Medicinal trees in smallholder

Medicinal trees in smallholder agroforestry systems: Assessing some factors influencing cultivation by farmers East of Mt. Kenya Jonathan Muriuki Kiura

Presentation summary The research problemResearch concept and objectivesStudy area and methodsResults and discussions Conclusions and recommendations2

The problem Indicator Austria EthiopiaKenyaMalawiRwandaTanzania Uganda Zambia Population (2010 est.) mio 8.4 79.5 38.6 15.6 10.4 43.2 31.813.3GDP (US$) trillions (2009)3852930552216128GDP per capita US$ (2009)38,7489361,5728591,0711,3581,2191,431Infant mortality rate (IMR) per 1000 births (2009)4.486.964.489.4112.472.676.992.7Under five mortality rate per 1000 births (2009)5.4145.3104.1131.8187.8118.4127.4157.0Maternal mortality rate per 100,000 live births (1999)NA87059011001100530510650Total fertility rate (2007/8)1.425.294.965.595.925.166.465.18Literacy rate (2007/8)99.035.973.671.864.972.373.670.6Life expectancy - years (2007/8)79.852.954.148.346.252.551.542.4Contraceptive use (%) (1999)NA15(05)46(09)41(08)36(08)26(05)24(06)41(07) Majority of Africa population is poor and ravaged by treatable diseases but can’t afford modern medicine

An example of malaria – quick facts Factor Statistic Annual number of malaria cases globally247 millionAnnual number of malaria deaths globally881000Number of malaria-endemic countries 109 Number of people at risk for malaria 3.3 billion Percent of global population at risk for malaria50% Percent of malaria deaths in Africa 91% Percent of malaria deaths in children under 5 85% Percent of symptomatic children under 5 in Africa treated with ACTs 3% Percent of at-risk people in Africa protected by Insecticide-Treated Nets26%Source: World Malaria Report 2008

Who offers treatment in Africa? Country Doctor : patient TMP : patientReferencesEritreaDoctors estimated at 120 in 1995Government of Eritrea, 1995 Ethiopia 1:33,000 World Bank, 1993 Kenya 1:7,142 (overall) 1:987 ( Urban - Mathare ) World Bank, 19931:833 (Mathare)1:378 (Kilungu)Good. 1987:Malawi1:50,0001:138Msonthi and Seyani, 1986Mozambique1:50,0001:200Green et al. 1994Sudan1:11,000-W Bank, 1993Swaziland1:10,0001: 100Green, 1985Hoff and Maseko,1986Tanzania1:33,0001:350-450 DSMW Bank, 1993, Swantz, 1984Uganda1:25,0001:708WBank, 1993, Amai, 1997Zambia1:11,000-World Bank, 1993Zimbabwe1:6,2501:234 (urban)1:956 (rural)World bank, 1993Gelfand et al. 1985

Medicine Plant Knowledge Phyto-medicines rely on two elements , plants (of which over 60% are perennial trees and shrubs) and the knowledge associated with their use. Either alone is useless. Traditional medicine, mainly herbal, has been substantially managing African health but is under threat6

Herbal medicine relying on wild plants collection is not viable because biodiversity in Africa is threatened by agriculture, urbanization etc and forests lost to below 10% in many countries (e.g. 1.7% in Kenya at present) Country area (x1000 hectares) FRA 2005 categories AustriaEthiopiaKenya Malawi RwandaTanzania Uganda Zambia Forest* 3862 13000 3522 3402 48035257362742452Forest and other wooded land39805765038442340254140013477745613Other land4293519811847260061926483461493328726Total land area82731096315691494082467883591971074339Inland water bodies1137991123244016761504394922Total area of country838611043058037118482634945092410475261% forest of total land area46.711.96.236.219.539.918.457.1% forest of total area of country46.111.86.128.718.237.315.056.47Extent of forest and other wooded land in Eastern Africa compared to Austria by 2005

With increasing trade and TM use, medicinal plant resource depletion is abundant and cultivation has been recommended as a possible solution But that is very easy for herbs (annuals) if appropriate germplasm and products markets are accessible. For trees and other long rotation woody perennials that poses a big challenge due to having to wait long and as long as wild resources are available and perceived to be a common good 8

9 Conservation through use under cultivation - What would be the ideal trend of growth in material supply as knowledge of use improves?

