Microsatellite Data Indicate an Extreme Founder Event with a Single Female Lineage in the Parasitoid Wasp Monodontomerus obscurus

TL;DR

A parasitoid wasp in Japan likely started from a single female, showing how extreme founder events can lead to population success.

Contribution

Provides rare empirical evidence of a population founded by a single female in a natural setting.

Findings

The wasp population in Japan was likely founded by a single female, as shown by extremely low genetic diversity.

Post-invasion mutations generated new genotypes in the population.

The species' mating system allows successful reproduction even in new habitats with minimal inbreeding effects.

Abstract

In general, when insects or other organisms invade a new habitat and establish a population, multiple individuals are required to arrive. When only a few individuals colonize a new area, they may fail to find mating partners, or inbreeding may produce genetically non-viable offspring. Monodontomerus obscurus is a parasitoid wasp that mainly attacks solitary bees. In Japan, this species was first recorded in 2000, and our field surveys showed that its parasitism rate has increased in recent years. By developing genetic markers and examining population structure and genetic variation, we found evidence that a population was founded by very few females, most likely a single female. Gregarious parasitoid wasps in the superfamily Chalcidoidea, including the present species, commonly mate with close relatives soon after emergence. This mating system may allow them to reproduce without searching for mates even in a new habitat, and to be less affected by the negative effects of inbreeding even after population establishment by a single individual.

How many founders are required for insects and other organisms to establish new populations is a fundamental question in invasion biology. We investigated the population establishment process of a parasitoid wasp, Monodontomerus sp., which was first recorded in Japan in 2000. Field surveys conducted in this study showed that the parasitism rate has been increasing in recent years. Morphological and molecular analyses suggested that the parasitoid species is M. obscurus, or a closely related lineage derived from it, which newly invaded Japan. To examine genetic variation during the early stage of invasion, we developed microsatellite DNA markers and conducted population genetic analyses. The results revealed extremely low genetic diversity: most loci were monomorphic, polymorphism was restricted to loci with long repeat motifs, and the allele frequencies of these loci were dominated by single alleles. A minimum spanning network based on microsatellite genotypes exhibited a star-like pattern. These results based on genome-wide microsatellite data indicate that the present population was founded by very few individuals, most likely a single female or an effectively single genetic lineage, and novel genotypes arose through post-invasion mutations. Our study provides rare empirical evidence for single-female founding under natural conditions, and highlights how species-specific life-history and genetic systems can enable successful invasion despite extreme bottlenecks.

Figures

Full text

1. Introduction

Biological invasions into new habitats pose a major challenge in insect biology and evolutionary ecology. Among the factors influencing invasion success, the number of founding individuals is considered one of the most critical determinants [1,2]. Once introduced into a new habitat, both the demographic characteristics and genetic composition of the founders can profoundly shape the trajectory of population establishment and subsequent population dynamics [3,4].

The Allee effect, which is a decrease in growth rate with declining population size or density, plays an important role in shaping the relationship between initial population size and the probability of establishment [5]. Insects are one of the organismal groups in which Allee effects are frequently documented [6,7,8]. In particular, low population density often leads to failures in finding mates [9,10]. Although the mated females of insects can reproduce without mating partners in newly colonized habitats, subsequent generations may still suffer mating failures when population density remains low during bottleneck events. In addition, low population density increases the rate of inbreeding, which generates genetic loads by elevating the homozygosity of recessive deleterious or lethal alleles that are normally masked in heterozygous individuals, resulting in inbreeding depression and potentially disrupting subsequent population establishment [11,12]. If these effects drive the growth rate to become negative, it further accelerates population decline, and can ultimately lead to extinction.

