Open-space bats are abundant and common in agricultural landscapes, and widely recognized for their effectiveness as natural suppressors of pest insects. However, there is currently a lack of comprehensive studies on the diets and diet partitioning of these bats, particularly in terms of ingested pest prey species. Here, we explored the diet and diet partitioning of open-space foraging bats using DNA metabarcoding. We hypothesized that open-space bats residing near intensive agricultural areas would primarily feed on pest insects. Additionally, considering their shared guild and foraging habitats, we anticipated the occurrence of diet partitioning as a mechanism of reducing competition. Our results demonstrate the effectiveness of DNA metabarcoding in bat diet analysis, revealing a high diet diversity and enabling prey species identification. The outcomes exposed 1,557 OTUs and at least 36 agricultural pest species were found in the diet of these aerial insectivorous bats, including Nilaparvata lugens, Nysius graminicola, Piezodorus hybneri, Mythimna impura, Sesamia inferens, and Spodoptera frugiperda, among others. This finding underscores the crucial role of the studied bat species in pest control across a variety of crops such as sugarcane, rice, and corn, among others. We also observed evidence of diet partitioning. Mops plicatus primarily preyed on Hemiptera (Pentatomidae and Delphacidae). Taphozous theobaldi predominantly fed on Lepidoptera (Noctuidae and Erebidae). Taphozous melanopogon primarily fed on Lepidoptera (Noctuidae, Erebidae, Crambidae, Pyralidae, Geometridae). Scotophilus heathii focused on Coleoptera (Hydrophilidae and Tenebrionidae). Niche overlap was high between T. theobaldi and T. melanopogon, while T. theobaldi also moderately overlapped with M. plicatus. Mops plicatus had lower degress of overlap with T. melanopogon and S. heathii. The lowest degrees of overlap were observed between S. heathii and T. melanopogon, as well as between S. heathii and T. theobaldi. Thus, these bats exhibited niche diet partitioning at the family level. Our study highlights the occurrence of diet overlap and partitioning among open-space bats, underscoring their crucial role as effective pest suppressors in Asian agricultural regions. Their widespread presence, sizable population, and consumption of key pests, including migratory species, emphasize their significance in pest suppression. Protecting these bats is essential for suppressing pest insects and aiding farmers in protecting their crop yields across the region.
Introduction
Studying wildlife-crop interactions is crucial for agricultural productivity and for understanding wildlife biology in anthropogenic landscapes (Federico et al., 2008; Aizpurua et al., 2018). Globally, pest animals can cause losses in annual crop production of more than 25% (Oerke, 2006). However, conventional agricultural practices often overlook the ecosystem services provided by pest-regulating animals. Integrated pest management (IPM) emphasizes the significance of biological pest control, aiming to reduce the use of pesticides while mitigating monetary losses from agricultural pests (Ehler, 2006).
Bats are recognized as highly effective natural suppressors of pest insects in agricultural landscapes (Boyles et al., 2011; Kunz et al., 2011; Maas et al., 2013). They provide essential ecosystem services by foraging on airborne insects above farmland, particularly when including crop-associated arthropods in their diet (Leelapaibul et al., 2005; Wanger et al., 2014; Maine and Boyles, 2015; Aizpurua et al., 2018; Srilopan et al., 2018). Open-space foraging insectivorous bats dominate local bat communities in agricultural areas and serve as excellent bioindicators due to their high trophic position in food webs, making them sensitive to anthropogenic factors (Kalko et al., 2008; Jones et al., 2009). Open-space bats are not only important in agricultural landscapes due to their adaptability to habitat fragmentation but also for providing valuable ecosystem services to farmers by consuming pest insects worldwide (Boyles et al., 2011; Kunz et al., 2011; Ghanem and Voigt 2012; Williams-Guillén et al., 2016). These bats are known for their high-altitude flights, which are associated with the migration of pest insects at high altitudes (Leelapaibul et al., 2005; McCracken et al., 2008; Krauel et al., 2018b; Srilopan et al., 2018; Nguyen et al., 2019). These bats have been reported to reduce production costs to crops like rice, corn, cotton, sugar cane, coffee, and cocoa (Wanger et al., 2014; Maine and Boyles, 2015; Puig-Montserrat et al., 2015). As these crops are usually grown in large continuous areas, open-space bats navigate these landscapes effectively (Schnitzler et al., 2003; Denzinger and Schnitzler, 2013).
In agricultural landscapes of central Thailand, bat activity is dominated by open-space species such as Mops plicatus, Taphozous melanopogon, T. theobaldi, and Scotophilus heathii (Suksai and Bumrungsri, 2020). These species forage in open spaces for airborne prey, and share similar wing morphologies with high aspect ratios and wing loadings (Zhu et al., 2012; Suksai and Bumrungsri, 2020). Bat species belonging to the same guild, such as open-space foragers, often exhibit niche differentiation in order to reduce competition (Denzinger et al., 2016). Ecological niche differentiation contributes to stable coexistence in the presence of competition, as it reduces the degree of competition for limited resources (Chesson, 2000). Previous studies have shown that local insectivorous bat species, especially closely related taxa, may differentiate along various niche dimensions, such as diet, to facilitate coexistence and reduce competition (Hooper and Brown, 1968; Siemers and Swift, 2006; Ruadreo et al., 2018). Limited data on dietary niches within the same bat guilds hinder our understanding of bat community partitioning (Bumrungsri et al., 2007). Detailed studies on resource use, particularly those that take advantage of new molecular approaches are needed to reveal insights into community structuring.
