INDIGENOUS AGROFORESTRY IN THE PERUVIAN AMAZON: BORA INDIAN MANAGEMENTr OF SNIDDEN FALLOWS WILLIAM M. DENEVAN, JOIN M. TREACY, JANIS B. ALCORN CHRISTINE JULIE DENSLOW PADOCH, and SALVADOR FLORES PAITAN

n recent years students nagement of tropical forest of Amazonia resurces: 1. farmers, but have ema- the diverse, multistoried rarely among colonist phasized swidden (shift- farmers. that some of ing cultivation However, it has received little at. the most successful food field) which protects the tention; producing adap- soil and allows brief mentions include: Denevan tations to the rain forest for habitat recovery (1971: habitat have under long fallow 508-509) for the Campa in been those of the (e.g., Conklin, 1957; eastern indigenous tribes, and Harris, Peru; Pose', 1982; 1983: 244. that consequently 1971); 2. the house g:arden, or we have much to learn dooryard 245) for the Kayap6 in central from these garden, also diverse and mul- Brazil, "ecosystem" people. "Refined Basso (1973: 34-35) for the Kalapalo over tistoried, but with :) large complement in millennia, Amazon Indian agri- of central Brazil; Eden (1980) culture tree crops ind with soil additives for the preserves the soils and from Andoke and Witoto the househoid retise, ash, and in the Colombian ecosystem ... If manure (e.g., Amazon; Smole the knowledge of in- Covich and (1976: 152-156) and digenous peoples can Nic-tcrson, 1966); and 3. the for he integrated with planting, protection, Harris (1971: 4S7, 489) and To. modern technological know-how, and harvesting of rres Espinoza then a trail side and campsite (19 0) for the Shuar new path for ecologically vegetation ("no- in eastern sound develop- medic aQricullrc"' Ecuador. L')me observers have or "forest field,"), assumed that all tha is involved is a ment of the Amazon will have been involNin- wild. semi-domesticated and return to abandoned found"(For similar (Pose, statements 1982: 18; 19S3: 225), domesticated ,widdens to search for other tropical 1983: 241-243).plants (e.g., Pose\ 1982; for residual regions see, for example, A related type of plant cultivation, crops left from the former Nigh and Na- management but indications are that tions, 1980; is the manipulation and ac­ Clarke, 1977; Eckholm, utilization tual management occurs, including 1982: of swidden fallows, a form plant. 34-35; and Klec, 1980). In par- of ing and protection as well agroforestry involving a combination as utilization ticular, Indian cultivation is of of certain useful characteriz- annual crops, perennial wild plants that appear ed by multiple cropping tree crops, and at various and interaction natural forest stages of fallow succession. with natural vegetation. regrowth. Swidden-fallo- The purpose of this Attention mlanage- paper is to examine has been di- ment apparenly the swidden fallows rected to several forms is widesprL among of an Amazon of traditional ma- Amazon tribes native group, the Bora of and some local mestizo eastern Peru, with the objective of de-

Wisconsin, William M. Denevan, Madison. His Ph. D. is from a cultural ecologist, is Professor the 'njhersitv of Cs!ifornia, Berkeley of Geography at the University and peasant agriculture in the in 1963. lie has carr;ed out of Amazon and the Anties in Peru, research on Indian Bolivia, Venezuela, and Ecuador. Geograph, University of Wisconsin, Address: Department of Madison, Wisconsin, U.S.A. 53706. John M. lie has ;rvious field experience Treacy is a doctoral student in in the Permian Andes geography at the University of Wisconsin, and in the Guayas Basin of Ecuador Madison. research in agricultural and is currently engaged in terracing in southern Peru. Janis of B. Alcorn, an etlinobotanist, received Texas, Austin. She has carried a Ph. D. in botany in 1982 out extensive field research on from the University Huasleca Indian plant use in Mexico. Christine Padoch received a Ph. D. in anthropology specialit from Columbia University h is land use, settlement, and demography, in 1978. 1hr and the particularly in South East Asia where Luin Daeh of East Kalimantan. She she has studied the Iban of Sarallak is currenily at the Instilutle of Economic Julie Denslow has a Bolany, New York Botanical ha, (ne l'I. I). in botany from the Unihersity Garden. field research on seconcla'.v successiun of Wisconsin, Madison. in ('olumhia and Costa Rica. She de la Amazonia Sid]-vador Flores Pait~n, Ingeniero Peruana in hluitos. le holds Agr6nomo, is Profesor de Suelos, Costa Rica, 1977. the degree of Magister Scientiae Ciencias Universidad Nacional Ile has been en-aged in research Agricolas y Forestales, Turrialhl, on soils, forestry, and agroforestrn in Iquitos and in 0'. rgion since 1972. !1o AnipilaCLI

346 378 0 -1844/84,'06-34-12 $ 03.00/0 NOV-DEC 1984, VOL. 9 NQ 6 )RIER'ELI1 fon tratine howk field, are grad l:, Noeri!"cr, but with the driest month Secondtry foreqt due to shifting ciii­ abandon>d This contrxts ,ith mot (August still having 133 millimeters. tivation since the Bora arriked there 50 studies of shifting cultivation whch focus Temperatures average around 26:C years ago. Hoever, high foret hecins on why 3 7 fields are abinlo;-ed, and which throughout tileyear (oINERN, 1976: ). at 20 minu'es walking distance from present a sharp distinction bctw cen the Brillo Nue,,o is situated beside ar. oxbow the village and extends north%%ard (s' field iden) and the abandoned field lake formed by the Yaguasyacu The across the Coloinb.t ilborder. (fallow). B For the ora there is no clear area is a hilly, dissected fluViei terrace Family fields are dis­ transition between swidden and fallow, interlaced with numerous seasonal pcrsed throughout the fore,,t siurround­ but rather a cont in uun, from a swidden streams. The soils are primarily deep itig the communal ,nloco (residence of dominated by cultivated plants to an old Lltisols (paleudults, tropohumults, and the vil ge tcraca or ceremonial head). fallow coisposed entirely of natural ve- tropudults). They itclt,c and yelliw, Fields atre often closely clustered be­ getation. Thirty-five years of more rnav clay soils, red and sandy soils, cause fa rmrers find it convenient to Visit be required before the latter coadition and glc' soils (tropaquods) in depres- several on one trip. Most plots are ac­ prevails. Abandonment is not a moment sion:,;. -he Blora prefer to farm the clay in time but rather a process overAgrofoestryisc(Gent-h979..thetinm,. soils and red sandy soils cessiblemaloca; within others15 minutesarc acrosson footthe fronYa- Agroforestrguasyacu ly and are reached by dugout receiving considerable attention as a There were J53 fa- an d ar ad by ae. Both prinmary andseodr potentiailly stable and ecologically viable milies living in the settlement. All are , widdens. Primary Cdescendents of tribal groups brouight forest are cleared form of tropical forest land use (King to t ribalro ps brg forest for sidn.Pinr and Chandlerto sites are recognized as more the mpacu rom the Igaraparan- fertile, while secondary forest (fallow) adwski,1981;dles, 1979; H t 9; Caquet:i reion of Colombia follow ing doski, 1981; SaIlas, 1979; Flart, 19S0; Peru's loss of a border war with that isclol,er at hand and more easily flled. Spurgeon, 1980). One of the major re­ comimendations of the recent U.S. Na- country in 193y4. They wcre resetled abou t years of age. There is bo­ tioaal Research Council (1982: 4, 5, on land eventually granted to themi by anical evidence, however, of secondary 146) report on tropical development is they retain community tit. (hice fort over -1)years of age. (Presence that the agroforestly systems of itn- studv was undertaken at Brillo Nuev0, of buried and surface postsherds i­ digenous people should be studied and rither than with a community Iong dicates previots occupation of the arva ri-corded before such knowledge is lo,,t. thed i t ha commuse of at unknown trines by unknown Indiani Vve believe that certain f'ature; of- Bo - ~established in its habitat. becMise Otfaresfarmers. swidden-fallow management can be in- by project member Sarador Flores.) The Blora say that a cor|,o~ltedl into systematic models o( The lBra are gradually tropical agroforestry beini as:.iiiiat- nlinitnutIi of ten y.iears of fallowing is systems. Indeed, an ed into Peruvian society throu-h . needed bef'ore a plot canl be cut and examination of Bora land use indicates sionaries, commerce, and access to planted anew. Most swidde;,s, howe,,er, that "agroforestry" is new in name only Pebas, Iquitos, and Pucallpa. Bora appear to be prepared from fallows 20 to native groups in the Amazon. Under villagers speak Spanish, wear mariu- years of age or older. For the Bora, denser populations in the past (Denevan. factured clothing, and market handicraft one indicator of a fallow ready to be 1976), large areas of Amazon fore,,t items and lumber. Bora subsistence, felled and cropped is a lack of shrub­ may actually have been ,tages ofCpro- however, retairns nrany of its traditioril by growth near ground level. ductile swidden fallows. sl,:ole hio:ic elements, wtith a reliance on widn Most fields are cut conponents wcre largely selected and .igriculture. house gardens, fallow ma- and burnt'. during the months of least nianaged, a condition Nigh and Naoetis e i hing, and rain; howvever a field can be prepared (1980) call "intermediate disturbance", hunting, in the high forest. Previous ac- any time the weather permits. Field and which Gordon (i969: 69; 1982: 73- counts of the Peruvian and Colombian sizes range from a fourth of a hectare "78) in Panama calls an "orchard-garden- Bora include Whiffen ('915), ie- to one hectare. Axes and machetes are thicket" or -tree garden". oez (1933), Forde (1994), Cirard tie only totols used for fell ing the forest. 1958), Gascht1 (1980,, Guyot ( 117 1, Cutting is otton accomplished itihin 1972, 1973, 1975a, 1975b), and Pl;- hours by cornmunity %ork tearns, bult Fle Research Area redes (1979). individtral fanlilies can cut a field ,.cr Field work was under- a period of seeral da,.s. Oiten small hills taken from July to December, 1981, in ire chosen a, tield sites, the h -hest part JuyecmbrBackground: llora Shifting Cultivation of the hill the tlora settlement of Brillo Nuevo on bcoinrue thle center of the the Yaguasyiacu river, a small affluent A brief survey ot ot- jeld. F.leni ',ecta on is .llo cd to if the Aripivacu river (between the agriculture ass conducted to " lk t I v :,'r tw or il'ree ,eeks lieftre i rn- Napo Pttumayo) Ind which joins t-'c fundaienl; dyinamics of the %. t;. ic1etise c~nitnrg, a conrirnir rita­ •\nazon at Pebas, 120 kilometers and to urder tand hotk ctit%'..t:,,n .. trC',t ehnitoie of swidden trnier;.

