* Corresponding author: Xiaodong Wu
The Effect of Grazing and Reclamation on the Rodent Community Stability in Alax Desert  
YUAN Shuai^{①, ②, ③†}, JI Yu^{①, ②, ③†}, FU Heping^{①, ②, ③†}, WU Xiaodong^{①, ②, ③}*, YANG Suwen^{①, ②, ③,} YUE Xiuxian^④
①.College of Grassland, Resources and Environment, Inner Mongolia Agricultural University, Hohhot 010011, China
②. Key Laboratory of Grassland Resources, Ministry of Education, Hohhot 010011, China
③. Rodent research center, Inner Mongolia Agricultural University, Hohhot 010011, China
④. The Institute of Forestry Monitoring and Planning of Inner Mongolia Autonomous Region, Hohhot, 010020, China
†Three authors contributed equally to this manuscript
                                                                                                    
Study on stability is an important content of ecology. It is the forefront of ecological environment and the focus issues of demand. Reclaiming and grazing are the main disturbances of desert ecosystem in China. Rodent is an important part of desert ecosystem. It plays an important role in the development of desert ecology theory and understanding the animal adaptation to arid environment. To understand the stability mechanism of rodent community under grazing disturbances is of great significance for further evaluation of the stability of desert ecosystem under human disturbances. In this study, the rodent community under different grazing disturbances in Alashan Desert was investigated by using live trapping from 2006-2011. The composition pattern, community diversity and variability of different life history strategies were analyzed. The results showed that the reclamation significantly reduce the rodent community diversity. Grazing do not make significant differences in the rodent community diversity, but the dominant species under different grazing disturbances changed. Reclamation have a great negative effect on the K-Stress tolerate-Slow strategist (KSS). With the increase of grazing intensity, the capture rate of KSS and r-Ruderal-Fast strategist (rRF) showes a decreasing trend, and the decline tread of rRF strategist is more significant. The cumulative biomass strengths in different processes of grazing exclusion, rotational grazing and over grazing are dominated by individual species. The cumulative abundance strengths is under the control of multiple species. Whereas the situation in reclamation area is contrast to this. The variability of KSS strategists in reclamation area is larger than any other areas. The variability of KSS strategist in grazing exclusion areas is the smallest. The abundance variability of rRF strategists is the largest in over grazing areas. From these results it can be seen that reclamation reduces the the rodent community stability by reducing the rodent community diversity. Reclamation reduces the the rodent community resistance by reducing the the advantages of KSS strategists, and through increasing variability of KSS strategists to increase the variability of the rodent community resistance; grazing by reducing the advantages of rRF strategists to reduce the rodent community resilience, and by increasing the variability of rRF strategists to increase the variability of the rodent community resilience.
Key words: bionomic strategies, community stability, rodent
* Corresponding author: Xiaodong Wu

1 第一章标题

Study on stability is an important content of ecology. It is the forefront of ecological environment and the focus issues of demand (Zhang Jiyi and Zhao Halin 2003). Facing a dramatic decline in global biodiversity, the question of stability has been the long-term concern of conservationists (Grimm V et al. 1992). MacArthur assumes a species produces abnormal variety of abundance for some reasons. If the other species abundance in this community has significant change because of the species, then the community is unstable. On the contrary, the smaller the other species affected by this species, the higher the stability of the community (MacArthur R 1955).
There are some debates on the definition of stability. May (1973) holds that stability is when the population is disturbed, timely return to the equilibrium value. But Holling (1973) considered that the degree of the disturbances to which a system can endure is the stability. Pimm (1984) denoted a characteristic as a “stability characteristic” if it can be assigned to one of the following conditions: recovery, variability, persistence and resistance. In these cases, different concepts of stability represent different stability characteristic, measures of different stability are quantified by measuring different stability characteristics. It is pointless to separate out a feature called the stability, because the stability covers the entire range of different characteristics (Sutherland J P 1981). Therefore, comprehensive evaluation from many aspects is needed in the research of community stability.
Resistance and resilience were considered to be two important components of the stability (King A W and Pimm S L 1983). MacArthur held that, if the abundance of other species in this community has significant changed because of the abnormal variety of a species, so that the community is unstable. On the other hand, the higher the community stability. This stability may be caused by two ways: first, the pattern of the interaction between species of the community; second, the stability of a species is inherent (MacArthur R 1955). But later, many researchers focus on the first approach, namely the relationship between diversity and stability as well as the interaction between species (Hairston N et al. 1968; King A W and Pimm S L 1983; Pimm S L 1984; Tilman D and Downing J A 1994), and few people to explore the inherent stability of species.
In the 1990s, the Chinese government has begun to encourage a large agricultural population migrated to the Alashan desert. With the increase of agricultural population, reclamation and overgrazing has become important human disturbances in this desert ecosystem. Rodents are not only the important part of the environment, but also the important indicator of environmental change in desert ecosystem (Flowerdew J R et al. 2004; Jones A L and Longland W S 1999), plays an important role in the development of desert ecology theory and understanding animal to adapt to the arid environment (Ojeda R A and Tabeni S 2009). Understanding the mechanism of rodent community stability has important significance for further evaluation of the stability of desert ecosystem in human disturbance. Studies on Chinese desert rodents are rare, especially in research on China desert rodent community stability. This study takes Alashan desert rodent community as the research object, discusses the impact that the different grazing and reclamation of disturbance on the stability of rodent community. We assume that: (1) Reclamation and overgrazing by reducing the rodent community diversity to reduce the the rodent community stability. (2) Reclamation and grazing by regulating the composition of different life history strategies in rodent community to change the resilience, resistance or variability of the desert rodent community.

