What works in the field? A comparison of different interviewing methods in ethnobotany with special reference to the use of photographs
ABSTRACT Ethnobotanists use a variety of interview techniques to collect ethnobotanical data. Drawing upon the results from a quantitative
ethnobotanical study in five Yuracaré and Trinitario communities in the Bolivian Amazon, the pros and cons of the following
methods are evaluated: (1) interviews in situ during transects, walk-in-the-woods, and homegarden sampling; and (2) interviews ex situ with fresh plant material, voucher specimens, or plant photographs as reference tools. Although the systematic use of plant
photographs for ethnobotanical interviews is poorly documented in literature, the results show that indigenoùs participants
in our study recognize significantly more plant species from photographs than from voucher specimens. It is argued that, especially
in remote and isolated study sites, photographs might be advantageous over voucher specimens.
- SourceAvailable from: Ulysses Paulino de Albuquerque[Show abstract] [Hide abstract]
ABSTRACT: This study compares the efficiency of identifying the plants in an area of semi-arid Northeast Brazil by methods that a) access the local knowledge used in ethnobotanical studies using semi-structured interviews conducted within the entire community, an inventory interview conducted with two participants using the previously collected vegetation inventory, and a participatory workshop presenting exsiccates and photographs to 32 people and b) inventory the vegetation (phytosociology) in locations with different histories of disturbance using rectangular plots and quadrant points.Journal of Ethnobiology and Ethnomedicine 06/2014; 10(1):48. · 2.42 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: We here tease apart the ethnopharmacological knowledge of plants in two Thai villages to determine to which degree the uses are particular to individual ethnic groups and to which degree they are part of a generalized and uniform set of widespread medicinal plants used over a large geographic range. We compared Karen and Lawa knowledge of medicinal plants in the Mae Cheam watershed of northern Thailand, where both ethnic groups have settled and share ecological conditions for resource extraction. We were interested in documenting the degree to which these two ethnic groups use the same or different medicinal plant species. The use of the same plant species by the two groups was considered a sign of uniform and cross-cultural local knowledge, whereas the use of different medicinal plants by each group was considered a sign of culturally specific local knowledge that developed within each ethnic group. We inventoried the plant species in different habitats around one Karen village and one Lawa village using stratified vegetation plots and using semi-structured questionnaires we interviewed 67 key informants regarding their use of plants for medicine. We then calculated the Fidelity Level FL (FL values near 100% for a species indicate that almost all use reports refer to the same way of using the species, whereas low FL values indicate that a species is used for many different purposes) and Cultural Importance Index CI (the sum of the proportion of informants that mention each of the use categories for a given species) to estimate the variation in medicinal plant use. We used Jaccard's Index JI (This index relates the number of shared species to the total number of species) to analyze the similarity of medicinal plant use between the two villages. A total of 103 species of medicinal plant species in 87 genera and 41 families were identified and they were used to cure 35 ailments. The FL of the medicinal plant species varied from 10-100%, was different for each ailment, and differed between the two ethnic groups. The most important medicinal plant species, those with the highest CI value, were not the same in the two villages. Costus speciosus, which is used to treat urinary infections and wounds in animals, had the highest CI value in the Karen village, whereas Sambucus javanica, which is used to treat wounds, fractures, bloat, and edema in humans, had the highest CI value in the Lawa village. Only 17 medicinal species (16.5%) were shared between the two villages. Methods of preparation and application were significantly different between the two villages, whereas the plant parts used, habit, and route of administration were similar. Our study demonstrates that ethnic groups that live in the same geographic area can have significantly different traditional knowledge systems for medicinal plants, at least when it comes to the species used and their preparation and medicinal application. We assume that differences in cultural history and background in the two villages led to differences in medicinal plant use, preparation, and application.Journal of ethnopharmacology 11/2013; · 2.32 Impact Factor
What Works in the Field? A Comparison of Different
Interviewing Methods in Ethnobotany with Special
Reference to the Use of Photographs'
EVERT THOMAS*'2, INA VANDEBROEK3, AND PATRICK VAN DAMME2
Z Laboratory of Tropical and Subtropical Agriculture and Ethnobotany, Ghent University, Coupure
links 653, B-9000 Ghent, Belgium
3 Institute of Economic Botany, The New York Botanical Garden, Bronx River Parkway at Fordham
Road, Bronx, NY 10458, USA*Corresponding author: firstname.lastname@example.org
WHAT WORKS IN THE FIELD? A COMPARISON OF DIFFERENT INTERVIEWING METHODS IN ETHNOBOTANY
WITH SPECIAL REFERENCE TO THE USE OF PHOTOGRAPHS. Ethnobotanists use a variety of interview
techniques to collect ethnobotanical data. Drawing upon the results from a quantitative
ethnobotanical study in five Yuracare and Trinitario communities in the Bolivian Amazon,
the pros and cons of the following methods are evaluated: (1) interviews in situ during
transects, walk-in-the-woods, and homegarden sampling; and (2) interviews ex situ with
fresh plant material, voucher specimens, or plant photographs as reference tools. Although
the systematic use of plant photographs for ethnobotanical interviews is poorly documented
in literature, the results show that indigenous participants in our study recognize signifi-
cantly more plant species from photographs than from voucher specimens. It is argued that,
especially in remote and isolated study sites, photographs might be advantageous over
Key Words: Field methods, photographs, voucher specimens, transect, walk-in-the-woods,
homegarden, Bolivia, Yuracare, Trinitario.
