Assessment of Multiple Citizen Science Methods and Carbon Footprint of Tourists in Two Australian Marine Parks

1. Introduction

Citizen science (CS) is the engagement of the public as volunteers in a scientific project alongside scientists. Such collaborations can improve the quantity and reach of data collections, at little additional cost to the project. These projects can range from small-scale research focused on a single species [1,2] to large-scale research of multiple species with both spatial and temporal components [3,4] These potential benefits can also integrate more broadly into the social sciences [5]. Citizen science projects that deliver high data density per location, wide geographic coverage, or regular observations over time, is considered particularly valuable [6]. Citizens can participate in every stage of the project, from problem identification, experimental design, data collection, data analysis and dissemination of the results [7,8,9]. One of the important aspects of CS is that it engages members of the public in science, and with the places and issues the science is designed to support. Given the current stresses faced by coral reefs worldwide, the benefits of citizen science are more important than ever before in tropical ecosystems.

The number of CS projects has increased worldwide over the last decade as public interest in the environment grows and technology allows more people to engage meaningfully in data collection, particularly in the natural environment [10]. As a result, science and management organisations, tourism companies, and expedition charters have increasingly started to incorporate citizen science into their activities [10,11,12]. However, marine CS participation has been lower than that for terrestrial systems [13,14], likely because of the challenges associated with field logistics, accessibility, equipment (or lack thereof), training, safety, and culture [8,15].

There are currently over 134 CS programs operating in Queensland, Australia, which are either run by management authorities, tourism operators, researchers, dive enthusiasts, or members of the general public [15,16,17,18,19]. An online survey of 1,145 marine end users reported that most citizen scientists tended to have a degree in science, were under 45 years old, enjoyed SCUBA diving, and had contributed to scientific research in the past [10].

It is standard practice to focus on one CS methodology per project [15,20], however, an increasing number of projects are employing multiple methods to enable a broader perspective and integrate these data sources [21]. Combining multiple CS tools not only increases the scope of data collected but can improve the probability of overall success in case one or more methods falls short of expectations. This is particularly valuable when data collection sites are remote and difficult to access.

Great Barrier Reef and Coral Sea

The Great Barrier Reef is the world's largest living organism, protected within the 344,400 km² Great Barrier Reef Marine Park (GBRMP), set along the north-eastern coast of Australia. Directly offshore and adjacent to the GBRMP is the Coral Sea Marine Park (CSMP), covering an even greater area of 989,836 km², making it the largest of Australia’s marine parks [22]. Both marine parks are considered biodiversity hotspots [22,23] and managed by State and Federal government agencies in Australia. Although the GBRMP has attracted significant research effort and funding over the last century, research in the CSMP has been comparatively limited, due to its remote location, situated at least 200 km from the Australian mainland [24].

The tourism company Coral Expeditions and the media company Australian Geographic have developed a series of annual expeditions with a program of scientific, cultural, historic, and nature-based activities facilitated by expert staff and guest lecturers [25]. The 2023 expedition promoted as the ‘Citizen science of the Great Barrier Reef expedition’ [26] the subject of this paper, had five aims: (1). Coral Expeditions passengers and staff learn about reef species, management and science (and underwater cultural heritage), (2). Coral Expeditions passengers and staff are scientific partners, actively involved in scientific endeavors to generate new knowledge or understanding about the Great Barrier Reef and Coral Sea, (3). CS projects deliver a genuine scientific outcome, including expanding our knowledge of species distribution and abundance, and disseminating that knowledge through reports and scientific papers, (4). Communication of the expedition to the general public and the scientific community through diverse channels including art, video, web, news and social media, and (5) Fundraising to support a carbon neutral expedition.

2. Materials and Methods

The research and ecotourism vessel Coral Discoverer departed Cairns on 23rd October 2023 and returned 14 days later on 6th November 2023. Six guest lecturers provided training in multiple CS methods to 40 guests. Guest lecturers and guests collected CS data during visits to 12 different islands, cays and reefs in the Great Barrier Reef Marine Park (9 locations) and Coral Sea Marine Park (3 locations) (Figure 1).

