Partial replacement of a total mixed ration with annual ryegrass herbage: effects on dairy cow dry matter intake and milk production

ABSTRACT

Studies with dairy cows receiving mixed ration and grazing annual pastures, which have low tiller density, are scarce. This study examined the effect of partial replacement of a total mixed ration (TMR) with annual temperate pasture. Treatments were ad libitum TMR (TMR₁₀₀), 75% ad libitum TMR + ryegrass (Lolium multiflorum ‘Maximus’) (pTMR₇₅), and 50% ad libitum TMR + ryegrass (pTMR₅₀). Twelve multiparous Holstein and F1 Jersey × Holstein cows were divided into six groups (experimental units), taking account of milk production (26.6 ± 4.55 kg/day), days-in-milk (129 ± 50.8) and body weight (546 ± 30.6 kg). Treatments were compared using two 3 × 3 Latin square arangements, comprising three 21-day periods (measurements during final 5 days). Herbage DM intake increased from 4.8 kg/day with pTMR₇₅ to 6.7 kg/day with pTMR₅₀. Total DM intake decreased from 19.4 kg/day (TMR₁₀₀), to 18.1 and 15.8 kg/day (pTMR₇₅ and pTMR₅₀, respectively). Milk production, energy corrected milk production (ECM), and milk fat content were similar between treatments, averaging 25.6, 28.4 kg/day and 44.6 g/kg, respectively. Ryegrass pastures were able to replace up to 50% of TMR offered to mid lactation dairy cows with no adverse effects on ECM production.

KEYWORDS:

• Intake
• milk production
• milk composition
• Lolium multiflorum

1. Introduction

Full-time grazing systems are normally unable to supply all of the energy requirements of lactating dairy cows (Kolver and Muller 1998), and often do not provide a constant supply of herbage throughout the year (Wilkinson et al. 2020). As a result, systems involving full time housing, in which cows are often offered a total mixed ration (TMR), are becoming increasingly common. However, giving housed cows access to grazing for part of the day may improve animal welfare (Arnott et al. 2017) and reduce feeding costs (White et al. 2002). Thus the adoption of ‘mixed systems’ (‘part housing-part grazing’) may provide a tool to help maintain individual cow total dry matter (DM) intakes and milk production, and overall stocking density (Phillips 1988; Bargo et al. 2003; Wales et al. 2013).

Total DM intake and milk production of dairy cows in mixed systems are influenced by herbage characteristics, including herbage quality and herbage allowance. For example, when cows receiving a partial TMR (pTMR) were given access to poorer quality herbage later in the grazing season (Soriano et al. 2001), or to a tropical herbage (Civiero et al. 2021), total DM intake and milk production were both reduced. However, when cows receiving pTMR grazed high quality herbage, without any limitations associated with sward structure and herbage allowance, total DM intake and milk production did not differ compared with cows receiving ad libitum TMR as the sole feed (Dall-Orsoletta et al. 2016).

Previous studies involving dairy cows grazing annual ryegrass (Lolium multiflorum Lam.) swards in a subtropical region have demonstrated that even when herbage allowance and herbage quality were high, herbage DM intake was lower than expected due to low pre-grazing herbage mass (Miguel et al. 2014, 2019). The low herbage mass associated with annual ryegrass (in comparison to perennial ryegrass species) has been attributed to its low tiller density, especially during the first grazing cycles (Miguel et al. 2014). Thus, DM intake of annual ryegrass swards appears to be predominantly limited by low herbage mass, regardless of herbage allowance. Nevertheless, to our best knowledge, the effect of including annual ryegrass pasture in dairy cow diets which are predominantly comprised of a TMR has not been studied previously.

In an earlier study in which cows from the same herd grazed a topical pasture, total DMI and milk production decreased by less than 10% when TMR intake was restricted by up to 50% of ad libitum (Civiero et al. 2021). Therefore we hypothesized that compared to cows offered TMR as the sole feed, cows given access to good quality annual ryegrass pastures could have their intakes of TMR restricted by up to 50% of ad libitum, while maintaining total DM intake and milk production. Thus the aim of this study was to assess the effect of including herbage in the diets of TMR-fed dairy cows, on total DM intake and milk production.

