The Development of Feeding and Management Strategies for Milk Production Systems with Very Contrasting Milk Supply Patterns

Location and site characteristics

This study was carried out at the Solohead Research Farm, Co. Tipperary, which is part of the Teagasc Dairy Production Research Centre. The experiment commenced in September 1995 and ended in January 1998. The experimental area was totally reseeded over the previous two years.

Experimental design and treatments

Three very contrasting calving patterns were compared in this experiment. A total of 78 Holstein-Friesian cows blocked for lactation number and milk yield were allocated to three treatment groups. There were 26 animals per treatment for Year 1 and Year 2 of the experiment. The three experimental treatments are outlined in Table 1.

Table 1. Details of the three experimental treatments

The experimental animals had an average breeding index (RBI95) of 113 in year 1 and an average breeding index (RBI95) of 116 in Year 2 of the experiment. In Year 1, 8 first lactation animals were randomly assigned to each of the three treatments, while in Year 2 there were six first lactation animals per treatment.

Farmlet design and grassland management

The site was reseeded in the previous two years and was almost 100% perennial ryegrass (Lolium perenne) at the start of the experiment. The seed mixture per hectare was Condessa (16.85 kg), Magella (5.61 kg), Tyrone (5.61 kg), Susi (1.11 kg) and Aran (1.11 kg). Each treatment had it's own farmlet comprising of 19 paddocks in Year 2 and 18 paddocks in Year 2. Paddock areas ranged from 0.49 ha. To 0.73 ha. The paddocks were grouped into blocks of 3 paddocks. The treatments were randomly assigned to the paddocks within each block. A rotational paddock grazing system was employed. For grazing, each paddock was subdivided to give a daily allowance of grass to the cows. Temporary electric fences were used to achieve greater control over the grazing area. Two 'entrances/exits' to each paddock were used to avoid 'poaching' in wet weather.

Grass production

Grass production was measured (in a separate experiment using 650 kg N fertiliser per ha at the Solohead Research Farm) using the method of Corrall and Fenlon (1978). The grass production for 1996 and 1997 are shown in Figure 1. Total grass production averaged 19.8 tonnes DM per hectare in 1996 and 16.7 tonnes DM per hectare in 1997. Therefore, total grass production was very high in 1996 and was especially so in June and August. Peak growth yields of 160 kg DM/ha/day in the week of 25th June 1996 was the highest ever recorded in this type of experiment in Ireland. In 1997, grass growth was above normal in March and early April and was below normal in May.

Grazing Management

Table 2 shows the grazing pattern for the two years of the experiment. Turnout date depended on ground conditions and not on grass supply. The turnout date in 1996 is the 'normal' turnout date for the Solohead Farm. Turnout was particularly early in 1997 due to good ground conditions in February and March. Grazing ceased in both years because of poor ground conditions and not because of restricted grass supply. When the ground score rises above 4.5 the soil surface becomes susceptible to damage by poaching. All animals were removed and housed when the ground score rose above this critical level.

Table 2. The dates on which grazing was commenced and ended in 1996 and 1997.

*Treatment A

Concentrate supplementation

Table 3 gives the concentrate input to the three treatments for the two years. In Year 1, the autumn-calving cows in Treatments A and AS were fed at a level of 8 kg concentrates/cow/day during the indoor period; this was gradually reduced after turnout to pasture and all cows were off concentrates by early May. The spring-calving cows in Treatments AS and S were fed at a level of 8-kg concentrates/cow/day from calving until turnout to pasture, this was gradually reduced and all cows were off concentrates by early May. The spring-calving cows in Treatments AS and S were given 2 kg concentrate/cow/day from October 11 onwards until drying off.

In Year 2, the autumn-calving cows in Treatments AS and S were given 6-kg concentrates/cow/day during the indoor period, this level was then gradually reduced and all cows were off concentrates by early April. The spring-calving cows in Treatments AS and S were fed at a level of 6-kg concentrates/cow/day from calving until turnout to pasture, this level of feeding was then gradually reduced and all cows were off concentrates by early April. Due to poor grass growing conditions in Year 2, the cows in Treatment A were supplemented with 3 kg concentrates/cow/day, while Treatment AS and S were supplemented with 5 kg of concentrate per cow per day from May 8th to May 30th. The spring-calving cows in Treatments AS and S were offered 2-kg concentrate/cow/day from September 12 onwards until drying off.

