Community Structure of Aquatic Insects Adapted to Lentic Water Environments, and Fine-Scale Analyses of Local Population Structures and the Genetic Structures of an Endangered Giant Water Bug Appasus japonicus

Environments such as floodplains and the marshlands of rivers, lakes and ponds, are important habitats for aquatic insects adapted to lentic water conditions. In addition, ponds and paddy fields artificially created for agriculture are also important alternative habitats for lotic water-dependent wildlife. In this study, we focused on aquatic insects in ponds in the Matsumoto Basin, located in the center of Japan. Although this is an urbanized area, aquatic animals adapted to floodplains inhabit it at a relatively high density for Japan. We conducted a multifaceted evaluation of the environments of the 33 ponds in this region and conducted a survey of the aquatic insect fauna inhabiting them. In this study, we conducted quantitative sampling, focusing on two insect orders adapted to large-scale lentic water environments (i.e., Heteroptera and Coleoptera), and observed five species of three families and 16 species of five families from the Matsumoto Basin, respectively. Within these species, eight endangered species were included. Furthermore, we carried out a genetic structure analysis for the giant water bug, Appasus japonicus, inhabiting these ponds in high density, and conducted a comparative evaluation of their genetic diversity between these ponds. A total of 530 specimens of A. japonicus were genetically analyzed for the mitochondrial DNA COI region, and 26 haplotypes were observed. The degree of genetic diversity between the ponds was clearly demonstrated. In addition, we discussed the wintering possibilities for the giant water bugs based on their corresponding surrounding environmental factors, and comprehensively discussed their “source−sink” relationships in this region. Therefore, this is a comprehensive study focused on the relevant environmental factors, diversification of their community structures, their population structures, and their genetic structure at a fine scale.


Introduction
Environments such as floodplains and the marshes of rivers, lakes, and ponds, are important habitats for aquatic insects adapted to lentic water conditions [1][2][3]. Such lentic water environments are very important not only for aquatic insects adapted to lentic water conditions, but also for many aquatic of types ranging from artificial pond set in concrete to places having a high naturalness. We were not able to include as research sites some ponds located on private properties for which we were unable to obtain permission from the landowners. Regarding population and genetic structure analyses, a preliminary assessment was undertaken collecting a wide variety of insects in the region in order to identify a suitable species as a target of study. From the results of the information gathered, the giant water bug A. japonicus, was selected as a target species.
Insects 2020, 11, x FOR PEER REVIEW 3 of 20 A total of 33 ponds were set as research sites over a wide area almost covering the entire Matsumoto Basin (Table 1; Figures 1 Figure S1). These 33 ponds include a number of types ranging from artificial pond set in concrete to places having a high naturalness. We were not able to include as research sites some ponds located on private properties for which we were unable to obtain permission from the landowners. Regarding population and genetic structure analyses, a preliminary assessment was undertaken collecting a wide variety of insects in the region in order to identify a suitable species as a target of study. From the results of the information gathered, the giant water bug A. japonicus, was selected as a target species.    In addition, with respect to other major lentic water adapted aquatic insects collected in the quantitative sampling, species identification was carried out in the field, and the number of each species collected was recorded at the same time. All specimens used in this study, including specimens that have undergone genetic analysis, are stored in the Natural Museum of Shinshu University (SHIN-Z No. 8001-8848).

Measurement of Environmental Factors at Each Pond
We measured thirteen environmental factors as listed below: (1) GPS location (latitude, longitude and altitude), (2) surface area of each pond, (3) perimeter of each pond, (4) percentage of concrete shoreline, (5) percentage of vegetated shoreline, (6) percentage of vegetated total pond surface area, (7) gradient of the pond bed in the zone 3 m from the shoreline (i.e., gradient based on depth measurements taken away from the shore toward the center at three arbitrary points around the pond), (8) opening rate of sky, (9) clarity of pond water, (10) water quality (i.e., concentration of the total nitrogen, T-N; ammonium nitrogen, NH 4 + -N or NH 3 + -N; phosphate-phosphorous, PO 4 3− ), (11) number of waterways flowing into the pond, (12) number of waterways flowing out of the pond, and (13) the usage conditions of the surrounding land (e.g., housing, forest, farmland, paddy fields). The analysis of factors (2) and (3), determination of the shape of each pond researched was based on geographical maps and/or aerial photographs of the Geospatial Information Authority of Japan, and then by analysis using the software Kashmir 3D ver. 9.2.9 [21]. The analysis of factor (5) evaluation was conducted by visually allocating 5 grades. Factor (9) was measured using a self-made transparency meter. Factor (10) was measured using three-pack test kits, WAK-TN-I, WAK-NH 4 (C), and KR-PO 4 (Kyoritsu Rika Kenkyu, Tokyo, Japan). Six samples of 100 mL pond water were sampled per pond and stored at 4 • C, and thereafter all samples were moved together a 25 • C room, and then measured in the 25 • C room after standing for 12 h. All of this water quality measurement work was carried out within 3 days from the time of water sampling.

