As a first step, we visualised participants’ responses using bar plots and stacked bar plots to explore general response distributions and highlight patterns across questions. Then, to statistically examine how demographic characteristics influenced respondents’ understanding, use, experiences and perceived value of the marine environment around South Africa, we used either ordinal or logistic regression models depending on question type.

We used ordinal regressions to analyse questions where participants selected a single response from an ordered set of categories (e.g. Likert-type questions). This approach allowed us to test how demographic moderators influenced the probability of selecting higher versus lower response categories. For multiple-response (select-all-that-apply) questions, we recoded each response option as a binary variable (1 = selected, 0 = not selected) and fitted separate logistic regression models for each. This approach enabled us to examine how respondent characteristics influenced the likelihood of selecting each specific response option, treating them as independent binary outcomes.

The respondents’ demographic characteristics included in both ordinal and logistic regressions were all categorical, namely: gender (2 levels: male, female; responses from other gender identities represented ~1% of the sample and were excluded from regression analyses due to low sample size), whether the province of residence was coastal or not (2 levels: inland, coastal), age (4 levels: 18- 34, 35-54, 55-74, >75), educational attainment (3 levels: high school, graduate, postgraduate; respondents with no qualifications represented ~1.5% of the sample and were excluded from regression analysis due to low sample size), employment status (3 levels: students, not employed, employed), and whether the respondent or their family had any employment relationship with the marine environment (2 levels: yes, no).

All model assumptions were checked. For ordinal regressions, we tested the proportional odds assumption, and discarded models with a Hessian value >10,000, which is a sign of non-identifiable models. For logistic regressions, the assumptions of a binary outcome and independence of observations were met, as each model was fitted independently per response option. No multicollinearity was detected (assessed using variance inflation factors), and the linearity of the logit was not assessed as all predictors were categorical.

All analyses were conducted in R (R Development Core Team, 2024). Ordinal regressions were run using the ordinal package. All R scripts used in this analysis are publicly available at: https://github.com/jordipages-repo.

Thematic coding was carried out on the responses to the open question ‘what is the first thing that comes to mind when you think of the ocean’ to identify common themes and trends within the data. An emergent thematic coding approach was adopted to allow for all possible topics and themes to be highlighted in this process. Responses were reviewed thoroughly to identify an initial set of emergent themes; each response was re-reviewed to ensure the themes were all relevant and did not need to be combined into broader thematic categories. This process of content analysis highlighted a range of topics, which will be explored in more detail later in the paper.

Given the prominence of the OL principles in broader OL discourse over the last two decades, respondents were asked to indicate how true they felt each of the seven OL principles was, as an initial assessment of the dimensions of knowledge and awareness. As shown in Figure 2, there were varying levels of agreement with OL principles. The highest level of agreement was observed for Principle 5 ‘The ocean supports a great diversity of life and ecosystems’, with almost total agreement (99%). Five of the remaining principles also received relatively high levels of endorsement, though the proportion of respondents selecting “completely true” and “mostly true” varied. Principle 1 - ‘The Earth has one big ocean with many features’ - received the lowest level of agreement, with 55% selecting “completely true” and 23% selecting “mostly true”.

Ordinal regression analyses revealed that agreement with several OL principles was significantly associated with demographic and contextual characteristics, although the strength and direction of effects varied across statements. Respondents with employment links to the marine sector showed higher odds of agreement with four out of the seven OL principles (see Supplementary Table 1). For example, they were more likely to agree that “The ocean is a major influence on weather and climate” (OR = 2.20, p = 0.003) and that “The ocean supports a great diversity of life and ecosystems” (OR = 2.46, p = 0.031). Other significant associations included higher odds of agreeing that “The ocean and humans are inextricably linked” among postgraduate respondents (OR = 4.35, p < 0.001). In contrast, respondents with graduate or postgraduate education were less likely to agree with the statement “The Earth has one big ocean with many features” (OR = 0.54 and 0.38, respectively). Gender and age also emerged as significant predictors. Male respondents were less likely to agree that “The ocean and life in the ocean shape the features of Earth” (OR = 0.61, p = 0.009), while older respondents (55–75 and 75+) showed significantly lower odds of agreeing that “The ocean is largely unexplored” (ORs ranging from 0.27 to 0.38).

