Citation
Impacts of Road Crossings on Headwater Streams

Material Information

Title:
Impacts of Road Crossings on Headwater Streams
Creator:
McMorrow, Shannon E.
Place of Publication:
[Gainesville, Fla.]
Publisher:
University of Florida
Publication Date:
Language:
english
Physical Description:
1 online resource (83 p.)

Thesis/Dissertation Information

Degree:
Master's ( M.S.)
Degree Grantor:
University of Florida
Degree Disciplines:
Environmental Engineering Sciences
Committee Chair:
Crisman, Thomas L.
Committee Members:
Wise, William R.
Brenner, Mark
Graduation Date:
5/1/2008

Subjects

Subjects / Keywords:
Aprons ( jstor )
Chromium ( jstor )
Lead ( jstor )
Nickel ( jstor )
Outliers ( jstor )
Riprap ( jstor )
Sediments ( jstor )
Stormwater ( jstor )
Streams ( jstor )
Zinc ( jstor )
Environmental Engineering Sciences -- Dissertations, Academic -- UF
ecology, metals, streams, urban
City of Gainesville ( local )
Genre:
Electronic Thesis or Dissertation
bibliography ( marcgt )
theses ( marcgt )
Environmental Engineering Sciences thesis, M.S.

Notes

Abstract:
Urban development and associated roads adversely affect stream ecosystems through altered hydrology and subsequent erosion and contamination. Florida population growth leads to urbanization of natural landscapes and associated road development. Knowledge of the impacts of road crossings on stream systems will help guide better road design. Impacts of road crossings on stream geomorphology, sediment particle size distribution, organic matter storage, and metal contamination were evaluated. Samples were taken up and downstream of the upper most road crossing at nine headwater streams of Gainesville, Florida. Overall, areas downstream of road crossings were characterized by narrower channels, greater bank slopes, increased fine sediment, and higher metal concentrations. Concrete aprons and riprap are structural best management practices (BMPs) designed to dissipate energy to prevent scour and erosion. The effectiveness of these BMPs was evaluated by comparing impacts of crossings at managed and non-managed sites. Concrete aprons and swales were effective at dissipating erosive energy and minimizing scour; however, they were ineffective at minimizing metal and fine sediment contamination. Riprap was not effective at minimizing scour, but was effective at trapping organic matter, fine sediments, and metals within the riprap and plunge pool. A more comprehensive management technique is needed to mitigate all the adverse affects of road crossings on stream ecosystems. ( en )
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis:
Thesis (M.S.)--University of Florida, 2008.
Local:
Adviser: Crisman, Thomas L.
Statement of Responsibility:
by Shannon E. McMorrow

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright by Shannon E. McMorrow. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
7/11/2008
Classification:
LD1780 2008 ( lcc )

Downloads

This item has the following downloads:


Full Text






2
S1.8
" 1.6
1.4
1.2
1
0 0.8
C 0.6
iZ 0.4
0.2
0


Summer


- 4- Upstream
- Downstream


Winter


Figure 3-14. Average percent fine organic matter according to position relative to road
crossings and sampling period.


II


m 4
03

o
S2
0

0


15m 10mm 5m Om Om 5m 10m 15m

Upstream Downstream


Figure 3-15. Average coarse organic matter according to location in the stream relative to the
road crossing.

























-0
0 -

-0.5
-3.00


-2.50


-2.00 -1.50


-1.00


Log Fine Sediment


-OO
0 -


-0.5 -
-3.00


-2.50 -2.00 -1.50 -1.00 -0.50


Log Fine Sediment


F

Figure 3-25. Continued


-0.50


0.00


0.00

















-*- Summer
-U-Winter


Downstream


Figure 3-20.


25

. 20

c 15

0 10

5

0


Average coarse organic matter according to season and location relative to the
road crossing at concrete apron sites.


Non-Managed


i~ mII


15m 10m 5m

Upstream


Om


Om


5m 10m

Downstream


Figure 3-21.


Average percent organic matter according to location relative to road crossings
and management type: A) non-managed B) concrete apron C) riprap.


* 12

10
C
a 8

0 6

S4
o
0
C 2


Upstream


N
N
N

N

N
N


15m









bank erosion (Forman and Alexander 1998). McBride and Booth (2005) found that urban

streams had greater cross-sectional dimensions compared to non-urban streams.

There was no effect of season on width-to-depth ratios; however, bank slope was

significantly affected by season with higher bank slopes in winter than summer. Bank slope

measurements are more sensitive to slight alterations than width-to-depth measurement. The

change in bank slope may be due to storm events between summer and winter sampling. During

storm events water is flushed into the stream often causing erosion of the bed and bank.

Zone of Influence

Repeated measures ANOVA was used to test for differences among transects at varying

distances from the road crossing (0, 5, 10 and 15 meters up and downstream of the road

crossing). There were no significant differences among transects (0, 5, 10 and 15 meters up and

downstream of the road crossing) in width-to-depth ratio (F7, 62 = 1.26; P = 0.29) or bank slope

(F7,62 = 1.66; P = 0.13).

Geomorphology downstream of road crossings is significantly different from upstream;

however, no clear zone of influence or significant differences among transects could be

identified. When comparisons were made between transects, statistical power was lowered and

differences were not significant.

Influence of Management Type

Repeated measures ANOVA revealed a significant effect of management type on width-

to-depth ratios (F2,68 = 6.52; P = 0.003). There was a significant difference in width-to-depth

ratios between up and downstream at non-managed (Fi, 22 = 5.84; P = 0.02) and riprap sites (Fi, 22

= 5.52; P = 0.03); however, at sites with concrete aprons, there were no significant differences

(F1, 22 = 0.04; P = 0.84) (Figure 3-4). Downstream sections at non-managed and riprap sites had

lower (respectively: X = 6.59; X = 4.03) width-to-depth ratios than upstream sections









Mason 1999; Beasley and Kneale 2002). Accumulation in sediments can lead to

bioaccumulation in macroinvertebrates and biomagnification through food webs (Power and

Chapman 1992; Beasley and Kneale 2002). Concentrations of heavy metals in sediments,

benthic macroinvertebrates, and fish of streams receiving road runoff are positively correlated

with traffic densities (Van Hassel et al. 1980; Marsalek et al. 1999).

Best Management Practices

"BMPs are control practices used for a given set of conditions to achieve satisfactory

water quality and quantity enhancement at a minimal cost" (DeWiest and Livingston 2002).

There are a number of structural and non-structural BMPs for minimizing effects of peak

discharge in streams during storm events. A broad-scale approach in Puget Sound found that

BMPs within a watershed had a very weak and even negative correlation with biological indices.

However, a more intense study indicated that structural BMPs help sustain aquatic biological

communities in highly urbanized areas. At moderate urbanization levels, there was less evidence

of benefit (Avolio 2003). Horner and Mar (1983) found that passage of runoff through vegetated

channels greatly reduced concentrations of toxic solids and heavy metals. In addition, vegetated

buffer strips between roads and streams act as filters (Forman and Alexander 1998).

There are a number of BMPs utilized in Gainesville, FL to minimize erosive impacts of

stormwater runoff on streams. Swales are waterways intended to convey stormwater with

minimal erosion. Concrete aprons are used to dissipate energy at outlets of pipes to prevent

scour and minimize erosion caused by stormwater. Riprap is an erosion-resistant ground cover

of large, loose, angular stones used both to protect the soil surface from erosion by runoff and to

slow water velocity, while enhancing the potential for infiltration. Finally, riprap is also used to

stabilize slopes with seepage problems (DeWiest and Livingston 2002).









al. 1996). Sample metal concentrations were compared to baseline, threshold and probable effect

levels. No samples exceeded any of the three comparison levels for copper or chromium. One

sample exceeded baseline concentrations for zinc, but did not exceed threshold levels. No

samples exceeded baseline concentrations for nickel; however, one sample exceeded the

threshold effect level. Six samples exceeded the baseline and threshold effect levels for lead,

and one sample (the outlier) exceeded probable effect levels. Beasley and Kneale (2004) found

sediments of headwater streams in the UK most often exceeded Ontario Ministry of Environment

(OME) toxicity standards for lead and zinc.

The six samples that exceeded threshold effect levels for lead and the one sample that

exceeded threshold levels for nickel came from four different sites. Six out of the seven samples

that exceeded threshold effect levels were downstream of road crossings. One transect upstream

had lead concentrations which exceeded threshold effect levels, and zinc concentrations that

exceeded baseline levels. Headwater streams are expected to be uncontaminated; however, four

out of nine sites sampled contained sediments that exceeded threshold effect levels for lead and

one for nickel.

Table 3-1. Baseline, threshold, and probable effect levels
Metal Baseline levels (mg/kg) Threshold level (mg/kg) Probable effect levels (mg/kg)
Copper 0.22-21.0 35.70 197.00
Chromium 0.89-80.7 37.30 90.00
Nickel 1.70-48.5 18.00 36.00
Lead 0.69-42.0 35.00 91.30
Zinc 0.89-29.6 123.00 315.00
Baseline levels for Florida surface soils were established by Chen et al. (1999). Threshold and
probable effect levels were established by Canadian Council of Ministers of the Environment
(2002).









CHAPTER 4
CONCLUSIONS

Florida is projected to surpass New York in total population by 2011. In order to support

this growing population, new roads will be built accompanied by greater traffic volume.

Understanding mechanisms by which road crossings influence streams can guide stormwater

management decisions. The purpose of this study was to document the direct impacts of road

crossings on stream ecosystems by comparing up and downstream sections and to assess the

influence of current best management practices on these impacts.

Impact of Road Crossings

Roads and other impervious surfaces reduce infiltration rates and increase peak discharge

during storm events, resulting in degradation of stream ecosystems. To quantify alterations,

comparisons were made between up and downstream sections of road crossings. Downstream

sections had lower width-to-depth ratios and higher bank slopes producing narrower, incised

channels than upstream sections. Furthermore, there was a significant difference in bank slope

between summer and winter sampling suggesting continued erosion over the six month period.

Increased peak discharge results in erosion of bed and bank material, thus changing stream

dimensions and altering the subsequent flow regime. Material eroded from the bed and bank are

suspended and transported, becoming a source of fine material for downstream sections.

Fine sediment distribution is determined by local geology and flow regime. Roads carry

fine sediment from the watershed to the stream. In areas of low flow, suspended fine sediments

are deposited in the streambed. Downstream sections had slightly higher proportions of fine

sediment indicating increased deposition downstream of crossings. However, fine sediment is

easily resuspended and carried through the stream channel and deposited farther downstream,

thus depositional zones may have been missed due to sampling design.









LIST OF FIGURES


Figure page

2-1 Map of Florida showing the location of Gainesville with sampled sites highlighted.
M ap created in ArcGIS 9.1 ....... ........................... ..........................................19

3-1 Average width-to-depth ratio for up and downstream sections of study streams .............45

3-2 Average bank slope up and downstream relative to road crossings ..............................45

3-3 Average bank slope for summer and winter sampling periods .....................................46

3-4 Average width-to-depth ratio up and downstream relative to road crossings by
m anagem ent type .................................... .. ........... .......... ...46

3-5 Average bank slope for up and downstream sections of streams according to
m anagem ent type .................................... .. ........... .......... ...47

3-6 Average bank slope for summer and winter sampling periods according to
m anagem ent type. ..................................... .. .......... .......... ...47

3-7 Average bank slope for summer and winter sampling periods according to location
within the stream relative to the road crossing at sites with concrete aprons ..................48

3-8 Average percent fine sediment up and downstream relative to road crossings. ................48

3-9 Average percent fine sediment (<0.063 mm) up and downstream of road crossings
according to m anagem ent type ................................................ .............................. 49

3-10 Average percent fine sediment (<0.063mm) relative to road crossings by
m anagem ent type. ..................................... .. ......... .............. .. 50

3-11 Average percent fine sediment relative to the road crossing seasonally at non-
m managed sites. ..............................................................................5 1

3-12 Average percent fine and coarse organic matter in study streams upstream and
dow nstream of road crossing. ............................................................................................52

3-13 Average percent fine and coarse organic matter by sampling period.............................52

3-14 Average percent fine organic matter according to position relative to road crossings
an d sam pling p eriod .......... ........................................................................... ..... .. .... 53

3-15 Average coarse organic matter according to location in the stream relative to the road
cro ssing .......................................................... .................................52

3-16 Average percent organic matter according to location relative to the road crossing by
m anagem ent type .................................... .. ........... .............. .. 54









meters upstream was significantly greater than 0 meters upstream and 0, 10 and 15 meters

downstream (Figure 3-10). At riprap sites, fine sediment 15 meters upstream was significantly

less than 15 meters downstream (Figure 3-10).

At non-managed sites, percent fine sediment was much greater downstream than

upstream. Fine sediment is washed into the stream during storm events and deposited

downstream. Upstream, fine sediment distribution seemed uninfluenced by the road. Fine

sediments were higher at individual downstream than upstream transects; however, no clear zone

of influence could be identified. Transects with highest percent fine sediment were different

between summer and winter suggesting that depositional zones are constantly changing.

At concrete apron sites percent fine sediment was higher up than downstream. This is

likely due to the swales at two concrete apron sites that directed road runoff both up and

downstream. Fine sediment may selectively deposit in upstream sections. In addition, transects

10 meters upstream had very high average percent fine sediment, likely due to one concrete

apron site that displayed wetland like characteristics with very slow flows and flocculent bed

material and thus a depositional zone for fine sediment.

At riprap sites, no differences existed in percent fine sediment between up and

downstream. Riprap slows flows and allows deposition of fine sediment to occur within plunge

pools and riprap of these sites (DeWiest and Livingston 2002). No clear trend in fine sediment

deposition across transects existed at riprap sites. Riprap seems to be the most effective

management practice for trapping fine material; however, during severe storm events, these areas

could be sources of fine sediment.

Organic Matter Storage

Benthic particulate organic matter was measured up and downstream of road crossings to

determine their influence on organic matter storage. In headwater streams, allochthonous










1.9
1.85
1.8
1.75
1.7
1.65
1.6
1.55
1.5


Summer


Winter


Average bank slope for summer and winter sampling periods (F1, 68 = 4.44; P =
0.04).








O upstream
--* downstream






Non-managed Concrete Apron Riprap

Average width-to-depth ratio up and downstream relative to road crossings by
management type: non-managed (F1, 22 = 5.84; P = 0.02), concrete apron (F1,22:
0.04; P = 0.84), riprap (Fi, 22 =5.52; P= 0.03).


Figure 3-3.


Figure 3-4.


I









CHAPTER 3
RESULTS AND DISCUSSION

Geomorphology

Downstream of road crossings, streams experience increased peak discharge during rain

events as additional water is routed into the stream via the road. These events increase erosion of

the stream bed and bank. To quantify changes in stream geomorphology caused by road

crossings, cross-sectional profiles were measured up and downstream of roads.

Upstream to Downstream Difference

Width-to-depth ratio. Repeated measures ANOVA was used to test for differences in

width-to-depth ratio between up and downstream sections for all streams. There were significant

(F1,68 = 6.23, P = 0.02) differences in width-to-depth ratios between up (X = 9.64) and

downstream (X = 7.00) (Figure 3-1). There were no significant (F1,68 = 0.00; P = 0.98)

differences between summer and winter sampling periods.

Bank slope. Repeated measures ANOVA was used to test for difference in bank slope

between up and downstream sections. There was a significant (F1,68 = 11.77; P = 0.001)

difference in bank slope between up (X = 1.28) and downstream (X = 2.00) (Figure 3-2). There

was also a significant (F1, 68 = 4.44; P = 0.04) effect of season (Figure 3-3) with summer slopes

being lower (X = 1.63) than winter slopes (X = 1.75).

Roads had a significant impact on the geomorphology of the stream. Downstream

sections of sampled streams had lower width-to-depth ratios and higher bank slopes than

upstream sections. Downstream channels were more narrow and incised than upstream sections.

Road crossings facilitate connectivity of impervious surfaces to the stream channel, causing

accelerated peak discharge following storm events (Wang et al. 2001) leading to channel bed and










Table 2-1. Study Sites in Gainesville, Florida.


eitp


Headwater


type


Management


Non-
1 Seep Non-
managed

Non-
2 Seep Nn
managed

Concrete
3 Seep Apron
Apron

4 Seep Riprap

5 Wetland Riprap

Concrete
6 Wetland
Apron

7 Seep Riprap

Concrete
8 Wetland
Apron

Non-
9 Seep Non-
managed


Traffic
Volume
(Vehicles/day)

4,147


2,107


16,004


31,000

9,347


14,920


300


17,200


16,004


Field Methods

Sites were sampled during two periods: Summer (18-26 July, 2006) and winter (19-24

February 2007). During both periods sediment cores were taken and geomorphologic profiles

were measured. In addition, during the winter sampling period, streams and roads were

characterized, and sediment samples were collected for metal analysis.

At non-managed sites, sediment sampling and geomorphologic measurements were

conducted at 0, 5, 10, and 15 meters upstream and downstream of the road crossing. For sites

with concrete aprons, samples were taken at 0, 5, 10, and 15 meters upstream of the road

crossing and 0, 5, 10, and 15 meters downstream of the concrete apron. For sites with riprap,


Position
relative
to road

Up
Down

Up
Down

Up
Down
Up
Down
Up
Down

Up
Down
Up
Down

Up
Down

Up
Down


Drainage
area (m2)

29,444
32,412
44,988
58,116

36,835
44,986
177,209
206,563
127,354
143,438
133,158
143,379
60,682
77,747
1,054,653
1,275,809
13,049
21,056


Stream
Slope

0.035
0.031
0.059

0.025

0.020
0.045
0.005
0.041
0.005
0.010
0.003
0.009
0.015
0.015

0.003
0.009
0.019
0.006


Base
Velocity
(m/s)

0.03
0.04
0.04
0.05

0.05
0.05
0.11
0.06
0.11
0.06
0.05
0.13
0.06
0.04

0.02
0.02
0.07
0.05









Zone of Influence

ANOVAs were used to test for difference in metal concentrations among transects at

varying distances (0, 5, 10, and 15 meter up and downstream) from road crossings. No

significant difference was noted among transects for any metals (Figure 3-24).

There were no visible trends in metal concentrations among transects with distance from

road crossings. Flow regime, organic matter, particle size, traffic densities and distance from the

road all influence stream metal concentrations (Horowitz 1991; Rhoads and Cahill 1999);

therefore, position in the stream relative to the road does not fully explain concentration

variability.

Relationships among Metals, Organic Matter, and Particle Size

Linear regressions were used to test for relationships between percent fine sediment and

metal concentrations. Relationships were significant for copper (R2 = 0.15, P = 0.008),

chromium (R2 = 0.21, P < 0.0001), nickel (R2 = 0.18, P = 0.0002), lead with outlier (R2 = 0.15, P

= 0.0008), lead without outlier (R2 = 0.27, P < 0.0001) and zinc (R2= 0.26, P < 0.0001). (Figure

3-25).

Linear regressions were used to test for relationships between organic matter and metal

concentrations. Relationships were significant for copper (R2 = 0.14, P = 0.001), chromium (R2

= 0.13, P = 0.002), nickel (R2 = 0.09, P = 0.01), and zinc (R2= 18.73, P = 0.0001), and

approached significance for lead without the outlier (R2 = 0.05, P = 0.09), but was not significant

with the outlier (R2 = 0.01, P = 0.33).

Metal binding capacity has great influence on the location of "hot spots" of metal

concentrations. The association among organic matter, particle size, and trace metals is well

documented and apparent in this study (Horowitz 1991; Estebe et al. 1997; Lee et al. 1997).

Smaller particles have a higher surface area to volume ratio, and therefore have increased










16
14
12
WE 10
OUp
0-. Down



(. 0

Non-Managed Concrete Apron Riprap
Management Type


Figure 3-9. Average percent fine sediment (<0.063 mm) up and downstream of road crossings
according to management type: non-managed (F, 70 = 16.72, P = 0.0001),
concrete apron (Fi,64 = 4.57, P = 0.04), and riprap (Fi,70 = 2.11, P = 0.15).
















-*- Summer
- ,- Winter


Upstream Downstream


Figure 3-18.


Average total organic matter by sampling period and location relative to the road
crossing at concrete apron sites.


-*- Summer
- -Winter


Upstream Downstream


Figure 3-19.


Average fine organic matter according to season and location relative to the road
crossing at concrete apron sites.









dissolved oxygen and habitat heterogeneity, thus acting as "ecological traps". Further research is

needed to test this hypothesis.

Concrete aprons and riprap seem to have their benefits and downfalls in terms of

minimizing the impact of roads on stream systems. Concrete aprons or the swales associated

with these sites minimize erosion of downstream sections; however, fine sediment and metal

contamination were not reduced. Riprap, allows deposition of fine sediment, organic matter, and

metals in close proximity to roads; however, this zone may become highly contaminated.

Observations during this study suggest future roads that cross streams include vegetated swales

as opposed to concrete swales. Vegetation provides infiltration capacity and reduction of peak

discharge, thus reducing erosion. Furthermore, vegetated swales encourage filtration processes

that reduce the load of fine sediment and heavy metals transported to streams.

Best management practices evaluated in this study were confined to areas directly

downstream of the road crossing; however, management within the watershed and better road

design would greatly minimize the impacts of the roads on streams. Generally, stormwater BMP

decisions are made after the roadway has already been designed. A better approach would be to

integrate stormwater management features into road designs.









sites, 0 meters downstream was significantly less than 15, 10, 5, and 0 meters upstream and 15

meters downstream, while 5 meters downstream was significantly less than 10 and 0 meters

upstream. At riprap sites, 15 meters upstream was significantly less than 10 meters upstream and

15 meters downstream (Figure 3-21).

Management type had a significant influence on organic matter storage. At non-managed

sites, total and fine organic matter was slightly greater upstream than downstream and coarse

organic matter was not significantly different. These differences may be due to selective

flushing of organic matter downstream; however, they could also be associated with local

depositional patterns and canopy cover.

At concrete apron sites, total, fine, and coarse organic matter were much greater upstream

than downstream. Coarse organic matter at individual downstream transects was significantly

less than upstream transects, but no clear zone of influence could be identified. The culvert at

concrete aprons may trap organic matter upstream, and increased peak discharge may flush

organic matter from downstream sections; however, swales associated with two concrete apron

sites may also influence organic matter depositional patterns. Upstream of one concrete apron

site displayed wetland like characteristics with very low flows and flocculent bed material.

Organic matter selectively deposits in areas of low flow (Allan 1995), resulting in elevated

organic matter concentrations in these areas.

At riprap sites, total and coarse organic matter were greater downstream than upstream.