Sampling for Population and Genetic Structure Analyses of A. japonicus
Sampling for population and genetic structure analysis was carried out for all of the 33 targeted ponds late in the autumn of 2015. First, the shoreline from which it was possible to perform sampling was equally divided into 6 sections, from which we carried out our quantitative sampling over the predetermined sampling period (i.e., three minutes sampling per zone × 6 zones). We mainly targeted large specimens of two orders (i.e., Heteroptera and Coleoptera), using D-frame sampling nets with a mesh size of 2 mm. Many water striders also inhabited those ponds we surveyed, but we excluded them in preference for the targeted insects, because they escaped frequently from the sampling net during our quantitative survey.
In this survey, when more than 20 individuals of A. japonicus were collected, 20 of them were used as genetic analysis specimens. If less than 20 individuals were collected, additional qualitative sampling to obtain the full 20 individuals was carried out. However, in some ponds, we could not eventually collect the 20 required specimens. With respect to other major lentic water adapted aquatic insects collected in the quantitative sampling, species identification was carried out in the field, and the number of each species collected was recorded at the same time. Due to the large proportion of endangered species collected, field identification was performed.

Survey on Overwintering Rate for A. japonicus
The same quantitative sampling was conducted in early spring of 2016 for the 29 ponds from which A. japonicus specimens were collected in the previous sampling in the late autumn of 2015 (Table S1). Based on these results, we determined the overwintering ponds in which A. japonicus was collected in both the late autumn of 2015 and the early spring of 2016. The ratio of the late autumn to the early spring quantitative sampling was regarded as the "overwintering success rate". However, in calculating the overwintering success rate, only those ponds, from which the number of A. japonicus collected in the late autumn research was greater than the median, were included (i.e., >10 specimens were collected in the late autumn of 2015). This overwintering success rate was used for the responsive variable of generalized linear mixed model (GLMM) analysis to reveal a relationship between overwintering success rate and the percentage of concrete shoreline.

Redundancy Analysis (RDA)
Redundancy Analysis (RDA) was conducted to examine the relationship between community structure and environmental factors, using the software R.3.3.2 [22]. Data on the species composition and their individual numbers of lentic water insects collected in each study site were used as the response variables. For the explanatory variables, we used a dataset of 7 environmental factors (altitude, surface area of each pond, percentage of vegetated shoreline, percentage of concrete shoreline, gradient of the pond bed in the zone 3 m from the shoreline, number of waterways flowing into the pond, concentration of the total nitrogen) excluding those that showed little difference among the survey environmental factors and the number of invasive alien species.

Generalized Linear Model (GLM) Analysis
A generalized linear model (GLM) analysis was conducted to examine the relationship between the population structure of each of the detected dominant species and the environmental factors, using the software R.3.3.2 [22]. Since the four species, A. japonicus, Ranatra chinensis Mayer, 1865, Ranatra unicolor Scott, 1874, Rhantus suturalis (MacLeay, 1825), were collected from more than half the number of surveyed ponds, and more than 100 individuals in total were collected, respectively (i.e., dominant species), they were subject to this GLM analysis. In addition, as for the species Laccotrephes japonensis Scott, 1874, although their collected number was not so large, they were also added to our analysis as they were collected from more than half the number of surveyed ponds. The relative number of collected individuals of each species in our quantitative sampling was used as the response variable. The measured environmental data and the number of collected invasive alien species were used as explanatory variables.

