Amazonian Fruit (Samanea tubulosa) in Dairy Cattle Diets: In Vitro Fermentation, Gas Production, and Digestibility

Simple Summary

The Amazonian tree Samanea tubulosa, locally known as the Bordão-de-velho, produces pods rich in nutrients that serve as valuable feed for cattle, particularly during periods of pasture scarcity. In this study, we evaluated how including these pods at varying levels in dairy cow diets affects in vitro ruminal fermentation and gas production. Four pod inclusion levels were tested (0, 100, 200, and 300 g/kg of dry matter). The presence of S. tubulosa enhanced fermentation efficiency and reduced gas production in lower-energy diets without affecting digestibility or methane production. These results suggest that S. tubulosa pod can be used as an alternative forage in ruminant production systems, supporting the sustainable use of native Amazonian resources.

Abstract

Edible trees, such as Bordão-de-velho (BVP; Samanea tubulosa), are being studied for their nutritional potential and the presence of bioactive compounds that influence ruminal fermentation. This study aimed to evaluate, using an in vitro assay, the effects of increasing the inclusion of BVP pods in dairy cow diets formulated with different energy levels. The experiment was conducted with eight treatments obtained from a 4 × 2 factorial arrangement, containing four levels of pod inclusion (0, 100, 200, and 300 g/kg dry matter) and two dietary energy levels (high and low). Increasing BVP levels resulted in a linear reduction in gas production from digestible organic matter in low-energy diets, without altering dry matter and organic matter digestibility or methane production. There was an increase in pH and in the acetate/propionate ratio, along with a reduction in the concentrations of short-chain fatty acids and isoacids. In conclusion, Samanea tubulosa pods improve fermentation efficiency and reduce gas production, making them a promising forage source for low-productivity animal diets.

1. Introduction

Innovative silvopastoral techniques using native forest trees are being studied to reduce deforestation [1]. Well-adapted tree species can provide several improvements, including animal shade and improved soil fertility. Moreover, several species produce fruits with high protein and energy content and are edible by ruminants [2]. Babaçu (Attalea speciosa), Baginha (Stryphnodendron pulcherrimum), Bordão-de-velho (Samanea tubulosa), Cajá (Spondias mombin), Jatobá (Hymenaea courbaril), and Jenipapo (Genipa americana) produce large amounts of pods with high nutritional content, high digestibility, and palatability [3].

Samanea tubulosa (Bordão-de-Velho) is notably important for animal nutrition in the Amazon, especially in the dry season when its fruits fall and are consumed by cattle. Nevertheless, daily intake recommendations, quality attributes, and anti-nutritional factor risks remain unclear. Bordão-de-Velho is a medium-sized legume tree with high forage potential. Its fleshy pods have sweet pulp [4] and are characterized by a high protein (149 g/kg crude protein) and energy content [3]. Additionally, BVP has secondary compounds with anti-methanogenic potentials, such as tannin and saponin [5]. Diets containing BVP may adversely affect fiber digestion, change the rumen fermentation process, and reduce methane production [6,7]. Due to considerable starch content (approximately 220 g/kg), BVP increases rumen propionic acid production [5]. For instance, Anastasook et al. [8] found that the supplementation with 60 g/kg of dry matter (DM) of BVP in the 40% roughage diet modified rumen fermentation and reduced methane emission by 9.25% in dairy steers. Use of a closely related species in the same genus, S. saman (up to 30% DM), in low-quality cross-breed diets decreased methane production linearly [5]. However, to the best of our knowledge, there is no study evaluating the increasing doses of BVP (S. tubulosa) in diets formulated to have different energy levels.

The in vitro fermentation technique is widely employed for assessing nutritional strategies to mitigate CH₄ production [9] and provides a flexible screening approach for alternative interventions [10]. We hypothesized that increased levels of BVP linearly improve fermentation profile and decrease methane production per g of digestible organic matter, especially in low-energy diets. Therefore, the present study aimed to evaluate the effects of increasing BVP in dairy cows’ diets containing different energy levels on in vitro fermentation profile, gas, and methane production.

