GrazeIn: a model of herbage intake and milk production for grazing dairy cows. 3. Simulations and external validation of the model

GrazeIn is a model for predicting herbage intake and milk production of grazing dairy cows. The objectives of this paper are to test its robustness according to a planned arrangement of grazing and feeding scenarios using a simulation procedure, and to investigate the precision of the predictions from an external validation procedure with independent data. Simulations show that the predicted effects of herbage allowance, herbage mass, herbage digestibility, concentrate supplementation, forage supplementation and daily time at pasture are consistent with current knowledge. The external validation of GrazeIn is investigated from a large dataset of twenty experiments representing 206 grazing herds, from five research centres within Western Europe. On average, mean actual and predicted values are 14·4 and 14·2 kg DM d^?1 for herbage intake and 22·7 and 24·7 kg d^?1 for milk production, respectively. The overall precision of the predictions, estimated by the mean prediction error, are 16% (i.e. 2·3 kg DM d^?1) and 14% (i.e. 3·1 kg d^?1) for herbage intake and milk production, respectively. It is concluded that the GrazeIn model is able to predict variations in herbage intake and milk production of grazing dairy cows in a realistic manner over a wide range of grazing management practices, rendering it suitable as a basis for decision support systems.     

GrazeIn: a model of herbage intake and milk production for grazing dairy cows. 1. Prediction of intake capacity,voluntary intake and milk production during lactation

The prediction of both food intake and milk production constitutes a major issue in ruminants. This article presents a model predicting voluntary dry matter intake and milk production by lactating cows fed indoors. This model, with an extension to predict herbage intake at grazing presented in a second article, is used in the Grazemore decision support system. The model is largely based on the INRA fill unit system, consisting of predicting separately the intake capacity of the cows and the fill value (ingestibility) of each feed. The intake capacity model considers potential milk production as a key component of voluntary feed intake. This potential milk production represents the energy requirement of the mammary gland, adjusted by protein supply when the protein availability is limiting. Actual milk production is predicted from the potential milk production and from the nutritional status of the cow. The law of response of milk production is a function of the difference between energy demand and actual energy intake, modulated by protein intake level. The simulation of experimental data from different feeding trials illustrates the performance of the model. This new model enables dynamic simulations of intake and milk production sensitive to feeding management during the whole lactation period.     

Evaluation of the GrazeIn model of grass dry‐matter intake and milk production prediction for dairy cows in temperate grass‐based production systems. 1–Sward characteristics and grazing management factors

This study evaluated the prediction accuracy of grass dry‐matter intake (GDMI) and milk yield predicted by the model GrazeIn using a database representing 522 grazing herds. The GrazeIn input variables under consideration were fill value (FV), grass energy content [Unité Fourragère Lait (UFL)], grass protein value [true protein absorbable in the small intestine when rumen fermen energy is limiting microbial protein synthesis in the rumen (PDIE)], pre‐grazing herbage mass (PGHM), daily herbage allowance (DHA) and concentrate supplementation. GrazeIn was evaluated using the relative prediction error (RPE). The mean actual GDMI and milk yields of grazing herds in the database ranged from 9·9–22·0 kg DM per cow d^?1 and 8·9–41·8 kg per cow d^?1, respectively. The accuracy of predictions for the total database estimated by RPE was 12·2 and 12·8% for GDMI and milk yield, respectively. The mean bias (predicted minus actual) for GDMI was ?0·3 kg DM per cow d^?1 and for milk yield was +0·9 kg per cow d^?1. GrazeIn predicted GDMI with a level of error <13·4% RPE for spring, summer and autumn. GrazeIn predicted milk yield in autumn (RPE = 17·6%) with a larger error in comparison with spring (RPE = 10·4%) and summer (RPE = 11·0%). Future studies should focus on the adaptation of GrazeIn to correct and improve the prediction of milk yield in autumn.     

