Reducing herbicide usage, minimizing grazing down time, and improving pasture productivity with banded application of herbicide during pasture renovation

Final Report for ONE03-006

Project Type: Partnership
Funds awarded in 2003: $9,980.00
Projected End Date: 12/31/2006
Matching Non-Federal Funds: $26,898.00
Region: Northeast
State: Pennsylvania
Project Leader:
David Hartman
Penn State University
Ronald Hoover
The Pennsylvania State University
Expand All

Project Information


There is great opportunity for increasing pasture productivity on many farms in the northeastern US. Improvements in pasture forage productivity often can result in per-animal or per-acre increases in animal production. A pasture renovation study was initiated during 2003 on six private dairy or livestock farms in central and northern Pennsylvania to investigate the potential for improving pasture productivity by establishing new or improved forage species. Plots of grass or grass/legume mixtures were no-till drilled during August 2004 into pastures where broadcasted or banded applications of non-selective herbicide were sprayed from a boom that was mounted on the front of the drill. One replication of twenty-one plots was drilled at each farm. Measures of performance included rate of establishment, stand density before and after several winters, and forage dry matter production. As a group, the ryegrasses and festuloliums established more rapidly and were more productive than orchardgrasses, fescues, meadow brome, and alfalfas. Extremely variable rainfall between the farms and inability to synchronize grazing events with farm owners prevented the collection of useful forage dry matter production during the latter part 2005 and all of 2006. All entries overwintered well; however, both of the two winters during the study period were relatively mild and winterhardiness of the forages was not tested.


A large percentage of open land in the northern areas of Pennsylvania is not well suited for row crop production due to land use limitations that include moderate to excessive slopes, shallow soils, variable drainage, and excessive stoniness. Many dairy and livestock producers in the region have adopted some form of grazing to make use of those acres. Most are well aware of the need to maintain forage quality and yields through rotational grazing. However, few regularly introduce new plant materials via pasture rotation, reseeding, or interseeding for the purpose of further improving the pasture and animal production on their farms. The prevailing attitude seems to be that plants already in the pasture are as good as anything that can be introduced.

Tillage as a part of pasture renovation on many of these sites can create soil erosion problems. Further, wholesale tillage that destroys all plants prior to renovation also results in a lengthy downtime until grazing can again occur. No-till drilling can minimize much of the soil loss problem; however, herbicide is often necessary to control some or all existing vegetation prior to successful no-till planting. Complete control of vegetation with herbicide likewise can result in long downtime until grazing commences. Of interest to the research team and farmers was the ability to enable satisfactory establishment of new seedlings through partial control of existing vegetation with band applications of herbicide. Soil conservation benefits and a reduction in the length of time that pastures are out of production might be realized when banding herbicide during pasture renovation.

This on-farm project was developed to gather useful local information that will support or disprove the idea that acceptable pasture renovation can be accomplished while minimizing the downtime and herbicide needed to successfully renovate pasture. Seven farms in the five counties of Tioga, Bradford, Susquehanna, Columbia, and Union Counties were identified as having interest in participating in this study. Local retail seed suppliers supported the effort through donations of all seed used in the study.

Strip plots were drilled and successfully established on six of these farms. A field day or evening producer meeting was held at each site to inform farmers, extension and other advisory personnel, and industry about the study and generate interest in the concept of improving pasture productivity through pasture renovation.

Project Objectives:

Based on discussions with some farmers regarding perceptions and/or needs to improve pasture productivity, several reasons why pasture renovation seldom occurs were identified. These include:
• Use of tillage on many acres of pasture in this region to facilitate renovation is environmentally unacceptable.
• The downtime (lost production) while renovating pasture might not be made up by improvements in future production.
• Banded applications of herbicide can be as effective as broadcasted application when attempting to establish new plant materials in a no-till renovation program.
• The species and varieties being promoted in the mid-Atlantic are not suited (winterhardy) for northern PA conditions.
• Some have tried new forages (especially ryegrasses) in the past with limited success.
• The expenses associated with renovation are perceived as too high by many graziers.