2. Materials and Methods

The present trial was performed at the Animal Nutrition Laboratory of the Center for Nuclear Energy in Agriculture (University of São Paulo, Piracicaba, Brazil) from January to March 2022.

2.1. Treatments and Experimental Design

Experimental diets followed a 4 × 2 factorial design to evaluate diet energy levels and BVP levels. Diets were formulated according to NRC [11] guidelines to meet requirements for high- (40 kg/d milk yield, 600 kg BW, 2.5 BSC, and 150 days in milk—DIM) and low (20 kg/d milk yield, 600 kg BW, 2.5 BSC, and 150 DIM)-production dairy cows. In addition, BVP was harvested in Alta Floresta (Mato Grosso State, Legal Amazon) between October and December 2020 and chemically characterized. Increasing levels of BVP (0, 100, 200, and 300 g/kg DM) were added to diets instead of corn silage, Cynodon hay, soybean meal, and citric pulp (

Table 1. Ingredients and chemical composition of experimental diets.

Item	High-Energy 1				Low Energy 1
	0 2	100 2	200 2	300 2	0 2	100 2	200 2	300 2
Ingredients, g/kg DM 3
Corn silage	460	370	280	190	0.0	0.0	0.0	0.0
BVP 4	0.0	100	200	300	0.0	100	200	300
Cynodon hay	30.0	20.0	10.0	0.0	500	400	300	200
Ground corn	241	248	255	262	252	214	176	138
Soybean meal	214	191	168	145	140	110	80.0	50.0
Citric pulp	45.0	30.0	15.0	0.0	90.0	80.0	70.0	60.0
Soybean hulls	0.0	31.0	62.0	93.0	0.0	78.0	156	234
Mineral premix	10.0	10.0	10.0	10.0	10.0	10.0	10.0	10.0
Urea	0.0	0.0	0.0	0.0	8.0	8.0	8.0	8.0
Chemical composition, g/kg DM
Neutral detergent fiber	347	347	347	347	488	488	488	488
Non-fiber carbohydrates	416	416	417	417	304	302	299	297
Acid detergent fiber	187	192	197	202	274	286	297	309
Lignin	3.1	3.9	4.7	5.5	5.2	5.7	6.2	6.6
Crude protein	166	166	166	166	150	150	150	150
Ether extract	30	27	25	22	32	28	25	22
RDP 5	107	105	104	102	109	107	105	103
RUP 6	60.0	61.3	62.7	64.0	41.0	43.3	45.7	48.0
TDN 7	700	683	667	650	640	623	607	590

¹ Diets formulated to high- (40 kg/d) and low (20 kg/d)-production dairy cows. ² Increasing levels of Bordão-de-Velho (Samanea tubulosa) pods in the diets. ³ Dry matter. ⁴ Bordão-de-Velho (Samanea tubulosa) pods. ⁵ Rumen degradable protein calculated according to [11]. ⁶ Rumen undegradable protein calculated according to NRC (2001). ⁷ Total digestible nutrients.

). Soybean hulls were used to adjust diets’ CP and neutral detergent fiber (NDF) levels. Diets were evaluated in a randomized block design with blocks corresponded to fluid donors (n = 3).

