Search

This repository contains files related to the publication: "European river typologies fail to capture trends in diatom, fish, and macrophyte community composition".

We study the effectiveness of several low-cost oblique illumination filters to improve overall image quality, in comparison with standard bright field imaging. For this purpose, a dataset composed of 3360 diatom images belonging to 21 taxa was acquired. Subjective and objective image quality assessments were done. The subjective evaluation was performed by a group of diatom experts by psychophysical test where resolution, focus, and contrast were assessed. Moreover, some objective nonreference image quality metrics were applied to the same image dataset to complete the study, together with the calculation of several texture features to analyze the effect of these filters in terms of textural properties. Both image quality evaluation methods, subjective and objective, showed better results for images acquired using these illumination filters in compari-son with the no filtered image. These promising results confirm that this kind of illumination filters can be a practical way to improve the image quality, thanks to the simple and low cost of the design and manufacturing process.

Diatom detection has been a challenging task for computer scientist and biologist during past years. In this work, the new state of art techniques based on the deep learning framework have been tested, in order to check whether they are suitable for this purpose. On the one hand, RCNNs (Region based Convolutional Neural Networks), which select candidate regions and applies a convolutional neural network and, on the other hand, YOLO (You Only Look Once), which applies a single neural network over the whole image, have been tested. The first one is able to reach poor results in out experimentation, with an average of 0.68 recall and some tricky aspects, as for example it is needed to apply a bounding box merging algorithm to get stable detections; but the second one gets remarkable results, with an average of 0.84 recall in the evaluation that have been carried out, and less aspects to take into account after the detection has been performed. Future work related to parameter tuning and processing are needed to increase the performance of deep learning in the detection task. However, as for classification it has been probed to provide succesfully performance.

There have been a number of studies that described a serial of type of teratology occurring in different diatom taxa and that highlight the relation between metal concentration and diatom deformities, but this subject still remain not deeply understood. The present study refers to the effect of metal pollution on the diatom Achnanthidium minutissimum s.l. by describing a new form of teratology. The samples were collected in a mine area, Rosia Montana, from Romania. We observed that, exposed to environmental stress, the frustule of diatom cells appeared altered in several ways, with abnormal forms occurring in different diatom species that presented deformed valve outlines, modifications of the raphe canal system, irregular striation or mixed teratologies. In a particular sampling location where A. minutissimum s.l. was identified as the dominant species, 20.53% of the individuals presented an unreported type of deformity. This kind of teratology affects the cingulum, the valvocopula more exactly, by becoming markedly undulate

We study the effectiveness of several low-cost oblique illumination filters to improve overall image quality, in comparison with standard bright field imaging. For this purpose, a dataset composed of 3360 diatom images belonging to 21 taxa was acquired. Subjective and objective image quality assessments were done. The subjective evaluation was performed by a group of diatom experts by psychophysical test where resolution, focus, and contrast were assessed. Moreover, some objective nonreference image quality metrics were applied to the same image dataset to complete the study, together with the calculation of several texture features to analyze the effect of these filters in terms of textural properties. Both image quality evaluation methods, subjective and objective, showed better results for images acquired using these illumination filters in compari-son with the no filtered image. These promising results confirm that this kind of illumination filters can be a practical way to improve the image quality, thanks to the simple and low cost of the design and manufacturing process.

Diatom detection has been a challenging task for computer scientist and biologist during past years. In this work, the new state of art techniques based on the deep learning framework have been tested, in order to check whether they are suitable for this purpose. On the one hand, RCNNs (Region based Convolutional Neural Networks), which select candidate regions and applies a convolutional neural network and, on the other hand, YOLO (You Only Look Once), which applies a single neural network over the whole image, have been tested. The first one is able to reach poor results in out experimentation, with an average of 0.68 recall and some tricky aspects, as for example it is needed to apply a bounding box merging algorithm to get stable detections; but the second one gets remarkable results, with an average of 0.84 recall in the evaluation that have been carried out, and less aspects to take into account after the detection has been performed. Future work related to parameter tuning and processing are needed to increase the performance of deep learning in the detection task. However, as for classification it has been probed to provide succesfully performance.

