sp12 by qingyunliuliu


									An Open Access Journal published by ICRISAT

                                      Climate Proofing Agricultural Research Investments

                          Peter Jones1, Andy Jarvis2, Glenn Hyman2, Steve Beebe2 and Douglas Pachico2

                  1 Waen Associates, Dolgellau, Wales, UK
                  2 Centro Internacional de Agricultura Tropical (CIAT), Cali, Colombia


         The authors wish to thank the reviewers of the paper for some telling criticism and some very useful hints.


         The case for impending climate change is now proven. Governments can decide, by their action or inaction, to

         what extent the change will occur; the International Agriculture Research Community (IARC) will have no say

         in this whatsoever. It is up to the IARC to try to maintain objectives in the face of the possible scenarios. In this

         paper we discuss the various types of agricultural research projects in terms of their time to fruition and the

         expected longevity of their results. We look at the information requirements for ensuring that project products

         have the necessary lifetimes to justify the investments in the research. We show that strategies differ depending

         on the type of research that is undertaken. Basic research into genetic traits and capacities within the available

         germplasm has to be planned in the long term with outcomes in mind. The vulnerability of the populations and

         agricultural systems that use developments from this basic research now places its priority setting in a changing

         climate and world concept. Ensuring that the germplasm is available for use has taken on a critical new

         importance with recent studies. Germplasm banks comprise a small fraction of what we will be relying on for

         the future. Well over 90% of useful genetic variability may still be in the wild. This has to be considered

         carefully in setting out research objectives. Plant breeders, who will put together the results of the basic research

         into useful packages, now have an uncertain target to aim for when regarding future climate conditions. They

         may not be able to choose their testing sites in present climates to target agricultural populations that will be

         using their products in the future. Agronomic and agricultural development projects face the most difficult task.

         How do we develop stable farming systems in an environment that is not only unstable, but also changing so

         slowly that the farmers cannot see, or even envisage, the changes. These are some examples of the problem. The

         paper sets out to categorise the types of research and information that will be necessary at all levels. We draw on

SAT eJournal | ejournal.icrisat.org                                                             December 2007 | Volume 4 | Issue 1
An Open Access Journal published by ICRISAT

         experience from the CGIAR system and from CIAT in particular. We show that a number of software tools have

         been developed that can address some of these problems.


         International agricultural research has developed into a complex system with scientists applying their

         specialities in many different ways to deliver goods to the farmers of the developing world. It does not

         comprise, and probably never has, just a group of plant breeders producing new varieties. The system runs the

         full gamut from germplasm conservation, both in the wild and in international collection, through understanding

         the effects of genes on plant physiology and pathology, plant breeding, agronomy, natural resource

         management, delivery of technologies to farmers, and up to agricultural systems design and development—from

         gene to sociology.

         Some years ago it became evident that the climate in which the IARCs were operating was about to change (see

         Jackson et al., 1990) and studies were made to evaluate the possible effects that this would have on international

         agriculture. Some of these were made within the CG system in collaboration between CIAT, ILRI, IPGRI and

         others (see, for example, Jarvis et al., 2001; Jones and Beebe, 2001; Jones and Thornton, 2002; 2003). In

         essence, these investigations were exploratory and illustrative; they were made to see if climate change would

         affect the way we had to operate if it occurred, and to estimate the magnitude of the effects.

         Although global agriculture contributes to climate change through methane emissions from livestock and rice

         paddies, CO2 emissions from deforestation and various oxides of nitrogen from fertilizer use, international

         agricultural research will always be a minor player in terms of mitigation. We need to come to terms with

         climate change in many aspect of what we do, not so that we can control it but so that we continue to work

         efficiently and effectively.

         The reality of climate change is evident (IPCC, 2007a) and the likely effects are broadly predicted, although still

         uncertain with regard to the nature, rate and extent to which such changes will occur (IPCC, 2007b). Also quite

         probable is the economic feasibility of controlling emissions sufficiently to limit the extent of change to

         whatever level world politicians can agree upon (HM Treasury, 2007). Whether they do this is, of course, pure

         speculation. These findings actually make our job much harder. It used to be relatively easy for climate

SAT eJournal | ejournal.icrisat.org                                                           December 2007 | Volume 4 | Issue 1
An Open Access Journal published by ICRISAT

         modellers to choose a scenario by merely making an intelligent guess at world demographics and economic

         progress; however, we now have to contend with the vagaries of the political process. It is therefore imperative

         that we have all the tools, improved and ready for use, to assist us in the task.

