Introduction
β-diversity (site-to-site variation in species composition) as a method to quantify the species assembly pattern has provide insights into the processes that create and maintain the compositional variation of natural communities (Whittaker, 1960,Moritz et al ., 2001; Grahamet al ., 2006), β-diversity of biological communities along biological gradients is a long debated focal topic for ecologists and bio-geographers (Routledge, 1977; Harrison et al. , 1992). Differences in β-diversity occurring along gradients are often used to infer variation in the processes of structuring communities. Recent studies suggested the driving forces of β-diversity was caused by a series of combined mechanisms that affect these local community assembly processes interactively with different relative contributions among bio-geographical regions (Mori & Seidl, 2018). Generally, Species relationships in the community can be conceived in terms of a multidimensional coordinate system, the axes of which are the various resource gradients, and those gradients usually can be categorized as focusing on two major components: biotic and abiotic processes (Sextonet al. ,2009).
In nature, the influence of biotic and abiotic factors shaping species community are often varying in time and spatial scale, and disentangling the relative influences of these factors is not always a straightforward process. Biotic interactions, which including interspecific competition, mutualism, and predation/host relationships, may modify either the resource availability or the local abiotic environment with potentially contrasting consequences on abundance, and both positive and negative interactions may affect species assemblage through inhibiting or facilitating species establishment, individual growth and population size (Holt 2009; Svenning et al. 2014). A good deal of examples from the ecological literatures documented that these interactions are not static in space and can be linked with the impacts of changing climate via different complicated ways (Braschler & Hill 2007; Gilman et al , 2010). And a part of these cases were elaborated as follows. Whittaker proposed a conceptual framework in which abiotic factors (temperature and precipitation) explained the distribution of terrestrial biomes of the world (Whittaker 1975). Further more, the idea that climate is the dominant factor shaping species distributions at broad scale is conceived to explain the correlation of climate and species occurrence patterns observed at a comparable spatial resolution (Woodward 1987). Different with the broad spatial scale, Soberón & Nakamura claimed that the pattern of fine spatial resolution is created by biotic interactions (Soberón & Nakamura 2009). This idea is also verified by Pearson & Dawson (2003), who stated that biotic interactions are expected to play a role in shaping species distributions only over local extents. All of these perspectives imply that the key point of the comparative influence of biotic and abiotic process for species assembly rely on scale.
From the literature, approximately 500,000 known species on the earth is herbivorous insects, which represent nearly a quarter of all terrestrial macroscopic biodiversity (Daly, Doyen, & Purcell, 1998; Southwood, 1973;). The intimate association with land plants (especially angiosperms) might be the dominating driving force of herbivorous insects’ extraordinary diversification (Farrell, 1998; Strong, Lawton, & Southwood, 1984; Marvaldi, Sequeira, O’Brien, & Farrell, 2002; Mitter, Farrell, & Wiegmann, 1988). There are a great deal of observations showing that most herbivorous insects only feed on one or few related plant species and showed narrow host range; adaptation towards different host-plant species can potentially generate ecological specialization in plant-feeding insects and, subsequently, species formation. If plant diversity influences insect speciation, we can expect that insect community composition will be strongly correlated with host-plant community composition. This should be observable at the level that insect community composition will co-vary with the host plant community composition.
For arthropod insect communities, substantial compositional change is prevalent between regions along abiotic gradients within regions (Hoisset al. , 2012; Pellissier et al. , 2013) and between habitats differing in plant community composition under the conditions that evolved host specialization (Siemann et al. , 1998; Schafferset al. , 2008). If insect speciation results from allopatric host shifts of specialized insect species, insect community composition variation should be associated with plant community composition variation in a predictable manner (Siemann et al. , 1998; Schaffers et al. , 2008; Pellissier et al. , 2013). However, although the matching patterns could indicate speciation through host shifts or allopatric insect speciation with subsequent host specialization, it could also indicate parallel responses to broad abiotic gradients. Therefore, patterns resulting from parallel responses to macroclimatic gradients are often difficult to distinguish from patterns resulting from evolutionary associations (Hawkins & Porter, 2003). A potential approach to tease these two factors apart is to examine patterns of association between plant and insect β-diversity at different spatial scales (Kemp et al. , 2017). If both patterns result from insect-host specialization, plant and insect β-diversity should be correlated at fine, as well as broad spatial scales. However, if patterns result from parallel responses to broad abiotic gradients, β-diversity patterns should only be correlated at broad spatial scales (Kemp et al. , 2017).
Longhorn beetles belongs to the order Coleoptera of the class Insecta, presenting a period in their life cycle (mainly larval stage) feeds on the xylem of plants or trees. It is estimated that there are more than 26000 species of longhorn beetles in the world, the relationships between longhorn beetles and their host plants are often quite specific, but there is a great range in the breadth of host tree species that might be used by the larvae of different species (Hanks, 1999). Longhorn beetle is not only an important component of ecological biodiversity, but also has vital effects in the sustaining and balancing of forest ecosystem. They are not only the decomposer of organic matter such as wood residues during forest renewal, but also the pollinators of some plant species. Since the diet habits of the longhorn beetle display attributes of host specificity, exploring the β-diversity of insect communities in different climatic zones and habitats will not only be helpful to understand the regional geography and renewal process of forest vegetation, but also contributes to understand the mechanism of maintenance and cycling of natural forest vegetation (Arias et al. 2008; Strong et al.1984; Wagner 2000).
The effect of plant diversity and structural heterogeneity of habitat on longhorn beetle assemblages within a local scale have been heavily studied (Meng et al. , 2013; Gatti et al. , 2018), but a cross regional comparison of β-diversity of longhorn beetle communities and their association with plant species along a broad environmental gradient has rarely been reported. Our specific goal of the study was to determine the contribution of tree species and phylogeny to the long horn beetle β-diversity distribution models along the increasing spatial extent and climatic gradient across Yunnan province in SW China, the Indo-Burma biodiversity hotspots of the eastern Himalayas (Myerset al. , 2000). We hypothesize that: (1) species β-diversity of the longhorn beetle community is comparable and scale dependent to that of plant species across spatial space from the tropical to temperate regions; (2) biotic (insect and host plant interaction) and abiotic (climatic gradients) factors collaboratively shaped longhorn beetle community composition along various spatial scales; (3) the relative importance of abiotic and biotic variation explaining the longhorn beetle community composition vary by spatial scale, and biotic interactions have prominent effect to longhorn beetle community composition at local scale while macroclimatic gradients impose the most control on it at macro-scale.