The Citizen Observation of Local Litter in coastal ECosysTems (COLLECT) project (2021-2022) is a citizen science initiative which aimed to acquire distribution and abundance data of coastal plastic litter in seven countries, in Africa (Benin, Cabo Verde, Côte d’Ivoire, Ghana, Morocco, Nigeria) and Asia (Malaysia). The COLLECT project consisted of training local students (15 - 18 years old) from second cycle institutions on sampling and analyzing macro-, meso- and microplastics (see size definitions in Methods) in beach sediments, using a quantitative assessment. The project further aimed to measure the impact of the citizen science intervention in ocean literacy, pro environmental behavior, and wellbeing of the participating students. The methodologies used in COLLECT followed best practices of citizen science projects on plastic litter quantification in aquatic environments (Rambonnet et al., 2019; Barnardo et al., 2020; de Rijck et al., 2020). The project ran under the auspices of the Partnership for Observation of the Global Ocean (POGO), a well-established network of oceanographic research institutions that work together to promote and implement global ocean observations through innovation, capacity development, outreach, and advocacy (Miloslavich et al., 2019; POGO, 2021). To reach out to a broad audience, COLLECT follows the ‘FAIR Guiding Principles for scientific data management and stewardship’ (Wilkinson et al., 2016), to guarantee the Findability, Accessibility, Interoperability, and Reuse (FAIR) (GFISCO, 2022) of all the digital assets produced within the project. The COLLECT project contributes to the United Nations’ Sustainable Development Goals (SDGs) by focusing on the sustainable use of the ocean (United Nations, 2022b).

The goal of the current work is to present and discuss the workflow ( Figure 1 ) of the establishment of the COLLECT project, along with its challenges and mitigation measures. We describe in detail the methodologies applied in sampling plastic litter and processing data ( Figure 2 ), and in assessing the impact of participating in the activities from a social sciences perspective. We further discuss the expected results and outcomes of the project, and offer recommendations for similar projects in citizen science observation of environmental litter.

To address the paucity of data on plastic litter distribution in Africa, the COLLECT (Citizen Observation of Local Litter in coastal ECosysTems) project recruited participants in six countries, namely Benin, Cabo Verde, Côte d’Ivoire, Ghana, Morocco, Nigeria ( Figure 3 ). The project included a Southeast Asian country, Malaysia, to demonstrate that the project could be expanded to other geographic areas, and that plastic litter samples could be collected and processed by school children globally, following the same procedures. In each country, one or two representatives from POGO member institutes in the region (focal points), organized the local activities, such as recruiting and training students and teachers, and leading the field sampling. Local focal points selected the sampling areas based on the following criteria: proximity to the research institute and the participating schools, sandy beaches located preferably at least 5 km away from an estuary or any other inland water bodies, accessibility by transportation vehicles, and with minimum impediments to ensure the health and safety of the participants ( Figures 2 , 3 ). Sampling events took place in rainy (or wet) versus dry seasons, or Autumn versus early Spring [October – December and January – April respectively], depending on the participating countries. Where needed, each focal point requested the necessary permission from local authorities for the field activities to take place.

The goal of COLLECT was to engage with ten second cycle schools and to recruit and train local students, preferably between the ages of 15 and 18, in beach litter sampling methodologies. Focal points managed recruiting tasks and communicated directly with participating schools, introducing the project and methodologies to the students and teachers, i.e., the citizen scientists. They also engaged with local media to promote the project, and managed all the local logistical arrangements. Each headteacher received an official letter from POGO, introducing the project and the planned activities ( Supplementary Information , Letters of Consent), and inviting a maximum of 50 students per school to participate. Due to the COVID restrictions in place, and to minimize risks, participants were assigned to one of two smaller groups (≤ 25), each of which carried out a dedicated field activity in each season. In cases where the students were underage (< 18 y.o.), parents were required to provide consent for their child’s participation in the COLLECT activities ( Supplementary information , Informed Consent Letter(s) to Parents). All students 16 y.o. or older had to provide additional personal consent forms to participate in all project related activities ( Supplementary Information , Informed Consent Letter(s) to Students, 16 y.o. or older; for the ethical committee approval, see below section Evaluation of the educational, behavioral, and well-being impact on students). In addition, consent from parents and schools was required to take photo or video recordings of students and to use it for educational and outreach purposes ( Supplementary Information , Informed Consent Letter(s) to Parents). Templates for invitation letters and consent forms are available in the supplementary information , in the working languages (English, French, Portuguese) of the project.

