A low nitrogen fertiliser rate in oat–pea intercrops does not impair N₂ fixation

ABSTRACT

Intercropping is commonly used in low-input systems but could also be a strategy for higher input systems. Three ratios of substitutive oat–pea intercrops were tested on a fertile soil in eastern Austria with application of a low nitrogen (N) fertiliser rate (6 g N m⁻²) versus an unfertilised control to assess the effect of intercropping and low N fertilisation on dinitrogen fixation (N_FIX). The above-ground dry matter (AGDM) and N yield of intercrops increased with N fertilisation, but the increase occurred only in oat, while pea was not affected by N fertilisation. Pure pea stands and intercrops with high pea share resulted in N sparing in the soil at harvest, as the soil mineral N was higher than in pure oat. Half of the applied amount of N was recovered by the AGDM of crops and half remained in the soil at harvest. The N_FIX per unit area was highest in pure pea. Intercropping considerably reduced N_FIX, especially in intercrops with low pea share. N_FIX per unit of AGDM of pea, however, was neither affected by intercropping nor by N fertilisation. Consequently, a low amount of N fertilisation of oat–pea intercrops on a fertile soil can increase overall performance of the system through increasing the performance of oat without impairing that of pea.

KEYWORDS:

• Avena sativa
• Pisum sativum
• dinitrogen fixation
• nitrogen yield
• nitrogen sparing

Introduction

Atmospheric dinitrogen (N₂) fixation (N_FIX) plays a critical role in crop production. Almost 20% (17 Tg N) of the worldwide needed N for grain crop production, which is about 100 Tg N, is supplied by N_FIX (Herridge and Rose 2000). In addition to the soil improvement through N_FIX, grain legumes are important break crops in cereal-based crop rotations for interrupting disease and pest cycles (Robson et al. 2002).

Grain legumes occupy only about 3% of the arable land in the European Union, of which about 30% are consumed as animal feed. This results in the need to import large amounts of protein feedstuffs, especially soy products (Häusling 2011). Enhancing grain legume production is therefore important for reducing the substantial deficit of protein sources in Europe (Henseler et al. 2013). Several strategies are possible here, among those are shifting from spring-sowing to autumn-sowing of grain legumes (Neugschwandtner et al. 2019), introduction of new grain legumes with better drought adaptation like chickpea (Neugschwandtner et al. 2013; Neugschwandtner, Wagentristl et al. 2015) and using grain legumes in intercropping systems with cereals (Neugschwandtner and Kaul 2014; 2015). Other strategies for enhancing legume productivity are Rhizobium inoculation and foliar fertilisation (Klimek-Kopyra et al. 2020).

Advantages of cereal-legume intercrops can include: a more efficient or complementary use of the limiting growth resources, a reduction of pest incidence, higher protein yields, an insurance against crop failure especially in areas with harsh environmental conditions and an improvement of soil fertility through N_FIX (Lithourgidis et al. 2011). In buckwheat-fenugreek intercrops, higher crop growth rates and a higher nutrient uptake compared to the pure crop stand has been observed (Salehi, Fallah, et al. 2018; Salehi, Mehdi, et al. 2018).

N_FIX in cereal-legume intercrops is the major source of nitrogen (N) under limiting N conditions (Fujita et al. 1992) and growing legumes in intercrops is an alternative and sustainable way of introducing N into low input agroecosystems (Fustec et al. 2010). N_FIX can be increased in intercrops as shown for grain legumes intercropped with barley, where the percentage of plant N derived from the atmosphere (%Ndfa) was 10–15% higher compared to the corresponding sole legume crops (Hauggaard-Nielsen et al. 2008). Legumes in intercrops rely much more strongly on N_FIX because an accompanying cereal partner is more competitive for soil inorganic N (Corre-Hellou et al. 2006).

Intercrops are therefore still extensively grown in traditional, labour-intensive, small-scale agricultural systems of developing countries (Anil et al. 1998). But there is also an increasing interest in intercropping in highly mechanised agriculture systems in temperate regions (Zając et al. 2014). In modern agricultural systems, they fit best for organic farming but might also be suitable in conventional cropping systems (Machado 2009). Anil et al. (1998) stated that the advantages of intercropping systems will be smaller in temperate regions than in tropical ones. But Bybee-Finley and Ryan (2019) also highlighted that intercropping can contribute to sustainable intensification of industrialised agricultural systems where it can play an important role in increasing productivity, stability and ecosystem services.

