In 1988, a marine mammal sanctuary was established around Banks Peninsula t
o reduce Hector's dolphin entanglement in commercial and recreational gill
nets. We describe the application of a multistate mark-recapture model to d
ata from photographic identification surveys to determine the effect of the
sanctuary on dolphin survival rates. The model allows sightings to occur i
n three geographic strata with free movement between strata. Although we fo
und evidence of both area and time effects on sighting probabilities we fou
nd no evidence that survival rates depended on area or time. Thus, there wa
s no evidence that dolphin survival rates increased following establishment
of the sanctuary. However, power of this test is shown to be low.