fed the chromic oxide diet compared to the control diet). In a later work on tilapia (Shiau and Liang, 1995), two levels of chromic oxide (5 and 20 g/kg) were incorporated into diets containing glucose or starch. Fish fed the glucose diet with 5 g/kg of chromic oxide had greater weight gain, feed efficiency ratio, protein efficiency ratio, and protein deposition than fish fed the glucose diet containing 20 g chromic oxide/

This work was supported by MEC grant PB96 –1488. To whom correspondence should be addressed.

0022-3166/99 $3.00 © 1999 American Society for Nutritional Sciences. Manuscript received 17 August 1998. Initial review completed 11 October 1998. Revision accepted 9 February 1999. 1053

1054

FERNANDEZ ET AL.

kg. The ingredient digestibilities estimated using 5 g chromic oxide/kg as the marker were greater than those estimated with 20 g chromic oxide/kg. Furthermore, fish fed the glucose diet with 5 g chromic oxide/kg had higher phosphofructokinase activity and lower tissue chromium concentration than fish fed the glucose diet with 20 g chromic oxide/kg. However, when fish were fed the raw cornstarch diets, the only effect of chromic oxide supplementation was a lowering of digestibility values. Apparent digestibilities of protein, lipid, carbohydrate, and dry matter were significantly higher for the starch diet supplemented with 5 g chromic oxide/kg than for the starch diet supplemented with 20 g chromic oxide/kg, although the magnitude of the differences observed were less than those in the glucose diets. In contrast, Ng and Wilson (1997) found that when channel catfish (Ictalurus punctatus) were fed diets with 33% glucose as the only carbohydrate source and eight levels of chromic oxide (from 0 to 10 g/kg), growth performances (weight gain, feed efficiency ratio and protein efficiency ratio), wholebody composition (percent moisture, crude protein, fat, and ash), and chromium concentration in the whole-fish carcass were not affected by the level of supplemental chromic oxide. They concluded that chromic oxide is sufficiently inert to be used as an external marker in digestibility studies in channel catfish. Finally, Shiau and Shy (1998), feeding diets with glucose as the only source of carbohydrate and eight levels of chromic oxide supplementation, presented further evidence that chromic oxide, at ;0.204 g/kg, induced maximal growth performances in tilapia. Given these conflicting results and the scarcity of data on this subject, the present study was performed to examine the possibility that the level of chromic oxide included in diets containing gelatinized cornstarch as the only carbohydrate source could affect nutrient digestibility, whole-body composition, growth performances, blood glucose, or the activity of the enzymes involved in carbohydrate and protein metabolism in gilthead sea bream (Sparus aurata). In this work, we used cornstarch because, in addition to providing energy that can spare protein, it can also act as a binding agent and is, therefore, the type of carbohydrate usually included in practical fish diets. However, there is less information on the effects of chromic oxide inclusion in diets that contain starch than in diets that contain glucose as the carbohydrate source. The results of Shiau and Chen (1993) and Shiau and Liang ( 1995) for glucose diets are contradicted by those of Ng and Wilson (1997) and by other studies by the same group (Shiau and Shy 1998), but no other work is available to confirm the results obtained by Shiau and Chen (1993) and Shiau and Liang (1995) for starch diets. We used a gelatinized form of starch because this form facilitates its digestibility by carnivorous fish, such as S. aurata (De Silva and Anderson 1995, Jobling 1994, Wilson 1994). To our knowledge, no work has been done to check the effect of chromium oxide inclusion in the diet when this form of starch is used as the carbohydrate source. We studied enzyme activities related to carbohydrates because chromium is an essential nutrient in vertebrates, with a role in carbohydrate metabolism, probably as a cofactor acting on insulin or on insulin receptors (Anderson 1997, Davis and Vincent 1997). Therefore, enhanced chromium absorption from chromium-supplemented diets could affect enzyme activities regulated by insulin.