2 Materials and Methods

2.1 Study site.
This study was conducted in the southern part of Alashan Zuo Qi at the eastern edge of the Tengger Desert, Inner Mongolia, China, (E104°10'-105°30', N37°24'-38°25') from April, 2006 to October, 2011. Our study area has a continental climate with cold and dry winters and warm summers. Annual temperatures range from -36 to 42 °C with a mean of 8.3 °C. Annual precipitation ranges from 45 to 215 mm, but about 70 percent falls from June to September. Potential evaporation ranges from 3000 to 4700 mm, and the annual frost-free period is 156 days. Approximately 5–15% of the ground is covered with shrubs, forbs, and some gramineous plants. Shrubs mainly consist of Zygophyllum xanthoxylon, Nitraria tangutorum, Caragana brachypoda, Ceratoides latens, Oxytropis aciphylla, Artemisia sphaerocephala and Artemisia xerophytica with Reaumuria soongorica as the dominant species. The major grasses/forbs species are Cleistogenes squarosa, Peqanum nigellastrum, Cynanchum komarovii, Salsola pestifer, Suaeda glauca, Bassia dasyphylla, Corispermum mongolicum, Artemisia dubia and Plantago lessingii (Wu Xiaodong et al.2016).
2.2 METHODS
An experiment, adopted a randomized block design with 3 blocks and four treatments including sheep light grazing, heavy grazing, grazing exclusion and farmland, were established in 2006, to assess the effects agricultural reclamation and livestock grazing on small mammal communities. Each block is 240 ha, and each treatment unit is 60 ha. Treatments including light grazing, successive grazing and grazing exclusion were established with standard sheep fencing (110 cm high). In light grazing sites, sheep grazing intensity was controlled within the range of 0.83 to 1.00 sheep per ha. In heavy grazing sites, sheep grazing intensity was controlled within the range of 3.75 to 4.23 sheep per ha. This was close to the common grazing intensity in the study area, but significantly higher than the Inner Mongolia government standards, which ranged from 0.603 to 1.120 sheep per ha in the study areas. Farmland which plant species similar with grazing exclusion was reclaimed in 1994, and a kind of shrub (Haloxylon ammodendron), sunflowers and corns were planted in farmland. A 7 × 8 trapping grid (0.96 ha) at a 15 m inter-trap distance was established at the center of each treatment units. Twelve trapping grids were established in study areas. One wire-mesh live trap (42 cm × 17 cm × 13cm) was placed at each trap station.
2.2.1 Trapping of rodent
Rodent were live trapped for 4 consecutive days at 4-week intervals from 2006 to 2011. Trapping was not run during winter (from November to March). Traps were baited with fresh pignuts, and checked twice (morning and afternoon) each day. Each captured individual was sexed, marked with a 1.5 g aluminum leg ring (0.4 cm diameter) with a unique identification number (ID) attached to the left hind foot, and weighed to the nearest 1 g. The sex, capture station, body weight, and reproductive condition of each capture were recorded. Males were considered in reproductive condition if they had scrotal testes. Females were considered reproductive if they possessed enlarged nipples surrounded with white mammary tissue, and a bulging abdomen. In order to avoid accidental death, traps were closed on extremely warm or rainy days, and trapping time was extended after extremely warm or rainy days to ensure 4 days of trapping in each month (Wu Xiaodong et al. 2016).
2.2.2 Segmentation of life history strategist
Study on the life history strategy includes two important concepts: fitness and treat-off. Reproduction and viability are the main content of fitness. Treat-off refers to the benefit that animal gain from a process, at the expense of losing the benefit in other process (Sun Ruyong 1997). The main purpose of this research is to clarify the patterns of different life history strategists of rodent community in different disturbed habitats. According to the segmentation of the life history strategists, this study analyzes the species richness and community structure of different life history strategists in different disturbed habitats. This study from the aspect of fitness to segment life history strategy. Based on Pianka’s classification principle on the r and K strategist (Pianka E R 1970) and Harvey’ classification principle (Read A F and Harvey P H 1989) on the fast-slow life history strategies, the 6 biological indicator of rodent in study areas include the body length, fetus number, litter size, trimester of pregnancy, reproductive period and longevity were counted (Lu Haoyong et al. 1988; Huang Wenji and Xu Shiju 1995; Zhang Zhibin and Wang Zuwang 1998; Wu Xiaodong et al. 2009; Ma Yong et al. 1987)(Table 2).
Cluster analysis was performed using unweighted pair-group method with arithmetic means (due to incomplete data of individual species, pregnancy is not included, Table 2). Because the hibernation in the high latitudes can significantly improve the survival rate of species under the adverse environmental conditions (Bieber C et al. 2011; Lebl K et al. 2011), so it will add hibernation or not into classification index. By clustering analysis, Alashan Gound Squirrel (Spermophilus alaschanicus), Mongolian five-toed Jerboa (Allactaga sibirica), Northern Three-toed Jerboa (Dipus sagitta) and Mongolian thick-tailed three-toed jerboa (Stylodipus andrewsi)with the larger size, long service life, short reproductive period and the fecundity is relatively weak, they are clustered into one category. This study classified this category as KSS strategists; and the relatively small size, long service life, long reproductive period and high fertility of other rodent species are clustered into the other hand, as rRF strategists(Fig. 1).
 
