Hua, Katheline, and Bureau, Dominique P. (2009) Development of a model to estimate digestible lipid content of salmonid fish feeds. Aquaculture, 286 (3-4). pp. 271-276.

PDF (Published Version) - Published Version Restricted to Repository staff only

 

47

1

Abstract

Fish oil is increasingly substituted by cost-effective lipid sources such as plant oils or animal fats in fish feeds. This results in wide variations of fatty acid profiles of formulated fish feeds and potentially substantial differences in the digestibility of dietary lipid. A meta-analysis of data from 16 studies with rainbow trout and Atlantic salmon suggests that variations in apparent digestibility of dietary lipid can be primarily explained by the proportion of saturated fatty acids (SAFA) in total fatty acids. A broken line analysis of the data from these studies suggests that SAFA can be incorporated in diets at levels below 23% (+/- 1%) of total fatty acids without negatively affecting lipid digestibility. Multiple regression analysis of data from these 16 studies was carried out to examine the relationship between digestible lipid content of the diet and inclusion levels of SAM monounsaturated fatty acid (MUFA), polyunsaturated fatty acid (PUFA), and water temperature. This analysis yielded the following model:

Digestible lipid content (% of diet) = 0.45 SAFA-0.08 SAFA(2) + 0.86 MUFA + 0.94 PUFA + 0.03 SAFA*MUFA + 0.04 SAFA*PUFA + 0.03 temperature*SAFA (P < 0.0001, R-2 = 0.98).

The results suggest that the apparent digestibility of different types of fatty acids differ significantly and the digestibility of SAFA cannot be assumed to be additive when estimating the digestible lipid content of fish feeds. A significant positive effect of water temperature was observed on SAFA, but not on MUFA and PUFA (P < 0.05). Comparison between model predictions and experimental observations from independent studies with salmonid fish indicated that the model was reliable (r(2) = 0.99, P < 0.0001). Comparison of model prediction with data from studies carried out with a variety of fish species also suggested that the model quite accurately predicted digestible lipid content of diets for other fish species (r(2) = 0.98, P < 0.0001). The present model can be a useful tool for predicting the digestible lipid content of fish feeds formulated with a variety of lipid sources.

Actions (Repository Staff Only)

Item Control Page