Flowers that cool themselves: Thermal ecology of summer-blooming thistles in hot Mediterranean environments

Flower exposure to high temperature reduces the production, viability, and performance of pollen, ovules, and seeds, which in turn impairs individual fecundity and risks the survival of populations. Autonomous floral cooling could alleviate the effects of flower exposure to harmful temperatures, yet investigations on thermal ecology of flowers in hot environments are needed to evaluate the reality, magnitude, and ecological significance of thermoregulatory cooling. This paper reports a study on the thermal ecology of the flower heads (=capitula) of 15 species of summer-blooming Asteraceae, tribe Cardueae, from hot-dry habitats in the southern Iberian Peninsula. Temperature inside (T_in) and outside (T_out) capitula were assessed under natural field conditions using two complementary sampling and measurement procedures, which provided information on the relationships between the two temperatures at the levels of individual capitula (“continuous recording”) and local plant populations (“instantaneous measurements”). Baselines for the T_in–T_out relationship in the absence of physiological activity were obtained by exposing dehydrated capitula to variable ambient temperatures in the field. To assess whether the co-flowering capitula of summer-blooming Asteraceae defined collectively a distinct thermal layer, the vertical distribution of capitula relative to the ground was quantified. Bees visiting capitula were watched and temperature of the air beside the visited capitulum was measured. Results were remarkably similar for all plant species. The capitula experienced high ambient temperatures during long periods, yet their interior was cooler than the air most of the time, with temperature differentials (ΔT = T_in − T_out) often approaching, and sometimes exceeding −10°C. The relationship between T_in and T_out was best described by a composite of one steep and one shallow linear relationship separated by a breakpoint (Ψ, interspecific range = 25–35°C). Capitula were only weakly thermoregulated when T_out < Ψ, but switched to closely thermoregulated cooling when T_out > Ψ. Narrow vertical distributions of capitula above the ground and similar cooling responses by all species resulted in a “refrigerated floral layer” where most bees foraged at T_out > Ψ and presumably visited cooled capitula. Thermoregulatory refrigeration of capitula (“thermal engineering”) can benefit not only plant reproduction by reducing pollen and ovule exposure to high temperatures during the summer but also the populations of bee pollinators and other floricolous insects.