Effect of sowing proportion on above- and below-ground competition in maize–soybean intercrops

Site description

Field experiments were conducted at the Changwu Experimental Station (35° 12′ N, 107° 40′ E, altitude 1200 m) located in Shaanxi Province, China. The experimental site was in the typical dryland farming area on the Loess Plateau. Annual precipitation in the area averaged 582 mm between 1957 and 2013, with a mean annual temperature of 9.7 °C over that period. Rainfall and temperature during the two study years are shown in Fig. S1. Soils were generally of the Calcaric Regosol group, according to the FAO/UNESCO soil classification system⁵², and were composed of 4% sand, 59% silt, and 37% clay⁵³. The 0–20 cm soil properties were the following: pH, 8.4; organic matter content, 11.8 g kg⁻¹; total N content, 0.87 g kg⁻¹; and Olsen-P, 14.4 mg kg⁻¹.

Experimental design and field management

Two-year experiment was arranged in a randomized complete block design with three replicate plots during 2012 and 2013 growing seasons^25,54_ENREF_53. The study was conducted using the soybean cultivar (Glycine max L.) cv. Zhonghuang 24 and the maize cultivar (Zea mays L.) cv. Zhengdan 958 grown in cereal–legume agricultural systems. Zhonghuang 24 was bred from Jilin 21 and fendou 31 × Zhongdou 19 (deposition number 2008003); Zhengdan 958 was the offspring of inbred Zheng 58 and Chang 7-2 (deposition number 20000009), which are approved in China. The cropping system treatments were as follows:

1. 1.

   Sole-cropped soybean (S).

2. 2.

   Sole-cropped maize (M).

3. 3.

   Two rows of maize intercropped with two rows of soybean (M2S2).

4. 4.

   Two rows of maize intercropped with four rows of soybean (M2S4).

Each plot measured 6 m × 4 m, with row spacing of 50 cm for maize and soybean both in sole crops and intercrops. Individual plants were spaced at 22 cm and 19 cm for maize and soybean, respectively, with one plant per stand for maize and two plants per stand for soybean to attain densities of 90,000 and 210,000 plants ha⁻¹, respectively. In 2012, seeds of maize and soybean were sown on 25 April and harvested on 28 September, and in 2013, seeds were sown on 20 April and harvested on 25 September. Before sowing, basal fertilizer was applied at a rate of 90 kg N ha⁻¹ as urea (46% N) and 150 kg P₂O₅ ha⁻¹ as superphosphate (12%, P₂O₅), and then additional fertilizers were uniformly spread in each plot, which were then ploughed into the 0–30 cm soil layer using a rotary tiller. All of the plots received 67.5 kg N ha⁻¹ as urea at the bell and silking stages using a hole-seeding machine. No irrigation was applied, and weeds were removed by hand when sighted. The research on plants complied with relevant institutional, national, and international guidelines and legislation.

Above- and below-ground measurements

The Pn was measured with a LI-6400 portable photosynthesis system (LI-COR Inc., Lincoln, NE, USA) from 9:00 to 11:00 h at 120 days after sowing, which corresponds to the milk stage in maize and full seed stage in soybean^7,13. We measured photosynthesis of ear leaves of maize, the first spreading leaves at the top of soybean in both the sole crops and intercrops. The Pn values were calculated as the sum of the mean readings for five leaves in each plot. The LAI values, DIFN were recorded using a Plant Canopy Analyzer (Li-2200, LiCor Inc., Lincoln, NE, USA) without direct sunlight at milk stage of maize. One above-canopy measurement and three below-canopy measurements at the soil surface were taken for four replicates in each plot. SPAD were collected using a hand-held dual wavelength meter (SPAD 502, Chlorophyll meter, Minolta Camera Co., Ltd., Japan) at milk stage of maize. Measurements were taken midway along the ear leaves of maize and the first spreading leaves at the top of soybean from five adjacent plants at the center of row in each plot.

