Comparative Biochemistry and Physiology Part B 136 (2003) 369–382

Captive roe deer (Capreolus capreolus) select for low amounts of

tannic acid but not quebracho: fluctuation of preferences and
potential benefits
Marcus Claussa,*, Karin Lasonb, Janin Gehrkeb, Matthias Lechner-Dollb, Jorns
¨ Fickelb,
c b
¨
Tilman Grune , W. Jurgen Streich
a
Institute of Animal Physiology, Physiological Chemistry and Animal Nutrition, Veterinaerstr. 13, Munich 80539, Germany
b
Institute of Zoo and Wildlife Research (IZW) Berlin, Berlin, Germany
c
´ Berlin, Germany
Neuroscientific Research Centre, Charite,

Received 22 May 2003; received in revised form 22 July 2003; accepted 25 July 2003

Abstract

Browsing ruminants have been shown to tolerate a certain amount of tannins in their natural diet, and preference trials
with captive roe deer (Capreolus capreolus) have suggested an active selection for a low dose of hydrolysable tannins.
In this study, we investigated the preference patterns for tannic acid, a source of hydrolysable tannins, and quebracho, a
source of condensed tannins, in a series of preference trials with captive roe deer over time, using a pelleted feed that
differed only in the respective tannin content. Additionally, two groups of four hand-raised roe deer fawns were fed
either a control or a 3% tannic-acid containing diet and physiological parameters were compared after 7.5 months. There
were large differences in preference patterns between the individual roe deer groups; quebracho was mostly avoided,
whereas tannic acid was actively included in the diet in differing, low proportions. However, one group consistently
preferred the quebracho diet over both the control or the tannic acid diet. For the tannic acid, the preference pattern
often revealed an initial period of high preference, followed by a stable period of a moderate preference. The fawns on
the tannic acid diet had a lower pellet intake and a higher relative mass gain than the fawns on the control diet;
differences in salivary tannin-binding capacity and in blood antioxidant status were below significance. These results are
the first indications of potential benefits of a low-dose tannin diet, which need further confirmation. The results of the
preference trials demonstrate that the time pattern of tannin intake is not constant, and pose the question about the
validity of short-term preference trials in general.
䊚 2003 Elsevier Inc. All rights reserved.

Keywords: Ruminant; Tannin; Preference trial; Salivary tannin-binding proteins; Mass gain; Food intake; Antioxidant status

1. Introduction tannins (Distel and Provenza, 1991; McArthur et

al., 1993) showed that ruminants generally seem
Ecological (Cooper and Owen-Smith, 1985; to select against tannins in their natural forages.
Cooper et al., 1988) and experimental work on However, Cooper and Owen-Smith (1985) found
a ‘threshold’ of 5% condensed tannin content
*Corresponding author. Tel.: q49-89-2180-2556; fax: q49-
below which they could not detect any influence
89-2180-3208.
E-mail address: clauss@tiph.vetmed.uni-muenchen.de of tannin concentration on feeding preference. For
(M. Clauss). duiker (Cephalophus) species, an avoidance of

1096-4959/03/$ - see front matter 䊚 2003 Elsevier Inc. All rights reserved.
doi:10.1016/S1096-4959(03)00244-6
370 M. Clauss et al. / Comparative Biochemistry and Physiology Part B 136 (2003) 369–382

tannins could not be demonstrated in trials with consisted of lucerne meal (39.8% of original ingre-
natural feedstuffs (Faurie and Perrin, 1993; Muller
¨ dient mass), crushed oats (39.1%), crushed wheat
et al., 1998). Additionally, Tixier et al. (1997) (7.6%), bran (9.9%), soy oil (0.6%), linseed oil
found that free-ranging roe deer (Capreolus (0.6%), a calcium-phosphorus-premix (1.2%) and
capreolus) selected a diet of higher tannin content a vitaminymineral-premix (1.2%). Prior to the
than the average tannin content of all available feeding experiments, this diet was fed for several
forage in the habitat. After a detailed analysis they weeks to the experimental animals. The tannin-
concluded that this could not simply be a side containing diets were produced using the basal
effect of a selection for plants with high soluble diet mix and adding 3 or 6% of original mass of
carbohydrate contents, which also had high tannin tannic acid (Merck, Darmstadt, Germany) or que-
levels. Experimental work by McArthur et al. bracho-sol (Otto Dille, Norderstedt, Germany).
(1993) supported the idea that browsing ruminants
(adapted to dietary tannins) will tolerate a certain 2.2. Preference trials
tannin load if that is the cost of ingesting otherwise
higher quality forage. Finally, Verheyden-Tixier et Three groups of adult captive roe deer and two
al. (1999) and Verheyden-Tixier and Duncan groups of captive hand-raised roe deer fawns were
(2000) performed preference trials with captive used for the preference trials. Animals were kept
roe deer groups with identical pelleted diets that at the field station of the Institute of Zoo and
differed only in their tannin content, which indi- Wildlife Research (IZW) Berlin, except one adult
cated that the animals actively included a certain roe deer group, which was kept at the Zoological
amount of tannins in their daily ration. Garden of Rostock, Germany. All animals had
Historically, the negative effects of tannins on been raised in captivity. Animals from two groups
ruminants have been emphasized (Kumar and died during the time in which the experiments for
Singh, 1984). However, positive effects of tannins this study were performed due to other reasons.
have been investigated more recently and have For details see Table 1. At the field station, animals
been reviewed for wild animals (Clauss, 2003) were kept in meadow enclosures and received, in
and comprise anthelmintic, antibacterial and addition to the basal or experimental pelleted diets,
antioxidative effects as well as a protection of high-quality meadow hay; at the zoo, they received
valuable nutrients from forestomach fermentation; additionally high-quality meadow hay and a vari-
these postive effects could account for a potential ety of fruits and vegetables. The pelleted diet was
selection for low doses of tannins. So far, poten- offered to the adult animals in plastic feed troughs
tially positive effects have not been demonstrated of identical colour. The position of the troughs
in wild ruminant species. was rotated arbitrarily on a daily basis. The amount
We adopted the trial design of Verheyden-Tixier of pellets offered was weighed, and after 24 h the
et al. (1999), using two different tannin sources leftovers were weighed again. The length of the
and increasing the time period of the individual preference trials varied between 12 and 44 days.
preference trials (which had, in their study, lasted In the case of the group ‘Adult2’, a series of
for 5–8 days with one exception of 15 days), as preference trials with tannic acid was performed
we wanted to test whether preference or avoidance to result in a continuous choice period of 76 days.
patterns varied with time. Additionally, we wanted The different preference trials are summarized in
to compare the effect of mid-term tannin feeding Table 2.
on measurable physiological parameters of roe deer
fawns by feeding them either a control or tannin— 2.3. Effect of tannin feeding
containing pelleted diet for 7.5 months.
In order to test the effect of tannin feeding, the
2. Methods two hand-raised fawn groups, with which at dif-
ferent stages preference trials were also performed,
2.1. Basal and experimental diets were allocated to different diet treatments. The
animals received the basal diet in August, 1999
A basic pelleted diet was formulated that met and were subjected to a preference trial (see Table
the nutritional requirements of captive roe deer 2); then, from 04.09.99 on, ‘Fawn1’ received the
and could be used as a sole feed. This diet basal diet only and were regarded as the ‘control
 M. Clauss et al. / Comparative Biochemistry and Physiology Part B 136 (2003) 369–382 371

