Keyhole Garden a la Tica: Organic and SustainableSara Arias, Emily Bissett, Constanza Carney, Lillie Dao, Alejandro Garcia, Zuhra MalikGlobalization and Community Health Field SchoolNational Science Foundation REU ProgramUniversity of South FloridaMonteverde InstituteMonteverde, Costa RicaMay 29 – August 3, 2013

TABLE OF cussion19Conclusion22Recommendations23Limitations of the Study24Acknowledgements25Research Team Biographies25Bibliography26Appendix I: Interview Questions, Focus Group Topics, and Surveys27Appendix II: Tables35Appendix III: Figures40Appendix IV: Plant Layout45Keywords: Food Insecurity, Gardening, Agriculture, Monteverde Costa Rica2

ABSTRACTThe economy of the Monteverde zone in Costa Rica has seen a significant shiftfrom an agricultural base to a focus on ecotourism. As a result, research shows that manyfamily members in the region, especially those involved in tourism, experience varyinglevels of food insecurity (Himmelgreen, 2006). Researchers from the Community Healthand Globalization Field School, a Research Experience for Undergraduates funded by theNational Science Foundation and organized through the University of South Florida,explored agricultural and gardening practices in the Monteverde region of Costa Ricathrough anthropological methods that focused on community input and involvement ( listing and pile sorting, interviews, surveys, and focus groups). Through thecombination of these findings and quantitative soil measurements, adaptations were madeto the keyhole garden technology. This low-cost, highly productive gardening methodwas originally implemented in an arid region of sub-Saharan Africa. Unlike this originallocation, Monteverde has an extremely wet climate, seeing an average of 118 inches ofrainfall per year. As a result, the key variable in the two experimental demonstrationgardens constructed on the Monteverde Institute grounds is the drainage layer.Researchers also altered the specific garden layers based off of recommendations madeby community members regarding locally available materials. Additionally, through theguidance of local agriculture experts, a roof was implemented above each of the gardens.Finally, researchers surveyed the public to determine the most desired plants to grow inthe region, adapted these results through the suggestions of community advisors, andplanted a variety of vegetables in the two gardens. These plants were organized in a rootsand shoots method in order to maximize garden productivity. The keyhole gardens are inan experimental phase and will be monitored throughout the year by Monteverde Institutestaff, so that the technology may be further adapted to the Monteverde climate andcommunity needs. Although the keyhole garden technology was adapted to improvefood security, it also provides many other benefits including educational and mentalhealth purposes. The research team hopes to partner with other community organizationsin the future, particularly schools and programs for the disabled.3

INTRODUCTIONData from a three-year research program funded by the National Science Foundation(NSF) suggest that food insecurity is a problem in the Monteverde zone of Costa Rica, anarea undergoing a shift from an agricultural economy to ecotourism. Researchers from theCommunity Health and Globalization Field School have explored ways to minimize issuesrelated to food insecurity in the Monteverde region through community involvement andempowerment. The NSF funded faculty from the Anthropology and Civil EngineeringDepartments at the University of South Florida investigated the relationship betweengardening and the issues related to food insecurity. The agricultural and gardening practicesof four communities in the Monteverde region of Costa Rica (see Figure 1 for communitymap) were explored through anthropological methods that focused on community input andinvolvement (e.g. free listing and pile sorting, interviews, surveys, and focus groups). Thesefindings led to the adaptation of the keyhole garden technology, originally developed in subSaharan Africa, to the climate and community needs of the Monteverde zone. The primarygoal for this project was to aid in the alleviation of issues surrounding food insecurity byapplying a community-based method to provide a reliable and accessible source of fresh andorganic produce.Figure 1. Map of Monteverde zone Tourist- Map.gif4

