**2. Material and method**

### **2.1 Study site**

*Plant Communities and Their Environment*

resource availability [15, 16].

physiological stress [13].

As one of the most important and largest terrestrial ecosystems in the world, grasslands cover 30% of the land surface and are mainly distributed in arid and semiarid regions [1]. Due to global climate change and human activity, such as heavy grazing, grasslands in this area have undergone desertification and even virtually disappeared in recent decades [2, 3], making restoration process urgent for degraded grasslands [4, 5]. Current studies about grassland restoration mainly focus on several key components: community composition and structure, species diversity, soil properties and vegetation succession process [6–10]. Grassland is considered very sensitive to climate changes [11–14] and also is influenced by soil

Compared with forest ecosystem and cropland ecosystem, aboveground net primary productivity (ANPP) of grasslands is highly temporally variable [16, 17]. Specifically, climate-driven variability in grassland productivity has important effects on the global carbon balance, ecosystem service delivery, profitability of pastoral livelihoods and the sustainability of grassland resources [11, 18, 19]. Many ecologists have analysed the impacts of annual precipitation and temperature on ANPP at regional and continental scales [17, 20–23], while numerous site-specific reports have indicated that interannual variability in ANPP is poorly or even not at all correlated with annual climate conditions [19, 24, 25]. Changes in precipitation or temperature during certain parts of the year have been proven to be more relevant drivers of ANPP than annual changes [26–29], and the impacts on vegetation production varied with seasons [13, 28, 30, 31]. For instance, warming in early spring increased grassland productivity by ameliorating cold temperature constraints on plant growth in northern mid- and high latitudes [32, 33] and advancing spring greening phenology [34–36]. Temperature increases in summer; however, it can depress productivity by reducing soil moisture and intensifying

The Loess Plateau of China has a total area of about 52 million hectares and is widely known for its fragile ecological environment, frequent severe droughts and problems with water runoff and soil erosion [37]. In recent years, the complicated landscape, frequent droughts and severe soil erosion have attracted worldwide attention and caused sustained deterioration of the ecosystem of this region. In contrast to numerous studies in the temperate grasslands of Inner Mongolia and the alpine grasslands of the Tibetan Plateau, very few reports are available on responses of grassland productivity to climate variability on the more arid Loess Plateau in China [3], especially with respect to responses to seasonal climatic variability. Restoration of the natural vegetation is regarded as the most effective method for

As a major determinant of nutrient cycling, litter decomposition is a fundamental process of grassland ecosystem functioning [39]. Decomposition traits of plant litters are affected by a number of factors, including litter quality, abiotic environment and soil organisms [40]. In general, plant litters with high C:N ratio and lignin concentration are supposed to have slow decomposition and nutrient immobilisation processes, whereas low C:N ratio and low lignin concentration contribute to fast decomposition and nutrient mineralisation processes. Decomposition traits of plant materials may vary with succession stages. For example, late-seral dominant grasses normally had high tissue N concentrations, low C:N ratios and lignin concentrations, which result into fast decomposition rate and enhanced nutrient

Most previous studies have focused on plant species richness and diversity in abandoned croplands following short-term grazing exclusion in China [8, 41, 42]. Few studies

changing the ecological environment of the Loess Plateau [7, 8, 38].

**1. Introduction**

**136**

mineralisation.

This study was conducted in Yunwu Mountain National Nature Reserve on the Loess Plateau (106°24′–106°28′ E, 36°13′–36°19′ N) (**Figure 1**) [45, 50].

**Figure 1.** *Location of experimental site.*

Grassland in this area was restored from grazing as a long-term ecological monitoring station since 1982. The elevation of this study area is 1800–2180 m and has a total area of 6660 hm<sup>2</sup> . The mean annual temperature is 7.01°C, and there are on average 137 frost-free days per year [49]. The mean annual precipitation is 425 mm, with 60–75% of rainfall falling during July–September. The mean annual evaporation is 1017–1739 mm. Snow cover depth in winters averaged 1.2 cm during the dormancy period. The vegetation type is typical steppe. *Gentianaceae*, *Stipa* and *Potentilla* are important plant components, and the main dominant species include *S. bungeana*, *Stipa grandis*, *S. przewalskyi*, *Thymus mongolicus*, *Artemisia sacrorum*, *Potentilla acaulis* and *Androsace erecta* [45]. Soil type is montane grey-cinnamon soil [45].

### **2.2 Experimental design and sampling**

## *2.2.1 Grassland ecological survey*

The grassland sites have been restored from grazing exclusion since 1982, and consequently goat grazing was excluded [45, 49, 50]. Three equal-sized transect of 300 × 100 m was established at the top, middle and down positions of the same slope, respectively. And, 15 quadrats (1 × 1 m) were established within each transect. The vegetation survey was carried out in mid- or late August each year during 1982–2011. Plant coverage, height, species abundance and plant density in each quadrat were measured. Aboveground parts of grassland plants were clipped and dried at 65°C for 48 h to determine aboveground biomass [43]. Plant roots of 0–120 cm soil layers were collected with a soil auger of 9 cm diameter, then were washed and dried to determine belowground biomass.

