- Email: [email protected]
Fermentation Characteristics of Corn Forage Ensiled in Mini-Silos
J. Dairy Sci. 87:4238–4246 © American Dairy Science Association, 2004.
Fermentation Characteristics of Corn Forage Ensiled in Mini-Silos D. J. R. Cherney,1 J. H. Cherney,2 and W. J. Cox2 1
Department of Animal Science, and Department of Crop and Soil Sciences, Cornell University, Ithaca, NY 14853
2
ABSTRACT
INTRODUCTION
To evaluate numerous experimental variables and their interactions involving different corn (Zea mays, L.) silage hybrids, scaled down mini-silos are necessary. Objectives of this study were to evaluate the influence of sample size on pH, NH3, and volatile fatty acid profile of 8 corn silage hybrids, selected to vary in fiber digestibility and ensiled in vacuum-sealed polyethylene bags for 90 d, and to assess the suitability of these mini-silos for detecting differences among corn silage samples. Hybrids were grown at the Cornell Teaching and Research Center located near Harford, NY, and harvested at a dry matter content of about 32% in the fall of 2002. Samples from 3 field replications of each hybrid were chipper-shredder chopped and vacuum-ensiled in bags with sample sizes of 50, 100, 200, 400, and 600 g. Increasing sample size resulted in decreased lactic acid, acetic acid, total acids, and NH3. Most of the difference among sample sizes occurred between the 50- and 100g sample sizes. Lactic acid:acetic acid ratio (3.1 ± 0.13) and pH (3.9 ± 0.08) did not vary among sample sizes. There was no detectable butyric acid in the samples. Fermentation characteristics suggested that all samples were well ensiled but that the fermentation profile of the 50-g samples differed the most from other sample sizes. Hybrids did vary in lactic acid, acetic acid, lactic acid:acetic acid, and pH. Differences among hybrids were also noted for dry matter and crude protein. Fieldchopped corn hybrids that were ensiled using mini-silos had higher acids than corresponding field-chopped corn hybrids ensiled in Ag-bags, in part due to no effluent escaping from the mini-silos. It is possible to use vacuum-sealed plastic bags to ensile corn, with samples as small as 200 g, and to use these mini-silos to assess differences among corn silage samples. Caution should be used when extrapolating mini-silo data to fieldscale ensiling. (Key words: laboratory silo, corn silage, fermentation)
A variety of traits and corn (Zea mays L.) silage processing techniques are being promoted for high milk production through improved forage quality. To evaluate numerous experimental variables and their interactions involving different corn silage hybrids, scaled down mini-silos are necessary. The entire contents of laboratory silos can be weighed, processed, and analyzed accurately. This can only be done under the assumption that the fermentation process is reasonably similar to that taking place in field-scale silos (Cherney and Cherney, 2003). Quality parameters that fluctuate during ensiling include temperature, pH, VFA, and aerobic stability (Anderson et al., 1989; Pitt, 1990; O’Kiely, 1993; May et al., 2001). Low-fiber silage with a pH < 4.2 is considered to be properly ensiled (Meiske et al., 1975; Cherney and Cherney, 2003). A survey of the limited number of studies directly comparing fermentation in field-scale and small-scale silos resulted in the conclusion that forage in both silo types did undergo similar fermentation (Meiske et al., 1975). Experimental silos enable experimental variables to be scaled down from field scale to experimental units that allow for multiple treatments and replications. Although the general ensiling process was similar for test tube, plastic pipe, and 4- to 8-tonne polyethyleneenclosed (clamp) silos, O’Kiely and Wilson (1991) concluded that there was potential for significant fermentation × silo type interactions when treatments that directly affected fermentation were being compared. Laboratory silos, however, are still considered a practical method of comparing a number of treatments (O’Kiely, 1993). Forage corn sealed in polyethylene bags has been used to demonstrate the ensiling process to undergraduate animal nutrition students (A.W. Bell, 2004, personal communication), but only pH and DM were assessed. The primary objectives of this study were to evaluate the influence of sample size on pH, NH3, and VFA profile of 8 corn silage hybrids, selected to vary in fiber digestibility and ensiled in vacuumsealed polyethylene bags for 90 d, and to assess the suitability of these mini-silos for detecting treatment differences among corn silage samples processed using a chipper-shredder. A secondary objective was to relate
Abbreviation key: HI = harvest index.
Received April 6, 2004. Accepted August 11, 2004. Corresponding author: D. J. R. Cherney; e-mail: [email protected].
4238
CORN FORAGE ENSILED IN MINI-SILOS
the fermentation characteristics of the chipper-shredder mini-silos to fermentation profiles obtained from fresh corn that is field-processed and ensiled in commercial field scale plastic bags or mini-silos. MATERIALS AND METHODS Silage corn hybrids Agway 5215, Agway 5496, Dekalb DKC 53-32, Dekalb RX 489, Mycogen BMR F407, Mycogen TMF 100, Pioneer 35P12, and Pioneer 36-B-08 were chosen to represent a range of fiber digestibility. Hybrids were planted on May 11, 2002, at a planting rate of about 32,400 plants/ha at the Teaching and Research Center located near Harford, NY, at a spacing of 76 cm in 12-row strips, replicated 3 times. The previous crop was alfalfa. About 224 kg/ha of 10-20-20 NPK fertilizer was applied through the corn planter, and about 84 kg/ ha of liquid N was applied at the 4-leaf stage of corn growth. Hybrids were harvested at a DM of about 32% in fall of 2002. For the primary objective, fresh corn samples were processed through a chipper-shredder (Mighty Mac LSC506, MacKissic, Inc., Parker Ford, PA). In the field, 3 plants were passed through a chipper-shredder each time a new hybrid was shredded to remove residual plant material that might still have been in the shredder. Six plants were then chipper-shredded and bagged separately in a tightly closed plastic bag to be used for ensiling in vacuum sealer bags. Shredded material from each hybrid × field replication was thoroughly mixed in a shallow plastic pan, resulting in a homogeneous mixture. Samples of 50, 100, 200, 300, 400, 500, and 600 g were vacuum-sealed into polyethylene bags (Nylon poly EVOH high barrier 3 mil 22 × 33 cm vacuum pouches), using a single chamber vacuum packaging machine (model KVP-420T with 512 mmHg vacuum level; Process Plus Food Processing Equipment Inc., Parker’s Prairie, MN) and stored for 90 d. Mini-silos were stored in the laboratory (21°C) in black plastic bags during the 90 d. Two subsamples from each hybrid × replicate were collected for determination of buffering capacity. To determine buffering capacity, 3 g of dry forage (60°C) was placed in 60 mL of distilled water and allowed to stand for 4 h. The pH was then measured and the samples titrated to pH 4.0 using 0.1 M lactic acid (Fisher and Burns, 1987). Four subsamples from each hybrid × replicate were collected for determination of sugars. Sugar was determined using a water-soluble sugar method (Hall et al., 1999). Another set of 6 plants was harvested at the same time to determine harvest index (HI), defined as the ratio of grain DM to total DM (Cox et al., 1993).
4239
The mini-silos were opened at the end of 90 d to determine pH and VFA content. Fifty grams of wet sample was homogenized with 500 mL of deionized water in a Waring blender (2 min) and filtered through 2 layers of cheesecloth before measuring pH; 8.35 mL of 1 M m-phosphoric acid was then added to 25 mL of the extract, which was frozen and later analyzed for VFA (procedure adapted from Wiseman and Irvin, 1957). Fermentation analysis of corn silage samples was performed by Dairy One (DHI Forage Testing Laboratory, Ithaca, NY). Samples were filtered through a disposable syringe filter [Millipore Millex 5.0 μm PVDF (Durapore) membrane, Millipore Corp., Billerica, MA] and analyzed for acetic, propionic, butyric, and isobutyric acids using gas chromatography (Supelco, 1990; procedures cited from Lombard and Dowell, Jr., 1962). Lactic acid was determined using an YSI 2700 SELECT Biochemistry Analyzer equipped with an L-lactate membrane. Crude protein (AOAC 976.06, 1990) and ammonia N (AOAC 941.04, 1990) were also determined. After samples for the above study were collected, corn forage was harvested with a commercial chopper (Claas-chopper, Jaguar 850; CLAAS of America, LLC, Columbus, IN; 21 mm setting). Field size for each hybrid × replication averaged 0.71 ± 0.017 ha. Corn forage was stored in separate plastic bags (Ag-bag International Ltd., Warrenton, OR) for each hybrid. Samples were treated with a lactic acid bacterial inoculant (Pioneer inoculant #1132) applied at 0.9 g/tonne through the forage harvester. Locations of each of the 3 field replications within each bag were noted. After 90 d, 4 samples were taken from random locations for each field replication in the Ag-bag using a Penn State hay core sampler. Samples were mixed, as described above, then immediately processed and analyzed as described above. Another set of mini-silo samples was prepared by sampling hybrid × replication described above as the Ag-bags were being filled. Mini-silos were then prepared as described above. Although we were primarily interested in the fermentation characteristics of the corn samples, Ag-bag samples were analyzed for DM, CP, NDF, and in vitro true digestibility to establish that there was a range in the chemical composition of the hybrids used. Samples were first dried at 60°C (NFTA, 1993), and then ground through a Wiley mill fitted with a 4-mm screen, followed by grinding with an Udy mill fitted with a 1-mm screen. Dry matter was determined by drying at 105°C for 24 h (NFTA, 1993). Nitrogen was determined by combustion (Leco Instruments, Inc., St. Joseph, MI) (AOAC 976.06, 1990) and multiplied by 6.25 to obtain CP. Neutral detergent fiber was analyzed according to Van Soest et al. (1991) using the ANKOM system for NDF. SulfiteJournal of Dairy Science Vol. 87, No. 12, 2004
