Effects of drought stress on root morphology and spatial distribution of soybean and adzuki bean

Hyen Chung CHUN,Sanghun LEE,Young Dae CHOI,Dong Hyeok GONG,Ki Youl JUNG

Crop Production Technology Research Division,National Institute of Crop Science,Miryang-si 50424,Republic of Korea

Abstract Due to global climate change,Korea is facing severe droughts that affect the planting and early vegetative periods of upland crops.Soybean and adzuki bean are important legume crops in Korea,so it is critical to understand their adaptations to water stress.This study investigated the changes in root morphological properties in soybean and adzuki bean and quantified the findings using fractal analysis.The experiment was performed at the National Institute of Crop Science in Miryang,Korea.Soybeans and adzuki beans were planted in test boxes and grown for 30 days.The boxes were filled with bed soil with various soil moisture treatments.Root images were obtained and scanned every two days,and the root properties were characterized by root length,depth and surface area,number of roots,and fractal parameters (fractal dimension and lacunarity).Root depth,length and surface area and the number of roots increased in both crops as the soil moisture content increased.The fractal dimension and lacunarity values increased as the soil moisture content increased.These results indicated that the greater the soil moisture,the more heterogeneous the root structure.Correlation analysis of the morphological properties and fractal parameters indicated that soybean and adzuki bean had different root structure developments.Both soybean and adzuki bean were sensitive to the amount of soil moisture in the early vegetative stage.Soybean required a soil moisture content greater than 70% of the field capacity to develop a full root structure,while adzuki bean required 100% of the field capacity.These results would be useful in understanding the responses of soybean and adzuki bean to water stress and managing irrigation during cultivation.

Keywords:soybean,adzuki bean,root,morphology,fractal dimension,soil water content

1.Introduction

Soybean (Glycine max(L.) Merr.) and adzuki bean (Vigna angularisL.) are representative legume crops and major food sources of protein in Korea (Kimet al.1988).These crops are used as the main ingredients in vegetable oil,fermented foods and processed foods in East Asian countries.The application of soybean and adzuki bean has widened across the food industry and other industrial circles.However,annual production of soybean and adzuki bean has decreased over the years because of decreases in arable lands and changes in environmental conditions,such as varying rainfall and temperature induced by global climate change (RDA 2018a,b).

Climate change has recently affected the rainfall patterns in Korea,and this has increased the occurrence of chronic drought from spring to summer (Parket al.2012).The average temperature in June increased from 23.3°C in 2014 to 25.1°C in 2018 and the average rainfall decreased from 213.8 mm in 2014 to 202.7 mm in 2018 (KSIS 2019).This climatic condition has affected the planting season and seedling stage of upland crops such as sesame,soybean and adzuki bean (Chunet al.2016).With the rainfall pattern change in Korea,the soybean and adzuki bean yields were reduced by 6.1 and 21.7%,respectively,in 2018 due to the drought during the seedling stage (KSIS 2019).

Drought damages crops by inhibiting and disrupting photosynthesis,which is the main mechanism of crop growth and yield (Dornboset al.1989).Studies on drought stress in crops have mainly focused on reproductive development because drought in the reproductive stage causes greater yield reduction than that in other stages (Turket al.1980).However,several studies found that drought stress on seedlings or the early vegetative stage somewhat affected crop growth and yield.Drought conditions at the soybean seedling or vegetative growth stages showed significant decreases in branch number and this led to a reduction in soybean yield (Fredericket al.2001).It is important to understand the drought response at the seedling and early vegetative stages.Therefore,the drought effects on soybean and adzuki bean during the planting and early growth stages should be investigated.

The effects of drought stress on soybean has been studied for years (Sinclair 1986;Earl 2002;Hatiet al.2006).Candoganet al.(2013) found that an increase in drought stress decreased soybean yield and Heet al.(2017)reported that water stress from the germination stage until maturity decreased the yield and water-use efficiency.These results showed that soybean was susceptible to changes in the soil moisture.However,there has been no study on the response of adzuki bean to drought conditions.Based on previous studies,it is reasonable to assume that the soil water contents in the early growth stage would affect the yields of both crops.

