Evaluating the effect of decadal trends in Dissolved Organic Carbon in regional acidification assessment of boreal lakes

Salar Valinia1, Bernard Cosby2, Martyn Futter1, Filip Moldan3, Jens Fölster1
1 Aquatic sciences and assessment, Swedish university of agricultural sciences
2 Centre for Environment and Hydrology
3 Swedish environmental research institute

O 1.12 in Long term trends in the functioning of ecosystems

15.07.2014, 11:20-11:40, H18

Credible estimates of reference conditions are the foundation for understanding the effects of human impact on the natural environment. The European Water Framework Directive (WFD) is based on classification of ecological status in relation an undisturbed state. Since undisturbed conditions rarely are found today, it relies on the idea that the undisturbed state for a given system can be determined through process based models or paleolimnological surveys. Models used to estimate reference conditions are calibrated to present-day monitoring data and may fail to capture trends that have occurred in the environment. When assessing acidification of surface waters using estimated changes in pH, it is important to include the effects of Dissolved Organic Carbon ([DOC]). Monitoring datasets do not extend far enough into history to include the reference period and has led to uncertainty in regional acidification assessment. However, historical concentrations can be reconstructed using Near Infrared Spectroscopy (NIRS) on sediment cores. By combining MAGIC geochemical modelling and NIRS inferred DOC trends we can evaluate the importance of decadal-scale variation of DOC in regional acidification assessment. We developed an empirical DOC model based on NIRS reconstructions and surface water monitoring data that uses site-specific catchment characteristics to estimate the trends in [DOC] between 1860-2010 at 23 sites in Sweden. We then tested the importance of DOC trends for estimating acidification by comparing three approaches for estimating changes in pH from 1860 to 2010 at the 23 sites: (i) assuming constant [DOC] based on 2009 concentrations at each site (∆DOCconst); (ii) estimating variable historical DOC using the trends from the empirical DOC model for each individual site; (∆DOCsite); and (iii) estimating variable historical DOC using the average trend estimated from the empirical model at all sites (∆DOCave). Substantial differences between pH reconstructions were observed depending on which estimates of [DOC] trend were used for MAGIC calibration. We assessed change in pH between reference condition and peak of acidification (1860-1980). The average ∆pHconst for the period 1860-1980 was 1.3 ± 0.9 (± std dev). ∆pHsite was 1.5 ± 0.9 for the period 1860-1980. The ∆pHave was 1.4 ± 0.9 for the period 1860-1980. These results indicate that reconstructions using variable [DOC] display higher ∆pH compared to the ∆pH estimated using constant [DOC] during peak acidification. Between the period 1860-2010 no pronounced changes were identified between the different approaches. Our study is the first of its kind to use long term reconstructions of DOC in combination with geochemical modelling to assess anthropogenic acidification on a regional scale. Based on our results, more credible estimates of ∆pH can be made which create a stronger basis for future predictions of pH and support WFD-related decision making.

Export as iCal: Export iCal

last modified 2014-06-19