Prior to the field activities, there was an educational in-person workshop in each school, led by the local focal points and their teams, and supported by the collaborating science and head teachers, to discuss the topic of plastic pollution, and to inform and train the students on the sampling procedures ( Figure 4 ). The workshop was an opportunity for students to interact directly with local scientific researchers and institutions. To ensure message consistency, focal points delivered the same presentation, scripted in the local language ( Supplementary Information , Information session). During each field campaign, and prior to the sampling activity, the students participated in a briefing led by focal points, to receive instructions and training on how to sample and/or collect information on plastic litter items and where each student received a simplified graphical illustration of the sampling procedure (Catarino et al., 2021c).

The goal of the COLLECT project was to sample macro- (> 25 mm), meso- (5 - 25 mm) (MSFD Technical Subgroup on Marine Litter, 2013; GESAMP, 2019) and microplastics (< 5 mm, sampled up to 1 mm) (Bosker et al., 2017b; Kiessling et al., 2021) in sandy beaches following quantitative procedures. To do so, we developed an extended and technical standard operating procedure (SOP) (Catarino et al., 2021d), with detailed instructions on materials and methods, and on sampling acquisition and processing, together with detailed recording sheets, and a section with health and safety recommendations (Cheshire et al., 2009; OSPAR, 2010; MSFD Technical Subgroup on Marine Litter, 2013; Bosker et al., 2017a; Bosker et al., 2017b; Barnardo et al., 2020). Both the extended SOP (Catarino et al., 2021d), including required materials, and the recording datasheets (Catarino et al., 2021e) were translated into the local school languages from English, i.e. French and Portuguese. All equipment was provided to the schools and students participating in the field activities [see detailed and technical SOP https://dx.doi.org/10.48470/38, (Catarino et al., 2021d)]. For sampling and processing activities, students were divided into groups (up to 5 groups per field campaign) of 5 – 6 students, and their work was mentored either by a teacher, researcher (e.g. focal point) or trained technical staff from the focal point institution. Sampling was done along a 50 m transect displayed at the high tide line, parallel to the water line, in six sampling spots displayed every 10 m. In each sampling spot, participants defined a 1 m radius circle (sampling spot area of 3.14 m²). Students recorded and classified all the macroplastic items visible on the sand surface ( Figure 2 ) according to the Guideline for Monitoring Marine Litter on the Beaches (OSPAR, 2010), which provides not only the name of the items in various languages (including English, French, Portuguese and Spanish), as well as images of the litter items, to assist in their identification and classification. In each circle, participants sampled the top 5 cm of sand of a 10 x 10 cm randomly displayed quadrat ( Figure 2 ). Sand was first sieved using a 25 mm mesh, and all macroplastics of smaller dimensions were collected. The sand was sieved once more using a 5 mm meshed sieve, and all mesoplastics and sieved sand were collected in separate zip bags until further processing. The sampling procedures were repeated in all sampling spots, along each group’s transect [see detailed and technical SOP https://dx.doi.org/10.48470/38, (Catarino et al., 2021d)].

All litter samples were transferred to the schools and air dried for 1 week or longer, before sample processing. Macroplastics collected from the circle surface and from the quadrat sand sample, and mesoplastics collected from the quadrat, were weighed and their length measured (to the nearest mm) wherever possible. The sand sample collected in the field was sieved using a 1 mm meshed sieve, with the aid of water (Catarino et al., 2021d). Collected particles were extracted from the remaining sediment and other materials using a simplified density separation method (supersaturated NaCl solution made using cooking salt) (Cutroneo et al., 2021). All microplastics were handpicked using metal spoons or tweezers, weighed (total weight of particles to the nearest g), and observed using a didactic microscope (purchased locally). The smaller macro-, the meso- and the microplastics were described and classified according to shape and color following the classification of the MSFD Technical Subgroup on Marine Litter (MSFD Technical Subgroup on Marine Litter, 2013). All items were displayed on a white surface and photographed (together with a ruler for scale reference). All meso- and microplastics were carefully wrapped in aluminum foil and stored until further analysis by collaborating researchers, while macroplastics were disposed according to local litter recycling and waste disposal guidelines (Catarino et al., 2021d).

