Ensiling natural meadow forage in « Tierra del Fuego » , Argentina

Livestock activity in the Argentinian «Tierra del Fuego» island is based on highly variable natural grasslands (NG). The conservation of fodder, despite the unfavorable environment, is restricted to hay; conservation as silage could be constrained by fodder quality and low temperatures. The objective was to assess the quality of NG silages fermented under contrasting storing conditions. Typical meadow forage was ensiled in minisilos and stored under shelter (Shelter) or in the field (Field) in a complete block design with repeated measures in time (t). Forage was chopped and inoculated with lactic acid bacteria plus enzymes, and harvested with 420 g dry matter (DM) kg fresh matter, 111 g crude protein kg DM, 665 g ash-free neutral detergent fibre kg DM, and 64 g water soluble carbohydrates kg DM. Shelter minisilos had higher metabolizable energy concentration at 236 d (PTrat×t = 0.03; Shelter = 10.2 ; Field = 9.6 MJ kg DM, P = 0.01), and lower DM losses (Shelter = –0.2; Campo = 22%, P = 0.02) and values of pH and N-NH3/total N (Field: 5.6 and 12% and Shelter: 4.4 and 6.8%, P ≤ 0.01). Fermentation acids concentration was similar for both treatments with preponderance of lactic acid, but acetic acid concentration increased over time (Pt ≤ 0.01). It was concluded that in «Tierra del Fuego», natural meadow forage quality is compatible with ensiling, but environmental conditions can limit the fermentation process. Additional key words: forage conservation, natural grasslands, Patagonia, silage.


Introduction
Livestock activity in Tierra del Fuego island (c.a.55°S and 65°W) is characterized, both in Argentine and Chilean sides, by extensive exploitation of sheep and beef cattle grazing natural steppes and meadows in an extreme agroecological environment with severe thermal, water and soil constrains (Jacob et al., 1999).There is a marked seasonality in forage production from late spring to summer, and a winter period with nil productivity, where much of the forage remains on the ground rotting under water, ice and snow.The geographical isolation limits the income of inputs, determining that livestock systems are highly dependent on the direct use of their own forage resources (natural or implanted).
Forage conservation is considered a priority tool to improve Argentine and Chilean livestock production systems in Patagonian region (Livraghi, 1996;Nilo Cavacevich, 2006).Widespread adoption of forage conservation technology would assist in avoiding massive waste of forage accumulated during the growing season, thereby reducing the recurrent forage supply imbalance.Concurrently, it would improve the nutritional status of animals and reduce the costs of winter food.
Although quantitatively limited, hay conserved as big round bales is the main form of preserving forage.However, the high soil moisture and environmental humidity, plus the typical strong winds undermine the efficiency of hay making.In contrast, silage minimizes the exposure of forage in the field, so that it could be more appropriate for the island conditions (Ballocchi, 1999).However, to our knowledge, there have not been local experiences in preserving forage as silage.Conservation as silage requires soluble carbohydrates, low buffer capacity, presence of sufficient amount of lactic acid bacteria (LAB) and anaerobic conditions (McDonald et al., 1991).Lactic acid bacteria requires between 5°C and 50°C to develop (Woolford, 1984), but in «Tierra del Fuego» daily temperatures during the harvesting season (January to April) fluctuate around 10°C, close to the minimum values required by LAB to grow.The aim of this work was to assess the quality of natural meadow forage silages fermented under contrasting storing conditions.

Experimental design
Six c.a. 100-kg minisilos were made with forage harvested from a typical meadow and were stored under shelter (Shelter) or directly in the field out-doors (Field) according to a complete block design (blocking by the time elapsed since cutting).