 
Fig.1  Clustr analysis of rodents with different bionomic strategies
Note: Cluster analysis was performed using UPGMA; A1-A7 denote respectively: Alashan Gound Squirrel Spermophilus alaschanicus, Mongolian five-toed Jerboa Allactaga sibirica, Northern Three-toed Jerboa Dipus sagitta, Mongolian thick-tailed three-toed Jerboa Stylodipus andrewsi, Southern Mongolian Jird Meriones meridianus, Mongolian Jird Meriones unguiculatus, Chinese Hamster Cricetulus barabensis, Eversman’s Hamster Cricetulus eversmanni, Alashan Gound Squirrel Phodopus roborovskii
2.2.3 Statistical analysis:
(1) W-statistical calculation formula is:
 
Where S is the number of species in the community, b_j is the cumulative biomass ratio of the j species before the biomass from the highest to the lowest; a_j is the cumulative abundance ratio of the j species before the abundance from the highest to the lowest. When W is close to 1, the dominant biomass of the community is gradually dominated by a single species, but the abundance of each species tend to be the same; when W is approaching to -1, the situation is just the opposite.
The diversity of the concept of stability, from a certain extent that they may contain many aspects of the content. Single out a particular feature is not suitable to evaluate the stability (Pimm S L 1984; Sutherland J P 1981). In order to understand the effects of different disturbances on the stability of desert rodent community, this study analyzed rodent community under different disturbances from the the aspects of community diversity, species abundance and the variability,resistance and recovery of population.
(2)Diversity indexes: Statistics on the species number of rodents and capture data of each species, and calculate the diversity index as follows: Species richness, and Shannon-Wiener index, Pielou evenness index, Simpson’s diversity index;
(3)Coefficient of variation: The measure of community stability is expressed by coefficient of variation (C.V) (Bai Yongfei and Chen Zhuozhong 2000):

(4)Significance test：
In order to assess the differences between rodent communities under different disturbance, the diversity index and W statistic of all rodents under the same disturbance were carried out to test the differences in the mixed effects model. All the analysis were performed with SAS9.2, with a significant level of a=0.05, using Tukey test for mean comparison.all values are given mean±1 standard error of mean,expect for special instructions（Standard error of means, sem).

3 Results and Analysis

3.1 Habitat suitability and rodent community structure under different disturbances: habitat suitability.
Habitat suitability refers to the possibility of an animal taking advantage of a particular habitat (Wang Y H et al. 2008). Vegetation characteristic is often used as a indicator of the suitability or quality of small mammals habitat (Jorgensen E E 2002). This study analysis vegetation characteristics under different disturbed habitats, from three aspects of habitat productivity, the intensity of disturbance and habitat concealment, to explore the desert rodent habitat suitability under different disturbance.
3.1.1 Habitat productivity
With the total biomass of plants (the sum of overground shrubs biomass and overground herbaceous biomass) to measure the productivity of animal habitat. The results showed that the total plant biomass of farmland area was significantly higher than that of other three disturbed areas. Although there was no statistically significant differences in total biomass between grazing exclusion areas, rotational grazing areas and over grazing areas, but the total biomass in grazing area was lower than that in grazing exclusion areas. The nutrition balanced value of reaumuria is low, there are significant differences in total biomass after the removal of the biomass of reaumuria (Wu Keshun et al. 2010), biomass was from high to low: grazing exclusion areas> rotational grazing areas> over grazing areas (Table 3).
 