Within the POGO network of member institutions, we have identified collaborating partner research groups with access to equipment to assess the polymer types of meso- and microplastics, such as Raman and Fourier transform infrared (FTIR) spectroscopy. To guarantee intercomparability of polymer identification methodologies applied by each research group, a series of standard particles was distributed amongst partners to perform a calibration exercise (blind identification test); the particles size ranged between 5 and 1 mm (CARAT, Belgium), and the polymer types were based on frequently sampled plastic items on beaches (Suaria et al., 2020; Everaert et al., 2022), namely polystyrene (PS), polyvinyl chloride (PVC), polypropylene (PP), nylon, polyethylene (PE) and polyethylene terephthalate (PET). Up to 100 randomly selected sub-samples of meso- and microplastics were then shipped to the collaborating institutes and analyzed for polymer identification.

To ensure the reliability of the data procedures, the SOP and datasheets were trialed in Ghana prior to the project by one of the project principal investigators and a team of professional and technical staff. Even though the participation of school students in citizen science plastic litter sampling campaigns is of comparable efficiency to researchers and trained teachers (Hidalgo-Ruz and Thiel, 2015; van der Velde et al., 2017), during this project we ensured that all steps of data acquisition and processing done by the students were supervised by trained technical staff and researchers, to have robust and reliable data (Hidalgo-Ruz and Thiel, 2015). Because the sampling did not target microplastics smaller than 1 mm, there were limited opportunities for sample contamination, as particles that could have been released from the use of plastic materials (e.g., zip lock bags used for sand and particles storage) and airborne fibers would most likely be smaller than 1 mm (Everaert et al., 2022), a size range outside the scope of this project (Bosker et al., 2017b). Using supersaturated sodium chloride solutions (density = 1.2 g/cm³) for microplastic extraction from field samples is not the most efficient salt to separate denser particles (e.g. PVC > 1.16 g/cm³, or PET > 1.37 g/cm³), when compared, for instance, to sodium polytungstate, zinc chloride or sodium iodide (MSFD Technical Subgroup on Marine Litter, 2013; Cutroneo et al., 2021). However, complex density separation techniques performed in a laboratory setting use salts that are either expensive and inaccessible to citizen science projects, or that may have associated health and safety risks (MSFD Technical Subgroup on Marine Litter, 2013; Cutroneo et al., 2021). Supersaturated solutions of NaCl are an accessible and safe alternative for any school activity, and their preparation requires only inexpensive and commercially available cooking salt (MSFD Technical Subgroup on Marine Litter, 2013; Cutroneo et al., 2021). All data recorded in paper format [see field and classroom datasheets in the COLLECT DMP, https://dx.doi.org/10.48470/40, (Catarino et al., 2021d)] were curated before insertion in digital table format (Microsoft Excel). Handwritten information was double checked by at least two independent users and all data were verified for consistency (Kiessling et al., 2019). The photos of plastic litter (from macro- to microplastics) were checked to confirm the suitability and reliability of the data inserted in the datasheets (Kiessling et al., 2019).

The social sciences component of the COLLECT project aimed to evaluate the educational, behavioral, and well-being impact of the citizen science intervention, by means of a pre- and post-assessment questionnaire (Severin et al., 2021). Students answered the same 15-minute questionnaire in their school language [available here: https://doi.org/10.17605/OSF.IO/VB8TX] prior to the training and again after the conclusion of their participation in the school activities (30 days on average). COLLECT’s social sciences component, including experimental design, questionnaires and data analysis plan, was preregistered in the Open Science Framework (OSF, see https://doi.org/10.17605/OSF.IO/VB8TX) registry prior to data collection (Severin et al., 2021), had its own data management plan (DMP) submitted to the DMPonline platform (dmponline.be, Belgium) [see section Data management and analysis], and was approved by the ethical committee of the Faculty of Psychology and Educational Sciences of Ghent University (ref: 2021/65). Parents and educators provided written informed consent for the students to participate in this study (see Supplementary Material for consent form templates). The questionnaire was prepared in English and translated to French and Portuguese, with each translation reviewed by at least a second native speaker. Schools conducted the questionnaire either on paper (African schools) or online (Malaysian school), using the Limesurvey platform (Limesurvey GmbH; www.limesurvey.org).