Genetic Analysis
As specimens for genetic analysis (mtDNA COI region) of A. japonicus, individuals collected by quantitative sampling, and individuals collected by supplementary qualitative sampling were used. However, although the designated sample of 20 individuals were analyzed properly in 25 ponds, the number collected did not reach 20 individuals for the remaining 8 ponds. The genetic analysis of A. japonicus, direct sequencing was performed in accordance with the same method utilized in our previous study [19,20].
Total genomic DNA was extracted from leg tissue, using the DNeasy Blood and Tissue Kit (QIAGEN, Hilden, Germany) according to the manufacturer's instructions. The total DNA was used for amplifying DNA fragments by polymerase chain reaction (PCR) with a set of universal primers  Table 1.
Phylogenetic analyses were performed using the Bayesian method [23], with MrBayes5D version 3.1.2 [24]; and the maximum likelihood method [25], using RAxML version 8.2.9 [26]. Appasus major Esaki 1934, which is a sister species of A. japonicus, were used as the outgroup for Bayesian and ML analyses. Nodal support was measured with the posterior probabilities in MrBayes5D, and 1000 bootstrap replicates in RAxML [27]. Prior to the ML and Bayesian phylogenetic estimations, the program Kakusan4 [28] was used to select an appropriate model based on Schwarz's Bayesian Information Criterion (BIC) [29]. Bayesian analysis used 100 million Markov Chain Monte Carlo (MCMC) cycles with a sampling frequency of 1000. To obtain a consensus tree, data from the initial 10 million cycles were discarded as burn-in.

Analysis between Pairwise Genetic and Geographic Distances
Mantel test was conducted in AIS (Alleles in Space) software [30] to assess the significance of isolation by distance (IBD) between populations with 1000 random permutations on matrices of p-distance values and the geographical distances.

Statistical Analyses Based on the Genetic Data Obtained for A. japonicus
Cluster analysis was applied as per the Ward's method applying the Euclidean distances to assess the similarity of the genetic structures of A. japonicus between study sites. The statistical package "mass" by the software R3.3.2 was used [22]. In this calculation, we applied the Chao Index, as one haplotype (H1) showed extremely high dominance [31,32].
In addition, we used nonmetric multidimensional scaling (NMDS) in PRIMER version 6 [33] to assess the similarity of the genetic structure of A. japonicus between study sites. We applied a dissimilarity matrix obtained from the Bray-Curtis dissimilarity calculation [34].

Quantitative Survey of Community Structure among Lentic Water Adapted Aquatic Insect Fauna
The results of our quantitative survey for 33 selected ponds located in the Matsumoto Basin are shown in Figure S1. In this research, only the main lentic water adapted aquatic insects were targeted, and water striders (i.e., Gerridae) were excluded from the survey. Some of the species collected were listed in one of the endangered species categories of the Red List of Japan and/or Nagano Prefecture. Heteropteran aquatic insects were collected more frequently. Among them, the frequencies of the giant water bug, Appasus japonicus, and a backswimmer, Notonecta triguttata Motschulsky, 1861, being collected, were extremely large. On the other hand, there were ponds from which no aquatic insects targeted in this study were collected (pond No. 13, 14 sites).
In addition, the result confirming the presence or absence of some major invasive alien fauna is also shown in Table S1. In each pond, very few species (i.e., 0-2 species) of invasive alien fauna were observed in this study. Table 1 shows the survey results of each environmental factor for the 33 ponds located in the Matsumoto Basin. Table S1 shows the number of individuals collected in the quantitative survey of A. japonicus. Specimens of A. japonicus were collected from 29 ponds in total, and the maximum number of specimens collected at a given time (18 min: 3 min × 6) was 99 (No. 26), and the minimum number of captures was 0 (No. 13,14,27,28). The GLM analysis for correlations between the numbers captured of A. japonicus and the corresponding environmental factors showed significant correlation with the following five factors: (2) surface area of each pond, (4) percentage of concrete shoreline, (5) percentage of vegetated shoreline, (6) percentage of vegetated total pond surface area, (7) gradient of the pond bed in the zone 3 m from the shoreline, (12) number of waterways flowing out of the pond.