Average long-term grain yields of peas in Austria are under conventional or organic production at 235 or 133 g m⁻² (Brückler et al. 2017). The protein content of pea grains, which was reported in a comparison with other grain legumes with 25.7%, is in a similar range with most grain legumes, but lower than that of lupines and soybean (cf. Vollmann 2016). Under temperate conditions in eastern Austria, grain N yields in pea pure crop stands ranging from 8 to 20.5 g m⁻² have been reported, which were considerably impaired in oat–pea intercrops (Neugschwandtner, Wagentristl et al. 2015; Neugschwandtner and Kaul 2015).

In temperate regions, mineral N is needed for achieving high yields (Moitzi et al. 2020). Pure grown pea takes up supplied mineral N efficiently, but low rates of fertiliser N do not decrease N_FIX especially when applied later during development, while high N rates impair N_FIX (Jensen 1986). Diverse results have been reported on the effects of N fertilisation on legumes in intercrops. Yield and yield components of oat–pea intercrops are affected by N fertilisation, and oat benefited more than pea (Neugschwandtner and Kaul 2014). Mineral N use in intercrops is reported to be more efficient than in monocropped grain legumes in temperate regions (Hauggaard-Nielsen et al. 2001) and the productivity of the intercropping system can be improved when the inhibitory effect of N fertilisation on nodulation and N_FIX is alleviated as shown for intercrops of maize with faba bean (Li et al. 2009).

The aim of this study was to assess different substitutive oat–pea intercrops as compared to pure stands of both crops grown on a fertile soil in temperate conditions of eastern Austria as affected by N fertilisation with focus on (a) biomass and N yields, (b) N_FIX and (c) residual N after harvest.

Material and methods

Environmental conditions

The experiment was carried out on a chernozem soil of alluvial origin (pH_CaCl2: 7.6, silty loam, soil organic carbon: 2.2–2.3%) in Raasdorf (48° 14ʹ N, 16° 33ʹ E), east of Vienna, Austria, on the edge of the Marchfeld plain, in 2010 and 2011. The soil mineral N in 0–0.9 m depth at sowing was 15.8 (2010) or 16.8 (2011) g N m⁻².

Precipitation during the vegetation period 2010 was highly above average from April until July, whereas in 2011, it was comparatively dry. In 2011, the temperature was also generally higher than in 2010. Details on the experimental site, the soil including macro- and micronutrients status, as well as climatic conditions in the experimental years are given in Neugschwandtner and Kaul (2016a, 2016b).

Experimental treatments and measurements

Pure stands of oat (cv. Effektiv) and pea (cv. Lessna) were established with 350 (oat) and 80 (pea) germinable seeds m⁻², respectively. Three substitutive intercrops were sown in March in replacement series consisting of following crop stands (oat:pea, %:%): 75:25, 50:50 and 25:75. Fertilisation consisted of an unfertilised control and a fertilisation treatment with 6 g N m⁻² applied as calcium ammonium nitrate (CAN, 27% N) in two equal splits, the first right after sowing and the second after tillering of oat. Harvest was performed in July.

Details on plot size, pre-crops, seedbed preparation, sowing and harvest (including dates), fertilising dates, plant protection, plant sampling and N determination in the above-ground dry matter (AGDM) are given in Neugschwandtner and Kaul (2014, 2015).

Assessing soil mineral N and N_FIX

Soil sampling after harvest was performed with soil probes (Puerckhauer type, core diameter: 20 mm) to a depth of 0.9 m. Samples were further divided into three layers: 0–0.3, 0.3–0.6, 0.6–0.9 m. Soil samples were extracted with 0.0125 M CaCl₂ in a soil extraction ratio of 1:4 (w/v) for 1 h using an overhead shaker and soil mineral N (SMN, as nitrate–nitrogen: NO₃–N) was determined photometrically (FIASTAR 5000, FOSS GmbH, Hamburg Germany) (ÖNORM L 1091, 2012). The gravimetric soil water content was obtained by drying the soil samples at 105°C for 24 h.

