ROTATIONAL GRAZING WITH DAIRY HEIFERS ON WICST:
Pasture Productivity and Quality

Janet Hedtcke ^[1] and Josh Posner ^[2]

INTRODUCTION

Traditionally, pasture has been left for land too poor for cultivation and was under a continuous grazing regime.  However, these same pastures can produce high quality, high yielding forage if properly managed.  Successful rotational grazing is based on a complex relationship between plants, animals and management and is more productive than continuous grazing by up to a 2000 lb dry matter/acre (Undersander et al, 1991).  There are varying degrees of intensity of rotational grazing based on frequency of movement from twice daily to weekly.  Movement that reflects the plant growth rate and not based on rigid time schedules is referred to as management intensive grazing (MIG).  Of the three dairy forage systems represented in the Wisconsin Integrated Cropping Systems Trial (WICST), the MIG system has the lowest input and is the most biologically diverse system.  Productivity is measured in terms of animal performance and forage production.  Long-term effects from grazing are also being monitored on the soil, the vegetation and water quality (Lakeland site).  This paper focuses on forage availability and forage quality of the pasture plots over the course of the WICST project.  Animal data can be found in the next paper of this report entitled ‘Rotational Grazing with Dairy Heifers on WICST: b. Animal performance summary’.

MATERIALS AND METHODS

Pastures were established at the Lakeland Agricultural Complex (LAC) and the Arlington Agricultural Research Station (ARS) in 1990.  Initially, species included red clover (Trifolium pratense), smooth bromegrass (Bromus inermis) and timothy (Phleum pratense); later, orchardgrass (Dactylis glomerata), perennial ryegrass (Lolium perenne)and reed canarygrass (Phalaris arundinacea) were also included (see Table 1 for seeding dates and rates).  At LAC, the forage was mechanically harvested in 1991 and grazing began in 1992.  At ARS, severe winterkill of the grasses required reseeding in 1992.  Slow establishment due to drought delayed grazing there until the spring of 1993.  Red clover was seeded (frost seeded at LAC and drilled at ARS) into the pastures on alternative years through 1995.  Thereafter, red clover was seeded at a reduced rate when deemed necessary.  Two of the paddocks at LAC (reps 1 and 4) were tilled and reseeded in August of 1996 to repair severe trampling damage, which occurred during the wet weather earlier that summer.  The fall seeding of 1996 there failed and there was more reseeding in the fall of 1997.  No grazing or mechanical harvest occurred on the plots at LAC in 1997.

Table 1.  WICST pasture seeding dates and rates.   Arlington Ag Research Station Lakeland Agricultural Complex Year Date Species (lb/acre) Date Species (lb/acre) 1990 23 April 'Marathon' red clover 7.0 30 May 'Marathon' red clover 6.1   'Badger' smooth bromegrass 8.0 'Badger' smooth bromegrass 3.0   'Toro' timothy 4.0 'Toro' timothy 3.5 1992 30 April Orchardgrass (early) 6.0 - - -     'Badger' smooth bromegrass 12.0 - - -     'Toro' timothy 6.0 - - -   31 July Orchardgrass (early) 4.5 - - - 1993 9 April 'Arlington' red clover 12.0 7 April 'Arlington' red clover 20.0 1995 17 March 'Arlington' red clover 15.0 24 March 'Arlington' red clover 18.0 1996 26 April 'Arlington' red clover 6.0 9 March 'Arlington' red clover1 18.0   - - - 22 April 'Arlington' red clover2 6.0   - - -   timothy 4.0   - - -   perennial ryegrass 3.0 1997 5 April 'Arlington' red clover 6.0 21 August reed canary grass 8.0   - - -   Orchardgrass (early) 3.0   - - -   'Arlington' red clover 8.0 1998 - - - 30 March 'Arlington' red clover3 15.0 2001 13 April 'Arlington' red clover 12.0 26 March 'Arlington' red clover 6.0 1 Seeded with no-till drill; otherwise broadcast seeded except for drilling with 1990 establishment. 2 Reps 1 and 2 3 Heavy seeding due to poor pasture conditions

