A total of 16 1-year-old male sika deer with similar body weight (BW; 63.25 ± 2.42 kg) were selected, with four animals per intake level. The deer were divided into four treatment groups, and the experiment was carried in four experimental stages. Each stage lasted for 12 days. The feeding levels of the four groups for deer were 40, 60, 80, and 100% of the recommended amount, respectively, following the nutrition requirements of feed for deer (1). To reduce the inaccuracy of the test data caused by the deer' body conditions, the deer were given a rest for 5 days and fed a normal nutrition level diet after each stage. The deer were fed two equal meals at 06:30 and 15:30 daily, and they can drink freely. The dietary composition and nutritional content of basal diets are shown in Table 1.

The digestibility of nutrients was measured by digestion trials. Deer were weighed at the beginning and end of the collection period. The feed refusals and feces were collected and weighed every day. Feces were collected every day for 3 days, stored at −20°C, then mixed and sampled again before chemical analysis, and dried at 65°C. Urine was collected in a bucket containing 20 mL of 10% concentrated sulfuric acid to avoid loss of nitrogen from urine. All the collected urine was weighed, and 3% of the daily urine output was sampled for analysis.

Four open-circuit respiration calorimetry chambers with a volume of 17.82 m³ (450 × 180 × 220 cm) were used in this study. In short, air conditioners and heaters were used to regulate the respiration chambers to maintain a constant temperature and humidity. A vacuum pump was used to continuously extract gas from the respiration chambers. The gas concentration in each respiration chamber was measured using an analyzer at a 3-min interval. O₂ was measured with a zirconia sensor, while CO₂ and CH₄ were measured with a non-dispersive infrared sensor in the analyzer.

The concentration of CH₄, CO₂, and O₂ was measured according to the method proposed by Tovar-Luna (3). Air was first analyzed for CH₄, followed by CO₂ and O₂. Before each test, analyzers were calibrated with standard gas mixtures (19.5% and 20.5% O₂, 0.0% and 1.5% CO₂, and 0.0% and 0.3% CH₄). The temperature and humidity in the calorimetry room were maintained at 20–23°C using an air conditioner at 50–55% using a dehumidifier, respectively (Whirlpool, Benton Harbor, MI).

An open-circuit respiratory heat measurement system was utilized at Deer Breeding Base of the Institute of Special Animal and Plant Science, CAAS (Jilin, China). A 4 × 4 Latin Square design was performed. Four deer were selected, with one deer put into one metabolism bin. After the 24-h adaptation period, CH₄ and CO₂ production of the individuals was measured for 24 consecutive hours. To avoid stress response in the deer in the metabolic cage, all animals had been trained previously. To reduce the inaccuracy of the test data caused by the deer' body conditions, the deer were given a rest for 5 days and fed a normal nutrition level diet after each stage.

The content of dry matter (DM), ash, Ca, and P in the feed, orts, and feces were analyzed by the method of AOAC (4). The concentrations of neutral detergent fiber (NDF) and acid detergent fiber (ADF) were determined according to the method described by Van Soest et al. (5). A bomb calorimeter (IKA C200, Germany) was used to measure gross energy (GE) in diets and feces. The GE of urine samples was measured with the method described by Deng et al. (6, 7). Carbon and nitrogen content in the feed, orts, feces, and urine was estimated in a C-N analyzer (Elementary Vario MAX CN, Germany).

Data were presented as means ± SD. The effects of feeding levels on the apparent digestibility of nutrients, energy values, energy balance, carbon-nitrogen balance, CH₄, and CO₂ emissions were analyzed using PROC GLM of SAS 8.0 (SAS Institute, Inc.; Cary, NC). The differences among the treatments were considered statistically significant with P < 0.05.

The effects of feed intake on BW, dry matter intake (DMI), and energy balance in the digestion and respirometry trials are present in Table 2. Different feeding levels had a significant impact on DM intake (P < 0.05). The feed levels significantly affected the BW of deer. The BW decreased significantly with the decrease of feed levels. As feed intake increased, total energy intake (GEI), fecal energy (FE), urine energy (UE), DE, ME, and methane energy significantly decreased (P < 0.05), while DE/GE, methane energy/GE significantly increased (P < 0.05). There was no significant difference between ME/GE and ME/DE when the deer experienced different levels of diets (P > 0.05).

The effects of feed intake level on CH₄ and CO₂ emissions are shown in Table 3. Feed intake level significantly affected CH₄ emission. CH₄ emission (L d⁻¹; L kg⁻¹ BW^0.75 d⁻¹) decreased (P < 0.05) as the feeding level decreased in the growing period. CH₄ emission (L kg⁻¹ DMI) showed an opposite trend (P < 0.05).

