Gaskill, B. N., Rohr, S., Lucas, J. R. et al. 2011. Working with what you've got: Changes in thermal preference and behavior in mice with or without nest material. Journal of Thermal Biology. 36(3), 193-199.

In laboratories mice are typically housed at ambient temperatures (Ta) of 20–24 °C, which are below their average preferred Ta of ≈30 °C. Adjusting laboratory Ta is not a solution because preferences differ depending on activity, time of the day, and gender. We tested the hypothesis that providing mice with nesting material will allow behavioral thermoregulation and reduce aversion to colder Ta. We housed C57BL/6J mice with and without nesting material in a set of 3 connected cages, each maintained at a different Ta(20, 25, or 30 °C). Mice were confined in and given free access to the Ta options to determine if thermotaxis or nest building was the primary mode of behavioral thermoregulation. As predicted, nesting material reduced aversion to 20 °C but the overall preference, in both treatments, was still 30 °C. Inactive and nesting behaviors were more likely to be seen in contact with nesting material while active behaviors were more likely to be observed when not in contact. Nest quality increased with decreasing Tawhen mice could not use thermotaxis but nest quality was uncorrelated with Ta when thermotaxis was possible. Males decreased nest quality with increasing temperatures but females showed no correlation. We conclude that nesting material does not alter thermal preferences for 30 °C when thermotaxis is possible, indicating thermotaxis as the primary mode of behavioral thermoregulation. However, when thermotaxis is not possible, mice adjust nest shape depending on the Ta. Nesting material appears to partially compensate for cooler Ta and is especially important when mice are inactive. Therefore, nesting material may be a solution to the mismatch between laboratory Ta and mouse thermal preferences.