MATERIALS AND METHODS Gilthead sea bream, with an initial weight of 15–18 g, were acquired from a fish factory (Aquadelt, San Carlos de la Rapita, Tarragona, Spain) and distributed in 12 260-L tanks, 18 fish per tank. The tanks were inside isothermic rooms, with the temperature regulated at 21 6 0.2°C, and the light-dark cycle programmed at 12:12 h. Each aquarium was connected to an independent, closed circuit, sea water system, so water and remains from each aquarium never mixed with those of other aquariums. Salinity was maintained at 38 6 0.5 g/L by the addition of distilled water when necessary. Water in the aquariums and connected systems were renewed at an rate of ;30% per wk and checked frequently for good quality conditions, such as pH and ammonia (Ferna´ndez et al. 1996 and 1998). The fish were fed the diet without chromic oxide (D0)3 (Table 1), at a daily ration equivalent to 2g/100g body weight provided in a single meal at 09:30 h. Every 7–10 d, all fish were weighed after being anesthetized with tricaine (MS-222), to adjust ration size. After 2 wk of acclimation to laboratory conditions, the fish were weighed. Two fish from each aquarium were taken for determination of initial composition, and the fish were fed one of the four diets shown in Table 1. The diets were identical except for the level of chromic oxide (0, 5, 10, and 20 g/kg), which was substituted for equivalent amounts of starch. These diets are referred to as D0, D5, D10, and D20, respectively. We avoided the use of any inert materials that could interfere with digestibility measurements. The ration level, feeding regimen, salinity, temperature, light-dark cycle, and weighing schedule were the same as during the acclimation period. After 3 wk, samples of feces were taken for 2–3 consecutive days by pipetting from the bottom of the aquariums. The feces were recovered in the afternoon (6 – 8 h after feeding) as soon as they were voided by the fish. The feces were filtered through fine nets (500 mm), immediately transferred to flasks, dried in an oven at 65°C for 24 – 48 h and stored at 220°C until analysis. The feces recovered on different days, but coming from the same aquarium, were pooled. These samples of voided feces (VF) were used in digestibility calculations. After 5 wk, fish were stripped under anesthesia (MS 222 diluted at 75 mg/L in sea water) to obtain fecal material (STR) for digestibility measurement. This sampling technique had the advantage of avoiding leakage that can occurs when the feces are voided to the water. After 6 wk, to conduct further digestibility and composition measurements, 10 fish from each aquarium were killed in the afternoon while anesthetized. Fish that were fed in the morning were killed in the afternoon at the time they usually voided feces. The fish were immediately dissected and the gut was emptied and its contents divided into sections. Samples from the intestine were separated into the anterior region (AR) samples, corresponding to the first 2–3 cm after stomach and pyloric caeca; the posterior region (PR), corresponding to the 2–3 cm before the rectum; and the rectum region (RR), corresponding to last 1–2 cm of the gut. Samples from the stomach were discarded. Samples from a given region of all fish in a given aquarium were pooled, dried in a oven (48 –96 h at 65°C) and stored at -20°C in sealed tubes until analysis. Once the gut was emptied, each fish was weighed for final fresh weight, dried in an oven at 60°C to constant weight (dry weight), and then also stored in sealed containers in a refrigerator at 220°C until analysis. After killing 10 fish for the above mentioned measurements, 5– 6 fish remained alive in each aquarium. These fish were starved for 24 h and then killed by cervical section after being anaesthetized (MS 222 at 75 mg/L). Blood was extracted from the dorsal aorta, and the samples from the 5– 6 fish in each aquarium were

3 Abbreviations used: ADC, apparent digestibility coefficient; ALT, alanine aminotransferase (EC 2.6.1.2); AR, samples obtained from the anterior region of intestine; ASAT, aspartate aminotransferase (EC 2.6.1.1); CI, condition index; DO–D20, diets with 0 –20 g/kg of chromic oxide; FBPase-1, fructose-1,6-bisphosphatase 1 (EC 3.1.3.11); FER, food efficiency ratio; MS-222, tricaine; NRE, nitrogen retention efficiency; PER, protein efficiency ratio; PFK-1, 6-phosphofructo 1-kinase (EC 2.7.1.11); PR, samples obtained from the posterior region of intestine; PRE, protein retention efficiency; RLW, relative liver weight; RR, samples obtained from the rectum; SGR, specific growth rate; STR, samples of intestine obtained by stripping; VF, voided feces obtained by pippeting.

EFFECTS OF CHROMIC OXIDE ON SPARUS AURATA

1055

TABLE 1 Formulation and composition of test diets Diets1

D0

D5

D10

D20

68.16 4.98 21.36 1.61 1.39 2.00 0.50

68.16 4.98 20.86 1.61 1.39 2.00 1.00

68.16 4.98 19.85 1.61 1.39 2.00 2.00

4.4 6 0.6

4.5 6 0.5

4.6 6 0.5

g/100g fresh diet Formulation Fish meal2 Fish oil3 Starch4 Vitamin mixture5 Mineral mixture6 Carrageenan Chromic oxide Proximate composition7 Moisture

68.16 4.98 21.86 1.61 1.39 2.00 0.00 4.5

6 0.6

g/100 g dry diet Protein Lipids Carbohydrates8 Ash Elemental composition9 Carbon Nitrogen Phosphorus Calcium Chromium

