Effects of Caffeine Doses on Rumen Fermentation Profile and Nutrient Digestibility Using a Lactating Cow Diet under Continuous Cultures Conditions

Simple Summary

This study explored the effects of different caffeine doses on nutrient digestion and fermentation in a simulated rumen environment. The main goal was to improve our understanding of how caffeine might influence these important metabolic processes in the rumen when a lactating cow diet is used. Here, we report that a dose of caffeine (50 ppm) improved the digestibility of dry/organic matter and starch while reducing the protozoa numbers and ammonia concentration, which are related to ruminal protein metabolism. However, higher caffeine doses (above 50 ppm) began to negatively impact these parameters. Our observations suggest that caffeine, a naturally occurring compound in several plants such as coffee, cacao, and tea, could potentially be used to enhance digestion and fermentation in ruminants. This research is valuable to our understanding of anaerobic fermentation in the rumen as it demonstrates the use of this plant secondary metabolite to enhance nutrient utilization and milk production and other benefits to management and the environment.

Abstract

Caffeine is a plant secondary metabolite, commonly known for its bioactivity properties. This molecule increases microbial activity during anaerobic digestion. The aim of this study was to determine the effects of caffeine doses on the rumen fermentation profile and nutrient digestibility when continuous culture fermenters were fed a lactating cow’s diet. We hypothesize that adding caffeine doses into continuous culture fermenters with a rumen fluid inoculum will not affect anaerobic fermentation or nutrient utilization. Fermenters were fed twice a day (at 0800 and 2000 h) with an experimental diet of high-producing dairy cows (53.55 g/DM day; Forage:Concentrate ratio, F:C of 40:60). Four levels of caffeine (0 ppm, 50 ppm, 100 ppm, and 150 ppm) were added to the diets as a treatment. The experiment was arranged in a randomized complete block design. Two blocks of four fermenters were run in two replicated periods of ten days. Statistical analyses were conducted in SAS version 9.4 for Windows (SAS Institute Inc., Cary, NC, USA) using the GLIMMIX procedure. The addition of caffeine at a 50 ppm dose on continuous culture fermentation decreased the protozoal counts (Diplodinium spp.) (p = 0.03) and ammonia concentrations (p < 0.05). A treatment of 50 ppm of caffeine increased the DM, OM, and starch digestibility (p < 0.05). When caffeine doses increased further than 50 ppm, the OM, DM, and starch digestibility decreased linearly (p = 0.01). The total volatile fatty acids and fatty acid proportions were unaffected. However, the acetate-to-propionate ratio (A:P) tended to decrease linearly among treatments (p = 0.07). The means of pH measurements and maximum pH had a significantly linear decrease effect (p < 0.01). Caffeine may represent a potential rumen fermentation modifier for use in lactating cow diets.

1. Introduction

Caffeine (1,3,7-trimethylxanthine) is a secondary plant metabolite that exists naturally in some plant seeds, nuts, or leaves. Plant secondary metabolites are compounds that are not part of the biochemical metabolism of plant growth and reproduction [1]. Caffeine, for example, plays an important role in plant defense; it acts as a natural pest repellent and insecticide [2,3,4], and bactericide [5,6].

By-products from the coffee and cocoa industry (with caffeine contents) have been demonstrated to increase microbial activity in a batch culture with rumen fluid, and cow manure [7]. Incremental bacterial growth and biogas production have been achieved when rumen fluid or cow manure inoculum was used in digestion studies with by-products from the coffee and cocoa industry [8,9]. The digestion of the soluble fraction of spent coffee grounds and cow manure showed a reduction in bacterial lag time in co-digestion (9 to 10 days) in comparison to control (10 to 12 days). Rapid bacterial growth occurred after a month of microbial adaptation in the co-digestion of coffee pulp and cow manure, resulting in the faster digestion of nutrients and a higher rate of CH₄ production in comparison with controls [8]. In addition, the co-digestion of cocoa shells and cow manure produced higher amounts of cumulative methane than cow manure digestion alone [9].