ClimateMonteverde topography offers an array of individual microclimates whichinclude: heavy rain, high winds, and a foggy atmosphere. Resting roughly at 1,400 meters(4,600 feet) above sea level, Monteverde is misty, humid, and windy, with a mean annualtemperature of 18 C (64 F) (Nadkarni, 2000). Annual rainfall averages approximately118 inches (3,000 mm). Humidity oscillates between 74% and 97% (Nadkarni, 2000).The climate in this region is cooler than in the lowlands and the area includes theMonteverde Cloud Forest Preserve, one of the world’s most threatened ecosystems(Nadkarni, 2000).Figure 2. Average seasonal Monteverde rainfall and Food Insecurity and TourismDue to the development of protected rainforests such as the Monteverde CloudForest Preserve, the Children’s Eternal Rainforest, and the Santa Elena RainforestPreserve, tourism has developed into an important part of the economy in the Monteverderegion. Tourism provides a large variety of jobs to men and women of different ages, thusincreasing the sources of income during a great part of the year (Himmelgreen, 2006). Asa result, tourism brings a mixture of both positive and negative effects to local residentsof the Monteverde zone. Himmelgreen conducted a three-year NSF funded project thatinvestigated food insecurity. The findings revealed that that the majority of families5

living in the Monteverde zone that were involved in tourism exhibit some degree of foodinsecurity. “Food insecurity exists whenever the availability of nutritional adequate andsafe foods or the ability to acquire acceptable foods in socially acceptable ways is limitedor uncertain” (Anderson, 1990). Studies conducted on 200 local residents in theMonteverde zone of Costa Rica show that more than half of the households in the sample(50.76%) exhibit some form of food insecurity. In a sample of 193 households, 132 or68.39% work in tourism (Himmelgreen, 2006). Results show that there seems to be acorrelation between those involved in the tourism industry and subsequent foodinsecurity.Agriculture and Nutrition in the ZoneAgriculture has long been the main source of income and sustenance for CostaRicans in the Monteverde Zone. However, in recent years involvement in tourism-relatedjobs has led to a decrease in food production in the community. Participation in familygardens or small farms, an activity that could buffer against food insecurity, has declinedat the local level (Himmelgreen, 2006). As tourism continues to thrive in the region thepopulation seems to be moving away from self-sustainable food production to anoverreliance on the tourism industry. Moreover, food inadequacy has taken a toll on thepopulation in terms of variety and quality. “As part of this change, the face of foodinsecurity in such settings has changed from food scarcity (i.e., limited amounts of foodavailable to individuals and communities) to food inadequacy in terms of variety andquality” (Himmelgreen, 2006). Wilkins, a local resident of the Monteverde region,highlighted that many residents who work in the ecotourism business often feel as thoughthey lack the time required to prepare meals, thus affecting their nutritional choices(Allen, 2012). In other words, they might choose a bag of potato chips to eat rather thanspend the time and energy to cook a typical and nutritious meal. Thus, while populationsmay indeed have more access to food in quantity, the variety and overall quality of thosefood items might be severely compromised. As a result of this shift, there is a trendtowards overconsumption of food high in fats and carbohydrates concurrent with thedecreased consumption of more nutritionally rich foods such as fruits and vegetables.These patterns manifest themselves in rising rates of obesity and BMI disproportion-6

factors that increase the individual’s risk for chronic health problems such ashypertension, diabetes, and cardiovascular disease (Himmelgreen, 2006).What is a keyhole garden?The keyhole garden is a compact technology adapted from sub-Saharan Africa tothe climate of the Monteverde zone. Each keyhole garden contains a basket in the centerfor organic kitchen waste to feed plants and replenish the soil's nutrients (KeyholeGarden Manual, 2009). The combination of the center basket and the entrance-way givesthe garden a keyhole shape when viewed from above (Figure 3). The garden uses anumber of layers to nourish the soil, making it more productive than a conventionalgarden. The garden is constructed from low-cost, recycled, reusable and local materials.The garden is two meters in diameter and one meter tall to allow for greater accessibilityand to limit pests. It is used to grow a variety of fruits and vegetables in the middle and issurrounded by medicinal/companion plants to repel pests.Figure 3. Overhead and side views of the basic keyhole garden model keyhole comp.jpgResearch Questions What current knowledge do community members have with regard to compostgardening? What motivations are needed to facilitate community involvement in compostgardening?7

