STRUCTURE,LIFE HISTORY AND SYSTEMATICS OF RHOICOSPHENIA (BACILLARIOPHYTA). I. THE VEGETATIVE CELL OF RH. CURVATA1

Rhoicosphenia Grun. has been placed by some authors in the monoraphid group with Achnanthes Bory and Cocconeis Ehrenb., and by others near Gomphonema Ehrenb. In order to clarify the systematic position of the genus, the morphology and anatomy of the vegetative cells of Rh. curvata (Kütz.) Grun. were investigated using light and electron microscopy. The structure and formation of the two types of valve are described, and the heterovalvy shown to be of a different type from that of the monoraphids; on the basis of raphe, valve and girdle structure a close relationship between these and Rhoicosphenia is unlikely. Rhoicosphenia shows many resemblances to Gomphonema but the types of pore occlusion present, coupled with apparently slight differences in the mucilage-secreting structures and the girdle, suggest that classification in the same family is unwise. The cryptic asymmetry of the valves, and in particular of the raphe system, is noted and explained with reference to their formation; with respect to this asymmetry two configurations of the valves can occur (named cis and trans types) and the distribution of these in raphid genera is discussed briefly. In view of the lack of evidence in raphid diatoms supporting a classification of bands into copulae and pleurae, it is recommended that this practice be suspended.     

VALVE AND BAND MORPHOLOGY OF SOME FRESHWATER DIATOMS. II. INTEGRATION OF VALVES AND BANDS IN NAVICULA CONFERVACEA VAR. CONFERVACEA12

Chemically cleaned and critical-point dried cells of a clonal culture were examined with scanning electron microscopy. Cells form filaments by valve-to-valve connections maintained by organic material which adheres to the central area of the valve face. Bending of filaments is probably restricted to some extent by the articulation of overlapping spatulate marginal spines with an adjacent underlapping set of much shorter spines (ridges), and with the mantle edge itself. Cell division results in three possible spine patterns for each cell: a set of overlapping and a set of underlapping spines; no overlapping sets of spines (two underlapping); or two sets of overlapping spines (no underlapping). Each filament inherits cells with spine set patterns in the ratio of 2 (with 1 set overlapping): 1 (with no sets overlapping): 1 (with 2 sets overlapping). Valvocopulae are shaped similarly to pleurae except that the partes exteriores of the valvocopulae are wider. The pars interior of both is delimited by an advalvar row of pores continuous around the cell apex. The pars exterior also has a row of pores, but it is median in the valvocopula and first pleura and does not continue around the cell apex. The valvocopulae always underlap the mantle and the pleurae always underlap their preceding band. The ends of both appeared attached, but may become free in acid-cleaned preparations. Bands alternate with each other so that the ends of the valvocopula attach to the first continuous apical portion of the first pleura; the ends of the first pleura attach in that same fashion to the second pleura but at the opposite apex; and all subsequent pleurae alternate in the same fashion with up to at least 13 pleurae/epicingulum. The continuous apical portion of each band is elevated so that a functional (but not structural) ligula is formed, with the continuous apical portion of alternate bands becoming adjacent and underlapping each other only in this region. The valvocopulae in a single cell, or of adjacent cells, may have their continuous apical ends on the same or on opposite apices. It is recommended that N. confervacea var. peregrina (W. Sm.) Grun. be merged with the nominate variety.     

LOCATION AND EXPRESSION OF FRUSTULINS IN THE PENNATE DIATOMS CYLINDROTHECA FUSIFORMIS, NAVICULA PELLICULOSA, AND NAVICULA SALINARUM (BACILLARIOPHYCEAE)

