INTRODUCTION
Cultivable area for food production in the EHR of India is declining and at the same time, burgeoning population growth is exerting tremendous pressure on limited land resources in order to produce more foods for meeting the food requirement of its inhabitants (Choudhary et al., Reference Choudhary, Anil, Kumar and Chauhan2014; Munda et al., Reference Munda, Das and Patel2009). Marginal input intensive rainfed agriculture of rice-fallow followed by maize-fallow dominates the food grain production systems in the region. The majority of farmers are marginal to small in land holdings (<0.5 ha), resource poor and marginal in farm mechanization. They grow rainfed maize during the rainy season (May–August) and crops are frequently severely infested with wide a range of weeds, subjecting crops to heavy weed competition for space, nutrients, water and resulting in low productivity (Shah et al., Reference Shah, Shroff, Patel and Usadadiya2011).
Maize is a heavy nutrient feeder crop and repeated mono-cropping of it over several years without periodic soil nutrient replenishment has caused decline in both soil fertility and productivity in the uplands of EHR of India. As a result, maize-based rainfed agriculture production system was unable to meet the diversified domestic needs of small farmers from the dwindling supply of new lands for cultivation and other limited resources in the region (Choudhary et al., Reference Choudhary, Kumar and Bhagawati2012; Ullah et al., Reference Ullah, Bhatti, Gurmani and Imran2007). These conditions necessitate a shift from mono/sole cropping to multiple cropping such as intercropping since it is considered an excellent strategy for crop intensification, utilization of light and other growth resources (Agegnehu et al., Reference Agegnehu, Ghizam and Sinebo2006; Lithourgidis et al., Reference Lithourgidis, Dordas, Damalas and Vlachostergios2011), improvement of soil fertility (Vesterager et al., Reference Vestrager, Nielsen and Jensen2008), absorbing excess labour (Marer et al., Reference Marer, Lingaraju and Shashidhara2007), increasing income (Chen et al., Reference Chen, Westcott, Neill, Wichman and Knox2004) and increasing production per unit land area and time (Agegnehu et al., Reference Agegnehu, Ghizam and Sinebo2006; Shah et al., Reference Shah, Shroff, Patel and Usadadiya2011). Some farmers in the region, however, have been traditionally practicing mixed cropping of as many as 35 crops involving cereals, legumes, vegetables and tuber crops in small patches to fulfil their household requirements for food and nutritional security (Murtem et al., Reference Murtem, Sinha and Dopum2008). Despite very high soil organic carbon content (>1%) in majority of the soils (Choudhury et al., Reference Choudhury, Mohapatra, Das, Das, Nongkhlaw, Fiyaz, Ngachan, Hazarika, Rajkhowa and Munda2013), the productivity of these crops is very low. The underlying cause for low productivity is due to incompatibility in crop type combinations and lack of appropriate row spacing/planting density. This results in overcrowding, which leads to severe competition for available resources including nutrients, water and light intensity among the component crops (Choudhary et al., Reference Choudhary, Kumar and Bhagawati2012). As a result, food and nutritional security still remains a challenge in the region (NEH, 2005).
Wider row spacing in maize can be used to grow short duration legumes as intercrops for sustaining soil health while achieving higher productivity (Singh et al., Reference Singh, Saad and Singh2008). Intercropping with legumes offers better soil coverage, smothers weeds, reduces nutrient leaching and fixes atmospheric nitrogen for its own use as well as use by companion crops (Fan et al., Reference Fan, Zhang, Song, Sun, Bao, Guo and Li2006; Lithourgidis et al., Reference Lithourgidis, Dordas, Damalas and Vlachostergios2011). Legumes also help with solibulization and mobilization of P and K from their respective pools, which become available to crops in the system. Suitable intercropping provides a yield advantage over sole cropping, because of the complimentary utilization of natural resources by the component crops (Lithourgidis et al., Reference Lithourgidis, Dordas, Damalas and Vlachostergios2011). However, success of intercropping mostly depends on proper combination of component crop species, the appropriate seed mixtures, planting densities and crop management practices (Lithourgidis et al., Reference Lithourgidis, Dordas, Damalas and Vlachostergios2011). Although, a recent shift in research priorities has helped to develop strategies to reduce crop–weed competition in cereals including maize based intercropping systems in some parts of the world (Agegnehu et al., Reference Agegnehu, Ghizam and Sinebo2006; Fan et al., Reference Fan, Zhang, Song, Sun, Bao, Guo and Li2006; Lithourgidis et al., Reference Lithourgidis, Dordas, Damalas and Vlachostergios2011; Singh et al., Reference Singh, Saad and Singh2008), no such information is available in the EHR of India. As such information on appropriate crop species combination, particularly legumes on pre-dominantly maize-based intercropping systems (e.g. optimum planting density or row spacing in the uplands) holds immense importance in improving the productivity, food grain vis-à-vis nutritional security of the small farm holdings in the EHR of India. Therefore, in the present investigation, we evaluated maize-based intercropping with legumes (soybean and groundnut) at different planting densities with respect to crop (and system) yield advantage, soil enrichment of major nutrients-NPK, crop uptake and weed smothering effect of intercrops on system productivity in the rainfed hilly ecosystem in the EHR of India.
