Applications of artificial neural networks predicting macroinvertebrates in freshwaters

To facilitate decision support in freshwater ecosystem protection and restoration management, habitat suitability models can be very valuable. Data driven methods such as artificial neural networks (ANNs) are particularly useful in this context, seen their time-efficient development and relatively high reliability. However, specialized and technical literature on neural network modelling offers a variety of model development criteria to select model architecture, training procedure, etc. This may lead to confusion among ecosystem modellers and managers regarding the optimal training and validation methodology. This paper focuses on the analysis of ANN development and application for predicting macroinvertebrate communities, a species group commonly used in freshwater assessment worldwide. This review reflects on the different aspects regarding model development and application based on a selection of 26 papers reporting the use of ANN models for the prediction of macroinvertebrates. This analysis revealed that the applied model training and validation methodologies can often be improved and moreover crucial steps in the modelling process are often poorly documented. Therefore, suggestions to improve model development, assessment and application in ecological river management are presented. In particular, data pre-processing determines to a high extent the reliability of the induced models and their predictive relevance. This also counts for the validation criteria, that need to be better tuned to the practical simulation requirements. Moreover, the use of sensitivity methods can help to extract knowledge on the habitat preference of species and allow peer-review by ecological experts. The selection of relevant input variables remains a critical challenge as well. Model coupling is a missing crucial step to link human activities, hydrology, physical habitat conditions, water quality and ecosystem status. This last aspect is probably the most valuable aspect to enable decision support in water management based on ANN models.     

Application of artificial neural networks for yeast cells classification

The present paper describes a method for automatic classification of yeast cells in four groups: active with oval form, budding, weakened and dead. This method can be used in the previously developed structural mathematical model of the yeast cultivation process described in [1].     

环境DNA技术在淡水底栖大型无脊椎动物多样性监测中的应用

环境DNA (eDNA)是指生物有机体在环境中(例如土壤、沉积物或水体)遗留下的DNA片段。eDNA技术是指从环境中提取DNA片段进行测序以及数据分析来反映环境中的物种或群落信息。与传统方法相比, eDNA技术具有高灵敏度、省时省力、无损伤等优点。目前, eDNA技术已成为一种新的水生生物监测方法, 主要应用于水生生物的多样性研究、濒危和稀有动物的物种状态及外来入侵动物扩散动态的监测等。本文从eDNA技术在水生生物多样性监测应用领域的发展历程、eDNA技术的操作流程以及其在监测淡水底栖大型无脊椎动物方面的应用进展、技术优势和局限性五个方面进行了综述。最后, 本文对eDNA技术在淡水底栖大型无脊椎动物多样性监测应用的发展趋势和前景作出展望。     

Assessing the impact of artificial summer drainage on the benthic macroinvertebrates in a freshwater wetland in northeast Italy

Hydrobiologia - Valle Mandriole is one of the two last remaining freshwater wetlands in the coastal area of Ravenna (NE Italy). In 2011, a management technique that involves the complete drainage...     

Effects of benthic macroinvertebrates on the exchange of solutes between sediments and freshwater

Infaunal macrobenthos (tubificid oligochaetes, chironomid larvae, and unionid bivalves) were studied in laboratory microcosms to determine their effects on freshwater sediment diagnesis and the exchange of solutes between sediments and water. Tubificids enhanced the flux of ammonium, bicarbonate, and silica from sediments. After the onset of anoxia, they decreased the flux of iron and phosphate. Chironomids increased the flux of nitrate, bicarbonate, and silica, but did not affect the flux of phosphate. Pore water concentrations were low within the irrigated burrowed zone for chemical species normally high in reduced sediments. Concentration gradients were less steep in the actively irrigated burrow zone, but radial diffusion to and from burrows, increased rates of organic decomposition, and enhanced diatom frustule dissolution rates result in enhanced mass transport from sediments. Data from an experiment with unionid clams demonstrated the presence of radial diffusion gradients. These clams enhanced the chloride and nitrate flux from sediments, decreased the bicarbonate flux, but did not affect the flux of either phosphate or silicate. Although the clams did not actively irrigate their burrows, their effect on sediments was similar to that of chironomids. Comparison of direct and indirect flux estimates showed that both types of estimates could be highly variable. In general, indirect flux estimates were higher than direct flux estimates.     

