One of the objectives of this study was to find ways to use the GIS data layers that were collected to answer large scale geographic questions about the Park's water and wetland resources. These questions include the vulnerability of lakes to acid precipitation, the potential effects of large scale disturbances on watershed functioning, and the use of the layers to identify the probable location of vulnerable ecosystems such as peatlands in the Park. Our GIS analysis sought to examine the following questions:

Atmospheric Deposition

What is the geographic distribution of lakes sensitive to atmospheric deposition, as measured by lake ANC, watershed geology and soils?

Is there a simple relationship between lake ANC and the soil or geology properties of the study area?

Lake Eutrophication (Productivity)

What are the properties of watersheds in the study area with respect to factors which contribute phosphorus to waterbodies: land use; landcover; shoreline development; and disturbance to the natural vegetation?

Is there a simple relationship between lake phosphorus concentrations and the watershed properties: land use; land cover; shoreline development; or disturbance to the natural vegetation?

Wetland Sensitivity

Can GIS be used to help locate peatlands (sensitive wetlands)?

Approach

The map coverages were used to evaluate two issues related to lakes and wetland watershed interactions. We examined lake/watershed sensitivity to atmospheric deposition by comparing lake acid neutralizing capacity (ANC) with watershed geology and soils. We examined lake eutrophication by using the 1989 residential development coverage and the APA land use coverage to compare lake phosphorus concentrations with watershed and shoreline development. To evaluate major landscape disturbances over the last century with respect to both lake eutrophication and acidification, we created a watershed disturbance composite from the history of major fire and storm disturbances. Lastly, we examined the question of wetland sensitivity to the addition of nutrients and acidity. From a review of the literature, we discuss identification of peatlands from other wetlands as well as atmospheric loading of nutrients and acidity to wetlands in the Adirondacks compared to northern Europe. The discussion on peatlands identification and peatlands sensitivity appears in Section III. Our findings on the other issues are discussed below.

ANC or acid neutralizing capacity is defined by Munson et al. 1990 as "a measurable parameter which indicates the net strong base in solution or net mineral acid if its value is negative. It is routinely measured by titration with mineral acid or base to an equivalence point (inflection point). The higher the ANC of a water, the more mineral acid is required to reduce the pH to a given equivalence point." Generally, ANC is used as an indicator of lake sensitivity to atmospheric deposition. In the Adirondack Park and elsewhere, it is considered a key indicator of aquatic ecosystem health (USEPA 1995, Moynihan and D'Amato 1997).

Of the total 1223 lakes identified in the Oswegatchie/Black study area, lake ANC data were available from the Adirondack Lakes Survey Corporation (1987) for 508 or 42% of the waters. These lake ANC values were categorized by lake status shown in Table IV.A.1. ANC values in the study area ranged from -79.6 to 2671 micro-equivalents per liter (ueq/l).

In ArcView, the lake categories were displayed across the study area by coding or coloring in the "immediate" lake watershed polygons only. For simple watersheds, these immediate lake watershed polygons represented the entire watershed. For complex watersheds, which were defined as watersheds which drained two or more ponded waters, however, it only represented the area directly draining into that lake, not the area contributing from lakes upstream. We used this immediate watershed approach for several reasons: first, for simplicity. Many of the watersheds are very large and complex, at times draining more than one hundred other lakes. It was therefore impossible to view them altogether using whole watersheds due to overlapping. Additionally, we believed that from a soils and geology perspective of lake ANC, these immediate watersheds would have the most influence on lake chemistry, so that it was reasonable to examine these first. We recognized, however, that this was unconventional from a hydrology standpoint.

Results by visual examination indicate that all four classes of lake ANCs are evenly distributed across the study area. For the entire study area, 688 out of a total of 1223 waters (56%) are simple watersheds. These simple watersheds appear evenly distributed throughout the study area and drain a total of 71,784 hectares or approximately 18% of the total Oswegatchie/Black study area.

The next step involved overlay of the bedrock ANC composite on the lake ANC polygon categories to determine any patterns between lake ANC and bedrock ANC. For example, would the low ANC lakes match with the occurrence of bedrock of low to no ANC capacity? The intersects were tabulated by area totals. Table IV.A.2 shows for each lake ANC class, the amount of immediate lake watershed area in each bedrock class.

