Computer-based tools for decision support in agroforestry: Current state and future needs

22 Pages·2013·0.54 MB·English

UUnniivveerrssiittyy ooff NNeebbrraasskkaa -- LLiinnccoollnn
DDiiggiittaallCCoommmmoonnss@@UUnniivveerrssiittyy ooff NNeebbrraasskkaa -- LLiinnccoollnn
U.S. Department of Agriculture: Forest Service --
USDA Forest Service / UNL Faculty Publications
National Agroforestry Center
April 2004
CCoommppuutteerr--bbaasseedd ttoooollss ffoorr ddeecciissiioonn ssuuppppoorrtt iinn aaggrrooffoorreessttrryy::
CCuurrrreenntt ssttaattee aanndd ffuuttuurree nneeeeddss
E. A. Ellis
University of Florida, PO Box 110410, Gainesville, FL
G. Bentrup
USDA National Agroforestry Center, USFS Rocky Mountain Research Station
M. M. Schoeneberger
2USDA National Agroforestry Center, USFS Rocky Mountain Research Station
Follow this and additional works at: https://digitalcommons.unl.edu/usdafsfacpub
Part of the Forest Sciences Commons
Ellis, E. A.; Bentrup, G.; and Schoeneberger, M. M., "Computer-based tools for decision support in
agroforestry: Current state and future needs" (2004). USDA Forest Service / UNL Faculty Publications. 4.
https://digitalcommons.unl.edu/usdafsfacpub/4
This Article is brought to you for free and open access by the U.S. Department of Agriculture: Forest Service --
National Agroforestry Center at DigitalCommons@University of Nebraska - Lincoln. It has been accepted for inclusion
in USDA Forest Service / UNL Faculty Publications by an authorized administrator of DigitalCommons@University of
Nebraska - Lincoln.
AgroforestrySystems 61: 401–421,2004. 401
©2004KluwerAcademicPublishers. PrintedintheNetherlands.
Computer-based tools for decision support in agroforestry: Current state
and future needs
E.A. Ellis1,∗, G.Bentrup2 and M.M.Schoeneberger2
1School of Forest Resources and Conservation, University of Florida, PO Box 110410, Gainesville, FL 32611-
0410, USA; 2USDA NationalAgroforestry Center, USFSRocky MountainResearch Station, UNL-East Campus,
∗
Lincoln,Nebraska68583–0822,USA; Authorforcorrespondence:e-mail:eaellis@ufl.edu
Keywords:Databases,DecisionSupportSystems,GeographicalInformationSystems,Models
Abstract
Successfuldesignofagroforestrypracticeshingesontheabilitytopulltogetherverydiverseandsometimeslarge
sets of information(i.e., biophysical, economic and social factors), and then implementing the synthesis of this
informationacrossseveralspatialscalesfromsitetolandscape.Agroforestry,byitsverynature,createscomplex
systems with impacts ranging from the site or practice level up to the landscape and beyond. Computer-based
DecisionSupportTools(DST)helptointegrateinformationtofacilitatethedecision-makingprocessthatdirects
development,acceptance,adoption,andmanagementaspectsinagroforestry.Computer-basedDSTsincludedata-
bases,geographicalinformationsystems,models,knowledge-baseorexpertsystems,and‘hybrid’decisionsupport
systems.ThesedifferentDSTsandtheirapplicationsinagroforestryresearchanddevelopmentaredescribedinthis
paper. Althoughagroforestrylacks the large research foundationof its agricultureand forestrycounterparts, the
developmentanduseofcomputer-basedtoolsinagroforestryhavebeensubstantialandareprojectedtoincreaseas
therecognitionoftheproductiveandprotective(service)rolesofthesetree-basedpracticesexpands.Theutilityof
theseandfuturetoolsfordecision-supportinagroforestrymusttakeintoaccountthelimitsofourcurrentscientific
information, the diversity of aspects (i.e. economic, social, and biophysical) that must be incorporated into the
planning and design process, and, most importantly, who the end-user of the tools will be. Incorporating these
tools into the design and planningprocesswill enhancethe capabilityof agroforestryto simultaneouslyachieve
environmentalprotectionandagriculturalproductiongoals.