         The paper now analyses a cross-section of the activities of the CG system to determine the possible effects of

         climate change on their outcomes and try to determine how far into the future we need to look to make sure that

         the research is targeted to the correct environment. It should be borne in mind that Global Circulation Models

         (GCMs) are not precise and are run under a range of possible scenarios for greenhouse gas emissions. The

         results vary between models and scenarios. The IPCC reports use up to 21 models to obtain a reasonable

         consensus picture. Most of these examples use the HADCM3 model and only look at one or two scenarios. The

         GCM output has been downscaled to take into account the local variation in soils and topography important in

         agricultural applications. This introduces further uncertainty. The results are presented to gain an impression of

         the magnitude and spatial variation of possible effects and should not be taken as accurate predictions.

         Project types and timelines

         We will now try to follow the process from gene to field. We look first at the basic building blocks we have

         available in the section entitled Germplasm conservation. We then look at the breeding process and varietal

         deployment to determine the lead-time for agricultural research. This involves three phases: Finding a useful

         trait, Creating a new variety and Lifetime of a variety. We then look at how to cope with this lead-time in a

         changing environment and, lastly, how to serve the farmers in the field.

         Germplasm conservation

         Genes from wild crop relatives are highly important to modern agriculture. The founder effect resulted when

         crops were first domesticated. Only certain individuals with desirable characteristics were selected; this has

         meant that many potentially useful genes were left behind in wild populations when domestication occurred.

         The key point for international agricultural research and development is that we are going to lose genes in the

         wild. These genes might be useful in combating climate change but that is not the point. We rely on those genes

SAT eJournal | ejournal.icrisat.org                                                           December 2007 | Volume 4 | Issue 1
An Open Access Journal published by ICRISAT

         for our everyday work regardless of climate change. Climate change is therefore impinging on how we do our

         job and has to be taken into account.

         For example, post harvest losses due to Bruchid bean weevils in Phaseolus vulgaris L. can be excessive. Jones

         and Beebe (2001) calculated that, given mid 1990s production and prices, these could cost in excess of US$1.2

         billion. Schoonhoven et al. (1983) found a novel seed protein in wild beans; it was named arcelin after the

         Mexican town of Arcelia, where it was found. The protein was further characterised by Romero et al. (1986),

         Osborn et al. (1986) and Osborn et al. (1988). Arcelin now provides an effective mechanism for Bruchid

         resistance and can be readily incorporated in new bean varieties.

         Many more useful genes may still exist, unidentified, in the wild populations but what will be their fate under

         climate change? Jones and Beebe (2001) used FloraMap® (Jones and Gladkov, 2001) to investigate the

         distribution of wild bean in Central America to the year 2055. They downscaled results from HADCM3 to a 1-

         km climate grid and created potential distributions for the present and for the year 2055. A set of 46 germplasm

         accessions from the region calibrated the model and Figure 1 shows the present distribution. Figure 2 shows

         how habitat will disappear over the next 50 years and Table 1 summarises the results. Wild bean habitat will

         essentially disappear in five of the countries.

         The peanut, Arachis hypogaea L. was domesticated more than 7000 years ago in South America (Dillehay et al.,

         2007) but until recently we did not know from which wild progenitors it came. Seijo et al. (2004) recently

         clarified that as being A. duranensis and A. ipaensis, both found in the border region between Bolivia, Paraguay

         and Argentina. A. hypogaea is highly susceptible to root knot nematode (Starr et al., 2002). Simpson and Starr

         (2001), in their registration article, described a new cultivar, COAN, that is resistant to this costly pest. COAN

         was derived from a complex interspecific amphiploid hybrid formed by first crossing A. cardenasii and A.

         diogoi and then crossing the F1 hybrid with A. batizicoi. This was then backcrossed using the commercial

         variety, Florunner, as the recurrent parent. The modern peanut therefore is not only the result of ancient crossing

         of wild parents but also can benefit immensely from the genes in the wild relatives found today.

         Jarvis et al. (2001) used the FloraMap® package to map the distribution of wild Arachis species in southern

         America and within the section Arachis (containing 17 of the closest relatives to the cultivated peanut) plotted

SAT eJournal | ejournal.icrisat.org                                                           December 2007 | Volume 4 | Issue 1
An Open Access Journal published by ICRISAT

         the species richness shown in Figure 3. They then used the climate grid derived from the HADCM3 model for

         scenario A1a to determine the species richness likely to be found in the year 2055 (Figure 4). The difference

         was marked. Arachis species can only spread naturally by up to 1 metre per year due to their hypogeal seed-

         bearing mechanism, so they cannot move rapidly in the face of climate change as can many other plants. The

         study found that all of the wild species used in the production of the cultivar COAN will be extinct under this

         scenario in 50 years (Table 2). Some 12 of the 17 species in this section will be extinct and a further four species

         will be dangerously threatened by that time.

