Maize-legume intercropping promote N uptake through changing the root spatial distribution,legume nodulation capacity,and soil N availability

ZHENG Ben-chuan ,ZHOU Ying ,CHEN Ping ,ZHANG Xiao-na,DU Qing,YANG Huan,WANG Xiao-chun,YANG Feng,XlAO Te,Ll Long,YANG Wen-yu,YONG Tai-wen

1 College of Agronomy,Sichuan Agricultural University,Chengdu 611130,P.R.China

2 Sichuan Engineering Research Center for Crop Strip Intercropping System/Key Laboratory of Crop Ecophysiology and Farming System in Southwest,Ministry of Agriculture and Rural Affairs,Chengdu 611130,P.R.China

3 College of Resources and Environmental Sciences,China Agricultural University,Beijing100093,P.R.China

Abstract Legume cultivars affect N uptake,component crop growth,and soil physical and chemical characteristics in maizelegume intercropping systems. However,how belowground interactions mediate root growth,N fixation,and nodulation of different legumes to affect N uptake is still unclear. Hence,a two-year experiment was conducted with five planting patterns,i.e.,maize-soybean strip intercropping (IMS),maize-peanut strip intercropping (IMP),and corresponding monocultures (monoculture maize (MM),monoculture soybean (MS),and monoculture peanut (MP)),and two N application rates,i.e.,no N fertilizer (N-) and conventional N fertilizer (N+),to examine relationships between N uptake and root distribution of crops,legume nodulation and soil N availability. Results showed that the averaged N uptake per unit area of intercrops was significantly lower than the corresponding monocultures. Compared with the monoculture system,the N uptake of the intercropping systems increased by 31.7-45.4% in IMS and by 7.4-12.2% in IMP,respectively. The N uptake per plant of intercropped maize and soybean significantly increased by 61.6 and 31.8%,and that of intercropped peanuts significantly decreased by 46.6% compared with the corresponding monocultures.Maize and soybean showed asymmetrical distribution of roots in strip intercropping systems. The root length density(RLD) and root surface area density (RSAD) of intercropped maize and soybean were significantly greater than that of the corresponding monocultures. The roots of intercropped peanuts were confined,which resulted in decreased RLD and RSAD compared with the monoculture. The nodule number and nodule fresh weight of soybean were significantly greater in IMS than in MS,and those of peanut were significantly lower in IMP than in MP. The soil protease,urease,and nitrate reductase activities of maize and soybean were significantly greater in IMS and IMP than in the corresponding monoculture,while the enzyme activities of peanut were significantly lower in IMP than in MP. The soil available N of maize and soybean was significantly greater increased in IMS and IMP than in the corresponding monocultures,while that of IMP was significantly lower than in MP. In summary,the IMS system was more beneficial to N uptake than the IMP system. The intercropping of maize and legumes can promote the N uptake of maize,thus reducing the need for N application and improving agricultural sustainability.

Keywords:maize-legume strip intercropping,nitrogen uptake,soil enzyme activity,soil available nitrogen,root length density

1.lntroduction

With the growing world population,it is becoming a challenge to meet the increased need for food production with limited farmland,and sustainable approaches to achieve food security are urgently needed. Intercropping is the simultaneous cultivation of two or more crops during the same season on the same piece of land (Matt and Elizabeth 1993;Yanget al.2014;Liuet al.2020).Previous studies have confirmed that intercropping systems have yield advantages over single crops (Awalet al.2006;Beedyet al.2010),increasing land-use and water-use efficiency (Choudharyet al.2016;Rahmanet al.2016),improving land equivalent ratio (LER) (Gaoet al.2014;Yanget al.2015),and reducing fertilizer input (Liuet al.2014;Yonget al.2014). Therefore,intercropping provides an approach to achieve sustainable agricultural development. In recent years,legume and non-legume intercropping systems,such as wheatchickpea (Baniket al.2006),barley-pea (Chapagain and Riseman 2014),wheat-maize-soybean (Chenet al.2015),and common bean-maize (Latatiet al.2016) have become increasingly popular worldwide due to their high productivity and harvest of multiple grains. The yield and nutrient uptake advantages of intercropping systems are due to both aboveground and underground interspecific interactions between intercropped species (Liet al.2006).

