Scientists have developed a method of combining microscopy with machine learning that allows them to follow multiple marine microorganisms throughout their entire lifespan, according to a study published today in eLife. Their work opens up new opportunities to further our knowledge of the lower food web in the ocean.
The majority of life in the ocean is microscopic and constitutes the base of the ocean food web, driving large-scale, global processes – for example, organisms such as phytoplanktons produce around half of the oxygen that we breathe. Despite the importance of this class of life, much of it remains a mystery.
For larger organisms and animals, we have a clearer understanding of their role in the food web, ‘who eats who’, and where biomass (the energy within living organisms) is transferred and how much is lost during consumption. For smaller oceanic microorganisms, observing these parameters is much more difficult, and depends on indirect bulk measurements and ensemble averages. While these methods provide accurate estimates of biomass fluxes in the food web, they are less useful for characterising the small-scale individual interactions that drive large-scale processes and do not allow continuous measurement of the same individual.
“We set out to obtain individual resolution for ecologically relevant microorganisms in the ocean, like plankton, so we can better understand the effect of these tiny organisms on larger-scale processes such as the global carbon cycle,” says first author Harshith Bachimanchi, PhD student at the Department of Physics, University of Gothenburg, Sweden.
Bachimanchi explains that continuous measurements of these microorganisms are possible through microscopy. In particular, holographic microscopy records holograms of cells under investigation in the form of interference patterns that have phase and amplitude information. Bachimanchi and colleagues note that, despite holographic microscopy’s capability to investigate the growth and feeding patterns of microplanktons continuously, this had not been properly put into practice, possibly due to the high cost of data acquisition and processing pipelines.
To remedy this, the team combined holographic microscopy with a deep learning AI program, which serves to circumvent long computational times and, once trained, allows the rapid determination of the three-dimensional position and dry mass (weight without water content) of individual microplanktons over an extended time period. This means that scientists can reliably estimate the growth rates of marine microbes in terms of dry mass increase and cell divisions, as well as measure trophic interactions – ‘who eats who’ – between species. The method is particularly useful for detailing these parameters in micro-zooplankton, one of the most important yet unknown primary consumers in the ocean.
The team tested this new method on nine plankton species of different food web levels that represent the major classes of microplankton. Unlike with previously reported methods, they found that they could follow and weigh single cells throughout their entire lifespan. This is especially useful when detailing micro-zooplankton and feeding events from both mixotrophic species (those that derive nourishment from different energy and carbon sources) and heterotrophic species (those that consume other organic matter) – as it allows feeding events to be quantitatively measured. Using their new method, Bachimanchi and colleagues were able to track and identify ‘prey’ and ‘predator’ cells and closely follow the transfer of mass from cell to cell.
The authors say the method has major advantages over previous techniques such as elemental analysis, including the fact it is non-destructive, minimally invasive and inexpensive. The team have made their coding freely available, with examples and demonstrations to enable other scientists to carry out further investigations.
“The marriage between holographic microscopy and deep learning has proven to be a strong complementary tool in marine ecology,” concludes senior author Giovanni Volpe, Professor at the Department of Physics, University of Gothenburg. “It allows us to accurately determine the three-dimensional position and dry mass of individual microorganisms and better understand their place in the food web. While holographic microscopy has previously been employed in marine sciences, the combination with deep learning algorithms makes it more versatile and much faster for scientists to use.”
The authors have made their code freely available at the GitHub repository: https://github.com/softmatterlab/Quantitative-Microplankton-Tracker
Media contacts
Emily Packer
eLife
e.packer@elifesciences.org
+441223855373George Litchfield
eLife
g.litchfield@elifesciences.org
About
eLife transforms research communication to create a future where a diverse, global community of scientists and researchers produces open and trusted results for the benefit of all. Independent, not-for-profit and supported by funders, we improve the way science is practised and shared. From the research we publish, to the tools we build, to the people we work with, we’ve earned a reputation for quality, integrity and the flexibility to bring about real change. eLife receives financial support and strategic guidance from the Howard Hughes Medical Institute, Knut and Alice Wallenberg Foundation, the Max Planck Society and Wellcome. Learn more at https://elifesciences.org/about.
To read the latest Ecology research published in eLife, visit https://elifesciences.org/subjects/ecology.
