Near-infrared evaluation of wet mink diets

Are fur farms a potential source of persistent organic pollutants or mercury to nearby freshwater ecosystems?

Farming of carnivorous animals for pelts potentially contaminates nearby ecosystems because animal feed and waste may contain persistent organic pollutants (POPs) and metals. Mink farms in Nova Scotia (NS), Canada, provide mink with feed partially composed of marine fish meal. To test whether mink farms potentially contribute contaminants to nearby lakes, we quantified organochlorine pesticides (OCP), polychlorinated biphenyls (PCB), and total mercury (THg) in mink/aquaculture feed, waste, and sediment collected from 14 lakes within rural southwest NS where mink farms are abundant and have operated for decades. Mercury, PCBs, dichlorodiphenyltrichloroethane (DDT), hexachlorocyclohexane (HCH), and dieldrin were present in mink/aquaculture feed and mink waste, indicating they are potential contaminant sources. Lakes with mink farms in their catchment exhibited significantly higher THg_flux than lakes downstream of mink farming activity and reference lakes (p < 0.0001) after the intensification of mink farming in 1980, indicating mink farming activity is likely associated with increased lacustrine THg_flux. Sedimentary ƩPCB_flux was elevated in lakes with mink farms in their catchments, suggesting possible PCB contributions from mink farming, local agriculture, and atmospheric deposition. Elevated ƩDDT in lakes near mink farms relative to reference lakes suggests a possible enrichment related to mink farming, although mixed land use and historical DDT usage related to forestry in the region complicates DDT source attribution. Maximum dieldrin_flux and HCH_flux in lake sediment occurred coeval with peak worldwide usage in the 1970s and are unlikely to be associated with local mink farming. Lakes with mink farming activities in their catchments were associated with increased THg_flux, ƩPCB_flux, and possibly ƩDDT_flux, suggesting a possible connection between marine fish meal, fur farms, and aquatic ecosystems in NS.

Feasibility of diode-array instruments to carry near-infrared spectroscopy from laboratory to feed process control

Near-infrared calibrations were developed for the instantaneous prediction of the chemical and ingredient composition of intact compound feeds. Two rather different instruments were compared (diode array vs grating monochromator). The grating monochromator was used in a static mode in the laboratory, whereas the diode-array instrumentbetter adapted to online analysiswas placed on a conveyor belt to simulate measurements at a feed mill plant. Modified partial least squares (MPLS) equations were developed using the same set of samples analyzed in the two instruments. Sample set 1 ( N = 398) was used to predict crude protein (CP) and crude fiber (CF), while sample set 2 ( N = 393) was used for the prediction of one macroingredient (sunflower meal, SFM) and one microingredient (mineral-vitamin premix, MVP). The standard error of cross-validation (SECV) and the coefficient of determination (R2) values for CF were better using the monochromator instrument. However, results obtained for CP, SFM, and MVP using the samples analyzed in the diode-array instrument showed similar or even greater accuracy than those obtained using samples analyzed in the grating monochromator. The excellent predictive ability [R2> 0.95; RPD (ratio of standard deviation to SECV) > 3] obtained for CP, CF, and SFM opens the way for the online use of NIRS diode-array instruments for surveillance and monitoring in the manufacture, processing, and marketing of compound feeds. R2, RPD, and SECV values for MVP showed similar performance for both instruments. Although RPD values did not reach the minimum recommended for quantitative analysis, results are encouraging for an ingredient present in feed compounds in such very low amounts.

Utility of near-infrared reflectance spectroscopy to predict nutrient composition and in vitro digestibility of total mixed rations

Total mixed ration (TMR) samples (n = 110) were analyzed for dry matter (DM), crude protein (CP), soluble CP, neutral detergent fiber (NDF), NDF CP, starch, ash, fat, total ethanol-soluble carbohydrate, and nonfiber carbohydrate (NFC). Rapidly and slowly degraded and undegraded in situ CP fractions and in vitro DM, organic matter, and NDF digestibility were determined on each TMR. The TMR were scanned using near-infrared reflectance spectroscopy (NIRS); spectra were retained with NIRS calibration and cross-validation statistics were determined using partial least squares regression methods. The CP, NDF, starch, in vitro DM, and in vitro indigestible NDF contents of TMR were predicted by NIRS with good degrees (R2 >0.85) of accuracy with proportionally low standard errors of prediction. Moderate utility of NIRS to predict the NFC (R2 = 0.83) and fat content (R2 = 0.81) of TMR was observed. Rapidly, slowly, and undegraded in situ CP fractions in TMR were not well predicted by NIRS. Similarly, soluble CP, NDF CP, total ethanol-soluble carbohydrate, and in vitro NDF digestibility (% of NDF) were not well predicted by NIRS. Ratios of nutrient range to reference laboratory method error were calculated and found to be positively related (R2 = 0.84) to NIRS predictability of a given TMR nutrient, suggesting some laboratory procedures were not precise enough to yield suitable NIRS predictions. Data suggest that NIRS has utility to predict basic nutrients such as CP, NDF, starch, NFC, and fat in TMR. However, difficulty was observed using NIRS in predicting key biological nutrients in TMR such as in situ CP fractions and in vitro NDF digestibility. Difficulty of NIRS in predicting these nutrients is related to the level of reference method error in relationship to the range of nutrient values in TMR, but other sources of prediction error may exist.

Near-infrared spectroscopy for measurement of total dietary fiber in homogenized meals

Near-infrared (NIR) reflectance spectroscopy was investigated as a method for prediction of total dietary fiber (TDF) in mixed meals. Meals were prepared for spectral analysis by homogenization only (HO), homogenization and drying (HD), and homogenization, drying, and defatting (HDF). The NIR spectra (400-2498 nm) were obtained with a dispersive NIR spectrometer. Total dietary fiber was determined in HDF samples by an enzymatic-gravimetric assay (AOAC 991.43), and values were calculated for HD and HO samples. Using multivariate analysis software and optimum processing, partial least squares models (n = 114) were developed to relate NIR spectra to the corresponding TDF values. The HO, HD, and HDF models predicted TDF in independent validation samples (n = 37) with a standard error of performance of 0.93% (range 0.7-8.4%), 1.90% (range 2.2-18.9%), and 1.45% (range 2.8-23.3%) and r(2) values of 0.89, 0.92, and 0.97, respectively. Compared with traditional analysis of TDF in mixed meals, which takes 4 days, NIR spectroscopy provides a faster method for screening samples for TDF. The accuracy of prediction was greatest for the HDF model followed by the HD model.