Table 1
Experimental roe deer groups (castr.scastrated), kept at the Institute of Zoo and Wildlife Research (IZW) or the Zoological Garden
of Rostock

Group no. Males (n) Females (n) Details Location

Adult1 4 (castr.) (until 22.04.99) 0 Hand-raised IZW
3 (castr.) (from 11.02.00 on) 0

Adult2 5 (castr.) (until 13.04.99) 0 Raised by IZW

4 (castr.) (until 29.04.99) 0 mother
2 (castr.) (from 30.04.99 on) 0

Adult3 1 3 Raised by mother Zoo Rostock

Fawn1 1 (castr.) 3 Hand-raised IZW

Fawn2 1 (castr.) 3 Hand-raised IZW

group’, and ‘Fawn2’ received the 3% tannic acid In April 2000 all animals were immobilized for
diet, for a total of 7.5 months. All animals had other veterinary procedures. On this occasion, body
been kept in one grass enclosure, which was masses were recorded, saliva samples were taken
divided when the two groups were defined. The from the oral cavity by cotton swabs, and blood
animals had been trained to step on an electronic samples were taken from the jugular vein. Some
scale while being hand-fed during bottle-feeding. blood and serum were transferred immediately on
Their readiness to step on the scales while an liquid nitrogen for later antioxidant analysis. Blood
investigator was nearby decreased progressively samples were analysed according to standard lab-
after weaning. Thus, masses were gathered sporad- oratory procedures. For the determination of
ically. Food intake was recorded by weighing the reduced glutathione (according to Beuter et al.,
offered amount of pelleted feed and the leftovers 1963) and oxidized glutathione (according to His-
after 24 h in August and September 1999 and in sin and Hilf, 1976) the blood was treated with a
January and April 2000. The amount of food was mixture of metaphosphoric acid, EDTA and NaCl.
always chosen so that there would be leftovers the After centrifugation, aliquots were taken for neu-
next day. Animals were screened for intestinal tralization with disodiumhydrogenphosphate fol-
parasites on a regular basis. lowed by addition of 5,59-dithio-2-nitrobenzoic

Table 2
Summary of performed preference trials: tannins used, trial lengths and pelleted diets used (TAstannic acid; Qsquebracho)

Tannin Group Days 0% 3% TA 6% TA 3% Q 6% Q c.f.

tested no. Figure
TA Adult1 16 x x x – – 6
TA Adult2 23 x x x – – 1
TA Adult2 28 x x – – – 1
TA Adult2 25 x – x – – 1
TA Fawn1 35 x x – – – 2
TA Fawn2 35 x x – – – –
Q Adult1 29 x – – x x 7
Q Adult2 29 x – – x x 4
Q Adult3 24 x – – x x 3
Q Fawn1 44 x – – x – 5
Q Fawn2 20 x – – x – –
TAyQ Adult1 12 x x – x – –
TAyQ Adult2 21 x x – x – –
TAyQ Fawn2 16 – x – x – –
372 M. Clauss et al. / Comparative Biochemistry and Physiology Part B 136 (2003) 369–382

Fig. 1. Daily diet choices of roe deer group ‘Adult2’ during a course of 76 days, during which different combinations of a basal diet
and a 3 and 6% tannic acid diet were offered (on two days measurements had to be discarded for technical reasons).

acid (DTNB). Reduced glutathione (GSH) was prepared prior to the experiment. Fifty microgram
determined after reaction with DTNB in a spectro- per millilitre tannin solution and 2, 10 and 20 mgy
photometer at 412 nm. For the determination of ml salivary protein solution were added to the pre-
oxidized glutathione (GSSG) the autoxidation of incubation mix. Two controls were utilised, one
GSH was stopped by addition of N-ethylmaleimi- without both tannin and salivary protein to give
de. After addition of sodium hydroxide GSSG was the highest possible value, the other without sali-
modified using ophtalaldehyde. GSSG was deter- vary protein to give the lowest possible value.
mined at a spectrofluorimeter (excitation: 350 nmy Samples were analysed on four plates. Higher
emission: 420 nm) using GSSG standards for A405 values reflect higher relative tannin-binding
quantification. Total cellular glutathione (S GSH) capacities of salivary proteins. For comparison, a
was calculated as GSHq2 GSSG. pooled sample of total saliva of seven adult roe
Saliva was separated from the cotton by centrif- deer kept on the same experimental station, which
ugation in salivettes (Sarstedt, Numbrecht,
¨ Ger- had been exposed to tannins in natural browse and
many) at 5500=g for 20 min at 8 8C and desalted apple pomace, were used.
applying centricons YM 3 (Millipore, Eschborn, The pellet intake of ‘Fawn1’ and ‘Fawn2’ was
Germany). Filters had a molecular mass cut-off of compared statistically by fitting a regression line
3 kDa. Protein concentration of the total saliva to each group (independent: number of days after
was determined by the BCA-method of Smith et starting the experiment) and comparing their inter-
al. (1985) which was adapted to microtitre plates. cepts using analysis of covariance (ANCOVA).
Bovine serum albumin (BSA) was used as a The difference in mass gain and blood parameters
standard. For the assay, 10 ml of BSA (0–2.5 mgy between the ‘Fawn1’ and ‘Fawn2’ samples was
ml) or desalted salivary proteins were mixed with tested for significance with an unpaired t-test. The
150 ml of the Smith Standard Working Reagent significance level was generally set to as0.05.
and incubated for 30 min at 37 8C. The absorbance All statistical calculations were performed with the
(A492 nm) was measured using a microtitre plate- SPSS 9.0 (SPSS, Chicago, IL) and STATISTICA
(StatSoft, Tulsa, OK) software packages.
reader (SLT Laborinstrumente, Crailsheim, Ger-
many). Tannin binding capacity was measured 3. Results
with help of the tannin-binding assay as described
3.1. Preference trials
previously (Fickel et al., 1998). The assay was
performed on microtitre plates. Trypsin and tannin The time period for which a preference trial was
stock solutions (1% tannin in 50% methanol) were conducted had distinctive effects on the outcome
 M. Clauss et al. / Comparative Biochemistry and Physiology Part B 136 (2003) 369–382 373

