Diatoms are unicells that share the feature of having a cell wall made of silicon dioxide. This opaline or glass frustule is composed of two parts (valves), which fit together with the help of a cingulum or set of girdle bands.
The taxonomy of diatoms is based in large part on the shape and structure of the siliceous valves. Although some species can be identified in the living state, and some genera were and have been erected based on the organization and fate of cytoplasmic organelles, species-level identification usually requires examination with oil immersion objectives of permanently-mounted specimens. To accomplish this, diatoms are "cleaned" to rid them of their organic components. Cleaning is done by oxidizing diatom material in potassium dichromate (van der Wurff 1954). After washing the cleaned material (which will appear white) several times in distilled water, it is dried onto a coverslip which is then mounted onto a glass slide with Hyrax or other mounting medium with a high refractive index (Naphrax).
Two major groups of diatoms are generally recognized: the centric diatoms exhibit radial symmetry (symmetry about a point) and have oogamous sexual reproduction, while the pennate diatoms are bilaterally symmetrical (symmetry about a line) and produce ameboid gametes that are morphologically similar but may be physiologically different. Chloroplasts of diatoms are variable, but consistent within most taxa. Chloroplasts may be many small discs, a condition found in most centric diatoms and some (araphid) pennates, or few large, plate-like chloroplasts are found in the majority of pennate taxa.
Diatoms live in most freshwater systems, including a wide range of pH, organic pollution, and temperatures. Cells occur singly or in colonies, some of which are visible to the naked eye. Mucilage is secreted by most diatoms and it covers the exterior of the frustule. Mucilage is also used to produce a diversity of growth forms/habits. Diatoms may grow at the ends of mucilaginous stalks (e.g. Achnanthes, Gomphonema) or within mucilaginous tubes (e.g. Encyonema). Mucilage is also employed to hold parts of different cells together in chains (e.g. Tabellaria, Diatoma) or in stellate colonies (Asterionella). Colony formation is also accomplished in diatoms by interdigitation of siliceous spines on the valve margins of adjoining cells (e.g. Fragilaria, Aulacoseira).
Through the life cycle of these diploid organisms (see Kociolek and Williams 1988 for a review) cell size decreases, as each valve of a frustule produces a smaller, complementary valve. Accompanying size diminution are changes in outline and, but less frequently, ornamentation (see Stoermer and Ladewski 1982; Stoermer et al. 1986 for examples). After sexual reproduction, which is accomplished through meiosis, the small zygote enlarges to form an auxospore, a mostly non-siliceous stage that has siliceous strips or scales covering it. The stage next produces a valve. The initial valve represents that largest valve in the cell cycle. Hence, sex accomplishes genetic variability and allow a cell line to regain maximum size before the next round of vegetative (mitotic) cell divisions. Diminution accompanies vegetative cell division until a certain size range is reached where, when environmental parameters are also right, sexual reproduction is induced. Most diatomists have considered the diminution of diatoms through their ontogeny a consequence of having rigid cell walls, but Lewis (1984) interprets the diminution cycle as a timing mechanism or adaptation ("sex clock") for sex in these unicells.
Although diatomists speak of vegetative cell division resulting in high degree of fidelity in diatom morphology, variability occurs within and between cell lines in a population and between populations. This has important implications for the identification of diatoms at the species and subspecific levels. The effect of cell division and the distribution of resulting sizes (e.g. ), variability in size, shape and ornamentation may occur within and between descendants of one or more individuals. Diatom taxonomists usually distinguish taxa based on a lack of intermediates between ranges of variability (see Theriot and Stoermer 1984; Kociolek and Stoermer 1991 for examples of this approach). However, heterovalvy may occur within individuals so that valves of a frustule possess the ornamentation of different "taxa" although no intermediate morphologies occur (e.g. Stoermer 1967). Obviously, in this case, the two "taxa" must be considered as one. In addition, differences in environmental conditions (e.g. salt concentrations, silica levels; e.g. Tuchman et al. 1984) can alter morphology within a taxon. All of these factors must be considered when attempting species-level identifications.
Critical examination of diatoms valves with light and electron microscopy has led to a proliferation of terms associated with the minutiae of valve morphology. Given the excellent, recent treatments of Round et al. (1990) and Krammer and Lange-Bertalot (1986, 1988, 1991) with regard to diatom ultrastructure, it seems redundant to provide a detailed treatment here. Terminology necessary to work through the keys and descriptions of genera is presented in the glossary.
Two taxonomic works have been mainstays in the libraries of diatomists in this country. Patrick and Reimer's (1966, 1975) excellent tomes, The Diatoms of the United States , which does not include centrics or most of the problematic keel-bearing taxa (e.g. Nitzschia) have been complemented by the classic work by Hustedt (1930). This latter work has been updated in a 4 volume set by Krammer and Lange-Bertalot (1986, 1988, 1991). Other works of utility to freshwater ecologists in identifying diatoms include Cleve-Euler (1951-1955), Hustedt (1927-1966), Simonsen (1989), Van Heurk (1880-1885) and Schmidt et al. (1854-1959).
The present work deviates somewhat from past taxonomic treatments of diatoms, particularly those included in large treatises of freshwater algae in North America (e.g. Smith 1950; Prescott 1951). These treatises relied heavily on other, older treatments of diatoms (e.g. Boyer 1927 in Smith's work) or gave diatoms only peripheral consideration (e.g. Prescott's work). The work presented herein represents our own treatment of the flora in the region, based on our own collections. In addition, the taxonomy of diatoms has undergone a revolution in the last 15 years, with many new taxa (particularly at the genus level) being proposed on ultrastructural characteristics. Many of these (e.g. Aulacoseira) have gained acceptance by diatom taxonomists, but have not found their way into general usage in the phycological literature. The work by Round et al. (1990), for all of its shortcomings (e.g. Kociolek 1991), has led to the recognition of many new taxa, several of which are followed in the present work (e.g. Luticola and Sellaphora, taxa considered part of Navicula in more classical treatments). We make reference to the older names and literature in the generic descriptions, so that ecological or distributional data may be gleaned for these taxa from the older literature. Other taxonomic proposals (the splitting or lumping of the araphid genera Fragilaria and Synedra; Williams and Round 1988, 1989; Lange-Bertalot 1991), which have serious implications for the way we deal with diatom taxonomy, have not been followed here, pending further arguments and evidence from both sides.