One of the principal tools used by ethnob-
otanists for obtaining ethnobotanical information
is the interview. Apart from the questions them-
selves, a great deal of variation in interviews can be
attributed to the plant prop that is used to stimu-
late answers from respondents. During interviews
in situ, living plants are used as a reference, and
questions are asked at the location where plants are
growing, including cultural as well as natural land-
scapes (e.g., Boom 1987; De Walt et al. 1999;
Phillips et al. 1994). By contrast, ex situ interviews
make use of harvested and/or artificial plant refer-
ences and are conducted physically away from the
plant's growing place sensu strictu, such as in the
home of a participant. Several scholars use fresh
1 Received 14 March 2007; revisions requested 23
April 2007; revision received 20 August 2007; accepted
22 August 2007.
material for ex situ interviews (e.g., Bennett et al.
2002; Davis and Yost 1983; Plotkin 1994), while
others resort to voucher specimens (e.g., Vande-
broek, Thomas, and AMETRAC 2003). Fewer re-
ports exist on the use of plant artifacts or handi-
crafts (Boom 1987), photographs (Alexiades 1996;
Martin 1995; Nguyen 2003), or line drawings
(Atran et at 2002; Ellen, personal communica-
tion) as stimuli for interviews.
Drawing upon our own research experience, in
the present paper we discuss the pros and cons of
in situ interviews during transect, walk-in-the-
woods, and homegarden sampling, and ex situ
interviews using fresh plant material, voucher
specimens, and photographs. A major concern re-
garding the use of photographs lies with the abil-
ity of participants to link abstract images to real-
ity. To the best of our knowledge, there exists no
literature in ethnobotanical research evaluating
plant recognition accuracy based on photographs
and voucher specimens as compared to in situ
Economic Botany, 61(4), 2007, pp. 376-384.
© 2007, by The New York Botanical Garden Press, Bronx, NY 10458-5126 U.S.A.
THOMAS ET AL.: PHOTOGRAPHS IN ETHNOBOTANY?
identification. Nguyen (2003) reports on people's
ability to identify photographed fruits, but no ref-
erence is made to their identification skills of fresh
fruits of these species. The null hypothesis from
which we depart is that a similar proportion of
plant species is recognized by participants from
photographs and voucher specimens.
Research was conducted in five Yuracare and
Trinitario communities from the Territorio Indi-
gena Parque National Isiboro-Secure (TIPNIS):
Sanandita, San Antonio, San Jose de la Angosta,
El Carmen de la Nueva Esperanza, and Tres de
Mayo. TIPNIS is officially a protected area since
1965. Its 12,000 km' surface area (between
16°23-16°40'S and 65°41-65°57'W) is cut in
half by the inaccurately-defined departmental
limit that separates Bolivia's departments of Beni
and Cochabamba. The participating communi-
ties are situated on the Cochabamba side of TIP-
NIS, at altitudes below 500m. The research area
is characterized by a warm subtropical climate
with mean annual temperature and precipitation
of about 27°C and 4000 mm, respectively (Rico
Pareja, Beetstra, and Rocha Torrez 2005). For a
detailed description of the tropical forest vegeta-
tion, see Vandebroek et al. (2004b) and Thomas
and Vandebroek (2006).
The participating communities are character-
ized by low accessibility (Vandebroek et al. 2004a).
Following a 3-5 hour drive in a cargo truck, the
nearest community is reached on foot after a 0.5-4
hour walk, the furthermost after a 5-8 hour walk,
depending on the location of the last truck stop
and weather conditions.