l , " i t e : llt" narearlnd I: rrr'i m m::-2 : . . ::: er ,:-e.:c. ,.i as . , :";zil [Ihe clo'est n:1t1,r .n ,1l,:,?',! .r ::.;:',e ..- . d - tec:, t)r,i c,,o IC iludirre sr,.ci .zon,.C",Xn ,irhs: t,,:tds '.r;.,' '[..0i ,':d other usefui trees rll kilmeters 75 di'stant. %aherean ,clhacrtm ini Peru, i.d tke schedle ,L conilonly Ict n or on the ede, -"n, .er 1 )ie'i f 2,757 millimeters kit pianting, ra,-'.csrrn , :d weedin,2. ,- tealv cleared' eilds: others smay 0CO",lc -reciptta i was recorded 11964-1972). eeth. these aIfect the eventural e Iruc-prtectedIn(Figure 1 here is a d>:.tncz seasonal distribution, tire and composition of the faI %%s The flora plant a ade '4Ith rams peaking from Decenier to (prins in Perul. Aimost the entire %arietv of crps: (Table I) hownever !he la n d tat ing from June through area of ',illagc land is in some staie of marin t ie is manioc. So ie 22 ,arret r-,

t k\ t)i(" l 4. VO I. NO b 247 ot ,5, ecl l bitter nnllioc ie knoN n by tl. or,., and . ne, lv phnted Lield brk,- Ie' with ni nioc cut tins spaced 51 to 8t0 centimeler-, :%Ilt. The Bora intersperse pinca ne., frut ittrec,,, and minor anntIal Cropi, anionest ti'e manioc. Bot'i seeds anti seedlmt-, of trees pl:ted. Ni nlMIIlln snicine for fruit trees i, soid to IM be bci\,,ecn I anud 2 meters. Ii \e'er. i as th" pIlintiniC neriod ma e xteid o--,­ t .' u5 4 L.' i c t re e ci, ', a : L .. , :' :l. a t- i; ' . " ' r0.{, ~v 4

iiaie crtcntl\. (ons"Cy ,t lytiale IcY!'I:T k ,. ,'° " "

- on i -i -p.ir .

'SoIlle "' );" ;Ill. ;I'C - i . i ,, V..' ittlin thle ficid. } ;I l i.C' rtt. r '" ' . , conl lton\I clstered o i h l:H.i I I ... . P .phr pclm-ittili. Areas ass. fro: b .: . IlLI boumdaries, or near trail, :aptc. r to be preferred site, fot t .i ,,.i-,'

!ice, Pitl,hu fron I to 2 nmeters .qii.Iic .,,, :.2 7q; arc ;iidi: into plantitIne bed' t lilK - ..­ oi11slcs ,cIccted ' B, accoldinle o , h .,-I, 1 Sriiloti or local soil \ ari,!l onA , ., k -- P0111 recogni/e V'ariosll'oil type-, b.=,d . -" - upon texture and . C'oc iLlio -. ,

Al\ayv planted in \%ell tctid-vd ro\s, ncit F::. t;,:i, of cippycing nmacimbo in one \ear old sw~idcn. trails and field entry a , Peailills. ii _:,C-,L' ­ ond or third year fields. arc pio::cJ - usin , a special inan.ieement iechiiiqL-,. N In a small are.i ilom \which nallioc hlis re.entlh been harvested, soil (pre\ious- . Is loorened by manioc cros\th and root ;" deciy i, ,'ithered and packed into ., - sevcera d.,en mounds incasurirt from t Ii'.,..

about 0. tn 1 meter square. A,,tIes _, bronght to the: f:,lds from hone cook- , ing fires a re mixed in with the soil its M.. U , " -, fert ilie,.r. Betwcen six and a dlozeii - .I , ii-ilnliappiC shelled peanuts, previously soaked V.,i... an ,ii ", overnigfht in an insecticide solution of , " crushed basil leaves and water to pre\ent " ,..t' . " ant predt ion, are planted in the : i'can , I. t t i -'it . mlounds. Frol two to four cuttin-s of pun I.S . "1 \ t Ilinioc .re placed laterally itio - N, fle sidesof tle monds. . y .. /- 44 " ora nanes , for s\widden - - - . . ;I baed aeI a field's c:pacitv :]; 'p', - &. '. / to .';,:1iicCulnioc. A field containing . . aplippk c . " ii ,' I1C I r l. pWodluCtiVe pl.a.tinti, I f ./ ._- m ,'', .dl,.d ai iMli1 c. As an ut'lihe rc la .­t -. - --- iN v . sjj.\5barsC .:ed and replltled, it bcc:ne s ;i kap:iwa, tie terin for a II'Ld , idin ' Ic-. prodiicti, e secondary -'. 2...... n. i.d...... iiif Ikic. l Ike conrc tss,0, cl.,:,tr:. of 1i;n11ioc the naxu iVttiI posibl.Vi-: nia;ioc is not loneer re­ pL. ted tile tci is term ed a i:. s Iiclh ration. lirst. cilstering fruit trees in tie - ill only support sparse, grassy secondary is ioii'hL.\ cqI: is.lcnt It l.loA field or field center or in areas of access allows growth. This may be due to local s0il ptlrnia. them to be easily harvested and weeded exhaustion or coMipa cltion, plant al­ llialii crop 1 in ition as file field latures. Second, heavily lelopathic effects, rein oval of seid II infhoenecs si, en.cat i .;t.-I. nt op- disttUbed or \weeded areas, particularly of secondarn speCic da:ing intense cul­ ticns and the pttcrn of foiest iegene- the coca and peanut zones, frequently tivation, or some coii;:ination of thsc.