Introduction small amount of land they occupy (Guo 2000; Nair
2001; Ruark et al. 2003). Realizing this potential is,
‘Few things disappoint a landowner more than however, a complex task of determining what op-
spendingmoney,time,andeffortonaprojectthat portunities, limitations, and trade-offs exist in each
fails ... especiallyone like agroforestry, where it situation, and of designing an agroforestry practice
canbeyearsbeforeproblemsbecomeapparent’ thatachievesthebestbalanceamongthem.Thereare
(DosskeyandWells2000). numerous impacts created by agroforestry plantings,
rangingfromintendedto nonintendedand, therefore,
Agroforestry, the deliberate integration of trees
ranging from detrimental to advantageous, occurring
into crop and livestock operations, has the potential
both on- and off-site, and varying over time. Con-
toachievemanyoftheenvironmental,economic,and
sequently, if agroforestryis to be a viable strategy in
socialobjectivesbeingdemandedfromworkingland-
promotingagroecosystemsustainability,thedecision-
scapes by landowners and society. By adding struc-
making process must incorporate many considera-
tural and functional diversity to the landscape, these
tions,notonlyatthepracticescalebutalsoatthelarger
tree-basedplantingscanperformecologicalfunctions
scales of farm, landscape, and watershed (Schoene-
that have significance far greater than the relatively
402
berger et al. 1994). Simply put, agroforestry creates solutions (Schmoldt and Rauscher 1996). There are
a complex system of interactions that must be man- fivemajorcategoriesofcomputer-basedtechnologies
agedformultipleobjectives,multiplealternativesand used for decision support: databases, geographical
multiplesocialinterestsand preferences,while being information systems (GIS), computer-based models,
appliedoverawiderangeoflandscapesandlandscape knowledge-basedorexpertsystems(KBS),andhybrid
features. systems(Table1).
The decision-making process involved in agro- Inthepastdecadetheuseofcomputer-basedtech-
forestryresearch,developmentandapplicationiscom- nologiesin agriculture, forestry, and naturalresource
posed of several components: the person or group managementhasbeenimpressive.Inthefieldofagri-
making the decision, the problem, the approach or culture,theuseofcomputerscanbeconsideredpartof
method to solve the problem, and the decision. De- theagriculturalrevolutionofthe20th century,advan-
cision support tools (DST) are a wide variety of cingscientificresearch,facilitatingfarmmanagement,
technologies that can be used to help integrate di- and improving production (Paarlberg and Paarlberg
verse and large sets of information. DSTs do not 2000). The developmentof crop models, expert sys-
replacethedecision-makingbythelandownerornat- temsforagriculturalmanagement,andprecisionfarm-
ural resource manager, but they do facilitate the ing that incorporates GIS has advanced considerably
decision-makingprocessbymakingtheplanningpro- andarebeingusedinmanyfarmingoperations(NRC
cess more informed and more objective (Grabaum 1997; Zazueta and Xin 1998; Ahuja et al. 2002).
and Meyer 1998). Although agroforestry, like most Forestry,comparedtoproductionagriculture,isoften
natural-resource management sciences, is character- facedwithmorecomplexandmultipleobjectiveman-
ized by high complexity of which we have limited agementscenarios.Theadoptionanduseofcomputers
understanding and data (Sanchez 1995; Nair 1998), for decision support in forestry have had to evolve
the science and application of agroforestry can be from simpler mathematic models used for harvesting
greatlyenhancedthroughtheuseofthesetools. scheduling to more complex computer DSTs used to
help make management decisions where timber pro-
duction must be balanced with wildlife conservation,
Computer-baseddecisionsupporttoolsinnatural water quality, recreation, and other objectives, often
resourcemanagement involving policy and social issues (Rauscher 1999).
Table2listssomeofthemorerecognizedDSTsused
Computers now play an integral role for information in agricultural and forest management and describes
managementanddecision-makinginalldisciplinesre- their degree of complexities and integration of the
latedtonaturalresourcemanagement.Constantaccre- majorcomputer-basedtechnologies.
tion of data and information on agriculture, forestry,
agroforestry, and natural resource management has Databases
created the need to synthesize, organize and manip-
ulate this growing knowledge base and facilitate its Databases are computer-based tools used to access
accessibility and use for education, research, and and query large quantities of data and information.
decision-making(Davis1988;SchmoldtandRauscher They are often key components within other DSTs.