2. Material and methods

The Ethics Committee of University of Santa Catarina State approved all procedures, with protocol number 4373090816.

2.1. Treatments, experimental design, and animals

The experiment was performed in in Lages, SC, Brazil (50.18°W, 27.47°S; 920 m above sea level) from June 19 to 21 August 2019. Twelve multiparous dairy cows (six Holstein and six Holstein × Jersey cows) were divided in six homogeneous groups, each of two cows (the experimental unit), according to (means ± standard deviations) milk production (26.6 ± 4.55 kg/day), days-in-milk (129 ± 50.8 days) and body weight (546 ± 30.6 kg) measured during the week prior to the experiment starting. Each pair of cows were randomly placed within two 3 × 3 Latin squares, with each experimental period 21 days in length: a 16-day adaptation period and a 5-day measurement period.

Three treatments were examined, as follows: a TMR diet offered at 100% of ad libitum intake (TMR₁₀₀), TMR at 75% of ad libitum intake + access to an annual ryegrass (Lolium multiflorum cv Maximus) herbage for grazing (pTMR₇₅), and TMR at 50% of ad libitum intake + access to annual ryegrass pasture for grazing (pTMR₅₀). The TMR was balanced to meet the net energy and metabolizable protein requirements of the cows, taking into account the average body weight, milk production, milk fat content, and days in milk at the beginning of the experiment, according to INRA (2007). The ingredients, the chemical composition and the nutritive value of the TMR are presented in Table 1.

Table 1. The ingredient composition, chemical composition and nutritive value of the mixed ration offered in the study.^a

Item	Content
Ingredients (g/kg DM)
Maize silage	600
Maize meal	260
Soybean meal	140
Chemical composition, g/kg DM
Dry matter, g/kg fresh	380
Organic matter	962
Crude protein	149
Neutral detergente fiber	340
Acid detergente fiber	173
Nonfiber carbohydrate	441
Nutritive value^b
OM digestibility, g/g DM	0.768
NEL, MJ/kg DM	7.02
PDIN, g/kg DM	97.8
PDIE, g/kg DM	99.0

^a Mineral supplement composition available in the feeding area and paddocks (on natural basis): 150 g/kg calcium, 78 g/kg phosphorus; 26 g/kg sulfur, 20 g/kg magnesium, 114 g/kg sodium, 100 mg/kg cobalt, 1500 mg/kg copper, 30 mg/kg chromium, 2000mg/kg iron, 80 mg/kg iodine, 2300 mg/kg manganese, 30 mg/kg selenium, 5000 mg/kg zinc, and 780 mg/kg fluorine.

^b Estimated from chemical analysis and via equations proposed by (INRA 2007).

NE_L: Net energy for lactation. PDIN: metabolizable protein when N is limiting for microbial synthesis in the rumen. PDIE: metabolizable protein when energy is limiting for microbial synthesis in the rumen.

During a 14 day pre-experimental period each pair of cows was offered the TMR ad-libitum. The average voluntary intake of each pair during the last 5 d of this period was then used to determine the quantity of TMR offered to the pTMR₇₅ and pTMR₅₀ treatment cows during the experiment. During the experiment (when they did not have access to pasture) cows were kept in pairs in an outdoor pen measuring 3 × 20 m. The TMR or pTMR was offered via covered outdoor feeders located at one end of this area, while cows had access to natural shade at the other end of the area. Cows on pTMR₇₅ and pTMR₅₀ had access to pasture for approximately 7 h/day, between morning and afternoon milking (from 08:00 h to 15:00 h). These cows were then offered the pTMR once a day following afternoon milking (13 h/day of access to pTMR, from 16:00 h to 07:00 h the following day) at 75% and 50% respectively of pre-experimental intake for each pair of cows. Uneaten TMR and pTMR from each pair of cows was removed and weighed once daily during the morning milking. Cows had continual access to water and a mineral supplement (Bovigold®, DSM Tortuga, São Paulo, Brazil) while in the feeding area and in the paddocks. The cow walking time from the milking parlour and feeding pen to the paddocks was not more than 5 min.