To date,many researchers have focused on the shoot response to drought rather than the root response (Whiteet al.1990;Huanget al.1997) because it is difficult to observe the root response in soil.However,the roots are the first responders to drought conditions.Previous studies showed that roots could sense water deficit conditions in soil and then signaled to reduce stomatal conductance and leaf growth (Gollanet al.1986;Davieset al.2002;Schachtman and Goodger 2008).Therefore,it is important to understand the relationship between deficit soil water conditions and root adaptation (Fukai and Cooper 1995;Priceet al.2002).Many studies have investigated the roots of crops or plants under water stress by measuring the root dimensions,such as root length,surface area or volume (Osonubi 1985;Zhang and Davies 1989;Liet al.2006;Caiet al.2017),the elongation rate (Whalleyet al.1998;Bengoughet al.2011),and the spatial structure (Fitter and Stickland 1992;Nielsenet al.1997;Wanget al.2009).These studies highlighted the need to understand the effects of water deficit stress on crop root systems and commonly concluded that crop roots were sensitive to water stress and root properties were affected by soil moisture conditions.Soybean roots under drought stress have been studied for the changes in root biomass(Francoet al.2011),diameter (Tanakaet al.2014) and root length (Thuet al.2014).Hoogenboomet al.(1987) found that soybean roots experiencing drought stress developed roots deeper into the soil than roots in well-watered conditions,which has been confirmed by previous studies.However,there has been no research on the changes in adzuki bean roots in drought conditions.

The architecture of the root system is also well known to be a determinant of root function in the acquisition of soil water (Yamauchiet al.1996;Fitter 2002).Root structure or root system,specifically when under stress,can be affected,which can be quantified by the spatial distribution analysis of roots.Fractal and lacunarity analyses are spatial distribution analyses that have been proven to be sensitive enough to characterize the heterogeneous structures or distributions of various objects (Pozdnyakovaet al.2005;Chunet al.2008).The fractal dimension measures the complexity in roughness or distribution,and lacunarity measures structural variation(Smithet al.1996).Fractal and lacunarity analyses have been applied in various fields,including ecological structures(rock,river,etc.) (Burrough 1981;Zuoet al.2009),neuron/gene structures (Sharon-Rubiniet al.2018),soil/aggregate structures (Zenget al.1996;Gimenezet al.2002),and plant identification (Brunoet al.2008;Chandra and Rani 2009).These analyses have also been applied to quantify plant root systems or structures (Tatsumiet al.1989;Fitter and Stickland 1992;Eghballet al.1993;Haucket al.2015).Wanget al.(2009) used fractal analysis to investigate rice root changes in response to drought stress and found that the fractal dimension characterized the root development affected by drought conditions.The authors concluded that the root structures tended to decrease in heterogeneity as the drought stress increased.However,there has been no study applying spatial analyses to soybean and adzuki bean roots under drought conditions.

This study hypothesized that differences in the soil moisture content would induce changes in the morphological properties and spatial distributions in soybean and adzuki bean roots in early development.The objectives of this study were to characterize the morphology and spatial distribution of soybean and adzuki bean roots and compare the root development between soybean and adzuki bean under various soil moisture contents.

2.Materials and methods

2.1.Crop and soil

The experiment was performed in a greenhouse located in Miryang,Gyeongnam,Korea from June to September 2019.The soybean cultivar was Dawon and the adzuki bean cultivar was Arari.Commercial horticultural bed soil was used to grow the soybean and adzuki bean.The bed soil consisted of coco peat 65–70%,peat moss 8–12%,vermiculite 10–14%,zeolite 3–5%,perlite 5–8%,watersoluble fertilizer,a microbial substance,and a humectant(Seoulbio,Korea).The physical and chemical properties of the soil are shown in Table 1.

