COMPARISON OF THREE SOIL EROSION CONTROL TREATMENTS

A Research Paper
Presented for the
Master of Science in Agriculture and Natural Resources
Degree
The University of Tennessee at Martin

Eatedal Alqusaireen
December 2012
 Acknowledgments

I would like to thank all the people that helped me finish this project. I am grateful to my

advisor, Dr. Sandy Mehlhorn, for her help and support during the project. Special thanks to Dr.

Barb Darroch for her close monitoring and editing of this work. I would also like to thank Dr.

Joey Mehlhorn and Scott Watson, the UT Martin Farm manager, for their help in the preparation

of rills and installation of treatments. I also thank the donors of the materials used in this project.

ii

 Abstract

Soil erosion is a concern that affects agriculture, wildlife and bodies of water. Soil

erosion can be avoided by maintaining a protective cover on the soil, creating a barrier to the

erosive agent, or by modifying the landscape to control runoff amounts and rates. The objective

of this study was to compare the effectiveness of three erosion control treatments on rill erosion

in a hill-slope area. Sweetgum (Liquidambar styraciflua) fruits (sweetgum balls) were compared

to riprap and sod to determine their effectiveness in controlling erosion, compared to a control

(no treatment). A randomized complete block design with three blocks was used. The study was

conducted on a Loring silt loam at The University of Tennessee at Martin campus between

February and October of 2012. Twelve rills were created to simulate erosion channels on a

hillside of 4.3% slope. Visual observations and before-after measurements of the rills were used

to evaluate the erosion levels of each treatment based on changes of the rills shapes and amount

of sedimentation. SAS was used to conduct analyses of variance on before-after measurements of

the rills depths and widths. Visual observations were consistent through all blocks for each

treatment. Results indicated that the sod was the most effective erosion control treatment,

followed by the riprap, and then by the sweetgum balls. Statistical analysis showed a significant

difference in the means among the erosion control treatments for some measurements. There

was significantly (P<0.05) less erosion (as measured by rills depth and width) in the sweetgum

ball treatment than in the control; therefore sweetgum balls were an acceptable treatment for rill

erosion.

iii

Table of Contents

Chapter 1: Introduction.................................................................................................................................1
Objective...................................................................................................................................................2
Chapter 2: Literature Review........................................................................................................................3
Sweetgum balls.........................................................................................................................................7
Sod............................................................................................................................................................9
Riprap......................................................................................................................................................10
Erosion Measurement.............................................................................................................................12
Chapter 3: Materials and Methods..............................................................................................................14
Location..................................................................................................................................................14
Treatment Materials and Experimental Design.......................................................................................14
Experiment Techniques..........................................................................................................................17
Data Collection and Assessment.............................................................................................................21
Chapter 4: Results and Discussion..............................................................................................................22
Visual Observation..................................................................................................................................22
February..............................................................................................................................................22
March..................................................................................................................................................22
April....................................................................................................................................................27
May.....................................................................................................................................................29
June, July and August.........................................................................................................................29
September...........................................................................................................................................29
October................................................................................................................................................29
Statistical Analysis..................................................................................................................................33
Chapter 5: Conclusion.................................................................................................................................40
References...................................................................................................................................................41

iv

List of Figures

Figure 1. The shape of rill erosion................................................................................................................4

Figure 2. Mulch from trimmed small plants and branches...........................................................................8

Figure 3. The spiky fruits produced by sweetgum trees...............................................................................8

Figure 4. Sod ready to be placed...................................................................................................................9

Figure 5. Riprap-Large pieces of limestone that can be used to control erosion........................................10

Figure 6. The location of the rills used in the study as seen from Google earth (3620'56.56" N
8851'41.61" W) ......................................................................................................................................... 15

Figure 7. A Magallen Mobile Mapper 6 GPS unit was used to map the rills used in the study................15

Figure 8. Map of the rills used in the study................................................................................................16

Figure 9. Layout of treatments in the experiment using a randomized complete block design..................16

Figure 10. a. The experiment was located on 4.3% slope; b. A furrower was used to create the rills;
c. Rills were approximately 30.4 cm (1 ft) deep and 60.8 cm (2 ft) wide and d. 6.08 m (20 ft) long........17

Figure 11. Placement of the sweetgum balls in the rill...............................................................................18

Figure 12. a. Geotextile filter placement within the rill, b. Riprap placement within the rill.....................19

Figure 13. Placement of Tifway Bermuda grass sod within the rill............................................................20

Figure 14. The four treatments in the first block at the beginning of the experiment: a. Sweetgum ball;
b.Tifway Bermuda grass sod; c. Control; d. Riprap....................................................................................20

Figure 15. Removing sweetgum balls (a) and filter under riprap (b) at the end of the experiment
(October 2, 2012)........................................................................................................................................20

Figure 16. Change in one sweetgum ball rill from February to June, 2012................................................23

Figure 17. Daily rainfall (inches) in February for Martin, TN (National Climatic Data
Center, 2012)...............................................................................................................................................24

Figure 18. Daily rainfall (inches) in March for Martin, TN (National Climatic Data Center, 2012).........24


Figure 19. Monthly rainfall in inches from February to October, 2012 for Martin, TN.............................25

Figure 20. Change in the control rill during March, 2012..........................................................................26

Figure 21. Change in the riprap rill during March, 2012............................................................................26

Figure 22. Change in the sweetgum balls rills during March, 2012...........................................................27

Figure 23. Change in the Tifway Bermuda grass sod rills during March, 2012.........................................27

Figure 24. Daily rainfall (inches) in April for Martin, TN (National Climatic Data Center, 2012)...........28

Figure 25. Change in all treatments rills as observed on April 8, 2012 a. Erosion at the edges of
sweetgum ball rill, b. Tifway Bermuda grass sod rill, c. Soil on the surface of riprap rill and d.
Sedimentation at the outlet of control rill...................................................................................................28

Figure 26. Daily rainfall (inches) in May for Martin, TN (National Climatic Data Center, 2012). ........... 30

Figure 27. Change in all treatment rills as observed on May 9, 2012: a. Sweetgum ball, b. Tifway
Bermuda grass sod, c. Control and d. Riprap..............................................................................................30

