The Rhizosphere Effect of Plant Kin Recognition: Morphological Responses Combined with Soil Nitrogen Cycling


 Aims Kin recognition has been used to explain plant interactions among siblings, but the morphological-based conclusions are various and the mechanism is still fuzzy. Here, we tested the rhizosphere effect of plant kin recognition based on soil nitrogen (N) cycling resulted from root exudates, combined with plant fitness, morphological and physiological performances to examine how plants respond to kin neighbors. Methods One factorial experimental design of relatedness including either sibling or strangers of Glycine max was constructed. After growing about three months, plant morphological traits including plant height, specific leaf area (SLA) and root length as well as plant biomass; physiological traits including root activity, nitrate reductase (NR) activity and contents of chlorophyll; plant N use efficiency of each individuals were measured. Moreover, the production rate of root exudates carbon (C) and N, soil microbial biomass C and N, as well as genes amoA-AOAs, amoA-AOBs, nifH, nirK, nirS and nosZ genes related with soil N were assayed. Finally, the abundances of soil archaea, bacteria and fungi were quantified. Results Our study showed significant higher plant fitness and physiological growth and N use efficiency in siblings than strangers. The root secreted C rather than secreted N was sensitive to kin identity of G. max. Moreover, higher root secreted C quantity of sibling also ignited increasing of soil microbial biomass especially the abundance of Archaea community, and the abundance of amoa-AOAs gene compared to stranger soils. Finally, siblings increased the supply of soil available N and N use efficiency compared to strangers. Conclusions The rhizosphere changes induced by root exudation resulted in increased fitness and greater resource use efficiency among siblings compared to strangers. These findings suggest that the rhizosphere effect of soil microbial changes and soil N cycling and transformation triggered by the root-exuded C, could be a potential underground feedback mechanism for multiple kin recognition responses.


Introduction
It has been suggested that plants can recognize their kin neighbours and non-self neighbours, and this recognition can have affect direct and inclusive tness for plants growing with relatives (Hamilton, 1964).
In addition to reducing competition, kin recognition could also be revealed through increasing resource use e ciency, especially when the tness is not available for direct measurements (Cheplick, 1992  The rhizosphere microbes can be stimulated by root exudates and thus accelerate soil carbon and nitrogen turnover, affecting the nutrient acquisition of soil microorganisms and plants (Eisenhauer et al, 2012). Numerous studies show that there is a direct relationship between the exudates of plant root system and the soil enzyme activity, because plant root exudates themselves contain enzymes and certainly have impact on the original soil enzyme activity (Dijkstra et al, 2007). Moreover, 20-60% that carbon xed by plants through photosynthesis is transported to plant roots, and 40%-70% of which is released into rhizosphere as root exudates (Kuzyakov et al, 2000). Therefore, a large amount of carbon and nitrogen accumulate in rhizosphere soil and become the substrate of microbial activities, which affects the growth and structure of microbial community (Singh et al, 2006) as well as the release and circulation of soil nutrients mediated by soil microbes (Dinkerlaker et al., 1989;Fisk et al., 2015). It was found that the increased carbon-nitrogen ratio of plant root exudates could increase the abundance of soil microorganisms, because higher carbon-nitrogen ratio would promote soil microorganisms to secret more extracellular enzymes to accelerate the decomposition of organic matter and provide carbon and nitrogen for the growth of microorganisms (Zhang et al., 2013). Additionally, root exudates are suggested to be kin signals among plant relatives, but it remains unclear which root exudates participate in kin recognition (Biedrzycki et al., 2010). Therefore, the study of root exudates could not only clarify how the kin recognition signals are passed, but also whether the microbial responses and soil enzyme effect are induced by root exudates response to relatedness.

Experiment Design and Arrangement
Seeds of Glycine max (L.) merr. were collected from 4 mother plants of one cultivar growing at least 20 meters far away in different plot of the same farm in Gansu Province. Before pollination, we selected some plants with similar size and bagged then with customized transparent waterproof bags to prevent cross-pollination. Those seeds were separated into family groups so that we could construct groups of siblings (seeds from the same mother) and strangers (seeds from different mothers). After germination of seeds in distilled water and growth for three days, two healthy and similarly sized seedlings (to eliminate the size effect) including either siblings (Siblings treatment) or strangers (Strangers treatment) from 4 mother plants were planted randomly as pairs in one pot (10 cm height and 20 cm diameter). The distance between two seedlings was 10 cm and there were 12 replicates of each treatment (experiment 1). At the same time, the same arrangement was repeated for root secretion collection (experiment 2).