To further explore the dimension of knowledge, respondents were asked to rate their familiarity with a series of ocean-related terms. Again, as with the OL principles, there was clear variation in the responses (see Figure 2). Overall, however, respondents considered themselves to have a moderate to high level of knowledge for most of the terms, including some of the more technical ones, such as ecosystem services or nature-based solutions. Climate change was the most well-known and understood term with 98% of respondents indicating they had at least heard of and had some understanding of the term; followed closely by biodiversity, sustainable fishing and sustainable development all of which had over 90% of respondents indicating moderate to high levels of knowledge and understanding. While the total number of people indicating some level of familiarity with Marine Protected Areas was high (93%), only 63% indicated that they knew the term and understood its meaning. The terms with the lowest levels of familiarity and knowledge, i.e. those that respondents had never heard of (illustrated by the pale green bars in Figure 2B), included blue carbon (38%), eutrophication (33%), marine citizenship (27%), carbon sequestration (25%), ocean acidification (18%) and ocean literacy (15%).

Ordinal regression analyses revealed that higher educational attainment (mainly postgraduate level) and employment ties to the marine sector were consistently associated with significantly higher odds of knowing and understanding all of these terms (except climate change and nature-based solutions). For example, odds of understanding carbon sequestration and eutrophication were over twice as high for those connected to the marine sector (OR = 2.02, p < 0.001 and OR = 2.64, p < 0.001 respectively, Supplementary Table 1). Male respondents had higher odds of reporting understanding of eutrophication and ocean acidification, while respondents aged 35–54 and 55–74 showed lower odds of familiarity with climate change, eutrophication, biodiversity, ocean acidification, and ecosystem services, compared to the youngest age group.

Overall, 44% of respondents indicated that they thought the health of the marine environment in South Africa was either poor or very poor, with 30% perceiving it to be good or very good (Figure 3A). Over a quarter indicated some form of uncertainty, either selecting a neutral answer (24%) or stating that they didn’t know (3%). Postgraduates showed higher odds of rating the health of the marine environment around South Africa as good or very good, although this effect was marginally significant (OR = 1.67, p = 0.058).

The majority (65%) of respondents indicated they had a high level of awareness (i.e. good or very good) about the challenges facing the global marine environment (Figure 3B). Less than 15% indicated their awareness to be poor. Respondents with an employment relationship with the marine sector had higher odds of assigning high ratings to their level of awareness, while the reverse was true for inland respondents.

Overall concern for the marine environment and its conservation was very high, with almost every respondent indicating it was either important or very important to them personally (98% in total). Further analysis showed that responses to this question were moderated by employment status and distance to the coast, with the odds of considering marine protection as important being lower for students, but higher for people living in provinces nearer to the coast.

In order to understand respondent awareness of the everyday role of the ocean and their perceptions of the benefits from the marine environment, respondents were asked to indicate which benefits they considered to be the most important to South Africa society. Respondents indicated the most important benefit to be provision of diverse habitats for marine biodiversity (49%) (Figure 3C), which also reflects the respondent level of agreement with the ocean literacy principles (Figure 1A). The next most frequently selected benefits were food (45%), oxygen production and carbon storage (34%), weather and climate control (28%). While varying levels of importance for other benefits are also summarised in Figure 3, it is worth noting that three out of four least frequently selected benefits are linked to different cultural uses, including traditional medicine, a sense of identity, and sites for baptism, all selected by 1% of respondents, although tourism and a place that provides spiritual upliftment was selected by a higher percentage of respondents, 25% and 9% respectively.

Logistic regressions revealed key demographic differences in how marine benefits were perceived, with gender, age, and education emerging as the most consistent moderators (Supplementary Table 1). Compared to women, men were more likely to identify food provision and recreation as key benefits, but significantly less likely to recognise the ocean’s role in supporting mental health and wellbeing, contributing to research and education, or providing habitat for biodiversity. Older respondents (aged 55–74 and 75+) were more likely to select benefits related to biodiversity, weather and climate regulation, and natural forms of coastal protection.