Figure 28. Daily rainfall (inches) in June for Martin, TN (National Climatic Data Center, 2012)............31

Figure 29. Daily rainfall (inches) in July for Martin, TN (National Climatic Data Center, 2012).............31

Figure 30. Daily rainfall (inches) in August for Martin, TN (National Climatic Data Center, 2012)........32

Figure 31. Final appearance of all treatment rills on September 29, 2012: a. Sweetgum ball, b. Tifway
Bermuda grass sod, c. Control and d. Riprap..............................................................................................32

Figure 32. Daily rainfall (inches) in September for Martin, TN (National Climatic Data Center, 2012)...33

Figure 33. Means of the final depth of the erosion control treatment rills..................................................34

Figure 34. Means of the final outlet width of the erosion control treatment rills.......................................35

Figure 35. Means of the difference between initial depth and final depth of the erosion control
treatment rills..............................................................................................................................................35

Figure 36. Means of the difference between final outlet width and initial outlet width of the erosion
control treatments rills ................................................................................................................................ 36

vi


Figure 37. Sweetgum ball rills shape after removing sweetgum balls. a. Rill shape after removal,
b. Sediments around sweetgum balls, and c. Erosion on the edge of rills..................................................37

Figure 38. Riprap rills shape after removing riprap. a. Rill shape, b. Sediment on the filter, and
c. Erosion at the outlet of the rill. ................................................................................................................ 39

Figure 39. Sweetgum ball properties: a. Spikes stick to the ground, and b. decay slowly.........................39

vii

 Chapter 1: Introduction

Soil erosion is the physical wearing of the earths surface by the action of water or wind.

It has been occurring for some 450 million years, since the first land plants formed the first soil.

There are two main types of soil erosion: geological and accelerated soil erosion. Geological soil

erosion happens at the same rate as soil is formed. Accelerated soil erosion is the loss of soil at a

much faster rate than it is formed (Favis-Mortlock, 2005)

Soil erosion is considered a serious problem all around the world. It is detrimental to

topsoil, which contains nutrients and organic matter. Loss of only 1/32 of an inch of topsoil can

represent a 5 ton per acre soil loss (USDA NRCS, 1996). This results in lower sustainability and

lower productive capacity of agriculture. During the last 40 years, nearly one-third of the world's

arable land has been lost by erosion and land continues to be lost at a rate of more than 24.7

million acres per year. In additions, soil erosion is considered to be the largest contributor to non-

point source pollution in the United States according to the federally mandated National

Pollution Discharge Elimination System (USEPA, 1997).

The rate of erosion from all cropland in Tennessee averaged 5.6 tons per acre per year

and 7.7 tons per acre per year on cultivated cropland (Denton, 2000). Compared to five years

earlier, this is down from 7.1 tons on all cropland and 9.1 tons on cultivated cropland. These

erosion levels are about half as high as 20 years ago. Tennessee has the highest rate of erosion of

cultivated cropland among the 50 states (Denton, 2000).

The rate and magnitude of soil erosion is affected by rainfall intensity and runoff, soil

erodibility, slope gradient and length, vegetation, and control treatments. Rill erosion results

when surface runoff concentrates, forming small yet well-defined channels (channels up to 30

cm deep). In many parts of the world, rill and gully erosion is the dominant form of water

erosion. Water erosion is most obvious on steep, convex landscape positions. (Shelton, 1987;

Agassi, 1996).

Soil erosion can be avoided by maintaining a protective cover on the soil, creating a

barrier to the erosive agent, and modifying the landscape to control runoff amounts and rates.

There are numerous treatments, combinations of treatments, and emerging products that may be

suitable for the site of erosion. Common treatments for erosion are mulching, vegetation,

terracing, riprap, matting, retaining walls, and reforestation (Rivas, 2006). There are different

ways to measure soil erosion. Visual, physical, chemical, and biological indicators can be used to

estimate soil surface stability or loss (USDA NRCS, 1996).

Objective

The objective of this study was to compare the effectiveness of three erosion control

treatments on rill erosion in a hill-slope area. Sweetgum (Liquidambar styraciflua) fruit

(sweetgum balls) was compared to riprap and sod treatments to determine their effectiveness in

controlling erosion, compared to a control rill (no treatment).

 Chapter 2: Literature Review

The effects of erosion impact two places: on-site (where the soil has become detached),

and off-site (where the eroded soil goes). Subsequent harm results in damage to plants, animals,

and humans (Mamo and Hain, 2005). Erosion removes topsoil, reduces levels of soil organic

matter, and contributes to the breakdown of soil structure (Figure 1). This creates a less favorable

environment for plant growth. In soils that have restrictions to root growth, erosion decreases

rooting depth, which decreases the amount of water, air, and nutrients available to plants (on-site

effect). Erosion removes surface soil, which often has the highest biological activity and greatest

amount of soil organic matter. Nutrients removed by erosion are no longer available to support

plant growth onsite, but can accumulate in water (off-site effect) where such problems as algal

blooms and lake eutrophication may occur (Favis-Mortlock, 2005). Deposition of eroded

materials can obstruct roadways and fill drainage channels. Sediment can damage fish habitat

and degrade water quality in streams, rivers, and lakes. In addition, blowing dust can affect

human health and create public safety hazards (USDA NRCS, 1996).

The simplest mathematical model for prediction of soil loss is the Universal Soil Loss

Equation (USLE), which has been frequently used around the world since it was developed by

American statistician W. H. Whichmeier in the 1960s. The USLE describes average annual soil

loss rates based on estimated and measured input data. The input data is divided into five

different factors: rainfall erosivity, soil erodibility, topography, crop management and

conservation practice (Agassi, 1996). The Revised Universal Soil Loss Equation (RUSLE),

which is a computerized version of USLE with improvements in many of the aforementioned

factors, was initially released for public use in 1992. RUSLE is used by numerous government

agencies and private businesses and individuals to assess the magnitude of rill

 Figure 1. The shape of rill erosion

Source: http://www.erosionpollution.com

erosion, to pin point situations where erosion is serious, and to guide development of plans to

control soil erosion (USDA ARS, 2009).

