Effects of a dual-purpose bacterial inoculant on the fermentation characteristics of high-moisture maize silage and dairy cattle performance

________________________________________________________________________________ Abstract The objective of this study was to investigate the effects of inoculating Lactisil Maize, a dual-purpose inoculant, on the fermentation characteristics and nutritive value of high-moisture maize silage and the performance of lactating cows receiving the silage in their diets. Whole-crop maize was harvested at 253 g dry matter (DM)/kg fresh crop. Maize silage was produced with or without Lactisil Maize and ensiled in two different bunkers. Eight multiparous lactating Holstein dairy cows were used in a replicated 2 × 2 Latin square experimental design and were fed total mixed rations that contained 230 g/kg of either inoculated or control maize silage. Inoculation did not affect the nutritive value or the aerobic stability of the maize silage, but increased the neutral detergent and acid detergent fibre fractions of the silage. However, inoculation increased the concentrations of acetic acid and lactic acid, but reduced ammonia N concentration compared to the control. Cows fed the Lactisil Maize-inoculated silage had a lower DM intake, milk yield, and yields of milk fat and milk protein were lower, compared to control. The digestibility of nutrients was not affected by inoculation. It was concluded that although Lactisil Maize inoculation improved the fermentation quality of high-moisture maize silage, it did not improve aerobic stability of the silage or the production performance of dairy cows compared to the control. _______________________________________________________________________________________


Introduction
Ensiling is a common preservation method for moist forage crops.It is based on the anaerobic fermentation of water-soluble carbohydrates (WSC) into lactic acid and volatile fatty acids (VFA) by lactic acid bacteria (LAB).Maize is one of the major forages used for ensiling due to its excellent nutritional quality and good ensiling properties (Allen et al., 2003).Maize silage is used as a source of effective fibre and fermentable energy that makes up to 50% of the forage source in the rations of dairy cows in many commercial dairies in the world and in Iran (NRC, 2001;Kowsar et al., 2008).However, maize crops grown in Iran are usually planted in July as a second crop.Therefore, it does not get sufficient time to mature and is ensiled with a low dry matter (DM) (20 to 25%) content (Khorvash et al., 2006).Moreover, after opening the silo for feeding, exposure to oxygen leads to aerobic deterioration of the silage.Maize silages are susceptible to spoilage because of higher concentrations of lactic acid and WSC (McDonald et al., 1991).This condition forces farmers to improve their management during crop harvesting and ensiling, e.g. using various types of silage additives including bacterial inoculants.The biological additives for silages are safe and easy to use, noncorrosive to machinery, do not pollute the environment and are regarded as natural products (Filya et al., 2000).During the last two decades, many microbial inoculants have been developed in order to improve the ensiling process and silage quality.Homofermentative and heterofermentative LAB inoculants are two types of inoculants commonly used in silage production.Homofermentative LAB such as Lactobacillus plantarum, generally increases lactic acid and decreases acetic acid, butyric acid and ammonia-nitrogen (NH 3 -N) levels, and the pH of the silage (Sheperd et al., 1995;Aksu et al., 2004).However, homofermentative LAB often increases aerobic deterioration of whole-crop cereal silages, probably because of insufficient VFAs to inhibit fungal growth (Weinberg et al., 1993;Filya et al., 2000;Kleinschmit & Kung, 2006b).
Heterofermentative LAB usually improves the aerobic stability of silage (Driehuis et al., 1999;Filya, 2001;Kung & Ranjit, 2001;Weinberg et al., 2002;Filya, 2003b).Lactobacillus buchneri is the main heterofermentative LAB inoculant most widely used during the ensiling of forages (Muck, 2008).Due to the fact that L. buchneri cannot be found in high quantities in most forages (Torriani et al., 1992), isolates of this bacterium have been identified and developed into silage inoculants (Muck, 1996).In recent years, dualpurpose inoculants have been developed to enhance both the efficiency of anaerobic fermentation and to improve aerobic stability (Hu et al., 2009).Lactisil Maize, a new commercial multi-species bacterial inoculant, consists of homofermentative (Enterococcus faecium M74, L. plantarum LS1, Lactobacillus casei and Pediococcus pentosaceus) and heterofermentative LAB (L.buchneri).For optimal fermentation, E. faecium and P. pentosaceus require a relatively high pH and, thus, start to produce lactate at a high pH in forage crops.This leads to a faster decrease in pH which provides better environmental conditions for L. plantarum and L. casei who require a lower pH for optimal fermentation.In the later phase of silage production, L. buchneri starts to produce acetic acid, resulting in an increase in aerobic stability of silages.The aim of the current study was to investigate the effects of Lactisil Maize on the chemical composition and fermentation characteristics of high-moisture maize silage and on the performance of lactating dairy cows.

