Four, multiparous mid-lactation (92 DIM: ±7SD) Holstein-Friesian crossbred cows (75% HF with 25% Thai native breed) were used in the experiment. Milk yield (pre-experiment) was 10±1 kg/d and the BW was 409±20 kg. Cows were randomly assigned according to a 4×4 Latin square design. The dietary treatments were: untreated, long form rice straw (LRS); urea-treated (5%), long form rice straw (5% ULRS); urea-treated (2.5%), long form rice straw (2.5% ULRS) and urea-treated (2.5%), chopped (4 cm) rice straw (2.5% UCRS). Cows were housed in individual pens and individually fed with concentrate diets at a ratio of concentrate to milk yield of 1:2, twice daily at 06:00 am and 16:00 pm after milking. All cows were fed ad libitum rice straw as a roughage source while allowing for 10% refusals. Chemical compositions of concentrate, rice straw and urea-treated rice straw are presented in Table 1. Clean fresh water and mineral blocks were available ad libitum. The experiment was conducted for four periods, each period lasting for 21 d, the first 14 d for feed intake measurements and the remaining 7 d for sample collection of feeds, feces, urine and milk. The BW was measured at the first and last days of sampling. Rice straw was chopped to a theoretical cut length of 4 cm prior to the 2.5% UCRS treatment. Urea-treated rice straw was prepared by using 5 or 2.5 kg of urea plus 100 kg water, sprayed onto 100 kg of straw, and then covered up for 10 days before directly feeding to the animals (Wanapat et al., 2009b).

Feed intakes and refusals were measured and recorded. Feeds were sampled daily during the collection period and were composited by period prior to analyses (n = 20). Feed and fecal samples were collected during the last 7 days of each period. Fresh fecal samples (about 500 g) were collected twice daily by rectal grab sampling (morning and afternoon) after milking for five consecutive days and composited for storage in the freezer. Composited samples were later dried at 60°C and ground (1 mm screen using Cyclotech Mill, Tecator) and analyzed using the standard methods of AOAC (1995) for dry matter (DM), crude protein (CP), ash, neutral detergent fiber (NDF), acid detergent fiber (ADF) (Van Soest et al., 1991) and acid-insoluble ash (AIA). AIA was used to estimate digestibility of nutrients (Van Keulen and Young, 1977).

Urine samples (about 100 mL) were collected by spot sampling (morning and afternoon) after milking. Urination was induced by manual stimulation of the vulva. Samples were analyzed for allantoin (IAEA, 1997) and creatinine (Hawk et al., 1976). The amount of microbial purines absorbed was calculated from purine derivative (PD) excretion based on the relationship derived by Chen and Gomes (1995). Microbial crude protein (MCP) (g/d) = 3.99×0.856×mmoles of purine derivatives excreted (Galo et al., 2003). The efficiency of microbial N synthesis (EMNS) [g of N/kg of organic matter (OM) digested in the rumen (OMDR)], assuming that rumen digestion = 65% of OM digestible in total tract (ARC, 1990).

Daily milk yield of each cow was recorded and individual milk samples were collected twice daily, morning and afternoon milkings, preserved with 2-bromo-2 nitropropane-1, 3-dial, and stored at 4°C until analysis. Milk composition including fat, protein, lactose, total solids, and solid-not-fat were determined by infrared methods using Milko-Scan 33 (Foss Electric, Hillerod, Denmark). Milk urea nitrogen (MUN) was determined using Sigma kits #640 (Sigma Diagnostics, St. Louis, MO). Economic return (feed cost, milk sale and margin over feed) were calculated.

At the end of each period, rumen fluid samples were collected by using a stomach tube and vacuum pump at 0 and 4 h post-feeding. Approximate 200 mL of rumen fluid was collected at each time from the middle part of the rumen using a 60 mL hand syringe. Temperature and pH of rumen fluid were measured using a portable pH and temperature meter (Hanna Instruments HI 8424 microcomputer, Singapore). Rumen fluid samples were then filtered through 4 layers of cheesecloth. Samples were used for NH₃-N analysis to which 5 mL of 1 M H₂SO₄ was added to 50 mL of rumen fluid. The mixture was centrifuged at 16,000×g for 15 min, and the supernatant was stored at −20°C before NH₃-N analysis using Kjeltech Auto 1030 Analyzer (Bremmer and Keeney, 1965) and VFA analysis using HPLC (Samuel et al., 1997).

A blood sample (about 10 mL) was collected from the jugular vein at the same time as rumen fluid sampling into tubes containing 12 mg of EDTA and plasma was separated by centrifugation at 500×g for 10 min at 4°C and stored at −20°C until analysis of blood urea nitrogen (BUN) according to Crocker (1967).

