There is an urgent need to develop fast and reliable methods to assess forest biodiversity. The diversity of tree-related microhabitats such as tree holes is an important measure of forest structural heterogeneity and level of conservation. Tree holes harbour diverse and specialized arthropod communities of conservation value. To survey invertebrates in tree holes the wood mould is extracted from the cavity, placed in berlesse funnels to collect any living organisms and the substrate is subsequently sieved to recover any body remains of arthropods. The identification of those body remains is challenging and often inaccurate. In the present study, we apply metabarcoding techniques on arthropod body remains sieved from wood mould excavated from tree holes to characterize arthropod communities. Our samples had low DNA concentrations and high contents of PCR inhibitors derived from cuticular pigments and traces of humic acid. We developed a protocol to bypass these methodological problems and successfully amplified and sequenced 130bp amplicons.