Survey on the Successful Overwintering Rate of the Giant Water Bug A. japonicus
The quantities collected of the giant water bug, A. japonicus, in the quantitative surveys of the autumn of 2015 and early spring of 2016, are shown in Table S1. The highest overwintering rate was observed at the No. 21

Results of Genetic Analysis of A. japonicus
From the 29 ponds where the specimens of A. japonicus were collected, 20 individuals were arbitrarily extracted from those specimens collected at each pond and genetically analyzed. With respect to the ponds where the number of captures was less than 20 individuals, as many of the specimens as possible were analyzed. Genetic analysis (mtDNA COI region) was performed on a total of 530 A. japonicus specimens, and 26 haplotypes were observed. The relationships between all observed 26 haplotypes are shown in Figure 3. The most dominant haplotype (i.e., H1 haplotype) was the overwhelming majority of specimens collected, and it was observed in all 29 ponds examined. The second and third most dominant haplotypes (H6 and H5 haplotypes) were also observed in many of the ponds (Table 2, Figure 3). The majority of haplotypes have a single step base substitution relationship from the dominant H1 haplotype, and they all had a substitution relationship of, at the most, 6 steps ( Figure 3). The composition of the haplotypes observed in each pond, and the results of haplotype and nucleotide diversity analysis at the intra-pond level are shown in

Results of Genetic Analysis of A. japonicus
From the 29 ponds where the specimens of A. japonicus were collected, 20 individuals were arbitrarily extracted from those specimens collected at each pond and genetically analyzed. With respect to the ponds where the number of captures was less than 20 individuals, as many of the specimens as possible were analyzed. Genetic analysis (mtDNA COI region) was performed on a total of 530 A. japonicus specimens, and 26 haplotypes were observed. The relationships between all observed 26 haplotypes are shown in Figure 3. The most dominant haplotype (i.e., H1 haplotype) was the overwhelming majority of specimens collected, and it was observed in all 29 ponds examined. The second and third most dominant haplotypes (H6 and H5 haplotypes) were also observed in many of the ponds (Table 2, Figure 3). The majority of haplotypes have a single step base substitution relationship from the dominant H1 haplotype, and they all had a substitution relationship of, at the most, 6 steps ( Figure 3). The composition of the haplotypes observed in each pond, and the results of haplotype and nucleotide diversity analysis at the intra-pond level are shown in Figure 4 H1 H2 H3 H4 H5 H6 H7 H8 H9 H10 H11 H12 H13 H14 H15 H16 H17 H18 H19 H20 H21 H22 H23 H24   Based on the results of these genetic analyses, to obtain the comparative genetic structures at the intra-pond level, a cluster analysis and an NMDS analysis were conducted, the results of which are shown in Figures 5 and 6. The result of the cluster analysis was largely divided into four clusters ( Figure 5). The respective geographical distribution of their ponds, constituting these four genetic clusters, is shown in Figures 4 and 7. It was observed that the geographically close ponds did not necessarily have a correspondingly similar genetic structure. On the other hand, NMDS analysis using the same data set showed a slightly different result from the cluster analysis ( Figure 6). Although the groups with the yellow and green backgrounds, were positioned as being different clusters from each other in the cluster analysis, they were shown to be almost the same group in this NMDS analysis. Regarding the remaining two clusters indicated by their red and blue background colors, their distinct grouping was also supported by the NMDS analysis. Figure 7 shows a radar chart showing the relationships between the environmental factors, which are suggested to correlate with the population structure, and genetic diversity of the A. japonicus populations. For the purposes of this study, each environmental factor was evaluated as being one of five grades. Although it is also described in the legend of Figure 7, the median value of each evaluated factor was given the grade of three. Then, the radar chart was prepared using the five grades of evaluation, with a maximum evaluation of the 5th grade, and a minimum evaluation of the 1st grade. That is, the larger the area and the more well-balanced the chart, the more optimal the environmental conditions for maintaining A. japonicus populations and their genetic diversity (e.g., pond No. 1, 18, 20, 22, 25 and 31 sites). Based on the results of these genetic analyses, to obtain the comparative genetic structures at the intra-pond level, a cluster analysis and an NMDS analysis were conducted, the results of which are shown in Figures 5 and 6. The result of the cluster analysis was largely divided into four clusters ( Figure 5). The respective geographical distribution of their ponds, constituting these four genetic clusters, is shown in Figures 4 and 7. It was observed that the geographically close ponds did not necessarily have a correspondingly similar genetic structure. On the other hand, NMDS analysis using the same data set showed a slightly different result from the cluster analysis ( Figure 6). Although the groups with the yellow and green backgrounds, were positioned as being different clusters from each other in the cluster analysis, they were shown to be almost the same group in this NMDS analysis. Regarding the remaining two clusters indicated by their red and blue background colors, their distinct grouping was also supported by the NMDS analysis. Figure 7 shows a radar chart showing the relationships between the environmental factors, which are suggested to correlate with the population structure, and genetic diversity of the A. japonicus populations. For the purposes of this study, each environmental factor was evaluated as being one of five grades. Although it is also described in the legend of Figure 7, the median value of each evaluated factor was given the grade of three. Then, the radar chart was prepared using the five grades of evaluation, with a maximum evaluation of the 5th grade, and a minimum evaluation of the 1st grade. That is, the larger the area and the more well-balanced the chart, the more optimal the environmental conditions for maintaining A. japonicus populations and their genetic diversity (e.g., pond No. 1, 18, 20, 22, 25 and 31 sites).

