This study was conducted from July 2012 to September 2012 at the Experimental Station of the Chinese Academy of Agricultural Sciences (CAAS) in Nankou, Beijing, China. The animals were kept in an enclosed facility equipped with heating radiators, and the mean temperature recorded inside the facility during the experimental period was 25.0°C (22.5°C to 28.0°C). The experimental procedures were approved by the Animal Ethics Committee of CAAS, and humane animal care and handling procedures were followed throughout the experiment.

Four 10-month-old Dorper×thin-tailed Han crossbred castrated lambs (62.3±1.9 kg body weight) were ruminally cannulated and used in this study 4 months after the surgery. In a 4×4 Latin-square experimental design, the lambs were randomly assigned to 4 dietary treatments in a 2×2 factorial arrangement: two levels of concentrate:forage (C:F, 40:60 and 60:40), and two levels of undegraded dietary protein (UDP, 35% and 50% of CP). The ingredients and chemical composition of the experimental diets are presented in Table 1. Lambs were fed twice daily at 08:00 h and 16:00 h for ad libitum intake, allowing for approximate 10% refusals. Clean water was available at all times. Each experimental period lasted for 19 d: The first 7 d were used for adaptation to dietary treatment. Lambs were then moved into individual crates for 12 d including 7 d of adaption and 5 d of a metabolism trial. All lambs were weighed at the beginning and end of the trial in each period.

In the metabolism trial, samples of feed were collected daily, combined, dried at 65°C for 72 h, and ground through a 1-mm sieve. Feed refusals were weighed, sampled, dried, ground, and combined for each lamb before analysis. Faeces were weighed daily, and then a sample of 10% was collected. The samples were pooled for each animal, dried at 65°C, and then ground through a 1-mm sieve before analysis. Urine from each lamb was also collected daily in a bucket containing 100 mL of 7.2 N H₂SO_4. The volume was measured and then diluted to 5 L with tap water, and a sample of 20 mL was collected, pooled for each animal, and stored at −20°C for analyzing total N and PD. Urinary spot samples were collected between 8:00 h and 9:30 h, 14:00 h and 15:30 h and 20:00 h and 21:30 h during the collection period. After collection of a urinary spot sample, 10 mL was sub-sampled, acidified with 1 mL of 10% H₂SO₄ and diluted four times with tap water. The whole of the diluted sub-sample was placed in a plastic bottle and kept at −20°C. Subsequent sub-samples were added into the same collection bottle. Urinary spot samples taken at different times from each animal were stored in separate bottles, thus three bottles were designated for the collection of urinary spot samples for each animal. The remaining urine was poured back into the collection bucket after taking a spot sample.

Dry matter was determined by drying samples in a forced-air oven at 135°C for 2 h (method 930.15; AOAC, 1990); Organic matter was measured as the difference between DM and the ash content (method 938.08; AOAC, 1990). Gross energy was measured using a bomb calorimeter (C200, IKA Works Inc., Staufen, Germany). Nitrogen was determined by the Kjeldahl method, using Se as a catalyst, and crude protein was calculated as 6.25×N. Dietary content of RDP was estimated using the in situ method by incubating samples in the rumen for 72 h. Feed preparation, nylon bags, rumen incubation program, and washing and drying procedures were completed as described by Volden and Harstad (1995). Ether extract (EE) was measured by weight loss of the dry matter on extraction with diethyl ether in Soxhlet extraction apparatuses for 8 h (method 920.85; AOAC, 1990). Neutral-detergent fibre (NDF) and acid-detergent fibre (ADF) were determined according to Van Soest et al. (1991) and Goering and Van Soest (1970), respectively. Urinary PD including allantoin, uric acid, xanthine, and hypoxanthine were quantified separately by the colorimetric method of Chen and Gomes (1995). Creatinine was measured by commercial kits (C011 Creatinine Kit, Jiancheng Bioengineering Institute, Nanjing, China) using an automated analyzer (ZY-310, Kehua Enterprise Development Co., Ltd, Shanghai, China).

The amount of microbial purines absorbed (X, mmol/d) corresponding to urinary PD (Y, mmol/d) was calculated based on the relationship derived by Chen and Gomes (1995), as

The supply of microbial N in gram per day was estimated as follows:

Where 0.83 is the digestibility of microbial purine, 70 is the N content of purines (mg N/mmol), 0.116 is the ratio of purine-N:total N in mixed rumen microbes.