What adaptations are necessary to implement the keyhole garden technology intothe Monteverde region and community?METHODSThe methods chosen for this study are interdisciplinary in nature and have mergedtogether techniques from the fields of anthropology, engineering, and public health. Thisexploratory study combined both qualitative and quantitative methods in order to attain aholistic understanding of the investigated variables and their relationship with each other.The methods implemented in this investigation are based on the operationalization of thefollowing variables: (1) generational knowledge on agriculture, (2) produce access andattainment, (3) current gardening practices, and (4) the level of interest in gardening. Themethods employed to operationalize these variables include participant observation, freelisting, pile sorting, focus groups, open-ended interviews, structured interviews, groundedtheory analyses, statistical analysis and a soil pH test.During the investigation in Monteverde, Costa Rica, all six researchers resided insix different homestays and attended daily classes and events at the Monteverde Institute(MVI). This allowed the researchers to conduct participant observation with the familiesof their homestays and the Monteverde Institute staff. Participant observation is a datacollection technique that requires the researcher to be present at, involved in, andrecording the routine daily activities with community members in the field setting. Thehomestays were all located in neighborhoods within the Santa Elena region ofMonteverde. Participant observation activities included: (1) observing what individualscomposted and (2) different forms of gardening within households and at the MVI.After obtaining informed consent, free lists and pile sorts (n 12) wereadministered with individuals that lived in the Monteverde region. Participants wereasked to free list the reasons why they would want a personal garden (see Appendix II:Table 4 for categories of reasons for having a personal garden). They were then asked torank the listed reasons on a scale of one to four on their level of importance (seeAppendix II: Table 4 for the value labels of each rank). Following the free list and pilesort activities, two focus groups were conducted: one was conducted in San Luis (n 6)8

and the second was conducted in Monteverde (n 8) (see Appendix I:2 for major topicsprovided by the focus group moderators). Open-ended interviews (n 3) were thenadministered with individuals who work in farming and agriculture in the Monteverderegion (see Appendix 1:1 for the interview questions). The nature of open-ended andunstructured interviews permits for an open exchange between the researcher and theparticipant in the study allowing for the researcher to build relations and rapport withcommunity members. This form of interviewing allows for the researcher to exploreareas, cultural domains, and topics of interest in great depth without presupposing anyspecific responses or conclusions (Benard, 2011). Additionally, oral questionnaires(n 53) were conducted at the 2013 San Luis Educational and Recreational Fair andsupermarkets and farmer’s markets located in Monteverde. Participants were asked torespond to as nearly identical a set of stimuli as possible (see Appendix I: 3 andAppendix I:4 for the full questionnaires). An interview schedule was employed to ensurereliability, validity and consistency by setting an explicit set of instructions tointerviewers who administer questionnaires orally. Monteverde Institute staff and the sixsprimary researchers recruited participants for this study by contacting Monteverde andSan Luis community members via phone, email, and in-person conversations. Finally, apH test was done to compare the acidity or alkalinity of compost soil that included acidicfood products and compost soil that did not contain acidic products.Data Analysis MethodsGrounded-theory analyses were implemented to record and manage the qualitativetextual data sets: interviews, focus group discussions, and open-ended responses derivedfrom the oral questionnaire. Grounded-theory is a set of systematic techniques fordiscovering pattern in human experiences that utilizes close, inductive examination ofunique cases combined with the application of deductive reasoning. The aim is to discovertheories- causal explanations involving the investigated variables- grounded in empiricaldata (Benard, 2011). The grounded-theory method relies on coding texts for themes thenanalyzing the themes for data. The texts were coded for different themes based on theoperationalized variables of the exploratory research: (1) existing knowledge on agriculture,(2) produce access and attainment, (3) current gardening practices, and (4) the level of9

