March 03, 2010
Poster on monitoring coral spawning using underwater fluorescence camera system was accepted into the PRAGMA 18 workshop.
Coral reefs are a diverse and dynamic environment, and they continue to flourish in activity even after the sun has set. Diving takes large amounts of time and can be treacherous during night time. Because of these issues, it is beneficial to setup long term underwater surveillance systems to monitor the behavior of the reef even during times the human eye can not. Data collected by sensors on the reef can often give indicators that an event of interest is occurring and so can be used to give an early warning to scientists that more direct observations need to be conducted.
During the summer of 2009 I went to the National Museum of Marine Biology and Aquarium in Taiwan as part of UCSD’s PRIME program. The project was a joint collaboration between CALIT2 and Scripps Institute of Oceanography (SIO) in San Diego, and the Taiwanese National Center for High-performance Computing (NCHC) and Museum of Marine Biology and Aquarium (NMMBA). The principle collaborators were Dr. Tony Fountain, the PI for the Open Source DataTurbine Initiative at UCSD CALIT2, Dr. Dimitri Deheyn and his PHD student Melissa Roth, researchers at the Marine Biology Research Division at SIO , Dr. Fang- Pang Lin , a researcher in information technology at NCHC and coral biologist Dr. Tony Fan from NMMBA.
The project used an excitation light and blocking filter to excite the fluorescent protein in Seriatopora Hystrix to capture the fluorescence on video. The fluorescent light allowed hard to see coral features to be visible at night, and was also used to study coral spawning events (which are normally difficult to record since they occur at night and involve coral larvae millimeters in diameter). The video was streamed in real time using Data Turbine to an image recognition program, developed by PRIME student Robert Chen at UCSD, to identify and count the coral larvae and store the results back into DataTurbine in real time.