         These are merely two examples that have come to our attention. Jarvis et al. (In press) made a similar study for

         African Vigna (including cowpea) and wild potato, finding extinction levels of 16% to 22% of species by 2055.

         Studies looking at extinction rates of other wild species find rates as high as 43% of all species (Malcolm,

         2006). In the case of agricultural biodiversity, the international germplasm banks may give us some hope that all

         may not be lost but is this enough?

         Lawrence et al. (1995) argued that relatively small samples are sufficient to capture a wide range of variation.

         They estimated that “A sample of about 172 plants, drawn at random from a population of a target species, is of

         sufficient size to conserve at a very high probability, all or very nearly all of the polymorphic genes that are

         segregating in the population, provided that their frequency is not less than 0.05” (Lawrence et al., 1995:1).

         Although wild relatives of some of the major crops may be conserved in sufficient numbers to fulfil these

         criteria, the reality is that many important wild gene pools are grossly underconserved in gene banks. The

         Arachis species used in this study were taken from the comprehensive catalogue of world Arachis germplasm

         compiled by Stalker et al. (2000). A. batizicoi had 23 accession points, five of which were herbarium specimens

         and only 12 were noted as viable germplasm held at Texas A&M University, North Carolina University and

         ICRISAT. A. cardenasii was represented by 28 points and A. digoi by 19 points, with 17 A. cardenasii and 5 A.

         digoi potentially viable germplasm accessions between the three institutions. In the case of P. vulgaris there are

         657 accessions in the CIAT germplasm bank ranging from Argentina to Mexico but only 27 of these were

         collected in Central America, which is often regarded as a possible centre of diversity. This is also cause for


SAT eJournal | ejournal.icrisat.org                                                            December 2007 | Volume 4 | Issue 1
An Open Access Journal published by ICRISAT

         In situ conservation does not look to be a good option in either case. If we are not to lose one of the essential

         components of our future task, it would appear that a stocktaking of germplasm collections and climate change

         analysis of the species viability in the field may have to be followed up by some judicious collecting work.

         Finding a new trait

         Climate change has little to do with this level of research. However, the timeframe is important; this phase is

         often referred to now as ‘pre-breeding’. The climate change implication is the lead-time that is involved. If we

         are going to see very different situations in the future we have to think of the lead-time that we have.

         When, in the late 1970s, CIAT decided to look for drought tolerance in Phaseolus vulgaris L. most bean

         scientists dismissed this as a wild hope. The common bean was well known to be drought susceptible; it was a

         weak plant that closed its stomata at the smallest sign of water stress. CIAT, nevertheless, started a screening of

         the germplasm bank and soon all the promising materials from the breeding programmes went through drought

         screening trials. This daring act of faith would not be permitted in the present day environment of short-term

         return to project funding.

         After about 8 years of screening, CIAT scientists eventually identified something significant, enough to warrant

         a serious physiological investigation. There were two candidate materials, BAT 477 and A21, with a similar

         growth habit but from different progenitors. Sponchiado et al. (1988) showed conclusively that the drought

         tolerance was due to deep rooting. This was an advance but only in situations where beans could put down deep


         The next advance had to wait for some years more. In the mid-1990s, Rao (2001) showed that another

         mechanism was involved. He noticed that in most drought-stressed bean plants the leaves became engorged with

         photosynthates, noticeably thickening. However, some plants retained the capability to translocate the

         photosynthates even under stress conditions. The scene was then set to put both of these traits into commercially

         acceptable varieties. Beebe et al. (2007) reported on the success of this effort and the new lines are now being

         distributed to breeding programmes throughout Central America.

SAT eJournal | ejournal.icrisat.org                                                            December 2007 | Volume 4 | Issue 1
An Open Access Journal published by ICRISAT

         Note that this research took over 20 years to complete. It probably only cost a few million dollars but throwing

         in a few million more would not necessarily speed it up by very much.

         Creating a new variety

         Often, creating a new variety is much quicker than finding the traits necessary to go into it but it may still take

         an appreciable amount of time. Two well-documented examples are presented. Hargrove and Coffman (2006)

         outline the development of the green revolution rice variety, IR8, and mention the development of 8156, the

         dwarf wheat line bred by Norman Borlaug. In both cases, the dwarf lines used as parents already were available.

         Hargrove and Coffman (2006) point out that the dwarf wheat used by Norman Borlaug had been found in Japan

         under General Macarthur’s Administration and had been sent to Washington in 1946. So the trait was there, it

         was just the bureaucracy of the post war period that caused delay. Orville Vogel sent the seed of Norin 10 to

         Borlaug after the Rockefeller Centre, later to become CIMMYT, was founded in Mexico. By the late 1960s, the

         cross 8156 was established as a leading variety in a number of countries—still almost 20 years from conception

         to fruition but fast by many standards.

