Marine Biophysics Unit (Satoshi Mitarai)

Unit outline

Satoshi Mitarai

Assistant Professor Satoshi Mitarai

satoshi at


  • Yuichi Nakajima, Researcher
  • Mary Grossmann, Researcher
  • Atsushi Fujimura, Researcher
  • Yukiko Murabayashi, Technician
  • Shohei Nakada, Technician
  • Yuta Nunome, Technician
  • Kazumi Inoha, Technician
  • Koichi Toda, Technician
  • Chihiro Arasaki, Research Administrator



Quantification of connectivity among coral reefs is key to understand global and longer-term changes of the coral ecosystem. Many important marine organisms have a characteristic life history with a relatively sessile adult stage and an obligate pelagic larval stage which can be from many days to several months. Due to the small size and limited swimming abilities of most marine larvae, ocean currents will be the predominant means connecting physically separated populations. Coupled with complex larval life history traits (e.g., different spawning seasons and planktonic larval durations among species), coastal circulation processes create different connectivity for different species, and thereby the different rates of gene flow are expected. Differential ability of corals to migrate will result in further changes to community structure, beyond the immediate effect of selective mortality caused by severe bleaching. In contrast, subpopulations on isolated islands or archipelagoes (e.g., Hawaii and Bermuda, possibly some of Okinawa Islands) may represent genetic isolation for virtually all coral reef species, with little input from other, distant localities. If isolated reefs bleach, recovery is likely to be far slower than in more central, interconnected populations.

In spite of the large body of literature on connectivity in marine populations, quantitative information on connectivity is sadly inadequate. There are no direct observations, other than at very short distances. Methods of estimating dispersal involve an analysis of either the dispersal process (the cause) or of the resultant population structure (the effect). For example, population genetics can provide information on the degree of exchange between populations, but it is aggregated and provides poor resolution of intra- and inter-annual variability with no clear information on specific source-to-destination links. Numerical models of larval dispersal draw on information of water flow and simplified (deterministic) information on larval behavior to obtain predictions of dispersal. These models are promising in that they can yield detailed estimates of connectivity and also resolve dispersal trajectories. However, predicted connectivity has yet to be fully examined against in-situ oceanographic data; furthermore, the theoretical framework behind it is too primitive. Therefore, simulated connectivity has to be viewed with some reservations.

Research Goals

We propose to develop real-time forecasting system for the coastal ocean circulation processes around Okinawa and implement a validated connectivity assessment tool, coupling available modeling and observation techniques. This modeling tool will …

  • 1. Help marine biologists and geneticists estimate rates of gene flow among coral reefs,
  • 2. Enable public agencies to build and evaluate marine protected areas based on population connectivity, and
  • 3. Provide marine ecologists with a modeling system for predicting the global and long-term dynamics of the coral reef ecosystem.

We combine expertise in the modeling of coastal circulation and in-situ physical and biological observations to create novel tools to assess population connectivity as well as a predictive tool to address management scenarios for Okinawa. This research combines modeled ocean currents and life history information (i.e., spawning season and duration, larval development time course, etc.) of the marine species of interest (both corals and coral-reef species) to determine the probability that larvae from a given site are transported to another site via a Lagrangian probability density function modeling procedure, backed up with available in-situ oceanographic data. This will provide a matrix of connectivity (i.e., connectivity for all possible combinations of coral reefs) that can be directly integrated into spatial population dynamics models of reef species to forecast the dynamics and structure of the coral reef ecosystem.

We will create state-of-the-art tools for connectivity analyses for Okinawa that can be implemented for public participation. This facilitates spatial ecosystem-based planning of oceans – not only for the designation of marine protected areas, but also for the risk assessment of pollutant discharge into the coral reef or the implementation of new energy generation facilities. These tools will provide a general model for applying oceanographic knowledge to reef management processes elsewhere around the world. Academically, the proposed work will provide new knowledge about larval connectivity in Okinawa, a validated framework for applying ocean observatory assets to marine biology and modeling environments for the coral ecosystem that can be easily applied to many on-going reef management research projects.