Macroinvertebrate Indices in Response to Urbanization

Travis County Stream Assessment

Introduction

Healthy aquatic environments are of fundamental importance for conservation, restoration and the sustainability of natural resources (Fusari and Fonseca-Gesser 2006; Allan 1998; Kerans 1994). Monitoring impacts of environmental changes on stream health is critical for analysis of ecological function, ecosystem services and the degree of impact a system has on the environment (Fusari 2006; Allan 1998; Kerans 1994). It is important to monitor and understand the distribution patterns in aquatic communities and to quantify biotic diversity to determine the response to local factors and urban impacts (Dar et al, 2014).

Macroinvertebrate community assemblages are commonly used to measure water quality because they consist of a variety of species and life cycles that are receptive to changes in the environment (Wang et al. 2003; Cao et al. 1996; Kerans 1994). Many species are sensitive to habitat degradation and have been found to respond to a range of urbanization effects on stream quality (Gilvear et al, 1999; Wang et al, 2003). The combined influences of hydrology, channel morphology and water quality have been shown to change macroinvertebrate assemblages (Wright et al. 1984; Cao et al. 1996). Sampling macroinvertebrates is an accurate and easy way to measure indicators of overall water quality and stream ecosystem function (Gilvear et al. 1999; Kerans 1994).

 Biological monitoring is done to evaluate the human impact on natural systems and resources. When streams and bodies of water are in an urban setting they are often subject to nonpoint pollution and increased sedimentation (Fore et al. 1996; Paul and Meyer 2001; Kerans 1994). Nonpoint sources of pollution and disturbance by urbanization are responsible for most of the contamination of streams (Barbour et al. 1996; Allan 1998; Paul and Meyer 2001). The distribution and species assemblages of macroinvertebrates are directly influenced by the sediment structure, amount of debris and organic pollutants (Dar et al. 2014; Paul and Meyer 2001; Hilsenhoff 1988). The level of organic matter that is deposited in the sediment is an important source of food for benthic macroinvertebrates, but high levels can lead to anoxic conditions, reducing survival and leading to mortality of organisms (Dar et al. 2014; Krueger and Waters 1983). Because of this macroinvertebrate assemblages reflect the ecological condition of aquatic habitats (Dar et al. 2014; Krueger and Waters 1983).

Traditional chemical methods for testing water and determining the effects of NPS pollution have been inadequate due to the difficult nature of tracing non-point source pollution, due to the variability in pollution over time sampling water sampling would have to be ongoing to understand what is happening in the stream. Macroinvertebrate sampling reflects the long-term conditions of streams (Barbour et al. 199; Barton and Metcalfe-Smith 1992).  However, biological monitoring however can show the effects of pollution and disturbance based on cumulative or single event disturbances (Barbour et al. 1996; Barton and Metcalfe-Smith 1992). The biological composition of streams is largely affected by physical and chemical factors affected by urbanization. Urbanization and impervious coverage alter watershed composition and lead to stream quality degradation (Wang et al. 2003; Benke et al. 1999).  There is extensive evidence that urban land use around stream ecosystems changes the physical, chemical and hydrological components of the system and alters macroinvertebrate assemblages (Allan 1998; Capitula et al. 2001; Chandler 1970).

Macroinvertebrates range from grazers, shredders, gatherers, filterers, and predators making them vital to the trophic web, and many species rely on the physical stream characteristics to get food (Wallace and Webster 1996). Macroinvertebrates that are at intermediate trophic levels are greatly affected by both bottom-up and top-down forces (Wallace and Webster 1996). Nutrient cycling, decomposition and primary productivity in streams are dependent on the biomass and diversity of macroinvertebrates (Wallace and Webster 1996). Changes in the habitat structure from human development often disturb the physical characteristics of streams that macroinvertebrates rely on.

