birds
Kurk, C.D.
Citation
Kurk, C. D. (2008, May 27). The bill of evolution : trophic adaptations in anseriform birds. Retrieved from https://hdl.handle.net/1887/12867
Version: Corrected Publisher’s Version
License: Licence agreement concerning inclusion of doctoral thesis in the Institutional Repository of the University of Leiden
Downloaded from: https://hdl.handle.net/1887/12867
Note: To cite this publication please use the final published version (if applicable).
T
The grazi how bite
C
ing mech e size, bit affe
Chapte
hanism in te rate a ect intake
er 5
n geese a nd amou e rate
and swan unt of sp
ns:
pill
In is d stu We can leu Fo siz gra Th ba ge co ery (B.
rat To req int an the gra tha
mammalian herb determined by th udy we investigat e performed graz nadensis), lesser ucopsis) on small, r all anatid specie e, but the increa azers loose signif e amount of gras rnacle geese no c ese and mute sw nsequence intak ythropus), but at .canadensis and te levels off.
determine the r quirements, inclu take rates per me d mallard (Anasp e oral cavity may azing tips of leave an quantity in thi
bivores, the intak he product of bite
te how these com zing experiments
white-fronted ge , 8 cm tall turfs.
es in our study in se is smaller as s ficant amounts of
ss lost increases w correlation betw wans bite size and e rate increases q
larger bite sizes C.olor) start to b
elative performa uding data from g etabolic weight. T platyrhynchos) sh y explain the relat es in barnacle ge is species.
Summary
ke rate of grazing e size and bite ra mponents contrib s with mute swan eese (Ansereryth
ntake rate increas pecies are larger f severed vegeta with bite size. In ween bite rate and
d bite rate show a quickly at small b the decrease in b balance the effec
ance of grazing, w grazing Anas spe The barnacle goo how relatively low tively low intake eese may reflect s
g on a small spati ate (Spalinger and
bute to the intake ns (Cygnusolor), hropus) and barn
ses proportionall r. Unlike in mamm
tion during trans both lesser whit d bite size was fo a negative correl bite sizes (B.leuc bite rate and the ct of increasing b
we scaled intake r cies. Geese and s ose, Eurasian wig
wer intake rates.
rates in the Anas selection for qua
al and temporal d Hobbs, 1992). I e rate in avian gr Canada geese (B acle geese (Bran
ly with increasing malian grazers, av sport through the
e-fronted and ound, but in Cana
ation. As a opsis and A.
increase of food ite size, and intak
rate to metabolic swans have simila
eon (Anaspenelo . The morphology s species, while lity of food rathe
scale n this razers.
ranta
nta
g bite vian e bill.
ada
d loss ke
c ar ope) y of er
To co un rem To foc gra sea (th 19 gra 19 alt bit inc 19 Mo int 19 dif siz Eh 19 Les An ha Pro Va is d Se al., Siz an go bit rel Gr cla Do tru 20 are
cope with the lo nsiderable amou it of time, determ maining for othe
understand the cussed on a smal azing on food sat arch for food item he amount of gra
92; Parsons et al azing animals ma 93a; Shipley et a ternatively by dec te, thereby increa crease in bite size 86; Laca et al., 19 orphological prop take rate (Anders
96; Illius and Gor fferences in body e of a bite (Gong rhardt, 1988; Shi 87).
ss detailed inform natidae (geese, sw ve quantified int op, 1991; Ridding
n der Wal et al., determined by va dinger and Ravel , 2003; Hupp et a ze of the bill may atid grazers. For ose species (Owe te size is the main lated with bill siz
azing as feeding ade; of which som onne-Goussé et a ue geese (Ansera 02; Livezey, 1996 e herbivorous, al
ow-energy conten unt of time feedin mines the amoun r activities that a
mechanism dete l temporal and s turated patches.
ms is negligible, i ss taken per bite ., 1994; Spalinge ay reduce time sp l., 1994; Spalinge creasing handling asing bite rate. A e leads to a declin
994; Prache, 199 perties of mamm sen and Sæther,
rdon, 1987). On b y size and the size g et al., 1996; Gor ipley et al., 1994)
mation is availab wans and ducks).
ake rate (Durant gton et al., 1997;
1998). Similar to ariation in handli ling, 1986) and b al., 1996; Stahl, 2 be an important bite rate, a nega en, 1980, but see n determinant of e on both intersp mechanism seem me Anas species r al., 2002; Livezey, andBranta) than 6b, 1997a; Soren though most spe
Introduction
nt of grasses, gra ng. Intake rate, i.e nt of time investe affect fitness.
ermining intake r patial scale: intak Assuming that o ntake rate is det e) and bite rate (G er and Hobbs, 199 pent feeding by c er et al., 1988; Sp g time, which is t As processing time
ne in bite rate (Fo 7; Wickstrom et malian grazers hav
1992; Demment both intra- (Illius e and shape of th rdon et al., 1996;
) as well as to the
le on the mechan A number of fie t et al., 2003; Has Rowcliffe et al., o mammalian gra ing time (Mayhew ite size (Black et 2001).
t morphological c ative relation with e Durant et al., 20 f intake rate, and pecific and intras ms to have evolve
represent the mo , 1991, 1997a). Sw
the dabbling duc son et al., 1999;
ecies secure their n
azing herbivores h e. the quantity o ed in feeding and ate in the field, s ke rate over shor n food saturated ermined by the Gross et al., 1993 92; Ungar and No cropping larger bi palinger and Hob the time needed e often depends orbes, 1988; Hud al., 1984; Wilmsh ve been related t
and Greenwood, et al., 1995) and he feeding appara
; Illius and Gordo e rate of biting (I
nism controlling ld studies on her ssall et al., 2001;
1999; Therkildse zers intake rate i w and Houston, 1
al., 1992; Cope e character contrib h bill length was 003). Durant et a d bite size appear specific (Cope et a
ed several times ost recent examp wans (Cygnus) ar cks (Brush, 1976;
Sraml et al., 199 r food from aqua
have to spend a f food consumed d the amount of t
several studies ha rt periods of activ patches time to product of bite s 3a; 1993b; Laca et
oy-Meir, 1988). H ites (Gross et al., bs, 1992), or to crop and proc on bite size, an dson and Watkins
hurst et al., 1999 to the performan , 1988; Gordon e d interspecific lev atus correspond on, 1987; Janis an llius and Gordon,
intake rate in gra rbivorous wildfow Lang and Black, 2 en and Madsen, 2 n anseriform gra 1998; Owen, 197 et al., 2005; Dura buting to intake ra
observed for sev l. (2003) found t rs to be positively
al., 2005) level.