         The miracle rice, IR8, was bred in a similar manner. FAO initiated the idea in 1949; it was known that short

         varieties existed. It was not until 1960, when Jennings and Wortman toured Asia and found Taichung Native 1

         (TN1), that the idea really prospered. Used in IRRI in 1962 with other dwarf rice lines in the crossing

         programme, Jennings made 38 crucial crosses. The eighth became IR8 in 1966 and was scheduled for release in

         the Philippines; fast work for a breeding programme but still 17 years from inception of the idea.

         Marker assisted selection, quantitative trait linkage methods and much smarter manipulation of DNA, even

         without direct genetic engineering, is speeding up the breeding process but, as we have seen above, the pre-

         breeding phase can be considerably longer than the time from crossing to fixed line.

         A variety may come into existence once a breeder has fixed the traits that are necessary but there is a long

         process of seed multiplication, acceptance and registration by National Agricultural Research programmes

         (NARs). This can take many more years. In modern times, the CGIAR centres do not release finished varieties

SAT eJournal | ejournal.icrisat.org                                                            December 2007 | Volume 4 | Issue 1
An Open Access Journal published by ICRISAT

         but send on promising materials to the NARs, possibly to be crossed again with local varieties to ensure local

         adaptation and acceptance.

         Lifetime of a new variety

         In some circumstances, varieties will last for many decades. The bean variety Jamapa was released in Mexico in

         1958, the result of selection within local landraces, and was a standard for nearly 40 years (Voysest, 2000).

         Others have shorter life spans, mainly due to inherent disease susceptibilities. A red-seeded climbing bean from

         Mexico, G2333, was introduced into Rwanda in the mid-1980s but within a decade had succumbed to Fusarium

         wilt. However, it would be a reasonable bet that a life span of 20 years was the norm.

         If we allow 20 years for conceptualisation and development (recognising that in the case of climate change,

         conceptualisation is still part of the challenge!), 6 years for dispersal and a 20-year life span in the field, then

         when we plan new breeding projects we should be looking at environments that will be in place in 46 years’

         time, not now. Thus the common concept that plant breeding can produce a new variety in a few years by

         making some crosses and fixing the traits is true but not really applicable to the problem. That time is only a

         small fraction of the total lifetime of a product from conception to use and through its lifetime.

         The same applies for new technology generated in agronomy, pest and disease control and other disciplines. If

         we look at the whole lifetime of the technology from conception through its useful period we are looking quite

         far into the future, and of course we all hope that any useful technology will have a long useful life.

         Planning to produce results still viable in the future

         In 1986, in Rome, a group of agricultural scientists came together to discuss the state of agro-ecological

         targeting; the results are reported by Bunting (1987). At this meeting there was no mention of climate change. It

         was thought that the environment was classifiable for the foreseeable future. However, the tools for the analysis

         were all laid out in this document, all we have to do now is apply them. The important point is that test sites

         reflect the environment for which the technology is intended in the future.

SAT eJournal | ejournal.icrisat.org                                                              December 2007 | Volume 4 | Issue 1
An Open Access Journal published by ICRISAT

         Where is there a homologue of my field anywhere else in the world? This is a question that not many farmers

         ask but many advisers would dearly like to know. To answer this, CIAT has developed the software package

         Homologue™ (Jones et al., 2005). This works with a generic FloraMap algorithm that can extrapolate

         homologous climate and soils from a single point. This allows us to map the area of environments with varying

         degrees of similarity to a test site. Hyman et al. (2007) performed a target area analysis for the Harvest Plus

         Challenge Programme. They constructed a target area index using human population, crop area and indices of

         micronutrient deficiencies, and mapped it to delineate critical areas for research in various crops. A random

         sample of pixels from this mapping was taken to create the target area homologue zones and this was mapped

         against the present testing sites to see if they fell within environments homologous with the critical target areas.

         Figure 5 shows the section for maize in Africa.

         For this paper, we modified Homologue to use the HADCM3 under scenario A1a for the year 2047 and

         calculated the homologues of the trial sites. Figure 6 shows the homologues of the climates of the present trial

         sites in the year 2047. It is evident that varieties selected now will have a very different adaptation range in 40

         years’ time. This, however, only tells us that these selection sites will not do for present maize areas in the far


         What about choosing selection sites that will still be representative of the target area into the future? This needs

         a different analysis but can still be done using Homologue.

         Figure 7 shows the homologues of the Harvest Plus Challenge Programme sites for maize environments in the

         year 2047 projected on the present day. One of the sites is already in a projected zone and two others are not far

         away. The message should not surprise anyone; to cope with more extreme environments in the future you need

         to test in more extreme environments now.

