Impact of Perkinsus olseni infection on a wild population of Manila clam Ruditapes philippinarum in Ariake Bay, Japan

Journal of Invertebrate Pathology 153 (2018) 134–144

Contents lists available at ScienceDirect

Journal of Invertebrate Pathology journal homepage: www.elsevier.com/locate/jip

Impact of Perkinsus olseni infection on a wild population of Manila clam Ruditapes philippinarum in Ariake Bay, Japan

T

Tsukasa Waki, Miki Takahashi, Tatsuya Eki, Masato Hiasa, Kousuke Umeda, Nanae Karakawa, ⁎ Tomoyoshi Yoshinaga Department of Aquatic Bioscience, Graduate School of Agricultural and Life Sciences, The University of Tokyo, Yahoi 1-1-1, Bunkyo-ku, Tokyo 113-8657, Japan

A R T I C LE I N FO

A B S T R A C T

Keywords: Perkinsosis Shellfishery Epidemiology Disease dynamics Seasonality Production decline

Many studies have addressed the production decline of Manila clam, Ruditapes philippinarum, in Japan, but infection of clams with Perkinsus olseni has received scarce attention. To evaluate the impact of P. olseni, infection levels and host density of a wild, unexploited clam population were monitored monthly or bimonthly on a tidal flat from June 2009 to January 2013. Real-time PCR analysis discriminating P. olseni and Perkinsus honshuensis detected only P. olseni in tested clams. The prevalence of P. olseni was 100% or nearly 100% in 7 cohorts throughout the study period, except in newly recruited clams. Infection intensity remained low and seldom reached 106 cells/g wet tissue in newly recruited clams until the month of September. Infection intensity reached 106 cells/g in autumn and remained high at 104–106 cells/g until each cohort of clams disappeared. Clam density began to decrease in the autumn when the infection intensities reached ca. 106 cells/g. Density was relatively stable in winter, increased in spring and decreased again in clams aged 1 year or older during summer and autumn in the following years. Survival of clams experimentally infected with P. olseni at ca. 106 cells/g and placed in a cage in the tidal flat for 1 or 2 months was significantly lower than survival of uninfected control clams. Our results suggested that heavy infection with P. olseni was a major cause of the clam density decrease, although other environmental and biological factors also may have contributed to the decline in density. In addition, uninfected clams were deployed in cages for 1–2 months from June 2010 to January 2013 and prevalence and infection intensity were monitored. Parasite transmission and infection progression increased in summer and autumn.

1. Introduction The Manila clam Ruditapes philippinarum is a common bivalve in tidal flats in Japan, and is important for recreational and licensed fisheries (Ito, 2002). In the early 1980s, Ariake Bay was a major site for Manila clam catch in Japan. According to the Food and Agriculture Organization of the United Nations (FAO) statistics, Manila clam production from the fishery and aquaculture in Japan was more than 100,000 tons from 1950 to 1986, the highest production worldwide. However, the annual catch dramatically declined to a minimum of 14,000 ton in 2015, probably because of the decline of the population in Ariake Bay since the late 1970s, and subsequent nationwide declines since the mid-1980s (Matsukawa et al., 2008). Environmental changes, overfishing, and predators, among other factors, are suspected causes of the decline in the Manila clam catch; however, the major causes of the decline are still unclear.

⁎

Protozoans in the genus Perkinsus, superphylum Alveolata (Siddall et al., 1997), infect mollusks (Villalba et al., 2004), and Perkinsus olseni is included in the list of notifiable mollusk diseases of the World Organization for Animal Health (OIE). In Japan, Perkinsus spp., primarily P. olseni, have been detected in most Manila clam populations since the mid-1990s (Choi et al., 2002; Hamaguchi et al., 1998, 2002; Momoyama and Taga, 2005; Sakai and Onodera, 2006; Takahashi et al., 2009; Umeda and Yoshinaga, 2012; Yoshinaga et al., 2010), except those on the north and east coasts of Hokkaido. Perkinsus spp. typically have two phases in their life cycle: the propagation phase in their hosts and the zoosporulation phase in seawater (Bordenave et al., 1995). Trophozoites propagate by repeated cell divisions in the host tissues. When trophozoites are exposed to anaerobic conditions generated by the death of the host, they transform into prezoosporangia. In seawater, the prezoosporangia transform into zoosporangia, which release infective zoospores (Bordenave et al., 1995).

Corresponding author. E-mail addresses: [email protected] (T. Waki), [email protected] (M. Takahashi), [email protected] (T. Eki), [email protected] (M. Hiasa), [email protected] (K. Umeda), [email protected]ff.go.jp (N. Karakawa), [email protected] (T. Yoshinaga). https://doi.org/10.1016/j.jip.2018.03.001 Received 24 February 2017; Received in revised form 1 March 2018; Accepted 2 March 2018 Available online 05 March 2018 0022-2011/ © 2018 Elsevier Inc. All rights reserved.

Journal of Invertebrate Pathology 153 (2018) 134–144

T. Waki et al.

Perkinsus honshuensis and P. olseni have been reported from exploited and natural populations of Manila clam in Japan and have occasionally co-infected wild Manila clams (Dungan and Reece, 2006; Umeda and Yoshinaga, 2012). In studies conducted before the first detection of P. honshuensis (Dungan and Reece, 2006), Perkinsus spp. in Manila clams in Japan were reported as unidentified Perkinsus sp. or P. olseni because neither the Ray’s fluid thioglycollate medium (RFTM) assay nor the PCR assay, both of which are genus specific, could distinguish the two species. Umeda and Yoshinaga (2012) developed a real-time PCR assay to distinguish P. olseni and P. honshuensis. They reported that P. olseni was predominant in all the tested Manila clam populations in Japan, and that the lower infection intensities of P. honshuensis were, therefore, negligible. Recently, infection with P. olseni has been suspected to be a major cause of the Manila clam decline in Japan. High infection intensity and prevalence were reported in many areas, including Ariake Bay where clam catch has declined (Choi et al., 2002; Park et al., 2008; Umeda and Yoshinaga, 2012). Previous challenge experiments demonstrated that infection with P. olseni had serious negative effects on the survival and physiology of the clams (Shimokawa et al., 2010; Waki et al., 2012; Waki and Yoshinaga, 2013, 2015), however, the negative impacts of P. olseni infection on the survival of Manila clams in wild unexploited or exploited populations remained unclear. The aim of this study was to evaluate the impact of P. olseni infection on wild Manila clam populations, and to quantify changes in population density of the wild clams after infection. We chose a survey site in a tidal flat of Ariake Bay where the annual Manila clam catch has dramatically decreased from 60,000 tons in the late 1970s to less than 10,000 tons in 2015. P. olseni infection, Manila clam population density and environmental factors were surveyed for 4.5 years in an area where clamming has been prohibited. We also examined the virulence of P. olseni under natural environmental conditions by placing experimentally challenged Manila clams, and unchallenged Perkinsus-free clams together in cages on the tidal flat. Additionally, seasonal fluctuations in parasite transmission were examined at the study site by placing uninfected clams at the site monthly or bimonthly and monitoring infection after 1–2 months of exposure.

