1. Introduction
Since the 1960s, the intensification of agriculture has led to the spread of such cropping systems, which are based on the cultivation of a couple of main crops in a simplified crop rotation, with the extensive use of agrochemical inputs [
1]. Crop specialization resulted in large productivity increase, and this mainstream was aimed at maximizing short-term yields and profits in developed countries [
2]. As a result, the cultivation of leguminous plants declined, and their role in facilitating soil fertility has become unused [
3,
4]. Due to the negative effects of climate change, maintaining the level of production requires greater investment and adaptability on the part of cultivated plants. The simplified sowing structure made the farmers vulnerable to the fluctuating market demands and increased the chance of developing an unfavorable pre-crop sequence [
5]. While a traditional cultivation system is based on the rotation of plants of a single genotype, biodiversity is a powerful management tool [
6]. Mixed cropping is a special kind of plant association, a spatially planned diversity, which is a simultaneous practical application of several ecological principles, such as diversity, competition, and facilitation. It usually means growing two or more plants at the same time at the same area [
7]. One of the greatest advantages of combining cereal and legume is breaking the overwhelming dominance of cereals by increasing the number of cultivated plants as a multiple cropping system. Due to the nitrogen-fixing ability of legumes, a natural nitrogen source is used, which is adapted for the needs of the companion plant, and this positive effect already prevails in the year of sowing.
Many authors agree that the most important advantage of mixed cropping is the more efficient use of growth resources than in pure stands [
8,
9,
10]. However, the effectiveness of each mixture may differ significantly and, in most cases, due to the complexity of the coordination, previous studies only provide rough guidelines [
11]. Most research works pay little attention to plant density and seeding rates of the species within the mixture and focus only on the final yield. In addition to the changing growing conditions, the choice of the appropriate seeding rate is still a question. According to Hauggaard-Nielsen and Jensen [
12], the composition of a plant association is primarily determined by the purpose of the cultivation and growing conditions. Due to the lower competitiveness of the leguminous companion plant, it may be necessary to use a higher seeding rate, which may even exceed the rate for pure stands [
8,
11]. In the case of mixture of two species, whether competition occurs depends on the plant density. Based on other findings of Hauggaard-Nielsen et al. [
13], the increased plant density in barley and pea association had a positive effect only on pea and severely limited the growth of barley. This is confirmed by Willey [
14], who believes that higher plant density increases the competition between the species and favors the dominant companion plant. In addition, Bedoussac et al. [
15] described that the selection of seeding rate depends on the leguminous species: In durum wheat–pea mixture, the sowing density of pea can be the same as the recommended rate for pure stands, while horse bean seeding rate must be halved in order to achieve a 50% grain ratio in the harvested crop. The proportion of the companion plants within the mixture is also an important factor for mixed cropping, as it directly affects competitiveness. In the experiment of Guiducci et al. [
16], in a temporary plant association of wheat and legumes, the competitive ability of pea, horse bean, and clover determined the level of nitrogen accumulation and nitrogen supply to wheat before their devitalization. Benincasa et al. [
17] investigated initial interference in horse bean/wheat and rapeseed/clover mixtures in greenhouse experiments. The best performance of the mixture is realized with a 1:3 seeding rate of the dominant horse bean and rapeseed to the companion plant. Furthermore, in vetch–oat mixed cropping, the competition between the companion plants increased above the threshold limit (10%) of oat [
18].
This paper is a part of a series of articles summarizing the results of the work of our research team. In our previous articles, we dealt with the parallel development process of the companion plants and the analysis of yield components [
19], then we determined the interaction between wheat and pea based on competition indices and economic benefits [
20]. The aim of this paper is to answer the following questions: (i) How much yield of winter wheat produces in pea mixed cropping compared to the monoculture? (ii) How the quality parameters of winter wheat change in plant association compared to pure stands? (iii) Does the choice of seeding rate and plant density affect the development of the final yield quantity and quality?
