Paving the way for improved insect metabarcoding

Paving the way for improved insect metabarcoding | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Paving the way for improved insect metabarcoding Roland Mühlethaler, Arne W. Lehmann, Sebastian Köthe, Helge Bruelheide, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7899842/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Metabarcoding is becoming an increasing popular method for broadscale insect monitoring. However, to complement or replace traditional insect monitoring approaches, the reliability of metabarcoding has to be confirmed. Therefore, we have evaluated the accuracy of species identifications of metabarcoding against the standard of morphology with binary classification in a confusion matrix to measure model performance. Within 12 German nature protected areas using a transect of five traps, metabarcoding found a total of 15,107 OTUs. Identifying 3096 individuals of three selected taxa by morphology, we found in total 151 species, compared with 130 species by metabarcoding. Species numbers for single traps differed substantially between methods, with a significant correspondence only found for Syrphidae (Diptera). Congruence at the species level was low, with sensitivity and precision below 50%, and even lower for Red List taxa. The match increased when aggregating traps and sites across Germany, or when comparing genera instead of species. Abundance curves strongly differed for species identified by both methods. So far, metabarcoding is lacking the necessary accuracy at the species level. We discuss possible causes for these inconsistencies and make suggestions for improvement for metabarcoding identification. Biological sciences/Ecology Earth and environmental sciences/Ecology Biological sciences/Evolution Biological sciences/Zoology biodiversity loss endangered species insect decline nature conservation Malaise traps species identification Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Introduction Global biodiversity is declining at an alarming rate (IPBES 2019), and terrestrial insects especially are affected (van Klink et al. 2020 ). This group has enormous diversity being the most species-rich taxon of Metazoa and plays important roles in almost all terrestrial ecosystems. The accurate identification of insect species is essential for many scientific fields including ecology, evolutionary biology, agriculture, and medical entomology (Eggleton 2020 ). The loss of insects has become the focus of many research projects, with a landmark study from Germany, where insect biomass declined by more than 75% over the past 27 years (Hallmann et al., 2017 ). In follow-up studies with identical methodology we found this trend to have continued in the last years, without any indication of being reverted (Mühlethaler et al. 2024 , Hallmann et al. 2025 ). To monitor insects, standardized methods, such as Malaise traps, are in widespread use, their broad sampling coverage of mainly flying individuals allows to document spatial and temporal changes (Skvarla et al. 2021 ). One easily accessible metric from malaise traps is insect biomass, which is correlated with species number, both within single taxonomic groups (Hallmann et al. 2021 ) and with overall insect richness (Köthe & Schneider et al. 2023). However, using biomass alone has limitations as it provides no specific answers which and where species are declining. Thus, reliable tools for species identification are needed, in particular to detect rare and/or Red List species. The whole field of nature protection legislation and conservation planning is to a high degree dependent on species records. Important examples at the global level are the Convention on International Trade in Endangered Species (CITES) and the IUCN Red Lists for endangered species, at the EU level the Birds and Habitats Directives and at the national level in Germany the Federal Conservation Act (Wirth et al. 2024 ). The traditional approach to morphology-based taxonomy is challenging, firstly because of the labor and time-consuming sorting of thousands of insects. For example, classifying 73 traps sampling for three years needed 15 years and 300 technicians in the Swedish Malaise Trap Project (Karlsson et al. 2020 ). Secondly, species determination skills are generally restricted to well-known taxa, while for the majority of species-rich taxon groups and for taxa with small body sizes no experts are available, and experienced taxonomists are a diminishing group at the risk of extinction (Hochkirch et al. 2022 ). Fourthly, cryptic species or dark taxa, unresolved taxonomies and the degradation of morphological features in wet samples are a challenge for correct identification. DNA-based molecular methods are a promising tool to address these issues, as DNA barcoding is easy to implement and allows rapid taxonomic identification (Hebert et al. 2003 ). Metabarcoding, where DNA is isolated and analyzed from a mixture of specimens and mass-amplified using high-throughput technologies, is increasingly used for species identification in bulk samples (Chua et al. 2023 ). Metabarcoding sequences are filtered to generate Amplicon Sequence Variants (ASV), which are subsequently clustered into Operational Taxonomic Units (OTU). By bioinformatical analysis and by comparisons with reference databases we may thus identify the taxa present in complex samples (Iwaszkiewicz-Eggebrecht et al. 2024 ). Metabarcoding has been tested in the German Malaise trapping programs (Geiger et al. 2016 ) and is employed to study insect diversity in large scale monitoring programs, namely the LTER (Buchner et al. 2025 , Remmel et al. 2024 ) and the DINA project (Köthe & Schneider et al. 2023, Thomas et al. 2025 ). Despite the huge benefits for mass assessment, metabarcoding has several limitations and not all problems for the integrity and correctness of the database entries and the fine-tuning of laboratory and sequencing methods have been solved yet. Two of the key problems for taxon identification with metabarcoding are the failure to detect taxa that are present in a sample (i.e. false negative records) and incorrect detection of taxa that are not present in the sample (i.e. false positive records). False negative records might arise from problems with non-perfect primer fitting, binding to non-target sequences, sequencing errors and random distribution of specimens in subsamples. Due to differences in body size, genome size, mitochondrial copy number, DNA extraction efficiency and PCR amplification efficiency PCR amplification rates differ among taxa (Luo et al., 2023 ). As the amount of DNA extracted differs from small versus large specimens, techniques like size sorting and hence increasing the representation of smaller species in the DNA extract might reduce part of the problem (Elbrecht et al. 2021 ). Nonetheless, the bias continues to exist for less abundant taxa and those with smaller body size and can lead to false negative records, altering species presence/absence assessments (Strutzenberger et al. 2024 ). Similarly, false positive records might result from wrong reference sequences in the databases or the picking of the most similar species by indiscriminately using bioinformatics tools like the Basic Local Alignment Search Tool (BLAST). As a consequence, metabarcoding lists might infer similar patterns of the overall community (Petsopoulos et al. 2024 , Buchner et al. 2025 ) but include inaccurate or incorrect taxonomic assignments, especially at the species level (Förster et al. 2023 , Li et al. 2024 , Remmel et al. 2024 , Strutzenberger et al. 2024 ). The DINA project (Diversity of Insects in Nature Protected Areas) which aims to investigate insect biodiversity in Germany (Lehmann et al. 2021 ) is one of the most prestigious follow-up projects of the landmark study on insect decline by Hallmann et al. ( 2017 ). For this purpose, Malaise traps were operated along transects in 21 representative nature protected areas across Germany (Lehmann et al. 2021 ), with a total of 105 traps operated simultaneously and assessed for biomass and species identification (Köthe & Schneider et al. 2023, Mühlethaler et al. 2024 , Hallmann et al. 2025 ). Within our project, we compared DNA metabarcoding for whole community species assessments (Thomas et al. 2025 ) with the gold standard of morphological sorting and identification by experts. Metabarcoding is still more effective with destructive homogenization than with lysis protocols (Wolany et al. 2024 ) but prevents any morphological inspection after metabarcoding. To overcome these restrictions, the Entomological Society Krefeld developed an insect sample fractionizer (Hörren et al. 2022 ) to enable subdividing of the samples. In our workflow, the biomassed samples are divided with the first half subjected to metabarcoding, while the second half is used for morphological analysis and subsequently stored. As it is impossible to cover all roughly 33,000 to 38,000 insect species in Germany from the estimated 20 million insects sampled by the DINA project (Lehmann et al., data on file), we concentrated on three insect groups: Diptera: Syrphidae, Hemiptera: Auchenorrhyncha and Orthoptera. Syrphidae provide important ecosystem services such as plant pollinators and larval predators. They are one of the main targets for Malaise trap sampling as adults tend to be good fliers (Hallmann et al. 2021 ). Around 470 species occur in Germany (Ssymank et al. 2011 ), which have very distinct habitat requirements, especially at the larval stage. Of the approximately 650 leafhopper and plant hopper species (Auchenorrhyncha, Hemiptera: Cicadomorpha & Fulgoromorpha) known from Germany (Nickel et al. 2016 ), around 50 percent occur in open grasslands (Mühlethaler et al. 2019 ). In addition, they can build up very high population densities of up to over 5,000 animals/m 2 . Many species have short wings and are therefore less frequently caught in Malaise traps but contribute a considerable fraction of the sampled biomass. Orthoptera (grasshoppers and crickets) are a small group in Germany of 83 established species (Poniatowski et al. 2024 ), particularly in grasslands, and have been intensively studied as indicators. In line with their behavior, they might be rare in Malaise trap catches but are relatively easy to determine (Horstkotte et al. 1991 ). Due to their large bodies, their exceptionally large genomes (Hawlitschek et al. 2023 ) and large numbers of fragments of mitochondrial DNA (mtDNA) that have been incorporated into the nuclear genome (NUMTs, Liu et al. 2024 ), they are often overrepresented in metabarcoding results (Remmel et al. 2024 , Thomas et al. 2025 ). We studied how precise metabarcoding reflects species numbers and species identifications. Therefore, we analyzed the performance in a binary classification framework. We constructed 2x2 confusion matrices at different spatial and taxonomic resolution to compare the true positive identifications (found both by morphology and metabarcoding) with the other three classes of false positives, false negatives and true negatives (found neither in the morphological nor the metabarcoding subsample) (Grandini et al. 2020 ). We expected a significant match between both methods, which can be expressed as sensitivity (i.e. the fraction of true positives in all morphologically identified taxa) or precision (i.e. the fraction of true positives in all taxa identified by metabarcoding). Moreover, we hypothesized that the match was improved when samples were aggregated across traps and sites, and comparisons were made at the genus rather than at the species level. Results In total, 15,107 OTUs from insects were detected by metabarcoding in the 12 nature protected areas with five traps along a transect in the given sampling period Including the taxa identified at the family, genus and species level, the number of OTUs for the three selected groups ranged from 110 for Syrphidae to 155 for Auchenorrhyncha to 1331 for Orthoptera. When focusing only on taxa at the species level, the numbers dropped to 52, 62 and 16 species, respectively (Table 1 , Suppl. Table 1). Table 1 Number of identified OTUs based on DNA metabarcoding from 12 nature protected areas, for all insects and for Syrphidae, Auchenorrhyncha and Orthoptera. In addition, the table shows the number of morphologically identified species and the total number of known species in Germany is given. RL = German Red List. Taxon OTUs / n= OTUs assigned to family level / n= OTUs assigned to genus level / n= OTUs assigned to species level / n= Species identified by barcoding / n= Families identified by morphology / n= Genera identified by morphology / n= Species identified by morphology / n= Species in Germany (*after RL) / n= All insects 15107 14493 11785 9063 2708 N/A N/A N/A 33466 Syrphidae 110 110 105 69 52 1 32 52 467 Auchenorrhyncha 155 155 139 98 62 7 55 91 645 Orthoptera 1331 1204 454 269 16 2 7 8 83 The 3096 individuals identified by morphology were assigned to 151 species, among them 52 species of Syrphidae (879 individuals), 91 species of Auchenorrhyncha (2175 individuals) and 8 species of Orthoptera (42 individuals) (Table 1 ). The total number across all sites was slightly higher but comparable to the 130 species identified by metabarcoding. However, when analyzing the patterns by trap within sites, the correspondence between both methods decreased considerably. Species richness determined by metabarcoding deviated strongly from species richness determined by morphology both, across all taxon groups and separately for Syrphidae, Auchenorrhyncha and Orthoptera (Fig. 1 , Suppl. Table 2a). In the mixed effects models that accounted for differences between sites, the slope in richness was significantly different from zero only for hoverflies, with 0.26 species found with metabarcoding for every species identified morphologically (Fig. 1 b, Suppl. Table 2b). Validating species lists from metabarcoding against morphological identification was not very successful. When aggregated by sites, the percentage of species identified morphologically compared those by metabarcoding was lower than 50% in all cases (Fig. 2 ). Precision, also known as positive predictive value of the matches divided by the number of species reported by metabarcoding, was highest for the Syrphidae with 30%, followed by Auchenorrhyncha with 27%, and lowest for the Orthoptera with 22% (Suppl. Table 4a-c). For Auchenorrhyncha there were two and for Orthoptera seven out of 12 localities without any species detection overlap between both methods. The general accuracy of the metabarcoding was high, with above 90% at the species level and when focusing on single traps (Suppl. Table 5: 1.1.-1.3). In turn, the error rate was reasonably low, which however, was largely driven by the high proportion of true negatives (TN), i.e. species not being encountered by either method. In turn, the proportion of true positives (TP) was rather low, both when expressed in relation to the false negatives (FN) as sensitivity (i.e. TP/(TP + FN)) or in relation to the false positives (FP) as precision (i.e. positive predicted value, TP/(TP + FP)). Values for both sensitivity and precision were below 50% for all three taxonomic groups at the species level (Suppl. Table 5: 1.1.-1.3, Suppl. Figure 1). Similarly low was the F1 score as harmonic mean between sensitivity and precision at the species level and focus on single traps, being 0.25 or lower (Fig. 3 ). With broadening the focus, i.e. combining all traps per site or all sites of the whole study, increased the F1 score, but never reaching 0.5. The same increase was observed, albeit at higher scores, when the analysis was done at the genus level, but also below 0.5 when focusing on single traps. Morphological examination enabled counting individuals per species. Likewise, metabarcoding resulted in read numbers which might to some degree related to the number of individuals. Concentrating on true positive species detected by both methods we compared the mean number of individuals (Fig. 4 a,b) with the mean read counts (Fig. 4 c,d). Ranking the species by decreasing abundance, the resulting abundance curves show little overlap, neither for Syrphidae (Fig. 4 a,c) nor Auchenorrhyncha (Fig. 4 b,d). For nature conservation issues, species listed on the Red list of endangered species are of special interest. In our sample, we detected 42 insect species endangered in Germany (Suppl. Table 6), either identified only by metabarcoding (21) by morphology (14) and by both methods (7). Accuracy in terms of sensitivity and precision was below 40% for Red list species for Syrphidae and Orthoptera (Fig. 6), except for nearly equal sensitivity in Auchenorrhyncha. Discussion Classical species identification based on morphological characters is highly accurate when carried out by experienced entomologists. Additional information such as the distribution of sex and larvae or wing morphs as indicator of migratory activity can add to the analysis of the insect populations. As a plus, the preserved specimen can be stored for future study after determination. Thus, we expect that morphological identification will remain the gold standard for the time being. However, the large amount of time required for specific identification represents a major barrier, combined with few or missing taxonomic experts and high-quality identification guides. Due to these problems, no one worldwide has yet succeeded in determining a complete annual sample series from a Malaise trap. Even analyzing only small parts of the total diversity is very time-consuming and therefore cost intensive. In contrast, identification based on metabarcoding is generally faster, cheaper and able to cover whole communities simultaneously (Chua et al. 2023 , Hawthorne et al. 2025 ). Even more, taxa difficult to identify morphologically and species with little morphological material in the samples (e.g. only individual body parts, endoparasites, intestinal contents) might be detected. Here, we made the comparison with the result of an unexpected poor or even lacking match between both methods. Thus, we have to reject our main hypothesis. The overlap in true positives, i.e. correctly identified species between both methods (true positive determinations = TP) was only around 30%. The problem of inaccurate or incorrect taxonomic assignments of metabarcoding has already been recognized for insects (Förster et al. 2023 , Li et al. 2024 , Remmel et al. 2024 , Strutzenberger et al. 2024 , Penel et al. 2025 ) and freshwater macroinvertebrates (Jones et al. 2025 ). However, compared to thousands of barcoding and metabarcoding studies published each year (Salis et al. 2024 ), the number of those evaluating metabarcoding data with morphology species identification is extremely small. As Iwaszkiewicz-Eggebrecht et al. ( 2024 ) stated, the metabarcoding scene concentrate largely