51.0 9.7 25.5 13.8

6 0.4 6 0.3 6 0.3 6 0.1

46.5 6 0.8 8.2 6 0.4 2.3 6 0.3 3.3 6 0.5 0.003 6 0.002

52.4 9.4 24.3 13.9

6 0.2 6 0.3 6 0.6 6 0.1

51.8 9.9 24.1 14.2

46.1 6 0.6 8.3 6 0.1 2.2 6 0.3 3.5 6 0.2 0.31 6 0.02

6 0.9 6 0.4 6 0.6 6 0.1

51.6 9.5 24.2 14.7

6 0.1 6 0.2 6 0.2 6 0.1

46.1 6 0.3 8.3 6 0.3 2.2 6 0.1 3.4 6 0.3 0.66 6 0.07

45.8 6 0.5 8.2 6 0.2 2.3 6 0.2 3.5 6 0.3 1.35 6 0.06

20.39

20.20

kJ/g dry weight Energy content Gross energy10

20.35

20.37

1 2 3 4 5

D0 –D20: Diets with chromic oxide inclusion levels from 0 to 20g/kg diet. Standardized Norwegian capelin meal (NORSE-LT94). Sogem Iberica (Barcelona, Spain). Cod liver oil with added vitamin A (all-trans retinol: 1,000 IU/g) and vitamin D (cholecalcipherol: 100 UI/g). Bonafont Quı´mica (Barcelona, Spain). Gelatinized corn starch. Campo Ebro (Zaragoza, Spain). Vitamin mixture B (per kg of diet): all-trans retinol, 3 mg; cholecalcipherol, 25 mg; all-rac–tocopherol acetate, 60 mg; menadione.HNaSO3, 10.6 mg; folic acid, 4.8 mg; nicotinic acid, 360 mg; riboflavin, 24 mg; thiamin.HCl, 30 mg; pyridoxine.HCl, 24 mg; cyanocobalamin, 0.12 mg; ascorbic acid, 200 mg; calcium pantothenate, 30 mg; biotin, 0,42 mg; choline chloride, 3.2 g; myo-inositol, 900 mg. 6 Mineral mixture (per kg of diet): HCaPO4.H2O, 8.83 g; CaCO3, 2.56 g; KCl, 1.25 g; MgO, 1.0 g; ZnO, 90 mg; FeCO3, 80 mg; MnO2, 12.5 mg; CuSO4, 7.5 mg; KI, 1.5 mg; Na2SeO3, 0.28 mg. 7 Values are means 6 SD, n 5 3. 8 Calculated by difference. 9 Values are means 6 SD, n 5 4. 10 Calculated from gross composition (protein: 24 kJ/g; lipid: 39 kJ/g; carbohydrate: 17 kJ/g).

pooled. Serum was extracted after coagulation and centrifugation and stored at 220°C for analysis. The liver was dissected, immediately frozen in liquid nitrogen, and kept at 280°C until used for assaying enzyme activities. Samples of diets, fish, and feces were analyzed for carbon, nitrogen, chromium, phosphorus, and calcium contents. Procedures and analytical techniques were as described elsewhere (Ferna´ndez et al. 1996 and 1998). Diets (n 5 4/diet), fish (n 5 4/aquarium), and feces (two replicate analysis per each pooled sample) were analyzed for carbon and nitrogen content with a CHN analyzer (Carlo Erba NA 1500, CE Instruments, Thermoquest Italia, Milan, Italy) and for P, Ca, and Cr content with an inductively coupled plasma spectrometer (Polyscan 61E, Thermojarrell Ash Corporation, Franklin, MA) after acid digestion of the samples. The digestion procedures followed closely the method of Furukawa and Tsukahara (1966). Samples of diets (n 5 3/diet) and fish (n 5 2– 4 fish/aquarium) were also analyzed for protein, lipid, and ash content, following standard procedures (Helrich 1990). Carbohydrates were calculated by difference. Apparent digestibility coefficient (ADC) of a given nutrient was calculated from the following equation:

ACD 5 100 2 100 z

S

D

% chromium in food % nutrient in feces z % chromium in feces % nutrient in food

For dry matter, the equation was: ADC 5 100 2 100 z

S

% chromium in food % chromium in feces

D

Growth rates were calculated for each aquarium as a specific growth coefficient (SGC) resulting from the following expression: SGR 5 ~ln Wf-ln Wi! z 100/t taken from Jobling (1994), where Wf is the mean final fresh weight for the fish in each aquarium, Wi is the mean initial fresh weight of the fish in the same aquarium, and t is time in days. Other parameters calculated for each aquarium were food efficiency ratio (FER) and protein efficiency ratio (PER), according to the following expressions: FER 5 g weight gain/g feed provided PER 5 g weight gain/g feed protein

FERNANDEZ ET AL.