Previous research has cataloged caffeine, tannins, and other polyphenols in coffee pulp as anti-nutritional components that may cause low feed intake, protein digestibility, and nitrogen retention when the coffee pulp is used as animal feed [10]. Despite the previous description of caffeine as a toxic component of by-products used as animal feed, caffeine increased the biogas production and decreased the microbial lag-phase by potentially increasing microbial activity during the anaerobic digestion of the food waste and cow manure mixture (80:20 ratio) [11]. These authors evaluated, for the first time to our knowledge, the effect of pure caffeine in anaerobic fermentation with cow manure as a microbial inoculum, utilizing a batch culture method to simulate digestion. One of the purposes of our study was to evaluate the caffeine bioactivity utilizing continuous culture fermenters with a ruminal inoculum. Continuous culture fermenters allow the stratification of feed particles, producing different solid and liquid turnover rates, which mimics the passage rate of the rumen [12].

Caffeine may inhibit bacterial proliferation by interacting with DNA repair mechanisms, such as the Salt Overly Sensitive (SOS) response pathway [13]. This pathway response is a regulatory mechanism produced when bacterial cells are exposed to stress conditions and have DNA damage [14]. This response induces cell mutations and antibiotic resistance [14]. Caffeine interacts with the UmuC gene, regulated by SOS [13]. Whitney and Weir (2015; [13]) described that caffeine exposure of Escherichia coli caused persistent damage. The caffeine interaction with cell DNA has mainly been tested in Escherichia coli [13,15,16]. Escherichia coli is a Gram-negative bacterium and facultatively anaerobic. Ruminal Gram-negative bacteria primarily produce acetate and some also produce butyrate [17]. Acetate and butyrate microbial metabolism within the rumen release hydrogen, which methanogens use in combination with carbon dioxide to synthesize methane [18]. If Gram-negative bacteria in the rumen are more sensitive to caffeine than Gram-positive bacteria, caffeine exposure could potentially shift the VFA profile, decreasing acetate and butyrate, and, consequently, methane, while increasing propionate.

In ruminants, inhibiting rumen methanogenesis not only enhances rumen fermentation but also improves energy efficiency and significantly reduces greenhouse gas emissions [1]. Remarkably, methane production in ruminants leads to a loss of 6.0 ± 1.18% of their energy intake [19], diminishing the metabolizable energy utilization. From an environmental perspective, methane from enteric fermentation in ruminants accounts for approximately 12% of the total methane emissions worldwide [20]. The literature available related to the co-digestion of caffeine-containing by-products with cow manure and rumen fluid reported that the major volatile fatty acid generated was acetate [21]. Saponins and essential oils have been shown to change the rumen microbial profile [22,23], enhancing rumen fermentation and nitrogen metabolism, and decreasing methane production [22,24,25]. However, the high diversity of essential oils and other compounds present in caffeine-containing plants, such as cacao, citrus, guarana, coffee, and their linages [26], complicates the determination of which specific component affects rumen fermentation. Thus, additional research about pure caffeine effects on rumen bacteria and rumen fermentation products is necessary

These studies demonstrate that caffeine-containing by-products increase microbial activity and, consequently, enhance biomass digestion. However, no rumen fermentation study has been conducted using pure caffeine. Thus, the aim of this study was to determine the effects of pure caffeine doses on the rumen fermentation profile and nutrient digestibility when dual continuous culture fermenters were used to simulate a lactating cow diet.

2. Materials and Methods

2.1. Experimental Design and Treatments

Eight continuous culture fermenters in a generalized randomized complete block design were inoculated with rumen fluid and fed with four treatments levels. Two blocks of four fermenters were run in two replicated periods of ten days. Each period consisted of ten days: seven for microbial adaptation to the diet, and data and sample collection during the last three days. The fresh inoculum was obtained from three fistulated Holstein dairy cows at the beginning of each period. Treatments were randomly assigned to a different fermenter at the beginning of each period to avoid differences between fermenters (

Table 1. Diet ingredients and nutrient composition used to feed continuous culture fermenters.

