WATERSHED PROTECTION OF THE ST. LAWRENCE RIVER WATERSHED
WITH SPECIAL CONSIDERATION TO LARGE WETLANDS AND LARGE LANDOWNERSHIP
PART ONE: THE ST. REGIS RIVER BASIN

On this page:
Sub-watershed Mapping
Watershed Mapping of Large Peatlands
Wetlands Mapping
Landowner Outreach

METHODS

The methods used to map watersheds and wetlands in the St. Regis River Basin were based on those used for previous Adirondack Park watershed projects (Roy et al. 1996, Roy et al. 1997, Primack et al., 2000). These maps and others created from existing data from the APA's Geographic Information System (GIS) were used in landowner meetings.

Sub-watershed Mapping
Sub-watersheds of the St. Regis River Basin were delineated for major rivers and ponds with unique pond identification numbers assigned by the Adirondack Lake Survey Corporation (ALSC) and the New York State Bureau of Fisheries (BOF). Watersheds were delineated on USGS 1:24000 or 1:25000 topographic maps using conventional methods (USDA 1977). Separate sub-watershed boundary determinations were made by two different people. A third person at the APA reconciled and verified the maps using 1:40000 NAPP or 1968 1:24000 panchromatic air photos and field checking. Final sub-watershed boundaries were transferred to clean USGS topographic maps and digitized, edge-matched, and joined into a seamless coverage.

Air photos and topographic maps were used for two independent determinations of flow pattern between sub-watersheds and directional arrows were drawn on paper copies of the boundary maps. Each paper map was used to digitally assign hierarchical flow direction labels to the GIS coverage of sub-watersheds using an ArcInfo AML (Arc Macro Language program) created by Primack (1997) (Appendix 8). The AML assigned each sub-watershed polygon a unique I.D. number in a data field called 'THIS#' in the polygon attribute table. A second field, called 'FLOWTO#', was created that contained the I.D. number of the receiving sub-watershed polygon. In this way, the path of water from one sub-watershed polygon to another could be followed through the entire St. Regis Watershed.

To uncover inconsistencies between flow direction assignments in the two independently produced flow-pattern maps, the polygon attribute tables were relationally joined using the 'THIS#' field. Where the 'FLOWTO#' field from one of the coverages did not match the 'FLOWTO#' field from the other coverage, the correct flow direction was ascertained by overlaying the sub-watershed polygon outlines over GeoTIFF images of 1:24000-scale USGS topographic maps (NYS GIS Clearinghouse 2000a) in a GIS. If evidence of flow direction could not be determined from the topographic maps, 1:40000 NAPP air photos or 1978 1:24000 panchromatic photos were viewed in stereo. In addition, the ALSC Pond database was a useful source that gave outlet information for most of the ponds in question (ALSC 1999). Differences between the two delineation efforts primarily occurred in areas where there was little topographical variation or where the pond outlet was uncertain, especially where the watershed was isolated and internally draining. After the flow direction hierarchy was finalized, regional watersheds were generated from the sub-watershed map using an ArcInfo AML (Appendix 9, Primack 1988) to select all watersheds flowing into a particular watershed and grouping them.

As a final step in the creation of the sub-watersheds coverage, a field was added to the polygon attribute table that contained the 11-digit USDA/SCS watershed codes (USDA/SCS 1980). This will allow the data set to be joined with other data sets that use the SCS codes for reference.

Watershed Mapping of Large Peatlands
In a related effort, detailed watersheds of the four largest boreal peatland complexes in the St. Regis River Basin were delineated using 1:40000-scale NAPP color infrared air photos using the same methods as for lake watershed mapping. These were: Madawaska Wetland, Waverly Bog, Osgood River Muskeg, and Spring Pond Bog.

Fig. 3
Figure 3. Watersheds were delineated separately for each section of peatland in the Osgood River Muskeg wetland complex.

The purpose of the effort was to determine the watershed area influencing the peatland portion(s) of the different wetland complexes. For some wetlands, such as the Osgood River Muskeg (Figure 3), rivers or streams flowing in well-defined beds or banks bisected the wetland. In this case, the bog mat was considered hydrologically isolated from the bisecting stream, and it was decided that the detailed watersheds of the wetland should be delineated separately for each portion of the mat. Although the each bog mat portion flows into the same river, their upslope contributing area is different. Watershed boundaries originated at the bog mat pour point, of which there could be many in the case of a raised peatland, and were drawn along the edges of bisecting streams based on differences in vegetation. Similar to the methods described for mapping sub-watersheds, the detailed watersheds of large wetlands were mapped independently by two people, compared and verified by a third, transferred to a 1:24000-scale basemap and then digitized. The final detailed watershed maps illustrated the hydrologic relationships between large wetlands and the surrounding landscape.

