Evaluation of buffer solutions and urea addition for estimating the in vitro digestibility of feeds

Abstract

The aim of this study was to compare the in vitro digestibility of dry matter (IVDMD) and neutral detergent fiber (IVNDFD) using 2 buffer solutions with or without urea addition. The study was comprised of 2 separate experiments. In both experiments, the treatments were composed of Kansas or McDougall’s buffer solutions with or without urea addition, according to a 2 × 2 factorial arrangement. In Exp. I, the IVDMD and IVNDFD of 25 forages and 25 concentrates were evaluated. Samples were incubated for 48 h using an artificial fermenter and nonwoven textile filter bags (100 g/m²). In this experiment, the repeatability and discriminatory power among samples were calculated within forage or concentrate samples, for each treatment. In Exp. II, Tifton hay and ground corn samples were incubated for 48 h. The pH and ammonia nitrogen (NH₃-N) concentration were measured after 0, 3, 6, 12, 18, 24, and 48 h of incubation. In Exp. I, the interaction between buffer solution and urea addition impacted the IVDMD and IVNDFD of forages (P < 0.05), with greater values being exhibited when using McDougall’s buffer with urea (P < 0.05). For concentrates, the effect of buffer and urea interaction did not affect IVDMD and IVNDFD (P > 0.05). However, greater IVDMD and IVNDFD were observed for McDougall’s buffer (P < 0.05), while urea addition increased IVDMD and IVFDFD estimates (P < 0.05) regardless of buffer solution used. In general, repeatability of the digestibility was better using McDougall’s buffer and improved when urea was added. Urea addition also decreased the discriminatory power among samples for both buffers. In Exp. II, a buffer solution × urea addition × incubation time interaction was detected (P < 0.05) for pH and NH₃-N in both Tifton hay and ground corn. Kansas buffer exhibited lower pH values with a greater decrease observed throughout incubation time when compared to McDougall’s buffer. The use of Kansas buffer with urea addition was the only treatment exhibiting NH₃-N accumulation throughout incubation. In conclusion, McDougall’s buffer provides both better conditions for in vitro fiber digestion and greater precision in digestibility estimates, and is recommended over Kansas buffer. In spite of urea addition increases the precision of in vitro digestibility estimates, it decreases discriminatory power among samples.

INTRODUCTION

Buffers of variable chemical compositions are available to perform rumen in vitro digestibility assays. However, differences in chemical compositions may influence the longevity of the buffer (Herod et al., 1978) and affect the fermentation rate of substrate (Mould et al., 2005b). McDougall’s buffer is primarily composed of sodium bicarbonate as the buffering compound (McDougall, 1948). However, this buffer requires CO₂ bubbling to achieve and maintain an adequate pH, which makes its utilization more expensive and laborious compared to other buffers. Kansas buffer is based on potassium phosphate as the main buffering compound and the pH adequacy is achieved by mixing only 2 solutions (Marten and Barnes, 1980). This simpler procedure for pH adjustment highlights Kansas buffer as a viable alternative that reduces costs and laboratory manipulation. Existing in vitro studies estimating the digestibility of feeds for ruminants and nonruminant herbivores have used either McDougall’s or Kansas buffer (Adesogan, 2005; Earing et al., 2010; Roth et al., 2016). Moreover, the addition of urea in the fermentation medium aims to provide nitrogen (N) for microbial inoculum, which, in turn, could decrease the influence of the donor animal’s diet on in vitro digestibility estimates (Marten and Barnes, 1980). However, N addition may decrease the discrimination power among samples in in vitro digestibility assays, due to a more prominent increase in the digestibility of less-digestible feeds. Thus, we hypothesized that using different buffer solutions alters the estimates of in vitro rumen digestibility and that addition of N from urea to the medium may decrease the discrimination among samples with regards to digestibility. Our objective was to evaluate the discrimination power among samples for in vitro ruminal digestibility, as well as to evaluate the changes in the fermentation medium during the in vitro incubations using Kansas or McDougall’s buffer solutions with or without urea addition.

MATERIALS AND METHODS

The study was carried out in the Animal Nutrition Laboratory of the Animal Science Department at the Universidade Federal de Viçosa, Viçosa, MG, Brazil. All procedures were approved by the Ethics Committee in Animal Use and Care at the Universidade Federal de Viçosa (CEUAP, protocol no. 061/2016).

