Development of appropriate Bt gene constructs and source inbred lines

One of the tenets of the use of Bt maize by CIMMYT is to provide transformed plants that carry only the gene of interest, i.e., plants that do not carry selectable markers, such as herbicide or antibiotic resistance. CIMMYT acquired Bt maize events from the private and public sectors, and has also synthesized other Bt genes with partners. Various Bt cry genes (cry1Ab, cry1Ac, cry1B, cry1E, and cry B-1Ab) have been used to transform the CML216 x CML72 hybrid maize with maize ubiquitin and rice actin promoters. Backcrosses were made to CML216 and the lines (T0 - T4) have shown high levels of resistance to stem borers.

Recently, development continued on second generation events that carry only the gene of interest. These “clean genes” do not carry the selectable Basta™ herbicide resistance (the bar gene) marker, thus addressing some of the concerns about this technology. The clean events are developed by using isolated Bt and bar gene sequences for transformation. The Bt and bar genes are co-transformed, and inserted separately into the maize genome, thus allowing separation of the two genes in the final product. By using this process, insect resistant, herbicide susceptible maize varieties are produced. Molecular characterization of all first generation Bt maize events has now been accomplished.

Efforts were made to identify Bt genes that are effective against each of the target stem borer species, and this was done by conducting insect bioassays of maize leaves (containing the different cry genes). Given the early state of biosafety in Kenya and the lack of proper infrastructure in the Kenya Agricultural Research Institute (KARI) to handle transgenic maize (in the greenhouse and the field), it was decided that the simplest procedure would be to import Bt maize leaves (that were grown in CIMMYT’s biosafety greenhouses in Mexico) into Kenya and perform leaf bioassays in the KARI-NARL Biotechnology Laboratories.

To introduce Bt maize leaves into Kenya, permission was sought from and granted by the National Biosafety Committee (NBC), and a permit issued by the Kenya Plant Health Inspectorate Service (KEPHIS), in accordance with Kenya’s biosafety regulations. Bioassays were conducted to evaluate the effect of various cry genes against each of the five stem borer species: Chilo partellus (Swinhoe), Busseola fusca Fuller, Sesamia calamistis Hampson, Eldana saccharina Walker, and Chilo orichalcociliellus Strand. Susceptibility of the different cry genes to the stem borers was assessed by determining the respective leaf areas consumed and the number of stem borer larvae that died after three days of leaf-feeding.

The Cry1Ab protein was the most effective against all the five stem borer species, while Cry1E was not effective against any of the stem borers. Chilo partellus was affected by all the Cry proteins, except Cry1E, while E.saccharina was the least affected by any of the Cry proteins (Figures 1 and 2). Chilo orichalcocilielus was most affected by Cry1Ab and Cry1B proteins. Sesamia calamistis was affected by Cry1Ab and Cry1Ab-1B proteins. None of the evaluated Cry proteins was effective in controlling B. fusca, hence the need to search for alternative effective cry genes for this stem borer. From results of these bioassays, a prospective effective control was identified for C. partellus, the most destructive and most widely distributed stem borer in Kenya.

The results of this laboratory study will later be verified in a biosafety greenhouse and in open quarantine in the field. Ground will soon be broken for a biosafety greenhouse at the KARI Biotechnology Center. Also, an open quarantine field site has been established at the KARI research site at Kiboko. No maize will be grown within 200 m of the fence of the open quarantine field. Maize within the quarantine field will be de-tasseled, to prevent inadvertent gene flow to other maize crops. The greenhouse studies will be followed by studies in the open quarantine, and it is expected that the information generated will allow better targeting of the development of Kenyan maize varieties with the appropriate combinations of genes for resistance to these stem borer species.

Development of locally adapted non transgenic and transgenic insect resistant maize germplasm

Host plant resistance is an approach to stem borer management, by which a plant itself is able to resist infestation and damage by pests. This control strategy is availed to farmers in the seed, a fact that ensures that the technology is inexpensive, safe, and that the farmers need not purchase more inputs to control stem borers. Use of stem borer resistant maize increases efficiency of farming by reducing or eliminating the expense of insecticides and reducing yield losses from stem borer damage. For resource poor small-scale farmers in developing countries, therefore, host plant resistance packaged into improved varieties will offer a practical and economic means of minimizing stem borer losses.

The IRMA project aims to develop stem borer resistant maize varieties by using two approaches: (1) search for sources of resistance and develop source germplasm for insect resistance, and, (2) search for insect resistance among elite germplasm, into which sources of Bt genes will be backcrossed, when they are available. The development of sources of stem borer resistant germplasm is based on utilizing genes and sources of resistance that already exist in the maize plant.

