No abstract

Low dissolved oxygen concentration in bottom layers of lakes and reservoirs usually indicates low water quality. In lakes, empirical models predicting anoxia are almost entirely based on the decay of plankton biomass, while in reservoirs recent findings suggest a prominent role of streamflow and load of organic carbon. This suggests a potential link between water quality in reservoirs and climate processes affecting streamflow. Here we support this hypothesis presenting evidence that both interannual climate variability and recent climate change, mainly consisting in a significant increase in potential evapotranspiration in the upstream basin, affected the oxygen content in a Mediterranean reservoir (Sau Reservoir, Spain). Using a 44-year monthly record, we found strong and consistent signatures of El Niñ o Southern Oscillation in the inflow and reservoir oxygen content. Spectral and wavelet techniques showed that the El Niñ o, streamflow, and reservoir oxygen content series oscillated in common periods, which coincided with the main El Niñ o variability modes. An empirical model explaining the annual oxygen content in the reservoir suggested that a decreasing streamflow trend reduced the oxygen content of the reservoir by about 20%, counteracting remediation measures implemented at the basin upstream the reservoir. Our results provide the first quantitative evidence of climate change effects on reservoir water quality using long-term instrumental data, and indicate that streamflow should be considered as a key variable in assessing climate change impact on reservoir water quality. These results are especially relevant in regions of the world where reservoirs are abundant and most climate models predict a decrease in runoff during the next decades. Both the expected trends and the sensitivity of reservoir water quality to global interannual climate variability should be considered for a correct management of water resources in the present and to design adaptation policies in the future.

No abstract

Here we describe a practical, step-by-step primer to scale-dependent correlation (SDC) analysis. The analysis of transitory processes is an important but often neglected topic in ecological studies because only a few statistical techniques appear to detect temporary features accurately enough. We introduce here the SDC analysis, a statistical and graphical method to study transitory processes at any temporal or spatial scale. SDC analysis, thanks to the combination of conventional procedures and simple well-known statistical techniques, becomes an improved time-domain analogue of wavelet analysis. We use several simple synthetic series to describe the method, a more complex example, full of transitory features, to compare SDC and wavelet analysis, and finally we analyze some selected ecological series to illustrate the methodology. The SDC analysis of time series of copepod abundances in the North Sea indicates that ENSO primarily is the main climatic driver of short-term changes in population dynamics. SDC also uncovers some long-term, unexpected features in the population. Similarly, the SDC analysis of Nicholson's blowflies data locates where the proposed models fail and provides new insights about the mechanism that drives the apparent vanishing of the population cycle during the second half of the series.