Identication of Bioactive Phytochemicals from Six Forest Plants: Insights into the Mechanistic Inhibition of Rumen Protozoa, Ammoniagenesis, and Α-Glucosidase

Rumen protozoa have a little contribution to the feed digestibility but Entodinium, the most predominant genus, is unfortunately culprit of the nitrogen utilization ineciency. To bridge the gap, antibiotics have been used to inhibit the rumen protozoa but unfortunately, due to the health concern, this could not be really applied at the farm level, especially in the organic farms where the use of natural plants is an utmost priority. Therefore, our study aimed at investigating the nutritional and functional properties of six forest plants for their potential as feed additives in animal husbandry. The plants were analyzed for major phytochemicals using reversed phase-HPLC analysis and then evaluated for their in-vitro suppressing effect on rumen protozoa, ammoniagenesis, and microbial α-glucosidase activity. For each plant, four doses (0, 0.7, 0.9, and 1.1 mg/mL culture uid) displayed in a complete randomized design were used. Quercetin, anthraquinone, 3-hydroxybenzoic acid, astragaloside, and myricetin were found to different extent (P ≤ 0.05) in the plant leaves. All the tested plants reduced total rumen protozoa counts but C. gigantea and B. rapa had the most inhibitory effect (P ≤ 0.05), inhibiting the rumen protozoa by 45.6 and 65.7%, respectively, at the dose of 1.1 mg/mL. Moreover, the scanning electron microscopy revealed the mechanistic disruption of the extracellular membrane of the protozoa, indicating their metabolic death pathway. Only C. gigantea inhibited the rumen protozoa in a proportion that also led to the reduction of the wasteful ammonia production (P ≤ 0.05). Besides, A. digitata and F. macrophylla have the higher inhibition rate (70%) of the microbial α-glucosidase activity at 100 µg/mL of crude extract. Overall, the plants showed promising results as functional feed additives although future research on bio-guided fractionation is needed to accurately identify the pure anti-protozoal bioactive compound(s).


Introduction
Ruminants can convert brous plant materials into meat and milk products for human consumption. However, feed e ciency remains a big challenge, especially nitrogen utilization e ciency (NUE), which is no more than 30% as most of the dietary nitrogen is excreted as urea and ammonia 1 , negatively in uencing the environment and increasing feeding cost. In particular, rumen protozoa engulf the cells of rumen microbes (primarily bacteria), which are the main source of metabolized protein for the host animals synthesized from the diet, and degrade the microbial protein into oligopeptides and free amino acids, both of which are converted to ammonia, lowering NUE 2 . To bridge the gap, antimicrobials have been used to improve feed utilization e ciency including NUE, but unfortunately, some side effects on both the animal products and human consumers have been reported 3. This situation led to the ban of antimicrobials as feed additives and calls for a renewed interest in using plants as a source of natural alternatives, especially in organic animal husbandry. A recent study 4 demonstrated that the leaves of Calotropis gigantea, which contains phytochemical groups, such as phenolics, avonoids, and alkaloids, could effectively inhibit rumen protozoa (by nearly 60%) and thereby decreasing the wasteful ammoniagenesis in vitro. More importantly, C. gigantea did not adversely affect fermentation characteristics. By far, no study has described the mechanism of inhibition of rumen protozoa by bioactive phytochemicals.
Using scanning and transmission electron microscopy, Park et al. 5 showed that antibiotics and inhibitors of protozoal digestive enzymes could destruct the cellular structure of Entodinium and other rumen protozoa both extracellularly and intracellularly.
High-producing ruminants are generally fed with diets rich in rapidly fermentable carbohydrates leading to an upsurge of volatile fatty acids (VFA) and consequently a dramatic decrease of the rumen pH, which in turn causes ruminal acidosis and associated metabolic disorders and depressed milk yields 6 . The rapid degradation of carbohydrates or polysaccharides is due to the enzymatic activity of microbial amylase and glucosidase, which needs to be regulated to mitigate rumen acidosis. To that end, acarbose has been used as a synthetic inhibitor of α-glucosidase to raise the rumen pH, but unfortunately, like other synthetic drugs, acarbose was associated with serious side complications 7 . In addition, the use of acarbose to alleviate ruminal acidosis also led to the increase of ammonia nitrogen (NH 3 -N) concentration 8 , which decreases NUE. Moreover, as a drug treat diabetes, acarbose is too expensive to be fed to ruminants. It becomes urgent to explore natural alternatives from plants as alternatives to inhibit microbial α-amylase and α-glucosidase enzymatic activity and lower the risk of rumen acidosis.