A detailed immunocytochemical and biochemical study of the location and expression of frustulins, a family of proteins associated with the frustules of diatoms, has been performed for Cylindrotheca fusiformis Reimann et Lewin, Navicula pelliculosa (Brébisson et Kützing) Hilse, and Navicula salinarum (Grunow) Husted. Immunocytochemistry revealed that frustulins, which share homologous epitopes but are different in size, were predominantly located in the organic casing. Based on timed immunolocalization experiments and Western blotting analysis of cell extracts obtained sequentially after repleting silicate to Si-synchronized cells, the continuous presence of the frustulins in the mature and parental organic casing of the examined species was observed. The frustulins of N. pelliculosa appeared as proteins similar to those of C. fusiformis, sharing identical epitopes. The extractions, however, yielded a markedly lower abundance of frustulins in N. pelliculosa. Peak concentrations of extracted frustulins appeared to be expressed just ahead of the silicification process in C. fusiformis, whereas the level of expression in N. pelliculosa increased along with maturation of the new valves. For N. salinarum, the presence of the frustulins could not be confirmed properly by Western blotting, most probably because of the small sample volumes, inefficient extraction, and a lower amount of homologous frustulins in the casing of this species. It is concluded that the frustulins of these species are not associated with the silicalemma of the newly formed silica deposition vesicles and therefore do not seem to be involved in the silicification process itself. Overall, the results imply a structural role of the frustulins in the casing of diatoms rather than a regulation of the silicification process.     

VARIATIONS IN THE SUBSTITUTED 3‐LINKED MANNANS CLOSELY ASSOCIATED WITH THE SILICIFIED WALLS OF DIATOMS1

The polysaccharides from cleaned frustules of the diatoms Pinnularia viridis (Nitzsch) Ehrenberg, Craspedostauros australis Cox, Thalassiosira pseudonana Hasle et Heimdal, and Nitzschia navis‐varingica Lundholm et Moestrup were extracted with hot alkali that dissolved the silica and were characterized by constituent sugar and linkage analyses. The polysaccharides from P. viridis were investigated further by permethylation, partitioning according to solubility, desulfation, and CD₃I‐methylation. Yields of carbohydrate in the hot alkali extracts ranged from 0.9% to 1.8% w/w based on the dry weight of the silica. Mannose was the dominant sugar in the polysaccharides from all four species (54–69 mol% of constituent sugars), although 14 other monosaccharides, including neutral sugars (glucose, galactose, xylose, arabinose, rhamnose, fucose), acidic sugars (glucuronic acid, galacturonic acid, 2‐O‐methylglucuronic acid), and O‐methylated neutral sugars (2‐O‐methylrhamnose, 3‐O‐methylrhamnose, 2,3‐di‐O‐methylrhamnose, 3‐O‐methylxylose, 4‐O‐methylxylose) were also detected in varying proportions among the four samples. The polysaccharides were predominantly composed of a 3‐linked mannopyranose backbone with a prevalence of linkage and/or substitution at O‐2 of the 3‐linked mannopyranosyl residues, and they were polyanionic, bearing uronic acid residues and/or sulfate esters. There were, however, species‐specific differences in the degree and position of substitution on the mannan backbone, the type and substitution patterns of the anionic substituents, and the type and linkage patterns of sugars other than mannose. Although definitive functions for these polysaccharides in diatom biology remain uncertain, a possible role in biosilicification is discussed.     

ULTRASTRUCTURE OF CHAETOCEROS MUELLERI (BACILLARIOPHYCEAE): AUXOSPORE,RESTING SPORE AND VEGETATIVE CELL MORPHOLOGY1

The structure of the frustule of auxospores, resting spores and vegetative cells of Chaetoceros muelleri Lemm. are described with LM and SEM. Vegetative frustules are relatively small and lightly silicified, are not united into filaments, and appear unornamented under LM and SEM. The setae are circular to subcircular in transverse section with spines and puncta arranged in a spiral pattern. The resting spore and auxospore frustules are more silicified than the vegetative frustules and appear unornamented under LM and SEM. The auxospores of C. muelleri were previously unknown.     

