Open Access
Knowl. Manag. Aquat. Ecosyst.
Number 421, 2020
Article Number 37
Number of page(s) 9
Published online 21 August 2020

© S. Nagayama et al., Published by EDP Sciences 2020

Licence Creative CommonsThis is an Open Access article distributed under the terms of the Creative Commons Attribution License CC-BY-ND (, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. If you remix, transform, or build upon the material, you may not distribute the modified material.

1 Introduction

Invasive species impact biotic/abiotic components, lead to newly-modulated ecosystems in invaded regions, and often involve biodiversity loss and ecosystem degradation (Strayer et al., 2006; Pyšek and Richardson, 2010). Predation is one of the most prominent behaviors of invasive animals that influence indigenous ecosystems. Predation by invasive animals directly affects the population structure and distribution of native prey species, resulting in changes in community compositions and biological interactions (Strayer et al., 2006; Kurle et al., 2008). Therefore, diet is basic information that will allow us to understand the degree of impact that invasive animals have on biological communities.

Coypu (Myocastor coypus), also known as nutria, is a large semi-aquatic rodent that originated from South America (LeBlanc, 1994). This species has been introduced into many other countries outside their native range for fur farming, and feral populations have been established in North America, Europe, and Asia owing to escapes and releases from fur farms (LeBlanc, 1994; Carter and Leonard, 2002; Hong et al., 2015). Currently, the coypu has been selected as part of the “100 of the world's worst invasive alien species” (Global Invasive Species Database, 2020).

The feeding habit of the coypu is almost entirely herbivorous (LeBlanc, 1994; Carter and Leonard, 2002). Therefore, it has been a concern that indigenous plant communities including endangered and threatened species would be destroyed in regions where coypus were introduced and are now feral (Carter and Leonard, 2002; Prigioni et al., 2005). There have been many studies on aquatic vegetation as the primary food source of coypus (Guichón et al., 2003; Prigioni et al., 2005; Colares et al., 2010), and damage to crops has also been a concern (LeBlanc, 1994; Carter and Leonard, 2002; Sone et al., 2006; Panzacchi et al., 2007; Egusa and Sakata, 2009; Hong et al., 2015).

In Japan, the coypu has been strongly suspected to prey on freshwater unionid mussels, which is an endangered taxon across Japan (Negishi et al., 2008b). Until now, there have only been brief descriptions from other countries, stating that coypu predation on mussels is an occasional and opportunistic behavior (Gosling and Baker, 1991; LeBlanc, 1994). In western Japan, some snapshots of coypu predation on mussels were taken in a pond at Okayama Prefecture in 1999 (Mori, 2002) and in the Yodo River in 2014 (Ishida et al., 2015). Middens, which are a pile of dead mussel shells suspected to be caused by coypu predation, were also found there (Mori, 2002; Ishida et al., 2015) and in the Kiso River, central Japan (Kume et al., 2012). Although these papers reported the characteristics of mussel shells (e.g., shell length and species) from middens as a diet habit of coypu, it was not sufficiently confirmed whether the middens were a result of coypu predation and whether coypu predation was common.

Unionid mussels play various functional roles in stagnant and running water ecosystems (Vaughn and Hakenkamp, 2001; Vaughn, 2018). Therefore, if coypu predation is intensive on mussels, populations of aquatic organisms other than mussels can be also influenced. In addition, coypu predation can further aggravate the deterioration of mussel populations already damaged by habitat degradation. Subsequently, bitterling and Sarcocheilichthys fishes can also be threatened because they require unionid mussels for reproduction (Kitamura, 2011, Kitamura and Uchiyama, 2020). The first purpose of this study is to elucidate whether exotic coypu predation on unionid mussels occurs in the Kiso River, Japan. We will examine the possible effects on mussel population structures if coypu predation on the mussels is confirmed. The second purpose is to examine the seasonal changes in the coypu diet in the Kiso River. Terrestrial plants and unionid mussels were the focus of the dietary analysis using fecal DNA metabarcoding.

2 Materials and methods

2.1 Study site

The study was conducted in a lowland segment of the Kiso River in central Japan (drainage area of 5275 km2), located in the temperate zone (Fig. 1). Based on the data from the nearest weather station over the last decade: mean daily air temperature was 15.7 °C and summer maximum and winter minimum daily air temperatures were 38.6 °C in August 2013 and −7.4 °C in February 2012, respectively. Usually, it snows for a few days every year and seldomly accumulates. The river bed slope of the study segment is approximately 0.02%, and the calculated flow rate from the data over the last decade ranged from approximately 80 m3 s−1 of the mean base flow to approximately 4600 m3 s−1 of the mean annual maximum (annual maximum during the last decade: 11,054 m3 s−1). Relatively high flows were observed in early summer from June to July (the Baiu season), and in early autumn from September to October (the typhoon season); while from November to May, flows were relatively low and stable (Negishi et al., 2012b). Levees were constructed along both sides of the study site, but floodplains and numerous floodplain waterbodies (FWBs) are present in the inter-levee zone. All FWBs become inundated, allowing them to connect to the main channel during floods >3200 m3 s−1 (Negishi et al., 2012a).