Some forest plants, such as the fruit pulp of Adansonia digitata (baobab), have been shown to inhibit α-glucosidase activity 9 . Those plants contain various bioactive and antimicrobial phytochemicals, and yet most of them have not been explored for their potential as feed or feed additives for ruminants. In a recent study 4 , we found that C. gigantea could markedly inhibit rumen protozoa and decrease NH 3 -N production in vitro. The objective of the present study was to analyze another six plants in comparison with C. gigantea for their potential to inhibit rumen protozoa, decrease ammonia production, and regulate the activity of α-glycosidase using in vitro rumen cultures. We hypothesized that some of those plants, such as Flemingia macrophylla (waras tree), Kalimeris indica (Indian aster), Brassica rapa subsp. chinensis (bok choy), Portulaca oleracea (common purslane), and C. gigantea, none of which had not yet been tested in animal nutrition, could be used as a source of protein and ber in typical diets of ruminants and also at the same time, as inhibitors of detrimental protozoa, ammoniagenesis, and starch digestion. This study aimed to test the above hypothesis by exploring the above forest plants through evaluating their inhibitory effect on total rumen protozoa and α-glucosidase and identifying their major phytochemicals for the mechanistic understanding of their effects.

Results
Proximate composition and total polyphenols. Crude protein, ether extract, ber, and ash were the main nutrients detected on a dry matter (DM) basis (Table  1). K. indica recorded the highest protein content, while F. macrophylla roots the lowest. Regarding the NDF/ADF content, F. macrophylla and C. gigantea had a higher value than the other plants. The ether extract yield was quite low for all the tested plants, and C. gigantea had the highest value. Ash content was found higher in P. oleracea and C. gigantea than in the other plants. The roots of F. macrophylla had the lowest ash content.
The plants had different yields of extracts and different content of phenolics, avonoids, and alkaloids. B. rapa subsp. chinensis (the tubers) had the highest crude extract yield, while F. macrophylla and P. oleracea recorded the lowest. Phenolic content was found the highest in F. macrophylla, followed by A. digitata and C. gigantea. B. rapa subsp. chinensis (tubers) and P. oleracea had the lowest phenolic content. The highest total avonoid content was found in C. gigantea and the lowest in B. rapa subsp. chinensis (tubers). Except for F. macrophylla that had the highest total alkaloid content, the other plants recorded the lowest proportion of alkaloids compared to the phenolic and avonoid content.
Quanti cation of some individual phytochemical compounds. Among the studied plants, A. digitata appeared to rich in 3-hydroxybenzoic acid, astragaloside, and myricetin as it had the highest content of those individual compounds, while K. indica was rich in 3-hydroxybenzoic acid, C. gigantea rich in quercetin, F. macrophylla, P. oleracea, and B. rapa subsp. chinensis rich in astragaloside (Table 2). F. macrophylla roots also contain a high level of 3-hydroxybenzoic acid.
Effects on rumen protozoal counts and ammoniagenesis of in vitro culture. All the tested plants inhibit the total protozoa counts though to a different extent ( Fig. 1). At the tested maximum dose of 1.1 mg/mL, B. rapa subsp. chinensis had a higher inhibitory effect with a 65.7% reduction than A. digitata with a 47.6% reduction and nally C. gigantea with a reduction by 45.6% (P< 0.05). K. indica, F. macrophylla, and P. oleracea suppressed the protozoal count by 24%, 9.5%, and 8%, respectively. Only C. gigantea and A. digitata decreased protozoa at the lowest dose tested (0.7 mg/mL) (P< 0.05). B. rapa subsp. chinensis, K. indica, P. oleracea, and A. digitata did not affect NH 3 -N production (P > 0.05) ( Table 3) but C. gigantea reduced the production of NH 3 -N, whereas F. macrophylla (leaves and roots) increased the ammoniagenesis (P< 0.05).
Disruption of the cells surface of Entodinium by the tested plants. The extracellular surface (pellicles) of Entodinium cells collapsed and wilted by the tested plants (Fig. 2). The normal longitudinal striations present on the cell surface disappeared after the exposure to B. rapa subsp. chinensis, K. indica, and A. digitata, but the treatments with C. gigantea, F. macrophylla or P. oleracea or the control did not destruct the striations.