Variable site-occupancy classification of N-linked glycosylation using artificial neural networks

A novel neural-network-based model has been developed for the prediction of N-linked glycosylation characteristics related to glycosylation site-occupancy. Intracellular oligosaccharide transfer to a polypeptide is known to be either robust or dependent upon culture conditions during pharmaceutical production. This glycan attachment is classified by the model as robust or variable and is based on an input of the polypeptide primary sequence around the site of glycosylation. The glycosylation model utilizes multiple recurrent neural networks followed by a perceptron classifier. The input length of the polypeptide chain around the site of glycosylation (glycosylation window) was optimized through multiple independent training sessions. Incorporation of five residues prior (n - 5) to the site of glycosylation (n) and four residues beyond (n + 4) the glycan attachment site led to optimal network performance. The size of the glycosylation window for site-occupancy determination is much larger than has been previously reported. This model was developed to evaluate the effects of theoretical polypeptide mutations on glycosylation site-occupancy characteristics. Following correct prediction of the model testing data set, 20 independent networks were used to predict site-occupancy characteristics of wild-type and mutants of the rabies virus glycoprotein (rgp). Simulation results strongly correlated with previously published experimental results (Kasturi, L.; Hegang, C.; Shakin-Eshleman, S. H. Regulation of N-linked core glycosylation: use of a site-directed mutagenesis approach to identify Asn-Xaa-Ser/Thr sequons that are poor oligosacchride acceptors. Biochem. J. 1997, 323, 415-419. Mellquist, J. L.; Kasturi, L.; Spitalnik, S. L.; Shakin-Eshleman, S. H. The amino acid following an Asn-X-Ser/Thr sequon is an important determinant of N-linked core glycosylation efficiency. Biochemistry 1998, 37, 6833-6837). Further simulations on purely theoretical sequences suggested that influences of charged residues were a subset of multiple mechanisms in the determination of glycosylation site-occupancy.     

Sleep classification in infants based on artificial neural networks.

The study reports on the possibility of classifying sleep stages in infants using an artificial neural network. The polygraphic data from 4 babies aged 6 weeks, 6 months and 1 year recorded over 8 hours were available for classification. From each baby 22 signals were recorded, digitized and stored on an optical disc. Subsets of these signals and additional calculated parameters were used to obtain data vectors, each of which represents an interval of 30 sec. For classification, two types of neural networks were used, a Multilayer Perceptron and a Learning Vector Quantizer. The teaching input for both networks was provided by a human expert. For the 6 sleep classes in babies aged 6 months, a 65% to 80% rate of correct classification (4 babies) was obtained for the testing data not previously seen.     

Comparisons among colonization of artificial substratum types and natural substratum by benthic macroinvertebrates

A field study was designed to concurrently evaluate differences in colonization by benthic macroinvertebrates on a range of artificial substratum types (single particles of natural rock or clay brick and baskets of natural substratum) after three colonization periods (1, 8 and 29 days). Fauna on the artificial substrata were compared to natural substratum and the effect of natural epilithic cover on colonization by zoobenthos was determined. Densities of total number of organisms and the seven most abundant taxa, total number of taxa and quantity of organic material were greater on the natural substratum than on the artificial substratum types. Relative abundances of taxa on pairs of the artificial substratum types, unlike pairs of each artificial substratum type and the natural substratum, were statistically correlated. Among the artificial substratum types densities of total number of organisms and about one-half of the most abundant taxa, total number of taxa and quantity of organic material were greatest in the substratum baskets. Natural epilithic cover on the single rock particles and substratum baskets affected the densities of total number of organisms and two of the seven most abundant taxa. These taxonomic groups were at approximately two to six-fold greater densities on the substrata with fine sediment. consistent patterns in densities of the zoobenthos on the substrata were found after each colonization period. In our study all measures of the macroinvertebrate assemblages (densities of each taxon, total number of organisms, total number of taxa and relative abundances of taxa), with few exceptions, were different between each artificial substratum type and natural substratum. This result showed the abundance and composition of the macroinvertebrate fauna on artificial substratum types were different from the natural substratum. Therefore, the choice of using artificial substrata instead of direct sampling of the natural substratum should be carefully made. Among the artificial substratum types relative abundances of taxa were similar on the single substratum particles and substratum baskets indicating single particles instead of baskets might be used to sample the zoobenthos. Investigators should consider the potential effect of the natural epilithic cover of substratum particles on colonization by zoobenthos when choosing the type of artificial substratum.     

