Comparison of In Vitro Fermentation Characteristics Among Five Maize Varieties

Abstract

Maize (Zea mays L.) silage in the irrigated and flat areas of Italy represents the most important large ruminant feed crop due to the high dry matter yield and nutritive value per hectare. The aim of the investigation was to evaluate the chemical composition and the in vitro fermentation patterns of five maize varieties (Tiesto, R700 1, MAS 78.T, DKC 7074 and KWS Kantico) freshly chopped and preserved via ensiling. The results indicated that the chemical composition was not significantly different among varieties. The substrates were incubated for 72 h with buffered rumen fluid collected from cow. The ensiling process slightly reduced gas production and fermentation kinetics, likely due to the consumption of soluble sugars during fermentation. Organic matter loss (OM loss) differed significantly (p < 0.01) among varieties in ensiled maize, correlating with their neutral detergent fiber (NDF) content. While total volatile fatty acid (VFA) production showed no significant differences between varieties, the buffer capacity ratio (BCR), an indicator of protein degradation, varied significantly. Ammonia production (NH₃) was significantly higher in ensiled samples, supporting previous findings that ensiling increases non-protein nitrogen (NPN) due to microbial proteolysis and plant enzyme activity. The gas production profiles and fermentation rates over time showed minor differences between fresh and ensiled samples, with fresh material exhibiting faster fermentation kinetics due to the presence of soluble sugars. These findings highlight the importance of evaluating maize silage quality to optimize ruminant nutrition and feed efficiency.

1. Introduction

Maize (Zea mays L.) silage in the irrigated and flat areas of Italy represents the most important large ruminant feed crop due to the high dry matter yield and nutritive value per hectare [1,2]. The widespread use of maize silage in ruminant diets is attributed to its ability to provide a consistent and energy-rich forage source, ensuring optimal animal performance and productivity. Indeed, high-quality forages play a crucial role in enhancing rumen fermentation features, which in turn supports overall animal well-being [3] and contributes to the production of high-quality animal-derived food products, such as milk and meat, with improved nutritional properties [4]. However, several factors (i.e., plant vegetative stage at harvesting, genetic, percentage of structural content and lignin, efficiency of compression, storage) could affect maize silage nutritive characteristics [5], thereby impacting its overall feeding value and fermentative profile, which are critical for accurately estimating its nutritional contribution to ruminant diets. One of the key challenges in maize silage production is determining the optimal harvesting time [6], as premature cutting can cause an excessive loss of nutrients from silo runoff. Additionally, harvesting too early often results in underdeveloped starch content in the grains, thereby reducing the overall energy concentration of the forage. Conversely, delaying harvest until later stages of ripening can lead to a decline in the silage’s nutritive value, characterized by lower protein content and an increase in crude fiber, which may negatively affect digestibility and feed efficiency [7]. Due to its great variability in terms of chemical composition, there could be a risk of formulating diets with an untruthful nutritional value when specific chemical analyses are not carried out. Inaccurate nutritional assessments may lead to imbalanced feed formulations, potentially requiring the inclusion of higher proportions of expensive concentrate feeds to compensate for nutrient deficiencies. This not only increases feed costs but may also impact the sustainability of livestock production systems. Therefore, precise evaluation of maize silage quality through regular testing is essential to optimize ruminant nutrition, enhance feed efficiency and ensure economic viability for farmers.