2.2. In Vitro Assay

Six rumen-cannulated sheep were used as donors and maintained on a standard maintenance diet. Rumen fluid was collected 4 h post-feeding to maximize microbial activity and fermentation potential. The collected fluid was strained through four layers of cheesecloth, pooled in equal proportions from all donors, and maintained at 39 °C under anaerobic conditions during transport to the laboratory. The mixed inoculum was prepared by combining rumen fluid with McDougall’s buffer at a 1:2 volume ratio, as is standard in in vitro fermentation protocols [12,13]. All incubations were conducted at 39 °C in sealed bottles filled with carbon dioxide. Substrate samples (1 g DM) were weighed, inserted in filter bags (Ankom F-57, Ankom Technology, Macedon, NY, USA), sealed, and incubated in 160 mL bottles with 75 mL of buffer and inoculum solution (1:3 proportion). For each inoculum, three bottles were incubated per treatment, and three additional blank bottles per inoculum, yielding 126 bottles in total. Bottles were sealed with a rubber stopper and incubated at 39 °C for 72 h. Gas pressure was measured at 0, 2, 4, 8, 16, 24, 30, 36, 48, 60, 72 h after the start of incubation, using a pressure transducer and recorder (Pressure Press 800, LANA, CENA/USP, Piracicaba, Brazil). Gas produced for each bottle was determined according to the equation “V (mL) = 6.142 psi + 0.045” (n = 138; p < 0.01; R² = 0.99), where V is the gas volume (mL) and psi is the measured pressure. Total gas production was determined by summing measured volumes.

Gas samples for methane determination were taken after 24, 48, and 72 h. Bottles were cooled (4 °C) post-incubation to halt fermentation. Filter bags were washed with neutral detergent and ashed in a muffle oven for four hours at 550 °C to determine degraded organic matter (DOM, g/kg). The ratio between the amount of degraded organic matter (mg) and the total gas production (mL) in 48 h was used as a partition factor (PF) to estimate microbial efficiency [14]. Effects of BVP tannins on ruminal fermentation were further evaluated by incubating 1 g substrate (BVP/diet) with or without PEG 4000 [15].

2.3. Chemical Analysis

The BVP substrate was chemically characterized according to AOAC [16], measuring DM, organic matter (OM), NDF [17], acid detergent fiber (ADF), and CP. Total tannins (TT) and condensed tannins (CT) were measured by the HCl-butanol method, according to Makkar [18], with results expressed as tannic acid (TT) and leucocyanidin (CT) equivalents. Non-fibrous carbohydrates (NFC) were estimated by the equation [19].

$$NFC=1000-ash-CP-NDF-EE$$

Ammonia-N levels were determined by the micro-Kjeldahl procedure [20]. Methane and short-chain fatty acids were evaluated using a gas chromatograph system (Shimadzu GC-2010, Shimadzu Corporation, Tokyo, Japan) according to Lima et al. [21].

2.4. Calculations and Statistical Analysis

Gas production through time (0, 2, 4, 8, 16, 24, 36, 48, 60, and 72 h of incubation) was analyzed using PROC NLIN and the following model, France et al. [22]:

$$Gas production \left(\mathrm{m}\mathrm{L}\right)=A \times \left(1-e^{\left[-b\times \left(t-L\right)-c\times \left(\sqrt{t}-\sqrt{L}\right)\right]}\right)$$

(1)

where t is the time of incubation, A is the potential of gas production, b is the fractional degradation rate, and L is the lag time. Effective gas production at 48 h of incubation (µ48) was assessed considering model parameters.

The biological effect of the tannins was quantified according to the following equation:

$$Biological effect={\displaystyle \frac{gas production in presence of PEG}{gas production in absence of PEG}}$$

(2)

Data was analyzed using PROC MIXED of SAS (9.4. version, SAS Inc., Cary, NC) and the following model:

$$Y_{ijk}=\mu + E_i+ {BVP}_j+ {E\times BVP}_{ij}+ i_k+ e_{ijk}$$

(3)

With $i_k~N(0,{\sigma }_k^2)$ and $e_{ijk}~N(0,{\sigma }_e^2)$; where $Y_{ijk}$ is the observed value of dependent variable; μ is the overall mean; E_i is the fixed effect of energy level (i = 1 and 2); BVP_j is the fixed effect of BVP level (j = 1 to 4); E × BVP_ij is the fixed interaction affect among energy and BVP effect; i_k is the random effect of inoculum (k = 1 to 3); e_ijk is the random residual error; N stands for Gaussian distribution; ${\sigma }_k^2$ is the variance of each inoculum; and ${\sigma }_e^2$ is the random residual variance. The BVP levels effects were studied using polynomial contrasts to evaluate linear and quadratic effects.

