Predicting the age of ancient <i>Thuja occidentalis</i> on cliffs

Abstract

In rocky, heterogeneous environments that support old-growth forests, the relationship between tree size and age is weaker than it is for trees growing in productive and homogeneous habitats. To assist in the management and conservation of ancient forests on rocky land of low productivity, it would be useful if the relationships among age, environmental heterogeneity, and morphological variability could be understood and used to develop predictive models of longevity so that extensive core sampling of trees would not be required. Here we sampled 296 mature Thuja occidentalis L. growing on limestone cliffs along the Niagara Escarpment, southern Ontario, Canada. We measured a variety of site conditions and morphological traits, including age, which varied from 51 to 1316 years. We then used redundancy analysis and multiple regression to model the relationships among age, morphology, growth rate, and environment, resulting in quantitative models predicting tree age from four subsets of variables. We subsequently tested the models on 60 additional trees not used to build the models and found that they predicted up to 78% of the variation in actual tree age. This approach could be adopted for use in other forest types to predict the age of trees without using tree-ring analysis.