Wetlands Mapping
Wetlands were delineated on 1:40000 NAPP aerial photo acetate overlays and labeled based on the National Wetlands Inventory (NWI) classification scheme (Cowardin et al. 1979). The NWI labeling system takes into account the fact that a wetland may be comprised of more than one cover type. Therefore, labels are assigned to the two most abundant cover types in the wetland (each with at least 30% areal coverage) in the fields 'Class1' and 'Class2,' with 'Class1' always being the taller structure (i.e. FOx, as a forested wetland would always be listed before SSx, a scrub-shrub wetland). Class designations were further modified by codes for hydrologic regime or other special conditions ('Regime' and 'Special' fields, respectively). Before transferring the delineations and labels to Mylar overlays on 1:24000 orthophoto basemap quadrangles using an IIS Stereo Zoom TransferScope they were examined by a second photo interpreter as a first QA/QC check. The wetland polygon Mylars were scanned as TIFF images and vectorized using the software R2V (Able Software Co., Lexington, MA). Vector lines for each file were printed and compared with the mylar polygon overlay. Labels were digitized onto the polygon coverage and sections of the full NWI label were entered into separate fields in the polygon attribute table (data fields: System, Class1, Class2, Regime, Special1, Special2, and Special3). Adding labels to the polygon coverage was used as the final QA/QC check because it allowed for missing or extraneous lines or polygons to be discovered and corrected. After the labels were checked against the original NAPP photo overlay labels, the digitized wetlands data files were considered complete (for additional QA/AC measures see
Appendix 6).

Although the St. Regis Watershed spans 19 7.5-minute quadrangles, wetlands were delineated on only 18 of the quads. This was because the aerial imagery for the Carry Falls Reservoir quadrangle was not available. The area of the St. Regis River Basin falling within the Carry Falls Reservoir quad is 4463 acres (1806 ha), or 1.3% of the watershed. If available, this area will be delineated during the St. Lawrence II project.

Landowner Outreach
To provide natural resource maps to landowners in the St. Regis River Basin, including sub-watershed and wetlands data, it was necessary first to select the landowners, and then to choose the data layers that would target their management concerns. The Adirondack Nature Conservancy and Land Trust (ANCLT) selected three landowners whose extensive properties are used for recreation, timber management, and wildlife habitat. Two of the properties, 1,606 acres (650 ha) and 21,244 acres (8,597 ha) in size, were already under conservation easement with ANCLT; the third, 2,540 (1,028 ha) in size, was considering an easement.

Existing and new GIS data layers were selected based on their relevance to the land use practices of the landowners and were made into thematic maps on a watershed-wide scale and property scale (Table 2) and provided in paper form in map packets and on CD-ROM. ANCLT digitized two of the property boundaries from Franklin County paper tax maps so that map data could be clipped to the land parcel; the third came from the Northern Forest Lands Inventory (NFLI, year unknown). Maps created from existing data included soils, slope, erosion potential, and NAPP orthoimagery. New GIS layers created for this project were sub-watersheds and wetlands.

APA and ANCLT held meetings at the APA in Ray Brook, NY with the two landowners already under conservation easement to discuss the map packets. The various APA and ANCLT attendees had different types of involvement with the project (some made the maps, some had worked closely with the landowner, some had visited the property before), which improved the quality of discussion.

The first meeting was made with a landowner who had GIS experience. His opinion on the maps was solicited in terms of their readability and interpretations, and improvements were made based on his suggestions. The meeting with the second landowner led to a follow-up meeting with him at his property for field checking. ANCLT met separately with the third landowner not yet under easement.

During the meetings the maps in the map packets were discussed in detail allowing enough time for the landowner to share his observations from the real-life landscape. Interpretations describing data development and the ways in which maps could be used as decision-making tools were provided in the map text. Legend colors used in maps of previous Adirondack Park watershed projects were used in this project for consistency.

Because of the widespread use of GIS as a land management tool, it was decided that the spatial data should be provided for the landowners in digital form on CD-ROM as well as in paper form. These were organized in a user-friendly way using hypertext pages that could be read in a web browser. JPEG images of the paper maps that could be re-printed were provided for users without access to GIS software. GIS users with access to the software ArcView or ArcInfo (Environmental Systems Research Institute, Redlands, CA) could work with the data to create new maps. For the most part, data on the CD-ROM was in ArcView shapefile format, however the Grid themes and some property boundary layers were in ArcInfo coverage format.

The methods used to provide existing data and new sub-watershed and wetlands data to landowners built upon the methods used in previous Adirondack Park watershed projects. Current GIS technology partnered with landowner meetings offered an excellent mode for data dissemination and outreach.

Table 2. Thematic Maps created for landowner outreach. Map packets were made for three different properties within the St. Regis watershed.