Fig. 1. Study stations at Taimei, Ariake Bay in Japan. A–F represent Stations. Gray lines represent depth contours.

2.1. Field survey

Cooperation) set 5 cm above the sediment surface. The salinity data logger was replaced with a new logger monthly or bi-monthly. To avoid attachment of sessile organisms on the sensor, the salinity logger was covered with a 300 µm-mesh nylon bag from July 2012 to November 1, 2012 and the sensor was coated with a chemical barnacle repellent (Annex, Ecowel, Tokyo, Japan) mixed with water-based paint from November 1, 2012 to January 30, 2013. We assumed that all clams determined to be positive in the RFTM assay were infected with P. olseni only (Umeda and Yoshinaga, 2012). Although some clams might have also been infected with P. honshuensis, the infection was considered to be negligible.

2.1.1. Study area Field sampling was conducted monthly or bimonthly from June 24, 2009 to January 30, 2013 at Station (St.) A (N32°52′46″, E130°30′13″) in the intertidal estuary zone of the tidal flat at Taimei, Kumamoto Prefecture, near the mouth of Kikuchi River, on the eastern side of Ariake Bay, Japan (Fig. 1). Aquaculture activities were absent in the area. Near St. A, clamming has been prohibited since 1989 to conserve Manila clams. At the station, we covered sediment with three polyethylene nets (37.5 mm mesh opening, 1 m × 2 m) from January 24, 2011 to January 30, 2013 to deter unexpected clamming. Sampling was conducted at an additional five stations: St. B (N32°52′59″, E130°30′24″), St. C (N32°52′34″, E130°30′7″), St. D (N32°53′2″, E130°30′13.31″), St. E (N32°52′54″, E130°30′6″), and St. F (N32°52′41″, E130°30′0″). These stations were set on a grid with spacing of ca. 500 m (Fig. 1). Sampling was conducted on June 5, 2012 (at St. B–F) and August 3, 2012 (at St. B and C) to determine the infection status of Manila clams in areas surrounding St. A. Sediment temperature was measured at St. A at 10- or 15-min intervals from June 27 to December 22, 2010, January 23 to December 16, 2011, and February 2012 to January 30, 2013 with a temperature data logger (Bobo water temp pro V2, Onset Computer Cooperation; Bourne, MA, USA) buried to a depth of 5 cm in sediment. Water salinity was measured at 15-min intervals from January 2012 to January 30, 2013 with a salinity data logger (H24-002, Onset Computer

2.1.2. Sampling procedure and analysis Sediment samples were collected in triplicate within a quadrat (20 cm × 20 cm × 10 cm, L × W × D) at each station during the ebb tide. Each sediment sample was separately sieved through a 2-mm mesh, and trapped sediment was collected. Manila clams with two shells (apparently alive) were collected from the sediment and kept on ice or at 4 °C for transportation and storage for 1–2 days. The shell length of the clams was measured and their shells subsequently opened. Clams with soft tissue were classified as live clams and those without were classified as newly dead clams. To determine infection levels, 30 live clams were selected so that they would roughly represent the sizes of clams from the sediment sample containing the largest number of live clams. Each live clam was individually subjected to the whole-body RFTM assay (Ray, 1952), according to Waki and Yoshinaga (2013), Choi et al. (1989), and Almeida et al. (1999). When fewer than 30 clams were collected in a sediment sample, all of the clams were assayed. To collect small Manila clams that passed through the 2-mm mesh (smaller than approximately 4 mm in shell length), sediment was sampled at depths of 0–1.5 cm and 1.5–5 cm in a small quadrat (10 cm × 10 cm × 5 cm, L × W × D) placed within the large quadrat (20 cm × 20 cm × 10 cm, L × W × D) at St. A from December 16, 2011 until the end of the survey. Sediment obtained from 0 to 1.5 cm depth with the 2-mm mesh was sieved and the samples were removed, while

2. Material and methods

135

Journal of Invertebrate Pathology 153 (2018) 134–144

T. Waki et al.

Table 1 Comparisons of survival and infection intensity between experimentally challenged Manila clams and unchallenged control clams placed in sediments in the survey area. Experimental period

Aug. 3 – Sep. 1

Sep. 1-Oct. 3

Oct. 3-Dec. 1

a b

Group

Manila clams before planting

Manila clams collected from the sediment

Number

Infection intensitya

Recoveryb (%)

Survival (%) in collected clams

Shell length (mm) of collected live clams upper: mean ± SD lower: range

Infection intensitya

Challenged

316

5.9–7.0

87.7

2.5

6.6–7.0

Control

316

0

83.2

85.6

13.1 ± 0.5 12.6–13.8 17.2 ± 1.3 12.8–20.3

Challenged

620

5.6–6.1

82.0

18.3

5.6–7.3

Control

620

0

79.5

75.2

12.4 ± 0.8 11.0–14.9 13.7 ± 1.1 11.1–17.2

Challenged

400

5.9–6.4

53.5

6.5

6.4–7.0

Control

400

2.1–2.2

62.5

87.2

12.4 ± 0.7 11.1–13.13 13.4 ± 1.0 11.1–15.8

1.3–4.9

3.0–5.6

2.3–5.5

log (N + 1) (cells/g wet tissue). (number of collected clams/number of clams placed in the field) × 100.

