Urbanization Effects on Streams

Urbanization poses a major threat to stream and watershed ecosystems. When we clear lands, remove wetlands, and pave roads, we increase overland runoff and increase the severity of river siltation and scour. By logging and removing vegetation close to streams, we alter stream temperature and remove valuable streamside nutrients, leading to changes in stream organism richness, health, and composition. The urban dust we create, laden with poisonous heavy metals, drops continuously from the sky. These are just a few of the factors that contribute to the degradation of urban environments. In a study from the mid 1990's, William J. Kleindl attempts to quantify the impact of urbanization on the biological health of streams in the Puget Sound lowlands of Washington State.

Kleindl uses percent impervious surface as a proxy for the amount of urbanization. He examined aerial and satellite images of the sample watersheds to estimate the areas or regions dedicated to different types of land use. Each type of land use had a pre-assigned impervious percentage. On the upper end of the scale, commercially developed areas were designated 90% impervious; on the lower end, agricultural and forest land were deemed 2% imperious. Land use area percentages, weighted by their impervious percentages, were then summed to get an overall watershed impervious percentage. Simple math, very effective.

In the study, macroinvertebrates were sampled at 31 sites in 18 different streams during September 1994 to determine urbanization effects on the streams' biotic integrity. Kleindl describes biotic or biological integrity as, "the capability of supporting and maintaining a balanced, integrated, adaptive community of organisms having a species composition and functional organization comparable to that of natural habitat in the region." The 18 streams represented a wide range of land use conditions in rural, suburban and urban watersheds. Data from 17 of the sites were used to create the Benthic Index of Biotic Integrity (B-IBI) and are presented here; data from the remaining 14 sites were used to test the B-IBI, and can be found in the original paper.

The sampled macroinvertebrates were tested for 9 attributes (metrics) that were proven to be sensitive to urbanization effects. Of these 9 metrics, 7 measured invertebrate composition and richness (Average Taxa Richness, Ephemeroptera Richness, Plecoptera Richness, Trichoptera Richness, Intolerant Taxa Richness, Long-lived Taxa Richness). Community structure was measured by 2 percentage metrics (Percent Planaria and Amphipoda Abundance, and Percent Tolerant Taxa). The final percentage metric (Percent Predator Richness) measured impact on feeding (trophic) levels. For each of these 9 metrics, data from the 17 sites were ranked in order from least impacted to most impacted. Streams that were least impacted were given a score of 5, intermediate streams were given a score of 3, and the most impacted streams were given a score of 1. (Assignments to categories were somewhat arbitrary.) The sum of these 9 metric scores provided a total B-IBI score for each stream.

Students might use the data provided here to address the following questions:

• Is percent impervious surface an overall good predictor of the biological health of a stream?
• For which percentages does there seem to be more variability in the data? How does this variability affect predictions for different percentages of impervious surface?
• What factors might contribute to the variability mentioned above, or is this simply random variability?
• Is there an equation whose graph fits the data well, and if so, how can you measure the goodness of fit?
• At any point along a best fitting equation, what would the slope measure?
• Which of the 9 individual metrics is most sensitive or intolerant to urbanization?
• Why do some of the metrics increase, while others decrease, as percent impervious surface increases?
• What are some other factors, besides impervious surface, that would lead to the decline in biological integrity of a stream? Considering your answer, is it fair to predict biological integrity using only impervious surface?
• Is it better to spend money improving areas with low biologic integrity or protecting areas with high biologic integrity?

Data source: Kleindl, William J. 1995. A Benthic Index of Biotic Integrity for Puget Sound Lowland Streams, Washington, USA. Thesis submitted for a Master of Science degree at the University of Washington in Seattle.

Explain the data set you chose and why you chose it. [text] Perform a linear regression on the data using Desmos or a spreadsheet program, such as Excel or Google Sheets (help performing linear regression can be found in Week 3). [graph] Share screenshots of your graph, regression equation and correlation coefficient.[put equation and r^2 on graph] Interpret both the slope and the y-intercept in this situation. What do these values mean in terms of your variables? [text] How confident are you in your model, that is what does r and r^2 values tell you? [text] Make a prediction based on your model. This means you should choose an x-value and find the corresponding y-value using your model. Then you should compare this to the actual data. [calculations using model and text]