Stressor tolerance values for benthic macroinvertebrates in Mississippi

Conceptually, tolerance values represent the relative capacity of aquatic organisms to survive and reproduce in the presence of known levels of stressors. Operationally, they represent the relative abundance and co-location of organisms and stressors. These numeric values are then used for calculating tolerance metrics. Defensibility of biological assessments using tolerance metrics is compromised if the origins of the tolerance values or technical foundations of metrics are unknown. To minimize circularity and maximize objectivity, we define stressed conditions using physical and chemical factors. Also, since single, isolated stressors in stream systems are rare, we used an approach that combines multiple physical and chemical characteristics into a single general stressor gradient. In this paper, we describe development of tolerance values for benthic macroinvertebrate taxa collected from 455 wadeable stream sites throughout Mississippi, USA, except the Alluvial Plain. Principal components analysis (PCA) was used to develop a gradient that incorporated direct (instream physical and chemical) and indirect (land use) stressors, which was then scaled from 0 to 10. Weighted averaging of the relative abundance of each taxon was used to assign tolerance values based on the point of greatest relative abundance along the stressor gradient. Tolerance values were derived for 324 of the 567 taxa collected from the study sites, and primarily represented sensitivity to agricultural influences including degradation of physical habitat and nutrient enrichment, the dominant stressors within the state. We suggest that this approach could be used in other areas of the country to develop new tolerance values, refine existing ones, and may be a useful approach for other taxonomic groups.     

Pattern recognition and classification of images of biological macromolecules using artificial neural networks.

The goal of this work was to analyze an image data set and to detect the structural variability within this set. Two algorithms for pattern recognition based on neural networks are presented, one that performs an unsupervised classification (the self-organizing map) and the other a supervised classification (the learning vector quantization). The approach has a direct impact in current strategies for structural determination from electron microscopic images of biological macromolecules. In this work we performed a classification of both aligned but heterogeneous image data sets as well as basically homogeneous but otherwise rotationally misaligned image populations, in the latter case completely avoiding the typical reference dependency of correlation-based alignment methods. A number of examples on chaperonins are presented. The approach is computationally fast and robust with respect to noise. Programs are available through ftp.     

Protein classification artificial neural system.

A neural network classification method is developed as an alternative approach to the large database search/organization problem. The system, termed Protein Classification Artificial Neural System (ProCANS), has been implemented on a Cray supercomputer for rapid superfamily classification of unknown proteins based on the information content of the neural interconnections. The system employs an n-gram hashing function that is similar to the k-tuple method for sequence encoding. A collection of modular back-propagation networks is used to store the large amount of sequence patterns. The system has been trained and tested with the first 2,148 of the 8,309 entries of the annotated Protein Identification Resource protein sequence database (release 29). The entries included the electron transfer proteins and the six enzyme groups (oxidoreductases, transferases, hydrolases, lyases, isomerases, and ligases), with a total of 620 superfamilies. After a total training time of seven Cray central processing unit (CPU) hours, the system has reached a predictive accuracy of 90%. The classification is fast (i.e., 0.1 Cray CPU second per sequence), as it only involves a forward-feeding through the networks. The classification time on a full-scale system embedded with all known superfamilies is estimated to be within 1 CPU second. Although the training time will grow linearly with the number of entries, the classification time is expected to remain low even if there is a 10-100-fold increase of sequence entries. The neural database, which consists of a set of weight matrices of the networks, together with the ProCANS software, can be ported to other computers and made available to the genome community. The rapid and accurate superfamily classification would be valuable to the organization of protein sequence databases and to the gene recognition in large sequencing projects.     

