1. Introduction
Tef (
Eragrotis tef (Zucc.) Trotter, also known as teff) is a cereal plant of Ethiopian origin, known as an agricultural crop for over 2000 years [
1]. Tef is a major source of food and feed in Ethiopia. It is a basic product providing a significant portion of the daily diet (bread) of the population and animal feed, occupying about 30% of the land used to cultivate grains in the country [
2]. Tef is an annual cereal plant with an upright stem; inflorescences are oat-like panicles which may be free or compact and comprise 2–12 flowers per spikelet [
3]. The flowers are bisexual with about 99% self-pollination. Tef has very small grains and 1000 seeds have a mass ranging from 0.25 to 0.35 g with varieties differing in grain colour from white to dark brown. Tef has a C4 photosynthetic pathway, which enables high productivity under high radiation and temperature. Tef can grow in a wide variety of environments (marginal lands inclusive) [
4,
5,
6]. The plant can adapt to low rainfall conditions (450–550 mm), and it can tolerate moisture stress [
3]; however, under severe drought there can be a great yield reduction (over 70%) [
2]. On the other hand, excess irrigation leads to plant lodging, causing loss of yields, and hampers mechanical harvest [
7,
8,
9,
10].
In recent years, many countries outside Ethiopia (Israel-inclusive) have increased interest in tef cultivation as a grain or fodder crop [
11,
12]. In Israel, the interest in tef is derived from various reasons. There is a demand of approximately 8000 tonnes of grains per annum by the Ethiopian ethnic community, while the Ethiopian government restricts the export of tef. There is also potential in domestic and export “health food” markets for tef kernels and products [
9]. Lastly, there is a great interest to diversify the field crops in general as well as the ruminant fodder crops in Israel. Dairy cows’ feeds in Israel are based on locally produced roughage (fibrous material), mostly wheat and maize, complemented with expensive imported concentrate (rich in energy and digestible nutrients). The introduction of tef as a new, high quality fodder crop will therefore contribute to the diversity of the forage crops.
Tef straw is preferred by livestock relative to other cereal straw, and hence could be used as livestock feed during the dry seasons when there is a shortage of feed, fetching a premium price [
3,
13]. Research has shown that the nutritional value of tef fodder is like other fodder used to make hay and silages [
14,
15] with a digestibility value similar to tropical grasses [
16]. Tef has a higher or similar crude protein (CP) content compared to other cereal grains [
17,
18] as well as high trace minerals such as calcium, phosphorus, iron, etc., and thiamine [
19]. Hence, it can be used as an alternative summer forage [
17].
Silage is one of the most reliable methods to preserve feedstuffs. Over the years, there has been a better understanding of the biochemistry and microbiology of the ensiling process paving way for development of numerous silage additives [
20,
21]. High nutritive value of silage depends on harvesting the crop at the proper stage of maturity, activity of plant enzymes, epiphytic microorganisms, and encouraging dominance of lactic acid bacteria (LAB) [
22]. Research has shown that the various additives had positive effects on silages as well as animal performance, although this greatly varied [
23,
24].
As part of a project that aims to assess tef as a multi-harvest summer crop for Israeli dairy cows, the objective of this study was to evaluate the ability to preserve tef as silage as well as assess the effects of cultivars, irrigation regime, and additives on tef silage.
4. Discussion
Tef is known worldwide as a multi-harvest crop with high production capacity and outstanding fodder quality [
14,
15,
18] that could be used as an alternative summer forage [
17]. We conducted an experiment to assess the possibility of ensiling tef and the effects of irrigation, genotype, and additives on tef silage. To the best of our knowledge, this is the first report of a study assessing the effects of these factors on tef silage.
On the day of opening the silage bags (day 0), they had a sweet smell, and the colour of the silage was greenish brown with no signs of moulds. This was in contradiction with a study of [
42] in guinea grass and cassava tops silages, in which the greenish silage was better than the greenish brown colour which was associated with moulds. The variation in the colour of the silage could probably depend on the ensiling material (plant species) and growth conditions.
