Performance of Alternative Straw Mulch Binding Agents Thesis (TPD1801069)

A significant portion of testing was completed using a wind tunnel to compare failure wind speeds of these products at different application rates and under a range of conditions. In general, wet straw was resistant to failure up to the maximum wind speed of 72-80 km h^-1 (45- 50 mi h^-1) even without tackifier. Products tested under gusty wind conditions (higher wind speed acceleration) failed at lower wind speeds than under steady conditions. Tackifier application rates below those recommended by the manufacturer were significantly less effective at withstanding wind, while applications beyond recommended did not always significantly improve stability for most products. Hydromulch products, made of paper and/or wood fiber, were as effective as asphalt in resisting failure, and some have a much lower material cost. Two smaller studies, outdoor and greenhouse, were conducted to determine the effect of tackifier products on grass growth, for species used by the North Carolina Department of Transportation. Neither study indicated negative impacts on grass establishment when these products were applied to straw. Overall, the lower cost hydromulches at 1120 kg ha^-1 (1,000 lb ac^-1) and the plant-based product, plantago, at 224 kg ha^-1 (200 lbs ac^-1) would be well suited for replacing emulsified asphalt on construction sites during the revegetation phase.

Performance of Alternative Straw Mulch Binding Agents

by

Maria Ann Polizzi

A thesis submitted to the Graduate Faculty of North Carolina State University

in partial fulfillment of the requirements for the degree of Master of Science

Soil Science

Raleigh, North Carolina 2018

APPROVED BY:

Dr. Richard McLaughlin                                 Dr. Grady Miller

Dr. Deanna Osmond

Biography:

Maria Ann Polizzi was born May 3, 1994 in Charlotte, North Carolina but lived much of her life in Weddington, a smaller suburb surrounding Charlotte. After graduating from Weddington High School, Maria attended North Carolina State University. Graduating cum laude, she received her B.S. in Environmental Technology and Management with a double minor in Biology and Soil Science in December, 2015. Following graduation, Maria decided to further her education and began graduate school in Soil Science at N.C. State under the direction of Dr. Richard A. McLaughlin.

Table of Contents:

LIST OF TABLES……………………………………………………………………………………………………………….. v

LIST OF FIGURES……………………………………………………………………………………………………………. vii

CHAPTER 1: LITERATURE REVIEW……………………………………………………………………………………… 1

Introduction…………………………………………………………………………………………………………. 1

Vegetation Establishment……………………………………………………………………………………… 2

The Effect of Increased Vegetation…………………………………………………………………………. 4

Temporary Erosion Control……………………………………………………………………………………. 4

Tackifier Products…………………………………………………………………………………………………. 7

Emulsified Asphalt………………………………………………………………………………………………… 9

Objectives…………………………………………………………………………………………………………. 10

CHAPTER 2: WIND TUNNEL TESTING………………………………………………………………………………. 12

Materials & Methods…………………………………………………………………………………………… 13

Wind Tunnel Design………………………………………………………………………………….. 13

Flow Pattern Determination……………………………………………………………………… 14

Box Preparation……………………………………………………………………………………….. 15

Test Set-Up……………………………………………………………………………………………… 16

During the Test………………………………………………………………………………………… 16

Treatments……………………………………………………………………………………………… 17

Data Analysis……………………………………………………………………………………………. 18

Results & Discussion…………………………………………………………………………………………… 19

Dry Straw and Steady Wind………………………………………………………………………. 19

Dry Straw and Gusty Wind………………………………………………………………………… 20

Ease of Application and Cost Analysis…………………………………………………………. 21

iii

Conclusion………………………………………………………………………………………………. 22

CHAPTER 3: FIELD VEGETATION PLOTS…………………………………………………………………………….. 23

Materials & Methods………………………………………………………………………………………….. 23

Lake Wheeler Field Laboratory Site……………………………………………………………. 24

Apex Active DOT Site………………………………………………………………………………… 27

Weather Data………………………………………………………………………………………….. 28

Image Processing using ArcGIS and Data Analysis………………………………………… 29

Results & Discussion……………………………………………………………………………………………. 31

Conclusion…………………………………………………………………………………………………………. 34

CHAPTER 4: GREENHOUSE TRIAL…………………………………………………………………………………….. 35

Materials & Methods…………………………………………………………………………………………… 35

Results & Discussion……………………………………………………………………………………………. 38

Tall Fescue………………………………………………………………………………………………. 38

Centipedegrass and Bermudagrass…………………………………………………………….. 40

Conclusion…………………………………………………………………………………………………………. 41

CHAPTER 5: CONCLUSIONS AND RECOMMENDATIONS……………………………………………………… 42

REFERENCES…………………………………………………………………………………………………………………. 44

TABLES………………………………………………………………………………………………………………………… 53

FIGURES………………………………………………………………………………………………………………………. 66

iv

List of Tables:

Table 1:           Legal timeframes for Department of Transportation site stabilization in

North Carolina (NCDOT (4), 2016)……………………………………………………………… 53

Table 2:           Tackifier cost on an area basis. Costs are calculated using pricing figures

from commercial retailers, specifically the suppliers for this study………………… 53

Table 3:           Products tested in this study as tackifiers……………………………………………………. 54

Table 4:           All tackifiers and application rates included in each test within the wind tunnel study. Application rates are shown as percentages of the

manufacturer’s recommended rate……………………………………………………………. 55

Table 5:           The manufacturer’s recommended rates for tested tackifier products…………… 56

Table 6:           ANOVA table for all tackifiers tested in the wind tunnel using all three tackifier application rates, under dry and steady conditions. These tackifiers include cellulose HM, plantago, Soiltac, Tornado Tack, wood fiber HM, bonded fiber HM, polyacrylamide and Tracer. The only

exceptions were emulsified asphalt and bare straw…………………………………….. 56

Table 7:           ANOVA table for all tackifiers under dry and steady conditions. These tackifiers include cellulose HM, emulsified asphalt, no tackifier, plantago,

Soiltac, Tornado Tack, wood fiber HM, bonded fiber HM, polyacrylamide

and Tracer. P- values of 0.05 or less are considered significant……………………… 57

Table 8:           Tackifier treatments under dry and steady conditions at two application rates of straw. Differences (p<0.05) in wind speed to failure are indicated between the two straw rates if the letters following the failure wind speed

are different……………………………………………………………………………………………. 57

Table 9:           Tackifier treatments under dry and steady conditions at three application rates tested for wind speed to failure. Since emulsified asphalt and “No Tack” are only applied at one rate, comparisons were made at all three rates of the other products. Differences (p<0.05) are

indicated if values do not have a common letter…………………………………………. 58

v

Table 10: ANOVA table for all tackifiers under dry and gusty conditions. Tackifiers include cellulose HM, emulsified asphalt, bonded fiber HM, no tackifier, polyacrylamide, plantago, Soiltac, Tornado Tack and wood fiber HM.

P-values less than 0.05 are considered significant………………………………………… 58

Table 11:          The effects of tackifier treatments on failure wind speed (km h^-1) under dry straw and gusty winds with an application of 2x the recommended rate. Differences (p<0.05) are indicated if values do not have a common

letter………………………………………………………………………………………………………. 59

Table 12:         Manufacturer and pricing information for grass seed, fertilizer and lime used in this project and on NC Department of Transportation (DOT)

construction sites from March 1^st through August 31^st…………………………………. 59

Table 13:         List of tackifiers used at the Lake Wheeler Field Laboratory for the outdoor vegetation tests, accompanied by the application rates at which

they were applied on this site……………………………………………………………………. 60

Table 14:         Grass cover from July 28^th 2016 aerial survey at the Lake Wheeler Field Laboratory. Four observations were recorded for each treatment and

no differences were found (p<0.05)…………………………………………………………… 61

Table 15:         Weights of 200 seeds from each grass species to be used in the

greenhouse trial………………………………………………………………………………………. 62

Table 16:         Greenhouse plot layout for treatments and blocks………………………………………. 63

Table 17:         ANOVA table for the repeated measures analysis of the greenhouse trial. It shows an interaction between tackifier and day, but no differences between tackifier treatments for grass blade count. This table represents tall fescue results only, and includes cellulose HM, emulsified asphalt, no tackifier, plantago, Soiltac, Tornado Tack, wood

fiber HM, bonded fiber HM and polyacrylamide………………………………………….. 64

Table 18:         Grass blade counts of tall fescue on days 16, 18, and 20 by tackifier.

Similar letters within columns are not different at p<0.05. Standard

deviations are shown in the columns marked S.D………………………………………… 64

Table 19:         Grass emergence for tall fescue for all evaluation times, with averaged values for each treatment day and tackifier. Significant differences

between tackifiers at each day are indicated with letters (p<0.05)………………… 65

vi

List of Figures:

Figure 1:          Illustration of tackifiers function with a straw cover. Yellow represents the straw and green the tackifier, which can both bind the straw

together (1) and bind the straw to the soil (2)……………………………………………… 66

Figure 2:          View of the wind tunnel through the section connecting the fan to the

tunnel, with laminar flow baffles……………………………………………………………….. 66

Figure 3:          View of the wind tunnel with the laminar flow baffle section attached…………… 67

Figure 4:          Bare soil box inside wind tunnel. Also shown is the permanently affixed

straw that has been glued to the bottom of the tunnel…………………………………. 67

Figure 5:          Illustration of potential wind flow paths with and without a base straw layer on the floor of the wind tunnel. The top figure illustrates flow paths with a base layer of straw, while the bottom figure illustrates flow

paths with no straw glued to the tunnel floor……………………………………………… 68

Figure 6:          (a) Smoke emitter affixed to the wind tunnel for testing the flow

pattern. (b) Red smoke shows wind pattern inside the tunnel……………………….. 69

Figure 7:          Fully operational wind tunnel with the fan properly attached and a bare soil box inside. The orange wind speed meter sits on top, exterior to the

box, with only the probe on the inside……………………………………………………….. 70

Figure 8:          A before (left) and after (right) view of a soil test box with an application

of hydraulic tackifier. This illustrates a “failure” with less than 50% of the

straw remaining………………………………………………………………………………………. 70

Figure 9:          Variation in failure wind speed of tackifier application rates (% of

manufacturer’s recommend rate) at two moisture conditions (wet and dry). Samples tested under wet conditions withstood significantly higher wind speeds than dry, and significant differences in tackifier application rate are shown with letters (p<0.05). Bars represent standard

error………………………………………………………………………………………………………. 71

Figure 10:        Effect at wet moisture conditions on tackifier and application rate. No statistical differences between any tackifiers or application rates

(p<0.05)………………………………………………………………………………………………….. 72

vii

Figure 11:        Distribution of wind speeds (km h^-1) throughout the wind tunnel. This figure shows the average of three replications and should be observed as if looking from the fan into the wind tunnel. The top set of columns shows the wind tunnel cross section that is closest to the fan, while the

bottom set of columns shows the furthest away. Height is measured from

the tray floor to the top of the tunnel…………………………………………………………. 73

Figure 12:        Application rate effect on failure wind speed under steady flow wind conditions. Statistical analyses for dry and steady results provided in

Table 6. Bars represent standard error……………………………………………………….. 74

Figure 13:        Differences in tackifier treatments at the recommended rate for dry straw and steady conditions. Differences (p<0.05) are indicated if

values do not have a common letter………………………………………………………….. 75

Figure 14:        Cost-benefit analysis for each tackifier under dry straw conditions, at three application rates, as reflected in the cost. Emulsified asphalt is shown as a straight line since it is only applied at one rate. The light blue box represents tackifier and application rates that would be

recommended to replace emulsified asphalt………………………………………………. 76

Figure 15:        Experimental treatments by block for the spring study…………………………………. 76