2.2.Experimental test box

Plastic boxes with thick edges were used to create an opaque container for soybean and adzuki bean planting(Fig.1).The upper side was opened to fill with the soil and grow the crops.The boxes had dimensions of 40 cm×21 cm×4 cm,and the total number of boxes was 65,including five replicates and five control boxes.Six moisture treatments were set,with soil moisture contents of 30% (B30box),40% (B40box),50% (B50box),60% (B60box),70% (B70box),and 100% (B100box) of the field capacity.The soil moisture contents were determined by weight saturation,and the water amounts were calculated to determine the target moisture content.Each box contained the same amount of soil (1 kg) and specific amounts of water were added to reach the target moisture levels (Table 2).The soil moisture of the control boxes was close to 100% of the field capacity throughout cultivation to avoid drought stress.The soybean and adzuki bean seeds were planted at 2–3 cm depths from the soil surface.No boxes except the control boxes had additional water after planting.The soil moisture contents were measured every hour by an SM 300 soil moisture sensor (WatchDog,Spectrum Technology,Inc.,USA) connected to the logger of a WatchDog 2400.

Table 2 Amount of water added to 1 kg of bed soil in a test box to achieve the target moisture content

Fig.1 Diagram of a test box for the root scan.

Table 1 Summary of bed soil chemical and physical properties1)

The detailed procedure for root scanning is shown in Fig.2.After planting,root scans were performed every 2 d for 30 d.After scanning,the roots in the scanned images were extracted from the background of the soil,and the extracted roots were converted into black and white images.The morphological properties of the roots in the scanned root images were measured using WinRHIZO Software(WinRHIZO Tron,Canada).This software measured the root length,the number of roots and the root surface area of each root image.The maximum root depths were also measured on the scanning date and were determined by measuring the depth of the main root.All roots were harvested after 30 d of planting.This crop cultivation procedure was repeated a second time,and a total of 130 boxes were analyzed to investigate the root growth.The roots from the B30,B40,B50,B60,and B70boxes withered around the 18th d after planting.Therefore,the root images from these boxes were extracted until the 20th d for the analyses.

Fig.2 The process of root scanning and black and white root image extraction.A,soybean root in a test box;B,test box scanning with a scanner;C,black and white root image extracted from image 1.

2.3.Fractal analysis

The digitized root images were converted into binary images and processed using fractal analysis with ImageJ (National Institutes of Health,USA).The image size was 2 819×3 779 pixels.The fractal analysis estimates heterogeneity by applying a linear regression of the box counts and box sizes.The box sizes ranged from 0 to 45% of the image sizes.A series of box sizes were applied to count the number of boxes containing roots.This process was repeated until the entire image was covered (Tarquiset al.2003).

The fractal dimensions (Db) were estimated by applying a linear regression of these relationships,and their slopes were the–Db values.The fractal dimensions from the root images were obtained from the number–size relationship(Gimenezet al.1997).

where N(C) is the number of roots at box sizeCand K is the number of initiators of the unit length and is assumed to be 1.

Lacunarity (Lac) is defined in terms of the mass of the related distributions.It calculates the variation in root pixel density at different box sizes:

where Lac is the lacunarity value over all of the box sizes,σis the standard deviation,andμis the mean number of roots boxes at box sizeC.

2.4.Statistical analysis

Statistical analysis of all the measured results were performed using analysis of variance (ANOVA) at the 95% significance level.All root measurements were subjected to a correlation analysis at the 95% significance level.The relationship between morphological properties and fractal parameters was determined by Pearson’s correlation analysis.All analyses were performed using SPSS Software (v.9.0.1,SPSS Inc.,USA).