The questionnaire was divided into three main topics: knowledge and awareness of marine litter, attitudes and behaviors towards marine litter and the environment in general, and nature connectedness and well-being. Questions regarding knowledge and awareness of marine litter were based on previous studies evaluating the impact of citizen science interventions on children or adolescents and perceptions of marine litter (Hartley et al., 2015; Wyles et al., 2017; Hartley et al., 2018; Locritani et al., 2019). Specifically, participants were evaluated on problem awareness, concern, and perceived causes and impacts of marine litter, with the use of Likert agreement scales. Perceived prevalence of plastic on the coastline was assessed with multiple-choice questions such as “what percentage of litter on the beach and in the sea do you think is plastic?” or “what do you think was the most common type of litter found on the coastline near you in 2020?”. An additional question on the objective knowledge of degradation time of plastic was included. The second topic of the questionnaire measured litter-reducing behaviors with items adapted from Hartley et al. (2015) and Locritani et al. (2019). Pro-environmental behavioral intentions were assessed by participants rating how often they think they will participate in beach clean-ups in the future (Wyles et al., 2017), as well as how often they will engage in more generic pro-environmental behaviors (e.g., “recycle”; “re-use plastic bags”). Participants’ attitudes towards beach litter removal were measured with items taken from Lucrezi and Digun-Aweto, 2020, e.g., “it is a responsibility of the local community to remove litter from the beach”. An additional item was also included: “only those who originally pollute the beach are responsible for removing litter from the beach”. General pro-environmental attitudes were measured with the revised version of the New Ecological Paradigm (NEP) scale (Dunlap et al., 2000) that is adapted for children and adolescents (Manoli et al., 2007). This 10-item scale assesses the extent to which one endorses the rights of nature, the possibility of an eco-crisis, and rejects human exemptionalism. The third and final topic of the questionnaire consisted of measuring the participants’ sense of connection with nature, with the use of the Nature Connection Index (NCI) (Richardson et al., 2019), which has been validated for both children and adults. Participants’ well-being was assessed with the Short Warwick Edinburgh Mental Well-Being Scale (SWEMWBS), developed by the Universities of Warwick, Edinburgh, and Leeds in conjunction with NHS Health Scotland (Stewart-Brown et al., 2009). The seven-item scale measures various aspects of well-being, such as social connection and sense of agency. A question on how happy participants felt was also added to evaluate well-being, with a ten-point scale going from extremely unhappy to extremely happy. Finally, the post-questionnaire included additional questions on how satisfied participants were with the COLLECT project, how worthwhile and meaningful it was to them, and which adjectives best described their experience of the project.

It is important to note that the items and scales were carefully chosen according to their established validity and reliability regarding the socio-demographics of our sample (i.e., being of young age and residing in African or Asian countries). For example, the NEP scale has been applied within an African context (Petegem and Blieck, 2006; Ogunbode, 2013), and the SWEMWBS has been validated in various cultural settings (Stewart-Brown et al., 2009). In addition, the items taken from Hartley et al., 2015; Locritani et al., 2019 were designed to be appropriate for children and adolescents. To ensure further validity of the survey questions, the survey was piloted with the students from Calabar, Nigeria, who could additionally indicate when they did not understand the question.

The COLLECT project had two data management plans (DMPs), one focused on plastic litter sampling and processing procedures (Catarino et al., 2021e) and another one focused on the component of social sciences (Severin et al., 2021). Each DMP, a document specifying how research data will be handled both during and after a research project, was submitted to the platform DMPonline (www.dmponline.be, Belgium). The DMPs identified the collection and origin of new data, types of data to be acquired, file formats and expected volume, metadata standards and the controlled vocabularies to be used, where data would be archived, and backup procedures. The DMP concerning the social sciences also contained information concerning General Data Protection Regulation (GDPR) under European Union (EU) law (https://eur-lex.europa.eu/eli/reg/2016/679/oj), which protects individuals with regard to the processing of personal data and on the free movement (i.e., transfer to third parties) of such data. Ethical aspects and additional security measures for data protection (e.g., a Data Protection Impact Assessment) were also included.