348 NOV I'H'C 10Q4, \( tI. 9 N"t, IM/Elzr 'IR '/ (See Uhl, et 19M, for a di:: ,-n df micror:.Ibitat preferen:ces of sccondlar,, scedlings in Amazonia). N The crop composition ( j of Bona fields c.:-i vary widely. Some 6 yr. field fields have an apparent low diversity index, planted only ,,ith manioc, pineap­ ple, and maize (ni.Jnly foi poultry), and : perhaps a few scattered piantains. Others o are rich ;n species and nuners and O f " -' feature complex zon ition. While a range Original field\ of options is to be expected in any swid- perimeter - l l'i, \ \1\ the twko den stvsteni (Denevan, 1971), reFo stro\,tu extrcmcs secnm to l:.e common in Bora swiddcns. A similar duality is noted by I I Harris (1971) for tribes in the Orinoco ,. region of Venezuela, wvhere fields seem -ar ' . ----- 'Frrc.. s B13 (B " , old be either primarily monocropped with to "pp\ I t, Old fallow - staples or polycropped with abundant O l subsidiary plants. In many of these cases Old fallow - Old peonut . eiatch / , 41. the crop composition in any one field 3l yrs - "­ c .. what a may in part be determined by D1) farmer hts available from other fields in /I I' Irame various ,4rages of development. Since a Bora f._milv may have six or more fields of different ages and crop mixtures,. di- Foresth Old c versify betv,,en fields fulfills the saine re r ....l 5. atcO[./, /patch/ function of assuring a supply of varied , ,-'nst. I . crops as does diversity within a single E- field. Another significant point regard- 7Tanti'c ­ ing crops is that simplified fitlds receive r field few visits after two or three y,ars of harvests, while diversified fie! is have longer lasting utility in the fallow stage. Cleaned fruit zone: 700 ni" approx. 2100 m approx. Bora Swidden Fallov,,

Fields of different ages were selected to examine vegetation Star apple -2S ' \Veedy plant ,strutures and the process of abandon­ ment. This paper examines plot. of three, Peach palm -28 Gua',a -2 five, six, nine, and 19 years of age from date of cutting. Each fi.:d was measured -ta'Old pineapple to determine its approximate size and percent canopy cover; the vegetation was OF Avocado -1 Forest inter- r described: and the owtners were IBarbasco-6Manioc viewed to record cropping histories, fiA- lov age, and to help inventory plants :1:Coca--6 found within the fields. The vegetation was sampled using the line-intercept method. In each selected field, subplots nf vecetation were identified for study. Fig. .. Map of five year old transitional/fruit orchard (Kaptluwa). These included plant communities in ,reas occasionally ,kceded and areas of ,in',eeded secondary vegetation. Each i practice. Nonetheless, a dlnitc handonment is similar to Manner's zcne was sampled by extending tvo 10- indel of ab.Ndonment is rea!edy 19a!i36a)devaluation of the swidden 361 elion he sien "eter long intercepts into the zone from a ed y 1 7.indomlv determined foints, Pltaritns an campar:: ..e.latin -l', r, erent .rr'..j cycle, which he describes as a te lines were collected and dentified v aed plots. "iccessional series partly regulated by their Bora names. In addition. Bora m- The swidden :.los human populations on the one hand. anidA farmants identified tseful plants. described elow, reflect a )iratietv ot ecological processes on the other". or transitional field, is chosen The plots are not managed succession desined to 5ole a k:inuuwa. to to head the sequence here because it re­ trictiv comparable in t'rms of relief, or shifting cultivator's diiemina of ho i,, presents a stage when human manage­ -od type, or planting histories. Hoyever, maintain field production in tile tw ht is still relatively intense and forest :inding a series of Iields with identical of the cropping cycle, while at the same ment tlstories and characteristics is impossible time permitting forest regeneration. A- regeneration is only just beginning.

349 Id;ERflffl1'R NOV-DEC 1984, VOL. 9 N9 6 1 ransitional Swidden: Three Years Old TABLE I

Figure 2 is a represen- COMMON BORA CULTIVATED AND PROTECTED ECONOMIC PLANIS tation of an enriched Bora swidden, cut Common Name lo a eSirtfc from 30-year old fallow, located five (English: Peruvian) Name Scientifi NameN m 1- -o--a minutes on foot from the settlement center. The field had developed nmiltiple canopies, featured complex ZOatioF. "Annaito: achi i ­ Bxa -rellana II. L and contained at least 20 Cuhigeni,. *Annona. ch:rint. : anon., taci,5he Annona cherimola I *Ass.. thte:.,: hu ,, . to,!tliul; Euterpe sp. F, C Guava. uvilla. macambo. .1nd peach p dn *Avoc ad.. -- Persea anrrcana F were the dominant trce spe,,all ia h:.CIVIMt,11 Ochrom:a sp. U rnealu~ine beiteen 3 and 4 nietvrs *l 1''Al.'. . EU , ujIwh Mir sp. F. L hce ht. 1ihe t.'s pros, iwc. 3ot perccil "Ii:::: muuc-iisl,...: ',,wit.:rn r p t_" " li;e , Ilb]ahaca -­ ()'ciw? tntcran;:huru F, U" field cover, but had not reached their tlrcr nit mit:itaaa pard Il.l., u rbot -ro, peal, yielding periods. Fruit-tree density de Pan riljahe Arncar;,us incia F in general was greatest near tile s Mithern C:111no mutsitsehe Poutrra cairmuto F end along the trail. The understory ot (alabash tree; pati ­ Crescentia culete U "Cashapona iiss ajCvo lrtariea sp. C manioc was spare bec:use tree roots "Cashew; cashu, marafirn afiadaiaihe Anacardium ucidntiae F and shade prevented replanted manioc *Cedar (tropical); cedro ­ Cedrela odtoraia C from full' developing. s illustrated in "Chambira niijihe Astrocaryum chaozbira C, It Figure 4 for a four-ear old field. A Chili pepper: aji - Capsicum sp. F ; Citr6n ­ Citrus sp. F small peanut. patch, also containing chili .Coo,, iipi Erthroixylon co'a M pepper and othetr minor crop plants, was "Cocona roollhe Solanum sp. F located near the northwest corner. Ba- Coconut - Cocos nucifera F nanas were more or less clustered on the CocoN11 hitina .dnm a Xanthosoma sp. F s(opal copathuallo niijillehe tvrnernea courbaril F, C southwest downslope corner. The field Cotton: alcod6n - Gos.rpiurn barbadense N1 was surrounded on three sides by 30-year Cowpca: chiclayo - Iigna unguiculata F old forest and on the south bv newer Culanrro ItrYngium f.ietidum F fields less than one year old. *(uinala alliuunhe ctCitriico Iriartca sp. or 1'irola sp. C The kapUtL,aa. or tr.Lt- l'1c-,l~te cfiinijcye Calathea allouia F sitional field, is a rapidly chinzmg ,niiw huito ­ Gempa amcricarim H mosaic of vegetation reflecting Bora imi- ' Guava; pacay tfiuitsihye, nagement techniques. Weeding. harvest- Guayaba ajivahe Inga p. F GuaaPsidiurn sp. F ing, and replanting manioc are perform- fluaca , barbasco a,'smihe Clihadium asperuin U ed in one small area at a time, produc- *tluacra pont aall.he Iriartea sp. C ing a pattern of different aged stand, of *lhuaransamana mtneco Jacaranda sp. C both manioc and associated secondary" mi;*tuicungo katuffbah tutsfibah Astrocar'um huicungo C katufibah Cyclanthus sp. H growth within the field. Weeds are often Lemon; lim6n - Citrus Union citratus F pulled out by the roots. In Figure 2 the Lemon grass; yerba luisa ­ Crmopogon F pineapple zone on the left is "eeded and *I-tanchama ptj1icfmico Olmedia sp. t1 the one in the center is nweeded. Ma cambo Aah Theobroma bicolor F, U Maize; mai7 ­ Zea mavs F Selective weeding, another %idespre.iJ .ianloc: yuca ,i:dnuwa (bitter) Afanihot esculenta F swidden technique, is practiced by the baajilri,a Bora. Seedlings of useful tree species (sweet) pcy6dmuwa are often spared: howvecr it is not (Cont.) axiomatic that all are left un.touched. In fields at thki stage, tree coppicing was readily obscrved (Fi­ gure 1). Some of the fruit trees in the field may have been coppicing trunks of were not burned during field prepa- manioc plants were small. As cutting­ trees planted in the field when it was a ration. Some 2 or 3 meters of unweeded are routinely thrust into the earth after chacra 30 years before. Inga species, field perimeter had been ceded to the harvesting the roots, manioc can continue useful as soil nitrogen fixers, are per- encroaching forest. The growth primarily to grow without forming much below, sistent coppicers, so abundant that they cois'isted of fast spreading vines and ground material. Manioc is also a per­ a.e nevertheless dispatched with machetes thin saplings. sistent plant; cuttings merely thrown aside as annoyances. Others, such as copal, will occasionally take root. Tel ce other resprout and are protected. This tree cultigens were origina:illy pL..,ted, of grows slowly, reaching harvestable ace Transitional!Orchard Fallow: which six were still clearly harvetablc: (edible fruits) within some 2( years. Five Years Old coca, caimito. peach palm, tuvilla, avo­ A bonus of useful coppicers appears to cado, and barbasco. be one advartage in clearing fallows for Sonic of the processes Zonation relting from new fields, outlined above were noted in this field, management was evident. The lare coca Because the field was also cut from a 30-year old fallow (Fi- patch was well v.cedcd and rn:6inraincd. periodically %ieeded, secondary growth gure 3), but at a later stage of de- A snmall coca patch was i,b:jdoned and had made little headway except for in- velopnient. The field contained a manioc empty, as was a toat patch. Seond.try vasion at the edges where fallen trees kapiuwva zone; how\ever the unharvested growth in both these abandoned areas