Fig. 2. Daily diet choices of roe deer group ‘Fawn1’ during a course of 35 days, during which a combination of a basal diet and a 3%
tannic acid diet were offered.

of the trial. For example, regarding the choices of increased up to 4.5% tannic acid initially and then
‘Adult2’ with respect to tannic acid (Fig. 1), after decreased gradually, seemingly stabilising at 2.0%
five days this group ingested a pelleted diet with again.
approximately 3.5% tannic acid. During the next The pattern observed in this group—an initial
six days, they chose higher tannin concentrations increase in tannin intake, followed by a decline
of 4–4.5%. In the following 10 days, however, and a stabilisation at a lower level—could be
there was a decrease in tannic acid preference and observed in other groups as well (c.f. Fig. 2). In
a stable choice pattern at a level of 3.0% tannic general, the preference pattern for quebracho
acid seemed to develop. In the ensuing experiment, seemed to follow a more uniform course, consist-
which lasted 28 days, the animals could only ing of an initial peak and a much faster reduction
choose between the basal diet and the 3% tannic of intake to generally lower overall tannin content
acid diet. After an initial drop, tannin intake (c.f. Figs. 3–5). However, even quebracho was
stabilized at approximately 2.0% tannic acid. Dur- never completely avoided. In the trial where they
ing the next experiment, which lasted 25 days, the could only choose between a 3 and a 6% quebra-
animals could only choose between the basal diet cho diet, ‘Fawn2’ had an average quebracho intake
and the 6% tannic acid diet. The preference pattern of 3.5% quebracho.

Fig. 3. Daily diet choices of roe deer group ‘Adult3’ during a Fig. 4. Daily diet choices of roe deer group ‘Adult2’ during a
course of 26 days, during which a combination of a basal diet course of 29 days, during which a combination of a basal diet
and a 3 and 6% quebracho diet were offered. and a 3 and 6% quebracho diet were offered.
374 M. Clauss et al. / Comparative Biochemistry and Physiology Part B 136 (2003) 369–382

Fig. 7. Daily diet choices of roe deer group ‘Adult1’ during a

Fig. 5. Daily diet choices of roe deer group ‘Fawn1’ during a course of 29 days, during which a combination of a basal diet
course of 44 days, during which a combination of a basal diet and a 3 and 6% quebracho diet were offered.
and a 3% quebracho diet were offered.

concentration for both the tannic acid and the

quebracho diet (Figs. 6 and 7).
When quebracho and tannic acid were directly
compared in choice trials, ‘Adult2’ and ‘Fawn2’
preferred tannic acid over quebracho (average
daily intakes: ‘Adult2’ 0.4% tannic acid and 0.1%
quebracho; ‘Fawn2’—only tannin pellets offered—
2.7% tannic acid and 0.3% quebracho). In contrast,
the deviating group ‘Adult1’ did not ingest any
basal or tannic acid pellets, but only the quebracho
pellets.
Obviously, the tannin content in the food taken
by the animals after starting the experiment did
Fig. 6. Daily diet choices of roe deer group ‘Adult1’ during a not randomly vary around a mean (Figs. 1–5).
course of 16 days, during which a combination of a basal diet Thus, the time sequence of the data has to be
and a 3 and 6% tannic acid diet were offered.
taken into account. Calculating regression lines
between the pellet intake and the tannin intake in
One group, ‘Adult1’, deviated from the other- order to estimate a mean tannin percentage (as
wise observed patterns. This group generally con- was done by Verheyden-Tixier and Duncan, 2000)
sumed lower amount of pelleted food, but would ignore this fact. Therefore, after visual
invariably chose to the highest possible tannin inspection of the adaptation curves, we determined

Table 3
Average selected daily tannic acid concentration in pelleted feeds in different roe deer groups. Averages calculated for the whole trial
period and for a period of ‘stable choice’

Roe deer Trial Average Average Average ‘Stable Average Average Average c.f. Figure
group period pellet tannic tannic choice pellet tannic acid tannic acid
(days) intake acid intake acid intake period’ intake intake intake
(gyanimal (gyanimal (% original (days) (gyanimal (gyanimal (% original
and day) and day) weight) and day) and day) weight)
Fawn1 35 386 5.0 1.3 17 558 6.5 1.2 2
Fawn2 35 312 3.9 1.2 not reached – – – not shown
Adult2 21 430 2.6 2.8 9 419 12.6 3.0 1ypart 1
Adult2 28 1138 21.8 1.9 13 1409 30.4 2.2 1ypart 2
Adult2 25 1208 40.1 3.2 6 783 15.2 2.0 1ypart 3
Adult1 16 114 6.8 app. 6 16 114 6.8 app. 6 6
 M. Clauss et al. / Comparative Biochemistry and Physiology Part B 136 (2003) 369–382 375

Table 4
Average selected daily quebracho concentration in pelleted feeds in different roe deer groups. Averages calculated for the whole trial
period and for a period of ‘stable choice’

Roe Trial Average Average Average ‘Stable Average Average Average c.f.
deer period pellet quebracho quebracho choice pellet quebracho quebracho Figure
group (days) intake intake intake period’ intake intake intake
(gyanimal (gyanimal (% original (days) (gyanimal (gyanimal (% original
and day) and day) mass) and day) and day) mass)
Fawn1 44 807 3.8 0.6 30 903 1.5 0.2 5
Fawn2 20 938 5.5 0.6 14 929 3.8 0.4 not shown
Adult2 29 557 6.8 1.2 23 565 7.3 1.3 4
Adult3 26 403 2.3 0.4 12 523 2.4 0.4 3
Adult1 29 534 31.7 app. 6 29 534 31.7 app. 6 7

Table 5
Average selected daily tannin concentration in pelleted feeds in different roe deer groups. Note that ‘Fawns 2’ could only choose
between the two tannin diets and was not offered the control pellets

