This study was conducted over the 2011 and 2012 growing seasons at the Loess Plateau Research Station of Lanzhou University (35°40′ N, 107°51′ E, elev. 1,298 m) located at Xifeng, Gansu Province, northwest China. The region is characterized as a semi-arid zone, with mean annual precipitation of 548 mm concentrated in July–September. Average annual pan evaporation is 1,504 mm, about three times higher than the precipitation. Average annual mean temperature is 8.3°C, and mean temperatures in the hottest (July) and coldest (January) months are 21.3°C and −5.3°C, respectively. Average annual solar radiation is 5,489 MJ m⁻². The mean length of the annual growing season is 255 d. The dominant soil type is sandy loam with an average field water-holding capacity of 0.22 kg·kg⁻¹ and a wilting point of 0.07 kg·kg⁻¹_. In this area, 0–10 cm soil organic carbon 9.6 g·kg⁻¹, available nitrogen 23.6 mg·kg⁻¹, available phosphorus 6.2 mg·kg⁻¹, total nitrogen 0.7 mg·kg⁻¹. The organic carbon of 10–20 cm soil was 10.2 mg·kg⁻¹, available nitrogen 31.5 mg·kg⁻¹ and available phosphorus 7.2 mg·kg⁻¹.

The seeding rate for monocultures of oat (cultivar No. 2 Qingyin) and common vetch (cultivar No. 3 Lanjian) was 165 (about 495 seeds m⁻²) and 105 kg ha⁻¹ (about 200 seeds m⁻²), respectively. The intercrops were sown in a replacement series using the following oat–common vetch seeding ratios (%): 80: 20, 67: 33, 50: 50, 33: 67 and 20: 80, corresponding to 396-40, 332-66, 248-100, 163-134 and 99-160 seeds m⁻², with row spacing of 30 cm. The seeds of both kinds were planted at the same time. The research methodology employed a randomized complete block approach consisting of four separate replications. The dimensions of the land were 4 meters by 6 meters. A 15-hp cultivator was used to prepare the seedbed down to a measurement of 20 cm in depth. Planting was carried out in a single fade using a 4-coulter plot drill and seeds were deposited at a depth of 4 cm. On 7 July 2011 and 12 July 2012, crops were seeded manually. Based on local fertilizer recommendations, nitrogen and phosphorus were applied prior to sowing at rates corresponding to 46 kg N and 100 kg P₂O₅ per hectare, respectively. Weeds were effectively controlled in all plots through human hoeing. The climatic conditions during the 2-year study period are shown in Figure 1.

Plants in a 0.75 m² sampling area within every plot were removed manually to ground level on October 25, 2011, and on October 21, 2012, when common vetch was at the pod-filling stages and oat was at the kernel milk stage. Samples were separated into component species and weighed. To assess the dry matter (DM) yield, subsamples (0.5 kg fresh weight) of each species from each plot were dried in a hot air draft oven at a constant temperature of 65°C until they reached a stable weight.

The sub-samples utilized for DM measurements were pulverized using a Wiley mill to achieve a particle size small enough to pass through a 1-mm screen. The Kjeldahl method was employed to ascertain the aggregate nitrogen content in intact plant samples (Black, 1965). The crude protein (CP) is determined by multiplying the amount of nitrogen by an amount of 6.25 (Jones, 1931). Crop protein yield (kg ha⁻¹) for this intercropping system was calculated as follows: Crop protein yield = N % oat × 6.25 × D M oat + N % vetch × 6.25 × D M vetch .

Multiple concurrence assessments were used to compute the possible advantages of intercropping and the influence of inter-species rivalry in the mixture. The LER was used as the criterion to assess the advantage of a mixed stand containing both oat and common vetch as desired species (Osiru and Willey, 1972). The LER was calculated as follows: L E R = L E R oat + L E R vetch L E R oat = Y O V Y O L E R vetch = Y V O Y V . where Y_O and Y_V are the yields of oat and common vetch in a pure stand, and Y_OV and Y_VO are the DM yields of oat and common vetch in any one mixture.

The relative crowding coefficient (K) quantifies the degree of dominance of one species over another in a mixture (Lithourgidis et al., 2011), and was calculated as follows: K = K oat × K vetch K oat = Y O V Z V O ( Y O − Y O V ) Z O V K vetch = Y V O Z O V ( Y V − Y V O ) Z V O

Z_OV and Z_VO represent the proportional proportions of oat and common vetch that are seeded in a mixture. When the multiplication of the two coefficients (K_oat and K_vetch) exceeds one, it results in a yield advantage. When the value of K is one, there is no advantage in terms of yield. There is a yield disadvantage when K is lower than one.