The rate and magnitude of soil erosion is affected by rainfall intensity and runoff, soil

erodibility, slope gradient and length, vegetation, and erosion control treatments (Wall et al.,

2003). Both rainfall and runoff factors must be considered in assessing a water erosion problem.

Raindrops on the soil surface can break down soil aggregates. Light soil aggregates can be easily

removed by raindrop splash and runoff water, while large soil aggregates need more raindrop

energy or runoff amounts to be moved. Raindrop splash, which is the movement of soil by

rainfall, is noticeable during short-duration, high-intensity thunderstorms. Runoff can occur

whenever there is excess water on a slope that cannot be absorbed into the soil or trapped on the

surface. Runoff from agricultural land may be greatest during spring months when the soils are

usually saturated, snow is melting and vegetative cover is minimal (Favis-Mortlock, 2005).

Soil erodibility is another factor that controls the rate and magnitude of soil erosion.

Erodibility is an estimate of the ability of soils to resist erosion, based on the physical

characteristics of each soil. Soils with faster infiltration rates, higher levels of organic matter and

improved soil structure have a greater resistance to erosion. The soils on eroded sites tend to be

more erodible than the original topsoils, because of their poorer structure and lower organic

matter content. Slope gradient and length also affect soil erosion: the steeper the slope of a field,

the greater the amount of soil erosion. As the slope length increases, the soil erosion by water

increases due to the greater accumulation of runoff (Agassi, 1996). However, vegetation cover

and plant residue protect the soil from raindrop impact and splash, which slows down the

movement of surface runoff and allows excess surface water to infiltrate. The effectiveness of

vegetative covers depends on the type, extent and quantity of cover. Partially incorporated

residues and residual roots are also important, as these provide channels that allow surface water

to move into the soil (Wall et al., 2003).

Signs of soil erosion include rills or cuts visible on the soil surface after rain or snowmelt,

soil accumulated at the bottom of slopes or depressions, soil on knolls that is lighter in color,

stones visible on hilltops, and crops buried with soil (Shelton, 1987). Soil erosion can be

categorized into several types: splash erosion, sheet erosion, rill erosion, gully erosion, and

tunnel erosion. Water erosion is most obvious on steep, convex landscape positions (Wall et al.,

2003; Agassi, 1996). In many parts of the world, rill and gully erosion are the dominant forms of

water erosion. Rills are shallow drainage lines resulting when surface runoff concentrates,

forming small yet well-defined channels (channels up to 30 cm deep) (Rivas, 2006). They

develop when surface water concentrates in depressions or low points and erodes the soil. Rill

erosion is common in bare agricultural land, particularly overgrazed land, and in freshly

cultivated soil where the soil structure has been loosened. The rills can usually be removed with

farm machinery. Rill erosion is often described as the intermediate stage between sheet erosion

and gully erosion (Heubeck and Schtt, 2007).

 Soil erosion can be avoided by maintaining a protective cover on the soil, creating a

barrier to the erosive agent, and modifying the landscape to control runoff amounts and rates.

Protective cover can absorb some of the energy of the water that causes the erosion. (Mamo and

Hain, 2005). There are numerous treatments, combinations of treatments, and emerging products

that may be suitable to control erosion. The general erosion-control treatments can be

categorized into grade-related, seed, fertilizer, soil amendments, soil stabilizers and tackifiers,

mulch, rolled erosion-control products, hard armor, and soil bioengineering. Common treatments

for erosion are mulching, vegetation, terracing, riprap, matting, retaining walls, and reforestation

(Rivas, 2006).

One of the most common erosion control treatments is mulching. Mulching is an

effective and cost efficient management practice for stabilizing construction sites, eroded banks,

and topdressing buffer plantings. The primary function is to protect the underlying soil from

erosion and keep soil out of a water body. Erosion control mats (ECM) were developed to

protect soil from erosion in areas that receive high traffic, exposure to the elements or occur on

steep slopes. One key to the effectiveness of ECM is the size and shape of the material. When

applied, the long and fairly thin material essentially weaves itself together and creates a kind of

blanket over the soil (USDA NRCS, 2012a). ECM trap plenty of moisture, similar to standard

bark mulch. Mulching treatments have a wide range of material application. The materials used

depend on the availability, cost, appearance and effect of materials on the environment and can

include organic residues, compost, rubber mulch, and plastic mulch. Organic residue might be

grass clippings, leaves, hay, straw, shells, woodchips, wool, and animal manure (Dehan 1996).

Mulch material varies from artificial material to plant residue. Figure 2 shows trimmed small

plants and branches that are used as mulch. Some fruits of hardwood trees, such as sweetgum

fruits, have the potential to be used as a mulch material.

Sweetgum balls

Sweetgum (Liquidambar styraciflua) is one of the most common hardwoods in the

eastern United States (USDA NRCS, 2012b). It is a shade tree that can be found from

Connecticut to Florida and in the mountains of Central America ranging from Mexico to

Panama. The trees range west all the way to Texas, Iowa and Oklahoma. Fruits of sweetgum are

spiky green balls that turn brown with age (Figure 3). They are just over one inch wide and

dangle on a long stalk. The sweetgum trees are known for their porcupine-like fruit or the

sweetgum pods that are often called Sweetgum balls. The mature spiny sweetgum balls remain

on the tree branches through the winter. They drop off intermittently in the spring, and need to be

raked up for several months. Many consider it a major annoyance in the Southern US, where the

trees are plentiful (USDA NRCS, 2012b). Awkward to feel underfoot, the sharp, spiky brown

orbs with stems attached cover otherwise tidy lawns and driveways in this part of the country and

fall into the street to be smashed flat by cars. They can also cause damage to lawn mowers (Jett,

2006). The sweetgum balls can be considered as a candidate for mulching material. Some

gardeners use them to help fill the bottom third of large flower pot containers to save potting soil

and to allow better drainage. Many people use them to keep cats and dogs from ruining their

landscapes. The seed pods take so long to decompose that they can be used as a soil amendment.

For lawns that are mostly clay or rocky, mixing the soil with sweetgum balls can help keep the

soil loose and aerated (Jett, 2006).

