Root production, turnover and respiration under two grassland types along an altitudinal gradient: influence of...

We have measured the rates of root production and death and of root respiration in situ under two grasslands along an altitudinal gradient in the northern Pennines, UK, represented by a lowland site at 171 m in an agricultural setting, and three upland sites between 480 and 845 m. One grassland was dominated by Festuca ovina and was on a brown earth soil; the other was dominated by Juncus squarrosus and Nardus stricta and occurred on a peaty gley. The natural altitudinal gradient was extended by transplantation. Although root biomass and root production (estimated using minirhizotrons) both showed pronounced seasonal peaks, there was no simple altitudinal gradient in either variable, and neither root production nor root death rate was a simple function of altitude. Increased root accumulation in summer was a function of change in the length of the growing season, not of soil temperature. Root populations in winter were similar at all sites, showing that increased production at some sites was accompanied by increased turnover, a conclusion confirmed by cohort analyses. Respiration rate, measured in the field by extracting roots and measuring respiration at field temperature in an incubator, was unrelated to temperature. The temperature sensitivity of respiration (expressed as the slope of a plot of log respiration rate against temperature) showed no simple seasonal or altitudinal pattern. Both root growth (under Festuca) and respiration rate were, however, closely related to radiation fluxes, averaged over the previous 10 days for growth and 2 days for respiration. The temperature sensitivity of respiration was a function of soil temperature at the time of measurement. These results show that root growth and the consequent input of carbon to soil in these communities is controlled by radiation flux not temperature, and that plants growing in these upland environments may acclimate strongly to low temperatures. Most carbon cycle models assume that carbon fluxes to soil are powerfully influenced by temperature, but that assumption is based largely on short-term studies and must be reassessed.