Intraspecific variations in shell calcification across thermal window and within constant temperatures: Experimental study on an intertidal gastropod Monetaria annulus

• Cowries have an ontogenetic stage exclusively spent in calcification.
• Thermal responses of shell growth rates were examined across the thermal window.
• Shell growth rates showed a typical unimodal curve against rearing temperature.
• Variance of shell growth rates increased with temperature rather than mean.
• Final shell thickness exhibited two peaks around 21 °C and 33 °C.

The relationship between temperature and shell growth was examined by rearing an intertidal cowrie species, Monetaria annulus (Gastropoda, Cypraeidae), at seven constant temperatures between 21 °C and 34 °C over four experimental years. Data on shell growth were collected during the callus-building stage, in which calcification exclusively occurs around the spiral shell constructed in the juvenile stage. Our experiments elucidated some clear trends among temperatures, although careful consideration is required to interpret the results obtained under the potential confounding factors between temperature and experimental year. Calcification was accelerated by increasing temperature in the range between 21 °C and 33 °C, and drastically slowed down at 34 °C probably due to heat stress. No relationship was found between body size and callus growth rates. The duration of callus-building indicated a negative exponential relationship with rearing temperature. However, no simple relationship was found between temperature and final shell thickness; the final shell thickness described two peaks around 21 °C and 30 °C. This finding strongly suggests that a physiological control by calcifiers cannot be ignored to explain the thermal response of calcification rates. Our results also suggest that considering the thermal seasonality as well as the difference in mean temperature, is required to explain the negative latitudinal clines in shell thickness frequently observed in molluscan species. We also focused on the variation among the individuals reared at the same temperatures. Statistical analyses suggested that the final shell thickness and the duration of calcification were better described by the normal and lognormal distributions, respectively. The distribution of shell growth rates did not considerably deviate from the normal distribution with respect to skewness and kurtosis. However, their heteroscedasticity among temperatures was inherent and significant; its variance positively depended not on the mean but on temperature. The observed pattern might be an evidence that possible stabilization selection on calcification rates is more relaxed at warmer temperatures.