Results showed that all four lake ANC classes contain both extremes of bedrock ANC capacities. No strong correlation between lake ANC and the occurrence of underlying bedrock of corresponding ANC capacities in their immediate watersheds was demonstrated.

For the purpose of trying to find a simple method to locate non-monitored lakes with low ANC using only the existing bedrock geology map, the results were disappointing, but understandable. It was expected that it would be necessary to include other factors influential to lake chemistry, such as surficial geology. Since similarly interpreted surficial geology data were not available, we decided to focus only on areas with thin soils using the general soils parent material composite. The following describes our approach.

Two categories were selected from the soils coverage to represent shallow depth to bedrock soils: Rock Outcrop Dominant and Shallow Soils Dominant. These were combined and positioned over the high ANC bedrock and again over the low ANC bedrock, each time plotting the location of each of the four lake ANC classes. We expected that if there was a strong correlation in shallow soils, then High to Infinite ANC bedrock would have presented a higher proportion of Adequate ANC lakes and the No to Low ANC bedrock would have presented a higher proportion of low ANC lakes. The results showed neither. Visually, both plots exhibited an even distribution of lake categories, suggesting no correlation. This indicated that when using these existing coverages, bedrock alone, even in areas where shallow soils predominate, cannot be expected to predict lake ANC. We recognize that differences in data scales, however, could also be a significant factor in the lack of correlation. We conclude that at this scale of available data, it was not possible to predict where the low or high ANC lakes would occur by using only bedrock geology at 1:250000. We also found that using additional existing soils data did not improve the predictability. We believe, however, that efforts to predict lake ANC are valuable, particularly in light of air quality policy implications for surface water quality protection in the Adirondack Park (USEPA 1995, Moynihan and D'Amato 1997). The identification of vulnerable surface waters in the absence of a complete census of lake chemistry would be a valuable tool in the midst of implementation of the Clean Air Act Amendments of 1990. As better data with finer resolution become available, these efforts should be continued and the results shared with others including EPA's Acid Rain Program staff.

Lake phosphorus with land use

Phosphorus is considered the limiting nutrient for algal growth in New York State lakes, and hence is a key indicator of lake productivity in the Adirondacks. It is analyzed in most lakes as total phosphorus (NYSDEC and Federation of Lake Associations, Inc. 1990).

Of the total 1223 lakes identified in the study area, lake phosphorus data was available for 508 (42%) waters (Adirondack Lakes Survey Corporation 1987). These summer lake phosphorus (P) data were categorized by lake trophic condition and the results are shown in Table IV.A.3. The values across the study area ranged from undetected to 0.233 parts per million (ppm).

Lake phosphorus categories were displayed across the study area using the "immediate" lake watershed polygons. Results by visual examination indicated that all three lake categories were evenly distributed across the study area.

We were interested in examining watersheds of lakes in the lowest phosphorus category because these are considered potentially the most sensitive to change from nutrient additions. We developed a method for quantifying watershed development to determine correlations or waters at risk using the available data coverages. We had an interest in both development on the immediate shoreline and development in the whole watershed. Because the study area contained a range of watershed sizes and complexities to consider, we began by examining a subset of lakes in the study area. We chose the 23 Adirondack Long Term Monitoring (ALTM) waters, which have the longest detailed record of surface water chemistry in the Oswegatchie/Black watershed. The ALTM watersheds are depicted on the poster graphic attached to this report.

We established the following procedure to estimate the amount of existing development along the shoreline of the lake.

1. Residential buffers (points) were selected from the 1994 residential areas coverage. A reselect was done for those built upon using the assessor's class code. A distance of 402.25 m (0.25 miles) from the shoreline was picked as a shore area limit based on an examination of existing development in 6 ALTM lakes using tax maps, aerial photos, and local knowledge of patterns of shoreline development.

2. Total shoreline length was determined from USGS surface water coverage (1:24000) which is part of APA Land Use and Development Plan coverage. The lake surface perimeter polygon was closed as needed either visually or using the default.

3. The points (from Step 1) were "pulled" to the nearest point along the shoreline. A radius length of 50' was assigned for each point, as representing 100' (30.8 meters) of developed shoreline.

4. The developed lengths were added and sum was divided by the overall shoreline length to get percent shoreline developed.

