J Comp Physiol B (1987) 157:67-76 Journal of Comparative .=.om~,, Systemic,

and Environ-

Physiology B . . . . " ' Physiology

�9 Springer-Verlag 1987

Digestion and metabolism of high-tannin Eucalyptus foliage by the brushtail possum (Trichosurus vulpecula) (Marsupialia: Phalangeridae)

W.J. Foley and I.D. Hume Department of Biochemistry, Microbiology and Nutrition, University of New England, Armidale, NSW 2351, Australia

Accepted September 1, 1986

Summary. The digestion and metabolism of Euca- lyptus melliodora foliage was studied in captive brushtail possums (Trichosurus vulpecula). The fo- liage was low in nitrogen and silica but high in lignified fibre and phenolics compared with diets consumed by most other herbivores. The high lig- nin content was suggested as the main cause of the low digestibility of E. melliodora cell walls (24%); microscopic observations of plant frag- ments in the caecum and faeces revealed few bacte- ria attached to lignified tissues. The conversion of digestible energy (0.34 MJ-kg -~ d-1) to meta- bolizable energy (0.26 MJ .kg -~ -1) was low compared to most other herbivores, probably be- cause of excretion of metabolites of leaf essential oils and phenolics in the urine. When the inhibitory effect of leaf tannins on fibre digestion was blocked by supplementing the animals with polyethylene glycol (PEG), intake of dry matter, metabolizable energy and digestible fibre increased. These effects were attributed to the reversal by PEG of tannin- microbial enzyme complexes. It was concluded that the gut-filling effect of a bulk of indigestible fibre is a major reason why the brushtail possum does not feed exclusively on Eucalyptus foliage in the wild.

Introduction

Eucalyptus foliage is an apparently abundant per- ennial food resource for several species of arboreal marsupials including the koala (Phascolarctos cin-

Abbreviations. A D F acid-detergent fibre; AL acid-lignin; DE digestible energy; D M dry matter; M E metabolizable energy; N D F neutral-detergent fibre; PEG polyethylene glycol

ereus) and greater glider (Petauroides volans); these species feed almost exclusively on eucalypt foliage (Marples 1973; Martin 1985). The brushtail pos- sum (Trichosurus vuIpecula), however, supplements a variable intake of eucalypt foliage with leaves from other species of trees and shrubs, as well as fruits, herbs and grasses (Kerle 1984). This sug- gests that the brushtail has a limited capacity to detoxify eucalypt allelochemicals (Freeland and Winter 1975) or a limited capacity to digest fibrous plant tissues.

The brushtail possum is perhaps the most wide- spread marsupial in Australia, and one of the few marsupials to have adapted successfully to the presence of European man. On the basis of its well developed caecum and proximal colon but simple stomach it is a hindgut fermenter (Hume 1982). Wellard and Hume (1981) found that brushtail possums were capable of extensive digestion of fibre, but the semi-purified fibre source used was finely ground and of a low lignin content. In con- trast, Eucalyptus foliage fibre is highly lignified (Ullrey et al. 1981; Cork and Pahl 1984). Also, significant quantities of essential oils and tannins are present (Southwell 1978; Fox and Macauley 1977). Both these groups of allelochemicals have the potential to influence the digestion of plant fibre by gut microbes (Oh et al. 1968; Griffiths and Jones 1977) and hence the intake of metaboliz- able energy (Cook et al. 1952).

This paper reports measurements of the diges- tion of several chemical components of Eucalyptus melliodora (yellow box) foliage and the intake: and excretion of energy by brushtail possums. Observa- tions were made of digesta remnants with a scan- ning electron microscope to identify the plant structures resistant to digestion. Also, the effect of tannins on foliage intake and digestion was as- sessed by supplementing animals with polyethylene

68 W.J. Foley and I.D. Hume: Digestion of Eucalyptus foliage by the brushtail possum

glycol, a compound known to reverse tannin-pro- tein complexes (Jones and Mangan 1977).

Materials and methods

Composition and digestibility of E. melliodora foliage

Experiments were conducted in all seasons of the year to mea- sure the intake, digestibility and metabolizable energy content of E. melliodora (yellow box) foliage when fed to brushtail pos- sums. The animals were caught in wire cage traps in eucalypt woodland near Armidale dominated by E. melliodora, E. eali- ginosa, E. viminalis and Angophora floribunda. Three or four mature male brushtails were housed individually in metabolism cages for at least three weeks prior to each experiment. Feed intakes were stable for at least 10 days prior to each 7-d collec- tion period. Details of the housing, light regime, feeding and collection procedures have been given by Foley and Hume (1987). Metabolizable energy intakes were calculated as digest- ible energy intake minus urinary energy output, but no account was taken of any energy lost as methane or heat of fermenta- tion.

Samples of leaves offered to the brushtails were collected and stored at - 10 ~ and later analysed for dry matter, nitro- gen, gross energy and phenolics (total phenolics, leucoantho- cyanidins and relative astringency) and fibre. Some samples were also analysed for crude lipid, total non-structural carbohy- drates and essential oils. Samples of leaves from various age classes on a single E. melliodora tree were collected, immediately frozen in liquid nitrogen and analysed as above to examine variation in leaf composition with age.