Statistical analyses were performed using the general linear procedure in PROC GLM of SAS (1996). Data were analyzed using the model Y_ijk = μ+M_i+A_j+P_k+ε_ijk, where Y_ijk= observation from treatment i, animal j and period k; μ, the overall mean, M_i = the mean effect of treatments (i = 1 to 4); A_j = the mean effect of animals (j = 1 to 4); P_k = the mean effect of periods (k = 1 to 4) and ε_ijk the residual error. Multiple comparisons among treatment means were performed by Duncan’s New Multiple Range Test (Steel and Torrie, 1980). Differences among means with p<0.05 were accepted as representing statistically significant differences.

The effect of physical forms of urea-treated rice straw on feed intake, nutrient intake and digestibility is presented in Table 2. Rice straw intake in terms of kg/d was greater for 2.5% ULRS and 2.5% UCRS, and the greatest was in the 5% ULRS (from 4.9 to 5.8 kg/d) (p<0.05). Therefore, total DM intake was subsequently improved (from 10.2 to 11.9 kg/d (p<0.01) for the urea-treated rice straw fed groups. Several studies (Wanapat et al., 1985; Oji et al., 2007; Mapato et al., 2010) on the use of urea as a source of ammonia for the straw treatment found an increase in the feed intake of ruminants. Similarly, Gunun and Wanapat (2012) reported that roughage intake was greater for urea-treated (3%) long form or chopped rice straw (5.7 kg/d) than for untreated long form or chopped rice straw (4.8 kg/d).

Feed intake will be restricted when ruminants receive rice straw (Distel et al., 1994) because the rate of digestion is slow and distension of the rumen wall may limit intake (Grovum, 1988). In addition to providing nitrogen, urea treatment supplied more moisture content than in the untreated rice straw and lead to the softening of roughage fiber with the resulting reduction of the mastication load as well as fermentation enhancement in the rumen of cows. In addition, nutrient intake and digestibility coefficients of OM, CP, NDF and ADF were increased (p<0.05) by urea-treated rice straw when compared with untreated rice straw. Prasad et al. (1998) reported that the nutrient digestibility of OM, CP and of the fiber fractions were higher for urea treated straw rations than for untreated straw fed to crossbred cows. Wanapat et al. (2009b) reported that urea-treated (5.5%) rice straw could improve nutrient digestibility in dairy cows. Moreover, Dutta et al. (2004) and Gunun and Wanapat (2012) reported that urea-treated rice straw could improve digestibility of NDF and ADF in goats or dairy steers. Increased nutrient digestibility was also reported (Nurfeta et al., 2009) as urea treatment of rice straw affected the cleavage of linkages between lignin, hemicelluloses and cellulose that caused the structural fibers to swell. Hence, it resulted in increased surface area available for rumen microbes to attack structural carbohydrates, and greater degradability of fiber (Wanapat and Cherdthong, 2009).

The effect of the physical forms of urea-treated rice straw on rumen fermentation and BUN are presented in Table 3. Ruminal pH is a very important variable to indicate function of the rumen ecology. In the present study, ruminal pH was not affected (p>0.05) by long or short forms of urea-treated rice straw. This observation is consistent with Schroeder et al. (2003). However, Allen (1997) found that forage particle length had the most influence on the range of ruminal pH. Zhao et al. (2009) reported that ruminal pH was positively correlated to particle size of roughage; as particle size decreases, ruminal pH decreases. These results clearly showed that the short form of rice straw increases rumination activity, then increases the salivary secretion flow into the rumen and thus buffers the ruminal pH resulting in the maintenance of normal rumen function, microorganisms and fiber digestibility. Ruminal pH of all diets was in the range of 6.7 to 6.9. These values were optimal for normal rumen fermentation, microorganisms and microbial synthesis of VFA and protein (Wanapat, 1999; Anantasook et al., 2013). Van Soest (1994) suggested that the pH range for optimal microbial activity was 6.2 to 7.2. In this study, reduced rice straw particle size had no effect on rumen pH. The concentrations of ruminal NH₃-N and BUN were increased (p<0.05) with urea treatment and was highest in the 2.5% UCRS. The NH₃-N concentration was increased with 2.5% UCRS which could be due to reduced particle size, and more attachment sites could lead to greater microbial colonization (Selim et al., 2004). Ruminal NH₃-N concentrations averaged 12.2 mg/dL and were close to the optimal ruminal NH₃-N range (12 to 17 mg/dL, Wanapat and Pimpa, 1999; Mapato et al., 2010; Lunsin et al., 2012) for rumen ecology, fermentation and optimal microbial growth (Satter and Slyter, 1974; Anantasook and Wanapat, 2012). Increasing rumen NH₃-N concentrations resulted in increasing concentrations of BUN when cows were fed with urea-treated rice straw. Concentrations of BUN are highly correlated to the NH₃ concentrations in the rumen (Wanapat et al., 2008).

