Soil nematode communities have been traditionally characterized through morphological examination, which is time-consuming and demands highly trained specialist in order to deliver reproducible results. In order to tackle these problems, molecular approaches have been developed for the characterization of soil nematode communities, and yet, their accuracy is controversial. In this study, we compared the characterization of soil nematode communities by the morphological method and by two molecular methods (that differed in primers and high-throughput sequencing platforms). Soil nematodes communities were quantified from four different land-use types (cropland, natural grass-shrubland, orange orchard, and semi-natural pine forest), encompassing 2 soil types (calcareous soil and red soil) with and without anthropogenic disturbance. Our results showed that the characterization of soil nematode community differed substantially between methodologies at both genus and trophic group level. We therefore argue that currently available molecular methodologies may not be accurate for quantitative evaluation of soil nematode communities as a standalone technique, and rather, the needs to be considered as complementary approaches. In order to match the precision with widely used microbial 16S rRNA techniques, we suggest that refinements in both primers sequences and quality alongside quantity of public-available sequences are needed. In addition, a large proportion of nematodes in the natural ecosystems could not be identified by the molecular methods, which indicates the need for enriching nematode sequences in GenBank. With respect to land-use effects, nematode abundances were higher in the highly disturbed cropland and orchard than in the undisturbed grass-shrubland and pine forest, perhaps because external inputs (inorganic or organic fertilizers) were added to the disturbed but not to the undisturbed systems. In each soil type, the absence of disturbance caused a shift in the soil decomposition pathway from bacterial-mediated to fungal-mediated. The cropland had higher nematode generic diversity and lower maturity and faunal structure than the grass-shrubland in the calcareous soil. The orchard had lower nematode diversity and higher maturity and faunal structure than the pine forest in the red soil. Altogether, these differences between disturbed and undisturbed ecosystems for each soil type might be due to different responses of low and high trophic groups to the input of chemical and organic fertilizer to the cropland and orchard, respectively.

传统上通过形态学检查来表征土壤线虫群落，这很费时，并且需要训练有素的专家才能提供可重复的结果。为了解决这些问题，已经开发了用于表征土壤线虫群落的分子方法，但是其准确性尚存争议。在这项研究中，我们通过形态学方法和两种分子方法（在引物和高通量测序平台上有所不同）比较了土壤线虫群落的特征。从四个不同的土地利用类型（农田，天然草丛，橙色果园和半天然松树林）对土壤线虫群落进行了定量，包括有和没有人为干扰的两种土壤类型（钙质土壤和红色土壤）。我们的结果表明，土壤线虫群落的特征在属和营养族水平上的方法学之间存在很大差异。因此，我们认为，作为一种独立的技术，当前可用的分子方法对于土壤线虫群落的定量评估可能并不准确，而是需要将其视为补充方法。为了使精度与广泛使用的微生物16S rRNA技术相匹配，我们建议在引物序列和质量以及可公开获得的序列数量上都需要改进。此外，自然生态系统中的大部分线虫无法通过分子方法鉴定，这表明需要在GenBank中丰富线虫序列。关于土地利用的影响，在高度受干扰的农田和果园中，线虫的丰度要比未受干扰的草丛和松树林中的线虫丰度高，这也许是因为外部输入（无机或有机肥料）已添加到受干扰的系统中，而不是未受干扰的系统。在每种土壤类型中，没有干扰会导致土壤分解途径从细菌介导转变为真菌介导。与石灰性土壤中的草灌木地相比，该农田具有较高的线虫通用多样性，且成熟度和动物区系较低。与红色土壤中的松树林相比，果园的线虫多样性更低，成熟度和动物区系更高。共，