Er

Da

b

c

d

e

a

ARRAA

KEENRRW

1

1

iiiabm

tc

((M

0d

Ecological Engineering 45 (2012) 30– 44

Contents lists available at ScienceDirect

Ecological Engineering

jo u r n al hom ep age: www.elsev ier .com/ locate /ec oleng

cological engineers ahead of their time: The functioning of pre-Columbianaised-field agriculture and its potential contributions to sustainability today

. Renarda,∗, J. Iriarteb, J.J. Birkc, S. Rostaind, B. Glaserc,e, D. McKeya

Université Montpellier II and Centre d’Ecologie Fonctionnelle et Evolutive, UMR 5175 CNRS, 1919 route de Mende, F-34293 Montpellier cedex 5, FranceDepartment of Archaeology, School of Humanities, Archaeology, and Earth Resources, University of Exeter, Laver Building, North Park Rd., Exeter EX4 4QE, United KingdomSoil Physics Group, University of Bayreuth, Universitätsstr. 30, Bayreuth 95447, GermanyArchéologie des Amériques, UMR 8096 CNRS, F-92323 Nanterre, FranceTerrestrial Biogeochemistry, Martin-Luther-University Halle-Wittenberg, von-Seckendorff-Platz 3, 06120 Halle, Germany

r t i c l e i n f o

rticle history:eceived 18 November 2010eceived in revised form 15 March 2011ccepted 17 March 2011vailable online 20 April 2011

eywords:cologically intensive agriculturecosystem engineerseotropical regionaised-field agriculture

a b s t r a c t

The need to reconcile food production, ecosystem services and biodiversity conservation has spurred thesearch for more sustainable ways of farming. Archaeology offers examples of prehistoric pathways toagricultural intensification that could be rich sources of inspiration for applying ecological engineering inagriculture today. We examine one set of techniques, pre-Columbian raised-field agriculture in wetlandsof Mesoamerica and South America. We point to gaps in knowledge at three levels. First, raised-fieldagriculture was conducted in a wide range of soils and climates. How different systems functioned waslikely to have been correspondingly diverse, but this variation is under-appreciated. At the scale of singlefarms, nutrient dynamics in raised-field systems likely included complexities quite unusual in ‘modern’agriculture, owing to the mixture of aerobic and waterlogged compartments, but data are scarce. Second,at the landscape level there is disagreement about whether fallow periods were necessary, and their

ehabilitationetlands

eventual roles are poorly understood. Current evidence suggests that self-organizing processes in fallowsmay have increased the sustainability of some raised-field farming systems in unusual ways. Third, thelabor-intensive nature of raised-field farming is held to limit its pertinence to today’s global problems,but its real labor costs are unknown. Furthermore, achieving sustainable intensive agriculture will requirecompensating farmers for ecosystem services they provide. Under a socioeconomic regime that does this,raised-field agriculture could have considerable practical application.

2ccieis2

. Introduction

.1. The search for sustainable agriculture

Over the past 10,000 years, humans have turned an ever-ncreasing part of the biosphere to food production. Humannventiveness in domesticating plants and animals, transform-ng environments and intensifying agriculture has appeared to

llow boundless increases in production. The Green Revolution,eginning in the 1960’s, led to the most recent and most dra-atic increases in agricultural production (e.g., Evenson and Gollin,

∗ Corresponding author at: Université Montpellier II and Centre d’Ecologie Fonc-ionnelle et Evolutive, UMR 5175 CNRS, 1919 route de Mende, F-34293 Montpellieredex 5, France. Tel.: +33 4 67 61 32 32; fax: +33 4 67 41 21 38.

E-mail addresses: delphinerenard@hotmail.fr (D. Renard), J.Iriarte@exeter.ac.ukJ. Iriarte), jago.birk@uni-bayreuth.de (J.J. Birk), stephen.rostain@mae.u-paris10.frS. Rostain), bruno.glaser@landw.uni-halle.de (B. Glaser), d mckey@hotmail.com (D.

cKey).

hi(Icia(tp(

925-8574/$ – see front matter © 2011 Elsevier B.V. All rights reserved.oi:10.1016/j.ecoleng.2011.03.007

© 2011 Elsevier B.V. All rights reserved.

003). However, by replacing internal controls on ecological pro-esses such as nutrient delivery and pest suppression with externalontrols such as fertilizers and pesticides, intensive industrial farm-ng, including the Green Revolution, divorced agriculture fromcology (Robertson and Swinton, 2005; Vandermeer, 2011). Its now well recognized that the Green Revolution brought sub-tantial environmental and social costs (Altieri, 2008; Griffon,002; Robertson and Swinton, 2005). This model of agricultureas degraded ecosystem services upon which life depends, includ-

ng water, carbon and nutrient cycles, and climate regulationRobertson and Swinton, 2005; Perfecto and Vandermeer, 2008).t has favored large-scale cultivators but forced many small-scaleultivators out of farming, helping drive an exodus to cities and anncrease in the number of urban poor (Spencer, 2000). Practicinggriculture as ecological and social ‘sacrifice’ is no longer tenable

Scherr and McNeely, 2008). The challenge of sustainable agricul-ure is to meet the food and fiber demands of a growing globalopulation while at the same time assuring ecosystem servicesHobbs et al., 2008).

l Engi

1

hlvptaaibMecemaodcRc(ah

1i

ced2wecr2oe(hwmdta

ttttpsTmlovmigw

1owaipa2iops

parc2hot

AaoGpf(aiefa2tetnpsFpaMdt

td1lswccaftfit

D. Renard et al. / Ecologica

.2. What science is needed?

The fraction of the earth’s land surface under agriculture isuge and increasing (Robertson and Swinton, 2005). Managing this

and in ways that maintain ecosystem services will require inno-ation in agriculture and in agricultural policy (Pretty, 2008). Theresent research and development climate, however, emphasizeshe genetic modification of crop plants, with much less attentionnd resources directed to agroecological innovation (Vanloquerennd Baret, 2009). There is an urgent need to apply ecology to adapt-ng, designing and managing agricultural landscapes that generateenefits for production, biodiversity and local people (Scherr andcNeely, 2008; Vandermeer, 2011). This will require advances in

cology (Robertson and Swinton, 2005), because very few agri-ultural systems are well understood at the ecosystem level. Onecosystem service (crop yield) has been emphasized to the detri-ent of others (e.g., the maintenance of biogeochemical cycles)

nd many ecological interactions important to the functioningf agroecosystems have garnered much less attention than theyeserve (Gliessman, 2007; Vandermeer, 2011), particularly in theontext of low-input agriculture (Drinkwater and Snapp, 2007).esearch is needed to identify the organisms and ecological pro-esses that play key roles in the functioning of agroecosystemsRobertson and Swinton, 2005) and ways must be found to explainnd promote the value of the services that ecosystems provide forumans.

.3. Folk knowledge, contemporary and past, as a source ofnspiration for agroecological innovation

Successful agroecological solutions are likely to be stronglyase-specific (Doolittle et al., 2002; Pretty, 2008). The folk knowl-dge of smallholder farmers is one source of inspiration for theiverse solutions that will be needed (Denevan, 1995; Altieri,008; Martin et al., 2010; Vandermeer, 2011). Often forced to findays to produce food in unfavorable environments, these farm-

rs have developed ingenious adaptations that do not depend onostly external inputs and that allow long-term use of limitedesources such as land, nutrients or water (e.g., Mollard and Walter,008). Because “traditional” agriculture is often considered to bef low productive potential, the knowledge of smallholder farm-rs is usually ascribed little relevance to solving today’s problemsErickson, 1992). However, productive potential of such systemsas often been underestimated (Vandermeer, 2011). Furthermore,hen comparisons are focused not solely on the production of aarketable commodity but on metrics that integrate total pro-

uction, ecosystem services and positive effects on biodiversity,raditional agriculture wins hands down over modern industrialgriculture (Altieri, 2008).

A subset of this knowledge that has attracted particular atten-ion is the “fossil” folk knowledge of past cultures. Long beforehe origin of Western-style industrial agriculture, many other cul-ures engaged in agricultural intensification, sometimes employingechniques that have virtually disappeared today. History andrehistory offer many examples of diverse trajectories to inten-ification (Denevan, 1995; Erickson, 1992; Guttmann-Bond, 2010;hurston and Fisher, 2006). By the time depth they offer and theultiplicity of cases permitting a comparative approach, archaeo-

ogical studies can provide unique insights into the sustainabilityf agricultural systems, the sources of their resilience, and theirulnerability, both to extrinsic environmental factors such as cli-

ate change (Kemp et al., 2006; Branch et al., 2007) and to factors

ntrinsic to societies (Janssen and Scheffer, 2004). The data sug-est that some of these prehistoric intensive agricultural systemsere sustainable for centuries (Armillas, 1971; Coe, 1964; Denevan,

(1PS

neering 45 (2012) 30– 44 31

995; Erickson, 1992), some having been used apparently continu-usly for over 1000 years (Gliessman, 1991; Mathewson, 1987). Itould be fanciful to suppose that any of these agricultural systems

re panaceas—after all, many have disappeared, and it is obviouslymportant to know the reasons why—or that they could be trans-osed intact into today’s very different social, techno-economicnd environmental contexts (e.g., Siemens, 2000; Lombardo et al.,011). Could the “fossil” local knowledge of extinct cultures, as

nferred from archaeological, geo-archaeological and archaeob-tanical studies, and from the results of rehabilitation experiments,rovide insights useful for devising sustainable agricultural inten-ification in the 21st century?

Archaeological studies of pre-Columbian Latin America haverovided two examples of once widely practiced, but now virtu-lly extinct, agricultural techniques that are considered to haveeal potential for contributing to the design of sustainable agroe-osystems today (Darch, 1988; Doolittle et al., 2002; Lehmann et al.,006; Siemens, 2000). The techniques employed appear not only toave provided novel solutions to strong environmental constraintsn food production, but also had effects on ecosystem functioninghat 21st century humans would consider positive.

The first of these examples is the discovery in forested parts ofmazonia of soils that were amended, intentionally or unintention-lly, with large amounts of charcoal (and with nutrient-containingrganic matter and ash) by pre-Columbian farmers (Glaser, 2007;laser et al., 2001; Glaser and Birk, in press). Their biochemicalroperties make these terra preta and terra mulata soils much moreertile than the oxisols and ultisols from which they are derivedGlaser, 2007; Glaser et al., 2001). Investigation of these soils byrchaeologists, soil scientists and ecologists has spurred burgeon-ng interest in biochar (Atkinson et al., 2010; Glaser, 2007; Glasert al., 2002; Lehmann, 2009; Roberts et al., 2010). In addition to con-erring greater fertility, biochar amendment is attracting interest as

way of storing carbon durably in soils (Glaser, 2007; Glaser et al.,002; Lehmann et al., 2006), thereby removing carbon dioxide fromhe atmosphere. Although the underlying mechanisms are unclear,ffects of biochar on soil biogeochemical processes may also leado reduced emission from soils of other greenhouse gases, such asitrous oxide (Yanai et al., 2007). Experimental studies of “terrareta nova”, inspired by archaeological and geo-archaeologicaltudies, are showing promising results (Steiner et al., 2007, 2008).urthermore, forest on and near abandoned pre-Columbian anthro-ogenic soils is richer in tree species (Clement and Junqueira, 2010)nd in agrobiodiversity (Clement et al., 2004; Junqueira et al., 2010;ajor et al., 2005) than that growing on nearby unmodified soils,

emonstrating the resilience of ecosystems to even several cen-uries of intense human occupation.

The second example of pre-Columbian agricultural techniqueshat have spurred interest in potential applications today is theiversity of forms of wetland agriculture (see Denevan et al.,987; Valdez, 2006; Sluyter, 1994; Denevan, 2001). During the

ate Holocene many seasonally flooded tropical savannas (andome lakeshore habitats) of South America and Meso-Americaere transformed into vast agricultural landscapes through the

onstruction of raised fields by Native Americans. Raised-field agri-ulture provided pre-Columbian farmers with better drainage, soileration, moisture retention during the dry season and increasedertility, as well as possibly easier weeding and harvest. In addi-ion, in some areas channels between raised fields were used forsh and turtle farming and as a renewable source of nutrients forhe soil. Estimates in the Beni savannas of the Bolivian Amazon

roughly the size of the UK) suggest that in this region alone up to

million ha of ancient raised earth platforms were constructed byre-Columbian cultures beginning around 400 BC (Erickson, 2006;aavedra, 2009). The extent of these constructions, which must

3 l Engi

hitiaahtl

tuetBoSneVoi1toB2D(pa1Mcbecoteit(ktrpb[

fimpCada1ttLmtHc

tieaG

1t

ih(affwtoWttcc[h(

ii1a(d1hs1itirisCr12

rWlositlt1Na

2 D. Renard et al. / Ecologica

ave developed over a long period, suggests that this was a form ofntensive agriculture that probably supported dense populations inhe basin of a very active river system over several centuries. Iron-cally, while today about 90% of local communities in this regionre impoverished, much of this ancient agricultural landscape liesbandoned or under-utilized. The contrast between the past, withigh population densities supported by sustainable agriculture, andhe present, with an impoverished, thinly scattered human popu-ation, is striking.

Nordenskiöld (2009 [1916]) was the first 20th century explorero remark upon raised fields, but they attracted no broad interestntil much later. Geographers and archaeologists have studied thecology and use of wetlands in the pre-Hispanic Americas sincehe 1960’s (Denevan, 1966, 1970, 2001, 2006; Luzzader-Beach andeach, 2006). The systems that Native American farmers devel-ped include the famous chinampas of the Basin of Mexico (e.g.,anders et al., 1979). Some form of wetland agriculture, culmi-ating in chinampas, has been continuously practiced in thesenvironments for perhaps 2000 years (Armillas, 1971; Coe, 1964).arious other raised-field systems were developed in the altiplanof Bolivia and Peru (Erickson, 1992, 2003; Kolata, 1996), in marshesn inter-Andean valleys (Bray et al., 1987; Knapp and Mothes,999; Wilson et al., 2002), and in seasonally flooded savannahs inhe lowlands of Mesoamerica and South America. Such savannahsccupy huge expanses of South America, in the Llanos de Mojos ofolivia (Denevan, 1966; Erickson, 1995; Saavedra, 2009; Walker,004), in the Llanos of Venezuela (Zucchi and Denevan, 1972;enevan and Zucchi, 1978; Spencer et al., 1994) and Colombia

Reichel-Dolmatoff and Reichel-Dolmatoff, 1974), and in the Mom-os Depression in Colombia (Plazas and Falchetti, 1990). Smallerreas are present in the Mexican states of Veracruz (Siemens et al.,988) and Tabasco (Gliessman, 1991), in the Maya lowlands ofexico and Belize (Gliessman et al., 1985; Sluyter, 1994), in the

oastal savannas of the Guianas (Rostain, 1994, 2008), in the Guayasasin of southeastern Ecuador (Stemper, 1987) and even in south-rn regions of Chile (Dillehay et al., 2007). Characterized by annualycles including rainy-season flooding, dry-season drought, andften widespread fires, many of these areas are thinly inhabitedoday, as noted above, and some are considered suitable only forxtensive cattle grazing. Raised-field agriculture was conductedn a great range of environments, from the highland plateaux ofhe Andes (Erickson, 1992; Kolata, 1996) and inter-Andean valleysBray et al., 1987; Wilson et al., 2002) to the hot lowlands, from thearstic Yucatan peninsula (Dunning et al., 2002) to the acid soils ofhe Llanos de Mojos (Boixadera et al., 2003; Hanagarth, 1993). Yetaised-field agriculture had virtually disappeared by the colonialeriod, with only scattered anecdotal descriptions of its techniquesy a few early chroniclers (De Las Casas, 1986 [1560]; Gumilla, 19631791]).

