Limitations to milk production from pasture.

Abstract

Limitations to Milk Production from Pasture W.J. Fulkerson and L. Trevaskis NSW Agriculture, Wollongbar Agricultural Institute, Wollongbar NSW 2477 Summary A milk yield of 20-25L per day from Friesian cows appears to be achievable from pasture as a sole feed. Genetic merit of stock grazing the pasture, availability of pasture and pasture species all influence the actual level of production, although to a surprisingly small extent. For example, even with extreme differences between cows in genetic merit (Australian Breeding Value for fat plus protein of 41 v 2kg/cow) there was only a 3.5L difference in daily milk yield/cow fi=om pasture. Recent studies in France with high producing cows have shown that available pasture (DM on offer) has to increase by 27kg DM/cow/day to increase milk yield by 2.6L fi=om 20.4L/cow/day. In extreme species comparisons, C4 grasses produced 5L milk/cow/day less than C, grasses whilst clover may give 3.5 L more, although these studies were with relatively low producing animals. Lifting the ceiling on production with the factors discussed above may be wasteful (increased feed availability) or may not be sustainable (pure clover swards). The potential exists to increase milk production from pasture by improving the proteincarbohydrate ratio, which is too high in most pastures, by supplementing cows with carbohydrates. However, there are practical problems in synchronising the availability of carbohydrates, in cows fed twice-a-day at milking, to promote use of excess dietary N from pastures. Two possible approaches to this problem are: l adjusting grazing times and perhaps species. Dependant on time of day, regrowth stage and season of the year, and within practical reality, soluble carbohydrate levels in ryegrass have been shown to vary from c2% to over 3 0% with protein levels usually tending in the opposite direction. This provides potential to manipulate the CP: WSC ratio substantially. Introduction What limits milk production capacity of dairy cows grazing pasture? This question is being increasingly asked by dairy farmers as they strive to improve financial margins by reducing costs or by increasing production, or both. Is it the physical limitation imposed by ingestion of pasture by the animal (see de Jong 1986) or is it within the rumen, or is milk production limited by a lack of certain nutrients in pasture? There is evidence that production per cow is limited to 20-25 L milk/day when temperate pasture is the sole feed. Van Soest (1982) claims that the primary determinant of intake is Neutral Detergent Fibre (NDF) content of feed. There is certainly a very significant positive linear relationship between forage digestibility and intake (Hodgson, 1977). However, vanvuuren (1993) has shown that with high quality forages, rumen capacity changes in a positive way with NDF content, indicating that NDF may not be the major limitation. According to Beever and Siddons (1986), production of milk from abundant pasture is limited by insufficient amounts of essential amino acids bypassing the rumen to the small intestine. However, there is little evidence for a production response of cows grazing pasture to two of the most limiting amino acids methionine and lysine - when fed in a form protected from degradation in the rurnen, at least in Australia (Kikuyu; L. Trevaskis, unpub.data.) and New Zealand (Ryegrass; M. van Houtert, pers. comm) when Feed a readily fermentable carbohydrate to stimulate microbial activity pre-grazing l Feed a slowly degrading source of carbohydrate to match the release of N from pasture during grazing. However, a better option may be to ensure a high level of non-structural carbohydrates in the pasture by Recent Advances in Animal Nutifiion in Australia 1997 University of New England, Armidale NSW 2351, Australia 160 Fulkerson, w. J. and Trevaskis, L. production is 20-28L milk/cow at peak lactation. This is in line with observations of Oldham (198 1) that, in animals whose demand for metabolisable protein/unit of metabolisable energy is around 6.5g MP/MJ of ME, the quantity and quality of amino acids in microbial protein alone should be enough to satisfy their needs and this equates to a milk production level of 2OL/cow/ day. In fact, Virtanen (1966) showed yields of milk up to 5OOOL/cow per lactation can be sustained on diets where the sole source of N is urea. Therefore on pasture-only diets, where dietary N levels may already be excessive, amino acid supplementation would not be expected to result in a production response in cows producing 20L milk/day or less. The more likely restriction would appear to be energy (Meijs, 198 1) and van Vuuren (1993) has calculated that, under ideal conditions, energy in pasture restricts milk