The recent advances in DNA-based dietary analysis provide detailed information on bat diets, capable of identifying individual pest species (Williams-Guillén et al., 2016). Improved methods of diet analysis may provide a clearer understanding of trophic resource partitioning among coexisting bat species (Razgour et al., 2011). Conventional techniques have sometimes failed to identify resource partitioning in morphologically similar bat species, but finer-scale evaluations of diet analysis can reveal distinct dietary niches (Jacobs and Barclay, 2009; Schoeman and Jacobs, 2011). DNA metabarcoding of bat faecal pellets, using mini-barcoding PCRs and high-throughput sequencing (HTS), is an effective approach to assess arthropod diversity (Bohmann et al., 2011) and has been successfully applied to address various aspects of bat diets, such as predator-prey interactions and resource partitioning (Clare et al., 2009; Bohmann et al., 2011; Razgour et al., 2011; Dodd et al., 2012; Mata et al., 2016; Vesterinen et al., 2016). However, few studies have explored the prevalence of pest species in landscapes primarily dominated by diverse annual crops, such as those found in Asia, along with the diet partitioning of sympatric bats. This study aims to improve our understanding of the dietary habits of prevalent open-space bats that inhabit agricultural landscapes, specifically examining the presence of pest insect species using DNA metabarcoding. Our objective is to examine the extent of resource partitioning among these bat species, considering potential differences in resource utilization within their ecological niches.
We hypothesize that open-space bats residing in agricultural areas might consume both local pest insects and migratory pest insects owing to the nature of their habitat. Additionally, we posit that bats sharing the same foraging guild and habitat might display diet partitioning. This hypothesis is based on the premise that when multiple bat species coexist in overlapping habitats, they may partition their diets to reduce competition for resources (Siemers and Swift, 2006; Ruadreo et al., 2018). By testing these hypotheses, we aim to gain insights into the foraging behaviour and dietary habits of open-space bats in agricultural landscapes.
Materials and Methods
Study Site and Species
This study was carried out in central Thailand (Lopburi and Ayutthaya provinces). Bats were captured using mist nets and hand nets as they returned to caves after foraging. The first roost, Khao Wong cave (15°02′N, 101°18′E), was inhabited by Mops plicatus and Taphozous theobaldi, while the second roost, Wat Tham Bot Yawat cave (15°03′N, 101°32′E), was a roosting site for T. melanopogon, located approximately 1.49 km away from the first cave. Additionally, a house-dwelling colony of Scotophilus heathii (14°18′N, 100°29′E) was found approximately 116 km away from both caves. The local landscape is dominated by sugarcane, corn, cassava plantations, and rice fields. Sugarcane is planted from October through February, taking 10–14 months to harvest (Manivong and Bourgois, 2017). Corn is planted in May–June and is harvested in 100–120 days, around September–October (Ekasingh et al., 2004). Cassava is planted March–May and harvested in January–March, 8–20 months after planting (Arthey et al., 2018). Rice is typically planted around May and harvested around December–January, depending on rainfall (Felkner et al., 2009). The region experiences three distinct seasons influenced by monsoon winds: dry, rainy, and cool dry. The dry season, from mid-February to mid-May, is dry and hot. The rainy season, lasting from mid-May to mid-October, brings abundant rainfall, especially in August and September. From mid-October to mid-February, the cool dry season prevails, characterized by cooler temperatures and dry weather (Ruadreo et al., 2018).
Open-space bats have narrow wings with high wing loading and aspect ratio for fast flight (Voigt et al., 2018). The wrinkle-lipped free-tailed bat (M. plicatus, Molossidae) is known for its foraging behaviour at high altitudes (Nguyen et al., 2019). This species typically forms large colonies in caves, ranging from 10,000 to several million individuals (Srilopan et al., 2018). Theobald's tomb bat (T. theobaldi, Emballonuridae) is known to roost in caves and forage above nearby forests. It has been observed foraging at altitudes of up to 800 m above ground (Roeleke et al., 2018). The black-bearded tomb bat (T. melanopogon, Emballonuridae) is typically found in hilly and forested areas, and this species roosts in caves, buildings, and temples, where they form sizable populations (Wei et al., 2008). The Greater Asiatic yellow house bat (S. heathii) is widespread and commonly found in various habitats, especially human settlements. It roosts in crevices, cracks in buildings, palm crowns, tree hollows, and among banana leaves, either alone or in colonies of up to 50 individuals (Shahbaz et al., 2014).
Sampling and Faecal Storage
We collected faeces from 70 individuals each of M. plicatus, T. theobaldi, and T. melanopogon. Additionally, we obtained 40 faecal samples from a roost of S. heathii by randomly collecting faces from bats in different spots within the roost. We ensured that the samples were collected from various locations to confidently ascertain that the pellets came from different individuals (Table 1). The bats were captured upon returning to the roost following foraging activities, ensuring prompt defecation. To prevent cross-contamination of samples, each bat was kept separately in a clean, single-use cotton bag for 15–20 minutes. They were then identified, sexed, and aged before being released. Faecal pellets were collected from the bags and stored in 1.5-ml collection tubes filled with silica gel granules and kept at room temperature (around 25°C) for up to 24 hours before being transferred to a freezer. All samples were kept dry and refrigerated (4–8°C) for 2 to 12 months until arrival at the laboratory. The experimental procedures and protocols received approval from the ‘Standard of Animal Research, Research and Development’ office, Prince of Songkla University (project license number: U1-06459-2560, protocol code: 2564-01-026).