         This example is extreme in that we have imposed the full 40-year lead-time. Presenting the results at a

         continental level, as we do, does not show the full utility of using the technique to choose relevant selection

         sites, which in theory can be used at high spatial precision and for any time-frame desired.

SAT eJournal | ejournal.icrisat.org                                                             December 2007 | Volume 4 | Issue 1
An Open Access Journal published by ICRISAT

         Unfortunately, not all future environments exist at present. The further we look into the future, the less present

         environments we have. However, for agricultural testing purposes, on a time-scale of 20 to 40 years we will be

         able to cover most of the growing regions of the tropics with present day homologues. Whether crops will grow

         in those homologues is debatable. In the case of completely new climates, it is doubtful that they will overlap

         with present day agriculture, although where they do so the plant breeders and agronomists will have a real

         challenge that is far and above that envisaged here.

         Farming systems and agricultural development

         Jones and Thornton (2003) made an early warning study of the effect of climate change. They used the DSSAT

         maize model to predict maize yield in Africa and Latin America for the year 2055. The study took 30-year runs

         for each 10-minute pixel and calculated the yield of appropriate generic cultivars for all possible maize-growing

         areas in the two continents. The overall result was that, given present day technology, there would be a 10%

         reduction in maize harvest.

         However, this result overshadowed a much harsher reality. Figure 8 shows a particular situation in the Horn of

         Africa; Figure 9 shows the expected yield change in the Great Lakes region. Unfortunately, in the processing of

         these images the colour for lakes has matched the colour for -2000 kg yield but this will be apparent on


         The 10% reduction may be the global deficit but look at what it is going to produce at a local scale. Farmers

         within 20 km of one another are going to go from bust to boom. There is going to be no grand recipe for a

         country or even a region to recover from climate change. It will be farm-to-farm depending on conditions, as it

         always was! That is not to say that regional predictions are not worth it; of course they are. How else do you get

         the message of the diversity down to the field?

         The absolute values of study like this are very questionable. They depend on the reliability of the GCM, the

         precision of the downscaling, the variation in local soil and topographical factors and also differences in local

         farming practices. One conclusion that we can draw from them is that the spatial variation will be paramount.

SAT eJournal | ejournal.icrisat.org                                                           December 2007 | Volume 4 | Issue 1
An Open Access Journal published by ICRISAT

         Farmers see the climate in terms of good years and bad years. In the past, this has been a random sequence of

         feast and famine. For many, the slow change in the average environment is still imperceptible, masked by the

         annual variance. This is now changing, it is getting faster but it will still be seen predominantly as the changing

         frequency of events. Figure 10 is a stylised representation of a random series of arbitrary events slowly rising

         over time. The trend would not be perceived by an individual living through the series as the mean trend but as

         the frequency of events above or below a threshold.

         How do we cope with farmer perception of climate change?

         Getting messages across to the farmer has changed in the last half century. In early years in the CG system, the

         emphasis was on the science of the crops, and the extension of results to farmers was the realm of the national

         programme scientists and their extension agents. The CGIAR invested hundreds of millions of dollars in

         training national institutions and their staff to this end. With the rise of capitalism and globalisation came the

         demise of many government-funded national agricultural research entities, hence CG Centre staff can now be

         found in the field working directly with farmers; a complete reversal but a good one. In the early years, it was

         almost impossible to get information from the field to the agro-ecological analysis sector. Today it is reckoned

         that contact with the farmer is absolutely necessary for the research process. To a large extent, the scientists in

         this process have been social scientists and agronomists trying to solve the problems of rural development with

         the tools at hand—the farmers’ knowledge of the environment and the knowledge, technologies and seeds

         coming from the CG centres.

         This has built up a wealth of information on how farmers cope with adversity. Now we have to put another level

         of analysis into practice because the situation is starting to change. How do farmers cope with a crop failure

         every 5 years? How will they cope when it is every 3 years? Why do they grow the crops they grow now? Why

         not others? Do they have lands that they graze; could they manage under extended grazing? Would they grow

         different crops if required? How will they be able to manage when cropping becomes impossible? How will

         they capitalize on any benefits that climate change may bring?

SAT eJournal | ejournal.icrisat.org                                                            December 2007 | Volume 4 | Issue 1
An Open Access Journal published by ICRISAT

         The questions for agriculturalists in the field are how did people cope with adversity in the past and how can

         that be applied to the future? This presupposes that someone will be telling them what it will be like in the

         future. We have some possible solutions that are extensions of current participatory research techniques.

         Farmers and/or extension agents could attend farm workshops where crop and integrated farm models are used

         to illustrate what might happen in the future. These types of meetings are already used in some places under

         present conditions.