In haplodiploid organisms, in which diploid females and haploid males develop from fertilized and unfertilized eggs, respectively, the genetic load caused by recessive deleterious or lethal alleles is expressed and purged in haploid males. Haplodiploid insects, such as hymenopteran species, and mites have been reported to suffer less from inbreeding depression than diploid species [13,14,15]. However, even haplodiploid organisms can still incur genetic load through inbreeding due to alleles expressed only in females [16,17]. Moreover, many hymenopteran species (e.g., Ichneumonoidea, Vespoidea, Formicoidea, and Apoidea) employ complementary sex determination (CSD), in which diploid females develop only when they are heterozygous at the sex-determining locus, whereas homozygotes develop into diploid males that are inviable or sterile [18,19,20]. In small populations, genetic variation at this locus becomes reduced, increasing the production of diploid males and thereby generating the genetic load, which further erodes genetic variation and ultimately results in population extinction [21,22].

Monodontomerus (Hymenoptera: Torymidae) is a genus of parasitoid wasp that mainly attacks various Lepidoptera and Hymenoptera. Five species of Monodontomerus have been recorded from Japan, and two of them (M. osmiae Kamijo, 1963 and M. obscurus Westwood, 1833) have been reported from solitary bees [23,24]. The giant leaf-cutting bee, Megachile sculpturalis Smith, 1853, native to East Asia, has invaded Europe and North America, where parasitism by M. obscurus has been documented. In its native range in Japan, while several parasitoid species have been recorded from M. sculpturalis, no parasitism by Monodontomerus had been reported during the 1900s [25,26,27]. In 2000, parasitism by Monodontomerus was first detected, and they were identified as M. obscurus [24]. Since then, Monodontomerus has been collected regularly from M. sculpturalis [28] and [J. A. personal observation].

The population of the Monodontomerus in Japan therefore provides an excellent opportunity to understand how underlying ecological and genetical features influence invasion and population establishment. In most gregarious and quasi-gregarious parasitoid wasps, including Monodontomerus species, mating occurs exclusively at emergence sites soon after eclosion [27], which may prevent mate-finding failure even in low-density populations. Monodontomerus species exhibit haplodiploid sex determination; however, unlike many other hymenopterans, species in the superfamily Chalcidoidea, including Monodontomerus, show sex-determination mechanisms other than CSD [19,20]. Consequently, they are likely to be less susceptible to genetic load caused by recessive deleterious or lethal alleles or by the production of diploid males. These features may enable invasion and population establishment by a very small number of founders. Nevertheless, even among haplodiploid parasitoid species lacking CSD, contrasting results have been reported, ranging from the absence to the presence of inbreeding depression during population bottlenecks [29,30,31].

In this study, we aimed to elucidate how a population was founded under natural conditions in a species of Monodontomerus in Japan. First, we conducted field surveys in Japanese populations to trace temporal changes in parasitism rates by collecting the nests of M. sculpturalis. Second, we examined morphological characteristics and mitochondrial DNA sequences to identify the species and infer its phylogenetic origin. Third, to examine recently founded populations with sufficient genetic resolution, we newly developed microsatellite DNA markers, which evolve more rapidly and are more suitable for detecting recent demographic events than other markers, such as mitochondrial DNA. Finally, using these markers, we investigated the genetic structure and variation of a population, and examined the demographic and genetic processes underlying the establishment of a newly founded population.

2. Materials and Methods

2.1. Field Sampling

To collect the host M. sculpturalis and its parasitoid Monodontomerus sp., we used bamboo traps, as M. sculpturalis builds its nests in gaps and crevices in the natural environment. The bamboo traps consisted of a bundle of 20 bamboo canes (200–300 mm in length and 10–15 mm in diameter) with each cane closed at one end by a node [32,33]. Each bamboo cane was split longitudinally into two halves and reassembled with vinyl tape, allowing the canes to be opened even after M. sculpturalis had constructed nests using sticky pine resin. We set up the bamboo traps in the Japanese mainland, including Hachioji city, Tokyo, in 2000–2006; Ninohe city, Iwate, in 2000–2006; Otu city, Shiga, in 2002–2014; Shizuoka city, Shizuoka, in 2009–2013; and Hiratsuka city, Kanagawa, in 2011–2025 (Table S1). We placed them in the field in July or early August and collected them between November and March, during which both the host and parasitoid wasps were diapausing as prepupae. After collection, the bamboo traps were brought back to the laboratory, where we opened them to inspect the nests of M. sculpturalis, and recorded the numbers of individuals parasitized and unparasitized by Monodontomerus sp., as well as their positions within the bamboo canes. Diapausing prepupae of Monodontomerus were allowed to continue development after the following spring, and emerged adults were preserved in 99.5% ethanol.

