Climate relationships with tree-ring width and δ13C of three Callitris species from semiarid woodlands in south-western Australia

Abstract

Proxy measures of climate based on tree rings can allow reconstruction of climate back past the limit of instrumental records, thereby improving understanding of natural climate variability. We assessed the dendroclimatic potential of tree-ring widths and δ13C of three broadly co-occurring species of Callitris in south-western Western Australia. Ring width chronologies of C. columellaris F.Muell., C. canescens (Parl.) S.T. Blake and C. preissii Miq. met standard measures of dendrochronological acceptability. For all three species, the Expressed Population Signal (EPS) was >0.93 and mean correlations between series in each chronology was >0.79. In contrast, δ13C chronologies were of poorer statistical quality (EPS ranged 0.59 to 0.88, mean correlations ranged 0.33 to 0.65) with both less year-to-year and lower-frequency information (lower mean sensitivity and standard deviation values). The dominant climatic signal in the ring width chronologies was related to rainfall and was strongest over the March–September season (correlations ranged 0.27 to 0.70, all P < 0.05). Consistent with the poorer quality of the δ13C chronologies compared with those from ring widths, tree-ring δ13C was also less strongly correlated with rainfall and rarely significant (P = 0.05). The weaker δ13C correlations may be due to a strong water conservation strategy by Callitris. Our analysis of the whole ring rather than latewood and low sampling effort may also have dampened the δ13C response. However, combining the ring width and δ13C chronologies using Principal Components Analysis did not enhance the extraction of a climatic signal. While the variance explained by the first principal component (PC) was high for all three species (76 to 87%), correlations between the first PC and rainfall were not stronger than for ring widths alone. Tree-ring δ13C, in conjunction with δ18O in particular, may nevertheless provide insight into physiological responses of Callitris to climate variation. However, dendroclimatic studies using Callitris to develop past rainfall records should focus on developing chronologies from ring widths. Further effort to find sites with old trees (250 years or more) should be undertaken and are likely to provide much needed information on past rainfall in Australia.