Spatial and temporal distribution of mung bean (Vigna radiata) and soybean (Glycine max) roots

Spatial and temporal distribution of roots of mung bean and soybean originated from different geographical backgrounds is an important scientific issue. The aim of this study was to research the spatial and temporal distribution of roots system of soybean cultivar ‘Hefeng55’ and mung bean cultivar ‘Jilv7’ which can elucidate differences between soybean roots and mung bean roots in the key spatial and temporal locations. The roots at V6, R2, R4, R5, R6, and R7 stages were collected to acquire data of root length, root surface area, root volume and root dry weight. 49.8%, 11.7%, 13.2%, 14.7% and 10.6% of soybean roots and 57.8%, 10.7%, 11.2%, 11.9% and 8.4% of mung bean roots were in 0-5, 5-10, 10-15, 15-20 and 20-25 cm horizontal soil layers, respectively; 79.2%, 11.5%, 4.3%, 1.8%, 1.1%, 1.0% and 1.1% of soybean roots and 70.0%, 12.3%, 8.0%, 3.0%, 1.6%, 1.7% and 3.4% of mung bean roots were in 0-20, 20-40, 40-60, 60-80, 80-100, 100-120 and 120-140 cm vertical soil layers, respectively. Compared with mung bean, soybean had a much larger root system during development. In horizontal direction, soybean root tended to be more laterally developed, but the distribution of mung bean root was more uniform in vertical direction. With a greater root surface area to weight ratio (AWR), mung bean had a finer root system than soybean. These findings can help to clarify the fourdimensional spatial and temporal distribution characteristics of legumes and may provide reference for production practice of soybean and mung bean in the future.


Introduction Introduction Introduction Introduction
Roots are an important organ of plants (Fang, 2011) which determines the ability of plants to absorb water and nutrients (Vamerali et al., 2003;Ehdaie et al., 2010). Roots of different crops have different distribution characteristics (Lynch, 1995;Benjamin and Nielsen, 2006). Comparing the temporal and spatial distribution characteristics of different crops' roots is beneficial to research the root structure differences and the adaptability of root systems to the soil environment among different crops (Gan, 2009;Fan et al., 2016).
The distribution characteristic of roots is the basic attribute during plant development (Atta et al., other plant parts (Eissenstat and Yanai, 2002;Waisel and Eshel, 2002). Liu et al. (2011) studied the distribution of pulses and found that surface area was mainly distributed in 0-60 cm soil layer. Benjamin and Nielsen (2006) reported 97% of the root dry weight of soybean and about 80% of the root dry weight of chickpea and field pea were in the surface 23 cm. Mitchell and Russell (1971) found that 90% or more of the root dry weight of soybean was concentrated in the upper 7.5 cm early in the season and in the upper 15 cm during the remainder of the season. Both soybean and mung bean are leguminous crops, which originated from different geographical backgrounds. Various workers have studied the root distribution of soybean in the last few years (Calonego et al., 2010;Farmaha et al., 2012), but the difference between these two crops' root distribution at different time and space points is still unclear. In this study, we used innovative horizontal and vertical devices to study the root distribution of soybean and mung bean in 0-5, 5-10, 10-15, 15-20 and 20-25 cm horizontal soil layers and in 0-20, 20-40, 40-60, 60-80, 80-100, 100-120 and 120-140 cm vertical soil layers, respectively. We hypothesized that the root system of soybean and mung bean have different trends in temporal distribution and different pattern in spatial distribution. It can give a deeper understanding of the four-dimensional spatial and temporal distribution characteristics of legume roots and may provide reference for breeding new cultivars of soybean and mung bean in the future.

Experimental site
The experiment was carried out at outdoor test site in National Coarse Cereals Engineering Research Centre, Daqing, China on June 5, 2015 (soybean) and June 5, 2016 (mung bean). The annual precipitation at the experimental site was 508.7 mm, the average annual temperature was 5.60 °C, the effective accumulated temperature was 2900-3000 °C, and the sunshine duration was 1158 h (Collected from Daqing Weather Station).

Experimental devices
There were two kinds of devices: horizontal device and vertical device ( Figure 1). The horizontal device was a cylindrical metal barrel with a diameter of 50 cm and a height of 50 cm, which was fixed with a cross steel frame with a length of 54 cm on both sides of the metal barrel. The inside of the metal barrel was equipped with a diameter of 10 cm, 20 cm, 30 cm and 40 cm of metal net. The distance between the metal net was 5 cm and the metal net were fixed on the cross-steel frame by nylon straps (Figure 2). The vertical device was a cylindrical plastic barrel with a diameter of 30 cm and a height of 150 cm ( Figure 3). In order to facilitate sampling, inside the vertical device was a plastic water belt of 30 cm in diameter and the soil was filled in the plastic water belt ( Figure 4). The lower end of the plastic water belt was sealed and four round holes were cut using scissors.           The soil was chernozem, with physical and chemical properties characterized by a pH of 7.8, effective phosphorus of 13.69 mg·kg -1 , alkali-hydrolyzed nitrogen of 134 mg·kg -1 , available potassium of 204 mg·kg -1 , and organic matter of 32.8 g·kg -1 . The soil was screened before pouring into the devices to remove grass root, tree root and large granular clods and stones. Then the soil was filled into the vertical device and horizontal device, respectively (1.15×102 kg·m -3 in density).

