Postprint version. Published in Journal of Thermal Biology, Volume 29, Issue 2, February 1, 2004, pages 91-96.
NOTE: At the time of publication, the author Emily Taylor was not yet affiliated with Cal Poly.
The definitive version is available at http://dx.doi.org/10.1016/j.jtherbio.2003.11.003.
We used intracoelomically implanted temperature dataloggers to obtain semi-continuous body temperature data and establish monthly thermal profiles for free-ranging rattlesnakes. We mimicked random and non-random point-sampling methods by selecting a single daily data point from all values or from restricted times of day to reflect common point-sampling constraints. Thermal profiles generated from point-sampling differed from those generated from semi-continuous sampling, and this difference was more apparent when point-sampling was non-random. We conclude that semi-continuous sampling provides a better estimate of thermal profiles, and that point-sampling methods are highly sensitive to deviations from true randomness.