Samenvatting

Wageningen Marine Research (WMR) conducts a variety of monitoring programs in Dutch waters. These monitoring programs are conducted to gain knowledge of trends, abundance and diversity of fish species in the Dutch rivers and lakes. Many monitoring programs pay special attention to diadromous fish which are only temporarily present in certain waterbodies. Mostly, these monitoring programs are conducted by fishermen which are contracted by WMR. Most of these fisherman own unique fishing rights for specific waterbodies. Since they are contracted to do research, no commercial fishing activities are allowed during monitoring and regulations defined in the permits are to be followed. Violations of these regulations and exclusion of fisherman cause a loss of monitoring locations since other fisherman do not own the fishing rights. Therefore, alternatives are needed to minimize risk of irregular trend monitoring data. However, further research is needed to test whether alternatives provide reliable trend monitoring data and whether data are sufficient to answer research questions. The goal of this project was to determine if remote camera monitoring is a suitable alternative to conventional trap monitoring. Camera boxes could remove the need for catching and handling fish and therefore prevent the risk of poaching.
A camera box was deployed at an existing monitoring location. Data was collected and analysed from 5th of April until 14th of May. The camera box was connected to a conventional monitoring trap. To determine accuracy, an additional collection net was attached at the end of the camera box to obtain data on actual catches for comparison. By determining species and abundance with obtained video recordings and comparing these to actual catches data, the accuracy and the pros and cons of camera monitoring were determined. Therefore, the main research question is:
What is the effectiveness of video monitoring compared to actual catches of conventional trap monitoring?
The results of the field experiments revealed that camera monitoring shows potential for counting individual fish. With the used set-up, a counting accuracy of 91% was reached. Camera monitoring also showed potential for classifying fish but this is highly dependent on the species. With the used setup, flatfish were difficult to classify. Certain species of round fish with similar physiological traits were also difficult to distinguish. Distinguishing silver eel from yellow eel using the video recordings did not seem possible. The overlap in similar physiological traits, and the lack of colour in the recordings, made it impossible to distinguish the two with certainty.
Camera monitoring shows potential as an alternative monitoring method but will require further research for improvements and fine tuning of this method.