Measures of cold tolerance in diploid and triploid Daphnia clones exposed to two temperatures

Despite decades of studies on the differential distribution of polyploid organisms, the causes of this pattern have yet to be elucidated. This study aimed to explore some of the possible physiological mechanisms explaining the differential northern distribution of polyploid clones of Daphnia pulex compared to the one of the diploid parental species. The critical thermal minimum (CTmin) was measured in 17 D. pulex clones of contrasted ploidy (diploid and triploid) and geographic origins (temperate and subarctic climates) reared under low and high temperatures (16 and 24 °C). Triploid clones had better cold tolerance (lower CTmin) than both sympatric and temperate diploid clones. No significant association was found between CTmin and body size nor with cell size. We suggest that triploids might express a cold shock resistant phenotype related to higher gene expression and/or fatty acid profiles. Cold tolerance can be viewed as one of the possible reasons for polyploid preponderance in subarctic climates.

Methods
Several diploid and triploid clones of Daphnia pulex from temperate and subarctic North American regions were reared for at least three generations in standardized conditions before the experiments to eliminate potential maternal effects (6 subarctic triploid clones, 4 subarctic diploid clones, 8 temperate diploid clones). Cold tolerance was tested by performing a critical thermal minimum (CTmin) challenge. A total of 543 daphnids (268 acclimated at 16°C and 275 at 24 °C respectively) from 18 clones were tested for cold tolerance. To find out which variables have a significant effect on cold tolerance, body size or cell area, linear mixed models were used (‘lme4’ (Bates et al. 2015) and ‘lmerTest’ (Kuznetsova et al. 2017) packages). Our experimental design was hierarchical, with each clone being replicated in three jars at each temperature and several individuals were sampled perjar. Clones and jar were computed as random factors. The same random effect structure was used for all models. Cell area and body size were standardized (mean = 0, SD = 1) to allow for a better interpretation of parameter estimates (see Schielzeth (2010) for more details).