Antidiabetic potential of the plants leaves. All the tested plants inhibited α-glucosidase activity, but different plants exhibited different effects, with B. rapa chinensis maintaining similar inhibition as its dose increased, while the other plants lost inhibition potency with the increased doses (Fig. 3). More intriguingly, C. gigantea, F. macrophylla, and K. indica stimulated the activity of α-glucosidase at their higher doses. Only F. macrophylla and A. digitata inhibited half of the α-glucosidase concentration, at the lowest amount of required extract (IC50 value). As expected, acarbose inhibited α-glucosidase in a dose-dependent manner. It remains to be determined if α-glucosidase is stimulated by other plant compounds when they reach their threshold concentration when su cient plant leave extracts are added.

Discussion
Many synthetic products are still used either as inhibitors of certain gut microbes and their enzymes or as growth promoters at different levels to control or regulate some physiological functions of food-producing animals. Unfortunately, those synthetic products are nowadays recognized to have adverse effects associated with global public health concerns. Therefore, many of those chemicals are being banned, and natural alternatives are highly sought after. This study investigated for the rst time some forest plants harvested in China, which have not yet been explored as modulators of ruminant nutrition. The study provided information on the basic nutritional composition, the major bioactive phytochemicals, and the inhibitory effect on total rumen protozoa, ammoniagenesis, and α-glucosidase activity of the selected plants.
The content of protein, NDF, ADF, and ash of K. indica, P. oleracea, B. rapa subsp. chinensis (tubers) were reported for the rst time in this study. Abiona et al. 10 found a protein content of 13.6% in A. digitata leaves collected in the state of Oyo in Nigeria. F. macrophylla foliage collected from the southern part of Vietnam contained 16% proteins, 64.7% NDF, and 53.4% ADF 11 . A. digitata, K. indica, C. gigantea, and P. oleracea had a protein content greater than 16%, and thus they could be used as a potential source of protein in the ration fed to dairy cows. Moreover, the fat-rich plants are involved in the insulation of organs for the maintenance of body temperature and cell function and they could be explored as sources of omega-3 and omega-6 fatty acids, which can promote health in both animals and humans, digestion, absorption, and transport of vitamins 12 . Flavonoids are a group of phenolic compounds widespread in plants and highly regarded for their health-promoting and disease-preventing potentials, anti-virus and anti-protozoal activities 13,14 . Future studies are needed to evaluate if they affect palatability, feed intake, or feed digestion in animal husbandry.
To the best of our knowledge, the tested plants have not yet so far been analyzed for the tested individual phytochemical compounds. Some studies just reported the qualitative screening of phytochemicals including avonols, terpenes, and cardenolide glycosides in C. gigantea 13,15,16 . A recent study also detected the presence of p-hydroxybenzoic acid, 4-O-β-d-galactopyranosyl-d-fructose, myrciacitrin IV, quinic acid, and derivatives of astragaloside/quinic acid in C. procera, a plant species of the same genus as Calotropis gigantea 17 . A. digitata contained quercetin 3-O-glucoside 18 , iridoid, phenylethanoid, and hydrocinnamic acid glycosides 19 .The bioactive compounds found in the tested plants have demonstrated functional values such as antioxidant, antiin ammatory, antidiabetic 17 and antimicrobial activities especially against pathogenic Gram-negative bacteria 20 . Moreover, the antimicrobials content found in the plants can be considered as a potential source of novel antimicrobial function required to manipulate the rumen microbiota, and thereby optimizing feed e ciency.