Aggressivity (A) was calculated as follows: A oat = Y O V Y O Z O V − Y V O Y V Z V O

If the value of A_oat is 0, it indicates that both crops have equal competitiveness. If the value of A_oat is positive, it indicates that oat is the dominating species. Conversely, if A_oat is negative, it suggests that oat is the species being dominated. Aggressivity of common vetch was determined in the same way. A vetch = Y V O Y V Z V O − Y O V Y O Z O V

Partial real yield loss (AYL_oat or AYL_vetch) represents the relative change in yield, either loss or gain, of each species when planted as intercrops compared to their yield in a monoculture. The AYL was calculated as follows (Banik et al., 2000): A Y L oat = Y O V ∕ Z O V Y O ∕ Z O − 1 A Y L vetch = Y V O ∕ Z V O Y V ∕ Z V − 1 A Y L = A Y L oat + A Y L vetch

The Monetary Advantage Index (MAI) offers insights into the economic benefits of the combining system. The MAI was calculated as follows: M A I = Y O V P oat + V V O P vetch × L E R − 1 L E R

A higher MAI score corresponds to a more profitable cropping scheme (Ghosh, 2004). The intercropping advantage (IA) was computed utilizing the subsequent equation (Banik et al., 2000). I A oat = A Y L oat × P oat I A vetch = A Y L vetch × P vetch I A = I A oat + I A vetch where P is the hay price, average hay procurement price per ton: common vetch = €55, oat = €43 (Lithourgidis et al., 2006).

An integrated analysis of variance (ANOVA) was conducted to independently examine the dry matter (DM), concentration, and yields of CP in both sole crops and intercrops, considering data from many years. Data were analyzed employing the mixed model feature in SPSS version 16.0 (SPSS Inc., Chicago, IL, United States) utilizing an autoregressive covariance structure. Year was a random effect, seeding ratio was a fixed effect. Bartlett’s test was used to examine the homogeneity of variances. The competition and economic indices were differentiated using the least significant difference (LSD) test with a significance level of 0.05.

The precipitation patterns observed throughout crop growth (July–October) in both 2011 and 2012 closely resembled the historical average but with some end-of-season variations (Figure 2). Since there were no significant effects between years, minor rainfall and temperature differences had no effect on DM or index measurements.

The analysis of variance for above ground dry matter (ABDM) data of oat and common vetch indicated notable variations exist among seed ratios. However, there were no significant effects on yield for either year or year x seed ratio (Table 1).

As expected, the ABDM yield was significantly higher for the oat monoculture than for the common vetch monoculture and the oat–common vetch intercrops (p < 0.05). As the ratio of common vetch increased in intercrops, the ABDM yield of oats decreased. For example, the oat ABDM was 6.51 t ha⁻¹ in the monoculture but decreased with higher proportions of common vetch in the mixture (2.15 and 2.24 t ha⁻¹ in the 33:67 and 20:80 oat–common vetch intercrops, respectively). Common vetch showed an ABDM advantage when sown in a mixture. Common vetch constituted 49 and 55% of the AGDM in the 33:67 and 20:80 oat–common vetch intercrops, respectively (Table 1).

We determined the CP concentration (%) and protein yield (kg ha⁻¹) for each of the monocrops and intercrops (Table 1). The lowest CP concentration was in the oat monoculture (8.12%), and the highest CP concentration was in the common vetch monoculture (14.88%). However, the common vetch monoculture had the lowest protein yield (496 kg ha⁻¹) because of its low DM yield (3.34 t ha⁻¹). In the intercrops, the CP percentage increased as the ratio of common vetch in the intercrops increased. The maximum protein output was in the 80:20 oat–common vetch intercrop (696 kg ha⁻¹). This was because of the high dry matter yield of the 80:20 oat–common vetch intercrop (6.39 t ha⁻¹).

The values for the competition indices K (Table 2) and A (Table 3) indicated that oat was the stronger competitor in the oat–common vetch intercrops. The values of A and K for oat differed significantly among seeding ratios. As evident from the formula for A, the values for A of oat and common vetch are numerically identical but differ in their sign. In this experiment, A_oat was positive (indicating dominance) in the 80:20, 67:33, and 50:50 oat–common vetch intercrops. The highest A_oat value was in the 20:80 oat–common vetch intercrop (Table 3). The values of K, which represent crowding, were notably higher for oat than for common vetch and did not show a smooth trend with respect to the range of seed ratios. However, the K_oat values were highest when oat was a minor or major mixture component, and lower when oat and common vetch were mixed in similar proportions (Table 2).

The values for AYL calculated for mixture components varied with seeding rate and were positive for oat and negative for common vetch, indicating a general yield loss for common vetch in intercrops. The LER_oat values exhibited a positive correlation with the proportion of oat in intercrops, while the LER_vetch values showed a positive correlation with the proportion of common vetch in intercrops. As year effects were not statistically significant, the average LER for the 2011 and 2012 growing seasons was used to calculate MAI. For the mixtures, both AYL and LER showed similar trends to that of K with respect to seeding ratios, with the lowest values for the 50:50 mixtures (Table 3).