Until the 1970s, most of the inter-levee floodplain comprised sand bars; however terrestrialization followed by tree establishment has rapidly progressed since the 1980s, resulting in most of the floodplain being covered by trees (Negishi et al., 2008a; Nagayama et al., 2015, 2017). FWBs became small but more abundant from the terrestrialization of the inter-levee floodplains, increasing the number of FWBs isolated from the main channel (Nagayama et al., 2015).

We selected four FWBs for this study, where there were frequent observations of coypus and their trails. Middens were also frequently found in these FWBs, which were suspected to be littered by coypus. Photographs of coypu preying on a mussel were first taken in one of the FWBs on April 11, 2017 (Fig. 2). The muskrat (Ondatra zibethicus), a closely related rodent species that commonly preys on mussels (Hanson et al., 1989; Neves and Odom, 1989; Owen et al., 2011), was absent in the study site. In these FWBs, three mussel taxa were present (Nodularia nipponensis, Lanceolaria oxyrhyncha, and Cristariini spp.; Lopes-Lima et al., 2020) and aquatic plants were scarce, but terrestrial plants were abundant.

thumbnail Fig. 1

Location of the study river and section. The photograph shows a part of the study section. The arrow denotes flow direction.

thumbnail Fig. 2

Photographic evidence of coypu predation on mussels (Nodularia nipponensis) in a floodplain waterbody (FWB 2) of the Kiso River, central Japan. These photographs were taken at the same position using the same camera trap. Left: coypu preying on a mussel. Right: dead mussel shell eaten by the coypu the previous night.

2.2 Collection of coypu feces

Coypu feces were collected from the four FWBs during December 2017 and February, May, and August 2018 to examine their diet using DNA metabarcoding. Two fecal samples that were recently deposited on the ground along the shore were collected at each sampling period from each FWB (32 feces in total). Collected feces were taken to a laboratory in cold storage, frozen on the same day, and stored for subsequent DNA analysis.

2.3 Fecal DNA metabarcoding

Each sample of coypu feces was freeze-dried using VD-250R Freeze Dryer (TAITEC, Saitama, Japan) and was crushed via a Shake Master Neo (BMS, Tokyo, Japan). The DNA was extracted from the crushed feces using the NucleoSpin Plant II Midi (TaKaRa Bio Inc., Shiga, Japan) according to the manufacturer's protocol. The extracted DNA samples were purified using AMPure XP (BECKMAN COULTER). The final volume of the extracted DNA samples was 100 µL, and the DNA samples were stored at −20 °C until PCR assay.

Amplicon libraries of rbcL and 18S rRNA genes were obtained via PCR amplification using universal primers for terrestrial plants (rbcL-F and −R) and bivalves (gClamF and R), respectively (Tab. 1). The first PCR was performed using each universal primer pair for terrestrial plants and bivalves. The total volume of the reaction sample was 10 µL: 1.0 µL of 10xEx buffer, 0.8 µL dNTPs, 0.5 µM of each primer, 2.0 µL DNA template, 0.1 µL Ex Taq HS (TaKaRa, Bio Inc.), and DDW. Four replicates of PCR amplification were performed for each sample. The thermal cycle profile for terrestrial plants was 94 °C for 2 min; 30 cycles of 94 °C for 30 s, 50 °C for 30 s, and 72 °C for 30 s; and 72 °C for 5 min. For bivalves, it was 95 °C for 3 min; 35 cycles of 98 °C for 20 s, 65 °C for 15 s, and 72 °C for 20 s; and 72 °C for 5 min. For each sample, the first PCR products were mixed for each sample and purified using AMPure XP to obtain a DNA template for the second PCR. The second PCR used 2nd-F and R primers (Tab. 1), including the index to identify samples and adapters for hybridizing on the surface of the Illumina flowcell for the MiSeq sequences. The total reaction volume of the second PCR was 10 µL: 1.0 µL of 10xEx buffer, 0.8 µL dNTPs, 0.5 µM of each primer, 2.0 µL DNA template, 0.1 µL ExTaq HS (TaKaRa, Bio Inc.), and DDW. The thermal cycle profile for the second PCR was 94 °C for 2 min; 10 cycles of 94 °C for 30 s, 60 °C for 30 s, and 72 °C for 30 s; and 72 °C for 5 min, for both terrestrial plants and bivalves. The second PCR products were then purified using AMPure XP. The negative control (DDW) was inserted for each step. All the indexed PCR products were pooled in equal volumes. Finally, the pooled library was purified via agarose gel electrophoresis. The purified library was then sequenced (2× 300bp paired-end) using an MiSeq v3 Reagent kit (Illumina, San Diego, CA, USA).