1996). The complexity of natural resource manage- The database DST consists of a database (a logically
ment,consideringthediversityofresources,interests, coherent collection of data) and a database manage-
objectives, constraints, and stakeholders involved, mentcomponent(thesoftwaresystem), whichallows
adds to the difficulty of making sound management a user to define, create, and maintain a database
decisions (Schmoldtand Rauscher 1996). Computer- (Mata-Toledo and Cushman 2000). Computer data-
based DSTs provide an effective means to compile basesareoftenimplementedasaRelationalDatabase
and sort out the medley of variables, information ManagementSystem(RDBMS),designedaroundthe
and knowledge (quantitative, qualitative, spatial, and mathematical concepts of relational algebra linking
heuristic) that managers must consider when mak- together two-dimensional data tables (Sanders 2000;
ing informedmanagementdecisions. In other words, SundericandWoodhead2001).Querystatementscan
DSTs synthesize the wide array of information and be developed, allowing users to search and analyze
offer a holistic approach for evaluating land and re- dataaswellasextractspecificinformationfromhuge
sourcemanagementproblemsandfindingappropriate datasets. This ability to extract pieces of informa-
403
Table1. Majorcategoriesofcomputer-baseddecisionsupporttechnologies.
Category Description
Databases Organizes andfacilitates themanagement andquerying oflargequantities ofdataand
information
GeographicalInformationSystems(GIS) Brings inageographic orspatial component toadatabase; manages, manipulates and
analyzesspatialdata
Computer-BasedModels Mathematicalcomputermodelsthatrepresentrealworldprocessesandpredictoutcomes
basedoninputscenarios
Knowledge-BasedorExpertSystems Adopts ‘Artificial Intelligence’ in the form of organizing, manipulating and obtaining
solutions using knowledge in the form qualitative statements, expert rules (i.e. rules
ofthumb)andacomputerlanguagerepresentationsystemforstoringandmanipulating
knowledge.
HybridSystems Integrates twoormoreoftheabovecomputer-basedtechnologies (e.g. (GIS,KBSand
Models)formoreversatile,efficientandcomprehensiveDSTs.
Table2. Computer-baseddecisionsupporttoolsusedinagriculturalandforestrymanagement.
Decisionsupporttool Description Reference
GOSSYM-COMAX Usedformanagement ofwater, nitrogen, herbicide andgrowthregulator in Reddyetal.2002;
cotton GertsisandWhisler1998;
GLYCM Soybeanproductionmodel Timlinetal.2002;
Acocketal.1997;
Manning1996
CERES Productionmodelforcropsinthetropicsandsubtropics RitchieandOtter1985;
Ahujaetal.2002
CROPGRO Productionmodelforcropsinthetropicsandsubtropics Booteetal.1998a;
Booteetal.1998b;
Ahujaetal.2002
DSSAT Packageofcrop-soilmodelstofacilitatetheevaluationandapplicationofdif- Jonesetal.1998;
ferentcroppingsystemsandtheinputandorganizationofrelevant scientific Jonesetal.2003
data
FORPLAN/SPECTRUM Optimization modelsforforestmanagement toevaluate financial efficiency, Field1984;
landallocationandresourcescheduling Kentetal.1991;
Rauscher1999
STEWPLAN Knowledge-based computer tool to assist landowners develop stewardship KnoppandTwery2003
plansbasedonforeststanddescriptions.
NED Hybriddecisionsupporttoolintegratingforestmodels,GIS,graphicvisualiz- Tweryetal.2000;
ationandaknowledgebaseformulti-useforestmanagement. Tweryetal.,2003
EMDS Landscape scale tool that integrates GIS, knowledge-based reasoning and Reynoldsetal.2002;
decisionmodelingtechnologiesforecosystemmanagementdecisionsupport. Rauscher1999
EPIC Erosion-ProductivityImpactCalculatordeterminesrelationshipsbetweensoil Jonesetal.1991;
erosionandcropproductivity Easterlingetal.1997
CO2FIX Estimates and evaluates dynamics of carbon in forest management and Maseraetal.2003
afforestationprojects
404
tionbasedonuser-specifiedcriteriamakesaRDBMS Knowledge-basedsystems
an excellent computer-based technology for decision
Knowledge-Based Systems (KBS) or expert systems
support.Amultitudeofnatural-resource-relateddata-
are part of the broad field of Artificial Intelligence
bases(i.e., ecosystems, flora, fauna, soils, hydrology,
(AI)involvingthecreationofcomputerprogramsthat
etc.) are now widely available and used in manage-
attempt to mimic human intelligence or reasoning,
mentdecisions.
