2.2 Experimental design
The experiment was established in May 2006 using a split-plot design,
with warming as the whole-plot factor and N as the subplot factor. The
area of each main plot was 3m×4m, and a buffer distance of 3 m was set
between adjacent main plots. The four treatments included the control,
warming, N addition, and the combination of warming and N addition, each
of which had six replicates. Starting from May 2006, each heated plot
was continuously heated using a 165cm×15cm MSR-2420 infrared radiator
(Kalgo Electronics, Bethlehem, PA). The infrared radiator was installed
2.25 m above the ground of the heating cell, and the output power was
set to 2000 W. Across the 13 years of the experiment, the radiator led
to an average warming of +1.4°C during the growing season
(May–September) and of +1.2°C during the non-growing season
(October–April) at a height of 0–2 cm above the surface (Ren et al.,
2021; Wu et al., 2020). In the control plot, a simulant radiator of the
same size was hung at the same height to simulate the shading effect of
the radiator. Once a year, around the third week of June, nitrogen was
added to the soil by adding 10 g m-2NH4NO3 before rainfall.