Introduction
The increasing complexity and intensity of anthropogenic impacts on marine systems has heightened the need for robust spatiotemporal physio-biogeochemical monitoring datasets to aid in natural resource damage assessments, conservation, and restoration. Unfortunately, the scientific merit of long-term monitoring programs is often undervalued (Lovett et al. 2007), funding availability is variable (Biber 2011), and support for management priorities can shift with political and/or social pressures, limiting long-term data collection and evaluation. Citizen science is being used more often to overcome these obstacles (Dalby et al. 2021), relying heavily on volunteerism to maximize survey effort and amplify message delivery to policy makers through diverse media channels. The definition of citizen science is broad (Robinson et al. 2018) and, for the purposes of this manuscript, will refer to the process by which any non-scientist collects data or uses the scientific method under the guidance or mentorship of a scientist (Bear 2016). Examples of successful citizen science spatiotemporal monitoring programs include documentation of corals (Branchini et al. 2015) and microplastic pollution (Forrest et al. 2019), investigating disease outbreaks (Harvell et al. 2019), and, perhaps most famously, bird monitoring (Butcher et al. 1990). Recent studies have shown that benefits of citizen science participation extend beyond the dataset to include the participant’s attitude towards and level of engagement in conservation (Lewandowski and Oberhauser 2017; Mason et al. 2020). Benefits to participants, coupled with substantial evidence in the literature for the value of citizen science in filling data gaps, illustrates the value of continued engagement in, and expansion of, such programs (Feldman et al. 2021).
Though a powerful force multiplier, citizen science programs are not always a straightforward solution to meeting data requirements. Startup effort can be significant, and even though a citizen program may produce more data, the cost sometimes proves to be greater than that of a traditional approach (Goldstein et al. 2014; Alfonso, Gharesifard, and When 2022). These costs may include recruitment, training, and outreach, with program administration personnel suggested to be around five and a half full-time employees for a large program (hundreds of volunteers; Locke et al. 2019). Retention of participants may also be challenging, though essential for the program’s data quality and longevity (Seymour and Haklay 2017). Sustaining involvement in long-term programs requires thoughtful engagement strategies and resource inputs to promote retention at critical phases of participation (Hyde et al. 2016; Tiago et al. 2017; Thiel et al. 2023). Costs and personnel time associated with managing large datasets characteristic of the monitoring required for citizen science is often overlooked, and obtaining software, providing broad access, ensuring data interoperability, obtaining timely data submittals, enacting quality assurance plans, and other logistical elements can prevent survey results from making their greatest impact (Schade and Tsinaraki 2016). The most frequently documented concern with citizen data may be its accuracy (Aceves-Bueno et al. 2017). Consideration of these potential pitfalls is essential to best navigate the development process and adaptively improve the program over time to garner trust from managers and policy makers who will be called upon to use the data.
Primary concerns for starting a long-term citizen science program include how to most efficiently find, train, and retain prospective citizen scientists to maximize data quality and project continuity. Investigation into existing participant demographics and motivations can help program leads better target groups that may harbor the greatest interest. The most common participants are often middle age or older, financially secure, and concerned about conservation (Domhnaill, Lyons, and Nolan 2020; Kam, Haklay, and Lorke 2021). Too heavy a focus on profiling prospective participants may harm the program, however, as it may reduce diversity and associated tangible outreach, engagement, and community-building benefits (Hermoso et al. 2019). Retention of participants is of arguably greater value because long-standing volunteers can assist in recruiting and training, display increased data collection proficiency, and be a gauge of long-term program satisfaction and/or changes to specific elements of a program. Appealing to motivations of participants may be a key to retention (Domroese and Johnson 2017; Hyde et al. 2016). Motivations may also change over time as participants are more invested in the science or feel that they are part of a culture and accepted by peers and professional scientists (Rotman et al. 2014; Larson et al. 2020). There is ample evidence in the literature suggesting citizen-provided data are often accurate and valuable for scientific goals (Kosmala et al. 2016) and long-term participation may reduce bias (Dambly et al. 2021). Recruitment and retention are prime challenges in operating citizen science projects, and effectively addressing those challenges may create feedback to benefit the program through longevity and higher data quality.
