The role of homegardens for in situ conservation of plant biodiversity in Holeta Town, Oromia National Regional State, Ethiopia

This study was conducted to assess the role of homegardens for in situ conservation of plant biodiversity in Holeta Town. In the garden data collection 75 homegardens were randomly selected. Ethnobotanical data were collected using homegarden observation, semi-structured interviews and market survey. Data were analyzed using preference ranking, direct matrix ranking, and descriptive statistics. A total of 112 plant species belonging to 93 genera and 43 families were identified which were classified to 14 functional groups. Out of which, 49 species (43%) were herbs, 32 species (29 %) were trees, 28 species (25 %) were shrubs, and (3%) species were climbers. Further analysis of the results showed that 70 species were cultivated, 35 were wild while 7 species were Ethiopia domesticated. Of the cultivated species, 41.07% were food crops and 58.93% were non-food crops. Family Fabaceae consisted of the highest number of species (11 species), whereas Ensete ventricosum was the most frequently occurring species (93.75%) in the homegardens of the area. Garden was managed by males (47.93%) and females (38.41%). Of the total plant species, 13% were medicinal plants, out of which 33.33% were nutraceutical plants. Direct matrix analysis showed that Juniperus procera was the most important versatile species followed by Cordia africana. These results indicate that homegardens studied play a crucial role in food security of the households and conservation in situ of these plant resources. However, insufficient agricultural support, smallsized gardens and the shifting of polycultural farming to few income generating food crops affect the diversity of species.


INTRODUCTION
Biodiversity is crucial for survival, health and well-being of humans.It is giving greater resilience to ecosystems and organisms (Qualset et al., 1995).Complex, diversified and highly traditional rooted part of plant biodiversity conservation and utilization is found in homegardens (Kumar and Nair, 2006;Zemede, 2004).
Homegardens are variously named in English language as agroforestry homegardens, backyard gardens, farmyard, roof top garden, homestead farms, gardens (Kumar and Nair, 2004).In Ethiopia, a very common Amharic vernacular name equivalent for the term homegarden is "Yeguaro-ersha", in Oromo vernacular language is "eddo" means a land at a backyard of a house (Zemede, 2001).
Homegarden is commonly defined as land use systems involving deliberate management of multipurpose trees and shrubs in intimate association with annual and perennial agricultural crops and invariably, livestock within the compounds of individual houses, the whole tree-cropanimal unit being intensively managed by family labour (Christanty, 1990;Kumar and Nair, 2004).Such systems are essentially man-made and reflect the wisdom of the traditional culture and ecological knowledge that have evolved over the years (Kumar and Nair, 2004;Abebe et al., 2010).
The high diversity of species in home gardens plays wide socioeconomic and ecological roles.because it is related with the production of food and other products such as firewood, fodders, spices, medicinal plants and ornamentals (Christanty, 1985), prevention of environmental deterioration commonly associated with monocultural production systems, income generating sites (Hoogerbuugge and Fresco, 1993) and in situ conservation of agrobiodiversty (FAO 2001;Watson and Eyzaguirre, 2002).Generally, homegardens serve as refuges to a number of plant species, particularly those not widely grown in the larger agroecosystems.Moreover, they are the place of enormous indigenous knowledge (Eyzaguirre and Linares, 2004).
Ethiopia is one of the eight world's centers of origin and diversity of agricultural products.The tremendous variety and complexity of genetic resources results from in situ conservation of plants traditionally grown in homegardens (Zemede, 2001;Kumar and Nair, 2004).However, homegardens are currently under threat of genetic erosion such as, the displacement of great variety of landraces by few high-yielding varieties, loss of traditional knowledge of cropping patterns and management practices, socioeconomic factors and drought (Kumar and Nair, 2004;Zemede, 2004).
In Ethiopia, inventory and documentation of home gardens are very few.It has been concentrated in south and southwestern parts of Ethiopia (Zemede and Zerihun, 1997;Tesfaye, 2005;Talemos Seta, 2007;Abebe et al., 2010).Thus, the study was initiated to increase the knowledge of homegardens of northwestern Ethiopia, specifically in west part of Ethiopia, Holeta Town region.We provided analysis regarding plant species, management practices of local people and their contribution for agrobiodiversity conservation in this region.

