Artificial reefs have been constructed and deployed in over 50 countries around the world to enhance the productivity of aquatic habitats and fishing experiences. In April 2013, two purpose-built concrete artificial reefs were deployed in Geographe Bay, Western Australia to provide additional fish habitat and increase upwelling and thus enhance recreational fishing opportunities. Due to the relatively high cost of planning, purchasing and deploying these structures, it is important to understand spatial and temporal usage of the reef by fish assemblages, in order to determine the extent to which fishing opportunities are actually enhanced. One potential method to reduce monitoring costs is to utilise volunteers from the general public to collect data, i.e. citizen science. The overall objective of this project was to determine whether recreational fishers, through a citizen science program, could potentially provide an effective means for monitoring artificial reefs.
Following a recruitment drive and underwater camera trial, a small number of recreational fishers were provided underwater drop video cameras and asked to record footage on the Bunbury and Dunsborough artificial reefs and also nearby natural reefs. Unfortunately, only very limited amounts of data (~1 hour) were received due to a combination of a lack of participation/engagement, unseasonal weather and the short timeframe of the project. However, enough videos were received to undertake a preliminary analysis of the differences in the characteristics of the fish faunas of the two types of reef, i.e. natural vs artificial. The results indicate that artificial reefs may potentially harbour greater numbers of species and a larger total maximum abundance. Multivariate statistical analyses did not detect any differences in the fish faunal compositions between natural and artificial reefs, which were likely due to the dominance of King Wrasse (Coris auricularis) on both reefs. Furthermore, large amount of variability between replicates caused by the differences in recording time, which, although standardised, was still an artefact in the resultant data and may have masked any ‘real’ differences among the reef types.
Given the limited data provided by the above monitoring program, a critical review of the citizen science components of the project was completed and a set of key recommendations for use in future projects using recreational fishers to collect video footage provided. These included: (i) enhancing the methods of contacting and recruiting volunteers to include social media and encouraging communication among participants, (ii) using a GoPro camera mounted on Baited Remote Underwater Video (BRUV) systems to ensure that the footage collected is of high quality (resolution), (iii) providing simplified and consistent instructions and (iv) ensuring regular communication and engagement with volunteers. A global literature review on citizen science and the benefits and limitations using this type of project for research purposes was also undertaken together with a brief description of such programs that have been or are being conducted in aquatic environments in WA.
To test the suggestions the BRUV systems with a GoPro camera constructed from readily available materials could be deployed by recreational fishers as a citizen science artificial reef monitoring tool, Ecotone Consulting built one of these units and deployed it randomly around the Dunsborough artificial reef. The resultant footage was found to be of much higher quality than that obtained using the drop camera. The GoPro videos were analysed to determine whether there was a difference in fish assemblages between artificial reef modules and the surrounding area, i.e. videos where the camera directly faced one or more of the artificial reef modules were vs those were no modules could be observed in the camera’s field of view. The results demonstrated that mean number of species and the mean number of benthic and epibenthic species were greater on footage recorded when the camera faced the modules. There was also a difference in the faunal composition, with 52.63% more recreational target species being found on artificial reefs than in surrounding areas. It was also concluded that the BRUV technology could be used, by citizen scientists, to monitor the fish faunas of artificial reefs.
Another potential method to reduce the cost of monitoring programs for the fish faunas of artificial reefs is to use citizens to analyse the footage as part of their studies. However, if such a program was to proceed using volunteers with limited experience, i.e. undergraduate students, it is important to ensure that the fish data extracted from the video is reliable. Thus, to investigate the impact of observer bias, the BRUV footage collected from the Dunsborough artificial reef was analysed by having multiple observers, with similar levels of marine science training and recreational fishing experience. It was found that whilst observers recorded similar mean numbers of species and total abundance counts, significant differences in species composition were detected. This was due to observers misidentifying members of particular families, i.e. the leatherjackets and trevallies. This suggests that, while observers with limited experience may be able to detect common species, misidentification of less common and/or less distinct species can lead to significant variation in the data due to observer bias. Therefore, if university students are to be used as part of any citizen science monitoring project, it is recommended that participants should receive additional training in species identification, and be subjected to an initial trial where their results are compared to that of a more experienced observer until a minimum similarity of 90% is consistently recorded.
The provision of BRUV footage from the Bunbury artificial reef by Ecotone Consulting allowed statistical analysis of fish faunas on both reefs to be undertaken to identify what level of information could be obtained using the BRUVs. Analysis of the data found both the mean number of species and total abundance were greater at the Dunsborough artificial reef and that there was also a significant difference in species composition. While, more data are required to provide a more accurate picture of any differences, this does highlight the fact that the BRUV footage can be employed to test for differences in the fish faunas of these reefs and possibly also nearby natural reef or other areas.
It is concluded that recreational fishers did not provide an effective means for monitoring artificial reefs during this project. This result, however, is a consequence of a lack of data stemming from an absence of volunteer engagement in a limited pilot project with a short time frame and unseasonal weather. This does not exclude the potential for using citizen scientists to monitor artificial reefs, following some changes in the methodology, technology and management of citizen science protocols, and thus it is possible to utilise recreational fishers as an effective means for monitoring artificial reefs. This project was subjected to restrictive and limiting factors but more importantly, discovered ways to overcome these issues by provided key recommendations on technology, methodologies and community engagement that should be followed to increase the effectiveness of using recreational fishers to provide sound scientific information in the future and these have been actively employed in a new citizen science program for monitoring the fish fauna of the Bunbury and Dunsborough artificial reefs Reef Vision.