The economic indices IA and MAI varied significantly with the seeding rate. Positive IA values were obtained for oat at seeding ratio 80:20 and 67:33 and the highest IA value was for the 80:20 oat–common vetch intercrop. The IA values for common vetch were all negative except that for the 20:80 mixture. This value was smaller than those obtained for oat so that the total IA followed the same pattern as that of IA_oat. The highest overall IA (0.254) was for the 80:20 oat–common vetch intercrop. Similar to the trends in yield (Table 1), the MAI showed positive values for the oat–common vetch intercrops with seeding ratios of 80:20, 67:33, and 20:80 (18.54, 5.90, and 35.51, respectively).

Intercropping is becoming more popular in sustainable agricultural methods, because of its potential to increase the efficiency of resource use. In previous studies on intercropping systems, it has been reported that oat dominated pea (Pisum sativum L.) (Neugschwandtner and Kaul, 2016), at 66:33 common vetch–barley (Hordeum vulgare L.) intercropping (Osman and Nersoyan, 1986), 55:45 common vetch–oat intercrop (Dhima et al., 2007) and 75:25 oat–faba bean (Vicia faba L.) intercropping (Dhima et al., 2014) had higher yield than their respective monocrops. These findings suggested that the maximum DM yields depend both on the species and on the seeding ratio. However, Another study found that the seeding ratios had no impact on the dry matter yields in combinations of pea and cereals, such as wheat or oat (Carr et al., 1998).

In this study, the contribution of common vetch to DM yield decreased with higher proportions of common vetch in the intercrops. This may have been because of competition between oat and common vetch in the intercrops, given that all the competition indices pointed out that oat was the prevailing species in the mixture. One possible explanation is that the tillering capacity of oat made it more competitive than common vetch (Lithourgidis et al., 2011). Another possible explanation is that oats grow faster initially and have a higher plant height than the intercropping common vetch. Putting the intercropped common vetch a disadvantage in terms of light (Feng et al., 2015).

The mechanisms that explain over-yielding intercropping are generally attributed to specific mutual complementarity (Dong et al., 2018) and beneficial interactions among species (facilitation) in resource use (Stomph et al., 2020). Cong et al. (2014) showed that facilitation of P, Fe, Mn and Zn acquisition is a potentially important cause of overyielding in annual intercropping systems. Cereal/legume intercropping improves phosphorus acquisition (Hinsinger et al., 2011).

The concentration of CP is a crucial nutritional key qualities of forage crops and is commonly utilised to assess forage systems, particularly intercropping systems (Yolcu et al., 2009). The CP concentration was shown to increase in intercrops due to the legume contribution (Bedoussac et al., 2014). In the Mediterranean environment, pure common vetch or higher proportions of common vetch intercropped with oat showed higher CP yields (Lithourgidis et al., 2006), because of the low proportion of protein in the ABDM of oat (Lithourgidis et al., 2011). However, Li et al. (2006) found no notable disparities in nitrogen intake between intercropping systems that combine legumes and cereals, and the traditional practice of growing these crops separately. The study found that the common vetch monocrop had the highest CP concentration at 14.88%. Another study found that CP yields were greatest with a 3:1 ratio of oats to common vetch, increasing 21.3 and 6.1% over oats alone and wild peas alone, respectively (Qu et al., 2022). However, in this semi-arid region under summer-sowing conditions, the highest CP yield (696 kg ha⁻¹) was in the oat–common vetch intercrop with an 80:20 seeding ratio. This was because of the high DM yield of this intercrop.

The LER is an indicator of the effectiveness of environmental resource use in intercropping compared to monocropping (Willey and Rao, 1980). Intercropping benefits species’ development and productivity when the LER is greater than one; Intercropping has a negative effect on the growth and yield of crops grown in mixtures when the LER is less than one (Reddy and Chetty, 1984; Ofori and Stern, 1987). Dhima et al. (2007) reported LER values ranging from 1.05 to 1.09 in a mix of common vetch with wheat, triticale (×Triticosecale Witt mark), barley, and oat at common vetch–cereal ratios of 55:45 and 65:35. LER for maize and cowpea (Vigna unguiculata L., Walp) intercropping exceeded 1 (1.91 and 1.53) and found that the intercrops were more stable than monocrops (Dimande et al., 2024). In another study, the land productivity of intercrops was 12–32% higher than those of monocrops (Chapagain and Riseman, 2014; Xu et al., 2021). In the present study, the oat–common vetch intercrop at a seeding ratio of 20:80 showed the highest LER value (1.17); This suggests that to attain the same yield as an intercrop, a solo cropping system would need to occupy an area that is 17% greater (Midya et al., 2005). Nassab et al. (2011) reported that a 67:33 mixture of maize and sunflower (Helianthus annuus L.) showed a higher LER value than those of other mixtures and monocrops. Wang et al. (2021) found that in oat and common vetch intercropping, LER was highest when oats were sown at a rate of 50%, which was 11–57% higher than other sowing rates. In the present study, the LER value for the 50:50 oat–common vetch intercrop was lower than one, suggesting strong competition between the two crops in this mixture.