All data preprocessing and analysis of MiSeq raw reads were performed using the following steps. The FASTX-Toolkit (version 0.0.14; was used to filter out low-quality reads; reads corresponding perfectly to the adapter sequence of the primers were selected, and then primer sequences of the reads were removed using the fastq_barcode_spliltter command. Subsequently, reads with <20 quality values were discarded using sickle tools, and short reads with ≤40 bp were also discarded with their pairs. The reads remaining after filtering were then assembled using FLASH when read pairs overlapped by ≥10 bp.

The pipeline program USEARCH (version 11.0.667; was used to obtain operational taxonomic units (OTUs), and taxa were assigned using the basic local alignment search tool (BLAST; Edgar, 2010). For each OTU, the top-10-hit taxa with ≥97% sequence similarity were listed, and then a taxon was assigned based on vegetation data obtained from the inter-levee floodplain of the studied river section (River environmental database released by the Ministry of Land, Infrastructure, Transport and Tourism in Japan: We were able to assign the family name of the freshwater mussel in this analysis, which was Unionidae including all mussel species observed in the studied river section. DNA sequences recovered from fecal metabarcoding can provide semi-quantitative information on diet composition because of the different digestibility for each prey species (Deagle et al., 2010). Therefore, an index number was given to each taxon based on the number of reads (0; <100 reads, 1; 100 to <1000 reads, 2; 1000 to <10,000 reads, 3; ≥10,000 reads) as a criterion to estimate the abundance of prey species/taxa.

Table 1

Universal primers for terrestrial plants (rbcL) and bivalves (gClam) in the first and second PCRs.

2.4 Survey of coypu middens and live mussels

Dead mussel shells in middens were collected monthly in one of the FWBs (FWB 1 shown in Tab. 2) during 2011 (survey dates: February 11, March 27, April 18, May 21, June 11, July 29). These data were obtained from a series of surveys by Kume et al. (2012), where they only used the data from February. Because coypu predation on mussels was detected in this study (see Sect. 3), these middens were considered to be derived from coypu predation. In addition, all middens were found in shallow area or on the ground along the shoreline of the FWB, where live mussels cannot be present (Fig. 3). Almost all dead mussel shells sampled were found in pairs (Fig. 3). These strongly indicated that dead shells found in middens were attributed to coypu predation.

Two investigators waded along the shore of the waterbodies and collected dead mussel shells when they found coypu middens (consisting of more than 4 paired shells). The collected dead shells were totaled for each taxonomic identity and shell length was measured. When we found one shell (valve), we searched a partner shell to avoid double counting and measuring. Dead shells were removed from the study site during every survey date. Therefore, the collected dead shells were newly deposited shells following the previous survey date. The number of dead shells collected on February 11 was considered as a cumulative collection over a certain period before the beginning of the study.

Data of live mussels obtained in the FWB 1 on May 25, 2007 (Negishi et al., 2012a) and May 28, 2018 (this study) were used to examine the possible effects of coypu predation on the population structures of live mussels. The method for the collection in 2018 followed that of the 2007 survey by Negishi et al. (2012a). The actual time spent for searching mussels in 2007 and 2018 was equal (115 min). Mussels were searched for bare-hands by personnel with dry suits in a belt transect (2–3 m wide) laid out along the longest axis of the waterbody. The survey progressed from one end to the other of a belt transect. When mussels were found, supporting personnel recorded taxonomic identities, measured shell length, and then released the mussels at the point of collection. The Kolmogorov-Smirnov test was performed to examine the difference between shell length distributions of live mussels between the 2007 and 2018.

Table 2

Index numbers of unionid mussel reads (Unionidae spp.) for each feces in each floodplain waterbody (FWB) based on fecal DNA metabarcoding for bivalves. 0: <100 reads, 1: 100 to <1000 reads, 2: 1000 to <10,000 reads, 3: ≥10,000 reads.

thumbnail Fig. 3

Photographs of middens deposited in nearshore area. An arrow on the left photograph indicates a midden. The right photograph shows a large midden found in our midden survey.

3 Results

3.1 Diet composition

Unionid mussels (Unionidae spp.) and 17 terrestrial plant taxa were detected as primary food by PCR amplification of coypu feces (Tabs. 2 and 3). Unionid mussels were detected from 7 feces: 1 in FWB 4 in February (the coldest winter), 3 in FWB 3 and 4 in May (spring), and 3 in FWB 3 and 4 in August (the warmest summer) with relatively higher index numbers (read numbers) (Tab. 2). Although mussels were not detected from the fecal samples of FWB 2 (Tab. 2), photographs of a coypu preying on mussel were taken in April, 2017 (Fig. 2). Mussels were also not detected in FWB 1, where the midden survey (2011) was conducted (see below).

The highest taxon number of terrestrial plants detected from feces was 10 in August, followed by 7 taxa in May and 6 taxa in December and February (Tab. 3). Volubile Rosa multiflora and perennial Carex spp. were detected in all sampling periods, with relatively higher index numbers in December and February (cold seasons) (Tab. 3). Some perennial plants such as Elymus spp., Poa sp., and Potentilla indica were only detected in the cold seasons, whereas some other perennial plants such as Achyranthes sp., Andropogon virginicus, Oenanthe javanica, and Pueraria montana were only detected in the warm seasons (May and August) (Tab. 3). Most of the other plants were annual plants that were only detected during warm seasons (Tab. 3).