A pre-expedition research plan with daily communication (lectures, newsletter, workshop, meetings, art) and field activities focused on four types of CS tools: biodiversity audits, reef health checks, heritage assessments, and social surveys, including ten specific CS tools [27] that were made available based on guest training, location, experience and interest (Table 1). We included an activity for artists to be involved in an expedition sketchbook to complement knowledge sharing, communication and learning.

2.1. Biodiversity Audit Tools

2.1.1. iNaturalist

Initiated in 2011, iNaturalist (www.inaturalist.org) is a multi-taxa platform that allows participants to contribute observations of any organism or traces thereof, along with associated spatiotemporal metadata to a centralized website. Observations are initially identified by the user or with the assistance of artificial intelligence computer vision suggestions and then identified and verified to high taxonomic resolution by the iNaturalist community. An observation is deemed “Research Grade” when it meets the site’s metadata quality criteria, and has two or more suggested identifications, more than two-thirds of which agree at a species level [28].

An iNaturalist project ‘Citizen science of the Great Barrier Reef and Coral Sea’ [29] was created with a focus on marine biota and spatial and temporal boundaries. Observers were restricted to the citizen scientists on the expedition. Identifiers were not restricted. T-tests were used to compare data between the GBRMP and CSMP.

2.1.2. eDNA

Water samples were collected at four sites (N = 4 replicates at each) in the GBRMP and CSMP between October 29, 2023, and November 4, 2023, to provide a complementary whole ecosystem biodiversity detections data set. For each water replicate, a sterile eDNA syringe mini kit containing a 30 mm x 1.2 μm cellulose acetate syringe filter with luer-lock inlet and outlet fitting was used (Wilderlab NZ Ltd., Wellington, New Zealand). The target water filtration volume was 2.5 L. Post-filtration, syringe filters were preserved with 350 μl of DNA/RNA Shield (Zymo Research Cat. No. R1200-125). Protocols followed [30,31,32].

2.1.3. Marine Vegetation Collection

Seagrass and algal specimens were targeted for collection to assist in the creation of a marine vegetation library for both marine reserves. Marine vegetation specimens vouchered in state of federal herbariums are severely lacking for the GBRMP and absent for the CSMP. These specimens will provide visual libraries, as well as a source for genetic tissues if requested by other researchers. All specimens collected were pat dried with a paper towel, before being placed in an herbarium press between multiple pieces of newspaper. Newspapers were replaced daily to ensure specimens did not stick to the paper, until the specimen was completely dry. Data on location, date, species, and collector's name were recorded in an herbarium data book, along with descriptions of where the specimen was collected as well as how many specimens.

2.2. Reef Health Check Tools

2.2.1. CoralWatch

Initiated in 2002 as a scientific research effort, CoralWatch is a global CS program that integrates education and global reef monitoring to examine coral bleaching. CoralWatch data has been collected in over 80 countries, from 2,245 different reefs. The main tool that CoralWatch participants use is the Coral Health Chart, which records changes in coral colour and represents a simple tool for citizens to monitor coral health. Records of coral colour changes over time provide data on coral health changes at the individual reef scale (Figure 2; [33]). We used the online CoralWatch analysis tool (https://coralwatch.org/monitoring/analyse-your-data/) to summarise our survey data. These data were visualised using a bar graph for coral colour scores, where a healthy reef generally scores 3 or more. We used a pie chart to show the percentages of the four coral types (Boulder, Branching, Plate, Soft), and a t-test was used to compare the data between the GBRMP and CSMP. Statistics were run in R version 4.3.1, with a measure of significance when p < 0.05.