3. Results

3.1. Chemical Composition of BVP and Biological Effect

The BVP samples (n = 10) presented 917 ± 9.06 g/kg DM; 907 ± 8.16 g/kg OM; 151 ± 13.5 g/kg CP; 426 ± 22.84 g/kg NDF; 258 ± 17.40 g/kg ADF; 123 ± 9.68 g/kg Lignin; 21.2 ± 2.27 g/kg NFC; 58.3 ± 6.73 g/kg TT; and 51.9 ± 7.61 g/kg CT (Table 1). There was no BVP impact (p = 0.78) on gas production (Figure 1a). Nevertheless, BVP linearly decreased (p = 0.04) tannin-induced biological effects on methane production (Figure 1b).

3.2. In Vitro Apparent Digestibility, Partition Factor, and Gas Production

The inclusion of BVP in lactating cows’ diets did not affect (p ≥ 0.66) DM and OM digestibility independent (p ≥ 0.44) of diet energy level (

Table 2. In vitro apparent digestibility, partition factor, and gas production of dairy cows’ diets containing energy levels and increasing BVP levels.

Item	Diet 1		BVP 2, g/kg DM				SEM	Probabilities 3
	High	Low	0	100	200	300		Diet	BVP	Lin.	Qua.	Diet × BVP	Diet × Lin.	Diet × Qua
DM 4 digestibility, g/kg	650	584	617	622	610	619	3.56	<0.01	0.71	0.89	0.82	0.57	0.7	0.26
OM 5 digestibility, g/kg	682	618	651	657	642	651	3.67	<0.01	0.60	0.66	0.82	0.44	0.48	0.22
Partition factor 6	4.09	3.93	3.89	4.01	3.96	4.18	0.04	<0.01	<0.01	<0.01	0.20	<0.01	<0.01	0.01
High			4.14	4.00	4.02	4.2	0.04		0.04	0.36	0.01
Low			3.64	4.01	3.9	4.16	0.04		<0.01	<0.01	0.33
GP:DOM 7, mL/g	213	221	223	216	220	210	2.22	<0.01	<0.01	<0.01	0.41	<0.01	<0.01	0.05
High			211	215	216	209	2.43		0.19	0.65	0.04
Low			235	217	223	211	2.43		<0.01	<0.01	0.41
CH4:DOM 8, mL/g	23.2	26.1	24.7	22.6	25.6	25.6	0.93	0.10	0.54	0.46	0.54	0.79	0.77	0.37

¹ Diets formulated to high (40 kg/d) and low (20 kg/d) production dairy cows. ² Increasing levels of Bordão-de-Velho (Samanea tubulosa) pods in the diets. ³ Probabilities. ⁴ Dry matter. ⁵ Organic matter. ⁶ Ratio between digestible organic matter (g) and gas production (mL). ⁷ Gas production (mL) and digestible organic matter (g) ratio. ⁸ Methane (mL) to digestible organic matter (g) ratio.

). There was (p < 0.01) BVP and energy content effect on the partition factor and gas production to digestible organic matter ratio. In low-energy diets, BVP linearly increased the partition factor and reduced the gas-to-DOM ratio. In high-energy, BVP presented a quadratic effect (p ≤ 0.04) in both variables. Furthermore, BVP and dietary energy did not significantly affect methane production per g of digestible OM.

3.3. In Vitro Fermentation

There was no energy and BVP interaction effect (p ≥ 0.44) on in vitro pH, SCFA, isoacids, and substrates for gas production and microbial growth (

Table 3. In vitro fermentation of dairy cows’ diets containing energy levels and increasing BVP levels.