The effect of C. gigantea on rumen protozoa has been recently evaluated at the genus level, and about half of the Entodinium population, which dominates the rumen protozoal community, was inhibited 4 . At the same time, the other genera of rumen protozoa, such as Isotricha, Epidinium, and Dasytricha, were maintained, and thus their contribution to ber digestion should not be compromised. Moreover, holotrich protozoa (Isotricha and Dasytricha), which have a lower predatory bacterial activity compared to the entodiniomorphids 21 , will ultimately have a lower impact on NH 3 -N concentration 22 and duodenal microbial protein ow 23 . When the protozoal population is reduced, the ciliate-associated prokaryotic cells (ectosymbionts) such as members of Acidobacteria and Actinobacteria 24 were also reduced, and this helps understand that the considerable limitation of the total removal of the whole protozoa population (defaunation) might lead to the tremendous decrease of the desired Actinobacteria and VFA concentration, especially butyrate 22 , which had been found reduced by 24% after 5h of in vivo fermentation in defaunation trials 25 . Indeed, Ampapon et al. 26 reported that supplementation of phytonutrients decreased the total protozoal counts by around 25%, increased amylolytic, proteolytic, and cellulolytic bacteria, enhanced ber digestibility, and reduced methane production. Similarly, Prevotella species that constitute one of the preferred preys of protozoa 24 were more abundant in the rumen of high-milk and high-proteins milk producing dairy cows 27 elucidating why protozoa contribute to lower down NUE 4 . Meanwhile, except for A. digitata, which has been used for its multiple health bene ts, the other tested plants still await further clinical con rmation or toxicological studies before applications in animal nutrition.
Studies on defaunated-refaunated animals using dietary chemical agents, rumen washing, animal isolation, and immunological approaches have been previously carried out to suppress the growth of rumen protozoa 28, 29 . However, none of those approaches is really practical at the farm level 30 . Chemicals like imidazole at 100 mmol/L inhibited rumen ciliate protozoa with the maintenance of NH 3 -N production after 24h of in vitro incubation of protozoal cultures 31 , which is consistent with our results using plants, but C. gigantea decreased the NH 3 -N concentration while F. macrophylla rather increased it. Besides, myricetin and quercetin were found in the tested plants, and these two compounds were shown to inhibit total rumen protozoa 32 . Future research using bioguided fractionation can help to con rm if these two compounds lonely or in association, are responsible for the observed inhibition of rumen protozoa by the plants.
The disruption of cell surface structure of protozoa has been reported by Zeitz et al. 33 as one of the signs of the dying rumen ciliate protozoa. Such a disruption of the cell surface structure leads to the loosened appearance of the lamentous glycocalyx, the non differentiation of chromatin and granular nucleoli, the accumulation of glycogen granules that obstructs the normal cell physical processes and ATP utilization 5,34 . Furthermore, the disruption of the protozoal cell surface will also quarrel the ectosymbiosis with the associated microbiota 35 , and the transport of soluble nutrients in unicellular parasitic protozoa will subsequently be compromised 36 . The hydrogen transporter will also be compromised which might contribute to a reduction of CH 4 release.
One of the well recognized and effective approaches for keeping the level of glucose from rapid rising in the serum is the inhibition of α-glucosidase with a speci c inhibitor, such as acarbose 37 . Although this does not apply to ruminant animals because they had little dietary sugars reaching the small intestines, digestion of starch after the feeding of high-concentrate diets can lead to rapid release of glucose, which is readily fermented to VFA, resulting in rumen acidosis. In lactating cows fed a high carbohydrate ration, the supplementation of acarbose prevented ruminal pH from lowering to the critical level of rumen acidosis 38 . Therefore, the tested plants, all of which can grow in dry or arid lands, may be used as alternatives of acarbose to lower the risk of or prevent rumen acidosis. An anti-glucosidase guided fractionation assays of the plants are needed to identify the component(s) responsible for the α-glucosidase inhibition. Meanwhile, some components have been found in the tested plants, and previous studies have ascribed the inhibitory effect of α-glucosidase activity to those components. Indeed, Mukhopadhyay and Prajapati 39 reported how the quercetin avonoid displayed inhibition of α-glucosidase activity, a remarkable hypoglycemic effect that improved and stabilized the secretion and regeneration of insulin in human pancreatic islets with no signi cant health hazards. Also, the administration of myricetin in diabetic rats resulted in a 50% decrease in hyperglycemia and an augmentation in hepatic glycogen and glucose-6-phosphate content 40 . This clearly demonstrated that the myricetin-rich plants tested in the present study could also have some anti-diabetic properties and deserve more attention as sources of functional feed and food. Indeed, the leaves were separated from the stems, sun-dried, and grounded to pass through a 2 mm mesh for nutritional analysis and a 0.3 mm mesh to be used as the feed of the protozoal cultures and subjected to crude extraction and phytochemical analysis (for F. macrophylla, both leaves and roots, for B. rapa chinensis, only its tuber). Finally, the ground plant samples were stored in glass bottles and kept at room temperature until further analysis. Nutritional analysis was done using standard methods 42  Plant sample preparation and extraction. The nely ground plant samples (0.3 mm particles) were individually subjected to chemical extraction according to the procedures described by Ayemele et al. 4 . Brie y, 15 g of each ground plant sample was combined with 450 mL of 80% ethanol. The mixture was subjected to ultrasonication using a Soniprep 150 Ultrasonicator (Hongxianglong Biotechnology Co., Ltd. Beijing, China) at 55°C for 45 min at 90% of its maximum power level. The sonicated samples were then centrifuged (10,000 rpm for 10 min), and the supernatants were collected and evaporated at 55°C in a rotary evaporator set at 85 rpm. Around 20 mL of the crude extract from each sample was freeze-dried at − 70°C for 72 h, and the nal extract was weighted.

