The bill of evolution : trophic adaptations in anseriform birds Kurk, C.D.

(1)

The bill of evolution : trophic adaptations in anseriform 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).

(2)

Filter-f

C

feeding p (

Chapte

performa (Anatida

er 6

ance in w e)

wildfowl

(3)

112 Wi mo for req sho aq du are pis an filt mu Filt as fee of pe du dif pe the filt lin res ton Th res fee

2

ildfowl (Anatidae orphology and th raging in differen quire different w own that speciali uatic feeding spe cks suggest that e necessary for th ston function of t

d filter-feeding. T ter-feeding perfo ute swan and com

ter-feeding perfo by the amount o eding ducks, the

the millet seeds r straining cycle cks than in grazin fference in volum

rformance seem eir elevated tong tered out food pa gual base. In graz sults in a less effi ngue in the same e results on filter sults on grazing p eding and grazing

e) exploit many d he exploitation of nt physical enviro ways of handling f ized grazing wild ecies. Morpholog the presence of he effective intra the tongue durin To demonstrate t ormance we inves mpared the resul ormance is deter of water and susp

goose species an drawn in at the t relative to body s ng wildfowl. Diffe me taken in per m s to be related to gue in a piston-lik

articles are trans zing geese the to

cient intake of w e way as vegetati r-feeding perform performance clea g in wildfowl.

Summary

ifferent food sou f food resources onments (aquatic food items. In a p fowl have a high gical and biomech

spines on the inn a-oral transport o g filter-feeding, r that high grazing stigated filter-fee lts to previous re mined by the per pended food part nd the mute swan tip of the bill. On

size is larger and erences in relativ movement cycle. T

o the different m ke way to genera ported simultane ongue is not eleva water, and seeds a on during grazing mance in the pre arly indicate a fun

urces. The relatio is poorly underst c versus terrestria previous study (ch er performance f hanical analysis o ner surface of the of vegetation, is i

resulting in a trad g performance is

eding of two goos sults on ducks.

rcentage of food ticles pumped th n are all able to r the other hand, the straining fre ve bill size only ex

The major cause mechanisms used.

te a flow of wate eously alongside

ated during filter are retained and g.

sent study in com nctional trade-off

Cha

onship between b tood. In particula al) may be expect hapter 5) we hav for grazing than of feeding in gees

e upper beak, wh ncompatible with de-off between g

associated with l se species and th

-items retained a rough the bill. Fi retain more than

the volume of w equency is higher xplain part of the

for the differenc . Duck species us er through the bil

the tongue to th r-feeding, which transported ove

mbination with th f between filter-

apter 6

bill ar,

ted to ve

se and hich

h the grazing

low he

as well lter-

95%

water r in e ce in

e ll and

e er the

he

(4)

Dif lim lin (Sc me is a Ru Fo the fitn suc Lac mi fro stu du sel fin we A m Ko tha tha lim (G for Wh the wa sep dir row Th fac su lam ge al., be wa gra

fferences in trop mited resources. A

ked to alternativ chluter, 2000b). W easures of fitness also open to que ubega, 2000).

r divergent selec e exploitation of ness) on alternat ch a trade-off mi ck (1971, 1974) s ght lead to partit om the water, by udies have docum

cks (Nummi, 199 lective uptake to

d such a relation ere more importa mechanical analy ooloos et al., 1989

at the separation an interlamellar d mit on the size of

urd, 2006). Howe raging performan hen ducks are ab ere is no clear be ater pumped thro

paration (i.e. the rectly depend on

w) (Kooloos et al e relationship be ct that terrestrial bmerged plants a mellae these are

ese and swans ge , 2003) in geese h tween the tongu ay of transporting ass. Although it m

hic morphologies At the population ve morphologies m

While it is difficu s, often the exact

stion (Arnold, 19 ction to operate m

one resource sho tive resources (Sc

ght occur.

suggested that di tioning of resour means of the co mented interspec 93; Nudds and Bo interspecific var nship (Nummi and

ant than lamellar ysis of the jaw ap 9) showed that d n between upper distance. With su the food retaine ever, this flexible nce, which is nec ble to set the low enefit for a wide i ough the bill per distance betwee

interlamellar spa ., 1989; Gurd, 20 etween morpholo

grazers like gees are also able to ‘f apparently not u enerate a water f has shown that f ue and palate and

g seeds collected may proof difficu

Introduction

s among species n level difference

may lead to dive lt to relate differ t relationship bet 983; Wake, 1992;

morphological ad ould decrease fe chluter, 2000a). H

fferences in bill m ces by selectively omb-like lamellae

cific differences i owlby, 1984; Guil riation in interlam

d Väänänen, 200 r density (Lagerqu

paratus and filte ucks can move th

and lower bill la uch an adjustable d, largely indepe e filter mechanism

essary for resour er limit of the ra interlamellar spa movement cycle en lamellae on th

acing (i.e. the dis 005).

