Tải bản đầy đủ - 0 (trang)
X. The Nutritive Value of Grazed Forage

X. The Nutritive Value of Grazed Forage

Tải bản đầy đủ - 0trang



The most commonly used techniques are based on a rearrangement of

Eq. (2):

Intake of forage (1)

= weight

of feces voided (F) X


(100- 7% digestibility of forage eaten (D)



and requires estimates to be made of (F) and (D); from (D) the metabolizable and net energy values of the forage eaten can also be calculated

to give estimates of nutrient intake (Eqs. 19 and 20). (See reviews by

Commonwealth Agricultural Bureaux, 196 I , p. 92; 1. W. McDonald,

1962; J. T. Reid, 1962.)

1. The Estimation of Fecal Output by Grazing Animals

The most precise method of measuring fecal output is undoubtedly by

total collection, using harness and fecal bags. However this method is

laborious, is difficult under extensive grazing conditions, and may interfere with animal behavior and production, so that most workers have preferred the indirect estimation of fecal production by the use of indigestible tracers. A known weight of the tracer is fed to each animal daily;

from the concentration of tracer in a sample of feces the weight of feces

which would contain the total weight of tracer is calculated as the total

output of feces. This requires (a) that the tracer is completely recovered

in the feces and (b) that the sample of feces analyzed is truly representative.

a . The Recovery of Tracers. Many different substances have been

examined as fecal tracers, but the most commonly used is still chromic

oxide, Cr203,which is cheap and nontoxic. Administration is easy with

the use of factory-produced capsules (Commonwealth Agricultural

Bureaux, I96 I , p. 93), as is also the analysis of the oxide in feces (Commonwealth Agricultural Bureaux, 196 I , p. 152). Unfortunately the fecal

recovery of this tracer has so often been found to diverge from the

theoretical 100 percent that the assumption of complete recovery, without confirmation under the conditions of the particular experiment, must

be questioned. Thus Kane et al. ( 1950) and Raymond and Minson ( 1955)

found mean recoveries close to 100 percent, whereas several other

workers, including Christian et al. (1965; 92.8 percent) and McGuire

et al. (1966; 94.2 percent) have found low recoveries. The reasons for

such low recoveries, particularly over the 66-day feeding period studied

by Christian et al. (1965). which must have eliminated “end errors,”

are still uncertain; they may include regurgitation and rejection of cap-



sules, an impurity of soluble chromate in the chromic oxide, and differential loss of fine particles of the oxide during milling of the feces sample

prior to analysis (see, e.g., P. H . Bailey et al., 1957). But they do indicate

that error, in an overestimation of fecal output, can arise unless an independent check on recovery is made within the experiment.

Attention has thus been given to other tracers, including techniques

based on the use of the rare earth elements, which have been shown to be

almost completely indigestible (Bell, 1964). With the use of radioactive

materials, only minute amounts of tracer are needed, and physical, rather

than chemical methods of fecal analysis can be used; furthermore the

rare earths appear to become firmly bound to plant materials and to pass

through the digestive tract with the undigested fiber residues (Bell,

1964) in contrast to chromic oxide, which appears to move with the liquid

digesta phase (Raymond and Minson, 1955). As a result the diurnal

variation of excretion of tracer in the feces seems to be reduced (Section X, A, 1 , b).

Radiocerium oxide has been used by Huston and Ellis (1968), but

dysprosium appears to have particular advantage (Ellis, 1968) as its

concentration in feces can be measured by activation analysis, so avoiding the need to feed radioactive material, with the consequent problem of

waste disposal. The fecal recovery of dysprosium was 99.8 percent, and

this tracer showed only a small diurnal variation in excretion (Ellis,

1968). Radiocerium also gave high recovery and low variation, and variation among animals was found to be less than with chromic oxide or

polyethylene glycol (Huston and Ellis, 1968). Further development of

the use of radioactive tracers is awaited with much interest.

b. The Pattern of Excretion of lndigestible Tracers. Where the tracer

can be thoroughly mixed with all the feed eaten it is fairly uniformly

excreted in the feces; but this is not possible with grazing livestock, and

here the pattern of dosing of the tracer, which is generally given once or

twice daily, is clearly reflected in a markedly nonuniform excretion in

the feces. Kane et al. ( I 952) studied this under indoor feeding conditions,

and proposed that a representative sample of feces could be obtained by

“grab” sampling directly from the animal’s rectum at fixed times each

day. Raymond and Minson ( 1 955) showed, however, that even under indoor conditions marked variations in excretion pattern could occur, depending among other factors on the type of feed and the level of feeding.