Table 1.
Number of samples collected from each bat species during each season
Laboratory Procedures
DNA was extracted from one dropping per study individual. The numbers of individuals and pellets analysed were carefully chosen to balance accuracy in diet estimation and cost considerations (Mata et al., 2019). A previous study has shown that pooling samples can provide poor estimates of diet diversity and frequency of occurrence. Mata et al. (2019) suggests that most variation in diet composition is associated with differences among individual bats, followed by the number of pellets per individual, and PCRs per pellet. Therefore, we chose to collect one sample from each individual rather than five samples from one individual to capture more individual variation and provided a more comprehensive understanding of the diet diversity within the population, despite the limited budget. We randomly selected a complete pellet that was not broken, ensuring it was of average size.
The extraction protocol followed methods described by Gonçalves et al. (2024), which involved an initial incubation period using a lysis buffer composed of 0.1 M Tris-HCl, 0.1 M EDTA, 0.01 M NaCl, 1% N-lauroylsarcosine, pH 7.5–8.0 (Maudet et al., 2002). Cell lysis, DNA precipitation, and washing steps were performed using E.Z.N.A. Tissue Kits (Omega Bio-Tek, Norcross, Georgia, USA). For each extraction, one faecal pellet was added to 650 µL of the lysis buffer. The samples were homogenized with a spatula, vortexed, and incubated in a dry bath at 70°C for 30 minutes. Following incubation, the samples were briefly centrifuged, and up to 500 µL of the supernatant was transferred to a new tube containing 25 µL of OB Protease. The remaining steps of the extraction protocol followed the recommendations provided with the E.Z.N.A. Tissue Kits, except that the DNA was eluted twice in separate 50-µL elutions. The DNA extraction process was carried out in batches, with each batch consisting of 23 samples and one negative control (without a faecal pellet). All extracted DNA was distributed across 96-well plates, with the final well in each plate designated for the PCR negative control.
The prey DNA was amplified using the FwhF2-R2n COI primer set (Vamos et al., 2017), which were modified to include Illumina adaptors. The Fwh2 primer was originally designed for amplifying freshwater invertebrates but has been shown to be effective in amplifying a wide range of terrestrial arthropods, including moths, ants, beetles, true bugs, spiders, and even spring-tails (Mata et al., 2021). PCR reactions consisted of 5 µL of Qiagen Multiplex Master Mix, 0.3 µL of each 10 nM primer (forward and reverse), 2.4 µL of water, and 1 µL of DNA extract. The cycling conditions started with an initial denaturation at 95°C for 15 minutes, followed by 40 cycles of denaturation at 95°C for 30 seconds, annealing at 50°C for 30 seconds, and extension at 72°C for 30 seconds. A final extension step was carried out at 60°C for 10 minutes. PCR products were diluted 1:3 with water and subjected to a second PCR reaction to incorporate 7 bp identification tags and Illumina P5 and P7 adaptors. The second PCR reactions and cycling conditions were similar to the first PCR, with the exception that KAPA HiFi HotStart ReadyMix (Rocher, KAPA Biosystems, Basel, Switzerland) was used, and only nine cycles of denaturation, annealing, and extension were performed, with an annealing temperature of 55°C. The PCR products were purified using Agencourt AMPure XP beads (Beckman Coulter, Brea, California, USA) and quantified using Nanodrop. They were then diluted to a concentration of 15 nM. The purified and normalized PCR products were pooled. The library was quantified using qPCR (KAPA Library Quant Kit qPCR Mix, Rocher) and diluted to a concentration of 4 nM. Sequencing was performed on a MiSeq platform using a v2 kit (500 cycles), targeting a coverage of 30k reads per PCR product. The target coverage was determined based on the taxa amplified by the marker. It is important to note that Fwh2 amplifies not only insects but also some fungi and vertebrates like bats, which may result in the loss of some reads to non-dietary items (Mata et al., 2021).
Bioinformatic Analysis
We aligned pairs of sequences using Obitools, then eliminated primer sequences. Sample information was added to each assembled read using the ‘ngsfilter’ command from OBITools (Boyer et al., 2016). Next, the reads were de-replicated into unique sequences using the ‘obiuniq’ command and merged into a single fasta file. Afterward, we employed VSEARCH (Rognes et al., 2016) to denoise the dataset, remove chimeric sequences, cluster sequences into Operating Taxonomic Units (OTUs) with 99% similarity, and map raw reads with the resulting OTUs. Further refinement of the obtained OTU table and sequences was performed using the R package LULU (Frøslev et al., 2017) to remove potential mtDNA nuclear copies and persisting PCR and sequencing errors. OTUs were taxonomically assigned using BOLDigger (Buchner and Leese, 2020), and the outcomes were subsequently scrutinized and refined through manual assessment. In scenarios where achieving species-level assignments was unattainable using BOLD, we performed supplementary manual BLAST searches on NCBI to ensure precise identification. Sequences displaying fewer than 90% similarity were categorized at the class level. Those with similarities ranging from 90% to 95% were assigned to the family level, while sequences with over 95% similarity were allocated to species or genus level. In cases where sequences matched multiple species within the same genus, the OTU was designated at the genus level (Collins and Cruickshank, 2013; Wray et al., 2018). When encountering matches with multiple genera or species, OTUs were assigned based on species records in Thailand. Each OTU was classified as either ‘diet’ (representing most arthropods) or ‘not diet’ (including fungi and internal and external parasites) based on its taxonomic identification. Samples with fewer than 100 reads attributed to dietary items were considered unsuccessful and excluded from further analysis. To minimize potential false interactions related to cross-talk (Edgar, 2016) and secondary predation (Deagle et al., 2019), taxa representing fewer than 1% of the total number of dietary reads in each sample were removed. After this filtering step, the remaining samples were used for subsequent analysis.
