Thermal variability alters climatic stress resistance and plastic responses in a globally invasive pest, the Mediterranean fruit fly (Ceratitis capitata)


Correspondence: E-mail:


Climatic means with different degrees of variability (δ) may change in the future and could significantly impact ectotherm species fitness. Thus, there is an increased interest in understanding the effects of changes in means and variances of temperature on traits of climatic stress resistance. Here, we examined short-term (within-generation) variation in mean temperature (23, 25, and 27 °C) at three levels of diel thermal fluctuations (δ = 1, 3, or 5 °C) on an invasive pest insect, the Mediterranean fruit fly, Ceratitis capitata (Wiedemann) (Diptera: Tephritidae). Using the adult flies, we address the hypothesis that temperature variability may affect the climatic stress resistance over and above changes in mean temperature at constant variability levels. We scored the traits of high- and low-thermal tolerance, high- and low-temperature acute hardening ability, water balance, and egg production under benign conditions after exposure to each of the nine experimental scenarios. Most importantly, results showed that temperature variance may have significant effects in addition to the changes in mean temperature for most traits scored. Although typical acclimation responses were detected for most of the traits under low variance conditions, high variance scenarios dramatically altered the outcomes, with poorer climatic stress resistance detected in some, but not all, traits. These results suggest that large temperature fluctuations might limit plastic responses which in turn could reduce the insect fitness. Increased mean temperatures in conjunction with increased temperature variability may therefore have stronger negative effects on this agricultural pest than elevated temperatures alone. The results of this study therefore have significant implications for understanding insect responses to climate change and suggest that analyses or simulations of only mean temperature variation may be inappropriate for predicting population-level responses under future climate change scenarios despite their widespread use.