Ecological and Evolutionary Analysis of Individual Base Data

The probability capturing all individuals in the population is hardly ever one. Unless recapture probabilities are high, consistently similar and show no bias towards any particular sex, age or size-group, this can lead to incorrect inferences of demographic parameters (ref). Mark-recapture models are designed to handle such data. When individuals are uniquely marked, the recapture probability can be estimated from those occasions in which an individuals was not caught but appeared in the data both before and after that occasion. This requires assuming individuals cannot leave the sampled area and return, which is the case in our study system (individuals can only leave over the downstream waterfall, but cannot return upstream).