At the same time, feed intake level also produced significant effects on CO₂ emission. The CO₂ emission (L d⁻¹; L kg⁻¹ BW^0.75 d⁻¹) decreased (P < 0.05) as the feeding level decreased. However, CO₂ emission (L kg⁻¹ DMI) showed an opposite trend (P < 0.01).

Table 4 shows the effects of feed intake level on daily C-N balance, retained energy, and heat production, fecal C, urinary, and retention C were significantly affected by the level of feed intake (P < 0.05). The above indicators showed a significant downward trend. The content of CO₂-C and CH₄-C significantly decreased (P < 0.05) as feed intake level decreased, but the apparent digestibility of C (digestible C) was not affected by the feed intake level (P > 0.05).

RE and HP significantly decreased with a decrease in feed intake (P < 0.05), so did the energy retention components REp and REf. NI, FN, UN, RN, and protein deposited significantly decreased (P < 0.05) as feed intake level decreased, but the apparent digestibility of N had no difference at different feeding levels (P > 0.05).

The effects of feed intake level on the apparent digestibility of dry matter (DM), organic matter (OM), acid detergent fiber (ADF), and neutral detergent fiber (NDF) are shown in Table 5. As the feed intake decreased, the digestibility of DM, OM, ADF, and NDF significantly increased (P < 0.05).

The estimated values of MEm, NEm, and NEm/MEm (Km) are shown in Table 6, and the linear relationship between logHP and MEI is also shown in Figure 1. The NEm value of the male sika deer was determined to be 223.62 kJ kg⁻¹ BW^0.75d⁻¹ by calculating the antilog of the regression intercept, and the MEm value was calculated to be 251.17 kJ kg⁻¹ BW^0.75d⁻¹ through iteration of the regression equation between Log HP on MEI until HP is equal to MEI. Meanwhile, the Km (NEm/MEm) value was calculated to be 0.89.

The linear relationship between RN and NI is shown in Table 7, Figure 2. Endogenous and metabolic loss of N was calculated as 251.8 mg kg⁻¹ BW^0.75d⁻¹ by estimating the intercepts of the linear regression of RN on NI. The NPm value was calculated to be 2.045 g kg⁻¹ BW^0.75d⁻¹ for growing male sika deer.

The CH₄ emission rate is a key factor used to assess the potential degree of global warming and to estimate enteric CH₄ estimation (12). The results of this study show that the CH₄ emission rate increased as the feeding level decreased for male sika deer during their growing period. It may be an important strategy to increase feeding during the above maintenance period to reduce enteric CH₄ emission (13), which is consistent with the results of this study. For Dorper crossbred ram lambs, CH₄ energy/GE increased but DE/GE, ME/GE, and ME/DE decreased in their growing period as the feeding level increased (6). These are consistent with the results for growing male sika deer in this study, except for ME/DE, which was not affected by feed intake. In this study, the ME/DE of male sika deer was 95.01%, which was higher than that of lamb (6), broiler (14), and sheep (15). This may be related to different dietary components and interspecies differences.

Flatt (16) found that DE and CH₄ energy of cows decreased but their urinary energy remained unchanged during their pregnancy. In contrast, Ferrell et al. (17) reported that any difference in energy digestibility of heifers due to pregnancy was insignificant compared to differences in feeding levels. The current data of male sika deer during their growing period suggest that the methane energy, urinary energy, fecal energy, and DE increased as the feed intake level increased. The results of this study are consistent with those reported above.

Blaxter (18) found that because the C-N balance method involved more analysis and measurement than the energy balance method, errors of about 30% may be encountered. As revealed by Kishan et al. (19), for male buffaloes and crossbred cattle, energy levels affected the excretion of C and N in urine, the C in urine was significantly related to DE intake (P < 0.01), but the urinary, CO₂, and CH₄ carbon outgo were not affected. There is also a correlation between N excretion and urinary C content (20). These are consistent with the results of this study. Furthermore, the results of this study show that the apparent digestibility of C was 52–63%, which is consistent with the results reported by Blaxter and Wainman (20). There is a certain difference between RC and RN calculated using the C-N balance method in this study. Graham (21) found that the RN calculated using the comparative slaughter method decreased as the number of gestation days increased. In this study, RN and RE decreased as the levels of feeding decrease. This is consistent with the results of Zhang et al. (15), Singh et al. (22), and George et al. (23).