ogy and particle r se, and aquatic g filter-feed’. Altho used to retain foo flow through the ood particles pas d are not transpo d during filter-fee lt to demonstrate

n

are believed to r es in the exploitat

rgent selection a ences in feeding tween trophic tra

Wainwright, 199 daptations that o

eding performan However, it is not

morphology amo y sieving differen e on the margins n the size of food llemain et al., 200 mellar spacing. Ot 01) or found that

uist and Ankney, er-feeding proces

he upper and low mellae during filt e filter ducks are endent of mean in m alone does not

rce partitioning to nge of food item acing. On the othe e decreases with i

he maxilla and ma stance between t

retention is furth grazers like swans ough geese and s od items. While, a e bill, X-ray analys ss to the oesopha orted along the m eding is very simil

e differences in f

reduce competiti tion of resources and adaptive radi performance to aits and recourse 91; Irschick, 2002 ffer an advantag nce (and ultimate t always apparen

ong filter-feeding nt sized food part of the bill. Some d items ingested

02) and related t ther studies did n

bill size and shap 1989).

ss (Zweers et al., wer bill in such a ter-feeding is larg

able to set a low nterlamellar dista t cause a trade-of

o evolve.

s that can be ret er hand the amo increasing lamell andible) but doe the lamellae with

her complicated b s feeding on swans do possess

as in dabbling du sis (Van der Leeu agus squeezed margins of the bill

lar to the transpo feeding performa

on for s

ation e use 2;

e in ely nt how

ducks ticles

by his not pe

1977;

way ger wer

ance ff in ained unt of lar

s not hin a

by the s ucks, uwet

l. This ort of ance

(5)

114 am be Tw Ko bil on lat the wa Th gra clip ton foo ch In to mo ov du wi we pe res

Exp

Filt ge pu Tw ind wa co an Bo av bir

4

mong dabbling du tween such dive wo morphological ooloos et al., 1989

l enables the ton e-way flow of wa teral sides of the

e margins of the ater and small foo

ese two characte azers have many pped vegetation ngue movements od items to the o apter 4).

a previous study metabolic weigh orphological and er-tongue transp ring filter-feedin ll exist. We hypo ell during filter-fe rformance of filt sults with previou

perimentalsetu ter-feeding trials ese (Brantaleuco rchased from a l wo mute swans an

doors aviary whe as held in an outd

nditions. When n d waterfowl pell oth training and e

iary, allowing bir rds were trained

ucks, one would e rse trophic group l traits seem to b 9; Van der Leeuw ngue to act as a p ater from the ant posterior part of bill to pass to the od items.

ers are clearly dif caudally pointin that is carried ba s during grazing.

oesophagus durin

we showed that ht is higher in gra biomechanical a port of grass is inc

g and that a trad thesize that both eeding than duck er-feeding in two us work on filter-

Ma

up

s were conducted opsis) and two le ocal trader.

nd the lesser whi ere birds had cont doors aviary and not engaged in fil ets) was availabl experimental tria ds to be tested in for 1-3 weeks to

expect a clear tra ps as anseriform be closely linked t w et al., 2003). Fir piston within the terior of the bill t f the tongue is th e oesophagus an fferent in grazing g papillae on the ackwards over th The lateral lingu ng filter-feeding a

t grazing perform zing geese and m analysis of feedin

compatible with de-off between gr h goose species a

species. In the p o goose species a -feeding anatinid

aterialsandMe

d with three mute sser white-fronte ite-fronted and b tinuous access to trials were cond lter-feeding trials e.

ls were performe ndividually. Imme get accustomed

ade-off in feeding grazers and filter to filter-feeding ( rst, a bald inner s slightly opened b to the posterior.

hought to allow fo d enables a cont g species. Instead e inside of the up he tongue with a al slits that allow are absent in graz

mance measured a mute swans than

g in geese and du the piston functi razing and filter-f and the mute swa present study, we and the mute swa ds (Kooloos et al.,

ethods

e swans (Cygnus

ed geese (Ansere barnacle geese w o a small pond. T ucted under othe s, ad libitum food ed within an encl ediately prior to to filter-feed fro

Cha

g performance r-feeding ducks.

(Zweers et al., 19 surface of the up beak and to gene Second, a slit on ood items diverte

inuous throughp d of a bald surface

per bill, which re series of rostro-c w the transport of zing species (see

as intake rate sca in 2 duck species ucks suggest that ion of the tongue feeding performa an will perform le e therefore asses

an, and compare , 1989).

olor), two barna erythropus), all were tested in an

he third mute sw erwise similar d (mixture of grai

losure inside the the experiments om a small tray.

apter 6

977;

per erate a

the ed to put of

e etain

caudal f small also

aled s. The t the e

ance ess

the e the

cle

wan ins

s, the

(6)

Th fol ge filt en 5 s a r foo To da Th an (fig str wit Du ma 1b Du co fee the ad pla the an

Fig fee exp me

e evening before lowing day betw ntly guided to th ter-feed from the closure. Birds we seconds were nee rate of one to thr

od was returned be able to comp ta on duck specie is tray was specif d the amount of gure 6.1). Part 1a rain only from pa th it, and is collec uring each feedin

ade. To avoid blu could be mecha uring the training ntinuous filter-fe eding instead of p e broader-billed s justed to secure aced underneath e geese, the relat

d adjustments to

gure6.1. Schematic eding experiments.

pelled water (and s echanical adaptor.