Figure 16:        The locations of each of the four soil samples (S1-S4) taken in each block

at the Lake Wheeler Field Laboratory…………………………………………………………. 77

Figure 17:        Photographs of the spring field test of tackifiers at the Lake Wheeler Field Laboratory, after a heavy rainstorm approximately one month after planting. The photograph on the left (a) shows the most extreme erosional damage, the top right (b) shows the plots affected by the damage and the bottom right (c) shows the entire grass stand on the

same date (May 18^th, 2016)……………………………………………………………………….. 77

Figure 18:        Final aerial image (top) from the spring vegetation installment at the Lake Wheeler Field Laboratory (July 28^th, 2016), after 3.5 months of grass growth. The image on the bottom was processed using ArcMap, and shows the vegetation cover in green, bare soil in brown, and plot

markers in red image……………………………………………………………………………….. 78

Figure 19:        Grass cover results from September 28^th, 2016 at the Lake Wheeler Field Laboratory. Differences (p<0.05) in grass cover (%) are indicated if

viii

values do not have a common letter. Letter “A” represents tackifier

treatments that had higher cover than tackifiers with a letter “B”…………………. 79

Figure 20:        Aerial image of the Lake Wheeler Field Laboratory vegetation study

(September 28^th, 2016)…………………………………………………………………………….. 79

Figure 21:        Aerial photograph of the Department of Transportation active site in

Apex, NC after Hurricane Matthew (October 31^st, 2016)………………………………. 80

Figure 22:        One of the gullies at the Apex site resulting from Hurricane Matthew,

with a rain gauge to depict the scale………………………………………………………….. 81

Figure 23:        Tackifier treatment plots located at the Apex Active site, September 29^th,

2016 …….………………………………………………………………………………………………………….81

Figure 24:        Grass cover results from the Apex, NC active Department of Transportation site on September 29^th, 2016. For each tackifier the upper edge of the box represents the 3^rd quartile (75^th percentile), the line inside the box represents the median (50^th percentile), the diamond represents the mean and the bottom edge of the box is the 1^st quartile (25^th percentile). There were no significant differences between any of

the tackifier treatments (p≤0.05)……………………………………………………………….. 82

Figure 25:        Photograph of the greenhouse trays prior to seed, straw and tackifier

application………………………………………………………………………………………………. 83

Figure 26:        Photographs of the greenhouse trial in progress, showing one bench (block). Photograph (a) shows one block within the greenhouse, (b) shows a Tornado Tack sample, (c) shows two emulsified asphalt samples, and (d) shows the trays located at the back of the greenhouse, in the 3^rd

block………………………………………………………………………………………………………. 84

Figure 27:        Grass emergence is shown for each tackifier on three sampling days where differences were evident. Treatments with different letters represent significant differences (p<0.05), and bars represent standard error. The color of the symbol signifies with which sampling date it corresponds. If the treatment does not have a letter, this means that it

is neither better nor worse than any other tackifier…………………………………….. 85

Figure 28:        Tall fescue blade count increases on days 9-11, 11-14 and 14-16. Significant differences are denoted with letters (p<0.05), and bars

ix

represent standard error. Treatments that do not have letters are

neither better nor worse than other tackifiers…………………………………………….. 86

Figure 29: Blade count results for centipedegrass and Bermudagrass on day 40, the final day that grass blade count was recorded. No significant differences (p<0.05) in grass blade count between tackifier treatments, but bars represent standard error…………………………………………………………………………… 87

Chapter 1: Literature Review

Introduction:

Soil erosion from either rainfall or wind can be a major problem on construction projects. The United States Environmental Protection Agency (USEPA) states that “soil loss rates from construction sites are 10 to 20 times that of agricultural lands” (USEPA, 2000). Once the vegetation is removed during the grading process, exposed soil can easily wash or blow away in storm events. Detached sediment then moves downwind or downslope, and can cause major problems for sensitive waterways or wetland areas. Many aquatic species are negatively affected by high turbidity, as it can reduce visibility, cause difficulty breathing, and bury rocky streambed environments where many organisms live (Grace, 2000). For this reason, the U.S. Environmental Protection Agency now mandates that graded soils may remain bare for no more than 14 days (USEPA, 2009).

The most susceptible time for sediment loss is immediately after grading (Bethlahmy and Kidd, 1966; Burroughs and King, 1989; King, 1984; Megahan, 1974; Megahan et al., 1991; Swift, 1985). Furthermore, roadside slopes may produce 70-90% of the sediment loss from construction projects (Swift, 1984b), with newly created slopes being most at risk, due to their loose, structureless nature (Grace, 2000). One of the best ways to reduce wind and water erosion on bare soil is to establish vegetation. According to the Alabama Forestry Commission (1993), establishment of vegetation is shown to reduce sediment loss from road side-slopes, and roots effectively hold soil particles together, where shoot (or above-ground) growth can lessen the impact of water droplets (Osborn, 1955). Furthermore, grasses and other vegetation

1

can slow wind speeds at the soil surface, and consequently limit particle detachment. Therefore, complete vegetation cover is ideal.

For vegetation establishment to occur, the grass seed requires a sheltered environment from damaging winds or rains that may wash the seed away. Straw mulch is typically used for standard highway seeding practice. Straw also reduces seed transport away from the site by intercepting rain drops and slowing sheet flows. Currently, emulsified asphalt is often used to prevent the straw from being blown away by wind, but it has a number of negative impacts: 1) It may pose environmental concerns, primarily potential water contamination, 2) it requires specialized equipment to apply, and 3) it can inhibit grass establishment at high application rates (Dudeck et al., 1970; McKee et al., 1964). Therefore the objective of this study was to find and compare potential tackifier products for their effectiveness at withstanding wind, cost efficiency, and ease of application. The overall goal was to determine if emulsified asphalt can be replaced by another tackifier that is as or more effective, less environmentally damaging, and potentially less expensive.

Vegetation Establishment:

On some landscapes it can be difficult to establish grass, with many perennial grasses potentially dying off within the first year (Brown and Gorres, 2011). Roadside soils often have low organic matter content, due to the removal of the surface horizon, subsequently low microbial activity, as well as high salt content from runoff after winter storms (Brown and Gorres, 2011). Varying climatic conditions, soil types, slopes, and other factors can greatly

2

affect the potential for successful grass establishment. Two of the main factors influencing germination are soil water content and the amount of seed-to-soil contact (Hauser, 1989). However, good contact can sometimes be difficult to achieve depending on the site conditions. One study tested the effects of water stress on tall fescue (Festuca arundinacea Schreb.) and found that at osmotic stress (0.76 MPa) reduced blade height and leaf extension rate (Thomas et al., 1998).

Different species are able to withstand different soil moisture and temperature levels (Swemmer et al., 2006). One option to take advantage of these differences is to apply a mixture from different species, which in turn reduces the risk of stand failure (Smale, 2004). As a result of similar research, the North Carolina Department of Transportation (NCDOT) uses an explicit seed specification to ensure seed diversity and quality (NCDOT (1), 2017). Different seed mixtures are recommended depending on the county where grass establishment is required (NCDOT (2), 2017). Additionally the NCDOT recommends either warm-season or cool-season turfgrass mixes depending on the time of year or region of the state. Cool-season grasses include tall fescue, hard fescue and Kentucky bluegrass (Poa pratensis), while warm-season grass contains centipedegrass (Eremochloa ophiuroides), bermudagrass (Cynodon dactylon), zoysiagrass (Zoysia japonica) and ‘Pensacola’ bahiagrass (Paspalum notatum) (NCDOT (3), 2017).

3

The Effect of Increased Vegetation:

Vegetation establishment is important, as it not only acts as a ground cover to protect the soil from erosion (Marques et al., 2007), but it can actually improve soil structure, infiltration, water holding capacity, and even decrease bulk density (Logsdon, 2013).

Additionally, in a study by Pan et al. (2006) it was found that grassy slopes can reduce runoff by 8%, and that bare slopes have 45 – 85% more sediment loss than grassy slopes. This is especially important on construction site soils, as many construction processes impair soil quality. By damaging soil structure and quality, grass establishment and growth may be hindered.

One of the main ways that construction damages soil is by compaction, often from heavy machinery, during the cut and fill process of grading (Gregory et al., 2006). Soil compaction increases soil strength and bulk density, decreases porosity, and creates smaller pore size distribution. This reduces the water and air flow throughout the soil, and can stress plants by limiting water availability for roots, and physically hindering root penetration (NRCS, 2000; Richard et al., 2001). Improved soil quality can help aid in the reduction of erosion, as increased infiltration allows water from a rain event to soak into the soil. This in-turn allows plants to receive more water, and possibly prevent the need for re-seeding due to drought stress.

Temporary Erosion Control:

Permanent stabilization using vegetation is generally the final goal for bare soil areas; however, until the vegetation becomes established, temporary measures must be taken to

4

prevent erosion. There are a variety of options including rolled erosion control products (such as erosion control blankets), hydromulches and loose straw, all of which perform the same general role. Each acts as a shield to protect the soil from raindrop impact and decrease the likelihood for detachment from soil aggregates (Gholami et al., 2014). In addition, mulches protect the soil from solar radiation, keeping the soil up to 20°C cooler than bare soil (Ross et al., 1985). According to Swanson et al. (1965), mulching was very effective in stabilizing slopes and preventing soil loss, and others found increased seed germination and growth when using mulch (Gilbert and Davis, 1967; Blaser, 1962). Similarly, Lemly et al. (1982) determined the effect of mulch on grass establishment, and found that with all five treatments (jute netting, excelsior, mulch blanket, wood chips and asphalt-tacked straw) each treatment promoted significantly greater grass cover than leaving the seeded soil bare.

Straw mulch may be used to hold grass seed in place during the germination phase. The NCDOT recommends 80% straw mulch coverage on all slopes (NCDOT (6), 1998), which helps to shade the seed, insulating it from extreme temperatures, and keeping the soil moist by reducing evaporation (Adams, 1966; Jordan, 1998; Grigg et al., 2006). McKee et al. (1964) found it to be one of the best mulching materials for revegetation compared to netting, hydromulches or a combination of products, particularly on steep slopes.

Wheat straw mulch was selected for this project due to its frequency of use, affordable price, ease of application and effectiveness in erosion reduction. In a study by Meyer et al. (1970), soil loss from straw-mulched plots on steep slopes (15%), was found to be approximately one third of that on bare soil areas. Straw mulch reduced erosion by 90%

5

compared to bare soil, with a significant increase in effectiveness with grass growth (Benik et al., 2003). Straw also functions to protect seeds during the germination phase by partially shading the seed, modify soil temperatures, and keeping the soil moist by preventing evaporation (Adams, 1966; Jordan, 1998). Additionally, as the wheat straw breaks down, it contributes organic matter to the soil, with a C:N ratio of approximately 80:1 (Dahmer, 2017).

Loose straw in bales is spread over the site manually or by using a straw blower. These blowers can disperse straw up to 50 feet depending on the size of the machinery, and distribute straw very quickly (NCDOT (3), 2017). Although the blanket form may prevent the straw from blowing away, it takes significantly more time to install, and is therefore less preferable when trying to cover large areas. NCDOT recommends 80% straw mulch coverage, or approximately 2240 – 4480 kg ha^-1 (2000 – 4000 lbs ac^-1) on all slopes, and tackifiers are applied to prevent the straw from blowing off the slope (NCDOT (6), 1998).

In order to avoid violations, all construction sites in North Carolina must abide by soil stabilization timeframes (Table 1) mandated by the Construction General Permit NCG 01 (Construction General Permit NCG 01, Section II.B.2). These timeframes indicate how long a site may remain bare before temporary or permanent stabilization must be in place, generally grass seed with straw mulch and tackifier. Different slopes require stabilization timeframes that correlate with their potential for soil loss; for instance, slopes steeper than 33% require stabilization within 7 days, whereas slopes shallower than 3:1 have up to 14 days.

6

Tackifier Products:

There have been many studies of the effectiveness of hydromulches (HM) as a cover during grass establishment but few hydromulches are used as a tackifiers. The main difference being that hydromulches as tackifiers use straw mulch as a base to decrease cost by applying less product, whereas hydromulches as a groundcover are applied at high application rates in order to provide adequate soil coverage. Although some studies have shown hydromulches to be effective without tackifiers (Emanual, 1976), other research has shown them to be ineffective. According to Faucette et al. (2005), hydroseeding provides limited soil coverage prior to vegetation growth, but may slow the drying period following application. However, by using straw as a base for each tackifier tested in this study, the concern for inferior soil coverage should be eliminated. Additionally, product costs would be greatly reduced by supplementing the tackifier with straw, as straw costs <$620 per hectare, as opposed to hydromulches which can range from $0.40 – $1.85 per kg, with a typical application rate of 2240 kg ha^-1 (2000 lbs ac^-1), for a total of $900 – $4150 per hectare (Table 2).