350 NOV-DEC 9S4, VOI. 9 N9 6 I2TEfRlflF9 was limited to short gra,,es, low herbs, TABE I (Cont.) and occasional seedlings of pioneer __ forest trees. A fruit zone extended the Common N..me length of the field along the trail. The (EnglLsh; Peruvian) Boa Name Scientific Name I Use understory constked of a viney thicket mixed wvith low herbs growing amongst Mauritia, moriche; aguaje ifi~jhe .iauritia fleiuosa F, U old pineapples and stray spindly manioc Orange; naranja Citrus sinensis F stems. This thicket formed an intermit- *Papaya Carica papaya F tent subcanopy 1.5 meters in height. *Peach palm, lejibaye; The overstory was primarily comprised pijuayo mSme Bactri.r gasipaes F, Ut of equal numbers of well spaced, pro- Peanut; man( mAtszica Arachir hypogaea F ductive caimito (3 to 5 meters in height Pineapple; pifia cudsiha A.r:nar comosus F 'Plantain; plitano -M ushsaNp. F and peach palm (8 to 10 meters in Pc maroia S:i6:,m malaccensis F height) prosiding a 25 percent canopy Rice; arroz -- Oriza sava F over the zone. ,hapaja Schelea sp. C Secondary vegetation *Shimbillo wachiSbowa Inga sp. F had swallowed about a third of the Soursop; gua:ibana - Annona muricata F plot. The -egrowth zone contain- *Star apple (see caimito) originaled Sugar cane; cala Saccharurn ollicinaridn F ed trees 10 to 15 meters in height and Sweet iotato camote c.itu Ipomoea batatas F measuring from 8 to 15 centimeters in Tangarine; tangerina - Citrus reticulata F diameter. Melastomataceae, Cecropia. Tobacco; tabaco Nicotiana tabacurn M Jacrnotida, and Inga were common. The *Umarf nimuhe Poraqueiha sericea F trees and abundant upper story vines Ungurabe, ungurahui - lessenia batana F formed a 100 percent canopy. The forest 'Uvilla ba.dcohe Pourounta cecropaetloa F o aYam: sacha-papa Diocorea trilidamacrocarpa F floor %as a dense tangle of herbs, in- Yarina t6kehilbah Phytelephas C eluding abundant Melastomataceae, Pi­ peraceae, and Araceae. Palms were few. In this transitional field, Common name, tentatively matched with scientific names from Soukup (1970); most pineapples, fruit trees, and other minor not collectcd. plants deemed useful were maintained. The pineapples may be harvested for up Use Code to five years; thereafter the fruits pro- F: Food and lleverage duced are small and bitter tasting. Visits C: Construction or Thatching to the field follow the ripening schedules H: Handicrafts and Dye., of the fruits, although visits for hunting U: Utilitarian also occurred periodically. The main ar- M: Nledic;nal and Drugs tivity besides harvesting fruit was weed­ ing. Coca was weeded every three Plants appearing in fallows months; the fruit trees and pineapples received a slash %keeding by macrete every three to four months. The farmer identified many useful plants, both in the veedy orchard and kaptl'Wa zones and in the reforested perimeter. The most immedi- ' ately harvestable species are vines and low herbs. These include utilitarian vines and ceremonial plants not now used by the Bora, including reeds, once used to make decorative noeplugs and flutes, and plants yielding body paints. Other useful, but not vet harvestable species were construction and other woods in the seedling stage. The relOrested zone contained a great numb-er of species. Thirty-four plants at'ie:ri( on tv, lo- Meter transects, 13 of 'Ahich skere con- - A, 1-.3:;

;idered useful. Six %kere con;ruction - woods; four provided materials for ,.cay­ .rg and dyeing baskets; and three vere fire oods. Most of the useful forest species in this fallow will not 1e harvestable for ten to 30 years. Rapidly Fig. 4. Photo of tour ,ear idtransirnal field (Kapuui. a). Residual manioc in the loreground; growing construction woods are har- uvilla tree to the right: a nacn:ambo branch juts out tcp left.