Microscopy of digesta fragments

Samples of foliage offered and digesta and faecal material were collected from three brushtail possums killed in conjunction with the experiments on essential oil metabolism described by Foley (1984). The samples were fixed in 3% glutaraldehyde in phosphate buffer (0.1 M, pH 7.2) for 24 h and thoroughly washed in the same buffer. Samples were then dehydrated in an aqueous ethanol series and dried by the critical point meth- od. After sputter coating with gold, the specimens were exam- ined in a Jeol SM Scanning Electron Microscope at an acceler- ating voltage of 25 keV.

Effect of tannins on intake and digestion of E. melliodora foliage

Polyethylene glycol 4000 (PEG) preferentially complexes with plant tannins in the presence of proteins (Jones and Mangan 1977); this compound was given to the brushtails in their drink- ing water to reverse the action of tannins in E. melliodora leaf. Preliminary experiments tested the acceptability of replacing the brushtails' drinking water with PEG solutions; all animals drank solutions of up to 15% PEG (w/v). Initially, mannitol solutions of comparable osmolarity to the PEG solutions were tested as controls but these were either refused or led to diahor- rea. Subsequently therefore, control animals were offered only tap water.

In Experiment 1, three brushtail possums were supple- mented with PEG at 100% of expected condensed tannin in- take. The concentration of PEG was varied so that differences among animals in their water intakes resulted in similar intakes of PEG. Animals were offered E. melliodora foliage ad libitum together with the PEG solution for 21 d. The first week was an adaptation period and collections of faeces and urine were made during the second and third weeks.

In Experiment 2, four brushtail possums were allocated be- tween two treatments. The two test animals were supplemented with PEG at a level of 100% of estimated condensed tannin intake. The other two animals received tap water only. After a 7-d adaptation period, faeces and urine were collected for a further 7 days. At the end of this collection period, the treat- ments were reversed and collections made for the last 7 days of a further two-week period.

In calculating apparent digestibilities of dry matter allow- ance was made for the PEG excreted in the faeces.

Analytical

The dry matter content of feed, feed residues, and faeces sam- ples was determined by drying a portion at 105 ~ to constant weight. A second sample of feed and feed residues was freeze- dried and a second sample of faeces was oven-dried at 55 ~ for chemical analysis. All samples were ground to pass a 1 mm s c r e e n .

The ash content of leaf samples was determined by burning samples in a muffle furnace at 500 ~ for 3 h. Total nitrogen was determined by the Kjeldahl technique of Ivan et al. (1974) and gross energy using a Gallenkamp adiabatic bomb calorime- ter standardized with benzoic acid. The crude lipid content of leaves was measured gravimetrically after extraction of lipid in 2:1 (v/v) chloroform:methanol (Folch et al. 1957). Total non-structural carbohydrates were extracted from dried, ground leaf with 0.2 N H2SO4 (Smith et al. 1964) and an ali- quot of the extract was analysed for glucose following Luchs- inger and Cornesky (1962).

Essential oils were extracted from frozen leaves by steam distillation (Hughes 1970) in an all-glass apparatus. The content of phenolic compounds in leaves was determined after extrac- tion of dried, ground tissue with hot 50% aqueous methanol (Swain 1979). The total phenolic content of this extract was determined with Folin-Ciocalteau reagent (Folin and Ciocaltau 1927). Leucoanthocyanidins were measured by heating a second portion of the methanolic extract in 5% butanolic HC1 for 2 h (Dement and Mooney 1974). The capacity of this extract to precipitate a protein (haemoglobin) was determined by a modification (Jones et al. 1976) of the method of Bate-Smith (1973). In all phenolic assays, standard curves were prepared from polyphenolic material isolated from E. melliodora leaves (Dement and Mooney 1974; Fox and Macauley 1977).

Neutral-detergent fibre (total cell walls) (NDF), acid-deter- .gent fibre (ADF) and acid-lignin (AL) in feed, feed residue and faeces samples were determined by the methods of Goering and Van Soest (1970). Cellulose and hemicellulose were taken to be the difference between ADF and AL, and N D F and ADF respectively. Although the inclusion of sodium sulphite in neutral-detergent extractions may remove some hemicellu- lose (Mould and Robbins 1981) it was included here to ensure more complete removal of protein residues.

PEG appeared to interfere with the determination of fibre by detergent methods. Addition of PEG to neutral-detergent extractions of faecal samples had no effect on the content of N D F but addition of PEG to acid-detergent extractions led to higher recoveries of ADF (Foley 1984). Similarly, complexes between PEG and euealypt tannins were soluble in neutral- detergent but insoluble in acid-detergent solutions. Because of this interference, only digestibilities of N D F are reported for the PEG supplementation experiments.

Statistical

Analysis of variance was used to test the significance of differ- ences between means of more than two experimental periods. Those means which bear none of the superscripts a--c in corn-

W.J. Foley and I.D. Hume : Digestion of Eucalyptus foliage by the brushtail possum 69

Table 1. Composition of E. melliodora foliage eaten by brushtail possums during different feeding experiments. Values except moisture and energy as %DM

Experiment

A B C D E F G (Winter '81) (Spring '81) (Summer '82) (Summer '82) (Summer '82) (Autumn '82) (Autumn '82)

Moisture(% wet weight)49.0 +0.1 50.1 -+0.2 50.1 -+0.1 49.8 -+0.3 50.2_+0.3 50.9 +0.6 51.7 -+0.4 Organic matter 96.2 -+0.2 96.1 -+0 .3 . . . . 97.0 -+0.2 Total nitrogen 1.58-+0.04 1.67_+0.2 1.62-+0.03 1.65-+0.04 1.67-+0.4 1.78-+0.02 t.43-+0.01 Total non-structural carbohydrate 8.9 +0.6 - 10.2 +0.3 . . . . Crude lipid 10.9 -+0.4 - 11.0 _+0.2 . . . . Essential oil 0.9 a - . . . . 1.2 a Totalphenols 27.9 _+0.3 29.1 +1.2 27.9 _+1.7 27.0 27.1 24.8 -+0.5 19.9 _4:0,9