The N_FIX per unit area at harvest was estimated by the extended difference method according to Stülpnagel (1982) for pure pea (Equation 1) and according Karpenstein-Machan and Stülpnagel (2000) for pea in intercrops (Equation 2) by comparing differences in N yield in the AGDM and differences in the depletion of SMN (0–0.9 m) between crop stands including pea (pure pea crop stands and all intercrops) and non-legume crop stands (pure oat crop stands): (1) ${\text{N}}_{\mathrm{F}\mathrm{I}\mathrm{X}}\left(\text{g N }{\text{m}}^{-2}\right)=(\mathrm{N}{\text{Y}}_{\mathrm{P}\mathrm{P}}-\text{N}{\text{Y}}_{\mathrm{O}})+(\text{SM}{\text{N}}_{\mathrm{P}\mathrm{P}}-\text{SM}{\text{N}}_{\mathrm{O}})$(1) (2) ${\text{N}}_{\mathrm{F}\mathrm{I}\mathrm{X}}\left(\text{g N }{\text{m}}^{-2}\right)=(\mathrm{N}{\text{Y}}_{\mathrm{I}\mathrm{C}}-\text{N}{\text{Y}}_{\mathrm{O}})+(\text{SM}{\text{N}}_{\mathrm{I}\mathrm{C}}-\text{SM}{\text{N}}_{\mathrm{O}})$(2) where NY_PP, NY_IC and NY_O are the quantities of N yield in AGDM and SMN_PP, SMN_IC and SMN_O are the amounts of SMN of crop stands at harvest of pure pea (PP), intercrops (IC) and pure oat (O). The lower removal of mineral nitrogen from soil cultivated with a legume compared to those with a cereal is called ‘nitrogen sparing’ (cf. Chalk et al. 1993).

N_FIX per unit of AGDM of pea was calculated according to equation 3: (3) ${\text{N}}_{\mathrm{F}\mathrm{I}\mathrm{X}}\left(\text{g }{\text{g}}^{-1}\text{pea}\right)={\text{N}}_{\mathrm{F}\mathrm{I}\mathrm{X}}\left(\text{g}\text{N}{\text{m}}^{-2}\right)/\text{AGDM of pea}\left(\text{g}{\text{m}}^{-2}\right)$(3)

Statistics

The experiments were established in a randomised complete block design with three replications. Statistical analyses were performed using SAS version 9.2 applying analysis of variance (PROC ANOVA) with subsequent multiple comparisons of means, which were separated by least significant differences when the F-test indicated a significance level of p < 0.05.

Results

Biomass yields, nitrogen concentrations and nitrogen yields

The mean value for biomass yields, N concentrations and N yields over all crop stands, fertilisation levels and years were as follows: AGDM of oat: 990 g m⁻², AGDM of pea: 411 g m⁻², AGDM of oat + pea: 1121 g m⁻², N % of the AGDM of oat: 1.21%, N % of the AGDM of pea: 2.46%, N yield of oat: 11.94 g m⁻², N yield of pea: 10.20 g m⁻², total N yield (i.e. the N yield of oat + pea): 17.71 g m⁻², Δ NY_OAT+PEA – NY_OAT: 5.49 g m⁻² (Table 1).

Table 1. Above-ground dry matter (AGDM), nitrogen (N) concentration, N yield and the difference of the N yield in the crop stands with pea compared to pure oat crop stands.