‘learn’ new information and tasks, and draw use-
ful conclusions aboutthe world around us (Patterson
Geographicalinformationsystems
1990). In a KBS, knowledge is defined as a ‘body
A GIS can be defined as a data management sys- offactsandprinciplesaccumulatedbyhumankindor
tem designed to input, store, retrieve, manipulate, the act, fact or state of knowing’ (Patterson 1990).
analyze, and display spatial data for the purposes of KBS are used to acquire, organize and manipulate
research and decision-making (De Mers 1997). In knowledge, often using heuristic rules, analogous to
a GIS, a database is associated with map features, ‘rules of thumb’ or ‘good judgments,’ to help make
and data values are geographicallyreferenced, so re- sounddeductions(Nikolopoulos1997).Often,experts
source managers can spatially represent information in the subject are used to define these rules; how-
such as soil types or plant communities. Since land ever, knowledgefor a KBS can be acquiredfrom the
use and a diversity of related disciplines (i.e., agri- literature, databases or other sources. A ‘knowledge
culture,forestry,ruralplanning,andconservation)all engineer’ extracts knowledge, information and data
dealwithspatialcharacteristicsoflandscapes(Lacher from experts and other sources and translates it into
1998), GIS has gained considerable use in land use programminglanguagessoacomputercanutilizeand
planning and natural-resource management, provid- reason with it (Nikolopoulos 1997; Patterson 1990;
ingaspatialframeworktoaidinthedecision-making Schmoldt and Rauscher 1996). With an appropriate
process(Zeiler1999). user interface, the user can input problem scenarios
Additional technologies are often associated with and make enquiries to find solutions (Nikolopoulos
GIS, such as Global Positioning Systems (GPS) and 1997).
remote sensing. GPS is a means for inputting spa-
tial data with real world coordinates into a GIS and Hybridsystems
has become an important tool for researchers locat-
Many DSTs today integrate a variety of computer-
ing and recording information in the field. Remote
decision supporttechnologiessuch as RDBMS, GIS,
sensinginvolvesusingspatialdatafromphotographic
Models and KBS (Davis 1988; Liebowitz 1990). In-
andsatelliteimages,andsoftwaretoolstoanalyzeand
creasingly, land-use planning and natural-resource
interpretthesedata.
managementDSTs are merging GIS and KBS to de-
velopveryeffectiveandefficientspatialplanningtools
Computer-basedmodels
(LohandRykiel1992).Nowadays,applicationdevel-
For the most part, computer-based models refer to opment programs, modeling tools and GIS software
the translation of data and information into a math- are designed to be compatible with other systems
ematical form using algorithms to represent a real andallowarelativelyeasyintegrationofthedifferent
world‘process’or‘system’andto forecastoutcomes computer-basedtechnologies.
of different scenarios. In the realm of environmental
andnatural-resourcerelatedfields,thesemodelstryto
Applicationsofcomputer-basedDSTsfor
mathematically represent ecologicalprocesses (Skid-
agroforestry
more 2002). Environmental models mathematically
define ecological interactions and processes between
Considerable advances have been made in research,
bioticandabioticcomponentsbasedoncurrentorpast
planninganddevelopmentforavarietyofagroforestry
conditionsorstatesofthesecomponents.Thegoalof
systems in a wide range of agroecological regions,
thesemodelsistoquantitativelypredictfuturestatesof
from tropical to temperate. Prior to 1991, computer
these components,servinga valuablerolein defining
use in agroforestry research began with the develop-
thekeyprocessesinagroforestrypractices(Pengetal.
ment of databases as aids in guiding plant selection
2002;Skidmore2002).
(Nair1998).Astheuseofagroforestryhasbroadened
405
to address such issues as climate change and crop different locations (Nair 1987; Oduol et al. 1988).
growth, carbon sequestration, biodiversity and even AFSI was apparentlydevelopedas a researchandin-
green infrastructure, so has the need to simulate formationtoolforresearchers,particularlyinICRAF.
agroforestry’slonger-termeffectsacrosslargerscales, No documentationcould be found about currentver-
further necessitating use of computer-based DSTs. sions or availability of AFSI. Unfortunately, many
These early DSTs used in agroforestry were gener- earlyDSTslike AFSIoftenfailto bemaintainedand
ally those already in place in the fields of agriculture upgradedandthereforefadewithtime.
and forestry. For instance, the effect of shelterbelts AnotherearlyagroforestrydatabasewastheMulti-
on maize productivity under hypothesized climate Purpose Tree and Shrub Database (MPTS), also de-
change scenarios was examined using the Erosion- veloped by ICRAF in 1991 (von Carlowitz et al.
