The research team examined gastrointestinal tract samples pooled from cultured specimens of both species raised in inland and semi-coastal pond systems in Kerala state. In total, the researchers identified 307 individual microplastic items across the sampled animals. Average contamination loads came out at 0.709 plus or minus 2 particles per gram of gastrointestinal tract tissue for the giant freshwater prawn, and 1.015 plus or minus 2 particles per gram for the white leg shrimp, meaning the Pacific white shrimp, though the smaller animal, carried the heavier per-gram burden. Particle sizes ranged from below 250 micrometers up to 5 millimeters, but the distribution was far from uniform. The dominant size fraction fell between 500 micrometers and 1 millimeter, a range that overlaps with particles small enough to be mistaken for food by grazing and filtering crustaceans yet large enough to lodge in digestive structures. Visually, the story was equally consistent: blue-colored microplastics with a fiber morphology dominated both species, accounting for 47.31 percent of recovered particles in the giant prawn and 46.28 percent in the white leg shrimp. Fibers of this kind are widely associated with the degradation of fishing nets, ropes, synthetic textiles, and, critically in an aquaculture context, the polymer-based materials used in pond linings, aeration equipment, and feed packaging.

Identifying what the particles were actually made of required more than a microscope. The team deployed attenuated total reflectance Fourier transform infrared spectroscopy, known as ATR-FTIR, alongside a confocal Raman microscope integrated with atomic force microscopy. These complementary techniques interrogate the vibrational fingerprints of polymers, allowing researchers to match observed spectra against reference libraries and assign each particle to a specific plastic chemistry. The analysis confirmed five polymer types: polyethylene, polystyrene, acrylonitrile-butadiene-styrene, commonly abbreviated ABS, polycarbonate, and polypropylene. The presence of polyethylene and polypropylene is unsurprising, since these are the most produced plastics on Earth and permeate packaging, tubing, and agricultural films. But the detection of ABS and polycarbonate carried more weight, because these engineering polymers are ranked among the more hazardous plastic families in published chemical-composition-based hazard assessments. Polycarbonate raises particular concern due to its association with bisphenol A monomers, while ABS can leach styrene and acrylonitrile residues.

To translate these polymer identities into a measure of danger, the researchers calculated a Polymer Hazard Index, or PHI, for each species. This index weights the observed polymer mixture by the toxicity scores assigned to each plastic type in the widely cited Lithner hazard ranking, producing a single number that reflects not just how much plastic an animal has ingested but how toxic that plastic is likely to be. The giant freshwater prawn scored higher on this metric, with a PHI of 29.06 compared with 22.92 for the white leg shrimp, a direct consequence of the ABS and polycarbonate found in its digestive tracts. On polymer toxicity alone, the prawn appeared to be the riskier meal. Yet the authors recognized that hazard rankings capture only one dimension of exposure, and this is where the study makes its most distinctive contribution.

The team then constructed an integrated Pollution Risk Index, or PRI, which folds together three independent variables: the total microplastic load in the animal, the Polymer Hazard Index, and the shape profile of the ingested particles. Shape matters because fibers, with their high aspect ratios and needle-like geometry, are considered more likely to cause physical irritation, penetrate tissue, and persist in the gut compared with fragments or films. When all three factors were combined, the ranking inverted. The white leg shrimp registered a PRI of 15.12, exceeding the giant prawn’s 13.72, because its higher ingestion rate of particles outweighed the prawn’s more hazardous polymer cocktail. The result is a cautionary lesson in risk assessment methodology: single-metric approaches can mislead, and a composite index that accounts for load, chemistry, and morphology simultaneously produces a fundamentally different picture of which farmed species poses the greater ecological and food safety concern.