within the anatid ple (wigeon spp., re closer related
; Donne-Goussé e 6). Like geese, sw tic plants, which
d per time ave ve
ize t al., Hence,
cess a s, 9).
nce of et al.,
el to the nd
,
azing wl
2001;
2000;
azers 72;
nt et ate in veral
hat y
d to the et al., wans
may
req ho sw Th fea the ch pre im mo tra de ha As eco un An by sho the
Exp Gr les Th an sw ge Th co co rem in ex the we allo pe
quire a somewha owever, all specie wans has not been
e time since dive atures related to e morphology of apter 4). The oro esence of spines portant role in th outh. Species lack ansport food, wh
r Leeuw et al., 20 ndling time and
feeding mechan ology of anserifo derstanding of th natidae. We comp examining bite s ort periods of act e lining of the up
perimentalsetu azing experimen sser white-fronte ree sets of trials d June 2005 and wans were used a
ese were examin e trials took plac mmercially availa nducted within a maining two mut an outside aviary periments, the b e training session ere cut from large owed to grow un renne with a sma
at different feedi es are able to sust
n studied.
ergence and degr grazing (chapter the tongue and opharyngeal morp
on the roof of th he retention of v
king such spines ich results in the 003). Such morph
indirectly intake nisms are importa orms (Pettifor et a he mechanism of pared the intake size, amount of le
tive grazing. Obs pper bill.
Ma
up
ts were performe ed geese, and two were conducted
the third betwee nd in the second ned.
ce in an inside avi able cereals and an enclosure insid
te swans and the y also with flowin irds were trained ns the actual mea er sods purchase ntil use (max. 20 c
all proportion of
ng technique the tain on swards. U
ree of adaptation r 2), but also in th lining of the bills phology of geese he mouth. These egetation during (Anas) use a diffe loss of a large pr hological differen
rate.
ant in understand al., 2000), this stu f grazing and its u rate of selected eaves lost during erved difference
aterialsandMe
ed with four mut o large Canada ge
, the first in Augu en June and Augu d set the barnacle
iary with flowing pellets). Both tra de the aviary, allo e Canada geese w
ng water and ad l d for several wee asurements of int ed from a garden
cm). The sods we Festucarubra.
en grazing on land Up to now, intake
n seem not only t he morphology o (Van der Leeuw e is particularly ch
spines are believ g the transport of erent transport m roportion of the nces are therefor
ding the foraging udy aims to broa underlying factor species of Anser, g food transport,
s were related to
ethods
te swan, two bar eese, all purchas ust 2004, the sec ust 2005. In the f e geese and the le
water and ad lib aining and experi owing individual were individually t
libitum food. Imm eks (1-3) to eat fr
take rate were p centre. The sods ere dominated by
d. In aviculture, e rate of grazing
to be reflected in f the oropharynx et al., 2003 and haracterized by t ved to play an f food through th mechanism to
cropped leaves ( re expected to af
g and population den our
rs in herbivorous , Branta and Cygn
and bite rate ove o the morphology
nacle geese, two ed from a local tr ond between Ap first set two mute esser white-front bitum food (mixtu mental sessions birds to be teste tested in an encl mediately prior to
om small turfs. A erformed. The tu s were put outsid y Poa spp. and Lo in
skull x, i.e.
he he
Van ffect
nus er
y of
o rader.
pril e ted ure of
were d. The osure o the After
urfs de and
olium
Th ins we Th fol gu 30 bit A t Ab (ba tw bir
Me
To 89 so ab an de To ne ea bir the fee Int los Bit of Du rep bit siz the as too loo cal
e afternoon prec side. The turfs we ere cut to ground e evening before lowing day betw ided to the enclo to 40 bites, or, t tes were taken fr trial lasted about bout ten trials per
arnacle en lesser wice a day. Trials u
rds.
easurements
characterize the .92 cm2 was clip rted in green and out 54°C to cons d ranged betwee nsity of the sods assess the amou arest 0.1 g befor ch trial the turf w rd) for 5 minutes
e total weight los eding bird.
take was then ca st during grazing.
te depth was calc the experiment a uring each trial a
play of these reco ting and to calcul
e (fresh weight p e number of bite
intake divided by ok much longer t ose of tillers with
lculations.
ceding a trial, fou ere secured to tra d level, and the re e a trial, food was ween 08:00 h and osure and allowe to minimize deple rom an area prev t 5 minutes, after r bird were cond -white fronted ge usually finished b
e turfs offered, le ped to trial heigh d dead material, w stant weight. The en 1064.28 g/m2
remained consta unt of grass remo re and after each was weighed and
and then re-weig ss during the tria
lculated as the b .
culated by measu and the height of video-camera re ordings were use
ate the duration per bite in g) of a s observed from y total time of gr to ingest than the h the roots still at
ur turfs were prep ays measuring 40 emaining grass w s removed from t 09:00 h. Five mi d to settle down etion effects on i iously grazed.
r which spilled gr ucted at a rate o eese), while the before 12:00, afte
eave density was ht and then cut to
weighed and the mean density of and 474.87 g/m2 ant (F3,23 = 2.23, p oved after grazing trial. To account placed in the ex ghed. The evapo l to derive the bi iomass removed uring the differen f the area grazed corded the activi ed to verify the co
of cropping and trial was calculat the video during razing, corrected e average of the t ttached, and wer
pared from the la 0x 15 cm (600cm was cut to 8 cm m the aviaries. Tria nutes before a tr . The animal was ntake, a trial was rass leaves were c f one trial a day f remaining birds w er which food wa
measured for ea o ground level. T e green material w
f fresh green leav
2. Throughout the p = 0.112).
g, each turf was w t for evaporative
perimental set u rative weight los omass removed minus the weigh nce between swa d.
ity of the grazing ounted bites, to d
transport per bit ted by dividing b g that trial. Intake
for excessive bit trial were mostly e excluded from
arge sod and put
2), inflorescences measured with a r
ls started the rial a bird was ge s allowed to graze
s ended as soon a collected.