Both of these can affect the rate of passage of digesta through the tract,

and so the time between chromic oxide dosing and the peak of excretion

in the feces. These authors concluded that “grab” sampling would thus

be invalid under grazing conditions, where the excretion pattern would



depend on the kind and amount of forage grazed, the very parameters

the technique is intended to measure. They proposed that feces should be

sampled directly from the sward being grazed, and details of sampling

methods have been described (Commonwealth Agricultural Bureaux,

I96 1 , p. 94). To permit the feces from different animals to be sampled

separately, each grazing animal is dosed daily with a capsule containing

polystyrene granules of a particular color, which can readily be identified

in the feces lying on the field (Commonwealth Agricultural Bureaux,

I96 I , p. 95). This technique is suitable only for cattle, as sheep are able

to chew the granules.

Another approach has been to minimize the extent of the diurnal variation in the content of tracer in the feces by feeding it in a form which

passes through the tract with the undigested residues. Corbett et a f .

( I 960) developed a kraft paper impregnated with chromic oxide, with

which a high proportion of the oxide remains associated with the feed

residues, rather than moving with the liquid digesta phase. They showed

a marked reduction in diurnal variation of oxide excretion in the feces

and, as expected, a change in the times of maximum and minimum excretion, compared with the oxide dosed in capsules. Cowlishaw and

Alder (1963) showed a similar improvement in uniformity of excretion

with the chromic oxide paper, but concluded that the greater ease of administration of commercially produced capsules justified the effort involved in sampling feces from the sward. The use of the rare earth elements, noted in the previous section, may well be a promising solution to

this problem.

2 . The Estimation of the Digestibility of Grazed Forage

The earliest applications of Eq. (24) made use of a determination of the

digestibility of forage cut from the sward being grazed. Observations

showed that this could vary markedly from the forage actually being

eaten by the grazing animals. These earlier studies are likely to have

underestimated the amount and digestibility of the forage grazed, because this forage tends to be the more digestible part of the forage on

offer - as indicated by the generally higher fiber content and lower protein content in forage sampled after grazing than before grazing.

However, estimates of the composition (digestibility) of forage grazed,

based on samples taken before and after grazing, are subject to considerable error (D. Brown, 1954). Thus attention has been given to

methods of estimating the digestibility of grazed forage based on chemical analysis of the feces produced. In the ratio techniques, digestibility

is calculated from the relative contents of a naturally occurring “indigesti-



ble” tracer in samples of “forage grazed” and in feces (summarized by

J. T. Reid, 1962). But because of the difficulty of obtaining relevant forage

samples, these methods have been superseded by the fecal index techniques (see Raymond et al., 1954). These were developed from observations by Lancaster ( 1949) from which it appeared that the nitrogen content of feces was correlated with the digestibility of the forage eaten;

thus the nitrogen content of the feces of grazing livestock should allow

an estimate of the digestibility of the forage selectively grazed from a

sward. Fecal contents of chromogen (J. T. Reid et al., I952), nitrogen and

fiber (Raymond et al., 1954), lignin and methoxyl (Richards et al., 1958),

phosphorus (Barrow and Lambourne, 1962), and copper, magnesium,

and silica (McManus et al., 1967) have all been shown to relate to forage

digestibility and these fecal index markers have been widely used in

grazing experiments. In a related technique Owen and lngleton (1963)

found that the intake of forage dry matter could be related directly to a

dissolved fecal fraction (feces extracted for 18 hours with 0.2 N hydrochloric acid at room temperature) thus:


Intake of forage (g.) = 250 17.8 X

weight of dissolved fecal fraction excreted (g.) (25)

Langlands and Corbett (1964) reported a less precise estimate by this

method than by the use of a fecal nitrogen relationship.

