Thermal reaction norms and the scale of temperature variation: latitudinal vulnerability of intertidal Nacellid limpets to climate change

The thermal reaction norms of 4 closely related intertidal Nacellid limpets, Antarctic (Nacella concinna), New Zealand (Cellana ornata), Australia (C. tramoserica) and Singapore (C. radiata), were compared across environments with different temperature magnitude, variability and predictability, to test their relative vulnerability to different scales of climate warming. Lethal limits were measured alongside a newly developed metric of “duration tenacity”, which was tested at different temperatures to calculate the thermal reaction norm of limpet adductor muscle fatigue. Except in C. tramoserica which had a wide optimum range with two break points, duration tenacity did not follow a typical aerobic capacity curve but was best described by a single break point at an optimum temperature. Thermal reaction norms were shifted to warmer temperatures in warmer environments; the optimum temperature for tenacity (Topt) increased from 1.0°C (N. concinna) to 14.3°C (C. ornata) to 18.0°C (an average for the optimum range of C. tramoserica) to 27.6°C (C. radiata). The temperature limits for duration tenacity of the 4 species were most consistently correlated with both maximum sea surface temperature and summer maximum in situ habitat logger temperature. Tropical C. radiata, which lives in the least variable and most predictable environment, generally had the lowest warming tolerance and thermal safety margin (WT and TSM; respectively the thermal buffer of CTmax and Topt over habitat temperature). However, the two temperate species, C. ornata and C. tramoserica, which live in a variable and seasonally unpredictable microhabitat, had the lowest TSM relative to in situ logger temperature. N. concinna which lives in the most variable, but seasonally predictable microhabitat, generally had the highest TSMs. Intertidal animals live at the highly variable interface between terrestrial and marine biomes and even small changes in the magnitude and predictability of their environment could markedly influence their future distributions.