The concentration of total VFA was increased (p<0.05) by 5% ULRS and 2.5% UCRS. There were variations in the proportions of acetic acid and propionic acid (p<0.05) in which acetic acid decreased while propionic acid increased, thus C2:C3 was subsequently lowered (p<0.05) in cows fed with urea-treated rice straw when compared with LRS. In addition, production of propionic acid was greater in 2.5% UCRS and were greatest in 5% and 2.5% ULRS. These results could be due to treatment of straw increasing the population of microbes responsible for carbohydrate degradation. Increases in ruminal propionate concentration have been reported by Wanapat et al. (2009b). Butyric acid was not influenced by dietary treatments (p>0.05). Total VFA concentrations in all treatments ranged from 105.6 to 128.8 mM and were similar to those reported by Wanapat et al. (2009a).

The effects of the physical forms of urea-treated rice straw on the excretion of urinary purine derivatives (PD) and microbial CP supply in dairy cows are reported in Table 4. Allantoin was the main PD detected in the urine of cows, and its excretion ranged from 132.2 to 186.9 mmol/d which was in the same range as the values previously reported for cows fed rice straw (145.2 to 258.0 mmol/d; Cherdthong et al., 2011). The PD excreted in urine originates from absorbed microbial purines and purine from animal tissues (Chen et al., 1992). Moreover, urinary PD excretion is used to predict or estimate ruminal microbial protein synthesis in ruminants (Yáñez Ruiz et al., 2004). In the present study, urinary allantoin excretion and absorption were greater (p<0.05) for 2.5% ULRS and 2.5% UCRS and was the greatest in the 5% ULRS, which led to the increase in microbial CP synthesis (p<0.05). The EMNS based on organic matter truly digested in the rumen was enhanced in the urea-treated rice straw fed group (p<0.05). As NH₃-N is a most important and major source for microbial CP synthesis in the rumen (Bryant, 1974), this result could be due to urea treatment of straw increasing NH₃-N concentration (12.8 to 15.0 mg/dL) and more NH₃-N for microbial utilization during ruminal fermentation. Urine creatinine was not affected by the physical forms of the rice straw or by treatment with urea (p>0.05). The MCP flow from the rumen as calculated from purine derivative excretion using the equation of Chen and Gomes (1995) ranged from 451.2 to 767.1 g/d, respectively.

The influence of physical forms of urea-treated rice straw on yields and composition of milk in dairy cows are shown in Table 5. Cows which received long or short forms of urea-treated rice straw had higher (p<0.05) milk production (milk yield and 3.5% FCM) than those fed on long form of untreated rice straw. Similarly, Vu et al. (1999) reported that urea-treated rice straw increased milk production. Wanapat et al. (2009b) reported that milk yield tended to be increased with urea-treated rice straw. However, Mapato et al. (2010) reported that urea-treated rice straw did not affect milk production of lactating dairy cows. Glucose is important for milk synthesis (Rigout et al., 2003) and in this study more energy and protein was provided by urea treatment of straw in the form of propionic acid and microbial protein in the rumen, thus there was an improved efficiency of milk synthesis and milk yield. Moreover, the total solids, fat, protein, lactose, solids-not-fat in milk were not influenced by dietary treatments (p>0.05). These results were in agreement with Dutta et al. (2004) who reported that milk composition was not affected by urea treatment of wheat straw. In contrast, Wanapat et al. (2009b) reported that 5.5% urea-treated rice straw gave greater milk protein and fat concentrations than untreated rice straw. Reduction in the particle size of roughages decrease rumination activity, fiber digestibility, ruminal pH, decrease ruminal acetate concentration and may precipitate milk fat depression (Grant et al., 1990). Krause and Combs (2003) found that a reduction of roughage particle size reduced milk fat percentage in diets containing both corn silage and alfalfa silage. Mertens (1997) reported that a reduction in physically effective fiber causes a greater reduction in milk fat percentage in diets with a low physically effective fiber and low rumen pH compared with diets with a high physical effective fiber content and a high rumen pH. In the present study, milk fat content was not affected by different physical forms of roughage. The greater fiber digestibility, ruminal pH (6.7) and acetate concentration may be a factor in maintaining milk fat.

Cows fed on 5% ULRS and 2.5% UCRS had higher (p<0.05) MUN than 2.5% ULRS and LRS. Similarly, Wanapat et al. (2009b) found that dairy cows fed on UTRS had increased MUN levels (14.5 to 15.8 mg/dL) and improved rumen fermentation through increased digestibility and intake of straw. MUN is the fraction of milk protein that is derived from BUN and it is proportional to the value of BUN. BUN was increased in urea-treated rice straw fed group, which led to greater MUN concentration in milk. MUN concentrations in all treatments ranged from 10.1 to 15.9 mg/dL and were close to optimal MUN range (5.6 to 17.8%, Roseler et al., 1993; Wanapat et al., 2009b; Mapato et al., 2010). Roughage and total feed cost was greater (p<0.01) in the 2.5% ULRS and 2.5% UCRS groups and was the greatest in the 5% ULRS. Economic analysis of milk sales clearly demonstrated that the higher margin over feed was obtained by urea-treated rice straw fed groups compared with untreated rice straw (p<0.01) (Table 5). Similarly, García-Martínez et al. (2009) reported that sale price of milk and profit was enhanced by urea treated maize straw.