2.2.2. Great Barrier Reef Census

Initiated in 2020, the Great Reef Census is an annual CS effort to survey the Great Barrier Reef between 1 October and 31 January, and so our expedition fell within this sampling window. The Great Reef Census collects tens of thousands of photos from hundreds of reefs across the GBR. An individual survey is made of up to 40 representative photos from a reef site, with each image capturing at least 5×5 m of a reefscape edge [34]). These images are uploaded to an online platform, where people from across the world identify the types of coral and the extent of coral cover per site [34].

2.2.3. Eye on the Reef

Initiated in 2007 by the Great Barrier Reef Marine Park Authority, the Eye on the Reef program aims to provide a multi-tiered approach to reef health monitoring and assessment. It is designed for people with little to moderate reef experience who can either snorkel confidently or SCUBA dive. The tiers we used were embedded within the Reef Health Impact Survey (RHIS), which assesses reef health in a series of five metre radius circles (a total of 78.5 m²), and the Rapid Monitoring Survey, which involves a 10 minute timed swim and count of 18 categories of animals: Sea cucumber, Giant clam, Anemonefish, Butterflyfish, Grazing herbivores, Cods and groupers, Coral trout (x2), Maori wrasse (x2), Turtle (x3), Shark (x3) and Crown-of-thorns starfish (x2) [20]. A GLM, with a negative binomial distribution to adjust for over-dispersion, was used to compare the data between the GBRMP and CSMP. Statistics were run in R version 4.3.1, with a measure of significance when p < 0.05.

2.2.4. Redmap

Initiated in 2009, Redmap (Range Extension Database & Mapping Project) is an online tool that invites Australians to share sightings of marine species that are uncommon or rare to their local area. The data collected aims to highlight areas and species that may be experiencing range expansion owing to environmental change so that future research may be focused in these areas. In Queensland waters there are currently 15 species of fish, four mammals, one reptile, and five sharks and rays that are the focus of this research, although additional species may be logged and reviewed [35].

2.2.5. Marine Debris

The Australian Marine Debris Initiative (AMDI) is run by the Tangaroa Blue Foundation and provides a scalable, collaborative framework to address marine debris impacting Australia from both domestic and international sources. Organised beach clean-ups and litter surveys are popular with volunteers. Community clean-up surveys provide data on the amount and different types of debris found, as well as hotspots where it is accumulating. Our intention in the use of this tool was to determine if there might be hotpots in the GBRMP and CSMP where debris was notable or accumulating.

2.3. Heritage

2.3.1. Photogrammetry

Photogrammetry involves taking overlapping photographs of an object, structure, or space and converting them into 2D or 3D digital models [36] From these models, accurate measurements of an object, reef or shipwreck can be made. Our intention in its use was to apply this tool to a historical shipwreck on a CSMP reef.

2.4. Social Survey

2.4.1. Questionnaire

Social surveys were conducted on four occasions (twice in the GBRMP and twice in the CSMP) between 29 October and 5 November 2023 by intercept interviews or by a self-administered online questionnaire. Survey metrics were categorised into four broad themes: knowledge of the reef, health of the reef, satisfaction with the trip, and CS. A total of 16 questions (Appendix D) were provided and included tick box and words for simplicity, as well as five- to ten-point Likert-type scales for statistical rigor. The surveys were completed by participants within 3 to 5 minutes. We used a multiple linear regression model to test if the perceived health of the reef and information received from CS significantly affected the satisfaction level of visitors to the CSMP and GBRMP. The model used was specified by Y = β0 + β1X1 + β2X2, where X1 is the health and X2 is the information received. This analysis of survey data aimed to provide an understanding of how these variables affected satisfaction levels. Statistical analysis was performed in R version 4.3.1, with a measure of significance when p < 0.05.