Pruning artificial neural networks using neural complexity measures

This paper describes a new method for pruning artificial neural networks, using a measure of the neural complexity of the neural network. This measure is used to determine the connections that should be pruned. The measure computes the information-theoretic complexity of a neural network, which is similar to, yet different from previous research on pruning. The method proposed here shows how overly large and complex networks can be reduced in size, whilst retaining learnt behaviour and fitness. The technique proposed here helps to discover a network topology that matches the complexity of the problem it is meant to solve. This novel pruning technique is tested in a robot control domain, simulating a racecar. It is shown, that the proposed pruning method is a significant improvement over the most commonly used pruning method Magnitude Based Pruning. Furthermore, some of the pruned networks prove to be faster learners than the benchmark network that they originate from. This means that this pruning method can also help to unleash hidden potential in a network, because the learning time decreases substantially for a pruned a network, due to the reduction of dimensionality of the network.     

Biological and ecological traits of benthic freshwater macroinvertebrates: relationships and definition of groups with similar traits

• 1 Relating species traits to habitat characteristics can provide important insights into the structure and functioning of stream communities. However, trade‐offs among species traits make it difficult to predict accurately the functional diversity of freshwater communities. Many authors have pointed to the value of working with groups of organisms as similar as possible in terms of relationships among traits and have called for definition of groups of organisms with similar suites of attributes.
• 2 We used multivariate analyses to examine separately the relationships among 11 biological traits and among 11 ecological traits of 472 benthic macroinvertebrate taxa (mainly genera). The main objective was to demonstrate (1) potential trade‐offs among traits; (2) the importance of the different traits to separate systematic units or functional groupings; and (3) uniform functional groups of taxa that should allow a more effective use of macroinvertebrate biological and ecological traits.
• 3 We defined eight groups and 15 subgroups according to a biological trait ordination which highlighted size (large to small), reproductive traits (K to r strategists), food (animal to plant material) and feeding habits (predator to scraper and/or deposit feeder) as ‘significant’ factors determining the ordination of taxa. This ordination partly preserved phylogenetic relationships among groups.
• 4 Seven ecological groups and 13 ecological subgroups included organisms with combinations of traits which should be successively more adequate in habitats from the main channel to temporary waters, and from the crenon to the potamic sections of rivers, and to systems situated outside the river floodplain. These gradients corresponded to a gradual shift from (1) rheophilic organisms that lived in the main channel of cold oligotrophic mountain streams to (2) animals that preferred eutrophic habitats of still or temporary waters in lowlands. The groups with similar ecological traits had a more diverse systematic structure than those with similar biological traits.
• 5 Monitoring and assessment tools for the management of water resources are generally more effective if they are based on a clear understanding of the mechanisms that lead to the presence or absence of species groups in the environment. We believe that groups with similar relationships among their species traits may be useful in developing tools that measure the functional diversity of communities.

     

A biotic index using benthic macroinvertebrates for Italian lakes

The relationship between benthic macroinvertebrates and environmental variables from 42 Italian lakes were analysed with the aim of developing a biotic index.In the investigated lakes, 570 species were found of which 373 belonged to the family of Chironomidae and 85 to the class of oligochaeta. Rare species, those present in less than 10 samples, were excluded from the analysis, leaving 57 species for data analysis.Multivariate analysis: canonical correlation analysis (CANON) and multivariate analysis of variance and covariance (MANCOVA) were carried out on a large database of 1060 sampling points, for which both environmental data (16 chemical and morphometric variables) and 57 species counts from soft bottom samples were available. In addition, a second dataset of 94 sites from small lakes sampled in 2005 in Northern Italy (Lombardy), was analysed for comparison.The data analysis (CANON) from the large database gave the following results:

• (1)the first canonical variate was related to conductivity, pH and alkalinity and accounted for 17% of the total variation;
• (2)the second canonical variate was related to total phosphorus, N-NH₄ and dissolved oxygen, and accounted for 15% of the total variation;
• (3)lake maximum depth, volume and water temperature were related to the third canonical variate, which accounted for 14% of the source of variation.

