Testing ecosystems – a suitable future for agricultural innovation? – some thoughts on Agriculture 4.0


A brief history of agricultural innovation

Agriculture being itself a revolution, an anthropogenic intervention in the cultivation of plants instead of a mere collection in the wild, it has also changed and evolved in many ways since its start, roughly twelve thousand years ago. Techniques as seed collection, crop domestication, land clearing, soil preparation with animal traction, irrigation with basins or channels, fertilization with animal litter developed through the centuries independently and have been used for millennials in many parts of the world. Basic breeding, i.e. selecting food plants with desirable characteristics whose seeds were used for subsequent generations, and crossing, especially for cattle, are also known techniques since ancient times. Large empires, as the Egyptians, the Persians or later the Romans developed large scale agriculture and long-distance exchanges.

Middle Ages saw the development of crop rotations, the improvement of soil tilling techniques through a better plow, harness and a broader use of metal in utensils, thus increasing productivity and efficiency of work. Knowledge was spread though Arabs between Europe and Asia, and monasteries were important centers for information collection and transmission.

Much later, a new phase started with the systematic selection instead of random reproduction, breeding and crossing in the 18th century during the British agricultural revolution, the understanding of the transmission of plant traits with Mendel, and the increased mechanical work capacity with the industrial revolution gave agriculture a new spurt for increasing production. Better understanding of basic principles of plant growth, transmission of traits and facilitated change of the environment like soil and water allowed for systematic improvements. Additionally, new species were introduced from further away, among others facilitated by the extension of colonial empires and their search for new resources, or by large human migration movements.

More recently, chemical and microbiological advances in the 19th and 20th century, both contributed to the understanding and intervention in plant diseases, fertilization, thus opening the era of the modern, industrial agriculture. Better developed meteorological predictions allowed for a more precise planning. Crops, well selected and with an optimized environment, practically managed with upscaled machines, uniform landscapes with large monocultures are some of the typical features of the agriculture of the last decades in industrialized countries. Economical aspects, as investments from governments, ownership distribution of the land and the expansion of big corporations selling seeds and chemical aids additionally shape the agricultural landscape.

The evolution of agriculture, therefore, cannot be separated from parallel political1, technical, social2, 3 and economical4 and even philosophical-spiritual evolutions and changes. Human interventions in nature and agricultural practices and their outcome are intrinsically linked.

If challenges, discoveries, shifts in human organization and accidental events have been triggers for changes in agriculture in the past, what are today’s triggers and offered solutions, in the dawn of the fourth agricultural revolution?

Nowadays, we have, in addition to the past techniques and knowledge, at our disposition an even broader range of resources. They include novel techniques and tools as electronical devices – from drones to smartphones and sensors to measures a wide range of parameters. Refined irrigation techniques, artificial lights, greenhouses and hydroponics further untie the dependency on weather, temperature, sun, and soil characteristics. In addition, precise breeding techniques and increasing knowledge about traits, viruses5, mutations and transmission, allowing to intervene actively in genetic material, enlarge the field of possibilities. Additionally, fast commercial exchanges, relatively abundant energy sources, cheap labor and economical resources unequally distributed influence preferred techniques and practices in different areas. Our knowledge about precise plant and animal metabolisms, and capacity to measure and model many factors that influence their growth allow us to tailor inputs to a suited output. In response to current challenges, climate smart agriculture6 is one aspect that is under study. In parallel, the latest advances in technology – Artificial Intelligence or the Internet of Things for instance – are about to enter that picture on big scale through investments from different stakeholders. Together all these aspects draw a global picture of agriculture and all its related activities.

With that progress – defined generally by an increased yield and food security – many drawbacks have become evident. While environmental consequences of past practices were real – effect on the climate7, 8, on rivers, on the ecosystem – but were not that clearly associated until now due to the lack of adapted tools, it is a different picture today. Indeed, the consequences of today’s agriculture are multiple and unequally distributed. They include the effects of pesticides on human health, the energy cost of the industrial agriculture facing oil peak and the effects of climate change9, deforestation10, water crisis, antibiotic resistance11 and biodiversity loss12, 13 . These issues became at the same time some of the many triggers for a new change in the approach to agriculture. The change seems even more urgent when the main argument in favor of the industrial agriculture is getting questioned: the capacity of current agriculture to provide food security, especially where lack of economic resources and more severe climate change effects add up14, 15. However, the rising awareness of the deleterious consequences of presumably effective, positive progress raises the question if all progress is suitable, and which one we should focus our efforts on, and depending on which criteria. With increasing means of action, our responsibility in making timely and well-thought choices is requested.