In Italy, maize is cultivated over an area of 578,417 ha, with a total production of 52,237,540 q [8], and about 86% of maize and its derivatives are used to feed livestock, a sector characterized by a strong orientation toward high-quality and certified products [9]. Dairy products in particular represent the most significant segment of the Italian food industry, accounting for more than 12% of the total revenue generated by the national agri-food sector. The economic value of dairy production reaches an estimated EUR 14.5 billion annually, reflecting the pivotal role of this industry in the country’s economy. Every year, Italian companies produce 1,000,000 tons of cheeses, 460,000 of which are Protected Origin Denomination (POD) products. In Italy, Lombardy represents the region with the highest number of dairy cattle farms, accounting for approximately 27% of the total national herd, the majority of which are concentrated in the Po Valley. However, one notable exception within this region is the Parmigiano Reggiano production area, where the use of silage is strictly prohibited by production regulations to maintain the specific characteristics and quality of the cheese [10]. Meanwhile, Campania emerges as the leading region for buffalo farming, accounting for 72% of Italy’s total buffalo population, which are mainly reared for producing the Mozzarella di Bufala DOP (Available online: https://www.mozzarelladop.it/ (accessed on 6 November 2024)) and, to a lesser extent, buffalo meat. Given the crucial role of maize silage in livestock feeding, understanding its nutritional characteristics is essential for formulating balanced diets that meet the animals’ feeding requirements [11]. Proper nutritional management not only ensures optimal animal health and productivity but also minimizes nutrient losses to the environment, thereby enhancing the sustainability of livestock operations [12]. The economic impact of feed costs in dairy production is significant, with estimates indicating that feed expenses account for approximately 45% of the total costs associated with producing 100 kg of milk. These costs are subject to fluctuations over time due to various factors, including the availability of raw materials, market dynamics and fluctuations in concentrate prices [13]. One of the key challenges associated with forage conservation is the inevitable reduction in nutritive value compared to fresh forage [14]. However, these losses can be minimized through proper harvesting, processing and storage techniques. Ensuring that silage is produced under optimal conditions, such as timely harvesting, adequate compaction and proper fermentation, helps preserve its nutritional integrity and enhances its digestibility for ruminants [15].

To assess the fermentation characteristics of forages, various analytical techniques have been developed, one of the most effective being the in vitro gas production technique (IVGPT) [16]. This method is widely used to estimate the potential rumen fermentation of different feedstuffs. IVGPT involves incubating a ground feed substrate with a microbial inoculum (typically rumen fluid or feces) in a buffered medium under anaerobic conditions at a constant temperature of 39 °C. The amount of gas produced over time provides valuable insights into the kinetics of microbial fermentation, allowing researchers to evaluate the digestibility, energy availability and overall nutritional value of various feed materials [17].

The primary objective of this study was to evaluate the in vitro fermentation patterns of five different maize varieties similar in genetic and agronomic characteristics, both in their freshly chopped form and after undergoing the ensiling process. By analyzing fermentation dynamics, this research aims to provide valuable information on how maize varieties commonly used in the south of Italy respond to preservation methods, ultimately contributing to the development of optimized feeding strategies for ruminant nutrition. We hypothesized that maize varieties exhibit different behaviors in terms of chemical composition and fermentation characteristics and kinetics.

2. Materials and Methods

2.1. Planting, Harvesting and Sample Preparation

In the experimental trial, five maize (Zea mays L.) varieties, named Tiesto (Var 1), R700 1 (Var 2), MAS 78.T (Var 3), DKC 7074 (Var 4) and KWS Kantico (Var 5), were used. All maize varieties were grown in 2022 in a plain area sited in Salerno Province, Italy (Longitude 14°58′00′′ E, Latitude 40°30′00′′ N, Altitude 10 m a.s.l., Mean value of temperature 16.8 °C and Rainfall 988 mm), fertilizing the soil (maintenance + enrichment) with P₂O₅ (200 kg/ha) and K₂O (250 kg/ha). The plants were harvested at the milk stage, with dry matter content ranging from 30 to 35%. Mowing was carried out between 15 and 20 cm from the ground to reduce the risk of contamination with the soil and with a shredding length not exceeding 1 cm. Freshly chopped whole plant material was bagged (ensiled) into a silobag (2.7 m diameter × 10 m length) by a silage bagging machine (SBM) (mod. TurboPress Apiesse). The experiment was conducted without the addition of any exogenous substances. The silos were opened after 40 days, and five samples were carefully scraped from each variety at 1 m. Subsequently, the samples were pooled for each variety. All fresh and ensiled samples were oven-dried at 65 °C (for 24 h) and ground to pass a 1 mm screen (BrabenderWiley mill, Brabender OHG, Duisburg, Germany). The chemical composition of all samples was determined in three replications for each variety by a portable NIRS device (AgriNIR^TM, Dinamica Generale, Poggio Rusco, Mantova, Italy) [18]. The chemical bonds associated with absorbance in the silage samples supported the evaluation of dry matter (DM), crude protein (CP), ash, starch, neutral detergent fiber (NDF) and acid detergent fiber (ADF). Total sugars (TS) were determined according to Kiatti et al. [19].