         Where a future homologue of an area exists within a reasonable distance of a participatory research area, a

         complementary group might be set up in the homologue. Farm visits or some other form of information flow

         could be set up from which the present area might benefit. This has the problem that it is not reciprocal, the

         group of farmers in the future homologue area will not benefit from a reversal of the information flow and

         would need a further future homologue to gain any benefit. Thus a cascade of information might be set up from

         future to present climates. Unfortunately, there will exist a final area for which there is no existing future



         Maintenance and improvement of agricultural production, even in the absence of climate change, needs:

                  1.     Reliable sources of genetic variation,

                  2.     Information to target investment in breeding programmes,

                  3.     Information to target the results of breeding programmes,

                  4.     Ways of delivering new technologies to the relevant farmers, and

                  5.     Methods of assisting farmers to adopt appropriate technologies.

         All of these stages are affected by climate change but with careful planning these effects can be minimised.

         Careful planning means access to information. Every scientist, when s/he initiates a project must think—“What

         is my target area? What will it be like at the termination of my project? What will it be like for the lifetime of

         the results?”

SAT eJournal | ejournal.icrisat.org                                                             December 2007 | Volume 4 | Issue 1
An Open Access Journal published by ICRISAT

         Thus, proofing the system against climate change does not mean a lot more research specifically on climate

         change added on to what we do now and kept separate, as would happen in a Challenge Programme. It is a basic

         necessity to all the research. What is needed that we do not have now is an efficient system for processing,

         analysing and distributing specific information, and modelling tools to help everyone plan for future change.

         This might look a little like the Consortium for Spatial Information but fully institutionalised and budgeted with

         trained people in each centre exchanging information and informing the centre scientists.

         The take-home line is that international agricultural research will be doing much the same under climate change

         scenarios but the undertaking has just become larger and harder to plan.

SAT eJournal | ejournal.icrisat.org                                                          December 2007 | Volume 4 | Issue 1
An Open Access Journal published by ICRISAT


         Beebe, S.E., Rao, I.M., Cajiao, C., Grajales, M. 2007. Selection for drought resistance in common bean also

         improves yield in phosphorus limited and favorable environments. Crop Sci. Submitted.

         Bunting, A.H. 1987. Agricultural environments: Characterisation, classification and mapping. CAB International,


         Dillehay, T.D., Rossen, J., Andres, T.C., Williams, D.E. 2007. Preceramic adoption of peanut, squash, and

         cotton in northern Peru. Science 316: 1890.

         Hargrove, T., Coffmann, W.R. 2006. Breeding history. International Rice Research Institute (IRRI) Rice Today


         HM Treasury. 2007. Stern review on the economics of climate change. http://www.hm-


         Hyman, G., Hodson, D., Jones, P., Fujisaka, S., Wood, S. 2007. Spatial analysis for the HarvestPlus global

         biofortification program. Internal report of the HarvestPlus Initiative. International Center for Tropical

         Agriculture (CIAT), Cali, Colombia. In preparation.

         IPCC (Intergovernmental Panel on Climate Change). 2007a. The physical basis of climate change. IPCC

         Working Group I. http://ipcc-wg1.ucar.edu/wg1/wg1-report.html

         IPCC (Intergovernmental Panel on Climate Change). 2007b. Climate change impacts, adaptation and

         vulnerability. IPCC Working Group II. http://www.ipcc-wg2.org/

         Jackson, M., Ford-Lloyd, B., Parry, M. (eds.). 1990. Climatic Change and Plant Genetic Resources. Belhaven

         Press, London and New York. 190 pp.

SAT eJournal | ejournal.icrisat.org                                                           December 2007 | Volume 4 | Issue 1
An Open Access Journal published by ICRISAT

         Jarvis, A., Jones, P., Mottram, G., Williams, D., Guarino, L., Ferguson, M. 2001. Predicting the impact of

         climate change on the distribution of plant genetic resources in wild peanuts. Poster presented at the Conference

         on Challenges of a Changing Earth – Global Change Open Science Conference, Amsterdam, 10-13 July 2001.

         Jarvis, A., Lane, A., Hijmans, R.H. In press. Impacts of climate change on crop wild relatives. Agric. Ecosyst.


         Jones, P.G., Beebe, S. 2001. Predicting the impact of climate change on the distribution of plant genetic

         resources in wild common bean (Phaseolus vulgaris L.) in Central America. Paper presented at the III

         International Conference on Geospatial Information in Agriculture and Forestry, Denver, Colorado.