tfl1(,L"EflUO NOV-DEC 1984, VOL. 9 N9 6 351 vestable, but they are so plentiful around TABLE I Erillo Nuevo the,' receive no special care. The Dora casually harvest useful her- SUCCESSION Ot IIARVFSTABLE PLANTS IN BORA FIELDS AND FALLOW baceous plants as needed. Stage l'lanted ltarveslable Spontaneous Harvesta'., Orchard Fallow: Six Years Old High forest. None Numercus high-forest This orchard fallow is construction, medici:i, mapped in Figure 5 and illustrated in utilitarian, handicraft Fig. S. Cut from primary forest, i is food its a:alt. :tltride. a sloping hill surrounded by Ne-M, p!anted Al! specie, develop:ng Dry firewood from unP..: ncvess field, on thiee sides. fi , r trees for hot fires The plot consisted o t\ko general \egetation communities: it New ield (Cimihe1): Corn. rice. cowpeas. Various tiseful early rci(i]Ll fruit orchard occupying about 3.9 n, succesi~onal species. one sixth of the original cleared area. in! abundant growth sur- ?slatuie field (fimihe); Manioc, some tubers, Abasidoned edge zor.e has ,, 9 noo. - 2 yrs. bananas, cocona, and useful vines, herbs. rounding the orchard. The original field other quick maturing crups. ,kas planted with over 26 crops, some of %\hichare tree species now surviling Transitional field Replanted manioc, Useful medicinals, wkapituwa); pineapples, peanuts, coca, utilitarian plants within within the orchard. Caimito was the 1-4 or 5 yrs. guava, caimito, uvilla, avocado, field and on edges. most numerous planted species, and cashew, barbasco, peppers, Seedlings of useful trees these trees measured 3 to 5 meters in tubers: trapped game. appear. Abandoned edges hcight. A canopy was formed by uvilla yield straight, tall saplings, including Cecropia (5 to 8 meters in height) and peach palm and Ochroma lagopus. (10 to 13 meters in height). Several 18­ meter tall Cecropia trecs dominated the Transitional fruit Peach palm. banana. Abundant regrowth in field. soft orchard area. The orchard had a 70 field (kaptiuwa); uvflla, climito, guava, Many useful 4-6 yrs. annatto. coca. some construction swoods atmi percent canopy cover and was well lit tt er: prOp.i9ules of firewoods. Palms appear, by .sin splash. Recent weedin:- had re- p1ineapple and oiter crop ; including Airoiarvum. Sstilted in an open floor of grassy ve- hunted and trapped game. Many vines: useful undlerstorv aroids. eetation covered with slashed ntulch. Harvesting fruit in such orchard fallows Stage Planted Ilarvestable Spontaneous Itlarvestablc is a casual pastime. The Bora use poles Orchard fallow (jia); Peach palm. some uvilla. Useful plants as above; equipped with vine loops on the ends to 6-12 yrs. macambo: propagules; self-seeding Inga. ensnare and pluck fruit-laden racemes hunted game. Prob;ibly most productive from high branches. Coca, however, had fallow stage suffered from shading and harvesting Forest fallos' (jia); Macambo, umarl, breadfruit, Self-seeding macambo and was reduced. Cuttings a:,. removed for 12-30 yrs. copal. uniari. High forest replanting in nearby new fields. There is successional species little evidence of manioc besides oc- appearing. Early casional stubble debris, successional species in gaps. Some useful hardwoods Growth surrounding,the becoming harvestable, e.g. orchard was topped by 25-meter tall cumala. Many large palms: Cecropia and Rubiaceae trees towering huicungo, chanibira, assai. over dense stands of 10-to-15 meter high ungurabe. trees, saplings and old pla:it-ins. There Old fallow; Umarl, macambo. Same as high forest zbove. was a thicker shrub understory contain- high forest Full maturity not reached ing members of Piperaceae and Ru- until 50 years or more. biaceac, shc rt saplings, and palm sprouts. The forest floor had accumulated a thin 0 Plants are identified in Table I. laver of leaf litter, and no grasses were present. .Anarray of useful spontaneous- Iy appearing species similar to those in the five-year field were present in the meters, and cultivated trees were few; The secondary regro\tth reeros th zone. hovsevcr, a vigorously growing unshaded was a \woojy thicket, 10 to 15 meters it! coca patch still remained. The patch height, vith many vines and su!l-,o:\ Orchard Fallow: Nine Years Old contained 82 evenly spaced. vell tended shrubs. Several useful trees includitt. bushes. Coca was clcarly the most va- cedar were on the field's perimeter. hc­ This fallow (Figure 6), luable crop available here. The owner cause this field is a dov'nriver site, 5Oi', cut from high forest, demonstrates how visits the field on a regular basis to and topography differ from the tip!.otd long a managed orchard-fallow succes- harvest the !eaves, and on those oc- sites nearer the settlement zone. 1!C sion can be matintaied, although size casions may refresh himself with uvilla, downriver sites are less s clldi:,ned, and is small and diversity is losw at this age. guava, and caimito foraged from the thus the secondary comn'tiniticS dffCr The orchard zone was only 250 square residual orchard, from the other fallows studied.

352 NOV - DEC 19S4, VOl . 0 N" 6 M7Ei;DE., Foret FalloN: Nineteen Years Old

This onie older iore-;t N New- field 3 ios. fallow (Figure 7) wa~S Surveyed forA useful tree species. The original swidden VY was cut from mature forest and, ac- .Fe r o t A cording to the owner, planted with at F r ot least I1 species, including several va- -I," --. __ rieties of fruit trees. OrrTz

Thle forest displayed - I'c ~ .1*.--- clear stratification. Low vegetation con- Tranet'1 .4 1,,I " sisted of herbaceou,; plants, including .. O ferns, measuring 30 centimeters to I Orchard meter in height. Above was a ,cond - .t Q / stratum of thin, straight saplings, 5 to 6 meters high, including many palis. I,. Seventy-five percent of the canopy was 4 yr. field I - - High forest provided by trees 15 to 18 meters in height, while emergent Cecropia and Ja',,,nild, both 25 meters tall. filled V, out the canopy. The forest floor was -0 Transect B Forest regrowth percent covered in leaf litter and walking was unhindered, except in small thicket- V filled gaps created by felling tices to A /,A 0Z harvest grubs. All individual trees A measuring 15 centimeters in circumfe­ rence within t transect 10 meters wide and 102 mete.rs long (length of the field) were tallied. Sonic 233 trees be- -1 longing to 82 species %kerecounted. Over High forest half the trees were single occurrences. Our informants iden- 1900 ni: approx. tified 22 useful trees in the transect, fit- Cleaned fruit zone: 300 m" ting into the following categories: a. Construction ma­ terials: 11 species, 25 individuals. in- Star apple -31 Annatto -2 eluding two varieties (three individuals) I of highlv valued cumala; 13 huicunco ' -5Umari "1 palms, used for general thatching, were )r also present. b. ledicinals: four Peach palm -I1 T CocaCoc

species, four individuals (not vet iden- F ied).a a 4 Forest

c. Food: two species. I1individtn'ls. consisting of eight ma- - ---..... cambos and three assai palms bearing Fig. 5. Map of sixyear old orchard fallow. edi'Ie fruits. d. Artisan material: one individual, a :Ive-bearing tree (not None of the aho.e trees, harvest after three or four years. mainly. yet identified). all of which are common, appeared to they say, because of soil depletion. but e. Utilitarian: four in- receive individtual attention. Ilhe cumalas also because of weed invasion.,-\band,n­ dividuals, four species. These included were not vet har estable. nor will they ment of fields planted almost eiitirelv in three palms from which ,Al is distilled, he for ,ahout a decade. Ihis old fallow manioc occurs within the space ot a %ear. and one tree from which Pitch is extract- ipp:irently receives few vtsits for col- However. if fields are polycropped with -:!cd iied to seal canoe :ls IAl :1'f Iection b'ur:-,cs.ut huntin trips and trees, weeds mav le the niior ob,t.Icle )et identified). grub tor12in ,refrequent. to extended Iield use..M~mntene:u-,r;H!. In :ddtion tLtcle .k C.'' front replantioig ntn11oc to Ae1ioh111 it least two other specie, if trees I ., 1 ladoe n lyi* encroaching secoidary veectat io :1' r:lt­ '-hien edible grubs are harvested. lhe ening tree cr:aps. With periodic ...'cdilIg. nl, apparent survivors of the prior The fora recognize that trees can renain producti,.e for several 0Xidden were the macambos, which were two ecological process-es, soil depletion years before disappearing into the ,e­ clustered within the transect 60 meters and secondary succession, must be con- condary forest, often succumbing to the *ownslope. These were h-srvested spora- fronted. They acknowledge that manioc effects of

