FACTORS INFLUENCING THE EFFECTIVENESS OF SINGLE AND DOUBLE LAYER WEED BARRIERS
IN SUPPRESSING GRASS GROWTH
Mayleen Farrington, Paul Gazda, and Kelly Slutz
Northern Arizona University
Abstract
This experiment attempted to assess factors influencing the effectiveness of weed barriers in suppressing grass
growth. A single layer woven polypropylene barrier was compared with a double layer barrier consisting of
woven polypropylene on top and polyethylene sheeting. The study also attempted to assess the effects on grass
growth of amount of water and drainage of excess water. Significant penetration of woven polypropylene barrier
by grass roots was observed in all test conditions. Grass growth increased as water amount increased, suggesting
an asymptotic relationship. Unexpected shortcomings in the experimental design prevented conclusions regarding
the effects of weed barrier type and drainage.
Introduction
This experiment was conducted in the Northern Arizona University (NAU) greenhouse from November 2012
to March 2013 to study factors affecting the efficacy of weed barriers in suppressing grass growth. Two weed
barrier configurations were studied: one a single layer of woven polypropylene, the other a double layer system
with woven polypropylene on top and polyethylene sheeting on the bottom. The effect of water amount and
adequate drainage of excess water were also studied.
The double layer weed barrier system was developed by one of us (Gazda) and had performed very effectively
on his property for over 15 years. The double layer system includes provisions to allow water penetration around
plants and trees, and prevent pooling of water at low points in the landscape. NAU adopted a modified version of
the double layer weed barrier system under rock mulch areas on the NAU campus in 2009 that did not include
provisions for water drainage, and used woven polypropylene instead of non-woven geotextile as the top layer. In
2010, the city of Sedona, Arizona installed the double barrier system under rock mulch in some of its highway
medians that successfully prevented weed growth.
In the summer of 2013, weeds were observed growing above the double layer weed barrier in some rock
mulch areas of the NAU campus after a month of unusually heavy summer rains. Since the double layer system
(using non-woven geotextile fabric instead of woven polypropylene as the top layer) had been completely
preventing weed growth on Gazda’s property and in the highway medians of Sedona, we decided to investigate
several parameters in an attempt to better understand why NAU’s double layer system had allowed significant
weed growth in some rock mulch areas on the NAU campus.
We suspected that lack of drainage in areas where the underlying soil formed bowls that retained rainwater
may have been giving weeds enough moisture to grow above the polyethylene layer. Our goal in this experiment
was to examine double weed barrier shapes that would either prevent water from accumulating on top of the
barrier by allowing it to drain around the edges, or allow water to accumulate by preventing it from draining
around the edges.
A secondary goal of this study was to compare the effectiveness in suppressing weed growth of a single layer
of woven weed barrier compared to the double layer system used at NAU.
Methodology
The experimental design examined the effect of three variables on grass growth: type of weed barrier (single
or double layer), shape of weed barrier (allowing or preventing water drainage over edges of weed barrier), and
amount of water applied to the grass.
Experimental trays contained a layer of soil approximately 1 to 1 ½ inches deep on the bottom, then a single or
double layer of weed barrier on top of the soil, then a thin layer of soil mixed with grass seed, then a 1 inch layer
of rock mulch mixed with a small amount of soil (described in more detail below).
Type: Two types of weed barrier systems were studied: a single layer weed barrier (Single) of DeWitt Weed
Barrier 20 year (1), 4.1 oz., woven polypropylene barrier; and a double layer barrier (Double) consisting of one
layer of impermeable clear 6 mil polyethylene sheeting under the same DeWitt weed barrier used for the Single
condition. The DeWitt weed barrier was the same as NAU used on its landscape, and is similar in design and
composition to other woven polypropylene weed barriers.
Shape: Barriers were installed in the experimental trays to either lie flat on the soil surface (Flat) to allow
water to drain over the edge of the barrier, or with the edges extending above the soil and resting on the sides of
the experimental trays (Cupped) to prevent water from draining over the edge.
The Flat condition allowed water to drain both through and around the edges of the Single layer barrier into
the soil below. When combined with the Double barrier condition, water could only drain over the edges of the
Flat barrier because the polyethylene sheeting below the woven weed barrier prevented water from penetrating the
barrier. In the Cupped condition, water could not drain into the soil below the
barrier because both barrier layers, including the polypropylene sheeting,
curved vertically against the edges of the experimental trays above the top of
the rock mulch (Figure 1).