ProductivityImpactCalculator(EPIC)cropmodel,an 1991).TheMPTSdatabase,developedforresearchers
agricultural model originally developed to determine and extension agents, helped to select the right tree
the relationship between soil erosion and crop pro- or shrub species for agroforestry practices, primarily
ductivity (Jones et al. 1991; Easterling et al. 1997). forthetropicsandsubtropics.MPTSDatabaseVersion
Even today, many of these models developedfor ag- 1.0 containedinformationfor 1093species including
riculture or forestry are still a first choice for use in site-specific requirements(e.g., soils), morphological
agroforestryexercises.CO2FIX,auser-friendlymodel and phenological descriptions, management charac-
for dynamically estimating the carbon sequestration teristics and environmental responses (Schroder and
potentialofforestmanagementandafforestationpro- Jaenicke1994).Asimpleclimatemodelwasincluded
jects, is readily adaptable for agroforestry (Masera to predict climatic conditions based on the input of
etal.2003). geographical coordinates, and tree and shrub spe-
Today we have several DSTs developed exclus- cies were selected via a database search or query
ivelyforagroforestryapplicationsforthepurposesof that matched the descriptors that the user selected.
selecting suitable species, identifying suitable lands, The descriptors included 19 different criteria cover-
modeling different systems and predicting outcomes ing aspects of location, climate and soil conditions,
of different scenarios. Several different types of products, environmental services, management and
computer-based DSTs that have been applied or are cultivation.Theuserwasalsoabletousebooleanop-
stronglyapplicabletoagroforestryresearch,planning erators (i.e., and, or, and not) to fine-tune the search
and development are listed in Table 3 and are dis- to their specific needs. References are also included
cussedinfurtherdetailbelow.Additionallytheinten- toprovidefurtherinformationonselectedagroforestry
deduses,targetedend-usersandcurrentstatusofthese species(SchroderandJaenicke1994).
majoragroforestryDSTsaresummarizedinTable4. The current and revised version of MPTS is now
the AgroforesTree Database (AFT). Unlike its pre-
Agroforestrydatabases decessor this database is more widely accessible and
available on the Internet and as a CD-ROM. It is a
An initial effort to use computers to manage agro- database managementsystem intendedfor use by re-
forestry data began in the late 1980s with the searchersandfieldworkerstoselectagroforestrytrees
AgroforestrySystemsInventoryDatabase(AFSI)de- that are being deliberately grown and managed for
veloped by the International Centre for Research in more than one output and expected to make signific-
Agroforestry (ICRAF), now the World Agroforestry ant economic and/or ecological impacts (Salim et al.
Centre. AFSI involved a global collection of data or 1998; World Agroforestry Centre 2003a). More than
informationonagroforestrysystemsusingaquestion- 300 species are incorporated into AFT with inform-
naire as the survey instrument. Data and information ation on ecology and distribution, propagation and
collected andenteredinto the database includedgen- management, functional uses and pest and diseases
eral description, geographical location, biophysical (Salim et al. 1998). With AFT, users are able to
characteristics, socioeconomicaspects, systemevalu- searchandselecttreesaccordingtogeographicalloca-
ation, components of the system and their uses, and tion,biophysicallimitsandothermanagementcriteria
identificationofresearchgaps.WithAFSI,theuseris selected. In addition, there are references, research
abletoquerythedatabase,extractinginformationsuch contacts, a seed supplier’s directory, images of trees,
asgeographicallocationsofdifferentagroforestrysys- and a glossary of terms to help agroforesters obtain
tems and the species found within these systems in vital information and make wise decisions concern-
406
Table3. Computer-baseddecisionsupporttoolsusedinagroforestry.
Decisionsupporttool Type Description Reference
AFSI Database Agroforestry system inventory database describ- Nair1987;Oduoletal.
(Agroforestry Systems Inventory inggeographiclocationandbiophysical,socioeco- 1988
Database) nomicandspeciescharacteristics
MPTS Database Multi-purposetreeandshrubdatabaseusedfortree SchroderandJaenicke
(MultipurposeTreeandShrub selectionandspeciesinformation 1994
Database)
AgroforesTreeDatabase Database InternetandCD-Romapplicationforreferenceand Salimetal.1998;
selectionguideofagroforestrytrees. WorldAgroforestry
Centre2003a
SubtropicalTreeandShrub Database On-linedatabaseonpotentialagroforestrytreeand Ellisetal.2003
Database shrubspeciesfortheAmericansubtropics.