Evaluation of the GrazeIn model of grass dry‐matter intake and milk production prediction for dairy cows in temperate grass‐based production systems. 2 – Animal characteristics

This study evaluated the prediction accuracy of grass dry‐matter intake (GDMI) and milk yield predicted by the GrazeIn model using a large database representing 8787 per cow GDMI measurements. In this study, the animal input variables (age, parity, week of lactation, potential peak milk yield, milk fat content, milk protein content, bodyweight, body condition score (BCS), week of conception, BCS at calving and calf birth weight) were investigated. The mean actual GDMI of the database was 15·9 kg DM per cow d^?1 and GrazeIn predicted a mean GDMI for the database of 15·5 kg DM per cow d^?1. The mean bias was ?0·4 kg DM per cow d^?1. GrazeIn predicted GDMI for the total database with an RPE of 15·5% at cow level. The mean actual daily milk yield of the database was 21·3 kg per cow d^?1 and GrazeIn predicted a daily milk yield for the database of 22·2 kg per cow d^?1. The mean bias was +0·9 kg per cow d^?1. GrazeIn predicted milk yield for the total database with an RPE of 16·7% at cow level. From the evaluation, GrazeIn predicted milk yield of all cows in late lactation with a larger level of error than in early and mid‐lactation. This error appears to be due to the persistency of the lactation curve used by the model, which results in a higher predicted milk yield in late lactation compared with the actual milk yield.     

Incorporating grazing behaviour measurements in models to predict herbage intake by grazing dairy cows

Models to predict herbage intake were constructed using 168 dairy cow records from three grazing experiments. Variables included fell into three categories: animal state, sward state and animal behaviour. Linear regression models of varying complexity were obtained by removing variables from the best fitting model to reflect progressive lack of information availability on farms. Thus, behavioural variables were removed first, followed by sward surface height and milk fat concentration. Models were subject to outlier analysis and collinearity tests. Equivalent models were constructed using ridge regression to minimize collinearity problems. They were tested using 20 Holstein–Friesian dairy cows continuously stocked on a perennial ryegrass sward. A `best practice' treatment [7 cm sward surface height (SSH), 6 kg day⁻¹ concentrate (C)] was used together with treatments of SSH5/C6, SSH7/C8, SSH7/C0 and SSH9/C6. The best model accounted for 0.37 of the variance in the estimation data and contained the following variables: concentrate intake, milk yield, milk fat concentration, days in milk, sward surface height and chewing rate while ruminating. Model performance against test data was generally poor. This was mainly because of consistent underprediction of herbage intake, caused in part by the higher average herbage intakes in the test data compared with the estimation data.     

Quantifying the influence of sward height, concentrate level and initial milk yield on the milk production and grazing behaviour of continuously stocked dairy cows

Two factorial design experiments were carried out in the spring of 1994 and 1995, each of 6 weeks, to quantify the effects of sward height (SH), concentrate level (CL) and initial milk yield (IMY) on milk production and grazing behaviour of continuously stocked dairy cows. In Experiment 1, forty‐five Holstein Friesian cows were in five groups with initial milk yields of 16·9, 21·1, 28·0, 31·5 and 35·5 kg d^–1, grazed sward heights were 3–5, 5–7 and 7–9 cm (LSH, MSH and HSH respectively), and concentrates were fed at rates of 0, 3 and 6 kg d^–1. In Experiment 2, 48 cows were in two groups with IMY of 21·3 and 35·5 kg d^–1, grazed sward heights were 3–5 and 7–9 cm (LSH and HSH), and concentrates were fed at 0 and 6 kg d^–1 and ad libitum. Multiple regression models were used to quantify the effects of the three variables on milk yield persistency (MYP), estimated herbage dry‐matter (DM) intake (HDMI), grazing time (GT) and rate of DM intake (RI). The partial regression coefficients showed that increased SH led to increased MYP (Experiment 1 P < 0·001, Experiment 2 P < 0·05), increased HDMI (P < 0·01, P < 0·01), increased GT (P < 0·001, P < 0·05) and increased RI (P < 0·001, P < 0·05). Increasing CL led to increased MYP (NS, P < 0·001), decreased HDMI (P < 0·001, P < 0·001), decreased GT (NS, P < 0·001) and decreased RI (P < 0·001, P < 0·001). Higher IMY level of cows decreased MYP (P < 0·001, P < 0·001), increased HDMI (P < 0·001, P < 0·001), increased GT (P < 0·001, P < 0·05) and increased RI (P < 0·05, P < 0·01). The models were highly significant (P < 0·001), and accounted for 0·48–0·87 of the total variance. The partial regression coefficients quantified the extent to which GT and RI by cows respond positively to higher IMY, and negatively to increased CL, but respond differently (GT declines in response to a higher RI) with increasing SH.     