4240
CHERNEY ET AL. Table 1. Dry matter, chemical, and fiber characteristics of corn silage hybrids ensiled in Ag-bags for 3 mo. Hybrid2 Variable1
AG5215
AG5496
BMR407
DK53-32
P35P12
P36B08
RX489
TMF100
SED3
DM, % NDF, % of DM IVTD, % of DM dNDF, % of DM CP, % of DM HI,4 % BC,4 mL of acid Sugar,4 % of DM
31.8abc 41.3bc 80.8a 53.6a 8.5ab 45.7b 8.3a 8.0a
30.3a 46.2c 81.1a 59.0b 9.1b 34.3a 10.0b 13.5bc
35.2c 34.9a 90.0c 71.2c 8.6ab 50.8c 8.2a 8.6a
34.4c 45.4c 82.9ab 62.3b 8.5ab 45.9b 9.4ab 9.6a
34.0abc 44.1bc 82.8ab 61.0b 8.9ab 44.4b 8.1a 10.18ab
35.6bc 42.4bc 82.1ab 57.8ab 8.9ab 43.3b 8.1a 12.42b
31.0ab 42.8bc 82.1ab 58.3ab 8.4a 35.8a 11.1b 15.4c
34.8c 39.0ab 84.4b 59.9b 9.1b 43.3b 9.6ab 8.89a
1.16 1.51 0.73 1.71 0.21 0.98 0.53 0.80
Means within a row with different superscripts differ (P ≤ 0.05). IVTD = In vitro true digestibility; dNDF = NDF digestibility; HI = harvest index; BC = buffering capacity. 2 AG5215 = Agway 5215; AG5496 = Agway 5496; BMR407 = Mycogen BMR F407; DK53-32 = Dekalb DKC 53-32; P35P12 = Pioneer 35-P-12; P36B08 = Pioneer 36-B-08; RX489 = Dekalb RX 489; TMF100= Mycogen TMF 100. 3 SED = Standard error of the differences of the least square means. 4 HI = Harvest index; BC = buffering capacity; determined from samples that were collected and dried at time of ensiling. a,b,c 1
and heat-stable alpha-amylase were used for NDF analysis of all samples. Samples were not corrected to an ash-free basis. In vitro fiber digestibility was determined according to Cherney et al. (1997), using the rumen buffer described by Marten and Barnes (1980) and using the Daisy II 200/220 in vitro incubator (ANKOM Technology, Fairport, NY) and the ANKOM 200/ 220 fiber analyzer (ANKOM Technology, Fairport, NY). The buffer contained urea. Ruminal fluid inoculum was obtained from a nonlactating, rumen-fistulated Holstein cow offered a medium quality orchardgrass (Dactylis glomerata L.) hay diet for ad libitum intake. Samples (0.25 g) were incubated for 48 h at 39°C, and undigested residues were treated with neutral detergent solution. The NDF digestibility was determined using the following formula: NDF digestibility = 100 − (NDF remaining at t = 48 h/NDF at t = 0 h) × 100. Particle sizes of silages from chipper-shredder and Claas-chopped mini-silos were analyzed using the Penn State Forage Particle separator (Kononoff et al., 2003). Particles were separated on 3 screens and a bottom pan, with screen hole sizes of 19, 8, and 1.18 mm. Material used for the 3 methods represented different sample populations. Therefore, they cannot be compared directly and were analyzed separately. For the chipper-shredder mini-silos, a split-plot in a randomized complete block design was used with 3 field replications. Hybrids were the main plots. Subplots consisted of the sample size. A split-plot ANOVA with repeated measures was used to test for statistical significance of treatment effects and interactions using PROC MIXED (Littell et al., 1996) in SAS, version 7.0 (SAS Inst., 1998). The model assumed that hybrid and sample size were fixed variables, and replication was considered a Journal of Dairy Science Vol. 87, No. 12, 2004
random variable. Sample DM was used as a covariate in the analysis. Separation of hybrid and sample size treatment means was accomplished using the TukeyKramer procedure for multiple comparisons (P ≤ 0.05). For the Claas-chopped mini-silos and Ag-bag material, a randomized design was used with 3 field replications. An ANOVA with repeated measures was used to test for statistical significance of treatment effects and interactions using PROC MIXED (Littell et al., 1996) in SAS, version 7.0 (SAS Inst., 1998). The model assumed that hybrid was a fixed variable, and replication was considered a random variable. Sample DM was used as a covariate. Separation of hybrid treatment means was accomplished using the Tukey-Kramer procedure for multiple comparisons (P ≤ 0.05). RESULTS AND DISCUSSION Ensiled corn hybrids exhibited a range in fiber, CP, digestibility, and other measures of forage quality (Table 1). Dry matter varied among hybrids, with a range of almost 5%, so DM was included as a covariate in the analysis of fermentation profiles. Hybrids were initially selected to vary in in vitro true digestibility and NDF digestibility. Based on the results observed here, we were successful in obtaining corn samples that varied in chemical composition. We needed this variation to test the ability of the mini-silo system to handle a range of corn silage samples. Size × hybrid interactions for the fermentation constituents measured were not significant (P > 0.05). Therefore, main effects will be discussed. Most of the differences among sample sizes occurred between the 50- and 100-g sample sizes (Figure 1). Increasing sam-
CORN FORAGE ENSILED IN MINI-SILOS
Figure 1. Influence of sample size on lactic acid, acetic acid, ammonia N, and pH. Means with different superscripts differ (P ≤ 0.05); comparisons do not include 50-g samples. pH did not differ among sample sizes (P > 0.05).
ple size from 50 to 100 g or more resulted in decreased lactic acid, acetic acid, total acids, and ammonia. There was very little or no propionic acid, butyric acid, or isobutyric acid in these samples, and sample sizes were not different (P > 0.05) in these constituents (data not shown). The reason for the large difference between the 50-g samples and the other sample sizes is not readily apparent. It is unlikely that sampling of the original fresh corn sample was the reason for the difference, as CV for the various sample sizes were similar (P > 0.05). Because of the large change in fermentation characteristics between 50- and 100-g samples, data was reanalyzed excluding the 50-g samples. There were still small but significant changes in acetic acid, lactic acid, and NH3 N between 100- and 200-g samples. Changes appeared to level off above 200 g for these constituents (Figure 1), as well as for total acids (6.68, 5.62, 5.33, and 5.44 % of DM, respectively, for 100-, 200-, 400-, and 600-g sample size, ± 0.31% SEM). There was a steady decrease (P < 0.05) in ammonia with increasing sample size (0.56, 0.52, 0.48, and 0.45% of DM, respectively, for 100-, 200-, 400-, and 600-g sample size, ± 0.38% SEM). The DM varied among sample sizes, but there was no apparent pattern (DM = 32.7, 32.4, 32.0, and 32.8% of DM, respectively, for 100-, 200-, 400-, and 600-g sample size, ± 0.78% SEM). The lactic:acetic acid ratio and CP did not vary among sample size. Variations in pH within a hybrid across sample size were generally less than 0.1 units and did not seem to follow a pattern. There were noticeable differences among hybrids. Kung and Shaver (2001) report that the typical range of pH for corn silage is 3.7 to 4.2. Silages were generally within that range regardless of
4241
hybrid or size of sample (Figure 2). One exception to this was the Dekalb RX 489 hybrid, which was always above pH 4.2, regardless of sample size. The Agway 5496 hybrid at sample sizes 400 and 600 g was also above pH 4.2 and tended to be the most variable of the hybrids in terms of pH across sample size. The Mycogen BMR F407 silage tended to be the lowest in pH, averaging about 3.7. There were no differences (P > 0.05) among hybrids in NH3, propionic acid, butryric acid, or isobutyric acid. Hybrids, however, did vary in lactic acid, acetic acid, lactic acid:acetic acid, total acids, and pH (Table 2). Differences among hybrids were noted for CP and DM. The Mycogen BMR F407 hybrid was highest in lactic acid, which would have contributed the most to lowering the pH. The Mycogen BMR F407 hybrid had the highest HI, which should result in the most substrate for the lactic acid bacteria (Table 1). Hybrids Agway 5496 and Dekalb RX 489, conversely, had the lowest HI. These 2 hybrids had the highest buffering capacities (Table 1) and the highest pH (Figure 2). Buffering capacity is generally associated with silage pH (McDonald et al., 1991); the association between buffering capacity and pH was high in this study (Figure 3; r = 0.69; P < 0.01). Sugar content of the hybrids was also associated with pH (Table 1; r = 0.63; P < 0.01). The positive association of sugars to pH is likely indicative of the low HI in hybrids with high sugars (Table 1). Russell (1986) noted that the nutritional value of corn stover was inversely related to grain yield and was due to a translocation of sugars from the stem to grain. The grain is a strong carbohydrate sink (Allen et al., 2003). Most of the total nonstructural carbohydrate in the grain is stored as starch, whereas in the stems it is stored as simple sugars. Typical concentrations on a DM basis range from 4 to 7% for lactic acid, 1 to 3% for acetic acid, <0.1% for propionic acid, 0% for butyric acid, and 5 to 7% for NH3 N (% of CP) (Kung and Shaver, 2001). Kung and Shaver (2001) also indicated that lactic acid should be 65 to 70% of total acids. Using these values as indicative of adequate ensiling, all sample sizes greater than 50 g fell within the range considered typical (Figure 1). The 50-g samples were higher in lactic acid than the typical range, which would be acceptable, but they were also higher in acetic acid and NH3 N concentrations. Some studies suggest that acetic acid concentrations above 5 to 6 % of DM will depress animal intake (BuchananSmith, 1990; Rook and Gill, 1990). Concentrations of NH3 N above 12 to 15% indicate excessive protein breakdown in the silage (Kung and Shaver, 2001). From our data, we conclude that the 50-g mini-silos are not representative of typical silage fermentation. Journal of Dairy Science Vol. 87, No. 12, 2004
4242
CHERNEY ET AL.
Figure 2. Influence of hybrid and sample size on pH. Vertical bars indicate standard error of the mean.
