Rethinking production systems: science for the land-based sector

Abstract

By 2030, New Zealand agribusiness will have accepted that there are biological limits to meeting burgeoning demands for safe, affordable food production, while mitigating adverse environmental impacts. The nature of scientific research will have changed significantly from agronomic field trial observation to one of understanding aimed at fundamentally rethinking and improving resource use efficiency. Top scientists, focusing innovatively on the processes involved, will have moved from historic to future criteria for optimisation. They will also have used process-based models to highlight the few sustainable ways to alter plant, animal and soil function safely. A priori proof of concept will be a matter of course and there will have been major progress in identifying genes controlling the fine detail of the balance of growth and resource use. This future depends on applying the same criteria for evaluating proficiency, productivity and minimising unproductive costs and waste, throughout the science system itself.

Keywords:

• emissions
• food
• funding
• intensification
• process-based modelling
• sustainability

Introduction

Feeding an increasing number of people sustainably is emerging as one of the greatest challenges facing mankind. Estimates vary as to how much more food will be required by 2050, from double the current quantities (e.g. Tilman et al. 2011) to 60% of current quantities if we improve the current food production system (e.g. ActionAidUSA 2013). Suggestions of how this might occur focus on stabilising land in production, closing the yield gap, eating less meat, using water and nutrients more strategically and reducing waste (e.g. Beddington 2010; Godfray et al. 2010; Foley et al. 2011; Ehrlich & Ehrlich 2013). One issue of importance is that estimates of how much production must increase have varied considerably over the years and now tend to be lower than in the past as the reality of intensification is included.

Intensification has increased productivity: in the 22 years between the periods 1990–1992 and 2012–2014, the proportion of the global population undernourished has decreased from 18.7% to 11.3% globally, and from 23.4% to 13.5% of the population in developing countries (FAO 2014, p. 40). However, the World Food Summit target of ‘halving the number of undernourished people in developing countries by 2015’, cannot be met (FAO 2014, p. 9).

To achieve this target and more requires concerted research effort (FAO 2009; Warr 2014), but rates of investment in agricultural research are slowing in many countries in spite of continuing high social rates of return to the investments (Alston & Pardey 2014), currently estimated at 9.8% per year (Hurley et al. 2014).

The goals

Although predicted food needs can be met, at least in theory, doing so while limiting adverse environmental impact—through, for instance, nitrogen release and greenhouse gas (GHG) emissions—is a huge challenge. The current estimate from the IPCC (2014) is a reduction in GHG by 40%–70% by 2050 in comparison with 2010 to prevent catastrophic warming.

The problem is that the agricultural system has complex connections with the major drivers of environmental degradation (Ehrlich & Ehrlich 2013). It is generally accepted that loss of soil services (such as soil carbon and nutrients) and the release of the substantial quantities of carbon sequestered beneath grassland to the atmosphere must be avoided. However, insights presented by process-based modelling suggest that increasing food production, maintaining soil carbon and reducing nitrogen leaching are incompatible goals (Parsons et al. 2013). From this it appears that we are approaching the limits of our capacity to make better use of biological and edaphic/environmental resources (Rowarth & Parsons 2014). What is now required is to find ways of making fundamental changes in the natural evolved biological efficiencies of resource use in cultivated plants and domesticated animals. Retention of inputs in products, compared with their loss/release in surplus into soils and the environment, is also required. This improvement in systemic ‘harvest index’ must avoid associated degradation of organic matter stocks in soils. These are considerable challenges and require unprecedented advances in understanding of the function of plants, animals and soils, and their interactions.

At the same time, energy and water infrastructure must be remodelled (Ehrlich & Ehrlich 2013). The challenge is clearly not for agricultural scientists alone and world market requirements must be a part of the evolution of the new thinking.

Current thinking

New Zealand agriculture has already moved into ‘precision’ application of inputs (e.g. Hedley et al. 2010; Yule & Grafton 2013). Suggestions that what is now required could involve reallocation of land surface to totally different species, or the reallocation of production between flat land and slopes, are just fine-tuning current systems while overlooking the problems with yield and environmental effects (e.g. erosion, nitrogen loss and GHG production).