1a 1b

e a trial, food was ween 08:00 h and e enclosure and e tray until it raise ere very fast in re

eded to empty th ree trials a day. T

to the birds.

pare filter-feeding es collected by K fically designed t water expelled a a and 1b serve as rt 1b. Expelled w cted into a tray ( g trial high-speed urring of the video nically adjusted t sessions of the g eeding was perfo

pecking, while at swans a simple r a similar position the tray to colle tive length of the o the tray of the s

c drawing of upper . 1a: reservoir, 1b:

seeds) roll down to

b 2

s removed from t 09:00 h. About f allowed to settle ed its head, after emoving the seed he tray. About fif Trials usually finis

g performance as ooloos et al. (198 to measure the a along the sides of s storage for wate water flows down

3).

d video-recording o-recordings as a to minimize the m geese we determ

rmed. This level f t the same time s

ectangular tray w n of the bill as in

ct expelled wate e bill that was ins swans were mad

r and side-view of e part where birds i o 3: collection tray.

1a 1

2

3

the aviaries. Tria five minutes befo e down. The anim r which it was gu ds from the wate fteen trials per bi hed before the a

s measured in th 89) we used the s

mount of seeds r f the bill. The tra er and food. A bi n a slope (2) carry

gs (50 fr/s) from a result of large h movement space mined the water le

forced the birds scooping out wat was used in which

the goose trials.

er and seeds. From serted into the w

e to achieve a sim

experimental tray u immerse bill tip, 2:

Part 1b can be red 1b

ls started the ore a trial a bird w mal was allowed t

ided out of the er, on average les

rd were conduct afternoon, after w

e present study w same tray and se retained by the b y consists of four rd was allowed t ying lost seeds alo

a lateral view we head movements e for the bill tips.

evel at which to forage by filte ter was prevented

h small plates co A plastic sheet w m video-recordin

ater was determ milar water level

used in the goose f slope along which duced in size by a

was to ss than ted at which

with et-up.

bird r parts

o ong

ere s, part

er- d. For uld be was ngs of

ined .

filter- h

(7)

116 Me

As see co fee ou de Se fee de Fra op use 1) 2) 3) 4)

Sta

To an ne ind sig spe Lev tes

6

easurements

in the filter-feed eds was suspend ncentration of m eding trial the ex tside the tray we termined by rew eds lost during fi eding tray were f termined. The se ame by frame rep penings and closin ed to characteris

the amount of w per cycle (pump the amount of the amount of entered the mo the amount of intake).

atisticalanalyses assess differenc d seeds entering sted ANOVA pro dividuals were ra gnificant (p < 0.05

ecies. When data vene’s test) value st was used when

ding experiments ded in 70 ml wate millet seeds was o pelled water from ere wiped with a weighing the tissu lter-feeding were filtered from the eeds were left to play of video reco ngs and the exact se filter-feeding p water pumped th p-performance),

water swallowed seeds retained b outh, expressed a seeds filtered me

s

es between spec g the bill per seco cedures (SPSS 12 ndom variables n 5), post-hoc tests a were not norma es were ln-transf n variances were

s of Kooloos et al er for the goose t offered in 200 ml m the collection t

tissue of known ue.

e collected and c water and the am dry at room tem ordings were use t duration of a fe performance:

hrough the beak d, expressed as p by the filter relati as percentage (fil easured as gram

cies in filter-feedi ond and per pump

2.0). Species were nested within spe s were performed ally distributed, o formed and subse still not equal.

. (1989), 1 gram ( trials. In the swan water. Immediat tray was weighed

weight and their counted. The see mount of water t mperature overnig

ed to count the n eeding trial. Seve measured as mil percentage per cy ve to the amoun ter-performance per second and

ing performance, ping cycle were u e considered as f ecies. When F-va d to attribute diff or had unequal v

equently analyse

Cha

(dry weight) of m n trials a similar

tely following a d. Spilled drops r weight was

ds remaining in t that remained wa

ght and then wei umber of beak ral measures we lilitre per second ycle,

t of seeds that e), and

per cycle (seed

, the amount of w used as input for fixed effect and alues proved to b

ferences to speci variances (tested ed. The Games-Ho

apter 6

millet

the as ighed.

re d and

water

e ific

with owell

(8)

Pu Th str or giv can wa go fro AN (F2

sig As we the sm Th spe 1.3 dif As thr

= 7 Th by sw

umpperformance e amount of wat raining movemen per movement c ven in table 6.1, t n be estimated b ater pumped thro

ose species. The onted geese pum NOVA shows that

2,4.042 = 72.713, p gnificantly from e there is a large d e scaled volume p e differences in t maller, and are no e frequency with ecies (table 6.1).