Research questions How do socio-economic factors influence the decisions by farmers to cultivate or conserve medicinal plants? How does ecology influence use and cultivation of medicinal tree species? 10Main hypothesis The level of medicinal tree cultivation (Mc) is a factor of germplasm availability (g), species ecology (e – climate, soil and competition), local disease burden perception with appropriate knowledge on use of medicinal trees (k), and availability of market for medicinal tree products (m). Mc = f(g, e, k, m, α )

Conceptual framework 11 General Conceptual Frame on farmer adoption of an agricultural practice Adapted from FAO (2001)

Smallholder production sub-system Household Consumption AlternativesSold to Markets Human CapitalProduction TechnologyAccess to inputs Other Products Income Fig. Conceptual framework showing some factors expected to influence cultivation of medicinal trees by smallholder farmers Germplasm (g) Medicinal trees (C) Timber, food, ethno veterinary, etc Clinical MedicineSelf treatment knowledge (k)Cultivation ecology (e)Motivation to plant(m)Niche defines quality and interaction with crops (opportunity cost to land and labourSeedling quality as well as access and cost may demotivateMotivates especially if not extractive harvestingIf alternative perceived better then only this path takenWild sourcesDemotivates depending on access and abundance12

Objectives 13 To collate the perspectives of farmers and herbalists on the factors influencing their preference and cultivation of tree species with medicinal value To assess the influence of local disease burden perception and knowledge of herbal treatment on the efforts by farmers and herbalists to cultivate medicinal trees To explore the contribution of farm grown herbal material to medicinal tree product markets and its effect on medicinal tree cultivation To explore how germplasm access by farmers and on-farm tree nurseries influence medicinal tree cultivation To explore motivational drivers of cultivation and the scope for herbalists’ and traders’ utilization of farm produced medicinal tree products

The study area 14 3 districts ( Embu, Mbeere and Meru central)Population density 100-500 persons / sq kmNine agroecological zones (LM5 to LH1)Rainfall – 500 -2600mm; Altitude 500 -2500maslSoils – varying from nitisols to ferrasols Mixed-crop and livestock agric systemsGood tree planting culture

Data collection methods Farmer group meetings - cultural domain analysis - 13 groups Individual interviews - analysis with SPSS 200 farmers60 herbalists60 nursery operators55 market players in 3 citiesSpecies abundance surveys in farms, forests and herbalist gardens - analysis with BiodiversityRPersonal observationsTriangulations - interview responses tested with empirical measurements15

Interview survey results Medicinal species present in farms and herbalist gardens Farms – 295 total species (trees – 45%, shrubs – 27% and herbs – 28%) Herbalists’ gardens – 203 total species (trees – 40%, shrubs – 27% and herbs – 33%)60 species known as medicinal by farmers but not recorded in any farm (22 trees, 26 shrubs and 12 herbs)Do farmers know more species than herbalists?????16

Factors influencing cultivation Factor influencing cultivation decision Farmers’ ratingHerbalists ranking Freq.% (n=200) Mean rate Freq % (n=60) Mean rank Knowledge of treatment 97 2.3 98 4.3 Access to medicinal products’ markets891.2982.9Germplasm availability801.2982.5Conservation of species that were getting scarce541.2984.0Species cultivation technology known811.3981.4Other uses of species191.321Herbalists knowledge issue was only a species treating many diseases17

Species highly preferred for cultivation Species Growth form Frequency (%) of preference byherbalists farmers Prunus africana Tree 56 26 Warburgia ugandensis Tree 56 7Aloe spp.*Herb4945Azadirachta indicaTree4047Olea europaea ssp africanaTree5118Strychnos henningsiiTree269Erythrina abyssinicaTree3510Myrsine melanophloeosTree2313Caesalpinia volkensiiShrub2614Zanthoxylum chalybeumTree126Senna didymobotryaTree9Ocotea usambarensisTree19Croton megalocarpusTree121118