LA* 22.2 -+0.7 23.6 -+1.5 20.0 -+1.8 19.0 20.0 17.6 -+0.3 16.8 -+0.4 RA* 4.0 _+0.7 3.5 -+0.2 2.7 _+0.8 2.1 3.0 3.1 -+0.9 3.0 4:0.9

Cell wall constituents:

NDF 29.6 _+0.4 26.6 _+0.6 27.7 -+0.4 28.2 28.8 30.3 _+0.1 29.5 -+0.3 Hemicellulose 5.3 +0.6 4.8 -+0.4 4.2 -+0.5 4.8 5.8 4.5 -+0.4 6.2 -+0,4

ADF 24.3 -+0.8 21.8 +0.5 23.5 -+0.2 23.4 23.0 25.9 -+0.3 23.3 -+0.2 Cellulose 14.2 -+0.6 12.8 +0.3 14.7 +0.2 14.7 13.0 16.1 -+0.5 11.1 -+0.3 Lignin 10.1 __+0.2 9.0 -+0.7 8.7 -+0.2 8.7 10.0 9.8 -+0.3 12.2 -+0.3 Residual ash 0.1 -+0.0 0.1 _+0.0 0.1 -+0.0 0.2 0.1 0.1 -+0.0 0.4 -+0.0

Gross energy(kJ.g -1) 20.6 -+0.1 20.6 -+0.2 20.9 -+0.1 21.5 21.1 21.0 -+0.0 21.1 -+0.2

LA: Leucoanthocyanidins; RA : Relative astringency; a Single bulked sample

mon are significantly different (P<0.05). Comparisons between two means were made using Student's t-test (Snedecor and Cochran 1967).

In the PEG experiments, when intakes were significantly different between treatments, analysis of covariance (AOCV: Snedecor and Cochran 1967) was used to adjust dependent means (e.g. digestibility). The low numbers of observations of each treatment in Experiment 2 reduced the sensitivity of tests for differences between treatments. Consequently, comparisons of AOCV-adjusted means were made between all PEG treat- ments and all other experiments in which no PEG was offered. This increased the power of the tests by increasing the degrees of freedom but it also increased between-experiment error. Also, in some cases, the slopes of the relationships between adjusted dependent and independent variables were not equal between the two treatments.

Results

Diet composition (Tables 1 and 2)

The ana lyses p r e s e n t e d in T a b l e 1 a c c o u n t e d for 9 0 - 9 1 % o f to ta l l ea f d ry m a t t e r ; n o d i s t inc t sea- sona l t r ends were a p p a r e n t . T o t a l pheno l i c s were the m a j o r so lub le c o m p o n e n t a n d l ign in c o m p r i s e d 3 0 - 4 0 % of the cell wal l f rac t ion . W i t h i n a s ingle i n d i v i d u a l tree (Tab le 2), o ld leaves t e n d e d to have lower c o n t e n t s o f m o i s t u r e , n i t r o g e n a n d n o n - s t r u c t u r a l c a r b o h y d r a t e s t h a n shoots , b u t h ighe r c o n t e n t s o f cell walls , a n d in pa r t i cu l a r , l ignin .

Table 2. Variation in leaf composition with age within a single E. melliodora tree. Values except moisture and energy as %DM

Constituent Shoots Mature leaves Old leaves

Moisture (% wet matter) 62.0 50.7 46.9 Organic matter 98.2 97.1 97.1 Total nitrogen 1.63 1.38 1.27 Total non-structural carbohydrates 11.6 10.8 9.2 Crude lipid 8.3 9.2 10.1 Essential oil 1.I 0.8 -

Cell wall constituents:

NDF 17.7 25.5 29.1 Hemicellulose 4.1 1.5 1.7

ADF 13.6 24.0 27.4 Cellulose 7.6 12.9 12.7 Lignin 6.0 11.1 14.7 Residual ash 0.1 0.1 0.1

Gross energy (kJ-g -1) 19.9 20.1 20.3

Digestibility of E. melliodora foliage (Tables 3 and 4)

There were n o s ign i f i can t d i f ferences b e t w e e n s u m - m e r a n d w i n t e r in the i n t a k e or d iges t ib i l i ty o f d ry m a t t e r ( D M ) or to ta l cell c o n t e n t s (Tab le 3). The b rush t a i l s c o n s u m e d 71 g D M per da y or 36 g- k g - ~ the m e a n a p p a r e n t d iges t ib i l i ty o f

70 W.J. Foley and I.D. Hume: Digestion of Eucalyptus foliage by the brushtail possum

Table 3. Intake and digestion of E, melliodora dry matter and cell contents by brushtail possums. Values are means • SE

Experiment Body mass (kg) Dry matter intake (g" kg - o,75. d - 1) Apparent digestibility (%)