Only one hybrid, Pioneer 36-B-08, fell outside the typical range for lactic acid when averaging 100- to 600g sample sizes (Table 2). Other variables were within the range considered typical. Lactic acid still represented 70% of total acids and pH was low. This would suggest that this hybrid was well ensiled. The data
for the hybrids suggests that the vacuum bag method produces silage that falls within the range considered typical. Thus, this method is sensitive enough to detect differences among hybrids in a laboratory setting where treatment differences are carefully controlled. Because the chipper-shredder method uses just 6 plants out of
Table 2. Fermentation analysis of chipper-chopped silage, no inoculant, after 90 d of storage.1 Effect (P ≤)3
Hybrid2 Variable Lactic acid, % of DM Acetic acid, % of DM Lactic:acetic, % of DM Total acids, % of DM pH CP, % of DM Ammonia, % of DM Ammonia N, % of CP DM, %
AG5215 ab
4.1 (0.354) 1.6b (0.06) 2.6a (0.25) 5.7a (0.36) 3.80bc (0.02) 8.0a (0.17) 0.50a (0.048) 6.31bc (0.501) 31.5ab (1.147)
AG5496 a
3.7 (0.35) 1.2a (0.06) 3.2ab (0.25) 5.0a (0.36) 4.14d (0.02) 9.3d (0.17) 0.48a (0.047) 5.21a (0.500) 31.5ab (1.14)
BMR407 c
5.4 (0.35) 1.2a (0.06) 4.4c (0.25) 6.6b (0.37) 3.72a (0.02) 7.9a (0.18) 0.51a (0.049) 6.45bc (0.513) 31.0ab (1.14)
DK53-32 abc
4.3 (0.35) 1.6b (0.06) 2.7a (0.25) 5.9ab (0.36) 3.85c (0.02) 8.5b (0.17) 0.58a (0.048) 6.72c (0.501) 33.3ab (1.14)
P35P12 bc
4.8 (0.35) 1.6b (0.06) 2.9a (0.25) 6.4b (0.36) 3.79ab (0.02) 8.8bc (0.17) 0.49a (0.048) 5.54ab (0.501) 33.3ab (1.14)
P36B08 a
3.9 (0.36) 1.5b (0.06) 2.5a (0.25) 5.4a (0.37) 3.82bc (0.02) 8.7bc (0.18) 0.48a (0.049) 5.45ab (0.513) 34.0b (1.14)
RX489 abc
4.3 (0.36) 1.2a (0.06) 3.7b (0.25) 5.4a (0.37) 4.28e (0.03) 8.6bc (0.18) 0.47a (0.049) 5.43ab (0.518) 30.9a (1.14)
TMF100 abc
4.2 (0.35) 1.6b (0.06) 2.6a (0.25) 5.8ab (0.38) 3.86c (0.02) 8.9cd (0.18) 0.49a (0.050) 5.44ab (0.525) 34.2b (1.14)
DM
H
S
HXS
0.01
0.04
0.03
0.99
0.03
0.01
0.01
0.97
0.01
0.01
0.50
0.99
0.01
0.01
0.01
0.99
0.07
0.01
0.05
0.53
0.01
0.01
0.47
0.79
0.26
0.60
0.03
0.63
0.68
0.11
0.03
0.65
...
0.09
0.01
0.21
Means within a row with different superscripts differ (P ≤ 0.05). Average of 200- to 600-g mini-silos, 3 replications. 2 AG5215 = Agway 5215; AG5496 = Agway 5496; BMR407 = Mycogen BMR F407; DK53-32 = Dekalb DKC 53-32; P35P12 = Pioneer 35P-12; P36B08 = Pioneer 36-B-08; RX489 = Dekalb RX 489; TMF100= Mycogen TMF 100. 3 Probability that effect is significant; DM = DM covariate; H = hybrid; S = sample size; HXS = sample × size interaction. 4 Standard error of the mean. a,b,c,d,e 1
Journal of Dairy Science Vol. 87, No. 12, 2004
4243
CORN FORAGE ENSILED IN MINI-SILOS
inoculant was added to the Claas-chopped material. Adding lactic acid bacteria to the Claas-chopped material would result in differences between this material and the chipper-shredded material. Particle size of chipper-shredder samples was smaller than corresponding Claas-chopped material (Figure 4). Chipper-shredder samples had less than 10% remaining above the 19-mm screen, whereas the Claaschopped samples averaged above 40%. On average, chipper-shredder samples retained more on screens smaller than 19 mm than did Claas-chopped samples (Figure 4). The finer chop of the chipper-shredder material may have altered the fermentation process. Chopping tends to increase microbial numbers (Bolsen, 1995), and the finer chop would have liberated more nutrients. Differences between the Claas-chopped forage ensiled in mini-silos (Table 3) and the Claas-chopped forage material in Ag-bags (Table 4) were generally less than those observed between chipper-shredder material (Table 2) and Claas-chopped material. Samples from Claas-chopped mini-silos tended to be higher in acids than Ag-bag samples. Effluent cannot escape from the vacuum-packed mini-silos, which would account for higher acids. Virtanen (1947) indicated that this effluent could lead to more extensive fermentation. O’Kiely and Wilson (1991) reported that lactic acid was higher in test-tube silages (no effluent escape) than in clamp silos. Hybrid ranking of the 3 methods for acids was
Figure 3. Relationship between pH and buffering capacity.
a field, there is likely to be some variation between the 6 sampled plants and the whole field, particularly when there is variation among the plants in the field. For that reason, caution must be exercised when extrapolating laboratory results to field studies. Although not directly comparable to the chippershredder samples, Claas-chopped inoculated samples appear to have more lactic acid, which would be expected (Tables 3 and 4), because a lactic acid bacterial
Table 3. Fermentation analysis of Claas-chopped silage, with inoculant, 600-g mini-silos after 90 d of storage.1 Effect (P ≤)3
Hybrid2 Variable Lactic acid, % of DM Acetic acid, % of DM Lactic:acetic, % of DM Total acids, % of DM pH CP, % of DM Ammonia, % of DM Ammonia N, % of CP DM, %
AG5215 a
9.0 (0.364) 1.8ab (0.09) 5.1a (0.35) 10.8ab (0.32) 3.80a (0.030) 8.2a (0.12) 0.64bc (0.022) 7.8b (0.28) 30.8b (0.64)
AG5496 a
7.5 (0.44) 1.5a (0.10) 5.2a (0.43) 8.9ab (0.39) 3.80a (0.037) 9.0b (0.15) 0.53a (0.028) 6.0a (0.34) 29.2b (0.64)
BMR407
DK53-32
a
a
8.5 (0.38) 1.8ab (0.09) 4.6a (0.37) 10.4ab (0.33) 3.73a (0.032) 8.3ab (0.13) 0.51a (0.024) 6.2a (0.29) 32.0bc (0.64)
8.9 (0.37) 1.8ab (0.08) 5.1a (0.36) 10.6ab (0.32) 3.80a (0.031) 8.5ab (0.13) 0.59ab (0.023) 6.9ab (0.29) 31.7abc (0.64)
P35P12 a
7.5 (0.36) 2.0b (0.08) 3.9a (0.35) 9.6a (0.32) 3.80a (0.030) 8.5ab (0.12) 0.66c (0.023) 7.8b (0.28) 30.9b (0.64)
P36B08 a
8.3 (0.36) 1.7ab (0.08) 5.0a (0.35) 10.0ab (0.32) 3.77a (0.030) 8.7ab (0.12) 0.55ab (0.023) 6.3a (0.28) 31.4abc (0.64)
RX489 a
8.2 (0.51) 1.9b (0.11) 4.4a (0.49) 10.2ab (0.44) 3.70a (0.042) 9.1b (0.17) 0.62abc (0.032) 6.9ab (0.38) 28.6ab (0.64)
TMF100 a
9.9 (0.56) 1.7ab (0.13) 5.7a (0.54) 11.7b (0.49) 3.87a (0.046) 8.9ab (0.19) 0.62abc (0.035) 7.0ab (0.43) 34.1c (0.64)
H
DM
0.05
0.07
0.02
0.92
0.15
0.31
0.02
0.05
0.13
0.09
0.01
0.07
0.01
0.49
0.01
0.19
0.01
...
Means within a row with different superscripts differ (P ≤ 0.05). Average of mini-silos, 3 replications. 2 AG5215 = Agway 5215; AG5496 = Agway 5496; BMR407 = Mycogen BMR F407; DK53-32 = Dekalb DKC 53-32; P35P12 = Pioneer 35P-12; P36B08 = Pioneer 36-B-08; RX489 = Dekalb RX 489; TMF100= Mycogen TMF 100. 3 Probability that effect is significant; H = hybrid; DM = DM covariate. 4 Standard error of the mean. a,b,c 1
Journal of Dairy Science Vol. 87, No. 12, 2004
4244
CHERNEY ET AL.
Table 4. Fermentation analysis of Claas-chopped silage, with inoculant, after 5 mo of storage in an Ag-bag.1 Effect (P ≤)3
Hybrid2 Variable
AG5215
AG5496
BMR407
DK53-32
P35P12
P36B08
RX489
TMF100
DM
Lactic acid, % of DM
6.8a (0.234) 1.1a (0.11) 6.1a (0.52) 7.9a (0.29) 3.82a (0.023) 8.1a (0.16) 0.44a (0.024) 5.4a (0.30) 29.9ab (0.59)
6.8a (0.32) 1.0a (0.15) 7.2a (0.72) 7.8a (0.41) 3.88a (0.030) 8.6ab (0.22) 0.52ab (0.033) 6.1a (0.42) 28.4a (0.59)
6.8a (0.24) 1.4a (0.11) 4.7a (0.53) 8.2a (0.30) 3.77a (0.023) 8.5ab (0.16) 0.48a (0.024) 5.6a (0.31) 33.2c (0.59)
7.9b (0.24) 1.3a (0.11) 6.3a (0.53) 9.2a (0.30) 3.77a (0.023) 8.3a (0.16) 0.51ab (0.024) 6.2a (0.31) 33.2c (0.59)
7.5ab (0.23) 1.3a (0.11) 6.1a (0.51) 8.8a (0.29) 3.78a (0.023) 8.6ab (0.16) 0.58b (0.023) 6.7a (0.30) 33.0c (0.59)
7.4ab (0.22) 1.3a (0.10) 5.4a (0.49) 8.7a (0.28) 3.78a (0.022) 8.3a (0.15) 0.52ab (0.023) 6.2a (0.29) 32.8c (0.59)
7.0ab (0.29) 0.9a (0.14) 7.9a (0.65) 7.9a (0.36) 3.91b (0.028) 8.2a (0.20) 0.54ab (0.030) 6.6a (0.38) 28.9a (0.59)
7.6ab (0.22) 1.2a (0.10) 6.6a (0.49) 8.8a (0.27) 3.91b (0.022) 8.9b (0.15) 0.61b (0.022) 6.9a (0.29) 32.7a (0.59)
0.07
Acetic acid, % of DM Lactic:acetic, % of DM Total acids, % of DM pH CP, % of DM Ammonia, % of DM Ammonia N, % of CP DM, %
0.01 0.03 0.02 0.06 0.29 0.03 0.11 ...
