04.12.2024
New Zealand Agriculture & Climate Change Conference 2024
Keynote address
New Zealand’s economy has been built on the backbone of agricultural research. Scobie’s often quoted 1986 calculation shows returns on investment annualised at 17% which highlights how much pastoral research has contributed to the growth in NZ productivity. When one looks at the range of research that has been conducted in the Crown Research Institutes (CRIs) and universities, much of this has required slow and cumulative efforts. Examples over the last 30-40 years include the genetic selection of livestock for beneficial traits, the discovery and selection of endophytes to promote the growth and protection of ryegrass and animal health, and the discovery that South American parasitoid wasps can suppress the destructive grassland pest, the Argentine stem weevil. All of this work is worth many hundreds of millions of dollars per year. The dairy industry has been the cornerstone of much of our economy particularly since the global financial crisis (GFC), but by and large, our economy remains largely based on commodity export trade with some relatively simple value-added products. There are of course exceptions, such as lactoferrin, and we also need to reflect on the lessons from A2 milk and whether in the future genetic approaches can produce foods or milk extracts with proven health claims. Within this environment we have seen how the pastoral sector has protected the economy during the emergence of a more entrepreneurial sector of our society and economy with the tech and innovation sector growing in strength.
The challenges that our pastoral agricultural sector now faces are clear. Given its contribution to the economy and to our rural and provincial societies it must adjust rapidly to the need to reduce methane production. The other greenhouse gases, nitrous oxide and carbon dioxide are also important, but we cannot ignore both the disproportionate forcing effects of methane. The Intergovernmental Panel on Climate Change (IPCC) has pointed to a compelling logic to giving priority to methane reduction. While there are many other sources of methane, both natural such as in tundra melting and anthropogenic as in fracking and in other forms of agriculture such as paddy rice, New Zealand must confront the burden of agricultural methane, both for existential and well as financial reasons. The costs of buying offshore credits to meet our Nationally Determined Contributions (NDCs) certainly creates risk to our near-term future.
I am no expert on ruminant physiology so there is nothing I can say that is particularly profound given who is in the room. But I have engaged peripherally on the question of agricultural greenhouse gases since 2008 when John Key as Leader of the Opposition talked to me about the need to establish an agricultural greenhouse gas research centre. In 2010, I chaired the scientific component of the inaugural meeting in Wellington to establish the Global Research Alliance and I chaired its collaborative grant funding mechanisms in its early years. But my focus today will be on where we go from here.
Internationally quite a lot of progress has been made meeting the challenge of ruminant methane production, but much of this progress is limited in its direct application to New Zealand in that our industry is based on pastoral grazing systems and not feedlots. Therefore, methane inhibitors such as 3-nitrooxypropanol (3-NOP), while starting to have real impacts on methane production in Europe, Latin America and Asia as a feed-additive; are not so here. Part of the reason for this is obviously the mode of application, but there have been other limitations in our system that seem inhibitory. By way of example our regulatory processes have for many years inhibited the company that developed (3-NOP) from working on long-acting forms for use in New Zealand. This highlights the country’s rather poor strategic approach that that, for a variety of reasons, has led to disparate decision-making. I could say much more about this as I initially linked New Zealand’s chemists to the company that developed (3-NOP) in order to create potential applications for our systems. I do not want to focus on this particular story but rather make the point that vital moonshot-type research requires a much more strategic approach towards regulation; the decisions of regulatory agencies are as critical to innovation and its adoption as the science itself. Contrast that with how quickly the system responded to Peter Beck in building a major aerospace industry.
I was in Europe last week. There I saw approaches that might get around the challenges that restrict the pastoral use of inhibitors. For example, could one manipulate the appetite preferences of calves so they would grow up with a preference to inhibitor containing pellets placed in the paddock? There are of course issues, beyond the scope of this talk. Again, I simply want to suggest the need to be innovative in thinking through how we can take the knowledge from feedlot systems and apply it domestically.