Throughout the history of the trial, dairy heifers were cycled 5-6 times per season through the four replicates of the pasture system at both sites.   Three grazing management strategies have been used.  From the beginning of the trial through 1995, each repetition was stocked with two 500# heifers and they were moved up and down the plot to simulate rotational grazing.  Excess forage was removed as hay (see Tables 2 and 3).  This system worked well statistically and replicated observations were taken each year on forage quality, animal weight gain and soil characteristics.  Unfortunately this system was too constrained for best management of the paddocks.  For the next four seasons (1996-1999) a put-and-take system was followed where twelve 500# heifers started as a cohort and moved through the four repetitions.  As feed became limiting after the early spring flush, some heifers were removed.  Little to no hay was taken during this period (see Tables 2 and 3).  It was decided however, that this system put too much pressure on the paddocks in the early season and had adversely affected weight gain of the heifers that were removed early.  This system also made it impossible to conduct economic analysis as once the heifers were removed they left the experiment, which is not the case when actually grazing heifers in a custom operation.  For the past three seasons (since 2000) we used a compromise system of putting only four heifers on the plots and treating them as a cohort.  This has resulted in some haying in the spring, but the paddocks appear to be in better health.  This translates into a stocking rate of 0.7 animal units/a (1 animal unit=1000 lbs).

Table 2.  Mechanical hay harvested on pasture plots over the course of the trial at ARS.
Year1	Date	Yield (tons DM/a)	Plot #/p>	Crude Protein (%)	RFV
1993	June 11	0.54	112,302,405	13.8	90
1994	May 31	1.08	All 4 plots2	13.3	103<
	October 21	1.10	All 4 plots2	16.7	125
1995	June 5	0.38	All 4 plots2	12.1	101
	October 9	0.99	All 4 plots2	16.9	93
1997	June 14	1.43	405	-	-
1998	May 12	1.68	302	19.4	145
2000	May 16	1.15	207	18.1	120
	June 6	1.68	302	9.6	82
2001	May 25	1.52	207	-	-
	June 30	1.04	405	10.7	89
2002	May 6	1.20	207	-	-
	May 16	0.94	405	18.7	132
	June 11	0.23	405	15.5	108

¹ No hay harvested in 1996 or 1999.  ² Only partial plot was taken as hay while the rest was grazed.

Table 3.  Mechanical hay harvested on pasture plots over the course of the trial at LAC.
Year1	Date	Yield (tons DM/a)	Plot #	Crude Protein (%)	RFV
1992	May 30	1.43	104,314,408	16.5	136
1994	June 16	0.39	All 4 plots2	12.1	99
1995	May 31	0.28	All 4 plots2	11.6	79
	June 23	0.32	104, 314	15.4	90
2000	June 20	1.59	408	16.6	-
2001	July 9	0.75	408	6.6	84
2002	June 9	1.28	408	13.9	102

¹ No hay harvested in 1993, 1996-1999.  ²  Only partial plot was taken as hay while the rest was grazed.

Since yield is difficult to measure in grazing systems, forage availability will be used to describe the pasture productivity.  Available forage differs from yield for two reasons: 1) to measure forage availability, forage is clipped to ground level (i.e. 2-in. stubble) but grazing height is rarely below 4 inches and is very selective and uneven, 2) at each successive cycle, senesced and refused/trampled forage remains and will be resampled thus overestimating yield and underestimated quality ^[3] .  

Samples were collected throughout the season from the paddock area to which the heifers were being moved (plots at Lakeland were not sampled consistently beyond 1999).  At the beginning of each grazing cycle, direct sampling (as discussed in CSSA special pub # 16, 1989) with a quadrat and sheers was used to estimate total available forage.  We have used a 0.25m² (0.5mx0.5m) quadrat but in the future we will use a 3-sided frame that measures 1.0x0.5m², as recommended by Mannetje and Jones (2000).  The 0.5m² sampled area, referred to as a strip, will allow more adequate sampling for most species. 

Forage was dried (in 60°C oven for at least 48 hr) and forage availability is expressed on a dry matter basis.  Samples were sent to the UW Forage Extension Lab in Moore Hall for quality analysis.  Using NIRS (ghay.eqa), forage quality was estimated for crude protein (CP), neutral detergent fiber (NDF), and acid detergent fiber (ADF).  Relative feed value (RFV) was calculated from NDF and ADF for each sample. 