yield to 27 L/cow/day. In practice, 5,500-6,OOOL milk/cow/lactation has been obtained from cows grazing well-managed ryegrass (Lolium perenne)lwhite clover (Trifolium repens) pasture as the sole feed (plus silage made fYom that pasture) (M. Blacklock, pers. comm.). Such pastures can also be the basis for a total yield of 8000-9OOOL milk/cow/lactation when there has been judicious use of concentrates (G. Hough, pers. comm.). These levels are equivalent to an average production of 20-22L milk/ cow/day fiom pasture, over the entire lactation. What possibilities exist for breaking through this production ceiling? The following factors may all have a major influence on potential milk production corn pasture. l Genetic merit of stock grazing pasture The parentage of the Australian Friesian dairy herd is becoming increasingly based on cows selected within a total mixed diet/feed lot system of farming in North America. These cows are not selected for foraging ability (bite size, grazing time), nor for good feet to walk the increasing distances required in larger and larger herds at pasture, nor tolerance to heat or other adverse weather conditions. It is true that progeny are tested under Australian conditions but perhaps we are simply comparing North American genes. A study at Wollongbar, underway for 2.5 years, aims to see if high genetic merit cows retain their production advantage if they are fed on pasture alone. Comparisons were made between farmlets whose pastures were initially similar but stocked with high or medium genetic merit cows. The results in Table 1 show that high genetic merit animals produced more on ryegrass/white clover pasture than medium genetic merit cows given the same feed availability. Thus, at constant body weight (and by inference, all nutrients for milk coming from pasture) high ABV cows (in the top 2-3 herds in NSW for ABV) produced 3.5L more milk/cow/day than medium genetic merit cows (ABV equivalent to 1988 national herd) without affecting protein content. The ability of high genetic merit cows to produce more was due to their ability/willingness to partition more feed to milk and to graze harder to achieve the higher intake. Post-grazing residues were more than 200kg DM/ha lower on the high genetic merit farmlet than the medium genetic merit farmlet. These results are consistant with studies in New Zealand (Anon, 1983; Grainger et al. 1985;) and inIreland (see Halleron, 1994), also on pasture. Genetic merit of stock grazing pasture; Pasture availability; Variation between pasture type/species in chemical composition and digestibility; Balancing nutrients deficient in pasture with appropriate supplements. l l l Table 1 Condition score, Iiveweight and production of Friesian cows of high and medium genetic merit in early-mid lactation at a time of constant Iiveweight. Limitations to milk production for pasture 761 Pasture availability Increasing pasture availability (kg DM/cow) increases intake and hence production per cow in a curvilinear fashion, but the utilisation of pasture offered is substantially reduced and unless steps can be taken to remedy this, the benefits may be negligible or negative. Recently, Peyraud et al. (1996) found that herbage available per cow had to rise f?om 19 to 46kg OM/cow/ day to achieve a rise in OM intake of 2.9 to 16.7 kg/cow and to raise milk yield from 20.4 to 23 .O kg/cow/day. The provision of an extra 27 kg OM per cow would be of dubious benefit unless other measures could be taken to offset the consequent wastefully low utilisation, such as using followers (dry stock). Pasture type/species There is a marked difference in milk production potential between C, (temperate) and C4 (tropical) grass pastures. A production ceiling of 12L/cow/day has been claimed for C, grasses (Cowan, 1975), but up to 15L/cow/day on kikuyu (Pennisetum clandestinum) has been achieved with appropriate management (Reeves et al. 1996). This management relies on developing a dense canopy of leaf with the stem being removed mechanically after grazing. The stem of kikuyu has a ME value of about 7.5 MJ/kg DM whilst the leaf may be up to 9.5 MJ/kg DM. Intake of kikuyu appears to be restricted to 13 kg DM/cow/day, probably due to the high NDF levels (6575% v 3545% for kikuyu and ryegrass, respectively) and perhaps by the low levels of fermentable carbohydrates (2-6% water soluble carbohydrates (WSC) plus starch) (Fulkerson et al. 1997). The sodium and phosphorus concentrations and the availability of calcium are also low (Reeves et al. 1996). Intake of short rotation ryegrass (L. multiflorum) by dairy cattle has been shown by Wilson (1966), to be higher than of perennial ryegrass although a difference in milk production has not been shown. The difference in intake may be due to higher levels of fermentable