Diet Ingredient (%DM)
Corn silage	42.8
Ground corn	17.7
Soybean meal	11.2
Citrus pulp	0.2
Beet pulp	0.2
SoyPlus	9.3
Soy hulls	17.0
Mineral mix	1.6
Nutrient Composition
% Dry Matter (DM) a	89.3
Crude Protein (CP), % DM a	17.1
Rumen Degradable Protein (RDP), % CP a	57.7
Rumen Undegradable Protein (RUP), % CP a	42.3
Neutral Detergent Fiber (NDF), % DM a	31.5
Acid Detergent Fiber (ADF), % DM a	20.1
Starch, % DM b	28.0
Sugar, % DM b	3.6
Soluble Fiber, % DM b	7.9
Fat, % DM a	3.1

^a Analysis of individual ingredients was conducted by Cumberland Valley Analytical, Waynesboro, PA, USA. ^b Estimated composition from diet simulation using individual ingredients.

). Fermenters were fed twice a day (at 0800 and 2000 h) with a total substrate of 53.55 g/DM day split evenly in AM and PM feeding. The forage source used for the experimental lactating cow diet was corn silage. The ratio of Forage:Concentrate (F:C) was 40:60 on a dry matter basis (Table 1) and the particle size of the substrate was ground uniformly to 2 mm. Four treatment levels at 0 ppm, 50 ppm, 100 ppm, and 150 ppm of pure caffeine (BAKER, Avantor Performance Materials, Allentown, PA, USA, Catalog # JTE268-7) were added to the diets as a treatment and were randomly assigned to the fermenters at the start of each period. Diets with treatment levels were prepared before the start of each period.

2.2. Continuous Culture Conditions

The fermenters’ design and operation were based on a previous model outlined by Teather et al. [12], with some modifications including the use of an overflow sidearm that angled downward at approximately 45° to facilitate emptying. All procedures involving surgical and animal care were approved by the Clemson University Institutional Animal Care and Use Committee. Around 1800 h, the rumen contents were collected from three rumen cannulated Holstein cows fed a 50% forage:50% concentrate diet and strained through two layers of cheesecloth into a prewarmed sealed container. The filtered rumen fluid was combined from all cows, mixed with a buffer in a 1:1 ratio according to the methods of [27], and purged with CO₂ until inoculation into the continuous culture fermenters. Moreover, the time from inoculum collection to fermenter inoculation did not exceed 60 min. Peristaltic pumps were calibrated and used to continuously influx 90 mL/h of buffer into the culture to maintain a liquid dilution rate from 10% to 12%. To maintain anaerobic conditions, carbon dioxide was continuously purged at a rate of 20 mL/min. Temperature was held at 39 °C by using a water bath that circulated warm water between the fermenters. Continuous pH meters measured culture values every 20 min throughout the ten-day periods and were reordered with data acquisition software (LabVIEW, Version 19.0f2; National Instruments, Austin, TX, USA). The culture pH was allowed to fluctuate during the day. The stirring rate was 45 rpm, allowing the stratification of particles (gases, middle layer of liquid and small particles, and lower layer of dense feed particles).