3.Results

3.1.Soil moisture contents

The soil moisture content was measured every hour during crop growth,and the measurements are shown in Fig.3.The control boxes had (40.2±4.21)% soil moisture content on the planting day,and water was added every day to avoid drought stress to the roots.The B100boxes initially had (41.4±5.16)% soil moisture,which was the highest among all the boxes,and the lowest moisture content was in the B30boxes at (14.0±4.10)%.Over time,soil moisture decreased by evaporation and uptake by the roots.The soil moisture content in the B100boxes was (7.2±3.01)%,and that of the B30boxes was (1.1±2.31)%,on the 20th d after planting.The soil moisture content in the B30boxes was less than 10% at 10 d after planting,and that of the B40boxes was less than 10% at 14 d after planting.The soil moisture content remained over 10% in the B100and B70boxes until the 18th d after planting.After 18 d,most of the root boxes except the control boxes had soil moisture contents less than 10%,so most of the roots in the root boxes withered.

Fig.3 Average of soil moisture content in the test boxes.DAP,days after planting.30–100%,B30–B100 box,respectively;no stress,the test boxes had no water stress.Bars mean SD (n=5).

3.2.Characteristics of root morphological features

The soybean and adzuki bean roots were scanned every 2 d to measure the root length,the number of roots,and the surface area and depth (Fig.4).The roots in the B30,B40,B50,and B60boxes withered due to lack of soil moisture around the 18th d.Therefore,the root morphological properties were measured until 20 d after planting.The root length,surface area and depth and the number of roots of soybean were not different in soil with different moisture contents until 6 d after planting (Fig.5,P>0.05).The roots from the control boxes and the B100boxes were greater than those from other soil moisture boxes from the 8th to the 30th d.The greatest root length was (34.95±1.15) cm,in the B100boxes,whereas the B30boxes had the smallest length at(12.49±3.82) cm (P=0.00).The root lengths in the control boxes,which were not significantly different from those in the B100and B70boxes (P>0.05),were statistically greater than those in the B60,B50and B40boxes (P=0.02).

Fig.4 Examples of black and white root images from scanned root images with no water stress after planting.A,soybean.B,adzuki bean.DAP,days after planting.

Fig.5 Average of morphological analyses of soybean root images in various soil moisture contents after planting.A,root length.B,number of roots.C,surface area of the roots.D,root depth from the soil surface.DAP,days after planting.No stress,the test boxes had no water stress;30–100%,B30–B100 box,respectively.Bars mean SD (n=5).

The roots in the control and B100boxes showed the greatest number of roots and surface area,whereas the roots from other soil moisture content boxes were not significantly different from each other.The roots from the control and B100boxes exhibited a continuous increase in the number of roots and surface area after planting.The roots from the other soil moisture content boxes showed increases in the number of roots and surface area until 12 d after planting and then decreased after 12 d.Overall,the roots from the control,B100and B70boxes had the greatest length,surface area,depth and number of roots (P<0.05).In contrast,the roots of the B30boxes had the smallest root length and the least depth during cultivation.

The morphological properties of the adzuki bean roots exhibited no differences across the different soil moisture contents until 8 d after planting,similar to soybean roots(Fig.6).The root length and depth in the control and B100boxes were significantly increased compared to those in the other soil moisture content boxes after 8 d.The root length of the B100boxes was the longest,at (32.40±3.2) cm on the 20th d after planting,whereas the roots in the B30boxes were the smallest at (8.71±2.65) cm (P=0.02).The roots in the control,B100and B70boxes steadily increased in length and depth until the 20th d,but the roots in the B60,B50,B40,and B30boxes showed decreases or no change 16 d after planting.The number of roots and the surface area were the greatest in the control and B100boxes compared to those in other soil moisture boxes.All roots from all soil moisture boxes exhibited steady increases in the number of roots and surface area until the 18th d and then wilted or withered 18 d after planting.

Fig.6 Average of morphological analyses of adzuki bean root images in various soil moisture contents after planting.A,root length.B,number of roots.C,surface area of the roots.D,root depth from the soil surface.DAP,days after planting.No stress,the test boxes had no water stress;30–100%,B30–B100 box,respectively.Bars mean SD (n=5).

The comparison between soybeans and adzuki beans showed that,in the control and B100boxes,the soybean roots had greater root length and depth but smaller number of roots and surface area than the adzuki bean roots.The morphological properties of the soybean roots from boxes with lower soil moisture content were not significantly different from those of the adzuki bean roots (P>0.05).