Along these lines, the IRMA project has evaluated 216 genotypes from CIMMYT and KARI, 42 multiple borer resistance (MBR) S4 lines and 500 inbred lines from CIMMYT Mexico. In the search for stem borer resistant elite germplasm, 330 maize OPVs and hybrids have been evaluated in different maize growing ecologies in Kenya. This germplasm has been evaluated for resistance to C. partellus and B. fusca stem borers through artificial infestation. The germplasm is also being screened for tolerance to local stresses such as drought, low-nitrogen, resistance to maize streak virus, Turcicum blight, leaf rust, and weevils in storage to ensure that insect resistance will be in good adapted genetic backgrounds.

Good and stem borer resistant inbred lines are being crossed to heterotic testers like CML78 and CML444, while combining ability studies are being done to identify lines with good specific and general combining abilities. Suitable hybrids will be made from lines with good specific combining abilities, while synthetics will be developed from promising lines with good general combining ability. Inbred lines are being recycled in order to develop elite locally-adapted germplasm. The project is identifying excellent sources, especially those carrying resistance to more than one stem borer species.

Investigations on potential impacts of Bt maize on non target arthropods

Knowledge of the environmental impacts of Bt gene-based stem borer resistance technology on nontarget organisms in the major maize cropping systems is essential for the safe deployment of Bt maize in Kenya. A prerequisite to studies on the nontarget effects of Bt maize is characterization of the maize cropping systems and associated arthropods in the major maize growing regions in Kenya.

Studies to identify and determine the relative abundance of the target and nontarget arthropods of Bt maize have been conducted in three maize growing regions: (1) Kilifi District in the lowland tropics in the Kenyan Coast, (2) Kakamega District in the moist transitional zone in western Kenya, and (3) Machakos District in the dry midaltitude zone in semiarid eastern Kenya. Similar work is ongoing in two other agro-ecologies: (1) Trans Nzoia District in the highland tropical zone in western Kenya, and (2) Embu District in the moist midaltitude zone in eastern Kenya. These studies were conducted in farmers’ fields, and different sampling methods were used for the various groups of arthropods.

Pit-fall traps were used for soil crawling arthropods, while the flying arthropods were captured using water and sticky traps. Catches from the various traps were recovered on a weekly basis and preserved in 70% ethanol for later identification. Arthropods on and in the maize plant were recovered by destructive sampling of randomly selected plants during vegetative, reproductive, and mature crop growth stages (Oloo, 1989). Stem borers recovered were reared singly in the laboratory for possible emergence of parasitoids. All the parasitoids and other arthropods recovered were identified and voucher specimens kept in the laboratory.

The stem borers that infested farmers’ maize crops, in descending order of abundance were in Kilifi, C. partellus, C. orichalcociliellus, S. calamistis and Cryptophlebia leucotreta (Meyrick); in Kakamega, B. fusca (Fuller), C. partellus, S. calamistis and C. leucotreta; and in Machakos, C. partellus, S. calamistis, and C. leucotreta. These findings agree with earlier studies at the Kenyan coast (Overholt et al., 1994) and in eastern Kenya (Songa et al., 2002a), and they suggest that in order to have an impact on stem borer damage in maize, Bt maize technology should be targeted at the most abundant species in each region.

The non target arthropods recovered included parasitoids of stem borers and a range of other arthropods. A majority of the parasitoids that were recovered from each of the three study sites were the larval type: Kilifi had the widest diversity, with six species: Chelonus curvimaculatus Cameron, Goniozus indicus Ashmead, Cotesia sesamiae (Cameron), Cotesia flavipes Cameron, Pediobius furvus Gahan and a Phorid. Machakos had three species: C. flavipes, C. sesamiae and P. furvus. Only two species were found in Kakamega: C. flavipes and Dentichasmias busseolae Heinrich. The exotic larval endo-parasitoid C. flavipes (the co-evolved natural enemy of C. partellus), was recovered from areas where releases had been made in Kilifi in 1993-94 (Overholt et al., 1997) and in Machakos in 1997 (Songa et al., 2001), respectively, thus showing good establishment and spread, and the need for studies on the nontarget effects of Bt maize on C. flavipes.