Table 3. The quantity (kg/cow) of the concentrate supplement offered to each of the three treatments for the two years.

Milk yield and composition

The milk yields, milk composition and lactation lengths for 1996 and 1997 are shown in Table 4.

Table 4. The effect of system of production on milk yield and milk composition

SEM = standard error of mean; NS, not significant (P<0.05); *P<0.05;

There were no statistically differences between treatments in milk yield in either year of the trial but there was a non-significant tendency towards higher milk production in Treatment A. Treatment A had a significantly longer lactation length in 1996. This longer lactation length accounts for some of the difference in milk yield but when all treatments had their lactations adjusted to a 305-day lactation, there was still a tendency towards higher milk yield with Treatment A. There were no significant differences in milk fat; protein and lactose yield in either year. Milk fat concentration was significantly lower (p<0.05) for Treatment S in Year 2, but not in year 1. When the two years results were combined Treatment S had a significantly lower (p<0.05) fat and protein yield and a significantly lower milk fat concentration (p<0.05). Differences between treatments in milk protein concentration in the two years combined were not statistically significant.

The effect of system of production on milk production by week of lactation is shown in Figure 2. The graph for 1997 is very similar to that for 1996. In early lactation in year 1 there was a higher peak milk production with Treatment S. Treatment A had higher milk production levels in the latter half of lactation and yields were more persistent until the end of lactation. There was an increase in milk production which coincided with turnout to pasture in Treatment A. Treatment S had the lowest milk production levels at the end of lactation in both years.

Treatment A and S had a highly seasonal milk supply pattern. Treatment S reached peak milk production at Week 35 of the experiment. Treatment AS had a more even or uniform supply pattern though it declined steadily after Week 50 of the experiment.

Grazing management in the early spring period

Spring grass has a higher feeding value than grass at any other part of the season. Autumn-calving cows are in mid-lactation at this time and are still producing high milk yields. The provision of early spring grass is an important consideration on farms with autumn-calving cows. The primary objectives are to give autumn-calving cows access to grazed grass as early as possible and to budget that grass until grass supply equals demand. Recent work by Maher et al. (1998) showed that 20-kg DM/cow/day was capable of supporting a high level of milk production and maintain optimum pasture utilisation. Higher levels of daily herbage allowances may result in only a small increase in milk production and under utilisation of pasture.

Grazing management in the mid-summer period

The supply of grass for grazing exceeds the demands of autumn-calving herds from mid-July onwards; this is because of the low herd demand for feed of the non-lactating cow. Grass growth rates are still relatively good at this time of year. The supply of grass on the farm should be monitored so as to avoid large surpluses (increase in farm cover) occurring. A larger area (16.4 ha) was conserved as surplus for Treatment A than for Treatment AS (13.1 ha) or Treatment S (12.6 ha) in both years of the study.

Autumn-calving cows generally have a higher body condition score at the end of lactation. It is important that they should not gain excess body condition during the dry period and that they calve down at a body condition score of 3.0 - 3.5. The strategy followed to restrict grass during the systems study was to graze the milking cows ahead of the dry cows. The pre- and post-grazing heights for the dry cows for both years are given in Table 5. The dry cows entered paddocks with varying progressing heights and grazed these paddocks to a pre determined target height of 6 cm. A post-grazing height of 6 cm resulted in dry cows being maintained at an optimum body condition score; and pasture quality was also improved for the remainder of the grazing season.

Table 5. Dry cow grazing measurements for 1996 and 1997.

The grazing management for lactating autumn-calving cows in the late autumn/early winter period is described in Experiment 2.

Reproductive performance

The data in Table 6 show the reproductive performance of all Spring and Autumn - calving cows. The effect of calving season on reproductive performance is also shown.

Table 6. The effect of calving season on reproductive performance combined for the two years of the experiment.

SEM = standard error of mean; NS, not significant (P<0.05); *P<0.05; ***P<0.001;

Differences in the calving to service interval and calving to conception interval between autumn and spring-calving cows were not significant but both followed the same trend, i.e., it were longest for autumncalving cows. The number of animals submitted for service in the first three weeks of the breeding season was significantly lower (p<0.001) for the autumn-calving cows (59%) than for the springcalving cows (87%). Differences in pregnancy rate to first and second service were not significant. The infertile rate was significantly higher (p<0.05) with autumn-calving cows (23%) than with spring-calving cows (10%).