2.3. Sample Collection and Analysis

Sample collection occurred from day 8 to day 10 of each ten-day period. The overflow was gathered in a 2 L Erlenmeyer flask placed on ice, with 10 mL of a 50% H₂SO₄ solution added to each flask to stop fermentation, ensuring accurate measurement of feed fermentation products. The overflow volume from each fermenter was measured every 12 h to calculate the total daily volume. Containers designated for each fermenter were used to collect 20% of the overflow sample and frozen at −20 °C from day 8 to day 10. Composite overflow samples were later thawed and homogenized, and 20% of the subsample was collected to analyze the dry matter, neutral detergent fiber, acid detergent fiber, and starch. On day 10, culture sample collection was performed every two hours from time 0 (before feeding) to time 12 (after feeding) for the analysis of volatile fatty acid, ammonia, and protozoa. The pH values from the culture were also measured manually with VWR^® SympHony^® pH Meter every two hours on day 10. Samples from culture and overflow were collected on day 10 to measure dry matter. Fermenters’ impeller revolutions were increased to 100 rpm to mix the cultures and ensure the uniformity of the sample on the last day of collection times.

Diet (input) and overflow (output) samples were analyzed for dry matter; samples were oven-dried at 60 °C for 48 h and corrected for laboratory DM at 102 °C. Samples for NDF and ADF analysis were ground at 2.2 mm. Fiber analysis was based on protocol resources from Ankom Technology, Macedon, NY, USA (based on AOAC method 973.18 and 2002.4; AOAC International, 2000; [28]). Diet and overflow samples were placed in a muffle furnace at 600 °C for six hours to burn all organic material, allowing for the measurement of the total inorganic components, or mineral contents. Ash sample analyses were performed as described by the Association Official Analytical Chemist (Method 942.05, [28]). Overflow samples were centrifuged at 40,000× g for 20 min at 4 °C, and the supernatant was removed for DM analysis. The organic matter (OM) values for diet and overflow samples were determined by subtracting the ash content from the total sample amount. Dry matter from the overflow was corrected by subtracting DM from the buffer that was pumped into the cultures. Organic matter, NDF, and ADF from the overflow were corrected for DM. Digestibility of DM, OM, NDF, and ADF was calculated for each nutrient as follows:

$${\displaystyle \frac{\left(\mathrm{I}\mathrm{n}\mathrm{p}\mathrm{u}\mathrm{t}-\mathrm{o}\mathrm{u}\mathrm{t}\mathrm{p}\mathrm{u}\mathrm{t}\right)}{\mathrm{I}\mathrm{n}\mathrm{p}\mathrm{u}\mathrm{t}}}\times 100$$

where input represents the nutrient contents in the diet and output are the nutrients remaining in the overflow.

Culture samples (4 mL) were pipetted into 15 mL polycarbonate centrifuge tubes containing 1 mL of 25% (w/v) metaphosphoric acid; the mixture was vortexed, and then centrifuged at 9000 rpm for 30 min at 4 °C [29]. The supernatant was used for volatile fatty acids (VFAs) and ammonia analysis. For sample preparation before VFA analysis, 0.5 mL of supernatant was combined (in duplicates) with 0.5 mL of ultra-pure water and 0.1 mL of internal standard (100 mM 2-Ethyl-butyric acid solution) in a 1.5 mL microcentrifuge tube. Samples were slowly vortexed, and, after standing in the refrigerator for 30 min, samples were centrifuged at 12,000 rpm for 12 min and filtered with a 0.45 µm syringe filter into a gas chromatography (GC) vial. VFA profiles were analyzed by GC according to the methods of [30]. The retention time of known standards was used to calibrate GC methods for VFA sample identification.

Modifications of ammonia analysis methods described in [31] were performed to accommodate sample and reagent amounts in a 96-well microplate and to read the solution absorbance at 625 nm of wavelength. Analysis to calculate starch concentration was based on the methods described by [32]. Incubated samples were pipetted in a 96-well microplate, and absorbance was read at 500 nm of wavelength.

For protozoa sample preparation, 4 mL of culture were transferred into a 10 mL vial containing 4 mL of methyl green–formaldehyde solution and stored at 4 °C for further analysis. Preserved protozoa sample was pipetted in a counting chamber (Fuchs–Rosenthal Bright-Line counting cell, 0.2 mm depth; Hausser Scientific, Horshamm, PA, USA) and placed in a microscope for quantification. Genera identification was conducted as described by [33].