MATERIALS AND METHODS
Description of experimental site, climate and soil
The experiment was conducted during the rainy season (April–August) of 2010 and 2011 at the Research Farm (27°95′N latitude and 94°76′E longitude, 660 m above MSL) of ICAR Research Complex for NEH Region, Basar, Arunachal Pradesh, India. The experimental site falls under humid sub-tropical climate. The daily temperature varies widely between a minimum of 4 °C and a maximum of 35 °C. The experimental site received an average annual rainfall of 2400 mm and the monsoon months (June to September) contributed more than 77% of it (Figure 1). The soil of the experimental site was silty loam in texture (sand: 33.4%, silt: 51.2% and clay: 15.4%), acidic in reaction (pH 5.3), and high in organic carbon (Walkley and Black, 13.2 g kg−1), low in available nitrogen (alkaline permanganate N, 193.8 kg ha−1), low in available phosphorus (Bray P, 8.4 kg ha−1) and available K (neutral normal ammonium acetate K, 210.5 kg ha−1) as earlier described by Choudhary et al. (Reference Choudhary, Anil, Kumar and Chauhan2014). Soils were moderate in water retention capacity (18–21% at 0.3 bars and 9–11% at 15 bars) and moderate in compaction level (bulk density: 1.42 mg m−3)
Figure 1. Monthly maximum temperature and rainfall during experimental period (2010 and 2011) at study area.
Experimental design and treatment imposition
The experiment was laid out in a randomized block design and replicated thrice with gross plot size of 4.8 m × 4.0 m. In the uplands of EHR, maize is the principal crop, a natural choice for farmers of the region. However, with the effort of several farmers’ participatory programmes conducted by the ICAR Research Complex for NEH Region along with other line department, soybean and groundnut has been successfully introduced in the region and farmers are adopting it but in small scale and mostly as single crop (either soybean or groundnut alone) in limited areas without any proper cropping system arrangement (Choudhary et al., Reference Choudhary, Anil, Kumar and Chauhan2014). This prompted us to choice of maize as the main crop with an intercrop of soybean/groundnut. Row proportions of intercrops, however, were devised from the existing maize planting density used in the region and the available information from the literature for soybean and groundnut in main land India (Mandal et al., Reference Mandal, Banerjee, Banerjee, Alipatra and Malik2014; Prasad and Brook, Reference Prasad and Brook2005).
There were nine treatment combinations namely T1: solitary maize [Zea mays (L.) cv. Allrounder) planted 90 cm × 20 cm apart (~55 500 plants ha−1), T2: solitary soybean [Glycine max (L.) cv. JS 335] planted 30 cm × 10 cm apart (~3 33 300 plants ha−1), T3: solitary groundnut [Arachis hypogea (L.) cv. JCGS 76] planted 30 cm × 10 cm apart (~3 33 300 plants ha−1), T4: maize with soybean at 1:1 rows (100:30 proportion), T5: maize with soybean at 1:2 rows (100:60 proportion) in additive series, T6: Maize with soybean at 1:5 rows (60:77 proportion) in replacement series, T7: maize with groundnut at 1:1 rows (100:30 proportion), T8: maize with groundnut at 1:2 rows (100:60 proportion) in additive series and T9: maize with groundnut at 1:5 rows (60:77 proportion) in replacement series. Row to row spacing for maize on T4, T5, T7 and T8 was 90 cm while in T6 and T9 it was 180 cm. In both additive and replacement series, plant to plant spacing was 20 cm. Based on the proportion of each crop, the corresponding planting density of crops was maintained in the gross plot area. Recommended doses of fertilizers-NPK for maize (40N, 60P2O5 and 40K2O kg ha−1), soybean (25N, 60 P2O5 and 40 K2O kg ha−1) and groundnut (25N, 60 P2O5 and 40 K2O kg ha−1) in the form of urea (46% N), single super phosphate (16% P2O5) and murate of potash (60% K2O) were applied as basal doses at the time of crop sowing. Forty days after sowing (DAS), an additional 40 kg N was applied to the maize crop (solitary and maize in intercrop). In the intercropping combinations (T4–T9), fertilizers were applied according to the proportion of main and intercrop populations.