Roe deer Trial Average Average Average Average Average Average

group period pellet tannic acid quebracho tannic acid quebracho total tannin
(days) intake intake intake intake intake intake (%
(gyanimal (gyanimal (gyanimal (% original (% original original
and day) and day) and day) mass) mass) mass)
Fawn2 16 560 15.2 1.6 2.7 0.3 3.0
Adult2 21 604 2.2 0.8 0.4 0.1 0.5
Adult1 12 375 0 11.2 0 3.0 3.0

for tannic acid, but significant differences (P-

0.001) between any other pair out of ‘Fawn1’,
‘Fawn2’, ‘Adult1’, ‘Adult2’(A). In a second
approach, the percentage tannic acid intake in the
three different periods (A–C) was analysed for
‘Adult2’. Again, post hoc tests not assuming equal
variances were performed. The difference was
significant beween period A and B (P-0.001),
but not between period A and C (Ps0.059) or B
and C (Ps0.089). Applying the same procedure
to the quebracho data, significant differences in
the level of tannin intake were found between all
Fig. 8. Tannin binding capacity (mean, S.D.) of salivary pro- pairs out of ‘Fawn1’, ‘Fawn2’, ‘Adult1’, ‘Adult2’,
teins of different roe deer groups in different analytical ‘Adult3’ (‘Fawn1’–’Fawn2’: Ps0.005; all other
concentrations. pairs: P-0.001) except for the pairs ‘Fawn1’–
’Adult3’ (Ps0.289) and ‘Fawn2’–’Adult3’ (Ps
a cut off-point for each curve describing the 1.0).
beginning of a relatively stable phase (Tables 3–
5) and ensured, by linear regression, that no 3.2. Effects of a tannic acid diet
significant slope occurred in these ‘stable choice
periods’. The percentage tannin intake of the 3.2.1. Salivary tannin-binding proteins
groups was then compared based on the stable The results of the comparison of the tannin
choice periods only. Multiple comparisons not binding capacity of the salivary proteins between
assuming equal variances (Levene test for equal the two roe deer fawn groups and the saliva from
variances: Ps0.024) revealed no difference in the adult roe deer are presented in Fig. 8. Because of
level of tannin intake between the two fawn groups the very small sample size, we did not perform
376 M. Clauss et al. / Comparative Biochemistry and Physiology Part B 136 (2003) 369–382

ANOVA with these data. The salivary proteins of

the juvenile roe deer groups from this study did
not achieve the same tannin binding capacity as
samples from adult roe deer. Differences between
the two juvenile groups were very slight, with
saliva from animals on the tannin diet having
slightly higher tannin-binding capacity.

3.2.2. Blood parameters (Table 6)

The tannin group had significantly lower hemo-
globin (Hb, Ps0.005) concentrations and hae-
matocrits (Hct, P-0.001) than the control group.
Among the serum biochemistry parameters, the
Fig. 9. Daily pellet intake (mean, S.D.) of roe deer fawn groups
tannin group had significantly lower glucose (Glc,
in different months (AugustySeptembersat the beginning,
Ps0.028) concentrations. While there was a gen- Januarysduring, Aprilsafter the experiment). The control
eral trend for tannin animals to have lower mineral group received the basal pelleted diet; the tannin groups the
concentrations, this was significant only for zinc same pelleted diet with the addition of tannic acid (3% original
(Zn, P-0.001). Among the antioxidant parameters mass). Intake was measured in August, September, January and
April on 31, 5, 7 and 12 consecutive days.
measured, there was a trend for tannin animals to
have a higher antioxidant status that was just below
significance (GSH: Ps0.051; S GSH: Ps0.050) control group weighed on average 19.8"0.2 kg,
(Table 6). and the tannin group 22.2"0.8 kg, with animal
T4 (16.1 kg) again excluded. In order to include
3.2.3. Body mass animal T4 in the data evaluation, the mass gains
During the monitoring of the body mass devel- during the tannin period were expressed as the
opment, it was obvious that one animal (T4), percentage of starting body mass, which removes
which underwent a prolonged period of diarrhea the effect of the absolute masses of the animals.
while being hand-raised, had a much lower body The average mass gain of the control group was
mass than the other seven animals. Average body 13.7"6.8% vs. 24.6"2.3% in the tannic acid
mass of the control group at the start of the trial group. The difference was significant (Ps0.023).
was 17.4"1.1 kg and of the tannin group Pellet intake data, ordered according to month, is
17.7"0.9 kg. Animal T4 (13.2 kg) is excluded diplayed in Fig. 9. The difference in pellet intake
from this average. At the end of the trial, the between the groups was analysed for two periods

Table 6
Selected blood parameters measured in the two roe deer fawn groups after a 7.5 month-feeding period (Cscontrol group; Tstannic
acid group). Hbswhole blood hemoglobin, Hctshaematocritypacked cell volume, Glcsserum glucose, APsserum alkaline phospha-
tase, TP stotal serum proteins, T3sserum triiodthyronin, T4sserum thyroxin, Fesserum iron, Znsserum zinc, GSHsreduced
glutathione, GSSGsoxidized glutathione, S GSHstotal cellular glutathione (GSHq2 GSSG)

Hb Hct Glc AP TP T3 T4 Fe Zn GSH GSSG S GSH

gydl % mmolyl Uyl gyl nmolyl nmolyl mgydl mgydl nmolymg Hb nmolymg Hb nmolymg Hb
C1 16.5 47.6 12.60 37 56 2.5 100.4 150 97 4.18 0.086 4.35
C2 – – 14.34 32 55 3.5 126.1 353 92 3.85 0.106 4.06
C3 15.7 47.6 15.57 63 57 3.4 136.4 237 101 3.48 0.080 3.64
C4 15.7 46.8 17.92 68 55 3.0 119.7 404 93 3.98 0.084 4.15
av. 16.0* 47.3* 15.11* 50 56 3.1 120.7 286 96* 3.87 0.089 4.05
S.D. 0.5 0.5 2.24 18 1 0.5 15.2 114 4 0.29 0.012 0.30
T1 14.6 44.4 13.38 58 58 3.2 101.7 221 70 4.19 0.095 4.38
T2 14.5 44.6 9.80 92 59 2.9 106.8 259 81 3.89 0.090 4.07
T3 14.8 44.5 9.63 78 57 3.3 110.7 188 66 4.56 0.106 4.77
T4 14.2 43.7 11.26 70 61 3.0 86.2 237 58 4.77 0.092 4.95
av. 14.5* 44.3* 11.02* 75 59 3.1 101.4 226 69* 4.35 0.096 4.54
SD 0.2 0.4 1.74 14 2 0.2 10.8 30 10 0.39 0.007 0.39
 M. Clauss et al. / Comparative Biochemistry and Physiology Part B 136 (2003) 369–382 377

