Modelling interactions between farmer practices and fattening pig performances with an individual-based model

Eco-Friendly Feed Formulation and On-Farm Feed Production as Ways to Reduce the Environmental Impacts of Pig Production Without Consequences on Animal Performance

Animal feeding has a major contribution to the environmental impacts of pig production. One potential way to mitigate such effects is to incorporate an assessment of these impacts in the feed formulation process. The objective of this study was to test the ability of innovative formulation methodologies to reduce the impacts of pig production while also taking into account possible effects on growth performance. We compared three different formulation methodologies: least-cost formulation, in accordance with standard practices on commercial farms; multiobjective (MO) formulation, which considered feed cost and environmental impacts as calculated by life cycle assessment (LCA); and MO formulation, which prioritized locally produced feed ingredients to reduce the impact of transport. Ninety-six pigs were distributed between three experimental groups, with pigs individually weighted and fed using an automatic feeding system from 40 to 115 kg body weight. Based on the experimental results, six categories of impacts were evaluated: climate change (CC), demand in non-renewable energy (NRE), acidification (AC), eutrophication (EU), land occupation (LO), and phosphorus demand (PD), at both feed plant gate and farm gate, with 1 kg of feed and 1 kg of live pig as functional units, respectively. At feed level, MO formulations reduced CC, NRE, AC, and PD impacts but sometimes increased LO and EU impacts. These formulations reduced the proportion of cereals and oil meals into feeds (feed ingredients with high impacts), while the proportion of alternative protein sources, like peas, faba beans, or high-protein agricultural coproducts increased (feed ingredients with low impacts). Overall, animal performance was not affected by the dietary treatment; because of this, the general pattern of results obtained with either MO formulation at farm gate was similar to that obtained at feed level. Thus, MO diet formulation represents an efficient way to reduce the environmental impacts of pig production without compromising animal performance.

Reducing environmental impacts of feed using multiobjective formulation: What benefits at the farm gate for pig and broiler production?

Feed production is the main contributor to several environmental impacts of livestock. To decrease environmental impacts of feed, those of feedstuffs should be considered during formulation. In particular, multiobjective feed formulation (MOF) can help reduce several environmental impacts simultaneously while keeping any increase in feed price moderate. The objective of this study was to assess environmental benefits of MOF at the farm gate for fattening pigs and broilers. For pigs, three feeding strategies were tested: classic 2-phase (2P), 2-phase with lower net energy content (2P-), and multiphase (MP). For broilers, two strategies were tested: classic 3-phase (3P) and 3-phase with higher digestible amino acid contents and lower metabolisable energy content (3P+). Diets were formulated using both least-cost formulation (LCF) and MOF, yielding six pig scenarios and four broiler scenarios. Environmental impacts at the farm gate were estimated using a modelling approach based on life cycle assessment. Indicators for six impact categories were then calculated: climate change (CC), cumulative non-renewable energy demand (CEDNR), acidification (AC), eutrophication (EU), land occupation (LO), and phosphorus demand (PD). As expected, MOF had lower farm-gate impacts than LCF (as much as -13%), but the degree of decrease varied by feeding strategy and impact. For pigs, MOF was equally effective in all strategies at reducing PD (-6 to -9%) and AC (-2%). In contrast, MOF was more effective in 2P and 2P- at decreasing CC (-5% to -7%), LO (-9% to -13%) and EU (-6% to -8%) than in MP (CC: -2%; LO: -4%; EU: -3%). The benefit of MOF was found greater in 2P (-7%) than in other pig strategies for CEDNR (-3 to +0%). For broilers, MOF was equally effective in both strategies tested at decreasing PD (-12%), AC (-2%), and EU (-4%). For CC and CEDNR, MOF was more effective in 3P (CC: -9%; CEDNR: -11%) than 3P+ (-6% for both impacts), but not for LO (+3% in 3P vs -1% in 3P+). These differences were due mainly to differences in animal performance (especially feed conversion ratio) among the strategies tested. Finally, in all scenarios, gross margin at the farm gate decreased with MOF comparatively to LCF (pigs: -3% to -11%); broilers: -7% to -11%). These results demonstrate the importance of comprehensive economic and environmental optimisation of feeding strategies by simultaneously considering feed impacts, animal performance, and manure management. To do so, further research is therefore required to develop new modelling tools.

Comparison of data analytics strategies in computer vision systems to predict pig body composition traits from 3D images

Computer vision systems (CVS) have been shown to be a powerful tool for the measurement of live pig body weight (BW) with no animal stress. With advances in precision farming, it is now possible to evaluate the growth performance of individual pigs more accurately. However, important traits such as muscle and fat deposition can still be evaluated only via ultrasound, computed tomography, or dual-energy x-ray absorptiometry. Therefore, the objectives of this study were: 1) to develop a CVS for prediction of live BW, muscle depth (MD), and back fat (BF) from top view 3D images of finishing pigs and 2) to compare the predictive ability of different approaches, such as traditional multiple linear regression, partial least squares, and machine learning techniques, including elastic networks, artificial neural networks, and deep learning (DL). A dataset containing over 12,000 images from 557 finishing pigs (average BW of 120 ± 12 kg) was split into training and testing sets using a 5-fold cross-validation (CV) technique so that 80% and 20% of the dataset were used for training and testing in each fold. Several image features, such as volume, area, length, widths, heights, polar image descriptors, and polar Fourier transforms, were extracted from the images and used as predictor variables in the different approaches evaluated. In addition, DL image encoders that take raw 3D images as input were also tested. This latter method achieved the best overall performance, with the lowest mean absolute scaled error (MASE) and root mean square error for all traits, and the highest predictive squared correlation (R2). The median predicted MASE achieved by this method was 2.69, 5.02, and 13.56, and R2 of 0.86, 0.50, and 0.45, for BW, MD, and BF, respectively. In conclusion, it was demonstrated that it is possible to successfully predict BW, MD, and BF via CVS on a fully automated setting using 3D images collected in farm conditions. Moreover, DL algorithms simplified and optimized the data analytics workflow, with raw 3D images used as direct inputs, without requiring prior image processing.

