Photosynthesis, biomass and fine root growth dynamics of soybean in walnut-soybean agroforestry system

Background: Agroforestry system is regarded as a promising practice in sustainable agricultural management. However, the effects of long-term tree-based intercropping on crop remain poorly understood, especially in the Loess Plateau. In this study, the impacts of photosynthetic and respiration rate were determined by the portable photosynthesis system (Li-6400), and the effects of the root growth dynamics of soybean in walnut-soybean intercropping system were measured by soil auger and WinRHIZO root analysis system, in the Loess Plateau. Results: The results showed that soybean reached the highest net photosynthetic rate during flowering period, with the net photosynthetic rate of intercropped soybean, which was 20.40µmol·m-2·s-1, significantly higher than that of its monocropped counterpart. Soybean biomass reached the maximum during the pod-bearing period, with intercropped soybean biomass being 25.49g, significantly higher than that of its monocropped counterpart. The mean diameter and increased density of soybean fine roots reduced along with increased soil depth. Both the diameter (0.43mm) and increased density (930cm/dm3) of intercropped soybean fine roots were evidently higher than those of monocropped soybean(0.35mm, 780cm/dm3). With increasing cropping years, fine roots of intercropped soybean tended to be mainly distributed in soil at a depth between 0 and 20cm from the fifth year. Conclusion: Collectively, compared with soybean monoculture, walnut soybean agroforestry system is more conducive to soybean growth in the Loess Plateau.


Background
The Loess Plateau is the cradle of the Chinese nation and the ancient civilization. As an artificial multispecies vegetation system, the agroforestry system entails different functional organs in the spatial niche, varied functional activity cycles on the temporal dimension, competition adaptability in term of biomorphic function and species distribution patterns on the spatial dimension. Theoretically, these give rise to a composite system. The competition and interplay between different species in a multi-species ecosystem determines the stability and sustainability of the system itself. Considering the potential benefits of tree-based intercropping systems, an appropriate agroforestry system can be a foundation for effective resource utilization. However, in actual applications, the benefits of agroforestry may disappear as tree shades can reduce crop photosynthesis and competition between root systems affect crop's absorption of water and nutrients.
Focusing on the walnut-soybean agroforestry system in the Loess Plateau of China, this paper explores the influence of soybean photosynthesis, as well as the competitive relationship between soybean and walnut root systems in order to provide feasible proposals for constructing highly efficient agroforestry systems and maximally increasing crop yields.

Results
The most evident difference between intercropped and monocropped soybean lies in a larger competitiveness of intercropped soybean in terms of transpiration rate, net photosynthesis rate and stomatal conductance during the flowering period (Table 1). For both intercropped and monocropped soybean, the transpiration rate, net photosynthesis rate and stomatal conductance increased and then decreased over time, reaching their maximum values during the flowering period.
The biomass of intercropping soybean was significantly heavier than that of monoculture soybean at podding stage (Table 2). At flowering stage, root biomass of intercropping soybean was significantly higher than that of monoculture soybean. For intercropping soybean and monoculture soybean, the fresh weight, dry weight of single plant, biomass of above ground and biomass of under ground of soybean increased with time, and reached the maximum at the podding stage of soybean.
The competition intensity between walnut and soybean decreased with the increase of soil depth (Table 3). In 0-20cm, walnut and soybean roots showed the greatest competition.
The soil moisture content gradually decreases with the increase of soybean growth period and soil depth (Figure 1). At 0-20cm, the moisture content of the walnut-soybean agroforestry system was significantly higher than that of the single-cropping soil at the flowering stage (P<0.05).
Both intercropped and monocropped soybean had only one root growth peak at the soil depth of 0-60cm, with the peaks being reached basically at the same time ( Figure 2). Starting from the 5th year, intercropped soybean had a significantly greater root growth peak (P<0.05) at 0-20 cm soil depth than its monocropped counterpart. Starting from the 50th after soybean planting, the monocropped soybean had a significantly greater root growth peak (P<0.05) at 20-60 cm soil depth than its intercropped counterpart.
There were no significant changes in soybean fine root length densities at 0-60 cm soil depth with the increasing planting years( Figure 3). The root length density of intercropped soybean at 0-20 cm soil depth increased with the increasing planting years. Starting from the 5th year, the root length density of intercropped soybean was significantly higher than its monocropped counterpart (P<0.05). The root length density of intercropped soybean at 20-60 cm soil depth decreased with the increasing cultivation years, and started to be significantly lower than that of the monocropped soybean (P<0.05) from the 7th year.
The mean fine root diameter of intercropped soybean at 0-20 cm soil depth was significantly larger than that of the control group( Figure 4). The mean fine root diameter of intercropped soybean during the flowering period was significantly larger than that of the control group (P<0.05), but the same diameter at 20-40 cm soil depth was smaller than that of the monocropped soybean during the sowing period. The mean fine root diameter of intercropped soybean at 40 cm soil depth was significantly smaller than that of the control group (P<0.05).

