This is the third of a three-part series of reporting on three plenary presentations aiming to ‘set the scene’ on the first day of a Science Forum 2016 organized by the Independent Science and Partnership Council (ISPC) of CGIAR and held in Addis Ababa, Ethiopia, 12–14 Apr 2016. The reports were prepared by the academics from 123helpme.org, who conducted their own researches in order to help with the developing of understanding of the topic in hand.
Following two presentations made to open the first day of Science Forum 2016, the first by development economist Stefan Dercon, of the University of Oxford and the UK’s Department for International Development (DFID), on the tenuous evidence linking agricultural research for development and rural poverty reduction (Rethinking pathways for rural prosperity: The agricultural challenge), and the second by gender expert Ruth Meinzen-Dick, of the International Food Policy Research Institute (IFPRI), on when agricultural development projects help close ‘gender asset gaps’ and when they do not (Closing gender gaps in control over assets for lasting development outcomes), was a presentation made by climate scientist Mark Howden.
Howden is director of the Climate Change Institute at the Australian National University. He has helped develop national greenhouse gas inventories and serves as vice chair of Working Group II of the Intergovernmental Panel on Climate Change (IPCC). His work focuses on innovative climate change adaptations for systems that we value: agriculture, food security, natural resources, ecosystems, biodiversity, energy, water and urban environments.
Howden’s presentation looked at the successes, opportunities and risks we’re facing and began with evidence of just how much the climate change challenge for agriculture is changing. The following is a transcript of Howden’s talk, lightly edited.
A figure from the IPCC’s 5th Assessment Report shows the proportion of studies reporting positive or negative changes in major crop yields over a 100-year period, from 2010 to 2109.
Impacts of climate change on major crop yields, with increases shown in blue and decreases in orange/brown. The intensity of the colour represents the positiveness or negativeness of those changes. So the bright blues, for example, represent significant increases in major crop yields.
What you can see is that the proportion of studies indicating increased global yields decreases over time, while the proportion of studies showing reduced global yields increases significantly over time, with up to 50–100 per cent yield reductions in the major crops—wheat, rice and maize—expected by 2109.
What jumps out at you is that after 2050 none of the blue bars, which show increases in agricultural yields, are for developing countries. So there are no positive analyses coming out of this meta-study—no agricultural benefits of climate change—for developing countries, where most of the world’s poor people live.
This is a worrying picture in terms of climate change as a function of increases in temperature, often reductions in rainfall, changes in evaporation, and increases in climate variability.
Of course, this is a very broad picture. It doesn’t go down to the detail needed to make national or local decisions. This analysis is also limited by high levels of variation and uncertainty. There remains a surprising degree of uncertainty, for example, among the projections of different agricultural crop models for any given scenario.
This also doesn’t do much justice to climate change adaptation, Only a very limited range of adaptations were assessed, although including adaptations in this kind of assessment would tend to reduce the negative impacts of climate change, and possibly increase the positive.
There are many issues that aren’t addressed in this sort of study. Pests and diseases are largely absent from these analyses. The long timescales of these analyses are often problematic, looking at 15, 20, 50 or 100 years ahead of time, whereas the planning horizon for many people is next year, or maybe five years from now. So the planning horizons of a lot of these analyses are much longer than those of the sorts of people we’re talking about helping in this meeting.
And, lastly, climate variability is very limited in terms of representation in the global climate models at the moment. While those climate models—when we actually put them through crop models—show very, very large increases in the variability of yields of major crops over time, they don’t actually show much change at the moment. But we already see significant variability occurring right now as a function of climate change. So we’re running a significant risk of underestimating the risk of climate change affects on these major crops.
And there are many limitations and gaps in these types of studies. Livestock, for example, is under-represented here, as are minor and orphan crops. (I’ve been told I have to call these ‘nutrient-dense, climate-resilient future crops’ to be politically correct now.) And there is a large omission, now starting to be corrected, of investigating the nutritional aspects of foodstuffs as a function of climate change. And there is starting to be some recognition of value chains.
There is almost total absence of understanding of the social norms and institutional arrangements that determine how individuals and villages and broader communities respond to climate change.
The stability dimension of food security, which is really crucial, is much, much less known and much, much less studied than the availability and access dimensions of food security, which focus on yield and which got critiqued severely by this morning’s first speaker (Stefan Dercon).
These gaps often align with the concerns of poor people and of less developed nations. So the biggest gaps are where we actually have the biggest problems. Which means we’re actually not well positioned as a science community to inform many of the discussions and decisions. My assessment is that the likelihood of rapidly closing those gaps is small, due lack of funds and interest and also lack of human resources.
Often what we’re trying to do here is to take research to an operational level, so that, for example, people will make different decisions on the basis of the research we do. A recent paper by Lacey et al. looked at this and compared what we do in climate adaptation work with, say, what’s done in the health sector. In the health sector, there are well-established ways of moving from clinical research to assessing treatment effectiveness to rolling out widespread medical care among members of a community.
As we read down the list in this slide from top to the bottom, we see that there are different types of research, different types of science and different types of methods. There are also different institutions. Each of the transitions between each of these boxes has distinct institutional arrangements that manage the transition of that information to safeguard the people we intend to help.
If you’re a mad scientist who developed a medical treatment and—without any of these safeguards—actually ran out and promoted it among a community, when in fact it was a very risky thing to do, you would probably end up in jail. Yet our research activities are often promoted as operationally appropriate with none of those checks and balances to mediate their usage being in place.
We must think more carefully about how we move agricultural research into operational modes. We need to think about the nature of the science that’s being done. When we’re trying to understand a problem, that science is going to be different from the science we use when we start trying to understand solutions. And that type of science is going to need to change again when we start to implement those solutions.
