Bia: Does It Work Or Not?

Thursday, September 12, 2013: 1:20 PM
Izard (Statehouse Convention Center)
F. Joseph Margraf Jr. , U. S. Geological Survey, Denver, CO
Kyle J. Hartman , Division of Forestry and Natural Resources, West Virginia University, Morgantown, WV
Andrew W. Hafs , Aquatic Biology, Bemidji State University, Bemidji, MN
Field measurement of the ecological condition of fish has been constrained by the lack of a means to directly determine in vivo fat (energy) content.  Fortunately, Bioelectrical Impedance Analysis (BIA) promises a simple, nondestructive means of estimating proximate composition (e.g. water, fat, protein) for field applications with fish.  However, while there have been several successful examples of the use of BIA in fish, there have been several instances where BIA appears not to work.  Herein, we demonstrate the utility of BIA for estimating the proximate composition of chum salmon Oncorhynchus keta from the Yukon River, Alaska.  From a sample of 86 fish, we were able to estimate fat content using BIA with 88% accuracy relative to the amounts measured using laboratory methods.  With this rather robust sample we were also able to demonstrate that a minimum sample size of about 50 fish and a range of fat content of about 30% (dry weight) are necessary to achieve successful results.  Failure to meet these minimum sampling criteria, and perhaps along with other procedural problems, explain the vast majority of failed attempts to use BIA to determine fat content in fish.  Along with improved procedures and standardized methods, these minimum sampling criteria should move the use of BIA to the fore in directly determining ecological condition of live fish in the field.