3.3.Characteristics of root spatial distribution analysis

All scanned images of the roots were analyzed using fractal analysis (Fig.7).The Db and Lac values were calculated from each root image according to eqs.(1) and (2).The fractal dimension values were determined by the slope of the regression line between the box counts with root appearance and box sizes.In general,non-linear regression has a smallerr2value,and this means that the analyzed objects have no fractal structure (Klinkenberg 1994).Ther2values of the regression lines of the roots from the 2nd to 6th d were less than 0.9,indicating non-linear regression.Therefore,the roots in early development did not have any fractal structure (Fig.8).Based on these results,Db and Lac values were calculated from the root images 8 d after planting.The soybean roots of the control and B100boxes had significantly greater Db values (1.7±0.12) than those from the other soil moisture content boxes 20 d after planting (Fig.9;P=0.00).The Lac values across different soil moisture contents were not statistically different until the 14th d (P>0.60).After the 14th d,the Lac values of the control and B100boxes were 0.72±0.20 and 0.83±0.21,respectively,whereas the Lac values of the B50and B30boxes were smaller than those from boxes with greater soil moisture contents (P=0.01).The roots from the B100boxes had greater Lac values than those from the boxes with other soil moisture contents on the 20th d after planting.

Fig.7 Examples of black and white root images in different soil water contents on the 20th d after planting.A,soybean.B,adzuki bean.30–100%,B30–B100 box,respectively.No stress means the test boxes had no water stress.

Fig.8 Example of fractal dimension results of root images on various days after planting (DAP).A,4th d after planting (DAP 4).B,8th d after planting (DAP 8).

Fig.9 Average of fractal analyses of soybean root images in various soil moisture contents after planting.A,fractal dimension(Db).B,lacunarity (Lac).DAP,days after planting.No stress,the test boxes had no water stress;30–100%,B30–B100 box,respectively.Bars mean SD (n=5).

Unlike soybean,the adzuki bean Db values showed a steady increase with time after planting (Fig.10).The adzuki bean Db values of the control and B100boxes were greater than those with other soil moisture contents from the 10th to the 20th d after planting (P<0.02).The adzuki bean Db values of the control and B100boxes were greater throughout the cultivation period.The Db values of the control and B100boxes were 1.28±0.33 and 1.33±0.07,respectively,on the 20th d after planting.There was no difference in the adzuki bean Db values for all soil moisture contents except for the control and B100boxes.The Lac value of the B100boxes was 0.88±0.18 on the 20th d after planting,which was greater than that of the boxes with other soil moisture contents (P=0.01).The Lac values were not statistically different among the boxes with other soil moisture contents.

Fig.10 Results of fractal analyses of adzuki bean root images in various soil moisture contents after planting.A,fractal dimension(Db).B,lacunarity (Lac).DAP,days after planting.No stress,the test boxes had no water stress;30–100%,B30–B100 box,respectively.Bars mean SD (n=5).

The soybean Db values were significantly different from the adzuki bean Db values 6 d after planting.Specifically,the soybean Db values of the control and B100boxes were greater than those of the adzuki beans.In contrast,the adzuki bean Lac values of the control and B100boxes were not different from the soybean Lac values throughout the growing period.The soybean Db and Lac values in the B60,B50,B40,and B30boxes were greater than those of the adzuki beans (P<0.05).

4.Discussion

The morphological characteristics of soybean and adzuki bean roots were quantified in this study.Various soil moisture contents were set from planting to investigate crop root response to water stress during the early growth stage.It is well known that root systems can change their morphological structure in response to the surrounding soil environment such as soil moisture (Zhanget al.2019).Root development in soybean and adzuki bean was found to be affected by the soil moisture content in this study.Root development was better under the control and 100% field capacity conditions than in soil with other moisture contents.Specifically,the roots were longer,deeper,more numerous and covered a wider surface area in the control boxes and the boxes with 100% field capacity of soil moisture than in any other boxes with lower soil mosture.As soil moisture decreased,the soybean and adzuki bean roots exhibited significantly slower and poorer development.Mackay and Barber (1985) investigated corn root growth from various soil textures under different soil moisture contents.As soil moisture increased from 0.22 to 0.27%,root length increased from 41 to 52% in all the different soil types.This conclusion was confirmed by research on different crops,such as maize (Lal 1978),lemon (Bouma and Bryla 2000)and sorghum (Monti and Zatta 2009).The results of the present study were consistent with those of previous studies.

