Biology subjects

Axtner, J.

Publications and source records attributed to Axtner, J..

5 recordsLinked to original sources

bioRxiv2021

imageseg: an R package for deep learning-based image segmentation

O_LIConvolutional neural networks (CNNs) and deep learning are powerful and robust tools for ecological applications. CNNs can perform very well in various tasks, especially for visual tasks and image data. Image segmentation (the classification of all pixels in images) is one such task and can for example be used to assess forest vertical and horizontal structure. While such methods have been suggested, widespread adoption in ecological research has been slow, likely due to technical difficulties in implementation of CNNs and lack of toolboxes for ecologists. C_LIO_LIHere, we present R package imageseg which implements a workflow for general-purpose image segmentation using CNNs and the U-Net architecture in R. The workflow covers data (pre)processing, model training, and predictions. We illustrate the utility of the package with two models for forest structural metrics: tree canopy density and understory vegetation density. We trained the models using large and diverse training data sets from a variety of forest types and biomes, consisting of 3288 canopy images (both canopy cover and hemispherical canopy closure photographs) and 1468 understory vegetation images. C_LIO_LIOverall classification accuracy of the models was high with a Dice score of 0.91 for the canopy model and 0.89 for the understory vegetation model (assessed with 821 and 367 images, respectively), indicating robustness to variation in input images and good generalization strength across forest types and biomes. C_LIO_LIThe package and its workflow allow simple yet powerful assessments of forest structural metrics using pre-trained models. Furthermore, the package facilitates custom image segmentation with multiple classes and based on color or grayscale images, e.g. in cell biology or for medical images. Our package is free, open source, and available from CRAN. It will enable easier and faster implementation of deep learning-based image segmentation within R for ecological applications and beyond. C_LI

ecology↗

bioRxiv2020

Non-invasive surveys of mammalian viruses using environmental DNA

O_LIEnvironmental DNA (eDNA) and invertebrate-derived DNA (iDNA) have been used to survey biodiversity non-invasively to mitigate difficulties of obtaining wildlife samples, particularly in remote areas or for rare species. Recently, eDNA/iDNA have been applied to monitor known wildlife pathogens, however, most wildlife pathogens are unknown and often evolutionarily divergent. C_LIO_LITo detect and identify known and novel mammalian viruses from eDNA/iDNA sources, we used a curated set of RNA oligonucleotides as viral baits in a hybridization capture system coupled with high throughput sequencing. C_LIO_LIWe detected multiple known and novel mammalian RNA and DNA viruses from multiple viral families from both waterhole eDNA and leech derived iDNA. Congruence was found between detected hosts and viruses identified in leeches and waterholes. C_LIO_LIOur results demonstrate that eDNA/iDNA samples represent an effective non-invasive resource for studying wildlife viral diversity and for detecting novel potentially zoonotic viruses prior to their emergence. C_LI

molecular biology↗

bioRxiv2019

Identifying conservation priorities in a defaunated tropical biodiversity hotspot

AimUnsustainable hunting is leading to widespread defaunation across the tropics. To mitigate against this threat with limited conservation resources, stakeholders must make decisions on where to focus anti-poaching activities. Identifying priority areas in a robust way allows decision-makers to target areas of conservation importance, therefore maximizing the impact of conservation interventions.\n\nLocationAnnamite mountains, Vietnam and Laos.\n\nMethodsWe conducted systematic landscape-scale surveys across five study sites (four protected areas, one unprotected area) using camera-trapping and leech-derived environmental DNA. We analyzed detections within a Bayesian multi-species occupancy framework to evaluate species responses to environmental and anthropogenic influences. Species responses were then used to predict occurrence to unsampled regions. We used predicted species richness maps and occurrence of endemic species to identify areas of conservation importance for targeted conservation interventions.\n\nResultsAnalyses showed that habitat-based covariates were uninformative. Our final model therefore incorporated three anthropogenic covariates as well as elevation, which reflects both ecological and anthropogenic factors. Conservation-priority species tended to found in areas that are more remote now or have been less accessible in the past, and at higher elevations. Predicted species richness was low and broadly similar across the sites, but slightly higher in the more remote site. Occupancy of the three endemic species showed a similar trend.\n\nMain conclusionIdentifying spatial patterns of biodiversity in heavily-defaunated landscapes may require novel methodological and analytical approaches. Our results indicate to build robust prediction maps it is beneficial to sample over large spatial scales, use multiple detection methods to increase detections for rare species, include anthropogenic covariates that capture different aspects of hunting pressure, and analyze data within a Bayesian multi-species framework. Our models further suggest that more remote areas should be prioritized for anti-poaching efforts to prevent the loss of rare and endemic species.

