Damage in the cornfield: blackbird (left), nutria (right). (Photos by F. C. Sehniid, left; Van T. Harris, right) 



April through September, to determine the total 

 impact of feeding at Sand Lake, and (8) study 

 of redwing production in the vicinity of the 

 Refuge. 



Breeding redwing populations in Arkansas. — 

 The breeding male red-winged blackbird popula- 

 tion of the Arkansas Grand Prairie was censused 

 in 1963. For this purpose, the area was divided 

 into two geographical strata, an inner block where, 

 rice culture is intensive, and an outer section where 

 rice growing is less common. Sampling units were 

 selected at random within each stratum : road- 

 side units 150 yards wide, and 5 miles long 

 (272.5 acres), and field units in the form of 10- 

 acre circular plots. 



Counts of males were obtained by the modified 

 roadside technique developed by Dr. O. II. Hewitt 

 of Cornell University, under contract with the 

 Bureau. Results for the entire 900-square-mile 

 Grand Prairie were extrapolated from counts made 

 on the stratified random samples. The breed- 

 ing male population was estimated at 44,000 birds 



in 1963. Average densities were 25 territorial 

 males per 100 acres for roadsides and 6 per 100 

 acres for fields. High roadside densities were at- 

 tributable mainly to the cover types found in the 

 commonly occurring irrigation ditches along the 

 roads. Roadside acreage was only about one- 

 seventh that of the fields, but contained 37 percent 

 of the birds, 63 percent of the birds being away 

 from the roadsides. These densities suggest 

 that efforts to control redwings of the Grand 

 Prairie should be concentrated in the wetland 

 habitat associated with ditches and reservoirs. 



Scare derives. — Crop growers have spent an esti- 

 mated $1 million to $2 million a year on scare 

 devices without obtaining clear information on 

 their value or effectiveness in protecting fields 

 from bird loss. In tests in northern Ohio, an auto- 

 matic exploder and a recorded alarm cry reduced 

 damage substantially for a radius of 600 feet 

 in each case. The tests further showed that 

 neither of the devices decreased in their effective- 

 ness during the test period. 



57 