ForestryCompendium Database Compilationofknowledgeonforestry,agroforestry CABI1998;
andplantations andinformationontreesforman- Kleineetal.2003
agementdecision-makingandspeciesselection
AgroforestrySystemSuitability in GIS Spatial analysis using climate, soil land use and Boothetal.1989;
Africa other spatial data alongside plant species data to Boothetal.1990;
determinespeciesandagroforestrysuitability Unruh and Lefebvre
1995
AgroforestrySystemSuitability in GIS SpatialanalysistodeterminesuitableareasofAn- Bydekerkeetal.1998
Ecuador nona cherimola agroforestry systems in Southern
Ecuador.
Agroforestry System Assessment GIS Spatialsuitabilityassessmentforwillowandforest BentrupandLeininger
inNebraska farmingagroforestrysystemsinaNebraskawater- 2002
shed
AgroforestryParklandsinBurkina GIS Spatialanalysisofdynamicsofagroforestrypark- Bernardand
Faso landsandspeciesdistributionduetohumanimpacts Depommier1997
HistoricalTransformationof GIS Spatialanalysisofcensusandgeomorphologicdata Paquetteand
AgroforestryLandscape toexploredynamicsofagroforestryin19thcentury Domon1997
inCanada Canadianlandscape
Field-levelspatialanalysisof GIS Spatial analysis using ground penetrating radar Joseetal.2001
temperateagroforestrysystem (GPR) to evaluate root biomass and distribution
andsoilnutrientcrop-treeinteractionsintemperate
alleycropping
AME ModelingTool Object-oriented tooltographically visualize, con- Muetzelfeldtand
(AgroforestryModeling struct,integrateandexchangeagroforestrymodels Taylor1997
Environment)
HyPAR Model Biophysicalmodelcombiningcropandforestmod- Mobbsetal.2001
elsandintegrating climate, hydrology, lightinter-
ception,waterandnutrientcompetition,andcarbon
allocationprocessesinagroforestrysystems
HyCAS Model Biophysical model for agroforestry systems with MatthewsandLawson
cassavasimulatingcompetitionforlight,waterand 1997
nutrientsincludingphosphoruscycles
407
Table3. Continued
Decisionsupporttool Type Description Reference
WaNuLCAS Model Biophysical modeloftree-crop interactions based VanNoordwijkand
(Water,NutrientandLight onabove and below-ground resource capture and Lusiana1999;
CaptureinAgroforestySystems) competitionofwater,nutrientsandlightunderdif- WorldAgroforestry
ferent management scenarios in agroforestry sys- Centre2003b
tems
SCUAF Model Nutrient cycling model predicts changes in soil YoungandMuraya
(SoilChangesunderagroforestry) conditions under different agroforestry systems 1990;
based on parameters of biophysical environment, Vermeulenetal.1993;
landuseandmanagement,plantgrowth,andplant– Menzetal.1997;
soilprocesses Macadongetal.1998;
Nelsonetal.1997
FALLOW Model Model to evaluate impacts of shifting cultivation WorldAgroforestry
(Forest,Agroforest,Low-value andGIS and fallow rotations at a landscape-scale evaluat- Centre2003c;
LandscapeorWasteland?) ing transitions in soil fertility, crop productivity, VanNoordwijk2002
biodiversityandcarbonstocks
BEAM Model Bioeconomicmodeltoassessphysicalandfinancial Willisetal.1993;
(Bio-economicAgroforestry performanceofagroforestrysystemsbasedontree WillisandThomas
Model) andcropbiometricandeconomicmodels 1997
AEM Model Economicmodeltoevaluate agroforestry incom- Middlemissand
(AgroforestryEstateModel) bination with other farm activities assessing ef- Knowles1996
fectsoftreeproductionandphysicalandfinancial
resourceson-farm
DESSAP Model Multi-objective linear programming model to as- Garcia-deCecaand
(AgroforestryPlanningModel) sess feasible agroforestry alternatives based on Gebremedhin1991
land,laborandcashconstraints
ATK KBS KBS to store, manipulate and analyze a variety Walkeretal.1995
(AgroforestryKnowledgeToolkit) of information and knowledge acquired on agro-
forestrysystems
AES KBS KBSusedheuristic knowledge orexpert‘rules of Warkentinetal.1990
(AgroforestryExpertSystem) thumb’ to determine optimal species and spacing
foralleycroppingsystemsinthetropics
AGFADOPT Decision KBSbasedondecisiontreesusedtoassessadop- Robotham1998
(AgroforestryAdoption TreeKBS tionofagroforestrybasedoneconomicandsocial
EvaluationTool) factorsfacedbysmall-scalefarmers
AgroforestryPlanningToolin Hybrid HybridDSTintegratingGISdata,regressionmod- Liuetal.1999
China GIS, Models els plus expert knowledge to assess biophysical,
andKBS socialandeconomicsuitabilityofPaulowniainter-
croppingagroforestrysystems