It is worth noting that moulds and yeast were not detected on the day of opening or 5 days later, which may indicate the high quality of the ensiling process (
Table S1); however, this phenomenon could be related to the fact that tef grass has remarkable low WSCs content, which did not encourage growth [
43]. The moulds and yeast mainly utilise sugars as a substrate [
20], and hence could hardly thrive on tef silage. This was also observed during the aerobic exposure of tef silage for five days with production of negligible carbon dioxide (data not shown) without any sign of aerobic spoilage. This is desirable as it minimises losses during the practical feed out phase of the silage.
The pH values at the opening time were lower in 75% irrigation than 100% irrigation regime (
Table 3). This could be explained by low moisture content which often leads to better silage fermentation by increasing the counts of LAB, which are essential for rapidly dropping pH, more homogenous silage, selectively inhibiting butyric acid fermentation which is associated with increased pH [
44,
45,
46]. The pH results are similar to the findings of [
47], who reported that generally grass silages have a pH of 4.3–4.7. They are also similar to the findings of [
46], who reported lower pH values in wilted grass in combination with LAB as an inoculant. The pH of tef silage was slightly greater than that of corn silage, which is usually less than 4 [
48,
49,
50,
51], but less than that reported in lucerne [
45], as well as less than that reported in ensiled high-moisture corn after 120 days as reported by [
52]. The pH of our tef silage was also lower than a mixture of Timothy and Meadow Fescue with additives [
53] as well as orchard grass and alfalfa [
46]. This could imply that each ensiling crop has different fermentation characteristics, and other factors should be considered in the assessment of the quality of silages. The differences in the pH values were also found between the four genotypes (
Table S1), which could be related to difference in their characteristics [
3,
13,
25,
54,
55,
56]. In addition, the lowest pH values were achieved with the addition of MOL + HI to the tef silage. The lowest pH in MOL + HI could be a result of the synergistic effect between molasses and HI leading to better and rapid fermentation. Assessment of pH after the aerobic stability experiment was even lower for all the additives. This could imply that tef silage has a good buffering capacity and low WSCs, which hamper the multiplication of aerobic microorganisms that could deteriorate the silage. In addition, higher concentrations of acetic acid and VFAs in MOL + HI could act as silage preservatives as described by [
57,
58,
59,
60].
In vitro DM digestibility values of tef silage (
Table 3) were almost similar to those of whole corn silage as reported by [
61]. The IVDMD results were also similar to alfalfa silage [
62] and five varieties of sorghum silage [
63]. These results imply that tef silage could be used like some of the most common silages. The IVDMD values obtained for the 75% irrigation treatment were greater than those of the 100% irrigation, possibly because of better fermentation with low moisture levels [
44,
45]. Looking at additives, MOL + HI had the greatest IVDMD, which could be attributed to the synergistic effect between molasses and HI and made nutrients more available for digestion as suggested by [
64]. This same trend was observed after the exposure to aerobic conditions for irrigation treatments with IVDMD still higher at 75% than 100% irrigation, though the IVDMD values were lower after 5 days of aerobic exposure. These slightly lower IVDMD values could be attributed to the slight DM loss during the aerobic exposure experiment. However, it was the opposite while looking at the additives; IVDMD was higher in CTL compared to others. This could be as a result of the increased microbial activity that led to more DM degradation in the other three additives than the CTL. The variation in IVDMD among genotypes could be as a result of the differences between the ratio of stems and leaves [
3,
13,
25,
54,
55,
56]. A high stem to leaf ratio could imply less digestibility, while a lower one could imply more digestibility because the leaves are less lignified and hence more digestible than the stems.
The IVDMD values below 50% in two genotypes (RTC-2 and RTC-361) at 100% irrigation following 5 days of aerobic exposure could be attributed to the high indigestible NDF content that was found in these genotypes. This is supported by the marginal differences of digestible NDF between 48 h vs. 240 h of incubation for the varieties as shown in (
Table 2).
LAB counts were lower in 75% than 100% irrigation, contradicting with the findings of [
44,
45] who found more LAB in wilted (lower moisture) silage with lactic acid producing bacteria. This could be attributed to more enzyme activity because most microbial and enzyme activities are controlled by moisture. In addition, the results of lower moisture (wilting) are inconclusive. Although the silage materials had similar DM contents, the materials from 100% had probably more tissue moisture from irrigation than the ones that originated from 75% irrigation. The difference in the genotypes could be attributed to the variation among the tef lines selected for the experiment, as tef is also diverse as reported by [
3,
13,
25,
54,
55,
56]. A higher LAB count in HI treatment relative to the other three treatments is most probably because of the LAB component of HI.
