«Biotechnology-Assisted Participatory Plant Breeding: Complement or Contradiction? PPB Monograph No. 3 Ann Mane Thro and Charlie Spillane 1 7 ...»
Melchinger et al, 1998; Van et al, 1998). The majority ofwork with QTLs is likely to concentrate on low QTL x E effects. However, a PPB project seeking to exploit high G x E effects for adaptation to a s pecific environment cou ld assemble germp lasm containing QTLs exhibiting high G x E effects from existing MAS efforts and test them.
Sorne form al breeders feel that, as recen t advances in MAS methods allow traditional plant breeding objectives to be met more efficiently, resources should become ava ilable for pursuing other goals that were previously considered too costly-inc1uding, perhaps, location-specific breeding (L. Sanint, M. Gale, K. Schmidt, pers. comms.). Stratcgic research to create thc necessary biotechnology applications could improve the cost:bencfit ratio of pla nt brecding targetted to the locationspecific necds of resource -poor farmcrs in dcveloping countries. In addition, geographical information systems (GIS) could be used to search for similar micro-environmcnts that might form part of the 'adaptation dornains' of varieties bred for local adaptation (G. LeC lerg, pers. comm.), enabling the results of location-specific PPB to be scaled up.
As we have a lready secn, where farmers are operating in heterogeneous, risk·prone, marginal environments, a single crop Síotechnology as a Set 01 Tools lor Fonnal and Informal Plant Breeding variety (or technology) is unlikely to meet all their needs (Chambers, 1983). In the past decade there has been a shift in research and extension practices towards providing a 'basket' oC options from which such farmers can choose according to their needs (Witcombe et al, 1996, 1998, 1999; Ashby and Sperling, 1994).
The reproductive processes of gennination, vegetative growth, Oowering, and secd maturation are- vital to resource-poor farmers.
Many minimize their risks by planting difIerent varieties or crops which mature at different times of the year, ensuring a steady supply of food (Gilbert, 1995). Farmers can be offered varieties with a mix of maturatLon periods and altematLve storage and processing characteristics. Intensive research is currently being conducted on the genetics of flowering time (Laurie, 1997). Biotechnology could be used to expand the range of varietal maturity options.
MAS can help breeders transfer the loei associated with maturity into otherwise desirable gene tic backgrounds with minimal alteratíon in other varietal characteristics (W. Beversdorf, pers. comm.). More genes and loei controlling flowering time will doubtless be identified over the next decade, and lmowledge generated on how they operate and interact. Other possibilities inelude the linkage of Oowering time genes to promoters so that flowering can be induced, shortening generatíon times. This will be especially useful in the early stages of breeding programs, when rapid progress needs to be made and demonstrated, and wherever there is a need to avoid continuing or imminent stresses (Laurie, 1997).
The relationship between Oowering time (heading date), crop adaptation, and yield is critica!. Clawson (1985) pointed out that tropical farmers orten use different colored varieties, which are associated with difIerent maturation periods. He concluded that fanners' adoption of modem varieties would accelerate if they were offered multiple high-yielding varieties of staple food crops of varying seed color and maturation periods.
At any rate, farmers may be unwilling to adopt any of the new options they are presented with unless they can easily distinguish them visually (S. Morin, K. Longley, pers. cornms.). A considerable arnount oC work on human cognition and the relationships between classificatLon, cultivation, and selection has recently been done. The model of 'selection for perceptual distinctiveness' developed by Boster (1985) suggests that, if farmers cannot distinguish between varieties, they will not be maintained in local fanning systems. At present, the Boster model applies mainly to root crops that reproduce vegetatively and is less relevant to out-breeding grain crops.
making it difficult for farmers to di stinguish bctween them (S. Morin, K.
Longley, pers. comms.). Many of th ese varielies dis play excelle nt qualities and in th eory offer farmers a much wi der choice. S ut lhis choice may not be exercised in practice if th e varie ties are not phenoty pically d istinct. Work is under way to adapt the Boster model to ri ce (Lon gley, 2000).
Molecular markers can be u sed tI) maintain or increase genetic diversity at a locus or range of loei that are neutral for agron omic lraits, while selecting for such traits al other non -neutralloci (Ribaut and Betran, 1999). This approach could be used lo maintain allelic series or a range of non -agronomic visual phenotypes (e.g" flower color, seed color) during the early stages of a breeding program, so a s to increa se the likelihood that the fmal products will be phenotypically distinct.