Means within a row with different superscripts differ (P ≤ 0.05). Average of 3 replications. 2 AG5215 = Agway 5215; AG5496 = Agway 5496; BMR407 = Mycogen BMR F407; DK53-32 = Dekalb DKC 53-32; P35P12 = Pioneer 35P-12; P36B08 = Pioneer 36-B-08; RX489 = Dekalb RX 489; TMF100= Mycogen TMF 100. 3 Probability that effect is significant; H = hybrid; DM = DM covariate. 4 Standard error of the mean. a,b,c 1
not similar, although all 3 methods resulted in adequate silage fermentation characteristics. Ranking of the hybrids for CP varied among the different methods, although in all 3 methods, the Agway 5215 hybrid ranked lowest and the Agway 5496 and Mycogen TMF 100 ranked highest. May et al. (2001) describe a vacuum bag system that they used to evaluate fungal communities associated with whole plant corn silage. In that study, they used
Figure 4. Influence of method of chopping on particle size expressed as % of DM remaining on screen. Journal of Dairy Science Vol. 87, No. 12, 2004
125 g of corn silage chopped into 20- to 30-mm lengths. The system was successfully used to study the influence of 2 different bacterial inoculants compared with an uninoculated control. In that study, bags were stored at room temperature for up to 3 months. We did not observe mold growth in any of the chipper-shredder mini-silo bags, suggesting that the bags, once sealed, are airtight. Three of the Claas-chopped bags did have a spot colony of mold, but mold was not prevalent throughout the bag. We observed no heating of the minisilos, possibly because the vacuum sealing removes most of the air. This would shorten the anaerobic fermentation phase, resulting in minimal heating. Some bags filled with gas after a few days, but none ruptured. Because fermentation characteristics were good, differences between types of bag may have been due to sampling. Haslemore and Holland (1981) reported that the minimum sample numbers required for representative sampling of field silos or stacks varied from 4 to 61 depending on the constituent. Analysis of pH required 4 samples to detect a difference of P < 0.05 between silages, whereas acetic acid required 32 samples and lactic acid 29 samples to detect the same difference (Haslemore and Holland, 1981). Cherney et al. (1996) suggested that selection of a smaller number of plants minimizes the problem of subsampling. Their data indicated that 2 plants would be required to detect a difference of 1.0% between hybrids in CP. Results from the present study support the conclusion that a 6-plant
CORN FORAGE ENSILED IN MINI-SILOS
sample is adequate to detect differences among corn silage samples ensiled in vacuum-sealed bags. It is recognized that a 6-plant sample may not generate the same values as when silage is sampled from the field. CONCLUSION O’Kiely and Wilson (1991) noted that interactions between laboratory silos and farm-scale silos could cause serious problems when treatments causing different fermentation patterns were being compared. Despite this potential difficulty, O’Kiely and Wilson (1991) concluded that laboratory silos were an accurate and reliable experimentation unit. Results from any type of laboratory silo should be field-tested using commercial silos before being applied at the producer level. There are inherent problems in all small-scale silo systems, with difficulty in obtaining some measures of silage quality, such as aerobic stability. Laboratory silos, however, are an accurate and reliable experimentation unit, if oxygen can be excluded to allow for consistent ensiling. Fermentation characteristics suggested that all samples in this study were adequately ensiled, but that the profile of the 50-g samples differed from other sample sizes. We conclude that it is possible to use vacuumsealed plastic bags to ensile corn, with samples as small as 200 g, to assess treatment differences when it is not practical to evaluate all treatments using commercial silos. ACKNOWLEDGMENTS This research was supported in part by the Cornell University Agricultural Experiment Station federal formula funds, Project No. NYC-1277455, received from Cooperative State Research, Education, and Extension Service, U.S. Department of Agriculture. Any opinions, findings, conclusions, or recommendations expressed in this publication are those of the authors and do not necessarily reflect the view of the U.S. Department of Agriculture. This study would not have been possible without the assistance of the field crew of the Cornell University Teaching and Research Facility. The assistance of Samuel Beer and T. W. Katsvairo is especially appreciated. REFERENCES Allen, M. S., J. G. Coors, and G. W. Roth. 2003. Corn silage. Pages 547–608 in Silage Science and Technology. D.R Buxton, R. Muck, and J. Harrison, ed. ASA, CSSA, and SSSA, Madison, WI. Anderson, R., H. I. Gracey, S. I. Kennedy, E. F. Unsworth, and R. W. J. Steen. 1989. Evaluation studies in the development of a commercial bacterial inoculant as an additive for grass silage. 1. Using pilot-scale tower silos. Grass and Forage Sci. 44:364–369.
4245
Association of Official Analytical Chemists. 1990. Official Methods of Analysis of the Association of Official Analytical Chemists. 15th ed. AOAC, Arlington, VA. Bolsen, K. K. 1995. Silage: Chapter 12. Silage: Basic Principles. Pages 163–176 in Forages, Vol. II: The Science of Grassland Agriculture. 5th ed. R. F. Barnes, C. J. Nelson, and D. Miller, ed. Iowa State University Press, Ames, IA. Buchanan-Smith, J. G. 1990. An investigation into palatability as a factor responsible for reduced intake of silage by sheep. Anim. Prod. 5:253–260. Cherney, D. J. R., M. J. Traxler, and J. B. Robertson. 1997. Use of Ankom fiber determination systems to determine digestibility. In Proc. NIRS Forage and Feed Testing Consortium Annual Conference, Madison, WI. Cherney, J. H., M. D. Casler, and D. J. R. Cherney. 1996. Sampling forage maize for quality. Can. J. Plant Sci. 76:93–99. Cherney, J. H., and D. J. R. Cherney. 2003. Assessing silage quality. Pages 141–198 in Silage Science and Technology. D.R. Buxton, R. Muck, and J. Harrison, ed. ASA, CSSA, and SSSA, Madison, WI. Cox, W. J., S. Kalonge, D. J. R. Cherney, and W. S. Reid. 1993. Growth, yield, and quality of forage maize under different nitrogen management practices. Agron. J. 85:341–347. Fisher, D. S., and J. C. Burns. 1987. Quality analysis of summerannual forages. II. Effects of forage carbohydrate constituents on silage fermentation. Agron. J. 79:242–248. Hall, M. B., W. H. Hoover, J. P. Jennings, and T. K. Miller Webster. 1999. A method for portioning neutral detergent soluble carbohydrates. J. Sci. Food Agric. 79:2079–2086. Haslemore, R. M., and R. Holland. 1981. Sampling and chemical composition of silage from a farm stack: Variation in moisture content and in degree of fermentation. J. Exp. Agric. 9:85–89. Konokoff, P. J., A. J. Heinrichs, and D. R. Buckmaster. 2003. Modification of the Penn State forage and total mixed ration particle separator and the effects of moisture content on its measurements. J. Dairy Sci. 86:1858–1863. Kung, L., and R. Shaver. 2001. Interpretation and use of silage fermentation analysis reports. Focus on Forage. 3:1–5. Littell, R. C., G. A. Milliken, W. W. Stroup, and R. D. Russell. 1996. SAS System for Mixed Models. SAS Institute Inc., Cary, NC. Lombard, G. L., and V. R. Dowell, Jr. 1962. Gas-Liquid Chromatography Analysis of the Acid Products of Bacteria (CDC Lab Manual). U.S. Dept. of Human Services, Public Health Service, Centers for Disease Control, Atlanta, GA. Marten, G. C., and R. F. Barnes. 1980. Prediction of energy digestibility of forages with in vitro rumen fermentation and fungal enzyme systems. Pages 61–71 in Standardization of Analytical Methodology for Feeds. W. J. Pigden, C. C. Balch, and M. Graham, ed. Int. Devel. Res. Ctr., Ottawa, ON, Canada. May, L. A., B. Smiley, and M. G. Schmidt. 2001. Comparative denaturing gradient gel electrophoresis analysis of fungal communities associated with whole plant corn silage. Can. J. Microbiol. 47:829–841. McDonald, P., N. Henderson, and S. Heron. 1991. The Biochemistry of Silage. 2nd ed. Chalcombe Publications, Marlow, UK. Meiske, J. C., J. G. Linn, and R. D. Goodrich. 1975. Types of laboratory silos and an evaluation of their usefulness. Pages 99–126 in Proc. 2nd Intl. Silage Res. Conf., National Silo Assoc., Inc. Waterloo, IA. National Forage Testing Association. 1993. National Forage Testing Association Recommends Forage Analyses Procedure. National Forage Testing Assoc., Omaha, NE. http://www.foragetesting.org/ lab_procedures.php. Accessed February 25, 2004. O’Kiely, P. 1993. Influence of a partially neutralised blend of aliphatic organic acids on fermentation, effluent production and aerobic stability of autumn grass silage. Irish J. Agric. Food Res. 32:13–26. O’Kiely, P., and R. K. Wilson. 1991. Comparison of three silo types used to study in-silo processes. Irish J. Agric. Res. 30:53–60. Pitt, R. E. 1990. Silage and Hay Preservation. NRAES-5, Northeast Regional Agricultural Engineering Service, Ithaca, NY. Journal of Dairy Science Vol. 87, No. 12, 2004
4246
CHERNEY ET AL.
Rook, A. J., and M. Gill. 1990. Prediction of the voluntary intake of grass silages by beef cattle. 1. Linear regression analyses. Anim. Prod. 50:425–438. Russell, J. R. 1986. Influence of harvest date on the nutritive value and ensiling characteristics of maize stover. Anim. Feed Sci. Technol. 14:11–27. SAS System, Version 7.0. 1998. SAS Institute Inc., Cary, NC. Supelco. 1990. Analyzing fatty acids by packed column gas chromatography. Supelco GC Bulletin 856A.
Journal of Dairy Science Vol. 87, No. 12, 2004
Van Soest, P. J., J. B. Robertson, and B. A. Lewis. 1991. Methods for dietary fiber, neutral detergent fiber, and nonstarch polysaccharides in relation to animal nutrition. J. Dairy Sci. 74:3583– 3597. Virtanen, A. I. 1947. Silage by the A.I.V. method. Economic Proc. of the Royal Dublin Society 3:311–343. Wisemann, H. G., and H. M. Irvin. 1957. Determination of organic acids in silage. J. Agric. Food Chem. 5:213–215.
