Exploring the Leaf Beetle Fauna (Coleoptera: Chrysomelidae) of an Ecuadorian Mountain Forest Using DNA Barcoding

Background
Tropical mountain forests are hotspots of biodiversity hosting a huge but little known diversity of insects that is endangered by habitat destruction and climate change. Therefore, rapid assessment approaches of insect diversity are urgently needed to complement slower traditional taxonomic approaches. We empirically compare different DNA-based species delimitation approaches for a rapid biodiversity assessment of hyperdiverse leaf beetle assemblages along an elevational gradient in southern Ecuador and explore their effect on species richness estimates.
Methodology/Principal Findings
Based on a COI barcode data set of 674 leaf beetle specimens (Coleoptera: Chrysomelidae) of 266 morphospecies from three sample sites in the Podocarpus National Park, we employed statistical parsimony analysis, distance-based clustering, GMYC- and PTP-modelling to delimit species-like units and compared them to morphology-based (parataxonomic) species identifications. The four different approaches for DNA-based species delimitation revealed highly similar numbers of molecular operational taxonomic units (MOTUs) (n = 284–289). Estimated total species richness was considerably higher than the sampled amount, 414 for morphospecies (Chao2) and 469–481 for the different MOTU types. Assemblages at different elevational levels (1000 vs. 2000 m) had similar species numbers but a very distinct species composition for all delimitation methods. Most species were found only at one elevation while this turnover pattern was even more pronounced for DNA-based delimitation.
Conclusions/Significance
Given the high congruence of DNA-based delimitation results, probably due to the sampling structure, our study suggests that when applied to species communities on a regionally limited level with high amount of rare species (i.e. ~50% singletons), the choice of species delimitation method can be of minor relevance for assessing species numbers and turnover in tropical insect communities. Therefore, DNA-based species delimitation is confirmed as a valuable tool for evaluating biodiversity of hyperdiverse insect communities, especially when exact taxonomic identifications are missing.

研究背景
热带山地森林是生物多样性热点区域，孕育着规模庞大却鲜为人知的昆虫类群，这些类群正遭受栖息地破坏与气候变化的双重威胁。传统分类学方法的评估速率较为缓慢，因此亟需快速昆虫多样性评估手段以作为补充。本研究针对厄瓜多尔南部沿海拔梯度分布的高度多样叶甲类群组合开展快速生物多样性评估，实证比较多种基于DNA序列的物种界定（DNA-based species delimitation）方法，并探讨这些方法对物种丰富度估算结果的影响。

材料与方法及主要研究结果
本研究基于采自罗汉松国家公园3个采样点的674头叶甲标本（鞘翅目:叶甲科（Coleoptera: Chrysomelidae））的COI条形码（COI barcode）数据集，这些标本隶属于266个形态种（morphospecies）。研究采用统计简约分析（statistical parsimony analysis）、基于距离的聚类（distance-based clustering）、GMYC模型与PTP模型（GMYC and PTP modelling）来界定类物种单元（species-like units），并将其与基于形态学的副分类学（parataxonomic）物种鉴定结果进行对比。四种基于DNA序列的物种界定方法得到的分子操作分类单元（MOTUs）数量极为接近（n=284~289）。估算得到的总物种丰富度远高于实际采样获得的物种数：形态种的Chao2估计量为414，不同MOTU类型的估计值为469~481。不同海拔梯度（1000米与2000米）的类群组合拥有相近的物种数量，但所有物种界定方法均显示二者的物种组成差异显著。多数物种仅在单一海拔位点被发现，而基于DNA序列的物种界定方法更凸显了这一物种更替模式。

结论与意义
鉴于基于DNA序列的物种界定结果具有高度一致性，这一现象大概率源于采样设计结构。本研究表明，当研究对象为区域尺度有限、稀有物种占比高（即约50%为单例序列（singletons））的物种群落时，物种界定方法的选择对热带昆虫群落的物种数量与更替模式评估影响较小。因此，本研究证实基于DNA序列的物种界定方法是评估高度多样昆虫群落生物多样性的有效工具，尤其适用于缺乏精准分类鉴定信息的研究场景。