In the intercrop mixtures, the K_oat values were higher than the K_vetch values, which indicates that oat was more competitively effective than vetch under these conditions. However, Dhima et al. (2007) report that K was higher for common vetch than for cereals in the mixture of vetch with wheat or triticale (65:35 common vetch–cereal). The K values exhibited a comparable pattern to the LER values in this investigation. In all mixtures, the K value was greater than one, indicating that there was a yield advantage of intercropping (Willey and Rao, 1980; Banik et al., 2000; Ghosh, 2004). The K value was above two in the 67:33 and 20:80 oat–common vetch mixtures, indicating a substantial yield increase from intercropping.

The A values indicated which species was dominant in the intercrops. Common vetch was the dominant species (as indicated by positive A_vetch values) only when it was the main component in the mixture (67% or 80%) (Table 3). Similar results have been reported previously for common vetch–wheat and common vetch–triticale mixtures (65:35) (Dhima et al., 2007). The study showed that the difference in aggression to competition ratios between oats and common wild pea was not significant under the mixed cropping system (Zhu et al., 2022). In the present study, oat was dominant in the oat–common vetch intercrops with seeding ratios of 80:20, 67:33, and 50:50. Cereals like maize, sorghum (Sorghum bicolor L) Moench, and pearl millet (Pennisetum americanum L) were also reported to be dominant in groundnut–cereal intercropping systems (Ghosh, 2004).

AYL offers more accurate data on both inter-and intraspecific rivalry among part crops and the actions of the various species in intercropping systems, in comparison to other competition indices (Banik et al., 2000). Positive AYL values signify a benefit, whereas negative AYL values signify a drawback in intercrops when the primary goal is to compare production on a per-plant basis (Zhang Q, et al., 2015). Dhima et al. (2007) reported negative AYL values for common vetch–triticale (65:35), common vetch–barley (65:35) and common vetch–oat (55:45) intercrops. Neugschwandtner and Kaul (2014) reported that 75:25, 50:50, and 25:75 oat–pea intercrops did not show yield advantages when grown in fertile soil. Takim (2012) studied maize–cowpea intercrops (67:33, 50:50, and 33:67) and reported negative AYL values for intercropped cowpea ranging from −0.257 to −0.813, indicating a yield loss of 25.7–81.3%, compared with that of a cowpea monocrop. In the present study, the AYL values for common vetch were negative and ranged from −0.044 to −0.138, indicating a yield loss of 4.4–13.8% in intercrops. Although the partial AYL of oat was positive except in the 50:50 oat–common vetch intercrop, the AYL was negative except in the 20:80 and 80:20 oat-common vetch intercrop, and this was insufficient to offset the decrease in crop productivity caused by the presence of common vetch in the mixture.

Dhima et al. (2007) reported a maximum MAI value of 13.47 for a 20:80 oat–common vetch mixture, while the maximum yield advantage was obtained for a 45:55 oat–common vetch mixture. In the present study, the most advantageous mixtures were the 20:80 oat–common vetch intercrop, followed by the 67:33 oat–common vetch intercrop (IA values of 0.254 and 0.99, respectively; Table 4). Lithourgidis et al. (2011) reported that a 20:80 cereal–pea intercrop provided the maximum economic profit, consistent with our results. In line with our findings, Ghosh (2004) observed that greater LER and K values were linked to notable economic advantages, as indicated by larger MAI values.

The outcomes of our study indicate that the practice of planting oat and common vetch together at various seeding ratios has an impact on the dry matter yield of each species, the degree of rivalry between the two species, and ultimately the economic viability of the intercropping system. By planting oats and common vetpea in different proportions, the method of increasing the yield of annual forage crops is realized to solve the problem of local feed shortage in winter and early spring. In future studies, we can explore the mixed sowing mechanism of annual forage crops and understand the principle of mixed sowing to increase yield, so as to select more mixed crops and clarify the breeding direction of future mixed crops.

On the Loess Plateau, oat–common vetch at a seeding ratio of 20:80 is the optimal intercrop in terms of balancing the nutritive value, competition between species, and economic returns. The results of this study illustrate how yield, competition, resource use, and economic indices can be used to compare different intercrops and identify which systems have the best overall value.