Dry matter Cell contents

A 2.38 _+ 0.10 39.7 • ].1 b 53.9-+ 2.4" 61.0 +_ 1.6 a B 2.33 _+ 0.16 37.2 • 0.8 ab 46.3___ 1.8 b 57.8_+ 1.1" C 2.59 _+ 0.25 33.3 ___ 2.7 a 50.1 • 2.6 a 60.3 +_ 3.3 a D 2.55 _ 0.27 30.8 -+_ 2.8 a 49.3 +_ 2.0 ab 60.7 +_ 2.7 a E 2.59+_0.25 32.1 -+ 1.2 a 50.3 • 1.1 a 61.6+_ 1.5 ~ Overall mean 2.49 _+ 0.12 35.5 +_ 1.1 50.9 -+ 1.0 60.5 -+ 0.9

Table 4. Intake and digestibility of components of E. melliodora cell walls by brushtail possums. Values are means_+ SE of 23 observations

Table 5. Intake of gross, digestible and metabolizable energy by brushtail possums fed E. melliodora foliage. Values are means +- SE

Constituent Intake Digestibility Experiment Intake (MJ-kg-~ 1) (g .kg-~ -1) (%)

Gross energy Digestible Metabolizable N D F (total cell walls) 10.1 • 0.4 26.5 ___ 2.4 energy energy

Hemicellulose 1.8 • 0.1 41.9 +- 4.0 Acid-detergent fibre 8.3 4- 0.3 22.8 _ 2.4 A 0.82 -[- 0 .02 b 0.40 -[- 0.028 0.30 • 0.02 b

Cellulose 5.0___0.1 31.5+_3.1 B 0.77+0.02 "b 0.31+0.01" 0.24___0.014 Acid-lignin 3.4+_0.2 9.4_+3.2 C 0.70+0.05 ~ 0.31+_0.01" 0.25• ~b

D 0.65_+0.06" 0.29_0.03 a 0.22___0.03" E 0.68_+0.03" 0.31_+0.01" 0.24_+0.01 ~ Overall mean 0.74_+ 0.02 0.34-+ 0.02 0.26-+ 0.01

this DM was 51%. There was no significant rela- tionship between DM digestibility and DM intake.

No seasonal trends were apparent in the intake or digestibility of cell-wall constituents (Table 4). The mean digestibility of N D F (total cell walls) was 27% (range 9-45%). The mean lignin digesti- bility was 9%. In two experiments, negative lignin digestibilities were recorded which suggested that lignin artefacts were present in the faeces.

Intake and excretion of energy (Table 5)

The digestibility of gross energy was 46% and the intake of digestible energy (DE) was 0.34 MJ. kg-~ -1. Retention of DE was 76%, so that intake of metabolizable energy (ME) was 0.26 MJ. kg-0.YS.d-1

Microscopy of digested foliage

Fragments of partly digested E. melliodora leaf taken from the brushtail stomach (Fig. 1 a) re- vealed some disruption of leaf structures but no micro-organisms were attached to any tissues. In contrast, fragments from the caecum and proximal colon (Fig. I b, c) showed extensive bacterial colo- nization of the cell walls with the aid of numerous thread-like structures (Fig. I d). Relatively few bac- teria were attached to more resistant structures such as vessel elements (Fig. 1 e), cuticle and epi-

dermis (Fig. 10, and these fragments appeared in- tact in the faeces.

Effect of PEG on intake and digestion of E. melliodora foliage

The intake and digestibility of dry matter and cell walls and the apparent digestibility of cell contents in brushtail possums supplemented with PEG are shown in Table 6. In Experiment 1, there were no significant differences in intake, excretion or diges- tibility of dry matter, cell contents or cell-wall con- stituents between collection periods.

In Experiment 2, there was a higher ( P < 0.05) intake of DM and N D F in those animals offered PEG solutions instead of normal drinking water. However, at similar levels of intake, there were no significant differences in DM digestibility be- tween treatments. There were trends (P < 0.11) to- ward higher digestibilities and intakes of digestible N D F in the PEG-supplemented animals. However, the digestibility of cell contents was lower ( P < 0.05) in the PEG-supplemented group than in the controls. The DE intake of the animals receiving PEG was higher (P<0.01) than in the control group.

In view of the trend towards higher N D F diges- tibilities in the PEG-supplemented group in Exper-

W.J. Foley and I.D. Hume: Digestion of Eucalyptus foliage by the brushtail possum 71

Fig. 1 a-f. Scanning electron micrographs of digesta fragments from the gut of brushtail possums fed E. melliodora foliage, a stomach contents showing epidermis and mesophyll intact (x 320). b, e caecal contents showing mesophyll cell walls with a high density of colonising bacteria including rods and cocci ( x 4,000). d caecal contents showing bacteria attached to cell wall by threadlike extracellular material ( x 9,400). e caecal contents showing leaf epidermis with minimal attaching bacteria, f faeces fragment showing mineral crystal (probably calcium oxalate) between undegraded spiral xylem elements ( x 4,000)

72 W.J. Foley and I.D. Hume: Digestion of Eucalyptus foliage by the brushtail possum

Table 6. Intake and digestibility of dry matter, cell walls, cell contents and energy of E. melliodora by brushtail possums supplemented with PEG. Values are means • SE

Experiment Treatment Intake (g.kg -~ -d -x) Apparent digestibility (%) Energy intake (MJ.kg -~ .d -~)

Dry matter Cell walls Dry matter Cell walls Cell contents Gross Digestible Metabolizable

