COlECCION HISTORfCA • 71? jZ7 ,e s, World.ng DocUment No. 47 ~ 'roPICS IN BREElJIN:; OF a:H'K)N BEAN 7-12 Noven1ber, 1988 Centro Internacional de Jlgricu1tura Tropical (CIAT) , cal!, Colombia Editor: S. Beebe IMPLICATIOOS FtlR BEI\N BREElJERS OF S'lUDIES 00 '!HE ORIGINS OF a::MOI' BFANS, Blaseolus vulgaris L. D. G. Dei:locIck and J. Tdlme* A revision is made of three lines of evidence (botanical, ardlaeological and biochemical results) for a lIll1tiple and in:lepement danestication of wild populations of cx:atllfJll bean in the Allericas. 'lhese danestication events took place at least 10.000 years ago frau different wild bean populations. In additión te a fOl.ll'ñer effect, a lawer number of populations caU.d llave been danesticated in Joi3soamerica versus the soothern Ames. sane li:mitations related te the fOl:lllation of gene poole and the unifonnity of the selective pressu:c:es are diecnssed. Pl:actical consequen::es for bean breed:in;J related to the genetic i.rxxJnptibility and a possible case of co-evolution are presentad. Part l. Origine of P:I!mn Ijoou l. Evjc'lia, Peru aro. Bolivia, the differenoes are less clear-cut. largar seed size aro. annualism have been set forth for considerirg the A1:gentinian fonns as feral (Gentry, 19(9), although sane of them were fourxi in rain forest habitats (Ibrkart aro. Brücher, 1953; Debouck, 1985). It is pe:dlaps too early to conclude about a sirgle place of origin for wild R. vulgaris on the basia of sud1 arguments, since sane true wild Phaseolys species sud1 as R. drlapaaanus piper or R. maculat us Sd:leele also have la:i:ge seeds (100 seed weight: 16 aro. 26 g respective1y) and were not dcanesticated. R. migrnrarpus Mart., an annual vine fi:a¡¡ the tropical deciduous forests of MeXioo, was not dcanesticated either. lis it ~ in Table 1, severa! dlaracters display a contimum alon:J the ran:Je of clist.ribution. Up to northen1 Peru (cajamarca), seeds tend to be small and roun::ied, while they are larger and paralle1-epipedic in the aborlgineus sensu stricto type fraII southe:rn Bolivia to A1:gentina. For materials from the southe:rn Andes, one can won:ier Whether it ia an adaptation to forest habitats urrler cloudy c1imates whe:re larger storage reseIVes are needed to gendnate. Lignified hypocotyl ia generally ohserved on old plants which can survive the fi1::st seed set in sane materials from MeXioo to Colcmbia. It is linked to the capacity for buds to l:egl:CM on the lower stem, a trait also observed in sane escapes (DGD and JSMM JI 2077 fran Queretaro). Bracteole size is scmewhat variable within a sirgle population; a clear-cut differenoe ia, howeVer, displayed 5 by the aborigineus form pl: esent fI:all scmhern Bolivia te Al:qentlna. CUriously, the ext:cemes of the range are earlier te flower than the wlid forms of Costa Rica anj Colanbia, the aborigineus being the earliest. Evidence fot the large range of distrib.Ition in the Americas anj the I1,::ultlological differences between the wild forms have been strengt:hened by recent explorations. 'Itlese facts have lead several authors (Evans, 1976; Heiser, 1979; VaOOerborght, 1986) te hypothesize two centers of bean domestication: lIIesoaIlerica anj the southem AOOes. Oí particular int:.erest is the fact that saoo IOOqilological traits separatin;J the wUd forms, such as seed size, were also fami of saoo significanoe te separate groups of cultivated larrlraces (Evans, 1976; Varx:lerborght, 1986). In other words, beans, both wUd anj cultivated fI:all a same origin, share te saoo extent sane similar c::haracteristics. We wi11 retum te this later. Te sumrnarize: '!he wild form of oalülAl bean has a large distribution in the Americas fI:aIl Chihuahua te San Luis. '1hroughout its distrib.Ition, there are mor¡;hological dlan:Jes whidl partly reflect its adaptation te different environments (e.g. sunny thickets in Jalisco versus rain forests in 'li.lcIlman) • Part of these moqilologica1 traits are shared between the wUd beans arxl their cultivated 0CRII'lteJ:part danesticated within the same area. We wi11 l10W revise sane ardlaeological records in:llcatin;J ancient anj probably :i:rrl.ependent danestications in the Americas. 1.2 AJ:phaeoloojrnJ evidences. Since we will in most cases deal with ancient plant remains, a11 our observations will be linked te their conservation ooniitions. ene wUl then not be smprised te firrl most 6 evidences in dzy amas am not in wet tropical America. In this regard, as it appears fn:m the data shown in Figure 2, na"ll:ds fran Cent:ral America am Colanbia a:te particularly lacking. 'lhis bias oould be partly oorzec\::.Ed in p.It.t.irq :oore E!!I{ilasis on pollen records am PtYtoliths (Pearsall, 1982 # Bozarth, 1986), b.lt this awroadl has just started. On the at::her harxl, one shruld not fm:t¡et that ardJaeoIogical eviéleooe is a non definitive axgun-ent, pendin;J the discaIIery of a mvel ancient site. ot:;her prdJl.ems in interpretation a:te relatEd to sbJdies of stratigra¡ily (F.rJ;Jel, 1963) am to the daninant use of 14C datirg met:hcd with ffiN coont:enteasure (qmeh et al., 1985). Bearing this in mi.rd am the fact that the data a:te still fragrnentary, the data of Figure 2 am the relatEd papers (Kaplan, 1956# I{apIan am MCNeish, 1960: Brooks et al., 1962; I{aplan, 1965: Kaplan, 1967; I{aplan et al., 1973; Ta:r.rago, 1980: Kaplan, 1981; ottonello am IDrardi, 1987: Pearsall, in press) in.:ücate that: '!he oldest records a:te fn:m HUachi.chocana, en the west side of the Quebrada de H1..mIahuaca, Jujuy, AJ:gentina, then follCMed by Guitarrero caves, Ancasb., Peru, am then by Tehuacan, Puebla, Mexi.oo. '!be ran:;Je of very old sites is large (7000 km between Tehuacan and HUachichocana) • '!he llDSt ancient beans a:te preceramic and fourrl :in a context of very :incipient agriculture (plant gatherirg!). As evidenced by Kaplan am Kaplan (1988), the American plant trilogy (maize, beans, squashes) does not exist when the beans a:te domesticatEd. 1 Tak:iIq into aco:llll1t the present distribution of the wild tonos, and 1 assuming that the Iatter did not change over the last millennia, the old ardlaeological sites tall very clase (100 -km or lesa) to their ran:;Je. 7 1 I '1be arc:::haeologica1 I:leans awear te be fully dalesticated, that is, their seeds do not show any transition fran the wild. '1hey also awear te be fully 1l1Odern. For exanple, seed sizes arxl color pattems are oanparable te 1900 t:ypes still present for saoo in the sama area. 