Survey on the Successful Overwintering Rate of the Giant Water Bug A. japonicus
The quantities collected of the giant water bug, A. japonicus, in the quantitative surveys of the autumn of 2015 and early spring of 2016, are shown in Table S1. The highest overwintering rate was observed at the No. 21  The most important requirement for ponds to function as sources is for the aquatic insects can successfully overwinter. In this study on overwintering of A. japonicus, it was found that the overwintering success rate varies greatly depending on the target ponds. In addition, it was clearly shown that the success rate of wintering has a strong negative correlation with the concrete revetment rate (Figure 8).

Survey on the successful overwintering rate of the giant water bug A. japonicus
The quantities collected of the giant water bug, A. japonicus, in the quantitative surveys of the autumn of 2015 and early spring of 2016, are shown in Table S1. The highest overwintering rate was observed at the No. 21 pond site (55.0%, 11/20). Thereafter in descending order there was; 41.7% (10/24)  The most important requirement for ponds to function as sources is for the aquatic insects can successfully overwinter. In this study on overwintering of A. japonicus, it was found that the overwintering success rate varies greatly depending on the target ponds. In addition, it was clearly shown that the success rate of wintering has a strong negative correlation with the concrete revetment rate (Figure. 8).   Regarding the five grades of evaluation of each factor, the median value of each evaluated factor was given the grade of three, with a maximum evaluation of the 5th grade, and a minimum evaluation of   Regarding the five grades of evaluation of each factor, the median value of each evaluated factor was given the grade of three, with a maximum evaluation of the 5th grade, and a minimum evaluation of Figure 7. Radar chart of each research pond showing the following environmental factors: population density and genetic diversity of Appasus japonicus vs. environmental factors at each pond (i.e., surface area, concrete revetment rate, gradient from the shore, number of inflowing waterways). Regarding the five grades of evaluation of each factor, the median value of each evaluated factor was given the grade of three, with a maximum evaluation of the 5th grade, and a minimum evaluation of the 1st grade. That is, the larger the area and the more well-balanced the chart, the more optimal the environmental conditions for maintaining A. japonicus populations and their genetic diversity.

