2.1. Site selection
The oyster production sites (i.e., oyster producers or buyers for further processing) used in this study were located on a sheltered bay in the north-west of the Republic of Ireland. Samples from two Pacific oyster-producing sites (i.e., Site 1 and Site 2) were collected for morphological and elemental analysis, while the LCAs were modelled using operational data from three Pacific oyster farms located along the West coast (i.e., Site 1, Site 3 and Site 4), Site 1 providing samples for the morphological and elemental analysis (Figure 1).
2.2. Ecosystem services methodology
2.2.1. Morphological and elemental analysis
For morphological and elemental analysis, farmers at each production site randomly harvested 15 individuals per market size category (i.e., small, 67.4-112.5 cm length; medium, 89.4-119.3 cm length; and large, 94.8-120.7 cm length) during different times of the winter season (i.e., February and March). Thus, there were a total of 45 samples per site. The following morphometric measurements were undertaken per oyster:
1. Total shell length, width, and depth; mm oyster-1;
2. Total wet weight; g oyster-1;
3. Shell wet weight and tissue wet weight; g oyster-1,
4. Dry tissue and shell weights; g oyster-1.
Among the 15 individuals morphologically assessed per site and size category, sets of 6 individuals were randomly selected and pooled for the elemental analysis [30]. Thus, for each site, 18 individuals were selected for elemental analysis. Tissue and shells from pooled individuals were dried in a fan assisted oven at 80 °C until a constant weight is achieved. Dried tissue and shells from pooled individuals were crushed using a mortar and pestle for dried tissue and a mill for dried shells. The pooled tissue and shell samples were then analysed for N and C content through an elemental CHN analyser (Flash smart elemental analyser, Thermo Fisher, Waltham, Massachusetts, United States). P content was measured through Inductively Coupled Plasma Optical Emission Spectrometry (700 series ICP-OES, Agilent, Santa Clara, California, United States). The results obtained as %C, %N, and %P in the dried tissue and shell samples were used to calculate: (i) the average %C, %N and %P per individual oyster (and separately the tissue and shell for each oyster), size category and site investigated; (ii) the average mass of C, N, and P removed per fresh individual oyster; and (iii) the average mass of C, N, and P removed per tonne of oysters harvested. Differences in the elemental analysis (i.e., %C, %N, %P) of Pacific oyster between the three size classes and the sites investigated were analysed using two-way ANOVA tests. A post-hoc Tukey's test was conducted on each dataset to discern significant differences between sizes and sites. Statistical significance was assigned when P<0.05. Limitations of the approaches used are discussed in Section 4.1.
2.2.2. National ecological impact
The morphological and elemental analysis results were then extrapolated to farm and national scale to obtain: a) the quantities of nutrients and carbon removed annually on each production site using the average annual production for the period 2015-2020 i.e., average annual production (tonne year -1) x N, P or C removed per tonne of fresh product (kg tonne -1); and b) national extrapolation of nutrients and carbon removed using the most recent estimated total annual production of Irish Pacific oysters [31], i.e., N, P or C removed per tonne of fresh product (kg tonne -1) x national production of Pacific oysters (tonne year-1).
An ecosystem services analysis of Pacific oyster farming was carried out to associate a monetary value with the nutrient remediation potential. This valuation of nutrient removal ES was calculated using the following median values for the removal of N (€18.9 kg-1) and P (€33.9 kg-1) [25]. These monetary values represent the theoretical cost of upgrading a wastewater treatment plant to remove one kg of N and P. Obviously, such values can vary between treatment plants depending on existing load, plant technology, discharge limits, plant size, etc. Nutrient valuation was also extrapolated nationally by applying the national production of Pacific oyster for 2022 [31]. Results were also equated to wastewater treatment plant performance in terms of population equivalent for N removal. A wastewater treatment plant, with secondary treatment, was estimated to remove, on average, 3.3 kg N person-1 year -1 [21]. This figure was applied to calculate the population equivalent where a wastewater treatment plant would remove the amount of N remediated (extrapolated as per the above) by pacific oyster farming in Ireland.
2.3. Life cycle assessment methodology
2.3.1. Goal and scope
LCA studies were undertaken on three Pacific oyster sites along Ireland’s West coast. A cradle-to-gate system boundary was used for the farming and on-site processing activities at each site. The systems boundaries included aquaculture infrastructure, seed procurement, consumable materials, energy production (electric and diesel), culture and harvesting, processing, and packaging. Waste management and treatment of waste materials and packaging are also included within the system boundaries. The functional units applied were one tonne of live oyster product (meat and shell), i.e., farm-to-gate. Each studied site produced, on average, 111 tonnes of oysters for the market annually. All sites used bags and trestles to grow their oysters, and oyster seed was purchased domestically (Figure 2).
2.3.2. Life cycle inventory
The life cycle inventory used primary data from the partner farms. Primary data was collected through questionnaires, interviews, and site visits. Energy, fuel, and consumables values were validated against bills and invoices where possible. Secondary data was collected from established life cycle databases such as Ecoinvent v3.10, Agri-footprint 6.3, and Agribalyse 3.0.1 to populate the life cycle inventories. The life cycle inventory of the present study covers all farm-based activities, infrastructures, and use of resources (Supplementary Table 5). The main transport vehicles used for daily farming activities at each site were a fleet of tractors and trailers. The trestles at each site were manufactured from 25 mm reinforced steel bars and weighed 18 kg per segment. The service life of the trestles was estimated to be 15 years. Oyster bags were made of high-density polyethene and weighed approximately 800 g per bag, with an average service life of 8 years.
2.3.3. Life cycle assessment
The life cycle impact assessment methodology was undertaken through the CML method [32] (Guinée, 2002). The following impact assessment categories were included: 100-year global warming potential (GWP, kg CO2 eq.), Acidification potential (AP, kg SO2 eq.), Eutrophication potential (EP, kg PO4 eq.), and Cumulative energy demand (CED,MJ), which assesses the degree of energy consumption associated with a production system [33]. These impact categories are the most concerning for aquaculture and shellfish production systems, as cited in many studies [34,35,36].
2.4. Life cycle assessment and ecosystem services
In this study, the elemental analysis results (i.e., N, P and C content in Pacific oyster shells) were adapted to LCA impact categories to estimate the ES provided by Pacific oyster aquaculture in Ireland. N and P content in the shell were converted to PO4 eq., a compatible form under the EP impact category. Characterisation factors of 0.42 and 3.07 were applied to convert N and P to PO4 eq., respectively [37]. To determine the net GWP of Pacific oyster farming, C content in oyster shells (i.e., amount of CO2 sequestered in the shell during biocalcification) was converted to CO2 eq [28]. The N, C, and P contained in the soft tissue were not included within the ES calculations as they are considered a short stage of the biogenic carbon cycle. On the contrary, shells can sequester nutrients for extended periods [13,15].