Summary on farmers and herbalists’ perceptions Herbalists preferred trees that treat more diseases and are scarce – farmers knowledge then markets Farmers in Mbeere influenced by germplasm availability than marketsMultiple use of species not very important to influence both farmers and herbalistsCultivation technology rated low – but factors such as appropriate niches and farm sizes importantWomen farmers rated knowledge, markets and multiple use higher than menTrees on farm correlated loosely with the frequency of species preferenceUsually one tree per household is enough for self treatment and neighbours can use19

Most socio-economically important diseases Disease Herbalist s’ score Herb RankFarmers’ scoreFarm RankMalaria 10.7 1 11.2 1 Typhoid 5.7 7 8.5 2 Respiratory problems8.337.93HIV/AIDS8.626.74Pneumonia7.046.15Hypertension5.295.06Tuberculosis5.964.47Diabetes6.154.28Back/bones/joints aches3.6164.29Cancers5.0114.010Measles3.5183.911Dental disorders5.382.820Rheumatism4.8133.815Amoeba4.8123.716Asthma5.1102.62420

Disease effect management by farmers Health management measure Percent (n =142) Preventive (ex ante risk minimising)232Clean drinking water30 Contribute to development of community health facilities 3 Good diets 30 Immunization through vaccination 3 Keep useful medicine in house 19 Keeping warm 4Medicinal plant conservation51Other preventive methods1Other traditional health practices3Personal and household hygiene57Public health training and practices1Use of mosquito nets31Treatment (ex post risk coping)32Off the counter medicine4Seek conventional medicine assistance3Use of herbal medicine25Grand Total264*21

Number of species used in treating important diseases 22

Highly ranked species in treatment of most important diseases 23 Amoebiasis Back/joint/ bone problemsCough/fluDental problemsDiabetes Malaria Pneumonia Rheumatism Typhoid Aloe sp 6 5 10 3 952788Azadirachta indica2151222126101525Caesalpinia volkensii5149112Dalbergia melanoxylon63211Erythrina abyssinica563125674Moringa oleifera12Myrsine melanophloeos1332Olea europaea1432617173Prunus africana4135311113Strychnos henningsii9212 Warburgia ugandensis 2 2 8 4 10 2 1

Farmers’ sources of knowledge on use of medicinal plants for disease treatment Information sources Frequency (%) of mention as source number: (N=200)1 2 3 Total Herbalists 25 0 0 25 Nursery operators 3003Media (newspapers, radios )65010Older relatives (parents, grandparents )5721482Neighbours716932Seminars1315Exchange programmes by NGOs2416No response25385Total10010010030024Most information passed through genealogy and herbalists contribution is low!

Who speaks about importance of medicinal tree cultivation to farmers? 1st 2nd 3rd Total Herbalists 11 0 0 11 Tree nursery operators 6 1 07Media (newspapers, radios)2214Older relatives (parents, grandparents )6106Neighbours24378 Development programmes (govt, NGOs 16 7 1 23 Medicinal tree product buyers* 1 1 0 1 Own initiative* 14 3 1 17 No response 45 84 95 25

So knowledge of medicinal tree species varies with socio-demographic categories Socio-economic factor 1 2 3 4 5 6 P - value Gender 12.6 13.2 0.551 Age8.410.112.914.914.912.70.002Education level16.113.212.511.16.6 0.012 District 9.2 15.1 13.9 0.000 First response to symptom of illness by family member 15.1 10.8 12.2 9.0 26 Key: Gender -1(Female), 2 (Male); Age in years – 1 (<25), 2 (25-35), 3 (35-45), 4 (55-65), 5 (>65); Level of education attained – 1 (not schooled), 2 (primary level), 3 (village polytechnic), 4 (secondary), 5 (post secondary); District – 1 (Embu), 2 (Mbeere), 3 (Meru Central); First response to ailment - 1 (find a medicinal plant), 2 (buy an over the counter drug), 3(consult a medical clinic or hospital), 4 (consult a herbalist) Number of species known increased with age, district harshness, and use but decreased with education level attained by respondent