Materials and methods:

Seven farmers in five central and northern Pennsylvania counties offered to host this study. Soil testing of pastures during spring 2003 determined that many pastures needed lime. Some pH levels were as low as 5.5. Lime was spread during 2003 and early 2004. Phosphorus and potassium levels were adequate for forage establishment. The project was divided into two smaller studies at each farm. The first consisted of an evaluation of banding herbicide (glyphosate with 4.5 pounds per gallon acid equivalent, applied at 44 oz per treated acre) with seven monocultures or binary mixtures. The other was a variety trial that contained fifteen grasses or grass/legume mixtures that were drilled after herbicide was broadcasted.

A ten foot wide no-till drill with openers on 7.5 inch spacing was fitted with a 30-gallon spray tank, an electrically operate pump, and a spray boom mounted ahead of the no-till coulters. The boom was configured with one nozzle centered on each double-disk opener. Boom height was maintained so that 3.5 to 4 inch wide bands were sprayed when equipped with 40-degree even spray flat fan nozzles and an evenly broadcast pattern resulted when 110-degree flat fan nozzles were used. The drill was calibrated for the numerous entries to be interseeded. The entries, seeding rates, and method of herbicide application are outlined in table 1.

Table 1. Seeding rates and method of herbicide application for pasture renovation plots that were no-till drilled during August 2004.
Species Variety Seeding Rate Herbicide
Part A. Ability to successfully establish behind banded applications of glyphosate*
—- lbs/acre —-
control plot nothing planted — —
Alfalfa + Orchardgrass Baralfa 42 IQ + Baridana 11.3 + 6.0 broadcasted
Alfalfa + Orchardgrass Baralfa 42 IQ + Baridana 11.3 + 6.0 banded
Alfalfa Baralfa 42 IQ 15.0 banded
Orchardgrass Baridana 16.0 banded
Birdsfoot trefoil Pardee 8.5 banded
Perennial ryegrass BG-34 20.0 banded
Festulolium Duo 13.7 banded

Part B. Variety comparison (all plots received broadcasted application of glyphosate*)
Alfalfa Ameristand 403 T 15.0 broadcasted
Kura clover Endura 8.5 broadcasted
Orchardgrass Tekapo 16.0 broadcasted
Extend 16.0 broadcasted
Perennial ryegrass Grand Daddy 30 broadcasted
Mara 30 broadcasted
Tivoli 30 broadcasted
Festulolium Perun 23 broadcasted
Spring Green 23 broadcasted
Tall fescue Bronson 20 broadcasted
Barolex 20 broadcasted
Meadow fescue Laura 40 broadcasted
Meadow brome Paddock 30 broadcasted
* glyphosate (4.5 pounds acid equivalent per gallon) applied at 44 oz/treated acre
Each plot was seeded with a single pass of the 10-foot drill. Only one replicate of each treatment was planted at each farm.

Nitrogen fertilizer was applied periodically during the 2005 and 2006 growing season. Three separate applications of fifty pounds nitrogen per acre were made during a relatively dry 2005 growing season while four applications at the same rate were made during the wetter and more favorable cool-season grass growing season of 2006. Fertilizer was distributed via a shoulder-mounted spinner spreader.

Visual observations were taken each fall and spring and plots were scored on a 1 to 10 scale for stand density with 10 being best. After the first or second grazing event, depending on the farm, a walk-behind power mower was used to clip refused forage on part of the plots to 2.5 inch high stubble. Prior to the second or third grazing, 0.5 square meter samples of forage were cut to a 2.5 inch height, collected, dried, and weighed. These estimates of dry matter were collected from the grass monoculture seedings planted behind broadcast applications of glyphosate. These data were subjected to analysis of variance and differences between means were determined by Duncan New Multiple Range test when appropriate.