Materials and Methods
Whole-crop maize (hybrid 700; Plant Breeding, Karaj, Iran) was harvested at 253 g DM/kg of fresh crop, and chopped to theoretical length of 2.5 cm, using a pool-type chopper (Model 965, Claas, Omaha, Netherland).The maize silage was produced with or without an inoculant, and treatments were the control (no inoculant) and inoculation with Lactisil Maize (Medipharm, Kågeröd, Sweden).Two bunker silos (with about 300 t capacity for fresh maize crops) were filled with the chopped maize, one with the inoculant, the other with inoculant-free water.The inoculant was applied layer by layer as the crop was unloaded from the silage wagon into the silos, followed by heavy rolling during the packing process.For each 50 t of crop, 500 g of inoculant was diluted into 200 L of water and sprayed over the freshly-chopped maize to obtain at least 1.5×10 5 colony-forming units (cfu) per g of fresh forage.For the control treatment, the same amount of inoculant-free water was applied.After 230 days of ensiling, both silos were opened and the silages were analyzed for chemical composition (Table 1) and fed to dairy cattle.
Eight multiparous, lactating Holstein cows (107 ± 15 d in milk and producing 41.8 ± 3 kg milk/d) were used in a replicated 2 × 2 Latin square experimental design with one of the two diets.The cows were allocated to squares based on days in milk, and randomly assigned to treatments within a square.They were offered the diets (Table 2) twice daily at 09:00 and 16:00 at a level to allow a maximum of 10% refusals.Diets were formulated based on the NRC (2001) recommendations to supply sufficient energy and protein for a 600 kg cow producing 40 kg/d of milk, containing 3.5% fat and 3.2% protein.The cows were housed in individual stalls at the dairy facilities of the Lavark Research Station (Isfahan University of Technology, Isfahan, Iran).The experiment consists of two periods of 21 days; the first 14 days for adaptation and the last 7 days for sampling and data collection.The individual stalls (4 × 4 m) were equipped with concrete feed bunkers.The cows had ad libitum access to fresh water and a salt block.Clean wood shavings and sand were used as bedding and replaced twice a day.The animals were handled according to the guidelines of the Iranian Council of Animal Care (1995) at the Dairy Facilities of the Lavark Research Station.
After 230 days of ensiling, about 15 kg of the silage was sampled at six sites from each silo, and was thoroughly mixed to obtain a representative sample of ca. 3 kg.About half of this sample was used for the determination of aerobic stability and the other half was frozen until further analysis for biochemical characteristics could be completed.At the beginning of each experimental period, the DM content of maize silages was determined prior to diet formulation.
The cows were milked three times daily in a milking parlour at 04:00, 12:00, and 20:00, with no provision of water or concentrate during milking.At the 12:00 milking session, the cows were allowed a 30 min exercise period outside, before being milked.During the last five days of sampling, the milk yield of all cows was recorded.The milk yield of each cow at each milking session, was measured using special graduated jars (Agri & SD Co., Frankfurt, Germany).Before each milking session, the cows were monitored for udder inflammation and the presence of milk clots in the nipples to ensure that the milk yield and composition were not affected by mastitis.Milk from individual cows was sampled at each milking session in pre-labelled 50 mL plastic vials and preserved with potassium dichromate until it could be analyzed for milk components.
Samples of the diets and orts were collected during days 14 -21 of each period.The feed intake was calculated as the difference between feed offered and orts collected.Grab faecal samples were taken daily from the rectum during data collection.The daily collected orts and faeces from each cow were composited.Feeds, orts and faecal samples were stored at -20 °C pending chemical analysis.Body weights were recorded on days 19 to 21 of each period at the same time each day.