Description of the study area
Holeta Town is situated at a distance of 31 km West of Addis Ababa and located at 9°02' N latitude and 38°29' E longitude.The Town has an area of 5550 ha.Holeta Town is found in Oromia National Regional State (ONRS) of Ethiopia.Holeta Town is found at an average altitude of 2449 m a.s.l.

Selection of study sites and informants
During a reconnaissance survey (September, 2010) of the study area, overall information was obtained.Consequently, four study sites were identified and selected based on the presence of traditional homegarden practices and various ethnic communities (languages).They are described as Welayita community site (W.C.S), Gurage community site (G.C.S), Oromo community site (O.C.S) and Mixed community site (M.C.S).
A total of 400 houses (100 for each site) were randomly selected to determine the frequency of homegardens and to identify their local positions (front-yard, back-yard, side-yard or others).To undergo data collection and analysis, 75 homegardens were selected randomly from houses that practice homegardening.A total of 12 knowledgeable key informants were systematically selected from different sites with the assistance of community elders and local developmental agents for ethnobotanical data collection.

Data collection
Ethnobotanical data were collected by using semi structured interviews, field observation, market survey and ranking and scoring methods (Martin, 1995;Cotton, 1996).Interviews and discussions were conducted in Amharic (the local language) using a checklist of topics.
Direct matrix ranking was calculated for five multipurpose tree species in order to assess their relative importance to local people and to generate a matrix that represents the views of the entire community.Based on their relative uses, selected twelve key informants were asked to assign use values for each plant (using the following category 5 = best, 4 = very good, 3 = good, 2 = less used, 1 = least used and 0 = not used).In the end, the results of twelve respondents were summed up to generate a matrix that represents the views of the entire community (Martin, 1995;Cotton, 1996).Preference ranking was calculated for six food crops to determine the relative importance to local people.This technique was employed to rank some selected homegarden species according to their significances.Based on their personal preference of efficacy, selected informants were asked to assign use values for each plant (Highest score (10), while the one with the least effectiveness given the lowest score (1).The numbers are summed for all respondents, giving an overall ranking for the items by the selected group of respondents (Martin, 1995;Cotton, 1996).

Floristic composition data and plant identification
Floristic composition data collection was conducted on 75 sample plots of 10 m × 10 m (100 m 2 ) were delimited in 75 representative homegardens giving a total of 7500 m 2 or 0.75 ha.Then, counts of each species (presence or absence) were conducted on each plot.Specimens of plants found to be the homegardens were collected and local names and habits of each plant were recorded with the help of key informants and development agents.Botanical names were established by comparing specimens with those at the National Herbarium, Science Faculty of Addis Ababa University using available floras.

Data analysis
Ethnobotanical data were analyzed and summarized using descripttive statistical methods (percentages) and floristic composition data were analyzed for species diversity using appropriate equations of the following parameters: Frequency, Frequency classes A-E, Density, Shannon and Wiener index (H'), evenness and species richness, and Sorensen's Index of similarity (Kent and Coker, 1992;Shannon and Wiener, 1949;Whittaker, 1972)

