Valeriya Komyakova

Large language models (LLMs) can ease the work of screening titles and abstracts for systematic reviews, but obtaining reliable results requires researchers to make practical choices about which LLMs to use, how to combine their scores into a ranking, and how far down that ranking to read. We aimed to identify a general-purpose workflow that screens accurately, minimises human review effort, and generalises across environmental literature corpora. We ran an ensemble of five open-source LLMs across ten human-annotated systematic reviews from the field of ecology and environmental science spanning 19,155 studies. We then asked: (1) how well an ensemble of LLMs ranks relevant papers above irrelevant ones, and (2) where a human reviewer should stop working down that ranked list. A four-LLM ensemble chosen without any labels came close, on every review, to the best ranking achievable with that review’s annotations (mean Average Precision 0.66 versus 0.68). We tested different rules for when to stop human review, finding that a single SAFE stopping rule chosen in advance recovered ≥ 95% of relevant records on all ten reviews while requiring a human to screen 56% of the corpus on average. Against the established open-source active-learning tool ASReview , this label-free ranking reached the same recall target at lower human workload on seven of the ten reviews (39% versus 42% of the corpus on average). The paper offers a complete workflow that can be adopted for new, unlabelled reviews, using open-source LLMs small enough to run on a high-end consumer laptop, and we provide it as an open-source R package.

The degradation of water quality is among the most pervasive, complex, and challenging problems that impact the world’s freshwater ecosystems and their biodiversity. Contamination pathways vary globally and involve a diverse array of pollutants that include simple (e.g., road salt) to more complex (e.g., pharmaceuticals) chemicals, as well as non-natural sources of heat, light, noise, silt, plastics, and also pathogens from wastewater or livestock. These pollutants can cause significant harm on their own, in mixtures or in combination with other stressors. Over geographically extensive areas affecting most river systems and many standing waters, pollutants have been released either legally, illegally or incidentally over inter-decadal to centennial timescales. Some pollutants and pathogens also affect people through morbidity, mortality, and economic loss, so that there is synergy in protecting freshwater biodiversity against pollution while delivering human well-being through the global Sustainable Development, ecosystem services, and climate change agenda. Existing approaches aimed at reducing water pollution are insufficient and require a previously unmatched step-change in approach to address the major impacts from both conventional and emerging pollutants, such as pharmaceuticals, pesticides and microplastics. Success stories on measures that address freshwater pollution are relatively limited such that available examples are often “works in progress” or represent partial success. Here, we review both challenges and potential solutions to freshwater pollution, focusing on case studies that range across (i) informing action through monitoring and understanding pollutant trends and ecosystem impacts; (ii) regulation and enforcement nationally and internationally to reduce pollutant loads, substitute harmful substances, and ensure that polluters pay; (iii) improving the management and remediation of pollutants at source, in situ, and across catchments, including scalable, nature-based solutions, and (iv) addressing the systemic drivers of pollution through education, incentives, and market mechanisms. Significant challenges remain, but water quality restoration in freshwater ecosystems will bring manifold benefits for freshwater biodiversity and human well-being at all scales from local to global.

Plastics are now ubiquitous in the environment and have been studied in wildlife and in ecosystems for more than 50 years. Measurement of size, shape and colour data for individual fragments of plastic is labour‐intensive, unreliable and prone to observer bias, particularly when it comes to assessment of colour, which relies on arbitrary and inconsistently defined colour categorisations. There is a clear need for a standard method for data collection on plastic pollution, particularly one that can be readily automated given the number of samples involved. This study describes a new method for standardised photography of marine plastics in the 1–100 mm size range (meso‐ and macro‐plastics), including colour correction to account for any image‐to‐image variation in lighting that may impact colour reproduction or apparent brightness. Automated image analysis is then applied to detect individual fragments of plastic for quantitative measurement of size, shape, and colour. The method was tested on 3793 fragments of debris ingested by Flesh‐footed Shearwaters (Ardenna carneipes) on Lord Howe Island, Australia, and compare results from photos taken in two separate locations using different equipment. Photos were acquired of up to 250 fragments at a time with a spatial resolution of 70 μm/pixel and were colour‐corrected using a reference chart to ensure accurate reproduction of colour. The automated image analysis pipeline was found to have a 98% success rate at detecting fragments, and the different size and shape parameters that can be outputted by the pipeline were compared in terms of usefulness. The evidence shown in this study should strongly encourage the uptake of this method for cataloguing macro‐scale plastic pollution, as it provides substantially higher quality data with accurate, reliable measurements of size, shape and colour for individual plastics that can be readily compared between disparate datasets.

Artificial reefs have many applications but are best known for their deployments to enhance recreational fisheries by creating new habitat in areas where natural reef is otherwise limited. The expectation is that fish assemblages will take up residence on artificial reefs and that these assemblages will become at least similar, if not more diverse and abundant, to those on natural reefs. Although designed, purpose-built artificial reefs are becoming more widely used in support of recreational fisheries and many of the historic issues have been resolved, conservation practitioners and managers still face challenges as to the type, number, and arrangement of structures and where to deploy them to maximize benefits and minimize risks. The ecological literature was reviewed to develop and enhance contemporary principles of artificial reef best practices for utilization. Our review identified optimal shapes, vertical relief, void spaces, and unit arrangements for increasing volumes and diversity of catch to recreational fishers and we provide a tool for identifying the least constrained areas for artificial reef deployment. We suggest; (a) monitoring of noncatch motivators in combination with quantitative indicators of the fishing activity (e.g., catch rate and effort) will provide the best understanding of success or failure of an artificial reef deployment; (b) choosing target species for informing purpose-built artificial reef designs to be reef-associated, demersal, philopatric, territorial, and obligatory reef species that are desired by local recreational fishers; and (c) considering the ecosystem services provided by artificial reefs beyond those associated with recreational fishing.

While there is increasing evidence for habitat specialization in coral reef fishes, the extent to which different corals support different fish communities is not well understood. Here we quantitatively assess the relative importance of different coral species in structuring fish communities and evaluate whether sampling scale and coral colony size affect the perceived strength of fish-habitat relationships. Fish communities present on colonies of eight coral species (Porites cylindrica, Echinopora horrida, Hydnophora rigida, Stylophora pistillata, Seriatopora hystrix, Acropora formosa, A. tenuis and A. millepora) were examined in the Lizard Island lagoon, Great Barrier Reef, Australia. Additionally, the differences in fish communities supported by three coral species (P. cylindrica, E. horrida, H. rigida) were investigated at three spatial scales of sampling (2x2 m, 1x1 m, 0.5x0.5 m). Substantial differences in fish communities were observed across the different coral species, with E. horrida and H. rigida supporting the most fish species and individuals. Coral species explained more of the variability in fish species richness (20.9–53.6%), than in fish abundance (0–15%). Most coral species supported distinctive fish communities, with dissimilarities ranging from 50 to 90%. For three focal coral species, a greater amount of total variation in fish species richness and fish abundance was evident at a larger scale of sampling. Together, these results indicate that the structure of reef fish communities is finely tuned to coral species. Loss of preferred coral species could have profound effects on reef fish biodiversity, potentially more so than would be predicted on the basis of declining coral cover alone.