We need to change not only the type of science but also the institutional arrangements that mediate the transitions among those different components, those different phases, of the innovation or solution pipeline.
In a sense, climate adaptation is a no-brainer. If you don’t adapt your farming system or your value chain to a changing environment, you’re either going to underperform or you’re going to increase the risk you’re exposed to. And neither of these are desirable, whether you’re a poor farmer or a businessperson.
So climate adaptation is just common sense. It should just be part of normal practice. But what we often find in the research domain is that climate adaptation is not dealt with from an operational perspective.
If we look at the literature that was summarized in that first figure, almost all of those adaptation analyses were single, were simple, were technical and were focused on short-term changes to existing systems. There’s very little in the analysis that takes into account the real-world nature of change, which is often complex, compound (many things changing at once) and highly contextual (one village or farmer will often do something different from the next). It’s also often strategic (forward thinking about succession planning or provision of funding for schoolchildren). It’s often tacit (understood rather than codified knowledge). And it’s often socially and institutionally mediated in all sorts of ways.
This is the nature of real-world adaptation. Yet it’s largely absent from adaptation studies. And of course we can’t look at adaptation by itself. It has to link with climate change mitigation, with gender concerns, with a whole range of other sustainable development agendas, which largely don’t get included.
Particularly, those studies focused on incremental changes to existing systems. And if you only focus on that, you may miss opportunities and incur risk through maladaptation. We need to think more broadly about systemic change (changes in the nature of a system) and transformative change (fundamental changes to the nature of a system). Sonja Vermeulen and her colleagues at the CGIAR Research Program on Climate Change, Agriculture and Food Security (CCAFS) have done some really good work on that, along with other people. The idea is to deal with adaptation in a comprehensive way.
Everett Rogers over half a century ago looked at adoption paths, focusing on particular characteristics: relative advantage, compatibility, manageable complexity, trialability and observability. All of these things are really difficult to do in a climate adaptation frame because we don’t have a time machine. We’re not Dr Who. We can’t travel into the future, do an experiment and come back and report on it. So we need to think carefully about how we actually frame climate change adaptation and make it relevant—real—for people.
To do that in Australia, we’re looking at frost risk. At the same time that we’re experiencing increased temperatures in southern Australia, the risk of frost is increasing. Decade by decade, the frost risk has changed. In the most recent decade, frost risk has changed fundamentally. If we followed best practice guidelines from extension officers, which use one-hundred-year climate data to estimate frost risk, we would be worse off than if we made changes based on the last decade. You are better off changing your system to better suit environmental changes decade by decade. Doing so incurs less risk and generates greater productivity and profits. So changing with the times is fundamentally important.
One of the things we’ve learned from the field is that when we first interact with a group of farmers, their focus is entirely on agronomic changes, such as planting dates or cultivation practices. But as we interact with a given group of farmers over time, the thing they focus on at the end is strategic business management—the ability to think carefully about different options and to make better decisions.
We can leapfrog a whole stack of such agronomic processes and instead train people to be better strategic decision-makers. That of course will often involve a real adaptation building exercise. So we need to start thinking about social norms—about information and social networks.
This is some work by Anne-Marie Dowd and others looking at the social networks of people who are making incremental versus transformational changes to their systems. They’re fundamentally different. Incremental adapters had very strong, very local networks, essentially confirming the way things are done and limiting ideas about what could be done because of those social norms. The transformational adapters had much weaker social networks, which were also often located at a distance. They weren’t constrained by the ways things were done and how people around them were saying they should be done.
The information networks of these two groups were completely the opposite. The incremental adapters were largely happy with the information they had; they weren’t going out and seeking it and the information they were getting was largely from local sources. They weren’t thinking about the long term or about options being explored in distant places. But the transformation adapters were. They were sponges for information and effective at pulling information that allowed them to make more complex and risky decisions.
If we don’t take those factors into account, we’ll miss some fundamentals of effective adaptation.
As we’ve heard from previous speakers, it’s really important to think about value chains and to think about adaptation along those value chains. Adaptation propagates change up as well as down a value chain. We have to think about adaptations systematically and coherently, and both up and down the value chain.
Part of adaptation research involves research ethics. We must think about how we, as a research community, operate effectively, and start to remove some of our practices, or transparently address some of our practices, that may increase the risks incurred by the people we’re trying to help.
Different groups providing adaptation information often have significant conflicts of interest. Take disciplinary bias, for example. If you go along to an economist and ask what adaptation you should make, you’ll get an economic answer. If you go to a plant physiologist, you’ll get a completely different answer. Go to a social scientist and you’ll get a completely different answer again.
This is unacceptable. If you went to a doctor and one doctor said you’ve got the flu and another said you’ve got pneumonia and a third said you’ve got cancer, you’d think there was a real problem with that system. Yet that’s what we scientists tend to do. We have to be highly transparent about the biases we bring to the advice we give the people we’re trying to help.
In summary, we must think about how to make our science ‘real’. It has to be relevant to the decisions that people are making. It also has to be useful to the things that they’re going to be doing next. And it still has to be robust—rigorous, repeatable, appropriate. It also has to be robust in terms of the outcomes we promise that our science will deliver. So we have to have much better monitoring and evaluation systems.
We need to think carefully about the options we present to people—naming the pros and the cons, the benefits and the risks—of those options in ways we don’t do now.
Finally, we have to learn to talk and work better with our partners. Every step along those agricultural-research-for-development transitions—from describing problems to determining solutions to implementing interventions—increases the critical role of partnerships. That means we have to think about power relationships in our research. Yes, we need to share knowledge. And yes, we need to share experiences. We also need to share power.
Real, robust options are what talk to our partners best. They are the pathways for change.
Story by Susan Macmillan, ILRI