The Yuracare speak an unclassified language
(Van Gijn 2006). In origin, they were hunter-
gatherers with limited practice of agriculture. TIP-
NIS is one of their original habitats (Querejazu
2005). The Trinitarios (a subgroup of the Mo-
jefios or Moxenos) belong to the Arawak language
family (Rico Pareja, Beetstra, and Rocha Torrez
2005). In the Bolivian municipality of Moxos
(Beni Department), the Mojefios reached a high
level of development in pre-Hispanic times. From
the nineteenth century onwards, many Trinitarios
started to migrate from Moxos in search of the
holy land (Loma Santa) as a response to land pres-
sure problems in their homeland area (Riester
1976) and ended up in TIPNIS. The participat-
ing Trinitario communities were established from
the 1970s onwards (Vandebroek et al. 2004a).
Acceptance of the present research project by
leaders of indigenous umbrella councils and the
participating communities was formalized by writ-
ten agreements between researchers, indigenous
representatives, and the Centro de Biodiversidad y
Genetica from the Universidad Mayor de San
For ecological quantification of the local
flora, Gentry's transect method was applied in
four nonpermanent forest transects, whereby all
plant individuals with dbh (diameter at breast
height) > 2.5 cm were sampled and measured in 10
consecutive 50 x 2 m2 transects (see Phillips and
Miller 2002). Every measured plant was recorded
as a unique "morphospecies," and a voucher collec-
tion was made if the taxon was encountered for the
first time, or in case its identity was uncertain (cf.
Phillips et al. 2003). We supplemented Gentry's
method with the collection of useful plants not ful-
filling protocol requirements of minimum dbh and
growing along (or near) the transect line (cf. Bern-
stein, Ellen, and Antaran 1997). This resulted in
the inclusion of useful herbs, small epiphytes, and
vines in the inventory. Transect sampling was com-
plemented with "walk-in-the-woods," whereby
participants are encouraged to actively lead field
trips and seek useful plants they know and/or use
(Phillips and Gentry 1993a). As a third sampling
strategy, useful plants were collected from home-
gardens upon indication by their owners.
During transect sampling and walk-in-the-
woods, accompanying participants were always
men, whereas mainly women escorted during
homegarden collections. This is a logical approach,
since most men spend more time in forests than
women who are frequently more knowledgeable
about nonforest or homegarden plants (Bernstein,
Ellen, and Antaran 1997; Kainer and Duryea
1992). Ethnobotanical interviews in situ were con-
ducted using all three sampling techniques with
one or two accompanying participants.
Plant samples were temporarily conserved
in alcohol on the same day of collection. Upon
arrival in the herbarium several weeks later,
plants were transferred immediately to a drier.
In this way, good quality voucher specimens
were obtained. Part of the voucher collection
was afterwards brought back to the participating
communities to be used as a prop during inter-
views. All plants were identified by the first au-
thor and international taxonomical specialists
and deposited in the national Bolivian herbaria
(BOLV and LPB).
In the village environment (ex situ), we inter-
viewed 22 participants (7 women and 15 men;
11 Yuracares and 11 Trinitarios), who were se-
lected through peer recommendations (see Davis
and Wagner 2003). Participants were presented
with fresh plant material, voucher specimens, and
photographs at distinct moments in time. Indi-
vidual semistructured interviews to obtain de-
tailed information on plant names and uses were
conducted in Spanish.
To evaluate plant recognition accuracy, five key
participants were encouraged to identify 50 useful
plant species each, from both photographs and
voucher specimens. All participants (three Trinitar-
ios and two Yuracares) were male and lived in
different communities. Female participants were
not included since less than 50 plants were, col-
lected with each of them. Only species with unique
and clearly distinguished vernacular names that had
been pointed out by each participant during prior
walk-in-the-woods were included. Different botani-
cal species were collected with each participant,
and therefore the nature of photographs and voucher
specimens differed between participants. Nonethe-
less, we made sure that each group of 50 specific
plant species was known to each participant, at least
in situ, as they had pointed out the plants to us.
Accuracy of recognition of plants from photo-
graphs or voucher specimens was evaluated based
on how often each participant mentioned the same
vernacular name on both occasions. Interviews were
conducted at least two months after the walk-in-
the-woods collection trips. Moreover, interview
events using either photographs or dried specimens
were separated by at least one month. These mea-
sures were taken to prevent participants from re-
membering collection trips or previous interviews
on the same species, and hence to avoid possible
bias. All interviews were conducted by the collector
of the plant specimens. SPSS 12.0 was used to
compare mean percentages of plant recognition
from photographs and voucher specimens.
Results and Discussion
IN SITU INTERVIEWS DURING TRANSECTS,
WALK-IN-THE-WOODS, AND HOMEGARDEN
After participant observation, the most direct
and reliable method to obtain ethnobotanical in-
formation is to interview participants in situ ("on
the spot during plant collection") (Alexiades
1996). Even in case plants are sterile, this method
allows participants to make use of a maximum of
decisive variables for plant recognition, such as
anatomical, physiological, morphological, architec-
tural, or ecological characteristics. Many of these
features are often lacking in plant artifacts used
during ex situ interviewing.
