A Comparison of Macroinvertebrate Diversity in Six Streams and Rivers

in Olympia, Lacey and Tumwater, Washington

Wendy Matson

Senior Seminar II

May 8, 2007

Final Draft


Table of Contents

Abstract 1

Introduction 2

Methods 5

Data Analysis 9

Results 10

Discussion 14

Acknowledgements 18

Literature Cited 20

Abstract

This study was conducted to compare the macroinvertebrate diversity among six streams and rivers in Olympia, Lacey and Tumwater, Washington. By examining the benthic macroinvertebrate population in streams and rivers, we can deduce the health and cleanliness of those water bodies. I sampled six streams; two streams from each city. In Olympia, I sampled Percival Creek and Mission Creek. In Lacey, I sampled Woodland Creek and Nisqually River and in Tumwater, I sampled Deschutes River and Dempsey Creek. When comparing the macroinvertebrate diversity among the six streams, I found that they were all statistically alike (f = 2.73; d.f. = 5; p = 0.071), while similar taxa among the streams like: Class Insecta, Phyla Annelida and Mollusca, and uncommon macroinvertebrates were statistically different among the six streams with p values greater than 0.05 for each group. General larvae (i.e. midgefly larvae, blackfly larvae, and fishfly larvae) was the only group that was statistically similar among the six streams (p = 0.400).


Introduction

Olympia, Lacey, and Tumwater, Washington are three cities located in the south Puget Sound area. A priority for these cities is to keep their water clean by testing the quality. Though water quality testing can be done several different ways, one way is to examine the benthic macroinvertebrate population. The common definition of benthic macroinvertebrate is: benthic, living on the bottom, usually in gravel or rocks; macro, large enough to see with the naked eye; and invertebrate, animals without backbones or vertebrate (SalmonWeb, 2000).

According to Carlisle et al. (2007) macroinvertebrate populations in streams and rivers can assist in the assessment of the overall health of the stream. To help understand which macroinvertebrates are more or less tolerant to pollutants, Anderson et al. (2005) sampled the Salina River in southern California, which had pesticides and fertilizers being directly pumped into it. They tested the effect of bifenthrin and permethrin, two pesticides, on macroinvertebrates in the river. For bifenthrin, the highest mean survival rate for any macroinvertebrate was 77% and the lowest was 0%; for permethrin the highest mean survival rate for any macroinvertebrate was 68% and the lowest was 48% (Anderson et al., 2005) thus concluding macroinvertebrates were sensitive to water disruptions, such as pollutants. Stonefly larvae, Plecoptera sp., mayfly larvae, Ephemeroptera sp., caddisfly larvae, Trichoptera sp., show the highest sensitivity to pollution, while midgefly larvae, Chironomidae sp. and aquatic worms, Oligochaete sp., show the highest tolerance to pollutants (Anderson et al., 2005).

According to Goodnight (1973) the sensitivity of stoneflies, caddisflies and mayflies is attributed to the fact that they are gill-breathing aquatic insects and are affected by conditions that alter or reduce the dissolved oxygen levels of the water, such as pollutants. Thus the association of caddisflies, mayflies, and stoneflies in a stream indicates high water quality. There are many factors that contribute to the existence of macroinvertebrates in streams: substrate, or the bottom of the river where macroinvertebrates attach and live, bridges or artificial walls in the stream, the width of the stream, and water temperature.

Hax and Golladay (1998) performed a study of the effect of flow disturbance on macroinvertebrates inhabiting sand and gravel substrates, as well as cobble substrates with woody debris. They wanted to find how disturbances, such as rainfall, would affect macroinvertebrates living on the two substrates. To clone the effects of heavy rain, Hax and Golladay (1998) diverted water through a certain part of stream, to increase water flow. The experiment was completed at two different sites, both within close proximity, but with two different substrates (sand/gravel and cobble/wood). They sampled before and after the water diversion and three invertebrate samples were collected. Their results found that on wood substrates there was no significant loss in total macroinvertebrate density, concluding that macroinvertebrate resistance was 10% greater on cobble/wood that sand/gravel (Hax and Golladay, 1998). Based on their results, macroinvertebrate richness and diversity can be directly correlated to the substrate of the stream.

