A comprehensive understanding of trophic interactions in terrestrial ecosystems is crucial for ecological research and conservation. Recent advances in non-invasive methods, such as environmental DNA (eDNA) metabarcoding, have enabled researchers to collect vast amounts of data on wild animal diets. However, sharing this data and metadata effectively and transparently presents new challenges. To address this, a new type of scholarly journal publication has emerged that aims to describe datasets rather than report research investigations. In this paper, we present a dataset of consumed prey species and parasites based on the metabarcoding of 113 faecal samples from the greater and lesser mouse-eared bats (Myotis myotis and Myotis blythii), along with a detailed description of the data sampling, laboratory analysis, and bioinformatics pipeline. Our dataset comprises 1018 unique Barcode Index Numbers (BINs) from 12 Classes and 43 Orders. In addition, we provide interactive Krona charts to visually summarize the taxonomic relationships and relative read abundance of the consumed prey species and parasites. This data can be used for meta-analysis, exploring new predator-prey and host-parasite interactions, studying inter- and intraspecific ecological interactions, and informing protected area management, among other applications. By sharing this dataset, we hope to encourage other researchers to use it to answer additional ecological questions and advance our understanding of trophic interactions in terrestrial ecosystems.

Bats (Chiroptera), metabarcoding, Myotis myotis, Myotis blythii, parasite-host interactions, predator-prey interactions

Bats play a crucial role in terrestrial ecosystems worldwide by occupying various ecological niches and exploiting a range of food sources including insects, vertebrates, blood, nectar, pollen and fruit (Simmons 2005; Kunz et al. 2011). Due to their ecological abundance, bats could be used as bioindicator species that provide quantitative information on the quality of ecosystems, enabling the tracking of environmental alterations (Russo et al. 2021). Therefore, high-quality and detailed information on the diets of bat species in areas of conservation importance is essential. Historically, bat diets have been studied through the morphological analysis of invasively collected gut content or of noninvasively collected faeces (Whitaker et al. 2009). These morphological methods are time-consuming, require specialized entomological knowledge, and often can only identify prey down to order level. In recent years, advances in DNA barcoding and metabarcoding using high-throughput sequencing revolutionized the study of animal diets by providing a powerful, accurate, and time- and cost-efficient tool that can often identify prey down to the species level. As a consequence, after 2012 metabarcoding became increasingly popular (Ando et al. 2020). However, the methodology is still developing, and mistakes can occur at any level from sample collection, through PCR amplification and the bioinformatics pipeline, to data interpretation (Alberdi et al. 2018; O’Rourke et al. 2020). Thus, the sharing of transparent, detailed and open protocols and data is crucial for promoting good practices and avoiding errors. Here, we provide the full methods for collecting, metabarcoding, and analyzing a dataset of faecal samples collected from the greater and the lesser mouse-eared bats (Myotis myotis and Myotis blythii) in Bulgaria.

Myotis myotis (Borkhausen, 1797) and Myotis blythii s.l. (Tomes, 1857; for summary and discussion on taxonomy and phylogeny of the species see Ruedi 2020), family Vespertilionidae, are closely related sibling bat species that coexist in parts of Europe and the Middle East. While both are listed as Least Concerned in IUCN (Coroiu et al. 2016; Juste and Paunović 2016), in Europe they are protected under the Agreement on the Conservation of Populations of European Bats (EUROBATS). In Bulgaria, both species are listed as Near Threatened in the Red List of Protected Species of the country (Golemanski and Peev 2015). On the Balkan Peninsula, both bat species live almost exclusively in caves, or on rare occasions in mines, where they form mixed maternity and hibernation colonies. Interspecific competition is avoided mainly by habitat selection and different foraging strategies (Arlettaz 1999). M. myotis predominantly forages in forests and agricultural land with open, accessible ground (Arlettaz 1999; Stidsholt et al. 2023). M. blythii, in contrast, tends to forage in steppe-like habitats with dense grass and small shrubs (Arlettaz 1999). Both species use echolocation for aerial hunting as well as passive listening for gleaning prey off the ground and vegetation (Arlettaz et al. 2001; Siemers and Güttinger 2006; Stidsholt et al. 2023). However, when gleaning, M. myotis listens for the rustling sounds of large walking prey (Siemers and Güttinger 2006) while M. blythii eavesdrops on the mating song of bushcrickets (Jones et al. 2011).

The diet of M. myotis is extensively studied throughout its range with morphological methods (Audet 1990; Beck 1995; Arlettaz et al. 1997a, b; Arlettaz 1999; Pereira et al. 2002; Zahn et al. 2006; Steck and Güttinger 2006; Graclik and Wasielewski 2012), while less is known about the diet of M. blythii (Arlettaz et al. 1997a, b; Arlettaz 1999). These studies cover well the geographical and ecological variation in the diet of the two bat species, but they have a low taxonomic resolution and mostly identify prey only down to the order. Recent metabarcoding studies provided higher taxonomic resolution, however, only two studies investigated the diet of M. myotis (Galan et al. 2018; Alberdi et al. 2020), and only one study the diet of M. blythii (Mata et al. 2021), which additionally sampled and analyzed both species as one species complex due to methodological limitations.