Despite the benefits of working in situ, some
disadvantages of transect interviewing are worth
mentioning. As discussed by Alexiades (1996),
relatively few participants can be involved in
transect sampling when time is an issue, poten-
tially hampering quantification of collected eth-
nobotanical data. Traditionally, scholars have per-
formed in situ transect interviews with a small
number of locally-recognized plant experts. Al-
though in some cases the plant knowledge of ex-
perts can be extrapolated to the entire commu-
nity, this is not universally valid. It has been
shown that the distribution of local knowledge is
not random but patterned according to age, gen-
der, or social status (Bernstein, Ellen, and An-
taran 1997; D'Andrade 1987; Kainer and Duryea
1992). Accordingly, Phillips and Gentry (1993a
and b) have argued that large, representative sam-
ples of participants from different knowledge
strata are more desirable if conclusions about the
breadth, range, or distribution of ethnobotanical
knowledge in the local population as a whole are
to be drawn. For a more detailed account on the
importance of sampling subjects, the reader is re-
ferred to Alexiades (1996) and Martin (1995).
Transect interviewing is a poor technique to
determine the general level of ethnobotanical
knowledge (expert versus nonexpert/lay knowl-
edge), particularly because it limits participation
of people with mobility issues due to physical dis-
abilities, social stigma, social role, bad vision, or
old age. For example, due to household responsi-
bilities or social stigma, women often can not
afford spending entire days in the woods with
(male) foreign researchers. Therefore, transect in-
terviewing necessarily has to be complemented
with ex situ interviews.
Another handicap of transect sampling in forests
is that at any given time the majority of sampled
plant individuals are sterile. According to Phillips
et al. (2003), the probability of an Amazonian
tree being fertile at any one point is less than 4%.
Although local people are often able to identify
THOMAS ET AL.: PHOTOGRAPHS IN ETHNOBOTANY?
the majority of sterile plant species, we have been
able to observe that mistakes are more frequent
than for fertile material.
The exact position of a transect can have a
significant effect on ethnobotanical results. For
instance, transects in disturbance habitats gener-
ally contain higher proportions of herbaceous and
pioneer species, transects in abandoned camps
and early and old-growth fallows are often rich in
exotic and/or managed plant species (see review in
Bennett 1992 and Voeks 2004), whereas in ma-
ture forest transects old-growth forest trees, lianas,
and bushes are predominant. Therefore, it is im-
portant to carefully consider the placement of
transects prior to research and to stimulate partic-
ipants to systematically indicate every useful
species they observe during transect sampling.
Contrary to transect sampling, where the in-
vestigators present species to participants, during
walk-in-the-woods ethnobotanical information is
collected about those plants that are specifically
indicated as being useful by accompanying par-
ticipants. Consequently, quality and reliability of
collected information are probably better than in
the case of transect interviewing. Collection pri-
ority is usually directed to fertile specimens, but
sterile plants should also be collected, as it might
be difficult to find them elsewhere later on.
The main disadvantage of the walk-in-the-
woods method is the lower ecological and ethno-
botanical sampling yield as compared to the tran-
sect method. During day-filling walk-in-the-woods
collection trips, we collected 10-20 specimens and
interviewed accompanying participants about their
uses. In an identical time span, we sampled two
50 x 2 m2 transects, thereby collecting ecological
and ethnobotanical information on 30-40 speci-
mens. Hence, in equal periods of time, more
quantitative data were obtained from transect sam-
pling than from the walk-in-the-woods technique.
During transect sampling, all plant individuals
fulfilling minimum dbh requirements are system-
atically shown to and discussed with participants,
while plants encountered during walk-in-the-
woods are dependent on the route followed, the
memory and perceptivity of the accompanying
participant, phenology, and coincidence. There-
fore, transect sampling favors detection of (incon-
spicuous) useful plant species that are often over-
looked during walk-in-the-woods (also see Van
Andel 2000). On the other hand, walk-in-the-
woods sampling is not bound to protocol require-
ments (such as minimum dbh) and includes vir-
tually all possible plant habits.
There are two additional limitations of the
walk-in-the-woods method, which are similar to
the transect method. These are the relatively low
number of local people who can participate over
a given time span and the need to work with par-
ticipants who are not limited in their mobility.