2.2. Herbage and grazing management

A three hectare block of annual ryegrass, which had been reseeded in April 2019, was used in the experiment. During the experimental period the average temperature was 11.8°C and the total rainfall was 197 mm (average = 3.1 mm/day). The 10-year average temperature and rainfall during the months of the experiment were 13.9°C and 107 mm, respectively. Before the first grazing cycle (when the ryegrass was at the third-leaf- stage) the experimental area was fertilized with 50 kg nitrogen (N)/ha, in the form of urea.

The pasture area was divided into two halves (one half for each Latin square), with one third of each half assigned to pTMR₇₅ and two thirds assigned to pTMR₅₀ (4 paddocks in total, 2 cows per paddock). Paddocks were strip grazed, with target pre- and post-grazing sward height of 24 and 12 cm, respectively (a target reduction in sward height of 50%). This target level of grazing intensity was adopted to ensure that herbage intakes were not restricted by herbage availability throughout the experiment (Delagarde et al. 2011). Fresh pasture was allocated daily, with the area allocated kept the same size as that allocated during the first week of the first experimental period. No other adjustment was necessary as actual pre- and post-grazing sward heights throughout the experiment were similar to the target heights. To minimize variation in herbage quality between periods, the cows did not regraze any area during the study (ie different areas were grazed during each period). The planned grazing areas were mown (and herbage removed) 14 days before the start of each measurement period to help achieve the target pre-grazing sward height and nutritive composition of the herbage offered.

2.3. Animal measurements

Individual cow milk yields were recorded twice daily (at 07:30 h and 15:30 h) while milk samples were collected at each milking during the last five days of each period via an electronic milk meter (Waikato Milking Systems, New Zealand) approved by the International Committee for Animal Recording (ICAR). Each milk sample was analyzed for fat, protein, and milk urea nitrogen [MUN] concentrations using infrared spectrophotometry (International Dairy Federation Standard 141C:2000).

Herbage DM intake of each pair of cows during grazing was estimated as the difference between the total pre- and post-grazing biomass (Lantinga et al., 2004) on each of the last 5 days of each period. Details of pre- and post-grazing biomass measurements are presented within the ‘Feed and Herbage Measurements’ section. The DM intake of TMR and pTMR of each pair of cows were also measured daily during the final five days of each period, as the difference between the quantity of DM offered and the quantity of DM remaining uneaten the following morning.

The time that each cow in the pTMR₇₅ and pTMR₅₀ groups spent grazing was measured during the last 5 days of each period via visual observations at 5 min intervals between 08:00 h and 15:00 h. Cows had been previously conditioned with human company, and the time spent observing each individual cow was no more than 10 s, with each cow recorded as either ‘grazing’ or ‘not grazing’ (Penning and Rutter 2004). On each occasion when ‘grazing’ was observed, this was assumed to represent a ‘5 min. grazing period’. The total time spent grazing was calculated as the sum of all occasions when ‘grazing’ was observed. Herbage intake rate (g DM/min) was estimated by dividing daily herbage intake by daily grazing time. Behaviour was not recorded when cows were being milked or when cows had access to the supplement.

The net energy for lactation (NE_L) balance was estimated for each pair of cows as the difference between the NE_L supply and NE_L requirements, as described by INRA (2007). Briefly, theoretical NE_L requirements were calculated from actual BW and 4% fat-corrected milk (FCM) production during the experiment, and net energy supply was calculated from the intake of herbage, corn silage, soybean meal and their concentrations of NE_L. The NE_L of all ration ingredients was calculated by taking into account specific equations based on their chemical composition, as proposed by INRA (2007).

2.4. Feed and herbage measurements

Samples of TMR were collected twice daily from day 15 to day 20 of each period, and combined to create a composite sample per period. Samples of the uneaten feed from each pair of cows was collected during the last 5 days of each period, and composited by pair for the 5-day period for analysis. All samples were dried in an oven for 72 h at 60°C and milled (Solab SL-31, Piracicaba, Brazil) through a 1-mm screen for subsequent chemical analyses.

