Intercropping organic melon and cowpea combined with return of crop residues increases yields and soil fertility

The growth of legumes, reduced tillage and addition of crop residues can be regarded as a good alternative in intercropping systems to increase soil organic matter, soil fertility and biodiversity while enhancing crop production and reducing the use of fertilizers. Despite the potential benefits, there is still a research gap about using the combination of cowpea and melon in intercropping to increase productivity and reduce external inputs. Thus, the aims of this study were to: i) assess if crop yield, crop quality and soil physicochemical properties can be improved by intercropping systems between melon (Cucumis melo L.) and cowpea (Vigna unguiculata (L.) Walp.) with reduced tillage and addition of crop residues, compared with a melon monoculture with intensive tillage and removal of crop residues, all grown under organic management; and ii) evaluated if cowpea grown as intercrop with fertilization reduced by 30% in the diversified plots can partially replace the use of fertilizers with no negative effects on total crop production. In this study we compared over three crop cycles monocrops with three different melon-cowpea intercropping patterns: mixed intercropping, row intercropping 1:1 (melon:cowpea) and row intercropping 2:1 (melon:cowpea). Our results, presented in this study, showed that intercropping systems, regardless of the pattern, kept soil organic C levels, while it significantly decreased in melon monoculture. Intercropping also significantly increased soil total N, available P and exchangeable K (0.13%, 62 mg·kg-1 and 387 mg·kg-1, respectively), compared to the melon monocrop (0.11%, 25 mg·kg-1 and 306 mg·kg-1). Total crop production was significantly higher under diversified systems, with land equivalent ratios > 1. Hence, the introduction of cowpea associated with melon, combined with reduced tillage and the incorporation of crop residues could be considered as a feasible strategy for sustainable agriculture, with environmental gains and economic savings for fertilizers and water.