Understanding the reasons for the abandonment of raised-eld agriculture in different regions of the Americas requiresore detailed archaeological and paleoecological research in each

articular region. However, in several regions the impact of post-olumbian diseases on agricultural populations may have played

major role in the demise of raised-field agriculture. These epi-emics and pandemics were arguably the most rapid, thorough,nd widespread to have occurred during the late Holocene (Crosby,972; Lovell, 1992). According to the estimates of some authors,hey may have resulted in the loss of as much as 80-95% ofhe agricultural population across the Neotropics (Dobyns, 1966;ovell and Lutz, 1995). Labor-demanding raised-field agriculture

ust have been significantly impacted by such substantial reduc-

ions of the labor force. In other cases, however (for example,ertenrits in coastal Suriname [Versteeg, 2008]), raised-field agri-ulture was abandoned several centuries earlier. In most sites,

drcr

neering 45 (2012) 30– 44

he timing and the cause(s) of abandonment of raised-field farm-ng are still unknown. Apart from the chinampas, with only fewxceptions raised fields are no longer used today by Amerindi-ns (Denevan and Schwerin, 1978; Grenand, 1981; Crews andliessman, 1991).

.4. Advantages of raised-field agriculture and its pertinence tooday’s problems

How pre-Columbian farmers managed to cultivate seeminglynhospitable wetland environments, sometimes for centuries,as long fascinated archaeologists, geographers and agronomistsDenevan, 1970; Lambert et al., 1984; Morris, 2004). Raised-fieldgriculture has been suggested to have several positive effects onood production. Nutrients were concentrated on cultivated sur-aces whose topsoil was made deeper with amendments. Wateras managed to avoid flooding and sometimes to promote irriga-

ion; embankments were sometimes constructed around groupsf raised fields to protect them from river overflow (Erickson andalker, 2009). The wetland component of the system is postulated

o have played roles in supplying nutrients to crops and some-imes food for people (Erickson, 1992, 2000; Kolata, 1996). Finally,rops were in some instances provided with more favorable micro-limates (greater frost tolerance in cold highland environmentsErickson, 1992; Kolata and Ortloff, 1989]). Researchers have alsoighlighted the apparent sustainability of raised-field agricultureCrews and Gliessman, 1991; Denevan, 1995).

This interest in raised-field agriculture has led to several exper-mental studies aimed at understanding how it worked and testingts proposed benefits (Gliessman, 1991; Erickson, 1995; Kolata,996; Barba et al., 2003; Morris, 2004; Saavedra, 2009), or eventtempting to rehabilitate it as a viable system of food productionMuse and Quinteros, 1987; Morris, 2004). These studies have oftenemonstrated high yields per unit land cultivated (Sanders et al.,979; Arce, 1993; Muse and Quinteros, 1987; Saavedra, 2009) andave shown how raised-field agriculture can enhance ecosystemervices such as nutrient retention in wetlands (Kolata and Ortloff,989; Biesboer et al., 1999; Carney et al., 1996), a topic of great

mportance today. Despite these demonstrated effects, however,here is still controversy about the relevance of raised-field farm-ng to agriculture in the 21st century. Although the productivity ofaised-field agriculture in terms of yield per unit land cultivateds demonstrated, the most extensive projects to rehabilitate theseystems are considered by their critics to have failed (Bandy, 2005;hapin, 1988). Meanwhile, proponents of many of these projectsemain optimistic about their potential (Denevan, 2001; Gliessman,991; Saavedra, 2009), and new projects are underway (Saavedra,009).

Why is there such uncertainty in interpreting the prospects forehabilitating raised-field agriculture for food production today?

e believe there have been two major problems. The first is aack of insight into how raised-field systems functioned in termsf their agroecology. This problem can be broken down into twoub-problems. (i) A failure to appreciate the diversity of ecolog-cal situations in which wetland agriculture was conducted, andhe corresponding diversity of adaptations and constraints, hased to overgeneralization and sometimes to transposition of his-orical models into inappropriate present-day contexts (Chapin,988; Lombardo et al., 2011). (ii) A failure to look beyond theeotropics has blinded workers to the existence of present-daynalogues of these systems that are demonstrably not only pro-

uctive but also economically viable, and that could help provideealistic guidelines for attempts to reconstruct raised-field agri-ulture in the Americas. The second major problem is that theesearchers and farmers involved in these experiments have had

l Engi

thtttocpemtmHwtoApfEtceet2

srdscseitlafiatiftcIcc

2a

2

stbaoiheh

2s

DHftbM(vt2vefflaRadvltlotiadrsvcLfcfi(

2a

sdnpetvogisktopK

D. Renard et al. / Ecologica

o confront sociocultural and economic contexts over which theyave little control, and which have often been highly unfavorableo the success of the experiments. Critics have emphasized whathey consider a lack of congruence between raised-field farmingechniques and today’s social, cultural and economic conditions,r even a failure on the part of the experimenters to take theseonditions into account, but the reality appears to be more com-lex. For example, Chapin (1988) notes that one Mexican attempt tostablish raised-field agriculture as a food-production system ulti-ately failed not because of agroecological constraints (although

hese were formidable) but because no provisions were made toarket the harvest, which rotted or was sold at give-away prices.e did not mention, however, that the experiment in questionas a state-run, large-scale, poorly planned top-down enterprise

hat tried to imitate a smaller experiment designed by agroecol-gists and conducted with more effective farmer participation.ccording to its designers, this smaller project was producingromising results before being interrupted by reduction of itsunding when the large-scale projects began (Gliessman, 1991).xperiments faced with such problems can hardly be expectedo yield conclusive results on the feasibility of raised-field agri-ulture today. Similar problems are faced by any move to adoptcological agriculture in today’s world, where both local knowl-dge and rural societies have been weakened or destroyed byhe hegemony of the industrial agricultural model (Vandermeer,011).

Rebuilding knowledge requires identifying gaps in our under-tanding. We identify several major gaps in our understanding ofaised-field agroecosystems, by analyzing their functioning at threeifferent levels. We first examine their ecological functioning at thecale of the local plot: How did humans engineer environments byonstructing and maintaining raised fields and the wetland matrixurrounding them, and what were the effects of this engineering oncological interactions in active wetland farms? Second, we exam-ne the system’s functioning at a larger scale of space and time,hat of the landscape over numerous cultivation cycles. Were fal-ows required? If they were, how did human engineering activitiesffect ecological interactions in fallows, and how did the integratedeld/fallow cycle function? Whether raised-field farming can actu-lly contribute to solving human problems depends on combininghese “hard agroecology” approaches with “soft agroecology”, tak-ng into account the sociocultural and economic context in whicharming activities are embedded (Dalgaard et al., 2003). Thus, in thehird and final part of this essay, we ask whether raised-field agri-ulture is adapted to today’s social and techno-economic context.n this third part we focus on the one aspect that is most frequentlyited as a factor likely to limit the pertinence of raised-field agri-ulture today, its labor-intensive nature.

. Towards a systems-level understanding of raised-fieldgriculture

.1. Raised-field agriculture as ecological engineering

Wetland agriculture offers great scope for engineering, becauselight variations in elevation dramatically affect drainage, which inurn affects the availability of water and oxygen, and thereby all theiological, chemical and physical processes that determine nutrientvailability (Moser et al., 2009). Moving relatively small amounts

f earth can thus create a mosaic of microenvironments differingn their ecological functioning. Farming systems can exploit thiseterogeneity in diverse ways. Furthermore, once created, the het-rogeneity produced by human earthmoving activities is likely toave long-lasting ecological effects.

(Iewe

neering 45 (2012) 30– 44 33

.1.1. Morphological diversity of raised fields and its functionalignificance

Raised fields exhibit great morphological diversity among (e.g.,enevan, 2001), and even within sites (e.g., McKey et al., 2010).owever, their initial height appears everywhere to have been

airly similar. Although mounds in many sites have been par-ially eroded, both archaeological data (e.g., profiles of raised fieldsuried by ash from the eruption of Quilotoa in Ecuador [Knapp andothes, 1999; Wilson et al., 2002]) and contemporary analogues

Vasey et al., 1984) show that planting surfaces were usually ele-ated about 0.5–1 m above the mean rainy-season water level ofhe floodplain or swamp. However, some raised fields are up to

m high (Darch, 1988) or even up to 3 m (Vasey et al., 1984). Suchariation could be related to local expectation of extreme floodvents, but no information exists to evaluate this hypothesis. Plat-orms varied in size, shape and orientation. Typical shapes rangerom linear and curvilinear ridges to square, rectangular or inter-ocking platforms, but round, rectangular and multi-sided fieldsre also known (Darch, 1988; Denevan, 1970; McKey et al., 2010).aised fields are often grouped in parallel series to form ladder-nd checkerboard-like arrangements, which are sometimes bor-ered by ditches or embankments. Size of individual fields alsoaries enormously. Ridged fields may be up to 25 m wide and 100 mong or longer and some round mounds are only 1–1.5 m in diame-er. Organization and orientation of fields appear engineered to suitocal hydrological conditions (Darch, 1988). For example, ridges areriented parallel to slopes to facilitate drainage and perpendicularo slopes to enhance water retention (McKey et al., 2010). Similarly,n some areas of the Beni, raised fields were built in flat areas (suchs old lake beach ridges or levees) and ditched fields, promotingrainage, on gentle slopes (Lombardo et al., 2011). In some cases,aised fields are combined with dykes, sluices or levees that gaveome control over the general water level (Darch, 1988). Canals ofarying widths, permanently or seasonally filled with water, werereated between the platforms. Depth of canals also varied. Aroundake Titicaca, most canals were at least 1 m below the present sur-ace and some were up to 2 m deep (Erickson, 1992). In other sites,anals were absent, or almost so, and the spaces between raisedelds constituted a shallow, usually only seasonally flooded, matrixMcKey et al., 2010).

.1.2. Sources of information about how raised-fieldgroecosystems functioned

Whereas the morphology of pre-Columbian raised-field land-capes has been extensively described, few studies provide actualata on their formation, chronology and use: “even for the chi-ampas of the Basin of Mexico, only four studies actually presentrimary data recovered through scientific investigation” (Beacht al., 2009). Twenty-five years ago, Vasey et al. (1984) lamentedhat while the morphology of raised fields had been well described,ery little work had been devoted to the ecological functioningf raised-field systems. Despite the progress reviewed here, largeaps remain. Data on processes underlying the productive capac-ty of raised-field systems, and the environmental services they areupposed to have ensured, are scarce and fragmentary. What wenow, or can surmise, about how raised-field farming functioned inhe past comes from three sources of information. First, archaeol-gists, geographers, geoarchaeologists and archaeobotanists haverovided some data on how these systems were constructed (e.g.,olata, 1993; McKey et al., 2010), how their soils were managed

Wilson et al., 2002), and what crops were grown on them (e.g.,

riarte et al., 2010; McKey et al., 2010; Pearsall, 1987; Siemenst al., 1988; Turner and Miksicek, 1984). Second, archaeologists,orking with ecologists, agronomists, soil scientists and local farm-

rs, have conducted experiments aimed at replicating raised-field

3 l Engi

ftETtmcCtt11aWwr

aogvtfw2st(eafitNceAcgrfrpaptant2

2

tenTerababis

2

pflftrPrsatTon

2

forWppFttbDioimom

wecl(meewLNatslBscAacoamb

4 D. Renard et al. / Ecologica

arming to study its workings and in some cases to rehabili-ate it as a viable method of food production (Erickson, 1992;rickson and Candler, 1989; Gliessman, 1991; Saavedra, 2009).hird, geographers and agronomists have reported data on con-emporary farming systems that are in at least some respects

odern analogues of ancient raised-field systems. Apart from thehinampas, which persist today in much altered form near Mexicoity (Torres-Lima et al., 1994), and less altered systems similaro chinampas in Tlaxcala, Mexico (Crews and Gliessman, 1991),he closest analogues are found in southern China (Kleinhenz,997; Luo and Lin, 1991), SE Asia (IIRR, 1990) and Oceania (Kirch,978). With the exception of a few early papers (e.g., Denevannd Turner, 1974; Vasey et al., 1984), the literature on these Old-orld systems appears to have been completely ignored by thoseorking on Neotropical raised fields and the prospects for their

e-establishment.Each of these data sources presents strong limitations. Whereas

rtifacts and ecofacts studied by modern archaeological meth-ds do reveal a wealth of information, one always wishes forreater resolution and richer detail. Experimental studies can pro-ide these, but there is always the possibility that environments,echniques, labor organization, plant material or all of these dif-er in important ways from those that existed centuries beforehen fields were being used (Gondard, 2006; Lombardo et al.,

011). Furthermore, some experiments in re-establishment haveuffered from methodological shortcomings that limit their per-inence as tests of the feasibility of raised-field agriculture todayLombardo et al., 2011). Finally, no contemporary systems arexact analogues of prehistoric systems. Agricultural systems in Asiand Oceania that combine cultivation of upland crops on raisedelds and of wetland crops (usually rice or taro) in ditches offerhe closest parallels. Pre-Columbian raised-field farmers in theeotropics appear to have had no waterlogging-tolerant wetlandrops similar to rice or taro. Like other “orphan systems” of marginalnvironments (Mollard and Walter, 2008), raised-field systems ofsia and Oceania, and some similar systems in Africa (e.g., riceultivation in reclaimed mangrove areas by the Diola of Sene-al, Pélissier, 2008 [1966]), have been comparatively neglected byesearchers, but the literature that does exist suggests they con-er distinct advantages (IIRR, 1990; Kleinhenz, 1997). Even paddyice farming systems, part of the life-support system for a largeart of the world’s population, were considered poorly knowns ecosystems 25 years ago (IRRI, 1985). Roger (1996) has sinceublished a remarkable synthesis of the considerable informationhat does exist. His treatment, emphasizing ecological interactionsnd how they are affected by agricultural intensification, suggestsumerous questions for research on Neotropical raised-field sys-ems. Some of these are addressed below (see Sections 2.2.3.2 and.2.3.3).

