Growth and absorption response of Perna viridis (Lin, 1758) to fish farm water quality in Valladolid, Carcar, Cebu, Philippines

The extractive species, green mussel (Perna viridis) has been integrated into the existing commercial finfish culture of Chanos chanos and Trachinotus blochii in Valladolid, Carcar Bay, Eastern Cebu, Philippines, following the Integrated Multitrophic Aquaculture (IMTA) concept. Two sampling sites (573 m apart) were selected; the experimental site, and control site. Biomass and growth rates of this species was measured monthly, along with physico-chemical parameters in the two sites, monitored for a period of up to one year. Results showed that P. viridis was adopted to the fish farming site where physico-chemical parameters like temperature, salinity, pH, dissolved oxygen were within the optimal range. Growth of P. viridis was mainly affected by predators and fouling organisms present during culture which reduced their potential biomass and growth rates. The introduction of green mussels in the fish farm appeared to have checked the adverse effects of fish culture activities daily inputs of commercial feeds, could led to the suspended solids. Transplanted green mussels grew into marketable after 7 months with average specific growth rate (SGR) reached highest level at 1.03% in February 2019. The use of IMTA concept, following the culture of green mussels within the cages area is found to work well in the culture of milkfish/pompano, and should be promoted and expanded in the Philippines.


Introduction
A study by Liutkus et al. (2012) shows that mussels can absorb up to 86% of feces (organic) and 90% of dissolved feeds (inorganic form) from salmon farming. The growth and biochemical profile of mussels are greatly affected by the environmental conditions in which they exist such as the type of organic matter from fish farm (biodeposition) (Sara et al., 2009). However, storms, tidal cycles, and current speed can resuspend bottom materials, increasing the concentration of seston in the upper portion, thus reducing the food available for bivalve filterfeeders at the bottom (Cranford et al., 2013;Karayücel et al., 2013). Furthermore, current speed and flow also affect phytoplankton transport, mussel clearance rate, and settling of organic and inorganic material on the cultured bivalves, while extreme waves can cause mortality and limit food intake (Buck, 2007). Complex interaction between all the aforementioned environmental parameters that influence growth and health of the animal must be considered. Bayne and Worrall (1980) showed that food quantity, quality, physiological availability, and temperature

Study site description
The capacity of the P. viridis in removing both suspended particles and dissolved inorganic nutrients was investigated within a fish farm located in Sitio Tawog, Valladolid, Carcar, south of Cebu as described in Figure 1a. Here, the green mussels were deployed in the vacant area between two rows of fish cages designated as left and right cages, shown in Figure 1b.

Collection and culture scheme of green mussel shells
Juvenile shells of P. viridis were obtained from bivalve collectors in the western part of Negros Occidental, Philippines on December 2018. About 5,000 pcs of the young shellfish specimens were brought to the experimental IMTA site in Valladolid, Carcar, Cebu where they were transferred to a nylon net bag and acclimatized for 1 week below the sea surface (30 cm deep) beside a floating plastic frame near the fish farm's ward house. The condition of the shells was checked daily prior to their deployment and dead shells were removed. After 1 week of acclimatization, at the start of culture, shell length of P. viridis individuals were measured again, after which they were transferred to nylon net bags following the design of Wong and Cheung (2001). The net bags containing 300 pcs of juvenile green mussels were tied with rope to the fiber frame of the fish cages (Figure 2a and 2b). This suspension culture technique was a modification of that of Langan (2009), to suit the experimental site, wherein each net bag was lined with multiple vertical branching ropes (0.6 m long abaca fiber) tied inside the net bag. Following the method of Azpeitia et al. (2018), individual mussel shells were carefully detached from the ropes and cleaned from encrusting organisms. All ropes and net bags were cleaned to remove fouling organisms twice a month as part of a standard maintenance protocol.

