INHSBull22No5


TAT OF JLLl r I 
DWIGHT H. GREE , Go'l.ler,ll)r 

DEPARTME T OF RE I TRATIO. A D EDU 

ATURAL HISTORY UR DI I IO 
TH EOOORE H. FRLSO . , Chirf 

Volume 22 BULLETI 

Preferential Rating of 

Duck Food Plants 

F R ANK C . B ELL R OSE, JR. 
HA RR Y G . A DE R 0 

• 
Printed by Authority of tfte Stale of llfinois 

URB A, ILLI or 
lvlay 19../-3 



Rice cut-gras (lursia or_vzoidn) growing in early fall along a tributary of the Illinois River. Research indicates that thi moist- oil plant out-
ranks all other uncultivated species as a source of food for migraror~• waterfowl in the Illinois River valle)". 



Preferential Rating of 

Duck Food Plants 
FRANK C. BELLR 

HARRY T. 

D RI G the past decade, the 
amount of research on the food 
plants used by migratory water-

fowl has increa ed notably. The ba i 
for much of this research ha been data 
obtained by the U. S. Biological Surve 
(. ucceeded by the . S. Fi, h and Wild-
life Service) through analyses of the con-
tent of everal thousand duck tom-
achs, collection of which wa begun in 
1901. The e data have recentl y been 
, ummarized by Martin ' hler (1939), 
who ba~ed their tudy on 7,998 tom-
ach , and by Cottam (1939) . 

Analyse of duck _ tomach, by the 
Biological Survey, other rganization 
and individual has prompted wildlife 
inve. tigators to attempt to ascertain the 
important duck food plant and their 
relative value in many regions of the 
l 1nited States. T n all these tud ies, each 
plant specie - has been judged solely on 
it use, as determined by lab ratory 
analy e of stomach . \ iVhile thi method 
a certain the important duck food 
plant it doe not reflect the relative 
values of these plant , for no con ider-
ation is given to the abundance of the 
plant specie in the area in which the 
stomachs have been collected. If in a 
certain area a plant species covering 
only 10 acres shows the ame volume of 
eed u e by duck a, another plant _ pe-

cies covering 100 acres, the two plant 
. hould not be considered of equal value 
as duck food sources; indication s are 
that the plant specie occupying 10 
acre is p tentially 10 times as aluable 
as the . pecie, occupying 100 acre . 

The hortcomi ng in the method of de-
~ermining the value of variou plants as 
food for waterfowl ha been evident to 
several investigators. For in tance, 

1artin & hler (1939) in commenting 

on the percentage and rankin g of du k 
fo d items in ·everal tables state that 
"it wi ll be de ·irable to have more ex-
tensive, careful field observation · to 
upplement present con clusi n -, which 

are founded primari ly on laborator) 
analy e ." Concerning the heav u.e 
made by duck f p ndweed and sedge 
eed , Pirnie (19 ) . ays that this may 

reflect abundance of the seeds rather 
than preference of duck for these , eeds. 
Mc tee (1918) state that " uperior 
availability after all is the guiding prin -
ciple in the ch ice of food by birds." 

The writers, * 193 -1940, had th ~ 
opportunity to make an inten ive study 

f the occurrence and use of duck food 
plants in the Illinoi, River valley. A 
preliminary report covering the results 
of the 1938 investigati n has been i , u t1 
(Bellro. e c ' Anders n 1940). B cause 
altered water levels in 1939 and 19--lO 
greatly changed the food resources of 
many b ttomland lake throu gh the tle-
velopment of extensive communities of 
moi t-soil plants, it i thought :les1rabl 
to ·ummarize at this time the data for 
all 3 years. 

Method of Rating Food Plant 

In the present paper, th authors have 
attempted to obtain a numerical rating 
of the value f certain wat rf , 1 plants 
by divi ling the per cent of use made by 
ducks of the plant part<s by the per cent 
of abundance of the plants. Per cent o f 
use data are based in th olumetri c 

* Fit:"ld \Vork rP14.:tli ng l o ol...,n l ... r,,•,·j Wtl~ Uont• J,y 
B<•llro. e, Lit n a. now \ ss,s t.«nl l ,a mc T d1111ri,1n uf t.111· 
Ill inois ' tural llislory Survc,y. \ 11.i lyses nl L1Jma<1, 
con t n t s wa~ ti, work nl \ndersuu . at ll w tim, • l• m pluy~rl 
by the urv y as Ju11i1Jr Bioln,gi~l un P roject. 2-R 01 t he 
Federal Aid in \\' ildlil Rest.or Li on Progrnm. carried 
on in coop rali n with th Slate Orpartmenl ol On.'Pr-
valion and lhe . :. Fish llnd \\"i ldhlc ervi e 

[ -07 ] 



ILLI. ·01 Tt: L HI TOR UR\'E B LLETI Yo/. 22, Ari. 5 

mca urcmcnt of th nnt nt. of 200 
S,!atht"rc I at duck lubs during 
19.19 and 19-lO hunting . ea:ons 

n m .). Table 1 re eal. the 
numb r. f . t machs and the arious 
-,p c1 nf du k: repre ented. The . am-
ple of dahhlin duck gizzard is f!lirl: 
proportionate to the occurrence ot the 

I. -Duck tomach e amined In 
tud . 

Dt i.. 

. lallard 
Pintail 
Gri:en-winged teal 
Rluc-wingcd teal 
Baldpatc 

adwall 
."hovelcr 
Canv back 
Rin neck 
Le er caur 
Other duck 

Total 

I~ 1~39 1 19-lO ITOT I, 

I 619 92 1 360 1.907 
222 40 39 669 
9 127 225 
51 59 110 

3 
16 
14 
72 
2 

17 
11 ,147 

49 2 
3 36 
23 39 
1 15 
9 81 
1 3 

10 6 33 
1,64 405 3 ,200 

different pecie of thi . type of duck in 
the 11linoi River region. The ample of 
di ing duck gizzard i not adequately 
repre. entative of the occurrence of thi 
waterfowl group in the region. 

on. idered in thi paper are only un-
ulti ated aquatic and moi . t-soil plants. 
orn, u. ually an important duck food 

plant in the Tllinoi River valley, is not 
included in the calculation . 

Per cent of abundance data are based 
up n the urface area in acre f the 
aquatic and moi t- oil bed. at the vari-
ous lake tudied; thi area in each case 

a obtained b. plottin the tati n 
b asp tion and rou h triangul ion on 
large cale ba e map and u ing a pla-
n imeter to mea ure the area o red b 
ach sp ci . (Bellro e 1941). able 2 

give: the total area of aquatic nd moi t-
soil e etation for the lake con idered 
in thi · paper, and the number of gizzar 
e amined from each area. 

An inde rating of 1.0 for a d plant 
indicate, that thi plant u ed ap-
pro imately in proportion to it bun-
dance. The larger the figure, the greater 
i the alue of the plant pecie a duck 
food. Conver. el , the maller the figur 
the lower i the value of the pecie . 

We do not wish to pretend that th 
inde figure Ii . ted here are the final and 
ab olute evaluation of the food plant 
enumerated. Habit of waterfowl, plant 
difference , plant habitat characteri ti 
and inaccuracie in determining abun-
dance of vegetation are factor tendin 
to prevent minutely e act rating figur • 

Duck, killed at any one lake ma r 
may not have obtained their I t meal 
there. Plant item in tomachs an 
banding data reveal that occasional) 
duck have fed 10 or 100 or more mil 
from the place in which they are hot. 

Food habit of variou .pecie 
ducks differ widely, too, so that c n 
tail may rank high at a certain place 
cau e of presence of man baldpat 
and gadwall , or nutgras e and pig 
weed may rank high at another pla 
becau e pintail and teal are abundan 
there. 

In ome in tance change in the rank 
ing of a plant from year to year are du 

Table 2.- umber of acre of duck food plant at variou lake in the Illinois RI 
valley and the number of duck stomachs examined from each area. 