.2. Agroecological functioning at the field level

What factors constrained food production in raised-field sys-ems, and how did farmers manage them? What were thenvironmental consequences of management, and what mecha-isms underlay crop production and other ecosystem services?he answers to these questions are probably as diverse as thenvironments in which raised fields were constructed. These envi-onments range from sea level to over 4000 m elevation, covering

great range of soils and climates. Some are in seasonally floodedasins, others in permanently wet habitats. Three key factors man-

ged by farmers to ensure production in raised-field systems haveeen identified, but their relative importance and the underly-

ng mechanisms are likely to be highly variable among differentystems.

gr1t

neering 45 (2012) 30– 44

.2.1. Frost reductionThis first factor could have been important only in the Alti-

lano. Tropical montane habitats undergo extreme temperatureuctuations between day and night. At high elevations, nocturnal

rosts may kill crops, even Andean species such as potatoes. Owingo its high thermal capacity, water acts as a temperature buffer,e-radiating during the night heat energy stored during the day.roximity of fields to bodies of water, such as the canals betweenaised fields, could thus protect crops against frosts. Knapp (1988)howed that minimum night-time air temperatures above canalsnd above raised fields around Lake Titicaca were 1–2 ◦C higherhan above areas without raised fields at similar elevation nearby.he magnitude of this effect should increase with increasing ratiof water to land in the zone. This may explain why ridge fields werearrow and canals were broad in this system (Gondard, 2006).

.2.2. Water managementFarmers constructed raised fields to create drained platforms

or planting crops, but their earthmoving activities often hadther additional purposes and consequences. In some regions,aised fields were associated with canals, causeways (Erickson and

alker, 2009), reservoirs, habitation mounds and structures inter-reted as fish weirs (Erickson, 2000). Water management thuslayed multiple roles in the strategies of some of these peoples.ish and other aquatic organisms are believed to have been impor-ant protein sources in some areas (Darch, 1988; Erickson, 2000), ashey are in some Asian paddy rice and mixed upland/wetland high-ed systems today (Guo and Bradshaw, 1993; Luo and Lin, 1991).espite the elaborate morphology of pre-Columbian earthworks

n some sites, it is important to note that most were dependentn rainwater, groundwater, or lakeshore habitats, with only lim-ted contribution of river water (Morris, 2004). There was thus in

any systems little need—or opportunity—for top-down controlf water flow and allocation; each group of fields could thus beanaged largely independently of others (Erickson, 2003).In addition to drainage, management also aimed at conserving

ater for long- and short-term droughts, using the stored water toxtend growing seasons (Erickson, 1992). In some sites, ditches andanals were deep enough to have exposed groundwater through ateast part of the dry season (Erickson, 1992), and hand irrigationsplash or bucket), as practiced today in the chinampas and other

odern analogues (Crews and Gliessman, 1991; Vasey et al., 1984),nabled cultivation into the dry season (Erickson, 1992). Work inxperimental raised fields has demonstrated the advantages suchater management could confer. Experimental raised fields near

ake Titicaca gave some yield during the severe 1982–1983 Elino drought, when other types of farms nearby yielded nothing,nd gave excellent yields in 1985–1986, when flooding devas-ated other types of farms (Erickson, 1992). Extending the growingeason may have been particularly important in raised-field sitesocated in hyperseasonal savannas, such as those in the Llanos ofolivia and Venezuela and the coastal savannas of the Guianas. Sea-onally fluctuating water levels must have conditioned agriculturalycles, with water shortage precluding farming in the dry season.s water levels receded, raised fields could have increased watervailability to crops by two mechanisms. First, high organic matterontent in mound soils, which would be expected under continu-us mulching (see next section), could maintain greater moisturevailability. Second, mound soils could receive water by capillaryovement (“wicking action”) from adjacent canals or the flooded

asin (Kleinhenz et al., 1996b). It has also been speculated that

roundwater could rise by capillary movement to infiltrate theoot zone of elevated planting surfaces from beneath (Erickson,992). This phenomenon is exploited by engineers in “subirriga-ion” systems (Allred et al., 2003). Although it has frequently been

l Engi

cs(ftsewe2wrs

phbhnbbatystfiearwgiwstirf

ostt(wDswbTapletPsa1us

ts

ia

2

eagfid(IcaadbWct

2ahwsgfcAmaprhhcbiFa2tifwvtT1rFcw

pmttc(

D. Renard et al. / Ecologica

onsidered important in Neotropical wetland agriculture, only onetudy has attempted to study this process in any detail. Crossley2004), working in modern-day chinampas, found little supportor an important role of subirrigation processes and noted “theremendous complexity of the soil and plant properties required forubirrigation to occur naturally, and function significantly”. How-ver, capillary rise is known to be an important factor affectingater availability to plants in a wide range of cultivated (Ayars

t al., 2006; Nosetto et al., 2009) and natural ecosystems (Döll et al.,003; Rodriguez-Iturbe et al., 2007) with shallow water tables, ande would find it surprising were it not to play an important role in

aised-field agroecosystems. This topic certainly deserves deepertudy.

Studies examining the hydraulic properties of soils of contem-orary high-bed systems in Asia suggest functional links betweenydrology and the optimal design of raised-field systems. Raisededs facilitate drainage when water level is high (by allowingorizontal removal of excessive soil water when vertical flow isegligible) and furrows increase the supply of water to plants oneds (by “wicking action”, i.e., capillary flow) when conditionsecome dry. Both effects are weaker in the center of beds, whichre quickly saturated in the rainy season and quickly dry out inhe dry season (Kleinhenz et al., 1996b). This helps explain whyield decreases from the edge to the center in some raised-fieldystems (Kleinhenz et al., 1996a). These processes should affecthe optimal dimensions of raised fields. If space allocation to raisedelds is to be maximized, they should be as wide as practical. How-ver, if soils are inundated for long periods by continuous rainfallnd if soil water is deficient during periods of limited rainfall, nar-ower fields are called for. Both effects of raised fields (drainage andicking) would be maximized in small, round raised fields, sug-

esting that this form would be optimal where seasonal extremesn flooding and drought are most marked. Data on flooding regimes

ithin and among different regions of pre-Columbian raised-fieldites are not of sufficient precision to test for the suggested correla-ion. However, the Llanos de Mojos, where large seasonal variationsn water level are known to occur, are not characterized by smallound raised fields; long, broad ridges appear to be much morerequent.

Perusal of the literature on wetland ecology leads to numerousther questions about hydrological aspects of raised-field con-truction and management. Raised-field construction increasedhe soil surface roughness of previously flatter landscapes. Didhis increase the depression storage capacity of the affected areasDarboux et al., 2002)? Many of the basins in which raised fieldsere constructed show great seasonal variation in water levels.id farmers cultivate different parts of the landscape at different

easons? It has usually been assumed that only the raised surfacesere cultivated (but see Parsons and Shlemon, 1987). Observations

y Gliessman (1992) among contemporary indigenous farmers inabasco, Mexico, call this assumption into question. He described

system in which maize is planted on higher ground around flood-rone areas during the wet season, then on lower ground as water

evels drop during the dry season. Pre-Columbian raised-field farm-rs could have similarly exploited the topographic heterogeneityhey created to extend cropping seasons in raised-field agriculture.resent-day Asian raised-field systems such as the sorjan croppingystem in Indonesia (Domingo and Hagerman, 1982; IIRR, 1990)nd permanent-high-bed systems in southern China (Kleinhenz,997; Luo and Lin, 1991) exploit such heterogeneity by growingpland crops, usually vegetables, on raised fields during the rainy

eason and wetland rice in the surrounding furrows or canals.

Considering the importance of hydrology in raised-field sys-ems, the absence of detailed treatments (and particularly ofurface water–groundwater interactions) constitutes a major gap

Ilrm

neering 45 (2012) 30– 44 35

n efforts to understand the functioning of these systems and topply this knowledge in agriculture today.

.2.3. Nutrient managementRaised-field farmers added nutrient-containing materials to the

levated planting surfaces they created. Observations of modernnalogues, and in some cases inferences from archaeological data,ive indications of the kinds of material that were added to raisedelds as nutrient sources. However, little work has examined theynamics of soil–crop–nutrient relations in any detail. Vasey et al.1984) noted this over 25 years ago and it is still largely true today.n raised-field environments, well-aerated and waterlogged soilompartments are both present, and vary in distribution in spacend time. Aquatic and terrestrial environments, and their ecotones,re all part of the whole system. These peculiarities should intro-uce complexity in nutrient dynamics in raised-field agriculture,ut we know very little about this aspect of their functioning.e will first review what is known about the kinds of nutrient-

ontaining amendments added to raised fields and then explorehe processes that may underly soil–crop–nutrient dynamics.

.2.3.1. Different types of materials added to raised fields. Materialsdded to raised fields likely came from several sources. Some mayave been imported from outside the raised-field system. Kitchenaste, fish remains, ash and charcoal from hearth fires are potential

ources, although these were more likely to be important in homeardens than in extensive raised-field systems relatively distantrom habitation sites. Farm animal manure is used in present-dayhinampas, and formerly human excrement was, too (Coe, 1964;rmillas, 1971). Wilson et al. (2002) interpreted some soil micro-orphological features as the possible result of animal manure

pplication in raised fields in Ecuador. Erickson (1994) noted theresence of abundant carbonized seeds of wild plants in ancientaised-field soils near Lake Titicaca. He postulated that these couldave been derived from burnt llama dung, which these people mayave used for cooking. However, as noted by Gondard (2006), thearbonized plant material could also have been produced by theurning of weeds gathered from agricultural fields, or by the burn-

ng of fallow vegetation (see Section 2.2.4). Finally, in one site inrench Guiana, farmers appear to have strip-mined topsoil fromdjacent areas and deposited it onto raised fields (McKey et al.,010). Raised fields in Cano Ventosidad in the Llanos Occiden-ales of Venezuela were also constructed with material broughtn from elsewhere, not from the canals near these raised sur-aces (Gondard, 2006). Other materials were derived from recyclingithin the raised-field system. Crop residues, for example, are

irtually universally incorporated into fields in analogous con-emporary systems (Crews and Gliessman, 1991; Denevan andurner, 1974; Domingo and Hagerman, 1982; IIRR, 1990; Kirch,978; Kleinhenz, 1997), and there is some evidence for incorpo-ation of maize leaf residues into pre-Columbian raised fields inrench Guiana (McKey et al., 2010; Iriarte et al., 2010). Addition ofrop residues as mulch confers the added advantage of suppressingeeds (Boucher et al., 1983; IIRR, 1990; Kirch, 1978).

In some systems, the most important source of nutrient-roviding material added to raised fields may have been the organicatter produced and accumulated in the aquatic component of

hese systems. Transferring vegetation and bottom sediments tohe cropping area is an inevitable consequence of maintaininganals, and such transfer is universal in extant analogous systemsCrews and Gliessman, 1991; Denevan, 2001; Vasey et al., 1984).

n present-day chinampas, organic matter-rich sediments accumu-ate in the wetland part of the system, particularly during eachainy season, and this muck is added to fields (Darch, 1988). Thisuck includes crop residues and other material lost in runoff from

3 l Engi

raantpirfammmtaaiFtrtdTcrntrifisrsp1a

2ttfnflgloeflbisfapaptmv(rtmpt

itdloofiagpGoucscwisp2

aesacidizatnfw

aeawitt

2Tyaa2absrei(Ag

6 D. Renard et al. / Ecologica

aised fields and material derived from aquatic plants. Macrophytesnd algae were abundant in canals of experimental raised fieldst Tiwanaku (Carney et al., 1993) and water hyacinth (Eichhor-ia crassipes) is abundant in the seasonally flooded savannas ofhe Llanos de Mojos (Gondard, 2006; Lee, 1997). Various macro-hytes are abundant in the contemporary chinampas-like system

n Tlaxcala, Mexico, and contribute to formation of mulch added toaised fields (Crews and Gliessman, 1991). The top 70 cm of coresrom ancient raised fields in Pulltrouser Swamp include significantmounts of water lily (Nymphaea sp.) pollen, substantiating theanuring hypothesis (Wiseman, 1983 in Darch, 1988). Macroalgae,icroalgae and aquatic and amphibious plants produce litter withuch lower C:N and C:P ratios that is decomposed more rapidly

han that produced by other plants (Enriquez et al., 1993). Bothlgae and aquatic macrophytes exhibit “luxury consumption” of Nnd P, assimilating these nutrients in excess of their needs and stor-ng them for use under nutrient-deficient conditions (Roger, 1996).inally, the aquatic compartment of raised-field systems may con-ain nitrogen-fixing bacteria and cyanobacteria. In experimentalaised fields around Lake Titicaca, nitrogen fixation by cyanobac-eria in the flooded canals furnished nitrogen- and mineral-richetritus that could be added to fields (Biesboer et al., 1999).he aquatic fern Azolla, associated with symbiotic nitrogen-fixingyanobacteria (Anabaena), is frequent in slack-water canals of someaised fields (Vasey et al., 1984) and in the seasonally flooded savan-as of the Llanos de Mojos (Gondard, 2006). The importance ofhe Azolla/Anabaena symbiosis to the nitrogen economy of paddyice systems has been demonstrated (IRRI, 1985; Roger, 1996), andt appears likely to have played a significant role in some raised-eld systems as well (Biesboer et al., 1999). Another nitrogen-fixingymbiosis, the actinorhizal association between Frankia and theiverine tree Alnus, is important in chinampas and in the similarystem in Tlaxcala. Alnus trees, frequently planted near the chinam-as platforms, supply nitrogen-rich litter (Crews and Gliessman,991). Litter from leguminous crops planted on platforms couldlso enhance nitrogen availability (Crews and Gliessman, 1991).