Inputs

Herbicide.  Generally, little to no herbicides have been used on the pasture with two exceptions:  At LAC 1997, Roundup Ultra (1.5 pt/a), and 2,4-D (1 pt/a) was applied during plot renovation and at ARS in July 2002 when ‘Stinger’ was spot sprayed to control Canada and bull thistle; approximately 1 gallon of solution (with 7 ml of product) was used for the 2.8 acres of pasture (or 2.5 ml stinger/a). 

Clipping.  Mechanical clipping with a rotary mower (to about 4” stubble height) has frequently been done through mid-July to remove seed heads and keep weeds in check thus improving the pasture stand and quality. 

N management.  Manure was added at 10 ton/a from 1990 to 1992 at both sites during plot establishment and again at LAC in 1996 and 1997 during plot renovation.  Synthetic N has occasionally been applied to the plots at a rate of 30-50 lbs N/a (1x in 1993, 1994, 1996, 2000).  Lakeland received synthetic N only once at 30 lb/a, in 1993.  For the most part, manure from the grazing heifers and the red clover (see Table 1 for seeding info) has provided enough N for adequate grass production.  To estimate the amount of nutrients in manure we have made some assumptions: 1) manure output for 500 lb heifer ~ 50 lb/d ‘as is’ basis, and 2) manure (both urine and feces) has about 10 lb N, 3 lb P₂O₅, and 8 lb K₂0 per ton (Cornell data).  Using a stocking rate of four- 500 lb heifers for 150d on 2.8 acres, about 55 lb N/a is excreted.  Although this isn’t an even spreading of nutrients, as heifers tend to linger around grain bins and water tank, the variability is reduced with rotational grazing vs. continuous grazing.

Pasture Management

Grazing details for years prior to 2001 have been reported in each of the previous WICST technical reports (2^nd-8^th).

Arlington 2001.  On April 25, paddocks were evaluated using a pasture condition score sheet compiled by Cosgrove et al. (1996).  All plots were in the ‘good’ category.  Late growth of legumes and low plant density reduced the score from the ‘ very good’ category.  To improve legume content, red clover was drilled in at 12 lb/a on April 13.  Some seedlings were beginning to emerge by late April.  Grazing began May 2 in 2001.    Plots were clipped at least once each by mid-July to remove mature seed heads and to increase light penetration for the newly seeded red clover.  Hay was mechanically harvested once at the end of May (plot 207, 1.52 ton dm/a) and once at the end of June (plot 405, 1.04 tons dm/a).  The excess forage harvested as hay was not needed later in the season as soil moisture was adequate for most of the season for timely regrowth. 

Continuing with the same management of the previous year, four dairy heifers flash grazed the paddocks the first month (36 days) of the grazing season until most clipping was done.  Thereafter, paddocks were divided into quarters and a back fence was erected to allow plant rest and regrowth.  Heifers were moved every 2 to 3 days to new section.  An alleyway was set up in each paddock to allow the heifers access to water at all times.  Approximately 1500 square feet of the paddock was used as this alleyway.  Throughout the grazing season, the heifers were cycled through the four paddocks an average of 5 times.  Grazing ended on October 2 for a total of 153 grazing days, about 30 days short of our goal.

Canada thistle is becoming an increasing problem in the pasture.  Hand weeding was done occasionally to remove thistles.  Fence line trimming was done in late May and again in mid-July to keep the electric fence clean, using a gas-powered, wheeled push trimmer.

Arlington 2002.  Plots were assessed and it was determined unnecessary to reseed red clover this year.  Grazing began May 1 using four 500 lb heifers.  Pastured heifers were given the entire plot for about 5 days and then rotated to a new plot.  After the first 10 days, the plots were subdivided into quarter section.  Only three plots were used for grazing in the first cycle and the other was mechanically harvested. 

All plots were clipped at least once after the spring flush.  Plot 207 and 407 were mechanically harvested once yielding 1.20 and 1.17 tons dm/a respectively.  Fence lines were trimmed in mid-June and did not need more trimming for the duration of the season.  Canada and bull thistles were present in each plot and sprayed with Stinger at the end of July. 

Plots were grazed 6-7 times during the season.  A dry summer and early fall resulted in low forage supply by mid-October.  Grazing ended on October 10 for a total of 163 grazing days. 

