Tooth damage in carnivores can reflect shifts in both diet and feeding habits, and in large carnivores it is associated with increased bone consumption. Variation in tooth condition in Icelandic arctic foxes, a mesocarnivore, was recorded from 854 individual foxes spanning 29 years. We hypothesized that annual climatic variations, which can influence food abundance and accessibility, will influence tooth condition by causing dietary shifts toward less edible prey. We examined tooth condition in relation to four climatic predictors: mean annual winter temperature, indices of both the El Niño anomaly and North Atlantic subpolar gyre (SPG), and the number of rain-on-snow days (ROS). We found unequivocal evidence for a strong effect of annual climate on tooth condition. Teeth of Icelandic foxes were in better condition when winter temperatures were higher, when the SPG was more positive, and when the number of ROS was low. We also found a substantial subregional effect with foxes from northeastern Iceland having lower tooth damage than those from two western sites. Contradicting our original hypothesis that foxes from northeastern Iceland, where foxes are known to scavenge on large mammal remains (e.g., sheep and horses), would show the highest tooth damage, we suggest that western coastal sites exhibited greater tooth damage because cold winter temperatures lowered the availability of seabirds, causing a shift in diet toward abrasive marine subsidies (e.g., bivalves) and frozen beach wrack. Our study shows that monitoring tooth breakage and wear can be a useful tool for evaluating the impact of climate on carnivore populations and that climate change may influence the condition and fitness of carnivores in complex and potentially conflicting ways.

We used the extensive collection of arctic fox lower mandibles, collected initially by the late Pall Hersteinsson, and now stored in the Icelandic Institute of Natural History. We recorded missing teeth from both the left and right mandibles (dentary bones) of each individual. The canine teeth were missing from all mandibles because they were extracted for age determination. Often, some of the incisors, the first premolar (PM1), and/or the third molar (M3) were also missing. Therefore, we focused on teeth that were most often present (PM2-PM4 and M1-M2), ten teeth in total. We examined every tooth for breakage and cracks, and calculated the fraction of damaged teeth in each specimen (i.e., out of ten teeth). We also scored the level of wear in the focal ten teeth using a five-stage wear score that ranged from minimal wear (stage 1) to heavy wear (stage 5).