Effects of environment and DBH on individual methylation level
Spatial variable, edaphic variables, topographic variables and DBH can explain over 40% of various methylation rate variations in recovering stands but less than 30% in native stands (Fig. 2a–g). Their contribution models were different in two populations. Contribution of soil factors (including alone and jointly with spatial factors and topography) to total methylation rate variation in two stands declined slowly, by 28.4% and 23.2% in recovering and native stands, respectively (Fig. 2d, 2g). However, the contribution of spatial factors changed greatly in different stands. Pure contribution and joint contribution of spatial factors to total methylation rate variation were 21.8% and 20.1% in recovering stands, but declined to 6.2% and 15.3%, respectively, in native stands. Contribution of topography to methylation increased from 3.2% in recovering stands to 13% in native stands.
For special edaphic factors, in recovering stands, most edaphic factors provided various degrees of contribution to full-methylation level variation (Fig. 2b), but only soil bulk density (BD), total nitrogen (TN), total phosphorus (TP), available phosphorus (AP) and elevation showed significant correlation with the variation of hemi-methylation levels (Fig. 2c). In native stands, the edaphic contribution to full-methylation level variation was mainly provided by total potassium (TK), TP, BD and elevation (Fig. 2e), and only TK and elevation had significant correlation with the level of hemi-methylation variation (Fig. 2f).
DBH did not have independent contribution to methylation rate in either recovering or native stands (Fig. 2b–g). However, in native stands, together with soil, topography and spatial factors affected 8.3% of the full-methylation rate variation, which was over one-third of all full-methylation variation interpreted by all four variables (Fig. 2e). Multiple regressions on distance matrices analysis (MRM) showed that DBH had a negative correlation with the hemi-methylation rate and positive correlation with the full-methylation rate (Fig. 3a). DBH also showed a significant inverse relationship with individual relative growth rates (Fig. 3b).