| 1. |
- Lafuente, Elvira, et al.
(författare)
-
Building on 150 Years of Knowledge : The Freshwater Isopod Asellus aquaticus as an Integrative Eco-Evolutionary Model System
- 2021
-
Ingår i: Frontiers in Ecology and Evolution. - : Frontiers Media SA. - 2296-701X. ; 9
-
Forskningsöversikt (refereegranskat)abstract
- Interactions between organisms and their environments are central to how biological diversity arises and how natural populations and ecosystems respond to environmental change. These interactions involve processes by which phenotypes are affected by or respond to external conditions (e.g., via phenotypic plasticity or natural selection) as well as processes by which organisms reciprocally interact with the environment (e.g., via eco-evolutionary feedbacks). Organism-environment interactions can be highly dynamic and operate on different hierarchical levels, from genes and phenotypes to populations, communities, and ecosystems. Therefore, the study of organism-environment interactions requires integrative approaches and model systems that are suitable for studies across different hierarchical levels. Here, we introduce the freshwater isopod Asellus aquaticus, a keystone species and an emerging invertebrate model system, as a prime candidate to address fundamental questions in ecology and evolution, and the interfaces therein. We review relevant fields of research that have used A. aquaticus and draft a set of specific scientific questions that can be answered using this species. Specifically, we propose that studies on A. aquaticus can help understanding (i) the influence of host-microbiome interactions on organismal and ecosystem function, (ii) the relevance of biotic interactions in ecosystem processes, and (iii) how ecological conditions and evolutionary forces facilitate phenotypic diversification.
|
|
| 2. |
- Lürig, Moritz D, et al.
(författare)
-
BioEncoder : A metric learning toolkit for comparative organismal biology
- 2024
-
Ingår i: Ecology Letters. - 1461-023X. ; 27:8
-
Tidskriftsartikel (refereegranskat)abstract
- In the realm of biological image analysis, deep learning (DL) has become a core toolkit, for example for segmentation and classification. However, conventional DL methods are challenged by large biodiversity datasets characterized by unbalanced classes and hard-to-distinguish phenotypic differences between them. Here we present BioEncoder, a user-friendly toolkit for metric learning, which overcomes these challenges by focussing on learning relationships between individual data points rather than on the separability of classes. BioEncoder is released as a Python package, created for ease of use and flexibility across diverse datasets. It features taxon-agnostic data loaders, custom augmentation options, and simple hyperparameter adjustments through text-based configuration files. The toolkit's significance lies in its potential to unlock new research avenues in biological image analysis while democratizing access to advanced deep metric learning techniques. BioEncoder focuses on the urgent need for toolkits bridging the gap between complex DL pipelines and practical applications in biological research.
|
|
| 3. |
- Lürig, Moritz D., et al.
(författare)
-
Computer Vision, Machine Learning, and the Promise of Phenomics in Ecology and Evolutionary Biology
- 2021
-
Ingår i: Frontiers in Ecology and Evolution. - : Frontiers Media SA. - 2296-701X. ; 9
-
Forskningsöversikt (refereegranskat)abstract
- For centuries, ecologists and evolutionary biologists have used images such as drawings, paintings and photographs to record and quantify the shapes and patterns of life. With the advent of digital imaging, biologists continue to collect image data at an ever-increasing rate. This immense body of data provides insight into a wide range of biological phenomena, including phenotypic diversity, population dynamics, mechanisms of divergence and adaptation, and evolutionary change. However, the rate of image acquisition frequently outpaces our capacity to manually extract meaningful information from images. Moreover, manual image analysis is low-throughput, difficult to reproduce, and typically measures only a few traits at a time. This has proven to be an impediment to the growing field of phenomics – the study of many phenotypic dimensions together. Computer vision (CV), the automated extraction and processing of information from digital images, provides the opportunity to alleviate this longstanding analytical bottleneck. In this review, we illustrate the capabilities of CV as an efficient and comprehensive method to collect phenomic data in ecological and evolutionary research. First, we briefly review phenomics, arguing that ecologists and evolutionary biologists can effectively capture phenomic-level data by taking pictures and analyzing them using CV. Next we describe the primary types of image-based data, review CV approaches for extracting them (including techniques that entail machine learning and others that do not), and identify the most common hurdles and pitfalls. Finally, we highlight recent successful implementations and promising future applications of CV in the study of phenotypes. In anticipation that CV will become a basic component of the biologist’s toolkit, our review is intended as an entry point for ecologists and evolutionary biologists that are interested in extracting phenotypic information from digital images.
|
|
| 4. |
- Lürig, Moritz D., et al.