In terms of awareness of perceived threats and challenges to the marine environment in South Africa, analysis found marine litter and plastic pollution to be selected as the most important threat by respondents (73%), followed closely by overfishing (72%) (Figure 3D). It is worth noting that lack of public understanding about the marine environment was selected by 59% of respondents, with concerns about a lack of political will (50%) and poor management (51%) also highlighted. Aspects relating to climate change and biodiversity loss (i.e. loss of marine species) were lower in the list of priorities, selected by 48% and 43% respectively.

Logistic regression models revealed that demographic characteristics significantly shaped perceptions of threats to the marine environment (Supplementary Table 1). Postgraduates were more likely to recognise scientific issues such as ocean acidification, habitat destruction, species loss or invasives, while unemployed were less likely to identify changes to ocean currents, coastal erosion and sea level rise as threats. Respondents with marine-related employment experience were more likely to identify changing ocean currents and poor management as a threat. Gender differences also emerged: men were less likely than women to perceive coastal and marine habitat destruction as a major threat.

When specifically asked about their views on climate change, 49% of respondents were of the view that climate change is driven by human activity, while 46% agreed that climate change is caused by a combination of both human activity and other natural processes (See Supplementary Figure 1). Furthermore, when asked about several statements about climate change, 89% respondents considered that climate action was urgently needed as true or mostly true, 73% considered true or mostly true that nature can help reduce climate change impacts, while only 23% considered true or mostly true that climate change is not the greatest threat to our natural world.

Ordinal regression models showed that these views were moderated by some demographic variables. Respondents with postgraduate education and those currently studying were less likely to agree with the statement that climate change is not the greatest threat to our natural world. Postgraduates also had higher odds of agreeing that climate action is urgently needed. In addition, respondents with employment ties to the marine sector were more likely to agree that nature can help to reduce climate change impacts, while unemployed respondents had lower odds of agreeing with this statement.

In an initial attempt to explore connection, an open question asked respondents to indicate the first word that comes to mind when they think of the ocean to provide insight into emotional responses and overall emoceans (McKinley et al., 2023), when considering the ocean (summarised in Figure 4).

While the thematic coding process allowed for the identification of key themes emerging from the data, it is interesting to note that several responses covered multiple themes, illustrating the multi-faceted and layered relationships held by respondents towards the ocean. Table 2 presents an overview of the top 10 key themes highlighted through the content analysis process, including examples of how these were referred to by respondents.

Emotional connections were further explored through questions which asked respondents to select the first three emotions that came to mind when they thought of the marine environment (Figure 5). Four of the five most frequently selected emotions that the ocean transmits to respondents were positive, while 7 of the 10 least chosen emotions were negative ones, highlighting a positive view of respondents towards the ocean. Specifically, the most frequently selected emotion was awe/wonder, a positive emotion chosen by 51% of respondents. Reflecting the concern indicated in other questions, concern, as a negative emotion, was the second most frequently selected option (39%) followed closely by curiosity and calm/relaxed (both selected by 34%). The emotions of happiness and excitement were positive until the 7th in the list, anxiety, with 15% respondents choosing it. Shame (6%), hopelessness (5%) and surprise (2%) were among the least frequently selected emotional responses. Other emotions highlighted by respondents in the open ‘other’ box also included gratitude, disappointment, fear of the deep ocean, but also a recognition of the complexity of human-ocean relationships with some stating that they experience ‘mixed emotions excited, worried about the future of our planet and encouraged’, while others expressed concern and anger about ocean issues, with one stating “I feel an unmitigated rage that the exquisite marine life and the coral reefs will be harmed or destroyed by oil and gas extraction and exploration. I think the ocean holds the key to the preservation of all life on the planet at this precarious juncture of the threat of global warming.”.