Description of meadow and ensiling technique
The forage was harvested from a fraction of a natural meadow located at the Misión Salesiana of Rio Grande («Tierra del Fuego», 53°42' 47.9" S, 67°49' 57.4" W and 7 m asl) at the 25 th January 2008 (south hemisphere summer).Aboveground biomass was estimated by cutting 10 random samples with scissors and using a rectangular frame of 0.2 m 2 (0.2 × 1 m).
Forage was cut and chopped directly with a grass harvester for silage (Mainero, Bell Ville Córdoba-Argentina) and inoculated manually (as indicated by the manufacturer: 4 L solution/ton of forage with 250 g inoculum/100 L water).The inoculant contained a mixture of enzymes (amylase and cellulase) and LAB (Lactobacillus plantarum, Streptococcus faecium, Pediococcus acidilactici; Sill-All, Alltech Biotechnology SRL, Pilar, Buenos Aires).The forage was carefully compacted into an appropriate section of silage storage bags (IPESA, Rio Chico, Rio Grande, Tierra del Fuego) and set in wooden minisilos c.a. 1 m 3 .Bags were subsequently sealed with tape, and minisilos were moved to their definite place of storage (Field or Shelter) and pressed with bags filled with gravel (c.a.70 kg m -2 ).Minisilos in Shelter treatment were kept in a big barn where machinery was sheltered, and no animal or source of heat was available.

Observations and sampling
Daily minimum and maximum temperatures from «Rio Grande» («Tierra del Fuego») were obtained from the National Meteorological Service.Temperature within the minisilos was monitored on days 1, 3, 111 and 236 (t 0 , t 1 , t 2 , t 3 corresponding to 25 th and 28 th January, 15 th May, and 17 th September 2008 respectively), using a digital thermometer (Maxthermo MD 3003 type K; Tainan, Taiwan) with a Inconel 600 probe of 0.6 m length, and reported as the temperature difference with respect to the environment temperature (recorded with the same instrument).
During harvesting, forage samples were taken, and then each minisilo was sampled at 111, 236 and 447 (t 4 , 16 th April 2009) days with a wool core sampler.All samples were preserved in plastic bags at -18°C until processed in the laboratory.
Volatile fatty acids (VFA, i.e. acetic, propionic, butyric and valeric acids) were measured on purified samples treated with ortophosphoric acid (2% v/v) at a rate of 0.2 mL per each 0.8 mL sample and later centrifuged for 10 min (×10,000 g).Determination of VFA was made by gas chromatography with a Konik-3000 equipment and a BP-20 column (Mark SGE) using N 2 as carrier and according to protocol recommended by the column manufacturer (adaptation, SUPELCO.«Analyzing fatty acid by packed column gas chromatography.G C. Bulletin 856 A»; Friggens et al., 1998).Lactic acid was measured calorimetrically according to Barnet (1951).
The metabolizable energy content (Mcal kg -1 DM) was estimated with the summative formula of Van Soest (Goering and Van Soest, 1970), and in vitro DM digestibility (ivDMD) was determined by the technique of Goering and Van Soest (1970).Simultaneously, in the same in vitro system, kinetics of organic matter digestion was described by the in vitro gas production technique proposed by Brooks and Theodorou (1997) and adapted by Wawrzkiewicz and Danelón (2004).

Statistical procedures
All variables, except for ivDMD, were analyzed according to a complete block design (blocking by the order in which minisilos were filled up) and using a model of repeated measures (t: date sampling) where treatment, block and minisilo were considered «fixed» factors (Proc mixed, SAS Institute, 2002).
Dry matter losses were estimated by mass difference between making (t 1 , 25 Incremental rate of gas production (GP) at 6 h, 12 h (cumulative GP at 12 h -GP at 6 h), 24 h (GP at 24 h -GP at 12 h), 48 h (GP at 48 h -GP at 24 h) and 72 h (GP at 72 h -GP at 48 h) was analyzed by ANOVA according to a complete block design.