The machine learning component added a further layer of analytical rigor. The researchers trained a random forest model, an ensemble method that builds hundreds of decision trees on random subsets of the data and aggregates their votes, to classify microplastic risk levels from the contamination dataset. The model achieved a classification accuracy of 91.3 percent, and, crucially, its internal feature importance analysis revealed which variables carried the most predictive power. Three emerged as key predictors: the species of the animal, the microplastic density in its tissues, and the polymer type of the ingested particles. Random forests have a long track record in ecological classification, valued for their robustness to noise, their resistance to overfitting, and their ability to capture nonlinear interactions among predictor variables that traditional statistical models miss. Their application here suggests that risk profiling in aquaculture could eventually become predictive rather than merely descriptive, allowing regulators to estimate contamination risk from a handful of measurable parameters without exhaustive particle-by-particle screening for every batch of farmed product.

The implications extend well beyond the ponds of Kollam. Global aquaculture now supplies more farmed aquatic animal protein than wild capture fisheries, and shrimp and prawn farming is among its fastest-growing and most export-oriented sectors. India is one of the world’s largest shrimp producers, and both species examined in this study anchor major industries: white leg shrimp dominates international seafood trade, while the giant freshwater prawn is a staple of domestic consumption and regional markets. Microplastics in farmed crustaceans therefore sit at the intersection of food security, rural livelihoods, and export economics. Previous research has established that farmed shrimp can ingest microplastics from feed, pond water, and sediments, and that plastic particles can accumulate across grow-out cycles, sometimes in an age-dependent fashion. The Kerala findings add a tropical freshwater dimension to a literature that has traditionally focused on marine systems, and they underscore that even managed, semi-closed production environments are not sealed off from plastic pollution.

The presence of microplastics in the gastrointestinal tracts of farmed prawns also raises questions about the pathway to human exposure. In crustaceans destined for market, the digestive tract is not always removed before consumption, particularly for smaller shrimp that are eaten whole or lightly processed. Experimental work has shown that crustacean digestion can fragment larger microplastics into nanoplastics, potentially increasing their bioavailability, and laboratory studies have documented physiological effects in shrimp ranging from altered gut microbiota to reduced immune competence and heightened vulnerability to pathogens such as white spot syndrome virus. Microplastic particles also act as vectors for other contaminants, adsorbing heavy metals, persistent organic pollutants, and antimicrobial residues from the surrounding water, which means the health burden of an ingested fiber may exceed that of the plastic itself. The authors of the current study note that their combined chemical and morphological dataset provides exactly the kind of granular information needed to begin estimating realistic dietary exposure levels for consumers.

What distinguishes this research is its methodological architecture. Rather than presenting abundance figures in isolation, the team layered multiple analytical instruments, two complementary risk indices, and a supervised learning classifier into a single workflow that could, in principle, be replicated in any aquaculture region. The combination of ATR-FTIR and confocal Raman spectroscopy with atomic force microscopy addresses a persistent weakness in the microplastics field, where visual identification alone has been shown to overestimate particle counts and misclassify non-plastic materials. Spectroscopic confirmation, as applied here, raises confidence that every counted particle is a genuine synthetic polymer, and open-source spectral libraries are making such approaches increasingly accessible to laboratories in producing countries. The random forest layer, meanwhile, converts static contamination snapshots into a predictive tool, aligning aquaculture monitoring with a broader movement in environmental science toward machine learning-assisted pollution assessment, from stormwater microplastics modeling to water quality criteria prediction.

The study received no dedicated external funding but benefited from infrastructure support under India’s DST-FIST program at Fatima Mata National College in Kollam, and the sampling relied on cooperation with local aquaculture operations in the region. The authors emphasize that their framework, combining polymer-resolved contamination data with integrated risk indices and machine learning classification, is intended as a transferable template for microplastic risk assessment in cultured aquatic food systems anywhere. As plastic production continues to climb and as demand for farmed aquatic protein grows in parallel, tools that can rapidly profile contamination and rank risk across species, farms, and regions will become essential for food safety regulators, certification schemes, and the aquaculture industry itself. The Kollam prawns may be small animals, but the analytical machinery now pointed at them signals where the science of food contamination monitoring is headed.