Evaluation of Methods for Determination of Bulk Density of Eight Kinds of Forage under Air-dry and Wet Conditions

The conditions of measurement for the determination of bulk density were evaluated to assess the bulkiness of 8 kinds of forage. The bulkiness of the forages was determined with 4 different sizes of forage samples with 7 different pressure application under air-dry and wet conditions. The dry bulk density (DBD) curvilinearly regressed with the pressure applied. The particle size of the samples and kinds of forage used in the present study did not affect changes in values of DBD determined under pressures over 20 g/cm 2 up to 200 g/cm 2 . The values of the wet bulk density (WBD) increased as an increment of particle size, but were not always regressed on the particle size of the 8 kinds of forage. The DBD determined on 8 mm particles showed a higher correlation coefficient with neutral detergent fiber (NDF) contents. The DBD may be a useful tool for the assessment of NDF in forage, when it is determined under condition of a pressure of 100 g/cm 2 or over with a particle size of 8 mm. The WBD may not be utilized for the direct measurement of the physical characteristics of forage, but may be required a thorough consideration on water solubility of forages. Further studies are needed to clarify the DBD contribution to the prediction of forage intake by ruminants. (Asian-Aust. J. Anim. Sci. 2003. Vol 16, No. 8 : 1126-1130)


INTRODUCTION
Intakes of forages by ruminants dependent upon diverse factors relating to the diet, animal and feeding condition (Forbes, 1995).Waldo (1986) stated that intake of temperate grasses and legumes was generally related to cell wall concentration which may have predicted gut fill as it did intake.The rumen fill becomes a primary function limiting the forage intake by ruminants (Conrad et al., 1964;Waldo, 1986).The rumen fill of sheep given a sole hay diet was reported to be 2.4 times of dry-matter intake, of which particles larger than 1.18 mm occupied more than 40% (Sekine et al., 1992).Okamoto et al. (1989) suggested that a large particle pool controlled the dry-matter intake of sheep.The distribution of particle size in a forage diet will affect the bulkiness of the corresponding forage, which thus influences the amount to be ingested by the animals.Seoane et al. (1982) reported that the forage intake was related to the density of feed.In general, forages contain a higher fiber fraction that is considered to relate to the physical characteristics of the forage.The physical characteristics of forage was inferred to influence upon the intake of the animals (Seoane et al., 1982, Mertens, 1987).Baile and Pfander (1967) found the relationship between voluntary bulk intake and density of ration.It was a remarkably close relation (Forbes, 1995).Forbes (1995) stated, however, that if more fibrous material has a lower bulk density, neutral detergent fiber (NDF) may be useful for predicting intake, since a few studies on density have been published.The measurement of density, however, is simple and less expensive compared with that of NDF.It requires only a graduated cylinder as described by Baile and Pfander (1967).Therefore, it is worth to establish a method to measure a simple index to express density or bulkiness of forage.
The present study was carried out to assess a simple but effective method for the determination of bulkiness to express one of the physical characteristics of forages.

MATERIALS AND METHODS
The forages used in the present study were alfalfa (Medicago sativa) hay (AL), Italian ryegrass (Lolium italicum) hay (IR), perennial ryegrass (Lolium perenne) hay (PR), timothy (Phreum pratense) hay (TI), bermudagrass (Cynodon dactylon) hay (BG), sudangrass (Sorghum sudanense) hay (SG), oats (Avena sativa) hay (OH) and rice (Oryza sativa) straw (RS).The bulkiness of the forage was expressed as a bulk density (g/mL).Four sizes of forage samples were prepared by grinding the forage with a cutting mill through 8, 4, 2 and 1 mm screens.Duplicated samples were then placed in a 100 mL calibrated glass cylinder with a 4.6 cm inner diameter and 16.6 cm 2 area of a horizontal section.Pressure was then applied at levels of 2, 5, 10, 20, 50, 100 and 200 g/cm 2 , respectively.Ten seconds after the pressure was applied, the volume was read by the graduation.The weight of sample was divided by volume to be expressed as dry bulk density (DBD, g/mL).The ground samples were also placed in a 200 mL calibrated resin cylinder with a 3.8 cm inner diameter and equipped with a drainage cock.Tap water was then added to the cylinder.The samples were kept standing for 8 to 10 minutes to have a good affinity to the water under a vacuum bell and then evacuated at a level of 2.4 kPa for 5 min.After evacuation, the water was drained to a level of 1 cm below the sample surface and the volume of the sample was determined.A wet bulk density (WBD, g/mL) was calculated in the same manner as the DBD.The proximate composition of forages was determined by the method of A.O. A.C. (1984).The determination of detergent fiber was done by the method described by Goering and Van Soest (1970).
Statistical analyses were done by the method described by Steel and Torrie (1960) by using the General AOV/AOCV procedure of the Statistix7 (Analytical Software, 2000).Regression analyses were also done by the method described by Steel and Torrie (1960).

