Behavioral Study: Methods

The Hawaiian Shrimp-Goby (Psilogobius mainlandi) (Baldwin 1972) is endemic to the Hawaiian Islands and forms an obligate relationship with two Alpheus sp. snapping shrimp – Alpheus rapax and Alpheus rapicida. The field study was conducted in Goby Bay on Coconut Island (in Kaneohe Bay, Oahu; Fig 1) during the months of October through December 2002 whereby only A. rapax was found to reside with the P. mainlandii. Goby bay consists of an approximately 100 m * 100 m semi-enclosed bay (Fig 2). It also lies on the leeward (eastern) side of Coconut island so is protected from prevailing trade winds. If winds change direction there is only about a 500 meter fetch due to the presence of the island of Oahu. Because of this, the bay is composed of silty-sandy sediments mixed in with some larger (approx 5 cm) coral rubble. The coral rubble provides some of the framework for the shrimp-goby holes.

Shrimp holes are easy to detect due to the presence of dark anoxic sediment surrounding a small hole. Only crab holes also have black sediment coming from them. They, however, are much larger and easily distinguished. When shrimp and gobies are scared, it takes them about five to ten minutes to re-emerge from their holes and continue activity. However, because it is difficult to determine whether or not the activity is normal while observing them, a small remote UW, black and white bullet camera was employed for the purposes of watching them. The camera was installed in an 18 inch tall plastic ring lacking a bottom or a top (Fig 3). The camera was able to move up or down the ring and stayed in the center. Because of the camera’s ability to move up or down in the ring, the field of view could change depending on the hole size. This allowed for better viewing while on land.

Behavioral observations were analyzed after the video capture was finished. Behaviors observed included a) shrimp bulldozing b) shrimp outside of the hole c) shrimp manipulating material outside of the hole d) goby in the hole. Each shrimp behavior was quantified over a 1000 second period (16.67 minutes). The camera was above each hole for 3 hours so that three individual 1000 second periods could be taken. Initially 8 holes with gobies and 8 holes without gobies were measured. However, only 7 goby holes and 2 shrimp holes were successfully analyzed due to video capture problems and unsuccessful removal of gobies.

To determine what normal shrimp behavior was, I measure shrimp in holes with gobies present. Because no holes were found with out gobies it was presumed that this was the ‘normal’ state. This data was compared to holes that had the gobies removed. Some problems arose when trying to remove the gobies. First, it was difficult to remove gobies without damaging their holes. To remove the gobies it was necessary to block the entrance to their hole, whereby they would freeze allowing relocation. Each holes damage varied. As soon as a hole’s goby was relocated the plastic ring was placed around the hole with the video monitor on it. Sometimes it would take up to two hours for a shrimp to fix their hole. Sometimes they would not fix their hole due to the severity of the damage. Because of this, several observation sessions of shrimp without gobies were thrown out. Only holes with minor damage and shrimp that appeared to be unaffected were used.