.2.3.2. Importance of nutrients from aquatic resources. However,he importance of aquatic resources as a nutrient source is likelyo have varied widely among systems, depending on at least fouractors. First, the potential was probably much greater in perma-ently flooded marsh and lakeshore habitats than in seasonallyooded wetlands (Lombardo et al., 2011). Fish, turtles, macroal-ae and aquatic macrophytes can persist only if the basin (or ateast some canals) is filled with standing water year-round. Withnly seasonal flooding, aquatic life is much more limited. How-ver, seasonally flooded basins do support microalgae, as well asooding-tolerant sedges and grasses, and biomass from these coulde added to raised fields. A second factor influencing the potential

mportance of the wetland component is the ratio of water to landurface. If the area covered by water is large relative to the sur-aces of planting platforms, greater nutrient concentration can bechieved (Vasey et al., 1984). Thus, in Altiplano raised fields, narrowlatforms and broad canals may have conferred not only a thermaldvantage but a fertility advantage as well. A third factor is theresence of uncultivated land near fields that can be “mined” foropsoil to supplement the cultivated surfaces (Vasey et al., 1984). If

aterial can be extracted from a sufficiently large area of unculti-ated land around the field complex and brought to cropping areassee for example McKey et al., 2010), the water-covered area sur-ounding fields may be irrelevant as a source of nutrients. Finally,

he productivity of the aquatic compartment is likely to vary enor-

ously among environments where raised-field agriculture wasracticed. Productivity of aquatic systems—and thus their impor-ance as a source of nutrients for raised fields—is likely to be higher

iaea

neering 45 (2012) 30– 44

n areas with relatively nutrient-rich young volcanic soils (such ashose in the Andes, the Valley of Mexico, and Tlaxcala), or soilserived from calcareous sedimentary rocks (e.g., in the Maya low-

ands), than in areas with old, highly weathered soils like thosef seasonally flooded South American savannas. The importancef nitrogen fixation is likely to show parallel variation. Nitrogenxation is energetically advantageous only if levels of phosphorusre adequate (Vitousek and Howarth, 1991). Furthermore, nitro-en fixation usually occurs at low rates in acid soils. In somerehistoric raised-field complexes in coastal savannas of Frenchuiana, the paucity of phosphorus and nitrogen is suggested by theccurrence of two genera of carnivorous plants, Drosera and Utric-laria (D. Renard et al., unpublished data). Under such oligotrophiconditions, the aquatic compartment may be less important as aource of nutrients for amending raised fields. There may also beonsiderable variation in fertility within raised-field regions. Thehite-water (sediment-rich) Mamore River runs from its source

n the Andes through the western part of the Beni, its overflowometimes affecting raised fields, whereas blackwater (sediment-oor) rivers run through the Beni’s eastern sector (Lombardo et al.,011).

Farmers in regions where the waters surrounding raised fieldsre oligotrophic may have been able to render them moreutrophic. For example, lower soil horizons are less acidic thanurface horizons, and deep ditches cut into them could reduce thecidity of water in ditches, facilitating nitrogen fixation if otheronditions allow it (Vasey et al., 1984; Gondard, 2006). However,t is uncertain how long this effect would last, because water initches would become more acidic as organic matter accumulates

n them and bases are leached out of newly exposed soil hori-ons (Vasey et al., 1984). Another way by which farmers in suchreas could render the aquatic compartments of raised-field sys-ems more eutrophic—and thereby enhance their role in recyclingutrients to field surfaces—is to import large quantities of nutrients

rom outside sources into the basins where raised-field agricultureas conducted.

However, work in paddy-rice systems shows that nitrogen fix-tion by bacteria and cyanobacteria can also play significant rolesven in acid soils (Roger, 1996). Furthermore, both pH and nutrientvailability can show great variation with seasonal drying and re-etting cycles (Roger, 1996), which could drive seasonal patterns

n nitrogen fixation. These and other complexities likely to charac-erize nutrient dynamics in raised-field agriculture are treated inhe following section.

.2.3.3. Nutrient dynamics in aquatic and terrestrial compartments.he ultimate determinant of plant performance and (in crop plants)ield is not the total amount of nutrients, but their availability. Theerobic or anaerobic status of environments influences the avail-bility of different nutrients in complex fashion (Chepkwony et al.,001; Kirk, 2004; Vasey et al., 1984). Raised-field systems combineerobic and anaerobic environments, and material is transferredetween the two. All of these facts should lead to very complexoil–crop–nutrient dynamics, but this aspect of the ecology ofaised-field agriculture is virtually unstudied. Literature on othercosystems with frequently waterlogged soils—paddy rice systemsn Asia (IRRI, 1985; Timsina and Connor, 2001), peatlands in EuropeOlde Venterink et al., 2002), flooded pampas in southern Southmerica (Rubio et al., 1997)—suggests the interest of closing thisap in our knowledge.

Phosphorus and nitrogen, the nutrients most likely to be limit-

ng to plant growth, also have the most distinctive behaviour undernaerobic conditions (Vasey et al., 1984). Waterlogging usuallynhances the availability of phosphorus, making it more solublend more diffusible, and flushing it from the organic P pool when

l Engi

sewayctioactplafReafbflthItlcl

N(tianaoan

csTeBrceeec

fipOesimlmebv

amti

scacwtumwbflapbtdd

pnac

2t

rtficciippliEBentarcptemwitsrnl

D. Renard et al. / Ecologica

oil microorganisms are killed under anaerobic conditions (Rubiot al., 1997). Low P availability is a frequent problem in highlyeathered ultisols and oxisols, where P is adsorbed to iron and

luminum oxides. Low P availability is also a problem in a family ofoung volcanic soils, andisols, where P is adsorbed to allophanelays. We could find no information on how anaerobic condi-ions affect this sorption. In all these soils, waterlogging could alsoncrease P availability simply by maintaining a large proportionf P in undecomposed organic matter, nutrients in which becomevailable to plants when muck is deposited on raised fields, whereonditions are aerobic. Waterlogged soils also offer much less resis-ance to root penetration, allowing plants to produce a greaterroportion of fine roots that more efficiently acquire nutrients of

ow mobility, such as P (Rubio et al., 1997). Plants that are toler-nt of waterlogged conditions may thus be able to extract more Prom soils than can plants growing in better-drained environments.ice is a good example. Its waterlogging-tolerant roots extract Pfficiently, reducing the need to apply P to wetland rice (Timsinand Connor, 2001). While the crops pre-Columbian raised-fieldarmers grew were insufficiently adapted to waterlogging to haveenefited directly from such effects, wild plants growing in theooded matrix could have benefited, increasing the overall quan-ity of P cycling in the system. Detritus from these plants couldave provided relatively P-rich mulch for addition to raised fields.

ncreased P availability through such mechanisms could be impor-ant, because P is often limiting to plant growth, and soils of ateast some abandoned raised-field complexes have generally lowontents of P (e.g., French Guiana, J.J. Birk and Bruno Glaser, unpub-ished results).

The dynamics of nitrogen add another level of complexity.itrogen availability is usually negatively affected by waterlogging

Timsina and Connor, 2001). Nitrogen accumulates in organic mat-er deposited in anaerobic conditions, and when this organic matters transported into aerobic conditions, nitrogen is rapidly miner-lized to nitrate. These dynamics are important in the supply ofitrogen to crops on raised fields (Vasey et al., 1984). As mentionedbove, biological fixation of atmospheric nitrogen is usually favorednly in systems where phosphorus is not limiting. Increased P avail-bility in raised-field systems should thus also increase the rate ofitrogen fixation.

Nutrient dynamics in raised-field agriculture not only affectrop production, they also affect the ecosystem services that can beupplied by these systems. In experimental raised fields near Lakeiticaca, Carney et al. (1993) showed that the efficient use of nutri-nts by these systems had important environmental consequences.y retaining nutrients and sediments, rehabilitated raised fieldseduced water pollution downstream. Raised-field agriculture thusonferred environmental benefits in addition to the economic ben-fits from crop production and maintenance of soil fertility. Thisxample suggests that further studies of these farming systems ascosystems could be rich with insights on how to reconcile agri-ultural productivity with ecosystem services.

Among the biggest gaps in our understanding of how raisedelds may have functioned as agroecosystems is the virtually com-lete absence of work on microbial communities in these systems.nce again, work on paddy-rice systems (Roger, 1996) suggestsxciting possibilities. Microbial biomass is an important potentialource of nutrients for plants, and flooding leads to an increasen total microbial biomass, in large part attributable to aquatic

icroalgae. However, nutrients in living microbes are immobi-ized and unavailable to plants, becoming available only when the

icrobes die and the nutrients they contain are mineralized. Nutri-nt release thus depends on microbial turnover. This idea forms thease of one explanation of the “Birch effect”, i.e., the frequent obser-ation that nutrient release increases in soils subjected to drying

cmva

neering 45 (2012) 30– 44 37

nd re-wetting cycles (Jarvis et al., 2007). According to this idea,icrobes that die as the soil dries are rapidly decomposed (and

heir nutrients mineralized) as decomposer populations rapidlyncrease with re-wetting.

The “Birch effect” and similar phenomena could be particularlyignificant in raised-field systems, where a diversity of mechanismsould enhance microbial turnover, mineralization and nutrientvailability to plants. In these systems, seasonal nutrient flushesould occur not only when dry soils are re-wetted, but also whenaterlogged soils become merely moist. For example, microalgae

hat accumulate during flooding die when floodwaters recede, andnder moist, well-aerated conditions their nutrients may then beineralized. Furthermore, in raised-field systems drying and re-etting cycles are produced not only by natural seasonal change,

ut also by human-mediated transfer of organic matter fromooded to well-drained compartments of the system. Biomass ofquatic microbes could lead to nutrient flushes as they are decom-osed in raised-field soils. Similarly, nutrient release to plants coulde enhanced by the accumulation of plant-derived organic mat-er under anaerobic, waterlogged conditions, followed by its rapidecomposition and mineralization once it is added to the well-rained raised fields.

These remarks are highly speculative, but what we know aboutaddy-rice systems (Roger, 1996) suggests that such speculation isot unfounded, and that studies of microbial ecology in raised-fieldgriculture are vital to understanding how these wetland agroe-osystems functioned.

.2.4. Functioning of raised-field systems at the landscape level:he role of fallow periods

Another large gap in our knowledge of how pre-Hispanicaised-field agriculture functioned is whether cultivation was con-inuous or whether fallows were incorporated, and if so, howeld/fallow cycles were organized in space and time. These areritical points in assessing the productivity of raised-field agri-ulture both in terms of land demand and of required labornput, and the level of use that is sustainable. The archaeolog-cal record is virtually silent on these points. Carbonized wildlant seeds present in ancient raised fields of Tiwanaku, inter-reted by Erickson (1994) as possibly the result of burning of

lama dung, could also reflect the burning of fallow vegetation dur-ng the preparation of a new cultivation cycle (Gondard, 2006).xperiments with rehabilitated raised fields give some insight.ased on results of these experiments, Erickson (2003) consid-rs that under good management, raised fields at Tiwanaku mayot have required long fallow cycles. In some of these rehabilita-ion experiments, however, yields have been reported to decreasefter 3–4 years, leading to the claim that fallows would have beenequired (Bandy, 2005). Bandy (2005) has suggested that fallowsould have been necessary at Tiwanaku to limit damage done byotato root nematodes. However, several objections can be raisedo this hypothesis. First—and this is a general problem with thexperiments that have been done—virtually the only thing that iseasured in these experiments is yield. We have no clue as tohat variables (e.g., pressure from pathogens or weeds, declin-

ng nutrient availability) might explain variation in yield. Second,here is no evidence for the importance of these pathogens in theseites. Third, pre-Hispanic farmers may have managed field envi-onments, crop populations, or both, in ways that obviated anyeed for fallow. Contemporary farming systems that are more or

ess analogous certainly include examples of successful permanent

ultivation over very long periods. The best-known, and probablyost durable, example is the chinampas. They have been cultivated

irtually continuously for centuries. While they are essentially lacustrine system and thus differ in many ways from extinct

3 l Engi

stCaetpfMta2

otptuftetcbFlOaphcw

creispalpwdpesfi

cfionfit2cesttarhr

2

hfofimncTmwtsodottcltaId

fihficcelp

ateaofset

2o

hahssmawm1ao

8 D. Renard et al. / Ecologica

ystems, most of which were in seasonally flooded environments,hey give interesting insights into the questions we examine here.hinampa soils contain several fungal species that limit prolifer-tion of crop pathogens, including both nematodes (Zuckermant al., 1994) and fungi (Lumsden et al., 1987). High organic mat-er content and greater biological activity are thought to favor theathogen-suppressive activity of chinampa soil. However, fallowsrequently do reduce the incidence of soil-borne crop pathogens.

ulching of raised fields with green manure from slashed vegeta-ion from the fallow could contribute to this effect, because plantslso contain diverse pathogen-suppressive metabolites (Chitwood,002).

Rehabilitation experiments have simply not been carried outver long enough periods to determine whether permanent cul-ivation was possible, and they have not measured agroecologicalarameters that could help us interpret their results. It seems likelyhat not all raised-field systems could have been cultivated contin-ously. First, some were characterized by soil of markedly lowerertility than in the Basin of Mexico, where chinampas were con-inuously cultivated. Second, most were located in highly seasonalnvironments, and were “fallow” each year for at least much ofhe dry season (though how long is unclear). Third, raised-fieldomplexes were likely not all built at the same time; a cycle ofuilding, use and reconstruction may have existed (Darch, 1988).allows may have been an integral part of such cycles. Fourth, fal-ows are included in some extant raised-field systems in Asia andceania. On Uvea, western Polynesia, “garden-island” raised fieldsre left fallow for unspecified periods after several successive crop-ing periods (Kirch, 1978). After fallow, they are rebuilt and the cuterbaceous fallow vegetation is incorporated into the soil. Signifi-antly, only a single area of the island with exceptionally good soilas cultivated continuously.

As in other agroecosystems, multi-year fallow periods betweenultivation cycles in raised-field farming systems could have playedoles in restoring soil fertility and in escaping pests. In tropical for-st environments, restoration of soil fertility during fallow periodss usually associated with substantial buildup of plant biomass, theource (after slashing and burning) of most of the nutrients thatermit growth of crops. In seasonally flooded savannas, biomassccumulates at much lower rates. However, traits peculiar to wet-and environments may have influenced the potential of falloweriods to restore soil fertility. During fallow periods, for example,ild plants adapted to waterlogging probably grew more abun-antly than when fields were actually in use. These plants arerobably better than the upland crops grown in raised fields inxtracting P from waterlogged soil and capturing it in biomass ortoring it in soil organic matter, which could be added to raisedelds in a new cultivation cycle.

