Experimental location
In the grasslands of Inner Mongolia, R. soongarica communities are found in scattered patches. An ancient lake basin containing an R. soongarica community, situated east of Erlianhot City, Inner Mongolia, China (43°23′16″–43°42′28″ N, 112°01′1″–112′01′78″ E; 910 m above sea level), was selected as the experimental site. The diameter of the ancient lake basin is approximately 15 km and the depth is 30 m. The soil conductivity (soil salinity) increases with decreasing elevation, which produces clear boundaries between the plant communities at the site (Fig. 1). Outside the ancient lake basin is the zonal plant community, which is dominated by Stipa glareosa. From the basin margin to the center, the plant communities are R. soongarica + S. glareosa, R. soongarica + Allium polyrhizum, pure R. soongarica, R. soongarica + Kalidium foliatum, and pure K. foliatum.
The mean monthly air temperature ranges from − 18.1°C in January to 23.3°C in July, with a mean annual temperature of 3.4°C. The mean annual precipitation (1971–2000) is 142.2 mm, 67% of which falls from June to August. The surface soil is composed of 66% sand, 20% silt, and 14% clay, with a mean pH of 7.1 outside the ancient lake basin. These natural plant communities are not fenced because they are used for grazing by domesticated sheep and camels. Grazing is the main land use method, and the intensity of grazing is relatively light.
We have the permission to collect the plant samples by the local government. The voucher specimen with the collect number(Wang 20080705) was identified by professor Liqing Wang and deposited in HIMC.
Experimental Design
Soil conductivity
There are four types of R. soongarica communities in the experimental area: R. soongarica + S. glareosa, R. soongarica + A. polyrhizum, pure R. soongarica, and R. soongarica + K. foliatum. Reaumuria soongarica coexists with the two herbs only—in the R. soongarica + S. glareosa and R. soongarica + A. polyrhizum communities. In order to understand the difference in soil conductivity (soil salinity) with different types of plant individuals in this coexistence community of herbs and shrubs, soil samples were collected from each of the three species in the R. soongarica + S. glareosa and R. soongarica + A. polyrhizum communities at depths of 0–5 cm, 5–15 cm, 15–25 cm, 25–35 cm, 35–45 cm and 45–55 cm. Each sampling was performed in triplicate. The total of soil samples number was108. Soil conductivity was measured using the conductance method (Multi 340i; WTW Xylem Analytics, Weilheim, Germany). The pure R. soongarica and R. soongarica + K. foliatum communities were not selected because they contained no herbs.
Salt secretion rate
To understand the differences in the salt secretion rate of R. soongarica with different soil conductivity in the different communities, 12 samples of the R. soongarica individuals were chosen randomly from the R. soongarica + S. glareosa, R. soongarica + A. polyrhizum, and pure R. soongarica communities. The selected individuals were divided into four groups (four × three). One branch per individual was randomly marked. The accumulated salt on the leaf surface of the selected branches was washed repeatedly with distilled water at the beginning of the experiment to ensure that the surfaces of the leaves were salt-free. The salt secretion rate was measured for three individuals on the first, third, fifth, and seventh days. The marked branches were soaked in 150 mL of distilled water to measure the conductance of the liquid using a conductivity meter (Multi 340i; WTW Xylem Analytics) and were removed from the stem to measure the weight of the fresh leaves using an electronic balance (YP1002N; INESA, Shanghai, China) at the end of the experiment. The leached water was collected in containers to prevent it from falling to the soil surface. The measurement of the conductance of the liquid was undertaken in July and the entire experiment was repeated in August 2009. Although the experiment was performed twice, owing to the destruction of some plants by animals, we only obtained two and three complete datasets in July and August, respectively.
Population distribution patterns
The amount of salt secretion was found to correlate positively with the crown size of R. soongarica. The salt secreted by the R. soongarica leaves was likely to increase the soil conductivity, and thus affect the growth of the herbs in the community. This could be demonstrated by the canopy area of R. soongarica and the distance between R. soongarica plants and the neighboring herbs. To investigate the effect of the salt secretion of the shrub on the herbs, the canopy diameter and distance to the nearest S. glareosa and A. polyrhizum plants were measured for 30 samples randomly selected R. soongarica individuals in the R. soongarica + A. polyrhizum community using a measuring tape. The R. soongarica + S. glareosa community was not selected because A. polyrhizum was rarely found in this community. The crown is assumed to be a perfect circle, and its area is calculated from the diameter.
Laboratory cultivation experiment
To verify the change in the competitiveness of different plants under different salt conditions, we conducted cultivation experiments in the laboratory. Reaumuria soongarica, S. glareosa, and A. polyrhizum seeds were collected in August 2008. In March 2009, the seeds were soaked in distilled water for 30 min and treated with 0.1% gibberellin to break the seed dormancy. The seeds were then placed between double filter papers and germinated in a dark environment in a Petri dish. The germinated seeds were transferred to a paper cup (12 cm in height, 7 cm top diameter, and 5 cm bottom diameter) that was filled with 300 g of river sand, which had been rinsed repeatedly with distilled water to remove any salt and nutrients from the soil, thus simulating quartz sand.
Based on the abundance of the three species observed in their natural communities, the ratio of herbs to shrubs was 6:1. Five combinations of three species were prepared: one combination consisted of six seedlings of S. glareosa arranged in a ring 1 cm from the cup wall, the second combination was similarly prepared but with six A. polyrhizum seedlings, and the third combination consisted of only one R. soongarica seedling placed in the center of the cup. In the mixed species combinations, six S. glareosa or A. polyrhizum seedlings were placed around one R. soongarica seedling in a cup.
Six NaCl salinity levels were prepared: 0, 200, 600, 1000, 1500, and 2000 µS/cm (soil conductivity). This salinity gradient was based on the soil conductivity measured in different natural communities. Each treatment was prepared in triplicate.
Based on local precipitation and soil nutrient levels, 5 mL of Hoagland nutrient solution was applied every 2 weeks and 10 mL of distilled water was added each time and four times (once per week) to each cup. The experiment lasted for 12 weeks. During the fifth week of the experiment, different amounts of NaCl (0, 0.18, 0.54, 0.9, 1.35, and 1.8 g) were added to each cup to obtain six salinity levels. The seedlings were cultivated in an artificial climate chamber (ZRX-1000 ESW; TESTMART, Hangzhou, China) at 25°C and 70% relative humidity with a 14 h/10 h light/dark regime.
At the end of the experiment, the entire plant was dried and weighed. The relative biomass of the three species in each treatment was calculated based on the biomass of the plant cultivated individually at 0 µS/cm soil conductivity as a control.
Statistical analysis
To test the difference in soil conductivity of the different plant individuals in this coexistence community of herbs and shrubs and the difference in the salt secretion rate of R. soongarica with different soil conductivity (in different communities), one-way analysis of variance (ANOVA) was used at every soil layer and each sampling time. Pearson’s correlation coefficient, which was used to characterize the population distribution patterns, was calculated for the distance between one of the two herbs (S. glareosa and A. polyrhizum) and R. soongarica and canopy diameter. ANOVA was used to test the effect of salt and competition on the three species. When significant salt or competition effects were observed, Duncan’s post-hoc tests were used to assess the differences among the species. Statistical tests were performed using SPSS version 15.0 for Windows (SPSS Inc., Chicago, IL, USA).
Material statement
This manuscript comply with institutional, national, or international guidelines and the Convention on the Trade in Endangered Species of Wild Fauna and Flora.