193 1939 1940 

ARF:A 
ere of 1 ... umber of of Acre of , umber of Acre 

\'egetation tomach \' egetation tomach Vegetation 

Ou~k Island 1,255 207 1,539 162 
lear and Chautauqua 
Lak.:s 1,441 35 2,022 531 2,389 134 

Crane Lak 425 293 466 342 531 142 
Cuba J land 329 2 9 441 355 
Oougla Lake 1,6 6 9 

oo e Pond 90 160 968 129 
Total 3,450 1.14-7 7,062 1,648 3,888 405 



Jlay, 19-1-3 BELLRO E & A ' DERso T: D K Fooo PLA ·T -1-1 9 

to change in eed production resulting 
from change in environment. u h 
change are well illu trat d b water 
hemp at Clear Lake. In 1940, becau ' e 
of water level changes that reduced eed 
yield, water hemp had an index rating 
only half a great as in the previou 
year. Among aquatic plant , giant bur-
reed, mar h martweed and ago pond-
weed are notable a pecie that vary 

greaciy in the yield f ed with an-
ance in envi r nmental condition ·. 

ln me year, , plant e<l . pre en t in 
abundance ma y be ina ces ·ible to al l c r 
' m ducks through the ab ence of 
water or thr ugh a depth of water toe 
great for dabbling pecies. In th r 
year, , , ee<ls of plant- not pre.,ent in 
tho e years may be available from pre -
iou years . F r e ample, in 193 , be-

Fig. 1. - Ri ce cut-gras (l eersia 01:v zoides ), or awgra s, in mo ·t y~ar and_ in mo_ l Illinoi s 
River valle\' habitats lead all other uncul tivated p lants in value a a ou rce of food to~ du ck · 

ot only i · this specie valuable for it seed, which it produce 111 fnir abundan c, but it · ro r-
sto k and hoots al o furni sh food for water fowl. 



2{ 

au e ot hi 1 h atcr, moi ·t-.oi l plants-
millt:t , martwecd ·, nut ra · and the 
lik " re aim > t n >ne i:tent in the 
lllinc 1. R1n:r vall y. Yet st ,mach 
nalvses n:n:aJed that · ed f th e 
pc .;e. f rmcd aim st one-fourth f the 

un ·ultivatcd food . taken. ting that 
th ,u. anJ · if ducks were feeding n ap-
parently barren mud flat. and in hallow 
water, we c mjc tured that the e water-
fowl were con urning seed depo ited 
from plant beds o curring there in 1936 
and 193i. 

To a. certain the validity of thi con-

Table 3.- Per cent of u e, per cent of 
bundanc , and inde value of aquatic and 

m i t· oil plant at certain lakes in the 
lllinoi River valley, 1938-40. Areas in-
cluded are Ii ted in table l. 

PL. T 

. ' utgr· ·e 
Large ced, nodding 

and other mart-
weeds ....... . 

J a pane e and wild 
millet 

iant bur-reed 
oontail .. 

Duck potato 
. larsh smartweed. 
Buttonbu h 
Lon rleaf pondwecd 
• pike ru h 
Te I gra · 
Water hemp 
. lar h cord gr, 
White waterlil · 
Sago pondwcc:d 
Ri ver bulru h 

m rican loru 
Pickerel weed 
~lar h mallow 
: uthcrn nai. d 
Wild ri c 
. mall pondwe d 
Longlcaf 

ammannia 
Other pl:tnt 

25 53 
6 19 

10 5 

11 76 
0 26 

15 09 
2 73 
6 60 
2 62 
J 19 
0 16 
1 99 
3 02 
0 05 
0 0 
0 55 
0 50 
0 35 

tr. 
tr. 
tr. 

0 00 
0 00 

0 00 
J 72 

PER 
CE T 

or 
As N-
OA ' E 

l 93 
0 5 
l SJ 

1 .09 

3.45 
0 .10 
7 91 
2 79 
7 .87 
1 13 t 
5 . 
0 .34 
002t 

. 0 
0 21 
0 74 
5 55 

26 02 
23 26 
0 01 
0 01 
0 12 
0 59 
0 05 

0 02 
•• 

J DEX 
AL E 

13 32 
10 .67 

7 .12 

4 .07 

3.41 
2 60 
1.91 
0 98 
0 .84 
med. 
0 .54 
0 47 

0 34 
0 .24 
0 10 
0 10 
0 02 
0 02 
low 
low 
low 

v. low 
v. low 

v. low 

•Ot.her than mar h mart.weed . tB u of difficulty 
f mea unng alrnndan · ot Lli plants, these figures are 

nut. a(curat hut. rcpr nt minimum abundance. ** o 
tigure bt-cau ol difficult.y of measurement. tr.- race. 
ab.- bundant, v, h,1\\=\'ery low. med.=medium. 

B LL Tl l'ol. 22. Art. 5 

jecture, in 1940 we collected from three 
place · mud ample aggreg ting 1 
. quare feet f . urface. Th e ampl 
• ielded 2,500 eed of Cyperus tr throrhi-
zos, 2,000 eed f C. slrigosus and SO 

ed f water hemp, Acnida tu/JerculaJa. 
o plant had grown on the e mud Rat 

for 3 year . \ e believe that our findin 
uh tantiate the premi e that large 

quantitie of eed depo ited in on year 
ma be available a food in ucceeding 
year in which germination i precluded 
b unfavorable environmental condi-
tion . 

The area of the variou plant bed , 
a determined from the map , are on) 
approximate. Jn certain place ther 
are sea onal change in the plant com-
munitie , change that re ult in error in 
calculating the abundance of the peci 

Table 4.- Per cent of use, per cent of 
abundance, and index value of aquatic 
and moist-soil plants at certain lakea in 
the Illinois River valley, 1938. Area in-
cluded are listed in table l. 

PLANT 

Rice cut-gra s ..... . 
Coontail. ......... . 

utgra ses ....... .. 
Marsh smartweed . . . 
Longleaf pondweed 
Buttonbush .... . .. . 
Teal grass . 
Water hemp ...... . 
Largeseed, nodding 

and other mart-
weeds* ......... . 

Duck potato ...... , 
ago pondweed 

American lotu 
Giant bur-reed 
River bulrush 
\l hite waterlily .. 
Mar h cotd gra 
• pike ru hes .. . 
Pickerelweed .. . 
outhern naiad .... . 

Mud plantain . . .. . 
Other plant (total 

of SO) 

PER 
CENT 

or 
UE 

2 .46 
22 .72 
11 . 73 
9 .61 
5 .85 
3 .49 
3 16 
2.17 

1.63 
1 .40 
0 .89 
0 38 
0 .35 
0 30 
0 .21 
0 . 15 
0 .05 
0 .00 
0 .00 
0 .00 

7 .45 

PER 
CENT 

OF 
Aeu -
DA CE 

0 . 73 
14 .20 

tr. 
14 .43 
13 .48 

ab. 
tr • 
tr. 

tr. 
7 .14 
8.81 

28 .30 
tr. 

9 .97 
0 .80 
0 .70 
1.04 
0 .40 

tr. 
tr. 

•• 

I DEX 
ALUE 

39 00 
1.68 
high 
0 .66 
0 .43 
low 
med. 
med. 

med. 
0 .20 
0 .10 
0 .01 
low 

0 .03 
0 .26 
0 .21 
0 .05 
v. low 
v. low 
v. low 

*Other than marsh amartweed. ••No fi1ure liv 
because of difficulty of meaaur ment. tr.• 
ab.=abundanl. med.=medium. v. low•very low. 



Ma11, 19./-3 BELLROSE & ANDER ON: Du K Fooo PL T 421 

involved. A noticeable error ccurred in 
the case of teal gra , Eragrostis hyp-
110ides, becau e millet , pigweed and 
other plant growing over thi s specie, 
when the area were urveyed made it 
virtually impo ible to plo t. 