Important issues facing rivers, streams, and lakes include erosion, the placement of man-made retaining walls, and bridges built in the middle of rivers and lakes. Schmude et al. (1998) stated that the development of macroinvertebrate colonization was dependent on many factors, including: heterogeneity (diverse nature), interstitial space (small openings or gaps) and substrate complexity (bumpy or smooth surfaces). They compared flat, brick walls to walls made of boulders and uneven surface by using concrete patio blocks to simulate the retaining walls, and cement balls in a wire basket to simulate rock riprap or boulders naturally found in the water. Schmude et al.’s (1998) results showed boulders and rocks found naturally in the stream or lake supported greater macroinvertebrate abundance than artificial or man-made walls. To ensure artificial walls did not negatively affect my results, I sampled at least 50 meters upstream or 200 meters downstream of any artificial structure (SalmonWeb, 2000).

Paller et al. (2006) tested whether there was a difference in macroinvertebrate richness, the number of different species found in a stream, among small, medium, and large rivers. Streams were measured by width. Small rivers were classified between 0 and 6 meters, medium rivers were 6 to 12 m, and large rivers were between 12 and 18 m wide. The study was conducted at 27 sites and streams ranged from 2-16 m wide. Macroinvertebrate sampling was done at seven to twelve evenly spaced transects across the stream perpendicular to the direction of water flow. Paller et al. (2006) found that there was a positive relationship between stream size and macroinvertebrate richness in small and medium streams. They also found that the evenness, or the relative proportion of those species found, showed a positive relationship with stream size in small and medium streams. However, there was a negative correlation between stream size and richness for medium and large streams.

Gallepp (1977) tested the response of caddisfly larvae to water temperature. He simulated two streams in the laboratory; used wooden dowels as attachment sites for the caddisflies, and controlled the water temperature by circulating antifreeze. Temperatures ranged from 4º Celsius to 30º Celsius. Gallepp (1977) found that caddisflies’ growth and population decrease with water temperatures below 8.5º Celsius and above 16º Celsius. I sampled from late January through mid-March and water temperatures at all six streams were measured below 8º Celsius. Due to seasonal low temperatures of this study, there may be low numbers of caddisflies.

My study tested the water quality of six local streams by collecting benthic macroinvertebrates. I compared the macroinvertebrate diversity among the six streams, while also comparing similar taxa like Class Insecta, Phyla Annelida and Mollusca, general larvae, and uncommon invertebrates. Although I compared similar taxa my goal was to determine if there was a significant (p < 0.05) difference in the macroinvertebrate diversity among the six streams. I sampled six streams in Olympia, Lacey and Tumwater, Washington; two streams from each city. In Olympia, I sampled Percival Creek and Mission Creek. In Lacey, I sampled Woodland Creek and Nisqually River and in Tumwater, I sampled Deschutes River and Dempsey Creek. I chose these rivers, because of their proximity to each other, being within the same county and the farthest distance was between two streams was Dempsey Creek and Nisqually River which are 32 kilometers apart, while Deschutes River and Percival Creek were the closest to each other, at 6.3 kilometers. Also all the streams and rivers had public access and were easily accessible and safe. I hypothesized there would not be differences in diversity among the six streams because of their proximity. I also hypothesized there will be a significant different (p < 0.05) between streams with sand and gravel substrate versus streams with cobble and wood substrate, since rivers with cobble and wood substrate should maintain a higher number of Class Insecta based on the results of Hax and Golladay (1998).

Methods

Locations of Streams and Data Collected

I sampled the macroinvertebrate population in six streams in Olympia, Lacey and Tumwater, Washington. I sampled two streams from each city over a six week period, sampling one stream per week. I tested to see whether there were differences among the three city’s macroinvertebrate populations. Beche et al. (2006) found that the abundance and composition of macroinvertebrates were sensitive to rainfall, so in order to limit variance in the data due to weather, I did not sample city’s streams consecutively. Instead, I randomly selected the order in which the streams were sampled (Table 1).

