Revisiting Bergmann's rule for saguaros (Carnegiea gigantea (Engelm.) Britt. and Rose): stem diameter patterns over space
The influence of winter temperatures and other climate variables are explored to determine which variables are associated with saguaro stem diameter and to determine if Bergmann's rule is applicable to saguaros. Location
The northern Sonoran Desert in Arizona, USA. Methods
Thirty saguaro populations were sampled (height, diameter, number of branches), and after adjustment for population height structure, mean relative thickness of saguaros was calculated for each plot (population). Fifty-seven climate variables were calculated for the thirty populations. Regression was run to determine which variables (cold winter, hot summer and precipitation) best predict relative thickness. Previous studies have demonstrated a significant positive relationship between winter precipitation and saguaro branching ( Drezner, 2003, Ecography, in press). To determine if relative thickness may be an artefact of branching (branches), partial correlation analysis was employed. Results
Mean March precipitation best predicts relative thickness. When only winter temperature variables are considered, none are significantly related to relative thickness. Relative thickness is not an artefact of branches. Main conclusions
(1) Rainfall, not temperature, best predicts saguaro stem thickness. In addition, despite the focus on summer rains in the literature, winter precipitation is the best predictor of thickness. (2) Bergmann's rule is not applicable to saguaro populations as has been previously suggested [e.g. Niering et al. (1963) Science142, 15], as thickness does not increase significantly with latitude. In addition, the suggested mechanism for Bergmann's rule, cold winter temperature, does not significantly predict saguaro stem diameters over the area studied. In some populations that experience high winter rainfall as well as cold temperatures, individuals likely derive thermal benefits from a larger stem diameter; however, the trend is not observed over the area studied and it does not appear to be adaptive.