The SWS was measured gravimetrically using a soil auger at 10 cm intervals over a depth of 100 cm and at 20 cm intervals over a depth of 200 cm at milk stage of maize for three replicates in each plot. The SWS was calculated for each plot in the 0–200 cm soil profile for the soil moisture using the following formula: SWS = SWC × SD × SBD, where SWC represents soil water content, SD represents soil depth, and SBD represents soil bulk density. Apparent water use during crop growth season was expressed as evapotranspiration (ET), which was determined according to the following formula: ET = ΔSWS + P, where ΔSWS is the change in soil water storage in the top 200 cm and P is the rainfall (mm) between planting and at milk stage in maize. The six adjacent plant samples were collected at milk stage of maize in the middle two rows of each plots (Fig. S2). The sampling included shoots and roots of maize and soybean. At the cotyledonary node, above-ground parts were separated from below-ground parts. Soil core samples (9 cm diameter × 15 cm) at the intra-row of crop were collected to a depth of 100 cm using an auger and separated in 10-cm sections to determine the root growth in sole-cropping and intercropping systems. The samples were exposed to 105 °C for 30 min and then dried to a constant weight at 75 °C. The oven-dried samples were put in small plastic bags after grinding. The study of N and P uptake are the most common among mineral elements^55,56. Concentrations of N and P in the plant dry matter were determined after digestion with H₂SO₄ and H₂O₂; N concentration was measured according to the Kjeldahl method²⁰, whereas P concentration was measured by the molybdenum-antimony anti-spectrophotometric method¹⁶. Crop N and P uptake were calculated by the actual above-ground biomass multiplied by plant tissue N and P concentrations. Grain yield was estimated at harvest from 6 m² for maize and soybean based on the average of three plot replicates.

Data analysis

The LER for assessment of land use advantage. LER is sum of ratio of intercrop to sole crop for maize and soybean yield⁵⁷:

where LER_m and LER_s are patial LER for maize and soybean, respectively. Y_im and Y_is are yields of maize and soybean under intercrops, respectively. Y_sm and Y_ss are the yield of maize and soybean under sole crop, respectively.

The water equivalent ratio (WER) was calculated to measure water use advantage of intercropping⁵⁸:

where WER_m and WER_s are patial WER for maize and soybean, respectively. ET_im and ET_is are ET of maize and soybean under intercrops, respectively. ET_sm and ET_ss are the ET of maize and soybean under sole crop, respectively.

All analyses were conducted in SPSS Statistics 17.0 (SPSS Inc., Chicago, IL, USA). Treatment means showing significant differences among different cropping systems were separated using one-way ANOVA or least significant difference (LSD) at a threshold of 5% to compare the effect of yield, above- and below-ground related parameters (Pn, LAI, SPAD, DIFN, SWS, N and P uptake) in different maize–soybean intercropping. The variation in Pn, LAI, SPAD, DIFN, SWS, N, and P uptake of crop, and the effects of cropping system × year were made using Univariate General Linear Models. Pearson’s correlation test was used to analyze between LER and above-and below-ground biomass of maize and soybean. The effects of above- and below-ground factors on biological yield were quantified, by calculating the contribution value of some key factors to yield. The effects of between above- (LAI, SPAD, DIFN) and below-ground (SWS, N and P uptake) competition on the biological yield and contribution rate were conducted by the linear regression model⁵⁹:

where Y represents biological yield, LAI represents leaf area index, SPAD represents chlorophyll, DIFN represents diffuse non interceptance, SWS represents soil water storage, N represents crop nitrogen uptake, P represents crop phosphorus uptake, X represents interaction for LAI, SPAD, DIFN, SWS, N, and P, and β₀, β₁, β₂, β₃, β₄, β₅, β₆ and β₇ represent the fitted parameters. The standard regression coefficients (Beta) of LAI, SPAD, DIFN, SWS, N, and P were determined on the basis of Eq. (1) to split their influence on the biological yield by the following equations:

where β₀′, β₁′, β₂′, β₃′, β₄′, and β₅′ represent the standard regression coefficients for LAI, SPAD, DIFN, SWS, N, and P. LAI′, SPAD′, DIFN′, SWS′, N′, and P′ represent the standard deviations for LAI, SPAD, DIFN, SWS, N, and P. Y′ is the standard deviation for the modeled biological yield.

Source: Ecology - nature.com