Ferrell (24) found that energy intake affected HP because of the metabolic activity of visceral organs. As the MEI of growing cattle increased, HP increased exponentially (25–27). As indicated by the results of Taylor and Turner (28), HP increased as the nutrient level increased, which is mainly due to the increased metabolism associated with energy retention. Analysis of energy metabolism for growing cattle indicated that HP increased exponentially with an increase in MEI level (25). It has been well-established that HP will increase significantly during pregnancy (29). In this study, MEI and HP gradually decreased, accompanied by a decrease in feeding levels, which may be due to the thermal effect of reduced feed. Meanwhile, ME intake decreased, leading to a decrease in HP (26, 30, 31). These are consistent with the findings in growing and fattening pigs reported by Zhang et al. (32).

The REp value is much lower than REf value. When deer were fed at 40% of the intake level, REp and REf reached their minimum values. These results are consistent with the findings in Hu sheep (15) and in arctic foxes (33). In this study, the decrease in the feed intake of male sika deer led to the reduction of NI, FN, and UN. As the feed intake decreased, RN gradually decreased from 0.38 to 0.20 g kg⁻¹ BW^0.75d⁻¹. This is consistent with the findings of Singh et al. (22).

The digestibility of nutrients in the rumen is the competition result between digestion and passing rate. Among them, the passing rate is positively correlated with dry matter intake (34). Degen and Young (35) found a correlation between increased digesta passage rates and increased feed intake. In this study, the digestibility values of C, N, DM, OM, ADF, and NDF increased significantly with a decrease in diet intake, indicating that dietary restriction can improve the digestibility and utilization of nutrients. The deer body promotes nutrient digestibility and utilization to meet maintenance requirements while feed intake is less. Deng et al. (6) found that the apparent energy digestibility (DE/GE) and metabolic rate (ME/GE) of lambs fed ad libitum were lower than lambs fed in a restricted manner. The results on male sika deer in this study are consistent with the above research results.

The logarithmic relationship between MEI and HP is often used to calculate NEm. HP is equal to NEm in the case of zero MEI (10, 26). Similarly, MEm can also be calculated by extrapolating HP where it is equal to MEI. The values of NEm and MEm calculated by the regression equation were 223.62 and 251.17 kJ kg⁻¹ BW^0.75d⁻¹, respectively. The Km was calculated to be 0.89 in this study. Li et al. (36) studied the energy metabolism of adult male sika deer and determined that the requirement of MEm of adult sika deer was 516 kJ kg⁻¹ BW^0.75d⁻¹. The result is greater than that in this study. The reason may be the different ages of the deer selected for experiments and the different physiological states of each period. Adult deer need more energy to maintain their growth and antler. In contrast, for deer in the growing period, more energy is used for the development and growth of their own bodies, and thus less MEm was observed in them. The Km value in this study is higher than that (0.707) reported by Li (36) in adult deer and that of lambs and sheep (14, 15, 32). This may also be related to the age of the selected deer. Luo et al. (37) reported that the difference in MEm seems to depend mainly on the change in NEm rather than Km, which can explain the difference in MEm requirements. The values of NEm and MEm vary from species to species. For animals, the species, physiological stages, environmental temperature, and feed composition also affect the values of NEm and Mem (14, 38). MEm was 768 kJ kg⁻¹ BW^0.75d⁻¹ and 501 kJ kg⁻¹ BW^0.75d⁻¹ at 18 and 24°C, respectively for adult female mink (38). In this study, male sika deer with a bodyweight of 56.43–64.26 kg were selected. Meanwhile, the temperature in the respiration chamber was 22°C, and the male sika deer in Northeast China during their growing period was chosen. These may be the main reasons why the NEm and MEm values are different from those in previous reports (36).

ARC (39) assumed that NPm equal to the amount of protein that can offset the loss of urine, feces, and dermal N, except for growing lambs because they do not consider dermal loss ARC. For lambs and sheep (15, 40), the regression equation between RN and daily NI is an effective method to obtain NPm through N-balance trails. The intercept of the regression equation represents the endogenous and metabolic N loss. According to the regression equation between the daily NI and RN of the male sika deer during their growing period, the values of NNm and NPm were estimated to be 251.8 mg kg⁻¹ BW^0.75d⁻¹ and 2.045 g kg⁻¹BW^0.75d⁻¹, respectively. Chizzotti et al. (26) reported that the NPm estimated according to the relationship between RN and daily NI using the comparative slaughter method is greater than the NPm determined based on the relationship between RN and daily NI using the N-balance method. This discrepancy may be due to losses of N that are not accounted for by the N balance method. The scurf protein represents about 20% of the maintenance requirement of the ARC system. However, no data was present in growing male sika deer. The N balance trails can reduce animal injuries and meet animal welfare requirements and is an important method for estimating net protein requirement.