At riprap sites, angular stones are in place to slow flow and dissipate erosive energy (DeWiest

and Livingston 2002). The rocks have the ability to trap organic matter and the slower flows

allow organic matter to deposit within the stream bed. There were significant differences in total

and coarse organic matter between individual upstream transects as well as differences in coarse
















E Upstream
U Downstream


Non-managed Concrete Apron Riprap


Figure 3-28. Continued









Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science

IMPACTS OF ROAD CROSSINGS ON HEADWATER STREAMS

By

Shannon E. McMorrow

May 2008

Chair: Thomas L. Crisman
Major: Environmental Engineering Sciences

Urban development and associated roads adversely affect stream ecosystems through

altered hydrology and subsequent erosion and contamination. Florida population growth leads to

urbanization of natural landscapes and associated road development. Knowledge of the impacts

of road crossings on stream systems will help guide better road design. Impacts of road

crossings on stream geomorphology, sediment particle size distribution, organic matter storage,

and metal contamination were evaluated. Overall, areas downstream of road crossings were

characterized by narrower channels, greater bank slopes, increased fine sediment, and higher

metal concentrations.

Concrete aprons and riprap are structural best management practices (BMPs) designed to

dissipate energy to prevent scour and erosion. The effectiveness of these BMPs was evaluated

by comparing impacts of crossings at managed and non-managed sites. Concrete aprons and

swales were effective at dissipating erosive energy and minimizing scour; however, they were

ineffective at minimizing metal and fine sediment contamination. Riprap was ineffective at

minimizing scour, but was effective at trapping organic matter, fine sediments, and metals within

the riprap and plunge pool. A more comprehensive management technique is needed to mitigate

all the adverse affects of road crossings on stream ecosystems.










4

3.5

3

2.5

2

1.5

1

0.5

0
Non-managed Concrete Apron Riprap


120
100
80
60
40
20
0


Non- Concrete Riprap (with
managed Apron outlier)


O Upstream
* Downstream


E Upstream
* Downstream


Riprap
(without
outlier)


D

Figure 3-28. Continued









Organic matter storage is determined by source and flow regime. There were no

significant differences in organic matter storage between up and downstream. There was a

significant relationship between fine sediment and organic matter storage because both are

influenced by flow regime and are deposited in areas of low flow.

Metals are deposited on road surfaces through leaks and wear of automobiles and carried

to the stream by stormwater runoff, preferentially adsorb to fine sediment and organic matter,

and are therefore concentrated in depositional areas. Downstream sections had higher

concentrations of chromium and lead. Lack of significant differences may be due to low sample

size or the fact that upstream sections may also be impacted by road runoff.

Overall differences between up and downstream sections suggest degradation of

downstream areas, expressed as more incised channels, steeper banks, more fine sediment, and

higher metal contamination. Upstream sections may be influenced by roads as well. At some

sites, high proportions of fine sediment and high concentrations of metals were found. There

was much variability because all of these variables interact with one another and are influenced

by additional factors such as flow regime, traffic density, land-use, and management practices.

Regardless of additional variables, there was evidence that overall downstream sections were

degraded.

Influence of Management Practices

Two in-stream, best management practices (BMPs) were in place at the onset of this

study: concrete aprons and riprap. Downstream of stormwater outlets, such as culverts, concrete

aprons and riprap are placed to dissipate energy to prevent scour and erosion. To assess the

efficiency of these BMPs, up and downstream of road crossings were compared among non-

managed, concrete apron and riprap sites. Areas downstream of non-managed sites display the











* Research question 1. How is sediment size distribution within small streams affected by
road crossings?

* Research question 2. How is geomorphology of small streams altered by road crossings?

* Research question 3. How is organic matter storage of small streams affected by road
crossings?

* Research question 4. How are sediment metal concentrations within small streams
affected by road crossings?

* Research question 5. How do current BMPs influence impacts of road crossings on small
streams, specifically in terms of stream morphology, sediment size distribution, organic
matter storage and sediment metal concentrations?









organic matter between individual up and downstream transects; however, no clear zone of

influence could be identified.

Organic matter transport and storage is influenced by flow regime. Differences in storage

may be a result of the road crossings and management type, but may also be associated with

stream geomorphology or canopy cover.

Fine Sediment and Organic Matter Relationship

Organic matter and fine sediment show similar depositional trends across study streams.

Linear regression was used to test for a relationship between organic matter and fine sediment in

stream bed material. There was a positive significant relationship between the log proportion

organic matter and log proportion fine sediment (R2= 0.33, P < 0.0001) (Figure 3-22).

The relationship between organic matter and fine sediment is well studied (Horowitz

1991; Rhoads and Cahill 1999). The deposition of fine sediments and organic matter is

dependent on flow regime and density (Allan 1995). They tend to be transported by fast moving

water and deposit in areas of low flow.

Metals

Road runoff contains contaminants including heavy metals associated with automobile

leaks and wear. Road crossings drain impervious surfaces to the stream network, and thus are

point sources for stormwater runoff. Copper, chromium, nickel, lead, and zinc were measured

up and downstream of the road crossing. One result from stream number four was an outlier with

respect to lead (550 mg/kg). It is unclear whether this was due to sampling error or a "hot spot"

of lead contamination; therefore, statistical tests involving lead were run twice, once with the

outlier included and once without.


























Figure 3-12.


SUp
I Down





Fine Coarse


Average percent fine and coarse organic matter in study streams upstream and
downstream of road crossing: fine (F1, 212= 0.49, P = 0.49) and coarse (F1, 212
0.64, P = 0.42).









O Summer
1 Winter


Fine Coarse


Figure 3-13.


Average percent fine and coarse organic matter by sampling period: fine (F1, 212
= 41.12, P < 0.0001) and coarse (F1,212= 32.63, P < 0.0001).









ACKNOWLEDGMENTS

I thank my family and friends for their unending love and support. I thank specifically

Marcus Griswold for his work and guidance on this project. I thank the volunteers who

accompanied in the field to collect data. I thank my parents for loving and supporting me

throughout my life and encouraging me to follow my dreams. I thank my major advisor Dr.

Thomas L. Crisman for guidance and support throughout my graduate experience. I thank Dr.

Mark Brenner and Dr. William Wise for their advice. I thank the Florida Stormwater

Association for awarding me their annual scholarship, which allowed me to evaluate metal

concentrations. And I thank Sean for being my rock and loving me no matter what.









for stream biota including alteration of habitat and interference with respiration, feeding, and

reproduction (Wood and Armitage 1997).

Zone of Influence

Repeated measures ANOVA revealed no significant (F,210 = 1.13, P = 0.34) differences

in percent fine sediment among transects at varying distances (0, 5, 10 and 15 meters up and

downstream) from road crossings. Roads act as a source of fine sediments; however, flow

regime drives sediment depositional processes (Allan 1995).

Influence of Management Type

Repeated measures ANOVA revealed a significant effect of management type on percent

fine sediment (F2, 210 = 24.83, P < 0.0001). Differences between up and downstream were

significant for non-managed (F, 70 = 16.72, P = 0.0001) and concrete apron sites (Fi, 64 = 4.57, P

= 0.04); however, not for riprap sites (F, 70 = 2.11, P = 0.15). At non-managed sites, percent

fine sediment was greater downstream (X = 12.70) than upstream (X = 1.93), while at concrete

apron sites, upstream (X = 9.96) was greater than downstream (X = 4.78) (Figure 3-9).

Repeated measures ANOVA was used to test for differences among transects at varying

distances (0, 5, 10 and 15 meters up and downstream) from roads according to management type.

There were significant differences among transects at non-managed (F7,64= 2.87, P = 0.01) and

riprap (F7, 64 = 2.15, P = 0.05) sites, and differences approached significance at concrete apron

sites (F7, 62 = 1.82, P = 0.10). Tukey post-hoc comparisons were used to test for pairwise

differences among all transects. At non-managed sites 0 meters downstream was significantly

greater than 10 meters upstream; and 5 meters downstream was significantly greater than 15, 10,

and 5 meters upstream (Figure 3-10). Effects of location were dependent on season at non-

managed sites (F7,64 = 3.95, P = 0.001), with "hot spots" of high percent fine sediment varying

between summer and winter sampling (Figure 3-11). At concrete apron sites, fine sediment 10









organic matter is a primary food base for macroinvertebrates. Culverts may disrupt natural

depositional patterns of organic matter. In addition, increased peak discharge associated with

road crossings, may result in the flushing of organic matter downstream.

Upstream to Downstream Differences

Repeated measures ANOVA was used to test for differences in percent organic matter

between up and downstream sections for all streams. No significant differences were found for

total (F1, 212 = 0.07, P = 0.79), fine (F1, 212 = 0.49, P = 0.49) and coarse (F1, 212 = 0.64, P = 0.42)

organic matter (Figure 3-12). There was a significant seasonal effect on total (F1,212 = 51.24, P <

0.0001), fine (F1,212 = 41.12, P < 0.0001) and coarse (F1,212= 32.63, P < 0.0001) organic matter,

with summer (respectively: x = 3.33, x = 0.96, x = 2.37) lower than winter (respectively: X =

4.87, x = 1.42, X = 3.46) (Figure 3-13). Seasonal effects were dependent on location within the

stream (F1,212 = 6.77, P = 0.01) for fine organic matter, with differences between summer and

winter being greater for downstream (respectively: X = 0.54, X = 1.07) than upstream

(respectively: X = 1.39, X = 1.76) (Figure 3-14).

Road crossings had no apparent influence on organic matter storage between up and

downstream. There was, however, a strong seasonal component, with winter organic matter

(fine and coarse) storage greater than summer. The greatest leaf fall in Florida usually occurs

between September and December, with a smaller peak in January and February (Roberts 2002),

resulting in more organic matter during these periods. Seasonal influence on fine organic matter

was stronger downstream than upstream, potentially due to summer storm events selectively

flushing fine organic matter from downstream sections.











I


[-7


Upstream Downstream Upstream Downstream


Non-managed


Concrete Apron


Upstream Downstream
Riprap


Figure 3-16.


Average percent organic matter according to location relative to the road crossing
by management type: total organic matter for non-managed (F1, 70 = 3.27, P =
0.08), concrete apron (F1, 70 = 10.77, P = 0.002) and riprap (F1, 70 = 6.29, P = 0.01)
sites.


Summer Winter Summer Winter Summer
Non-managed Concrete Apron Rip


Winter


rap


E Coarse
* Fine


Figure 3-17. Average percent organic matter according to sampling period and management
type: total organic matter at non-managed (F1,70= 10.51, P = 0.002), concrete
apron (F, 70= 12.49, P = 0.0007) and riprap (F, 70= 32.19, P < 0.0001) sites.


O Coarse
* Fine


GL~










TABLE OF CONTENTS

page

A CK N O W LED G M EN T S ................................................................. ........... ............. .....

L IS T O F T A B L E S .................................................................................7

LIST OF FIGURES .................................. .. ..... ..... ................. .8

A B S T R A C T ......... ....................... ............................................................ 10

CHAPTER

1 INTRODUCTION ............... .............................. ........................ .... 11

Effects of Urbanization on Stream Ecosystems............................................ ................11
Effects of Roads on Stream Ecosystem s ........................................ .......................... 13
H y drology ..............................................................................................13
G e o m o rp h o lo g y ......................................................................................................... 13
H habitat Structure ............................................... 14
M etal Contam nation ................................ ................... ....... .. ........ .... 14
S tu d y O b je c tiv e s ..................................................................................................................... 1 6

2 M A TER IA L S A N D M ETH O D S ........................................ ............................................18

Site Selection and D description ..................................................................... ............... 18
S ite S ele ctio n ........................................... ............................... 1 8
Site Description ................................. ................................ ......... 18
G IS W watershed D elineation ............. ..................................................... ............... 18
F ie ld M eth o d s ................................................................................................................... 2 0
G eom orphology ............ ....................................................................... ......21
Sedim ent .................... ........................ ..................... ............... 21
M etals .........................................................21
Stream C characterization .............................................................22
R o ad C h aracterization ............................................................................................... 2 2
Laboratory Methods................................. ......... 22
Sedim ent Particle Size D istribution........................................... 22
Organic M matter ................. ................................................. ........ 23
Metals ..........................................23
D ata A n aly sis ................................................................................................. ...............2 3
G eom orp h ology .............................. ......................................................2 4
Sedim ent Particle Size D istribution........................................... 24
O organic m matter ............... ...................................................................................24
Statistical A nalyses................................................... 25









Statistical Analyses

The following statistical analyses were performed for width-to-depth ratios, bank slopes,

fine sediment content, and total, fine, and coarse organic matter content. Repeated measures

Analysis of Variances (ANOVAs) were used to test for effects of position relative to road

crossings (up versus down), season (summer versus winter), management type (non-managed,

concrete apron, and riprap) and interactions of these variables across all streams and then by

management type. Repeated measures ANOVAs were also used to test for effect of position in

the stream (0, 5, 10, and 15 meters up and downstream), season, management type and

interactions of these variable across all streams and then by management type.

T-tests were performed to test for difference in metal concentrations between up and

downstream sections. ANOVA tests were used to test for differences in metal concentrations

among transects at varying distances (0, 5, 10 and 15 meters up and downstream) from road

crossings. Tukey post-hoc comparisons were used to test for pairwise differences among all

transects for significant ANOVA results. Linear regressions were used to test for linear

relationships between metal concentrations and percent organic matter, percent fine sediment

(<63-tm), and traffic densities.









Study Objectives

The shift in hydrologic regime associated with urbanization is magnified both by roads

and connectivity of impervious surfaces to streams (Alberti et al. 2007). Although studies on the

effects of urbanization are common (reviewed in Paul and Meyer 2001), few have examined the

impact of urban road crossings on stream ecosystems. The literature is lacking in four major

areas: 1) Impacts of urban roads broadly, 2) Impacts of urban roads at the local scale, 3)

Impacts of urban roads in Florida, and 4) Influence of best management practices on such

impacts.

The majority of the research has examined the impacts of road crossings on streams in

forested landscapes (e.g. Montgomery 1994; Madej 2001). The impacts of roads in an urban

context are much different than in forested landscapes, and the majority of literature is more

closely related to the overall impact of urbanization and is not specific to roads (May et al. 1997;

Konrad 2000; McBride 2001). The majority of the limited research on the impact of urban roads

on streams has been conducted in the temperate zone (Van Hassel et al. 1980; Marsalek 1990;

Perdikaki and Mason 1999; Forman and Deblinger 2000; Gidding et al. 2001; Grapetine et al.

2004), with streams very different from those of Florida. Florida has unique topography,

geology, and climate, and its streams display sandy bottoms and low flow (Whitney et al. 2004).

Finally, impacts of roads on streams have not addressed effectiveness of best management

practices. Efforts have obviously been made to mitigate the impact of road crossings on streams

through BMPs; however, research on their effectiveness has been neglected.

This study investigated the direct influence of road crossings on a stream ecosystem by

measuring channel morphology, particle size, organic matter storage, and metal accumulation.

The effectiveness of best management practices downstream of road crossings in relation to

these variables was also considered.




























-3 -2.5 -2 -1.5 -1 -0.5


Log Fine Sediment


Figure 3-22.


Linear relationship between log proportion organic matter and log proportion fine
sediment (R2 0.33, P < 0.0001).


-0.5

-1

-1.5

-2

-2.5

-3


-3.5
-3.5
































To my family and friends without whose support I could not have done it.










Non-Managed


30
.E 25

C E 20
(D E
iE 0 15 -

V 10 -
5

0. 0
15 10 5 0

Upstream


0 5 10 15

Downstream


Concrete Apron


25
E
20
(/ E
SE 15

-0 10
,-
S 5

0
a.


Figure 3-10. Average percent fine sediment (<0.063mm) relative to road crossings by
management type.









expected degradation associated with roads, with incised channels, steeper bank slopes, more

fine sediment, and higher metal concentrations.

Concrete aprons and swales associated with two of the study sites seemed to dissipate

erosive energy resulting in less difference in channel morphology between up and downstream

sections. Swales route water off the road, providing more storage and slowing flows, thus

dissipating the increased peak discharge during storm events that causes erosion. Concrete

aprons seem to be effective at dissipating erosive energy and minimizing downstream erosion;

however, after closely examining the results, sites with swales display less difference in

geomorphology between up and downstream. Investigation of the influence of swales on road

crossing impacts is needed to verify these results. Concrete apron sites also exhibited both much

higher organic matter and fine sediment concentrations upstream than downstream and elevated

levels of metals. Downstream sections were not significantly different from upstream in metal

concentrations, suggesting that both were equally influenced by metals in road runoff Although

management at these sites is effective at minimizing erosion, up and downstream sections are

compromised by road crossings through metal and fine sediment contamination.

Riprap sites displayed the least differences between up and downstream. The latter were

more incised and had higher bank slopes than upstream. Riprap seems inefficient at dissipating

erosive forces. Fine sediment and metal concentration were not significantly different between

up and downstream, likely due to deposition within the plunge pool and riprap. Organic matter

was higher downstream than upstream, likely due to slower flow allowing deposition. Riprap

seems to be the most efficient management practice for trapping organic matter, fine sediments,

and metals associated with these materials, but may become highly contaminated due to selective

deposition. These areas may also be selectively utilized by macroinvertebrates due to increased









watershed to filter nutrients, sediment, and other contaminants before entering the stream. Total

impervious surface area within a drainage basin has often been used as an indicator of urban

stream impairment (Klein 1979; Paul and Meyer 2001; Robson and Beech 2006).

Due to increased impervious surface area, there is a shift in stream hydrologic regime

indicated by an increase in peak discharge during storm events, commonly referred to as "flashy

hydrology." This leads to increased erosion and suspended sediment, resulting in movement of

fine sediments and organic matter downstream, coupled with increased turbidity (Bilby et al.

1989). Grain size distribution in urban streams tends to shift toward smaller particles; however

in highly urbanized streams, smaller grain sizes are selectively removed during storm events

(McBride and Booth 2005). The degree of sedimentation can alter channel morphology and

aquatic habitat (Forman and Alexander 1998). McBride and Booth (2005) found that urban

streams have greater cross-sectional dimensions, increased erosion rates, and simplified

morphology compared to non-urban streams.

Urbanization can also increase heavy metal concentrations in streams, especially lead,

nickel, zinc, copper, chromium, and cadmium that are concentrated in road runoff (Forman and

Alexander 1998; Gidding et al. 2001). Such heavy metals are detrimental to stream biota and

can be incorporated into food webs leading to bioaccumulation and biomagnification in longer-

lived species (Beasley and Kneale 2002).

Urbanization within a drainage basin can also have dramatic impacts on stream

ecosystem structure and function. Not only the degree of urbanization (Klein 1979; Paul and

Meyer 2001; Robson and Beech 2006) but the configuration may have differential impacts on

streams. Road density, number of crossings, and connectivity of impervious surface areas within









Concrete Apron


35
S30
S25
2 20
c 15
0 10
5
0


10m 5m
Upstream


Om Om


5m 10m
Downstream


Riprap


C

Figure 3-21. Continued


15m


15m


S30
S25
0 20
c 15
0 10
5
0


* Coarse
m Fine


T T


I i OTO a it


-
























Upstream Downstream

Average width-to-depth ratio for up and downstream sections of study streams
(F1,68= 6.23, P = 0.02).


Upstream


Downstream


Average bank slope up and downstream relative to road crossings (F1,68 = 11.77;
P = 0.001).


Figure 3-1.


2.5

2

1.5

1

0.5

0


Figure 3-2.


I









Willson, J. D. and M. E. Dorcas. 2003. Effects of habitat disturbance on stream salamanders:
implications for buffer zones and watershed management. Conservation Biology 17:763-
771.

Wood, P. J. and P. D. Armitage. 1997. Biological effects of fine sediment in the lotic
environment. Environmental Management 21:203-217.









Zone of Influence

Repeated measures ANOVA was used to test for differences among transects at varying

distances from road crossings (0, 5, 10, and 15 meters up and downstream). There were no

significant differences for total (F7, 206 = 1.01, P = 0.43) or fine organic matter (F7, 206 = 0.47, P =

0.86); however, there were significant differences for coarse organic matter (F7, 206 = 2.20, P =

0.04). Tukey post-hoc comparisons were used to test for pairwise differences among all

transects, and resulted in a significant difference in coarse organic matter between 0 meters

upstream (X = 3.76) and 0 meters downstream (X = 1.28) (Figure 3-15).

There is no clear trend in organic matter storage associated with location in the stream

relative to the road. Coarse organic matter was greater 0 meters upstream than 0 meters

downstream; however, no other significant differences existed. Variability in organic matter

storage is likely due to flow and local depositional patterns, as opposed to proximity to road

crossings.

Influence of Management Type

Management type had a significant influence on total (F1,212 = 27.14, P < 0.0001), fine

(Fi, 212 = 38.09, P < 0.0001) and coarse (F,212 = 19.79, P < 0.0001) organic matter. Differences

in total organic matter between up and downstream approached significance (Fi, 70 = 3.27, P =

0.08) at non-managed sites and were significantly different at concrete apron (Fi,70 = 10.77, P=

0.002) and riprap sites (F, 70= 6.29, P = 0.01). At non-managed and concrete apron sites, total

organic matter was higher upstream (respectively: X = 2.42, X = 11.56) than downstream (X =

2.27, X = 4.99); however, at riprap sites, total organic matter was lower upstream (X = 1.37)

than downstream (X = 2.00) (Figure 3-16). There was a significant seasonal effect on total

organic matter at non-managed (F1, 70 = 10.51, P = 0.002), concrete apron (F1, 70 = 12.49, P =









LIST OF TABLES

Table page

2-1. Study Sites in G ainesville, Florida........................................................ ............... 20

3-1. Baseline, threshold, and probable effect levels...................................... ............... 39

3-2. Primary sources of heavy metals in road runoff ..........................................................42
















O Upstream
* Downstream


Non-managed Concrete Apron Riprap


Average bank slope for up and downstream sections of streams according to
management type: non-managed (F1, 22 = 16.66; P = 0.0005), concrete apron (F1,
22 = 0.23; P = 0.63), and riprap (F1,22 = 3.63; P = 0.07).







T" Summer
U Winter






Non-managed Concrete Apron Riprap


Average bank slope for summer and winter sampling periods according to
management type: non-managed (F1, 22 = 12.63; P = 0.002).


) 2.5
0
n

m 1.5

(0 1

q 0.5

0


Figure 3-5.



3.5


0




Figure 3-6.









(respectively: X = 12.69; X = 6.89). Width-to-depth ratios were not significantly different

among transects (0, 5, 10 and 15 m up and downstream of road crossing) for non-managed,

concrete apron or riprap sites (respectively: F7, 16 = 1.13; P = 0.39; F7, 16 = 0.19; P = 0.98; F7, 16 =

1.85; P = 0.15).

Repeated measures ANOVA revealed a significant effect of management type on bank

slope (F1, 68 = 4.47; P = 0.02). Bank slope was significantly different between up (X = 0.99) and

downstream sections (X = 2.24) at non-managed sites (F1, 22 = 16.66; P = 0.0005); differences

between up (X = 1.64) and downstream sections (X = 2.39) approached significance at riprap

(F1, 22 = 3.63; P = 0.07) sites; however, there was no significant difference at concrete apron sites

(F1, 22 = 0.23; P = 0.63) (Figure 3-5). Seasonal effects on bank slope were not significant at

managed sites (concrete aprons and riprap); however, there was a significant difference in bank

slope at non-managed sites between summer (X = 1.49) and winter (X = 1.75) sampling periods

(F1, 22 = 12.63; P = 0.002) (Figure 3-6).