Item	Diet 1		BVP 2, g/kg				SEM	Probabilities 3
	High	Low	0	100	200	300		Diet	BVP.	Lin.	Qua.	Diet × BVP	Diet × Lin.	Diet × Qua
pH	6.69	6.78	6.68	6.74	6.74	6.76	0.012	<0.01	0.11	0.03	0.37	0.42	0.12	0.74
NH3-N 4, %	0.052	0.055	0.052	0.053	0.053	0.053	0.0071	<0.01	0.02	0.01	0.69	<0.01	0.75	<0.01
High			0.048	0.055	0.053	0.051	0.0080		<0.01	0.03	0.01
Low			0.055	0.052	0.053	0.057	0.0080		0.01	0.08	<0.01
SCFA 5, mmol/L	176	162	173	174	166	165	1.6	<0.01	0.06	0.02	0.65	0.21	0.08	0.49
C2:C3 6 ratio	0.80	0.98	0.84	0.86	0.92	0.94	0.015	<0.01	<0.01	<0.01	0.96	0.09	0.02	0.70
High			0.78	0.78	0.82	0.84	0.016		0.01	<0.01	0.75
Low			0.91	0.95	1.01	1.04	0.016		<0.01	<0.01	0.81
Isoacids, mmol/L	4.33	4.12	4.37	4.48	4.18	3.87	0.087	0.07	0.01	<0.01	0.08	0.66	0.85	0.37
Substrate to gas, mg	252	237	249	250	239	250	1.7	<0.01	0.04	0.01	0.56	0.18	0.07	0.47
Substrate to microbials, mg	333	287	302	308	308	322	3.8	<0.01	0.06	0.01	0.44	0.21	0.07	0.28

¹ Diets formulated to high (40 kg/d) and low (20 kg/d) production dairy cows. ² Increasing levels of Bordão-de-Velho (Samanea tubulosa) pods in the diets. ³ Probabilities. ⁴ Amonnia nitrogen. ⁵ Short-chain fatty acids. ⁶ Acetate to propionate ratio.

). BVP linearly increased (p ≤ 0.03) in vitro pH and substrates for microbial growth, but decreased (p ≤ 0.02) SCFA, isoacids, and substrates for gas production. Analyses of acetate/propionate ratios indicated a BVP × energy level interaction (p = 0.09), attributable to a greater linear increase in low-energy than in high-energy diets. Intermediary BVP level increased NH₃-N content in high-energy diets, while decreasing it in low-energy diets.

3.4. Kinetics of Gas Production

Dietary energy level did not affect (p ≥ 0.12) kinetic parameters of gas production (

Table 4. In vitro gas production kinetics of dairy cows’ diets containing energy levels and increasing BVP levels.

Item	Diet 1		BVP 2, g/kg				SEM	Probabilities 3
	High	Low	0	100	200	300		Diet	BVP.	Lin.	Qua.	Diet × BVP	Diet × Lin.	Diet × Qua
A 4	191	186	204	199	177	173	5.2	0.34	<0.01	<0.01	0.98	0.21	0.07	0.33
B 5	0.039	0.036	0.034	0.032	0.042	0.042	0.0014	0.12	<0.01	<0.01	0.82	0.10	0.25	0.03
C 6	−0.018	−0.012	−0.022	−0.008	−0.017	−0.013	0.0036	0.17	0.21	0.46	0.29	0.42	0.87	0.19
L 7	0.164	0.097	0.115	0.145	0.137	0.125	0.0432	0.12	0.95	0.90	0.61	0.87	0.97	0.45
µ48 8	157	147	157	154	150	147	2.6	0.01	0.20	0.04	0.96	0.24	0.14	0.49

¹ Diets formulated to high (40 kg/d) and low (20 kg/d) production dairy cows. ² Increasing levels of Bordão-de-Velho (Samanea tubulosa) pods in the diets. ³ Probabilities. ⁴ Potential of gas production. ^5,6 Fractional degradation rates. ⁷ Lag time. ⁸ Effective gas production at 48 h of incubation.

; Figure 2a,b). BVP linearly increased (p < 0.01) B-fraction and reduced A-fraction. High-energy diets showed higher 48 h gas production than low-energy diets, and BVP addition linearly decreased (p = 0.04) gas volume at 48 h.

