Discussion
Our results reveal that changes in the composition of insect herbivore communities track changes in plant species community composition and plant phylogenetic turnover at multiple spatial scales within the Yunnan province. Three lines of evidence suggest that the positive association of plant and insect turnover partly results from insect host specialization. First, from Fig. 2-I we can see that both plant and insect communities exhibit high levels of β-diversity across remarkably short spatial scales (i.e., β1: 40-100 m). This is not a pattern that is expected if compositional co-variation of these groups results from shared bio-geographical histories or parallel responses to climatic gradients, but instead points to host specificity as the underlying process, suggesting this is the scale where biotic interaction plays an important role in the community assemblage through ecological processes (also see Chesson, 2000). Second, we can see that pure plant species composition as a general control on wood boring longhorn beetle community composition accounted for 10% percent of the explained variation (Fig. 4-II). This explanation rate means that after removing the effect of geographic distance and environment, the insects and plants are still positively correlated, indicating that host plant specificity of Cerambycidae is one of the driving factors of the presented pattern. Finally, from Table 2 we can see that beetles standardized Shanoon diversity is significantly positively correlated with standardized plant Chao1 and Shannon diversity.
To further confirm the relationship between plants and insects, it is important to improve the estimation of host specificity and species richness through correction for plant phylogenetic relatedness (Ødegaardet al. , 2005). Our study demonstrates that the phylogenetic diversity of plant communities had a strong positive influence on the species richness of Cerambycidae communities, and the pure effect of this on the Cerambycidae distribution was 6% percent (Fig. 4-I). The strong influence of plant phylogenetic composition on insect community composition suggests that insects tend to either be specialized on plant hosts at a supra-specific level (Novotny et al. , 2002), or that insects are selecting host plants on the basis of phylogenetically conserved traits (Ødegaard, 2006; Whitfield et al. , 2012). There is growing appreciation that long-term co-evolutionary interaction across trophic levels plays a major role in driving the composition and structure of communities (Futuyma & Agrawal 2009; Dinnage et al. , 2012). Evidence of these reciprocal evolutionary pathways (Wheatet al. , 2007; Fordyce 2010; Wagner 2000) has supported the view that phylogenetic information is embedded in the way plant–insect interactions shape community structure and evolution (Agrawal, 2007). Our study provides the first evidence in a natural system that plant phylogenetic β-diversity is highly correlated with Cerambycidae β-diversity, suggesting that the turnover in host-plant lineage structures Cerambycidae assembly in space.
In addition to the plant species composition and phylogeny determining the community assemblage of wood boring longhorn beetles, the environment also played an important role in influencing wood boring longhorn beetle community assemblage. From Fig. 2-I we can see that the relationship between plant and insect β-diversity holds at both regional (i.e., δ1: 250-300 km, δ2: 500 km) and local (i.e., β1: 40-100 m) scales. The local scale co-occurrence pattern is not likely resulted from the environmental gradients cause it varied slightly, however, with the increase of the spatial extent, the matched pattern gradually disappeared but recurrent at macro-scale, which lurking with the mechanism that the effects of biotic interaction is gradually averaged out with the scale aggregation, however, the effect of macroclimatic gradient become more and more obvious, which revealed that the highly associated pattern at coarse-scale might be the result of parallel responses of insects and plants to macroclimatic gradients. Second, from the NMDS analysis (Fig. 3), ELE, MTCM, and AMT together explained 45% of the variation in beetle species composition. All of these suggested that climate gradients impose constraints on Cerambycidae species richness and abundance. Third, the isolated effect of environment to Cerambycidea distribution accounted for 5%, which had a distinct advantage compared with spatial distance and the latter only accounted for 1% as the pure effect (Fig. 4). Finally, the best linear mixed effect model showed that beetles standardized Shanoon diversity is significantly positively correlated with standardized AHR, AMH, MTCM and MTWM. But negatively correlated with standardized ATR (Table 2).
From the above analysis, we can conclude that the dominating mechanisms of insects and tree distribution patterns are related differently at different spatial scales. On the macro/regional scale, the environmental factors, including AMT and MTCM, induced by elevation and latitudinal gradient are the major driving forces on longhorn beetle β-diversity. However, on the micro/local scale, tree diversity and phylogenetic relationship affect the β-diversity much more. Considering the huge possibility of β-diversity scaling relationships varying widely across disparate organisms with different dispersal ability (Soininen et al. , 2007), spatial scales are the most important prerequisites to quantify the relative effects of environmental factors on longhorn beetle species distribution in the present study. As the results showed (Fig. 2-II & III), β-diversity had no significant differences for trees between the tropics and subtropics, and for beetles between the subtropics and the temperate regions at the regional scale. When focusing on local scale comparisons, a completely different pattern emerged. Mechanistically, β-diversity at local and regional scales is typically ascribed to differing processes operating at different spatial scales. These mechanisms can help inform our a priori expectations of how β-diversity might vary among sampling units drawn at each of these scales. Usually, environmental filtering should dominate over dispersal limitation at larger spatial scales, and the local assemblages are mainly influenced by ecological processes such as species interactions, stochastic occupancy, resource specificity, and niche requirements (Whittaker et al ., 2001; Ricklefs, 2004; Hortal et al ., 2010). The climate gradients along the latitude are known to be generally unvarying and result in stabilized patterns, whereas the different ecological processes interact with each other at a local scale leading to the β-diversity of various organism exhibiting divergent distribution patterns. Secondly, the results showed that the dispersal capacity would affect the ability of organisms to colonize suitable environments. Trees, compared with insects, showed that low average rates of dispersal can be expected to show lower average geographic range sizes and higher rates of local endemism, resulting in higher rates of species turnover from local to regional scales (Qian, 2009; Baselga et al. , 2012). Compared with the highest β-diversity of tree organisms that disperse via seeds (Soininen et al. , 2007), short-lived insect assemblages, including wood boring beetles, would not be affected by dispersal limitation to some extent (Novotny et al. , 2007; Hulcr et al. , 2008).
In conclusion, we compared the β-diversity in wood boring longhorn beetles and plant communities from the tropics to the cold temperate regions. The prerequisites of insect and plant interaction for determining the wood boring longhorn beetles existence at any spatial scales are definite and clear. Irrespective of whether from a micro/local or macro/regional spatial scale, β-diversity of longhorn beetles was related to plant richness and phylogenetic diversity to some extent, while at micro/local scale, longhorn beetles exhibits its strongest biotic niche relations (in affecting /or being affected by plant species), but with the increase of spatial resolution, their relationship are expected to be averaged out at landscape scale, and macroclimatic heterogeneity will dominate the community assemblage processes. Besides, the relative lower β-diversity of beetles than that of the plant communities from the tropics to temperate regions was controlled by their stronger dispersal ability.