Pre- and post-grazing compressed sward heights were measured daily using a rising plate meter (F200 model, Farmworks, Feilding, New Zealand). To calibrate herbage mass against the plate meter reading, at the start of each experimental period, eighteen pre- and post-grazing samples were cut at ground level using scissors from eighteen 0.1 m² quadrats (nine in each of the pTMR₇₅ and pTMR₅₀ paddocks) in the ‘footprint’ of areas where the plate meter had been used to record herbage height. Individual samples were dried in an oven for 72 h at 60°C, and equations subsequently developed to predict the actual herbage mass that had been offered during the study. Owing to the relatively high R-squared value and lower residual standard deviation values, the pregrazing herbage mass was estimated according to equations generated within each period, while postgrazing herbage mass was estimated according to a general equation developed using data from all three periods (Table 2).

Table 2. Pre- and postgrazing herbage mass (HM) estimates as a function of sward height on a annual ryegrass pasture (Lolium multiflorum ‘Maximus’) grazed by dairy cows receiving mixed rations.

Item	Equation	n	R^2	rsd
Pregrazing HM
Period 1	y = 63 x−54	18	0.82	237.6
Period 2	y = 83 x + 38	18	0.81	289.4
Period 3	y = 75 x + 151	18	0.78	336.8
General	y = 63 x + 286	54	0.71	356.2
Postgrazing HM
Period 1	y = 105 x−412	18	0.80	298.9
Period 2	y = 123 x−371	18	0.83	294.7
Period 3	y = 122 x−469	18	0.82	455.3
General	y = 118 x−423	54	0.83	261.9

y = Pre- or post-grazing above-ground herbage mass (kg DM/ha); x = pre- or post-grazing sward height (cm, compressed); rsd = residual standard deviation.

Actual pre- and postgrazing sward heights were measured daily throughout the experiment using a 0.5-m sward stick (Barthram 1986) by averaging the first contact of 100 readings taken randomly throughout the area allocated for grazing by each pair of cows. The pregrazing extended heights of the tallest leaf blade and sheath were measured on 100 randomly selected tillers on days 16, 18, and 20 of each period, and the postgrazing leaf and sheath extended heights were measured on days 18, 20, and 22 of each period on 100 randomly selected tillers per treatment. The morphological and chemical compositions of the swards offered to each pair of cows were determined on days 17, 19, and 21 of each period. Prior to cows being given access to pasture, twenty randomly selected handfuls of herbage (∼1000 g fresh) were cut at ground level, and a composite sample created. This composite sample was separated into two subsamples. One subsample was used to estimate the chemical composition of the herbage apparently selected for grazing, as follows: the harvested grass sample was cut at the post-grazing extended tiller height and the cut portion dried in an oven for 72 h at 60°C with forced ventilation, and stored for chemical analyses. The second subsample was used for morphological classification (ryegrass only). The ryegrass (∼300 g fresh) was separated into leaf (lamina + sheath), stem and senescent material. Each component was dried in an oven for 72 h at 60°C to determine the morphological composition of the pasture on a DM basis.

2.5. Chemical analyses of feeds

The DM content of all feeds was determined by drying samples at 105°C for 24 h. The ash content was measured by combustion in a muffle furnace at 550°C for 4 h, and the OM was quantified based on the mass difference. The total N content was measured by the Dumas combustion method 968.06 (AOAC 2019) using the Leco FP 528 equipment (LC, Leco Corporation, Saint Joseph, EUA). The neutral detergent fiber (NDF) concentration was assessed according to (Mertens 2002), except that the samples were weighed in filter bags and treated with neutral detergent in an ANKOM A220 system (ANKOM Technology, Macedon, NY, U.S.A.). This analysis included alpha-amylase and residual ash but did not include sodium sulfite. The concentration of acid detergent fibre (ADF) was analyzed according to Method 973.18 of the AOAC (AOAC 2019).