on laboratory-specific proof of concepts instead of external validation as calibration standards. So far, the focus has more been on standardization of molecular tools (Chua et al. 2023 , Hawthorne et al. 2025 ) or a unified reporting of lab procedures (Iwaszkiewicz-Eggebrecht et al. 2024 ). Here we will discuss mainly two aspects: (a) the main reasons for the poor model fit between metabarcoding and morphology, (b) the next steps to improve it. Deviations in metabarcoding species detection from morphological determinations was nearly equal in both directions; the failure to detect species present in the samples (false negatives = FN) and the reporting of species not present (false positives = FP). False negative as the failure to detect species actually present in the sample might arise from problems with laboratory routines (Iwaszkiewicz-Eggebrecht et al. 2024 ), bioinformatic (Penel et al. 2025 ) or unequal separation of subsamples. In our study, only one methodological approach of DNA metabarcoding was applied for all preparatory steps in the laboratory, like homogenization, amount of tissue used, PCR steps, selection of primers and the sequencing depth. It can be assumed that this comparison would yield different results if the laboratory practice were carried out differently, e.g. using a higher tissue quantity or multiple use of small tissue quantities, a higher sequencing depth and different bioinformatics pipeline. It is well known that DNA extraction is more successful in some species than others (Elbrecht et al. 2021 , Zizka et al. 2022 ). Primer fitting especially for species-rich taxon groups is a difficult task as non-perfect primers might bind to non-target sequences and produce sequencing errors. A special challenge is related to the variance in body size between species. As the amount of DNA extracted differs from small versus large specimens, mitochondrial copy number, DNA extraction and PCR amplification rates differ among taxa (Luo et al., 2023 ). In common agreement, we used size sorting (Elbrecht et al. 2021 ), and hence, increased the representation of smaller species in our DNA extracts (Zizka et al. 2022 ), which might have reduced the problem. Nonetheless, the bias would remain for less abundant taxa and those with smaller body size and can lead to false negative records, altering species presence/absence assessments (Strutzenberger et al. 2024 ). This is also reflected in our analysis of Red List species, which show an even weaker model fit (Fig. 5 ), likely because of their rarity. Furthermore, metabarcoding can suffer from a level of heterogeneity in DNA degradation between samples. Small variations in DNA quality, associated with the stochastic process of PCR amplification of mixed samples, were the largest error source and accounted for 60% of false negative metabarcoding identifications in beetles (Penel et al. 2025 ). With our bioinformatics pipeline 60% of the OTUs could be assigned to the species level (Table 1 ). This is far from perfect, and the generation of more complete DNA barcode databases should be a major goal (Chua et al. 2023 ). Indeed, the sensitivity was nearly doubled for Syrphidae, Auchenorrhyncha and Orthoptera (Suppl. Figure 1, Suppl. Table 5) when shifting our analysis from the species to the genus level, supporting our second hypothesis. Even while barcoding of German insects has quite progressed, a lack of reference sequences in the database against which the blasting had been done might be one reason for the many false negatives. Another coherent issue is the reliance on the assumption of a barcoding gap, i.e. that an universal similarity threshold can be used for species assignment (Hebert et al. 2003 ). Genetic distances vary considerable intra- and interspecifically and do even overlap, which results in errors when using a fixed threshold for molecular species detection (Creedy et al. 2022 ). Flexible filtering approaches based on multiple thresholds (Arribas et al. 2021 , Penel et al. 2025 ) in combination with broader phylogenetic coverage will probably increase the proportion of true positives in the future. The order Orthoptera is especially challenging for metabarcoding as they have enormous genome sizes (Hawlitschek et al. 2023 ) and large numbers of NUMTs (fragments of mitochondrial DNA in the nuclear genome, Liu et al. 2024 ). Despite proper filtering of the pseudogenes the OTUs of Orthoptera are regularly overreported (Remmel et al. 2024 , Buchner et al. 2025 ) even when they have not been in the target group of insects (Kortmann et al. 2025 ). Our metabarcoding analysis suffered from the same problem with OTUs for Orthoptera, which accounted for 9% of all OTUs in our samples, while the 83 species of Orthoptera reported for Germany comprise only about 0.2%cent of the total of German insect species (Table 1 ). A further problem with Central European Orthoptera is the relaxed species resolution with barcoding: up to five species share a barcode and this sharing extends even cross genera (Hawlitschek et al. 2017 ). For the time being, it might be wise to exclude Orthoptera from metabarcoding analyses either by upfront filtering or even better by manually removing the easy to detect and large individuals from the samples prior to metabarcoding. To allow for morphological inspection and long-term storage of controls, we fractionized our samples with a splitter into two parts (Hörren et al. 2022 ). Even if we assume an ideal randomization, species that are only contained in few individuals might be present in the first and missing in the second subsample. Such an uneven distribution might contribute to the mismatch when comparing identifications based on metabarcoding from the first half with morphology from the second half. This explanation might apply to taxa with only a very few individuals in the sample. However, we found the mismatch also to be pertinent for frequent species. If the poor match of metabarcoding had been caused by dividing the samples, we would expect a major increase in overlap when shifting the focus from single traps to the combined analysis of five traps at a site. As hypothesized, pooling the samples across sites increased the match, with the sensitivity (true positives divided by TP plus false negatives) increasing from 31, 19 and 17% to 39, 28 and 28% for the three taxa, respectively, and further to 44, 36 and 30%, when summing all 60 traps form all sites across Germany. However, even pooling did not remove the large proportion of false negative species identifications. In our view, sample fractionization is an appropriate method and superior to morphological determination before sequencing (Remmel et al. 2024 ) and have the additional advantage of keeping controls for later reinspection (Hawthorne et al. 2025 ). We see a great potential for better filtering, better barcoding databases and appropriate measures for species with low abundance to reduce the number of false negatives while increasing the number of true positives. False positive species in turn were detected by metabarcoding with a similarly high rate as false negatives of around 30%. This corresponds to a 21 and 22% false positive rate reported for insects from Germany (Remmel et al. 2024 ) and beetles across France (Penel et al. 2025 ). Such false positive rates might mainly be influenced by data filtering and inaccuracy of current databases (Jones et al. 2025 ). A key issue are errors in the databases. A prominent example are co-amplified bacteria from the gut or soma like Wolbachia (Mioduchowska et al. 2018 ). Contamination can practically occur at any stage of the laboratory process, and despite adapted filtering NUMTs remain a problem (Hebert et al. 2023 ). Due to the GBOL (German barcoding of life) initiative (Geiger et al. 2016 ) the species coverage for insects is actually around 57 percent for Germany. However, the intraspecific genetic diversity might not be satisfyingly reflected. Any lack in the sequence databases bears the risk for misinterpretation of sequences and picking related, but wrong species with bioinformatics tools like the Basic Local Alignment Search Tool (BLAST). We do not expect that setting a threshold to deliminate sequences with low read number would reduce the number of false positives because in our study read numbers were not well reflected in number of insect individuals. The use of taxon group-wise thresholds (Arribas et al. 2021 ), more complex filters (Noguerales et al. 2023 ) and the incorporation of morphological controls in combination with better curated reference databases are necessary (Penel et al. 2025 ). It is argued, while exact species identification is unsatisfactory, the number of taxa is sufficiently recovered by metabarcoding (Salis et al. 2024 ). For bees, true bugs, butterflies and Syrphidae caught in four malaise traps near to Frankfurt, Remmel et al. ( 2024 ) found community richness and taxonomic composition to be similar between methods. Consequently, metabarcoding lists might infer similar patterns of the overall community (Petsopoulos et al. 2024 ). This view depends largely on our expectations of congruence; in our study a significant relationship for species richness was only encountered for the Syrphidae (Fig. 1 b), while strongly deviating from the 1:1 expectation, but not for Orthoptera, Auchenorrhyncha or the combination of all. The overreporting of species numbers by metabarcoding up to a certain threshold (in our study seven in Syrphidae) and an underreporting above that threshold is congruent with the patterns found for French beetles (Penel et al. 2025 ). While Remmel et al. ( 2024 ) carried out their comparison across four traps operated in three seasons and considered the twelve samples independent, we accounted for the spatial structure in our analysis. This means that a potential correspondence between both identification methods found by Remmel et al. ( 2024 ) can be a statistical artefact and might be driven by the varying numbers of insects among seasons. In conclusion, when accounting for independence of samples, metabarcoding is not reliable for assessing species richness. Even if a worldwide overview of published data stated a robust outcome across 99 studies, the overall consistency between metabarcoding and morphological species identification was only 49% (Salis et al. 2024 ), i.e. lower than by chance. It has been argued that abundances can be derived from read numbers obtained from metabarcoding (e.g. Salis et a. 2024). However, estimating abundances of certain taxa is difficult as metabarcoding detect and amplify gene sequences with different sensitivity (Shaffer et al. 2025 ). When restricting our analysis to true positive taxa that were encountered both by the morphological and metabarcoding, our species rank abundance curves had little in common (Fig. 4 ). It is obvious that estimation of abundance data from metabarcoding needs more research (Shaffer et al. 2025 ). The common metabarcoding approach with the use of a single DNA fragment for species identification remains inherently problematic as the 313 bp long stretch of the COI DNA region is not unique enough to separate all taxa. To improve sensitivity and precision of the method the use of multiple or at least two gene regions for the implementation of multiplex PCR approaches are proposed (Iwaszkiewicz-Eggebrecht et al. 2024 ). The method can also gain from better mock samples included as controls. As shown for zooplankton, mock samples including previously determined specimens with known biomass will allow more accurate species detection by metabarcoding and might also be used as internal standard for quantitative assessments (Ershova et al. 2023 ). Hence, we will include predetermined individuals in future reevaluation of our metabarcoding. Another elegant way can be the addition of DNA for synthetic cytochrome oxidase spike proteins, acting as internal standards and allow better between sample comparisons (Iwaszkiewicz-Eggebrecht et al. 2024 ). In conclusion, we will need further validation like ours to increase the model fit by increasing the number of true positives in metabarcoding. Metabarcoding studies should mandatorily add morphological validations to their records which is steadily increasing in plankton studies (Neri et al. 2025, Weydmann-Zwolicka et al. 2024 ) and sparsely in insects (Buchner et al. 2025 , Penel et al. 2025 ), to enable a comparison with actual species lists. For central Europe a predefined set of species might be used as morphological standard, with Syrphidae being a promising group for validation. Integrated approaches combining metabarcoding with morphological assessments and advanced image-based techniques (Svenning et al. 2025 ) are on the horizon. For the time being, special attention should be given to metabarcoding results for legal conservation purposes. Here, metabarcoding is currently lacking the necessary accuracy at the species level, in particular for Red list species, and morphological identification will remain the gold standard. Material and Methods Insect sampling Within the project DINA, the insect species diversity and insect biomass were investigated in 21 nature reserves (Natura 2000 areas) in Germany (Lehmann et al. 2021 ). The sample sites were selected after a spatial analysis of all 8836 nature conservation areas based on GIS analyses. The requirements for the study sites were grassland-dominated habitat types with adjacent or integrated arable land. After cooperation with local authorities and landowners, 21 study sites were finally selected. At each site, transects were equipped with five Malaise traps for mass sampling of insects, starting on the arable land (trap 1), the transition zone (trap 2) and reaching into the protected area (traps 3 to 5) with distances of 25 m between the individual traps. The traps ran continuously from April to September 2020 and 2021 with a two-week collection interval to obtain phenological data and the potential to record species with short flight times (for details on the design see Lehmann et al. 2021 ). The Malaise traps were produced in a standardized way by the Entomological Society Krefeld and installed in the field following a defined protocol (Hallmann et al. 2017 ) to ensure data comparability with past and future insect monitoring studies. Flying insects were collected into 1000-ml polyethylene bottles which were emptied on average every 14 days. The data presented here were collected in a two-week interval in 2020 (May 16 to June 2, see Suppl. Table 3). Here we compare the results from 12 locations from this period for which full sample sets were available including the metabarcoding (Köthe & Schneider et al. 2023). Insect samples collected in the Malaise traps were split into two parts by the Entomological Society Krefeld (EVK) using a specially developed sample divider (Hörren et al. 2022 ) and subjected to metabarcoding at the Alexander Koenig Zoological Research Museum (ZFMK; see Köthe & Schneider et al. 2023, Thomas et al. 2025 ), while the second half was morphologically sorted at The Nature and Biodiversity Conservation Union (NABU). We selected three insect groups which are suitable bioindicators for determining the quality of grassland dominated habitat types. Morphological species identification The sample subset stored for morphological analysis was sorted into insect orders, with the main groups being Diptera, Hymenoptera, Coleoptera, Lepidoptera, and Hemiptera (Lehmann GUC et al. data on file). Our three selected target groups (Diptera: Syrphidae, Hemiptera: Auchenorrhyncha, and Orthoptera) were sorted and morphologically identified by experts (Orthoptera, Syrphidae: Arne W. Lehmann, Auchenorrhyncha: Roland Mühlethaler) under Stereomicroscopes of the types Citoval 2 (Zeiss Jena, Germany). Determinations were made with the support of standard literature (Auchenorrhyncha: Biedermann & Niedringhaus ( 2004 ), Holzinger ( 2003 ), Orthoptera adults: Horstkotte et al. 1991 , Orthoptera nymphs: Thommen et al. 2021, and Syrphidae: van Veen 2010). The private collections of Lehmann and Mühlethaler were used for comparison and genital extractions (Syrphidae) or dissections (Auchenorrhyncha) were performed when necessary. Identified specimens were labelled and stored in separate jars. The material will be permanently deposited at the collections of the Entomological Society Krefeld, Germany. Species numbers and nature protection status for Germany were extracted from the newest Red Lists (Auchenorrhyncha: Nickel et al. 2016 , Orthoptera: Poniatowski et al. 2024 , Syrphidae: Ssymank et al. 2011 ). Insect metabarcoding Each sample was fractioned into two size classes referred to as S (small, 4 mm) to avoid the underrepresentation of species with small biomass. These subsamples were dried until complete evaporation of ethanol, homogenized and lysated. Lysates of both size fractions were used and subsequently merged into constant proportions (90% of size class S with 10% of size class L (Elbrecht et al. 2021 ). The eluted DNA was checked for quality on an agarose gel. 84 tissue samples were processed on a 96 well spin column plate and complemented with 12 negative controls (only lysis buffer (ATL, Qiagen) and 10% Proteinase K (Qiagen)). We applied a two-step PCR protocol using standard Illumina Nextera primers for dual indexing of samples, with fwhF2 forward (Vamos et al. 2017 ) and Fol_degen_rev reverse (Yu et al. 2012 ) primer. With a Quantus fluorometer (Promega, Madison, USA) on a Fragment Analyzer (Agilent Technologies, Santa Clara, CA, USA) the library concentration was measured, and the pool was sent for sequencing on a Novaseq SP platform at CeGaT GmbH (Tübingen, Germany). Data analysis was performed on merged paired-end reads with a length of 303–323 bp as implemented in JAMP v0.78 (Elbrecht 2019 ). The taxonomic assignment of the molecular units was carried out by comparison with a reference database for Arthropoda, generated by a beta version of the Taxalogue (Noll et al. 2023 ) with sequences from BOLD (Barcode of Life Data System) (Ratnasingham and Hebert 2007 ), NCBI GenBank (Clark et al. 2016 ) and GBOL (German Barcode of Life) (Geiger et al. 2016 ) and at least 85% sequence similarity. Validation against the standard of morphological species identification To compile the consensus species lists the metabarcoding results was filtered to include only results for the order Orthoptera, the suborder Auchenorrhyncha (Hemiptera) and the family Syrphidae (Diptera). OTUs assigned to species were merged but kept in the consensus list for clarity (Suppl.Table 1 ). Results from the metabarcoding and morphological identifications were combined with three foci (the single Malaise trap, the combined five traps of a transect of a single area, and for all 12 areas across Germany). OTUs that only allowed identification up to the genus, family or order level were also summarized (Suppl.Table 1 ). Statistical analysis The statistical analysis was performed with Rx64 4.4.2–4.5.1 (R Core Team, 2024-25) embedded in RStudio (RStudio Team 2024-25). We used the R programs for Linear mixed effect models (lmer: Kuznetsova et al. 2017 ), Generalized Additive Models (mgcv: Wood 2017 ) and ranking of abundance curves (matrixStats: Bengtsson et al. 2025 ). Binary model classification was done with a confusion matrix (caret: Kuhn 2008 ). It is based on a 2 x 2 matrix with the correctly predicted positive (True Positive TP = identified by both morphology and metabarcoding), and negative instances (True negatives TN = neither identified by morphology nor metabarcoding), and the type I error of incorrectly predicted positive (False Positive FP) and the type II error of positive identifications incorrectly classified as negative (False Negative FN). Based on these four cases several scores are established, with sensitivity (TP / TP + FN) being the true positive rate and precision (TP / TP + FP) the positive predicted value. The F1 score is a metric for evaluating the performance of classification models without considering the true negatives. It is the harmonic mean of precision and sensitivity (F1 score = 2 * (Precision*Sensitivity) / (Precision + Sensitivity)), providing a balance between the two metrics. The F1 score ranges from 0 to 1, where 1 indicates perfect precision and sensitivity, and 0 the worst performance (Grandini et al. 2020 . Declarations Competing interests The author(s) declare no competing interests. Funding This study was funded within the DINA project (Diversity of Insects in Nature protected Areas) by the Federal Ministry of Education and Research (BMBF) under the grant number FKZ 01LC1901 and handled by the VDI project management organization. The BfN (Bundesamt für Naturschutz) financially supported the analysis of the probes and the determination of some taxa. Author Contribution R.M. contributed substantially to the conception of the work, acquired data, interpreted the data, drafted the work and substantially revised it; A.W.L. contributed to the design of the work, acquired data, analyzed and interpreted the data, drafted the work and substantially revised it; S.K. contributed substantially to the conception of the work, acquired data, analyzed the data and drafted the work; H.B. analyzed the data, interpreted the data, and substantially revised the work; G.U.C.L. contributed substantially to the conception and design of the work, acquired, analyzed and interpreted the data, and substantially revised the work. All authors have approved the submitted version and agreed to be personally accountable. Acknowledgement We thank the Federal Ministry of Education and Research (BMBF) for founding the DINA project and the VDI for project management organization for handling (grant number FKZ 01LC1901). The BfN (Bundesamt für Naturschutz) gave financial support for the analysis of the probes and the determination of some taxa, we are grateful. Our study very much benefitted from workshop discussions in the sMon project (Trend analysis of biodiversity data in Germany), a strategic project of the German Centre for Integrative Biodiversity Research (iDiv) Halle- Jena- Leipzig. We would also like to thank Michael R. Wilson (UK) for his linguistic revision and comments on the manuscript. Our gratitude goes to the farmers and authorities for granting licenses to set up Malaise traps and take samples. Without the dedicated support of our more than 45 volunteers and citizen scientists caring for the Malaise traps, we could not have carried out the project. Data Availability The metabarcoding data are already published and available at the Digital Cataloque of the Leibniz Institute for the Analysis of Biodiversity Change (https://collections.leibniz-lib.de) under the accession numbers listed in TableS11 of Thomas et al. (2025).All data are presented in the Supplementary material section. References Arribas, P., Andújar, C., Salces-Castellano, A., Emerson, B. C. & Vogler, A. P. The limited spatial scale of dispersal in soil arthropods revealed with whole-community haplotype-level metabarcoding. Mol. Ecol. 30 (1), 48–61. 10.1111/mec.15591 (2021). Bengtsson, H. et al. matrixStats: Functions that Apply to Rows and Columns of Matrices (and to Vectors)1.5.0. (2025). 10.32614/CRAN.package.matrixStats Biedermann, R. & Niedringhaus, R. Die Zikaden Deutschlands: Bestimmungstafeln für alle Arten (WABV-Fründ, 2004). Buchner, D. et al. Upscaling biodiversity monitoring: Metabarcoding estimates 31,846 insect species from Malaise traps across Germany. Mol. Ecol. Resour. 25 , e14023. https://doi.org/10.1111/1755-0998.14023 (2025). Chua, P. Y. S. et al. Future of DNA-based insect monitoring. Trends Genet. 39 , 531–544. 10.1016/j.tig.2023.02.012 (2023). Clark, K., Karsch-Mizrachi, I., Lipman, D. J., Ostell, J. & Sayers, E. W. GenBank. Nucleic Acids Res. 44 (D1), D67–72. 10.1093/nar/gkv1276 (2016). Creedy, T. J. et al. Coming of age for COI metabarcoding of whole organism community DNA: towards bioinformatic harmonisation. Mol. Ecol. Resour. 22 , 847–861. https://doi.org/10.1111/1755-0998.13502 (2022). Eggleton, P. The state of the World’s insects. Annu. Rev. Environ. Resour. 45 , 61–82. 10.1146/annurev-environ-012420-050035 (2020). Elbrecht, V. JAMP v.0.78. (2019). https://github.com/VascoElbrecht/JAMP Elbrecht, V. et al. Pooling size sorted Malaise trap fractions to maximize taxon recovery with metabarcoding. PeerJ 9 , e12177. https://doi.org/10.7717/peerj.12177 (2021). Ershova, E. A., Wangensteen, O. S. & Falkenhaug, T. Mock samples resolve biases in diversity estimates and quantitative interpretation of zooplankton metabarcoding data. Marine Biodivers. 53 , 66. https://doi.org/10.1007/s12526-023-01372-x (2023). Förster, T., Creutzburg, F., Anton, E., Weigel, A. & Hartmann, M. Metabarcoding versus morphologische Identifizierung: der Herausforderung. gewachsen? Entomol. Z. 133 , 103–116 (2023). Geiger, M. F. et al. How to tackle the molecular species inventory for an industrialized nation—lessons from the first phase of the German Barcode of Life initiative GBOL (2012–2015). Genome 59 , 661–670. https://doi.org/10.1139/gen-2015-0185 (2016). Grandini, M., Bagli, E. & Visani, G. Metrics for multi-class classification: an overview. arXiv preprint arXiv:2008.05756. (2020). Hallmann, C. A. et al. More than 75 percent decline over 27 years in total flying insect biomass in protected areas. PLoS One . 12 , e0185809. https://doi.org/10.1371/journal.pone.0185809 (2017). Hallmann, C. A., Ssymank, A., Sorg, M., de Kroon, H. & Jongejans, E. Insect biomass decline scaled to species diversity: General patterns derived from a hoverfly community. PNAS 118 (2), e2002554117. https://doi.org/10.1073/pnas.2002554117 (2021). Hallmann, C. A. et al. Weather anomalies cannot explain insect decline. Nature 639 , E7–E11. https://doi.org/10.1038/s41586-024-08528-0 (2025). Hawlitschek, O. et al. DNA barcoding of crickets, katydids, and grasshoppers (Orthoptera) from Central Europe with focus on Austria, Germany, and Switzerland. Mol. Ecol. Resour. 17 , 1037–1053. https://doi.org/10.1111/1755-0998.12638 (2017). Hawlitschek, O. et al. New estimates of genome size in Orthoptera and their evolutionary implications. PLoS One . 18 (3), e0275551 (2023). Hawthorne, B. S. J., Cuff, J. P., Collins, L. E. & Evans, D. M. Metabarcoding advances agricultural invertebrate biomonitoring by enhancing resolution, increasing throughput and facilitating network inference. Agric. For. Entomol. 27 (1), 50–66. https://doi.org/10.1111/afe.12628 (2025). Hebert, P. D. N., Cywinska, A., Ball, S. L. & de Waard, J. R. Biological identifications through DNA barcodes. Proceedings of the Royal Society B: Biological Sciences 270: 313–321. (2003). 10.1098/rspb.2002.2218 Hebert, P. D. N., Bock, D. G. & Prosser, S. W. J. Interrogating 1000 insect genomes for NUMTs: a risk assessment for estimates of species richness. PLoS One . 18 (6), e0286620. 10.1371/journal.pone.0286620 (2023). Hochkirch, A. et al. European Red List of Insect Taxonomists (Publication Office of the European Union, 2022). Holzinger, W. E. The Auchenorrhyncha of Central Europe - Die Zikaden Mitteleuropas. Brill (2003). Hörren, T. et al. Development of an insect sample fractionizer for biodiversity research. bioRxiv 2022.11.04.515206. (2022). https://doi.org/10.1101/2022.11.04.515206 Horstkotte, J., Lorenz, C. & Wendler, A. Heuschrecken - Bestimmung, Verbreitung, Lebensräume und Gefährdung aller in Deutschland vorkommenden Arten. 9ed. DJN-Hamburg (1991). IPBES & the global assessment report on biodiversity and ecosystem services of the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services. : Summary for policymakers of. Díaz S, Settele J, Brondízio ES, Ngo HT, Guèze M,Agard J, Arneth A, Balvanera P, Brauman KA, Butchart SHM, Chan KMA, Garibaldi LA, Ichii K, Liu J, Subramanian SM, Midgley GF, Miloslavich P, Molnár Z, Obura D, Pfaff A, Polasky S, Purvis A, Razzaque J, Reyers B, Chowdhury RR, Shin YJ, Visseren-Hamakers IJ, Willis KJ, Zayas CN (eds.) IPBES secretariat, Bonn, Germany. (2019). https://doi.org/10.5281/zenodo.3553579 Iwaszkiewicz-Eggebrecht, E., Zizka, V. & Lynggaard, C. Three steps towards comparability and standardization among molecular methods for characterizing insect communities. Philosophical Trans. Royal Soc. B: Biol. Sci. 379 , 20230118. https://doi.org/10.1098/rstb.2023.0118 (2024). Jones, J. I. et al. Sources of uncertainty in DNA metabarcoding of whole communities: Implications for its use in biomonitoring. Methods Ecol. Evol. 16 , 1658–1673. https://doi.org/10.1111/2041-210X.70093 (2025). Karlsson, D. et al. The Swedish Malaise Trap Project: A 15 Year Retrospective on a Countrywide Insect Inventory. Biodivers. Data J. 8 , e47255. https://doi.org/10.3897/bdj.8.e47255 (2020). Kortmann, M. et al. A shortcut to sample coverage standardization in metabarcoding data provides new insights into land-use effects on insect diversity. Proceedings of the Royal Society B: Biological Sciences 292:20242927. (2025). http://doi.org/10.1098/rspb.2024.2927 Köthe, S. et al. Improving insect conservation management through insect monitoring and stakeholder involvement. Biodivers. Conserv. 32 , 691–713. https://doi.org/10.1007/s10531-022-02519-1 (2023). Kuhn, M. Building Predictive Models in R Using the caret Package. J. Stat. Softw. 28 (5), 1–26. 10.18637/jss.v028.i05 (2008). Kuznetsova, A., Brockhoff, P. B. & Christensen, R. H. B. lmerTest Package: Tests in Linear Mixed Effects Models. J. Stat. Softw. 82 (13), 1–26. 10.18637/jss.v082.i13 (2017). Lehmann, G. U. C. et al. Diversity of Insects in Nature protected Areas (DINA): an interdisciplinary German research project. Biodivers. Conserv. 30 , 2605–2614. https://doi.org/10.1007/s10531-021-02209-4 (2021). Li, Y. et al. Combining environmental DNA and remote sensing for efficient, fine-scale mapping of arthropod biodiversity. Philosophical Trans. Royal Soc. B: Biol. Sci. 379 , 20230123. https://doi.org/10.1098/rstb.2023.0123 (2024). Liu, X. et al. Genomic and transcriptomic perspectives on the origin and evolution of NUMTs in Orthoptera. Mol. Phylogenet. Evol. 201 , 108221. https://doi.org/10.1016/j.ympev.2024.108221 (2024). Luo, M., Ji, Y., Warton, D. & Yu, D. W. Extracting abundance information from DNA -based data. Mol. Ecol. Resour. 23 , 174–189. https://doi.org/10.1111/1755-0998.13703 (2023). Mioduchowska, M., Czyż, M. J., Gołdyn, B., Kur, J. & Sell, J. Instances of erroneous DNA barcoding of metazoan invertebrates: Are universal cox1 gene primers too universal? PLoS ONE . 13 (6), e0199609. https://doi.org/10.1371/journal.pone.0199609 (2018). Mühlethaler, R., Holzinger, W. E., Nickel, H. & Wachmann, E. Die Zikaden Deutschlands, Österreichs und der Schweiz (Quelle & Meyer, 2019). Mühlethaler, R. et al. No recovery in the biomass of flying insects over the last decade in German nature protected areas. Ecol. Evol. 14 , e11182. https://doi.org/10.1002/ece3.11182 (2024). Neri, F. et al. Comparative analysis of phytoplankton diversity using microscopy and metabarcoding: insights from an eLTER station in the Northern Adriatic Sea. Hydrobiologia 852 , 169–183. https://doi.org/10.1007/s10750-024-05692-2 (2024). Nickel, H. et al. Rote Liste und Gesamtartenliste der Zikaden (Hemiptera: Auchenorrhyncha) Deutschlands. In: Gruttke H, Balzer S, Binot-Hafke M, Haupt H, Hofbauer N, Ludwig G, Matzke-Hajek G, Ries M (Red.): Rote Liste gefährdeter Tiere, Pflanzen und Pilze Deutschlands, Band 4: Wirbellose Tiere (Teil 2). Naturschutz und Biologische Vielfalt 70(4): 249–298. (2016). Noguerales, V. et al. Community metabarcoding reveals the relative role of environmental filtering and spatial processes in metacommunity dynamics of soil microarthropods across a mosaic of montane forests. Mol. Ecol. 32 , 6110–6128. https://doi.org/10.1111/mec.16275 (2023). Noll, N. W., Scherber, C. & Schäffler, L. taxalogue: a toolkit to create comprehensive CO1 reference databases. PeerJ 11 , e16253. 10.7717/peerj.16253 (2023). Penel, B. et al. The best of two worlds: toward large-scale monitoring of biodiversity combining COI metabarcoding and optimized parataxonomic validation. Ecography 2025 (6), e07699. https://doi.org/10.1111/ecog.07699 (2025). Petsopoulos, D. et al. Identifying archived insect bulk samples using DNA metabarcoding: A case study using the long-term Rothamsted Insect Survey. Environ. DNA . 6 (3), e542. https://doi.org/10.1002/edn3.542 (2024). Poniatowski, D. et al. Rote Liste und Gesamtartenliste der Heuschrecken und Fangschrecken (Orthoptera et Mantodea) Deutschlands. Naturschutz und Biologische Vielfalt . 170 (7), 1–88 (2024). Ratnasingham, S. & Hebert, P. D. N. bold: The Barcode of Life Data System. Mol. Ecol. Notes . 7 , 355–364. https://doi.org/10.1111/j.1471-8286.2007.01678.x (2007). http://www.barcodinglife.org Core Team, R. (2024–2025) R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. < https://www.R-project.org/ Remmel, N. et al. DNA metabarcoding and morphological identification reveal similar richness, taxonomic composition and body size patterns among flying insect communities. Insect Conserv. Divers. 17 (3), 449–463. https://doi.org/10.1111/icad.12710 (2024). RStudio, T. -25) RStudio: Integrated Development Environment for R (Version 2024.12.0–2025.05.1) [Computer software] (RStudio, PBC, 2024). http://www.rstudio.com/ Salis, R., Sunde, J., Gubonin, N., Franzén, M. & Forsman, A. Performance of DNA metabarcoding, standard barcoding and morphological approaches in the identification of insect biodiversity. Mol. Ecol. Resour. 24 , e14018. https://doi.org/10.1111/1755-0998.14018 (2024). Shaffer, M. R. et al. Observation Bias in Metabarcoding. Molecular Ecology Resources, early view:e14119. (2025). https://doi.org/10.1111/1755-0998.14119 Skvarla, M. J., Larson, J. L., Fisher, J. R. & Dowling, A. P. G. A review of terrestrial and canopy Malaise traps. Ann. Entomol. Soc. Am. 114 (1), 27–47. https://doi.org/10.1093/aesa/saaa044 (2021). Ssymank, A., Doczkal, D., Rennwald, K. & Dziock, F. Rote Liste und Gesamtartenliste der Schwebfliegen (Diptera: Syrphidae) Deutschlands. In: Binot-Hafke M, Balzer S, Becker N, Gruttke H, Haupt H, Hofbauer N, Ludwig G, Matzke-Hajek G, Strauch M (ed.): Rote Liste der gefährdeten Tiere, Pflanzen und Pilze Deutschlands. Band 3: Wirbellose Tiere (Teil 1). Bonn (Bundesamt für Naturschutz). Naturschutz und Biologische Vielfalt 70 (3): 13–83. (2011). Strutzenberger, P. et al. DNA metabarcoding of light trap samples vs. morphological species identification. Ecol. Entomol. 49 (2), 245–256. https://doi.org/10.1111/een.13297 (2024). Svenning, A. et al. A General Method for Detection and Segmentation of Terrestrial Arthropods in Images. bioRxiv 2025.04.08.647223. (2025). https://doi.org/10.1101/2025.04.08.647223 Taberlet, P., Coissac, E., Pompanon, F., Brochmann, C. & Willerslev, E. Towards next-generation biodiversity assessment using DNA metabarcoding. Mol. Ecol. 21 , 2045–2050. 10.1111/j.1365-294X.2012.05470.x (2012). Thomas, L. J. et al. Identifying conservation hotspots and assessing species commonness and rarity: Baseline arthropod diversity in German nature reserves via national Malaise trap monitoring. Insect Conserv. Divers. 18 (4), 510–525. https://doi.org/10.1111/icad.12818 (2025). Thommen, D. Jugendstadien der Heuschrecken der Schweiz - Der Bestimmungsführer für Heuschrecken-Nymphen (Haupt, 2021). Vamos, E., Elbrecht, V. & Leese, F. Short COI markers for freshwater macroinvertebrate metabarcoding. Metabarcoding Metagenomics . 1 , e14625. https://doi.org/10.3897/mbmg.1.14625 (2017). van Klink, R. et al. Meta-analysis reveals declines in terrestrial but increases in freshwater insect abundances. Science 368 , 417–420. https://doi.org/10.1126/science.aax9931 (2020). van Veen, M. P. Hoverflies of Northwest Europe: identification keys to the Syrphidae. 2nd edition. KNNV uitgeverij. (2020). 10.13140/RG.2.1.3697.4567 Weydmann-Zwolicka, A., Dąbrowska, A. M., Mioduchowska, M. & Zwolicki, A. Comparison of DNA metabarcoding and microscopy in analysing planktonic protists from the European Arctic. Marine Biodivers. 54 , 44. https://doi.org/10.1007/s12526-024-01436-6 (2024). Wirth, C., Bruelheide, H., Farwig, N., Marx, J. M. & Settele, J. Faktencheck Artenvielfalt. Bestandsaufnahme und Perspektiven für den Erhalt der biologischen Vielfalt in Deutschland (oekom, 2024). Wolany, L., Enss, J., Leese, F. & Buchner, D. Homogenization of insect bulk samples yields more comprehensive yet comparable biodiversity data than non-destructive lysis. Metabarcoding Metagenomics . 8 , e129814. https://doi.org/10.3897/mbmg.8.129814 (2024). Wood, S. : mgcv: Mixed GAM Computation Vehicle with Automatic Smoothness Estimation. (2017). 10.32614/CRAN.package.mgcv Yu, D. W. et al. Biodiversity soup: metabarcoding of arthropods for rapid biodiversity assessment and biomonitoring. Methods Ecol. Evol. 3 , 613–623. https://doi.org/10.1111/j.2041-210X.2012.00198.x (2012). Zizka, V. M. A. et al. Repeated subsamples during DNA extraction reveal increased diversity estimates in DNA metabarcoding of Malaise traps. Ecol. Evol. 12 , e9502. https://doi.org/10.1002/ece3.9502 (2022). Additional Declarations No competing interests reported. Supplementary Files PavingthewayformetabarcodingSupplementarymaterial.