For statistical analysis, we analyzed the data using the statistical software R version 4.4.2 [34]. To test whether parasitism tended to occur at particular positions of hosts within bamboo canes, we conducted a randomization test (1000 iterations), in which host positions were resampled and the parasitism rate at the specific position was compared with that observed in the actual data. Temporal changes in parasitism rates across years were tested using generalized linear mixed models with a binomial error distribution, including bamboo traps as a random effect.

2.2. Species Identification Based on Morphological and Molecular Data

We identified Monodontomerus species using individuals collected from Hiratsuka according to Graham (1992) and Zerova and Seryogina (2002) [35,36]. In addition, we sequenced the mitochondrial cytochrome oxidase subunit I (COI) gene region from individuals of the Hiratsuka population, and analyzed their phylogenetic relationships using available sequence data of Monodontomerus species retrieved from the DNA Data Bank of Japan (DDBJ)/the European Nucleotide Archive (ENA)/GenBank and the Barcode of Life Data (BOLD) Systems.

Total genomic DNA was extracted from eight female individuals using a DNeasy Blood & Tissue Kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions. Polymerase chain reaction (PCR) amplification was performed using the primers LCO1490 and HCO2198 [37] and Ex Taq DNA polymerase (TaKaRa Bio, Kusatsu, Japan). The PCR composition and thermal cycling profile have been described elsewhere [38]. After purification of PCR products using ExoSAP-IT Express (Thermo Fisher Scientific, Waltham, MA, USA), sequencing was outsourced to FASMAC (Atsugi, Japan). Sequencing reactions were performed bidirectionally using the same primers as for PCR with the BigDye Terminator v3.1 Cycle Sequencing Kit (Thermo Fisher Scientific, Waltham, MA, USA), followed by purification with the BigDye XTerminator Purification Kit (Thermo Fisher Scientific, Waltham, MA, USA), and analysis on an Applied Biosystems 3730xl DNA Analyzer (Thermo Fisher Scientific, Waltham, MA, USA). Sequences obtained from both strands were assembled using GeneStudio Professional version 2.2.0.0 (GeneStudio Inc., Suwanee, GA, USA). We retrieved available COI sequence data for previously identified Monodontomerus species from the National Center for Biotechnology Information (NCBI) and the Barcode of Life Data Systems (BOLD). Including the samples sequenced in the present study, multiple sequence alignment was performed using the ClustalW program [39], and pairwise p-distances among species were calculated and a maximum likelihood (ML) phylogenetic tree was constructed using MEGA11 version 11.0.13 [40], where Anneckeida sp. 1 (GenBank accession no. MF956301) was used as the outgroup [41].

2.3. Development of Microsatellite DNA Markers and Population Genetic Analysis

Genomic DNA was extracted from a single female individual collected at Hiratsuka in 2023, using the methods described above. Microsatellite markers were developed by a commercial service provider (Seibutsu Giken Inc., Sagamihara, Japan). Briefly, the genomic DNA was used to construct paired-end sequencing libraries with the MGIEasy Fast FS DNA Library Prep Set v2.0 (MGI Tech, Shenzhen, China), followed by circularization, DNA nanoball (DNB) preparation, and paired-end (150 bp) sequencing on a DNBSEQ-T7 platform (MGI Tech, Shenzhen, China). Reads were de novo assembled using MaSuRCA v3.4.2, genome completeness was evaluated with BUSCO v5.4.7 (eukaryota dataset), and microsatellite loci were identified using QDD v3.1.2 [42].

