The dynamics and demography of roots were followed for 5 years that spanned wet and drought periods in native, semiarid shortgrass steppe grassland exposed to ambient and elevated atmospheric CO2 treatments. Elevated compared with ambient CO2 concentrations resulted in greater root-length growth (+52%), root-length losses (+37%), and total pool sizes (+41%). The greater standing pool of roots under elevated compared with ambient CO2 was because of the greater number of roots (+35%), not because individuals were longer. Loss rates increased relatively less than growth rates because life spans were longer (+41%). The diameter of roots was larger under elevated compared with ambient CO2 only in the upper soil profile. Elevated CO2 affected root architecture through increased branching.
Growth-to-loss ratio regressions to time of equilibrium indicate very long turnover times of 5.8, 7.0, and 5.3 years for control, ambient, and elevated CO2, respectively. Production was greater under elevated compared with ambient CO2 both below- and aboveground, and the above- to belowground ratios did not differ between treatments. However, estimates of belowground production differed among methods of calculation using minirhizotron data, as well as between minirhizotron and root-ingrowth methods. Users of minirhizotrons may need to consider equilibration in terms of both new growth and disappearance, rather than just growth.
Large temporal pulses of root initiation and termination rates of entire individuals were observed (analogous to birth–death rates), and precipitation explained more of the variance in root initiation than termination. There was a dampening of the pulsing in root initiation and termination under elevated CO2 during both wet and dry periods, which may be because of conservation of soil water reducing the suddenness of wet pulses and duration and severity of dry pulses. However, a very low degree of synchrony was observed between growth and disappearance (production and decomposition).