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7899842","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":547866863,"identity":"917bdb0d-e8ea-417c-9e81-0db2a34241a2","order_by":0,"name":"Roland Mühlethaler","email":"","orcid":"","institution":"NABU (The Nature And Biodiversity Conservation Union)","correspondingAuthor":false,"prefix":"","firstName":"Roland","middleName":"","lastName":"Mühlethaler","suffix":""},{"id":547866864,"identity":"2f45030f-1d72-4e0d-a30b-c64d167fbb64","order_by":1,"name":"Arne W. Lehmann","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Arne","middleName":"W.","lastName":"Lehmann","suffix":""},{"id":547866865,"identity":"b39b3d92-bddc-4f21-bd79-4a8af7cb6a2d","order_by":2,"name":"Sebastian Köthe","email":"","orcid":"","institution":"NABU (The Nature And Biodiversity Conservation Union)","correspondingAuthor":false,"prefix":"","firstName":"Sebastian","middleName":"","lastName":"Köthe","suffix":""},{"id":547866866,"identity":"d33a8dae-dbd1-41f8-9e25-895fa85008dc","order_by":3,"name":"Helge Bruelheide","email":"","orcid":"","institution":"Martin Luther University Halle Wittenberg","correspondingAuthor":false,"prefix":"","firstName":"Helge","middleName":"","lastName":"Bruelheide","suffix":""},{"id":547866867,"identity":"83946d4f-50e1-4705-aef8-0e8181b8ffc4","order_by":4,"name":"Gerlind U.C. Lehmann","email":"data:image/png;base64,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","orcid":"","institution":"NABU (The Nature And Biodiversity Conservation Union)","correspondingAuthor":true,"prefix":"","firstName":"Gerlind","middleName":"U.C.","lastName":"Lehmann","suffix":""}],"badges":[],"createdAt":"2025-10-19 16:38:06","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7899842/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7899842/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":96529397,"identity":"3e1baf4e-485e-4a4b-a7bc-2ae868bdfb2e","added_by":"auto","created_at":"2025-11-22 15:07:12","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":556927,"visible":true,"origin":"","legend":"","description":"","filename":"Pavingthewayformetabarcodingv3.docx","url":"https://assets-eu.researchsquare.com/files/rs-7899842/v1/169aec0efdf9bd91a9e97e90.docx"},{"id":96529395,"identity":"e8a92988-68ab-4793-8578-5a5e0d67f633","added_by":"auto","created_at":"2025-11-22 15:07:12","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":8024,"visible":true,"origin":"","legend":"","description":"","filename":"c6d62d815f02484d8de3ea87aca29efe.json","url":"https://assets-eu.researchsquare.com/files/rs-7899842/v1/0faf5edac7ed980bada83a06.json"},{"id":96529398,"identity":"c302bf49-3bea-4fb0-9fca-191a815cd765","added_by":"auto","created_at":"2025-11-22 15:07:12","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":715444,"visible":true,"origin":"","legend":"","description":"","filename":"PavingthewayformetabarcodingSupplementarymaterial.docx","url":"https://assets-eu.researchsquare.com/files/rs-7899842/v1/1207f4c6832bc3e8abfb9564.docx"},{"id":96529404,"identity":"b87d4506-66f0-4a7e-9f58-a44bef0efe35","added_by":"auto","created_at":"2025-11-22 15:07:12","extension":"xml","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":153615,"visible":true,"origin":"","legend":"","description":"","filename":"c6d62d815f02484d8de3ea87aca29efe1enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7899842/v1/ae47aebf7896ed2adb465027.xml"},{"id":96529413,"identity":"8a75283a-1116-48c4-8ab6-7157f58d5912","added_by":"auto","created_at":"2025-11-22 15:07:12","extension":"png","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":154612,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7899842/v1/bf5bf34738eb55b77e573fe5.png"},{"id":96604637,"identity":"12ca69c4-580c-4e2a-9091-c96ead7d2a72","added_by":"auto","created_at":"2025-11-24 09:14:26","extension":"png","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":29208,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7899842/v1/7ee30eaeac8bf65a6ebe6cd9.png"},{"id":96604373,"identity":"520711b4-b043-4ccf-8456-1b5f92374899","added_by":"auto","created_at":"2025-11-24 09:13:46","extension":"png","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":75778,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7899842/v1/78718f7ea856078ffe192352.png"},{"id":96529410,"identity":"8102a621-febd-4837-ab30-2f390ae5606c","added_by":"auto","created_at":"2025-11-22 15:07:12","extension":"png","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":180308,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7899842/v1/cf0224c096d29350880122b9.png"},{"id":96603471,"identity":"1cf4950b-527d-4762-879e-31feef757f25","added_by":"auto","created_at":"2025-11-24 09:09:26","extension":"png","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":46927,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7899842/v1/e49019f83c5d6385f836de47.png"},{"id":96529408,"identity":"09d04a09-7459-4131-ba98-fba8f3547ac5","added_by":"auto","created_at":"2025-11-22 15:07:12","extension":"png","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":15870,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7899842/v1/294e00523a26118e0a6f91df.png"},{"id":96529411,"identity":"e46a0e4c-c301-4847-bc7f-15e0f49492db","added_by":"auto","created_at":"2025-11-22 15:07:12","extension":"png","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":22947,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7899842/v1/055e0dc94b0ca7408de28788.png"},{"id":96529402,"identity":"b1df337d-e60c-471e-99a8-fadb1c4bbe1d","added_by":"auto","created_at":"2025-11-22 15:07:12","extension":"png","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":58046,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7899842/v1/d023ed6d57e3879a76a2da02.png"},{"id":96529412,"identity":"d2e99ccb-544e-4d39-b9b3-1ce3f285e4fb","added_by":"auto","created_at":"2025-11-22 15:07:12","extension":"xml","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":152145,"visible":true,"origin":"","legend":"","description":"","filename":"c6d62d815f02484d8de3ea87aca29efe1structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7899842/v1/43b61840014c37cc315ac00e.xml"},{"id":96604070,"identity":"2781657e-a08d-4991-8fd1-d5d77ffd115c","added_by":"auto","created_at":"2025-11-24 09:12:38","extension":"html","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":163015,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7899842/v1/1051ee6ef829418dc4c83996.html"},{"id":96529393,"identity":"7a94a3e7-479e-48db-86bd-4a35bba017b2","added_by":"auto","created_at":"2025-11-22 15:07:12","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":162954,"visible":true,"origin":"","legend":"\u003cp\u003eSpecies richness compared between morphological and metabarcoding identification for (a) all combined species, and separated for (b) Syrphidae, (c) Auchenorrhyncha, and (d) Orthoptera. The dashed line shows the bisecting 1:1 relationship for comparison. The bold line is the linear regression from a mixed effects model, using site as random slope factor (see Suppl. Table 2b). The slope estimate was significant only for hoverflies (b) significantly deviates from zero (0.26 ± 0.13, t\u003csub\u003e55.68 \u003c/sub\u003e= 2.07, \u003cstrong\u003ep = 0.044\u003c/strong\u003e). For list of site names see Suppl. Table 3.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-7899842/v1/1717f5f3726a5ae7016d9fb4.png"},{"id":96604490,"identity":"30ccb5e5-e8b4-47c9-a9db-fc0e91b87b6b","added_by":"auto","created_at":"2025-11-24 09:14:02","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":85781,"visible":true,"origin":"","legend":"\u003cp\u003eComparison between the number of species detected by both methods (true positive = TP, dark color), only by metabarcoding (false positive = FP, rich color) or only by morphology (false negative = FN, light color) for a) Syrphidae (blue), b) Auchenorrhyncha (orange) and c) Orthoptera (green) aggregated for the 12 sites (list of site names Suppl. Table 3).\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-7899842/v1/e937208b9bbf93bcdcb00500.png"},{"id":96604725,"identity":"74ee13e7-0866-479a-a866-44a3125065f6","added_by":"auto","created_at":"2025-11-24 09:14:42","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":85645,"visible":true,"origin":"","legend":"\u003cp\u003eRelationship of the F1score, the harmonic mean between sensitivity and precision, as a function of narrowing the focus from the whole study (all) to site to traps, separately for the taxonomic levels of species (solid lines, triangles) and genera (dotted lines, circles) for Syrphidae (blue), Auchenorrhyncha (orange), and Orthoptera (green).\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-7899842/v1/5c41c46ab700eece2f75dfa2.png"},{"id":96605109,"identity":"0689fea2-b955-4724-ae44-9c31cfbdfa0d","added_by":"auto","created_at":"2025-11-24 09:18:27","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":191407,"visible":true,"origin":"","legend":"\u003cp\u003eBoxplots for number of individuals (a and b) and number of reads from metabarcoding (c and d), summing up all values at the species level across all traps per site and including only true positive species detected with both methods. Boxplots show the median and quantiles across the 12 sites. Species are arranged according to declining median numbers of individuals identified by morphology. a) and c) Syrphidae b) and d) Auchenorrhyncha.\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-7899842/v1/84ced9c4c131b8080c5a1a2c.png"},{"id":96529406,"identity":"a83ee7e4-611e-4834-8c3c-8acaa5a5cafa","added_by":"auto","created_at":"2025-11-22 15:07:12","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":35988,"visible":true,"origin":"","legend":"\u003cp\u003eSensitivity (dark color) and precision (lite color) separated into all species and for Red list species (red stripes) stratified for Syrphidae, Orthoptera and Auchenorrhyncha shows the general lower fit for Red list species.\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-7899842/v1/64edc0bab6f9fa71bea1f5bd.png"},{"id":96920374,"identity":"100da18b-39e4-41e7-bc61-5c4975e7aa46","added_by":"auto","created_at":"2025-11-27 14:15:07","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1225677,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7899842/v1/88177f08-cf2f-4c1d-b90c-b830a525a347.pdf"},{"id":96529399,"identity":"e5dad275-f2b7-4dcf-875a-1e0ed4a0fc23","added_by":"auto","created_at":"2025-11-22 15:07:12","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":715444,"visible":true,"origin":"","legend":"","description":"","filename":"PavingthewayformetabarcodingSupplementarymaterial.docx","url":"https://assets-eu.researchsquare.com/files/rs-7899842/v1/8d25d1a98c9ae274a2656703.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Paving the way for improved insect metabarcoding","fulltext":[{"header":"Introduction","content":"\u003cp\u003eGlobal biodiversity is declining at an alarming rate (IPBES 2019), and terrestrial insects especially are affected (van Klink et al. \u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). This group has enormous diversity being the most species-rich taxon of Metazoa and plays important roles in almost all terrestrial ecosystems. The accurate identification of insect species is essential for many scientific fields including ecology, evolutionary biology, agriculture, and medical entomology (Eggleton \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2020\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe loss of insects has become the focus of many research projects, with a landmark study from Germany, where insect biomass declined by more than 75% over the past 27 years (Hallmann et al., \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). In follow-up studies with identical methodology we found this trend to have continued in the last years, without any indication of being reverted (M\u0026uuml;hlethaler et al. \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2024\u003c/span\u003e, Hallmann et al. \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eTo monitor insects, standardized methods, such as Malaise traps, are in widespread use, their broad sampling coverage of mainly flying individuals allows to document spatial and temporal changes (Skvarla et al. \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). One easily accessible metric from malaise traps is insect biomass, which is correlated with species number, both within single taxonomic groups (Hallmann et al. \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) and with overall insect richness (K\u0026ouml;the \u0026amp; Schneider et al. 2023). However, using biomass alone has limitations as it provides no specific answers which and where species are declining. Thus, reliable tools for species identification are needed, in particular to detect rare and/or Red List species. The whole field of nature protection legislation and conservation planning is to a high degree dependent on species records. Important examples at the global level are the Convention on International Trade in Endangered Species (CITES) and the IUCN Red Lists for endangered species, at the EU level the Birds and Habitats Directives and at the national level in Germany the Federal Conservation Act (Wirth et al. \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe traditional approach to morphology-based taxonomy is challenging, firstly because of the labor and time-consuming sorting of thousands of insects. For example, classifying 73 traps sampling for three years needed 15 years and 300 technicians in the Swedish Malaise Trap Project (Karlsson et al. \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Secondly, species determination skills are generally restricted to well-known taxa, while for the majority of species-rich taxon groups and for taxa with small body sizes no experts are available, and experienced taxonomists are a diminishing group at the risk of extinction (Hochkirch et al. \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Fourthly, cryptic species or dark taxa, unresolved taxonomies and the degradation of morphological features in wet samples are a challenge for correct identification.\u003c/p\u003e\u003cp\u003eDNA-based molecular methods are a promising tool to address these issues, as DNA barcoding is easy to implement and allows rapid taxonomic identification (Hebert et al. \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2003\u003c/span\u003e). Metabarcoding, where DNA is isolated and analyzed from a mixture of specimens and mass-amplified using high-throughput technologies, is increasingly used for species identification in bulk samples (Chua et al. \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Metabarcoding sequences are filtered to generate Amplicon Sequence Variants (ASV), which are subsequently clustered into Operational Taxonomic Units (OTU). By bioinformatical analysis and by comparisons with reference databases we may thus identify the taxa present in complex samples (Iwaszkiewicz-Eggebrecht et al. \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Metabarcoding has been tested in the German Malaise trapping programs (Geiger et al. \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2016\u003c/span\u003e) and is employed to study insect diversity in large scale monitoring programs, namely the LTER (Buchner et al. \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2025\u003c/span\u003e, Remmel et al. \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) and the DINA project (K\u0026ouml;the \u0026amp; Schneider et al. 2023, Thomas et al. \u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Despite the huge benefits for mass assessment, metabarcoding has several limitations and not all problems for the integrity and correctness of the database entries and the fine-tuning of laboratory and sequencing methods have been solved yet.\u003c/p\u003e\u003cp\u003eTwo of the key problems for taxon identification with metabarcoding are the failure to detect taxa that are present in a sample (i.e. false negative records) and incorrect detection of taxa that are not present in the sample (i.e. false positive records). False negative records might arise from problems with non-perfect primer fitting, binding to non-target sequences, sequencing errors and random distribution of specimens in subsamples. Due to differences in body size, genome size, mitochondrial copy number, DNA extraction efficiency and PCR amplification efficiency PCR amplification rates differ among taxa (Luo et al., \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). As the amount of DNA extracted differs from small versus large specimens, techniques like size sorting and hence increasing the representation of smaller species in the DNA extract might reduce part of the problem (Elbrecht et al. \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Nonetheless, the bias continues to exist for less abundant taxa and those with smaller body size and can lead to false negative records, altering species presence/absence assessments (Strutzenberger et al. \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Similarly, false positive records might result from wrong reference sequences in the databases or the picking of the most similar species by indiscriminately using bioinformatics tools like the Basic Local Alignment Search Tool (BLAST). As a consequence, metabarcoding lists might infer similar patterns of the overall community (Petsopoulos et al. \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2024\u003c/span\u003e, Buchner et al. \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) but include inaccurate or incorrect taxonomic assignments, especially at the species level (F\u0026ouml;rster et al. \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2023\u003c/span\u003e, Li et al. \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2024\u003c/span\u003e, Remmel et al. \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2024\u003c/span\u003e, Strutzenberger et al. \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe DINA project (Diversity of Insects in Nature Protected Areas) which aims to investigate insect biodiversity in Germany (Lehmann et al. \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) is one of the most prestigious follow-up projects of the landmark study on insect decline by Hallmann et al. (\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). For this purpose, Malaise traps were operated along transects in 21 representative nature protected areas across Germany (Lehmann et al. \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), with a total of 105 traps operated simultaneously and assessed for biomass and species identification (K\u0026ouml;the \u0026amp; Schneider et al. 2023, M\u0026uuml;hlethaler et al. \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2024\u003c/span\u003e, Hallmann et al. \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eWithin our project, we compared DNA metabarcoding for whole community species assessments (Thomas et al. \u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) with the gold standard of morphological sorting and identification by experts. Metabarcoding is still more effective with destructive homogenization than with lysis protocols (Wolany et al. \u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) but prevents any morphological inspection after metabarcoding. To overcome these restrictions, the Entomological Society Krefeld developed an insect sample fractionizer (H\u0026ouml;rren et al. \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) to enable subdividing of the samples. In our workflow, the biomassed samples are divided with the first half subjected to metabarcoding, while the second half is used for morphological analysis and subsequently stored.