Among the identified microsatellite loci, we evaluated the amplification and polymorphism of each designed primer pair. First, loci were selected using relaxed criteria, retaining dinucleotide and trinucleotide repeat motifs with at least seven repeat units. By this screening, we selected 24 dinucleotide and 48 trinucleotide microsatellites with repeat numbers ranging from 7 to 18. Second, we further filtered the identified microsatellite loci to retain those with the longest repeat motifs. Consequently, we retained 89 dinucleotide loci with 19–32 repeat units and 7 trinucleotide loci with 16–27 repeat units. For the polymorphism evaluation, we used 17 female individuals from Hiratsuka population (Table S3), from which DNA was extracted from the head and thorax to avoid contamination by male sperm. PCR was performed using the M13-tail technique [43], with the PCR composition and thermal cycling profile described elsewhere [38]. The amplified PCR fragments were electrophoresed using an ABI 3130 capillary sequencer (Thermo Fisher Scientific, Waltham, MA, USA), and fluorescence profiles were analyzed using the Peak Scanner software version 2.0 (Thermo Fisher Scientific, Waltham, MA, USA).

For the population genetic analyses described below, we selected polymorphic loci with a large number of alleles and relatively even allele frequency distributions among the individuals examined. The analyses were conducted using 52 female individuals collected from the Hiratsuka population between 2016 and 2023 (Table S3). Linkage disequilibrium (LD) among the selected loci was tested using GENEPOP version 4.7 [44]. Significance thresholds for multiple tests across all pairwise combinations of loci were corrected using the false discovery rate (FDR) [45]. To assess whether significant LD among the loci reflected physical linkage or genome-wide non-independence due to inbreeding or population subdivision, the standardized index of association ( $[eqn]$ ) was calculated in R using the “poppr” package [46].

Observed and expected heterozygosities (HO and HE), inbreeding coefficient (FIS) according to Weir and Cockerham (1984) [47], and tests for deviations from Hardy–Weinberg equilibrium (HWE) were performed using GENEPOP version 4.7 [44]. The multiple p values for all the loci were corrected using FDR. Genetic relationships among individuals were examined using Bruvo’s distance, which is specifically designed for microsatellite data and incorporates a stepwise mutation model [48]. A minimum spanning network (MSN) was constructed from the obtained distance matrix to visualize genetic relationships among individuals, with connections between nodes weighted by Bruvo’s distance. Bruvo’s distance was calculated and the MSN was constructed using the R package “poppr” [46], and network visualization and layout optimization were performed using the R package “igraph” [49].

3. Results

3.1. Parasitism in the Field

Monodontomerus sp. was collected while parasitizing prepupae of M. sculpturalis or its social parasite Coelioxys fenestrata Smith, 1873 (Table S2). While the host bees constructed brood cells sequentially within the bamboo canes, resulting in multiple potential hosts arranged in a series, Monodontomerus sp. showed a strong tendency to attack the host located either at the innermost or the outermost position within the cane (Table S2; randomization test, p < 0.001). Until the 2000s or the early 2010s, parasitism by Monodontomerus sp. was not detected or remained low at all the sites investigated, with parasitism rates of at most 5% per host and 9% per bamboo cane (Table S1). However, at the site in Hiratsuka, where surveys were continued beyond the late 2010s, parasitism rates began to increase around 2018 (Figure 1; Table S1). At this site, parasitism rates increased significantly across years, both per host and per bamboo cane (χ^2^1 = 35.0, p < 0.001 and χ^2^1 = 47.1, p < 0.001, respectively).

3.2. Species Identification and Phylogenetic Relationship

Our morphological examination indicated that the Monodontomerus species in the present study was M. obscurus. The specimens exhibited the following morphological features: frenal area on scutellum with superficial striae; propodeum with a strong median carina, which is not divided at base; basal cell on forewing with a setal row below the submarginal vein; forewing with infuscate area around stigmal vein; ovipositor sheath as long as metasoma, 1.5–1.9 times as long as hind tibia. We were able to examine the morphology of both male and female individuals, and the male morphology clearly distinguished this species from M. osmiae. The specimens were also morphologically distinct from M. laticornis Grissell & Zerova, 1985 and M. rugulosus Thomson, 1876, which have been recorded as being reared from Megachilidae species in the Old World [36].