'Ibera is thus a noticeable stabiliq.- in these landraoes, arxl an out:st.am.in;J continuiq.- by the fanners in g:r::ow.in:J them. Further evidence about continuiq.- is shown by the different dates at the sama place (e.g., Huachichocana reportad by Tat:xago, 1980; Oca"'I'Q repotted by Kaplan arxl M:::Neish, 1960). '1be above mentioned elenents are in favor of an irrlependent danestication of wild I:leans in the IOOUntainous regions oí Mesoaloorica arxl SOUth J\merica where they were grow.in:J. However, with the beginni.tJ;J of ceramics, I:leans became a major fcx:xi plant, arxl long befare Coll.lJl1OO.s, seed trade arxl crop diffusion took place. In ancient Petu, by at least 4000 B.P., the I:leans lOClVed fmm the highlarrls te the coast (Picke:rsgill arxl Heiser, 1978; Pearsall, 1978: Pearsall, in presa). In ancient Mesoanerlca, the I:leans lOClVed fran northwestem Mexico 1nto the southwestern USA (by 2000 B.P.: Kaplan, 1956; Kaplan arxl Kaplan, 1988) arxl pxobably fran there te the eastern USA (by 1000 B.P.: Ford, 1985). How far they changed dur.in:J these precolumbian seed lOClVenents will be discllssed later. 1.3 Biochemi.cal evidences. 'lhere is no evidence that aboUt: 10.000 years ago early agriculturists aOO¡or animals took an active part in the dispersa! of wild I:leans. Rather, the I:leans wexe pxobably self-dispersed. In preceramic contexts (aprox. 8000 years B.P.), in both Ocanp> arxl Guila Naquitz (Kaplan arxl McNeish, 1960: Kaplan, 1985, respectively) I plant gatherers collected wild I:leans, but these were of -hypogeal gexmination, perhaps wild 1? cocci.neus arxl 1? aaxacam.1S respectively. Also because wild bean seeds contain antinutritional factors sudl as };tlytOOaemagglutinins (Jaffe and Brücher, 1968), one can assume that the present distribution of wild;J? vulgaris is. broadly similar te that of that periodo 8 If one CXlUl.d fiIrl a biodlemica1 marker with the following praperties, present in both the wild populations arrl their cultivated ooont:erparts, it would be of great help in clarliying the dmesti.catioo pt:0CleSS: Geclgra¡túc polymoq:.hism: several forms of the marker should exist t:h:rcu;¡hout the ran;Je of distribution of the p.ttative anoest:or(s). High heritabillty arrl oatplexity of each of the variants of the marker, so that the.re is little probability that the sama variant will be prodll.ced several times during the histol:y of the crop. Environmental stability: the expression of the variants is not influenced by the growing conditions. stability facing the danestication prooess: the marker is not influenced by the danestication process, the selection pressures being on other characters. Gepts (1984, 1988) dEm:mstrated that ¡ilaseolin, the major seed storage protein, is a useful marker for evolutionary studies in;f. vulgaris. As ge:t:nplasm has been made available by explorations in latin America, larrlraces and wild populations have been examined for their ¡;tlaseolin characteristics (Gepts et al., 1986; Gepts & Bliss, 1986: Del:Joock et al., 1988; Koenig et al., in press; 'ltllme & Del:Joock, in press). '!he results are shown in Table 2 arrl Figure 3. Although the ran;Je of wild ;f. vulgaris is not fully covered yet, ene can sea that: 1) 'Ihere are llOre ¡ilaseolin types in the wild mataríaIs than in the cultivated enes. As discussed below, this shows that ooly a portioo of the genetic diversity existing in the wild forms was danesticated. On the other haIrl, in the case the actual wild beans will just be feral or weedy types, they would not express such a lcn:ger diversity in ¡;tlaseolin types. 9 2) 'lhere is a good otrres¡:x>Mence between the nativa latñraces of a particular country. aIXl the wlld beans p¡:g;ent there. 'lhis was expected given the relativa stabllity of latñraces as shown by the archaeological records (see abJ\re) aIXl the roleollar conp1exity of p:¡aseolin causing its uniqueness (Gepts, 1988). F\.Irthenrore the presenoe of "odd" types in a particular country can be explained by varietal introduction (e.g. the "cacahuate" types in Mexico with 'T' p:¡aseolin, or the "panami.tos" in coastal Peru with ·s' ¡:haseolin). 'lhere is also evidence of early seed movenv.mts, prdJably preo:>lunt>ian, with the ext:ension of the 'B' ¡:haseo1in cultivars into Central America. 3) More variabllity exists in the southem Ames as ~ te the nort:hem Ames aIXl Mexico both in wlld pcp.¡lations aIXl in cultivated materials. 1\ssI.lltti.n;r oc::mron bean orlglnated in Mexico (as its has been clainal by Gentry, 1969; Heiser, 1979), it wouJ.d be difficult te explain such a diversity (tIlis statement is valid fer the last 10.000 years). Orle shruld teke into acxxru.nt, however, the polymorp,úsm of the 'M' phaseolin in w~ld types of Mesoallerica, so far absen:t in any cultivated type. Another piece of evidence aI:x::Alt. a separate d.ctoostication in MesoaIoorica vs the southern Ames is glven by isozyrne anaiysis (Schinkel et al., 1988) I since they are coded by genes (probab1y unl:inked) irñepen:lent of ¡:haseolin. Analyses perfonned on latin America laIXlraces reveal contrasting isozyrne proflles fer the two regions. Similar results were obtained for a collection of Malawian gen¡plasm (Spred:J.er, 198sa) once the accessions were traced back te their canter of erigin in the Americas using pbaseolin types. Two daninant alternate isozyrnes patterns were reported for 6 enzyrnes where polymor¡;hlsms were abseI:ved. '!be daninant Arñean pattern was p¡:g;ent at a frequency of 76%, Wile the Mesoaroerican one was of 15%. '!be renaining 9% were distriWted between the Arñean gene pool (8%) aIXl the MesoaInerican one (1%), with variants of the respective 10 daninant patterns (Spred:ler, 1988b) • 'Ihese frequencies for Malawi genplasm "WOUld lead ale to tlúnk: that, at least for the Andean gene pool, it would be possible to recogn1ze subunits, a corx:lusion also read:led by Sdúnkel et al. (1988) work:iIg on American cultivars. Fbr a better definition of races, it woold be very useful to sc::reen additional primitive American varieties as well as pcp1lat1ons of wild beans. '!he fact that nme pol~ has been foon:i .in snap beans (t"ee?en, 1984) also suggests a need to inclu:ie nme loe1 when evaluatirg wild pcp1lations. '!he presen::le of rore J:ilaseolin types .in wild pcp1lations vs cultivated larx'lraces mises questions aba.lt the danestication prooess, given the cbaracteristics of ~l.in as a l'M.rker. Iet us JXM djsaJSS ~ of them. 2. '!he fO!.!!'lier effect of cicm¡¡stication We will adopt the cc. ..