Limits and risks – how to evaluate them?


Changes in the agricultural landscape are associated to political, technical, social1 and economical evolutions and changes, both upstream and downstream. The economical impact16-21 of environmental services and their loss has been now measured for years and is starting to be considered at its full measure. Few voices rose to evaluate beforehand the impact of technologies on the social aspects of agriculture on a larger scale22. And while social and economic aspects are essential to consider, they also depend themselves on the physical, biological consequences of agriculture on climate, the ecosystems and health.

Those consequences are often more visible on large scale, on a longer term, so when a practice is widespread, the consequences have a greater impact. So, if those consequences are negative, the negative impact will be greater. The most dangerous scenarios are either the long-term-large-scale or the explosion of the unexpected, both possibly overlapping. The first is for techniques that have positive short-term consequences, that are used on large areas in many parts of the world and that eventually bear negative effect that appear only after a long term. Often, it is difficult to link the consequences to their causes, paradoxically needing large scale negative effects to be able to measure and evaluate significantly the chain of causality. That is the process for long-term (chronical) toxicity, and in the case of agriculture, the health effects of pesticides on farmers, the bioaccumulation of the insecticide DDT and its degradation products or the destruction of pollinators with pesticides13, 23, for example. The other risky bet is made with changes that influence key processes, thus a change in the node can potentially influence all chains linked to that knot, provoking effects are mostly unpredictable and unforeseen with dramatic changes as a consequence. This could be the case of water soluble molecules that do not degrade, or degrade into more toxic, long-lasting components: water being the node, as it is essential in all living processes, a solvent and reagent in many mineral, chemical, physical and geochemical processes, that molecule can potentially spread, and interact at all levels of the water circle. This could also be the case in genetical intervention, when key functions, or polyvalent genetic expressions are touched. The concert of simultaneous expressions and their physiological consequences are not always predictable. The effect is even greater when the scale of observation is increased: If the consequences on the organism itself are already hazardous, what is predictable in terms of horizontal gene transmission or ecological population dynamics, for example? How much is under our control? Is it possible to intervene and until what point?

Known vs unknown risks – how much do we control? how much is a controlled risk and what is an acceptable risk?

Risks are easier to evaluate as a single risk, independent from other interacting factors. For example, the effect of one chemical on plant growth, on microbial occurrence24, or on tumor incidence, the mobility of an agent in the water circle, the effect of a pest on a crop filed, the resistance of one genetic variation towards a virus25 to name a few.

However, the risks are often more complex than that, since the phenomena mostly occur in parallel and influence each other. For instance, one pesticide primarily impacts the weed, but can also spread into the water, influence other organisms unintendedly, as pollinators, mammals etc., have effects on health and accumulate on the long term in the food chain, and disrupting the viability of organisms that are geographically distant and not directly affected. The effects on one organism can cause a cascade of effects on a whole ecosystem with higher, accumulated risks at the end. This is the more real picture of pollinator decline26, 27, not only due to chemical agents, but to an overall approach of agriculture as an industry that is built on symptom cure, the practicability of landscape uniformization and an unilateral economical motivation. Similarly, the consequences of genetical modifications: what other changes are induced with the change of one metabolic pathway is often hard to say. Further, a successful change that increases a suited trait of one species can completely change its ecological niche and behavior compared to the wild type, forcing to evaluate risks in terms of population dynamics as for a new, potentially invasive, species.

The concept of time-bomb is a term that comes with this type of intervention.  Many useful models are developed to evaluate one effect with more factors, from meteorology to epidemiology and ecology. But per definition, a model can only be as good as the data that is given as an input. Models do depend on an understanding of the parameters that have to be measured, questioned, evaluated and chosen, and on the paradigms that underly the model design. Good datasets coupled with deeper understanding of a problematic allow for efficient models and tools development in order to respond to current issues28. Therefore, the understanding of multiple interacting factors in connection with agriculture and the corresponding datasets are a necessity equally for risk assessment and for future innovation.

Why ecosystems?