Prefecture, and used to make Perkinsus-free residual tissue to produce uninfected control clams in the experimental challenges. The Manila clams in the Bay were known to be free from Perkinsus infection. Perkinsus infection had not been detected in these clams in previous studies, or in our analyses that were conducted until 2017 (Hamaguchi et al., 2002; Nishihara, 2010; Shimokawa et al., 2010; Waki and Yoshinaga, 2013, 2015). Manila clams (11–14 mm in shell length) produced in a hatchery of Oita Prefectural Agriculture, Forestry and Fisheries Research Center (Bungotakada City, Oita Prefecture, Japan) using wild clams captured from a nearby tidal flat as broodstock, were obtained for these challenge experiments. The hatchery-raised clams were examined for Perkinsus infection by the whole-body RFTM assay before each challenge. No infection was detected in 20 clams before the first and the second experiments. Although a low-intensity infection (less than 102 cells/g wet tissue) was detected in two of 20 clams in the third challenge experiment, we proceeded with their use because the infection intensities were very low. Prezoosporangia (1.6 × 105–4.1 × 106 cells) produced from wild Manila clams obtained from Ariake Bay in the RFTM were incubated in seawater to produce a zoospore suspension, as described by Waki and Yoshinaga (2013). Hatchery-raised clams (260–620 individuals) were exposed to the zoospore suspension generated from the prezoosporangia (1.7 × 105–1.0 × 106 cells) for 24 or 48 h, as described by Waki and Yoshinaga (2013). The numbers of clams, number of prezoosporangia used to generate zoospores, incubation periods for zoosporulation, and period of exposure to zoospore suspensions varied among the three challenge experiments because of the varying numbers of clams and P. olseni prezoosporangia available for each challenge. Manila clams used as controls were immersed for 24 or 48 h in a suspension of tissue residue from Perkinsus-free clams purchased from Akkeshi Bay, as described by Waki and Yoshinaga (2013). After exposure, the challenged and control clams were labelled with different colored pens and kept in floating baskets in a seawater aquarium (22–25 °C) for 1–5 days until being transported to St. A. Equal numbers of challenged and control clams (200–400 individuals for each group) were placed in a covered plastic cage (60 cm × 25 cm × 25 cm, L × W × D, 3-mm mesh), which was partially buried to 15-cm depth in the sediment at St. A, and left for 1–2 months: August 3 to September 1, 2012, September 1 to October 3, 2012, and October 3 to December 1, 2012 (Table 1). The Manila clams were removed from the cage and kept on ice or at 4 °C for transportation and storage. They were examined within 2 days of collection. The numbers of the recovered, surviving, and dead Manila clams were

clams passing through the mesh were collected separately. Sediment obtained from 1.5 to 5 cm depth was filtered through 2-mm mesh and then through 1-mm mesh, and the samples trapped on each mesh were collected separately. Sediments passing through the 1-mm mesh were not sampled, because a preliminary investigation found no small clams passing through this mesh size (ca. 1–2 mm in shell length) in sediment from 1.5 to 5 cm depth. The samples were kept on ice or at 4 °C for transportation and were stored for 1–4 days. Small clams of identifiable size (> 500 µm in shell length) were sampled from sediments under a stereomicroscope. Live clams and newly dead clams with both shells were collected, and infection intensities in the live clams were determined by the RFTM assay after measuring shell length. The remaining sediment in the large quadrat was sampled using the 2-mm mesh, and clam samples were processed in the same way as those collected before November 2011. Manila clams collected in each month were separated into several cohorts for data analyses of population dynamics, as described by Aizawa and Takiguchi (1999). The frequency distributions of shell length were determined from triplicate samples at each sampling time. Each frequency distribution was divided into plural distributions of cohorts, assuming that each cohort had a normal frequency distribution in shell length. We calculated the mean shell length and standard deviation, and the density of each obtained cohort. When examining the infection status, each clam was allocated to a cohort for analysis. When the distributions of two cohorts overlapped, clams larger than and smaller than the cross point were allocated to the older cohort and younger cohort, respectively. Clams smaller than 2 mm in shell length were regarded as one cohort because such small clams did not exhibit a normal distribution in their frequency distribution. To survey the infection status surrounding St. A, we sampled Manila clams from Stations B–F using the larger quadrat and the 2-mm mesh. The samples trapped on the 2-mm mesh were collected and processed. 2.2. Placement of experimentally challenged Manila clams in field sediments The virulence of P. olseni infection in the natural environment was evaluated three times by placing experimentally challenged Manila clams and naive unchallenged clams in field cages at St. A for 1–2 months. Naturally infected wild clams (21–37 mm in shell length) from Ariake Bay were used as the infection source in experimental challenges. Perkinsus-free Manila clams of commercial size (36–45 mm in shell length) were purchased from Akkeshi Bay, Hokkaido 136

Journal of Invertebrate Pathology 153 (2018) 134–144

T. Waki et al.

counted to calculate recovery percentage (no. of dead and live clams collected at sampling time/no. of clams placed × 100) and survival percentage (no. of live clams collected at sampling date/no. of clams placed × 100), ignoring the clams lost in the field. The shell lengths of the surviving clams were measured to calculate the growth rates for each group. The infection levels of P. olseni and P. honshuensis in four clams that were experimentally challenged and collected from the field in the first experiment and two clams in the second experiment were evaluated by real-time PCR according to Umeda and Yoshinaga (2012). The remaining challenged clams and 30 of the control Manila clams were analyzed using the RFTM assay to determine the infection intensity in each experiment. In parallel with these field experiments, equal numbers of challenged and control clams (84–400 individuals) were maintained in floating baskets (25 cm × 15 cm × 15 cm, L × W × D) in a 1-ton recirculating artificial seawater aquarium equipped with a biological filter and a temperature controller (salinity: 30‰, temperature: 28 °C, 25 °C, and 18–22 °C for the first, second, and third experiments, respectively). These temperature conditions were similar to those at the field site (Fig. 2). Ten milliliters of commercially supplied diatoms (Chaetoceros calcitrans, approximately 6 × 105 cells/mL; Sunculture, Nissin Marintech, Aichi Prefecture, Japan) was added to the aquarium each day as food. Clams were sampled at intervals. At the end of the first and second experiments, four of five and three of 12 challenged clams, respectively, were analyzed by real-time PCR (Umeda and Yoshinaga, 2012) and the remaining challenged and control clams were analyzed by the RFTM assay. Kaplan–Meier survival curves were constructed to analyze survival.

Table 2 Manila clams placed and then recovered in natural transmission experiment. Date of placement

Date of recovery

Experiment periods (days)

No. of planted clamsa

2010 Jun 27 Jul 26 Aug 25 Sep 28 Oct 29 Nov 25 Dec 22

2010 Jul 26 Aug 25 Sep 28 Oct 29 Nov 25 Dec 22 2011 Jan 24

29 30 34 31 27 27 33

20–30 20–30 20–30 20–30 20–30 20–30 20–30

4 17 0 21 20 6 16

26

20–30

27

31 33 31 25 32 29 31 31 32

20–30 30 30 30 30 30 30 30 30

30

21 14 20 3 2 19 27 21 30

58

30

30

30

24 32 27 30 28 31 29 32 29 30

30 30 30 30 30 30 30 30 30 30

30 30 29 30 30 24 30 30 30 30

28 27 28 30 30 5 26 13 29 30

60

30

27

27

2011 Jan 24 Feb 19 Apr 16 May 19 Jun 19 Jul 14 Aug 15 Sep 13 Oct 14 Nov 14 Dec 16 2012 Feb 12 Mar 8 Apr 9 May 6 Jun 5 Jul 3 Aug 3 Sep 1 Oct. 3 Nov. 1