The analysis of the database from small lakes revealed that in the plane of the first two canonical variates dissolved oxygen was inversely correlated with the depth of sampling and that the second canonical variate showed an inverse relationship between transparency and nutrients.MANCOVA found evidence for differences between lake types and sampling years, confirmed the CANON results and allowed the quantification of the contribution of different target variables on species composition.A trophic status index was calculated ranking percentage oxygen saturation (O₂ %sat), transparency (transp) and total phosphorus (TP). The means of O₂ %sat, transp and TP weighted by the species abundances were calculated to have optimum values for each species; a benthic quality index was then obtained by multiplying the weighted means of O₂ %sat, transp and TP by species abundances and dividing the product by the total number of specimens present at each station. A significant correlation was observed between trophic status and the benthic quality index. A good agreement was also observed between the indexes calculated using the large database and the second dataset on small lakes.     

String matching artificial neural networks.

Three artificial neural networks (ANNs) are proposed for solving a variety of on- and off-line string matching problems. The ANN structure employed as the building block of these ANNs is derived from the harmony theory (HT) ANN, whereby the resulting string matching ANNs are characterized by fast match-mismatch decisions, low computational complexity, and activation values of the ANN output nodes that can be used as indicators of substitution, insertion (addition) and deletion spelling errors.     

A method for subsampling unsorted benthic macroinvertebrates by weight

A simple method for subsampling unsorted benthic macroinvertebrates by weight is described for different types of samples obtained from lentic and lotic environments. It is especially useful for samples containing large amounts of filamentous algae that preclude the use of conventional subsampling methods. The method provided random dispersions of individuals in the original samples. Overall variability of the subsamples was low for artificial substrate and catastrophic drift samples. Variability was higher for regular drift samples, which had the lowest numbers of individuals of the three sample types. The method produced considerable savings in time spent sorting. Subsampling approaches for community level studies are discussed.     

Cardiovascular control with artificial neural networks

This paper describes the use of artificial neural networks to model cardiovascular autonomic control in a study of the hemodynamic changes associated with space flight. Cardiovascular system models were created including four parameters: heart rate, contractility, peripheral resistance, and venous tone. Artificial neural networks were then designed and trained. A technique known as backpropagation networking was used and the results of the application of this technique to heart rate control are presented and discussed.     

Stream mosses as chemically-defended refugia for freshwater macroinvertebrates

Marine and terrestrial studies show that small, sedentary herbivores that utilize plants as both food and habitat can gain enemy-free space by living on hosts that are chemically defended from larger, generalist consumers. Although large herbivores are increasingly recognized as important consumers of macrophytes in freshwater communities, the potential indirect effects of herbivory on plant-associated macroinvertebrates have rarely been studied. Here, we show that the large, generalist consumers in a riverine system, Canada geese, Branta canadensis , and crayfish, Procambarus spiculifer , both selectively consumed riverweed, Podostemum ceratophyllum , over an aquatic moss, Fontinalis novae-angliae, even though moss comprised 89% of the total plant biomass on riverine rocky shoals. Moss supported twice as many plant-associated macroinvertebrates as riverweed, suggesting that it might provide a spatial refuge from consumption by these larger consumers. Bioassay-guided fractionation of moss extracts led to the isolation of a C₁₈ acetylenic acid, octadeca-9,12-dien-6-ynoic acid, that deterred crayfish feeding. In contrast to results with Canada geese and crayfish, both the amphipod Crangonyx gracilis and the isopod Asellus aquaticus consumed significant amounts of moss but rejected riverweed in laboratory feeding assays. Moreover, neither amphipod nor isopod feeding was deterred by the crude organic extract of Fontinalis , suggesting that these mesograzers tolerate or circumvent the chemical defenses that deterred larger consumers. Thus, herbivory by large, generalist herbivores may drive freshwater plant community structure towards chemically defended plants and favor the ecological specialization of smaller, less mobile herbivores on unpalatable hosts that represent enemy-free space.