~ pL developed by Iadiz.i.ns:k;y (1985) and thus define t;he foun:ler effect as the reduc:tion .in genetic variability when iOOividuals giviT'g rise to a new pcp1lation only carry part of the original genetic variability. '!he foun:ler effect .in bean danestication cx::JUl.d be considered urrler the followiT'g aspects: 1 '!he number of iOOividuals taken flXlll the wild pop'llations and .i.ncluded .in the dc:llreStication plOOOElSS was small. 2) Gene flOlrl between the dc:llreSticated stocks arn their wild counte:t:parts was limited. II 2.1 'lbe limited initial gene stock. Evidence that parts of the geogra¡;trlc rarg-e were t:oodhed is foord in the ¡t¡aseolin patterns of the cultivated types vs the wild ones (Table 2). In Mesoo:merica, the );ilaseolin type fourd so far in the cultivated materials is 'S', while the polynx:n:prlc ard cxmoon 'M' type has not been foord in any cultivated genotype eJCalIIined to date. ene CXlUld think that the ¡;ilaseolin 'M' VIOOld be li.nJ«ld to a negative dlaracter rejected durin:;J the danesticatian process: it has been transferred in experimental breedin;J liDes without e:xpressin:J any negative dlaracter :r:egardin;J to yie1d, etc. (F, miss, pera. ocmn.). In the sarthern Ames, the types '1' ard 'J' have not been foord in cultivated materials so far, while the 'T', 'C', 'H' are present in both wild ard cultivated genotypes. 'lbe fact that the 'A' type (fourd only in Ayacucho so far) is present in a 1aoorace ard not in a wild population is, to us, more illustrative of the i.rlcu!p1ete survey for both cultivated ard wild materials fram the sarthern Ames than anythin:;J e1se. In ex1:llIÚn:J Figure 1, it is obvious that the rarge of the wlld populations is not fully oovered yet, ard that perhaps only 10% of all the Alrerican Iaooraces !:lave been proaessed. As sean in Tab1e 2 ard Figure 3, the situation in Mesoo:merica CXlUld be different fran the situation in the sarthern Ames, Le. a very few wild populations CXlUld have been danesticated in Mex1co while severa! wild populations ovar a broader rarge of distributian oould have been danesticated in a zone fran mid-Peru to norl:hwestern Argentina. As indicated by Gepts (1984, 1988), the t:ruly wild types in Mex1co showin:;J no sign of i:ntrogression fran cultivated forms are concentrated in west:ero Mex1co which CXlUld then be a place of bean danestication. Orle shruld, however, keep in min:l that when those stWies were nade little was known aboot wild beans in the region fran Chiapas to Nicaragua, 'Where tbey are known to be present ard still to be collectad. 'lbe possibility of firdin:;J more places of bean danestication in Mió:Ue Alrerica thus remains apen. 'lb identify a sin:;Jle site of danestication in the sarthern Ames is, for the time beÍDCJ, more difficult ard pe:rtlaps irrelevant. After considerin:J the 12 latitudinal extension of the wild beans in SaIt:h AlIerica, Briicher (1968) arñ Harlan (1971) dalbted that there was a spec1fic center of origin for the rouiiUOtl bean, b.lt were th.inkirq abaut a II1Il1tiple danestication there. !&:>re detai.led infcn:ma.tions CXlUld be d:ltained looJd.n;J at the mi1:odJorxkia1 lliA but this awroacb is just beginnin;J to be deve1oped. Ano\:her piace of evidenoe that part of the wild pop:ilations arñ petllaps just a few plant:s were inclOOed into the dnn!E!stication process is shown by 'WIOrk on brudúd resist:aooe. 'l1lousarJ:h¡ of all.tivated genatypes llave been SCIeened for :resistance to that pest, b.lt not a siD;Jle ene has shown :resistance (Sdlc:xlnhoven arñ cardona, 1982). HoWever, a :reduoed I1lllllber of wild pcpllations of J¡l. vulgaris f:r:an Mexico llave shown :resistance (Schoonhoven et al., 1983). It was dem::lnstrated that the :resistance, an antibiosis react1on, was linked to the presence of the protein arcelin (Osbom ei: al., 1986, 1988). Scma polymoqirism exists in aroelin (Osbom et al., 1986), those types with the highest levels of resistance beiD;J colloentrated in Guen:ero arñ Michoacan, Mexico~, as claimed by Gepts (1984, 1988), danestication did not take place. '!he reason why no souroe of b:rudúd resistance was fourrl in all.tivated bean may s.inply be because the trait of interest, the right aroelln type, was not present in the danest1cated pcpllation(s). Although evidenoe lB fragmenta:t:y, it thus se liS that a f8JI pop:ilations alor.g their lor.g range. in the Anm'icas were in fact da.nesticated. Although evidenoe is even roore scarce, it alBo seems reasonable to conolude that just a f8JI plant:s within those pcpllations actually participated in the danest1cation process. wild beans are still haIvested as emergency foad in same parts of the .!\OOes (Argentina: Briicher, 1954¡ Pe:ru: Debouck arñ Tohme, 1988). Imy particular variant (e.g. colored seed, non dehiscent pod) might have been notioed arñ selected by early agriall.turiSts. since many traite of high anthropic value are highly heritable (see Table 3), PLU;jle&S in selectinJ for those traite wauld have been rapid. 13 As pointed rut by Iadizinzky (198S), there was no nee.i to oont:in.te to grow wild types once a llOre suitable genotype had been abtained (altl:loogh we wi11 c:atmnt 011 that later 011). As is still dale today in traditional agriculture in Peru, farme:rs frequently exdlaIY;Je seeds on an individual basis. When they see an interestirg plant they do not llave, they pick up a few seeds (in sane cases just one), walk back to their huts, aro plant them next season to see how they grow. 'lhi.s form of selection oould elq)lain the diffusion of the 's' types throughrut ~oanerica starti.n;J fran a few plants once dc.m:lsticated in westan M'axÍco. 'lhi.s is also illustrated in ColaIbia: alt.