\u003c/p\u003e\u003cp\u003eAs it is impossible to cover all roughly 33,000 to 38,000 insect species in Germany from the estimated 20\u0026nbsp;million insects sampled by the DINA project (Lehmann et al., data on file), we concentrated on three insect groups: Diptera: Syrphidae, Hemiptera: Auchenorrhyncha and Orthoptera.\u003c/p\u003e\u003cp\u003eSyrphidae provide important ecosystem services such as plant pollinators and larval predators. They are one of the main targets for Malaise trap sampling as adults tend to be good fliers (Hallmann et al. \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Around 470 species occur in Germany (Ssymank et al. \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2011\u003c/span\u003e), which have very distinct habitat requirements, especially at the larval stage.\u003c/p\u003e\u003cp\u003eOf the approximately 650 leafhopper and plant hopper species (Auchenorrhyncha, Hemiptera: Cicadomorpha \u0026amp; Fulgoromorpha) known from Germany (Nickel et al. \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2016\u003c/span\u003e), around 50 percent occur in open grasslands (M\u0026uuml;hlethaler et al. \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). In addition, they can build up very high population densities of up to over 5,000 animals/m\u003csup\u003e2\u003c/sup\u003e. Many species have short wings and are therefore less frequently caught in Malaise traps but contribute a considerable fraction of the sampled biomass.\u003c/p\u003e\u003cp\u003eOrthoptera (grasshoppers and crickets) are a small group in Germany of 83 established species (Poniatowski et al. \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), particularly in grasslands, and have been intensively studied as indicators. In line with their behavior, they might be rare in Malaise trap catches but are relatively easy to determine (Horstkotte et al. \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e1991\u003c/span\u003e). Due to their large bodies, their exceptionally large genomes (Hawlitschek et al. \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) and large numbers of fragments of mitochondrial DNA (mtDNA) that have been incorporated into the nuclear genome (NUMTs, Liu et al. \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), they are often overrepresented in metabarcoding results (Remmel et al. \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2024\u003c/span\u003e, Thomas et al. \u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eWe studied how precise metabarcoding reflects species numbers and species identifications. Therefore, we analyzed the performance in a binary classification framework. We constructed 2x2 confusion matrices at different spatial and taxonomic resolution to compare the true positive identifications (found both by morphology and metabarcoding) with the other three classes of false positives, false negatives and true negatives (found neither in the morphological nor the metabarcoding subsample) (Grandini et al. \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2020\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eWe expected a significant match between both methods, which can be expressed as sensitivity (i.e. the fraction of true positives in all morphologically identified taxa) or precision (i.e. the fraction of true positives in all taxa identified by metabarcoding). Moreover, we hypothesized that the match was improved when samples were aggregated across traps and sites, and comparisons were made at the genus rather than at the species level.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003eIn total, 15,107 OTUs from insects were detected by metabarcoding in the 12 nature protected areas with five traps along a transect in the given sampling period Including the taxa identified at the family, genus and species level, the number of OTUs for the three selected groups ranged from 110 for Syrphidae to 155 for Auchenorrhyncha to 1331 for Orthoptera. When focusing only on taxa at the species level, the numbers dropped to 52, 62 and 16 species, respectively (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, Suppl. Table\u0026nbsp;1).\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eNumber of identified OTUs based on DNA metabarcoding from 12 nature protected areas, for all insects and for Syrphidae, Auchenorrhyncha and Orthoptera. In addition, the table shows the number of morphologically identified species and the total number of known species in Germany is given. RL\u0026thinsp;=\u0026thinsp;German Red List.\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"10\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c10\" colnum=\"10\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTaxon\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eOTUs\u003c/p\u003e\u003cp\u003e/ n=\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eOTUs assigned to family level / n=\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eOTUs assigned to genus level / n=\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003eOTUs assigned to species level / n=\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c6\"\u003e\u003cp\u003eSpecies identified by barcoding \u003c/p\u003e\u003cp\u003e/ n=\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c7\"\u003e\u003cp\u003eFamilies identified by morphology\u003c/p\u003e\u003cp\u003e/ n=\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c8\"\u003e\u003cp\u003eGenera identified by morphology\u003c/p\u003e\u003cp\u003e/ n=\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c9\"\u003e\u003cp\u003eSpecies identified by morphology\u003c/p\u003e\u003cp\u003e/ n=\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c10\"\u003e\u003cp\u003eSpecies in Germany \u003c/p\u003e\u003cp\u003e(*after RL) / n=\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAll insects\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e15107\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e14493\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e11785\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e9063\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e2708\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eN/A\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003eN/A\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003eN/A\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e\u003cp\u003e33466\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSyrphidae\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e110\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e110\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e105\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e69\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e52\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e32\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e52\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e\u003cp\u003e467\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAuchenorrhyncha\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e155\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e155\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e139\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e98\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e62\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e7\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e55\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e91\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e\u003cp\u003e645\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eOrthoptera\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e1331\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e1204\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e454\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e269\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e16\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e7\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e8\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c10\"\u003e\u003cp\u003e83\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eThe 3096 individuals identified by morphology were assigned to 151 species, among them 52 species of Syrphidae (879 individuals), 91 species of Auchenorrhyncha (2175 individuals) and 8 species of Orthoptera (42 individuals) (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). The total number across all sites was slightly higher but comparable to the 130 species identified by metabarcoding. However, when analyzing the patterns by trap within sites, the correspondence between both methods decreased considerably. Species richness determined by metabarcoding deviated strongly from species richness determined by morphology both, across all taxon groups and separately for Syrphidae, Auchenorrhyncha and Orthoptera (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, Suppl. Table\u0026nbsp;2a). In the mixed effects models that accounted for differences between sites, the slope in richness was significantly different from zero only for hoverflies, with 0.26 species found with metabarcoding for every species identified morphologically (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb, Suppl. Table\u0026nbsp;2b).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eValidating species lists from metabarcoding against morphological identification was not very successful. When aggregated by sites, the percentage of species identified morphologically compared those by metabarcoding was lower than 50% in all cases (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). Precision, also known as positive predictive value of the matches divided by the number of species reported by metabarcoding, was highest for the Syrphidae with 30%, followed by Auchenorrhyncha with 27%, and lowest for the Orthoptera with 22% (Suppl. Table\u0026nbsp;4a-c). For Auchenorrhyncha there were two and for Orthoptera seven out of 12 localities without any species detection overlap between both methods.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eThe general accuracy of the metabarcoding was high, with above 90% at the species level and when focusing on single traps (Suppl. Table\u0026nbsp;5: 1.1.-1.3). In turn, the error rate was reasonably low, which however, was largely driven by the high proportion of true negatives (TN), i.e. species not being encountered by either method. In turn, the proportion of true positives (TP) was rather low, both when expressed in relation to the false negatives (FN) as sensitivity (i.e. TP/(TP\u0026thinsp;+\u0026thinsp;FN)) or in relation to the false positives (FP) as precision (i.e. positive predicted value, TP/(TP\u0026thinsp;+\u0026thinsp;FP)). Values for both sensitivity and precision were below 50% for all three taxonomic groups at the species level (Suppl. Table\u0026nbsp;5: 1.1.-1.3, Suppl. Figure\u0026nbsp;1). Similarly low was the F1 score as harmonic mean between sensitivity and precision at the species level and focus on single traps, being 0.25 or lower (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). With broadening the focus, i.e. combining all traps per site or all sites of the whole study, increased the F1 score, but never reaching 0.5. The same increase was observed, albeit at higher scores, when the analysis was done at the genus level, but also below 0.5 when focusing on single traps.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eMorphological examination enabled counting individuals per species. Likewise, metabarcoding resulted in read numbers which might to some degree related to the number of individuals. Concentrating on true positive species detected by both methods we compared the mean number of individuals (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea,b) with the mean read counts (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ec,d). Ranking the species by decreasing abundance, the resulting abundance curves show little overlap, neither for Syrphidae (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea,c) nor Auchenorrhyncha (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eb,d).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eFor nature conservation issues, species listed on the Red list of endangered species are of special interest. In our sample, we detected 42 insect species endangered in Germany (Suppl. Table\u0026nbsp;6), either identified only by metabarcoding (21) by morphology (14) and by both methods (7). Accuracy in terms of sensitivity and precision was below 40% for Red list species for Syrphidae and Orthoptera (Fig.\u0026nbsp;6), except for nearly equal sensitivity in Auchenorrhyncha.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eClassical species identification based on morphological characters is highly accurate when carried out by experienced entomologists. Additional information such as the distribution of sex and larvae or wing morphs as indicator of migratory activity can add to the analysis of the insect populations. As a plus, the preserved specimen can be stored for future study after determination. Thus, we expect that morphological identification will remain the gold standard for the time being. However, the large amount of time required for specific identification represents a major barrier, combined with few or missing taxonomic experts and high-quality identification guides. Due to these problems, no one worldwide has yet succeeded in determining a complete annual sample series from a Malaise trap. Even analyzing only small parts of the total diversity is very time-consuming and therefore cost intensive.\u003c/p\u003e\u003cp\u003eIn contrast, identification based on metabarcoding is generally faster, cheaper and able to cover whole communities simultaneously (Chua et al. \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2023\u003c/span\u003e, Hawthorne et al. \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Even more, taxa difficult to identify morphologically and species with little morphological material in the samples (e.g. only individual body parts, endoparasites, intestinal contents) might be detected. Here, we made the comparison with the result of an unexpected poor or even lacking match between both methods. Thus, we have to reject our main hypothesis. The overlap in true positives, i.e. correctly identified species between both methods (true positive determinations\u0026thinsp;=\u0026thinsp;TP) was only around 30%.\u003c/p\u003e\u003cp\u003eThe problem of inaccurate or incorrect taxonomic assignments of metabarcoding has already been recognized for insects (F\u0026ouml;rster et al. \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2023\u003c/span\u003e, Li et al. \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2024\u003c/span\u003e, Remmel et al. \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2024\u003c/span\u003e, Strutzenberger et al. \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2024\u003c/span\u003e, Penel et al. \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) and freshwater macroinvertebrates (Jones et al. \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). However, compared to thousands of barcoding and metabarcoding studies published each year (Salis et al. \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), the number of those evaluating metabarcoding data with morphology species identification is extremely small. As Iwaszkiewicz-Eggebrecht et al. (\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) stated, the metabarcoding scene concentrate largely on laboratory-specific proof of concepts instead of external validation as calibration standards. So far, the focus has more been on standardization of molecular tools (Chua et al. \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2023\u003c/span\u003e, Hawthorne et al. \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) or a unified reporting of lab procedures (Iwaszkiewicz-Eggebrecht et al. \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eHere we will discuss mainly two aspects: (a) the main reasons for the poor model fit between metabarcoding and morphology, (b) the next steps to improve it. Deviations in metabarcoding species detection from morphological determinations was nearly equal in both directions; the failure to detect species present in the samples (false negatives\u0026thinsp;=\u0026thinsp;FN) and the reporting of species not present (false positives\u0026thinsp;=\u0026thinsp;FP). False negative as the failure to detect species actually present in the sample might arise from problems with laboratory routines (Iwaszkiewicz-Eggebrecht et al. \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), bioinformatic (Penel et al. \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) or unequal separation of subsamples. In our study, only one methodological approach of DNA metabarcoding was applied for all preparatory steps in the laboratory, like homogenization, amount of tissue used, PCR steps, selection of primers and the sequencing depth. It can be assumed that this comparison would yield different results if the laboratory practice were carried out differently, e.g. using a higher tissue quantity or multiple use of small tissue quantities, a higher sequencing depth and different bioinformatics pipeline. It is well known that DNA extraction is more successful in some species than others (Elbrecht et al. \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2021\u003c/span\u003e, Zizka et al. \u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Primer fitting especially for species-rich taxon groups is a difficult task as non-perfect primers might bind to non-target sequences and produce sequencing errors. A special challenge is related to the variance in body size between species. As the amount of DNA extracted differs from small versus large specimens, mitochondrial copy number, DNA extraction and PCR amplification rates differ among taxa (Luo et al., \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). In common agreement, we used size sorting (Elbrecht et al. \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), and hence, increased the representation of smaller species in our DNA extracts (Zizka et al. \u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), which might have reduced the problem. Nonetheless, the bias would remain for less abundant taxa and those with smaller body size and can lead to false negative records, altering species presence/absence assessments (Strutzenberger et al. \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). This is also reflected in our analysis of Red List species, which show an even weaker model fit (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e), likely because of their rarity. Furthermore, metabarcoding can suffer from a level of heterogeneity in DNA degradation between samples. Small variations in DNA quality, associated with the stochastic process of PCR amplification of mixed samples, were the largest error source and accounted for 60% of false negative metabarcoding identifications in beetles (Penel et al. \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eWith our bioinformatics pipeline 60% of the OTUs could be assigned to the species level (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). This is far from perfect, and the generation of more complete DNA barcode databases should be a major goal (Chua et al. \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Indeed, the sensitivity was nearly doubled for Syrphidae, Auchenorrhyncha and Orthoptera (Suppl. Figure\u0026nbsp;1, Suppl. Table\u0026nbsp;5) when shifting our analysis from the species to the genus level, supporting our second hypothesis. Even while barcoding of German insects has quite progressed, a lack of reference sequences in the database against which the blasting had been done might be one reason for the many false negatives.