Summary DNA metabarcoding holds great promise for the assessment of macroinvertebrates in stream ecosystems. However, few large‐scale studies have compared the performance of DNA metabarcoding with that of routine morphological identification. We performed metabarcoding using four primer sets on macroinvertebrate samples from 18 stream sites across Finland. The samples were collected in 2013 and identified based on morphology as part of a Finnish stream monitoring program. Specimens were morphologically classified, following standardised protocols, to the lowest taxonomic level for which identification was feasible in the routine national monitoring. DNA metabarcoding identified more than twice the number of taxa than the morphology‐based protocol, and also yielded a higher taxonomic resolution. For each sample, we detected more taxa by metabarcoding than by the morphological method, and all four primer sets exhibited comparably good performance. Sequence read abundance and the number of specimens per taxon (a proxy for biomass) were significantly correlated in each sample, although the adjusted R2 values were low. With a few exceptions, the ecological status assessment metrics calculated from morphological and DNA metabarcoding datasets were similar. Given the recent reduction in sequencing costs, metabarcoding is currently approximately as expensive as morphology‐based identification. Using samples obtained in the field, we demonstrated that DNA metabarcoding can achieve comparable assessment results to current protocols relying on morphological identification. Thus, metabarcoding represents a feasible and reliable method to identify macroinvertebrates in stream bioassessment, and offers powerful advantage over morphological identification in providing identification for taxonomic groups that are unfeasible to identify in routine protocols. To unlock the full potential of DNA metabarcoding for ecosystem assessment, however, it will be necessary to address key problems with current laboratory protocols and reference databases.

Assessment of ecological status for the European Water Framework Directive (WFD) is based on "Biological Quality Elements" (BQEs), namely phytoplankton, benthic flora, benthic invertebrates and fish. Morphological identification of these organisms is a time-consuming and expensive procedure. Here, we assess the options for complementing and, perhaps, replacing morphological identification with procedures using eDNA, metabarcoding or similar approaches. We rate the applicability of DNA-based identification for the individual BQEs and water categories (rivers, lakes, transitional and coastal waters) against eleven criteria, summarised under the headlines representativeness (for example suitability of current sampling methods for DNA-based identification, errors from DNA-based species detection), sensitivity (for example capability to detect sensitive taxa, unassigned reads), precision of DNA-based identification (knowledge about uncertainty), comparability with conventional approaches (for example sensitivity of metrics to differences in DNA-based identification), cost effectiveness and environmental impact. Overall, suitability of DNA-based identification is particularly high for fish, as eDNA is a well-suited sampling approach which can replace expensive and potentially harmful methods such as gill-netting, trawling or electrofishing. Furthermore, there are attempts to replace absolute by relative abundance in metric calculations. For invertebrates and phytobenthos, the main challenges include the modification of indices and completing barcode libraries. For phytoplankton, the barcode libraries are even more problematic, due to the high taxonomic diversity in plankton samples. If current assessment concepts are kept, DNA-based identification is least appropriate for macrophytes (rivers, lakes) and angiosperms/macroalgae (transitional and coastal waters), which are surveyed rather than sampled. We discuss general implications of implementing DNA-based identification into standard ecological assessment, in particular considering any adaptations to the WFD that may be required to facilitate the transition to molecular data.

No abstract

Most animal species on Earth are insects, and recent reports suggest that their abundance is in drastic decline. Although these reports come from a wide range of insect taxa and regions, the evidence to assess the extent of the phenomenon is sparse. Insect populations are challenging to study, and most monitoring methods are labor intensive and inefficient. Advances in computer vision and deep learning provide potential new solutions to this global challenge. Cameras and other sensors can effectively, continuously, and noninvasively perform entomological observations throughout diurnal and seasonal cycles. The physical appearance of specimens can also be captured by automated imaging in the laboratory. When trained on these data, deep learning models can provide estimates of insect abundance, biomass, and diversity. Further, deep learning models can quantify variation in phenotypic traits, behavior, and interactions. Here, we connect recent developments in deep learning and computer vision to the urgent demand for more cost-efficient monitoring of insects and other invertebrates. We present examples of sensor-based monitoring of insects. We show how deep learning tools can be applied to exceptionally large datasets to derive ecological information and discuss the challenges that lie ahead for the implementation of such solutions in entomology. We identify four focal areas, which will facilitate this transformation: 1) validation of image-based taxonomic identification; 2) generation of sufficient training data; 3) development of public, curated reference databases; and 4) solutions to integrate deep learning and molecular tools.

18 19 1) DNA metabarcoding holds great promise for the assessment of macroinvertebrates in stream ecosystems. However, few 20 large-scale studies have compared the performance of DNA metabarcoding with that of routine morphological identification. 212) We performed metabarcoding using four primer sets on macroinvertebrate samples from 18 stream sites across Finland. 22The samples were collected in 2013 and identified based on morphology as part of a Finnish stream monitoring program. 23Specimens were morphologically classified, following standardised protocols, to the lowest taxonomic level for which 24 identification was feasible in the routine national monitoring. 253) DNA metabarcoding identified more than twice the number of taxa than the morphology-based protocol, and also yielded 26 a higher taxonomic resolution. For each sample, we detected more taxa by metabarcoding than by the morphological 27 method, and all four primer sets exhibited comparably good performance. Sequence read abundance and the number of 28 specimens per taxon (a proxy for biomass) were significantly correlated in each sample, although the adjusted R 2 were low. 29With a few exceptions, the ecological status assessment metrics calculated from morphological and DNA metabarcoding 30 datasets were similar. Given the recent reduction in sequencing costs, metabarcoding is currently approximately as 31 expensive as morphology-based identification. 32 4) Using samples obtained in the field, we demonstrated that DNA metabarcoding can achieve comparable assessment 33 results to current protocols relying on morphological identification. Thus, metabarcoding represents a feasible and reliable 34 method to identify macroinvertebrates in stream bioassessment, and offers powerful advantage over morphological 35 identification in providing identification for taxonomic groups that are unfeasible to identify in routine protocols. To unlock 36 the full potential of DNA metabarcoding for ecosystem assessment, however, it will be necessary to address key problems 37 with current laboratory protocols and reference databases. 38 39 40