Bite force is a key performance trait in lizards since biting is involved in many ecologically relevant tasks, including foraging, fighting, and mating. Several factors have been previously suggested to impact bite force in lizards, such as head morphology (proximate factors), or diet, intraspecific competition, and habitat characteristics (ultimate factors). However, these have been generally investigated separately and mostly at the interspecific level. We tested which factors drive variation in bite force at the population level and to what extent. Our study includes 20 populations of two closely-related lacertid species, Podarcis melisellensis and Podarcis sicula, which inhabit islands in the Adriatic. We found that lizards with more forceful bites have relatively wider and taller heads, and consume more hard prey and plant material. Island isolation correlates with bite force, likely by driving the resource availability. Bite force is only poorly explained by proxies of intraspecific competition. The linear distance from a large island and the proportion of difficult-to-reduce food items consumed are the ultimate factors that explain most of the variation in bite force. Our findings suggest that the way in which morphological variation affects bite force is species-specific, likely reflecting the different selective pressures operating on the two species.

In-vivo bite force was measured using an isometric Kistler force transducer (type 9203) connected to a Kistler charge amplifier (type 5995, Kistler Inc., Winterthur, Switzerland). Lizards were made to bite the plates of the transducer, left bare, at least five times, and the greatest bite force across the five trials was retained as an estimate of an individual’s maximum bite force. Gape angle was standardized across trials by adjusting the distance between the bite plates for each individual to maintain gape angle constant at around 30°. Bite position was standardized by assuring that the tips of the jaws of each lizard were up against the metal stop mounted on the device. Bite force was log₁₀-transformed before statistical analyses.

The maximum hardness of each food item was estimated using the regression equations proposed by Herrel & O’Reilly (2006) that relate prey length and hardness, according to the item hardness category.

We measured all individuals using digital calipers (Mitutoyo absolute digimatic; ± 0.01mm; see Supplementary Fig. 1) and recorded snout-vent length (SVL) and linear head dimensions including head length (HL), head width (HW), head height (HH), lower jaw length (LJL), quadrate to jaw tip length (QT), and coronoid to jaw tip length (CT). Two other functionally relevant variables were calculated: the in-lever for jaw opening (open = LJL – QT) and the in-lever for jaw closing (close= QT – CT).

For both species separately, a principal component analysis (PCA) was run on head dimensions. The contribution of each specimen along the three first principal components (PC) was extracted and used to calculate the mean contribution of each sex of each population on these axes. The sexual dimorphism in head dimensions (SDh) for each site was determined as follows:

where m_i and f_i refer respectively to the mean contribution of the males and the females of the population of interest along with the PC_i. Two other proxies for intraspecific competition were estimated: the proportion of individuals which missed the longest toe on one of the hind feet and the proportion of individuals with a regenerated tail.

Data has been log-10 transformed, expect for the proportions which have been arcsin transformed.