A Study of Benthic Macroinvertebrates in the Salmon River, New York, in Relation to Hydrologic Regime and Channel Gradient

John L. Hallock Jr.

What are benthic macroinvertebrates?
     These are the insects, crustaceans and other small invertebrates dwelling on and under the rocks and sediment of streams and lakes.
      In streams, the benthic macroinvertebrate community is dominated typically by insects such as mayflies, stoneflies, and caddisflies.

Why Study Benthic Macroinvertebrates?
They are relatively immobile and short-lived (less than a year to several years), and are sensitive to changes in their environment. They thus make good indicators of habitat conditions, water quality, and overall ecosystem productivity - both from place to place and over time.

By consuming algae, land-derived materials such as leaves, and detritus, and being eaten by fish, waterfowl and even mammals, benthic macroinvertebrates are the major intermediaries of stream foodwebs.

Premises of the Study - Watershed and Reach Influences
What is a watershed? The surface and subsurface area that water lands on and flows through toward a common outlet (confluence with another stream or lake).

The physical and chemical characteristics of streams are reflections of the characteristics of the watershed producing the stream. How big is the watershed? Is the channel steep? Is forested or agricultural? Is it subject to frequent storms? Is the geology dominated by bedrock outcrops or glacial till? Are the soils thick? Is there a lot of woody debris?

The answers to such questions determine the character of the stream, its volume and velocity of water flow, the character of the rock/sediment of the bottom, and the growth of algae, and input and retention of leaves and other organic matter on the stream bottom. The character of a stream can change over short distances (termed reaches), making the overall character of a stream the sum of both local and larger-scale factors.

Macroinvertebrates do not know what watersheds or stream reaches are. They are aware only of their local habitat - the temperature, dissolved oxygen, water velocity, type of bottom, availability of food, etc.

These things are determined by the character of the watershed and reach, though. As watershed or reach conditions change in certain ways, so will the local water temperature, amount of shade or sun, water velocity, etc. If the change in local environment is great enough, changes in the macroinvertebrate community should result.

How do Reach and Watershed Factors Affect Habitat? - Fast vs. Slow

Food - Higher water velocities associated with floods and /or steeper channels can dislodge leaves and other detritus, or keep them from settling on the bottom, and also scour algae from rocks. When present on or beneath the rocks, these materials serve as food for certain macroinvertebrates, thus acting as the base of the food chain in streams - the source of all food. Less food should mean fewer invertebrates.

Physical habitat - When water velocities are consistently fast enough, macoinvertebrates that are not adapted to hold position in swift current will be rarer, and the community may be dominated by species with low profiles to reduce drag, or ones that can anchor themselves to the rocks in those areas.

How do Reach and Watershed Factors Affect Habitat?

Open vs. Shaded
Shading of streams by trees or hillsides fosters cooler water temperatures, while extensive open stretches upstream of a given point foster warmer temperatures due to warming by the sun. Some groups of macroinvertebrates are adapted to warmer waters compared to others.

Open waters receive more sunlight, potentially fostering greater photosynthesis and algal production on the stream bottom. One might expect greater numbers or mass of macroinvertebrates specialized to feed on algae from the rocks in such areas.

A stream reach with extensive forest cover upstream will also receive greater leaf inputs than a section with fewer trees upstream. In the former section one would expect to find more macroinvertebrates specialized to feed on leaves and pieces thereof.

Rough vs. Smooth, and the retention of leaves and detritus
     While a source of leaves is the important first step in providing food for many macroinvertebrates, the leaves must settle or be retained in a stream section to be available as food. Large, stable rocks or branches can catch leaves directly, and also create eddies and localized areas of slower water that allow leaves to settle into crevices - where they are less susceptible to washing away.

All else being equal, a stream reach with a rougher channel or more branches should retain more leaves and organic matter, and thus produce more invertebrates.

Background on the Salmon River System
The Salmon River in New York State drains an approximately 285 square mile watershed encompassing portions of Oswego, Lewis, and Jefferson counties. It comes off the Tug Hill Plateau region to eventually empty into the Eastern shore of Lake Ontario near Pulaski, New York. The river was dammed in 1914 and 1930 for hydroelectric power generation, creating the Upper and Lower Salmon River Reservoirs, respectively. The area is largely rural in character, with agriculture and hardwood deciduous forest dominating the landscape.

Water released from the dams is drawn from the upper layers of the reservoirs, the same layers warmed by extensive exposure to sunlight. Top-release reservoir systems such as this generally result in water below the dams being warmer than water in reaches above the reservoirs during spring through fall.