Does farmer’s knowledge influence cultivation 27

Summary on farmers knowledge on TM and its influence on med tree planting No difference in rating of disease economic importance between farmers and herbalists rate -same as hospitals Medicinal trees play role in household health Herbalists use more of wild species while farmers use agroforestry species moreFarmers learn about medicinal trees from relatives and cultivation mainly own initiativeThe medicinal species present in farms influenced more by the species known little by the perception of the socioeconomic importance of diseases28

Medicinal plant markets and cultivation of medicinal trees 29 Business categories Freq % (n=55) Av Trade period Av % annual growth Av no Species traded Av % volume purchased Av %of volume wild Final products 36 11 424 76929Herbal Clinic401715864559Pre-processing241710053672 Grand Total 100 15 241 6 53 51 Herbal clinics Pre-processors Final products

Sources of herbal materials in markets Species Growth habit Freq % (n=55) Av Trade period Av annual trade (Kg) Av annual growth % Av % from farm Demand trend Aloe spp Shrub 51 14 28633355RisingAzadirachta indicaTree441469349688RisingWarburgia ugandensisTree 24 11 333 231 44 Rising Eucalyptus spp Tree 22 10 117 600 98 Rising Prunus africana Tree 22 14 408 255 73 Rising Urtica dioica Herb 20 8 943 1122 30 Const Ekebergia capensis Tree 13 22 105 32 5 Rising Zanthoxylum gillettii Tree 13 15 175 109 0 Rising Albizia anthelmintica Tree 9 20 77 75 0 Rising Kigelia africana Tree 9 5 269 214 30 Const Moringa oleifera Tree 9 5 463 864 100 Rising Croton megalocarpus Tree 7 20 168 189 95 Rising Rhamnus prinoides Tree 5 6 104 150 33 Rising Senna didymobotrya Shrub 5 12 80 199 33 Const The numbers show the average per trader for each parameter; n=55 30

Traders’ preference for source of medicinal plant materials 31 Preferred source Reason for preference% of respondents giving reasonProcurement approachFarms (27% of respondents)Natural resource conservation5Own collection 39%Purchased 61%Good tree husbandry in farms 11 Species authenticity in farms4 Species scarcity in the wild 7 To create market / future sources 2 Total 29 Natural forests and woodlands (69% of respondents)Customary preference in practice2Own collection 51%Purchased 49%Mature plants and grown in rich substrate16Less contamination/interference15Little or no cost of procurement7Total76No preference (4% of resp)Depends on distance and costs2Own collection 100%No reason given24

But farmers reported little connection to markets Species Number of farmers Plant part sold Where sold Markhamia lutea 1 Roots Herbalist/neighbours (1) Myrsine melanophloeos 3 Seeds Vendors (1); Market (2); Neighbours (1) Olea europaea2CuttingsVendors (2)Osyris lanceolata1Whole plantVendors (1)Warburgia ugandensis1BarkHerbalist (1)32Trade in medicinal trees was rising but farmers were not participating in any significant mannerMost of indigenous tree species were traded collected from the wild – threatsTwo thirds of traders who purchased materials preferred farm sourced materials – mostly in the final products categoryGetting materials at little or no costs contributes to more wild collection – distance may discourage but not tested in this studySummary on trade

Focus on tree seedling sources District No of nurseries Av size (m”)Space with trees (%)Embu2010393Mbeere204794Meru Central2054479Total/Average60231 89 33 District Planted in garden Given away sold or given free Exclusively soldGrand TotalEmbu775523Mbeere30205Meru Central522312Total1588840Herbalists with nurseries (numbers are % n = 60)