F I(+PEG) 44.9+_1.9 13.6+_0.6 51.9_+2.0 46.9_+2.4 54.0_+2.3 2(+PEG) 44.5_+3.3 13.5• 54.1___1.5 53.3+1.4 54.5•

G 3(+PEG) 52.3_+2.6 15.4• 44.5• 43.9• 44.1• 4 ( -PEG) 43.0• 12.6-t-0.7 46.0• 29.7+3.7 52.8•

0.94_+0.04 0.52+_0.03 0.43_+0.03 0.94_+0.07 0.56+0.05 0.48•

1.10• 0.54-+0.02 0.44• 0.91+_0.06 0.38+_0.04 0.28_+0.06

Overall (Treats 1-3) 47.8_+1.9 14.3+_0.5 49.6• 48.1_+2.7 50.2• 1.00_+0.04 0.54+0.02 0.45+0.02 mean

Table 7. Means of all feeding experiments with brushtail possums supplemented with or lacking PEG. Intake of chemical constituents in g.kg -~ .d -1, energy as MJ.kg -~ . d -< Values are mean_+ SE

Parameter + PEG - PEG Significance of difference between means at similar levels of intake

Dry matter intake 48.7 • 1.9 36.8 • 1.2 - Dry matter

digestibility (%) n.s. Cell wall intake Cell wall digestibility (%) P< 0.01 Cell contents intake Cell contents

digestibility (%) P < 0.01 Non-dietary faecal

nitrogen n.s. Gross energy intake Digestible energy P < 0.001 Metabolizable energy P < 0.001

49.6 ___ 1.6 50.0 +0.6 14.3 +_0.5 10.6 _+0.4 48.1 • 27.1 _+2.1 33.4 _+1.4 26.3 +0.9

50.2 • 59.1 •

0.31 -t-0.01 0.25___0.01 1.00 • 0.04 0.77 _ 0.07 0.54 ___ 0.02 0.35 • 0.01 0.45 +_ 0.02 0.27 • 0.01

iment 2, data from all experiments in which PEG was offered were pooled and compared with all other experiments. Table 7 summarises the means of the intake and digestibility of DM, NDF and cell contents.

There were no significant differences in DM digestibility between treatments, but both NDF di- gestibility (P<0.01) and the intake of digestible NDF (P<0.001) were higher in the PEG-supple- mented animals. The digestibility of cell contents was lower (P < 0.05) in the PEG group as a result of a higher (P<0.05) faecal excretion, but the in- take of digestible cell contents was still higher (P < 0.05) in the PEG-supplemented animals. The in- take of both DE and ME was higher (P<0.001) in the PEG-supplemented animals, largely as a re- sult of their higher intakes of gross energy.

Discussion

Analyses of the chemical composition of E. mellio- dora foliage accounted for 91-92% of the dry mat- ter. Much of the remaining fraction probably con-

sists of pectins, water-soluble carbohydrates asso- ciated with the plant cell wall. Although no mea- surements have been made of the pectin content of leaves of any eucalypt species, Mould and Rob- bins (1981) found pectin contents of 6-11% in leaves of a range of North American trees. On the other hand there may be some overlap between measured components. In particular, the crude lip- id fraction will include monoterpenes from the leaf essential oil fraction.

In common with leaves from other eucalypts, E. melliodora was low in nitrogen and silica (resid- ual ash) compared with other foliage (Ullrey et al. 1981; Cork and Pahl 1984). The content of cell walls was at the low end of the range of values found in mature tree foliage, but this fibre was highly lignified. Although E. melliodora (and other eucalypts (Cork and Pahl 1984)) contains high lev- els of crude lipid, the inclusion of essential oils in this fraction reduces its potential value to the animal (Cook et al. 1952; Foley 1984). The pheno- lic fraction contained mainly condensed tannins, but Hillis (1966) found other phenolic monomers

W.J. Foley and I.D. Hume : Digestion of Eucalyptus foliage by the brushtail possum 73

and hydrolysable tannins in a chromatographic ex- amination of E. melliodora.

The E. melliodora foliage eaten by the brushtail possums varied remarkably little in composition between seasons of the year (Table 1), despite the higher nitrogen content and lower fibre content of shoots compared with mature and old leaves (Table 2). This is because of the preference of brushtails for mature foliage, an observation that accords with that of Freeland and Winter (1975) on free-living brushtails in an open woodland habi- tat. The uniformity in the composition of the fo- liage eaten throughout the year is reflected in the relative uniformity of the values recorded for the intake and digestibility of dry matter, cell contents (Table 3) and cell walls (Table 4) of E. melliodora, and the intakes of gross, digestible and metaboliz- able energy (Table 5) between seasons.

Digestibility of the dry matter and cell walls of E. melliodora by brushtails was low compared with grass and forage diets by both foregut and hindgut fermenters. However, fibre digestibility was similar to that found in comparable-sized her- bivores fed browse. For example, mountain hares (Lepus timidus) digested 25-30% of the ADF of willow or blueberry twigs (Pehrson 1983a). In general, the digestibility of browse diets by rumi- nants is markedly lower than that of herbage of similar fibre content (Wilson 1977; Robbins et al. 1975; Mould and Robbins 1982). This seems to be associated with the higher lignin content of browse; lignin is the major cell wall component limiting digestibility (Van Soest 1982). Similarly, the relatively low lignin content of the semi-puri- fied diets used in Wellard and Hume's (1981) study of brushtails probably explains why fibre digestibi- lity was markedly higher than on the foliage diets used here.