Observations
Grain (and pod) yield, production efficiency and maize equivalent yield
Grain yield of maize, soybean and groundnut were recorded from an area of 2.4 m × 3.0 m. Cobs and pods were harvested manually at physiological maturity. Grain yield was recorded at 14% moisture content. PE implies per day productivity of the whole system and was estimated as the ratio of yield and the days taken to reach the harvesting stage (crop duration). Since the study area is predominantly under rainfed agriculture, PE will provide additional information in choosing suitable cropping pattern for increasing temporal land productivity though crop intensification (Choudhary and Kumar, Reference Choudhary and Kumar2013). Biological yield is the sum of above ground biomass harvested during the cropping season (sum of grain and stover yield of maize and intercrop in the intercropping system). Yields of intercrops were converted into MEY considering the experimental years (2011–12) minimum support price (MSP) in Indian rupees (INR) for maize, soybean and groundnut. The MEY was calculated as follows:
MEY (kg ha−1) = Maize yield in intercropping system + [intercrop yield × market price of intercrop (INR kg−1)/price of maize (INR kg−1)]
Competition indices
Land equivalent ratio (LER) was estimated as the land required under monoculture to produce equal yield as that of intercropping system. It was calculated as suggested by Willey (Reference Willey1979):
$$\begin{equation*}
\rm LER = \left\{ {La + Lb} \right\}.
\end{equation*}$$
$$\begin{equation*}
\rm La = \left( {\frac{{Yab}}{{Yaa}}} \right)
\end{equation*}$$
$$\begin{equation*}
\rm La = \left( {\frac{{Yba}}{{Ybb}}} \right),
\end{equation*}$$
where La and Lb are the LERs for individual crops, Yab and Yba are the individual crop yields in intercropping and Yaa and Ybb are the individual crop yields in sole cropping. If LER was more than 1, then there was a yield advantage.
Relative crowding coefficient (K) is a measure of the relative dominance of one species over the other in intercropping. It was calculated as suggested by de Wit (Reference de Wit1960) and adopted by John and Mini (Reference John and Mini2005):
$$\begin{equation*}
\rm K = \left( {Ka \times Kb} \right).
\end{equation*}$$
$$\begin{equation*}
\rm Ka = \left( {\frac{{Yab}}{{Yaa - Yab}}} \right)
\end{equation*}$$
$$\begin{equation*}
\rm Kb = \left( {\frac{{Yba}}{{Ybb - Yba}}} \right),
\end{equation*}$$
where Ka and Kb are the relative crowding coefficients of maize and intercrops, respectively. When the product of the two coefficients (Ka × Kb) is greater than 1, there is a yield advantage, if the value of K is 1, then there is no yield advantage and if less than 1, there is a disadvantage in yield.
Monetary advantage index (MAI) evaluates the yield of all crops in different intercropping combinations as well as the sole cropping system and their economic return in terms of monetary value, including the total profitability of the land. The higher the index value, the more profitable the cropping system is. It is expressed as
$$\begin{equation*}
\rm MAI = \left( {Pab + Pba} \right)\times \frac{{LER - 1}}{{LER}}.
\end{equation*}$$
where, Pab = Pa × Yab; Pba = Pb × Yba; Pa = price of species ‘a’ and Pb = price of species ‘b’. The prevailing market price for the produce was INR 9000/ton for maize; INR 16 500/ton for soybean and INR 27 000/ton for groundnut.
Measurement of weed parameters
Weeds were measured in randomly selected plots from an area of 1.0 m2 at 60 DAS and segregated in to grasses, sedges and broadleaves. After counting, weed roots and above ground biomasses were separated, dried in an oven at 65 °C for 48 h. During drying, the total weed dry biomass was recorded until a constant weight was obtained. The data on weeds were subjected to square root transformation (√x+0.5) to normalize their distribution. Original weed values (grasses, sedges and broadleaves) of density and dry biomass are also presented in parenthesis for better clarity/comparison. WSE was calculated as follows:
WSE (%) = (Weed dry biomass of solitary maize − weed dry biomass of intercrop)/ Weed dry biomass of solitary maize
Nutrient uptake and removal by crops and weeds
Tissue nutrient contents (N, P & K) of weeds and crops were measured through standard laboratory procedures (Chapman and Pratta, Reference Chapman and Pratta1961; Richards, Reference Richards1968). Their respective uptakes were computed by multiplying the tissue nutrient contents with weed dry biomass for weeds, grain and stover in crops and finally summed up to derive total NPK uptake by the component crops. The uptake of component crops was summed to calculate the total uptake of the intercropping system.
The nutrient balance for N, P and K were quantified as a difference of nutrient inputs and removal (Blaise et al., Reference Blaise, Bonde and Chaudhary2005). Nutrient budget of N, P and K were estimated based on the initial and final soil status, nutrient added into the soil through fertilization and biologically fixed-N, whereas nutrients uptake by the crops (above and below ground biomass) and weeds were considered as nutrient removal or depletion. To estimate biological nitrogen fixation, leguminous component crops were sown solely with normal recommendations as explained in treatment imposition. The amount of nitrogen fixed by soybean and groundnut were then estimated using the standard method (Hauser, Reference Hauser1987).
Statistical analysis
Different parameters were statistically analysed using PROC GLM procedure of SAS Version 9.2 (SAS Institute Inc., Carry, NC, USA). The significance of treatment effects was determined by the F-test. The significance of the difference between means of two treatments was tested using least significant difference (LSD) at 5% probability level. In most of the cases, the effect of year and/or year × intercropping was significant; therefore, the results were presented separately for years.
