Application of Meta-Analysis and Machine Learning Methods to the Prediction of Methane Production from In Vitro Mixed Ruminal Micro-Organism Fermentation.

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.

Prediction of enteric methane production, yield, and intensity in dairy cattle using an intercontinental database.

Enteric methane (CH4 ) production from cattle contributes to global greenhouse gas emissions. Measurement of enteric CH4 is complex, expensive, and impractical at large scales; therefore, models are commonly used to predict CH4 production. However, building robust prediction models requires extensive data from animals under different management systems worldwide. The objectives of this study were to (1) collate a global database of enteric CH4 production from individual lactating dairy cattle; (2) determine the availability of key variables for predicting enteric CH4 production (g/day per cow), yield [g/kg dry matter intake (DMI)], and intensity (g/kg energy corrected milk) and their respective relationships; (3) develop intercontinental and regional models and cross-validate their performance; and (4) assess the trade-off between availability of on-farm inputs and CH4 prediction accuracy. The intercontinental database covered Europe (EU), the United States (US), and Australia (AU). A sequential approach was taken by incrementally adding key variables to develop models with increasing complexity. Methane emissions were predicted by fitting linear mixed models. Within model categories, an intercontinental model with the most available independent variables performed best with root mean square prediction error (RMSPE) as a percentage of mean observed value of 16.6%, 14.7%, and 19.8% for intercontinental, EU, and United States regions, respectively. Less complex models requiring only DMI had predictive ability comparable to complex models. Enteric CH4 production, yield, and intensity prediction models developed on an intercontinental basis had similar performance across regions, however, intercepts and slopes were different with implications for prediction. Revised CH4 emission conversion factors for specific regions are required to improve CH4 production estimates in national inventories. In conclusion, information on DMI is required for good prediction, and other factors such as dietary neutral detergent fiber (NDF) concentration, improve the prediction. For enteric CH4 yield and intensity prediction, information on milk yield and composition is required for better estimation.

Development and validation of an UHPLC-MS/MS method for the determination of mycotoxins in grass silages.

An ultra-high-performance liquid chromatography-tandem mass spectrometry (UHPLC-MS/MS) multi-mycotoxin analytical method was developed to simultaneously identify and quantify 20 mycotoxins in grass silages, inclusive of mycotoxins that are currently regulated in European Union feeds. Extraction of mycotoxins from dried grass silages was performed using of a modified QuEChERS extraction employing an acidified aqueous extraction (0.1 N HCl) with no further clean-up. Following chromatographic separation, analytes were detected using a fast polarity-switching MS/MS method that allowed both positive and negative ions to be analysed from a single injection, thus the reducing time and cost of analysis. The limits of detection and quantification ranged between 3 µg kg(-1) DM (aflatoxin B1, beauvericin and enniatin A and A1) and 200 µg kg(-1) DM (deoxynivalenol), and between 10 µg kg(-1) DM (aflatoxin B1, beauvericin and enniatin A1) and 500 µg kg(-1) DM (deoxynivalenol), respectively. Inter-assay accuracy and precision ranged between 90% and 107% and between 3.9% and 15.0% CV, respectively. The accuracy of the method was assessed through the application to a range of incurred samples in an inter-laboratory study.

Ensiling of seaweed for a seaweed biofuel industry.

Effective biogas production from seaweed necessitates harvest at times of peak quality of biomass and low-loss preservation for year-around supply. Ensiling of five seaweed species and storage up to 90days was investigated as a method to preserve the methane yield potential. Adequate acidification by natural lactic acid fermentation was difficult due to low rapidly fermentable carbohydrate contents, high buffering capacities and low initial numbers of lactic acid bacteria. Nevertheless, products of silage fermentation increased methane yields by up to 28% and compensated for volatile solid losses during ensiling. Preservation of the original methane yield potential was achieved for four of five seaweed species, provided that silage effluent is collected and utilised. 10-28% of the ensiled biomass was released as effluent with methane yields of 218-423LNkg(-1) VS. If further optimised, ensiling represents an effective method of preservation crucial for an efficient seaweed biofuel industry.

Use of a mannitol rich ensiled grass press juice (EGPJ) as a sole carbon source for polyhydroxyalkanoates (PHAs) production through high cell density cultivation.

This study demonstrates the use of a mannitol rich ensiled grass press juice (EGPJ) as a renewable carbon substrate for polyhydroxyalkanoates (PHA) production in shaking flask experiments and fed-batch stirred tank reactor cultivations. Fed-batch cultivations of Burkholderia sacchari IPT101 using EGPJ as sole carbon source produced 44.5 g/L CDW containing 33% polyhydroxybutyrate (PHB) in 36 h, while Pseudomonas chlororaphis IMD555 produced a CDW of 37 g/L containing 10% of medium chain length polyhydroxyalkanoates (mcl-PHA) in 34 h. PHB and mcl-PHA extracted from B. sacchari IPT101 and P. chlororaphis IMD555, grown on EGPJ, had a molecular weight of 548 kg/mol and 115.4 kg/mol, respectively. While mcl-PHA can be produced from EGPJ, PHB production is more interesting as there is a 4-fold higher volumetric productivity compared to mcl-PHA.

