Prediction of Fermentation Parameters During Solid-State Fermentation of Lemongrass Feedstuffs Using NIRS Combined with Spectral Correction Techniques

Authors

• Alis Syarmalis, Master Student of Animal Science, Faculty of Agriculture, Universitas Syiah Kuala, Darussalam-Banda Aceh, 23111, IndonesiaFollow
• Agus Arip Munawar, Research Center for Innovation and Feed Technology, Universitas Syiah Kuala, Banda Aceh, 23111, IndonesiaFollow
• Indra Wahyudi, Research Center for Innovation and Feed Technology, Universitas Syiah Kuala, Banda Aceh, 23111, IndonesiaFollow
• Zulfikar Amir, Research Center for Innovation and Feed Technology, Universitas Syiah Kuala, Banda Aceh, 23111, IndonesiaFollow
• Samadi Samadi, Research Center for Innovation and Feed Technology, Universitas Syiah Kuala, Banda Aceh, 23111, IndonesiaFollow

Author ORCID Identifier

Agus Arip Munawar, https://orcid.org/0000-0003-3951-4386

Samadi Samadi, https://orcid.org/0000-0003-1669-2585

Abstract

Lignocellulosic residues such as lemongrass (Cymbopogon nardus L.) waste have strong potential to be reused as livestock feed, helping improve sustainability and reduce environmental waste. However, their high lignin and crude fiber contents limit digestibility, making pretreatment necessary. One effective approach is solid-state fermentation (SSF) using white-rot fungi to enhance nutritional quality. During SSF, pH and moisture content are key indicators of process performance, but conventional measurement methods are slow and inefficient for real-time monitoring. In this study, near-infrared spectroscopy (NIRS) combined with partial least squares regression (PLSR) was evaluated as a rapid and non-destructive method to predict pH and moisture content in fermented citronella waste. Four spectral pretreatment techniques were compared: Peak Normalization (PN), Mean Normalization (MN), Multiplicative Scatter Correction (MSC), and Savitzky-Golay Smoothing (SGS). Mean Normalization delivered the best prediction accuracy for moisture content (R2 = 0.99; RPD = 22.5), while Savitzky-Golay Smoothing produced the most accurate model for pH prediction (R2 > 0.99). PN and MSC also improved spectral consistency across different fermentation stages. Overall, the results demonstrate that NIRS is a fast, accurate, and non-desctructive tool for monitoring the SSF process of citronella waste, with strong potential for real-time process control in lignocellulosic biomass-based feed production systems.

Recommended Citation

Syarmalis, Alis; Munawar, Agus Arip; Wahyudi, Indra; Amir, Zulfikar; and Samadi, Samadi (2026) "Prediction of Fermentation Parameters During Solid-State Fermentation of Lemongrass Feedstuffs Using NIRS Combined with Spectral Correction Techniques," Jurnal Ilmu Ternak Universitas Padjadjaran: Vol. 26: No. 2, Article 4.
DOI: https://doi.org/10.24198/jit.v26i2.1452
Available at: https://journal.unpad.ac.id/jitunpad/vol26/iss2/4