Root interactions may play an important role in underground interspecific interactions (Xiaet al.2013).Yonget al.(2015) reported that both grain yield and aboveground N uptake of wheat and maize without root barriers were significantly higher than with root barriers.The root system is the main organ through which plants uptake and utilize water and nutrients (Zhanget al.2017).Crop growth and yield are affected by root proliferation(Fanet al.2016;Ramamoorthyet al.2017) and crops can increase soil nutrient uptakeviaroot proliferation in nutrient-enriched regions (Chilundoet al.2017). Root length density (RLD),root weight density (RWD) and root surface area density (RSAD) can be used to quantify the crop root extension and distribution (Renet al.2017;Liuet al.2020). Previous studies have indicated that intercropping can both promote and restrict root growth and modify root distribution (Renet al.2017;Liuet al.2020). In the maize-soybean strip intercropping system,the RLD of intercrops is increased by intercropping. The roots of intercropped maize have been shown to have an asymmetric distribution and the roots of intercropped soybean have been found to be restricted mainly to the zone near the plants (Gaoet al.2010). Similarly,fewer roots have been found in intercropped maize in the wheat-maize intercropping system than in the corresponding monocultures (Liet al.2006;Maet al.2019). In the maize-peanut intercropping system,the specific root length (SRL) of both intercropped maize and peanut was decreased (Gaoet al.2016). However,their RSAD was significantly increased. The RSAD of peanut was significantly higher in the neighboring row of maize and peanut than in the inter-row of peanut,showing an asymmetric distribution (Gaoet al.2016). The changes to root distribution caused by intercropping may affect the ability of crops to uptake the nutrients and water necessary to sustain plant growth and needs further study.

The yield and nutrient uptake advantages of intercropping systems are due to biological nitrogen fixation (BNF) of the legume component. BNF is essential for the development of a sustainable agriculture system,but it is usually inhibited by N fertilization (Fanet al.2006). Previous studies have shown that intercropping can alleviate the inhibitory effect of N application on nodulation and BNF by the legume. Liet al.(2009)reported that in the faba bean-maize intercropping system,nodule biomass and N derived from the atmosphere (Ndfa%) of intercropped faba bean were enhanced at different growth stages compared with the faba bean monoculture. Huet al.(2016) found that in the maize-peanut strip intercropping system,nodule biomass and Ndfa% of intercropped pea were significantly increased by 99 and 35% respectively,compared with the pea monoculture. Compared with the monoculture faba bean,however,intercropping did not significantly increase the Ndfa% either in the wheat-faba bean intercropping system or the maize-faba bean intercropping system (Fanet al.2006). The different BNF responses observed in these studies may be due to the interspecific interactions.Further studies are needed to fully understand nodulation and BNF responses in the different cereal-legume intercropping systems.

Maize,soybean,and peanut are major grains and oil crops worldwide,playing an important role in ensuring national food security. Previous studies have shown that there are differences in N-uptake in different maizelegume intercropping systems. Yuanet al.(2018) found N-uptake was increased in intercropped maize and decreased in intercropped peanut in a maize-peanut intercropping system compared to the monocultures.However,Gaoet al.(2020) found that N-uptake of both maize and peanut were decreased in a maize-peanut replacement strip intercropping system. Intercropping increased the N-uptake of both maize and soybean in an additive maize-soybean relay strip intercropping system compared to the monocultures (Fuet al.2019).Intercropping was found to improve soil nutrient supply capacity by enhancing the soil enzyme activities and soil bacteria abundance (Liet al.2018). The above studies focused on N-uptake,soil enzyme activity and the soil bacterial community independently and there is no systematic analysis of the inter-relationships between these processes. Intercropping can affect root spatial distribution and BNF (Fanet al.2006;Liet al.2006;Gaoet al.2010;Xiaet al.2013;Liuet al.2017,2020;Yonget al.2018b),and these effects may be important for N uptake. Further,the effects and functions of different legumes differ in the maize-legume intercropping system.The reasons for different N uptake levels between maize-soybean strip intercropping and maize-peanut strip intercropping,and the importance of underground interactions contributing to these differences still need further clarification. We hypothesized that:(i) differences in root spatial distribution and root characteristics between the different maize-legume strip intercropping systems contributed to differences in crop N-uptake;(ii) crop root growth regulated nodulation capacity of legumes and rhizosphere soil N availability to improve soil N availability in maize-legume strip intercropping systems. We investigated crop root spatial distribution,legume nodules and the soil available N levels in two maize-legume intercropping systems. The objectives of the present study were to:(i) examine and compare the crop root distribution and root characteristics in different maizelegume strip intercropping systems;(ii) explore legume nodulation in response to change in root distribution;and(iii) determine the rhizosphere soil enzyme activity and non-rhizosphere soil available N levels in different maizelegumes strip intercropping systems,and elucidate the possible N uptake mechanisms involved.

2.Materials and methods

2.1.Experimental site

The experiment was performed in Renshou County(30°16′N,104°00′E),Sichuan Province,Southwest China,during the 2017 and 2018 cropping seasons (from April to November). The climate of the experimental site is humid subtropical,with an annual temperature of 17.4°C and annual precipitation of 1 009 mm (Appendix A). The soil is an anthrosol with a clay loam texture,and the chemical characteristics of the topsoil are as follows:14.19 g kg-1of organic matter,1.22 g kg-1of total N,1.95 g kg-1of total P,26.06 g kg-1of total K,and average pH of 8.18.