I was given the title for this address pointing to a focus on the role of new and advanced technologies in dealing with enteric greenhouse gas (GHG) eructation. However, I cannot attempt in a few minutes to consider all such possibilities or indeed all the ways that GHG reduction in agriculture may be addressed. In preparing for this talk I decided to read some recent views on ruminant methane production and what became abundantly clear is the complexity of the rumen and its physiology.
Methane production is not simply a matter of fermentation in the rumen producing hydrogen that feed a methanogenic bacterium that captures the hydrogen and releases it as methane as a byproduct of the way they produce energy for their own survival and reproduction. The system involves multiple classes of bacteria, as well as the very difficult to work on anaerobic archaea. Other bacteria in the rumen can also use hydrogen to produce acetate and proprionate and do not produce methane. Then to add to the complexity there are fungi and especially ciliate protozoa in the rumen that also affect the ecosystem. In turn the ecosystem is affected by many other factors including the genetics and physiological state of the host animal, by the forage it eats and its developmental history.
Such complexity has a number of implications. The first is why does it exist? Ruminants rely absolutely on fermentation non-digestible carbohydrates to produce the short fatty acids that are the dominant energy supply. This is the ruminant equivalent of our glucose and so evolution has ensured redundancy in the system and thus there has been no selection pressure against methane eructation. Secondly, evolution has equally created complex equilibria in the rumen with coexistence of many species and types of microorganisms as well as the genetics of the host ruminant. We are only beginning to understand how the microbial complex communities work and then most of that understanding has been with simpler aerobic systems. Thus, it is inevitable that perturbation of the system is more likely to have transient, than persistent, effects in that equilibrium via various pathways will be restored. Hence single perturbations will, as the many thousands of papers in the scientific literature show, have unequal and inconsistent effects and designing what might be an effective intervention in a farming system is very complex. Here inhibitors, if they are safe, may be the only real silver bullets, if we know how to use them. Further, at the end of the day, no intervention is worthwhile if it is not affordable, scalable and most importantly enhances, rather than inhibits productivity. Indeed, at the end of the day, we have seen in the sheep industry that enhancing productivity per animal the best way to get stock numbers down thereby reduce GHG loads.
But back to the complexity of the rumen and dealing with its microbially driven physiology, there is a raft of claims of agents and forages both, raw or pretreated, that might affect methane production. This list is long including saponins, pretreated forages, new forages, tannins, lipids and oils, bromoforms and algal extracts all associated with various claims of their actions in the complex ecosystem of the rumen. Reading such claims was totally confusing. Then there are the synthetic inhibitors such as 3-NOP that have been synthesised and that directly inhibit an enzyme in the final step of the methane production by the anaerobes – in particular, that conferred by the methyl co-enzyme M reductase.
Alternatively, there is a focus on manipulating the rumen microbial flora by vaccination of the animal, by probiotics or potentially inoculation with desirable strains. Then there are difficult to study bacteria and archea such as Quinella, that appear to compete for hydrogen and do not produce methane. Last week in Europe I saw claims of forage manipulation that markedly reduce methane production, enhanced productivity and was directly related to the much greater abundance of Quinella. One important fact we know is if we are to manipulate the rumen microbiome it is more likely to have lasting effect if done very early in the host’s life.
And then on top of all this we then have variation in the host animals. There has been compelling evidence of animal selection for lower levels of methane production but the genetic or epigenetic basis of this is unclear clear. Is it in the genetics of the host itself or the community in its rumen or both? To date the polygenetic and perhaps multispecies nature of such a response remains unclear.
But in all this complexity one tool stands out; that is the potential of artificial intelligence to make sense of a very complex system of systems. It is the type of data that traditional systems analyses are likely to be totally confounded by. Artitificial Intelligence (AI) is much more than ChatGPT and there are multiple forms of AI that can be used both to synthesise the enormous amount of available evidence from all sorts of sources and to construct complex models to define what might be most effective targets for serious efforts at focused intervention. Two weeks ago, I was at a meeting in London on the use of AI in science and the methane-free cow was highlighted as a place to go. It is likely that multiple interventions will be needed just as in humans dealing with many diseases need multiple elements to be addressed in parallel. I worry that we persist with thinking only in terms of a silver bullet approach when evolution tells us it may have to be a multipronged attack. New Zealand is woefully behind in developing and using AI – here would be a singular application that merits focus.