To understand the patterns of macroinvertebrate assemblage and identify the possible factors influencing water quality its important to measure surrounding habitat vegetation and canopy coverage and impervious coverage from urban development. Spatial scale and location of land disturbance have the strongest influence on aquatic ecosystems (Schiff and Benoit 2007; Paul and Meyer 2001; Fore et al. 1996). A critical level of 5%-7% impervious coverage is the identified threshold for stream health (Schiff and Benoit 2007). Comparing macroinvertebrate assemblage with urbanization and changes in habitat structure will outline the threshold in Travis county and identify areas that are degraded.

Water quality and macroinvertebrate assemblages decline with the increase of imperviousness and level off in a constant state of impairment at 10% (Schiff and Benoit 2007; Paul & Meyer 2001). This illustrates the need for multiscale watershed management and stream bed monitoring in urban settings (Schiff and Benoit 2007; Fore et al. 1996).

In streams with good water quality, you would generally expect to find a greater diversity of macroinvertebrate groups, which would include groups that are both sensitive and tolerant to pollution (Camargo 1993). As organic pollution increases, dissolved oxygen levels within the stream will fluctuate more extremely, and fewer organisms that are pollution-sensitive will be found (Armitage et al. 1983). As organic pollution continues to increase, you would expect pollution-tolerant macroinvertebrates to become more dominant, while other sensitive and semi-sensitive organisms will not be able to survive (Armitage et al. 1983).

Population growth often has lasting impacts on air and water quality because of increased development and impervious coverage and a decrease in bank stability, vegetation and canopy cover (Cao et al. 1996). The increase or fluctuation of lead, bacteria, pesticides, iron, copper, alkalinity, pH, hardness and chlorine can have lasting damage on water quality and watersheds. These pollutants and changes in chemistry are often introduced into the water with the increase of traffic and human presence. It is important to be able to measure and understand stream water quality and the functionality of the ecosystem in response to human presence (Camargo 1993).

Urbanization and human impact are rapidly increasing in Travis County which lies in a critical recharge zone for the Edwards Aquifer, home to sensitive and endangered species. This makes the city of Austin an ideal location to conduct this research on how urbanization affects water quality and macroinvertebrate assemblages.  Quantifying the diversity and abundance of macroinvertebrates will allow us to better understand the effects of increased organic and heavy metal pollutants on water quality in and around the city. It also important for the health and richness of the lake that all these streams flow into.

This study determined the stream bed stability of 4 different streams in Travis County measuring the biotic indices used to classify the degree of pollution in the aquatic systems and using water quality tests for common pollutants. I hypothesized that benthic diversity and abundance of sensitive species (biotic indices) would be greater where water quality is higher. I also hypothesized that water quality would be lower in more urban areas such as streams that enter the lake through downtown Austin.

Methods

Study Site:

 This study was conducted in Travis county and within Austin TX, which sits above the Edwards Aquifer.  Travis county is 1023 mi² and Austin is 272 mi². Population growth rate over 15 years is expected to be nearly 50 percent (Tretter et al. 2013). This makes Austin susceptible to increased habitat disturbance and increased impervious coverage from continued development and urban sprawl.

Streams studied within Travis County included Barton Creek at Lost Creek and Sculpture Falls, Bee Creek at Wild Basin and Westlake Drive crossing, Shoal Creek at 29^th and at West Ave & 5^th St And Waller Creek at East Woods Neighborhood Park and at 15^th St. Each sample site was a perennial wadable stream that contained similar physical conditions. These physical characteristics were riffles, stream flow velocity at least 0.3 meters per second and stream substrate composed of varying gravel size. Each site was sampled and tested twice, first between March 15- April 23rd and then between Sept 15th- Oct 23rd.

Sampling Methods:

I sampled 4 streams in the Travis County area using macroinvertebrates as bio-assessment indicators for water quality by analyzing the biological indices: species richness, density and family biotic index (FBI) (Xu et al. 2014, Capitula et al. 2001, Chandler 1970). Analysis was done by examining the occurrence of macroinvertebrates taxa in different water quality levels, pollutant sensitive species were used for the bio-assessment (Xu et al. 2014; Capitula et al. 2001; Chandler 1970).