PLANTGRO Hybrid Plantation and agroforestry species selection tool Booth1996;
(PlantationandAgroforestry GIS/KBS integratesGISandexpertsystemonplantgrowth HackettandVanclay
SpeciesSelectionTool) 2003
SEADSS Hybrid Landscapeandsite-scaleagroforestryplanningand Ellisetal.2003
(SoutheasternAgroforestry Database, species selection DST for landowners and exten-
DecisionSupportSystem) GIS,KBS sion agents of Southeast US that integrates GIS,
treeandshrubdatabaseandexpertknowledge
408
Table3. Continued
Decisionsupporttool Type Description Reference
ConservationBufferPlanning Hybrid Suite of GIS, economic models and visualization Bentrupetal.2003
Tools for Western Cornbelt Re- GIS/Models/ toolforlandownersandresourcemanagerstoeval-
gion,USA Visualization uate agroforestry strategies in Midwest Cornbelt
regionoftheUSA
ing the use and selection of agroforestrytrees(Salim and Lefebvre (1995) performed a similar GIS ap-
et al. 1998). Even though the AgroforesTree Data- plication for sub-Saharan Africa to determine areas
base is recognized and linked to a variety of rural suitable for different agroforestry systems. Integrat-
andagriculturaldevelopmentWebsites,itsspecificuse ing ICRAF’s agroforestry database with spatial data
andimpactonagroforestryresearchanddevelopment on geographic regions, climate and land uses in the
projectsaredifficulttoassessatthisstage. region,theirapplicationwasabletomapoutpotential
Althoughnot solely for agroforestry, the Forestry regionsfor21specifictypesofagroforestrysystems.
Compendiumisanextremelyusefuldatabaseforagro- Most of the past agroforestry GIS applications
forestry research and planning. The development of mentioned above have been research-oriented. The
the Forestry Compendium was undertaken by both Southeastern Agroforestry Decision Support System
theCommonwealthAgriculturalBureauInternational (SEADSS),developedrecentlybytheCenterforSub-
(CABI)andInternationalUnionofForestryResearch tropical Agroforestry (CSTAF) at the University of
Organizations (IUFRO) and consists of a compila- Floridabringson-lineGIScapabilitiesdirectlytoex-
tionofknowledgeonmultipurposeforestry,including tensionagentsandlandowners; it offerscountysoils,
agroforestry, plantations, and natural forest manage- land use and other spatial data for selecting suitable
ment (CABI 1998; Kleine et al. 2003). The com- tree and shrub species in a specified location (Ellis
pendiumgivesinformationaboutwhattreescouldbe etal.2003).TheUSDANationalAgroforestryCenter
plantedina particularenvironmentandforwhatpur- (NAC)iscurrentlyusingGIStofacilitateconservation
poses, how they will perform, how they should be bufferplanningintheWesternCornBeltecoregionin
managed, and provides current documents available the central United States (Bentrup et al. 2000). GIS-
regardingeachspecies(Kleineetal.2003).ASpecies guided assessments, derived from publicly available
Selection Module aids in decision-makingfor select- datasets, are being used to evaluate four key issues
ing suitable species according to a variety of criteria oftheWestern Cornbelt:biodiversity,soilprotection,
includinggeographicallocation, climate, typeofuse, water quality, and agroforestry products. By com-
andothermanagementoptions(Kleineetal.2003). bining these assessments, information is generated
for use in identifying opportunities and constraints
GISapplicationsinagroforestry on the landscape where multiple benefits from con-
servation buffers, especially agroforestry plantings,
Considering that GIS technology is widely available
can be achieved (Bentrup et al. 2000). Utilizing the
and affordabletoday and the fact that agroforestryis
agroforestryproductassessments (Bentrup and Lein-
directly dependent upon spatial characteristics, it is
inger 2002) in conjunction with the riparian buffer
logicalto expectto haveseveralagroforestry-specific
connectivityassessments,areaswereidentifiedwhere
GISDSTs;buttherealityisthatonlyafewareavail-
riparian forest buffers could be located to improve
able. An early GIS application compiled information
habitatconnectivitywhileofferinglandownerstheop-
on 173 species including their descriptions, soil and
tionto growwoodyfloralsforprofit(G. Bentrupand
climate preferences, and management characteristics
T. Kellerman, presentation to 8th North American
for Africa (Booth et al. 1989). This application al-
AgroforestryConference,June2003).