         Jones, P.G., Gladkov, A. 2001. FloraMap Version 1.01. A computer tool for predicting the distribution of plants

         and other organisms in the wild. Centro Internacional de Agricultura Tropical (CIAT) CD-ROM series, Cali,

         Colombia..CD-ROM + Guide (104 p.)

         Jones, P.G., Thornton, P.K. 2002. Spatial modeling of risk in natural resource management. Conserv. Ecol. 5(2):

         27. [online] URL: http://www.consecol.org/vol5/iss2/art27

         Jones, P.G., Thornton, P.K. 2003. The potential impacts of climate change on maize production in Africa and

         Latin America in 2055. Global Environ. Change 13:51-59.

         Jones, P.G., Díaz, W., Cock, J.H. 2005. Homologue™. A computer system for identifying similar environments

         throughout the tropical world. Centro Internacional de Agricultura Tropical (CIAT) CD-ROM series. CIAT

         publication no. 342. CIAT, Cali, CO. Manual 100 p + CD-ROM.

         Lawrence, M.J., Marshall, D.F., Davies, P. 1995. Genetics of genetic conservation. 1. Sample-size when

         collecting germplasm. Euphytica 84: 89-99.

         Malcolm, J.R. 2006. Global warming and extinctions of endemic species from biodiversity hotspots. Conserv.

         Biol. 20 (2), 538-548.

SAT eJournal | ejournal.icrisat.org                                                         December 2007 | Volume 4 | Issue 1
An Open Access Journal published by ICRISAT

         Osborn, T.C., Blake, T., Gepts, P., Bliss, F.A. 1986. Bean arcelin. 2. Genetic variation, inheritance and linkage

         relationships of a novel seed protein of Phaseolus vulgaris L. Theor. Appl. Genet. 71:847-855.

         Osborn, T.C., Alexander, D.C., Sun, S.S.M., Cardona, C., Bliss, F.A. 1988. Insecticidal activity and lectin

         homology of arcelin seed protein. Science 240:207-210.

         Rao, I.M. 2001. Role of physiology in improving crop adaptation to abiotic stresses in the tropics: The case of

         common bean and tropical forages. In M. Pessarakli (ed.). Handbook of Plant and Crop Physiology. Marcel Dekker,

         Inc., New York, USA. p. 583-613.

         Romero, A.J., Yandell, B.S., Bliss, F.A. 1986. Bean arcelin. 1. Inheritance of a novel seed protein of Phaseolus

         vulgaris L. and its effect on seed composition. Theor. Appl. Genet. 72:123-128.

         Schoonhoven, A.V., Cardona, C., Valor, J. 1983. Resistance to the bean weevil and the Mexican bean weevil

         (Coleoptera: Bruchidae) in non-cultivated common bean accessions. J. Econ. Entomol. 76:1255-1259.

         Seijo, J.G., Lavia, G.I., Fernández, A., Krapovickas, A., Ducasse, D., Moscone, E.A. 2004. Physical mapping of

         the 5S and 18S–25S rRNA genes by FISH as evidence that Arachis duranensis and A. ipaensis are the wild

         diploid progenitors of A. hypogaea (Leguminosae), Am. J. Bot. 91:1294-1303.

         Simpson, C.E., Starr, J.L. 2001. Registration of ‘COAN’ peanut. Crop Sci. 41:918.

         Sponchiado, B.N., White, J.W., Castillo, J.A., Jones, P.G. 1988. Root growth of four common bean cultivars in

         relation to drought tolerance in environments with contrasting soil types. Expl. Agric. 25:249-257.

         Stalker, H.T., Ferguson, M.E., Valls, J.F.M., Pittman, R.N., Simpson, C.E., Bramel P. 2000. Catalog of Arachis

         Germplasm Collections. Available online at


SAT eJournal | ejournal.icrisat.org                                                            December 2007 | Volume 4 | Issue 1
An Open Access Journal published by ICRISAT

         Starr, J.L., Morgan, E.R., Simpson, C.E. 2002. Management of the peanut root-knot nematode, Meloidogyne

         arenaria, with host resistance. Plant Health Progress. Available online at


         Voysest, O. 2000. Mejoramiento genético del frijol (Phaseolus vulgaris L.): legado de variedades de América

         Latina 1930-1999. Centro Internacional de Agricultura Tropical, Cali, Colombia.

