Community-led restoration operates at the intersection of ecological feasibility and social acceptability. In the marine realm, restoration is challenging due to gaps in ecological knowledge on how and where to restore lost ecosystems and limited public engagement that provides social license for restoration. The restoration of lost oyster reefs provides a prime example because these ecosystems have been degraded to functional extinction on many coastlines, resulting in limited knowledge on their restoration potential and generational amnesia among communities that these ecosystems ever existed. To generate an evidence-based and social license for future restoration work, we engaged high school students and coastal residents in research on where to restore lost oyster reefs in South Australia’s iconic Coffin Bay. Using a mixed methods approach, we aimed to understand (1) the motivation of high school students to participate in restoration research, (2) to quantify ecological responses to habitat provision (oyster and biodiversity recruitment) to identify appropriate restoration sites, and (3) assess the response of residents’ to the ecological outcomes, including their willingness to support future restoration efforts. The high school students anticipated personal benefits (e.g., new experiences, career development), environmental benefits (e.g., nature connection), and benefited the local community (e.g., recreational activities). Students received SCUBA diving certification that enabled them to deploy 28 restoration units (shell baskets) at 8 sites throughout Coffin Bay. This experiment was retrieved after 3 months to reveal high-density recruitment of oysters and biodiversity at all sites, key environmental indicators for identifying suitable sites for restoration. Most residents engaged with the results expressed surprise in the ecological outcomes (the density of native oyster recruitment and associated diversity of marine life) and were very supportive of more oyster restoration occurring (91% of respondents). This study demonstrates that ecological feasibility and social licence are not sequential hurdles to be overcome independently, but mutually reinforcing processes that can be co-generated through community-based research. These results show that when restoration is designed as a socio-ecological learning system, rather than a technical intervention alone, it can unlock local stewardship, political momentum, and generate restoration-ready knowledge.

Study overview

This social-ecological research describes the initial phase of an ongoing program to restore native Ostrea angasi oyster reefs across South Australia through community-led restoration. In this paper, we focus on the program’s initiation in Coffin Bay, including project conceptualisation, study co-design, student engagement, ecological fieldwork, and community response. Our social analyses focus on student motivations to participate and residents’ responses to our ecological data because these provide insight into how people connect with local environments and how they perceive restoration (Fox & Cundill, 2018; Furness, 2021). Ecologically, our analyses focus on oyster recruitment dynamics and biodiversity enhancement because these provide key knowledge to inform the location and timing of restoration efforts (McAfee & Connell 2020). First, we provide a brief overview of the program’s development to date, followed by detailed methods for each phase.

This project was conceptualised through collaboration between Coffin Bay residents, local business owners, marine parks managers and m, and marine scientists. A marine ecologist and a local NT and SCUBA diving business owner (co-authors D. McAfee and M. Katz, respectively) coordinated the project’s development through a series of online and face-to-face meetings with key stakeholders. To enable knowledge sharing between the community and the project leads, these meetings included a catered community forum that approximately 80 residents attended (~16% of Coffin Bay’s residents), and targeted meetings with existing conservation groups (Lower Eyre Coastcare), oyster farmers, and State Government marine park managers. These meetings contributed to the co-design of the ecological field experiment, including site identification using local ecological knowledge. Meetings with the marine parks managers led to the inclusion of experimental sites within Sanctuary Zones to assess their influence on biodiversity recruitment (key knowledge for government managers). This collaborative approach ensured the research generated data relevant to both oyster restoration (e.g., spatial distribution of oyster recruitment) and local stakeholders, aligning with guidance for integrating ecological and social dimensions in marine restoration (Saunders et al. 2024).

Once the ecological experiment was designed and scientific research permits received (MR00234-1, D. McAfee), regional high school students were engaged with the project through scientific presentations at two local high schools. Following a recruitment and training process (described in section 2.3), 10 high school students worked alongside diving professionals to construct and deploy the field experiment across eight sites within Coffin Bay (Fig. 1).  After three months, researchers retrieved and processed the experiment with the assistance of the same high school students.  The ecological results were shared with the broader community through a poster displayed on local community notice boards and at a public nature festival, and were shared through social media. Community responses were then recorded via online surveys (described in section 2.5). Finally, outcomes from both the ecological and social data were presented to government stakeholders to inform the next phase of this restoration project, which is ongoing. All human data was collected following human ethics approval from the University of Adelaide (H-2021-140).

 Study location

The ecological research was conducted in the sheltered temperate waters of Coffin Bay, on South Australia’s Eyre Peninsula. The Eyre Peninsula is a large agricultural region with a sparse population (0.25 people km^-2), known for its coastal towns, recreational fishing, seafood industries, and mining. The waters of Coffin Bay are within the Thorny Passage Marine Park that includes multiple Sanctuary Zones that exclude all fishing activity (Fig. 1). The bay’s tidal regime includes mixed semidiurnal and diurnal patterns influenced by several narrow inlets that restrict water movement, resulting in slow flushing and variable water flow across the various bays (Kampf & Ellis 015).

