Winter grazing and dead material effects on early spring pasture production

ABSTRACT

Improved performance of the national ewe flock along with the displacement of finishing farms into hill country has increased pressure on spring pasture production. Increasing spring pasture production to cope with this increased demand is considered one of the main areas of improvement needed in hill country farming. A trial was set up in the South Island of New Zealand to determine the response of spring pasture production to two winter grazing intensities across three levels of sward dead material. Total spring pasture production was unaffected by winter grazing intensity. Winter dead material levels significantly affected total spring pasture production. Plots with high dead material levels (>60%) grew 657 kg DM/ha less than plots with low dead material levels (<30%) over the spring period. Reducing the amount of dead material in the sward is recommended as the shading of grass tillers by dead material reduces growth of new pasture.

KEYWORDS:

• Winter grazing management
• pasture growth
• pasture quality
• dead material

Introduction

The continuing expansion of the dairy sector in New Zealand has displaced traditional sheep and beef finishing properties into hill country that historically accommodated breeding operations with very little stock finishing. Pasture production in these environments, particularly in winter and spring, is suppressed by lower temperatures (Cossens & Radcliffe 1978) that occur at higher altitudes. Pasture production decreases by an estimated 900 kg DM/ha for every 100 m increase in altitude for perennial ryegrass (Lolium perenne L.) and white clover (Trifolium repens L.) based pastures and by an estimated 450 kg DM/ha for every 100 m in altitude for browntop (Agrostis capillaris L.) dominant pastures (Cossens 1983). These environments are therefore more challenging than lowland environments to finish livestock (Copland & Stevens 2012) due to lower and later spring peak pasture production.

Compounding this issue is the significant increases in the performance, both reproductive and liveweight gain, of the national sheep flock over the past 20 years (Burtt 2015) that has increased feed demand on farm. Improved fecundity alongside an increase in mature weight of ewes has placed increasing pressure on late winter and early spring feed supply due to an increase in energy requirements for maintenance, gestation and lactation (Wall et al. 2012).

Management options that improve feed supply in early spring include the use of autumn and winter applied nitrogen (N) fertiliser, pasture management tools to aid the farmer in pasture control and the use of tactical winter grazing to ensure an optimised pasture sward in spring. Nitrogen fertiliser application can increase pasture production in hill country with response rates ranging from 2.0 to 32.7 kg DM/kg N applied (Ledgard et al. 1983; Lambert et al. 2003; Gillingham et al. 2007). The use of winter grazing management has also been well defined in terms of the effect of grazing intensity on pasture production on ryegrass-based pastures with Harris and Brown (1970) demonstrating that hard grazing in winter was detrimental to spring pasture production. Intense grazing during winter can also been demonstrated to have on-going effects by delaying the on-set of the spring growth period by up to six weeks (Harris & Brown 1970; Black 1975, 1978). More recent research demonstrated that maintaining higher pasture covers during winter resulted in increased pasture growth in late winter and spring (Coutinho et al. 1998).

What is poorly understood currently is what effect the level of dead material in winter has on spring pasture production. Dead material can be a significant issue on New Zealand hill country farms following summer and autumn. Farmers generally use a mob stocking grazing management to help reduce the amount of dead material in the pasture sward in the hope of increasing pasture production and hence stock performance.

To gain a better understanding of the implication of these trials were set up at two sites in the South Island of New Zealand to assess the impact of levels of winter dead material and winter grazing intensity had on spring pasture production.