St. Regis Watershed-wide Maps

Map Title Scale Data Layers Depicted Description
St. Regis Watershed Elevation 1:175,000 •10m Digital Elevation Model (DEM)
•1:250,000 Hydrography
•St. Regis watershed boundary
•Adirondack Park boundary within the watershed
•Franklin-St. Lawrence county boundary
•Towns
•Major Roads
•Property boundary
Shaded relief map of watershed. Used to orient observer to property location within the watershed.
Adirondack Park Land Classification in the St. Regis Watershed 1:175,000 •Adirondack Park Land Use and Development Plan Map
•1:250,000 Hydrography
•St. Regis watershed boundary
•Adirondack Park boundary within the watershed
•Franklin-St. Lawrence county boundary
•Towns
•Major Roads
•Property boundary
A map of land classifications in the watershed. How does land classification on the property compare to the whole watershed? How are neighboring landowners managing their property? Acreage and percent of each land class provided in the legend.
Sub-watersheds of the St. Regis Watershed 1:175,000 •10m DEM
•1:250,000 Hydrography
•Sub-catchment map created for this project
•St. Regis watershed boundary
•Adirondack Park boundary within the watershed
•Franklin-St. Lawrence county boundary
•Towns
•Major Roads
•Property boundary
Illustrates what might be affecting the property upstream, or how the land use practices on the property might affect downstream areas. DEM (displayed with shaded relief) and hydrography layers help to show which way the watersheds drain.
Predicted Nitrogen Deposition in the St. Regis Watershed 1:300,000 •Nitrogen Deposition model (931m pixel size)
•St. Regis watershed Boundary
•Property Boundary
Shows nitrogen deposition trends in the watershed. The map was taken from a larger digital map of deposition in the northeastern U.S. created by Ollinger et. al (1993). The data were created from a simple model using elevation and precipitation to predict yearly N deposition. The map gives a wider perspective on factors potentially affecting timber stands and water quality on the property.
Predicted Nitrogen Deposition in Adirondack Park 1:1,200,000 •Nitrogen Deposition model (931m pixel size)
•Adirondack Park Boundary
•St. Regis watershed Boundary
•Property Boundary
Displays nitrogen deposition trends throughout the entire Adirondack Park. See preceding map’s description.

Property-specific Maps

Map title Scale Data Layers Depicted Description
Sub-watersheds of the St. Regis River Basin in the vicinity of [Property Name] variable depending on property size (1:150,000-1:40,000) •10m DEM - shaded relief
•1:250,000 Hydrography
•Sub-catchment map created for this project.
•St. Regis watershed boundary
•Property boundary
Displays watersheds in proximity to the property, both upstream and downstream of the property.
[Property Name] Elevation variable depending on property size (1:70,000-1:20,000) •10m DEM - shaded relief
•USGS 1:24,000 Hydrography
•Property Boundary
Orients observer to the property with a landscape-wide view.
[Property Name] Slope variable depending on property size (1:70,000-1:20,000) •Slope, derived from 10m DEM - shaded relief
•USGS 1:24,000 Hydrography
•Property Boundary
Slope % classified into 5 range intervals: 0-3%, 3-8%, 8-15%, 15-25%, 25+%. Acreage and percent of each range interval on the property provided in the legend. Associated table derived from USDA-NRCS Slope Potentials and Limitations ratings provided for interpretation (USDA-NRCS 1999).
[Property Name] Soils variable depending on property size (1:70,000-1:20,000) •Adirondack Park MesoSoils Map (orig. scale 1:62,500, pixel size 63.6m)
•USGS 1:24,000 Hydrography
•Property Boundary
Shows generalized soils/soil associations occurring on the property. Acreage and percent of each soil type on the property provided in the legend. USDA-NRCS Official Soil Series Descriptions provided for interpretation (USDA-NRCS 1998).
[Property Name] Soil Erodibility variable depending on property size (1:70,000-1:20,000) •Adirondack Park MesoSoils Map displayed by erosion potential rating (orig. scale 1:62,500, pixel size 63.6m)
•USGS 1:24,000 Hydrography
•Property Boundary
Shows areas where erosion potential is high and care may be necessary for timber harvesting and recreational activities. Soil erosion potential is calculated from the soil K-factor (soil susceptibility to erosion and rate of runoff) multiplied by slope. Classified into 5 ratings: very severe, severe, moderate, slight, low.
[Property Name] Forest Productivity variable depending on property size (1:70,000-1:20,000) •Adirondack Park MesoSoils Map displayed by forest productivity rating (orig. scale 1:62,500, pixel size 63.6m)
•USGS 1:24,000 Hydrography
•Property Boundary
Provides observer with an understanding of productive/less productive/non-commercial soil locations on the property. Productivity ratings developed by Craul (1987). Based on soil profile drainage, organic horizon thickness, effective rooting depth, texture of the profile, degree of profile development, and profile stoniness.
[Property Name] Orthoimages variable depending on property size (1:70,000-1:20,000) •USGS NAPP 1994-1997 digital orthophoto quarter quadrangles (1m pixel ground resolution)
•Property Boundary
Detailed view of the property at a specific point in time: forest composition, wetlands, beaver dams, buildings, roads, skid trails, etc.
[Property Name] Land Cover variable depending on property size (1:70,000-1:20,000) •30m NYS GAP Analysis Map
•Property Boundary
Land cover data derived from LANDSAT imagery by the NYS GAP Analysis program at Cornell University (Smith et al. 2000).

Continue to next section of St. Regis Report: Results
Return to table of contents
References Cited