Coffin Bay is part of the ancestral lands of the Nauo people, who fished these coastal waters for thousands of years before European settlement. Large oyster reef ecosystems formed by Ostrea angasi characterised the seafloor at the time of European settlement in South Australia. The first colonial settlement of Coffin Bay (early 1840s) was called “Oyster Town” due to the productive oyster dredge fishery it supported (Alleway & Connell 2015). Coffin Bay had some of the largest oyster reefs in South Australia, but decades of overharvesting drove the native oyster population to functional extinction, and the oyster dredge fishery collapsed within a century (Alleway & Connell 2015). The combined impacts of oyster dredging, which removed all the shell substrate that oysters need to recruit and regenerate reefs, and coastal land use change that resulted in sediment runoff that smothered the remaining shell, mean that there has been very little natural reef recovery since dredging ended (McAfee et al. 2022a). Coffin Bay has maintained its oyster industry through the highly successful aquaculture of introduced Pacific oysters (Magallana gigas) that continues to underpin the local economy. This historical context highlights the long-term socio-economic and ecological significance of oysters to the region, and the opportunity for community-engaged restoration of native oysters.

Recruitment and analysis of student motivation to participate

To engage local high school students in the project, 10 students (~15-17 years old) were recruited to participate following scientific presentations given by the project leads at their schools (Cummins Area High School, Port Lincoln High School). Presentations described the ecological role of oysters, current restoration efforts, and the project’s aims. The students were encouraged to apply to participate in the preparation and deployment of the field experiment, a process that included the students' SCUBA diving training and PADI Open Water Diver certification at no cost to the students (training was funded by an Environmental Citizen Science Small Grant from the South Australian Government, E. Katz). To apply, students were asked to submit a 1-page essay on why they would like to be involved in the project, which was submitted with parental consent. Ten students were accepted into the program and completed their diving certification before assisting researchers with the underwater experiment (Section 2.4).

Student motivations were analysed using inductive thematic coding to identify recurring motivational themes across all essays (Williams & Moser, 2019). Essays included multiple motivations that were categorised into three emergent themes related to (1) Personal Benefits, (2) Environmental Benefits, and (3) Community & Social Opportunities (Pichler et al. 2025). All motivations were captured; however, each motivational word or phrase was assigned to a single theme to avoid double-counting. We tallied the number of times each motivation was mentioned and used a word cloud generator to visually represent each motivation’s relative contribution (font size increasing with the frequency of use). This process provides insight into the varied and multifaceted reasons that motivate students to engage in restoration and helps to demonstrate how youth participation can contribute to generating social license for restoration initiatives (Bela et al. 2016; Kelly et al. 2019). While ocean literacy was not the primary focus, the process of students reflecting on their motivations to participate can contribute to a deeper understanding of marine ecosystems and human–ocean relationships (Jefferson et al. 2015; O'Brien et al. 2023).

 Ecological restoration experiment

Experimental design

To assess the feasibility and biodiversity benefits of restoring oyster reefs and identify priority sites for future restoration efforts, the ecological research was designed to assess where oyster larvae and marine biodiversity recruitment were greatest to oyster habitat “restoration units”. Additionally, to assess the influence of spatial closures aimed to protect marine biodiversity, we utilised four no fishing Sanctuary Zones within Coffin Bay, complimented by four adjacent Control sites outside sanctuary boundaries (n = 8 sites total; Fig. 1). Within each Sanctuary Zone and Control site, four and three restoration units were respectively positioned on sandy seafloor by the school students, equalling a total of n = 28 restoration units deployed across eight sites.

Each “restoration unit” consisted of 100 clean O. angasi oyster shells (shell height: ~10–15 cm) placed inside a plastic mesh crate (L×W×H: 40×20×15 cm, open top) that enabled water and small animals to freely move through the shell substrata. This shell layer within each restoration unit represents the non-living structure of oyster reefs that provide habitat for recruiting oyster larvae and associated organisms (McAfee et al. 2024). The dead oyster shells, which were collected from historical reef sites that had long been covered by sediment, were collected by the school students on SCUBA by gently excavating them by hand from the sediment surface. Any organisms present were removed by hand and s, and sand holes were cleaned by high-pressure hose before being dried and sterilised in the sun. Thereafter, shells were randomly assigned to the 28 restoration units*.* Each restoration unit had two vertically-attached settlement panels (15 × 15 cm concrete board; Williams et al. 2024) to provide a standardised unit with which to assess oyster recruitment and size among sites. To maximise the likelihood of natural oyster recruitment, restoration units were deployed during the known oyster recruitment season (January to April; McAfee & Connell 2020). All units were deployed on the same day and were collected after three months at the end of the recruitment season for immediate onshore processing.