Advantages and disadvantages of interviews in
homegardens are basically similar to those of
walk-in-the-woods. The main difference is that
homegardens, as cultural landscapes, are usually
situated at short distances from the home envi-
ronment. This favors local people's participation
in research by alternating interview sessions with
household tasks. Due to shorter distances and
lower time investment, homegarden sampling is
better suited to include representatives from all
knowledge strata of a population, allowing less
mobile people to participate and facilitating indi-
vidually based in situ interviews with multiple
Specific plant species are easier to find in home-
gardens as compared to natural vegetation, since
in the former they are cultivated, tolerated, or
protected (Bennett 1992). Although some home-
gardens may represent a significant proportion of
the local flora (Albuquerque, Andrade, and Ca-
ballero 2005), they often contain higher ratios of
herbaceous and exotic plants (Thomas, unpub-
lished data; Voeks 2004). Therefore, homegarden
sampling necessarily needs to be supplemented
with sampling of the natural landscape to comple-
ment ethnobotanical data.
Ex SITU INTERVIEWS WITH FRESH PLANT
MATERIAL, VOUCHER SPECIMENS,
The primary advantage of ex situ interviewing
("away from the collection site of plants") is that,
in a given period of time, more participants can
be involved as compared to in situ methods. Be-
cause interviews are usually held within the vil-
lage or home environment, it is easy to include
less mobile people. Moreover, when showing arti-
fact plant material to participants at a central
place in the village, interest is often drawn from
passersby and children. Such groupings of people
frequently reveal additional ethnobotanical infor-
mation (Alexiades 1996; Plotkin 1994).
Ex situ interviews are less time-consuming and
favor short interview sessions, varying from min-
utes to hours, as opposed to transect and walk-in-
the-woods sampling that usually take at least sev-
eral hours to an entire day. This allows for a
dynamic and efficient use of participants' time.
However, ex situ interviewing also holds limi-
tations. Recognition of plant species is more prob-
lematic than in situ, since many botanical and
ecological details are missing in artifact plant ma-
terial. These data should be provided verbally by
the interviewer. The collector of the plant mate-
rial is the most qualified person to do so, as (s)he
knows best under which conditions the plant was
growing. Other interviewers have to rely entirely
on field notes from the collector, which might in-
troduce some degree of bias in data collection.
One of the most important factors for the suc-
cess of ex situ interviewing concerns the selection
of appropriate artifact plant material. In the fol-
lowing paragraphs, the strengths and weaknesses of
using fresh plant material, voucher specimens, and
photographs as reference tools will be discussed.
Logically, the best reference tool to promote
positive plant identification during ex situ inter-
views is freshly-collected plant material. This
would be the best choice for people with limited
mobility and participants with limited eyesight,
since they are able to touch, smell, and see plant
samples from all angles. However, as pointed out
by Alexiades (1996), it may not be evident for par-
ticipants to recognize small sections of large shrubs
or trees, and therefore even fresh material is likely
to be recognized less easily ex situ as compared to
A problem with fresh material is that it wilts
rapidly, which is certainly the case in hot and
humid climates. Therefore, plants should be
pressed and dried, or soaked in alcohol, as soon as
possible after collection, consequently limiting
time available for interviewing. In this respect it is
wise to collect extra duplicates, including various
specimens for immediate conservation, and one or
two specimens designated for interviewing.
Since good artificial light sources are frequently
nonexistent in remote tropical settings, daylight is
a prerequisite for plant interviews. Therefore, day-
filling activities such as transect sampling are often
difficult to combine afterwards with interviews
based on fresh material. The walk-in-the-woods
method offers more flexibility when kept short and
conducted in the morning. The afternoon can
then be used to interview people about collected
plant material that is still fresh for a few hours.
The use of fresh plants seems less appropriate
when one strives to gather extensive quantitative
ethnobotanical data. With fresh material, the
number of participants living in distant commu-
nities can only be increased through repeated col-
lections of the same species. This requires a higher
sampling investment, which may not always be
possible due to constraining factors such as plant
phenology, abundance, and accessibility.
Voucher Specimens and Color Photographs
The pros and cons of voucher specimens and
photographs for use as ethnobotanical reference
tools are best understood when discussed in com-
parison. To allow a transparent evaluation of both
methods, we first describe some key techniques
regarding the production of photographs as plant
Taking, Developing, and Preserving Pictures
With the digital era in full expansion, taking
pictures has never been easier. Storage possibili-
ties of pictures have become unlimited, photo
quality can be checked immediately, and rela-
tively cheap digital cameras are available that pro-
vide astonishing possibilities. Moreover, when the
camera is protected with a waterproof case, one
does not have to worry about humidity and rain.
Pictures should always be taken of fresh mate-
rial, preferably in situ. It is important to include as
much detail about important plant characteristics
as possible, such as details of flowers, fruits (in-
cluding cross sections), bark aspect, stem cuttings
(including exudates), associated fauna (e.g., sym-
biotic ants), and/or flora (e.g., hosts of [hemi] par-
asites), etc. Branches of large trees, obtained
through tree climbing or by means of telescopic se-
cateurs, are best photographed on the ground.