ACTIVE GLIDING MOTILITY IN AN ARAPHID MARINE DIATOM,ARDISSONEA (FORMERLY SYNEDRA) CRYSTALLINA1

Active gliding movement over long distances was observed and filmed in the marine pennate diatom Ardissonea (Synedra) crystallina (Agardh) Kütz. Typical speeds measured ca. 1–2 μm-s^?1. Motion wax often smooth and steady; however, discontinuous jerky motions and rolling movements were common. Motion, was associated with secretion of twin or, less commonly, single straight trails of mucilage from one end of the cell. In a few instances, reversal in direction was related to cessation of mucilage secretion at one end and commencement at the other. Temporary cessation of movement due to an obstruction was accompanied by a build-up of mucilage at one end of the cell. Mucilage was apparently secreted at two specific sites at each end of the cell and was stained by alcian blue. Persistent trails were visible under scanning electron microscopy (SEM). SEM confirmed that cells had no raphes or labiate processes. The apparent site of secretion was a deep groove formed at the junction of the valve and valvocopula (first girdle band) at each end of the cell. Transmission electron microscopy confirmed the presence of mucilage vesicles in the cytoplasm, but these were not in any manner obviously related to secretion nor was any morphological structure associated with secretion. Cells often become epiphytic through secretion of a terminal stipe. Both stipe secretion and movement may involve the same structural differentiation of the frustule. These results demonstrate a previously unrecorded type of diatom motility. The mechanism, involves mucilage secretion and appears similar to that seen, for example, in some other algae such as the desmids (green algae).     

VALVE AND BAND MORPHOLOGY OF SOME FRESHWATER DIATOMS. III. PRE- AND POST-AUXOSPORE FRUSTULES AND THE INITIAL CELL OF MELOSIRA ROESEANA1

Frustules of a clonal culture of Melosira roeseana Rabenh. were examined with light and scanning electron microscopy. Vegetative valves in the post-auxospore (full size) stage exhibit a larger width/length ratio than those in the pre-auxospore (size-reduced) stage. Cells form chains by linking spines of adjacent valves which occur at the periphery of the valve face-mantle junction. Three or jour large pores occur at the center of the valve face, with the diameter of each pore tapering from the inner to the outer valve surface; these pores are often occluded by siliceous processes. Features of M. roeseana, not shown previously for Melosira, include a “stepped” mantle, on only one of the two valves resulting from the same cell division, flattened processes attached to short siliceous stalks on the valve face, disk-like processes on the mantle, and an open girdle band with up to eight antiligulae. Siliceous scales on the surface of the initial cell are remnants of the auxospore wall. The epivalve of the initial cell is larger in diameter than the hypovalve, and both valves lack linking spines and a step on the valve surface. The initial, cell epicingulum consists of only two bands; the hypocingulum has up to seven. Initial cells with four or more hypocingular bands divide to form new post-auxospore filaments. Melosira roeseana should not be included in the genus Melosira as it is presently defined by the type species, M. nurnmuloides C. Ag. Major differences include irregular linking spines, a closed pseudoloculate valve construction, and labiate processes on the valve face and mantle of M. nummuloides, compared with well-defined linking spines, a valve constructed of a basal siliceous layer perforated by poroid areolae, and labiate processes lacking on the valve of M. roeseana.     

FUNCTIONAL VALVE MORPHOLOGY IN SOME SURIRELLA SPECIES (BACILLARIOPHYCEAE) AND A COMPARISON WITH THE NAVICULACEAE1

On the basis of results of stratigraphic and comparative morphological studies on the diatom frustule, the Surirellaceae is generally assumed to be the endpoint of the evolution of the Pennales. The present study shows that a line of development, based on frustule construction and which parallels the search for optimum design of comparable elements in engineering, can be traced from the Naviculaceae to the Surirellaceae. In both cases lightweight construction is achieved through economy of material and energy expenditure. This leads to structural stability and in the case of the diatom valve, a larger area for metabolic exchange. From the functional-morphological point of view, three construction principles can be distinguished in the genus Surirella: 1. valves with pennate costal framework, raphe keels and fibulae (Surirella gemma group); 2. frustules where all supporting elements are in the form of corrugations, with raphe keel and fibulae (Pinnatae, Fastuosae, Surirella striatula group); and 3. as in 2, but with true alae with alar canals (Robustae) instead of keels with fibulae.