1.1. Collection and preservation of specimens
Biological specimens of different varieties are collected and preserved to gauge the diversity existing in the area. Physical representatives or "types" are used to help track down and, if possible, resolve inconsistencies due to different backgrounds of informants (e.g., migration history, ethnic affiliation, occupational history) or to physical variations of the biological specimens themselves caused by differences in growing conditions (e.g., soil, rainfall, temperature, pests, and pathogens). Although the issue of correspondence between local nomenclature and existing varieties cannot be fully resolved without biochemical assays (see, for example, Quiros et al. 1990) or genetic mapping, the preserved specimens are always helpful as a reference point.
Collection of specimens
1. Walk around the village and talk with farmers about the crop, what varieties exist, where they are found, why people cultivate them, what varieties existed but have disappeared and for what reasons. Assume the role of pupil or apprentice and allow the local people to point out, name, and give samples of varieties being talked about. Some farmers make better informants than others. Be open to opportunities and try not to be rigid.
2. Secure at least four cuttings of each variety, one for drying and preservation and three for cultivation in local germplasm collections and at the experiment station (or stations).
3. In the case of root crops, secure, whenever possible, three roots per variety: two pieces for photographing (one whole and one sliced to show color of flesh), and one for planting.
Preservation of specimens
1. Treat specimens as you would most herbarium specimens. After collection, remove any trace of surface moisture, then press specimens between layers of newspapers and bind tightly in wooden frames (Fig. 2).
2. In the absence of an oven - dry under the hot sun for 4-5 days, changing newspapers daily to prevent mold growth. If an oven is accessible, oven-dry samples at low temperatures (approximately 150-200 degrees F) for 2-3 days.
3. Mount dried specimens on thick white herbarium paper. Label carefully, filling the data sheet completely. At the same time, enter on index cards morphological evaluations of the particular variety (Fig. 3) (Huaman 1988).
1.2. Rapid appraisal/preliminary participant observation.
Rapid (actually not-so-rapid) community appraisal coupled with preliminary participant observation is undertaken during the first month of fieldwork to arrive at a working knowledge of the agricultural system and the different "players" or "actors" involved. Rapid appraisal of the RRA genre achieves the goal of quick information gathering but has been criticized for being "dirty" in that only the most superficial (and at times incorrect) information yields itself to a quick reconnaissance. Preliminary participant observation, a stock-in-trade of anthropologists, involves immersion in the community to understand local ways of doing things, thinking, and being. It has, however, been dismissed by some development workers as too cumbersome, inefficient, and "unscientific." Complementing each other, the two can familiarize the researcher with local patterns of subsistence and exchange, provide numerous opportunities to establish rapport with the community, and help in framing meaningful questions for the benchmark socioeconomic survey and subsequent ethnographic procedures.
Places to visit are local markets, home gardens, settlements of natives and migrants, trading stations, and sites of home enterprise or small-scale food processing. Investigators should come prepared with a loosely structured (ideally, memorized) interview schedule. Some specific topics to explore are the following:
1. Information on diversity
3. Information on cropping systems
This should be attempted only after some time (at least one month) has elapsed from one's entry into the community. Otherwise, one runs the risk of asking totally out-of-context questions that strain the patience of the respondents and provide no meaningful information on the subject.
A benchmark survey is a structured way of finding out about the farming system in the area. It also provides valuable insights into the degree and bases of internal differentiation in the community. The latter is important in deciding how many key informants to choose for the more intensive tasks and what categories (depending on the parameters of internal differentiation) these informants should represent.
1. Depending on intent, the researcher can use close-ended or open-ended questions. The former will be easier to code during data analysis but is categorical, and therefore, restrictive, whereas the latter is more difficult to analyze but yields a lot of unexpected information.
2. The survey should be in the local dialect. If translated from English by a native speaker, the translation should be counterchecked by asking another native bilingual speaker to translate it back into English.
3. The interview schedule for a benchmark survey should be pretested on a limited but comparable sample to anticipate and correct any vague or problematic question.
4. Local assistants or enumerators hired for administering the survey should first be trained in the methodology of the survey and overall goals of the research. Objectives and expected results need not be detailed, however, as these might unconsciously influence the way they ask questions and record answers. The minimum number of enumerators needed to accomplish the task within the time allotted should be employed to assure greater quality control in the data generated.
5. Economy in personnel also forces the researcher to limit the number of questions to shorten the interview, and thus avoid the negative effects of long and tiring questionnaires (refered to as interview fatigue).