Environmental Factors Affecting the Community and Population Structure of Aquatic Insects
In this study we targeted 33 ponds in a single basin, and compared the pond environments, the fauna and species diversity of aquatic insects, and the genetic structure and genetic diversity of a giant water bug, Appasus japonicus (Tables 1 and 2, Table S1, Figures 1-8, Figures S1 and S2). This is the first case study in which a detailed fine-scale survey has been conducted on aquatic insects that inhabit ponds in Japan. Although in a previous study no genetic analysis was conducted, it did investigate the effects of various environmental factors in irrigation ponds on their lentic aquatic insect community structures, using the "Canonical Correspondence Analysis (CCA)" method [35]. In the previous study, the species richness of emergent plants in irrigation ponds was evaluated as being the highest contributing factor towards the community structure of the lentic aquatic insects present. The Spearman's rank correlation coefficient test reported a negative correlation between the number of aquatic insects and the percentage of concreted shoreline, and a positive correlation to the total percentage of vegetated pond surface area [35]. In the previous study, the order Odonata (i.e., dragonflies and damselflies) was included for targets with heteropteran and coleopteran aquatic insects. Although the odonatan species accounted for an especially large proportion in the community, it was a somewhat different study target to the lentic water insect community targeted in our research. However, the pattern of a high correlation between the percentages of concrete shoreline to the corresponding community structures of aquatic insects showed agreement between the previous study and our study. In addition, the result of the "Redundancy Analysis (RDA)" in this study indicated that the gradient of the pond bed to the shoreline was the environmental factor having the largest contribution toward the community structure of the corresponding lentic water inhabiting insects. Therefore, it is generally suggested that the shoreline environments and vegetation play an extremely important role in terms of the species diversity of lentic aquatic insects.
On the other hand, in our study of water scorpions, Laccotrephes japonensis, more than expected were captured in those ponds where the percentage of concrete shoreline was somewhat higher (ca. 60%). This trend was also reflected in previous studies. In such cases of ponds having a high percentage of concrete shoreline as described above, where overhanging land-based plants reached the water surface, water scorpions were collected in such microhabitats. The percentage of vegetated shoreline and vegetated total pond surface area measured in this study included cases where overhanging land plants reached the water surface, and it covered not only herbaceous plants but also woody plants. In our study, since water scorpions were collected from 19 ponds (ca. 58%), they may have originated from several robust habitats that exist abundantly in the surroundings.

Environmental Factors That Influence the Intra-Specific Structure in Several Dominant Species of Aquatic Insects
As for the dominant aquatic insect species (i.e., the belostomatid water bugs, A. japonicus, Ranatra chinensis, Ranatra unicolor, Rhantus suturalis and L. japonensis), and with respect to the environmental factors assessed using the generalized linear model (GLM), it revealed that the relationship between the number of individuals captured relative to the corresponding environmental factors, tended to vary depending on the species.
The analyses of the relationships between the community structures of aquatic insects with the environmental factors, were expected to reveal that the shoreline environment was likely to be an extremely important factor for all of the examined species of aquatic insects. However, the significance of the shoreline gradient and/or the percentage of concrete shoreline was recognized in the only two species, i.e., A. japonicus, L. japonensis. Both of these species are endangered species listed on the "Red List" [13], and so it was suggested that both species are susceptible to environmental changes due to artificial modification of the shoreline environments. Especially for A. japonicus, concrete revetment decrease the overwintering success rate (Figure 8).
The interspecific differences with regard to the influence of the surrounding environments observed for each species in this study may reflect the differences in their dispersal capability (i.e., A. japonicus and L. japonensis vs. R. chinensis and R. unicolor). Appasus japonicus and L. japonensis have a low dispersal ability compared to the Ranatra species, and their distribution pattern tends to be more isolated and scattered. It is thought that they are more susceptible to artificial influences in their habitat environments. On the other hand, the remaining two species of the genus Ranatra have a relatively high dispersal ability, whereby it is possible to move between various water bodies, and they are thought less likely to be affected by changes in their habitat.
Among the four heteropteran water bugs which were collected in this study, only one species, R. unicolor exhibited patterns of environmental preference for each environmental factor contrary to the other three species. In addition, only within A. japonicus, was significance observed between the population density and the number of inflowing water sources. As the number of waterways flowing into the pond increased, the population density tended to decrease (Table 1 and Table S1). In such cases where the ponds are located in mountainous areas, it may be caused by the decreased temperature of the ponds due to the inflow of the cold mountain water. It was reported that the developmental speed of the Appasus bugs was strongly influenced by water temperature [36]. Unfortunately, in this study, we did not collect water temperature variance data between the surveyed ponds, and so we cannot discuss this matter more deeply than this. However, such differences in the degree of influence of each environmental factor may suggest basic patterns of habitat differentiation (i.e., some niche differentiation) among these water bug species.
Regarding the relationships with other species groups (i.e., the interspecific interaction), only in Rhantus beetles (Colymbetinae) was a relationship observed whereby the number of individuals collected decreased significantly when the number of invasive alien species collected together at the same location increased ( Figure S1). Among all the species studied, although they are all prey to other species, since the targeted four heteropteran bugs are less mobile than the Rhantus beetles, they are less likely to be found by predators, so they are less susceptible to invasive alien species rather than the Rhantus. Thus it is thought that it is less likely for the four heteropterans to be affected by predation pressure.