interest in gardening. Data were also analyzed using SPSS statistical analysis and softwarepackage. A Chi-Square test was run with two variables: (1) The frequency that one getsproduce from the farm and (2) age categories.RESULTSCommunity Interest in GardeningIn order to adapt the keyhole gardening method to the Monteverde zone,researchers sought to understand community members’ interests in gardening andpotential challenges associated with gardening practices. Throughout theseconversations, several key themes emerged from qualitative data analysis. The initialinvestigation led to evident reasons for garden participation through the use of the freelisting and pile sorting method. The five most common answers consisted of thefollowing: produce quality (23.73%), food access (15.25%), selling (10.17%), therapeutic(8.47%), and saving money (8.47%) (see Figure 4 for the categories of reasons for havinga personal garden). Although “produce quality” and “food access” were repeated mostoften by participants, in individual pile sorting, these responses received the lowestrankings of importance (see Appendix III: Figure 2 for the value labels of each rank).One participant listed both not having to purchase produce and not having to travel to thesupermarket as reasons to garden, but ranked them on opposite ends of the spectrum. Sheexplained that she “can go to the grocery store easily, but not having to buy produce isvery important.”Figure 4. Free list/ pile sorting of Monteverde locals (n 12): reasons why a resident ofthe zone would have a personal gardenConvenience5%SocialAesthetic 5%Passion ucational5%Medicinal7%Save Money9% Therapeutic9%Food Access15%Selling10%10

Additionally, the oral questionnaires conducted at the 2013 San Luis Educationaland Recreational Fair offered insight into individuals’ levels of interest in theconstruction of a keyhole garden. When asked how interested one would be in assistingwith the construction of a community keyhole garden, most participants (82.14%)indicated that they were either interested or very interested. Similarly, the majority of therespondents (89.29%) stated that they were interested or very interested in constructing ahome keyhole garden. However, when asked how interested one might be in assistingwith financing a community keyhole garden, fewer participants (65.52%) stated that theywere interested or very interested. This variation highlights an interest in gardening, buta hesitation in financial aspects required of a keyhole garden.Both the MVI focus group and the oral questionnaires (those administered toparticipants in the Educational and Recreational Fair and consumers at the local farmer’smarket and supermarkets in Santa Elena) provided community members with theopportunity to share perceived benefits, disadvantages and personal impediments to theconstruction of a home keyhole garden. Qualitative coding revealed that the threegreatest benefits concerned the environmentally-friendly focus, functional structure, andself-sufficiency inherent in growing one’s own produce (see Figure 5 for keyhole gardenbenefit frequencies). Participants who highlighted environmental benefits specificallyfocused on both the use of excess organic material in the form of compost and theproduction of organic crops. One focus group participant stated that the compost basketin the center of the keyhole garden “allows for the reuse of the nutrients, such as theorganic peels, so nothing is wasted.” These benefits may imply motivations key tocommunity participation in gardening as a form of improving food security.11

Figure 5. Perceived benefits of a keyhole garden identified by Monteverde residents insupermarket, farmer’s market, and San Luis Educational and Recreational fair surveysand focus group discussion (n 61)181614121086420The majority of the perceived disadvantages were derived from questionsregarding the structural integrity of the keyhole garden against environmental factorsspecific to Monteverde. Community input also indicated other key concerns related tothe compost and the pests that it would attract, general maintenance, and a lack ofknowledge on how to build the specific garden (see Figure 6 for keyhole gardendisadvantage frequencies). General maintenance apprehensions centered on questions oftime and required upkeep needs such as weeding and watering the garden. Additionally,focus group participants discussed a perceived lack of productivity. One member stated,“If I want to produce a large amount, [the garden] seems very small.” In adapting thekeyhole garden for the Monteverde region, the changes were made based on theseperceived disadvantages.12

Figure 6. Perceived disadvantages of keyhole gardens expressed by MV residents insupermarket, farmer’s market, and San Luis Educational and Recreational Fair surveysand focus group discussion (n 61)*6543210*Responses such as "None" removed to diminish skewed results but included in Appendix II: Table 6The individual impediments to building a keyhole garden remained minimal, withmost answers (32.26%) being “none.” Upon removing “none” as an answer, the top threeimpediments were: the physical effort required in building the garden (28.57%), a lack ofspecific keyhole garden knowledge (23.81%), and a lack of time (23.81%) (see AppendixIII: Figure 8 for keyhole garden impediment frequencies). Once the experimentalkeyhole gardens are tested against the environmental factors in the zone, its productivitytested, and a building manual made, this may assuage some doubts from the communityon the keyhole gardens effectiveness.Generational Knowledge of AgricultureThe passing down of agricultural knowledge from older generations to youngercommunity members has been identified in past research as a changing variable and hasbecome a relevant topic to our investigative study in Monteverde. This was noted in bothquantitative and qualitative data. During the San Luis focus group, one participant shareda personal experience in which he has viewed the slow disappearance of the art form ofsugar cane processing due to the time consuming nature, a lack of education in thetraditional manner, and the apparent preference for the use of new technologies.Additionally, all five elder participants (ages 55-70) of the free listing and pile sorting13

