FIGURE 1 (Continued).

this limit the diatoms seem unable to rejuvenate themselves, and they continue to divide until they reach a stage at which cell division is no longer possible. There are reports in the literature of taxa which do divide without a simultaneous size reduction (Drebes, 1977), and some species seem to be able to multiply at their lower size limit without further size reduction for an extended length of time (E. Syvertsen, personal observations). For most species, however, aux-ospore formation is a necessary and normal occurrence in their life cycle. It may take place as a vegetative event or as the result of sexual reproduction.

All diatoms are diplonts with a meiosis at the end of the gametogenesis. The zygote develops into an auxospore. In the centric diatoms, sexual reproduc-

tion is by oogamy with flagellated male gametes, while most pennate diatoms are morphologically isogamous lacking a flagellated stage. A few araphid species have been shown to be anisogamous (Drebes, 1977) and are considered to represent a transitional stage between centric and pennate diatoms. Since the diatoms are thought to have evolved from centrics toward pennates (see also Geological age), this implies a development from oogamy toward isogamy, contrary to what is considered the normal evolution in other groups.

One peculiar consequence of the diatom mode of vegetative multiplication is the possible "eternal life" of larger valves, resulting from the fact that one of the daughter cells inherits one of the parent cell thecae. Nothing seems to be known about how many times a theca can be "reused" in this way. Theoretically, it could go on indefinitely if the cells are not destroyed by external factors. But even if the thecae are reused only a few times, the valve morphology of certain cells may reflect earlier events in the life of the population. As discussed below, a vegetative cell may have a resting spore epitheca documenting previous resting spore formation. In the same way, various valve morphotypes may be found in a population of a species capable of developing morphologically distinct forms in response to environmental influences.


Many diatoms are heterovalvate, i.e., the two valves of a frustule are dissimilar. This is most prominent within the family Achnanthaceae, where the cells have one valve with a raphe and the other without a raphe, and in the Cymatosiraceae, where one valve has a tubular process and the other does not,

Chain-forming" species with cells linked together by siliceous structures may, in addition, have separation valves. These valves are morphologically different from the valves within the chain. Thus, Cymatosira lorenziana has four morphologically distinct types of valves: a separation valve with a tubular process, a separation valve without a tubular process, and intercalary valves with and without a process, respectively. In the genera Bacteriastrum, Chae-toceros, Paralia, and Skeletonema, the intercalary valves of the chains are all alike and different from the separation valves (Fryxell, 1976; Crawford, 1979).

Another type of heteromorphy may be found with species in which the morphology varies in response to changes in the environment. These morphotypes are generally considered to be forms of the species. During and after environmental changes specimens may be found which have two different valves reflecting different environmental conditions. This type of morphological adaptation has been found in Thalassiosira rotula. In this species, the valve morphology changes in response to variations in temperature and the girdle morphology changes in response to available nutrients (Syvertsen, 1977).

The diatom resting spores are first and foremost recognized by their heavily silicified frustules. The resting spore morphology of some species is similar to that of the corresponding vegetative cells, whereas in other species, the resting spores and the vegetative cells differ drastically (Syvertsen, 1979, 1985).

Diatom resting spores are normally formed as a response to unfavorable environmental conditions, and germination occurs when the conditions improve (see Hargraves & French, 1983, for a review). Resting spore formation is common in centric, but rare in pennate marine planktonic diatoms. Whereas resting spores of several centric marine planktonic diatoms germinate in culture within a few days, the freshwater benthic pennate species, Eunotia soleirolii (Kiitzing) Rabenhorst, requires a dormancy of several weeks (von Stosch & Fecher, 1979) before germination. Acbnanthes taeniata and Fragilariopsis oceanica are pennate marine planktonic diatoms known to form resting spores; whether a dormancy period is present in these species is unknown.