Statistical Analysis
For statistical analysis, presence/absence matrices were constructed for each sample at three identification levels: species, family, and order. Additionally, we calculated the frequency of occurrence for each prey taxon at the species, family, and order levels for each bat species. The frequency of occurrence was determined by dividing the number of faecal samples in which a specific species, family, or order was detected by the total number of samples collected for each bat species.
All analyses were performed with R (version 4.2.2) and R Studio. We tested for differences in prey richness, niche width, and prey composition among bat species diet, as well as overall niche overlap among bats. To test for differences in prey average richness between bat species, we conducted a generalized linear mixed model with a Poisson error distribution using the ‘glm' function from the R package ‘lme4'. The variable ‘season' was included as a random factor to account for potential temporal effects. The significance of the generalized linear mixed model was assessed using the ‘Anova’ function from the ‘car’ package. We compared the niche breadth of the four bat species at the species, family, and order levels by estimating richness values using Hill numbers base on twice the number of samples of the species with the smallest sample size. This analysis was conducted using the ‘iNEXT’ function from the R package ‘iNEXT’. Differences between estimates were deemed significant if the 95% confidence intervals (CIs) around the richness estimates did not overlap. To examine the influence of bat species on the diversity of prey species composition within each sample, we performed a permutational multivariate analysis of variance (PERMANOVA) based on presence/ absence data and the Jaccard index. Significance was assessed using 9,999 random permutations (Anderson, 2001). To identify the most distinct prey items, we conducted a SIMPER analysis by calculating the percentage similarities within samples attributed to bat species that contributed to these similarities. This analysis was based on Jaccard’s similarities, using the frequency of prey occurrences in samples as a measure of abundance. To assess trophic niche overlap among the four bat species, we calculated dietary niche similarity using Pianka’s index, which is based on the frequency of occurrence of prey insects. This index ranges from 0 to 1, with a value of 1 indicating complete overlap in dietary niche among the bat species. The interaction network between each bat species and insect family was constructed and analysed using the R package ‘bipartite’. The strength of interactions was assessed based on the number of samples in which each insect family was detected.
Analysis of Agricultural Pest Insects
We assessed the pest status of prey taxa identified to species level or as a species complex (i.e., when two or three species of the same genus were equally matched) with high support (> 99% identity). We focused on agricultural insect pests based on a comprehensive literature survey (Sukonthaphirom Na Phatthalung et al., 2021; http://ippc.acfs.go.th/pest; http://www.ricethailand.go.th/rkb3/Insect.htm; https://www.si.edu/spotlight/buginfo/diseases). We also evaluated the pest status of migratory insect prey based on a comprehensive review of the literature (Fox, 1978; Farrow and McDonald, 1987; Riley et al., 1987, 1995; Chapman et al., 2002; Song, 2005; Wood et al., 2009; Margam et al., 2011; Fu et al., 2015; Longbottom et al., 2017; Huang et al., 2020; Shen et al., 2020; Zeng et al., 2020; Chen et al., 2022; Wang et al., 2022; Atieli et al., 2023; Chatterjee et al., 2023; Huang et al., 2023).
Results
A total of 1,557 operational taxonomic units (OTUs) were obtained from all samples. Arthropods considered to be diet constituted a significant proportion of the OTUs (49.45% — 770). In contrast, Chiroptera OTUs, not part of the diet, contributed to a smaller fraction, approximately 0.83% (13). The remaining OTUs were either fungi, parasites, or other non-dietary items. Among the studied bat species, M. plicatus exhibited a remarkably diverse diet, consuming 216 OTUs from 67 families and 11 orders. The primary prey items were predominantly from the order Hemiptera (28.6%), specifically, the families Pentatomidae (9.1%), and Delphacidae (7.7%). The order Diptera (16.8%) was primarily composed of the families Limoniidae (5.5%) and Culicidae (2.5%). The order Lepidoptera (14.3%) consisted of the families Noctuidae (3.2%) and Erebidae (2.3%). Additionally, the order Blattodea (12.1%) was primarily attributed to the family Termitidae (10.9%). The order Coleoptera (11.1%) was mostly comprised of the families Tenebrionidae (3%) and Carabidae (1.4%). These orders served as the primary contributors to the diet of M. plicatus (Fig. 1).
The diet of T. theobaldi included 110 OTUs across 42 families and 11 orders. The primary prey items were predominantly derived from the order Lepidoptera (40.3%), with notable contributions from the families Noctuidae (16.6%) and Erebidae (12.8%). Another significant order was Orthoptera (29.7%), primarily comprising the families Gryllidae (11.0%), Gryllotalpidae (6.9%), and Tettigoniidae (5.5%). The order Blattodea (11.7%), was represented mainly by the family Termitidae (10.7%). The order Diptera (5.5%) had the family Chironomidae (3.1%) as the main contributor. Additionally, the order Coleoptera (4.5%), was comprised largely by the family Bostrichidae (1.7%) (Fig. 1).
