Research Document - 2011/025

Sources of Bias and Uncertainty in Seal Diet Composition: Hard Part and Fatty Acid Analysis

By W.D. Bowen

Abstract

Diet estimation in marine mammals relies mostly on indirect methods. The most common methods and still widely used are the recovery of hard parts from stomach contents, intestines, and faeces. Several chemical methods also have been developed, including quantitative fatty acid signature analysis (QFASA). Both of these approaches have been used to estimate the diet of seals. Although based on different assumptions and methods, both approaches are subject to sources of variation and to potential biases. Experimental evidence shows that digestion strongly influences both the number and size of hard parts that can be recovery in stomachs, intestines and faeces. Number correction factors (NCF) and digestion coefficients have been developed to reduce the biases caused by the effects of digestion on hard parts recovered from faeces. Although more work needs to be done on sources of variation in the correction factors, experiment evidence clearly shows that reasonable estimates of diet are dependent on the application of such corrections. The use of fatty acids depends on prey having distinct fatty acids signatures and the application of calibration coefficients to account for differential metabolism of prey fatty acids which influence their deposition in predator fat stores, such as blubber. Quantitative estimates of diet composition are made using QFASA, a statistical model that estimates which prey species and amounts must have been eaten to account for the fatty composition of the predator. Experimental studies indicate that generally estimates of diet can be accurately made, but these studies also reveal that significant errors of the magnitude also seen using hard parts can occur. False positive and false negative errors can occur with both approaches and obtaining a representative sample from which to infer diet may be the most significant challenge in estimating diet regardless of the approach used. Further experiments are needed to better understand the sources of variation in otolith erosion and the effect of variation in calibration coefficients, number of prey seals and the fatty acid set used in QFASA on the accuracy of estimates.

Complete document (PDF, 165 KB)

Accessibility Notice

This document is available in PDF format. If the document is not accessible to you, please contact the Secretariat to obtain another appropriate format, such as regular print, large print, Braille or audio version.