Íloogh the early Colanbian In:lians began with selecting bean pcpilations with 'B' ¡;:haseolin (Gepts aro Bliss, 1986), they were eager to ilrport 'S' types in northem Colanbia aro 'T' types in southe:rn Colanbia. Another evldence of tbis form of selectlon ls suggested by the work 011 bruchid resi.stance. A possible explanation of the absenoe of resi.stance in any cultivated qenoytpe would be that the rlqht populations were not t:ouched by dc.m:lsticatirg farme:rs. One shcW.d also renaober that riqht a.rcelins occur in low frequencies in the original wlid bean pcpilations (artllIl'rl 20%: Osbo:rn et al., 1986). Glven tbis, the few plante with interestirg seed, pod, or plant characteristics would not include lIOSt probably the a.rcelin conferrirg resi.stance to bruchids. A remainirg question is to conslder whether there was (or is) any genetic flow fran wild pcpilations aro¡or early cultivare in recently danesticated stocks. 2.2 '!be gene flow between bean genotypes. '!be prOOlem of a possible genetic flow to increase diversity in cultivated }? vulqaris, aro so to limit the founder effect, can be considered in two aspects: first, the gene exchan:Je durirg daoostication between synpatric wild forme aro primitive cultivare, aro secord the crossirg between cultivated genotypes 14 cutside t.he rarqe of t.he wUd forms. But we should first eJaUI1i.ne whether t.he gene flow is ¡:hysically possible based on the rata of natural ootcrossilg in R. YUlgar18. C!:oss-pollination apparently exists in the wUd forms. For Mexican materials, wa:u arñ wall (1975) arñ Van:1erlxn:ght (1982) lepol:t.ed high levels of ootcrossilg (3 to 50%), although Briidler (1988) spoke about Han outspoken cleist:ogamy'l far the R. aborigineus forms. COnflictilg figures are also lepol:t.ed far cul.tivated R. vulgar18 genot.ypes (see Table 4). Although these results need further oonfitmation particul.arly with respect: lcx::ation effect:s, ecological oonlitions, arñ to pollinatilg agents (carpenter arñ l:l\.mi:)le bees seem quite active in this xegard: Delgado et al., 1988 arñ persa1al abservations), it a¡:p:!arS that cleistogamy 18 pethaps not as constant as it was thought befare arñ that high rates of cross-pollination (> 10%) cool.d be fourxi in both wUd arñ cul.tivated a:nniOn bean. C!:ossilg between wUd forms arñ primitive cul.tivars can pethaps be inferred from what can be seen now where both are still growin:J together. lI!tiOnJ Blaseo1us species, wUd R. vulgaris behaves as a weed in sorne plaoos, enterilg human macle ar distu:ttled habitats (in Mexioo: Delgado et al., 1988; in Peru: Dehouck, 1987; Dehouck arñ Tobme, 1988; in Bolivia, Berglur.d-Bl::ÜCher arñ Br:üd:ler, 1976). ihere 18 thus a possibility to finó both the wUd forms arñ sorne traditional cultivars growin:J together. 'lhis situation has been stu:lioo reoently in Olzoo, Peru (DeboUck aro. Tobme, 1988; Debouck et al., in press). ihere, the fa:rmers eat t.he wUd aro. t.he weedy types arñ keep the large grain type "Amarillo Gigante" to be sold on SUnday:raatXet.. In:ieed, beside the wUd aro. the tJ:uly cul.tivated types, a dozen of weedy or intennediate types were also foorrl in the same field aro. scm¡¡times eaten on t.he farm. 'lbat situation arñ the catIIelts made by the fanners loo these authors (Debouck et al., in press) to hypothesize a Itweed-crop CXlIPlex" alter Harlan (1965) aro. Harlan arñ de Wet (1965), whern tha weedy types result from crosses between t.he wUd aro. the cul.tivated 15 forros. similar situation an:l subsequent use by mm llave been dleerved in other crops incllñin;J, patato (Sdlmiediche et al., 1980), maize (Wilkes, 1977) ard dill.i pe¡;p!rS (Pickersgill, 1981) • 'Ihere are thus st:nm;J irxtications that the wild types llave enriched the primitive lardraees, ard apparently t:his is sti1l occurrit:q in 1988 in parts of the rar:qe! As shown in Table 3, lOClSt of the ''wild'' dlaracters are daninant in the progenies of natural hybrids involvin;J cu1tivated an:l wild forros. As pointed out by radizinsky (1985), t:his could lead to the formation of n:me weedy mees on the one him:i an:l to the rejection of UI'I.Wlmted types by traditional f8l."llmS on the other. In the case of beans, one womers, 1'latlever, how st:J:'c:n;J is that rejection, since many traditional lardraces of Pero an:l Bolivia still exhibit a st:J:'c:n;J pod dElhiscence (900 aleo Brücher, 1988). On the other him:i, in one place in OlzOO, Pero, Debouck ard 'l'cilI!e (1988) observed that f8l."llmS consumed weedy types on the fa.."1n. As a concllñin;J remark, one will note that crossin;J between wild ard cu1tivated forros is takin;J place within a restricted geogra¡:túca1 area resultin;J perhaps in a closer m:>rPJologica1 l.i.keness between the lardraces ard their wild ancestors (900 also VaOO.ertlorght, 1986). 'lhis could contr:i.bute to the formation of gene pools too, as we will discuss it below. erossin;J between lardraces in cr outside the rar:qe of wild ancestors could be another way to reduce the foorxier effeot. It can aleo be studied in sec::on::lary oenters. In traditional lardraces of Malawi, natural outorossin;J occurs at a low frequency (arourd 1%, Martin an:l JIdams, 1987), but t:his is high ernlgh to prcduoe recanbinants that can be selected by f8l."llmS. How far does outcrossin;J allow the c:reation of new variability? '!be work by Sprecher (1988b) on the same Malawi oolleotion Wicates that new genetic variability a¡;pears but principally within gene pools an:l ~ little between gene poole (partly due to male sterility an:l other genetic dist.uJ:Panoes) • It \VOOld then be interestin;J to quantify the intercl:oesin;J between gene poole wnere they were in ian times is obt:ained fran ardlaeological ani bioc:hemical stu:iioo, ani stu:iies al cultural contact:s. '1his survey cannot be exhaustive, wi11 only present a few qenerallll:l'lE!ltB1ts. Related ro Mesocmarica: 1) 'Ihere are sane intlcations (Soustelle, 1979) that the 01mec:s had CXItIlIerCial contact:s with ClaXaca ani Guerrero, ani fran their hanelani in Veracruz up ro southwestem Costa Rica (1500-2000 years BP). '1his could explain the diftusion ef sane tropical black beans, :frequent a.roun:l the Gulf ef MeXico ani neighboorÍD;J regions. 