Logistic regression models revealed that several emotional responses to the ocean were significantly associated with demographic characteristics (Supplementary Table 1). Respondents aged 55–74 had higher odds of selecting “anxiety” and “awe/wonder”, but lower odds of selecting “curiosity” and “inspired”. Inland residents were more likely than coastal residents to report feeling “calm/relaxed” when thinking about the ocean. Educational level and employment status also showed associations: graduates were less likely to report “calm/relaxed”, and students showed higher odds of selecting “curiosity” and lower odds of selecting “inspired”. In terms of gender differences, male respondents were more likely than female respondents to select “excitement” and “frustration”, and less likely to select “inspired”.

When asked about how they access and experience the marine environment, most respondents stated they had visited the marine environment in the previous 12 months (87%) (Figure 6A). More specifically, when asked about the places people tend to visit, analysis found coastal or seaside towns to be more commonly visited daily (29%), followed by sandy beaches (4%) and estuaries (4%). Overall, coastal towns, rocky shores and sandy beaches were the types of marine environments visited most frequently. However, respondent characteristics influenced the frequency of visits to the different marine environments. Respondents living in inland provinces had significantly lower odds of frequently visiting all types of marine environments. In contrast, respondents who were not employed had higher odds of visiting sandy beaches, rocky shores, estuaries and coastal/seaside towns than employed ones. Males had higher odds of visiting the open sea more frequently, while older age groups were less likely to visit the open sea frequently. Respondents with ties to the marine sector were more likely to frequently visit Marine Protected Areas (see Supplementary Table 1).

When asked what sort of recreational activities they participate in, the most frequently selected options included sea swimming (52%), walking with/without a dog (78%), beach cleaning (43%), and wildlife/bird watching (42%). As shown in Figures 6B, there was clearly a broad spread of activities being carried out by respondents. Participation in specific marine recreational and cultural activities varied significantly with demographic and contextual characteristics (Supplementary Table 1). Older age groups showed strong negative associations with scuba diving, hiking, beach cleaning and marine-themed attractions. Living inland, as opposed to coastal areas, was negatively associated with most marine activities, including beach cleaning, walking with a dog, rock pooling, sea swimming, and running near the coast. High levels of education were associated with higher odds of selecting snorkelling, running, walking with or without a dog, and hiking. Male respondents had substantially higher odds of reporting sea fishing, surfing/body boarding, and scuba diving, and were less likely to report beach cleaning or wildlife watching. Finally, respondents with a personal or family employment link to the marine sector were more likely to select scuba diving, snorkelling, and photography/videography at the coast.

In addition to understanding how people use the marine environment in South Africa, questions were also posed to respondents to assess the barriers to access experienced (Figure 6C). For just under 25%, no barriers were experienced over the previous 12 months (indicated by those who selected Not applicable); however, it is important to note that barriers to access and experience were noted by most respondents (see Figure 6C). The most frequently noted barrier was being too busy at work (20%) and living too far away from accessible marine spaces (18%). Cost was recognised by 15% of respondents as a barrier to their ability to access these places. Several demographic and contextual factors were associated with barriers to visiting the marine environment, according to logistic regressions (Supplementary Table 1). Inland respondents had significantly higher odds of citing distance-related barriers, including “takes too long to get there/live too far away” and “cost/too expensive”. Age was also associated with certain constraints: respondents aged 35–54 were less likely to report being “too busy at work” or concerned about crime, and those aged 55–74 were less likely to cite “bad/poor weather” as a barrier. Respondents who were not employed were less likely to report being “too busy at work”. Finally, those with a connection to the marine sector were less likely to cite distance (“takes too long to get there/live too far away”) as a barrier.

Reasons for visiting the marine environment were found to be varied (Figure 6D). Overall, the vast majority of respondents agreed or strongly agreed with all the reasons to visit the marine environment, except for the ocean being important for them culturally (with < 50% of agreement). The most agreed upon reasons for visiting the ocean were its beauty, feeling connected to nature, having had memorable experiences there, and several reasons related to the ocean improving respondents’ wellbeing, such as visiting the marine environment being good for their spirit, clearing their head, feeling healthier, freer, among others.

