This study was conducted in accordance with internationally accepted ethical standards for research and reporting. All procedures complied with the ARRIVE 2.0 guidelines for transparent reporting of scientific studies and adhered to the principles outlined in the EU Directive 2010/63/EU on the protection of animals used for scientific purposes. Ethical clearance was obtained from the Ethical Committee of the Scientific Council, Faculty of Agricultural Sciences, Universidad Central “Marta Abreu” de Las Villas, Cuba (Protocol No. 57/2021; approved on March 14, 2021).

The experimental design involved exclusively in vitro analyses of plant-based feed materials, and therefore no live animals were subjected to experimental procedures, handling, or invasive sampling. Consequently, no animal welfare risks were identified. All laboratory procedures were conducted under standard biosafety conditions in accordance with institutional laboratory safety protocols.

Plant material (T. diversifolia) was collected from agricultural land in Napo Province, Ecuador, with full compliance to national agricultural and environmental regulations. The collection did not involve protected species or restricted areas, and no specific permits were required under local legislation. Handling, processing, and disposal of plant materials followed established environmental safety guidelines to minimize ecological impact.

The study design, data handling, and reporting were conducted in accordance with principles of scientific integrity, reproducibility, and transparency. No human participants or personal data were involved in this study.

The study was conducted from March to July 2025 in Napo Province, Tena Canton, Misahuallí Parish (Santo Urku community), Ecuador (0°57′01″S; 77°51′46″W; 565 m a.s.l.). The region has a humid tropical climate (Af, Köppen classification), with an annual rainfall of approximately 3800 mm, a mean temperature of 24°C–26°C, and relative humidity of around 85%. Soils are sandy loam, acidic (pH 4.5–5.5), and of moderate fertility.

In vitro digestibility tests were performed at the Natural Products Laboratory, Universidad Regional Amazónica Ikiam (Tena, Ecuador). Harvesting was carried out at a cutting height of 50 cm from the ground [9], and the evaluation of green and dry matter production was conducted using the method proposed by Rodríguez and Torres [10].

The bromatological characterization of the meal was conducted using 2 kg samples of green forage, analyzed in triplicate for each treatment. Samples were dried using a vertical solar dryer at 55°C ± 2°C for 72 h until reaching approximately 10% residual moisture. The dried material was ground using a Willey mill (Thomas Scientific, USA) with a 1 mm sieve, homogenized, and stored in airtight polyethylene bags for subsequent analyses.

Proximate analysis was performed according to U. Florida [11, 12], including dry matter (DM), organic matter (OM), crude protein (CP; Kjeldahl method, N × 6.25), ether extract (EE), and nitrogen-free extract (NFE). Fiber fractionation (neutral detergent fiber [NDF] and acid detergent fiber [ADF]) was determined according to Van Soest et al. [13]. Additionally, analyses followed AOAC International (2016) methods: DM (934.01), CP (976.05), EE (920.39), and ash (942.05). Lignin content was determined using permanganate oxidation.

Gross energy (GE), ME, and digestible energy (DE) were estimated using predictive equations based on nutritional composition (Table 1) [14, 15]. Mineral composition, including phosphorus (P), potassium (K), calcium (Ca), magnesium (Mg), iron (Fe), copper (Cu), zinc (Zn), and manganese (Mn), was analyzed using atomic absorption spectrophotometry [16].

Sixteen amino acids were quantified by high-performance liquid chromatography (HPLC; Agilent 1100 system, Ecuachemlab Cía. Ltda., Ecuador) following acid hydrolysis (6 N HCl, 110°C, 24 h). Tryptophan was analyzed separately using alkaline hydrolysis with fluorescence detection. Data were expressed as percentage of sample dry weight (% w/w).

The study was conducted at the Natural Products Laboratory of Universidad Regional Amazónica Ikiam using five treatments based on formulated diets containing different inclusion levels of T. diversifolia meal (10%, 15%, 20%, and 25%), along with a control treatment (0%). Each treatment was evaluated with four replicates under a completely randomized design.

Proximate chemical composition parameters of all treatments (DM, EE, CP, and OM) were determined (Table 2). Based on in vitro OM digestibility values, DE and NE were estimated using predictive equations described by [17]. In addition, ME for growing pigs and finishing pigs was calculated according to the models proposed by previous studies [18, 19] (Table 3).

Five experimental diets were formulated containing 0%, 10%, 15%, 20%, and 25% T. diversifolia meal as a partial replacement for soybean meal and fish meal. The control (basal) diet was formulated using corn meal, soybean meal, fish meal, wheat bran, and a vitamin–mineral premix (Table 4). All diets were designed to be iso-nitrogenous and iso-energetic, meeting the nutrient requirements for growing pigs according to NRC (National Research Council, 2012) [20].

Statistical analyses were performed using R software version 4.3.1 (R Core Team, 2023, Vienna, Austria) under a completely randomized design with one-way analysis of variance. Data are expressed as mean ± standard deviation, and significance was declared at p < 0.05 using Tukey’s post hoc test.