Experiment I—Obtaining Estimates of In Vitro Digestibility

Experiment I was designed to evaluate the in vitro digestibility of dry matter (IVDMD) and neutral detergent fiber (IVNDFD) of forage and concentrate feeds using Kansas or McDougall’s buffer solutions with or without urea addition.

Samples.

A total of 25 samples were evaluated in the forage group: 19 fresh [Panicum maximum cv. Mombaça (n = 1), Saccharum sp. (n = 4), Brachiaria sp. (n = 10), Cajanus cajan (n = 1), Calopogonium mucunoides (n = 1), Leucaena lecocephala (n = 1), and Neonotonia wightii (n = 1)], 5 hay [Cynodon sp. (n = 3), Stylosanthes guianensis cv. Campo Grande (n = 1), and Medicago sativa (n = 1)], and 1 silage (Arachis pintoi). Fresh forages and silage samples were oven-dried at 55 °C for 72 h. Also, a total of 25 samples were evaluated in the concentrate group: soybean hulls (n = 2), ground flint corn (n = 5), corn gluten meal (n = 1), citrus pulp (n = 1), rice bran (n = 1), sunflower meal (n = 1), soybean meal (n = 2), wheat bran (n = 1), ground sorghum (n = 2), and commercial concentrate mixtures (n = 9).

All samples were milled through a 1-mm screen sieve and stored until chemical analysis and in vitro incubations. Sample dry matter (DM; method INCT-CA G-003/1), crude protein (CP; method INCT-CA N-001/1), and neutral detergent fiber (aNDF; method INCT-CA F-002/1) contents were analyzed according to the standard analytical procedures of the Brazilian National Institute of Science and Technology in Animal Science (Detmann et al., 2012; Table 1). The NDF was analyzed using heat-stable amylase (Termamyl 2X, Novozymes, São Paulo, Brazil), omitting sodium sulfite and expressed inclusive of residual ash. More details about individual samples are presented as an e-supplement.

Table 1.

Descriptive statistics of crude protein (CP) and neutral detergent fiber (NDF) concentrations (% dry matter) of forage and concentrate samples used in the Exp. I

Group	Statistics
	Minimum	Maximum	Mean	Standard deviation
Forages (n = 25)
CP	2.08	20.1	7.02	4.05
NDF	24.7	71.0	58.3	11.4
Concentrates (n = 25)
CP	4.87	59.0	17.3	13.2
NDF	10.3	62.6	23.2	11.3

Treatments.

Treatments included combinations of McDougall’s (McDougall, 1948) or Kansas (Marten and Barnes, 1980; Silva and Queiroz, 2002) buffer solutions, with or without urea addition, according to a 2 × 2 factorial arrangement. Buffer solutions were prepared as described by McDougall (1948) and Silva and Queiroz (2002) (Table 2). In brief, for McDougall’s buffer solution, pH adjustment was performed by bubbling CO₂ until the buffer pH reached 6.8. For Kansas buffer solution, which was composed by 2 solutions (A and B), pH was adjusted to 6.8 by mixing the 2 solutions at ratios of 5:1 and 7.7:1, with and without urea addition, respectively (Table 2). For McDougall’s buffer, urea addition was conducted by adding 5 mL of urea solution (55 g/liter) in 300 mL buffer solution prior to pH adjustment (Silva and Queiroz, 2002). For Kansas buffer, urea was added (0.5 g/liter) during solution A preparation (Silva and Queiroz, 2002). All procedures were completed 24 h prior to each in vitro incubation. The buffer solutions were stored into a climate-controlled room (39 °C) until incubation began.

Table 2.

Final composition of the evaluated buffer solutions with or without urea addition

Component, g/liter	Buffer solution
	McDougall		Kansas1
	Without urea	With urea	Without urea	With urea
NaHCO3	9.80	9.80	–	–
Anhydrous Na2HPO4	3.71	3.71	–	–
KCl	0.57	0.57	–	–
NaCl	0.47	0.47	0.42	0.44
Heptahydrate MgSO4	0.12	0.12	0.42	0.44
Dihydrate CaCl2	0.05	0.05	0.08	0.09
Anhydrous KH2PO4	–	–	8.33	8.85
Na2CO3	–	–	2.50	1.72
Nonahydrate Na2S	–	–	0.17	0.11
Urea	–	0.90	–	0.50

¹Reported values are final concentrations. Kansas buffer solution was composed by mixture of 2 solutions (A and B). Solution A: anhydrous KH₂PO₄ (10.0 g/liter), heptahydrate MgSO₄ (0.5 g/liter), NaCl (0.5 g/liter), and dihydrate CaCl₂ (0.1 g/liter). Solution B: Na₂CO₃ (15.0 g/liter), and nonahydrate Na₂S (1 g/liter). In the treatment with urea addition, 0.5 g/liter of urea was added in the solution A. Differences in composition with or without urea are due to different proportions of solutions A and B (5:1 and 7.7:1, respectively) for the adequate pH adjustment.

