Ecological Agriculture Projects Logo

EAP Publications | Virtual Library | Magazine Rack | Search | What's new

Join the Ecological  Solutions Roundtable


Mulching for Weed Control

Park managers in the City of Berkeley, California wanted to know how effective mulching could be in suppressing weed growth in shrub beds and street median strips. They were also interested in knowing how mulching would compare with use of pre-emergent herbicides to control common weeds in urban settings.

To answer these questions, experiments comparing the effectiveness of two mulch materials as well as the effects of a mulch compared to a preI emergent herbicide, were initiated in November, I 1981. The testing implemented by the author and colleagues Dr. William Olkowski, Dr. William Klitz and Field Assistant Deborah Miller. Testing was carried out in a park shrub bed and in a street median strip in order to provide data based on actual field conditions.

The two mulch materials tested were Ceda-chips, a hard, wood chip approximately 2 1/2 long by 1/2" wider, and medium-grade redwood bark, a product with individual units averaging 1/2" by 3 /4" in size. In addition , the weed -suppressing ability of hardwood (Ceda-chip) mulch was compared to that of a pre-emergent herbicide, oryzalin (Surflan).

This report summarizes the results of these two weed control tests, and offers some recommendations to urban pest managers interested in using mulches as a weed control tactic.



Twenty test plots, each averaging 110 ft2 in size were established in a shrub bed located in Berkeley's Ohlone Park. The shrub bed had been renovated and planted with a variety of perennial ground-covers, shrubs and trees approximately one year prior to this experiment. Standard landscape installation procedures had been followed by the landscape contractor. At the time the mulch experiment was initiated, there was a high population of weeds growing in the shrub bed.

Weed species growing in the plots were identified and then removed by hand. The majority of weed species appearing in the plots during the course of this project were annuals although some perennial weeds were present. (See Table 1 for a list of weed species present in the plots.) When the perennial weeds were pulled, particular care was taken to thoroughly remove all underground roots.

Three mulch treatments plus a check, replicated five times, were installed in 20 plots in standard randomized block design. The treatments were as follows:

a. Ceda-chips applied at a depth of 2 inches;

b. Ceda-chips applied at a depth of 4 inches;

c. redwood bark mulch applied at a depth of 2 inches;

d. check (no treatment).

The redwood bark and wood chip mulches were installed at a depth of 2" to coincide with the standard used by many landscape architects and contractors in the design and installation of plantings in city parks. An additional treatment of 4" of wood chips was included in the test to determine if significant improvement in weed control could be achieved by doubling the depth of the mulch material. Funds were not available to include a treatment of redwood bark at 4". The check plots received no treatments other than the initial hand weeding.

The plots were monitored at designated intervals (generally monthly) during the period January to August, 1982. This period of time coincided with the spring/summer weed season. Monitoring included three major elements: 1 ) A visual estimate of relative percent cover of weeds in each plot. This monitoring method was selected as a means of developing data on aesthetic injury levels since aesthetic damage ( rather than economic damage) from weed growth is the problem of main concern in public parks. 2) Identification and recording of all weed species growing in the plots. 3) A count of individual weed plants growing in a 1.33 ft2 sample area in each plot.


At the end of the monitoring period, plots with 4" of wood chip mulch contained an average of 7.496 weed cover; plots containing a 2" layer of wood chips showed an average of 11.936 weed cover; plots containing 2" of redwood bark mulch averaged 12.4% weed cover; and plots receiving no mulch averaged 30.296 weed cover,


The data shows that all mulch treatments were significantly more effective at weed suppression than was no treatment at the 99% confidence level. In assessing the relative performance of wood chips versus redwood bark, there was no statistically significant difference among treatments during the seven month monitoring period. However, subsequent regular observations of the plots between August, 1982 and January, 1984 indicate that the plots containing 4 " of wood chips have remained virtually free of weed growth while plots containing 2" of redwood bark appear to have greater than 3096 weed cover. Plots with 2" of wood chip mulch have slightly more weed growth than the plots with 4" of mulch. The Ceda-chips mulch appears to have decomposed very little in the three years it has covered the plots while the redwood bark has decomposed markedly.