There were four main types of tackifiers that were tested in this study: 1) wood-based, hydraulically applied mulches (cellulose HM, bonded fiber flexible growth media, Tornado Tack, and wood fiber HM, 2) plant-based glue (plantago), 3) flocculants (polyacrylamide), and 4) organic soil stabilizers (Soiltac and Tracer) (Table 3).

Wood fiber products as tackifiers are less expensive, often because they come from recycled materials, equally as effective and more environmentally acceptable in comparison to emulsified asphalt. (Kay 1978, Brown and Hallman, 1984). Many of these hydromulches obtain

7

their fibers from the wood of aspen or alder trees, while others use recycled paper (Keammerer, 1988). These hydromulches are also typically dyed a bright green or blue color as a visual aid for even and full coverage over the site (Keammerer, 1988).

Linear polyacrylamide (PAM), although not typically used as a tackifier, was included in this study since it is used for erosion control on constructions sites and in irrigated agriculture (Sojka et al., 2007), and therefore could provide two functions if it worked as a tackifier.

Polyacrylamide used in erosion control, and in this project, is a linear anionic polymer which becomes sticky when wet and may hold straw. Linear PAM is also one of the ingredients in Tracer.

Plantago, is a plant byproduct that has been ground into a fine powder and which becomes sticky when combined with water. Due to its inexpensive and environmentally friendly nature, plantago was included in this study.

Soil stabilizers, oftentimes organic emulsions and frequently used as paints or glues, have been successful tackifiers due to their field longevity resulting from their resistance to

ultraviolet radiation (Keammerer, 1988). TRACER™ (Epic Manufacturing, Greenwood, DE, USA) tackifier, used in this study, is one example of this type of soil stabilizer. It comes in a powdered form, which is then added to water to create the tackifier for straw. It is described as “a water soluble blend containing linear, anionic, copolymer of acrylamide and sodium acrylate, with less than 0.05% acrylamide monomer and blended with one or more polysaccharides such as pre- gelatinized starch in conjunction with an inorganic salt used as a cross linking agent” (Braun, Material Safety Data Sheet (MSDS)). The powdered form is greenish-brown and when mixed

8

with water it transitions to a brighter green color. It is not currently listed as a carcinogen by the International Agency for Research on Cancer (IARC), National Toxicology Program (NTP), Occupational Safety and Health Administration (OSHA), or the Association Advancing Occupational and Environmental Health (ACGIH). Soiltac®, another organic soil stabilizer used in this study, is manufactured by Soilworks® (Soilworks LLC, Scottsdale, AZ, USA) for the purpose of dust control. It is composed of 55% synthetic vinyl copolymer dispersion and 45% water,  with no known carcinogens or environmentally hazardous chemical components (Soiltac MSDS).

The products selected for this study can all be applied with a standard hydroseeder, which are commonly used on construction sites. Basic information about the products in this study is provided in Table 3.

Emulsified Asphalt

Emulsified asphalt is used as a straw tackifier, primarily due to its availability and affordable cost of the straw mulch base. By emulsifying the asphalt, it creates a product that is easier to apply, and the water will evaporate after application leaving only the asphalt behind as the tackifier (TDOT, 2006).

However, despite its widespread use it is important to note some concerns about this product. Although the application is less involved than laying asphalt for pavement, the process is still difficult because the tackifier must be kept between 40° – 70° C (104° -158° F) during

application (TDOT, 2006). This raises concerns as some seed species can be damaged by high

9

temperatures (Corbineau et al., 2002) and contact with the hot asphalt could kill the seeds. Even if the asphalt cools quickly after initial application, its dark black color will absorb sunlight and therefore continue to warm the emerging seedlings, or could prevent sunlight from reaching the ground surface at all. Asphalt tackifier produced inferior grass growth compared to seven other mulching treatments on fill slopes (Grace, 2000), and poor grass stands have been observed in areas with generous asphalt application (McKee, 1965).

Additionally, asphalt is dangerous to animals who may come into contact with the substance and are unable to remove it from their fur (Sittig and Pohanish, 2002). For the workers applying tackifiers, there are short- and long-term exposure risks in the form of burns and skin cancer (Sittig and Pohanish, 2002). Personnel handling asphalt are required to wear personal protective equipment when applying product to protect themselves from any risks associated with it (Sittig and Pohanish, 2002).

Objectives

Due to the erosive nature of bare slopes, it is imperative to work towards permanent stabilization after construction projects. Grass is a common form of permanent stabilization, while straw mulch is used in combination with tackifier as a temporary stabilization until the grass grows. In this project we compare the current straw tackifier, emulsified asphalt, to other alternative tackifiers in an attempt to find a product that could be used as its replacement.

Emulsified asphalt is problematic due to environmental concerns, high price and difficult application process, which an alternative tackifier could alleviate. To determine if an alternative tackifier would be an appropriate replacement, it must be able to withstand storm-level wind

10

and have no negative impacts on grass growth. Additionally, to be a recommended replacement, the product should be less expensive and easier to apply.

11

Chapter 2: Wind Tunnel Testing

The wind tunnel used in this study was built for the specific purpose of testing tackifier effectiveness on wheat straw. The goal was to find which tackifier products were most successful at withstanding laminar wind, and preventing straw mulch failure. We hypothesized that straw would be able to withstand higher wind speeds when using a tackifier, but we did not have any specific products that were expected to work better than others. The ultimate goal was to find any alternative tackifier products that were at least as effective at tacking straw as emulsified asphalt, less environmentally detrimental, and at a lower cost.

Building a wind tunnel to fit the exact specifications for this project was an additional endeavor. It was important that this tunnel was portable, and therefore relatively small, created a laminar wind pattern, and reached wind speeds representative of moderate storm events. Many portable wind tunnels found in the literature were designed for testing soil erodibility, and therefore have the actual ground surface as the “working section” of their tunnel (Pelt and Zobeck, 2013). In this project, however, the focus was on straw failure and therefore a fully enclosed tunnel was preferable. The tunnel created in this project allowed for full sample trays of soil, straw and tackifier to be freely moved in and out of the wind tunnel.

12

Methods:

Wind Tunnel Design:

The wind tunnel in this study was an example of an open design tunnel (Advanced Thermal Systems, 2009), meaning that ambient air enters the wind tunnel and then exits into ambient air after passing through the testing chamber (Advanced Thermal Systems, 2009 ). The dimensions are approximately 2.4 x 0.61 x 0.3 m (8 x 2 x 1 ft), with four wheels for easy mobility. The fan used was a Tempest Technology Direct-Drive Gas Power Blower (Tempest Technology Corporation, Fresno, CA), with a 4.1 kW (5.5 horsepower) engine, 53 cm (21 in) blades, dimensions of 63.5 x 53.3 x 66.7 cm (25 x 21 x 26.25 in), and total weight at 40.8 kg (90 lbs). It achieved maximum wind speeds of 72 to 80 km h^-1 (45-50 mi h^-1), and minimums between 24-32 km h^-1 (15- 20 mi h^-1). It was attached to the testing tunnel via a segmented wooden frame with baffles to create a laminar flow pattern (Figures 2 and 3). This is important as flow quality often determines the effectiveness of a wind tunnel (Hernández et al., 2013).

Flow quality includes three main factors: maximum achievable speed, flow uniformity and turbulence level (Hernández et al., 2013). According to Pelt and Zobeck (2013) there are also three components to a portable wind tunnel: 1) a self-contained power source or engine, 2) a fan, and 3) transportation of wind from the fan to the working section, either an actual soil area, or an enclosed testing space.

The floor of the tunnel was covered in a base layer of wheat straw glued in place to simulate bare soil with straw applied to it. In the center was a 0.3 x 1.22 m (1 x 4 ft) cutout for removable wood boxes that served as the test plots (Figure 4). The boxes sat flush with the floor of the tunnel to prevent interference with the air flow. The purpose was to create an

13

environment similar to straw applied to a slope and to avoid any discontinuity in the straw layer, as illustrated in Figure 5.

Flow Pattern Determination:

To examine the flow pattern within the wind tunnel a series of videos were taken using colorful smoke emitters (Burst Wire-Pull Smoke Grenades, Enola Gaye Company, Pahrump, NV, USA). These emitters were fastened to the wind tunnel such that either one or two emitters were suspended in the middle of the tunnel entrance (Figures 6a and b). Once they were securely fastened, the plug was pulled, video camera started and the fan was set to produce 32 km h^-1 (20 mi h^-1) winds, or the lowest speed. The wind speed was increased slowly, and the test was only concluded when the smoke had run out. These videos were then used to determine the general wind flow pattern within the tunnel, and elucidate any areas of turbulence.

Additionally, results from the smoke tests indicated an even flow distribution throughout the tunnel, as well as a laminar flow pattern. This was determined through visual observation both live and by reviewing video recordings. Additionally, a more qualitative measurement was taken to determine the wind speed distribution. Using the anemometer, wind speed was measured in various places throughout the tunnel, including 3 depths, with 5 points at each of the three cross sections (Figure 11), for a total of 45 measurements per replication. Overall for the lowest wind speed (approximately 32 km h^-1) the standard deviation is 0.98 km h^-1 on

14

average, and 2.3 km h^-1 for the highest wind speed (80 km h^-1). This suggested that in general the speed is reasonably uniform throughout the tunnel.

Box Preparation:

Each box was 0.27 x 1.18 m (0.88 x 3.88 ft) for a total of 1.041 m² (3.414 ft²). The boxes were filled with a clay loam soil, 42% sand, 31% silt and 28% clay, which was smoothed to create an even surface. Each box then received 2240 or 4482 kg ha^-1 (2,000 or 4,000 lbs ac^-1) straw, depending on the test, which was spread by hand evenly over the surface of the soil.

After straw was applied, the tackifier was applied evenly and by hand. Each tackifier was tested at 50%, 100%, and 200% of the manufacturer’s recommend rates (Figure 7), except for the emulsified asphalt which was applied by a contractor. Since emulsified asphalt does not have a specific application rate, it is generally applied by sight, and therefore we could not halve or double the recommended rate. To account for variability within the contractor applied emulsified asphalt, we chose the most average looking boxes out of 18 total trays, to be included in the study.

After tackifier application, each soil box was left for a minimum of two days after the liquid tackifier application, which allowed time for moisture to evaporate. The second set of tests was conducted to simulate wet conditions, such as during a rain event. For these, after two days of drying after tackifier application, the boxes were placed in a rainfall simulator for ten minutes, and then placed in the wind tunnel for testing ten minutes later. The rainfall simulator averaged approximately 5 cm (2 in) of water per hour.

15

Test Set-Up:

The tunnel was stored in a shed at the LWFL (Lake Wheeler Field Laboratory) site and was transported into an open, covered area for testing. There are three main parts: the fan, the stand for the fan, and the tunnel itself. Both the fan and tunnel have wheels and can be rolled out of the shed and into the covered area chosen for running these tests. The base for the fan can easily be carried and is necessary as it allows the fan to be on the same horizontal plane as the tunnel. The soil boxes were also stored in the shed and transported to the wind tunnel.

Boxes were placed in the wind tunnel and run individually.