Introduction
Conventional agriculture in the last decades has caused water scarcity, low biodiversity, soil and water pollution, high levels of greenhouse gas (GHG) emissions, loss of soil organic matter (SOM), high erosion rates and high incidence of pests and diseases (FAO 2017;JRC 2012). In this context, the priority in modern agriculture is maintaining productivity and social welfare while minimizing negative environmental impact (Duhamel and Vandenkoornhuyse 2013). Organic farming, which generally does not allow the use of synthetic inputs, is regarded as a suitable alternative to enhance the sustainability of modern agriculture while decreasing environmental impacts (Bedoussac et al. 2015). In a well-managed organic farming system, that promotes a circular economy with the recycling of crop residues and reduced tillage, soil health is enhanced, which entails producing the highest yields possible, in a sustainable, ecofriendly manner (Sofia et al. 2006). In fact, many studies have shown that the transition from conventional to organic farming can lead to high yields after a settle down period of ~5 years (Badgley et al. 2007;Seufert et al. 2012). Nevertheless, despite the benefits of organic farming compared to conventional farming, many organic systems, mostly in horticulture, are based on monocultures. Monocropping has had the objective of increasing the economic efficiency of agrifood systems, leading to a sharp decline in within-field and landscape crop diversity (Messéan et al. 2021). Thus, monocultures frequently cause land degradation by depletion of soil organic matter, reduction of below-and aboveground biodiversity, water pollution by use of fertilizers (even in organic systems) and increased pest/disease incidence by reduction of natural enemies (Morugán-Coronado et al. 2022Vanino et al. 2022). In addition, monocropping is associated with economic threats owing to the low resilience to variability in prices, markets, climate and pests/diseases due to the dependence on a single crop (Barnes et al. 2015). As a consequence, in recent years, there has been a growing interest in crop diversification as a strategy to counteract the negative effects of monocropping, through the association of different species in time and space and an optimized use of resources (Brooker et al. 2015). Crop diversification in organic farming can include one, several or all of the following: (i) diversification of crops grown on the same land in successive growing seasons (crop rotation); (ii) diversification of crops grown on the same land within a growing season (multiple cropping); and (iii) diversification of crops grown in proximity in the same field (row intercropping, strip intercropping, mixed intercropping) (Mao et al. 2015). This strategy, if properly designed, can help farmers enhance nutrient availability, increase overall crop production, and potentially reduce the need for external fertilizer inputs, contributing to decreased production costs and improving gross margins (Alcon et al. 2020). Nonetheless, farmers willing to adopt crop diversification may lack technical knowledge or locally-adapted minor crop varieties, mainly due to low research investments capacities (Messéan et al. 2021). Thus, to speed up the transition toward diversified agro-systems, new scientific and technical knowledge is needed.
Combining organic farming and intercropping could be proposed as a global response to the challenges of future agriculture, offering potential improvements in productivity, resource use efficiency, and environmental sustainability. The use of intercropping, when species are properly selected, can lead to an overall production per area greater than monocrops, with values of Land Equivalent Ratio (LER) >1. LER is defined as the ratio of the area under sole cropping to the area under intercropping needed to give equal amounts of yield at the same management level (Mead and Willey 1980). The increase in land productivity by intercropping can be explained by the fact that crop diversification can increase the availability of nutrients and the resistance and resilience of the agroecosystem to drought and disease incidences (Lin 2011). This is achieved through the complementarity in the use of resources and ecological facilitation processes between the species (Franco et al. 2015). Nonetheless, not all intercropping systems constitute improvements, it is important to not combine species that compete too much for nutrients, water, or sunlight (Lithourgidis et al. 2011).
Melon crops (Cucumis melo L.) are among the most common crops cultivated in regions with warm and hot climates, and are commonly grown as monocultures even in organic management (Aldoshin et al. 2020). Melon monocrops, like many other monocrops, tends to reduce SOM (Sánchez-Navarro et al. 2019a), and soil microbial and invertebrate biodiversity, contributing to soil and water pollution by pesticides and excessive use of fertilizers (Hijri et al. 2006;Jiao et al. 2011).Legumes are of a particular interest in organic farming where N availability is often limiting especially in the absence of livestock (Christophe David et al. 2005). In arid environments, the legume crop cowpea (Vigna unguiculata (L.) Walp.) is normally used because of its adaptability, low fertility requirements, and its active rhizodeposition, improving soil fertility (Amorim et al. 2022). Active rhizodeposition can stimulate soil microbial communities, contributing to increase soil fertility by SOM mineralization and/or the solubilization of many soil nutrients such as K, P, Fe and Ca, which can be unavailable for plants owing to chemical precipitation in basic soils (Latati et al. 2016;Łukowiak et al. 2016). The complexity and biodiversity of the microbial community structure and its functionality is linked with the diversity in the crop associations (Njeru et al. 2015), and so it is essential to properly define the most effective crop combinations in intercropping to stimulate microbial communities that mobilize nutrients and decrease the incidence of soil-born diseases, and so allowing to reduce the use of external inputs. Intercropping in maize and other cereals is becoming very common, and legumes are preferably chosen in cereal-based intercropping system (Maitra et al. 2021). Grain and legumes associations are characterized by increasing SOM and reducing GHG emissions compared to their monocrops (Qin et al. 2013). In this sense, intercropping between melon and cowpea under organic management could contribute significantly to overcoming the challenges of developing both productive and environmentally friendly agricultural systems for melon cultivation. The use of rotations as a diversified cropping system is quite common in scientific literature, but there is only little information on the use of intercropping in vegetables production (Bacchi et al. 2021;Baumann et al. 2002). In this sense, there is not a complete knowledge about the mechanisms of complementarity and facilitation between economically profitable crop associations in horticulture to increase productivity and the delivery of ecosystem services.
The aims of this study were to: i) assess if crop yield, crop quality and soil physicochemical properties can be improved by: i) intercropping systems between melon (Cucumis melo L.) and cowpea (Vigna unguiculata (L.) Walp.) combined with reduced tillage and addition of crop residues compared with a melon monoculture with intensive tillage, all grown under organic management; and ii) assess if cowpea grown as intercrop with fertilization reduced by 30% in the diversified plots can partially replace the use of fertilizers with no negative effects on total crop production ( Figure 1).
We hypothesized that the inclusion of cowpea in intercropping with melon under organic management can allow the decrease in fertilizers rate and have a positive effect on melon yield by stimulation of microbial communities that makes nutrients more available, besides increasing the overall production by the harvest of a new commodity. Owing to the active rhizosphere of legumes, and their ability to fix atmospheric N, we assume that intercropped systems should show higher availability of soil nutrients and content of soil organic carbon than the melon monocrop, despite the reduction in external fertilizers.