Fermentation Characteristics of Corn Forage Ensiled in Mini-Silos D. J. R. Cherney,1 J. H. Cherney,2 and W. J. Cox2 1
Department of Animal Science, and Department of Crop and Soil Sciences, Cornell University, Ithaca, NY 14853
2
ABSTRACT
INTRODUCTION
To evaluate numerous experimental variables and their interactions involving different corn (Zea mays, L.) silage hybrids, scaled down mini-silos are necessary. Objectives of this study were to evaluate the influence of sample size on pH, NH3, and volatile fatty acid profile of 8 corn silage hybrids, selected to vary in fiber digestibility and ensiled in vacuum-sealed polyethylene bags for 90 d, and to assess the suitability of these mini-silos for detecting differences among corn silage samples. Hybrids were grown at the Cornell Teaching and Research Center located near Harford, NY, and harvested at a dry matter content of about 32% in the fall of 2002. Samples from 3 field replications of each hybrid were chipper-shredder chopped and vacuum-ensiled in bags with sample sizes of 50, 100, 200, 400, and 600 g. Increasing sample size resulted in decreased lactic acid, acetic acid, total acids, and NH3. Most of the difference among sample sizes occurred between the 50- and 100g sample sizes. Lactic acid:acetic acid ratio (3.1 ± 0.13) and pH (3.9 ± 0.08) did not vary among sample sizes. There was no detectable butyric acid in the samples. Fermentation characteristics suggested that all samples were well ensiled but that the fermentation profile of the 50-g samples differed the most from other sample sizes. Hybrids did vary in lactic acid, acetic acid, lactic acid:acetic acid, and pH. Differences among hybrids were also noted for dry matter and crude protein. Fieldchopped corn hybrids that were ensiled using mini-silos had higher acids than corresponding field-chopped corn hybrids ensiled in Ag-bags, in part due to no effluent escaping from the mini-silos. It is possible to use vacuum-sealed plastic bags to ensile corn, with samples as small as 200 g, and to use these mini-silos to assess differences among corn silage samples. Caution should be used when extrapolating mini-silo data to fieldscale ensiling. (Key words: laboratory silo, corn silage, fermentation)
A variety of traits and corn (Zea mays L.) silage processing techniques are being promoted for high milk production through improved forage quality. To evaluate numerous experimental variables and their interactions involving different corn silage hybrids, scaled down mini-silos are necessary. The entire contents of laboratory silos can be weighed, processed, and analyzed accurately. This can only be done under the assumption that the fermentation process is reasonably similar to that taking place in field-scale silos (Cherney and Cherney, 2003). Quality parameters that fluctuate during ensiling include temperature, pH, VFA, and aerobic stability (Anderson et al., 1989; Pitt, 1990; O’Kiely, 1993; May et al., 2001). Low-fiber silage with a pH < 4.2 is considered to be properly ensiled (Meiske et al., 1975; Cherney and Cherney, 2003). A survey of the limited number of studies directly comparing fermentation in field-scale and small-scale silos resulted in the conclusion that forage in both silo types did undergo similar fermentation (Meiske et al., 1975). Experimental silos enable experimental variables to be scaled down from field scale to experimental units that allow for multiple treatments and replications. Although the general ensiling process was similar for test tube, plastic pipe, and 4- to 8-tonne polyethyleneenclosed (clamp) silos, O’Kiely and Wilson (1991) concluded that there was potential for significant fermentation × silo type interactions when treatments that directly affected fermentation were being compared. Laboratory silos, however, are still considered a practical method of comparing a number of treatments (O’Kiely, 1993). Forage corn sealed in polyethylene bags has been used to demonstrate the ensiling process to undergraduate animal nutrition students (A.W. Bell, 2004, personal communication), but only pH and DM were assessed. The primary objectives of this study were to evaluate the influence of sample size on pH, NH3, and VFA profile of 8 corn silage hybrids, selected to vary in fiber digestibility and ensiled in vacuumsealed polyethylene bags for 90 d, and to assess the suitability of these mini-silos for detecting treatment differences among corn silage samples processed using a chipper-shredder. A secondary objective was to relate
Abbreviation key: HI = harvest index.
Received April 6, 2004. Accepted August 11, 2004. Corresponding author: D. J. R. Cherney; e-mail: [email protected].
4238
CORN FORAGE ENSILED IN MINI-SILOS
the fermentation characteristics of the chipper-shredder mini-silos to fermentation profiles obtained from fresh corn that is field-processed and ensiled in commercial field scale plastic bags or mini-silos. MATERIALS AND METHODS Silage corn hybrids Agway 5215, Agway 5496, Dekalb DKC 53-32, Dekalb RX 489, Mycogen BMR F407, Mycogen TMF 100, Pioneer 35P12, and Pioneer 36-B-08 were chosen to represent a range of fiber digestibility. Hybrids were planted on May 11, 2002, at a planting rate of about 32,400 plants/ha at the Teaching and Research Center located near Harford, NY, at a spacing of 76 cm in 12-row strips, replicated 3 times. The previous crop was alfalfa. About 224 kg/ha of 10-20-20 NPK fertilizer was applied through the corn planter, and about 84 kg/ ha of liquid N was applied at the 4-leaf stage of corn growth. Hybrids were harvested at a DM of about 32% in fall of 2002. For the primary objective, fresh corn samples were processed through a chipper-shredder (Mighty Mac LSC506, MacKissic, Inc., Parker Ford, PA). In the field, 3 plants were passed through a chipper-shredder each time a new hybrid was shredded to remove residual plant material that might still have been in the shredder. Six plants were then chipper-shredded and bagged separately in a tightly closed plastic bag to be used for ensiling in vacuum sealer bags. Shredded material from each hybrid × field replication was thoroughly mixed in a shallow plastic pan, resulting in a homogeneous mixture. Samples of 50, 100, 200, 300, 400, 500, and 600 g were vacuum-sealed into polyethylene bags (Nylon poly EVOH high barrier 3 mil 22 × 33 cm vacuum pouches), using a single chamber vacuum packaging machine (model KVP-420T with 512 mmHg vacuum level; Process Plus Food Processing Equipment Inc., Parker’s Prairie, MN) and stored for 90 d. Mini-silos were stored in the laboratory (21°C) in black plastic bags during the 90 d. Two subsamples from each hybrid × replicate were collected for determination of buffering capacity. To determine buffering capacity, 3 g of dry forage (60°C) was placed in 60 mL of distilled water and allowed to stand for 4 h. The pH was then measured and the samples titrated to pH 4.0 using 0.1 M lactic acid (Fisher and Burns, 1987). Four subsamples from each hybrid × replicate were collected for determination of sugars. Sugar was determined using a water-soluble sugar method (Hall et al., 1999). Another set of 6 plants was harvested at the same time to determine harvest index (HI), defined as the ratio of grain DM to total DM (Cox et al., 1993).
4239
The mini-silos were opened at the end of 90 d to determine pH and VFA content. Fifty grams of wet sample was homogenized with 500 mL of deionized water in a Waring blender (2 min) and filtered through 2 layers of cheesecloth before measuring pH; 8.35 mL of 1 M m-phosphoric acid was then added to 25 mL of the extract, which was frozen and later analyzed for VFA (procedure adapted from Wiseman and Irvin, 1957). Fermentation analysis of corn silage samples was performed by Dairy One (DHI Forage Testing Laboratory, Ithaca, NY). Samples were filtered through a disposable syringe filter [Millipore Millex 5.0 μm PVDF (Durapore) membrane, Millipore Corp., Billerica, MA] and analyzed for acetic, propionic, butyric, and isobutyric acids using gas chromatography (Supelco, 1990; procedures cited from Lombard and Dowell, Jr., 1962). Lactic acid was determined using an YSI 2700 SELECT Biochemistry Analyzer equipped with an L-lactate membrane. Crude protein (AOAC 976.06, 1990) and ammonia N (AOAC 941.04, 1990) were also determined. After samples for the above study were collected, corn forage was harvested with a commercial chopper (Claas-chopper, Jaguar 850; CLAAS of America, LLC, Columbus, IN; 21 mm setting). Field size for each hybrid × replication averaged 0.71 ± 0.017 ha. Corn forage was stored in separate plastic bags (Ag-bag International Ltd., Warrenton, OR) for each hybrid. Samples were treated with a lactic acid bacterial inoculant (Pioneer inoculant #1132) applied at 0.9 g/tonne through the forage harvester. Locations of each of the 3 field replications within each bag were noted. After 90 d, 4 samples were taken from random locations for each field replication in the Ag-bag using a Penn State hay core sampler. Samples were mixed, as described above, then immediately processed and analyzed as described above. Another set of mini-silo samples was prepared by sampling hybrid × replication described above as the Ag-bags were being filled. Mini-silos were then prepared as described above. Although we were primarily interested in the fermentation characteristics of the corn samples, Ag-bag samples were analyzed for DM, CP, NDF, and in vitro true digestibility to establish that there was a range in the chemical composition of the hybrids used. Samples were first dried at 60°C (NFTA, 1993), and then ground through a Wiley mill fitted with a 4-mm screen, followed by grinding with an Udy mill fitted with a 1-mm screen. Dry matter was determined by drying at 105°C for 24 h (NFTA, 1993). Nitrogen was determined by combustion (Leco Instruments, Inc., St. Joseph, MI) (AOAC 976.06, 1990) and multiplied by 6.25 to obtain CP. Neutral detergent fiber was analyzed according to Van Soest et al. (1991) using the ANKOM system for NDF. SulfiteJournal of Dairy Science Vol. 87, No. 12, 2004