2.5. Carbon Footprint

An integrated tool to calculate plane, vessel, and car travel for the complete footprint of the expedition was not available. We instead included seven separate inputs for fuel use and greenhouse gas calculations: large vessel, medium vessel, small vessel, generator, international flights, domestic flights, and car travel. We measured distance and fuel use for the large vessel, and estimated fuel use for medium and smaller vessels, using a multiplication factor of 2.68 [37] to convert to greenhouse gas tonnes. We estimated the number of international and domestic passengers as well the staff on board and applied an average for flight times to and from Cairns. We noted that established integrated tools assume ¼ tonne CO₂ equivalent per passenger per hour flying [38]. We therefore assumed an average of two hours for domestic staff and passengers, as well as 25 hours for international passengers. We assumed one tonne of greenhouse gas emissions for every 1,000 km of car travel. We noted that the average cost of an eligible carbon offset in Australia was $25 AUD per tonne of CO₂ abated.

3. Results

A summary of the dates and locations visited during the expedition as well as the type and number of CS data collected is provided in Table 2.

3.1. Biodiversity Audit Tools

3.1.1. iNaturalist

In total 5,390 CS observations (individual data entries) were downloaded from the expedition project. These observations were made by 25 people with the top five observers recording 74 % of all observations. 10,660 identifications (species identifications by other users) were made by 188 people with the top five people recording 41% of all identifications. There were on average 7.9 +0.4 research grade observations per species (median = 4 records per species), and a range from a single observation for 44 % of the species observed, to 85 observations for Acropora spp. (Elkhorn, and Staghorn Corals). 268 species (47 %) were uniquely observed in the GBRMP and 137 species (24 %) were uniquely observed in the CSMP; 165 species (29 %) were observed in both marine parks (Figure 3).

Of the 5,390 observations, 85 % (4,586) were identified to species, of which 98 % (4,478) were categorized as ‘Research Grade’, thus contributing 570 species (Table 3 The largest proportion and diversity of observations were ray finned fishes (78.4 %, 391 species), followed by molluscs (5.8 %, 58 species), corals (4.7 %, 56 species), and sea cucumbers (3.0 %, 14 species). The remaining 8 % of observations were made up of birds (2.2 %, 4 species), sharks (1.2 %, 7 species), crustaceans (1.1 %, 11 species), sea turtles (1.0 %, green sea turtle or Chelonia mydas only), sea stars (0.9 %, 7 species), algae (0.8 %, 11 species), ascidians (0.4 %, 2 species), sponges (0.1 %, 3 species), hydrozoans (0.1 %, 2 species), and terrestrial plants (0.1 %, 3 species) (Figure 4a and b).

There was no difference in the number of observations between the two marine parks (GLM: t.ratio₁₃ = -0.834, p = 0.4193), however, there were significantly more observations of ray finned fish compared to anemones (GLM: t.ratio₁₃ = -4.770, p = 0.0151), ascidians (GLM: t.ratio₁₃ = -4.507, p = 0.0232), sponges (GLM: t.ratio₁₃ = 4.770, p = 0.0151), and terrestrial plants (GLM: t.ratio₁₃ = 5.033, p = 0.0098). There was no difference in the number of species per taxa observed between the two marine parks (GLM: t.ratio₁₃ = -1.397; p = 0.1859), however, there was significantly more species observed of ray finned fish compared to all other taxa, except molluscs and corals (GLM: p = 0.0016; Supplementary Material 1). There were also significantly more molluscs and coral species observed than the other taxonomic subgroups (i.e., birds, sea turtles, ascidians, anemones, sponges, and terrestrial plants; GLM: p < 0.001) The top five observed species of fish, and all other taxa for the GBRMP and CSMP are listed in Table 3.

3.1.2. eDNA

In total, 7,507 amplicon sequence variants (ASVs) were generated in the full dataset (i.e., all ASVs considered in each assay independently) through bioinformatic processing. Our dataset included 876 unique taxa once it was aggregated across all assays, of which 787 taxa were associated with aquatic environments. Based on the aggregated dataset, 193 (22.0 %) of our DNA sequences were identified as fish, 153 (17.5 %) as bacteria, 90 (10.3 %) as plants; all other taxonomic groups ranged between 7.1 % and < 1% (Figure 5).

