The influences of four types of soil on the growth, physiological and biochemical characteristics of Lycoris aurea (L’ Her.) Herb

Based on the characteristics of Lycoris aurea (L. aurea) natural distribution and local soil types, we selected four representative types of soil, including humus soil, sandy soil, garden soil and yellow-brown soil, for conducting the cultivation experiments to investigate key soil factors influencing its growth and development and to select the soil types suitable for cultivating it. We found that there existed significant differences in the contents of mineral elements and the activities of soil enzymes (urease, phosphatase, sucrase and catalase) etc. Among which, the contents of organic matters, alkali-hydrolysable nitrogen, Ca and Mg as well as the activities of soil enzymes in humus soil were the highest ones. In yellow-brown soil, except for Fe, the values of all the other items were the lowest ones. Net photosynthetic rate (Pn), biomass and lycorine content in humus soil were all the highest ones, which were increased by 31.02, 69.39 and 55.79%, respectively, as compared to those of yellow-brown soil. Stepwise multiple regression analysis and path analysis indicated that alkali-hydrolysable nitrogen, and Ca etc. were key soil factors influencing Pn, biomass and lycorine content of L. aurea. Thus, humus soil can be used as medium suitable for artificial cultivation of L. aurea.

biosynthesis and accumulation of alkaloids in Lupinus angustifolius. Ca 2+ plays important roles in sequestration and signaling in regulating the activities of chloroplasts 9,10 . Plants require K + for important intracellular physiological functions, including photosynthesis and nutrient transport 11 . Soil enzymes are one type of the most important biological components of the soil ecosystem. They play an important role in organic matter decomposition and nutrient cycling 12 . For instance, the hydroxylases (e.g. urease and sucrase etc.) can hydrolyze the macromolecules, such as proteins and polysaccharides, to form the simpler and smaller molecules that are easily absorbed by plants and to accelerate the nitrogen cycle and carbon cycle within the soil ecosystem. The activities of soil enzymes are closely related to soil physicochemical properties, soil types, and fertilizer application, cultivation and other agricultural measures 13,14 . Alkaloids are an important class of plant secondary metabolites and the result of the interactions between plants and their environments (both biotic and abiotic) during the long-term evolution process 15,16 . Different types of soil possess different textures and physiochemical properties while the demands of different types of medicinal herb plants for suitable soil conditions are quite different. Thus, the types of soil for cultivation of medicinal herb plants should be selected according to the particular physiological requirements of the particular plants 6,17 . Currently, most of the studies on L. aurea have mainly focused on the such aspects as biological evolution 18,19 , chemical compositions [20][21][22] , physiology and biochemistry [23][24][25] , pharmacology and pharmacodynamics 2,26,27 . In term of cultivation, Zeng et al. 28 reported that L. aurea preferred the environments of shading, humidity, pleasantly cool, ventilation and penetrating light and had no strict requirement for soil type, but it grew better in sandy loam and calcific soil etc. that were fertile, porous, and rich in organic matter. However, the studies on the effects of soil conditions on the growth and development of L. aurea and the accumulation of the medicinal components have been barely available. Thus, it is necessary to select suitable soil conditions for artificial cultivation of L. aurea. In this study, based on the characteristics of its natural distribution patterns, we selected four representative types of soil with different textures and physiochemical properties for conducting the controlled experiments on the cultivation of L. aurea, aiming to study the correlations of the key soil factors with its growth, development and accumulation of medicinal component for providing the experimental basis for artificial cultivation of L. aurea.