RESULTS AND DISCUSSION
The chemical composition of the forages was presented in Table 1.The contents of acid detergent lignin (ADL) averaged 49±12 g/kg of dry matter (DM) for all of the forages, with the highest value for alfalfa hay (75 g/kg DM) and the lowest for bermudagrass hay (35 g/kg DM).The contents of neutral detergent fiber (NDF) averaged 673±87 g/kg DM, with the highest value for rice straw (769 g/kg DM) and the lowest for alfalfa hay (481 g/kg DM).For crude protein, the highest value was 147 g/kg DM for alfalfa and the lowest 37 g/kg DM for rice straw.The mean was 76±35 g/kg DM for all of the forages.
3.6×10 -4 ±0.5×10 -4 for forages used in the present study.Thus, it is inferred that the particle size of the samples and kinds of forage used in the present study did not affect changes in the values of DBD determined under pressures over 20 g/cm 2 up to 200 g/cm 2 .Correlation analyses revealed that there was significant correlation between NDF in forage and DBD determined for the particle sizes of 4 mm (r=-0.818,p<0.05, for the pressure of 20 g/cm 2 , r=-0.835, p<0.05, for the pressure of 50 g/cm 2 , r=-0.829, p<0.05, for the pressure of 100 g/cm 2 , and r=-0.887,p<0.01, for the pressure of 200 g/cm 2 ) and 8 mm (r=-0.862,p<0.01, for the pressure of 20 g/cm 2 , r=-0.916, p<0.01, for the pressure of 50 g/cm 2 , r=-0.938, p<0.01, for the pressure of 100 g/cm 2 , and r=-0.931,p<0.01, for the pressure of 200 g/cm 2 ), irrespective of the pressure applied, but not always significant for the particle sizes of 2 and 1 mm.The Values of DBD were not necessarily consistent to the contents of ADL for forages studied nor to the chemical composition determined in the present study.The DBD determined on 8 mm particles showed a higher significant correlation with NDF contents under the pressure of 100 g/cm 2 and 200 g/cm 2 (p<0.01)compared with those determined by 4 mm particles (p<0.05).When the particle sizes were reduced less than 4 mm, no significant correlation was found between DBD values and NDF contents.Comminuting forage for the measurement of DBD to a certain extent may loose the relationship between the physical characteristics and chemical composition of the forage.There was no significant correlation found between crude protein contents and DBD.Van Soest (1965) reported that in terms of chemical composition, the only consistent effect on voluntary intake is that of the total fibrous fraction, cell-wall constituents.It has been also stated that NDF may be a useful tool for predicting intake (Forbes, 1995).Thus, the determination of DBD for forages is expected to be a simple measure for the estimation of NDF contents in the forage, which may result in one of simple indexes for the prediction of forage intake by ruminants.Considering the results of the present study, the bulk density may be a useful tool for assessment of NDF in forage.
Although the particle sizes of 4 and 8 mm showed a significant correlation between NDF and DBD, the level of significance was higher in 8-mm particles (p<0.01)than 4 mm particles (p<0.05).A higher pressure applied over 20 g/cm 2 resulted in a higher coefficient of correlation in 8-mm particles, in which the pressure over 100 g/cm 2 showed a higher coefficient of correlation compared with those of 50 g/cm 2 or less.Therefore, it is inferred that 8-mm size is to be selected for the particle size of the samples, and the pressure over 100 g/cm 2 may be preferable, even though the DBD values were not affected by the size of samples under the pressure over 20 g/cm 2 .
The values of WBD increased as an increment of the particle size.Regression analyses of WBD revealed that SG and OH failed to have a statistically significant regression of WBD on the particle size studied in the present study.The Values of WBD were not necessarily consistent to the contents of NDF for forages studied nor to the other chemical compositions determined in the present study.Wattiaux et al. (1993) have found that partial solubilization of dry-matter in distilled water resulted in losses of mass and volume of samples for alfalfa hay, alfalfa silage and bromegrass hay.They indicated that the soluble fraction contributed more to the mass than to the volume of feed particles.The determination of bulk density by soaking in tap water may have resulted in the diverse extent and degree of water solubility in the hays used in the present study.Thus, the values of WBD may be affected by the difference in quantities of dry matter solubilized in water among hays.The measurement of WBD may not be utilized as an index for physical characteristics.Therefore, the measurement should be carried out under a precisely predetermined procedure when WBD is to be utilized as an index for physical characteristics of forage.
It is concluded that the measurement of DBD should be carried out at a pressure greater than 20 g/cm 2 as possible to reduce a possibility of a curvilinear change in DBD determination when DBD is to be utilized as an index for the physical characteristics of forage.It is preferable that the DBD is determined under the pressure over 100 g/cm 2 with the particle size of 8 mm.
The WBD may not be utilized for the direct measurement of the physical characteristics of forage.It may require a thorough consideration on the water solubility of forages.Further studies are needed to clarify a possibility of contribution by DBD to what extent the quantity of forage intake by ruminants may be predicted by the physical characteristics of forage.

Table 1 .
Mean values of chemical composition of eight kinds of

Table 2 .
Regression equations of DBD values (Y, g/mL) on the pressure applied (X, g/cm 2 ) for eight kinds of forage