A. From April 2001 to March 2010, I was enrolled in the Faculty of Science and Graduate School of Science at Chiba University, from undergraduate to doctoral program.
In April 2010, I joined the Forest Insect Ecology Laboratory, Department of Forest Entomology at the Incorporated Administrative Agency, Forestry and Forest Products Research Institute (now National Research and Development Agency, Forest Research and Management Organization) in Tsukuba City, Ibaraki Prefecture. From April 2011, I conducted research on eelgrass beds in the Macrophyte and Tidal Habitat Group at the Coastal Fisheries and Environment Division of National Research Institute of Fisheries and Environment of Inland Sea (FEIS), Incorporated Administrative Agency, Fisheries Research Agency (now National Research and Development Agency, Fisheries Research Agency) in Hiroshima Prefecture. After that, from July 2011, I pursued laboratory research on biodiversity at the Graduate School of Agricultural and Life Sciences, University of Tokyo, before joining JAMSTEC (Japan Agency for Marine-Earth Science and Technology) in July 2012, where I currently work.
A. Initially, I was interested in JAMSTEC since I had been working with them on a project estimating the biodiversity of Japan, funded by the Ministry of the Environment. Later, when JAMSTEC launched the Tohoku Ecosystem-Associated Marine Sciences (TEAMS) project and announced that it was recruiting personnel, I applied because I wanted to utilize the distribution results of the organisms that I had estimated so far and witness the resurgence of the ecosystem.
A. Firstly, one aspect involves visualizing biodiversity and identifying critical marine areas. For example, I conduct estimates to determine where and what types of organisms are present throughout Japan, examining the species distribution modelling along Japan's coastlines. Based on these estimation of biodiversity, I identify important marine areas around Japan. The Ministry of the Environment also conducts selection processes for Ecologically or Biologically Significant Marine Area (EBSAs) in Japan, referencing the results of this research. Additionally, I analyze not only biodiversity nationwide but also focus on specific regions such as the Southern Honshu or Tohoku area, estimating changes in coral reefs and seagrass/macro algae beds due to climate change. I also analyze the relationship between genetic diversity in coral and its relevance to potential of marine protected areas.
Secondly, I evaluate the "services" and "values" of ecosystems and their socioeconomic relationships. While biodiversity is important, conveying its significance to humans can be challenging. Alongside assessing the intrinsic value of organisms and ecosystems, I evaluate their "economic value" and "value to human society." This includes assessing the value of seafood as food, as well as the value of leisure activities such as tidal flat exploration/clamming and recreational swimming, along with the value of water purification and CO₂ sequestration.
Furthermore, from a socio-economic scenario perspective, I examine how these services might change in the future. For example, I study how changes in Japan's population distribution, such as urban concentration versus rural dispersion, might affect the natural environment.
Finally, I engage in deep-sea biological observation and data extraction. Our center conducts surveys of deep-sea organisms and I, utilizing methods such as deep learning to automatically extract organisms from images of deep-sea taken by Remotely Operated Vehicle (ROV) and Autonomous Underwater Vehicle (AUV). We investigate the numbers and quantities of organisms and monitor changes in their distribution. Such distribution data are also used under the “Cross-ministerial Strategic Innovation Promotion Program (SIP) Developing Innovative Technologies for Exploration of Deep-Sea Resources” for examining methodologies for assessing the impact of seabed resource development. Additionally, I utilize satellite imagery and deep learning techniques to extract spatial extent on seagrass beds, my original area of expertise, not limited to the deep sea.
Click here for the Interview with Dr. Igarashi (link)
A. "Habitat" refers to the living environment of a species. The habitat model estimates the habitat of organisms based on the hypothesis that each species has a range of environments in which it can live, using data on the distribution of organisms and environmental factors. There is actually a debate about whether the hypothesis of "suitable habitat existence" is correct, with another theory suggesting that organisms are distributed randomly in tropical regions. However, it is known that habitat varies depending on environmental factors such as water temperature, on a broad spatial scale especially in temperate regions. For example, it has been observed that skipjack tuna (Katsuo) moves their habitat according to ocean currents. Therefore, habitat models using various statistical methods are employed to predict the distribution of organisms.