It may eem that the di crepancie 
enumerated above outweigh the bene-
fit derived from thi method of evalu-
ation. However, we believe that, de-
spite it limitations, the u e-abundance 
rating give a far more nearl y accurate 
picture of the food value of a plant than 
do percentage ha ed olely upon the 
amount of food taken. 

Becau e of the different facto r that 
may affect the availability of seed o r 
other plant part of a particular specie 

Table 5.- Per cent of use, per cent of 
abundance, and index value of aquatic and 
moist-soiJ plants at certain lake in the 
Illinois River valley, 1939. Areas induded 
are listed in table 2. 

P ER PER 

CENT CENT I DEX 
PLANT OF OF VALUE 

UE ABUN-

DAN E 

\\ alter' mille t 9 37 0 26 36 04 
Rice cut-gras 22 .25 1 47 15 13 

1utgrasse .. 10 . 3 0 . 72 1- 04 
Japanese and wild 

millet ..... . .. 1-l . 82 5 .4-1- 2 72 
Duck potato . . . . . 3 .94 1 4 2 66 
Largeseed, nodding 

and other mart-
weed* .... . . . ... 3 .13 tr. high 

pike ru he . 0 25 0 10 2 50 
Buttonbush . . .... 2 64 1.16t med. 
Giant bur-reed 0 44 0 20 2 20 
Coon tail 13 .62 6 2 2 19 
Mar h smartweed 7 .75 6 .73 1 15 
Teal grass . . . .. 1 54 tr. med. 
Longleaf pondweed 2 .11 2 .24 0 .94-
Water hemp 1.65 2 .22 0 74-

ago pondweed . 0 .38 6 .14 0 06 
American lotus . . . 0 -1-3 25 29 0 02 
River bulru h .... 0 .6" 37.47 0 02 
Pickerel weed . ... tr. 0 03 low 
Mar h mallow .. tr. 0 .02 low 
White waterlily tr. 1 15 10\ 
\\'ild rice 0 .00 l. 22 low 
Mar h cord gra 0 00 0 26 low 
outhern naiad . . ... 0 .00 0 .12 low 

Other plant 4 20 ** .. . ... 

in an one year, it eem advisable tn jn-
clude yearly table a well a. a general 
ummar table. mpari:;on between 

year · will ai<l in determining the 111-
flu ence faltered habitat condition on 
the food value f plant pecie . 

Ratings of Food Plants Studied 

The value, a. fi od fi r duck , of the 
m ist- oi l and aquatic plants com monly 
occurring in the Jil in i. River alley i 
ba ·ed on data presented in table 3- 6. 

Rice cut-gra , L eersia o,~v~oidt's, fig. 
1, utrank all other . pecie 111 fo d 
value fi r the -year p riod, table 3. I n 
1938 it ranked first by a wi<le margm, 
table 4. In 1939 it ranked ·econd to 
Walter' millet, table 5; in 1940 it wa 
lightly behind J a pane e and wild mil-

let combined, table 6. Thi · variance in 

Table 6.- Per cent of u e, per cent of 
abundance, and index value of aquatic and 
moist- oil plant at certain lake in the 
Illinois River valley, 1940. Area included 
are Ii ted in table 2. 

PER P ER 

CE T CE:-.rT J 1¥DEX 
P LANT OF or VAi.i E 

E 
AB l ,.__ 

DANC:E 

J apanese and wild 
millet 21 2 2 92 7 2 

Rice cut-gra s . 2 72 4- I " 6 92 
Walter' miJlet 7 6,1- 1 63 --1 6 
Large eed, noddin 

and other mart-
weed * .... . . 3 94 2 -1-3 
utgras es -l l 2 06 

Duck potato 1 4-1 0 97 
Coontai l .... " 16 0 70 
Spike rushes 0 13 0 60 
Button6u h 2 o- 0 3-
Teal gras 0 0 t 
'Nater hemp 2 OJ 0 30 
Longleaf pondweed 5 5 0 25 

ago pondweed .. 1 5 OU 
Marsh martweed 4- 06 0 I 
Ri ver bulru h 20 4 0 02 
American lo u I+ 27 0 00-1-

outhern naiad 0 22 lo 
mall pondweed 20 low 

Longleaf 

: · I 0 00 0 07 low ammannia 
Other , man)· 6 00 * 

*Other than mar h marlweed. "'* To figure given *Other than mar h manweed. • · 'u figur given 
becau•Q of d1"fficu.lty of measurement. lr.= Lrace. m d.= d"ffi It r mea urem nt t'I Lrumum "-ure - becau e u1 , cu y o " ""' · medium. tMinimum figure. 



cau c 1i high \\ atcr moist-soil plants-
m1llet. , martw t.'d nut ras. e · and the 
like \\Crl· almost n ne ·istent in the 
lllin ,1. l{ ,,·cr \'alley. Yet . tomach 
analv . r vealed that . eed of these 
p ·it:s f rme~ aim . t one-fourt_h of the 

uncult1rnt d to is taken. ting that 
th u ands of duck w re feeding n ap-
parentl v barrt:n mud Rat and in hall w 
water, ~e ·onjectur d that the e water-
fowl were ·c nsuming seed dep ited 
fr m plant bed~ c urring there in 1936 
and 19 7. 

To ascertain the alidit f thi con-

Tabl 3.- Per ent of u e, per cent of 
abundance, and inde value of aquatic and 
moi t- oil plant at certain lakes in the 
JIiin i River valley, 1938-40. Areas in-
luded are Ii ted in table 2. 

Pt.A T 

Ri ce cut-gra . . . . . . 25 53 
Walter' millet 6. 19 
. ' utgra e 10 . S 
Lar-e ced, nodding 

and other mart-
weed • ....... . 

J apane ·e and wild 
millet 

ia nt bur-reed 
Coonrnil . 
Duck potato . 
t\l ar h martwced. 
Buttonbu h ... 
Longb1f pondweed . 
'pikc ru he 

T ea l gr· 
Water hemp 

lar h cord gra 
White waterli ly 
. ago p nd eed 
Riverbulru h 

merican lo u · 
l'ickt>rdweed 
l\ l ar h mallow 

outht>rn naiad 
Wild ri ce 

mall pondwecd 
l.ongleaf 

ammanma 
ther I lanr 

-l .-l-l 

11 76 
0 .26 

15 09 
2.73 
6 60 
2.62 
3 19 
0 16 
1.99 
3 02 
0 OS 
0 0 
0 55 
0 so 
0 .35 

tr. 
tr. 
tr. 

0 00 
0 00 

0 00 
3 72 

PER 
CE NT 

or 
As N-

OAN E 

1.93 
0 .58 
1.53 

l 09 

3.45 
0 .10 
7.91 
2.79 
7. 7 
1.13 t 
5.88 
0 .34 
002t 
8. 0 
0 .21 
0 74 
5 .55 

26 02 
23 .26 
0 01 
0 01 
0 12 
0 59 
0 OS 

0 02 ... 

INDEX 
AL E 

13 .32 
10 .67 
7 . 12 

4 .07 

3 .41 
2.60 
1.91 
0 .98 
0 84 
med. 
0 54 
0.47 

0 .34 
0 .24 
0 .10 
0 .10 
0 .02 
0 .02 
low 
low 
low 

v. low 
v. low 

v. low 

Lh r Lhan m:-,r h mart weed . tBecau e of difficulty 
of m a urinR abundan e or Lh e plant , these figures are 
not accural bUL repr enl minimum abu.odat1ce. •• o 
figur :a u of difficully of mea urement. Lr.=trace. au. · bundant. v. lo - •ery I w. med .=medium. 

R\'EY 8 LLETI f' ol. 22, Art. 5 

jecture, in 1940 we collected from three 
pla e mud ample aggregating 1 
. quare feet of urface. The e ample 

ielded 2,500 eed of C_vperus erylhrorhi-
zo.r, 2,000 eed of C . .rtrigo.rus and 550 
·eed of water hemp, Acnida luherculata. 

o plant had grown on the e mud flat 
for 3 year . \\ e believe that our finding 
ubstantiate the premi e that large 

quantitie of eed depo ited in one year 
may be available as food in ucceeding 
year in which germination i precluded 
by unfavorable environmental condi-
tion . 