Study site and experimental design
This study was carried out in Cartagena, South-East of Spain, at Tomás Ferro Experimental Farm of the Universidad Politécnica de Cartagena (37° 41` N; 0° 57` E). Climate is semiarid Mediterranean with a total annual precipitation of 275 mm and a mean annual temperature of 18 °C. Annual potential evapotranspiration surpasses 900 mm. Soil is classified as Haplic Calcisol (loamic, hypercalcic) (Food and Agriculture Organization of the United Nations 2014), with clay loam texture.
We compared a melon monocrop (Cucumis melo L.) and a cowpea monocrop (Vigna unguiculata (L.) Walp.) with intensive tillage and elimination of crop residues, with different melon-cowpea intercropping systems, with reduced tillage and addition of crop residues. The study was conducted over three separate summer crop cycles, each lasting one growing season (from May to August) during the years of 2018, 2019, and 2020. Based on intercropping, reduced tillage and addition of crop residues we wanted to propose an alternative cropping system to current intensive melon monocultures in organic systems. The intercropping systems were: i) mixed intercropping (alternation within the same row of melon and cowpea plants), row intercropping 1:1 (combination of alternate rows of melon -cowpea) and row intercropping 2:1 (combination of two rows of the melon crop and one row of cowpea). Tillage was performed at the beginning of each crop cycle. For the monocrops, we used chisel plow as a traditional practice in the region, which involved plowing the soil to a depth of 30-40 cm. Afterwards, beds were shaped into elevated ridges by double mold-board, and only the tops of the ridges were cultivated. In the intercropped systems, we employed reduced tillage, which involved shallower chisel plowing at a depth of 15-20 cm, followed by double mold-board to make the ridges. Thus, the main difference in tillage treatments between the monocrops and intercropped systems was the depth of tillage, with the aim of reducing soil disturbance in the intercropped systems. In both monocrop systems, melon and cowpea, the crop was mowed after harvest and crop residues used to feed animals, as traditionally performed in the area. In the intercropped systems, and approx. one month after harvest to ensure that all plants from both crops were totally dry, the crop residues were incorporated into the soil up to 15 cm with a chisel plower as a strategy to increase soil organic matter. The amount of melon biomass incorporated in each plot was 663 kg ha -1 , while the amount of cowpea biomass incorporated was 1298 kg ha -1 for row 1:1, 642 kg ha -1 for row 2:1 and 182 kg ha -1 for mix intercropping. After this, soil was let aside until next season (from September to April), with implementation of no treatment. In all systems, compost was added annually at the beginning of each cycle (April), as a traditional practice in the region, with a dose of 14,000 kg ha -1 . This compost derived from sheep manures and had the following characteristics: moisture = 20%; pH = 6.5; electrical conductivity = 20 dS m −1 ; total nitrogen content (Nt) = 1.1 %; C:N ratio = 17.4; P 2 O 5 content = 1.7 % and K 2 O content = 1.9 % AGRIORGAN. Melon monocrop received the equivalent of 3000 kg ha -1 of organic fertilizer NORGAN (plan-based fertilizer with 45% humic and fulvic acids, 3.2% N, 7% K 2 O, Fyneco SL, Spain), cowpea monocrop received the equivalent of 1875 kg ha -1 of NORGAN. Intercropping systems received 30% less fertilizer quantity than the melon monocrop to verify a saving in the use of fertilizers as a result of the development of the legume. This reduction rate in fertilizer application was based on previous research on the N contribution of legumes cultivation in soil (Sánchez-Navarro et al. 2020, 2019b, 2019a. No herbicides were added, and the weed control was done by hand-hoeing. All crops were drip irrigated and grown under organic management. The irrigation was scheduled according to climatic conditions, crop coefficient and evapotranspiration rate. The average irrigation amount was 3016 m 3 ha -1 per crop cycle for all treatments. Thus, with the same quantity of water used to produce melon under monocroppig, we performed two associated crops in intercropping systems. We followed a completely randomised experimental design. Treatments were randomly setup in plots of 120 m 2 (12 m x 10 m) established in triplicate. Melon seedlings were planted in a density of 0.4 plants m -2 , with a spacing of 200 cm between rows and 120 cm between plants in all plots (monocultures and intercropped plots). Cowpea was sown in a density of 5 plants m -2 in the monoculture, with a spacing of 100 cm between rows and 20 cm between plants. Cowpea seeds were sown between two rows of melon in the row intercropped systems, spacing 100 cm between melon and cowpea rows, and 20 cm between cowpea plants. Density of cowpea plants was 2.5 plants m -2 and 1.5 plants m -2 in the row 1:1 and row 2:1 systems, respectively. In the mixed intercropping system, cowpea, in a density of 0.4 plants m -2 , was sown in all melon rows between two melon plants, with a spacing of 200 cm between rows and 120 cm between plants. So, density of melon was the same in the different treatments, but the density of cowpea changed. Crop cycles lasted from May to August. Harvest of these two species is manual in commercial farms, and there was no need for special machinery. The area where this experiment was performed was set aside for the previous two years, with development of spontaneous cover crops. Previously, it was dedicated to the growth of fava bean and broccoli in rotations under organic management for five years.

Soil and plant sampling
Soil was sampled at the beginning of the trial (2017), and at end of each cycle (2018, 2019 and 2020). All plots were sampled at 0-10 cm and 10-30 cm depth. Two composite soil samples derived from 5 sampling points per plot were collected avoiding the border effect. Thus, soil samples from 10-30 cm depth in monocropping included the tillage depth, but for intercropping, samples were a mixture of tillage depth and undisturbed soil. Soil was collected in the crop line, between two plants. Soil moisture was recorded at sampling time using a ProCheck and 5TM sensors (Decagon Devices,USA). Each sample was divided into two aliquots. First aliquot was air dried for 7 days, sieved < 2 mm and kept at room temperature. The other aliquot was immediately kept at 4ºC for and analysed for NH 4 + and NO 3 within the following five days.
Melon crop yield was determined by weighing all the fruits per plot when they were ripe and ready for consumption. With regard to cowpea yield, all the pods in each plot were harvested when the seeds were dried at the end of the crop cycle. External rows in each plot were discarded to avoid the border effect. Cowpea yield was also devided by the density of cowpea in each plot to assess the influence of interspecific competition with melon in each intercropping pattern.Total production was calculated by the addition of melon and cowpea yields in each plot. Sugar content in melon was determined using a refractometer (Brix degrees). Mean melon weight, number of melons, Brix degrees, the weight of 100 cowpea seeds and seed protein content were recorded as crop quality parameters. Land equivalent ratio (LER) was used to compare intercropping and monocrop yields, calculated following equation 1 defined like in Maitra et al. (2021).
where, Yab is the yield of "a" crop grown in association with "b" crop and Yba is the yield of "b" crop grown in association with "a" crop. Yaa and Ybb represent the yields of "a" and "b" crops grown in a pure stand, respectively.