Soybean roots exhibited no decrease in morphological properties with decreases in soil moisture until 70% of the field capacity.In contrast,the length,depth,surface area,and number of adzuki bean roots decreased as soil moisture decreased from 100% of the field capacity.There have been no studies comparing soybean and adzuki bean under drought stress conditions.However,the sensitivity of soybean and adzuki bean to water stress was measured under excessive soil moisture conditions.Chunet al.(2018) tested excessive soil moisture stress on soybean and adzuki bean and concluded that soybean was less sensitive to water stress than adzuki bean.Based on the results of root growth in this study,in the case of drought conditions,adzuki bean roots were more sensitive to soil water reduction than soybean roots that were not affected by soil water reductions to 70% of the field capacity.Among the morphological properties,the root length and depth values of soybean were greater than those of adzuki bean,whereas adzuki bean had a greater number of roots than soybean.These results suggested that the adzuki bean roots were more widely distributed in a shallower location than soybean roots.A previous study found that plants with deeper root development at the seedling stage tended to survive drought stress better than plants with shallower root development (Padilla and Pugnaire 2007).Based on these findings,adzuki bean was impacted more by drought stress than soybean,as the former had shorter and shallower roots than the latter.

Many studies have reported that crop roots tend to grow deeper into the soil under drought conditions (Benjamin and Nielsen 2006;Kashiwagiet al.2006;Caiet al.2017).These conclusions were based on a common experimental condition;applying water stress at least two weeks after planting.Guoet al.(2013) investigated the germination rate and root growth of wheat by applying deficit water stress from planting.As a result,they concluded that deficit water stress from planting was the main factor limiting the growth and development of wheat roots and stalks.Sadeghian and Yavari (2004) reported that the drought level and root growth were inversely related if the drought occurred during planting and the early growth stages.The results of the current study also confirmed that greater drought stress led to less root growth during early root development.These findings suggested that poor development at the seedling stage would decrease crop productivity,consistent with the conclusion of a previous study (Monti and Zatta 2009).

Based on the Db and Lac results from soybean and adzuki bean roots,the root structures were affected by the soil moisture content.As the soil moisture increased,the Db and Lac values of the soybean and adzuki bean increased.Tatsumiet al.(1989) reported that the shapes of the root systems were closely related to the fractal dimension.The Db parameter indicates the degree of root-growing directions based on the Db calculation.If the roots spread out in more diverse directions,the Db value increases.An increase in the Db value represents the development of a more complex structure.Therefore,greater soil moisture resulted in more complex root structures.Fitter and Stickland (1992)measured the fractal dimension in grass roots for 30 days and found that Db values could be an index of the complexity of the root system.In this study,the Db values showed that higher soil moisture induced more complex root structures.In addition,soybean developed more complex root structures than adzuki bean.The Lac parameter indicates the degree of uniform distribution of the roots (Plotnicket al.1996).If the roots spread out equally in every direction,the Lac values decrease.Greater Lac values from higher soil moisture content indicated that the roots developed mainly in one or two directions,whereas fine roots did not develop in all directions.As drought stress increased,the Lac values decreased,indicating that the roots stretched out evenly.Thus,these results indicated that the main root development was poor,and the development of fine roots was relatively better.Wanget al.(2009) calculated the Db values of rice roots under drought conditions and found a decrease in the Db values as the drought stress increased.Pierceet al.(2012) reported that an increase in the soil moisture content led to an increase in the Db in plant roots.So far,there has been no research on the characteristics of the Lac values from roots under drought stress.However,the Db and Lac parameter definitions and previous studies on Db have reasonably inferred that an increase in soil moisture content results in increases in root development and the complexity of the root structure.Based on these findings,it is reasonable to conclude that drought stress reduced the complexity of root structure while increased the fine root development.Moreover,soybean developed more main roots and complex root structure than adzuki bean under drought stress.