I collected a water sample at each site on each sampling date to test for lead, heavy metals (Pb, Fe, Cu) bacteria, pesticides, alkalinity, pH, hardness, chlorine, nitrates and nitrites using a water quality test kit (Test Assured, COMINHKPR136163) and dissolved oxygen (DO) levels with an ORP monitor.

Macroinvertebrate were collected following the procedures described by Hilsenhoff (1987).  I collected benthic macroinvertebrate samples using a rectangular net frame, 600-micron mesh kick by holding the net frame firmly against the stream bottom and disturbing the substrate upstream (approximately a full arm’s length) from the net with my feet, digging deeply into the substrate to dislodge macroinvertebrates from the streambed for approximately 3 minutes. The macroinvertebrates washed downstream into the net and I made sure that the plume of silt that resulted from disturbing the substrate flowed into the net, as this plume also contained the dislodged invertebrates.

Sampling effort, number of macroinvertebrates / sample volume:

More than 100 macroinvertebrates were collected per sample. I inspected to ensure that more than 100 macroinvertebrates had been collected. If it was determined that insufficient numbers of macroinvertebrates were captured after initial sampling efforts, I repeated my method for a second period of equal duration.

Specimen Handling and Preservation:

After I collected the macroinvertebrate sample, I rinsed the sediment from the net being careful not to lose the organisms captured. The macroinvertebrates were transferred to a tightly sealing glass wide-mouth jar labeled properly. I preserved my samples with 80 proof ethanol, filling the sample jar to the top with the alcohol solution and gently inverting the container several times to thoroughly mix the sample and preservative. Within 24 hours I poured off the alcohol solution and refilled the jar with fresh 80-85% alcohol. I identified all specimens to family and order; A Guide to Common Freshwater Invertebrates of North America by J. Reese Voshell Jr (Voshell & Wright 2002).

Measuring Landscape structure

 Bank vegetation was measured by 30m Linear transects on both banks at each site.  vegetation cover was recorded when it began and ended along the line. Canopy cover and impervious coverage were determined by creating a .5-mile buffer using ArcGIS to over lap layers that were available in the open source database.

Analysis of data

I calculated species richness and Shannon diversity index for each sample site.  I evaluated the distribution and density of pollutant tolerant or intolerant species by running an EPT analysis. The EPT Index is named for three orders of aquatic insects that are common in the benthic macroinvertebrate community: Ephemeroptera (mayflies), Plecoptera (stoneflies), and Trichoptera (caddisflies) (EPT). EPT scores are 0-1 Very poor; 2 poor; 3-4 fair, 5+ good.  I ran a regression to test whether there was a significant p value between EPT and impervious coverage percent for each site.  I ran a principal component analysis (PCA) to test the relationships between the water quality indicators of lead, heavy metals (Pb, Fe, Cu) bacteria, pesticides, alkalinity, pH, hardness, chlorine, nitrates and nitrites and DO. I ran a multiple regression separately comparing EPT, Richness and Shannon Index with canopy cover, bank vegetation and impervious coverage.

Results

EPT was significantly lower where the percent impervious coverage was highest (Regression; F_(1,25)=31, 52, p < 0.0001). There was a significant increase of EPT with the increase of bank vegetation (Multiple Regression; R2 =.89, F_(2,24)=114, p<.0001) (Fig. 2).  There was a significant decrease of EPT with the increase of impervious coverage percent (Multiple Regression; R2 =.89, F_(2,24)=114, p=.006) (Fig. 3.)  There was a significant increase of Richness with the increase of bank vegetation (Multiple Regression; R2 =.94, F_(2,24)=214, p<.0001) (Fig. 2.) There was significant decrease in richness with the increase of impervious coverage percent (Multiple Regression; R2 =.94, F_(2,24)=214, p=0.03) ( Fig. 3). There was a significant increase of Shannon diversity index with the increase of bank vegetation (Multiple Regression; R2 =.94, F_(2,24)=248, p<.0001) (Fig. 2). There was a significant decrease in Shannon Diversity Index with the increase of impervious coverage percent (Multiple Regression; R2 =.94, F_(2,24)=248, p=.0005) (Fig. 3). there were no significant differences in water quality indicators of lead, heavy metals (Pb, Fe, Cu) bacteria, pesticides, alkalinity, pH, hardness, chlorine, nitrates and nitrites or DO across all sites. Canopy cover was not a significant predictor variable for richness, EPT or Shannon Diversity Index.