;rEREJEfR'A NOV-DEC 1984, VOL. 9 N9 6 33

K Our observations in­ dic:ate that the most productive ldloa stage is between about Iour and 2 vears. N Before four years, fruit trees are \,. not yet producing or have limited pro- Peach , palm-6 Old falio ducton. After 12.. years. management il applfl.7 "p minimal and many of the smaller useful DIP Star apple -7. plnts are shaded out. Harvesting of V t1vila­ some pecies continues, howeser. for uip to 0'.1-s 20 to 30 or more ears Another ini- 1Yarana -5 portnt characteristic is se0sonalit,. Te -1 fa various flora lruit-tic e spcI , .' \1Xic. st- I qiinti yll 111Clos a s a e: pro uc :1" . .", , A nu)her (t tree -1 .orest Fvdar regrowth species planted In lor; fiCldS aitc, h,.0 ,- l a/ ever. adapted to growing in dense se- T' ld I / / condarv forests. Umari and macambo ( trt / Qn)n are common cultivated trees found in old fallows, either growing alone ,r in groups. These survivors of swidden orchards are Valued components of Bora faliows. At 20 or 30 .ears of age mot Old fallow fruit trees cannot be easily harvested; howevcr the Bora occasionally gather 23o0 m* appro. the fallen fruits. A valuable function of CIaned frut 1one: 250 m: fallen fruit ik that they attract garie animals. It is common to find an umari Fig. 6. Map of nine year old orchard fallow. fruit on the forest floor \ ithi tooth marks of a majdis (Cunicuius paca) or other browser. For this eason older purittas planting ha rtl ood. and usefkul palms in seem to follow this approach. Bananas are good hunting grIoinds. Thre process of abart- swidden and in fallomed fields. do well in low shaddv areas, where lHe­ d o n mc n t a n d fo re s t I-liabn- proess r e g-e n ra tion c lea ­lyl f mosticoniai ni obviousplantsp a t s arer e alsoa example ois common.c uvillam o which. T'teT e has a spatial aspect. While successionary Phascd Abandonment: Implications matches its ubiquitous cousin, the C­ processes are complex, there is a tendency for Agroforestn' cropit. Guava is also in the same genus towards a pattern of centripetal forest as itsit a serni-doinesticateds m d o icate analog,anl g nth regrowth which might be explained large- the There are similarities be- shimbillo. Further research may reveal ly as a result of the history of weeding. tseen comilex swidden systems and agro- other similarities between naturally ap- Harvesting and wecding of manioc holds focstrv systems (1-ccht. 1982). Agrofo- pearing species and cultigens %hichcould regrowth at bay. Once a manioc zone is restry combines the production of trees be incorporated into swidden agro­ abandoned, terrain is t cradually surrcn- and o wer crops on the same unit land forestry type models. dered to the forest and the field shrinks (King aind Chandler, 1978), a strategy Another feature of 1ora in size. essentially identical to sidden-fallow sw ddens that could be useful it agro- Abandonment is also m;inace:nent. Both systems rely on '., forestry design is the use of space. flora relatkd to how harvesting proceeds se- succession of tree crops following te tree clustering according to local to­ quentially from grain producing annuals harvests of short-term cultigens. pographical conditions suggests that (rice and maize) to root crops and pi- Viewed in this fashion, slope and terrain should he considered neapples to fruit trees and spontarncolisls flora agricu~lture converts to an agro- when planning agroforestry plots. More appearing utilitarian trees and vines, forestry system during the e:rly stages important, slowly abandoning ground to Tabie 11 shows the suc- of forest fallow. rhe enriched swidden secondary forest may be a sound strate­ cession of harvestable plants in Bora to fallow sequence closely resembles the gy for tropical farming. There is no fields and fallows. While the Flora re- natural succession analog approach to reason to think that agroforestrv plots cognize many useful fallos plants, many tropical agroforestry outlined by 1tart should have 100 percent planted stand­ go unhlarvested and are essentially ie- ( 1980; also Uhl, 1983: 78-79). Hart ing biomass. Managed forest rcgrowth glected. The main reason for this is that suggests that select cultigens be placed could provide useful products, as well high forest, from where sturdy construc- in the niches nornially occupied bv as '-anopy cover for the soil and a source tion woods and vines :ire harvested, is common early successional species. File of stored nutrients for \shen the forest still a short walk away. At present !itost analog plants would have CVrossth strut- is cleared to begin the s\\1Jden and :aro­ plants used for handicrafts, for example, tures and resource requeriments similar forestry cycle anew. ame taken from high forest. Nevertheless, to those of their weedy cotiterparts. Swidden-fallow agrofo­ as the high forest frontier becomes more Thus, rice or nmiizc replaces early annual restry, enriched \sith tree clops plant­ depleted and distant, secondary growth species, bananas replace wide-leafed ed in, areas of forest rcgro\lh, could species become more important. There lIcliconia, and late ;!ppearing tree crops approximate the "tree-irden" Iodeli of is evidence that this is occurring. The mimic early successional tree species. silviculture a\hieh itay have been a Bora have recently beccme inte-ested in Whether by accident or design, tile Bora pre-Columbian ariculitiral adaptation

354 NOV - DEC 19P4, \()L. 9 NO 6 JRU-ERO'R) ,ft in the Caribbe:n lo'Alandi of Colonmbia. Central America, and the Maya region (Gordon, 1982). This model involves a trail combination of overstory fruit trees and sub-canopy woody shrubs, interspersed with open areas of s%iddens containing A maize. bainanais, ianioc, and other crops. JI Svstemaitic swvidden-fallow agroforestrvy would have a fruit orchard core, or se­ ries of cores. but these would be emi- 5 10ldao braced by areas of regenerated forest. \A 1 Slope 'Ilie forest, in turn, could be enriched hv a variety Of Useful anlgspecies As able to compete in the viney sub-canopy, or later on as canopy species (fruit, timber) in high-forest fallow. Timber Forest regrowth species woull be appropriate iate-fallow \ enrichment trees. Over a large area,r\'\ swidden-fallow agroforestry would re- ' semble Gordon's model. It would be more a thicket and If-is a plantation.A Furthermore, the gro%\. rate of manag- Old falo \ , ed successions may be as fast or faster than natural successions (Uhl, 1983: t\\%X 4 / 79).

Swidden-Fallow Products N Forest regrowth \ \ The cumulative dietary contiil ution of fruits and nuts. even 'A %%hen harvested casually, may be sig- "' "I \ ., nificant. Certainly they provide a con­ tinuing (seasonal) variety of minerals. fats, and vitamins to tropical diets do­ minated by roots and tubers rich in carbohydrates. Some trees. moreover. 3500 m2 approx. can provide major staples. The peach palim. very important to the Bora for its Fu:t and heart, can compete with maize \ Clustered as a nutritious food (Hunter, 1969; 10 m. wide transect macamo Johannessen, 1966). In addition, plant i\ \ products useful for beverages, con- Forest A diments. construction, tools. orugs, and medicines are of more than minor im­ portance to villaie societies and eco- 233 Individual trees 4.7 cms. dia. or more notiaes. ..s with major and 82 Species identified by Bora informants minor natural forest products, those of widden fallows frequently reach markets 22 Useful species beyond the villa,e, at regional, national and international levels. Even remete I traditional cultivators ar illing and Fig. 7. Map of nineteen year old forest faltow. ihle to respond to market opporluni ies for forest products and manage those products accordingly. Pelzer (1978: att of forest products is problematic. Tro- 2' 6) arguecs that to Bora households. The considerable a larg-e flercenta~ . deputtdei h lii '" rubber, black repper. :opra. coi:ec. . !,:,. ir .rd ,)her.- export trade in the l ,itos nd benzoin harvested for cash in be floamte down river to mrk:1. One area provides an outlet for traditional Southeast .\sia comes from smail-hoader a Ton , e tprc.,sIIxI,,Jrms d -L.mning items Nuch as h .mnccks. bas. baskets. .viddens through intercropping "inwhat or protecting tropical cedar ceLilins in bo,,i and ornaments. On the other isordinarily thought of as the 'fallow' their swiddens and f.llokks.antic::uin a hand. the marketing of perishable food neriod of the sivddens". Tne ultimate iibstanti d cash otiumn for ther :hildrn items does constitute a difficult problem "lccess of aeroforestry systems will 3I10 e -s laci r. rhe use of o%.oLden-falow for remote villages such as Brillo 'Nuevo. lepend on such cash cropping. products. )uch ispdh and liarar. rers. especially in view of the poorly de elop­ For isolated communi- tree bark, and dyes. tor he m.nufacture ed processing and marketing facilities in ties such as Brillo Nuevo, cash cropping of handicraft items can brim an income the region. Toasted macambo nuts. a

MU1EREnE'fl NOV-DEC 1984, VOL. 9 N9 6 355 §h.