with many consecutive measurements. After fitting lation, incidentally. In two feeding trials of our
regression lines, the comparison of intercepts study, a ‘stable choice period’ was not reached
revealed a significant lower pellet intake for the (Table 4). In the three experiments in which both
tannin group (days 1–34: P-0.001, days 270– tannin sources were offered simultaneously, no
281: Ps0.001). ‘adaptation period’ was evident (Table 5); these
trials had been always performed after the trials
4. Discussion with the individual tannin substances (c.f. Table
2).
According to Hagerman et al. (1992), results It is difficult to compare these results directly
derived from experimental feeding of tannic acid to those presented by Verheyden-Tixier and Dun-
or quebracho are not necessarily the same as those can (2000). In that study, only a few juvenile
that would be achieved, if natural tannin sources animals were included, which were not evaluated
were used (which are, however, generally not separately from adults. The trials were conducted
available). Even if these two commercially avail- for 5–8 days only (with one exception of 15
able tannin sources are included in feeds in iden- days), and fluctuations in intake patterns were not
tical amounts, as in our trial, differences in the presented. The results were evaluated by means of
amount of biochemically active tannin substances a regression analysis of the daily total amount of
will occur between the two preparations: whereas pellets consumed by all groups vs. the daily
the tannic acid used was of a certified pureness, amount of tannin consumed, which resulted in an
quebracho may contain as little as 50% effective inclusion of the tannin source at a level of 2.8%
tannins (Robbins et al., 1991). As analyses on the of the pelleted food mixture. For our data, such a
substances used in this study could not be per- regression analysis cannot be performed, as the
formed, differences between the two quebracho daily intake data points are clearly not independent
sources must be regarded as mainly qualitative. (c.f. Figs. 1–7), and because there were significant
differences in the preference pattern of the different
4.1. Preference trials animal groups tested. Thus, our results confirm the
conclusion that roe deer do not avoid dietary
The distinctively fluctuating patterns of prefer- tannins (in this case tannic acidyquebracho), but
ence in the preference trials present an important from our data we cannot conclude that there is a
problem for the interpretation of the results. Ide- species-specific level of active tannin selection
ally, only results of trials should be evaluated that that is precisely regulated. Rather, large differences
were performed long enough for a stable pattern between animal groups were observed.
to become evident—be it a constant level of It was demonstrated in roe deer (Tixier et al.,
selection or a consistently repeated preference 1998) and white-tailed deer (Odocoileus virgini-
cycle. For the feeding trials of this study, this anus, Spalinger et al., 1997) that hand-raised fawns
demand could not be considered to be fulfilled by do not differn in their dietary preferences from
all trials (c.f. Fig. 3). The importance of the adult, mother-raised individuals. In this respect,
distinction between an ‘adjustment stage’ of vari- generally lower tannic acid intake of our fawn
able time length and a ‘representative choice stage’ groups should not solely be regarded as an effect
can be demonstrated if the average ingested tannin of their young age, but rather as another manifes-
concentrations of the complete trials or of the tation of individual differences between animal
final, stabilized choice stage, are compared (Tables groups. For free-ranging roe deer and mule and
3 and 4). In some cases, higher initial tannin black-tailed deer (Odocoileus hemionus spp.),
intakes were levelled out in the average calculation large individual differences in forage preferences
due to the length of the feeding trial (e.g. ‘Fawn1’ have been described in free-ranging animals (Vos-
on tannic acid, Fig. 2 or ‘Adult2’ on quebracho, er-Huber and Nievergelt, 1975; Bartmann and Car-
Fig. 4), but in other cases a high initial tannin penter, 1982; Gillingham and Bunnell 1989). In
intake could lead to overestimations if the total preference trials with captive roe deer with diets
average was calculated (e.g. ‘Fawn1’ on quebra- of different protein and starch content and a single
cho, Fig. 5). In ‘Adult3’ on quebracho (Fig. 3), trial period of four days, Rusterholz and Turner
the initial refusal and the subsequent high intake (1978) found large individual differences in pref-
levelled each other out in the total average calcu- erence patterns and changes in preference patterns
378 M. Clauss et al. / Comparative Biochemistry and Physiology Part B 136 (2003) 369–382

of the same individuals if the trials were repeated al., 1994). In experiments with sheep (Krueger et
after a certain time period. Evidently, results of al., 1974), taste was identified as the most impor-
such trials must be interpreted with caution. tant sense for forage selection, with all other senses
Robbins et al. (1991) and Hagerman et al. playing only supportive roles. The use of smell for
(1992) showed that when quebracho was added to food preference decisions will depend on the type
a pelleted diet, protein digestibility decreased in of feeds available for choosing. In this respect, it
mule deer and even more dramatically in sheep. is impossible to decide whether the low proportion
However, the addition of tannic acid did not, in of quebracho in the intake of most animals was a
those trials, lead to a reduction in protein digesti- result of a constant ‘sampling’ of all available
bility. Similar observations regarding tannic acid foods, or a deliberate inclusion of this low propor-
had already been made by Hawkins (1955). Given tion, although the latter interpretation may seem
the different effect of tannic acid and quebracho more unlikely. A constant sampling behaviour has
on digestion, the general pattern of a fast decline
also been described for roe deer by Kossak (1981)
of quebracho preference (c.f. Figs. 4 and 5) in
and Tixier et al. (1998). The tendency for tannic
contrast to a generally much more gradual reduc-
acid to be included in comparatively higher pro-
tion in tannic acid preference, seems understanda-
ble; it is in accord with the aversion learning portions could, however, be interpreted as an
theory of Provenza et al. (1990) that the develop- indication of an active selection.
ment of a digestive malaise leads to a fast aversion In general, the following conclusions regarding
of the according foodstuff. In this respect, the the selection of dietary tannins in the animals
preference behaviour of ‘Adult1’, who consistently investigated in this study can be made: (1) Neither
chose the highest quebracho concentration availa- tannic acid nor quebracho were entirely avoided.
ble to them, is difficult to understand. From the (2) Tannic acid seems to have been included in
time necessary until a relative rejection of the the diet to a higher degree than quebracho. In the
tannic acid pellets occurred, one can assume that direct comparisons, this preference pattern was
the proportions of tannic acid ingested did not confirmed in two of three cases. However, one
cause instantaneous malaise in roe deer. Would adult group preferred quebracho to tannic acid. (3)
that have been the case, one would have expected The fawn groups used in this study generally
a rejection of the tannin-containing pellet mix included tannic acid in lesser proportions than the
within the first few days (Provenza et al., 1990), adult animal groups, however, the intake of que-
as displayed by most groups against the quebracho bracho did not differ significantly between the
diet. However, one can only speculate if the fawns and ‘Adult3’. (4) Significant differences in
observed pattern is the result of a delayed negative the proportion of tannins ingested between the
feedback, or the lack of positive reinforcement of individual animal groups do not allow the postu-
the behavior. A tempting interpretation of the lation of a ‘species pattern’. (5) A pattern of an
observed fluctuation in tannic acid intake is that initially increasing preference of tannin-containing
the astringent, ‘woody’ tannin taste and smell is food with a subsequent decline seemed to occur
habitually used by these captive animals to identify
in several of the trials.
potentially beneficial diet items, i.e. browse, and
The length of time period for which a preference
that it takes a while to un-learn this association;
trial is performed will influence the results. Con-
when the animals realize that the digestive effect
of the ‘tasty’ food equals that of the regular food, ventionally, preference trials with both domestic
the particular preference ceases to peak. This and wild animals are carried out for much shorter
would also explain why the ‘woody’ taste of the periods of time (e.g. Schwartz et al., 1980; Prov-
comparatively harmless tannic acid is finally pre- enza et al., 1990; Duncan et al., 1994; Kirschner
ferred over that of the malaise-producing quebra- et al., 1999; Verheyden-Tixier and Duncan, 2000;
cho. Similar fluctuations of preferences were found Alm et al., 2002; De Rosa et al., 2002). In the
in a similar trial conducted with captive giraffes view of the individual differences, and the fluctu-
and tannic acid (Clauss et al., 2003). ation of preference within individual animals, the
Observations on the use of gustation or olfaction results of short-duration preference trials may have
for the selection of preferred feeds in deer is to be interpreted with caution, and should ideally
controversial (Schwartz et al., 1980; Duncan et be confirmed by long-term trials.
 M. Clauss et al. / Comparative Biochemistry and Physiology Part B 136 (2003) 369–382 379