3.1.2 The degree of habitat disturbed.
Measuring the degree of habitat disturbance with stocking capacity, grazing frequency and to: stocking capacity and grazing frequency in each habitat has been mentioned in the previous paragraph, this analysis focuses on the plant diversity of disturbed habitat. Shannon-Wiener index of shrubs has significant difference among different disturbance, which are affected by yearly, from the high to low: rotational grazing areas>grazing exclusion areas>over grazing areas>reclamation areas (Table 3). Moderate disturbance hypothesis (Connell J H 1978) suggests that moderate disturbances may result in higher plant diversity, so the grazing areas may be subject to moderate disturbances, while over grazing areas and reclamation areas are relatively heavy. In addition to frequent agricultural operation in reclamation areas may lead to disturbance in the rodent community has intensified.
3.1.3 Habitat concealment.
Use the height, coverage and density of plant and other indicators to measure the habitat concealment. There were significant difference in height,coverage and density of plant in different disturbed habitats, which are affected by yearly. Shrub height as follows: reclamation areas> grazing exclusion areas> rotational grazing areas> over grazing areas.Shrub coverage cover grazing overgrazing is the highest, the lowest is in over grazing areas; The shrub density was in the order of: grazing exclusion areas> rotational grazing areas> over grazing areas> reclamation areas, but there was no significant difference between rotational grazing areas and over grazing areas (Table 3). Herbaceous height was as follows: reclamation areas> grazing exclusion areas> rotational grazing areas> over grazing areas, there is no significant difference between grazing exclusion areas and rotational grazing areas; Herbaceous height in reclamation areas is the highest, the lowest is in grazing exclusion areas; Herbaceous height in reclamation areas is significant higher than that in other three disturbed habitat, although no significant difference in herbaceous coverage between other three disturbed habitat, which in grazing exclusion areas still higher than over grazing areas.
3.1.4 Structure of community of rodent: community composition of rodent.
2006-2011 were distributed live cage 75264 cage days, capturing rodents 3 families, 7genera, 9 kinds of 5129 rodents. There are three species of Dipodidae, including: Northern Three-toed Jerboa (Dipus sagitta), Mongolian five-toed Jerboa (Allactaga sibirica)and Mongolian feather-drenghill (Stylodipus andrewsi); 5 species of Cricetidae, including: Midday jird (Meriones meridianus), Mongolian gerbil (Meriones unguiculatus), Desert hamstero (Phodopus roborovskii), Chinese Hamster (Cricetulus barabensis) and (Cricetulus eversmanni); 1 species of Sciuridae: Alashan Gound Squirrel (Spermophilus alaschanicus). According to the type of exercise can be divided into: two-legged species of three species of rodents, including Three-toed Jerboa, Mongolian five-toed Jerboa and Mongolian feather-drenghill four-legged species of rodents,including Midday jird, Mongolian gerbil, Desert hamstero, Chinese Hamsterand Cricetulus eversmanni; between the bipedal and quadruped activities of the 1 species, for the Alashan Gound Squirrel. According to the activity time of rodents, there were 2 species of daytime rodents in the study area, which were Mongolian gerbil and Alashan Gound Squirrel. The remaining 7 rodents were nocturnal rodents. The total capture rate was significantly different among different disturbances, and that of the grazing exclusion areas and reclamation areas were significantly higher than that of the reclamation area. The total capture rate in the grazing exclusion areas was significantly higher than over grazing areas (Fig.2).
 
3.1.5 Determination of dominant species.
In this study, on the basis of the rodent capture rate to determine the dominant species. The rodent communities with different structures were formed under different disturbances, and the reclamation area was a rodent community dominated by Midday jird. The grazing exclusion areas was a rodent community dominated by Three-toed Jerboa and Midday jird. The rotational grazing areas was a rodent community dominated by Three-toed Jerboa and Mongolian five-toed Jerboa. The over grazing areas was a rodent community dominated by Mongolian Five-toed Jerboa and Three-toed Jerboa (Table 4).
3.2 The composition of life history strategist in the different disturbed community
According to the classification of the life history strategy, this study calculated the species richness and and the structure of community of the different life history strategy in different disturbed habitats. From the species richness of different strategist: with the same disturbance, the species richness of the rRS strategist is significantly more than that of the KSS strategist in the reclamation area; although there is no significant difference between the species richness of rRS and KSS strategist in the other three areas, but the species richness of the KSS strategist is still more than that of rRS strategist in all areas (Fig.3). Under different disturbances, there are no significant difference between the species richness of the rRS strategist in all areas, but the species richness of KSS strategy in the reclamation area is significantly lower than that of other districts(Fig.4).
From the composition proportion of rRS/KSS strategist: Under the same disturbance, the proportion of rRS strategist was significantly higher than that of KSS strategist in reclamation area, and the other three areas is exactly the opposite, and significance successively increase from grazing exclusion areas, rotational grazing areas to over grazing areas. (P= 0.034,0.003,0.0001, Fig.3). Under different disturbance, the capture rate of rRS the strategy is highest in the grazing area, the lowest in the pastoral areas; the capture rate of KSS strategy in the reclamation area is the lowest, there is no significant difference in the other three areas(Fig.4).
 
