A roadmap for biomonitoring in the 21st century: Merging methods into metrics via ecological networks

Cuff, Jordan P.; Deivarajan Suresh, Mukilan; Dopson, Matthew E.G.; Hawthorne, Ben S.J.; Howells, Thomas; Kitson, James J.N.; Miller, Kyle A.; Xin, Tong; Evans, Darren M.

doi:10.1016/bs.aecr.2023.09.002

Cited by 3 publications

(1 citation statement)

References 221 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…TensorFlow, PyTorch and Caret), online services (e.g. Edge Impulse, AutoML and AnIML), review articles (Cuff et al, 2023;Høye et al, 2021;Ryo, 2023;Tuia et al, 2022;Weinstein, 2018) and training programmes (Cole et al, 2023) to assist with the training of custom models, there is a notable lack of method guides to assist ecologists with the unique challenges of computer vision classification for ecological and taxonomic purposes specifically.…”

Section: Introductionmentioning

confidence: 99%

A gentle introduction to computer vision‐based specimen classification in ecological datasets

Blair,

Gaynor,

Palmer

et al. 2024

Journal of Animal Ecology

View full text Add to dashboard Cite

Classifying specimens is a critical component of ecological research, biodiversity monitoring and conservation. However, manual classification can be prohibitively time‐consuming and expensive, limiting how much data a project can afford to process. Computer vision, a form of machine learning, can help overcome these problems by rapidly, automatically and accurately classifying images of specimens. Given the diversity of animal species and contexts in which images are captured, there is no universal classifier for all species and use cases. As such, ecologists often need to train their own models. While numerous software programs exist to support this process, ecologists need a fundamental understanding of how computer vision works to select appropriate model workflows based on their specific use case, data types, computing resources and desired performance capabilities. Ecologists may also face characteristic quirks of ecological datasets, such as long‐tail distributions, ‘unknown’ species, similarity between species and polymorphism within species, which impact the efficacy of computer vision. Despite growing interest in computer vision for ecology, there are few resources available to help ecologists face the challenges they are likely to encounter. Here, we present a gentle introduction for species classification using computer vision. In this manuscript and associated GitHub repository, we demonstrate how to prepare training data, basic model training procedures, and methods for model evaluation and selection. Throughout, we explore specific considerations ecologists should make when training classification models, such as data domains, feature extractors and class imbalances. With these basics, ecologists can adjust their workflows to achieve research goals and/or account for uncertainty in downstream analysis. Our goal is to provide guidance for ecologists for getting started in or improving their use of machine learning for visual classification tasks.

show abstract

Section: Introductionmentioning

confidence: 99%

A gentle introduction to computer vision‐based specimen classification in ecological datasets

Blair,

Gaynor,

Palmer

et al. 2024

Journal of Animal Ecology

View full text Add to dashboard Cite

show abstract

Metabarcoding advances agricultural invertebrate biomonitoring by enhancing resolution, increasing throughput and facilitating network inference

Hawthorne,

Cuff,

Collins

et al. 2024

Agri and Forest Entomology

View full text Add to dashboard Cite

Biomonitoring of agriculturally important insects is increasingly vital given our need to understand: (a) the severity of impacts by pests and pathogens on crop yield and health and (b) the impact of environmental change and land management on insects, in line with sustainable development and global conservation targets. Traditional entomological traps remain an important part of the biomonitoring toolbox, but sample processing is laborious and introduces latency, and accuracy can be variable. The integration of molecular techniques such as environmental DNA and DNA metabarcoding into insect biomonitoring has gained increasing attention, but the advantages of doing so, the kind of data this can generate, and how easily and effectively molecular analyses can be integrated with the diverse types of entomological traps currently used remains relatively unclear. In this review, we examine how combining DNA metabarcoding with a range of conventional and unconventional entomological sampling techniques can advance biomonitoring in a way that is useful to researchers and practitioners. We highlight some of the key challenges and how to mitigate them, using examples of its integration with different sampling methods from the literature (e.g., interception, pitfall and sticky traps) to demonstrate efficacy and suitability. We discuss how metabarcoding data can be used to infer ecological networks, emphasizing the importance of this as a framework for understanding species interactions and ecosystem functioning for more effective and descriptive biomonitoring. Finally, future advances in biomonitoring are highlighted, alongside recommendations of best practice for researchers both new to and experienced in invertebrate biomonitoring with metabarcoding.

show abstract

Biomonitoring of biocontrol across the full annual cycle in temperate climates: Post‐harvest, winter and early‐season interaction data and methodological considerations for its collection

Cuff,

Gajski,

Michalko

et al. 2024

Agri and Forest Entomology

View full text Add to dashboard Cite

Conservation biocontrol, the regulation of crop pests by naturally occurring biocontrol agents (e.g., predators and parasitoids), is predominantly monitored throughout periods of primary crop growth when pests exert the most observable impact on yields. Pest‐focused agricultural biomonitoring often overlooks post‐harvest, winter and even early‐season biocontrol, despite the significant predator–pest interactions during these periods that profoundly affect pest abundance and, consequently, crop yields. Rapid advances in biomonitoring, particularly in the detection of predator–pest interactions that underpin biocontrol, provide an opportunity to reconsider how and when we monitor these interactions. Advances in agricultural biomonitoring must transcend methodological innovation and encompass conceptual changes in the monitoring of agricultural systems. Here, we assess existing evidence supporting the importance of periods beyond primary crop growth for biocontrol and how predator–pest interactions are likely to evolve during these periods, subsequently influencing pest population dynamics during the primary crop growth period. We advocate for a greater concerted effort to establish continuous monitoring of biocontrol interactions, particularly beyond primary crop growth periods in temperate climates. To facilitate this, we also summarise the methodological approaches that can make it possible and explore how extending sampling across the full annual cycle might impact the practicalities and outcomes of these approaches. Year‐round monitoring of biocontrol interactions, both in crops and adjacent semi‐natural habitats, will provide a previously intractable understanding of predator–pest dynamics, offering significant potential to enhance our ability to optimise and manipulate these systems. This would manifest in reduced crop yield losses, pest infestation rates and disease transmission, with concomitant long‐term financial, environmental and land‐use benefits.

show abstract

A roadmap for biomonitoring in the 21st century: Merging methods into metrics via ecological networks

Cited by 3 publications

References 221 publications

A gentle introduction to computer vision‐based specimen classification in ecological datasets

A gentle introduction to computer vision‐based specimen classification in ecological datasets

Metabarcoding advances agricultural invertebrate biomonitoring by enhancing resolution, increasing throughput and facilitating network inference

Biomonitoring of biocontrol across the full annual cycle in temperate climates: Post‐harvest, winter and early‐season interaction data and methodological considerations for its collection

Contact Info

Product

Resources

About