Molecular analysis of the COI region showed that all eight individuals investigated shared a single haplotype (the sequence data was deposited in DDBJ/ENA/GenBank under the accession number LC915967). Among the available sequences of Monodontomerus species, pairwise p-distances were calculated and an ML phylogenetic tree was constructed (Table S4 and Figure 2). The haplotype in the present species was most closely related to M. cf. osmiae collected in South Korea, with a percent identity of 98.4% (Table S4 and Figure 2). In contrast, the percent identity with M. obscurus from a European population was only 86.9%. In the present study, we treat the Monodontomerus species parasitizing M. sculpturalis in Japanese populations as M. obscurus based on morphological identification. However, this taxonomic assignment may require reconsideration in future studies.

3.3. Microsatellite DNA Analysis

From the genome assembly, primer pairs were designed for a total of 6049 dinucleotide and 378 trinucleotide microsatellite loci with at least seven repeat units. First, among loci retained under the relaxed criteria, we tested amplification and polymorphism for 24 dinucleotide and 48 trinucleotide loci and confirmed successful amplification for 21 dinucleotide and 45 trinucleotide loci. However, all the loci were monomorphic except one, in which only two individuals were heterozygous with the same alleles. Second, we applied a more stringent filtering step by prioritizing loci with the longest repeat motifs from all identified loci, and tested 89 dinucleotide and 6 trinucleotide loci. Among these, 88 dinucleotide and 6 trinucleotide loci were successfully amplified, and polymorphism (two to five alleles) was detected in 51 dinucleotide and 3 trinucleotide loci (Table S5; the sequence data of the microsatellite loci with polymorphism were deposited in DDBJ/ENA/GenBank under the accession numbers LC915913–LC915966). Notably, although these loci had polymorphism, allele frequencies were strongly dominated by single alleles, whose frequencies were never lower than 0.68 (mean = 0.85 across loci; Table S5).

From these polymorphic loci, we selected 22 loci that were suitable for subsequent population genetic analyses with multiplexing (Table 1). We tested LD for all pairwise combinations of these 22 loci, and significant LD was detected between several pairs of loci even after FDR correction (Table S6). Such significant LD can arise not only from genetic linkage between loci located adjacently on the same chromosome but also from non-random mating such as inbreeding or genetic drift in subdivided populations [50,51]. The standardized index of association was significantly greater than zero ( $[eqn]$ = 0.133; p < 0.001), indicating genome-wide linkage disequilibrium rather than disequilibrium restricted to specific pairs of loci. This pattern suggests that the observed LD most likely reflects inbreeding and population subdivision, which is consistent with the biology of the species and the results below.

When the 22 screened loci were examined with an expanded sample of 52 individuals, the allele frequency distributions remained strongly skewed toward single alleles, with frequencies of the most common alleles being at least 0.70 (mean = 0.81; Table 1). For every locus examined, observed heterozygosity (HO) was consistently lower than expected heterozygosity (HE), leading to high inbreeding coefficients (F) and significant deviations from Hardy–Weinberg equilibrium (Table 1), which indicates strong inbreeding. An MSN based on Bruvo’s distance showed that six individuals with completely identical genotypes across all loci occupied a central position, with all other individuals radiating from this central genotype (Figure 3). These six individuals located at the center were homozygous at all the loci with the most frequent alleles in this population.

4. Discussion

Our field data, together with historical observational records [24,25,27], suggest that the Japanese population of Monodontomerus parasitizing M. sculpturalis was established prior to 2000 and, after persisting at low density for an extended period, has expanded in abundance in recent years. Our microsatellite analyses of a population revealed an almost complete lack of genetic diversity, except at loci with long repeat motifs, where allele frequencies were strongly skewed, and the MSN exhibited a star-like pattern. Taken together, these genome-wide microsatellite results indicate that the population was founded by very few individuals, most parsimoniously a single mated female. The absence of genetic diversity reflects limited variation in the founders and the mating males, polymorphism restricted to long repeat motifs with rare alleles likely reflects post-invasion mutation, and the star-like genealogical structure indicates recent radiation from the founder genotype.