382, p = 0.246), b fferent from each the differences i rough the bill per 70.785, p = 0.001 e small amounts the birds (table wallowed (F4.01, 2 = e

ter that is expelle nts may be expre cycle. Values for v together with lite by multiplying the ough the bill per

difference betwe mps just slightly m

t species differ sig

= 0.001). Post-ho each other (all p <

difference in bod pumped per cycle

he amount of wa o longer significan h which the bill o

Individuals withi but the small diffe

h other (F2,4.152 = 9 in frequency amo r second (water p 1), and all species

of water lost du 6.1). There is no

= 0.339, p = 0.73)

Results

ed at the caudal r ssed as the amou volume per cycle erature data on d e frequency with cycle is 5 times la een the two goos more water per cy

gnificantly more oc tests (Bonferro

< 0.01).

y size between t e to body size (ta ater pumped thro nt (F4.042, 2 = 1.778 pens and closes ( in species do not erences in freque 92.493, p < 0.001 ong species are s pumped per cycle s differ among ea

ring each trial ar significant differ among species.

rims of the bills d unt of water pum e for the goose an duck species. Volu the volume per c arger in mute sw se species is sma ycle than the barn

than individuals oni) show that al

he two goose spe able 6.1). After ac

ough per cycle ha 8, p = 0.279).

(11-14 Hz) is rela t differ in strainin ency among spec 1).

small the volume e x frequency) is ach other.

e considered to h ence in the perce

during a series of mped per unit of t

nd swan species a ume per unit of t cycle. The volum wans than in the t all: the lesser whi nacle geese. A ne

within a species l species differ

ecies and mute s ccounting for bod ave become muc

atively constant w ng frequency (F4,9

cies are significan of water pumpe also significant (

have been swallo entage of water

cyclic time are time

e of two

te- ested

swans, dy size ch

within

92 = ntly ed

F2,4.05

owed

(9)

Table6.1. Parameters of filter-feeding performance expressed as averages with standard deviations. N+

Body mass (kg)#Gape (mm) Freq. (Hz)

Vol. per cycle (ml)

Vol. per cycle relative to body weight (ml/kg) Percentage of water swallowed (%) Filter performance (%)

No. seeds per ml

Seeds per cycle and body weight (mg/kg) Rate of seed ingestion (mg/s)

Seeds ingested per metabolic weight and unit of time (mg/kg0.75 .s) Lesser white- fronted goose

43 2.07 12.0 ± 1.0 (n = 6) 13.5 ±0.6 0.26 ±0.06 0.13 ± 0.038.41 ± 3.6499.7 ±1.49.7 ±2.97.6 ± 2.0211.9 ± 53.4122.2 ± 31.0 Barnacle goose 30 2.14 14.6 ± 1.7 (n = 12) 12.4 ±0.4 0.22 ±0.04 0.10 ± 0.0210.10 ± 3.84 99.9 ± 0.313.5 ±3.88.8 ± 2.2233.2 ± 56.5131.8 ± 31.9 Mute swan 26 8.67 18.4 ± 0.7 (n = 7)

11.2 ±0.5 1.24 ±0.28 0.14 ± 0.039.25 ± 5.7799.4 ± 0.96.4 ±2.35.6 ± 1.9543.9 ± 178.2107.65 ± 35.3 Literature data* Mallard1.04 5.3 ± 1.518.0 0.58 ± 0.13 (n = 49) 0.57 ± 0.1395.2 ± 3.3 (n = 36) 5.4 ±1.98.9$ ± 2.3 (n = 49) 167 ± 43.6 (n = 49) 162.5 ± 42.3 (n = 49) Wigeon 13 0.63 ? 22.0 0.42 ± 0.11 0.66 ± 0.175.41 ± 1.82 (n = 14) 98.3 ± 2.1 (n = 14) 6.1 ± ?? Tufted duck 0.77 5.0 20.0 0.60 ± 0.27 (n = 34) 0.77 ± 0.3596.4 ± 3.0 (n = 22) 7.2 ±3.415.2$ ± 5.7 (n = 34) 233 ± 87.2 (n = 34) 283.5 ± 106.1 (n = 34) Northern shoveler0.64 5.5 ± 1.013.0 0.63 ± 0.21 (n = 51) 0.99 ± 0.3392.0 ± 5.9 (n = 33) 5.1 ±2.315.1$ ± 7.1 (n = 51)

125 ± 59.3 (n = 51) 175 ± 82.9 (n = 51) +: number of trials, #: weight of birds from literature, *: Kooloos et al. 1989, van der Leeuw et al. 2003, $: seed mass (3.07 mg, av. diam. = 1.8 mm) estimated from reported diameter range and density calculated from data (6.4 mg; av diam = 2.3 mm).

(10)

Fil Th am wit sec tha be 0.0 wit To sec sm fre res

In filt co spe sam sho spe eff int int et to 19 tha sw Th (Ta Gr Nu Tré to Alt the

terperformance e filter performa mount of millet se

th the water exp cond are listed in an the two goose tween the mute 071). This is due t thin species (F4,92

account for the cond to metabol maller and are not

equency are smal sults for the amo

the present stud ter-feeders, the l mpare the result ecialized filter-fe me high efficienc oveler (Anasclyp ecies and mute s ficiency with whi terlamellar distan terlamellar distan al., 2003). The m retain approxima 89). The white-fr an 1.2 mm and e wallowed togethe e sizes of seeds m aylor, 1978; Dirsc

uenhagen and Fr ummi, 1993; Bald éca, 1986; Silveir 5 mm and thus v though no data is ey would not con

eandseedintake ance, i.e., the amo

eeds sucked in, is elled (table 6.1).