Ecosystem inspired agriculture – in its simplest form as a two-element polyculture – raises enthusiasms and criticisms equally. The defenders stress the potential for increased resilience29, the use of natural predators and the better soil protection30, especially for perennial crops, whereas the opponents criticize the unpracticality with respect to machinery from the soil preparation to harvest cycle and the poor evidence of sufficient yields and economic viability. Nevertheless, whatever point of view is put forward, ecology plays a role and is both influenced by the agricultural practices and an influencer of the agricultural output. The growing attention that pollinator decline got in scientific publications13, 16, 17, 26, 31 and in mainstream media32 shows the increasing awareness of the interlinked dynamics of agriculture and environment.

Therefore, in any case, studying ecosystem interactions in the context of agriculture is a necessity, both in the optic of not harming it and optimizing yield on the long term.

Ecosystems and their dynamics have been studied in a perspective of description, conservation or risk assessment in the context of climate, or socioeconomics, or to study the negative impact of agriculture on native and surrounding nature. These findings and studies are very valuable, as they give insights in important parameters as population stability33 – smaller populations are more prone to extinction,  or diversity losses – that decrease the risk of survival as the connectivity to organisms is lost34. Additionally, other factors as their homogeneity or heterogeneity, their spatial distribution, also influence their survival and thriving34-36, especially when facing challenges. Furthermore, the place and role of an organism relative to others defines the modularity of an ecosystem and thus it influences the risk of perturbation or extinction.  Further, seemingly insignificant features as day-night temperature differences can impact blooming of plants37 and affect a cascade of organisms depending on the functionality of the affected one. For instance, a recent model showed that “random node and specialist extinctions are always detrimental for the system, and the response of the system is approximately proportional to the percentage of loss of species”, while other dynamics can sustain a larger extinction percentage before the system totally collapses 38. Reciprocally, it is primordial to understand precisely those principles and switch points when the aim is to build a functioning ecosystem from scratch – or from an incomplete one. Indeed, including the ecosystem way of thinking in agriculture means using its productive and resilient dynamics for an optimized, sustainable yield. Therefore, the study of ecosystems and the knowledge of its dynamics and nodes, of the fine-tuned regulations of organisms to the larger network of interactions, is equally essential to plan effective conservation measures and to design agricultural innovation.

Thinking innovation in terms of ecosystems not only opens up a new field of agricultural research and practice, unlocking its potential39. It also forces the inclusion of other aspects of human activity40 as transportation, health care41, education42, infrastructure and energy use to name a few, in order to reach an efficient output – stability and resilience for health and food security. With its performance closely linked to climate and its changes, agricultural innovation needs similar and complementary strategies43. Indeed, the coordination of actions on a large scale is a necessity and would be hardly doable without the involvement of all in a fair way, and understanding the leverage points is essential44, 45. The topic of pollination is a good example of how a mismatch in strategy of different actors and stakeholders can miss the recovery of the ecosystem service by pollinators46.

Knowing well the interactions is all the more important as well-meant additions to agricultural systems can turn out to be deleterious if the underlying ecological dynamics were poorly understood, thus motivating to abandon all together the path of that innovation with potential, rather that correcting its application and context.


In a nutshell


Innovation is always happening, and changing needs, even more with regards to increasing challenges as climate change, therefore the evaluation beforehand of the consequences of a seemingly promising discovery or technique is a necessity. In a globalized world, with a closing window for the field of action against irreparable damages on nature as we know (and don’t yet know) it, the risk assessment is even more crucial, as every innovation, potentially spreading fast and far will have a large impact, or on the contrary, an expected impact does not happen due to lacking complementary actions. This impact, with both its destructive and regenerative side, is an aspect that needs better understanding and a stronger voice. Ecology is a factor that cannot be ignored and should be best included while thinking of agricultural innovation of any kind. Indeed, biodiversity is a key element, even if not fully estimated and understood, that bears a huge potential for future solutions that we cannot afford to neglect. Similarly, conservation efforts should therefore not be a palliative measure but an inherent part of the innovation process.

There is a real and urgent need to understand ecosystem interactions and to develop methods to study their dynamics, possible disruptions and synergetic effects, both to evaluate risks and to plan wisely and efficiently future developments in agriculture.

Thinking in terms of interlinked ecosystems not only opens up a new field of agricultural research and practice, unlocking its potential. It also forces the inclusion of other aspects of human activity, in order to reach an efficient output – stability and resilience for health and food security. Indeed, the coordination of actions on a large scale is a necessity and would be hardly doable without a broad, coordinated involvement of all actors, in an equilibrated way.