In all these studies the precision of each technique was assessed in

terms of the statistical correlation between forage digestibility and fecal

composition, based on indoor feeding experiments with cut forages. The

aim was to develop a relationship based on a wide range of forages, so as

to be generally applicable, yet with low errors of prediction (Raymond

et al., 1954). Arnold and Dudzinski ( 1963) tested several increasingly

complex statistical relationships with only a marginal reduction in prediction errors. McManus et al. ( 1 967) found that a multiple fecal index

relationship based on fecal contents of nitrogen, copper, and silica gave

slightly lower errors than one based only on fecal nitrogen. Hennaux

et al. ( 1967) found similar precision of estimation of forage digestibility

using eithep chromogen ratio or fecal chromogen techniques. L. E. Davis

et al. (1 967), however, found better prediction using fecal nitrogen because of the low recovery of forage chromogen in the feces, particularly

with lucerne: fecal nitrogen reflected differences in digestibility between

Bromus and lucerne which were not picked up by fecal chromogen, indicating that the latter might not be valid in comparisons between forage


It is now evident, however, that other, possibly more serious, errors



may be involved in the practical application of fecal index methods.

First, there is the possibility that “application” errors may arise when

an indoor-based regression is applied in the field, because of the different

levels of forage intake likely in the two situations (Raymond et al., 1956).

Indoor experiments with the same forage fed at different levels of intake

indicated that fecal nitrogen and chromogen contents may change in a

self-compensating way, but that regressions based on fecal contents of

fiber, lignin, and methoxyl may introduce error if forage intake in the

field differs from that indoors. J. B. Hutton and Jury ( I 964) have also

found that the statistical errors of fecal index relationships based on

forages fed under ad libitum conditions are greater than under the restricted intake conditions of most of the earlier digestibility experiments

( e g . Raymond et al., 1954).

Second, it was found that the assumption in the regression analyses

that the errors were randomly distributed was incorrect, and that different populations of feeds showed a biased distribution about the “general” regression line. Thus Greenhalgh and Corbett ( 1960), Minson and

Kemp ( 196 1 ), and Lambourne and Reardon ( 1963) showed considerable

differences in the relationships between forage digestibility and fecal

nitrogen content with the same forage species at different seasons of

the year, or at different levels of nitrogen fertilization, and similar differences occur in relation to fecal chromogen content. As a result the

feces from animals grazing pasture heavily fertilized with nitrogen contain more nitrogen than feces from less heavily fertilized pasture; this

would indicate a higher level of digestibility for the fertilized forage, but

this is not found in practice (Minson and Kemp, 196 I). This indicated the

need for “restricted” regressions, each based on forage cut from the

sward under study, in order to reduce error due to bias. The continuous

digestibility experiment, using fresh forage cut daily from the sward, was

considered appropriate for this purpose (Greenhalgh and Corbett, 1960;

Commonwealth Agricultural Bureaux, 1961, p. 97), but the need for

such associated indoor feeding must limit the situations in which this

technique can be applied.

More serious, however, is the recent evidence which questions the

basic premise of the fecal index technique, that a relationship based on

the feeding of cut forage indoors can be used to predict the digestibility

of that part of the forage that the grazing animal selects in the jield.

This was indicated by the results of Minson and Kemp ( 196 I ) that forages

of the same digestibility, but containing different levels of nitrogen, give

feces of different nitrogen contents. Lambourne and Reardon ( 1 962)

showed that the “stem” and “leaf’ fractions of a sward may produce



quite different digestibi1ity:fecal nitrogen relationships (Fig. 8), and

Pearce et al. (1962) found different relationships for the “top” and

“bottom” fractions of a sward, harvested and fed separately in digesti-







% N in fecal O.M.