Research results and discussion:

The moist conditions at the time of seeding and during the subsequent two months were ideal for germination and establishment of forages. While most plots at six farms looked good going into the winter, the plots at one farm did not establish well. It appeared that the seeding depth was too shallow for the shalely hilltop soils at that site. It was decided during early 2005 to abandon the testing at that location.

Initial observations of the banded herbicide plots were very encouraging. While sprayed strips of existing vegetation were destroyed by the herbicide, the environmental conditions that were favorable for seedling establishment in September and October were also very favorable for recovery and spreading of the plants that were not sprayed. Large areas of many banded plots were had recovered remarkably well by spring 2005. Since the contribution to dry matter from existing plants was comparable or greater than that from the new seedlings in many plots, the banded herbicide plots were not sampled for dry matter production during 2005 or 2006.

Visual observations of the plots during November 2004 revealed that the perennial ryegrasses (PRG) and festuloliums (FEST) established most rapidly, as measured by plant density in the row and fall growth vigor. The tall fescues (TF) were somewhat slower and these were followed closely by the orchardgrasses (OG). There was some alfalfa, but not nearly as much as was expected by that time. There was very little birdsfoot trefoil (BFT), meadow fescue, and meadow brome. The kura clover never established at any location.

The 2004/05 and 2005/06 winters were average to mild for the region. In 2005, dry early spring conditions resulted in soil moistures less than normal and the lack of rainfall through April and May resulted in slightly-to-greatly reduced dry matter production. Dry matter measurements from pastures prior to the second or third grazing (depending on host farmer’s initial spring turnout and grazing intensity) were very much reduced due to the lack of moisture (table 2). Although yields for this one regrowth cycle varied greatly between the grass entries, the differences were not significant.

Table 2. Spring 2005 dry matter yields from grasses seeded where glyphosate was broadcasted.
Species Variety Dry matter yield*
—- pounds per acre —-
Orchardgrass Tekapo 655
Extend 778
Perennial ryegrass Grand Daddy 929
Mara 1148
Tivoli 1134
Festulolium Perun 985
Spring Green 1045
Tall fescue Bronson 924
Barolex 764
Meadow fescue Laura 794
Meadow brome Paddock 945
* differences between entries were not different as determined by analysis of variance (p<.05).

Additional dry matter yield data were not collected during the summer and fall of 2005 due to the persistent dry weather. Overall, rainfall during 2006 was greater, however the distribution between participating farms was very uneven. Grazing cycles were likewise very asynchronous between farms, preventing the acquisition of useful dry matter production data.

The plot conditions scores determined by visual observation during November 2005 were very different between the entries (table 3). The plots of PRG and FEST entries, as a group, were in better condition after the first year than those of OG and TF. The meadow fescue and meadow brome at most locations and for the overall average had the lowest score of the eleven grass entries in part B. The one exception was inconsistent stands of Perun FEST across the six locations.

Table 3. Visual estimates of relative plot condition after the first year of grazing (November 2005.)
Species Variety Condition score*
(0 to 10 scale, 10 is best)
Orchardgrass Tekapo 6.6 c
Extend 7.2 a b c
Perennial ryegrass Grand Daddy 8.3 a
Mara 8.2 a b
Tivoli 8.0 a b
Festulolium Perun 6.1 c d
Spring Green 6.9 b c
Tall fescue Bronson 6.4 c
Barolex 6.4 c
Meadow fescue Laura 4.8 d
Meadow brome Paddock 6.0 c d
* differences between means followed by the same letter are not significantly different as determined by Duncan’s new multiple range test (p<.05).

The plot condition scores during the early spring green-up of 2006 are listed in Table 4. The relative rankings did not change very much. There were a few exceptions. Tekapo OG and Tivoli PRG did not look as good in the spring as they did during the previous fall. Paddock meadow brome also had a slightly lower score in the spring compared to it previous fall rating.

