Soybean is one of the most expensive and limiting feed ingredients in diet formulations; however, in pig farming, it represents the main source of protein. The production and supply of soybean are critical steps due to their environmental impact and feed/food competition for land use. Therefore, research is focusing on finding alternatives to replace soybean partially or totally. However, alternative ingredients should ensure similar growth performance, carcass traits, and meat quality characteristics compared to conventional soybean-based diets. The objective of this review was to evaluate the impact of different alternative protein sources to soybean in pig nutrition and their effects on growth performance, carcass, and meat quality traits. The review process was performed on Scopus^®, and it considered research findings published from 2012 to the present on the Sus scrofa species. Articles without a control group fed with soybean were discarded. The main alternative protein sources identified were other legumes and distillers’ dried grain with solubles (fish and animal proteins, oilseed by- and co-products). Interesting innovative protein sources included by-products from other industries (residues), microalgae and insects. Nevertheless, in dietary formulations, close attention must be paid to address the nutritional requirements, balance the supply of amino acids, avoid anti-nutritional or toxic compounds occasionally present in alternative protein sources, as well as determine the availability of protein feed in specific geographical areas.

This study evaluates the potential of Fourier-Transform Near Infrared Spectroscopy (FT-NIRS) to estimate the chemical composition of fresh natural pastures of Tuscany without previous drying and grinding. Chemical composition of herbage samples is determined by applying usual chemistry. FT-NIRS calibration and cross-validation were developed applying spectra pre-treatment and two statistical models: partial least square regression and principal component regression. The results are evaluated in terms of coefficients of determination (R²), root mean square error (RMSE) and residual prediction deviation (RPD). Calibration results, using partial least square models, obtained a R² in calibration greater than 0.95 for dry matter and crude protein, intermediate values (>0.75) for the fibre fraction and lower results for ash and crude fat (<0.75). The chemometric analysis shows lower results using principal component regression than partial least square models, although dry matter and acid detergent fibre obtained relatively high R² in calibration (0.876 and 0.863, respectively). Cross-validation achieved both lower R² and higher errors than calibration. Despite the wide variability of the data set, the results suggest that coupling FT-NIRS with partial least squares analysis allows us to estimate some chemical parameters of natural pastures, while the use of principal component regression models needs further evaluation.

This study evaluates the potential of Fourier transformation near-infrared reflectance spectroscopy to estimate the nutritional value and the chemical composition of natural pastures. Variability from all samples of pastures available is considered in order to assess the applicability of the calibration models in the future predictions. Chemical components (dry matter, crude protein, ash, ether extract, crude fibre, fibrous fractions) of grass samples were determined by applying official methods, and milk and meat forage units were calculated. Calibration and validation models were developed between chemical–nutritional parameters and NIRS spectral data using partial least square regression (PLS). The capacity of methods has been achieved using two validation approaches: the first using an independent dataset for prediction and the second by cross-validation process. The results are evaluated in term of coefficient of determination, root-mean-square error and residual prediction deviation. Despite the wide variability of the data set, the results of FT-NIRS have been able to estimate the chemical composition of natural and naturalised pasture with good accuracy and precision, while for nutritional value parameters, a further evaluation may be useful.

Near-infrared spectroscopy (NIRS) and closed spectroscopy methods have been applied to analyse the quality of forage and animal feed. However, grasslands are linked to variability factors (e.g., site, year, occurring species, etc.) which restrict the prediction capacity of the NIRS. The aim of this study is to test the Fourier transform NIRS application in order to determine the chemical characteristics of fresh, undried and unground samples of grassland located in north-central Apennine. The results indicated the success of FT-NIRS models for dry matter (DM), crude protein (CP), acid detergent fibre (ADF), neutral detergent fibre (NDF) and acid detergent lignin (ADL) on fresh grassland samples (R2 > 0.90, in validation). The model can be used to quantitatively determine CP and ADF (residual prediction deviation-RPD > 3 and range error ratio- RER > 10), followed by DM and NDF that maintain a RER > 10, and are sufficient for screening for the lignin fraction (RPD = 2.4 and RER = 8.8). On the contrary, models for both lipid and ash seem not to be usable at a practical level. The success of FT-NIRS quantification for the main chemical parameters is promising from the practical point of view considering both the absence of samples preparation and the importance of these parameters for diet formulation.