FIG. 8. Relationship between percent digestibility of forage organic matter and percent

nitrogen in fecal organic matter (O.M.), for whole-cut forages and for leaf-rich and stemrich fractions separated from these forages. As fecal nitrogen decreases from 3.6 percent

to 2.6 percent, the whole-plant regression indicates a fall in digestibility of herbage grazed

from 8 1 percent to 73 percent ( x to y). The actual fall in digestibility, as the type of feed

selected changes from leafy to stemmy, would be much less, i.e., from x’ toy’, as indicated

by the dashed line. (From Lambourne and Reardon, 1962; Raymond, 1966b.)

bility experiments. The effect of this is illustrated in Fig. 8; use of a regression equation based on whole forage plants will indicate a much

greater effect of selective grazing on the digestibility of forage eaten than

is in fact the case. Confirmation of this conclusion has been provided by

studies with fistulated animals (below). These show that the earlier concept that the digestibility of grazed forage decreases markedly on the

successive days of grazing on a paddock is often erroneous. This must

mean for instance that the fecal index technique cannot validly be used

to study the effects of grazing pressure on forage selection or intake.

As Langlands (1 967b) concludes, the appropriate fecal index relationship must be derived from herbage similar to that being selected by

each particular group of animals-in which case a less indirect method

of estimating forage digestibility must surely be possible. “This must

mean that, even in experiments where ‘local’ regressions have been used,

incorrect estimates may have been made of the effects of selective graz-



ing, whilst there must be even greater doubt of the validity of the results

of the large number of (grazing) experiments which have used ‘general’

regressions” (Raymond, 1966b).

3 . The U s e of Fistulate Techniques

This conclusion has stimulated investigation into other methods of

studying the forage intake of grazing animals, and in particular the use of

fistulated animals for obtaining samples of the herbage actually being

selected, on which botanical and chemical determinations can be made.

The establishment of esophageal fistulas in ruminants was first described by Torell ( 1954), and their earlier experimental use was reviewed by Van Dyne and Torell (1 964). The more general application of

the technique was advanded by the development of the split-plug closure

technique, which reduces the hazards of the initial operation and simplifies the handling of the animals in the field (McManus et al., 1962). The

technique was initially used for species-selection studies (see below),

but many workers have also used it to estimate the nutritive value of

the herbage grazed (e.g., Bedell, 1967; Bredon and Torell, 1967). I t has

been important, then, to examine how far the extrusa sample collected

through the fistula is fully representative of the feed eaten, and whether

the chemical composition of the sample differs from that of the feed.

D. L. Bath et al. ( 1956) and Grimes et al. ( 196.5) have shown reasonable

agreement between the botanical composition of herbage eaten and that

measured on the extrusa sample; but greater differences can occur in

chemical composition, due both to the contamination of the sample with

saliva before it can be collected and to the preparation of the sample for

analysis. Thus Lombard and van Schalkwyk ( 1963) and C. M. Campbell

et af. (1968) found higher ash contents in extrusa samples than in the

herbage eaten, and Hoehne et al. ( 1967) found higher ash, phosphate, and

chloride levels, lower nitrogen and water-soluble carbohydrate contents,

but similar contents of crude fiber and lignin. Gonzalez Gonzalez and

Lambourne ( 1 967) showed that when a hay-saliva mix was dried, the

measured in vitro digestibility of the hay increased by up to 4 percent

units, indicating the importance of removing saliva contamination before drying extrusa samples. Hoehne et al. ( 1 967) suggested that this

should be by squeezing the sample to remove excess saliva before it is

dried; Robards and Wilson (1967) have shown that washing the sample to

remove saliva reduced in vitro digestibility by as much as 7 units. Thus

the method generally adopted is gentle squeezing of the sample in a

muslin bag, but the validity of any procedure used must be checked

against forage samples wherever possible.