Our morphological identification indicated that the present Monodontomerus species was M. obscurus, whereas the molecular data of the COI region showed that the percent identity between the present species and M. obscurus from a European population was only 86.9%. Morphological evidence excluded the possibility that the present species corresponds to any other previously described Monodontomerus species, including M. osmiae, suggesting that the origin of the present species is not attributable to other species previously distributed in Japan. This indicates that M. obscurus, or a closely related species derived from it, likely invaded Japan and subsequently established a population. The present species is relatively close to M. cf. osmiae from South Korea, showing 98.4% identity in the COI sequence, suggesting that they may represent the same species. The morphological similarity between females of M. obscurus and M. osmiae is consistent with this interpretation [35]. Including the present species, the classification of closely related species within the genus Monodontomerus is expected to be clarified by future studies integrating detailed morphological examinations and molecular analyses.

Polymorphism analyses of the microsatellite loci we developed revealed that genetic diversity in the present species is extremely low. When loci were selected using standard screening criteria, genetic diversity was almost completely absent. This level of genetic diversity is remarkably low even in comparison with other parasitoid wasp species in the superfamily Chalcidoidea that commonly experience inbreeding [32,33]. In microsatellites, loci with a larger number of repeat motifs are known to have higher mutation rates [52,53]. Accordingly, in our analyses, polymorphism was detected only when loci with long repeat motifs were examined.

These results suggest that the genetic variation observed in the present study did not originate from native populations carried by the founders, but instead arose through mutation from founder alleles after invasion. Conversely, the absence of genetic variation at the time of invasion is likely attributable to the population having been founded by a single individual or an effectively single genetic lineage. Monodontomerus species frequently experience inbreeding [27], as also shown in the present study. Repeated inbreeding may therefore have resulted in the founding female that was homozygous at all loci, with her mating partner also possessing identical alleles. This interpretation is further supported by the strong skew in allele frequencies toward a specific allele and by the star-shaped MSN. These patterns suggest that individuals carrying the most frequent alleles (corresponding to the genotype located at the center in the MSN) share the same genotype as the founder and her mate, from which novel alleles have subsequently arisen through mutation and radiated outward.

A large number of studies have reported that new populations are often initiated by a restricted number of founders despite experiencing bottleneck events. However, populations founded by single individuals are quite rare. As notable exceptions, successful population establishment by a single individual has been documented under artificial introduction programs for biological control in a chrysomelid beetle, a psyllid bug, and a dryinid parasitoid wasp [29,54,55]. Molecular analyses have also provided evidence for single-individual founding events in solitary and social bees [56,57]. However, in contrast to the present study, these studies showed that no novel alleles were detected in newly established populations, indicating that these populations consisted solely of alleles carried by the founders. Specifically, the honey bee Apis cerana that invaded Australia exhibits extreme polyandry, maintaining genetic diversity through alleles contributed by several dozens of the mating partners of the single female founder [57]. In addition, selection favoring equal allele frequencies to avoid genetic load is expected to sustain the variation introduced by the founders [58,59]. In the present study, because microsatellite loci are considered neutral with respect to adaptation and genetic load caused by bottlenecks is unlikely to be severe in the present species, alleles newly generated by mutation are expected to fluctuate randomly owing to genetic drift and to remain at low frequencies currently.

In conclusion, the present study shows that the population of M. obscurus was most likely founded by a single female under an extreme founder event. Despite severe bottleneck events after invasion, the present species is expected to have suffered little from the negative effects, owing to the biology that involves frequent inbreeding and the genetic system that reduces the genetic loads. Taken together, these findings demonstrate that extreme founder events do not necessarily preclude successful population establishment under natural conditions, and that species-specific ecological and genetic characteristics can play a key role in shaping invasion outcomes.