Water Amount: Three water amounts were calculated based on the
average rainfall in Flagstaff, Arizona during the month of July from 20112013 as reported by the Weather Underground website
(www.wunderground.com). The experimental trays were filled to the average
July rainfall depth, then poured into beakers to determine the number of fluid
ounces. Average rainfall was determined to be 37.38 oz. per week which was
then divided into three applications per week. Three water amounts were
used in the experiment: half of average amount (0.5X) calculated to be 6.25
oz. per application; average amount (1.0X) calculated to be 12.5 oz. per
application; or twice the average amount (2.0X) calculated to be 25 oz. A
Figure 1: Tray with cupped double
barrier. Edges of barrier materials
can be seen above rock mulch. Tray
number is in the upper left corner.
watering can with sprinkler head was used to apply measured water.
Control: Control treatments did not have weed barrier or rock mulch, but were otherwise prepared as described
below.
The treatment combinations were set up and the grasses grown in 30 translucent plastic trays with inside
dimensions measuring approximately 14 H x 9 W x 5.5 D inches. The trays were drilled with small drainage
holes. Each tray received 10 loosely packed, level measuring cups of sterile soil mixed with vermiculite (the
standard soil available at the NAU greenhouse). For cupped trays, barrier material was cut into 12.5” x 16.5”
rectangles and placed on top of soil with the edges going up along the sides of the trays (Figure 1). Flat trays had
8” x 12.75” material laid flat on top of the soil. Once barriers were in place, each tray (including controls) was
sprinkled with 92g of soil mixed with 1.5g of grass seed (Warner’s Native Arizona Turf Mix, a seed mix used on
the NAU campus). Barrier trays then had a one inch layer of rock mulch material placed evenly across the soil to
simulate the rock mulch that typically covers weed barrier on the NAU campus. Finally, each tray (including
controls) received another 92g of soil applied with a flour sifter to replicate wind-blown material.
A wooden craft stick was placed in each tray with marks to designate water amount to be applied. Trays were
divided into two replicates for each barrier system and water level (Table 1).
Tray
number
1, 2
Barrier Type
Double
Barrier
Shape
Cupped
Water
Amount
2.0X
Tray
number
17, 18
Barrier
Type
Double
Barrier
Shape
Flat
Water
Amount
1.0X
3, 4
Double
Cupped
0.5X
19, 20
Single
Flat
2.0X
5, 6
Double
Cupped
1.0X
21, 22
Single
Flat
0.5X
7, 8
Single
Cupped
2.0X
23, 24
Single
Flat
1.0X
9, 10
Single
Cupped
0.5X
25, 26
None
N/A
2.0X
11, 12
Single
Cupped
1.0X
27, 28
None
N/A
0.5X
13, 14
Double
Flat
2.0X
29, 30
None
N/A
1.0X
15, 16
Double
Flat
0.5X
Table 1: Assignment of trays to experimental conditions.
Trays were placed on a greenhouse bench according to a random number
generator (Figure 2), and were watered three times per week (every 2 – 3
days) from November 1, 2013 to March 17, 2014. The greenhouse
maintained temperatures between 65F – 75F (+/- 2). Humidity was not set
but was similar to the outdoors at approximately 40% - 50%. Humidity
spiked during watering. Photos were taken intermittently during the growing
process.
When most of the grass had died, the barriers were carefully lifted out of
the trays to photograph root penetration of the barriers. In all but the Double
Cupped treatment, most of the soil below the barrier lifted out of the tray
with the barrier due to extensive root growth through and/or around the edges
of the barrier. After photographing, the grass was harvested by tearing it at
Figure 2: Randomly placed trays in
greenhouse.
soil level to separate root growth from shoot growth. The grass was placed in labeled paper bags, weighed, and
oven dried for 7 days, periodically being reweighed to assess moisture loss. Final weights were used to compare
above-ground biomass between replicates.
Results
Grass roots penetrated the woven polypropylene weed barriers in all replicates containing barriers. In the
Double barrier treatments with the Cupped shape, where the roots could only grow between the woven barrier and
the polyethylene layer and could not grow around the edges because of the cupped shape of the barrier, abundant
root growth through the woven barrier was observed at the 1.0X and 2.0X water amounts (Figure 3), with
moderate root growth at
the 0.5X water amount. In
the Double barrier
treatments with the Flat
shape, root growth similar
to the Double Cupped
treatment was observed
between the woven barrier
and the polyethylene layer.
However, extensive root
growth into soil below the
weed barrier was also
observed due to roots
Figure 3: Root growth through woven
weed barrier in Double Cupped
treatment.
Figure 4: Root growth around edges of
double weed barrier (into soil below) in
Double Flat treatment.
growing around the edge of
the weed barrier (Figure 4).