\u003c/p\u003e\u003cp\u003eAnother coherent issue is the reliance on the assumption of a barcoding gap, i.e. that an universal similarity threshold can be used for species assignment (Hebert et al. \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2003\u003c/span\u003e). Genetic distances vary considerable intra- and interspecifically and do even overlap, which results in errors when using a fixed threshold for molecular species detection (Creedy et al. \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Flexible filtering approaches based on multiple thresholds (Arribas et al. \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2021\u003c/span\u003e, Penel et al. \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) in combination with broader phylogenetic coverage will probably increase the proportion of true positives in the future.\u003c/p\u003e\u003cp\u003eThe order Orthoptera is especially challenging for metabarcoding as they have enormous genome sizes (Hawlitschek et al. \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) and large numbers of NUMTs (fragments of mitochondrial DNA in the nuclear genome, Liu et al. \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Despite proper filtering of the pseudogenes the OTUs of Orthoptera are regularly overreported (Remmel et al. \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2024\u003c/span\u003e, Buchner et al. \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) even when they have not been in the target group of insects (Kortmann et al. \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Our metabarcoding analysis suffered from the same problem with OTUs for Orthoptera, which accounted for 9% of all OTUs in our samples, while the 83 species of Orthoptera reported for Germany comprise only about 0.2%cent of the total of German insect species (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). A further problem with Central European Orthoptera is the relaxed species resolution with barcoding: up to five species share a barcode and this sharing extends even cross genera (Hawlitschek et al. \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). For the time being, it might be wise to exclude Orthoptera from metabarcoding analyses either by upfront filtering or even better by manually removing the easy to detect and large individuals from the samples prior to metabarcoding.\u003c/p\u003e\u003cp\u003eTo allow for morphological inspection and long-term storage of controls, we fractionized our samples with a splitter into two parts (H\u0026ouml;rren et al. \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Even if we assume an ideal randomization, species that are only contained in few individuals might be present in the first and missing in the second subsample. Such an uneven distribution might contribute to the mismatch when comparing identifications based on metabarcoding from the first half with morphology from the second half. This explanation might apply to taxa with only a very few individuals in the sample. However, we found the mismatch also to be pertinent for frequent species. If the poor match of metabarcoding had been caused by dividing the samples, we would expect a major increase in overlap when shifting the focus from single traps to the combined analysis of five traps at a site. As hypothesized, pooling the samples across sites increased the match, with the sensitivity (true positives divided by TP plus false negatives) increasing from 31, 19 and 17% to 39, 28 and 28% for the three taxa, respectively, and further to 44, 36 and 30%, when summing all 60 traps form all sites across Germany. However, even pooling did not remove the large proportion of false negative species identifications. In our view, sample fractionization is an appropriate method and superior to morphological determination before sequencing (Remmel et al. \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) and have the additional advantage of keeping controls for later reinspection (Hawthorne et al. \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). We see a great potential for better filtering, better barcoding databases and appropriate measures for species with low abundance to reduce the number of false negatives while increasing the number of true positives.\u003c/p\u003e\u003cp\u003eFalse positive species in turn were detected by metabarcoding with a similarly high rate as false negatives of around 30%. This corresponds to a 21 and 22% false positive rate reported for insects from Germany (Remmel et al. \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) and beetles across France (Penel et al. \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Such false positive rates might mainly be influenced by data filtering and inaccuracy of current databases (Jones et al. \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). A key issue are errors in the databases. A prominent example are co-amplified bacteria from the gut or soma like Wolbachia (Mioduchowska et al. \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Contamination can practically occur at any stage of the laboratory process, and despite adapted filtering NUMTs remain a problem (Hebert et al. \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Due to the GBOL (German barcoding of life) initiative (Geiger et al. \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2016\u003c/span\u003e) the species coverage for insects is actually around 57 percent for Germany. However, the intraspecific genetic diversity might not be satisfyingly reflected. Any lack in the sequence databases bears the risk for misinterpretation of sequences and picking related, but wrong species with bioinformatics tools like the Basic Local Alignment Search Tool (BLAST). We do not expect that setting a threshold to deliminate sequences with low read number would reduce the number of false positives because in our study read numbers were not well reflected in number of insect individuals. The use of taxon group-wise thresholds (Arribas et al. \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), more complex filters (Noguerales et al. \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) and the incorporation of morphological controls in combination with better curated reference databases are necessary (Penel et al. \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eIt is argued, while exact species identification is unsatisfactory, the number of taxa is sufficiently recovered by metabarcoding (Salis et al. \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). For bees, true bugs, butterflies and Syrphidae caught in four malaise traps near to Frankfurt, Remmel et al. (\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) found community richness and taxonomic composition to be similar between methods. Consequently, metabarcoding lists might infer similar patterns of the overall community (Petsopoulos et al. \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). This view depends largely on our expectations of congruence; in our study a significant relationship for species richness was only encountered for the Syrphidae (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb), while strongly deviating from the 1:1 expectation, but not for Orthoptera, Auchenorrhyncha or the combination of all. The overreporting of species numbers by metabarcoding up to a certain threshold (in our study seven in Syrphidae) and an underreporting above that threshold is congruent with the patterns found for French beetles (Penel et al. \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). While Remmel et al. (\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) carried out their comparison across four traps operated in three seasons and considered the twelve samples independent, we accounted for the spatial structure in our analysis. This means that a potential correspondence between both identification methods found by Remmel et al. (\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) can be a statistical artefact and might be driven by the varying numbers of insects among seasons. In conclusion, when accounting for independence of samples, metabarcoding is not reliable for assessing species richness. Even if a worldwide overview of published data stated a robust outcome across 99 studies, the overall consistency between metabarcoding and morphological species identification was only 49% (Salis et al. \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), i.e. lower than by chance.\u003c/p\u003e\u003cp\u003eIt has been argued that abundances can be derived from read numbers obtained from metabarcoding (e.g. Salis et a. 2024). However, estimating abundances of certain taxa is difficult as metabarcoding detect and amplify gene sequences with different sensitivity (Shaffer et al. \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). When restricting our analysis to true positive taxa that were encountered both by the morphological and metabarcoding, our species rank abundance curves had little in common (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). It is obvious that estimation of abundance data from metabarcoding needs more research (Shaffer et al. \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe common metabarcoding approach with the use of a single DNA fragment for species identification remains inherently problematic as the 313 bp long stretch of the COI DNA region is not unique enough to separate all taxa. To improve sensitivity and precision of the method the use of multiple or at least two gene regions for the implementation of multiplex PCR approaches are proposed (Iwaszkiewicz-Eggebrecht et al. \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). The method can also gain from better mock samples included as controls. As shown for zooplankton, mock samples including previously determined specimens with known biomass will allow more accurate species detection by metabarcoding and might also be used as internal standard for quantitative assessments (Ershova et al. \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eHence, we will include predetermined individuals in future reevaluation of our metabarcoding. Another elegant way can be the addition of DNA for synthetic cytochrome oxidase spike proteins, acting as internal standards and allow better between sample comparisons (Iwaszkiewicz-Eggebrecht et al. \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eIn conclusion, we will need further validation like ours to increase the model fit by increasing the number of true positives in metabarcoding. Metabarcoding studies should mandatorily add morphological validations to their records which is steadily increasing in plankton studies (Neri et al. 2025, Weydmann-Zwolicka et al. \u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) and sparsely in insects (Buchner et al. \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2025\u003c/span\u003e, Penel et al. \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2025\u003c/span\u003e), to enable a comparison with actual species lists. For central Europe a predefined set of species might be used as morphological standard, with Syrphidae being a promising group for validation. Integrated approaches combining metabarcoding with morphological assessments and advanced image-based techniques (Svenning et al. \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) are on the horizon. For the time being, special attention should be given to metabarcoding results for legal conservation purposes. Here, metabarcoding is currently lacking the necessary accuracy at the species level, in particular for Red list species, and morphological identification will remain the gold standard.\u003c/p\u003e"},{"header":"Material and Methods","content":"\u003cdiv id=\"Sec5\" class=\"Section2\"\u003e\u003ch2\u003eInsect sampling\u003c/h2\u003e\u003cp\u003eWithin the project DINA, the insect species diversity and insect biomass were investigated in 21 nature reserves (Natura 2000 areas) in Germany (Lehmann et al. \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). The sample sites were selected after a spatial analysis of all 8836 nature conservation areas based on GIS analyses. The requirements for the study sites were grassland-dominated habitat types with adjacent or integrated arable land. After cooperation with local authorities and landowners, 21 study sites were finally selected. At each site, transects were equipped with five Malaise traps for mass sampling of insects, starting on the arable land (trap 1), the transition zone (trap 2) and reaching into the protected area (traps 3 to 5) with distances of 25 m between the individual traps. The traps ran continuously from April to September 2020 and 2021 with a two-week collection interval to obtain phenological data and the potential to record species with short flight times (for details on the design see Lehmann et al. \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). The Malaise traps were produced in a standardized way by the Entomological Society Krefeld and installed in the field following a defined protocol (Hallmann et al. \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) to ensure data comparability with past and future insect monitoring studies. Flying insects were collected into 1000-ml polyethylene bottles which were emptied on average every 14 days.\u003c/p\u003e\u003cp\u003eThe data presented here were collected in a two-week interval in 2020 (May 16 to June 2, see Suppl. Table\u0026nbsp;3). Here we compare the results from 12 locations from this period for which full sample sets were available including the metabarcoding (K\u0026ouml;the \u0026amp; Schneider et al. 2023).\u003c/p\u003e\u003cp\u003eInsect samples collected in the Malaise traps were split into two parts by the Entomological Society Krefeld (EVK) using a specially developed sample divider (H\u0026ouml;rren et al. \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) and subjected to metabarcoding at the Alexander Koenig Zoological Research Museum (ZFMK; see K\u0026ouml;the \u0026amp; Schneider et al. 2023, Thomas et al. \u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e2025\u003c/span\u003e), while the second half was morphologically sorted at The Nature and Biodiversity Conservation Union (NABU). We selected three insect groups which are suitable bioindicators for determining the quality of grassland dominated habitat types.\u003c/p\u003e\u003c/div\u003e\n\u003ch3\u003eMorphological species identification\u003c/h3\u003e\n\u003cp\u003eThe sample subset stored for morphological analysis was sorted into insect orders, with the main groups being Diptera, Hymenoptera, Coleoptera, Lepidoptera, and Hemiptera (Lehmann GUC et al. data on file). Our three selected target groups (Diptera: Syrphidae, Hemiptera: Auchenorrhyncha, and Orthoptera) were sorted and morphologically identified by experts (Orthoptera, Syrphidae: Arne W. Lehmann, Auchenorrhyncha: Roland M\u0026uuml;hlethaler) under Stereomicroscopes of the types Citoval 2 (Zeiss Jena, Germany). Determinations were made with the support of standard literature (Auchenorrhyncha: Biedermann \u0026amp; Niedringhaus (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2004\u003c/span\u003e), Holzinger (\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2003\u003c/span\u003e), Orthoptera adults: Horstkotte et al. \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e1991\u003c/span\u003e, Orthoptera nymphs: Thommen et al. 2021, and Syrphidae: van Veen 2010). The private collections of Lehmann and M\u0026uuml;hlethaler were used for comparison and genital extractions (Syrphidae) or dissections (Auchenorrhyncha) were performed when necessary. Identified specimens were labelled and stored in separate jars. The material will be permanently deposited at the collections of the Entomological Society Krefeld, Germany. Species numbers and nature protection status for Germany were extracted from the newest Red Lists (Auchenorrhyncha: Nickel et al. \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2016\u003c/span\u003e, Orthoptera: Poniatowski et al. \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2024\u003c/span\u003e, Syrphidae: Ssymank et al. \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2011\u003c/span\u003e).\u003c/p\u003e\n\u003ch3\u003eInsect metabarcoding\u003c/h3\u003e\n\u003cp\u003eEach sample was fractioned into two size classes referred to as S (small, \u0026lt; 4 mm) or L (large,\u0026gt;4 mm) to avoid the underrepresentation of species with small biomass. These subsamples were dried until complete evaporation of ethanol, homogenized and lysated. Lysates of both size fractions were used and subsequently merged into constant proportions (90% of size class S with 10% of size class L (Elbrecht et al. \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). The eluted DNA was checked for quality on an agarose gel. 84 tissue samples were processed on a 96 well spin column plate and complemented with 12 negative controls (only lysis buffer (ATL, Qiagen) and 10% Proteinase K (Qiagen)). We applied a two-step PCR protocol using standard Illumina Nextera primers for dual indexing of samples, with fwhF2 forward (Vamos et al. \u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) and Fol_degen_rev reverse (Yu et al. \u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e2012\u003c/span\u003e) primer. With a Quantus fluorometer (Promega, Madison, USA) on a Fragment Analyzer (Agilent Technologies, Santa Clara, CA, USA) the library concentration was measured, and the pool was sent for sequencing on a Novaseq SP platform at CeGaT GmbH (T\u0026uuml;bingen, Germany). Data analysis was performed on merged paired-end reads with a length of 303\u0026ndash;323 bp as implemented in JAMP v0.78 (Elbrecht \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). The taxonomic assignment of the molecular units was carried out by comparison with a reference database for Arthropoda, generated by a beta version of the Taxalogue (Noll et al. \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) with sequences from BOLD (Barcode of Life Data System) (Ratnasingham and Hebert \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2007\u003c/span\u003e), NCBI GenBank (Clark et al. \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2016\u003c/span\u003e) and GBOL (German Barcode of Life) (Geiger et al. \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2016\u003c/span\u003e) and at least 85% sequence similarity.