1056

The retention efficiency for protein (PRE) and nitrogen (NRE) in a given aquarium were also calculated according to the equations: PRE 5 g protein gain z 100/g feed protein NRE 5 g nitrogen gain z 100/g feed nitrogen where the nutrient gain (protein or N) was calculated from weight increase and the nutrient content (protein or N) of the initial and final samples of fish. The feed nutrient was calculated from the food provided and the content of food in either protein or nitrogen. For each fish, the relative liver weight (RLW) and the condition index (CI) were also calculated for the initial and final samples, using the following equations: RLW 5 liver weight~ g! z 100/bod y wt ~ g! CI 5 bod y wt~ g! z 100/ @bod y length ~cm!# 3 Serum glucose was determined using a commercial kit (GLUC GDH, Roche Diagnostics, F. Hoffmann-La Roche Ltd., Basel, Switzerland) for measurement in a COBAS MIRA S analyzer (Roche Diagnostics). For determining liver enzyme activities, crude extracts were obtained by centrifugation (at 15,000 x g for 20 min) of powdered frozen liver homogenized (1/5, wt/v) in 50 mmol Tris-HCl/L, pH 7.5; 4 mmol EDTA/L, 50 mmol NaF/L, 0.5 mmol PMSF/L, 1 mmol/L DTT, and 250 mmol sucrose/L using a Polytron homogenizer (PTA-7). The assays for 6-phosphofructo 1-kinase (PFK-1, EC 2.7.1.11), fructose-1,6-bisphosphatase 1 (EC 3.1.3.11), pyruvate kinase (EC 2.7.1.40), glucose 6-phosphate dehydrogenase (EC 1.1.1.49) and 6-phosphogluconate dehydrogenase (EC 1.1.1.43) activities and total protein were modified for measurement in a COBAS MIRA S analyzer, based on previously described procedures (Bonamusa et al. 1992). Alanine aminotransferase (ALT, EC 2.6.1.2) and aspartate aminotransferase (ASAT, EC 2.6.1.1) activities were assayed with kits from Roche for routine determinations using the COBAS MIRA S automatic analyzer. All enzyme assays were carried out at 30°C and measured at 340 nm. The data were analyzed by ANOVA. For ADC data, we applied a two-way ANOVA (the effects tested were type of fecal sample and diet) taking the pooled samples from each aquarium as the unit of measurement (n 5 3). For the rest of the data (fish composition, growth performances, and enzyme activities) we applied a one-way ANOVA (the effect tested was diet) also taking the mean value for each aquarium as a single number for ANOVA (n53). In both cases, individual mean differences were determined by Duncan’s multiple range test (Duncan 1955). In all cases, we used a computer program (SuperANOVA, Abacus Concepts, Berkeley, CA). Linear regressions were also calculated with the same program.

RESULTS The type of fecal sample significantly affected digestibility values (Table 2). For carbon, nitrogen, and dry matter, there was a significant increase in ADC along the gut, generally in the order AR , PR 5 STR , RR , VF. For phosphorus, ADC was lower in AR samples than in other samples. For calcium, ADC differences were found including AR ,, PR, PR , STR, and RR , VF, but PR 5 RR, RR 5 STR, and STR 5 VF. As reported in previous studies (Ferna´ndez et al. 1996 and 1998), we concluded that PR, RR, and STR samples are the best representatives of overall ADC. For a given type of sample, the chromium level in the diet did not affect carbon, nitrogen, or dry matter digestibilities (Table 2). However, ADC for phosphorus and calcium were affected by the level of chromium in the diet, with significantly higher values in the fish fed D5 than in those fed D20. For calcium, the difference in ADC between fish fed D5 and D10 was also significant. For RR samples there was an inverse relationship between chromium level in the diet and fish Ca and P ADC values, as follows:

