robust design and animals removed from the pop

[Guest]

Hi All,

I an analyzing drift fence/pitfall trap data for amphibians. Specifically, I am marking METAMORPHS (thousands of them). I am trying to calculate temporary emigration. For different reasons, usually death in the buckets, some animals are removed from the population.

My problem is that MARK is ADDING not SUBTRACTING this data when I enter it in this format for robust design:

.
.
.
001000000000 124;
001000000000 -2;
001100000000 -7;
001100000000 21;
001100010000 1;
001010000000 1;
001000000001 1;
.
.
.

Sometimes I just have 1 individual die, but sometimes, esp during peak emergence, I may have >100 die. Therefore, merely removing these dead individuals may bias my data on the low side, esp during this peak emergence.

Any suggestions to fix this problem??

[gwhite]

Froggirltx:

When you say "adding", I assume that you were thinking these animals would not appear in the analysis. Obviously, these animals are and should be included in the analysis. The effect of entering a negative frequency is to remove the animal from the likelihood after that capture, but the animal is included in the likelihood prior to the last capture. In the case of animals only captured once and removed immediately, the effect is to increase M(t+1), and affect the initial capture probability p, but to not affect any further parameters. Thus, the population estimates will include these removed animals, but they will not affect survival rates.

Gary