Soil analyses
Bulk density was measured using the cylinder method, with cores of 5 cm of diameter and 5 cm length. Soil electrical conductivity (EC) and pH were calculated using deionized water (1:5 and 1:2.5, respectively). Soil organic carbon (SOC) and total nitrogen (Nt) were analyzed by using an elemental CHN (CHN 628, Leco). Particulate organic carbon (POC) was determined according to the method by Cambardella and Elliott (1992). Soil aggregation (SA) was determined using the wet sieving method of air-dried soil (Elliott 1986). The mean weight diameter (MWD) were calculated according to the method by Kemper and Rosenau(1986): where x i is the mean diameter of size class (large macroaggregates > 2; small macroaggregates 2-0.25 mm; microaggregates 0.25-0.053 mm; silt and clay < 0.053 mm) and w i is the proportion of the total sample mass in the corresponding size fraction after deducting the mass of stones. Available P was analyzed using the Burriel-Hernando method (Díez 1982), using Burriel-Hernando solution (0.2 g CaCO 3 , 0.17 g MgCO 3 , 5 mL glacial acetic acid and 0.2 mL H 2 SO 4 in 2 L deionized water) in a 1:25 soil:extractant ratio. Cation exchange capacity (CEC) was determined using BaCl 2 as exchangeable salt (Álvaro Fuentes et al. 2019). Exchangeable Na, Ca, K and Mg were measured in the BaCl 2 extract from CEC. Available B was extracted with deionized water (1:5 w/v) at 50 ºC (Tang et al. 1997). Bioavailable oligoelements (Fe, Mn, Cu and Zn) were extracted using the chelating agent DTPA (1:2 w/v) (Álvaro Fuentes et al. 2019). P, Ca, Mg, Na, K, B, Zn, Cu, Mn and Fe concentrations were determined using ICP-MS (Agilent 7900). Ammonium (NH 4 + ) was extracted with 2M KCl in a 1:10 soil:extractant ratio (Keeney and Nelson 1982) and colorimetrically measured (Kandeler and Gerber 1988). NO 3 − was extracted with deionized water in a 1:10 soil:extractant ratio and measured by ion chromatography (Metrohm 861).

Statistical analysis
Data were checked to ensure normal distribution using the Shapiro Wilk test at P < 0.05 and log-transformed when necessary to ensure normal distribution. Homoscedasticity was checked by the Levene test. Data were submitted to three-way repeated measures ANOVA, with year (2019, 2020 and 2021) as within-subject factor, and treatment (monocrops and intercropping patterns) and depth (0-10 and 10-30 cm) as between-subject factors. The relationships among the properties were studied using multiple regression and Pearson's correlations. A principal components analysis (PCA) was performed with all data to study the structure of x i w i dependence and correlation established among the variables studied. Statistical analyses were performed in R version 4.1.2 for Windows.

Crop yields and crop quality properties
Mean melon weight and the number of melons per hectare showed no significant differences between intercropping and monocrop systems with an average value of 2.88 kg and 8795 melons·ha -1 , respectively (Table 1). However, there was a trend to increase the number of melons with all intercropping systems. Melon crop yield was significantly highest (p < 0.05) under intercropped systems compared to monocrop, with no significant difference between intercropping patterns (Table 1). Cowpea crop yield showed significantly highest (p < 0.05) values in the monocrop system. Mixed intercropping and row intercropping 1:1 showed the lowest (p < 0.05) cowpea yields. However, with regard to the cowpea yield per unit of plant density, there were no significant differences between treatments (p > 0.05), although mixed intercropping and row intercropping 2:1 tended to increase this quotient, indicating higher grain production per plant. Total production was significantly higher under diversified systems for the three crop cycles, with no significant differences between the three intercrop patterns (p < 0.01). LER ranged between 1.45 and 1.92 in the three crop cycles, with no significant differences between intercropping patterns, but with high variability between years. Protein content of cowpea seeds was significantly higher under row intercropping 2:1, and lowest under monocrop (p < 0.01). The weight of 100 seeds and brix degrees were not significantly affected by diversification, with an average value of 23 g and 11.19 %, respectively. The interaction year x diversification was significant for the mean melon weight, cowpea yield, protein content of cowpea seeds and the weight of 100 seeds.