2.4. Statistical Analysis

Statistical analyses were conducted in SAS version 9.4 for Windows (SAS Institute Inc., Cary, NC, USA) using the GLIMMIX procedure. The experiment was a randomized complete block design, and response variables were analyzed using the following model:

Y_ijk = μ + C_i + P_j + CP_ij + F_k + e_ijk

where Y_ijk is the dependent variable, μ is the overall mean, C_i is the fixed effect for caffeine levels (i = 1 to 4), P_j is the fixed effect of the period (j = 1 to 2), CP is the interaction of C and P, F_k is the random effect of the fermenters within blocks (k = 1 to 8), and eijk is the residual error. The fixed effect of time, the interactions of time with the period, time with treatment, and the time with period and treatment were added in the model for repeated measures. For repeated-measures analysis, the first order of autoregressive structure, AR (1), was used. Linear and quadratic contrasts were performed to evaluate the effects of caffeine doses. Pairwise comparisons between treatments were conducted using PDIFF option. Least square means are presented in tables, and evidence for statistical significance was declared at p ≤ 0.05, while trends for main effects and interactions are discussed at 0.10 ≥ p > 0.05.

3. Results and Discussion

3.1. Caffeine Effects on Nutrient Digestibility

Table 2. Nutrient apparent digestibility values in response to caffeine doses (0, 50, 100, and 150 ppm) in continuous culture fermenters.

	Caffeine Concentrations, ppm *					p-Value
Digestibility (% DM)	0	50	100	150	SE	Linear	Quadratic
DM	65.5 a	66.3 a	64.0 ab	62.8 b	0.46	0.01	0.14
OM	72.3 ab	73.2 a	71.3 ab	70.3 b	0.37	0.01	0.12
NDF	54.8	57.5	55.7	54.0	1.20	0.51	0.12
ADF	51.1	52.5	51.7	49.4	0.92	0.28	0.13
Starch	97.9 a	98.0 a	97.8 ab	97.8 b	0.03	0.01	0.13

* Means in the same row, followed by different superscripts (^a and ^b), are significantly different (p < 0.05).

shows the nutrient apparent digestibility coefficients. Both the DM and OM digestibility exhibited a significant linear decrease (p = 0.01) as caffeine doses exceeded 50 ppm. At 50 ppm of caffeine, fermenters had the highest apparent digestibility values for NDF and acid ADF, although the differences among treatments were not statistically significant. Additionally, increasing caffeine doses led to a significant linear decrease in starch digestibility (p = 0.01).

In our study, we used pure caffeine, a compound found in coffee and coffee by-products. Prabhudessai et al. (2009; [7]) found that caffeine doses of 50 and 100 ppm significantly increased biogas production, indicating enhanced biomass digestion. Our results with the 50 ppm caffeine treatment align with these findings. This suggests a dose–response relationship to caffeine, as we observed a linear decrease in digestibility at higher doses of 100 and 150 ppm. A decrease in OM and DM digestibility was also observed in a previous study that included coffee by-products in ruminants [25]. However, this study did not test purified caffeine but used an ingredient substitution approach with a coffee grounds addition that resulted in higher fiber concentrations. The replacement of timothy and alfalfa hay with 10% and 20% of wet coffee grounds on a DM basis has been shown to linearly decrease the digestibility of OM (p < 0.01) and DM (p < 0.01) in sheep [34]. High-quality forages have a lower proportion of structural carbohydrates and more digestible organic matter content, which may increase DM and OM digestibility [35]. The content of NDF and ADF in wet coffee grounds is approximately 68.8% and 54.8%, respectively, which is higher than the amount found in alfalfa (NDF of 41.1% and ADF of 31.1%) and timothy (NDF of 60.8% and ADF of 39.6%) [34]. When high-quality forage is replaced with an ingredient with a higher amount of structural carbohydrates, and a lower amount of readily fermentative carbohydrates, the digestibility of DM, OM, and starch is expected to decrease.