During the test:

Each steady flow test was started at 24-32 km h^-1 (15-20 mph), the lowest wind speed, and the throttle was adjusted to result in an 8 km h^-1 (5 mph) increase in speed every minute. For gusted flow tests, wind speeds began at 40 km h^-1 (25 mi h^-1) and were increased by 16 km h^-1 (10 mi h^-1) every 20 seconds. Additionally, instead of increasing the wind speed while the fan was running, like in the steady flow test, the fan was turned off after each 20 seconds, and then started again from a 0 km h^-1. This created a faster acceleration of wind speed in the gusty tests, as compared with the steady flow. Testing was completed when either a failure occurred or maximum wind speed (72-80 km h^-1 (45-50 mph)) was reached. Each test was monitored with a Pitot tube anemometer (Extech HD350: Anemometer and Differential Manometer, Extech Instruments, Boston, MA) as well as a video camera. The gauge recorded wind speed continuously and showed the current wind speed on a display during the testing. The video

16

camera recorded both the wind tunnel and the wind gauge, which was later used as a reference when needed. During the first 15 seconds of each video a sign was placed in view to denote which test was being run. In addition, any notable observations during the test were also recorded by hand, and the failure wind speed was documented. If the straw/tackifier combination was not a full failure, a brief description was added, to provide a more detailed analysis. A failure was defined as the point at which more than 50% of the straw had blown off (Figure 8). After the test was completed each box was photographed to record the amount of failure.

Treatments:

Wind tunnel tests were conducted under a variety of conditions to determine the effects on the performance of the tackifiers. These conditions included tackifier application rate, straw application rate, wet vs. dry straw, and gusty vs. steady wind (Table 4). Application rates were tested under dry and steady wind conditions at the rate recommended by the manufacturer as well as at half and double that rate (Table 5), while straw application was tested at 2240 kg ha^-1 (2000 lbs ac^-1) and 4480 kg ha^-1 (4000 lbs ac^-1). Dry samples were tested directly after the two day drying period. Wet samples were dried for two days and then placed under a rainfall simulator (5 cm h^-1) for ten minutes prior to testing, with only the higher straw rate included. Steady wind testing involved increasing wind speed by 8 km h^-1 (5 mi h^-1) every minute until failure. Gusty tests were increased by 16 km h^-1 (10 mi h^-1) every twenty seconds, with the fan being turned off between wind speed increases. For instance, if the starting wind

17

speed was 40 km h^-1 (25 mi h^-1) and no failure occurred, the fan would be stopped and restarted at 56 km h^-1 (35 mi h^-1) , so a 0 – 56 km h^-1 (0 – 35 mi h^-1) gust was created. Only the 4480 kg ha^-1 (4000 lbs ac^-1) straw rate was included in the gusty testing.

Data Analysis:

After collection, raw data was entered into excel spreadsheets before statistical analysis using SAS (version 9.4, SAS Institute Inc., Cary, NC, USA). In SAS, one-way analysis of variance (ANOVA) was performed on all data (p ≤ 0.05), which compares the means of independent groups, and showed whether the means are significantly different. The significant differences found using this method are shown in many of the tables and figures, and are expressed using lower case letters. Additionally, two-factor, factorial ANOVA was used to elucidate interactions between variables for the dry and steady tests. The reason why this analysis was performed only under these conditions is because it was the only test that included all variables. Both straw rates were tested only under dry and steady conditions, along with all tackifier rates, whereas wet tests or dry and gusty tests did not include all of these rates. However, due to

emulsified asphalt and “no tack” only being applied at one application rate under all conditions, even the dry and steady test had to be adjusted slightly in order to reduce statistical error. This was done by creating a reduced model that only compared tackifier, straw rate, tackifier rate, replication, and the interaction between tackifier and straw rate, leaving out the interaction between tackifier and tackifier application rate. Additionally, in order to accurately determine an interaction between tackifier and tackifier application rate, emulsified asphalt and “no tack”

18

were removed from the data set, and the interaction was then calculated with the seven remaining tackifiers.

Results and Discussion:

Some of the general results were that higher tackifier application rates performed better than lower application rates under dry and steady conditions (Figure 9), with a two- factor interaction p-value of <0.0001. Wet straw was able withstand higher wind speeds than dry straw. In fact, the other factors, including tackifier, tackifier application rate, and wind condition had little effect when the straw was wet, and in many tests the maximum wind speed (80 km h^-1 [50 mi h^-1]) was reached with no failure (Figure 10). A statistical interaction was also found between tackifier type and straw application rate (p = 0.0003), indicating that the higher straw rate resulted in different tackifier performance under dry and steady conditions.

Dry Straw and Steady Wind:

The effects of tackifier application rate for dry and steady tests were consistent with that of all dry samples, where 200% of the recommended rate was better at withstanding wind than 100%, which was better than 50%. There was a significant interaction between application rate and tackifier with a p-value of less than 0.0001 (Table 6) as well as an interaction between tackifier and straw application rate (p= 0.0004) (Table 7). Six tackifier treatments were unaffected by straw application rate, while cellulose HM and Tornado Tack performed better at the 2240 kg ha^-1 (2000 lbs ac^-1) straw application rate (Table 8). When no tackifier was applied, the higher straw rate resisted winds up to 8 km h^-1 higher than the lower rate.

19

At 200% application rate the average wind speed to failure is 70 km h^-1 (43.4 mi h^-1), whereas 100% and 50% are 57.5 (35.7) and 46.3 km h^-1 (28.8 mi h^-1) respectively. Differences between tackifier treatments were found at all tackifier application rates. At 50% application rate, bonded fiber HM, Tornado Tack, and emulsified asphalt could all withstand higher wind speeds than plantago, while only bonded fiber HM and emulsified asphalt could withstand higher wind speeds than bare straw and plantago. At the half rate, Soiltac, wood fiber HM, PAM, Tracer, plantago and bare straw performed worse than emulsified asphalt (Table 9).

However, in this comparison, emulsified asphalt was applied at its recommended rate (the only rate tested throughout the study), while all other tackifiers were applied at half the recommended rate. Cellulose HM, bonded fiber HM, Tornado Tack, wood fiber HM, Soiltac and emulsified asphalt all performed better than PAM, plantago, Tracer and bare straw samples at the recommended rate (Table 9 and Figure 13). At the 200% application rate, cellulose HM, bonded fiber HM, Tornado Tack, wood fiber HM, emulsified asphalt and plantago all performed better than Tracer and bare straw, while PAM was neither better nor worse (Table 9).

Dry Straw and Gusty Wind:

Samples run in gusty wind flow conditions had no differences in failure wind speed among tackifier treatments when all rates were included (Table 10). In general, samples tested under gusty conditions failed at lower wind speeds than under steady conditions. On average, for dry samples, steady tests failed at 55.7 (34.6) vs. 51.1 (31.7 mi h^-1) km h^-1 under gusty conditions. Furthermore, there were no differences at the recommended rate under gusty conditions. At the 200% application rate, plantago and bonded fiber HM withstood higher wind

20

speeds than bare straw and PAM, while all other tackifiers performed neither better nor worse (Table 11). Additionally, there was no significant interaction between tackifier and application rate under dry and gusty conditions.

Ease of Application and Cost Analysis:

Another focus for this project is to find tackifier products that are both easy to apply and cost efficient. All tackifiers used in this study were tested for their ability to be applied via a hydroseeder, and all are capable of being sprayed through a 3.8 cm (1 ½ in.) nozzle. This was determined through brief testing or according to the manufacturers when information was available. The hydroseeder used was a TurfMaker 420 (TurfMaker Corp., Rowlett, TX) with a 3.8 cm (1 ½ in.) hose attachment. The only product not applicable for hydraulic application was emulsified asphalt, which must be applied with specialized equipment to keep the mixture heated. All products require that the spray tank be cleaned thoroughly after application; however, PAM and plantago require special attention. We observed that plantago biodegrades rather quickly after application, so material left in the tank is likely to also degrade quickly. PAM becomes very viscous and sticky as it dries and this could clog the hydroseeder plumbing unless it is thoroughly rinsed.

There are a wide range of material costs depending on the type and brand of the tackifier, with national brand name hydromulches being relatively expensive, and off brand or alternative tackifier products being much cheaper (Table 2). A cost-benefit graph (Figure 14) compares product prices to their effectiveness at withstanding wind, which can be used to

21

determine the least cost to attain the greatest resistance to failure. From this analysis, either plantago at 200% the recommended rate or cellulose HM and wood fiber HM at the recommended rate appear to be the best options for cost and performance.

Conclusion:

The main results from this study show that there are differences between tackifiers, with hydromulches holding the straw to higher wind speeds on average than PAM, Soiltac and no tack treatments. Plantago is as effective as hydromulches when applied at twice the recommended rate, but is ineffective at the recommended rate. Under dry and steady conditions, tackifier effectiveness in holding straw in place increases with tackifier application rate, and that effectiveness decreases with higher straw applications. Additionally, wet straw was found to be highly resistant to failure when exposed to wind even without a tackifier.

Lastly, straw failures occurred at lower wind speeds under gusty conditions, with rapidly increasing wind speeds, compared to steady winds in the wind tunnel.

22

Chapter 3: Field Vegetation Plots

The goal of the field vegetation study was to determine if any tackifiers which were successful in preventing straw failure in the wind tunnel testing affected grass germination or growth. We expected that emulsified asphalt would negatively impact growth more than other tackifiers, due to its dark color and ability to obscure light penetration to the ground surface.

Studies by Grace (2000) and McKee (1965) found that emulsified asphalt performed significantly worse than other mulch treatments, and that high applications of asphalt can hinder growth. Previous tests of hydromulches alone at a full erosion control rate suggested they can inhibit grass growth (Lee, 2012). This study allowed for natural weather patterns, soil conditions and wind to influence the study and create a simulation of actual construction site conditions.

Methods:

Two locations were used for the grass establishment study, one being the LWFL in Raleigh, North Carolina, and the other was a NCDOT project site in Apex, NC. The LWFL plots were replicated in the spring and fall, whereas the Apex site was only tested in the fall. The LWFL site was initially prepared by scraping the existing grass off with a motor grader then rotary tilling the area 12-15 cm (5-6 in) deep. The area sloped southward at approximately 4%. The Apex site was a fill slope on the inside curve of a highway interchange under construction, with approximately 10% slope facing northeast.

23

Lake Wheeler Field Laboratory Site

The grass establishment study at LWFL was conducted twice, one beginning in mid-April 2016, and the other in late September. Prior to test initiation, the entire site 6.1 x 30.5 m (20 x 100 sq. ft.) was scraped to remove the top layer of soil, and create more realistic construction site conditions. Next it received DOT specification lime and fertilizer applications followed by tillage. Fertilizer was applied at 560 kg ha^-1 (500 lbs/ac) of 10-20-20 (N-P-K) and 4482 kg ha^-1 (4000 lbs ac^-1) of lime. The grass seed mix is specified under the NCDOT eastern North Carolina seeding and mulching requirements for “shoulder and median areas”, with slight variations depending on the season. The two seasonal mixes are composed of 56 kg ha^-1 (50 lbs/ac) tall fescue, 11 kg ha^-1 (10 lbs/ac) centipedegrass, and either 28 kg ha^-1 (25 lbs/ac) bermudagrass (hulled) or 39 kg ha^-1 (35 lbs/ac) burmudagrass (unhulled) for March 1- August 31 or September 1- February 28, respectively (NDOT (6), 2016). Since all the testing conducted in this study occurred between August 1^st and June 1^st, only the fall mix was used.

After the grass seed was applied using hand-held rotary spreaders on April 4^th and September 15^th 2016, wheat straw was applied with a commercial straw blower to create an even distribution of approximately 2240 to 4480 kg ha^-1 (2000 – 4000 lbs/ac). An acceptable application, according to the NCDOT, allows some sunlight to reach the soil, while still partially shading the ground, which helps to reduce erosion and conserve soil moisture (NCDOT (5), 1998). Since each bale of straw is of variable weight, this range is rather wide, and most contractors apply straw by sight, rather than a set number of bales or weight. However, based on average bale weight, 247 bales was approximately the correct amount for one hectare (100

24

bales per acre). A cost breakdown for each grass type is included in Table 12, along with the specific brand and variety information. All pricing information is based upon the rates that were paid during this project, so certain products may be less expensive if purchased in bulk. Under this pricing scheme, and the estimated 247 bales of straw at $6 per bale, the total cost prior to tackifier application was approximately $4,500 per hectare.

Fifteen tackifier treatments were applied during the spring replication, and 16 treatments in the fall (Table 13). Tackifier treatments were selected based on the wind tunnel testing to select tackifiers and rates that were effective at withstanding wind speeds of over 48 km h^-1 (30 mi h^-1). For this reason, some tackifiers were tested at multiple application rates.

