Not clear if they increase exponentially or linearly with temperature. Both fits are significant. Samples fit well when plotting the lnK versus 1/T. But rates are unstable and thus we have to set a criteria about which rate represents better the reaction to the new temperature. That happens when "a really rate" is measured. Because if I integrate the plots I am not sure if I have a rate but the total heat dissipated by the reaction as well as the CO2 accumulated by the reaction. If measured on accumulated basis then both heat and CO2 increases exponentially with temperature but are those rates? Under these paradoxes probably Q10 be more convenient for CO2 given as soil scientists do but then CO2 may be not convenient for Arrhenius adjustments if they do not represent a really rate. Anyway the Ea values obtained are fine by the calorimetric procedure but I would like to compare them with the Q10 measured by the reponse of the heat rate to temperature.
If the accumulated values are all taken over the same time interval then they are equivalent to a rate. If these increase linearly with temperature, then it is easy. The Q10 can be obtained from the slope of a plot of the rates versus temperature. The slope is just the change in rate divided by the temperature range, i.e. change in rate per change in temperature. Then multiply by 10 to get Q10.