Phylogenetic trees based on the ASVs illustrated the relationship between the taxonomic groups and taxa recorded independently in the GBRMP and CSMP (Supplementary Material 2) based on seawater samples collected at two locations and four replicates per location (N = 8 for GBRMP and N = 8 for CSMP).

3.1.3. Marine Vegetation Collection

A total of 19 algal and one seagrass herbarium specimens were collected on the expedition; 15 algal samples from six islands and reefs in the GBRMP, and four algal samples and one seagrass sample from the Herald Cays in the CSMP (Table 4).

2.2. Reef Health Check Tools

3.2.1. CoralWatch

A comparison of the 239 coral surveys from the GBRMP and 138 coral surveys from the CSMP found that, on average, both marine reserves had slightly more corals scoring below a healthy score of 3 (GBRMP average = 2.9 + 0.06, with 55 % scoring less than 3; CSMP = average 2.7 + 0.07, with 58 % scoring less than 3) (Figure 6a and b). Coral health was significantly worse in the CSMP compared to the GBRMP, based on CoralWatch data (t-test: t = 2.2399, df = 321.34, p = 0.03). The morphology of corals was evenly split across the four main categories for the GBR (branching = 17.1 %, plate = 32.6 %, boulder = 29.1 %, and soft = 21.2 %), whereas the CSMP had nearly equal branching (49.5 %) and boulder (37.6 %) corals, fewer soft corals (11.9 %), and almost no plate corals (1 %) recorded (Figure 6c and d).

3.2.2. Great Barrier Reef Census

We collected 184 photos across six coral reefs in the GBRMP over seven separate surveys (two reefs were surveyed twice). The data was submitted online to the Great Barrier Reef census website (https://greatreefcensus.org/). As of the date of publication, the data has not yet been analysed and is not publicly available.

3.2.3. Eye on the Reef

We conducted a total of 22 surveys and recorded 2,518 biota from 18 categories. The most abundant categories were grazing herbivores (47 %), butterflyfish (34 %), and sea cucumbers (11 %). No data was recorded for five categories: Green turtle, Hawksbill turtle, Other turtle, Blacktip shark, and Crown of thorns starfish (Figure 7).

The primary differences in biota between the two regions were the greater number of sea cucumbers observed in the GBRMP (GLM.nb: Z.ratio = -5.838, p < 0.001) and lesser number of butterflyfish (GLM.nb: Z.ratio = 4.170, p = 0.0145). Although there was a trend for more giant clams in the GBRMP, this was not statistically different (GLM.nb: Z.ratio = -3.709, p = 0.0770) (Figure 7).

3.2.4. Marine Debris

Multiple full garbage bags of marine debris were collected either floating on the water’s surface or washed up on the beach of sand cays and islands in both marine parks. Some notable items were multiple intact fluorescent 3ft light bulbs on South Herald Cay in the CSMP. Unfortunately, due to a lack of interest and time, we did not weigh or collate data on the marine debris collected.

3.3. Heritage Assessment

Photographs along a 50 m transect were processed to create a summary of the coral habitats and remnants of the historic shipwreck Foam, which wrecked on Myrmidon Reef in 1893. A previously unknown anchor was recorded during the survey and reported to the Queensland Department of Environment, Science and Innovation (DESI) as well as the Department of Climate Change, Energy, the Environment and Water (DCCEEW).

3.4. Social Survey

3.4.1. Questionnaire

A total of 106 surveys were completed by citizen scientists; 53 were completed within our expedition time in the GBRMP and 53 were completed within our expedition time in the CSMP. We collated all data for survey participants related to their knowledge of coral reefs. The majority (64 %) of citizen scientists reported that they had low knowledge of coral reefs and self-identified as novices (scores 1 to 4). A small percentage (12 %) of citizen scientists reported that they had a high knowledge of coral reefs and considered themselves experts (scores 8 to 10).