Bora delicacy, could have market poten! "A tial.Palms such as Jessenia and Mauritia, I Ecoloi potential sources of edible oils (Ba- (Berk, lick. 1982), are common in Amazon Wrke,'A aence. forests and could be integrated into agroforestry models. The history of the uR. Gi Amazon has been one of commercial 363-31 harvesting of forest products (quinine, Conllin, I copal, sarsaparilla, barbasco, palm heart, (Romi Brazil nuts, rubber, timber). Much of ch, that history involved the destruction of C Sdic important resources by urhiie harvest- dAelli ing practices and the economic and so-m cial exploitation of indigenous peoples. Denevan. Sustainable and equitable' procedures are in1h, possible, and trade in forest produces can 4. be enhanced by incorporating forest Lil.evan,; comrrpulti species of com.ercial value into agro- (ulaii forestry s'ystems. Such commercial orien- Anaer tation would, of course, necessitate not Wise, only the, development of, specific agro- DSchlip, forestry designs and techniques, but also i ArTc appropriate processing, transportation, Fig. 8. Photo of six year old orchard fallow. Uvilla and peach palm are present Note o eI credit, and marketing facilities. The well lit floor. Dubtoi,; economic possibilities for Amazonian " Cont plants are :vast (Myers, 1983). An ar- Wor gument might well be made that the po- Ame. tential value of marketable production harvests of fruits, spontaneously appear- 1984). We have much to learn about Eckholmn from sustained-yield agroforestry plots, ing species, and even animals continue these systems. rianW including swidden fallows', can be sig- until the forest is removed for further 6. Agroforestry drawing nificantly greater per year per hectare cropping. on traditional management methods and ,IEden, M than that from cattle ranching or shift- 3. There exists an iden- combining planted species and natural . 1. Ftt ing cultivation. tifiable sequence from original forest secondary vegetation could be an eco­ with some economic plants present, to a logically appropriate and economically Sod. swidden with numerous individual eco- viable alternative to destructive short- 509- Coniclusions nomic plants present, to an orchard fal- fallow ,shifting cultivation in tropical Forde, low or agroforestry phase combining -1reas. The ideal model would provide Am The Bora process of managed economic plants and natural food crops during the swidden stage and j swidden abandonment is in reality a con- vegetation, to a forest fallow in which cash crops and otherr products during G h' version of a short-term croppingsystem economic plants are fewer but still the fallow stage. The cash crop percn- 1, into a longer-term agroforestry system, present in greater numbers than in the nials should be relatively fast maturing cicst Some general conclusions regarding original forest. Likewise, there is a co- species which can be harvested by :. tern abandonment and fallow management rresponding sequence in the proportions around 10 years so that the cycle can pert are possible. of biomass which are cultivated or ma- be renewedas soon as possible. Such a che I. Fallowing is multi- naged, spontaneous economic, and model would help fullfill the need for Gaschlos, purpose. The secondary forest'is not only spontaneous non-economic. sustained production of food and other a nutrient storage forftecropp b 4, Research is needed needed products and simultaneously do , an important niche for secondary crops on analog species -with growth architec- minimal damage to a fragile environ- Act and useul spontaneously appering tures and nutrient requeriments adapted ment. an, plnts.; We identified 133 different to secondary forest environments. REFERENCES usefulspecies in Bora fallows. We pro- REFERENCES Girard, be 5. Swidden-fallow ma­ pose that an appropriate designation J. r (" established to account for enriched fal- nagement is not unique to Amazo-i___s~n._Alcorn, bog an) B. (Austin,(1984): HuasiecUniversity Mayan of "Te,.aEthno- s :' lows, We observed it with the peasants ,Press); ...U .. .Gordon' . a characteristic which may be Indians. .i': common in tropical swidden systems, in Tamshiyacua,-,easb wies~eauprivert from Iquitos.-:;, ~ Ito Balick,-e, M. . J. (19812):.(18) PalmasP neotrmpicales,e p : . t f r ptrr The term "orchard fallow" could be used appears to be widespread in Africa (D Nuevas fuentes de acites comestibles I to describe the structural and functional Schlippe, 1956: 215-216; Dubois, 1979) Interciencia, 7: 25-29. G( aspects of traditional agroforestry. In a and in the Pacific, including the Philip- Basso, E. B. (1973): The 'Kalpato India-u Gordot subsequent "forest-fallow" stage, eco- pines (Conklin, 1957: 125-126; oraci6n, of Central Brazil (New Yorl, liolt, Ri- Sit nomic plants are still present but are 1963), New Guinea (Clarke, 1971: 82- nehart, and Winston). 6 more dkpersed, fewer in number, and 84, 138-139; Hyndman, 1982), and Mi- Budowski, G. (1981): The place of agro.fo­ less managed, cronesia (Yen, 1974). It may once have restry in managing tropical forests. In: Guyot, 2. Viewed properly, a been common in Middle America (Gor-, F. etrgen (d,), Tropical Forsts: U1l. d ~iddn ste s nvercomletlyban do, 169,198), nd t i stll racic- lization and Conserv'adon (New I lavel', tu nmplYale University School of Fotestry), IT, (c doned as a resource zone. Secondary ed by the Huastec in Mexico (Alcorn, 181-194. .