T. diversifolia flour had a CP content of 318.127 (g/kg DM), with high levels of crude fiber (366.146 g/kg DM) and ash (176.471 g/kg DM). EE content was low (10.444 g/kg DM), while the neutral detergent fiber (NDF) and ADF contents were 534.214 (g/kg DM) and 533.013 (g/kg DM), respectively. Lignin reached 292.917 (g/kg DM), indicating a structurally complex fibrous fraction. In addition, the sample had appreciable levels of minerals such as calcium (32.413 g/kg DM), potassium (39.616 g/kg DM), and phosphorus (7.203 g/kg DM).

Chromatographic analysis revealed a total amino acid concentration of 12.38% (Figure 1). The most abundant amino acids were aspartic acid (3.37%), serine (1.46%), leucine (1.35%), valine (1.08%), and glycine (0.93%). Significant amounts of lysine (0.85%) were detected, whereas methionine was low (0.26%). Tryptophan was detected in minimal concentrations (0.0004%), and neither isoleucine nor glutamic acid were detected.

The results of the chemical analysis (Table 5) of experimental diets with different levels of T. diversifolia flour inclusion showed no statistically significant differences (p > 0.05) in DM, OM, EE, or ADF contents among treatments. However, significant differences were observed in CP content (p = 0.03) and NDF content (p = 0.02).

Crude protein content showed a progressive increase from the control treatment (173.9 g/kg DM) to treatment T3 (20%), reaching a maximum value of 178.38 g/kg DM, followed by a decrease in T4 (171.68 g/kg DM). Post hoc analysis indicated that T3 differed significantly from the treatment with the highest inclusion level (T4), whereas treatments T1, T2, and the control exhibited intermediate values.

In terms of NDF, treatment T3 (20%) showed the highest value (201.23 g/kg DM), which was significantly higher than T2 (15%) with 188.78 g/kg DM. The remaining treatments did not show significant differences among them.

Total starch content (ST) ranged from 390.83 g/kg DM (T4) to 405.88 g/kg DM (T3), with no statistical differences (p = 0.4019). EE values remained constant (29.98–31.38 g/kg DM), with no evidence of treatment effects (p = 0.9033). DM and OM also remained stable among treatments (p > 0.4), indicating that the inclusion of T. diversifolia meal up to 25% does not significantly affect these fractions.

Energy trends derived from the in vitro digestibility assays (Figure 2) indicated a gradual numerical decline in DE, ME, and NE values as the inclusion level of T. diversifolia increased from 0% to 25%.

The in vitro digestibility of DM and OM showed a numerical reduction with increasing levels of T. diversifolia inclusion (Table 6). However, no significant differences (p > 0.05) were observed among the 10%–25% inclusion treatments, indicating that the reduction pattern was not strictly progressive. The control diet (0%) exhibited slightly higher digestibility coefficients, followed by minor decreases in the experimental diets, although the overall variation remained within an acceptable range.

Correspondingly, the estimated energy values (DE, ME, and NE) did not differ statistically (p > 0.05) among treatments, although a mild numerical downward trend was observed with increasing inclusion levels. These findings suggest that T. diversifolia meal can be incorporated up to 25% in pig diets without significantly compromising in vitro digestibility or energy content.

T. diversifolia flour has a nutritional profile that stands out for its high CP content (318.127 g/kg DM), making it a promising alternative in animal diets, especially in production systems seeking accessible and sustainable protein sources. This value falls within the range reported by various authors, who have documented concentrations between 264.100 g/kg DM and 348.100 g/kg DM depending on factors such as the physiological age of the plant, cutting frequency, and agroecological conditions [21].

The crude fiber content was high (366.146 g/kg DM), as were the NDF (534.214 g/kg DM) and ADF (533.013 g/kg DM) values, reflecting a considerable structural fraction. These values suggest limited DM digestibility, especially for monogastric animals, although they could be functional as effective fiber in ruminants [4]. Lignin reached 292.917 (g/kg DM), a value that compromises the digestibility of the fibrous fraction by forming lignocellulosic complexes that are resistant to enzymatic action [22]. It has been observed that lignin levels above 180.100 (g/kg DM) can significantly reduce nutrient degradability in monogastric animals [23], and therefore this value must be carefully considered when formulating diets.

The low EE content (10.444 g/kg DM) is characteristic of shrubby forages, limiting their energy contribution via lipids. However, the energy balance can be maintained by including other energy-rich ingredients in the diet [24]. Ash accounted for 176.471 (g/kg DM), indicating significant mineral richness. This value is often associated with the presence of essential macro- and microminerals, as evidenced by the levels of calcium (32.413 g/kg DM), potassium (39.616 g/kg DM), and phosphorus (7.203 g/kg DM). The calcium content is particularly relevant for the diets of growing and lactating animals [25], while potassium and phosphorus play fundamental roles in osmotic regulation and energy metabolism [26].

The magnesium level (4.682 g/kg DM) and values for trace elements such as iron (0.267 g/kg DM), zinc (0.113 g/kg DM), manganese (0.068 g/kg DM), and copper (0.003 g/kg DM) are adequate and comparable to other tropical forage species [27]. The bioavailability of these elements must be evaluated based on their interaction with other dietary components, especially fiber and phytates [28]. Although the copper content was low, it can be supplemented with external sources according to the requirements of the animal species [29].