Ca ADC ~RR! 5 55.4 2 1.57 Cr~r 5 2 0.93, P , 0.008! P ADC ~RR! 5 79.0 2 0.96 Cr ~r 5 20.91, P , 0.03! Dry weight percentage was not affected by the level of chromic oxide present in the diet (Table 3). The same was true for carbon, nitrogen, protein, lipid, and carbohydrate (Table 3). However, significant differences in carcass composition were found for calcium, phosphorus, and ash concentrations (Table 3). The highest concentrations of Ca, P, and ash were found in fish fed D5. In general, concentrations were lower in fish fed D0, D10, and D20, but they were not different from those in fish analyzed before the experiment began. Chromium concentration (Table 3) was lower in the fish fed the unsupplemented diet than in the fish fed the chromium-supplemented diets (except D20) or the fish analyzed at the start of the experiment (Initial). However, no differences in chromium concentration were found in the fish fed the three chromium-supplemented diets. No significant differences were found between initial (25.4 6 1.3 g) or final (50.6 6 3.7g) fish weight, nor between growth rates (SGR 5 1.61 6 0.09) of gilthead sea bream fed the experimental diets. No significant differences were found in food efficiency ratios (1.01 6 0.06), protein efficiency ratios (1.96 6 0.13), protein retention efficiency (34.4 6 2.3), nitrogen retention efficiency (36.8 6 4.0), the relative liver weight (1.81 6 0.2), or the condition index (2.13 6 0.07). The chromium level in the diet did not affect blood glucose after a postprandial period of 24 h (mean of 3.54 6 0.36 mmol/L). The only significant difference in enzyme activity was found for alanine transaminase, which was lower in the fish fed D10 (765 6 98 mU/mg protein) than in the fish fed the control diet (892 6 77 mU/mg protein). Values for fish fed D5 and D20 were intermediate. A similar pattern was obtained for ASAT, but differences between fish fed D0 (974 6 140 mU/mg protein) and those fed D10 (807 6 305 mU/mg protein) were not significant (P 5 0.19). DISCUSSION Our results demonstrate that chromic oxide supplementation, at the levels usually employed for digestibility determinations, caused some significant differences in food digestibility, body composition, and liver enzyme activities related to carbohydrate metabolism in gilthead sea bream. We found no differences in ADC of elements characteristic of the organic fraction of the food, such as carbon and nitrogen, caused by dietary chromic oxide. These results are consistent with those obtained in tilapia by Shiau and Shy (1998) for protein, lipid, and dry matter digestibilities. However, our results contradict those of Shiau and Liang (1995) and Tacon and Rodrigues (1984), detailed in the introduction. We found that fish fed diets with different chromium concentrations had ADC differences for mineral components, such as calcium and phosphorus, with a maximum ADC for the diet with the lowest chromic oxide concentration (in this case, 5 g/kg). A possible reason for these differences could be a direct interference of chromic oxide with the absorption of mineral salts, such as calcium phosphate. However, because marine fish drink sea water for osmoregulation (Evans 1993), it is possible that the high level of chromic oxide in the diet caused an increase in the amount of sea water drunk by the fish and, therefore, in the mineral salts entering the gut (Ca is particularly abundant in sea water). Because calcium and phosphates constitute a large propor-

EFFECTS OF CHROMIC OXIDE ON SPARUS AURATA

1057

TABLE 2 Apparent digestibilities of diets with different chromic oxide levels measured at different regions of intestine and voided feces of gilthead sea bream1 Sample3 Diet2 Carbon D5 D10 D20 All diets Nitrogen D5 D10 D20 All diets Phosphorus D5 D10 D20 All diets Calcium D5 D10 D20 All diets Dry matter D5 D10 D20 All diets

