Search

The phytoplankton succession and related abiotic variables were investigated within the frame of a monitoring programme. Results from 1985 to 1991 are presented. For the prymnesiophytePhaeocystis globosa, an increase in the maximum annual abundance was observed. The appearance of this species in the phytoplankton succession followed an annually recurrent pattern. Among the starting conditions for the bloom, a coincidence with the annual DIN/P maximum was most obvious. Quantitatively, temperature and salinity showed the lowest variability at the onset of the blooms. Nutrient uptake during the increasing phase of the bloom appeared to affect nitrate concentrations more significantly than those of ammonia and phosphate.

The photosynthetic performance of macroalgae isolated in Antarctica was studied in the laboratory. Species investigated were the brown algae Himantothallus grandifolius, Desmarestia anceps, Ascoseira mirabilis, the red algae Palmaria decipiens, Iridaea cordata, Gigartina skottsbergii, and the green algae Enteromorpha bulbosa, Acrosiphonia arcta, Ulothrix subflaccida and U. implexa. Unialgal cultures of the brown and red algae were maintained at 0°C, the green algae were cultivated at 10°C. IK values were between 18 and 53 μmol m−2 s−1 characteristic or low light adapted algae. Only the two Ulothrix species showed higher IK values between 70 and 74 μmol m−2 s−1. Photosynthesis compensated dark respiration at very low photon fluence rates between 1.6 and 10.6 μmol m−2 s−1. Values of α were high: between 0.4 and 1.1 μmol O2 g−1 FW h−1 (μmol m−2 s−1)−1 in the brown and red algae and between 2.1 and 4.9 μmol O2 g−1 FW h−1 (μmol m−2 s−1)−1 in the green algal species. At 0°C Pmax values of the brown and red algae ranged from 6.8 to 19.1 μmol O2 g−1 FW h−1 and were similarly high or higher than those of comparable Arctic-cold temperate species. Optimum temperatures for photosynthesis were 5 to 10°C in A. mirabilis, 10°C in H. grandifolius, 15°C in G. skottsbergii and 20°C or higher in D. anceps and I. cordata. P: R ratios strongly decreased in most brown and red algae with increasing temperatures due to different Q10 values for photosynthesis (1.4 to 2.5) and dark respiration (2.5 to 4.1). These features indicate considerable physiological adaptation to the prevailing low light conditions and temperatures of Antarctic waters. In this respect the lower depth distribution limits and the northern distribution boundaries of these species partly depend on the physiological properties described here.

This work introduces a tree structured neural network model for topology preserving vector quantization with one-dimensional Kohonen chains. The leaves of the tree are the chains, each of which quantizes a subspace of the input space. Topological defects can effectively be detected and splitting of the chain at that location results in a growing of the tree structure and increase of topology preservation. Additionally, the chains are able to grow and shrink in order to approximate user defined criteria. Advantages over existing dynamic network models are the flexible tree structure, the total lack of global parameters or calculations as well as the simulation and retrieval speed due to the network structure. Different levels of generalization and prototypicality are naturally observed.

To compare four data-driven approaches to automated tumor grading based on morphometric data. Apart from the statistical procedure of linear discriminant analysis, three other approaches from the field of neural computing were evaluated.The numerical basis of this study was computed tomography-guided, stereotactically obtained astrocytoma biopsies from 86 patients colored with a combination of Feulgen and immunhistochemical Ki-67 (MIB1) staining. In these biopsies the cell nuclei in four consecutive fields of vision were evaluated morphometrically and the following parameters determined: relative nuclei area, secant lengths of the minimal spanning trees and relative volume-weighted mean nuclear volumes of the proliferating nuclei.Based on the analysis of these morphometric features, the multivariate-generated HOM grading system provides the highest correct grading rates (90%), whereas the two widely employed qualitative histologic grading systems for astrocytomas yield correct grading rates of about 60%. For automated tumor grading all approaches yield similar grading results; however, back-propagation networks provide reliable results only following an extensive training phase, which requires the use of a supercomputer. All other neurocomputing models can be run on simple UNIX workstations (ATT, U.S.A).In contrast to discriminant analysis, backpropagation and Kohonen networks, the newly developed neural network architecture model of self-editing nearest neighbor nets (SEN3) provides incremental learning; i.e., the training phase does not need to be restarted each time when there is further information to learn. Trained SEN3 networks can be considered ready-to-use knowledge bases and are appropriate to integrating further morphometric data in a dynamic process that enhances the diagnostic power of such a network.