Each treatment was replicated four times in a complete, randomized block design (Figure 15), with Tornado Tack at the 2X rate replacing the second plot of the “No Tack” treatment in each block for the fall study. Replication in the spring and fall study had the same plot layout within the blocks; however, the entire testing area was shifted approximately five feet before initiating the fall testing. This shift ensured that no plot would be replicated at the same location as the previous test. Each plot was 0.46 x 0.46 m (5 x 5 ft.), so by moving the testing area, no plot was on the same area as in the previous test. All plots were marked with colored flags to denote boundaries and to designate tackifier treatments.

The tackifiers were carefully applied by hand immediately following the straw application, to prevent it from being blown away. A commercial operator applied the emulsified asphalt to those plots using their standard equipment. The plots were photographed periodically, either from the ground or from an unmanned aerial vehicle (drone), to track the

25

grass germination and growth. Spray paint or other markers were used to indicate plot boundaries that would be visible from above. The aerial photographs were then processed using ArcGIS (version 10.3.1; ESRI, Redlands, CA, USA) to determine the vegetation cover for each plot. Ground photos were not used in any analysis, but were helpful for documenting growth progress. Aerial photographs were taken on June 22^nd (~2 months after planting) and July 28^th (~3 months after), 2016 for the spring replication and on September 28^th (~2 weeks after) and October 18^th, 2016 (~1 month after) for the fall. The main reason for the difference in time elapsed from planting and photographing was the slow grass growth during the spring replication. It’s possible that long gaps between rainfalls may have caused this slow development. From the weather station we installed at the LWFL site, two gaps in rain events were found in the month of June, one was eight days long and the other six, occurring only two days after the first. However, no specific time interval was set to record grass establishment for these sites, and aerial photographs were taken frequently as grass growth progressed. The vegetation was evaluated once the vegetation cover was relatively constant.

Lastly, to determine the variability in soil chemical properties across the site, soil samples were taken from four locations within each block (Figure 16) and sent to Brookside Laboratories, Inc. (New Bremen, OH, USA) for testing. The analyses included exchangeable cations (Ca, Mg, K, and Na), anions (soluble sulfur and P2O5), extractable minor elements (B, Fe, Mn, Cu, Zn, Al, etc.), total exchange capacity (ME / 100 g), organic matter, and pH (H2O 1:1).

Organic matter was determined using the loss on ignition procedure at 360 °C (Schulte and Hopkins, 1996), total exchange capacity by summation (Ross, 1995), and the extractant used for most cations and anions was Mehlich III (Mehlich, 1984). Additionally, we received a soil

26

analysis report which showed carbon and nitrogen content, as well as C: N ratio on samples with nitrogen higher than 0.05%. Carbon and nitrogen were determined by combustion analysis (McGeehan and Naylor, 1988).

Apex Active DOT Site

Four tackifiers which were at least partially successful in the wind tunnel tests and which represented different product types were selected for testing at the Apex construction site.

These tackifiers were plantago (200% application rate), cellulose HM (100%), Soiltac (100%) and PAM (200%), which were applied at 8.68 (2.29), 43.46 (11.48), 17.83 (4.71) and 215.65 L ha^-1

(56.97 gal ha^-1), respectively. Plantago was applied at 224 kg ha^-1 with a mixing rate of 48 g L^-1. Cellulose HM was applied at 1120 kg ha^-1 with a mixing rate of 48 g L^-1. Soiltac was applied at 243 L ha^-1 with a mixing ratio for water and tackifier of 40:1, and PAM was applied at 112 kg ha^{– 1} with a mixing rate of 0.99 g L^-1. Fertilizer, lime, grass seed and straw were applied by the DOT contractor. Plots were 3 x 6 m (10 x 20 ft) and were installed at the top of the short slope below the pavement. Due to the larger plot size compared to the plots at LWFL, cellulose HM, Soiltac and PAM were applied with a hydroseeder (TurfMaker 420, TurfMaker Corp., Rowlett, TX). To determine the application rates for each product, the spray rate was measured by spraying into a bucket and the spray time was calculated to apply the appropriate amount of product.

Plantago was applied by hand due to its lower application rate. An additional 0.93 x 1.89 m² (10 X 20 ft²) area, adjacent to the 12 test plots, was tackified with emulsified asphalt applied by the

27

contractor, and used for comparison of grass growth. Tackifier was applied on September 16^th, 2016 and aerial photographs of this site were taken on September 29^th and October 31^st, 2016.

Weather Data:

Weather data was recorded on the LWFL site using the HOBO U30 USB Weather Station Starter Kit (Onset, Bourne, MA, USA), with a 0.025 cm (0.01 in) Rain Gauge (2m cable) Smart Sensor (Onset, Bourne, MA, USA) attached. Rainfall during the study appeared to be relatively normal during the spring iteration, with a total rainfall in June of 12.1 cm (4.77 in) and 11.3 cm (4.46 in) in July, where the average in Raleigh was 10.3 (4.06 in) and 11.4 cm (4.49 in), respectively. August was slightly lower than average with 8.1 (3.2 in) as opposed to the average

10.7 cm (4.21 in) (U.S. Climate Data, 2017). Since the HOBO weather station at LWFL was not set up until mid-May, a total precipitation value was not collected for that month. However, according to the North Carolina Climate Retrieval and Observations Network of the Southeast Database from the Lake Wheeler Rd. Field Laboratory, the total precipitation for May, 2016 was

21.9 cm (8.6 in) (North Carolina State Climate Office, 2017). One particularly large storm did occur during the month of May causing some erosion at the LWFL site; however, only block two, and minimally block 1, were affected, so with three replications remaining, we were still able to test differences with the remaining three blocks (Figure 17). During the fall, however, precipitation values were much higher than average. The HOBO weather station at LWFL measured 18.3 cm (7.21 in) of rain during September and 23 cm (9.04 in) in October, approximately 7.1 (2.8 in) and 13.9 cm (5.46 in) higher than average, respectively (U.S. Climate Data, 2017). A large amount of the precipitation recorded for October was due to Hurricane

28

Matthew, which resulted in problematic erosion at the Apex site as previously mentioned. Grass stands at LWFL were seemingly unaffected by the storm.

The LWFL HOBO weather station also provided wind speed data, monitoring both continuous wind speed (steady flow), and wind gusts. The average steady wind speed during this study (June – October) was only 3.57 km h^-1 (2.22 mi h^-1) while the average wind gust speed was 9.11 km h^-1 (5.66 mi h^-1). According to Weather Spark (2017), for Raleigh, NC, seasonal variation in wind speed does occur, with the highest wind speeds occurring in March (6.4 km h^-1 or 4 mi h^-1 on average), and the lowest speeds in August (4 km h^-1 [2.5 mi h^-1]). These average speeds are low, and would not pose a threat to straw mulch failure; however, under stormy conditions wind speeds increase and damaging gusts can occur. The maximum wind speed recorded, using the HOBO weather station, at LWFL under steady conditions was approximately

30.6 km h^-1 (19 mi h^-1), whereas the maximum gust speed was 62.8 km h^-1 (39 mi h^-1). Wind data was also collected from the Apex field site, and values for steady and gusted winds were similar to those at LWFL, with average speeds at 3.7 and 8.9 km h^-1 (2.3 and 5.5 mi h^-1) for steady and gusted, respectively, and maximum gust speeds in the 45 – 55 km h^-1 (30-mi h^-1) range. At these speeds straw mulch failure could occur, even with some of the tackifiers tested in this project, but there was no evidence that it did.

Image Processing Using ArcGIS and Data Analysis:

After collecting an aerial image from the field, the photograph was cropped to show only the plot area, and saved in .tif format. Next, after starting ArcMap (Esri, version 10.3.1),

29

the cropped photograph was opened using “Add Data”. However, in order to analyze individual

plots within the photo a new shape file was created to isolate each plot. This was done by opening ArcCatalog and selecting “New” and then “Shapefile”. This prompted a window to appear, with a drop down menu to select the feature type. “Polygon” was selected before

clicking “Ok” to create the new shapefile. Once the shapefile was created, and “Start Editing” had been selected, the “Create Features” tool was used to outline the entire testing area, either using the selection for rectangle or polygon. Next the “Cut Polygon” tool was selected to divide the total testing area into individual treatment plots. Each plot was then labeled by opening the attribute table, adding a field and naming each square with its respective treatment.

The training sample manager was used to sample a selection of pixels, specifically sampling colors for grass and for bare soil areas. At least 15 to 20 samples were taken for each color (grass, soil, spray paint, etc.) to ensure accuracy and guarantee that each color was represented fully.

Once each color has been carefully sampled, a visual display was created using

“Interactive Supervised Classification”, under “Classification”. The display helped to determine if more color samples were needed, or if the display accurately matched the original image (Figure 18). This type of analysis is used commonly to classify various different parts of an image (Parece et al., N.D.), such as distinguishing between land and water or trees from grass. When the display became an accurate representation of the original grass and soil areas, the image was saved as a grid file using the “Export Data” selection. Upon reopening the file in grid format the “Tabulate Area” tool was selected in ArcToolbox under “Spatial Analyst” then “Zonal”. It is

30

important at this point that the projection was defined as “State Plane (ft.)”, which was accomplished by selecting “Define Projection” under “Data Management” and “Projections and Transformations”. The final output table was saved as a text file where it was then imported into excel in order to sum row values and calculate the percentage of green color (representing vegetation) that appeared in each plot.

From excel the data was imported into SAS 9.4 for one-way ANOVA, where p-values of

0.05 or less indicated significant differences. Nested and two-way ANOVA were not included in this analysis since many tackifiers were only tested at one application rate, instead of all three, which limited degrees of freedom.

Results and Discussion:

During the spring evaluation at LWFL, no differences in grass cover were found between any of the tackifier treatments, application rates or blocks (Table 14). Additionally, rates (50, 100 or 200%) of application also did not affect grass stands. The aerial image used for this analysis was collected on July 29th, 2016, approximately 3 months after grass seed was planted, and on average the vegetation cover was 61%.

Lemly (1982) compared asphalt-tackified straw to jute netting, mulch blanket, wood chips and stapled excelsior blanket and found that over a 3 month period tall fescue cover was approximately 75% for the asphalt treatment, which was similar to our 79% grass cover on July 29^th, also for emulsified asphalt. Although this study is not applicable to other tackifiers, it

31

shows growth consistency to another study for the emulsified asphalt treatment, and the general range for grass cover after a three month time period.

During the fall evaluation at LWFL, there were differences in blocks, tackifiers and application rates. In the data collected 13 days after seeding emulsified asphalt, cellulose HM, plantago, Soiltac and Tornado Tack had higher vegetation cover than bonded fiber HM, wood fiber HM and “No Tack” plots when including all rates (Figure 19). At the 50% application rate, bonded fiber HM had a higher vegetation cover than wood fiber HM with 9.5% cover compared to 3.9%. Emulsified asphalt had more grass coverage than wood fiber HM, bonded fiber HM and straw alone at the 100% rate, and at two times the recommended rate there were no differences in grass establishment. However, any differences in ground cover disappeared by 33 days after seeding, when the average grass cover was 56%.

Although there were some distinctions in grass cover between tackifiers under the recommended and 50% recommended rates, it seems likely that these effects were not due to inhibition by tackifier, but rather due to other factors. There were no differences at the 200% tackifier application rate, suggesting that the tackifiers were not inhibiting growth. No differences occurred in the spring iteration or after the initial few weeks in the fall. It is also possible that since only a few weeks had passed when the differences were found between tackifiers in the fall, these tackifiers may be positively affecting grass growth, instead of other tackifiers inhibiting growth. It is imperative that the soil retain its moisture during seed germination and early growth (Ayers, 1952), and therefore it could be possible that certain tackifiers offer more initial moisture or are better able to retain moisture. Tilley and St. John (2013) compared Nebraska sedge (Carex nebrascensis) seed application methods, either dry

32

broadcast or hydroseeded (using Turbo Tack tackifier), and seedbed preparation method, traditional vs. slurry method. Both methods were prepared on unlined ponds, with the traditional method including tillage and soil packing to the point where a boot print would sink no deeper than 13 mm. The slurry method, however, involved flooding the pond and allowing it to dry just enough for foot traffic, before planting. No differences were found between the seed application methods under traditional seedbed preparation, but with the slurry method they found that after one month areas seeded by hydroseeder had approximately 250% more plants and after three months that figure had increased to 465% more plants. It is possible that these results could also be related to soil moisture due to hydromulch cover.

