Age of regrowth as a factor affecting the nutritive value of hay of kikuyu grass (Pennisetum clandestinum) offered to lambs

Feeds, sheep and experimental design

Kikuyu grass hay was prepared by cutting the plants at a height of 10 cm from the soil surface after 50, 70 and 90 d of regrowth, between December 2004 and February 2005, from a pasture previously established in Lages, Santa Catarina, Brazil [27°49′S, 50°35′W; 937 m above sea level (asl)]. Cut material was field-dried, baled and stored in a barn. Data on rainfall, temperature and light conditions during the periods of growth of the herbage were obtained from local weather stations and are shown in Figure 1. The chemical composition of the hays is shown in Table 1.

Two experiments were carried out with Polwarth × Texel crossbred lambs housed in metabolic cages. In Experiment 1 six lambs with a mean live weight (LW) of 35 kg (s.e.m. 3·0) were used in a replicated 3 × 3 Latin square design to evaluate intake, digestibility, rumen microbial protein synthesis and N retention. The three treatments were the three ages of regrowth of the hay, which was offered ad libitum. Three of these lambs were fitted with a permanent rumen cannula (siliconized PVC, 35 cm in length, 10 mm in external diameter and 7 mm in internal diameter) to measure fermentation variables in the rumen and the other three were used to estimate the kinetics of passage of the particulate phase through the digestive tract. Each of these sets of variables was studied using a single 3 × 3 Latin square. Experiment 2 was carried out in a 3 × 3 Latin square design with three lambs with a mean LW of 33 (3·0) kg, surgically implanted with permanent indwelling catheters in the portal and mesenteric veins to measure oxygen uptake by PDV. Surgical procedures were carried out under general anaesthesia and followed those described by Katz and Bergman (1969a). For arterial blood collection, one of the carotid arteries was surgically elevated in order to be placed closer to the skin. In all experiments diets consisted of hay only and were offered at 08:00 and 17:00 to provide 100 to 200 g of refusals. A commercial mineral mixture containing per kg: Ca: 100 g, P: 45 g, S: 4·12 g, Na: 205 g, Co: 25 mg, Cu: 450 mg, Fe: 1500 mg, I: 50 mg, Mn: 1000 mg, Se: 9 mg, Zn: 2520 mg and F: 450 mg, was fed at 10 g kg⁻¹ of the dry matter (DM) intake of hay. Animals were treated in accordance with the guidelines of the Animal Care and Ethics Committee of the Universidade Federal de Santa Maria (Santa Maria, RS, Brazil).

Experiment 1

Prior to the start of the experiment, lambs received hay of kikuyu grass ad libitum for approximately 21 d. Thereafter, the experiment was conducted in three 15-d periods, each including a 10-d adaptation period and a 5-d period for data and sample collection. The hay offered and refusals, and faeces output, were recorded daily during the 5-d collection periods. Samples of refusals and faeces were collected daily and those of hay once in each collection period. All samples were dried at 55°C for at least 72 h in a forced-draught oven, ground through a 1-mm screen and stored until analysis. All urine was collected daily during each 5-d collection period in buckets containing 100 mL of 200 mL H₂SO₄ L⁻¹ H₂O. The volume was measured and a sample of 10 mL L⁻¹ of the total volume was stored at −20°C until analysis. Samples of refusals, faeces and urine were pooled on a 5-d basis within each experimental period. Supply of microbial N to the small intestine was calculated from the urinary output of purine derivatives, according to Chen and Gomes (1995).

The estimation of passage of the particulate phase through the digestive tract was performed by the use of chromium-mordanted fibre (CMF) prepared for each hay type, according to the method of Udén et al. (1980). After giving each animal a dose of CMF (approximately 20 g of CMF containing 7 g chromium kg⁻¹ DM), faeces samples were collected directly from the rectum at 0, 12, 24, 30, 34, 38, 42, 48, 60, 72, 96, 120 and 144 h, oven-dried at 55°C for at least 72 h in a forced-draught oven, ground through a 1-mm screen and stored until analysis. Parameters describing the rate of passage of CMF were estimated by the analysis of individual curves of excretion of chromium in the faeces according to the mathematical model of Grovum and Williams (1973). proc nlin of SAS (2002) with the iterative process of Marquardt was used for these calculations.