(författare)
-
Dietary-based developmental plasticity affects juvenile survival in an aquatic detritivore
- 2021
-
Ingår i: Proceedings of the Royal Society B: Biological Sciences. - : The Royal Society. - 0962-8452 .- 1471-2954. ; 288:1945
-
Tidskriftsartikel (refereegranskat)abstract
- Developmental plasticity is ubiquitous in natural populations, but the underlying causes and fitness consequences are poorly understood. For consumers, nutritional variation of juvenile diets is probably associated with plasticity in developmental rates, but little is known about how diet quality can affect phenotypic trajectories in ways that might influence survival to maturity and lifetime reproductive output. Here, we tested how the diet quality of a freshwater detritivorous isopod (Asellus aquaticus), in terms of elemental ratios of diet (i.e. carbon: nitrogen: phosphorus; C: N: P), can affect (i) developmental rates of body size and pigmentation and (ii) variation in juvenile survival. We reared 1047 individuals, in a full-sib split-family design (29 families), on either a high- (low C: P, C: N) or low-quality (high C: P, C: N) diet, and quantified developmental trajectories of body size and pigmentation for every individual over 12 weeks. Our diet contrast caused strong divergence in the developmental rates of pigmentation but not growth, culminating in a distribution of adult pigmentation spanning the broad range of phenotypes observed both within and among natural populations. Under low-quality diet, we found highest survival at intermediate growth and pigmentation rates. By contrast, survival under high-quality diet survival increased continuously with pigmentation rate, with longest lifespans at intermediate growth rates and high pigmentation rates. Building on previous work which suggests that visual predation mediates the evolution of cryptic pigmentation in A. aquaticus, our study shows how diet quality and composition can generate substantial phenotypic variation by affecting rates of growth and pigmentation during development in the absence of predation.
|
|
| 5. |
- Lürig, Moritz D.
(författare)
-
phenopype : A phenotyping pipeline for Python
- 2022
-
Ingår i: Methods in Ecology and Evolution. - 2041-210X. ; 13:3, s. 569-576
-
Tidskriftsartikel (refereegranskat)abstract
- Digital images are an intuitive way to capture, store and analyse organismal phenotypes. Many biologists are taking images to collect high-dimensional phenotypic information from specimens to investigate complex ecological, evolutionary and developmental phenomena, such as relationships between trait diversity and ecosystem function, multivariate natural selection or developmental plasticity. As a consequence, images are being collected at ever-increasing rates, but extraction of the contained phenotypic information poses a veritable analytical bottleneck. phenopype is a high-throughput phenotyping pipeline for the programming language Python that aims at alleviating this bottleneck. The package facilitates immediate extraction of high-dimensional phenotypic data from digital images with low levels of background noise and complexity. At the core, phenopype provides functions for rapid signal processing-based image preprocessing and segmentation, data extraction, as well as visualization and data export. This functionality is provided by wrapping low-level computer vision libraries (such as OpenCV) into accessible functions to facilitate scientific image analysis. In addition, phenopype provides a project management ecosystem to streamline data collection and to increase reproducibility. phenopype offers two different workflows that support users during different stages of scientific image analysis. The low-throughput workflow uses regular Python syntax and has greater flexibility at the cost of reproducibility, which is suitable for prototyping during the initial stages of a research project. The high-throughput workflow allows users to specify and store image-specific settings for analysis in human-readable YAML format, and then execute all functions in one step by means of an interactive parser. This approach facilitates rapid program-user interactions during batch processing, and greatly increases scientific reproducibility. Overall, phenopype intends to make the features of powerful but technically involved low-level CV libraries available to biologists with little or no Python coding experience. Therefore, phenopype is aiming to augment, rather than replace the utility of existing Python CV libraries, allowing biologists to focus on rapid and reproducible data collection. Furthermore, image annotations produced by phenopype can be used as training data, thus presenting a stepping stone towards the application of deep learning architectures.
|
|
| 6. |
- Moosmann, Marvin, et al.
(författare)
-
On the evolution of trophic position
- 2021
-
Ingår i: Ecology Letters. - : Wiley. - 1461-023X .- 1461-0248. ; 24:12, s. 2549-2562
-
Tidskriftsartikel (refereegranskat)abstract
- The trophic structure of food webs is primarily determined by the variation in trophic position among species and individuals. Temporal dynamics of food web structure are central to our understanding of energy and nutrient fluxes in changing environments, but little is known about how evolutionary processes shape trophic position variation in natural populations. We propose that trophic position, whose expression depends on both environmental and genetic determinants of the diet variation in individual consumers, is a quantitative trait that can evolve via natural selection. Such evolution can occur either when trophic position is correlated with other heritable morphological and behavioural traits under selection, or when trophic position is a target of selection, which is possible if the fitness effects of prey items are heterogeneously distributed along food chains. Recognising trophic position as an evolving trait, whose expression depends on the food web context, provides an important conceptual link between behavioural foraging theory and food web dynamics, and a useful starting point for the integration of ecological and evolutionary studies of trophic position.
|
|
| 7. |
- Russo, Stefania, et al.
(författare)
-
The value of human data annotation for machine learning based anomaly detection in environmental systems
- 2021
-
Ingår i: Water Research. - : Elsevier BV. - 0043-1354. ; 206
-
Tidskriftsartikel (refereegranskat)abstract
- Anomaly detection is the process of identifying unexpected data samples in datasets. Automated anomaly detection is either performed using supervised machine learning models, which require a labelled dataset for their calibration, or unsupervised models, which do not require labels. While academic research has produced a vast array of tools and machine learning models for automated anomaly detection, the research community focused on environmental systems still lacks a comparative analysis that is simultaneously comprehensive, objective, and systematic. This knowledge gap is addressed for the first time in this study, where 15 different supervised and unsupervised anomaly detection models are evaluated on 5 different environmental datasets from engineered and natural aquatic systems. To this end, anomaly detection performance, labelling efforts, as well as the impact of model and algorithm tuning are taken into account. As a result, our analysis reveals the relative strengths and weaknesses of the different approaches in an objective manner without bias for any particular paradigm in machine learning. Most importantly, our results show that expert-based data annotation is extremely valuable for anomaly detection based on machine learning.
|
|