Some of the challenges and future directions for New Zealand agriculture were discussed in the book Future food farming (Emerson & Rowarth 2009), the most extreme being fenceless farming (Archer 2009). Since then, New Zealand food production discussion has ranged from destocking and ‘back to the 1980s’ (e.g. M. Joy, Massey University, pers. comm. Charles Fleming Medal Tour, Hamilton, 25 June) to total housing of cows (Piddock 2014). ‘Agritechture’ has been proposed by Julian Cribb (author of The coming famine): giant floating greenhouses and translucent vertical urban farms providing half the world's food by 2050 (Cribb 2014). However, such systems would require large amounts of energy: desert agriculture complexes supported by ‘nuclear agro-industrial complexes’ have not yet eventuated because sufficient cheap energy has never been produced by nuclear power to enable large-scale agriculture to develop in that direction (Ehrlich & Ehrlich 2013). Of further note is that agriculture has not moved towards feeding people protein extracted from leaves or bacteria grown on petroleum (Ehrlich & Ehrlich 2013).

Picking systems as solutions is risky; there are already examples of failures caused by focusing on one factor (e.g. species or trait, Chapman et al. in press) to provide the solutions (see Parsons et al. 2011 for explanation of plant breeding failures) and many examples (e.g. groundnuts in Kenya and Jatropha in India) of solutions that turn out to be failures.

Sustainable intensification

A framework for sustainability formulated by Smyth & Dumanski (1995) was adopted by the International Soil Science community during the 1990s. Sustainable intensification is the ‘new’ version, which outlines that ‘strategic intensification elevates yields in under-yielding nations and so greatly reduces land clearing and GHG emissions’, plus preserves biodiversity (Tilman et al. 2011, Conclusion). However, although sustainable intensification denotes a goal, it does not specify how this goal should be attained or which agricultural techniques should be employed. Garnett et al. (2013) suggest that ‘diverse approaches should be tested and assessed, taking biophysical and social contexts into account’ (p. 33).

The above suggestion points to a fundamental science approach. In previous proposals of ‘solutions’, what has been missing is proof of concept, a key characteristic of which is a highly critical experimental design, whether at a molecular, pot or field scale. Proof of concept can remove areas of research or fine-tune the goals (e.g. Parsons et al. 2011). It can also involve field data.

Fundamental research can have critically valuable practical value. Investigating ‘proof of concept’ (Parsons et al. 2011, 2013) in the first instance avoids wasting vast amounts of money and time in following the wrong path or trying to prove that a hunch actually works in practice. Focusing on proof of concept assists in overcoming the dogmatism that can skew research direction created when people, not necessarily the scientists involved, are ‘sure that it will work’.

Computer models

Computer models have long been used to bring together large amounts of information, gain insights and advance the design e.g. of better aircraft, power plants and not least of biological systems. There is, however, an important distinction between the kind of empirical ‘data-’ and ‘observation-’ driven models that have become favoured by the agricultural industry and the kind of process-based approaches used in engineering and physics. It is the latter that are increasingly needed for exploring sustainable food production options. Biological systems are extremely complex, dynamic and potentially even more demanding than engineering systems because they involve ‘adaptive materials’. There is a capacity for changes in function within the organisms themselves, as well as through changes in the balance of functional types in the community. Biological systems also face a changing climate.

In pastoral agriculture, empirical models have prevailed in attempts to examine nitrogen inputs and emissions/releases. Their reliance on past observations of the very thing they ‘predict’ is reassuring to industry and policy. However, this makes them inappropriate for exploring the future (for which empirical data are lacking) or even for understanding situations beyond those from which the input data were derived. After modelling data from 264 farms, Anastasiadis & Kerr (2013) could explain less than half of the variation in the nitrogen use efficiency of dairy farms using geophysical factors, specific mitigation technologies and practices that move emissions across farms such as wintering-off animals.

By contrast, dynamic, process-based models are available (e.g. Thornley 1998), although their use in New Zealand has become less common. Unlike empirical models, process-based models can be used to identify just what function needs to be changed to alter the fundamental efficiency of resource use of the agricultural ecosystem, and what would be the long-term and dynamic consequences. They are thus appropriate for ‘rethinking’ the options for agriculture and our land-based industries, and in identifying what fundamental research is needed.

It is perhaps the attention to ‘theory’ that makes process-based models less popular to applied industry, but their capacity to explore topical, practical and serious issues is evident.