4.2. Effects of a tannic acid diet zinc in the tannin group could be attributed to a
possible influence of tannic acid on zinc absorp-
4.2.1. Saliva tion. The results on the antioxidant parameters
Fickel et al. (1998) demonstrated that the parot- GSH and S GSH could indicate that the ingestion
id saliva of roe deer contains specific, tannin- of tannic acid stabilizes the antioxidative metabo-
binding proteins—just like the saliva of mule deer, lism. Since the GSSG concentrations remain
the other deer species previously examined (Austin unchanged, no direct effect on the free radical
et al., 1989; Robbins et al., 1991). In mule deer, production can be assumed, whereas higher GSH
tannin-binding proteins could not be modulated by and consequently higher S GSH are a clear hint
the manipulation of dietary tannin concentrations on some improvement in the glutathione synthesis.
(Austin et al., 1989). As mule and roe deer are Positive effects of tannins on the antioxidant status
strict browsers, the presence of salivary tannin- of animals and man have attracted attention recent-
binding proteins could be part of the overall ly (Tebib et al., 1997; Bravo, 1998; Hassig
¨ et al.,
genetic makeup of these species. The difference 1999). The effect of tannin feeding on wild ani-
observed in the salivary tannin-binding capactiy mals in this respect, and the consequences of a
between the pooled sample from adult animals and lack of dietary tannins in captive environments,
both juvenile groups is astonishing. One could try should be further investigated.
to interpret this finding as an indication of a
gradual induction of tannin-binding proteins during 4.2.3. Body mass gain and food intake
ontogeny by constant consumption of tannin-con- The fact that the tannin animals increased their
taining forages; however, the comparison of the body mass sigificantly more than the control group
control and the tannin juvenile group does not indicates that the tannic acid supplement increased
strongly support this argument. Whether differenc- the food conversion efficiency in these animals.
es in tannin-binding capacity exist between ani- At the moment, three possible explanations for the
mals of different age classes would need to be increased food efficiency are available:
verified with a larger sample size.
1. As tannins can reduce the parasite burden (But-
4.2.2. Blood parameters ter et al., 2000; Kabasa et al., 2000; Athanasia-
Although the external conditions as day time or dou et al., 2001), this might result in a lesser
immobilisation stress were identical for all ani- energy consumption by parasites and anti-para-
mals, the small sample size and the lack of a site defence, making more nutrients available
crossover design requires a cautious interpretation for growth. However, all study animals were
of these results. Tannins are known to chelate regularly dewormed, and faecal flotations
dietary iron and thus reduce iron absorption (Gil- revealed no parasites throughout the entire study
looly et al., 1983). The lower hemoglobin concen- period. Therefore, other explanations are more
trations of the tannin animals are probably a direct likely.
result of this effect and thus indicate that the tannic 2. Low levels of tannins can reduce the protein
acid had a physiological effect regardless of the degradation in the forestomach of ruminants and
presence of salivary tannin-binding proteins. For thereby increase the flow of more valuable
iron absorption, serum iron is not considered a amino acids to the small intestine (Driedger and
suitable parameter, and accordingly, this parameter Hatfield, 1972; Barry and Manley, 1984 Wagh-
did not differ between the treatment groups in orn et al., 1987; Liu et al., 1998; Frutos et al.,
spite of the differences in hemoglobin concentra- 2000). Even mass gains have been described
tion. There are conflicting results on effects of (Kaitho et al., 1998). It is also speculated that
tannins on the availability of other minerals (Kies tannins in forages can protect soluble carbohy-
and Umoren, 1989; Jansman et al., 1993; Vaquero drates from rumen fermentation and, therefore,
et al., 1994). The availability of zinc has been increase their nutritive value for ruminants (Bar-
shown to be either decreased Greger and Lyle, ry and Manley, 1984; Sreerangaraju et al.,
1987) or not affected (Flanagan et al., 1985; 2000). These effects have mainly been observed
Jansman et al., 1993) by the presence of tannin- for condensed tannins. Whether the hydrolysa-
containing compounds. The reduction in serum ble compounds of tannic acid could perform a
380 M. Clauss et al. / Comparative Biochemistry and Physiology Part B 136 (2003) 369–382

protein- or soluble carbohydrate-protecting different gastrointestinal nematodes of sheep: in vitro and