for the smaller ge were offered a tu as returned to the
ach sod. An area o The cut vegetatio
was then dried at ves was 667.26 g/
e experiments le
weighed to the weight loss, prio p (fenced off for ss was subtracted from the turf by ht of the grass lea ard height at the
g bird. Frame by f determine the ra te. The average b iomass removed e rate was determ
e durations. Bite y due to the teari
time related s ruler.
ntly e for
as
eese urf
e
of n was t /m2 ave
or to the d from
the aves
start
frame ate of bite
by mined
s that ing
Sta
Da rou Wa To bit an sig use
Int In Int ind co sw tha is r ge fro
Bit Bit su (ta rat va the sho In rat exc no wit thi wh
atisticalanalyses ata were ln-transf utine (S)MATR (v arton and Weber assess differenc te rate nested AN
d individuals wer gnificant (p < 0.05 ed when variance
takerate
table 5.1 intake r take rate differs s dividuals (F6, 124 = mpletely due to t wan #3. All other i at each species d roughly twice as ese intake rate m onted geese have
tesize
te size (i.e. amou bsequently small able 5.1). Species te the difference riability in the m e analysis makes ow that each spe all four anatid sp te and bite size (f cept for two of th ot significantly dif
th a single comm is common slope hite-fronted and s
formed and analy v1, Falster et al. (2
r (2002).
es between spec NOVA procedures re random variab 5), post-hoc tests es were not equa
rate and the fact significantly betw
10.871, p = 0.00 the low intake ra individuals do no differs from all ot high as in the Ca more than double
e an intake rate in
nt of grass remov ler in Canada goo differ significant between individ ute swan data. R the variation of ecies differs from pecies measured, figure 5.1). On th he four mute swa fferent among ind mon slope (1.09 9
e y-intercepts diff barnacle geese h
ysed with SPSS 1 2003), which imp cies in intake rate s were used. Spe bles nested withi s were performed al.
Results
ors underlying in ween species (F3,
00). The statistic ate for mute swa ot differ within th
her species (all p nada geese, exce es the intake rate ntermediate of th
ved per bite) is la ose, lesser white- tly in bite size (F3
duals (F6, 124 = 13.9 emoving mute sw individuals withi m all other species
, a clear positive he individual leve ans. A model II re dividuals (p = 0.1 5% CI 0.96-1.23) fer significantly (F have similar inter
2.0 and the stand plements the algo e, bite size, amou
cies were consid n species. When d. The Games-Ho
ntake rate are list
6.06 = 16.97, p = 0 cal significance of n #2 and the high he same species.
p < 0.000). In the ept for individual e in barnacle gees
hose in Canada a
argest in the mut -fronted goose a
3, 6.05 =60.90, p < 0 955, p = 0.0000) i wan #1 and #2 or n species not sign s (all p < 0.000).
relationship is fo l this correlation egression shows 1489) and that th for all birds (figu F9, 124 = 146.281, rcepts, but ln inta
dardized major a orithms develope unt of loss per bit
ered as fixed effe F-values proved owell method wa
ted per individua 0.0024) and betw f individual variat h intake rate for
Post hoc tests sh mute swan intak #2, and in Canad se. The lesser wh and barnacle gees
te swan, and nd barnacle goos 0.0001). As for in is the result of th r swan #3 and #4 nificant. Post hoc ound between int is always signific that the slopes a e data are consis ure 5.1, table 5.2) p = 0.000). Lesse ake rate for the s
axis ed by te and
ect to be s
l.
ween tion is
mute how ke rate
da hite-
se.
se ntake he 4 from
c tests take cant are
stent ). For er
ame
bit rat
Tab of Int
In Cy
# Cy
#2 Cy
#3 Cy
#4 Br ca Br ca A er A er Br
# Br
#2
Tab rat
Sp Cy Br A Br
te size is lower in tes relative to the
ble5.1. Averages w intake rate per ind ake rate = (bite siz
ndividual ygnusolor 1
ygnusolor 2
ygnusolor 3
ygnusolor 4
ranta
anadensis #1 ranta
anadensis #2 nser
rythropus #1 nser
rythropus #2 rantaleucopsis 1
rantaleucopsis 2
ble5.2. Pearson co te. Ln intake rate =
pecies ygnusolor rantacanadensis nsererythropus rantaleucopsis
the two Canada eir bite size than
with standard devia ividual of anatid sp e- amount of lost l
Intake rate (mg min-1)
B
14051.07 (2751.41)
4 (1 6933.09 (1353.99)
3 ( 18135.35 (3854.37)
9 (2 12548.18 (5472.23)
7 (2 6676.43 (1553.32)
1 ( 7011.12 (1125.26)
1 ( 4877.93 (1620.02)
4238.28 (1482.51) (
2518.45 (1024.43) (
2584.03 (637.96)
orrelation-coefficie slope * ln bite size
Pearson’s r s 0.630 ** 1 0.826 ** 0 0.858 ** 1 0.919 ** 1
geese (p = 0.000 all other species
ation between bra pecies studied.
eaves per bite) * b
Bite size (mg)
Loss bite ( 492.66
111.20)
159 (53.
388.71 69.27)
164 (41.
922.58 239.47)
179 (47.
765.97 249.02)
199 (63.
153.94 37.05)
10.
(6.1 149.64
29.36)
8.7 (3.4 63.15
(8.18)
2.1 (1.6 65.65
19.20)
4.0 (3.4 37.60
13.85)
3.2 (2.4 37.12
(7.16)
3.3 (1.7
ent and SMA result e + intercept.
slope Lower CI 1.011 0.824 0.910 0.723 1.327 1.053 1.125 0.937
0). The mute swa s (p = 0.000).
ckets of intake rate bite rate.