Cluster analysis of fasciolosis in dairy cow herds in Munster province of Ireland and detection of major climatic and environmental predictors of the exposure risk.

Fasciolosis caused by Fasciola hepatica is a widespread parasitic disease in cattle farms. The aim of this study was to detect clusters of fasciolosis in dairy cow herds in Munster Province, Ireland and to identify significant climatic and environmental predictors of the exposure risk. In total, 1,292 dairy herds across Munster was sampled in September 2012 providing a single bulk tank milk (BTM) sample. The analysis of samples by an in-house antibody-detection enzyme-linked immunosorbent assay (ELISA), showed that 65% of the dairy herds (n = 842) had been exposed to F. hepatica. Using the Getis-Ord Gi* statistic, 16 high-risk and 24 low-risk (P <0.01) clusters of fasciolosis were identified. The spatial distribution of high-risk clusters was more dispersed and mainly located in the northern and western regions of Munster compared to the low-risk clusters that were mostly concentrated in the southern and eastern regions. The most significant classes of variables that could reflect the difference between high-risk and low-risk clusters were the total number of wet-days and rain-days, rainfall, the normalized difference vegetation index (NDVI), temperature and soil type. There was a bigger proportion of well-drained soils among the low-risk clusters, whereas poorly drained soils were more common among the high-risk clusters. These results stress the role of precipitation, grazing, temperature and drainage on the life cycle of F. hepatica in the temperate Irish climate. The findings of this study highlight the importance of cluster analysis for identifying significant differences in climatic and environmental variables between high-risk and low-risk clusters of fasciolosis in Irish dairy herds.

Spatial analysis and risk mapping of Fasciola hepatica infection in dairy herds in Ireland.

Fasciolosis is generally a subclinical infection of dairy cows and can cause marked economic losses. This study investigated the prevalence and spatial distribution of fasciolosis in dairy cow herds in Ireland using an in-house antibodydetection enzyme-linked immunosorbent assay applied to bulk tank milk (BTM) samples collected during the autumn of 2012. A total of 5,116 BTM samples were collected from 4,602 different herds, with 514 farmers submitting BTM samples in two consecutive months. Analysis of the BTM samples showed that 82% (n = 3,764) of the dairy herds had been exposed to Fasciola hepatica. A total of 108 variables, including averaged climatic data for the period 1981-2010 and contemporary meteorological data for the year 2012, such as soil, subsoil, land cover and habitat maps, were investigated for a possible role as predictor of fasciolosis. Using mainly climatic variables as the major predictors, a model of the predicted risk of fasciolosis was created by Random Forest modelling that had 95% sensitivity and 100% specificity. The most important predictors in descending order of importance were: average of annual total number of rain-days for the period 1981-2010, total rainfall during September, winter and autumn of 2012, average of annual total number of wet-days for the period 1981- 2010 and annual mean temperature of 2012. The findings of this study confirm the high prevalence of fasciolosis in Irish dairy herds and suggest that specific weather and environmental risk factors support a robust and precise distribution model.

The effects of farm management practices on liver fluke prevalence and the current internal parasite control measures employed on Irish dairy farms.

Fasciolosis caused by Fasciola hepatica is responsible for major production losses in cattle farms. The objectives of this study were to assess the effect of farm management practices on liver fluke prevalence on Irish dairy farms and to document the current control measures against parasitic diseases. In total, 369 dairy farms throughout Ireland were sampled from October to December 2013, each providing a single bulk tank milk (BTM) sample for liver fluke antibody-detection ELISA testing and completing a questionnaire on their farm management. The analysis of samples showed that cows on 78% (n=288) of dairy farms had been exposed to liver fluke. There was a difference (P<0.05) between farms where cows were positive or negative for liver fluke antibodies in (a) the total number of adult dairy cows in herds, (b) the number of adult dairy cows contributing to BTM samples, and (c) the size of the total area of grassland, with positive farms having larger numbers in each case. There was no difference (P>0.05) between positive and negative farms in (a) the grazing of dry cows together with replacement cows, (b) whether or not grazed grassland was mowed for conservation, (c) the type of drinking water provision system, (d) spreading of cattle manure on grassland or (e) for grazing season length (GSL; mean=262.5 days). Also, there were differences (P<0.001) between drainage statuses for GSL with farms on good drainage having longer GSL than moderately drained farms. The GSL for dairy cows on farms with good drainage was 11 days longer than for those with moderate drainage (P<0.001). The percentage of farmers that used an active ingredient during the non-lactating period against liver fluke, gastrointestinal nematodes, lungworm, and rumen fluke was 96%, 85%, 77% and 90%, respectively. Albendazole was the most frequently used active ingredient for treatment against gastrointestinal nematodes (57%), liver fluke (40%) and lungworm (47%), respectively. There was a difference (P<0.05) in the use of triclabendazole and albendazole between positive and negative farms, with triclabendazole use being more common in positive farms. This study highlighted differences in dairy management practices between Irish farms with dairy herds exposed or not exposed to liver fluke and stressed the need of fine-scale mapping of the disease patterns even at farm level to increase the accuracy of risk models. Also, comprehensive advice and professional support services to farmers on appropriate farm management practices are very important for an effective anthelmintic control strategy.

