Biology subjects

Dainys, J.

Publications and source records attributed to Dainys, J..

5 recordsLinked to original sources

bioRxiv2022

A machine learning based image classification method to estimate fish sizes from images without a specified reference object

Most fish populations around the world are unassessed and their status is unknown. Size based methods could provide fast and transparent assessments, but they require information on fish sizes. Citizen science programs, social media and smart phone applications generate millions of georeferenced fish images globally. Machine learning based fish species and size identification could help turn these images into valuable data for population status assessments. We present a machine learning, image classification based method to identify fish size classes from photos of anglers holding fish. To train the model we group images into ten 5-10 cm size classes, similar to classes used in underwater visual fish surveys. The model was trained using 2602 images from angler citizen science platforms MyCatch and FishSizeProject. Although the number of images was limited, the model achieved an overall accuracy of ~50%. Importantly, the misidentification of size classes was consistent across 20 separate model training rounds, each conducted with an independent, random allocation of images for training and test datasets. Our method suggests that photo based fish size class identification is feasible, and that prediction uncertainty should be incorporated into subsequent analysis, as it is done with fish ageing errors in fisheries stock assessments.

ecology↗

bioRxiv2022

Status and future perspectives for pikeperch (Sander lucioperca) stocks in Europe

Pikeperch (Sander lucioperca) is a European fresh and brackish water piscivorous fish, important as both a key predator and a valuable commercial and recreational fisheries species. There are concerns that some stocks are depleted due to overfishing and environmental changes. We review data collection and population assessments currently used for nine pikeperch stocks across six European countries and apply a unified assessment framework to evaluate population status and trends. For this we first standardised commercial, scientific, and recreational catch-per-unit-effort (CPUE) and catch time series and then applied Bayesian surplus production models. Our results showed that three stocks (including two in the Baltic Sea) were strongly depleted, with estimated biomasses considerably lower than the biomass at maximum sustainable yield (Bmsy). Other stocks were either close or higher than their estimated Bmsy. Looking at the trends, we find that four stocks (Lake Oulujarvi, Kvadofjarden, Lake Peipsi and Lipno) showed increasing biomass trends and two (Curonian Lagoon, Galtfjarden) had a strong decline in biomass. In most cases the stocks with clear signs of recovery were also those for which strong management strategies have been implemented. We find that, despite pikeperch being one of the most valuable inland fisheries, formalised stock assessments and regular surveys remain rare. Importantly, although most stocks are strongly targeted by recreational fishing, estimates of recreational catch are highly uncertain. We conclude that data limited stock assessment methods are useful for assessing fish population status and highlight an urgent need to improve pikeperch scientific monitoring and assessment of recreational catches.

ecology↗

bioRxiv2022

High-resolution app data reveal sustained increases in recreational fishing effort in Europe during and after COVID-19 lockdowns

It is manifest that COVID-19 lockdowns extensively impacted human interactions with natural ecosystems. One example is recreational fishing, an activity which involves nearly 1 in 10 people in developed countries. Fishing licence sales and direct observations at popular angling locations suggest that recreational fishing effort increased substantially during lockdowns. However, the extent and duration of this increase remain largely unknown due to a lack of objective data. We used four years (2018 to 2021) of anonymous, high-resolution data from a personal fish-finder device to explore the impact of COVID-19 lockdowns on recreational fishing effort in four European countries (Lithuania, the Czech Republic, Denmark, and Germany). We show that device use and, by extension, angling effort increased 1.2-3.8 fold during March-May 2020 and remained elevated even at the end of 2021 in all countries except Denmark. Fishing during the first lockdown also became more frequent during weekdays. Statistical models with the full set of fixed (weekdays, lockdown, population) and random (season, year, administrative unit) factors typically explained 50-70% of the variation, suggesting that device use and angling effort were relatively consistent and predictable through space and time. Our study demonstrates that recreational fishing behaviour can change substantially and rapidly in response to societal shifts, with profound ecological, human well-being and economic implications. We also show the potential of angler devices and smartphone applications to supply data for high-resolution fishing effort analysis and encourage more extensive science and industry collaborations to take advantage of this information. Significance statementRecreational fishing is a popular and widespread activity with ecological, social and economic impacts, though problematic to assess and manage due to a paucity of information regarding effort and catch. Here, we use high-resolution data from a personal angler sonar device to show how the COVID-19 pandemic changed angler behaviour and fishing effort across Europe. We demonstrate that angling effort doubled and remained higher at the end of 2021 than before the first lockdowns. Such rapid and profound changes could have significant consequences for aquatic ecosystems, possibly requiring new management approaches. We encourage the adoption of novel data from angler devices, citizen science, and more active science-industry collaborations to improve recreational fishing assessment and management.

ecology↗

bioRxiv2022

A scalable open-source framework for machine learning based image collection, annotation and classification: a case study for automatic fish species identification

Citizen science platforms, social media and multiple smart phone applications enable collection of large amounts of georeferenced images. This provides a huge opportunity in biodiversity and ecological research, but also creates challenges for efficient data handling and processing. Recreational and small-scale fisheries is one of the fields that could be revolutionised by efficient, widely accessible and machine learning based processing of georeferenced images. The majority of non-commercial inland and coastal fisheries are considered data poor and are rarely assessed, yet they provide multiple societal benefits and can have large ecological impacts. Given that large quantities of fish observations and images are being collected by fishers every day, artificial intelligence (AI) and computer vision applications offer a great opportunity to improve data collection, automate analyses and inform management. Yet, to date, many AI image analysis applications in fisheries are focused on the commercial sector and are not publicly available for community use. In this study we present an open-source modular framework for large scale image storage, handling, annotation and automatic classification, using cost- and labour-efficient methodologies. The tool is based on TensorFlow Lite Model Maker library and includes data augmentation and transfer learning techniques, applied to different convolutional neural network models. We demonstrate the implementation of this framework in an example case study for automatic fish species identification from images taken through a recreational fishing smartphone application. The framework presented here is highly customisable for further advancement and community based image collection and annotation.

ecology↗

bioRxiv2022

Impacts of recreational angling on fish population recovery after a commercial fishing ban

It is often assumed that recreational fishing has negligible impact on fish stocks compared to commercial fishing. Yet, for inland water bodies in densely populated areas, this is unlikely to be true. In this study we demonstrate remarkably variable stock recovery rates among different fish species with similar life histories in a large productive inland freshwater ecosystem (Kaunas Reservoir, Lithuania), where all commercial fishing has been banned since 2013. We conducted over 900 surveys of recreational anglers during a period of four years (2016 to 2021) to assess recreational fishing catches. These surveys are combined with drone and fishfinder device-based assessment of recreational fishing effort. Fish population recovery rates were assessed using standardised catch per unit effort time series. We show that recreational fishing is having a major impact in retarding the recovery of predatory species, such as pikeperch and perch. In contrast, recovery of roach, rarely caught by anglers, has been remarkably rapid and the species is now dominating the ecosystem. Our study demonstrates that recreational fishing can have strong impacts on some fish species, alter relative species composition and potentially change ecosystem state and dynamics.

ecology↗