Dear Ecolog-L, 

Wilson et al. (2000; see citation below) compared two methods (using a
paired design), namely line transects and point counts, for collecting
data on migratory landbirds in forested wetlands.  In their publication
they stated the following: 

"...  Our goal was to compare species composition and abundance within each
forest stand using these two survey methods. However, we did not overlay
point counts on transects because of the potential to bias detections when
observers had prior knowledge of species identification and location.
Thus, we alternated survey methods between the parallel survey lines on
subsequent visits. We varied the order in which we conducted survey
methods to ensure that each method was applied to each survey line and
that each method preceded the other on approximately half the visits.
 ..."

I hope that is enough of a quote for my purposes.  What I am most
interested in is their decision to not overlay the point counts on the
line transects because this would bias detections due to prior knowledge
of individuals locations.  Obviously this increased their sampling
effort and (for now) let me say I was wondering if there is a way to
control for this bias so the point counts and line transects could be
overlain.  

It has been suggested to me that this bias can be controlled for
statistically by taking the difference between each line transect and
the overlain pooled point counts.  I was told this eliminates the
requirement of "independence".  After further consideration of this
guidance I remain unconvinced, because I believe a biased sample remains
a biased sample regardless of the statistical procedure performed upon
it.  The point count almost certainly would improve detectability of
individuals on the first and last 150m of the line transect, and I
believe any "difference" computed between them would be flawed by that
bias.  

Thank you for your thoughts, 
David 

Wilson, Twedt, and Elliott (2000).  COMPARISON OF LINE TRANSECTS AND
POINT COUNTS FOR MONITORING SPRING MIGRATION IN FORESTED WETLANDS.  J.
Field Ornithol., 71(2):345–355

Reply via email to