Growth monitoring
A total of forty-five pieces or 15% of the cultured green mussels were randomly sampled from two culture nets (20 and 25 pcs each from two selected nets) for monthly growth measurements, namely: 1) total shell length, 2) shell width, and 3) whole wet weight (shell plus meat). The shell length and width were measured using a plastic ruler (fixed in a plastic cardboard) and a plastic vernier caliper while shell weight was measured using a weighing balance. Another batch of 45 pcs of juvenile P. viridis from another culture unit was also measured. The growth of green mussels deployed besides the fish cages (i.e. left and right of the cages) were physically compared in terms of their biomass, shell length and width, after which they were returned back to the culture nets. All measurements were conducted in the morning (9 to 10 AM). Percent mortality and survival rates were monitored weekly in the entire duration of the study. Dead and lost shells were replaced with new ones from the reserved stock maintained near the ward house.
Growth performance of the green mussels was calculated using the formula of Reid et al. (2018):

SGR (%biomass ÷ d) = 100 x LN (Wf -Wi)/d
Where; Wf and Wi are the final and initial wet weights of the shells sampled in each station (g); d = the duration of the experiment in days; LN = normal logarithm.

Water quality monitoring, including physico-chemical parameters
Water samples for nutrient analysis (ammonia, nitrate and phosphate) and for Total Dissolved Solids (TDS) and Total Suspended Solids (TSS) were taken at 2.5 m deep using a water sampler, from both the experimental and control sites, for the whole duration of the study. These were done in two phasesthe first phase or a total of 4 months (August to November 2018) was before the deployment of extractive species (pre-IMTA) to serve as baseline data; and the second phase or a total of 7 months (January to July 2019) during the IMTA implementation. The samples, in 1.5 L volumes for nutrients and 1.5 L for TDS and TSS, were taken between 1:00 to 2:00 PM. The samples from the first phase (pre-IMTA) were collected into polyethylene and glass bottles, placed in a chest cooler and brought to the Water Quality First Analytical Services and Technical Cooperative (F.A.S.T.) Laboratories in Highway, Mandaue City, Cebu, within less than 24 h from collection. Water samples from the second phase, instead of submitting to FAST Laboratories (which charged higher analysis fees), were submitted to the USC Water Laboratory, in Talamban Campus, Cebu City. In both laboratories, water samples were analyzed using colorimetric and gravimetric methods.
Water samples for the physico-chemical parameter measurements were also taken from 2.5 m depth in both the experimental and control sites twice a month. Water temperature was immediately measured in situ twice a month. Water temperature and dissolved oxygen (using YSI 550A DO meter), salinity (using a digital refractometer -Trans Instruments), and pH (using a digital pH meter, American Marine Inc. brand), were also measured. Four to five replicates of water samples were randomly obtained at 2.5 m depth within the fish cage perimeter.

Fish feed profile and composition analysis
Proximate analysis content profile of fish feed where examined as presented in Table 4. Meat sample from January and another batch in July 2019 of the green mussel weighing 200 g each were submitted for proximate analysis of protein, glycogen and ash content to the Laboratory of the Department of Science and Technology (DOST) -7, Lahug, Cebu City Philippines.

Statistical Analysis
All water quality parameters were analyzed using F comparison between the experimental and control sites with an independent two-tailed test application to check the differences between the nutrients in the two sites. All data are presented herein as means ± standard deviation (SD). Physico-chemical parameters namely DO, temperature, salinity, and pH were analyzed for minimum and maximum values.  Figure 3). The negative SGR value in March 2019 was due to predation of the still young mussels but SGRs recovered in the succeeding months of April to July 2019. Monthly average biomass (n = 90), on the other hand, increased from an initial weight of 225.40 +2.92 g in January to reach a maximum weight of 447.80 ±4.20 g in July or after 7 months ( Figure 4). Although P. viridis had consistent positive growth rates, mortality ©2021 The Authors. Journal of Nature and Applied Research by NAT ARK www.natark.com | 42 was observed starting April which could be attributed to an abundance of predatory crabs and blister worms, and the proliferation of sponges inside the culture nets which could have competed for food for this filter-feeding mollusk. These factors (i.e. predation and competition) resulted to an overall P. viridis mortality of 21.9% for 7 months, with highest single month mortality rate of 6.7% obtained in July, and affected the biomass production for that month as shown in Figure 4.  Table 1 shows the specific growth rate of green mussel spats in terms of shell length (SL) and shell width (SW) from its initial size measured in December 2018 or during pre-IMTA phase and during IMTA implementation from January to July 2019 parallel the fish culture. A sudden increase in shell size was observed in March and May followed by a slight decrease in shell length June. Highest shell size in terms of width (SW) and length (SL) (2016) and Layugan et al. (2018) obtained a maximum size of over 40 mm after 6 months. Results in this study for 6 months alone (January to June 2019) showed a comparable growth and have reached SL of 49.70 ±4.16 mm in a random selection of shells from the net bags. Based on personal interviews with a local grower and supplier of green mussels in Pasil Market, Cebu City, green mussels can have better growth when deployed or positioned near the river mouths, where there is abundance of organic particulates as their food. However, more research is needed to support this claim and whether it is also applicable to Carcar Bay area. Up to what extent will this maximum growth of bivalves (in terms of commercial size) can be achieved in the IMTA site remains to be determined, which is beyond the scope of the present study.