There have been a number of studies that described a serial of type of teratology occurring in different diatom taxa and that highlight the relation between metal concentration and diatom deformities, but this subject still remain not deeply understood. The present study refers to the effect of metal pollution on the diatom Achnanthidium minutissimum s.l. by describing a new form of teratology. The samples were collected in a mine area, Rosia Montana, from Romania. We observed that, exposed to environmental stress, the frustule of diatom cells appeared altered in several ways, with abnormal forms occurring in different diatom species that presented deformed valve outlines, modifications of the raphe canal system, irregular striation or mixed teratologies. In a particular sampling location where A. minutissimum s.l. was identified as the dominant species, 20.53% of the individuals presented an unreported type of deformity. This kind of teratology affects the cingulum, the valvocopula more exactly, by becoming markedly undulate

Diatom detection has been a challenging task for computer scientist and biologist during past years. In this work, the new state of art techniques based on the deep learning framework have been tested, in order to check whether they are suitable for this purpose. On the one hand, RCNNs (Region based Convolutional Neural Networks), which select candidate regions and applies a convolutional neural network and, on the other hand, YOLO (You Only Look Once), which applies a single neural network over the whole image, have been tested. The first one is able to reach poor results in out experimentation, with an average of 0.68 recall and some tricky aspects, as for example it is needed to apply a bounding box merging algorithm to get stable detections; but the second one gets remarkable results, with an average of 0.84 recall in the evaluation that have been carried out, and less aspects to take into account after the detection has been performed. Future work related to parameter tuning and processing are needed to increase the performance of deep learning in the detection task. However, as for classification it has been probed to provide succesfully performance.

We study the effectiveness of several low-cost oblique illumination filters to improve overall image quality, in comparison with standard bright field imaging. For this purpose, a dataset composed of 3360 diatom images belonging to 21 taxa was acquired. Subjective and objective image quality assessments were done. The subjective evaluation was performed by a group of diatom experts by psychophysical test where resolution, focus, and contrast were assessed. Moreover, some objective nonreference image quality metrics were applied to the same image dataset to complete the study, together with the calculation of several texture features to analyze the effect of these filters in terms of textural properties. Both image quality evaluation methods, subjective and objective, showed better results for images acquired using these illumination filters in comparison with the no filtered image. These promising results confirm that this kind of illumination filters can be a practical way to improve the image quality, thanks to the simple and low cost of the design and manufacturing process.

This paper deals with automatic taxa identification based on machine learning methods. The aim is therefore to automatically classify diatoms, in terms of pattern recognition terminology. Diatoms are a kind of algae microorganism with high biodiversity at the species level, which are useful for water quality assessment. The most relevant features for diatom description and classification have been selected using an extensive dataset of 80 taxa with a minimum of 100 samples/taxon augmented to 300 samples/taxon. In addition to published morphological, statistical and textural descriptors, a new textural descriptor, Local Binary Patterns (LBP), to characterize the diatom’s valves, and a log Gabor implementation not tested before for this purpose are introduced in this paper. Results show an overall accuracy of 98.11% using bagging decision trees and combinations of descriptors. Finally, some phycological features of diatoms that are still difficult to integrate in computer systems are discussed for future work.

Diatoms, a kind of algae microorganisms with several species, are quite useful for water quality determination, one of the hottest topics in applied biology nowadays. At the same time, deep learning and convolutional neural networks (CNN) are becoming an extensively used technique for image classification in a variety of problems. This paper approaches diatom classification with this technique, in order to demonstrate whether it is suitable for solving the classification problem. An extensive dataset was specifically collected (80 types, 100 samples/type) for this study. The dataset covers different illumination conditions and it was computationally augmented to more than 160,000 samples. After that, CNNs were applied over datasets pre-processed with different image processing techniques. An overall accuracy of 99% is obtained for the 80-class problem and different kinds of images (brightfield, normalized). Results were compared to previous presented classification techniques with different number of samples. As far as the authors know, this is the first time that CNNs are applied to diatom classification.