activity designated a passion for agriculture or a method to educate youth as a veryimportant reason to garden. Focus group participants from younger age groups did notindicate education on and a passion for agriculture as reasons to garden.Two quantitative SPSS statistical analyses of the oral questionnaire datadetermined moderate correlations between age and the frequency by which one obtainshis or her produce from a farm. This correlation highlights the change in farmparticipation between different generations. The Pearson Correlation test revealed a twotailed correlation with a significance value of 0.033 and an r2 value of 0.397 whichindicates weak, but present correlation between the two variables. The Chi-Square testalso revealed that these variables are not independent of one another. Questionnaireparticipants within the age ranges of 18-50 years were least likely to obtain their producefrom farms while people greater than 50 years of age were more likely to obtain theirproduce from farms (see Appendix II: Table 9 for Pearson Correlation and Chi-Squareanalyses). Based off of these findings, keyhole garden education materials andintervention targets may need to be tailored to a zone that is experiencing a transitionaway from generational agricultural knowledge.Produce Access and AttainmentResearch was aimed at examining local interest in specific fruit and vegetableproduce. Investigation methods implemented the use of oral questionnaires, focus groups,and open-ended interviews. Employing these methods enabled researchers to work withcommunity members to better understand which types of produce are consumed andthose that are desirable to plant in a home garden, as well as local experience of plantcultivation in Monteverde and San Luis. Participant perception and concern for producewas significant in the determination of plants that would be both feasible to cultivate andgreatly utilized by the local community.Survey data generated the top ten plants that local participants would choose togrow in a home garden. Questionnaire responses, collected from the San LuisEducational and Recreational Fair, indicated that the following plants were mostfrequently desired for cultivation. In order of most to least frequent, participants chose: 1)14

lettuce, 2) tomato, 3) cilantro, 4) sweet chili, 5) cabbage, 6) beans, 7) corn, 8) orange, 9)yucca, and 10) ayote (see Figure 7 for the top ten plants most desired to grow).PercentageFigure 7. Top ten products desired to cultivate as determined by individuals at San LuisEducational and Recreational Fair (n 31), vegetable, or herbAdditionally, oral questionnaires were implemented to review the top 10 produceconsumed by the sample population. Questionnaires were conducted at two supermarkets,the Farmer’s market, and the San Luis Educational and Recreational fair. Responses from53 participants, determined that our sample population consumed the following produce,from most to least frequent. 1) tomato, 2) lettuce, 3) carrots, 4) potato, 5) mango, 6)onion, 7) papaya, 8) chayote, 9) cilantro and 10)pineapple (see Appendix III: Figure 4 forthe top ten plants most consumed).Frequency comparison between produce consumption and participant’s desire tocultivate certain plants were functional in the understanding of production capability anddietary interests. Tomato emerged as the most highly consumed vegetable overall and thesecond most highly desired to cultivate. Lettuce was a plant that more than half ofparticipants would like to grow; our sample population also frequently consumed it. Adifferent trend was identified for produce such as papaya, mango, pineapple, and potato.Levels of consumption and desire for cultivation offered contrasting frequencies. 24% ofrespondents listed papaya, for instance, as one of the top 5 produce that they consumed,15