The diversity of arthropod families recovered from traps in the different maize cropping systems was 69, 67, and 59 in Kilifi, Kakamega, and Machakos, respectively. Out of the wide range of arthropods recovered, five categories of nontarget arthropods of interest have been identified, including potential biological control agents, pollinators, and decomposers of organic material in the soil. (Table 2). The arthropods that were most frequently recovered from the maize plants in Kakamega, Kilifi, and Machakos were Formicidae (ants), Forficulidae (earwigs), Blattidae (cockroaches) and Araneida (spiders).

Formicidae and Forficulidae are known to be predators of stem borer eggs and larvae (Oloo, 1989). Ladybird beetles, which are known to be predators of C. partellus eggs (Dwumfour et al., 1991), were also recovered, with Cheilomenes sulphurea (Olivier) being the most common species, especially at Kakamega Towards the end of this year, the arthropod characterization studies in all five major maize growing regions in Kenya will be complete; after which, internationally recognized criteria will be used to select the key arthropods on which the nontarget effects of Bt maize will be studied. However, some prominent arthropods such as C. flavipes, will be tested before then.

One of the activities in the arthropod characterization studies is the establishment of a reference collection. A reference collection comprises of voucher specimens of the various specific taxa collected from the field in the study sites. The different groups of taxa are preserved using appropriate preservation methods, including (1) dry collection, (2i) wet collection, and (3) a pictorial database. The reference collection is expected to serve as a technical reference during the monitoring of effects of Bt maize in the field, later in the IRMA project cycle. Interaction between Bt maize and the parasitoid Campoletis sonorensis in controlling the armyworm Spodoptera frugiperda

A study was conducted at CIMMYT headquarters in Mexico where a biosafety greenhouse and laboratory are available to conduct transgenic trials. The objectives of the study were to develop a methodology to test the effects of Bt maize on biological control agents and to identify a potential synergism between a Bt maize and a parasitoid wasp, which attacks a species of armyworm Spodoptera frugiperda.

S. frugiperda is similar to the armyworm species found in the Kenyan Spodoptera exigua. Two experimental protocols were used: (1) nochoice study: in which the armyworm and the parasitoid wasp, C. sonorensis, were placed on only maize leaves with Bt (CryIAb toxin) or on maize leaves without Bt, and (2) a free choice study, in which the wasps were placed inside a netting that contained both Bt and non-Bt maize infested with armyworm.

The observations made from the study were (1) the rate of parasitism in the Bt and non-Bt- maize was similar, with both types of maize resulting in a peak rate of parasitism of around 45% at 10 days after armyworm placement on the plants (Figure 3), and (2) there was a higher rate of parasitism on Bt maize following the peak, with 30% parasitism observed on day 12 versus only 20% for the non-Bt maize.

A possible reason for this difference in parasitism may be the reduced growth rate of the armyworm when feeding on the Bt maize. The average weight of the armyworm on day 12 was 9.7 for those collected from Bt maize versus 16.6 mg for those on non-Bt- maize. These results show that the wasp has a longer period of time to attack the armyworm feeding on the Bt maize, as once the armyworm reaches the third larval instar, it is too large to be successfully attacked by the parasitoid. In this regard, the Bt maize is enhancing the efficiency of the wasp in controlling the armyworm, even though the Bt maize is not directly controlling the armyworm pest.

Another possible advantage to this system is that with a reduced growth rate, the rate of cannibalism, which is known to occur in armyworm species, is reduced on the Bt maize. This means that the armyworms, which have been attacked and have the parasitoid larvae developing within them, could likely have a great probability of escaping cannibalism by neighboring armyworms and therefore facilitate higher parasitoid populations. This hypothesis has not yet been tested but will be the subject of future testing.

This study has now established a protocol for testing the interaction between parasitic wasps and transgenic maize to quantify their impact on the control of secondary pests of maize, such as the armyworm. Once a biosafety greenhouse is in place at NARL, these types of studies will be continued in order to quantify the impact of Bt maize on biological control agents and other nontarget organisms found in Kenya.

Development of appropriate insect resistance management strategies for resource poor farmers in Kenya One of the concerns about deploying Bt maize technology is the likelihood of development of resistance to the Bt toxins by the target stem borer species. However, the rate of evolution of this resistance can be slowed or stopped through the implementation of appropriate resistance management strategies. Mechanisms of insect resistance management (IRM) are based on three principals: (1) diversification of mortality factors (multiple mortality factors) (Georghiu, 1972), (2) reduction of selection pressure from each mortality factor (Wharlon and Norris, 1996), and (3) maintenance of susceptible pest individuals by providing refugia (Wharlon and Norris, 1996).