Many biological objects are barely distinguished with the brightfield microscope because they appear transparent, translucent and colourless. One simple way to make such specimens visible without compromising contrast and resolution is by controlling the amount and the directionality of the illumination light. Oblique illumination is an old technique described by many scientists and microscopists that however has been largely neglected in favour of other alternative methods. Oblique lighting is created by illuminating the sample by only a portion of the light coming from the condenser. If properly used it can improve the resolution and contrast of transparent specimens such as diatoms. In this paper a quantitative evaluation of OL in brigthfield microscopy is presented. Several feature descriptors were selected for characterising contrast and sharpness showing that in general OL provides better performance for distinguishing minute details compared to other lighting modalities. Oblique lighting is capable to produce directionally shadowed differential contrast images allowing to observe phase details in a similar way to differential contrast images (DIC) but at lower cost. The main advantage of OL is that the resolution of the light microscope can be increased by effectively doubling the angular aperture. OL appears as a cost-effective technique both for the amateur and professional scientist that can be used as a replacement of DIC or phase contrast when resources are scarce.

[EN] Diatoms, a kind of algae microorganisms with several species, are quite useful for water quality determination, one of the hottest topics in applied biology nowadays. At the same time, deep learning and convolutional neural networks (CNN) are becoming an extensively used technique for image classification in a variety of problems. This paper approaches diatom classification with this technique, in order to demonstrate whether it is suitable for solving the classification problem. An extensive dataset was specifically collected (80 types, 100 samples/type) for this study. The dataset covers different illumination conditions and it was computationally augmented to more than 160,000 samples. After that, CNNs were applied over datasets pre-processed with different image processing techniques. An overall accuracy of 99% is obtained for the 80-class problem and different kinds of images (brightfield, normalized). Results were compared to previous presented classification techniques with different number of samples. As far as the authors know, this is the first time that CNNs are applied to diatom classification

[EN] Accurate taxonomic resolution in light microscopy analyses of microalgae is essential to achieve high quality, comparable results in both floristic analyses and biomonitoring studies. A number of closely related diatom taxa have been detected to date co-occurring within benthic diatom assemblages, sharing many morphological, morphometrical and ecological characteristics. In this contribution, we analysed the hypothesis that, where a large sample size (number of individuals) is available, common morphometrical parameters (valve length, width and stria density) are sufficient to achieve a correct identification to the species level. We focused on some common diatom taxa belonging to the genus Gomphonema. More than 400 valves and frustules were photographed in valve view and measured using Fiji software. Several statistical tools (mixture and discriminant analysis, k-means clustering, classification trees, etc.) were explored to test whether mere morphometry, independently of other valve features, leads to correct identifications, when compared to identifications made by experts. In view of the results obtained, morphometry-based determination in diatom taxonomy is discouraged

[EN] This paper deals with automatic taxa identification based on machine learning methods. The aim is therefore to automatically classify diatoms, in terms of pattern recognition terminology. Diatoms are a kind of algae microorganism with high biodiversity at the species level, which are useful for water quality assessment. The most relevant features for diatom description and classification have been selected using an extensive dataset of 80 taxa with a minimum of 100 samples/taxon augmented to 300 samples/taxon. In addition to published morphological, statistical and textural descriptors, a new textural descriptor, Local Binary Patterns (LBP), to characterize the diatom’s valves, and a log Gabor implementation not tested before for this purpose are introduced in this paper. Results show an overall accuracy of 98.11% using bagging decision trees and combinations of descriptors. Finally, some phycological features of diatoms that are still difficult to integrate in computer systems are discussed for future work