AR

PR

RR

STR

VF

All samples

79.2 6 80.1 6 80.6 6 80.0 6

7.9 1.5 2.0 3.8a

86.5 6 87.9 6 88.1 6 87.5 6

2.1 1.8 1.0 1.5b

90.4 6 89.2 6 90.8 6 90.1 6

1.1 0.7 1.1 1.1c

89.6 6 1.5 86.2 6 0.8 90.0 6 2.6 88.4 6 2.3bc

94.9 6 0.4 94.7 6 0.2 95.1 6 0.3 94.9 6 0.3d

88.8 6 5.9 88.1 6 5.0 89.5 6 5.2

83.6 6 84.1 6 84.2 6 84.0 6

4.8 1.1 1.0 2.3a

88.6 6 89.6 6 89.5 6 89.2 6

1.5 1.5 0.3 1.1b

92.3 6 90.7 6 91.5 6 91.5 6

0.4 0.7 0.7 0.9c

91.0 6 0.9 87.9 6 0.4 91.0 6 2.6 89.8 6 2.0b

96.0 6 0.4 95.8 6 0.4 95.8 6 0.3 95.9 6 0.3d

90.8 6 4.5 90.0 6 4.2 90.9 6 4.2

60.3 6 11 54.6 6 8.6 55.0 6 11 56.6 6 8.6a

69.2 6 2.3 70.5 6 11 64.4 6 0.9 68.0 6 5.9b

73.6 6 71.4 6 59.5 6 68.1 6

5.2 5.0 3.6 7.7b

75.6 6 0.6 71.9 6 6.3 70.4 6 3.0 72.9 6 4.5b

69.1 6 3.8 69.8 6 2.5 67.3 6 1.4 68.8 6 2.6b

70.0 6 6.7y 68.2 6 8.4xy 63.7 6 7.0x

21.4 6 16 8.8 6 15 9.7 6 15 13.3 6 13a

36.9 6 2.0 36.8 6 15 25.0 6 2.0 32.9 6 9.3b

48.9 6 4.8 37.6 6 6.7 24.6 6 3.8 37.0 6 12bc

49.8 6 2.7 39.8 6 8.0 42.1 6 0.2 44.1 6 6.6cd

51.1 6 6.8 44.9 6 2.4 46.4 6 1.8 47.5 6 4.6d

43.1 6 13y 35.0 6 15x 31.1 6 15x

70.1 6 69.4 6 68.9 6 69.5 6

78.2 6 79.8 6 77.2 6 78.4 6

83.6 6 81.5 6 80.2 6 81.8 6

79.7 6 1.3 72.8 6 1.6 74.7 6 2.9 75.6 6 3.2b

88.5 6 1.1 86.9 6 1.4 87.0 6 0.4 87.5 6 1.2d

80.5 6 7.0 78.4 6 7.0 78.4 6 6.8

9.2 2.9 3.7 4.6a

2.7 3.8 1.3 2.5b

1.9 1.3 1.6 2.0c

ANOVA (P value) Source

df

C

N

P

Ca

DM

Sample Diet Sample 3 Diet

4 2 8

0.0001 0.39 0.78

0.0001 0.40 0.46

0.0015 0.057 0.82

0.0001 0.007 0.63

0.0001 0.14 0.66

1 Values a means 6 SD, n 5 3 (aquariums per diet). For “all diets” rows, values with common letters (a,b,c,d) are not significantly different, (P . 0.05). For a given column and variable, values with common letters (x,y) are not different (P . 0.05). 2 Chromic oxide levels: 5g/ kg diet (D5), 10g/ kg diet (D10), and 20g/kg diet (D20). 3 AR: anterior region of intestine; PR: posterior region of intestine; RR: Rectum; STR: stripped samples; VF: voided feces. AR, PR, and RR are pooled samples that correspond to 10 fish per aquarium that were killed at the end of the experiment. STR and VF are pooled samples that correspond to all the live fish (n 5 16) present in each aquarium and that were collected after 3 wk (VF samples) or 5 wk (STR samples) of feeding the fish the different diets.

tion of the ash consumed, our results again do not agree with those of Tacon and Rodrigues (1984) cited above. When using chromic oxide as marker for ADC calculations, they found significant ADC differences for ash ADC among fish fed diets supplemented with 5, 10, and 20 g chromic oxide/kg, but their higher ADC appeared in the fish fed the diet supplemented with 20 g chromic oxide/kg, whereas the lowest corresponded to those fed the diet supplemented with 10 g chromic oxide/ kg. We found that the dry weight concentration of organic compounds, such as protein and lipids, or main elements, such as carbon and nitrogen, were unaffected in gilthead sea bream by the chromium content of the diet (Table 3). The protein and lipid data are consistent with those of Ng and Wilson (1997) and Shiau and Shy (1998). Our data suggest that fish fed D5 retain more Ca, P, and ash than those fed diets with either higher or lower levels of chromic oxide. Data of Shiau and Shy (1998) also confirm a maximum ash content in tilapia fed diets supplemented with 5 g chromic oxide/kg, although this was the maximum level of

chromic oxide they used, and the results are not directly comparable because they included glucose instead of gelatinized starch as the carbohydrate source in their diets. The diet with 5 g chromic oxide/kg was that with the highest ADC for Ca and P, and digestibility and deposition may be related. However, apart from assimilation, the deposition of mineral salts depends on other functions, such as excretion, which can be regulated by the fish (Evans 1993). Therefore, we do not know whether there is a causal connection or a simple coincidence between these two processes. Fish fed the chromium-supplemented diets had higher chromium contents than those fed the nonsupplemented diet (Table 3). One possible explanation is that chromic oxide is not an inert marker, and that there is substantial absorption of chromium, in the form of chromic oxide or other chromium derivatives produced by digestion, through the intestinal wall. Another explanation, suggested by Ng and Wilson (1997) to explain the results of Shiau and Liang (1995), is the possible incorporation through the fish’s gills of the chromium present

FERNANDEZ ET AL.

