In Vitro Neutral Detergent Cellulase Method and Chemical Composition to Predict In Vivo Fermentable Organic Matter of Roughages

Simple Summary Various methods such as in situ, gas production and enzymatic methods are exercised to estimate the in vivo fermentable organic matter (FOM). However, each of these methods has its limitations. The in vivo method with fistulated animals for FOM determination is expensive, laborious and negatively affects animal welfare. Similarly, the in situ method also requires rumen fluid and is costly. However, enzymatic methods eliminate the need for fistulated animals and are comparatively simple, cheaper, faster, have greater repeatability, and also ensure the standardization of the process. Additionally, in situ technique can be disregarded as a standard method to test the accuracy of other techniques in cases where in vivo testing is not feasible. Therefore, in the current study, we compared the in situ nylon bag technique with the in vitro neutral detergent cellulase method and chemical composition to estimate in vivo FOM of roughages. Abstract In Vivo fermentable organic matter (FOM) reflects the energy production and the potential of rumen’s microbial protein synthesis. However, the in vivo method with fistulated animals for FOM measurement compromises animal welfare and is laborious as well as expensive. Although the alternative in situ nylon bag technique has been widely used, it is also costly and requires rumen liquor. Therefore, the present study was performed to compare the in situ nylon bag technique with the in vitro neutral detergent cellulase (NDC) method or chemical composition to estimate in vivo FOM of roughages. For this purpose, we selected 12 roughages, including six each from forages and crop residues. Our results have shown the strong correlation equations between FOMin situ and FOMNDC of forages (n = 6; R2 = 0.79), crop residues (n = 6; R2 = 0.80), and roughages (n = 12; R2 = 0.84), respectively. Moreover, there were also strong correlations between the chemical composition of roughages and FOMin situ (n = 12; R2 = 0.84–0.93) or FOMNDC (n = 12; R2 = 0.79–0.89). In conclusion, the in vitro NDC method and chemical composition were alternatives to in situ nylon bag technique for predicting in vivo FOM of roughages in the current experiment.


Introduction
In Vivo fermentable organic matter (FOM) of roughages reflects the rumen's energy production and the potential of microbial protein synthesis [1]. The determination of in vivo FOM of forages with fistulated animals is costly, time-consuming, and negatively affects animal welfare [2]. On the other hand, in the French small intestine digestible protein system, FOM is calculated from organic matter total tract digestibility (OMD) [3,4].
Previous studies have shown that the methods of in situ nylon bag, gas production, pepsin-cellulase (PC), Tilley and Terry, and neutral detergent cellulase (NDC) are well correlated with in vivo OMD [5][6][7][8][9]. Although the in situ nylon bag technique has been widely used, this method requires fistulated animals and is also expensive [5]. Other existing in vitro methods, such as Tilley and Terry [5] and gas production [7], also need rumen liquor. Compared to rumen liquor-based methods, enzymatic methods eliminate the need for fistulated animals, are simpler, cheaper, faster, have greater repeatability, and ensure the standardization of the process [1,9]. Moreover, the PC method requires regular evaluation to ensure accurate results, while the NDC method has not been updated [6] and is faster than the pretreatment of samples with pepsin [10]. Therefore, it seems that NDC method may be an attractive alternative for predicting in vivo FOM. Moreover, it was shown that in vitro digestibility determined by rumen fluid or enzymes was superior to chemical characteristics [5,11]; however, the chemical composition method is considered the simpler, faster, and cheaper method [12]. Thus, the chemical composition method has also attempted to predict in vivo FOM in the current study. Kitessa et al. [3] reported that the in situ nylon bag technique could be disregarded as a standard method to test the accuracy of other techniques in cases where in vivo testing is not feasible. Additionally, the in situ nylon bag technique with the Bang-Bang (BB) apparatus was inexpensive, portable, and easy to operate to measure the rumen degradation characteristics of feedstuffs than traditional steel chain or flexible plastic tubes for binding the bags used in the in situ nylon bag technique [13]. Therefore, in the current study, we considered the in situ nylon bag technique with the BB apparatus as the reference method.
To our knowledge, no study has performed an in vitro NDC method to predict in vivo FOM, and the correlation of chemical composition to predict in vivo FOM is also controversial and unclear. This study aimed to determine the possibility of predicting in vivo FOM of roughages using the in vitro NDC method or chemical composition. Thus, we hypothesized that the in vitro NDC method and chemical composition are suitable methods for an alternative in vivo FOM of roughages.

