Four grazing paddocks from each of the two partnership farms were monitored throughout the season in order to assess pasture growth rates, pre- and post-pasture mass, net harvested mass, percent utilization, percent grass and legume, and pasture quality. Although only twenty miles apart, the two farms vary in soil type, pasture composition and mass, stocking rate and utilization.
Although both farms use managed intensive grazing as a method of pasturing their dairy animals (usually moving their milk cows to new paddocks between each milking), Shelburne Farm uses the least amount of pasture (an average of 39%) whereas, Moultrup Farm uses an average of 50% of their pasture each grazing. The main reasons for this difference are stocking rate and soil type. Shelburne Farms has a larger land base to work with and, in order to optimize milk production, applies less grazing pressure. The other factor that affected utilization was the unusually high rainfall during the summer of 2006. Many of the soils at Shelburne Farm are moderate to poorly drained and the farm decided to lightly graze in order to avoid serious pugging of the fields. During the heaviest rainfall period in late May/early June, the farm actually pulled cows off pasture for a couple weeks.
Moultrup farm has a more limited land base and well drained soils; therefore, they need to fully utilize their pasture without sacrificing milk production. If anything, the excessive rainfall was helpful on these soils producing pasture growth rates exceeding 90 lbs. of dry matter per acre per day.
Legume content varied across paddocks but most varied with season. Generally, legume content increased as the season progressed with highest levels in July and August. The recently overseeded paddocks did not show any higher levels of legume content as compared to the non-overseeded ones.
For a more complete report of this project including pictures, figures and extensive tables, to the Pasture and Grazing section of the Vermont Crops and Soils Homepage at
Feed planning on pasture is an important management strategy for improving profitability for pasture-based dairy farms. Prescribed grazing plans as required by the USDA Natural Resources and Conservation Service for their programs, rely on feed budgeting information (Kevin Kaija, personal communication). Yet, little on-farm information is available concerning pasture productivity and forage quality in Vermont and the New England states that can be used for planning purposes.
Variables that influence pasture productivity and quality include soil productivity, weather conditions during the growing season (particularly rainfall), grazing heights and species content. Legume content is particularly important. Not only do legumes fix atmospheric nitrogen reducing the need and expense of N fertilizers, legumes also improve the quality of pasture by increasing crude protein, energy and some minerals in the feed. Because of a higher growth rate in the summer as compared to cool season grasses, legumes also improve the seasonal distribution of pasture.
Farmers need practical ways to collect on-farm data that can be used for feed planning purposes. Measuring dry matter yields either by quadrat sampling or some indirect approach such as the rising plate meter can be very labor intensive and is usually confined to controlled research studies. Farmers generally don’t have the time or resources to do such measurements on-farm. However, with good daily record keeping and the right type of pasture design, farmers can use their own animals to evaluate the productivity of their pastures.
- To collect quantitative information on pasture production and quality for a variety of soil types and species mixtures.
To compare a method of on-farm record-based assessments as a way to estimated pasture yield and availability as compared to more intensive hand sampling.
To demonstrate how farmers can collect their own data using daily record keeping and turn that data into feed planning information.
To evaluate and demonstrate the impact of legume content on forage availability and quality through the grazing season by comparing frost-seeded clovers to the existing pasture.
Four grazing paddocks from two Vermont farms were monitored throughout the 2006 season in order to assess pasture growth rates, pre and post pasture mass, net mass, percent utilization, botanical composition, and pasture quality. The two farms participating in this project included Shelburne Farms of Shelburne, VT and Moultrup Farm of Richmond, VT. Both farms utilize a management intensive grazing system such that lactating cows are moved to new pasture between every milking. Shelburne Farms milks Brown Swiss cows and is located near Lake Champlain. Their pastures are on a variety of soils ranging from well-drained silt loams to moderate and poorly drained clays. The Moultrup farm milks Jersey cows and is located about 20 miles east of Shelburne and is adjacent to the Huntington River. Most of the pasture soils are fine and very fine sandy loam soils.
Pasture Mass and Dry Matter Yield – The day before and the day after each treatment paddock was grazed, a pre- and post-grazing measure of pasture yield was determined using the acrylic pasture plate method taking the mean and standard deviation of 25 measurements per paddock. The difference between pre and post grazing yield was calculated as net pasture yield. Percent utilization was calculated by dividing net yield by pre-grazing yield times 100.
Pasture Quality and Botanical Composition – During each pre-grazing sampling, half of the 25 sites were hand sampled using a 2×2 ft wire quadrat placed over the plate meter. Forage was removed with hand clippers to a height similar to what the cows are observed to graze. A composite from all samples was thoroughly mixed and sub-sampled for quality analysis. Samples were sent to the University of Vermont Agricultural Testing Lab and analyzed for crude protein, soluble protein, acid detergent fiber (ADF), neutral detergent fiber (NDF), digestible NDF, net energy of lactation and minerals. A second sub-sample was hand separated into grasses, legumes, and forbs. Each was dried and weighed to determine percent botanical composition.