Results
Aboveground forage biomass yielded 11,245 kg DM ha -1 (SD = 3,618 kg ha -1 ).The meadow was dominated by Alopecurus magallanicus and Hordeum pubiflorum, with a quantitatively lower presence of Carex sp., Poa pratensis and Deschampsia flexuosa.The dominant species at harvest (25 th January 2008) were in flowering state.
Minimum and maximum weekly mean temperatures are presented in Figure 1.Difference between minisilos and the outside temperature fell quickly from c.a. 2°C to 0°C during the first three days after sealed.Minisilos kept under Shelter had better color (green-yellowish) and odor (acidic) than those stored outdoors, which also had mycelium development.These differences were coincident with DM losses, as indoor minisilos had losses (no different from zero, P > 0.05), which were significantly lower than those obtained from units kept outdoors (-0.2 and 22 % respectively, P = 0.02).
Except for ADF OM , Cel, lactic acid and VFA (P > 0.05), most variables had differences between forage ensiled under Shelter or in the Field (Table 1).Original forage CP content was 111 g kg -1 DM (SEM = 4.2), and it increased slightly throughout the fermentation process (P = 0.03).
Contents of ADL SA and ash increased over the process of silage.Direct ash mean content rose from 82 g kg -1 DM (t 1 ) up to 94 and 86 g kg -1 DM at 236 (t 3 ) d of fermentation, respectively for Field and Shelter storing systems (P t = 0.0003).Similarly, direct ADL SA mean increased from 28 g kg -1 MS (t 1 ) to 36 and 31 g kg -1 DM respectively for Field and Shelter at t 3 .Furthermore, silage from Shelter treatment had lower aNDF OM content (P = 0.06) than Field ones, and lower Hemi content (P = 0.02), appearing dramatically reduced with respect to the Field treatment by the last sampling time (P Trat×t = 0.02; 337 and 301 g Hemi kg -1 DM at t 3 for Field and Shelter respectively).However, aNDF OM / ADL SA and Hemi/ADL SA ratios indicated that there only existed a time effect (P t ≤ 0.02; aNDF OM /ADL SA ratio: 23b, 25a and 20c g g -1 ; Hemi/ADL SA ratio: 11b, 12a and 10c g g -1 ; for t 1 , t 2 and t 3 respectively.Numbers followed by different letters differ, P < 0.05) without noticeable differences between treatments (P>0.05).
Silage samples from minisilos kept indoors yielded 9% (P > 0.05) more cumulative gas production than those stored in the Field (Fig. 2), but the greatest difference was brought about by the gas produced during the first 6 h of incubation (Field = 23, Shelter = 50 mL kg -1 DM, P = 0.06), as can be clearly appreciated by the hourly  rates of gas production (later GP rate differences between treatments were non-significant, P > 0.15).
In agreement with the chemical composition, fermentation products also showed effects associated with the form of storage, as pH and N-NH 3 (expressed as % of total N, TN) were higher for the silage obtained from minisilos kept in the Field (respectively, Field: 5.6 and 12% and Shelter: 4.4 and 6.8 %, P < 0.01; Table 1).Both variables presented a similar pattern (Fig. 3), indicating a loss of acidity beyond 111 d of fermentation in the Field treatment.For both variables, differences between treatments only became apparent as fermentation progressed.
Fermentation acids concentration were similar for both treatments (Table 1), and although the lactic/acetic ratio indicated a strong preponderance of lactic fermentation, a reduction of this ratio over time (15 to 5 g kg -1 DM declining in 230 d) was observed due to the increment in acetic acid concentration.

