Protein, Fat and Moisture Analyses of Fresh Fishmeal with an Antaris II FT-NIR Analyzer

Significance of the topic

Fishmeal is a globally important high-protein ingredient for aquaculture and animal feeds. Rapid, accurate measurement of protein, fat and moisture in fishmeal is essential for process control, product quality and economic optimization. Traditional reference methods (Kjeldahl for protein, Soxhlet for fat, and loss-on-drying or Karl Fischer for moisture) are accurate but slow, labor intensive, solvent- and waste-generating, and poorly suited for near-real-time process monitoring. Demonstrating a validated FT-NIR alternative addresses these limitations and enables faster quality decisions, reduced chemical use and safer laboratory conditions.

Objectives and study overview

The study evaluated whether a single FT-NIR measurement using Thermo Scientific Antaris analyzers can replace multiple conventional assays for fresh fishmeal. Objectives were to develop robust calibrations for protein, fat, moisture (water), ash and ammonia-related parameters using production-derived samples, to quantify method performance versus primary reference techniques, and to demonstrate transferability from a laboratory Antaris II system to an Antaris MX process analyzer.

Applied methodology and instrumentation

The analytical protocol used production samples to incorporate real process variability (seasonal and geographic). Key experimental and chemometric choices:

• Sampling and presentation: samples taken directly from production; spectra acquired on Antaris II MDS with a sample cup spinner and petri dish. Pressing samples in the petri dish improved spectral consistency and prediction performance.
• Spectral acquisition: range 9000–4000 cm-1, resolution 16 cm-1, 32 co-added scans, typical collection time ~12 s per spectrum.
• Pretreatment: first derivative spectra with Norris smoothing (segment length = 3, gap = 3).
• Multivariate calibration: Partial Least Squares (PLS) regression using the full spectral range; calibration built from 247 standards representing process variability.
• Transfer: calibration developed on Antaris II lab system and subsequently implemented on Antaris MX process analyzer.

Used instrumentation

• Thermo Scientific Antaris II FT-NIR analyzer with Method Development Sampling (MDS) system, sample cup spinner and petri dish sampling accessory.
• Thermo Scientific Antaris MX FT-NIR process analyzer for process deployment.
• Reference laboratory methods used for calibration targets: Kjeldahl for protein, Soxhlet extraction for fat, and loss-on-drying/Karl Fischer-type moisture determinations; ash and ammonia quantified by standard reference protocols.

Main results and discussion

Calibrations showed strong agreement with reference methods across the analytes studied. Summary of calibration performance (PLS models built on full spectral range, numbers of latent variables varied by analyte):

• Protein: correlation coefficient ~0.96; root-mean-square error of calibration (RMSEC) about 0.57 (units consistent with % points); cross-validation RMSECV ≈ 0.71; calibration range ~68.7–72.3.
• Fat: correlation coefficient ~0.984; RMSEC ≈ 0.27; RMSECV ≈ 0.31; range ~8.3–10.5.
• Ammonia (NH3): correlation coefficient ~0.97; RMSEC ≈ 0.012; RMSECV ≈ 0.015; range ~0.12–0.25.
• Ash: correlation coefficient ~0.935; RMSEC ≈ 0.605; RMSECV ≈ 0.74; range ~12.0–15.7.
• Water (moisture): correlation coefficient ~0.985; RMSEC ≈ 0.242; RMSECV ≈ 0.32; range ~5.6–9.3.

The models achieved high linearity (R > 0.935 for all analytes) and practical prediction errors suitable for routine quality control. Using production-derived standards ensured the calibrations captured process heterogeneity (seasonal and catch-area effects). Simple sample presentation improvements (pressing into petri dishes) reduced scatter and improved model robustness. The chemometric workflow (derivative pretreatment + PLS) effectively compensated baseline and scattering effects common in ground heterogeneous materials.

Benefits and practical applications

Replacing traditional assays with FT-NIR measurement yields multiple operational advantages:

• Speed: spectra collected in seconds and multivariate predictions returned in real time, enabling near-line or inline monitoring and faster process control decisions.
• Safety and sustainability: eliminates hazardous solvents and large volumes of chemical reagents, reducing laboratory waste and operator exposure.
• Cost and throughput: reduced labor and consumables costs, higher sample throughput compared with Kjeldahl and Soxhlet workflows.
• Transferability: demonstrated migration of calibration from laboratory Antaris II to Antaris MX process analyzer supports deployment into production environments.

Future trends and potential uses

Opportunities to extend and strengthen FT-NIR application in fishmeal and related feed matrices include:

• Robustness enhancement: implementation of advanced preprocessing, wavelength selection, and routine transfer protocols to maintain calibration stability across instruments and over time.
• Model maintenance: use of adaptive or incremental calibration strategies to accommodate raw-material shifts (seasonality, species mix) and instrument drift.
• Data-driven augmentation: application of machine learning methods (nonlinear models, ensemble approaches) where spectral–property relationships are complex.
• Process integration: inline FT-NIR combined with process analytical control systems to enable closed-loop process optimization (e.g., moisture control during drying, blend adjustments for target protein/fat).
• Expanded analyte panels: extension to detect contaminants, fatty-acid proxies, or quality indices, and coupling with other sensors (e.g., mid-IR, Raman) for complementary information.

Conclusion

The Antaris FT-NIR approach provides accurate, precise and fast quantification of protein, fat, moisture, ash and ammonia-related parameters in fresh fishmeal. Calibrations developed on production samples and pretreated with first-derivative/Norris smoothing combined with PLS regression showed strong agreement with reference methods and were transferable to a process analyzer. FT-NIR offers a practical alternative to laborious, solvent-intensive reference assays, enabling rapid quality control and improved process monitoring in fishmeal production.

References

1. Wiertz M. Protein, Fat and Moisture Analyses of Fresh Fishmeal with an Antaris II FT-NIR Analyzer. Thermo Fisher Scientific Application Note 51873, 2010.