3.2. Caffeine Effects on Volatile Fatty Acids, pH, and Ammonia

The total volatile fatty acid concentration, volatile fatty acid proportions, and ammonia concentration are presented in

Table 3. Volatile fatty acids and ammonia produced from continuous culture fermenters in response to caffeine doses (0, 50, 100, and 150 ppm).

	Caffeine Concentrations *					p-Value
Item	0 ppm	50 ppm	100 ppm	150 ppm	SE	Linear	Quadratic
Total VFA, mM	76.52	75.79	88.46	78.76	6.66	0.46	0.54
VFA, mol/100 mol
Acetic acid	39.36	41.45	40.10	42.84	2.17	0.32	0.87
Propionic acid	23.23	26.71	30.46	27.04	2.86	0.16	0.16
Butyric acid	21.65	18.84	18.26	18.64	2.35	0.39	0.52
Isovaleric acid	0.98	1.00	1.25	1.21	0.33	0.53	0.93
Valeric acid	11.95	10.28	8.09	8.39	1.52	0.09	0.53
Caproic acid	2.75	1.64	1.77	1.86	0.41	0.20	0.18
Isoacids	0.10	0.08	0.08	0.01	0.05	0.27	0.63
Ammonia, µM	3.98 d	2.67 a	2.90 b	3.48 c	0.08	<0.01	<0.001

* Means in the same row, followed by different superscripts (^{a, b, c}, and ^d), are significantly different (p < 0.05).

. Interestingly, while the total volatile fatty acid concentrations remained unaffected by varying caffeine doses, the valeric acid concentrations showed a tendency to decrease linearly as caffeine doses increased (Table 3; p = 0.09). This trend suggests a potential shift in the microbial profile, likely due to caffeine’s unique pro- and anti-microbial properties [7]. Similar effects on rumen microbial profiles have been observed with other plant secondary metabolites [22,23].

The change in volatile fatty acids affects microbial activity, and, consequently, the rate of ammonia utilization for the de novo synthesis of amino acids. In our study, adding caffeine at doses of 50, 100, and 150 ppm resulted in a significant quadratic increase in ammonia concentration (Table 3; p < 0.01), indicating a shift in the microbial profile. Notably, the 50 ppm caffeine dose had the most substantial impact, significantly reducing ammonia production over the 12 h of sampling compared to the fermenters with no caffeine addition (p < 0.05). Other plant secondary metabolites have been shown to decrease ruminal hyper-ammonia-producing bacteria (HAB), responsible for the deamination of amino acids to ammonia in the protein degradation pathway [36]. Daily supplementation with 10.5 mL of eucalyptus oil (80% purity) and 7.35 g of licorice root powder reduced ruminal ammonia production by up to 50% in buffalos [37]. Additionally, the increment of saponin doses (2.0 mL and 4.0 mL) on batch cultures with rumen fluid resulted in a significant decrease (p < 0.01) in ammonia levels [38]. The reduction of ruminal ammonia is correlated with improved protein efficiency and feed efficiency utilization by utilizing ammonia for the de novo synthesis of amino acids. Protein synthesis is enhanced when rumen microbes are provided with enough N and protein sources to meet their requirements [39]. Moreover, if readily available sources of carbohydrates are not present for microbes to capture N sources, microbial growth is limited and ammonia concentrations spike [40]. In our continuous culture experiment, we observed a quadratic effect for DM, OM, and starch digestibility in fermenters receiving increasing caffeine doses, with the 50 ppm treatment showing the highest apparent digestibility coefficients and the lowest ammonia concentration.