Materials And Methods
Quanti cation of total phenolic, avonoid, and alkaloid of the crude extract. The crude extract samples were analyzed for total phenolic compounds (TPC), total avonoid compounds (TFC), and total alkaloid content (TAC) as previously described 4,44,45 . The content (per gram of plant leaves) of TPC, TFC, and TAC were respectively estimated as µg of gallic acid equivalent (µg GAE/g), µg of rutin equivalent (µg RE/g), and µg of aconitine equivalent (µg AE/g).
Reversed-phase HPLC for quanti cation of individual plant metabolites. Major secondary metabolites in the crude extract samples were analyzed using reversed-phase HPLC (RP-HPLC) as described by Ayemele et al. 4 . Brie y, the analysis consisted of separation, detection, and quanti cation of the compounds using the Agilent ZORBAX EclipsPlus C-18 reversed-phase column, a diode array detector, and an autosampler (Agilent Technologies, 1290 In nity II, Palo Alto, CA, USA). Spectral data were recorded from 200 to 800 nm, and the chromatograms of the compounds of interest were monitored at 326 nm. The chromatographic peaks of the plant compounds were con rmed by comparing the retention time with those of the following reference standards: 3hydroxybenzoic acid, myricetin, astragaloside, quercetin, and anthraquinone. The content of individual compounds was expressed as µg/g of plant samples (DM).
In vitro culture of rumen protozoa and evaluation of plant effects Collection of rumen uid and preparation of concentrated rumen protozoa. Animals were handled and cared for following the guidelines approved by the Animal Care Advisory Committee of the Chinese Academy of Agricultural Sciences that approved the rumen uid collection method. Rumen uid was collected from three cannulated dairy cows and transported to the laboratory within 1.5 hours, following the procedures as previously reported 4 . Immediately after the rumen uid was brought to the laboratory, the protozoa cells were sedimented under a continuous O 2 -free CO 2 ux, and the top portion of the rumen uid was removed. The remaining rumen uid with concentrated protozoal cells was used as the inoculum in the in vitro culture experiments and evaluation of the plants.
In vitro culture of rumen protozoa with plant supplementation. The in-vitro culture technique of rumen protozoa was done as reported by Ayemele et al. 4 Brie y, 8 mL of medium 43 was dispensed into individual glass culture tubes containing 100 mg of the same TMR fed to the lactating cows that donated the rumen uid. Each of the ground plant samples was added to the glass culture tubes at four doses (0, 0.7, 0.9, and 1.1 mg/mL culture uid) with each dose having three replicates (n = 3). Finally, 2 mL of the concentrated rumen protozoa were added under a continuous ux of O 2 -free CO 2 . The tubes were incubated at 39°C for 24 h. Each in vitro culture was sampled for protozoa counting and detection of NH 3 -N concentration.
Microscopic counting of rumen protozoal cells. The protozoal cells were morphologically identi ed and counted microscopically using the procedures previously described 4,5 . Brie y, 0.3 mL of each in vitro culture sample was combined with 0.3 mL of 18.5% formalin to x the protozoal cells. Then, 30 µL of a brilliant green dye solution was added to stain the protozoal cells. Finally, 1.4 mL of 30% glycerol was added to each tube and mixed. The cells were counted using a Sedgewick-Rafter counting chamber (Thomas Scienti c, Swedesboro, NJ) under a light microscope at 10X magni cation.