n table 6.1. Per cy e species. Howev

swan and the tw to relatively large

2 = 18.174, p < 0.

effect of body m ic weight (M^0.75).

t significant (F4.01

ll the results for t ount of seeds per

y we assess the p esser white-front ts with previous s eeders are able to cy as dabbling du peata) retains 90- swan scored over

ch food items are nce. The smallest nce (Kooloos et a mallard (Anasplat ately 60% of seed ronted goose (An

xpel water from er with 74% of the

most commonly e chl, 1969; Euliss a redrickson, 1990;

dwin and Lovvorn ra, 1998; Green e very similar to the

s available on gee nsume these seed e

ount of millet see s very high for all

The amounts of ycle swans draw ver, a nested ANO wo goose species

e variation amon 001).

mass we scaled th Clearly, the diffe

12, 2 = 0.518, p = 0 the amount of se

cycle.

Discussion

performance of f ted goose, the ba studies in duck sp o retain seeds wit cks. While a filte -100% of the see r 99% on the sam e retained depen t food items are r al., 1989; Mott, 19

tyrhynchos) and t ds with a diamet nseralbifrons) is u

the bill at the sa e water entering exploited by filte and Harris, 1987;

; Afton et al., 199 n, 1994; Rogers a et al., 2002; Guille

e size range gees ese foraging on s ds when available

eds retained as p species. Less tha seed retained pe in about 3 times OVA showed that is not significant g individuals, wh

e amount of seed erences between 0.631). As the diff eeds per second a

filter-feeding in t arnacle goose an pecies. The data th a diameter of er-feeding special eds (Kooloos et al me seed in our stu nds on the size of retained by speci 994; Guillemain e tufted duck (Ayth er of 0.7 – 1.2 m unable to retain f me time. Such sm

the bill (Van der er-feeding duck s Gammonley and 91; Marchant and nd Korschgen, 19 emain et al., 2002 se are able to ret seeds in water it e (Sedinger, pers

percentage of the an 1% is lost alon er cycle and per

more millet seed t this difference t (F2,4.011 = 5.516, hich differ signific

ds ingested per n species only bec

ferences in strain are very similar t

hree non-special nd the mute swan

show that non- 2.3 ± 0.2 mm wit list like the north l., 1989), the goo udy. In ducks the f the food item a ies with the smal et al., 2002; Figue hyafuligula) are

m (Kooloos et al.

food items small mall food items a r Leeuw et al., 20 pecies in the field d Heitmeyer, 199 d Higgins, 1993;

996; Petrie, 1996 2) are in the rang ain effectively.

is highly unlikely s. comm.).

e ng

ds p = cantly

come ning

o the

ized n, and

th the hern ose

nd on lest erola able .,

er re 03).

d 90;

6;

ge of 1 that

(11)

120 Th

Th be al., use X-r ton foo an in bil Alt 19 op the int wa ton the pre so- Du po lam Du bil fee are int To the scr lin to

0

emechanism

e high efficiency cause a mechani , 1977; Kooloos e e very different t ray film recording ngue (see below) od follows a path d, during strainin the way food is t l opening and clo though tongue m 89; Zweers et al.

pens, the tongue e ventral side of t to an anterior and ater and food item

ngue protracts an e bulges to the b evented from en -called ‘lingual cu uring the next mo osterior bill cavity mellae (figure 6.2 ucks, as most bird l in such a way th eding is greater t e able to set a low terlamellar distan

be able to conti e rims of the bill.

rapers lining the gual cushion. Thi stop and swallow

with which relat ical analysis of th et al., 1989; Van d techniques to ret

gs of filter-feedin ) and than swallo h through the ora ng, transported a transported is ass osing.

movements are di , 1977) suggest t is retracted while the upper bill. In d a posterior sec ms into the anter nd the lingual bu

ack of the tongue tering the oesop ushion’), which re ovement cycle wh y is reduced and w 2).

ds, have a movab hat the separatio han interlamella wer limit on the s nce (Gurd, 2006).

nue filter-feeding A further transp tongue, which m is enables filter-f w.

tively large food i he jaw apparatus der Leeuw et al., tain seeds. Kinem ng show that in g owed, while in th al cavity to the rim alongside the ton

sociated with diff ifficult to analyse the following sce

e part of the tong this position the tion. The coordin rior bill cavity. W

lges are depresse e. The water that hagus by the ele emains elevated hen the tongue r water is forced o

ble upper jaw and on between uppe r distance (figure size of the food r .

g food must be tr port to the oesop move food items t feeding ducks to

items are retaine and filter-feedin 2003) suggests t matical analysis of eese seeds are tr e mallard and wi ms of the bill whe

gue to the oesop ferences in tongu e kinematical stud nario for filter-fe gue (lingual bulge e lingual bulges d

nated action of to When the bills star

ed, forcing water t is transported b vated posterior p throughout succ retracts again the ut through the sp

d are able to mov er and lower bill la

e 6.3). With this a retained that is la ransported away hagus is mediate through a groove continue feeding