Taphozous melanopogon consumed a total of 137 OTUs from 34 families and eight orders. The main prey order observed was Lepidoptera (58%), accounting for a significant proportion of the diet. The most common lepidopteran families were Noctuidae (25.6%), followed by Erebidae (13.9%), Crambidae (6.8%), Pyralidae (3.2%), and Geometridae (2.1%). Another significant order in the diet was Orthoptera (30.3%), represented by the families Gryllotalpidae (9.6%), Gryllidae (9.3%), Tettigoniidae (6.1%), and Acrididae (3.6%). The order Diptera (3.9%), was primarily represented by the family Limoniidae (1.8%) (Fig. 1).
The diet of S. heathii included a total of 126 OTUs across 47 families and 12 orders. The primary order observed was Coleoptera (44.8%) which was represented mostly by the families Hydrophilidae (11.2%) and Tenebrionidae (9.4%). The order Hemiptera (18.8%) also played a substantial role in the diet, with the families Pentatomidae (8.0%) and Notonectidae (2.2%) being the most abundant. The OTUs in the order Lepidoptera (10.8%) were primarily within the families Gelechiidae (4.0%) and Tortricidae (1.4%). Additionally, the order Diptera (9.4%) was represented by the families Chironomidae (2.2%) and Culicidae (1.8%) while the order Psocodea (5.1%) was primarily represented by the family Liposcelididae (4.7%) (Fig. 1).
Rarefaction curves indicated that the dietary items of M. plicatus and T. melanopogon, approached saturation, while those of T. theobaldi and S. heathii did not seem to be saturated. Predicted family richness estimates indicated that more samples were needed for the latter species. At the family level, M. plicatus and S. heathii showed the widest niches, while T. theobaldi and T. melanopogon exhibited the lowest (Fig. 2). Confidence intervals (95%) indicated significant differences in niche width between M. plicatus and both Taphozous species, as well as between S. heathii and T. melanopogon. Analysis of prey richness per sample indicated significant differences among certain bat species. Specifically, T. melanopogon and T. theobaldi exhibited significantly lower prey richness compared to M. plicatus (P < 0.001), while S. heatii showed no significant difference (P = 0.235).
The bat species in this study showed different levels of dietary overlap. Taphozous theobaldi and T. melanopogon had the highest overlap (0.75), indicating similar food preferences. Taphozous theobaldi also had moderate overlap with M. plicatus (0.41). Mops plicatus had lower overlap values with T. melanopogon (0.25) and S. heathii (0.24). The lowest degrees of overlap were observed between S. heathii and T. melanopogon (0.07) and S. heathii and T. theobaldi (0.1) (Fig. 3).
Diet composition at the family level displayed significant differences between bat species, where the species factor showed a significant effect on the composition (F-value = 9.401, d.f. = 3, R2 = 0.112, P = 0.001). SIMPER tests revealed that M. plicatus and S. heathii exhibited differences across 27 families. For instance, M. plicatus consumed more Cicadellidae, Delphacidae, and Pentatomidae from the order Hemiptera, while S. heathii consumed more Cerambycidae, Hydrophilidae, and Tenebrionidae from the order Coleoptera. Mops plicatus displayed differences in diet composition across 35 families compared to T. melanopogon, and across 34 families compared to T. theobaldi. For example, M. plicatus consumed higher amounts of Chrysomelidae and Staphylinidae from the order Coleoptera, along with Limoniidae and Muscidae from the order Diptera, and Delphacidae and Lygaeidae from the order Hemiptera. On the other hand, T. melanopogon had a greater intake of Crambidae and Noctuidae from the order Lepidoptera. Similarly, S. heathii showed differences across 25 families compared to T. melanopogon, and across 24 families compared to T. theobaldi. As an example, S. heathii consumed larger quantities of Hydrophilidae, Tenebrionidae, and Cerambycidae from the order Coleoptera, along with Micropezidae from the order Diptera, and Belostomatidae, Notonectidae, and Flatidae from the order Hemiptera. In contrast, T. melanopogon exhibited a higher consumption of Noctuidae from the order Lepidoptera and Gryllotalpidae from the order Orthoptera. Additionally, there were differences in diet composition across 22 families between T. melanopogon and T. theobaldi. For instance, T. melanopogon showed greater consumption of Erebidae, Noctuidae, and Pyralidae from the order Lepidoptera, as well as Gryllotalpidae and Tettigoniidae from the order Orthoptera. Conversely, T. theobaldi exhibited higher consumption of Baetidae from the order Ephemeroptera, along with Aphididae and Psyllidae from the order Hemiptera, and Mantidae from the order Mantodea (Fig. 1 and the Appendix).
Fig. 1.
Interaction network showing the dietary niches of the four open-space bat species studied in central Thailand. Link width represents the percent frequency of occurrence of each prey family/order detected in each bat sample
Fig. 2.
Rarefaction and extrapolation curves of prey richness for the studied bat species: M. plicatus, S. heathii, T. melanopogon, and T. theobaldi
Fig. 3.
