Agriculture and Human Values 17: 169–179, 2000.
   © 2000 Kluwer Academic Publishers. Printed in the Netherlands.


Identifying ecosystem services using multiple methods: Lessons from the
mangrove wetlands of Yucatan, Mexico

Michael D. Kaplowitz
Department of Resource Development, Michigan State University, East Lansing, Michigan, USA

Accepted in revised form October 10, 1999

Abstract. The failure to properly account for the total value of environmental and natural resources results in
socially undesirable overexploitation and degradation of complex ecosystems such as mangrove wetlands. How-
ever, most ecosystem valuation research too often focuses on the question of “what is the value” and not enough
on “what people value.” Nonmarket valuation practitioners have used qualitative approaches in their work for
some time. Yet, the relative strengths and weaknesses of different qualitative methods have been more the subject
of speculation than systematic research. The statistical examination of focus group and individual interview data
on ecosystem services illustrates that the two methods generate important but different ecosystem service data.
Further, the data show that the use of multiple data collection methods offers a more robust understanding of what
people value.

Key words: Focus groups, Interviews, Nonmarket valuation, Qualitative methods, Statistical analysis

Michael D. Kaplowitz is an Assistant Professor in the Department of Resource Development at Michigan State
University. This paper is based on original research made possible, in part, by support by the Inter-American
Foundation and the Organization of American States.


Introduction                          that are not well-captured in markets (Aylward and
                                Barbier, 1992; Barbier et al., 1997; Carson, 1998;
The failure to properly account for the total value      Mitsch and Gosselink, 1993). In particular, the
of environmental and natural resources results in       value of wetland ecosystems may be especially great
socially undesirable overexploitation and degradation     in developing countries where efficient markets for
of complex ecosystems such as mangrove wetlands        wetland services do not exist (Aylward and Barbier,
(Clark, 1996; Farnsworth and Ellison, 1997; Hamilton      1992; Barbier et al., 1997; Carson, 1998). However,
et al., 1989; Spaninks and van Beukering, 1997).        the availability of valuation methods for estimating
Complex environmental and natural resources, such as      wetland economic values does not necessarily mean
the Yucatán’s mangrove wetlands, represent substan-      that the pertinent resources services are identified and
tial sources of cultural, intergenerational, environ-     included in wetland ecosystem valuation studies and
mental, and economic wealth (Aylward and Barbier,       policy decisions.
1992; Bann, 1997; Barbier, 1994; Barbier et al.,          The reported research examines two relatively
1997; Carson, 1998; Perrings, 1995). However, most       inexpensive research methods for helping researchers
ecosystem valuation research is “too focused on the      identify relevant ecosystem services associated with a
question of ‘what is the value’ and not enough on       mangrove wetland. Using focus groups and individual
what, in particular, people value” (Swallow et al.,      interviews, the researcher explored what local resource
1998). There is a need for resource valuation research     beneficiaries associate with the mangrove wetland of
to identify the range and relative importance of the      Chelém Lagoon. The study identifies the particular
components of ecosystem value rather than merely        mangrove wetland services important and relevant to
estimate some value for a particular ecosystem service.    the inhabitants of two communities along the coastal
  Despite this need for understanding the com-        fringe west of Progresso, Mexico. The study demon-
ponents of ecosystem value, it is prohibitively        strates that the use of both focus groups and in-
expensive and unrealistic to conduct detailed empirical    depth individual interviews can lead to a more robust
nonmarket valuation studies of each ecosystem. The       understanding of what people value about a shared
need for ecosystem valuation information is especially     ecosystem. Furthermore, the study addresses a gap in
great for those public good services of ecosystems       the resource valuation literature by using an empirical
170                      MICHAEL D. KAPLOWITZ