There were no significant differences among transects (0, 5, 10 and 15m up and

downstream of road crossing) for bank slope for non-managed, concrete apron, or riprap sites

(respectively: F7, 16 = 2.12; P = 0.11; F7, 16 = 0.25; P = 0.96; F7, 16 = 2.11; P = 0.10). Seasonal

effects on bank slope were dependent on location within the stream for concrete apron sites (F7,

16 = 5.77; P = 0.002) (Figure 3-7).

Management type had a significant influence on stream geomorphology. Concrete

aprons and riprap are intended to slow flow directly downstream of road crossings to minimize

erosion of the stream (DeWiest and Livingston 2002). Downstream sections at non-managed

and riprap sites were more incised and had higher bank slopes than downstream sections;

however, up and downstream sections of concrete apron sites were not significantly different.









IMPACTS OF ROAD CROSSINGS ON HEADWATER STREAMS


By

SHANNON E. MCMORROW


















A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE

UNIVERSITY OF FLORIDA

2008









significant for lead (R2= 0.60, P = 0.01) and nickel (R2 = 0.44, P = 0.05), and approached

significance for chromium (R2 = 0.37, P = 0.08); however, were not so for copper or zinc.

(Figure 3-27)

Van Hassel et al. (1980) and Marsalek et al. (1999) also found that traffic densities

influence metal accumulation in stream sediment. The primary sources of heavy metals in runoff

are automobiles (Table 3-2); however, metals associated with building roofing and siding

materials are also transported to the stream via the road (Davis et al. 2001). Nickel and lead

exhibit the strongest correlation to traffic densities. The greatest current source of lead is brick

walls, followed by wet deposition (Davis et al. 2001). Levels of lead in stormwater runoff are

often low; therefore, high levels found in these streams may be due to historic use of leaded

gasoline (Turer et al. 2001). Nickel is associated with diesel fuel, oil, metal plating, brake

linings and asphalt paving (Grant et al. 2002). The other metals are also associated with

automobiles (Table 3-2) and correlations with traffic densities are apparent. Although there are

no clear trends in metal concentrations by location in the stream, traffic densities associated with

each site, explain some of the variability.

Table 3-2. Primary sources of heavy metals in road runoff.
Metals Sources
Chromium Tire wear, brake pads, combustion of oils, and insecticides
Metal plating, bearing and bushing wear, moving engine parts, brake lining wear,
r fungicides and insecticides
Lead Leaded gasoline and tire wear
Nicl Diesel fuel and gasoline, lubricating oil, metal plating, bushing wear, brake lining wear, and
asphalt paving
Zinc Tire wear, motor oil, grease
Adapted from Grant et al. 2003.

Influence of Management Type

Management type had a significant effect on metal concentrations for copper (F2,68

4.01, P = 0,02) and zinc (F2, 68 = 4.70, P = 0.01); approached significance for chromium (F2, 68
















0
-0.5
-1

-1.5
-2


-2.5
-2.50


-2.00


-1.50


-1.00


Log Organic Matter


3
2.5
2
1.5
1
0.5
0
-0.5


-2.50


-2.00


-1.50


-1.00


Log Organic Matter


D



Figure 3-26 Continued


*t .


-0.50


0.00


-0.50


0.00


,w









aqueous hexametaphosphate to disaggregate clay particles (10 mL / 25 g sample). It was then

passed through a 63-tlm sieve to separate silts and clays from coarser sediment. Both retained

fractions were dried between 60-70C for 48 hours and ashed in a muffle furnace at 550C for 5

hours. Upon drying, samples were weighed to 0.001 g. Sediment was then passed through a

nest of 2 sieves (2-mm and 63-[tm), and each fraction was weighed to 0.001 g (Plumb 1981).

Organic Matter

To determine how roads influence organic matter storage within the sediments, sediment

samples were homogenized and a 5-cc sub-sample of the sediment was taken to calculate organic

matter content. It was wet sieved through 1 mm and 63-[tm sieves to separate coarse from fine

organic matter. Each fraction was dried in an oven at 60-70C until dry (48 hours), weighed to

0.001 g and ashed in a muffle furnace at 550C for 5 hours. Upon drying, the samples were

placed in a decondenser and reweighed.

Metals

Metal samples sent to the laboratory were analyzed in wet material, therefore, moisture

content of the duplicate samples was measured and metal concentrations were converted to dry

weight. Each sediment sample was homogenized and three replicate 5-cc sub-samples were

taken to measure moisture content. Samples were weighed to the nearest 0.001 g, dried for 48

hours between 60-70C, and re-weighed to the nearest 0.001 g.

Data Analysis

All statistical analyses were performed using SAS 8.02 software. Shapiro-Wilks

normality tests were used to test for normal data distributions. Width-to-depth ratio, bank slopes,

fine sediment, organic matter, and metal concentrations were loglO transformed for further




















02



.0 0O E

7:)




















v













Clj
19


E CA
01h
\ ;CA


C)
C11 CA



hC
~~C j








CA

~-ln a








Ccj



AJ;i




zs






19Z









2.90, P = 0.06) and nickel (F2,68 = 2.86, P = 0.06); but was not significant for lead with (F2,68

0.15, P = 0.86) or without (F2,68 = 0.88, P = 0.42) the outlier.

T-tests were used to test for difference between up and downstream metal concentrations

by management type. At non-managed sites, there were significant differences for nickel (T = -

2.78, P = 0.01) and lead (T = -2.13, P = 0.05). Differences approached significance for

chromium (T = -1.83, P = 0.08) and copper (T = -1.90, P = 0.07); and were not significant for

zinc (T = -1.66, P = 0.12). Differences between up and downstream were not significant for any

metals at concrete apron and riprap sites. (Figure 3-28)

ANOVAs were used to test for differences in metal concentrations among transects at

varying distances (0, 5, 10, 15 meters up and downstream) from road crossings according to

management type. There were no significant differences among transects for non-managed,

concrete apron, or riprap sites.

Metal concentrations were higher downstream than upstream at non-managed sites;

however, at concrete apron and riprap sites, differences were not significant. Concrete aprons

and riprap slow flow directly downstream of road crossings, allowing materials to deposit closer

to the road (DeWiest and Livingston 2002), thus minimizing the downstream zone of influence.

At concrete apron sites, copper and zinc were in higher concentrations upstream than

downstream. At two of the concrete apron sites, there were concrete swales that routed runoff

both up and downstream. Swales are waterways intended to convey stormwater to the stream

with minimal erosion (DeWiest and Livingston 2002). Erosion may be minimized; however,

contamination of upstream sections may be facilitated. This may be an additional factor

influencing metal concentration distribution within streams. For riprap sites, finer sediments,

organic matter and metals adsorbed to those materials would deposit within the plunge pool and























I i



0 10000 20000 30000 40000
Traffic Density (Vehicles/Day)


1.2
1
0.8
0.6
0.4
0.2
0
-0.2


-U.4


0


10000 20000 30000 40000
Traffic density (Vehicles/ Day


0 Upstream



Dow nstream



Linear (Dow nstream)



Linear (Upstream)




Figure 3-27. Continued


^ 0 U.
5









BIOGRAPHICAL SKETCH

Shannon McMorrow grew up in Fairfield, Connecticut, where she developed a love for

both math and science. Upon graduation from high school, Shannon moved to Gainesville,

Florida, to start her college career majoring in zoology at the University of Florida. During this

time she designed and implemented an independent study project investigating the sexual

behavior of Florida flag fish. She also decided to minor in environmental sciences to expand her

educational scope. Shannon graduated summa cum laude in May of 2004.

Shannon next volunteered at a wild life rehabilitation center in Friday Harbor, Washington,

for three months. There she helped care for a variety of animals including harbor seals. While

in Washington, Shannon realized she wanted to return to school to study applied ecology. In

January 2005, Shannon started graduate school back at the University of Florida at the Center for

Wetlands. Shannon's research interests included but were not limited to investigation and

mitigation of impacts of humans on wildlife and the environment. This interest led to the

completion of this research during her graduate career.

In addition to scientific research, Shannon was able to teach at varying levels during

graduate school. She taught an introductory biology lab as well as participated in the SPICE

program where she co-taught seventh grade science. This experience helped her improve her

communication and presentational skills. Upon completion of her master's degree, Shannon

hopes to continue research on the impacts of human development on the environment in the

hopes of mitigating these effects.









CHAPTER 1
INTRODUCTION

Currently, 15-20% of land area in the U.S. is impacted by roads (Forman and Alexander

1998). Florida is the fourth most populous state and has increased by 9% from 2000 to 2004. In

order to support this growing population, more roads will be built and accompanied by greater

traffic volume throughout the state. Thus, it is important to understand the impacts of roads on

ecosystems in order to design and manage both current and projected roads appropriately for the

benefit of humans and the environment.

Effects of Urbanization on Stream Ecosystems

Urbanization increases impervious surfaces causing in lowered infiltration and

evapotransporation rates and increased runoff and peak discharge of water following storm

events (Bilby et al. 1989; McBride and Booth 2005). In addition, urbanization often causes

thinning, if not complete removal, of riparian buffers around stream ecosystems that act as a

filter for nutrients and sediment prior to entering a stream (Alberti et al. 2007). The loss of

riparian vegetation leads to a more open canopy, which increases temperature and creates

unsuitable habitat for some wildlife such as frogs and salamanders (Wilson and Dorcas 2003).

Riparian vegetation also provides allochthonous leaf litter and woody debris that drive food web

dynamics, of which the latter helps protect the streambed from erosive forces (Booth et al. 1997).

Alteration of the riparian buffer can have severe impacts on stream ecosystems, causing shifts in

structure and function.

Modification of the landscape associated with urban development includes: vegetation

clearing, soil compaction, ditching and draining, and creation of impervious surfaces, all of

which alter runoff Urban landscapes have lower water storage capacity, resulting in runoff

dominated by overland flow (McBride and Booth 2005). This results in reduced ability of the









DeWiest, D. R. and E. H. Livingston. 2002. Florida Erosion and Sediment Control Inspector's
Manual. Florida Department of Environmental Protection, Stormwater/ Nonpoint Source
Management, in cooperation with Florida Department of Transportation.

Estebe, A., H. Boudries, J. M. Mouchel, and D. R. Thevenot. 1997. Urban runoff impacts on
particulate metal and hydrocarbon concentration in River Seine: suspended solid and
sediment transport. Water Science and Technology 36:185-193.

Forman, R. T. T. and L. E. Alexander. 1998. Roads and their major ecological effects. Annual
Review of Ecology and Systematics 29:207-231.

Forman, R. T. T. and R. D. Deblinger. 2000. The ecological road-effect zone of a Massachusetts
(U.S.A.) suburban highway. Conservation Biology 14:36-46.

Gidding, E. M., M. I. Homberger, and H. K. Hadley. 2001. Trace-metal concentrations in
sediment and water and health of aquatic macroinvertebrate communities of streams near
Park City, Summit County, Utah. U.S. Geological Survey Water-Resource Investigation
Report 01-4213.

Gordon, N. D., T. A. McMahon, B. L. Finlayson, C. J. Gippel, R. J. Nathan. 2005. Stream
Hydrology: an introduction for ecologists second edition. John Wiley and Sons. West
Sussex.

Grant, S. B., N. V. Rekhi, N. R. Pise, and R. L. Reeves. 2003. A review of the contaminants and
toxicity associated with particles in stormwater runoff. Prepared for California
Department of Transportation; CTSW-RT-D3-059.73.15.

Grapentine, L., W. Rochfort, and J. Marsalek. 2004. Benthic responses to wet-weather
discharges in urban streams in southern Ontario. Water Quality Resources Journal of
Canada 39:374-391.

Horner, R. R. and Mar, B. W. 1983. Guide for assessing water quality impacts of highway
operations and maintenance. Transportation Resources Record 948:31-39.

Horowitz, A. J. 1991. A Primer on Sediment Trace-element Chemistry, 2nd ed. Lewis, Chelsea,
MI.

Jones, J. A., F. J. Swanson, B. C. Wemple, and K. U. Snyder. 2000. Effects of Roads on
hydrology, geomorphology, and distribution patches in stream networks. Conservation
Biology 14:76-85.

Klein, R. D. 1979. Urbanization and stream quality impairment. Water Resources Bulletin
15:948-963.

Kristan, W. B. 2003. The role of habitat selection behavior in population dynamics: source-sink
systems and ecological traps. Oikos 103:457-468.









samples were taken at 0, 5, 10, and 15 meters upstream of the crossing, directly after the road

crossing, within the riprap, and at 0, 5, 10 and 15 meters downstream of the riprap.

Geomorphology

Cross-sectional profiles of the stream were created to assess incision and stream widening

downstream of road crossings due to increased erosion. Using a laser level (LaserMark MP5

five beam self-leveling laser), a cross-sectional profile of the stream bed to the high water mark

was mapped. Stakes were hammered into the ground at the high water channel edge at sampling

locations. Profiles were measured perpendicular to stream flow, and depth to stream bed was

measured at 20-cm intervals (Gordon et al. 2005).

Sediment

Sediment samples were taken to determine particle size distribution and organic matter

storage up and downstream of road crossings. At each transect, three replicate sediment samples

were taken to a 5-cm depth with a 5.1-cm internal diameter clear PVC hand corer. Sediment

samples were then transferred to labeled 1-quart zip-lock plastic bags and stored on ice for

transport to the laboratory where they were kept at 4C.

Metals

Sediment samples were taken for metal analysis at three locations across each transect. A

plastic shovel was kept in a 10% HC1 acid bath for 7 days prior to sampling and stored in a

plastic bag with 10% HC1 acid between samplings. A composite sample of approximately 50

mL of wet sediment was collected in acid washed jars. Samples were sent to Corner Stone

Laboratory in Memphis, Tennessee, and analyzed for lead, zinc, copper, chromium, and nickel

via Inductively Coupled Plasma-Mass Spectrometry (ICP-MS). A duplicate sediment sample

was taken to measure moisture content. Samples were transferred to labeled 1 quart zip-lock

plastic bags and stored on ice for transport to the laboratory where they were kept at 4C.









0.0007) and riprap (F1,70= 32.19, P < 0.0001) sites, with summer organic matter (respectively:

X = 1.50; X = 7.30; X = 1.19) lower than winter (respectively: X = 3.20, X = 9.24, X = 2.18)

(Figure 3-17). Seasonal effects on total organic matter were dependent on location within the

stream (Fi, 70 = 5.89, P = 0.02) for concrete apron sites. Upstream, total organic matter was

greater in summer (X = 12.87) than winter (X = 10.24), while downstream, total organic matter

the opposite was true (Figure 3-18).

Differences in fine organic matter between up and downstream approached significance

at non-managed sites (F1, 70 = 3.79, P = 0.06), were significant at concrete apron (F1, 70 = 6.92, P

0.01) sites, and were not significant at riprap (Fi, 70 = 0.12, P = 0.73) sites. Fine organic matter at

non-managed and concrete apron sites was higher upstream (respectively: X = 1.22, X = 3.15)

than downstream (respectively: X = 0.86, X = 1.19) (Figure 3-16). There were seasonal effects

on fine organic matter at non-managed (F1, 70 = 11.44, P = 0.001), concrete apron (F1, 70 = 16.87,

P = 0.001), and riprap (F1,70 = 14.27, P = 0.003) sites, with summer fine organic matter

(respectively: X = 0.58, X = 2.00, X = 0.32) lower than winter (X = 1.50, X = 2.35, X = 0.40)

(Figure 3-17). Seasonal effects on fine organic matter were dependent on location within the

stream (Fi, 70 = 8.07, P = 0.006) for concrete apron sites. For upstream, fine organic matter was

greater in summer (X = 3.42) than winter (X = 2.88); however, downstream fine organic matter

was greater in winter (X = 1.82) than summer (X = 0.57) (Figure 3-19).

There were no significant differences in coarse organic matter between up and

downstream for non-managed sites (Fi,70 = 0.68, P = 0.41); however, there were significant

differences at concrete apron (Fi,70 = 14.52, P = 0.0003) and riprap sites (Fi,70 = 4.87, P = 0.03).

At concrete apron sites, coarse organic matter was higher upstream (X = 8.41) than downstream

(X = 3.79); and at riprap sites, coarse organic matter was lower upstream (X = 1.01) than











4.5

4

3.5

3

2.5

2

1.5

1

0.5

0


Non-managed Concrete Apron
Non-managed Concrete Apron


O Upstream
U Downstream


Riprap


O Upstream
* Downstream


Non-managed


Figure 3-28.


Concrete Apron Riprap


Average metal concentrations relative to roads and management type.
Significant differences at non-managed sites for nickel (T = -2.78, P = 0.01) and
lead (T = -2.13, P = 0.05).


1l


-

-

-

-

-

-


_L









aquatic ecosystems (Forman and Alexander 1998; Ruediger and Ruediger 1999). Increased fine

sediment deposition can interfere with macroinvertebrate and fish respiration, feeding, and

reproduction, and can alter habitat availability (Wood and Armitage 1997).

Habitat Structure

Altered hydrology associated with roads can affect stream habitat directly downstream.

Roads interrupt natural debris flows, and flashy hydrology results in removal of organic matter,

woody debris, and substrate roughness of the stream bed (Jones et al. 2000; McBride and Booth

2005). Sedimentation caused by roads can alter substrate distribution, thus affecting habitat

availability (McBride and Booth 2005).

The floodplain can also be altered through rearrangement of channels, logs, branches,

boulders, and fine-sediment deposits (Forman and Alexander 1998). Road crossings alter stream

migration across its floodplain and affect flow rates, pool-riffle sequences, and habitat structure

(Forman and Alexander 1998). Local scour removes habitat forming debris and eliminates

roughness elements essential for fish and other aquatic organisms (McBride and Booth 2005).

Stream segments downstream of roads are more channelized and have less habitat complexity

than upstream (Avolio 2003). McBride and Booth (2005) found that habitat heterogeneity within

a stream was greater in moderately urban streams compared with highly urbanized or rural

streams.

Metal Contamination

Urbanization increases heavy metal concentrations in streams, especially lead, zinc,

copper, nickel, chromium, and cadmium, which are in high concentrations in road runoff

(Forman and Alexander 1998; Gidding et al. 2001). These toxic metals are associated with tire

and brake lining wear and combustion of lubricating oils (Makepeace et al. 1995). These metals

tend to accumulate in sediment and may not be detectable in water samples (Perdikaki and













1
0.5

Z0 0
-0.5 -*
o -1
--J
-1.5
-2 ..*

-2.5
-3.00 -2.50 -2.00 -1.50 -1.00 -0.50 0.00
Log Fine Sediment
C

3

2.5

2 *

S 1.5
--J

** *p*


0 *

-0.5
-3.00 -2.50 -2.00 -1.50 -1.00 -0.50 0.00
Log Fine Sediment
D

Figure 3-25. Continued









CHAPTER 2
MATERIALS AND METHODS

Site Selection and Description

Site Selection

Potential sites were selected using road network and creek maps for Gainesville obtained

through www.cityofgainesville.org. Selected sites were first order streams at their first road

crossings. Following field evaluation, sites with severe channelization and degradation were

eliminated, as were ephemeral streams. First order streams that were mowed to the channel edge

or had anthropogenic alterations not associated with roads were eliminated. Streams were also

selected based on management type since two BMPs were of interest: concrete aprons and

riprap. Nine road crossings were selected for sampling; three were unmanaged, three had

concrete aprons downstream, and three had riprap downstream.

Site Description

Sampling was conducted in Gainesville, Florida (29.40 N, 82.20 W), in north-central

Florida (Figure 2-1). Streams are primarily sandy bottom; however, in deeply eroded portions,

clay (Hawthorn formation) or even limestone (Ocala formation) are exposed. Climate is

characterized by a long wet summer and short dry winter with an average annual temperature of

20.20C and average annual precipitation of 131.6 cm (Climate-zone.com). Sampling sites were

at the first road crossings in nine headwater streams (Table 2-1).

GIS Watershed Delineation

Watershed analysis was done to assess drainage basin area using ArcGIS 9.1. The

drainage basin was delineated using 5-foot contour lines for Alachua County, obtained from

Florida Geographic Database Library (FGDL). All data layers were projected in

NAD_1983_Stateplane.
















I c0





0 10000 20000 30000 40000
Traffic density (Vehicles/ Day


f ,





0 10000 20000 30000 40000
Traffic density (Vehicles/ Day


Figure 3-27.


0 10000 20000 30000 40000
Traffic density (Vehicleslday)


0 10000 20000 30000 40000
Traffic density (Vehicles/ Day)


Linear regression between traffic densities and log transformed metal
concentrations: a) copper (upstream: R2 0.22, P = 0.20; downstream: R2= 0.40,
P = 0.07) b) chromium (upstream: R2 = 0.37, P = 0.08; downstream: R2 = 0.21, P
=0.21) c) nickel (upstream: R2= 0.44, P = 0.05; downstream: R2= 0.67, P=
0.007) d) lead with outlier (upstream: R2 = 0.60, P = 0.01; downstream: R2
0.66, P = 0.008) e) lead without outlier (downstream: R2 = 0.45, P = 0.05) f) zinc
(upstream: R2= 0.22, P = 0.20; downstream: R2= 0.37, P = 0.08).










5

0
S4
3.5
-- 3
5 2.5
2
o 1.5
W-,
LUni-


0
o
o
0 15








.2 6
- 5
C
S4
c 3
o


L.
o

c 15
C-


10 5

Upstream


0 0


5 10
Downstream


Figure 3-24. Metal concentrations relative to roads: A) copper B) chromium C) nickel D) lead
with outlier E) lead without outlier F) zinc.


2


15


TliI/


I I I I I I L









Power, E. A., and P. M. Chapman. 1992. Assessing sediment quality. Pages 1-18 in G. A.
Burton Jr., editor. Sediment Toxicity Assessment. Lewis Publishers, Inc., Boca Raton,
Florida, USA.

Rhoads, B. L. and R. A. Cahill. 1999. Geomorphological assessment of sediment contamination
in an urban stream system. Applied Geochemistry 14:459-483.

Robson, M., K. Spence, and L. Beech. 2006. Stream quality in a small urbanized catchment.
Science of the Total Environment 357:194-207.

Roberts, C. R. 2002. Riparian tree association and storage, transport, and processing of
particulate organic matter in subtropical streams. Ph.D. Dissertation, University of
Florida.

Ruediger, B. and B. Ruediger. 1999. The effects of highways on trout and salmon rivers and
streams in the Western U.S. Pages 151-160 in Proceedings of the third international
conference on wildlife ecology and transportation. Florida Department of Transportation,
Tallahassee, Florida, USA.