Optimisation of digester performance with increasing organic loading rate for mono- and co-digestion of grass silage and dairy slurry.

This study investigated the feasibility of mono-digesting grass silage, dairy slurry and the co-digestion of the two substrates at a range of concentrations with a specific focus on digester performance while increasing organic loading rate (OLR). The results show that the higher the proportion of grass silage in the substrate mix the higher the specific methane yield (SMY) achieved. Optimum conditions were assessed for 100% grass silage at an OLR of 3.5 kg VS m(-3) d(-1) generating a SMY of 398 L CH4 kg(-1) VS equating to a biomethane efficiency of 1.0. For co-digestion of grass silage with 20% dairy slurry the optimum condition was noted at an OLR of 4.0 kg VS m(-3) d(-1) generating a SMY of 349L CH4 kg(-1) VS and a biomethane efficiency of 1.01. Hydraulic retention times of less than 20 days proved to be a limiting factor in the operation of farm digesters.

The effect of trace element addition to mono-digestion of grass silage at high organic loading rates.

This study investigated the effect of trace element addition to mono-digestion of grass silage at high organic loading rates. Two continuous reactors were compared. The first mono-digested grass silage whilst the second operated in co-digestion, 80% grass silage with 20% dairy slurry (VS basis). The reactors were run for 65weeks with a further 5weeks taken for trace element supplementation for the mono-digestion of grass silage. The co-digestion reactor reported a higher biomethane efficiency (1.01) than mono-digestion (0.90) at an OLR of 4.0kgVSm(-3)d(-1) prior to addition of trace elements. Addition of cobalt, iron and nickel, led to an increase in the SMY in mono-digestion of grass silage by 12% to 404LCH4kg(-1)VS and attained a biomethane efficiency of 1.01.

The potential for biomethane from grass and slurry to satisfy renewable energy targets.

A biomethane potential (BMP) assessment of grass silage yielded 107 m(3)CH4 t(-1). Long term mono-digestion of grass silage can suffer due to a deficiency in essential nutrients; this may be overcome by co-digesting with slurry. Mono-digestion of slurry achieved a low yield of 16 m(3)CH4 t(-1). BMP assessments at a range of co-digestion ratios indicated methane yields were between 4% and 11% lower than the values calculated from mono-digestion. This paper suggests that co-digestion of the majority of slurry produced from dairy cows in Ireland with grass silage quantities equivalent to 1.1% of grassland on a 50:50 volatile solids basis would generate over 10% renewable energy supply in transport (RES-T). The industry proposed would equate to 170 digesters each treating 10,000 t a(-1) of grass silage and 40,000 t a(-1) of slurry from dairy cows.

Anaerobic methane production from five common grassland species at sequential stages of maturity.

Since grass will likely be a dominant feedstock for on-farm anaerobic digestion in Northwest Europe, changes in the chemical composition of five common grass species with advancing harvest date in the primary growth were investigated and related to specific CH(4) yields. The increase in fibre components with advancing harvest date had a negative impact on the specific CH(4) yield (253 and 225 Nl CH(4) kg(-1) VS for 12 May and 7 July harvests, respectively), and this impact was similar across the five grass species. At common growth stages, only small differences in herbage digestibility was observed between the grass species and this was reflected in similar specific CH(4) yields; however, the 26% lower area-specific CH(4) yield of the cocksfoot variety (Dactylis glomerata L. var. Pizza) would make it the most expensive of the five grass species to produce and the least suitable for anaerobic digestion.

Morphological and molecular characterisation of Penicillium roqueforti and P. paneum isolated from baled grass silage.