Tables and Figures

Figure 1. Bar plot of EPT score compared with percent impervious coverage (Regression; F_(1,25)=31, 52, p < 0.0001). 

Figure 2. Scatter plot of EPT Index, richness and Shannon Index compared with Bank Vegetation Percent (Multiple Regression EPT; F_(2,24)=114, p<.0001), (Multiple Regression richness; F_(2,24)=248, p<.0001), (Multiple Regression Shannon Index; F_(2,24)=248, p<.0001)

Figure 2. Scatter plot of EPT Index, richness and Shannon Index compared with impervious coverage percent (Multiple Regression EPT; F_(2,24)=214, p<.0001), (Multiple Regression richness; F_(2,24)=214, p=0.03), (Multiple Regression Shannon Index; F_(2,24)=248, p=.0005)

Discussion

The results supported my hypothesis that water quality would be greater in less urban and developed areas and low water quality would be found in more developed areas. The EPT index was a good indicator of urban pollution having a higher score in less urban areas and a lower score in more developed areas. Impervious coverage and bank vegetation proved to be significant predictor variables of water quality while canopy cover although not significant in this study did show a positive correlation with water quality. The EPT, Shannon Diversity Index and richness all were significantly negatively correlated with the increase of impervious coverage.

Looking at the EPT specifically compared with impervious coverage confirm the absence of organisms susceptible to habitat degradation above 10% impervious coverage (Schiff and Benoit 2007), with impairment starting at 7%.  Any site that was above 10% impervious coverage had an EPT score that suggests low water quality. Alternatively, EPT, Shannon Diversity Index and richness all increased with bank vegetation which would also be more prevalent in areas with less impervious coverage. Health communities where located upstream in areas with more vegetation and canopy cover, while impaired communities were closer to the inner city with increased impervious coverage.

 Although other factors of physical stream habitat quality were not measured in this study, it is likely that the decrease in canopy cover reduces the density of woody debris which is a primary food source for many macroinvertebrates. Bank Vegetation being a significant predictor is likely due to bank stability. Sedimentation and erosion are significant factors affecting macroinvertebrates and streams with poor bank stability are highly susceptible to both. Bank vegetation and canopy cover were obvious decreased with the increase of impervious coverage. Much of the scientific literature looking at stream health and impervious coverage indicates a presence a threshold at 10%, this study is in agreement with the findings from previous research that indicate biotic health and physical quality decline with increased impervious coverage.

 The mechanisms of degradation above 10% impervious coverage include poor water drainage that disrupts the natural flow of water to low areas and increasing the semination and erosion. This influence on the geomorphology variables combined with mixed NPS pollutants lead to long-term stream degradation. At a site-specific scale, the removal of bank and woody vegetation fragments riparian habitats limiting flow reduction and NPS pollution; directly altering habitat by reducing debris, organic matter and increasing sedimentation.

 Previous studies and my data suggest that impervious coverage should be minimized to less that 10% to properly protect aquatic biota in small streams. The protection and restoration of aquatic ecosystems should incorporate a multiscale approach and consider stream bank integrity and surrounding impervious coverage. City planners and local laws should consider the implications of not treating water on site to reduce the influx of runoff and NPS attenuation.

