Recommendations

Timing of surveys

In general species diversity varies from season to season, and these patterns may also vary between substrates. However, flow-related factors override these patterns by washing away loosely-attached species during spates and scouring floods (Stevenson, 1990 and CEN, 2003) and the early recolonizers may not reflect water quality accurately. For these reasons, surveys should not be conducted for three to four weeks after a major storm. However, local knowledge is essential if the potential of the diatom-based monitoring is to be maximised. At least one sample per site per year is required for surveillance of water quality. This should be taken either at a time of low flow or when the highest concentration of pollution is expected. The rationale for the time of sampling will vary from region to region. In the peninsular part of India, diatom communities are at the peak of their development in winter to early summer (Karthick et al., 2009). In peninsular and North East hill regions west flowing rivers should be sampled during the post-south west monsoon season from September to February and east flowing rivers should be sampled during post-north east monsoon December to February. The Himalayan streams should be sampled during pre-monsoon, which is before June or post-monsoon during November to January to avoid the influence of monsoon rain on the stream diatom communities.

Site selection

The number and location of sampling sites should be selected so as to provide representative samples, preferably where marked changes in water quality are likely to occur or where there are important river uses, for example confluences, major discharges or abstractions. If sampling is intended to monitor the effects of discharges, sampling both upstream and downstream of discharge points should be carried out. Sampling should extend for an appropriate distance to assess the effects of pollution on the river (CEN, 2003 and Taylor et al., 2005). Basically site selection can be considered at three different scales: (i) a very broad scale, concerned with location of the sampling sites within a catchment, (ii) an intermediate scale concerned with selection of an site for sampling within a designated area, and (iii) a fine scale concerned with determining the precise areas within a reach from where samples should be collected.

Where benthic diatoms are being added to pre-existing surveys, then sample sites are likely to be located close to existing survey stations to aid comparisons between diatoms and other types of environmental information. Such a process is likely to limit the choice of sampling station to a zone of approximately 100 m. Selection of an appropriate reach within this stretch will be determined partly by the choice of substratum. It is important to take notes on the site in standard format to aid subsequent data interpretation. Local situations and needs will determine the precise nature of the information that is required, but the following points should be borne in mind:

1. A detailed description of the site is required on the first visit, on subsequent visits notes of unusual occurrences must be made;
2. Details (including sketch maps) of the site location with map references from the Survey of India topographical sheets (preferably 1:25,000/ 1:50,000). This information should include co-ordinates (Latitude,  Longitude, altitude) and name of village/hamlet;
3. Design of field record form to record all parameters at site; this includes: name of sampler, measurements of channel width and depth, estimates of substratum composition, cover of filamentous algae and other macrophytes, extent of bank side shading, existence of small check dams in upstream and, if known, time since last spate or major rain event;
4. A photographic record as an aid to data interpretation. On subsequent visits, records may be limited to major changes that have occurred since the previous visit, and any variations in sampling protocol employed.

Sites for stream biomonitoring should be in a ‘riffle’, where the water is flowing over stones (Round, 1993), with a current velocity >20cm/sec (CEN, 2003). However, ‘runs’ and ‘glides’ with suitable substrata are also suitable (DARES, 2009). Sampling in ‘riffles’ or areas of moderate or high water velocity ensures continuous exchange of the water surrounding the algae and prevents the build-up of a local chemical environment. Selection of cobbles from pools and ponded areas is to be avoided for sampling for the assessment of water quality (Kelly et al., 1998). The above recommendations have, however, been made for wadeable rivers and are not necessarily applicable to deep rivers. In deep rivers it is often too hazardous to wade in very far. A suitable sample may be collected near to the bank of such a river, provided the river is flowing over the substratum in question. This situation is typical of the lower reaches of all Indian westerly flowing rivers and the delta sections of the easterly flowing rivers, which are not wadeable, cobbles or other substrata may be collected close to the riverbank from ‘riffles’ with flowing water or where flow is >20cm/sec (Fore and Grafe, 2002). This is based on the assumption that the flowing water at the edge of the main stream (littoral zone) be of the same physical and chemical quality as that in the main stream.