The area of the variou plant bed , 
a determined from the maps, are only 
approximate. In certain place there 
are sea onal change in the plant com-
muni tie , changes that re ult in error in 
calculating the abundance of the specie 

Table 4.-Per cent of use, per cent of 
abundance, and index value of aquatic 
and moist-soil plants at certain lakes in 
the Illinois River valley, 1938. Area in-
cluded are listed in table 2. 

PER PER 
CENT CENT INDEX 

PLANT OF OF VALUE 
U E AeuN-

DANCE 

Rice cut-grass .. .... 28 .46 0 . 73 39 .00 
Coontail . .. . . .... .. 22 .72 14 .20 1.68 
Nu tgrasses ......... 11 . 73 tr. high 
Marsh smartweed . .. 9 .61 14 .43 0 .66 
Longleaf pondweed . 5.85 13 .48 0 .43 
Buttonbush .. ...... 3 .49 ab. low 
Teal grass .... . .... 3 .16 tr. med. 
Water hemp .. ..... 2 . 17 tr. med. 
Largeseed, nodding 

and other mart-
weeds* ..... . .. .. 1.63 tr. med. 

Duck potato ....... 1.40 7.14 0 .20 
ago pondweed .. ... 0 .89 8.81 0 .10 

American lotu ..... 0 .38 28 .30 0 .01 
Giant bur-reed ..... 0 .35 tr. low 
River bulru h , ..... 0 .30 9.97 0 .03 
White waterlil y ... 0 .21 0.80 0 .26 
Marsh cord gras ... 0 .15 0 .70 0 .21 

pike ru hes . . ' .. . 0 05 1.04 0 .05 
Pickerel weed ....... 0 .00 0 .40 v. low 
outhern naiad . . . . 0 .00 tr . v. low 

Mud plantain . . . . . 0 .00 tr . v. low 
Other plants (total 

of 50) .. , ..... .. . 7.45 •• .. .. 
•OLher than marsh smartweed. ••No figure given 

because of difficulty of measurement. tr.=trace. 
a b.=abundant. med.= medium. v. low-very low. 



May, 1943 BELLROSE & ANDER ON: D CK Foon PLA Ts 421 

involved. A noticeable error occurred in 
the ca e of teal gra , Eragrostis l1yp-
noides, because millet , pigweed and 
other plants growing over thi pecies 
when the areas were surveyed made it 
virtually impo sible to plot. 

It may eem that the di crepancie 
enumerated above outweigh the bene-
fit derived from thi method of evalu-
ation. However, we believe that, de-
pite its limitation , the u e-abundance 

rating gives a far more nearly accurate 
picture of the food value of a plant than 
do percentage ba ed olely upon the 
amount of food taken. 

Becau e of the different factors that 
may affect the availability of eeds or 
other plant part of a particular pecie 

Table 5.- Per cent of use, per cent of 
abundance, and index value of aquatic and 
moist-soil plants at certain lakes in the 
Illinois River valley, 1939. Areas included 
are listed in table 2. 

PER PER 
CENT CE T I DEX 

PLANT OF OF VAL E 
SE AB N-

DA CE 

" ' alter' millet . , . . 9.37 0 .26 36 04 
Rice cut-grass . . .... 22 .25 1 4 7 15 .13 

utgra es .... . ... 10 .83 0 . 72 15 04 
Japane e and wild 

millet . .. - .. ... 14 82 5 -1--l 2 . 72 
Duck potato . 3 .94 1.-1-8 2.66 
Largeseed, nodding 

and other smart-
weeds* .. .... .. 3 . 13 tr. high 

pike rushes .. . .. 0 .25 0 . to 2.50 
Button bush ....... 2.64 l. l6t med . 
Giant bur-reed . ... 0 .-14 0 .20 2.20 
Coontail ......... 13 62 6 .2 2 . 19 
Marsh smartweed 7.75 6 .73 1 15 
Teal grass .... . . 1 5-! tr. med. 
Longleaf pondweed 2 . 11 2 24 0 .94 
Water hemp 1.65 2 22 0 . 74 

ago pondweed ... 0 3 6 . 1-1, 0 06 
American lotu .... 0 43 25 .29 0 02 
River bulru h 0 .65 37.-1-7 0 .02 
Pickerelweed ...... tr. 0 .03 low 
Marsh mallow . .. tr. 0 .02 low 
White waterlily . .. tr. 1. 15 low 
Wild rice ......... 0 .00 I . 22 low 
Marsh cord gra 0 .00 0 .26 low 
Southern naiad ... 0 00 0 . 12 low 
Other plant ....... 4 20 ** 

in any one year, it eem advi ' able to in-
clude yearly tables a well a a general 
ummary table. Compari on between 

year will aid in determining the 111-
fluences of altered habitat condition on 
the food alue of plant specie . 

Ratings of Food Plants Studied 

The value, a food for duck , f the 
moi t- oil and aquatic plant commonly 
occurring in the Illinoi River valley i. 
based on data pre ented in table 3- 6. 

Rice cut-grass, L eersia oryzoides, fig. 
1, outrank ' all other pecie in food 
value for the 3-year period, table 3. In 
1938 it ranked first by a wide margin, 
table 4. In 1939 it ranked second to 
Walter' millet, table 5; in 1940 it wa 
lightly behind J apane e and wild mil-

let combined, table 6. Thi variance in 

Tab1e 6.- Per cent of use, per cent of 
abundance and index value of aquatic and 
moist-soil plants at certain lakes in the 
Illinois River valley, 1940. Areas included 
are listed in table 2. 

PER PER 
CENT CEwr l NDEX 

PLA T OF OF VALUE 
U 0 E As N-

DAN E 

J apanese and wild 
millet 21 26 2 92 7 2 

Rice cut-gras .. 28 . 72 -1 ts 6 92 
Walter's millet .... 7 .M 1 63 4 6 
Large eed, nodding 

and other mart-
weed* .... 9.56 3 .9-l 2 43 
utgrasses 8 6 -1 . I 2 06 

Duck potato 1 37 1 41 0 97 
Coon tail ..... 3.60 5. 16 0 . 70 
Spike rushes ...... 0 .0 0 . 13 0 60 
Buttonbu h 0 71 2 OS 0 .35 
T eal gra 0 .97 0 0 t 
V.'ater hemp 53 28 03 0 30 
Longleaf pondweed l 3 5 5 0 25 

ago pondweed .. 0 3 1 ·s 0 2-! 
Mar h martweed 0 73 -! 06 0 1 
River bulru h . ... 0 .39 20 .3-l 0 02 

meri can lotu .... 0 05 l4 27 0 004 
outhern nai ad 0 00 0 22 low 

Small pondweed 0 00 0 .20 low 
Longleaf 

ammanma . 0 00 0 .07 low 
Other, man y 6 .00 ** 

*Other than mar h ma rtweed. ** figure given 
because of difficulty of measurement. tr.- race. med.= *Other than marsh smart weed. U No figure given 
medium. tMinimum figure. b~au e of difficulty of mea urement. tMinimum figu re. 



,L2 •. ATl' R L HISTORY R\'EY B LLETI. T T"ol. 22, Art. 