 Figure 2. Mulch from trimmed small plants and branches.

Source: http://www.cuckooforcoconuts.com

Figure 3. The spiky fruits produced by sweetgum trees.

Source: http://en.wikipedia.org

Sod
One of the most effective erosion control materials available is grass sod. Sodding is a

permanent form of erosion control that involves laying a permanent cover of grass sod on

exposed soils. Because of the high density of grass blades, thick root mass and total weight, sod

stops erosion immediately. In addition to stabilizing soils, sodding can reduce the velocity of

storm water runoff. Sodding can provide immediate vegetative cover for critical areas and

stabilize areas that cannot be vegetated by seed. It can also stabilize channels or swales that

convey concentrated flows and can reduce flow velocities. Selection of grass in the sod is

primarily determined by region, availability, and intended use. For example, in Alabama, the

choice of species is limited to Bermuda grass (Cynodon dactylon), zoysia (Zoysia), centipede

(Eremochloa ophiuroides), St. Augustine (Stenotaphrum secundatum), tall fescue (Festuca

arundinacea), and bahiagrass (Paspalum notatum ) (Pitt, 2003). According to a project at The

Pennsylvania State University in 1986, grassed areas established with turfgrass sod are up to 15

times more effective in controlling runoff than seed established grass, even after three years

(Beard and Green, 1994). Figure 4 shows a pallet of sod ready to be used in landscaping or

erosion control.

Figure 4. Sod ready to be placed.

Source: http://www.sodding.com

Riprap

An efficient erosion control treatment for sloped sites (such as river banks, shorelines,

creeks, highway ditches, etc) is riprap. Riprap is permanent erosion control a resistant

protective layer of rock intended to prevent soil erosion in areas of concentrated flow, turbulence

or wave energy. There are many specifications for riprap. The amount and size of rock needed

depend on each site's exposure to erosion. Many states recommend using 30.48 cm (12-inch)

diameter, or larger, natural rock for most moderate erosion areas. Large pieces of limestone are

commonly used to control erosion (Figure 5) (PESC, 2010). A filter must be placed under the

riprap to prevent water from removing the underlying soil material through the voids in the

riprap. The filter material under the rock can be fabric, gravel, crushed stone or small rock. The

size of the filter material depends on the site, but it should be smaller than the riprap. If gravel or

other material isn't available, a filter cloth can be used (Best Management Practices Manual.

2010).

Figure 5. Riprap-Large pieces of limestone that can be used to control erosion.

Source: https://www.landscapemulch.com

10

 There are many soil erosion treatment studies that are found in the literature. Treatments

vary based on type of soil, topography, climate, and land use. Materials used depend on type and

abundance of useful raw material.

Benik et al. (2003) applied five treatments on a slope of a newly constructed highway:

wood fiber blanket, straw with coconut blanket, straw blanket, a bonded-fiber matrix, and disk-

anchored straw mulch. The largest soil erosion was observed in a bare area with no treatment.

The soil erosion there was ten times more than that from the straw-mulch treatment.

China is one of the countries that suffer most from erosion. The Chinese have conducted

a large amount of research and work in this area. For example, Xu et al. (2010) evaluated the

effects of revegetation on eroded soil biochemical and biological properties in China. Artificial

revegetation can effectively enhance the productivity of degraded soil caused by erosion in

subtropical areas. Six artificial revegetation treatments were used in the experiment: crenate

gugertree (20 years), Chinese fir (20 years), Chinese fir (12 years), grapefruit (12 years), annual

ryegrass (12 years) and the sixth was without treatment. Among the species used in the study,

crenate gugertree and ryegrass enhanced soil biological properties better than the other species.

Faucette et al. (2007) examined different erosion control practices such as wood mulch

blankets, yard waste compost, blending wood and compost at different percents, and straw

blanket with polyacrylamide (PAM). Faucette et al. (2007) concluded that the greater the

percent of compost used in an erosion control blanket, the lower the total runoff and slower the

rate of the runoff. In addition, the compost blanket was the best erosion control practice,

followed by the blended compost and wood blanket.

11

Erosion Measurement

There are many reasons to measure erosion: to assess erosion for an erosion inventory,

for scientific erosion research, to develop and evaluate erosion control technology, to develop

erosion prediction technology, to allocate conservation resources and to develop policies and

regulations (Stroosnijder, 2003).

Soil erosion can be measured in different ways. Visual, physical, chemical, and biological

indicators can be used to estimate soil surface stability or loss (USDA NRCS, 1996). Visual

indicators include comparisons of aerial photographs taken over time, presence of moss and

algae (crypotogams), presence of crusts in desert or arid soils, changes in soil horizon thickness,

and deposition of soil at field boundaries. Physical indicators include measurements of aggregate

stability, and increasing depth of channels, rills, and gullies. Chemical indicators of erosion

include decreases in soil organic matter content, increases in calcium carbonate content at the

surface (provided large amounts of calcium carbonate exist in subsurface layers) and changes in

cation-exchange capacity (CEC). Finally, the biological indicators of erosion are decreased

microbial biomass, lower rate of respiration in the soil, and slower decomposition of plant

residues (USDA NRCS, 1996).

Shit and Maiti (2012) designed a facility to quantify the effects of grass roots on the

erodibility of lateritic topsoil by concentrated flow in Paschim Medinipur, West Bengal, India.

The slope gradient and flow rate were controlled. The concentrated flow erosion was tested for

three lateritic topsoils (bare, scattered grass-root-permeated and densely grass-root-permeated)

and exposed to two slopes (25 and 35%) by using a hydraulic flume. Their findings showed that

there was a significant negative exponential relationship between relative soil detachment and

12

root density. They concluded that the formula based on root density has the potential to improve

methodology for assessing soil detachment rate in concentrated flow for lateritic topsoil.

Geographic information system (GIS) techniques can also be used to evaluate and

monitor soil erosion. Digital analyses of spatial and climatic data were used to estimate soil

erosion on a basin wide scale for Mourganis catchment, Kalabaka area, in cental Greece (Tsimi

et al. 2012). The GIS database included the following geospatial data: land elevation, cities,

river network, road network, geology, and land use. The spatial features were digitized:

catchment boundaries, streams, geology formations, elevation contours, and elevation. The input

data included slope angle of the ground surface, elevation, rainfall, land cover, and geology. The

digital elevation model was constructed by combining elevation information from photogram-

metrically extracted contours and elevation points. The results from the soil erosion model were

compared with a photo-interpretation from Google Earth (Tsimi et al. 2012).