Table 3

Mean index numbers (N = 8) from reads of each taxon, based on the fecal DNA metabarcoding for terrestrial plants. Taxa with a total index number >0.5 from the four survey periods are shown as primary foods.

3.2 Size structures of live and dead mussels

Eight-to-fourteen middens were found in each of the midden survey dates in 2011, and a total of 918 dead mussel shells were found: 513 L. oxyrhyncha, 384 N. nipponensis, and 21 Cristariini spp. From live mussel surveys, 134 (26 L. oxyrhyncha, 96 N. nipponensis, and 12 Cristariini spp.) and 87 (54 L. oxyrhyncha, 28 N. nipponensis, and 5 Cristariini spp.) mussels were captured in 2007 and in 2018, respectively.

Size structures of the two dominant live mussel taxa were statistically different between 2007 and 2018 (L. oxyrhyncha: p < 0.001, N. nipponensis: p = 0.001) (Fig. 4). Mean shell lengths of L. oxyrhyncha and N. nipponensis decreased from 116.7 mm in 2007 to 88.8 mm in 2018, and from 61.9 mm in 2007 to 53.6 mm in 2018, respectively. The proportion of <100 mm L. oxyrhyncha increased from 0.27 in 2007 to 0.74 in 2018. Additionally, the proportion of <60 mm N. nipponensis also increased from 0.36 in 2007 to 0.71 in 2018 (Fig. 4).

Mean shell length of dead L. oxyrhyncha was 110.6–118.7 mm from February to June 2011, and finally decreased to 99.0 mm in July 2011 (Fig. 4). The proportion of ≥100 mm shells of dead L. oxyrhyncha decreased over time (0.84 in February, 0.72 in March, 0.81 in April, 0.74 in May, 0.67 in June, and 0.50 in July). The number of shells of dead L. oxyrhyncha also decreased over time during the midden survey year. Mean shell length of dead N. nipponensis was similar during the midden survey year (51.3–56.1 mm). Overall, the shell length of N. nipponensis revealed a hump-shaped distribution with a peak of medium ranges. There was no evidence for the high proportion of large shells (≥60 mm) of dead N. nipponensis (0.22–0.35), and for the decrease in the number of shells of dead mussels during the midden survey year. However, large shells of dead mussels were constantly found throughout the midden survey year.

thumbnail Fig. 4

Size structures of live mussels (open bar) in 2007 and 2018, and dead mussel shells (dark gray bar) during the midden survey of 2011. Dashed lines denote mean shell length of the species in each survey date.

4 Discussion

4.1 Coypu predation on mussels

The coypu has been well known as an herbivorous rodent in its native range (Colares et al., 2010) and some invaded regions (Abbas, 1991; Prigioni et al., 2005). However, the results of fecal DNA metabarcoding revealed that invasive coypu in the Kiso River preyed on freshwater unionid mussels from the coldest winter to the hottest summer. Until now, the predation of mussels by coypus has been considered as an occasional and opportunistic behavior (Gosling and Baker, 1991; LeBlanc, 1994), and thus, there has not been any report on this predation event with scientific verification. In Japan, there have been some previous reports of coypu preying on mussels in central and western Honshu (Mori, 2002; Kume et al., 2012; Ishida et al., 2015). However, these reports only investigated dead mussel shells in middens, presumed to be littered by coypus. This study is the first report that verifies common, non-occasional predation of native unionid mussels by coypus in an invaded site.

Diet shifting, where herbivores begin to prey on animals in invaded sites, has also been observed in other rodent species such as the black rat (Rattus rattus) (Kawakami et al., 2010; Shiels et al., 2014). This rat preyed on seabirds, insects, frogs, snakes, and plants in invaded islands (Yabe et al., 2009; Shiels et al., 2014). Such a diet shift of rodent species from herbivory to omnivory may be induced by the changes in the environment they inhabit. Coypu predation on mussels may not be that surprising when considering the common muskrat's predation on freshwater mussels and plants (Haag, 2012). In the Czech Republic, coypu predation is considered as one of the main factors of recent decline of mussel populations (Beran, 2019). We should recognize the predation risks of coypu on freshwater mussels to fully understand the damage to ecosystems in invaded regions, including not only Japan but also North America and Europe.

Coypu preying on mussels might be triggered by the lack of food items during winter and might subsequently continue through other seasons and beyond generations. The lower taxon number of terrestrial plants in winter (6 taxa) identified as primary food by fecal DNA metabarcoding, is likely attributed to the absence of annual plant species in winter. Some reports indicated that plant roots and rhizomes are the primary food diet of coypu during winter (Abbas, 1991; Gosling and Baker, 1991). Compared to the amount of effort necessary to dig into the ground, mussels are likely to be easily available prey for coypu. Moreover, it was reported that muskrat predation on mussels increased during winter when plant material was scarce (reviewed in Zahner-Meike and Hanson, 2001). These imply that coypu likely learned to prey on mussels in winter owing to necessity.