Due to the prevailing economic limitations and lack of technical support from both the public and private sector in Africa, an IRM strategy for Africa would have to be different from that for the developed countries. A proposed IRM strategy for Africa would include the following strategies: (1) the Bt maize to be released would have carry more than one resistance mechanism (pyramiding of genes), and (2) the refugia requirements would have to be met through existing cropping systems, and the refugia crops must be economically viable and socially acceptable in order to increase the probability of adoption of the IRM strategies by the farmers.

The germplasm to be backcrossed to the Bt maize source line will have resistance developed using conventional means. Another approach that has been used to reduce the evolution of resistance to stem borers is to develop varieties that carry multiple forms of resistance, for example multiple Bt genes and combinations of Bt genes as well as conventional resistance. In these instances, a borer population would have to develop multiple resistances rather than single gene resistance to one Bt gene. Crop mixes are common in Africa, often involving thick-stemmed grass species such as sorghum and millets as well as fodder species like napier grass.

It was desirable to determine the appropriate crops/grasses for refugia and how they should be provided in time and space, in order to substantially retard the development of resistance. The refugia area depends on the crop and the selected type of refugia treatment. Most stem borers of maize and sorghum are polyphagous and have several graminaceous and other wild hosts in addition to cultivated crops. Wild host plants of stem borers have been documented by various researchers (Ingram, 1958; Seshu Reddy, 1983; Khan et al. 1997; Songa et al., 2002c). The most important alternative hosts of the major stem borers (B. fusca, C. partellus, S. calamistis, and E. saccharina) are reported to be cultivated sorghum, Sorghum versicolour, Sorghum arundinaceum, napier grass (Pennisetum purpureum) and Hyperrhenia rufa (Khan et al., 1997).

Although stem borers oviposit heavily on some grasses, only a few grasses are favorable for them to complete their life cycles (Huttler, 1996). It is therefore important to select alternative hosts with economic value, e.g., high-yielding livestock feeds or food crops, which fit in the farming systems where the Bt maize will be planted. Studies on development rates of different stem borers have been done by Khan et al., (1997) on maize and a few napier grasses, mainly for Chilo partellus. However, there is need to study the development and survival rates of the common stem borer species in various agroecological zones on grasses of economic value. It is important to recommend to farmers the cutting regimes for napier grass and other forages based on the development time, to avoid harvesting napier before the pests complete their life cycle. Studies were initiated to develop IRM strategies for the different agroecological zones of Kenya, based on existing cropping systems. The studies included (1) surveys to identify and determine the importance of various potential wild and crop alternate hosts of stem borers in the major maize growing agroecological zones of Kenya, and (2) studies to evaluate the suitability of 30 potential alternate host for oviposition, development, and survival. After evaluating the 30 alternate hosts, preliminary results showed Columbus and Sudan grasses as the most effective refugia for C. partellus and B. fusca. Sorghum was the best host for C. partellus and B. fusca, given the high number of moth exit holes and the numerous tillers. Napier grasses attracted oviposition, but were not good hosts for larval development.

Gene flow is the movement of genes between plants of the same species. This is particularly found in crosspollinating crops such as maize, in which wind and insects disperse pollen to varying distances. The average farm size in Kenya is often small a factor that could result in pollen movement to several neighboring farms. Since the Bt gene is a dominant gene, and because the Bt gene should confer a selective advantage for resisting stem borers (the ear sizes will be larger), farmers will most likely select Bt maize ears during their selection process. There is need to understand the dynamics of gene flow in maize cropping systems in Kenya.

Research is underway to estimate the distance that pollen travels and to assess the methods that farmers use to select seeds, with respect to the relative location in the field. Most farmers in Kenya recycle seed for planting the following season. This has several implications for IRM. Unlike developed countries where farmers sign licensing contracts at the time of seed purchase, farmers in developing countries are not likely to report the breakdown of resistance. Therefore, techniques must be developed that will enable the early detection of resistance development so that steps can be taken to replace the technology in a timely manner to avoid a complete breakdown of resistance. Screening technologies should be inexpensive and sensitive enough to detect shifts in the insect populations in a timely manner.

A sampling protocol must also be developed to ensure that representative samples are taken from the major maize growing regions, especially those that have a high adoption rate of Bt technology. Agronomic studies will commence when insect resistant maize varieties are available. Seed production strategies will also be developed when insect resistant maize varieties are available.