The soils in the blocks were not different in organic matter, pH, soluble sulfur, sodium, potassium and magnesium. Block 2 had significantly more carbon than blocks 3 and 4, with 0.573% carbon as opposed to 0.328% and 0.338% respectively. Calcium and total exchange capacity also showed differences between blocks, where block 2 had higher values than block 4 for both variables. These results also provided general information about the soil for this study as a whole, with the site average pH of 7.54 and 1.3% organic matter. However, none of these variations in soil fertility seemed to affect the results in any obvious way. During the spring replication no differences were found between blocks, and in the fall replication the block effects changed over time.

The Apex site was prematurely terminated due to a severe erosion during Hurricane Matthew on October 8^th (Figures 21, 22), during which the estimated rainfall at the site was >20 cm (8 in). The erosion during this storm event was due primarily to the lack of a curb to protect the slope from the paved road runoff. However, one set of aerial photographs (Figure 23) was

33

taken prior to Hurricane Matthew 13 days after planting, which was processed through the GIS software in the same manner as the LWFL plots. There were no differences in grass cover between any of the tackifier treatments (Figure 24). This indicated that these tackifier treatments did not have different effects on grass germination or growth.

Conclusion:

The main conclusion from the field plot study was that none of the alternative tackifiers seemed to have a negative impact on grass germination or growth compared to the emulsified asphalt standard. In addition, under the weather conditions experienced during the three tests, there were no failures of the tackifiers in maintaining the straw cover. Due to the lack of statistical differences (especially expected differences, such as emulsified asphalt) as well as limited grass growth, additional testing was needed to validate these results.

34

Chapter 4: Greenhouse Trial

The objective of this study was to expand upon the outdoor vegetation testing of the potential for tackifiers to inhibit grass growth, but under more controlled conditions. Each sample used the same soil, in the same sized tray, received the same amount of water, approximately the same amount of light, and did not experience any wind. This study also delved deeper into growth inhibition by separating the grass species in order to observe any differences for each species. After observing few differences during the outdoor vegetation study, tackifiers were hypothesized to have no effect on grass growth.

Methods:

The greenhouse bioassay study was initiated in December 2016. Instead of using a grass seed mix, each grass species was tested individually so that any species-specific effects could be isolated. The same grass species, tall fescue (Raptor II), bermudagrass (hulled, RFLB)) and centipedegrass, were used in the greenhouse trial as in the outdoor vegetation study, and were planted at a rate of 200 seeds per tray. All grass seed was purchased from Wyatt-Quarles Seed Company (Garner, NC, USA), where all seed is tested for purity and quality. The slotted flats, or trays, were purchased from Corr Farm Supply in Smithfield, NC, with dimensions of 25.4 x 52.1 x 6.35 cm (10 x 20.5 x 2.5 in) for a total of 84.03 m³ (42.71 ft³) (Figure 25).

After the trays were filled with potting soil (Pro-line C/B, Jolly Gardner, Oldcastle, Inc., Atlanta, GA, USA), grass seed, straw, and tackifier application all occurred on December 12^th,

35

2016 (Figure 26). No specific density of soil was used, but all trays were filled completely. To simplify the 200 seed per tray application rate, 200 seeds of each species were counted out 3 times and weighed to determine average weight for that number of seeds (Table 15). Seeds were then added to the trays by weight to approximate 200 seeds. Due to the small size of the bermudagrass and centipedegrass seeds, approximately 0.5 ounces of sterile sand was mixed in with the 200 seeds per tray to aid in even application. Seed was applied to the surface of the soil and was gently raked in by hand, followed by straw at an equivalent rate of 4400 kg ha^-1 (4000 lbs ac^-1).

The tackifiers included in this study were selected based on their successful reduction in straw failure in the wind tunnel testing. These included plantago, cellulose HM, wood fiber HM, bonded fiber HM, PAM, Soiltac, Tornado Tack, emulsified asphalt plus a no tackifier control. All were applied by hand at double the recommended rate. Each treatment was replicated three times with 3 grass species and 9 tackifiers, for a total of 81 trays. Due to a temperature gradient in the greenhouse, a randomized, complete block design was implemented to cover the expected ranges in temperature (Table 16). The application rate was accidentally doubled for all plantago treatments and for two replications of Soiltac (Table 16). The emulsified asphalt was applied by a contractor three days after all other tackifiers. To prevent grass growth prior to asphalt application, the asphalt-tackified trays were not watered until they received the tackifier, and all data was adjusted to correct the delayed application. This correction was applied by shifting the results backwards by two days, meaning that data collected on day 18, for example, would be recorded as being collected on day 16 instead.

36

The trays received 335 mL of water per day (0.1 in/day) throughout the test period. Blade height and count measurements were taken on alternating days. Grass blade count included fully matured blades as well as barely visible, newly emerged blades. Blade height was measured on the tallest three blades in each tray and averaged. If less than three blades were present, the average was calculated using only the number of blades that were present. And grass emergence was calculated by dividing the number of grass blades per tray by 200 (the estimated total number of seeds planted).

Data was recorded for all grass species until January 2^nd, 2017, 22 days after planting. At this time tall fescue was nearing 100% emergence; however, centipedegrass and bermudagrass had only produced a few blades per tray. Watering ceased on January 2^nd; however on January 11th centipedegrass and bermudagrass appeared to have continued growing, so to accommodate this additional growth watering was resumed and the test was extended until January 24th (44 days) for these species. Measurement frequency for blade count was decreased during this time due to the slow growth, and height data was only recorded during the final day of testing. Blade counts were taken every three days, ending on January 20th, and on January 24th blade height was recorded for the last time.</sup>

Data was entered into an Excel (Microsoft Office, 2010) and statistical analyses were performed using SAS 9.4. Comparisons were made between tackifiers, grass species, and blocks, with significant differences determined using one-way ANOVA (p ≤ 0.05). However, due to limited growth for the bermudagrass and centipedegrass, no interactions were analyzed between tackifiers and grass species. Additionally, a compound symmetry repeated measures analysis was used to determine changes in variance over time. Compound symmetry was

37

chosen as it had the lowest AIC (Akaike information criterion) value of 1269. AIC demonstrates the quality of a model and helps to determine which model is the best fit for the data.

Results and Discussion:

Tall Fescue:

Overall, there were few significant differences in grass establishment due to tackifier treatments effects. Differences were evident on only 37% of the eight days when blade counts were performed. Average maximum blade height, and the increase in blade count also had few differences (12 and 29% of days, respectively), during the course of the testing. Seedling emergence was different for 62% of the measurement days, but the numeric differences were relatively small. Overall, only 26% of the measurements for tall fescue had differences in tackifier treatments, suggesting that tall fescue growth was relatively unaffected by the tackifier applied to the straw. There were no differences in blade count between tackifiers

when all days were included (p ≤0.05), nor was there an interaction between tackifier and day; however, there was an day effect (p &lt; 0.0001)(Table 17).

Differences in blade count did not occur until days 16, 18 and 20 (Table 18). On day 16, Tornado Tack HM had more grass blades than wood fiber HM, cellulose HM and bonded fiber HM. On day 18, Tornado Tack had a higher blade count than cellulose HM and bonded fiber HM. By day 20 the only difference in blade count was between Tornado Tack and cellulose HM with 158 blades. The maximum average blade height was only different on day 10, with Tornado Tack having taller blades than emulsified asphalt. The average blade height for

38

Tornado Tack on day 10 was 80.2 cm, whereas the average for emulsified asphalt was 32 cm, and the mean for all tackifiers was 61.2 cm.

Differences in tall fescue grass emergence were evident on days 9, 11, 16, 18 and 20 (Figure 27 and Table 19). Early emergence (days 9 and 11) was generally much worse for the emulsified asphalt than the other tackifiers. Later (days 16, 18, 20) Tornado Tack HM, tended to have higher emergence than the other tackifiers, but emergence in the emulsified asphalt caught up with all but the Tornado Tack HM. Differences were relatively narrow toward the end of the test, ranging from 67 -77% emergence for tall fescue. Barkley et al. (1965), compared grass emergence when mulched with straw, Turfiber (a wood cellulose fiber), saw dust, an elastomeric polymer emulsion called Soilset, and no mulch, finding it lower in the emulsion or no mulch treatments. All three grass species tested (Kentucky 31 fescue, Kentucky bluegrass, and redtop) had similar responses to the mulching treatments. These results seem related as Soilset and emulsified asphalt are both black emulsions, and although varying in composition, both had lower grass emergence. Although tall fescue was not specifically tested in the Barkley et al. experiment, the effect of tackifiers on emergence was consistent with the results from this study.

The change in number of grass blades between the measurement days indicated the rate of growth. Differences were found between days 9 and 11 as well as 11 and 14, where growth in the emulsified asphalt treatments was much less between days 9 and 11, and significantly more than other tackifiers between days 11 and 14 (Figure 28). This result may be because emulsified asphalt was one day behind the other tackifiers, and that rapid growth generally occurs at a certain stage in the emergence process. On days 9-11 only 14 additional

39

grass blade emerged in the emulsified asphalt treatment compared to 62 averaged among the other treatments. However, between days 11 and 14, grass emergence increased by 108 blades in the emulsified asphalt treatment compared to only 25 blades averaged across all other treatments.

Centipedegrass and Bermudagrass:

Both centipedegrass and bermudagrass species had much slower growth than tall fescue. Even after extending the test for an additional 3 weeks, the growth still did not reach the same level, with only 43 blades out of 200 (21.6%) emerged on day 40 for bermudagrass and less than 4 (1.9%) for centipedegrass on average (Figure 29). Due to this low growth, no comparisons were made for the centipedegrass species. No significant differences were found between any of the tackifier treatments for the bermudagrass.

The slower growth of both centipedegrass and bermudagrass was most likely due to the fact that these grasses are warm season grasses, whereas tall fescue is a cool-season perennial. Since the testing was conducted in December, the bermudagrass and centipedegrass did not receive enough light, due to shorter day lengths, and therefore exhibited stunted growth. This is likely due to fescue having greater growth potential under winter sunlight conditions compared to warm-season grasses (McCarty, 2001). Decreased light and photosynthetic input can cause reduced carbohydrate storage and therefore sparse stands, as seen in this study (Barrios et al., 1986 and Beard, 1969).

40

Conclusion:

The results from this study were similar to the outdoor vegetation studyin that tackifier type had little effect on grass emergence or growth. For tall fescue, there was an interaction between tackifier treatment and day, and differences between tackifiers were found on certain days. However, at most measurement periods there were no differences between grass blade count or maximum blade height. Centipedegrass and Bermudagrass had limited growth due to short winter days, so full conclusions cannot be made about these species. However, when combining results from the outdoor vegetation study, which included centipedegrass and Bermudagrass, and this greenhouse trial, it does not appear that tackifier type has a negative effect on grass emergence.

41

Chapter 5: Conclusions and Recommendations

The main goal of this study was to determine viable straw tackifier options to replace emulsified asphalt. Specifically, a replacement tackifier would need to meet four main qualifications: 1) effectively withstand wind and rain events, 2) quick and easy application, 3) no interference with grass germination or growth, and 4) similar or lower cost. All tackifiers tested in this project met the ease of application qualification, as all are capable of hydraulic application with a hydroseeder, and each other aspect was tested in the three studies conducted in this project.