RESULTS
Maize–legume intercropping on productivity
In the intercropping systems, grain yield of intercrops were influenced by the type of companion crops as well as planting densities (proportion of rows), and as a result, yield varied significantly among the intercrops with different row proportions. During both years (2010 and 2011), maize grain yield was significantly (p < 0.05) higher in solitary maize (3887–4110 kg ha−1) but was statistically (p < 0.05) comparable with intercropping of maize–soybean at 1:2 row proportion. The lowest maize grain yield was recorded for the intercropping of maize–groundnut at 1:5 rows (Table 1). However, intercropping of maize with soybean recorded a higher yield of maize relative to maize–groundnut intercropping across all planting densities. Soybean and groundnut recorded maximum yields when grown as sole crops compared to intercropping with maize across all planting density combinations. The lowest yield of both crops was recorded for the intercropping with maize at 1:1 row proportions (soybean: 380–485 kg ha−1 and groundnut: 349–371 kg ha−1) but the yield increased significantly (p < 0.05) with an increase in the planting density of both crops (soybean & groundnut) at 1:5 row proportions.
Table 1. Effect of maize–legume intercropping on grain and stover yield of sole and intercrops, biological and maize equivalent yield.
*T1: solitary maize, T2: solitary soybean, T3: solitary groundnut, T4: maize with soybean at 1:1 rows, T5: maize with soybean at 1:2 rows, T6: maize with soybean at 1:5 rows, T7: maize with groundnut at 1:1 rows, T8: maize with groundnut at 1:2 rows, T9: maize with groundnut at 1:5 rows. Means in the column followed by common letters (a-h) are statistically non-significant at 5% level of significance.
Stover yield of maize was higher with 1:1 rows of maize–soybean followed by 1:2 rows of maize–soybean and the lowest was recorded with 1:5 rows of maize–groundnut. Stover yield of intercrops (soybean & groundnut) during both the years (2010–2011) increased significantly (p < 0.05) with the increase in row proportions of intercrops from 1:1 to 1:5 maize–soybean/groundnut. The biological yield was highest for the intercropping of maize with soybean/groundnut at 1:2 rows followed by 1:1 row proportions. MEY was highest with intercropping of maize with groundnut at 1:5 rows (5725–6712 kg ha−1) followed by 1:2 (5485–6009 kg ha−1) row proportions. Intercropping of maize with soybean at 1:5 row proportions also produced comparable MEY (5294–5931 kg ha−1) with maize–groundnut at 1:2 row proportions. Among all the treatment combinations, solitary soybean during both years produced the lowest MEY (3304–3569 kg ha−1) (Table 1).
During 2010, the PE was highest with intercropping of maize with groundnut at 1:5 row proportions (50.5 kg ha−1 day−1), but in 2011, PE was highest with solitary groundnut (53.8 kg ha−1 day−1) followed by 1:2 row proportions of maize–groundnut (42.2–45.5 kg ha−1 day−1). Sole planting of groundnut, however, produced comparable PE (43.9–53.8 kg ha−1 day−1). Among all the treatment combinations, solitary soybean has the lowest PE during 2010 and solitary maize during 2011 (Figure 2).
Figure 2. Production efficiency as influenced by maize–legume intercropping.
Maize–legume intercropping on competition index
LER was significantly (p < 0.05) influenced by intercropping of maize with soybean and groundnut at various row proportions (Table 2). The highest LER was recorded with intercropping of maize–soybean followed by maize–groundnut at 1:5 proportions and the lowest was estimated with 1:1 of maize–groundnut. Intercropping of maize with legumes (soybean & groundnut) resulted in improvement of LER by 17.3–53.3% during 2010 and 5.7–50.2% in 2011 over sole maize crop. Similarly, relative crowding coefficient (K) was measured higher in intercropping of maize with soybean and groundnut at 1:2 row proportions compared to intercropping at 1:1 and 1:5 row proportions and solitary crops (Table 2). Estimated MAI reflected significantly (p < 0.05) higher monetary return with intercropping of maize–groundnut at 1:5 row proportions (12 687–21 006 INR ha−1) followed by 1:5 of maize–soybean (15 938–19 243 INR ha−1) and 1:2 of maize-groundnut (12 392–15 786 INR ha−1) than the solitary maize. Estimated LER, K and MAI, were higher during 2010 compared to 2011, mostly due to higher rainfall occurrence that led to higher grain yield under rainfed conditions in the EHR.
Table 2. Effect of intercropping on land equivalent ratio, relative crowding coefficient, monitory advantage index.
*T4: maize with soybean at 1:1 rows, T5: maize with soybean at 1:2 rows, T6: maize with soybean at 1:5 rows, T7: maize with groundnut at 1:1 rows, T8: maize with groundnut at 1:2 rows, T9: maize with groundnut at 1:5 rows. Means in the column followed by common letters (a-d) are statistically non-significant at 5% level of significance.
