Ecology of Headwater Streams in Prospect Hill Tract

• Investigators: Emery Boose, Elizabeth Colburn, Bridget Collins, William Sobczak
• Contact: Elizabeth Colburn
• Start date: 2003-06-01
• End date: ongoing
• Location: Prospect Hill Tract (Harvard Forest)
• Latitude: +42.53 to +42.55
• Longitude: -72.20 to -72.17
• Elevation: 280 to 420 meters
• Taxa:
• Keywords: aquatic invertebrates, carbon cycling, headwater streams, hydrology, intermittent streams, stream ecology, stream salamanders, streamflow, water quality
• Abstract:

  We have initiated long-term monitoring of streamflow, water quality, litter inputs, decomposition, periphyton, and aquatic fauna in headwater streams at the Harvard Forest's Prospect Hill Tract. Key research questions include the following. (1) How do differences in forest composition and historic land use influence stream hydrology, chemistry, ecosystem functions, and fauna? (2) How do streamflow differences between adjacent headwater streams within the same watershed translate to differences in ecosystem functions and aquatic fauna? (3) How stable are aquatic communities of headwater streams in relation to year-to-year variations in streamflows? (4) How will changes in forest composition, in response to invasion by the hemlock woolly adelgid, affect headwater streams? (5) To what extent to headwaters influence water quality and ecosystem processes in downstream, perennial stream reaches?

• Methods:

  Stream mapping

  Each stream channel is flagged at 20-m intervals. 0 represents the most downstream point on the study stream; successive numbers represent 20-m distances upstream from the 0 point to the seasonal head of flow.

  Watershed assessments

  Baseline field assessments from the 0 point to the top of the watershed include watershed delineation; quantification of dominant trees in the watershed using standard plotless methods applied in ongoing ecosystem studies of the effects of the hemlock woolly adelgid; assessment of watershed land uses; and, in hemlock stands, assessment of the status of hemlock woolly adelgid infestation and impacts.

  Instream habitat

  Using a modification of the field procedures established by Ohio EPA (2002) for assessment of headwater streams, we record habitat characteristics such as flow, gradient, substrate (percent boulder, cobble, sand, leaf dams, etc.), channel conditions (pool, riffle, cascade, etc.), riparian vegetation, and adjacent land use for each 20-m reach of each tributary.

  Water quality

  In a lower reach and at the head of flow we are continuously monitoring temperature on the stream bank and in the channel with Thermochron© i-buttons (Dallas Semi-conductor Company, Dallas, TX). Periodic sampling of water quality includes: temperature, specific conductance, and pH (YSI Meter Model 63); dissolved oxygen saturation (YSI Meter Model 57); Cl, NO3, PO4, SO4, Ca, K, Mg, Na, using a Lachat QuikChem 8500 with integrated ion chromatograph (Lachat Instruments, Milwaukee, WI); and DOC via high temperature combustion) (Sobczak and Findlay 2002). In summer, 2004, a transect of PVC wells was installed across upper PHA for comparison of water quality in subsurface and surface sections of the stream.

  Streamflow

  Flow conditions including water depth and flooded channel width are noted on each sampling date.

  Weekly monitoring of streamflow in a subset of the streams in summer, 2003, involved recording flow characteristics in each 20-m reach (e.g., continuous flow, flow interspersed with stagnant water, pools interspersed with waterless stretches, dry channel).

  In fall, 2004, stream gauges were installed in three tributaries, Nelson Brook A and E (BGA and BGE) and Bigelow Brook west (PHA). In these tributaries, periodic measurement of water depth at the gauges is providing more detailed information on stream discharge patterns. Automation of the gauges is planned for 2006.

  Carbon sources

  Periphyton biomass is measured on ceramic tiles placed into each tributary and sampled quarterly (Sobczak 1996, Steinman and Lamberti 1996). Leaf litter inputs are being measured using litter baskets (n = 5+ per stand type, depending on watershed area), placed pointing upward (for direct leaf-fall) and sideways (to account for lateral transport). Cover by mosses and aquatic plants is sampled along transects perpendicular to the stream channel in summer.

  Fauna

  In a subset of the study streams, we randomly established quadrats for macroinvertebrate sampling and transects for intensive visual searches for salamanders in 50-m stream reaches.

  Habitat characteristics of the quadrats and transects are quantified using the same procedures followed for reach-level habitat assessments. Each salamander transect is 1 m wide and with a length of 4x the mean stream width at the reach. Where the channel is more than 1 m wide, the side of stream to be followed is determined randomly.

  In reaches without standing water, leaf-dams, woody debris, and rocks are carefully searched and bank vegetation and refugia are examined for the presence of stream salamanders. Where water is present, potential refugia are lifted and the water swept with a D-frame aquatic net. All salamanders are identified and released back into the stream.

  Macroinvertebrates are sampled in separate, randomly located, 1 m2 quadrats in each 50-m reach by repeatedly stirring the substrate and sweeping dislodged organisms into a D-frame aquatic net. The samples are decanted into a white pan filled with water and picked live in the field. With the exception of amphibians, which are identified and released, all macroscopic organisms captured in the quadrats are preserved in 70 percent ethanol and returned to the laboratory for identification. Trophic classifications of identified organisms are based on data in Merritt and Cummins (1996).

• Related datasets: HF070 HF095 HF099