Wanget al.(2015) found a strong relationship between the fractal parameters and certain soil properties from reconstructed soils and Walket al.(2004) reported a relationship between fractal parameters and root properties.In this study,all measured parameters (root depth,length,surface area,and the number of roots)and fractal parameters (Db and Lac) were applied to a correlation analysis (Tables 3 and 4).The results showedthat the soybean Db values were significantly related to root length and depth,whereas the Lac values had a stronger relationship with the number of roots.These results implied that the heterogeneity or complexity of root structure was influenced by the development of the main root,whereas the Lac values were influenced by the development of fine roots.The Db and Lac values of adzuki bean had stronger correlations with root depth,surface area and the number of roots.These results showed that the structural heterogeneity of adzuki bean was not affected by the main root but mainly by the fine roots.In other words,the fine roots,rather than the main root,of adzuki bean were the factor influencing its root structural heterogeneity.

Table 3 Values of Pearson’s correlation coefficient between the morphological properties (length,depth,surface area,and the number of roots) and fractal parameters (fractal dimension(Db) and lacunarity (Lac)) of soybean roots

Table 4 Values of Pearson’s correlation coefficient betweenthe morphological properties (length,depth,surface area,and the number of roots) and fractal parameters (fractal dimension(Db) and lacunarity (Lac)) of adzuki bean roots

Crop growth,such as leaf and stalk growth,is strongly related to root growth at the seedling stage.The ability to vigorously uptake water and nutrients from the roots promotes crop processes,including photosynthesis,translocation and mineral nutrition.These processes result in fertile yields.Under deficit water conditions,crops or plants focus on root growth and slow down the growth of leaves and stalks (Hsiao and Xu 2000).Continuous water stress would eventually reduce root growth and crop growth-related processes.Pritchardet al.(1990) developed a yield model related to root growth and root development.Based on the results of this model,low root growth results in low yield.Frensch and Hsiao (1995) predicted that water stress at the seedling stage ofZea maysL.reduced root elongation,nutrient uptake,and yield.Previous studies showed that severe water stress led to a reduction in root growth,stalk and leaf growth processes,and yield.Based on the results of this study,it is reasonable to conclude that drought stress at the seedling stage would affect root growth,stalk growth,and yield if water stress to the roots was not relieved.These results may be very helpful in developing more complete plant root structures and guiding water management techniques for soybean and adzuki bean from sowing to the early development period.

5.Conclusion

Soil moisture content affected the root development of soybean and adzuki bean during the sowing period.Soybean showed greater main root development than fine root development under sufficient soil moisture conditions at the early growing stage.As the soil water deficit stress increased,the development of both the soybean main root and fine root decreased.Adzuki bean roots also followed a similar trend in the main root development as water stress increased.Soybean roots were not affected by water stress until the soil moisture decreased to 70% of the field capacity at planting,whereas the adzuki bean roots were affected by water deficit once the soil moisture was less than 100%of the field capacity.However,the fine adzuki bean roots did not decrease under drought stress.These findings indicated that soybean and adzuki bean have different patterns of root development,even though they are in the same legume family.In addition,the root response to variations in soil moisture was significantly different between soybean and adzuki bean.The fractal analyses quantified and distinguished the root development between soybean and adzuki bean.The spatial analysis may be useful in distinguishing the changes in different crops’ root structure caused by various soil moisture contents.The results of this study suggest that conservative water management may be useful in achieving optimal soybean and adzuki bean root development and minimizing crop losses due to drought.

Acknowledgements

This research was performed and funded by an Agenda Project of the Rural Development Administration(PJ 013482022020),Republic of Korea.

Declaration of competing interest

The authors declare that they have no conflict of interest.