Maize–legume intercropping on weeds dynamics
The weed density, biomass and WSE were significantly (p < 0.05) influenced by intercropping of maize with legumes in various row proportions (Table 3). During the crop growth periods (at 60 DAS), the field was infested with broadleaf weeds (Ageratum conyzoides, Chromolaena odorata, Galinsoga parviflora & Boreria hispida), grasses (Digitaria sanguinalis & Cynodon dactylon) and sedge (Cyperus rotundus) during both years of experimentation (2010 and 2011). Galinsoga parviflora was the most dominant weed with a relative density of 31–32% followed by Ageratum conyzoides (26–27%), Chromolaen odorata (15–19%), Digiteria sanguinalis (12–13%) and Cyperus rotundus (6–7%). The measured weed density was significantly (p < 0.05) higher in solitary maize followed by intercropping of maize–groundnut and maize–soybean at 1:1 row proportions. The lowest weed density was recorded in solitary soybean (71.6–73.6% less) followed by solitary groundnut (62.6–63.7% less). Due to higher weed density, weed biomass was also significantly (p < 0.05) higher in solitary maize as compared to solitary soybean (66.3–67.2%) and groundnut (55.8–56.8%). Weed control efficiency expressed as WSE, was, by and large, inversely related to weed density and weed biomass. As a result, it was higher with solitary soybean (WSE: 66.2–67.1%) followed by intercropping of maize with soybean at 1:5 row proportions (55.6–57.1%) and groundnut (55.1–56.7%). It was noticed that 1:5 row proportions of maize–legume has better WSE than 1:1 to 1:2 row proportions.
Table 3. Effect of by maize–legume intercropping on density, biomass and weed smothering efficiency.
*T1: solitary maize, T2: solitary soybean, T3: solitary groundnut, T4: maize with soybean at 1:1 rows, T5: maize with soybean at 1:2 rows, T6: maize with soybean at 1:5 rows, T7: maize with groundnut at 1:1 rows, T8: maize with groundnut at 1:2 rows, T9: maize with groundnut at 1:5 rows; † figures in parenthesis represents original weed density; WSE: weed smothering efficiency. Means in the column followed by common letters (a–f) are statistically non-significant at 5% level of significance.
Maize–legume intercropping on nutrient budgeting
Table 4 elucidated that available nitrogen (N) balance was significantly (p < 0.05) influenced by the type of intercrops and row proportions. The nitrogen content of the soil was improved with the inclusion of legumes as intercrops with maize at higher row proportions. The maximum N gain was obtained with sole groundnut and their respective higher row proportions in the intercropping. Higher the row proportion of groundnut, maximum residual/leftover of N in soil was measured. The balance of the recorded N after two year of experimentation was positive with 1:5 of maize–groundnut (73.2 kg N ha−1) followed by 1:2 of maize–groundnut and 1:5 of maize–soybean. Intercropping of maize–soybean with 1:1 row proportion exhausted 47.6 kg N ha−1 followed by solitary maize (42.7 kg N ha−1) and soybean (16.5 kg N ha−1) from its initial status and measured a negative N balance. The trend was almost similar during both the years (2010 and 2011).
Table 4. Effect of maize–legume intercropping on nitrogen (kg ha−1) balance including crop uptake.
* T1: solitary maize, T2: solitary soybean, T3: solitary groundnut, T4: maize with soybean at 1:1 rows, T5: maize with soybean at 1:2 rows,T6: maize with soybean at 1:5 rows, T7: maize with groundnut at 1:1 rows, T8: maize with groundnut at 1:2 rows, T9: maize with groundnut at 1:5 rows; †Fixation was estimated by {final status of N-[(initial soil status+ additional supply)-(uptake by crop+uptake by weeds)]}. Means in the column followed by common letters (a-g) are statistically non-significant at 5% level of significance.
Similar to N, available phosphorus (P) and potash (K) balance were significantly (p < 0.05) influenced by intercrops and row proportions. Compared to its initial level, the available P was improved by 1.5 kg P ha−1 with solitary soybean and amount in 1:5 row proportions of maize–soybean (Table 5). It was observed that with the increase in row proportions of intercrop, availability of P also improved gradually. However, solitary maize had a negative balance of P (−0.2 kg P ha−1). Solitary maize exhausted the available P and resulted in a lower availability P in the soil. The available K also followed a similar trend of P and, compared to its initial level, it was improved by 31.6 kg K ha−1 with solitary soybean and by 28.2 kg K ha−1 with 1:5 row proportions of maize–soybean (Table 6). However, K availability with solitary maize was exhausted by 5.3 kg K ha−1 compared to its initial level. The trend was similar during both 2010 and 2011.
Table 5. Effect of maize–legume intercropping on phosphorus (kg ha−1) balance including crop uptake.
* T1: solitary maize, T2: solitary soybean, T3: solitary groundnut, T4: maize with soybean at 1:1 rows, T5: maize with soybean at 1:2 rows,T6: maize with soybean at 1:5 rows, T7: maize with groundnut at 1:1 rows, T8: maize with groundnut at 1:2 rows, T9: maize with groundnut at 1:5 rows; †Fixation was estimated by {final status of N−[(initial soil status+ additional supply)−(uptake by crop+ uptake by weeds)]}. Means in the column followed by common letters (a–f) are statistically non-significant at 5% level of significance.