Farmer-frlendly specialized collections?
The provision of a range of existing vari eties to inte rested farmers is an important func tion for genebanks (FAO, 1996). The practica l difficu lty of screenin g large numbers of germplasm accessions wiU be felt just as acutely by farmers as by formal plant breeders, or even more so. To make screc n ing cheaper and easier, many genebanks h ave established core collections, designed to represent a crop 's ma.x.imum genetic diversity throu gh the mínimum possible number of accessio n s (Hodgkin el al.
1995). At least 63 different core collec tions of 51 crops have been established worldwide (Spillane el al, 1999). Plant breeders and biotechnologists havc, in add ition, developed specialized experimental coll ections, such as near-isogenic hnes and special ge netic stocks, to facilitate their research.
There has becn Httle systematic thinking about how these s pecialized collections migh t be adapted to meet the needs of PPB.
Several end-user oriented variations on th e concept of s pecialized collection s have been proposed, but not yet tested (e.g., van Hintum, 1999). Van Hintum et a l (2000) have developed an on-hne selector which allows u sers to define their own collection (see www.cpro.dlo.nl/cgn/coreeoll/usercore.htm).
Alternatively, after farmers have de fined their criteria, breeders could search germpl asm collections for corresponding genotypes a nd assemble them into source populations for farmer breeders. For example, the collection of caSSRva clones being developed by a coopc rative of small-scale farmers in coastal Ecuador (see Box 10) will, at the farmers' request, inelude material from CIAT breeding populations. These materials are be ing selected by a CIAT breeder according to criteria specified by the farmers, which inc1ude high yield, drought toleran ce, and good processing quality. GIS are an additional tool that can be used to support the assemb ly of sets of a ccessions adapted to specific environmental variables (D. Wood, pers. comm.).
Biotechnology as a Set of Tools for Formal and Informal Plant Breeding
Tools ror Promoting Recombination
Conunentators vary in their liiews on the optimal amount of recombination, or mutabilily in its larges t sense, that should be included in PPB. Sorne Ceel that m ethods derilicd in th e laboratoT)' may nOl be superior to evolutionary processes in the Cield (J. Jiggins, pers.
comm.). Others, however, such as Simmonds (1979), h ave feIt that the limitations to recombination have beeo one of the major con strrunts to selection efTorts by both formal and informal breeders.
Creating endogenous genetic variation
Farmer-led PPB is likely to face constraints in accessing and¡or managing new genetie variatian from ou tside the farming system. The faet that formal breeders have made considerable progress u sing endogenous genetic variation- variation available in limited or do sed breeding populations-alone may be highly significant far farmer-led eIToTts (Leng, 1974; Wych and Rasmussen, 1983; HallaueT, 1986;
Mac Key, 1986; Dudley and Lambert, 1992 ; Manninen and Nissila, 1997; Rasmussen and Phillips, 1997).
Chemical treatment or nuclear irradiation have been used to induce mutabons fOT !he purposes of crop impTovement (FAOjIAEA, 1986).
Commonly used mutagenic chemicals like EMS introduce point mutations, while X-ray irradiation leads to gross chromosomal changes.
Because these techniques do not distinguish between human and plant DNA, highly controlled experimental conditions are required to protect users. For this and several other reasans, these methods could not easily be u sed by farmers.
Anoth er mechanism for inducing rnutage nesis is transposition (Wessler, 1988; Peterson, 1993). This relies on transposons, which are naturally occurring genetic elernents (i.e., pieces of DNA) that move around the genome of most plan t species. Transposons generate new genetic variation as they move. The rate at which different transposons move through particular genomes vanes widely, and with it the rate at which variation occurs (Levy and Walbot, 1990). A recent study of mruze demonstrated the importan ce oC transpasition in generating gene tic variation (Fischer et al, 1995).
The advent of increasingly sophisticated and controllable transposon mutagen esis techniques h as already revolutionized plant molecular biology research (Sundaresan, 1996; lzawa et al, 1997). In sorne plant species (e.g., Arabidopsis, maize, and rice), these techniques a re now being used as experimen lal tools by biotechnologists, primarily to identify genes and/or phenotypes through insertional mutagen esis (Sundaresan, 1996 ; Izawa et al, 1997). They are províng more accurate and potentially useful than previous mutagen esis approaches. In theoT)', they could eventually be used to help fa rmers genera te, augment Biotechnology-Assisled PPB: Complement or Contradiction ?
or 'release' u sefuJ variation within local germplasm (R. Jefferson, pcrs.comm.).