2.6. Statistical analyses

The dependent variables were analyzed using analysis of variance using the function PROC MIXED in the software SAS (2002, version 9.4, SAS Institute, Cary, NC). The animal variables, which were averaged per group (pair) of cows and period (n = 18), were analyzed using the following model: ${\mathrm{Y}}_{\mathrm{ijk}}=\mu +\mathrm{squar}{\mathrm{e}}_i+\mathrm{perio}{\mathrm{d}}_j+\mathrm{treatmen}{\mathrm{t}}_k+\mathrm{squar}{\mathrm{e}}_i\times \mathrm{treatmen}{\mathrm{t}}_k+\mathrm{grou}{\mathrm{p}}_{l\left(i\right)}+e_{\mathrm{ijk}}$where Y_ijk, μ, square_i, period_j, treatment_k, square_i × treatment_k, group_l(i) and e_ijkl represent the analyzed variable, the overall mean, the fixed effects of the square, the fixed effects of period, the fixed effects of treatment, the fixed effects of square × treatment interaction, the random effect of group (pair of cows) nested in square and the residual error, respectively. The fixed effect of treatment × period interaction was not significant for any animal variable, and was removed from the model.

The herbage variables were averaged per paddock and period (n = 12) and analyzed using the following model: ${\mathrm{Y}}_{\mathrm{ij}}=\mu +\mathrm{perio}{\mathrm{d}}_i+\mathrm{treatmen}{\mathrm{t}}_j+e_{i\mathrm{j}}.$where Y_ij, µ, period_i, treatment_j and e_ij represent the analyzed variable, the overall mean, the random effect of period_, the fixed effect of the treatment and the residual error, respectively.

The variables were tested via orthogonal polynomial contrasts to determine the linear and quadratic effects of the proportion of TMR restriction in the diet. The least square means were considered as significantly different if P < 0.05, P values between 0.05 and 0.10 were considered trends and SEM were reported to describe variations.

3. Results

Pre- and post-grazing herbage mass of annual ryegrass averaged 1804 and 906 kg DM/ha, respectively (Table 3). Pre- and post-grazing sward height were similar with both treatments and averaged 24.5 and 11.4 cm, respectively. The morphological composition of ryegrass was similar in paddocks grazed by both pTMR groups, with the leaf component averaging 770 g/kg DM. Herbage sampled above the grazing height with pTMR₇₅ and pTMR₅₀ was similar in composition, with an average crude protein (CP), NDF and ADF content of 303, 495 and 199 g/kg DM, respectively. Throughout the experiment the herbage sampled above the grazing height had an average CP content (in periods 1, 2 and 3, respectively) of 318, 308 and 285 g/kg DM (data not shown), an average NDF content of 468, 491 and 526 g/kg DM, and an average OM digestibility of 0.82, 0.81 and 0.79 g/g DM in periods 1, 2 and 3, respectively. The OM digestibility, energetic value and metabolizable protein content of herbage sampled above the grazing height were also similar between treatments, averaging 0.82 g/g DM, 7.0 MJ NE_L/kg DM and 120 g/kg DM, respectively.

Table 3. Pre- and post-grazing sward characteristics, sward morphology, chemical composition and nutritive value of annual ryegrass (‘Maximus’) grazed by dairy cows receiving mixed rations.

	Treatment^a		SEM	P-value
	pTMR75	pTMR50
Pre-grazing characteristics
Herbage mass, kg DM/ha	1895	1713	11.2	0.087
Sward height, cm^b	25.8	23.1	1.44	0.113
Extended tiller height, mm	439	418	5.0	0.247
Extended sheath height, mm	89.1	86.0	2.06	0.295
Post-grazing characteristics
Herbage mass, kg DM/ha	915	902	7.7	0.712
Sward height, cm^b	11.5	11.3	0.75	0.743
Extended tiller height, mm	177	184	5.5	0.711
Extended sheath height, mm	79.0	70.8	2.99	0.211
Herbage allowance, kg DM/day
Above ground level	20.4	30.5	0.50	0.001
Live leaf	15.4	23.9	1.07	0.001
Morphological composition, g/kg DM
Leaves (lamina + sheath)	755	783	16.9	0.194
Stems	208	190	11.4	0.103
Dead material	37.5	27.0	15.0	0.482
Chemical composition, g/kg DM
Dry matter, g/kg fresh	112	111	8.9	0.358
Organic matter	878	866	11.0	0.138
Crude protein	300	306	15.6	0.429
Neutral detergente fiber	494	496	12.5	0.619
Acid detergente fiber	198	199	11.4	0.856
Nutritive value
OM digestibility	0.83	0.81	0.018	0.419
NEL, MJ/kg DM	7.0	6.9	1.38	0.617
PDIN, g/kg DM^c	203	201	4.9	0.772
PDIE, g/kg DM^d	122	119	2.8	0.626

^a pTMR75 = 75% ad libitum TMR intake + grazing herbage after the morning milking (7 h/day); pTMR₅₀ = 50% ad libitum TMR intake + grazing herbage after the morning milking (7 h/day). Net energy for lactation.