Three types of resting spores can be distinguished: exogenous resting spore—the mature resting spore is not physically in contact with a parent cell theca; semiendogenous resting spore—the spore hypovalve is enclosed within one of the parent cell thecae; and endogenous resting spore—the whole spore is enclosed within the parent cell frustule. Normally two or more exogenous resting spores [e.g., a chain of 13 resting spores of Detonula confervacae (Syvertsen, 1979)], two semiendogenous resting spores [e.g., Thalassiosira aus-tralis (Syvertsen, 1985)], and one endogenous resting spore [e.g., Chaetoceros spp. (Hargraves, 1979)] are formed. All three types were found in clonal cultures of Thalassiosira nordenskioldii and T. antarctica with the semiendogenous type as the most common (Syvertsen, 1979).

Resting spore morphology is a more constant, specific feature than the type and mode of formation and, thus, is of greater taxonomic value. Until disproven by Syvertsen (1979) for centric diatoms and by von Stosch & Fecher (1979) for pennate diatoms, it was generally believed that resting spores had no girdle and thus differed from vegetative cells. Among the centric diatoms, the general trend seems to be that resting spores of species, possibly early in the phylogenetic diatom system (e.g., Thalassiosira and Stellarima), have a girdle and are often morphologically similar to the vegetative cells. Resting spores of species in the possibly more advanced part of the system (e.g., Bacteria-strum and Chaetoceros) are usually very different from the vegetative frustules and often lack a girdle. This seems to coincide with a suggested development from exogenous or semiendogenous toward endogenous resting spores (Syvertsen, 1979). On the other hand, phylogenetically advanced pennate diatoms, e.g., Achnanthes taeniata and Fragilariopsis oceanica, form resting spores with a girdle. A special case occurs when resting spores are formed within aux-ospores. This takes place, for instance, in Chaetoceros eibenii (von Stosch et

vegetative cells resting cells

Da semiendogenous resting spores endogenous resting spores

exogenous resting spore

semiendogenous resting spore

Thalassiosira nordenskioeldii

endogenous resting spore parent cell epivalve

- - rudimentary valve primary resting spore valve secondary resting spore valve rudimentary valve

'parent cell hypovalve

FIGURE 2 Formation of vegetative cells, resting cells, and resting spores from a vegetative parent cell. Thalassiosira nordenskioeldii is an example of a species forming all three types of resting spores.

al., 1973) and in Leptocylindrus danicus and L. minimus (Hargraves, 1990). Unlike other diatoms known so far, the resting spores are an obligate part of the life cycle of L. danicus (French &c Hargraves, 1985).

Resting spore formation includes two cytokinetic mitoses (von Stosch et al., 1973), where one or both may be unequal. Depending on the degree of dissimilarity, the rudimentary cells may or may not be visible (Syvertsen, 1979). In terms of morphology, the rudimentary valves are often intermediate between vegetative and resting spore valves, but may be sufficiently different to risk being described as separate species unless their origin is known [vide Thalassiosira australis (Syvertsen, 1985)].

The two valves of a resting spore may be similar or distinctly different. Often the first valve formed (primary resting spore valve) is more similar to the valves of the vegetative cells than is the second valve (secondary resting spore valve). Thus, during resting spore formation at least four morphologically different valve types may be found which can easily be and probably often have been identified as belonging to different species. These valve types are (1) normal vegetative valves, (2) rudimentary valves, (3) primary resting spore valves, and (4) secondary resting spore valves. In addition, intermediate valve types between those mentioned and representing various degrees of development are often seen (E. Syvertsen, personal observations). This diversity of valve types belonging to one and the same species calls for caution in identification work using cleaned diatom material.

Resting spores germinate in two ways, according to whether or not they have a girdle. Spores with a girdle germinate to form two new vegetative cells where the resting spore thecae serve as epithecae [e.g., Thalassiosira (E. Syvertsen, personal observations)], while spores lacking a girdle shed the spore valves in the process of vegetative cell formation, as with Bacteriastrum and Chaetoceros (von Stosch et al., 1973). In the first case, chains formed after resting spore germination have the resting spore valves as epivalves on the end cells, and these cells are thus heterovalvate.

Was this article helpful?

0 0

Post a comment