12 



TEAL GRASS, Eragrostis h ypnoldes (Lain.j BSP. , V.ATER liEMP' 

 or FIGVi^'EED, Acnida tuberculata Moq.', and SMARTw'EEDS, I^ lyr'^'i'iu^ ^ s::^.. ■ 

 tT^E.or" cliT.z'.'.'vyil^x-i.'.r'.j-i) f?.c:-i^";: tiie nutgrasses, were very scarce in 

 the'lTrinois"^RTv"er~regTon' in 1938, because of unfavorable water lev- 

 els.- Yet tl.'ey formed, respectively, 3.16, 2.17 and 1.63 per cent of 

 the native food plants taken from the 1,147 duck stomachs examined. 

 Vve believe that miany seeds of these moist soil plants were deposited 

 in 1936 and 1937, when the plants were abundant, and that waterfowl 

 feeding on the barren mud .flats in 1938 picked up the seeds. A 

 ■sm.all sample of mud collected from barren mud flats in 1940 yielded 

 550 seeds cl v/ater hemp. ' , - 



MARSI-I SMARTWEED, P olygon!^ MulxlenbergJi (I.Ieisn. ) \cats., 

 rated higher than longleaf pondweed, but below coontail, for the 

 four areas as a whole; its index rating was 0.66, about one-third 

 that of coontail. This smartv\feed rarely produces seed when growing 

 on dry soil. In 1938, seed was produced by all beds in the areas 

 considered.. Marsh smartweed was of more value in the Crane Lake 

 area, where a greater scarcity." of natural foods existed, than in .the 

 other ' areas. 



LONGLEAF POKDV.'EED, Pota moge ton am.ericanus C. & S.", liad an 

 index rating for the four areas of O.'liT, indicating that it v/as 

 about two-thirds as valuable as -marsh smartweed. This pondweed was 

 •scarce in the Cuba Island, Crane Lake and Duck Island areas. At 

 Lake Chautauqua, where it ranked fourth in preference, it was the 

 most abundant species, due to stabilized water levels there." In the 

 Crane' Lake region, it rated as th'e second best duck food plant. The 

 high p'ercentage of longleaf pondweed recorded, for the gizzards col- 

 lected "at Duc.k Island was evidently due to the f.act that many of the 

 ducks killed there had previously fed in nearby Lake Chautauqua. 



GIANT BURREED, Sparganium eurycarpu m Engelm. , had an ex- 

 tensive' distribution, but- its occurrence-was limited to small, scat- 

 tered p.atches. 'Ae believe' that it forined about 1.0 per cent of 'the 

 vegetation. This would indicate that burreed was not preferred to 

 longleaf pondweed b.ut was .a better, food than white vi/aterlily or duck 

 potato. 



BUTTONBUSK, Cephalanthus occidentalis L. , was approximately 

 as abundant in the Illinois River region in 1933 as was river bul- 

 rush. It was, hovvever, seldom within the mapping area, occurring for 

 tlie most part inside the sliore line.. Seeds of the buttonbush 

 amounted to- 3.49 per cent of the native plant foods fo'uhd in the duck 

 gizzards examined. This figure probably entitles it to be ranked in 

 1938 as a better, duck food than wlxite waterlily or duck potato. It 

 should be noted that in the stomachs taken from Duck Island and Cuba 

 Island, where natural foods abounded, little buttonbush seed was 

 found. However, in the gizzards taken from Crane' Lake, where there 

 was a dearth of good, natural food, buttonbusli seed amounted to 8.55 

 per cent of the native plant food's. 



WHITE VvATERLILY, C astalia tuberosa (Paine) Greene, had an 

 index rating for the four areas of 0.26, almost half that of long- 

 leaf pondweed. ' -However, we are hesitant to believe that this water- 

 lily is as valuable as indicated. Its limited distribution and tlie 

 hlghcontent of a few gizzards may have distorted its value. In the 

 Crane Lake region, where the most extensive area of wLiite waterlily 

 occurred and other natural foods were scarce, the index ratirgwas 0. 19. 



