Introduction
Assessments of environmental conditions in rivers and streams using diatoms have a long history in which two basic conceptual approaches emerged. First, based on the work of Kolkwitz and Marsson (1908), autecological indices were developed to infer levels of pollution based on the species composition of assemblages and the ecological preferences and tolerances of taxa (e.g. Butcher, 1947; Fjerdingstad, 1950; Zelinka and Marvan, 1961; Lowe 1974; Lange-Bertalot, 1979). Second, Patrick's early monitoring studies (Patrick, 1949; Patrick et al., 1954; Patrick and Strawbridge, 1963) relied primarily on diatom diversity as a general indicator of river health (i.e. ecological integrity), because species composition of assemblages varied seasonally and species diversity varied less. The conceptual differences in these two approaches really address two different goals for environmental assessments, one inferring pollution levels and the other determining biodiversity, a more valued ecological attribute (Stevenson, 2006). Thus, the concepts and tools for assessing ecosystem health and diagnosing causes of impairment in aquatic habitats, particularly rivers and streams, were established and developed between ∼50 and 100 years ago.
Today, diatoms are being used to assess ecological conditions in streams and rivers around the world (Asai, 1996; Kelly et al., 1998; Wu, 1999; Lobo et al., 2004; Wang et al., 2005; Chessman et al., 2007; Taylor et al., 2007; Porter et al., 2008). They have become valuable elements in large-scale national and international assessment programs of the United States and Europe (e.g. Kelly et al. 2009a).