The use of herbage nitrogen status to optimize herbage composition and intake and to minimize nitrogen excretion: an assessment of grazing management flexibility for dairy cows

A framework for managing rotationally grazed pastures for dairy cattle which enables the cows’ energy and protein requirements to be met while simultaneously limiting the amount of N excreted in order to reduce N losses is described. The first objective is achieved by ensuring that lamina mass and the N concentration of herbage do not limit herbage intake or feeding value. The second objective is achieved by limiting N fertilizer supply or increasing the interval between defoliations to reduce the N concentration of herbage. Lower and upper thresholds for the N concentration of herbage and lamina mass were estimated from published data. The method is illustrated using two vegetative regrowths (beginning and end of spring) in a cutting experiment with two fertilizer treatments, 0 or 120 kg N ha^?1 (?N and +N), and early or late cutting. Decreasing N supply led to a reduction in grazing management flexibility, i.e. the defoliation interval ranges which were compatible with the required sward characteristics (minimum lamina mass and N concentration of lamina) for herbage intake and to meet the protein requirements of dairy cows. Aiming for the upper threshold N concentration of herbage increased the minimum interval between defoliations only for the +N treatments. Nevertheless, grazing management flexibility remained the highest for the +N treatments.     

A model describing the utilization of herbage by cattle in a rotational grazing system

A model that describes the utilization of herbage by cattle in a rotational grazing system is presented. The model considers swards as being structured into two phases: a long phase associated with old dung patches, and a short phase. Both phases are treated as consisting of a series of horizontal strata corresponding in depth to a bite depth. The sward is divided into feeding stations consisting of either the long or the short phase. In each, only the surface stratum is available for grazing at each time step. At any time step, the individuals of a herd of cattle, distributed at random, encounter the entire range of strata. The rate of intake of each member of the herd depends on the intake properties of the stratum that it has encountered. The number of cattle that encounter each stratum type is variable so that the mean intake per member of the herd is the weighted mean. The core feature of the model is the simulation of the change over time in the frequency distribution of exposed stratum types and the distribution of grazing across this range of strata. The members of the herd are assumed to select a feeding station based on preference for leafiness of the encountered stata and the phase. The decision to graze or not is based on the comparison between the current vs. the previous feeding station. Model parameter values were based on published data. The proportion of leaf and bulk density of a phase or the strata were determined from an analysis of a sample of sward profiles. Using bite dimension, bite weight, biting rate, search time, feeding station area and selective behaviour, it was possible to simulate sward depletion that is very similar to the observed data from grazed paddocks in experiments in south‐east Ireland. The model of herbage utilization adequately described the changes in intake and sward structure during grazing and it was concluded that it was suitable for use as part of a simulation of a grazing system.     