Week 1 / Week 2 / Week 3 / Week 4 / Week 5 / Week 6
City:
Approximate Location / Lacey
SMU Campus / Olympia
Priest Point Park / Tumwater
Pioneer Park / Lacey
6th Street Nisqually / Olympia
SPSCC
Campus / Tumwater
88th Ave.
Little Rock
Stream
Sampled / Woodland
Creek / Mission
Creek / Deschutes River / Nisqually
River / Percival
Creek / Dempsey Creek

In Olympia, I sampled Mission Creek and Percival Creek, in Lacey, Nisqually River and Woodland Creek and in Tumwater, Deschutes River and Dempsey Creek. All of the sites were located in vegetative wooded areas, however, Mission Creek was surrounded by the most vegetation and isolated from the road and pedestrians, while Percival Creek, was also surrounded by vegetation, but was located in the middle of South Puget Sound Community College campus on a high traffic trail. Water temperatures ranged from 4°C to 8°C and air temperatures ranged from 5°C to 10°C. The weather ranged from clear with sunshine to light rain. When it was raining heavily, sampling was postponed to ensure that the process of sampling was completed as consistently and accurately as the previous samples taken.

Methods Completed at Sampling Site

When choosing the spot in the stream to sample, I looked for a riffle. A riffle is a fast moving portion of the river, where the water breaks over rocks (SalmonWeb, 2000). The riffle needed to be at least 20 meters long. The randomization in my experiment was the 20 meter area of the riffle chosen. Randomization was important because it created an un-biased sample and measured a fair variability between replicates (Heath, 2005). I could not randomly choose an area of a stream, because of the specific needs of a 20 meter riffle and because the water had to be shallow enough in which to walk. I took three replicate samples from each stream, and each replicate was separated by at least 5 meters (SalmonWeb, 2000). I avoided bridges and other man-made structures, because uneven natural ground supports higher numbers of macroinvertebrates than block or man-made structures (Schmude et al., 1998).

Once I reached each sampling site, the stream name, date, time, air temperature and water temperature were recorded (SalmonWeb, 2000). The site was set up to be as efficient as possible once the sample was removed from the water. A plastic sheet was laid out on the ground, a 500 micron brass sieve used to sieve the sample, screened rinsing jug, rubber spatula, two sets of forceps, plastic scoop, a bucket, 70% denatured ethanol in a plastic bottle obtained from Saint Martin's University Chemistry Department and an empty 500 ml plastic bottle (final bottle) for macroinvertebrate transfer to laboratory from site, were placed on the plastic sheet.

My first replicate was at the farthest point downstream, and the third was the farthest upstream. The first site was flagged to show where I sampled. Approaching the sampling site from downstream, I placed a fine mesh net, (500 µm) or Surber sampler, downstream of the first replicate. My sampling partner held the sampler in place and, using a watch with a second hand, recorded the time. Using both hands, I vigorously stirred the substrate (bottom of the river) for 60 seconds. At the end of the time, any large rocks that remained in the sampling area were brushed off, removed from the water and placed in the bucket (SalmonWeb, 2000). Before removing the Surber sampler from the creek, water was used to rinse the outside of the net. With the plastic scoop, water was poured from the outside to help release any macroinvertebrates attached to the net. Once the net was clean, the sampler was removed from the water, placed over the plastic sheet, and the sampler’s end cup, or cod end was removed. I poured the contents of the cod end into the 500 micron brass sieve and rinsed it out with water from the screened jug, always pouring the cod end out through the sieve. The Surber sampler was then turned inside out and examined for any remaining invertebrates clinging to the sides. Any macroinvertebrates found were removed and placed directly into the final bottle. Once the cod end and Surber sampler had been successfully cleaned, they were placed to the side for use in the next replicate. The rocks in the bucket were then examined for macroinvertebrates. Once again, if invertebrates were found, they were placed directly into the final bottle. Large rocks in the sieve were also examined at this time. When all the large rocks had been removed, the contents of the sieve were scraped into the final bottle until both sampling partners were convinced that no macroinvertebrates remained in the sieve. Once this step was completed, the final bottle was filled with 70% ethanol, making sure that the ethanol covered all of the contents in the bottle. The final bottle was labeled with replicate number, stream name and date and was set aside (SalmonWeb, 2000).