3.3 W-statistics of rodent community under different disturbances
Showed by W-statistics: The W statistical values of reclamation area and over grazing area fluctuated greatly; smaller fluctuations in the rotational grazing area; the W statistics value with minimum fluctuation in grazing area; which indicates that grazing and reclamation exacerbated the number and biomass of rodent community changes. The W statistical values of reclamation area is less than 0 in most years; The W statistical values is larger than 0 in most years in rest areas. The analysis of variance showed the W statistical values of over grazing areas is the highest, but the lowest of the reclamation,under the four disturbances. This shows that cumulative biomass of grazing exclusion areas,rotational grazing areas and over grazing areas is dominated by individual species richness, and the cumulative abundance advantage under the control of multiple species, the situation of reclamation area is opposite. KSS strategist in grazing exclusion areas and grazing areas hold a dominant position, and in the over grazing areas is more significant.
3.4 Rodent community stability under different disturbances
3.4.1 Rodent community diversity under different disturbances
The analysis of variance showed that the number of species in reclamation areas is significantly lower than the other three disturbed area, but no significant difference between grazing exclusion areas, rotational grazing areas and over grazing areas. The Shannon-Wiener index and Simpson’s diversity index of reclamation areas is significantly lower than that of the other three disturbed area. Both index show the following sort: grazing exclusion areas<rotational grazing areas<over grazing areas, but this difference is not statistically significant. Although the evenness index was not significant between the four disturbances, but the lowest value was found in the reclamation area, the other three kinds of disturbance sort share the same with the sort of Shannon-Wiener index and Simpson’s diversity index(Table 5).
 
3.4.2 The variability of rodent communities under different disturbance
  The measure of community variability is expressed by coefficient of variation (Bai Yongfei and Chen Zuozhong 2000). The variability of dominant species richness of rodent communities under different disturbance is the lowest, non dominant species richness in large variability. Over grazing exclusion areas, rotational grazing areas and over grazing areas, low coefficient of variation for some of the larger size species, large variation mainly occurred in the smaller size species. Over reclamation areas, variability of all rodent were higher in more than 70%. S.A and C.E are rare species in the region, the capture rate is few, and greater variability (Table 6).
In different components of strategist in community, the coefficient of variation of richness and species number of KSS strategist in reclamation areas are all greater than that of rRF strategist, and other disturbed area opposite to this case. The variability of richness and species number of KSS strategist are all the lowest in grazing exclusion areas; the richness variability of rRF strategist is highest in over grazing area; the species number variability of rRF strategist is highest in grazing exclusion areas and rotational grazing areas. Under different disturbance, the maximum coefficient of variation of total species richness of rodent communities found in the reclamation areas, the minimum found in rotational grazing areas. Under different disturbance, the maximum coefficient of variation of species number of rodent communities found in the rotational grazing areas, the minimum found in grazing exclusion areas(Table 7).
 