Shelton, L. R. and P. D. Capel. 1994. Guidelines for collecting, processing samples of stream
bed sediment for analysis of trace elements and organic contaminants for the National
Water-Quality Assessment Program. U.S. Geological Survey Open-File Report 94-455,
20.

Smith, S. L, D. D. MacDonald, K. A. Keenleyside, C. G. Ingersoll, and J. Field. 1996. A
preliminary evaluation of sediment quality assessment values for freshwater ecosystems.
Journal of Great Lakes Resources 22:624-638.

Sutherland, R. A. 2000. Bed sediment-associated trace metals in an urban stream, Oahu, Hawaii.
Environmental Geology 39:611-627.

Turer, D., J. B. Maynard, and J. J. Salone. 2001. Heavy metal contamination in soils of urban
highways: comparison between runoff and soil concentrations at Cincinnati, Ohio.
Water, Air, and Soil Pollution 132:293-314.

Van Hassel, J. H, J. J. Ney, and D. L. Garling. 1980. Heavy metals in a stream ecosystem at sites
near highways. Transactions of the American Fisheries Society 109:636-43.

Wang, L., J. Lyons, P. Kanehl, and R. Bannerman. 2001. Impacts of urbanization on stream
habitat and fish across multiple spatial scales. Environmental Management 28:255-266.

Waters, T. F. 1995. Sediment in Streams: sources, biological effects, and control. American
Fisheries Society Monograph 7, Bethesda, Maryland, USA.

Whitney, E., D. B. Means, and A. Rudloe. 2004. Priceless Florida: Natural Ecosystems and
Native Species. Pineapple Press Inc. Sarasota, Florida, USA.









Stream Characterization

Stream slope, base flow velocity, and canopy cover were measured. Using a laser level,

elevations were measured at each transect and slopes were calculated by dividing the change in

elevation by the horizontal distance. Average slopes were calculated for upstream and

downstream sections. Velocity was measured once during winter sampling, with a Marsh-

McBirney Flomate 2000. Canopy cover was estimated above the thalweg at each transect using

a densitometer.

Road Characterization

Roads were characterized to determine what features of roads influence the impact of

road crossings on the stream. Area of the road draining to the stream, traffic volume, and the

mode by which water enters the stream were documented. Gainesville urban traffic counts were

obtained from the Florida Department of Transportation. These are determined in a three year

rotation; therefore, data for some sites are more current than others. The data range was 2003-

2005. Site number seven did not have traffic counts because it is a residential road; therefore, to

estimate an average daily traffic count, vehicles crossing the stream were tallied for two hours

during sampling and multiplied by twelve. The surface area draining from a road to a stream

was calculated using 5-foot contour line, creek and road network data layers in ArcGIS 9.1.

Some roads had storm drains that directed water from the road directly to the stream, while at

other roads water ran off the road into a concrete swale along the road before entering the stream.

Therefore, the means by which water entered the stream was classified.

Laboratory Methods

Sediment Particle Size Distribution

To examine how the road influences sediment particle size distribution, samples were

homogenized, and a sub-sample of approximately 100 g was taken, which was then treated with









The concrete apron appears to minimize erosion of downstream sections. Two concrete apron

sites also had swales that routed water from the road to both the up and downstream sections of

the stream. Swales may also play a role in minimizing erosion in these streams.

Seasonal influence on bank slope was significant only at non-managed sites, with higher

bank slopes in winter. BMPs may be effective at preventing bank erosion following storm

events. The seasonal effect on bank slope was dependent on the location within the stream for

concrete apron sites; however, no clear pattern could be identified and was most likely not

associated with the road crossing.

Sediment Particle Size Distribution

Roads are sources of fine sediment that during storm events are washed into streams and

deposited in areas of low flow (Allan 1995). To measure fine sediment distribution, the percent

of sediment less than 63 [tm was calculated for each benthic sample.

Upstream to Downstream Differences

Repeated measures ANOVA was used to test for differences in percent fine sediments

between up and downstream sections for all streams. Differences approached significance (F1,

210 = 3.58, P = 0.06), with downstream (X = 6.52) being greater than upstream (X = 4.45)

(Figure 3-8). There was no difference between summer and winter sampling periods.

Downstream sections had slightly higher percent fine sediment than upstream. The grain

size distribution of urban stream sediment tends to shift towards smaller particle sizes; however,

in densely urban areas, fine particles are selectively removed during storm events (McBride and

Booth 2005). Sediment size distribution is primarily determined by geology of the drainage area

and current velocity (Allan 1995), but roads also act as sources of fine sediment to streams

(Forman and Alexander 1998). Increased deposition of fine sediments has negative implications









the drainage basin may be better indicators of stream impairment than total impervious surface

area (Bledsoe and Watson 2000; Wang et al. 2001; Alberti et al. 2007).

Effects of Roads on Stream Ecosystems

Roads facilitate urbanization and are often key stressors on aquatic systems. They

promote connectivity of the urban matrix to stream networks and are often points of stormwater

discharge to the stream. Roads influence stream hydrology, which in turn alters stream

geomorphology and habitat availability. Roads are also sources of contaminants such as heavy

metals that are flushed into the stream during storms.

Hydrology

Runoff, storage capacity, hydroperiod, stream velocity and depth, groundwater

infiltration and aquifer recharge are hydrologic factors altered by roads and other impervious

surfaces. Impervious surfaces lead to accelerated peak discharge following storm events and

lower base flows (Wang et al. 2001). Increased runoff often results in downstream flooding

(Bilby et al. 1989; Waters 1995; Grapentine et al. 2004). Hydrologic changes associated with

road runoff can modify stream morphology, habitat and water quality (Forman and Alexander

1998).

Geomorphology

Increased runoff and peak water discharge associated with roads lead to increased stream

erosion, and altered channel bed levels and suspended sediment (Brown 1982; Bilby et al. 1989;

Grapentine et al. 2004). As a result, transport of fine sediment and debris downstream increases

(Bilby et al. 1989). Road geometry and maintenance, as well as soil properties and vegetative

cover adjacent to the road influence sedimentation (Forman and Alexander 1998). The degree of

sedimentation can alter channel morphology and aquatic habitat. Although a natural process,

increased sedimentation caused by urbanization and roads can have detrimental effects on
























-2.00


-1.50


-1.00


Log Fine Sediment


-2.00


-1.50


-1.00


Log Fine Sediment


Figure 3-25. Relationship between metal concentrations and fine sediment A) copper (R2 =
0.15, P = 0.008), B) chromium (R2 = 0.21, P < 0.0001), C) nickel (R2 = 0.18, P
0.0002), D) lead with outlier (R2 = 0.15, P = 0.0008), E) lead without outlier (R2
0.27, P < 0.0001) and F) zinc (R2= 0.26, P < 0.0001)


-0.5

-1
-I
-3


3.00


-2.50


-0.50


0.00


1

0.5

0

-0.5


-1.5 -
-3.00


-2.50


-0.50


0.00


~* ......









potential for adsorption of trace metals. In contaminated areas, trace metal concentrations

increase with decreasing particle size (Rhoads and Cahill 1999). Trace metals, specifically zinc

and copper, also have a high affinity to adsorb to organic matter (Krupadam et. al. 2007). The

relationships between organic matter and metals relative to particle size and metals were

significant; but not strong. A number of factors influenced metal concentrations within the

stream, and organic matter and fine sediment concentrations explain some variability in this

study. Both are correlated due to flow regime (Allan 1995); therefore, independent effects of

particle size and organic matter on metal concentrations cannot be separated.

Proximity to a source of metals such as a road is an important factor; however, "hot

spots" of metal contamination exist due to preferential adsorption of trace metals and

depositional processes associated with these materials (Rhoads and Cahill 1999). Some studies

investigating metal concentrations within streams concentrated their sampling efforts on areas of

deposition (Shelton and Capel 1994; Sutherland 2000). In addition, they only ran chemical

analyses on the portion less than 63jtm. This sampling strategy likely identifies "hot spots" of

metal contamination. These areas are where streams are most vulnerable and likely areas where

remediation may be most effective.

Influence of Traffic Density on Metal Concentrations

Linear regressions were used to test for relationships between average metal

concentrations up and downstream of road crossings and traffic densities. Relationships

downstream were significant for nickel (R2 = 0.67, P = 0.007) and lead with (R2 = 0.66, P =

0.008) and without the outlier (R2 = 0.45, P = 0.05), and approached significance for copper (R2

= 0.40, P = 0.07) and zinc (R2 = 0.37, P = 0.08), but were not significant for chromium (R2

0.21, P = 0.21). Relationships between traffic densities and metal concentrations upstream were













45

40

S35
E

g 30
0

25
C
0
0
C 20
0

S15


Copper Chromium


Figure 3-23.


Nickel Lead (with Lead
outlier) (without
outlier)


OUp
U Down


Zinc


Average metal concentrations upstream and downstream of road crossings for all
streams combined: copper (t = 0.25, P = 0.80), chromium (t = -2.04, P = 0.05),
nickel (t = -1.53, P = 0.13), lead (with outlier) (t = -1.96, P = 0.05), lead (without
outlier) (t = -1.67, P = 0.10), and zinc (t = -1.30, P = 0.20).


n ~-r ~ II II n















O Summer
SWinter


15m 10m 5m Om Om 5m 10m 15m
Upstream Downstream


Average bank slope for summer and winter sampling periods according to
location within the stream relative to the road crossing at sites with concrete
aprons.


Upstream Downstream


Average percent fine sediment up and downstream relative to road crossings (Fi,
210 = 3.58, P = 0.06).


S3
2 2.5
19 2
S1.5
|) 1
o 0.5
S0


Figure 3-7.


Figure 3-8.









analysis to ensure normality. An alpha level of 0.05 was used as a threshold for determining

significant effects.

Geomorphology

Width-to-depth ratios were calculated for each transect by dividing the width of the

channel at bank full by the average channel depth (Gordon et al. 2002). Bank slopes were

calculated for each transect by taking average bank slopes of the right and left side of the stream.

Sediment Particle Size Distribution

Fine sediment content was calculated as the proportion of mass in each sediment sample

less than 63 |tm (Allan 1995).

Organic matter

Fine and coarse organic matter content were calculated using Equation 2-1.

% Organic Matter = (DW60- DW550)/DW60 (2-1)

DW60 = dry weight after drying at 600C; DW550 = ashed weight after drying at 550C.

Total organic matter content was calculated by combining fine and coarse organic matter and

using the same formula.

Metals

Laboratory metal concentrations in the sediment were measured as received (wet weight);

therefore, moisture content was calculated for each sediment sample using Equation 2-2.

% Moisture = ((WW DW)/ WW) x 100 (2-2)

WW = wet weight; DW60 = dry weight after drying at 60C

Moisture content was averaged for each sample, and metal concentrations were calculated on a

dry weight basis for comparison using Equation 2-3.

[metal]d = [metal]w / (1- Proportion Moisture) (2-3)

[metal]d = dry weight metal concentrations; [metal]w = wet weight metal concentrations










45
S40
r 35
0 30
S25
5 20
c 15
oS 10
S5
- 0


Upstream Downstream


2 12

E 10
.2 8

6
4-
O 2
N 0
15 10 5 0

Upstream
F

Figure 3-24. Continued


0 5 10 15Downstream
Downstream









riprap downstream of the crossing, thus reducing metal concentrations downstream and

restricting the zone of contamination. The plunge pool and riprap sections may become highly

contaminated over time and require further management. These areas may also be selectively

utilized by macroinvertebrates due to increased dissolved oxygen and habitat heterogeneity, thus

acting as "ecological traps." Ecological traps occur when habitat selection and suitability are

dissociated (Kristan 2003). Further investigation of invertebrate communities and metal

accumulation would be needed to test this hypothesis.










0)
6
E
S5
0
S4
S3
S2
0 1

S0
z


140
120
E
C 100
o
S80
5 60
C60
0 40
*o 20
- 0


TiT T T


15 10 5 0 0 5 10 15

Upstream Downstream


15 10 5 0 0 5 10 15

Upstream Downstream


Figure 3-24. Continued









LIST OF REFERENCES


Alberti, M., D. Booth, K. Hill, B. Coburn, C. Avolio, S. Coe, and D. Spirandelli. 2007. The
impact of urban patterns on aquatic ecosystems: an empirical analysis in Puget Lowland
sub-basins. Landscape and Urban Planning 80:345-361.

Allan, D. 1995. Stream Ecology: Structure and Function of Running Waters. Chapman and Hall,
London, England.

Avolio, C. M. 2003. The local impacts of road crossings on Puget Lowland creeks. M.S. Thesis,
University of Washington.

Beasley, G. and P. Kneale. 2002. Reviewing the impact of metals and PAHs on
macroinvertebrates in urban watercourses. Progress in Physical Geography 26:236-270.

Beasley, G. and P. Kneale. 2004. Assessment of heavy metal and PAH contamination of urban
streambed sediments on macroinvertebrates. Water, Air, and Soil Pollution: Focus
4:1573-2940.

Bilby R. E., K. Sullivan, and S. H. Duncan. 1989. The generation and fate of road-surface
sediment in forested watersheds in southwestern Washington. Forest Science 35:453-68.

Bledsoe, B. P. and C. C. Watson. 2001. Effects of urbanization on channel instability. Journal of
the American Water Resources Association 37:255-270.

Booth, D. B., D. R. Montgomery, and J. Bethel. 1997. Large woody debris in urban streams of
Pacific northwest. Pages 179-197 in Proceedings of the Conference on Effects of
Watershed Development and Management on Aquatic Ecosystems. American Society of
Civil Engineers, Snowbird, Utah, USA.

Brown, S. A. 1982. Prediction of channel bed grade changes at highway stream crossings.
Transportation Research Record 896:1-7.

Canadian Council of Ministers of the Environment. 2001. Canadian Environmental Quality
Guidelines. Winnipeg, Manitoba.

Chen, M., L. Q. Ma, and W.G. Harris. 1999. Baseline concentrations of 15 trace elements in
Florida surface soils. Journal of Environmental Quality 28:1173-1181.

Climate-zone.com. 2003. Data based on the CIA World Fact Book. zone.com/climate/united-states/florida/gainesville/>

Davis, A. P., M. Shokouhian, and S. Ni. 2001. Loading estimates of lead, copper, cadmium, and
zinc in urban runoff from specific sources. Chemosphere 44:997-1009.


































2008 Shannon E. McMorrow









3 RESULTS AND D ISCU SSION .................................................. ............................... 26

G eom orphology ...........................................................................26
Upstream to Downstream Difference....... .. .......... ......................................... .....26
Zone of Influence ............... ....... ....................................... ............... ...27
Influence of M management Type............................................................. .................27
Sedim ent Particle Size D istribution............................................... ............................. 29
Upstream to Downstream Differences ........................................ ........................ 29
Z one of Influence .............................................................. .................30
Influence of M management Type............................................................. .................. 30
O rg anic M matter Storag e............ ........ ................................................................ .. .... .. .. .. 3 1
Upstream to Downstream Differences ........................................ ........................ 32
Z one of Influence .............................................................. .................33
Influence of M anagem ent Type........................................... ........................... ..... 33
Fine Sediment and Organic Matter Relationship ................................. ................ 37
M etals ....................... ....... ............................... ......... 37
Up to D ownstream Differences ......................................................... .............. 38
Zone of Influence ................ .................. ............ .......... .. ................... .......... 40
Relationships among Metals, Organic Matter, and Particle Size ..................................40
Influence of Traffic Density on Metal Concentrations......................... ...................41

4 C O N C L U SIO N S ................. ......................................... ........ ........ ..... .... ...... .. 74

Im pact of R oad C rossings............................ .............................................. .......................74
Influence of M anagem ent Practices......................... .............. ................ ............... 75

L IST O F R E FE R E N C E S ............................ .................................... ...................... ............... 78

B IO G R A PH IC A L SK E T C H .............................................................................. .....................83























6









Up to Downstream Differences

A t-test was used to test for differences in metal concentrations between up and

downstream locations for all streams combined. There was a significant difference for

chromium (t = -2.04, P = 0.05), with downstream concentrations (X = 2.63) greater than

upstream (X = 1.92); however, there were no significant differences for copper (t = 0.25, P =

0.80), nickel (t = -1.53, P = 0.13), or zinc (t = -1.30, P = 0.20). When the outlier was included in

statistical analyses, there was a significant difference (t = -1.96, P = 0.05) in lead concentrations

between up (X = 7.02) and downstream (X = 29.60); however, when it was removed, the

difference was not significant (t = -1.67, P = 0.10). (Figure 3-23).

Average metal concentrations downstream were higher than upstream for chromium and

lead. Such variability in heavy metal concentrations is unclear. Perdikaki and Mason (1999)

also found no significant difference in lead or zinc concentrations up and downstream of a main

road. Metals from road runoff may be depositing both up and downstream of road crossings.

Alternatively, certain metals may be more mobile than others, Rhoads and Cahill (1999) found

that concentrations of nickel and chromium were lower in the suspended load than bed material,

suggesting that these metals are relatively immobile, while the opposite was true for copper and

zinc.

Establishing baseline levels for heavy metals is difficult considering atmospheric

depositional sources; however, Chen et al. (1999) did so for potentially toxic metals based on

448 Florida surface soils, and the Canadian Council of Ministers of the Environment (2001)

established threshold and probable effect levels for sediments (Table 3-1). Threshold effect

levels (TEL) are concentrations below which adverse effects are rare. Probable effect levels

(PEL) are concentrations above which adverse effects are expected to occur frequently (Smith et
























-0.5 -
-2.50


-2.00


-1.50


-1.00


Log Organic Matter


o -- ** <' #4 *

-0.5
-2.50 -2.00 -1.50 -1.00 -0.50


Log Organic Matter


F

Figure 3-26. Continued


* ** *


-0.50


0.00


0.00











Riprap


0 6

E 5





vv 2

2 1

(. 0
15 10 5 0

Upstream

Figure 3-10. Continued



E 25

0 20
0
V
15

E 10

o 05 5,

L. 0
O i n n


0 5 10 15

Downstream


- -A- winter
+-summer


Figure 3-11. Average percent fine sediment relative to the road crossing seasonally at non-
managed sites.









downstream (X = 1.64) (Figure 3-16). Seasonal differences approached significance at non-

managed sites (Fi, 70 = 2.85, P = 0.10) and were significant at concrete apron (Fi, 70 = 12.26, P =

0.0008) and riprap (Fi,70= 21.18, P < 0.0001) sites, with summer coarse organic matter

(respectively: X = 0.92, X = 5.31, X = 0.87) lower than winter (respectively: X = 1.70,X =

6.89, X = 1.78) (Figure 3-17). Seasonal effects on coarse organic matter were dependent on

location within the stream (Fi, 70 = 6.09, P= 0.01) for concrete apron sites. For upstream, coarse

organic matter was greater in summer (X = 9.45) than winter (X = 7.36); however, for

downstream fine organic matter was greater in winter (X = 6.41) than summer (X = 1.17)

(Figure 3-20).

Repeated measures ANOVA was used to test for differences among transects at varying

distances (0, 5, 10, and 15 meters up and downstream) from the road crossing by management

type. There were no significant differences among transects at non-managed sites (F7,64 = 1.10,

P = 0.37); however, there were significant differences among transects at concrete apron (F7,64=

2.42, P = 0.03) and riprap (F2,64= 2.67, P = 0.02) sites. Tukey post-hoc comparisons were used

to test for pairwise differences among transects. At concrete apron sites, total organic matter 10

meters upstream was significantly greater than at 0 and 5 meters downstream, and 0 meters

upstream was significantly greater than at 0 meters downstream. At riprap sites, total organic

matter at 15 meters upstream was significantly less than at 10 meters upstream and 10 meters

downstream (Figure 3-21). There were no significant differences among transects for fine

organic matter at non-managed, concrete apron, or riprap sites (Figure 3-21).

There were no significant differences among transects for coarse organic matter at non-

managed sites (F7,64= 0.50, P = 0.83); however, there were significant differences at concrete

apron (F7, 64 = 4.60, P = 0.0003) and riprap (F7, 64 = 3.17, P = 0.006) sites. At concrete apron









Krupadam, R. J., R. Ahuja, and S. R. Wate. 2007. Heavy metal binding fractions in the sediment
of the Godavari Estuary, east coast of India. Environmental Modeling and Assessment
12:145-155.

Lee, D. K., J. C. Touray, P. Bailif, and J. P. Ildeonse. 1997. Heavy metal contamination of
settling particles in a retention pond along the A71 motorway in Sologne, France.
Science of the Total Environment 201:1-15.

Madej, M. A. 2001. Erosion and sediment delivery following removal of forest roads. Earth
Surface Processes and Landforms 26:175-190.

Makepeace, D. K., Smith, D. W. and Stanley, S. J. 1995. Urban stormwater quality: summary of
contaminant data. Critical Reviews in Environmental Science and Technology 25:93-
139.

Marsalek, J., Q. Rochfort, B. Brownlee, T. Mayer, and M. Servos. 1999. An explanatory study of
urban runoff toxicity. Water Science and Technology 39:33-39.

May, C. W., R. R. Homer, J. R. Karr, B. W. Mar, and E. B. Welch. 1997. Effects of urbanization
on small streams in the Puget Sound Lowland ecoregion. Watershed Protection
Techniques 2:483-494.

Mcbride, M. 2001. Spatial effects of urbanization on physical conditions in Puget Sound
Lowland streams. Water Resources Series, Technical Report No. 177, Department of
Civil and Environmental Engineering, University of Washington, Seattle, Washington.

McBride, M. and D. B. Booth. 2005. Urban impacts on physical stream condition: effects of
spatial scale connectivity and longitudinal trends. Journal of the American Water
Resources Association 41:565- 580.

Montgomery, D. 1994. Road surface drainage, channel initiation, and slope instability. Water
Resources Research 30:192-93.

Paul, M. J. and J. L. Meyer. 2001. Streams in the urban landscape. Annual Review of Ecology
and Systematics 32:333-365.

Perdikaki, K. and C. F. Mason. 1999. Impacts of road run-off on receiving streams in Eastern
England. Pergamon 33:1627-1633.

Plumb, R. H. Jr. 1981. Procedures for handling and chemical analysis of sediment and water
samples. Contract EPA-4805572010. U.S. Environmental Protection Agency/Corps of
Engineers Technical Committee on Criteria for Dredged and Fill Material, Washington,
D.C.




