On the last day of each period, rumen fluid was sampled at 4 h post-feeding to monitor rumen fermentation, by using a stomach tube.The initial 400 mL of the fluid aspirated was discarded to minimize saliva contamination.The pH of the second portion was measured immediately, using a mobile pH meter (HI8314, Hanna Instruments, ClujNapoca, Romania), and 10 mL of the fluid was preserved with 1 mL of 5% sulphuric acid and frozen at -20 °C for VFA analysis.
The analysis on the pre-ensiled maize, silage, feeds, orts and faecal samples were done in triplicate.Dry matter was determined in a drying oven for 72 h at 60 °C.The dried samples were ground to pass through a 1 mm screen using a Wiley mill (Arthur H. Thomas Co., Philadelphia).Organic matter was determined by incinerating the material at 550 °C for 12 h.Total nitrogen (N) was determined by using the Kjeldahl method (Kjeltec 1030 Auto Analyzer, Tecator, Höganäs, Sweden).Neutral detergent fibre (NDF) was determined according to the method of Van Soest et al. (1991) and the concentration of acid detergent fibre (ADF) as described by AOAC (1990).
Twenty gram of fresh maize or maize silage was mixed with 180 mL distilled water and blended for 30 s in a blender (GS253-450, Sunny, China).This was filtered through two No. 1 filter papers (Whatman, Maidstone, U.K.) to obtain the extract which was used to measure the organic acid and ammonia-N (NH 3 -N) concentrations in the silage extracts (Filya, 2003a).Immediately after extraction, the pH was measured, using a portable pH meter (HI8314, Hanna Instruments, ClujNapoca, Romania).Ammonia-N concentration was measured by distillation in a Kjeldahl auto analyzer (Filya, 2003a).Volatile fatty acid concentrations in the silage extract and in ruminal fluid were determined by using gas chromatography (WCOT Fused Silica Capillary, CHROMPACK CP. 9002, Model No. CP-9002, Serial No. 94 77 B, Vulcanusweg 259, AM DELFT, the Netherlands).Lactic acid was determined according to the technique of Barker & Summerson (1941) and WSC, using the phenol-sulphuric acid method of Dubois et al. (1956).
Milk samples were analyzed for lactose, fat and protein on a Milk-O-Scan (134 BN Foss Electric, Hillerod, Denmark).The milk composition was calculated as the average of the triple samples, using the portion of daily production at milking time as a weighting factor (Bal et al., 2000).Acid detergent insoluble ash (ADIA) concentration was measured in the composited diet and faecal samples, and used as an internal marker to calculate the apparent total-tract digestibility of the nutrients (Nikkhah et al., 2004).
Aerobic stability was defined as the number of hours that it took the silage to rise 2 °C above that of the ambient temperature, as described by Nishino et al. (2003).After opening the silos, about 1500 g of silage samples were loosely packed in open plastic containers, and a thermometer was placed in the geometric centre of the silage mass.The containers were covered with two layers of sterile cheese cloth to minimize drying, kept at room temperature (25 °C) and the temperature was measured at 2-h intervals.
Data, for chemical composition and fermentation characteristics of the silage, were analyzed in a completely randomized design (CRD), using the GLM procedure of SAS (2003).In vivo data were analyzed using the MIXED procedure of SAS (2003).The fixed effects of treatment and period, and the random effects of square and cow within a square were included in the model.For all variables the sampling was repeated over time (dry matter intake (DMI), milk yield and milk composition).The effect of time was included in the REPEATED statement of the model.The compound symmetric covariance structure was selected, based on the lowest Akaike Information Criterion (AIC).All differences were declared significant if P <0.05 and trends were discussed at P <0.10.