Understanding the relationship between investment of resources and the recruitment and retention of participants is valuable for any leader seeking to create a citizen science project. Tulloch et al. (2013) evaluates the degree of investment required to achieve specific benefits in citizen science programs on a macro scale, identifying critical elements to maximize effectiveness of the data, but not a granular level essential for application to the delicate developmental stage. Better understanding of startup requirements for citizen science programs, particularly specific elements of program design and the level and type of effort and resources required over an initial period, would allow leaders to determine if their available resources are sufficient and to clarify expectations. Also, it would enable participants to visualize an expected trajectory as the program develops.
In this case study, we document the relationship between outreach effort (e.g., presentations, in-person survey tutorials) and participation across multiple groups in a novel citizen science-driven rockfish survey in the Salish Sea over seven years. This work is conducted on SCUBA in cold water. Diving itself presents a barrier to participation, and it is no surprise that less than a third of all marine related citizen science publications from 2014–2018 required SCUBA (Sandahl and Tøttrup 2020). Some may consider the cold waters of the Salish Sea a hindrance; however, research has shown little difference in interest for marine citizen science projects between tropical and temperate regions (Hermoso, Narváez, and Thiel 2021) and that divers in temperate regions may be more experienced, knowledgeable, and willing to participate (Lucrezi et al. 2018b). This detailed case study of program development from infancy through to a stable and well-respected program (Obaza et al. 2023; Lowry et al. 2024) that delivers actionable scientific data will allow scientists starting citizen-science data efforts to better understand and estimate resource requirements.
Materials and Methods
Survey program
In the inland waters of the state of Washington, USA, and the province of British Columbia, Canada, a region known as the Salish Sea, a long history of exploitation has resulted in substantial population declines for multiple rockfish species (genus Sebastes; Palsson et al. 2009; Williams, Levin, and Palsson 2010; Drake et al. 2010; COSEWIC 2020). In some cases, these declines have been dramatic enough to result in federal listing under the Endangered Species Act (ESA, USA) or the Species at Risk Act (SARA, Canada). The National Oceanic and Atmospheric Administration’s (NOAA) National Marine Fisheries Service (NMFS) ESA recovery plan for yelloweye rockfish (Sebastes ruberrimus) and bocaccio (S. paucispinis) in the Puget Sound/Georgia Basin includes the need to survey for young-of-the-year (YOY) rockfish to evaluate little-understood recruitment dynamics (NMFS 2017). Given the large area of interest, erratic recruitment patterns exhibited by Sebastes spp. (Field and Ralston 2005; Field et al. 2021), and abundance of divers in the region (Northern Economics, Inc. 2016), a citizen scientist SCUBA-based monitoring program in which participants could survey far more sites at a greater frequency than professionals was well suited to this need (Browning et al. 2013). There is already precedent in this region for citizens assisting in rockfish recovery (Sawchuk et al. 2015; Andrews et al. 2019). In 2014, prior to finalization of a recovery plan, a survey methodology was developed based on strategies proposed by Browning et al. (2013) to balance the tradeoff between robust data collection and survey simplicity (Parsons, Lukyanenko, and Wiersma 2011). The result was a timed roving survey for four morphological categories using two traits (deep versus elongate body and presence or absence of a dorsal spot) of YOY rockfish (individuals <10 cm) separated by qualitative habitat types and depth bins (Andrews et al. 2019). Once trained (described herein), volunteers are free to collect data at any time and any site within the Salish Sea and enter their results through an online portal. Regular feedback is provided through email and personal interfacing with project leads to maintain quality assurance and control. Photos or videos are encouraged as evidence of accurate YOY sightings.