In this study, however, winter predation on mussels by coypu was detected from only one fecal sample in February, but with a high index number. In general, food items detected from feces reflect the diet from within a few days (Deagle et al., 2010). Therefore, our results of fecal DNA metabarcoding might underestimate the degree of coypu predation on mussels during winter. Whenever we visited the study FWBs, coypu trails and middens were always observed. Further studies with more sampling effort will be necessary to elucidate coypu predation on mussels throughout the year.

Short-term but intensive predation on newly established mussel populations by the muskrat was observed in a lake in western New York State, USA (Diggins and Stewart, 2000). Similarly, the intensive predation on a mussel population by invaded coypu was also reported in a pond in Okayama Prefecture, western Japan (Mori, 2002). These studies imply that intensive coypu predation on mussels began suddenly when both populations encountered one another. A large number of mussels were consumed in 2011 in the FWB 1 site (see also Kume et al., 2012). However, the DNA metabarcoding did not detect mussels from coypu feces in 2018, although coypu middens were still observed in the FWB 1. This may indicate that intensive predation on mussels by coypu discontinued in the FWB 1; likely owing to the decline in large mussel availability caused by previous intensive predation by coypu. When the availability of mussels becomes high, intensive predation may restart in the FWB 1.

4.2 Impact of coypu predation on unionid mussels

The results of midden surveys showed that large-sized L. oxyrhyncha and medium-sized N. nipponensis tended to be consumed by coypus, likely resulting in an alteration of mussel population structures. Such trends are well known as a feeding strategy of muskrats, which promises high energy benefits, and thereby heavily alters mussel population structures (Hanson et al., 1989; Neves and Odom, 1989; Diggins and Stewart, 2000; Owen et al., 2011). We also observed in our study that the size structures of both mussel species significantly changed from 2007 to 2018, with reductions in mean shell lengths and the proportions of large-sized individuals.

If size-selective predation by coypu occurs commonly, this might cause a deterioration in population size of mussels through the decrease in mature adult mussels. Mature mussel density is an important factor for fecundity rate (Downing et al., 1993; McLain and Ross, 2005). Decrease in their abundance could affect reproductive success, followed by diminishing mussel population size. Possibly, coypu predation does not necessarily cause extirpation of mussel populations because predation pressure gradually decreases when mussel density becomes low, as observed in L. oxyrhyncha in our study. However, as long as coypu predation continues, the recovery of the mussel population could be inhibited.

Coypu predation on unionid mussels could indirectly influence bitterling fish populations. In the study site (Kiso River) for example, the deepbody bitterling (also known as the Itasenpara bitterling; Acheilognathus longipinnis), which is a national monument and national endangered species of wild fauna in Japan, is present. A. longipinnis deposits eggs inside the branchial cavity of host unionid mussels in autumn, and hatched larvae spend time there until the next spring when they emerge from the host mussels (Uehara, 2007; Nishio et al., 2015; Kitamura and Uchiyama, 2020). Unionid mussels, N. nipponensis and L. oxyrhyncha, are available host mussels for A. longipinnis (Uehara, 2007; Kitamura et al., 2009). In the Kiso River, the A. longipinnis population is already threatened by the altered river environment (Sagawa et al., 2011; Nagayama et al., 2017), and coypu preying on these mussels could further aggravate the status of A. longipinnis populations, and other bitterling populations in Japanese rivers.

4.3 Seasonally flexible diet of coypu

The high seasonal variation in the coypu diet consisting of plant species has been reported in both native and invaded ranges (Abbas, 1991; Prigioni et al., 2005; Colares et al., 2010). These seasonally flexible diets were also detected from our fecal DNA metabarcoding analysis for terrestrial plants, with a greater richness in plant predation during warm periods. The coypu in our site preferred Carex spp. and Rosa multiflora, which were consumed throughout the year. Some perennial plants were only consumed in warm seasons, likely because their flowers and leaves might be preferred. Because perennial Elymus spp. and Potentilla indica were consumed only in winter, they might be involuntary food during periods when annual plants are absent. Annual plants that were consumed as primary foods might be seasonally favorable foods.

Coypu primarily consumed aquatic plants in both native and invaded sites (Guichón et al., 2003; Prigioni et al., 2005; Colares et al., 2010). In this study, aquatic plants were not targeted in the fecal DNA metabarcoding analysis. However, our results generally represent the diet of coypu in this study site. Coypus consume terrestrial plants when aquatic plants are scarce (Guichón et al., 2003; Prigioni et al., 2005). For example, terrestrial plants were sufficiently consumed by coypus inhabiting agricultural lands, where aquatic plants were scarcer (Abbas, 1991). The FWBs in our study were present in terrestrialized floodplains (see Sect. 2.1). Trees surrounded the FWBs, the aquatic-terrestrial transition zone with wet conditions was very narrow, and overall water transparency was low. Therefore, aquatic plants were scarce and could not be used as primary food by coypu in our study site.

