Study Area and Sites
The lower reaches of the Schuylkill flow through a mix of agricultural, industrial and urban areas.
The Schuylkill River, which drains ~4950 km2 in southeastern Pennsylvania, flows for 211 km through several population centers (e.g., Pottsville, Reading, Pottstown, Phoenixville, Norristown) before its confluence with the Delaware River at Philadelphia. The Schuylkill basin is composed primarily of forest and agricultural land (~85% combined), with most remaining land classified as urban. The watershed has a complex topography and geology as it flows through three physiographic regions (i.e., the Ridge and Valley Province, New England Province (Reading Prong), and Piedmont Province).
In 1996, we identified 19 study sites that were located on most of the major Schuylkill River tributaries and included streams of varying sizes, physiographies, and land uses. Together, watersheds upstream of these sites represent 57% of total basin area. Watershed land use varied among sites: agricultural land ranged from 5% (Headwaters Mainstem Schuylkill) to 66% (Tulpehocken Creek); forested land ranged from 21% (Skippack Creek) to 75-77% (Angelica Creek, Headwaters Mainstem Schuylkill, West Branch Schuylkill); urban land cover was high for Wissahickon Creek (49%), Valley Creek (42%), Skippack Creek (24%), and East Branch Perkiomen (12%), but only 0.6-7% for most other sites; and mining operations were 7-9% of land cover for the West Branch Schuylkill and Headwaters Mainstem Schuylkill, 1.5-2.5% for Little Schuylkill and Valley Creek, and rare or absent for all other basins.
There are over 250 dams distributed throughout the Schuylkill River basin, including in our study streams (Poten 1996). These are generally low head dams that create relatively small impoundments, but there are three large reservoirs in the basin: Green Lane Reservoir on Lower Perkiomen Creek, Blue Marsh Reservoir on Tulpehocken Creek, Lake Ontelaunee on Maiden Creek. Most dams on the study streams were located either on upstream tributaries or >10 km from the actual study reach. The Site 12 was 150 m downstream from a low-head impoundment and several km downstream of Blue Marsh Reservoir while Site 17 was downstream of several small impoundments.
Each of these 19 sites integrates the upstream activities within each watershed, and represents water quality in the tributaries just before they enter the mainstem Schuylkill River. This assessment provides an overall evaluation of efforts to protect, mitigate, and restore these tributaries and watersheds. It also provides important perspective when examining new sites. These sites were sampled annually to provide a measure of annual variation in the stream communities and its impact on stream assessments, to generate the data needed to assess long-term changes in stream conditions, and provide simultaneous data to provide perspective when evaluating data from other sites. Beginning in 2001, we worked with local watershed associations and other interested parties to identify additional sites (12-31 per year, 127 total) to provide spatially detailed information in specific watersheds in the southwest (2001 and 2006), upper (2002), northeast (2003 and 2006), northwest (2004), Perkiomen (2005 and 2006), and northwest, southwest and northeast (2007) tributaries.
Sample Collection and Processing
Macroinvertebrate samples were collected annually (between late March and late April) for 12 years (1996-2007) at the 19 Long-Term study sites. The other 127 sites were sampled only once, between late March and late April from 2001 to 2007. Five quantitative samples were collected randomly from a single riffle at each site using a modified Hess sampler (0.088 m2 sample area; 500 µm mesh net). During sample collection, the substrate was disturbed and dislodged macroinvertebrates were collected in the sampler's net. Each stone (large gravel to cobble) in the sampling area was scrubbed with a brush and then inspected for any remaining attached macroinvertebrates, which were then removed by hand. Although substrates were generally similar across all samples (mostly small to large cobble), the quantity of leaves, woody debris, and sand/silt associated with the cobbles varied among sites and years.
After collection, macroinvertebrate samples were preserved in 95% Ethanol, transported to the laboratory, and split into four or more equal subsamples. Subsampling reduced the number of individuals examined from each sample towards our target of 200 individuals per subsample. Because macroinvertebrate densities varied greatly across sites and years, the subsample size processed varied from whole samples to 1/16 of a sample. Macroinvertebrates were removed from subsamples under magnification and identified to family (most insects) or order/class (most non-insects) using keys in Peckarsky et al. (1990) and Merritt and Cummins (1996). The samples were processed in the laboratory by 88 college students (5-14 summer interns per year) recruited from 43 colleges or universities. The students were provided initial training in macroinvertebrate identification techniques, and they quickly acquired the skills needed to collect data with taxonomic content similar to family-level data commonly collected by state and federal agencies. All identifications and counts in every sample were rechecked after initial processing.
Water samples were collected at each of the study sites (one sample per site) during baseflow. Nitrate (NO3), total dissolved phosphorus (TDP), pH, conductivity, and alkalinity were measured for each sample.