Characteristics of Genetic Structure Analysis of A. japonicus
All 26 haplotypes observed in our genetic structure analysis of A. japonicus in the Matsumoto Basin were grouped in "Clade I" among the three intraspecific lineages known so far [20]. This is an expected result when considering the regional nature of the Matsumoto Basin. However, it was interesting that so many haplotypes were observed within a single basin ( Table 2, Figures 4 and 5). Therefore, we think that the Matsumoto Basin populations constitute a robust regional group, in a habitat which is extremely well-suited to A. japonicus. In view of the situation wherein extinction is common across wide areas of the Japanese Archipelago, there are many lentic water environments in which A. japonicus thrives in considerably high density within the Matsumoto Basin. In fact, even though our survey was carried out to include several ponds with concrete shoreline as a target, the proportion of ponds in which A. japonicus could be confirmed was still high (i.e., 29/33 ponds).
Cluster analysis and NMDS analysis based on the genetic structure of A. japonicus, whereby ponds were classified into four pond groups, however, ponds within a relatively close range in the Matsumoto Basin did not always belong to the same group (Figures 5-7). This is because not only is dispersion occurring between neighboring ponds but also migration between ponds across the entire Matsumoto Basin is occurring, and as a result corresponding gene flow.
However, the result of the haplotype network analysis showed the tendency toward a typical "simultaneous dissipation" pattern, with the core dominant haplotype being the H1 haplotype ( Figure 3). That is, in several past glacial periods, it is thought that they underwent a bottleneck situation. Such derived haplotypes from the H1 haplotype, had a relationship in which 22 single base substitutions were detected, indicating a relatively recent divergence era. In addition, A. japonicus was found present in many ponds, it seems occasional migration between ponds seem to be occurring. In other words, it is inferred that the entire Matsumoto Basin functions as a typical "Metapopulation" of A. japonicus (Figure 4 and Figure S2). japonicus was found present in many ponds, it seems occasional migration between ponds seem to be occurring. In other words, it is inferred that the entire Matsumoto Basin functions as a typical "Metapopulation" of A. japonicus (Figure 4 and Figure S2).

"Source−Sink" Relationships at the Intra-Specific Group Level
Statistical analyses based on environmental factors, population density and genetic diversity of A. japonicus were carried out, based on the assumption that they would likely be significant to the source−sink relationships observable within the Matsumoto Basin for A. japonicus. First, among 33 targeted ponds, A. japonicus were collected in high density in some ponds (i.e., No. 1, 16, 20, 25 and 26 ponds), but not in other ponds by quantitative survey (i.e., No. 13, 14, 27 and 28 ponds; Figures 4  and 7). Furthermore, by adding in the viewpoint of genetic diversity, some ponds with high density and high genetic diversity, that is, with high robustness were found (i.e., No. 1 and 25 ponds; Figures  4 and 7). It was considered that such particular ponds in the Matsumoto Basin, which were judged to be particularly excellent habitats, took on the function as "sources" in the "Metapopulation" system, and were the hub of movement and dispersion in the horizontal direction.
On the other hand, there were several ponds with a large decrease in density the following spring, despite many A. japonicus inhabitants the previous autumn (i.e., No. 3, 6, 15, 31). These ponds are considered "sinks", like the low-density ponds originally (ponds with low overwintering success rate are likely to be typical "sink" ponds; Figures 7 and 8). These are considered temporary habitats used in the limited seasons from late spring to autumn. Therefore, within the fine-scaled geographical area of the Matsumoto Basin, many lentic habitats of A. japonicus connected by the "source−sink" relationships are close to each other, and the "Metapopulation" of the entire region is maintained well.