Recreational fishing and snowmobiling contribute substantially to the local economy. Below-dam sections of the river receive sizable migrations of landlocked anadromous salmonids - steelhead trout, and coho and chinook salmon. All are introduced species. There is currently much interest and research devoted to the potential for reestablishing the native Atlantic Salmon (the fish for which the river is named) to the river.

Potential Influence of Reservoirs

Flow - The Salmon River, like other streams draining the Tug Hill, is generally a "flashy" system - the water rises and falls quickly in response to rain events. Regulation of the Salmon River involves "chopping off" the top of the hydrograph - letting out less water than comes in from upper sections during very high flows, and often releasing more than comes in during lower flows. Example hydrograph for Salmon River vs. Sandy Creek

While this generally makes the lower river subject to less-severe high flows than it would be if it were unregulated, it is subject to elevated flows for a greater period of time each year than above-reservoir areas. On four weekends each summer, below-reservoir sections are subject to "white-water releases" for kayakers and canoeists, during which the river flow can increase several fold. These releases can be seen in the 1998 hydrograph.

There is thus potential for greater flushing of leaves, detritus and algae from the bottom at below-dam sections, because elevated flows create higher velocity water.

Temperature - The higher water temperatures present below the reservoirs much of the year may cause observable differences in macroinvertebrate community when compared to upper sites.

Goals and Hypotheses of Study
The original purpose of this study was to survey the benthic macroinvertebrate communities of riffle areas at sites above and below the reservoir system. The dams' influence on the timing and level of river flow, and on water temperature and potentially chemistry may have altered the stream conditions enough to affect the macroinvertebrate community.This plan was later modified to include an investigation of the effect of channel gradient which affects water velocity on invertebrates.

Hypotheses:
Sites above the reservoirs will have a greater density and mass of macroinvertebrate types feeding on leaves, detritus, and algae, due to a lower frequency of elevated flows.
On any one side of the reservoirs, higher gradient riffle sites will be associated with fewer individuals and less mass of non-filter-feeding macroinvertebrates than lower gradient sites, due to greater flushing of food from these higher-velocity habitats.

Methods:
I sampled six relatively open riffle areas bordered by deciduous forest found near public fishing access locations. Three upper and two lower sites were added during sampling round five to include a more even range of channel gradient on each side of the reservoirs. The sites are listed, upstream to downstream, by road crossing, with site numbers in parentheses below.
Above Reservoir Sites - N. Osceola Rd. (6), Ryan Rd. (11,9,5), Harvester Mill Rd. (10), Waterbury Rd. (4)
Below Reservoir Sites -Rt. 22(3), Rt. 52 (2), Sheepskin Rd. (8,1), Lehigh Rd. (7).

Map of the entire Salmon River watershed

Data/Samples collected:

Benthic macroinvertebrates - Collected by surber sampler every 6-12 weeks (10/99, 12/99, 1/2000, 3/2000, 6/2000) with 3-4 replicate sub-samples per site.

Physical parameters - Substrate size-class (proportion of gravel, pebble, cobble, boulder), depth, and velocity were recorded for each sub-sample site where benthos was collected. Stream width was measured at summer base flow, with the exception of site 7 (335 cubic feet per second instead of 185 cfs). Channel gradient was measured with a handheld clinometer.
Chemical parameters - Nitrogen, phosphorus, dissolved organic carbon, alkalinity, pH, dissolved oxygen, conductivity, chloride, sulfate were normally collected within several weeks of benthic samples.

Results:
Data from processed benthic samples and associated analyses will be reported here as it becomes available over the next few months.

The mean maximum summer water temperature during 2000 was higher at above-dam (upper) than below-dam (lower) sites ( 23.75 oC and 18.75 oC, respectively). Stream reach gradient spanned a similar range on both sides of the reservoirs. Nitrate was greater at upper sites on all occasions except in March, with means of 0.302 and 0.274 mg/l, respectively. Total alkalinity was greater at upper sites than lower sites on all occasions, with means of 48.85 mg/l and 37.8 mg/l, respectively. Dissolved organic carbon was higher at lower than upper sites on all dates, with means of 4.04 mg/l and 3.06 mg/l, respectively. Total phosphorus was also greater at lower sites than upper sites on all dates, with means of 5.22 mg/l and 2.06 mg/l, respectively.

The chemistry data represent the mean of 5 sampling periods, and are in mg/l.

About The Author: John L. Hallock Jr.

Originally from Rochester, NY region. Academic and scientific interests include aquatic ecology and influence of watershed characteristics, ecology and natural history in general. Other interests include history, writing, skiing and backpacking.

Education - Bachelors Degree in natural resources from Cornell University - 1991.