Experiment design, species and seeding
Soybean cultivar 'Hefeng55' and mung bean cultivar 'Jilv7' were planted at five seeds per barrel separately in 48 barrels which included 24 horizontal devices and 24 vertical devices. Two seedlings were retained, and grown with four replications per growth stage and type of device.
In horizontal devices, the centre of the cross section of the column was taken as the starting point to obtain root samples in the horizontal direction i.e. 0-5 cm, 5-10 cm, 10-15 cm, 15-20 cm and 20-25 cm layers. In vertical devices, the upper soil surface was taken as the starting point to obtain soil samples with root in 0-20 cm, 20-40 cm, 40-60 cm, 60-80 cm, 80-100 cm, 100-120 cm and 120-140 cm soil layers in vertical direction. The plants were clipped at cotyledons by scissors before sampling. Soil samples containing root were soaked in a plastic bucket filled with water until the soil became soft and then filtered. The obtained root samples were washed with clean tap water and then placed in a plastic, sealable bag, and the bag was placed in a refrigerator for further use.

Data collection
The harvested root samples were placed in a clear glass tray filled with water. The roots were washed to remove soil particles and other dirt that could hamper efficient scanning of root samples. The glass tray was placed on a scanner (Epson V700) and digital images were generated at 400 dpi. Digital image analysis of root samples was conducted using WinRHIZO (version 2014a, Reagent Instruments Inc., Quebec, Canada) and the data included root length, root surface area and root volume, from which root length density (RLD), root surface area density (RSAD) and root volume density (RVD) were estimated as follows: RLD = L/V0 RSAD = S/V0 RVD = V/V0 V0 = πr 2 h where V is the root volume, L is root length, S is root surface area, V0 is the soil volume, r is the radius, and h is the height.
After scanning, the roots were removed from glass tray and subsequently were placed in an oven at 105 °C for 2 hours, then drying to constant weight in 75 °C oven. The dry weight of roots was obtained by analytical balance and the root dry weight density (RDWD) was estimated as: RDWD = M/V0 V0 = πr 2 h where M is the root dry weight.

Statistical analysis
Differences between soybean and mung bean roots were determined by LSD test by SPSS 22.