Structure of homegardens in Holeta Town
The survey of 400 houses in Holeta Town indicated that 342 (85.50%) households were practicing homegardening.
The positions of homegardens are varying in type.The same case was reported by Zemede (2002), andTalemos Seta (2007).The size and diversity of species in the study area have positive correlations.With increase in holding size, more variations in species composition were encountered.Similarly, Tesfaye (2005) reported the same result in the study of diversity in homegarden agroforestry system of southern Ethiopia.
Homegardens in the study area are composed of trees, shrubs, herbs, climbing plants and food crops in different strata.They consist of trees approximately 10 to 15 m on the upper strata (Eucalyptus camaldulensis, Erythrina brucei, Cupressus lusitanica).Fruit crops (Prunus persica) and Ensete ventricosum encompassed the middle strata.Herbaceous plants up to 1 m from the ground strata (Brassica carinata, Cymbopogon citrates, Beta vulgaris, Brassica carinata, Brassica oleracea, Daucus carota, Lycopersicon esculentum and Ocimum basilicum).However, the horizontal structure of the species declines as one goes from homegarden to the out fields.This was also reported by Tesfaye (2005) in coffee-enset-based Sidama homegardens and Talemos (2007) enset-based homegardens in Welayita.

Plant diversity and composition of the homegardens of Holeta Town
A total of 112 plant species were identified and documented from the study area (Appendix I).These plant species were classified into 93 genera and 43 families.The commonly represented families were Fabaceae which contains 11 species, followed by Rutaceae and Poaceae in the second rank, which contain 8 species each, and Solanaceae in the third with 7 plant species.
The richest homegarden contained 47 species; whereas, the poorest garden contained 4 species and the mean was 22 species per homegarden.Among the recorded species, only 34 species (26.79%) were found in all study sites and 5 species in only two homegardens.From 112 plants species identified, 6.25% were indigenous plants such as E. ventricosum, Juniperus procera (Appendix I); 35 species were wild plants which grow, and 70 species were cultivated crops.
The homegarden flora is composed of both food and non-food plants, accounting for 41.07% and 58.93% of the total of species respectively.Among the food crops, 19 species (17%) were fruit species, 15 species (13%) were vegetables and 6% pulses & cereals and ranked 1 to 3 in that order.On the other hand, from non-food components of the garden grown species, medicinal plants were 13% and construction & building plants were 12% and miscellaneous consisting of 5% ranked 1 to 3, respectively.
From the total number of species recorded in the study area, E. ventricosum (93.75%) was the most frequent species, followed by E. camaldulensis (90.63%).The species distributed in the frequency classes indicated 6.22% of high frequency values species occurred in higher frequency classes A and B. These classes include E. ventricosum, E. camaldulensis, Justicia shimperiana, C. edulis, Solanum tuberosum and R. prinoides.The remaining species were distributed in frequency classes C, D and E in ascending order containing 13.39%, 20.54% and 59.85% in that order a total of 93.78%, that indicated that more than half of species showed low frequencies.These results indicated that homegardens play a vital role in in situ conservation of agrobiodiversity (Lamprecht, 1989).
The value of Shannon-Wiener diversity index of sites ranges from 3.016 to 3.28 (Table 1).Naturally the Shannon's index value varies from 1.5 to 3.5 and rarely exceeds 4.5 (Kent and Coker, 1992).
Sorenson similarity index of the study area ranged from 0.206 -0.346 or below 0.5 (Table 2), indicating the existence of low similarities/high species diversity among the recognized sites (Table 3).It may be related to the existence of diversified ethnic society in the study area, dissimilar habits for growing plants and preference of food crops as illustrated in (Table 4).

Factors that affect homegarden plant diversity
According to the results of semi-structured interviews, the diversity and productivity of homegardens in the study area were mainly affected by lack of agricultural support/ extension service (81.25%).Disease and pests are the main biological factors of the Welayita and Gurage community sites, which damaged market and nonmarket crops like B. oleracea, Prunus persica, and S. tuberosum.

Food plants in Holeta Town homegardens for nutrition and food security
Out of 112 plant species identified in the study area, 43 species distributed among 36 genera and 17 families were documented as food plants which accounted for 41.07% of the total.Among these, 45.50% were fruits, 30.23% vegetables, 13.95% pulses & cereals, 10.32% tubers and roots and 4% were spices.The fruits are the most usable parts (39.96%) and roots were least usable parts (4.35%) of the food plants.The diversity of food crops of the study area had significant role to increase nutritional and income status of the local people.Consequently, P. persica the most preferred and Allium sativum was the least preferred food crop (Table 4).