2) CUltural oontacts have also been a:rgued a.roun:l 1500 years BP (snarskis, 1985) between Colombia ani Panana. ani Costa Rica. '1his could explain the diftusion ef sane 'B' Ji!aseolin cultivars, eriginally fran Boyaca, Colanbia te Costa Ricá ani even Guaten'ala (Gepts ani Bliss, 1986; Koenig et al., in press) • on the reverse, the sama contact:s could explain the presenoe ef '$' cultivars in Southeastem Brazil. 3) Tributes i?f'yed ro the Aztecs would have presentad another OWOrtunity fer exchar:}oo ef large amounts of soods in prehispanic !SE S oamerica (Torres, (985). 'lbe dcminant rule of this group -would llave n contributed on a large scale to the diffusion of •S ' ¡:iJaseolin cultivars in Mesoamerica (see Gepts et al., 1986). Rel.ated to Soot:h America: 1} 'llle presence of a sieva type of lima bean in the Uf.PElIs levels at Huaca Prieta (nort:hem coast of Pe:ru) (Towle, 1961) indicates cultural contacta at aboot 2500 years BP between that part of Pe:ru an:l Colanbia or pe:dJaps MesoanJilrica mere those beans originated (pickersgill an:i Heiser, 1978). 2) 'llle presence of large seeded lima beans in Norcinti, Bolivia (Debouck, 1988a) outside the ~ of distril:ution of their wild relatives (Debouck et al., 1987) wwld indicate an introduc:tion there an:l thus cultural contact:s between southern Bolivia an:l westem Pe:ru. 3} 'llle presence of 'T' P'laseolin cultivars in southern Colonbia (Nariño; mere wild beans are not known) would indicate int:roduction fran the southern Andes mere these types of ¡:iJaseolin are UJI'.UOll both in the wild forms an:l landraoes (Gepts et al., 1986: Tol:me an:l Debouck, in presa) • Little infornation is available aboot the geneti.c dla.rges oocurred in the roaterials during these millenia. As jOOged on seed characters, roany grain colora were already selected incltñing yellow (Kaplan, 1980). By 800 years B.P., at Ocanpo, Tamaulipas, bush types as well as a satEWhat snap bean are already present (Kaplan an:l Me Neish, 1960). 3.2 Travels within the Arnericas afier Colmnbus relatfrl to Mesoarnerica. 1) A lllOVeroent already perceptible durinJ the last precolumbian perlad will continue in the Greater Southwest with nm:e beans enterinJ that part of the USA (carter, 1945: Kaplan, 1956). 18 2) Soon after the Spanish Conquest, there was an active excharge of gernplasm between Mexico and Peru (E. Hemardez X., pers. catm.). 'Ibis could explain the preserx::e of 'T' ¡iJaseolin cultivars in Mexico sudl as oertain "cacahuat.es" or "Japones". In the ClfPJSite direction there is the small seeded whi.te variety in Peru oaUed "Panamite" with 'S' J;ilaseoJin (Gepts, 1984). 3) Haiti, part of Hispaniola, servin:;J as a stcp fer the spanish galleons on their way te Eurt:pe, :te !dile soon a secon.:1ary oenter of diversity, with a special en¡ilasis on earliness CC. Messiaen, pers. catm.), resulti.rq in the "R:mpadour" grtXlp with 'T' ¡iJaseolin (Gepts et al., 1988). Again little infonnation is available. One can hypot:hesize that as the beans were moved towaros higher latitudes especially in the northem h~, sane selection was nade for day neutral cultivars, Le. able te bloom umer lon;¡er days. AIso pert¡aps for bush types, as the growin:;J season l:leclcIoos shorter at higher latitudes. 3.3. l'ravels frgn and ootside the Amerirna aftel:" ColU!!bJS. Columbus discovered central Ainerioa in 1492. IA3SS than 60 years later c:x::tlilUl bean was gIWn ex:tensively in westem EUrope (I.eón, 1987; Briíd:ler, 1988). Fran there, it was distr:i.buted nuch further: saltheastern Europe and the Middle East, Iran and Irxtia, other places in Asia, Afrioa and back te the l\merioas. Iet us illustrate a few cases. 1) Most of the varieties in Europe CCIlE! fran the swthem Anies displayin:;J 'T' and 'C' phasolins (Gepts and Bliss, 1988). 'Ibis raises the followin:;J questions: - Were they better adaptated ro lon;¡er ¡tl.otoperiods arrljor cooler tenperatures? - oid early travelers prefer them for their laJ:qer grain size? Was there a higher mutation rate for snap bean characteristics in 19 these gene pools" 2) SeVeral Ellropean varieties were brought back to the AIIerlcas, particularly the USA a:OO Argentina. SeVeral specific cases are well documented (Gept:s et al., 1988). 3) Beans were also :introduaed sllooessfully in Africa, both d.i.rectJ.y frau Brazil (Evans, 1976) a:OO indirectly frau the soot:hel:n Ardes throJgh Europa since the An:'lean types of ¡;ilaseolin are daninant in Africa (Gept:s a:OO Bliss, 1988). Because of the ext:ension of the geograpbic ranga, day neutral cultivara were developed. Also to mention is the awearance of several recessive charact:ars (waxy pods, yellow pods,ve:ry lcn::J pods) aooonling to specific salaction pressures in the secoOOary oenter. Part 2. Inplications for Breeders In the first part, we llave presant:ed three gra:¡ps of facta about the origin of the tumOll bean, 1'. vulgaris, Vlhich are: origina of the c:amon bean are 1I1lltiple a:OO the different danestications about 10.000 yaars ago were than imeperrlant• .D:::snestication was not neutral towa.I1:is genetic divarsity included into the danesticated stoc:ks: a founler effect took placa with :regional differances, also with differances in int:ensity. Genetic dlanges at a few loei resultirg in new charact:ars of high antJ:ll:qlÍc significance (related to seed, pod, growh habit) ~ early in bean crq:¡ history. 20 Iet us I'ICIW' d;sruss sane of their consequenoes. 1. Gene txlOls formation aro in:x:l!patibíli.t;y prOOlens Irrleperdent daoostications usin;J nm:r;ilo1oqically different wild pc:p¡lations loo te formation of m:qiloloqically distinct groups of cultivated genotypes. As we llave seen abolle aro as has been d:lserved in other crops (Harlan aro de Wet, 1965; Johns aro Keen, 1986; wilkes, 1977), there is a stron;J possibUity that crosses between wild am,tor weedy pcpulations aro early cultivan; contributed te the li.keness between these two. MultUocal danestications had other consequences: ¡;tlysical isolation of groups of early cultivan; fran each other for t:h.wsands of years (2000-4000 years as can be deduced fran the above menUoned figures) loo te their geneUc isolation aro the formation of gene pools. If so, one should experience prdllens of genetic incarpatibílity 'When c:cossin;J between gene pools aro first of all between regions of origino In Table 5, sane of these crosses are reported. Although this survey is far fran bein;J conplete, it a¡:pears that c:cossin;J between small seeded Mesoamerican cultivan; aro large seeded southern Arrlean cultivan; is likely to faH with little or no trile recx:anbination. As we llave seen above an1 in Sprecher (1988b), in Africa, While crosses occur, genetic barriera prevent the formation of recx:anbinants between the gene pools brought there an1 planted toqether. In crosses investigated so far, F1 hybrid weakness is a¡;:parently controllOO by 2 conplementary genes: DLl (for dosage depen:ient lethal, Shii et al., 1980) attributed te Mesoamerica aro DL2 te the southern An:ies. 