2.0 Methods
To date, few studies have investigated how the regulatory shift from the
CWR to the NWPR has impacted US watersheds (Meyer & Robertson, 2019;
Walsh & Ward, 2019; Walsh and Ward, In Review). We take NYS as a case
study (Figure 1), providing an assessment in the northeastern US that
spans multiple ecoregions and drains to multiple endpoints (including
the Great Lakes, Susquehanna and Delaware River basins, and the Atlantic
Ocean). The approach used in this study, following prior work of Walsh
& Ward (2019, In Review) and Meyer & Robertson (2019), leverages
several publicly available datasets to classify the jurisdictional
status of wetlands in New York. We quantified the federal protections of
more than 373,000 non-marine wetlands present in the US Fish and
Wildlife Service’s National Wetland Inventory (NWI), covering over
876,000 ha across the state of New York (U.S. Fish and Wildlife Service,
2021). We excluded wetlands classified as estuarine or marine in NWI
records. Wetlands smaller than 0.01 ha in size were also omitted from
further analysis. We selected this threshold to minimize the inclusion
of small wetland fragments adjacent to larger contiguous wetland
polygons. We summarize wetland protection statuses state-wide and at a
smaller regional scale (defined by USGS HUC–6 basins; Figure 1). Given
that most investigations of wetland policy are performed at the
watershed scale, our analysis allows for an evaluation of policy
outcomes across a range of broader spatial scales.
As the jurisdictional status of wetlands primarily stems from their
proximity to streams, the identification of the location of the stream
network is a key task in the evaluation of WOTUS protections. Previous
approaches (e.g., Walsh & Ward, 2019) used topographically-based flow
accumulation methods to estimate the spatial extent of the stream
network. We instead chose to define the stream network using the USGS
National Hydrography Dataset (NHD; 1:24,000 scale, Model Version 2.2.1),
as we could not ensure the reliability and accuracy of a
topographically-generated network at the geographic scale of NYS (U.S.
Geological Survey, 2020). While the NHD may underestimate stream network
extent (Elmore et al., 2013), the NHD does provide a consistent and
validated record of the location of streams within our study area and
across most regions in the US (Meyer & Robertson, 2019). We used NHD
waterways as the basis for the location of the stream network in both
the CWR and NWPR scenarios. This may result in a conservative estimate
of the jurisdictional stream network under the CWR as the version of the
NHD used here only maps perennial and intermittent streams in NYS (U.S.
Geological Survey, 2020). Thus, we assume the NHD network is
representative of the network with flow during a typical year for the
NWPR scenario and serves as a conservative lower bound for the stream
network in the CWR scenario.
To define the protection status of each wetland within NYS, we
translated the regulatory procedures of the CWR and NWPR into a set of
geospatial analysis steps within ArcGIS Pro (Version 2.6.0). Using this
process, we classified each NWI wetland as jurisdictional,
non-jurisdictional, or requiring the application of the significant
nexus test (abbreviated as ‘significant nexus’) for each regulation
based on the spatial relationships of each wetland to the stream
network. Under the CWR, wetlands located (1) within 100 ft (30.5 m) of a
jurisdictional water, (2) within both 1500 ft (457.2 m) of a
jurisdictional water and the 100-year floodplain, or (3) within 1500 ft
(457.2 m) of a traditionally navigable waterway are considered
jurisdictional by rule. As US wetland regulations are based around
imperial system units, we report these values first, followed by SI
units in parentheses. We applied a spatially continuous floodplain map
(Woznicki et al., 2019) to estimate the location of the 100-year
floodplain in our study area, as the commonly used Federal Emergency
Management Agency (FEMA) floodplain dataset was incomplete across NYS.
To identify the extent of traditionally navigable waterways, we used
publicly available streamflow information and channel geometry from the
USGS to create discharge-depth regression relationships across New York
State. We defined ‘traditionally navigable’ waterways as those receiving
at least a median of 0.30 m of inundation per year (following Walsh and
Ward, 2019). All wetlands that do not satisfy the aforementioned
jurisdictional criteria but are located within 4000 ft (1219.2 m) of
another jurisdictional water are subject to the significant nexus test
under the CWR. Given the challenge in applying the case-by-case
significant nexus test to broad spatial scales, we were unable to
determine the precise extent of jurisdictional wetlands when evaluating
the CWR. Instead, depending on the results of site-specific significant
nexus determinations made by the USEPA and USACE, the percentage of
jurisdictional wetlands under the CWR may vary considerably. Wetlands
located further than 4000 ft (1219.2 m) away from a jurisdictional water
are considered non-jurisdictional by rule and are afforded no protection
by the CWR. While not directly assessed in this study, the CWR also
considers the impacts of ‘similarly situated waters’, effectively
assessing aggregate impacts from several wetlands.
In contrast to the buffering approach used by the CWR, the NWPR defines
protections based on a surface-water connection between wetlands and the
jurisdictional stream network. The rule eliminates the significant nexus
test, instead classifying wetlands as either jurisdictional or
non-jurisdictional. For the NWPR scenario, wetlands were considered
jurisdictional only if they directly intersected the jurisdictional
stream network, simulating a surface-water connection. While we
acknowledge that this intersection approach does not necessarily
guarantee federal jurisdiction based on the NWPR, as we cannot confirm a
continuous surface connection at the scale of this analysis, we take
this as a reasonable estimate for protection.