but only 3.4% wished to grow papaya. Contrastingly, although 17.2% of respondents atthe San Luis Educational and Recreational Fair listed beans as a plant they would like togrow, only 3.4% indicated beans to be one of the top 5 produce they consume each week.During the investigation of local produce interest and attainment qualitative datawere acquired via through focus groups and interviews. Reoccurring themes in ourresearch study included feasibility and scale of specific vegetable and fruit cultivation.Although questionnaire data indicated a high desire for tomato, interview data presentedreoccurring concern for the cultivation of tomato and sweet chili. When the topic ofspecific plant cultivation was addressed, three out of four interviewees discussedenvironment specific complications related to tomato cultivation. Concerns for tomatoproduction primarily pertained to the high volume of water that the Monteverde zoneexperiences. These participants all discussed excess moisture, plant plagues, and resultingdifficulty of plant cultivation. Celery, carrots, beans, green beans, and cilantro werestated in interviews and focus groups to be highly consumed and successfully grown withthe Monteverde region.Based on these research results, lettuce, celery, cilantro, green beans, onion,carrots, round zucchini, radish, oregano, rosemary and rue will be cultivated in theMonteverde Institute demonstration keyhole gardens. These plants were organized in thegarden to highlight permaculture practices in companion planting as well as a method ofalternating root and shoot vegetables to maximize productivity (see Appendix IV for theplanting layout).Local Gardening Knowledge and PracticesInterviews with agricultural experts in the area, as well as focus groups fromMonteverde and San Luis, revealed current gardening practices regarding pest control,environmental factors (rain/wind), and composting processes common to the region (seeAppendix II: Table 3 for the major themes highlighted in interviews and focus groups).With regard to different forms of pest control, both interviews and focus groupsmentioned a chili and garlic spray mixture adequate for small gardens. The use ofmedicinal plants, such as rue, rosemary and oregano, were also mentioned as a deterrent16

against pests. These three plants (rue, rosemary and oregano) will be implemented aroundthe perimeter of the two keyhole gardens, while the chili/ garlic recipe will be madeavailable to the MVI staff for future maintenance of garden. In the Monteverde focusgroup, specifically, concerns were raised over the pests that may be attracted to thecompost basket within the keyhole garden, such as birds and vermin. One suggestion wasto implement a cover and a fine mesh screen around the exposed compost baskets.Currently, the compost baskets are covered by recycled sacks, which will be replaced byfine mesh screens.Although previous research indicated significant environmental factors relevantto the Monteverde zone, specific issues such heavy rainfall, winds, and condensationfrom fog (especially during the rainy season of October-December) were highlightedduring our interviews and focus groups. Some suggestions offered from interviews andSan Luis focus group discussions was to implement a roof, either from lamina or saranand the use of light bulbs to increase the amount of heat to plants. One participant fromthe Monteverde focus group suggested using a greenhouse white plastic to enclose theentire garden, in order to retain heat during the winter months and provide protectionagainst condensation and winds. For the keyhole garden, both the saran and thegreenhouse clear plastic were implemented, with saran on the sides and plastic on the topto reduce the amount of rainwater in the garden and protect against wind and fog, whilestill allowing sufficient Ultra-Violet light.The significant rainfall added an additional element to the adaptation of thekeyhole garden. In the original design of the African keyhole garden water filtration anddrainage was not an issue due to the arid climate of sub-Saharan Africa. However,Monteverde poses the opposite problem in that, during the rainy season especially; thereis an abundance of rainfall. For this reason, the research team determined that theexperimental variable between the two demonstration keyhole gardens would be thedrainage layer. The first drainage mechanism involves inverted glass bottles that weredonated by the Monteverde Institute (Image 1). These bottles were surrounded by dirt andapproximately one finger width was left between bottles to allow for filtration. The dirtpadding also acts as a pseudo-sponge so that, in the event of an earthquake, the bottles17

will not shatter. The second drainage system incorporated small rocks that were leftoverfrom the exterior construction (Image 1). These rocks of varying shapes and sizes wereplaced with small distances between to allow for filtration. Over the coming year, plantgrowth and water filtration will aid in determining which drainage system functions best:the uniform bottles or the irregular rocks.Image 1. (left) Drainage mechanism using inverted glass bottles; (right) Drainagemechanism using irregular rocks of varying

Monteverde zone of Costa Rica show that more than half of the households in the sample (50.76%) exhibit some form of food insecurity. In a sample of 193 households, 132 or 68.39% work in tourism (Himmelgreen, 2006). Results show that there seems to be a correlation between those involve