Indirect Method for Determining Legume Content – a point count method utilizing 24 equally spaced 3/8” holes drilled into the acrylic rising plate was found to correlate well to hand separations and was also used throughout the season to estimate legume content. Counts were made on each of the 25 rising plate points per paddock measurement.
On Farm Records – in 2006 and 2007, both farms kept daily records of cow numbers, feed regime, paddocks grazed, and milk production. This data was used in both years to estimate daily pasture dry matter intake which was then used to calculate harvested production of the targeted pastures in the project.
Growth and Pasture Utilization – Both farms in this project utilized managed intensive grazing as a method for pasturing their dairy animals (usually moving their milk cows to new paddocks between each milking). Pre-grazing mass was usually above 2500 lbs. per acre, more than adequate for optimum dry matter intake. The Moultrup Farm utilized an average of 50% of their pasture each grazing. Shelburne Farm had poorer utilization partially due to the excessive rain in 2006 along with their poorly drained soils.
Pasture growth rates ranged from 38 to 98 lbs. per acre per day across farms and pastures. The high amounts of rainfall in June seemed to enhance the pasture growth at the Moultrup farm, which has a predominately course texture soil; whereas, the growth rate at Shelburne Farm was more variable across paddocks, particularly in the first half of the season when there was excessive rainfall.
Legume Content – Legume content varied across paddocks but mostly varied with season. Generally, legume content increased as the season progressed with highest levels in August and September running between 30 and 45 percent on a dry matter basis.
Generally, legume content was only associated with higher pasture quality, as expressed by crude protein content, acid detergent fiber and neutral detergent fiber, when it exceeded 30 percent of the botanical mixture; however, the relationship varied greatly and was also influenced by time of year. Soluble protein was not affected by legume content. Calcium was the only mineral that significantly correlated with legume content having higher levels with more legume as expected. Phosphorus and potassium can be taken up as similar rates by legumes or grasses, so we did not expect to see major differences. Control research studies have shown that magnesium levels can be improved with higher legume pasture; however, we did not find this in this project perhaps due to the high variation in legume content and the generally low levels we measured.
Frost seeding on the partner farms did not appear to make a significant difference in legume content in the first year after overseeding except for one set of paddocks at the Moultrup Farm. Middle A, which had been overseeded in the spring did have a consistently higher legume content than its adjacent paddock, Middle B which had not been overseeded. These paddocks were intensively grazed starting in May and the grasses were never allowed to grow very tall. On the other hand, Jim South, which had been overseeded, and its adjacent paddock, Jim North were not different in legume content. This may be explained by the fact that these paddocks were harvested for hay in the first cutting, and the legumes never got a chance to grow and develop. The wet fields at Shelburne Farms were difficult to graze in the first half of the season resulting in a higher amount of grass cover and low legume content. In addition, the four fields at Shelburne Farms varied quite a bit more in soil type, drainage, size and distance as compared to the Moultrup pastures; therefore, it was not a conducive situation to make a good comparison.
On-Farm Records – Using on-farm records to estimate pasture yield at the Moultrup farm showed a very similar result to intensive hand sampling utilizing an acrylic pasture plate method taking the mean and standard deviation of 25 measurements per paddock. There was far less consistency or similarity between methods at Shelburne Farm. This could have been caused by errors in calibrating the rising plate on pasture that was very lush and wet due to excessive rain. Also, the farm only grazed these pastures during the day and not at night, so the time to get through the pasture ranged from one half to three days. This could also have lead toward errors in estimating the exact time the animals were on the pasture grazing.
As a result of this project, we were able to demonstrate practical methods for assessing pasture yields and legume content at two pasture walks in 2007. We conducted an in-service training in 2007 for agricultural professionals (extension, NRCS, Certified Crop Advisors) to demonstrate the use of these tools for conducting pasture condition scoring.
The data we collected from these farms was also used by Dr. Woody Lane, a nationally recognized sheep nutritionist, when he conducted nutritional workshops in Vermont and surrounding states in August of 2007.
Education & Outreach Activities and Participation Summary
- The results of this project were presented at the 2007 Vermont Grazing Conference to a group of 35 advanced grazers. The data from the project created a lot of discussion about the impact of grazing management and stocking rate on forage production and quality.
A poster of the project was presented at Annual Meeting of the American Forage and Grassland Council in June 2007. As a result of that meeting, Dr. Woody Lane, a nationally recognized sheep nutritionist, was interested in the data from the project and has used it since in his nutritional workshops.
An annual report of the project was posted on the Vermont Crops and Soils website in June of 2007.
This has since been updated to a final report (http://pss.uvm.edu/vtcrops/articles/SAREPartner_Pasture_Summary_Report.pdf)
Data we collected from this project has also been used in the Forage and Pasture Management Course taught at the University of Vermont.
The two farms that partnered in this project continue to keep good grazing records. We were able to collected pasture data for 2007 using their records.
Areas needing additional study
A future outcome of this project will be to coordinate with USDA-NRCS as they collect on-farm records from pasture-based farms enrolled in the Grassland Reserve Program and use that data to develop a pasture productivity database based on soil type and species composition. This data will then be helpful in developing pasture management plans.