Results and discussion
The transplanted green mussel grew into marketable size after 7 months based on weight and shell size (length and width) with lesser mortality after 4 months onward. The site where the mussels originated has a significant impact on bivalve mortality (Fuentes et al., 2000) because it will create more efforts for the transplanted species to adjust in the new culture environment. Site suitability according to Sallih (2005) was very important for the culture of bivalves. Consequently, the growth of the bivalves in the IMTA set up can be affected by the conditions of surrounding environment within the bay, with significant contribution from the river runoffs containing anthropogenic wastes. Seawater quality depends on several internal (autochthonous) and external (allochthonous) factors, such as particulate matters that are brought down into the bay during rainfall, freshwater inflow carried by tidal movement and biological activities. These factors have been found to contribute to water quality in the coastal environment of the southeast coast of India during summer (Satpathy et al., 2010).
More detailed analysis of planktons in the site should be able to provide detailed information about the composition and quantity of food organisms present in the area. A site with high diatom concentration is considered good for green mussel farming, while an area with high turbidity, low salinity and low dissolved oxygen should be avoided (Sing and Ransangan, 2019). The culture of bivalves close to the fish farm has many benefits because it will buffer estuaries and coastal ocean waters against excessive phytoplankton blooms, counteracting the symptoms of eutrophication; they may also remove inorganic sediments from suspension, counteracting coastal water turbidity according to National Research Council of the United States (2010). However, in Cavite, Philippines, a decline in mussel production was due to water pollution (Layugan et al., 2018), which means that the optimal range of water quality for mussels to survive has to be observed. The extractive species being added to the monoculture of finfish (e.g. Chanos chanos and Trachinotus blochii), has been proven to use the waste products in the form of dissolved nutrients (nitrate, ammonia, and phosphate) and suspended solids. The efficiency of such waste extraction would depend, to some degree, on the positioning of the extractive species vis-à-vis to fed species. Being filter-feeders it was found in this study that they were better positioned in the water column (at 2.5 m deep) where the fed fish are more concentrated, to allow for the removal of suspended organic particles.
To attain sustainability, fish farming operators must know the kind of fish species to be cultured, feeds being supplied, feed ingredients, and temperature of the seawater (Blancheton et al., 2007 and2009;Schneider et al., 2005Schneider et al., , 2006Schneider et al., , and 2010. It is worth noting that such information could be helpful to sea farming industries in facilitating the proper adoption technique specific to tropical countries like the Philippines. It is the responsibility of the aquaculture industry to enhance the culture techniques and to discover the equitable balance between economic return and sustainability of the cultured species. There is no doubt, however, that these efforts are being pursued to meet the ecological carrying capacity of the environment where the species is cultured. Sustainability is not an endpoint, but rather a trajectory of constant improvement (Hargreaves, 2011). This initial culture development can work as a basis for future ventures to integrate several species of different trophic levels in a holistic approach. Yet, the proximity of IMTA market and consumers is also a factor to be thought of, among others.