Temporal distribution of total root dry weight
The distribution of total root dry weight showed that 'Hefeng55' had significantly greater root dry weight than 'Jilv7' during all growth stages in horizontal devices, and total root dry weight of 'Hefeng55' were significantly greater than 'Jilv7' except for V6 in vertical devices (Figure 8).
Root dry weight density as estimated by vertical device: 'Hefeng55' had significantly greater root dry weight density than 'Jilv7' in 0-20 cm soil layer during all growth stages, in 20-40 cm soil layer at R2, R4, R5, R6, R7 and in 40-60 cm soil layer at R6. The percentage of root dry weight in 0-20, 20-40, 40-60, 60-80, 80-100, 100-120 and 120-140 cm vertical soil layers to total root dry weight of 'Hefeng55' were 79.2%, 11.5%, 4.3%, 1.8%, 1.1%, 1.0% and 1.1%, respectively, and the percentage of root dry weight in 0-20, 20-40, 40-60, 60-80, 80-100, 100-120 and 120-140 cm vertical soil layers to total root dry weight of 'Jilv7' were 70.0%, 12.3%, 8.0%, 3.0%, 1.6%, 1.7% and 3.4%, respectively (Table 8) Table  Table Table  Table 7 7 7 7. . . . Root dry weight density (g · m -3 ) of soybean cultivar 'Hefeng55' and mung bean cultivar 'Jilv7' in different horizontal soil layers at V6, R2, R4, R5, R6 and R7 growth stages Data represent average ± standard error. Distinct letters in the row indicate significant differences. Significant at the 0.05 probability level.  Table  Table Table  Table 8 8 8 8. . . . Root dry weight density (g · m -3 ) of soybean cultivar 'Hefeng55' and mung bean cultivar 'Jilv7' in different vertical soil layers at V6, R2, R4, R5, R6 and R7 growth stages Data represent average ± standard error. Distinct letters in the row indicate significant differences. Significant at the 0.05 probability level Root surface area to root weight ratio In horizontal devices, 'Jilv7' had greater root surface area to root weight ratio (AWR) except for R7, and a significantly greater AWR of Jilv7 was found at V6 than 'Hefeng55'. In vertical devices, 'Jilv7' had significantly greater AWR than 'Hefeng55' during all growth stages (Figure 9).   In this research, the variation and trend of horizontal distribution of total root length of soybean and mung bean were identical during all growth stages, but soybean had greater total root length than mung bean. In horizontal direction, both soybean and mung bean had the largest proportion of root length in 15-20 cm soil layer followed by 10-15, 20-25 and 5-10 cm soil layers, and the minimum proportion of root length was found in 0-5 cm soil layer. Root length of soybean in vertical 0-20 cm and 20-40 cm soil layers reached 56.8% and 23.2% of the total root length, respectively, which were higher than 51.2% and 22.9% of mung bean in the same soil layer. This is similar to the finding of Gao et al. (2010). However, both the ratio of root length to total root length of mung bean in vertical 40-60 and 60-80 cm soil were greater than those of soybean. In soil layers below 80 cm in vertical direction, the root length percentages of soybean and mung bean were similar. The root length of soybean and mung bean were mainly concentrated in 0-40 cm vertical soil layer, and root length of mung bean was more evenly distributed in the vertical direction compared with soybean.
The root length density of the crop could be used to reflect extension and distribution of crop root (Adiku et al., 2001;Zhu, 2010;Liu et al., 2011). With the deepening of the soil layer, root length density of soybean and mung bean decreased gradually. One possible explanation for this phenomenon is that mechanical impedance limited the extension of the root system (Logsdon et al., 1987).
The distribution of crop root volume is critical to growth, development and yield formation of crop (Rao and Ito, 1998). The trend of total root volume of mung bean was more stable than that of soybean, but the total root volume of soybean was greater than that of mung bean during all growth stages. Spatial distribution of root showed that the root volume of mung bean was more concentrated in 0-5 cm horizontal soil layer compared with soybean, but soybean root volume tended to develop laterally. Both soybean (26.6%) and mung bean (23.9%) had the largest proportion of root volume in horizontal 15-20 cm soil layer. The ratio of root volume to total root volume in horizontal 0-5 cm soil layer of mung bean was greater than that of soybean, but the proportion of soybean in other horizontal soil layers were greater than those of mung bean. In vertical direction, soybean and mung bean were similar in root volume percentage in 0-20 cm vertical soil layer, and both of them had the greatest proportion in this layer. This is similar to the finding of . The root volume percentage of soybean in vertical 20-40 cm soil layer was higher than that of mung bean, but in the subsequent 40-80 cm vertical soil layer, mung bean had a larger root volume percentage compared with soybean. This showed that the difference of root volume distribution between soybean and mung bean was mainly concentrated in the upper middle soil layers in vertical direction. Compared with mung bean, soybean had a larger proportion of root volume in the upper soil layer, while proportion of mung bean root volume was higher than that of soybean in the middle soil layer.
In horizontal spatial distribution, the root surface area of soybean and mung bean were mainly distributed in the soil layer of 10-25 cm. Compared with soybean, mung bean had a larger percentage of root surface area in horizontal 0-10 cm soil layer. In vertical direction, soybean and mung bean had a similar distribution of root surface area in the upper, middle and lower layers. For root dry weight distribution, the dry root weight of soybean (49.8%) and mung bean (57.8%) were mainly distributed in 0-5 cm horizontal soil layer, and both dry weight of soybean and mung bean in vertical 0-20 cm soil layer reached more than 70% of the total dry weight, more than 82% of total root dry weight in vertical 0-40 cm soil layer. This is similar to the finding of Mitchell and Russell (1971) who found the highest proportion of root dry weight density in vertical 0.23 m soil layer.
A low AWR indicates either a thicker root system or roots with higher specific density (Benjamin and Nielsen, 2006). Mung bean had greater AWR than soybean during all growth stages. This showed mung bean had a finer root system or roots with lower specific density. From V6 to R7, soybean and mung bean AWR decreased indicating a thickening or densification of the root material.
In horizontal devices, we found that the maximum of total root length, total root surface area, and total root volume of soybean and mung bean were at R5. However, in vertical device, the maximum of total root length, total root surface area, and total root volume of soybean were at R5. But for mung bean, the maximum of total root length and total root volume were found at R4 and the maximum total root surface was at R2. The reason may be that limiting the extension of roots in horizontal direction accelerated the aging process of mung bean roots, but soybean roots showed stronger adaptability than mung bean.

Conclusions Conclusions Conclusions
Compared with mung bean, soybean had a much larger root system during development; In horizontal direction, root system was mainly concentrated in the 0-5 cm soil layer, but soybean root tended to be more laterally developed compared with mung bean. In vertical direction, the distribution of mung bean root was more uniform than that of soybean; With a greater AWR, mung bean had a finer root system than soybean.
Authors' Contribution HZ: Collection of samples, data collection and analysis, article writing; DZ and NF: Guidance on methods. All authors read and approved the final manuscript.