Table 6. Effect of maize–legume intercropping on potassium (kg ha−1) balance including crop uptake.
* T1: solitary maize,T2: solitary soybean, T3: solitary groundnut, T4: maize with soybean at 1:1 rows, T5: maize with soybean at 1:2 rows,T6: maize with soybean at 1:5 rows, T7: maize with groundnut at 1:1 rows, T8: maize with groundnut at 1:2 rows, T9: maize with groundnut at 1:5 rows.; †Fixation was estimated by {final status of N−[(initial soil status+ additional supply)−(uptake by crop+uptake by weeds)]}. Means in the column followed by common letters (a–h) are statistically non-significant at 5% level of significance.
Similarly, intercropping of maize with legumes at different row proportions significantly (p < 0.05) influenced the nutrient uptake. The most N and K were taken up by 1:5 of maize–soybean and least with solitary groundnut. But, higher P was taken up by 1:5 of maize–groundnut followed by 1:5 of maize–soybean and least uptake of K was with solitary maize (Tables 4–6). However, nutrient mining by weeds followed the inverse trend of nutrient uptake by the crop. Estimated nutrient removal by weeds was highest in solitary maize (9.2–11.5 kg N ha−1; 3.4–4.2 kg P ha−1 and 10.9–13.2 kg K ha−1), and lowest in solitary soybean (4.4–5.7 kg N ha−1; 2.3–2.8 kg P ha−1 and 5.3–6.5 kg K ha−1). In intercropping, higher row proportions of soybean and groundnut, resulted in a significant (p < 0.05) reduction in the observed uptake of nutrients-NPK by weed biomasses (Tables 4–6). It was also noticed that, most N and K were mined by weeds in solitary maize followed by 1:1 of maize–groundnut at 1:1 row proportions. But, most P was mined with 1:1 of maize–groundnut followed by 1:1 of maize–soybean and least with solitary soybean.
DISCUSSION
Maize–legume intercropping on productivity and production efficiency
A fundamental aspect of intercropping is to avoid unfavourable intra- or inter-specific competitions while facilitating inter-specific complementation for increasing the growth and survivals of intercrops. In the present study, relative to the conventional solitary planting of maize, intercropping system resulted in higher system productivity and other subsidiary benefits including soil enrichment of major nutrients-NPK. The observed benefits of the intercropping system significantly improved the overall yield expressed in terms of MEY, mostly due to the synergistic effect of crop intensification and higher plant density per unit area. The highest MEY, observed with maize–groundnut at 1:5 followed closely by 1:2 rows and then 1:5 of maize–soybean, might be due to a higher yield of intercrops in the intercropping system and also because of the relatively high market price of the companion crop (legumes) produces. With the inclusion of legumes (groundnut & soybean) with maize at different row proportions (1:1 to 1:5), the MEY was improved by considerably (17–63.3% in 2010 and 9.5–47.3% in 2011) over solitary maize. Sani et al. (Reference Sani, Oluwasemire and Mohammed2008) also demonstrated that with the reduction in plant density in solitary crops, yield is reduced; on the contrary, increase in plant density increased the individual production, but system productivity is largely dependent on the yield of component crops.
The yield data illustrated how intercropping effectively reduced the risk of low productivity, particularly in the socio-economically fragile hilly ecosystem with predominantly rainfed agriculture, vulnerable to abrupt rainfall change. Low maize yields in intercrops were often compensated by higher legume yields, and vice versa. In solitary maize, higher maize yield was mainly due to less competition for sunlight, space, water and nutrients. This resulted in reduction of maize yield at varying magnitude (2.2–9.5%) with different row proportions (1:1 to 1:5). Similar findings were also reported by Hussain et al. (Reference Hussain, Shamsi, Khan, Akbar and Shah2003) and Haque et al. (Reference Haque, Sharma and Prasad2008). However, additional significant yield from the intercrops compensated the marginal loss of maize yield and even superseded the system productivity in intercropping than the solitary maize crop.
Maize–legume intercropping on competition indices
Competition indices are often used for judging the efficiency and profitability of intercropping systems. The LER of an intercropping system provides precise assessment of the competitive relationship between the component crops. It also indicates better adjustment of row proportions for suitable use of land with the higher system productivity (Hayder et al., Reference Hayder, Mumtaz, Khan and Khan2003). The higher LER observed with 1:5 row proportion of maize–soybean followed by maize–groundnut intercropping indicated that 1:5 row proportions strongly influenced the crop productivity and utilised the land area more efficiently compared to solitary and other intercropping systems. Higher relative crowding coefficient (K) at 1:2 rows of maize–soybean depicted that this proportion was more competitive for input resource use efficiency. Similarly, considerably higher MAI affirmed the income advantages in the intercropping system over solitary system. The positive exponential relationship between LER and K with higher coefficient (R 2 = 0.66) as well as quadratic relationship between LER and MAI (R 2 = 0.81) further affirmed the added advantage of intercropping in productivity and PE. Similarly, K followed the linear relationship with MAI (R 2 = 0.79). The relationship among different competition indices thus varied from linear to non-linear in nature. In separate studies, similar findings were also reported by Hayder et al., (Reference Hayder, Mumtaz, Khan and Khan2003) and Esmaeil et al., (Reference Esmaeil, Dabbagh, Mohammadi, Shakiba, Ghassemi-Golezani, Aharizad and Shekari2010).