Literature Cited:

• Allan J. 1998. Ekologia wód plynacych, Stream Ecology. Structure and Function of Running Waters. Wyd. Nauk. PWN, Warszawa.
• Armitage P, Moss D, Wright J, Furse M. 1983. The performance of the new biological water quality score system based on macroinvertebrates over a wide range of unpolluted running water sites. Water Restoration 17: 333-347.
• Barbour M, Gerritsen J, Griffith G, Frydenborg R, McCarron E, White J & Bastian M. 1996. A framework for biological criteria for Florida streams using benthic macroinvertebrates. Journal of the North American Benthological Society 15:2, 185-211.
• Barton D, Metcalfe-Smith J. 1992. A comparison of sampling techniques and summary indices for assessment of water quality in the Yamaska river Quebec based on benthic macroinvertebrates. Environ Monit Assess 21: 225-244.
• Benke A, Huryn A, Smock L, Wallace J. 1999. Length-mass relationships for freshwater macroinvertebrates in North America with particular reference to the Southeastern United States. Journal of the North American Benthological Society 18:3, 308-343 
• Camargo J. 1993. Macrobenthic surveys as a valuable tool for assessing freshwater quality in the Iberian Peninsula. Environ Monit Assess 24: 71-90.
• Cao Y, Bark A, Williams W. 1996. Measuring the responses of macroinvertebrate communities to water pollution: a comparison of multivariate approaches, biotic and diversity indices. Hydrobiologia 341: 1-19.
• Capitulo A, Tangorra M, Ocon C. 2001. Use of benthic macroinvertebrates to assess the biological status of Pampean streams in Argentina. Aquatic Ecology 35: 109-119.
• Chandler R. 1970. A biological approach to water quality management. Water Pollution Control 69: 415.
• Dar N, Pandit A, Ganai B. 2014. Factors affecting the distribution patterns of aquatic macrophytes. Limnological Review 14: 75-81
• Fore L, Karr J, Wisseman R. 1996. Assessing invertebrate responses to human activities: evaluating alternative approaches. Journal of the North American Benthological Society 15: 212-231.
• Fusari L, Fonseca-Gessner A. 2006. Environmental assessment of two small reservoirs in southeastern Brazil, using macroinvertebrate community metrics. Acta Limnologica Brasiliensia, 18: 89-99.
• Gilvear D, Karr J, Chu E. 1999. Restoring life in running waters: better biological monitoring. Water, Air, and Soil Pollution, 116: 666-667.
• Gołdyn, Szpakowska, Świerk, Domek, Buxakowski, Dondajewska, Sajnóg. 2018. Influence of stormwater runoff on macroinvertebrates in a small urban river and a reservoir. Science of the Total Environment, 625: 743-751.
• Hilsenhoff W. 1988. Rapid field assessment of organic pollution with a family-level biotic index. Journal of North American Benthological Society 7: 65–68.
• Hilsenhoff W. 1987. An improved biotic index of Organic Stream Pollution. The Great Lakes Entomologist: Vol 20: No. 1, Article 7.
• Kerans B. 1994. Freshwater biomonitoring and benthic macroinvertebrates. Ecology 75: 267.
• Paul M, Meyer J. 2001. Streams in the urban landscape. Annual Review of Ecology and Systematics 32: 333-365. 
• Schiff R, Benoit G. 2007. Effects of impervious cover at multiple spatial scales on coastal watershed streams. Journal of the American Water Resources Association 43: 712-730.
• Tretter E, Gibbs D, Krueger R., & Macleod G. 2013. Sustainability and Neoliberal Urban Development: The Environment, Crime and the Remaking of Austin’s Downtown. Urban Studies, 50:(11), 2222-2237.
• Voshell J, Wright A. 2002. A guide to common freshwater invertebrates of North America. Blacksburg, Va.: McDonald & Woodward Pub.
• Wang, Lyons, Rasmussen, Seelbach. 2003. Watershed, reach, and riparian influences on stream fish assemblages in the Northern Lakes and Forest Ecoregion, U.S.A. Canadian Journal of Fisheries and Aquatic Sciences 60: 491-505.
• Wright F, Furse M, Armitage P. 1993. RIVPACS- A technique for evaluating the Biological quality of rivers in the U.K. Freshwater biology 3: 15-25
• Wallace B, Webster J. 1996. The role of macroinvertebrates in stream ecosystem function. Annual Review of Entomology 41:1, 115-139 
• Xu M, Wang Z, Duan X, Pan B. 2014. Effects of pollution on macroinvertebrates and water quality bio-assessment. Hydrobiologia, 729: 247-259.