Comparing soil microarthropod communities derived directly from soil DNA metabarcoding with those from morphological assessment in a drought-prone and irrigated pine forest

Springtails (Collembola) and mites (Acari) are soil microarthropods, one of the most diverse animal groups in soils. They play a crucial role in organic matter cycling and are active throughout the food web as decomposers, bacterivores, fungivores, and carnivores. Only little is known about how these groups might respond to shifts in water availability, for example in the context of global change. Here, we investigated how soil microarthropods responded to long-term irrigation in a drought-prone Scots pine (Pinus sylvestris) forest in southern Switzerland. After more than a decade of doubling the annual rainfall, irrigation improved not only tree vitality but also soil quality, and with shifts in bacteria and fungi reflecting changes from oligotrophic to copiotrophic conditions. We assessed soil microarthropods with two approaches: (1) directly by soil DNA metabarcoding and (2) by morphological assessment after extraction of the animals with Macfadyen funnels. Another main aim of that study was to compare the results with the two approaches. The dominant Collembola genus in both assessment approaches was Parisotoma. The dominant Sarcoptiformes genus was Oppiella whereas Geolaelaps was the dominant Mesostigmata genus in both assessment approaches. Only the metabarcoding approach detected Trombidiformes genera, and only one genus, Microtydeus, had a classification confidence >80 %. The abundance and alpha-diversity of Collembola and Acari did not change significantly as a result of the irrigation treatment, regardless of the assessment approaches applied. In contrast, microarthropod beta-diversity showed significantly shifts for Collembola and Acari, and for the Collembola order Entomobryomorpha and the Acari orders Sarcoptiformes, Mesostigmata, and Trombidiformes. A Procrustes analysis comparing the two assessment approaches indicated a significant effect of the irrigation treatment for the mite order Sarcoptiformes and a nearly significant effect for Collembola. Using indicator species analysis a Parisotoma species was the only Collembola taxon in the metabarcoding assessment that was strongly associated with the irrigation treatment. With the morphological assessment, Parisotoma notabilis and Lepidocyrtus sp. were significantly associated with irrigation. For Acari, only the morphological approach let to Licnodamaeus pulcherrimus as a negative indicator taxon for irrigation. By using the morphospecies lists as a reference for validation and comparing it with the species list obtained through metabarcoding, we found that only a small percentage of Collembola and Acari morphospecies overlapped. The metabarcoding approach detected taxa that were not observed with the morphological assessment, such as Neelipleona, Symphypleona, or Trombidiformes. Due to the complexity of the taxa and the lack of comprehensive taxonomic identification and reference databases, identification at the species level is hardly possible. Further efforts to enrich the microarthropod reference database are urgently needed.