Gut microbiome composition differences among breeds impact feed efficiency in swine

BACKGROUND

Feed efficiency is a crucial parameter in swine production, given both its economic and environmental impact. The gut microbiota plays an essential role in nutrient digestibility and is, therefore, likely to affect feed efficiency. This study aimed to characterize feed efficiency, fatness traits, and gut microbiome composition in three major breeds of domesticated swine and investigate a possible link between feed efficiency and gut microbiota composition.

RESULTS

Average daily feed intake (ADFI), average daily gain (ADG), feed conversion ratio (FCR), residual feed intake (RFI), backfat, loin depth, and intramuscular fat of 615 pigs belonging to the Duroc (DR), Landrace (LR), and Large White (LW) breeds were measured. Gut microbiota composition was characterized by 16S rRNA gene sequencing. Orthogonal contrasts between paternal line (DR) and maternal lines (LR+LW) and between the two maternal lines (LR versus LW) were performed. Average daily feed intake and ADG were statistically different with DR having lower ADFI and ADG compared to LR and LW. Landrace and LW had a similar ADG and RFI, with higher ADFI and FCR for LW. Alpha diversity was higher in the fecal microbial communities of LR pigs than in those of DR and LW pigs for all time points considered. Duroc communities had significantly higher proportional representation of the Catenibacterium and Clostridium genera compared to LR and LW, while LR pigs had significantly higher proportions of Bacteroides than LW for all time points considered. Amplicon sequence variants from multiple genera (including Anaerovibrio, Bacteroides, Blautia, Clostridium, Dorea, Eubacterium, Faecalibacterium, Lactobacillus, Oscillibacter, and Ruminococcus) were found to be significantly associated with feed efficiency, regardless of the time point considered.

CONCLUSIONS

In this study, we characterized differences in the composition of the fecal microbiota of three commercially relevant breeds of swine, both over time and between breeds. Correlations between different microbiome compositions and feed efficiency were established. This suggests that the microbial community may contribute to shaping host productive parameters. Moreover, our study provides important insights into how the intestinal microbial community might influence host energy harvesting capacity. A deeper understanding of this process may allow us to modulate the gut microbiome in order to raise more efficient animals. Video Abstract.

Effects of interactions between feeding practices, animal health and farm infrastructure on technical, economic and environmental performances of a pig-fattening unit

European pig production faces economic and environmental challenges. Modelling can help farmers simulate and understand how changes in their management practices affect the efficiency of their production system. We developed an individual-based model of a pig-fattening unit that considers individual variability in performance among pigs, farmers' feeding practices and animal management and estimates environmental impacts (using life cycle assessment) and economic results of the unit. We previously demonstrated that this model provides reliable estimates of farm performance for different combinations of management practices, pig types and building characteristics. The objectives of this study were to quantify how interactions between feeding practices and animal management influence fattening unit results in healthy or impaired health conditions using the model. A virtual experiment was designed to evaluate effects of interactions between feeding practices, health status of the pig herd and infrastructure constraints on the technical performance, economic results and environmental impacts of the unit. The virtual experiment consisted of 96 scenarios, which combined chosen values of 6 input parameters of the model: batch interval (35 days and 7 days), use or non-use of a buffer room to manage the lightest pigs, feed rationing (ad libitum and restricted) and sequence plans (two-phase (2P), daily-phase (DP)), scale at which the feeding plan is applied (i.e. room, pen and individual) and health status of the pig herd (i.e. healthy v. impaired). Variance analysis was used to test effects of the factors in these 96 scenarios, and multivariate data analyses were used to classify the scenarios. Healthy populations obtained on average higher economic results (e.g. gross margin of 11.20 v. 1.50 €/pig) and lower environmental impacts (e.g. 2.24 v. 2.38 kg CO2-eq/kg pig live weight gain) than the population with impaired health. With 35 days batch interval and DP feeding, populations with impaired health reached gross margin similar to healthy populations with 2P ad libitum feeding and 7 days batch interval. Restricted, DP and individual feeding plans improved the economic and environmental performances of the unit for both health statuses. This study highlighted that health status of the pig herd is the main factor that affects technical, economic and environmental performances of a pig-fattening unit, and that adequate feeding strategies and animal management can compensate, to some extent, the effects of impaired health on environmental impacts but not on gross margin.

A novel automated system to acquire biometric and morphological measurements and predict body weight of pigs via 3D computer vision

Computer vision applications in livestock are appealing since they enable measurement of traits of interest without the need to directly interact with the animals. This allows the possibility of multiple measurements of traits of interest with minimal animal stress. In the current study, an automated computer vision system was devised and evaluated for extraction of features of interest, as body measurements and shape descriptors, and prediction of body weight in pigs. From the 655 pigs that had data collected 580 had more than 5 frames recorded and were used for development of the predictive models. The cross-validation for the models developed with data from nursery and finishing pigs presented an R2 ranging from 0.86 (random selected image) to 0.94 (median of images truncated on the third quartile), whereas with the dataset without nursery pigs, the R2 estimates ranged from 0.70 (random selected image) to 0.84 (median of images truncated on the third quartile). However, overall the mean absolute error was lower for the models fitted without data on nursery animals. From the body measures extracted from the image, body volume, area, and length were the most informative for prediction of body weight. The inclusion of the remaining body measurements (width and heights) or shape descriptors to the model promoted significant improvement of the predictions, whereas the further inclusion of sex and line effects were not significant.