Niche overlap among open-space bats in central Thailand, with overlap values ranging from 0–1. SH — S. heathi, MP — M. plicatus, TM — T. melanopogon, TT — T. theobaldi
We found at least 36 agricultural pest species in the diet of these open-space bats (Table 2). Mops plicatus and T. melanopogon consumed the highest number of pests (20), followed by T. theobaldi (18), and S. heathii (6). Additionally, these bats were observed foraging on at least 22 species of migratory insect pests, including: Culex tritaeniorhynchus, Nephotettix virescens, Nilaparvata lugens, Sogatella furcifera, Nezara viridula, Cnaphalocrocis patnalis, C. poeyalis, Maruca vitrata, Ostrinia furnacalis, Scirpophaga incertulas, Achaea janata, Anomis flava, Mythimna formosana, M. impura, Sesamia inferens, Spodoptera exigua, S. frugiperda, S. litura, S. mauritia, Plutella xylostella, Cyrtacanthacris tatarica, and Locusta migratoria — Table 2). Certain migratory pest insects were consumed by all four sympatric open-space bat species, including the pests, e.g., M. impura, S. inferens, and S. frugiperda (Table 2).
The family Termitidae (notably, Macrotermes annandalei, Odontotermes sp. 1, and Odontotermes sp. 2) was consumed at a high frequency of occurrence by most of the bat species. Additionally, in the diet of M. plicatus, the family Delphacidae (notably, N. lugens and S. furcifera) was found to be consumed frequently. On the other hand, T. theobaldi and T. melanopogon primarily consumed prey from the family Noctuidae (notably, M. impura, S. inferens, S. frugiperda, and S. litura) (Table 2).
Discussion
Insectivorous bats are well-known as important predators of arthropod pests (Aizpurua et al., 2018). Our DNA metabarcoding results showed that the open-space bats, M. plicatus, S. heathii, T. melanopogon, and T. theobaldi, have diverse diets that include economically damaging insect pests that harm crops and serve as disease vectors. We identified at least 36 insect species known to damage agricultural crops and carry human diseases. These pests are known to affect a variety of crops, including sugarcane (e.g., Macrotermes annandalei, Chilo tumidicostalis), rice (e.g., N. lugens, S. furcifera), and corn (e.g., S. frugiperda, M. impura), as well as other plants such as cotton, mung beans, grapes, okra, dendrobium, asparagus, yard-long beans, cruciferous vegetables, sunflowers, sorg-hum, soybeans, sesame, castor beans, groundnuts or peanuts, chili, garden peas, tomatoes, and mangosteen (Sukonthaphirom Na Phatthalung et al., 2021). Additionally, these bats also consumed Culex tritaeniorhynchus and Culex vishnui, which are known to be the primary vectors of Japanese encephalitis (Karthika et al., 2018).
In accordance with previous research, M. plicatus was found to primarily feed on rice insect pests, including N. lugens, S. furcifera, N. virescens, and S. incertulas, as well as disease vectors such as Culex gelicus, C. tritaeniorhynchus, and C. vishnui (Leelapaibul et al., 2005; Srilopan et al., 2018; Thongjued, 2019). In contrast, there have been no previous reports of insect pests in the diets of T. theobaldi, T. melanopogon, and S. heathii. However, our study revealed that these three species consume various insect pests, including Mythimna formosana, M. impura, S. inferens, C. tumidicostalis, Cnaphalocrocis exigua, C. patnalis, and S. frugiperda, among others. These findings suggest that these bat species might also play important roles in pest suppression within agricultural landscapes, similar to M. plicatus.
Our results demonstrate the effectiveness of DNA metabarcoding in bat diet analysis, revealing substantially high prey diversity compared to previous studies conducted in Thailand (Leelapaibul et al., 2005; Weterings et al., 2015; Ruadreo et al., 2018; Srilopan et al., 2018; Thongjued, 2019). Mops plicatus exhibited the most diverse diet, consuming 216 OTUs from 67 families and 11 orders. The diet primarily consisted of prey items from the order Hemiptera, followed by Diptera and Lepidoptera. This finding corresponds to the results of previous studies, including those that identified prey from faecal pellets via microscopic analysis, which identified Hemiptera, Coleoptera, Lepidoptera, and Diptera as the main components within the diet of M. plicatus (Leelapaibul et al., 2005; Srilopan et al., 2018). In contrast, research conducted by Ruadreo and colleagues (2018) and Thongjued (2019) reported Diptera as the primary prey, followed by Hemiptera, Lepidoptera, and Coleoptera. Thongjued (2019) used direct PCR-DGGE and found that while Diptera constituted the primary diet, Hemiptera also made up a substantial proportion, with minimal differences between them. Ruadreo and colleagues (2018) used stable isotopes for diet analysis. However, this method is more suitable for providing a broad overview of the ecological niche rather than detailed dietary information (Roswag et al., 2018). The observed differences in results are likely due in part to the different methods employed. Additionally, our results showed that M. plicatus feeds mainly on insects in Hemiptera, whereas Tadarida brasiliensis primarily consumes Lepidoptera, despite both being molossid bats with similar behaviour. This difference may be due to the distinct regions and main crops each species inhabits (Krauel et al., 2018a). In Thailand, Hemipterans are the main pests of rice, which is widely cultivated in central Thailand (Srilopan et al., 2018). Nevertheless, previous studies have consistently shown that M. plicatus primarily consumes migratory pest insects, such as N. lugens and S. furcifera, despite variations in methodology (Leelapaibul et al., 2005; Srilopan et al., 2018; Thongjued, 2019). Similarly, T. brasiliensis, which preys on migratory insects (Krauel et al., 2018a). This high-flying bat provides potential for investigating the flows of migratory insects across the landscape in Texas (Krauel et al., 2018a).