carbohydrates (Fulkerson et al. 1994) and preferential selection for the short rotation ryegrass if they have a choice (W Fulkerson, unpub data). Cows giving rather low production (13.4 L milk/d) on abundant ryegrass produced about 25% more milk (16.7L milk/d) on pure white clover pastures, with milk fat plus protein 35% higher. These production increases were associated with a 33% increase in DM intake (Rogers et al. 1982). The higher intake of clover is believed to be due to its lower cell wall content and higher levels of protein and cell constituents (Dermarquilly and Jarrige, 1973) and can be expected to result in a higher amino acid flow from the rumen. Beever et al. (1986) showed that the quantity of amino acids entering the duodenum was 30% higher in dairy cows fed white clover than those fed ryegrass. A comparison of the rates of degradation of N for these three pasture species is shown in Figure 1 and reflects their milk production potential. Balancing nutrients in pasture In ideal dairy pastures, the rate of microbial growth in the rumen is dependent on the availability of protein (N) and carbohydrates (energy) (Van Vuuren, 1993). Protein content is nearly always too high whilst the levels of non-structural carbohydrates (free sugars, tictosans and starch) are too low. This results in high rumen ammonia levels and inefficient use of N which is reflected in high urine N loss and milk urea levels. For example, the mean milk urea levels in cows grazing pasture in Australia is over 400 mg/L (L Trevaskis unpub. data) compared to 150-300 mg/L for cows receiving a completely balanced ration in North American feed-lot dairies. Excess dietary protein intakes leading to high rumen NH, levels, can potentially reduce production, because of the energy required to synthesize and excrete urea (Blaxter, 1962). Sometimes @rod and Butler, 1993), but not always (Howard et al. 1987), high protein intake may reduce reproductive performance, although this has not been shown in Australia under grazing conditions (L. Trevaskis, unpub. data). Supplementing with carbohydrate Work by Hoover and Stokes (199 1) indicates that the ratio (g/g) of non-structural carbohydrates to degradable intake protein should be about 2.0 to optimise microbial activity. Apart from the ratio of carbohydrate to protein in the feed, studies by Sinclair et al. (1993) have shown the importance of synchronising the availability of these three feed components for the rumen microbes in stall-fed sheep. 762 Fulkerson, w. J. and Trevaskis, L. The provision of additional carbohydrates to match (synchronise with) the release of the high levels of N liberated from pasture at grazing is difficult. Supplements are usually fed twice-a-day at milking whilst the most intense grazing activity is for 4h following milking (Cowan 1975). The option of moving cows to and from feeding stalls at this time, to increase frequency of supplementary feeding, completely compromises effective grazing management and would probably negate any benefit. One option may be to feed a carbohydrate source which has its highest rate of degradation coinciding with the peak period when protein N in pasture is being liberated by microbes in the rumen following grazing as shown in Figure 2. An alternative proposal may be fermentable carbohydrates in the dairy at to increase microbial numbers to cope of forage at the subsequent grazing. effective, the following conditions may l to feed readily milking in order with the inflow For this to be need to apply: The level of soluble carbohydrates in the grass would need to be adequate to maintain a larger microbial population. This may not be the case with young grass (Fulkerson and Slack, 1994). Work by Hesbell(1979) has shown that up to 60% of rumen microbes die within 2h if they have an inadequate supply of energy. Hence the time between concentrate feeding and start of grazing would have to be less than 2h. l Generally, however, the effect on N-utilisation of feeding carbohydrate-based concentrates to cows grazing pasture has been small. There may be two explanations for this. Firstly, that the efficiency of microbial synthesis from pasture is already high (J. Peyraud, pers.comm.). Secondly, if the metabolisable protein received from pasture diets is adequate for milk production of up to 2OL/cow/day, any increased flow of metabolisable protein to the small intestine would be absorbed and could lead to increased absorption and catabolism of amino acids and increased urinary N output (Peyraud et al. 1996). Figure 2 Schematic representation of the degradation of pasture N in the rumen in a 4h grazing period after feeding a carbohydrate-based concentrate with carbohydrate degradability characteristics to match N availability in the rumen. Changes in crude protein (CP) to water