'lhe F1 hybrids DLl dll DL2 dl2 give cripplOO plants at high tent;Jeratures, lesa abnormal planta at lower tent;Jeratures. The fact that sane crosses between 'S' an1 'T' phaseolins in both cultivated an1 wild parents can give a normal offsprin;J would discard a direct linkage between the genes codirY;r for phaseolin aro the expression of Fl weakness. But, on the reversa, DLl 21 an:i DL2 are respectively unsepa:rable fran certain ¡;ilaseolin types. 'lbís was irrlil:ectl.y evidenced by Sprecher (l988a) en a small sanple incluii.n;J the starrlard lines for ¡;ilaseolin types an:i 27 lines producing Fl hybrid weakness. 'Ihese materials displayed altemate isozyme profiles bebNeen the Mesoamerican (wiht 'S' an:i 'B' ¡;ilaseolins) an:i the AlxJean gene pools (with 'T', 'e', 'A' an:i 'H' ¡;ilaseolins). Now the fact that criR;>led plants can also be observed in crosses with 'S' arrl 'B' ¡;ilaseolin types would indicate that there would be llDre genes responsible for Fl weakness. Although OOviwsly llDre data are needed, it would be indicativa of an incipient separation of the Colanbian materials with 'B' fran the Mesoamerican stocks, despite their s:imil.arity in isozym:¡ profiles (Sdrlnkel et al., 1988). Once llDre, we need l!Dre data aJ:x::ut crossability prablems in beans, also because of its practical cx:nsequences. But there is increasing evidence that the ext:reIres of the ran;¡e when cultivated are not always easy te cross. It wou1d be of interest te test the fo1lowing hypothesis: Are there cases of incottpatibility in 'T' ¡;ilaseolin types? Are there other cases of incottpatibility between gern.úne Mesoamerican an:i Colanbian cultivars? Is the incottpatibility a result of danestication? Although the type of ¡;ilaseolin can be very useful in predicting sucoessful crosses, it !loes not constitute a definitiva insurance. It would be therefore useful to develop other (llDlecular?) markers. On the other haIXi, as theSe hypotheses are progr ssively examined, it wlll be possible to increase the variability by crossing bebNeen gene pools. Although difficult (sea aboV'e), it could be rewarding on the Ion;)" run, as indicated by the larger ntm1ber of loci 1Nhere polyxrorphism is expressed in snap beans (Weeden, 1984: Sprecher, 1988a). 22 2. Are we facim a case of co=eyp1ution betvieen the c:rgp arrl its biotic arrl abiotic envirornnent? If as we hypot:hesbe above early lan:haoes were daDesticated in isolate plots arrl were sepa.rated fran each other for t:housa:rrls of years, then there would be st.rorg cases of co-evolution l:ieb.veen these larrlraces arrl fue agents eithar biotic or abiotic present in their environment. In other words, one shoold firxi st:J:OII~ correlations between sane particular lan:haoes arrl sane strains of diseases, sane elilnatic si:xesses, etc. 'Ibis approach : in I:leans, su;¡gested by Gepts arrl miss (1985), has just started arrl evi.dEinces are aCCUlllllatirg. We will considar two cases: diseases arrl RhizobiUlll. stavely (1984 arrl 1988) reported about fue variability of rust in the US, shcMirq the d1fferential. susceptibility of bean cultivare to rust :races: sane snap bean cultivare highly susceptible to US :races reacted in the sama way, suggest:i.rq a 1tind of specific host-pat:hcqen reaction. a:mpuesto Negro Clrlmalte.nan;Jo, originally a mixture of black I:leans fran Guatemala, shcMed the highest level of resistance to a large series of mees (stavely, 1986), perhaps because of its larger genetical basis. Iater on, other lines were tested (Stavely, 1988) arrl PI 181996 arrl 189013 bot:h fran Guatemala were also highly resistan!: to US:races. Another interestirg cultivar Ecuador 299 has a 'S I ¡tJaseolin, as does carp.¡esto Negro Clrlmalt:enan:Jo (P. Gepts, pare. comm.). After a electrq;iloretical study on 55 isolates of angular lea! spot, Correa (1987) foun:i that the resultirq two patterns of four enzyIOOS cwl.d be associated with the 1tind of beans they infect, one pattern be:i.rq foun:i on large seeded types arrl the other be:i.rq fomrl on small seeded types. A confimation is needed to be sure that one pattern is foun:i only on large seeded materials in a specific ~c zone arrl vice versa. rut fuese two examples indicate that pathogen variability is not distrihrt:ed at rarrlom. In this case, many agronomical practicas arrl screening methods 23 should be revised. Ren!1Íe aro Kenp (1983a, b) showed that when testinq severa! bean cultivars for their efficien:::y in nitrogen fixatian usirg a sirgle strain of Rhizobitnn, fue effect of cultivars was significant, as was the interaction strain-cultivars. If these resulta prove' true on wre materials (wild beans for .instanc:e), nitrogen fixation could be :i.nproved by dloosirg the right c:ounteqlarts. 3. Practica! consegue.noes for breedin;r aro gernplasm management '!he aboIre mentioned resulta, although fragmentary in S. lIner. Antiquity 27(3):356--369. Brudler, H. 1954. Atgentinien, urtJ.eimat unserer bd:men. l.Jmsctlau in Wiss. u Tedmik 54(1) :14-15. Brucher, H. 1968. Die evolution der gartenl::d:me Blaseolus vulgaris L. aus der sudamerikanisdlen wild l::dme 1!. aboriaineus Burk. ArqM. Botanik. 42{3-4):119-128. Brucher, H. 1988. '!he wild ancestor of lbaseolus vulgaris in South lInerica. p. 185-214. In P. Gepts (ed.). Genetic resources of Blaseolus beans. Kluwer Academic l?I.1blishers, [):¡rdrecht, Hollaro.. Brunner, B.R. aro. J.S. Beaver. 1988. Estimation of rutcrossing of dry beans in Puerto Riex> Ann. Rept. Bean Inpro\IeIIent COc.p. 31:42-42. rurkart, A. 1941. Sobre la exisc:encia de razas silvestres de "lbaseolus vulgaris y "Blaseolus lunatus" en el Norte Al:gentino. Resoluc. y ResUme.nes Botánica, Primara Reunión Atgentina de l\groncm1a, Buenos Aires: 52. Burkart, A. aro. H. Brudler. 