Among the variables affecting the quantity of nutrients thatould be stored during fallow periods, frequency and intensity ofres may be among the most important. Fires mineralize N, P andther nutrients. Although some loss occurs through erosion, savan-ah soils appear quite efficient at retaining nutrients released byre (Kellman et al., 1985). However, fires also result in substan-ial loss to the atmosphere of N (through volatilization [Wan et al.,001]) and of P (as aerosols [Mahowald et al., 2005]). Frequent firesould also reduce vegetation cover and root density, lowering thefficiency of the plant community in capturing nutrients, and intoring them underground where they are less susceptible to losshrough fire (Kaufmann et al., 1994). Frequent, intense fires couldhus severely impact long-term primary productivity and limit the

ccumulation of soil organic matter. Savannas are fire-prone envi-onments, and limiting the frequency and intensity of fires mayave been an important aspect of the management of fallows byaised-field farmers.

tire

neering 45 (2012) 30– 44

.2.5. Self-organizing processes in fallow raised-field complexesAs McKey et al. (2010) outlined, self-organizing processes could

ave greatly enhanced the capacity of fallow periods to restore soilertility in maintaining human-initiated concentration of resourcesn the raised fields during fallows. Our data from “fossil” raised-eld landscapes of French Guianan coastal savannas show thatany organisms that have important “engineering” activities,

otably social insects, earthworms and woody plants, are stronglyoncentrated on the well-drained soils of abandoned raised fields.heir actions—transport of organic and mineral matter to mounds,odification of the properties of soils of abandoned raised fields inays that reduce their erodibility—can contribute to compensate

he non-negligible erosion we have demonstrated in this land-cape (McKey et al., 2010). Thus the concentration of resourcesn raised fields initiated by humans continues after field aban-onment (e.g., in fallows), driven by engineering activities of otherrganisms. Such engineering activities appear to have contributedo maintaining pre-Columbian raised fields for centuries followingheir abandonment. Such “outsourcing” of the task of resource-oncentration to natural engineers could also have reduced theabor costs of restoring raised fields in each cultivation cycle, con-ributing to the resilience of the system. The limited informationvailable suggests similar dynamics in other raised-field systems.n the Llanos de Mojos, termite nests are concentrated on aban-oned raised fields (see Plate 14b in Denevan, 1966).

The dynamics described by McKey et al. (2010) concern raised-eld landscapes that were abandoned centuries ago. It is not knownow long it took for such dynamics to be established after raisedelds were abandoned. How long did it take for organisms toolonize abandoned raised fields, and in what sequence did theyolonize? How much did abandoned raised fields erode beforengineer-generated feedbacks kicked in? Alternatively, were ateast some organisms with marked engineering activities alreadyresent in active fields?

According to Byers et al. (2006), “manipulating biotic inter-ctions to provide desired services and thus reduce or eliminatehe need for external inputs is fundamental to the practice ofcologically sound agriculture.” Positive feedbacks between thections of the farmers who constructed mounds and those of therganisms that maintain them illustrate an often strived for, butrequently elusive, objective of ecological engineering: exploitingynergies between actions of human engineers and those of naturalcosystem engineers, so that ecosystems self-organize to fit withechnology (Odum and Odum, 2003).

.2.6. Functioning of raised-field agriculture: socioeconomicrganization

Knowing why past cultures adopted raised-field agriculture,ow these farming systems were organized socially, technicallynd economically, and why they were eventually abandoned, couldelp us assess whether it is feasible and desirable to re-introduceome of their elements today. Here again there is great debate andubstantial uncertainty. The debate focuses mostly on the systemost intensively studied by archaeologists, the enormous area of

bandoned raised fields around Lake Titicaca, in Peru and Bolivia,hich were cultivated beginning about 3000 years ago and formeduch of the subsistence basis for the Tiwanaku empire (Kolata,

996). Why did raised-field agriculture in this area originate? Somerchaeologists view raised-field agriculture as a textbook examplef the process of induced intensification treated in the influen-

ial work of Ester Boserup (1965). According to her theory, peoplenitially practiced a more extensive type of agriculture requiringelatively little labor. With increasing population pressure, thesextensive methods no longer met food demands. People thus had

l Engi

nf

elhtasdcbs

piamrpctwcHmcsws

acbplvcb

3

ra

3

tetpqatSutatsbt

2mdAtrdco“Coaft

wrgwclladefl

ecttptt((diefifbri5soss

3p

tSoaSt

D. Renard et al. / Ecologica

o other choice than to adopt more time-demanding intensivearming systems (Bandy, 2005).

According to another view, forced labor imposed by a governinglite to produce surpluses may have been as important as “popu-ation pressure” in driving agricultural intensification. Top-downierarchical control by a central authority may have been necessaryo organize the construction and maintenance of raised fields andssociated earthworks (Kolata and Ortloff, 1989). This view alsouggests hypotheses about why raised-field systems were aban-oned, considering their fall as the inevitable consequence of theollapse of the empires on which they depended. Collapse may haveeen triggered by environmental change, by forces internal to theociety, or by the two in combination.

Erickson (1992) offers a contrasting viewpoint on all theseoints. According to his view, people adopted raised-field farm-

ng driven by preference, not by necessity. For example, in theltiplano, people may have preferred to live in lakeshore environ-ents because they were richer in wild resources, and adopted

aised-field agriculture as the best way to farm in the kinds oflaces they preferred to live. In keeping with this idea, no strongentral authority would have been necessary to develop and main-ain earthworks. As pointed out above (see Morris, 2004), theseere not irrigation systems that required control of water allo-

ation. Each set of raised fields functioned largely independently.ypotheses about strong top-down control in “hydraulic societies”ay thus have little pertinence here (Stanish, 1994). “Bottom-up”

ooperation among family groups or neighbors would have beenufficient to construct and maintain earthworks. Again in keepingith these ideas, raised fields continued to be farmed, at a smaller

cale, after the collapse of the Tiwanaku empire (Gray, 1992).Answers to all these questions can help us assess the desirability

nd feasibility of rehabilitating raised-field agriculture in the 21stentury. Systems requiring strong top-down control are likely toe both unfeasible and undesirable. Systems in which increasedroductivity per unit land comes at the cost of greatly increased

abor per unit of yield, without a corresponding increase in thealuation of labor, is also not a desirable future. Raised-field farmingan be part of a sustainable future only if people choose to adopt itecause it creates viable rural livelihoods.

. Raised-field agriculture: is it pertinent to today’s world?

Is it desirable, and feasible, to attempt to rehabilitate ancientaised fields, or to promote the development of new raised-fieldgriculture in similar environments?

.1. Why promote further agricultural conversion of wetlands?

Wetlands are important in biodiversity conservation and con-ribute significantly to global ecosystem services. Wetlands areverywhere under threat, and the principal threat is conversiono agricultural use (Daniels and Cumming, 2008). Why make thisroblem worse than it already is? Two points are relevant to thisuestion. First, many wetlands are already degraded. Raised-fieldgriculture could be integrated into projects aimed at restoringheir value for conservation and for providing ecosystem services.econd, maintaining “pristine” wetlands that exclude humans isnlikely to be an option, except for a relatively few integrally pro-ected areas. Even globally important wetlands such as the Pantanalre under enormous pressure (Junk and Nunes de Cunha, 2005). In

he view of one writer, “if wetlands are to be maintained in someemblance of ecological and hydrological health, they will have toe utilized for both traditional and modern sustainable uses (eveno the extent of some degradation) or they will be lost” (Smardon,

lwet

neering 45 (2012) 30– 44 39

006). Aside from putting wetlands under rice monoculture,ost attempts to develop agriculture in wetlands have envisaged

raining them for cattle ranching or for planting upland crops.griculture has usually worked against wetlands rather than with

hem (Rijsberman and de Silva, 2006). By working with wetlands,aised-field agriculture could contribute to reconciling food pro-uction and environmental quality (Robertson and Swinton, 2005),reating “working wetlands” that meet multiple needs (Dialoguen Water, Food and Environment; www.iwmi.org/dialogue) andfloodplain-friendly” development alternatives (Junk and Nunes deunha, 2005). Raised-field agriculture, modern agroecology basedn it, and cultural- and ecological-based tourism activities aimedt experiencing and interpreting culture/environment interactionsrom the perspectives of past peoples could be integrated into sus-ainable use schemes (Smardon, 2006).

Climate change stands to have dramatic effects on tropicaletlands (Mitsch et al., 2009). However, in environments where

aised-field agriculture remains possible, it could help reducereenhouse gas emission and thereby combat climate change. Byorking with wetlands, instead of against them, raised-field agri-

ulture could help maintain tropical wetlands and their soils, whichock away large quantities of carbon from the atmosphere for mil-ennia (Jungkunst and Fiedler, 2007). Seasonal tropical wetlandsre fire-prone environments (Mitsch et al., 2009). By limiting fireuring fallows, raised-field farmers could further reduce carbonmissions. Furthermore, fallow raised fields provide habitat forooding-intolerant components of wild biodiversity.

Raised-field agriculture could have further environmental ben-fits. In the Beni savannas of the Llanos de Mojos of Bolivia, workarried out by Oxfam and the Fundación Kenneth Lee has shownhat in addition to providing a form of alternative food securityo the poorest and most vulnerable segments of the region’s ruralopulations, raised-field agriculture could reduce the environmen-al and economic impact caused by the severe flooding regime ofhe Mamoré and Beni rivers associated with strong La Nina eventsCEPAL, 2008; Latrubesse et al., 2010). As in the Lake Titicaca basinErickson, 2003), raised fields help put crops beyond the reach ofevastating floods. However, flooding regimes are more variable

n these dynamic river basins than in lacustrine environments, andxtreme floods would still impact lower-lying fields. Finally, raised-eld agriculture offers an alternative to the clearing of tropical

orest for slash-and-burn agriculture. Preliminary results of reha-ilitation experiments indicate that raised-field systems reduceisks associated with floods and are extremely productive, produc-ng two to three maize crops per year with annual yields of up to

t ha−1 (Saavedra, 2009). However, we are just beginning to under-tand how these agricultural systems function within the contextf the larger fluvial environment and these experiments are stillmall-scale and short-term, preventing proper assessment of theirustainability.

.2. What factors could limit the adoption of this way ofracticing agriculture?

Some critics of attempts to rehabilitate raised-field agricul-ure consider them to have failed (Bandy, 2005; Chapin, 1988).cholars who expressed early enthusiasm are now quick to pointut that any lessons pre-Columbian raised-field farming has forgriculture today must be very carefully drawn (Denevan, 2001;iemens, 2004). However, the difficulties encountered in attemptso rehabilitate raised-field farming are simply examples of a prob-

em faced in any move to ecological agriculture today: “Today’s

orld is filled with urban workers while what we need are ruralcologists. . .the destruction of rural society has taken with ithe knowledge base and labor force that will be needed for the

4 l Engi

tb

irhafitpo(lm(efitfFolefilo

ithmueittsCameeneoepgtilp

3a

wiiaiae

afe

slusfvthstnwmoyttot(2rShWdpihd

cicirtw1pamocturcp1

jilpdi

0 D. Renard et al. / Ecologica

ransformation” (Vandermeer, 2011). If ecological agriculture is toe successful anywhere, these paradoxes have to be resolved.

The most comprehensive analyses of why most farmersnvolved in rehabilitation attempts have not permanently adoptedaised-field farming as a production strategy are offered by theardiest champions of these experiments. Both Erickson (2003)nd Gliessman (1991) cogently discuss the sustainability of raised-eld agriculture, pointing out that the problem is not simply one ofechnology, soil fertility, or labor requirements, but includes com-lexities that other critics have not acknowledged. High labor costf construction is often cited as a principal drawback of raised fieldse.g., Bandy, 2005). However, although the initial construction ofarge blocks of raised fields requires considerable labor, total labor

ay be relatively low when spread out over years of cultivationErickson, 2003; Mathewson, 1987; see Section 3.3), particularly ifcosystem engineers contribute to raised-field maintenance duringallows (McKey et al., 2010). “The most important factor explain-ng non-adoption by farmers”, Erickson (2003) concludes, “is thathe social, political, and economic environment today is differentrom that when the raised fields were first constructed and used.”or example, competing labor demands exist today that reduce thepportunity for initiating raised-field construction. Furthermore,ivestock is now a significant source of income for Quechua farm-rs, who must decide between using land for rehabilitating raisedelds or for grazing livestock. Political instability also led to irregu-

ar funding for rehabilitation experiments, leading to abandonmentf many fields.

Chapin (1988) and Gliessman (1991) present strongly contrast-ng analyses of rehabilitation experiments in Mexico. Accordingo the former author, the seductive chinampas model from theighlands was inappropriately transposed into lowland environ-ents in Veracruz and Tabasco: “In both cases, the stated and

nstated objectives of project managers had little fit with the inter-sts and needs of the farmers. The two projects were designed andmplemented by outside technicians without significant local par-icipation, and both rapidly fell apart when ‘beneficiaries’ failedo cooperate” (Chapin, 1988). Gliessman (1991), who did exten-ive field work in the area on a third project not mentioned byhapin (1988), paints a very different picture. Projects in whichgroecologists cooperated with local people had begun to accu-ulate the data needed to adapt the highland model to lowland

nvironments (e.g., Boucher et al., 1983; Gliessman, 1992). Thexperiments showed that achieving high yields would take time:ewly built platforms took 2 years to accumulate adequate lev-ls of organic matter and available nutrients and a further 3 yearsr more to develop good agricultural soil. Unfortunately, thesexperiments were terminated too soon. Inspired by these fledglingrojects, government officials decided to initiate a large-scale pro-ram. According to Gliessman (1991), this last project (which washe principal target of Chapin’s [1988] critique) was top-downn its approach and generally poorly designed. Before failing, ited to reduced funding and interruption of the promising smallerrojects, bringing them down as well.

.3. Labor requirements and the potential of raised-fieldgriculture to support viable rural livelihoods

The pertinent question about raised-field agriculture is nothether it can work in today’s world, but whether it can work

n tomorrow’s world. How agriculture is conducted must be rad-cally rethought. There are currently few signs that governments

re ready to encourage food production strategies that help mit-gate and adapt to climate change, conserve biodiversity andchieve social justice. It is thus more important than ever thatcologists continue to argue for the necessity of these goals,

Att

neering 45 (2012) 30– 44

nd test ideas about how to accomplish them. Does raised-fieldarming have a place in the new kinds of agriculture that mustmerge?

The labor-intensive nature of raised-field farming has been con-idered a major drawback to its adoption. However, the amount ofabor that was actually needed to construct raised-field systems isnknown. In archaeology, estimating time and labor costs of con-tructing raised fields and other earthworks is a difficult enterprise,raught with uncertainty concerning many variables, including theolume of earth moved, the amount of work required to do it withhe tools available, the number of people available to do the work,ow long their workdays were, and how much maintenance theystem required once constructed. Researchers vary in the atten-ion they give to each of these variables (estimates for each areot always made explicit), and estimates are expressed in differentays, rendering comparison difficult. Arco and Abrams (2006) esti-ate that construction of the entire chinampa system in the Basin

f Mexico (6500 ha) took 25 million worker-days, spread over 40ears. This amounts to 1712 worker-days per year, but the size ofhe labor force was not estimated. Mathewson (1987) estimatedhat the labor necessary to produce and maintain the 50,000 haf raised-field complexes in the Guayas Basin, Ecuador, amountedo only 12 days per year per family from the start of the system1600 B.C.) to its end (1500 A.D.). Turner and Harrison (1983, pp.59–260) estimate that between 710 and 3266 worker-years wereequired for construction of the 310 ha of raised fields in Pulltrouserwamp. Assuming a labor force of 100 workers, construction wouldave taken from 7.1 to 32.7 years. Using a wide range of estimates,alker (2004, pp. 43–47) argues that the raised fields in the Llanos

e Mojos could have been constructed by a small number of peo-le. For example, he estimates that the largest of the raised fields

n northwestern Mojos, which cover about 1.8 ha each, could eachave been built by a group of 100 people in a single episode of 20ays.