2.2.Experimental design and crop management

A two-factor split-plot experimental design was performed with three replicates. The main factor was N application rate,i.e.,no N fertilizer (N-) and conventional N fertilizer (N+),and the sub-factor was planting pattern,i.e.,monoculture maize (MM),monoculture soybean(MS),monoculture peanut (MP),maize-soybean strip intercropping (IMS),and maize-peanut strip intercropping(IMP). The plot size was 5.8 m×6.0 m. The density of maize was 100 000 plants ha-1for MM,with the row spacing and inter-plant spacing of 0.5 and 0.2 m,respectively. The density of the legume was 200 000 plants ha-1for both MS and MP,with row and inter-plant spacings of 0.5 and 0.1 m,respectively. Two rows of maize were replaced with two rows of legumes in the two maize-legume strip intercropping systems. The density of intercropped maize,i.e.,maize in IMS (IMS/M) and maize in IMP (IMP/M) was half that of maize in MM at 50 000 plants ha-1,and the density of intercropped soybean (IMS/S) and peanut (IMP/P) was half of the corresponding monocultures at 100 000 plants ha-1(Fig.1).

Fig.1 Diagram showing the arrangement of planting patterns and root sampling sites in the field experiments. A,monoculture maize.B,monoculture soybean. C,monoculture peanut. D,maize-soybean strip intercropping. E,maize-peanut strip intercropping.

The conventional N rate was 240 kg N ha-1for MM and 80 kg N ha-1for MS and MP in the monocropping systems. In intercropping systems,the amount of N applied depends on the proportions of the crops planted,compared to the corresponding monocultures. The total N rate was 160 kg N ha-1for the intercropping systems,including 120 kg N ha-1for IM and 40 kg N ha-1for IS or IP. The N fertilizer for maize was divided into two applications. The amount of base fertilizer for all maize was 80 kg N ha?1,and the topdressing for MM was 160 kg N ha?1while that for intercropped maize was 40 kg N ha-1.The N fertilizer for soybean and peanut was applied in a single application. Basal P and K fertilizers were applied to all planting patterns at rates of 120 kg P2O5ha-1and 100 kg K2O ha-1.

Maize (ZeamaysL.cv.Xianyu 335) was sown manually on Apr 8,2017,and Apr 5,2018,and harvested on Aug 4,2017,and Aug 1,2018. Soybean (GlycinemaxL.Merr.cv.Nandou 12) was sown manually on Jun 9,2017,and Jun 5,2018,and harvested on Nov 1,2017,and Nov 5,2018. Peanut (ArachishypogaeaL.cv.Tianfu 18) was sown manually on Apr 7,2017,and Apr 7,2018,and harvested on Sept 13,2017,and Sept 10,2018.

2.3.Plant N content and N uptake

In the 2017 and 2018 field experiments,four maize and six soybean and peanut plants from each monoculture and intercropped plot were collected at each harvest stage. Plant samples were dried first at 105°C for 30 min and then at 75°C to constant weight. They were then ground and passed through a 60-mesh (0.25 mm)stainless steel sieve. The N content was determined with a Cleverchem Anna Random Access Analyser instrument(DeChem-Tech.GmbH,Hamburg,Germany). The plant N content was measured by the sulfuric acid-sodium salicylate method (Liuet al.2020).

The N uptake of intercropping relative to the monoculture system (ΔNU) was calculated as follows (Fanet al.2020):

whereNUicis the total N uptake of maize and legume in the intercropping systems.NUmmandNUmlare the N uptakes of monoculture maize and monoculture legume,respectively.FmandFlare the ratios of maize and legume plants in intercropping systems,respectively. The positive or negative value of ΔNU represents the increase or decrease in the amount of N uptake in intercropping relative to the monoculture crop,respectively.

2.4.Root growth

Root samples in intercropped and monoculture plots were collected at the silking stage (Jun 17,2017;Jun 13,2018) for maize,the full-bloom stage (Aug 4,2017;Aug 1,2018) for soybean,and the beginning of flowering stage (May 28,2017;May 30,2018) for peanut in 2017 and 2018 using an auger to minimize damage to the plots (Xiaet al.2013). Soil cores (10 cm in diameter)were collected at 20-cm intervals to a maximum depth of 100 cm to determine vertical root distribution. To determine the spatial distribution of roots as affected by planting pattern,soil cores were collected from under the row center below plants and 25 cm away from the center towards the nearest row in a single direction from the maize (or soybean or peanut) monoculture (Fig.1). Three soil cores were sampled in monocultures (Fig.1-A and B) and five soil cores were sampled in the intercropping systems (Fig.1-C).Soil cores were stored in plastic bags until the roots could be recovered. All soil cores were shaken vigorously and continuously,and passed through a sieve (0.2-mm mesh,20 cm in diameter and 5 cm in height) until the roots of maize and other crops could be distinguished by their different colors,textures,and branching patterns,and recovered using very fine tweezers. The roots of maize,soybean,and peanut were white,off-white,and brown,respectively. Separated root fractions were scanned with an Epson expression 10000XL scanner and then analyzed using the Win-RHIZOTMProgram to provide total root length and total root surface area data for each crop.

