4.

ETHNOBOTANICAL DATA COLLECTION & ANALYTICAL METHODS

1. Ethnobotanical data collection methods

•In the past, ethnobotanical studies have largely focused on provision of list of medicinal
plants with their corresponding uses (Table 1).

•There was no much attention given to how data should be collected.

•As a result:
•Data were not suitable for quantitative analysis &
•Results were not amenable for comparisons

•But today, are collected in an explicit way so as to make them suitable for quantitative
analysis & hypothesis testing.
 Table1. List of medicinal plants used by the Zay people, Ethiopia

Name Family Local name Part used Disease treated

Achyranthes aspera Amaranthaceae Ambule (Zay) leaves Skin wound

Allium sativum Alliaceae Nech shinkurt (Am.) bulb Chest pain

Croton macrostachyus Euphorbiaceae Bekanisa (Orom.) leaves Retained placenta

Cucumis ficifolius Cucurbitaceae Holoto (Orom.) root Chest pain

Ocimum urticifolium Lamiaceae Chebicha (Orom.) leaves MICH

Ruta chalepensis Rutaceae Tena adam (Am.) leaves GI complaint

Solanum incanum Solanaceae Hidi (Orom) root GI complaint

Vernonia amygdalina Asteraceae Ibicha (Zay) leaves Skin wound

Withania somnifera Solanaceae Kumo (Orom,) root Typhoid

1. Ethnobotanical data collection methods

The following are the commonly used methods to collect ethnobotanical data.
1. Interview
2. Observation
3. Guided field walk
4. Simulation
5. Market survey
6. Focus-group discussion
1.1. Interview techniques

•Interview is the commonly used method to collect ethnobotanical data.

•It requires interaction in the form of talking between two or more persons.

1. Individual interview = 1 Interviewer + 1 informant

- Suitable for quantitative data collection
- Suitable for expressing personal views & to speak freely.

2. Group interview = 1 interviewer + two or more informants [rapid]

•Good interview requires good rapport (trust) with informants.

Based on the degree of control over the process by interviewer, Interview
can be grouped into four types:

- Informal interview
- Unstructured interview
- Semi-structured, and
- Structured
1.1.1. Informal interview

•No framework established (no framed topic)

•Researcher only takes notes after casual conversation.
•It lacks any structure or control
•More useful at early stages of a study where there is little knowledge about the area &
thus not possible to raise specific topics.
1.1.2. Unstructured interview

•Interview goes within established framework

•Get informant on topic with minimum control
•Useful exercise to build initial rapport with informants
•Useful to develop guidelines/information for semi-structured or structured interview
1.1.3. Semi-structured interview

•It is conducted based on an interview guide.

•You have list of topics or questions for discussion (Table 2).

•The guide gives room to entertain any relevant subjects brought up by interviewee.
• Important tool for collection of both quantitative & qualitative data.
- Quantitative – because based on list of selected topics
- Qualitative – because it openly discusses topics not included in the list
[List of items/questions for semi-structured interview]
Table 2. Medicinal Plants of the Bench, Sheko and Meinit People in Ethiopia
1. Local name of the reported medicinal plant
2. Morphological description of the plant
3. Illness treated using with the plant
4. What do you think are the causes of the illness
5. What are the major symptoms of the illness?
6. Plant part/parts of the medicinal plant used
7. How is the plant part collected? (Including the amount collected at a time)
8. Season where the plant is collected
9. Detailed preparation of the remedy
10. Mode of remedy administration
11. Amount of the remedy used (dose)
12. Does the dose differ among males, females and children?
13. Any noticeable side effect?
14. Conditions that forbid taking the medicine (e.g. being a pregnant)
15. Do you store the medicine? If yes, how and for how long?
1.1.4. Structured interview

•It is based on set of fixed questions, often in questionnaire form & needs specific answer.

•No room for further discussion

•Such data is suitable to collect quantitative data & conduct quantification & statistical
analysis.

•It is appropriate to conduct this type of interview at later stage of the study where we know
what specific questions to ask.
1.2. Observation

•Data collected based on researcher’s observation of human-plant interactions.

•Researcher visits homes, market places, crop fields, etc. & takes notes.

•The researches may also raise relevant questions for clarification.

•There are two types of observation

- Passive observation – simple observation

- Participant observation ---researcher participates in activities & rituals

• Its limitation is that It only captures a brief moment in the whole process of interaction.
1.3. Guided field walk/walk in the woods

•It is a combination of observation & interview methods.

•The researcher is guided by the interviewee through areas of interest for observation.

•Every time the interviewee comes across a particular plant, he tries to explain.

•Important to note growing site, habitat & other related species of interest.

•It is, however, tiring, especially for older people who are knowledgeable, & also time
consuming.
1.4. Simulation

•Data is collected by asking participants to re-enact/re-create events/activities no longer

practiced or rarely practiced.

•Valid in demonstrating how plant remedies are prepared and administered.