Growth Conditions and Management
The soil in planting pots contained 50% sand and 50% humus. We also creat six pots without plants as a soil control. Seedlings were then grown in a greenhouse at the Institute of Genetics and Developmental Biology, Chinese Academy of Sciences, Beijing, China. The temperature of the glasshouse was 28°C in the daytime and 18°C at night time, with a light:dark photoperiod of 16:8 h. The photosynthetic photon uency rate was 180 µM m − 2 s − 1 with relative humidity of approximately 60%. Pots were watered every three days to maintain soil moisture but without fertilization.

Sample Collection
After growing for three months in the greenhouse, plants in experiment 1 were harvested from each plot and 50 g of soil that around to the root system were collected and sieved to through 2mm aperture sieves as rhizosphere soil samples of each individual (45 g were stored at 4°C and 5g were stored at − 80°C respectively). Roots were washed for a while and put into 0. . The solutions in funnels were pumped out and the funnels were washed twice with the C-free nutrient and pumped out and both were discarded after which we repeated process, and collected all solutions in funnels and dried them under − 60°C using a freeze dryer. Then the sediments were diluted to 10 ml with sterile water for analysis C and N of root secretion. After the collection of root exudates, the fresh root and leaves were washed clearly by distilled water and stored at 4°C for physiological measurements of plant root activity, chlorophyll content, NO 3 − content and nitrate reductase (NR) activity both in siblings and strangers.

Measurement of Indicators
(1) Phenotypic Index plant height, the length of main roots and seed number were measured directly by a ruler and counting.
The shoot and root biomass were weighted separately after drying. Two medium sized leaves from each plant were used to measure the average leaf area using a scanner (Epson perfection V700 PHOTO, Long Beach, CA, USA). The speci c leaf area (SLA) was calculated by the ratio of leaf area and leaf biomass. (3) Physiological Index Root activity was determined by the triphenylte trazolium chloride (TTC) reduction method (Islam et al., 2007). Brie y, roots of siblings and strangers were reacted with 0.  (Lichtenthaler et al., 1998). As to the NO 3 − content, the lateral roots including the root tips were washed clearly by distilled water, weighted 1 g and grinded in 20% acetic acid solution, diluted with the solution to 20 mL. Then we added 0.4 g of the mixed powder (including 100 g of BaSO 4 , 2 g of alpha-naphthylamine,   , 1987). In brief, petri dish containing 10 g fresh soil sample was put into a vacuum dryer, with a beaker containing chloroform without ethanol and some glass balls, and a small beaker containing dilute sodium hydroxide solution. After the lid is closed, the sealing is con rmed and the vacuum pump is used to pump the vacuum to chloroform boiling for 5min, covered the vacuum dryer with a black plastic bag for 24h to sampling. After the fumigation, all the soil was transferred to a 250ml conical ask, and 80ml 0.5 M K 2 SO 4 was added, shaken for 1h. At the same time, extraction of 10 g soil without fumigation was performed with the same procedure. Total organic C and N in extracts were measured on a TOC analyzer (Milti N/C 2100S). The microbial biomass was calculated as fumigated soil microbial biomass C/N minus unfumigated soil microbial biomass C/N.

Statistics
Statistical signi cance was assessed using SPSS 16 (SPSS Inc., Chicago, IL, USA). After the normal distribution for the residuals and the homoscedasticity of the residual variance were tested, ANOVA was used to compare the differences of tness, biomass, morphology and physiology traits, soil N forms, enzyme activity, gene abundance and alpha diversity between siblings and strangers. A p-value of less than 0.05 was considered to be signi cant.

