The yield of wheat in Henan Province is the highest in the world. Henan province (China) is located at 30 to 36 degree of north latitude, which belongs to warm temperate transition continental monsoon climate and is suitable for wheat growth. A total of 17 wheat samples were collected from 16 cities in Henan Province during mid-to-late May at the fully ripe stage (Table 1). The chemical constituents of 17 wheat cultivar samples are analyzed and presented in Table 2.

The procedures of this experiment were verified and approved by the Institutional Animal Care and Use Committee of Shangqiu Normal University (2023-1109). This experiment was executed in the animal experimental base of Shangqiu Normal University (Shangqiu, China).

Fifty-one Duroc × (Landrace × Yorkshire) growing barrows (body weight of 30.1 ± 1.8 kg) were cultivated in temperature-controlled room, and environment temperature was ranged from 20 to 25 °C. Each pig was cultivated in metabolism crates (1.4 × 0.7 × 0.6 m). Barrows were randomly allotted to an 17 × 2 incomplete Latin Square design with 17 diets and two consecutive 12-d periods. Each diet included 3 barrows in each period with 6 replications per diet. Dietary composition was shown in Table 3. The chemical constituents of diets were measured and showed in Table 4. Pigs received an equal amount of feed equivalent to 4% of body weight, and feed and water were available ad libitum. The diet ingredient were ground by a 2.5-mm screen (hammer mill). The feed was divided into two equal portions and administered at 8:00 h and 16:00 h daily. The experiment was lasted for 12 d (7 d for adaptive phase and 5 d for samples collection). The experimental samples and protocol were based on previous studies (8). From 8 d to 12 d, feces were gathered and stored at −20 °C to suppress degradation of the samples caused by microorganism. At last, the collected samples for feces were weighed, dissolved, and then homogenized thoroughly. Feces sample of 300 g was dried at 65 °C for 72 h. The 50 mL of 6 N HCl was added into buckets to gather urine sample. We measured the volume of collected urine daily, and 20% of urine was stored at −20 °C.

The detection of dry matter (DM), starch, CP, ash, calcium (Ca), and phosphorus (P) is carried out according to a previous report (9). The ether extract (EE) of samples was determined as previously described (10). The NDF and ADF were analyzed according to McGhee and Stein (11). The GE of all samples were analyzed by an Automatic Isoperibol Oxygen Bomb Calorimeter (Parr 1281 Calorimeter, Moline, IL).

The ATTD of nutrients, DE, and ME of wheat samples were calculated according to the equation described by previous research (12):

In which DE_d and ME_d are the DE and ME in wheat diets (MJ/kg of DM); GE_i is the amount of GE ingested for each pig (MJ of DM); F_i represented the actual amount of feed intake; the amount of GE in the feces and urine of each pig (MJ of DM) are represented by GE_f and GE_u; DE_w and ME_w are the content of DE and ME in each wheat sample (MJ/kg of DM); DN and FN are the concentrations of nutrients in diets and feces; F_o is the output of feces.

Data analysis was executed by SAS 9.2. A pig was an experimental unit. The diet was regarded as the fixed effect, and the period and animal were the random effects. The correlation and stepwise regression analysis between chemical compositions and DE, ME were analyzed by the CORR and REG procedures (13). The LSMEANS procedure was executed to calculate the mean values. Tukey’s multiple range test was used to separate statistical differences. The significance level was set at p < 0.05. The equations were considered as the best-fit prediction equations of DE and ME when the R² is greatest and the residual standard deviation (RSD) is smallest (14).

In this study, the chemical composition and bulk density were presented in Table 2 on a DM basis. CP, starch, and crude fat are essential sources of nitrogen and energy to promote pig growth. The previous study indicated that the contents of CP, starch, and EE were ranged from 12.40 to 17.40%, 54.10 to 58.40%, 1.40 to 1.80% in 12 samples of wheat from Canada Prairie and Western (15). Zhao et al. (14) found that the contents of CP, starch, and EE were ranged from 14.25 to 16.14%, 53.43 to 57.52%, 1.30 to 1.34% in 2 samples of wheat. The results of the CP and EE contents mentioned above were consistent with our data. We found that the contents of CP and EE were ranged from 11.26 to 17.63%, 1.39 to 1.93% in 17 samples of wheat. Moreover, there was a considerable variation in the CP content within different cultivars of wheat (11.26 to 17.63%; CV = 9.00%). This indicated that although the varieties and origins were different, the variability in the energy value of wheat and its primary determinants were similar. Zhang et al. (4) reported that GE, CP, EE, NDF, and ADF is 17.24 MJ/kg, 14.07, 1.44, 15.21, 2.45% in extruded wheat, respectively. Apart from the inherent varietal differences, subsequent processing technologies (such as extrusion and pelleting) also influence the ultimate nutritional value of wheat.