Participating groups
Volunteer recruitment was selective, as the program did not have many resources or existing citizen science infrastructure from which to draw. Instead of visiting as many dive organizations as possible, groups with goals aligned towards citizen science or an interest in rockfish recovery were targeted. Interested groups attended a presentation to train new divers in rockfish identification, survey methodologies, and the purpose and goals of the program. Instructive dives were subsequently performed by project leads or experienced citizen surveyors with all new participants, and as needed for all participants, to review data collection methods and results. Project leads provided one-on-one feedback following dives and ongoing guidance on data entry.
At the writing of this manuscript, more than a dozen different groups have participated in data collection. A subset of the most consistently participating groups are included herein because they provide the most meaningful data on program development. The organizations documented are diverse. Emerald Sea Dive Club (ESDC) includes more than two dozen recreational divers with considerable natural history and local dive site knowledge. Harbor WildWatch (HWW) is a nonprofit environmental education institution with 53 volunteer divers and experience conducting citizen science. Pacific Marine Research (PMR) is also a nonprofit education organization, with two staff divers familiar with local species from live underwater video presentations and touch-tank specimen collection dives. Point Defiance Zoo and Aquarium (PDZA) has several staff members accomplished in subtidal surveys, and a group of volunteers diving in the field (13 divers total), whose activities could be expanded to include these surveys. Though NOAA is responsible for leading this effort, one employee and one contractor from this large agency not directly involved with rockfish recovery but capable with SCUBA-based surveys and rockfish identification participated and were counted as their own group. No partner groups were funded for their participation in this effort.
Outreach effort
Outreach events performed by project leads for the five consistently participating groups (“outreach effort”) were logged in a spreadsheet (Supplemental File 1: Appendix A), summed over each year, and compared with the total number of dives each group completed (“survey dives”) in that year, for 2015–2021. An outreach event includes any meeting, presentation or dive with project leads. Outreach logged for the program but not included in the present analysis consists of 26 social media posts, one internet article, two infographics, one report, and one peer-reviewed publication (Andrews et al. 2019). These items were not included because tracking engagement from social media or internet traffic statistics and applying results to any particular group is beyond the scope of this study. The relationship between outreach and survey effort was quantified using a linear regression. Because project leads have noticed that long-term relationships are also important in driving participation, a cumulative effort regression was calculated. An analysis of variance on the interaction effect of cumulative outreach effort against survey dives by survey group was used to compare regression slopes among groups using the lsmeans package (v2.30-0; Lenth 2016) in R (R Core Team, 2023). A Shannon-Wiener diversity index was calculated on unique participants within each group and year combination to determine if participation grew consistently across groups. Though this metric is traditionally applied towards biodiversity, it is well suited to identify spread of effort across individuals within a group in this study.
Results
Outreach effort ranged from one to six events a year for each group except for one group (PDZA), which did not receive any outreach in 2020. Annual median values of outreach events ranged from one to four (Table 1). A new group started each year from 2015 to 2019, ordered as: HWW, ESDC, PMR, PDZA, and NOAA. The first outreach event for every group was a meeting or presentation to introduce the program, followed by a mixture of in-person meetings and training dives. From 2020 to 2021, almost all outreach events were presentations given by video call and dives were planned to allow participants to remain socially distanced before, during, and after diving in accordance with COVID-19 safety protocols.
Table 1
Participating groups in the YOY rockfish survey effort along with median number of outreach events (outreach effort) and number of survey dives conducted by each group. P-values refer to linear regression of cumulative outreach events against annual survey dives in each year of program’s participation.
| GROUP | START YEAR | MEDIAN OUTREACH EVENTS/YEAR | MEDIAN SURVEYS/YEAR | REGRESSION EQUATION | p-VALUE |
|---|---|---|---|---|---|
| Harbor WildWatch (HWW) | 2015 | 2 | 17 | 1.47x+3.55 | 0.25 |
| Emerald Sea Dive Club (ESDC) | 2016 | 2 | 18 | 2.98x–4.09 | 0.04 |
| Pacific Marine Research (PMR) | 2017 | 4 | 9 | 2.09x–6.45 | 0.03 |
| Point Defiance Zoo and Aquarium (PDZA) | 2018 | 1 | 2 | 10.1x–23 | 0.06 |
| NOAA | 2019 | 1 | 2 | 5.33x–6.86 | 0.05 |
Three groups collected no data in their first year of receiving outreach, and no group completed more than four dives in their first year of receiving outreach effort from project leads. All groups but one (HWW) completed a majority of annual dives in the final year recorded for this study. After two years and at least three outreach events, all groups dramatically increased participation. The number of survey dives was significantly related to outreach effort for all groups combined (α = 0.05; Figure 1) but only for three of the five groups, individually (Figure 2; Table 1). Most groups increased survey dives each year, with the exception of HWW that peaked in its third year in the program, then slightly declined to a consistent level in the most recent three years (Figure 3). Slopes for more recently joining groups were sequentially higher with the exception of ESDC and PMR, though did not differ significantly (all p > 0.05).