There were a number of alternative tackifiers that were able to effectively withstand dry and steady winds at or above 56 km h^-1 (35 mi h^-1), including cellulose HM, bonded fiber HM, Tornado Tack, wood fiber HM, and Soiltac at the recommended rate. Plantago also achieved good wind resistance at 2X the recommended rate. Under dry and gusty conditions, a number of products withstood winds of 56 km h^-1 (35 mi h^-1) or greater when applied at 200% of the recommended rate: plantago (42 mph), BF (42 mph), cellulose HM (38 mph), wood fiber HM (35 mph), and Tornado Tack (35 mph). A tackifier with an ability to withstand both steady and gusty winds is important, as gusty winds, common during storms or when frontal systems pass an area, were shown to cause straw failure at lower wind speeds. Although there was variation in the performance of these products, most of them provided similar protection from wind erosion to emulsified asphalt. Tracer and polyacrylamide were generally ineffective under most conditions, while Soiltac was only moderately effective.

42

There was little evidence that the tested tackifiers had much effect on grass growth, and early growth was generally better with these compared to emulsified asphalt.

The hydromulches were generally effective at the current recommended rate of 1,140 kg ha^-1 (1,000 lb ac^-1), and plantago was effective at the 2X (225 kg ha^-1 or 200 lbs ac^-1) rate.

Comparing the cost per acre, plantago costs approximately $465 per hectare at the high rate. Cost for the recommended rates of cellulose HM (Country Boy) were $450 ha^-1, Tornado Tack HM were $410 ha^-1, wood fiber HM (Conwed 1000) were $615 ha^-1, and bonded fiber HM (FlexTerra) were $2070 ha^-1. Since all of these were equally effective, the lower cost products plantago, Tornado Tack and cellulose HM would be recommended to replace emulsified asphalt. Compared to the asphalt tackifier these products are all easier to apply, have fewer environmental concerns, do not hinder vegetation emergence or growth and are significantly less expensive. Each of these products offers protection from straw mulch failure at wind speeds up to 68 km h^-1 (42 mi h^-1) under dry and steady conditions and 56 km h^-1 (35 mi h^-1) under dry and gusty conditions.

43

References:

Adams, J.E. 1966. Influence of mulches on runoff, erosion, and soil moisture depletion. Soil Sci Soc Am J. 30:110-114.

Advanced Thermal Solutions, Inc. 2012. Some basic principles of wind tunnel design. https://www.qats.com/cms/2012/07/17/some-basic-principles-of-wind-tunnel-design/

Alabama Forestry Commission. 1993. Alabama’s best management practices for forestry. Alabama Forestry Commission: Montgomery, AL. http://www.forestry.alabama.gov/BMPIndex.aspx?bv=2&amp;s=1%20.

Ayers, A. D. 1952. Seed germination as affected by soil moisture and salinity1. Agron J. 44:82- 84. doi:10.2134/agronj1952.00021962004400020006x.

Barkley, D. G., R.E. Blaser, and R.E. Schmidt. 1965. Effect of mulches on microclimate and turf establishment. Agron J. 57:189-192. doi:10.2134/agronj1965.00021962005700020016x

Barrios, E. P., F.J. Sundstrom, D. Babcock, and L. Leger. 1986. Quality and yield response of four warm-season lawngrasses to shade conditions. Agron J. 78:270-273. doi:10.2134/agronj1986.00021962007800020011x

Beard, J.B. 1969. Turfgrass shade adaptation. p. 273-282. In E.C. Robers (ed.) Proc. of the 1^st Int. Turfgrass Res. Conf., Alf Smith and Co., Bradford, UK. archive.lib.msu.edu/tic/turfx/article/2002jan6.pdf.

Bethlahmy, N., and W.J. Kidd Jr. 1966. Controlling soil movement from steep road fills. INT-45. USDA Forest Service, Intermountain Research Station: Ogden, UT.

Blaser, R. E. 1962. Soil mulches for grassing. p. 15-20. In: Roadside development. Pub. 1030. Highway Research Board, NAS-NRC.

44

Braun, George F. 2003. Material safety data sheet- Reinco Tracer Tackifier. https://www.epicmanufacturing.com/images/pdf/Reinco…/MSDS_Tracer_Tackifier.pdf

Brofas, G., and C. Varelides. 2000. Hydro-seeding and mulching for establishing vegetation on mining spoils in greece. Land Degrad Dev. 11(4):375-382. doi:10.1002/1099- 145x(200007/08)11:4&lt;375::aid-ldr400&gt;3.3.co;2-7.

Brown, D. and R. Hallman. 2009. Reclaiming disturbed lands. U.S. Department of Agriculture, Forest Service, Equipment Development Center.

Brown, R.N., and J.H. Gorres. 2011. The use of soil amendments to improve survival of roadside grasses”. HortScience 46(10):1404-1410.

Burroughs, E.R. Jr., and J.G. King. 1989. Reduction of soil erosion on forest roads. INT-264. USDA Forest Service, Intermountain Research Station: Ogden, UT. https://www.fs.fed.us/rm/pubs_int/int_gtr264.pdf.

U.S. Climate Data. 2017. Climate Raleigh – North

Carolina. www.usclimatedata.com/climate/raleigh/north-carolina/united-states/usnc0558.

Corbineau, F., C. Gay-Mathieu, D. Vinel and D. Côme. 2002. Decrease in sunflower (Helianthus annuus) seed viability caused by high temperature as related to energy metabolism, membrane damage and lipid composition. Physiologia Plantarum, 116: 489–496. doi:10.1034/j.1399- 3054.2002.1160407.x.

Dahmer, Adam. n.d. Understanding the carbon: nitrogen ratio (C:N). Advance Cover Crops. www.advancecovercrops.com/understanding-the-carbon-nitrogen-ratio-cn/.

Dudeck, A.E., N.P. Swanson, L.N. Mielke and A.R. Dedrick. 1970. Mulches for grass establishment on fill slopes. Agron J. 60(6): 810-812.

45

Emanual, David M. 1976. Hydromulch: a potential use for hardwood bark residue. Research Note NE- 226. Upper Darby, PA. U.S. Dept. of Agriculture, Forest Service, Northeastern Forest Experiment Station.

Faucette, L.B., C.F. Jordan, L.M. Risse, M. Cabrera, D.C. Coleman and L.T. West. 2005. Evaluation of storm water from compost and conventional erosion control practices in construction activities. J Soil Water Conserv. 60(6):288-297.

Gholami, L., K. Banasik, S.H. Sadeghi, A.K. Darvishan and L. Hejduk. 2014. Effectiveness of straw mulch on infiltration, splash erosion, runoff and sediment in laboratory conditions. Journal of Water and Land Development, 22(1): 51-60. doi:http://dx.doi.org/10.2478/jwld-2014-0022

Gilbert, W. B., and D.L. Davis. 1967. An investigation of critical problems of establishing a satisfactory sod cover along North Carolina highways. School of Engineering, N.C.S.U. and N.C. Div. of Highways. Final Report ERD-lOO-S, Raleigh, N.C.

Gregory, J.H., M.D. Dukes, P.H. Jones, and G.L. Miller. 2006. Effect of urban soil compaction on infiltration rate. J Soil Water Conserv. 61(3):117-124.

Grace, J.M, III. 2000. Forest road sideslopes and soil conservation techniques. J Soil Water Conserv. 55: 96-101.

Grigg, A.H., G.J. Sheridan, A.B. Pearce and D.R. Mulligan. 2006. The effect of organic mulch amendments on the physical and chemical properties and revegetation success of a saline-sodic minespoil from Central Queensland, Australia. Aust J Soil Res. 44(2):97-105.

Hauser, V. 1989. Improving grass seedling establishment. J Soil Water Conserv. 44(2):153-156. http://www.jswconline.org/content/44/2/153.abstract.

Hernández, M.A.G., A.I. Lopez, A.A. Jarzabek, J.M. Perales, Y. Wu and S. Xiaoxiao. 2013. Design Methodology for a Quick and Low-Cost Wind Tunnel, Wind Tunnel Designs and Their Diverse Engineering Applications, Dr. Noor Ahmed (Ed.), InTech, DOI: 10.5772/54169. Available from:

46

https://www.intechopen.com/books/wind-tunnel-designs-and-their-diverse-engineering- applications/design-methodology-for-a-quick-and-low-cost-wind-tunnel

Jordan, C.F. 1998. Working with nature: resource management for sustainability. Hardwood Academic Publishers, Amsterdam, Netherlands.

Kay, Burgess L. 1978. Mulch and chemical stabilizers for land reclamation in dry regions. In Reclamation of Drastically Disturbed Lands, (F.W. Schaller and P. Sutton eds.), Am Soc Agron, Crop Sci Soc Am J., and Soil Sci Soc Am J., Madison, WI, p. 467 – 483.

Keammerer, W.R., L.F. Brown and Colorado Water Resources Research Institute. 1988. High altitude revegetation workshop. Vol. 8. dspace.library.colostate.edu/bitstream/handle/10217/3138/is_59.pdf?sequence=1#page=250.

Krichen, K., H.B. Mariem and M. Chaieb. 2014. Ecophysiological requirements on seed germination of a Mediterranean perennial grass (Stipa tenacissima L.) under controlled temperatures and water stress. S Af J Bot. 94: 210-217.

Lee, G. 2012. Comparison of five hydromulches and polyacrylamide to straw for erosion control and vegetation establishment on steep slopes. MS Thesis, North Carolina State University, Department of Soil Science. http://catalog.lib.ncsu.edu/record/NCSU2713517.

Lemly, A. Dennis. 1982. Erosion control at construction sites on red clay soils. Environ Manage. 6(4):343-52. https://link.springer.com/article/10.1007/BF01875066.

Logsdon, S.D. 2013. Root effects on soil properties and processes: synthesis and future research needs. In: Timlin T, Ahuja LR (eds) Enhancing understanding and quantification of soil–root growth interactions. Adv Agric Syst Model 4. ASA, CSSA, SSSA, Madison, WI. p. 173-196. doi:10.2134/advagricsystmodel4.c8

Marques, M.J., R. Bienes, L. Jiménez and R. Pérez-Rodríguez. 2007. Effect of vegetal cover on runoff and soil erosion under light intensity events. Rainfall Simulation Over USLE Plots. The Science of the total environment 378 (1-2):161-165.

47

McCarty, L.B. 2001. Best golf course management practices. Prentice Hall, Upper Saddle River, NJ.

McGeehan, S.L., and D.V. Naylor. 1988. Automated instrumental analysis of carbon and nitrogen in plant and soil samples. Commun. Soil Sci. Plant Anal. 19:493-505.

McKee, W.H. Jr., R.E. Blaser and D.G. Barkley. 1964. Mulches for steep cut slopes. Highway Research Board. 53:35-42.

McKee, W.H. Jr., A.J. Powell Jr., R.B. Cooper and R.E. Blaser. 1965. Microclimate conditions found on highway slope facings as related to adaptation of species. Highway Research Board 93:38-43.

Megahan, W.F. 1974. Erosion over time on severely disturbed granitic soils: a model. INT-156. USDA Forest Service, Intermountain Forest and Range Experiment Station: Ogden, UT.

Megahan, W.F., S.B. Monsen and M.D. Wilson. 1991. Probability of sediment yields from surface erosion on granitic roadfills in Idaho. J Environ Qual. 20:53-60.

Mehlich, A. 1984. Mehlich-3 soil test extractant: A modification of Mehlich-2 extractant. Commun. Soil Sci. Plant Anal. 15:1409-1416.

NCDOT (1), Division of Highways, Roadside Environmental Unit. n.d. NCDOT general seed specification for seed quality. www.ncdot.gov/doh/operations/dp_chief_eng/roadside/soil_water/pdf/SeedQuality.pdf.

NCDOT (2), Division of Highways, Roadside Environmental Unit. n.d. Seed mix, crimping, guardrail seed to be used by county. www.ncdot.gov/doh/operations/dp_chief_eng/roadside/soil_water/pdf/SeedMixTable.pdf.

48

NCDOT (3), Division of Highways, Roadside Environmental Unit. n.d. Vegetation management section. www.ncdot.gov/doh/operations/dp_chief_eng/roadside/vegetation/materials/turf.html.

NCDOT (4), Division of Highways. n.d. Soil stabilization timeframes. https://www.ncdot.gov/doh/…/dp…/StabilizationGuidelinesMaintenance_detail.pdf.