	AGDM			N concentration		N yield			Δ NYOAT+PEA – NYOAT^A
	oat	pea	oat + pea	oat	pea	oat	pea	oat + pea
	g m^−2			%		g m^−2			g m^−2
Crop stand (oat:pea, %:%)
100:0	1158^a		1158	1.15^c		13.32^a		13.32^d
75:25	1096^a	64^d	1160	1.17^bc	2.39	12.93^ab	1.56^d	14.49^cd	1.17^c
50:50	952^b	190^c	1142	1.25^ab	2.46	11.87^b	4.65^c	16.52^bc	3.20^bc
25:75	756^c	339^b	1095	1.27^a	2.50	9.63^c	8.36^b	17.99^b	4.67^b
0:100		1052^a	1052		2.48		26.23^a	26.23^a	12.91^a
Fertilisation (g N m^−2)
0	903^b	404	1045^b	1.14^b	2.46	10.21^b	10.04	16.20^b	6.50^a
6	1078^a	419	1197^a	1.28^a	2.46	13.67^a	10.36	19.22^a	4.50^b
Year
2010	1026	419	1156^a	1.22	2.67^a	12.47^a	11.30^a	19.01^a	6.49^a
2011	955	404	1086^b	1.20	2.34^b	11.40^b	9.10^b	16.40^b	4.48^b
ANOVA
Crop stand (C)	***	***		*		***	***	***	***
Fertilisation (F)	***		***	***		***		***	*
Year (Y)			*		***	*	*	***	*
C × Y							***	***	**
F × Y				*

Notes: Different letters indicate significant differences between means. Significant effects at p < 0.05 (*), p < 0.01 (**) and p < 0.001 (***).

^A NY_OAT+PEA = N yield of pure pea crop stands and intercrops; NY_OAT = N yield of pure oat crop stands. No significant interactions were found for C × F and C × F × Y.

AGDM was similar in the pure crop stands of both crops, but in intercrops, oat was the dominant partner outcompeting pea. This resulted in a low decrease for oat but a high decrease for pea of the AGDM with a decreasing share in the intercrops. For example, in 50:50 intercrops, the AGDM of oat was 82.2% and that of pea 18.1% compared to the corresponding pure crop stands. The total AGDM (oat + pea) did not differ between crop stands. N fertilisation increased the AGDM and total AGDM of oat but not that of pea. Total AGDM was higher in 2010 than in 2011.

N concentration in AGDM of oat in the intercrops was higher compared to the pure oat crop stand with 50% oat at 9.4% and with 25% at 10.6%. It increased with N fertilisation in both years but significantly only in 2011 (in % N; 0 vs. 6 g N m⁻² – 2010: 1.19 vs. 1.26; 2011: 1.10 vs. 1.30; LSD = 0.09) (Table 3). N concentration in AGDM of pea was not affected by either crop stand or fertilisation, though it was higher in 2010 than in 2011, while that of oat did not differ between years.

The N yield of oat slightly decreased and that of pea sharply decreased with a declining share of the crops on the intercrops. For example, in 50:50 intercrops, the N yield of oat was 89.1% and that of pea 17.7% compared to the corresponding pure crop stands. Total N yield was highest in the pure pea crop stands and decreased with lower pea shares in the intercrops. It did not differ between pure oat crop stands and 75:25 intercrops. The total N yield of pea was higher in 2010 than in 2011 with no differences between years for other crop stands (Table 3). The N yield of oat was higher than that of pea in all intercrops, e.g. in 50:50 intercrops, the total N yield consisted of 71.8% oat and 28.2% pea. N fertilisation increased the N yield of oat but not that of pea. Total N yield with N fertilisation was 3.02 g m⁻² higher than in the control (mean over all crop stands and years).

The mean calculated difference (Δ) (over fertiliser levels and years) of N yield in crops stands with pea (pure pea crop stands and all intercrops; NY_OAT+PEA) and pure oat crop stands (NY_OAT) was highest in pure pea crops. It decreased with a decreasing share of pea in the intercrops as follows: 75%, 50% or 25% of pea resulted in 36.2%, 24.8% or 9.1% of the pure pea crops stands, respectively. The Δ NY_OAT+PEA – NY_OAT decreased with N fertilisation by 31.1%. It was significantly higher for the pure pea crop stand in 2010 than in 2011, whereas no differences occurred between the intercrops between the years (Table 3).

Gravimetric water content

The H₂O content at harvest was higher in 2010 than in 2011 in the different soil layers across all crop stands and fertilisation levels; 0–0.3 m: 18.5% vs. 16.6%, 0.3–0.6 m: 17.6% vs. 12.6%, 0.6–0.9 m: 12.8% vs. 7.3%. In 0.3–0.6 m, there was a significant sowing ratio × year interaction with a higher H₂O content in pea than in other treatments in 2011 but no differences between treatments in 2010 (data not shown). In 0–0.3 m and in 0.6–0.9 m, there were no differences between sowing ratios. N fertilisation did not affect the H₂O content in any of the three layers.