Figure 1
Linear regression of cumulative outreach effort (number of outreach events completed by project leads) against survey dives for all participating groups combined. Each point represents a single year of this program from 2015–2021.
Figure 2
Linear regressions of outreach effort against survey dives by group. Each point represents a single year of this program from 2015–2021. Regression equations and p values are reported in Table 1. ESDC: Emerald Sea Dive Club; HWW: Harbor WildWatch; PMR: Pacific Marine Research; NOAA: National Oceanic and Atmospheric Administration; PDZA: Point Defiance Zoo and Aquarium.
Figure 3
Number of survey dives completed by each group over time. ESDC: Emerald Sea Dive Club; HWW: Harbor WildWatch; PMR: Pacific Marine Research; NOAA: National Oceanic and Atmospheric Administration; PDZA: Point Defiance Zoo and Aquarium.
Groups split into three broad categories related to participation. NOAA and PMR never had more than two participants, ESDC and PDZA exhibited an intermediate range of seven to ten participants at their peak, and HWW had fifteen or more participants for most of their involvement. The number of participants either remained constant or slightly increased for all groups except HWW, which increased dramatically over the first three years, then declined over the following three to a number closer to intermediate participation (Figure 4). Shannon-Wiener diversity, a proxy for spread of survey effort within a group, increased over time for all groups besides HWW, which increased for three years and then declined the following three to a level similar to other groups (Figure 5).
Figure 4
Number of participants that conducted survey dives affiliated with each group over time. Three groups collected no data during their first year and therefore have no data reported for that time. ESDC: Emerald Sea Dive Club; HWW: Harbor WildWatch; PMR: Pacific Marine Research; NOAA: National Oceanic and Atmospheric Administration; PDZA: Point Defiance Zoo and Aquarium.
Figure 5
Shannon-Wiener diversity index for each group over time, where this metric may be used as a proxy for effort spread across participants within each program. ESDC: Emerald Sea Dive Club; HWW: Harbor WildWatch; PMR: Pacific Marine Research; NOAA: National Oceanic and Atmospheric Administration; PDZA: Point Defiance Zoo and Aquarium.
Increases in effort brought tangible value to the dataset beyond the intrinsic benefits of program growth. Eleven unique sites were surveyed in 2015, while sixty-four unique sites were surveyed in 2021, with four visited in both years (Figure 6). The greater spatial extent of survey effort captured a higher diversity of morphological groups, whereas only deep bodied YOY lacking a dorsal spot were recorded in 2015. The combination of greater spatial and temporal survey coverage captured a brief high recruitment event for elongate bodied YOY with a dorsal spot (S. flavidus and S. melanops) in Admiralty Inlet in March of 2021 (Figure 6).
Figure 6
Map of survey effort between 2015 and 2021, where each point represents at least one survey. Pie charts illustrate the change in morphological group abundance coinciding with increased effort. A comparison of encounter rate of elongate bodied young-of-the-year with a dorsal spot in Admiralty Inlet are presented for 2015–2020, all months in 2021 except March (*), and March 2021. Error bars are standard error.