13

 Chapter 3: Materials and Methods

Location

This project was conducted at The University of Tennessee at Martin campus between

February and October of 2012. The soil type for the study area was a Loring silt loam with 2 to 5

percent slopes; it is considered to be an eroded soil (USDA NRCS, 2012c). The selected study

area is located at 3620'56.56" N 8851'41.61" W north of campus next to the motor pool

(Figure 6).

A geographic information system (GIS) was used to map the 12 rills used in the

experiment. The global positioning coordinates of the rills were measured using a Magallen

Mobile Mapper 6 (Figure 7) with Arc Pad software. The data were taken on March 25, 2012. The

XY coordinates of the rills were measured using points features. Each rill has four points located

at each corner. After entering these feature points into the GIS software, they were saved as

polyline features and edited. A map of the experiment was created using ArcGIS (Figure 8).

Treatment Materials and Experimental Design

Four treatments were used: sweetgum ball mulch, Bermuda grass sod, riprap, and no-

treatment as a control. Sweetgum balls were collected locally from nearby trees and farms and

stored in bins near the site. A net was used to wrap the sweetgum balls inside the rills to prevent

sliding or moving. Riprap was donated and delivered to the site. The average diameter of the

riprap was 25.4 cm (10 inches) and it was composed of limestone rock. Geotextile fabric filter

was placed below the riprap to prevent erosion below the rocks. Tifway Bermuda grass sod was

donated by McCurdy Sod Farm and was placed within the rills. A randomized complete block

design (RCBD) with three blocks was used for this project (Figure 9).

14

 Motor
Pool

Figure 6. The location of the rills used in the study as seen from Google earth (3620'56.56" N
8851'41.61" W)

Figure 7. A Magallen Mobile Mapper 6 GPS unit was used to map the rills used in the study.

15

 Figure 8. Map of the rills used in the study.

Gumball

Control

Gumball
Riprap

Gumball

Control
Control

Riprap
Riprap
Sod

Sod
Sod

Block 1 Block 2 Block 3

Direction
Gradient

Figure 9. Layout of treatments in the experiment using a randomized complete block design.

16

Experiment Techniques

The experiment was located on a hillside with an average slope of 4.3% (Figure 10a).

Erosion channels were simulated on February 24, 2012 by creating several rills along the hill

slope using a furrower attached to a tractor (Figure 10b). Each rill was 30.4 cm (1ft) deep and

60.8 cm (2ft) wide (Figure 10c). The rills were 6.08 m (20 feet) long (Figure 10d).

a b

c d

Figure 10. a. The experiment was located on 4.3% slope; b. A furrower was used to create
the rills; c. Rills were approximately 30.4 cm (1 ft) deep and 60.8 cm (2 ft) wide and
d. 6.08 m (20 ft) long.

17

 Sweetgum balls were collected locally and held in place using a plastic net fixed to the

bed of the rill. The sweetgum balls were placed by putting a net on the ground of the rill and

filling with more than 0.17 m3 (six cubic feet) of sweetgum balls for each rill. The net was then

tied together, and four stakes were put along the rill to hold the netted sweetgum ball to the

ground (Figure 11).

Riprap treatment installation followed Tennessee division of water control and United

States army corps of engineers guidelines (Tennessee Department of Transportation, 2001). The

average diameter of rocks used along the rill was 2504 cm (10 inches). Geotextile fabric filter

from Preen was placed below the rocks to prevent erosion below the rocks (Figure 12).

Netaround
Sweetgumball

Figure 11. Placement of the sweetgum balls in the rill.

18

 a b

Figure 12. a. Geotextile filter placement within the rill, b. Riprap placement within the rill.

The third treatment in the experiment was the use of Tifway Bermuda grass sod that was

laid along the bed of the rill (Figure 13). Three wooden stakes were placed through the rill to

hold the sod in place until a root system could be established.

Figure 14 shows all four treatments in the first block at the beginning of the experiment

(on February 24, 2012).

The final step in the project was removing two treatments in all blocks at the end of the

experiment. Sweetgum balls and riprap were removed from the rills on October 2, 2012 to

observe the change of the rills bed. Figure 15 shows removing the Sweetgum ball and the riprap

with the fabric filter.

19

 Figure 13. Placement of Tifway Bermuda grass sod within the rill.

a b c d

Figure 14. The four treatments in the first block at the beginning of the experiment:
a. Sweetgum ball; b.Tifway Bermuda grass sod; c. Control; d. Riprap.

a b

Figure 15. Removing sweetgum balls (a) and filter under

riprap (b) at the end of the experiment (October 2, 2012).

20

Data Collection and Assessment

Data were collected to assess the erosion caused by rainfall and runoff from February to October

of 2012. Rill erosion treatments were observed on a weekly basis and immediately after rainfall.

Visual observation for erosion were conducted and documented. Observations included taking

photos for each treatment at critical locations along the rill. The treated rills were compared to

the control rill (no treatment) for all blocks. In addition to visual observations, measurements of

depth and width of the rills were taken at both beginning and end of the project period. The

width of the rills was measured at both inlet and outlet of the rill. The depth of the rill was

measured at five different sections along the rill and then averaged. The measurements were

taken for sweetgum ball, riprap, and control rills. No measurements were taken for Tifway

Bermuda grass sod because it was hard to remove at the end of the project. ANOVA tests for

RCBD were run for all measured variables for sweetgum ball, riprap, and control rills. The tested

variables were initial depth, final depth, initial inlet width, final inlet width, initial outlet width,

final outlet width, the difference between initial depth and final depth, the difference between

final inlet width and initial inlet width, and the difference between final outlet width and initial

outlet width.