A range of bacterial types attached to meso- phyll cell walls in the caecum confirmed that the hindgut was the principal site of microbial diges- tion of fibre in the brushtail possum. In contrast, few bacteria were attached to the more highly ligni- fled structures such as xylem vessel elements. These observations support the idea that the intractabi- lity of lignified tissues was one of the principal reasons for the low digestibility of E. melliodora cell walls.

Digestibility of fibre was similar to values ob- tained in koalas fed E. punctata foliage (Cork et al. 1983) but lower than in ringtail possums (Pseudo- cheirus peregrinus) fed E. andrewsii foliage (Chil- cott and Hume 1984). The high digestibilities found in the ringtail may be due in part to its practice of caecotrophy (coprophagy) (Chilcott

and Hume 1985). Although no measurements of fibre digestibility were made in non-caecotrophic ringtails, Pehrson (1983b) showed that prevention of caecotrophy in mountain hares led to a signifi- cant reduction in digestibility of the ADF of' blue- berry shoots.

The brushtail possums maintained body mass on E. melliodora foliage, suggesting that energy in- takes were close to maintenance. ME intakes were 1.5 times the standard metabolic rate (SMR) of adult brushtails (Dawson and Hulbert 1970). ME intakes of adult ringtail possums fed E. andrewsii foliage under similar conditions were also 1.5 times SMR (Chilcott and Hume 1984). In both possum species ME as a proportion of DE is low compared with other herbivores fed forage diets. This is large- ly due to the energy content of dietary essential oils, phenolics and their metabolites excreted in the urine.

As well as effects on urinary energy loss, allelo- chemicals in eucalypt foliage may also affect mi- crobial activity and hence fibre digestion. Essential oils similar to those found in E. melliodora (Foley . 1984) have been shown to inhibit microbial activity in ruminants (Oh et al. 1967, 1968). However, in the brushtail possum the majority of ingested oils is absorbed before reaching the hindgut. Oh et al. (1967) found that microbial activity in the rumen of deer was inhibited at terpene concentrations 20 times that found in the brushtail hindgut; it is thus unlikely that essential oils had a significant effect on microbial activity in the brushtail possums (Fo- ley 1984).

In contrast, results from the experiments with PEG suggested that E. melliodora tannins de- pressed digestibility of plant cell walls. The higher intake of digestible cell walls with PEG supplemen- tation may be due to the removal of the inhibitory effects of tannins on microbial cellulases. A similar explanation was advanced by Barry and Duncan (1984) to explain the low digestibility of the cellu- lose and hemicellulose of high-tannin varieties of Lotus pedunculatus by sheep; as in the present study, supplementation with PEG resulted in in- creased intakes of organic matter and increased digestibilities of fibre.

The inhibition of cellulases by tannins in vitro is well documented (e.g. Mandels and Reese 1'965; Lyford et al. 1967; Griffiths and Jones 1977). However, by the time tannins reach the hindgut they may already have been complexed with di- etary proteins, or with proteins from the gut wall including enzymes such as proteases, and so would not be free to complex with microbial cellulases in the caecum and colon. There are at least tlhree

74 W.J. Foley and I.D. Hume: Digestion of Eucalyptus foliage by the brushtail possum

possible explanations for tannins binding to cellu- lases in the hindgut of the brushtail possum. First, the tannins responsible for the inhibition could be specific for cellulases (Hagerman and Butler 1981). Second, condensed tannins could occur bound to the cellulose and pectins of the cell wall (Zucker 1983); this would help to ensure a high specificity for microbial enzymes. Third, tannins could be re- leased from protein-tannin complexes either by bacterial degradation (O'Brien et al. 1986) or by a change in pH (Jones and Mangan 1977) in the hindgut.

The higher digestibility of cell walls on the PEG treatment, together with the higher intake of DM, led to higher intakes of DE and ME. About half the increase in DE intake was as digestible cell walls. Since there was no change in urinary energy loss with PEG supplementation the additional en- ergy absorbed was highly metabolizable and mostly retained; some may have been lost as meth- ane and heat of fermentation, but these sources of energy loss were not measured here.

In contrast to the increase in digestibility of cell walls in response to PEG supplementation, di- gestibility of cell contents declined (Table 7). This can be explained by a combination of an increased microbial biomass in the hindgut consequent upon an enhanced supply of energy from the fermenta- tion of cell walls, and the lack of a mechanism in the brushtail's hindgut to selectively retain fluid and fine particles (including bacteria) in the cae- cum (Foley and Hume 1987). The result was an increased loss of microbial matter in the faeces. This is reflected in a trend for non-dietary faecal nitrogen (which in herbivores is composed mainly of microbial nitrogen (Mason 1969)) to increase (by 24%) on the PEG treatments (Table 7).

The net effect of the decrease in digestibility of cell contents and the concomitant increase in digestibility of cell walls (Table 7) was a similar apparent digestibility of dry matter in the presence or absence of PEG in the drinking water. However, the net effect on intake of dry matter (an increase of 32%) was probably dominated by the increase in cell wall digestibility, with any change in cell contents digestibility expected to have a negligible effect.