Presence of highly preferred species in nurseries Species Growth habitFreq % (n = 60)Av no seedlings Av seedlings supplied beforeAv trend DemandPrunus africanaT 25 211 1111H Azadirachta indica T 13 37 43HOlea europaeaT122921477HAloe spp.H810161HCroton megalocarpusT8746CJuniperus proceraT84062223HHagenia abyssinicaT74108HCroton macrostachyusT5087CMarkhamia luteaT530593CMyrsinne melanophloeosT56683438HAcacia xanthophloeaT3533CBridelia micranthaT3025H34

Sources of medicinal trees in farms Species % farmsNRNeighboursTree nurseriesWildlingsAloe sp5217 6 5 64Azadirachta indica 27 9 4 59 6Croton macrostachyus244821729Prunus africana232422250Senna didymobotrya21672221Croton megalocarpus20334835Erythrina abyssinica20683325Tithonia diversifolia19248062Olea europaea174131235Psidium guajava161034823Solanum incanum1396004Terminalia brownii11810514Ocimum suave10805010Zanthoxylum chalybeum1090001035

Summary on seedling sources Farmers were mainly planting medicinal trees from wildlings Demand for medicinal species lower than timber and higher than fruits and fodder in Meru; lower than fruits and fodder in Mbeere and fruits in EmbuDemand for medicinal seedlings higher than supply in nurseries – but not all speciesNeed investment in diversifying germplasm in both private and herbalist nurseries36

Key species abundance surveys (focus on 30 most preferred) Forests and woodlands (herbalist in team) Species abundance Age/size distributionsEvidence of harvesting method damage37Farms In twenty farms and ten herbalist gardens in each district Key species abundance Age/size class distributions relating to regeneration method Niche in the farm Other competing household or market uses of the species

38

Species abundance survey results The thirty species were whose abundance was measured include Albizia gummiferaFicus sycomorusRhamnus priniodesAloe sp Ficus thonningii Ricinus communis Azadirachta indica Kigelia africana Senna didymobotrya Brideria micrantha Leonotis mollissimaSolanum incanumCeasalpinia volkensiiMoringa oleiferaStrychnos henningsiiCordia africanaOcotea usambarensisTithonia diversifoliaCroton macrostachyusMyrsine melanophloeosVepris nobilisCroton megalocarpusOsyris lanceolataWarburgia ugandensisDalbergia melanoxylonOlea europaea ssp africana Zanthoxylum chalybeumErythrina abyssinicaPrunus africanaZanthoxylum usambarense39

General abundance of species in surveyed farms, forests and herbalists’ gardens Rank Farms % PropHerbalists‘ gardens% PropForests% Prop 1 Eucalyptus spp 11.3Lantana camara 16.4 Sizygium guinense 8.7 2 Grevillea robusta 10.0 Catha edulis 9.2Mugiru 7.83Catha edulis 9.4Solanum incanum 9.1Mwenyuka 6.44Solanum incanum 9.1Erythrina abyssinica 6.8Mukwethe 2.95Acacia tortilis 6.0Leucaena spp 6.4Mutengerethe 2.96Acacia spp 3.2Grevillea robusta 3.2Aspilia africana 2.77Acacia brevispica 3.2Indigofera lupatana 3.1Gnidia subcordata 2.68Lantana camara 2.9Acacia nilotica 2.9Lantana camara 2.49Tithonia diversifolia 2.9Acacia tortilis 2.6Murieni 2.310Aloe spp 2.6Maytenus senegalensis 2.6Ocimum suave 1.840

30 top species accumulation curves 41 Forest and woodlands Mbeere Embu MeruSmallholder farms Mbeere Embu Meru Herbalist gardens/farms Mbeere Embu Meru More abundance in forests and woodlands in Mbeere than Embu and Meru Herbalists in Embu and Meru plant more – response to scarcity Not much difference in abundance in smallholder farms in the three districts but smallholders generally plant less 20 25 20 201020Species richnessSpecies richnessSpecies richnessSitesSites

30 top species Renyi profiles 42 Forests and districtsHerbalists and districts Farms and districts Combined Forests and distance RP Forests and distance SAC >5 km from village <5 km from village Mbeere Embu MeruMbeere Embu MeruSpecies richness25H-alphaH-alphaH-alphaH-alphaH-alpha