2. Materials and Methods
2.1. Experimental Design
The present experiment was conducted over a period of three growing seasons (2020–2023) at the Applied Agronomy Research Station of the Hungarian University of Agriculture and Life Sciences in Szeged-Öthalom. The Research Station is located in the Great Plain in Hungary at 46.291° N, 20.088° E, 70 m above the sea level. We used the mixed cropping type of intercropping, where the mixed plots were arranged in a randomized block design with four replicates. Each plot measured 10 square meters (1 × 10 m). The soil type of the sample area was identified as meadow chernozem soil, which is characterized by a slightly alkaline reaction (pH = 7.9), and the organic matter is 2.8–3.2% (0–20 cm depth), P: 0.24 g kg
−1, K: 0.2 g kg
−1, NO
3− + NO
2: 0.24 g kg
−1. In each growing season, the previous crop was winter wheat. Although all the seeds were covered, the pea seeds were not inoculated. The sample area was not fertilized apart from multi-nutrient autumn fertilizer before sowing at a rate of 200 kg ha
−1 (N:P
2O
5:K
2O 8:15:15). Seedbed preparation was based on ploughing (20 cm), disk harrowing (15 cm), and cultivator (5–10 cm). The mixed cropping consisted of a combination of three winter wheat varieties (GK Szilárd, Cellule, GK Csillag) and one winter pea variety (Aviron), with three different seeding rates in accordance with the local standard cropping practice. The 100% sowing rate for wheat meant 5 million seed ha
−1, and 1 million seed ha
−1 for winter peas. We proceeded in the same way with the 50% and 75% seeding density. The species were cultivated at all seeding rates; thus, we had a total of 27 associated plots to which we added another 12 pure-sown plots. Thus, the total number of parcels were 156, each block separated by a cultivation path (3 m). The exact seeding rates are summarized in
Table 1.
GK Szilárd is a medium-ripening, awnless, and mill-quality winter wheat variety. This variety is notable for its high yield potential, its ability to thrive within a range of environmental conditions, and its robust stem structure. Cellule is also a mid-ripening wheat variety with strong tillering characteristics with awn. It exhibited high yield stability and nutrient utilization capacity even under unfavorable conditions. GK Csillag can be categorized as an early-maturing, awnless winter wheat variety. Thanks to its excellent adaptability to environmental conditions is a key factor contributing to its capacity to provide a high and balanced yield. It can be threshed easily. Aviron is a semi-leafless, medium-ripening winter pea variety with excellent winter hardiness and disease resistance.
The sowing was executed using a Wintersteiger Plotseeder Plotman machine (Wintersteiger GmbH, Ried, Austria) with a row width of 12.5 cm, at a depth of 4–6 cm. A one-stage harvest was conducted utilizing a Wintersteiger Nurserymaster plot combine (Wintersteiger GmbH, Ried, Austria), which was adapted to the full ripening stage of winter wheat (BBCH 89-92). After that, the crop was cleaned by a Pfeuffer SLN3 Sample Cleaner, and the quality parameters were measured with an Infratec
TM NIR Grain Analyser (Foss, Hilleroed, Denmark). In the case of winter wheat, the following parameters were determined: crude protein, gluten, Zeleny index, and W-value. In addition, the protein content of winter peas was ascertained. The sowing and harvesting dates of mixed cropping are given in
Table 2.
2.2. Weather Conditions
The annual precipitation and temperature data can be seen in
Figure 1.
In the first growing season (2020/2021), it was rainy weather at the time of sowing, which had a favorable effect on the initial development of both companion plants. Later, the drier November and December and the lack of winter snow caused a decrease in soil water balance. April 2021 was not optimal in terms of the water needs of peas, but the rainfall in May proved to be enough for its development, which was facilitated the differentiation of the wheat ears. We measured very little precipitation in June, which mainly affected wheat sensitively. In contrast, the second growing season (2021/2022) was characterized by extreme, uneven rainfall distribution. The vegetation period started with much less rainfall than usual, which was critical for the development of both plants despite the heavy rains in December. Then, a strong lack of precipitation that occurred during the pea flowering period affected the plant in its most sensitive phase. This phenomenon had a negative effect on the formation of the number of ears, too. The wetter April and May were more favorable than this, which covered the maximum water demand for the development of the companion plants. The rainfall during the grain-filling period of wheat was less than the optimal level. The third growing season had the least amount of precipitation at the time of sowing. This initial disadvantage was significantly alleviated by the rainfall of the winter months. After that, another slightly drier period began in the spring of 2023, which did not benefit either species. Although the rainfall distribution was not as extreme as the previous year, the lack of rainfall occurred in the critical stages of the companion plants.