To go further

  1. Les Agrocarburants Industriels Provoquent la Faim et la Pauvreté. Cahiers de La Via Campesina. 2009;1.
  2. Conférence Paysanne Internationale: Stop Aux Accaparements de Terres. Les cahiers de La Via Campesina. 2012.
  3. Hall R, Hornby D, Lawry S, Leopold A, Mtero F and Samii C. The Impact of Land Property Rights Interventions on Agricultural Productivity in Developing Countries: A Systematic Review. the campbell collaboration. 2012.
  4. Trostle R. Global Agricultural Supply and Demand: Factors Contributing to the Recent Increase in Food Commodity Prices. 2008.
  6. Negra C, Vermeulen SJ, Barioni LG, Mamo T, Melville P and Tadesse M. Brazil, Ethiopia, and New Zealand lead the way on climate-smart agriculture. Agriculture & Food Security. 2014;4.
  7. Koch A, Brierley C, Maslin MM and Lewis SL. Earth system impacts of the European arrival and Great Dying in the Americas after 1492. Quaternary Science Reviews. 2019;207:13-36.
  8. Gilgen A, Wilkenskjeld S, Kaplan JO, Kühn T and Lohmann U. Did the Roman Empire affect European climate? A new look at the effects of land use and anthropogenic aerosol emissions. Clim Past Discuss. in review, 2019.
  9. Springmann M, Clark M, Mason-D’Croz D, Wiebe K, Bodirsky BL, Lassaletta L, de Vries W, Vermeulen SJ, Herrero M, Carlson KM, Jonell M, Troell M, DeClerck F, Gordon LJ, Zurayk R, Scarborough P, Rayner M, Loken B, Fanzo J, Godfray HCJ, Tilman D, Rockström J and Willett W. Options for keeping the food system within environmental limits. Nature. 2018;562:519-525.
  10. DeFries RS, Rudel T, Uriarte M and Hansen M. Deforestation driven by urban population growth and agricultural trade in the twenty-first century. Nature Geoscience Letters. 2010.
  11. McKenna M. Antibiotics set to flood Florida’s troubled orange orchards. Nature. 2019;567:302-303.
  12. IUCN. IUCN Red List of Threatened Species.
  13. Allen-Wardell GP, Bernhardt R and Bitner A. The potential consequences of pollinator declines on conservation of biodiversity and stability of food crop yields. Conservation Biology. 1998;12:8.
  14. UN. Wake up before it is too late – Make agriculture truly sustainable now for food security in a changing climate – Trade and Environment Review 2013. Paper presented at: United Nations Conference on Trade and Development (UNCTAD); 2013; Geneva.
  15. Searchinger T. Creating a Sustainable Food Future ; A menu of solutions to sustainably feed more than 9 billion people by 2050. World Resources Report 2013–14: Interim Findings. 2015.
  16. Gallai N, Salles JM, Settele J and Vaissiere BE. Economic valuation of the vulnerability of world agriculture confronted with pollinator decline. Ecological Economics. 2009;68:810.
  17. Kevan PG and Phillips TP. The economic impacts of pollinator declines: An approach to assessing the consequences. Conservation Ecology. 2001;5:8.
  18. Pimentel D, Wilson C, McCullum C, Huang R, Dwen P, Flack J, Tran Q, Saltman T and Cliff B. Economic and environmental benefits of biodiversity. Bioscience. 1997;47:747.
  19. D. Levin M. Value of Bee Pollination to U.S. Agriculture; 1983.
  20. Carreck N and Williams IH. The economic value of bees in the UK. Bee World. 1998;79:115.
  21. Morse RA and Calderone NW. The value of honey bees as pollinators of US crops in 2000. Bee Culture. 2001;128:1.
  22. Rose DC and Chilvers J. Agriculture 4.0: Broadening Responsible Innovation in an Era of Smart Farming. Frontiers in Sustainable Food Systems. 2018;2.
  23. Johansen CA. Pesticides and pollinators. Annual Review of Entomology. 1977;22:177.
  24. Fox JE, Gulledge J, Engelhaupt E, Burow ME and McLachlan JA. Pesticides reduce symbiotic efficiency of nitrogen-fixing rhizobia and host plants. Proceedings of the National Academy of Sciences. 2007;104:10282-10287.
  25. Virus lurking inside banana genome has been destroyed with CRISPR. newscientist. 2019.
