2.1. Study sites and sampling
The study site was located in Xiaosugan Lake Provincial Nature Reserve for Migratory Birds, Aksay County, Jiuquan City, Gansu Province, China. It is located in the Huahaizi Plateau Basin between Altyn Tagh Mountain, Danghe South Mountain and Saishiteng Mountain on the northern margin of Qaidam Basin on the Qinghai-Tibet Plateau (39.22°–39.35° N, 94.45° –94.59° E) (Fig.1), at altitudes of 2807–2808m. The study area is characterized by an inland alpine and semi-arid climate, with an annual average temperature of 2.0 °C, an average annual precipitation of 77.6 mm, an average annual evaporation of 1964.8 mm, and an average frost-free period of 90 days, prevailing gale weather, high frequency of sandstorms, drying degree of 30 (Li et al., 2019).
The Sugan Lake water system is an independent water system at the central and northwest end of the Qaidam inner water system. It originates from the Da and Xiao Haerteng Rivers in the Nanshan of the Danghe River, and flows underground through the mountain passes, flows into the Da and Xiao Sugan Lakes in the Haizi Basin, and then flows into the Da and Xiao Sugan Lakes. Finally, it is consumed by lake evaporation and evapotranspiration in the spring overflow zone. According to the classification standard of Chinese soil classification system in 1992, the main soil types in the study area are meadow soil, meadow swamp soil and saline-alkali soil. The dominant plant species are Saussurea salsa, Leymus secalinus, Phragmites australis, Glaux maritima, Triglochin palustre, Triglochin maritimum, Suaeda glauca, Blysmus sinocompressus, and Salicornia europaea. (Li et al., 2019)
2.2 Experimental method and design
From 15 to 27 August 2020, based on the field investigation, an area with relatively representative habitat gradient change was selected along the northwest shore of Xiaosugan Lake in the direction of the vertical water area. A rectangular transect (800 m in length × 200 m in width) was arranged as the study sample. The plant community in this transect was dominated by Saussurea salsa, accompanied by plants such as Leymus secalinus and Salicornia europaea.
First, the groundwater depth was measured (the groundwater was dug with a shovel at intervals of 50 m from near the lakeshore in each transect, and measured after the water level stabilized). Then, based on the distance from the lakeshore and the variation of groundwater depth, the transect was divided into three plots (Plot I, Plot II and Plot III, Fig.1). They were as follows, 1) Plot I, The near water area, 0-200 m from the water body, groundwater depth 0.30-0.50 m. There was no accumulation of salt and alkali in the soil surface, and the soil moisture was saturated almost all the year round. The main plants in this region are Triglochin maritimum, Glaux maritima, Saussurea salsa and so on. 2) Plot II, the middle water area, 300-500 m away from the water body, groundwater depth 0.50-1.00 m. This area was seasonally flooded and soil water is seasonally saturated. The accumulation of salt and alkali in the soil surface was stronger than Plot I. The vegetation was patchlike, with Saussurea salsa, Glaux maritima and Leymus secalinus growing together. 3) Plot III, the far water area, 600-800 m from the water body, groundwater depth 1.00-1.50 m. The accumulation of salt and alkali in the soil surface was strongest. Vegetation was distributed in aggregated patches with low height and high density. The main plants were Saussurea salsa and Leymus secalinus.
Third, based on this gradient of the Saussurea salsa population in the sample, six 1 m × 1 m quadrats were established in each plot, for an overall total of 18 quadrats. Fourth, the community traits (height and coverage) were determined. Fifth, 6 well-growing Saussurea salsa plants were selected from each plot. After measuring the height of the plants, the aboveground parts were cut off and put into numbered envelopes. The root of each Saussurea salsa was completely dug after measuring the root depth and branch angle by using the trench method and the whole root excavation method. The root was placed with a mesh screen (aperture = 0.25mm), and the adjacent water source was searched to clean the soil and debris attached to the root. The soil was put into the numbered self-sealing bag respectively, and the root was brought back to the laboratory together with the aboveground part to measure the biomass and other traits. Finally, the aboveground parts of all plants in each plot were cut and transported to the laboratory. The aboveground biomass of each plot was measured after drying in an oven at 80 °C for 48 h.
