A long-debated question in ecology is whether the hyper-diversity of tropical plant-feeding insects is a direct consequence of high tropical plant diversity and/or should be attributed to increases in host plant specialization. To address this debate, we used the longhorn beetle as a study system because their larval stages feed on the xylems of trees and lianas. We hypothesized that longhorn beetles show higher host-specificity in tropical forests than in other forests; alternatively, the high longhorn beetle diversity in the tropics may simply be owing to more diverse host plants. We therefore designed an investigation in tropical and subtropical forests to test these hypotheses. We adapted several analyses (i.e., non-metric multidimensional scaling analysis, alpha-diversity, beta-dissimilarity indices comparisons, and variation partitioning based on redundancy analysis) to compare the species diversity of plants and longhorn beetles in different forests. Our results show that both the plant and beetle species in the tropical and subtropical areas were well-stratified (non-metric multidimensional scaling analysis). The beetle alpha-diversity in the tropical forests was significantly higher than that in the subtropical forests, but the plant alpha-diversity in the two types of forests were not significantly different. The beta-dissimilarity comparison showed that the plant species exerted a significant influence on beetle compositional assemblage in the tropical forests, but not in the subtropical forests. Finally, the variation partitioning results showed that both plant species and plant phylogenetic beta-diversity possessed significant explanatory power for beetle assemblage composition in the tropical forests, but not in the subtropical forests. We conclude that wood-boring longhorn beetles show higher host-specificity in tropical forests than in subtropical forests, and the high diversity of wood-boring longhorn beetles in tropical forests might be explained to a large extent by their more finely partitioned diet-breadth.

We used the longhorn beetle and plant to study the difference of alpha-diversity, beta-diversity and host-pecificity of both taxa in tropial and subtropical forest. We hypothesized that longhorn beetles show higher host-specificity in tropical forests than in other forests; alternatively, the high longhorn beetle diversity in the tropics may simply be owing to more diverse host plants. We therefore designed an investigation in tropical and subtropical forests to test these hypotheses. We adapted several analyses (i.e., non-metric multidimensional scaling analysis, alpha-diversity, beta-dissimilarity indices comparisons, and variation partitioning based on redundancy analysis) to compare the species diversity of plants and longhorn beetles in different forests.