-2.50


*
rr *
~ I, ~~~
* **
** *
*
+ +
*


-2.00


-1.50


-1.00


-0.50


0.00


Log Organic Matter


1

0.5

0

-0.5


-1.5
-2.50


-2.00


-1.50


-1.00


-0.50


0.00


Log Organic Matter
B

Figure 3-26. Relationship between metal concentrations and organic matter: A) copper (R2
0.14, P = 0.001), B) chromium (R2= 0.13, P = 0.002), C) nickel (R2= 0.09, P
0.01), D) lead (R2= 0.01, P = 0.33), E) lead without outlier (R2= 0.05, P = 0.09),
and F) zinc (R2= 0.19, P = 0.0001).


t o

. #









3-17 Average percent organic matter according to sampling period and management type .....54

3-18 Average total organic matter by sampling period and location relative to the road
crossing at concrete apron sites............................................................... .....................55

3-19 Average fine organic matter according to season and location relative to the road
crossing at concrete apron sites............................................................... .....................55

3-20 Average coarse organic matter according to season and location relative to the road
crossing at concrete apron sites............................................................... .....................56

3-21 Average percent organic matter according to location relative to road crossings and
m anagem ent type. ..................................... .. ......... .............. .. 56

3-22 Linear relationship between log proportion organic matter and log proportion fine
sedim ent. ................... ......... .............. .............. ............. .......... .................. 58

3-23 Average metal concentrations upstream and downstream of road crossings for all
stream s com b in ed .................................................... ................. 59

3-24 M etal concentrations relative to roads ...................................................... .............. 60

3-25 Relationship between metal concentrations and fine sediment ......................................63

3-26 Relationship between metal concentrations and organic matter.....................................66

3-27 Linear regression between traffic densities and log transformed metal concentrations. ...69

3-28 Average metal concentrations relative to roads and management type.............................71




Full Text

PAGE 1

1 IMPACTS OF ROAD CROSSINGS ON HEADWATER STREAMS By SHANNON E. MCMORROW A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2008

PAGE 2

2 2008 Shannon E. McMorrow

PAGE 3

3 To my family and friends without whos e support I could not have done it.

PAGE 4

4 ACKNOWLEDGMENTS I thank m y family and friends for their unendi ng love and support. I thank specifically Marcus Griswold for his work and guidance on this project. I tha nk the volunteers who accompanied in the field to collect data. I thank my parents for loving and supporting me throughout my life and encouraging me to follow my dreams. I thank my major advisor Dr. Thomas L. Crisman for guidance and support thr oughout my graduate experience. I thank Dr. Mark Brenner and Dr. William Wise for their advice. I thank the Florida Stormwater Association for awarding me their annual scho larship, which allowed me to evaluate metal concentrations. And I thank Sean for be ing my rock and loving me no matter what.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........7 LIST OF FIGURES.........................................................................................................................8 ABSTRACT...................................................................................................................................10 CHAP TER 1 INTRODUCTION..................................................................................................................11 Effects of Urbanization on Stream Ecosystems...................................................................... 11 Effects of Roads on Stream Ecosystems................................................................................ 13 Hydrology........................................................................................................................13 Geomorphology............................................................................................................... 13 Habitat Structure.............................................................................................................. 14 Metal Contamination.......................................................................................................14 Study Objectives.....................................................................................................................16 2 MATERIALS AND METHODS........................................................................................... 18 Site Selection and Description................................................................................................18 Site Selection...................................................................................................................18 Site Description...............................................................................................................18 GIS Watershed Delineation............................................................................................. 18 Field Methods.........................................................................................................................20 Geomorphology............................................................................................................... 21 Sediment..........................................................................................................................21 Metals..............................................................................................................................21 Stream Characterization.................................................................................................. 22 Road Characterization..................................................................................................... 22 Laboratory Methods............................................................................................................. ...22 Sediment Particle Size Distribution................................................................................. 22 Organic Matter.................................................................................................................23 Metals..............................................................................................................................23 Data Analysis..........................................................................................................................23 Geomorphology............................................................................................................... 24 Sediment Particle Size Distribution................................................................................. 24 Organic matter................................................................................................................. 24 Statistical Analyses.......................................................................................................... 25

PAGE 6

6 3 RESULTS AND DISCUSSION............................................................................................. 26 Geomorphology.................................................................................................................. ....26 Upstream to Downstream Difference.............................................................................. 26 Zone of Influence............................................................................................................27 Influence of Management Type....................................................................................... 27 Sediment Particle Size Distribution........................................................................................ 29 Upstream to Downstream Differences............................................................................ 29 Zone of Influence............................................................................................................30 Influence of Management Type....................................................................................... 30 Organic Matter Storage...........................................................................................................31 Upstream to Downstream Differences............................................................................ 32 Zone of Influence............................................................................................................33 Influence of Management Type....................................................................................... 33 Fine Sediment and Orga nic Matter Relationship ............................................................ 37 Metals.....................................................................................................................................37 Up to Downstream Differences....................................................................................... 38 Zone of Influence............................................................................................................40 Relationships among Metals, Organi c Matter, and Particle Size .................................... 40 Influence of Traffic Density on Metal Concentrations....................................................41 4 CONCLUSIONS.................................................................................................................... 74 Impact of Road Crossings....................................................................................................... 74 Influence of Management Practices........................................................................................75 LIST OF REFERENCES...............................................................................................................78 BIOGRAPHICAL SKETCH.........................................................................................................83

PAGE 7

7 LIST OF TABLES Table page 2-1. Study Sites in Gainesville, Florida..................................................................................... 20 3-1. Baseline, threshold, and probable effect levels ................................................................. 39 3-2. Primary sources of heavy metals in road runoff...............................................................42

PAGE 8

8 LIST OF FIGURES Figure page 2-1 Map of Florida showing the location of Gainesville with sampled sites highlighted. Map created in ArcGIS 9.1................................................................................................19 3-1 Average width-to-depth ratio for up and downstream sections of study streams ............. 45 3-2 Average bank slope up and downstream relative to road crossings ................................. 45 3-3 Average bank slope for summ er and winter sampling periods...................................... 46 3-4 Average width-to-depth ratio up and downs tream relative to road crossings by m anagement type...............................................................................................................46 3-5 Average ban k slope for up and downstream sections of streams according to management type...............................................................................................................47 3-6 Average ban k slope for summer and winter sampling periods according to management type...............................................................................................................47 3-7 Average ban k slope for summer and wint er sampling periods according to location within the stream relative to the road crossing at site s with concrete aprons.................... 48 3-8 Average percent fine sedim ent up and dow nstream relative to road crossings................. 48 3-9 Average percent fine sedim ent (<0.063 mm ) up and downstream of road crossings according to mana gement type.......................................................................................... 49 3-10 Average percent fine sedim ent (<0.063 mm) relative to road crossings by management type...............................................................................................................50 3-11 Average percent fine sed iment relative to the road crossing seasonally at nonmanaged sites.................................................................................................................. ...51 3-12 Average percent fine and coarse organic m atter in study streams upstream and downstream of road crossing.............................................................................................52 3-13 Average percent fine and coarse organic m atter by sampling period................................ 52 3-14 Average percent fine org anic matter accordi ng to position relative to road crossings and sampling period........................................................................................................... 53 3-15 Average coarse organ ic matter according to lo cation in the stream relative to the road crossing...................................................................................................................... ....52 3-16 Average percent org anic matter according to location relative to the road crossing by management type...............................................................................................................54

PAGE 9

9 3-17 Average percent organic m atter according to sampling period and management type..... 54 3-18 Average to tal organic matter by sampling period and location relative to the road crossing at concrete apron sites.......................................................................................... 55 3-19 Average fin e organic matter according to s eason and location re lative to the road crossing at concrete apron sites.......................................................................................... 55 3-20 Average coarse organ ic matter according to season and location relative to the road crossing at concrete apron sites.......................................................................................... 56 3-21 Average percent org anic matter according to location relative to road crossings and management type...............................................................................................................56 3-22 Linear relationship betw een log proportion organic matter and log proportion fine sediment....................................................................................................................... ......58 3-23 Average m etal concentrations upstream and downstream of road crossings for all streams combined............................................................................................................... 59 3-24 Metal concentrations relative to roads ............................................................................... 60 3-25 Relationship between m etal concen trations and fine sediment......................................... 63 3-26 Relationship between m etal concen trations and organic matter........................................ 66 3-27 Linear regression between tra ffic densities and log transfor m ed metal concentrations.... 69 3-28 Average m etal concentrations relati ve to roads and management type............................. 71

PAGE 10

10 Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science IMPACTS OF ROAD CROSSINGS ON HEADWATER STREAMS By Shannon E. McMorrow May 2008 Chair: Thomas L. Crisman Major: Environmental Engineering Sciences Urban development and associated roads a dversely affect stream ecosystems through altered hydrology and subsequent erosion and contam ination. Florida population growth leads to urbanization of natural landscapes and associated road development. K nowledge of the impacts of road crossings on stream systems will help guide better ro ad design. Impacts of road crossings on stream geomorphology, sediment part icle size distribution, or ganic matter storage, and metal contamination were evaluated. Overal l, areas downstream of road crossings were characterized by narrower channels greater bank slopes, increased fine sediment, and higher metal concentrations. Concrete aprons and riprap are structural be st management practices (BMPs) designed to dissipate energy to prevent scour and erosion. The effectivene ss of these BMPs was evaluated by comparing impacts of crossings at managed and non-managed sites. Concrete aprons and swales were effective at dissipating erosive en ergy and minimizing scour; however, they were ineffective at minimizing metal and fine sedime nt contamination. Riprap was ineffective at minimizing scour, but was effective at trapping orga nic matter, fine sediments, and metals within the riprap and plunge pool. A more comprehensive management technique is needed to mitigate all the adverse affects of road crossings on stream ecosystems.

PAGE 11

11 CHAPTER 1 INTRODUCTION Currently, 15-20% of land area in the U.S. is impacted by ro ads (Forman and Alexander 1998). Florida is the fourth most populous stat e and has increased by 9% from 2000 to 2004. In order to support this growing population, more roads will be built and accompanied by greater traffic volume throughout the state. Thus, it is important to understand the impacts of roads on ecosystems in order to design and manage both cu rrent and projected road s appropriately for the benefit of humans and the environment. Effects of Urbanization on Stream Ecosystems Urbanizatio n increases impervious surfaces causing in lowered infiltration and evapotransporation rates and increased runoff and peak discharge of water following storm events (Bilby et al. 1989; McBr ide and Booth 2005). In addition, urbanization often causes thinning, if not complete removal, of riparian buffers around stream eco systems that act as a filter for nutrients and sediment prior to enteri ng a stream (Alberti et al. 2007). The loss of riparian vegetation leads to a more open ca nopy, which increases temperature and creates unsuitable habitat for some wild life such as frogs and salamanders (Wilson and Dorcas 2003). Riparian vegetation also provide s allochthonous leaf litter and w oody debris that drive food web dynamics, of which the latter helps protect the streambed from erosiv e forces (Booth et al. 1997). Alteration of the riparian buffer can have severe impacts on stream ecosystems, causing shifts in structure and function. Modification of the landscape associated w ith urban development includes: vegetation clearing, soil compaction, ditchi ng and draining, and creation of impervious surfaces, all of which alter runoff. Urban landscapes have lo wer water storage capacity, resulting in runoff dominated by overland flow (McBri de and Booth 2005). This resu lts in reduced ability of the

PAGE 12

12 watershed to filter nutrients, sediment, and other contaminants before entering the stream. Total impervious surface area within a dr ainage basin has often been us ed as an indicator of urban stream impairment (Klein 1979; Paul and Meyer 2001; Robson and Beech 2006). Due to increased impervious surface area, th ere is a shift in stream hydrologic regime indicated by an increase in peak discharge during storm events, commonly referred to as flashy hydrology. This leads to increased erosion and suspended sediment, resulting in movement of fine sediments and organic matter downstream, c oupled with increased turbidity (Bilby et al. 1989). Grain size distribution in urban streams te nds to shift toward smaller particles; however in highly urbanized streams, smaller grain sizes are selectively removed during storm events (McBride and Booth 2005). The degree of sedimentation can alter channel morphology and aquatic habitat (Forman and Alexander 1998) McBride and Booth ( 2005) found that urban streams have greater cross-se ctional dimensions, increased erosion rates, and simplified morphology compared to non-urban streams. Urbanization can also increase heavy metal conc entrations in streams, especially lead, nickel, zinc, copper, chromium, and cadmium that are concentrated in road runoff (Forman and Alexander 1998; Gidding et al. 2001). Such he avy metals are detrimental to stream biota and can be incorporated into food webs leading to bioaccumulation and biomagnification in longerlived species (Beasle y and Kneale 2002). Urbanization within a drainage basin can also have dramatic impacts on stream ecosystem structure and functi on. Not only the degree of urba nization (Klein 1979; Paul and Meyer 2001; Robson and Beech 2006) but the conf iguration may have differential impacts on streams. Road density, number of crossings, an d connectivity of impervi ous surface areas within

PAGE 13

13 the drainage basin may be better indicators of st ream impairment than total impervious surface area (Bledsoe and Watson 2000; Wang et al. 2001; Alberti et al. 2007). Effects of Roads on Stream Ecosystems Roads facilitate urbanization and are often key stressors on aquatic system s. They promote connectivity of the urban matrix to stre am networks and are often points of stormwater discharge to the stream. Roads influence st ream hydrology, which in turn alters stream geomorphology and habitat availability. Roads are also sources of contaminants such as heavy metals that are flushed into the stream during storms. Hydrology Runoff, storage capacity, hydroperiod, stream velocity and depth, groundwater infiltration and aquifer recharge are hydrologic factors altered by roads and other impervious surfaces. Impervious surfaces lead to accelerated peak discha rge following storm events and lower base flows (Wang et al. 2001). Increa sed runoff often results in downstream flooding (Bilby et al. 1989; Waters 1995; Grapentine et al. 2004). Hydrol ogic changes associated with road runoff can modify stream morphology, habi tat and water quality (Forman and Alexander 1998). Geomorphology Increased runoff and peak water discharge associ ated with ro ads lead to increased stream erosion, and altered channel bed levels and suspended sediment (Brown 1982; Bilby et al. 1989; Grapentine et al. 2004). As a re sult, transport of fine sediment and debris downstream increases (Bilby et al. 1989). Road geometry and maintenan ce, as well as soil properties and vegetative cover adjacent to the road influence sedimentation (Forman and Alexander 1998). The degree of sedimentation can alter channel morphology and a quatic habitat. Although a natural process, increased sedimentation caused by urbanization and roads can have detrimental effects on

PAGE 14

14 aquatic ecosystems (Forman and Alexander 1998; Ru ediger and Ruediger 1999). Increased fine sediment deposition can interfere with macroinve rtebrate and fish re spiration, feeding, and reproduction, and can alter habitat av ailability (Wood and Armitage 1997). Habitat Structure Altered hydrology associated with roads can affect stream ha bitat directly downstream. Roads interrupt natural debris flows, and flashy hydrology results in removal of organic matter, woody debris, and substrate roughness of the str eam bed (Jones et al. 2000; McBride and Booth 2005). Sedimentation caused by roads can alter subs trate distribution, thus affecting habitat availability (McBride and Booth 2005). The floodplain can also be al tered through rearrangement of channels, logs, branches, boulders, and fine-sediment deposits (Forman and Alexander 1998). Road crossings alter stream migration across its floodplain and affect flow rates, pool-riffle sequences, and habitat structure (Forman and Alexander 1998). Lo cal scour removes habitat forming debris and eliminates roughness elements essential for fish and other aquatic organisms (McBride and Booth 2005). Stream segments downstream of roads are more channelized and have less habitat complexity than upstream (Avolio 2003). McBride and Booth (2005) found that habitat heterogeneity within a stream was greater in modera tely urban streams compared w ith highly urbanized or rural streams. Metal Contamination Urbanization increases heavy m etal concentrat ions in streams, especially lead, zinc, copper, nickel, chromium, and cadmium, which ar e in high concentrations in road runoff (Forman and Alexander 1998; Gidding et al. 2001). These toxic meta ls are associated with tire and brake lining wear and combustion of lubrica ting oils (Makepeace et al. 1995). These metals tend to accumulate in sediment and may not be detectable in water samples (Perdikaki and

PAGE 15

15 Mason 1999; Beasley and Kneal e 2002). Accumulation in sediments can lead to bioaccumulation in macroinvertebrates and bi omagnification through food webs (Power and Chapman 1992; Beasley and Kneale 2002). Concentr ations of heavy metals in sediments, benthic macroinvertebrates, and fish of str eams receiving road runoff are positively correlated with traffic densities (Van Hassel et al. 1980; Marsalek et al. 1999). Best Management Practices BMPs are control practices us ed for a given set of conditi ons to achieve satisfactory water quality and quantity enhancement at a mi nimal cost (DeWiest and Livingston 2002). There are a number of structur al and non-structural BMPs fo r minimizing effects of peak discharge in streams during storm events. A broad-scale approach in Puget Sound found that BMPs within a watershed had a very weak and even negative correlation with biological indices. However, a more intense study indicated that st ructural BMPs help sustain aquatic biological communities in highly urbanized areas. At modera te urbanization levels, there was less evidence of benefit (Avolio 2003). Horner and Mar (1983) found that pa ssage of runoff through vegetated channels greatly reduced concentrations of toxic solids and heavy metals. In addition, vegetated buffer strips between roads and streams act as filters (Forman and Alexander 1998). There are a number of BMPs utilized in Gainesville, FL to minimize erosive impacts of stormwater runoff on streams. Swales are wa terways intended to convey stormwater with minimal erosion. Concrete aprons are used to di ssipate energy at outlets of pipes to prevent scour and minimize erosion caused by stormwater. Riprap is an erosio n-resistant ground cover of large, loose, angular stones used both to protect the soil surface from erosion by runoff and to slow water velocity, while enhancing the potential fo r infiltration. Finally, ri prap is also used to stabilize slopes with seepage pr oblems (DeWiest and Livingston 2002).

PAGE 16

16 Study Objectives The shift in hydrologic regim e associated wi th urbanization is magnified both by roads and connectivity of impervious surfaces to str eams (Alberti et al. 2007). Although studies on the effects of urbanization are common (reviewed in Paul and Meyer 2001), few have examined the impact of urban road crossings on stream ecosystems. The literature is lacking in four major areas: 1) Impacts of urban roads broadly, 2) Impacts of urban roads at the local scale, 3) Impacts of urban roads in Florida, and 4) In fluence of best management practices on such impacts. The majority of the research has examined the impacts of road crossings on streams in forested landscapes (e.g. Montgomery 1994; Made j 2001). The impacts of roads in an urban context are much different than in forested land scapes, and the majority of literature is more closely related to the overall imp act of urbanization and is not spec ific to roads (M ay et al. 1997; Konrad 2000; McBride 2001). The majority of the limited research on the im pact of urban roads on streams has been conducted in the temperate zone (Van Hassel et al. 1980; Marsalek 1990; Perdikaki and Mason 1999; Forman and Deblinge r 2000; Gidding et al. 2001; Grapetine et al. 2004), with streams very differe nt from those of Florida. Florida has unique topography, geology, and climate, and its stream s display sandy bottoms and low flow (Whitney et al. 2004). Finally, impacts of roads on streams have not addressed effectiveness of best management practices. Efforts have obviously been made to mitigate the impact of road crossings on streams through BMPs; however, research on their effectiveness has been neglected. This study investigated the di rect influence of road cro ssings on a stream ecosystem by measuring channel morphology, particle size, or ganic matter storage, and metal accumulation. The effectiveness of best manage ment practices downstream of road crossings in relation to these variables was also considered.

PAGE 17

17 Research question 1. How is sediment size distribution within small streams affected by road crossings? Research question 2. How is geomorphology of small streams altered by road crossings? Research question 3. How is organic matter storage of small streams affected by road crossings? Research question 4. How are sediment metal concentrations within small streams affected by road crossings? Research question 5. How do current BMPs influence imp acts of road crossings on small streams, specifically in terms of stream morphology, sediment size distribution, organic matter storage and sediment metal concentrations?

PAGE 18

18 CHAPTER 2 MATERIALS AND METHODS Site Selection and Description Site Selection Potential sites were selected using road netw ork and creek m aps for Gainesville obtained through www.cityofgainesville.org Selected sites were first or der s treams at their first road crossings. Following field evaluation, sites with severe channelization and degradation were eliminated, as were ephemeral streams. First order streams that were mowed to the channel edge or had anthropogenic alterations not associated with roads were el iminated. Streams were also selected based on management type since two BM Ps were of interest: concrete aprons and riprap. Nine road crossings were selected for sampling; three were unmanaged, three had concrete aprons downstream, and three had riprap downstream. Site Description Sa mpling was conducted in Gainesville, Fl orida (29.40 N, 82.20 W), in north-central Florida (Figure 2-1). Streams are primarily sa ndy bottom; however, in deeply eroded portions, clay (Hawthorn formation) or even limestone (Ocala formation) are exposed. Climate is characterized by a long wet summer and short dry winter with an average annual temperature of 20.2C and average annual precipita tion of 131.6 cm (Climate-zone.com). Sampling sites were at the first road crossings in nine headwater streams (Table 2-1). GIS Watershed Delineation Watershed analysis was done to assess dr ainage basin area using ArcGIS 9.1. The drainage basin was delineated using 5-foot contour lines fo r Alachua County, obtained from Florida Geographic Database Library (FGDL). All data layers were projected in NAD_1983_Stateplane.

PAGE 19

19 Figure 2-1. Map of Florida showing the lo cation of Gainesville with sampled sites highlighted. Map created in ArcGIS 9.1.

PAGE 20

20 Table 2-1. Study Sites in Ga inesville, Florida. Site Headwater type Management Traffic Volume (Vehicles/day) Position relative to road Drainage area (m2) Stream Slope Base Velocity (m/s) 1 Seep Nonmanaged 4,147 Up 29,444 0.035 0.03 Down 32,412 0.031 0.04 2 Seep Nonmanaged 2,107 Up 44,988 0.059 0.04 Down 58,116 0.025 0.05 3 Seep Concrete Apron 16,004 Up 36,835 0.020 0.05 Down 44,986 0.045 0.05 4 Seep Riprap 31,000 Up 177,209 0.005 0.11 Down 206,563 0.041 0.06 5 Wetland Riprap 9,347 Up 127,354 0.005 0.11 Down 143,438 0.010 0.06 6 Wetland Concrete Apron 14,920 Up 133,158 0.003 0.05 Down 143,379 0.009 0.13 7 Seep Riprap 300 Up 60,682 0.015 0.06 Down 77,747 0.015 0.04 8 Wetland Concrete Apron 17,200 Up 1,054,653 0.003 0.02 Down 1,275,809 0.009 0.02 9 Seep Nonmanaged 16,004 Up 13,049 0.019 0.07 Down 21,056 0.006 0.05 Field Methods Sites were sam pled during two periods: Su mmer (18-26 July, 2006) and winter (19-24 February 2007). During both periods sediment cores were taken and geomorphologic profiles were measured. In addition, during the winter sampling period, streams and roads were characterized, and sediment samples were collected for metal analysis. At non-managed sites, sediment sampling and geomorphologic measurements were conducted at 0, 5, 10, and 15 meters upstream and downstream of the road crossing. For sites with concrete aprons, samples were taken at 0, 5, 10, and 15 meters upstream of the road crossing and 0, 5, 10, and 15 meters downstream of the concrete apron. For sites with riprap,

PAGE 21

21 samples were taken at 0, 5, 10, and 15 meters upstr eam of the crossing, directly after the road crossing, within the riprap, and at 0, 5, 10 and 15 meters downstream of the riprap. Geomorphology Cross-sectional profiles of the stream were created to assess incision and stream widening downstream of road crossings due to increased erosion. Using a laser level (LaserMark MP5 five beam self-leveling laser), a cross-sectional profile of the stream bed to the high water mark was mapped. Stakes were hammered into the gr ound at the high water channel edge at sampling locations. Profiles were measured perpendicular to stream flow, and depth to stream bed was measured at 20-cm intervals (Gordon et al. 2005). Sediment Sedim ent samples were taken to determine particle size distribution and organic matter storage up and downstream of road crossings. At each transect, three replicate sediment samples were taken to a 5-cm depth with a 5.1-cm inte rnal diameter clear PVC hand corer. Sediment samples were then transferred to labeled 1-quart zip-lock plastic bags and stored on ice for transport to the laboratory wh ere they were kept at 4 C. Metals Sedim ent samples were taken for metal analysis at three locations across each transect. A plastic shovel was kept in a 10% HCl acid bath for 7 days prior to sampling and stored in a plastic bag with 10% HCl acid between samplings A composite sample of approximately 50 mL of wet sediment was collected in acid washed jars. Samples were sent to Corner Stone Laboratory in Memphis, Tennessee, and analyzed for lead, zinc, copper, chromium, and nickel via Inductively Coupled Plasma-Mass Spectrometr y (ICP-MS). A duplicate sediment sample was taken to measure moisture content. Sample s were transferred to labeled 1 quart zip-lock plastic bags and stored on i ce for transport to the laborator y where they were kept at 4 C.