The acetate-to-propionate ratio (A:P) tended to decrease linearly among treatments (Figure 1A; p = 0.07). Interestingly, acetate slightly decreased at 6 h after treatment, while propionate increased. The A:P ratio was the lowest on fermenters receiving 100 ppm of caffeine and tended to be lower compared to the control (Figure 1B; p = 0.09). Some secondary plant metabolites have also been shown to decrease A:P without impacting the total VFA production [41]. These findings align with Prabhudessai et al. (2009; [7]), who reported that adding 100 ppm of caffeine to the anaerobic fermentation of food waste with 8% total solids (TS) resulted in the highest biogas production (408.5 mL/g TS), compared to the control (0 ppm) which produced 182.5 mL/g TS. Prabhudessai et al. (2009; [7]) concluded that caffeine potentially increases microbial activity when added to anaerobic fermentation. Furthermore, other secondary plant metabolites, such as saponins, have been shown to increase rumen propionate production after treatment by rechanneling hydrogen from methane to propionate pathways [1]. These observations related to A:P demonstrate that the fermentation profile and microbial dynamics were modified; however, the overall fermentation rate remained consistent across treatments.

Rumen fluid pH values are presented in

Table 4. Postprandial pH profile in continuous culture fermenters in response to caffeine doses (0, 50, 100, and 150 ppm).

	Caffeine Concentrations *					p-Value
Culture pH	0 ppm	50 ppm	100 ppm	150 ppm	SE	Linear	Quadratic
pH	6.07 a	6.11 a	6.00 b	5.99 b	0.02	<0.01	0.15
Maximum pH	6.37 ab	6.40 a	6.30 bc	6.28 c	0.02	<0.01	0.16
Minimum pH	5.74	5.78	5.73	5.67	0.04	0.18	0.09
pH, h < 6.0	5.00 b	4.50 b	8.00 a	7.00 a	0.35	<0.01	0.39

* Means in the same row, followed by different superscripts (^{a, b,} and ^c), are significantly different (p < 0.05).

. The means of pH measurements and maximum pH had a significantly linear decrease (p < 0.01). However, the minimum pH had a quadratic tendency (p = 0.09) with the 50 ppm treatment resulting in the highest minimum pH. A linear effect was also observed when the pH values were below 6.0 (p < 0.01), with the 0 and 50 ppm caffeine addition not being different among each other but expending significantly less time in pH below 6 than the 100 and 150 ppm treatments. The addition of spent coffee water in co-digestions with cow manure resulted in a pH reduction (pH = 5) during the first days of the fermentation period and incremented the cumulative biogas (2558 mL) in comparison with cow manure digestions (pH = 7 during the first days of fermentation; 993 mL of cumulative biogas) [42]. This pH reduction after the addition of coffee by-products is attributed to the increased acids produced by enhanced microbial fermentation. The co-digestion of coffee pulp with cow manure resulted in a lower initial pH (4.5) than the pH (6.5) of cow manure digestion alone [8]. These results highlight how coffee by-products can influence pH levels and microbial activity, enhancing fermentation and biogas production.

It is important to note that coffee or cocoa by-products have other components that may impact anaerobic fermentation, possibly by the synergistic action of caffeine and other components. For example, coffee intake in humans selectively modulates the colonic microbiota populations and increases the production of acetate and butyrate due to the presence of chlorogenic acid and other components such as caffeine [43]. However, when caffeine-containing by-products are used in anaerobic fermentation, different outcomes can be obtained due to their complex composition. Widjaja et al. (2017; [21]) reported a low rate of digestion of total solids and volatile solids during the first 30 days of co-digestion of coffee pulp with cow manure and rumen fluid. Moreover, Luz et al. (2017; [42]) reported a reduction in bacterial lag time on the co-digestion of the soluble fraction of spent coffee grounds and cow manure (9 to 10 days) in comparison to control (10 to 12 days). Thus, the easier hydrolysis of the soluble fraction of spent coffee grounds in co-digestion leads to an increment of the volatile fatty acid concentration, a lower pH, and, consequently, a faster digestion process [42].

3.3. Caffeine Effects on Protozoa Populations

Total protozoa counts were not affected by caffeine doses (

Table 5. Protozoal population in continuous culture fermenters in response to caffeine doses (0, 50, 100, and 150 ppm).