4240
CHERNEY ET AL. Table 1. Dry matter, chemical, and fiber characteristics of corn silage hybrids ensiled in Ag-bags for 3 mo. Hybrid2 Variable1
AG5215
AG5496
BMR407
DK53-32
P35P12
P36B08
RX489
TMF100
SED3
DM, % NDF, % of DM IVTD, % of DM dNDF, % of DM CP, % of DM HI,4 % BC,4 mL of acid Sugar,4 % of DM
31.8abc 41.3bc 80.8a 53.6a 8.5ab 45.7b 8.3a 8.0a
30.3a 46.2c 81.1a 59.0b 9.1b 34.3a 10.0b 13.5bc
35.2c 34.9a 90.0c 71.2c 8.6ab 50.8c 8.2a 8.6a
34.4c 45.4c 82.9ab 62.3b 8.5ab 45.9b 9.4ab 9.6a
34.0abc 44.1bc 82.8ab 61.0b 8.9ab 44.4b 8.1a 10.18ab
35.6bc 42.4bc 82.1ab 57.8ab 8.9ab 43.3b 8.1a 12.42b
31.0ab 42.8bc 82.1ab 58.3ab 8.4a 35.8a 11.1b 15.4c
34.8c 39.0ab 84.4b 59.9b 9.1b 43.3b 9.6ab 8.89a
1.16 1.51 0.73 1.71 0.21 0.98 0.53 0.80
Means within a row with different superscripts differ (P ≤ 0.05). IVTD = In vitro true digestibility; dNDF = NDF digestibility; HI = harvest index; BC = buffering capacity. 2 AG5215 = Agway 5215; AG5496 = Agway 5496; BMR407 = Mycogen BMR F407; DK53-32 = Dekalb DKC 53-32; P35P12 = Pioneer 35-P-12; P36B08 = Pioneer 36-B-08; RX489 = Dekalb RX 489; TMF100= Mycogen TMF 100. 3 SED = Standard error of the differences of the least square means. 4 HI = Harvest index; BC = buffering capacity; determined from samples that were collected and dried at time of ensiling. a,b,c 1
and heat-stable alpha-amylase were used for NDF analysis of all samples. Samples were not corrected to an ash-free basis. In vitro fiber digestibility was determined according to Cherney et al. (1997), using the rumen buffer described by Marten and Barnes (1980) and using the Daisy II 200/220 in vitro incubator (ANKOM Technology, Fairport, NY) and the ANKOM 200/ 220 fiber analyzer (ANKOM Technology, Fairport, NY). The buffer contained urea. Ruminal fluid inoculum was obtained from a nonlactating, rumen-fistulated Holstein cow offered a medium quality orchardgrass (Dactylis glomerata L.) hay diet for ad libitum intake. Samples (0.25 g) were incubated for 48 h at 39°C, and undigested residues were treated with neutral detergent solution. The NDF digestibility was determined using the following formula: NDF digestibility = 100 − (NDF remaining at t = 48 h/NDF at t = 0 h) × 100. Particle sizes of silages from chipper-shredder and Claas-chopped mini-silos were analyzed using the Penn State Forage Particle separator (Kononoff et al., 2003). Particles were separated on 3 screens and a bottom pan, with screen hole sizes of 19, 8, and 1.18 mm. Material used for the 3 methods represented different sample populations. Therefore, they cannot be compared directly and were analyzed separately. For the chipper-shredder mini-silos, a split-plot in a randomized complete block design was used with 3 field replications. Hybrids were the main plots. Subplots consisted of the sample size. A split-plot ANOVA with repeated measures was used to test for statistical significance of treatment effects and interactions using PROC MIXED (Littell et al., 1996) in SAS, version 7.0 (SAS Inst., 1998). The model assumed that hybrid and sample size were fixed variables, and replication was considered a Journal of Dairy Science Vol. 87, No. 12, 2004
random variable. Sample DM was used as a covariate in the analysis. Separation of hybrid and sample size treatment means was accomplished using the TukeyKramer procedure for multiple comparisons (P ≤ 0.05). For the Claas-chopped mini-silos and Ag-bag material, a randomized design was used with 3 field replications. An ANOVA with repeated measures was used to test for statistical significance of treatment effects and interactions using PROC MIXED (Littell et al., 1996) in SAS, version 7.0 (SAS Inst., 1998). The model assumed that hybrid was a fixed variable, and replication was considered a random variable. Sample DM was used as a covariate. Separation of hybrid treatment means was accomplished using the Tukey-Kramer procedure for multiple comparisons (P ≤ 0.05). RESULTS AND DISCUSSION Ensiled corn hybrids exhibited a range in fiber, CP, digestibility, and other measures of forage quality (Table 1). Dry matter varied among hybrids, with a range of almost 5%, so DM was included as a covariate in the analysis of fermentation profiles. Hybrids were initially selected to vary in in vitro true digestibility and NDF digestibility. Based on the results observed here, we were successful in obtaining corn samples that varied in chemical composition. We needed this variation to test the ability of the mini-silo system to handle a range of corn silage samples. Size × hybrid interactions for the fermentation constituents measured were not significant (P > 0.05). Therefore, main effects will be discussed. Most of the differences among sample sizes occurred between the 50- and 100-g sample sizes (Figure 1). Increasing sam-
CORN FORAGE ENSILED IN MINI-SILOS
Figure 1. Influence of sample size on lactic acid, acetic acid, ammonia N, and pH. Means with different superscripts differ (P ≤ 0.05); comparisons do not include 50-g samples. pH did not differ among sample sizes (P > 0.05).
ple size from 50 to 100 g or more resulted in decreased lactic acid, acetic acid, total acids, and ammonia. There was very little or no propionic acid, butyric acid, or isobutyric acid in these samples, and sample sizes were not different (P > 0.05) in these constituents (data not shown). The reason for the large difference between the 50-g samples and the other sample sizes is not readily apparent. It is unlikely that sampling of the original fresh corn sample was the reason for the difference, as CV for the various sample sizes were similar (P > 0.05). Because of the large change in fermentation characteristics between 50- and 100-g samples, data was reanalyzed excluding the 50-g samples. There were still small but significant changes in acetic acid, lactic acid, and NH3 N between 100- and 200-g samples. Changes appeared to level off above 200 g for these constituents (Figure 1), as well as for total acids (6.68, 5.62, 5.33, and 5.44 % of DM, respectively, for 100-, 200-, 400-, and 600-g sample size, ± 0.31% SEM). There was a steady decrease (P < 0.05) in ammonia with increasing sample size (0.56, 0.52, 0.48, and 0.45% of DM, respectively, for 100-, 200-, 400-, and 600-g sample size, ± 0.38% SEM). The DM varied among sample sizes, but there was no apparent pattern (DM = 32.7, 32.4, 32.0, and 32.8% of DM, respectively, for 100-, 200-, 400-, and 600-g sample size, ± 0.78% SEM). The lactic:acetic acid ratio and CP did not vary among sample size. Variations in pH within a hybrid across sample size were generally less than 0.1 units and did not seem to follow a pattern. There were noticeable differences among hybrids. Kung and Shaver (2001) report that the typical range of pH for corn silage is 3.7 to 4.2. Silages were generally within that range regardless of
4241
hybrid or size of sample (Figure 2). One exception to this was the Dekalb RX 489 hybrid, which was always above pH 4.2, regardless of sample size. The Agway 5496 hybrid at sample sizes 400 and 600 g was also above pH 4.2 and tended to be the most variable of the hybrids in terms of pH across sample size. The Mycogen BMR F407 silage tended to be the lowest in pH, averaging about 3.7. There were no differences (P > 0.05) among hybrids in NH3, propionic acid, butryric acid, or isobutyric acid. Hybrids, however, did vary in lactic acid, acetic acid, lactic acid:acetic acid, total acids, and pH (Table 2). Differences among hybrids were noted for CP and DM. The Mycogen BMR F407 hybrid was highest in lactic acid, which would have contributed the most to lowering the pH. The Mycogen BMR F407 hybrid had the highest HI, which should result in the most substrate for the lactic acid bacteria (Table 1). Hybrids Agway 5496 and Dekalb RX 489, conversely, had the lowest HI. These 2 hybrids had the highest buffering capacities (Table 1) and the highest pH (Figure 2). Buffering capacity is generally associated with silage pH (McDonald et al., 1991); the association between buffering capacity and pH was high in this study (Figure 3; r = 0.69; P < 0.01). Sugar content of the hybrids was also associated with pH (Table 1; r = 0.63; P < 0.01). The positive association of sugars to pH is likely indicative of the low HI in hybrids with high sugars (Table 1). Russell (1986) noted that the nutritional value of corn stover was inversely related to grain yield and was due to a translocation of sugars from the stem to grain. The grain is a strong carbohydrate sink (Allen et al., 2003). Most of the total nonstructural carbohydrate in the grain is stored as starch, whereas in the stems it is stored as simple sugars. Typical concentrations on a DM basis range from 4 to 7% for lactic acid, 1 to 3% for acetic acid, <0.1% for propionic acid, 0% for butyric acid, and 5 to 7% for NH3 N (% of CP) (Kung and Shaver, 2001). Kung and Shaver (2001) also indicated that lactic acid should be 65 to 70% of total acids. Using these values as indicative of adequate ensiling, all sample sizes greater than 50 g fell within the range considered typical (Figure 1). The 50-g samples were higher in lactic acid than the typical range, which would be acceptable, but they were also higher in acetic acid and NH3 N concentrations. Some studies suggest that acetic acid concentrations above 5 to 6 % of DM will depress animal intake (BuchananSmith, 1990; Rook and Gill, 1990). Concentrations of NH3 N above 12 to 15% indicate excessive protein breakdown in the silage (Kung and Shaver, 2001). From our data, we conclude that the 50-g mini-silos are not representative of typical silage fermentation. Journal of Dairy Science Vol. 87, No. 12, 2004
4242
CHERNEY ET AL.
Figure 2. Influence of hybrid and sample size on pH. Vertical bars indicate standard error of the mean.
