MATERIALS AND METHODS

Ethical approval

The study was approved by the Animal Care and Use Committee of Rajamangala University of Technology Isan, Thailand (approval no. 01-67-002). According to the National Research Council of Thailand Guidelines for Ethics of Animal Experimentation, approval was required for rumen fluid collection because the primary objective of this study involved the laboratory evaluation of ruminant feeds. All procedures were conducted in accordance with the applicable regulations and guidelines. The study adhered to the Animal Research: Reporting of in vivo Experiments 2.0 guidelines. Throughout the experimental period, all animals were monitored daily for clinical signs of mimosine toxicity and adverse health effects, including alopecia, excessive salivation, lethargy, and abnormal weight loss. No adverse clinical signs were observed during the study.

Study period and location

The experiment was conducted during the rainy season in August 2024 in Roi Et Province, Thailand. Fresh young leaves of L. leucocephala and rhizomes of Z. officinale were harvested locally at approximately 45 days of age. The feeding trial and laboratory analyses were performed under tropical conditions in Thailand.

Study design

Four female Thai native beef cattle with an initial body weight (BW) of 230 ± 10 kg and approximately 2 years of age were randomly assigned to four dietary treatments according to a 4 × 4 Latin square design. The treatments consisted of a control group without LGP supplementation and groups receiving LGP at 50, 100, and 150 g/head/day.

The concentrate allowance was provided at 1.0% of BW on a DM basis. To ensure accurate feeding rates throughout the study, the daily concentrate allocation for each animal was recalculated according to individual BW measured at the beginning and end of each experimental period. Because of the short 21-day duration of each period, average daily gain was not evaluated as a primary parameter; however, no significant BW loss was observed.

Animals received LGP according to their assigned treatments and were fed two equal portions of locally sourced rice straw (Oryza sativa L.) at 07:00 and 16:00 h daily. The basal diet was formulated to meet the maintenance and growth requirements of Thai native beef cattle. The chemical compositions of the concentrate mixture, rice straw, and LGP are presented in Table 1. Based on the daily intake proportions, the nutrient composition of the total diet provided sufficient energy and protein to meet the maintenance requirements of the experimental animals.

Each experimental period lasted 21 days, comprising 14 days for dietary adaptation and 7 days for sample collection and nutrient digestibility measurements. During the experiment, the cattle were housed individually in well-ventilated stalls measuring approximately 2 × 3 m, with concrete floors covered with rubber mats, under natural daylight and ventilation. Clean drinking water was available ad libitum throughout the experiment.

Before the start of the trial, all animals underwent a comprehensive health examination, including routine vaccination against Foot and Mouth Disease and treatment for internal and external parasites with ivermectin to ensure optimal health.

The chemical compositions and nutrient contents of the experimental diets and LGP are listed in Table 1.

Items | Concentrate | Rice straw | LGP
Feed ingredients (% as fed) |  |  | 
Cassava pulp | 25.0 | – | –
Cassava chip | 21.5 | – | –
Palm meal | 34.0 | – | –
Rice bran | 14.5 | – | –
Soybean meal | 0.5 | – | –
Urea | 1.5 | – | –
Sulfur | 1.0 | – | –
Mineral mix | 1.0 | – | –
Salt | 1.0 | – | –
Leucaena leaves meal | – | – | 75.0
Ginger powder | – | – | 15.0
Cassava chip | – | – | 9.0
Molasses | – | – | 1.0
Chemical composition (%) |  |  | 
Dry matter (DM) | 92.5 | 93.0 | 89.5
Organic matter | 92.7 | 91.5 | 85.2
Ash | 7.3 | 8.5 | 14.8
Crude protein | 14.7 | 2.2 | 24.6
Neutral detergent fiber | 28.3 | 75.5 | 32.5
Acid detergent fiber | 15.1 | 47.4 | 25.2
Condensed tannins | – | – | 6.8
Flavonoids | – | – | 2.4

Preparation of phytogenic pellets

Fresh young leaves of L. leucocephala and rhizomes of Z. officinale were harvested locally in Roi Et Province, Thailand. Following collection, the leaves and chopped ginger rhizomes were dried in a hot-air oven at 60°C for 48 h until a constant weight was achieved. The dried materials were then ground using a Cyclotec Mill (Tecator, Hoganas, Sweden) and passed through a 1-mm screen.

Pellets were prepared by combining 75% L. leucocephala leaves, 15% ginger powder, 9% cassava chips, and 1% molasses using a pellet machine. This ratio of active phytogenic ingredients was selected to provide an optimal supply of CT and essential oils to effectively modulate rumen fermentation without impairing microbial nutrient degradability. Following pellet formation, pellets measuring approximately 6-8 mm in diameter were sun-dried for 2-3 days to reduce moisture content and ensure stability during storage and feeding.