Total VFA concentrations were lower in tef silage than corn silage [
65]. The concentrations in MOL + HI were consistently higher in individual VFAs as well as total VFA compared to the other additives. This could imply the synergy between HI and molasses [
28]. Certain VFAs, such as butyric and acetic acid, have been implicated in depressing feed intake, although there are some inconsistences [
45], hence their lower concentrations might be beneficial. This same trend was maintained during the aerobic stability treatment. However, the VFAs were lower probably because some were lost during the aerobic stability exposure.
Lactic acid was less than 1% in all four treatments, but it was higher in MOL and MOL + HI [
28]. This could be attributed to molasses acting as a source of energy for the LAB producing more LA [
20] and the synergistic effect between MOL + HI. In comparison between tef silage, LA in other silages from corn, legumes, and grasses was higher (2% or greater) [
47,
65,
66], or almost similar in whole-maize silage (about 1%) and high-moisture corn (0.5–2%) [
47]. LA concentration increased during the aerobic stability experiment and was greater at 75% irrigation, which could be attributed to less moisture leading to rapid fermentation activity as well as low WSC [
43], which could not support aerobic microorganisms to break down the LA. Phenotypic and genotypic differences could still be the cause of the variation in LA concentrations among the genotypes [
3,
13,
25,
54,
55,
56].
Water-soluble carbohydrates in tef silage was similar to that in high moisture corn silage (<0.10%) as reported by [
52], but less than that of corn ensiled with potassium sorbate [
65] and those reported on the effect of damaged ears of corn and additives on corn by [
66]. WSCs were greater in 75% irrigation. This could be due to lower moisture content leading to better silage fermentation characteristics [
67,
68,
69]. The difference among genotypes could be attributed to the tef diversity reported by [
3,
13,
25,
54,
55,
56]. The lower WSC contents among treatments (0.04 to 0.16 g/100 g DM) and particularly lower in HI and MOL + HI could be attributed to more efficient ensiling where the microorganisms used the sugars as their source of energy. This phenomenon is desirable because the microorganisms’ mass would serve as a rich source of proteins to the animals. The difference among genotypes could still be attributed to variation between of the tef lines. However, the WSC were still lower in HI, MOL, and MOL + HI, which could be attributed to utilisation of the sugars for microbial survival and growth.
Ammonia nitrogen concentration in silages was less in 75% than 100% irrigation. Ammonia nitrogen is a result of proteolysis of amines and other sources of proteins [
20]. High N-NH
3 concentrations could depict less efficient fermentation, and lower ones could depict the conversion of the N-NH
3 and its derivatives into microbial proteins, which benefits livestock and prevents N wastage [
70]. The variation in the N-NH
3 concentration could still be attributed to diversity [
3,
13,
25,
54,
55,
56], whereas the low concentrations in MOL + HI would imply more efficient utilisation of N-NH
3 and its derivatives due to the synergy between MOL + HI, as the HI would provide microorganisms for fermentation and molasses would serve as source of available energy. It is also worth noting that there was even more N-NH
3 at 100% irrigation during the aerobic stability experiment. This increase could be attributed to the likely greater moisture content of the plant material in 100% than 75% irrigation. More moisture content would imply more enzyme and microbial activity leading to more proteolysis; hence, increase in N-NH
3 concentration and the variation among genotypes could also still be as a result of the diversity in the tef germplasm.
The OM content of tef silages was slightly more in 100% than 75% irrigation, this could be attributed to the faster fermentation where there is more moisture averting losses [
47]. The slightly lower OM content of 75% irrigation and lower OM content in HI, MOL and MOL + HI relative to the control could be because the OM was used by the microorganisms. There was a slight variation among genotypes which could still be attributed to differences between lines [
3,
13,
25,
54,
55,
56]. This trend was maintained during the aerobic stability experiment.