4 Discussion and Conclusion

4.1 Habitat suitability
Reclamation significantly increased the total biomass of the habitat, and the impact of grazing on the total biomass of the plant was not significant. However, there was a significant difference in the total biomass of the plant among different grazing treatments after removal of the reaumuria with lower nutrient balance. The productivity of each disturbance habitat is highest in the reclamation area. There was no significant difference in productivity among grazing exclusion areas, rolational grazing areas and over grazing areas, but the over grazing significantly reduced the plant biomass with higher nutrient balance. This indicates that grazing does not affect the total biomass of the plant but affects plants with high nutritional value. Some studies have found that livestock grazing in grassland tends to choose high-quality plants rather than low palatability plants (Peco B et al.1978). Fleischner (1994)also argues that overgrazing has led to a decline in the quality of pastures that have dominated the species that are resistant to grazing. This study therefore supports the idea that livestock grazing affects the quality of the food rather than the total biomass of the plant. Reclamation and overgrazing significantly reduced the diversity of plants, but the plant diversity in the rotational grazing area was higher than that in the grazing exclusion area. The intermediate disturbance hypothesis suggests that moderate interference suppresses the competitive exclusion of dominant species and thus has a higher diversity than severe and mild interference (Coonell J H 1978). In this study, the effects of interference on plant diversity support the intermediate disturbance hypothesis. Overgrazing reduces the height and coverage of plants and the density of shrubs. According to the diversity of plants, the interference intensity of each disturbed areas in the reclamation area is the largest, next is in over grazing areas, and the grazing exclusion areas and rotational grazing areas are the smallest. According to the habitat productivity, food quality and concealment,animals in overgrazing areas are subject to the greatest environmental pressure, followed by rotational grazing, grazing exclusion, and reclamation (Fig.6).
4.2 Stability and diversity
The diversity-stability hypothesis suggests that the higher the species diversity in the community, the more stable the community (MacArthur R 1955). Approximately 70% of the studies support the view that higher diversity leads to low-variation communities (Lves A R and Carpenter S R 2007) And May thinks the simple community is more stable (May R M 1973). This view has been supported by some scholars (Gardner M R and Ashby W R 1970; Goodman D 1975; Gilpin M E 1975). In addition, some scholars believe that the two views do not conflict. Tilman (1996) holds that when the external interference damage some species, the number of other not injured species who competing with those injured species in the community increased. This compensation increases the stability of the total biomass of the community, but also makes the species abundance more varied. He argues that this conclusion supports May's consideration of the impact of diversity on the population and the apply diversity-stability hypotheses to communities and ecology processes. Studies by McCann (2000) showed that diversity usually contributes to stability, but diversity is not the driving force of stability. The ability of a community to accommodate species or functional groups with different responsiveness limits stability. He argues that strong interactions in the randomly constructed communities mentioned by May, strong interactions are not necessarily coupled with weak interactions (this weak interaction mitigates the potential of unstable factors), resulting in complex and violent change diverse communities. MacArthur's hypothesis, however, builds up many channels for weak interactions, thereby mitigating the instability caused by strong consumer-resource interactions. So the two hypotheses do not conflict, but their emphasis on different priorities. There may be a complex nonlinear relationship between diversity and stability, rather than a simple increase or decrease relationship (Hairston N et al.1968; King A W et al.1983; Grime J P.1998). In this study, the reclamation area had a significantly lower diversity of rodent community, so the stability of the rodent community in the reclamation area was low. The effects of grazing on the stability of the rodent community may come from species with different responsiveness or response of functional group to disturbances.
4.3 Stability and resistance, resilience and variability
Resistance and resilience are considered to be two important components of stability (Tilman D and Downing J A 1994). MacArthur argues that this community is unstable if the abnormal changes in the abundance of a species significantly affect the abundance of other species in the community. On the contrary, the higher the community stability. This stability may be caused by two ways: first, the pattern of the interaction between the species of the community; second, the stability of the species is inherent (MacArthur R 1955). But then many researchers have focused on the first path, that is, the relationship between the diversity and stability of the community and the interaction between species (Hairston N et al. 1968; King A W et al.1983; Pimm S L 1984; Tilman D and Downing J A 1994), and few have explored the inherent stability of species. McNaughton argues that species with different adaptation patterns, whose abundance volatility, may be a mechanism for communities to remain stable in a changing environment. The adaptation pattern here is likely to be related to the life history of the species (MacNaughton S J 1977). Studying how different systems are dominated by creatures of different life history can help solve the problem of resistance and resilience (Pimm S L 1984).
Pianka (1970) argues that the increase in animal size has reduced environmental resistance in many ways; larger creatures have less potential predators. Those living creatures older than one year have to deal with the physical or biological conditions of the whole range in order to survive, and for those living less than one year, they will only experience part of the short time range. So the longer life, larger size organisms can better buffer the changes in the environment, their population size is not as smaller, relatively short life of biological changes as dramatic. While stable populations also provide the necessary conditions for stable communities. Studies in the same study area have shown that the survival rate of the large squirrel is less affected by the small size of the Meridian gerbil (Yuan Shuai 2013). Mammals with fast strategies have relatively high reproductive elasticity and lower adult survival elasticity. While slow-minded mammals have lower reproductive elasticity and higher adult or juvenile survival elasticity. Since the decline in adult survival is not conducive to the use of slow-minded mammalian populations, the recovery of the slow strategy will be slower once the outside interference exceeds its ability to resist environmental resistance (Heppell S S et al. 2000). MacArthur (1967) argues that high fecundity strategists can recover through frequent non-density control deaths through higher population growth rates. Therefore, KSS strategists have higher resistance and lower resilience to changes in the external environment; rRF strategists have lower resistance and higher resilience to the external environment (MacArthur R H and Wilson E O 1967; Pianka E R 1970; Heppell S S et al. 2000). Lepš and others Argue that two different stability may be determined by the life strategy and external variables of the dominant species (Lep J et al. 1982). The increase in KSS strategy in the community means an increase in community resistance, whereas an increase in rRF strategists means an increase in community resilience. The variability of community stability reflects the effect of external variables on the community^[44]. In this study, the reclamation disturbance significantly reduced the number of species, the catch ratio, and the capture rate of the KSS strategy. Therefore, reclamation had a significant negative effect on KSS, so the reclamation area had lower community resistance, The variability of the resistance is also greater than the other three disturbed areas. With the increase of grazing intensity, the capture rate of KSS and rRF strategists showed a decreasing trend, and the decline trend of rRF strategists was more significant. Which led to the KSS strategists in the community in the dominant advantage gradually increased. So grazing has greater impact on rRF strategists than KSS strategists (Fig.6). Because the overgrazing areas has lower rRF strategy abundance than other regions, it has lower community resilience, and this resilience variability is larger than the other three disturbed zones. So reclamation reduces the resistance of the rodent community. There was no significant difference in species richness and population abundance between KSS strategists and other disturbance areas, and the variability of this resistance was not very different among the three disturbances, so the effect of grazing on community resistance was not great. Because the overgrazing area has a lower rRF strategists abundance than the grazing exclusion areas, it has lower community resilience, and this resilience variability is larger than the other two disturbed areas. So grazing reduces the resilience of the community.
In summary, grazing exacerbates the variability of the rodent community by increasing the variability of the number of KSS strategies and the species number; significantly reducing the resilience of the rodent community by significantly reducing the abundance of the rRF strategy; By increasing the variability of the resistance and reducing the resilience of the large community weaken the stability of the rodent community. Because different species have different sensitivities to disturbances, it is difficult to further explain the mechanism hidden in this response (Ma Keming et al. 2004). If the division of these sensitive species from the perspective of the life history strategy, rather than the classification of species, may lead to some common things, that is, under different disturbances, these sensitive species to take what kind of life strategy.
 