In the Single barrier
treatments with Cupped
shape, abundant root growth
was observed in the soil
below the weed barrier at the
1.0X and 2.0X water
amounts (Figure 5), with
moderate root growth at the
0.5X water amount. Because
of the cupped shape, the roots
could only have grown into
Figure 5: Root growth through
woven weed barrier in Single
Cupped treatment.
Figure 6: Root growth through and
around edges of woven barrier in Single
Flat treatment.
the soil by penetrating the woven barrier. In the Single barrier treatments with Flat shape, abundant root growth
was observed in the soil at the 1.0X and 2.0X water amounts (Figure 6), with moderate root growth at the 0.5X
water amount. Because there was not a polyethylene layer in these treatments to cleanly separate the soil from the
woven barrier, it was not possible to directly observe how much of the root growth had penetrated the barrier
versus grown around the edges of the barrier.
Looking at the grass weight data, the most obvious effect was that of water amount on grass weight. Figure 7
shows pictorially the effects of water amount on grass weight. As water amount increased, grass growth
increased. At half the normal amount of water (0.5X) the grass was very sparse in all treatment conditions. The
Figure 7: Typical weed growth at 0.5X (left), 1.0X (center), 2.0X (right) water amounts. These examples are from
the Single layer, Flat shape treatment, but are typical of the trends observed as water amount varied in all
treatments.
control treatments, which had no weed barrier or rock mulch had no grass growth at the 0.5X water level. It was
not clear whether the zero growth in the control treatments at the 0.5X level was due to absence of weed barrier or
absence of rock mulch, or both. Unfortunately, we realized too late that we did not have adequate control
conditions to draw conclusions about the effect of the weed barrier and rock mulch on grass growth. Additional
treatments having weed barrier and no rock mulch, and rock mulch and no weed barrier should have been
included in the study. We concluded that the single control condition of no weed barrier and no rock mulch would
not contribute to an understanding of the data. Therefore, we excluded the control treatments from the graphs and
statistical analyses below.
Figure 8: Grass weight by water amount.
The normal (1.0X) and twice normal (2.0X) water amounts showed significantly higher grass weights than the
0.5X level (Figure 8). An analysis of variance showed the effect of water amount to be significant at the p = .0002
level. The least squares means differences Tukey HSD showed that the 0.5X grass weighed significantly less than
the 1.0X and 2.0X treatments, but that the 1.0X and 2.0X treatments did not differ significantly from each other.
Barrier type and barrier shape did not have a significant effect on grass weight by themselves, but both showed a
significant interaction with water amount. Barrier shape interacted significantly with water amount with p < .0001
(Figure 9). Effect tests showed the difference between Cupped and Flat shapes as water amount varied was
Figure 9: Grass weight by barrier type and water amount.
significant at the p=.0119 level. Figure 9 shows little difference between the shapes at the 0.5X water amount, but
at the 1.0X water amount, the Flat shape yielded an average of 5.37 grams of grass compared with 1.98 grams for
the Cupped shape. At the 2.0X water level, the Flat shape yielded an average of 4.92 grams of grass compared
with 3.64 grams for the Cupped shape.
Barrier type interacted significantly with water amount with p = .0004 (Figure 10). However, effect tests
showed the difference between Single and Double barriers as water amount varied failed to reach the level of
significance (p=.197). Figure 10 shows little difference between the barrier types at the 0.5X and 1.0X water
Figure 10: Grass weight by barrier type and water amount.
amounts, but at the 2.0X water amount, the Single barrier yielded an average of 5.06 grams of grass compared
with 3.50 grams for the Double barrier.
Discussion
Although this study revealed shortcomings in the experimental design over the course of the experiment that
prevented an assessment of all of the effects we had hoped to measure, there were several significant findings.
Grass roots penetrate woven weed barriers
Grass roots penetrated the woven weed barriers in all replicates containing barriers. Root penetration through
woven barrier was clearly evident in the Cupped Double barrier treatment where roots grew between the woven
barrier and polyethylene sheeting where no soil obscured the view of the roots. In the Cupped Single barrier
treatments, the abundant root growth observed in the soil below the weed barrier could only have occurred by
penetrating the barrier, since the cupped shape did not allow roots to grow over the edge of the barrier.
Woven polypropylene weed barriers, such as the DeWitt woven barrier employed in this study, are widely
used in landscaping. One of us (Gazda) has had several discussions with professional landscapers about weed
barrier, and they have expressed a common view that woven weed barriers are initially effective in preventing
weed growth, but over time, silt accumulates on top of the woven barriers and weeds grow in the silt on top of the
barrier. None of the landscapers seemed to hold the view that the weed roots actually penetrate the weed barrier
into the soil below.
Given the clear evidence of abundant root penetration through woven weed barrier in this study, it seems more
likely that woven weed barriers fail when enough silt accumulates on top of the barrier to allow seed germination.