2.2. Inoculum Preparation, Inoculation and Incubation

The substrates were evaluated in vitro using the gas production technique according to Calabrò at al. [17], performing two gas runs for each substrate on two consecutive days. Incubations were conducted in 120 mL serum flasks at 39 °C under anaerobic conditions. Four replicates were used for each substrate to measure gas production and to estimate OM disappearance at the end of the fermentation (72 h). In addition, four negative controls (buffered medium and rumen fluid only) were included. Starting twelve hours before inoculation, each fermentation flask containing 0.5 g of substrate and 89 mL of anaerobic medium [20] was closed with a butyl rubber stopper and an aluminum crimp seal and warmed at 39 °C in an incubator. Rumen liquor was obtained before the morning feed (7.30 a.m.) at an EU-authorized slaughterhouse from three adult Frisian dairy cows’ (650.2 ± 18.0 kg body weight) fed diet (NDF 45.5% DM and crude protein 12.5% DM). Rumen contents were obtained from multiple sites within the rumen and collected into a pre-warmed thermos flask. In the laboratory, the rumen contents of the three donor animals were pooled and shaken with a stirrer and strained through four layers of cheesecloth under CO₂. The liquid part of the rumen content was homogenized in a blender for 60 s under CO₂ and held at 39 °C until use. A syringe fitted with a 21-gauge needle was used to inject 5 mL of rumen fluid into each flask. Following inoculation, the displaced gas was allowed to escape, and flasks were placed in an incubator at 39 °C for 72 h.

2.3. Measurements of Gas and Residual Substrate

The gas production was recorded 0, 2, 4, 6, 9, 12, 15, 18, 21, 24, 28, 32, 36, 42, 48, 54, 60, 66 and 72 h after inoculation. Initial readings were taken at two-hour intervals due to the rapid rate of gas production. Gas measurements were taken using a manual pressure transducer (Cole and Palmer Instrument Co., Vernon Hills, IL, USA) calibrated to atmospheric pressure and connected to a Luer-lock three-way stopcock. At the end of the incubation, the flasks were opened, and from the contents of each flask, the pH was determined with a pH meter (model 3030 Alessandrini Instrument glass electrode, Jenway, Dunmow, UK) to verify the correct trend of the fermentation process. Samples (5 mL) were collected for the determination of volatile fatty acids (VFAs) and ammonia (NH₃-N). OM degradability (OM loss) was determined by filtering the residues using sintered glass crucibles (Gooch, porosity #2) under vacuum, drying for 5 h at 103 °C and ashing overnight at 550 °C. The cumulative gas volumes obtained at 72 h were related to the quantity of OM incubated (OMCV, mL/g) and gas released from controls. In addition, control flask fermentation residues were used to correct OM degradation. The data obtained from the cumulative gas production were fitted to the mathematical multiphasic model of Groot et al. [21]:

$$G(\mathrm{t})={\displaystyle \frac{A}{1+\frac{B^C}{t^C}}}$$

where G represents the amount of gas (mL) produced per g of incubated OM at time t after incubation; A (mL/g) denotes its asymptotic or potential value (at t = ∞); B (h) is the time after incubation at which A/2 has been formed; and C is a constant determining the sharpness of the switching characteristic of the gas production profile. Assuming a linear relationship between gas production and substrate fermentation, the time to reach the maximum rate (t_RM, h) and the maximum rate itself (R_M, %/h) can be calculated with the second derivate [21]:

t_RM = B × [(C − 1)/(C + 1)]^1/C

R_M = [A × B^C × C × t_RM^(−C−1)]/[(1 + B^C) × t_RM^(−C)]².