Discussion
Temperature regime was within the normal range observed in the region, and forage quality was similar to that reported for temperate grasses in terms of CP, aNDF OM and ADF OM , but the content of ADL SA was substantially lower (26 g kg -1 DM in this work) (Jaurena and Danelón, 2006).The CP contents in the fresh material and silage (c.a.110 g kg -1 DM) presented a slight increase without biological importance between harvest and t 3 sampling (236 d), and was comparable to previous reports for meadows with dominance of Carex sp.(Cabeza and Livraghi, 2004).
The WSC contents in the original forage (64 g kg -1 DM) was within the minimum (0-80 g kg -1 DM) suggested for grasses for a proper silage fermentation according to Woolford (1984), but below the threshold of 7 g kg -1 fresh matter suggested by Haigh (1990), hence it can be argued that the forage sugar content could have reduced the ensiling fermentation.Rapid reduction of internal temperatures during the first three days of storage indicated a rapid limitation of respiratory activity brought about by the proper sealing of the minisilos.In-silo DM losses (-0.2% and 22% respectively, for Shelter and Field treatments) were in agreement with other authors reports (25%, Bastiman and Altman, 1985;18%, Jaurena and Pichard, 1999;3-25%, McDonald et al., 1991;22%, Moore and Kennedy, 1994;0-43%, Ruppel et al., 1995;16%, Watson and Nash, 1960), and clearly indicated the technical convenience of the Shelter treatment.During storage, disappearance of non-structural organic matter components induced the increments in ADL SA and ash concentrations, which is in concordance with the above discussed DM losses.Furthermore, aNDF OM /ADL SA and Hemi/ADL SA ratios signaled a similar pattern between treatments, i.e. structural carbohydrates were not affected by treatments, but by t 3 some remotion of Hemi was detected for both.Consequently, it is apparent that aFDN OM and Hemi reduction by t 3 (on a DM bassis) in Shelter treatment was probably brought about by the differential disappearance of non-structural components between treatments.
Ensiling forage under shelter led to better fermentation of material as indicated by pH (4.03 at t 2 ) and N-NH 4 concentration, on the contrary forage samples taken from outdoors minisilos did not reach stability, as indicated by the higher pH and N-NH 4 concentration and beyond 111 d they suffered a drastic increase in acetic acid concentration, pH and proteolysis.The rise in acetic acid concentration could have been associated with the activity of yeasts or acetic bacteria that tend to start the process of deterioration in many silages when exposed to air (Driehuis et al., 1999).However, the production of acetic acid has also been reported from the fermentation of lactic acid by L. buchneri at pH as low as 3.8, or L. bifermentans at pH above 4 (Driehuis et al., 1999), and by propionic bacteria that can proliferate in silages with pH above 4.8 (Weinberg and Muck, 1996).Indeed, future research about ensiling natural forages in this environment will require of a complete description of microbial populations.
Overall, ensiling under shelter contributed to achieve a better stabilization of the silage without apparent ivDMD superiority, in agreement with previous results showing that fermentative profile is not necessarily associated with digestibility (Haigh, 1995).This difference was probably associated with the preservation of fermentable carbohydrates in forage ensiled under shelter, as suggested by the gas production yield during the early hours of in vitro incubation.
As final conclusions, the conservation of natural forage as silage is a promising alternative for «Tierra del Fuego» livestock production systems.This work intended to assess the feasibility of ensiling forage from natural meadows under two systems with contrasting degrees of protection from the environment, considering that the temperature could constrain microbial development during the ensiling process.
The original quality of the forage despite the advanced degree of maturity at harvest was acceptable for ensiling and comparable to that obtained in others regions of the country.With respect to the ensiling process, these preliminary and exploratory results are consistent with the hypothesis tested, as forage ensiled under shelter developed a better fermentation, nutritive quality descriptors (ME concentration) and reached stability.On the contrary, forage ensiled in minisilos exposed to the outdoors environmental conditions did not reach proper stability.
These results support the idea that ensiling could play a key role in enhancing livestock production systems, and that local research on this issue is imperative.

Figure 1 .
Figure 1.Weekly minimum and maximum mean temperatures for Rio Grande meteorological station (Argentina) during 2008.

Figure 2 .Figure 3 .
Figure 2. (a) Cumulative and (b) hourly gas production rate of meadow silages stored in field (full line) or under shelter (dotted line).

Table 1 .
Chemical composition of natural meadow forage stored in the Field or under Shelter.Adjusted means, expressed in g kg -1 DM, unless stated otherwise