method to compare the outcomes of group discus-       1993). Examples of natural resource use values include
sions with individual interviews concerning ecosystem    camping, hunting, wood collection, fishing, farming,
services (Chilton and Hutchinson, 1999).           as well as such things as breathing clean air. Values
  First, the paper reviews some of the natural       independent of in situ activities have been called
resource services attributed to mangrove ecosystems.     passive use or nonuse values. Examples of nonuse
Next, the paper describes how valuation research       values include the value of knowing the resource
has used qualitative research methods in some valu-     simply exists, the value some people attribute to
ation studies of natural resources. The paper then      some potential use of the resource, and the value
describes the research design and method that was      of knowing that future generations will have the
used to test the hypothesis that focus groups and indi-   resource (Freeman, 1993). In order to properly account
vidual interviews help researchers identify substan-     for the total value of ecosystems in their decision-
tially similar ecosystem services associated with a     making, policymakers should understand the extent
mangrove wetland. The research results are presented     and magnitude of use and nonuse values associated
before discussing the implications of the findings on     with the resource.
the usefulness of multiple methods, ways to improve
valuation studies, and the ability of statistical analysis  Qualitative methods and valuation studies
to shed light on the significance of qualitative data.
                               Social scientists in diverse fields of study regularly use
                               qualitative methods as comprehensive research tools
                               and as important components in designing and imple-
Background
                               menting reliable research studies (Krueger, 1994;
                               Morgan, 1997; Schwarz, 1997; Sudman et al., 1996;
Mangrove wetland values
                               Weiss, 1994). Studies for estimating the economic
The term mangrove refers to a number of tree species     value of environmental and natural resources range
capable of living in saltwater or salty soils. Mangroves   from market or behavior-based methods to direct
                               methods such as contingent valuation (CV) studies.1
and their ecosystems are found in intertidal areas
of sheltered coastlines called lagoons and estuaries.    For some time, resource valuation researchers have
Ecologically, mangrove wetlands maintain high levels     been advised to consider using focus group interviews
of biological productivity; export nutrients to outside   as well as individual interviews for questionnaire
waters; and provide habitat for valuable plant and      pretesting and development (Mitchell and Carson,
animal species (Clark, 1996). Mangrove ecosystems      1989). Despite some initial skepticism of the utility
are also important to the subsistence livelihood of     of qualitative methods for designing nonmarket valu-
tropical coastal communities (Hamilton et al., 1989;     ation studies (e.g., Arrow et al., 1993), focus groups
Hamilton and Snedaker, 1984). Mangrove ecosys-        have been increasingly recognized and relied upon as
tems potentially provide an array of important indirect   important aspects of resource valuation questionnaire
services – prevention of storm damage, flood and       design and evaluation (Carson and Mitchell, 1993;
water control, support of fisheries, waste absorption,    Schkade and Payne, 1994; Chilton and Hutchinson,
recreation, and transport (Barbier, 1994; Barbier et     1999; Hutchinson et al., 1995). Individual interviews
al., 1997). Mangrove ecosystems may be directly       have also been reported to provide efficient means for
exploited by extracting goods such as fish, agricul-     collecting information on beneficiaries’ use and under-
ture, wildlife, wood, and fresh water (Bann, 1997;      standing of mangrove ecosystems at the local level
Bennet and Reynolds, 1993; Farnsworth and Ellison,      (Kovacs, 1999).
1997; Hirsch and Mauser, 1992; Kunstadter et al.,        Work by cognitive psychologists and survey
1985; Ruitenbeek, 1992). Additionally, mangrove       method researchers underscore the value of quali-
wetlands have also been said to be significant sources    tative research methods for questionnaire design
of nonuse benefits that do not flow from direct use of     (Schwarz, 1997; Sudman et al., 1996). These same
the ecosystem (Aylward and Barbier, 1992; Barbier,      researchers point out that one qualitative research
1994; Barbier et al., 1997).                 method alone may be insufficient to learn about
  Mangrove ecosystems, like other complex environ-     respondents’ resource use and understanding. Some
mental and natural resources, are potential sources of    researchers suggest that focus groups and individual
an array of use and nonuse values (Barbier, 1994;      interviews may lead to the discovery of different infor-
Barbier et al., 1997; Carson, 1998; Hamilton et       mation (De Jong and Schellens, 1998; Kitzinger,
al., 1989). While not dependant upon entry directly     1994a, 1994b). Other researchers assert that focus
into markets, use values require that some in situ      group research should be combined with other types of
activity takes place that benefits individuals (Freeman,   research, including individual interviews, to triangu-
                                                        171
                    IDENTIFYING ECOSYSTEM SERVICES

