Abstract
The wintering areas for humpback whales within the Great Barrier Reef World Heritage Area (GBRWHA) have been poorly defined, mainly because of the large size of the area which prohibits broad-scale surveys. This information gap was addressed by applying predictive spatial habitat modelling using presence-only sighting data from an opportunistic sightings database. The model identified high habitat suitability for breeding humpback whales in the southern GBRWHA, which decreased as latitude decreased. However, predictive habitat modelling is seldom validated and the accuracy of models is often unchecked. We recently validated this predictive model by conducting a dedicated line transect aerial survey that subsampled three regions in the GBRWHA predicted to represent areas of low, medium and high habitat suitability. The distribution and relative abundance of whales was investigated in relation to environmental variables using GIS and generalized additive models (GAMs). Data from the dedicated survey supports the predictive habitat model, with areas of high density closely reflecting areas of high habitat suitability identified by the predictive model. Encounter rates from the aerial survey were highest (0.04 per sq. km) in the southern GBRWHA and lowest (0.002 per sq. km) in the northern GBRWHA, according to un-modelled data. Calving areas were not separate from mating areas, and groups containing calves were distributed throughout the entire GBRWHA within the same range of groups sighted without calves. The area of highest density of whales on the breeding grounds corresponded to an offshore area adjacent to two coastal cities undergoing major port expansions, and within the GBRWHA inner shipping route. There are many proposed and several approved port expansions along the coastline adjoining the GBRWHA. With an associated increase in shipping activity and a rapidly recovering population of whales, ship strikes with breeding humpback whales are likely to be an emerging issue in Australia.