effect in the rumen prior to dissolution of the in vivo studies. Vet. Parasitol. 99, 205–219.
Austin, P.J., Suchar, L.A., Robbins, C.T., Hagerman, A.E.,
associated complexes in the lower gastrointes- 1989. Tannin binding proteins in the saliva of deer and their
tinal tract remains to be demonstrated. absence in the saliva of sheep and cattle. J. Chem. Ecol. 15,
3. The tannic acid could have led to a generally 1335–1347.
reduced metabolic rate of the animals, thus Barry, T.N., Manley, T.R., 1984. The role of condensed tannins
making more energy available for growth. The in the nutritional value of Lotus pedunculatus for sheep. 2.
Quantitative digestion of carbohydrates and proteins. Br. J.
reduced metabolic rate could have been reflect- Nutr. 51, 493–504.
ed in the lower glucose values detected in the Bartmann, R.M., Carpenter, L.H., 1982. Effects of foraging
tannic acid animals (Table 6) and their in experience on food selectivity of tame mule deer. J. Wildl.
generally lower thyroxin values (although the Manage. 46, 813–818.
latter were not significantly different from the Beuter, E., Duron, O., Kelly, B.M., 1963. Improved method
for the determination of blood glutathione. J. Lab. Clin.
control group, Ps0.083). We did not find Med. 61, 882–888.
literature indications of a suppressing effect of Bravo, L., 1998. Polyphenolics: chemistry, dietary sources,
tannic acid on thyroid hormone levels. There is metabolism and nutritional significance. Nutr. Rev. 56,
some evidence that flavonoids might decrease 317–333.
thyroid hormone levels (Schroeder van der Elst Butter, N.L., Dawson, J.M., Wakelin, D., Buttery, P.J., 2000.
Effect of dietary tannin and protein concentration on nem-
et al., 1991; Yin et al., 1999), but thyroid- atode infection in lambs. J. Agric. Sci. 134, 89–99.
hormone-inducing effects of flavonoids are also Clauss, M., 2003 Tannins in the nutrition of wild animals: a
discussed (Skibola and Smith, 2000). review. In: Fidgett, A., Clauss, M., Ganslosser, U., Hatt,
J.M., Nijboer, J. (Eds.), Zoo animal nutrition, Vol. II. Furth,
¨
These possibilities should be further investigated Germany: Filander Verlag, pp. 51–87.
in completely controlled trials. In this trial, even Clauss, M., Flach, E.J., Tack, C., Lechner-Doll, M., Streich,
W.J., Hatt, J.M., 2003. Food intake, preference patterns and
if the offered hay and the enclosure vegetation did
digestion coefficients in captive giraffe (Giraffa camelopar-
not differ in quantity or quality, an increased intake dalis) offered a tannin-containing pelleted diet. Verh. Ber.
by one animal group could not be ruled out. Erkr. Zootiere 41, 333–342.
Our study confirms that roe deer do not neces- Cooper, S.M., Owen-Smith, N., 1985. Condensed tannins deter
sarily avoid tannin substances consistently, and feeding by browsing ruminants in a South African savanna.
Oecologia 67, 142–146.
that a certain amount of tannin is included in the
Cooper, S.M., Owen-Smith, N., Bryant, J.P., 1988. Foliage
diet in simple choice situations. Potentially, posi- acceptability to browsing ruminants in relation to seasonal
tive effects of the tannins might include higher changes in the leaf chemistry of woody plants in a South
feed conversion efficiency. However, the large African savanna. Oecologia 75, 336–342.
differences observed between individuals makes a De Rosa, G., Moio, L., Napoletano, F., Grasso, F., Subitosi,
precise regulation of tannin intake at a species- L., Bordi, A., 2002. Influence of flavor on goat feeding
preferences. J. Chem. Ecol. 28, 269–281.
specific level unlikely. Distel, R.A., Provenza, F.D., 1991. Experience early in life
affects voluntary intake of blackbrush by goats. J. Chem.
Acknowledgments Ecol. 17, 431–450.
Driedger, A., Hatfield, E.E., 1972. Influence of tannins on the
nutritive value of soybean meal for ruminants. J. Anim. Sci.
We thank P. Baumann for his support in captive 34, 465–468.
roe deer husbandry and management, K. Manner ¨ Duncan, A.J., Hartley, S.E., Iason, G.R., 1994. The effect of
for his support in producing the pelleted roe deer monoterpene concentrations in Sitka spruce (Picea sitchen-
sis) on the browsing behaviour of red deer (Cervus elaphus).
diet, and the Zoological Garden of Rostock for the Can. J. Zool. 72, 1715–1720.
support of this study. Faurie, A.S., Perrin, M.R., 1993. Diet selection and utilization
in blue duikers and red duikers. Afr. J. Zool. 107, 287–299.
¨
Fickel, J., Goritz, F., Joest, B.A., Hildebrandt, T., Hofmann,
References
R.R., Breves, G., 1998. Analysis of parotid and mixed saliva
in roe deer (Capreolus capreolus). J. Comp. Physiol. B
Alm, U., Birgersson, B., Leimar, O., 2002. The effect of food 168, 257–264.
quality and relative abundance on food choice in fallow Flanagan, P.R., Cluett, J., Chamberlain, M.J., Valberg, L.S.,
deer. Anim. Behav. 64, 439–445. 1985. Dual-isotope method for determination of human zinc
Athanasiadou, S., Kyriazakis, I., Jackson, F., Coop, R.L., 2001. absorption: the use of a test meal of turkey meat. J. Nutr.
Direct anthelmintic effects of condensed tannins towards 115, 111–122.
 M. Clauss et al. / Comparative Biochemistry and Physiology Part B 136 (2003) 369–382 381