Important value (IV) was used to describe the importance of species in grassland community during the restoration process. Shannon-Wiener index was used to indicate diversity and evenness of plant community [50]. All indices were calculated according to 8 and 43.

Important value (IV)

\*\*Important\*\* value (IV)

$$IV = \frac{RH + RC + RA + RF}{4} \tag{1}$$

where IV is the important value, RH is the relative height, RC is the relative coverage, RA is the relative abundance and RF is the relative frequency.

Diversity index (H), using Shannon-Wiener index

$$H = -\sum\_{i=1}^{S} P\_i \ln P\_i \tag{2}$$

**139**

productivity.

*Responses of Community Structure, Productivity and Turnover Traits to Long-Term Grazing…*

In early October of 2013, the leaf litterbags of three *Stipa* species were transferred to grassland site restored for 23 years. Four plots of 10 × 10 m were established, and seven leaf litterbags of each *Stipa* species were placed on the soil surface and secured in place with iron nails on each of four plots. Four leaf litterbags of each *Stipa* species were harvested after 1, 3, 6, 9, 12, 18 and 24 months of

In the laboratory, leaf litters were removed from bags, cleaned to remove any extraneous material and weighed after drying at 65°C for 48 h. Leaf litters were analysed for carbon (C), nitrogen (N) and phosphorus (P). C was determined by oxidation with potassium dichromate in a heated oil bath. N was determined by the

According to [55], decomposition rate (k) of leaf litters was estimated by the

where X is the remaining mass, X0 is the initial mass and t is the decaying time (year). Based on the nutrient concentration and remaining mass, we further calculated nutrient accumulation index (NAI) for C, N and P of leaf litters during decomposi-

> *Xt* × *Ct X*<sup>0</sup> × *C*<sup>0</sup>

where X0 and C0 indicate initial leaf litter mass and chemical element concentration, respectively. Xt and Ct indicate remaining leaf litter mass and chemical

All data in the paper are presented as mean ± standard error. A two-way analysis of variance was conducted to determine the effects of decomposition time, species and their interaction on decomposition rate, nutrient concentration and NAI of leaf litters. A linear mixed model was used to examine correlations of vegetative indices with restoration time, productivity with climate variables and remaining mass with decomposition time. Significant differences of all statistical tests were estimated at a significance level of P < 0.05. All statistical analyses were performed using SPSS

Partial least squares (PLS) regression was used to analyse the responses of grassland productivity to variation in daily temperature and precipitation during all 365 days of the year based on data for 1992–2011 [58, 59]. The two major outputs of PLS analysis are the variable importance in the projection (VIP) and standardised model coefficients. The VIP threshold for considering variables as important is often set to 0.8 [60]. The standardised model coefficients indicate the strength and direction of the impacts of each variable in the PLS model. The root-mean-square errors (RMSE) of the regression analyses were calculated to determine the accuracy of the PLS model. In the PLS analyses, periods with VIP greater than 0.8 and high absolute values of model coefficients represent the relevant phases influencing grassland productivity. Positive model coefficients indicate that increasing temperature or precipitation during the respective period should increase ANPP, while negative model coefficients imply negative impacts on

element concentration after a period of time t (year), respectively.

= *e*<sup>−</sup>*kt* (3)

× 100% (4)

*X*0

semimicro Kjeldahl method. P was determined by Olsen method [54].

*DOI: http://dx.doi.org/10.5772/intechopen.85306*

negative exponential decay function:

tion process [56, 57]:

**2.3 Data analyses**

18.0 (SPSS Inc., Chicago, IL, USA).

\_\_\_*<sup>X</sup>*

*NAI* = \_\_\_\_\_\_

incubation.

where S is the total species number of a quadrat and Pi is the relative importance value of species i.

### *2.2.2 Litter decomposition experiment*

Considering the difficulty of gathering sufficient senesced leaves, leaves of three *Stipa* species (*S. bungeana*, *S. grandis* and *S. przewalskyi*) were collected in August of 2013 and then dried at 40°C as decomposition materials, according to other decomposition studies [51–53]. Leaf litters were cut into pieces of 10 cm in length and enclosed in nylon bag (15 g bag<sup>−</sup><sup>1</sup> , 15 × 10 cm, 0.15 mm mesh).

*Responses of Community Structure, Productivity and Turnover Traits to Long-Term Grazing… DOI: http://dx.doi.org/10.5772/intechopen.85306*

In early October of 2013, the leaf litterbags of three *Stipa* species were transferred to grassland site restored for 23 years. Four plots of 10 × 10 m were established, and seven leaf litterbags of each *Stipa* species were placed on the soil surface and secured in place with iron nails on each of four plots. Four leaf litterbags of each *Stipa* species were harvested after 1, 3, 6, 9, 12, 18 and 24 months of incubation.

In the laboratory, leaf litters were removed from bags, cleaned to remove any extraneous material and weighed after drying at 65°C for 48 h. Leaf litters were analysed for carbon (C), nitrogen (N) and phosphorus (P). C was determined by oxidation with potassium dichromate in a heated oil bath. N was determined by the semimicro Kjeldahl method. P was determined by Olsen method [54].

