Skip to main content
ARS Home » Research » Publications at this Location » Publication #125326


item Windham, William
item Lawrence, Kurt
item Buhr, Richard - Jeff
item Smith, Douglas

Submitted to: Near Infrared Spectroscopy International Conference Proceedings
Publication Type: Proceedings
Publication Acceptance Date: 6/10/2001
Publication Date: 9/12/2002
Citation: Windham, W.R., Park, B., Lawrence, K.C., Buhr, R.J., Smith, D.P. 2002. Selection of visible/nir wavelengths for characterizing fecal and ingesta contamination of poultry carcasses. Near Infrared Spectroscopy International Conference Proceedings.

Interpretive Summary: To ensure a healthy and safe meat supply to consumers, the Food Safety Inspection Service (FSIS) established a zero tolerance standard for feces on the surfaces of animal carcasses during slaughter. The reasoning for the zero tolerance is to minimize the likelihood of contamination of meat and poultry with microbial pathogens. Visual observation currently verifies compliance with zero tolerance in meat processing establishments. We investigated visible near infrared spectroscopy as a potential objective method for discriminating between feces and uncontaminated poultry breast skin. The results of this research show that the visible light region can separate feces from the uncontaminated skin. Important visible wavelengths can be derived from this research and implemented in an imaging system for identification of fecal surface contaminates on poultry carcasses. Results may be incorporated with ongoing research to ultimately design a system to detect fecal contamination on poultry carcasses during commercial processing.

Technical Abstract: Ingesta and fecal contamination on a poultry carcass is a food safety hazard due to potential microbiological contamination. A visible/near- infrared (NIR) spectrometer was used to discriminate among pure ingesta and fecal material, breast skin contaminated with ingesta or fecal material and uncontaminated breast skin. Birds were fed isocaloric diets formulated with either maize, milo, or wheat and soybean meal for protein requirements. Following completion of the feeding period (14 days), the birds were humanely processed and eviscerated to obtain ingesta from the crop or proventriculus and feces from the duodenum, ceca, and colon portion of the digestive tract. Pure feces and ingesta, breast skin, and contaminated breast skin were scanned from 400 to 2500 nm and analyzed from 400 to 900 nm. Principal component analysis (PCA) of reflectance spectra was used to discriminate between contaminates and uncontaminated breast skin. Results indicate that visible (400 to 760 nm) and NIR 760-900 nm) spectra can detect contaminates. From PCA analysis, key wavelengths were identified for discrimination of uncontaminated skin from contaminates based the evaluation of loadings weights.