COS 165-6 - Plant roots in boreal peatlands under whole-ecosystem warming and elevated [CO2] track nutrients, not water

Abstract: Peatlands cover less than 3% of the global land surface but store more than one third of global soil carbon in deep deposits of peat. The interplay between the rooting depth distribution of peatland plants and peat biogeochemistry can impact peat accumulation but is understudied. Furthermore, climate change - elevated [CO2] and warming, can alter the depth at which roots are found in the peat profile. Previous work in uplands has shown that elevated [CO2] tends to increase rooting depth distribution. Additionally, rising temperatures can increase evapo-transpiration, depressing water table levels in saturated peatlands, possibly expanding the depth of oxic peat available to exploration by plant roots. In contrast, warming-induced mortality of Sphagnum mosses at the peat surface and corresponding increases in available nutrients may drive plants to allocate more of their roots in shallower portions of the peat profile. Peatlands comprise a number of co-occurring plant functional types with differing strategies for water and nutrient acquisition. We predicted that roots will grow deeper with elevated [CO2] and higher temperatures (via depressed water tables). Alternatively, increased nutrient availability in surface peat at higher temperatures may lead to shallow roots. We predict that these responses will be strongest in the plant functional types with the highest growth rates.

We tested these hypotheses at the Spruce and Peatland Responses Under Changing Environments (SPRUCE) experiment, a whole-ecosystem warming (WEW) and elevated CO2 experiment occurring in an ombrotrophic boreal bog in northern Minnesota, USA. WEW treatments ranging from 0 to +9 °C above ambient temperatures and increasing in 2.25 °C steps were initiated in 2015. These WEW treatments are crossed by ambient and elevated (+500 ppm above ambient) [CO2], resulting in a total of ten treatment plots and two untreated control plots. Rooting depth distributions were estimated from images collected at 3+ time points annually from minirhizotron tubes between 2018-2021. Elevated [CO2] did not lead to deeper root distributions, nor did plant functional types with high growth rates respond more strongly than others. Higher WEW treatments lead to shallower root distributions under ambient, but not elevated, [CO2], matching patterns in nutrient concentrations in surface peat. In contrast, the water table uniformly declined with WEW across [CO2] treatments. Thus we attribute these differential responses of rooting depth distribution to WEW at differing [CO2] to shifts in nutrient concentration and not changes in the height of the water table.