Processing restoration units

 Immediately following collection, researchers and school students identified marine organisms within the restoration units to the lowest possible taxonomic level. Large and readily identifiable organisms were immediately released, while small invertebrates and cryptic species were preserved in 70% ethanol for later identification in the laboratory under a stereo microscope. Identified species were uploaded to the iNaturalist database to cross-reference their spatial distribution with other records (www.inaturalist.org; an open-access network of experts and citizen scientists recording and verifying biodiversity).

Oyster recruitment was quantified on both the er oyster settlement panels. Recruitment to oyster shells was immediately assessed for each restoration unit using magnifying glasses. In the laboratory, oyster recruits on the settlement panels were counted under disseca tion microscope and categorised by shell length (<2 mm, 2–5 mm, 5–10 mm, 10–15 mm, 15–20 mm, >20 mm; Fig. S1) as a proxy for settlement timing. For the settlement panels, oyster counts were standardised per m² and averaged between the two panels on each restoration unit to provide a single value per replicate unit (i.e., n = 4 per Sanctuary Site and n = 3 per Control site).

Statistical analysis of oyster recruitment and biodiversity

To assess factors influencing oyster recruitment and biodiversity, we used a combination of negative binomial generalised linear models (GLMs), Wilcoxon rank sum tests (for non-parametric comparisons where data did not meet assumptions of normality), and permutational multivariate analysis of variance (PERMANOVA). Due to the loss of two restoration units at the Eely Point Control site, both Eely Point sites were excluded from statistical analysis to avoid an unbalanced design, although Eely Point data were still visualised within figures to avoid loss of informative data. Therefore, GLMs were applied to evaluate the effects of Site (3 levels: Dutton Bay, Kellidie Bay, Yangie Bay) and Area (2 levels, fixed: Sanctuary vs. Control), with post-hoc comparisons to identify the direction of differences. All statistical analyses were conducted using R (v.4.3.1; R Core Team 2023) using packages: MASS for negative binomial models, vegan for NMDS and community structure, ggplot2 for visualisation, and emmeans for post-hoc pairwise comparisons.

Oyster recruitment, species richness, and organism abundance data (excluding oysters) were highly variable and violated the assumptions of ANOVA (normality and equal variance). Therefore, GLMs were used to account for overdispersion, confirmed by comparing the residual deviance to the degrees of freedom in initial Poisson models. Full models included Site (3 levels), Area (2 levels), and their interaction (Site × Area). Post-hoc pairwise comparisons were conducted with Tukey-adjusted p-values (emmeans package). Species diversity was further assessed using Shannon-Weiner and Simpson diversity indices that account for species richness, evenness, and dominance. Indices were calculated per basket for comparison between sites and protection areas using Wilcoxon rank-sum tests.

Differences in community composition among Sites, Areas, and their interaction were assessed using PERMANOVA via the adonis2() function, with pairwise post-hoc comparisons explored using pairwise.adonis(). Before analysis, species abundance data were square root-transformed and Wisconsin double-standardised to reduce the influence of dominant taxa. Differences were visualised using NMDS on Bray-Curtis dissimilarity, performed using the metaMDS() function in the vegan package in R, based on the full species-by-site matrix.

Public survey design and analysis

Following processing of the ecological data, Coffin Bay residents were engaged with the main ecological outcomes through face-to-face presentations by the project leaders and through information sheets distributed via social media platforms, local community group newsletters, and via posters positioned in public places (e.g., noticeboards). Accompanying the information sheets was a QR code for residents to access a short online survey designed to gather community responses to the project outcomes and people’s likelihood of supporting future restoration activities (Appendix 2). Because we were interested in perceptions of the restoration work and not public opinion on Sanctuary Zones, which are politically divisive in this region (Kirkman 2013), the ted data did not mention the use of the Sanctuary Zones. The survey was hosted by SurveyMonkey and was accessible for two months (October – December 2024), during which time n = 44 members of the public completed the survey (~7% of localents).

The survey consisted of four components that could be answered quickly (i.e., multiple-choice and Likert-scale responses), and an open-ended question to allow participants to elaborate on their thoughts (Appendix 2). Briefly, the survey consisted of questions related to: (1) demographic information, including participant age group and postcode to confirm residency; (2) which aspect of the data most surprised the participant (e.g., oyster recruitment outcomes; amount of biodiversity; no surprises); (3) likelihood of supporting larger restoration work (using a five-point Likert-scale approach); and (4) an open-ended question to enable additional comments.

Outcomes of the multiple-choice and Likert-scale responses were summarised using descriptive statistics (proportions, percentages). Open-ended responses were assessed to identify recurring ideas, concerns, or novel suggestions. Additionally, ANOVA was used to test whether age group or postcode location influenced responses to the question: “How likely would you be to support larger oyster reef restoration in Coffin Bay?”