Lakeland 2001.  Heifers were put on plots on May 1, 2001.  Continuing with the same management of the previous year, four dairy heifers flash grazed plot 104 for the first 14 days.  Then, plots were subdivided and heifers grazed each plot from 15-30 days.  Heifers cycled through each plot three times during the season.  No legume was seeded this spring.  Hay was harvested on July 9 from plot 408 (2.68 ton dm/a) but this hay was not fed back to heifers.  Paddocks were sectioned and heifers were given new sections every four to five days.  The grazing season lasted for 158 days.

Lakeland 2002.  Four heifers were leased from a neighboring farm because the Walworth County farm’s dairy herd was sold last year.  There was no confinement group to serve as a control.  Grazing began in mid-May and ended in early November for a total of 172 grazing days.  Hay was harvested from plot 408 as round bales (1.28 ton dm/a) and fed back later in the summer.

RESULTS AND DISCUSSION

Long-term Productivity

Since the pasture plots were established, they have produced well with about a ton of dry matter/acre available on average at each sampling period at each site (Fig. 1).  This figure was built by pooling the harvest quadrat data (0.25m², 3-4 subsamples/date) from the 9 years at ARS and the 5 years at LAC (1992-1995, 1999; no reliable sampling at LAC after 1999).  Each point represents the mean of 4-16 sampling dates that fell within the indicated 10-day period across years.  The growth curves at each site follow the typical cool-season grass growth pattern with highest production in spring/fall and lowest in mid-summer, growth being related to temperature and moisture.  At ARS, the growth distribution was a bit more even that at LAC although both sites experienced the ‘summer slump’, or reduced forage production in late June and July.  The flat lay of the paddocks at ARS allows most of the rainfall to infiltrate instead of running off and the excellent water holding capacity of the soil is a factor of sustained production during the dry summer months.  At LAC, reduced summer yields may have been due to a combination of factors such as higher slopes, slower infiltration, spring trampling damage due to slowly draining, heavy soils, and hotter, drier summer weather at this site.

1 no yield data for 1996 at ARS

2 Sampling period denotation: May 1 = May 1-10th, May 2 = May 11-20th, May 3 = May 21-31st, etc.

Long-term Quality

Forage quality has been very good with protein levels fairly steady within the season (Fig. 2).  Average crude protein was 18.4% across sites and peaked at 21% when grass regrowth was leafy and legumes were prevalent.  Generally, protein levels dropped as low as 14% during mid-June as grasses matured.  Relative feed value, averaged 123 across sites and years ranging from 100 to 160 (Fig. 3).  Arlington tended to have higher quality grass than LAC at most sampling points perhaps due to a more favorable environment at ARS.  It should be noted that RFV is based on alfalfa and not grasses.  Therefore, it isn’t quite appropriate to compare alfalfa to grass based on RFV.  Future analyses will be on RFQ.

 

¹ Sampling period denotation: May 1 = May 1-10th, May 2 = May 11-20th, May 3 = May 21-31st, etc.

¹ Sampling period denotation: May 1 = May 1-10th, may 2 = May 11-20th, my 3 = May 21-31st, etc.

CONCLUSION

Rotational grazing offers a low input system with very good forage quality and productivity.  Close attention to forage growth, animal condition, and weather conditions is important to maintain high forage production and quality.  Excess forage growth should be harvested as hay or clipped early in the season to maintain high quality and intake.  Quality will vary with plant maturity but post-grazing clipping can help keep the forage quality consistently high.  Red clover has been reseeded about every other year, which has helped maintain pasture quality.  By managing both forage and heifer needs, we can realize the potential of rotational grazing as a viable alternative to conventional heifer rearing. 

LITERATURE CITED

Cosgrove, D., D.Undersander, and M. Davis.  1996.  Determining pasture condition.  UW Extension Bulletin A3667.

Marten, C.G. (ed).  1989.  Grazing Research: Design, Methodology and Analysis.  CSSA Special Publication #16. 

Pasture and Range research techniques. 1962.  Prepared by joint committee of ASA, ADSA, ASAP, and ASRM.  Ithaca, NY.  Comstock Pub. Associates.  242 pp.

Mannetje, L.‘t, and R.M. Jones (eds).  2000.  Field and laboratory methods for grassland and animal production research.  New York, New York CABI public.  447 pp.

Undersander, D.J., B. Albert, P. Porter, A. Crossley, and N. Martin.  1991. Pasture for Profit: A guide to rotational grazing.  University of WI- Extension publication A3529.