\u003c/p\u003e\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\u003ch2\u003eValidation against the standard of morphological species identification\u003c/h2\u003e\u003cp\u003eTo compile the consensus species lists the metabarcoding results was filtered to include only results for the order Orthoptera, the suborder Auchenorrhyncha (Hemiptera) and the family Syrphidae (Diptera). OTUs assigned to species were merged but kept in the consensus list for clarity (Suppl.Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Results from the metabarcoding and morphological identifications were combined with three foci (the single Malaise trap, the combined five traps of a transect of a single area, and for all 12 areas across Germany). OTUs that only allowed identification up to the genus, family or order level were also summarized (Suppl.Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec9\" class=\"Section2\"\u003e\u003ch2\u003eStatistical analysis\u003c/h2\u003e\u003cp\u003eThe statistical analysis was performed with Rx64 4.4.2\u0026ndash;4.5.1 (R Core Team, 2024-25) embedded in RStudio (RStudio Team 2024-25). We used the R programs for Linear mixed effect models (lmer: Kuznetsova et al. \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2017\u003c/span\u003e), Generalized Additive Models (mgcv: Wood \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) and ranking of abundance curves (matrixStats: Bengtsson et al. \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eBinary model classification was done with a confusion matrix (caret: Kuhn \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2008\u003c/span\u003e). It is based on a 2 x 2 matrix with the correctly predicted positive (True Positive TP\u0026thinsp;=\u0026thinsp;identified by both morphology and metabarcoding), and negative instances (True negatives TN\u0026thinsp;=\u0026thinsp;neither identified by morphology nor metabarcoding), and the type I error of incorrectly predicted positive (False Positive FP) and the type II error of positive identifications incorrectly classified as negative (False Negative FN). Based on these four cases several scores are established, with sensitivity (TP / TP\u0026thinsp;+\u0026thinsp;FN) being the true positive rate and precision (TP / TP\u0026thinsp;+\u0026thinsp;FP) the positive predicted value. The F1 score is a metric for evaluating the performance of classification models without considering the true negatives. It is the harmonic mean of precision and sensitivity (F1 score\u0026thinsp;=\u0026thinsp;2 * (Precision*Sensitivity) / (Precision\u0026thinsp;+\u0026thinsp;Sensitivity)), providing a balance between the two metrics. The F1 score ranges from 0 to 1, where 1 indicates perfect precision and sensitivity, and 0 the worst performance (Grandini et al. \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2020\u003c/span\u003e.\u003c/p\u003e\u003c/div\u003e"},{"header":"Declarations","content":"\u003ch2\u003eCompeting interests\u003c/h2\u003e\u003cp\u003eThe author(s) declare no competing interests.\u003c/p\u003e\u003ch2\u003eFunding\u003c/h2\u003e\u003cp\u003eThis study was funded within the DINA project (Diversity of Insects in Nature protected Areas) by the Federal Ministry of Education and Research (BMBF) under the grant number FKZ 01LC1901 and handled by the VDI project management organization. The BfN (Bundesamt f\u0026uuml;r Naturschutz) financially supported the analysis of the probes and the determination of some taxa.\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eR.M. contributed substantially to the conception of the work, acquired data, interpreted the data, drafted the work and substantially revised it; A.W.L. contributed to the design of the work, acquired data, analyzed and interpreted the data, drafted the work and substantially revised it; S.K. contributed substantially to the conception of the work, acquired data, analyzed the data and drafted the work; H.B. analyzed the data, interpreted the data, and substantially revised the work; G.U.C.L. contributed substantially to the conception and design of the work, acquired, analyzed and interpreted the data, and substantially revised the work. All authors have approved the submitted version and agreed to be personally accountable.\u003c/p\u003e\u003ch2\u003eAcknowledgement\u003c/h2\u003e\u003cp\u003eWe thank the Federal Ministry of Education and Research (BMBF) for founding the DINA project and the VDI for project management organization for handling (grant number FKZ 01LC1901). The BfN (Bundesamt f\u0026uuml;r Naturschutz) gave financial support for the analysis of the probes and the determination of some taxa, we are grateful. Our study very much benefitted from workshop discussions in the sMon project (Trend analysis of biodiversity data in Germany), a strategic project of the German Centre for Integrative Biodiversity Research (iDiv) Halle- Jena- Leipzig. We would also like to thank Michael R. Wilson (UK) for his linguistic revision and comments on the manuscript. Our gratitude goes to the farmers and authorities for granting licenses to set up Malaise traps and take samples. Without the dedicated support of our more than 45 volunteers and citizen scientists caring for the Malaise traps, we could not have carried out the project.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eThe metabarcoding data are already published and available at the Digital Cataloque of the Leibniz Institute for the Analysis of Biodiversity Change (https://collections.leibniz-lib.de) under the accession numbers listed in TableS11 of Thomas et al. (2025).All data are presented in the Supplementary material section.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eArribas, P., And\u0026uacute;jar, C., Salces-Castellano, A., Emerson, B. C. \u0026amp; Vogler, A. P. The limited spatial scale of dispersal in soil arthropods revealed with whole-community haplotype-level metabarcoding. \u003cem\u003eMol. Ecol.\u003c/em\u003e \u003cb\u003e30\u003c/b\u003e (1), 48\u0026ndash;61. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/mec.15591\u003c/span\u003e\u003cspan address=\"10.1111/mec.15591\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2021).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBengtsson, H. et al. matrixStats: Functions that Apply to Rows and Columns of Matrices (and to Vectors)1.5.0. (2025). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.32614/CRAN.package.matrixStats\u003c/span\u003e\u003cspan address=\"10.32614/CRAN.package.matrixStats\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBiedermann, R. \u0026amp; Niedringhaus, R. \u003cem\u003eDie Zikaden Deutschlands: Bestimmungstafeln f\u0026uuml;r alle Arten\u003c/em\u003e (WABV-Fr\u0026uuml;nd, 2004).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBuchner, D. et al. Upscaling biodiversity monitoring: Metabarcoding estimates 31,846 insect species from Malaise traps across Germany. \u003cem\u003eMol. Ecol. Resour.\u003c/em\u003e \u003cb\u003e25\u003c/b\u003e, e14023. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/1755-0998.14023\u003c/span\u003e\u003cspan address=\"10.1111/1755-0998.14023\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2025).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChua, P. Y. S. et al. Future of DNA-based insect monitoring. \u003cem\u003eTrends Genet.\u003c/em\u003e \u003cb\u003e39\u003c/b\u003e, 531\u0026ndash;544. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.tig.2023.02.012\u003c/span\u003e\u003cspan address=\"10.1016/j.tig.2023.02.012\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eClark, K., Karsch-Mizrachi, I., Lipman, D. J., Ostell, J. \u0026amp; Sayers, E. W. GenBank. \u003cem\u003eNucleic Acids Res.\u003c/em\u003e \u003cb\u003e44\u003c/b\u003e (D1), D67\u0026ndash;72. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/nar/gkv1276\u003c/span\u003e\u003cspan address=\"10.1093/nar/gkv1276\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2016).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCreedy, T. J. et al. Coming of age for COI metabarcoding of whole organism community DNA: towards bioinformatic harmonisation. \u003cem\u003eMol. Ecol. Resour.\u003c/em\u003e \u003cb\u003e22\u003c/b\u003e, 847\u0026ndash;861. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/1755-0998.13502\u003c/span\u003e\u003cspan address=\"10.1111/1755-0998.13502\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2022).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eEggleton, P. The state of the World\u0026rsquo;s insects. \u003cem\u003eAnnu. Rev. Environ. Resour.\u003c/em\u003e \u003cb\u003e45\u003c/b\u003e, 61\u0026ndash;82. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1146/annurev-environ-012420-050035\u003c/span\u003e\u003cspan address=\"10.1146/annurev-environ-012420-050035\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eElbrecht, V. JAMP v.0.78. (2019). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://github.com/VascoElbrecht/JAMP\u003c/span\u003e\u003cspan address=\"https://github.com/VascoElbrecht/JAMP\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eElbrecht, V. et al. Pooling size sorted Malaise trap fractions to maximize taxon recovery with metabarcoding. \u003cem\u003ePeerJ\u003c/em\u003e \u003cb\u003e9\u003c/b\u003e, e12177. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.7717/peerj.12177\u003c/span\u003e\u003cspan address=\"10.7717/peerj.12177\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2021).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eErshova, E. A., Wangensteen, O. S. \u0026amp; Falkenhaug, T. Mock samples resolve biases in diversity estimates and quantitative interpretation of zooplankton metabarcoding data. \u003cem\u003eMarine Biodivers.\u003c/em\u003e \u003cb\u003e53\u003c/b\u003e, 66. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s12526-023-01372-x\u003c/span\u003e\u003cspan address=\"10.1007/s12526-023-01372-x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eF\u0026ouml;rster, T., Creutzburg, F., Anton, E., Weigel, A. \u0026amp; Hartmann, M. Metabarcoding versus morphologische Identifizierung: der Herausforderung. \u003cem\u003egewachsen? Entomol. Z.\u003c/em\u003e \u003cb\u003e133\u003c/b\u003e, 103\u0026ndash;116 (2023).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGeiger, M. F. et al. How to tackle the molecular species inventory for an industrialized nation\u0026mdash;lessons from the first phase of the German Barcode of Life initiative GBOL (2012\u0026ndash;2015). \u003cem\u003eGenome\u003c/em\u003e \u003cb\u003e59\u003c/b\u003e, 661\u0026ndash;670. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1139/gen-2015-0185\u003c/span\u003e\u003cspan address=\"10.1139/gen-2015-0185\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2016).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGrandini, M., Bagli, E. \u0026amp; Visani, G. Metrics for multi-class classification: an overview. arXiv preprint arXiv:2008.05756. (2020).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHallmann, C. A. et al. More than 75 percent decline over 27 years in total flying insect biomass in protected areas. \u003cem\u003ePLoS One\u003c/em\u003e. \u003cb\u003e12\u003c/b\u003e, e0185809. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1371/journal.pone.0185809\u003c/span\u003e\u003cspan address=\"10.1371/journal.pone.0185809\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2017).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHallmann, C. A., Ssymank, A., Sorg, M., de Kroon, H. \u0026amp; Jongejans, E. Insect biomass decline scaled to species diversity: General patterns derived from a hoverfly community. \u003cem\u003ePNAS\u003c/em\u003e \u003cb\u003e118\u003c/b\u003e (2), e2002554117. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1073/pnas.2002554117\u003c/span\u003e\u003cspan address=\"10.1073/pnas.2002554117\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2021).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHallmann, C. A. et al. Weather anomalies cannot explain insect decline. \u003cem\u003eNature\u003c/em\u003e \u003cb\u003e639\u003c/b\u003e, E7\u0026ndash;E11. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41586-024-08528-0\u003c/span\u003e\u003cspan address=\"10.1038/s41586-024-08528-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2025).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHawlitschek, O. et al. DNA barcoding of crickets, katydids, and grasshoppers (Orthoptera) from Central Europe with focus on Austria, Germany, and Switzerland. \u003cem\u003eMol. Ecol. Resour.\u003c/em\u003e \u003cb\u003e17\u003c/b\u003e, 1037\u0026ndash;1053. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/1755-0998.12638\u003c/span\u003e\u003cspan address=\"10.1111/1755-0998.12638\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2017).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHawlitschek, O. et al. New estimates of genome size in Orthoptera and their evolutionary implications. \u003cem\u003ePLoS One\u003c/em\u003e. \u003cb\u003e18\u003c/b\u003e (3), e0275551 (2023).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHawthorne, B. S. J., Cuff, J. P., Collins, L. E. \u0026amp; Evans, D. M. Metabarcoding advances agricultural invertebrate biomonitoring by enhancing resolution, increasing throughput and facilitating network inference. \u003cem\u003eAgric. For. Entomol.\u003c/em\u003e \u003cb\u003e27\u003c/b\u003e (1), 50\u0026ndash;66. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/afe.12628\u003c/span\u003e\u003cspan address=\"10.1111/afe.12628\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2025).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHebert, P. D. N., Cywinska, A., Ball, S. L. \u0026amp; de Waard, J. R. Biological identifications through DNA barcodes. Proceedings of the Royal Society B: Biological Sciences 270: 313\u0026ndash;321. (2003). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1098/rspb.2002.2218\u003c/span\u003e\u003cspan address=\"10.1098/rspb.2002.2218\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHebert, P. D. N., Bock, D. G. \u0026amp; Prosser, S. W. J. Interrogating 1000 insect genomes for NUMTs: a risk assessment for estimates of species richness. \u003cem\u003ePLoS One\u003c/em\u003e. \u003cb\u003e18\u003c/b\u003e (6), e0286620. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1371/journal.pone.0286620\u003c/span\u003e\u003cspan address=\"10.1371/journal.pone.0286620\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHochkirch, A. et al. \u003cem\u003eEuropean Red List of Insect Taxonomists\u003c/em\u003e (Publication Office of the European Union, 2022).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHolzinger, W. E. The Auchenorrhyncha of Central Europe - Die Zikaden Mitteleuropas. Brill (2003).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eH\u0026ouml;rren, T. et al. Development of an insect sample fractionizer for biodiversity research. bioRxiv 2022.11.04.515206. (2022). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1101/2022.11.04.515206\u003c/span\u003e\u003cspan address=\"10.1101/2022.11.04.515206\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHorstkotte, J., Lorenz, C. \u0026amp; Wendler, A. Heuschrecken - Bestimmung, Verbreitung, Lebensr\u0026auml;ume und Gef\u0026auml;hrdung aller in Deutschland vorkommenden Arten. 9ed. DJN-Hamburg (1991).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eIPBES \u0026amp; the global assessment report on biodiversity and ecosystem services of the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services. : Summary for policymakers of. D\u0026iacute;az S, Settele J, Brond\u0026iacute;zio ES, Ngo HT, Gu\u0026egrave;ze M,Agard J, Arneth A, Balvanera P, Brauman KA, Butchart SHM, Chan KMA, Garibaldi LA, Ichii K, Liu J, Subramanian SM, Midgley GF, Miloslavich P, Moln\u0026aacute;r Z, Obura D, Pfaff A, Polasky S, Purvis A, Razzaque J, Reyers B, Chowdhury RR, Shin YJ, Visseren-Hamakers IJ, Willis KJ, Zayas CN (eds.) IPBES secretariat, Bonn, Germany. (2019). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.5281/zenodo.3553579\u003c/span\u003e\u003cspan address=\"10.5281/zenodo.3553579\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eIwaszkiewicz-Eggebrecht, E., Zizka, V. \u0026amp; Lynggaard, C. Three steps towards comparability and standardization among molecular methods for characterizing insect communities. \u003cem\u003ePhilosophical Trans. Royal Soc. B: Biol. Sci.\u003c/em\u003e \u003cb\u003e379\u003c/b\u003e, 20230118. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1098/rstb.2023.0118\u003c/span\u003e\u003cspan address=\"10.1098/rstb.2023.0118\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2024).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eJones, J. I. et al. Sources of uncertainty in DNA metabarcoding of whole communities: Implications for its use in biomonitoring. \u003cem\u003eMethods Ecol. Evol.\u003c/em\u003e \u003cb\u003e16\u003c/b\u003e, 1658\u0026ndash;1673. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/2041-210X.70093\u003c/span\u003e\u003cspan address=\"10.1111/2041-210X.70093\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2025).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKarlsson, D. et al. The Swedish Malaise Trap Project: A 15 Year Retrospective on a Countrywide Insect Inventory. \u003cem\u003eBiodivers. Data J.\u003c/em\u003e \u003cb\u003e8\u003c/b\u003e, e47255. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3897/bdj.8.e47255\u003c/span\u003e\u003cspan address=\"10.3897/bdj.8.e47255\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKortmann, M. et al. A shortcut to sample coverage standardization in metabarcoding data provides new insights into land-use effects on insect diversity. Proceedings of the Royal Society B: Biological Sciences 292:20242927. (2025). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://doi.org/10.1098/rspb.2024.2927\u003c/span\u003e\u003cspan address=\"10.1098/rspb.2024.2927\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eK\u0026ouml;the, S. et al. Improving insect conservation management through insect monitoring and stakeholder involvement. \u003cem\u003eBiodivers. Conserv.