Environ. Sci. Technol. 2003, 37, 3085-3089
Calcium Losses from a Forested Catchment in South-Central Ontario, Canada SHAUN A. WATMOUGH* AND PETER J. DILLON Environmental and Resource Studies, Trent University, 1600 West Bank Drive, Peterborough, Ontario, K9J 7B8, Canada
Mass balance studies at a number of calibrated watersheds in eastern North America suggest that large losses of Ca from soil have occurred due to acid deposition. However, there is considerable controversy over whether losses have occurred from the exchangeable pool or whether there are other sources of Ca that have not been considered. Mass balance calculations at a small, calibrated catchment (PC1) in south-central Ontario also indicate that large losses of Ca have occurred over the past 2 decades. If the exchangeable Ca pool has declined by approximately 40% between 1983/1984 and 1998/1999 but the exchangeable Mg pool has remained relatively stable, these changes should be evident in streamwater chemistry. The slope of the buffer curve relating Ca to ∑acid anions (FCa) decreased significantly over the study period. We estimate that FCa decreased by 12-24% between 1983 and 1999, although the level of decrease is affected by changes in hydrology. During the same period, there was no significant change in FMg, and the annual volume-weighted Ca:Mg (equiv/equiv) ratio in streamwater decreased from ∼2.8 to ∼2.1. Measured changes in streamwater chemistry translate into Ca losses from the exchangeable pool of approximately 250-570 mequiv/m2 over the study period, which is similar to estimated losses based on soil measurements (∼425 mequiv/m2) and mass balance calculations (∼380430 mequiv/m2). The magnitude of Ca loss cannot be explained by decreases in Ca deposition or decreased Ca weathering. Large Ca losses from the soil exchangeable pool have serious implications for future forest health and the recovery of streamwater from acid deposition.
Introduction In eastern North America there is widespread concern that large Ca losses from soil are occurring due to a combination of high levels of acid deposition, tree growth, and forest harvesting (1-6). Decreases in exchangeable Ca may lead to reduced tree growth, increased forest dieback, a decline in freeze tolerance, and increased susceptibility to drought and disease (7-11). In addition, declining base cation pools in soil, of which Ca is generally the dominant component, may delay the chemical recovery of surface waters to reductions in S deposition (12-14). Results from several nutrient mass balance studies conducted at the watershed scale indicate that large Ca losses from soil have occurred, although these * Corresonding author phone: (705)748-1011, ext. 1647; fax: (705)748-1569; e-mail:
[email protected]. 10.1021/es034034t CCC: $25.00 Published on Web 06/18/2003
2003 American Chemical Society
estimations have seldom been verified by soil measurements. This lack of field verification may simply be due to the inherent variability of soil chemistry, which may obscure even moderate changes in the exchangeable pool. However, alternate sources of Ca have also been proposed to balance the Ca budget, including additional inputs of Ca by mycorrhizal mining of Ca in minerals (15) and dissolution of calcium oxalate (16). A watershed-scale base cation mass balance study at Plastic Lake (PC1) in south-central Ontario indicated that large Ca losses have occurred since the early 1980s, although losses of Mg were much smaller (6, 17). These estimates were supported by repeated field measurements that revealed statistically significant decreases in soil pH and exchangeable Ca, but not Mg, despite a large variation in soil measurements. If substantial losses have occurred from the exchangeable soil pool (i.e., there are no other unaccounted for Ca sources), then these changes should be reflected in streamwater chemistry. In this study, we examine whether changes in PC1 stream chemistry over a 16-yr period are consistent with large losses of Ca but not Mg from the soil exchangeable pool.
Theory We adopt an approach that is described in detail by Kirchner and colleagues (12, 18), which uses a simple mechanistic method for predicting catchment response to changing anion concentrations directly from runoff chemistry. Predicted responses (FSBC) for base cations to changes in ∑acids in streamwater can be calculated as:
FSBC )
dSBC
∑acids
)
d
4[Ca2+] + 4[Mg2+] + [Na+] + [K+] + [NH4+]
∑z
2 i
Ci
where dSBC is the change in concentration of base cations, d∑acids is the change in the sum of strong acid plus organic anions (OA-), zi is the ionic valence, and Ci is the concentration of all ions that directly or indirectly participate in heterogeneous equilibrium reactions within the catchment (in µmol/L) and [ ] denote concentration (also in µmol/L). This approach has been successfully applied to 18 intensively monitored catchments in south-central Ontario, where FSBC values were between ∼0.75 and 0.95 and reliable FSBC responses to drought-induced increases in ∑acid could be predicted by individual predrought measurements (19). Similarly, the response of individual cations Ca and Mg (Fy) (or buffer slope) to changes in anion export can be calculated as:
Fy )
zy2[y]
∑z
2 i
(2)
Ci
where [y] and zy are the concentration (µmol/L) and valence, respectively, of the base cation of interest. At PC1, SO4, OA-, and Cl are the dominant anions in streamwater. Nitrate does not contribute greatly (