All materials and data generated during the COLLECT project comply with the FAIR (Findability, Accessibility, Interoperability, and Reuse) principles of digital assets (Wilkinson et al., 2016). Materials and data will be archived using the Marine Data Archive (MDA, www.mda.vliz.be, VLIZ, Belgium), and searchable via a Digital Object Identifier (DOI, www.doi.org) The Integrated Marine Information System (IMIS, www.vliz.be/en/imis, VLIZ, Belgium). The macroplastics litter was classified following the OSPAR guidelines for monitoring marine litter on beaches (OSPAR, 2010). The meso- and microplastics items were classified according to the type, shape and color reported by the Marine Strategy Framework Directive (MSFD) technical Subgroup on marine Litter (MSFD Technical Subgroup on Marine Litter, 2013). All the data generated will be made open access and unrestricted after completion of the project. This means that data will be freely available at no charge to third parties, submitting it to the The European Marine Observation and Data Network (EMODnet) central beach litter database, operated by EMODnet Chemistry (www.emodnet-chemistry.eu). A link to the data portal will be included in the metadata record in IMIS.

During the design phase of the project, international data and metadata standards were consulted to comply with dataset requirements for the targeted data portal, ensuring interoperability and facilitating exchange and re-use of the data. Common controlled vocabularies, such as the OSPAR guidelines and BODC Vocabularies (British Oceanographic Data Centre, www.bodc.ac.uk/resources/vocabularies), and file format (Excel spreadsheet file), were used for consistency and interoperability, enabling records to be interpreted by computers. The data quality control procedure was initiated locally by the focal points and overseen by the data management collaborators based at VLIZ (Belgium). The latter ensure the quality and completeness of the data, before datasets are made available through international repositories. After the initial quality checks and formatting of the data according to the requirements from EMODnet, the EMODnet Beach Litter Format Validator will be used to check correctness and approval of the generated files. Quality control procedures will be described in the metadata of the dataset (Fraisl et al., 2022).

Analysis of the sampled plastic litter data will be done using the free software R (R Core Team, 2022). We will assess differences in the diversity and abundance of macroplastic items observed in each season and collection site. We will further investigate the size frequency distribution of meso- and microplastics and diversity of these particles according to shape, color and polymer types at each collection site. Absence of litter observations was considered as data observations, if sampling was performed, and will be expressed as 0 items per sampling unit.

The analysis of the social sciences data will be carried out with the use of the software IBM SPSS Statistics (Version 28). We will evaluate changes in knowledge and awareness of marine litter, attitudes and behaviors towards marine litter and the environment in general, and well-being, using a within-subject design with pre- and post-assessment. Potential differences amongst countries, ages, and genders will be explored (Severin et al., 2021).

The COLLECT communications plan focused on the development of tools to support the collaborating schools, reach out to local communities, and to provide visibility for the project activities and progress (Seeyave et al., 2017). It consisted of creating a visual identity, establishing social media profiles on appropriate public platforms, engaging with local media outlets, producing educational materials, and compiling media clippings to assess impact ( Table 1 ). The COLLECT visual identity, created in collaboration with a consultancy company (Science Crunchers, Portugal), comprises the project logo and the branding guidelines for visuals, colors and fonts applied across all project materials. Through application of the guidelines, communication activities (e.g., presentations, banners, documentation, etc.) were consistent in all participating countries. A webpage (pogo-ocean.org/innovation-in-ocean-observing/activities/collect-citizen-observation-of-local-litter-in-coastal-ecosystems) lists the main information about the project (e.g., goals, strategy, outcomes, team, collaborators, contacts) and social media presence on various platforms (Twitter, Instagram, Facebook) ensures dissemination of the project goal and activities to a broad audience. The project educational materials, namely a simplified illustrated sampling protocol (Catarino et al., 2021c) and its video companion (YouTube, shorturl.at/gyGHP) (Catarino et al., 2021b),developed in collaboration with the same consultancy as the visual identity, were based on the extended technical sampling protocol and produced or translated into the working languages (i.e., English, French, Portuguese and for the illustrated versions additionally in Spanish). The COLLECT communication plan included a component focused on national/local advertisement/engagement via traditional media (radio, newspaper and television). All media clippings were compiled to assess media impact and demonstrate outreach (see Supplementary Information ). The results of the project are being shared with participating schools, teachers and students via webinars organized by the local focal points, and via a poster or booklet to be distributed to the headteachers.