Transposon mutagenesis techniqu es can generate a lleles associated with a gain or a 1058 of fun ction for many phenotypic traits an d have been primarily used to date in the iden tificatíon of the loci associated with specific traits. At prescnt a research group in Wageningen is using these tcchniqu es to over-express, mis-express or ectopically express ca ndidate tra nsgenes a t different locations in the genome in order to gen erate n ew phenotypes (A. Perelra, pe rs. comm.J. While most available tran sposon techn ique s are s uitable only for laboratory-based line selcction a nd screenin g, the tech n iques cu rren tly under developrnent \Vil! enable selection and scree ning to be done in experimen tal Cields. Il is likely that field-Ievel techniqu es such as promoter perturbation, gene knockouts, or activation tagging could be developed or adapted for use to generate ge netic variabon for PPB an d PVS program s.
So rne co mmenta tors [ee! that 'randoro' mu ta genesis approaches of this kind will nol be usefu l to farmer-breeders because they wiU genera le mo re )unk' variatíon than farmers can han die (D. Duvick, pers. cornm.). They su ggest that sorne pre-screening for desirable phenotypes would have to be done by fo rma l researchers befare farme rs would be inte¡-ested. The poteritial of transposon systems for ge neratin g ge netic gain could probably be empirica Uy tested against conven tionaJ breeding techniqu es. However, biosafety regulatio n s make it unlikely that farmers will be allowed to experiment at Cield level with transgenic transposon mutage nesis techniques.
Controlltng recombinatton rates
Another way of increas ing endogenous ge netic varíation is through o pti.mizin g the process of recombination. This issue is considered by sorne to h ave been neglected in plant breeding compared to the techniques of selec tion and isolation (Simmonds, 19791_ Recognizing tha t a high degree of genetic variabili ty is required for major evolutio nary adva nces, Stebbin s (1 959) a rgu cd that, \vhcn-endogenous mu ta tion rates a re low, genetic recom bina tio n is the most likely source of such variabi.lity and that recombin ation-genera ted diversity could be maxirnized by hybridization between populations with d¡fferent a dap tive norms. Reco mbination within the sequ en ce of a single gen e and epistastic effec ts-the effecls of one gen e on another- have been ident.ified as a potentially important source of new genetic variability in the development of elite germplasm (Schnable e t a l, 1998; Rasmusse n a nd Phill ips, 1997)_ For ¡nstance, the generation of new s pecificities through unequal crossing-over within co mplex resistance genes d urin g recombination has been demonstrated, to date rn ain ly in mode l systcms such as the Zea mays-Puccinia s orghi interaction (Pryor and Ell is, 1993).
Biotechnology as a Ser 01 TooLs lar Formal and Informal Planl BTeeding
The level of recombination in farmer-Ied PPB is likcly to be far from optimal for the purposes of generating endogenou s genetic variation.
Increasing it could help.¡Ha nson, 1959a, 1959b; Rieseberg et al, 1996).
but n ot always to the same degree. Crop plant genomes difTer in their 'permeability' as regards the introgression of different genes or chromosomal regions, whether by wide-cross recombination with wild relatives or whe n crossed with other dornesticates in the prirnary genepool ¡e.g.• Rieseberg et al, 1996)_ Mating strategies have a significant efTect on recombination rates.
They rnay be important for genetic enhancement or pre -breeding, especially where the re sources lo conducl marker-assisted introgression a re not a va ila ble (Tanksley el al, 1989). lmproving fa rmers' ma ting strategics could prove cost-effective in PPB programs, ¡Spilla n e and Ge pts. 2000).
Molecula r mapping efforts are like ly to increase knowledge of the genomics of recombin ation ra tes, both within and between crop genepools. The existenee of genes that influence c rossability in many species indicates that the presence or absence of these genes in farrners' populations may affeet reeombination ra tes as well as interspecific hybridization (e.g., Luo el al, 1996). For instan ce, the genes la 1, kr2, kr3, and kr4 found in wheat cultivars such as Chinese Spring (and in sorne Chinese landraces) are known lo facilitate crossability with species of olher genera (Luo et al, 1996; Jiang el al, 1994).