**3. Discussion**

The main fouling species of *C. gigas* aquaculture in the studied bay were sponges, algae, and *M. galloprovincialis* regardless of sampling site and depth, a finding that was partially inconsistent with some previous aquaculture studies conducted in Japan, which reported that mussels were dominant [25–27]. Mazouni et al. [28] reported that the predominant fouling organisms on *C. giga* clusters were ascidians. Royer et al. [29] reported that *C. gigas* clusters were mostly fouled by barnacles. Rodriguez et al. [22] found that ascidians, bryozoans, sponges, hydrozoans, and algae were the predominant colonisers on oyster farming beds, indicating that biofouling communities differ compositionally, even for the same host species [30]. This can be explained by the difference of environmental factors and climate among previous reports and the current study. Spatiotemporally variable factors, such as water temperature [22], the season [28], and larval supply from the benthic community [31], influence the settlement, abundance, and community structure in oyster farms.

While the negative effects of fouling organisms on oysters have been widely reported [23,32], some studies have indicated no significant effect of fouling organisms on cultivated oysters [29,33]. These inconsistent findings may indicate that the effects of fouling organisms on host oysters are species-specific and depend on focal factors (e.g., growth rate, survival rate, etc.) [30]. Similarly, in this study, the effects of sponges, macroalgae, and *M. galloprovincialis* on the fatty acid content and condition of *C. gigas* were different.

Sponges and macroalgal mixtures seem to have decreased the DHA content of *C. gigas* individuals. It is known that *C. gigas* have a poor ability to carry out biosynthesis of DHA [34,35]. Thus, the reduction of DHA in *C. gigas* implies that the dietary intake of DHA sources was reduced. DHA is abundant in some algal classes, such as dinoflagellates [36] and haptophyta [37]. For Shizugawa bay near farm A in July 2017, dinoflagellates were detected; however, haptophyta were not observed (Sakamaki, unpublished data, Table S1). Since sponges are suspension feeders [38], dietary competition for dinoflagellates between oysters and sponges is one of the possible mechanisms for the reduction of DHA in oysters. However, sponges generally ingest smaller particles, and their main dietary sources are known to be bacteria [39]. In addition, sponges can meet their dietary requirements through ingesting particles of <1 μm in diameter [40]. Therefore, competition is unlikely to explain the reduction of DHA in oysters, because most dinoflagellates are larger than 1 μm in diameter.

For macroalgae, allelopathy could be a possible mechanism for the reduction of DHA in oysters. Brown algae and *Ulva* spp., which were dominant macroalgae in our oyster aquaculture, have been demonstrated to significantly reduce the growth of dinoflagellates through allelopathy [41,42]. This hypothesis could explain the reduction of DHA in oysters if we assume that the allelopathy was more effective for dinoflagellates. In fact, speciesspecific allelopathy effects have been demonstrated, including *Ulva* and dinoflagellates [43]. However, further research is needed to clarify the mechanisms behind the DHA reduction in *C. gigas* when sponges and macroalgae act as fouling organisms.

A mixture of sponges and macroalgae also seem to have reduced the amount of EPA and DHA in *C. gigas* clusters. Since a negative relationship between the total weight of *C. gigas* in a cluster and the relative weight of the sponge and macroalgal mixture was detected, the observed reduction of EPA and DHA in the oyster clusters can be explained by the reduction of the total biomass of the oysters, as the relative weight of the sponges and macroalgae increased.

*M. galloprovincialis* seem to have decreased the EPA content and CI of *C. gigas*. Although *C. gigas* can biosynthesise EPA from its precursors, its conversion efficiency is not enough to meet its requirement, and the EPA content in *C. gigas* mainly represents dietary EPA [35]. Thus, the amount of EPA in *C. gigas* seemed to depend mainly on their dietary intake, rather than on their own biosynthesis. Although EPA is abundant in diatoms, Cryptophyceae, and Rhodophyceae [36], the main EPA source for the assessed *C. gigas* was diatoms, because diatoms were dominant near our study sites (Sakamaki, unpublished data, Table S1). Furthermore, the observed significant positive relationship between EPA and 16:1n7/16:0 in oysters (Figure 5), which have been used as diatom markers, indicated that the main origin of EPA in the oysters were diatoms [44]. Pernet et al. [45] reported high bivalve growth rates during diatom bloom periods, and a feeding experiment by Piveteau et al. [46] also demonstrated an increase in the condition index of *C. gigas* when feeding on diatoms. In addition, a positive relationship was found between the EPA content and growth of a mussel species, *M. edulis* [47]. This was further demonstrated by the significant positive relationships between CI and EPA content in both bivalve species in our study. These findings support the idea that diatoms are a high-quality dietary source for both *C. gigas* and *M. galloprovincialis* and also indicate that there is probably a high competition potential for diatoms between *C. gigas* and *M. galloprovincialis*. Therefore, the condition index of *C. gigas* can be reduced as a result of competition with *M. galloprovincialis* for diatoms, with a consequent reduction of EPA acquisition.

