15 1. Manipulative experiments show a decrease in dryland biological soil crust cover and al-16 tered species composition under climate change. However, the underlying mechanisms are 17 not fully understood, and long-term interacting effects of different drivers are largely un-18 known due to the short-term nature of the studies conducted so far. 192. We addressed this gap and successfully parameterized a process-based model for the bi-20 ocrust-forming lichen Diploschistes diacapsis as a common and globally distributed rep-21 resentative of biocrust communities to quantify how changing atmospheric CO2, tempera-22 ture, rainfall amount and relative humidity affect its photosynthetic activity and cover. We 23 also mimicked a long-term manipulative climate change experiment to understand the 24 mechanisms underlying observed patterns in the field. 25 3. The model reproduced observed experimental findings: warming reduced lichen cover 26 whereas less rainfall had no effect. This warming effect was caused by the associated de-27 crease in relative humidity and non-rainfall water inputs as major water sources for li-28 chens. Warming alone, however, increased cover because higher temperatures promoted 29 photosynthesis during the cool morning hours with high lichen activity. When combined, 30 climate variables showed non-additive effects on lichen cover, and fertilization effects of 31 CO2 leveled off with decreasing levels of relative humidity. 32 4. Synthesis. Our results show that a decrease in relative humidity, rather than an increase in 33 temperature may be the key factor for the survival of dryland lichens under climate 34 change and that CO2 fertilization effects might be offset by a reduction in non-rainfall 35 water inputs in the future. Because of a global trend towards warmer and thus drier air, 36 3 this will affect lichen-dominated dryland biocrust communities and their role in regulat-37 ing ecosystem functions, worldwide. 38 39 40