2.4. Fermentation End Products

For VFA determination, samples of fermenting liquors were collected at the end of incubation and centrifuged at 12,000× g for 10 min at 4 °C (Universal 32R centrifuge, Hettich Furn Tech Division DIY, Nussloch, Germany). VFAs were measured via gas chromatography (Thermo Quest 8000top Italia SpA, Rodano, Milan, Italy; fused-silica capillary column, 30 m, 0.25 mm ID, 0.25 m film thickness), using an external standard solution composed of acetic, propionic, butyric, isobutyric, valeric and isovaleric acids. The branched-chain fatty acid ratio (BCR) was calculated as follows: (isobutyric acid + isovaleric acid)/tVFAs. The non-glycogenic ratio (NGR) was calculated as follows: [(acetic + 2 ∗ butyric)/propionic)]. For the NH₃-N analysis, 5 mL of the sample was taken to which 5 mL of 10% trichloroacetic acid (TCA) was added. NH₃-N concentration was determined using the indophenol method, as described by Searle [22]. The NH₃-N concentrations were corrected with the values of the blank and the standard samples. The correction for the blank (bottle without substrate) was performed by subtracting the amount of NH₃-N produced by the blank from the observed NH₃-N value of every bottle.

2.5. Statistical Analysis

The chemical compositions of the five maize varieties, both fresh and ensiled, were compared using ANOVA. Differences within the same variety in fresh and ensiled forms were also analyzed using ANOVA. The fermentation parameters were subjected to a statistical analysis to detect the ensiling influence within the five varieties using the following model:

y_ijk = μ + E_i + V_j + E × V_ij + ε_ijk

where y is the dependent variable, μ is the general mean, E is the ensiling effect (i = fresh and ensiled), V denotes the maize varieties (j = 1–5), E × V is the interaction among the ensiled varieties, and ε is the error term. The gas run replication was not significantly different and was therefore not taken into consideration. For the statistical analysis, the GLM procedure (SAS, 2000) was used, fixing the significance level at p < 0.01.

3. Results and Discussion

The chemical composition of the five maize varieties, in both their fresh and ensiled forms, is presented in

Table 1. Chemical composition (g·kg⁻¹) of the five maize varieties, fresh and ensiled.

Variety	DM	CP	Ash	Starch	NDF	ADF	TS
Fresh
Tiesto	290 ± 21	73 ± 3	45 ± 4	325 ± 22	428 ± 38	243 ± 21	105 ± 12
R700 1	300 ± 22	74 ± 2	40 ± 7	311 ± 24	395 ± 36	228 ± 26	79 ± 8
MAS 78.T	285 ± 19	72 ± 3	42 ± 5	318 ± 23	398 ± 41	228 ± 28	70 ± 9
DKC 7074	298 ± 20	78 ± 4	44 ± 4	294 ± 19	401 ± 40	238 ± 28	76 ± 11
KWS Kantico	301 ± 23	71 ± 3	41 ± 4	314 ± 19	402 ± 40	227 ± 30	70 ± 10
Ensiled
Tiesto	303 ± 21	79 ± 2	44 ± 5	324 ± 24	428 ± 42	231 ± 21	<12
R700 1	303 ± 22	81 ± 3	46 ± 5	309 ± 27	443 ± 43	242 ± 22	<12
MAS 78.T	332 ± 23	83 ± 3	44 ± 7	328 ± 28	423 ± 38	232 ± 22	<12
DKC 7074	322 ± 25	86 ± 4	47 ± 5	303 ± 28	434 ± 37	242 ± 25	<12
KWS Kantico	332 ± 21	76 ± 3	45 ± 4	335 ± 22	421 ± 37	230 ± 19	<12

DM: dry matter; CP: crude protein; NDF: neutral detergent fiber; ADF: acid detergent fiber; TS: total sugars.

. In general, all the parameters are similar to the findings of Crovetto et al. [23] in four Italian varieties of forage maize (Constanza, Proxima, Frassino, Eleonora), which differ in the type of starch; nevertheless, some differences were observed compared to our previous data [18]. Indeed, in the work of Zicarelli et al. [18], maize silage samples were obtained from three different areas of the Campania region and scored higher DM and lower CP content compared to the present trial. No significant differences were observed among varieties. This suggests that, under similar growing and harvesting conditions, the selected varieties exhibit comparable nutritional profiles, which is advantageous for livestock feed formulation, as it ensures reliability and predictability in diet planning. As expected, the ensiling process led to substantial changes in the composition of the maize; in particular, significant differences (p < 0.001) were observed for the total sugars (TS), which in large part disappeared. This phenomenon can be attributed primarily to the microbial fermentation of carbohydrates [24], which are converted into lactic acid and volatile fatty acids (VFAs) during the ensiling process. Consequently, the relative concentration of most other components increased by 6–8%.

