Participatory rural appraisal investigations on beekeeping in remote Himalayan villages of West Nepal (2002)

Participatory rural appraisal investigations on beekeeping in remote Himalayan villages of West Nepal (2002)
Saville Naomi M. (1), Upadhaya S.N.(2) and Acharya N.P. (3)

1 GPO Box 8975, EPC 1514, Kathmandu, Nepal,
2 ICIMOD, GPO Box 3226, Kathmandu, Nepal,
3 Surya Social Service Society (4S), Chandanath ward no. 6, Jumla, Nepal

Paper for the International Aid session of the 7th IBRA Conference on Tropical bees: management and diversity & 5th Asian Apicultural Association Conference, 19-25 March 2000, Chang Mai, Thailand

1 Introduction
The use of participatory methods in development, particularly Participatory Rural Appraisal (PRA) has become accepted practice in development projects (see Pretty et al 1995, Chambers 1992a and b for descriptions of the tools and discussion of their ideology). In beekeeping development projects though few have taken up the approach. Current beekeeping development practice often puts emphasis on technical issues such as introduction of movable combs, with little reference to social and cultural factors. Consultants making short visits to beekeeping projects may have insufficient time and cultural orientation to address sociological aspects and for learning from indigenous beekeepers. This can turn the development process into a one way channel for the 'dumping' of outsider ideas and technologies on communities that may or may not be ready to accept them. In such cases, inhibition of two way learning and of synthesis of indigenous and outsider / modern ideas leads to failure of projects.

2 Methods
Participatory Action Research (PAR) to address means of conducting appropriate and participatory beekeeping development in very remote areas of the Himalaya, was carried out in Jumla district of Mid-Western Nepal between 1995 and 1999 (Saville 2000a, Saville and Upadhaya 2000, Saville, Upadhaya and Acharya in press). Participant Observation method was used to integrate into the local community, for establishing trust with project beneficiaries and accessing indigenous technical knowledge (Saville 2000b). Having trained local practitioners in PRA (local project and NGO staff and 'farmer-trainers' of the local beekeepers organisation) selected PRA exercises were conducted with Jumla communities. Use of 'farmer-trainers' to conduct PRA exercises was innovative. It allowed PRA to be organised cheaply and easily by people who all speak the same dialect and deal with similar problems and deepened understanding of villagers' situations for extension workers and farmers together. Use of visualisation was especially important as literacy skills are severely limited in the district.

Applications of PRA were used as follows:

Preference ranking of income sources, to determine the relative importance to farmers of beekeeping relative to other income generation or other agricultural tasks.
Preference ranking of hive types to determine the preferred beehive in Jumla.
Venn diagrams to show village leadership in relation to the beekeeping group members or relationships between existing institutions.
Flow charts to show the perceived effects of beekeeping enterprise in the community.
Seasonal calendars of forage, agricultural and beekeeping activities.
Resource maps showing hive baiting sites, forest, bee forage, crops and other resources.
Daily and seasonal work schedules of women and men
Participatory monitoring and evaluation where the extent that trainees felt they had understood different components of the training was evaluated by beekeeping group members, using visualisation of the phases of the moon.

For reasons of brevity, examples of preference ranking of income sources only are explained in detail here. Exercises were conducted by calling a meeting in the village and asking community members to help in drawing up charts. Participants generally enjoyed the exercises and gave up valuable time to take part. Farmers were asked to draw things or represent things on the charts themselves using materials easily available and familiar to them. Wherever possible the PRA practitioners helped to demonstrate how to start only.

Preference ranking exercises were comprised of two parts: pair-wise ranking and matrix ranking. Pair-wise ranking gave the relative preference for one income source over another but failed to show why. The matrix ranking of reasons for preference showed why one income source is preferred over another and which sources had many minor reasons for preference and which one major reason for preference.

First the income sources to be included in the ranking exercises were listed and a matrix table scratched into the ground. Different materials were accumulated by the villagers and agreed upon to represent different crops or activities. For example wool was used to represent sheep, straw to represent wheat, old combs to represent bees and so on. In the first ranking exercise each item was compared with each other on a pair-wise basis, the preferred item being placed in the matrix cell. Where wheat was preferred, the chart had a predominance of wheat straw, whereas where sheep husbandry was preferred, the chart had a predominance of wool, and so on. This resulted in a matrix that was clear for literate and illiterate participants to understand.