Soil physicochemical properties
SOC was not significantly affected by crop diversification ( Figure 2A), with no significant differences among treatments in the evolution of three years period, with an average value of 1.18% for 0-10 cm and 1.14% for 10-30 cm. However, SOC tended to decrease (p < 0.05) with melon monocrop in both depths, while cowpea monocrop or intercrops maintained SOC level at similar values during all crop cycles (p > 0.05). Nt showed significant differences among treatments (p < 0.01), with significantly highest values in cowpea monocrop and intercropped systems compared to melon monocrop in both depths, despite reducing N fertilization under intercropping ( Figure 2B). Nt showed an average value of 0.11% for melon monocrop at both depths and 0.13% and 0.12% for intercropping systems for 0-10 cm and 10-30 cm respectively. In general, Nt was significantly higher in the surface (p < 0.01). POC showed significant differences among treatments (p < 0.01) in the experimental period ( Figure 2C), with highest values in the cowpea monocrop (0-10 cm) with an average value of 2.75 g·kg -1 . Comparing to melon monocrop, POC showed no significant differences with intercropping systems except for the last year of study, where there was a significant increase (p < 0.01) in the intercropping systems at both depths, with an average value of 1.78 g·kg -1 and 2.32 g·kg -1 at 0-10 cm for melon monocrop and intercropping systems, respectively, and 1.33 g·kg -1 and 2.10 g·kg -1 at 10-30 cm for melon monocrop and intercropping systems, respectively (p > 0.05). Furthermore, POC was significantly higher (p < 0.05) at 10-30 cm depth in the cowpea monocrop and intercrops than in the melon monocrop. The monocrop systems had significantly lower content of available P in soil compared to the intercrop systems (P<0.001, Figure 3A), with no significant differences between melon and cowpea monocrops. There were no significant differences between intercropping patterns with regard to available P content. Before the establishment of the field trial, soil had high quantity of NO 3 -( Figure 3B), but owing to the optimized fertilization program developed during this assay, NO 3 values decreased. However, despite the significant decrease after the first cycle, soil NO 3 levels significantly increased in plots with cowpea monocrop and mix intercrop at 0-10 cm compared to the other treatments (p < 0.05) ( Figure 3B). NO 3 was significantly lower at 10-30 cm depth in intercropping systems (p < 0.05) compared to monocrops. With regard to available K, intercropping systems significantly increased the content of available K at both depths compared to both monocrops (p < 0.001), with averages values of 306.69 mg·kg -1 for melon monocrop, 331.29 mg·kg -1 for cowpea monocrop and 387.16 mg·kg -1 for intercropping systems at 0-10 cm, and 290.59 mg·kg -1 for melon monocrop, 317.05 mg·kg -1 for cowpea monocrop and 374.76 mg·kg -1 for intercropping systems at 10-30 cm (p < 0.05) ( Figure 3C). There were no significant differences between intercropping patterns.
During the experimental period, crop diversification significantly influenced EC, CEC, Na, Ca and Mg, with no significant effects on NH 4 + and MWD (Tables 2 and  4). EC and CEC were significantly lowest (p < 0.01 and p < 0.001, respectively) under melon monocrop, with increased values with the growth of cowpea, as sole crop or under intercropping. Exchangeable Ca significantly increased with cowpea monocrop at both depths (p < 0.01). Exchangeable Mg significantly increased with cowpea monoculture and all intercropping patterns compared to melon monocrop, with no significant differences between intercrop patterns (p < 0.001).
With regard to micronutrients (Tables 3 and 4), intercropping, independently of the pattern, significantly increased Mn, Cu, Zn and B compared to monocrops (p < 0.001) at both depths. Crop year significantly affected all soil properties except for NH 4 + (Table 4), suggesting some annual shifts depending on the specific crop cycle (p < 0.001). In this sense, the interaction of year and diversification (Y x D) was significant for NH 4 + , EC, Ca, Mg, Fe, Cu and Zn (Table 4). The interaction of year and soil depth (Y x d) was also significant for NH 4 + , MWD, Na and Mg. Soil pH was not affected by any of the studied factors, with average values for all systems of 8.41 at both depths (data not shown).

Interrelationships between crop and soil properties
The PCA performed with the soil physicochemical properties, crop yields and crop quality properties showed that 63.95% of the total variation could be explained by the first three PCs (Figure 4; Table 5). PC1 (30.98% of the variation) was related to Brix degree, total production, melon yield, mean melon weight and number of melons, and negatively related to Ca, Fe and EC, and separated cowpea from the other systems. PC2 (18.69% of the variation) was related to K, Nt, P, Na and SOC, and separated melon monocrop from intercrops, showing that these soil properties were the ones more affected by intercropping compared to melon monocrop. PC3 (14.28% of the variation) was related with Cu, LER, Mn, Zn and pH, separating cowpea monocrop from intercropping plots.