Kin Responses for Plant Characteristics
The results of ANOVA showed that siblings allocated more biomass to seeds in focal plants living with siblings than with strangers (P = 0.025), i.e. increasing the tness of kin groups. Moreover, there was higher shoot biomass (P = 0.009) in siblings than strangers (Fig. 1b), but no signi cant difference for root biomass (P = 0.179). Plant height (P = 0.011) and SLA (P = 0.041), corrected for biomass were greater in siblings than in strangers ( Table 1). The root length showed no difference between both groups (Fig. 1a). Siblings showed higher chlorophyll content (P = 0.024) and nitrate reductase activity (P = 0.019), but lower root activity (P = 0.062) compared to strangers ( Fig. 2c; Table 1). Although the plant N uptake rate ( Fig. 2a; Table 1) and the concentration of NO 3 − in plant roots (P = 0.480) did not differ between siblings and strangers of G. max (Fig. 2b; Table 1), higher total N use e ciency (lower N accumulation) (P = 0.048) of siblings than strangers suggested increased nitrogen use e ciency of siblings.
The root secretion rate of C (P = 0.007) in root exudates of G. max was signi cantly higher in siblings than strangers. However, the root secretion of N (P = 0.252) in exudates was not signi cantly different between siblings and strangers ( Fig. 3b; Table 1).

Kin Effects on Soil Characteristics and Soil Microorganisms
Siblings showed signi cant higher microbial carbon content (MC, P = 0.014) than stranger rhizosphere soils compare to microbial nitrogen (MN) (Fig. 3a). However, siblings maintained lower soil NO 3 − content than strangers (P = 0.008) compare to NH 4 + content of G. max (Fig. 3c; Table 1).
The qPCR results of functional genes related to nitrogen cycle showed that signi cant higher amoa-AOAs abundance in sibling than stranger soils (P = 0.08; Fig. 4a). However, there showed no signi cant differences on abundances of genes amoA-AOBs, nirS, nirK and nosZ between sibling and stranger soils (Fig. 4). In terms of the microbial community, sibling soils of G. max had higher archaea abundance (P = 0.004) than strangers compare to the bacteria and fungi abundance ( Fig. 5; Table 1).