Maize–legume intercropping on weed and nutrient removal
Intercropping of maize with soybean/groundnut at all row proportions reduced the weed density and weed biomass compared to the solitary maize. The decrease in available space and light in intercrops due to more crop canopy cover resulted in reduction in weed density and biomass, thus, registered higher WSE. In the uplands of the EHR, due to high rainfall, richness in biodiversity and higher soil organic carbon (Choudhury et al., Reference Choudhury, Mohapatra, Das, Das, Nongkhlaw, Fiyaz, Ngachan, Hazarika, Rajkhowa and Munda2013), vigorous weed growth and its infestation is one of the major causes of low productivity (<1.5 t ha−1) of the predominantly maize-based, rainfed agriculture (Choudhary et al., Reference Choudhary, Anil, Kumar and Chauhan2014). There are reports that due to the heavy weed infestation, the yield loss of rainfed, maize-based system ranges from 26 to 70.5% and, even in extreme cases, partial to complete crop failure is encountered (Singh et al., Reference Singh, Prabhukumar, Sairam and Kumar2009). Thus, introducing intercrops, particularly leguminous soybean and groundnut in maize–based system with 1:5 row proportions will be one of the feasible options for improving individual crop yield as well as system productivity. The higher WSE in intercropping helped to control the unproductive nutrient removal by weeds and thus, increased the availability of nutrients to the crops in the intercropping system. Similar results of higher weed suppression capacity (Haque et al., Reference Haque, Sharma and Prasad2008; Tripathi et al., Reference Tripathi, Kumar, Nath and Yadav2008), enrichment of soil with N and its higher availability in intercropping of legumes with cereals (maize/wheat) were also reported by other workers (Amossé et al., Reference Amossé, Jeuffroy, Mary and David2014). Removal of nutrients-NPK by weeds statistically (p < 0.05) differed with different row proportions. Among the cereals, maize is one of the nutrient exhaustive crops and this was also reflected in our study. The highest uptake of major nutrients-NPK was recorded from solitary maize followed by 1:1 of maize–groundnut and the least K removal was observed with solitary groundnut. Contrary to our reported results of less removal of K, Srinivasarao et al. (Reference Srinivasarao, Kundu, Venkateshwarulu, Lal, Singh, Balaguravaiah, Vijaysankarbabu, Vittal, Reddy and Manideep2013) reported higher K removal by groundnut, even below the critical limit of soil. This difference might be due to their intensive experimentation of monocropping of groundnut for long periods (20 years) in rainfed semi-arid ecosystem compared to our two year old less extensive intercropping of groundnut with maize under humid sub-tropical climate with 4–5 times higher rainfall than the semi-arid environment.
Maize–legume intercropping on nutrient budget
The nitrogen balance study revealed that for cultivation of solitary maize, soil N status was reduced by >20 kg ha−1 compared to the initial N level. It was also noticed that the inclusion of legume crops, particularly soybean either in solitary maize or with maize at 1:1 row proportion also exhausted the N level. Contrary to soybean, solitary groundnut increased the available N in soil but in 1:1 row with maize, declining trend similar to that with soybean was observed. More crop uptake and less fixation in solitary soybean as well as maize compared to groundnut resulted in this differential responses. Exploitation of mineral nutrients by maize also encourages the legumes to fix atmospheric N. Groundnut was more responsive to this and possibly contributed much N to the system. Therefore, groundnut has a higher potential to supply N to the maize–legume intercropping system (Mucheru-Muna et al., Reference Mucheru-Muna, Pypers, Mugendi, Kunngu, Mugwe, Merckx and Vanlauwe2010). N fixation by the soybean could not compensate for the N removed through the crop produces. Nitrogen content in soybean grain is very high (6.2%) and this mostly happens due to high rate of translocation of N to the grain by soybean. To meet this requirement, N uptake by soybean was higher than groundnut and thus left less N for the soil system. It has also been noted that a large amount of N is leached during the cropping season in maize–legume systems (Gentry et al., Reference Gentry, David, Below, Royer and Mclsaac2009), whereas, groundnut with better (below and above ground) plant architecture, including surface coverage had the advantage of conserving N within the system. Similar results were also corroborated by Sanginga, (Reference Sanginga2003) and Mucheru-Muna et al. (Reference Mucheru-Muna, Pypers, Mugendi, Kunngu, Mugwe, Merckx and Vanlauwe2010) in separate studies.
