Accuracy of Moose Age Determinations from Canine and Incisor Cementum Annuli

We sent 76 canines and 77 incisors (I1) from 84 known-age moose (Alces alces) >= 2 years old sampled from near Fairbanks, Alaska, USA (2003-2011) to Matson's Laboratory (Milltown, MT) to test G. Matson's accuracy rate in estimating moose ages. To estimate ages, G. Matson counted annuli in the cementum of root tips using a Giemsa-staining technique and assumed a birth date of 1 June. We originally radiocollared moose at 9 months of age, and we extracted teeth upon death. Estimated moose ages averaged 7.0 years using canines and 6.9 using I1 teeth (range = 2-16 year), and known ages of each sample averaged 7.1 years. The accuracy rate among 76 canines was 74% and improved to 95% when ignoring errors within 1 year of the known age; comparative results among 77 I1 teeth were 66% and 94%. By far the most frequent error was a 1-year underestimate in age, particularly for moose that died in July and August, which included the seasonal transition period associated with completing peripheral annuli formation. After controlling for -1-year errors associated with the seasonal transition period, we found evidence for errors accumulating with age. We found no significant difference in accuracy based on which tooth was sectioned. However, G. Matson observed more individually distinct annuli and regular deposition patterns in canines, compared with incisors. Thus, we recommend the more easily extracted canine for estimating moose age. (C) 2015 The Wildlife Society.