Effect of altering the grazing interval on growth and utilization of grass herbage and performance of dairy cows under rotational grazing

The effects of short grazing intervals in the early part of the grazing season on the growth and utilization of grass herbage, and the performance of grazing dairy cows, in a rotational grazing system were examined. Seventy-six cows were allocated to two grazing treatments: a normal rotation treatment (20-d rotations for the first 60 d) and a short rotation treatment (12-, 8-, 8-, 8-, 12- and 12-day rotations). Thereafter, both treatments had the same grazing interval and over the season as a whole both treatments received the same amount of nitrogen fertilizer and were stocked at the same rate. The short rotation treatment significantly reduced pre- and post-grazing sward heights and pre-grazing herbage mass in May and June. Total herbage production was significantly lower on the short than the normal rotation treatment as a result of a significant reduction in the growth rate of herbage in May and June. The short rotation treatment had a significantly lower milk output per cow. Grazing shorter swards, as a result of shorter rotations, significantly reduced herbage intake, reflecting reductions in intake per bite, grazing time and total bites per day. Treatment had no significant effect on herbage quality or pre- and post-grazing sward height in August and September, despite the increased grazing severity in May and June with the short rotations. The severity of rotation length in this instance had a detrimental impact on animal performance, whereas a more modest reduction in grazing interval may control herbage production, without reducing animal performance.     

The effects of offering a range of forage and concentrate supplements on milk production and dry matter intake of grazing dairy cows

An experiment was undertaken to examine the effect of supplement type on herbage intake, total dry matter (DM) intake, animal performance and nitrogen utilization with grazing dairy cows. Twenty‐four spring‐calving dairy cows were allocated to one of six treatments in a partially balanced changeover design with five periods of four weeks. The six treatments were no supplement (NONE), or supplementation with either grass silage (GS), whole‐crop wheat silage (WS), maize silage (MS), rapidly degradable concentrate (RC) or slowly degradable concentrate (SC). Cows were rotationally grazed with a mean herbage allowance of 20·5 kg DM per cow per day, measured above 4 cm. Forage supplements were offered for approximately 2 h immediately after each morning milking, with cows on NONE, RC and SC treatments returning to the grazing paddock immediately after milking. Cows on treatment MS had a significantly higher supplement DM intake than the other treatments but a significantly lower grass DM intake than the other treatments, resulting in no significant difference in total DM intake when compared with cows on treatments WS, RC and SC. Concentrate type had no significant effect on herbage intake, milk yield, milk composition or yield of milk components. The yield of milk fat and milk protein was significantly higher on treatments MS, RC and SC compared with treatments NONE, GS and WS. The results indicate that despite a relatively high substitution rate, maize silage can be a useful supplement for the grazing dairy cow.     

The effect of initial spring grazing date and subsequent stocking rate on the grazing management, grass dry matter intake and milk production of dairy cows in summer

The objective of this study was to investigate the effects of an early (February; F) or delayed (April; A) primary spring grazing date and two stocking rates, high (H) and medium (M), on the grazing management, dry matter (DM) intake of grass herbage and milk production of spring‐calving dairy cows grazing a perennial ryegrass sward in the subsequent summer. Sixty‐four Holstein‐Friesian dairy cows (mean of 58 d in milk) were assigned to one of four grazing treatments (n = 16) which were imposed from 12 April to 3 July 2004. Cows on the early spring‐grazing treatment were grazed at 5·5 cows ha^?1 (treatment FH) and 4·5 cows ha^?1 (treatment FM) while cows on the late‐grazing treatment were grazed at 6·4 cows ha^?1 (treatment AH) and 5·5 cows ha^?1 (treatment AM). The organic matter digestibility and crude protein concentration of the grass herbage were higher on the early‐grazing treatment than on the late‐grazing treatment. The cows on the FM treatment had significantly (P < 0·001) higher milk (24·5 kg), solids‐corrected milk (22·5 kg), fat (P < 0·01, 918 g) and protein (831 g) yields than the other three treatments. Cows on the FM treatment had a higher (P < 0·001) DM intake of grass herbage by 2·3 kg DM per cow per day than cows on the AH treatment, which had a DM intake significantly lower than all other treatments (15·2 kg DM per cow per day). The results of the present study showed that grazing in early spring has a positive effect on herbage quality in subsequent grazing rotations. The study also concluded that early spring‐grazed swards stocked at a medium stocking rate (4·5 cows ha^?1; FM) resulted in the highest DM intake of grass herbage and milk production.     