Materials and methods

Sites

The field trial was undertaken at two sites, AgResearch Invermay Research Centre (site 1), Mosgiel, New Zealand (45°85′S, 170°39′E) and at AgResearch Winchmore Research Centre (site 2), Ashburton, New Zealand (45°79′S, 171°78′E) over the winter–spring period of 2012. Site 1 utilised a four-year-old perennial ryegrass and white clover pasture on a south west facing aspect with an average slope of 18°. The soil was a Warepa silt loam (New Zealand soil classification, mottled Fragic Pallic soil; (Hewitt 2010)) with a high soil fertility (pH 6.1 ± 0.03, Olsen P 39.7 ± 2.44, sulphate-S (SO₄-S) 10.9 ± 0.72, Quick test K 5.8 ± 0.73, organic matter 7.8% ± 0.15 and carbon:nitrogen 13.6 ± 0.15 (mean ± s.e., n = 18). Site 1 was at an elevation of 100 m above sea level with a mean annual rainfall and temperature of 830 mm and 9°C respectively. Site 2 utilised a three-year-old perennial ryegrass and white clover pasture on flat land. The soil was a Lismore stoney silt loam (New Zealand soil classification, Pallic Orthic Brown soil; (Hewitt 2010)) with a moderate soil fertility (pH 5.9 ± 0.05, Olsen P 25.7 ± 1.40, sulphate-S (SO₄-S) 6.7 ± 0.80, and a Quick test K 6.0 ± 0.34). Site 2 was at an elevation of 150 m above sea level with a mean annual rainfall and temperature of 710 mm and 11°C respectively.

Treatment design

A randomised design was used with a factorial treatment structure of three winter pasture dead material levels (high (50–70% dead material), medium (30–50%), low (10–30%)) × two winter grazing intensities (lax (1200 kg  DM/ha) and intense (600 kg DM/ha)) replicated three times at each site. The plots were identified in late July within a single paddock at each site that had been rotationally grazed with ewes from early summer through to late autumn. Plots were 10 × 10 m at site 1 and 30 × 30 m at site 2. Visual identification of plots were used initially and verified by botanical analysis. Plots were fenced using temporary electrified netting and randomly allocated to either a 600 or 1200 kg DM/ha targeted winter grazing residual. Ewes were introduced on 7 August into the plots adjusting numbers depending on the pre-grazing pasture mass with the intent to achieve the target grazing residual over a 24-hour period. The plots and ewes were checked on a regular basis with sheep being removed when either the desired grazing residual was reached or if there was a risk of excessive pasture damage through treading.

During the spring period the plots were grazed approximately fortnightly depending on weather, starting from 10 September through to 6 November at site 1 (five spring grazing events) and from 11 September through to 1 November at site 2 (four spring grazing events). A 14-day grazing interval was chosen to simulate the defoliation pattern exhibited when continuous grazing (Hodgson 1966) as practiced by sheep farmers in spring. Ewe numbers were adjusted according to pre-grazing pasture height. The target grazing residual was 1200 kg DM/ha for the spring regardless of previous treatments. If the weather during the grazing event was predicted to be wet and pugging of pastures was possible the grazing was delayed until the risk of pugging had diminished.

Measurements

Herbage mass was assessed pre-and post-grazing per plot using a F400 electronic rising plate metre (Farmworks Ltd., Fielding, New Zealand) ensuring that no less than 40 readings were recorded across each plot by walking in an even zig-zag. Compressed height was then converted into pasture mass using individual calibrated equations for pre- and post-grazing plots at each site. These equations were derived from the randomly placed quadrats described below from a composite of all cut samples during the experimental period.

In each plot four randomly placed 0.15 m² (site 1) or 0.25 m² (site 2) quadrats, one in each quarter, were used to assess pasture cover, morphological components and botanical composition. Quadrats were cut to ground level using electric hand shears and collected at every grazing event for both sites pre- and post-grazing. Within each plot the samples from each quarter were bulked together and transferred to a washing facility to remove any soil contaminants. The samples were then drip dried to remove excess water prior to weighing and then sub-divided into samples for dry matter determination and morphological and botanical composition analysis. The sub-sample for dry matter determination was dried at 90°C in a fan forced oven for 48 hours prior to weighing to determine herbage mass (kg  DM/ha). Morphological and botanical composition sub-samples were reduced to 400 pieces using the quartering method (Lynch 1966) prior to dissection into five categories: green grass leaf, grass stem, legume, weeds and dead material. The wet dissection samples were then dried at 90°C in a fan forced oven for 48 hours prior to weighing to determine proportions of the plant components in the pasture sward.

Statistics

The dry matter growth in each of the spring growth periods (i.e. between grazing events) and herbage composition at each measurement date were analysed with analyses of variance with residual plots checked for constancy of variance and consistency with normality. Winter grazing intensity and dead matter level and their interaction were used as fixed terms with each site analysed separately. All statistical analysis was carried out using Genstat (Version 16).