lowed usersto querythe databaseand generatemaps
GIS-guided agroforestry suitability analysis will
showing the climatic suitability for different species.
only improve as spatial data and computer resources
At a regional scale, Booth et al. (1990) created a
become more accessible. Many states and countries
similarapplicationforZimbabwe,demonstratinghow
already are assembling internet-accessible GIS data
GIS applications can be done at many scales. Unruh
409
Table4. Uses,targetedend-usersandcurrentstatusofmajordecisionsupporttoolsusedinAgroforestry.
Decisionsupporttool Intendeduse Targetedend- Currentstatusandavailability
users
AFSI Generalagroforestry Researchers Nocurrentversionsoravailability
(AgroforestrySystems research & planning for andICRAF
InventoryDatabase) ICRAF
MPTS Species selection for agro- Researchers UpgradedtoAgroforesTree
(Multipurpose Tree and forestryresearch&planning & Extension
ShrubDatabase) Agents,
Foresters
AgroforesTreeDatabase Species selection for agro- Researchers CD-ROM 1998 and currently available on-line from
forestryresearch&planning & WorldAgroforestryCentre
(World-wide) Fieldworkers http://www.worldagroforestrycentre.org/Sites/TreeDBS/
AFT/AFT.htm
Subtropical Tree and Species selection and in- Landowners, Currently under development and evaluation. Avail-
ShrubDatabase formation for agroforestry Extension able on-line from Center for Subtropical Agroforestry
extension,planning&devel- Agents, http://cstaf.ifas.ufl.edu/tree&shrubdb.asp
opment (American Subtrop- Researchers
ics&Caribbean)
ForestryCompendium Species selection and in- Foresters, CD-ROM 2003 and internet version available through
Database formationforforestry, agro- Policy CABIInternational
forestry and plantation re- Makers,Con- http://www.cabi.org/compendia/fc/index.asp
searchandplanningandde- servationists,
velopment(World-wide) Consultants,
Extensionists
AME Developmentofagroforestry Researchers NowSIMILEforbuildinggeneralecologymodelsavail-
(Agroforestry Modeling modelsforresearch ablefromSimulistics
Environment) http://simulistics.com/
HyPARModel Researchonbiophysicalpro- Researchers HyPARv4.5available through Center forEcology and
cessandinteractionsinagro- Hydrology,Edinburgh,UK
forestrysystems http://www.nbu.ac.uk/hypar/
HyCASModel Researchonbiophysicalpro- Researchers HyCASavailablethroughCranfieldUniversity,UK
cessandinteractions incas- http://www.silsoe.cranfield.ac.uk/iwe/research/hycas.htm
savaagroforestrysystems
WaNuLCASModel Researchonbiophysicalpro- Researchers WaNuLCASv2.11AvailablethroughWorldAgroforestry
(Water,Nutrient cesses and interactions in Centre
and Light Capture in agroforestrysystems http://www.worldagroforestrycentre.org/sea/Products/
AgroforestrySystems) AFModels/WaNulCAS/index.htm
SCUAFModel Environmental evaluation of Researchers SCUAFv4.0available throughCentre forResource and
(SoilChangesunder agroforestry systems used EnvironmentalStudies,AustralianNationalUniversity
agroforestry) for research and develop- http://incres.anu.edu.au/imperata/imp-mods.htm
mentprojects
FALLOWModel Impact assessment on land- Researchers FALLOWavailablethroughWorldAgroforestryCentre
(Forest,Agroforest, scape dynamics due to so- http://www.worldagroforestrycentre.org/sea/Products/
Low-valueLandscape cioeconomic and land-use AFModels/FALLOW/Fallow.htm
orWasteland?) changes