4.3. Diatom indices
In the R. Alwin, the ‘‘control’’ stream for this parti-cular study, TDI-UK and TDI-D values were fairly constant throughout the study, with mean values of 37 and 1.8 respectively (Table 1, Fig. 4). Values of both indices were also similar in Usway Burn at the start of the study, despite the floristic differences mentioned earlier. However, from mid-June onwards, the increase in Nitzschia mentioned earlier led to an increase in both indices as well as in the total number of motile valves recorded in counts (‘‘% motile’’: Fig. 5). Both indices started to decrease from their peak values at about the same time as aerial fertilisation started and by the end of the study in early October, values in both rivers were, again, similar. By comparison, values of the TI were slightly higher for Usway Burn than the River Alwin throughout the study but there was little temporal change in the TI at either site.
5. Discussion
Results presented in this paper have a number of implications for those designing programs to monitor the effects of nutrient control, and interpreting their results. Data from the R. Alwin (Table 1) indicate the scale of short-term variation to be expected in diatom assemblages and indices in rivers assumed to be in a ‘‘steady-state’’ (i.e. with no known environmental per-turbations during the study period). This will, in turn, determine the sensitivity of indices such as the TDI to nutrient fluxes. By contrast, results from Usway Burn demonstrate the problems inherent in interpreting results when changes due to a ‘‘predicted’’ environ-mental change (aerial fertilisation in this case) are over-lain by changes due to an unexpected pertubation. That this change was genuinely unexpected means that much of the reasoning in this discussion is post hoc, but this is a necessary part of reactive studies (cf Sabater, 2000).
5.1. Reasons for floristic changes in Usway Burn
Changes observed in Usway Burn fall into two groups: those that occurred before, and those that occurred after the onset of aerial fertilisation. Those that occurred in advance of aerial fertilisation were unexpected, within the initial context of the study. The shift that was observed was from a flora dominated by