late or corroborate research findings (Bryman, 1988;    Design and procedure
Morgan, 1996). Multiple qualitative methods such as
                             The research design allowed for examination of
focus groups and one-on-one interviews may be useful
                             the collected data across interview type, gender,
for revealing a wide range of local beneficiaries’ ideas
                             and community (see Figure 1). Research assistants
about and conception of complex environmental and
                             canvassed randomly selected sections of the target
natural resources (e.g., Carson et al., 1994; Chilton
                             communities at staggered times of day to recruit partic-
et al., 1998; Hutchinson et al., 1995). Thus there is a
                             ipants. The focus groups were comprised of between
need for research that compares the outcomes of focus
                             four and seven individuals of the same gender from the
group and individual interviews regarding the respond-
                             same village. No respondent or their family members
ents’ understanding of complex ecosystems (Chilton
                             participated in more than one focus group or inter-
and Hutchinson, 1999).
                             view. The focus groups and individual interviews were
                             designed and implemented following the generally
                             accepted practices of Morgan4 (1996, 1997, 1998)
Research design and method
                             and Weiss5 (1994) respectively. A Mexican profes-
                             sional moderator using a specially prepared discussion
Research question
                             guide conducted the focus groups and individual inter-
The relative strength and weakness of particular quali-  views. All focus group and individual interviews were
tative research methods “has been more the subject    tape-recorded and subsequently transcribed.
of speculation than systematic research” (Morgan,
1997: 13). A few researchers have explored differ-    Qualitative analysis
ences in focus group and individual interview infor-
                             The qualitative data analysis allowed the researchers
mation (De Jong and Schellens, 1998; Kitzinger,
                             to (1) discover themes, (2) consider the choice and
1994a, 1994b). Qualitative methods may be used
                             meanings of words, (3) consider the context(s) of
successfully to learn from local beneficiaries how they
                             data collection, and (4) consider the consistency of
use, perceive, and value environmental and natural
                             responses (Krueger, 1994). Although work remains
resources (Mandondo, 1997). Studies also show that
                             in developing uniform guidelines and rules for the
resource beneficiaries’ ideas about natural resources
                             qualitative coding and analysis process (Fredricks and
may differ from those of scientists and so-called
                             Miller, 1997), the researcher attempted to systemat-
experts (Talawar and Rhoades, 1998). This reported
                             ically reveal elements of respondents’ experience and
research examines the research hypothesis that focus
                             perceptions. The qualitative analysis did not produce
groups and individual interviews, all else being equal,
                             simple counts of things, but rather “fractured” the
reveal similar sets of information about a shared
                             data and rearrange it into categories that facilitated
mangrove ecosystem.
                             understanding the data and comparing the data within
                             and between categories (Maxwell, 1996; Strauss and
Participants
                             Corbin, 1990). After the transcripts were read, the
The communities of Chelém and Chuburná, Mexico      analyst used memos (researcher’s notes and observa-
are located along a 15-kilometer stretch of coastal    tions), categorizing strategies (coding and thematic
fringe that borders the Gulf of Mexico on one side    analysis), and contextualizing strategies (narrative
and Chelém Lagoon on the other. These villages are    analysis and individual case studies).
comprised of families that have traditionally relied     The 12 focus group and 19 individual interview
upon the natural resources of the region, including    transcripts resulted in more than 500 pages of text.
the mangrove wetland, for their subsistence and live-   An iterative, grounded theory approach (Strauss and
lihood. Focus group interviews and individual in-     Corbin, 1990) was used to code the transcripts. First,
depth interviews were conducted with residents of     almost every word of a randomly selected subset
these communities as part of a study evaluating the    of transcripts was coded (open coding). Next a set
importance of mangrove wetlands in Yucatán, Mexico.    of thematic or summary codes was developed (axial
Chelém and Chuburná share similar socio-economic     coding). When no new open codes were necessary to
characteristics and have roughly 475 and 215 house-    code additional transcripts, all of the study’s transcripts
holds respectively (Instituto Nacional de Estadística   were axial coded. The final iteration of coding the text,
Geografía e Informática (INEGI), 1992). A total of 97   selective coding, focused on organizing the data into
year-round residents from the two communities were    36 categories relevant to respondents’ resource use,
interviewed in one of 12 focus groups2 or 19 individual  value, understanding, perception, and control of the
in-depth interviews.3                   ecosystem. The reported research is one means for
172                      MICHAEL D. KAPLOWITZ




                       Figure 1. Research design.

