Introduction
Remoras are a family of teleost fishes (Osteichthyes: Carangiformes: Echeneidae) that attach to other marine animals via an adhesive disc, derived from dorsal fin elements, located on top of their head (Britz and Johnson Reference Britz and Johnson2012). The oval-shaped disc is composed of bony, flat lamellae that can independently rotate, surrounded by a fleshy lip, which can be used like a suction cup to attach to organisms with a variety of skin textures. Remoras have been observed attached to a wide range of species, including sharks, rays, turtles, dolphins, whales, bony fishes (including billfishes), and even humans (Andrade Reference Andrade2007; Brunnschweiler et al. Reference Brunnschweiler and Sazima2008; de Sola Reference De Sola1932; Fertl et al. Reference Fertl and Landry1999; Kenaley et al. Reference Kenaley, Stote, Ludt and Chakrabarty2019; Nicholson-Jack et al. Reference Nicholson-Jack, Harris, Ballard, Turner and Stevens2021; O’Toole Reference O’Toole2002; Sazima and Grossman Reference Sazima and Grossman2006; Sazima Reference Sazima2018).
It can be difficult to classify the relationship between a remora and its host (Parmentier and Michel Reference Parmentier and Michel2013). Some interactions are mutualistic in nature (i.e., the remora consumes parasites from its host, benefiting both), while some are commensal (the remora eats scraps from its host’s prey, which does not benefit but does not harm the host). Of particular interest for this study, Echeneis naucrates (common name: Sharksucker) is known to prefer ostraccid and scarid fishes as hosts while a juvenile, later transitioning to larger hosts like reef sharks (O’Toole Reference O’Toole2002).
It is difficult to conclusively document how long remoras remain attached to a specific individual host because observations are sporadic and opportunistic. Remoras were identified four times on the same individual dolphin over a two-month period (Sazima Reference Sazima2018), but it is unknown if they remained attached the entire time because they were not observed every day. It is more common to observe that a host often has a remora attached to it over a long period of time, though it is impossible to determine if it was the same individual remora (see Brunnschweiler et al. Reference Brunnschweiler and Sazima2008, which documented repeated observations of a remora on a blubberlip snapper, Lutjanus rivulatus, at a popular dive site). One of the challenges in understanding remora attachment duration is the ability to maintain long-term underwater visual records.
Underwater video systems are a common tool for documenting the presence, abundance, and behaviour of marine life (Cappo et al. Reference Cappo, Speare and De’ath2004; Harvey et al. Reference Harvey, McLean, Frusher, Haywood, Newman and Williams2013; Whitmarsh et al. Reference Whitmarsh, Fairweather and Huveneers2017). Unlike direct observation by divers, videos allow for the non-invasive observation of natural behaviours (Coleman and Burge Reference Coleman and Burge2021; Zarco-Perello and Enríquez Reference Zarco-Perello and Enríquez2019). Videos have previously been used to describe the ecological interactions between fishes and fishing gear (Renchen et al. Reference Renchen, Pittman and Brandt2012), foraging behaviour and dietary preference (Lee et al. Reference Lee, Huang, Chen and Lin2016), behavioural responses across gradients of predator abundance (Goetze et al. Reference Goetze, Januchowski‐Hartley, Claudet, Langlois, Wilson and Jupiter2017; Sherman et al. Reference Sherman, Heupel, Moore, Chin and Simpfendorfer2020a), reproductive behaviour (Esteve Reference Esteve2005), predation (Crook and Davoren Reference Crook and Davoren2014), and behaviourally mediated indirect interactions (Coleman and Burge Reference Coleman and Burge2021; Sherman et al. Reference Sherman, Heupel, Moore, Chin and Simpfendorfer2020a). Symbiotic interactions have also been observed using underwater video systems, including the use of cleaning stations (Grutter et al. Reference Grutter, Murphy and Choat2003), scraping and chafing behaviours (Grossman and Sazima Reference Grossman and Sazima2009; Williams et al. Reference Williams, Anstett, Muñoz, Chisholm, Fallows, Green, Higuera Rivas, Skomal, Winton and Hammerschlag2022), and the attachment of remoras (Flammang et al. Reference Flammang, Marras, Anderson, Lehmkuhl, Mukherjee, Cade and Goldbogen2020). Baited remote underwater video systems (BRUVS) are increasingly used for documenting the relative abundance of sharks (MacNeil et al. Reference MacNeil, Chapman, Heupel, Simpfendorfer, Heithaus, Meekan and Cinner2020) and bony fishes (Burge et al. Reference Burge, Atack, Andrews, Binder, Hart, Wood and Jagannathan2012). However, even in complex habitats, the use of bait can influence the behaviour of fishes (Coghlan et al. Reference Coghlan, McLean, Harvey and Langlois2017; Harvey et al. Reference Harvey, Cappo, Butler, Hall and Kendrick2007). More recently, underwater videos have been used to identify individuals to increase count accuracy (i.e., Sherman et al. Reference Sherman, Chin, Heupel and Simpfendorfer2018), estimate movement between adjacent video stations (i.e., Benjamins et al. Reference Benjamins, Fox, Last and McCarty2018), and resight individuals over longer periods (i.e., Gore et al. Reference Gore, Ormond, Clarke, Kohler, Millar and Brooks2020).