In response to the survey question ‘What CS tools did you use today?” the participants used between 0 (19 %) and 4 (3.9 %) CS tools, averaging 1.22 CS tools per day. The most frequently used CS tool in both the GBRMP and CSMP was iNaturalist (52.7 %; Figure 8) followed by Coral Watch (16.3 %) and GBR Census (10.9 %). No survey respondents reported using RedMap during the survey (Figure 8).

A majority (89%) of the participants reported high levels of satisfaction (scores 8 to 10) in response to a question about their overall satisfaction with the day’s activities (Figure 9). We found no significant difference between levels of satisfaction for trips in the CSMP versus the GBRMP. The model's overall fit was statistically significant (F_2,101 = 22.62, p < 0.001), suggesting that the model explained a significant portion of the variance in satisfaction levels. The adjusted R² value of 0.2957 illustrated that our model accounted for approximately 30% of the satisfaction level variability, highlighting the included predictors’ impact. It was found that both health and information (β = 0.124, p < 0.001) significantly predicted satisfaction levels.

A direct quote from one of the tourists who rated the quality of the information delivered as 9/10 stated, “I’d just like to comment on the excellent standard of education, general manner and interpersonal skills of the group leaders. I’m happy such people are leading the way on reef education and ecology. Very impressive!!”

3.5. Public Awareness

The expedition aimed to build an understanding of coral reefs as well as the diversity of local and global threats to coral reef habitats and resident species to empower individual and community stewardship. The public awareness of the CS expedition was difficult to quantify, and so we instead recorded metrics for direct and indirect public awareness. The direct impact on 340 people included the 90 passengers and staff as well as the 250 citizen scientists, identifiers, collaborators, and partners. The indirect awareness was estimated as 701,000 from a combination of advertising (reach of 500,000 people), media (100,000 people), social media (100,000 people), family and friends (500 people), and conference presentations (500 people).

3.6. Carbon Footprint

The trip distance was 1,154 nautical miles, with 26.97 litres of diesel consumed per nautical mile for a total of 31,146 litres. We estimated that the large vessel used the equivalent of 83.5 tonnes of CO₂. We estimated that the four small vessels used approximately 1,560 litres of unleaded fuel equivalent to 4.2 tonnes of CO₂. We assumed that the expedition generator produced the equivalent of 3 tonnes of CO₂. We assumed that the 10 international passenger flights produced the equivalent of 125 tonnes of CO₂ and the 50 domestic passenger and crew flights produced the equivalent of 50 tonnes of CO₂. The total of these seven inputs was 268.7 tonnes or 4.48 tonnes of CO₂ per person.

4. Discussion

Citizen Science (CS) projects involve members of the general public as active participants in research. While some advocates hope that CS can increase scientific knowledge production (“productivity view”), others emphasize that it may bridge a perceived gap between science and the broader society (“democratization view”) [39]. We discuss how the first view was achieved in this project and the second view had challenges including behavior, financial and policy barriers.

Successful in Achieving Scientific Aims of the Expedition

The 2023 CS expedition to the GBRMP and CSMP directly engaged dozens of citizen scientists, many with no previous experience of the Great Barrier Reef or Coral Sea, in a far-ranging data collection and CS awareness-building effort. Despite challenges from cyclones, weather, people, and equipment, we generated data to support each of our four aims.

Citizen scientists increased their knowledge and understanding of coral reef environments and the challenges that they face in a warming world. By the end of the trip, citizen scientists had developed theoretical and practical knowledge of ten different marine CS tools as well as learned the common and scientific names of hundreds of species of birds, fish, molluscs, and algae. With this suite of tools, citizen scientists contributed over 10,000 new data points (including thousands of valuable digital images and eDNA detection and sequencing data) to Australia’s reef monitoring effort. Seagrass and algae samples were collected for donation to national and state herbariums with assistance from citizen scientists. In addition, the expedition team and citizen scientists have continued to communicate the findings through CS conference attendance, the production of an iNaturalist field guide [40] and participating as co-authors of this manuscript. The expedition team communicated the results of the expedition widely, including through the production of over 100 artworks, two media releases, conferences, presentations, and an estimated 100 social media articles, reaching an estimated 700,000 people.