1953. Blaseolus aborigineus Illrkart, die llllbnabliche aniine stammfonn der Kultur Bdme. Der Zud1ter 23(3):65-72. carter, a.F. 1945. Plant geograrby aro. culture history in the lInerican Southwest. Vik:ing Fund l?I.1blications in Anthropology 5:1-140. Correa, V.F.J. 1987. Pathogenic variation, production of toxic netabolites, aro. i~ analysis in BlaeoiAAriopsis griseola (Sacc.) Ferr. lb.D thesis, Michigan state university, East Lansing Michigan, USA, 154 W. COyne, O.P. 1965. A genetic study of "crippled" IIDqilology resembling virus synptoms in Blaseolus vulgaris L. J. Hered. 56 (4) :162. 26 Crispin, M. 1960. cruzamiento natural en el frijol. Agricultura Técnica México 11:38-39. Debcuck, D.G. 1985. Trip report to Peru am Al:gentina. Int:em. Board tor Plant Genetic Fesources, Ralle, ltaly .l\GrojIBroR, 85j123:3lp. Debcuck, D.G. 1986a. Ibaseolus gernplasm collection in west:ern Guatanala, C.A. Intemational Board for Plant Genetic Resan:oes, Ralle, AGEG/IBroR, 86/40:3Op. Debcuck, D.G. 1986b. BIaseolus gernplasm collection in cajamarca. am Amazonas, Peru. Intemational Board for Plant Genetic Fesources, Ralle, ltaly, .l\Gro/IBroR, 86j161:38p. Debcuck, D.G. 1987. Recolección de gerJlq)lasma de Rlaseolus en el C'altro Y Cent.ro-Sur del Perú. Intemational Board for Plant Genetic Resan:oes, Rome, ltaly, AGrojIBroR, 87/112: 36p. Debcuck, D.G. 198sa. Recolección de Gernoplasma de RlaseolUS en Bolivia. C'altro Intemacional de Agricultura Tropical, cali, Colati>ia, ~, 24 pp. Debcuck, D.G. 1988b. Rlaseolus gemplasm explo:r:ation. p. 3-29. In P. Gepts (ed.). Genetic resouroes of Blaseolus béans. Kluwer Academic rublishers, Dordrecht, Hollam, p.3-29. Debcuck, D.G., R. Araya, W.G. Gonzalez am J. Td:Jme. 1988. Presencia de formas silvestres de Rlaseolus VUlgaris L. en Costa Rica. Actas XXXIV ReUnión AnUal Prograna cent:roa:mericano para el Mejoramiento de CUltivos Alimenticios, PCOIlCA, San José, Costa Rica, 21-25 marzo 1988, p. 75. Debcuck, D.G., F.M. Gamarra, A.V. Ortiz am J. Td:Jme. Presenoe of a wild-weed-crop calP1ex in Ibaseolus VUlgaris L. in Peru? Am. Rept. Sean I:nprovement Coop., su1:tnitted. Debcuck, D.G., J.H. Liiían Jara, S.A. canpana am J.H. De la cruz Rojas. 1987. ObseIvations on the danestication ef Ibaseolus lunatus L. Plant Genetic Resources Newsl. 7Q: 26-32. Debcuck, D.G. am J. Td:Jme. 1988. Recolección de Gernoplasma de Blaseolus en el Centro-SUr del Perú. Centro Internacional de Agricultura Tropical, Cali, Colombia, Mi.meografe, 23 pp. 27 Delgado, S.A., A. Bonet aro P. Gept:s. 1988. '!he wild relative of Bl.aseolus vulgaris in Middle Anv:lrica •. p. 163-184. In P. Gept:s (Ed.). Genetic resouroes of Bl.aseolus beans. Kluwer Academic l'I.lbli.shers, Dordrecht, Hollaro. En;Jel, F.A. 1963. Illtations a Paide du radio-carlJane 14, et probleoos de la préhistoire du Pé'rou. J. SOCo Am!ricanistes 52:101-132. Evans, A.M. 1976. Beans RIaseolus SW. (Ieguminosae- Papilionatae). p. 168-172. In N.W. s:i.ntoords (ed.). Evolution of crq:¡ plants. Iorgman, IDndon, New York. Ford, R. 1985. Patterns of prehistoric foad production in north Anv:lrica. p. 341-364. In R.l. Ford (ed.). Prehistoric foad production in north Atoorica. univ. Middgan, Mus. An1:lu:qlol. Papers, No. 75. Gentry, H.S. 1969. origin of the CUiiOCll1 !:lean, ¡haseolus vulgaris. Econ. Bot. 23(1);55-69. Gept:s, P.L. 1984. Nutritional aro evolutionary ilIplications of ¡;tlaseolin seEd protein variability in CCiI1Irol1!:lean (Bl.aseolus vulgaris L.) Eh.D. thesis, univ. wisoonsin, Madison, USA, 209 };:p. Gept:s P. 1988. RIaseolin as an evolutionary mrker. p. 215-241. In P. Gept:s. (Ed.). Genetic resouroes of Bl.aseolus beans. Kluwer Academic l'I.lblishers, D:>rdrec.ht, Holland. Gept:s, P. aro F.A. Bliss. 1985. Fl hybrid 'Weakness in the CUtUilJll bean; differential geographic origin suggest two gene pools in cultivated !:lean gernplasm. J. Hered. 76;447-450. Gept:s, P.L. aro F .A. Bliss. 1986. RIaseolin variability anDng wild aro cultivated COImPIl beans (Bl.aseolus vulgaris) fran Colanbia. Econ. Bot. 40(4):469-478. Gept:s, P. aro F.A. Bliss. 1988. Dissemination pathways of 100rs and pod characters in 2»&'91us vulgaris L. Genetica 16:177-294. Rennie, R.J. and G.A. :Kenpt. 1983a. N2-fiJaltion in field beans quantified by ~ isotcpe di1ution. (1) Effect of strains of llhizchlum tilaseoli . .I\gron. J. 75:640-644. Rennie, R.J. and C.A. KeIIp. 1983b. N -fiJaltion in field beans quantified 2 by ~ isotcpe dilution. (2) Effect of cultivars of beans. .I\gron. J. 75:645-649 Roas, E.E. and M.S. centner. 1984. llections. Ann. Rept. Sean Improvement coop. 27:68-70. Sdlinkel, C., S.P. Singh and P. Gepts. 1988. Pat.te:ms of genetic diversity in BIaseo1us beans. Ann. Rept. Sean Illprovement Cocp. 31:40-41- Schm.iediche, P.E., J .G. Hawkes and C.M. Oc:hoa. 1980. Breedirq of the cultivated patato species So1anum X juzepczukii 8.Ik. and So1anum X curti1ob.nn Juz. et 'alk. l. A study of the natural variation of .2. X iuzepczukii, Q. X curtilob.nn and their wild progenitor, Q. acaule Bitt .. Eu¡;hytica 29:685-704. Sdloonhoven, A.v. and C. cardona. 1982. I.ow levels of resistance to the nexican bean weevil in dJ::y beans. J. EcDn. EI'rt:o!;¡pl. 75:567-569. Sdloonhoven, A. v., c. cardona and J. Valor. 1983. Resistance to the Sean and the Mexican Sean weevil (O:llec::ptera: Brud:lidae) in non cultivated CUiiliOll bean aClCtsions. J. EcDn. Er'It:l. 76(6) :1255-1259. Shii, C.T., M.C. Mok, S.R. Temple and D.W.S. Mok. 1980. Expression of developrental ¡ bnormalitit.es in hubrids of 8Jaseolus vuJ teris L. J. Hered. 71(4):2:8-222. 32 Sin;Jh, S.P. and J .A. Gutieriez. 