per (mg)
Loss pe bite (%) 9.85
93)
32.23 (7.57) 4.21
13)
42.35 (7.25) 9.49
54)
20.32 (6.27) 9.52
77)
27.95 (10.88) 73
15)
6.93 (3.45) 73
49)
6.11 (3.04) 5
1)
3.34 (2.41) 07
44)
6.13 (4.54) 23
40)
8.18 (5.64) 35
73)
9.06 (4.64)
s on ln-transforme
Upper
CI Intercep 1.240 3.004 1.146 4.266 1.671 2.836 1.350 3.765
ns have lower in
e and the determin
r )
Bite rate (min-1)
42.42 (7.95) 31.18 (2.95) 26.02 (7.35)
)
22.62 (4.94) 47.00 (5.12) 50.58 (6.71) 78.62 (18.55)
68.24 (8.08) 72.47 (7.29) 76.78 (12.61)
ed bite size and inta
pt Intercept slope = 1.09
2.533 3.394 3.835 3.908
take
nants
n
24
9
13
13
12
15
13
13
12
10
ake
n 59 27 26 22
Bit Bit rem sw df all tes tes
Tab ind
Sp Cy Br An Br
lnintakerate(mgmin1)
tedepth
te size may be de maining lengths o wan and the three
= 3, p < 0.001). T p < 0.001). Cana st p = 0.19), but b st, both p < 0.001
ble5.3. Average bo dividuals used in th
pecies
ygnusolor rantaCanadensis
nsererythropus rantaleucopsis
2 3 4
ln remo
7 8 9 10 11
ln intake rate (mg min-1)
etermined by bea of the initially 8 c e goose species u The mute swan cl da and lesser wh both bite slightly 1).
ody mass, length o is study. Numbers
Body (n = 4) 9.22
(n = 2) 3.88 (n = 2) 2.07 (n = 2) 2.14
5 6 7
oved per bite (mg)
Figure5 (mg) ag transfo with co (except Legend Ansere
ak width or beak cm tall leaves afte used in this study
learly has the larg hite-fronted gees
deeper than the
of the bill (gape) an in brackets indicat
y size (kg) B 2 (1.13)
8 (0.82) 7 (0.11) 4 (0.17)
8
5.1. Amount of gra gainst intake rate (m
rmed. Lines repres ommon slope but d t for A.erythropus : : Cygnusolor, * erythropus, +: Bran
length (table 5.3 er a bite clearly d y (figure 5.2; Krus
gest bite depth ( se clip similar leng e barnacle goose
d width at rostral p te standard deviati
Bill length (mm) 105.25 (14.06)
63.81 (2.13) 39.50 (0.57) 36.43 (0.40)
ass removed per bi mg min-1), both ln sent a model II regr
ifferent intercepts and B.leucopsis).
*: Brantacanadens taleucopsis.
3) and bite depth.
differ between m skal Wallis 2 = 12 Mann-Whitney U gths (Mann-Whit (Mann-Whitney
part of the bill of th ions.
Bill width (m 35.45 (0.90 22.05 (0.28 15.69 (0.01 16.45 (0.92 te ression
is, {:
. The mute
28.61, U-tests
tney U U
he
m) 0) 8) 1) 2)
Lhd()
Los Sp ne in sig tha tha dif Th mu ind Ca Th inc Ho an siz Ho are are co Slo be sw int
Length removed (cm)
ssafterclipping illing of severed sted ANOVA on t amount of food l gnificant differen
at mute swans lo an lesser white-f ffer among each e amount of gras ute swan, which dividual #3 to 42%
nada geese, 4.7 % ese data suggest creasing bite size owever, the relat d species level (t es within a speci owever, in a num e positive and so e not significantly
nsistent with a si opes vary for indi lower than for o wan are not differ tercepts differ sig
1,5 2,0 2,5 3,0 3,5 4,0 4,5
Cygnus olor
leaves during tra the ln transforme lost during food t ce between indiv oose the most (al
ronted geese (p <
other (p < 0.993) ss leaves spilled a looses about 30%
% for individual #
% in lesser white t a progressive in across species, w ionship between table 5.4). The am
es is limited.
ber of animals th metimes close to y different amon ingle common slo ividuals within sp other species (tab rent from that of gnificantly (F9, 124
s Branta
canadensis A ery
ansport through t ed weight of gras transport (F3, 6.48
viduals (F6, 124 = 1 l p < 0.001), and
< 0.000) and barn ).
as a percentage o
% of the amount
#2. The three goo e-fronted geese a
crease in the am which may also b bite size and foo mount of grass lo he correlation is s o p = 0.05. A mod g individuals (p = ope (1.71, 95% C pecies, and althou ble 5.4), the slope individuals of ot
= 22.970, p = 0.0 Anser
thropus
Branta
leucopsis
Figure5.2 the stand expressed initially 8 specified
the mouth occurs ss leaves lost sho
= 459.15, p < 0.0 .45, p = 0.202). P
that Canada gee nacle geese (p < 0 of the amount cli
clipped, ranging ose species spill m
nd 8.6 % in barn mount of food lost
be present within od loss is weak at st varies widely a significant and al del II regression s
= 0.1230) and tha I 1.444-2.029) fo ugh the slope for es for most indiv her species. For t 000). As for the sl s
2. Average and 2 ti dard error of bite si d as the length of t cm tall leaves rem per species.
s in all four speci ows that species d
001) but there is Post hoc tests sho se loose more gr 0.000), which do ipped is highest i
from 20% for much less; 6.4 %
acle geese (table t during transpor n species or indivi t both the individ and the range of l other correlatio shows that the slo at the data are or all birds.
r mute swans ten iduals of the mut the common slop opes, the differe
imes ize, the moved,
es. A differ
no ow rass o not
n the in e 5.1).
rt with iduals.
dual bite ons
opes
nds to te pe y- ences
in va Th ve We siz see ave ind Th int mi dif 0.9 dif mu ge
Tab Ln
Sp Cy B A B
Bit As ex mu be in
= 0 spe ind
y-intercept do no riation.
e y-intercept of t ry similar to the i e therefore conc
e, and that loss i ems similar for in erage bite size an dividual birds (ln
e progressive los take rate with inc nus loss) shows t fferent among ind 936-1.147) y-inte fference in increa ute swans intake ese, but only 2.4
ble5.4. Pearson co loss = slope * ln bi
pecies
ygnusolor rantacanadensis
nsererythropus rantaleucopsis
terateandhand the decrease in plained by the pr ust (by definition tween bite size a bite rate with bit 0.016, ln bite rate ecies take relativ dividual level this
ot show a consist the mute swan is intercept for Can lude that the loss ncreases expone ndividual birds an nd average loss p loss = 1.49 * ln b ss of grass with in creasing bite size
that again the slo dividuals (p = 0.1 ercepts still differ ase in intake rate rate increases 4 times slower wit
orrelation-coefficie te size + intercept.
