Quantitative joint evaluation of sheep enteric methane emissions and faecal dry matter and nitrogen excretion

Abstract

Specific nutrition and management interventions at farm level that aim at reducing enteric methane (CH4) production from sheep may not reduce excretion of dry matter and N, the main source of excreta CH4 production and nitrous oxide. The objectives were a) to determine the relationship of dietary variables and animal characteristics to both enteric CH4 emissions and N excretion from sheep fed ryegrass or forage brassica diets, and b) to jointly predict enteric CH4 emissions, and dry matter and N excretion from sheep based on feed intake, dietary nutrient contents and animal characteristics. A database containing 270 indirect respiration calorimetry and total excreta collection records of eight sheep studies from New Zealand was used. To explain variation in CH4 production and N excretion, all records were assigned to four categories based on CH4 emissions and N excretion being lesser or greater than the means of the entire database, viz. high CH4-high N, high CH4-low N, low CH4-high N and low CH4-low N. A redundancy analysis indicated the two low N categories were positively related to dry matter digestibility. Subsequently, dry matter digestibility was positively related to brassica forages and negatively related to ryegrass forages. Univariate modelling indicated that CH4 (g d−1) was best predicted by dry matter intake (DMI), faecal dry matter excretion (g d−1) by DMI and dietary neutral detergent fibre content, and N excretion (g d−1) by DMI and dietary crude protein content, with dry matter digestibility also contributing to the best prediction of faecal and total N. A bivariate multiple regression model for jointly predicting CH4 emission and total N excretion, and a trivariate model for also predicting faecal dry matter were developed. Key covariates included DMI, dietary neutral detergent fibre and crude protein contents, whereas DMI was replaced by body weight in the bivariate yield model (g [kg DMI]-1), and not selected for the trivariate yield model. Regression coefficients of dietary crude protein content obtained through multivariate modelling had opposite signs for CH4 emission and N excretion, indicating a trade-off. However, this trade-off was not supported by univariate modelling. Therefore, evidence for a trade-off between enteric CH4 and N excretion from sheep is limited.