Only one hybrid, Pioneer 36-B-08, fell outside the typical range for lactic acid when averaging 100- to 600g sample sizes (Table 2). Other variables were within the range considered typical. Lactic acid still represented 70% of total acids and pH was low. This would suggest that this hybrid was well ensiled. The data
for the hybrids suggests that the vacuum bag method produces silage that falls within the range considered typical. Thus, this method is sensitive enough to detect differences among hybrids in a laboratory setting where treatment differences are carefully controlled. Because the chipper-shredder method uses just 6 plants out of
Table 2. Fermentation analysis of chipper-chopped silage, no inoculant, after 90 d of storage.1 Effect (P ≤)3
Hybrid2 Variable Lactic acid, % of DM Acetic acid, % of DM Lactic:acetic, % of DM Total acids, % of DM pH CP, % of DM Ammonia, % of DM Ammonia N, % of CP DM, %
AG5215 ab
4.1 (0.354) 1.6b (0.06) 2.6a (0.25) 5.7a (0.36) 3.80bc (0.02) 8.0a (0.17) 0.50a (0.048) 6.31bc (0.501) 31.5ab (1.147)
AG5496 a
3.7 (0.35) 1.2a (0.06) 3.2ab (0.25) 5.0a (0.36) 4.14d (0.02) 9.3d (0.17) 0.48a (0.047) 5.21a (0.500) 31.5ab (1.14)
BMR407 c
5.4 (0.35) 1.2a (0.06) 4.4c (0.25) 6.6b (0.37) 3.72a (0.02) 7.9a (0.18) 0.51a (0.049) 6.45bc (0.513) 31.0ab (1.14)
DK53-32 abc
4.3 (0.35) 1.6b (0.06) 2.7a (0.25) 5.9ab (0.36) 3.85c (0.02) 8.5b (0.17) 0.58a (0.048) 6.72c (0.501) 33.3ab (1.14)
P35P12 bc
4.8 (0.35) 1.6b (0.06) 2.9a (0.25) 6.4b (0.36) 3.79ab (0.02) 8.8bc (0.17) 0.49a (0.048) 5.54ab (0.501) 33.3ab (1.14)
P36B08 a
3.9 (0.36) 1.5b (0.06) 2.5a (0.25) 5.4a (0.37) 3.82bc (0.02) 8.7bc (0.18) 0.48a (0.049) 5.45ab (0.513) 34.0b (1.14)
RX489 abc
4.3 (0.36) 1.2a (0.06) 3.7b (0.25) 5.4a (0.37) 4.28e (0.03) 8.6bc (0.18) 0.47a (0.049) 5.43ab (0.518) 30.9a (1.14)
TMF100 abc
4.2 (0.35) 1.6b (0.06) 2.6a (0.25) 5.8ab (0.38) 3.86c (0.02) 8.9cd (0.18) 0.49a (0.050) 5.44ab (0.525) 34.2b (1.14)
DM
H
S
HXS
0.01
0.04
0.03
0.99
0.03
0.01
0.01
0.97
0.01
0.01
0.50
0.99
0.01
0.01
0.01
0.99
0.07
0.01
0.05
0.53
0.01
0.01
0.47
0.79
0.26
0.60
0.03
0.63
0.68
0.11
0.03
0.65
...
0.09
0.01
0.21
Means within a row with different superscripts differ (P ≤ 0.05). Average of 200- to 600-g mini-silos, 3 replications. 2 AG5215 = Agway 5215; AG5496 = Agway 5496; BMR407 = Mycogen BMR F407; DK53-32 = Dekalb DKC 53-32; P35P12 = Pioneer 35P-12; P36B08 = Pioneer 36-B-08; RX489 = Dekalb RX 489; TMF100= Mycogen TMF 100. 3 Probability that effect is significant; DM = DM covariate; H = hybrid; S = sample size; HXS = sample × size interaction. 4 Standard error of the mean. a,b,c,d,e 1
Journal of Dairy Science Vol. 87, No. 12, 2004
4243
CORN FORAGE ENSILED IN MINI-SILOS
inoculant was added to the Claas-chopped material. Adding lactic acid bacteria to the Claas-chopped material would result in differences between this material and the chipper-shredded material. Particle size of chipper-shredder samples was smaller than corresponding Claas-chopped material (Figure 4). Chipper-shredder samples had less than 10% remaining above the 19-mm screen, whereas the Claaschopped samples averaged above 40%. On average, chipper-shredder samples retained more on screens smaller than 19 mm than did Claas-chopped samples (Figure 4). The finer chop of the chipper-shredder material may have altered the fermentation process. Chopping tends to increase microbial numbers (Bolsen, 1995), and the finer chop would have liberated more nutrients. Differences between the Claas-chopped forage ensiled in mini-silos (Table 3) and the Claas-chopped forage material in Ag-bags (Table 4) were generally less than those observed between chipper-shredder material (Table 2) and Claas-chopped material. Samples from Claas-chopped mini-silos tended to be higher in acids than Ag-bag samples. Effluent cannot escape from the vacuum-packed mini-silos, which would account for higher acids. Virtanen (1947) indicated that this effluent could lead to more extensive fermentation. O’Kiely and Wilson (1991) reported that lactic acid was higher in test-tube silages (no effluent escape) than in clamp silos. Hybrid ranking of the 3 methods for acids was
Figure 3. Relationship between pH and buffering capacity.
a field, there is likely to be some variation between the 6 sampled plants and the whole field, particularly when there is variation among the plants in the field. For that reason, caution must be exercised when extrapolating laboratory results to field studies. Although not directly comparable to the chippershredder samples, Claas-chopped inoculated samples appear to have more lactic acid, which would be expected (Tables 3 and 4), because a lactic acid bacterial
Table 3. Fermentation analysis of Claas-chopped silage, with inoculant, 600-g mini-silos after 90 d of storage.1 Effect (P ≤)3
Hybrid2 Variable Lactic acid, % of DM Acetic acid, % of DM Lactic:acetic, % of DM Total acids, % of DM pH CP, % of DM Ammonia, % of DM Ammonia N, % of CP DM, %
AG5215 a
9.0 (0.364) 1.8ab (0.09) 5.1a (0.35) 10.8ab (0.32) 3.80a (0.030) 8.2a (0.12) 0.64bc (0.022) 7.8b (0.28) 30.8b (0.64)
AG5496 a
7.5 (0.44) 1.5a (0.10) 5.2a (0.43) 8.9ab (0.39) 3.80a (0.037) 9.0b (0.15) 0.53a (0.028) 6.0a (0.34) 29.2b (0.64)
BMR407
DK53-32
a
a
8.5 (0.38) 1.8ab (0.09) 4.6a (0.37) 10.4ab (0.33) 3.73a (0.032) 8.3ab (0.13) 0.51a (0.024) 6.2a (0.29) 32.0bc (0.64)
8.9 (0.37) 1.8ab (0.08) 5.1a (0.36) 10.6ab (0.32) 3.80a (0.031) 8.5ab (0.13) 0.59ab (0.023) 6.9ab (0.29) 31.7abc (0.64)
P35P12 a
7.5 (0.36) 2.0b (0.08) 3.9a (0.35) 9.6a (0.32) 3.80a (0.030) 8.5ab (0.12) 0.66c (0.023) 7.8b (0.28) 30.9b (0.64)
P36B08 a
8.3 (0.36) 1.7ab (0.08) 5.0a (0.35) 10.0ab (0.32) 3.77a (0.030) 8.7ab (0.12) 0.55ab (0.023) 6.3a (0.28) 31.4abc (0.64)
RX489 a
8.2 (0.51) 1.9b (0.11) 4.4a (0.49) 10.2ab (0.44) 3.70a (0.042) 9.1b (0.17) 0.62abc (0.032) 6.9ab (0.38) 28.6ab (0.64)
TMF100 a
9.9 (0.56) 1.7ab (0.13) 5.7a (0.54) 11.7b (0.49) 3.87a (0.046) 8.9ab (0.19) 0.62abc (0.035) 7.0ab (0.43) 34.1c (0.64)
H
DM
0.05
0.07
0.02
0.92
0.15
0.31
0.02
0.05
0.13
0.09
0.01
0.07
0.01
0.49
0.01
0.19
0.01
...
Means within a row with different superscripts differ (P ≤ 0.05). Average of mini-silos, 3 replications. 2 AG5215 = Agway 5215; AG5496 = Agway 5496; BMR407 = Mycogen BMR F407; DK53-32 = Dekalb DKC 53-32; P35P12 = Pioneer 35P-12; P36B08 = Pioneer 36-B-08; RX489 = Dekalb RX 489; TMF100= Mycogen TMF 100. 3 Probability that effect is significant; H = hybrid; DM = DM covariate. 4 Standard error of the mean. a,b,c 1
Journal of Dairy Science Vol. 87, No. 12, 2004
4244
CHERNEY ET AL.
Table 4. Fermentation analysis of Claas-chopped silage, with inoculant, after 5 mo of storage in an Ag-bag.1 Effect (P ≤)3
Hybrid2 Variable
AG5215
AG5496
BMR407
DK53-32
P35P12
P36B08
RX489
TMF100
DM
Lactic acid, % of DM
6.8a (0.234) 1.1a (0.11) 6.1a (0.52) 7.9a (0.29) 3.82a (0.023) 8.1a (0.16) 0.44a (0.024) 5.4a (0.30) 29.9ab (0.59)
6.8a (0.32) 1.0a (0.15) 7.2a (0.72) 7.8a (0.41) 3.88a (0.030) 8.6ab (0.22) 0.52ab (0.033) 6.1a (0.42) 28.4a (0.59)
6.8a (0.24) 1.4a (0.11) 4.7a (0.53) 8.2a (0.30) 3.77a (0.023) 8.5ab (0.16) 0.48a (0.024) 5.6a (0.31) 33.2c (0.59)
7.9b (0.24) 1.3a (0.11) 6.3a (0.53) 9.2a (0.30) 3.77a (0.023) 8.3a (0.16) 0.51ab (0.024) 6.2a (0.31) 33.2c (0.59)
7.5ab (0.23) 1.3a (0.11) 6.1a (0.51) 8.8a (0.29) 3.78a (0.023) 8.6ab (0.16) 0.58b (0.023) 6.7a (0.30) 33.0c (0.59)
7.4ab (0.22) 1.3a (0.10) 5.4a (0.49) 8.7a (0.28) 3.78a (0.022) 8.3a (0.15) 0.52ab (0.023) 6.2a (0.29) 32.8c (0.59)
7.0ab (0.29) 0.9a (0.14) 7.9a (0.65) 7.9a (0.36) 3.91b (0.028) 8.2a (0.20) 0.54ab (0.030) 6.6a (0.38) 28.9a (0.59)
7.6ab (0.22) 1.2a (0.10) 6.6a (0.49) 8.8a (0.27) 3.91b (0.022) 8.9b (0.15) 0.61b (0.022) 6.9a (0.29) 32.7a (0.59)
0.07
Acetic acid, % of DM Lactic:acetic, % of DM Total acids, % of DM pH CP, % of DM Ammonia, % of DM Ammonia N, % of CP DM, %
0.01 0.03 0.02 0.06 0.29 0.03 0.11 ...