In terms of temperature, wheat was close to the optimal level during the entire development process in the first growing season (2020/2021). The flowering of peas started earlier than expected in mixed cropping. As a result, the temperature at the time of flowering was lower than the optimal level. None of the species surpassed the critical upper threshold value, which would have negatively affected the development of wheat and pea. In the second growing season (2021/2022), the winter was colder than the previous year, but the average temperature did not fall below the freezing point. February was slightly warmer than usual, but this did not cause a significant change in the development of the companion plants. On the other hand, high temperatures were experienced during the flowering and ripening of wheat and in many cases lasted for a long time. Wheat and pea also began to dry from below. The last growing season (2022/2023) was moderately warm and typical for the season. The warm days appeared again this year at the end of May, but this warming was gradual, which left enough time for the plants to acclimatize.
2.3. Statistical Analysis
The data of the wheat yield and quality values were analyzed statistically using the method of multivariate analysis of variance (MANOVA) in a three-way random block design with dependent variables yield, protein, gluten, Zeleny index, W-value as well as with the factors ‘variety’ (of winter wheat), Zpi (the seeding rate of winter peas in the mixture), and Zwi (the seeding rate of winter wheat in the mixture) and the three growing seasons as blocks. In each growing season, the experiment was based on a randomized block design, with 4 replications and 27 mixed cropping plots. The overall MANOVA was evaluated based on the unexplained variance rate expressed by Wilk’s lambda. Following a significant MANOVA result, subsequent three-way random block design univariate ANOVA tests were performed with Bonferroni’s correction to avoid Type I error probability inflation. To ensure the normality requirements of the linear model, W-value was subjected to an ln transformation. The normality of the model residuals was accepted based on the absolute values of their skewness and kurtosis, which were all below 1. The homogeneity of variances was tested using Levene’s test (
p > 0.05). Finally, for the significant factors, post hoc tests were carried out using the Games–Howell method: The homogeneous groups of varieties were separated for all Zpi × Zwi combinations, while Zwi and Zpi levels were compared for all variety × Zpi and variety × Zwi combinations. The results are presented separately for the three years. All statistical analyses were performed using the statistical software IBM SPSS v.29 (IBM SPSS, Armonk, NY, USA, 2022) [
21].
4. Discussion
Although mixed cropping is a widely used cultivation method in the world, in practice it still receives more emphasis in developing countries [
22]. Despite the many advantages of its application, the direction of agronomic research is largely limited to a single crop or a single cultivar [
1,
2], since a plant association raises more complex problems compared to a monoculture [
11].
One of these issues is the yield advantage of mixed cropping, which many studies consider a key feature, especially under low nitrogen fertilizer supply [
13,
23]. In contrast, according to Willey [
14], the yield of intercrops may exceed the total yield of the individual species grown separately. In the research work of Pelzer et al. [
24], the yield of the low-input wheat and pea mixed cropping only approached the level of the conventionally fertilized pure stands of cereal crops. The results of these studies are consistent to a certain degree with our experiment, where only a limited number of combinations surpassed the values of pure stands. In the first growing season (2020/2021), the highest yield was achieved with the GK Csillag variety (control + 24%). Meanwhile, in the second growing season (2021/2022), there was a yield surplus for GK Szilárd and Cellule wheat varieties, but the maximum value was represented by GK Szilárd (control + 13%). On the other hand, there was no significant difference between the yield values of the wheat varieties, which was also confirmed by the statistical analysis. In our experiment, we did not use nitrogen fertilizer, so the analysis purely shows the performance of mixed cropping. In those plant associations where nitrogen fertilization takes place, the yield-increasing effect is predominantly observed in wheat, while there is a substantial decline in pea yield [
25]. In terms of the final crop, the total yield represents at least the same value as pure stands, but the ratio of the companion plants within the mixture shifts in the direction of wheat. A similar result is obtained if the mineral nitrogen supply in the soil is already at a high level, or if the competitive balance between the companion plants is upset due to overcrowding. In the second case, the high plant density increases the competition between the companion plants, a phenomenon that primarily favors the dominant species [
14]. A decline in the proportion of legumes within the mixture has been demonstrated to result in a decrease in nitrogen fixation [
26]. The presence of competitive hierarchy not only reduces the benefits of legumes but also impairs the efficiency of the association. In the case of wheat, the backwardness is not necessarily visible, yet it has the capacity to contribute to the formation of yield quantity and quality, and it is evident that the plant allocates a significant proportion of its energy expenditure to competition. This phenomenon was realized in our experiment, which becomes visible with the change in the proportion of wheat within the harvested crop from the original seeding rate (
Figure 2).