Age and Dbh class comparisons 43 More planting by herbalists in the lower age classes than farmers More lower size classes in farms than forestsBut farmers only 30% of the species were said to be primarily for medicinal use by farmers compared to 66% by herbalists

Is there potential for herbalists and traders to use farm-grown herbal material in future? 44 Cluster analysis based on ecological preferences for herbal medicine raw material sources by herbalists and traders ItemParameterCluster centres 1 2 3 4 General ecological perception Believes ecology affects medicine quality 2 2 2 2Prefers farm (1) or forest (2) source2112Prefers humid (1) or dry (2) source2202Prefers cool (1) or warm (2) source2202Perception of farm as only sourcePrefers isolated (1) or many (2) trees1101Prefers fertile (1) or infertile (2) site 1101Prefers open (1) or shaded (2) sites1101Preference for improved medicinal tree ideotype*Mean score for fast growth rate3432Mean score for resilience with constant harvesting2333Mean score for high chemical composition2234Mean score for high biomass production4121Percent of respondents in clustersHerbalists (n=60)15221743Traders (n=55) 5125222Total32231033Options for ideotype improvement preference:- 1. Least important; 2. A bit important; 3. Important; 4. Most important

Summary findings on species abundance Highly preferred medicinal trees were more abundant and even in herbalist gardens than farms and forests There were more lower age and size classes in herbalist gardens hence diversity may increase in future Herbalists planted more in areas where diversity was less in forestsThere were no specific niches that mimicked forests that herbalists preferred to plant medicinal treesMore herbalists and traders preferred medicinal trees sourced from forests but not necessarily farm niches that mimicked forest situationsUpto 67% of the current traders and herbalists can switch preference to farm grown herbal material if forest trees were not very accessible45

General conclusions Farmers maintain medicinal tree species on farms for household health insurance – only one tree is enough per household The more trees known the more conserved Herbalists not engaged in conservation advocacy and young and educated farmers least informedHerbalists cultivation is increasing as a response to scarcity – good entry to ensure diversity in farmsTrade in medicinal trees’ products is growing and could stimulate cultivation Strategies needed to support nurseries in dry areas as current abundance will disappear fastEmpirical measurements supported survey responses46

Recommendations - actions Involve herbalists and tree nursery operators in extension on medicinal trees informationFurther development of markets and link to farmersPolicy incentives to promote cultivation and discourage wild collection Germplasm conservation and production linking herbalists and nursery operatorsPolicies to develop arid areas as future sources of medicinal tree materialFurther research on influence of various cultivation approaches on medicinal tree active component concentration 47

48 Multi-stakeholder approach needed to collate and share information with farmers on Useful medicinal species for what diseases Markets needs (MIS)High quality germplasm sourcesAppropriate cultivation technologies

Recommendations - top species for domestication priority – matrix ranked Azadirachta indica Aloe spWarburgia ugandensisCaesalpinia volkensiiPrunus africanaZanthoxylum chalybeumStrychnos henningsiiSenna didymobotryaMoringa oleiferaDalbergia melanoxylon49 Leonotis mollissima Croton macrostachyus Croton megalocarpus Olea europaea ssp africanaPsidium guajavaOsyris lanceolataPlectranthus barbatusErythrina abyssinicaRhamnus prinoidesFagaropsis angolensis

50 Senna didymobotrya Azadirachta indica Dalbergia melanoxylon Olea europaea ssp africana Warburgia ugandensis Aloe sp Moringa oleifera Zanthoxylum chalybeum

Acknowledgements Prof. Gerhard Glatzel Prof. Christian Vogl Profs. From IFE : Gratzer, Hager and others ICRAF senior staff: Drs. T. Simons, S. Franzel, R. Jamnadass OEAD ICRAF administration and GRP1 colleagues Fellow students and IFS staff My family (Esther, Grace, Victor) Extended family and friends et al

52 And many thanks to you all for attending and listening‘If many little people, in many little places, do many little things, they can change the face of the earth.’