Do sexual differences in life strategies make male lizards more susceptible to parasite infection?
Description
1. Female and male hosts may maximise their fitness by differential evolution of strategies that compensate for the costs of parasite infections. The resulting sexual dimorphism might be apparent in differential relationships between parasite load and body condition, potentially reflecting differences in energy allocation to anti-parasitic defences. For example, male lacertids with high body condition may produce many offspring while being intensely parasitised. In contrast, female lacertids may express a different outcome of the trade-offs between body condition and immunity, aiming to better protect themselves from the harm of parasites. 2. We predicted that females would have fewer parasites than males and a lower body condition across parasitaemia levels, because they would invest resources into parasite defence to mitigate the costs of infection. In contrast, the male strategy to maximise access to females would imply some level of parasite tolerance and, thus, higher parasitaemia. 3. We analysed the relationship between body condition of lizards and parasitaemias of Karyolysus and Schellackia, two genera of blood parasites with different phylogenetic origin, in 565 females and 899 males belonging to 10 species of the Lacertidae (Squamata) that were sampled during a period of 12 years across 34 sampling sites in southwestern Europe. 4. The results concerning the Karyolysus infections were consistent with the predictions, with males having similar body condition across parasitaemia levels even though they had higher infection intensities than females. On the other hand, females with higher levels of Karyolysus parasitaemia had lower body condition. This is consistent with the prediction that different life strategies of male and female lacertids can explain the infection patterns of Karyolysus. In contrast, the parasitaemia of Schellackia was consistently low in both male and female hosts, with no significant effect on body condition of lizards, suggesting that lizards of both sexes maintain this parasite bellow a pathogenic threshold.
Files
Steps to reproduce
To test for a relationship between lizard body condition and parasitaemia, while controlling for the phylogenetic relationships among lizard species, phylogenetic generalized least square regression models (PGLSRM) were used. This method uses a maximum likelihood modelling approach to estimate the phylogenetically corrected partial correlation between the variables of interest (Freckleton et al., 2002). We used function pglm3.3 and libraries “MASS”, “mvtnorm”, and “ape”. A phylogenetic tree was built that included every individual lizard as a terminal branch and that had equal branch length polytomies for individuals within populations and equal branch length polytomies for populations within species (for a complete list of populations within each lizard species, see Table 1). Genetic relations and distances among the 10 lizard species were based on García-Porta et al. (2019). Considering that the smallest genetic distance among any pair of our 10 species was 0.03246 (namely between P. bocagei and P. guadarramae; García-Porta et al., 2019), the genetic distance among individuals within populations was arbitrarily set to 0.00001, and the genetic distance among populations within species to 0.0005. Although these values were arbitrary, choosing different genetic distances (for example, 0.0001 or 0.000001 among individuals, and 0.005 or 0.00005 among populations) resulted in qualitatively identical results (Tables S4 and S5). The built phylogenetic tree (Annex I in Supplementary Material) was included in the model as a design matrix with phylogenetic dependence (lambda parameter) set to 1. We included as factors in the model the year of capture, host sex, the log10-transformed parasitaemias of Karyolysus and Schellackia, and the three-way interaction between sex and the two log10-transformed parasitaemias of Karyolysus and Schellackia. The Julian date of capture (z-standardized) as proxy to the reproductive status of the lizards that might as well influence body condition, which was calculated independently for every year, and the two PC microclimate variables were included as covariates in the PGLSRM. We compared the results of the PGLSRM with those of a generalized linear mixed model (GLMM) with gamma error distribution and log link function, using the library ‘lme4’ to confirm the consistency of the results when controlling for phylogenetic inertia versus the categorical random effects of species and sampling site. We checked the GLMM for collinearity using the function ‘check_collinearity’ of library ‘performance’. It included the species nested within sampling site as a random term, the fixed factors host sex, log10-transformed parasitaemia of Karyolysus and Schellackia, capture year, and lizard species, and the three-way interaction between sex and the two log10-transformed parasitaemias of Karyolysus and Schellackia. The z-standardized Julian date and the two PC_microclimate variables were also modelled here as covariates.