Conclusions
In this study, a quantitative survey of aquatic insects was carried out in lentic environments (i.e., 33 ponds) in the Matsumoto Basin located in the center of the Japanese Archipelago. In addition, a supplementary qualitative survey was also conducted. Various environmental factors were measured in all ponds. As a result of this, we observed five species of three families of Heteroptera and 16 species of five families of Coleoptera, including eight endangered species. Furthermore, we carried out a genetic structure analysis for the giant water bug, Appasus japonicus, inhabiting these ponds in high density and conducted a comparative evaluation of their genetic diversity between these ponds. A total of 530 specimens of A. japonicus were genetically analyzed for the mitochondrial DNA COI region, and 26 haplotypes were observed. The degree of genetic

"Source−Sink" Relationships at the Intra-Specific Group Level
Statistical analyses based on environmental factors, population density and genetic diversity of A. japonicus were carried out, based on the assumption that they would likely be significant to the source−sink relationships observable within the Matsumoto Basin for A. japonicus. First, among 33 targeted ponds, A. japonicus were collected in high density in some ponds (i.e., No. 1, 16, 20, 25 and 26 ponds), but not in other ponds by quantitative survey (i.e., No. 13, 14, 27 and 28 ponds; Figures 4  and 7). Furthermore, by adding in the viewpoint of genetic diversity, some ponds with high density and high genetic diversity, that is, with high robustness were found (i.e., No. 1 and 25 ponds; Figures 4  and 7). It was considered that such particular ponds in the Matsumoto Basin, which were judged to be particularly excellent habitats, took on the function as "sources" in the "Metapopulation" system, and were the hub of movement and dispersion in the horizontal direction.
On the other hand, there were several ponds with a large decrease in density the following spring, despite many A. japonicus inhabitants the previous autumn (i.e., No. 3, 6, 15 and 31 ponds). These ponds are considered "sinks", like the low-density ponds originally (ponds with low overwintering success rate are likely to be typical "sink" ponds; Figures 7 and 8). These are considered temporary habitats used in the limited seasons from late spring to autumn. Therefore, within the fine-scaled geographical area of the Matsumoto Basin, many lentic habitats of A. japonicus connected by the "source−sink" relationships are close to each other, and the "Metapopulation" of the entire region is maintained well.

Conclusions
In this study, a quantitative survey of aquatic insects was carried out in lentic environments (i.e., total of 33 ponds) in the Matsumoto Basin located in the center of the Japanese Archipelago. In addition, a supplementary qualitative survey was also conducted. Various environmental factors were measured in all ponds. As a result of this, we observed five species of three families of Heteroptera and 16 species of five families of Coleoptera, including eight endangered species. Furthermore, we carried out a genetic structure analysis for the giant water bug, Appasus japonicus, inhabiting these ponds in high density and conducted a comparative evaluation of their genetic diversity between these ponds. A total of 530 specimens of A. japonicus were genetically analyzed for the mitochondrial DNA COI region, and 26 haplotypes were observed. The degree of genetic diversity between the ponds was clearly demonstrated. In addition, we discussed the overwintering possibilities for the giant water bugs, based on their corresponding surrounding environmental factors, and comprehensively discussed their "source−sink" relationships in this region. Therefore, this is a comprehensive study focused on the relevant environmental factors, diversification of their community structures, their population structures, and their genetic structure at a fine scale. The clarity of source−sink relationships is very important in terms of the assessment of important bodies of water for the conservation of A. japonicus, which is also an endangered species. In particular, conservation of this carnivorous aquatic insect, also contributes to conservation of the ecological balance with respect to the many benthic species living under their "umbrella" within their robust habitat environments.
Supplementary Materials: The following are available online at http://www.mdpi.com/2075-4450/11/6/389/s1, Table S1: List of major lentic aquatic insects collected in the quantitative survey (early spring of 2016/autumn of 2015), and the observation of notable invasive alien fauna: observed (+) or unobserved (−). Figure S1: Pictures of all of study sites. Figure S2: Relationships between pairwise genetic distances (uncorrected p-distance) and pairwise geographical distances of the giant water bug, Appasus japonicus (530 individuals; correlation of genetic and geographical distances: r = −0.0021520714. Probability of observing a correlation greater than or equal to observed: p = 0.5704295704).