example, styrofoam and "periphytometers" that hold glass or Plexiglas slides or coverslips). Plastic rods, flagging tape, and artificial aquarium plants also have been used to simulate epiphytic microhabitats (Aloi, 1990). In addition, nutrient-diffusing substrates can be used to assess nitrogen or phosphorus limitation in streams and offer the potential for assessing the effects of herbicides and other water-quality constituents. Examples of nutrient-diffusing substrates include clay pots or saucers (Fairchild and Lowe, 1984; Fairchild and others, 1985) and nutrient-enriched sand substrates (Pringle and Bowers, 1984; Pringle, 1987).
Artificial substrates have a number of limitations that should be considered in relation to the primary objectives of NAWQA ecological surveys, or secondary objectives within a NAWQA study unit. For example, (1) artificial substrates require a minimum of two trips to the sampling reach (installation and retrieval), separated by an interval of colonization time that can vary with season, discharge, temperature, and other environmental variables; (2) they can be susceptible to loss from vandalism or from washout during periods of storm-water discharge; (3) periphyton communities that develop on introduced substrates can be biased toward algal taxa that were actively immigrating or colonizing at the time of substrate placement and might not reflect the types or the relative abundance of algal taxa present on natural substrates (Lamberti and Resh, 1985); and (4) artificial substrates can provide a less sensitive indication of changes or differences in water quality associated with land disturbance, including increases in soil erosion and sediment transport or relative changes in the distribution of periphyton microhabitats within a sampling reach.
However, artificial substrates can be considered for obtaining quantitative periphyton samples (1) from large nonwadeable rivers, or sampling reaches where quantitative samples cannot be collected from natural substrates, (2) when uniformity of substrate surface is an important consideration for water-quality interpretations, or (3) when time-series information, such as determination of algal species-growth rates, is a secondary objective of the ecological survey (for example, a case study to investigate eutrophication processes or the effects of herbicides or other toxicants on primary production).
The preferred sampling method under these circumstances would be to use benthic artificial substrates that serve as a model for natural substrates. For example, rocks with a mineralogic composition similar to that of natural substrates can be introduced as artificial substrates into sampling reaches where rocks do not occur or where the predominant substrate is coarse sand or gravel. Alternative sampling methods include the use of (1) unglazed clay tiles, flagstone, or frosted glass (or acrylic) as an a rtificial substrate for rocks (Lowe and Gale, 1980; Tuchman and Stevenson, 1980); (2) Plexiglas rods (Goldsborough and others, 1986) or plastic flagging tape (Lethbridge and others, 1988) as artificial substitutes for the stems or leaves of aquatic macrop hytes; or (3) wooden dowels (Millie and Lowe, 1983) or pre-weathered wood stakes to simulate snag or epidendric microhabitats.
Suspended artificial substrates can be useful for certain water-quality studies because they (1) provide a uniform substrate surface for algal colonization, (2) normalize among-site differences relative to stream depth and light availability, (3) can redu ce variance among replicate samples from a reach, and (4) are less subject to relative changes in stream stage during the colonization period. However, algal biomass and community structure on artificial substrates are often very different from that observed on natural substrates in the sampling reach. Examples of suspended artificial substrates include glass slides held in a buoyant holder (Patrick and others, 1954; Britton and Greeson, 1988), nylon lines or cords