[EN] Diatoms are important organisms in freshwater ecosystems due to their position as primary producers and therefore, analyzing their assemblages provides relevant information on ecosystem functioning. Diatoms have historically been identified based on morphological traits, which is time-consuming and requires well-trained specialists. Nevertheless, DNA barcoding offers an alternative approach to overcome some limitations of the morphological method. Here, we assess if both approaches are comparable methods to study patterns and mechanisms (including environmental filtering and dispersal limitation) of epiphytic diatom metacommunities using a comprehensive dataset from 22 Mediterranean ponds at different taxonomic resolutions. We used a fragment of rbcL barcode gene combined with High-Throughput Sequencing to infer diatom community composition. The overall degree of correspondence between both approaches was assessed by Procrustean rotation analysis and Procrustean randomization tests, whereas the role of local environmental variables and geographical distances was studied using a comprehensive combination of BIOENV, Mantel tests and distance-based redundancy analysis. Our results showed a relatively poor correspondence in the compositional variation of diatom metacommunity between both approaches. We speculate that the incompleteness of the reference database and the bioinformatics processing are the biases most likely affecting the molecular approach, whereas the limited counting effort and the presence of cryptic species are presumably the major biases related with the morphological method. On the other hand, variation in diatom community composition detected with both approaches was strongly related to the environmental template, which may be related with the narrow community-environment relationships in diatoms. Nevertheless, we found no significant relationship between compositional variation and geographical distances. Overall, our work shows the complement

[EN] Benthic diatoms have traditionally been used as bioindicators of aquatic ecosystems. Because diatom-based monitoring of water quality is required by European legislation, molecular-based methods had emerged as useful alternatives to classical methods based on morphological identification using light microscopy. The aim of this study was to test the reliability of DNA metabarcoding combined with High-Throughput Sequencing (HTS) techniques in the bioassessment of the trophic status of 22 Mediterranean shallow ponds in NW Spain. For each pond, the Trophic Diatom Index (TDI) was calculated from inventories obtained by identification using light microscopy (LM) followed by high-throughput sequencing (HTS) at the molecular level. Ponds were subsequently classified into five water quality classes. The results showed a good correspondence between both methods, especially after applying a correction factor that depended on the biovolume of the cells. This correspondence led to the assignment to the same quality class in 59% of the ponds. The determination and quantification of valves or DNA sequences was one of the main pitfalls, which mainly included those related to the variability in the relative abundances of some species. Accordingly, ponds with similar relative abundances for the dominant species were assigned to the same quality class. Moreover, other difficulties leading the discrepancies were the misidentification of some species due to the presence of semi-cryptic taxa, the incompleteness of the reference database and the bioinformatic protocol. Thus, the validation of DNA-based methods for the identification of freshwater diatoms represents an important goal, as an alternative to using traditional methods in Mediterranean shallow ponds

[EN] Diatoms are microalgae, known to present several characteristics that make them excellent bioindicators. In this study, we analyzed the epiphytic diatom communities of ponds, with the aim of exploring the genetic variability explained by physical, chemical or geographical variables. We selected 77 lakes within the Spanish part of the Duero River Basin characterized by wide chemical variability. In each lake, physical parameters were measured in situ and water samples were taken to determine nutrient levels in the laboratory. Diatom communities were collected and identified by DNA-metabarcoding, amplifying a genetic barcode of rbcL gene. We analyzed the phylogenetic signal for several environmental parameters at the genus level for the three most representative genera of these lakes. Our results indicate significant phylogenetic signals that differ from a genus to another. Processes related to spatial scales appear to be the main predictors controlling the genetic diversity of diatom communities in these ponds, with genetic variants differing among geographical areas for a given species. These molecular studies confirm the enormous diversity present in the diatom communities that inhabit these ecosystems, but further research is needed to apply diatom-based techniques in ponds monitoring or conservation programs. Our results reinforce the view that considering not only morphological but also molecular and ecological characters are necessary to describe new living diatom species, whenever possible, especially given the high number of species described each year.