However, at higher row proportions of soybean and groundnut with maize (1:2 to 1:5) despite higher N uptake, soil N level increased and this was mostly due to the higher amount of external N application in the form of fertilizer and additional supplementation in the form of N-fixation. Inclusion of legumes as an intercrop supplied N to the companion crop by increasing soil N level from initial status by atmospheric N-fixation (Fan et al., Reference Fan, Zhang, Song, Sun, Bao, Guo and Li2006) and by the decomposition of fallen leaf litter of legumes (Ncube et al., Reference Ncube, Twomlow, Van Wijk, Dimes and Giller2007). But, N-fixation by the legumes alone cannot compensate for the N removed through the produces. N balances largely depended on additional supply through fertilizer and fixation, uptake by crop and weeds and other losses that occur, mainly through leaching. Therefore, a moderate N application, targeted for the cereal crop, was necessary to sustain yields in the long run, especially when solitary maize was planted or for initial development of intercrops. Similar views were also opined by other researchers (Li et al., Reference Li, Sun, Zhang, Li, Rengel, Yang and Tang2001; Li et al., Reference Li, Zhang, Li, Christie, Sun, Yang and Tang2003; Sanginga, Reference Sanginga2003). Intercropping of deep rooted (>70 cm) maize with shallow rooted (<60 cm) legumes (soybean & groundnut) might have encouraged the maize to explore further for nutrients and seek soil N from different soil depths (Hauggaard-Nielsen et al., 2001). Intercropping of maize with shallow rooted soybean and groundnut at varying planting densities and proportions in the rainfed hilly ecosystem of the EHR of India provided better and more efficient root architecture modified for exploration of soil moisture as well as mineralizable plant nutrient elements essential for crop uptake. This was reflected in the higher grain yield (including MEY) observed in intercropping compared to solitary cropping despite comparable management practices (i.e. rainfed). Previous studies also reported higher nutrient uptake in intercropping with pea-barley (Hauggaard-Nielsen et al., 2001) and maize–legume (Woomer et al., Reference Woomer, Langat and Tungani2004) over respective solitary crops.
Higher MEY under the intercropping system helped increase uptake of P and K. However, it was noticed that solitary maize did not influence the P status, while K was exhausted by 5.3 kg K ha−1. Unlike N, solitary soybean and groundnut improved both the P and K status of soil by varying magnitudes. Inclusion of legumes with maize also significantly (p < 0.05) improved the P and K status with all the row proportions and intercrops. This might be due to acidification of rhizosphere which mobilized the P and K in different row proportions of intercrops. Similar results were corroborated in faba bean by Li et al. (Reference Li, Li, Sun, Zhou, Bao, Gang and Zhang2007; Reference Li, Shen, Zhang, Marschner, Cawthray and Rengel2010). The nutrient balance sheet indicated that, N uptake by soybean was higher to supplement grain N content, which reduced the left over residue of N, and thus resulted in soil N depletion. In groundnut N uptake was relatively less, fixation was higher and thus left more residual N was left in the soil resulting in a positive balance over the years. Thus, adequate supplementation of N in soybean (either sole or intercrop) based system is essential for better grain quality. Due to continuous removal of crop residues including fallen legume (soybean & groundnut) after harvest nutrient enrichment of soils was marginal (Ncube et al., Reference Ncube, Twomlow, Van Wijk, Dimes and Giller2007).
CONCLUSIONS
Dominance of marginal-input, intensive, solitary maize in the vast uplands of high rainfall receiving EHR is one of the major reasons for low cropping intensity (<130%), heavy weed infestation, decline in soil nutrients and low productivity (<1.5 t ha−1). The investigation revealed that intercropping, particularly involving leguminous intercrops (soybean & groundnut) in maize-based system holds significant potential to overcome these shortfalls. Inclusion of groundnut and soybean as intercrops with maize at higher row proportions (1:2 to 1:5) increased the system productivity by 40–60% as well as improving the soil fertility status and the availability of major nutrients-NPK. Leguminous intercrops, particularly groundnut also enriched the soil with atmospheric N through fixation, reduced the weed infestation with very high wide smothering efficiency (>40%) and thus, avoided unproductive nutrient removal by weed uptake. Initial external supplementation of fertilizer N in solitary soybean or as intercrop at lower row proportions (1:1) with maize-based systems will ensure higher soil N availability and better grain quality. Inclusion of leguminous intercrops in predominantly solitary maize-based systems, thus hold the potential to reduce the risk of crop failure, whilst simultaneously improving the land and system productivity vis-à-vis food grain security in the rainfed hilly ecosystem of the EHR and other similar hilly agro-ecosystems of the world.
COMPLIANCE WITH ETHICAL STANDARDS
The authors declare that they have no conflict of interest.
Acknowledgement
Authors are very greatful to Dr. Lindsay Todman, Soil Scientist at the Division of Sustainable Soils and Grassland System, Rothamsted Research for her help in checking the English language (proof reading) of the manuscript.