Pearson’s r s
0.479 *** 0
0.213 2
0.368 2
0.604 ** 1
dlingtime
intake rate with rogressive loss of ) be the consequ and bite rate is w te size across spe e = -0.336 * ln bit vely larger bites b s relationship is le
tent pattern acro s low but individu nada geese or les
s of grass during entially (exponen nd across species per species (n = 4 bite size – 4.64; r
ncreasing bite siz (see before). An opes of the ln tra 1010) and that fo r significantly (F9,1
e with bite size be times slower wit th net bite size.
ent and SMA result .
lope Lower CI 0.808 0.642 2.105 1.421 2.785 1.900 1.720 1.196
increasing bite si f grass with incre uence of a reduce weak at the individ ecies (table 5.1; ln te size + 5.571 95 but at a relatively
ess clear.
oss species but re ual swans show y ser white-fronte food transport o t 1.71) with bite s. Note that the r 4) has a similar slo
= 0.957; p = 0.04 e only partly exp n analysis of net b nsformed data a r the common sl
124 = 91.612, p = 0 etween species h
th increasing bite
s on ln-transforme
Upper
CI Intercep 1.018 -0.036 3.119 -8.258 4.082 -10.304 2.474 -4.824
ize (and body size easing bite size, th ed bite rate. Agai
dual level, but th n transformed da 5% CI slope: -0.56 y lower rate than
epresent individu y-intercepts that a
d geese.
only depends on b size. This relation elationship betw ope as for the 43).
plains the decreas bite size (bite size re not significant ope (1.036, 95%
0.000). However, has become small e size than in bar
ed bite size and foo
t Intercept slope = 1.71 -5.810 (-6.296) (-4.625) -4.798
e) is only partly he remaining var in the relationshi here is a clear dec ata: r = -0.984, n 66 /-0.199). Large
small species. At al are bite nship ween
se in e
tly CI , the ler. In nacle
od loss.
n
59 27 26 22
riation ip crease
= 4, p er t the
As be an dif co 0.0 5.3
Tab Ln
Sp Cy B A B
none of the indi tween bite size a alysed separately fference in slope
mmon slope -0.7 000). Bite rate dr 3).
ble5.5.Pearsonco biterate=slope*
pecies
ygnusolor rantacanadensis
nsererythropus
rantaleucopsis
2 3
l 2,5
3,0 3,5 4,0 4,5 5,0
ln bite rate (min -1)
viduals of lesser and bite rate, the y (table 5.5). Am for the relations 778, 95% CI -0.65 ops faster with b
orrelationcoefficie lnbitesize+interc
Pearson’s r s
0.653 *** -0 0.470 ** -0
0.319 0
0.179 0
4 5 6 7
n bite size (mg)
white-fronted an e data of the mut
ong these 6 indiv hip between ln b 0 / -0.930) but th bite size in mute s
ntandSMAresults cept.
lope Lower CI
0.855 -1.044 0.582 -0.831 0.876 0.593 0.429 0.275
Figure grass transf regres and B bite s erythr Legen {: An
7 8
nd barnacle gees te swan and Cana
viduals there was bite rate versus ln he two species di swans than in Ca
sonlntransformed
Upper
CI Intercep
-0.700 8.879 -0.408 6.794 1.292 0.651 0.669 2.768
e5.3. Bite rate (mi removed per bite formed. Lines repre ssion with commo Brantacanadensis.
ize and bite rate w ropus or in Branta
nd: : Cygnusolor, nsererythropus, +:
e show a correla ada geese were s no significant n bite size (p = 0.0
iffer in intercept nada geese (figu
dbitesizeandbite
t
Intercept slope = -0.778 8.390 7.776 (7.463) (7.083)
n-1) against amoun (mg), both ln esent the model II n slope for Cygnus
No correlation bet was found either in
leucopsis.
*: Brantacanaden Brantaleucopsis.
tion
0635, (p = re
rate.
n
59 27 26 22
nt of
olor tween
Anser
nsis,
Th spe of oe bit A n
6.02
p = du wa the int ge
Tab tim
Sp Cy B A B
Va spe Sti on op pe sit
Int Ou co exc int wh ge
e inverse relation ecies and at least a an increase in t sophagus, but al te).
nested ANOVA sh
23 = 17.500, p = 0
= 0.056). In post h ration of croppin as longest in the
e least time to ap termediate value
ese species.
ble5.6. Averages w me and transport ti
pecies
ygnusolor rantacanadensis
nsererythropus
rantaleucopsis
riation in foragin ecies are better t llman et al., 2000 the foraging site ptimised. Quantify riods of active gr uations.
takerate
ur experiments sh nsiderably betwe cept for one out take rate in Cana hite-fronted gees
ese.
nship between b t partly within sp the time needed so from an increa hows that specie
.002), but transp hoc tests both va ng a bite from an
mute swan (table pprehend and tra es that are signific
with standard devia me per anatid spec
Cropping time 1.08 ± 0.46 ± 0.35 ± 0.34 ±
ng performance b than others at ha 0). Grazing anatid e, where searchin ying the underlyi razing as in the p
howed that grazi een anatid specie
of the four indiv da geese is abou se have an intake
ite size and bite pecies, and individ
to transport an ase in time inves s (and individuals port time is just a
ariables follow th 8 cm turf of gras e 5.6). Lesser wh ansport a bite of cantly different f
ations between bra cies studied.
e per bite (s) T
± 0.40
± 0.10
± 0.08
± 0.07
Discussion
between species andling and consu
d species spend m ng time is minimi
ing processes of t resent study may
ng intake rates o es. Mute swans in
iduals tested, tha t twice as high co e rate intermedia
rate, which seem duals within spec increasing amou sted in cropping ( s) clearly differ in bove significance he same pattern a ss and transporti
ite-fronted and b grass, while Cana rom mute swans
ackets of handling
ransporting time p 0.98 ± 0.39 0.79 ± 0.12 0.50 ± 0.11 0.49 ± 0.08
emerges from sit uming food item most of the time ized and processi their grazing mec y thus be a prope
on sods of 8 cm h ngest twice as m an individuals of ompared to barn te of those of Ca
ms to exist across cies, may be the
nt of food to the (i.e. time to sever n cropping time ( e level (F3, 6023 = 4 across species: th
ng the food colle barnacle geese sp ada geeseshowe s and the two sm
time split into crop
per bite (s) n
9 59
2 29
1 26
8 22
tuations where s s (Caldow et al., available for fora ing food items m chanism over sho er approach to fie
eight differ uch per unit of ti Canada geese. In nacle geese. Lesse nada and barnac
result r a (F3,
4.473, he ected
pent ed
aller
pping
9 9 6 2
some 1999;
aging may be
ort eld
ime, n turn,
er cle
Co use Pro the fin ba the mi fou wa an dif al.