NCDOT (5), Division of Highways, Roadside Environmental Unit. 1998. Soil stabilization level 2: skill block training guide REU 155. https://www.ncdot.gov/doh/operations/dp_chief_eng/roadside/vegetation/training/manuals/ Skillblock2R8REU155.doc+&amp;cd=1&amp;hl=en&amp;ct=clnk&amp;gl=us.

NCDOT (6), Division of Highways, Roadside Environmental Unit. 1998. Soil stabilization level 3: skill block training guide REU 156. 21 Aug. https://www.ncdot.gov/doh/…/dp…/training/manuals/SkillBlock3R10REU156.doc.

NCDOT (7), Division of Highways, Roadside Environmental Unit. 2016. Stabilization requirements. https://connect.ncdot.gov/letting/…/Z-06- 19…/DG00207PropErosionControl.pdf.

North Carolina Department of Environment and Natural Resources, Division of Water Quality. 2011. Construction General Permit NCG 01, Section II.B.2. http://portal.ncdenr.org/c/document_library/get_file?uuid=d15e7ee8-cb50-443f-9dcf- 049b001337ac&amp;groupId=38364.

Osborn, B. 1955. Effectiveness of cover in reducing soil splash by raindrop impact. J Soil Water Conserv. 10(1):70-76.

Pan, C., Z. Shangguan and T. Lei. 2006. Influences of grass and moss on runoff and sediment yield on sloped loess surfaces under simulated rainfall. Hydrol. Process., 20: 3815–3824. doi:10.1002/hyp.6158

49

Parece, T., J. Campbell and J. McGee. N.D. Remote sensing in an ArcMap environment. Chapter 19: Classification of a landsat image (supervised). VirginiaView. http://virginiaview.cnre.vt.edu/tutorial/Chapter_19_Classification%20of%20a%20Landsat%20I mage%20%28Supervised%29.pdf

Pelt, R.S.V. and T.M. Zobeck. 2013. Portable Wind Tunnels for Field Testing of Soils and Natural Surfaces. Wind tunnel designs and their diverse engineering applications. Intech. http://dx.doi.org/10.5772/54141.

Richard, G., I. Cousin, J.F. Sillon, A. Bruand and J. Gue’rif. 2001. Effect of compaction on soil porosity: consequences on hydraulic properties. Eur J Soil Sci. 52:49-58.

Ross, D. 1995. Recommended soil tests for determining exchange capacity. p. 62-69 In J.T. Sims and A. Wolf (eds.) Recommended soil testing procedures for the northeastern United States.

Northeastern Regional Bulletin #493. Ag Experiment Station, University of Delaware, Newark, DE.

Ross P.J, J. Williams and R.L. Mccown. 1985. Soil temperature and the energy balance of vegetative mulch in the semi-arid tropics. I. Static analysis of the radiation balance. Aust J Soil Res. 23(4):493-514.

Schulte, E.E., and B.G. Hopkins. 1996. Estimation of soil organic matter by weight Loss- OnIgnition. p. 21-32. In: Soil organic matter: Analysis and interpretation. (ed.) F.R. Magdoff,

M.A. Tabatabai, and E.A. Hanlon, Jr. Special publication No. 46. Soil Sci. Soc. Am. Madison, WI.

Sittig, M., R.P. Pohanish. 2002. Asphalt Fumes. Sittig’s Handbook of Toxic and Hazardous Chemicals and Carcinogens (4th Edition). William Andrew Publishing/Noyes.

Smale, M., M.R. Bellon, D. Jarvis and B. Sthaphit. 2004. Economic concepts for designing policies to conserve crop genetic resources on farms. Genet Resour Crop Ev. 51:121–135.

Soiltac safety data sheet. 2015. Scottsdale, AZ. /media/23415/soiltac-material-data-safety-sheets-en-.pdf.

50

Sojka, R.E., D.L. Bjorneberg, J.A. Entry, R.D. Lentz and W.J. Orts. 2007. Polyacrylamide in agriculture and environmental land management. San Diego: Elsevier Academic Press Inc. 92:75-162.

State Climate Office of North Carolina. 2017 Lake Wheeler Rd Field Lab, NC (LAKE): monthly data retrieval. CRONOS Database. climate.ncsu.edu/cronos/?station=LAKE&amp;temporal=monthly.

Swanson, N. P., A.R. Dedrick, H.E. Weakly and H.R. Haise. 1965. Evaluation of mulches for water erosion control. Trans Am Soc Agric Eng. 8:438-440.

Swemmer, A.M., A.K. Knapp and M.D. Smith. 2006. Growth responses of two dominant C4 grass species to altered water availability. Int J Plant Sci. 167(5):1001-1010.

Swift, L.W. Jr. 1984. Soil loss from roadbeds and cut and fill slopes in the southern Appalachian Mountains. South J Appl For. 8(4):209-215.

Swift, L.W. Jr. 1985. Forest road design to minimize erosion in the southern Appalachians. In:

B.G Blackman (Ed.) Proceedings Forest and Water Quality: A mid-South symposium. University of Arkansas: Little Rock, AR. 141-151.

Texas Department of Transportation, Construction Division. 2006. Asphalt emulsions. ftp://ftp.dot.state.tx.us/pub/txdot-info/cst/AsphaltEmulsions.pdf.

Thomas, H., A.R. James, and M.W. Humphreys. 1998. Effects of water stress on leaf growth in tall fescue, Italian ryegrass and their hybrid: rheological properties of expansion zones of leaves, measured on growing and killed tissue. J Exp Bot. 50(331): 221-231.

Tilley, Derek J., and St. John Loren. 2013. Hydroseeding improves field establishment of Nebraska sedge regardless of seed treatment. Home Organization Selection. Project Muse. Web. 10 July 2017. http://npj.uwpress.org/content/14/2/89.abstract.

51

U.S. Department of Agriculture Natural Resources Conservation Service (USDA-NRCS). 2000. Urban soil compaction. Urban Technical Note No. 2. U.S. Department of Agriculture, Natural Resources Conservation Service, Soil Quality Institute, Auburn, Alabama.

U.S. Environmental Protection Agency (USEPA). 2000. Stormwater Phase II Final Rule: construction site runoff control minimum control measure. Office of Water (4203). EPA 833-F- 00-008, Fact Sheet 2.6. U.S. EPA, Washington, D.C.

U.S. Environmental Protection Agency (USEPA). 2009. Construction and development, final effluent guidelines. U.S. EPA, Washington, D.C. https://www.epa.gov/eg/construction-and- development-effluent-guidelines.

WeatherSpark.com. 2017. Average weather in Raleigh, North Carolina, United States, year round – Weather Spark. weatherspark.com/y/20170/Average-Weather-in-Raleigh-North- Carolina-United-States

52

Tables:

Table 1: Legal timeframes for Department of Transportation site stabilization in North Carolina (NCDOT (4), 2016)

53

Table 3: Products tested in this study as tackifiers.

54

55

56

57

58

59

60

61

Table 15: Weights of 200 seeds from each grass species to be used in the greenhouse trial.

62

Table 16: Greenhouse plot layout for treatments and blocks.

Key: PAM = polyacrylamide, NT = no tack, BF = bonded fiber FGM, WF = wood fiber HM, CH = cellulose HM, ST = Soiltac, Plan = Plantago, Em = emulsified asphalt, TT = Tornado Tack HM, Cent = centipedegrass , Tf = tall fescue, Berm = Bermudagrass, (x2) = double tackifier application

Green – Block 1 Blue – Block 2 Purple – Block 3

63

64

65

Figures:

Figure 1: Illustration of tackifiers function with a straw cover. Yellow represents the straw and green the tackifier, which can both bind the straw together (1) and bind the straw to the soil (2).

Figure 2: View of the wind tunnel through the section connecting the fan to the tunnel, with laminar flow baffles.

66

Figure 3: View of the wind tunnel with the laminar flow baffle section attached.

Figure 4: Bare soil box inside wind tunnel. Also shown is the permanently affixed straw that has been glued to the bottom of the tunnel.

67

Figure 5: Illustration of potential wind flow paths with and without a base straw layer on the floor of the wind tunnel. The top figure illustrates flow paths with a base layer of straw, while the bottom figure illustrates flow paths with no straw glued to the tunnel floor.

68

Figure 6: (a) Smoke emitter affixed to the wind tunnel for testing the flow pattern. (b) Red smoke shows wind pattern inside the tunnel.

69

Figure 7: Fully operational wind tunnel with the fan properly attached and a bare soil box inside. The orange wind speed meter sits on top, exterior to the box, with only the probe on the inside.

Figure 8: A before (left) and after (right) view of a soil test box with an application of

hydraulic tackifier. This illustrates a “failure” with less than 50% of the straw remaining.

70

Figure 9: Variation in failure wind speed of tackifier application rates (% of manufacturer’s recommend rate) at two moisture conditions (wet and dry). Significant differences in tackifier application rate are shown with letters (p&lt;0.05). Bars represent standard error.

71

Figure 10: Effect or tackifier and application rate on failure wind speed under wet moisture conditions. There were no statistical differences between any tackifiers or application rates (p&lt;0.05).

PAM= Polyacrylimide, CHM = Cellulose HM, BF= Bonded Fiber HM, TT = Tornado Tack, and

72

73

Figure 12: Application rate effect on failure wind speed under steady flow wind conditions. Statistical analyses for dry and steady results provided in Table 6. Bars represent standard errors.

74

Figure 13: Differences in tackifier treatments at the recommended rate for dry straw and steady conditions. Differences (p&lt;0.05) are indicated if values do not have a common letter. Key: No Tack = No Tackifier, PAM = Polyacrylamide

75

Figure 14: Cost-benefit analysis for each tackifier under dry straw conditions, at three application rates, as reflected in the cost. Emulsified asphalt is shown as a straight line since it is only applied at one rate. The light blue box represents tackifier and application rates that would be recommended to replace emulsified asphalt.

76

Figure 16: The locations of each of the four soil samples (S1-S4) taken in each block at the Lake Wheeler Field Laboratory.

Figure 17: Photographs of the spring field test of tackifiers at the Lake Wheeler Field Laboratory, after a heavy rainstorm approximately one month after planting. The photograph on the left (a) shows the most extreme erosional damage, the top right (b) shows the plots affected by the damage and the bottom right (c) shows the entire grass stand on the same date (May 18^th, 2016).

77

Figure 18: Final aerial image (top) from the spring vegetation installment at Lake Wheeler Field Laboratory (July 28^th, 2016), after 3.5 months of grass growth. The image on the bottom was processed using ArcMap, and shows the vegetation cover in green, bare soil in brown, and plot markers in red image.

78

Figure 19: Grass cover results from September 28^th, 2016 at the Lake Wheeler Field Laboratory. Differences (p&lt;0.05) in grass cover (%) are indicated if values do not have a common letter. Letter “A” represents tackifier treatments that had higher cover than tackifiers with a letter “B”.

Figure 20: Aerial image of the Lake Wheeler Field Laboratory vegetation study (September 28^th, 2016).

79

Figure 21: Aerial photograph of the Department of Transportation active site in Apex, NC after Hurricane Matthew (October 31^st, 2016).

80

Figure 22: One of the gullies at the Apex site resulting from Hurricane Matthew, with a rain gauge to depict the scale.

Figure 23: Tackifier treatment plots located at the Apex Active site, September 29^th, 2016.

81

Figure 24: Grass cover results from the Apex, NC active Department of Transportation site on September 29^th, 2016. For each tackifier the upper edge of the box represents the 3^rd quartile (75^th percentile), the line inside the box represents the median (50^th percentile), the diamond represents the mean and the bottom edge of the box is the 1^st quartile (25^th percentile). There were no differences (p&lt;0.05) among tackifier treatments.

82

Figure 25: Photograph of the greenhouse trays prior to seed, straw and tackifier application.

83

Figure 26: Photographs of the greenhouse trial in progress, showing one bench (block). Photograph (a) shows one block within the greenhouse, (b) shows a Tornado Tack sample, (c) shows two emulsified asphalt samples, and (d) shows the trays located at the back of the greenhouse, in the 3^rd block.

84

85

86