PAGE 22

22 Stream Characterization Stream slope, base flow velocity, and canopy co ver were measured. Using a laser level, elevations were measured at each transect and slopes were calculated by dividing the change in elevation by the horizontal distance. Averag e slopes were calculated for upstream and downstream sections. Velocity was measur ed once during winter sampling, with a MarshMcBirney Flomate 2000 Canopy cover was estimated above th e thalweg at each transect using a densitometer. Road Characterization Roads were characterized to determ ine what features of roads infl uence the impact of road crossings on the stream. Area of the road draining to the stream, traffic volume, and the mode by which water enters the stream were doc umented. Gainesville urban traffic counts were obtained from the Florida Department of Transpor tation. These are determined in a three year rotation; therefore, data for so me sites are more current than others. The data range was 20032005. Site number seven did not have traffic counts because it is a residential road; therefore, to estimate an average daily traffic count, vehicles crossing the stream were tallied for two hours during sampling and multiplied by twelve. The surface area draining from a road to a stream was calculated using 5-foot contour line, creek and road network data layers in ArcGIS 9.1. Some roads had storm drains that directed water from the road directly to the stream, while at other roads water ran off the road into a concrete swale along the ro ad before entering the stream. Therefore, the means by which water entered the stream was classified. Laboratory Methods Sediment Particle Size Distribution To exam ine how the road influences sedime nt particle size dist ribution, samples were homogenized, and a sub-sample of approximately 100 g was taken, which was then treated with

PAGE 23

23 aqueous hexametaphosphate to disaggregate clay part icles (10 mL / 25 g sample). It was then passed through a 63m sieve to separate silts and clays fr om coarser sediment Both retained fractions were dried between 60-70C for 48 hours and ashed in a muffle furnace at 550 C for 5 hours. Upon drying, samples were weighed to 0.001 g. Sediment was then passed through a nest of 2 sieves (2-mm and 63m), and each fraction was weighed to 0.001 g (Plumb 1981). Organic Matter To determ ine how roads influence organic matte r storage within the sediments, sediment samples were homogenized and a 5-cc sub-sample of the sediment was taken to calculate organic matter content. It was wet sieved through 1 mm and 63m sieves to separate coarse from fine organic matter. Each fraction was dried in an oven at 60-70 C until dry (48 hours), weighed to 0.001 g and ashed in a muffle furnace at 550 C for 5 hours. Upon drying, the samples were placed in a decondenser and reweighed. Metals Meta l samples sent to the laboratory were analyzed in wet material, therefore, moisture content of the duplicate samples was measured a nd metal concentrations were converted to dry weight. Each sediment sample was homogenized and three replicate 5-cc sub-samples were taken to measure moisture content. Samples were weighed to the nearest 0.001 g, dried for 48 hours between 60-70C, and re-weighed to the nearest 0.001 g. Data Analysis All statistical analyses were performe d using SAS 8.02 software. Shapiro-W ilks normality tests were used to test for normal data distributions. Width-to-depth ratio, bank slopes, fine sediment, organic matter, and metal con centrations were log10 tr ansformed for further

PAGE 24

24 analysis to ensure normality. An alpha level of 0.05 was used as a threshold for determining significant effects. Geomorphology W idth-to-depth ratios were calculated for each transect by dividing the width of the channel at bank full by the average channel depth (Gordon et al. 2002). Bank slopes were calculated for each transect by taking average bank slope s of the right and left side of the stream. Sediment Particle Size Distribution Fine sedim ent content was calculated as the proportion of mass in each sediment sample less than 63 m (Allan 1995). Organic matter Fine and coarse organic matter conten t were calculated using Equation 2-1. % Organic Matter = (DW60DW550)/DW60 (2-1) DW60 = dry weight after drying at 60 C; DW550 = ashed weight after drying at 550 C. Total organic matter content was calculated by combining fine and coarse organic matter and using the same formula. Metals Laboratory metal concentrations in the sedime nt were measured as received (wet weight); therefore, moisture content was calculated for each sediment sample using Equation 2-2. % Moisture = ((WW DW)/ WW) x 100 (2-2) WW = wet weight; DW60 = dry weight after drying at 60 C Moisture content was averaged for each sample, and metal concentrations were calculated on a dry weight basis for comparison using Equation 2-3. [metal]d = [metal]w / (1Proportion Moisture) (2-3) [metal]d = dry weight metal concentrations; [metal]w = wet weight metal concentrations

PAGE 25

25 Statistical Analyses The following statistical analyses were perform ed for width-to-depth ratios, bank slopes, fine sediment content, and total, fine, and coar se organic matter content. Repeated measures Analysis of Variances (ANOVAs) were used to test for effects of position relative to road crossings (up versus down), s eason (summer versus winter), management type (non-managed, concrete apron, and riprap) and interactions of these variables across all streams and then by management type. Repeated measures ANOVAs were also used to test for effect of position in the stream (0, 5, 10, and 15 meters up and downstream), season, management type and interactions of these variable across all streams and then by management type. T-tests were performed to test for differe nce in metal concentrations between up and downstream sections. ANOVA tests were used to test for differences in metal concentrations among transects at varying distances (0, 5, 10 and 15 meters up and downstream) from road crossings. Tukey post-hoc comparisons were used to test for pairwise differences among all transects for significant ANOVA results. Linear regressions were used to test for linear relationships between metal con centrations and percent organic matter, percent fine sediment (<63m), and traffic densities.

PAGE 26

26 CHAPTER 3 RESULTS AND DISCUSSION Geomorphology Downstream of road crossings, stream s expe rience increased peak discharge during rain events as additional water is routed into the stream via the road. These events increase erosion of the stream bed and bank. To quantify changes in stream geomorphology caused by road crossings, cross-sectional profiles were measured up and downstream of roads. Upstream to Downstream Difference Width-to -depth ratio. Repeated measures ANOVA was used to test for differences in width-to-depth ratio between up a nd downstream sections for all str eams. There were significant (F1,68 = 6.23, P = 0.02) differences in widthto-depth ratios between up (X = 9.64) and downstream (X = 7.00) (Figure 3-1). Th ere were no significant (F1,68 = 0.00; P = 0.98) differences between summer a nd winter sampling periods. Bank slope. Repeated measures ANOVA was used to test for difference in bank slope between up and downstream sections. There was a significant (F1,68 = 11.77; P = 0.001) difference in bank slope between up (X = 1.28) and downstream (X = 2.00) (Figure 3-2). There was also a significant (F1, 68 = 4.44; P = 0.04) effect of season (F igure 3-3) with summer slopes being lower (X = 1.63) than winter slopes (X = 1.75). Roads had a significant impact on the geom orphology of the stream. Downstream sections of sampled streams had lower widt h-to-depth ratios and higher bank slopes than upstream sections. Downstream channels were mo re narrow and incised than upstream sections. Road crossings facilitate connectivity of imperv ious surfaces to the stream channel, causing accelerated peak discharge following storm events (Wang et al. 2001) leading to channel bed and

PAGE 27

27 bank erosion (Forman and Alexander 1998). McBride and Booth ( 2005) found that urban streams had greater cross-sectional dimensi ons compared to non-urban streams. There was no effect of season on widthto-depth ratios; however, bank slope was significantly affected by season w ith higher bank slopes in winter than summer. Bank slope measurements are more sensitive to slight alte rations than width-to-depth measurement. The change in bank slope may be due to storm events between summer and winter sampling. During storm events water is flushed into the stream often causing erosion of the bed and bank. Zone of Influence Repeated measures AN OVA was used to test for differences among transects at varying distances from the road crossing (0, 5, 10 and 15 meters up and downstream of the road crossing). There were no significant differen ces among transects (0, 5, 10 and 15 meters up and downstream of the road crossing ) in width-to-depth ratio (F7, 62 = 1.26; P = 0.29) or bank slope (F7, 62 = 1.66; P = 0.13). Geomorphology downstream of road crossings is significantly different from upstream; however, no clear zone of influence or signi ficant differences among transects could be identified. When comparisons were made betwee n transects, statistical power was lowered and differences were not significant. Influence of Management Type Repeated measures ANOVA revealed a significant effect of m anagement type on widthto-depth ratios (F2, 68 = 6.52; P = 0.003). There was a significant difference in width-to-depth ratios between up and downstream at non-managed (F1, 22 = 5.84; P = 0.02) and riprap sites (F1, 22 = 5.52; P = 0.03); however, at sites with concrete aprons, there were no significant differences (F1, 22 = 0.04; P = 0.84) (Figure 3-4). Downstream sections at non-managed and riprap sites had lower (respectively: X = 6.59; X = 4.03) width-to-depth ra tios than upstream sections

PAGE 28

28 (respectively: X = 12.69; X = 6.89). Width-to-depth ratios were not significantly different among transects (0, 5, 10 and 15 m up and downs tream of road crossing) for non-managed, concrete apron or riprap sites (respectively: F7, 16 = 1.13; P = 0.39; F7, 16 = 0.19; P = 0.98; F7, 16 = 1.85; P = 0.15). Repeated measures ANOVA revealed a significant effect of management type on bank slope (F1, 68 = 4.47; P = 0.02). Bank slope was significantly different between up (X = 0.99) and downstream sections (X = 2.24) at non-managed sites (F1, 22 = 16.66; P = 0.0005); differences between up (X = 1.64) and downstream sections (X = 2.39) approached significance at riprap (F1, 22 = 3.63; P = 0.07) sites; however, there was no si gnificant difference at concrete apron sites (F1, 22 = 0.23; P = 0.63) (Figure 3-5). Seasonal eff ects on bank slope were not significant at managed sites (concrete aprons and riprap); however, there was a significant difference in bank slope at non-managed sites between summer (X = 1.49) and winter (X = 1.75) sampling periods (F1, 22 = 12.63; P = 0.002) (Figure 3-6). There were no significant differences among transects (0, 5, 10 and 15m up and downstream of road crossing) for bank slope fo r non-managed, concrete apron, or riprap sites (respectively: F7, 16 = 2.12; P = 0.11; F7, 16 = 0.25; P = 0.96; F7, 16 = 2.11; P = 0.10). Seasonal effects on bank slope were depende nt on location within the stream for concrete apron sites (F7, 16 = 5.77; P = 0.002) (Figure 3-7). Management type had a significant influence on stream geomorphology. Concrete aprons and riprap are intended to slow flow di rectly downstream of road crossings to minimize erosion of the stream (DeWiest and Livingst on 2002). Downstream sections at non-managed and riprap sites were more incised and ha d higher bank slopes than downstream sections; however, up and downstream sections of concrete ap ron sites were not sign ificantly different.

PAGE 29

29 The concrete apron appears to minimize erosion of downstream sections. Two concrete apron sites also had swales that routed water from the road to both the up and downstream sections of the stream. Swales may also play a role in minimizing erosion in these streams. Seasonal influence on bank slope was significant only at non-managed sites, with higher bank slopes in winter. BMPs may be effective at preventing bank erosion following storm events. The seasonal effect on bank slope was de pendent on the location w ithin the stream for concrete apron sites; however, no clear pattern could be iden tified and was most likely not associated with the road crossing. Sediment Particle Size Distribution Roads are sources of fine sedim ent that duri ng storm events are washed into streams and deposited in areas of low flow (Allan 1995). To measure fine sediment distribution, the percent of sediment less than 63 m was calculated for each benthic sample. Upstream to Downstream Differences Repeated measures ANOVA was used to test for differences in percent fine sedim ents between up and downstream sections for all stream s. Differences approached significance (F1, 210 = 3.58, P = 0.06), with downstream (X = 6.52) being greater than upstream (X = 4.45) (Figure 3-8). There was no difference betw een summer and winter sampling periods. Downstream sections had slightly higher per cent fine sediment than upstream. The grain size distribution of urban stream sediment tends to shift towards smaller particle sizes; however, in densely urban areas, fine particles are selec tively removed during storm events (McBride and Booth 2005). Sediment size distribution is primar ily determined by geology of the drainage area and current velocity (Allan 1995), but roads also act as sources of fine sediment to streams (Forman and Alexander 1998). Increased deposition of fine sediments has negative implications

PAGE 30

30 for stream biota including alteration of habitat and interference with respiration, feeding, and reproduction (Wood and Armitage 1997). Zone of Influence Repeated measures ANOVA revealed no significant (F1, 210 = 1.13, P = 0.34) differences in percent fine sediment among transects at varying distances (0, 5, 10 and 15 meters up and downstream) from road crossings. Roads act as a source of fine sedi ments; however, flow regime drives sediment depos itional processes (Allan 1995). Influence of Management Type Repeated measures ANOVA revealed a signific ant effect of m anagement type on percent fine sediment (F2, 210 = 24.83, P < 0.0001). Differences between up and downstream were significant for non-managed (F1, 70 = 16.72, P = 0.0001) and concrete apron sites (F1, 64 = 4.57, P = 0.04); however, not for riprap sites (F1, 70 = 2.11, P = 0.15). At non-managed sites, percent fine sediment was greater downstream (X = 12.70) than upstream (X = 1.93), while at concrete apron sites, upstream (X = 9.96) was greater than downstream (X = 4.78) (Figure 3-9). Repeated measures ANOVA was used to test for differences among transects at varying distances (0, 5, 10 and 15 meters up and downstream) from roads according to management type. There were significant differences among transects at non-managed (F7, 64 = 2.87, P = 0.01) and riprap (F7, 64 = 2.15, P = 0.05) sites, and differences approached significance at concrete apron sites (F7, 62 = 1.82, P = 0.10). Tukey post-hoc comparisons were used to test for pairwise differences among all transects. At non-manage d sites 0 meters downstream was significantly greater than 10 meters upstream; and 5 meters downstream was si gnificantly greater than 15, 10, and 5 meters upstream (Figure 3-10). Effect s of location were dependent on season at nonmanaged sites (F7, 64 = 3.95, P = 0.001), with hot spots of high percent fine sediment varying between summer and winter sampling (Figure 3-11). At concrete ap ron sites, fine sediment 10

PAGE 31

31 meters upstream was significantly greater than 0 meters upstream and 0, 10 and 15 meters downstream (Figure 3-10). At riprap sites, fi ne sediment 15 meters upstream was significantly less than 15 meters downstream (Figure 3-10). At non-managed sites, percent fine sediment was much greater downstream than upstream. Fine sediment is washed into th e stream during storm events and deposited downstream. Upstream, fine sediment distribu tion seemed uninfluenced by the road. Fine sediments were higher at individual downstream than upstream tr ansects; however, no clear zone of influence could be identified. Transects with highest percent fine se diment were different between summer and winter suggesting that depositional zones are constantly changing At concrete apron sites per cent fine sediment was higher up than downstream. This is likely due to the swales at two concrete apr on sites that directed road runoff both up and downstream. Fine sediment may se lectively deposit in upstream s ections. In addition, transects 10 meters upstream had very high average percen t fine sediment, likely due to one concrete apron site that displayed wetland like characteristics with very slow flows and flocculent bed material and thus a depositional zone for fine sediment. At riprap sites, no differences existed in percent fine sediment between up and downstream. Riprap slows flows and allows depos ition of fine sediment to occur within plunge pools and riprap of these sites (DeWiest and Livings ton 2002). No clear trend in fine sediment deposition across transects existed at riprap sites. Riprap seems to be the most effective management practice for trapping fine material; ho wever, during severe storm events, these areas could be sources of fine sediment. Organic Matter Storage Benthic particulate organic m atter was measur ed up and downstream of road crossings to determine their influence on organic matter storage. In headwater streams, allochthonous

PAGE 32

32 organic matter is a primary food base for macr oinvertebrates. Culverts may disrupt natural depositional patterns of organic matter. In addition, increased peak discharge associated with road crossings, may result in the flushing of organic matter downstream. Upstream to Downstream Differences Repeated measures ANOVA was used to test for differences in percent organic m atter between up and downstream sections for all stream s. No significant differences were found for total (F1, 212 = 0.07, P = 0.79), fine (F1, 212 = 0.49, P = 0.49) and coarse (F1, 212 = 0.64, P = 0.42) organic matter (Figure 3-12). There was a significant seasona l effect on total (F1, 212 = 51.24, P < 0.0001), fine (F1, 212 = 41.12, P < 0.0001) and coarse (F1, 212 = 32.63, P < 0.0001) organic matter, with summer (respectively: x = 3.33, x = 0.96, x = 2.37) lower than winter (respectively: X = 4.87, x = 1.42, X = 3.46) (Figure 3-13). Seasonal effect s were dependent on location within the stream (F1, 212 = 6.77, P = 0.01) for fine organic matter, with differences between summer and winter being greater for dow nstream (respectively: X = 0.54, X = 1.07) than upstream (respectively: X = 1.39, X = 1.76) (Figure 3-14). Road crossings had no apparent influen ce on organic matter storage between up and downstream. There was, however, a strong seas onal component, with winter organic matter (fine and coarse) storage greater than summer. Th e greatest leaf fall in Florida usually occurs between September and December, with a smaller peak in January and February (Roberts 2002), resulting in more organic matter during these periods. Seasonal influence on fine organic matter was stronger downstream than upstream, potentially due to summer storm events selectively flushing fine organic matter from downstream sections.

PAGE 33

33 Zone of Influence Repeated measures AN OVA was used to test for differences among transects at varying distances from road crossings (0, 5, 10, and 15 meters up and downstream). There were no significant differences for total (F7, 206 = 1.01, P = 0.43) or fine organic matter (F7, 206 = 0.47, P = 0.86); however, there were significant diffe rences for coarse organic matter (F7, 206 = 2.20, P = 0.04). Tukey post-hoc comparisons were used to test for pairwise differences among all transects, and resulted in a significant differe nce in coarse organic matter between 0 meters upstream (X = 3.76) and 0 meters downstream (X = 1.28) (Figure 3-15). There is no clear trend in organic matter stor age associated with location in the stream relative to the road. Coarse organic matter was greater 0 meters upstream than 0 meters downstream; however, no other significant differen ces existed. Variability in organic matter storage is likely due to flow a nd local depositional patterns, as opposed to proximity to road crossings. Influence of Management Type Managem ent type had a signi ficant influence on total (F1, 212 = 27.14, P < 0.0001), fine (F1, 212 = 38.09, P < 0.0001) and coarse (F1, 212 = 19.79, P < 0.0001) organic matter. Differences in total organic matter between up and downstream approached significance (F1, 70 = 3.27, P = 0.08) at non-managed sites and were signi ficantly different at concrete apron (F1, 70 = 10.77, P = 0.002) and riprap sites (F1, 70 = 6.29, P = 0.01). At non-managed and concrete apron sites, total organic matter was higher upstream (respectively: X = 2.42, X = 11.56) than downstream (X = 2.27, X = 4.99); however, at riprap sites, to tal organic matter was lower upstream (X = 1.37) than downstream (X = 2.00) (Figure 3-16). There was a significant seasonal effect on total organic matter at non-managed (F1, 70 = 10.51, P = 0.002), concrete apron (F1, 70 = 12.49, P =

PAGE 34

34 0.0007) and riprap (F1, 70 = 32.19, P < 0.0001) sites, with summer organic matter (respectively: X = 1.50; X = 7.30; X = 1.19) lower than wint er (respectively: X = 3.20, X = 9.24, X = 2.18) (Figure 3-17). Seasonal effect s on total organic matter were de pendent on location within the stream (F1, 70 = 5.89, P = 0.02) for concrete apron sites. Upstream, total organic matter was greater in summer (X = 12.87) than winter (X = 10.24), while downstream, total organic matter the opposite was true (Figure 3-18). Differences in fine organic matter between up and downstream approached significance at non-managed sites (F1, 70 = 3.79, P = 0.06), were significa nt at concrete apron (F1, 70 = 6.92, P = 0.01) sites, and were not significant at riprap (F1, 70 = 0.12, P = 0.73) sites. Fi ne organic matter at non-managed and concrete apron sites wa s higher upstream (respectively: X = 1.22, X = 3.15) than downstream (respectively: X = 0.86, X = 1.19) (Figure 3-16). There were seasonal effects on fine organic matter at non-managed (F1, 70 = 11.44, P = 0.001), concrete apron (F1, 70 = 16.87, P = 0.001), and riprap (F1, 70 = 14.27, P = 0.003) sites, with summer fine organic matter (respectively: X = 0.58, X = 2.00, X = 0.32) lower than winter (X = 1.50, X = 2.35, X = 0.40) (Figure 3-17). Seasonal effects on fine organi c matter were dependent on location within the stream (F1, 70 = 8.07, P = 0.006) for concrete apron sites. For upstream, fine organic matter was greater in summer (X = 3.42) than winter (X = 2.88); however, downstream fine organic matter was greater in winter (X = 1.82) than summer (X = 0.57) (Figure 3-19). There were no significant di fferences in coarse organic matter between up and downstream for non-managed sites (F1, 70 = 0.68, P = 0.41); however, there were significant differences at concrete apron (F1, 70 = 14.52, P = 0.0003) and riprap sites (F1, 70 = 4.87, P = 0.03). At concrete apron sites, coarse organic matter was higher upstream (X = 8.41) than downstream (X = 3.79); and at riprap sites, coarse organic matter was lower upstream (X = 1.01) than