1984. Geograj;hlcal distr.i.bution of the D~ and ~ genes causin;:¡ hybrid dwarfism in 1ibaseolus vu]9ªris L., their association with seed siza, and their significaIXJe to bxeedi.rJ;J. ~ca 33:337-345. Snarskis, M.J. 1985. '!he o::qarative icono!Jla¡iJ.y of metal work. and other media in prec:olumbian Costa Rica. p. 121-136. In Precolumbian American Metallurgy, 45th International Ol!:gl ss of Americanists, Banoo de la Replblica, Bogota. SWstelle, J. 1979. les Olmeques. Arthaw, Paris, 221 p. Sprecher, S.L. 198sa. lsozyme gekJtype diffmeIkAS between the large seeda! and smallseeded gene pools in Blaseolus vulgaris L. Ann. Rept. Bean Dtprovement coop. 31:36-37. Sprecher, S.L. 1988b. Allozyme differenUation between gene pools in COllilWh bean (Blaseolus vulgaris L. ) , with special reference to Malawian germplasm. lb.D. thesis, Michigan state University, East Lansin;:¡ Michigan, USA, 207 p. Stavely, J .R. 1984. Pathcgenic specialization in Urgnyces ¡:tIaseoli in the United states and rust resistance in beans. Plant disease 68(2) :95-99. stavely, J.R. 1986. Pathogenic variabUity, resistanc:e ~, and progresa t:owards developin;:¡ atable resistance to bean rusto Ann. Rept. Bean D~ Coop. 29:24-25. stave1y, J.R. 1918. Rust resistanc:e in beans: the plant intro:iuction oollection al' a resouroe and resistanc:e developnent. Ann. Rept. Bean Dtprovement (tlop. 31: 64-65 Stoetzer, H.A.!. 1984. Natural cross-pollination in bean in Ethiopia. Ann. Rept. B an Dtprovem. Coop. 27:99-100. Tarrago, M.N. 19: lO. El proceso de agriculturizaciÓll en el noroeste argentino, lona valliserrana. p. 181-217. In. Actas de V 0l!:t;JIE!S<.l Nacional de ArqUeología Argentina, TClIW 1. Universidad Nacional de San Juán, Instituto de Investigaciones ArqUeológicas y Museo, San Juan, Arger tina. 33 ToI:mle, J. arx:l. D.G. Debouck. Towards a broader area of daoostication of the O:;¡,dlLlll bean, :RIaseolus vulgaris L., in the sout:he:m Andes? lInn. Rept. Bean l'lrpr\::NeIIeI O:lop., subnitted. Torres, B.W. 1985. ras plantas útiles en el Méxicx> antiguo según las fuentes del siglo XVI. p. 53-128. In T. Rojas Rubiela arx:l. W.T. san::lers (eds.). Historia de la agricultura, época prehlspánica siglo XVI. Vol 1, Instituto Nacional de AIIlJ:tpJl.ogia e Historia, México, DF, México. 'J.Wle, M.A. 1961. 'llle EI:hoobotany of Pre-a:>1Ullbian Pero. viJd.rg F\m:i Publ. ~. 30:1-180. TIlclrght, T. 1982. Seed increase arx:l. evaluation of the wild !ihaseolus vulgaris gen¡plasm. centro Internacional de Jlgricultura Tropical, cali, Colatbia, m:i.meogra¡;ned: 69 pp. Vanierllorght, T. 1986. L'étude de la variabilité dlez le haricot cxmmm (!ihaseolus vulgaris) par l'utilisation de méthodes statistiques multivariées appliquées a une banque de données. 'Ihese de D::lctorat, Faculté des Seiences Agronaniques de l'Etat, Gembloux, Belgique, 229p. Wall, J.R. arx:l. S.W. Wall. 1975. Isozyme pol~ in the stWy of evolution in the JitJaseolus yulgaris - 1'. cxx:cineus cx:mplex in ~co. p. 287-305. In C.L. Markert (oo.). Isozymes, Vol. 1, Genetics arx:l. EVolution. Academic Press, New York. i1e eclen, N. F. 1984. oist.in;¡uishing amc:n;J white seeded bean cultivars by means of al10zyne genotypes. El.:q::hytica 33:199-208. Wells, W.C., W.H. Isan arx:l. J.G. waines. 1988. outcrossing rates of six ccmnon bean linee. CI:'op Sei. 28:177-178. wilkes, H.G. 1977. Hybridization of maize arx:l. teosinte in ~co arx:l. Guat:errala arx:l. the inprovement of maize. Econ. Bot. 31:254-293. 34 'rabIe l. Sane 1OO:r:¡ho-¡:ilysiologica1 characters of wUd l? vulgaris obseIved in Palmira arxi Popayan, CoIalilia. 100 Seed HypocotyI BracteoIe D!ys te Origin toeight texture sizejshape flowering (g) Mexico 5-8 Lignified Iatge, avate 40-50 QJaternala 6-9 Lignified Iatge, orbicular 50-65 Costa Rica 5-6 Hel:baceous? Iatge, avate >100 CoIalilia 8-12 Lignified Iatge, avate 65-75 Peru Nort:he:m 15-16 Lignified? Medium te Iarge, 55-65 avate Southe:rn 10-16 Hel:bacea.!S Iatge, lanceoIate 45-55 Bolivia 9-12 Herbaoeoos Iatge, Ianceolate 40-50 & small, trianJular AI1;¡entina 12-16 Herbaceous SmalI, trianJular 35-40 'rabIe 2. Rlaseolin types of wild pcpllations arxi Iarxiraoes of~. vulgaris fra¡¡ diffe.r:ent latin Anerican cnmtries. wild ¡;q;w.ations Mexico M, S S, Sd, T 1, 4 QJatenaIa M, S S, S 1, 4 costa Rica M S, S 3, 1, 4 CoIanbia 01, B S, T, e, B 2, 4 Peru North 1 T, e, H, s 1, 4 Center T,e T, e, H, A, S 5, 1 SOUth T,e,K T, e, H, s 5, 1 Bolivia T T, e, s 5, 1 AI1;¡entina (NW) T, H, e, J T, H 4, 1 Sources: 1. Gepts et al., 1986 2. Gepts arxi Bliss, 1986 3. Debouck, et al., 1988 4. Koenig et al., in press 5. Tahme arxi Debouck, in press 35 Table 3. Sane mxphologica1 traits of high anthropic significance arrl their heritabilty. Trait Nl:mi:ler of Loei Inheritance Reference Dehiscent:/oon 2or3 Daninant/recessive Prakken, 1934 ¡ dehiscent pcñ I.sakay, 1988 ColoredjpJre whit.e seed 9 Daninant/recessive I.sakay, 1988 seed I:ndet.erminate¡detel:minate 1 Daninant/recessive Hortal, 1915; grcMth habit miss, 1971 Interncde len;¡th: lon:y 1 Daninant/recessive I.sakay, 1988 sho:rt lower int:eJ:nodes ~ Table 4. Estimated c:utcrossing rates for a.tltivated F. vulgaris Place Rate e%) Be:r:keley, California 0.7 Mack1e ar.d Smith, 1935 O!Ivis, California 0-0.007 'l\lcker ar.d Hard..i.rY:J, 1975 Irvine, California 9-69 wells et al., 1988 Lilorgwe, Malawi 0.8 Martin ar.d 1\dams, 1987 Awassa, Eth1cpia 3-5 stoetzer, 1984 several. placas, Mexico 1-4 crispin, 1960 Cbap:l.n;¡o, Mexico 1-3 MirarXIa (blin, 1971 Mayaguez, ruerto Rico 0.04-17.6 B:r:unner ar.d Beaver, 1988 37 Tabla 5. San>e crosses of cultivated ,f. yulgaris from differen!; origins an::i problems encount:ered. origins I:baseolin 5ympt:ansj Referenc:e types abnormalities south~ican TxS crippled plants Coyne, 1965 via Kenya via USA Guatema1 ajBolivia SxT Chlorcsis an::i Sh.ii et al, 1980 an::i its reciprocal lethality south Andea:ly'Mesoamican TxS J::lW'arf plants singb. an::i Gutlérrez, 1984 via 'lUr:Key via BrazU .... 00 south An:'IeanjMesoalcan TxS Absence of roots Gepts an::i Bliss, 1985 via USA via BrazU (also for a review) GuatemalajChlle SxC cripples Temple, 1977 (unpublished) costa Rioa,Isouth Alñean SxT criW1es Temple, 1977 (unpublished) via Haiti Mex:ioojCOlanbia SxB cripples Temple, 1977 (unpublished) Figures Fig. 1. Dist:ribution of seed accessions of wild Blaseolus vulgarls L. as of 1988. Fig. 2. A:rdlaeloqical fimings of Blaseolus yulgariS L.: plaoes am years before presento Fig. 3. :Alaseolin types in wild Blaseolus vulgarls L. as of 1988. 39 1200 lOS" 900 75° 600 45~ ° ,N \ _. UNITEO STATES ./ 300, '\:-k---'-- -"" ( 300 ~ ____T !.'!'! .. ~_ ____~ ~f~~v:~\~ ~-r-'- -- \) _______ .:. ________________ ._ _ . _______________________ _ ... ,:.. 7 CUBA~::. .DDMICAN O\ .""'''-........- ,1 cosTA RICA - Ir- (1( "'i' PANAMA \,_.:::::'e:ztJELA ! ~ .&. ' ) t CTOCVMBIA<' ."- .- I\"r ) , ,... .. 1.,r,) " .. .::. ... "" ... ' 00f .. "'" - ".. 00 ~ ..E /c uÁIb ciR_.'" i JI Ir-,