Scanning electron microscopy for protozoal cells surface. The samples of cultured rumen protozoa cells were prepared for scanning electron microscopy following the procedures previously described et al. 4 . Brie y, protozoal cells were pelleted by centrifugation at 500 × g for 5 min from 1 mL of each in vitro culture. Then, the cells were xed with 3% glutaraldehyde and subsequently rinsed twice with a potassium phosphate buffer (0.1 M, pH 7.2). A sequential washing of protozoal cells with 30, 50, 70, 90, and 100% ethanol helped to dehydrate the cells, followed by washing in acetone and then hexamethyldisilazane. The cells were then spatter-coated with platinum and viewed on a Hitachi S-4700 (Hitachi America, Ltd.) after ve hours of drying with a ux of CO 2 .
Detection of ammonia nitrogen concentration. NH 3 -N concentrations of the culture uid were determined using the colorimetric method described by Chaney and Marbach 46 . Brie y, 500µL and 400µL of the solutions A and B, respectively, were successively added and mixed with 8µL of each sample, and then incubated for 30min at 37°C. In a 96-well microplate, 200 µL was dispensed to each well and optical absorbance was determined at 550nm wavelength using a spectrophotometer (Thermo Fisher Scienti c, Waltham, USA). A serial dilution of ammonia solution (1,2,4,8,16, and 32 mg/dL) served as the external standard.
α-glucosidase inhibitory assay. The α-glucosidase inhibition assay was performed according to the procedures of Obaroakpo et al. 37 and Rengasamy et al. 47 with minor modi cation. Brie y, α-glucosidase obtained from Saccharomyces cerevisiae was dissolved in 0.1 M potassium phosphate buffer (PBS, pH 6.8) and used as the enzyme stock solution (0.1 Unit/mL). The substrate for the enzyme activity assay was p-nitrophenyl-a-D-glucopyranoside (pNPG) dissolved in the same PBS buffer (stock concentration of pNPG: 0.375 mM). Inhibition to the yeast α-glucosidase was determined at ve concentrations ranging from 100 to 500 µg/mL with an increment of 100 µg/mL of each plant extract in PBS, using 96-wells microplates. Acarbose, which is an inhibitor of α-glucosidase, served as the positive control at the above concentrations. Each α-glucosidase assay reaction contained 20 µL of each plant extract solution, 20 µL of αglucosidase enzyme solution, and 40 µL of pNPG. The mixture was incubated at 37°C for 40 min, and then 80 µL of 0.2 M sodium carbonate in PBS was added to terminate the reaction. Optical absorbance was read using a spectrophotometer (Thermo Fisher Scienti c, Waltham, USA) at 405 nm. One negative control was included in parallel that contained no plant extract or acarbose. Each assay reaction had three replicates. The inhibition (%) was calculated using the following equation: where A control is the absorbance of the negative control (no inhibitor), and A sample is the absorbance of each plant extract.
The concentration of each crude extract that inhibited the α-glucosidase activity by 50% under the assay conditions was de ned as IC 50 expressed in μg/mL. Statistical Data Analysis. The data were analyzed in a complete randomized design with one-way ANOVA using the PROC GLM procedure of SAS 9.4 (SAS Institute, Cary, NC, USA) to compare the means of nutrients and phytochemicals among the different plants and the protozoal cell counts among the different doses of each plant sample. The means of α-glucosidase inhibitory activity were also compared among the different doses within each plant. Orthogonal polynomial contrast was done to determine the linear and/or quadratic effects of plant doses on protozoal cells and NH 3 -N concentration. Signi cance was declared at P< 0.05. The GC/MS Translator B.07.17 was used to convert the Chemstat HPLC data les into MassHunter les for qualitative and quantitative analysis of the compounds in the plant extracts.

Conclusion
The present study demonstrated how new forest plants could be used as potential feed additives for animal husbandry. The tested plants contained bioactive compounds that could suppress the rumen ciliate protozoal population, potentially decreasing the intra-ruminal protein recycling. In addition, the plants were able to inhibit the α-glucosidase enzyme, and thereby potentially mitigating the risk of ruminal acidosis in high-producing cows fed high-concentrate diets. Overall, the leaves of the tested plants contain high levels of protein, especially the leaves of A. digitata, K. indica, P. oleracea, and C. gigantea, making them new sources of feed for ruminants. However, future studies are needed to evaluate other important nutritional traits, such as digestibility and palatability, and to accurately identify the anti-protozoal phytochemical(s) using the bio-guided fractionation assays.   Means with different superscripts within a row differ (P < 0.05).