\u003c/em\u003e \u003cb\u003e32\u003c/b\u003e, 691\u0026ndash;713. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10531-022-02519-1\u003c/span\u003e\u003cspan address=\"10.1007/s10531-022-02519-1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKuhn, M. Building Predictive Models in R Using the caret Package. \u003cem\u003eJ. Stat. Softw.\u003c/em\u003e \u003cb\u003e28\u003c/b\u003e (5), 1\u0026ndash;26. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.18637/jss.v028.i05\u003c/span\u003e\u003cspan address=\"10.18637/jss.v028.i05\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2008).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKuznetsova, A., Brockhoff, P. B. \u0026amp; Christensen, R. H. B. lmerTest Package: Tests in Linear Mixed Effects Models. \u003cem\u003eJ. Stat. Softw.\u003c/em\u003e \u003cb\u003e82\u003c/b\u003e (13), 1\u0026ndash;26. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.18637/jss.v082.i13\u003c/span\u003e\u003cspan address=\"10.18637/jss.v082.i13\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2017).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLehmann, G. U. C. et al. Diversity of Insects in Nature protected Areas (DINA): an interdisciplinary German research project. \u003cem\u003eBiodivers. Conserv.\u003c/em\u003e \u003cb\u003e30\u003c/b\u003e, 2605\u0026ndash;2614. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10531-021-02209-4\u003c/span\u003e\u003cspan address=\"10.1007/s10531-021-02209-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2021).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLi, Y. et al. Combining environmental DNA and remote sensing for efficient, fine-scale mapping of arthropod biodiversity. \u003cem\u003ePhilosophical Trans. Royal Soc. B: Biol. Sci.\u003c/em\u003e \u003cb\u003e379\u003c/b\u003e, 20230123. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1098/rstb.2023.0123\u003c/span\u003e\u003cspan address=\"10.1098/rstb.2023.0123\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2024).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLiu, X. et al. Genomic and transcriptomic perspectives on the origin and evolution of NUMTs in Orthoptera. \u003cem\u003eMol. Phylogenet. Evol.\u003c/em\u003e \u003cb\u003e201\u003c/b\u003e, 108221. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.ympev.2024.108221\u003c/span\u003e\u003cspan address=\"10.1016/j.ympev.2024.108221\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2024).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLuo, M., Ji, Y., Warton, D. \u0026amp; Yu, D. W. Extracting abundance information from DNA -based data. \u003cem\u003eMol. Ecol. Resour.\u003c/em\u003e \u003cb\u003e23\u003c/b\u003e, 174\u0026ndash;189. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/1755-0998.13703\u003c/span\u003e\u003cspan address=\"10.1111/1755-0998.13703\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMioduchowska, M., Czyż, M. J., Gołdyn, B., Kur, J. \u0026amp; Sell, J. Instances of erroneous DNA barcoding of metazoan invertebrates: Are universal \u003cem\u003ecox1\u003c/em\u003e gene primers too universal? \u003cem\u003ePLoS ONE\u003c/em\u003e. \u003cb\u003e13\u003c/b\u003e (6), e0199609. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1371/journal.pone.0199609\u003c/span\u003e\u003cspan address=\"10.1371/journal.pone.0199609\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2018).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eM\u0026uuml;hlethaler, R., Holzinger, W. E., Nickel, H. \u0026amp; Wachmann, E. \u003cem\u003eDie Zikaden Deutschlands, \u0026Ouml;sterreichs und der Schweiz\u003c/em\u003e (Quelle \u0026amp; Meyer, 2019).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eM\u0026uuml;hlethaler, R. et al. No recovery in the biomass of flying insects over the last decade in German nature protected areas. \u003cem\u003eEcol. Evol.\u003c/em\u003e \u003cb\u003e14\u003c/b\u003e, e11182. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/ece3.11182\u003c/span\u003e\u003cspan address=\"10.1002/ece3.11182\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2024).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eNeri, F. et al. Comparative analysis of phytoplankton diversity using microscopy and metabarcoding: insights from an eLTER station in the Northern Adriatic Sea. \u003cem\u003eHydrobiologia\u003c/em\u003e \u003cb\u003e852\u003c/b\u003e, 169\u0026ndash;183. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10750-024-05692-2\u003c/span\u003e\u003cspan address=\"10.1007/s10750-024-05692-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2024).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eNickel, H. et al. Rote Liste und Gesamtartenliste der Zikaden (Hemiptera: Auchenorrhyncha) Deutschlands. In: Gruttke H, Balzer S, Binot-Hafke M, Haupt H, Hofbauer N, Ludwig G, Matzke-Hajek G, Ries M (Red.): Rote Liste gef\u0026auml;hrdeter Tiere, Pflanzen und Pilze Deutschlands, Band 4: Wirbellose Tiere (Teil 2). Naturschutz und Biologische Vielfalt 70(4): 249\u0026ndash;298. (2016).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eNoguerales, V. et al. Community metabarcoding reveals the relative role of environmental filtering and spatial processes in metacommunity dynamics of soil microarthropods across a mosaic of montane forests. \u003cem\u003eMol. Ecol.\u003c/em\u003e \u003cb\u003e32\u003c/b\u003e, 6110\u0026ndash;6128. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/mec.16275\u003c/span\u003e\u003cspan address=\"10.1111/mec.16275\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eNoll, N. W., Scherber, C. \u0026amp; Sch\u0026auml;ffler, L. taxalogue: a toolkit to create comprehensive CO1 reference databases. \u003cem\u003ePeerJ\u003c/em\u003e \u003cb\u003e11\u003c/b\u003e, e16253. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.7717/peerj.16253\u003c/span\u003e\u003cspan address=\"10.7717/peerj.16253\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePenel, B. et al. The best of two worlds: toward large-scale monitoring of biodiversity combining COI metabarcoding and optimized parataxonomic validation. \u003cem\u003eEcography\u003c/em\u003e \u003cb\u003e2025\u003c/b\u003e (6), e07699. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/ecog.07699\u003c/span\u003e\u003cspan address=\"10.1111/ecog.07699\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2025).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePetsopoulos, D. et al. Identifying archived insect bulk samples using DNA metabarcoding: A case study using the long-term Rothamsted Insect Survey. \u003cem\u003eEnviron. DNA\u003c/em\u003e. \u003cb\u003e6\u003c/b\u003e (3), e542. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/edn3.542\u003c/span\u003e\u003cspan address=\"10.1002/edn3.542\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2024).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePoniatowski, D. et al. Rote Liste und Gesamtartenliste der Heuschrecken und Fangschrecken (Orthoptera et Mantodea) Deutschlands. \u003cem\u003eNaturschutz und Biologische Vielfalt\u003c/em\u003e. \u003cb\u003e170\u003c/b\u003e (7), 1\u0026ndash;88 (2024).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRatnasingham, S. \u0026amp; Hebert, P. D. N. bold: The Barcode of Life Data System. \u003cem\u003eMol. Ecol. Notes\u003c/em\u003e. \u003cb\u003e7\u003c/b\u003e, 355\u0026ndash;364. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/j.1471-8286.2007.01678.x\u003c/span\u003e\u003cspan address=\"10.1111/j.1471-8286.2007.01678.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2007). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.barcodinglife.org\u003c/span\u003e\u003cspan address=\"http://www.barcodinglife.org\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCore Team, R. (2024\u0026ndash;2025) R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. \u0026lt;\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.R-project.org/\u003c/span\u003e\u003cspan address=\"https://www.R-project.org/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRemmel, N. et al. DNA metabarcoding and morphological identification reveal similar richness, taxonomic composition and body size patterns among flying insect communities. \u003cem\u003eInsect Conserv. Divers.\u003c/em\u003e \u003cb\u003e17\u003c/b\u003e (3), 449\u0026ndash;463. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/icad.12710\u003c/span\u003e\u003cspan address=\"10.1111/icad.12710\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2024).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRStudio, T. \u003cem\u003e-25) RStudio: Integrated Development Environment for R (Version 2024.12.0\u0026ndash;2025.05.1) [Computer software]\u003c/em\u003e (RStudio, PBC, 2024). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.rstudio.com/\u003c/span\u003e\u003cspan address=\"http://www.rstudio.com/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSalis, R., Sunde, J., Gubonin, N., Franz\u0026eacute;n, M. \u0026amp; Forsman, A. Performance of DNA metabarcoding, standard barcoding and morphological approaches in the identification of insect biodiversity. \u003cem\u003eMol. Ecol. Resour.\u003c/em\u003e \u003cb\u003e24\u003c/b\u003e, e14018. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/1755-0998.14018\u003c/span\u003e\u003cspan address=\"10.1111/1755-0998.14018\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2024).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eShaffer, M. R. et al. Observation Bias in Metabarcoding. Molecular Ecology Resources, early view:e14119. (2025). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/1755-0998.14119\u003c/span\u003e\u003cspan address=\"10.1111/1755-0998.14119\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSkvarla, M. J., Larson, J. L., Fisher, J. R. \u0026amp; Dowling, A. P. G. A review of terrestrial and canopy Malaise traps. \u003cem\u003eAnn. Entomol. Soc. Am.\u003c/em\u003e \u003cb\u003e114\u003c/b\u003e (1), 27\u0026ndash;47. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1093/aesa/saaa044\u003c/span\u003e\u003cspan address=\"10.1093/aesa/saaa044\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2021).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSsymank, A., Doczkal, D., Rennwald, K. \u0026amp; Dziock, F. Rote Liste und Gesamtartenliste der Schwebfliegen (Diptera: Syrphidae) Deutschlands. In: Binot-Hafke M, Balzer S, Becker N, Gruttke H, Haupt H, Hofbauer N, Ludwig G, Matzke-Hajek G, Strauch M (ed.): Rote Liste der gef\u0026auml;hrdeten Tiere, Pflanzen und Pilze Deutschlands. Band 3: Wirbellose Tiere (Teil 1). Bonn (Bundesamt f\u0026uuml;r Naturschutz). Naturschutz und Biologische Vielfalt 70 (3): 13\u0026ndash;83. (2011).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eStrutzenberger, P. et al. DNA metabarcoding of light trap samples vs. morphological species identification. \u003cem\u003eEcol. Entomol.\u003c/em\u003e \u003cb\u003e49\u003c/b\u003e (2), 245\u0026ndash;256. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/een.13297\u003c/span\u003e\u003cspan address=\"10.1111/een.13297\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2024).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSvenning, A. et al. A General Method for Detection and Segmentation of Terrestrial Arthropods in Images. bioRxiv 2025.04.08.647223. (2025). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1101/2025.04.08.647223\u003c/span\u003e\u003cspan address=\"10.1101/2025.04.08.647223\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTaberlet, P., Coissac, E., Pompanon, F., Brochmann, C. \u0026amp; Willerslev, E. Towards next-generation biodiversity assessment using DNA metabarcoding. \u003cem\u003eMol. Ecol.\u003c/em\u003e \u003cb\u003e21\u003c/b\u003e, 2045\u0026ndash;2050. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/j.1365-294X.2012.05470.x\u003c/span\u003e\u003cspan address=\"10.1111/j.1365-294X.2012.05470.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2012).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eThomas, L. J. et al. Identifying conservation hotspots and assessing species commonness and rarity: Baseline arthropod diversity in German nature reserves via national Malaise trap monitoring. \u003cem\u003eInsect Conserv. Divers.\u003c/em\u003e \u003cb\u003e18\u003c/b\u003e (4), 510\u0026ndash;525. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/icad.12818\u003c/span\u003e\u003cspan address=\"10.1111/icad.12818\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2025).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eThommen, D. \u003cem\u003eJugendstadien der Heuschrecken der Schweiz - Der Bestimmungsf\u0026uuml;hrer f\u0026uuml;r Heuschrecken-Nymphen\u003c/em\u003e (Haupt, 2021).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eVamos, E., Elbrecht, V. \u0026amp; Leese, F. Short COI markers for freshwater macroinvertebrate metabarcoding. \u003cem\u003eMetabarcoding Metagenomics\u003c/em\u003e. \u003cb\u003e1\u003c/b\u003e, e14625. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3897/mbmg.1.14625\u003c/span\u003e\u003cspan address=\"10.3897/mbmg.1.14625\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2017).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003evan Klink, R. et al. Meta-analysis reveals declines in terrestrial but increases in freshwater insect abundances. \u003cem\u003eScience\u003c/em\u003e \u003cb\u003e368\u003c/b\u003e, 417\u0026ndash;420. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1126/science.aax9931\u003c/span\u003e\u003cspan address=\"10.1126/science.aax9931\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003evan Veen, M. P. Hoverflies of Northwest Europe: identification keys to the Syrphidae. 2nd edition. KNNV uitgeverij. (2020). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.13140/RG.2.1.3697.4567\u003c/span\u003e\u003cspan address=\"10.13140/RG.2.1.3697.4567\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWeydmann-Zwolicka, A., Dąbrowska, A. M., Mioduchowska, M. \u0026amp; Zwolicki, A. Comparison of DNA metabarcoding and microscopy in analysing planktonic protists from the European Arctic. \u003cem\u003eMarine Biodivers.\u003c/em\u003e \u003cb\u003e54\u003c/b\u003e, 44. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s12526-024-01436-6\u003c/span\u003e\u003cspan address=\"10.1007/s12526-024-01436-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2024).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWirth, C., Bruelheide, H., Farwig, N., Marx, J. M. \u0026amp; Settele, J. \u003cem\u003eFaktencheck Artenvielfalt. Bestandsaufnahme und Perspektiven f\u0026uuml;r den Erhalt der biologischen Vielfalt in Deutschland\u003c/em\u003e (oekom, 2024).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWolany, L., Enss, J., Leese, F. \u0026amp; Buchner, D. Homogenization of insect bulk samples yields more comprehensive yet comparable biodiversity data than non-destructive lysis. \u003cem\u003eMetabarcoding Metagenomics\u003c/em\u003e. \u003cb\u003e8\u003c/b\u003e, e129814. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3897/mbmg.8.129814\u003c/span\u003e\u003cspan address=\"10.3897/mbmg.8.129814\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2024).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWood, S. : mgcv: Mixed GAM Computation Vehicle with Automatic Smoothness Estimation. (2017). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.32614/CRAN.package.mgcv\u003c/span\u003e\u003cspan address=\"10.32614/CRAN.package.mgcv\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eYu, D. W. et al. Biodiversity soup: metabarcoding of arthropods for rapid biodiversity assessment and biomonitoring. \u003cem\u003eMethods Ecol. Evol.\u003c/em\u003e \u003cb\u003e3\u003c/b\u003e, 613\u0026ndash;623. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/j.2041-210X.2012.00198.x\u003c/span\u003e\u003cspan address=\"10.1111/j.2041-210X.2012.00198.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2012).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZizka, V. M. A. et al. Repeated subsamples during DNA extraction reveal increased diversity estimates in DNA metabarcoding of Malaise traps. \u003cem\u003eEcol. Evol.\u003c/em\u003e \u003cb\u003e12\u003c/b\u003e, e9502. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/ece3.9502\u003c/span\u003e\u003cspan address=\"10.1002/ece3.9502\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2022).\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"biodiversity loss, endangered species, insect decline, nature conservation, Malaise traps, species identification","lastPublishedDoi":"10.21203/rs.3.rs-7899842/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7899842/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eMetabarcoding is becoming an increasing popular method for broadscale insect monitoring. However, to complement or replace traditional insect monitoring approaches, the reliability of metabarcoding has to be confirmed. Therefore, we have evaluated the accuracy of species identifications of metabarcoding against the standard of morphology with binary classification in a confusion matrix to measure model performance. Within 12 German nature protected areas using a transect of five traps, metabarcoding found a total of 15,107 OTUs. Identifying 3096 individuals of three selected taxa by morphology, we found in total 151 species, compared with 130 species by metabarcoding. Species numbers for single traps differed substantially between methods, with a significant correspondence only found for Syrphidae (Diptera). Congruence at the species level was low, with sensitivity and precision below 50%, and even lower for Red List taxa. The match increased when aggregating traps and sites across Germany, or when comparing genera instead of species. Abundance curves strongly differed for species identified by both methods. So far, metabarcoding is lacking the necessary accuracy at the species level. We discuss possible causes for these inconsistencies and make suggestions for improvement for metabarcoding identification.\u003c/p\u003e","manuscriptTitle":"Paving the way for improved insect metabarcoding","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-11-22 15:07:07","doi":"10.21203/rs.3.rs-7899842/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"eef7196d-0a6b-42ea-866d-69e9f21025a3","owner":[],"postedDate":"November 22nd, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":58292477,"name":"Biological sciences/Ecology"},{"id":58292478,"name":"Earth and environmental sciences/Ecology"},{"id":58292479,"name":"Biological sciences/Evolution"},{"id":58292480,"name":"Biological sciences/Zoology"}],"tags":[],"updatedAt":"2025-11-27T05:53:10+00:00","versionOfRecord":[],"versionCreatedAt":"2025-11-22 15:07:07","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7899842","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7899842","identity":"rs-7899842","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}