Automatic classification of field-collected dinoflagellates by artificial neural network

Automatic taxonomic categorisation of 23 species of dinoflagellates was demonstrated using field-collected specimens. These dinoflagellates have been responsible for the majority of toxic and noxious phytoplankton blooms which have occurred in the coastal waters of the European Union in recent years and make severe impact on the aquaculture industry. The performance by human 'expert' ecologists/taxonomists in identifying these species was compared to that achieved by 2 artificial neural network classifiers (multilayer perceptron and radial basis function networks) and 2 other statistical techniques, k-Nearest Neighbour and Quadratic Discriminant Analysis. The neural network classifiers outperform the classical statistical techniques. Over extended trials, the human experts averaged 85% while the radial basis network achieved a best performance of 83%, the multilayer perceptron 66%, k-Nearest Neighbour 60%, and the Quadratic Discriminant Analysis 56%.

Analysis of classification accuracy for pre-filtered multichannel remote sensing data

Noise is one of the main factors degrading the quality of original multichannel remote sensing data and its presence influences classification efficiency, object detection, etc. Thus, pre-filtering is often used to remove noise and improve the solving of final tasks of multichannel remote sensing. Recent studies indicate that a classical model of additive noise is not adequate enough for images formed by modern multichannel sensors operating in visible and infrared bands. However, this fact is often ignored by researchers designing noise removal methods and algorithms. Because of this, we focus on the classification of multichannel remote sensing images in the case of signal-dependent noise present in component images. Three approaches to filtering of multichannel images for the considered noise model are analysed, all based on discrete cosine transform in blocks. The study is carried out not only in terms of conventional efficiency metrics used in filtering (MSE) but also in terms of multichannel data classification accuracy (probability of correct classification, confusion matrix). The proposed classification system combines the pre-processing stage where a DCT-based filter processes the blocks of the multichannel remote sensing image and the classification stage. Two modern classifiers are employed, radial basis function neural network and support vector machines. Simulations are carried out for three-channel image of Landsat TM sensor. Different cases of learning are considered: using noise-free samples of the test multichannel image, the noisy multichannel image and the pre-filtered one. It is shown that the use of the pre-filtered image for training produces better classification in comparison to the case of learning for the noisy image. It is demonstrated that the best results for both groups of quantitative criteria are provided if a proposed 3D discrete cosine transform filter equipped by variance stabilizing transform is applied. The classification results obtained for data pre-filtered in different ways are in agreement for both considered classifiers. Comparison of classifier performance is carried out as well. The radial basis neural network classifier is less sensitive to noise in original images, but after pre-filtering the performance of both classifiers is approximately the same.

The assessment of a global marine ecosystem model on the basis of emergent properties and ecosystem function: a case study with ERSEM

Ecosystem models are often assessed using quantitative metrics of absolute ecosystem state, but these model-data comparisons are disproportionately vulnerable to discrepancies in the location of important circulation features. An alternative method is to demonstrate the models capacity to represent ecosystem function; the emergence of a coherent natural relationship in a simulation indicates that the model may have an appropriate representation of the ecosystem functions that lead to the emergent relationship. Furthermore, as emergent properties are large-scale properties of the system, model validation with emergent properties is possible even when there is very little or no appropriate data for the region under study, or when the hydrodynamic component of the model differs significantly from that observed in nature at the same location and time. A selection of published meta-analyses are used to establish the validity of a complex marine ecosystem model and to demonstrate the power of validation with emergent properties. These relationships include the phytoplankton community structure, the ratio of carbon to chlorophyll in phytoplankton and particulate organic matter, the ratio of particulate organic carbon to particulate organic nitrogen and the stoichiometric balance of the ecosystem. These metrics can also inform aspects of the marine ecosystem model not available from traditional quantitative and qualitative methods. For instance, these emergent properties can be used to validate the design decisions of the model, such as the range of phytoplankton functional types and their behaviour, the stoichiometric flexibility with regards to each nutrient, and the choice of fixed or variable carbon to nitrogen ratios.

The biodiscovery potential of marine bacteria: an investigation of phylogeny and function

A collection of marine bacteria isolated from a temperate coastal zone has been screened in a programme of biodiscovery. A total of 34 enzymes with biotechnological potential were screened in 374 isolates of marine bacteria. Only two enzymes were found in all isolates while the majority of enzyme activities were present in a smaller proportion of the isolates. A cluster analysis demonstrated no significant correlation between taxonomy and enzyme function. However, there was evidence of co-occurrence of some enzyme activity in the same isolate. In this study marine Proteobacteria had a higher complement of enzymes with biodiscovery potential than Actinobacteria; this contrasts with the terrestrial environment where the Actinobacteria phylum is a proven source of enzymes with important industrial applications. In addition, a number of novel enzyme functions were more abundant in this marine culture collection than would be expected on the basis of knowledge from terrestrial bacteria. There is a strong case for future investigation of marine bacteria as a source for biodiscovery.

Comparative analysis European wide marine ecosystem shifts - a large scale approach for developing the basis for ecosystem-based management

Abrupt and rapid ecosystem shifts (where major reorganizations of food-web and community structures occur), commonly termed regime shifts, are changes between contrasting and persisting states of ecosystem structure and function. These shifts have been increasingly reported for exploited marine ecosystems around the world from the North Pacific to the North Atlantic. Understanding the drivers and mechanisms leading to marine ecosystem shifts is crucial in developing adaptive management strategies to achieve sustainable exploitation of marine ecosystems. An international workshop on a comparative approach to analysing these marine ecosystem shifts was held at Hamburg University, Institute for Hydrobiology and Fisheries Science, Germany on 1-3 November 2010. Twenty-seven scientists from 14 countries attended the meeting, representing specialists from seven marine regions, including the Baltic Sea, the North Sea, the Barents Sea, the Black Sea, the Mediterranean Sea, the Bay of Biscay and the Scotian Shelf off the Canadian East coast. The goal of the workshop was to conduct the first large-scale comparison of marine ecosystem regime shifts across multiple regional areas, in order to support the development of ecosystem-based management strategies.