•Technique is valid as long as participants are able to remember accurately what to do.
1.5. Market survey

•Many markets in cities and towns have

sections where medicinal plants are sold.
These sections are good source of
ethnobotanical information.

•Information from markets may be gathered

through direct observation & interviews.
1.6. Focus-group discussion (FGD)

•FDG is a facilitated discussion of a group of people

consisting of 5-7 participants from similar
backgrounds/experiences on specific topic of the interest.

•FGD is guided by a moderator (facilitator) who introduces

topics for discussion.

•FDG encourages discussions within participants in the

group.
•Questions used in FGD are open-ended ones.

•Length of time of discussion should not normally exceed two hours.

•Useful in developing relevant research hypotheses

•Helps in the formulation of appropriate questions for more structured, larger scale
surveys.

• useful to understand and solve unexpected problems in interventions.

2. Analytical tools used in ethnobotanical data analysis

•The tools are mainly adapted from the fields of anthropology & ecology.

2.1. Free-listing
•Involves asking community members to list any plant used for particular purpose, e.g. plants
used to treat taeniasis (see table next page).

•The assumption is that more significant plants are the ones mentioned earlier in the list.

•By assigning numerical values according to their position in the list, an index of
saliency/prominence is calculated for the plants.
 Free listing analysis of five plants used against taeniasis in Ethiopia
(data from a single informant)

Name of plant remedy Inverted rank/Total listed Salience value

listed by informant

1. Kossao zaf 5/5 1

2. Meterre 4/5 0.8
3. Enkoko 3/5 0.6
4. Yeduba fire 2/5 0.4
5. Qechemo 1/5 0.2
2.2. Preference ranking
•Used to identify most important/useful plants in a given community.
•Involves asking informants individually (up to 10), chosen for exercise, to rank selected
items/species (5-7) & according to a single selected criterion, e.g. degree of potency.

•Most preferred ones are assigned with the highest number depending on number of plants
compared, & least preferred ones with the lowest number (1).

• Numbers are finally summed for all respondents/informants to produce an overall ranking
(Table 3).

•It is convenient when only few items (species) are compared at a time.
 Table 3. Ranking of six Zay medicinal plants based on their degree of local scarcity

Key informants (coded A to J) Total Rank

score

List of medicinal plants A B C D E F G H I J

Acacia sieberiana 3 3 2 2 2 2 1 2 2 4 23 5th

Acokanthera schimperi 5 4 6 6 5 5 6 6 5 5 53 1st

Asparagus africanus 1 1 3 3 1 6 2 5 3 2 27 4th

Carissa edulis 4 6 4 5 4 3 3 3 4 1 37 3rd

Cordia monoica 6 5 5 4 6 4 5 4 6 6 51 2nd

Euclea schimperi 2 2 1 1 3 1 4 1 1 3 19 6th

Taken from Mirutse et al. (2003)

2.3. Direct matrix ranking

•It is used to identify plants of multipurpose use.

•It is a more complex method as it involves asking informant to order given items (5-7) by
considering several attributes at a time (e.g. for medicinal plants: availability, potency,
taste, side effect).

•For each attribute, most preferred item is assigned with the highest number, and the least
preferred one with number ‘1’.

•Total score of several individual responses for each species are summed to get grand
total (S1+S2+S3…) (Table 4)

•Direct matrix can also be done as a group exercise in which participants rank each item
through consensus or vote (but not convenient for statistical treatment).
Table 4. Results of a single person’s direct matrix ranking of four species on seven use
criteria

Name of the tree species

Use (attribute) Eucalyptus Palm Acacia Pine

Fuel wood 4 1 2 3
Building 4 1 2 3
Fruit 1 4 2 3
Medicine 4 1 3 2
Fodder 3 - 4 2
Shade 4 3 1 2
Charcoal 2 - 3 4
Total score 22 10 17 19
Rank 1st 4th 3rd 2nd

Taken from Martin (1995).

2.4. Pair-wise ranking

•Used to evaluate the degree of preference or level of importance of selected plants/parts

of plants (5-10).

•All possible combinations of pairs items (species) are organized & presented to individual
informants in random order.

•Total number of possible pairs is calculated using the formula:

n(n-1)/2, where n= number of items compared

•Informants are then asked to order items in each pair based on a of given criterion.

•Each item in a pair is assigned ‘1’ if more useful, or ‘0’ if less useful.

•Total order is obtained by summing the numbers for each item in a pair (Table 5).
 Table 5. pair-wise comparisons of four useful fruit trees
(banana, lemon, orange and pear)

Bold = more useful (1); Normal = less useful (0)

Sequential Randomized

banana, lemon lemon, orange

banana, orange orange, pear

banana, pear banana, orange

lemon, orange lemon, pear

lemon, pear banana, Lemon

orange, pear banana, pear

Taken from Cotton, 1996

2.5. Triadic comparisons

•It is similar to that of pair-wise ranking except that the number of items to be compared at a
time are three.