Discussion
According to the preliminary data published by the General Administration of Customs of China in 2016, soybean imports have exhibited a growing trend over recent years, and soybean is proven to generate more economic benefits than other crops (Peng 2009). To alleviate poverty and soybean import pressure, the Chinese government has started to encourage soybean cultivation among farmers. Exploring the interspecies competitive pressure exerted on soybean in the walnut-soybean agroforestry system is of great significance.
Shading influences crop's net photosynthesis rate and thus its yields. Nurembe (2012) found that the walnut-soybean agroforestry system had a limited shading influence on soybean, which is consistent with our experimental results. A study conducted by Miller (2014) indicated that competition over water resources represented a key factor influencing crop yields in the walnut-soybean agroforestry system. An appropriate level of shading effectively prevents soil moisture evaporation, increases soil 5 moisture concentration, contributes to the increase in net photosynthesis rate and thus improves the biomass of soybean (Luo 1995). Such a conclusion is also supported by the results of this research.
According to the study of , yields of intercropped soybean decreased by 29% compared with those of monocropped soybean. This may be attributable to the relatively thinner soil layer in Zhang 's experiment site located in the Taihang Mountain region, which can result in unfixed ecological niche of soybean roots and thus higher pressure of interroot competition. Results from this study showed that the walnut-soybean agroforestry system can effectively improve soybean yields.
The growth and root length density peaks of intercropped soybean at 0-20 cm soil depth were higher than those of its monocropped counterpart, and the growth and fine-root length density peak values of intercropped soybean at 20-60 cm soil depth were lower than its monocropped counterpart, which are consistent with findings of  and Wang (2011). Yang (2007) argued that root growth was associated with soil moisture, water obtained by soybean from the soil would be predominantly used for root growth, and insufficient soil moisture would stimulate root growth. Interspecies competition between the root systems of a walnut-soybean agroforestry system is quite intense at 0-20 cm soil depth, which stimulates, to some extent, the growth of soybean roots. Guan (2001) and Li (2010) asserted that the walnut-soybean agroforestry system effectively increased the moisture content of the 20-60 cm soil layer, and consequently, the root length density of monocropped soybean would be higher than that of the intercropped soybean at this soil depth. As such, the walnut-soybean agroforestry system can effectively improve the root length density of soybean and increase the crop's absorption and utilization of water from soil.
In a walnut-soybean agroforestry system, soybean root systems could achieve rapid growth within a short period of time, thus gaining an advantage in the interspecies competition. Results from this study showed that with the increasing planting years in a walnut-soybean agroforestry system, the central distribution of roots exhibits a declining trend for walnut and a climbing trend for soybean.
The results of this research are consistent with those of Fan (1999), Ma (2009) and .
Walnut is reported to produce a secretion called juglone, which is a toxic substance that can affect the growth of maize plants (Jose 2016). The study of walnut juglone by Angel (1993) and Jose (1998) 6 showed that the concentration of walnut juglone affected the photosynthetic efficiency of soybean.
