Comments on a negative appraisal of taxidermic mounts as tools for studies of ecological energetics.

not discussed by these authors is that the experimental errors Walsberg and Wolf (1996) have recently questioned the value generated by taxidermic mounts can originate from two physiof taxidermic mounts for quantifying the thermal environment cal sources, namely, the biological envelope and the metallic of animals in terms of standard operative temperature (see core of the mount. The variations due to the envelope can Bakken 1992). They point out that, because of difficulties and result from several factors, including natural ones, traceable to costs associated with their construction and calibration, taxiindividual differences in skin and coat properties, and artidermic mounts are typically used in small numbers (often one) factual ones, stemming from the taxidermic treatments and and without standardization against live animals. By doing the artificial posture and arrangement of hairs or feathers. so, users assume that mounts, owing to judicious selection of These factors are discussed extensively by Walsberg and Wolf animals and careful taxidermic work, are instruments accurate (1996). enough to allow direct extrapolation of the collected data to Because the empty metallic shell that constitutes the core ‘‘average’’ living subjects. Walsberg and Wolf (1996) have of a taxidermic mount is such a simple physical system in tested this assumption by comparing values of standard operacomparison to feathered or hairy skin, most authors, including tive temperature determined from series of mounts to values Walsberg and Wolf (1996), assume that its contribution to the obtained from three species (one passerine bird and two roexperimental errors of the mounts data can be ignored. This dents) exposed to similar combinations of wind speeds and null assumption requires that the temperature of the shell be solar radiation intensities at constant air temperature (157C). uniform and therefore accurately measurable via a sensor (usuTheir data sets from the two mammals showed substantial ally located near the center of the shell). Copper is preferred intermount variability as well as poor agreement between over aluminum to build the shells, because its thermal conducmounts and living animals. These results led the authors to tivity is 70% higher. Walsberg and Wolf ’s (1996) most accurate make a negative appraisal of the technique, which, if warranted, mounts (bird) had a copper shell, and their most erratic ones casts serious doubts on the value of many studies in ecological (mammals) had an aluminum shell. It is, however, difficult to energetics. assess the importance of this factor, since it interacts with For this appraisal to be acceptable as a general conclusion other shell characteristics such as wall thickness (unspecified on the use of taxidermic mounts, it should meet the following in Walsberg and Wolf [1996]). four criteria. First, the measurements made on taxidermic In heated mounts, however, achieving internal temperature mounts must come from properly constructed units. Second, uniformity is complicated by the presence of heating elements. the species used for mounts-to-animals comparisons must conThese elements necessarily maintain a temperature gradient stitute a representative sample of species used in studies of between their surface and that of the dissipating shell. This ecological energetics. Third, these comparisons must be recreates a potential for increased thermal heterogeneity of the stricted to experimental conditions where they are theoretically shell, as well as for direct overheating of the sensor through justified. Finally, the statistical parameters used to describe the free convection and radiation. To minimize such sources of