Here, we provide a detailed description of the metabarcoding analysis of the faeces and of the diet of 113 individual bats (60 M. myotis and 53 M. blythii) collected in an area with high biodiversity value (Cimatti et al. 2021). The dataset contains 1018 Barcode Index Numbers (BIN) species from 12 classes and 43 orders, ranging from prey species to various ecto- and endo-parasites (including mites, fleas, tapeworms, roundworms, and others). The BIN System clusters sequences using well established algorithms to produce operational taxonomic units that closely correspond to species. Interestingly, the presence of molluscan and annelid species (such as Pomatias rivulare, Lumbricus rubellus, and Eisenia fetida) suggests that the data could contain information on further trophic interactions from prey species of predatory carabid beetles or other arthropods, that in turn have been consumed by the bats. The re-using potential of our data set varies from meta-analysis of the diets of insectivorous bats to investigations of predator-prey or host-parasite interactions and interspecific food webs and ecological interactions, and to the management of protected areas.

In summary, the dataset we present in this paper is a valuable resource that can aid in advancing ecological research and conservation efforts. We hope that by sharing our data, we can contribute to a more collaborative and transparent research environment that will lead to more effective conservation and management of terrestrial ecosystems.

The presented dataset is a comprehensive collection of 1018 BIN species belonging to 12 Classes and 43 Orders. The interactive Krona charts, based on BOLD taxonomy, provide a useful tool for visualizing the dataset (Fig. 2). These charts present the proportions of the sequence counts of respective species/taxonomic groups relative to all counts. However, it is essential to acknowledge that the sequence counts obtained from the samples do not necessarily correspond to the biomass of the original sample due to inherent methodological limitations and considerable variability in species sizes (Elbrecht et al. 2017; Lamb et al. 2019). Therefore, the relative read abundance (RRA, Deagle et al. 2019) presented in the Krona charts should only be considered a visual guide to the taxonomic diversity and relative abundance of the reads, and not an indication of the actual consumed biomass of the respective species.

Insects made up the largest proportion of the detected species. The observed differences in the RRA from carabid beetles between M. myotis (46%) and M. blythii (7%) aligns with the differences in diet, foraging style and habitat of these species (Arlettaz et al. 1997a, b; Arlettaz 1999). However, the differences in the RRA of other taxa shown in the Krona charts were less pronounced between the two bat species, with an important exception being Orthoptera. In M. myotis samples, the majority of Orthopteran species belonged to the family Gryllidae (crickets), while in M. blythii samples, the records primarily came from the family Tettigoniidae (bush-crickets), including species such as Tettigonia, Phaneroptera, Pholidoptera, Decticus, Poecilimon, and Isophya. This observation aligns with the known differences in foraging strategies between the two species, where M. myotis primarily utilizes rustling sounds produced by crickets, while M. blythii eavesdrops on the mating songs of bush-crickets (Arlettaz et al. 1997b; Jones et al. 2011; Stidsholt et al. 2023).

Importantly, since the RRA does not accurately represent actual biomass abundance, it is crucial to complement this data with other research techniques, such as biologging (Stidsholt et al. 2023), behavioral experiments (Stat et al. 2019; Jones et al. 2020) and ecological studies (Evans et al. 2016; Sato et al. 2021; Andriollo et al. 2021), to enhance the reliability and scope of the results. For instance, Stidsholt et al. (2023) utilized miniature airborne tags attached to wild foraging M. myotis individuals from the same colony as used in this study. The number of attacks were identified and categorized as either aerial or ground captures. Out of 3917 recorded prey attacks, approximately two-thirds were ground attacks. The results revealed that nearly all ground attacks were triggered by rustling sounds from carabid beetles, while fewer than 1% of ground attacks were initiated by mating calls of bush-crickets. These additional findings help place the results of the metabarcoding analysis into a more biologically relevant context.

Notably, in addition to the bats’ prey species, the provided dataset also includes reads from various ecto- and endo-parasites, such as ticks (Ixodes), mites (Mesostigmata and Sarcoptiformes), roundworms (Strongylida and Rhabditida), and other parasite species. Furthermore, we identified molluscs (Gastropoda) and worms (Annelida) in the samples, including Pomatias rivulare, Lumbricus rubellus, and Eisenia fetida, which were likely consumed by predatory carabid beetles or other arthropods that were then consumed by the bats. Moreover, the presence of species from the roundworm genus Steinernema, which are known to parasitize mole crickets and other bat prey, suggests that the dataset also contains parasites of the bats’ prey species. This comprehensive dataset thus offers valuable insights into the diversity and abundance of the parasites, the prey and their associated species of the greater and the lesser mouse-eared bats.

We thank Martin Georgiev, Theresa Hügel, Kathrin Dimitrova, and the entire field team of the Siemers Bat Research Station (field seasons 2017 and 2018) for their invaluable assistance with sample collection, methodological, logistical, and mental support. We acknowledge the Max Planck Institute for Ornithology and the National Museum of Natural History, Sofia for infrastructure and support.