Results
Comparison in physicochemical properties among different types of cultivation soil. As shown in Table 1, there were significant differences in pH value, the contents of soil moisture, organic matter, alkali-hydrolysable nitrogen, rapidly available phosphorus, rapidly available kalium, Ca and Mg etc. in different types of soil. Among which, the contents of organic matter, alkali-hydrolysable nitrogen, Ca and Mg were the richest ones in humus soil, which were 22.64, 7.99, 11.66 and 5.88 times those of yellow-brown soil, the poorest ones. The Fe content in yellow-brown soil was the highest one. The contents of soil moisture, rapidly available phosphorus, rapidly available kalium, Zn, Mn and Cu in garden soil were higher but its Mo content was extremely low. The Mo content in sandy soil was the highest one while the contents of the remaining compositions were between the other types of soil. The humus soil, sandy soil and garden soil were all alkalescent while yellow-brown soil was acidic.
Comparison and analysis on major agronomic trials of L. aurea among different types of cultivation soil. Different types of cultivation soil had different impacts on the major agronomic trials of L. aurea ( Table 2). The biomass performances, the bulb diameter, floral axis height, leaf length and leaf width were all the highest ones in humus soil, which were increased by 69. 39, 14.50, 11.74, 15.72 and 8.06%, respectively, as compared to those in yellow-brown soil in which their performances were poorest. The differences in these parameters between two types of soil were statistically significant (P < 0.05). Their performances in sandy soil and garden soil were between those of the humus soil and yellow-brown soil. Comparison and analysis on photosynthetic parameters of L. aurea among different types of cultivation soil. Different types of cultivation soil had different effects on photosynthetic parameters of L. aurea (Table 3). Among which, the net photosynthetic rate (P n ), chlorophyll content, transpiration rate (T r ), intercellular CO 2 concentration (C i ) and stomatal conductance (G s ) were all the highest ones in humus soil. Except for the lowest T r value in sandy soil, the performances of all the remaining parameters in yellow-brown soil were the poorest ones. Compared to those in yellow-brown soil, P n and chlorophyll content were significantly increased by 31.02 and 25.32%, respectively (P < 0.01).
Comparison and analysis on the activities of soil enzymes among different types of cultivation soil. As shown in Table 4, there existed differences, to certain extend, in the activities of soil enzymes of L. aurea among four types of cultivation soil. Among which, the activities of soil enzymes in humus soil were the highest ones whereas those in yellow-brown soil were the lowest ones. Compared to those in yellow-brown soil, the activities of urease, sucrase, phosphatase and catalase were significantly increased by 9.63, 1.64, 4.03 and 1.95 times, respectively (P < 0.01). Among the soil enzymes tested, the activities of both urease and sucrase in four types of soil were also higher whereas the activity of catalase was the lowest one.
Comparison and analysis on the lycorine content of L. aurea among different types of cultivation soil. The chromatogram of the bulb sample of L. aurea in cultivation soil were shown in Fig. 1b. The lycorine content of L. aurea in four types of cultivation soil was in the order from high to low as follows: humus soil (1.48 mg g −1 DW) > sandy soil (1.35 mg g −1 DW) > garden soil (1.27 mg g −1 DW) > yellow-brown soil (0.95 mg g −1 DW). Among which, the lycorine content of L. aurea in humus soil was significantly increased by 55.79%, as compared to that in yellow-brown soil (P < 0.01).

Analysis on key soil factors significantly influencing the medicinal quality of L. aurea.
Stepwide multiple regression analysis on soil factors significantly influencing the medicinal quality of L. aurea. The measured values of the soil nutrients and mineral elements were taken as the soil factor group while the measured values of P n , biomass and lycorine content were taken as the L. aurea medicinal quality group. The soil factors, including  Table 2. Comparison of main agronomic characters of L. aurea in four types of soil. Data followed by different letters within the same column are significantly different (p < 0.05). Mean ± SM represents their standards of error, n = 15. DW represents dry weight.