The in vitro fermentation parameters and the end products of the five maize varieties, both in their fresh and ensiled forms, are depicted in

Table 2. Fermentation kinetics of the fresh and ensiled maize per variety at 72 h.

Variety	OMCV	B	tRM	RM	OM Loss	pH
	mL·g−1	h	h	%/h	%
Fresh
Tiesto	302	20.7	12.5	10.24	76.4	6.51
R700 1	292	20.6	12.7	9.96	76.9	6.57
MAS 78.T	282	20.9	13.4	9.61	77.4	6.56
DKC 7074	281	22.1	13.0	8.85	74.9	6.54
KWS Kantico	296	21.3	13.9	9.98	77.0	6.56
MSE	75	0.60	0.33	0.35	6.76	0.004
Ensiled
Tiesto	286	21.3	15.1	10.12	78.1 AB	6.52
R700 1	281	21.1	14.6	9.84	74.6 B	6.51
MAS 78.T	291	21.5	15.1	10.09	76.0 AB	6.54
DKC 7074	272	22.3	14.3	8.95	77.5 AB	6.49
KWS Kantico	292	20.4	14.4	10.7	79.7 A	6.48
MSE	195	1.42	0.49	0.73	2.66	0.002
Ensiling effect	NS	NS	NS	NS	NS	NS

OMCV: cumulative gas volumes obtained related to organic matter (OM) incubated; B: the time after incubation at which half of potential gas has been formed; t_RM: the time to reach the maximum rate; R_M: the maximum rate. MSE: mean square error. Means with different capital letters in the same column differ significantly at p < 0.01. NS: not significant.

and

Table 3. Fermentation end products of the fresh and ensiled maize per variety at 72 h.

Variety	Acetic	Propionic	Butyric	VFA	BCR	NGR	NH3-N
	mM·g−1
Fresh
Tiesto	4.68	2.38	0.836	8.31	0.097 AB	6.76	1.73
R700 1	4.52	2.63	0.827	8.36	0.088 B	6.58	1.71
MAS 78.T	4.63	2.79	0.822	8.63	0.09 AB	6.70	1.71
DKC 7074	4.44	2.74	0.826	8.45	0.105 A	6.58	1.70
KWS Kantico	4.54	2.85	0.847	8.61	0.087 B	6.63	1.68
MSE	0.034	0.036	0.003	0.074	0.00003	0.083	0.005
Ensiled
Tiesto	4.29	2.46	0.999	8.15	0.095 B	6.71	1.91
R700 1	4.25	2.56	0.941	8.19	0.105 B	6.58	1.90
MAS 78.T	4.32	2.36	0.952	8.05	0.103 B	6.64	1.92
DKC 7074	4.49	2.54	0.935	8.38	0.098 B	6.76	1.90
KWS Kantico	4.35	2.48	1.074	8.46	0.126 A	7.07	1.84
MSE	0.031	0.065	0.01	0.11	0.00006	0.17	0.003
Ensiling effect	NS	NS	NS	NS	NS	NS	NS

VFA: volatile fatty acid; BCR: branched-chain ratio; NGR: non-glycogenic glucogenic ratio. NH₃-N: nitrogen ammonia. MSE: mean square error. Means with different capital letters in the same column differ significantly at p < 0.01. NS: not significant.

, respectively. The gas produced after 72 h (OMCV) is on average equal to 283 mL·g⁻¹, which is slightly lower than the 323 mL·g⁻¹ observed by De Boever et al. [25] as a mean of 30 samples of maize silage originating from 11 different varieties but higher than the asymptotic gas production of 246 mL·g⁻¹ measured by Lovet et al. [26] in 12 forage maize cultivars. For all the substrates tested, the ensiling process resulted in a slight reduction in gas production (OMCV). Additionally, fermentation kinetic parameters B and t_RM were slightly affected by the ensiling process. Although these changes were not statistically significant (p > 0.01), a trend toward longer fermentation times was observed. All these differences indicate a slowing down of the fermentation process. This is most likely due to the well-known utilization of soluble sugars during the ensiling process. Among the analyzed parameters, OM loss was the only variable that showed a statistically significant difference (p < 0.01) among varieties in their ensiled form. Specifically, variety 5 (KWS Kantico) exhibited the highest OM loss, while variety 2 (R700 1) showed the lowest values. This difference aligns with their respective NDF and starch content. This parameter, approaching around 80%, was quite similar to the value observed by De Boever et al. [25]. This parameter, corresponding to in vivo organic matter degradability, is essential for assigning nutritional value to feedstuffs and therefore formulating balanced diets to be administered to animals. Another important aspect of fermentation is the pH stability of the incubation medium, as it directly influences microbial activity and digestion efficiency. The pH values recorded after 72 h of incubation ranged from 6.48 to 6.57. These values indicate that the buffering capacity of the medium was consistently sufficient to maintain the fermentation environment within the range required to ensure a favorable environment for cellulolytic bacteria [27].