Most underwater video surveys rely on short-term manual deployments of camera systems of approximately one hour per deployment. Recent advances in technology have reduced the size and cost of units while increasing resolution and battery life, enabling longer-term deployments of underwater camera systems that can film continuously for weeks at a time or longer (e.g., 12-h long deployments, Boom et al Reference Boom, Huang, Beyan, Spampinato, Palazzo, He and Fisher2012, Reference Boom, He, Palazzo, Huang, Beyan, Chou and Fisher2014). This has allowed the observation of new behaviours previously inaccessible to most scientists, including observing hagfish defensive behaviour in the deep sea (Zintzen et al. Reference Zintzen, Roberts, Anderson, Stewart, Struthers and Harvey2011), multi-species behaviourally-mediated indirect interactions (Coleman and Burge Reference Coleman and Burge2021), and community structure (e.g., elasmobranch assemblages in shallow seas; Currey-Randall et al. Reference Currey-Randall, Cappo, Simpfendorfer, Farabaugh and Heupel2020; Simpfendorfer et al. Reference Simpfendorfer2023).
Coral City Camera is a collaborative art and science installation piece. It consists of an underwater video camera situated in waters off the coast of Miami, FL (USA) at approximately 3 m depth and continuously streams live video of a shallow coastal coral reef for free public viewing (viewable at CoralCityCamera.com). Coral City Camera is maintained by Coral Morphologic and has been in operation since 2020. No one person is constantly watching, but at any given time, someone from the community very likely is. It is one of a growing number of free live streams set up in ecologically interesting areas globally, including a watering hole in Kenya (https://explore.org/livecams/african-wildlife/african-animal-lookout-camera), a major flyway for migratory birds in Wisconsin, USA (https://explore.org/livecams/raptor-resource-project/mississippi-river-flyway-cam), and polar bear habitat in Churchill, Canada (https://explore.org/livecams/three-bears/polar-bear-cam).
One of the goals of the Coral City Camera project was to stimulate an interest in natural history and community science, and to make Miami residents aware of and interested in the biodiversity immediately off their coast. The associated Facebook group includes, as of July 1, 2023, over 500 members, comprising scientists, SCUBA divers, members of the South Florida community, and people from at least 93 countries all over the world (Appendix 1), who have reported that they simply enjoy having a livestream of a coral reef to view during their day (Pers. Comm. to author K.M.). Demographic data provided by YouTube suggests that a majority of users are 45 years or older (Appendix 1). Demographic data provided by Facebook suggests that most viewers are based in the United States, and within the United States, many viewers are within Florida, but there are viewers from around the world.
Volunteer observers can record short clips of wildlife swimming by the camera, which, as of July 2023, have included over 200 species of sharks, squid, sea turtles, manatees, and coral reef fishes (Appendix 2). These clips have been archived (privately but accessible by request) by the Coral City Camera team, and some are also shared on Coral City Camera’s social media pages. As of July 2023, the livestream has more than 2.3 million views, averaging 32 min per view, and 13,609 people have subscribed to the YouTube channel. Though far less widely used than the YouTube livestream or the Facebook community, some users also communicate via a dedicated Discord channel.
An unexpected but delightful emergent property of Coral City Camera’s public engagement is the recognition of identifiable individual animals who are resident on the reef. The most popular of these is a yellowtail parrotfish, Sparisoma rubripinne (Valenciennes, 1840), named “Ramón” by the Coral City Camera community, who has inspired fan-created artwork and even merchandise (Figure 1).
Examples of Ramón’s celebrity and fan base, including an embroidery by Dr. Elyse Mowle, sister of author K.M. (image courtesy Elyse Mowle), (B) a mug, and (C) a t-shirt available on the Coral City Camera store (images courtesy Coral City Camera/coauthor C.F.).