Results

Monthly rainfall and average monthly temperatures for the two sites are shown in Figure 1.

Figure 1. Monthly rainfall and average monthly air temperature for the two trial sites for the duration of the research period. Data were recorded at Winchmore electronic weather station (National Institute of Water and Atmosphere Research, New Zealand, http://cliflo.niwa,co.nz) and Invermay Agricultural Centre.

Pre-grazing dead material averaged 61%, 47%, and 28% (p < .001) for the high, medium and low dead material plots respectively at site 1 at the start of the trial (July) and 75%, 50% and 5% (p < .001) for the high, medium and low dead material plots respectively for site 2 (Figure 2).

Figure 2. Percentage of dead material in the pre-grazing pasture swards at site 1 (a) and site 2 (b) for the three winter dead material treatments over spring.

The winter grazing of the plots resulted in average pasture residuals of 890 and 653 kg DM/ha for the intensely grazed plots at sites 1 and 2 respectively and 1378 and 1874 kg DM/ha at sites 1 and 2 respectively for the laxly grazed plots.

Pasture production as determined by the plate metre did not differ significantly over the first two measurement periods in early spring at either site regardless of winter grazing intensity or the quantity of winter dead material (Tables 1 and 2). However, the quantity of winter dead material did significantly affect the amount of pasture grown in periods 3–5 (11 October–19 November) and 4 (3–20 November) at sites 1 and 2 respectively (Tables 1 and 2). The effect culminated in total pasture production for spring that was significantly lower (p < .001 and 0.006 for sites 1 and 2 respectively) in plots that had high quantities of dead material in winter compared to those that had low and medium levels of dead material. Laxly grazed plots grew significantly (p = .014) more grass in the fifth measurement period (8 November–19 November) than intensely grazed plots at site 1 but there were no significant differences detected between winter grazing treatments at site 2. However, there were no significant differences for total spring production between intense and laxly grazed swards at both sites.

Table 1. Site 1 spring pasture production (kg DM/ha) of swards subjected to direct grazing of different intensity and dead material levels.

		Sampling dates (period)					Spring total
	Treatment	24 September (1)	8 October (2)	23 October (3)	5 November (4)	19 November (5)
Winter grazing intensity	Intense	326	239	387	435	474^a	1862
	Lax	225	272	416	466	565^b	1945
	SED	66.4	49.2	35.4	28.9	31.7	77
	p-value	.154	.519	.436	.304	.014	.306
Dead matter	High	309	167	343^a	389^a	438^a	1647^a
	Medium	262	295	361^a	444^a	494^a	1857^b
	Low	256	304	500^b	519^b	627^b	2206^c
	SED	81.3	60.3	43.3	35.4	38.8	94.4
	p-value	.785	.075	.007	.011	.001	<.001

Note: Differing superscripts within columns indicate significant differences (P < 0.05).

Table 2. Site 2 spring pasture production (kg DM/ha) of swards subjected to direct grazing of different intensity and dead material levels.

		Sampling dates (period)				Spring total
	Treatment	27 September (1)	15 October (2)	30 October (3)	20 November (4)
Winter grazing intensity	Intense	843	982	918	1097	3840
	Lax	905	868	941	1245	3959
	SED	76.8	55.7	127.9	68.1	189.0
	p-value	.435	.063	.859	.051	.541
Dead matter	High	772	846	846	905^a	3369^a
	Medium	903	948	1010	1346^b	4206^b
	Low	948	980	934	1263^b	4124^b
	SED	94.1	68.2	156.6	83.4	231.5
	p-value	.194	.166	.592	<.001	.006

Note: Differing superscripts within columns indicate significant differences (P < 0.05).

There were no differences in the amount of dead material between the two winter grazing intensities at both sites with the one exception being the post-grazing residual at site 2 where the laxly grazed swards had significantly (p = .006) more dead material than their intensely grazed counterparts.