In general, the total VFA production showed no differences among varieties. This suggests that despite compositional variations among the different cultivars, the overall fermentation process followed a similar pattern, leading to a relatively uniform production of VFAs, the primary energy source for ruminants. However, notable differences were observed when analyzing the breakdown of protein, as reflected by the BCR. This parameter can be useful in formulating a diet for animals, as it is an index of protein degradation at the rumen level, despite the moderate protein intake of maize silage. In fresh maize samples, the BCR values were significantly higher (p < 0.01) in variety 4 (DKC 7074) and lower in variety 5 (KWS Kantico), indicating that protein degradation occurred at different rates among the varieties. Interestingly, when maize was ensiled, variety 5 exhibited the highest BCR values (p < 0.01). These results are also in accordance with the protein content of these varieties. According to Iommelli et al. [28], differences in CP degradation can also be observed according to different genotype and treatment. The ammonia produced was significantly higher in the ensiled substrates compared to fresh maize. The effects observed on BCR and NH₃-N can be explained by the correlation between these parameters and the higher crude protein content of the silage in comparison to the fresh material. According to Van Soest [29], ensiling generally leads to higher ammonia levels than those found in the fresh plant. This may result from the proteolysis by microbes that occurs during fermentation [30]. In the cut forage, the plant proteases and peptidases are also active and are the principal enzymes responsible for the conversion of true protein to NPN in ensiled feeds [31]. However, an alternative and more likely explanation is that, because of the disappearance of part of the substrate during ensiling, notably soluble sugars, less energy is available for the capture of NH₃ by the rumen microbes.

Figure 1 and Figure 2 show the gas production profiles and fermentation rates over time for the fresh and ensiled material, Panel A and Panel B, respectively. The curves resulted from the average values recorded at each time point for every fermentation bottle, providing a comprehensive representation of the overall fermentation dynamics. The overall shape of the gas production curves remained relatively consistent between fresh and ensiled samples, indicating that the general fermentation pattern was not drastically altered by the ensiling process. The fermentation rate closely followed the trends in gas production, with minor variations among the different maize varieties. In general, fresh samples exhibited a slightly higher fermentation rate compared to their ensiled counterparts. This difference can be attributed to the presence of soluble sugars in the fresh material, which serve as readily available energy sources for microbial fermentation.

4. Conclusions

The present study, in accordance with previous findings, confirmed that the chemical composition and fermentation characteristics differed between fresh and ensiled maize in all the different varieties tested, demonstrating that the ensiling process significantly alters both substrate availability and fermentation kinetics. While the ensiling process depletes soluble sugars, it enhances ammonia production, suggesting an increase in crude protein degradation during the storage. The reduction in gas production and fermentation rate in the ensiled material further indicates a shift in microbial substrate utilization, primarily due to the absence of readily fermentable carbohydrates. The variations in organic matter loss and branched-chain ratio (BCR) among varieties highlight the influence of fiber and protein content on fermentation dynamics. Overall, these results emphasize the need for detailed nutritional assessments of maize silage to ensure optimal dietary formulations for large ruminants. Future research should further explore the impact of variety-specific characteristics on silage digestibility and efficiency in ruminant diets, with a particular focus on balancing different energy sources (e.g., starch and structural carbohydrates) and availability. The results obtained can be considered valid despite the limitations with respect to in vivo methods (i.e., closed vs. open and dynamic system). However, the data obtained should be confirmed with tests to be carried out directly in vivo.