The light regime and velocity of water are two physical factors, which determine the diatom community in the streams. The light regime can affect both diatom community structure (Cox, 1984; Kawecka, 1985 and Kawecka, 1986) as well as physiological processes (Guasch and Sabater, 1995) influencing organism’s response to pollutants (Guasch, 1997). Samples that are to be compared should be collected from sites with similar light regimes. Heavily shaded areas are to be avoided, unless it a characteristic of the system under study. Areas very close to the bank should also be avoided because of possible varied water quality and an increased sediment influence on diatom community because of the lower velocity. Stream survey is done by wading, which in turn sets an upper limit on the sampling depth, performance of diatom indices is not affected up to 0.5 m depth (Elber et al., 1992) as long as this is still within the euphotic zone. Water velocity from 1 to 16 cm/sec has no effect on indices (Antoine and Benson-Evans, 1982). Higher water velocity may often lead to changes in the growth rate and relative abundance of species (Antoine and Benson-Evans, 1982 and Wendker, 1992), and also a decrease in species diversity (Lindstrøm and Traaen, 1972 and Rolland, 1997).

Choice of substratum

Diatoms can be found growing on most submerged surfaces; however, the composition of the community varies, depending upon the substratum chosen. Ideally, a single substratum should be used at all sites included in a survey. According to Round (Round, 1993), diatoms form distinct assemblages that occur closely associated with particular microhabitats, e.g. on sediments (epipelon), sand (epipsammon), gravel, stone and bedrock (epilithon) and macrophytic plants (epiphyton). Care needs to be taken not to contaminate the target community with species from other microhabitats when sampling as each substratum has distinct diatom assemblages. Sampling well colonised substrata could minimize the error in inferring ecological conditions. Diatom communities are detected on substrata by feel (slimy or mucilaginous) or seen as a thin golden-brown film covering substrata. In some conditions or at certain times of the year this film may become thicker and much more noticeable. Diatoms also colonise artifacts and waste materials (e.g. plastic bags, pieces of wood). Samples may be collected from such substrata when all other alternatives are absent.

Diatom community structures are governed by substrata associations (Reavie and Smol, 1997). However, major influences on community composition are disturbance (mainly from floods), resource supply (mainly from inorganic nutrients) and, to a lesser extent, grazing (Biggs et al., 1998). It is advisable to carry out an exploratory survey to gain knowledge about the flora growing in different habitats before initiating a detailed study. Stratified random sampling strategies will be appropriate for a site with abundant substrata for detailed investigations.  For example, some workers do not recommend the flora living on silt (‘epipelon’) for routine monitoring, as its high organic content tends to favour taxa with a higher ‘saprobity’ than the prevailing water quality of the river (Vizinet, 1995);

Epilithon

Round (1993), Kelly et al., (1998), Prygiel et al., (2002) and Taylor (2004) consider cobbles and small boulders as the preferred substratum for monitoring diatoms in the riverine environment, promoting the universal applicability of diatom indices and for routine water quality monitoring. The flora at a particular site does not depend on the type of stone sampled (Kelly et al., 1998), although broad aspects of catchment geology do have a pronounced influence on the flora (Lay and Ward, 1987 and Maier, 1994). In addition, this substratum type is preferred for the following reasons:

• Availability of epilithic substrata throughout the length of a river across all seasons (Kelly et al., 1998);
• Easy to collect;
• Clear understanding of diatom ecology and the performance of major diatom-based indices on this substratum (Round, 1993 and Eloranta and Kwandrans, 1996).

It is preferable to collect samples from five or more cobbles (>64≤256mm diameter) or small boulders (>256mm) from a reach of at least 10m ⁶⁵ in the river or stream. Remove any loosely attached surface contamination (e.g. organic debris) by agitating the substratum briefly in the stream water. Place the substrata in a tray, along with approximately 50 ml of river water. Diatoms are removed by vigorously scrubbing the upper surface (the side exposed to flowing water) of the substratum with a small brush (e.g. toothbrush) to dislodge the diatom community. Only the upper side of boulders should be scrubbed to avoid contamination with sediment that might be present on the lower portions of the boulders. A knife or other sharp instrument can be more effective for removing firmly-attached diatoms, but will be less efficient at penetrating crevices on rough surfaces, which may cause more damage to frustules and may lead to more rock particles being transferred to the sample. Care should be taken to avoid instrument contamination between sites by washing the tooth brush, knife and the plastic tray in clean river water and rubbing the tooth brush on a clean surface both before and after taking the diatom sample in order to minimise the contamination. However, it is unlikely that there will be any quantitative difference in the results obtained. Replace the substratum in the stream, and repeat the process for the other replicate substrata. Transfer the contents of the tray to the sample bottle of approximately 125ml. The contents of the bottle should be brown and turbid due to the presence of diatoms. Label the sample bottle with details relevant to the sample.