PHOTO 2. PRA materials being collected by farmers to make a seasonal calendar.
PHOTO 3. Farmer from Shipti taking part in a PRA preference ranking exercise.
PHOTO 4. Lekhpur farmers with their income source preference ranking chart completed.

After pair-wise ranking, a rank score was calculated for each item. Reasons for each income source receiving this score were listed and then a new matrix table with reasons along the top and items being ranked down the side was drawn. Participants then ranked each item in relation to the reason given, giving it a score out of 10. This allowed the participants and PRA practitioners to analyse why one item was preferred over the other and which reasons turned out to be more important than others were.

3. Results
3.1 Analysis of the benefits of beekeeping relative to other income sources, using PRA preference ranking

In order to demonstrate the kind of results that can be obtained, one example is presented in detail from participants of the villages of Phungra and Bare in a remote part of Jumla district (Figures 1 and 2).

Figure 1. Results of PRA pair-wise preference ranking of income sources with farmers of Bare & Phungra, a remote part of Jumla district, conducted in July 1988.

a) Pair-wise ranking matrix

b) Scores and reasons given for preference or lack of preference from pair-wise ranking

Figure 2. Results of PRA matrix ranking of the reasons for preference of different income sources with farmers of Bare & Phungra, a remote part of Jumla district, conducted in July 1988.

Bare and Phungra farmers ranked bees 6th out of 14 income sources in pair-wise ranking, but 3rd in the matrix ranking of reasons. Multi-purpose uses of bees and their products created an overall higher score when ranking reasons. People of this area make business by seasonal migration to the plains or India to sell local produce. Some operate mules for transporting goods. Grain crops were most important since these were the source of daily food. Sheep ranked second equally with paid jobs though paid jobs are difficult to come by, especially for people living so far from the bazaar (1 full days walk). Abundance of pasture, cash income from manure and meat, and benefits from sheep and goats' wool for making blankets and clothes for use at home and for sale made sheep the most important livestock. Cows ranked equally with business since they provide so many nutritious, sacred and useful products (milk, yoghurt, manure) and bulls for ploughing. Beekeeping was particularly valued for the medicinal and nutritional effects of honey, the fact that anyone can keep bees and it involves relatively little work. Crafts (e.g. woollen blankets and cloth), vegetables, horses, mules, casual labour, buffaloes, chickens and rabbits ranked below bees (in order of importance).

Figure 3.1 Summary of the beekeeping rank scores in comparison with other income sources from nine different villages in Jumla district.

Legend : Pair-wise ranking scores are shown above & matrix ranking scores are shown below in brackets. + Codes for castes: B = Bahun or Brahmin; Ch = Chettri; Th = Thakuri, K = Kami. 1 - listed in order of highest to lowest rank. 2Alt. Range: Altitude indicated in terms of whether the village lies: above- (2600-2700m); much above- ( 2700-3000m); or below- (2300-2600) the rice growing threshold. Comparison of total matrix score with beekeeping's score indicates its importance.

Combined results from 10 similar preference ranking matrices to those shown in Figures 1 and 2 are presented in Figure 3. The relative importance of beekeeping in the different villages varied with altitude and the caste of the people. Although bees are necessary for improved apple production and quality at the rice-growing altitudes of <2600m, insecticide usage on rice and vegetable crops and the lack of uncultivated land with wild bee forage, limit the advantage of beekeeping at these altitudes. Above the rice-growing threshold, barley, potatoes and wheat are most important staple crops, but land is of poor quality and usually on steep and rocky terrain. This limits cultivation and makes sheep and goat grazing more profitable. In these higher altitudes >2600m beekeeping takes on greater importance because of abundant forage and lack of alternative income or food production.

Although beekeeping is a valuable income source for Jumla farmers, it is a sideline income generation activity providing nutritional and medicinal supplements for families, not daily staple food. Only Phungra farmers ranked beekeeping highest, but this result was probably biased by enthusiasm of the community for the beekeeping project. In most of the villages sheep and goats were more profitable than beekeeping and in some cases cows also ranked higher. Buffaloes usually ranked lower than beekeeping, along with vegetables, chickens and rabbits. In villages where people had access to paid jobs (e.g. Gajrangkot) agricultural activities took a lower rank, but in most communities very few or no people from the village were employed in jobs and hence subsistence farming activities ranked higher. In several villages (e.g. Bare, Gajrangkot and Mallapani), business (trading between India and Nepal ) ranked higher than agricultural activities and beekeeping.

