Why is metabolic energy budgeting not more widely used?

This editorial sets out my personal professional journey with metabolic energy budgeting (MEB) and concludes that the methodology deserves to be more widely used. I compiled my first metabolic energy budget for calculation of feed requirements of ruminants grazing pasture around 30 years ago when a colleague left our Department at Massey University, New Zealand, and I took over his teaching. The student assignment for my colleague's section of the course presented details of cow numbers on a dairy farm in each month of the year, their body weights and milk production on the one hand and the pasture growth and available supplementary feed sources and quantities on the other hand. Class members were required to compile, in a Microsoft Excel spreadsheet, a feed budget for the farm showing the animal feed requirements calculated using metabolic energy equations and the combinations of different feeds used to meet those requirements in different months of the year, all in common units of kg DM ha−1 day−1. To decide if I would adopt my colleague's assignment in my own teaching of his course, or develop an alternative of my own, I set aside an evening after dinner and completed the assignment myself. This was a revelation. I found that through completing the assignment, I understood the role of the various feeds at different times of the year and their integration within the farm system in a way I never had before. The feed budget demonstrated the value of higher energy feeds such as fresh pasture or brewer's grain (assuming that the diet was balanced for nutrient components like protein) in lactation, whereas lower energy feeds such as grass silage could be used for body weight maintenance in winter. To balance the budget across the months, it was necessary to include in the calculation a representation of how autumn- and early-winter-grown pasture is stockpiled as increased herbage mass (pasture cover), and then released back to animals in the first 6 weeks of lactation, when cow feed requirements exceed herbage accumulation rates in most New Zealand dairy farm systems. Based on how much I had learned by completing the calculations, I incorporated the assignment into my own teaching and over the next 20 years, some 150–200 members of my agriculture and veterinary classes each year prepared whole-year feed demand and supply calculations with a monthly time step for farms that we visited. I soon found that preparing a MEB for a farm provided a foundation for engaging conversations with the farmers because the calculation made me aware of nuances and seasonal feed supply pressure points that the farmers understood through practical experience over multiple years on their farm, but a casual visitor might be unaware of. While metabolic energy calculations in North America and Europe more often use units of total Joules or Calories for a specified time period, we carried out our calculations on a per unit area basis. This expresses feed requirements in the same units as herbage accumulation for easy comparison and intuitive appreciation of seasonal feed surplus and deficit in a farm system and removes the effect of farm size on total feed requirements, thus providing for easier comparison between farms. Metabolic energy calculations readily allow quantification of the flow of energy from capture of sunlight by vegetation, to intake of herbage by animals, to product extracted from a farm system for human use (Figure 1). By contrast with the 115 GJ ME ha−1 year−1 feed grown in a fertilized pasture in New Zealand's warm temperate climate, the herbage harvested annually in an experimental lamb-rearing system in Huanxian County, Gansu China (mean annual precipitation 359 mm, mean annual evaporation 1993 mm, frost-free period 125 days, low soil fertility status), was 2–5 GJ ME ha−1 year−1 (Chen et al., 2010). Such visualization of energy flow could readily be used in environmentally sensitive areas like the Qinghai–Tibet Plateau to explore and determine ecologically or environmentally sustainable carrying capacities for guidance of both farmers and government administrators involved in regional planning. The MEB can also represent the gain or loss of animal body weight as a way to temporarily store surplus energy or meet animal feed requirements when herbage accumulation rates are insufficient to do so. Initially, I imagined that metabolic energy calculations would be very approximate, but experience soon showed that when a metabolic energy representation of an animal grazing system is prepared with care and attention to detail, calculated herbage intake by animals in the system will be within 2%–5% of actual intake. Once aware of the power and versatility of metabolic energy budgeting to quantify farm system components and enhance understanding of the system, I found myself using the methodology regularly in my own professional work and publications. For example, in the experiments reported by Matthew et al. (2017), herbage mass of farm paddocks was measured weekly, so the herbage accumulation rate could be determined from accumulation on ungrazed paddocks within the rotation. Meanwhile, the sheep in the experiment were weighed periodically, so output from metabolic energy calculations of herbage intake based on sheep body weight and body weight change could be checked against field data. When 6 years' historical data for rotation lengths of 16, 48, and 72 days were analyzed by MEB, the average across years for herbage consumption by sheep in the experiments during the period from early May to late September was quantified as 4640, 5090, and 5563 kg DM ha−1, respectively, with a standard error of means of 143 kg ha−1. By contrast, a recent meta-analysis of the accuracy of the n-alkane fecal marker method for measuring animal intake (Andriarimalala et al., 2020) reported high R2 values, but coefficients of variation of 8%–16%. It is salutary to note that a measurement uncertainty of 10%–15% for herbage intake of animals would exceed the size of the treatment effect in many feed-comparison experiments and render the treatment effect in the experiment statistically undetectable. The lower errors in metabolic energy determination of herbage intake by animals arise because a key calculation parameter, animal body weight, can be measured with precision, and because pastoral systems are mostly “energy-limited.” However, in certain situations, such as during prolonged summer drought, metabolizable protein may become “first-limiting” and metabolic energy calculations may be less accurate (R. Vibart, Personal Communication). Similarly, in stall-feeding systems based on concentrate feeds, more common in the United States and Europe, more detailed nutrient balancing calculations are required. For MEB calculations at the farm system level, various highly evolved commercial packages are available in Australasia. One example favored in New Zealand industry extension circles for farm system optimization is FARMAX (Bryant et al., 2010), which was originally launched as STOCKPOL for sheep and beef farms in the 1980s (Marshall et al., 1991) and is now available in both sheep and beef, and dairy versions. A second package, often used by farm-systems researchers, is APSIM (Holzworth et al., 2014), while Bryant and Snow (2008) review others. The fertilizer nutrient budgeting package OVERSEER also performs animal MEB calculations in the background as part of the prediction of leaching losses of N and P from animal dung and urine returns. Previously, another prominent package was UDDER for dairy farm systems (Uribe et al., 1996). Further, it is not necessary to use commercial packages to obtain publication-quality output from MEB calculations. Matthew et al. (2010) used self-built Excel spreadsheets to demonstrate differences in feed conversion efficiency (milk energy as a % of metabolizable energy in feed consumed) for alternative farm system configurations involving differing grazing management strategies; this study also noted that values for feed consumed in the modeled farm systems from the self-built spreadsheets generally differed from those of OVERSEER by less than 5%. Similarly, Duranton and Matthew (2018) modeled hill country sheep–beef farm systems with or without a proportion of the farm planted in plantain, and reported in their Table S3 that values for annual herbage consumption generated by self-built Excel spreadsheets differed from those generated by FARMAX by less than 4%. A further example of the application of self-built Excel spreadsheets is the PhD research of Gobilik (2017), who used farmer diaries and other records to explore the evolution of New Zealand sheep–beef farm systems over the period 1980–2010 (Gobilik et al., 2021) and to define the key characteristics of tropical beef production systems in Sabah, Malaysia. Malaysian beef production systems studied included grazed pasture, cut and carry feeding of forage to animals indoors, and grazing within oil-palm plantations (Gobilik et al., 2023). In conclusion, modern commercial MEB packages are highly evolved and sophisticated, and have a substantial body of data and experience underpinning them. Self-built Excel spreadsheets are also accurate and effective. Given this, it is hard to understand why the MEB methodology has not been more widely adopted, where relevant, across disciplines within the agricultural and ecological research arena, as statistical techniques like analysis of variance or technologies like remote sensing have been.