The influence of strain of Holstein-Friesian dairy cow and pasture-based feeding system on grazing behaviour, intake and milk production

A comparative study of grazing behaviour, herbage intake and milk production of three strains of Holstein‐Friesian dairy cow was conducted using three grass‐based feeding systems over two years. The three strains of Holstein‐Friesian cows were: high production North American (HP), high durability North American (HD) and New Zealand (NZ). The three grass‐based feeding systems were: high grass allowance (MP), high concentrate (HC) and high stocking rate (HS). In each year seventy‐two pluriparous cows, divided equally between strains of Holstein‐Friesian and feeding systems were used. Strain of Holstein‐Friesian cow and feeding system had significant effects on grazing behaviour, dry matter (DM) intake and milk production. The NZ strain had the longest grazing time while the HD strain had the shortest. The grazing time of cows in the HC system was shorter than those in both the HS and MP systems. There was a significant strain of Holstein‐Friesian cow by feeding system interaction for DM intake of grass herbage and milk production. The NZ strain had the highest substitution rate with the HP strain having the lowest. Hence, response in milk production to concentrate was much greater with the HP than the NZ strain. Reduction in milk yield as a consequence of a higher stocking rate (MP vs. HS system) was, however, greater for the HP and HD strains compared with the NZ strain. The results suggest that differences in grazing behaviour are important in influencing DM intake and milk production.     

Continuous and rotational grazing of dairy cows – the interactions of grazing system with level of milk yield, sward height and concentrate level

An experiment was conducted to test the hypothesis that for cows with high levels of milk yield, rotational grazing produces higher milk yields than continuous grazing. The comparison of grazing systems was made at two levels of milk yield (initially 20·3 and 32·5 kg d^?1), and interactions with sward height and concentrate level were also examined. The study used 48 multiparous Holstein Friesian cows over a period of 62 d. Mean milk yield, its persistency and composition, live weight, body condition score and liveweight gain were not significantly affected by grazing system at either level of milk yield. There were no significant interactions between grazing system and sward height or concentrate level for any milk production measurement. Mean estimated herbage and total dry matter (DM) intake (P < 0·01), grazing time (P < 0·05) and ruminating time (P < 0·01) were significantly greater on the continuous grazing system. The cows in the higher milk yield group and those grazed at the higher sward height had a significantly (P < 0·05) higher estimated daily herbage DM intake and rate of herbage intake on the continuous grazing system than those on the rotational grazing system. There was no evidence to support the hypothesis that rotational grazing systems support higher levels of milk production than continuous grazing for cows of high milk yield. The shorter grazing time on the rotational grazing system indicated that cows may anticipate the timing of the daily movement of the electric fence, and this reduces their time spent grazing residual herbage.     

Effects of perennial ryegrass (Lolium perenne L.) cultivars on herbage production, nutritional quality and herbage intake of grazing dairy cows