soluble carbohydrates (WSC) ratio with time of grazing It seems more appropriate to improve the ratio of CP to WSC in the plant itself. Studies at Wollongbar and elsewhere have shown large differences in this ratio dependent on time of grazing. Time of day The level of non-structural carbohydrate in the pasture plant is the result of the balance between gains from photosynthesis and loss through respiration. As a consequence, WSC levels are lowest at sunrise, after respiration during the night, and highest in late afternoon. Diurnal changes in the levels of WSC in the leaf of perennial, and `annual', ryegrasses (Fulkerson et al. 1997) and kikuyu grass (Reeves et al. 1996) in the subtropical environment of Wollongbar are shown in Figure 3. The absolute values for ryegrass are relatively low because samples were taken in early autumn. Overall, there is about a 0.5% rise in WSCkour during daylight hours. Thus, a cow eating 15kg DM of perennial ryegrass from 3 to 6 pm would ingest 0.8 lkg more WSC than if she grazed her pasture allocation at 5 to 8am There is Figure 3 Percent water soluble carbohydrates in the leaf of annual tyegrass (M), perennial ryegrass (V) and kikuyu (1) estimated for samples taken between 06.00 and 18.00h. Limitations to milk production for pasture 163 some potential to take advantage of this by providing a new block of pasture after the PM, rather than AM, milking. Intake is always highest when a fresh block of pasture is given and declines with time (Walker and Heitschmidt, 1989). Incoming solar radiation Leaf carbohydrate levels can be markedly depressed by cloudy weather as shown in Figure 4. There is a very close relationship between solar radiation, sunlight hours and WSC in both leaf and stem. The WSC in leaf appears to fluctuate less than in the stem-consistent with the function of the stem as a storage organ. The fall in WSC in cloudy weather may be a factor responsible for the decline in milk yield observed in the field after prolonged cloudy weather and carbohydrate supplementation might then be beneficial. Regrowth There are major changes in the carbohydrate and protein contents of grass with regrowth time and this is affected by season, as shown by the information from Wollongbar in Figure 5. In June, the ratio of CP to WSC changes from 4: 1 to 1:2 as ryegrass regrows to 3 leaves/tiller. Clear skies at this time of year ensure ample incoming solar radiation, and the cool nights minimise carbohydrate loss through respiration. In September, with higher temperatures, the change in ratio of CP and WSC with regrowth is not as marked, while in November there is no clear pattern. The time scale in Figure 5 is related to leaf number/ tiller in the knowledge that as ryegrass expands 3 leaves per tiller, and then each new leaf initiated is balanced by senescence of the oldest leaf In this way regrowth curves can be validly compared between seasons. Delaying defoliation to optimise the CP:WSC ratio will not reduce sward quality provided plants do not produce more than 3 leaves/tiller. The levels of NDF in ryegrass do not change significantly during the regrowth cycle in June (mean38%, SE& 1.4) or September(mean41.3k 1.14%). Season The seasonal variation in CP and WSC in a ryegrass/ white clover pasture is in accordance with the assumptions previously outlined (see Figure 6). The ratio of CP:WSC is lowest in early spring and highest from late spring to autumn. It is conceivable that leaf WSC may be 2% or less for samples of young grass (1 leafftiller) taken in the morning in autumn to well over 30% for mature grass (x3 leaves/tiller) in the afternoon in early spring. At the same time, protein changes in the reverse manner resulting in extremes of nutritive value. `Grass ain't Grass'. Figure 4 The percent WSC in the dry matter of perennial ryegrass stubble (m) and leaf (0) measured over an 18 d period. Corresponding mean values are given for solar radiation, (W/m*/1 00) (0) during the periods of sunlight of varying duration (V). Figure 5 Percentages of crude protein (CP) (0) and WSC (V) and the CP:WSC ratio (m) for leaf of perennial ryegrass taken at 3h after sunrise in July (A), September (B) and November (C). 