Soil mineral nitrogen and dinitrogen fixation

Mean values for SMN, N sparing and N_FIX across all crop stands, fertilisation levels and years were as follows: SMN (0–30 cm): 2.59 g m⁻², SMN (30–60 cm): 1.41 g m⁻², SMN (60–90 cm): 0.74 g m⁻², SMN (0–90 cm): 4.74 g m⁻², N sparing (0–30 cm): 0.88 g m⁻², N sparing (30–60 cm): 0.83 g m⁻², N sparing (60–90 cm): 0.63 g m⁻², N sparing (0–90 cm): 2.34 g m⁻², N_FIX: 7.83 g m⁻² and 22.9 mg g^–1 pea (Table 2).

Table 2. Soil mineral nitrogen, nitrogen (N) sparing and N fixation (N_FIX) on the area and plant dry matter level (pea).

	Soil mineral nitrogen				N sparing				NFIX	NFIX
m	0–0.3	0.3–0.6	0.6–0.9	0–0.9	0–0.3	0.3–0.6	0.6–0.9	0–0.9
	g m^−2				g m^−2				g m^−2	mg g^−1 pea
Crop stand (oat:pea, %:%)
100:0	1.88^b	0.75^b	0.24^c	2.88^c
75:25	1.82^b	0.70^b	0.49^bc	3.00^c	−0.06^b	−0.06^b	0.25^b	0.13^b	1.30^c	26.8
50:50	2.12^b	0.90^b	0.44^bc	3.46^bc	0.23^b	0.15^b	0.20^b	0.58^b	3.78^bc	20.3
25:75	3.36^ab	1.62^b	0.94^b	5.92^ab	1.48^a	0.87^b	0.70^ab	3.05^ab	7.72^b	28.1
0:100	3.76^a	3.11^a	1.61^a	8.48^a	1.88^a	2.36^a	1.37^a	5.60^a	18.51^a	16.7
Fertilisation (g N m^−2)
0	1.95^b	0.87^b	0.53^b	3.35^b	−0.18^b	0.61^b	0.46	0.89^b	7.39	18.0
6	3.23^a	1.96^a	0.96^a	6.15^a	1.94^a	1.05^a	0.80	3.79^a	8.26	27.9
Year
2010	2.10	1.30	0.86	4.27	1.30	1.24	1.08^a	3.62^a	10.11^a	25.0
2011	3.07	1.53	0.62	5.23	0.46	0.42	0.18^b	1.06^b	5.54^b	20.9
ANOVA
Crop stand (C)	*	**	**	**	*	**	**	*	***
Fertilisation (F)	*	*	*	**	**	*		*
Year (Y)							**	*	**
C × Y			**				*		**

Notes: Different letters indicate significant differences between means. Significant effects at p < 0.05 (*), p < 0.01 (**) and p < 0.001 (***). No significant interactions were found for C × F, F × Y and C × F × Y.

SMN at harvest was higher in 0–0.3 m in pure pea than in the pure oat crop stands as well as in 75:25 and 50:50 intercrops, with 25:75 intercrops showing intermediate values. In 0.3–0.6 m, SMN was higher in pure pea crops compared to all other crop stands. In 0.6–0.9 m, SMN was higher in 2010 in pure pea than in pure oat and in 75:25 and 50:50 intercrops, with 25:75 intercrops showing intermediate values, whereas in 2011 there were no differences between crop stands (Table 3).

Table 3. Interactions of crop stand × year.

Crop stands	N yield				Δ NYOAT+PEA – NYOAT^B		SMN^C		N sparing		NFIX
(oat:pea, %:%)	pea		oat+pea				0.6–0.9 m		0.6–0.9 m
	g m^−2		g m^−2		g m^−2		g m^−2		g m^−2		g m^−2
	2010	2011	2010	2011	2010	2011	2010	2011	2010	2011	2010	2011
100:0			13.83	12.82			0.00	0.48
75:25	2.08	1.04	14.84	14.14	1.02	1.32	0.27	0.70	0.27	0.22	0.75	1.85
50:50	4.30	5.01	16.36	16.68	2.54	3.86	0.36	0.52	0.36	0.04	3.59	3.97
25:75	8.02	8.70	19.25	16.73	5.42	3.91	1.15	0.73	1.15	0.25	10.70	4.73
0:100	30.81	21.64	30.81	21.64	16.99	8.82	2.54	0.68	2.54	0.20	25.40	11.61
LSD^A	3.55		2.94		3.81		0.91		1.01		6.29

^A LSD = least significant difference at p < 0.05.