2.3. Seasonal fluctuations in natural transmission of Perkinsus olseni This experiment was designed to determine the seasonal fluctuations in the transmission of P. olseni. A covered plastic cage (30 cm × 30 cm × 40 cm, L × W × D, 5-mm mesh) was partially buried to approximately 35-cm depth in the sand at St. A. Approximately 30 uninfected Manila clams of commercial size (31–54 mm in shell length) obtained from Akkeshi Bay were placed in the cage monthly or bimonthly during sampling events from June 2010 to December 2012 (Table 2). The clams were recovered at the next regular sampling time and kept on ice or at 4 °C for transportation to the laboratory. The gills were analyzed using the RFTM assay to determine the prevalence and intensity of infection. Data were not obtained in September 2009 because all clams in the cage had died. In addition, Manila clams could not be obtained from

Dec. 1

Feb 19 Mar 22 May 19 Jun 19 Jul 14 Aug 15 Sep 13 Oct 14 Nov 14 Dec 16 2012 Feb 12 Mar 8 Apr 9 May 6 Jun 5 Jul 3 Aug 3 Sep 1 Oct 3 Nov 1 Dec 1 2013 Jan 30

No. of collected clamsb

No. of surviving clams

a Number of Manila clams in cage was not counted precisely from June 2010 to February 2011, and ranged between 20 and 30. b Number of Manila clams collected from cage was not counted from Aug. 2010 to Oct 2011; only surviving clams were counted.

Akkeshi Bay in March 2011 because of The Great East Japan Earthquake. The data obtained in February 2012 represented the infection status of clams left in the cage for 2 months (from December 2011 to February 2012) because regular sampling was not conducted in January 2012. 3. Results 3.1. Field survey The sediment temperature ranged from ca. 2 °C to 15 °C in winter and from ca. 25 °C to 37 °C in summer (Fig. 2). Unfortunately, sediment temperature was not recorded from June to December 2011 because we failed to configure the logger. Water salinity varied within the range of 9.4–37.6‰ at St. A (Fig. 2). The salinity level was relatively low in July, possibly because of heavy rainfall during the rainy season in Japan. We failed to measure water salinity during May, June, September, and October because the sensor was covered with barnacles. Manila clams were collected until July 2012 (Fig. 3). After August 2012, the density of Manila clams dramatically declined in the area, probably because of flooding of the Kikuchi River after heavy rainfall in mid-July (Ministry of Land, Infrastructure, Transport and Tourism Japan, Fig. 4). This decrease in density prevented the cohort analysis. The shell length frequency distributions of live Manila clams are shown in Fig. 3. The cohort analysis revealed that seven cohorts appeared at St. A during the survey. The three cohorts that appeared in

Fig. 2. Sediment temperature and water salinity at Station A.

137

Journal of Invertebrate Pathology 153 (2018) 134–144

T. Waki et al.

Fig. 3. Frequency distributions and deduced normal distributions for shell length of live Manila clams sampled at St. A between June 2009 and January 2013. Dates are expressed in yyyy.mm.dd format. Vertical bars represent SD. N represents number of live Manila clams sampled.

cohorts, respectively (Fig. 5). The clams in these three cohorts appeared from June to August. Their densities peaked several months after sampling began, and peak density ranged from approximately 250 to 1000 individuals/m2 (Fig. 5B). The density of each cohort subsequently declined during late summer and autumn, and increased in the following winter and spring. The increase during winter and spring resulted from the appearance of young clams that were recruited in winter and spring, indicating that we could not completely distinguish between younger clams and the older three cohorts during this period. Subsequently, the density of the three cohorts began to decline in early summer (from June or July). Although young Manila clams that were recruited in winter and spring could not be distinguished from the three older cohorts, there was a clear pattern in the dynamics of the three cohorts; in each cohort, the density of clams decreased from late summer to autumn in the first year and from early summer to autumn in the second year. In July 2012, the density of Manila clams rapidly declined, probably because of the freshet that affected the sampling

Fig. 4. Seasonal fluctuations in flow of Kikuchi River. Asterisk represents flooding event.

the summer of 2009, 2010, and 2011 and were monitored for more than 12 months were designated as the 2009, 2010, and 2011 summer

138

Journal of Invertebrate Pathology 153 (2018) 134–144

T. Waki et al.

Fig. 5. Seasonal variations in shell length (A), density (B), and infection intensity (C–F) of Manila clam cohorts at St. A. Vertical bars in A represent SD. C, D, and E represent infection intensities of 2009 summer cohort, 2010 summer cohort, and 2011 summer cohort, respectively. F represents infection intensity of 2008 cohort, 2009 autumn cohort, 2012 summer cohorts 1 and 2, and clams of unknown cohort.

sampled only in July 2012 (Fig. 5A and B). These minor cohorts showed similar trends in density as those of the major cohorts. That is, the density declined from late summer to autumn in the first year and from early summer to autumn in the second year. Most of the data obtained for the minor cohorts were fragmented and insufficient for full interpretation. The infection intensity remained low and seldom reached 106 cells/

area. Other minor cohorts were also found during the survey (Fig. 5A and F). Clams in the 2008 cohort (probably the summer or autumn cohort) were present at the start of the survey in June 2009 (Fig. 5A). The 2009 autumn cohort was collected from November 2009 to February 2010 (Fig. 5A and B). Clams in the 2012 summer cohort-1 were collected from April to June 2012, and those in the 2012 summer cohort-2 were 139

Journal of Invertebrate Pathology 153 (2018) 134–144

T. Waki et al.

g wet tissue in newly recruited clams until September of the first year in the 2009, 2010, and 2011 cohorts, but increased to ca. 104–106 cells/g wet tissue in the following autumn (Fig. 5C–E). Subsequently, the infection intensities remained within that range, except in uninfected or lightly infected clams. Low infection intensities (< 103 cells/g wet tissue) were detected again between January and June in the second year. This was probably caused by the appearance of young clams, because the periods coincided with the increase in the density of the three cohorts in the second year, and cure or recovery from P. olseni infection (i.e., a decrease in the infection intensity of P. olseni in Manila clams) has never been detected in any challenge experiments previously reported (Shimokawa et al., 2010; Waki et al., 2012; Waki and Yoshinaga, 2013, 2015) studies. The disappearance of heavily infected Manila clams with infection intensity > 106 cells/g wet tissue was roughly simultaneous with the decrease in clam density. Heavily infected clams disappeared after late autumn in the first year and between June and August in the second year in the 2009 summer cohort, after late autumn in the first year and between July and October in the second year in the 2010 summer cohort, and between April and June in the second year in the 2011 summer cohort. The shell length distributions of newly dead clams did not fit a normal distribution (Fig. 6). However, it was possible to identify the cohort of dead clams by comparing the shell length frequency distribution of newly dead clams with that of live clams collected at the same time. Newly dead clams belonging to the 2010 summer cohort (< 10 mm shell length in July 2010) and the 2011 summer cohort (ca. 10 mm shell length) appeared from July 2010 to December 2010 and from May 2011 to October 2011, respectively. Newly dead small juveniles with < 1 mm shell length were found in November 2012, but live juveniles of the same size were not detected. All Manila clams sampled at St. B–F were infected, and showed infection intensities ranging from 104 to 106 cells/g wet tissue (Fig. 7). There was no clear difference in infection intensities among the stations in June 5, 2012 (ANCOVA, p > 0.05). Infection intensities in August 3, 2012 were not statistically analyzed due to the insufficient number of clams, but did not show much difference from those in August. Thus, the infection intensities at St. A were considered to be representative of those in the surrounding area, although the clam densities differed substantially among the stations (Fig. 8).