Data for intake, nitrogen balance, urinary PD, creatinine, and PDC index in urine collected during 24 h were analyzed as an 4×4 Latin square using the MIXED procedure of SAS (version 9.1; SAS Institute, Inc., Cary, USA). The effect of dietary treatments (C:F and UDP) and period were considered as fixed effect and lamb was considered as random effect. Observations of urinary PD, creatinine, and PDC index calculated from spot urine samples were analyzed as repeated measures. The statistical model used included C:F, UDP, C:F×UDP, period, time, C:F×time, UDP ×time, and C:F×UDP×time as fixed effects, and lamb as a random effect. Data for each variable were analyzed using compound symmetry, unstructured, and autoregressive covariance structures, and the one that produced the minimum Akaike’s information criterion was chosen. Statistical significance was accepted when p<0.05.

The DM and OM intakes (Table 2) were not affected by dietary C:F (p>0.05), but increased with decreasing level of UDP (p<0.01). The DM intake in our study ranged from 1.4 to 1.7 kg/d. These observations were in accordance with those of Kiran and Mutsvangwa (2007) who found an increasing dietary RDP level from 60% of CP to 70% of CP increased DM intake in growing lambs (ranging from 1.3 to 1.6 kg/d). Feed intake is affected by factors including palatability, feeding frequency, and nutrient levels. The impact of supplementation of RDP on forage intake is likely related to improvements in fiber digestion and, subsequently, rate of passage (Arroquy et al., 2004). It is also possible that RDP supplementation affected passage rate and intake by impacting gastrointestinal motility (Egan and Moir, 1965). Similar to our results, Cantalapiedra et al. (2008) and Ramos et al. (2009) found no effect of dietary C:F (30:70 and 70:30) on DM or OM intake in goats and sheep fed diets beyond maintenance level, respectively.

All lambs maintained a positive nitrogen balance in the current study (Table 2). Nitrogen intake was not affected by dietary C:F (p>0.05), but increased as level of UDP decreased (p<0.05). As the diets used in the present study were iso-nitrogenous, N intake followed the same pattern as DM intake.

Fecal N was not affected by dietary treatments (p>0.05). However, Urinary N was reduced (p<0.05) with increasing dietary C:F, but increased as UDP level decreased (p<0.05). Nitrogen in feces is the sum of undigested dietary N, microbial fecal N, and endogenous N, while urinary N excretion was a combined consequence of the absorption of ammonia N from the rumen and recycling of that N back into the gut, which was in turn related to the composition of the diet and the ruminal utilization efficiency of dietary N (Moorby et al., 2006). In the current study, fecal N was stable for all groups and took approximate 42% of N intake while the proportion of urinary N to N intake decreased from 45% to 27% as dietary C:F increased, and this proportion increased by approximate 7% as level of UDP decreased.

A significant change in urinary N rather than fecal N in the present study indicated that the main route for output of dietary N in excess of animal requirements was via the urine. As the experimental diets were designed as iso-caloric and iso-nitrogenous, the main dietary factor contributing to the difference in urinary N between lamb fed diets differing in C:F could be the structure of the carbohydrates. Diets higher in NFC can provide more available energy to rumen microbes than diets higher in NDF, which in turn can better utilize degraded protein and reduce the N output in urine. On the other hand, it was reported (Reynal and Broderick, 2005) that protein degradability had a limited effect on fecal N output but had a major effect on urine N output because of the excess N in the rumen from diets with a higher proportion of RDP. The imbalance of N availability and N capture by microbes in diets with excessive degradable protein contents resulted in surplus N in the rumen which was converted to urea and excreted via the urine (Kebreab et al., 2002).

Excretion of PD and creatinine in total urine, PDC index, and microbial N (MN) are shown in Table 3. In the current study, allantoin was the dominant PD, then followed by uric acid and xanthine plus hypoxanthine. Similar patterns were also reported by Lindberg (1989) and Belenguer et al. (2002) in goats. Urinary excretion of allantoin and total PD increased (p<0.01) as dietary C:F increased, while uric acid and xanthine plus hypoxanthine was not affected by dietary treatments (p>0.05). This could be partly explained by the relatively lower concentrations of uric acid (averaged 8.8%) and xanthine plus hypoxanthine (averaged 6.3%) than allantoin (averaged 84.8%) in urine.

Creatinine is an indicator of body protein turnover and creatinine in urine is excreted in proportion to live weight (Brody, 1964). As body weight of lambs in our study was similar, a relatively constant urinary creatinine output was observed (0.33±0.02 mmol/kg BW^0.75). This value is higher than that reported by Balcells et al. (1991) in ewes (0.29 mmol/kg BW^0.75) but lower than that reported by Chen et al. (1995) also in ewes (0.47 mmol/kg BW^0.75). It is conceivable that between breed/species differences in musculature and variations in body weight could account for these discrepancies (George et al., 2006). The ratio of PD:C was corrected for metabolic body weight to facilitate comparison between animals of different body weight, and this W^0.75-corrected PD:C ratio is referred to as ‘PDC index’ (Guerouali et al., 2004). Dietary effects on PDC index in the current study followed the same pattern as urinary PD, and a linear correlation existed between total PD excretion and PDC index (R² = 0.93; Figure 1).