Interestingly, the most frequent respondent characteristic moderating the agreement to the reasons for visiting the marine environment, was education attainment, followed by gender, age, and having an employment relationship with the marine sector in South Africa (See Supplementary Table 1). Specifically, respondents with higher qualifications (i.e. graduates and postgraduates) had higher odds of agreeing with several statements, mostly related to the wellbeing benefits of visiting the ocean. In addition, being a male decreased the odds of agreeing with several of the reasons for visiting the ocean, such as most of the statements related to the wellbeing and psychological restoration benefits of visiting the ocean, as well as those related to nature connection, feeling part of something greater than themselves, several linked to memories, and other deep feelings in general. Finally, having an employment relationship with the marine environment increased the odds of agreeing with statements related to past experiences in the ocean and feeling they can contribute to improving the marine environment. Contrary to expectation, respondents from inland provinces had higher odds of agreeing with a feeling they can contribute to taking care of the marine environment.

Communication has long been recognised as a central tenet of ocean literacy, and remains one of the 10 dimensions proposed by McKinley et al., 2023. When asked where they get information about the marine environment from (Figure 7A), the most frequently selected option was films and nature/wildlife documentaries (49%), followed closely by social media (39%), and then different forms of news materials (33%). Public events and places were also selected by a number (e.g. 28% of respondents selected aquariums/zoos and visitor centres). Despite OL emerging from formal education, only 19% of respondents indicated that this is where they got their information, while government agencies were among the least frequently selected (4%).

The logistic regressions identified strong demographic differences in the perceived sources of information about the marine environment. Education level had the most consistent influence: postgraduates were more likely to cite formal sources like scientists, seminars, and news, but less likely to cite social media, television, documentaries or community events as their source of information. Older respondents were less likely to report formal education and social media, instead favouring traditional sources such as TV and print media. Respondents with coastal ties were more likely to use institutional sources, including scientists and government agencies. Gender differences also emerged, with males more likely to cite scientists and government agencies, and less likely to report aquariums and museums.

In terms of the sources of information most trusted by respondents, it is interesting to note that these sources differed from those most frequently used (See Figure 7B). Scientists were found to be most trusted by respondents (62%), followed by aquariums/zoos/museums (46%) and then films and wildlife/Nature documentaries (41%). Despite being the second most common source of information, social media was only selected by 4% as being among their most trusted sources. It is also of note that government agencies were chosen by only 4% of respondents. Respondents’ education, age, and connection through employment shaped their trust towards the information sources used (Supplementary Table 1). Specifically, postgraduates showed strong preference for formal sources such as scientists, seminars, and education institutions, while being less trusting of informal sources like TV/radio, or nature documentaries. Trust in formal education declined significantly with age, while older respondents increasingly favoured campaigns and books/magazines. Those with coastal ties showed greater trust in seminars and scientists, but less in TV/radio, news and books/magazines.

With a growing focus on how ocean literacy can lead to improved behaviour change and action, respondents were asked to indicate how much impact they think their lifestyle has on the marine environment. Interestingly, 59% of respondents indicated that they thought they had a positive impact on the marine environment, 26% indicated a negative impact and only 15% felt that they had no impact at all. While it is not possible to ascertain what kind of impacts respondents were referring to when they indicated that they had a positive impact, the number respondents indicating no impact at all indicates lower levels of awareness and understanding of the interconnections and impacts between society and the ocean on an everyday level than would be hoped for. Among respondent characteristics, education was the only significant moderator of self-attributed impact on the marine environment. Specifically, respondents with postgraduate education had lower odds of perceiving their lifestyle as having a positive impact on the marine environment, compared to those with lower levels of education.

In order to understand the behaviours and actions that people already undertake, a series of questions were posed to respondents to explore this dimension of ocean literacy. The majority of respondents were found to have already made changes to their lifestyle to benefit the marine environment (59%) (Figure 8A). Of those who reported that they had already made changes, the most commonly undertaken activities included signing petitions about marine issues (53%), volunteering time to participate in beach clean-ups and other activities (52%) and trying to encourage others to do more to protect the marine environment (44%). The less common activities included attending protests or demonstrations, with 10% indicating that this was something they had done, while 12% indicated they had donated money.