On day 15 of each period, samples of rumen fluid were collected at 0, 1, 2, 3, 4, 6 and 8 h after the morning meal and filtered through a 50-μm nylon filter. The pH was immediately measured and two 18-mL samples were taken. Two millilitres of 7·2 n H₂SO₄ was added to one sample and 2 mL of 500 g L⁻¹ trichloroacetic acid (TCA) to the other. Samples were centrifuged (4000 × g for 20 min) at room temperature and the supernatants collected and stored frozen at −20°C. Pellets were discarded. The supernatant of TCA-acidified samples was assumed to contain free amino acids and short-chain peptides (<10 units) while the pellet was assumed to comprise protein and long-chain peptides (Greenberg and Shipe, 1979).

Experiment 2

After a period of approximately 3 weeks before and 1 week after surgery, when the sheep were already housed and receiving experimental hay ad libitum, Experiment 2 was carried out in three periods of 8 d. The first 7 d were used for adaptation to the hay, and the last day of each experimental period was used for sample collection. In the morning of the last day of each experimental period, a temporary catheter [Insyte 20 gauges, 3 cm in length × 1·1 mm internal diameter (i.d.); G.A. Becton Dickenson, Maylar, France], attached to an extension with a three-way valve, was introduced into the carotid artery. Hay was offered and kept available for the sheep for only 1 h. After this period, refusals were removed and weighed. Immediately thereafter, portal blood flow (PBF) was measured by the downstream dilution of p-aminohippurate (PAH) solution (15 g L⁻¹, pH 7·4) initially primed with 10 mL followed by continuous infusion at a rate of 1 mL min⁻¹. Arterial and portal blood samples were simultaneously taken in heparinized syringes hourly from 30 min after PAH infusion. Between sampling intervals, catheters were kept heparinized with a physiological solution containing 20 IU mL⁻¹ of heparin. One sample (10 mL) was used for analysis of packed cell volume (by micro-centrifugation), and for analysis of haemoglobin and PAH concentrations. The other sample (2 mL) was obtained anoxically for immediate measurement of oxygen concentration by an automatic system of blood gas analysis (AVL 990; AVL Co, Graz, Austria).

Portal blood flow was estimated by the portal dilution of PAH (Katz and Bergman, 1969b). The concentration of oxygen in the blood indicated the sum of the oxygen associated with haemoglobin plus the oxygen dissolved in the blood aqueous phase, calculated according to Huntington and Tyrell (1985). Net PDV oxygen flux was calculated by multiplying portal blood flow by the difference between portal and arterial concentrations of oxygen. The amount of aerobic heat produced by the PDV was estimated from the resulting use of oxygen, taking into account a heat equivalent of 20·4 kJ L⁻¹ of oxygen uptake (Huntington and Tyrell, 1985).

Chemical analyses

Dry matter content was determined by drying samples at 105°C for at least 8 h. Ash was determined by combusting the samples at 550°C for 2 h. Total N was assayed by the Kjeldahl method (Method 984.13; AOAC, 1997), modified as described by Kozloski et al. (2003). Analysis of the concentration of neutral-detergent fibre (NDF) included ash but did not include either alpha amylase or sodium sulphite. This analysis was performed according to Mertens (2002) except that samples were weighed in polyester filter bags (Komarek, 1993) and treated with neutral detergent in an autoclave at 110°C for 1 h. Concentrations of acid-detergent fibre (ADF) and sulphuric-acid lignin (ADL) were analysed according to Method 973.18 (AOAC, 1997) except that asbestos was not included. Acid-detergent insoluble N (ADIN) and neutral-detergent insoluble N (NDIN) were analysed according to Licitra et al. (1996). Concentration of ether extract (EE) was determined in a reflux system with ethyl ether, at 180°C for 2 h (Soxtherm, Gerhardt, Germany). Non-fibre carbohydrate concentrations (NFC) were calculated as: 1000 − [(NDF − (NDIN × 6·25)) + (N ×6·25) + EE + ash], according to Van Soest et al. (1991). True digestibility of organic matter (OMTD) was estimated according to Mulligan et al. (2001), considering that only the NDF fraction of the faeces originated from the feed (Van Soest, 1994). Heat of combustion (H) was measured, using a bomb calorimeter (Parr Adiabatic Calorimeter, Moline, IL, USA). Intake of digestible energy (DE) was calculated as: DE (kJ d⁻¹) = (DM intake (g d⁻¹) × hay H (kJ g⁻¹ DM)) − (faecal DM (g d⁻¹) × faecal H (kJ g⁻¹ DM)). Concentration of chromium in samples of faeces was determined by atomic absorption spectrometry. For this analysis, samples were previously combusted at 550°C for 2 h and chromium was solubilized with a sulphuric acid and perchloric acid solution at 200°C (Czarnocki et al., 1961).