It is interesting to compare these estimates with constructionosts that have been measured for a present-day Asian system thats a fairly close analogue. The sorjan cropping system of Indonesiaombines growing upland crops on raised beds and lowland cropsn the sink (IIRR, 1990). In a trial experiment, constructing theaised beds and sinks (50 cm from the bottom of the sink to theop of the bed; beds 3.5 m wide, sinks 3 m wide) required 4479orker-hours of labor per ha of sorjan (Domingo and Hagerman,

982). Assuming a 6-h workday, this amounts to 746 worker-dayser ha. Again, while comparison is difficult, these measured valuesppear substantially lower than most of the archaeologists’ esti-ates. Although this initial labor cost may still be considered high,

nce constructed sorjan beds are fairly permanent structures. Theironstruction should thus be considered as an investment ratherhan a simple labor cost. Labor requirement could be reduced by these of earthmoving machines to construct raised fields, as in someehabilitation experiments (Denevan, 2001; Saavedra, 2009), butare must be taken to avoid destroying by compaction the physicalroperties that help make soils favorable for agriculture (Chapin,988).

If we succeed in devising mechanisms by which farmers areustly compensated for labor that provides ecosystem servicesn addition to food production, then labor-intensiveness may noonger be an obstacle. It could in fact confer important advantages,roviding jobs for rural populations and thereby reversing the exo-us to cities driven by the lack of gainful rural employment in

ndustrial agriculture.

The chinampas of Mexico City provide an instructive example.

fter many centuries, the system is still part of the region’s agricul-ural picture, although it is much altered, even degraded, and underhreat (Torres-Lima et al., 1994). Adapting to modern pressures,

l Engi

ctcrsse1

4

rfifidsTtacslclttpa

ntrtwiWbapil(isDpataopbt

tletocLss

tho

A

CzEStttwC

R

A

A

A

A

A

A

A

B

B

B

B

B

B

B

B

B

B

C

C

D. Renard et al. / Ecologica

hinamperos grow as their most important crop cut flowers, ratherhan staple foods. In addition to research on the ecological prin-iples applied in management—a classical theme in chinampasesearch (Armillas, 1971; Coe, 1964)—what is needed now is under-tanding of the relationships between agroecological principles andocioeconomic strategies addressing urban development, regionalmployment, and environmental concerns (Torres-Lima et al.,994).

. Conclusion

Raised-field agriculture in the Neotropical region could yet beesurrected. Data from archaeological and contemporary raised-eld systems suggest how their functioning at the local plot scalets food production to ecosystem processes, but experiments con-ucted so far to determine the feasibility of re-establishing suchystems have encountered a number of methodological problems.heir focus has been on measuring yield, rather than studyinghe ecological interactions that underpin yield. Furthermore, thegroecological diversity of raised-field systems has not been appre-iated. Raised-field systems function differently under differentoils and climates. Except for those in the richest environments, fal-ow periods are probably necessary, but experiments have not beenonducted over long enough times to assess the importance of fal-ows and the roles they play. Experiments have also been too shorto explore how raised-field agriculture can be adapted not onlyo local biophysical environments but also to social, economic, andolitical contexts—or how the social capital necessary for ecologicalgriculture can be reconstructed.

We believe that two distinct kinds of experimental studies areeeded. First, there is still a great need for research aiming simplyo fill the large gaps in our knowledge about how pre-Columbianaised-field farming systems may have worked, without pressureo demonstrate their usefulness in some real-life situation todayithin the time frame of a typical 3- to-5-year project cycle. There

s great diversity among systems, and great depth within each.e have barely scratched the surface, and this knowledge will

e necessary for experiments of the second type, those aimedt re-establishing raised-field agriculture as a sustainable food-roduction system today and in environments in the near future

n which resources such as land, fertilizers and water will be moreimited. In research aimed at clarifying how raised fields workedand work), we must look further afield than we have so far fornspiration from present-day systems. Dynamic sorjan croppingystems in Asia may have as much to tell us today as the chinampas.etailed studies of energy and nutrient flow in highly integratedresent-day field/pond agroecosystems in southern China (Guond Bradshaw, 1993; Luo and Lin, 1991) may yield more insightshan the coarse-grained inferences possible from archaeologicalnd archaeobotanical studies. We must understand the agroecol-gy of raised-field farming systems in enough detail that, instead ofroposing a preconceived, prepackaged model to farmers, we wille able to propose à la carte those pieces that might be suited toheir needs in any one of a diversity of situations today.

For experiments aimed at re-establishing raised-field agricul-ure as a sustainable food-production system, the single overridingesson from previous work is that the people affected by thesexperiments must be involved in the fundamental decisions fromhe start. The solutions that emerge must fit their needs, not thether way around. The labor-intensive nature of raised-field agri-

ulture may make it unsuitable in many present-day contexts inatin America, but if conditions are created that favor ecologicallyustainable intensification of agriculture by reconstructing theocial capital on which such agriculture rests and by compensating

D

neering 45 (2012) 30– 44 41

he labor of farmers who conduct it, raised-field agriculture couldave broad application. Rigorous studies of raised-field agroecol-gy are required if we are to be ready.

cknowledgements

This study was funded by two interdisciplinary programs of theNRS (INEE, Institut National d’Ecologie et Environnement), “Ama-onie” and “Ingénièrie Ecologique”. Personnel of the laboratorycoFog (UMR L3MA, CNRS), Gaëlle Fornet of the CNRS Guyane andandrine Richard of the Centre Spatial Guyanais, all in Kourou, arehanked for their logistical assistance. We thank the organizers ofhe EECA symposium and the editors of Ecological Engineering forhe invitation to contribute to this special issue. The manuscriptas greatly improved by the perceptive comments of Charleslement (INPA, Manaus, Brazil) and of an anonymous reviewer.

eferences

llred, B.J., Brown, L.C., Fausey, N.R., Cooper, R.L., Clevenger, W.B., Prill, G.L., La Barge,G.A., Thornton, C., Riethman, D.T., Chester, P.W., Czartoski, B.J., 2003. Water tablemanagement to enhance crop yields in a wetland reservoir subirrigation system.Appl. Eng. Agric. 19, 407–421.

ltieri, M.A., 2008. Small Farms as a Planetary Ecological Asset: Five Key ReasonsWhy we Should Support the Revitalisation of Small Farms in the Global South.Third World Network, Penang, Malaysia.

rce, J., 1993. Evaluación y comparación de rendimientos de cuatro cultivos en tresanchuras de camellones (campos elevados) en la Estación Biológica del Beni(Prov. Ballivián, Dpto. Beni). PhD thesis, Department of Agronomy, UniversidadTécnica del Beni, Trinidad.

rco, L.J., Abrams, E.M., 2006. An essay on energetics: the construction of the Aztecchinampa system. Antiquity 80, 906–918.

rmillas, P., 1971. Gardens on swamps. Archaeological research verifies historicaldata on Aztec land reclamation in the valley of Mexico. Science 174, 653–661.

tkinson, C.J., Fitzgerald, J.D., Hipps, N.A., 2010. Potential mechanisms for achievingagricultural benefits from biochar application to temperate soils: a review. PlantSoil, doi:10.1007/s11104-010-0464-5.

yars, J.E., Christen, E.W., Soppe, R.W., Meyer, W.S., 2006. The resource potential ofin-situ shallow ground water use in irrigated agriculture: a review. Irrig. Sci. 24,147–160.

andy, M.S., 2005. Energetic efficiency and political expediency in Titicaca Basinraised field agriculture. J. Anthropol. Archaeol. 24, 271–296.

arba, J., Canal, E., García, E., Jordà, E., Miró, M., Pastó, E., Playà, R., Romero, I., Via,M., Woynarovich, E., 2003. Moxos: Una Limnocultura. Cultura y Medio Naturalen la Amazonia Boliviana. CEAM, Barcelona.

each, T., Luzzadder-Beach, S., Dunning, N., Jones, J., Lohse, J., Guderjan, T., Bozarth, S.,Millspaugh, S., Bhattacharya, T., 2009. A review of human and natural changes inMaya Lowland wetlands over the Holocene. Quaternary Sci. Rev. 28, 1710–1724.

iesboer, D.D., Binford, M., Kolata, A., 1999. Nitrogen fixation in soils and canals ofrehabilitated raised-fields of the Bolivian Altiplano. Biotropica 31, 255–267.

oixadera, J., Poch, R.M., Garcíıa-González, M.T., Vizcayno, C., 2003. Hydromorphicand clay-related processes in soils from the Llanos de Moxos (northern Bolivia).Catena 54, 403–424.

oserup, E., 1965. The Conditions of Agricultural Growth. Earthscan Publications,London (paperback edition, 1993).

oucher, D.H., Espinosa, M.J., Romero, B.S., Gliessman, S.R., 1983. Out-of-seasonplanting of grain legumes as green manure for a tropical raised-field agroe-cosystem. Biol. Agric. Hort. 1, 127–133.

ranch, N.P., Kemp, R.A., Silva, B., Meddens, F.M., Williams, A., Kendall, A., VivancoPomacanchari, C., 2007. Testing the sustainability and sensitivity to climaticchange of terrace agricultural systems in the Peruvian Andes: a pilot study. J.Archaeol. Sci. 34, 1–9.

ray, W., Herrera, L., Schrimpff, M.C., Botero, P., Monsalve, J.G., 1987. The ancientagricultural landscape of Calima. In: Denevan, W., Mathewson, D., Knapp, G.(Eds.), Pre-Hispanic Agricultural Fields in the Andean Region, Part I. BritishArchaeological Reports. International Series Oxford, UK, pp. 443–457.

yers, J.E., Cuddington, K., Jones, C.G., Talley, T.S., Hastings, A., Lambrinos, J.G., Crooks,J.A., Wilson, W.G., 2006. Using ecosystem engineers to restore ecological sys-tems. Trends Ecol. Evol. 21, 493–500.

arney, H.J., Binford, M.W., Kolata, A.L., Marín, R.R., Goldman, C.R., 1993. Nutrientand sediment retention in Andean raised-field agriculture. Nature 364, 131–133.

arney, H.J., Binford, M.W., Kolata, A.L., 1996. Nutrient fluxes and retention inAndean raised-field agriculture. Implications for long-term sustainability. In:

Kolata, A. (Ed.), Tiwanaku and its Hinterlands: Archaeology and Paleoecology inAndean Civilization, vol 1. Smithsonian Institution Press, Washington D.C., pp.169–179.

e Las Casas, B., 1986 [1560]. Historia de las Indias. Biblioteca Ayacucho, Caracas,Venezuela.

4 l Engi

C

C

C

C

C

C

CC

C

C

D

D

D

D

D

D

D

D

D

D

D

D

D

D

D

D

D

D

D

D

E

E

E

E

E

E

E

E

E

E

G

G

G

G

G

G

G

G

G

G

G

G

G

G

G

H

H

I

I

I

J

J

2 D. Renard et al. / Ecologica

EPAL, 2008. Evaluacion del Impacto Acumulado y Adicional Ocasionado por LaNina, Bolivia 2008. Naciones Unidas. Comision Economica para America Latinay el Caribe.

hapin, M., 1988. The seduction of models: Chinampa agriculture in Mexico. Grass-roots Dev. 12, 8–17.

hepkwony, C.K., Haynes, R.J., Swift, R.S., Harrison, R., 2001. Mineralization of soilorganic P induced by drying and rewetting as a source of plant-available P inlimed and unlimed samples of an acid soil. Plant Soil 234, 83–90.

hitwood, D.J., 2002. Phytochemical based strategies for nematode control. Annu.Rev. Phytopathol. 40, 221–249.

lement, C.R., Junqueira, A., 2010. Between a pristine myth and an impoverishedfuture. Biotropica, doi:10.1111/j.1744-7429.2010.00674.x.

lement, C.R., McCann, J.M., Smith, N.J.H., 2004. Agrobiodiversity in Amazônia andits relationship with dark earths. In: Lehmann, J., Kern, D., Glaser, B., Woods,W. (Eds.), Amazonian Dark Earths: Origin, Properties and Management. KluwerAcademic Publisher, Dordrecht, pp. 159–178.

oe, M.D., 1964. The chinampas of Mexico. Sci. Am. 211, 90–98.rews, T.E., Gliessman, S.R., 1991. Raised field agriculture in Tlaxcala, Mexico: an

ecosystem perspective on maintenance of soil fertility. Am. J. Altern. Agric. 6,9–16.

rosby, A.W., 1972. The Columbian Exchange: Biological and Cultural Consequencesof 1492. Greenwood Press, Westport, CT, USA.

rossley, P.L., 2004. Sub-irrigation in wetland agriculture. Agric. Hum. Values 21,191–205.

algaard, T., Hutchings, N.J., Porter, J.R., 2003. Agroecology, scaling and interdisci-plinarity. Agric. Ecosyst. Environ. 100, 39–51.

aniels, A.E., Cumming, G.S., 2008. Conversion or conservation? Understanding wet-land change in Northwest Costa Rica. Ecol. Appl. 18 (1), 49–63.

arboux, F., Gascuel-Odoux, C., Davy, P., 2002. Effects of surface water storage bysoil roughness on overland-flow generation. Earth Surf. Proc. Land. 27, 223–233.

arch, J.P., 1988. Drained field agriculture in tropical Latin America: parallels frompast to present. J. Biogeogr. 15, 87–95.

enevan, W.M., 1966. The Aboriginal Cultural Geography of the Llanos de Mojos ofBolivia. University of California Press, Berkeley and Los Angeles.

enevan, W.M., 1970. Aboriginal drained-field cultivation in the Americas: pre-Columbian reclamation of wet lands was widespread in the savannas andhighlands of Latin America. Science 169, 647–654.

enevan, W.M., 1995. Prehistorical agricultural methods as models for sustainabil-ity. Adv. Plant Pathol. 11, 21–43.

enevan, W.M., 2001. Cultivated Landscapes of Native Amazonia and the Andes.Oxford University Press, New York.

enevan, W.M., 2006. Una perspectiva histórica sobre el descubrimiento de camposelevados (camellones) prehispánicos en Sud América. In: Valdez, F. (Ed.), Agri-cultura Ancestral. Camellones y Albarradas. Contexto Social, Usos y Retos delPasado y del Presente. Ediciones Abya-Yala, Quito, Ecuador, pp. 17–24.

enevan, W., Mathewson, D., Knapp, G. (Eds.), 1987. Pre-Hispanic Agricultural Fieldsin the Andean Region, Part I. British Archaeological Reports. International SeriesOxford, UK.

enevan, W.M., Schwerin, K.H., 1978. Adaptive strategies in Karinya subsistence,Venezuelan Llanos. Fundación La Salle, Caracas. Antropológica 50, 3–91.

enevan, W.M., Turner, B.L., 1974. Forms, functions and associations of raised fieldsin the old world tropics. J. Trop. Geogr. 39, 24–33.

enevan, W., Zucchi, A., 1978. Ridged-field excavations in the Central OrinocoLlanos, Venezuela. In: Browman, D.L. (Ed.), Advances in Andean Archaeology.Mouton Publishers, Oxford, UK, pp. 235–245.

illehay, T.D., Pino, M., Bonzani, R., Silva, C., Wallner, J., Le Quesne, C., 2007. Cul-tivated wetlands and emerging complexity in south-central Chile and longdistance effects of climate change. Antiquity 81, 949–960.

obyns, H.F., 1966. Estimating aboriginal American population: an appraisal of tech-niques with a new hemispheric estimate. Curr. Anthropol. 7, 395–416.