Physico-chemical factors
Physico-chemical parameters showed monthly variations at the water column (i.e., at 2.5 m deep) where the mussels were cultured in the two culture sites (Table 2, Figure 5). The monthly water temperature over a 12-month monitoring period, i.e. from August 2018 to July 2019, in the experimental and control sites had average values of 29.14 ±0.16 and 28.86 ±0.42 °C, respectively. The monthly average temperatures were 29.14 °C and 28.86 °C in experimental and control site, respectively. However, at 2.5 m, the water temperatures could be as low as 26.62 o C in the control site which was situated at the deeper portion of Carcar Bay. Generally, all these values are within the optimal temperature requirements of the green mussel P. viridis. Water temperature is one of the most important water quality parameters in an aquatic environment because it determines certain biological activities (Gupta, 2005). floods (Upadhyay, 1988). The slightly higher values of above normal pH in the control site could be due to organic pollutants from human settlements (Tan and Ransangan, 2015) being delivered from the river outlet located 575 m away from the experimental site of the fish farm.
On the other hand, low pH could be attributed to the organic effluents from land (Sany et al., 2014). Considering the excessive introduction of CO 2 into the oceans, the pH of seawater is expected to shift towards the acidic side although current model calculated pH show values to still hover around 8.2 with a variation between 8.08 and 8.33 (Marion et al., 2011).
In terms of salinity, an average value of 35.01 ±0.23 psu was obtained in the control site, wherein a maximum value of 35.70 ±0.27 psu was observed in July and a minimum of 34.50 ±0.00 psu was observed in October 2018. This high salinity range was recorded during the hotter period of the year in the control site. In the experimental site, an average salinity of 34.72 ±0.07 psu was obtained, with minimum and maximum values of 34.00 ±0.00 and 35.00 ±0.00 psu were obtained, respectively. Salinity in both experimental and control site ranged 34 and 35 psu which are well within the normal salinity range for marine waters and the optimal growth requirements for P. viridis (Rajagopal et al., 2006;Tan and Ransangan, 2016). A slight salinity drop in both sites in December 2018 and January 2019 could be due to the successive heavy rains before the measurements were being conducted. These salinity values hovered around 35 psu in the both sites, but with few instances where salinity could go up to 36 psu for certain months, even during and after heavy rains. Fish farm sites in Carcar Bay are in quite deep water as compared to other sites in the province of Cebu where there are also fish farming activities, thus leaving a little chance for a rapid salinity drops, except from the river outlet of the bay, which is facing less than 1 km from the experimental site. Table 2: Physico-chemical parameters (random replicates, n= 8 to 10 ±SD per month) before and during deployment of green mussels at 2.5m depth in the experiment site and in the control site (no green mussels deployed).
DENR Administrative Order (DAO) 2016-08-WQG under the classification of the marine waters in the Philippines for its commercial use placed water quality standard to have a pH range of 6.5 -8.5 and dissolved oxygen of 5.0 mg/L. An Ideal range of several physicochemical factors for the culture of green mussel Perna (as Mytilus) edulis as found in neighboring Malaysia, including dissolved oxygen at more than 8 mg/L, salinity of between 27 and 32, water temperature of 26-32 o C, and pH of 7.9 to 8.2 (Sivalingam, 1977). The aquaculture sites of Carcar Bay have values falling within these ranges.
In terms of dissolved oxygen (DO) concentration, readings varied between the two sampling sites wherein highest value of 6.23 ±0.36 mg/L was obtained in December 2018 and lowest value of 5.12 ±0.14 mg/L obtained in August 2018 in the experimental site. These months were before the deployment of the green mussels (pre-IMTA). On the other hand, highest value of 6.56 ±0.24 and lowest value of 5.24 ±0.19 mg/L were obtained in October and December 2018, respectively. The occurrence of successive rains and two typhoons in Central Visayas during the study period could have altered the dissolved oxygen concentrations due to stronger waves compared to other times of the year in 2018. Strong waves also occurred on certain days in May 2019. Consequently, the growth of bivalves in the experimental site could have been affected by the prevailing environmental conditions on the bay together with inputs of wastes from the nearby river. Dissolved oxygen (DO) levels in both sites which went from 5.05 -6.77 mg/L conformed to the threshold set by DENR DAO 2016-08 to be at least 5 mg/L. Adequate DO is essential for fish farm operation (Price et al., 2015). The highest DO readings occurred in December 2018, while the low DO reading of 5.12 mg/L occurred in September 2018. If dissolved oxygen values were also to be used as an indicator of flushing activity, its higher average value in the control site would show that this was the case compared to the experimental site where its average value was lower. This can also be seen in the average pH value over an 11-month period where the experimental site had lower average pH than the control site. Moreover, DO remained high in the control site at deeper waters, indicating no issues on this parameter that would deprive the cultured animal species -both the fed species (fish) and the extractive ones -with oxygen that usually happens in more stagnant or less dynamic water bodies (Price et al., 2015).

