CANVAS-BACK DUCK FOOD. 
H. L. SKAVLEM . * * 
**Abstract of a paper read at the April meeting of the Wis- 
consin Natural History Society (1904) and published in full in 
The Northwestern Sportsman, 1905, pp. 161-164, 236-240. 
Ornithologists have been accustomed for many years to copy 
Wilson's description of food habits of the canvas-back duck with 
reference to its fondness for "wild-celery," Vallisnerta spiralis, 
and the good effect of the same on the flavor of the flesh. The 
writer shot canvas-back duck on Lake Koshkonong for a quarter 
of a century, believing that the Vallisneria was the bait which at- 
tracted the birds to the lake, but in locating the celery beds found 
that the duck usually chose other spots where no celery was to 
be found, although examination of their stomach always revealed 
"buds" having a strong taste of celery presumably derived from 
the Vallisneria. Very often they would settle on the great sub- 
merged meadows of pondweed (Naidaceae) locally called moss or 
sedge. 
During the season of 1895 the writer discovered in examin- 
ing masses of this floating pondweed that the so-called "celery- 
buds" were none other than the terminal buds of this plant. This 
explained their indifference to the Vallisneria and fondness for 
localities where the pond-weed (Potamogeton pectinatus) was 
abundant since the latter forms the real food of the ducks. 
An examination of the literature on the subject reveals a sur- 
prising number of incorrect assumptions and erroneous state- 
ments on the part of various authors, regarding the food of the 
canvas-back. Wilson states that the ducks feed on a plant "said 
to be a species of Vallisneria" growing on fresh water shoals, and 
that they dive to tear out and obtain the roots which have the ap- 
pearance of small celery. It appears that Wilson had in mind the 
common "eel-grass" (Zostera marina) and not Vallisneria at all. 
Nuttall (Amer. Ornith., Ill p. 29) corrects Wilson's error and 
states that the birds feed on the tender parts of the root of the 
"sea-wrack." Audubon in his great work overlooks Nuttall's 
correction and repeats Wilson's remarks on Vallisneria. Chap- 
man (Hand-book of Birds of Eastern N. Amer., 6th Edit. (1903), 
p. Ill) describes the favorable effect of wild celery on the flavor 
of the canvas-back. If for "wild celery" we substitute Naidaceae 
(which includes Zostera (eel-grass), Ruppia (ditch-grass) and the 
species of Potamogeton (pond-weed), Zizania (wild-rice), Hydro- 
cardiacme (to which Vallisneria belongs) we have the pro- 
per list of food plants. At Lake Koshkonong it was found that 
"celery-buds" (Potamogeton pectinatus) formed from 60% to 
80% of the ducks' food. 
108 