Frutos, P., Hervas, G., Giraldez, F.J., Fernandez, M., Mantecon, McArthur, C., Robbins, C.T., Hagerman, A.E., Hanley, T.A.,
A.R., 2000. Digestive utilization of quebracho-treated soya 1993. Diet selection by a ruminant generalist browser in
bean meals in sheep. J. Agric. Sci. 134, 101–108. relation to plant chemistry. Can. J. Zool. 71, 2236–2243.
Gillingham, M.P., Bunnell, F.L., 1989. Effects of learning on ¨
Muller, P.J., Hart, J.A., Robertson, J.B., 1998. Food choice,
food selection and searching behaviour of deer. Can. J. digestion and body size among African rainforest duikers
Zool. 67, 24–32. fed natural fruit diets. Proc. Comp. Nutr. Soc. 2, 145–149.
Gillooly, M., Bothwell, T.H., Torrance, J.D., MacPahil, A.P., Provenza, F.D., Burritt, E.A., Clausen, T.P., Bryant, J.P., Rei-
Derman, D.P., Bezwoda, W.R., et al., 1983. The effect of chardt, P.B., Distel, R.A., 1990. Conditioned flavor aversion:
organic acids, phytates and polyphenolics on the absorption a mechanism for goats to avoid condensed tannins in
of iron from vegetables. Br. J. Nutr. 49, 331–342. blackbrush. Am. Nat. 136, 810–828.
Greger, J.L., Lyle, B.J., 1987. Iron, copper and zinc metabolism Robbins, C.T., Hagerman, A.E., Austin, P.J., McArthur, C.,
of rats fed various levels and types of tea. J. Nutr. 118, Hanley, T.A., 1991. Variation in mammalian physiological
52–60. responses to a condensed tannin and its ecological implica-
¨
Hassig, A., Liang, W.X., Schwabl, H., Stampfli, K., 1999. tions. J. Mammal. 72, 480–486.
Flavonoids and tannins: plant-based antioxidants with vita- Rusterholz, M., Turner, D.C., 1978. Experiments on ‘nutritional
min character. Med. Hypoth. 52, 479–481. wisdom’ of roe deer. Rev. Suisse Zool. 85, 718–730.
Hagerman, A.E., Robbins, C.T., Weerasuriya, Y., Wilson, T.C., Schroeder van der Elst, J.P., van der Heide, D., Kohrle, J.,
McArthur, C., 1992. Tannin chemistry in relation to diges- 1991. In vivo effects of flavonoid EMD 21388 on thyroid
tion. J. Range Manage. 45, 57–62. hormone secretion and metabolism in rats. Am. J. Physiol.
Hawkins, G.E., 1955. Consumption and digestibility of Les- 261, E227–E232.
pedeza sericea hay and alfalfa hay plus gallotannin. J. Dairy. Schwartz, C.C., Regelin, W.L., Nagy, J.G., 1980. Deer prefer-
Sci. 38, 237–243. ence for juniper forage and volatile oil treated foods. J.
Hissin, P.J., Hilf, R., 1976. A fluorimetric method for deter- Wildl. Manage. 44, 114–120.
mination of oxidized and reduced glutathione. Anal. Bioch-
Skibola, C.F., Smith, M.T., 2000. Potential health impacts of
em. 74, 77–82.
excessive flavonoid intake. Free Radical Biol. Med. 29,
Jansman, A.J.M., Houdijk, J.G.M., Verstegen, M.W.A., 1993. 375–383.
Effects of condensed tannins in faba beans on the availabi-
Smith, P.K., Krohn, R.I., Mallia, A.K., Gartner, F.H., Proven-
lity of minerals in pigs. In: Schlemmer, G. (Ed.), Bioavail-
cano, M.D., Fujimoto, E.K., et al., 1985. Measurement of
ability ’93. Nutritional, Chemical, and Food Processing
protein using bicinchoninic acid. Analyt. Biochem. 150,
Implications of Nutrient Availability. Fed. Europ. Chem.
76–85.
Soc. 2, 48–52.
Spalinger, D.E., Cooper, S.M., Martin, D.J., Shipley, L.A.,
Kabasa, J.D., Opuda-Asibo, J., ter Meulen, U., 2000. The
effect of oral administration of polyethylene glycol on faecal 1997. Is social learning an important influence on foraging
helminth egg counts in pregnang goats grazed on browse behavior in white-tailed deer? J. Wildl. Manage. 61,
containing condensed tannins. Trop. Anim. Health Prod. 32, 611–621.
73–86. Sreerangaraju, G., Krishnamoorthy, U., Kailas, M.M., 2000.
Kaitho, R.J., Umunna, N.N., Nsahlai, I.V., Tamminga, S., Evaluation of bengal gram (Cicer arietinum) husk as a
Van Bruchem, J., 1998. Utilization of browse supplements source of tannin and its interference in rumen and post-
with varying tannin levels by Ethiopian Menz sheep. 1. rumen nutrient digestion in sheep. Anim. Feed Sci. Technol.
Intake, digestibility and live weight changes. Agroforestry 85, 131–138.
Syst. 39, 145–159. Tebib, K., Rouanet, J.M., Besancon, P., 1997. Antioxidant
Kies, C., Umoren, J., 1989. Inhibition of copper bioutilization: effects of dietary polymeric grape seed tannins in tissues of
fiber, lead, phytate and tannins. Adv. Exp. Med. Biol. 258, rats fed a high cholesterol-vitamin E-deficient diet. Food
81–93. Chem. 59, 135–141.
Kirschner, A.C., Putnam, L.A., Calvin, A.A., Irlbeck, N.A., Tixier, H., Duncan, P., Scehovic, J., Yani, A., Gleizes, M.,
1999. Browse species preference and palatability of Colobus Lila, M., 1997. Food selection by European roe deer
guereza kukuyuensis at the Denver Zoological Gardens. (Capreolus capreolus): effects of plant chemistry, and con-
Proc. Nutr. Adv. Group Conf. 3, 35–45. sequences for the nutritional value of their diets. J. Zool.
Kossak, S., 1981. Development of food habits in roe deer. Lond. 242, 229–245.
Acta Theriol. 26, 483–494. Tixier, H., Maizeret, C., Duncan, P., Bertrand, R., Poirel, C.,
Krueger, W.C., Laycock, W.A., Price, D.A., 1974. Relation- Roger, M., 1998. Development of feeding selectivity in roe
ships of taste, smell, sight and touch to forage selection. J. deer. Behav. Proc. 43, 33–42.
Range Manage. 27, 258–262. Vaquero, M.P., Veldhuizen, M., van Dokkum, W.,
Kumar, R., Singh, M., 1984. Tannins, their adverse role in van den Hamer, C.J.A., Schaafsma, G., 1994. Copper avail-
ruminant nutrition. J. Agric. Food Chem. 32, 447–453. ability from breakfasts containing tea. Influence of the
Liu, F.Y., Hodgson, J., Barry, T.N., 1998. Effects of grazing addition of milk. J. Sci. Food Agric. 64, 475–481.
sequence and condensed tannins on ingestive behaviour, Verheyden-Tixier, H., Duncan, P., 2000. Selection for small
herbage intake, and performance of lambs grazing Yorkshire amounts of hydrolysable tannins by a concentrate-selecting
fog pasture. N. Z. J. Agric. Res. 41, 359–366. mammalian herbivore. J. Chem. Ecol. 26, 351–358.
382 M. Clauss et al. / Comparative Biochemistry and Physiology Part B 136 (2003) 369–382

Verheyden-Tixier, H., Duncan, P., Guillon, N., 1999. Roe deer Waghorn, G.C., Ulyatt, M.J., John, A., Fisher, M.T., 1987. The
choose to include small amounts of tannins in their diet. effect of condensed tannins on the site of digestion of amino
´ Juilllet 1999, 147–152.
Entreties de Chize, acids and other nutrients in sheep. Br. J. Nutr. 57, 115–126.
Voser-Huber, M.L., Nievergelt, B., 1975. Das Futterwahlver- Yin, F., Giuliano, A.E., Van Herle, A.J., 1999. Growth inhibi-
halten des Rehes in einem voralpinem Revier. Z. Jagdwiss. tory effects of flavonoids in human thyroid cancer cell lines.
21, 197–215. Thyroid 9, 369–376.