trying to understand the significance of what the quali-   concerning resource beneficiaries’ social conflicts and
tative research revealed about local beneficiaries’ use,   discussion of socially sensitive topics may be found
perception, and understanding of Chelém Lagoon.       elsewhere (Kaplowitz, 1998, 1999; Kaplowitz and
                               Hoehn, 1998).
Operationalizing hypothesis test                 While for some qualitative researchers, summary
                               discursive reports of their findings (e.g., consumer
If focus groups and individual interviews concerning     preferences among brands of a product) are sufficient,
respondents’ relationships with a local mangrove       other researchers rigorously test their research hypoth-
ecosystem yield similar data on beneficiaries’ percep-    eses with a statistical analysis of collected qualitative
tions and appreciation of ecosystem services, one      data (e.g., De Jong and Schellens, 1998; Krippen-
would expect, all else being equal, that transcripts     dorff, 1980). The statistical analysis of qualitative data
of those sessions would evidence a similar set of      has been found to be both possible and helpful (De
data on such services. That is, it would be reason-     Jong and Schellens, 1998; Krippendorff, 1980; Weber,
able to expect that a uniform process of coding the     1990). Differences in focus group and individual inter-
focus group and individual interview transcripts would    view data of text evaluation exercises have been tested
result in similar distributions of codes that capture    using code frequencies, t-tests, and analysis of vari-
use and nonuse services associated with the mangrove     ance (De Jong and Schellens, 1998). Similarly, the
ecosystem. Likewise, if there were particular use      analysis of manifest attributes of text and accompany-
and nonuse services of importance to resource bene-     ing inferential attributes has been performed using
ficiaries, one might expect that systematic analysis of    various counts, percentages, and statistical measures
the focus group and individual interview data would     (Gray and Denstein, 1998). In their recent analysis of
evidence a higher frequency of such codes. This paper    focus group data collected in anticipation of a contin-
focuses on testing the hypothesis that focus groups and   gent valuation studies, Chilton and Hutchinson (1999)
individual interviews reveal substantially similar infor-  “quasi-quantified” qualitative data to test divergence of
mation concerning resource services associated with a    respondent and researchers definitions of goods.
mangrove ecosystem. Presentation of research findings
                                                           173
                     IDENTIFYING ECOSYSTEM SERVICES

                      Table 1. Ecosystem service variables.

    Topic variable   Example                              % sessions rasing topic
                                              Focus groups Individual
                                              interviews

    Beauty       Wetland is beautiful; a pretty place to see; enjoy the views   100      11
    Chivita      Melongena melongena; small shellfish collected; as food and in   100      95
             commerce
    Crab        Collected as bait; frozen for use during 2 month octopus season  92      42
    Lagoon fishing   We fish in lagoon; people come to fish in wetland; there are nets  92      90
             day and night at lagoon entrance
    Salt extraction  Used to be salt here; salt ponds once lucrative; construction   92      37
             destroyed salt business
    Shrimp       Seawater brings shrimp; when shrimp here, all fish for them;    75      16
             not as many shrimp as in past
    Nongame species  Flamingoes; crocodiles; heron; turtles; seagulls          67      42
    Ducks       Ducks sometimes here; few locals benefit; need permit to hunt    42      42
             ducks
    Recreation     Take guests for ride there; sometimes picnic there; celebrate   42      32
             Mass there annually
    Storm protection  Can protect boats from storm; helps if water rises;        42      16
    Wood        Some collect wood for fires; not much wood collection lately    17      5



  The research reported here created and used         by respondents during the focus groups and indi-
discrete variables grounded in economic theory that       vidual interviews. Table 1 illustrates the 11 ecosystem
were derived from the iterative reading, analysis, and     service variables that resulted from the coding and
coding of the transcripts. These discrete variables       variable transformation process of the focus group
recorded those instances that focus group discussions      and individual interview data. Table 1 also presents
and individual interviews raised topics concerning       some examples of representative references and the
wetland ecosystem services associated with Chelém        percentage of focus group and individual interview
Lagoon. For example, the variable Lagoon fishing,        sessions that raised each topic. As can be seen, most
recorded discussion of fishing for corvina, mullet, or      of the services discussed by participants are extractive
other species in the lagoon. Such variables accom-       or consumptive use services (e.g., crab, shrimp, and
modated wide-ranges of discussion topics as well as       wood collection). Some of the services discussed are
allowed the coded transcript data to be subsequently      nonconsumptive uses (e.g., recreation, storm protec-
analyzed using statistical software. The research        tion). A few ecosystem discussed by participants
question was operationalized to statistically test the     appear to be noncunsumptive uses but arguably may
null hypothesis that respondents’ discussions of the      evidence some nonuse value (e.g., beauty, nongame
wetland ecosystem raised the same wetland services       species).
equally during focus groups and individual interview        Table 2 illustrates the relative ranking of frequen-
sessions.                            cies for the ecosystem services variables for the focus
                                group and individual interview data. It illustrates, for
                                example, that wetland beauty was raised during every
Results                             focus group discussion (rank 1), but was only the
                                seventh most frequent service topic raised during indi-
The focus group and individual interview transcript       vidual interviews (rank 7). While perhaps a similar
data were transformed into 12 summary variables to       range of ecosystem services were discussed in the
test the research hypothesis. One summary variable,       focus groups and in the individual interviews, not
Interview type, records the type of interview (e.g.,      every individual interview or focus group raised the
focus group or individual interview) associated with      entire range of mangrove services. However, apparent
each case of coded data. The other 11 summary vari-       differences in aggregate frequencies alone however are
ables capture those wetland ecosystem services raised      insufficient to support or reject the null hypothesis.
174                          MICHAEL D. KAPLOWITZ