To preserve photographs for long-term use
during interviewing, it is worthwhile to plasticize
them. The number of photographs per (A4/of-
fice) sheet can range from 1-9 on each side, de-
pending on desired clarity (Fig. 1). According to
our experience, good qualitative and durable re-
sults are obtained with high-resolution deskjet
color prints on normal paper that are plasticized
afterwards. This approach results in a significant
reduction of costs as compared to commercial de-
velopment of photographs. Nevertheless, prepar-
ing photographic sheets can represent a higher
THOMAS ET AL.: PHOTOGRAPHS IN ETHNOBOTANY?
Fig. 1. Example of a full-color photo sheet used for ethnobotanical interviews (actual size 21 cm x 29.7 cm);
plasticized together with another photo sheet on the reverse side. (1: Geophila macropoda [Ruiz and Pavon] DC.;
2: Pycnoporus sanguineus [L. ex Fr.] Murril; 3: jacaratia digitata [Poepp. and Endl.] Solms; 4: Mouriri cauliflora
Mart. ex DC.)
time investment and cost as compared to prepar-
ing voucher specimens.
However, in spite of the many advantages of
digital technology, it should be pointed out that
the quality of pictures can only be as good as the
photographer's experience with this medium.
Therefore, training in and experience with (macro)
photography prior to engaging in ethnobotanical
field work are recommended.
Use of Color Photographs Versus Voucher Specimens
as Ethnobotanical Props
Photographs are much lighter and more com-
pact than voucher specimens. This is particularly
important when working with large numbers of
plant specimens, as was the case in our own re-
search (> 1500 specimens). When large distances
have to be traveled in uncomfortable public
transport and on foot, heavy and oversized lug-
gage is to be avoided. Moreover, photographs are
relatively robust when properly plasticized.
Voucher specimens by contrast, are fragile and
have to be protected against rot, insect attack,
and from getting wet during rains or when cross-
ing rivers (cf. Martin 1995).
Ethnobotanists who have worked with dried
specimens are familiar with the unintentional
breaking of specimens by participants during in-
terview sessions, often creating the necessity to re-
place specimens by new collections. Photographic
sheets are much easier to handle, and one can
look through the sheets for certain pictures with
relative ease. Even children can take them in hand
without causing any damage, which creates an op-
portunity to collect additional information.
Photographs provide good representation possi-
bilities of plant species. Oversized plant parts (e.g.,
palm leafs, large fruits) are generally easily cap-
tured as images, while they necessarily must be cut
into pieces when working with dried specimens.
Also, typical fresh characteristics like plant exu-
dates can be represented in photographs, but are
generally not present in dried specimens.
Photographs facilitate simultaneous presenta-
tion of different phenological phases that are en-
countered throughout the fieldwork period. For
dried specimens, this is usually only possible by
showing various specimens. Furthermore, photo-
graphs provide natural—albeit two-dimensional—
representations of three-dimensional plants.
Through pressing and drying, plants become flat-
tened, "two-dimensionalized" specimens that un-
avoidably lose much of their natural aspect (Alexi-
ades 1996; Martin 1995).
A characteristic of photographs that can be
seen either as positive or negative is the ability to
magnify or reduce reality. Magnifying details that
are crucial for identification by indigenous people
can promote recognition considerably. On , the
other hand, a representation scale that is different
from 1:1 can also complicate identification, par-
ticularly for participants who are unfamiliar with
media such as photography or television. This
issue can be countered by including a scaling ref-
erence, such as a person standing next to an en-
tire plant in photographs, or a hand holding a de-
picted plant part.
A disadvantage of photographs is that recog-
nition of plants is strictly visual, whereas in
dried species at least some natural characteristics
like roughness, hairs, and odors are still present.
Also, details that are often not represented in
photographs should in theory be visible on dried
specimens, as far as they did not vanish during
the conservation process. To avoid misidentifica-
tions, photographs or voucher specimens of ster-
ile plant species can generally not be used for
Picture angle and focus largely depend on deci-
sions made by the photographer, and therefore
plant characteristics that are actually portrayed
can differ substantially from those that people use
for identification. After all, "visual data are not
`what the camera can record but. . . what the eye
can see"' (Emmison and Smith 2000). Therefore,
it is advisable to ask people which specific charac-
teristics they use for plant identification prior to
taking pictures. Another option is to let local
people take the picture, or at least show the re-
searcher how they would do so (Reyes-Garcia,
Plant recognition: Color Photographs
or Voucher Specimens?