Although these species have the potential to compete for diatoms, the CI of *M. galloprovincialis* was not negatively affected by the presence of *C. gigas*. This indicates that the dietary competition between *C. gigas* and *M. galloprovincialis* is not balanced. Although both *C. gigas* and *M. galloprovincialis* preferentially selected larger particles (>5 μm) in their diet, they did not necessarily need to compete, because *M. galloprovincialis* can also utilise smaller particles (<2 μm), which are not retained by *C. gigas* [48]. In fact, the invasion of *C. gigas* did not negatively affect local populations of the mussels *M. edulis* in Limfjord,

Denmark, even though *C. gigas* were considered to have a competitive advantage owing to their higher filtration rate [49]. Contrary to this, our results clearly demonstrated that *M. galloprovincialis* has an advantage in dietary competition over *C. gigas*. There are two possible reasons for this. First, more than 97% of the EPA was distributed in particles of >2 μm near farm A in Shizugawa Bay [50], and dietary segregation by utilising small particles (<2 μm) by *M. galloprovincialis* was not valid in our study fields. The second reason was the vertical distribution of *C. gigas* and *M. galloprovincialis* in the cluster. *M. galloprovincialis* develops on the shells of *C. gigas* in Shizugawa Bay since oyster spats are artificially settled on the surface of scallop shells and grow before the scallop shells are put in the bay. As *M. galloprovincialis* settles on the surfaces of *C. gigas* shells, *M. galloprovincialis* has a spatial advantage in terms of feeding on diatoms before *C. gigas*. A portion of the diatoms ingested by the bivalves can survive [51], indicating that the faecal material of *M. galloprovincialis*, including diatoms, may supply the *C. gigas*, which are located inside the cluster. However, faecal material generally contains less EPA than suspended matter [50]. Therefore, *M. galloprovincialis* fouling on *C. gigas* could have substantial negative effects on the *C. gigas* in oyster aquaculture farms, in terms of EPA acquisition.

**Figure 5.** The location of sampling points in Shizugawa Bay, Japan.

Although our data indicated that fouling organisms possibly reduce the EPA and DHA content of *C. gigas*, it should be noted that other environmental factors can also affect the EPA and DHA content of *C. gigas*. For instance, water temperature [52] and the reproductive cycle [53] are known to affect fatty acid profiles of oysters. In addition, the total amount and quality of supplied food sources, especially diatoms and dinoflagellate, may influence the EPA and DHA content in *C. gigas*. In this study, *C. gigas* were all collected with two-year-old oysters with similar sizes on the same day, which indicates that the effects of water temperature and the reproductive cycle can be assumed to not produce the difference of EPA and DHA content. Unfortunately, as we did not investigate the supplied food at each sampling point, further study is required to understand the effect of food availability. Furthermore, the fatty acid content of oysters changes spatially and seasonally [54,55], and this could be associated with composition and the amounts of fouling organisms. Therefore, long-term monitoring in different sites is an effective way for the comprehensive understanding of the effects of fouling organisms on the EPA and DHA content in cultivated species.

Although EPA and DHA are essential fatty acids for humans [5,6], the effect of fouling organisms on the EPA and DHA content in cultivated host species has not been compre-

hensively evaluated before. Our findings demonstrated a reduction of EPA and DHA in the cultivated oyster *C. gigas* likely due to fouling organisms. This can devalue the quality of the oysters as an aquaculture product. Removing fouling mussels is empirically known to reduce their negative impact on oyster growth [25,27]. Our results support the idea that the current efforts to remove fouling mussels from oyster clusters in the study region [56], which include hot water treatment and physical removal, are expected to enhance the content of EPA and DHA in the oysters.

#### **4. Materials and Methods**

### *4.1. Study Site*

This study was conducted in the inner part of Shizugawa bay, located on the northeast side of Honshu Island, Japan (38.65◦ N, 141.50◦ E; Figure 5). The area of the bay is 46.8 km2, and the average and the maximum depth at the bay mouth is 30 and 54 m, respectively. In our observations in 2015, the annual range of seawater temperatures at a depth of 2 m was approximately 5–21 ◦C. The Pacific oyster *C. gigas* is one of the major aquaculture products in this bay, and longline oyster suspension facilities are distributed in the inner parts of the bay, in which the depth ranges from approximately 10–30 m. Oyster spats are artificially settled on scallop shells. Then, oyster clusters growing on scallop shells are tied at ~0.4 m intervals to ropes of approximately 8 to 10 m in length, and the ropes are vertically suspended from ~100 m longlines that are horizontally sustained by floating buoys. There are ~400 oyster farming longline facilities in the bay, which is based on the information provided by a local fishery cooperative.