Nutrient profile
Nutrient concentrations showed monthly fluctuations during the IMTA implementation from January to July 2019 with the deployment of green mussels in the experimental site ( Figure 6). High monthly average concentrations of 0.057 mg/L for ammonia, 0.264 mg/L for nitrate and 0.034 mg/L for phosphate were observed in the experimental site as compared to the control site where lower concentrations of 0.043, 0.157 and 0.022 mg/L were observed for ammonia, nitrate and phosphate, respectively (Table 3). High ammonia and phosphate compounds could be related to increased water temperature during summer period and the prevalence of anoxic conditions in deep waters would give rise to organic compound decomposition that produce ammonia and phosphate compounds (Markou et al., 2007;Spears et al., 2008). Generally, 0.02 mg/L value of ammonia in water is harmless (PHILMINAQ, 2008). Mussels cultured on one side of the fish cages ("Left side") (in Figure 1b) showed better condition and growth profile at the time of sampling and final harvest of July 2019 than its "right side" counterpart, supporting the hypothesis of a more stable parameter for P. viridis growth. In our opinion, in order to see more marked differences between sites, the distance from the river mouth of the bay should be considered. Based on our interviews to local farmers, P. viridis grow better when the bivalve is deployed or positioned near the river due to an abundance of organic particles serving as food. Whether maximum growths for mussels are achievable in the study site is a function of many biophysical parameters, not pursued in this study. However, knowledge of seawater nutrient changes prior to and during the study can be useful to frame up expectations of nutrient extraction and to assist in the interpretation of the augmented growth of this bivalve (Petersen et al., 2019), and a tracer data as a means to infer nutrient reduction is also needed. Absorption efficiency is a common goal of IMTA involving different species such as combining the sea cucumber (Parastichopus californicus) with Pacific oyster (Crassostrea gigas) (Paltzat et al., 2008), sea cucumber (Cucumaria frondosa) and blue mussel (Mytilus edulis) with salmon (Salmo salar) (Reid et al., 2010;Nelson et al., 2012) and C. gigas with sea bass (Dicentrarchus labrax) (Lefebvre et al., 2000).
Although inorganic nutrients in the form of ammonia, nitrate and phosphate showed little variations in average values between the experimental and control sites (Table 3), fluctuations were observed on a monthly basis before and during IMTA implementation in both experimental and control sites ( Figure 6). Phosphate had the highest concentration in August 2018 (0.78 mg/L), which was before the implementation of IMTA in the experimental site. It was then taken over by nitrate during IMTA implementation in March (0.547 mg/L) and April 2019 (0.376 mg/L) as the dominant nutrient with highest concentration in the experimental and control sites. The increase in nitrate from March and April 2019, matched the peak production periods of the cultured milkfishes in both experimental and control cages entailing the increased amount of commercial feeds provided to the fish. Ammonia, which is among the waste products of the cultured aquatic animals, remained low before IMTA (0.027 mg/L in August -November 2018). As indicated, it rose up higher than phosphate during the IMTA ©2021 The Authors. Journal of Nature and Applied Research by NAT ARK www.natark.com | 48 implementation (0.058 mg/L from January to July 2019) in the experimental site ( Figure 6), even when the extractive species was deployed. Generally, 0.02 mg/L value of ammonia in water is harmless (PHILMINAQ, 2008). At 2.5 m water column, all three nutrient compounds had high average concentrations at the experimental than in the control site (Table 3). The study sites can be contrasted more in terms of their hydrodynamics function which could explain the difference in their water quality parameters. For instance, the INCA circular fish cages (as control site) being positioned in a deeper portion (about 35 m deep), has a more dynamic water exchange and efficient flushing capacity for wastes coming from the fed fish. As it was being indicated by the lower concentration of nutrients, except for ammonia near the surface water.
In contrast, the experimental site was set in a relatively shallower depths of only 13 to 15 m during low tide. In this site, the nitrate and phosphate have relatively higher average concentrations (0.226 and 0.149 mg/L, respectively) over an 11-month study period than the average concentrations of the same nutrients (0.215 and 0.085 mg/L, respectively) in the control site in the same period. An F comparison test between the two sites showed a significant difference for ammonia and phosphate but not in nitrate concentrations. A study on the biogeochemical responses on the removal of the mariculture structures in Tapong Bay, Taiwan, explained the added inputs of nutrients into the lagoon from external sources aside from the mariculture structures (i.e., discharges from land-based wastes) by the river into the sea (Hung et al., 2008). This situation could be possible to happen in Carcar Bay, where the fish farm is located close to a river. Based on the F comparison for these two organic nutrients, the TDS was slightly reduced (Fstat = 0.751) than the TSS (Fstat = 3.043) in the experimental site which could be attributed to the deployment of the IMTA species in this site. The proximate composition of the experimental IMTA species P. viridis was shown in Table 4. The live weight increment of the green mussel for 7 month culture period obtained 5.38. Their crude protein, and ash content increment were; 1.1 and 1.67. However, the total fat content of the cultures were almost non-detectable (<0.35). A non-significant tendency for an increased percentage of crude protein with an increased dietary protein and body weight level was observed.   Based on the F comparison for these two organic nutrients, the TDS was slightly reduced (Fstat = 0.751) than the TSS (Fstat = 3.043) in the experimental site which could be attributed to the deployment of the IMTA species in this site. The proximate composition of the experimental IMTA species P. viridis was shown in Table 4. The live weight increment of the green mussel for 7 month culture period obtained 5.38. Their crude protein, and ash content increment were; 1.1 and 1.67. However, the total fat content of the cultures were almost non-detectable (<0.35).
A non-significant tendency for an increased percentage of crude protein with an increased dietary protein and body weight level was observed.  The biomass of fish in the control site, which were placed in INCA cages, was 3 to 4 times higher in stocking density (based on company information) and therefore requires more feeds than those in the experimental site, the released wastes coming from the cages are expected to be enormous. However, the average concentrations of these nutrients were relatively low in the control site which, as explained above, could be attributed to the location of this site being in the deeper portion of Carcar Bay. Inspite of the overall relatively high concentrations of ammonia, phosphate, and nitrate (Fstat = 1.139, 1.607, and 2.195, respectively) in the experimental site than in the control site, it would appear from this study that P. viridis as a biofilter was working for these nutrients, especially if the monthly nutrient differences and feeds supplication between the experimental and control sites were being compared.