The percentage of the pre-grazed sward made up of dead material for each of the winter dead material groups remained significantly different at the end of the trial (p = .011 and p = .030 for site 1 and 2 respectively). Plots that had high and medium levels of dead material in the winter reduced from an average of 68% and 48% of dead material down to 26% and 19% respectively (Figure 2).

Winter grazing intensity and winter dead material level had no effect on legume or weed production during the spring period (p > .05). Reproductive stem and seed head was not present in the sward until the final grazing period and was unaffected by winter grazing intensity. However, there was a significant difference in the percentage of reproductive material making up the sward between the winter dead material levels at both sites with more reproductive material being present in the swards with lower dead material during winter (Figure 3).

Figure 3. Percentage of the last post-grazing sward sample (late spring, 18 November) made up by reproductive material (stem and seed head) for the three winter dead material levels at site 1 (Invermay) and site 2 (Winchmore).

Discussion

The build-up of dead material in a pasture sward during late spring and summer in laxly grazed swards (Korte & Sheath 1978) affects total sward quality (Litherland et al. 2002) and hence animal intake (van Soest 1994). This study shows that the growing potential of the pasture during a critical feed deficit period in the following spring also decreases as dead material increases (Tables 1 and 2). Ensuring that dead material is removed from accounting for >60% to below 30% of the sward in the winter allowed an extra 660 kg DM/ha to be grown over the entire spring period. This equates to being able to carry an estimated extra 3.9 twin-bearing ewes per hectare over the eight-week spring period using the feed requirements given in Geenty and Rattray (1987), assuming that the ewes are a 60 kg liveweight and parturition occurs at the start of the given eight-week period.

Winter grazing severity in this experiment did not affect early spring pasture production as similarly demonstrated by Lee et al. (2005). The lack of an effect is most likely due to the pasture having a WSC (water soluble carbohydrate) store in the leaf sheath left after grazing that is adequate to support regrowth regardless of the above ground dry matter. In previous trials that demonstrated a difference in pasture growth between grazing severities (Harris & Brown 1970) the pasture was continually defoliated to low levels which could have depleted the WSC store, therefore reducing future regrowth.

The decrease in pasture production over the spring period from the high dead material plots could potentially be explained by the shading of the tillers during winter. While tiller densities were not recorded in this trial, previous research has indicated that tillers that develop in winter are important for providing a dense vegetative sward for the main spring growth period (Korte 1986). The high amount of dead material during the autumn and winter period may have shaded the base of the sward decreasing the light intensity and therefore reducing tiller development (Hunt & Field 1979) and density, resulting in lower pasture production during the following spring period. This is reinforced by the increased amount of reproductive material being present in the low dead material plots (Figure 3). Korte (1986) found that 30% of tillers that were present prior to winter and 6% of those that developed during winter became reproductive in spring if they were still present. This would indicate that more live tillers were present in the low dead material swards resulting in higher occurrences of reproductive material but also greater growth.

Achieving a winter sward with low dead material requires specific grazing management during the previous spring and summer period to control the reproductive phase of the grass species present. In perennial ryegrass-dominated pastures, close grazing in late spring that removes the apical meristems on flowering tillers prevents stem development and consequent build-up of dead material, creating dense leafy swards in summer (Korte 1982) that produce more growth in autumn and winter (Korte et al. 1984). Farmers that utilise an autumn saved pasture cycle to ensure feed during winter can decrease the amount of the dead material in the sward by ensuring that grazing management during the spring period is effective at removing the apical meristems and thus prevent build-up of seed head and consequent dead material in the sward.

However the majority of hill country pastures are dominated by a range of other grass species that flower over a longer period making it hard to develop a grazing plan that ensures that the critical plants are grazed at optimal times in a rotational grazing situation. If farmers are not able to prevent the build-up of the dead material in the sward in late spring they may have an opportunity to remove it prior to the winter period. Waiting until the winter period especially if using a one off grazing event, may impact significantly on spring pasture production, as demonstrated by these results.

Acknowledgements

The authors would also like to acknowledge the AgResearch Farm Systems and Farm field teams.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

The authors would like to acknowledge Ministry of Business, Innovation and Employment, DairyNZ, Fonterra Co-Operative Group, Beef + Lamb New Zealand and DCANZ for funding.