Soil types
Humus pH (X 1 ), soil moisture (X 2 ), organic matter (X 3 ), alkali-hydrolysable nitrogen (X 4 ), rapidly available phosphorus (X 5 ), rapidly available kalium (X 6 ), Ca (X 7 ), Mn (X 8 ), Fe (X 9 ), Na (X 10 ), Zn (X 11 ), Cu (X 12 ), Mn (X 13 ), and Mo (X 14 ), were taken as the independent variables, i.e. soil factor group while the leaf P n (Y 1 ), biomass (Y 2 ) and lycorine content (Y 3 ) of L. aurea were taken as the dependent variables, i.e. medicinal quality group. The soil factors were selected with stepwide multiple regression method. The stepwide multiple regression equations between the L. aurea medicinal quality and significant soil factors were formulated ( Table 5). As shown in Table 5, the significant soil factors influencing leaf P n were alkali-hydrolysable nitrogen (X 4 ) and Mo(X 14 ) while the significant soil factor influencing the biomass was alkali-hydrolysable nitrogen (X 4 ); the significant soil factors influencing the lycorine content were alkali-hydrolysable nitrogen (X 4 ), Ca (X 7 ) and soil moisture (X 2 ). Thus, the significant soil factors influencing the medicinal quality of L. aurea are alkali-hydrolysable nitrogen, Ca, Mo and soil moisture.
Path analysis on the key soil factors significantly influencing P n , biomass and lycorine content of L. aurea. In order to further confirm the key soil factors significantly influencing P n , biomass and lycorine content of L. aurea, SPSS statistics analysis was conducted on the significant soil factors and the results were presented in Table 6. As shown in Table 6, the effects of alkali-hydrolysable nitrogen on P n , biomass and lycorine content of L. aurea were the greatest ones with determination coefficients of 0.927, 0.976 and 0.833, respectively, indicating that among these factors, alkali-hydrolysable nitrogen is the most significant one; The determination coefficient for the effect of Ca on lycorine content of L. aurea was 0.666, which was ranked the second place among the factors tested, indicating that Ca has important effect on the accumulation of lycorine in L. aurea. Mo had little direct effect on leaf P n of L. aurea (R 2 = 0.015); Soil moisture had negative effect on the lycorine content as its determination coefficient was negative, indicating that soil moisture is a limiting factor. Thus, the key soil factors significantly influencing P n , biomass and lycorine content of L. aurea were alkaline-hydrolysable nitrogen and Ca etc.

