Differentiation of Pregnant Shiba Goats Using Plasma Amino Acid Concentrations and Mathematical Analysis

Abstract

Development of convenient techniques to diagnose pregnancy status is important for efficient animal production. Considering that pregnancy is an exciting biological event in the maternal body, the metabolism of amino acids during this period might be different from other periods in life. Pregnant animals would then be distinguished from non-pregnant ones, if relative differences in plasma amino acid concentrations could be identified. In the present investigation, a linear discriminant analysis was performed to characterize specific and possibly unique aspects of plasma amino acid concentrations in pregnant animals. Pregnant group (PG, n=5) and non-pregnant group (NPG, n=5) of Shiba goats were subjected to this study at the Experimental Station for Bio-Animal Science, The University of Tokyo. On approximately 50^th day after mating (day 0 = day of estrus), the animals' pregnancy status was diagnosed by using an ultrasonography. In addition to body weight (BW) measurements, blood samples, from which plasma amino acid concentrations were determined, were obtained every week via the jugular vein from two groups of animals. Because variations of plasma amino acid concentrations (Ile, Leu, Glu and Tau) and BW in PG among pregnant days after mating were high (P<0.01), they were adjusted to the average days passed after mating (adjusted to 70^th, 77^th, 84^th, 91^st, 98^th, 105^th and 112^th days). An effect of pregnant status was significant only for Glu (P<0.01). Correlations between BW and plasma amino acid concentrations of PG and NPG were low, suggesting that concentrations of these plasma amino acids were not affected directly by BW gain. Groups of PG and NPG were discriminated significantly using a linear discriminant model with plasma amino acid concentrations in the middle period of pregnancy (P<0.05). These results suggest that the pregnancy status can be evaluated easily and at low cost by using mathematical models and suitable weighted variables of biological factors such as plasma amino acid concentrations.