1. Introduction
Community assembly mechanisms have always been a topic of ecological research, and natural communities are generally believed to be structured by a set of processes (Chase et al., 2014; Levine et al., 2017; Wang et al, 2021). Niche theory-based determinative processes , including the influence of the abiotic environment on fitness (Wang et al, 2021; e.g. habitat filtering) and biotic interactions, in particular interspecific competition (Leibold, 1998), and neutral theory-based stochastic processes , including spatial dispersal limitation, demographic stochasticity, and ecological drift (Hubbell, 2005; Zhou & Zhang, 2008; Chase & Myers, 2011) have been regarded as two primary ecological mechanisms driving community assembly (Li et al., 2019). The relative importance of these processes tends to vary among ecosystems (Liu et al., 2013; Jiang et al., 2018), spatial scales (Zhang et al., 2021), community succession (Csecserits et al., 2021) and even in different environments, especially extreme environments (Wang et al., 2021). For example, deterministic processes may play a greater role than stochastic processes in adverse environments (Chase & Myers, 2011). Interspecific interactions and density-dependent mechanisms should be strongest at the neighborhood scale where individual organisms interact, and environmental filtering should be stronger than interspecific interactions at the habitat scale (Cavender-Bares et al., 2009; Purschke et al., 2017).
Plant functional traits are usually used as proxies to determine whether different tree species have different ecological strategies for resource capture and reproduction (McGill et al., 2006; Baraloto et al., 2012; Adler et al., 2013; Liu et al., 2020). Analyses of the distribution of trait values within communities yield insights of the ecological processes constraining their assembly (Kraft et al., 2008; Paine et al., 2011). If the niches of two species overlap, it is generally expected that the two species are similar in a range of functional traits, and vice versa (Westoby & Wright, 2006). Based on competition theory, higher similarity in functional traits for a community could lead to an increased intensity of interactions among neighboring individuals (Uriarte et al., 2004; Paine et al., 2011; Funk et al., 2016). Consequently, communities with scattered trait values are primarily shaped by niche differentiation, whereas environmental filtering is the dominant process shaping ecological communities when trait value range is narrower than predicted (Paine et al., 2011). Thus, based on the fact that functional traits could represent the key aspects of physiology, investigating the variation in functional traits at the species level (i.e., intraspecific and interspecific variation) and at the community level could be beneficial for a deeper understanding of how physiological processes shape the assembly of ecological communities. Plant leaves are critical organs for the exchange of matter and energy with the photosynthetic organs of plants, and several biological processes such as plant growth, survival, reproduction, and ecosystem function are fully influenced by leaf parameters (e.g., leaf area, length, and dry mass) (Surya et al., 2020). Leaf functional traits are sensitive to changes in environmental factors. They can adjust resource utilization strategies to adapt to different habitats through trade-offs of various traits, which can reflect the driving mechanisms of the environment on community assembly (Wright et al., 2004; Tian et al., 2016; Zhang et al., 2020).
Although functional traits could provide a way to infer prevailing ecological process information based on morphological, physiological, and ecological characteristics, plant functional traits are not only affected by environmental factors but also by species evolution history (Swenson, 2013). Species coexisting in the same habitat might be relatives sharing common functional traits influenced by evolutionarily conserved or perhaps distant relatives adopting convergent traits to adapt to the habitat (Cavender-Bares et al., 2009). Thus, phylogeny and functional traits do not necessarily present similar information and patterns (Cadotte et al., 2019), and testing the phylogenetic signals of functional traits is a necessary key step to more accurately infer the mechanism of community assembly (Baraloto et al., 2012; Cheng et al., 2019). Phylogeny is an indirect estimation of ecological similarity based on species affinity and an estimation of the impact of historical factors on the existing community (Swenson et al., 2013; Jiang et al., 2018). Therefore, the combined analysis of phylogeny and functional traits can not only reveal the impact of community evolutionary history and functional traits simultaneously on the current community ecological process (Webb et al., 2002; Zhou et al., 2021), but also contribute to revealing the ecological processes responsible for evolution and functional assembly (Zhou et al., 2021). In other words, combined trait-based and phylogenetic-based approaches is a powerful way to detect community assembly processes (Kraft & Ackerly, 2010; Gianuca et al., 2017; Li et al., 2019).
Subtropical region of China holds the largest evergreen broadleaved forest in the world and harbors abundant seed plants and endemic species (Xu et al., 2017), which play an important role in biodiversity protection and carbon balance. However, due to the prolonged and frequent anthropogenic interferences, vegetation degradation is severe, and ecological problems are prominent in this area. Research based on the typical community structure and ecological processes in zonal vegetation has become an important means of vegetation restoration and reconstruction (Zhao et al., 2015; Ouyang et al., 2016; Zhang et al., 2020). As one of the typical types of evergreen broad-leaved forests in the subtropical China, Lithocarpus glaber–Cyclobalanopsis glaucaforests exhibit high species diversity, stable community structure, and high ecosystem function and service values (Zhao et al., 2015). To study the assembly processes and mechanisms, we carried out a series of studies focusing on species composition, spatial patterns, and effects of topographic and soil factors on woody species assembly in this long-term monitoring community. We found that aggregation was the major spatial pattern and environment (topographic and soil factors) explained 28.10% of the species assembly (Zhao et al., 2015), and it is apparent that species-based approaches to understand community assembly are limited. Here, we re-examined this issue by integrating functional and phylogeny-based approaches to explore the community assembly processes. The degree to which patterns of functional traits and phylogenetic dispersion may be easily explained by the abiotic environment and spatial relationship has been also discussed. We propose that a variance partitioning approach can be applied to simultaneously address these challenges (Durate et al., 2013; Zhang et al., 2021). The degree to which the abiotic environment, dispersal limitation, and/or their joint effects affect trait dispersion should be determined by partitioning variance in trait dispersion into pure environmental, spatial and joint effects. We also measured 9 leaf functional traits and phylogenetic data of 18 dominant woody species with important value >1.00% and accurate topographic and edaphic data sets to address the following: (i) To clarify the variation and trade-off relationship of leaf functional traits; (ii) To explore the effects of phylogeny and environment on trait variation; (iii) To disentangle the relative importance of niche and neutral processes shaping community assembly at a fine scale.