However, the study of Jose (1998) showed that when the concentration of walnut juglone was lower than 10-6M, the concentration of walnut juglone would not affect the photosynthesis of soybean. The study of BÖHM(2006BÖHM( , 2010 on Walnut juglone to promote the lignification of soybean root system showed that the low concentration of walnut juglone (< 1 μ M) could inhibit the lignification of soybean root system and promote the growth of soybean root system. However, similar result that walnut juglone inhibited the photosynthesis and root growth of soybean has not been found in this research, which may be due to the low concentration of walnut juglone in the soil under field conditions. To sum up, due to the interspecies competition between the root systems in a walnutsoybean agroforestry system, the roots of intercropped soybean tend to be distributed in the 0-20 cm soil layer. In the 5th operating year of the walnut-soybean agroforestry system, a unique ecological niche of soybean roots starts to take shape in the soil.
Results from this study showed that the mean diameter of soybean fine roots reaches its maximum value during the flowering period, which may be due to the fact that the flowering season falls in late July when rainfall in the Qishan region reaches its maximum intensity, in the 0-60 cm soil layer.
Abundant soil moisture reduces the pressure of interspecies competition between soybean roots (Wang 2007), explaining why the mean diameter of soybean fine roots reaches its maximum value during the flowering period. In the 0-60 cm soil layer, the mean diameters of both intercropped and monocropped soybean reduce along with the increasing soil depth, which may be due to decreased water and nutrients with the increase in soil depth . In the 0-20 cm soil layer, the mean root diameter of intercropped soybean significantly increases compared with that of its monocropped counterpart. This may be caused by the phenomenon that the remnant pore spaces of walnut roots amid their turnover processes are utilized by soybean roots, and a relatively loose soil environment contributes to the increase of mean diameter of soybean fine roots. These are consistent with findings of Xu (2012) and Wang (2014). In the 40-60 cm soil layer, the mean root diameter of intercropped soybean is significantly lower than that of its monocropped counterpart. Soybean fine roots are mostly distributed in the 0-20 cm soil layer (Xun 2013). However, a study conducted by Wang (2018) 7 explored the mean fine root diameter of wheat in a walnut-wheat agroforestry system, finding that in the 0-60 cm soil layer, the mean fine root diameter of intercropped wheat is larger than that of its monocropped counterpart. This may be because that the walnut-soybean agroforestry system cannot give rise to a stable ecological niche of roots as the walnut-wheat agroforestry system does.
Therefore, the smaller mean fine root diameter of intercropped soybean in the 40-60 cm soil layer is a product of a survival strategy adopted by soybean roots in order to cope with the fierce interspecies competition (Sun 2002, Myers 2005, Yan 2013). In short, the walnut-soybean agroforestry systemwhose operation originated from late July when interspecies competition was the lowest -can provide favorable soil conditions for roots of intercropped soybean, which further contributes to soybean roots' general water and nutrient absorption from soil in the 0-60 cm soil layer.