A similar problem may arise when forage samples are collected manually via a rumen fistula: this technique has been used by, among others,

Ridley et al. ( 1 963) and Tayler and Deriaz (1963), but it is applicable

only to cattle because of limitation in the diameter of the cannula with

sheep. The observer has to walk alongside the animal to make the collection, whereas with an esophageal fistula the extrusa samples are collected in a bag tied under the animal’s neck. An advantage of sampling

from the rumen is that the sample can be related directly to the particular

sub-area of pasture being grazed; measurement of the in vitro digestibility

of such samples has provided valuable information on the effects of management and grazing pressure on selective grazing, which differs significantly from earlier evidence based on the fecal index technique (Tayler

and Deriaz, 1963). The digestibility of the forage eaten by animals

grazing immature forage changes relatively little as the sward is grazed

down, despite a marked decrease in the nitrogen content in the feces produced. Where uneaten residues remain from a previous grazing, these

form a low-digestibility component which dilutes the high digestibility

of the new growth. If this residue is in a lower horizon than the new

growth, cattle can select forage of high digestibility, but little selection

occurs where the old and new growth are intimately mixed within the

forage on offer, or where high grazing pressure is imposed so that selection is difficult.

There are several limitations in the use of fistulated animals for estimating the digestibility of “forage grazed” and in the prediction of forage

intake from Eq. (24). First, there is some uncertainty as to the validity

of the digestibility estimate, because of the possible effect of saliva contamination, noted above. Second, in vitro digestibility is a parameter of a

forage measured under standard conditions. It is not necessarily the same

as the in vivo digestibility of the forage by the particular animals being

studied, yet it is this measure of digestibility which is required in Eq.

(24). Finally, extrusa samples are collected only at intervals from a proportion of the animals within a flock or herd, and intensive sampling may

be necessary to obtain a representative sample. Langlands (1967b) has

thus suggested that a “local” fecal index regression should be calculated

between the in vitro digestibility of the extrusa samples and the chemical

composition (nitrogen and cellulose) of the feces voided by the fistulated

animals. He showed different relationships for whole forage fed indoors

and for forage selected in the field as would be expected from Fig. 8. The

latter relationship should allow corrections for selection on the basis of

contents of nitrogen and cellulose in feces from the nonfistulated animals

within the herd. A multiple-regression approach to the same problem has

been described by Arnold and Dudzinski ( 1967a).



An alternative method of using extrusa samples has been described by

Cook et al. (1963), the contents of lignin or silica in the extrusa samples,

and in the corresponding feces, being measured, and the digestibility of

the forage eaten estimated by the ratio method. This method may well

have advantage over the direct measurement of in vitro digestibility on

the extrusa sample, although it relies on the assumption of 100% indigestibility of the lignin or silica, or the use of an indigestibility coefficient measured in vivo on herbage cut from the sward under test.

For precise results it is also necessary for fistulated animals to be

adapted to the pasture being examined, before they can be considered to

give a representative sample of the forage grazed. Thus Langlands

( 1967a) found that fistulated sheep newly moved onto a ryegrass-clover

pasture gave extrusa samples containing 3.8% nitrogen, compared with

3 . 3 3 % nitrogen in extrusa samples from sheep already adapted to the

pasture. This indicates that fistulated sheep should not be moved, as

tester animals, between different experimental groups.

A novel approach, outlined by Wilkins ( 1966), is now being examined.

This is based on the premise that there is a maximum extent to which the

cellulose in a given forage can be digested (potential digestibility) which

is controlled by the cellulose structure and degree of lignification. Cellulose digestibility in vivo will be less than this either because the feed

passes through the digestive tract too rapidly for all the cellulose to be

digested, or because the rumen medium is deficient in nutrients needed

for efficient microbial digestion, or because rumen pH is not optimal for

cellulose digestion. As a result the feces still contain cellulose which

can be digested by an in vitro incubation. The level of cellulose digestibility in vivo (CD) is related to the potential cellulose digestibility

(PCD, determined in a 6-day in vitro incubation) as under:

CD =

100 (PCD - FCD)

100 - FCD

where FCD is the potential digestibility of the cellulose in the feces. The

intake of cellulose can then be calculated from the quantity of cellulose

excreted in the feces:

Intake of cellulose =


x fecal production



x % cellulose in feces

Analysis of the potential cellulose digestibility in a sample of forage eaten,

collected via an esophageal fistula, and of the potential cellulose digestibility in the feces, permits calculation of C D from Eq. ( 2 5 ) . Substitution



of this value for C D in Eq. (26) gives an estimate of the intake of cellulose. From the percentage content of cellulose in the forage sample the

intake of forage dry matter or organic matter can be calculated:

