on the study area in sample period idoes not permanently emigrate from the
study area between sample periods iandi1. Given these two sets of para-
meters, we would model the above encounter history as:
P(0 1 1 1 3 0 | release in period 2)s 2 f 2 s 3 f 3 (1f 4 ).
Thes 2 parameter denotes the probability of survival between sample periods
2 and 3, f 2 corresponds to fidelity during the time interval from period 2 to 3, s 3 and
f 3 correspond to survival and fidelity, respectively, between 3 and 4, and (1f 4 )
is the probability that a bird alive in the study area in period 4 permanently
emigrates before period 5.
As noted above, some field studies involve sampling multiple locations at each
sample period. Birds are caught, radio-tagged, and released at one or more of the
locations, and each location is searched with a radio receiver for marked birds
at each sample period. Such studies permit estimation of the probabilities of
moving among the different study locations. Encounter data from such a study
must specify not only fate with respect to presence in the study system (consist-
ing of all sampled locations), death, and emigration from the study system, but
also location within the study system. Assume a simple system in which birds are
radio-tagged and sampled at two locations, A and B. Encounter histories can be
written with numbers reflecting fate and letters reflecting location. So encounter
history 0 A1 A1 B1 B3 indicates a bird marked in location A at period 2, located
again in location A at period 3, located in location B at period 4 and not located
in either A or B (departed the study system) in period 5 (the notation B3 simply
indicates that the bird was last detected at location B). If we again assume that
detection probability for radio-marked birds within the study system is 1, then
the following parameters are needed to model encounter history data:
siRprobability that a bird in location R at period ithat does not
permanently emigrate the study system survives until period i1;
ψiRSprobability that a bird in location R at period iand alive in the
study system at period i1 is in location S in period i1;
fiRprobability that a bird in location R at period idoes not permanently
emigrate from the study system between periods iandi1.
Given these definitions, we can write the probability for the above example
capture history as:
s 2 Af 2 A(1 2 AB)s 3 Af 3 A 3 AB(1f 4 B).P(0 A1 A1 B1 B3release in location A, period 2)130 |Estimating survival and movement