A major concern when using photographs for
ethnobotanical interviewing is the ability of partic-
ipants to recognize plant species from pictures. Re-
sults from the recognition test presented in Table 1
show that our participants recognized 92-96% of
photographs of plants that they had indicated and
named during prior walk-in-the-woods collection
trips. For voucher specimens, the recognition rate
was significantly lower and ranged from 68-86%
(Mann-Whitney test; p < 0.01). These results con-
firm the usefulness of photographs in ethnobotan-
ical interviews, as has been recently pointed out by
Nguyen (2003). This does not imply that we claim
these results to be universally valid. For example,
Reyes-Garcia et al. (2003) have argued that Boli-
vian Tsimane' people had difficulties identifying
objects through pictures. Therefore, further re-
search on people's ability to recognize plants from
photographs will be necessary to substantiate our
claim on a larger scale.
If time is not an issue, the most appropriate
way to carry out a (quantitative) ethnobotanical
study would be to conduct in situ interviews dur-
ing transect, walk-in-the-woods, or homegarden
sampling, which can partly be combined with ex
situ interviews using fresh plant material. In this
TABLE 1. NUMBER (#) AND PERCENTAGE (%) OF 50 PLANT SPECIES RECOGNIZED BY FIVE PARTICIPANTS
BASED ON PHOTOGRAPHS AND VOUCHER SPECIMENS.
Recognition photographs? #
Recognition voucher specimens?
THOMAS ET AL.: PHOTOGRAPHS IN ETHNOBOTANY?
way, participants are presented only with those
plant props that they are used to seeing in real life.
However, when time, financial, and/or human re-
sources are limited, researchers often have to turn
to using artificial representations of plant species
such as voucher specimens and/or plant photo-
graphs for optimizing return on time investment
and quantification of ethnobotanical data.
Our findings demonstrate the usefulness of
photographs in ethnobotanical interviews. Fur-
thermore, we suggest that in some cases using
photographs as props for interviews can be a bet-
ter alternative than voucher specimens, particu-
larly when research is conducted in remote and
isolated study sites. However, in order to general-
ize this conclusion, people's ability to recognize
plant species from photographs should be evalu-
ated on a larger scale with different ethnic groups
and communities, and with representatives from
all knowledge strata.
The present research was financed by a doctoral
research grant of the Bijzonder Onderzoeksfonds
of Ghent University to Evert Thomas (Grant
Number: B/03801/01). Logistic support in Bo-
livia was provided by the Centre of Biodiversity
and Genetics and the Herbarium Martin Carde-
nas of the Universidad Mayor de San Simon in
Cochabamba. Special thanks are due to all inhabi-
tants of the indigenous communities San Jose de
la Angosta, San Antonio, El Carmen de la Nueva
Esperanza, Tres de Mayo, and Sanandita for their
kind assistance in this project.
Albuquerque, U. P., L. H. C. Andrade, and J. Ca-
ballero. 2005. Structure and Floristics of Homegar-
dens in Northeastern Brazil. Journal of Arid Envi-
Alexiades, M. N. 1996. Collecting Ethnobotanical Data:
An Introduction to Basic Concepts and Techniques.
Pages 53-94 in M. N. Alexiades, ed. Selected Guide-
lines for Ethnobotanical Research: A Field Manual.
The New York Botanical Garden, New York
Atran, S., D. Medin, N. Ross, E. Lynch, V. Vapnarsky,
E. U. Ek', J. Coley, C. Timura, and M. Baran.
2002. Folkecology, Cultural Epidemiology, and the
Spirit of the Commons. Current Anthropology
Bennett, B. 1992. Plants and People of the Amazonian
Rainforest: The Role of Ethnobotany in Sustainable
Development. Bioscience 42:599-607.
M. A. Baker, and P. Gomez Andrade, eds.
2002. Ethnobotany of the Shuar of Eastern Ecua-
dor. Advances in Economic Botany 14.
Bernstein, J. H., R. Ellen, and B. B. Antaran. 1997.
The Use of Plot Surveys for the Study of Ethnobo-
tanical Knowledge: A Brunei Dusun Example.
Journal of Ethnobiology 17:69-96.
Boom, B. 1987. Ethnobotany of the Chacobo Indians.
Advances in Economic Botany 4:1-68.
D'Andrade, R. G. 1987. Modal Responses and Cul-
tural Expertise. American Behavioral Scientist
Davis, A., and J. R. Wagner 2003. Who Knows? On
the Importance of Identifying "Experts" When Re-
searching Local Ecological Knowledge. Human
Davis, E. W., and J. A. Yost 1983. The Ethnobotany of
the Waorani of Eastern Ecuador. Botanical Mu-
seum Leaflets 29:159-217.