of go Vu is 2 me To int bit rat bit spe lar inc los a p rat
Bit
Lik siz siz an co ba bit yie on stu ba ba (Ha
omparable data fr ed in this study is op and Black, 199 erefore difficult t
dings on dry wei sed on the ratio e study of Durant
nute on 8 cm tal und in the presen ater content, as N d Van der Graaf fferences in swar (2003). It may w digestion results ose data on dige ulink, 1992), whic 2.9 times too low entioned study h
account for spec take rate by relat te rate to intake r te increases prop te. The rate at wh ecies. Intake rate rger Canada goos crease implies a c st. The difference progressive loss o te with increasing
tesize
ke mammalian he e of the mouth ( e with increasing atid species in th mpare bite sizes.
rnacle geese in t te on grass of 8 c elds bite sizes of 3
8 cm tall turfs of udy (34 mg). Inte
rnacle geese, but rnacle geese on t assall et al., 2001
rom literature fo s scarce and only 98). In the study to compare with ght of grass leav found in the pres t et al. (2003) had l swards, which i nt study. This diff Naujeck and Hill ( et al. (2006) foun rd density betwee well be that the in
s in an underestim stibility are avail ch show that the w. Correcting for t
ad an intake rate cies-specific feat ting bite size, am rate as well as to portionally with in
hich intake rate i e increases much se, and intake rat complex relation es in rate of incre of food during tra
g bite size (see be
erbivores, which Gordon et al., 19 g size of the bill (o he study of Duran . Based on the co he study of Dura m, but correcting 31 mg. Cope et a f 32 mg. Both val restingly, lesser w t do take larger b taller swards is th 1; Van der Wal et
r the anatid spec y available for the by Prop et al. (19 our experiment.
es. Assuming a w sent study (x = 83 d an intake rate o s considerably lo ference is unlikel (2003)) estimated nd a conversion f
en the present st ndirect determina mation of the tru
able (Prop and B digestibility of 14 this inconsistenc e of 2763 mg, ver ures of intake rat ount of food lost o each other. For ncreasing amoun ncreases with bit faster in the two te is lowest in the ship between bit ease in intake rate
ansport through t elow).
show a positive r 996), the anatids
or body size). A s nt et al. (2003). O onversion rate of nt et al. (2003) in g for their low es al. (2005) found a lues are very clos white-fronted ge bites. One possib hat these birds h t al., 1998), called
cies, grass species e barnacle goose 998) grass densit Durant et al. (20 water content of 8
3.0, s.d. = 1.9, n = of about 1735 m ower than the 255
y to be due eithe d a water conten factor of 46% for tudy and those re ation of intake th
e intake rate. On lack, 1998; Prop 4% as estimated y, barnacle geese ry close to our ow te, we examined t during transpor
all anatid species nt of grass remov
te size, however, o smaller goose s e mute swan. Thi te size, bite rate a
e with bite size a the bill and partl
relationship betw in our study show similar relationsh Only a few studies
83% for fresh to ngest about 22 m stimate of digesti an average bite si se to the bite size eese have similar le explanation fo ave difficulty gra d the ‘spaghetti e
s and sward heig (Durant et al., 20 y was very low, a 003) based their
83% of fresh wei
= 25) barnacle ge g fresh weight pe 50 mg per minut er to our estimat nt of grasses of 72
Festucarubra,o eported in Duran hrough the estim nly for the barnac
et al., 2005; Prop by Durant et al. ( e in the previous wn estimate.
the mechanics o t through the bil s in our study, int ved from the turf differs among th species than in th
s proportional and amount of fo
re partly explain y by a decrease i
ween bite size an w an increase in hip was found for s are available to o dry matter the mg fresh weight p
bility (see above ize for barnacle g e determined in o
bill dimensions a or the small bite s asping the long le effect’ by R. Dren
ght 003;
and ght, eese in
er e e for 2%, or to nt et ation cle p and (2003)
of l and take
per he he ood
ed by n bite
d the bite r the o per
) geese
our as size of eaves nt
(pe wh po lar pu be inc Lan the bit bil (ex oft tim A t ba de
Los Aft oe bu ve alt Alt dim tak tra sev siz da the Va pla pa tra for oe em rel for 19 sho Ho
ers. com., see als hite-fronted gees ossibility is that th rger bites, either
lling movement tween the bills d crease when swa ng and Black, 200 e limiting effect o te size. In the mu
l dimensions and xponent 0.948, 9 ten uses more th mes.
third explanation rnacle geese is th pth below).
ssoffoodduring ter cropping a bit sophagus. Transp t the effect is no ry close to the le ternating forward though experime mensions of bill a kes as much time ansport the same veral variables in e and amount of ta suggest that t e mechanics of fo
n der Leeuw et a ays an important pillae on the roo ansported efficien
rward again, afte sophagus. In the mploys a complet leased after a jer rward movement
85) or ‘catch-and ort papillae are p owever, as bite si
so Bignal, 1984; V se does not seem he barnacle goos
because they can of the head or be during the pull. N
rd height increas 01; Van der Graa of bite or pulling tes swans and Ca d sward characte
5% CI 0.787/1.14 an one pull to re n for the differenc
hat the latter sele
gtransport
te food has to be port time seems t very strong. Th vel for significan d and backward m ental data are not
and tongue, the t e to transport an e amount of food ncrease in magnit f food lost is very he amount of foo ood transport thr al. (2003) suggest role in the efficie f of the mouth m ntly. In geese, th er a backward mo e mallard (Anasp
tely different tran rky backward mo
t of the head wit d-throw’ mechan present (chapter ze increases the
Van der Graaf, 20 m to be affected b e is not able to p n not generate th ecause they are u ote that in the ba ses above 8 cm (C f et al., 2006). An force on bite size anada geese crop ristics) increases 42). Our video-re emove the graspe ce in bite size bet ect for the high q
e transported dow to increase with e differences in t ce p = 0.05. Gras movements of th t available, it see transport movem
amount of food d along a long bill tude, including bi y similar across sp
od lost before sw rough the bill.