Four perennial ryegrass (Lolium perenne L.) cultivars were compared for differences in herbage production, nutritive value and herbage intake of dry matter (DM) during the summers of 2002 and 2003. Two paddocks were sown with pure stands of four cultivars in a randomized block design with three replicates. Each plot was subdivided into fourteen subplots (22 m × 6 m) which were grazed by one cow during 24 h. Twelve lactating dairy cows were assigned to one cultivar for a period of 2 weeks in a 4 × 4 Latin square experimental design; the experiment lasted 8 weeks in each year. Sward structure (sward surface height, DM yield, green leaf mass, bulk density and tiller density) and morphological characteristics were measured. The ash, neutral‐detergent fibre, acid‐detergent lignin, crude protein and water‐soluble carbohydrate concentrations, and in vitro digestibility of the herbage were measured. The sward was also examined for infestation by crown rust (Puccinia coronata f. sp. lolii). Herbage intake of dairy cows was estimated using the n‐alkane technique. Cultivar differences for all sward structural characteristics were found except for bulk density and tiller density in 2003. Cultivars differed for proportions of pseudostem, stem (in 2003 only) and dead material. The chemical composition of the herbage was different among cultivars, with the water‐soluble carbohydrate concentration showing large variation (>0·35). Cultivars differed in susceptibility to crown rust. Herbage intake differed among cultivars in 2002 (>2 kg DM) but not in 2003. Herbage intake was positively associated with sward height, DM yield and green leaf mass. Canopy morphology did not affect herbage intake. Crown rust affected herbage intake negatively. It was concluded that options for breeders to select for higher intake were limited. High‐yielding cultivars and cultivars highly resistant to crown rust were positively related with a high herbage intake.     

Concentrate supplementation of grazing dairy cows. 2. Effect of concentrate composition on herbage intake and milk production

The effect of feeding either traditional concentrates containing starch or high quality fibrous concentrates on the performance of grazing dairy cows was examined in a trial in which cows were given concentrates with either 350 g starch and sugars (kg dry matter (DM))^-1 (high-starch) or 100 g starch and sugars (kg DM)^-1 (high-fibre). The swards used consisted predominantly of perennial ryegrass and were usually aftermaths following cutting. Each area was grazed for 3 or 4 d at each grazing and a two-machine sward-cutting technique was used for estimating herbage intake.
The effect of concentrate composition on the herbage intake of grazing cows at a high daily herbage allowance of 28 kg OM above 4 cm cutting height was investigated in 1983 and 1984. With 54 kg OM d^-1 of high-starch concentrates the mean herbage intake was 11·5 kg OM d^-1 per cow while cows fed 5.3 kg d^-1 of high-fibre concentrates consumed on average 12–6 kg OM d^-1. The mean substitution rate of herbage by concentrates was reduced from 0·45 kg herbage OM (kg concentrate OM)^-1 with the high-starch concentrate to 0·21 with the high-fibre concentrates.
The effect of the treatments on milk production was studied in 1984. The cows consumed 5·5 kg OM d^-1 as concentrates and grazed at a lower herbage allowance of 19 kg OM above 4 cm cutting height. With high-fibre concentrates milk production and 4% fat-corrected milk production were 13 and 1·8 kg d^-1, respectively, higher than with the high-starch treatment. The daily live weight gain with the high-starch concentrates was 0·17 kg per cow more than with the high-fibre concentrates.     

Pattern of herbage intake rate and bite dimensions of rotationally grazed dairy cows as sward height declines

To allow improved prediction of daily herbage intake of dairy cows in rotational grazing systems, intake behaviour was assessed throughout the day in 24‐h paddocks. Herbage intake in 16 lactating Holstein–Friesian cows was assessed using the short‐term (1‐h) weight gain method at four predetermined natural meal times throughout the day (early morning, T1; late morning, T2; mid‐afternoon, T3; and early evening, T4). The study comprised two 4‐day experiments, each with a cross‐over design of four blocks. In both experiments, cows grazed a 24‐h paddock daily, and the effect of the immediately previous grazing experience on intake behaviour was investigated throughout the day, taking account of daily fluctuations in the short‐term physiological condition of the cows. Experiment 1 was carried out to investigate overall grazing behaviour during meals as a sward is progressively depleted during the day, with intake being assessed within the paddock and, hence, on a depleted sward. Experiment 2 similarly investigated the effect of sward depletion and physiological condition throughout the day on intake, but cows were removed to fresh, undefoliated swards during intake measurement periods; thus, intake rate was not influenced by differences in sward condition. Intake behaviour from both experiments was compared to establish the effect on herbage intake of changes in sward state and non‐sward factors. In Experiment 1, sward surface height, available herbage mass, proportion of leaf and green leaf mass declined as the day progressed. Bite mass declined with sward depletion, and mean intake rate was 1·64 kg dry matter (DM) h^–1, which was significantly lower at T3 (P < 0·01) than during other meals. In Experiment 2, plot sward conditions did not change throughout the day, and intake behaviour also remained constant, with a mean intake rate of 2·11 kg DM h^–1. Mean bite depth as a proportion of pregrazing extended tiller height was constant throughout the day (mean 0·32). The results show that, although cows grazed throughout the day on progressively depleted swards, indicative of rotationally grazed paddocks (Experiment 1), bite mass declined linearly and intake behaviour was variable. However, where intake was assessed on high‐quality, undefoliated swards (Experiment 2), intake behaviour was similar regardless of the time of day and the immediately previous experience. There was some indication of an interaction between the effects of the sward and the physiological condition of the animal on herbage intake.     