Root length density (RLD) is the root length per unit soil volume (cm cm-3),which was calculated using the following formula:

where L is the root length (cm) and V is the volume of the soil sample (1 570 cm3).

Root surface area density (RSAD) is the root surface area per unit soil volume (cm cm-3) which was calculated using the following formula:

where S is the root surface area (cm2) and V is the volume of the soil sample (1 570 cm3).

RLD contours were prepared using the Surfer ver.8.0 Program. RSAD stacked columns were prepared using the OriginPro 2017 Program.

2.5.Legume nodulation capacity

The nodulation capacity of legumes was assessed by measuring the nodule number and fresh weight. In each replicate,four representative plants were dug out to determine nodule number and weight according to the study of Yonget al.(2018a). Soybean nodules were collected at the full bloom stage (Aug 4,2017;Aug 1,2018),the full pod stage (Sept 16,2017;Sept 14,2018),and the full seed stage (Oct 9,2017;Oct 10,2018). Peanut nodules were collected at the beginning of the flowering stage (May 28,2017;May 30,2018) and the beginning of the seed formation stage (Jul 2,2017;Jul 7,2018).

2.6.Soil enzyme activity and soil available N

In 2018,rhizosphere soil samples (0-20 cm depth)were collected at the silking stage for maize,at the fullbloom stage for soybean,and at the beginning of the flowering stage for peanut to investigate soil enzyme activity. The roots were carefully extracted from the soil cores and shaken gently to remove non-rhizosphere soil.The rhizosphere soil,which was closely attached to the roots,was carefully brushed down and collected,and then sieved through a 20-mesh (0.0841 mm) stainless steel sieve. The soil samples were stored at -80°C prior to soil enzyme activity analysis. Soil protease,urease,and nitrate reductase activity kits (ELISA,Nanjing Jiancheng Bioengineering Institute,China) were used to measure enzyme activity. Non-rhizosphere soil samples were collected at the milk stage and full-maturity stage for maize,at the full-pod stage and full-maturity stage for soybean,and at the beginning of pod stage and fullmaturity stage for peanut to determine soil available N. Roots were carefully excavated from the soil and gently shaken to remove loosely attached soil which was designated non-rhizosphere soil. Soil samples were air-dried and passed through a 100-mesh (0.147 mm)sieve for determination of available N. The soil available nitrogen was determined by the alkali hydrolyzable method (Liet al.2018).

2.7.Statistical analysis

The two-way ANOVA analysis was performed to determine if there were significant differences among treatment means and interactions,and Fisher’s least significant difference test (LSD,α=0.05) was used to determine significant differences between individual treatment means. All analyses were performed with SPSS v.22 and Microsoft Excel. SigmaPlot14.0,Origin 2017,and Surfer ver.8.0 were used to draw the figures.

3.Results

3.1.N uptake

Significant differences in N uptake per unit area of crop were observed under different N application rates and planting patterns (Table 1). Compared with the monocultures,N uptake per unit area of the intercrops was significantly reduced. Compared with the MM,the averaged N uptake per unit area of IMS/M decreased by 18.6% under N-and 24.8% under N+,and that of IMP/M decreased by 11.3% under N-and 22.0% under N+,respectively.Compared with the MS,the N uptake per unit area of IMS/S significantly decreased by 30.4% under N-and 37.8%under N+,respectively. Similarly,the N uptake per unit area of IMP/P significantly decreased by 75.5% under N-and 71.2% under N+compared with the MP,respectively. We found that the N uptake per plant of maize and soybean was significantly greater in intercropping systems than in the corresponding monocultures. However,the N uptake per plant of IMP/P significantly decreased compared with MP (Appendix B). As shown in Table 2,in IMS system N uptake in the N-and N+treatments was 385.2 and 468.3 kg ha-1,respectively,and in the monoculture crops was 266.0 and 356.7 kg ha-1respectively (average of the years 2017 and 2018). The N uptake of IMS was 45.4 and 31.7% greater in the N-and N+treatments,respectively,than in the monoculture crops. In the IMP system,differences in N uptake were observed between the two cropping seasons. In 2017,N uptake was lower by 7.2% in the N-treatment and by 7.9% in the N+treatment than in the monocultures,but greater by 31.6 and by 22.7% in the N-and N+treatments respectively in 2018.

Table 1 Effects of N application rate and planting pattern on N uptake per unit area of crop (kg ha-1)

Table 2 Nitrogen uptake in maize-legume strip intercropping systems and monoculture systems at full-maturity stage

3.2.Root length density (RLD)

The RLD of MM (Fig.2) and MS (Fig.3) showed symmetrical distribution in both the N-and N+treatments.However,asymmetric distribution of roots was observed in IMS/M,IMP/M,and IMS/S (Figs.2 and 3). The maize roots extended into legume (soybean or peanut) stand underneath the space and even between the inner-rows of legume in IMS/M and IMP/M (Fig.2). When intercropped with maize,soybean roots were biased towards maize(Fig.3). However,peanut roots were confined and only distributed under the peanut plant space both in MP and IMP/P (Fig.4).

