The influence of urbanization on stream temperature regimes

© 2017 Manuja Udeshi LekammudiyanseStreams draining urban catchments are relatively warmer than non-urban streams. Catchment urbanization and subsequent stormwater drainage systems are likely a primary cause of this warming of urban streams, however, the extent of these causes and their relative contribution to the degradation of thermal regimes still remain unclear. Addressing this knowledge gap is important to provide guidelines for thermal regime restoration measures. Stream temperature data were collected continuously at a 15 minute time step from temperature loggers deployed across 82 sites around the city of Melbourne (population ~5 million people), in south-eastern Australia, over a 3 year period. After screening for periods with zero flow and removing logger errors, 53 sites with >80% of good quality data over a >30 month period were selected to calculate a set of temperature metrics that indicate the spatial patterns of thermal regimes (mean, mean daily maximum, mean daily minimum, mean diel change, maximum weekly average temperature, mean summer, mean winter, summer maxima and number of days exceeding threshold levels of 24, and 29C per year). To indicate temporal dynamics, the maximum amplitude of temperature surges during storm events was calculated (a surge was defined as >1 C temperature change recorded over 30 minute intervals). All temperature metrics were then linearly regressed with catchment and climatic variables to identify the main drivers of surges in temperature. The thermal regime of the study streams was also examined with reference to the thermal tolerances of native fish occurring in the region. The majority of streams in the study were unlikely to exceed tolerable temperature limits for most Australian native fish species. Spatial variation in temperature (~88%) were most plausibly predicted by models that included attenuated impervious cover (a measure of urban drainage systems), attenuated forest cover (a measure of riparian deforestation), mean annual discharge volume and mean annual air temperature. Stream temperature surges were observed most commonly after rain events and depending on the antecedent weather, the amplitude of surges was either positive or negative. Positive surges (i.e. increases in temperature) were most commonly observed in autumn, with fewer positive surges in summer than in winter. Negative surges (i.e. decreases in temperature) were most common in spring and summer, and least common in winter and autumn. The variations in the amplitude of negative surges (~74%) were explained by the attenuated imperious cover; however, the drivers of spatial variations in positive surges remained unclear. Surges are small and less common and thus, the potential degradation of stream ecosystem cannot be expected from temperature surges. Therefore, temperature surges were appeared not to be an important aspect of urban stormwater impacts on stream ecosystems in Melbourne water region. The temperature models suggest that increasing the extent of drainage systems and riparian deforestation in future climate change would have a potential to limit thermal habitats suitable for temperature-sensitive fish species. Moreover, riparian reforestation was shown not to be able to overcome the influence of drainage systems