Overall, the nutritional profile of T. diversifolia flour indicates interesting potential as a protein–mineral supplement in diets for ruminants and, to a lesser extent, for monogastric animals, provided that its high-fiber and lignin content is considered. Its use could be particularly suitable in silvopastoral systems or as a component of multinutritional mixtures [30]. However, further in vivo or in vitro digestibility studies are recommended to confirm its effective nutritional value, as well as acceptance and palatability tests in the target species [31].

Chromatographic analysis of T. diversifolia flour revealed a total amino acid content of 12.38%, confirming its potential as a protein source for animal feed systems in tropical areas. This value significantly exceeds the contribution of conventional tropical grasses (Cenchrus clandestinum: 8–10% CP) and is comparable to the range reported in forage legumes such as Leucaena leucocephala (18%–25%) [5].

Among essential amino acids, leucine and valine predominated, with values of 1.35% and 1.08%, respectively. These amino acids are essential for muscle metabolism and protein synthesis, particularly in growing animals [32]. The concentration of lysine (0.85%) is also noteworthy, as it is commonly a limiting amino acid in plant-based diets, highlighting the potential of T. diversifolia as a complementary ingredient in monogastric diets [33].

However, low concentrations of methionine (0.26%) and an almost negligible presence of tryptophan (0.0004%) were observed, along with the absence of isoleucine and glutamic acid, which represents a nutritional limitation. The deficiency of methionine, an essential sulfur-containing amino acid, can compromise protein utilization efficiency [34]. Therefore, it is advisable to formulate diets that compensate for this deficiency, either through methionine-rich ingredients such as treated soybean meal or through synthetic supplementation [35].

Among non-essential amino acids, aspartic acid (3.37%), serine (1.46%), and glycine (0.93%) were predominant. These compounds play key roles as metabolic precursors and in the synthesis of nucleotides, collagen, and other structural components [36–38].

Overall, the amino acid profile suggests that T. diversifolia is a valuable resource for diet formulation; however, its use must be carefully balanced to address deficiencies in essential amino acids. Integration of this ingredient into animal diets should be supported by evaluations of true protein digestibility and productive performance, particularly in species sensitive to amino acid imbalances [39].

The results of the chemical analysis (Table 5) indicate that the progressive inclusion of T. diversifolia flour up to 25% did not significantly affect the DM, OM, EE, or ADF contents (p > 0.05), which is consistent with previous findings [6]. This behavior suggests that the incorporation of this forage does not substantially alter the structural or non-DE fractions of the diet [40]. However, significant effects were observed for CP (p = 0.03) and NDF (p = 0.02), which are critical parameters in diet formulation.

CP content increased up to treatment T3 (20%), reaching 178.38 g/kg DM, followed by a decrease in T4 (25%). This pattern suggests that moderate inclusion levels may enhance protein intake, possibly due to the high concentration of soluble nitrogen in the leaves of this species. The reduction observed in T4 may be attributed to protein dilution associated with increased fiber content, a phenomenon also reported in diets with high inclusion of fibrous forages [41]. Post hoc analysis confirmed a significant difference between T3 and T4, indicating that 20% may represent an optimal inclusion level for maximizing protein content.

NDF content was highest in T3 (201.23 g/kg DM), significantly exceeding that of T2 (188.78 g/kg DM). This increase in effective fiber may contribute to satiety in monogastric animals and stimulate rumination in ruminants [42], although it may also reduce digestibility. Despite variations in CP and NDF, total starch content remained stable (p = 0.40), suggesting that non-structural carbohydrate fractions were not affected by T. diversifolia inclusion [43]. EE content also remained constant (p = 0.90), confirming the limited lipid contribution of this forage.

In vitro DM digestibility showed a significant decrease (p = 0.0128) with increasing inclusion levels, particularly in treatment T3 (31.68%) compared with the control (42.88%). Although no significant differences were observed in in vitro digestibility of OM or in estimated energy values (DE, MEcr, MEfc, and NE), a general numerical decline (p > 0.1) was evident as inclusion levels increased. This reduction may be associated with increased structural fiber and lignified fractions, which limit enzymatic degradation and nutrient availability [44]. According to Weimer [45], elevated lignin and hemicellulose levels reduce cell wall degradability and negatively impact digestive efficiency.

Despite this trend, treatments T1 (10%) and T2 (15%) maintained moderate energy losses and acceptable digestibility coefficients (>36%) (Table 6), indicating that low to moderate inclusion levels may be viable without significantly compromising nutritional value [46]. This finding is relevant for practical applications, as it supports the use of T. diversifolia as a functional and economical feed component without adversely affecting animal performance [47]. Although DE and NE values did not differ significantly (p > 0.45), the control diet maintained the highest values (16.83 kcal/kg and 9.44 kcal/kg, respectively), likely reflecting the higher efficiency of conventional diets lacking high-fiber components [48].