In our study, higher starch content was observed compared with those of the previous reports (14, 15). The variation of wheat starch and CP contents were caused by different varieties, growing season and regions (16). Many previous studies reported that a considerable variation of CP was caused by different climatic conditions rather than soil quality or genetic factors in wheat sample (17, 18). The abundant soil moisture and low environment temperature during the ripening of the wheat results in lower content of protein, whereas the CP content of wheat was higher in the warm weather conditions (15). Starch constitutes the principal component of wheat and serves as the substrate for sugar provision during fermentation. Given the growing demand for wheat as a renewable resource, research aimed at developing wheat varieties with elevated starch content may attract keen interest (19). The content of wheat carbohydrates could provide energy and determine the pattern and efficiency of energy release (20). In addition, protein-riched wheat was preferentially utilized for breadmaking and the requirement of society (21, 22). Whereas wheat, as a feedstuff and renewable resource for biofuel production, eventually contains large amounts of starch (23).

The level of wheat fiber components was affected by the cultivars and growing environment. Many previous studies reported that NDF and ADF fractions were ranged from 16.34 to 25.10% and 2.70 to 5.10% in different cultivars and harvest times of wheat (4, 16, 24, 25). Other researches demonstrated that NDF content and ADF content were from 10.99–15.83%, 2.04–4.70%, which were consistent with our study (12, 14, 26). We found that there were large variations in the NDF concentration (10.82 to 29.20%; CV = 34.60%), and the ADF content (2.33 to 3.33%) was lower compared with the reports mentioned above. Climatic temperature has a great influence on the content of NDF and ADF of wheat. The level of fiber component in wheat in winter was lower than in summer (12). The nutrient absorption capacity of wheat root was decreased in low-temperature, moreover, part of the nutrients in wheat are used to resist cold stress (27). The above findings demonstrated that the interplay between genetic attributes and environmental factors determine nutritional composition of wheat, and ultimately results in a predictable and comparable range of available energy content for swine (see Table 5).

The ATTD of GE, OM, CP, NDF, ADF, EE and Ca showed significant differences (p < 0.05) (Table 6). And the ATTD of GE, DM, OM, CP, NDF, ADF, and EE ranged from 87.30 to 90.77%, 87.76 to 90.37%, 88.63 to 91.06%, 81.73 to 90.44%, 49.04 to 66.14%, 18.19 to 46.31%, 27.07 to 57.13% among 17 wheat cultivars, respectively. The ATTD of GE in our result was similar to previous studies, which reported that the ATTD of GE in the diets ranged from 86.58 to 90.79% in different wheat cultivars (4, 12, 15). An earlier research reported a lower ATTD of GE and DM (ranged from 76.2 to 87.0% and 79.1 to 89.4%) in 20 different wheat grains in growing pigs (8). The different ATTD of GE and nutrients were associated with various chemical composition of wheat grains grown in various areas and climates, for example, the ATTD of GE in winter wheat is higher than that of the spring (12, 28). Previous research demonstrated that the ATTD of DM, NDF, ADF, and CP ranged from 87.47–88.54%, 56.78–64.21%, 25.80–40.74%, and 83.63–87.58% (14), which were consistent with our study.

In the present study, the DE and ME contents ranged from 16.06 to 16.88 MJ/kg DM and 15.65 to 16.36 MJ/kg DM in 17 wheat cultivars, respectively (Table 7). In agreement with our findings, previous study demonstrated that the DE and ME contents were ranged from 16.40 to 17.02 MJ/kg DM and 15.73 to 16.49 MJ/kg DM in 12 wheat cultivars (12). Other research demonstrated that the DE of 5 wheat cultivars was from 16.19 to 16.99 MJ/kg DM (15). Similarly to the above researches, another study indicated that the DE content of 50 wheat cultivars (10 varieties each grown at five sites) were ranged from 15.49–16.95 MJ/kg (24). The DE content of No. 5 wheat sample is lowest, which may be due to its lower GE and CP content. On the other hand, high starch content of No. 5 wheat sample hindered the interaction between digestive enzymes and substrates (29, 30). The ATTD of CP, GE, and DM in No. 10 and 11 wheat samples were higher, and NDF content was lower, which may mainly contribute to greater DE and ME. However, in this study, we found that the DE content was higher than the values (13.44 to 13.71 MJ/kg) reported by Zhao et al. (14). This may due to higher ATTD of GE and EE content. Because fats facilitate digestion and reduce the flow rate of chyme (31), meanwhile, lower NDF content decreases endogenous losses of protein and fat, resulting in increasing of the excretion of organic matter (32, 33). In addition, carbohydrates are the largest source of energy for wheat, and carbohydrate digestibility largely determines the DE and ME contents of wheat (11). The discrepant results of the DE and ME contents may attribute to the differences in chemical composition and physical characteristics resulting from different cultivars and growing environment of wheat (34, 35) (see Table 8).