The habitats of offshore organisms may be somewhat difficult to imagine, so let me provide an example of the zonal distribution of organisms on seawalls and tidal flats that you can see at beaches in Tokyo Metropolitan wards, including West Beach in Kasai Marine Park, Tokyo Port Wild Bird Park, the Omori Nori Museum, and Hama-rikyu Gardens. When observing the seawall, you can see that barnacles are attached higher up, oysters are attached to areas sometime submerged in water, and sponges are attached even lower. This demonstrates how the organisms inhabiting different depths vary. Due to various factors such as water temperature, predators, and aridity, we can define the "niche," which is the range of environmental conditions where organisms can be distributed, and the "habitat," which is the suitable living environment, based on water depth in this case.
〈 Dr. Yamakita illustrating the distribution of organisms living on the seawall 〉
A. While data analysis may conjure images of programming, more than 90% of the work actually involves data extraction and preprocessing. In the data extraction process, tasks include visually confirming organisms depicted in vast amounts of past underwater footage of the seafloor and extracting information from paper documents to determine the organisms present in specific locations. Constructing the database also requires significant effort, involving tasks such as formatting and cleaning up data. Naming organisms is particularly challenging. In the world of biology, organisms are assigned scientific names as common identifiers. However, with recent advances in taxonomy and the introduction of genetic techniques, scientific names of marine organisms frequently change, species may split, or different species may be grouped together, necessitating adjustments in this regard.
The simplest method in the process of creating habitat models is to indicate the distribution range of organisms (latitude, longitude, or range of water temperature). While various machine learning techniques are now available, we primarily employ estimation using the maximum entropy method, which can analyze data even without explicit absence data, as obtaining such data can be challenging.
〈 Dr. Yamakita illustrating the distribution of organisms living on the seawall 〉
A. The primary challenge lies in the nature of the data. For instance, taking "oysters" as an example, scattering oysters across various tidal flat environments and observing where they survive can reveal suitable habitats. Ideally, when constructing the actual model, randomly sampling across various environments is desirable. However, real-world data often contain various biases and errors, making it challenging to address these issues. Additionally, the information about the target organisms, such as their growth stages or seasonal variations, can vary in response to the environment. Depending on the extent of detailed information available, adjustments may be necessary for variables such as the temporal resolution of the data used.
A. The Earth Simulator produces outputs of global-scale models used in assessments by the Intergovernmental Panel on Climate Change (IPCC) regarding climate change. These outputs are recalculated to achieve higher resolution specifically for the waters surrounding Japan. As a result, daily estimates of ocean currents and water temperatures are generated. While there is a considerable amount of archived daily estimates, we aggregate this data to a level that can be handled by a regular personal computer.
For the habitat modelling of Splendid alfonsino (Kinme-Dai), in particular, we initially couldn't extract data from actual footage in the first year of the project. Therefore, we relied on estimates from information in existing global open-access database of ocean biodiversity record. However, by using data extracted from actual footage this year, we were able to observe the distribution with higher resolution. Additionally, with the availability of "absence data" now, we are planning to consider whether the accuracy of the model will improve and evaluating the extent of variation by using different models.
One challenge regarding offshore areas is that the information on the distribution of organisms is not comprehensive at all. While conducting surveys on ships is important, it is often challenging to access these areas. Therefore, we are considering installing cameras on the seabed for long-term monitoring to verify how marine organisms are behaving and interacting. By doing so, we believe we can build more deductive models based on the ecology and physiology of fish, rather than relying solely on fragmented distribution information.
Furthermore, understanding human interventions poses a challenge. In conjunction with socio-economic scenarios, it will be necessary to model how natural environments are changing and interacting with society in the future.
〈 At the meeting room in the JAMSTEC Tokyo Office 〉A. Tokyo is situated near various types of seas, from Tokyo Bay to the waters around the Izu and Ogasawara Islands. I encourage you to take advantage of this proximity and explore the diverse marine environments, from tidal flats to offshore areas. By actually visiting these locations and getting to know the sea firsthand, as well as utilizing resources like JAMSTEC's E-Library of Deep-sea Images (J-EDI) database for underwater imagery, I hope you can deepen your understanding of the ocean.
The official website of the Tokyo Metropolitan Government