These sightings were made between May 25 and June 17, 2009, probably, representing the earlier migrants arriving for the breeding season off Brazil, which normally occurs between July and November (Siciliano 1997; Morete et al. 2007). All four sighting records were made opportunistically from different sites of the island (Fig. 1). During two occasions, a closer approximation to the whales was possible using an inflatable boat. Behavioural descriptions were based on the method ad libitum, collected from the boat and underwater observations by free diving.

The bottom contact behaviour observed in cetaceans is mainly related to cleaning or feeding functions. This behaviour was already related to feeding strategy for bottlenose in Bahamas (Rossbach and Herzing 1997), for estuarine dolphins in Brazil (Rossi-Santos and Wedekin 2006) and humpback whales in North Atlantic (Hain et al. 1995). However, the feeding hypothesis is less probable due to the rare records of feeding by humpback whales in Brazilian waters (Alves et al. 2009), considered only as a breeding ground (Zerbini et al. 2004).

On the other hand, there are records of cetaceans rubbing their skin on sand, mud, pebble and limestone substrate near shore, a behaviour associated with moult (Smith et al. 1992) and hygienic functions (Dudzinski 1998). According to Smith et al. (1992), the seasonal variation of water temperature accelerates the turnover of superficial layers in belugas, where the whales rub themselves on the bottom, a process that appears to help the moult. This process is so far unknown in other cetaceans, but might have happened in the humpback whale here reported, since it probably had just migrated from the cold Antarctic waters to the warm tropical Trindade coast. Moreover, the flipper slap behaviour initially observed for this forth individual could be also related to the cleaning/moult process as well. This behaviour is commonly considered a form of communication, in order to maintain acoustic contact between humpback whales (Herman and Tavolga 1980). However, the expected low numbers of whales around Trindade in that period decreases the possibility of communication purposes. Despite such uncertain reason of the behaviour, this novel record comes to increase the behavioural repertoire for the humpback whales in Brazilian waters and may represent an additional cleaning strategy.

Despite several studies conducted in shallow waters of Brazil (e.g. Abrolhos Bank), no information of a similar behaviour was observed among humpback whales yet, where just the breaching behaviour is suggested for cleaning function in Brazil (Morete et al. 2003). Thus, these observations also call attention for the possibilities of underwater behavioural studies at Trindade Island, favoured by the short platform and clear water that it presents. Whether Trindade Island is a cleaning station to the humpback whales in its migratory route, or whether the bottom contact behaviour is just an unusual behaviour, it is still need clarification, and future work is necessary to a better assessment and understanding of their behavioural repertoire.