Participation in marine-related activism varied significantly across demographic groups (SM Excel). Respondents aged 35–54 and 55–74 were consistently less likely than younger respondents to report engaging in behaviours such as persuading others, posting marine content online, participating in citizen science, or joining environmental organisations. Inland residents were significantly less likely to volunteer time (e.g. beach clean-ups) compared to coastal residents. In contrast, those with a personal or family connection to the marine sector showed markedly higher odds of engaging in a wide range of pro-environmental actions, including volunteering, citizen science, posting content, and persuading others, and were less likely to report doing “none of the above”. Postgraduate respondents were more likely to participate in science and research projects, and students were more likely to report doing “none of the above”. Gender also played a role: male respondents were more likely to report participation in citizen science, but less likely to report signing petitions or making lifestyle changes to benefit the marine environment.

With individual purchasing behaviour commonly identified as a mechanism for expressing ocean literacy and marine citizenship (Nuojua et al., 2024; Garcia-Vazquez et al., 2022), questions were included which sought to explore patterns in purchasing behaviour among respondents. As shown in Figures 8B, the majority of respondents had made some changes to their purchasing and consumption behaviours, with reduction in use of single use plastic (85%), re-using plastic (74%), increased recycling (69%) and reduced use of disposable products, such as plates or cutlery (62%) commonly noted, although it is worth noting that every option was selected by at least 40% of respondents.

Logistic regressions showed that males were less likely to report 5 out of 7 of pro-environmental behaviours than women (Supplementary Table 1). In contrast, postgraduates were more likely to report having recycled more, reduced single use plastics and avoided excessive packaging in the last 12 months compared to those with lower education levels. Relating to this, there were also some questions posed relating to respondents’ views on ocean governance and management in South Africa, relating to the dimensions of activism and ideas of where responsibility for management lies. When asked their views on the most important activities addressing issues facing South Africa’s marine environment, tightening controls over commercial activities, such as oil/gas exploration was identified most frequently by respondents (56%). This was closely followed by tightening controls to reduce pollution from industry/agriculture (47%), restricting and regulating fisheries activities (47%) and expansion of Marine Protected Areas (36%) (Figure 8C). Building human-ocean relationships and supporting more people to engage with the ocean was also highlighted by 30%. Support for specific marine conservation actions varied across demographic groups (Supplementary Table 1). Male respondents showed higher odds of supporting restrictions on fisheries and commitments to safeguard habitats but were less likely to support expanding ocean engagement through education. Respondents with employment links to the marine sector had higher odds of supporting restrictions on fisheries and behavioural change to reduce marine pollution. Respondents aged 55–74 were less likely than younger groups to support increased protection through Marine Protected Areas.

In terms of the actions that the South African government should prioritise for the marine environment (Figure 8D), respondents most frequently selected better enforcement of existing regulations (61%), designation and implementation of more MPAs (50%), and more effective management of marine pollution (44%). Of lower priority to respondents were items relating to blue economy sectors, including better use of oil and gas resources (9%), more equitable allocation of quotas (7%), and, of least importance, overall expansion of the blue economy (6%). Support for specific government actions to manage the marine environment differed by age, education, and connection to the marine sector (Supplementary Table 1). Respondents aged 35–54 were less likely to support the creation of more Marine Protected Areas, while those aged 55–74 were less likely to support stricter fishing regulations, but more likely to support better protection of South Africa’s fish resources. Postgraduate respondents showed lower odds of supporting stricter fishing regulations, and respondents with a personal or family link to the marine sector were less likely to select better management of pollution.

Finally, in relation to the statement above regarding designation of more MPAs, and with protecting marine areas high on the broader conservation and management agenda, respondents were asked to indicate their level of support for the creation of MPAs in South Africa. Analysis found there to be almost complete support (97%). Among respondent characteristics, type of employment was the only significant moderator of this question, with students having lower odds of strongly supporting MPAs in South Africa, compared to unemployed or employed respondents.