5 Conclusion

The present study demonstrated that feral coypus prey on freshwater unionid mussels in Japan. In our study site, the exotic coypu can inhibit the increase in mussel populations. It is possible that several small mussel populations have been seriously threatened by feral coypus at other local sites in Japan. Unionid mussels play various functional roles in freshwater ecosystems (Vaughn and Hakenkamp, 2001; Vaughn, 2018). Therefore, the impacts on freshwater ecosystems by exotic coypu may be higher and broader than expected in Japan and possibly also in other countries where coypus are invasive. The common feeding on various terrestrial plants observed in this study further amplifies this concern. The black rat that invaded oceanic islands shifted their diet from herbivory to omnivory, with many animal species as a part of their common diet (Yabe et al., 2009; Shiels et al., 2014). Our results imply that the exotic coypu may also shift its diet and consume animals other than mussels, such as crustaceans and amphibians. Future studies should exhaustively examine the diet of exotic coypu to better understand its impacts on communities and ecosystems.


The authors extend special thanks to Dr. Junjiro N. Negishi at the Hokkaido University for providing previous data of live mussels in the Kiso River and to Dr. Y. Onoda and Dr. M. Sueyoshi at the Public Works Research Institute for their invaluable assistance in conducting the field survey. We would like to thank two anonymous reviewers for helpful comments on this manuscript. A part of this study was supported by the research fund for the Kiso River provided by the Ministry of Land, Infrastructure, Transport and Tourism of Japan.


