The Teagasc Farming for a Better Future event held in conjunction with the Irish Farmers Journal took place on Tuesday 30 August at Johnstown Castle. On the day, the Irish Farmers Journal held two practical demonstrations, one of which showcased results from a trial carried out where three different methods of over-sowing clover in an existing grass sward were compared.


The three methods included the farm’s standard Rauch Axis twin-disc fertiliser spreader (lowest cost), a 3m Einböck grass harrow with a mounted air-seeder (medium cost) (also belonging to the farm) and a 3m Erth Agriseeder direct drill (higher cost).

No method likes thatch or thrash in the sward, while the last two have the mechanical design to help cut through it or tear at least some of it out.

The direct drill uses an angled disc to cut a narrow slot in the soil, where the seed is placed in direct contact with the soil. \ Philip Doyle

As per its design, the fertiliser spreader will broadcast the seed, but once set up correctly, the success of the method is down to preparation of the sward – it needs to be open and grazed tightly, with the aim of achieving soil-seed contact.

The tines on the grass harrow create a small amount of tilth for soil-seed contact which is what we are always looking for. The mounted air seeder then accurately distributes the seed. Giving the sward any more than one pass would run the risk of chitting unwanted dormant weed seeds.

Meanwhile, the direct drill uses an angled disc to cut a narrow slot in the soil, where the seed is placed in direct contact with the soil using the trailing coulter.

Regardless of the method, rolling is strongly encouraged to lock in soil seed contact. This is done most easily by putting livestock in the paddock, or by using a roller once the dung pats have dried. All methods are successful to a certain extent when carried out correctly.


Trial plots using each of the three methods were sown on 27 April. To keep variables to a minimum, a bare paddock with a cover of 600kgDM/ha with limited pre-existing clover was chosen. Each machine was calibrated to sow the naked seed at a rate of 5kg/ha. The variety mixture used was a 50:50 mix of AberHerald and AberSwan (medium-leaf white clover).

After each respective plot was sown, the paddock was then left four days so that dung patches would dry ahead of rolling. Following rolling, 2,600 gallons/ac of dirty water was applied to help establishment. After 18 days the paddock was grazed at a cover of 1,000kg/DM/ha to let light down to the new clover seedlings to encourage growth.

The fertiliser spreader will broadcast the seed, but once set up correctly, the success of the method is down to preparation of the sward \ Philip Doyle

The paddock then received 16kg N/ha, half its usual N application which was the case for the remainder of the summer to help the clover establish within the sward. After the first grazing, the paddock re-entered the normal grazing rotation.


To measure accurate results in terms of clover distribution and clover content within the sward, three clover counts were taken. The first was on 22 April (prior to over-sowing), the second on 2 June and the third on 13 July.

Clover distribution is a way of examining where clover is present in the sward. If there is one plant or 100 plants in one area it is marked as having clover.

Regardless of the method, clover distribution increased significantly since over-sowing. However, on the third count a 13.3% greater distribution was observed in the plots sown by the Einböck and Erth drill, which was put down to the accuracy of the metering units used on both machines.

The clover content percentages show how clover content has increased since over-sowing. The first count in April is where the smallest difference in results was observed across each of the three methods. After every count the difference between each method became larger.

On the second count, the Einböck showed just shy of three times the clover plants per m2 compared to the broadcast fertiliser spreader method and 8.69 more clover plants per m2 when compared with the direct drill.

On the third count the latter changed. Here, the direct drill method counted 590.93 plants per m2 compared to 507 and 202 from the harrow/air-seeder and broadcast fertiliser spreader methods respectively. In terms of percentages, the difference in plants per m2 between the Einböck and Erth drill after the final count was 16.5%. With correct management, clover will multiply through its nodes so if we were to measure this again, it’s likely plant counts would have further increased.

Getting good establishment

Before undertaking any over- or re-seeding job, it is important to examine the paddock or field to assess the chances of success.

The grass harrow tines create a small amount of tilth for soil seed contact followed by mounted air seeder which accurately distributes the seed. \ Philip Doyle

Clover needs a pH up around 6.5 and also high levels of P and K. The establishment cost is too big to risk failure. Soil pH should really be corrected in the previous season. Target a time slot when frequent rain is likely post sowing.

Assess the surface of the soil for:

  • The level of thrash or thatch present.
  • The tightness of the soil surface.
  • Thatch on the surface could interfere with the ability of seeds to get onto the soil to establish while very tight soil could limit the ability of young clover roots to penetrate. Check this using a fork or spade.

    Once the seeds are sown and rolled or walked in, allow time for germination. The grass should be grazed before the young clover plants can be caught by grazing animals. This will remove some of the competition for light.

    Grass growth should be slowed in the early weeks to help the clover establish. This means less or preferably no N until the clover is well established.