Five species of lichens, the majority members of a soil-crust community ( Cladonia convoluta, Diploschistes muscorum, Fulgensia fulgens, Lecanora muralis, Squamarina lentigera) showed seasonal changes of temperature sensitivity of their dark respiration (DR) to such an extent that several substantially met the definition of full acclimation, i.e. near identical DR under different nocturnal temperature conditions during the course of the year. C. convoluta, for example, had maximal DR at 5 degrees C of -0.42, -1.11 and -0.09 nmol CO(2) g(-1) s(-1) in autumn, winter, and summer, respectively, a tenfold range. However, at the mean night temperatures for the same three seasons, 9.7 degrees C, 4.2 degrees C and 13.6 degrees C, maximal DR were almost identical at -1.11, -0.93, and -1.45 nmol CO(2) g(-1) s(-1). The information was extracted from measurements using automatic cuvettes that continuously recorded a sample lichen's gas exchange every 30 min under near-natural conditions. The longest period (for L. muralis) covered 15 months and 22,000 data sets whilst, for the other species studied, data blocks were available throughout the calendar year. The acclimation of DR means that maximal net carbon fixation rates remain substantially similar throughout the year and are not depressed by increased carbon loss by respiration in warmer seasons. This is especially important for lichens because of their normally high rate of DR compared to net photosynthesis. We suggest that lichens, especially soil-crust species, could be a suitable model for fungi generally, a group of organisms for which little is known about temperature acclimation because of the great difficulty in separating the organism from its growth medium. Fungi, whether saprophytic, symbiotic or parasitic, including soil lichens, are important components of soil ecosystems and contribute much of the respired CO(2) from these systems. Temperature acclimation by fungi would mean that expected increases in carbon losses caused by global climate warming from soil ecosystems might not be as extensive as first thought. This would ameliorate this positive feedback loop present in some climate models and might substantially lower the predicted warming.