Unsuccessful in Achieving the Carbon Footprint Aims of the Expedition

We were not able to leverage the carbon footprint and offset activity with concurrent fundraising to support a carbon-neutral expedition. This was surprising as recent research by Andre et al. [41] indicates that 56.1 % of Australians (and 69.0 % of the global population) were willing to donate 1 % ($900 AUD) of their income to tackle climate change. Our estimate of 268.7 tonnes of CO₂ for 60 passengers and staff is equivalent to 4.47 tonnes of CO₂ (estimated at $112 AUD) per person. This indicates a major difference in what people say they can do and what they will do. A comparison of the expedition footprint of 4.47 tonnes of CO₂ to that of the average Australian generating a carbon footprint of about 15 tonnes of CO₂ per year indicates high impact [42]. We discussed offsets during the expedition and there was a general mistrust of the process. Even though climate change is the biggest threat to the future of coral reefs worldwide [43], there is currently no relevant climate change or carbon offset policy for companies such as Coral Expeditions, Australian Geographic and Reef Ecologic or citizen scientists or visitors to the GBRMP or CSMP. Achieving climate-neutrality in any endeavor is a three-step process that involves measuring the climate footprint, reducing emissions as much as possible, and offsetting the balance through direct action or funding. A citizen scientist, business, tourism industry, or government-led Environmental Management Charge or Reef Trust Offsets program focused on climate impact and offsets may provide a solution.

Reef Health

The expedition team found that overall, the health of the GBRMP and CSMP during the expedition (late October to early November 2023) was good, which we quantified based on extensive observations of healthy corals, a diversity of species, and few observations of threats such as Crown of Thorns starfish, coral bleaching, or marine debris. The difference in coral morphology between the two marine parks is also an indication of continual recovery from regular cyclonic activity and back-to-back bleaching events in the CSMP [24], providing valuable data on differences between the two marine parks for managers. Given the bleaching being observed across large parts of the GBRMP and CSMP in the austral summer of 2023-2024, this information is particularly valuable. Until now, there have been few scientific comparisons between the two marine parks, and our research is one of the first to compare communities of fish, molluscs, and other biota. We also identified new records of fish, corals, and molluscs, along with significant observations of spawning behavior of fish in the GBRMP and mating turtles in the CSMP. An important heritage discovery by citizen scientists was a previously unknown anchor from the historic shipwreck Foam. This was a surprising result as the wreck was rediscovered on 10 October 1982 and inspected by archaeologists nine times between 1982 and 2018 [44] with considerable effort, time, and technology committed to these inspections. We hypothesise that cyclones between 2015 and 2023 reduced coral cover and sand to partially uncover the previously hidden anchor.

Citizen Science Methods and Practices

Sampaio and Rault [11] suggest that citizen-led initiatives should be encouraged by governmental and non-governmental agencies, along with other stakeholders to strengthen the connection between the public and the scientific community. We agree, but with qualifications, as CS data has inherent advantages and limitations. The CS advantages include the ability to fill data gaps in regions of scarcity and expand monitoring beyond the constraints of research or authority budgets [45]. Another advantage is the ability to communicate science and inspire the public by using a diversity of communication tools. The CS limitations include a lack of reliability and rigor in the data generated, though this can be mitigated by providing educational material in lectures and field workshops alongside supervision by trained scientists prior to and during data collection. As an example with the iNaturalist CS method, recent research indicates that the accuracy of Research Grade taxonomic identifications is 95%, but this varies between taxa [46]. A further insight from Leona Kustra (one of the citizen scientists on the expedition and co-author of this paper), was that “when multiple citizen science platforms are introduced and used by scientists, it can be inspiring for citizens to explore each and determine which ones best suit and motivate them. Offering multiple marine citizen science tools and having the citizens try each tool alongside scientists works to increase confidence amongst citizens which then are more likely to adopt at least one citizen science tool. Since marine citizen science is underrepresented, increased offerings and having scientists introduce, use and support citizens in their learning and use of citizen science platforms may increase acceptance, contributions and advocacy”.