The proportion of wheat in the final crop in each growing season increased by an equivalent amount to the decrease in pea yield. There was only one mixture in 2020/2021, when the proportion of wheat became less than the original seeding rate. According to Bedoussac et al. [
15], this compensation enables peas to produce yield despite the strong dominance of wheat.
Another key issue of mixed cropping is increased yield stability [
15,
27], where several plants growing simultaneously is a guarantee against yield loss even under unfavorable growing conditions and provides greater financial stability compared to the monoculture [
11]. Due to the distinct growing seasons of the companion plants, at least one of them can be grown successfully, which is not possible in pure stands [
28]. According to Pelzer et al. [
24], yield stability between the growing seasons depends on several factors, such as soil climatic conditions, seeding rate, sowing time, the proportion of the companion plants within the mixture or nitrogen distribution. In the present experiment, no significant deviation in yield was observed between the first two years, although the second year was characterized by severe drought, and the symptoms of heat stress were observed in the entire standing crop. This phenomenon suggests that the number and duration of hot days did not occur during the critical stage of wheat; consequently, the decline in yield was not of significant proportion. Moreover, in the second growing season (2021/2022), a yield surplus appeared for the two medium-ripening wheat varieties (GK Csillag was just below the control level), which indicates that the specific microclimate of mixed cropping provided a kind of protection against unfavorable environmental conditions. In the third year, the recurrent absence of precipitation resulted in a significant decline in yield, regardless of the wheat variety. It can be assumed that the factors affecting yield fluctuations could have caused a lower yield together, rather than separately.
In recent years, the paradigm of crop cultivation has evolved. The emphasis gradually shifted from the pursuit of quantity to quality. In the contemporary context, a cultivation method that meets the criteria of sustainability is considered effective in the long term. One of the basic pillars of this is high-quality plant production [
5]. In winter wheat and winter pea mixed sowing, there was a mixture every year whose quality parameters exceeded the value of pure stands. However, this surplus did not occur with the same wheat variety. In the case of protein, the Cellule/Aviron 50:100 mixture just exceeded the value of pure stands in the first year, while in the third year it achieved the best value of all (pure stands + 11%). In the second year, outstanding results were attained by the GK Szilárd/Aviron combination with the 50:75 and 50:100 seeding rates. The same wheat variety selection process was carried out for the Zeleny index and W-value, with the difference that in the first year, the W-value showed a significant difference between the Cellule and GK Szilárd varieties, in favor of Cellule. In the second year, the baking values were much higher than in the monoculture (Zeleny index: + 10%, W-value +25%). High values in the third year occurred in almost all mixtures, which was also true for gluten, but this quality parameter differed from the others in terms of wheat variety selection. In the first and the third years, the aforementioned Cellule/Aviron combination in the same seeding rate (50:100 and 75:50) also appeared among the mixtures exceeding pure stands, but the GK Csillag/Aviron 50:75 combination achieved even higher values (pure stands + 3% and 39%). In these years, we were able to prove a statistical difference between wheat varieties in favor of GK Csillag. We observed a similar significance in the second year, but already with the Cellule variety. In this drought year, the Cellule/Aviron 50:50 mixture proved to be better than the others.
In summary, the composition of our experiment with a high wheat seeding rate was beneficial for baking qualities; the experimental results showed consistently high values for certain mixtures even under unfavorable conditions. However, it should be noted that the performance of mixed cropping fell short of the level that could have been achieved by reducing the seeding rate of the dominant wheat species in favor of peas, as Benincasa et al. [
17] also experienced with a mixture of horse bean and wheat. This contradicts the statement of Bedoussiac et al. [
15] that the sowing density of peas in a mixture can be equal to the value of pure stands. With regard to the further direction of research work, the question is to what extent the quality parameters of the companion plants would change within the mixture by increasing the proportion of peas.