CO2 and temperature interactions buffer flowering time changes over the last century

Background/Question/Methods

A variety of field studies have documented accelerations in flowering time during recent, warmer decades relative to earlier, cooler decades. The majority of these studies have attributed earlier flowering to increasing temperatures, while ignoring the potential direct effects of rising atmospheric [CO₂]. When accounting for flowering time shifts, it is challenging to distinguish between the simultaneous effects of rising [CO₂] and temperature within natural systems. These factors, however, can be delineated in controlled studies to determine both their independent and interactive effects. We grew twelve field-collected accessions of Arabidopsis thaliana originating from different latitudes in controlled growth chambers that allowed for reductions in ambient [CO₂]. We used a full factorial design with CO₂ treatments representing pre-industrial (270 ppm) and modern (380 ppm) periods, and temperature treatments representing changes in the northern temperate zone across the same time period, respectively (25.0, 26.3 °C). Plants were grown at long days with non-limiting nutrients, water, and light. We measured time to flowering, plant growth rate, and total plant mass at flowering.

Results/Conclusions

We show that the assumption that rising temperatures lead to faster plant growth rates, and ultimately earlier flowering, may be overly simplistic due to strong interactions between CO₂ and temperature. When comparing responses at full pre-industrial and full modern conditions, we did not detect a change in flowering time within any of the accessions. This occurred despite a significant increase in growth rate among accessions at the modern versus pre-industrial condition. The lack of a flowering time response was due to a compensating increase in total plant mass at flowering that occurred in all accessions grown at modern versus pre-industrial conditions. When only [CO₂] was increased without a change in temperature, this compensating effect disappeared. In that case, flowering time was accelerated relative to the full pre-industrial condition due to a significant increase in growth rate, with no simultaneous change in plant mass at flowering. Taken together, our data show that increases in [CO₂] since the pre-industrial period may have played a role in accelerating flowering times. This response may have been eliminated or dampened, however, in the presence of higher temperatures due to shifts in plant size at flowering. Perhaps this outcome helps to explain why accelerated flowering is not more pronounced in field studies where both [CO₂] and temperature have changed over the last century.