  26. Kremen C, Williams NM and Thorp RW. Crop pollination from native bees at risk from agricultural intensification. Proceedings of the National Academy of Sciences. 2002;99:16812.
  27. Liu ZJ and Zhang FT. Issues in sweet cherry cultivation. Shanxi Fruit. 2002;1:57.
  28. Humphreys AM, Govaerts R, Ficinski SZ, Nic Lughadha E and Vorontsova MS. Global dataset shows geography and life form predict modern plant extinction and rediscovery. Nature Ecology & Evolution. 2019.
  29. Matsushita K, Yamane F and Asano K. Linkage between crop diversity and agro-ecosystem resilience: Nonmonotonic agricultural response under alternate regimes. Ecological Economics. 2016;126:23-31.
  30. Blanco Sepúlveda R and Aguilar Carrillo A. Soil erosion and erosion thresholds in an agroforestry system of coffee (Coffea arabica) and mixed shade trees (Inga spp and Musa spp) in Northern Nicaragua. Agriculture, Ecosystems & Environment. 2015;210:25-35.
  31. Aizen MA and Feinsinger P. Habitat fragmentation, native insect pollinators, and feral honeybees in Argentine Chaco Serrano. Ecological Applications. 1994;4:378.
  32. Adding Value to Agriculture with Pollination Research.
  33. Sasaki T, Furukawa T, Iwasaki Y, Seto M and Mori AS. Perspectives for ecosystem management based on ecosystem resilience and ecological thresholds against multiple and stochastic disturbances. Ecological Indicators. 2015;57:395-408.
  34. Yang G, Wagg C, Veresoglou SD, Hempel S and Rillig MC. How Soil Biota Drive Ecosystem Stability. Trends in Plant Science. 2018;23:1057-1067.
  35. NECSI. Understanding Biodiversity. 2019.
  36. Allan E, Weisser W, Weigelt A, Roscher C, Fischer M and Hillebrand H. More diverse plant communities have higher functioning over time due to turnover in complementary dominant species. Proceedings of the National Academy of Sciences. 2017.
  37. Frastelli D. Due gradi in più in 40 anni: l’allarme degli albicocchi. Il Tirreno. 2019.
  38. Sheykhali S, Fernández-Gracia J, Traveset A and Eguíluz VM. Extinction-induced community reorganization in bipartite networks. Applied Network Science. 2019; arXiv:1905.13007 [physics.soc-ph].
  39. Boisselet T. Missing the potential of evolutionary ecology. 2018.
  40. Zawadzka JE, Corstanje R, Fookes J, Nichols J and Harris J. Operationalizing the ecosystems approach: Assessing the environmental impact of major infrastructure development. Ecological Indicators. 2017;78:75-84.
  41. Kirjavainen PV, Karvonen AM, Adams RI, Täubel M, Roponen M, Tuoresmäki P, Loss G, Jayaprakash B, Depner M, Ege MJ, Renz H, Pfefferle PI, Schaub B, Lauener R, Hyvärinen A, Knight R, Heederik DJJ, von Mutius E and Pekkanen J. Farm-like indoor microbiota in non-farm homes protects children from asthma development. Nature Medicine. 2019.
  42. Pasiona SP. How to Engage the Youth, Teach Agriculture in Senior High School. Agriculture Monthly. 2017.
  43. Lamontagne JR, Reed PM, Marangoni G, Keller K and Garner GG. Robust abatement pathways to tolerable climate futures require immediate global action. Nature Climate Change. 2019;9:290-294.
  44. Davis K, Chadès, I, Rhodes, JR, Bode, M. General rules for environmental management to prioritise social-ecological systems research based on a value of information approach. J Appl Ecol. 2019.
  45. Callaghan CT, Major RE, Lyons MB, Martin JM, Wilshire JH, Kingsford RT and Cornwell WK. Using citizen science data to define and track restoration targets in urban areas. Journal of Applied Ecology. 0.
  46. Breeze TD, Vaissière BE, Bommarco R, Petanidou T, Seraphides N, Kozák L, Scheper J, Biesmeijer JC, Kleijn D, Gyldenkærne S, Moretti M, Holzschuh A, Steffan-Dewenter I, Stout JC, Pärtel M, Zobel M and Potts SG. Agricultural Policies Exacerbate Honeybee Pollination Service Supply-Demand Mismatches Across Europe. PLoS ONE. 2014;9.