,·alut' 1s bn ught about larg ly by local 
habitat change . From tudies made of 
seed produ ·tion in 1941 b · L w ' Bell-
ro. e (m~.) 1t I appar nt that dry . ii 
greatly lo er the seed pr duct1 ity f 
ri e cut-gra .. ; furthermore, unle there 
i~ an in h or tw of water o er the root-
tocks during the fall, tho:e items, which 

als are a wat rf wl f od, are not a ail-
able to ducks. At se era) lake· where 
inde: alues of ri e cut-gra were ob-
tained throughout the 3- ear period, 
this plant ranked fir tat five and second 

Fi~. 2. -\\'alter' mi ll et (Echinocliloa Wal-
leri) outranks Japanese millet in certain habi-
tat a a food for migratory waterfowl in the 
Illinois River region. While it favorable posi -
tion may be due in mo t in tances to greater 
eed production, it may be due at times to a 

diet preference on the part of pintail and 
teal . 

at four area . check of the en iron-
mental ondition of the. e nine are 
re ealed that the fir t five were more 

Fig. 3.- utgra es (Cyperus spp.), a 
group ranking fifth in value among Illinoi 
River valley duck food plants, produce larg 
quantities of minute seeds. traw-colored 
cyperu (C. strigosus) is pictured here. 

favorable habitat for the plant than 
were the others. 

Walter's millet, Echinochloa Walteri, 
fig. 2, approaches rice cut-grass in value 
according to the index figures in tabl 
3. It i to be expected that, like man 
other plants, this species varies in valu 
with the year and habitat, its value de 
pending on its own seed production an 
the availability of other food . Althoug 
Martin & Uhler (1939) regard the small-
er seeds of this species less importan 
as a duck food than those of wild millet 
E. crusgalli, we believe that Walter' 
millet may locally, in certain years, b 
the more valuable species. For exam 
pie, at Clear Lake in 1940, Walter's mil .. 
let was slightly less abundant than wil 
and Japanese millets combined. How-
ever, Walter's millet amounted to 16.5 
per cent of the native food plant items 
taken from the stomachs of ducks sho 
at thi s lake, while wild and Japane d 
millet totaled only 9.24 per cent. 

We believe that the greater value o 
Walter's millet may lie in a generall 
heavier eed yield of this plant or in the 
fact that pintails and teals prefer th 
maJI eeds of this pecies to the large 

ones of the other millets. We know tha 
pin tail and teal diet are made up o 
mailer seeds than are those of mallards 

baldpate , gadwall and the like, an 
that the number of pintail and te 



BELLROSE & A DERso 1 : DucK Fooo PLANT 423 

tomachs analyzed from Clear Lake 
as proportionally larger than at lake 
t which Walter' millet did n t ranks 
igh. Low & Bellro e (m .) found that, 

·n 1941, wild and Japanese millet pro-
1:iuced considerably greater quantitie 
of eed per unit of area than did\\ alter' 

Fig. 4. - Largeseed s martweed (Polygo 1111111 
pe1111sy!va 11ic11m), in contrast to nodding smart-
wecd, has thick, erect pike . This and other 
moist- oil martwecd con titute a group that 
appears to be slightly more valuable than the 
nutgras e , but les valuable than the mill ets, 
a a ource of food for migratory waterfowl. 

millet. Howev r, in that year, high 
water o retarded growth of \\ alter's 
millet that plants were only 2 to S feet 
high, wherea in 1939 and 1940 they 
were 7 t 9 feet high. The tal.ler plant. 
had con iderably larger inflore cences 
and therefore produced much more seed. 

Nut11rasses, fig. 3, are in third posi-
tion in table 3. Although the index 

rating derived from data obtained in all 
3 year place thi s gr up, Cypems ery-
throrhizos, C. strigosus and C. esculentus, 
third in value, we believe that an inac-
curacy in figures re ulted from the fact 
that waterfowl fed on eed depo ited 
prior to 1938 on mud flat that were 
bare in that year and that were therefore 
recorded a upporting no nutgrass 
plant . All evidence point to the fact 
that a omewhat imilar ituation oc-
cured in 1939, when water prevented 
plant from appearJJ1g on many area . 
In 1940, water level were low earlier 
than in the 2 preceding year , so that by 
fall all mud flats were covered by a rank 
growth of vegetati n. \Ve believe that 

Fi~. 5.- Nodding martweed (Po!ygonu m 
!apttthijolium ) grows on mud fl at a nd other 
moist places. Its long, drooping, de n ely 
flow ered pikes di s tingui h it from other mart-
wed. 

the index value a gi en in 1940 for thi s 
pecie is the mo t nearly accurate, 

table 6. This rating placed the nut-
gra e ab ve duck potato in value- the 
numerical value wa over twice a great, 



-tH lLu 01s AT RAL Hr TORY R EY BULLET( r ol. 22, Art. 5 

in f ·t -but b lo th gr up con i ting 
prin ,pally ,f nodding, large- eed and 
swamp ·martw d. 

Th rdati\'c alue f the three specie 
f nutgrasse may be judged by the fol-

lowing facts: G_vperus trytln-or/1i:os, 
repre. ent d by 56.5 per cent of the total 
nutgrass s ed found in duck gizzard , 
was mu h m re abundant than C. slri-
gosw represented by 19.0 per cent of 
the nutgras. . eed; C. esmlenlus, how-
ever made up 24.5 per cent of the total 
nutgrass eed, e en though Jes. abun-
dant than C. slrigosus. This would in-
di ate that C. csmlcnlus wa slightly 
b tter than C. e1:)1t/1rorhizos, which, in 
turn, wa. better than C. strigosus. 

Moist-soil smartweeds, consisting 
f large eed ·martweed, Polygonwn 

pemuylt•animm, fig. 4, nodding mart-
weed, P. lapathijolium, fig. 5, wamp 
·martweed, P. lzydropiperoides, and 
min r quantitie of other specie , rank 
fourth for the 3-year period, table 3. 
However, here also the ame factors 
prevail that were re pon ible for an 
err r in the index value of the nut-
gra. e. : in many place seeds deposited 

Fi~. 6.- J apanc e millet (Erl1inocliloa fru-
mrnlacea) under favorable growing condition 
i nc of the greate t eed producers among the 
du ck f od plant o ( t he Illinois River valler. 
Because it eed production i greatly depe nd-
ent n growi ng co ndition , it comparative 
value range from excel lent to fair in the cour e 
of several yea r . 

in one year were con urned the next, 
when germination did not occur and 
plants were not recorded. The 1940 
inde rating of 2.43 in table 6 i probably 
the mo t near) accurate. It place 
the e pecie. lightly above the nut-
grasse. in value. 

Wild and Japanese millets, Echi-
nochloa cnugalli and E.frumentacea, fig. 
6, appear fifth in order of preference in 
table 3, covering 1938, 1939 and 1940. 
For ome rea on, waterfowl did not ob-

Fi~. 7.-Giant bur-reed (Sparganium tury-
cm·pum) ranks sixth as a duck food plant in the 
lllinois River region. Its value here is con-
siderably higher than it is generally accorded 
el ewhere. The globose heads are composed of 
nutlike, beaked eeds that are eaten by ducks. 

tain appreciable amount of seed of 
these pecies from mud flats in 1938. 
Had they done so, the status of the e 
millet would undoubtedly have been 
raised above that of the smartweeds. 
Table 6 show that Japanese and wild 
millet headed the list in 1940, when 
they were lightly better than rice cut-
gra s. 

As in other species, millets vary in 
seed yield with habitat conditions. 
Furthermore, time of planting greatly 
affects eed production. Japanese mil-
let sown in the Illinois River valley after 



lllaJ, 1943 BELLROSE & ANDERSON; DUCK F O PLA1 TS 425 

/ 
Fig. 8.-Coontail (Ceratoplzyllum demersum ) rank below the moi t- oil plants in va lue a. 

a source of food for waterfowl, but leads the trul y aquatic pecie . Ducks feed mainly on its 
eaves and stems. 

Aug. 1 often fails to ripen before fro t. 
On the other hand, this millet sown in 

une may mature, the eed hatter and 
germinate the ame summer. In the 
Illinoi River vaUey this second growth 
of millet ha never matured before frost . 
\,\ e do not know just how operative the 
above condition were in Jowering the 
value of this pecies in 1939, but we 
know that the water receded in many 
lake basin 2 week earlier in 1940 than 
jn 1939 to give the millet a longer grow-
mg season. 

