Comparison of Cell-In-Series and Meso-Scale Physical Habitat Sampling for the Interpretation of Spatiotemporal Variation of Stream Water Quality

Abstract

Although scale dependence of ecological patterns is conceptually recognised, the studies involving quantitative assessments are rare and rudimentary. Here we evaluate spatiotemporal variation of water quality using two sampling scales (approaches): cell-in-series (CIS) and meso-scale physical habitats (MPH). CIS has its origins in probabilistic sampling and relatively simple. It also reported to be suitable for streams with advective transport. MPH approach is relatively novel for water quality assessments and it considers medium scale morphological units such as pools, riffles, glides, etc. for sampling. Sampling was carried out in the short and steep Tseng Lan Shue stream, during Spring and Summer of 2012. The stream is subject to regulation and various anthropogenic inputs, but with irregular occurrence. For each season, observations were carried out during periods with no influence of severe weather events (typical state) as well as after a rainfall (flushed state). The response variables including water chlorophyll, turbidity, dissolved oxygen, nitrate, nitrite, ammoniacal nitrogen and soluble reactive phosphorous were checked against a set of hydro-environmental variables. These included: stream velocity, width, depth, and slope, bankfull dimensions, and substrate conditions. Relationships among variables were evidenced using redundancy analysis. In general, the water quality parameters showed an irregular variation in the longitudinal direction of the stream. Response and hydro-environmental variables based on two best axes showed a 41% of variance in spring response data in the MPH approach. For CIS it was around 44%. But in flashed floods these were observed to be 60% and 35 % for MPH and CIS, respectively. Similar trend was observed in summer where explanatory power based on CIS was higher for typical state but otherwise for the flashed state. Furthermore, significant environmental variable(s) for respective cases changed with the scales being used: substrate conditions for CIS and stream width and slope for MPS. This study shows that MPH approach is more suitable than CIS as a modelling tool when the stream has less anthropogenic loads. We conclude that the explanatory powers of the MPH and CIS scales (approaches) could be useful in providing a quantitative definition on identifying a “pristine stream”.