Table 2.
List of arthropod pest species preyed upon by open-space bats in central Thailand (M. plicatus — MP, T. melanopogon — TM, T. theobaldi — TT, and S. heathii — SH), along with the crops or hosts affected by each species and the frequency of occurrence in the diet of each bat species (* = migratory pest). Please also note that Culex tritaeniorhynchus and C. vishnui are arbovirus vectors
Our dietary analysis also revealed that the congeners T. theobaldi and T. melanopogon have similar diets. Taphozous theobaldi consumed 110 OTUs from 42 families and 11 orders. Similarly, the diet of T. melanopogon comprised 137 OTUs from 34 families and eight orders. Their primary prey items were predominantly from the order Lepidoptera, followed by Orthoptera, Blattodea, and Diptera. This is also supported by previous studies examining the diets of T. melanopogon and T. longimanus, where Lepidoptera was the primary prey, followed by Coleoptera, Diptera, and Orthoptera (Srinivasulu and Srinivasulu, 2005; Wei et al., 2008). Additionally, T. theobaldi, known for high-altitude foraging (Roeleke et al., 2018), exhibited a preference for Lepidoptera, especially the Noctuidae family, in every sampling period. We assume from their diets that both species primarily feed on migratory insects such as Noctuidae at high altitudes. This behaviour may be similar to that of T. brasiliensis, an open-space bat known for consuming migratory noctuid moths (Krauel et al., 2018b). In contrast, a previous study conducted in our study area that use stable isotopes reported that T. theobaldi primarily consumed Diptera and Lepidoptera during the dry season and shifted to a Diptera dominant diet in the wet season, while T. melanopogon mainly consumed Diptera in both seasons (Ruadreo et al., 2018). However, the observed variations in the diets of T. theobaldi and T. melanopogon between our study and prior research could be attributed to several factors, such as the time of data collection, which may impact the availability of insects, and differences in the methods used (Dai et al., 2023).
Dietary analysis of S. heathii revealed the consumption of 126 OTUs from 47 families and 12 orders. The primary order observed was Coleoptera, followed by Hemiptera, Lepidoptera, and Diptera. This finding is consistent with a previous study in India, where Coleoptera were predominantly found in the guano of S. heathii (Misra and Elangovan, 2016). Additionally, Coleoptera, Hemiptera, Diptera, and Lepidoptera were recorded in the diet of a smaller congener, S. kuhlii (Srinivasulu et al., 2010; Zhu and Zhang, 2012; Misra and Elangovan, 2016).
The dietary analysis of these four bat species, reveals varying patterns of dietary breadth and specialization. Mops plicatus exhibits the broadest dietary breadth, with a primary focus on Hemiptera but also including Diptera and Lepidoptera. In contrast, T. theobaldi and T. melanopogon show degrees of specialization towards Lepidoptera, along with other orders such as Orthoptera, Blattodea, and Diptera, albeit with fewer total OTUs and narrower taxonomic ranges compared to M. plicatus. Scotophilus heathii, demonstrates a moderate dietary breadth with a primary focus on Coleoptera, supplemented by Hemiptera, Lepidoptera, and Diptera. These findings suggest that while some species like M. plicatus maintain a broader dietary spectrum encompassing multiple prey types, others like T. theobaldi and T. melanopogon exhibit more specialized feeding behaviours, possibly influenced by habitat preferences and available prey resources (Dai et al., 2023), indicating niche partitioning among these bat species.
Regarding dietary niche partitioning of these four sympatric insectivorous bats, our observations revealed a varying degree of diet overlap and partition. Specifically, we observed notable instances of diet overlap, such as all four bat species consuming M. annandalei, a termite known to swarm abundantly during the night time a few days before or after rainfall (Peppuy et al., 2004). This finding suggests that these bats may share a preference for this termite, which becomes readily available during specific weather conditions. In the analysis of each bat pair, T. theobaldi and T. melanopogon exhibited a substantial overlap in insect prey at the family level, and thus a preference for similar food items. This may be notable as they are sibling species and co-occur throughout their distributions (Arrizabalaga-Escudero et al., 2018). We observed that the dietary niche width and composition of both species were remarkably similar, characterized by the consumption of Noctuidae and Erebidae from the order Lepidoptera. Noctuidae and Erebidae are well-documented as common agricultural pests (Crabo et al., 2013; Rakkasikorn et al., 2019; Sukonthaphirom Na Phatthalung et al., 2021). These bats may share agricultural migratory pest insects, particularly moths that migrate at high altitudes (Lee and McCracken, 2005; Müller et al., 2012), as evidenced by the presence of such pests in their diets. However, differences in diet composition across 22 families were observed. Taphozous melanopogon showed greater consumption of Noctuidae (Lepidoptera) and Gryllotalpidae (Orthoptera), while T. theobaldi showed higher consumption of Baetidae (Ephemeroptera), Aphididae, and Psyllidae (Hemiptera). Our study suggests specific prey differences might contribute to the partitioning of their diet. This is supported by previous research showing that, even though their prey belong to the same orders of Arthropoda, the exact species of insects available to the bats likely varies (Rolfe et al., 2014). When migratory insect pests are not abundant, these bats may focus on different prey while foraging in distinct habitats. Previous research aligns with our findings, illustrating varied habitat preferences. Bats often exploit different habitats and are consequently exposed to various species of insects while foraging (Rolfe et al., 2014). Taphozous melanopogon tends to forage more in agricultural areas and edge forests where Noctuidae and Gryllotalpidae are widespread, while T. theobaldi seems to prefer settlements and water bodies where Baetidae are prevalent (Zahoor et al., 2003; Kubendran et al., 2017; Suksai and Bumrungsri, 2020; Holusa and Kalab, 2023).