Important homegarden tree species with multiple uses
Homegarden owners and other people in Holeta Town have the tradition of using various tree species found in their homegardens for different purposes.The results of 12 key informants using direct matrix ranking in the four study sites showed that tree species have versatile uses (Table 5).The tree species were chosen according to the informant's consensus.Thus, Juniperus procera showed a total score of 239 (79.67%) and ranked first, Cordia africana and Olea europaea with a total of 227 (75.33%) and 219 (73%) second and third positions, respectively.

Homegarden plant species of HoletaTown with medicinal values
The traditions of planting nutraceutical plants and wild plant species in homegardens for medicinal purposes have a vital role to in situ conservation of agrobiodiversity.From a total of plant species (112) identified in the present study, 13% were used as traditional medicinal plants and distributed among 15 genera in 9 families (Appendix II).Among these, 33.33% were nutraceutical plants (Table 6) and 60% were wild plants.The findings obtained were similar to that of Belachew et al. (2003) that reported 133 plant species grown in the 'Gamo' The seeds are ground, mixed with lemon & water & are taken orally to cure constipation & diarrhea / rubbed on the skin to treat skin rash south Ethiopia homegardens of which 18 were medicinal plants.

Indigenous knowledge and homegarden management practices in Holeta Town
For maintenance, the diversity of homegarden species in Holeta Town depends on various indigenous management activities.Belachew et al. (2003) and Talemos (2007) confirmed the same results.
Homegarden owners of Welaiyta and Gurage communities sites developed indigenous knowledge about the preparation, classification and cultivating of E. ventricosum.E. ventricosum is a versatile crop that is used for food, fodder, medicinal and other uses (Table 7).In regions surrounding the study site, Zemede and Zerihun (1997) reported similar perception with regard to the functions of this valuable crop.
One of the best aspects of indigenous knowledge in the study area is work division in managing homegardens.Females managed (38.41%) vegetables, spices, and medicinal plants by planting, weeding, watering and selling (Christianity, 1990).Males participated (47.93%) cultivating cash crop plants and digging, designing, finding seed and seedling.Zemede (2002) remarked that the male family head is often accountable for designing homegarden structure, identifying suitable places for positioning the major crops, monitoring and strongly impacting the structure and direction of homegarden development.Our results confirmed homegardens are useful for the maintenance of good health in developing countries as also reported by UNICEF (1982).This indigenous know-ledge is also important in the development of modern medicines as reported by Dawit et al. (2003) and Fisseha (2007).

CONCLUSION
The results of this study indicated that homegardens in Holeta Town had high species diversity and a rich floristic composition that is worthy of in situ conservation of plant biodiversity, trial sites of new variety of income source vegetables and other species.In addition, homegardens provide significant contributions for the gardener and the society as source of supplementary food, medicinal functions, and income.However, insufficient agricultural support, small sized garden and disease and pests affect the diversity of species.If these challenges receive attention by concerned institutions and researchers, the hotspot will maintain its existing biodiversity and traditional management systems on a sustainable basis in the future.

Table 1 .
Species numbers, Shannon Wiener diversity index and species evenness foreach study site.

Table 2 .
Sorenson similarity index of the homegardens of the study area.

Table 3 .
Frequency of factors that affect diversity of homegardens.

Table 4 .
Simple preferences ranking for widely used food crops in homegardens: 10-for most valuable, 1-for least valuable.

Table 5 .
Results of direct matrix ranking of five homegarden tree species and its six major uses.According to the 12 key informants.

Table 6 .
Medicinal and food (nutraceutical) plant species found in the study area.

Table 7 .
Eight landraces of Ensete ventricosum recorded in Gurage community site.