Cha

ed is remarkable ng process (Zwee that ducks and ge f high speed vide ransported over t igeon (Anaspene ere seeds are ret phagus. This diffe ue movement du dies (Kooloos et a eeding. When the

es) is elevated ag ivide the oral cav ongue and bills d rt to close again t r and food items backwards is part of the tongu cessive pump-cyc e volume of the

pace between th

ve the upper and amellae during fi adjustable filter d arger than y from the filter a ed by spines and e along the side o g without the nec

apter 6

rs et eese eo and

the elope) tained erence uring

al., e bill gainst vity

raws the

over ue (the cles.

he

lower ilter- ducks

area at of the

cessity

(12)

ent ling Foo the cyc

Fig fro and are Mo

ter the oral cavity a gual surface. Durin od items on top of e oral cavity. As th cles, in which the c

gure6.3. Cross-sect om the water flow.

d the ventral mand e retained between odified after Kooloo

Specialized fil feeder

and food items fro ng closing of the bil the lingual surface e cushion remains ushion is depresse

tion of bills and ton Indicated are the m dibular lamellae (C) n the inner surface

os et al. (1989).

mand A

B

C ter-

bill opening

bill closing

om a previous cycle ls, the tongue prot e are retained by th

elevated, several c ed and food items s

ngue of the mallard maxillary lamellae ). Keratin elements of the upper bill a

tongue

dible

maxilla Grazer

Figure6.2. Sc of movement bills of the filt of specialized grazing (right) Upper figures lower figures:

(indicated by species). Arro indicate direc tongues.

Filter-feeding bill coincides w lingual bulges tongue) and e (caudal part o retraction of t food are draw from a previo bulges and cu During closing depressed and moving under food items.

Grazing specie tongue retrac but elevated c e are carried backw tracts with elevate he caudally directe collection cycles ar swallowed.

d at the level wher (A), the dorsal man s are indicated by b and the dorsal man

hematic represent cycles of tongue a ter-feeding mechan

filter-feeding (left ) wildfowl species.

: opening of the bi : closing of the bill arrows in between ws underneath bil tion of movement

species: opening o with elevation of t s (rostral part of the elevated lingual cus of the tongue) and

the tongue. Water wn into the bill, and us cycle (in betwee shion) is expelled.

g the lingual bulges d the tongue protr rneath the water a

es: during opening cts with depressed cushion. Water and wards on top of the

d bulges and cushi ed spines on the ce re followed by tran

re food items are fi ndibular lamellae ( black areas. Food it ndibular lamellae (a

tation and

nisms ) and

ll,

n ls

of the

of the he e shion

and d water

en the

s are racts, nd

the bulges d food e

ion.

eiling of nsport

iltered (B),

tems arrow).

(13)

122 Alt an the sim 20 fee do ph Du ret an ex the oe ele div are Du bil wit spi ret cu ele to

Wa

Pa ma the err Alt Pe sho Up de the ava rel Cy

Th les mu

2

though the exact d the mute swan e upper beak and milar to the one d 03). The mechan eding mechanism omestic goose typ

ase.

uring the collectio traction of the to d food items ent pulsion of water e elevated lingua sophagus. During evated, while wat verted to the side e retained by pre uring the transpo l and tongue. Du th bill opening, a ines on the uppe traction and dep shion to the oeso evated during the the one used du

aterlost

rt of the water p ay represent a tru e bill and lost for rors due to weigh ternatively, a sign rformance exper ows that only aft p to this age gosli r Leeuwet al., 20 e present experim ailable seems to lated to the use o yclevolumeandb e amount of wat sser white-fronte ute swan. Howev

t mechanism of fi n has not been st d the absence of described for the nism of filter-feed m in ducks. Instea

pically has two se on phase, openin ongue, but in con ter the bill and at from the preced al cushion may se g closing of the b ter and food item e of the bill but fo essing them again rt phase, there is ring the transpor and when the ton er bill. The food is ression when the ophagus, the ling e tongue protract

ring grazing (Van

rovided to the bi ue loss. In most t r collection after t

hing very small a nificant amount o riments on filter- ter 4 weeks of ag ngs still swallow 003). It is not clea ments is actually

ingest less water of the ‘under-ton billsize

ter pumped throu ed goose and barn ver, the mute swa

ilter-feeding in th udied, the prese a lingual groove domestic goose ding in the dome ad of a continuou eparate phases, a ng of the bill occu trast to the malla t the same time t ding movement cy erve to prevent w bills, the tongue p ms are transporte ollow a more me nst the spines on s a shift in phase rt phase protract ngue moves forw s transported furt e bills are closed.

gual cushion is de tion (figure 6.2).

n der Leeuw et al

irds is not recove trials some drops the end of a filte mounts of water of water may hav

feeding in an ont ge the adult filter- 83% of the wate ar whether the to

ingested, but the r than the geese ngue’ transport m

ugh the bill per m nacle goose is ve an is 4 times larg

he two goose spe nce of spines on suggests that the (Anseranser; Va stic goose is diffe us process, filter-f a collection phase urs simultaneousl ard the bulges ar the large lingual r ycle at the rims o water from runnin protracts and the ed over the tongu edial course over

the inside of the between the mo tion of the elevat ard the food item ther backward d To transport foo epressed during t This transport m ., 2003).