•The total number of possible triads are calculated using the formula:

[n(n-1) (n-2)]/6, where n = number of items compared

•Each item in a triad is assigned ‘2’ if most useful, ‘1’ if less useful, or ‘0’ if least useful.

•Total order is obtained by summing the numbers for each item (Table 6).
Table 6. Triadic comparisons of four useful fruit trees
Bold = most favoured (2); Normal = less favoured (1)& italics = least favoured (0)

Triads
Sequential Randomised
banana, lemon, orange banana, lemon, pear
banana, lemon, pear banana, orange, pear
banana, orange, pear lemon, orange, pear
lemon, orange, pear banana, lemon, orange

Taken from Cotton, 1996

2.6. Informant consensus

A. Simple calculation of informants agreement in percept (crude)

•By this method, a relative importance of a plant for any use is calculated directly from the
degree of consensus in informants’ responses.

•The importance of different plants or uses is assessed based on the proportion of

informants who independently reported a given use or who claim to have used a plant in a
specific way (Table 7)

•According to some authors, medicinal plants with higher informant consensus values are
more likely to be biologically active and efficacious.
 Table 7. List of medicinal plants reported as remedies against particular types
of disease by five or more informants in southern Tigray

Scientific name Local name Type of ailment treated No of informant

Achyranthes Aspera mechelo, inflammation of the eye (cattle) 18 (9%)

gurbe
Nicotiana tabacum timbaho leech infestation (cattle) 13 (6.4%)
Impatiens rothii gurshit lungworm pneumonia (cattle) 11 (5.4%)
Aloe sp. ire bloat (cattle) 9 (4.5%)
Calpurnia aurea hitsawts, mange mites (cattle, sheep, goat 7 (3.5%)
cherecnchah and equine)
lice infestation (chickens) 6 (3%)
Ricinus communis gulee external wound infection (cattle) 6 (3%)
Meriandra bengalensis mesaguh lung worm pneumonia (cattle, 5 (2.5%)
goat, sheep)
Rumex nervosus digele strangles 5 (2.5%)
Silene macrosolen saero-saero, strangles 5 (2.5%)
wogert
Senecio myriocephalus tsaeda-qotsli limping, physical damage (cattle) 5 (2.5%)

Taken from Mirutse et al. (2003)

B. Determination of fidelity level (FL) values (refined)
• FL value is determined for every reported medicinal plant, to measure degree of
agreement among informants in its use against a given specific ailment.
• FL is calculated using the formula:
• FL = (Ip /Iu) × 100, where
Ip = the number of citations (informants) who reported the utilization of medicinal plants
against a specific ailment, and
Iu = the total number of citations (informants) who mentioned the same plant against any
ailment.
2.7. Informant consensus factor (FIC)

•FIC measures the degree of agreement among the different informants interviewed
concerning the use of plants species for treating an illness category.

•FIC is calculated using the formula:

Fic = Nur – Nt / Nt – 1, where

nur = the number of use reports (citations) for a particular disease category, and
nt = the number of taxa used for a particular disease category
2.8. Estimation of total plant-use diversity

•Here, the Shinnnon-Wiever diversity index (used to calculated species diversity) was
adopted by Begossi (1996) to calculate plant-use diversity.

•It is used to compare plant use diversity among:

- Different local populations in different environments

- Different social categories (gender, age, education level, religion, …)

• Plant-use diversity (H) is calculated using the following index:

H = - Σpi log pi (to the base e or 10)

where pi = proportion of the number of citations per species out of the total citations
2.9. Cluster analysis

•It is a multivariate technique, useful to partition a set of heterogeneous group of informants

into relatively homogenous sub-groups based on their patterns of responses.

•In ethnobotany, it is used to show similarity/dissimilarity of peoples (informants) based on

their responses to well defined questions, influenced by different factors.

e.g., the similarity in people’s preferences to certain plant species.

•Obtain the basic data matrix (Table 8), &

•Draw the tree or dendrogram using a computer software based on resemblance matrix
(Figure 1).
 Table 8. Basic data matrix for paired comparison of wood species: rank values of six
species used for wood curving obtained from sixteen Kenyan respondents

Sp. Respondents

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 416

S1 1 1 1 2 2 1 2 1 1 1 1 1 1 4 1 2

S2 5 5 5 5 4 5 4 5 5 2 5 5 5 5 5 5

S3 3 3 1 2 3 4 5 4 4 2 4 4 4 1 3 3

S4 2 2 1 1 1 2 1 3 2 5 2 2 2 1 2 1

S5 4 4 1 4 4 3 3 2 3 2 3 3 3 1 3 4

Taken from Höft et al. (1999).

Figure 1. dendrogram based on similarity matrix of sixteen respondents in the paired
comparison of six species used for wood curving (Hoft etal., 1999)
Thank you!