Seasonal variation in the contribution of different behavioural mechanisms to lizard thermoregulation

1. We studied seasonal changes in the thermoregulatory behaviour of the lacertid lizard Psammodroums algirus in a Mediterranean evergreen forest. Body temperatures (T-b), environmental operative temperatures (T-e) and upper and lower limits of the selected thermal range (T-sel) were lower in May than in July. 2. On average, mean deviations of T-b from T-sel (0.7degreesC in both seasons) were much smaller than those of T-e (8.3degreesC in both seasons). Thus both the accuracy (average difference between T-b and T-sel) and effectiveness (the extent to which T-b are closer than T-e to T-sel) of thermoregulation were high, and similar in both seasons. 3. However the thermoregulatory contribution of two distinct behavioural mechanisms varied markedly between seasons. Daily activity contributed significantly to the regulation of T-b in May (when a population of T-e thermometers matching lizard activity patterns would be, on average, 1.0degreesC closer to T-sel than were randomly available T-e), but not in July (when such a population would be only 0.2degreesC closer to T-sel than were randomly available T-e). The selection of sun-shade patches, the contribution of which was larger than that of daily activity in both seasons, was more important in July (when it produced a distribution of T-e that would be, on average, 3.1degreesC closer to T-sel than were randomly distributed T-e) than in May (when a population of thermometers matching the lizards' pattern of exposure to sunlight would be 1.3degreesC closer to T-sel than were randomly available T-e). 4.These changes are discussed in the light of seasonal differences in the daily patterns of among-patch variation in T-e. In spring, lizard activity was low in the early morning because even the selection of sunlit patches was of limited utility to attain T-b within T-sel; in summer, lizards could remain active at midday, despite low overall thermal suitability, by selecting shaded patches. Thus the contribution of patch selection to thermoregulation was important in the early basking period of both seasons, and at summer midday hours. 5. Our data suggest that shuttling between sun and shade, rather than selecting sun or shade, may be an additional mechanism of behavioural thermoregulation, the importance of which would be greatest at times of day when lizards use patches at random (e.g. spring midday hours), and that their mean T-b is closer to the grand mean of full sun and full shade T-e than to the mean equilibrium T-e within any type of patch.