Effects of Forest Ditch Cleaning on Surface Water Quality

Abstract: Ditch cleaning (DC) is recommended by authorities to Swedish landowners to maintain forest productivity after forest harvest by lowering groundwater tables of catchment soils. Knowledge regarding the impact of this practice on surface water quality is limited, and results from previous research vary. Therefore, this master thesis aims to contribute to the knowledge gap by evaluating the impact of DC in relation to the impact of clear-cutting (CC) on ditch water quality. Additionally, I attempt to investigate whether impacts of DC on water quality can be predicted from catchment characteristics. The study was based on synoptic sampling of water chemistry conducted according to a paired design with 25 cleaned and 25 uncleaned reference ditches located along the east coast of mid-Sweden. The sites were further selected to investigate effects of CC forest harvest, with 25 sites located in direct connection to CC areas and 25 sites located in forested areas (unpaired). Sampling was conducted at three different occasions for 25 chemical variables including standard water chemistry, mercury and dissolved greenhouse gases. The results showed that six out of the 25 chemical variables analysed were significantly different between DC and reference (R) sites, namely pH, sulphate (SO4), SUVA254, carbon dioxide (CO2-C), methane (CH4-C), and nitrous dioxide (N2O-N). A lowered groundwater table and more deeper flow paths following DC is suggested as the main cause for the obsereved differences in chemical composition. In contrast, following CC higher concentrations of many chemical variables could be linked to more superficial groundwater flow with many of the variables typically linked to organic matter and nutrients. No interaction effect was observed between treatments, meaning that DC does not enhance the impact of CC. However, CO2-C was significantly lower in DC compared to R, but significantly higher in CC compared to F, indicating that DC counteracts CC in some respects. Finally, the predictability of water quality following DC based on catchment characteristics proved weak. Only ΔSO4 and ΔCO2-C models showed predictive relevance, but the low explanatory power of the models suggest that these models should be used with caution and seen as indications rather than used for predictions. The results of this thesis provide important information on how water quality could be affected by DC. The multi-variable inclusion of the current study is an important knowledge basis that can be used as a starting point for more in-depth evaluations of the effects of DC on different water quality aspects.