Conclusions
The study has shown that P. viridis suspended near the fish farm cages are capable of utilizing the excess organic and inorganic nutrients. Despite in the presence of predators and fouling organisms, these green mussels that have grown into a harvestable size after a 7-month period based on body weights, length and width of shells, can be an augmentation to the existing monoculture fish farm. In view of the positive growth performance with no adverse effects on or significant changes in the water quality of the surrounding fish farm waters, green mussels have generally been shown to work well as an IMTA combination within the finfish culture. However, factors that could affect, such as predators, water quality, and excess nutrients from feeds, will need to be closely monitored so that the environmental conditions will be within the optimal requirements of the cultured species. Finally, for the IMTA to work well, maintenance of water quality is of paramount importance and should be constantly monitored both spatially and temporally within the farm sites.

Recommendations
The goal of commercial aquaculture production is to produce great number of the produce of a given size within the shortest possible time. In this case, a variety of considerations such as, economic, environmental effects, and social benefits should be considered for effective mass ©2021 The Authors. Journal of Nature and Applied Research by NAT ARK www.natark.com | 52 culture and production of IMTA species. Based on the findings of this study, the following are recommended: 1) The use of green mussels which normally are not naturally obtained in Cebu, and has to be brought from the neighboring provinces (like Negros and Bohol areas) for this study, took so much time, effort and cost to transport, therefore the use of locally occurring filter feeders should be explored. This will not only address stress of the organisms that is usually associated with long transportation but acclimatization may no longer be necessary.
2) Corollary to the above, research on the population structure of the selected filter-feeding species (e.g. wild mussels) must be conducted to assess their potential genetic consequences of interbreeding between farmed and wild stocks. These studies should be done before the mass production of the wild stock to their new environment as part of the IMTA system.
3) The incorporation of modern technologies (e.g. wireless sensors) for real-time water quality monitoring in the aquaculture system is becoming a necessity and is highly recommended for a timely response in case of water quality disruptions in the aquaculture site.