Abstract
Given the expense and time required to monitor marine mammal populations effectively, approaches that fully exploit the resulting data certainly are warranted. We employed a two‐step modeling approach to estimate key demographic parameters, including immigration, from aerial surveys of manatees ( Trichechus manatus latirostris ) in the Northwest management unit of Florida. Abundances of adults and calves were predicted by multivariate adaptive regression spline ( MARS ) models, after accounting for heterogeneous detection rates caused by variable environmental conditions. The resulting predictions were incorporated into a stage‐structured, deterministic model that used an inverse method to estimate parameters with and without immigration. The model without immigration estimated mean survival probabilities of 0.966, 0.923, and 0.794 for adults, subadults and calves, respectively, with a per capita reproductive rate of 0.135. These parameter estimates yielded an overall mean population growth rate of approximately 1.037, which is comparable to rates from mark‐recapture studies. When we added an immigration term that accounted for the greater slope in adult counts since 1999, as identified by the MARS model, the estimated per capita reproductive rate was 0.122, with survival probabilities for adults, subadults and calves of 0.926, 0.920, and 0.833, respectively. These rates were coupled with an estimated mean winter immigration rate corresponding to roughly 5.2% of the adult and subadult population. In this latter scenario, the number of manatees in the core population of the Northwest management unit was predicted to remain constant, with a population growth rate near one, and additional manatees counted during aerial surveys were deemed to be immigrants. While further studies could certainly expound on the potential effects of migrants on population indices, we present this first published immigration estimate for wintering manatees in northwest Florida.