Although mimosine and SP concentrations were not quantified because of laboratory limitations, previous studies have demonstrated that the combination of sun drying, oven drying at 60°C, and heat generated during pellet production substantially reduces mimosine concentrations to safe levels. Cassava chips and molasses were incorporated primarily as pellet binders and highly fermentable energy sources, and their nutrient contributions were inherently included in the proximate composition of the final pellets (Table 1). To maintain stability and prolong shelf life, the pellets were stored in sealed moisture-proof containers at room temperature throughout the feeding trial.

Sample collection and chemical analyses

Feeds were weighed daily to determine the intake of concentrate, rice straw, and LGP for each dietary treatment. Feed, fecal, and urine samples were collected during the last 7 days of each experimental period. Following oven drying, samples were ground to pass through a 1-mm screen and analyzed for DM, ash, and CP according to Association of Official Analytical Chemists procedures [14]. Acid-insoluble ash was determined according to Van Keulen and Young [15]. Acid detergent fiber (ADF) and neutral detergent fiber contents were analyzed according to the method of Van Soest et al. [16].

The concentrations of CT and flavonoids in LGP were determined using the Folin-Ciocalteu reagent by measuring absorbance at 765 nm [17].

Spot urine samples were obtained by manual stimulation of the vulva to induce urination, whereas fecal samples were collected directly from the rectum. Samples were collected at 0 and 4 h after the morning feeding. Total daily urine output was estimated from creatinine concentration, assuming a constant creatinine excretion rate of 0.88 mmol/kg BW^0.75. Microbial purine absorption was estimated from the excretion of purine derivatives. Allantoin and creatinine concentrations were measured by high-performance liquid chromatography, and microbial nitrogen supply was calculated using the equations, assumptions, and recovery factors described by Chen and Gomes [18].

Rumen fluid and jugular blood samples were collected at 0 and 4 h after feeding. To minimize stress and discomfort, rumen fluid was collected using a soft flexible stomach tube connected to a vacuum pump, with the sample volume limited to 200 mL. Immediately after collection, ruminal temperature and pH were measured using a handheld pH meter (HI8424 Microcomputer; Hanna Instruments, Woonsocket, RI, USA).

Rumen fluid was filtered through four layers of cheesecloth and divided into two portions. In the first portion, NH3-N and VFA concentrations were determined by mixing 45 mL of rumen fluid with 5 mL of 1 M H2SO4 followed by centrifugation at 1,600 × g for 15 min. The NH3-N concentration was determined by the micro-Kjeldahl method [14], whereas VFA concentrations were analyzed using high-performance liquid chromatography [19].

CH4 production was estimated from VFA concentrations according to Moss et al. [20] using the following equation:

CH4 production (mol/100 mol total VFA) = 0.45 (acetate) − 0.275 (propionate) + 0.40 (butyrate).

For microbial enumeration, 1 mL of filtered rumen fluid was immediately transferred into 9 mL of 10% formalin solution to preserve microbial cells. Bacterial and protozoal populations were counted directly using a hemocytometer under a phase-contrast microscope, as described by Galyean [21]. To ensure accuracy and reliability, microscopic counts were performed in duplicate, and at least 10 microscopic fields were examined for each sample. To ensure precision and reproducibility, all chemical and microbial analyses were conducted in triplicate.

Blood samples (5 mL) were collected from the jugular vein at 0 and 4 h after feeding, placed in tubes containing ethylenediaminetetraacetic acid, and analyzed for blood urea nitrogen (BUN) according to Crocker [22].

Statistical analysis

Before statistical analysis, the assumptions of analysis of variance were evaluated. Residual normality was assessed using the Shapiro-Wilk test, and homogeneity of variances was examined to verify the suitability of the statistical model. The use of four animals in a 4 × 4 Latin square design represents standard practice for in vivo rumen metabolism studies and balances statistical power with the ethical principle of minimizing animal use.

Potential outliers were evaluated using studentized residuals, and no observations exceeded the ±3 standard deviation threshold. Therefore, all observations were retained for the final analysis.

Data were analyzed using the MIXED procedure in SAS version 9.4 (SAS Institute Inc., Cary, NC, USA) with a variance-covariance structure based on a 4 × 4 Latin square design. The statistical model was:

Yijk = μ + Ai + Pj + Tk + eijk

where Yijk represents the observation, μ is the overall mean, Ai is the random effect of animal (i = 1, 2, 3, and 4), Pj is the random effect of experimental period (j = 1, 2, 3, and 4), Tk is the fixed effect of treatment (k = 1, 2, 3, and 4), and eijk is the residual error.

Differences among treatment means were determined using Tukey's multiple comparison procedure [23], and significance was declared at p < 0.05. Orthogonal polynomial contrasts were used to evaluate linear and quadratic treatment responses.