Crop production and quality
Crops have different needs, so it is especially important to combine them in the right way to obtain yield improvements. Melon crop yield was not affected by interspecific competence with cowpea, since cowpea growth, associated to the addition of crop residues, increased soil fertility despite reducing external fertilization, contributing to maintain high melon crop yields. High variability between years in melon crop yield could be explained by differences in genotypes provided every year by the nursery, although the variety is the same, and different environmental parameters between years (Mohamad et al. 2010). With these regards, comparing average values of weather indicators for the three crop cycles (Supplementary Table S1), we can confirm this last assertion, with a possible influence of weather conditions on melon yield. In this sense, the lowest melon    crop yield (2020) coincided with the year with the lowest precipitation and lowest average temperature. The year with the highest melon crop yield (2019) was the year with the highest precipitation and lowest hours of sun, radiation and evapotranspiration. In our experiment, the highest cowpea yield in the monocrop could be explained by the high density of the plants compared to the other systems. However, the relationship between cowpea yield and plant density tended to be higher in those systems where intraspecific competition is lower (i.e. lower density of cowpea plants: mix intercropping and row intercropping 2:1). Thus, cowpea seems to be more productive growing in association with melon than close to other cowpea plants, highlighting the efficiency of species associations to enhance productivity. Although interspecific facilitation studies in intercropping remain scarce, facilitative interactions between species can enhance the availability of N, P, water resources and other essential resources like Zn and Fe (Homulle et al. 2022). Factors like density of plants or growth state can have an influence in the facilitative interactions, like for example in N transfer (Islam and Adjesiwor 2017). Thus, in the intercropping systems, the density of cowpea was lower to that in the monoculture, and production was lower, but when expressed based on plant density, data indicates a facilitation process between both species that has led to higher yields per unit of plant. The highest content in protein of cowpea seeds under the row intercrop 2:1 may be explained by the lowest interspecific competence with melon, increasing the quality of cowpea. As far as we are concerned, these melon-cowpea intercropping systems with incorporation of crop residues and reduced tillage, have not been previously studied, but this strategy could be an important choice for sustainable horticulture management. In this line, total production (melon + cowpea) was higher in intercropping systems despite using less fertilizers and the same quantity of water, with no negative effect on quality properties or nutritional characteristics. The three tested patterns of intercropping have shown LER values > 1, which confirms the high efficiency of the proposed intercropping system. Hence, by using the same quantity of water for irrigation and decreasing the amount of fertilizers by 30%, the introduction of cowpea associated to melon incremented the overall land production. This could be considered as a sustainable strategy from environmental, social and economic point of view, with economic savings for water and fertilizers. There are many reasons why intercropping and recycling of residues are important for the future of agriculture, like for example, improving interspecific facilitation, minimizing N and P inputs and the competition for available resources, improving efficiently in the use of resources, influencing in the presence of weeds and increasing yield productivity for non-legume companion crops (Chamkhi et al. 2022;Maitra 2019  with all intercropping patterns, and so, to take a decision on the most suitable pattern, only production values and management practices should be considered. Thus, it would depend on farmer interests depending on market demands for the products and the facility to implement the different patterns in the farms.

Soil organic matter, organic carbon, and total nitrogen
The implementation of intercropping systems has a profound impact on the dynamics of SOM, SOC and Nt. Our investigation indicates that intercropping can effectively maintain the decreasing trend of SOC content seen under melon monocrop cultivation. By integrating practices such as reduced tillage and crop residue incorporation, these systems can, at the very least, prevent the loss of SOC attributed to vegetable production under a Mediterranean climate. This loss has been known to have detrimental effects on the sustainability of agroecosystems (Sánchez-Navarro et al. 2019a).Longer periods would be needed to assess if the proposed cropping systems can even reverse this trend and contribute to soil C sequestration and storage. In this line, previous studies have reported that intercropping can increase soil C sequestration in the long term .
The type of crop residue can also affect SOC, as demonstrated by an 11-year field study of rotations with rice in southeast China, with increases in labile organic C with implementation of winter crops (Chen et al. 2016). In addition, intercropping systems, with addition of crop residues, were even able to increase Nt, despite decrease external fertilization, as previously reported in other studies which used legumes as intercropping species (Amossé et al. 2014;Baldé et al. 2020;Chander et al. 1998).

