Streambed sediment composition and deposition in a forested stream: Spatial and temporal analysis

Assessment of fish habitat quality often includes the measurement of substrate composition and fine sediment deposition in streams. The natural spatial (localized, inter-reach) and temporal (event-specific, seasonal, annual) variability of both sediment deposition and streambed composition were investigated in Catamaran Brook (New Brunswick, Canada). A grab-type sampler (i.e. scoop) and Whitlock-Vibert boxes were used to investigate grain size distribution of streambed material and deposition of fine sediments over time. Results showed that the scoop sampler was a good method to quantify intra-site and inter-site grain-size distribution, especially as it relates to the streambed composition and changes thereof. In Catamaran Brook, scoop samples showed that gravel constituted the highest percentage of the total dry weight with a median of 38.8%, for seven sites monitored in three different study reaches. The median percentage of fines (i.e. material < 2 mm) was found to be 10.8% of total weight. Intra-site variation in substrate grain size distribution was also assessed using the scoop sampler. Significant differences in the distribution of fine sediments within a study reach as well as among habitat types (riffles, runs, etc.) were found, which reiterates the importance of taking multiple samples. To observe changes in fine sediment, a subsample of the scoop sampler limited to grain sizes < 32 mm was used to reduce some of the variability in weight percentages due to large substrate material, e.g. cobbles. The percentage of fines within this subsample varied between 15.2% and 20.0% (overall median of 18.6%). Statistical analyses showed that the temporal variability in grain size distribution was significant and was mainly associated with hydrological events such as midwinter thaw and summer peak flows. Fredle indices (f(i)), an index for assessing survival of incubating alevins, were calculated in Catamaran Brook and varied between 2.9 and 13.1, with marked spatial and temporal variability. There was a general decrease in f(i) in the autumn of 1996, which may have been related to a redistribution of fines within the drainage basin. Whitlock-Vibert boxes appeared to be more efficient for determining storage rates than for substrate composition analyses. The median storage rate in Catamaran Brook was 40 kg/m(2)/year. Fine sand median storage rate was 9.4 kg/m(2)/year. Our study indicates that scoop samplers and Whitlock-Vibert boxes should both be used, in order to assess sediment deposition and variation in grain size distribution simultaneously. Copyright (c) 2005 John Wiley & Sons, Ltd.