Some challenges and opportunities for grazing dairy cows on temperate pastures

Grazing plays an important role in milk production in most regions of the world. In this review, some challenges to the grazing cow are discussed together with opportunities for future improvement. We focus on daily feed intake, efficiency of pasture utilization, output of milk per head, environmental impact of grazing and the nutritional quality to humans of milk produced from dairy cows in contrasting production systems. Challenges are discussed in the context of a trend towards increased size of individual herds and include limited and variable levels of daily herbage consumption, lower levels of milk output per cow, excessive excretion of nitrogenous compounds and requirements for minimal periods of grazing regardless of production system. A major challenge is to engage more farmers in making appropriate adjustments to their grazing management. In relation to product quality, the main challenge is to demonstrate enhanced nutritional/processing benefits of milk from grazed cows. Opportunities include more accurate diet formulations, supplementation of grazed pasture to match macro- and micronutrient supply with animal requirement and plant breeding. The application of robotics and artificial intelligence to pasture management will assist in matching daily supply to animal requirement. Wider consumer recognition of the perceived enhanced nutritional value of milk from grazed cows, together with greater appreciation of the animal health, welfare and behavioural benefits of grazing should contribute to the future sustainability of demand for milk from dairy cows on pasture.     

Effect of replacing time available for grazing with time available for eating maize silage and soyabean meal on milk yield and feeding behaviour in dairy cows

Low rates of herbage dry matter (DM) intake impose limits on total daily DM intake in grazing dairy cows. The objective of this study was to increase total daily DM intake and milk production by restricting daily time available for grazing (TAG) and replacing it with time available for eating a maize silage/soyabean meal (TAMS) diet indoors. The treatments (TAG + TAMS) were 20 + 0, 19 + 1, 10 + 10 and 5 + 15 h. Measurements were made of milk production, intake and feeding behaviour. The interactions of TAG + TAMS treatments with sward height (SH) and concentrate level (CL) were also examined. Two experiments, each lasting 42 days, were carried out in spring ( Experiment 1 ) and autumn ( Experiment 2 ) using forty‐eight and twenty‐four Holstein‐Friesian cows respectively. Treatments were arranged in a factorial design with TAG + TAMS treatments, SH ( Experiment 1 only) and CL as the independent variables and a TAG + TAMS of 20 h. Reducing TAG and increasing TAMS significantly reduced estimated herbage DM intake and significantly increased maize silage/soyabean meal intake in both experiments, but there were no significant main effects of TAG + TAMS treatments on milk yield (mean, 27·4 and 25·5 kg d^?1 for Experiments 1 and 2 respectively), and yield of milk constituents. Increasing SH ( Experiment 1 ) and CL ( Experiments 1 and 2 ) significantly increased milk yield. In Experiment 1 , there was a significant interaction between TAG + TAMS treatments and SH with the taller sward height of 8–10 cm and the 20 + 0 treatment having the highest milk yield (29·7 kg d^?1) and the 5 + 15 treatment the lowest (27·2 kg d^?1), whereas at the lower sward height of 4–6 cm, milk yield was lowest on the 20 + 0 treatment (25·5 kg d^?1) with the other three treatments being higher (mean, 26·9 kg d^?1). Replacing TAG with TAMS significantly increased liveweight gain in Experiment 1 but not in Experiment 2 . Estimated rates of intake of herbage were lower in the autumn experiment ( Experiment 2 , 9·6 g DM min ^?1) than in the spring experiment ( Experiment 1 , 29·4 g DM min ^?1) but rates of intake of maize silage were higher in the autumn (112·4 g DM min^?1) than in the spring (72·5 g DM min^?1). In conclusion, in spring the response to replacing TAG with TAMS was dependent on sward conditions with the highest milk fat plus protein yield being on the 20 + 0 treatment at the high sward height and on the 19 + 1 treatment at the low sward height. The high liveweight gain of the 5 + 15 treatment could be an important means of restoring body condition in grazing lactating cows. In autumn, intakes of herbage were low in spite of its high estimated nutritive value with all treatments having a similar level of performance.     