g wet tissue) and the prevalence was < 60% (Fig. 9). There were no significant differences in shell length between the challenged and control group at the end of the trial (t-test, p > 0.05). P. olseni was the only Perkinsus sp. detected by real-time PCR in the clams that were experimentally challenged and collected from the field, at infection intensities of 104.9–108.6 cells/g wet tissue. P. honshuensis was not detected in any clams. 3.3. Seasonal fluctuations in transmission of P. olseni Placing uninfected clams in cages in the field and collecting them the following month showed that prevalence typically reached 100% in summer and autumn while it was much lower (0–60%) in winter and spring (Fig. 10). The geometric mean infection intensity was higher in summer and autumn and lower in winter and spring (Fig. 10). 4. Discussion Previous field studies have focused on negative effects of P. olseni infection on Manila clam populations (Choi et al., 2002; Park et al., 2008; Yoshinaga et al., 2010), however, the nature of the effects was unclear because of the short survey periods and because the clam populations were not divided into cohorts. Our previous challenge experiments showed that mortality began to occur when mean infection intensities reached ca. 106 cells/g wet tissue in juvenile clams (approximately 10 mm shell length) (Shimokawa et al., 2010; Waki and Yoshinaga, 2013; Waki et al., 2012). These findings indicated that P. olseni infection intensity at 106 cells/g wet tissue or higher is lethal for juvenile Manila clams. Several lines of evidence from our current field survey also suggested that an infection intensity of ca. 106 cells/g wet tissue or higher is lethal to Manila clams: (1) Clams with infection intensities of more than 106 cells/g wet tissue were seldom found in samples, (2) Clam densities decreased when the highest infection intensities reached ca. 106 cells/g wet tissue in their first autumn in the three major cohorts and, (3) Clams with infection intensity close to 106 cell/g wet tissue and higher disappeared when the clam density declined from early summer to autumn in following years in the three cohorts. In addition, many newly dead Manila clams were found during the period of declining clam densities, indicating that the decrease in density was not caused primarily by predation or by removal due to water movements. If predation or removal had been the major causes, newly dead clams would have disappeared from the area. Instead, our findings suggested that mortality was associated with infection to some extent, although other factors (e.g., high water temperatures in summer and autumn, and exhaustion after spawning) may also have contributed. The significantly higher mortality of experimentally challenged clams (at ca. 106 cells/g wet tissue or higher) than unchallenged clams when placed in the sediment supports the lethal effect of infection at this intensity. Interestingly, although Manila clams with much higher infection intensities (107–108 cells/g wet tissue) were produced in aquaria in previous studies (Shimokawa et al., 2010; Waki and Yoshinaga, 2013; Waki et al., 2012), we found few with such high infection intensities in the present field survey. It is likely that clams with very high infection intensities do not survive under natural conditions, which are harsher than aquaria conditions. In addition, we did not detect other Perkinsus species, including P. honshuensis, in this study. This result suggested that the majority of Perkinsus spp. in Ariake Bay was P. olseni, as reported by Umeda et al. (2013). Our results suggested that heavy infection of P. olseni caused the death of host Manila clams in Ariake Bay. It is difficult to estimate to what extent the death of very heavily infected clams corresponded with the decrease in live clam density and clam resources, although our challenge experiments revealed high mortality rates of heavily infected clams. There are many factors that could affect the clam death in natural conditions. In addition, sensitivity differs between hatchery-raised

3.2. Placing of experimentally challenged Manila clams in sediments Manila clams experimentally challenged and prepared for placement in sediments were infected with P. olseni within the range of 105.6–107.0 cells/g wet tissue (Table 1), although the challenge doses and periods varied among challenges. None of the unchallenged clams was infected, except for two of 20 clams in the third trial. After 1–2 months in the field cage, all clams in challenged and control groups were infected at the ranges of 105.6–107.3 cells/g wet tissue and 101.3–105.6 cells/g wet tissue, respectively (Table 1). The challenged clams showed significantly lower survival and smaller shell length than the control clams in all trials (χ2-test and t-test, respectively, p < 0.05). Approximately 50–80% of the clams were recovered in both challenged and control groups in the first, second, and third trial (Table 1). The survival percentages at the end experiments were 2.5%–18.3% and 75.2%–87.2% in challenged and unchallenged groups, respectively (Table 1). In the aquaria experiments, the challenged clams exhibited significantly lower survival than the control group in all trials (log-rank test, p < 0.05, Fig. 9). The survival percentages of the challenged and control groups were 7.0% and 53.2%, 2.4% and 58.8%, and 4.5% and 29.3% in the first, second, and third trials, respectively (Fig. 9). In the challenged groups, infection intensities were ca. 106 cells/g wet tissue at 0 d (Fig. 9). Survival in the challenged groups began to decrease when the infection intensities exceeded 106 cells/g wet tissue. The controls in the aquaria were lightly infected with P. olseni (< 103 cells/ 140

Journal of Invertebrate Pathology 153 (2018) 134–144

T. Waki et al.

Fig. 6. Frequency distributions in shell length of newly dead Manila clams sampled at St. A between April 2010 and January 2013. Dates are expressed in yyyy.mm.dd format. Vertical bars represent SD. Deduced normal distributions of live clams at the same sampling times are represented by solid and broken lines. N represents number of newly dead Manila clams sampled.

Yoshinaga et al. (2010) reported that P. olseni infection had little effect on the filtration activity and high temperature tolerance of Manila clams, and found no clear relationships between the condition index and infection intensities of P. olseni in clams of commercial size

clams and wild clams. To understand the effect of P. olseni on clam density and resources, further surveys of Manila clam population dynamics should be conducted in areas where the infection level of P. olseni is much lower than that in Ariake Bay.