ecology↗

bioRxiv2018

Shifting up a gear with iDNA: from mammal detection events to standardized surveys

O_LIInvertebrate-derived DNA (iDNA), in combination with high throughput sequencing, has been proposed as a cost-efficient and powerful tool to survey vertebrate species. Previous studies, however, have only provided evidence that vertebrates can be detected using iDNA, but have not taken the next step of placing these detection events within a statistical framework that allows for robust biodiversity assessments.\nC_LIO_LIHere, we compare concurrent iDNA and camera-trap surveys. Leeches were repeatedly collected in close vicinity to 64 camera-trap stations in Sabah, Malaysian Borneo. We analyze iDNA-derived mammalian detection events in a modern occupancy model that accounts for imperfect detection and compare the results with those from occupancy models parameterized with camera-trap-derived detection events. We also combine leech-iDNA and camera-trap data in a single occupancy model.\nC_LIO_LIWe found consistent estimates of occupancy probabilities produced by our camera-trap and leech datasets. This indicates that the metabarcoding of leech-iDNA method provides reasonable estimates of occupancy and can be a suitable method for studying and monitoring mammal species in tropical rainforests. However, we also show that a more extensive collection of leeches would be needed to assess mammal biodiversity with a similar robustness as with camera traps. As certain taxa were only detected in leeches, we see great potential in complementing camera-trap studies with the iDNA approach, as long as the collection of leeches follows a similar robust and standardized sampling scheme.\nC_LIO_LISynthesis and applications. The approach we describe here is not restricted to the processing of leech samples, but can be used for the analysis of other iDNA and environmental DNA (eDNA) data. Our study is the first step to shift the application of e/iDNA studies from opportunistic ad-hoc collections to systematic surveys required for long-term wildlife populations and biodiversity monitoring programs.\nC_LI

ecology↗

bioRxiv2018

An efficient and improved laboratory workflow and tetrapod database for larger scale eDNA studies

BackgroundThe use of environmental DNA, eDNA, for species detection via metabarcoding is growing rapidly. We present a co-designed lab workflow and bioinformatic pipeline to mitigate the two most important risks of eDNA: sample contamination and taxonomic mis-assignment. These risks arise from the need for PCR amplification to detect the trace amounts of DNA combined with the necessity of using short target regions due to DNA degradation. FindingsOur high-throughput workflow minimises these risks via a four-step strategy: (1) technical replication with two PCR replicates and two extraction replicates; (2) using multi-markers (12S, 16S, CytB); (3) a twin-tagging, two-step PCR protocol;(4) use of the probabilistic taxonomic assignment method PROTAX, which can account for incomplete reference databases. As annotation errors in the reference sequences can result in taxonomic mis-assignment, we supply a protocol for curating sequence datasets. For some taxonomic groups and some markers, curation resulted in over 50% of sequences being deleted from public reference databases, due to (1) limited overlap between our target amplicon and reference sequences; (2) mislabelling of reference sequences; (3) redundancy. Finally, we provide a bioinformatic pipeline to process amplicons and conduct PROTAX assignment and tested it on an invertebrate derived DNA (iDNA) dataset from 1532 leeches from Sabah, Malaysia. Twin-tagging allowed us to detect and exclude sequences with non-matching tags. The smallest DNA fragment (16S) amplified most frequently for all samples, but was less powerful for discriminating at species rank. Using a stringent and lax acceptance criteria we found 162 (stringent) and 190 (lax) vertebrate detections of 95 (stringent) and 109 (lax) leech samples. ConclusionsOur metabarcoding workflow should help research groups increase the robustness of their results and therefore facilitate wider usage of e/iDNA, which is turning into a valuable source of ecological and conservation information on tetrapods.

molecular biology↗