In vitro incubations and digestibility calculations.

Six 48-h in vitro incubations were conducted, with the first and last 3 incubations assigned for forage and concentrate evaluations, respectively. In vitro digestibility assays were conducted using a TE-150 artificial fermenter (Tecnal Equipamentos Científicos, Piracicaba, SP, Brazil) (Silva et al., 2017). Samples were incubated using nonwoven textile filter bags (100 g/m²) of 4 × 4.5 cm (Valente et al., 2011). In each incubation period, four 500-mg aliquots per sample were weighed and stored in heat-sealed filter bags.

To reduce variation among incubations, only 1 animal was used as a rumen inoculum donor. A rumen cannulated heifer, fed a sugarcane and concentrate (220 g CP/kg DM)-based diet with a forage-to-concentrate ratio of 80:20 was used. The heifer had free access to water and mineral mixture (90 g/kg of phosphorus), and was adapted to the diet for 14 d prior to rumen inoculum collections (Machado et al., 2016). The ruminal inoculum (liquid and solid) collections were performed at several points in the rumen shortly before the beginning of each incubation. Ruminal inoculum was stored in preheated (39 °C) thermal bottles and then mixed for a few seconds using a blender to homogenize liquid and solid phases. The fluid was then filtered through 4 layers of cheesecloth. The steps from rumen inoculum collection and incubation onset were conducted within 20 min in a climate-controlled room (39 °C).

The artificial fermenter possessed 4 jars (3,200 mL), and each jar randomly received 1 treatment. Moreover, each jar received 25 filter bags with samples and 1 blank filter bag. In each jar, 400 mL of ruminal inoculum and 1,600 mL of buffer solution were added. Carbon dioxide was flushed into the head space of each jar, which was closed and placed into the preheated (39 °C) artificial fermenter. After 48 h of incubation, the filter bags were washed using hot distilled water (90 °C) until the water became clear, and bags were gently pressed to remove gases. Then, to estimate the apparently undigested DM residue, filter bags were oven-dried (55 °C/24 h and 105 °C/16 h, sequentially), placed in a desiccator, and weighed.

For the IVNDFD evaluations, filter bags containing the incubation residues were placed into polypropylene screw-capped flasks (120 mL; autoclavable universal collection vial, Bioplast 2605, Porto Alegre, RS, Brazil) with 80 mL of neutral detergent solution and 500 µL of a heat-stable amylase (Termamyl 2X, Novozymes, São Paulo, Brazil). Flasks containing the filter bags were closed and autoclaved for 1 h at 105 °C (Barbosa et al., 2015). After that, the filter bags were washed using hot distilled water (90 °C) and acetone. The drying and weighing procedures were performed as previously described.

The IVDMD and IVNDFD (D, g/kg) were calculated as:

$$D=\frac{M-(R-B)}{M}\times 1,000,$$ (1)

where M is the incubated mass of DM or aNDF (g); R is the undigested residue of DM or NDF (g); and B is the DM or aNDF residue in blank filter bags (g).

Experiment II—Monitoring Buffering Capacity

Experiment II was designed to evaluate the buffering capacity of McDougall’s and Kansas buffer solutions with or without urea addition.

Samples.

Samples of Tifton 85 (Cynodon sp.) hay (113 g CP and 687 g NDF/kg DM) and ground flint corn (49.1 g CP and 197 g NDF/kg DM) were used. Both samples were chosen for being feeds with high aNDF and starch content, respectively.

Treatments.

As described for Exp. I, the treatments involved combinations of McDougall’s or Kansas buffer solutions, with or without urea addition, according to a 2 × 2 factorial arrangement (Table 2).

In vitro incubations and measurements.