356. NOV - DEC 1984, VOL, 9 N9 6 ...ER ... I) IRERL Wr'.VEi(97); Place a'u'PepleY'.nVGtyot, M. (1974): LUaThalson 'des indien5~ Paredes, P."O0, (1979): Bora-iooOans-"' Eology oy ae Nw G inean Commity Bora et NMiraia. (Etudes sur la territoire .Estudo Einoldgico de las Coin'undades I(Berkeley, University of California Press). ci l'habitat dans l'ouesto .amazonicn). Nativas del Rio Anipiyacu, Bob eliazo­ ' eiaVk,W. ' C. '(1977): The structure of per- Journal de Ia Sociil des A,nericanistes, nas (Iquitos, .Ministerio de Agricultura), manonce: The relevance, of self-subsis- " 61141.176. Peizer, K. J. (1978): Swidden cultivadon in tence communities for world ecosystem Guyot, M. (1975a): Le systeme cultral Dora- Southeast Asia: Historical, ecological, and management.. n:"T. P. Bayliss.Smith and Mirafa, In: P. Centlivres (ed,), Culture. economic perspectives. In: P. Kunstadter, R. G. Feachsm (eds.), Subsistence and sur Prulls et Evolution de Milieu Forestler E. ,C. Chapman,,and. S. Sabahasri (eds., Sur-vival (London, Academic Press), pp. en Amazonie de Nord-Ouest (Geneva, Fah..e,-, in the Forest (Honolulu, Uni-.,' 363-384. Muse d'Ethnographie),' pp. 93-109. versitq.Press of Hawaii), pp. 271-286. onklin, H. C. (1957): Hlanuujo Agriculture Guyot, N1. (1975b): Mission chez Ics Indiens Posey, D. A. (1982): The keepers ,of the (Rome, FAO). Bora et Mirafia (Amazonie colombienne, forest. Garden 6 (1:) 18-24. ovich, A. P. and N. H. Nickerson (1966): 1969-70). In Amazonie. Nord-Ouest (Neu. Posey_ D. A. (1983): Indigenous, ecological ':dwellingStudios ofclearings, cultivated Darien, plants Panama. in ChocoEco­ chatel, Muse d'Ethnographie), pp. 17-28. knuwledge and development of the Ama. dwllin clcean, 0: ie, P . Harris, D. R. (1971): The'ecology of swidden zon. In: E. Moran (ed.), The Dilemma nomccultivation in the Upper Orinoco rain of Anazonian Development (Boulder, nevan. W, M. (1971): Campa subsitene forest, Venezuela. Th, Geographical Re- Wtstview Press), pp. 225-257. in the Gran Pajonal, eastern Peru. The view, 61: 475-495. Salas, G. de las (ed.) (1979): Workshop: Geographical Review, 61: 496-518. Agro-/orestry Systems in, Latin Amer;ca 'nevan, W. M,ninlovppoanttedein (1976): The aboriginal p0- analog approach to the n designr offasuc­a c (Turrialba, CATIE). pulation"of Amazonia. In: W. M. DeneVan cessidnal crop system for tropical forest Smole, W. J. (1976): The Yanoama Indians: (ed.), The Native Population of the environments. Blotropica, 12 (Supplement, A Cultural Geography (Austin, University Americas in 1492 (Madison, University of Tropical Succession): 73-83. of Texas Press). Wisconsin Press), pp. 205-234. S.~hlappe, P, (1956): Shifting Cultivation in Hecht. S. B. (1982): Agroforestry in the Soukup, J. (1970): Vocabulario de los Non­ Amazon Basin: practice, theory and limits bres Vulgares de la Flora Peruana (Lima, Africa (London, Routledge and Kegan of a promising land use. In:. S. B. Hecht Colegio Salesiano). Paul). (ed.), Amazonia: Agrlculturei\and, Land Spurgeon, D. (1980): Agroforestry: A pro­ Dubois, J. (1979):' Aspects of agroforestry Us' Research (Cali, Centro Iternacional rising system of improved land ma. systems used in Moyombe and !Congo (Zaire).-In: G. de Ins Salas Lower(ed.), dor Agricultura Tropical), pp j331-371. nagementag men for Latin America. 'Inter­ Congksop:( aiore Gnr dLasSystems (ed Hunter, R. 3. (1969): The lack bf acceptance ciencia, 5: 176-178. Workshop: Agro.,orestry Systems in Latin of the pejibaye palm ' and a:!relative cOm- Torres Espinoza, W. (1980): Prcticas agro­ N America (Turrialba, CATIE), pp. 84-90. parison of its productivity to that of maize. pecuarias en Ia Amazonia ocuaturiana. In: Eckholm,Svironmient E. P.,.(1982): and Human Down Needs to (NewEarth: York. En- Economic Botany,J'r 23: 237245. Consulta Cientifica Subregional sabre las iW. W. Norton). Nd(Hyndman, D. C. (1982): Biotope gradient in a Actividades de Corte y Quema en ei Eco­ d . . Norto).A trdt ow r ke H diversified New Guinea subsistence system. sistema del'Bosque Tropical (Iquitos, Man den., NM.J. (1980): A traditional agro-system : Human Ecology, 10: 219.259. and the Biosphere Program), pp. 37-53. in the Amazon region of Colombia. I: J. I. Furtado (ed.), Tropical Ecology and De- Semlnario, A. (1933): Breve estudia UhI, dwnnezC. (1983): You can keep a good forest velopment (Kuala Lumpur, International sobre la tribu Bora. Revista Universtarla down. Natural History 92 (c): 6979. Society of Tropical Ecology), Vol. I, pp. (Cuzco), 22: 173-191. Uhl, C., K. Clark, H'. C. Clark, and P. 509-514. Johannessen, C. L. (1966): Pejibayes in com- Nfurphy (1981): Early plant succession Forde, C. D. (1934): The Bora of the western mercial production. Turrialba, 16: 181- after cutting and burning in the Upper Rio Negro region of the Amazon Basin. Amazon, forest. In: C. D. Forde, Habitat, 187, Journal of Ecology, 69: 631-649. Economy and Society (London, Methuen), King, K. F. S. and N. T. Chandler (1978): Whiffen, T. W. (1915): The North-West pp. 131-147, The Wasted Lands: The Program of Work AmaonT ( 191) ,Gasch., J. (1979): Cultivo de corte y quema of the IC.\F (Nairobi, International Amazons (London). y evoluci6n del media forestal en el nor- Council for Research in Agroforestry). Yen, D. E. (1974): Arboriculture in the sub. oeste del Amazonas: ecologia de los sis- Klee, G. (ed). (1980): World Systes o247-287. sistence of Santa Cruz, Solomon Islands. temnas do cultivo indigenas en la selva Traditional Resources Management (New peruana (Paris, Centre National de la Re- York, Halsted Press). cherche Scientifique), mimeo, 24 pp..YokHatdPrs) crcheJ cient ,_...'mimeo, 24 . Manners, H. 1. (1981): Ecological succession .Gasch, J. (1980): El'estudio couparativo de in new and old swiddens of montane ACKINOWLEDGEMENTS los sistemas de cultivo nativos y su in- P N Gui, H.. gy, 9: pacto sobre el bosque amaz6nico. In: 359 377 The Bora agroforestry project was funded Consula Clenti/lica Subregional sobre las Acltl'edades de Corte y Quema en el Eco. Myers, N. (1983): A Wealth of Wild Species: by the UNESCO Man and the Biosphere ststenlade Bosque Tropical (Iquitos, Man Storehouse for Human. Welfare (Boulder, Program (mAB), 1981-1983, under an agre­ and the Biosphere Program), pp. 61-74. Westview Press). ement with the Universiy of .Wisconsin, Girard, R. (1958): Los Bora. In: R. :Girard, National Research Council (1982): Ecological Madison and the Universidad 'Nai ional de la I lndlos Selvdticos de la Amazonia Peruana Aspects of Development in the lIhanid Amazonia Peruana in Iquitos. We are grateful (Mexico, Libro Mex), pp. 85-124. Tropics (Washington, D. C,, National to Manuel Mibeco, the Bora village leader in ordon, B L. (1969): Anthropogcography and Academy Press). Brillo Nuevo, for his cooperation and con. Rain/orest Ecology in Bocar del Toro Nigh, R. B. and J. D. Nations (1980): siderable assistance to our fitld work. Plant Province, Panama. Office of Naval Re. search Report (Berkeley, Department of Tropical rainforcstq. The Bulletin o;) he collections, facilitated by the Iquitos herbarium Geography, Univerity of California). Atomic Scientists, 36 (31: 12.19. supervised by Franklin Ayala, are still in the Gordon, B, L. (1982): .4 Panana Forest and osrRN (1976): Inveniarw, Eialwiacait e late- process of idenufication. Salvador Flores and Shore: Natural History and A.merindian gracidi de los Recursos 'Naturales de la his assistants are continuing the project, con­ Culture in llocaT del Toro (Pacific Grove, Zona Iquitos, Nauta. Requena y Colma centrating on older swidden fallows, as well Boxwood Press). ' Angamos (Lima, Oficina Naicional de Eva­ luaci6n de Recursos Naturales), as managing etperimental agrofarestry plots uyot 1., (1971): Recherches etnographiques in Iquitos. An early version of this paper was dans les bassins des rios Caqueta et Pu- faci6n, T. (1963): Kaingin igriculture among presentedat ' 4 " " " o- re • tumayo, Amazonie colomblenne: le. Bora the Bukidnond of south eastcrn Negros ented at te 44th Iternatona C.ngress (compte i'endu de mission). Journal de la Philippines. Tropical Geoqraphy, 17: 213- of Americanists in Manchester, England in !' Socidt; des .-lnterlcanlstes, 58: 275-283. 224. 1982.

MKRCERU NOV. DEC 1984, VOL. 9 N'6 1- 357