Production and behavioural responses of high- and low-yielding dairy cows to different periods of access to grazing or to a maize silage and soyabean meal diet fed indoors

The study examined whether high‐yielding cows grazing pasture respond differently from low‐yielding cows in milk production and feeding behaviour, to increasing the time made available for eating a maize silage and soyabean meal (TAMS) diet indoors and reducing the time available for grazing (TAG). Two experiments, each lasting 42 d, were carried out in spring (Experiment 1) and autumn (Experiment 2) using Holstein‐Friesian cows at two different levels of milk yield (MYL). Milk production and feeding behaviour were examined for TAG + TAMS systems of 19 h, TAG plus 1 h TAMS (19 + 1), and 5 h TAG plus 15 h TAMS (5 + 15). There were two levels of concentrate (0 and 6 kg d^?1), and in the spring experiment two sward heights (4–6 and 8–10 cm) were also studied. Milk yield, persistency of milk yield, liveweight change and estimated total DM intake were significantly higher on the 5 + 15 than on the 19 + 1 grazing system in Experiment 1 but not in Experiment 2. There were no significant interactions of TAG + TAMS treatment with MYL for any production or behavioural measurements except for maize silage feeding time, where high MYL cows spent a significantly greater time eating maize silage than low MYL cows on the 5 + 15 treatment but not on the 19 + 1 treatment. It can be concluded that high‐ and low‐yielding cows respond similarly in milk production and feeding behaviour to different combinations of TAG and TAMS. In autumn, estimated daily intakes of herbage were lower on both grazing treatments relative to spring, resulting from lower rates of herbage intake with no compensatory increase in grazing time. In contrast, rates of intake of maize silage were higher in autumn especially on the 19 + 1 system. These results may imply a change of preference from herbage to maize silage between spring and autumn.     

A simulation model of a dairy forage system to evaluate feeding management strategies with spring rotational grazing

Abstract To counteract a decrease in the availability of grazing for feeding dairy cows in France, a simulation model is proposed in the paper, which combines decision and agronomic submodels to study forage system management strategies compatible with spring grazing use. Nine strategies were tested with the model using a sequence of 16 climatic years. Three of these strategies come from a survey in south‐west France and six others were designed with research scientists or farm advisors. The strategies differ in the duration of maize silage feeding, the area dedicated to maize silage and the area dedicated to grass silage. Results from simulation models show that the consequence of a large and constant maize silage area is a high maize silage overstock if there is an early turnout or a high grass silage overstock if turnout is late. The consequence of a low maize area is a high grazing duration combined in some years with feeding shortages. Strategies that have no feeding shortage and a low level of maize and grass silage overstock have a high grazing duration and have no constant maize or grass silage areas. The solution proposed to avoid climatic risk, and its consequences on feeding, is based on two procedures: use of reserve areas for production and allowing the production programme to be modified to take into account fresh information, especially weather records.