Abstract :
Understanding community restoration state and the corresponding assembly mechanism is helpful to assess the restoration measures and predict community dynamics. We collected plots by shrub cover (low, medium and high) in three Caragana mircophylla shrub-encroached grasslands which were fenced since 1979, 1983, and 2003 (fencing duration) in the northern China, to explore the effect of fencing duration and shrub cover on the community restoration by vegetation investigation and phylogenetic approach. There were significant differences in community composition among different fencing duration or shrub cover treatments. Species richness in the site of fencing since 1979 or in the plots of high shrub cover was relatively higher than that in any other sites or plots. By phylogenetic analysis, functional traits were phylogenetically convergent. Based on the standardized effect sizes of mean pairwise distance (SESMPD) ranged from -1.96 to 1.96 in six out of nine plots, which suggested that stochastic processes dominated community assembly. SESMPD were lower than -1.96 in the rest three plots which indicated that competitive exclusion drove community assembly. These results indicated that the increase of fencing duration or shrub cover could enhance competitive exclusion. The present findings highlighted the importance of shrub in influencing the community composition and community assembly, supporting that shrub-encroached grassland is another stable state in the semi-arid northern China. Therefore, it is essential to distinct shrub-encroached grassland from degraded grasslands when formulating relevant conservation and management measures in the semi-arid regions.
Key words : Fencing duration; Shrub cover; Community assembly; Phylogenetic patern; Shrub-encroached grassland1 INTRODUCTION
Understanding the mechanisms underlying community assembly and predicting community dynamics has long been a central problem in ecology (Kraft et al., 2007; Dong et al., 2019), especially in the context of restoration communities (Jones et al., 2019). Proposed mechanisms range from niche-based assembly rules (Diamond 1975) to neutral assembly (Hubbell, 2001). Early workers relied on taxonomic ranks (e.g., species-to-genus ratios; Elton 1946) to infer ecological processes that structure communities. Recently, the availability of molecular phylogenies has provided a new method for more accurately measuring evolutionary relatedness, and has led to a rapidly growing body of research in this field (Weeb, 2002). A local community can be considered as a subset of a regional pool. Processes contribute to local community assembly (Karft et al., 2007; Liu et al., 2018), including environmental filtering, competitive exclusion and stochastic processes, and outcomes of these processes may be reflected in patterns of relatedness within a community (Cornwell et al., 2006).
Recently, assessing phylogenetic conservatism, i.e., assessing relationships between phylogenetic relatedness and ecological similarities, is usually done before give a conclusion by phylogenetic approach, because more and more ecologists found that ecological traits are not always conserved in evolution but convergent sometimes (Cavender-Bares et al., 2004; Liu et al., 2018). Phylogentic conservatism indicated that the closely related species tend to be similar of ecological and functional traits (conserved traits). Under this condition, phylogenetic clustering pattern suggests environmental filtering, phylogenetic overdispersal indicates competitive exclusion, and phylogenetic random pattern implies neutral processes dominating the community assembly (Webb et al. 2002; Kembel & Hubbell, 2006; Liu et al., 2018). However, when ecological traits are convergent, phylogenetic clustering suggests competitive exclusion, and phylogenetic overdispersal indicates environmental filtering driving the community assembly processes (Karft et al., 2007; Liu et al., 2018).
Due to global changes, large areas of terrestrial communities are degrading, especially in arid and semi-arid regions (Eldridge et al., 2011; Chen et al., 2014; Dong et al., 2019). For example, in the semi-arid Inner Mongolia Steppe in the northern China, large areas of native climax communities, Stipa grandisLeymus chinensis communities, degraded or even thicketized because of drought caused by climate change and over-grazing by cattle or/and sheep, with the loss of plant diversity and the decline of ecosystem functioning (Peng et al., 2013; Dong et al., 2019). In response, a number of restoration practices were carried out to restore the semi-arid grassland vegetation, and fencing is regarded as an effective management practice (Wu et al., 2009). Therefore, many researches have been focused on the restoration processes (or secondary succession) of different communities, including forest (Muscarella et al., 2015; Liu et al., 2018; Pinho et al., 2018), wetlands (Lozada-Gobilard et al., 2019), grasslands (Conradi et al., 2017; Barber et al., 2019; Dong et al., 2019) and abandoned agricultural fields (Dinnage, 2009).
Secondary succession is community assembly in action with changes in species richness as well as community composition, and therefore, it offers a chance to test the relative importance of different assembly processes along restoration duration and its consequences on the structure and functioning of plant communities (Pinho et al., 2018). As for the secondary succession of non-shrub-encroached degraded grasslands, both arrivals of propagules from outside and germinations from soil seed bank are very important at the the beginning of restoration, thereby, both random processes (arrival) and environmental filtering (conditions for germination) are expected to dominate the community assembly during the early stage of succession (Meiners et al., 2015; Marteinsdóttir et al., 2018). As successions proceed, competitive exclusion increasingly influences community assembly due to the improvement of abiotic conditions (Chen and Tang, 2016) and interspecific competition (Purschke et al., 2013; Dong et al., 2019). Consequently, the community gradually tends to be relatively stable (Zhang et al., 2019). However, there is still a lack of research on the community assmbly mechanisms in secondary succession of shrub-encroached grassland considering the significant differences between non-shrub-encroached degraded grassland and shrub-encroached grassland. Although some reseachers pointed out that shrubs can facilitate the establishment of surrounding herbaceous communities (Koyama et al., 2015) by reducing abiotic stress (Maestre et al., 2009; Soliveres and Eldridge, 2014), many studies comfirmed that shrub encroachment influenced the cycling of nutrient such as nitrogen, carbon and surface water and energy flow (Peng et al., 2013), causing grassland patches or degradation (Xiong et al., 2004; Zhou et al., 2018; Pierce et al., 2019).
For shrub-encroached grassland, shrub formed different cover at the landscape scale along the degree of shrub encroachment (Eldridge et al., 2011; Peng et al., 2013; Soliveres and Eldridge, 2014). On one hand, the increase of shrub cover across entire landscapes may affect the heterogeneity of soil nutrients and light availability, provide more niche and promote plant differentiation (Breshears, 2006; Soliveres and Eldridge, 2014). On the other hand, microclimatic conditions (Soliveres et al., 2012; Zhou et al., 2018) affected by shrubs could act as environmental filters for plant communities. So far, no research has explored the impact of shrub encroachment on the assembly of plant community in shrub-encroached grasslands although some studies have shown that shrub encorachment has changed the plant composition, community structure and abiotic environment of local habitat (Eldridge et al., 2011; Peng et al., 2013).
In this study, we chose communities treated by different fencing duration and shrub cover in the Caragana mircophyllashrub-encroached grassland in the northern China, and used phylogenetic distances to assess assembly mechanisms along fencing duration and shrub cover gradient. We aimed to answer the following questions: 1) did the increase in fencing duration or shrub cover promote the restoration of herbaceous communities? 2) were key traits of species in the community phylogenetically conserved? and 3) how did fencing duration and shrub cover affect phylogenetic structure of herbaceous community and which process(es) played important roles in affecting community assembly.