       Table 2. Rank of service frequencies.         Table 3. Focus group and individual interview data asso-
                                  ciations.
   Rank   Focus groups     Individual interview
                                              Interview type χ 2
                                  Topic                    P    Odds
   1    Beauty        Chivita                        Group Indiv.         ratio
        Chivita
                                                     23.77a 0.001
                                  Beauty      Yes 12    2           9.50
   2    Crab         Lagoon fishing                  No 0     17
        Lagoon fishing                    Chivita     Yes 12    18       n.s.
        Salt extraction                            No 0     1
                                                      7.62a 0.006 15.12
   3    Shrimp        Crab              Crab       Yes 11    8
                                           No 1     11
                  Ducks
                  Nongame species         Fishing     Yes 11    17       n.s.
                                           No 1     2
   4    Nongame species   Salt extraction
                                                      9.08a 0.003 18.86
                                  Salt extract   Yes 11    7
   5    Ducks        Recreation                    No 1     12
        Recreation
                                                     10.87a 0.001 16.00
                                  Shrimp      Yes  9   3
        Storm protection                           No   3   16
   6    Wood         Shrimp             Nongame     Yes  8   8       n.s.
                  Storm protection                 No   4   11
                                  Ducks      Yes  5   8       n.s.
   7               Beauty
                                           No   7   11
   8               Wood
                                  Recreation    Yes  5   6       n.s.
                                           No   7   13
                                  Storm protection Yes  5   3       n.s.
Absolute differences may be statistically insignificant                No   7   16
when sample size, proportions, expected frequencies,
                                  Wood       Yes 2     1       n.s.
and distributions are taken into account.                      No 10    18       n.s.
  As a result, crosstabulation analysis of each
ecosystem service variable with the interview type         a d.f. = 1, N = 31
variable was generated to test the null hypothesis that,
in the sample population, the same percentage of focus      services by focus group and individual interview data,
groups and individual interviews raised each wetland       it should be remembered that 4 times as many people
service for discussion (see Table 3). Table 3 illus-       participated in focus groups (78) than in individual
trates the Pearson chi-square test of the distribution      interviews (19). All else being equal, if there are differ-
of observed instances that focus groups and indi-        ences in focus group and individual interview data
vidual interviews raised each ecosystem service topic      that are a linear function of number of people, the
against the null hypothesis that each interview type       expected odds ratios should be closer to 4. However,
results in the same frequency of the topic being raised.     the observed odds ratios are two to four times that. This
The null hypothesis was rejected for four variables –      suggests that more than the larger numbers of partic-
Beauty (P < 0.001), Crab (P < 0.006), Salt Extraction      ipants in focus groups is responsible for the increased
(P < 0.003), and Shrimp (P < 0.001). To examine         frequency that certain topics were raised by focus
the strength of the association of interview type with      groups.
respondents’ raising the particular ecosystem service        Four of the six most frequently mentioned topics
in discussion, odds ratios were computed. Table 3        differed significantly in the frequency in which focus
shows, it is about 9 times more likely that a focus group    groups and individual interviews raised them in discus-
of local resource beneficiaries raises the topic of the      sions. Therefore, the research findings support the
mangrove ecosystem’s beauty than an individual inter-      rejection of the null hypothesis. The data show
view. The topics of crab collection, salt extraction, and    that focus groups and individual interviews revealed
fishing for shrimp are respectively 15, 19, and 16 times     significantly different ecosystem service information.
more likely to be raised in focus groups than raised by
individual during one-on-one interviews.
  To further appreciate the significant differences
observed in the frequencies of discussion of ecosystem
                                                        175
                     IDENTIFYING ECOSYSTEM SERVICES