Our experience is that multiple CS methods and tools increased participation, data collection, and enjoyment, ultimately resulting in a better understanding of the marine environment. When used together, methods and tools that measure reef health, such as Eye on the Reef, Great Barrier Reef Census, and CoralWatch provided replication and increased confidence in the data. Use of multiple tools from one vessel and a coordinated project compared to individual tools from multiple vessels and disconnected projects, arguably reduces carbon footprint and increases sustainability.

CS involving community participants can improve knowledge of marine parks by providing supplementary information for management. The addition of CS is complementary to traditional research and monitoring programs and can therefore increase the availability of data on free-to-access data repositories. This is especially notable when the data collected is from remote locations, where traditional research and monitoring can be sparse due to logistical and budget constraints. It also provides participants with an opportunity for a hands-on, enriching experience, and understanding of marine park values.

Of the ten CS tools available during the 2023 Citizen Science of the Great Barrier Reef expedition, the most popular was iNaturalist. Citizen scientists found it easy to use, and the extensive support of online taxonomic experts supporting identifications meant that the data quality was high. Over one-hundred times more fish observations were made with iNaturalist than with Eye on the Reef. Roberts et al. [12] reported that iNaturalist recorded 1.2 to 5.5 times more fish species than structured surveys resulting in significantly greater annual species richness estimates. The expedition observed and recorded 452 fish species, about a third of the total estimated 1,500 - 1,625 fish species known to the region [47] though the ability to make inferences about species abundance with iNaturalist data remains limited [48].

The application of iNaturalist and eDNA metabarcoding in tandem facilitated the detection of a large range of biologically and economically important taxa, as well as those that may be under threat. However, there were also gaps in our taxonomic assignments using these tools. A major shortcoming of the eDNA approach is the poor coverage of molluscan and holothurian species using existing metabarcoding assays. Indeed, although based on eDNA sampling in a subtropical estuary in Queensland, Australia, [31] found low detection rates for oysters and mussels using nearly the same set of metabarcoding assays that we applied here, which the authors attributed to primer bias, poor resolution of the available assays, and incomplete genetic reference databases. Similarly, although we did detect Cardiidae in 2 out of 16 replicates at Orpheus Island in the GBRMP, a taxonomic family that includes Giant Clams, our low rate of detection with eDNA does not reflect their diversity at several locations that we visited on the expedition. If we were interested specifically in this family at the expense of all the rest of the flora and fauna in our sample, we could redesign and reapply more targeted genetic assays. This refinement is facilitated by the fact that all the eDNA samples were archived (frozen at -80 °C) in perpetuity and therefore available for re-examination in the years to come.

5. Conclusions

We propose four recommendations to leverage the lessons learned from the 2023 Citizen Science of the Great Barrier Reef and Coral Sea expedition to assist future marine CS and sustainability activities.

1. Review Strategy. Review the Queensland citizen science strategy (Queensland Government 2021) with a specific focus on marine CS in Queensland and adjacent Australian waters applied in an ecotourism context. Allocate resources to encourage integration and prevent duplication of effort among scientists, citizen scientists, tourism operators, government and indigenous rangers. Encourage citizen science participation in education, field work, data entry and analysis, and the communication of results.

2. Measure changes over time. Repeat the expedition in 2025 (2 years) and 5 years (2028) to compare CS methods, the biodiversity detected, and reef health.

3. Drive Sustainability. Individual researchers lead by example and ensure a carbon neutral footprint. For tourists and businesses, we recommend an Environmental Management Charge or Reef Trust Offset or fee based on our estimated carbon impact for passengers and staff to ensure future expeditions are carbon neutral.

4. Encourage Participation. Work with all sectors of the reef community to encourage people to take part in collaborative CS to help better understand the reef, the challenges that all reefs face, and the sustainable solutions that can be brought to bear to preserve them for the future.