öll, P., Kaspar, F., Lehner, B., 2003. A global hydrological model for deriving wateravailability indicators: model tuning and validation. J. Hydrol. 270, 105–134.

omingo, A.A., Hagerman, H.H., 1982. Sorjan cropping system trial in irrigated wetland conditions. Philippine J. Crop Sci. 7, 154–161.

oolittle, W.E., Sluyter, A., Perramond, E.P., Crossley, P.L., Lambert, D.P., 2002. Feed-ing a growing population on an increasingly fragile environment. In: Knapp, G.(Ed.), Latin America in the 21st Century: Challenges and Solutions. University ofTexas Press, Austin, pp. 45–72.

rinkwater, L.E., Snapp, S.S., 2007. Understanding and managing the rhizosphere inagroecosystems. In: Cardon, Z.G., Whitbeck, J.L. (Eds.), The Rhizosphere. An Eco-logical Perspective. Elsevier Academic Press, San Diego, CA, USA, pp. 127–154.

unning, N.P., Luzzadder-Beach, S., Beach, T., Jones, J.G., Scarborough, V., Culbert,T.P., 2002. Arising from the Bajos: the evolution of a Neotropical landscape andthe rise of Maya civilization. Ann. Assoc. Am. Geogr. 92, 267–283.

nriquez, S., Duarte, C.M., Sand-Jensen, K., 1993. Patterns in decomposition ratesamong photosynthetic organisms: the importance of detritus C:N:P content.Oecologia 94, 457–471.

rickson, C., 1992. Applied archaeology and rural development: archaeology’spotential contribution to development. J. Steward Anthropol. Soc. 20, 1–16.

rickson, C.L., 1994. Methodological considerations in the study of ancient Andeanfield systems. In: Miller, N.K., Gleason, K.L. (Eds.), The Archaeology of Gardenand Field. University of Pennsylvania Press, Philadelphia, pp. 111–152.

rickson, C., 1995. Archaeological methods for the study of ancient landscapes ofthe Llanos de Mojos in the Bolivian Amazon. In: Stahl, P. (Ed.), Archaeology in

J

J

neering 45 (2012) 30– 44

the Lowland American Tropics. Current Analytical Methods and Recent Appli-cations. Cambridge University Press, Cambridge, UK, pp. 66–95.

rickson, C.L., 2000. An artificial landscape-scale fishery in the Bolivian Amazon.Nature 408, 190–193.

rickson, C.L., 2003. Agricultural landscapes as world heritage: raised field agricul-ture in Bolivia and Peru. In: Teutonico, J.-M., Matero, F. (Eds.), Managing Change:Sustainable Approaches to the Conservation of the Built Environment. GettyConservation Institute, Los Angeles, pp. 181–204.

rickson, C., 2006. The domesticated landscapes of the Bolivian Amazon. In: Balée,W., Erickson, C. (Eds.), Time and Complexity in Historical Ecology: Studies in theNeotropical Lowlands. Columbia University Press, New York, pp. 235–278.

rickson, C.L., Candler, K.L., 1989. Raised fields and sustainable agriculture in theLake Titicaca Basin of Peru. In: Browder, J.O. (Ed.), Fragile Lands of Latin America:Strategies for Sustainable Development. Westview Press, Boulder, pp. 230–248.

rickson, C., Walker, J., 2009. Precolumbian causeways and canals as landesque cap-ital. In: Snead, J.E., Erickson, C.L., Darling, J.A. (Eds.), Landscape of Movement.Trails, Paths, and Roads in Anthropological Perspective. University of Pennsyl-vania Museum of Archaeology and Anthropology, Philadelphia, pp. 232–252.

venson, R.E., Gollin, D., 2003. Assessing the impact of the green revolution, 1960 to2000. Science 300, 758–762.

laser, B., 2007. Prehistorically modified soils of central Amazonia: a model forsustainable agriculture in the twenty-first century. Phil. Trans. R. Soc. B 362,187–196.

laser, B., Birk, J.J. State of the scientific knowledge on properties and genesis ofAnthropogenic Dark Earths in Central Amazonia (terra preta de Índio). Geochim.Cosmochim. Acta, doi:10.1016/j.gca.2010.11.029, in press.

laser, B., Haumaier, L., Guggenberger, G., Zech, W., 2001. The Terra Pretaphenomenon – a model for sustainable agriculture in the humid tropics. Natur-wissenschaften 88, 37–41.

laser, B., Lehmann, J., Zech, W., 2002. Ameliorating physical and chemical proper-ties of highly weathered soils in the tropics with charcoal: a review. Biol. Fertil.Soils 35, 219–230.

liessman, S.R., 1991. Ecological basis of traditional management of wetlands intropical Mexico: Learning from agroecosystem models. In: Oldfield, M.L., Alcorn,J.B. (Eds.), Biodiversity: Culture, Conservation and Ecodevelopment. WestviewPress, Boulder, CO, pp. 211–229.

liessman, S.R., 1992. Agroecology in the tropics: achieving a balance between landuse and preservation. Environ. Manage. 16, 681–689.

liessman, S.R., 2007. Agroecology. The Ecology of Sustainable Food Systems. CRCPress, Boca Raton, FL, USA.

liessman, S.R., Turner, B.L., Rosado May, F.J., Amador, M.F., 1985. Ancient raisedfield agriculture in the Maya lowlands of southeastern Mexico. In: Farrington, I.S.(Ed.), Prehistoric Intensive Agriculture in the Tropics. BAR International Series232, Oxford, UK, pp. 97–111.

ondard, P., 2006. Campos elevados en llanuras húmedas. Del modelado al paisaje.Camellones, waru warus o pijales. In: Valdez, F. (Ed.), Agricultura Ancestral.Camellones y Albarradas. Contexto Social, Usos y Retos del Pasado y del Presente.Ediciones Abya-Yala, Quito, Ecuador, pp. 25–53.

ray, G., 1992. Beyond state collapse: rural history, raised fields, and pastoralism inthe South Andes. Am. Anthropol. 94, 882–904.

renand, P., 1981. Agriculture sur brûlis et changements culturels: le cas des Indi-ens Wayãpi et Palikur de Guyane. J. d’agriculture traditionnelle et de botaniqueappliquée 28, 23–31.

riffon, M., 2002. Révolution verte, revolution doublement verte. Mondes enDéveloppement 30, 117–139.

umilla, J., 1963 [1791]. The Orinoco Illustrated and Defended. Biblioteca de laAcademia Nacional de la Historia, Caracas (translated from Spanish).

uo, J.Y., Bradshaw, A.D., 1993. The flow of nutrients and energy through a Chinesefarming system. J. Appl. Ecol. 30, 86–94.

uttmann-Bond, E., 2010. Sustainability out of the past: how archaeology can savethe planet. World Archaeol. 42, 355–366.

anagarth, W., 1993. Acerca de la Geoecologia de las Sabanas del Beni en el Noroestede Bolivia. Instituto de Ecologia, La Paz, Bolivia.

obbs, P.R., Sayre, K., Gupta, R., 2008. The role of conservation agriculture in sus-tainable agriculture. Phil. Trans. R. Soc. B 363, 543–555.

IRR, 1990. Low-external Input Rice Production (LIRP): Technology Information Kit.IIRR (International Institute of Rural Reconstruction), Cavite, Philippines.

riarte, J., Glaser, B., Watling, J., Wainwright, A., Birks, J., Renard, D., Rostain, S.,McKey, D., 2010. Late Holocene Neotropical agricultural landscapes: Phytolithand stable carbon isotope analysis of raised fields from French Guianan coastalsavannahs. J. Archaeol. Sci. 37, 2984–2994.

RRI, 1985. Wetland Soils: Characterization, Classification and Utilization. Interna-tional Rice Research Institute, Los Banos, Philippines.

anssen, M.A., Scheffer, M., 2004. Overexploitation of renewable resources byancient societies and the role of sunk-cost effects. Ecol. Soc. 9 (1), 6,http://www.ecologyandsociety.org/vol9/iss1/art6/main.html.

arvis, P., Rey, A., Petsikos, C., Wingate, L., Rayment, M., Pereira, J., Banza, J., David,J., Miglietta, F., Borghetti, M., Manca, G., Valentini, R., 2007. Drying and wettingof Mediterranean soils stimulates decomposition and carbon dioxide emission:the “Birch effect”. Tree Physiol. 27, 929–940.

ungkunst, H.F., Fiedler, S., 2007. Latitudinal differentiated water table control ofcarbon dioxide, methane and nitrous oxide fluxes from hydromorphic soils:feedbacks to climate change. Glob. Change Biol. 13, 2668–2683.

unk, W.J., Nunes de Cunha, C., 2005. Pantanal: a large South American wetland at acrossroads. Ecol. Eng. 24, 391–401.

l Engi

J

K

K

K

K

KK

K

K

K

K

K

K

K

L

L

L

L

L

L

L

L

L

L

L

M

M

M

M

M

M

MM

M

M

N

N

O

O

P

P

P

P

P

P

R

R

R

R

R

R

R

R

R

SS

S

S

S

S

S

D. Renard et al. / Ecologica

unqueira, A.B., Shepard Jr., G.H., Clement, C.R., 2010. Secondary forests on anthro-pogenic soils in Brazilian Amazonia conserve agrobiodiversity. Biodivers.Conserv. 19, 1933–1961.

aufmann, J.B., Cummings, D.L., Ward, D.E., 1994. Relationships of fire, biomass andnutrient dynamics along a vegetation gradient in the Brazilian cerrado. J. Ecol.82, 519–531.

ellman, M., Miyanishi, K., Hiebert, P., 1985. Nutrient retention by savanna ecosys-tems: II. Retention after fire. J. Ecol. 73, 953–962.

emp, R., Branch, N., Silva, B., Meddens, F., Williams, A., Kendall, A., Vivanco, C., 2006.Pedosedimentary, cultural and environmental significance of paleosols withinpre-Hispanic agricultural terraces in the southern Peruvian Andes. QuaternaryInt. 158, 13–22.

irch, P.V., 1978. Indigenous agriculture on Uvea (western Polynesia). Econ. Bot. 32,157–181.

irk, G., 2004. The Biogeochemistry of Submerged Soils. Wiley, Chichester, UK.leinhenz, V., 1997. Technologies for Sustainable Vegetable Production in the Trop-

ical Lowlands. Herbert Utz Verlag Wissenschaft, Munich.leinhenz, V., Midmore, D.J., Schnitzler, W.H., 1996a. High bed systems for off-

season vegetable production in the tropics and subtropics. Agric. Rural Dev. 3,60–62.

leinhenz, V., Schnitzler, W.H., Midmore, D.J., 1996b. Permanent high-bed technol-ogy to improve intensive vegetable production in tropical lowlands – effect onwater stress, root distribution and yield. Horticultural Sci. Kertészeti Tudomány28, 90–98.

napp, G.W., 1988. Ecología Cultural Prehispánica del Ecuador. Biblioteca deGeografia Ecuatoriana. Ediciones del Banco Central del Ecuador, Quito.

napp, G.W., Mothes, P.A., 1999. Quilotoa ash and human settlements in the Equa-torial Andes. In: Mothes, P. (Dir.), Actividad volcánica y Pueblos Precolombinosen el Ecuador. Ediciones Abla-Yala, Quito, pp. 139–155.

olata, A., 1993. The Tiwanaku. Portrait of an Andean Civilization. Blackwell, Cam-bridge.

olata, A., 1996. Tiwanaku and Its Hinterland: Archaeology and Paleoecology ofan Andean Civilization. Volume 1: Agroecology. Smithsonian Institution Press,Washington, DC.

olata, A.L., Ortloff, C., 1989. Thermal analysis of Tiwanaku raised field systems inthe lake Titicaca basin of Bolivia. J. Archaeol. Sci. 16, 233–263.

ambert, J.D.H., Siemens, A.H., Arnason, J.T., 1984. Ancient Maya drained field agri-culture: its possible application today in the New River floodplain, Belize. C.A.Agric. Ecosyst. Environ. 11, 67–84.

atrubesse, E.M., Baker, P.A., Argollo, J., 2010. Geomorphology of natural hazardsand human-induced disasters in Bolivia. Dev. Earth Surf. Proc. 13, 181–194.

ee, K., 1997. Apuntes sobre las obras hidrauclicas prehispánicas de las llanuras deMoxos. Paititi 11 (1), 24–26, Casa de la Cultura de Beni y Pando, Trinidad, Bolivia.

ehmann, J., 2009. Terra preta nova – where to from here? In: Woods, W.I., Teix-eira, W.G., Lehmann, J., Steiner, C., WinklerPrins, A.M.G.A., Rebellato, L. (Eds.),Amazonian Dark Earths: Wim Sombroek’s Vision. Springer, Berlin, pp. 473–486.

ehmann, J., Gaunt, J., Rondon, M., 2006. Bio-char sequestration in terrestrialecosystems—a review. Mitig. Adapt. Strat. Global Change 11, 395–419.

ombardo, U., Canal-Beeby, E., Fehr, S., Veit, H., 2011. Raised fields in the BolivianAmazonia: a prehistoric green revolution or a flood risk mitigation strategy? J.Archaeol. Sci. 38, 502–512.

ovell, W.G., 1992. “Heavy shadows and black night”: disease and depopulation incolonial Spanish America. Ann. Assoc. Am. Geogr. 82, 426–443.