ered after the tria s of water were v

r-feeding trial. A r may accumulate

ve been swallowe togenetic series o -feeding mechan er along with the otal amount of w e only duck spec

and swan (table mechanism in stra

movement cycle o ery similar, but m

er than the goos

Cha

ecies used in this the inner surface ey use a mechan an der Leeuw et a

erent from the fil feeding in the e and a transport ly with a large re depressed. Wa retraction causes of the bill. As in d ng into the e lingual bulges a

ue. Food items ar the lingual bulge e upper bill.

ovement cycles o ted tongue coinci ms are held in pla

uring tongue od over the lingu tongue retraction mechanism is iden

al. Some of this w vigorously shaken lso measuremen e over a trial.

ed by the birds.

of the domestic g nism starts opera

food particles (V water not recover ies for which dat 6.1). This may be aining ducks.

or unit of time by uch lower than i e species.

apter 6

study e of

ism al.,

lter- t

ater s the ducks,

re re not es and f the ides ace by

al n, and ntical

water n of

t

goose ting.

Van red in

a is e

y the n the

(14)

Aft we stu spe sw no ab ma It h tha vo vo ge clo of mu me pu vo ba 0.5

Fig goo sid

volume of water expelled Volumeofwaterexpelledper

ter accounting fo eight pump perfo udies on the grea ecies (Kooloos et wan have a much on-specialized (An

solute volume pu allard.

has been demons an filter-feeding A

lume per cycle m lume is large eno ese and swan, an osed bill of a mall the size expected ute swan) / 1.04 easured bill volum

mped volume pe lume. Measured sed on geometri 58) = 26% of the

gure6.4. Volume o ose species and the

e.

0 0.2 0.4 0.6 0.8 1 1.2 1.4

lesser white- fronted goose p per cycle per body mass (ml/kg)

lesser white fronte goose Volume of water expelled per cycle per body mass (ml/kg)

or the size differe ormance is very s ater white-fronte t al., 1989) shows

lower pump perf naspenelope) filt umped per cycle

strated that gees Anas species (ch may be expected.

ough to explain th nd smaller ducks.

lard and a mute s d for its body size (weight mallard) me is only 65/13 er cycle is howev

intake of mute s c scaling with res expected cycle-v

f water expelled pe e mute swan. Figur r

- d e

barnacle goose

mu sw r

- d e

barnacle goose

mu sw

ence by scaling w imilar in geese a d goose (Van der s that, scaled to b formance than b ter-feeding ducks in the mute swa

se and swans hav apter 2), and the However, it is un he large differen . A rough compar swan indicates th e. Geometric sca

= 8.33 larger bill ml = 5 times larg er lower than the swan divided by t spect to the mall volume.

er movement cycle res from literature ute

wan ute m wan

m

ater intake per p nd swan. A comp r Leeuw et al., 20 body size, goose oth specialized (A s (figure 6.4 and t

n is only 2.1 time

ve relatively shor erefore a relative nlikely that this d

ce in pump perfo rison between th hat the bill of a m ling would predic l volume for the ger. The measure e expected 60 % the expected inta

ard (table 6.1) eq

e of the bill scaled g data for ducks are mallard wigeon mallard wigeon

pump cycle to bod parison with prev 003) and three an species and the m Anasclypeata) a table 6.1). The es larger than in t

ter and narrowe ly smaller pumpe difference in bill

ormance betwee he outer volume o mute swan is only

ct an 8.67 (weigh mute swan, while ed difference in

reduction from b ake of mute swan quals 1.24 /(8.33

geometrically for t e depicted at the rig

tufted duck

norther shovele tufted

duck

northe shovel

dy vious natid mute nd the

r bills ed

n the of the y 60 % ht

e the bill ns

*

two ght- rn er ern

er

(15)

124 Th rel dif the 20 cav ex allo tot su ma As in co Al wh vid ge the tw len rel

Pu We an me Lar no cu the ele are ton shi mo de ton typ in ba ton

4

e difference in vo lative bill size ma fference in the ex e study of Kooloo 05). Values of cy vity were higher perimental set u owed to submerg tal bill length). U bmerged, and at ay depend less on

the geese and sw the study of Koo mparison among ternatively, gees hich reduces the deo recordings su

ometrically scale e power 1/3 time wice as large as ex ngth, while in fac latively short bills

umpcapacityand e believe that the d geese and swa echanisms and m

rge pieces of veg ot be transported shion in both stra e upper bill bears evating the tongu e pressed against ngue movement ifted with respec ovement of the b pressed during p ngue from prope pe of tongue mov

a ‘leaky’ pump w ld and food item ngue to operate

olume per cycle t ay be due to eithe xtent to which th os et al. (1989) m cle volumes obta than those meas p. To be able to c ge only the most nder natural con a more acute an n suction force a wans in the prese

loos et al. (1989) g the species.

