Low humidity and hypersalinity reduce cold tolerance in mangroves

Macroclimatic changes are expected to radically alter coastal wetland ecosystems in the coming century. The trajectory of the response to climate warming may differ based on other concomitantly changing abiotic variables such as soil salinity and relative humidity. Thus, understanding plant responses to multiple interacting stressors is required to accurately predict coastal wetland shifts under climate change. The ongoing poleward shift of mangrove range limits has been linked with a reduction in freeze events, yet interactions between low temperature and other abiotic stressors remain underexplored. We grew two common mangroves (Avicennia germinans and Rhizophora mangle, n = 1222) from propagules for 10 months in environmental growth chambers under experimentally manipulated temperature, salinity, and relative humidity treatments that reflected the range of conditions these species experience in the field. We measured variation in growth and physiological characteristics before, during, and after low temperature exposure. For both species, resistance and resilience to low temperature stress were mediated by salinity and relative humidity conditions. Chronic chilling at 10 °C caused widespread reduction in seedling stem elongation rate, altered leaf gas exchange rates, and increased mortality, particularly under high salinity and low humidity conditions. Additional exposure to an overnight freeze (−4 °C) had relatively minor impacts. Five months after exposure to low temperatures, some R. mangle exhibited the capacity to recover from severe cold damage, but only under optimal humidity and salinity conditions. Although A. germinans were generally more resistant to low temperature stress, severely damaged plants did not recover, even in low salinity and high humidity conditions. We contend that current and future mangrove range limits are the result of interactions between multiple abiotic stressors including temperature, salinity, and relative humidity. Consequently, future modelling approaches to predicting range shifts under climate change need to consider multiple concomitantly changing abiotic variables and their interactions