Epiphyton

In the absence of cobbles or small boulders, emergent or submerged macrophytes — such as Typha sp., Phyla sp., Phragmites sp., Nymphoides sp., Cyperus sp., Eichhornia sp. etc. may be sampled for diatoms. Habitat preference of diatom should be considered while sampling in aquatic vegetation. There are important differences in diatom flora between the epiphyton on bryophytes, emergent macrophytes and submerged macrophytes as well as spatial and temporal differences inherent in sampling a dynamic substratum.

In general, submerged macrophytes are preferred over emergent ones for routine monitoring purposes; however, there is considerable diversity in the manner in which the plants are subsequently treated, from digesting entire portions of the plant to scraping or brushing stems, leaves and roots to squeezing or shaking the plant to dislodge the epiphyton (Porter et al., 1993); Comparative analysis of samples collected from macrophytes and from stones was done by Lenoir and Coste (1994) for rivers in the Rhin-Meuse basins in France. IPS (Indice de Polluosensibilité/Specific Pollution sensitivity Index) (CEMAGREF, 1982) scores varied by less than 2 units on a scale of 1–20 for samples from stones, compared to those from submerged plants. A similar difference in scores was observed for trophic indices for diatom flora on epiphyton and epilithon from lakes in Germany (Hofmann, 1994) and rivers in Finland (Kwandrans et al., 1998).

Sampling from macrophyte substrata: Sampling from emergent macrophyte substrata should be achieved as follows. The emergent macrophyte stem is cut above the water line. A plastic bottle is then inverted over the remainder of the stem and the stem is cut slightly above the point where it emerges from the sediment. The bottle is then inverted and brought to the bank. This procedure needs to be repeated until five stems have been collected (CEN, 2003). The scrubbing and removal of the diatom communities can then proceed in a similar fashion to that described above for solid substrata. Submerged macrophytes can be sampled by selecting replicates from five different plants growing in the main flow of the river. Each replicate, consisting of a single stem plus associated branches of the plant from the lowest healthy leaves to the tip, should be placed in a plastic bag together with 50ml of stream water. Diatoms should be present as a brown film associated with the macrophytes. The plants should be shaken vigorously in the plastic bag and the resulting brown suspension poured into a sample bottle (DARES, 2009 and Taylor et al., 2005).

Artificial substrata

Artificial substrata may be introduced in the euphotic zone of deeper rivers and lakes (with fine silt and sand as substratum) for ecological studies and for bioindication (Iserentant and Blancke, 1986; Cattaneo and Amireault, 1992 and Lane et al., 2003). The following points should be kept in mind when using artificial substrata:

1. A minimum exposure time of four weeks is recommended (Cattaneo and Amireault, 1992 and Hürlimann and Schanz, 1993), although this period is dependent on the trophic status and other parameters like shade and temperature of the water (Guasch et al., 1997). Exposure time must be constant for comparisons between sites.
2. Although glass slides have been widely used in the past (Butcher, 1932 and Patrick et al., 1954), Cattaneo and Amireault, (1992) highlight the need for near ‘natural’ surfaces such as unglazed tiles, (Coring, 1993), or polypropylene rope, frayed at the ends as a substratum and  staked to the river bed (Salden, 1978; Snoeijs and Simenstad, 1995 and Goldsmith, 1997);
3. Artificial substrata should not interfere with the activities of legitimate users of the river (e.g. nearby habitants) or attract the attention of passers-by. Normally, artificial substrata are introduced in secluded stretches of river away from footpaths and bridges to minimize losses or tampering with substrata. Involving nearby schools and local people in monitoring programs help in creating awareness and minimize loss of substrata.
4. The smooth surfaces of some artificial substrata require proper positioning to prevent sloughing off of the diatom film.
5. Results from studies using introduced substrata cannot be interpreted unless full details of methods are reported.
6. Care should be taken while interpreting as the bias towards fast growing diatoms that is not a ‘climax’ community (Round, 1993).