Conclusion
In conclusion, we argue that with the increasing planting years, a fixed ecological niche for soybean roots will be established in the walnut-soybean agroforestry system. The interspecies competition of soybean roots in the walnut-soybean agroforestry system is favorable, to some extent, for root growth, which further facilitates soybean roots in the intercropping system to absorb water from soil, thereby improving soybean yields.

Plant materials and site description
The research area is located in the southern part of the Loess Plateau, belonging to the temperate continental monsoon climate. The experimental areas of the study was located in Zhangjiagou village (34°19 '36 "N, 107°37' 16" E) and Liujiagou village (34°19 '41 "N, 107°38' 40" E) in Qishan county, Shaanxi Province. The annual average sunshine hours are 2185h, the average temperature is 11.6℃, the extreme maximum temperature is 41.4℃, the extreme minimum temperature is -20.6℃, and the frost-free period is 198 days. The mean annual precipitation is 679mm, which can basically meet the needs of crop growth. However, due to the influence of monsoon climate, the annual seasonal distribution is not uniform. The precipitation in June to September accounts for about 70% of the annual precipitation. In the test area, the soil was mainly made of loess mother material, with organic 8 matter content of 2.45%, total nitrogen content of 0.116% and total phosphorus content of 0.098% (IUSS 2015).
Soybean is the major crop in the researched area. The crop under tree canopies selected in this research was soybean (Qindou 8). Xiangling, a local walnut variety that is most widely planted in the area, was selected. The walnut trees were planted in 2016 (1a) at 3m×6m row spacing. The width of soybean alleys was 4.8m, while that of walnut tree strips was 1.2m. Soybean was planted on June 6, 2017 at 30cm*30cm row spacing. The planting density of monocropped soybean was the same as that of its intercropped counterpart. An experimental plot used to measure dynamic changes in root length density of tree seedlings in different planting years was established in 2007, which underwent the same arrangements as in the experimental field described above.

Photosynthetic and respiration rate sampling
On three consecutive sunny days, the net photosynthesis rate (Pn), stomatal conductance (Gs), respiratory rate (Tr) and intercellular CO 2 concentration on the second fully expanded leaf of soybean, which were selected during 9:00-11:00 a.m., were measured with the portable photosynthesis system 100g of soil was weighed and place it in the oven at 105±2℃to dry the constant weight (W). Samples were taken during the branching, flowering and pod bearing periods, respectively.

See formula 1 in the supplementary files.
Collection and measurement of photosynthetic data Root sampling Root samples were taken with a soil auger. The total sampling depth was 60cm, with 10cm for each layer. Three sampling points distributed in a triangle form, far away from fringe areas, were established in the sample plot. Samples were taken during the branching, flowering and pod bearing periods, respectively. 9 Soybean fine roots were prepared through soaking, washing and separating procedures. Fine roots with a diameter less than 2mm were selected. WinRHIZO root analysis system was applied to obtain the morphological indicators of fine roots, such as root length density and fine root diameter.
After the fixed sample plots were established, a minirhizotron system (BTC-100 Borescope Root Ecology Monitoring System) was employed to carry out in-situ observation of seasonal growth dynamics of roots. Root imaging data were collected every 10 days over an entire growth cycle. Root length parameter was obtained using the Winrhizo-tron-MF to calculate root length density, which was followed by a calculation of fine root length for the observation period. Pianka(1973) niche overlap formula (symmetric α method) is adopted to calculate the competition intensity index among walnut soybeans, and the formula is shown as follows:

See formula 2 in the supplementary files.
A ih = A hi , and A ih ≤1, r is depth of soil layer, and r≤3, P ij and P hj are respectively the proportion of   T(walnut-soybean) CK (soybean) Table 3 The competition between the root of walnut and soybean in agroforestry system  The relative soil moisture content at different growth stages, different soil depths and different tillage methods. The same asterisk indicates that, through variance analysis, the age of agriculture and forestry compound has a significant influence, p<0.05.
17 Figure 2 The relative density of increased root in soybean under different growing period, different soil depth and different tillage mode. Asterisks indicate that the effect of agroforestry age is significant through ANOVA at p<0.05.
18 Figure 3 The relative density of soybean root under different growing period, different soil depth and different tillage mode. Asterisks indicate that the effect of agroforestry age is significant through ANOVA at p<0.05.

Figure 4
The relative average diameter of thin root in soybean under different growing period, different soil depth and different tillage mode. Asterisks indicate that the effect of agroforestry age is significant through ANOVA at p<0.05.

Supplementary Files
This is a list of supplementary files associated with this preprint. Click to download. formulas.docx