se and swans may volume of the or uggests the oppo ed up to the size o

es gape) the gape xpected. Note tha t the bills of gees s only further red

dtransportmech e difference in cy n species on the morphological ada getation (grass, w along the tongu aining ducks and s spine-like struct ue, while the tong t the roof of the

in straining duck ct to the rostro-ca bill. In straining sp protraction. In ge erly acting as a pis

vement is used. E with a reduced pu s are not trapped as a closed valve

that remains afte er a relatively sm he bill is immerse may underestimat ained from a biom sured experiment capture expelled t rostral part of th

ditions ducks ma ngle to the water nd become large ent study were te ) an underestima

y use relatively s ral cavity. Howev osite (table 6.1). W

of the geese and es measured in g

at this compariso se and mute swa duce the expecte

hanism

ycle volume betw other hand is mo aptation of the u waterplants), but e cushion but mu grazing geese. In tures to facilitate gue is protracted mouth and retain ks the elevation-d audal movement pecies the tongue

ese the presence ston during strain Elevation of the t ump capacity. In d

d between the to e-system (Zweers

er accounting for maller opening of d. The cycle volu te the true pump mechanical mode

tally. This is prob water and seeds heir bills (approx ay feed with at le surface. In this p er than in the exp ested in a similar ate of cycle volum

maller gapes dur ver, a rough estim When the gape o

swan species (ra eese and the mu on assumes geom

n are relatively s ed gape.

ween duck specie ost likely related pper bill used to also large seeds ust be transporte n geese and mute e transport of pie d and the bills are ned by the spines depression move t of the tongue an

e is elevated dur e of spines preve ning, even when tongue during ret ducks, the inside ongue and upper

et al., 1977).

Cha

r the difference in the bill (gape), o mes of the ducks p-capacity (Gurd, el of the bill and o bably the result o

s, the birds were imately one third ast half of their b position cycle vol perimental set up r situation as the me does not affec ring filter-feeding mate of gape from

f the mallard is atio of body weig ute swans are alm metrical scaling of shorter. However

s on the one han to different tran filter-feed or gra (Kooloos, 1986), ed over the tongu

e swans the insid eces of vegetatio e closing, food ite

s. Compared to t ement of the tong

nd the opening-c ing retraction an nts an elevated

the duck-strainin traction would re of the upper bill r bill, allowing the

apter 6

n r a s in oral of the

d of bill

ume p used.

ducks ct the g, m ghts to most

f bill r,

nd, nsport

aze.

can ue de of

n. By ems the

gue is closing

d ng esult l is e

(16)

A co filt du wa cu inc of du

Str On bil tim Ko co fre rel no str the fee op rea An filt op filt

Int A c see pro the Th mi mm an to

further disadvan ntinuous through ter-feeding exper

ring a series of m ater and a furthe

shion. During the crease, and some a particular amo cks.

rainingfrequency ne way to compe

l movements, an me. Straining freq ooloos et al., 1989 ntrast to expecta equencies than m lationship betwe orthern shoveler h

raining frequency e forces that are eding. The flow o pening. Drag force

action force of th n increase in bill s ter-feeding frequ pener muscles co ter-feeding frequ

take

comparison of se eds per ml water obably an effect e supply tray dur e comparison of llet used in our e m). Although the serine birds com (metabolic) body

ntage of eating la hput of food to th riments in geese movements with r series of bill mo e collection phase etimes seeds are ount of seeds may

y

nsate for a decre nd to increase the quencies vary litt

9), and are largel ation, the geese a most ducks. Altho

en the size of the has a bill that is a y is almost half th generated by the of water along the es are proportion he water is propo size will increase uency. Interesting mpared to non-g uency.

eed intake per str r pumped throug of an uneven dis ring the experime

relative intake ra experiments was

difference in siz pared to ducks, d y mass.

rge food items m he oesophagus is and mute swan c immersed bill, af ovements is used e the number of observed to drop y therefore take

ease in pump-cap e amount of wate le within individu y independent o and mute swans ough the data are e bill and strainin almost twice as la hat of the wigeon e bill movements e bill (drag) and t nal to both area o ortional to the dis both forces and gly, grazing Anser grazing species (c

raining cycle show gh the bill than th stribution and de

ent.

ate (mg/cycle.kg) larger than used e contributes to duck species still

may be that for la s not possible. Th clearly show that fter which the he d to transport the seeds on the ton p back into the w longer in Anserin

pacity is to increa er passing throug uals and species ( f food size (Koolo in the present st e limited, there ap ng frequency, esp

arge as that of th n. Such a relation s through the wa the displacement of the bill and ve splaced mass and

may therefore be riformes have rel chapter 3), which

ws that the goose he mute swan and livery of seeds th ) is further biased d for the duck spe

an overestimatio have a higher in

rge food items a he video-images o

t food is collected ead is lifted from e food over the to ngue is seen to water again. Inges

nae species than

ase the frequency gh the bill per un

(present study;

oos et al., 1989).

udy show lower ppears to be a pecially in ducks.

he wigeon, and its nship may result f ter during filter- t of water resist j locity squared; th d to its accelerati

e at the expense latively small jaw may further limi

e species take in d ducks. This is hat are drawn in f

d by the fact that ecies (2.3 versus on of intake rate take rate when s

of the d

the ongue stion

in

y of it of

In

The s from jaw he

on.

of w

it