The increase in DM intake by the brushtails as a result of PEG supplementation brought their intakes up to a level similar to those found in greater gliders, koalas, and ringtail possums (Foley 1984; Cork et al. 1983; Chilcott and Hume 1984). It was suggested by Foley and Hume (1987) that one reason why the brushtail possum feeds to only a limited extent on Eucalyptus foliage, with its

highly lignified tissues, is that it lacks a mechanism in the hindgut to selectively retain fluid and fine particles, including bacteria, in the caecum. The effect of this mechanism is to concentrate digestive effort on the potentially more fermentable frac- tions of the digesta. In the brushtail possum the mean retention time of both particulate and fluid digesta is close to 50 h (Foley and Hume 1987). In contrast, in the greater glider, ringtail possum and koala, coarse particles are eliminated more rapidly than fine particles and fluid (Hume et al. 1984). For instance, the mean retention time of coarse particles in the ringtail possum is 35 h, but that of fluid is 63 h (Chilcott and Hume 1985).

Degradation of tannin-microbial enzyme com- plexes in the hindgut is probably highly dependent on the long digesta retention times found in arbor- eal marsupials. In O'Brien's et al. (1986) study with ringtail possums it was only after the cell wall was degraded that the caecal bacteria were able to at- tack the tanned cytoplasm elements. Hence, mate- rial would have to be retained in the caecum long enough for cell wall digestion to occur and then for a further period for these complexes to be bro- ken down. In our captive brushtail possums once the effects of tannins were countered with PEG, digestion of fibre in the hindgut was more extensive and the limitations to eucalypt foliage intake by the brushtail imposed by its lack of a selective re- tention mechanism in the hindgut were at least partly overcome. In the wild the brushtail possum copes with the lack of selective retention of fluid and fine particles in the caecum by spending a sig- nificant proportion of total feeding time on the ground eating plants from the herb layer (Freeland and Winter 1975; Kerle 1984), even though this increases the risk of predation. This solution by the brushtail possum to high-tannin Eucalyptus fo- liage is thus a behavioural rather than a physiologi- cal solution.

Before the importance of dietary tannins can be assessed more completely in animals such as the brushtail possum, more information is needed on the structure of eucalypt tannins. The informa- tion needed includes the molecular weights and structures of tannins from various eucalypt species, the factors that affect the binding of various pro- teins, the stability of tannin complexes at various pH's and the fate of tannins degraded in the hind- gut.

Acknowledgements. The authors wish to thank Mrs R. Busby and Mr P. Garlick for technical assistance. The study was funded by grants from National Geographic Society and the Australian Research Grants Scheme. W.J.F. was supported by a University of New England Postgraduate Research Award.

W.J. Foley and I.D. Hume: Digestion of Eucalyptus foliage by the brushtail possum 75

The animals were held under the provisions of Licence A158 from the National Parks and Wildlife Service of N.S.W.

R e f e r e n c e s

Barry TN, Duncan SJ (1984) The role of condensed tannins in the nutritional value of Lotus pedunculatus for sheep. 1. Voluntary intake. Br J Nutr 51:485-491

Bate-Smith EC (1973) Haemanalysis of tannins: the concept of relative astringency. Phytochem 12: 907-912

Chilcott M J, Hurne ID (1984) Digestion of Eucalyptus andrewsii foliage by the common ringtail possum, Pseudocheirus pere- grinus. Aust J Zool 32:605-613

Chilcott M J, Hume 1D (1985) Coprophagy and selective reten- tion of fluid digesta: their role in the nutrition of the com- mon ringtail possum, Pseudocheirus peregrinus. Aust J Zool 33:1-15

Cook CW, Stoddart LA, Harris LE (1952) Determining the digestibility and metabolizable energy of winter range plants by sheep. J Anim Sci 11 : 578-590

Cork S J, Hume ID, Dawson TJ (1983) Digestion and metabo- lism of a natural foliar diet (Eucalyptus punctata) by an arboreal marsupial, the koala (Phascolarctos cinereus). J Comp Physiol 153:181-190

Cork S J, Pahl L (1984) The possible influence of nutritional factors on diet and habitat selection by the ringtail possum. In: Smith AP, Hume ID (eds) Possums and Gliders. Austra- lian Mammal Society, Sydney, pp 269-276

Dawson TJ, Hulbert AJ (1970) Standard metabolism, body temperature, and surface areas of Australian marsupials. Am J Physiol 218:1233-1238

Dement WA, Mooney HA (1974) Seasonal variation in the production of tannins and cyanogenic glycosides in the chapparal shrub Heteromeles arbutifolia. Oecologia 15:65-76

Folch J, Lees N, Sloane-Stanley CH (1957) A simple method for the isolation and purification of total lipids from animal tissues. J Biol Chem 226:49%509

Foley WJ (1984) The utilization of Eucalyptus foliage by the greater glider (Petauroides volans) and the brushtail possum (Trichosurus vulpecula). PhD thesis, University of New Eng- land, Armidale

Foley WJ, Hume ID (1987) Passage of digesta markers in two species of arboreal folivorous marsupials - the greater glider (Petauroides volans) and the brushtail possum (Trichosurus vulpecula). Physiol Zool (in press)

Folin O, Ciocalteau V (1927) On tryosine and tryptophan deter- mination in proteins. J Biol Chem 73 : 627-650

Fox LR, Macauley BJ (1977) Insect grazing on Eucalyptus in response to variation in leaf tannins and nitrogen. Oecologia 29:145M6

Freeland WJ, Winter JW (1975) Evolutionary consequences of eating: Trichosurus vulpecula and the genus Eucalyptus. J Chem Ecol 1:439-455

Goering HK, Van Soest PJ (1970) Forage fiber analyses. Agri- culture Handbook No. 379. US Dept Agriculture

Griffiths DW, Jones DIH (1977) Cellulase inhibition by tannins in the testa of field beans (Vicia faba). J Sci Fd Agric 28 : 983-989