In the two-year field experiments,the RLD of crops decreased with soil depth under both monoculture and intercropping (Figs.2-4). Higher RLD was observed in the top layers of soil (0-20 cm) for all crops (Figs.2-4).The RLD of maize declined in the treatment order IMP/M＞IMS/M＞MM at most soil depths. The maize roots were vertically distributed at the 0-60 cm soil layers (Fig.2).The RLD of intercropped maize,recorded at interrows between the maize and legumes (-25 cm),was greater than that for the maize inner-row (25 cm) inboth N treatments (Fig.2). RLD was higher in IMP/M than in IMS/M at the legume planting point under the N+treatment (Fig.2). The RLD of soybean was greater in IMS/S than MS. The soybean roots were vertically distributed at the 0-40 cm soil layer (Fig.3). The RLD of intercropped soybean at the inter-row position between maize and soybean (25 cm) was higher than that at the soybean inner-row (-25 cm) in both N treatments (Fig.3).The RLD of peanut was greater in IMP/P than in MP at all soil depths. The peanut roots were mainly distributed at the 0-20 cm soil layer (Fig.4).

Fig.2 Spatial root length density (cm cm-3) distribution of maize at the flowering stage. MM,monoculture maize;IMS/M,intercropped maize with soybean;IMP/M,intercropped maize with peanut. N-,no N fertilizer;N+,conventional N fertilizer. The color scale shows the value of root length density (cm cm-3). Plants in green are those whose root distributions are shown in the corresponding panels and plants in gray show the positions of the co-cropped species.

Fig.3 Spatial root length density (cm cm-3) distribution of soybean at the flowering stage. MS,Monoculture soybean;IMS/S,intercropped soybean with maize. N-,no N fertilizer;N+,conventional N fertilizer. The color scale shows the value of root length density (cm cm-3 soil volume). Plants in green are those whose root distributions are shown in the corresponding panels and plants in gray show the positions of the co-cropped species.

Fig.4 Spatial root length density (cm cm-3) distribution of peanut at the flowering stage. MP,monoculture peanut;IMP/P,intercropped peanut with maize. N-,no N fertilizer;N+,conventional N fertilizer. The color scale shows the value of root length density (cm cm-3 soil volume). Plants in green are those whose root distributions are shown in the corresponding panels and plants in gray show the positions of the co-cropped species.

3.3.Root surface area density (RSAD)

At most soil depths,the RSAD of maize was greater in IMS/M and IMP/M than that in MM in both the N-and N+treatments (Fig.5). The total RSAD of maize in the maize row (0 cm) was significantly greater by 21.5% (average of two years) in IMS/M and by 24.9% (average of two years)in IMP/M than in MM (Fig.5). The total RSAD of maize in the maize row (0 cm) was greater by 10.8% (average of two years) in IMP/M than in IMS/M in the N+treatment(Fig.5),while it was less by 7.6% (average of two years)in IMP/M than in IMS/M in the N-treatment (Fig.5). The total RSAD of IMS/S in the soybean row (0 cm) was greater by 15.9 and 17.9% in the N-and N+treatments,respectively (average of two years),than in MS (Fig.6).Further,the RSAD of IMS/S roots at the inter-row (25 cm)was higher by 116.7% (average of two years) than that at the location between the soybean and maize rows (Fig.6).Regarding peanut,the total RSAD was significantly lower by 18.6% (average of two years) in IMP/P than in MP(Fig.7).

Fig.5 Root surface area density (cm2 cm-3) distribution of maize at the flowering stage. MM,monoculture maize;IMS/M,intercropped maize with soybean;IMP/M,intercropped maize with peanut. N-,no N fertilizer;N+,conventional N fertilizer. The color scale shows the soil depth. Error bar is SD value (n=3).

Fig.6 Root surface area density (cm2 cm-3) distribution of soybean at the flowering stage. MS,monoculture soybean;IMS/S,intercropped soybean with maize. N-,no N fertilizer;N+,conventional N fertilizer. The color scale is the value of soil depth. Error bar is SD value (n=3).

Fig.7 Root surface area density (cm2 cm-3) distribution of peanut at the flowering stage. MP,monoculture peanut;IMP/P,intercropped peanut with maize;N-,no N fertilizer;N+,conventional N fertilizer. The color scale is the value of soil depth. Error bar is SD value (n=3).