Table 4
Results (F-value) of three-way repeated measures ANOVA of soil properties shown in Tables 2 and 3 Cowpea fixes atmospheric N and supplies it to associated plants like melon and at the same time provides soil shading to conserve water moisture and avoid fruit insolation (Munisse et al. 2012). Beyond biological N fixation, when the legume biomass decomposes, additional N is released into the system and made available to subsequent crops. This additional N input in intercropping systems have contributed to replace external fertilizers, with no negative effect on melon crop yield. Thus, not only the growth of cowpea in intercropping but also the incorporation of crop residues into the soil may have contributed to this increase in soil fertility, as previously reported (Bhuiyan and Zaman 1996) N-rich root exudates may stimulate microbial populations (Fustec et al. 2011). Furthermore, legumes secrete larger amounts of carboxylates than other plant families per unit of soil volume (Neumann and Römheld 1999;Pearse et al. 2006), phosphatases (Makoi et al. 2010;Nuruzzaman et al. 2006) and protons (Tang et al. 1997). The release of all these root exudates activates microbial activity, which can also increase humification processes and so stabilize SOM content, using the crop residues incorporated in the soil in the intercropping systems as a source, not present in the melon monocrop (Kuzyakov 2010;Sanchez et al. 2004). This strategy and processes can explain why SOC was higher compared to melon monocrop at the end of the experiment in all intercropping systems. SOM pools, which are composed of many essential elements, can improve soil chemical, physical and biological processes as a result of release of essential nutrients by mineralization or desorption, stable aggregation with mineral particles and release of enzymes that will allow microbial growth (Lal 2014). POC showed no significant differences with intercropping systems except in the last year of study, where its content increased in all intercropping systems comparing to monocrop at both depths. Thus, the deposition of crop residues and root exudates in the intercropping system by the development of two different plant species promoted higher levels of POC, directly influencing the availability of this highly labile substrate for microorganisms. So, the interaction between plant species seems to be a key factor to enhance these labile organic sources. This intensified root activity by intercropping systems, and consequently the increase in rhizodeposition, improves SOC content and has a positive impact on nutrient storage and on the soil exchange capacity (Balesdent et al. 2011;Farrar et al. 2003). In this line, the increasing trend of EC with intercropping can also be explained with the increase of exchangeable cations through increments in SOC compared to melon monocrop (Solly et al. 2020).

Nitrogen losses
While intercropping systems can increase SOC and Nt, they may also influence nitrogen losses, particularly nitrate leaching and N 2 O emissions. Although the incorporation of crop residues into the soil can lead to an increase in soil N, it can potentially result in increased N 2 O emissions and nitrate leaching due to the higher contents of easily available C and N (Chen et al. 2013;Lehtinen et al. 2014). However, in our study, we did not observe enhanced N 2 O emissions or nitrate leaching (data not published). In fact, the reduction of fertilization rate by 30% may have outlined the effect of crop residue addition, since microorganisms need to degrade these compounds to obtain available N, not easily added as external synthetic fertilizer. In this line, Sainju and Singh (1997) observed that inorganic N leaching from non-legume crop residues ranged 29-94%, while from legume residues, N leaching was 6-48%. In addition, Reckling et al. (2016) observed that cropping systems with legumes reduced N 2 O emissions with slightly lower nitrate leaching. Biological N is incorporated by mineralization of residues after harvest or through root exudation, and so, a significant amount of N can be added to the soil via the legume component of an intercropping system (Laberge et al. 2009).

Phosphorus and other nutrients
In general, the cowpea associated with melon in intercropping systems improved P availability, independently of the association pattern. Previous studies have also proven that intercropping systems with legumes can increase soil P availability (Li et al. 2018;Ngwira et al. 2012;Tang et al. 2021). In fact, metabolic energy processes in which P is involved are important in symbiotic N 2 fixation, and that is why legumes are efficient at solubilizing P in soil (Li et al. 2014;Sulieman and Schulze 2010). Soil microorganisms change P from insoluble forms immobilized by precipitation (Mg-P or Ca-P) to soluble forms using organic acids, and so, availability to plants are enhanced through biological processes in the soil (Arcand and Schneider 2006;Kubicek and Druzhinina 2007;Rodríguez and Fraga 1999). In our context, rhizodeposition processes and organic compounds derived from crop residues are the essential part activating microbial population in the intercropping systems affecting available P in the soil (Barros et al. 2007;Hannam et al. 2006). Hence, the incorporation of crop residues and the inclusion of cowpea in the intercropping system increased N content of soil through biological N fixation processes, these active processes activated beneficial microorganisms and therefore there was an increasing in P availability. It is important to highlight that soil available P only increased in intercropping systems, not with monocrops, neither melon nor cowpea. So, the interaction between both species may have activated beneficial microorganisms associated to P solubilization. In this line, Cuartero et al. (2022) studied the bacterial community of the soil samples coming from the first sampling of the present study, and confirmed that intercropping systems, independently of the pattern, increased the abundance of Pseudomonas, Bacillus, Streptomyces and Sphingomonas, which can be considered phosphate-solubilizing bacteria. Nonetheless, the incorporation of crop residues in the intercropping systems may have also contributed to recycle nutrients and increase their availability in soil, with the return of available nutrients uptaken by crops into the soil. Thus, the combination between the association of two crops and the incorporation of crop residues can positively contribute to increase soil fertility, turning into a more productive cropping system than monocrops with removal of crop residues. In this line, previous studies showed that the incorporation of cowpea contained considerable amounts of other nutrients besides N, such as K, Ca, Mg, Zn, Mn and Fe (Bhuiyan and Zaman 1996;Kaledhonkar et al. 2018).
As observed for P, the incorporation of crop residues and the association of melon and cowpea was needed to increase soil exchangeable K, with the stimulation and growth of beneficial bacteria that are able to solubilize K from insoluble and inaccessible forms (Cuartero et al. 2022). Plant rhizosphere through its microorganisms take an important part in K-solubilizing and in the natural potassium cycle (Han and Lee 2005;. Previous intercropping studies have also reported increases of soil available P and K with the implementation of intercropping (Guo et al. 2021;Zhang et al. 2018) and incorporation of crop residues (Medina et al. 2015). Thus, the incorporation of crop residues in intercropping systems emphasizes the potential benefits of combining sustainable practices, such as intercropping and crop residue management, for maintaining and improving microbial activity, soil quality and agroecosystem productivity (Vinther et al. 2004).
In addition to the release of nutrients and C by rhizodeposition, cowpea roots can also play a major role in modifying the chemical composition of the rhizosphere and transforming unavailable Ca and Mg into available (Messaoudi et al. 2020). That could explain our results with Ca higher in the cowpea monocrop, and the highest concentrations of Mg in cowpea monocrop and intercropping systems in our study. With this regard, Xue et al. (2016) also reported highest available soil Zn in cereal/legume intercropping systems. Mn and Cu showed a significantly correlation with Nt, POC and Zn, which suggests that changes in rhizosphere microbial activity is also associated to the dynamics of these nutrients.

