Table 2.

Predicted performance of ten functions in monocultures of four forage species and the four-species equi-proportional mixture at three N fertilisation treatments (N50: 50 kg N ha⁻¹ yr⁻¹, N150: 150 kg N ha⁻¹ yr⁻¹, N450: 450 kg N ha⁻¹ yr⁻¹). Data averaged across years. Values are in % of the maximal performance per function (max) at a single year over the three-year experiment and N fertilisation treatments.

Function	Maximal performance per function (max)	N treatment	Monocultures‡					Equi-proportional mixture
			Performance (% of max)					Performance	Diversity effects§ (% of max)
			Lp	Dg	Tp	Tr	Average	(% of max)	DBGL effect	Lp∙Dg effect	Tp∙Tr effect
Yield	19.96 (t ha−1 yr−1)	N50	30.4	33.9	49.3	48.6	40.6	79.0	33.8***	2.1†	2.5†
		N150	44.3	44.9	60.2	47.0	49.1	79.2	25.5***	2.1†	2.5†
		N450	60.5	67.8	64.0	54.5	61.7	88.1	21.8***	2.1†	2.5†
		s.e.¶	4.47	4.47	4.47	4.47	2.52	2.63	4.02	1.31	1.31
SDYield	4.60 (t ha−1 yr−1)	N50	31.3	20.8	81.0	48.9	45.5	24.7	− 15.5†	− 3.1ns	− 2.2ns
		N150	17.4	23.7	58.8	37.6	34.4	31.7	2.6ns	− 3.1ns	− 2.2ns
		N450	47.4	51.4	54.0	40.0	48.2	15.5	− 27.4***	− 3.1ns	− 2.2ns
		s.e	7.27	7.27	7.27	7.27	4.10	4.28	6.53	2.13	2.13
Stability	–#	N50	41.6	59.7	24.5	37.7	40.9	68.4	18.9*	6.0**	2.7ns
		N150	64.1	53.9	37.4	45.9	50.3	63.6	4.6ns	6.0**	2.7ns
		N450	42.1	45.1	40.9	49.3	44.3	79.8	26.8**	6.0**	2.7ns
		s.e	7.22	7.22	7.22	7.22	4.07	4.26	6.53	2.11	2.11
Weed biomass	–#	N50	40.9	51.6	87.9	75.7	64.0	28.4	− 27.5**	− 0.8ns	− 7.3***
		N150	25.9	46.4	96.1	80.9	62.3	23.0	− 31.2***	− 0.8ns	− 7.3***
		N450	19.7	61.1	79.0	65.1	56.2	36.8	− 11.3ns	− 0.8ns	− 7.3***
		s.e	7.39	7.39	7.39	7.39	4.18	4.37	6.71	2.15	2.15
Nsym	374.4 (t ha−1 yr−1)	N50	7.4	6.5	45.3	56.5	28.9	67.8	37.3***	1.6ns	0.1ns
		N150	9.2	< 0.1	65.6	46.7	29.6	58.2	26.9***	1.6ns	0.1ns
		N450	2.2	1.2	36.3	31.6	17.8	29.6	10.2ns	1.6ns	0.1ns
		s.e	6.62	6.62	6.62	6.62	3.75	3.91	6.00	1.90	1.90
N efficiency	10.54	N50	26.1	29.2	58.7	73.9	46.9	85.3	37.3***	1.5ns	− 0.4ns
		N150	13.3	8.7	28.1	25.6	18.9	27.7	7.7***	1.5ns	− 0.4ns
		N450	7.5	8.3	9.9	10.0	8.9	11.4	1.5ns	1.5ns	− 0.4ns
		s.e	3.46	3.46	3.46	3.46	1.96	2.05	3.17	1.00	1.00
NO3 soil solution	–#	N50	13.4	21.6	52.6	75.1	40.7	29.4	− 11.3ns	–‡‡	–
		N150	34.9	8.0	76.6	79.5	49.8	26.9	− 22.9**	–	–
		N450	76.6	81.2	90.3	97.3	86.3	84.5	− 1.8ns	–	–
		s.e	9.09	9.09	9.09	9.09	5.13	5.99	8.93	–	–
Crude protein content	322.1 (g kg−1 DM)	N50	41.4	42.2	59.6	76.4	54.9	55.7	2.0ns	1.2ns	− 2.3*
		N150	44.2	40.2	70.6	78.2	58.3	53.4	− 3.8†	1.2ns	− 2.3*
		N450	59.5	59.1	70.2	84.0	68.2	58.8	− 8.3*	1.2ns	− 2.3*
		s.e	3.56	3.56	3.56	3.56	2.02	2.10	3.22	1.01	1.01
OM digestibility	671.0 (g kg−1 DM)	N50	93.3	88.9	83.5	92.2	89.5	88.6	− 0.6ns	− 0.6†	0.3ns
		N150	96.8	88.7	86.2	91.6	90.8	89.0	− 1.5†	− 0.6†	0.3ns
		N450	95.7	90.4	85.0	93.9	91.3	92.5	1.6ns	− 0.6†	0.3ns
		s.e	1.22	1.22	1.22	1.22	0.69	0.72	1.09	0.36	0.36
Metabolisable energy	10.77 (MJ kg−1 DM)	N50	89.4	84.3	82.2	92.3	87.1	86.6	− 0.1ns	− 0.4ns	< 0.1ns
		N150	92.5	83.3	86.0	92.2	88.5	86.2	− 1.9*	− 0.4ns	< 0.1ns
		N450	94.2	88.5	85.5	95.0	90.8	91.2	0.7ns	− 0.4ns	< 0.1ns
		s.e	1.39	1.39	1.39	1.39	0.78	0.82	1.25	0.40	0.40

***P ≤ 0.001, ** P ≤ 0.01, * P ≤ 0.05, ^† P ≤ 0.1, ns: not significant.

^‡Lp: L. perenne, Dg: D. glomerata, Tp: T. pratense, Tr: T. repens.

^§The diversity effect is calculated as the difference between the performance of the equi-proportional four-species mixture and the average of monocultures and is split into the effects of mixing grass and legume species (D_BGL effect), mixing the two grasses Lp and Dg (Lp∙Dg effect), and mixing the two legumes Tp and Tr (Tp∙Tr effect).

^¶Standard errors (s.e.) were calculated from the weighted average of variances at each N fertilisation level.

^#Stability, weed biomass, and NO₃ in soil solution were natural log transformed and subsequently scaled to range between 0 and 100%; thus, scaling with the maximal functional performance is not reasonably possible. See Fig. 1 for back-transformed values on the linear scale.

^‡‡Lp∙Dg and Tp∙Tr effects were omitted for NO₃ in soil solution due to singularity in the design matrix.