Assessment of the impact of insect resistant maize varieties in Kenyan agricultural systems Before deployment of Bt maize in Kenya, it is important to determine the potential demand for this technology. Also, in order to develop technologies with high adoption values, it is important to have a clear understanding of the existing farming systems of the target areas (Songa et al., 2002b). For these reasons, participatory rural appraisals (PRA) were conducted in the five major maize growing regions in Kenya. Discussions were held with more than 900 farmers in 43 discussion groups.

The PRAs showed that a majority of the farmers in the lowpotential areas mainly grew local maize varieties; this included the lowland tropical the dry midaltitudes, and the moist midaltitudes. Most farmers in the high-potential areas (these include the moist-transitional and the highlands zones) grew improved varieties. The most important selection criteria were early maturity, yield, and drought tolerance, followed by tolerance to field and storage pests.

The major constraints to maize production were availability of cash, lack of technical know-how, and availability of good quality seed. The major pest problems were stem borers and weevils. Farmers showed a keen interest in new insect resistant varieties, as long as they fit in with their selection criteria, even if they are moderately more expensive.

However, since seed supply and quality are problems, the quality of seed and its distribution need to be guaranteed. A study was conducted to better understand the Kenyan maize sector. The study was based on a literature review, analysis of production data from the Ministry of Agriculture, and interviews with resource people and farmers. These studies showed that most restrictions on maize marketing have been lifted, markets for fertilizer and pesticides are liberalized ,and that these inputs are now widely available. However, the infrastructure is still poor, thus keeping transport costs high. Also, market information and access to rural credit markets remain problematic.

To estimate maize yield losses due to stem borers in Kenya, interviews were conducted in the five maize growing regions in Kenya; this yielded important data on farmers’ perceptions of the on-farm yield losses due to stem borer damage in maize. Based on the farmers’ estimates, the average national yield loss due to stem borers was estimated to be around 12.9%, amounting to 0.39 million tons, with an average annual value of US$ 76 million (De Groote, 2002).

In both 2000 and 2001, on-farm studies were conducted in 27 villages, on a total of 135 farms in order to determine the yield losses due to stem borer damage under natural infestation. In each of the fields, two adjacent maize plots, each 100 m2 , were established. One plot was treated with a systemic insecticide, while the other one was unprotected; the yield difference between the two plots provided an estimate of the loss due to stem borers. The results of the fieldwork were remarkably consistent with the farmers’ estimates.

Based on the onfarm field results from the five regions, the national yield loss due to stem borer damage in maize was estimated at 13.5%, a value of US$ 80 million. Of the total losses, the large majority fell in the high-potential zones, the moist transitional and highland zones. The results showed losses in the field between 9% in the highlands and 20% in the dry transitional zone. Crop loss assessment will be a continuous exercise in IRMA project to ascertain the loss experienced by farmers.

Technology transfer and capacity building.

Technology transfer is important in the development and deployment of new technology. Human and institutional capacity building in local institutions, are critical to success and sustainability of the technology. Training of KARI scientists was done through visits to Mexico and on-site in Kenya on genetic engineering, management of biosafety and entomology laboratories, and on how to conduct insect bioassays. Others were exposed to biosafety regulatory systems in Mexico. Training was also conducted on impact assessment and general methods in breeding and entomology.

Where new technology is being developed and disseminated, communication is important for education and creating public awareness. Considerable effort has been given to creating dialogue and raising public awareness about biotechnology in general and Bt gene-based stem borer resistance in particular. The main thrusts of the communication work at this juncture are to demystify the technology as much as possible, and to convey that project scientists are intent on doing their work thoroughly and with great care (no short-cuts).

This has been achieved through the following avenues: (1) Stakeholder meetings to engage a range of stakeholders on key issues and developments in the project; (2) print materials to inform diverse audiences about insect resistant maize technologies, such as public relations materials; (3) IRMA web site; (4) establishment of good media relations, including having journalists participate in workshops with scientists on “working with the media”; newspaper clipping service.

In the near future, as possible deployment grows closer, the emphasis will shift to reaching farmers with more practical information about the technology; this would be by working with extension agents, developing appropriate materials, and utilizing other communication channels that farmers access. Another key item is the documentation of the project, the thought being that the lessons learned by the project in Kenya might be helpful to other African countries that may want to pursue the use of GM crops.

This documentation is being maintained through (1) publications that track the project’s progress, especially the IRMA Project Document series, (2) a quarterly IRMA Updates Newsletter), (3) conference and seminar papers, (4) raw video footage of key interviews, processes, and TV coverage, and (5) the news clipping service.