21

 Chapter 4: Results and Discussion

Visual Observation

Visual assessment to evaluate erosion levels was based on the observed change of the

rills shapes. Also, assessment was based on the relative amount of sedimentation at the outlet of

the rills for each treatment compared to that of the control rills (no treatment). The final

assessment for each rill treatment was rated from high to no change relative to the erosion level

of the control rill. The results were summarized by treatment in each block separately and an

average for all blocks during the whole period. All treatments had consistent changes throughout

all blocks. The block number in the discussion will not be mentioned because all the blocks had

similar changes.

The photos in Figure 16 show an example of the changes in sweetgum ball rills over the

data collection period. Pictures were taken at different time periods after heavy rain falls. Similar

photos were used to compare the other treated rills.

February

Since rills were prepared on February 24th, there was no change in the rills shapes and

sedimentation levels during February, although there was one heavy rainfall during that period

(Figure 17).

March

March was the critical period for this study where most changes in the rills occurred,

especially for the control rill. There were three storms with heavy rainfall and more than four

times of low intensity rainfall or scattered showers (Figure 18), and the maximum rainfall

occurred during this month (5.57 inches) (Figure 19). The control treatments had the biggest

change in the shape of the rill and in the amount of sedimentation. By the first week of March,

22

 Feb 29 Mar 17 Mar 6 Mar 10

Mar 25 Apr 8 Apr 17 May 1

May 9 May 15 Jun 3

Figure 16. Change in one sweetgum ball rills from February to June, 2012.

23

 Figure 17. Daily rainfall (inches) in February for Martin, TN

(National Climatic Data Center, 2012).

Figure 18. Daily rainfall (inches) in March for Martin, TN

(National Climatic Data Center, 2012).

24

Figure 19. Monthly rainfall in inches from February to October, 2012 for Martin, TN.

the sedimentation was obvious in the outlet of the control rill (Figure 20d). Also, on March 17th

and 25th, sedimentation increased and a small groove was formed in the middle of the rill

(Figure 20c,d).

The rate of the erosion in the other treatments was not as obvious as in the control rills.

The erosion rate in the riprap rill was low in March. There were only small changes at the outlet

of the rills (Figure 21). The same observation was noted for the level of erosion at the sweetgum

ball rills (Figure 22). In the sweetgum balls rills, there were some grooves at the outlet of the rills

by the end of March. There were no signs of erosion in the sod rills in all blocks and they did not

show any changes to the shape of the rills nor any sedimentation (Figure 23).

25

 Mar6 Mar10

Mar17 Mar25

Figure 20. Change in the control rill during March, 2012.

Mar6 Mar25

Figure 21. Change in the riprap rill during March, 2012.

26

 Mar6 Mar25

Figure 22. Change in the sweetgum balls rills during March, 2012.

Mar6 Mar25

Figure 23. Change in the Tifway Bermuda grass sod rills during March, 2012.

April

April had one heavy rainfall storm and three events with low rainfall intensity (Figure

24). The total rainfall in April was 2.1 inches (Figure 19). There were only slight changes in all

treated rills in April (Figure 25). In the control rill, sedimentation increased and the depth of the

grooves in the rills beds also increased. There were some changes in the sweetgum ball rills

with some erosion noted, especially at the edges. The soil on the surface of riprap rocks was

evidence of some erosion, especially at the edges of the rills. No observed changes were noticed

in the sod rills.

27

 Figure 24. Daily rainfall (inches) in April for Martin, TN

(National Climatic Data Center, 2012).

a b

c d

Figure 25. Change in all treatments rills as observed on April 8, 2012 a. Erosion at
the edges of sweetgum ball rill, b. Tifway Bermuda grass sod rill, c. Soil on the
surface of riprap rill and d. Sedimentation at the outlet of control rill.

28

May

Three rainfall events occurred in May with low intensity (Figure 26). The rate of erosion

was high, similar to that in April. The erosion of the sweetgum ball rill at the edge and outlet

was noticeable (Figure 27). The sod rills had no signs of erosion. On the other hand, there was a

change at the bottom part of the edges of the riprap rills. One third of the control rill was filled

with sediment and the groove in the middle of the rill was bigger than the previous month.

June, July and August

During June and July, there were scattered showers of rainfall with low intensity (Figure

28 and 29). August had few events of low intensity rainfall and one medium intensity rainfall

event that occurred at the end of August (Figure 30). The mean temperature in the Martin area

was 5.8 F above normal for the entire summer. This broke the previous record set in 1936.

Precipitation was 12.41" below normal. It was third driest summer since 1936 (Tennessee

Climatological Service, 2012). There was no change in the rills after the month of May until the

end of August.

September

In September there were minor changes in the rills shapes. The final shape of the rills

can be seen in Figure 31. Weeds had grown tall during the summer and there was a need for

application of weed killer and trimming. Because of the weeds, changes in the rills appearance

were difficult to detect even though there were high rainfall intensity events during this month

(Figure 31).

October

The data collection process ended on October 2nd where the sweetgum ball and the riprap

were removed.

29

 Figure 26. Daily rainfall (inches) in May for Martin, TN

(National Climatic Data Center, 2012).

a b

c d

Figure 27. Change in all treatment rills as observed on May 9, 2012:

a. Sweetgum ball, b. Tifway Bermuda grass sod, c. Control and d. Riprap.

30

Figure 28. Daily rainfall (inches) in June for Martin, TN

(National Climatic Data Center, 2012).

Figure 29. Daily rainfall (inches) in July for Martin, TN

(National Climatic Data Center, 2012).

31

 .
Figure 30. Daily rainfall (inches) in August for Martin, TN
(National Climatic Data Center, 2012).

a b

c d

Figure 31. Final appearance of all treatment rills on September 29, 2012:
a. Sweetgum ball, b. Tifway Bermuda grass sod, c. Control and d. Riprap.
32

 Figure 32. Daily rainfall (inches) in September for Martin, TN

(National Climatic Data Center, 2012).

The following discussion of the results is a comparison between the data collected in

February (beginning of study) and that in October (ends of study). This comparison was done by

statistical analysis. Sod couldnt be removed and showed no change in the rills shape, which

means no observed erosion happened or the erosion was controlled.