ted that the morp ent transport of may determine th ese papillae reta ovement to trans latyrhynchos) suc nsport mechanism
vement of the he h open bills (iner nism (Kripp, 1933 4) and form a ro number of leave
006). However, th by this ‘spaghetti- produce the force he force to break unable to keep th arnacle geese bit Cope et al., 2005 nother observatio e is the increase pping time of ind almost proportio cordings showed ed leaves, resultin tween lesser whi quality parts of th
wn the length of bill length (body transport time be ss is transported t he tongue (Van de ems likely that wit ments also becom along a short bill . However, with ite size. The relat pecies and within wallowing only de
phology of the in severed vegetati he amount of veg in food items wh sport food in the ch papillae are ab m during grazing, ead and caught a rtial transport, (B )). In most grazin ugh inner lining o es of grass that m
he bite size of les -effect’. A second es required to tak k the grass during he bite clamped te size does not
; Durant et al., 20 on that may indic in cropping time dividual birds (con
onal with bite siz d that mute swan ng in longer crop ite-fronted and he forage (see bit
the bill to the y size) across spec
etween species a through the bill b er Leeuw et al., 2 th increasing line me larger, and tha
l as it takes to increasing body s tionship between n individuals. The epends on bite siz
nterior of the mo on. The presence getation that can hen the tongue m direction of the bsent. This specie , in which food is again during the Bramble and Wak
ng anatid species of the mouth.
make direct conta sser
d ke g the
003;
cate with nstant ze ns pping
te
cies, are
by 2003).
ear at it
size n bite ese
ze and
uth e of be moves
es s ke,
many ct
wit res inn du go the ind the pa ap ton ma thr pa rel co ob Lar or tra am pe (20 pre siz los mo et the me bo pa roo
Bit Fo ne be Ac low fre co Lan gra
th the roof of the sulting in longer t ner lining of the u ring transport, b ose species invol e amount of food dividuals (20%-42 e caudal part of t rt of the mouth.
parently the pap ngue and to avoi allard. We did no row’-mechanism rt of the bill and latively large and nditions grazing bserved to drop a rge losses during
‘catch-and-throw ansport large foo mounts of grass a rcent of the amo 003). Note that s esent study. Mal e and food loss f ss of 12 percent i outh in this speci al., 2003; Zweers e mouth and use echanisms for fee old palate results
late may limit th of of the mouth,
terate
od intake rate no gative relationsh tween species.
ross species the w frequency (33 b equency (74 bites
mparable height ng and Black (200 ass species). In le
e mouth during f transport times a upper bill, all thre ut losses were le lves loss as well i d severed per bit 2%). Compared to the mouth and lo
This makes the r pillae and ridges a d the energetica ot observe the ch in the mute swa the small amoun d variable food lo
mute (Sears, 198 large proportion g the transport of
w’ transport mec d items, is also u re not caught wit ount of grass clipp uch losses are re lards take 125 m found in our stud n the mallard ins es is believed to s et al., 1977). Du d as a piston to g eding suggest tha in an inefficient e performance o creating a ‘leaky
ot only depends o hip between bite
large bite sizes ta bites/min), while s/min). Bite rates are similar to th 01): 70-90/min, V esser white-front
ood transport (in and a larger loss ee goose species ess than 10%. Wh s unknown. Mut e, but the percen o geese, the mut ow transverse rid roof of the mouth are sufficient to t
lly expensive ine aracteristic head ans. However, the nt of papillae in t
sses found in mu 89) and Bewick’s
n of the vegetatio f grass were also chanism, which is sed for the trans thin the bills and ped on tall turfs w elatively high com
g of grass per bit dy (ln loss = 1.71 *
stead of 40%. The be characteristic uring straining th generate a water at there is a trade
inertial transport of straining becau y’ pump.
on bite size but a size and bite rate aken by the mute e barnacle geese
s of barnacle gee e rate found in o Van der Graaf et ed and barnacle
nstead of other le of food. In spite o
examined spilt s hether grazing in e swans spill on a ntage food lost va te swan has relat
ges instead of pa h less rough than transport clipped rtial transport m d and jaw movem
e absence of pap he posterior half ute swans. Under swans (Rees in S on cropped.
reported for the s normally emplo sport of grass. Fo d are lost for inge was found by Va mpared to the los te. From the relat
* bite size – 5.74 e absence of pap c for straining spe
e tongue is press rflow through the e-off between st t mechanism for use the tongue do
also on bite rate.
e, but this relatio e swans are colle collect much sm se in other studie our study (Durant
al. (2006): 60/mi geese no correla
eaves) may decre of the papillae on some of the leave field conditions i average about 30 aries among ively few papillae apillae in the rost n in geese, but d grass leaves wit
echanism of the ments of the ‘catc pillae in the anter f may contribute r field
ears 1989) were
e mallard. The ine oyed by the malla llowing a ‘throw’
stion. A loss of 4 n der Leeuwet a sses reported in t tionship between 3), one would ex pillae on the roof ecies (Van der Le sed against the ro e beak. The differ
raining and grazi grass, while a ro oes not properly
Our data sugges onship seems to d ected at a relative aller bites at a hi es on swards of t et al. (2003): 89
in on a narrow le ation between bit
ease, n the es
in 0% of e in tral h the ch and rior
to the also
ertial ard to
’ large 0 l.
the n bite xpect a
of the eeuw
oof of rent ng. A ugh fit the
t a differ ely
gh 9/min;
eaved te
rat sho inc acr spe sm in of ch spe rat im ins co Wh als rat wh ge sho Th ma an rat all po tim rel
intakerate(mg/min)
te and bite size w ow a negative co creasing bite size ross species (exp ecies does have c mall bite sizes, the bite size, and int bite size with inc anges much less ecies in our study te. A more accura portance in thes stantaneous bite
rrelation betwee hen the relations so applied to less te becomes highe hat one would ex
nerate more forc ould go down, an e force required ammalian grazers
d transport time tes in these anim grazing anatid sp ositively related to me seems to be in lationship as foun
0 200
0 3000 6000 9000 12000 15000
intake rate (mg/min)
*
+
was found. In Can orrelation. In thes (exponent -0.77 ponent -0.336). Th
consequences fo e decrease in bite
ake rate levels of creasing body siz
at large bite size y variation in bite ate control of bit
e species. An acc rate remains con en bite rate and b
ship between bite er white-fronted er for the same b xpect for species
ce than smaller s nd bite rate shou to sever a bite is s. Therefore, cro is proportional t mals (Spalinger an pecies. In mute s o bite size. In sm ndependent of b nd in the mute sw
400 600 800
bite size (mg)
nada geese and th se two species bi 78) then one wou he fast decrease r intake rate. Wh e rate at large bit ff (figure 5.4). Th e and average bit es than at small b
e size would resu te size, close to it curate control of nstant for all spe bite size in the tw
e rate and bite si d and barnacle go bite size as specie of varying body s pecies to crop a
ld go up.
s not considered pping time is ass to bite size in the nd Hobbs, 1992).
wans cropping ti all geese, like in t ite size, although wans.