Giant bur-reed, 1parganium eur_v-
carpum, fig. 7, with an index value in 
table 3 of 2.60, rank a a much better 
duck food plant in the Illinoi River val-
ley than it is generally believed to be. 
Studies by Low Bellro e (m .) in 1941 
on the seed yield of thi pecies reveal 
that it produced about three time a 
much eed in area with table water 
levels a in area with emi table water 
leveL. In the variou types of areas 
combined, giant bur-reed in 1941 pro-
duced more seed per unit of area than 
either largeseed or nodding martweed. 

Coontail, Cei-atophyllum demer.rum 
fig. 8, ranking lightly below giant bur-
reed in value, table 3, is an excellent 

Fig. 9. - Duck potato( , ,1git1arit1 latifolia ) i 
valued bv duck more for it eed than for its 
large, de~ pl r buri ed tuber . 



I I.LI • • I • AT R L HI TOR y R\'EY B LLETI f'ol. 22, Art. 5 

fn d fo r bal !pate and gadwalls. It 
furn, he· lirtl -.e i, but du k ' mak e -
t n, 1ve use nf the leave. and stem_. 

Teal gra Emgroslis hypnoidrs, ha 
a tatu that cann t be judged by any 
111 I figur deriv d from u ·e and 
abundan data recorded for this spe-
1e , table , 4 5 and 6. Thi . i. evident 

fr lm th fa t that in 193 and 1939 it 
formed r . pecti ely .16 and 1.54 per 
cent o f the fi d taken, , hile 111 a few 
_ mall pat h _ f plant were . een n ex-

tensi e mud flat around everal lake . 
In all probability, duck. in tho e year 
obtained eeds depo ited by growth of 
previou -eason . Thi. mall, procum-
ben t plant in 1940 wa overgrown b 
pigweed , milJet and martweed to 
uch an extent that accurate mapping 
f the area covered by the numerou 

s attered patche wa impo ible. 
From it relatively high u e and ap-

parently low occurrence, we believe that 
teal gra rank above duck potato in 

li'.ig. 10. ;\f arsh martweed (Polygonum !vlulz/enbergii), one of the few martweeds that 
grow 1n wate r, ca n readil y be identified by it bright pink blossom . It is not so valuable a duck 
toad plant a large eed or nodding martweeds because it does not produce so much seed. 



iUay, 1943 BELLROSE & A DERSON: DucK Fooo PLA TTs 427 

(_ 

----
Fig. 11.- Longleaf pondweed (Pota mogeto n americanus) rank eleventh among Jllinois River 

alley plants a a ource of duck food. It ranks above sago pondweed apparently becau e it 
roduce more eed. 

alue, but below the nutgra group. 
eed production studies (Low & Bell-
ose m .) revealed that in 1941 teal 
rass produced about half a much eed 
s Cyperus erythrorhizos and almo t as 
uch as did duck potato per area unit. 
Duck potato, Sagittaria !atijolia, fig. 

, has an index value, table 3, about two-
fth a great as that of giant bur-reed. 
ncidentally, in 1941, bur-reed pro-
uced about two-fifth more eed per 
nit of area than did duck potato. For 
he most part, the large tubers of duck 

tato, often a foot or more under-
round, are not available as duck food. 
owever, at times canvasback and 

ingneck have ucceeded in obtaining 
umerou tuber , which con iderably 
aised the value of duck potato for cer-

tain areas. 
Marsh smartweed, Polygonum Muh-

~enbergii, fig. 10, ha in table 3 an index 
alue of 0.84, which place it slightly 

below duck potato. Beds of thi pe-
cie will not produce eed when growing 

ut of water; optimum production oc-
cur in water 12 to 18 inche deep. In 
everal area , bed produced no eed be-

cau e of a lack of water. Thi pecies 
would rank somewhat higher if all the 
beds that failed to produce eed were ex-
cluded from the calculations. 

Lonllleaf pondweed, Potamogeton 
americanus, fig. 11, rank below marsh 
martweed in index value, table 3, de-
pite the fact it outproduces the latter in 

seed yield by a wide margin. This may 
indicate that, because of the greater 
depth of water at which the plants grow, 
eed of aquatic plants are less accessi-

ble to dabbling ducks than are tho e of 
mar_ h plants. 

Buttonbush, Cephalcmthus occiden-
talis, eldom grows in the area that 
were mapped, occurring usually as a 
part of, or within, the horeline. 
Since it i also a woody pecie , it 
abundance wa not determined for most 
area . However, in the Crane Lake 
region, its abundance wa determined in 
1939 and 1940. When there wa a 
dearth of other duck food in this area, 
1939, the index rating was 0.94; when 
there was an abundance of other food , 
1940, the index rating dropped to 0.12. 
A study f the food con umption on 



,L le. LI. ' 01 RAL HI TORY URVEY BULLETJ l'ol. 22, Art. 

o her areas sub.-tan tiate the Crane 
Lake oh er a t ion . eed pr duction 
,f buttonbu h is n t known to ar ' 
r atl v from year to year; ther fo re, we 

must ~ondude that duck · do not reli . h 
rh ·eeds and that the e seed are . uffi-
cientlv palatable to be taken e ten. ive-
h· nlv wh n other fo d are lacking. 
· pik ru he , Elt'ochm·is pp., include 

main ly t he Elrocl1m·is palustris group. 
T he e ·hort, r und- temmed pecie , 
growing on moi ·t oil r in . hall w water 
may be e. cellent duck foods in . ome 
. ear and po r fo d in other , a illu. -
t rated by the inde value in table 4, 5 
and 6. Tables 5 and 6 how that, in 
1939 and 1940, pike ru he formed 
about the ame per cent of vegetation at 
the lake . tudied. Yet in the fir t year 
the ·eed of the ·e pecie consumed by 
duck wa three time a great as in the 

Fig. ll. - \\' a ter hemp (Acnida tuberculaJa ) 
hcttcr know n a pigweed , i generall y a largeJ 
coarse moi t-. oil plant. 1t i of lower value as 
• · urce of duck food than are mo t other 
specie· growing on the lllinoi Ri ver mud Rat . 

econd ear. We are at a lo 
count f~r thi variance. The 3- ear in-
de rating fi r the e pecie plac their 
duck food value slight! below that o 
the longleaf pondweed, table 3. 

Water hemp, ,1cnida tuoerculata, fig 
12, a large, coar e herb growing on drier 
. oil ites than other moi t- oil plant 
has mall • eed , pin-headed in hape 
that may not alway be as available 
seed. of the other pecie of thi group 
It index value for 1938-40 i 0.34 
table 3. The data include eed pro-
duced prior to 1938, a year in which onl 
a very few plant of this pecie grew i 
the areas studied. The 1940 rating o 
0.30 is probably the best indicator of it 
true value, table 6. Such moist- o· 
plan ts as smartweeds, millets and ric 
cut-grass are apparently about 10 to 
times as good sources of duck food as i 
water hemp. 

Marsh cord grass, Spartina Michau 
iana, may be a fair duck food, as how 
in table 3, with an index value of 0.2 
This figure places it in a better positio 
as a duck food in the Illinois River valJe 
than it is customarily accorded else 
where. 

Sago pondweed, Potamogeton pecti 
natus, fig. 13, is generalJy regarded as on 
of the most excelJent duck foods on t 
North American continent. Howeve 
it falls far short of this ideal in the Illi 
nois River valley. Tables 4, 5 and 
show that its 1938 index rating was 0.1 
its 1939 rating was 0.06 and its 19 
rating was 0.24. Table 3 shows a 3-y 
index value of 0.10. The low status 
this highly rated species in the Illino' 
River region is due to its low seed yiel 
here. Very little foliage and few tube 
of this species were found in the 3,2 
gizzards analyzed. 

The apparently low eed yield of s 
pondweed was substantiated by studi 
made in 1941 by Low & Bellrose (ms. 
which revealed that this plant produc 
less seed per area unit than 23 oth 
aquatic and moist-soil plants in the Ill' 
nois River valley. 