These observations suggest that the long-term weed control value of using 4" of a hard, wood mulch such as Ceda-chips may not become statistically significant until approximately 12 to 18 months after initial application of mulch. However, the difference may be visually (or aesthetically) apparent within 12 months after mulch is applied.


To maximize the weed suppressing effects of mulch, the following procedures are recommended:

1. Begin the mulching program at the onset of the Spring/Summer weed season (usually early March) or the Fall/Winter weed season (usually at the onset of fall rains in October or November ) . At these times, mechanical or physical weed removal is made easier by the moist soil conditions. In addition, most weeds are in the seedling stage and easiest to remove or kill with least toxic methods.

2. Organize the weeding process so that 2" to 4" of a hard, wood mulch such as Ceda-chips is applied within a few days of weed removal. The effectiveness of the mulch in suppressing weeds will be diminished if several weeks or months go by between weed removal and application of mulch.

3. Monitor mulched areas so that any regenerating perennial weeds that survived the initial weeding and come up through the mulch are pulled out or spot-treated before they reach a large size. The mulch will retain soil moisture and make physical removal of most perennial reeds quite easy to accomplish. Mulched areas also should be monitored so that mulch displaced by park users or animals can be restored to proper depth and location.

4. When applying mulch in beds containing established shrubs or trees, keep mulch a minimum of six inches away from stems or trunks. This is necessary to prevent the mulch from causing an increase of moisture at the base of plants which can lead to plant disease problems.

Based on the evidence from the experimental plots at Ohlone Park, the process outlined above can result in less than 8% weed growth in mulched plots in the first year. Regular (although informal) observations of plots at Ohlone Park three years after the mulch was applied indicate that weed growth has been kept to less than 18% of total plot area in the beds containing 4" of wood mulch. This level of weed growth is well below aesthetic or economic injury levels for typical urban park shrub bed conditions.

It should be noted, however, that the plots at Ohlone Park did not contain high populations of mature perennial weed species at the onset of the experiment. The long-term weed suppression achieved by this mulching program probably would not hold true in areas were weed species are primarily perennials unless great care was taken to thoroughly remove or kill perennial roots and crowns prior to application of the mulch.


In a similar experiment, the weed suppressing performance of 4" of Ceda-chips was compared with that of the pre-emergent herbicide oryzalin (Surflan) on weeds growing in street median strips during the spring/summer weed season.


Twenty-four plots, each averaging 60 ft2 in size were established on two median strips along San Pablo Avenue in Berkeley. Weed species growing in the plots were identified and found to consist primarily of established perennials.

During hand weed removal, an effort was made to completely remove roots of perennial weeds such as kikuyugrass, field bindweed and nutgrass but this did not prove to be possible in many of the plots. It was concluded that perennial weed roots remaining in the plots were so evenly distributed throughout all plots that they would not compromise the experimental design.

All plots contained one or more clumps of ornamental 'Peter Pan' Agapanthus. A visual estimate of the area covered by these ornamentals in each plot was made and recorded.

Three treatments plus a check were applied to the plots in standard randomized block design. Each treatment and check was replicated six times. The treatments were as follows:

a. Wood chip mulch applied at a depth of 4";

b. 4 "of wood chip mulch plus oryzalin herbicide;

c. oryzalin; and

d. check (no treatment).

The plots were monitored during the 4" of wood chip mulch plus oryzalin her spring/summer weed season. Monitoring included the same major elements described above.


At the end of the monitoring period, plots with 4" of wood mulch contained an average of 33.3% weed cover; plots containing 4" of wood mulch plus oryzalin showed an average of 25.296 weed cover; plots treated with oryzalin alone averaged 85,2% weed cover; and plots receiving no treatment (checks) averaged 9S.8% weed cover,

Copyright 199.

Info Request | Services | Become EAP Member | Site Map

Give us your comments about the EAP site

Ecological Agriculture Projects, McGill University (Macdonald Campus)
Ste-Anne-de-Bellevue, QC,  H9X 3V9 Canada
Telephone:          (514)-398-7771
Fax:                     (514)-398-7621


To report problems or otherwise comment on the structure of this site, send mail to the Webmaster