In urine samples, allantoin and uric acid concentrations were determined according to Chen and Gomes (1995). Uric acid was determined using a commercial kit (LABTEST, Lagoa Santa MG, Brazil) after xanthine and hypoxanthine were converted to uric acid with xanthine oxidase. Thus, the uric acid values were the sum of uric acid, xanthine and hypoxanthine and the total purine derivatives (PD) were the sum of uric acid and allantoin.

Rumen fluid samples acidified with H₂SO₄ were analysed for ammonia (Weatherburn, 1967) and sugars (Dubois et al., 1956). The TCA-acidified samples were analysed for amino acids (Palmer and Peters, 1969) before and after hydrolysis with 6 n HCl (2 mL of sample and 2 mL of 6 n HCl), at 120°C for 24 h, in an autoclave. Peptide concentrations were calculated as the difference between concentrations of amino acids before and after hydrolysis. In blood samples, haemoglobin was determined using the iron cyanide method (LABTEST kit) and PAH concentration was analysed as described by Huntington (1982).

Statistical analyses

Experiment 1

Data of intake, digestibility, N retention and microbial protein synthesis in the rumen (Experiment 1) were analysed using the GLM option of SAS (2002) according to the model:

where A, S, P, D are sheep, Latin Square, period and treatment effects, respectively, and μ is the overall mean and e_ijkl is residual error. Analysis of data of kinetics of the particulate phase in the digestive tract from Experiment 1 and of intake from Experiment 2 did not include the Latin Square effect.

Data derived from rumen (Experiment 1) and blood samples (Experiment 2) collected at each sampling interval were analysed with the mixed procedure of SAS (2002) for repeated measures according to model:

where T is time, D × T is treatment × time interaction, A(P × D) is the between-sheep random effect and e_ijkl is the within-sheep residual error. The effect of regrowth age was also analysed by regression. As appropriate, Pearson’s correlations were performed between variables.

Experiment 1

Intake of components of the hay decreased linearly with increased age of regrowth of kikuyu grass (P < 0·05), except intakes of indigestible NDF and ADF which were similar for all treatments (Table 2). Apparent digestibility of DM, OM and NDF, as well as OM true digestibility, was similar for all treatments. Nitrogen intake and retention decreased linearly (P < 0·05) with increased age of regrowth (Table 3) while apparent and true N digestibility, as well as urinary N excretion, microbial protein synthesis and efficiency of microbial protein synthesis in the rumen were similar for all treatments.

There was no treatment × time interaction for any rumen variable. The pH, sugar, amino acid and peptide concentrations of rumen fluid were similar for all treatments (Table 4) while ammonia-N concentration in rumen fluid decreased linearly with increased age of regrowth (P < 0·05). The rate of passage of the particulate phase through the reticulum–rumen (PRrr) and the caecum–proximal colon (PRcc), and, consequently, the mean retention time of particles in these compartments, were not affected by age of regrowth (Table 5).

Experiment 2

Organic matter intake, portal blood flow, oxygen uptake, heat production and heat production as a proportion of DE intake were not affected by age of regrowth (Table 6). Heat production represented on average 0·039 of daily DE intake.