According to [55], decomposition rate (k) of leaf litters was estimated by the negative exponential decay function:

$$\frac{X}{X\_0} = e^{-kt} \tag{3}$$

where X is the remaining mass, X0 is the initial mass and t is the decaying time (year).

Based on the nutrient concentration and remaining mass, we further calculated nutrient accumulation index (NAI) for C, N and P of leaf litters during decomposition process [56, 57]:

$$\text{NAI} = \frac{X\_t \times C\_t}{X\_0 \times C\_0} \times \text{100}\text{\%} \tag{4}$$

where X0 and C0 indicate initial leaf litter mass and chemical element concentration, respectively. Xt and Ct indicate remaining leaf litter mass and chemical element concentration after a period of time t (year), respectively.

### **2.3 Data analyses**

*Plant Communities and Their Environment*

and has a total area of 6660 hm<sup>2</sup>

[45]. Soil type is montane grey-cinnamon soil [45].

**2.2 Experimental design and sampling**

*2.2.1 Grassland ecological survey*

lated according to 8 and 43. Important value (IV)

value of species i.

biomass.

Grassland in this area was restored from grazing as a long-term ecological monitoring station since 1982. The elevation of this study area is 1800–2180 m

there are on average 137 frost-free days per year [49]. The mean annual precipitation is 425 mm, with 60–75% of rainfall falling during July–September. The mean annual evaporation is 1017–1739 mm. Snow cover depth in winters averaged 1.2 cm during the dormancy period. The vegetation type is typical steppe. *Gentianaceae*, *Stipa* and *Potentilla* are important plant components, and the main dominant species include *S. bungeana*, *Stipa grandis*, *S. przewalskyi*, *Thymus mongolicus*, *Artemisia sacrorum*, *Potentilla acaulis* and *Androsace erecta*

The grassland sites have been restored from grazing exclusion since 1982, and consequently goat grazing was excluded [45, 49, 50]. Three equal-sized transect of 300 × 100 m was established at the top, middle and down positions of the same slope, respectively. And, 15 quadrats (1 × 1 m) were established within each transect. The vegetation survey was carried out in mid- or late August each year during 1982–2011. Plant coverage, height, species abundance and plant density in each quadrat were measured. Aboveground parts of grassland plants were clipped and dried at 65°C for 48 h to determine aboveground biomass [43]. Plant roots of 0–120 cm soil layers were collected with a soil auger of 9 cm diameter, then were washed and dried to determine belowground

Important value (IV) was used to describe the importance of species in grassland community during the restoration process. Shannon-Wiener index was used to indicate diversity and evenness of plant community [50]. All indices were calcu-

*IV* <sup>=</sup> *RH* <sup>+</sup> *RC* <sup>+</sup> *RA* <sup>+</sup> *RF* \_\_\_\_\_\_\_\_\_\_\_\_\_ 4 (1)

where IV is the important value, RH is the relative height, RC is the relative

*i*=1 *S*

where S is the total species number of a quadrat and Pi is the relative importance

Considering the difficulty of gathering sufficient senesced leaves, leaves of three *Stipa* species (*S. bungeana*, *S. grandis* and *S. przewalskyi*) were collected in August of 2013 and then dried at 40°C as decomposition materials, according to other decomposition studies [51–53]. Leaf litters were cut into pieces of 10 cm in length

, 15 × 10 cm, 0.15 mm mesh).

*Pi* ln*Pi* (2)

coverage, RA is the relative abundance and RF is the relative frequency.

Diversity index (H), using Shannon-Wiener index

*H* = −∑

*2.2.2 Litter decomposition experiment*

and enclosed in nylon bag (15 g bag<sup>−</sup><sup>1</sup>

. The mean annual temperature is 7.01°C, and

**138**

All data in the paper are presented as mean ± standard error. A two-way analysis of variance was conducted to determine the effects of decomposition time, species and their interaction on decomposition rate, nutrient concentration and NAI of leaf litters. A linear mixed model was used to examine correlations of vegetative indices with restoration time, productivity with climate variables and remaining mass with decomposition time. Significant differences of all statistical tests were estimated at a significance level of P < 0.05. All statistical analyses were performed using SPSS 18.0 (SPSS Inc., Chicago, IL, USA).

Partial least squares (PLS) regression was used to analyse the responses of grassland productivity to variation in daily temperature and precipitation during all 365 days of the year based on data for 1992–2011 [58, 59]. The two major outputs of PLS analysis are the variable importance in the projection (VIP) and standardised model coefficients. The VIP threshold for considering variables as important is often set to 0.8 [60]. The standardised model coefficients indicate the strength and direction of the impacts of each variable in the PLS model. The root-mean-square errors (RMSE) of the regression analyses were calculated to determine the accuracy of the PLS model. In the PLS analyses, periods with VIP greater than 0.8 and high absolute values of model coefficients represent the relevant phases influencing grassland productivity. Positive model coefficients indicate that increasing temperature or precipitation during the respective period should increase ANPP, while negative model coefficients imply negative impacts on productivity.
