This is an open access article under the terms of the Creat ive Commo ns Attri bution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.

Rensch's rule relates to a pattern whereby sexual size dimorphism is more female‐biased in small‐sized species and more male‐biased in large‐sized ones. We collected literature and museum data on the body size of males and females belonging to 4032 lizard species, as well as data on their reproductive modes and clutch sizes. We used phylogenetic comparative analyses, and general linear mixed models, to test Rensch's rule and examined how reproductive mode and clutch size affect sexual size dimorphism. Sexual size dimorphism was independent of clutch size in lizard species with variable clutch sizes and in oviparous lizards. Large litters were associated with female‐biased sexual dimorphism in viviparous and in scincomorph lizards. Inference regarding Rensch's rule depended on the analytical method used to identify it. The widely used, but less conservative, reduced major axis regression usually support Rensch's rule while ordinary least squares regressions mostly show isometric relationships. The rule tended to apply more to oviparous than to viviparous lizards. We infer that Rensch's rule is, at best, a weak pattern in lizards. This is especially true in viviparous lineages where females reproduce infrequently and therefore evolve large sizes to maximise fecundity, resulting in female‐biased dimorphism.

Toe fringe is the most typical morphological feature of lizards adapted to sandy environments, and it is simple in shape, can evolve repeatedly, and has a high degree of repetition; therefore, this feature is suitable for testing the adaptive convergence suggested by form-environment correlations. Phrynocephalus mystaceus mainly lives in dune habitats, has a developed bilateral toe fringe, and exhibits fast sand-burying behavior for predator avoidance. We tested the effects of resecting the medial and bilateral toe fringes on the locomotor performance and sand-burying performance of P. mystaceus. The results showed that the maximum sprint speed and acceleration on sand substrate did not significantly differ under different conditions (P > 0.05). Sand-burying performance scores of the unresected individuals were significantly greater than those of the resected individuals (P < 0.05). A partial least squares (PLS) regression analysis showed that the relative area of toe fringe was the main factor affecting the sand-burying performance of unresected P. mystaceus. For lizards without fringe, the PLS regression showed that the swinging index of the hind-limb was the main factor affecting the sand-burying performance of the lizard. A comparison of the swinging indexes of the hind-limb of the lizard under three states revealed that under the unresected states, the frequency of the swinging of the hind-limb was significantly higher than those of lizards with resected bilateral fringes, further indicating that the lizards compensated for the loss of fringe by increasing the time and frequency of swinging of the hind-limb. A path analysis also showed that the fringe affected the sand-burying performance of P. mystaceus not only directly but also indirectly by affecting the frequency of the swinging of the hind-limb. After the bilateral toe fringe was removed, a significant negative correlation between locomotor and sand-burying performance was observed (P < 0.05). Taken together, these results provide experimental evidence that toe fringe is positively associated with the sand-burying performance of P. mystaceus.