^B NY_OAT+PEA = N yield of pure pea crop stands and intercrops; NY_OAT = pure oat crop stands.

^C SMN = soil mineral nitrogen.

Consequently, in the whole soil profile (0–0.9 m) SMN was ranked as follows: pure pea ≥ 25:75 ≥ 50:50 ≥ 75:25, pure oat. SMN in pure pea was 2.94-fold higher than in pure oat. SMN in intercrops was higher compared to pure oat: 1.04-fold for 75:25, 1.20-fold for 50:50 and 2.06-fold for 25:75. In the individual soil layers, SMN of pure pea was by 1.78-fold (0–0.3 m), 4.47-fold (0.3–0.6 m), 6.71-fold (0.6–0.9 m) higher compared to pure oat (mean over fertilisation levels and years).

The SMN was higher in all three layers with N fertilisation by 1.66-fold (0–0.3 m), 2.25-fold (0.3–0.6 m), 1.83-fold (0.6–0.9 m), and in the whole soil profile (0–0.9 m) by 1.84-fold compared to the control (mean over all crop stands and years). SMN in the whole soil profile (0–0.9 m) was higher by 2.80 g m⁻² with N fertilisation than in the control (mean over all crop stands and years), which is roughly half of the applied fertiliser N.

Nitrogen sparing was higher in pure pea and in 25:75 intercrops than in 50:50 and 75:25 intercrops in 0–0.3 m. In 0.3–0.6 m, it was higher in pure pea than in all intercrops. In 0.6–0.9 m, it was higher for pure pea than for intercrops in 2010, whereas no differences were observed between crop stands in 2011 (Table 3). In the complete soil profile, pure pea had a higher N sparing than 50:50 and 75:25 intercrops, with 25:75 intercrops showing intermediate values. With N fertilisation, values of N sparing showed a pattern similar to SMN with higher values than the control for the layers of 0–0.3 m and 0.3–0.6 m and the whole soil profile, but not for 0.6–0.9 m.

N_FIX on the area level was ranked among the crop stands as follows: 2010 – pure pea > 25:75 > 50:50, 75:25; 2011 – pure pea > all intercrops. N_FIX of pure pea was higher in 2010 than in 2011, with no differences between years among the different intercrops (Table 3). Over fertilisation levels and years, mean N fixation of intercrops compared to pure pea attained just 41.7% (25:75), 20.4% (50:50) and 7.0% (75:25). N_FIX per area did not differ between fertilisation levels, and N_FIX per unit of AGDM of pea did not differ between crops stands, fertilisation levels and years.

Discussion

Oat did outcompete pea in the intercrops. It achieved both a higher AGDM yield (compared to its share on the intercrops) and a higher N concentration in the AGDM with a lower sowing ratio. Both AGDM yield and N concentration in the AGDM increased with N fertilisation. Contrary to that, these parameters of pea were not affected by fertilisation. Consequently, the proportion of pea in intercrops decreased with N fertilisation. The same observation has been reported for pea in intercrops with barley, oat or wheat (Jensen 1996; Ghaley et al. 2005; Noworolnik and Dworakowski 2010), but in all those studies, also the grain yield of pea decreased with N fertilisation. Whereas in a pot experiment, where competition for space might have not been as limiting as under field conditions, Lošák (2007) observed for narrow-leaf lupin an increase of pod plant⁻¹, thousand kernel weight and single plant grain yield with N fertilisation.

Higher yields of oat in intercrops and with fertilisation might be due to the low uptake of SMN of legumes at early growth, which can reduce their competitiveness with other species, e.g. in intercrops (Dayoub et al. 2017). Cereals at early growth have a competitive advantage especially with N fertilisation due to a faster initial growth, as shown for barley-pea intercrops (Andersen et al. 2005). In barley-pea intercrops, also a faster root growth of barley than of pea has been shown to enhance the nitrogen uptake of barley (Hauggaard-Nielsen et al. 2001).