Community science (often referred to as citizen science or participatory science) refers to projects or activities that engage both professional scientists and those with no formal background in STEM (Dickinson et al. Reference Dickinson, Zuckerberg and Bonter2010; Kostadinova Reference Kostadinova2011). These engagements encourage crowdsourcing, enabling scientists to share the workload with volunteers and inviting non-scientists to contribute valid and novel contributions to the field (Bonney et al. Reference Bonney, Cooper, Dickinson, Kelling, Phillips, Rosenberg and Shirk2009; Kawrykow et al. Reference Kawrykow, Roumanis, Kam, Kwak, Leung, Wu, Zarour, players, Sarmenta, Blanchette and Waldispühl2012; Khatib et al. Reference Khatib, Cooper, Tyka, Xu, Makedon, Popović, Baker and Players2011a, Reference Khatib, DiMaio, Group, Group, Cooper, Kazmierczyk, Gilski, Krzywda, Zabranska, Pichova, Thompson, Popović, Jaskolski and Baker2011b). While such projects have traditionally referred to ecological, biological, or environmental projects with volunteers local to a specific region (Silvertown Reference Silvertown2009), a growing number are now conducted entirely online with global participation throughout a range of disciplines (i.e., the Zooniverse, www.zooniverse.org,Reed et al. Reference Reed, Rodriguez and Rickhoff2012).
In this article, we show how a novel use of the underwater camera provided by Coral City Camera has revealed a unique natural history observation: the possible long-term association of a sharksucker Echeneis naucrates Linnaeus 1758 with a male yellowtail parrotfish Sparisoma rubripinne.
Methods
Real-time observation of the livestreaming webcam by an enthusiastic online community of largely non-experts resulted in the observation that Ramón, the yellowtail parrotfish, frequently had a sharksucker attached, apparently causing some discomfort (i.e., Ramón was regularly observed scraping against rocks and benthos, presumably to dislodge the sharksucker, and occasionally was seen with scraping injuries presumably caused by this behaviour). Because the sharksucker gave Ramón the appearance of wearing a hat, the sharksucker was dubbed “Fedora” by the Coral City Camera community. The community observed and discussed this interaction, which brought it to the attention of scientists.
To confirm that Ramón, the parrotfish, and Fedora, the sharksucker, were the same individuals over time, author K.M. collated photos of whenever the pair were cited, and author C.S.S. took relative measurements from these. Through a period of 283 days (June 11, 2020 to March 21, 2021, inclusive), this generated 129 screenshots of Ramón and Fedora together. From November 9, 2020, to the last day observed, there was not a full week where Ramón and Fedora were not observed together. Screenshots were taken at times when both fish were lateral to the camera and as straight as possible. Screenshots were not necessarily all the same dimensions, but as it was the relative size of the sharksucker to the parrotfish that was measured, not absolute size, this should not affect the results. Each screenshot of the two fish was imported into Adobe Photoshop for measurement. Ramón’s fork length (FL) was measured, and Fedora’s total length (TL) was measured in pixels using the measurement tool (Figure 2). Each image was assigned a quality (good n = 53, okay n = 52, bad n = 24). Good quality images were those where the pair was side-on to the camera and straight; okay images included the pair at an angle or slightly curved; and bad quality images included the pair facing the camera or too far away for clear measurements. Bad quality images were not included in analyses, leaving 105 usable images from 94 unique days. Including some images with slight curvature only changed the R2 value from 0.96 to 0.94.
A representative example measurement of Ramón and Fedora, image taken from a Coral City Camera livestream used with permission of coral city camera (coauthor C.F.).
We acknowledge that this measurement method is not ideal. However, as this was an opportunistic study based on the observations of the public, we used the best method available. We did not have stereo-video to accurately measure, nor did we have any in-frame reference measurements, as the camera had moved locations throughout the study period. Instead, we used relative size (measured in pixels). We feel this method is suitable based on the high level of precision (high R2 value) regardless of the fishes’ distance from the camera.
Results
An analysis of archived Coral City Camera footage found that Ramón, the yellowtail parrotfish, had Fedora, the remora, attached on at least 94 unique days from summer 2020 to spring 2021 (Figure 3). When plotted against each other, the relative size of the two fish remained constant with an R 2 value of 0.96 (Figure 4). When analysed separately by year, the R 2 values are 0.93 (2020) and 0.98 (2021), likely because Fedora was smaller and growing faster earlier in this interaction. With this high level of correlation and the lack of observations of remoras attached to other teleost fishes in the videos taken at this location, we feel confident that this pair of fish is likely the same individuals throughout the observation period.
A photo of Ramón and Fedora from November 2020, and one from March 2021.
The relationship between observed Ramón fork length (FL; pixels) and observed Fedora total length (TL; pixels) from 105 screenshots (r2 = 0.9555; y = 0.4731x –7.542).