4. Discussion

BVP supplementation was hypothesized to improve the fermentation profile and decrease methane production. Observed effects included linear increases in C2:C3 ratio and partition factor in low-energy diets and decreased SCFA concentration. Although BVP decreased methane production in the bioassay, no significant effects on methane were found across treatments.

The absence of BVP effects on DM and OM digestibility agrees with previous observations by Salazar et al. [5]. Lack of BVP impact may be attributed to dietary formulation, with BVP serving as a fiber source: low-energy diets contained 500 g/kg DM and high-energy diets 490 g/kg DM of total forage plus BVP. While BVP attenuated gas production potential (A-fraction kinetics) and reduced 48 h gas volume, prior work suggests in vitro gas is closely linked to SCFA profiles and acetate/butyrate ratios [14]. Increased gas in high-energy diets follows established patterns. BVP addition decreased 48 h gas volume at progressively higher inclusion levels, likely due to tannin content [18].

Despite raising C2:C3 ratios, BVP reduced SCFA production in the present study. Higher BVP levels increased tannin concentrations and partition factors in low-energy diets, in line with Bueno et al. [23]. Tannins have been scrutinized for their anti-nutritional effects but are now actively investigated as modulators of rumen fermentation for methane reduction [24]. Multiple studies document that BVP reduces ruminal methane synthesis via condensed tannins [5,8,25]. Linear decreases in methane observed with PEG exposure illustrate BVP’s anti-methanogenic capacity, though overall methane was unchanged. Methane formation depends on hydrogen availability, and BVP-induced propionate synthesis can reduce hydrogen [26,27]. Despite reductions in gas and previous reports of methane suppression by BVP, the present results found no methane effect, possibly due to rumen microorganisms’ ability to degrade polyphenolic polymers and mitigate tannin impact [24].

BVP addition and low-energy level increased in vitro pH, which is favorable for fiber degradation and microbial proliferation [28]. Fiber is necessary for acetate production and prevents ruminal acidification [29]. In the current study, BVP replaced forage sources and linearly increased pH and acetate synthesis. These findings support greater efficacy for BVP versus other forages tested. Similar results for acetate production were noted by Argôlo et al. [30] using S. tubulosa extracts in goat diets.

The use of sheep rumen fluid is justified by established in vitro methodology, as the supporting literature confirms its broad comparability to bovine inoculum for assessing general fermentation profiles [31,32]. We acknowledge, however, that while the overall approach is valid, species-specific differences exist. These variations, especially in ammonia concentration and the microbial community structure, may influence the absolute magnitude of certain endpoints, including methane production. Readers should thus interpret the precise scale of these results with appropriate caution when extrapolating directly to bovine systems, while the relative treatment effects remain robustly assessed. During in vitro fermentation assay, no ammonia-N absorption or recycling occurs [33]. Ammonia-N reflects the net balance between microbial degradation and uptake [34]. Intermediate BVP level decreased NH₃-N content in low-energy diets, likely due to tannin protecting crude protein from ruminal degradation and ammoniacal nitrogen release, increasing duodenal protein supply [24,35]. In high-energy diets, BVP increased NH₃-N content in the in vitro fermentation, possibly reflecting reduced N uptake due to decreased SCFA production.

Although substrate-to-gas values differed among diets, maximum gas production did not, indicating that the treatments influenced fermentation partitioning rather than the overall extent of substrate degradation. The substrate-to-gas ratio reflects how fermented substrate is distributed between gaseous (CO₂ and CH₄) and non-gaseous end products (VFA and microbial biomass). Thus, diets that promoted higher substrate-to-gas ratios likely redirected a greater proportion of degraded substrate toward VFA production and microbial synthesis without altering the asymptotic gas volume. This distinction in fermentation efficiency explains why shifts in substrate partitioning occurred independently of changes in maximum gas production.

5. Conclusions

The addition of BVP in dairy cow diets decreases gas and short-chain fatty acids production, increases C2:C3 ratio, and does not affect in vitro methane production.