One in vitro incubation was performed for 48 h according to the recommendations of Tilley and Terry (1963), with some adaptations. In brief, bottles were sealed and gas was released at 3, 6, 9, 12, 18, 24, 30, 36, and 42 h of incubation with needle instead of Bunsen valves. For each treatment, twenty-one 250-mg aliquots per sample were weighed into 60 mL penicillin bottles. Then, 10 mL of ruminal inoculum and 40 mL of buffer solution were added into each bottle. The head space was then flushed with CO₂ and bottles were sealed with rubber caps and aluminum seals. The bottles were kept on a stirring table (40 rpm) in a climate-controlled room (39 °C).

The pH and ammonia-N (NH₃-N) concentrations were measured at 0, 3, 6, 12, 18, 24, and 48 h of incubation. Each time, 3 bottles from each treatment and sample were randomly selected. The content of each bottle was transferred to 50-mL conical polypropylene screw top tubes, and pH was measured using a digital potentiometer (Quimis, São Paulo, Brazil). Then, 1 mL of a H₂SO₄ solution (500 mL/liter) was added into each tube, and all tubes were kept at 4 °C until NH₃-N analysis. Ammonia N was estimated by colorimetric reactions catalyzed by indophenol (method INCT-CA N-006/1; Detmann et al., 2012).

Statistical Analysis

For Exp. I, statistical analysis was performed separately for forages and concentrates using the MIXED procedure of SAS 9.4 (SAS Inst. Inc., Cary, NC), according to the model:

$$Y_{ijklm}=\mu +A_i+B_j+T_k+U_l+T{U}_{kl}+{\epsilon }_{ijklm},$$ (2)

where µ is the general constant; A_i is the random effect of sample i; B_j is the random effect of incubation j; T_k is the fixed effect of buffer solution k; U_l is the fixed effect of urea addition l; TU_kl is the fixed effect of interaction between buffer solution and urea addition; and ε_ijklm is the random error.

For outlier evaluation, the INFLUENCE statement of MIXED procedure was used and residuals (observations) with restricted likelihood distance higher than 1 were removed. Then, another variance analysis was performed following the previously described model.

To evaluate specific variances of IVDMD and IVNDFD results, the estimates obtained for forages and concentrates were analyzed separately for each treatment, following the model:

$$Y_{ijk}=\mu +A_i+B_j+{\epsilon }_{ijk},$$ (3)

where µ is the general constant; A_i is the random effect of sample i; B_j is the random effect of incubation j; and ε_ijk is the random error.

Estimates of the residual variance (${\widehat{\sigma }}_{\epsilon }^2$) and the variance among samples (${\widehat{\sigma }}_a^2$) were obtained by the restricted maximum likelihood method. The standardized repeatability (r, lower is better) was calculated as:

$$r=\frac{\sqrt{{\widehat{\sigma }}_{\epsilon }^2}}{\overline{Y}}\times 100,$$ (4)

where r is the standardized repeatability (%); ${\widehat{\sigma }}_{\epsilon }^2$ is the residual variance; and $\overline{Y}$ is the average digestibility.

The relative standard deviation among samples (SDas) was used as an indicator of the power of discrimination among samples (higher is better) within treatments (or methods), which was calculated as:

$$SDas=\frac{\sqrt{{\widehat{\sigma }}_a^2}}{\overline{Y}}\times 100,$$ (5)

where SDas is the relative standard deviation among samples (%); ${\widehat{\sigma }}_a^2$ is the variance among samples; and $\overline{Y}$ is the average digestibility.

For pH values and NH₃-N concentration in Exp. II, an analysis of variance was performed for each feed using the MIXED procedure of SAS 9.4, according the model:

$$Y_{ijkl}=\mu +T_i+U_j+T{U}_{ij}+t_k+t{T}_{ki}+t{U}_{kj}+tT{U}_{kij}+{\epsilon }_{ijkl},$$ (6)

where µ is the general constant; T_i is the fixed effect of buffer solution i; U_j is the fixed effect of urea addition j; t_k is the fixed effect of incubation time k; TU_ij, tT_ki, tU_kj, and tTU_kij are the fixed effects of interactions between buffer, urea addition, and time; and ε_ijkl is the random error.

When a significant interaction was detected, the results were analyzed using the SLICE statement of the MIXED procedure. In all procedures, 0.05 was used as the critical level for type I error occurrence.