  • Abbas A. 1991. Feeding strategy of coypu (Myocastor coypus) in central western France. J Zool 224: 385–401. [Google Scholar]
  • Beran L. 2019. Distribution and recent status of freshwater mussels of family Unionidae (Bivalvia) in the Czech Republic. Knowl Manag Aquat Ecosyst 420: 45. [CrossRef] [Google Scholar]
  • Carter J, Leonard BP. 2002. A review of the literature on the worldwide distribution, spread of, and efforts to eradicate the coypu (Myocastor coypus). Wildl Soc Bull 30: 162–175. [Google Scholar]
  • Colares IG, Oliveira RNV, Oliveira RM, Colares EP. 2010. Feeding habits of coypu (Myocastor coypus Molina 1978) in the wetlands of the Southern region of Brazil. Annals Braz Acad Sci 82: 671–678. [CrossRef] [Google Scholar]
  • Deagle BE, Chiaradia A, McInnes J, Jarman SN. 2010. Pyrosequencing faecal DNA to determine diet of little penguins: is what goes in what comes out? Conserv Genet 11: 2039–2048. [Google Scholar]
  • Diggins TP, Stewart KM. 2000. Evidence of large change in unionid mussel abundance from selective muskrat predation, as inferred by shell remains left on shore. Int Rev Hydrobiol 85: 505–520. [Google Scholar]
  • Downing JA, Rochon Y, Pérusse M, Harvey H. 1993. Spatial aggregation, body size, and reproductive success in the freshwater mussel Elliptio complanate . J N Am Benthol Soc 12: 148–156. [CrossRef] [Google Scholar]
  • Edgar RC. 2010. Search and clustering orders of magnitude faster than BLAST. Bioinformatics 26: 2460–2461. [CrossRef] [PubMed] [Google Scholar]
  • Egusa S, Sakata H. 2009. Status of coypu control in Hyogo Prefecture. Jpn J Limnol 70: 273–276. (In Japanese with English summary) [CrossRef] [Google Scholar]
  • Global Invasive Species Database. 2020. Accessed on 30 April 2020. [Google Scholar]
  • Gosling LM, Baker SJ. 1991. Coypu. In Corbet GB, Harris S, eds. The handbook of British mammals. Oxford: Blackwell, 267–275. [Google Scholar]
  • Guichón ML, Benítez VB, Abba A, Borgnia M, Cassini MH. 2003. Foraging behaviour of coypus Myocastor coypus: why do coypus consume aquatic plants? Acta Oecol 24: 241–246. [CrossRef] [Google Scholar]
  • Haag WR. 2012. North American Freshwater Mussels. New York: Cambridge University Press, 505 p. [Google Scholar]
  • Hanson JM, Mackay WC, Prepas EE. 1989. Effect of size-selective predation by muskrats (Ondatra zebithicus) on a population of unionid clams (Anodonta grandis simpsoniana). J Anim Ecol 58: 15–28. [Google Scholar]
  • Hong S, Do Y, Kim JY, Kim D-K, Joo G-J. 2015. Distribution, spread and habitat preferences of nutria (Myocastor coypus) invading the lower Nakdong River, South Korea. Biol Invasions 17: 1485–1496. [Google Scholar]
  • Ishida S, Kimura S, Karasawa T, Okazaki K, Hoshino T, Nagayasu N. 2015. Predation on unionid bivalves by the nutria Myocastor coypus in the Yodogawa River and its characteristics inferred from dead shell samples. Bull Osaka Mus Nat Hist 69: 29–40. (In Japanese with English summary) [Google Scholar]
  • Kawakami K, Horikoshi K, Suzuki H, Sasaki T. 2010. Impacts of predation by the invasive black rat Rattus rattus on the Bulwer's petrel Bulweria bulwerii in the Bonin Islands, Japan. In Kawakami K, Okochi I, eds. Restoring the Oceanic Island Ecosystem. Tokyo: Springer, 51–55. [CrossRef] [Google Scholar]
  • Kitamura J, Negishi JN, Nishio M, Sagawa S, Akino J, Aoki S. 2009. Host mussel utilization of the Itasenpara bitterling (Acheilognathus longipinnis) in the Moo River in Himi. Jpn Ichthyol Res 56: 296–300. [CrossRef] [Google Scholar]
  • Kitamura J. 2011. Host mussel utilization by Sarcocheilichthys variegatus variegatus (Cyprinidae, Sarcocheilichthyinae) in a drainage ditch and the Harai River of the Kushida River system, Japan. Jpn J Ichthyol 58: 195–198 (In Japanese with English summary). [Google Scholar]
  • Kitamura J, Uchiyama R. 2020. Bitterling fishes of Japan − Natural history and Culture. Tokyo: Yama-kei Publishers, 223 p. (In Japanese) [Google Scholar]
  • Kume M, Onoda Y, Negishi JN, Sagawa S, Nagayama S, Kayaba Y. 2012. Feeding damage by exotic species, nutria (Myocastor coypus), to unionid mussels in a floodplain water-body of the Kiso River, Japan. Biol Inland Waters 27: 41–47 (in Japanese with English summary). [Google Scholar]
  • Kurle CM, Croll DA, Tershy BR. 2008. Introduced rats indirectly change marine rocky intertidal communities from algae- to invertebrate-dominated. PNAS 105: 3800–3804. [CrossRef] [Google Scholar]
  • LeBlanc DJ. 1994. Nutria. In: Timm RM, ed. Prevention and control of wildlife damage, University of Nebraska − Lincoln, Nebraska, B71–B80. [Google Scholar]
  • Lopes-Lima M, Hattori A, Kondo T, et al. 2020. Freshwater mussels (Bivalvia: Unionidae) from the rising sun (Far East Asia): phylogeny, systematics, and distribution. Mol Phylogenet Evol 146: 106755. [Google Scholar]
  • McLain DC, Ross MR. 2005. Reproduction based on local patch size of Alasmidonta heterodon and dispersal by its darter host in the Mill River, Massachusetts, USA. J N Am Benthol Soc 24: 139–147. [CrossRef] [Google Scholar]
  • Mori I. 2002. A mass predation of unionid clams (Anodonta woodiana) by the feral nutria population. Bull Okayama Pref Nat Conserv Cent 10: 63–67 (in Japanese with English summary). [Google Scholar]
  • Nagayama S, Harada M, Kayaba Y. 2015. Can floodplains be recovered by flood-channel excavation? An example from Japanese lowland rivers. Ecol Civ Eng 17: 67–77 (in Japanese with English summary). [CrossRef] [Google Scholar]
  • Nagayama S, Tashiro T, Kitamura J. 2017. Inland Water Landscape: Structural and Functional Changes in the Ecosystem. In Shimizu H, Takatori C, Kawaguchi N, eds. Labor Forces and Landscape Management − Japanese Case Studies. Singapore: Springer, 107–120. [CrossRef] [Google Scholar]
  • Negishi JN, Kayaba Y, Sagawa S. 2008a. Ecological consequences of changing riverscape: terrestrialization of floodplain and freshwater mussels. Civ Eng J 50: 38–41 (in Japanese). [Google Scholar]
  • Negishi JN, Kayaba Y, Tsukahara K, Miwa Y. 2008b. Unionoid mussels as imperiled indicator organisms: habitat degradation processes and restoration approaches. Ecol Civ Eng 11: 195–211 (in Japanese with English summary). [CrossRef] [Google Scholar]
  • Negishi JN, Sagawa S, Kayaba Y, Sanada S, Kume M, Miyashita T. 2012a. Mussel responses to flood pulse frequency: the importance of local habitat. Freshw Biol 57: 1500–1511. [Google Scholar]
  • Negishi JN, Sagawa S, Sanada S, Kume M, Ohmori T, Miyashita T, Kayaba Y. 2012b. Using airborne scanning laser altimetry (LiDAR) to estimate surface connectivity of floodplain water bodies. River Res Appl 28: 258–267. [Google Scholar]
  • Neves RJ, Odom MC. 1989. Muskrat predation on endangered freshwater mussels in Virginia. J Wildl Manag 53: 934–941. [CrossRef] [Google Scholar]
  • Nishio M, Kawamoto T, Kawakami R, Edo K, Yamazaki Y. 2015. Life history and reproductive ecology of the endangered Itasenpara bitterling Acheilognathus longipinnis (Cyprinidae) in the Himi region, central Japan. J Fish Biol 87: 616–633. [CrossRef] [PubMed] [Google Scholar]
  • Owen CT, Mcgregor MA, Cobbs GA, Alexander JE Jr. 2011. Muskrat predation on a diverse unionid mussel community: impacts of prey species composition, size and shape. Freshw Biol 56: 554–564. [Google Scholar]
  • Panzacchi M, Bertolino S, Cocchi R, Genovesi P. 2007. Population control of coypu Myocastor coypus in Italy compared to eradication in UK: a cost-benefit analysis. Wildl Biol 13: 159–171. [CrossRef] [Google Scholar]
  • Pyšek P, Richardson DM. 2010. Invasive species, environmental change and management, and health. Annu Rev Environ Resour 35: 25–55. [Google Scholar]
  • Prigioni C, Balestrieri A, Remonti L. 2005. Food habits of the coypu, Myocastor coypus, and its impact on aquatic vegetation in a freshwater habitat of NW Italy. Folia Zool 54: 269–277. [Google Scholar]
  • Sagawa S, Kayaba Y, Kume M, Mori S. 2011. Comprehension of floodplain in the Kiso River and its restoration project. Civ Eng J 53: 6–9 (in Japanese). [Google Scholar]
  • Shiels AB, Pitt WC, Sugihara RT, Witmer GW. 2014. Biology and impacts of Pacific Island invasive species. 11. Rattus rattus, the black rat (Rodentia: Muridae). Pac Sci 68: 145–184. [CrossRef] [Google Scholar]
  • Sone K, Koyasu K, Kobayashi S, Tanaka S, Oda S. 2006. Agricultural damage caused by feral coypus (Myocastor coypus) in Aichi Prefecture, Japan. Mamm Sci 46: 151–159 (in Japanese with English summary). [Google Scholar]
  • Strayer DL, Eviner VT, Jeschke JM, Pace ML. 2006. Understanding the long-term effects of species invasions. Trends Ecol Evol 21: 645–651. [CrossRef] [PubMed] [Google Scholar]
  • Uehara K. 2007. Studies on methods for breeding the endangered Itasenpara bitterling, Acheilognathus longipinnis . Dissertation, Kinki University. [Google Scholar]
  • Vaughn CC. 2018. Ecosystem services provided by freshwater mussels. Hydrobiologia 810: 15–27. [Google Scholar]
  • Vaughn CC, Hakenkamp CC. 2001. The functional role of burrowing bivalves in freshwater ecosystems. Freshw Biol 46: 1431–1446. [Google Scholar]
  • Yabe T, Hashimoto T, Takiguchi M, Aoki M, Kawakami K. 2009. Seabirds in the stomach contents of black rats Rattus rattus on Higashijima, the Ogasawara (Bonin) Islands, Japan. Mar Ornithol 37: 293–295. [Google Scholar]
  • Zahner-Meike E, Hanson JM. 2001. Effect of muskrat predation on naiads. In Bauer G, Wächtler K, eds. Ecology and evolution of the freshwater mussels Unionoida, Ecological Studies 145, Springer, Berlin, pp 163–184. [CrossRef] [Google Scholar]