Discussion
Previous studies have suggested that kin recognition could occur generally in nature, but morphologically based studies have obtained varied conclusions (see to File et al., 2012). We would argue that changes in performance or tness in theory likely indicate the tness consequences of changes in adaptive traits, hence changes in potentially adaptive traits (such as root allocation and C exudation) could indicate kin recognition because of which without resource differences suggesting a response to a signal or cue. The current study determined plant kin recognition of G. max by not only including the tness, but also including morphological traits, physiological performances, root exudation, and N uptake as well as microbial biomass C and N, functional genes related to nitri cation and denitri cation. Integrating these ndings, we demonstrated that kin recognition is more nutrient related than previously thought.
Plant tness (Cheplick, 1992(Cheplick, , 2004File et al., 2012) and morphological responses involved in resource capture (Dudley & File, 2007;Murphy & Dudley, 2009;Bhatt et al., 2010;Biedrzycki et al., 2010;Biernaskie, 2011) in our current study suggested that plants in kin groups show bene t in plant growth e ciency (Lepik et al., 2012). In particular, siblings of Glycine max increased their SLA compared to strangers, which was known to re ect the e cient light capture of plants resulting from mutual shading (Ballare et al. 1994; Gri th and Sultan 2005), suggested a cooperative behaviour through increasing the e ciency of light acquisition (Lepik et al. 2012;Crepy and Casal 2015;Li and Xu, 2017). This was also supported by higher chlorophyll content in siblings compare to strangers, which contributes to potential photosynthetic e ciency (Melis, 2009). These changes can also explain better increased size and growth, e. g, height and shoot biomass of siblings living together (Till-Bottraud and de Villemereuil, 2016).
The lack of signi cant differences in root length and root biomass allocation suggests that there was no kin response in these belowground traits. However, lower root activity of siblings than strangers suggests lower underground competition for soil nutrients or water (Islam et al., 2007). Higher use e ciency calculated by total N content in siblings than strangers suggested that kin bene t may achieved through increasing the nutrient use e ciency (Lepik et al., 2012) of siblings compared to strangers. Thus, this may suggest that a mutual accommodation was happened among siblings while there was more competition among strangers belowground. This was further supported by higher NR activity in siblings than in strangers, implying higher potential NO 3 − -N transformed to NO 2 − -N and then amino acid for plant growth (Barber, 1984;Baker and Hall, 1988). Thus, the kin recognition responses in N uptake and use e ciency, complementary with the phenotypical performance, may support growing with kin as increasing resource capture e ciency and energy saving of siblings compare to strangers of G. max. To summarize, such distinct difference between above-and below-ground parts is ascribed to the ability of legumes such as G. max to use atmospheric N 2 through symbiotic rhizobium. As a result, N is not so limited for G. max but it requires more C to fuel biological N xation, thus allocating more C to shoots to enhance photosynthesis. This is supported by similar nitrogenase activity between siblings and strangers ( Fig. 4b) as well as higher chlorophyll content in siblings than in strangers (Fig. 2b).
It has been demonstrated in Arabidopsis is that plants in kin groups could recognize kin or stranger neighbor through root exudates (Semchenko et al., 2014;Biedrzycki et al. 2010). However, the root exudates are complex, and what substances play key roles on this process was still unknown. We did analyze the main substance of root secreted-C and N and found that root secreted-C rather than secreted-N was sensitive to kin identity of G. max, but this is not evidence that root secreted-C provides kin signals. But what is most important is that, plants in kin groups produce more root secreted-C when facing siblings then strangers, which could be a direct available carbon source for rhizosphere microorganisms to activate the growth of rhizosphere microbes and activity (Toberman et al., 2011) and elicit rhizosphere priming effect (Kuzyakov, 2010). This could not only accelerate soil organic matter decomposition (Kuzyakov, 2002;Hopkins et al., 2013;Cheng et al., 2014), but also increase the supply of soil-available N for plants. Thus, root secreted-C could induce rhizosphere microbes changes, and those change may also create feedbacks in the soil nutrient cycle, especially the N mode of occurrence in soil and N absorption of plants, and thus in uence the plant growth or competition. This hypothesized mechanism is supported by higher root exudation of C and N, higher microbial biomass C, higher AOA in siblings than in strangers in our study. Lower soil NO 3 − concentrations in siblings than in strangers are ascribed to higher root uptake, which is supported by higher NR and uptake rates.
Speci cally, the N transformation model in sibling plants of G. max tended to convert more organic N into nitrate N (the amoa-AOA was higher in sibling than stranger soils) for plant absorption and utilization, and consequently the nitrate N in rhizosphere soil was lower in sibling than stranger soils. Soil N conversion was mainly actuated by soil nitri cation process regulated by ammonia oxidizing bacteria and ammonia-oxidizing archaea, and the denitri cation process related with denitrifying bacteria, as well as microbial N xation process. The results exhibited that the abundances of soil archaea and amoA-AOAs genes contributing to soil nitri cation were signi cantly increased in sibling soils compare to stranger soils, while the abundances of genes of NirK, NirS and NosZ contributed to soil denitri cation, and gene of nifH contributing to soil N xation did not signi cantly differed between siblings and strangers of G. max. This could be easily explained because denitri cation is generally considered to be intense in an anaerobic environment, but our experimental soil did not create a ooded condition. The result also suggested that compere to soil bacteria and fungi, the signi cant difference in abundance of soil archaea between sibling and strangers suggested which plays a signi cant roles on soil nitri cation process resulted from plant kin interaction, and plant growing with siblings could have an advantage from soil nitri cation process and N retention compared to strangers of G. max. Certainly, we should not deny the function of soil bacteria and fungus related with N cycling in plant kin interactions. Therefore, subsequent work should analyze the functional ora and community structure of soil microbes including archaea, bacteria and fungi at the classi cation level to nd the important functional ora involved in nutrient conversion affected by plant kin interactions.

Conclusion
In summary, our results showed the bene ts of growing with kin in G. max were re ected by increased tness and resource use e ciency instead of competition among siblings compare to strangers. Moreover, we found evidence that the rhizosphere effects of soil microbial changes and soil N cycling and transformation triggered by the root secreted C, could be a potential underground feedback mechanism of plant kin recognition. We have found that soil archaea community of microorganisms was sensitive both on abundance and function in kin interactions, but we did not identify the functional strains of these communities at the level of a family or genus, which still need further study by high throughput sequencing to reveal more information.
Declarations ACKNOWLEDGEMENT