This research was conducted to compare the accuracy of two mathematical models to predict calcium (Ca) metabolism in sheep supplemented with different Ca sources. Twenty Brazilian male sheep averaging 7 to 8 months of age were fed a basal diet containing corn, soybean meal, hydrolyzed sugarcane bagasse, urea, mono-ammonium phosphate, mineral mixture and one of the following Ca sources: limestone (LM), alfalfa hay (AH), dicalcium phosphate (DP), oyster shell meal (SM), and citrus pulp (CP). On day 22 each animal received an intravenous injection of 7.7 MBq of radio-calcium (45Ca) followed by collection of blood, feces, and urine every 24 h for 7 days. Total Ca and radioactivity were measured in all samples. At the end of the collection period, animals were slaughtered and samples of liver, heart, kidney, muscles and 12th rib were taken for analysis. Calcium metabolism was evaluated by the models proposed by Vitti et al. (2000) and Fernandez (1995). A completely randomized design was used and means were compared by the Duncan test. No significant differences were observed for Ca intake among treatments. However, total fecal excretion of Ca differed and was lowest in animals fed LM. Calcium balance was negative in sheep receiving AH and CP because Ca absorption might have been compromised by the high contents of pectin and oxalate in these two supplements, respectively. Flows of Ca between the central pool and the digestive tract showed the greatest absorption of Ca on animals fed LM. Diets supplemented with LM, DP, and SM had highest Ca availability with values of 65.38, 34.12 and 32.43%, respectively. The models differed in the predictions of Ca flows among blood, bone and soft tissues but showed similar values for Ca balance in the different compartments. These discrepant results may be explained by structural difference between models.
metabolism; model; radio-calcium; sheep