3.4.Root nodules

Intercropping and N application significantly affected the number and fresh weight of legume root nodules at the different growth stages (Tables 3 and 4). The averaged number of nodules,over both years,in IMS/S was significantly greater than in MS,by 34.4% at the full bloom stage,63.9% at the full pod stage,and 170.9%at the full seed stage in the N-treatment,and by 50.6%at the full bloom stage,63.2% at the full pod stage,and 120.0% at the full seed stage in the N+treatment(Table 3). Similarly,the fresh weight of nodules in IMS/S was significantly greater by 74.7% at the full bloom stage,71.4% at the full pod stage,and 156.1% at the full seed stage in the N-treatment,and notably by 81.3%at the full bloom stage,80.6% at the full pod stage,and 99.0% at the full seed stage in the N+treatment,than in MS (Table 4). However,the number and fresh weight of nodules decreased in IMP/P compared MP (Tables 3 and 4). The number of nodules in IMP/P was significantly lower by 33.4 and 39.0% in the N-and N+treatments respectively at the flowering stage in 2017,while the differences in fresh weight between IMP/P and MP were not significant. At the seed-setting stage of peanut,the number of nodules in IMP was significantly lower by 33.0 and 39.0% (average of two years) in the N-and N+treatments respectively,than in MP (Table 3). The fresh weight of root nodules in IMP/P was significantly lower than in MP by 33.7 and 56.4% (average of two years) in the N-and N+treatments,respectively (Table 4).

3.5.Soil enzymes activities

Soil enzyme activities were significantly greater in IMS/M,IMP/M (Fig.8-A,D,and G),and IMS/S (Fig.8-B,E,and H) than in the corresponding monocultures,but were significantly lower in IMP/P (Fig.8-C,F,and I). Soil protease,urease,and nitrate reductase activities of maize were significantly greater by 7.8,12.4,and 6.3% in IMS/M and by 15.6,14.7,and 9.5% in IMP/M than in MM in the N-treatment,respectively. Regarding the N+treatment,the soil protease,urease,and nitrate reductase activities of maize were significantly greater by respectively 14.9,30.8,and 2.8% in IMS/M and 17.6,61.7,and 5.6% in IMP/M than in MM (Fig.8-A,D,and G). The protease activity of IMP/M was significantly greater by 7.3% than that of IMS/M in the N-treatment (Fig.8-A). The urease activity of IMP/M was significantly greater by 23.6%than in IMS/M (Fig.8-D). Compared with corresponding monocultures,the soil protease,urease,and nitrate reductase activities of IMS/S was significantly enhanced by 20.5,37.6,and 6.5% in the N-treatment,and by 6.0,17.1,and 8.3% in the N+treatment (Fig.8-B,E,and H),while their activities in IMP/P were lower by 13.9,26.5,and 2.9% respectively in the N-treatment,and by 5.1,15.5,and 7.6% respectively in the N+treatment (Fig.8-C,F,and I).

Fig.8 Effect of planting pattern and N application on soil enzyme activities at the flowering stage in the 2018 cropping season.MM,monoculture maize;MS,monoculture soybean;MP,monoculture peanut;IMS/M,intercropped maize with soybean;IMP/M,intercropped maize with peanut;IMS/S,intercropped soybean with maize;IMP/P,intercropped peanut with maize. N-,no N fertilizer;N+,conventional N fertilizer. Different lower-case letters indicate significant differences under different planting patterns(LSD,P＜0.05). The asterisk (＊) and (＊＊),and (ns) indicate significant difference (P＜0.05),highly significant difference (P＜0.01),and no significant difference (P＞0.05) between different N application rates. Error bar is SD value (n=3).

3.6.Soil available N

The differences in soil available N between the intercropping systems and corresponding monocultures were significant at different growth stages (Fig.9).The soil available N of IMS/M was significantly greater than MM by 28.5% at the milk stage and by 16.8% at the full-maturity stage,and that of IMP/M was significantly greater than MM by 37.5% at the milk stage and by 28.8% at the full-maturity stage in the N+treatment (Fig.9-A and B). The soil available N of IMP/M was 7.0 and 10.3% greater in the N-and N+treatments respectively than with IMS/M (Fig.9-A and B). The soil available N of IMS/S was greater than MS by 6.3% at the full-pod stage and 11.8% at the full-maturity stage in the N+treatment (Fig.9-C and D). The soil available N content of IMP/P was significantly lower by 14.0% at the beginning flowering stage and 18.3% at the full-maturity stage than with the MP (Fig.9-E and F).

Fig.9 Effect of planting pattern and N application on soil available nitrogen at the different growth stages in 2018 cropping season.MM,monoculture maize;MS,monoculture soybean;MP,monoculture peanut;IMS/M,intercropped maize with soybean;IMP/M,intercropped maize with peanut;IMS/S,intercropped soybean with maize;IMP/P,intercropped peanut with maize. N-,no N fertilizer;N+,conventional N fertilizer. Different lower-case letters indicate significant differences under different planting patterns(LSD,P＜0.05). The asterisk (＊) and (＊＊),and (ns) indicate significant difference (P＜0.05),highly significant difference (P＜0.01),and no significant difference (P＞0.05) between different N application rates. Error bar is SD value (n=3).