^b Measured with a rising plate meter (F200 model, Farmworks, Feilding, New Zealand).

^c PDIN: metabolizable protein when N is limiting for microbial synthesis in the rumen (INRA 2007).

^d PDIE: metabolizable protein when energy is limiting for microbial synthesis in the rumen (INRA 2007).

The DM intake of the mixed ration was 10.3 kg/day less with pTMR₅₀ (P < 0.01) compared to TMR₁₀₀, whereas total DM intake decreased by only 3.6 kg/day (Table 4). Herbage DM intake increased (P < 0.001) by 1.9 kg/day from TMR₇₅ to TMR₅₀. For each kg of herbage DM consumed, intake of mixed ration decreased by 1.3 and 1.6 kg DM/day with treatments pTMR₇₅ and pTMR₅₀, respectively. Crude protein and NDF content of the consumed diet increased as the proportion of herbage in the diet increased. Diets TMR₁₀₀, pTMR₇₅ and pTMR₅₀ supplied 113%, 104% and 92% of the cows energy requirements, respectively. Cows on pTMR₅₀ had a 20.4% longer grazing time (P < 0.01) and 15.5% higher rate of DM intake (P < 0.05) than those on TMR₇₅.

Table 4. Dry matter intake, grazing behavior and chemical composition of the diet of dairy cows offered mixed rations, with or without grazing access to an annual ryegrass (‘Maximus’).

	Treatment^a			SEM	P-value
	TMR100	pTMR75	pTMR50		ANOVA	Linear	Quadratic
DM intake, kg/day
Total	19.4	18.1	15.8	0.382	0.001	0.001	0.044
Herbage	–	4.8	6.7	0.037	0.001	–	–
Mixed ration	19.4	13.3	9.1	0.324	0.001	0.001	< 0.001
Grazing behavior
Grazing time, min/day	–	244	294	4.922	0.004	–	–
DMI rate, g/min^b	–	19.9	22.8	1.516	0.014	–	–
Chemical composition of diet (g/kg DM)
Organic matter	953	933	916	0.8	0.001	0.001	0.110
Crude protein	149	189	208	1.7	0.001	0.001	<0.001
Neutral detergent fiber	366	416	427	1.6	0.001	0.001	<0.001
Acid detergente fiber	220	218	214	0.9	0.001	0.006	0.286
NEL supply^c, MJ/day	138	127	111	0.96	0.001	0.001	0.024
NEL balance^d, MJ/day	15.4	2.50	−10.9	2.96	0.001	0.001	0.879
NEL balance^e, %	113	104	92.3	2.52	0.001	0.001	0.736

^a TMR₁₀₀ = Total mixed ration ad libitum; TMR₇₅ = 75% of TMR ad libitum intake + grazing pasture after morning milking (7 h/day); TMR₅₀ = 50% of TMR ad libitum intake + grazing pasture after morning milking (7 h/day). NE_L= Net energy for lactation.

^b Herbage DM intake rate.

^c Net energy for lactation supply.

^d Net energy for lactation balance (NE_L supply – NE_L requirements).

^e % NE_L requirements.

Neither milk production nor milk fat content differed between treatments, averaging 25.6 kg/day and 44.6 g/kg, respectively (Table 5). Milk protein content decreased (P < 0.05) from 34.2 g/kg on TMR₁₀₀ to 33.8 and 33.1 g/kg in pTMR₇₅ and pTMR₅₀ treatments, respectively. Neither milk fat production nor milk protein production were affected by treatment. The MUN concentration was higher (P < 0.001) in cows offered the pTMR treatments compared with cows offered TMR₁₀₀.