Results and Discussion
The chemical composition and fermentation characteristics of pre-ensiled whole-crop maize and experimental silages are shown in Table 1.Ensiling increased (P <0.05) the concentration of NDF and ADF, while reducing that of DM, WSC, CP and the pH, compared with the fresh maize crop.The concentration of lactic acid (P <0.01) and acetic acid (P <0.05) increased with inoculation compared to the control.The results are in agreement with Filya (2003a) and Kleinschmit & Kung (2006c), who reported a greater concentration of lactic and acetic acid in maize silages in response to L. buchneri or L. buchneri + L. plantarum.In general, a high content of lactic acid was expected in the silages due to the high concentration of WSC in maize crops.This is confirmed by the substantial lower levels of residual WSC in the silages compared with that found in the respective fresh maize forages.The concentration of butyric acid was increased (P <0.01) in response to inoculation with LM due to the heterofermentative activity of L. buchneri.Both silages, with and without inoculants, had good fermentation qualities, and the pH in both the control and Lactisil Maize treatment was below 4.Although a greater concentration of lactic acid was found in Lactisil Maize-silage, a higher pH (P <0.01) was recorded compared to the control.This may be partly related to a higher VFA concentration in these silages and a higher PKa value in VFA in comparison with lactic acid.Also, the results of previous studies have demonstrated greater concentrations of 1,2-propanediol (Taylor & Kung, 2002;Nishino et al., 2004;Kleinschmit & Kung, 2006a), ethanol and mannitol (Taylor & Kung, 2002;Kleinschmit & Kung, 2006a) in silages inoculated with L. buchneri, which may explain the higher pH of inoculated silages in the current study.In line with our study, silages with L. buchneri inoculants had a higher pH than the control or L. plantarum-inoculated silages in laboratory experiments conducted by Filya (2003b) and Kleinschmit & Kung (2006c), and field studies by Mari et al. (2009) and Kristensen et al. (2010).
The lower NH 3 -N concentration in the Lactisil Maize-treated silage (P <0.05) suggests that the inoculant reduced proteolysis, consistent with Filya (2003b) and Driehuis et al. (1999) who reported reduced NH 3 -N with L. buchneri + L. plantarum inoculation compared to untreated silages.Homofermentative bacteria such as L. plantarum usually accelerate the drop in pH at the beginning of ensiling (Weinberg & Muck, 1996;Driehuis et al., 1997).McDonald et al. (1991) and Zahiroddini et al. (2004) reported that as soon as the silage pH falls rapidly after ensiling, the aerobic microorganisms and plant enzymes are inhibited, which results in reduced proteolysis.
Lactic acid and WSC are nutrient substrates for fungi and are consequently responsible for the aerobic deterioration of silages (McDonald et al., 1991;Ohmomo et al., 2002).On the other hand, the antifungal effects of acetic acid on fungus are well documented (Kung & Ranjit, 2001;Nishino et al., 2003;Weinberg et al., 2011).In the current study, both lactic and acetic acid concentrations were increased (P <0.05) with Lactisil Maize inoculation compared to the control, but the ratios of lactic to acetic acid were similar between treatments.Moreover, low levels of acetic acid (<20 g/kg DM) were found in Lactisil Maizeenriched silage and as a result, aerobic stability was not affected (P >0.05) by the treatments.
The composition of the experimental diets, feed intake, ruminal fermentation, total tract nutrient digestibility and milk production data are shown in Tables 2 and 3. Cows fed inoculated silage had a lower (P <0.05) DMI than those fed uninoculated silage.The reduced DMI with Lactisil Maize may have resulted  from the extensive fermentation and greater silage end-products (McAllister et al., 1998;Buxton et al., 2003).Moreover, Nishino et al. (2007) reported increased concentrations of biogenic amines (tyramine and putrescine) in maize silage treated with L. buchneri, which can limit DMI (Lingaas & Tveit, 1992).The reduced milk yield (P <0.01) could be attributed to a decreased net energy intake due to the lower DMI in cows fed diets containing Lactisil Maize-inoculated silage.This was confirmed by a lower concentration of total VFA (P <0.05) in the rumens of these cows.Furthermore, the reduced yields of milk fat and protein in cows fed the Lactisil Maize-inoculated silage were related to the lower actual milk yield as the levels of milk fat and protein were not affected (P >0.05).Homofermentative LAB inoculants usually decrease milk fat and protein due to the high production of lactic acid and consequently decreasing the ratio of lipogenic VFAs (acetic acid and butyric acid) to glucogenic (propionate) VFAs (Huhtanen et al., 2003).The combination of homo-and heterofermentative LAB, in this study, produced milk with a similar (P >0.05) composition compared to the control.This can be partially explained by the similarity in molar proportions of the VFA in the rumen and the similar ratio of lipogenic VFA to glucogenic VFA.The lack (P >0.05) of significant effects in treatment on the chemical composition of maize silage, confirms the similarity in the total tract, apparent digestibility of nutrients and molar proportion of VFA in the rumen content between the two cow groups.The lack of effects in the Lactisil Maize-inoculated silage treatment, on total digestibility of nutrients, is in agreement with other reports (Kung et al., 1990;McAllister et al., 1998;Filya, 2003a).

Conclusion
The inoculation of high-moisture maize with Lactisil Maize had no effect on the chemical composition of the silage, but altered the fermentation characteristics.Inoculated silages had greater concentrations of lactic acid and acetic acid, and less ammonia N compared to the control.Cows fed diets containing Lactisil Maize-inoculated silage consumed less feed and consequently produced less milk, whereas nutrient digestibility and milk composition were not affected.These results indicated that inoculating with dualpurpose LAB inoculants in high-moisture maize silage improved only the fermentation quality of the silage, while cow performance, feed efficiency and milk composition were not affected.

Table 2
Feed ingredients and chemical composition of the experimental diets (n = 3)

Table 3
Dry matter intake, animal performance and apparent total tract nutrient digestibility (n = 3)