Research Feature
Could seasonal changes in water quality lead to seasonal shifts in drinking water treatment requirements?

Key messages

  • The molecular character of natural organic matter (NOM) changed with the seasons, indicating a shift in source, and a potential seasonality to treatment requirements. Specifically, wet (fall/winter) season antecedent wetness conditions linked stream water to distant NOM source pools; this more forest-like NOM has a stronger spectral absorbance coefficient at 254 nm, which was well correlated to disinfection byproduct formation-potential (total trihalomethanes, r = 0.988; and total haloacetic acids, r = 0.993). These results indicate that source water could have a greater drinking water treatment demand during the wet season than during relatively dry (summer) months, when NOM origins were more local.

  • NOM is indeed important for contaminant transport: sample sub-sets were evaluated for total metals content, and dissolved organic carbon (DOC) was well correlated with several metals: mercury (Hg R²=0.99), aluminum (Al R²=0.81), iron (Fe R²=0.72).

  • Sub-surface geological features (parent material) played an important role in surface water quality dynamics and was found to be relatively more important than historical harvest area as a driver for NOM quantity and quality (and relatively less important than seasonal conditions).

  • Streamflow conditions were linked to variations in stream water quality: approximately 80% of the time, DOC concentration peaked with stream stage.

Summary

When and how does aqueous natural organic matter (NOM) vary across a second-growth forested watershed, and what can it tell us about drinking source water and its treatability? To answer these questions, we analysed more than 400 stream water samples collected over sixteen months. Analyses focused on two evaluations of stream NOM: how much is there (measured as concentration of dissolved organic carbon, the backbone of NOM), and what is its character (measured by how the molecules interact with light). Results were used to evaluate patterns over space and time, and to identify the main drivers for those observations.

We found that historical logging (e.g., area harvested, average tree age) had no consistent bearing on the concentration or character of NOM in streams. Instead, subsurface geological features and the weather were key drivers for observed NOM dynamics. The concentration and character of NOM in streams was tied to dominant geological formations (e.g., metamorphic parent materials) underlying the watersheds.

Streamflow and associated wet conditions were found to be important variables in determining concentration and character of stream NOM. Based on the character of NOM, we found that the source pools (i.e., origin) of stream NOM varied over time. During the drier, warmer, summer months, NOM was generated more in-stream (e.g., micro-biological processes like algae
production); and as the wet seasons progressed, NOM showed increasingly terrestrial characteristics (e.g., forest-like structure). Evaluation of NOM dynamics with streamflow showed that the summer-time NOM of in-stream origin was diluted and flushed through the system early in the wet season, and that the forest-like NOM increasingly enriched stream waters as the wet season progressed. This speaks to hydrologic connectivity: essentially, wetness creates pathways to areas relatively far from the stream banks. When it was cold enough to freeze and snow, streamflow was reduced as were NOM concentrations.

Drinking water treatment requirements differ for water with NOM of more forest-like or more in-stream origins. Based on the seasonal shifts in stream water quality, we anticipate there could be parallel seasonal shifts in drinking water treatment requirements. Furthermore, results confirmed that NOM plays a role in contaminant transport: the concentration of DOC was correlated with several metals of aquatic importance (e.g. total mercury, aluminum, iron), and wet-season NOM molecular character (i.e., shape and structure) was correlated to potentially harmful chlorinated disinfection byproducts. Given the seasonal variability in hydrologic connectivity and the role of aqueous NOM in contaminant transport and treatment requirements (e.g., oxidant demand), understanding variable dynamics of stream NOM across a drinking water supply area can help to inform drinking water treatability and land-use planning. 
Photo: Working onsite with partner X.

Contributors

Bill Floyd and Mark Johnson – academic co-supervisors, VIU and UBC

NSERC Canadian Graduate Scholarship Masters Award (CGS-M)

Capital Regional District (CRD) Watershed Protection and Management Division, Integrated Water Services:  Tobi Gardner, Kathy Haesevoets, Annette Constabel, Joel Ussery, Patrick McCoubrey, Ryan Biggs, Burn Hemus, Christoph Moch, Jessica Dupuis, Devon Barnes, and Athina Connor
University of Waterloo:
Mike Stone, Dana Herriman, Monica Emelko

University of Alberta:
Suzanne Tank, Uldis Silins, Axel Anderson

Stewart Butler, VIU (field installations partner)

Ali Bishop, VIU (GIS review and support)

Emily Mistick, UBC (research colleague)