Acknowledgements

This experiment was funded by the National Natural Science Foundation of China (Grant#31602003, 31772667 and 31560669) and Ministry of Education Key Laboratory of Grassland Resources. We wish to acknowledge XXX for revising the manuscript. Thanks are also due to Muha Cha, Yanjing Han, Fushun Zhang, Jin E and Shengnan Ji for data collection, Ganghui Zhuang, Hongjun Gan, Guohai Sang and Manliang Zhao for assistance in the field.  
 

Literature Cited

Bieber,C.,R.Juškaitis.,C.Turbill.,AND T.Ruf.2011. High survival during hibernation affects onset and timing of reproduction. Oecologia 169:1-12.
Bai Yongfei,and Chen Zuozhong.2000. Effects of long-term variability of plant species and functional groups on stability of a Leymus Chinensis community in the Xilin River Basin,Inner Mongolia.Acta Phytoecologica Sinica 24:641-647.
Connell,J.H.1978. Diversity in tropical rain forests and coral reefs. Science 199:1302-1310.
Clarke,K.R.1990. Comparisons of dominance curves. Journal of Experimental Marine Biology and Ecology 138:143-157.
Flowerdew,J.R.,R.F.Shore.,S.M.C.Poulton.,and T.H.Sparks.2004. Live trapping to monitor small mammals in Britain. Mammal Review 34:31-50.
Fleischner,T.L.1994. Ecological costs of livestock grazing in western North America. Conservation Biology 8:629—644.
Grimm,V.,E.Schmidt,and C.Wissel.1992. On the application of stability concepts in ecology. Ecological Modelling 63:143-161.
Goodman,D.1975. The theory of diversity-stability relationships in ecology. Quarterly Review of Biology 50:237-266.
Gilpin,M.E.1975. Stability of feasible predator-prey systems. Nature 254:137-139.
Gardner,M.R.,and W.R.Ashby.1970. Connectance of Large Dynamic (Cybernetic) Systems: Critical Values for Stability. Nature 228:784-784.
Holling,C.S.1973. Resilience and Stability of Ecological Systems. Annual Review of Ecology and Systematics 4:1-23.
Hairston,N.,J.Allan.,R.Colwell,D.Futuyma.,J.Howell.,M.Lubin.,J.Mathias.,and J.Vandermeer.1968. The relationship between species diversity and stability: an experimental approach with protozoa and bacteria. Ecology 49:1091-1101.
Huang Wenji,and Xu Shiju.1995. Rodent of china. Fudan University press.Shanghai,China.
Heppell,S.S.,H.Caswell.,and L.B.Crowder.2000. Life histories and elasticity patterns: perturbation analysis for species with minimal demographic data. Ecology 81:654-665.
Ives,A.R.,and S.R.Carpenter.2007. Stability and diversity of ecosystems. Science,  317:58—62.
Jiyi Zhang,and Halin Zhao.2003. Review on the study of vegetation stability. Chinese Journal of Ecology 22:42-48.
Jones,A.L., and W.S.Longland.1999. Effects of cattle grazing on salt desert rodent communities. The American midland naturalist 141:1-11.
Jorgensen,E.E.2002. Small mammals: consequences of stochastic data variation for modeling indicators of habitat suitability for a well-studied resource. Ecological Indicators 1:313-321.
King,A.W., and S.L.Pimm.1983. Complexity, diversity, and stability: a reconciliation of theoretical and empirical results. American Naturalist 122:229-239.
Lu Haoyong,Ma Yong,and Zhao Guizhi.1988. Classification,prediction and control of rodent damage. China agriculture press. Beijing,China.
Lebl,K.,C.Bieber.,P.Adamík.,J.Fietz.,P.Morris.,A.Pilastro., and T.Ruf . 2011. Survival rates in a small hibernator, the edible dormouse: a comparison across Europe. Ecography 34:683-692.
Lep,J.,J.Osbornová-Kosinová.,and M.Rejmánek.1982. Community Stability, Complexity and Species Life History Strategies. Vegetatio 50:53-63.
MacArthur,R.1995. Fluctuations of Animal Populations and a Measure of Community Stability. Ecology 36:533-536.