RESULTS

Experiment I—In Vitro Digestibility, Repeatability, and Discrimination Among Samples

There was an interaction between buffer solution and urea addition on forage IVDMD (P < 0.001) and IVNDFD (P = 0.037), in which urea addition increased both IVDMD and IVNDFD when McDougall’s solution was used, while no effects of urea were observed with Kansas buffer (Fig. 1). For concentrate samples, the interaction between buffer and urea addition was not observed for either IVDMD (P = 0.109) or IVNDFD (P = 0.493). McDougall’s buffer presented greater (P < 0.01) IVDMD and IVNDFD when compared to Kansas buffer (Fig. 1). On the other hand, urea addition increased (P < 0.01) both IVDMD and IVNDFD in concentrate samples (Fig. 1). There was no effect of the different incubations (P > 0.05) on IVDMD and IVNDFD for both forages and concentrates.

Figure 1.

In vitro digestibility of dry matter (IVDMD) and neutral detergent fiber (IVNDFD) of forages or concentrates using Kansas or McDougall’s buffer solutions with or without urea addition.

Values for the r and SDas of forages and concentrates are presented in Fig. 2. Generally, for both forage and concentrate samples, McDougall’s solution exhibited better repeatability values for IVDMD and IVNDFD. Better r values for forage IVDMD and IVNDFD were observed by adding urea to both buffers. For concentrate, urea addition decreased r in all cases except IVDMD using McDougall’s buffer. The SDas for forages were similar among treatments for IVDMD and IVNDFD. For concentrates, SDas were quite similar among treatments for IVDMD, with higher values for IVNDFD using Kansas buffer. Generally, urea addition decreased SDas for IVNDFD.

Figure 2.

Standardized repeatability and relative standard deviation among samples (SDas) of in vitro digestibility of dry matter (IVDMD) and neutral detergent fiber (IVNDFD) in forages or concentrates using Kansas (K) or McDougall’s (M) buffer solutions with (U) or without urea addition.

Experiment II—Buffering Capacity and Ammonia Nitrogen Concentration

A buffer solution × urea addition × incubation time interaction was detected (P ≤ 0.04) on pH and NH₃-N (Fig. 3) for both Tifton hay and ground corn. For both feeds, Kansas buffer had lowest pH values (P < 0.05). The addition of urea only increased pH (P < 0.05) in Kansas buffer. Regardless of feed, decreases in pH were more profound during incubation in Kansas than McDougall’s buffer (P < 0.01). After 3 h of incubation, the NH₃-N concentrations were greater for Kansas buffer with added urea, reaching 29.6 ± 1.69 mg/dL at 48 h of incubation. For the other treatments, NH₃-N accumulation was not observed for either Tifton hay or ground corn, and NH₃-N values at 48 h of incubation were lower than 15 mg/dL.

Figure 3.

In vitro pH and ammonia nitrogen (NH₃-N) concentration at different incubation times of Tifton hay or ground corn incubated using Kansas (K) or McDougall’s (M) buffer solutions with (U) or without urea addition (SEM for pH: 0.0258 for Tifton hay, and 0.0434 for ground corn).

DISCUSSION

Digestibility Estimates and Buffers Characteristics

The significant effect of interaction between buffer solution and urea addition on IVDMD and IVNDFD for forage samples reinforces the interference of fermentation medium conditions in digestibility (Mould et al., 2005a; Detmann et al., 2008). However, this interaction did not affect IVDMD and IVNDFD for concentrate samples. The N and carbohydrates availability for microbial growth are the main factors affecting the in vitro digestibility estimates (Mould et al., 2005b). Under in vitro conditions, these substrates are primarily derived from the incubated diet or feed and buffered rumen fluid. Moreover, forages typically present lower nutrient availability than concentrates, making microbes more dependent on buffered rumen fluid substrates. Since NH₃-N is the primary N substrate for rumen fibrolytic bacteria (Russell, 2002), the added urea likely provided greater NH₃-N availability for fiber digestion (Costa et al., 2008; Detmann et al., 2009). The absence of an increase in forage IVNDFD when adding urea to Kansas buffer seems to be a consequence of its lower buffering capacity and pH values (Fig. 3). Although urea resulted in increased pH, likely due to the production of ammonia bicarbonate from CO₂ in the fermentation medium, its addition did not avoid a greater pH decrease along incubation times when included in Kansas buffer. For concentrates, the greater allowance of fermentable substrates from the feed, coupled with a greater N availability, can lead to a greater microbial activity, which explains the increased IVDMD and IVNDFD observed upon adding urea to both buffer solutions.

