The offspring size-number trade-off is a crucial concept in life-history theory, offering key insights into animal reproductive strategies. Our study examines the relationship between reproductive characteristics, morphological traits, and metabolism in Phrynocephalus lizards across 10 species around the Qinghai-Tibetan Plateau. Reproductive output and morphological and metabolic differentiations were analyzed using Phylogenetic Generalized Least Squares (PGLS) and phylogenetic ANOVA. The results show no significant differences in reproductive traits between oviparous and viviparous species. Snout-vent length and standard metabolic rate positively correlated with offspring mass, while no correlation was found with offspring number. The lack of a trade-off between offspring size and number suggests that larger females invest more in offspring mass than in offspring number. These were inconsistent with the classic prediction that females give priority to adjusting the number rather than the size of their offspring, enabling us to understand the evolution of the reproductive strategy in reptiles.

Data collection

The morphological variables include snout-vent length (SVL), head length (HL), abdomen length (AL), head width (HW), snout length (SL), snout width (SW), fore-limb length (FLL), and hind-limb length (HLL). HL, HW, SL, SW, FLL, and HLL refer to the length of protruding parts of the body. To minimize experimental errors, we used the same caliper with an accuracy of 0.01 mm for measurements and had the same experimenter perform the operations and readings of the data.

Phylogenetic analysis

Based on previous studies and our field work in recent years, we compiled 10 published mtDNA cytochrome b (cyt b) genes (GenBank IDs AY053901, AY053906, AY053915, AY053936, AY053946, AY053967, AY053976, AY053991, KF691627, KF691632). To construct a phylogeny for Phrynocephalus clades, we used MEGA-XI (Tamura et al., 2021) to align selected mtDNA cyt b genes with one gene from the outgroup Paralaudakia lehmanni (GenBank ID KF691618). Phylogenetic trees were constructed separately by using maximum likelihood (ML), implemented in PhyloSuite v1.2.1 (Figure 2; Zhang et al., 2020).

Statistical analysis

All statistical analyses were performed with R 4.3.1 (R Development Core Team, 2023). Phylogenetic generalized least squares (PGLS) regression and phylogenetic analysis of variance (phyANOVA) were conducted using CAPER 1.0.3 (Orme et al., 2023) and geomorph 4.0.7.99(Baken et al., 2021) to analyze the correlations between phenotypic traits, metabolism, and reproductive traits. We standardized all phenotypic data by dividing each trait by snout-vent length (SVL) to remove the influence of individual size. To mitigate collinearity in phenotypic data, PCA was conducted on the morphological traits related to head size and limb size. Separate principal components were extracted to represent head size and limb size. PGLS was conducted to account for the nonindependence of data due to the shared evolutionary history among species. The strength of phylogenetic signal in phenotypic traits was indicated by Pagel’s (1999) lambda (λ). If λ = 0, it suggests no significant correlation with phylogeny; if 0 < λ < 1, it indicates some degree of correlation with phylogeny; if λ = 1, it signifies a strong association with phylogeny.