Discussion
L. aurea has been used as a traditional Chinese medicinal herb plant to treat several important diseases. However, the increasing demand on L. aurea is contradictory to the limited supply source of wild L. aurea due to the deterioration of its inhabitant environment and over-artificial digging. One of the effective ways to resolve this contradiction is to artificially cultivate it in large-scale. Selection of appropriate types of soil for artificial cultivation of L. aurea is an essential step toward this solution. In the present study, we selected four representative types of soil, i.e. humus soil, sandy soil, garden soil and yellow-brown soil, to investigate their effects on the growth, development and accumulation of alkaloids of L. aurea. We determined that humus soil could be the suitable soil type for artificial cultivation of L. aurea, as supported by several lines of evidence as follows: (a) Humus soil contained the most abundant organic matter, alkali-hydrolysable nitrogen, Ca and Mg; (b) Humus soil displayed the best performances in several important agronomic trials, including biomass, the bulb diameter, floral axis height, leaf    6 reported that the types and texture of soil were closely related to the growth and development of medicinal plants and that loam soil was the relatively ideal type of soil for the cultivation of root/stem-types of medicinal plants. The results obtained from this study have indicated that the humus soil displays the best comprehensive performances in both agronomic trials and physiological and biochemical characteristics including P n , biomass and lycorine content of L. aurea grown among four different types of soil tested, followed by those of sandy soil, garden soil, and yellow-brown soil in the order from high to low. The differences in comprehensive performances are partially due to the significant differences in texture, pH value and organic matter etc. among the types of soil tested. In this study, measurement of the general physicochemical properties of cultivation soils revealed that humus soil was rich in organic matter content with its looser texture and better permeability. Humus soil, sandy soil and garden soil were all alkalescent while yellow-brown soil was acidic. The correlation analysis revealed that the soil pH value displayed a positive correlation with the lycorine content of L. aurea (Supplementary Information), indicating that the alkalescent soil is favorable for the accumulation of lycorine of L. aurea. This result was consistent with that obtained by Chao et al. 30 , who reported that the alkaline soil in North China was favorable for the accumulation of alkaloids while the acidic soil and yellow-brown soil in South China were unfavorable for the accumulation of alkaloids. Furthermore, path analysis indicated that value for the direct effect of soil moisture on the lycorine content was negative, implying that high soil moisture content may be unfavorable for the accumulation of alkaloids such as lycorine in L. aurea. This result was consistent with those obtained by El-Shazly et al. 16 and Bustamante et al. 31 , who reported that drought environment could enhance the biosynthesis of plant alkaloids. The rich organic matter that was constantly decomposed in humus soil can provide the stable supply of nitrogen nutrients etc. for the growth of plants. The physicochemical properties of humus soil, i.e. rich organic matter, alkalescent pH value, the looser texture and thus, lower soil moisture, are favorable for the growth, biosynthesis and accumulation of alkaloids such as lycorine of L. aurea.
Deficiency or shortage of any of the nutritional elements (e.g. N, P, K, Ca, and Mg etc.) will certainly affect the normal growth and development as well as the internal and external qualities of plants 7 . Nitrogen is the most important element among all the nutritional elements required by plants 32 . For instance, the biosynthetic processes of alkaloids require nitrogen involvement. The increased, adequate or surplus nitrogen source was found to be favorable for the biosynthesis of alkaloids in Larkspur 33 . The present study indicated that the contents of alkaline-hydrolysable nitrogen and Ca in humus soil were higher than those in the poorest yellow-brown soil. Its biomass and the lycorine content were increased substantially. Stepwide multiple regression analysis and path analysis indicated that alkaline-hydrolysable nitrogen was the most important soil factor, and Ca is the secondary factor, implying that the higher contents of alkaline-hydrolysable nitrogen and Ca in humus soil are favorable not only for the growth and development of L. aurea but also for the accumulation of alkaloids including lycorine. These results also further confirm that nitrogen nutrient and Ca are the important environment factors stimulating plant growth and the biosynthesis of alkaloids. NO 3 -N and NH 4 -N are two major forms of nitrogen nutrients that are absorbed and utilized by plants. The effectiveness of these two forms of nitrogen element on the growth and development of plants are dependent on the types of plants, the concentrations of NO 3 -N and NH 4 -N and their ratio. The absorption, transport and assimilation during the metabolism processes and the effects on the growth, development and physiological processes are significantly different 34 . During the cultivation of crops, nitrogen nutrient and water supply are two very important controlling factors. Thus, how to maximize the effects of nitrogen nutrients, water and Ca 2+ on stimulation of the growth, development, and the accumulation of alkaloids such as lycorine of L. aurea and the underlying regulatory mechanisms remain to be further investigated.
The microorganisms inhabiting in soil (e.g. bacteria and fungi etc.) play extremely important roles in the formation of soil fertility and the inter-conversion of plant nutrients and also affect the permeability of root cells and root metabolism. They can modify the root secretion and change the rhizosphere nutrients 35,36 . In the present study, we found that among four types of soil tested, there existed significant differences in the activities of soil enzymes, including urease and sucrase. Among four types of soil, the humus soil was rich in nutrients, such as organic matter, The activities of its soil enzymes were also higher. But in the poorest yellow-brown soil, the  Table 6. Path analysis on key soil factors significantiy influencing P n , biomass and lycorine content of L. aurea. R 2 represents coefficient of determination.
mean activities of soil enzymes were lower. The higher activities of these soil enzymes in the humus soil may be mainly attributed to the higher abundance and activities of soil microorganisms, likely due to the favorable soil conditions for their growth. The high activities of these soil enzymes and active soil microorganisms can continuously drive the degradation and mineralization of soil organic matter and provide the stable supply of nitrogen nutrients etc. for the growth of plants and thus, they are favorable for the growth, development and formation and accumulation of alkaloids, including lycorine of medicinal plants.

Conclusion
In this study, we found that among four representative types of soil tested, the humus soil displayed the best comprehensive performances in terms of the agronomic, physiological and biochemical characteristics including P n , biomass and lycorine content of L. aurea, followed by those of sandy soil, garden soil, and yellow-brown soil in the order from high to low. This humus soil contained higher levels of organic matter, the activities of soil enzymes and mineral elements such as alkali-hydrolysable nitrogen, rapidly available phosphorus, and Ca etc. Its texture was looser and its permeability was quite good. Thus, this type of humus soil was suitable for artificial cultivation of L. aurea. Stepwise multiple regression analysis and path analysis indicated that the key soil factors significantly influencing P n , biomass and lycorine content of L. aurea were alkaline-hydrolysable nitrogen and Ca etc. Soil moisture was a limiting factor, implying that high soil moisture content may be unfavorable for the accumulation of lycorine of L. aurea. Our findings provide not only the guidance for conducting artificial cultivation of L.aurea, but also the methods for accumulation of alkaloids including lycorine of medicinal plants. Measurement of other factors of the soil samples. The contents of alkali-hydrolysable nitrogen, rapidly available phosphorus, rapidly available kalium were determined with diffusion method, NaHCO 3 extraction-Mo-Sb colorimetric method, and NH 4 OAc extraction-flame spectrophotometry, respectively. The organic matters were determined with potassium dichromate oxidation heating method. The soil water content was measured with oven-drying methd 38 . The pH value was determined by using of PHS-3C precision acidity meter.