Table 5. Milk production, milk composition and body weight of dairy cows offered mixed rations with or without grazing access to an annual ryegrass sward (‘Maximus’).

	Treatment^a			SEM	P-value
	TMR100	pTMR75	pTMR50		ANOVA	Linear	Quadratic
Milk production, kg/day	24.8	26.7	25.2	1.31	0.118	0.413	0.108
4% FCM production, kg/day^b	26.5	27.9	26.7	0.81	0.695	0.795	0.422
ECM, kg/day^c	28.1	28.9	28.3	1.32	0.183	0.807	0.071
Milk fat, g/kg	45.5	43.1	45.3	0.87	0.084	0.841	0.028
Milk protein, g/kg	34.2	33.8	33.1	0.73	0.024	0.007	0.704
Milk fat production, g/day	1122	1157	1184	60.3	0.338	0.148	0.910
Milk protein production, g/day	846	898	873	39.2	0.091	0.242	0.059
MUN, mg/L	17.0	19.1	20.8	0.40	0.001	0.001	0.656
Body weight. kg	559	555	556	10.69	0.688	0.483	0.623

^a TMR₁₀₀ = Total mixed ration ad libitum; pTMR₇₅ = 75% ad libitum TMR intake + grazing herbage after the morning milking (7 h/day); pTMR₅₀ = 50% ad libitum TMR intake + grazing herbage after the morning milking (7 h/day).

^b 4% fat-corrected milk production.

^c ECM, calculated as kg milk production × (37.6 × fat (g/kg) + 20.9 × protein (g/kg) + 948) / 3.13.

4. Discussion

The main objective of this study was to assess the effect of including a typical annual ryegrass herbage, with high nutrive value, but low pre-grazing herbage mass, in the diets of TMR-fed dairy cows. Mean pre-grazing herbage mass was less than 1900 kg DM/ha, as planned, with this value similar to that found in previous studies where herbage DM intake was lower than expected (Miguel et al. 2014, 2019). In addition, the OM digestibility and energetic value of herbage averaged 0.82 and 7.0 MJ NE_L/kg DM, respectively. Fresh grass with an OM digestibility greater than 0.81 and an energetic value greater than 6.8 MJ NE_L/kg DM may be classified as ‘good to excellent’ quality (Peyraud and Delagarde 2013).

4.1. Dry matter intake and grazing behaviour

The reduction in total DM intake with the pTMR₅₀ treatment is in agreement with that observed in a recent meta-analysis (Brito et al. 2022), where DM intake decreased by an average of 9.5% in cows offered a pTMR plus access to grazing, compared to cows offered a TMR only. The reduction in DM intake in the current study is unlikely to be due to the quality of herbage offered, as the relatively low NDF and high CP content of the herbage suggests that quality was good. Indeed, Vibart et al. (2008) found that total DM intake was not reduced when TMR-fed dairy cows had access to high-quality pastures. However, a reduction in pre-grazing herbage mass has been shown to reduce herbage intake (Pérez-Prieto et al. 2013), even with good quality pastures. For example, (Pérez-Prieto and Delagarde 2013) demonstrated that herbage DM intake decreased linearly when pre-grazing herbage mass (calculated above ground level) decreased from 5500 to 2500 kg DM/ha. In the current study pre-grazing herbage mass averaged only 1800 kg DM ha, even though pre-grazing sward heights were over 23 cm, a reflection of the low tiller density of annual ryegrass. Additionally, (Ferris 2007) reported that once sward height is reduced by grazing, DM intake tended to be lower on swards of low bulk density compared with swards of high bulk density. Thus, it is likely that the reduction in total DM intake with the pTMR₇₅ and pTMR₅₀ treatments is due in part to sward structure restricting intakes.

The reduction in total DM intake with the progressive inclusion of annual ryegrass in the diet may also be explained in part by the relatively high moisture content of fresh herbage. The effect of herbage moisture content on herbage DM intake has been demonstrated by (Cabrera Estrada et al. 2004). In this study, when housed dairy cows were offered fresh herbage, increasing the DM content of fresh herbage from 120 to 300 g/kg resulted in an increase in herbage DM intake of 130 g with each percentage unit increase in herbage DM content. Similarly, (Mendoza et al. 2016) attributed a reduction in total DM intake with the progressive inclusion of fresh hebage in the diets of dairy cows offered a TMR to the relatively low DM content (around 150 g/kg) of the herbage offered. In the current experiment, herbage DM content did not exceed 112 g/kg.

