. 2020 Oct 28;10:18451. doi: 10.1038/s41598-020-75532-5

Table 2.

Drivers of significant trends in response variables identified via linear mixed effect modelling in the years 2004, 2009, 2014 (monthly resolution, January to December).

Variable	Stats	Interc	N_dep	S_dep	Year	Temp	Precip	Discharge	df N	R²
DIN^log	Slope	43.789	0.002		− 0.022	− 0.031	− 0.001	0.015	115	0.92
DIN^log	p	< 0.001	0.001		< 0.001	< 0.001	0.045	< 0.001	180	0.92
TON^log	Slope	1.624		− 0.001		0.009		0.013	117	0.74
TON^log	p	< 0.001		0.003		0.010		< 0.001	180	0.74
DIN:TP^sqrt	Slope	162.356	0.015		− 0.080	− 0.122	− 0.006		116	0.70
DIN:TP^sqrt	p	0.008	< 0.001		0.009	< 0.001	0.028		180	0.70
DIN:TON^sqrt	Slope	60.827		− 0.001	− 0.030	− 0.023	0.000	0.000	115	0.63
DIN:TON^sqrt	p	< 0.001		< 0.001	< 0.001	< 0.001	0.016	0.803	180	0.63
TOC:TON^sqrt	Slope	71.271			− 0.033	− 0.018		0.004	117	0.20
TOC:TON^sqrt	p	< 0.001			< 0.001	< 0.001		0.001	180	0.20

Statistical results (Stats) are presented as slope, p value, and R². For abbreviations: Interc. = Intercept, N_dep = N deposition, S_dep = S deposition, Temp = Temperature, Precip = Precipitation. Transformation of response variables (log- or square-root) to ensure normality is indicated by superscripts, “log” and “sqrt”, respectively. For detailed information on respective models see Supplementary Information (Table S2–S6).