In one occupational (so-called 'lower') caste village called Dandakot, labour ranked highest because the occupational castes are skilled in wood working, masonry, jewellery making, metal work and agricultural work (especially ploughing for men). For these people bees ranked near the bottom. However rice and vegetables ranked lower because occupational caste people lack land. In stark contrast, in the neighbouring Brahmin village Simkhada, ('patron' community for the 'client' occupational caste village) bees ranked highly. Beekeeping is customary amongst the Brahmins but not amongst the occupational castes, who previously were forbidden to keep bees because people believed that the bees would abscond if they touched hives.

Even though beekeeping ranked lower than staple foods in the pair-wise ranking exercises, often it scored highest or amongst the highest in matrix ranking of reasons. The reason for the high score for beekeeping is that there are multiple benefits from beekeeping. Benefits perceived by the farmers of Jumla (most of whom had been trained by the project) are summarised below:

provides medicine for the family;
less time consuming than other activities;
cheap;
traditional pastime in which many people are already experienced;
benefits everyone through pollination (improving crop yields and quality);
provides honey needed for religious rites and rites of passage (especially marriage);
provides vitamin-rich foods;
provides cash income;
requires little investment;
suited to the local conditions;
provides tasty good quality food;
can be done by anyone;
doesn't require very specialised skills;
more profitable for the time and money invested than other activities.

3.2 Summary of experiences with applying other PRA tools in Jumla district

Although detailed results of other PRA exercises could not be presented here, these were valuable in different ways as follows:

Preference ranking of different types of hives showed which of the different types of hives being tested by the project together with farmers were seen to be most appropriate by Jumla farmers. Adaptations of traditional log and wall hives were preferred to thin wooden frame hives and straw hives (see Saville, Upadhaya and Acharya in press for details). Once farmers are familiar with different kinds of beekeeping technology available, it is enlightening to conduct preference ranking of the different technologies in order to choose which one to adopt.
Seasonal calendars of agricultural and beekeeping activities allowed the project to time training and other activities to times when farmers were free, or to the traditional timing for that activity. The calendars also accessed indigenous knowledge in beekeeping and agricultural practices.
Daily and seasonal work schedules of women and men were used to decide timing of training and extension activities both on a seasonal and time of day basis. Usually meetings with farmers had to start early in the morning (e.g. 6am) or in the evening before or after eating if people were to be able to give up time. For women it was often difficult to find any time to meet due to their extremely heavy workload.
Forage calendars showed villagers' impressions of the most important bee flowers at different times. This had implications in choosing apiary locations and in terms of planning forage management through plantation of selected species.
Resource maps showing hive baiting sites, forest, bee forage, crops and other resources helped community members and extension workers to assess together what resources might limit beekeeping activities. Whether the community had access to forest for making hives or placing bait hives, whether forage availability might be limiting, and the kind of agricultural production and its potential impact on bees could be assessed. For example areas with emphasis on rice and vegetable production might prove less suitable for bees due to insecticide usage and lack of forage, compared with areas used for maize, mustard, apples, buckwheat or sheep grazing
Flow charts visualised through pictures the potential effects of having an active beekeeping group in the community.

The charts showed:

how income could be gained from honey, beeswax, hives and bee colonies and how such income could be used (eg. on school fees, clothes, making a new house, buying land, etc.);
medicinal applications of beeswax, honey, pollen and bee stings;
uses of honey in religious ceremonies and rites of passage (especially marriage);
use of beeswax (for lost wax casting) for exchange with metal objects;
applications of beeswax to make creams and candles;
pollination benefits in terms of increased yields of apples, mustard, maize, etc.

These charts were very helpful in discussing why it could be advantageous to have a beekeeping group, or to liven up theoretical training on the benefits of beekeeping.

Venn diagrams, to show village leadership in relation to the beekeeping group members, were less useful in Jumla in the face of sensitive political issues. Rather than visualising the roles of individual community leaders, Venn diagrams had more application in analysis of existing institutions or groups in communities and their relationship to one another.