The greater herbage intake with pTMR₅₀ compared to pTMR₇₅ was a consequence of the greater time spent grazing and a higher DM intake rate. This result may be explained in that cows receiving the lowest pTMR supply also consumed the lowest NE_L supply from mixture. Reductions in grazing time due to increased energy supplied from corn silage based supplements have been widely reported (Pérez-Prieto et al. 2011; Wright et al. 2016; Ribeiro-Filho et al. 2021). In the same way, reductions in both daily grazing time and herbage intake rate have been observed when TMR supply was increased (Pérez-Ramirez et al. 2008; Civiero et al. 2021), while fresh forage intake rate was reduced when dairy cows had a greater intake of a mixed ration (Mendoza et al. 2018).

4.2. Milk production and milk composition

Despite the reduction in total DM intake, the inclusion of herbage in the diet of TMR-fed dairy cow did not decrease milk production, thus partially confirming the main hypothesis of this study. However, despite the reduction in intake with pTMR₇₅ (compared to TMR₁₀₀) this diet still supplied 104% of energy required for milk production. In contrast, pTMR₅₀ provided only 92% of energy requirements for milk production, suggesting that milk production may have been maintained by mobilization of body tissue reserves. Given the short-term nature of the experimental periods, it was not possible to accurately assess changes in body tissue. However, previous studies have demonstrated that cows with similar energy defecit at mid-lactation can recovery their body condition during the final third of lactation and the first 4 weeks of the dry period, without any negative effect on performance and reproduction during the subsequent lactation (Roche et al. 2006, 2017). Additionally, when the energy supply is greater than 90% of requirements, there is a reduced likelihood of metabolic disorders (Overton and Waldron 2004). In another study involving similar treatments (Pastorini et al. 2019), milk production decreased (−10%) when dairy cows were offered a diet comprising 50% of ad libitum TMR intake + fresh annual ryegrass provided in feeders. The apparently discrepance between studies may be explained in part by the difference in days-in-milk and milk production, with the study by Pastorini et al. (2019) involving early lactation dairy cows (average of 31 kg milk/day). While the current study involved mid-lactation cows (average of 25.5 kg milk/day).

The lack of difference in milk fat content between treatments may be explained by the relatively narrow range in the NDF content of the diets consumed (366–427 g/kg DM). According to the values proposed by Stockdale (1997), milk fat concentration is unaffected when the NDF content of the total diet is between 250 and 400 g/kg DM. The relatively high milk fat content observed in this study (44.6 g/kg) can be also explained by the breed characteristics, and is in agreement with the milk fat content reported in two previous studies involving cows from the same herd (Dall-Orsoletta et al. 2019; Civiero et al. 2021).

As milk production did not decrease with the progressive inclusion of grazed herbage in the diet, the observed reduction in milk protein content was not sufficient to reduce milk protein production. It has been shown that reducing energy supply decreases milk protein production (Nousiainen et al. 2004). Thus, it appears that the energy supply greater than 90% of energy requirements in pTMR₅₀ is close to the threashold to avoid a reduction in daily milk protein production. Finally, the higher MUN concentration in treatments with pasture inclusion reflects the greater CP content of the herbage compared with the mixed ration.

5. Conclusion

Giving mid-lactation cows access to high quality fresh pasture allowed the amount of TMR offered to be reduced by 50% with no loss in milk production. However, pasture access reduced total DMI, which forced cows into negative energy balance. The reduction in total DM intake with herbage access is likely due to the low tiller density, low herbage mass and low herbage DM content of the annual ryegrass.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported in part by Conselho Nacional de Desenvolvimento Cientifico e Tecnologico, Brasil (finance code 311107/2022-2), Fundacao de Amparo a Pesquisa e Inovacao do Estado de Santa Catarina (finance code 2021 TR 813), and Coordenacao de Aperfeicoamento de Pessoal de Nivel Superior, Brasil (finance code 001).