This research represents a multi-disciplinary and multi-institutional effort to extrapolate research, monitoring, and modeling results, including physical, chemical, and biological findings from intensively-studied lakes to the regional population of acid-sensitive Adirondack lakes. Extrapolation was based on the statistical frame of EPA’s Environmental Monitoring and Assessment Program (EMAP). Intensively-studied sites were drawn from RPI’s Adirondack Effects Assessment Program (AEAP) and the New York State Department of Environmental Conservation’s Adirondack Long-term Monitoring Project (ALTM). A total of 70 watersheds were included in this effort, which involved field sampling to develop a statistically-representative soils database and model projections using the MAGIC and PnET-BGC models to classify lakes according to their sensitivity to change in atmospheric sulfur (S) and nitrogen (N) deposition.
We studied soil characteristics at 199 locations within 44 statistically-selected Adirondack lake-watersheds, plus 26 additional watersheds that are included in long-term lakewater monitoring programs. The statistically-selected watersheds were chosen to be representative of Adirondack watersheds containing lakes larger than 1 ha and deeper than 1 m that have lakewater acid neutralizing capacity (ANC) less than or equal to 200 μeq/L. Results of soil analyses and model projections of lakewater chemistry were extrapolated to the watersheds of 1,320 low-ANC lakes. In general, the concentrations of exchangeable base cations, base saturation, and soil pH were low. More than 75% of the target lakes received drainage from watersheds having average soil B horizon exchangeable Ca concentrations less than 0.52 cmolc/kg, base saturation (BS) less than 10.3%, and pH (H2O) less than 4.5. These data provide a baseline against which to compare future changes in regional soil chemistry. In addition, the soil data provided input for aquatic effects models used to project future changes in surface water chemistry and biological conditions.
Lake water chemical recovery responses have been indicated in ongoing lakewater monitoring databases and in modeling results reported here. However, our modeling results further suggested that, for many Adirondack lakes, chemical recovery might fail to continue in the future. We simulated that low-ANC lakes would actually reacidify under emissions control regulations in place at the time of development of this modeling effort. Both models suggested that most of the Adirondack lakes that are currently lowest in ANC (≤ 20 μeq/L) would begin to reacidify within approximately the next two decades under the Base Case scenario. This reacidification was attributable to projected continued declines in mineral soil BS within the lake watersheds. Continued chemical recovery was suggested, however, under additional emissions controls.
We developed empirical relationships between lakewater ANC and the species richness of zooplankton and fish, based on available data. These relationships were then applied to PnET-BGC model hindcast and forecast projections to generate estimates of the extent to which changes in species richness might accompany projected chemical changes. Using the empirical relationships between zooplankton species richness and lake ANC, the median inferred loss of zooplankton species from 1850 to 1990 was 2, with some lakes inferred to have lost up to 6 species.
Ignoring other factors that might influence habitat quality, we estimated that the median EMAP study lake had lakewater acid-base chemistry consistent with the presence of five fish species in 1850, prior to the onset of air pollution. Twenty percent of the lake population was estimated to have pre-industrial lakewater ANC consistent with supporting fewer than 4.1 fish species. By 1990, these median and 20th percentile values for estimated fish species richness had been reduced to 4.6 and 2.0 species, respectively. None of the emissions control scenarios suggested that the median Adirondack lake would gain more than 0.4 fish species by 2100, based on application of this empirical relationship to the PnET-BGC projections of future lakewater chemistry. However, 20% of the lakes (those most acid-sensitive) were estimated to change ANC to an extent consistent with a further loss of 1.3 fish species by 2100 under the Base Case, and a gain of 0.9 and 1.5 species under the Moderate and Aggressive Additional Emissions Control scenarios, respectively.
Model output comparison between MAGIC and PnET-BGC focused on site-by-site comparisons of simulation outputs, regional comparisons using cumulative distribution functions, and patterns of historical and future simulated ANC in comparison with variations in sulfur deposition and baseline (1990) values for ANC and mineral soil BS. In general, PnET-BGC estimated less historical acidification and less future chemical recovery as compared with MAGIC. This inter-model difference was most pronounced for lakes having ANC between about 50 and 150 μeq/L.
The MAGIC and PnET-BGC models differed somewhat in their assessment of how representative each of the modeled ALTM lakes are compared with the overall population of Adirondack lakes. For each modeled long-term monitoring lake, we estimated the percentage of lakes in the overall population that were simulated by each model to be more acid-sensitive than the subject lake. Both models estimated that the modeled ALTM lakes were largely among the lakes in the population that had acidified most between 1850 and 1990. Both models estimated that virtually all of the modeled ALTM lakes were in the top 50% of acid sensitivity compared with the 1,829 Adirondack lakes in the EMAP statistical frame. This result was found for projections of both past acidification and future recovery.
The results of this research will allow fuller utilization of data from on-going chemical and biological monitoring and process-level studies. A mechanism is provided for regionalization of findings. This approach was accomplished by developing/refining relationships among watershed characteristics, chemical change, and biological responses to changing levels of acid deposition. Such information is important for the management of the ecosystems in New York that are most responsive to changes in acid deposition.