The morphological and molecular features of Penicillium roqueforti and P. paneum isolated from baled grass silage were characterised. A total of 315 isolates were investigated, comprising 237 P. roqueforti and 78 P. paneum isolates randomly selected from more than 900 Penicillium colonies cultured from bales. The macromorphological features of both species broadly agreed with the literature, but the micromorphological features differed in some respects. When observed using SEM, P. roqueforti and P. paneum had finely roughened conidia, and conidiophores, phialides and conidia of P. paneum were each larger than those of P. roqueforti. Based on the phylogenetic analysis of partial sequences of beta-tubulin and acetyl co-enzyme A (CoA) synthetase genes, P. roqueforti and P. paneum isolates were found to be monophyletic species.

Mycotoxins and other secondary metabolites produced in vitro by Penicillium paneum Frisvad and Penicillium roqueforti Thom isolated from baled grass silage in Ireland.

Secondary metabolites produced by Penicillium paneum and Penicillium roqueforti from baled grass silage were analyzed. A total of 157 isolates were investigated, comprising 78 P. paneum and 79 P. roqueforti isolates randomly selected from more than 900 colonies cultured from bales. The findings mostly agreed with the literature, although some metabolites were not consistently produced by either fungus. Roquefortine C, marcfortine A, and andrastin A were consistently produced, whereas PR toxin and patulin were not. Five silage samples were screened for fungal metabolites, with two visually moldy samples containing up to 20 mg/kg of roquefortine C, mycophenolic acid, and andrastin A along with minor quantities (0.1-5 mg/kg) of roquefortines A, B, and D, festuclavine, marcfortine A, and agroclavine. Three visually nonmoldy samples contained low amounts of mycophenolic acid and andrastin A. The ability of both molds to produce a diverse range of secondary metabolites in vitro and in silage should be a concern to livestock producers.

Monitoring Lactobacillus plantarum in grass silages with the aid of 16S rDNA-based quantitative real-time PCR assays.

Ensiling plant material with the aid of lactic acid bacteria (LAB) is a common agricultural practice for conserving forages independently of the time point of harvest. Despite ensiling being a natural process, it can be improved by the treatment of the harvested forage with starter cultures before storage. Within this context, Lactobacillus plantarum (L. plantarum) is the most frequently used LAB in commercially available starter cultures. In order to enable the monitoring of the population dynamics of L. plantarum in silage, methods for species-specific detection based on the 16S ribosomal DNA (rDNA) sequence were developed by applying a quantitative real-time polymerase chain reaction (QRT-PCR) approach. The QRT-PCR assay was also applied to estimate the development of the L. plantarum population within experimental grass silages. In addition, a multiplex QRT-PCR assay was developed to estimate the amount of L. plantarum 16S rDNA in relation to total bacterial 16S rDNA. This multiplex QRT-PCR assay was applied to monitor the influence of different silage additives on the L. plantarum population.

Alteration of the carbon and nitrogen stable isotope composition of beef by substitution of grass silage with maize silage.

This study investigated the effect of substituting grass silage (C3 photosynthetic plant product) with maize silage (C4 photosynthetic plant product) on the natural abundance carbon (delta13C) and nitrogen (delta15N) stable isotope composition of bovine muscle tissue. Forty-five continental crossbred heifers were assigned to one of three diets consisting of 3 kg of a barley-based concentrate plus grass silage, maize silage or an equal mixture (dry matter basis) of grass silage and maize silage, fed ad libitum, for 167 days. Substitution resulted in less negative delta13C values (P<0.001) in lipid-free muscle and in lipid, and also a lower delta15N (P<0.001) in lipid-free muscle. Feeding of maize silage was clearly reflected in the delta13C of muscle, with each 10% difference in the dietary C4 carbon intake resulting in a 0.9 to 1.0 per thousand shift of delta13C in lipid-free muscle and a 1.0 to 1.2 per thousand in lipid. Minimum detectable mean differences (95% confidence, power 0.80, n=15) in this experiment were about 0.5 per thousand and 1.0 per thousand for delta13C of lipid-free muscle and lipid, respectively, and about 0.5 per thousand for delta15N of lipid-free muscle. The power analysis presented here is useful for estimating minimum isotopic differences that can be detected between any two groups of beef samples with a given number of replicates. It is concluded that carbon stable isotope ratio analysis of meat can be used to quantify C3/C4 dietary constituents in beef production.

Fungi isolated from contaminated baled grass silage on farms in the Irish Midlands.

The incidence of fungal growth on baled grass silage was recorded on 35 farms in the Irish Midlands in 2003. Fungal colonies were visible on 58 of 64 bales examined and the number of colonies per bale ranged from 1 to 12. On average, 5% of bale surface areas were affected. The fungus most prevalent on bales was Penicillium roqueforti, present on 86% of bales and representing 52% of all isolates. Other moulds isolated were Penicillium paneum, Geotrichum, Fusarium and mucoraceous species. Schizophyllum commune was observed protruding through the plastic film on bales on 17 of the 35 farms.