	Caffeine Concentrations *					p-Value
Protozoa 102/mL	0 ppm	50 ppm	100 ppm	150 ppm	SE	Linear	Quadratic
Total Protozoa	67.30	58.48	67.30	63.84	5.47	0.95	0.62
Entodinium spp.	24.44	25.89	26.90	27.68	3.43	0.28	0.88
Epidinium spp.	17.30	16.41	16.29	16.52	1.01	0.56	0.63
Daysitricha spp.	7.03	4.80	8.82	6.36	1.55	0.78	0.94
Isotricha spp.	1.12	0.78	2.12	1.23	0.40	0.34	0.48
Diplodinium spp.	16.96 a	9.93 b	12.83 ab	12.05 ab	1.32	0.03	0.10
Ophryosoloex spp.	0.45	0.67	0.33	0.00	0.23	0.09	0.19

* Means in the same row, followed by different superscripts (^a and ^b), are significantly different (p < 0.05).

). It is important to note that continuous culture fermenters are not capable of maintaining similar ruminal protozoa counts when compared to host animals [44]. However, in the present experiment, the numbers observed are similar to other continuous culture reports [45,46]. The addition of caffeine doses resulted in a significant linear reduction for Diplodinium spp. (p = 0.03). Moreover, the addition of the 50 ppm dose was the only that resulted in a significant difference from the control. Moreover, Ophryosoloex spp. resulted in a linear decrease trend (p = 0.09). Interestingly, the presence of Ophryosoloex spp. resulted in a mean of 0 when the caffeine doses were increased to 150 ppm.

These findings suggest that increasing caffeine doses in dairy cows’ experimental diet may inhibit the growth of Ophryosoloex spp. in rumen fluid fermentation. The effect of plant secondary metabolites on the ruminal population has been studied in search of natural methods to decrease methane production by ruminal defaunation. The addition of 2 g/d of anise extract (an essential oil) to four Holstein heifers for 21 days resulted in a decrease in protozoal counts (Entodinium spp. and Holotrichs), leading to the reduction in ruminal ammonia production [47]. Regarding the effects of saponins as well as essential oils, saponins have been reported to decrease ruminal protozoal counts and ammonia concentrations [41,48]. Ruminal protozoa play an important role in ammonia production by the proteolysis and deamination of bacterial and dietary nitrogen [17,49]. Because protozoa predates bacteria, the ruminal bacterial protein decreases, ammonia production increases and, as a consequence, the N efficiency decreases [50]. When ruminal defaunation occurs, ammonia production is decreased, and more bacteria are able to utilize ruminal nitrogen for the de novo synthesis of amino acids [49]. The decrease in protozoa counts may be a reason for the reduction in ammonia concentration in the present study and the numerical increase in total VFAs due to a possible reduction in bacteria predation.

4. Conclusions

In conclusion, the addition of caffeine affected the apparent digestibility coefficients and fermentation profile, yet not the fermentation rate. In this experiment, the addition of caffeine had linear and quadratic increases in digestibility and the fermentation profile parameters with a consistent pattern resulting in the enhancement of these parameters with the 50 ppm treatment. Adding 50 ppm of caffeine to continuous culture fermenters reduced the ammonia concentration and decreased the protozoal counts (Diplodinium spp.) in a linear fashion. This dose also improved the digestibility of DM, OM, and starch. However, increasing the caffeine dose beyond 50 ppm led to a decrease in the digestibility of OM, DM, and starch. These results showed an optimal dose for the caffeine addition of 50 ppm. Caffeine may be a potential natural additive in ruminant animals for its benefits on nutrient utilization and microbial population enhancement. Further in vitro trials may be necessary to confirm our observations, followed by in vivo trials to evaluate the caffeine effects, particularly in N efficiency, microbial profiling, VFA, and animal production.