Hagerman AE, Butler LG (1981) The specficity of proantho- cyauidin-protein interactions. J Biol Chem 256:4494~4497

Hillis WE (1966) Variation in polyphenol composition within species of Eucalyptus L'Herit. Phytochem 5 : 541-556

Hughes A (1970) A modified receiver for heavier than water essential oils. Chem Ind (London) 48:1536

Hume ID (1982) Digestive Physiology and Nutrition of Marsu- pials. Cambridge University Press, Cambridge

Hume ID, Foley WJ, Chilcott MJ (1984) Physiological mecha- nisms of foliage digestion in the greater glider and ringtail possum (Marsupialia: Pseudocheiridae). In: SmiLh AP, Hume ID (eds) Possums and gliders. Australian Mammal Society, Sydney, pp 247-251

Ivan M, Clack D J, White GJ (1974) Improved nitrogen distilla- tion apparatus. Lab Pract 23:184--185

Jones WT, Mangan JL (1977) Complexes of the condensed tannins of sainfon (Onobrychis viciifolia Scop.) with fraction 1 leaf protein and with submaxillary mucoprotein and their reversal by polyethylene glycol and pH. J Sci Fd Agric 28 : 126-136

Jones WT, Broadhurst RB, Lyttleton JW (1976) The condensed tannins of pasture legume species. Phytochernistry 15 : 1407-1409

Kerle JA (1984) Variation in the ecology of Trichosurus: its adaptive significance. In: Smith AP, Hume ID (eds) Pos- sums and gliders. Australian Mammal Society, Sydney, pp 115-128

Luchsinger WW, Cornesky RA (1962) Reducing power by the dinitrosalicylic acid method. Anal Biochem 4: 346-347

Lyford S J, Smart WWG, Bell TA (1967) Inhibition of rumen cellulose digestion by extracts of Sericea lespedeza. J Anita Sci 26:632-637

Mandels M, Reese ET (1965) Inhibition of cellulases. Annu Rev Phytopath 3 : 85-102

Marples TG (1973) Studies on the marsupial glider Schoino- bates volans (Kerr). IV. Feeding biology. Aust J Zool 21:213 216

Martin RW (1985) Overbrowsing and decline of a population of the koala, Phascolarctos cinereus, in Victoria. I. Food preference and food tree defoliation. Aust Wildl Res 12:355-366

Mason VC (1969) Some observations on the distribution and origin of nitrogen in sheep faeces. J Agric Sci Camb 73 : 99-111

Mould ED, Robbins CT (1981) Evaluation of detergent analy- sis in estimating nutritional value of browse. J Wildl Mgmt 45 : 937-947

Mould ED, Robbins CT (1982) Digestive capabilities in elk compared to white-tailed deer. J Wildl Mgmt 46:22-29

O'Brien TP, Lomdahl A, Sanson G (1986) Preliminary micro- scopic investigations of the digesta derived from foliage of Eucalyptus ovata (Labill.) in the digestive tract of the com- mon ringtail possum, Pseudocheirus peregrinus (Marsupia- lia). Aust J Zool 34:157-176

Oh HK, Sakai T, Jones MB, Longhurst WM (1967) Effect of various essential oils isolated from Douglas fir needles upon sheep and deer rumen microbial activity. Appl Micro- biol 15:777-784

Oh HK, Jones MB, Longhurst WM (1968) Comparison of ru- men microbial inhibition resulting from various essential oils isolated from relatively unpalatable plant species. Appl Microbiol 16: 39-44

Pehrson A (1983a) Digestibility and retention of food compo- nents in caged mountain hares Lepus timidus during the winter. Holarctic Ecol 6:395-403

Pehrson A (1983b) Cacecotrophy in caged mountain hares. J Zool Loud 199:563-574

Robbins CT, Van Soest PJ, Mautz WW, Moen AN (1975) Feed analyses and digestion with reference to white-tailed deer. J Wildl Mgmt 39:67-79

Smith D, Paulsen GM, Ragus CA (1964) Extraction of total available carbohydrates from grass and legume tissue. ]Plant Physiol 39 : 96~-962

Snedecor GW, Cochran WG (1967) Statistical methods (6th edn). Iowa State University Press, Ames, Iowa

76 W.J. Foley and I.D. Hume : Digestion of Eucalyptus foliage by the brushtail possum

Southwell IA (1978) Essential oil content of koala food trees. In: Bergin TJ (ed) The koala. Zoological Parks Board of NSW, Sydney, pp 62-74

Swain T (1979) Tannins and lignins. In: Rosenthal GA, Janzen DH (eds) Herbivores - their interaction with secondary plant metabolites. Academic Press, New York, pp 657-682

Ullrey DE, Robinson PT, Whetter PA (1981) Composition of preferred and rejected Eucalyptus browse offered to captive koalas, Phascolarctos cinereus (Marsupialia). Aust J Zool 29: 839-846

Van Soest PJ (1982) Nutritional ecology of the ruminant. O and B Books, Corvallis, Oregon

Wellard GA, Hume ID (t981) Digestion and digesta passage in the brushtail possum, Trichosurus vulpecula (Kerr). Aust J Zool 29:157-166

Wilson AD (1977) The digestibility and voluntary intake of the leaves of trees and shrubs by sheep and goats. Aust J Agric Res 28 : 501 508

Zucker WV (1983) Tannins: does structure determine function? An ecological perspective. Am Nat 121:335-356