SAT eJournal | ejournal.icrisat.org                                                        December 2007 | Volume 4 | Issue 1
An Open Access Journal published by ICRISAT

         Figure 1. Present day distribution of wild Phaseolus vulgaris L. environments in Central America



                                      El Salvador

                                                              Cost Rica

SAT eJournal | ejournal.icrisat.org                                                  December 2007 | Volume 4 | Issue 1
An Open Access Journal published by ICRISAT

         Figure 2. Predicted distribution of wild P. vulgaris environments in the year 2055

SAT eJournal | ejournal.icrisat.org                                                    December 2007 | Volume 4 | Issue 1
An Open Access Journal published by ICRISAT

         Figure 3. Species richness for wild Arachis section Arachis in southern America in the year 2000

SAT eJournal | ejournal.icrisat.org                                                   December 2007 | Volume 4 | Issue 1
An Open Access Journal published by ICRISAT

         Figure 4. Species richness for wild Arachis section Arachis in southern America in the year 2055

SAT eJournal | ejournal.icrisat.org                                                    December 2007 | Volume 4 | Issue 1
An Open Access Journal published by ICRISAT

         Figure 5. Maize testing sites in Africa (black dots). The yellow–red graded areas show the present day

         extent of the Harvest Plus maize target area homologues

SAT eJournal | ejournal.icrisat.org                                                    December 2007 | Volume 4 | Issue 1
An Open Access Journal published by ICRISAT

         Figure 6. Maize trial sites in Africa with the areas of adaptation to the present environment at the sites in

         the year 2047

SAT eJournal | ejournal.icrisat.org                                                      December 2007 | Volume 4 | Issue 1
An Open Access Journal published by ICRISAT

         Figure 7. Maize trial sites in Africa with the present day homologues for environments at the sites in the

         year 2047

SAT eJournal | ejournal.icrisat.org                                                     December 2007 | Volume 4 | Issue 1
An Open Access Journal published by ICRISAT

         Figure 8. Modelled maize yields in the Horn of Africa and their changes to 2055

                  Yield kg ha-1                           Change kg-1

                  500                                     <-2000

                 1000                                       -1000

                 1500                                        -250
                 2500                                      +1000

               >2500                                       >2000

SAT eJournal | ejournal.icrisat.org                                                   December 2007 | Volume 4 | Issue 1
An Open Access Journal published by ICRISAT

         Figure 9. Change of maize yields in the Great Lakes region from present day to 2055

                                                         Increase in highlands

                                                                                          Change kg-1

                           Sharp decrease in lowlands                                      <-2000





SAT eJournal | ejournal.icrisat.org                                                  December 2007 | Volume 4 | Issue 1
An Open Access Journal published by ICRISAT

         Figure 10. The slow perception of a change that is real but masked by random events. (Adapted from a

         presentation by Barry Smit of the University of Guelph)

                               Random Series gently rising
                                         Number of high events
                                     5 increases
           Value (arbitrary units)


                                     3                                             Comfort

                                                           Number of low events decreases
                                         0     5      10     15      20      25   30        35        40
                                                              Time (years)

SAT eJournal | ejournal.icrisat.org                                                       December 2007 | Volume 4 | Issue 1
An Open Access Journal published by ICRISAT

         Table 1. Bean habitat (km2)a in 2000 and 2055 in different land typesb.

                                            2000                                2055               % reduction
                                 Appropriate Inappropriate        Appropriate     Inappropriate
             Mexico                61542          84708             33696              92430               45
             Guatemala            200538         149760            140634             130572               30
             El Salvador           17550          42354               2340             14040               87
             Honduras              92664          49608               1170              4212               99
             Nicaragua            270738         106236             20124               10296              93
             Costa Rica            22230          59904             86580               22698              61
             Panama                  4212             0                  0                  0             100

           a.     From pixels with probability greater than 0.5.
           b.     Appropriate land types include all cropland, pasture and wild shrub land. Inappropriate land types
                  include urban areas, water bodies and closed forest areas.

SAT eJournal | ejournal.icrisat.org                                                         December 2007 | Volume 4 | Issue 1
An Open Access Journal published by ICRISAT

         Table 2. Expected change in the distribution of wild Arachis species in South America to the year 2055

         using data from Hadley III under scenario A1. Species used in the production of the cultivar COAN are

         shown in colour

                                           Change in area of Predicted state in
                                            distribution (%)       2055

            batizocoi                               -100                Extinct
            cardenasii                              -100                Extinct
            correntina                              -100                Extinct
            decora                                  -100                Extinct
            diogoi                                  -100                Extinct
            duranensis                               -91              Threatened
            glandulifera                             -17                 Stable
            helodes                                 -100                Extinct
            hoehnii                                 -100                Extinct
            kempff-mercadoi                          -69            Near-Threatened
            kuhlmannii                              -100                Extinct
            magna                                   -100                Extinct
            microsperma                             -100                Extinct
            palustris                               -100                Extinct
            praecox                                 -100                Extinct
            stenosperma                              -86              Threatened
            villosa                                  -51            Near-Threatened

SAT eJournal | ejournal.icrisat.org                                                    December 2007 | Volume 4 | Issue 1

To top