Once germination occurs, weed roots penetrate the woven barrier into the soil below where sufficient moisture
and nutrients are available to sustain their growth. The barrier would also tend to preserve moisture in the soil
below by inhibiting evaporation.
Because woven weed barriers have small spaces between the weaves, they are more likely to be effective in
preventing the growth of plants with taproot systems, such as dandelion, rather than plants with fibrous root
systems such as grasses and plantain.
Effect of Water Amount on Grass Growth
The water amounts chosen for this study were meant to simulate half of normal, normal, and twice normal
rainfall for Flagstaff, Arizona during the month of July. Data analysis showed a large and highly statistically
significant increase in grass growth (as measured by weight) as water varied from half of normal to normal, and a
much smaller, statistically insignificant, difference as water varied from normal to twice normal. This suggests
that the amount of water that the grasses in this experiment could utilize in their growth process has an upper
limit. The graph in this study suggests that further increases in water amount may have a diminishing effect. Since
measuring the effect of water on grass growth was not a primary goal of this study, we did not conduct a literature
review to determine if such studies had previously been reported.
Barrier Shape and Water Amount
The strongest observed effect related to the weed barriers was that the Cupped shape, which prevented grass
roots from growing over the edges of the barrier into the soil below, was more effective in suppressing grass
growth than the Flat shape. This was likely due to the ability of grass roots to grow over the edges of the barrier
into the soil below with the Flat shaped barrier. This view was reinforced by observations of the soil under the
weed barriers when the grass was harvested. In the 0.5X water treatment, little moisture was observed beneath the
flat barriers and relatively little grass survived, so root growth over the edges would not have contributed much
additional moisture for the grass. In the 1.0X and 2.0X water treatments, substantial moisture was observed in the
soil beneath the barriers and extensive root growth was observed in the soil near the edges of the containers. In
those treatments, significant additional moisture was available to roots that grew over the edges of the barrier. The
significantly higher grass weight in the Flat Single barrier condition is consistent with the fact that with the single
barrier, roots could reach the moist soil below by growing both through and over the edges of the barrier, but with
the Flat Double barrier, the roots could only grow over the edges, so fewer roots could reach the moist soil below
the barrier.
Barrier Type and Water Amount
Although barrier type (Single versus Double layer) interacted significantly with water amount, the difference
between Single and Double layers failed to reach significance. The data do suggest, however, that the Double
barrier may limit growth sooner than Single barrier. There was essentially no increase in grass weight in Double
barrier treatments as water amount increased from 1.0X to 2.0X, whereas grass weight continued to increase in
Single barrier treatments as water amount increased from 1.0X to 2.0X.
Areas for Further Study
The design of this experiment was intended to primarily evaluate the role of adequate drainage in the
performance of the double weed barrier used at NAU by comparing a Flat barrier in which water could drain over
the edges with a Cup shaped barrier in which the barrier edges extended above the top of the soil and rock mulch
to prevent drainage. Although extensive grass root growth over the edges in the Flat barrier treatment confounded
results and prevented conclusions regarding drainage, a number of opportunities for further research were
identified based on observations made in this study.
Drainage: Attempting to assess the role of adequate drainage in a greenhouse setting if fraught with
challenges. Even though drainage holes were drilled in the small trays used for the experimental conditions, it did
not adequately simulate the absorption of excess water into deep soil as would happen in a natural outdoor setting.
To assess the role of drainage in the performance of the double layer weed barrier would likely require several
outdoor plots, some with no drainage allowed through the polypropylene layer, and others with specific areas, as
prescribed in the double barrier specification, where the polyethylene layer is omitted to allow adequate drainage.
The role of rock mulch: The control treatments in this study, which had no weed barrier or rock mulch showed
no grass growth at the 0.5X water level, whereas all treatments with rock mulch cover showed some growth at the
0.5X water level. This suggests that rock mulch, which is often used as a ground cover, may encourage weed
growth, perhaps by suppressing moisture evaporation from the soil. A study to assess the relationship between
rock mulch and weed growth could provide valuable insights.
Woven versus non-woven top layer in double barrier: The primary role of the top layer of the double layer
weed barrier system is to protect the lower polyethylene layer from UV damage and puncture. Whether the woven
weed barrier used by NAU decreases the effectiveness of the double layer system by trapping more moisture
between the layers than the non-woven geotextile is an important question to answer. It seems likely that the
woven barrier provides fewer spaces for water to evaporate than the non-woven geotextile. But whether that is the
case awaits a future experiment.
References
1. DeWitt fabric weed barrier http://www.dewittcompany.com/products.html#