Means within a row with different superscripts differ (P ≤ 0.05). Average of 3 replications. 2 AG5215 = Agway 5215; AG5496 = Agway 5496; BMR407 = Mycogen BMR F407; DK53-32 = Dekalb DKC 53-32; P35P12 = Pioneer 35P-12; P36B08 = Pioneer 36-B-08; RX489 = Dekalb RX 489; TMF100= Mycogen TMF 100. 3 Probability that effect is significant; H = hybrid; DM = DM covariate. 4 Standard error of the mean. a,b,c 1
not similar, although all 3 methods resulted in adequate silage fermentation characteristics. Ranking of the hybrids for CP varied among the different methods, although in all 3 methods, the Agway 5215 hybrid ranked lowest and the Agway 5496 and Mycogen TMF 100 ranked highest. May et al. (2001) describe a vacuum bag system that they used to evaluate fungal communities associated with whole plant corn silage. In that study, they used
Figure 4. Influence of method of chopping on particle size expressed as % of DM remaining on screen. Journal of Dairy Science Vol. 87, No. 12, 2004
125 g of corn silage chopped into 20- to 30-mm lengths. The system was successfully used to study the influence of 2 different bacterial inoculants compared with an uninoculated control. In that study, bags were stored at room temperature for up to 3 months. We did not observe mold growth in any of the chipper-shredder mini-silo bags, suggesting that the bags, once sealed, are airtight. Three of the Claas-chopped bags did have a spot colony of mold, but mold was not prevalent throughout the bag. We observed no heating of the minisilos, possibly because the vacuum sealing removes most of the air. This would shorten the anaerobic fermentation phase, resulting in minimal heating. Some bags filled with gas after a few days, but none ruptured. Because fermentation characteristics were good, differences between types of bag may have been due to sampling. Haslemore and Holland (1981) reported that the minimum sample numbers required for representative sampling of field silos or stacks varied from 4 to 61 depending on the constituent. Analysis of pH required 4 samples to detect a difference of P < 0.05 between silages, whereas acetic acid required 32 samples and lactic acid 29 samples to detect the same difference (Haslemore and Holland, 1981). Cherney et al. (1996) suggested that selection of a smaller number of plants minimizes the problem of subsampling. Their data indicated that 2 plants would be required to detect a difference of 1.0% between hybrids in CP. Results from the present study support the conclusion that a 6-plant
CORN FORAGE ENSILED IN MINI-SILOS
sample is adequate to detect differences among corn silage samples ensiled in vacuum-sealed bags. It is recognized that a 6-plant sample may not generate the same values as when silage is sampled from the field. CONCLUSION O’Kiely and Wilson (1991) noted that interactions between laboratory silos and farm-scale silos could cause serious problems when treatments causing different fermentation patterns were being compared. Despite this potential difficulty, O’Kiely and Wilson (1991) concluded that laboratory silos were an accurate and reliable experimentation unit. Results from any type of laboratory silo should be field-tested using commercial silos before being applied at the producer level. There are inherent problems in all small-scale silo systems, with difficulty in obtaining some measures of silage quality, such as aerobic stability. Laboratory silos, however, are an accurate and reliable experimentation unit, if oxygen can be excluded to allow for consistent ensiling. Fermentation characteristics suggested that all samples in this study were adequately ensiled, but that the profile of the 50-g samples differed from other sample sizes. We conclude that it is possible to use vacuumsealed plastic bags to ensile corn, with samples as small as 200 g, to assess treatment differences when it is not practical to evaluate all treatments using commercial silos. ACKNOWLEDGMENTS This research was supported in part by the Cornell University Agricultural Experiment Station federal formula funds, Project No. NYC-1277455, received from Cooperative State Research, Education, and Extension Service, U.S. Department of Agriculture. Any opinions, findings, conclusions, or recommendations expressed in this publication are those of the authors and do not necessarily reflect the view of the U.S. Department of Agriculture. This study would not have been possible without the assistance of the field crew of the Cornell University Teaching and Research Facility. The assistance of Samuel Beer and T. W. Katsvairo is especially appreciated. REFERENCES Allen, M. S., J. G. Coors, and G. W. Roth. 2003. Corn silage. Pages 547–608 in Silage Science and Technology. D.R Buxton, R. Muck, and J. Harrison, ed. ASA, CSSA, and SSSA, Madison, WI. Anderson, R., H. I. Gracey, S. I. Kennedy, E. F. Unsworth, and R. W. J. Steen. 1989. Evaluation studies in the development of a commercial bacterial inoculant as an additive for grass silage. 1. Using pilot-scale tower silos. Grass and Forage Sci. 44:364–369.
4245
Association of Official Analytical Chemists. 1990. Official Methods of Analysis of the Association of Official Analytical Chemists. 15th ed. AOAC, Arlington, VA. Bolsen, K. K. 1995. Silage: Chapter 12. Silage: Basic Principles. Pages 163–176 in Forages, Vol. II: The Science of Grassland Agriculture. 5th ed. R. F. Barnes, C. J. Nelson, and D. Miller, ed. Iowa State University Press, Ames, IA. Buchanan-Smith, J. G. 1990. An investigation into palatability as a factor responsible for reduced intake of silage by sheep. Anim. Prod. 5:253–260. Cherney, D. J. R., M. J. Traxler, and J. B. Robertson. 1997. Use of Ankom fiber determination systems to determine digestibility. In Proc. NIRS Forage and Feed Testing Consortium Annual Conference, Madison, WI. Cherney, J. H., M. D. Casler, and D. J. R. Cherney. 1996. Sampling forage maize for quality. Can. J. Plant Sci. 76:93–99. Cherney, J. H., and D. J. R. Cherney. 2003. Assessing silage quality. Pages 141–198 in Silage Science and Technology. D.R. Buxton, R. Muck, and J. Harrison, ed. ASA, CSSA, and SSSA, Madison, WI. Cox, W. J., S. Kalonge, D. J. R. Cherney, and W. S. Reid. 1993. Growth, yield, and quality of forage maize under different nitrogen management practices. Agron. J. 85:341–347. Fisher, D. S., and J. C. Burns. 1987. Quality analysis of summerannual forages. II. Effects of forage carbohydrate constituents on silage fermentation. Agron. J. 79:242–248. Hall, M. B., W. H. Hoover, J. P. Jennings, and T. K. Miller Webster. 1999. A method for portioning neutral detergent soluble carbohydrates. J. Sci. Food Agric. 79:2079–2086. Haslemore, R. M., and R. Holland. 1981. Sampling and chemical composition of silage from a farm stack: Variation in moisture content and in degree of fermentation. J. Exp. Agric. 9:85–89. Konokoff, P. J., A. J. Heinrichs, and D. R. Buckmaster. 2003. Modification of the Penn State forage and total mixed ration particle separator and the effects of moisture content on its measurements. J. Dairy Sci. 86:1858–1863. Kung, L., and R. Shaver. 2001. Interpretation and use of silage fermentation analysis reports. Focus on Forage. 3:1–5. Littell, R. C., G. A. Milliken, W. W. Stroup, and R. D. Russell. 1996. SAS System for Mixed Models. SAS Institute Inc., Cary, NC. Lombard, G. L., and V. R. Dowell, Jr. 1962. Gas-Liquid Chromatography Analysis of the Acid Products of Bacteria (CDC Lab Manual). U.S. Dept. of Human Services, Public Health Service, Centers for Disease Control, Atlanta, GA. Marten, G. C., and R. F. Barnes. 1980. Prediction of energy digestibility of forages with in vitro rumen fermentation and fungal enzyme systems. Pages 61–71 in Standardization of Analytical Methodology for Feeds. W. J. Pigden, C. C. Balch, and M. Graham, ed. Int. Devel. Res. Ctr., Ottawa, ON, Canada. May, L. A., B. Smiley, and M. G. Schmidt. 2001. Comparative denaturing gradient gel electrophoresis analysis of fungal communities associated with whole plant corn silage. Can. J. Microbiol. 47:829–841. McDonald, P., N. Henderson, and S. Heron. 1991. The Biochemistry of Silage. 2nd ed. Chalcombe Publications, Marlow, UK. Meiske, J. C., J. G. Linn, and R. D. Goodrich. 1975. Types of laboratory silos and an evaluation of their usefulness. Pages 99–126 in Proc. 2nd Intl. Silage Res. Conf., National Silo Assoc., Inc. Waterloo, IA. National Forage Testing Association. 1993. National Forage Testing Association Recommends Forage Analyses Procedure. National Forage Testing Assoc., Omaha, NE. http://www.foragetesting.org/ lab_procedures.php. Accessed February 25, 2004. O’Kiely, P. 1993. Influence of a partially neutralised blend of aliphatic organic acids on fermentation, effluent production and aerobic stability of autumn grass silage. Irish J. Agric. Food Res. 32:13–26. O’Kiely, P., and R. K. Wilson. 1991. Comparison of three silo types used to study in-silo processes. Irish J. Agric. Res. 30:53–60. Pitt, R. E. 1990. Silage and Hay Preservation. NRAES-5, Northeast Regional Agricultural Engineering Service, Ithaca, NY. Journal of Dairy Science Vol. 87, No. 12, 2004
4246
CHERNEY ET AL.
Rook, A. J., and M. Gill. 1990. Prediction of the voluntary intake of grass silages by beef cattle. 1. Linear regression analyses. Anim. Prod. 50:425–438. Russell, J. R. 1986. Influence of harvest date on the nutritive value and ensiling characteristics of maize stover. Anim. Feed Sci. Technol. 14:11–27. SAS System, Version 7.0. 1998. SAS Institute Inc., Cary, NC. Supelco. 1990. Analyzing fatty acids by packed column gas chromatography. Supelco GC Bulletin 856A.
Journal of Dairy Science Vol. 87, No. 12, 2004
Van Soest, P. J., J. B. Robertson, and B. A. Lewis. 1991. Methods for dietary fiber, neutral detergent fiber, and nonstarch polysaccharides in relation to animal nutrition. J. Dairy Sci. 74:3583– 3597. Virtanen, A. I. 1947. Silage by the A.I.V. method. Economic Proc. of the Royal Dublin Society 3:311–343. Wisemann, H. G., and H. M. Irvin. 1957. Determination of organic acids in silage. J. Agric. Food Chem. 5:213–215.