ovell, W.G., Lutz, C.H., 1995. Demography and Empire: A Guide to the Populationhistory of Spanish Central America, 1500–1821. Westview Press, Boulder, CO,USA.

umsden, R.D., Garcia, E.R., Lewis, J.A., Frías-T, G.A., 1987. Suppression of damping-off caused by Pythium spp. in soil from the indigenous Mexican chinampaagricultural system. Soil Biol. Biochem. 19, 501–508.

uo, S.M., Lin, R.J., 1991. High bed-low ditch system in the Pearl River Delta, SouthChina. Agric. Ecosyst. Environ. 36, 101–109.

uzzader-Beach, S., Beach, T., 2006. Wetlands as the intersection of soils, water,and indigenous human society in the Americas. In: McNeill, J.R., Winiwarter,V. (Eds.), Soils and Societies: Perspectives from Environmental History. WhiteHorse Press, UK, pp. 91–117.

ahowald, N., Artaxo, P., Baker, A.R., Jickells, T.D., Okin, G.S., Randerson, J.T.,Townsend, A.R., 2005. Impacts of biomass burning emissions and land usechange on Amazonian atmospheric phosphorus cycling and deposition. GlobalBiogeochem. Cycles 19, GB4030, doi:10.1029/2005GB002541.

ajor, J., DiTommaso, A., Lehmann, J., Falcão, N.P.S., 2005. Weed dynamics on Ama-zonian Dark Earth and adjacent soils of Brazil. Agric. Ecosyst. Environ. 111,1–12.

artin, J.F., Roy, E.D., Diemont, S.A.W., Ferguson, B.G., 2010. Traditional Ecologi-cal Knowledge (TEK): ideas, inspiration, and designs for ecological engineering.Ecol. Eng. 36, 839–849.

athewson, K., 1987. Estimating labor inputs for the Guayas raised fields: initial con-siderations. In: Denevan, W.M., Mathewson, D., Knapp, G. (Eds.), Pre-HispanicAgricultural Fields in the Andean Region, Part I. British Archaeological Reports.International Series 359(i), Oxford, UK, pp. 321–336.

cKey, D., Rostain, S., Iriarte, J., Glaser, B., Birk, J.J., Holst, I., Renard, D., 2010. Pre-

Columbian agricultural landscapes, ecosystem engineers, and self-organizedpatchiness in Amazonia. PNAS 107 (17), 7823–7828.

itsch, W.J., Nahlik, A., Wolski, P., Bernal, B., Zhang, L., Ramberg, L., 2009. Tropi-cal wetlands: seasonal hydrologic pulsing, carbon sequestration, and methaneemissions. Wetland Ecol. Manage. 18, 573–586.

S

S

neering 45 (2012) 30– 44 43

ollard, E., Walter, A., 2008. Agricultures Singulières. IRD, Paris.orris, A.S., 2004. Raised Field Technology: the Raised Fields Projects Around Lake

Titicaca. King’s SOAS Studies in Development Geography. Ashgate PublishingLimited, Aldershot, UK.

oser, K.F., Ahn, C., Noe, G.B., 2009. The influence of microtopography on soil nutri-ents in created mitigation wetlands. Restor. Ecol. 17, 641–651.

use, M., Quinteros, F., 1987. Experimentos de reactivación de campos elevados,Penón del Río, Guayas, Ecuador. In: Denevan, W.M., Mathewson, D., Knapp,G. (Eds.), Pre-Hispanic Agricultural Fields in the Andean Region, Part I. BritishArchaeological Reports. International Series 359(i), Oxford, UK, pp. 249–266.

ordenskiöld, E., 2009 [1916]. Indian adaptations in flooded regions of South Amer-ica. J. Latin Am. Geogr. 8, 209-224.

osetto, M.D., Jobbágy, E.G., Jackson, R.B., Sznaider, G.A., 2009. Reciprocal influenceof crops and shallow ground water in sandy landscapes of the Inland Pampas.Field Crop. Res. 113, 138–148.

dum, H.T., Odum, B., 2003. Concepts and methods of ecological engineering. Ecol.Eng. 20, 339–361.

lde Venterink, H., Pieterse, N.M., Belgers, J.D.M., Wassen, M.J., de Ruiter, P.C., 2002.N, P and K budgets along nutrient availability and productivity gradients inwetlands. Ecol. Appl. 12 (4), 1010–1026.

arsons, J.J., Shlemon, R., 1987. Mapping and dating the prehistoric raised fields ofthe Guayas Basin, Ecuador. In: Denevan, W., Mathewson, K. and Knapp, G. (Eds.),Pre-Hispanic Agricultural Fields in the Andean Region, Part I. British Archaeo-logical Reports 359(ii), Oxford, pp. 207–216.

earsall, D.M., 1987. Evidence for preHispanic maize cultivation on raised fieldsat Penón del Rio, Guayas, Ecuador. In: Denevan, W.M., Mathewson, D., Knapp,G. (Eds.), Pre-Hispanic Agricultural Fields in the Andean Region, Part I. BritishArchaeological Reports. International Series 359(i), Oxford, UK, pp. 279–295.

élissier, P., 2008 (1966). Les paysans du Sénégal. Les civilisations agrairesdu Cayor à la Casamance. Electronic version of the book of the same title(Fabrègue, Saint-Yrieix,:France). Dakar, UCAD – Département d’Histoire,juin 2008. http://www.histoire-ucad.org/archives/index.php?option=com remository&Itemid=60&func=select&id=22 (Consulted Nov 16, 2010).

erfecto, I., Vandermeer, J., 2008. Biodiversity conservation in tropical agroe-cosystems. A new conservation paradigm. Ann. N. Y. Acad. Sci. 1134, 173–200.

lazas, C., Falchetti, A.M., 1990. Manejo Hidráulico Zenú. Ingenierías Prehispánicas.Fondo FEN Colombia. Instituto Colombiano de Antropología, Bogotá.

retty, J., 2008. Agricultural sustainability: concepts, principles and evidence. Phil.Trans. R. Soc. B 363, 447–465.

eichel-Dolmatoff, G., Reichel-Dolmatoff, A., 1974. Un sistema de agricultura pre-historica de los Llanos Orientales. Rev. Colomb. Antropol. 13, 189–200.

ijsberman, F., de Silva, S., 2006. Sustainable agriculture and wetlands. In: Verho-even, J.T.A., Beltman, B., Bobbink, R., Whigham, D.F. (Eds.), Wetlands and NaturalResource Management. Ecological Studies, vol. 190. Springer, Heidelberg, pp.33–52.

oberts, K.G., Bloy, B.A., Joseph, S., Scott, N.R., Lehmann, J., 2010. Life cycle assess-ment of biochar systems: estimating the energetic, economic, and climatechange potential. Environ. Sci. Technol. 44, 827–833.

obertson, G.P., Swinton, S.M., 2005. Reconciling agricultural productivity and envi-ronmental integrity: a grand challenge for agriculture. Front. Ecol. Environ. 3,38–46.

odriguez-Iturbe, I., D’Odorico, P., Laio, F., Ridolfi, L., Tamea, S., 2007. Challengesin humid land ecohydrology: interactions of water table and unsaturatedzone with climate, soil, and vegetation. Water Resour. Res. 43, W09301,doi:10.1029/2007WR006073.

oger, P.A., 1996. Biology and Management of the Floodwater Ecosystem in Rice-fields. IRRI/ORSTOM, Manila, Philippines.

ostain, S., 1994. L’occupation amérindienne ancienne du littoral de Guyane.TDM129. Editions d’ORSTOM. 2 volumes.

ostain, S., 2008. Agricultural earthworks on the French Guiana coast. In: Sil-verman, H., Isbell, W. (Eds.), Handbook of South American Archaeology.Springer/Kluwer/Plenum, New York, pp. 217–233.

ubio, G., Oesterheld, M., Alvarez, C.R., Lavado, R.S., 1997. Mechanisms for theincrease in phosphorus uptake of waterlogged plants: soil phosphorus avail-ability, root morphology and uptake kinetics. Oecologia 112, 150–155.

aavedra, O., 2009. Culturas Hidráulicas de la Amazonía Boliviana. Oxfam, La Paz.anders, W.T., Parsons, J.R., Santley, R., 1979. The Basin of Mexico: Ecological Pro-

cesses in the Evolution of Civilization. Academic Press, New York.cherr, S.J., McNeely, J.A., 2008. Biodiversity conservation and agricultural sustain-

ability: towards a new paradigm of ‘ecoagriculture’ landscapes. Phil. Trans. R.Soc. B 363, 477–494.

iemens, A.H., 2000. Resistance to Agricultural Globalization: Walking South on aNorthbound Train? Revista del CESLA, University of Warsaw, Poland.

iemens, A.H., 2004. Modeling the tropical wetland landscape and adaptations.Agric. Hum. Values 21, 243–254.

iemens, A.H., Hebda, R.J., Navarrete Hernández, M., Piperno, D.R., Stein, J.K., ZolaBaez, M.G., 1988. Evidence for a cultivar and a chronology from patterned wet-lands in central Veracruz, Mexico. Science 242, 105–107.

luyter, A., 1994. Intensive wetland agriculture in Mesoamerica: space, time, and

form. Ann. Assoc. Am. Geogr. 84, 557–584.

mardon, R.C., 2006. Heritage values and fucntions of wetlands in Southern Mexico.Landscape Urban Plan. 74 (3–4), 296–312.

pencer, C.S., 2000. Food scarcity, rural poverty, and agricultural development inLatin America: issues and evidence. Cult. Agric. 22, 1–14.

4 l Engi

S

S

S

S

S

T

T

T

T

T

V

V

V

V

V

V

W

W

W

W

Y

4 D. Renard et al. / Ecologica

pencer, C.S., Redmond, E.M., Rinaldi, M., 1994. Drained fields at La Tigra, VenezuelanLlanos: a regional perspective. Latin Am. Antiq. 5, 95–110.

tanish, C., 1994. The hydraulic hypothesis revisited: Lake Titicaca basin raised fieldsin theoretical perspective. Latin Am. Antiq. 5 (4), 312–332.

teiner, C., Glaser, B., Teixeira, W., Lehmann, J., Blum, W.E.H., Zech, W., 2008.Nitrogen retention and plant uptake on a highly weathered central Amazo-nian Ferralsol amended with compost and charcoal. J. Plant Nutr. Soil Sci. 171,893–899.

teiner, C., Teixeira, W.G., Lehmann, J., Nehls, T., de Macedo, J.L.V., Blum, W.E.H., Zech,W., 2007. Long term effects of manure, charcoal and mineral fertilization on cropproduction and fertility on a highly weathered Central Amazonian upland soil.Plant Soil 291, 275–290.

temper, D.M., 1987. Raised fields and agricultural production, AD 1400-1600,Rio Duale, Guayas, Ecuador. In: Denevan, W.M., Mathewson, D., Knapp, G.(Eds.), Pre-Hispanic Agricultural Fields in the Andean Region, Part I. BritishArchaeological Reports. International Series 359(i), Oxford, UK, pp. 297–320.

hurston, T.L., Fisher, C.T. (Eds.), 2006. Seeking a Richer Harvest. The Archaeol-ogy of Subsistence Intensification, Innovation and Change. Springer ScientificPublishing, New York.

imsina, J., Connor, D.J., 2001. Productivity and management of rice-wheat croppingsystems: issues and challenges. Field Crop Res. 69, 93–132.

orres-Lima, P., Canabal-Cristiani, B., Burela-Rueda, G., 1994. Urban sustainable agri-culture: the paradox of the chinampa system in Mexico City. Agric. Hum. Values11, 37–46.

urner, B.L., Harrison, P.T., 1983. Pulltrouser Swamps: Ancient Maya Habitat, Agri-

culture and Settlement in Northern Belize. University of Texas Press, Austin.

urner I.I., B.L., Miksicek, C.H., 1984. Economic plant species associated with prehis-toric agriculture in the Maya Lowlands. Econ. Bot. 38, 179–193.

aldez, F. (Ed.), 2006. Agricultura Ancestral Camellones y Barradas. Contexto Social,Usos y Retos del Pasado y Presente. Ediciones Abya-Yala, Quito, Ecuador.

Z

Z

neering 45 (2012) 30– 44

andermeer, J.H., 2011. The Ecology of Agroecosystems. Jones and Bartlett, Sudbury,MA, USA.

anloqueren, G., Baret, P.V., 2009. How agricultural research systems shape a tech-nological regime that develops genetic engineering but locks out agroecologicalinnovations. Res. Policy 38, 971–983.

asey, D.E., Harris, D.R., Olson, G.W., Spriggs, M.J.T., Turner II, B.L., 1984. The role ofstanding water and water-logged soils in raised-field, drained-field and island-bed agriculture. Singapore J. Trop. Geogr. 5 (1), 63–72.

ersteeg, A.H., 2008. Barrancoid and Arauquinoid mound builders in coastalSuriname. In: Silverman, H., Isbell, W. (Eds.), Handbook of South AmericanArchaeology. Springer/Kluwer/Plenum, New York, pp. 303–318.

itousek, P.M., Howarth, R.W., 1991. Nitrogen limitation on land and in the sea: howcan it occur? Biogeochemistry 13, 87–115.

alker, J.H., 2004. Agricultural Change in the Bolivian Amazon. University of Pitts-burgh Latin American Archaeology Publications No. 13. University of Pittsburgh.

an, S., Hui, D., Luo, Y., 2001. Fire effects on nitrogen pools and dynamics in terres-trial ecosystems: a meta-analysis. Ecol. Appl. 11, 1349–1365.

ilson, C., Simpson, I.A., Currie, E.J., 2002. Soil management in pre-Hispanic raisedfield systems: micromorphological evidence from Hacienda Zuleta, Ecuador.Geoarchaeology 17 (3), 261–283.

iseman, F.M., 1983. Analysis of pollen from the fields at pulltrouser swamp. In:Turner, B.L., Harrison, P.D. (Eds.), Pulltrouser Swamp: Ancient Maya Habitat,Agriculture and Settlement in Northern Benize. University of Texas Press, Austin,pp. 105–119.

anai, Y., Toyota, K., Okazaki, M., 2007. Effects of charcoal addition on N2O emis-sions from soil resulting from rewetting air-dried soil in short-term laboratory

experiments. Soil Sci. Plant Nutr. 53, 181–188.

ucchi, A., Denevan, W.M., 1972. Campos elevados e historia cultural prehispánicaen los llanos occidentales de Venezuela. Montalban (Caracas) 9, 565–736.

uckerman, B.M., Matheny, M., Acosta, N., 1994. Control of plant-parasitic nema-todes by a nematicidal strain of Aspergillus niger. J. Chem. Ecol. 20 (1), 33–43.