Similar to the higher AGDM with N fertilisation, also a higher N yield of N fertilised intercrops was achieved solely by the increase of the N yield of oat. Ghaley et al. (2005) have also reported for wheat-pea intercrops an increase of the grain N yield of wheat with increasing N supply, whereby up to 90% of the total intercrop fertiliser N could be acquired by wheat; additionally, they reported a decrease of grain N yield of pea with increasing N supply, which we cannot confirm.

Intercrops of durum wheat and winter pea were less efficient in the production of yield and N yield than their respective sole crops at high N fertiliser rates (Bedoussac and Justes 2010b), as with more available N the growth complementarity over time of wheat and pea was poorer and the better early growth of wheat outcompeted pea (Bedoussac and Justes 2010a). In our experiment, the low N rate did not reduce the competitiveness of intercrops compared to sole crops.

Over every crop stand and all years, the total N yield was higher with N fertilisation by a mean of 3.02 g m⁻² than in the control. Thereby, 50% of applied N could be taken up in the total AGDM.

Advantages of intercropping may arise from complementary utilisation of growth resources by using crops with differences in canopy structure, nutrient requirements and rooting ability (Lithourgidis et al. 2011). As the H₂O content at harvest just differed in one soil layer (0.3–0.6 m) in one year, soil water depletion did not much differ between crop stands. N fertilisation also did not reduce H₂O content at harvest, although it caused a higher AGDM yield, which suggests a better water use efficiency of fertilised crops (Wang et al. 2018).

The SMN in the whole soil profile at harvest was considerably lower in pure oat and intercrops with medium and high oat share than in pure pea. On the same location, an about three times higher SMN at harvest was observed after faba bean than after wheat and the highest total differences of SMN between the grain legume and the cereal occurred as in this experiment in the soil layer of 0.3–0.6 m (Neugschwandtner, Ziegler et al. 2015). SMN was only higher in the intercrops with a high pea share than in pure oat. Consequently, N sparing compared to pure oat was low in intercrops with medium and high oat share, while N sparing was as high in the upper soil layer as in pure pea in the intercrops with high pea share. Fertilisation resulted in a higher SMN in all soil layers. Partly similar to these observations, Neumann et al. (2007) observed no differences in the SMN of oat–pea intercrops and pure oat, especially in deeper soil layers, from which the intercropped oat took up more SMN than pea.

Over all crop stands and years, the SMN in the whole soil profile (0–0.9 m) was higher with N fertilisation than in the control by a mean of 2.80 g m⁻². Thereby, 47% of applied N remained in the soil after harvest.

Higher SMN with grain legumes compared to non-legumes after harvest persists up to winter (Kaul 2004). This (and more N in crop residues) can cause positive yield effects as shown for two succeeding cereals (Senaratne and Hardarson 1988) but can also cause higher N leaching (Plaza-Bonilla et al. 2015). Lower residual SMN in intercrops, as also shown in maize-faba bean intercrops compared to pure faba bean (Stoltz and Nadeau 2014) and oat–pea intercrops compared to pure pea (Neumann et al. 2007), may thereby reduce the risk of N leaching.

N_FIX per unit area was considerably lower in all intercrops than in pure pea and especially in intercrops with low pea share due to both the lower NY of pea and the lower N sparing in intercrops compared to pure pea. A lower N_FIX per unit area of intercrops than for pure legume crop stands often occurs due to decreased legume population densities and increased competition for light and nutrients by the non-legume (Van Kessel and Hartley 2000). In intercrops with low and medium pea share, the calculated N_FIX mainly arose from higher N_FIX in AGDM (Δ NY_OAT+PEA – NY_OAT); whereas, in the intercrops with highest pea share, also N sparing had a high contribution to N_FIX. In a comparison of winter and spring grain legumes, the considerably higher N_FIX of winter grain legumes arose solely from the higher N_FIX in AGDM, whereas N sparing did not differ between winter and spring grain legumes (Neugschwandtner, Ziegler et al. 2015).