Crude protein content in silages was slightly less in 100% than 75% irrigation. This could be because more moisture increases enzyme activity [
47], which could have led to more proteolysis than in 75% irrigation. This was demonstrated by the higher N-NH
3 concentration at 100% vs. 75% irrigation at both day 0 (3.5 vs. 5.0 mgdL
−1) and day 5 (2.5 vs. 5.6 mgdL
−1, respectively). A higher N-NH
3 concentration in silage is undesirable because it is an indicator of protein degradation as suggested by [
20]. It is worth noting that although higher in 100% than 75% irrigation, the N-NH
3 concentration was still below what is usually observed in other silages (8–15%) as reported by [
70]. This could imply less protein loss which is desirable in silages. The slight variation among genotypes may still be attributed to tef germplasm diversity [
3,
13,
25,
54,
55,
56]. A similar trend was observed during the aerobic stability exposure. The CP was higher in HI, MOL, and MOL + HI than the control during the aerobic stability experiment. The higher CP could be as a result of microbial proteins accumulated as a result of the additives supplement [
64,
71].
The cell wall carbohydrates content (NDF, ADF, hemicellulose, and cellulose) in silages and raw material were consistently higher in 100% than 75% irrigation. However, lower fibre content in silage could be attributed to utilisation by microorganisms toward more efficient fermentation [
44,
45]. Research has shown that the next source of fermentation microorganisms after depletion of WSCs is hemicellulose [
20]. Low fibre contents and higher digestible material in silage are desirable because they encourage feed intake and reduce the amount of indigestible fibre. Differences of fibre content among genotypes may still be because of the differences between tef lines. A comparison among silages with additives showed that MOL + HI and HI had less fibre content, which could be indicative of more fibre utilization and efficient fermentation. These results are similar to [
64] who reported better fermentation parameters with additives. It is worth noting that after the aerobic stability experiment, the cellulose content was slightly higher in the other three additives treatments relative to the CTL. This could be either because of more DM breakdown leading to concentration of cellulose or more stable fermentation conditions leading to less loss of cellulose.
The ADL content in silages were almost thrice higher in both irrigation treatments than the original ensiling material, but higher in 100% than 75% irrigation. This could be attributed to increased DM loss at more moisture content due to increased microbial and enzyme activity [
47] increasing the concentration of ADL. The ADL content was even greater at 100% irrigation after aerobic stability experiment implying more DM loss, whereas it was rather less at 75% irrigation, which could depict better silage fermentation [
44,
45,
67,
69] minimizing accelerated DM loss. However, due to a small loss (2–4%) in DM in our experiment, the threefold increase in ADL could not be explained by DM loss, but could be rather a result of artefacts because ADL is hardly degraded or produced during ensiling. Considering the genotypic effect on tef silage, there were variations among genotypes which could be related to the diversity in the tef germplasm, and a further comparison among additives showed that a higher content was seen in the CTL relative to the other treatments. These high ADL contents could be attributed to a more accelerated DM loss in the CTL than other treatments because the additives often mitigate DM losses as reported by [
72].
The DM losses were minimal (up to a maximum of 2% and 4% on day 0 and day 5, respectively) This is less than what was observed in other silage crops such as corn, ryegrass and wheat that had between 4 to 11% DM loss [
41]. The DM losses in tef silage are also much lower than what was postulated (15 to 30%) by [
73]. The low DM losses in tef silage could be attributed the lack of moulds and yeast that are usually implicated in silage deterioration, proper DM content (approximately 35%) during ensiling, and creation of anaerobic conditions [
20,
22,
73] achieved by vacuum sealing. The lower DM loss in tef silage would imply a reduction in wastage and more efficient utilisation of resources. However, the methodology used herein (vacuum-sealed bags) could be a good reason for the minimum DM losses.
A comparison between the two irrigation treatments showed that most of the silage parameters were better at 75% than 100% irrigation. This could be attributed to more rapid vegetative growth and higher yield [
74] at 100% irrigation, which would need support structures (lignin). Our analysis of silage raw materials showed that lignin content was higher in 100% irrigation than 75%. Lignification usually reduces the quality of the forage [
74,
75]. This could be one of the reasons that better silage parameters was achieved at 75% than 100% irrigation. In addition, having better silage parameters at 75% irrigation is desirable, as this would save of the forage production costs (less water usage).