Our findings revealed moderate dietary overlap between T. theobaldi and M. plicatus, possibly due to their shared characteristics as high-altitude aerial insectivores that forage in open spaces, coexist in sympatric areas, and share roosts in the same cave (Ruadreo et al., 2018; Kemp et al., 2019). However, despite this overlap, our study clearly identified diet partitioning between them. For example, T. theobaldi exhibited a preference for pest insects from the order Lepidoptera, while M. plicatus primarily targeted those from the order Hemiptera (Leelapaibul et al., 2005; Srinivasulu and Srinivasulu, 2005; Wei et al., 2008; Srilopan et al., 2018). This distinction may be linked to the size difference between the bats, as T. theobaldi, being larger than M. plicatus (Ruadreo et al., 2018), shows a preference for larger moths over smaller planthoppers.
Moreover, M. plicatus exhibited lower overlap with T. melanopogon and S. heathii, indicating differences in their diet compositions. Mops plicatus mainly consumed pest insects from the order Hemiptera, T. melanopogon primarily targeted pest insects from the order Lepidoptera, and S. heathii predominantly consumed insects from the order Coleoptera (Leelapaibul et al., 2005; Srinivasulu and Srinivasulu, 2005; Wei et al., 2008; Misra and Elangovan, 2016; Srilopan et al., 2018). Previous studies have suggested that Mops plicatus flies at different altitudes compared to T. melanopogon and S. heathii (Roeleke et al., 2018; Nguyen et al., 2019). This variation in flight altitude may correspond to the insects targeted by each bat during foraging (Nguyen et al., 2019), leading to distinct diets, a phenomenon referred to as resource partitioning.
Scotophilus heathii has relatively low overlap with other open space bat. The diet of S. heathii exhibited lower diversity in pest insects compared to other species. However, Culex vishnui appeared notably frequently in its diet, particularly compared to other pest insects. This trend could be attributed to its habit of roosting in human houses within urban areas, reflecting its foraging preferences in such environments.
We acknowledge several limitations that are worthy of discussion. First, the collection of bat faeces from the house-dwelling colony of S. heathii posed challenges, restricting us to a single collection event. The obtained faeces were not fresh, potentially influencing prey species detection. Second, the house-dwelling colony of S. heathii was geographically distant from the other sampled roosts, with bats potentially exposed to distinct habitats and environmental factors. Third, we faced constraints in collecting and analysing insects in their natural habitat to serve as reference for dietary analysis. The cost associated with DNA metabarcoding, while continuing to become more affordable, remain a limiting factor in undertaking comprehensive insect analysis.
Dietary analysis of four insectivorous bats, M. plicatus, T. theobaldi, T. melanopogon, and S. heathii, emphasizing their crucial role in agricultural landscapes through the consumption of numerous pest insects, including migratory species that threaten crops. Our results indicate dietary partitioning among these sympatric species: M. plicatus primarily consumes Hemipterans, T. theobaldi and T. melanopogon favor lepidopterans, while S. heathii prefers coleopterans. This partitioning is essential for their coexistence and highlights their collective importance in pest suppression. While our study focused on the family level, a more detailed analysis at the species level could reveal even greater dietary distinctions. Future research should delve deeper into the specific diets of these bats and quantify the ecosystem services they offer. These findings underscore the significant ecosystem services provided by insectivorous bats. Communicating these insights to stakeholders involved in bat conservation is crucial to protect these species and the benefits they provide to agriculture.
Acknowledgements
This research project is supported by National Research Council of Thailand (NRCT): NRCT5-RGJ63019-162. We thank Phongsin Inpaen, Kamonporn Srilopan, Tusita Srilopan, and Sunet Nalisa for help during the field work. We also thank CTM Team, CIBIO/InBIO for their assistance in the laboratory. We extend our special thanks to Alyssa Stewart for her valuable comments on the manuscript. We thank Christian Voigt, Jennifer Krauel and the members of the Small Mammal, Bird and Spider Research Unit who helped develop ideas. Additionally, we appreciate the support of the Faculty of Science, Prince of Songkla University and CIBIO-InBIO, Faculty of Sciences, University of Porto. VAM was funded by Fundação para Ciência e a Tecnologia (FCT) through an individual research contract (doi: 10.54499/2020.02547.CEECIND/CP1601/CP1649/CT0004).
© Museum and Institute of Zoology PAS
Author Contribution Statement
SB: research concept and design, critical revision and final approval of the article; VAM: research concept and design, data analysis and interpretation, critical revision and final approval of the article; SS: research concept and design, collection and/or assembly of data, data analysis and interpretation, writing the article and its final approval.
Literature Cited
Appendices
Appendix
Statistically significant differences in prey items at the family level from 14 prey orders among aerial insectivorous bat species’ diets. Abbreviations for the bat species; MP — M. plicatus, SH — S. heathii, TM — T. melanopogon, TB — T. theobaldi. Different colors in the table represented statistically significant differences: 
P < 0.001, 
P < 0.01, 
P < 0.05, 
P < 0.1, 
non-significant