Cite this article as: Nagayama S, Kume M, Oota M, Mizushima K, Mori S. 2020. Common coypu predation on unionid mussels and terrestrial plants in an invaded Japanese river. Knowl. Manag. Aquat. Ecosyst., 421, 37.

All Tables

Table 1

Universal primers for terrestrial plants (rbcL) and bivalves (gClam) in the first and second PCRs.

Table 2

Index numbers of unionid mussel reads (Unionidae spp.) for each feces in each floodplain waterbody (FWB) based on fecal DNA metabarcoding for bivalves. 0: <100 reads, 1: 100 to <1000 reads, 2: 1000 to <10,000 reads, 3: ≥10,000 reads.

Table 3

Mean index numbers (N = 8) from reads of each taxon, based on the fecal DNA metabarcoding for terrestrial plants. Taxa with a total index number >0.5 from the four survey periods are shown as primary foods.

All Figures

thumbnail Fig. 1

Location of the study river and section. The photograph shows a part of the study section. The arrow denotes flow direction.

In the text
thumbnail Fig. 2

Photographic evidence of coypu predation on mussels (Nodularia nipponensis) in a floodplain waterbody (FWB 2) of the Kiso River, central Japan. These photographs were taken at the same position using the same camera trap. Left: coypu preying on a mussel. Right: dead mussel shell eaten by the coypu the previous night.

In the text
thumbnail Fig. 3

Photographs of middens deposited in nearshore area. An arrow on the left photograph indicates a midden. The right photograph shows a large midden found in our midden survey.

In the text
thumbnail Fig. 4

Size structures of live mussels (open bar) in 2007 and 2018, and dead mussel shells (dark gray bar) during the midden survey of 2011. Dashed lines denote mean shell length of the species in each survey date.

In the text

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.