In vitro gas production systems are utilized to screen feed ingredients for inclusion in ruminant diets. However, not all in vitro systems are set up to measure methane (CH4) production, nor do all publications report in vitro CH4. Therefore, the objective of this study was to develop models to predict in vitro CH4 production from total gas and volatile fatty acid (VFA) production data and to identify the major drivers of CH4 production in these systems. Meta-analysis and machine learning (ML) methodologies were applied to a database of 354 data points from 11 studies to predict CH4 production from total gas production, apparent DM digestibility (DMD), final pH, feed type (forage or concentrate), and acetate, propionate, butyrate and valerate production. Model evaluation was performed on an internal dataset of 107 data points. Meta-analysis results indicate that equations containing DMD, total VFA production, propionate, feed type and valerate resulted in best predictability of CH4 on the internal evaluation dataset. The ML models far exceeded the predictability achieved using meta-analysis, but further evaluation on an external database would be required to assess generalization ability on unrelated data. Between the ML methodologies assessed, artificial neural networks and support vector regression resulted in very similar predictability, but differed in fitting, as assessed by behaviour analysis. The models developed can be utilized to estimate CH4 emissions in vitro.
Gas production profiles typically show a monotonically increasing monophasic pattern. However, atypical gas production profiles exist whereby at least two consecutive phases of gas production or additional extraneous features that distort the typical profile are present. Such profiles are more likely to occur with the use of a fecal inoculum and are much less well described. The presence of multiple phases or non-descript extraneous features makes it difficult to apply directly recommended modeling approaches such as standard response functions or classical growth functions. To overcome such difficulties, extensions of the Mitscherlich equation and a numerical modeling option also based on the Mitscherlich are explored. The numerical modeling option uses an estimate of relative rate obtained from the smoothed data profile and an estimate of maximum gas produced together with any lag time information drawn from the raw data to construct a simple Mitscherlich equation. In summary, this article illustrates the analysis of atypical gas production profiles obtained using a fecal inoculum and explores the methodology of numerical modeling to reconstruct equivalent typical growth-like trends.
Two models are proposed to describe atypical biphasic gas production profiles obtained from in vitro digestibility studies. The models are extensions of the standard Mitscherlich equation, comprising either two Mitscherlich terms or one Mitscherlich and one linear term. Two models that describe typical monophasic gas production curves, the standard Mitscherlich and the France model [a generalised Mitscherlich (root-t) equation], were assessed for comparison. Models were fitted to 25 gas production profiles resulting from incubating feedstuffs with faecal inocula from equines. Seventeen profiles displayed atypical biphasic patterns while the other eight displayed typical monophasic patterns. Models were evaluated using statistical measures of goodness-of-fit and by analysis of residuals. Good agreement was found between observed atypical profiles values and fitted values obtained with the two biphasic models, and both can revert to a simple Mitscherlich allowing them to describe typical monophasic profiles. The models contain kinetic fermentation parameters that can be used in conjunction with substrate degradability information and digesta passage rate to calculate extent of substrate degradation in the rumen or hindgut. Thus, models link the in vitro gas production technique to nutrient supply in the animal by providing information relating to digestion and nutritive value of feedstuffs.
Two new models, based upon the principles promulgated by Baranyi and co-workers are presented and resulting growth functions evaluated based upon their ability to mimic bacterial growth of the fish pathogen Flavobacterium psychrophilum. These growth functions make use of a dampening function to suppress potential growth, represented by a logistic, and are derived from rate:state differential equations. Dampening effects are represented by a rectangular hyperbola or a simple exponential, incorporated into a logistic differential equation and solved analytically resulting in two newly derived growth equations, viz. logistic × hyperbola (log × hyp) and logistic × exponential (log × exp). These characteristics result in flexible and robust growth functions that can be expressed as equations with biologically meaningful parameters. The newly derived functions (log × hyp and log × exp), along with the Baranyi (BAR), simple logistic (LOG) and its modified form (MLOG) were evaluated based upon examination of residuals and measures of goodness-of-fit and cross-validation. Using these criteria, log × hyp, log × exp and BAR performed better than, or at least equally well as, LOG and MLOG. In contrast with log × exp and BAR, log × hyp can be easily manipulated mathematically allowing for simple algebraic expressions for time and microbial biomass at inflexion point, in addition to maximum and scaled maximum growth rates.