Sample Collection
Twelve roughage samples, including six each from forages and crop residues, were collected from five main beef cattle breeding areas in China. Specifically, the samples were langsdorff small reed (Hulunbuir City, Inner Mongolia), alfalfa, avena nuda straw, corn straw (Zuoyun County, Shanxi Province), mixed forage, oat straw, hulless barley straw, rapeseed straw (Hezuo City, Gansu Province), lolium perenne (Xiahe County, Gansu Province), poa annua (Nagqu Prefecture, Tibet Autonomous Region), alpine kobresia (Maizhokunggar County, Tibet Autonomous Region), and barley straw (Haiyan County, Qinghai Province). Mixed forage was collected from the natural meadow. Three samples were selected from each forage and crop residue for further analysis.

In Situ Nylon Bag Technique
Three Angus steers fitted with a permanent rumen cannula (450 ± 15 kg) were utilized to determine the effective rumen degradability of the organic matter of roughages with the nylon bag technique within the BB apparatus [13]. Animals were fed a total mixed ration at 8:00 h and 16:00 h according to NRC 2000 to meet the ME requirement for 1.3 × maintenance [15], and ad libitum access to water and a mineral block was provided. This animal experiment was approved by the China Agricultural University Animal Care and Use Committee (AW28059102-1, Beijing, China). The sample was weighed (4.00 ± 0.01 g) in nylon bags (80 × 140 mm) with a pore size of 37 µm, and then suspended in the rumen before morning feeding. Bags were removed after 6, 24, 48, 72, 96, 120, and 144 h of incubation. After removal, the bags were immediately immersed in cold water to stop fermentation, washed six times (1 min/rinse) in a washing machine until the water was clean, then dried at 65 • C for 48 h. Rumen degradation kinetics was determined using the following equations.
Rumen dynamic degradation rate: Effective degradability: Here, y is the rumen degradation rate at time t, a, b, c, t, and k stand for the rapidly degraded fraction (%), the slowly degraded fraction (%), the degradation rate constant at which b is degraded (%/h), incubation time, and rumen passage rate at 0.0253%/h, respectively. The effective degradability (ED) of organic matter was regarded as the rumen fermentable organic matter (FOM in situ ) [16].

In Vitro Neutral Detergent-Cellulase Plus Amylase Method
The in vitro organic matter digestibility was determined by NDC plus amylase method using small nylon bags (37 µm, 25 × 60 mm). Samples (0.50 ± 0.001 g) were transferred to small nylon bags, sealed (FR-300B, Blueberry, Shanghai, China), boiled in neutral detergent solution with heat-stable α-amylase for 75 min (ANKOM A220, ANKOM Technology Corp., Macedon, NY, USA), then oven-dried at 65 • C for 48 h. For the cellulase buffer solution preparation, we incubated 20 g cellulase [17] in a 1 L acetic acid buffer (pH 4.8) for 1 h at 40 • C. Two small nylon bags were randomly placed in 100-mL culture tubes and prewarmed overnight at 39 • C. The next morning, cellulase buffer solution (80 mL) was added into each preheated 100-mL culture tube using an automatic pump and then incubated for 24 h in a 40 ± 2 • C water at 40-60 rpm (DSHZ-300A, Jiangsu, China). Following incubation, the bags were immersed in cold water to stop fermentation, washed as in situ nylon bag technique, and oven-dried at 65 • C for 48 h.

Statistical Analysis
We analyzed the chemical composition and in situ organic matter degradation characteristics of samples using a two-tailed Student's t-test [18]. Then we compared the FOM in situ and FOM NDC of samples using the general linear model (GLM) procedure of SAS version 9.0 (SAS Institute Inc., Cary, NC, USA): y ij = µ + a i + e ij , where y ij is jth observation value at the ith level of factor a, µ is the population mean, a i is the treatment effect at level i of factor a, and e ij is the individual random residual error. The data of the prediction equations were analyzed using the linear regression procedure (REG) of SAS version 9.0 (SAS Institute Inc., Cary, NC, USA): y = b 0 + b 1 x 1 + b 2 x 2 + · · · + b m x m , where y is the dependent variable and x 1 , x 2 , x 3 , · · · x m represent m independent variables. Statistical significance was set at p < 0.05.

Chemical Compositions of Roughages
The data regarding chemical analysis for contents (%) of DM, OM, EE, CP, CF, NDF, ADF, Ash, and NFE are presented in Table 1. Compared with crop residues, EE, CP and NFE of forages were higher (p < 0.01), but CF, NDF, and ADF were lower (p < 0.01).