4.Discussion

Numerous studies have reported that intercropping enhances crop N uptake,i.e.,a N uptake by a nonlegume crop is increased when intercropped with a legume (Waghmaref and Singh 1984;Stern 1993;Yonget al.2015;Malungaet al.2017). In the present study the averaged N uptake per unit area of maize was significantly lower in IMS and in IMP than in the corresponding monocultures. The averaged N uptake per unit area of soybean and peanut were also significantly lower. The possible reason for the decrease in N uptake per unit area of intercrops was the lower crop density. In our present study,the density of intercrops was half of the corresponding monocultures. Compared with the corresponding monocultures,the N uptake per plant of maize and soybean significantly increased in intercropping,while that of peanut decreased in intercropping(Appendix B). However,the ΔNU was different.The ΔNU in IMS and in IMP was greater than in the corresponding monocultures,and the N uptake advantages were greater in IMS than in IMP.

4.1.Effects of root spatial distribution on N uptake

The N uptake of maize per plant was significantly greater when intercropped with soybean and peanut than when grown alone. Compared with the corresponding monocultures,the N uptake per plant of intercropped soybean was significantly enhanced,while that of intercropped peanut was significantly reduced. The different N uptakes result from interspecific facilitation by shaping root growth and root spatial distribution (Gaoet al.2010;Renet al.2017;Liuet al.2020). The root system facilitates resource absorption and utilization (Zhanget al.2017).In intercropping systems,the interspecific competitive use of nutrients regulates root spatial distribution to address the availability of soil nutrients (Yuet al.2014;Yonget al.2015). Previous studies have shown that intercropping can change the spatial distribution of crop roots (Gaoet al.2010;Liuet al.2020). In this study,the roots of maize and soybean showed an asymmetric distribution in the intercropping system compared with the corresponding monocultures.Maize roots extended underneath legume plants and even between the inner rows of both soybean and peanut. Similarly,the roots of soybean proliferated toward maize rows and occupied the soil layers in maize-soybean intercropping system at the full-bloom stage of soybean. The changed root distribution enhanced the root absorption range of maize and soybean in the maize-soybean system. Further,the RSAD was higher in the top layer (0-20 cm) than in lower soil layers. The RSAD of maize and soybean increased when intercropped compared with the corresponding monocultures (Figs.5 and 6),which provided a larger absorption area for nutrients for the component crops. A well-developed fine root system and optimized root distribution can improve nutrient uptake by crops (Liuet al.2020). The optimized root distribution helps to efficiently use the soil resource,and a higher root surface area increases the efficiency of acquiring nutrients. This is probably the reason for the efficient N use in the maize-soybean system.In contrast,peanut roots were restricted to the zone near the plants in the maize-peanut system (Fig.4).Although the RLD of peanut increased in the maizepeanut system,the RSAD of peanut was significantly reduced in comparison with peanut in monoculture(Fig.7). These results suggest that there were less fine roots in the maize-peanut system than where peanut was grown alone,resulting in a smaller surface area for nutrient acquisition and adverse effects on N uptake. Thus,N uptake per plant of peanut was significantly reduced in the maize-peanut system compared with the peanut monoculture (Appendix B).The differences in root growth may be due to the interactive effect between aboveground growth and root proliferation (Xiaet al.2013). A good root system promotes aboveground growth by providing sufficient nutrients (Liuet al.2020). However,root growth,nutrient acquisition and transport consume energy.The carbohydrates needed for root growth,nutrient acquisition and transport are mainly provided by the aboveground parts of the plant (Philipson 1988;G?ttlicheret al.2010). In the present study,the aboveground biomass of maize was significantly higher in the maizelegume systems than in monoculture (Appendix C). Thus maize in intercropping accumulated more photosynthetic product than in monoculture,which could provide more carbohydrates to the roots and promote greater root growth and nutrient acquisition. After maize harvest,biomass accumulation in soybean when intercropped with maize was greater than when grown alone (Appendix C). In contrast,peanut biomass accumulation was significantly lower in the intercropping system (Appendix C) and the reduced photosynthetic product accumulation would have disadvantaged root growth and soil nutrient acquisition in comparison with peanut in monoculture (Wanget al.2013).In summary,good growth of aboveground promotes root growth by providing sufficient energy. Furthermore,the asymmetric distribution of roots contributed to the roots occupying more soil layers in intercropping than in the corresponding monocultures,and increasing RLD and RSAD strengthened nutrient acquisition ability of maize and soybean in intercropping,thereby promoting crop N uptake.