The metabolizable energy values derived from distinct wheat cultivars, heterogeneous experimental protocols, and divergent computational methodologies exhibit substantial variability. Regression models predicated on wheat chemical composition have been established as a recognized and standardized predictive paradigm. Previous studies have indicated that both ADF and NDF content are negatively correlated with DE of wheat, which is attributed to the reduced nutrient digestibility caused by the anti-nutritive properties of fibrous components (36, 37). Consequently, ADF and NDF are widely regarded as reliable predictors of available energy in wheat, due to their negative impact on the digestibility of nutrients (38, 39). Apart from macronutrients, wheat contains a diverse array of compounds such as phenolic acids, flavonoids, and carotenoids. Advanced analytical techniques can elucidate the variability in the content and composition of these compounds in different wheat varieties, growing environments, and processing methods (40). The variability of the compounds could be utilized to predict the nutritional quality and metabolic characteristics of wheat in animal diets. NDF and ADF contribute to a higher heat increment during fermentation or digestion processes and physically impede the digestion of other nutrients. Hence, they are routinely incorporated as negative predictors in net energy prediction equations (41). On the other hand, a prediction equation was developed by Son et al. (7) and Huang et al. (23) through nutritional composition analysis of the diet (including corn, soybean meal, and wheat), wherein NDF and ADF were identified as the principal predictive factors. Consequently, neutral detergent fiber (NDF) and acid detergent fiber (ADF) consistently serve as fundamental predictive variables in energy prediction models of diets with high fiber content. Previous research investigated the energy prediction model of high-fiber wheat bran and derived the prediction equation: NE = 11.16 − 0.09NDF (R² = 0.71) and NE = 10.51 − 0.27ADF (R² = 0.68), however, the regression equation is NE = −21.65 + 1.20 GE + 0.35 × CP (R² = 0.94) and the R² was elevated (0.65 vs. 0.94) when the GE and CP was regarded as predictors of regression equation (38). Therefore, CP and gross energy GE serve as the dominant predictors in energy prediction models. In this study, there were no significant correlations between ADF or NDF and DE content (Figure 1). This different results is likely because the samples were whole wheat grains in our study, which have a lower and less variable fiber content compared to bran. The limited range of fiber variation was insufficient to exert a dominant anti-nutritive effect on energy digestibility, allowing the positive effects of CP and GE to be the primary determinants. The result could be associated with a faint variation of NDF and ADF contents in wheat samples. In summary, ADF and NDF did not serve as predictive factors in our study.

GE was significantly negatively correlated with starch (p < 0.05). Compared to protein and fat, starch is the primary component in wheat with a relatively low energy density. When the starch content is high in wheat, the levels of other components are decline, particularly protein and fat, which has an extremely high GE. Meanwhile, the formation of starch-protein interactions contributes to reduction GE (42). The CP content was positively related (p < 0.05) to GE, DE, and ME concentrations in wheat. The most suitable predictor for DE was the combination of CP in the present study (Figure 1), which was consistent with a previous study that reported the variation of DE in different wheat samples was mainly caused by the content of CP (35). We analyzed data and gained a regression equation of DE = 14.6609 + 0.1216 CP. Previous research demonstrated that GE could precisely predict the DE content of wheat, and gained a regression equation of DE = −1,384 + 1.1 GE + 0.65 bulk weight (R² = 0.83) (12). Thus, based on the equation, the GE was added as a predictor of regression equation on account of positive correlation between GE and CP content of wheat. We gained a regression equation of DE = 26.6394 − 0.6783 GE + 0.1618 CP and found that the R² was elevated (0.65 vs. 0.72), and the analytical method referred to previous research (43). Differences in predictive factors and weighting coefficients could originate from seasonal variations and regional differences in wheat sample collection. The above findings validate the universality of crude protein, gross energy, and bulk density in predicting the energy value of wheat once again.

In this study, the most suitable predictor for ME was the DE content of wheat, which was consistent with some previous studies (44–46). In this study, regression equation is ME = 3.1923 + 0.7741 DE (R² = 0.73). Huang et al. (23) demonstrated that regression equation is ME = 0.92 DE + 0.44 (R² = 0.96) in mixed diet (include corn, soybean, and wheat shorts). In addition, Zhao et al. (14) found that regression equation is ME = 0.9716 × DE (MJ/kg) + 0.0636 (R² = 0.99) in mixed diet (corn, wheat, and rice). The difference in R² between the above results and our findings might be attributed to high corn content in diets. Corn, as a energy-dense ingredient with relatively uniform composition, minimizes variation in key predictors like starch content, delivering tighter statistical constraints and higher R². Previous study reported starch was the main component for predicting energy metabolism (47). Thus, the starch and bulk density was added as predictors of regression equation in this study based on the above information. We gain regression equation of ME = −0.3869 + 0.7788 DE + 0.0336 starch + 0.0020 bulk density (R² = 0.81). The prediction accuracy of ME was improved when starch and bulk density were incorporated into the prediction model as the second and third predictive factors.

Further research will focus on elucidating the interaction effect between fiber composition and other nutrients (e.g., arabinoxylans and β-glucans).