Materials and Methods
Measurement of major agronomic trails. Because L. aurea has the following characteristics: its flowers and leaves do not appear at the same time and it has summer dormancy. Its flowers blossom out in August, its leaf development starts in September and its vigorous growth stage is in December. Thus, the measurements of its agronomic trails in different types of cultivation soil were conducted in two stages. Its floral axis height was measured in August 2014 while its morphological parameters, including leaf length, leaf width and the bulb size were measured in December of the same year. The entire plants, including roots, leaves and bulb, were collected and dried by baking in oven to the constant weight and its biomass was weighted. The bulb samples were ground into powder with grinder (60 meshes) and stored under dry condition for subsequent analysis. Five healthy and strong plants were randomly selected from each sampling site and used as the measured subjects. Three repeat experiments were set.
Measurement of photosynthetic characteristics. Photosynthetic parameters, such as leaf P n , T r , G s , and C i, and other physiological factors were measured by using of Li-6400 portable photosynthesis measurement system with a red-blue light source (Li-cor, USA) under saturating light of 1 000 μ mol m −2 s −1 . The net photosynthetic rates were measured at least 30 min after the attainment of the temperature 39 . Given that an "afternoon relaxation of photosynthesis" phenomenon exists in L. aurea, measurement of photosynthesis was conducted in morning time.
The same positions of the leaves of five randomly selected L. aurea were used to measure P n during 9:00-11:00 am in December 2014 when the plant was in nutrition period 24 . Immediately after that, the corresponding leaves were collected and extracted in 95% ethanol and measured by spectrophotometric method (DU-800 spectrophotometer, Beckman Coulter, Inc.) at the wavelengths 665 and 649 nm. Contents of chlorophyll a and b were calculated by using the method of Lichtenthaler 40 .
Assays of activities of soil enzymes. In the middle 10 days of each month from January to December 2013, the soil samples at 2 cm below the surface soil nearby the root system of L.aurea were collected from various sampling sites. The enzymatic activities in these soil samples were assayed with the method reported by Guan et al. 41 as follows: Urease activity was assayed with indophenol blue colorimetric method and expressed as the amount (in mg) of NH 3 -N/g of dried soil produced within 24 h incubation at 37 °C. Sucrase activity was assayed with 3,5-dinitrosalicylic acid colorimetry and expressed as amount (in mg) of glucose produced within 24 h incubation at 37 °C. Phosphotase activity was assayed with alkaline phosphatase colorimetric method and expressed as the amount (in mg) of phenol produced/g of dried soil within 24 h incubation at 37 °C. Catalase activity was assayed with potassium permanganate titration method and expressed as the volume (in mL) of consumption of 0.1 N KMnO 4 /g of dried soil within 20 min incubation at 37 °C. The mean annual activity of each of these enzymes was calculated. Sample preparation and measurement of lycorine content. The sieved samples of the bulbs were dried at 65 °C to constant weight. The sample was extracted with Soxhlet method 24 . The sample solution was separated as described above and lycorine content was calculated using the peak area according to the linear regression equation.

Measurement of lycorine content. Conditions used in assays with
Data analysis. Data analysis was performed with Statistical Product and Service Solutions(SPSS). The correlation analysis was conducted with Pearson correlation coefficient method. The key soil factors influencing P n , biomass and lycorine content of L. aurea were determined with multiple regression analysis 42 and path analysis 43,44 . The multiple regression analysis was performed with a stepwide method to sequentially include variables in the model, using the following pre-established criteria: inclusion of a variable when its level of significance was < 0.05 (p in < 0.05), exclusion of a variable when its level of significance was > 0.10. This method selects significant variables one by one, and every time a new variable is included, the rest of those previously selected are examined to check if any of them may be removed from the model. The significance of coefficients was evaluated by a t-test. A p-value < 0.05 was considered significant.