44 
Fishery Bulletin 109(1 ) 
sion in energy flow and proximate composition studies 
on spatial and temporal scales that were previously 
impractical. At the individual level, BIA will permit 
repeated measures on the same individual during the 
course of investigation, yielding better tracking of ener- 
getics components and improved precision in bioenergetic 
models. At the population level, BIA will permit assess- 
ment of the condition of cohorts over time and permit 
detailed comparisons across cohorts, and temporal and 
spatial scales. At the community level, BIA will permit 
the evaluation of growth and energy-flow dynamics 
across species that may elucidate community dynamics 
that were previously unknown, or permit correlation of 
condition with outbreaks of disease. This approach also 
has potential for the nonlethal study of threatened or 
endangered species by the use of models developed for 
closely related species. In order for the application of 
BIA to reach its potential, sources of error that affect R 
and X c measurements need to be continually identified 
and analyzed. 
Sources of error include different electrode needle 
locations, procedure, user experience, time periods be- 
tween death and impedance measurements, and tem- 
perature. Measuring impedance in the same anatomical 
location of the fish is critical to obtaining accurate and 
reproducible impedance measurements. When electrodes 
are not placed in the same anatomical location of fish, 
incomparable and inaccurate results are obtained, but 
which location is best is still a question. Variability in 
R and X c measurements increased with the distance of 
electrode placement from the control. Impedance values 
can change for two reasons: 1) the distance between 
electrodes is directly proportional to the electrical vol- 
ume (e.g., R in Table 1) and consequently, halving the 
distance between electrodes leads to reduced values 
of R and X c ; and 2) when electrodes are placed in dif- 
ferent locations on the fish, different tissue types are 
represented, and moving electrodes from the dorsal 
side of the fish to the ventral side will not only change 
the distance between electrodes, but it will also reflect 
different tissue types. The dorsal side is mainly muscle 
and the ventral consists of peritoneal tissue and organs. 
Changing the distance will change the R and X c values, 
and changing the electrode location will change the 
tissue types that are being measured. Sensitivity to 
tissue types is consistent with Geddes and Baker (1967) 
who reported different impedance values with differ- 
ent tissue types (i.e., skeletal muscle, liver, and kidney 
tissues). Therefore, electrodes can be re-inserted into 
the same holes or moved slightly dorsally or ventrally 
as long as the same tissue type is measured and the 
distance stays relatively similar. Impedance measure- 
ments are dependent on the anatomical location of the 
needle electrodes. 
Sources of error caused by procedural deviation can 
also be avoided by standardizing protocols. Measures 
of R and X c are affected by covariates such as needle 
depth, needle size, and conductive surfaces where the 
measurements are taken. Minimizing these errors can 
be accomplished by inserting needles to a uniform 
depth, blot drying the fish before measurements are 
taken, taking measurements on a nonconductive board, 
and by using the same gauge of needle electrodes. If 
procedures are not standardized, R and X c change as 
electrical currents are altered by procedural changes. 
For example, changing the needle depth or size will 
change the needle surface area that is in contact with 
the tissue. Because smaller surface areas present more 
resistance to the electrical current than larger ones, 
R and X c values will change. Similarly, taking imped- 
ance measurements on a conductive board offers the 
electrical current a less resistant route. Ohm’s law 
states that when electrical currents are offered a less 