But AI has other potential. The Google company DeepMind’s inventor Dennis Hassabis, this year received the Nobel prize for using AI to predict the shape of proteins and in their public database Alpha Fold; here there is information on hundreds of thousands of proteins and their three-dimensional probabilistic structure. Why is this important is that enzymes are proteins and if we know the shape of a protein, then AI can also design molecules that will bind to target enzymes in ways that block its action. Similarly, it is also possible to use AI-based analytic approaches to explore for potential toxicity. This is now the primary form of drug design in human medicine, and we should be using it to think about methane inhibition. Can we find a non-toxic molecule that can be used in a pasture system either directly or through including in grasses say through endophytes to meet our needs? A third use of AI will be to do what I do not think has yet been done to – to make a multispecies polygenetic analysis of what are we selecting for between host and rumen inhabitants. And all this leads on to the other two technologies we need rapidly to get back on track with – synthetic biology and generic engineering.
I will not say much about synthetic biology except it is moving at pace. Essentially it is the combination of information technologies such as AI with molecular and cellular biology to produce viruses or bacteria that have designed properties. It may sound futuristic but if AI holds its promise could we design a bacterium that has the properties of Quinella but is readily cultured and able to be inoculated in the calf? More immediately we have genetic possibilities. CRISPR (clustered regularly interspaced short palindromic repeats) is the most exact of the bioengineering techniques and it holds so much promise for agriculture. While there have been some early and crude efforts with transgenesis to increase metabolisable energy in ryegrasses, CRISPR offer so many opportunities alongside more traditional genetic techniques. Can we change the soil microbiota, can we use endophytes and inhibitor delivery mechanisms? Further to this, can we manipulate the rumen microbiota to preferentially produce propionate or even release inhibitors? Once we know what we are selecting for in low-methane producing animals, can we use genetic analyses to accelerate the processes? Notably CRISPR is now being used widely in human medicine and mRNA vaccines (such as that of used against covid).
We have been inhibited for two decades by our attitudes to modern molecular biology. But attitudes and indeed the regulatory field are changing. We no longer can pretend to meet the needs of sustaining an agricultural sector in its current form without exploring these technologies. Here there are, of course, many issues – human, animal and environmental safety being at the top of the lists. We face these issues with any chemical put into the food chain and we have well-developed mechanisms to deal with such material. The global experience with GM and GE has not revealed any of the effects that early objectors had claimed. We continue to manipulate genes of plants all the time. I was at the International Atomic Energy Agency in Vienna last week. They still are extensively using radiation induced random mutagenesis followed by selection as the basis of a good deal of plant development especially in the developing world.
Social license is key in the food industry, and this not only relates to the use of the technologies but also in market responses. Thus, many countries are now recognising that they must enhance science literacy and science education as AI and life science technologies become critical to our futures. The sad state of science education in New Zealand is an inditement on our education system.
Solving complex problems requires systems approaches and the existential and escalating impacts of global warming cannot be shrugged off as someone else’s problem. The first world has created the problem and as part of the first world we must make our contribution to addressing it. Trade-offs will be needed, and the nature of those trade-offs affect many interests – hence why it is so difficult to confront. Technologies will not solve all of the issues, but for New Zealand avoiding addressing the methane challenge will simply create other fiscal challenges if we are to meet our NDCs.
In the interest of time, I have not talked about the other GHGs from agriculture, but the same technologies will have value in reducing nitrous oxide emissions.
In Europe and the USA, we are now seeing major claims of massive reductions in methane production in the next decade – but these are referring largely feedlot-based systems. However, we can probably no longer truly claim to be the most greenhouse gas-efficient milk producers in the world. New Zealand must greatly accelerate its research capacities in advanced technologies with the methane question being one obvious reason why. At the same time, we need to greatly expand our international science collaborations as we are risk of being left behind. Right now, we face the imminent prospect of non-tariff barriers levied against our pastoral products based on the levels of methane emission by our livestock. Science and science-based innovation are far more central to our future as a nation than has been realised governments over the last the 25 years.