Additionally, on many occasions, Ramón was recorded attempting to dislodge Fedora. Rapid turns, flips, and scraping against coral and the seafloor were observed. On several occasions, Fedora was observed closely following and reattaching himself to Ramón, providing further evidence that we are observing the same remora on the same parrotfish repeatedly. (Supplementary Files: Video clips of Ramón dislodging Fedora and Fedora reattaching to Ramón.)
After Fedora was successfully dislodged, Ramón had scars visible for several months and was repeatedly resighted on Coral City Camera until February 2022, nearly a year later. He has not been seen since, and it is unknown whether he was outcompeted as the dominant male yellowtail in the local reef system and moved on, or if he died.
Discussion
The observation of a potential long-term association between Ramón, the yellowtail parrotfish, and Fedora, the sharksucker, is ecologically interesting in its own right, as, to the best of our knowledge, it is the longest interaction between a sharksucker and any host recorded in the scientific literature. However, this observation was only possible due to a combination of technological advances that allowed for Coral City Camera’s continuous underwater livestream and the dedicated community of volunteer community scientists who watch and record observations every day.
The engagement of Coral City Camera’s virtual community highlights the role that community scientists can have in contributing to our understanding of biology and ecology. One of the most time-consuming aspects of underwater video-based research is not deploying the cameras but analysing the enormous volume of resulting video footage. For example, a recent global-scale analysis of coral reef sharks included nearly 3 years of raw video files, all of which were manually viewed and annotated for analysis (Simpfendorfer et al. Reference Simpfendorfer2023). For Coral City Camera’s highly engaged community, the observation and analysis time is distributed over hundreds of enthusiastic participants, which means that novel features and emergent properties can be more easily identified than could be done using traditional methods with few observers each analysing thousands of hours of video recordings. We note here that the observation of a long-term association between Ramón and Fedora was not the result of hypothesis-based experimentation but a novel observation resulting from community who viewed the footage, with no guidance or training from participating scientists. Once the observation was brought to our attention (via community members flagging another observation of Ramón with a remora on a daily log of interesting observations on the livestream), the scientific team performed the analysis, and community members were not involved in data analysis.
Analyses based on observations from community members have been responsible for discoveries in other fields, including monitoring arthropod-based disease vectors (Cull Reference Cull2021), exploring predator–prey dynamics (Hernandez et al. Reference Hernandez, Masonick and Weirauch2019) and tracking invasive species (Maistrello et al. Reference Maistrello, Dioli, Dutto, Volani, Pasquali and Gilioli2018). In these cases, and in our own example, a community of interested observers was able to approach large data sets through a sense of gameplay and genuine interest. This “gamification” of science has been shown in other systems to be important in getting across complex scientific information and in community building (Drew et al Reference Drew, Sardelis and Davis‐Berg2017; Hinde et al Reference Hinde, Amorim, Brokaw, Burt, Casillas, Chen and Anderson2021). Projects like this can be a conduit towards democratising science and promoting universal, equitable access to scientific data and information.
Underwater cameras can be used to document a wide variety of ecological interactions and biological observations. They can easily be used to quantify presence/absence or population trends (MaxN, Campbell et al., Reference Campbell, Salisbury, Caillouet, Driggers and Kilfoil2018) of particular species, and if this changes seasonally or as a response to stressors or new management regimes (i.e., are there more of one species in the summer than in the fall, are there more of that species in a fully protected area compared to a nearby partially protected area). A series of cameras deployed throughout the range of a species can allow for comparison of that species’ behaviour across a range of environmental conditions or management regimes, and they will undoubtedly allow for additional unexpected observations that we cannot even predict. Future research should expound on additional possible uses of these incredible tools and the passionate online communities of observers that they attract.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S0025315425100416.
Acknowledgements
The authors would like to thank the Coral City Camera community of fans and volunteers for making this fascinating ecological observation possible. Images and video clips are freely available online through the Coral City Camera YouTube page, and the specific images and analysis used here will be shared upon reasonable request to the corresponding author. We would also like to thank Dr. Tom Mowle, father of author K.M., for his suggestion of the name Fedora, which was excitedly accepted by the Coral City Camera community.
Author contributions
D.S.S. conceived the project and took the lead on writing and editing. C.F. and K.M. provided key data and context through Coral City Camera, which C.F. runs. C.S.S. took the lead on analysis and, along with author EB, provided context on the use of underwater cameras. B.E.F. provided expertise on remora biology and ecology, and authors J.D. and M.C.M. provided expertise on community science and online communities. All authors contributed to writing and editing.
Funding
The authors have no funding to declare.
Competing interests
The authors have no conflicts of interest to declare.