Table 4. Visual estimates of relative plot condition during the spring of the second year of the study (April 2006.)
Species Variety Condition score*
(0 to 10 scale, 10 is best)
Orchardgrass Tekapo 5.2 c d
Extend 7.1 a b
Perennial ryegrass Grand Daddy 8.0 a
Mara 7.8 a
Tivoli 7.1 a b
Festulolium Perun 6.2 b c
Spring Green 6.9 a b
Tall fescue Bronson 6.9 a b
Barolex 6.4 b
Meadow fescue Laura 4.8 d
Meadow brome Paddock 5.3 c d
* differences between means followed by the same letter are not significantly different as determined by Duncan’s new multiple range test (p<.05).

Plot condition scores during November 2006 were similar or slightly lower when compared to the spring scores for nearly all entries. However, Tekapo OG and Barolex TF were found to have slightly improved scores from the beginning to the end of the 2006 growing season.

Table 5. Visual estimates of relative plot condition during the fall of the second year of the study (November 2006.)
Species Variety Condition score*
(0 to 10 scale, 10 is best)
Orchardgrass Tekapo 5.4 d e
Extend 6.6 a b c
Perennial ryegrass Grand Daddy 7.2 a b
Mara 7.4 a
Tivoli 6.9 a b c
Festulolium Perun 6.3 b c d
Spring Green 6.6 a b c
Tall fescue Bronson 6.0 c d
Barolex 6.8 a b c
Meadow fescue Laura 4.6 e
Meadow brome Paddock 4.8 e
* differences between means followed by the same letter are not significantly different as determined by Duncan’s new multiple range test (p<.05).

Research conclusions:

The anticipated benefits from band spraying herbicide during renovation instead of broadcasting it were threefold. Leaving some vegetation between newly drilled rows should reduce soil erosion during establishment. A second anticipated advantage was a reduction in the downtime between planting and the restart of grazing. The third was that less herbicide would be needed to accomplish the task. Neither of the first two ideas resulted in benefits to the system. Leaving half the existing vegetation alive did not improve soil conserving properties of the sites; a dead sod during the first year after being killed serves as a very effective mulch and effectively holds soil in place. While there was more vegetation on plots that were strip killed the previous fall, both strip and broadcast killed plots were ready to graze at about the same time in the spring. Surprisingly, some of the older, well established bluegrasses, clovers, orchardgrasses, and fescues did not die. They were able to recover from the broadcast applications of glyphosate. The recovery was extensive enough that by the second season, the contribution from the volunteer and non-controlled plants was as great as or greater than that from drilled species. The savings in cost of the herbicide was probably not worth the difficulty involved making a band application system to fit a drill.

The project raised awareness for the potential to improve pasture productivity. Some of the ideas tested in this project may have little value to some graziers, while others will be more readily adopted. However, the exercise of working with farmers on their farms, organizing field days and presenting concepts related to increasing pasture forage production, and simply raising awareness that one might consider some form of renovation to improve one’s pasture situation should encourage farmers to think more about their pastures. Stimulating some to try some new ideas will have made this project and the effort to accomplish it worthwhile.

Participation Summary

Education & Outreach Activities and Participation Summary

Participation Summary

Education/outreach description:

The outreach associated with this project has been extensive. A field day, evening meeting, or farm tour had taken place at each of the six successful test farms. Collectively, the project team hosted 132 participants at seven events. At three more formal winter meetings during 2006 and early 2007, approximately 210 participants attended and listened to presentations about the project. The project was highlighted in two articles published in Lancaster Farming, the most widely circulated farm press publication in Pennsylvania and the Mid-Atlantic States. The circulation for this weekly publication is approaching 60,000. Also, the results from the study will be made available on the Penn State University’s On-Farm Research website at

Any opinions, findings, conclusions, or recommendations expressed in this publication are those of the author(s) and do not necessarily reflect the view of the U.S. Department of Agriculture or SARE.