April and June, 2014. UCSC. Thermal preferences and parasites of Sceloporus occidentalis bocourtii.
We captured 77 adult (males = 47, females = 30) coast range fence lizards, Sceloporus occidentalis bocourtii (Squamata: Phrynosomatidae), at Santa Cruz County, CA, USA (36.9852, -122.0614) during two sampling campaigns in 2014. In mid April we captured 42 lizards and in early June 35 lizards. All lizards were caught using a noose and were transported to the lab facilities in a cooler. The sex of the individuals was determined by the presence of enlarged post-anal scales (Cox et al., 2005; Megía-Palma et al., 2018). Snout to vent length (SVL) of the lizards was measured with a ruler to the nearest millimeter and their mass to the nearest gram using a digital scale. A body condition index was estimated using the Scaled Mass index (SMi). This estimation accounts for the growth effect on body size as well as for the scaling relationship between mass and body length (Peig and Green, 2009; 2010), and it is calculated using the standardized major-axis of the mass on body length (following Bohonak, 2004). A second mass measurement of the lizards captured in April (N = 42) was taken 11 days post-capture. Thus, during this period, the lizards were housed individually in plastic terraria with a damp substrate of peat moss and sand, and were provided with water and food (crickets dusted with vitamins and calcium) ad libitum. The terraria were kept in an environmental chamber with a 12L:12D photoperiod provided by natural spectrum fluorescent and ultraviolet lights and a cycling thermal regime. We used one empty terrarium with substrate to register standard temperatures 24 h/day every 5 minutes during 11 days with a data logger (hobo U23-002, Onset, Cape Cod, Massachusetts, USA). We set one probe of the data logger in one corner of the cage under the heat source (i.e., 40 Watt bulb) and other probe in the opposite corner. Within two days of capture lizards were allowed to select temperatures in experimental thermal gradients (i.e., 48-25 ºC) and their body temperature was recorded every minute during the 120 minute trial (as in Paranjpe et al., 2013). Preferred temperature (Tp) was later calculated as the arithmetic mean of measured body temperatures (Pough and Gans, 1982). Standard deviations of body temperatures (SD) were also calculated. Thus, lizards with high SD have low precision of thermoregulation (Paranjpe et al., 2013). Maximum and minimum body temperatures of each individual were also registered as measures of voluntarily selected temperatures (Tmax.v, and Tmin.v). In addition to these measurements of thermal preference, we also investigated the parasitic infections of these lizards. Tick load was quantified within the first day of capture and at day 11 by the same person (RM-P). In addition, following the methods described in Megía-Palma et al. (2018), the same person (RM-P) diagnosed chronic infections by hematic (i.e., Lankesterella occidentalis) and intestinal (i.e., Acroeimeria sceloporis) coccidians only once at the beginning of the trials.
Steps to reproduce
We investigated different parasitic infections, and their interactions, as predictors on both lizards’ thermal preferences in the lab and mass loss. As a preliminary analysis and to ascertain independence among the thermal variables used here (i.e., Tp, SD, Tmax.v, and Tmin.v), we ran a correlation matrix for these parameters. Then, we performed general linear analyses including month of capture, sex, SVL, SMi, tick load, presence of intestinal (A. sceloporis) and hematic (L. occidentalis) coccidians, and the interactions A. sceloporis*L. occidentalis, tick load*A. sceloporis, and tick load*L. occidentalis as predictors of Tp, SD, maximum, and minimum temperatures in separate analyses. We also studied the within-individual body mass change as a function of the different parasites investigated here. For this, we ran a general linear mixed analysis (GLMM) where the individual was set as random factor and the initial and final mass measurements as response variables. Body size, presence/absence of A. sceloporis and L. occidentalis, and tick load were set as predictors. A preliminary AICc comparison analysis discarded a more complex model with the interactions between the different parasitic infections (i.e., ΔAICc of the simpler model = -51). In addition, generalized linear models (GLMs) were used to investigate tick load, and prevalence (absence/presence) of Acroeimeria, and Lankesterella. For this, we fitted a GLM with negative binomial distribution linked to a log function with tick load as response and month, sex, SVL, SMi, and presence of Acroeimeria and Lankesterella as predictors. Furthermore, we ran a non-parametric Wilcoxon matched pairs test to compare tick load of the lizards maintained in captivity between the beginning and the end of the experiment. We also fitted two separated GLMs with binomial distribution linked to a logit function with Acroeimeria and Lankesterella as response and month, sex, SVL, SMi, and tick load as predictors. We confirmed normality of all models after testing their deviance residuals with Shapiro-Wilk’s normality tests. In addition, we controlled the observed heteroscedasticity of model residuals by applying a “sandwich” correction (Zeileis, 2006). Thus, we present the coefficients for corrected models. All analyses were performed in R v.3.4.3 (R core Team, 2017).