2.2 Experimental design
The experiment was a completely randomized block design, with 3
replicates in MSSL. Eight N addition levels were 0
(control),30,60,90,120,150, 180 and 210 kg N
ha-1, respectively. The area of each plot was 9
m2 (3 by 3 m), and there were 24 plots in total. There
was a 1-m gap between adjacent plots with separation maintained by
cement boards installed to a depth of 10 cm in the soil. The experiment
sites had been fenced and they had never received fertilizer before the
experiment was started.
Nitrogen was added as urea (the content of pure N 46.2%), when theL. chinensis resumed growth in the middle of May once a year, N
fertilizer was applied to the surface of the L. chinensisgrassland in the early evening since 2009. After N addition, 10 mm water
was slowly and evenly sprayed to each plot to make urea granule quickly
dissolved and reduce the loss of fertilizer caused by volatilization.
The same field management was applied to each plot, and maintainedL. chinensis natural growth without human interference.