164 Fulkerson, w. J. and Trevaskis, L. Figure 6 Percentages of crude protein (t) and WSC (1) in samples of perennial tyegrass plucked pre-grazing over a two year period. Samples were plucked to simulate grazing height for milking cows and were fitted using Table Curve (Jandel Scientific, San Rafeal, California, USA). Federation, Uppsala PL. Husbandv 28, 3341. Anon (1983). High BI Cows graze harder. &iv Exporter March pp. 30-32. Beever, D.E. and Siddons, R.C. (1986). Digestion and metabolism in the grazing ruminant. In: Control of Digestion and Metabolism in Ruminants. pp. 479-497 (Eds. L.P. Milligan, WL Grovum and A. Dobson). Prentice-Hall: Englewood Cliffs, N. J. Beever, D.E., Lasada, H.R, Garnell, S.B., Evans, R.J. and Harnes, M.J. (1986). Effect of forage species and season on nutrient digestion and supply in grazing cattle. British Journal of Nutrition 56,209-225. Blaxter, K.L. (1962). The Energy Metabolism of Ruminants. pp. 522. Hutchinson & Co: London. Cowan, T. (1975). Grazing time and pattern of grazing of Friesian cows on a tropical grass-legume pasture. Australian Journal of Experimental Agriculture and Animal Husbandry 15,32-37. Elrod, C.C., and Butler, W.R. (1993). Reduction of fertility and alteration of uterine pH in heifers fed ruminally degradable protein. Journal ofAnimal Science 71, 694-701. Fulkerson, W.J., Slack, K., and Hough, G (1997). Nutrients in ryegrass (Lolium spp), white clover (Trifolium repens) pastures in relation to season and stage of regrowth in a subtropical environment Australian Journal of Agricultural Research (in press). Fulkerson, W.J., and Slack, K. (1995). Leafnumber as a criterion for determining defoliation time for Lolium perenne II. Effect of defoliation frequency and height. Grass and Forage Science 50, M-20. Fulkerson, W.J., Slack, K., and Lowe, K.J. (1994). Variation in the response of lolium genotypes to defoliation. Australian Journal ofAgricultural Research. 45, 1309-1317. Grainger, C., Davey, A. W.F. and Holmes, C.W. (1985). Performance of Friesian Cows with high and low breeding indexes. Stall feeding and grazing experiments and performance during the whole lactation. Animal Production. 40,379-388. de Jong, A. (1986). The role of metabolites and hormone as feed backs in the control of food intake. In: Ruminants in control of digestion and metabolism in ruminants pp. 459-478 (Eds. L.P. Milligan, W.L. Grovum and A. Dobson). Prentice - Hall: Englewood Cliffs, N.J. Demarquilly, C. and Jarrige, R. (1973). The comparative nutritive value of grasses and legumes. Proceedings of the 5th. General meeting of the European Grassland Halleron, R. (1994). Maximising your herds genetic potential. Dairy Farmer December, p. 82. Limitations to milk production for pasture 165 Henzell, R.B. (1979). Efficiency of growth by ruminal bacteria. Federation Proceedings 38,2707-27 12. Hodgson, J. (1977). Factors limiting herbage intake by the grazing animal. Proceeding of the International meeting on Animal Production porn Temperate Grasslands. Dublin, June 1977. pp. 70-75. Hoover, W.H., and Stokes, S.R. (1991). Balancing carbohydrates and proteins for optimum rumen microbial yield. Journal of Dairy Science 74,363O. Howard, H. J., Aalseth, E.P., Adams, GD., Bush, L. J., McNew, R.W., and Davison, L.J. (1987). Influence of dietary protein on reproductive performance of dairy cows. Journal of Dairy Science 70,1563-l 57 1. Meijs, J.A.C. (198 1). Herbage Intake by Grazing Dairy Cows. Agricultural research report (V.L.0) No. 909. Pudoc: Wageningen. Oldham, J.D. (198 1). Amino acid requirements for lactation in high-yielding dairy cows. In: Recent Advances in AnimalNutrition - 1980, pp. 33-65 (Ed. I.S. Haresign). London: Butterworths. Peyraud, J.L. Cameron, E.A., Wade, M.H. and Lemaire, G (1996). The effect of daily herbage allowance, herbage mass and animal factors upon herbage intake by grazing dairy cows. Annales de Zootechnic 45,201217. Reeves, M., Fulkerson, W-J., and Kellaway, R.C. (1996). A comparison of three techniques to determine the herbage intake of dairy cows grazing kikuyu (Pennisetum clandestinum) pasture. Australian Journal of kperimental Agriculture 36,23-30. Rogers, GL., Porter, R.H.D. and Robinson, I. (1982). Comparison of ryegrass and white clover for milk production. In `Dairy Production corn Pasture'. Proceedings of Conference of the New Zealand and Australian Societies ofAnimal Production, Hamilton, New Zealand. Sinclair, L.A., Garnsworthy, P.C., Newbold, J.R., and Buttery, P.J. (1993). Effect of synchronizing the rate of dietary energy and nitrogen release on rumen fermentation and microbial protein synthesis in sheep. Journal ofAgricultural Science, Cambridge 120,251-263. van Soest, P. J. (1982). Nutritional Ecology of Ruminants. pp. 374. 0. & B. books, Inc.: Corvallis, OR. van Vuuren, A.M. (1993). Digestion and Nitrogen Metabolism of Grass-fed Dairy Cows. Ph.D. Thesis 134 pp. Wageningen: Netherlands. Virtanen, A.I. (1966). Milk production of cows on protein fkee feeds. Science 153,1603. Walker, J. W. and Heitschmidt, R.K. (1989). Some effects of a rotational grazing treatment on cattle grazing behaviour. Journal of Range Management 42,337339. Wilson, GF. (1966). Ryegrass varieties in relation to dairy cattle performances. II: The influence of ryegrass varieties on intake, digestibility and on some characteristics of rumen fermentation. New Zealand Journal of Agricultural Research 9,1053-l 063.