Plants are foundational to terrestrial ecosystems and because they are sessile, are particularly reliant on physiological plasticity to respond to weather extremes. However, diversity in molecular or transcriptomic responses to temperature extremes is not well described across plants from contrasting environments. Beyond molecular responses, photosystem II (PSII) thermal tolerance traits are widely used to assay plant thermal tolerance. To explore options for improving the prediction of thermal tolerance capacity we investigated variation in the transcriptomic stress responses of 20 native Australian plants species from varied environments, using de novo transcriptome assemblies and 188 RNA-seq libraries. We documented gene expression responses for biological processes, to both hot and cold temperature treatments, that were consistent with transcriptomic stress responses seen in model species. The magnitude of some responses showed differentiation between the species from contrasting arid, alpine and temperate biomes. This variation among biomes indicated that post heat exposure, alpine and temperate species had greater shifts in expression than arid species. Changes in the median expression of biological processes were also compared to plasticity in PSII heat and cold tolerance traits. Gene expression responses showed some expected relationships with PSII thermal tolerance plasticity, but these two response types appeared to be mostly independent. Our findings demonstrate the potential for using conserved gene expression plasticity to characterize the sensitivity or capacity of plants from diverse taxa to respond to temperature extremes. The varied combinations of molecular and physiological responses of plants to temperature stimuli could help define successful strategies under future climates.

Global warming is already affecting plant phenology, growth and reproduction. A wide range of evidence indicates warming effects on reproductive and vegetative traits, as well as phenology, but seldom do studies assess these traits in concert and across the whole of a plant's life cycle, particularly in wild species. Further, while there is evidence that these effects vary between species little is known about the extent of within-species variation for plant persistence under future warming scenarios. We assessed trait variation in response to warming in Oreomyrrhis eriopoda, an Australian native montane herb, in which within-species variation in germination strategy and growth characteristics has been demonstrated. We quantified associations between developmental trajectories and population-level variation in germination timing and examined whether the next-generation traits are altered by maternal growth conditions. Warming effects were expressed in different traits during different developmental stages. The effect of warming varied as a function of germination strategy, but germination strategy itself was conserved across generations. Thus, we conclude that understand the response of wild species to warming takes a whole-of-life perspective and attention to ecologically significant patterns of within species variation.

Quantifying the magnitude of phenotypic plasticity to compare among species, populations, cultivars, or genotypes is important for revealing the ecological and evolutionary significance of plastic responses to various abiotic and biotic factors. Commonly used plasticity estimators have occasionally been found to generate different species’ plasticity rankings. However, we do not know how frequent this incongruence is or the factors that influence the occurrence thereof; nor do we know which plasticity estimator is more reliable. We first addressed these problems using a theoretical framework, revealing inherent conflicts between the reaction norm slope and plasticity indices, and the conditions affecting these conflicts. We then empirically tested the effects of the estimators on interspecific plasticity differences by reanalyzing 1248 sets of relevant data, confirming the predictions derived from our theoretical framework. Finally, we show through theoretical analyses that the reaction norm slope is more reliable than plasticity indices for interspecific comparisons.