In Situ Organic Matter Degradation
In Situ organic matter degradation characteristics of roughages are shown in Table 2. Compared with crop residues, a (p < 0.01), b (p < 0.01), c (p = 0.01), and FOM in situ (p < 0.01) were higher in the forages.

Comparison of FOM NDC and FOM in situ
The FOM NDC and FOM in situ were presented in Table 4. For forages and crop residues, we only obtained a difference in alfalfa (p = 0.02).

Correlation Analysis between FOM NDC and FOM in situ of Forages, Crop Residues, and Roughages
The correlation analysis between FOM NDC and FOM in situ of forages (a), crop residues (b), and roughages (c) is presented in Figure 2. Our results show the strong correlation equations between FOM in situ and FOM NDC of forages (n = 6; R 2 = 0.79), crop residues (n = 6; R 2 = 0.80), and roughages (n = 12; R 2 = 0.84), respectively.

Discussion
In Vivo FOM depends on rumen dynamic processes, while in vivo OMD depends on the digestion of OM in the total digestive tract, and a reduction in the rumen can be compensated with the enhancement of the fermentation of the hindgut. Therefore, in vivo FOM's determination is more difficult than that of in vivo OMD; the measurement of in vivo FOM is also less precise than in vivo OMD. Consistently, the differences between alternative methods will probably be more pronounced when correlated with in vivo FOM than with in vivo OMD [1].
It was shown in a previous report that the in vitro method is the most inaccurate in predicting the digestibility of some low-quality feeds, such as hay and grain by-products [3]. The low-quality feed has higher contents of insoluble material, which up to some extent, determines the digestibility of a feed. When the roughage is treated with neutral detergent solution, the soluble carbohydrates, pectin, proteins, and other soluble components dissolve, leaving insoluble cell walls, which can be degraded by cellulase [10]. Therefore, in our study, we selected the NDC method. Consequently, to combine the characteristics of beef roughage, we used low-quality roughages as experimental material, which is in line with previous findings [19].
Although studies have suggested that the feedstuff's chemical composition cannot be used to predict OMD satisfactorily [3,20,21]. On the contrary, a good relationship between chemical composition and FOM in situ or FOM NDC of roughages was observed in the current study. These findings suggest that it is possible to predict in vivo FOM by the chemical compositions of a sample. Moreover, research has shown that chemical composition is related to DMI, DMI is negatively related to NDF, and NDF associated with gastrointestinal filling [22], and a strong relationship between NDF and ADF is well established, which can explain that equations based on ADF and NDF are accurate enough to predict FOM in situ or FM NDC of roughages. Additionally, the prediction of EE, CP, and CF in the same equation was more accurate than that of CF or ADF alone, suggesting that using chemical composition to predict in vivo FOM may be more correlated with fiber components. In the future, with increasing the number of samples, more accurate prediction equations are expected to be obtained.
The findings of our study showed a strong correlation between FOM NDC and FOM in situ of roughages (R 2 = 0.84). Consistently, Mary et al. [10] had documented the correlation between NDC and in vivo values and seemed to be consistent with those obtained by Tilley and Terry techniques. Moreover, that relationship was increased with the number of samples. Givens et al. [23] also reported a higher R 2 value between in vivo OMD and in vitro OMD in the spring forage than in the autumn forage. Future samples collected only once to construct the equation will not be adequate; the collection of more samples at different times of the year is suggested.
It may be noted that enzyme-based FOM measurements did not take into account possible interactions between microbial species present in the rumen. Although the nylon bag technique feed samples are digested in the actual rumen environment, many of the procedures involved in this method have not been standardized, and the range of variability within laboratories may be greater than all other in vitro methods [3]. Therefore, different methods have their advantages and disadvantages, but as long as the correlation between methods is strong, they might be replaced by each other.

Conclusions
Altogether, it was concluded that there is a strong relationship between FOM values obtained with the in situ nylon bag technique and the in vitro NDC method, which FOM in situ = 0.77 FOM NDC + 9.90 (forages, R 2 = 0.79), FOM in situ = 0.88 FOM NDC + 5.57 (crop residues, R 2 = 0.80), FOM in situ = 0.81 FOM NDC + 7.92 (roughages, R 2 = 0.84). Moreover, the FOM predicted by chemical composition correlates very well with the in situ nylon bag technique (R 2 = 0.84-0.93) and in vitro NDC method (R 2 = 0.79-0.89), respectively. Therefore, the NDC method and chemical composition seem adequate to develop equations to predict the in vivo FOM. However, more research is warranted to develop local equations considering specific pasture types, environmental conditions such as harvest season, and phenological stages, and to corroborate such predictions with in vivo data where conditions permit.