Contrasting to the N fixation per area, the N_FIX per unit of AGDM of pea did not differ between crop stands, with a mean of 22.9 mg N g⁻¹ pea. Peoples et al. (2009) reported N_FIX values per dry matter of legumes on the whole plant basis, i.e. in shoots and nodulated roots, of 30–40 mg N g⁻¹. Many studies report that %Ndfa is higher in intercrops than in pure legumes (e.g. Bedoussac and Justes 2011), as N uptake by the non-legume forces the legume to rely more on N_FIX (Hauggaard-Nielsen et al. 2009). This complementary N use especially occurs in low N input systems (Bedoussac and Justes 2011). As our study was conducted on a fertile soil, this might be the reason why intercrops failed to achieve a higher grain yield than pure stands (cf. Neugschwandtner and Kaul 2014) and also complementary N use and N_FIX might not have been as important for the pea as it could have been on a poor N soil.

With N fertilisation, the N_FIX in the AGDM decreased but N sparing increased, thus there was no difference in the calculated N_FIX values between fertilisation treatments. In a meta-analysis of grain legume-cereal intercrops, even small doses of N (0.1–5 g N m⁻²) have been reported to affect N_FIX (Rodriguez et al. 2020). For example, in wheat-faba bean intercrops, %Ndfa was higher without N fertilisation than in pure faba bean but did not differ between crop stands with N fertilisation (Fan et al. 2006). In contrast, Cowell et al. (1989) reported that N fertilisation affected %Ndfa of pure but not of intercropped legumes.

Similar values of N_FIX per unit of AGDM of pea in our experiments indicate, that %Ndfa did not differ between crop stands. As intercropping is more suited to low N input systems than to highly fertilised systems (Bedoussac and Justes 2010b), its benefits might not have been fully exploited on a fertile soil. Naudin et al. (2010) reported for wheat-pea intercrops, that N_FIX is mainly affected by crop growth rather than %Ndfa and also Fan et al. (2006) stated that N_FIX is more closely correlated with yield than with %Ndfa. Thus, also intercropping might be a good strategy for fertile soils, when the focus is laid on yield maximisation, whereby N_FIX can be maximised (Fan et al. 2006). The optimal strategy for N fertilisation of intercrops should therefore attempt not to affect the performance of the legumes too strongly (cf. Naudin et al. 2010; Bedoussac and Justes 2010b).

A low nitrogen fertiliser rate of oat–pea intercrops increased AGDM and N yield of intercrops through increasing both parameters of oat, whereas pea was not affected by N fertilisation. One half of the applied N could be recovered by the AGDM of crops and the other half remained in the soil at harvest. N_FIX per unit of AGDM of pea was not affected by intercropping or by N fertilisation. Consequently, a low amount of N fertilisation of oat–pea intercrops on a fertile soil can increase the overall performance of the intercropping system due to the positive effect of oat whereas productivity of pea is not affected.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

Reinhard W. Neugschwandtner

Reinhard W. Neugschwandtner is an Assistant Professor at the University of Natural Resources and Life Sciences, Vienna. His research interests include grain legumes, intercropping, underutilized crops, soil tillage and long-term field experiments.

Hans-Peter Kaul

Hans-Peter Kaul is a Professor at the University of Natural Resources and Life Sciences, Vienna. His research interests include agronomy, underutilized crops and crop modelling.

Gerhard Moitzi

Gerhard Moitzi is a Senior Scientist at the University of Natural Resources and Life Sciences, Vienna. His research interests include agricultural engineering, soil tillage and energy efficiency of farming systems.

Agnieszka Klimek-Kopyra

Agnieszka Klimek-Kopyra is an Assistant Professor at the University of Agriculture in Krakow. Her research interests include sustainable crop production, intercropping, legumes and nitrogen fixation.

Tomáš Lošák

Tomáš Lošák is a Professor at the Mendel University in Brno. His research interests include plant nutrition and fertilization.

Helmut Wagentristl

Helmut Wagentristl is an Assistant Professor and head of the Experimental Farm Groß-Enzersdorf of the University of Natural Resources and Life Sciences, Vienna. His research interests include agronomy, soil tillage and long-term field experiments.