DeWalt, S. J., G. Bourdy, L. R. Chavez de Michel, and
C. Quenevo. 1999. Ethnobotany of the Tacana:
Quantitative Inventories of Two Permanent Plots of
Northwestern Bolivia. Economic Botany 53:
Emmison, M., and P. Smith. 2000. Researching the
Visual: Images, Objects, Contexts and Interactions
in Social and Cultural Inquiry. Sage, London.
Kainer, K. A., and M. L. Duryea. 1992. Tapping
Women's Knowledge: Plant Resource Use in Ex-
tractive Reserves, Acre, Brazil. Economic Botany
Martin, G. J. 1995. Ethnobotany: A Methods Manual.
Chapman and Hall, London.
Nguyen M. L. T. 2003. Comparison of Food Plant
Knowledge between Urban Vietnamese Living in
Vietnam and in Hawai'i. Economic Botany
Phillips, O. L., and A. H. Gentry. 1993a. The Useful
Plants of Tambopata, Peru: I. Statistical Hypotheses
Tests with a New Quantitative Technique. Eco-
nomic Botany 47:15-32.
and . 1993b. The Useful Plants of
Tambopata, Peru: II. Additional Hypotheses Testing
in Quantitative Ethnobotany. Economic Botany
C. Reynel, P Wilkin, and B. C.
Galvez-Durand. 1994. Quantitative Ethnobotany
and Amazonian Conservation. Conservation Biol-
and J. Miller 2002. Global Patterns of Forest
Diversity: The Dataset of Alwyn H. Gentry. Mono-
graphs in Systematic Botany 89. Missouri Botanical
Garden, St. Louis, Missouri.
R. Vasquez Martinez, P. N6fiez Vargas, A.
Lorenzo Monteagudo, M.-E. Chuspe Zans, W.
Galiano Sanchez, A. Pena Cruz, M. Timana, M.
Yli-Halla, and S. Rose. 2003. Efficient Plot-Based
Floristic Assessment of Tropical Forests. Journal of
Tropical Ecology 19:629-645.
Plotkin, M. J. 1994. Tales of a Shaman's Apprentice.
An Ethnobotanist Searches for New Medicines in
the Amazon Rainforest. Penguin Books, New York.
Querejazu, R. L. 2005. La Cultura de los Yuracares, su
Habitat y su Proceso de Cambio. Impresiones
Poligraf, Cochabamba, Bolivia.
Reyes-Garcia, V., E. Byron, V. Vadez, R. Godoy, L.
Apaza, E. P. Limache, W. R. Leonard, and D.
Wilkie. 2003. Measuring Culture as Shared
Knowledge: Do Data Collection Formats Matter?
Cultural Knowledge of Plant Uses among Tsi-
mane' Amerindians, Bolivia. Field Methods
Rico Pareja, A., T. Beetstra, and F. M. Rocha Torrez.
2005. Atlas del Tropico de Cochabamba. Talleres
Graficos "KIPUS," Cochabamba, Bolivia.
Riester, J. ed. 1976. En Busca de la Loma Santa. Edito-
rial Los Amigos del Libro, La Paz—Cochabamba,
Thomas, E., and I. Vandebroek. 2006. Guia de Plan-
tas Medicinales de los Yuracares y Trinitarios del
Territorio Indigena Parque Nacional Isiboro-
Secure, Bolivia. Imprenta Sirena, Santa Cruz,
Van Andel, T. 2000. Non-Timber Forest Products of
the North-West District of Guyana. Part I.
Tropenbos-Guayana Series 8a, Utrecht University,
Van Gijn, E. 2006. A Grammar of Yurakare. Nijmegen
University, The Netherlands.
Vandebroek, I., J.-B. Calewaert, S. De Jonckheere, S.
Sanca, L. Semo, P. Van Damme, L. Van Puyvelde,
and N. De Kimpe. 2004a. Use of Medicinal Plants
and Pharmaceuticals by Indigenous Communities
in the Bolivian Andes and Amazon. Bulletin of the
World Health Organization 82:243-250.
P. Van Damme, L. Van Puyvelde, S. Arrazola,
and N. De Kimpe. 2004b. A Comparison of Tradi-
tional Healers' Medicinal Plant Knowledge in the
Bolivian Andes and Amazon. Social Science and
E. Thomas, and AMETRAC. 2003. Plantas
Medicinales para Ia Atencion Primaria de la Salud.
El Conocimiento de Ocho Medicos Tradicionales
de Apillapampa (Bolivia). Industrias Graficas Ser-
rano, Cochabamba, Bolivia.
Voeks, R. A. 2004. Disturbance Pharmacopoeias:
Medicine and Myth from the Humid Tropics. An-
nals of the Association of American Geographers