White waterlily, Castalia tuoeros 
has an index figure of 0.10 for the 3-ye 
period, table 3. However, its val 
may vary from year to year and pla 
to place, depending on its own seed yie 
and on the presence or absence of oth 



Ha J1, 1943 BELLROSE & ANDERSON: DucK Fooo PLA, 'T 429 

Fig. 13.- Sago pondweed (Potamogeton pectinatus) has the reputation of being one of the 
e t waterfowl food plants in North America. However, in the Jllinois River va lley it ranks 
ixteenth among 25 species or g roups. Its low racing in Illinois appea r attributable to its low 
eed yie ld in thi area and to the fact t hat it foliage i eldom fou nd in duck gizzard here. 

ood resource . A serious paucity of 
ther duck food plants in 1938 may have 
ccounted for its unusually high rating 
f 0.26 in that year, table 4; when ther 
ood resources were greater, 1939, it in-
ex rating wa less than 0.02, table 5. 
t Crane Lake in 1938, white waterlily 

ated 0.19, but it dropped to 0.01 in 
alue the following year. This wa 
oubtles ly due to the inhibition of 
ruiting cau ed by low water. The 
hove data rank this pecie as poor 
o fair in food value, somewhat higher 
han the American lotus. 

Pickerelweed, Pontederia cordata, fig. 
4, i not given a numerical value in 
able 3. It wa impo si ble to secu re an 
ndex figure for thi s plant covering the 
-year study period because of the in-

ite imal amount of seed con urned 
y duck . An index value of 0.03, de-
ived from data obtained in 1939 at 
rane Lake, places pickerelwee<l for that 
ear and area above river bulrush and 
merican lotus. If its low value i 
dequately portrayed by thi small 

Fig. 14.-Fickerelweed (Po11lederia cordata) 
i of dou b fo l value a a du ck food plant in 
spite of the fact chat it i one of the top-ranking 
seed producers in t he llli noi River region. lts 
blue Rowers and heart-s haped leaves di tin-
gui h it from du ck potato, which has \ hite 
blo oms and arrowhead- haped leaves. 



.po I I.I .I. ' OI . AT R I. HI TORY UR\' EV B LLETI • I' al. 22, Art. 5 

mplc, then th I w use mu t be due t 
lo palatability rather than la k of 
a variability for this s~ ie · i.- one f the 
top-ranking seed produ er , a ·h< n hr 

1 41 study (L, (; ' Bcllrose m . . ). 
River bulru h, . 'cirpus jiu itztilis, a 
arse, d minant mar h plant frequent-

I · f, rm. dense bed. of SO to 700 acre. in 
th Illmoi Ri er valley. nl rarely in 
this area due. it fruit and then nl in 
:mall pat he. ; seed ar -eldom avail-
able f r fi od. Thi. circum tance ac-

unts f r th fact that alth ugh river 
bulru. h fi rmed over 26 per cent of the 

egetati n for the 3- ear tudy period, 
it ed accounted for only 0.50 per cent 
f the uncultivated plant part con-

·umed b waterfowl; it index value for 
the 3- ear period i 0.02, table 3. At 
Lake hautauqua, in 1938, when river 
bulru h bed pr duced more than the 
u ual am unt of eed, the index value of 
thi plant f, r the area wa 0.23. 

ince river bulru h cover exten ive 
area that might well be occupied by 
m re valuable food producer , it mu t 
be regarded a ne f the most perni-
ci u weed in many waterfowl habitat 
of the lllin i River valley. 

American lotus, ' elumbo lutea, fig. 
15, i next to river bulru h in abundance 
in lake adjacent to the Illinoi River. 

nlike thi bulru h, however, lotu pro-
duce a fair amount of eed. Yet it 
index rating i 0.02, table 3, the same as 
river bulru h. It light value a a 
duck food and it dominance over many 

ther aquatic plant of greater value 
make thi ,pecie a weed in the migra-
t ry waterfowl habitat of the Illinoi 
River valley. 

It low value a a duck fi od plant 
during ctober and ovember mu t be 
attributed to the unpalatabilit of the 
hard, nutlike eed . Field ob. ervation. 
indicate that before the ·eed fully 
ripen, in late , ugu t and early eptem -
ber, wood ducks feed extensivel on 
them. At that time the pericarp and 
cotyled n of the eed are soft. 

Marsh mallow, Hibiscus militaris, is 
not generally regarded as a waterfowl 
food plant. However, in the 3-year in-
ve tigation period it averaged 0.01 per 
cent f the vegetation of the mar he 
tudied, table 3, and a few eed were 

con urned by duck . 

Southern naiad, ajas guadalupm-
si.r, i apparentl. a poor duck fi din th 
Illinoi Ri er valley. While it formed 
0.12 per cent of the vegetation on th 
areas ·tudied, table 3, onl a matterin 
of seed - was found in the gizzard 
analyzed. Martin hler (19 9), in 
commenting on the value of the north-
ern and :outhern naiad. in the countr 
a a whole, term them e cellent duck 
food. 

Wild rice, Zizania aquatica, fig. 16, i 
often regarded a the food upreme for 
waterfowl. Thi may well be the ca e 
in region where it abundance i 
measured in thou and of acre . How-
ever, de pite the fact that it formed be 
of 3 to 81 acre in . everal Illinoi River 
valley lake , none of it eed were found 
in any of the duck stomach analyzed 
table 3. A possible explanation for th 
ab ence of eed may be that few wer 
available, for in Augu t and eptember 
thou and of red-winged blackbird 
were observed feeding on the ripening 
eed. Seed that escape the blackbird 

may be inaccessible among the vegeta-
tion debris and muck of the lake bot-

Fi~. 15.- American lotus (Ne/um/Jo lulea), o, 
yorkey nut as it is often called locally, is seldo 
utilized by duck de pite its abundance. I 
low value as a food plant i due probably to the 
unpalatability of its hard, nutlike seeds. 



JI.lay, 19./-3 BELLROSE & ANDERSON: D CK Fooo PLA T 431 

toms; most of the eed fall a month be-
fore the greatest numbers of ducks ar-
rive. McAtee (1939, p. 33) in comment-
ing on the value of wild rice tate : 
"Thi plant ha a great reputation a a 
producer of food for wildfowl-too high 
a rating, perhap , con idering it local 
and ea 'onal availability." 

Small pondweed, Potamogeton pusif-
lus, i another species that i generally 

Fi~. 16.-\Vild rice (Zizania aquatica) is not 
a valuable duck food plant in the Illinois River 
valley; few of the eeds are available during the 
eason of the principal waterfowl Right in the 

fall months. 

considered a good waterfowl food plant. 
Although it amounted to 0.05 per cent 
of the vegetation, table 3, no foliage or 
seeds were found in any of the stomachs 
analyzed. Studie f eed yield in 1941 
(Low & Bellro em .) revealed that it wa 
one of the lowe t producer . We have 
noted that the vegetative part , after 
fruiting time in late July and early 
August, generalJy disintegrate. Whether 
disintegration i caused by green algae, 
by the competition of c ontail and 
southern naiad, or by some unkn wn 
factor, we do n t know. 

Longleaf ammannia, dmmannia 
coccinea, is a moi t-soil plant that ha 
occurred fairly abundantly at Clear 
Lake. N evidence wa obtained that 
thi specie wa ever u. ed as fo d by 
duck , table 3. 

Plant parts of a large number of other 
specie were consumed in infinite imal 
amounts; likewise ome other pecies 

were found in the field in amounts too 
small to tabulate, table 3. 

Discussion 

That the true value to waterfowl of 
the variou aquatic and moi t-soil plant 
cannot be determined olely from the 
use made of them by the bird i evident 
~fter consideration of a few out tanding 
instance . Based olely on u e, Walter' 
miUet ranks ixth in value among duck 
fo d plant of the Illinoi . River valley, 
slightly below mar h smartweed, table 3. 
When abundance a well a u e i con-
idered, Walter' millet is econd in 

rank, nearly 13 time as great in value 
as mar h martweed, which place 
ninth in value, table 3. In another in-
stance, based on u e only, coon tail rank 
econd, three-fifth a valuable as the 

leading rice cut-gras ; however, after 
the abundance figure i considered, 
coontail dr p t seventh place, with 
about one-seventh the value of rice cut-
gra . Giant bur-reed, ace rding to u e 
made by ducks, rank ixteenth in value. 
When the meager occurrence of the bur-
reed is taken into con ideration, it 
jump to ixth in value. 