May,R.M.1973. Stability in Randomly Fluctuating Versus Deterministic Environments. The American Naturalist 107:621-650.
Ma Yong, Wang Fenggui, Jin Shanke,and Li Sihua.1987. Glires(Rodents and lagomorphs) of northern Xinjiang and their zoogeographical distribution.China Science Publishing & Media ltd.Beijing,China.
McCann,K.S.2000. The diversity-stability debate. Nature 405:228-233.
Grime,J.P.1998. Benefits of plant diversity to ecosystems: immediate, filter and founder effects. Journal of Ecology 86:902-910.
McNaughton,S.J.1977. Diversity and stability of ecological communities: a comment on the role of empiricism in ecology. American Naturalist 111:515-525.
MacArthur,R.H.,and E.O.Wilson.1967. The theory of island biogeography. Princeton University Press. Princeton, New Jersey.
Ma Keming,Fu Bojie,and Chen Liding.2004.Research advances on contemporary landscape ecology.Pp 534-553 in Progresses and perspective of ecological research..
Ojeda,R.A., and S.Tabeni. 2009. The mammals of the Monte Desert revisited. Journal of Arid Environments 73:173-181.
Pimm,S.L.1984. The complexity and stability of ecosystems. Nature 307:321-326.
Pianka,E.R.1970. On r and K selection. The American Naturalist 104:592-597.
Peco,B.,I.de.Pablos.,J.Traba., and C.Levassor.2005. The effect of grazing abandonment on species composition and functional traits: the case of dehesa grasslands. Basic and Applied Ecology 6:175-183.
Read,A.F., and P.H.Harvey.1989. Life history differences among the eutherian radiations. Journal of Zoology 219:329-353.
Suther,l.and J.P.1981. The fouling community at Beaufort, North Carolina: a study in stability. American Naturalist 118:499-519.
Sun Ruyong.1997. Life history strategy. Bulletin of Biology 32:2-4.
Tilman,D. and  J.A.Downing.1994. Biodiversity and stability in grasslands. Nature 367:363-365.
Tilman,D.1996. Biodiversity: Population Versus Ecosystem Stability. Ecology 77:350-363.
Wu Xiaodong，Yuan Shuai，Fu Heping，Zhang Xiaodong， Zhang Fushun, and Gao Quanrong.2016. Responses of dominant rodentspecies to climate change in different disturbed habitats in the Alashan desert. Acta Ecologica Sinica  36:1765-1773.
Wu Xiaodong, Fu Heping,and Yang Zelong.2009. Study on rodents in typical semi-desert and desert areas in China. China Science Publishing & Media ltd.Beijing,China.
Wang,Y.H., K.C.Yang.,C. Bridgman., and L.K.Lin.2008. Habitat suitability modelling to correlate gene flow with landscape connectivity. Landscape Ecology 23:989-1000.
Wu Keshun,Fu Hua, Zhang Xueying,Niu Decao,Talateng.2010. Seasonal Dynamics and Nutritional Equilibrium Valuation of Eight Kinds of Forage Nutrients in Alashan Desert.Arid Zone Research 27:257-262.
Yuan Shuai.2013. Response mechanism of desert rodent community and dominant species to different distuebances. Ph.D. dissertation,Inner Mongolia Agricultural University.Inner Mongolia,China.
Zhang Zhibin,and Wang Zuwang.1998. Ecology and management of rodent pests in agriculture.China ocean press.Beijing,China.
 
 
修改：把题目、作者都加上
      把图表放到文中对应位置
      参考文献及引用格式参考Journal of Mammalogy

 
Table 2 Biological features of captured rodents species

Note:”-”:missing data.
 
Table3  Vegetation difference among different disturbed habitats

Note：sem：Standard error of means
 
Table 4  Total proportion of each rodent species in community in 6 years scale (%)

Note：sem：standard error of means
 
Table 5  Difference in rodent community diversity among disturbances

Note：sem：Standard error of means
 
Table 6  Variability of rodent populations subjected to different disturbances （%）

 
 
Table 7  Variability of rodent communities subjected to different disturbances （%）