Microbial N was higher for lambs fed 60:40, C:F diet comparing with those fed 40:60 diet (p<0.05), but no effect of UDP level was observed (p>0.05). This result was in accordance with that reported by Pina et al. (2009) in beef heifers. Microbial protein produced in the rumen was found to always be higher when the diet contained higher proportion of nonstructural carbohydrates in DM (Hoover and Stokes, 1991). It is likely that N supply for microbial growth was not limiting under current dietary regime, thus UDP level did not affect MN synthesis estimated using urinary PD excretion. Previous studies in goats (Cantalapiedra et al., 2008) and sheep (Ramos et al., 2009) have shown that increasing the proportion of concentrate in diet could result in a higher microbial yield. The higher MN synthesis could be the result of the higher availability of rumen fermentable carbohydrates, as the rate and extent of carbohydrate fermentation are two of the major factors controlling energy available for microbial growth (Russell et al., 1992). However, Russell et al. (1992) reported that at a NDF content lower than 200 g/kg DM, the MN yield was reduced by 2.5% to 3.0% per unit decrease of NDF. All the diets used in the current study were above the limit of this NDF level.

Concentrations of urinary PD, creatinine, and PDC index in spot urine are shown in Table 4. Allantoin accounted for 84% of total PD then followed uric acid (14%) and xanthine plus hypoxanthine (2%). Concentrations of total PD (p<0.05), allantoin (p<0.05), uric acid (p<0.05), and PDC index increased (p<0.05) with increasing level of dietary C:F, but were not affected by sampling time (p>0.05). Concentration of creatinine was also unaffected by sampling time (p>0.05). All above entries were not affected by level of UDP (p>0.05). These results indicate that when a given feed is offered ad libitum to animals with varying body weights, the PD concentration and PDC index in urinary spot samples may vary within narrow bounds despite intake varying quite widely. Similarly to Chen et al. (1992), who reported that insignificant diurnal fluctuations in the plasma allantoin concentration and urinary PD/C ratio occurred in steers fed twice daily.

A linear correlation existed between the daily PD excretion and PDC index in each sampling period (R² = 0.63, 0.83, and 0.80 for 08:00 to 09:30, 14:00 to 15:30, and 20:00 to 21:30 h, respectively). When the PDC index from three spot urine samples were averaged, an overall correlation was established between daily PD excretion and averaged PDC index for spot urine (R² = 0.88; Figure 2). A similar correlation was also found in sheep (R² = 0.81) when spot urine was collected hourly (Chen et al., 1995). Chen et al. (2004) reviewed that the coefficient of variation (CV) for the PD/creatinine ratio from a single spot sample was 11% to 16% and if the mean of four samples was used, the CV would be reduced to about 4%. Thus multiple samples are required to obtain a measurement representative of the dietary regime. However, it is inconvenient to collect spot urine samples intensively (e.g. every hour) under practical circumstances otherwise; the advantage of using spot samples would be compromised. From the current study, we observed a weaker relationship between daily PD excretion and PDC index calculated from spot urine collected between 08:00 to 09:30 compared with that between 14:00 to 15:30 h or 20:00 to 21:30 h. Similarly, Fujihara et al. (2005) found that when feed was given ad libitum and spot urine were collected hourly, the highest correlation between daily total PD excretion and PD concentration occurred in spot urine collected 7 to 8 h after morning feeding in sheep fed twice daily (09:00 and 17:00). For animals fed ad libitum, the supply of microbial protein at the post-ruminal gut will have peak periods depending on meal periods, and the size of each peak should be proportional to the meal size (Nsahlai et al., 2000). Low sensitivity in predicting total PD excretion from spot urine collected between 08:00 to 09:30 h may indicate that this time span fell, or partly fell in the lag phase for the hydration and colonization of feed particles by rumen microbes during which time the supply of MN as well as microbial purine bases cannot be properly reflected by the concentration of PD in spot urine. Thus the collection of spot urine should be avoided immediately after feeding because when feed is given ad libitum, animals do not eat continuously (Sauvant et al., 1996), but tend to consume feed in large meals, as we observed in current study. Therefore, an appropriate lag phase between the time of feeding and sampling should be determined so that the sampling time can coincides with the peak concentration of PD in the spot urine.