Abstract:

1. A small population of coastal bottlenose dolphins (Tursiops truncatus gephyreus) inhabits the inland waters of Laguna, southern Brazil. A subset of this population interacts with artisanal fishermen. This specialized foraging tactic structures dolphin society into “cooperatives”—participants of the interaction—and “non-cooperatives”—non-participants.

2. Between 2012 and 2015, a bridge was constructed in Laguna over an important core area for dolphins. Photo-identification and georeferenced data collected on boat surveys, conducted both before and during bridge construction, were used to evaluate changes in both “cooperative” and “non-cooperative” dolphin distribution.

3. Changes in dolphin distribution were analysed with kernel densities and hurdle models. A grid of 120 1-km² cells was used to model dolphin distribution and the minimum distance from the grid cell centre to the bridge and to the area of gillnet use were defined as explanatory variables of human activities. Habitat descriptor (depth and distance) from the lagoon margin were also considered in the model procedure.  

4. Dolphin distribution patterns shifted between periods. A core area used by “non-cooperative” dolphins near the bridge construction works disappeared. The effects of habitat descriptor and anthropogenic activity on dolphin distribution also differed between periods. Before bridge construction, the abundance of “non-cooperative” dolphin was higher close to the bridge area (p < 0.05). During bridge construction, the presence of “cooperative” and “non-cooperative” dolphins decreased significantly with the distance from gillnet fishing activity (p < 0.01; p < 0.001).   

5. This study highlights the importance of accounting for individual variations in response when assessing the effects of a habitat disturbance, or when implementing conservation plans.

Keywords: anthropic impacts, bottlenose dolphin, conservation, distribution, habitat disturbance, hurdle models