1058

TABLE 3 Composition of fish fed diets with different levels of chromic oxide for 6 wk1 Fish group2

Initial

D0

D5

D10

D20

g/100g fresh weight Dry weight

29.4 6 1.0

30.4 6 0.1

29.2 6 0.1

30.0 6 0.5

29.2 6 0.1

g/100g dry weight Carbon Nitrogen Phosphorus Calcium Protein Lipid Ash Carbohydrate3

50.0 9.3 2.4 3.6 56.1 26.9 14.0 3.0

6 1.9 6 0.2 6 0.2ab 6 0.1ab 6 1.4 6 2.9 6 0.4b 6 0.4

52.0 9.5 2.2 3.2 56.7 27.1 12.8 3.4

6 0.1 6 0.1 6 0.03a 6 0.1a 6 1.1 6 0.2 6 0.4a 6 0.6

50.3 9.6 2.5 3.9 57.3 26.6 14.0 2.1

6 0.2 6 0.1 6 0.06b 6 0.09b 6 0.8 6 1.2 6 0.7b 6 1.5

50.5 9.6 2.3 3.4 56.6 26.7 13.2 3.5

6 1.1 6 0.7 6 0.03a 6 0.05a 6 1.4 6 2.3 6 0.6ab 6 1.0

51.1 9.8 2.2 3.3 57.6 26.4 12.9 3.1

6 0.2 6 0.4 6 0.2a 6 0.4a 6 2.4 6 3.2 6 0.8a 6 1.0

mg/kg dry weight Chromium

3.46 6 0.4c

2.01 6 0.3a

2.68 6 0.3b

3.03 6 0.3bc

2.53 6 0.4ab

1 Values are the mean 6 SD, n 5 3 aquariums. For each aquarium, 4 fish carcasses were analyzed individually, and the mean was calculated. The mean for each three aquariums corresponding to a given diet was then computed and analyzed for significant differences among diet treatments. For a given row, means with the same letters are not significantly different (P . 0.05). 2 “Initial” corresponds to fish analyzed at the start of the experiment; “D0 –D20,” corresponds to fish fed for 6 wk diets supplemented with 0 –20 g chromic oxide/kg, respectively. 3 Calculated by difference (100 2 [protein 1 lipid 1 ash]) for each fish analyzed and then calculating the mean for each three aquariums corresponding to a given diet treatment.

in the aquarium water resulting from the fish voiding their chromium-containing feces in the aquarium water. Even with an efficient turnover of the aquarium water, the chromium concentration could be higher than in the aquariums containing fish fed the control diet. A third explanation could be that the chromium content of the fish follows the same pattern that we found for other inorganic nutrients (calcium, phosphate, ashes), increasing its concentration in the fish fed the diets supplemented with chromic oxide, with a maximum at a chromic oxide level around 5–10 g/kg. This increase could have more to do with a higher retention of the natural chromium present in the diet than with the absorption of the supplemented chromic oxide. The fish analyzed at the start of the experiment had levels of chromium that were even higher than those found in the fish fed any of the chromium diets. However, this was true also for calcium, phosphorus, and ash. This would indicate a natural decrease in the mineral salt concentration of dry matter as the fish grow from ;25 to ;50g. There are indications that mineral content of the fish decreases with size (Shearer 1984). This natural decrease would be balanced by chromic oxide supplementation of the diet. Our results for specific growth rate, the food efficiency ratio, and the protein efficiency ratio are consistent with those of Shiau and Liang (1995), who did not find any differences for these variables between tilapia fed diets containing raw cornstarch as the carbohydrate source and 5 or 20 g chromic oxide/kg supplementation, although they did find significant differences between fish fed glucose diets supplemented with these same levels of chromic oxide. Growth performances measured in this work were obtained in fish fed diets containing gelatinized cornstarch as the only carbohydrate source, but our results for weight gain, FER, and PER, are consistent with those of Ng and Wilson (1997) and Shiau and Shy (1988) obtained with fish fed glucose diets.

Our finding that PFK-1 did not differ among groups is in agreement with data of Shiau and Chen (1993) who fed tilapia raw cornstarch diets with 0 and 20 g chromic oxide/kg, but differs from the results found by these same authors using tilapia fed glucose diets, where significant differences in PFK-1 activity were found. In conclusion, the inclusion of chromic oxide in the diet of Sparus aurata, at 5, 10, and 20 g/kg, does not have any effect on digestibility, body composition, or growth performance of the primary organic constituents (carbon, nitrogen, dry matter, protein, and lipid), but affects digestibility and body composition of mineral constituents, such as phosphorus, calcium, chromium, and total ash, which generally seem to be utilized better at the inclusion level of 5 g chromic oxide/ kg when compared to both the control diet without chromic oxide or diets with higher levels of chromic oxide inclusion. ACKNOWLEDGMENTS We thank the staff of the Servicios Cientı´fico Tecnicos of the University of Barcelona for their help with the analysis of C, N, Cr, Ca, P, and ash. We also thank Cuidados Para Animales (CPA, Nestle´ group) for providing and analyzing the diets. We are also grateful to the Instituto de Ciencias del Mar of Barcelona for provision of sea water.

