188 
In analyzing the various habitat types to determine whether the individual 
types also had a consistent ratio, it was found that four types in 1951 were consistent 
-- large lakes, small lakes and ponds, irrigation canals, and irrigation ditches. 
In 1952, five types showed a consistent ratio -- irrigation canals, irrigation ditches, 
river, creeks, and small lakes and ponds (Table III). The other types not mentioned 
above had too much variation between coverages to be consistent. Therefore, on 
those areas wherein the ratio was found to be consistent it would be possible, given 
the same observer, to multiply the results of an air coverage of a given type by the 
ratio for that type to obtain an estimate of the ground count. As can be noted in 
Table III, the ratio not only varies in size between the two years, but also some 
types which were consistent one year were not the next. This can also be attributed 
to the difference in observers. 
Table II - Comparison of Air and Ground Counts by Coverages, 195l.and 1952 



Coinparison Date ne Observed Territories Ratio: 
951 Air Ground Ground/Air 
Air II: Ground IV Mid -May 78 161 2.064 
Air III: Ground VII Late June | 64 134 2.093 
and VIII (average) 
Air IV: Ground IX Mid-July 40 101 2.525 
Average Ratio - 1951 2.231 
Chi-square = .8487 Good fit - average ratio may be used, 
Chi-square .5= 1.386 
1952 
Air I: Ground II Late April 96 149 1,55 
Air II: Ground III Early May 86 130 1.51 
AirlII: Ground IV Late May 67 119 1,78 
Air IV: Ground V Early June 71 104 1.46 
Air V: Ground VII Early July 60 80 1,33 
Average Ratio 1952 1.52 
Chi-square = 1.704 Good fit.- average ratio may be used. 
Chi-square .5 = 3.357 

It was determined that in 1951, pairs on drain ditches were most difficult 
to see, while those on irrigation canals were easiest observed. In 1952, drains were 
again the most difficult area to observe; however, this year small lakes and ponds 
afforded easiest observations. It appears that aerial counts on small lakes and ponds 
can tally more than ground counts, hence the reason for a ratio less than 1 in Table III. 
The general trend for both years was that aerial counting was easier on those areas 
where high-growing vegetation was not present to interfere with the actual observations, 
and most difficult on the habitats with heavy permanent cover as drains, rivers, and 
creeks. 