Results for the 6 ALTM lakes that had shoreline in private ownership showed percent shoreline developed as residential ranged from 1 to 15% (Table IV.A.4). Next, total watershed development was evaluated. Using existing data we found the simplest approach was to consider the percent watershed as Forest Preserve lands. Generally, with few exceptions, lands in the New York State Forest Preserve are not developed nor can they be developed for residential use. We chose, therefore, to use percent area of watershed as Forest Preserve as an indicator of watershed development. This was done using the APA Land Use and Development Map coverage. For the 23 ALTM waters within the Oswegatchie/Black study area, 17 were located entirely within Forest Preserve lands (Table IV.A.4).

This evaluation was extended to the 171 oligotrophic waters found within the study area. These oligotrophic waters defined by a phosphorus concentration of less than 0.010 ppm, represent the lakes potentially most vulnerable to nutrient additions. Of these, we found 81 waters or 46% had watersheds located entirely within the Forest Preserve. For all 171 oligotrophic waters, the average amount of watershed in Forest Preserve was high at 75%, ranging from 1.5 to 100 percent. A significant number of waters (105) contained a high proportion (90% or more) of their watersheds in Forest Preserve.

We concluded that low phosphorus oligotrophic lakes are widely distributed within the Oswegatchie/Black study area. A significant number of them contain a high proportion of Forest Preserve Land in their watersheds. Only 6 out of the 23 ALTM lakes found in the study area contain shoreline in private ownership, and the total percent of residential shoreline development on those lakes is low.

We expect to continue refining this method of quantifying existing residential development on shoreline as it is applied to other Adirondack waters.

Major landscape disturbances of lake watersheds

In the Adirondacks, major landscape disturbances such as fires and storm blowdowns continue to be discussed in the mix of determinants of lake chemistry (Sullivan et al. 1996). The landscape disturbance composite, discussed in Section II.O of this report, was a combination of major episodes of forest disturbance during the last century, including logging, fire and the blowdown caused by the storm of November 25, 1950. The disturbance composite was examined with the lakes watershed coverage, the wetlands coverage and the available ALSC lake chemistry data.

For the Oswegatchie/Black watershed, only a relatively small portion (25,786 hectares or 6.5%) of the total study area was considered undisturbed. This undisturbed area touched upon parts of 180 different watersheds, but only 12 watersheds were entirely undisturbed. We were able to identify these, determine total watershed areas and search for the availability of ALSC lake chemistry data. The 12 undisturbed watersheds were relatively small ranging from 2.6 to 131.3 ha with an average area of 42.6 ha. Ten ponds had ALSC lake chemistry data. For these lakes, ANC ranged from -56 to 149 ueq/l and dissolved organic carbon ranged from 3.2 to 8.8 mg/l. ANC values for all the ALSC waters in the study area ranged from -68.7 to 2671.0 ueq/1. DOC values for the study area ranged from 0.2 to 32.8 mg/l. From the wetlands coverage, wetland area as percent total watershed for the ten ponds ranged from 0 to 29%.

Adirondack Long Term Monitoring (ALTM) Watershed Characteristics

A matrix of watershed attributes was prepared for the 23 ALTM waters from each major coverage grouping that was created as part of this study (Table IV.A.4). The matrix is also included as part of the graphic insert attached to this report. The purpose of this exercise was to illustrate what types of data could be extracted from the coverages in this catalog. To evaluate watershed landscape effects on both lake and wetland water quality, we tried to develop meaningful development indices, particularly along lake shorelines. We expect that watershed boundaries for wetlands will be available when the Digital Elevation Models (DEM) are done for the Adirondacks.

The ALTM lakes exhibit a wide range of lake ANC from -45.8 for Squash Pond to 149.2 ueq/l for Windfall Pond. Windfall Pond is also the smallest watershed at 41 ha, with Lake Rondaxe the largest at 14156 ha. From the Oswegatchie/Black Phase I study, percent watershed as wetland for the ALTM lakes was not strongly correlated with either lake dissolved organic carbon (DOC) or lake ANC.

Using the TM data, percent watershed as forested ranged from 64 to 91%. The forested categories selected were Deciduous, Conifer and Mixed. All of the ALTM lake watersheds drained some proportion of State Forest Preserve lands, ranging from 41 to 100%. Nearly half of the watersheds are located entirely on State lands. We expect that more work will be done with these study waters and the other watersheds as the catalog of coverages gets distributed and used. Some studies have already been initiated.