Predicting white spruce cone crops in the boreal forests of southern and central Yukon.

White spruce ( Picea glauca (Moench) Voss) cone crops were measured at five regional centers in southern and central Yukon for 30 years at one site from 1986 to 2015 and at four other sites during 9 to 11 years to select the best climatic model that uses cues from growing season temperature and rainfall to predict the size of cone crops. We evaluated six climatic models that use summer temperature and rainfall of years t - 1 and t - 2 to predict cone crops in year t. July temperatures provided the best predictors of white spruce cone crops, and no rainfall variable was related to the size of cone crops. We explored three variants of July temperatures: mean temperature, degree-days > 5 °C, and maximum temperatures. For each of these, we used the Δ T model that uses the difference in the July temperature measures of years t - 1 and t - 2. We compared the resulting six models with corrected Akaike's information criterion (AICc) to determine their relative predictive performance. The best model combined Δ T measures of degree-days > 5 °C and the four highest daily maximum July temperatures with R2 = 0.65. By comparison, the Δ T model involving only mean July temperatures was less successful ( R2 = 0.49). There was good regional synchrony ( rp = 0.7 to 0.8) in high cone crops over southern and central Yukon during 1986 to 2015. [ABSTRACT FROM AUTHOR]