It is of critical concern for New Zealand that any mention of ‘fundamental science’, ‘theory’, ‘complexity and dynamics’ is not seen in future as antithetical to making practical progress. Seeing agriculture as something that has a comfortable rural familiarity is not only thwarting progress at present, but runs the risk of diverting the next generations of bright and advanced thinkers from attending to one of mankind's greatest problems.

The science system

Current capacity to improve the use of biological and edaphic/environmental resources in New Zealand is being constrained by capability and capacity in the science system (Rowarth et al. 2014). Given the unique nature of our resources, relying on ‘fast-follow’ solutions to these problems based on research and development (R&D) developed overseas would be extremely risky, verging on irresponsible. The New Zealand agribusiness industries, from farm to fork, must be supported in order to maintain the bulk of the export economy.

Step one is to focus on science excellence, with track record a key element; experienced excellent scientists with secure funding can then enable development of the next generation of scientists. Step two is to review the funding model where the competitive system has been wound back in theory through core funding, but essentially recreated within the research organisations themselves. Step three is redressing the balance between public good and sector interests’ investment to ensure a balance between economic, social and environmental goals. Step four is to increase research funding, as indicated by the Organisation for Economic Cooperation and Development and repeated governments—but increasing funding must be accompanied by changes to the funding model.

New Zealand's biggest challenge for science is to educate, value and retain motivated and able people, with proven scientific talent, from the lab-bench to the boardroom. Scientists that are immersed in their discipline, and working with others similarly immersed in theirs, are the ones most likely to be able to create what is needed for the future. They need to be allowed to have and to follow their own initiatives. This is not curiosity-driven research (as funded by the Marsden Fund); this is research focused on advances for the future, aligned with national goals.

Futures

An important point of sustainability is economic viability (Smyth & Dumanski 1995). The current pressure to produce more for less is the driver for intensification, but environmental limits are constraining production. For instance, nitrogen caps in the sensitive Taupo catchment have increased on-farm costs by 45% (E. van Reenen, Beef+Lamb, pers. comm. 27 November 2014). However, most consumers are not prepared to pay for sustainable food; the cheap option will almost always be preferred (Sustainable Food Supply Commission 2014). In addition, because it is inexpensive, it can be thrown away; it is estimated that 40% of food is wasted (Foley et al. 2011).

A major focus on maximising resource use efficiency through recovering and revaluing every component of the food chain—hence minimising waste—is vital. For instance, bioactive compounds have been identified in the lipid fraction of agro-industrial waste (da Silva & Jorge 2014); waste recovery increases the value of the original production.

Ehrlich & Ehrlich (2013) have suggested that natural scientists should collaborate with social scientists to develop ways to stimulate ‘foresight intelligence’ that ‘looks ahead and guides cultural change towards increased socio-economic resilience’ (p. 5). Thus, foresight intelligence assists with providing long-term analysis and planning that markets cannot supply. It is, however, increasingly obvious that the problems for food production are in energy: energy for growing plants in contained areas (e.g. agrotechture and deserts) as well as energy for making fertiliser and crop and animal protectants. It is vital that natural scientists work with engineers.

A further prospect is that food will be regarded as a global resource (Bebbington 2010), putting water-rich nations such as New Zealand at an advantage. Approximately a quarter of food is traded globally; New Zealand exports more than 90% of its production and so potentially has added value with embodied water. The advent of precision irrigation means that this water, once on farm, can be used efficiently (Hedley et al. 2010).

Conclusions

Overall, the calculations indicate that current technologies applied to less-developed countries will enable the population to be fed to a better state of nutrition than at present. However, overcoming the limitations will take considerable effort in enabling the technologies to be affordable. In addition, there is the challenge of minimising environmental impact. Traditional routes involving field trials to ascertain ‘which plants grow better’ cannot deliver what is now required; the tools are too broad-brush. Fundamental science is vital to make the fundamental changes that we need. What is proposed in New Zealand is research leading to a step change in efficiency. Good scientific research is a conditio sine qua non for progress; advances cannot be made without some sort of research. Funding the research process and allowing scientists with a track record to lead their research for public good should be the way to the required future.

Acknowledgements

Thanks to the many scientists and researchers with whom we have discussed ideas over the past 30 years.