157 
be used: (1) A direct comparison between broods per square mile and pairs per 
square mile on identical transects; and (2) A comparison between species composi- 
tion of broods and of pairs. Both methods give only an index to relative productivity 
from one year to another. 
In both of the Dakotas certain transects were covered five times in order to 
arrive at as accurate a breeding pair and brood population as possible and also to 
measure changes in the lone drake and pair ratio during the nesting season. Only in 
South Dakota was this type of coverage made in 1949. Results are shown in Table 5, 
Table 5, -- Breeding Pair and Brood Production --- 1949 and 1950 Transects - 
C, D, and E; South Dakota 


Number of Pairs Number of Broods 
May 4-5 May 16-17 July 27 ~— Aug. 8 Pairs per Brood 
Species 1949 1950 1949 1950 1949 1950 
B-w. teal 535 425 90 78 5.94 5.45 
Mallard 156 129 16 13 9.75 9.92 
Pintail 43 57 4 12 10,75 4,75 
Gadwall 219 ; 126 25 10 8.76 12. 60 
Shoveler 36 21 3 2 12.00 10.50 
Redhead 41 26 3 6 13. 67 4.33 
Ruddy duck 44 47 = 4 --- 11.75 
Baldpate 14 3 1 2 14. 00 1.50 
All others 31 16 6 2 5.17 8.00 
Total 1119 850 148 * 129 7.56 - 6.59 
Another way of expressing this success ratio is by way of a formula: 
Number of broods per square mile X Average number per brood . - Index to success. 
Number of pairs per square mile seen on spring transects 
Applying the above formula to 1949 and 1950 data on Transects C. D. & E, in South 
Dakota, we find: 
6.46 X 7.61 (over number per brood; all species) - 1. 006 
48.86 
1949: 
5.63 X 6.90 (over number per brood; all species) = 1. 046 f 4% 
37.12 
1950: 
Thus we find from the fore-going Table 5 and from the formula that although 
there were actually fewer pairs and broods counted in 1950 than in 1949 the success 
of nesting pairs was indicated to be greater in 1950. In other words, the production 
of ducklings per pair was greater in 1950 than it was in 1949. This was in the Waubay 
Hills area only. The formula takes into consideration brood sizes, which was not the 
case in Table 5. Due to a smaller brood average in 1950 the table indicates even 
better success per pair than the formula, 
Applying the formula to data from the Leola Hills of South Dakota we find 
the following: 

2.89 X 8,02 = 3.17 X 7.73 = 
July 12, 1949 a July 24, 19 43.33 