LITERATURE CITED Anderson, R. A. (1997) Nutritional factors influencing the glucose/insulin system: Chromium. J. Am. Coll. Nutr. 16: 404 – 410. Bonamusa, L., Garcı´a de Frutos, P., Ferna´ndez, F. & Baanante, I. V. (1992) Nutritional effects on key glycolitic-gluconeogenic enzyme activities and metabolite levels in the liver of the teleost fish Sparus aurata. Mol. Mar. Biol. Biotech 1: 113–125. Bondi, A. A. (1987) Animal Nutrition. John Wiley and Sons Ltd. Chichester, U.K. Bowen, S. H. (1978) Chromic oxide in assimilation studies-a caution. Trans. Am. Fish. Soc. 107: 755–756.

EFFECTS OF CHROMIC OXIDE ON SPARUS AURATA Davis, C. M. & Vincent, J. B. (1997) Chromium in carbohydrate and lipid metabolism. J. Biol. Inorg. Chem. 2: 675– 679. De Silva, S. & Anderson, T. A. (1995) Fish Nutrition in Aquaculture. Chapman and Hall, London, UK. Duncan, D. (1955) Multiple range tests and multiple range F tests. Biometrics 11: 1– 42. Evans, D. H. (1993) Osmotic and ionic regulation. In: The Physiology of Fishes (Evans, D. H., ed.), pp. 315–341. CRC Press, Boca Raton, FL. Ferna´ndez, F., Miquel, A. G., Cumplido, L. R., Guinea, J. & Ros, E. (1996) Comparisons of faecal collection methods for digestibility determinations in gilthead sea bream. J. Fish Biol. 49: 735–738. Ferna´ndez, F., Miquel, A. G., Guinea, J. & Martı´nez, R. (1998) Digestion and digestibility in gilthead sea bream (Sparus aurata): The effect of diet composition and ration size. Aquaculture 166: 67– 84. Furukawa, A. & Tsukahara, H. (1966) On the acid digestion method for the determination of chromic oxide as an index substance in the study of digestibility of fish feed. Bull. Jpn. Soc. Sci. Fish. 32: 502–506. Helrich, K., ed. (1990) Official Methods of Analysis, 15 th ed. Association of Official Analytical Chemists, Arlington, VA. Jobling, M. (1994) Fish Bioenergetics. Chapman & Hall, London, UK. Ng, W.-K. & Wilson, R. P. (1997) Chromic oxide inclusion in the diet does not affect glucose utilization or chromium retention by channel catfish, Ictalurus punctatus. J. Nutr. 127: 2357–2362. Ringo, E. (1993) Does chromic oxide (Cr2O3) affect faecal lipid and intestinal

1059

bacterial flora in Arctic charr: Salvelinus alpinus (L.)?. Aquacult. Fish. Manage. 24:767–776. Shearer, K. D. (1984) Changes in elemental composition of hatchery-reared rainbow trout, Salmo gairdneri, associated with growth and reproduction. Can. J. Fish. Aquat. Sci. 41: 1592–1600. Shiau, S. Y. & Chen, M. J. (1993) Carbohydrate utilization by tilapia (Oreochromis niloticus x O. aureus) as influenced by different chromium sources. J. Nutr. 123: 1747–1753. Shiau, S. Y. & Liang, H. S. (1995) Carbohydrate utilization and digestibility by Tilapia, Oreochromis niloticus x O. aureus, are affected by chromic oxide inclusion in the diet. J. Nutr. 125: 976 –982. Shiau, S. Y. & Shy, S. M. (1998) Dietary chromic oxide inclusion level required to maximize glucose utilization in hybrid tilapia. Oreochromis niloticus x O. aureus. Aquaculture 161: 357–364. Tacon, A.G.J & Rodrigues, A.M.P. (1984) Comparison of chromic oxide, crude fiber, polyethylene and acid-insoluble ash as dietary markers for the estimation of apparent digestibility coefficients in rainbow trout. Aquaculture 43: 391–399. Talbot, C. (1985) Laboratory methods in fish feeding and nutritional studies. In: Fish Energetics: New Perspectives (Tytler, P. & Calow, P., eds.), pp. 125–154. Croom Helm, London, UK. Wilson, R. P. (1994) Review: Utilization of carbohydrate by fish. Aquaculture 124: 67– 80.