2.2 1 Measurement of root morphological characteristics and organs biomass of Saussurea salsa
In the laboratory, Saussurea salsa roots were put on the root disk of a root scanner with a small amount of water to make it spread out. After scanning, the root was stored in a graphic file to the computer, and the average diameter, root length and root forks (RF) of the root were measured with WIN-Rhizo (Pro 5.0) software (Regent In-struments, Canada). RF refers to the number of forks per unit root length. Within the same unit length, the higher the number of forks, the stronger the ability of the root to absorb nutrients or water. The formula is, RF (No.·cm-1) =The number of RF (No.)/ root length (cm) (Liu et al., 2010). Then, the root and leaf parts of Saussurea salsa were put into the numbered envelope bags and placed in the oven. They were measured after drying in an oven at 80 °C for 48 h. Root-shoot ratio = underground biomass/aboveground biomass, and specific root length (cm/g) = root length (cm)/root biomass (g).
2.2.2 Measurement of fractal characteristics of Saussurea Salsa Roots
The roots of Saussurea salsa were analyzed with WIN-Rhizo (Pro 5.0) software (Regent Instrumentation, Canada), obtaining squares with side length of r and the number of squares cut by roots in root distribution maps of each plot. As the square length r decreases gradually, the Nr cut by the root system increases gradually. After obtaining the corresponding Nr values at different levels of r, the graph was drawn with lgr and lgNr as abscissa and ordinate respectively. The equation of the regression line is,
The negative number of the slope of the regression line is the fractal dimension (FD), and lgK is the fractal abundance (FA) (Ketipearachchi and Tatsumi, 2000). When r=1, the FA has a maximum value, lgNr =lgK, FD=0, that is, there is no branching, which does not actually exist.
2.2.3 Measurement of soil moisture content, soil bulk density, soil salinity and pH
In the wetland community survey, soil samples were randomly collected of the 1 m × 1 m × 0.5 m (length × width × depth) soil profile using a ring knife (200 cm3). Five soil layers each of 10 cm depth were collected, with three replicates per layer. The replicate soil samples were mixed while fresh, then oven-dried at 105 °C for 12 h. The soil moisture content (SMC) and soil bulk density (SBD) were calculated for each soil layer.
The salt content of the soil samples was measured using the EC method. A sample (10 g) of air-dried soil sieved with 2 mm mesh was weighed at room temperature, then 50 mL CO2-free distilled water was added (water: soil mass 5:1, v/w). The leachate was collected and placed in an oscillator for 5 min. Using a vacuum filtration system, the suspended soil slurry was slowly poured into the funnel until filtration was completed. The filtrate was poured into conical flasks for later use (Li et al., 2019). A portable conductivity meter (DDS-11C, Shanghai Lei Magnetic Instrument Factory, Shanghai, China) and portable soil pH meter (ST3100, Ohaus Instruments Co., Ltd, Shanghai, China) were used to measure the EC and pH of the leachate, respectively, with three replicates per sample, and the average was calculated.
2.3 Statistical analysis
Microsoft Excel 2019 was used to organize all the original data of the experiment, and the root shoot ratio, stem mass ratio, leaf mass ratio and root mass ratio of various Saussurea Salsa were calculated respectively. Secondly, the correlation between geometric characteristics of roots and fractal structural parameters of Saussurea Salsa under habitat gradient were analyzed in the R 4.0.2 software. Then, the experimental data of FD and FA of Saussurea Salsa were converted by log 10 to make them conform to normal distribution. At last, the relationship between FD and FA of roots of Saussurea Salsa was analyzed by linear regression analysis. Plot with origin 2016 and SigmaPlot 10.0 software. The mean and standard error (SE) of five replicates were obtained for each measurement.
Table 1 Common parameters and their abbreviations
Parameter
|
Abbreviations
|
Units
|
Soil moisture content
|
SMC
|
%
|
Soil electrical conductivity
|
EC
|
ms·cm–1
|
Soil bulk density
|
SBD
|
g/cm3
|
Aboveground biomass
|
AB
|
g·m–2
|
Average height
|
AH
|
cm
|
Density
|
D
|
strain·m-2
|
Coverage
|
C
|
%
|
Root depth
|
Rdep
|
cm
|
Stem biomass
|
SM
|
g
|
Leaf biomass
|
LM
|
g
|
Root biomass
|
RM
|
g
|
Root surface area
|
RSA
|
cm2
|
Total Root length
|
TRL
|
cm
|
Root average diameter
|
RAD
|
mm
|
Specific root length
|
SRL
|
cm/g
|
Root forks
|
RF
|
No./cm
|
Root branch angle
|
RBA
|
°
|
Fractal dimension
|
FD
|
No dimension
|
Fractal abundance
|
FA
|
No dimension
|
Root-shoot ratio
|
RSR
|
No dimension
|