Impacts of different tillage depths
Different tillage depth (30/40 cm for monocrops and 15/20 cm for intercropping systems) could have also influenced the difference in soil properties between systems. In this line, P and K could form a stratified layer over several years for a depth higher than tillage, increasing their availability in no till horizons, as observed in our study (Garcia et al. 2007;Wright et al. 2007). In reduced tillage systems it has been shown an increase in POC, CEC, Zn and Mg availability (Chan et al. 2002;Guimarães et al. 2016;Hickman 2007), as in our study, with higher values of these properties in depth in intercropping systems with no tillage compared to monocrops. However, the highest value of these properties also in the cowpea monocrop, with tillage up to 30/40 cm may indicate that crop residues incorporation and root activity may be more responsible than tillage for this effect. In this line, Koga and Tsuji (2009) highlighted the positive effects of reduced tillage, crop residue management and manure application on crop yields and soil carbon sequestration in an Andisol in northern Japan.

Conclusions
The aim of our study was to assess the viability of intercropping systems between melon and cowpea combined with reduced tillage and the addition of crop residues as an agroecological alternative strategy for improving crop yield, crop quality and soil physicochemical properties while reducing the use of fertilizers. Our results have shown that the alternative system for melon growth based on intercropping with cowpea, reduced tillage and addition of crop residues can be an agroecological viable alternative for sustainable agriculture, with a reduction of 30% in fertilizers rate application. Total production, melon crop yield, soil Nt, P, K and some micronutrients increased under intercropping compared to melon monocrop. This reinforces the notion that introducing a legume species such as cowpea in intercropping systems with melon is an efficient approach in order to improve soil quality, enhance overall productivity and decrease the use of external inputs such as fertilizers and water. The interaction between both species may have stimulated beneficial microorganisms improving soil fertility and so promoting high crop yields. Thus, all crop and soil indicators used suggest that intercropping melon with cowpea is a more sustainable system than the sole cropping of melon, regardless of intercropping pattern. Nonetheless, the presence of two different species with different characteristics and needs can make harvest difficult. This is because crop density is higher owing to the presence of two species, and the harvest of one crop can damage the other crop if not carefully made. Hence intercropping should be adapted to each farm, with the need for training to properly implement these treatments.
Acknowledgements This research was funded by the Project Aso-ciaHortus granted by the Spanish Ministry of Science and Innovation (AGL2017-83975-R). M. Marcos-Pérez acknowledges the financial support from the Spanish Ministry of Science, Innovation and Universities through the "Ayudas para contratos predoctorales para la formación de doctores 2018" Program [PRE2018-085702]. We also acknowledge the technicians of the Tomás Ferro Experimental Farm of the Universidad Politécnica de Cartagena for their support in sowing/planting, monitoring the crops and harvest.
Authors' contributions Raúl Zornoza, Silvia Martinez-Martinez and Virginia Sánchez-Navarro contributed to the study conception and design. Material preparation, data collection and analysis were performed by Mariano Marcos-Pérez, Raúl Zornoza, Virginia Sánchez-Navarro, Silvia Martinez-Martinez, Eloísa García and María Martínez-Mena. The first draft of the manuscript was written by Mariano Marcos-Pérez, Raúl Zornoza commented on previous versions of the manuscript. All authors read and approved the final manuscript.
Funding Open Access funding provided thanks to the CRUE-CSIC agreement with Springer Nature. This research was funded by the Project AsociaHortus granted by the Spanish Ministry of Science and Innovation (AGL2017-83975-R). M. Marcos-Pérez received financial support from the Spanish Ministry of Science, Innovation and Universities through the "Ayudas para contratos predoctorales para la formación de doctores 2018" Program [PRE2018-085702] Data availability The manuscript has no associated data.

Consent for publication Not applicable
Conflicts of interest There are no competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.