Figure5 calculat loss as f maxima Average symbols and A.e rate. In change Legend:
Anserer
1000
he mute swans b te rate drops mu uld expect from th
in bite rate with hile intake rate in te sizes starts to b
is may explain th te size (e.g., mut ite sizes. For the ult in relatively la
ts maximum, may bite size while th cies would expla wo smallest specie
ze for Canada ge oose data, the res es get larger. This size. As larger spe bite of the same
to constrain inta umed to be inde e mechanistic mo These assumptio me and transpor the Spalinger-Ho h individual birds
5.4. Intake rate as f ted from bite rate a found in Muteswa al intake rate and c e bite sizes per spe s. Small changes in erythropus results B.canadensis, and in bite size affects : : Cygnusolor, * rythropus, +: Brant
bite size and bite uch faster with
he same relation bite size within ncreases quickly a
balance the incre he increase in var te swan). Intake r two smaller goo rge changes in in y be of more he variation in
in the absence o es.
eese and mute sw sults show that b s is consistent wit ecies are able to
size, cropping tim
ke rate in large pendent of bite s odel describing in
ons may not hold rting time are obbs model, crop
may show the sa
function of bite siz and bite size minus an. Lines indicate
corresponding bite ecies are indicated n bite size in B.leuc in large changes in d in particular in C.
intake rate very lit : Brantacanadensi taleucopsis.
rate ship at ease riation rate ose ntake
f a
wan is bite
th me
size take d for
ping ame
e, s food
size.
by copsis
intake olor a ttle.
is, {:
Bit
Lar bit Alt litt gra pro bo Lan be for Da the de Su sw sw Ha inc de an qu qu en Ba su (Ae Va ge ge ge ter ov ma ov en for oth un bre (Pr en Pro to
tedepthanddai rger bite sizes ma te. We did not fin though bite dept tle among the thr ass leaves. On th oportion of the le ody mass. The sm
ng et al. ,2001 an related to the am raging strategy.
aily food intake is e gut, and quality
creases from the mmers and Critc ward (Delagardee ward declines with
assall et al., 2001;
creases tensile st cline in digestibil d thus to body m ality of vegetable ality of food may abling them to in rnacle geese are perficial layer of erts et al., 1996;
n der Graaf, 200 ese, but clip a lar ese. Apart from m ese discussed ab rm foraging strat er rate may cons aximize the long- er rate in the gut ergy maximizers rage long enough her activities con gulate foraging, eeding season, w rop, 2004). In add able them to fee op and Vulink, 19
time-minimizers
ilyenergyrequire ay be the result o nd a clear relation h roughly increas ree goose specie e other hand less ength of the gras mall bite depth of nd Therkildsen an mount of force re
s limited by eithe y of the food is of e top to the base hley, 1990), whil et al., 2000). In ad
h increasing swar
; Summers and C rength of leafs, r lity. As digestibili mass (Bruinzeel et e food (e.g. Allpo y be of less impo ngest about 75%
the smallest ana natural swards, w Durant et al., 200 6). Lesser white- rger proportion o mechanical diffe bove, this differen
egies. While plan strain intake on t -term rate of ene t constrains long- on the one hand h to obtain energ ntributing to an in
in geese this phe where individual g dition, geese are ed on low digestib 992) and thereby s. In our study, ba
ements
of larger bite dep nship between b ses with body siz s, and only mute ser white-fronted ss leaves than ba
the barnacle gee nd Madsen, 1999 equired to severe
r the short-term f primary import of a sward (Dela e fibre content in ddition, several s rd height (e.g. Bo Critchley, 1990).N
requiring a greate ity is positively re t al., 1997), large ort, 1991) than sm
rtance than for t of the length of atid species in thi which contain th 04; Loonen and B fronted geese ar of the length of th
rences in the hea nce in short-term nt biomass may c he long-term. It i ergy intake (Steph
-term intake. The d and time-minim gy requirements a ncrease in fitness enomenon has be geese consistentl shown to be abl ble food in summ y revert their fora arnacle geese ma
pth and/or severi ite depth and bil ze across species, e swans take a cle
d geese seem to rnacle geese, wh ese is very similar 9. These differenc e a bite, but also
rate of intake or ance. Protein/nu agarde et al.; 200 ncreases towards studies have show os et al., 2005; Du Not only does hig er effort for seve elated to length o e Anatidae are les mall species. For
he three smaller 8 cm tall sods.
is study and they e least fibre and Bos, 2003; Prins a re similar in body
he grass leafs com ad-neck apparatu m rate of food inta constrain short-te is assumed that a hen and Krebs, 19 e two extremes a mizers on the oth and thereby free s. Although this t een observed as ly follow one of t e to prolong rete mer (Prins et al., 1 aging strategy fro ay behave like en
ng more leaves p l (or body) size.
, it varies relative early larger part o
clip a larger hich have a simila
r to values report ces in bite depth to food quality a
r the turn-over ra utrient content
0, Duru, 2003;
s the base of the wn that quality o urant et al., 2004 gh fibre content
rance, it also cau of the digestive sy ss susceptible to
the large mute sw goose species, y exploit only the
the most nutrien and Ydenberg, 19 y mass to barnacl
mpared to barna us in the barnacle ake may reflect lo erm intake rate, t animals strive to 986), and that tu are represented b er hand, which o time to devote t heory is based on well during the the two strategie ention times, whi 1981; Prop et al., om energy-maxim
ergy-maximizers per ely
of the r ted by
may and
ate in
f the 4;
uses a ystem the wan
nts 985;
e cle e
ong- turn- urn
by only
to n es ich 2005;
mizers ,