the marked animals suffer greater mortality than unmarked animals between the two sampling periods.
|
Percent mortality
|
0%
|
20%
|
40%
|
|
#SIMULATIONS
|
1,000
|
1,000
|
1,000
|
|
Average estimate
|
10,021.0
|
8,3928
|
6789.2
|
|
#misses
|
48
|
1,000
|
1,000
|
|
Average deviation of estimate
|
308.9
|
1,607.7
|
3,210.8
|
|
Average width of CI
|
1,572.3
|
1,146.8
|
772.8
|
What is the effect of this variable on the population size estimate? On the number of misses? Are there any other interesting patterns?
What can you do to minimize this bias in the real world? Can you solve the problem by increasing S1 and S2?