Statistical Analysis

The results of the ANOVA tests indicated that four variables had a significant (P<0.05)

difference (Table 1). These variables were the final depth (Figure 33), final outlet width (Figure

34), the difference between initial depth and final depth (Figure 35), and the difference between

final outlet width and initial outlet width (Figure 36). Sweetgum ball rills had less erosion than

the control rills but more erosion than the riprap rills. The mean of the difference between final

outlet width and initial outlet width for sweetgum ball (4.1 cm) was significantly less than the

control (10.2 cm) but not significantly different from the mean difference for riprap (6.1 cm;

Table 1, Figure 36 ). Also, there was no difference between initial depth and final depth for

33

Table 1. Means of all variables for the sweetgum ball, riprap, and control treatments.

Initial Final inlet Final outlet

Initial Final Initial Final
Initial Final depth- width width-
Treatments inlet inlet outlet outlet
depth depth Final initial inlet initial outlet
width width width width
depth width width
-------------------------------------------------------- cm -------------------------------------------------------
Sweetgum ball 21.7a* 16.7ab 42.7a 46.7a 43.7a 47.8c 5.0 a 4.1 a 4.1 b

Riprap 19.8 a 19.8 a 53.9a 56.9a 48.8a 54.9b 0.0 b 3.1 a 6.1 b

Control 20.8 a 14.5 b 44.7a 48.8a 49.8a 59.9a 6.3 a 4.1 a 10.2 a

*Within the columns, means followed by the same letter are not significantly different (at P =
0.05) using Ryan-Einot-Gabriel-Welsch multiple range test.

Figure 33. Means of the final depth of the erosion control treatment rills.
Means labeled by the same letter are not significantly different (P=0.05)
using Ryan-Einot-Gabriel-Welsch multiple range test.

34

 Figure 34. Means of the final outlet width of the erosion control treatment
rills.Means labeled by the same letter are not significantly different
(P=0.05) using Ryan-Einot-Gabriel-Welsch multiple range test.

Figure 35. Means of the difference between initial depth and final depth of the erosion
control treatment rills. Means labeled by the same letter are not significantly
different (P=0.05) using Ryan-Einot-Gabriel-Welsch multiple range test.

35

 Figure 36. Means of the difference between final outlet width and initial outlet width of
the erosion control treatments rills. Means labeled by the same letter are not significantly
different (P=0.05) using Ryan-Einot-Gabriel-Welsch multiple range test.

riprap, and there was a difference in depth for both sweetgum ball and control rills. The

difference in depth for sweetgum ball (5 cm) was less than control rill (6.3 cm) but these means

were not significantly different (Table 1, Figure 35).

Control rills showed a difference in the shape of the rills and the amount of sediments

when compared at the beginning and at the end of this project. The average depth of the control

rill was 14.5 cm (0.48 ft) at the end of the project compared to the 20.8 cm (0.68 ft) at the

beginning of the project period (Table 1).

The pictures in Figure 37 show the shape of the bed of the rills after removing the

sweetgum balls. The average depth of the rills after removing the sweetgum balls was decreased

in all blocks. For example, the average depth of the rill before placing the sweetgum ball was

21.6 cm (0.7 ft) and the outlet depth of the rill after removing the sweetgum balls was 16.7 cm

(0.55 ft; Table 1). There was some sediment around the removed sweetgum balls as shown in

36

 a b c

Figure 37. Sweetgum ball rills shape after removing sweetgum balls. a. Rill shape after
removal, b. Sediments around sweetgum balls, and c. Erosion on the edge of rills.

Figure 37b. Also, Figure 37c shows the breakdown or the erosion of the soil at the edges of the

rill. There was some erosion at edges of sweetgum ball rills where soil was transported to the

bottom of the bed.

The pictures in Figure 38 show the shape of the bed of the rills after removing the riprap.

There was sediment in the riprap fabric filter after removing the riprap (Figure 38b). The width

of the riprap rill was increased at the outlet from 48.8 cm to 54.9 cm (1.6 ft to 1.8 ft; Table 1).

Figure 38c shows the erosion on the edge of the rill at the outlet. There was no difference in the

depth and the shape of the bed of the riprap rill. The amount of sediment in the fabric filter

indicates that erosion occurred at the edges since the depth had not changed.

The observed and measured erosion changes that occurred in the sweetgum ball rills were

smaller than the changes in the control (no treatment) rills. Similar studies in the literature

confirmed the effectiveness of mulch material similar to sweetgum balls as an erosion control

37

treatment. Benik et al. (2003) observed that soil erosion was ten times larger in bare areas than in

straw-mulch treated areas. Also, Faucette et al. (2007) concluded that the greater the percent of

compost used in an erosion control blanket, the lower the total runoff and slower the rate of the

runoff, which leads to reduced erosion.

Sweetgum balls had stuck to the bed of the rill and stayed in its place during the entire

period of the experiment. The spikes on the Sweetgum ball helped the netted mulch to stick to

the ground and prevented it from sliding (Figure 39a). After more than six months, the

Sweetgum balls were not decayed (Figure 39b), which proved the property of slow decay (Jett,

2006). Sweetgum balls showed desirable characteristics as a mulch material suitable for erosion

control. The sweetgum ball treatments had an advantage over riprap because agricultural

operations can be done on top of the treatment. This can include such operations as mowing and

placing desired vegetation cover on site. Also, grass was able to grow up through the sweetgum

balls to further stabilize the soil and help reduce erosion.

At the end of this study, the rate of erosion in all rill treatments was ranked from highest

to lowest as follows: control, sweetgum ball, riprap and finally sod.

38

 a b c

Figure 38. Riprap rills shape after removing riprap. a. Rill shape, b. Sediment on the filter, and
c. Erosion at the outlet of the rill.

a b

Figure 39. Sweetgum ball properties: a. Spikes stick to the ground, and b. decay slowly.

39

 Chapter 5: Conclusion

In conclusion, this study explored the effectiveness of three erosion control treatments. A

new proposed treatment was the use of sweetgum balls as a mulch-based erosion control

treatment. It was compared to riprap, sod, and to the control (no treatment). All three erosion

control treatments were effective in reducing erosion compared to the control (no treatment).