Discussion
Outreach is critical for citizen scientist participation and this case study shows several important factors consistent across groups. First, participation more than doubled between years two and three for every group. The number of outreach events across years did not hold constant for each group, so it is impossible to tease apart the interaction between time and outreach events. Regardless, it is clear that a period of relationship building is required before substantial participation occurs. Second, the regression slopes are mostly steeper the later a group begins collecting data. The consistency in this relationship suggests more rapid participation may not be a function of the group itself, but several factors, including, but not limited to, longevity of the program, the strength and consistency of participation of citizens and program leads, and regular feedback on program goals and results (Hansen and Bonny 2021). The dive community, even in a large metropolitan area, is small. Divers talk amongst divers and a fun, constructive citizen science program can be communicated and build readiness to participate (Lucrezi et al. 2018a). If a program is creating attractive products, whether they be social media videos, applications to management or scientific achievements, or unique ways to socialize in activities of interest, citizens may more readily be willing to commit their time (Alender 2016). In this case, it also is likely that project leads improved their messaging and outreach technique over time while not deviating from overall goals (Schulwitz et al. 2021), with added input from and direct experience with participants. A database improved by this additional effort, capable of recording a greater diversity of rockfish species and answering questions about rockfish recruitment (e.g., a high-recruitment event in March 2021), makes its own argument for further effort. Whatever the cause, outreach, as a measure of time spent meaningfully engaging citizens, works within and across groups as an investment, and a slow start to a citizen science program may be viewed as a period of start-up investment, not lack of interest.
Unlike the consistent change in survey output by groups, individual participation within groups varied substantially. HWW, a non-profit organization, exhibited the highest individual participation and survey diversity, but their data input declined in recent years. No other group exhibited a decline in participation or diversity, nor have they neared the individual participation peak of HWW. Despite fewer participants, other groups (notably PMR and NOAA) have reached similar survey output as HWW during their peak year. Even if groups react similarly to outreach, that does not mean their composition or individual effort is uniform. HWW engages a larger volunteer base, and greater participation would be expected to exhibit greater variability, hence ebbs and flows of engagement (Hyde et al. 2016). Smaller groups that are led by a handful of highly motivated individuals may have less variability (Seymour and Haklay 2017). No group has worked on this project as long as HWW, either, and ebbs and flows of interest should be expected over time. HWW surveys are also led by a single individual, and competing demands for any individual’s time affect the entire organization’s effort. Members of ESDC noticed that a variety of personality types were useful in maximizing their contributions. One archetype was effective at driving excitement and early interest, originally brought project leads into the club to speak, and participated in the majority of the early surveys. However, they may not exhibit organizational and reliability traits. The more process-oriented members arranged club dives and organized group efforts for a larger long-term impact. While it may be too much effort and outside the bounds of club interactions to run members through personality assessment testing, it is useful to keep in mind that it may take multiple personality types to construct a lasting volunteer program. Identification of these personality types within groups may occur during the relationship-building stage and may contribute to large participation jumps. Project leads should be receptive to these archetypes and allow them to guide their participation and program construction. Though survey output of participating groups may appear similar on a macro scale, engagement and expectations should be tailored to the characteristics of each specific organization.
While not separately quantified as part of this study, conducting dives as outreach and training is likely an influential factor in promoting participation. Leadership that is willing to get in cold water alongside participants, demonstrate dive and scientific proficiency, and do so repeatedly, will gain respect (Yanamandra-Fisher et al. 2020). Leading by example has been documented as beneficial in other environmental fields as well (Boiral, Talbot, and Paillé 2013). The unique challenges of organizing a citizen science project that require diving in cold water may be countered by leadership willing to participate alongside participants and do not preclude broader application of these results.