PAGE 35

35 downstream (X = 1.64) (Figure 3-16). Seasonal diffe rences approached significance at nonmanaged sites (F1, 70 = 2.85, P = 0.10) and were signi ficant at concrete apron (F1, 70 = 12.26, P = 0.0008) and riprap (F1, 70 = 21.18, P < 0.0001) sites, with summer coarse organic matter (respectively: X = 0.92, X = 5.31, X = 0.87) lower than winter (respectively: X = 1.70, X = 6.89, X = 1.78) (Figure 3-17). Seasonal effects on coarse organic matter were dependent on location within the stream (F1, 70 = 6.09, P = 0.01) for concrete apron sites. For upstream, coarse organic matter was greater in summer (X = 9.45) than winter (X = 7.36); however, for downstream fine organic matter was greater in winter (X = 6.41) than summer (X = 1.17) (Figure 3-20). Repeated measures ANOVA was used to test for differences among transects at varying distances (0, 5, 10, and 15 meters up and downstr eam) from the road crossing by management type. There were no significant differen ces among transects at non-managed sites (F7, 64 = 1.10, P = 0.37); however, there were significant diffe rences among transects at concrete apron (F7, 64 = 2.42, P = 0.03) and riprap (F2, 64 = 2.67, P = 0.02) sites. Tukey post-hoc comparisons were used to test for pairwise differences among transects. At concrete apron site s, total organic matter 10 meters upstream was significantly greater than at 0 and 5 meters downstream, and 0 meters upstream was significantly greater than at 0 mete rs downstream. At riprap sites, total organic matter at 15 meters upstream was significantly le ss than at 10 meters upstream and 10 meters downstream (Figure 3-21). There were no si gnificant differences among transects for fine organic matter at non-managed, concrete ap ron, or riprap sites (Figure 3-21). There were no significant diffe rences among transects for coarse organic matter at nonmanaged sites (F7, 64 = 0.50, P = 0.83); however, there were si gnificant differences at concrete apron (F7, 64 = 4.60, P = 0.0003) and riprap (F7, 64 = 3.17, P = 0.006) sites. At concrete apron

PAGE 36

36 sites, 0 meters downstream was significantly less than 15, 10, 5, and 0 meters upstream and 15 meters downstream, while 5 meters downstream was significantly less than 10 and 0 meters upstream. At riprap sites, 15 meters upstream wa s significantly less than 10 meters upstream and 15 meters downstream (Figure 3-21). Management type had a signifi cant influence on organic matter storage. At non-managed sites, total and fine organic matter was slightly greater upstream than downstream and coarse organic matter was not significan tly different. These differences may be due to selective flushing of organic matter downstream; however, they could also be associated with local depositional patterns and canopy cover. At concrete apron sites, tota l, fine, and coarse organic matter were much greater upstream than downstream. Coarse organic matter at i ndividual downstream transe cts was significantly less than upstream transects, but no clear zone of influence could be identified. The culvert at concrete aprons may trap organic matter upstream, and increased peak discharge may flush organic matter from downstream sections; however swales associated with two concrete apron sites may also influence organic matter depositiona l patterns. Upstream of one concrete apron site displayed wetland like characteristics with very low flows and flocculent bed material. Organic matter selectively deposits in areas of low flow (Allan 1995), resulting in elevated organic matter concentrations in these areas. At riprap sites, total and co arse organic matter were greater downstream than upstream. At riprap sites, angular stones are in place to slow flow and di ssipate erosive energy (DeWiest and Livingston 2002). The rocks have the ability to trap organic matter and the slower flows allow organic matter to deposit within the stream bed. There were significan t differences in total and coarse organic matter between individual upstream transects as we ll as differences in coarse

PAGE 37

37 organic matter between indivi dual up and downstream transect s; however, no clear zone of influence could be identified. Organic matter transport and storage is influen ced by flow regime. Differences in storage may be a result of the road crossings and management type, but may also be associated with stream geomorphology or canopy cover. Fine Sediment and Organic Matter Relationship Organic m atter and fine sediment show simila r depositional trends across study streams. Linear regression was used to te st for a relationship between organic matter and fine sediment in stream bed material. There was a positive significant relationship between the log proportion organic matter and log pr oportion fine sediment (R2 = 0.33, P < 0.0001) (Figure 3-22). The relationship between organic matter and fine sediment is well studied (Horowitz 1991; Rhoads and Cahill 1999). The deposition of fine sediments and organic matter is dependent on flow regime and density (Allan 1995). They tend to be transported by fast moving water and deposit in areas of low flow. Metals Road runoff contains contam inants including heavy metals associated with automobile leaks and wear. Road crossings drain impervious surfaces to the stream network, and thus are point sources for stormwater runoff. Copper, chromium, nickel, lead, and zinc were measured up and downstream of the road crossing. One result fr om stream number four was an outlier with respect to lead (550 mg/kg). It is unclear whether this was due to sampling error or a hot spot of lead contamination; therefore, statistical te sts involving lead were run twice, once with the outlier included and once without.

PAGE 38

38 Up to Downstream Differences A t-test was used to test for differences in metal concentrations between up and downstream locations for all streams combin ed. There was a significant difference for chromium (t = -2.04, P = 0.05), with downstream concentrations (X = 2.63) greater than upstream (X = 1.92); however, there were no significan t differences for copper (t = 0.25, P = 0.80), nickel (t = -1.53, P = 0.13), or zinc (t = -1.30, P = 0.20). When the outlier was included in statistical analyses, there was a significant differe nce (t = -1.96, P = 0.05) in lead concentrations between up (X = 7.02) and downstream (X = 29.60); however, when it was removed, the difference was not significant (t = -1.67, P = 0.10). (Figure 3-23). Average metal concentrations downstream we re higher than upstream for chromium and lead. Such variability in heavy metal concentr ations is unclear. Perdikaki and Mason (1999) also found no significant difference in lead or zinc concentrati ons up and downstream of a main road. Metals from road runoff may be depositing both up and downstrea m of road crossings. Alternatively, certain metals may be more mobi le than others, Rhoads and Cahill (1999) found that concentrations of nickel and chromium were lower in the suspended load than bed material, suggesting that these metals are relatively i mmobile, while the opposite was true for copper and zinc. Establishing baseline levels for heavy me tals is difficult considering atmospheric depositional sources; however, Chen et al. (1999) did so for potentially toxic metals based on 448 Florida surface soils, and the Canadian Coun cil of Ministers of the Environment (2001) established threshold and probable effect levels for sediments (T able 3-1). Threshold effect levels (TEL) are concentrations below which adve rse effects are rare. Probable effect levels (PEL) are concentrations above wh ich adverse effects are expected to occur frequently (Smith et

PAGE 39

39 al. 1996). Sample metal concentrations were comp ared to baseline, threshold and probable effect levels. No samples exceeded any of the three comparison levels for copper or chromium. One sample exceeded baseline concentrations for zi nc, but did not exceed threshold levels. No samples exceeded baseline concentrations fo r nickel; however, one sample exceeded the threshold effect level. Six samples exceeded the baseline and threshold effect levels for lead, and one sample (the outlier) exceeded probable ef fect levels. Beasley and Kneale (2004) found sediments of headwater streams in the UK most often exceeded Ontario Ministry of Environment (OME) toxicity standards for lead and zinc. The six samples that exceeded threshold effect levels for lead and the one sample that exceeded threshold levels for nickel came from four different sites. Six out of the seven samples that exceeded threshold effect levels were downstream of road crossings. One transect upstream had lead concentrations which exceeded threshold effect levels, and zinc concentrations that exceeded baseline levels. Headwater streams are expected to be uncontaminated; however, four out of nine sites sampled contained sediments that exceeded threshold effect levels for lead and one for nickel. Table 3-1. Baseline, threshold, and probable effect levels Metal Baseline levels (mg/kg) Threshold level (mg/kg) Probable effect levels (mg/kg) Copper 0.22-21.0 35.70 197.00 Chromium 0.89-80.7 37.30 90.00 Nickel 1.70-48.5 18.00 36.00 Lead 0.69-42.0 35.00 91.30 Zinc 0.89-29.6 123.00 315.00 Baseline levels for Florida surface soils were es tablished by Chen et al. (1999). Threshold and probable effect levels were established by Cana dian Council of Minister s of the Environment (2002).

PAGE 40

40 Zone of Influence ANOVAs were used to test for difference in m etal concentrations among transects at varying distances (0, 5, 10, and 15 meter up and downstream) from road crossings. No significant difference was noted among transects for any metals (Figure 3-24). There were no visible trends in metal con centrations among transects with distance from road crossings. Flow regime, organic matter, particle size, traffic densities and distance from the road all influence stream me tal concentrations (Horowitz 1991; Rhoads and Cahill 1999); therefore, position in the stream relative to the road does not fully explain concentration variability. Relationships among Metals, Orga nic Matter, and Particle Siz e Linear regressions were used to test for re lationships between per cent fine sediment and metal concentrations. Relationshi ps were significant for copper (R2 = 0.15, P = 0.008), chromium (R2 = 0.21, P < 0.0001), nickel (R2 = 0.18, P = 0.0002), lead with outlier (R2 = 0.15, P = 0.0008), lead without outlier (R2 = 0.27, P < 0.0001) and zinc (R2 = 0.26, P < 0.0001). (Figure 3-25). Linear regressions were used to test for re lationships between orga nic matter and metal concentrations. Relationship s were significant for copper (R2 = 0.14, P = 0.001), chromium (R2 = 0.13, P = 0.002), nickel (R2 = 0.09, P = 0.01), and zinc (R2 = 18.73, P = 0.0001), and approached significance for lead without the outlier (R2 = 0.05, P = 0.09), but was not significant with the outlier (R2 = 0.01, P = 0.33). Metal binding capacity has grea t influence on the location of hot spots of metal concentrations. The association among organic ma tter, particle size, and trace metals is well documented and apparent in this study (Horowitz 1991; Estebe et al. 1997 ; Lee et al. 1997). Smaller particles have a higher surface area to volume ratio, and therefore have increased

PAGE 41

41 potential for adsorption of trace metals. In co ntaminated areas, trace metal concentrations increase with decreasing particle size (Rhoads and Cahill 1999). Trace metals, specifically zinc and copper, also have a high affi nity to adsorb to organic matte r (Krupadam et. al. 2007). The relationships between organic matter and metals relative to particle size and metals were significant; but not strong. A nu mber of factors influenced metal concentrations within the stream, and organic matter and fine sediment c oncentrations explain some variability in this study. Both are correlated due to flow regime (Allan 1995); therefore, independent effects of particle size and organic matter on meta l concentrations cannot be separated. Proximity to a source of metals such as a road is an important factor; however, hot spots of metal contamination exist due to preferential adsorption of trace metals and depositional processes associated with these ma terials (Rhoads and Cahill 1999). Some studies investigating metal concentrations within stream s concentrated their sampling efforts on areas of deposition (Shelton and Capel 1994; Sutherland 2 000). In addition, they only ran chemical analyses on the portion less than 63 m. This sampling strategy likel y identifies hot spots of metal contamination. These areas are where stre ams are most vulnerable and likely areas where remediation may be most effective. Influence of Traffic Density on Metal Concentrations Linear regressions were used to test f or relationships between average metal concentrations up and downstream of road cros sings and traffic densities. Relationships downstream were significant for nickel (R2 = 0.67, P = 0.007) and lead with (R2 = 0.66, P = 0.008) and without the outlier (R2 = 0.45, P = 0.05), and approached significance for copper (R2 = 0.40, P = 0.07) and zinc (R2 = 0.37, P = 0.08), but were not significant for chromium (R2 = 0.21, P = 0.21). Relationships between traffic densit ies and metal concentrations upstream were

PAGE 42

42 significant for lead (R2 = 0.60, P = 0.01) and nickel (R2 = 0.44, P = 0.05), and approached significance for chromium (R2 = 0.37, P = 0.08); however, were not so for copper or zinc. (Figure 3-27) Van Hassel et al. (1980) and Ma rsalek et al. (1999) also found that traffic densities influence metal accumulation in stream sediment. The primary sources of heavy metals in runoff are automobiles (Table 3-2); however, metals associated with build ing roofing and siding materials are also transported to the stream via the road (Davis et al. 2001). Nickel and lead exhibit the strongest correlation to traffic densities. The greatest current source of lead is brick walls, followed by wet deposition (Davis et al. 2001). Levels of lead in stormwater runoff are often low; therefore, high leve ls found in these streams may be due to historic use of leaded gasoline (Turer et al. 2001). Ni ckel is associated with diesel fuel, oil, metal plating, brake linings and asphalt paving (Grant et al. 2002). The other metals are also associated with automobiles (Table 3-2) and correlations with traffic densities are appa rent. Although there are no clear trends in metal concentrations by location in the stream, traffic densities associated with each site, explain some of the variability. Table 3-2. Primary sources of h eavy metals in road runoff. Metals Sources Chromium Tire wear, brake pads, combustion of oils, and insecticides Copper Metal plating, bearing and bushing wear, moving engine parts, brake lining wear, fungicides and insecticides Lead Leaded gasoline and tire wear Nickel Diesel fuel and gasoline, lubricating oil, me tal plating, bushing wear, brake lining wear, and asphalt paving Zinc Tire wear, motor oil, grease Adapted from Grant et al. 2003. Influence of Management Type Management type had a significant effect on metal concentrations for copper (F2, 68 = 4.01, P = 0,02) and zinc (F2, 68 = 4.70, P = 0.01); approached significance for chromium (F2, 68 =

PAGE 43

43 2.90, P = 0.06) and nickel (F2, 68 = 2.86, P = 0.06); but was not significant for lead with (F2, 68 = 0.15, P = 0.86) or without (F2, 68 = 0.88, P = 0.42) the outlier. T-tests were used to test for difference between up and downstream metal concentrations by management type. At non-managed sites, there were significant differen ces for nickel (T = 2.78, P = 0.01) and lead (T = -2.13, P = 0.05). Differences approached significance for chromium (T = -1.83, P = 0.08) and copper (T = -1.90, P = 0.07); and were not significant for zinc (T = -1.66, P = 0.12). Differences between up and downstream were not significant for any metals at concrete apron and riprap sites. (Figure 3-28) ANOVAs were used to test for differences in metal concentrati ons among transects at varying distances (0, 5, 10, 15 meters up and dow nstream) from road crossings according to management type. There were no significan t differences among transects for non-managed, concrete apron, or riprap sites. Metal concentrations were higher downstream than upstr eam at non-managed sites; however, at concrete apron and ri prap sites, differences were not significant. Concrete aprons and riprap slow flow directly dow nstream of road crossings, allowing materials to deposit closer to the road (DeWiest and Livingston 2002), thus minimizing the downstream zone of influence. At concrete apron sites, coppe r and zinc were in higher co ncentrations upstream than downstream. At two of the concre te apron sites, there were conc rete swales that routed runoff both up and downstream. Swales are waterways in tended to convey stormwater to the stream with minimal erosion (DeWiest and Livingst on 2002). Erosion may be minimized; however, contamination of upstream sections may be facilitated. This may be an additional factor influencing metal concentration distribution within streams. For riprap sites, finer sediments, organic matter and metals adsorbed to those materials would deposit wi thin the plunge pool and

PAGE 44

44 riprap downstream of the crossing, thus reduc ing metal concentrations downstream and restricting the zone of contamination. The plunge pool and riprap sections may become highly contaminated over time and require further mana gement. These areas may also be selectively utilized by macroinvertebrates due to increased dissolved oxygen and habitat heterogeneity, thus acting as ecological traps. Ecological traps o ccur when habitat selection and suitability are dissociated (Kristan 2003). Fu rther investigation of invert ebrate communities and metal accumulation would be needed to test this hypothesis.

PAGE 45

45 0 2 4 6 8 10 12 Upstream DownstreamAverage width-to-depth ratio Figure 3-1. Average width-to-depth ratio fo r up and downstream sections of study streams (F1,68 = 6.23, P = 0.02). 0 0.5 1 1.5 2 2.5 Upstream DownstreamAverage Bank Slope Figure 3-2. Average bank slope up and downstream relative to road crossings (F1,68 = 11.77; P = 0.001).

PAGE 46

46 1.5 1.55 1.6 1.65 1.7 1.75 1.8 1.85 1.9 Summer WinterAverage Bank Slope Figure 3-3. Average bank slope for summer and winter sampling periods (F1, 68 = 4.44; P = 0.04). 0 2 4 6 8 10 12 14 16 Non-managedConcrete ApronRiprapAverage Width-to-Depth Ratio upstream downstream Figure 3-4. Average width-to-depth ratio up and downstream relative to road crossings by management type: non-managed (F1, 22 = 5.84; P = 0.02), concrete apron (F1, 22 = 0.04; P = 0.84), riprap (F1, 22 = 5.52; P = 0.03).

PAGE 47

47 0 0.5 1 1.5 2 2.5 3Non-managedConcrete ApronRiprapAverage Bank Slope Upstream Downstream Figure 3-5. Average bank slope for up and downstream sections of streams according to management type: non-managed (F1, 22 = 16.66; P = 0.0005), concrete apron (F1, 22 = 0.23; P = 0.63), and riprap (F1, 22 = 3.63; P = 0.07). 0 0.5 1 1.5 2 2.5 3 3.5 Non-managedConcrete ApronRiprapAverage Bank Slope Summer Winter Figure 3-6. Average bank sl ope for summer and winter sampling periods according to management type: non-managed (F1, 22 = 12.63; P = 0.002).

PAGE 48

48 0 0.5 1 1.5 2 2.5 3 15m10m5m0m0m5m10m15m Upstream Downstream Average Bank Slope Summer Winter Figure 3-7. Average bank sl ope for summer and winter sa mpling periods according to location within the stream relative to th e road crossing at sites with concrete aprons. 0 1 2 3 4 5 6 7 8 Upstream DownstreamPercent (%) Fine Sediment (0.063mm) Figure 3-8. Average percent fine sediment up and downstream relative to road crossings (F1, 210 = 3.58, P = 0.06).

PAGE 49

49 0 2 4 6 8 10 12 14 16 Non-ManagedConcrete ApronRiprap Management TypePercent (%) Fine Sediment (<0.063 mm) Up Down Figure 3-9. Average percent fine sediment (<0.063 mm) up and downstream of road crossings according to management type: non-managed (F1, 70 = 16.72, P = 0.0001), concrete apron (F1, 64 = 4.57, P = 0.04), and riprap (F1, 70 = 2.11, P = 0.15).

PAGE 50

50 Concrete Apron 0 5 10 15 20 25 151050 051015 Upstream Downstream Percent (%) Fine Sediment (<0.063mm) Figure 3-10. Average percent fine sedime nt (<0.063mm) relative to road crossings by management type. Non-Managed 0 5 10 15 20 25 30 151050051015 Upstream Downstream Percent (%) Fine Sediment (<0.063mm)

PAGE 51

51 Riprap 0 1 2 3 4 5 6 151050051015 Upstream Downstream Percent (%) Fine Sediment (<0.063mm) Figure 3-10. Continued 0 5 10 15 20 25 151050 051015 Upstream Downstream % Fine Sediment (< 0.063mm) winter summer Figure 3-11. Average percent fine sediment re lative to the road crossing seasonally at nonmanaged sites.

PAGE 52

52 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 Fine Coarse% Organic Matter Up Down Figure 3-12. Average percent fine and coar se organic matter in study streams upstream and downstream of road crossing: fine (F1, 212 = 0.49, P = 0.49) and coarse (F1, 212 = 0.64, P = 0.42). 0 0.5 1 1.5 2 2.5 3 3.5 4 Fine Coarse% Organic Matter Summer Winter Figure 3-13. Average percent fine and coarse organic matter by sampling period: fine (F1, 212 = 41.12, P < 0.0001) and coarse (F1, 212 = 32.63, P < 0.0001).

PAGE 53

53 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 Summer Winter% Fine Organic Matter Upstream Downstream Figure 3-14. Average percent fine organic matter according to position relative to road crossings and sampling period. 0 1 2 3 4 5 6 7 15m10m5m0m 0m5m10m15m Upstream Downstream % Coarse Organic Matter Figure 3-15. Average coarse organic matter according to location in the stream relative to the road crossing.

PAGE 54

54 0 2 4 6 8 10 12 14UpstreamDownstreamUpstreamDownstreamUpstreamDownstream Non-managed Concrete Apron Riprap % Organic Matter Coarse Fine Figure 3-16. Average percent organic matter acco rding to location relative to the road crossing by management type: total organic matter for non-managed (F1, 70 = 3.27, P = 0.08), concrete apron (F1, 70 = 10.77, P = 0.002) and riprap (F1, 70 = 6.29, P = 0.01) sites. 0 2 4 6 8 10 12 SummerWinterSummerWinterSummerWinter Non-managedConcrete Apron Riprap % Organic Matter Coarse Fine Figure 3-17. Average percent organic matter according to sampling period and management type: total organic matter at non-managed (F1, 70 = 10.51, P = 0.002), concrete apron (F1, 70 = 12.49, P = 0.0007) and riprap (F1, 70 = 32.19, P < 0.0001) sites.

PAGE 55

55 0 2 4 6 8 10 12 14 16 18 UpstreamDownstream% Total Organic Matter Summer Winter Figure 3-18. Average total organic matter by sa mpling period and location relative to the road crossing at concrete apron sites. 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 Upstream Downstream% Fine Organic Matter Summer Winter Figure 3-19. Average fine organic matter according to season and location relative to the road crossing at concrete apron sites.

PAGE 56

56 0 2 4 6 8 10 12 14 Upstream Downstream% Coarse Organic Matter Summer Winter Figure 3-20. Average coarse organic matter acco rding to season and location relative to the road crossing at c oncrete apron sites. A Non-Managed 0 5 10 15 20 25 30 35 15m10m5m0m 0m5m10m15m Upstream Downstream % Organic Matter Figure 3-21. Average percent organic matter a ccording to location relative to road crossings and management type: A) non-managed B) concrete apron C) riprap.

PAGE 57

57 B Concrete Apron 0 5 10 15 20 25 30 35 15m10m5m0m0m5m10m15m Upstream Downstream % Organic Matter C Riprap 0 5 10 15 20 25 30 35 15m10m5m0m0m5m10m15m Upstream Downstream % Organic Matter Coarse Fine Figure 3-21. Continued

PAGE 58

58 -3.5 -3 -2.5 -2 -1.5 -1 -0.5 0 -3.5-3-2.5-2-1.5-1-0.50 Log Fine SedimentLog Organic Matter Figure 3-22. Linear relationship between log proportion organic matter and log proportion fine sediment (R2 = 0.33, P < 0.0001).

PAGE 59

59 0 5 10 15 20 25 30 35 40 45 50 CopperChromiumNickelLead (with outlier) Lead (without outlier) ZincMetal Concentrations mg/kg Up Down Figure 3-23. Average metal concentrations upstr eam and downstream of road crossings for all streams combined: copper (t = 0.25, P = 0.80), chromium (t = -2.04, P = 0.05), nickel (t = -1.53, P = 0.13), lead (with outlier) (t = -1.96, P = 0.05), lead (without outlier) (t = -1.67, P = 0.10), a nd zinc (t = -1.30, P = 0.20).