According to the visual observations of the rills, the sod was the most effective erosion treatment

followed by the riprap, and then followed closely by the sweetgum balls. The measurements of

rill depth and width before and after the installation of the treatments were tested statistically

using SAS. The statistical results showed a significant difference among the erosion control

treatments. Sweetgum ball treatment was considered as an acceptable and efficient treatment to

control erosion, similar to the other treatments in this study.

40

 References

Agassi, M . 1996. Soil Erosion, Conservation, and Rehabilitation. Marcel Dekker,Inc. New
York, NY. Available at http://www.tucson.ars.ag.gov. (Accessed Feb 4, 2012).

Beard, J. B., and R. L. Green. 1994. The role of turfgrasses in environmental protection and
their benefits to humans. J. Environ. Qual. 23:452-460.

Benik, S. R., B. N. Wilson, D. D. Biesboer, B. Hansen, and D. Stenlund. 2003. Evaluation of

erosion control products using natural rainfall events. J. of Soil and Water Conservation
58( 2).

Best Management Practices Manual. 2010. Riprap. Hamilton County Tennessee. Available at:
http://www.hamiltontn.gov. (Accessed Feb 4, 2012).

Dehan, R. 1996. Mulching for Erosion Control. Government of Prince Edward Island. Canada.
Available at: http://www.gov.pe.ca Last updated Jul, 2004. (Accessed Feb 3, 2012).

Denton, P. 2000. Tennessee soil erosion picture fuzzy. Southeast Farm Press. Available at:
http://southeastfarmpress.com. (Accessed Nov 12, 2012).

Favis-Mortlock, D.T. 2005. Soil Erosion. Available at: http://www.soilerosion.net. Last updated
Feb, 2008. (Accessed Sep 3, 2012).

Faucette, L. B., J. Governo, C. F. Jordan, B. G. Lockaby, H. F. Carino, and R. Governo. 2007.

Erosion control and storm water quality from straw with PAM, mulch, and composite
blanket of varying particle size. J. Soil Water Conserv. 62(6).

Jett, J. W. 2006. Sweetgum ball control. Extension Service, West Virginia University. Available
at: http://www.wvu.edu. (Accessed Nov 12, 2012).

Heubeck, Ch., and B. Schtt. 2007. Types of erosion. GeoLearning project, Department of Earth
Sciences, Freie Universitt Berlin. Available at: http://www.geo.fu-berlin.de.last
Updated 01/25/2008. (Accessed Feb 4, 2012).

Mamo, M., and P. Hain. 2005. Erosion Control Measures. J. of Nat. Resour. and Life Sci. Edu.
Available at: http://passel.unl.edu. ) (Accessed Feb 2, 2012).

National Climatic Data Center, 2012. Available at: http://www7.ncdc.noaa.gov. (Accessed Nov
15, 2012.

PESC (Permanent erosion and sediments control). 2010. Design guide lines. Montana
departments of transportation. Available at:http://www.mdt.mt.gov. (Accessed Feb 2,
2012).

41

Pitt, R., 2003.Construction site erosion and sediment controls. Planning, design and performance,
Chapter 7: Vegetation for erosion and sediment control. Available at:
http://rpitt.eng.ua.edu. (Accessed Sep 3, 2012).

Rivas, T. 2006. Erosion control treatment selection guide. USDA Forest Service. Available at:
http://www.fs.fed.us. (Accessed Nov 12, 2012).

Shelton, I.J. 1987. Soil Erosion Still a Threat to Our Soils. Ontario Ministry of Agriculture,
Food, and Rural Affairs. Available at: http://www.omafra.gov. (Accessed on Feb 2,
2012).

Shit P. K., and R .Maiti. 2012. Effects of plant root density on the erodibility of lateritic topsoil
by simulated flume experiment. International Journal Forest, Soil Erosion, 2(3): 137-
142.

Stroosnijder, L., 2003. Measurement of erosion: is it possible? In: D. Gabriels and W. Cornelis.
Available at: http://www.ldd.wur.nl. (Accessed Sep 3, 2012).

Tennessee Climatological Service. 2012. Available at: http://climate.tennessee.edu.(Accessed

Sep 3, 2012).

Tennessee Department of Transportation, Design Division, Design Guidelines. 2001. Available

at: http://www.tdot.state.tn.us. Last update May 2011. ( Accessed Nov 12,2012).

Tsimi, C. A. Ganas, D. Dimoyiannis, S. Valmis, and E. Lekkas. 2012. Catchment-wide estimate

of single storm interrill soil erosion using an aggregate instability index model based on
geographic information system. Nat. Hazards. 62:863-875. DOI 10.1007/s1 1069-0114-8.
Available at: http://www.elekkas.gr

United States Department of Agriculture (USDA), Agriculture Research Service (ARS). 2009.
USLE History. Available at: http://www.ars.usda.gov. (Accessed Nov 12, 2012).

USDA Natural Resources Conservation Service (NRCS). 1996. Soil Quality Resource Concerns:
Soil Erosion. Available at: http://soils.usda.gov. (Accessed Feb 3. 2012).

USDA Natural Resources Conservation Service (NRCS). 2012a. Erosion Control Mats
(ECM).Available at: http://www.fs.fed.us. . (Accessed Nov 12, 2012).

USDA Natural Resources Conservation Service (NRCS). 2012b. Plant Profile. Available at
http://plants.usda.gov. (Accessed Sep. 3, 2012).

USDA Natural Resources Conservation Service. 2012c. Web Soil Survey. Available at:
http://websoilsurvey.nrcs.usda.gov. (Accessed Nov 13, 2012).

42

United States Environmental Protection Agency (USEPA). 1997. Innovative Uses of Composite:
Erosion Control, Turf Remediation and Landscaping. EPA 530-F-97-043. Washington,
DC: USEPA.

Wall, G., C.S. Baldwin , and I.J. Shelton. 2003. Soil Erosion - Causes and Effects. Ontario
Ministry of agriculture, food, and rural affairs. Available at: http://www.omafra.gov.
(Accessed Feb 2, 2012).

Xu, Q., P. Jiang, and H.Wang. 2010. Improvement of biochemical and biological properties of
eroded red soil by artificial revegetation. J. Soils Sediments 10:255262.

43