Participatory monitoring and evaluation was extremely useful for assessing which aspects of the training and extension programme had been successful and which aspects needed further follow-up. The exercises formed a useful refresher to training by assisting farmers to recall what they had learnt. All training and extension programmes should adopt such evaluation methods if training is to be assessed fairly from the point of view of trainees.

Discussion
PRA preference ranking exercises comparing beekeeping with other activities enhance understanding of their relative importance for extension workers and farmers together. Those villages that view beekeeping as a worthwhile income source are more likely to succeed in developing beekeeping than those communities where alternatives are valued most highly. PRA exercises generally strengthen participants' capacity to analyse their situation. In training, PRA can be used to demonstrate the multiple benefits of beekeeping in terms of health, nutrition, pollination and traditional uses such as in religious rites, as well as its potential for cash income.

Gender considerations need to be made when conducting PRA exercises (Mosse 1993, Gurung 1995). Due to inequity in workload and decision-making power between women and men in Jumla (see Gurung 1995 and Kharel et al 1996 for supporting data) it was not possible to access women's opinions as easily as men's. Often women were too busy to be able to take part, and in mixed groups they were not free to express themselves. In such situations women PRA practitioners are needed to work separately with groups of women on their own. Had women-only exercises been possible with preference ranking the exercises may have shown quite different results.

Conclusions
Various PRA tools can be employed in the process of implementing beekeeping projects, from feasibility studies at the outset, in training and extension, and in participatory monitoring and evaluation of project activities. Careful analysis of PRA with village people allows deeper understanding of the relative importance of beekeeping in the community and the potential for developing it. Wider practice of PRA tools could substantially improve the quality of beekeeping extension work.

References

Chambers R. 1992.a. Rural appraisal: rapid, relaxed and participatory. Institute of development Studies. Discussion Paper 311.
Chambers, R. 1992.b. Relaxed and Participatory Rural Appraisal on practical approaches and methods. Notes for participants in the workshop held in Chang Mai on Nov. 19-20, 1992: Social Research Institute, Chang Mai, Thailand.
Gurung, J. D. 1995. Participatory approaches to agricultural technology promotion with women in the Hills of Nepal. ICIMOD Discussion Paper Series No. MFS 95/4.
Kharel, B.P.; Thapa, R.B.; van Dijk, N.; Laats, H.; Basnyat, B.B. 1996. Learning from people. Natural Resource Management Systems in Humla and Kalikot districts. Nepalgunj: Karnali Local Development Programme SNV- Nepal. 91 pp.
Mosse, D. 1993. Authority, gender and knowledge: theoretical reflections on the practice of Participatory Rural Appraisal. ODI Agricultural Administration Network Paper 44. 28 pp.
Pretty J.N, Gujit I., Thompson J. and Scoones I. 1995. Participatory Learning and Action - A trainer's guide. IIED. London. 267 pp.
Saville, Naomi M., 2000a. "Farmer-participatory extension in Jumla, Western Nepal." In Asian Bees and Beekeeping: Progress of Research and Development. (Eds: Matsuka, M. et al.) pp. 230-236. New Delhi: Oxford and IBH.
Saville, Naomi M., 2000b. Getting a buzz out of participation. Participation Vol. 1. (1) pp. 3-8. (NEPAN, Kathmandu, Nepal).
Saville, Naomi M.; and Upadhaya, S.N., 2000 "Indigenous knowledge of beekeeping in Jumla, Western Nepal." In Asian Bees and Beekeeping: Progress of Research and Development. (Eds: Matsuka, M. et al.) pp. 248-251. New Delhi: Oxford and IBH.
Saville, Naomi M.; Upadhaya, S.N.; and Acharya, N.P. (in press) Farmer-Participatory Action Research into Beehive Design in the Remote Area of Jumla, Nepal. ICIMOD Discussion Paper Series.

Acknowledgements
The authors would like to thank the following for their assistance with this work: all the farmer trainers of HIBA and 4S local NGOs in Jumla, Karna Bir Sunar and Kaali Bahadur Thapa for assistance with data collection and project implementation; Jumla farmers who took part in PRA exercises and participatory action research; Karnali Technical School for providing office and apiary locations; ICIMOD for project management; Austroprojekt and the Government of Austria for funding the project and for employing the 1st author from October 1996 to October 1999; DFID for employing the 1st author from April 1995 to October 1996.