The program presented in this study, though supported by a United States federal agency, had little citizen science infrastructure at its launch. Because the program is new, these results are most applicable to novel citizen science efforts. Small size and novelty may be considered a detriment to recruitment, but it also confers advantages that are not conveyed within the presented data that may be instructive. A small leadership team must externalize a portion of survey planning to the participants. Beyond basic guidelines, site selection was left up to participants. Not only did this allow participants to survey in areas of their interest, but it illustrated their sense of equality with professional scientists, often seen as a barrier to entry (Martin et al. 2016; Lewandowski et al. 2017) and discernably improved the database (e.g., substantial spatial effort to record a high-recruitment event in 2021). Regular communication between participants and project leads, which may not be possible in larger programs, allows for near-real-time feedback and furthers the sense of value for participants (Robinson et al. 2021; Thiel et al. 2023). This relationship allows the experienced citizen divers to offer even further input on the program in the form of recruitment strategies, tweaks in methodology, and additional research questions, creating positive feedback for participation (Darch 2017). Project leads remained the same throughout the study period, eliminating the need for a novel relationship to be created, and furthering trust between participants and program leadership. The smaller size of a project, along with consistent leadership, allows some decentralization to create a more bottom-up than top-down program that comes closer to realizing the full benefits of a talented and experienced team of citizen scientists.
Interpretation of participation results via program-specific traits is necessary to put this study in context, though the process here falls short of the rigor found in numerous studies designed specifically to classify participation motivations through interviews and surveys. The chief utility of these results is not in answering the “why” of participation, but the “how” of starting a successful new program. Programs featured in the scientific literature understandably focus on successes, but many efforts are not developed to the point of reaching their goals (Westreicher et al. 2021). Expectation management and construction of realistic timelines are integral to successful citizen science and are both quantified in this study. Incorporation of the lessons documented in this case study into subsequent projects will increase the likelihood of success and long-term sustainability of data collection efforts.
Conclusions
This study presents a granular view of growing citizen science participation from consistent outreach effort in a cold-water diving project. Though the manner of data collection requires specialized equipment and training, the messages provided are likely applicable to the broader citizen science community. Perseverance through early relationship-building stages is necessary regardless of group composition. Project leads should be prepared to invest substantial resources into outreach, including extensive training and data collection alongside participants, before expecting meaningful data yields. Programs do not need large amounts of existing infrastructure, but do benefit from consistent messaging, adapting to a decentralized structure to enable participants, and an improved database to show tangible value of effort. These lessons, centered around patience, effort, and flexibility, should be adequately communicated to funders or supervisors to clarify expectations and allow time for success. By coalescing expectations and requirements from this case-study on cold-water diving, future citizen science projects will be more likely to create a sustainable program that engages citizens as vital partners in conservation.
Data Accessibility Statement
Data from this study are available upon request, but no personally identifiable information will be shared as part of that dataset.
Supplemental File
The supplemental file for this article can be found as follows:
Ethics and Consent
This work is authorized via the NOAA Citizen Science and Crowdsourcing information collections submitted to the United States Office of Management and Budget, of which this program is a part. Data collected are authorized under 5 U.S.C. § 301, Departmental regulations which authorizes the operations of an executive agency, including the creation, custodianship, maintenance and distribution of records, and 15 U.S.C. 1512, Powers and duties of the Department. Disclosure of this information is permitted under the Privacy Act of 1974 (5 U.S.C. Section 552a) to be shared among Department staff for work-related purposes. Disclosure of this information is also subject to all of the published routine uses as identified in the Privacy Act System of Records Notice COMMERCE/NOAA-11.
Acknowledgements
The authors wish to thank all divers who collected data for this project as well as Chris Yates and Lynne Barre for project support, and the Seattle Aquarium for building outreach material. Lisa Glover with the NOAA Central Library conducted a thorough literature collection to support this manuscript.
Funding Information
The SeaDoc Society graciously covered publication fees. Work funded under NOAA grants NFA-NFAPO-2018-2005418 and WCRO-2021-2006805. This manuscript is dedicated to Brian Bugge and Jerry Dollar, two inspirational surveyors that passed too soon.
Competing Interests
The authors have no competing interests to declare.
Author Contributions
Obaza collected and analyzed the data, conducted outreach and wrote the manuscript. Bird collected data, conducted outreach, and wrote the manuscript. Gaydos, Wilken, Selbitschka, Troyer, Peltonen, and Berger conducted outreach and helped write manuscript. Selleck conducted outreach and collected data. Lowry wrote the manuscript and guided the project. Tonnes conceived the project, conducted outreach, and wrote the manuscript.