141

Journal of Invertebrate Pathology 153 (2018) 134–144

T. Waki et al.

Fig. 9. Kaplan–Meier survival and infection intensities of Manila clams maintained in floating baskets in laboratory aquarium after experimental challenges. A-1, B-1, and C-1 show survival curves in the first, second, and third experiments, respectively. Black lines and gray lines represent challenged and control groups, respectively. A-2, B-2, and C-2 show infection intensities in first, second, and third experiments respectively. Closed squares and open circles represent infection intensities in challenged and control groups, respectively. Fig. 7. Infection intensities and shell length of Manila clams. A–F represent clams collected from St. A–F on June 5, 2012; G, H, and I represent clams collected from St. A, B, and C on August 3, 2012.

Fig. 10. Variation in the values of prevalence (line plot) and geometric mean infection intensity (columns) estimated from clams that had been placed in cages in the study area one month earlier. Asterisks indicate lack of data. The clams collected in February 2012 had been placed in the field two months earlier.

consumption of trophozoites of P. marinus infecting host oysters, and reported that 2.3 × 106 cells/g wet tissue and higher would likely result in the death of the host. Although the host and parasite in those studies differ from those studied here, those findings provide support that heavy infection with Perkinsus species can lead to the death of the host. In the present study, experimentally challenged Manila clams exposed to the natural environment in field cages had a lower growth rate than that of unchallenged control clams. Recently, we found that infection with P. olseni led to decreased growth rate and body mass index, as well as reduced filtration activities in experimentally challenged clams (Waki and Yoshinaga, 2018). Thus, the slower growth rate of the challenged clams in the present study likely resulted from poorer feeding because of decreased filtration activity. Nutrient absorption by trophozoites in clam tissues may also have affected the growth of the clams (Choi et al., 1989). The greatest decrease in density of the three major cohorts occurred in summer and autumn. In summer, the sediment temperature often exceeded 30 °C and occasionally reached 38 °C. The high temperature in summer probably increased the negative effects of infection on clam survival, and this effect lasted until autumn. Waki and Yoshinaga (2013) also showed that the lethal effects of P. olseni infection on survival of Manila clams increased as the temperature increased from 18 °C to 30 °C. No relationship was found between infection level and salinity, although Waki and Yoshinaga (2015) reported that the virulence of P. olseni was lower at salinity < 22‰ than at 30‰. In the present study, salinity fluctuated seasonally and occasionally fell below 22‰. In general, Manila clams are not tolerant to salinity as low as 22‰ for a

Fig. 8. Frequency distributions of shell length of live Manila clams. A–F represent clams collected at St. A–F on June 5, 2012 and G, H, and I represent clams collected from St. A, B, and C on August 3, 2012. Vertical bars represent SD. N represents number of live Manila clams collected.

(≥30 mm shell length) heavily infected with 104–107 cells/g gill weight. The infection intensities in the gill were reported to be close to those in the whole body (Choi et al., 2002). They suggested that the impact of P. olseni infection was not significant on Manila clams of that size. When adult and juvenile Manila clams (18–22 mm and 3–6 mm, respectively, in the shell length) were experimentally challenged with P. olseni, adults were found to be more tolerant than juveniles (Waki and Yoshinaga, 2013). Thus, the impact of infection is likely to be stronger on juveniles than on larger clams. Only a small proportion of the clams sampled in the present study had shell lengths of 30 mm or longer (Fig. 3). The inconsistency of the impact of P. olseni reported between Yoshinaga et al. (2010) and the present study can be explained by the difference in the sizes of clams examined in the two studies. Bushek et al. (2012) reported that mortality of eastern oysters (Crassostrea virginica) infected with Perkinsus marinus increased in areas where infection intensities were > 1.0 on the Mackin scale (estimated using histological sections) and suggested this was the lethal infection intensity of the parasite. Choi et al. (1989) estimated the energy 142

Journal of Invertebrate Pathology 153 (2018) 134–144

T. Waki et al.

both live and newly dead Manila clams were seldom sampled, the majority of Manila clams may have been washed outside of St. A by water movements, died under the continuous low salinity during the freshet, or both, as Manila clams began to die after 3 days in freshwater at around 20 °C (Kurashige, 1943). The generation of prezoosporangia in clams that died under low salinity may have contributed to new infections. Nevertheless, the extended period of infection acquisition after the depletion of the clam population needs to be explained. To assess the origin of the infection after the depletion of infected clams during this period, the transmigration and survival of the zoospores in the water surrounding the area should be investigated. The shell length distributions of newly dead clams did not fit a normal distribution. Choi (1963) reported that dead Manila clams were transported further than live clams by water movements in the field. Thus, it is likely that dead clams sampled in this study were carried by water movements, making the shell length distributions difficult to analyze. However, the effect of clam migration was not strong enough to explain the trends in the appearance of dead clams. In clams in aquaria, the difference in survival between challenged and unchallenged groups was lower in the third trial than in the first and second trials. The reason for this is unclear. The survival of clams may have been affected by their history, including their rearing environment after they hatched and settled in the hatchery facility, as well as by infection. The results of this study provide evidence for the negative effect of P. olseni on the wild population of Manila clams at Ariake Bay. Infection of the clams by P. olseni has been detected at sites throughout Japan; Hamaguchi et al. (2002) reported a prevalence of > 80% in 17 out of 68 locations. Other studies have reported P. olseni prevalence of > 80% from many other Manila clam populations in Japan (Ikeura, 2002; Momoyama and Taga, 2005; Nishihara, 2010; Park et al., 2008; Sakai and Onodera, 2006; Umeda and Yoshinaga, 2012). P. olseni prevalence of up to 100% was detected in this study. Clams heavily infected with P. olseni at 105 cell tissue−1 (gills only or whole soft tissue) and higher have been found at several sites in western Japan (Park et al., 2008; Umeda and Yoshinaga, 2012; Yoshinaga et al., 2010) and in this study. Many other Manila clam populations, especially those with high levels of P. olseni infection, are likely to be impacted. Considering the wide distribution of P. olseni in Japan, it is highly probable that this parasite is one of the major causes of the decline in Manila clam populations in Japan since the 1980s. Many field and laboratory studies have been conducted on Manila clams in Japan, especially to clarify the causes of declines in clam populations and to enhance clam stocks. However, most studies have focused on factors such as environmental changes, overfishing, and occurrence of new predators (Hirano et al., 2010; Matsuda et al., 2008; Matsukawa et al., 2008; Watanabe, 2007). The negative impact of P. olseni on the survival and physiology of individual clams has been repeatedly demonstrated by laboratory-based challenge experiments. The present survey and experiments on a wild clam population have clearly demonstrated the negative effects of the parasite. Considering the high prevalence and intensity of the parasite in many Manila clam populations in Japan, much more attention should be paid to P. olseni in field and physiological studies.