4.2.Effects of legume nodulation on N uptake

Cereal-legume intercropping can increase N input through BNF and promote N transfer from legume to cereal(Senaratneet al.1993;Yonget al.2013;Liuet al.2017),and almost half of the N demand in the lifespan for legume growth can be met by BNF (Salvagiottiet al.2008). In the cereal-legume intercropping system,the competitive N use of cereal can decrease soil N and increase nodule number and dry weight (Liet al.2001). Besides,the rhizosphere exudates of cereal can promote legume nodulation and strengthen the BNF ability (Liet al.2016).In the present study,the nodule number and nodule fresh weight significantly increased by 34.4-170.9%and 74.7-156.1% in IMS/S than that in MS (Tables 3 and 4). This implied that intercropped with maize not only increases the number of soybean root nodules but also increases the size of root nodules. Thus,it was beneficial to improve the N fixation ability of intercropped soybean.The nodule number quickly decreased at the full-seed stage as compared to that at the full-pod stage in MS,while that of IMS/S slowly decreased at the full-seed stage as compared to that at the full-pod stage (Table 3). This implied that maize-soybean strip intercropping delayed soybean nodules senescence to enhance the ability of continuous N fixation of soybean. Namely,parts of the fixed N were transferred for maize uptake,and the other fixed N was used for soybean growth (Yonget al.2015).On the contrary,the number and fresh weight of nodules significantly decreased by 13.8-39.0% and 13.0-56.4%in IMP/P compared with MP (Tables 3 and 4). Probably because that symbiosis is a high energy-consuming activity,in which photosynthates are used as an energy source to drive processes (Tjepkema and Winship 1980).The root of intercropped peanuts was suppressed by maize during the long coexistence period compared with MP (Figs.4 and 7),which led to insufficient nutrients and water acquisition to peanut growth demand. Therefore,peanut biomass significantly decreased in IMP compared with MP (Appendix C). Indeed,strong competitive use of light and nutrients have adverse effects on peanut growth in maize-peanut intercropping (Awalet al.2006),because of that heavy shading decreasing the photosynthetic product accumulation (Pons and Pearcy 1994;Callan and Kennedy 1995). This is the probable reason why the nodule number,nodule weight,and N uptake of peanut decreased in IMP.

4.3.Effects of soil N availability on N uptake

Soil available N of maize and soybean was significantly increased in intercropping system,compared with the corresponding monocultures (Fig.9-A-D). Previous studies have documented that intercropping can enhance the availability of soil nutrients (Makindeet al.2007;Maet al.2017). On the one hand,root exudates directly improve nutrient availability (Baiset al.2006). On the other hand,root exudates regulate soil microbial activity to accelerate the transformation of mineral nutrients,because root exudates are the carbon source of soil microorganisms (Baudoinet al.2003;Badri and Vivanco 2009). Soil enzymes are intimately involved in the cycling of nutrients (Lalandeet al.2000),and the influence of intercropping on soil enzymes cannot be ignored (Zhouet al.2011;Maet al.2017;Chenet al.2018). In the present study,the rhizosphere soil enzyme activities of protease,urease,and nitrate reductase in IMS/M and IMP/M were significantly enhanced compared with those of MM (Fig.8-A,D,and G),promoting the maize soil N cycle and enhancing soil N availability. Soil available N was increased by 16.8-28.5% in IMS/M and 28.8-37.5%in IMP/M compared with that of MM (Fig.9-A and B).Enzyme activities were higher in IMP/M than those in IMS/M (Fig.8-A,D,and G),which is probably the reason for soil available N being greater in the maize-peanut system than in the maize soybean system. The soil available N was significantly enhanced when soybean was intercropped with maize compared for MS. In contrast,soil available N was significantly reduced when peanut was intercropped with maize compared with peanut grown alone. This difference may have been due to the distribution of maize roots (Fig.2). The asymmetric distribution of maize roots intensified the interspecific N competition between maize and peanut,which resulted in decreased soil available N to peanut. Furthermore,the soil enzyme activities of protease,urease,and nitrate reductase were significantly decreased in IMP/P compared with that of MP (Fig.8-C,F,and I). This would have contributed to reduced soil available N in IMP/P. Therefore,soil N supply of intercropped peanuts was insufficient and plant N uptake was significantly decreased.

5.Conclusion

There are significant differences in N uptake of crops under different maize-legume strip intercropping systems.Compared with the corresponding monocultures,the N uptake per unit area of intercrops was considerably decreased. In contrast,in the present study,the N uptake of the intercropping system was increased by 31.7-45.4%in IMS and 7.4-12.2% in IMP,respectively,compared with the corresponding monocultures. Compared to the monocultures,the roots of maize and soybean in the intercropping system had an asymmetrical distribution,and increased RLD and RSAD. However,the roots of intercropped peanut were confined,which resulted in decreased RLD and RSAD when compared with peanut monoculture. In the intercropping systems,nodulation capacity was stronger in soybean than in peanut.Furthermore,soil enzyme activities (i.e.,protease,urease,and nitrate reductase) and soil available N of maize and soybean were significantly greater in intercropping system than in the corresponding monocultures,whereas they were significantly lower in intercropped peanut than in the monocultures. In conclusion,strip intercropping with legumes can increase the N uptake capacity of maize by improving root spatial distribution soil enzyme activity and soil N availability.

Acknowledgements

The research was supported by the National Natural Science Foundation of China (31872856) and the National Key Research and Development Program of China(2016YFD030020205).

Declaration of competing interest

The authors declare that they have no conflict of interest.

Appendicesassociated with this paper are available on http://www.ChinaAgriSci.com/V2/En/appendix.htm