High-resolution observation of phytoplankton is critical to understanding their responses to fine-scale environmental variability, short-term disturbances, and changing aquatic ecosystems. However, near-real-time analyses of natural waters rarely encounter "clean" phytoplankton samples. Suspended sediments, detritus, aggregates, and non-autotrophic organisms occupy the same sampling volume, obscuring phytoplankton signals and complicating interpretation of bulk optical or single-modality measurements. Thus, a key challenge for high-resolution plankton monitoring is not merely increasing sampling frequency, but selectively resolving phytoplankton within complex, particle-rich backgrounds.
We present OptoSieve, a multimodal imaging technology for high-content observation of suspended phytoplankton. OptoSieve simultaneously acquires co-registered color scattering and multispectral autofluorescence images targeting chlorophyll a (Chl a), phycocyanin (PC), and phycoerythrin (PE). Each particle is characterized by morphology, intrinsic color, pigment composition, and spatial pigment distribution at the single-cell level. Combining pigment-specific fluorescence with structural scattering enables distinction of Chl a-containing phytoplankton from non-fluorescent particles and separation of pigment-based functional groups.
Validation with cultured microalgae demonstrated three capabilities. First, multimodal imaging discriminated morphologically similar but pigmentarily distinct species (Chlorella, Microcystis, Porphyridium) via co-registered PC/PE/Chl a fingerprints. Second, the system resolved sub-population physiological heterogeneity: under osmotic stress, Porphyridium cruentum exhibited differential swelling and selective PE dilution—a transient response invisible to bulk fluorometry. Third, white-light scattering redundancy ensured robust morphological imaging despite wavelength-specific pigment absorption.
Applied to untreated samples from a pond, a river, and coastal seawater, OptoSieve resolved particle composition across water bodies. Analysis of >1000 image sets per sample revealed ecologically coherent patterns: the pond showed highest phytoplankton abundance and PC-bearing fraction (consistent with cyanobacterial dominance); the river contained abundant non-phytoplankton particulates; coastal water had the lowest particle load. Single-cell observations further identified candidate taxa—including a euglenoid with a distinctive eyespot, spiral cyanobacterial filaments, and cryptophytes with asymmetric intracellular PE distribution—via combined morphology and pigment signatures.
We will report the development of this novel technology and expect to transition it into a deployable instrument, making it broadly available for process-oriented understanding by plankton ecologists studying fine-scale hydrographic features and short-term disturbances.