PAGE 60

60 A 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 151050 051015 Upstream Downstream Copper Concentration mg/kg B 0 1 2 3 4 5 6 7 8 151050 051015 Upstream Downstream Chromium concentration mg/kg Figure 3-24. Metal concentrations relative to roads: A) copper B) chromium C) nickel D) lead with outlier E) lead wi thout outlier F) zinc.

PAGE 61

61 C 0 1 2 3 4 5 6 151050051015 Upstream Downstream Nickel Concentration mg/kg D 0 20 40 60 80 100 120 140 151050051015 Upstream Downstream Lead concentration mg/kg Figure 3-24. Continued

PAGE 62

62 E 0 5 10 15 20 25 30 35 40 45 151050051015 Upstream Downstream Lead Concentration mg/kg F 0 2 4 6 8 10 12 151050 051015 Upstream Downstream Zinc Concentration mg/kg Figure 3-24. Continued

PAGE 63

63 A -1 -0.5 0 0.5 1 1.5 -3.00-2.50-2.00-1.50-1.00-0.500.00 Log Fine SedimentLog Copper B -1.5 -1 -0.5 0 0.5 1 1.5 2 -3.00-2.50-2.00-1.50-1.00-0.500.00 Log Fine SedimentLog Chromium Figure 3-25. Relationship between metal concentrations and fine se diment A) copper (R2 = 0.15, P = 0.008), B) chromium (R2 = 0.21, P < 0.0001), C) nickel (R2 = 0.18, P = 0.0002), D) lead with outlier (R2 = 0.15, P = 0.0008), E) le ad without outlier (R2 = 0.27, P < 0.0001) and F) zinc (R2 = 0.26, P < 0.0001)

PAGE 64

64 C -2.5 -2 -1.5 -1 -0.5 0 0.5 1 1.5 -3.00-2.50-2.00-1.50-1.00-0.500.00 Log Fine SedimentLog Nickel D -0.5 0 0.5 1 1.5 2 2.5 3 -3.00-2.50-2.00-1.50-1.00-0.500.00 Log Fine SedimentLog Lead Figure 3-25. Continued

PAGE 65

65 E -0.5 0 0.5 1 1.5 2 2.5 -3.00-2.50-2.00-1.50-1.00-0.500.00 Log Fine SedimentLog Lead F -0.5 0 0.5 1 1.5 2 -3.00-2.50-2.00-1.50-1.00-0.500.00 Log Fine SedimentLog Zinc Figure 3-25. Continued

PAGE 66

66 A -1 -0.5 0 0.5 1 1.5 -2.50-2.00-1.50-1.00-0.500.00 Log Organic MatterLog Copper B -1.5 -1 -0.5 0 0.5 1 1.5 2 -2.50-2.00-1.50-1.00-0.500.00 Log Organic MatterLog Chromium Figure 3-26. Relationship between metal concentrations and organic matter: A) copper (R2 = 0.14, P = 0.001), B) chromium (R2 = 0.13, P = 0.002), C) nickel (R2 = 0.09, P = 0.01), D) lead (R2 = 0.01, P = 0.33), E) l ead without outlier (R2 = 0.05, P = 0.09), and F) zinc (R2 = 0.19, P = 0.0001).

PAGE 67

67 C -2.5 -2 -1.5 -1 -0.5 0 0.5 1 1.5 -2.50-2.00-1.50-1.00-0.500.00 Log Organic MatterLog Nickel D -0.5 0 0.5 1 1.5 2 2.5 3 -2.50-2.00-1.50-1.00-0.500.00 Log Organic MatterLog Lead Figure 3-26 Continued

PAGE 68

68 E -0.5 0 0.5 1 1.5 2 2.5 -2.50-2.00-1.50-1.00-0.500.00 Log Organic MatterLog Lead F -0.5 0 0.5 1 1.5 2 -2.50-2.00-1.50-1.00-0.500.00 Log Organic MatterLog Zinc Figure 3-26. Continued

PAGE 69

69 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 1.2 010000200003000040000 Traffic density (Vehicles/ DayLog Copper -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 010000200003000040000 Traffic density (Vehicles/day)Log Chromium A B -2 -1.5 -1 -0.5 0 0.5 1 010000200003000040000 Traffic density (Vehicles/ Day Log Nickel 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 010000200003000040000 Traffic density (Vehicles/ Day)Log Lead C D Figure 3-27. Linear regression between tr affic densities and l og transformed metal concentrations: a) copper (upstream: R2 = 0.22, P = 0.20; downstream: R2 = 0.40, P = 0.07) b) chromium (upstream: R2 = 0.37, P = 0.08; downstream: R2 = 0.21, P = 0.21) c) nickel (upstream: R2 = 0.44, P = 0.05; downstream: R2 = 0.67, P= 0.007) d) lead with outlier (upstream: R2 = 0.60, P = 0.01; downstream: R2 = 0.66, P = 0.008) e) lead without outlier (downstream: R2 = 0.45, P = 0.05) f) zinc (upstream: R2 = 0.22, P = 0.20; downstream: R2 = 0.37, P = 0.08).

PAGE 70

70 0 0.5 1 1.5 2 2.5 010000200003000040000 Traffic Density (Vehicles/Day)Log Lead -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 1.2 010000200003000040000 Traffic density (Vehicles/ DayLog Zinc E F -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 02000040000 Traffic density (Vehicles/ Day Upstream Downstream Linear (Downstream) Linear (Upstream) Figure 3-27. Continued

PAGE 71

71 A 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 Non-managedConcrete ApronRiprapConcentration of Copper mg/kg Upstream Downstream B 0 1 2 3 4 5 6 7 Non-managedConcrete ApronRiprapChromium concentration mg/kg Upstream Downstream Figure 3-28. Average metal concentrations relative to roads and management type. Significant differences at non-managed si tes for nickel (T = -2.78, P = 0.01) and lead (T = -2.13, P = 0.05).

PAGE 72

72 C 0 0.5 1 1.5 2 2.5 3 3.5 4 Non-managedConcrete ApronRiprapNickel concentration mg/kg Upstream Downstream D 0 20 40 60 80 100 120 Nonmanaged Concrete Apron Riprap (with outlier) Riprap (without outlier)Lead Concentration mg/kg Upstream Downstream Figure 3-28. Continued

PAGE 73

73 E 0 1 2 3 4 5 6 7 8 9 Non-managedConcrete ApronRiprapZinc concentration mg/kg Upstream Downstream Figure 3-28. Continued

PAGE 74

74 CHAPTER 4 CONCLUSIONS Florida is projected to surpass New York in total population by 2011. In order to support this growing population, new roads will be built accompanied by greater traffic volume. Understanding mechanisms by whic h road crossings influence streams can guide stormwater management decisions. The purpose of this stud y was to document the direct impacts of road crossings on stream ecosystems by comparing up and downstream sections and to assess the influence of current best management practices on these impacts. Impact of Road Crossings Roads and other im pervious surfaces reduce in filtration rates and increase peak discharge during storm events, resulting in degradation of stream ecosystem s. To quantify alterations, comparisons were made between up and downstream sections of road crossings. Downstream sections had lower width-to-depth ratios and higher bank slopes produc ing narrower, incised channels than upstream sections. Furthermore, there was a significant difference in bank slope between summer and winter sampling suggesting continued erosion over the six month period. Increased peak discharge results in erosion of bed and bank material, thus changing stream dimensions and altering the subsequent flow regi me. Material eroded from the bed and bank are suspended and transported, becoming a source of fine material for downstream sections. Fine sediment distribution is determined by local geology and flow regime. Roads carry fine sediment from the watershed to the stream. In areas of low flow, suspended fine sediments are deposited in the streambed. Downstream sec tions had slightly higher proportions of fine sediment indicating increased deposition downstream of crossings. However, fine sediment is easily resuspended and carried through the stream channel and deposited farther downstream, thus depositional zones may have been missed due to sampling design.

PAGE 75

75 Organic matter storage is determined by source and flow regime. There were no significant differences in organic matter stor age between up and downstream. There was a significant relationship between fine sediment and organic matter storage because both are influenced by flow regime and are deposited in areas of low flow. Metals are deposited on road surfaces through leaks and wear of automobiles and carried to the stream by stormwater runoff, preferentially adsorb to fine sediment and organic matter, and are therefore concentrated in depositional areas. Downstream sections had higher concentrations of chromium and lead. Lack of significant differences may be due to low sample size or the fact that upstream sections may also be impacted by road runoff. Overall differences between up and downstr eam sections suggest degradation of downstream areas, expressed as more incised cha nnels, steeper banks, more fine sediment, and higher metal contamination. Upstream sections ma y be influenced by roads as well. At some sites, high proportions of fine sediment and high concentrations of metals were found. There was much variability because all of these variable s interact with one another and are influenced by additional factors such as flow regime, traffi c density, land-use, and management practices. Regardless of additional variables, there was ev idence that overall downstream sections were degraded. Influence of Management Practices Two in-stream, best m anagement practices (B MPs) were in place at the onset of this study: concrete aprons and riprap. Downstream of stormwater outlets, such as culverts, concrete aprons and riprap are placed to dissipate ener gy to prevent scour and erosion. To assess the efficiency of these BMPs, up and downstream of road crossings were compared among nonmanaged, concrete apron and riprap sites. Ar eas downstream of non-mana ged sites display the

PAGE 76

76 expected degradation associated with roads, wi th incised channels, steeper bank slopes, more fine sediment, and higher metal concentrations. Concrete aprons and swales as sociated with two of the st udy sites seemed to dissipate erosive energy resulting in less difference in channel morphology betw een up and downstream sections. Swales route water off the road, pr oviding more storage and slowing flows, thus dissipating the increased peak di scharge during storm events that causes erosion. Concrete aprons seem to be effective at dissipating er osive energy and minimizing downstream erosion; however, after closely examining the results, si tes with swales display less difference in geomorphology between up and downstream. Investig ation of the influence of swales on road crossing impacts is needed to verify these results. Concrete apron sites also exhibited both much higher organic matter and fine sediment concentr ations upstream than downstream and elevated levels of metals. Downstream sections were not significantly different from upstream in metal concentrations, suggesting that bo th were equally influenced by metals in road runoff. Although management at these sites is effective at mi nimizing erosion, up and downstream sections are compromised by road crossings through meta l and fine sediment contamination. Riprap sites displayed the least differences between up and downstream. The latter were more incised and had higher bank slopes than upst ream. Riprap seems inefficient at dissipating erosive forces. Fine sediment and metal concen tration were not significa ntly different between up and downstream, likely due to deposition within the plunge pool and riprap. Organic matter was higher downstream than upstream, likely due to slower flow allowing deposition. Riprap seems to be the most efficient management practice for trapping organic matter, fine sediments, and metals associated with these materials, but may become highly contaminated due to selective deposition. These areas may also be selectively utilized by macroinvertebr ates due to increased

PAGE 77

77 dissolved oxygen and habitat hetero geneity, thus acting as ecological traps. Further research is needed to test this hypothesis. Concrete aprons and riprap seem to have their benefits and downfalls in terms of minimizing the impact of roads on stream systems. Concrete aprons or the swales associated with these sites minimize eros ion of downstream sections; howe ver, fine sediment and metal contamination were not reduced. Riprap, allows deposition of fine sediment, organic matter, and metals in close proximity to roads; however, this zone may become highly contaminated. Observations during this study suggest future road s that cross streams include vegetated swales as opposed to concrete swales. Vegetation provides infiltration capacity and reduction of peak discharge, thus reducing erosion. Furthermore, vegetated swales encour age filtration processes that reduce the load of fine sediment a nd heavy metals transported to streams. Best management practices evaluated in this study were confined to areas directly downstream of the road crossing ; however, management within the watershed and better road design would greatly minimize the impacts of the roads on streams. Generally, stormwater BMP decisions are made after the roadway has already been designed. A better approach would be to integrate stormwater management features into road designs.

PAGE 78

78 LIST OF REFERENCES Alberti, M., D. Booth, K. Hill, B. Coburn, C. Avolio, S. Coe, and D. Spirandelli. 2007. The impact of urban patterns on aquatic ecosystems : an empirical analysis in Puget Lowland sub-basins. Landscape and Urban Planning 80:345-361. Allan, D. 1995. Stream Ecology: Structure and Function of Running Waters. Chapman and Hall, London, England. Avolio, C. M. 2003. The local impacts of road cr ossings on Puget Lowland creeks. M.S. Thesis, University of Washington. Beasley, G. and P. Kneale. 2002. Reviewing the impact of metals and PAHs on macroinvertebrates in urban watercourses. Progress in Physical Geography 26:236-270. Beasley, G. and P. Kneale. 2004. Assessment of heavy metal and PAH contamination of urban streambed sediments on macroinvertebrates. Water, Air, and Soil Pollution: Focus 4:1573-2940. Bilby R. E., K. Sullivan, and S. H. Dunca n. 1989. The generation and fate of road-surface sediment in forested watersheds in sout hwestern Washington. Forest Science 35:453. Bledsoe, B. P. and C. C. Watson. 2001. Effects of urbanization on channel instability. Journal of the American Water Resour ces Association 37:255-270. Booth, D. B., D. R. Montgomery, and J. Bethel 1997. Large woody debris in urban streams of Pacific northwest. Pages 179-197 in Proceedings of the Conference on Effects of Watershed Development and Management on A quatic Ecosystems. American Society of Civil Engineers, Snowbird, Utah, USA. Brown, S. A. 1982. Prediction of channel bed gr ade changes at highway stream crossings. Transportation Research Record 896:1. Canadian Council of Ministers of the Envir onment. 2001. Canadian Environmental Quality Guidelines. Winnipeg, Manitoba. Chen, M., L. Q. Ma, and W.G. Harris. 1999. Base line concentrations of 15 trace elements in Florida surface soils. Journal of Environmental Quality 28:1173-1181. Climate-zone.com. 2003. Data based on the CIA World Fact Book. Davis, A. P., M. Shokouhian, and S. Ni. 2001. Load ing estimates of lead, copper, cadmium, and zinc in urban runoff from specifi c sources. Chemosphere 44:997-1009.

PAGE 79

79 DeWiest, D. R. and E. H. Livingston. 2002. Flor ida Erosion and Sediment Control Inspectors Manual. Florida Department of Environm ental Protection, Stormwater/ Nonpoint Source Management, in cooperation with Flor ida Department of Transportation. Estebe, A., H. Boudries, J. M. Mouchel, and D. R. Thevenot. 1997. Urban runoff impacts on particulate metal and hydrocarbon concentratio n in River Seine: suspended solid and sediment transport. Water Science and Technology 36:185-193. Forman, R. T. T. and L. E. Alexander. 1998. Ro ads and their major ecological effects. Annual Review of Ecology and Systematics 29:207-231. Forman, R. T. T. and R. D. Deblinger. 2000. The ecological road-effect z one of a Massachusetts (U.S.A.) suburban highway. Conservation Biology 14:36-46. Gidding, E. M., M. I. Hornberger, and H. K. Hadley. 2001. Trace-metal concentrations in sediment and water and health of aquatic m acroinvertebrate communities of streams near Park City, Summit County, Utah. U.S. Geol ogical Survey Water-Resource Investigation Report 01-4213. Gordon, N. D., T. A. McMahon, B. L. Finlayson, C. J. Gippel, R. J. Nathan. 2005. Stream Hydrology: an introduction for ecologists sec ond edition. John Wiley and Sons. West Sussex. Grant, S. B., N. V. Rekhi, N. R. Pise, and R. L. Reeves. 2003. A review of the contaminants and toxicity associated with particles in stor mwater runoff. Prepared for California Department of Transportation; CTSW-RT-D3-059.73.15. Grapentine, L., W. Rochfort, and J. Marsalek. 2004. Benthi c responses to wet-weather discharges in urban streams in southern Ontario. Water Quality Resources Journal of Canada 39:374-391. Horner, R. R. and Mar, B. W. 1983. Guide for assessing water quality impacts of highway operations and maintenance. Transp ortation Resources Record 948:31-39. Horowitz, A. J. 1991. A Primer on Sediment Tr ace-element Chemistry, 2nd ed. Lewis, Chelsea, MI. Jones, J. A., F. J. Swanson, B. C. Wemple, and K. U. Snyder. 2000. Effects of Roads on hydrology, geomorphology, and distribution patche s in stream networks. Conservation Biology 14:76-85. Klein, R. D. 1979. Urbanization and stream qua lity impairment. Water Resources Bulletin 15:948. Kristan, W. B. 2003. The role of habitat selection behavior in population dynamics: source-sink systems and ecological tr aps. Oikos 103:457-468.

PAGE 80

80 Krupadam, R. J., R. Ahuja, and S. R. Wate. 2007. Heavy metal bindi ng fractions in the sediment of the Godavari Estuary, eas t coast of India. Environm ental Modeling and Assessment 12:145-155. Lee, D. K., J. C. Touray, P. Bailif, and J. P. Ildeonse. 1997. Heavy metal contamination of settling particles in a retention pond along the A71 motorway in Sologne, France. Science of the Total Environment 201:1-15. Madej, M. A. 2001. Erosion and sediment delivery following removal of forest roads. Earth Surface Processes and Landforms 26:175-190. Makepeace, D. K., Smith, D. W. and Stanley, S. J. 1995. Urban stormwater quality: summary of contaminant data. Critical Reviews in E nvironmental Science and Technology 25:93 139. Marsalek, J., Q. Rochfort, B. Brownlee, T. Mayer, and M. Servos. 1999. An explanatory study of urban runoff toxicity. Water Science and Technology 39:33-39. May, C. W., R. R. Horner, J. R. Karr, B. W. Mar, and E. B. Welch. 1997. Effects of urbanization on small streams in the Puget Sound Lowland ecoregion. Watershed Protection Techniques 2:483-494. Mcbride, M. 2001. Spatial effects of urbani zation on physical conditions in Puget Sound Lowland streams. Water Resources Series Technical Report N o. 177, Department of Civil and Environmental Engineering, Univer sity of Washington, Seattle, Washington. McBride, M. and D. B. Booth. 2005. Urban impact s on physical stream condition: effects of spatial scale connectivity a nd longitudinal trends. Journal of the American Water Resources Association 41:565580. Montgomery, D. 1994. Road surface drainage, cha nnel initiation, and slope instability. Water Resources Research 30:192. Paul, M. J. and J. L. Meyer. 2001. Streams in the urban landscape. Annual Review of Ecology and Systematics 32:333. Perdikaki, K. and C. F. Mason. 1999. Impacts of road run-off on receiving streams in Eastern England. Pergamon 33:1627-1633. Plumb, R. H. Jr. 1981. Procedures for handling an d chemical analysis of sediment and water samples. Contract EPA-4805572010. U.S. Enviro nmental Protection Agency/Corps of Engineers Technical Committee on Criteria for Dredged and Fill Material, Washington, D.C.

PAGE 81

81 Power, E. A., and P. M. Chapman. 1992. Assessing sediment quality. Pages 1-18 in G. A. Burton Jr., editor. Sediment Toxicity Assessm ent. Lewis Publishers, Inc., Boca Raton, Florida, USA. Rhoads, B. L. and R. A. Cahill. 1999. Geomor phological assessment of sediment contamination in an urban stream system. Applied Geochemistry 14:459-483. Robson, M., K. Spence, and L. Beech. 2006. Stream quality in a small urbanized catchment. Science of the Total Environment 357:194-207. Roberts, C. R. 2002. Riparian tree associati on and storage, transpor t, and processing of particulate organic matter in subtropical streams. Ph.D. Dissertation, University of Florida. Ruediger, B. and B. Ruediger. 1999. The effects of highways on trout and salmon rivers and streams in the Western U.S. Pages 151-160 in Proceedings of the third international conference on wildlife ecology and transportation. Florida Department of Transportation, Tallahassee, Florida, USA. Shelton, L. R. and P. D. Capel. 1994. Guidelines for collecting, processing samples of stream bed sediment for analysis of trace elements and organic contaminants for the National Water-Quality Assessment Program. U.S. Ge ological Survey Open -File Report 94-455, 20. Smith, S. L, D. D. MacDonald, K. A. Keenleyside, C. G. Ingersoll, and J. Field. 1996. A preliminary evaluation of sediment quality a ssessment values for freshwater ecosystems. Journal of Great Lakes Resources 22:624. Sutherland, R. A. 2000. Bed sediment -associated trace metals in an urban stream, Oahu, Hawaii. Environmental Geology 39:611-627. Turer, D., J. B. Maynard, and J. J. Salone. 2001. Heavy metal contamination in soils of urban highways: comparison between runoff and so il concentrations at Cincinnati, Ohio. Water, Air, and Soil Pollution 132:293-314. Van Hassel, J. H, J. J. Ney, and D. L. Garling. 1980. Heavy metals in a st ream ecosystem at sites near highways. Transactions of th e American Fisheries Society 109:636. Wang, L., J. Lyons, P. Kanehl, and R. Banne rman. 2001. Impacts of urbanization on stream habitat and fish across multiple spatial scales. Environmental Management 28:255-266. Waters, T. F. 1995. Sediment in Streams: sources biological effects, a nd control. American Fisheries Society Monograph 7, Bethesda, Maryland, USA. Whitney, E., D. B. Means, and A. Rudloe. 2004. Priceless Florida: Natural Ecosystems and Native Species. Pineapple Press Inc. Sarasota, Florida, USA.

PAGE 82

82 Willson, J. D. and M. E. Dorcas. 2003. Effects of habitat disturbance on stream salamanders: implications for buffer zones and watershed management. C onservation Biology 17:763771. Wood, P. J. and P. D. Armitage. 1997. Biologica l effects of fine sediment in the lotic environment. Environmental Management 21:203-217.

PAGE 83

83 BIOGRAPHICAL SKETCH Shannon McMorrow grew up in Fairfield, Connecticut, where she developed a love f or both math and science. Upon graduation fr om high school, Shannon moved to Gainesville, Florida, to start her college career majoring in zo ology at the University of Florida. During this time she designed and implemented an indepe ndent study project inve stigating the sexual behavior of Florida flag fish. She also decided to minor in environmental sciences to expand her educational scope. Shannon graduated summa cum laude in May of 2004. Shannon next volunteered at a wild life rehabili tation center in Frid ay Harbor, Washington, for three months. There she helped care for a va riety of animals including harbor seals. While in Washington, Shannon realized she wanted to re turn to school to st udy applied ecology. In January 2005, Shannon started graduate school back at the University of Florida at the Center for Wetlands. Shannons research interests included but were not limited to investigation and mitigation of impacts of humans on wildlife and the environment. This interest led to the completion of this research dur ing her graduate career. In addition to scientific research, Shannon wa s able to teach at varying levels during graduate school. She taught an introductory bi ology lab as well as participated in the SPICE program where she co-taught seve nth grade science. This expe rience helped her improve her communication and presentational skills. Upon completion of her masters degree, Shannon hopes to continue research on the impacts of human development on the environment in the hopes of mitigating these effects.