Overall, the results demonstrate greater IVDMD and IVNDFD for McDougall’s solution than for Kansas buffer, which indicates greater microbial activity when using McDougall’s buffer. Moreover, those effects were more prominent for IVNDFD than IVDMD, allows us to infer that differences in digestibility estimates are more related to constraints on fiber digestion. Furthermore, Kansas buffer exhibited lower pH than McDougall’s buffer, which can significantly impair fiber digestion. Although the majority of pH values were above 6.0, which represents a critical low limit recommended for proper fibrolytic activity (Hoover, 1986), the Kansas buffer experienced a greater pH decrease over time, resulting in values below or approaching 6.0 after 24 h of incubation for both Tifton hay and ground corn. Congruent with the results of the present study, previous in vitro studies have demonstrated the lower buffering capacity of phosphate (e.g., Kansas) over bicarbonate-based buffers (e.g., McDougall) (Turner and Hodgetts, 1955; Herod et al., 1978; Kohn and Dunlap, 1998). It should be also emphasized the McDougall’s buffer is more similar to a in vivo environment than Kansas buffer, as it is based on the chemical composition of ruminant saliva (McDougall, 1948).

Within the observed pH range, 2 additional buffer characteristics that could affect pH and buffering capacity should be also considered. First, the effect of CO₂ bubbling for pH adjustment in McDougall’s buffer is important, as carbon dioxide bubbling is a common step in buffer solution preparation (McDougall, 1948; Pittman and Bryant, 1964; Menke and Steingass, 1988). This is important not only for pH adequacy (McDougall, 1948; Van Soest and Robertson, 1985), but also for the removal of dissolved oxygen (Bryant, 1972), which could increase redox potential of buffered rumen fluid and compromise microbial growth (Van Soest, 1994). Moreover, rumen microbes cannot grow without CO₂, which is important for carboxylation reactions of amino acid synthesis (Van Soest and Robertson, 1985). Therefore, the absence of CO₂ bubbling in Kansas buffer can be an additional limitation. Second, the presence of sodium sulfide in Kansas buffer, which has a strong redox power, could negatively affect anaerobic microbial growth. This argument is supported by the formation of toxic intermediate compounds (e.g., sulfur compounds) and the precipitation of metal ions (Fukushima et al., 2003).

For forages, urea increased digestibility, particularly for IVNDFD when using McDougall’s buffer. However, NH₃-N accumulated over time when urea was added to the Kansas buffer, which was not observed for McDougall’s buffer. Such a pattern indicates a limited NH₃-N uptake by microbes when Kansas buffer was used. In other words, urea hydrolysis did occur. However, N assimilation by microorganisms was limited, which reflected no increase in microbial activity and IVNDFD when using Kansas buffer.

Standardized Repeatability and Relative Standard Deviation Among Samples

An adequate buffer solution should provide lower r values and greater SDas, which indicates greater precision and discrimination power among samples, respectively. In this sense, lower overall r values for McDougall’s solution over Kansas buffer, and similar SDas between them, highlight McDougall’s solution as the best option. The only notable exception was the greater SDas for concentrate IVNDFD when using Kansas buffer; however, this could be the result of its limited buffering capacity, mainly with concentrates.

The observed reduction in r could be viewed as a positive effect of urea in laboratory procedures. In this sense, the use of McDougall’s buffer with urea addition had predominantly lower r values, indicating that random variations among replicates decreased, leading to more precise IVDMD and IVNDFD estimates. However, urea addition predominantly decreased SDas for both buffers. In spite of addition of N compounds to in vitro fermentation medium has been suggested (Pittman and Bryant, 1964; Silva and Queiroz, 2002), some authors have pointed out a proper rumen inoculum should provide sufficient N compounds for microbial growth (Tilley and Terry, 1963), and N addition should not be recommended for estimating the in vitro digestion kinetics of sole feeds diets (Mertens, 2005). In this way, the donor animal diet cannot have limited N concentration in the diet. As observed in the present study, N addition to in vitro assays could reduce discriminatory power among samples. This is most likely due to increased digestibility estimates of lower-quality feeds, and a consequent decrease in the differences between high and low digestibility feeds. This can be a negative effect, primarily when in vitro assays aim to compare feeds with regards of IVDMD and IVNDFD. Thus, urea addition may represent an adverse factor for this type of in vitro study.

CONCLUSIONS

McDougall’s buffer provides a better in vitro environment for fiber digestion and is recommended over Kansas buffer for more precise estimation of feed digestibility. Moreover, urea addition increases the precision of in vitro digestibility estimates, but decreases the discriminatory power among samples.