Although many extraneous and di-
ver. e factors have prevented u from 
obtaining exact value for duck food 
plant , we believe that, by con idering 
both the abundance and use of such 
plant , it i p ssible to a certain more 
nearly the true value of plant specie , as 
food for waterfowl, than by u ing data 
ba ed lely on the quantity of the 
item taken. 

What determine the duck food value 
of various aquatic plants? Logically, 
availability and palatability are two 
most important fact r . 'vVe believe 
evidence di cu sed in thi paper how 
that availability, a mea ured b food 
yield and acce. sibility, determine the 
value of most plant generally con-
. idered a source · of duck food. It 
shoLJd be noted that, in many specie. , 
seed yield and value g hand in hand. 
In ther pecies, depth of water evi-
dent! affect availability, through mak-
ing the food source le , , or more, ea ily 
acce sible to duck ·, e pecia]ly dabbling 
duck . v"\ e may tentatively a ume that 
eed of moi t- oil plant are more 



( LI.I. ·01 . . . TURAL HI T RY 'R\'E B LLETI. • l'ol. 22, Art. 

ea ·ilv acces 1blc than tho. c f emergent 
aquari .-pel:ies, which, in turn are mor 
ea. ih· acce. ible than fl ating or ub-
merged aquati plant . 

That palatability play an important 
r ,le in determining the i od value f 
-ev ral ·pecie i. al ·o quite e ident. 
Both the pi ker lw ed and butt nbu. h 
are heavy e d producers, the fi d i. 
fairlr ac e · ·ibl , and yet the inde value 
i. I w. :\merican lotu i a medium fo d 
pr ducer, the :eed are fairly acce ible, 
and vet th inde value i · verv low. 
PaJatability in the . peci~ mu t ·be the 
op rative factor in determining their 
meager u e by duck . 

Summary 

mpl ying data ha ed upon the oc-
currence of the plant. a well a upon 
con umption by duck of the plant 
part , the writer have attempted to 
pre:ent an accurate apprai aJ of the 
value of certain plant of the IJlinoi 
River valley a waterfowl food ource . 

B_ , di iding the per cent of u e made 
br duck by the per cent of abundance 

f the important aquatic and m i. t- oil 
plant. ccurring in the alley, it wa 
p .. ible t ecure an index figure of 

alue for each f the. e , pecie . \\ e be-
lie e the figure obtained to be far more 
reliable indi ator. of value a waterfowl 
fo d ource: than are data deri ed only 
from plant part taken by the bird . 

umerou factor , . uch a eed pro-
duction and acce ibilit., influence 
. ource.· of upply that vary with the 
habitat and year. Taking cognizance 
f the e variable (a · di cu ed under 

each pecie ) make it eem advisable 
t judge the value of each pecie on the 
figure, f, r no ne year, but to u e the 
inde alue for the ear or year that 
seem mo ·t t pica!. Thi procedure 
place 25 plant · or gr up in the follow-
ing descending order of alue a. food 
f, r duck in the lllin i. River valley, 
193 , 19 9 and 1940. Table 3, 4, 5 
and 6 gi e c mparative aJue for pe-
cies and year . . 

Ex ELLE T D CK F oo o RCE 

1. Rice cut-gra ·, Leersia oryzoides 
(Linnaeu ) wartz 

2. Walter'. millet, Ed,inod,loa Wal-
leri (Pur h) a h 

3. Wild and J a pane e millet , Echi-
nochloa crusga/li (Linnae ) Beau 

oi. and E . .frummtaua (Ro 
burgh) Link, re pecti el 

4. Moi t- oil martweed 
a. Lar e eed mart eed Polygo-

n um pennsylvanicum Linnaeu 
b. odding mart eed, Polygo 

num lapalhijolium Linnaeu 
c. wamp martweed, Polygonum 

hydropiperoides Michau 
d. Mi cellaneou , Po/ygonum pp. 

5. utgra e 
a. Chufa, Cyperus euulentus Lin 

naeu 
b. Red-rooted cyperu , Cyperus 

erythrorhizos Muhlenberg 
c. tra w-colored cyperu , Cyperus 

strigosus Linnaeu 

Gooo D CK Fooo o RCE 

6. Giant bur-reed, parganium eury-
carpum Engelmann 

7. Coontail, Ceratophyllum demersum 
Linnaeu 

8. Teal gra , Eragrostis hypnoidu 
(Lamarck) Britton, Stern 
Poggenberg 

9. Duck potato agillaria latifoli11 

10. 

11. 

Willdenow 
Mar h martweed, Polygonum 
Muhlenbe1-gii (Mei ner) Watson 
Longleaf pondweed, Potamogeto11 
americanus Chami o & chlech-
tendal 

FAIR D CK Fooo So RCE 

12. Buttonbu h, Cephalanthus oui-
dentalis Linnaeu 

13. pike rushe , principal1y Eleo. 
charis paltu11·is (Linnaeu ) Roemer 

chulte 
14. \\ ater hemp, dcnida tubercu/a/11 

Moquin-Tandon 
15. Marsh cord gra s, "partina Mic!,.. 

au.\·iana Hitchcock 
16. ago pondweed, Potamogelon pee-. 

tinalus Linnaeu 
17. White waterlily, Castalia tu/Je-

1'osa (Paine) Greene 



May, 19-1-3 B ELLROSE & AN DER 0 

PooR DucK Fooo So RC£ 

18. River bulrush, Scirpus jluuiatifis 
(Torrey) Gray 

19. American lotus, Nelumbo futea 
(Willdenow) Persoon 

20. Pickerelweed, Pontederia cordata 
Linnaeus 

21. Marsh mallow, Hibiscus mifitaris 
Cavanilles 

22. Southern naiad, ajas guadalu-
pensis (Sprengel) Morong 

DucK Fooo PLANTS 433 

23. Wild nee, Zizania aquatica Lin-
naeu 

24. Small pondweed, Potamogelon P"-
sillus L innaeu 

25. Long-leaved ammannia, Amman-
nia coccinea Rottboell 

It houJd be noted that the five lead-
ing plants, or group , are moist- oil 
pecie . The moist- oil plan ts a a group 

are better eed-yielders than the truly 
aquatic plants and their eed are more 
readily available to most duck . 

LITERATURE CITED 

Bellrose, Frank C., Jr. 
19-11. Duck food plant of the Illinois River 

valley. Ill. at. Hi t. urv. Bui. 
21 (8) :237-80. Fronti . +35 figs. 

Bellrose, Frank C., and Harry G. Anderson 
1940. Preliminary report on availability 

and u e of waterfowl food plants in 
the Jllinoi River valley. Ill. Nat. 
Hi t. Surv. Biol. 1otes 15. 14 pp. 
(Mimeographed. ) 

Cottam, Clarence 
1939. Food habits of orth American div-

ing ducks. U. S. Dept. Ag. Tech. 
Bui. 643. 139 pp., i llu . 

Martin, A. C., and F. M. hler 
1939. Food of game duck in the United 

States and Canada. U. . Dept. Ag. 
T ech. Bui. 634. 156 pp., illu . 

McAtee, W. L. 
1918. Food habit of the mallard du ck in 

the nited States. U. S. Dept. Ag. 
Bui. 720. 35 pp., illus. 

1939. Wildfowl food plants. Collegiate 
Press, Inc., Ame, Iowa. 14lpp., illus. 

Pirnie, Miles D. 
1935. Michigan waterfowl management. 

Mich. Dept. Con ., Lansing. 328 pp., 
illus.