Intake offorage= intake of cellulose X


100 - content of cellulose in forage



This may appear to be a complex procedure. But it is based solely on

the determination in feed and feces samples of the content and potential

digestibility of a chemical entity which can be analyzed simply and with

considerable precision (cellulose by the method of Crampton and Maynard, 1938). More important, it appears to compensate for the animal

factors (level of intake, effect of feed supplements, parasite damage)

which, by affecting the in vivo digestibility of a feed, may invalidate both

fecal index techniques and the use of in vitro digestibility determinations

on fistula extrusa samples. It does, however, make the assumption that

the process of digestion in vivo does not render potentially digestible in

the feces cellulose which was not measured as potentially digestible in

the feed. Despite this, it appears to offer the most logically valid approach

to the estimation of grazing intake that has yet been proposed.





The original use of the esophageal fistula by D. L. Bath et al. (1956)

was to measure the botanical composition of grazed forage, and the technique has been widely used for that purpose to replace the earlier swardsampling techniques. This has been more successful with cattle than with

sheep, presumably because sheep masticate feed more completely before

it is swallowed. Ridley et al. ( 1 963) were able to distinguish smoothstalked meadowgrass from cocksfoot and tall fescue in herbage sampled

from the rumen of cattle grazing a mixed sward, whereas Cook et al.

(1958) were not able to distinguish between species grazed by sheep

from cultivated pasture. However with rangeland forages Cook et

al. (1967) were able to study species selection by both cattle and sheep

and to show differences in preference between the two classes of stock.

The technique has also been used successfully by Van Dyne and

Heady (1965) and Leigh and Mulham (1966) on rangeland vegetation

and by Hodge and Doyle ( 1967) on cultivated pasture. The latter workers

were primarily concerned with selection of either ryegrass or clover from

a mixed sward, and this degree of separation was readily achieved in

extrusa samples taken from lambs and yearling sheep. In nutritional



terms selection of different plant fractions may often be as important as

interspecies selection, and Tayler and Deriaz ( 1 963) were able to draw

important conclusions from samples of leaf, stem, and dead and senescent

forage separated from samples collected via a rumen fistula. This separation was essentially quantitative, whereas species separations must

normally be accepted as qualitative, because of the considerable part of

an extrusa sample which cannot be identified by species-but which

could be allocated to a leaf or stem separate.

The same restriction to a qualitative botanical separation also applies

to the microscopic examination of feces samples for residues of epidermis

and cuticle, characteristic of the different species in the plant community

being grazed (e.g., Hercus, 1963; Stewart, 1967).



It is evident from the consideration of the possible errors involved,

discussed in the previous sections, that some reservation must be applied

to many published estimates of the nutrient intake by grazing livestock.

However, where the experimenter has used a “local,” rather than a

“general,” fecal index relationship, the results are likely to be of reasonable validity.

I . The Energy Requirements of Grazing Animals

1. W. McDonald (1968) has emphasized the considerable assumptions

implicit in the extrapolation, to field conditions, of energy standards

based on controlled calorimetric studies. Most of the experiments in

which fecal techniques have been used to estimate grazing intake indicate

that the maintenance energy requirement of grazing livestock is considerably higher than of stall-fed animals (Coop and Hill, 1962; Langlands et al., 1963), particularly when the amount of pasture available is

restricted (Lambourne and Reardon, 1963). Clapperton (1 964) showed

that maintenance requirement is increased by walking activity; Blaxter

( 1 964) calculated that this was unlikely to increase the requirement by

more than 18 percent, and that environmental effects, such as wind, rain,

and air temperature would also be important. Graham ( I 966) has developed equations for calculating the maintenance requirement of grazing

livestock in terms of locomotion, grazing energy, and environmental

stress, and Lambourne and Reardon ( 1963) suggested that heat production may be increased by endocrine stimulation initiated by stress when

the supply of forage is inadequate.

Tài liệu bạn tìm kiếm đã sẵn sàng tải về

X. The Nutritive Value of Grazed Forage

Tải bản đầy đủ ngay(0 tr)