long period of time. They endure unfavorably low salinity by tightly closing their shells. It is unlikely that the infection levels at our site were affected by salinity. Perkinsus infection was not detected in Manila clams in the 2012 summer cohort-1, which appeared from April to June only. As small clams that passed through 2-mm sieve were collected only from December 2011, the clams in the 2012 summer cohort-1 were much smaller than newly recruited clams in other cohorts that were collected after late June, July, and early August (Fig. 6A and F). The absence of infection in the 2012 summer cohort-1 can be attributed to both the small clam size and the low intensity of new infections. There may be a minimal clam size for acquiring new infections. The trends in population density were similar for the minor cohorts and the major cohorts. The Manila clams present at the start of the survey in June 2009 (the 2009 autumn cohort) probably settled in 2008. The 2009 autumn cohort possibly interfused into the 2009 summer cohort in March 2010. Most clams disappeared from St. A in August 2012 because of the flooding of the Kikuchi River after heavy rain. Such disappearances of Manila clams associated with flooding of the Kikuchi River have been recorded repeatedly; in June 1979, June 1982, July 1987, July 1990, and July 1997 (Tezuka et al., 1980; Hirata and Nakamura, 1991; Nakahara and Nasu, 2002; Nakahara and Tobase, 2008; Ikushima et al., 2008). Therefore, the decrease in clam densities caused by floods is not the major cause of the continuous decline in the Manila clam population that began in the late 1980s. The prevalence of Perkinsus in uninfected Manila clams caged in the tidal flat represented the frequency of newly acquired Perkinsus infection during course of the experiment. However, infection intensity reflected trophozoite propagation in the hosts as well, which depends on the ambient temperature. The major period for new P. olseni infections was from summer to autumn at the present survey site, shown by increased prevalence and infection intensities in the uninfected clams placed in the field cages during these seasons. Bordenave et al. (1995) suggested that the zoospore is the infective stage to clams, and is released from zoosporangia that develop in the decomposed tissues of dead clams. Therefore, mortality of the infected clams in summer and autumn likely resulted in enormous numbers of zoospores that contributed to the high prevalence infection in the caged clams. In addition, the high infection intensity in summer and autumn was likely due to rapid propagation of P. olseni trophozoites in host tissue under high temperatures (La Peyre et al., 2008; Umeda et al., 2013). Nishihara (2010) reported that Manila clams challenged by immersion in a zoospore suspension in seawater for 24 h were infected at 25 °C but not at 20 °C. However, we observed that Perkinsus infection also occurred in caged clams during autumn and winter when the sediment temperature was lower than 20 °C. In the Nishihara (2010) challenge experiment, the immersion period may have been too short for the zoospores to successfully invade at temperatures below 20 °C. The prevalence and infection intensity showed similar seasonal dynamics, which is consistent with the influence of temperature in the propagation of trophozoites through the host tissues. Temperature thus increases infection intensity, favors host death, and leads to increase the production of zoospores and disease transmission. In Japan, juvenile Manila clams are released for commercial clamming and stock enhancement in areas where wild Manila clam populations have been depleted. Uninfected or mildly infected juveniles can be obtained from some of these areas (Hamaguchi et al., 2002). If juveniles are released during winter and spring when new transmission of P. olseni is rare, the juveniles may avoid P. olseni infection until the next summer, and mortality may remain low. Of course, since many factors other than P. olseni infection can affect juvenile survival, it is important to select the best time to release juveniles to reduce mortality. Perkinsus transmission and infection occurred even after August 2012 until November, although most of the wild Manila clams disappeared from St. A because of the heavy rainfall in July 2012. Since

Acknowledgements The authors thank Oita Prefectural Agriculture, Forestry and Fisheries Research Center for supplying hatchery-raised Manila clams. The authors thank anonymous reviewers for their comments much contributing to improvement of this article. Funding: This study was supported by Japan Society for Promotion of Science KAKENHI Grant Number 13J08152, 25252036 and 22380106.

143

Journal of Invertebrate Pathology 153 (2018) 134–144

T. Waki et al.

Appendix A. Supplementary material

low salinity on survival, hemolymph osmolality and tissue water content of the Manila clam Ruditapes philippinarum. Aquac. Sci. 56, 127–136 (in Japanese). Momoyama, K., Taga, S., 2005. Detection of a parasitic protozoa Perkinsus sp. in the clam Ruditapes philippinarum collected from the tidal flats along the Seto Inland Sea in Yamaguchi Prefecture. Bull. Yamaguchi Prefectural Fish. Res. Center 3, 111–117. Nakahara, Y., Nasu, H., 2002. Report from the coastal areas of Ariake Sound, Kumamoto Prefecture; the main fisheries ground for the clam Ruditapes philippinarum populations in Japan. Jpn. J. Bentholog. 57, 139–144. Nakahara, Y., Tobase, N., 2008. Mortality of Manila clam Ruditapes philippinarum in the estuary of Kikuchi River on July 1997. Report of Kumamoto Prefectural Fisheries Research Center, vol. 8, pp. 81–88 (in Japanese). Nishihara, Y., 2010. Infection of protozoan Perkinsus in the short necked clam (Ruditapes philippinarum) on the Hokkaido coastal region and the infection examination. Sci. Rep. Hokkaido Fisheries Exp. Stn. 77, 83–88 (In Japanese). Park, K.I., Tsutsumi, H., Hong, J.S., Choi, K.S., 2008. Pathology survey of the shortneck clam Ruditapes philippinarum occurring on sandy tidal flats along the coast of Ariake Bay, Kyushu, Japan. J. Invert. Pathol. 99, 212–219. Ray, S.M., 1952. A culture technique for the diagnosis of infections with Dermocystidium marinum Mackin, Owen, and Collier in oysters. Science 116, 360–361. Sakai, K., Onodera, J., 2006. Epizootiological investigation on Perkinsus Protozoan (Apicomplexa) infection in Manila Clam Ruditapes philippinarum in Miyagi Prefecture. Jpn. Miyagi Pref. Rep. Fish Sci. 6, 77–81 (in Japanese). Shimokawa, J., Yoshinaga, T., Ogawa, K., 2010. Experimental evaluation of the pathogenicity of Perkinsus olseni in juvenile Manila clams Ruditapes philippinarum. J. Invertebr. Pathol. 105, 347–351. Siddall, M.E., Reece, K.S., Graves, J.E., Burreson, E.M., 1997. Total evidence refutes the inclusion of Perkinsus species in the phylum Apicomplexa. Parasitology 115, 165–176. Takahashi, M., Yoshinaga, T., Waki, T., Shimokawa, J., Ogawa, K., 2009. Development of a PCR-RFLP method for differentiation of Perkinsus olseni and P. honshuensis in the Manila clam Ruditapes philippinarum. Fish Pathol. 44 (4), 185–188. Tezuka, H., Kugita, Y., Yoshida, K., Abe, H., Harada, M., 1980. Research of mortality of Manila clam in estuary region of Kikuchi River. Report of Kumamoto Prefectural Laver Research Laboratory, vol. 24, pp. 289–247. Umeda, K., Yoshinaga, T., 2012. Development of real-time PCR assays for discrimination and quantification of two Perkinsus spp. in the Manila clam Ruditapes philippinarum. Dis. Aquat. Org. 99, 215–225. Umeda, K., Shimokawa, J., Yoshinaga, T., 2013. Effects of temperature and salinity on the in vitro proliferation of trophozoites and the development of zoosporangia in Perkinsus olseni and P. honshuensis, both infecting Manila clam. Fish Pathol. 48, 13–15. Villalba, A., Reece, K.S., Camino Ordás, M., Casas, S.M., Figueras, A., 2004. Perkinsosis in molluscs: a review. Aquat. Living Resour. 17, 411–432. Waki, T., Yoshinaga, T., 2013. Experimental challenges of juvenile and adult Manila clams with the protozoan Perkinsus olseni at different temperatures. Fish. Sci. 79, 779–786. Waki, T., Yoshinaga, T., 2015. Suppressive effects of low salinity and low temperature on in-vivo propagation of the protozoan Perkinsus olseni in Manila clam. Fish Pathol. 50 (1), 16–22. Waki, T., Yoshinaga, T., 2018. Experimental evaluation of the impact of Perkinsus olseni on the physiological activities of juvenile Manila clam. J. Shellfish Res. (in press). Waki, T., Shimokawa, J., Watanabe, S., Yoshinaga, T., Ogawa, K., 2012. Experimental challenges of wild Manila clams with Perkinsus species isolated from naturally infected wild Manila clams. J. Invert. Pathol. 111, 50–55. Watanabe, S., 2007. Causes of decline in Manila clam production and attempts for recovery in Japan. Mon Oceanogr. 39, 229–233 (in Japanese). Yoshinaga, T., Watanabe, S., Waki, T., Aoki, S., Ogawa, K., 2010. Influence of Perkinsus infection on the physiology and behavior of adult Manila clam Ruditapes philippinarum. Fish Pathol. 45, 151–157.

Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jip.2018.03.001. References Aizawa, Y., Takiguchi, N., 1999. Consideration of the methods for estimating the agecomposition from the length frequency data with MS-Excel. Bull. Jpn. Soc. Sci. Fish. 63, 205–214 (in Japanese). Almeida, M., Berthe, F., Thébault, A., Dinis, M.T., 1999. Whole clam culture as a quantitative diagnostic procedure of Perkinsus atlanticus (Apicomplexa, Perkinsea) in clams Ruditapes decussates. Aquaculture 177, 325–332. Bordenave, S.A., Vigario, A.M., Ruano, F., Domart-Coulon, I., Doumenc, D., 1995. In vitro sporulation of the clam pathogen Perkinsus atlanticus (Apicomplexa, Perkinsea) under various environmental conditions. J. Shellfish Res. 14, 469–475. Bushek, D., Ford, S.E., Burt, I., 2012. Long-term patterns of an estuarine pathogen along a salinity gradient. J. Marine Res. 70, 225–251. Choi, K.S., Wilson, E.A., Lewis, D.H., Powell, E.N., Ray, S.M., 1989. The energetic cost of Perkinsus marinus parasitism in oysters: quantification of the thioglycollate method. J. Shellfish Res. 8, 125–131. Choi, K.S., Park, K.I., Lee, K.W., Matsuoka, K., 2002. Infection intensity, prevalence, and histopathology of Perkinsus sp. in the Manila clam, Ruditapes philippinarum, in Isahaya Bay. J. Shellfish Res. 21, 119–125. Choi, S., 1963. Transmission of Manila clam. The Aquiculture 11, 13–24. Dungan, C.F., Reece, K.S., 2006. In vitro propagation of two Perkinsus spp. parasites from Japanese Manila clams Venerupis philippinarum and description of Perkinsus honshuensis n. sp. J. Eukaryot. Microbiol. 53, 316–326. Hamaguchi, M., Suzuki, N., Usuki, H., Ishioka, H., 1998. Perkinsus protozoan infection in short-necked clam Tapes (=Ruditapes) philippinarum in Japan. Fish Pathol. 33, 473–480. Hamaguchi, M., Sasaki, M., Usuki, H., 2002. Prevelance of a Perkinsus protozoan in the clam Ruditapes philippinarum in Japan. Jpn. J. Benthol. 57, 168–176 (in Japanese). Hirata, I., Nakamura, Y., 1991. Research for management of protected waters. Report of Kumamoto Prefectural Fisheries Research Center, pp. 50–51 (in Japanese). Hirano, K., Higano, J., Nakata, H., Shinagawa, A., Fujiya, T., Tokunaga, M., Kogo, K., 2010. An experiment for preventing mass mortality of cultured short-neck clams due to hypoxia formation during summer in Isahaya Bay. Fish. Eng. 47, 53–62 (In Japanese). Ikeura, S., 2002. Retention of Perkinsus spp. in Japanese short-neck clam at Buzen Sea. Bull. Fukuoka Fish. Mar. Tec. Res. 12, 127–129 (In Japanese). Ikushima, N., Nasu, H., Jinnai, Y., Nakahara, Y., Tobase, H., 2008. Distribution of Manila clam in estuary of Kikuchi River (Nameishi area) from 1996 to 2006. Report of Kumamoto Prefectural Fisheries Research Center, vol. 8, pp. 47–58 (in Japanese). Ito, H., 2002. What kind of animal is the Clam Ruditapes philippinarum?-Introduction to its ecology and fishery. Jpn. J. Bentholog. 57, 134–138 (in Japanese). Kurashige, E., 1943. Experimental observations on the lethal salinity of Paphia philoppinarum, Adams et Reeve. J. Oceanogr. Soc. Jpn. 1, 29–43. La Peyre, M.L., Casas, S.M., Villalba, A., Peyre, J.L., 2008. Determination of the effects of temperature on viability, metabolic activity and proliferation of two Perkinsus species, and its significance to understanding seasonal cycles of perkinsosis. Parasitology 135, 505–519. Matsukawa, Y., Cho, N., Katayama, S., Kamio, K., 2008. Factors responsible for the drastic catch decline of the Manila clam Ruditapes philippinarum in Japan. Nippon Suisan Gakkaishi 74, 137–143 (in Japanese). Matsuda, M., Shinagawa, A., Higano, J., Fujii, A., Hirano, K., Ishikawa, J., 2008. Effects of

144