Discussion                         to discover different information about ecosystem
                              services than the individual interviews. Had only
The two methods, focus groups and individual inter-     focus group information been collected and relied
views, do not reveal equal sets of information nor     upon, it would haven been reasonable to believe
do they rank ecosystem services comparably. While      that wetland beauty was extremely important to local
the data illustrate that resource beneficiaries asso-    beneficiaries (perhaps on a par with lagoon fishing).
ciate a variety of ecosystem services with complex     Conversely, had researchers only relied upon indi-
ecosystems, most of the services discussed tended      vidual interview data, wetland beauty and nonuse
to be extractive uses of the ecosystem. The focus      values might have easily been dismissed as unim-
groups and individual interviews were dominated by     portant or beyond the apprehension of respondents.
discussion of lagoon fishing of one type or another.     However, using both individual interview and focus
Although wetland beauty was raised in all of the focus   group data revealed that wetland beauty was signifi-
groups, the low frequency of its discussion by indi-    cant to individuals but only accessible after a dynamic
vidual interviews seems to more accurately reflect      exchange of information. The focus groups seem to
individual beneficiaries’ relative appreciation for non-   have provided a dynamic that allowed respondents to
consumptive and nonuse values of the ecosystem.       identify and discuss nonconsumptive and, at times,
This is no surprise given the economic difficulties     nonuse ecosystem services such as wetland beauty.
facing the communities and Mexico as a whole. The      This finding is in line with the recent work by cognitive
focus groups and individual interviews were replete     psychologists that shows that increased interaction and
with discussions of the difficulty for providing for     exchange of information improves respondents’ under-
one’s family. Increasing commercial fishing pressure     standing of complex ideas (Schwarz, 1997; Schwarz
in the Gulf of Mexico has decimated the once rich      and Sudman, 1995; Sudman et al., 1996). This result,
coastal fishing resource. Local beneficiaries increas-    researchers learning of different and complimentary
ingly rely upon the lagoon and its mangrove ecosystem    ecosystem services using focus groups and individual
for subsistence. Therefore, it is no surprise that     interviews, clearly supports the desirability of using
consumptive use services predominate conversations     multiple methods to corroborate qualitative research
about the ecosystem.                    findings in future ecosystem valuation work (Bryman,
  Only 4 of the 11 wetland services discussed by     1988; Morgan, 1996).
participants were non-extractive in nature – Beauty,
Nongame species, Recreation, and Storm Protection.     Implications for valuation research
The relatively low frequencies associated with the use
                              The results also underscore the difficulty of designing
services of storm protection and recreation in both
                              studies and instruments for estimating the total
focus groups and individual interviews support the
                              economic value of a complex ecosystem. Valuing
notion that these services are not particularly signifi-
                              nonmarket and nonuse services associated with natural
cant to most residents. The other two non-extractive
                              resources, especially in developing countries, seems to
services, Beauty and Nongame species, arguably
                              require extra care. While the study supports the notion
capture some respondents’ recognition and appreci-
                              that nonconsumptive and nonuse values may be signifi-
ation of nonuse ecosystem services. While wetland
                              cant for wetland ecosystems in developing countries
beauty and the presence of nongame species in the
                              (Aylward and Barbier, 1992), the data reveal the poten-
ecosystem may be classified by some as use values
                              tial import of using multiple qualitative methods for
because of the benefits derived from in situ enjoyment
                              identifying potential values to be measured.
of these services, these variables also capture partici-
                                In the case at hand, local resource beneficiaries
pants’ expressed sentiments that wetland beauty and
                              seemed better able to identify and appreciate noncon-
diversity should be preserved for future generations.
                              sumptive and nonuse values in focus group discussions
                              rather than in individual interviews. Since valuation
Value of multiple methods
                              methods such as contingent valuation or contingent
It appears important that wetland beauty was ranked     ranking rely upon individuals, not in groups, making
first by groups and seventh by individuals. The statis-   trade-off choices to reveal nonuse and total economic
tically significant difference in the frequency that focus  values, the findings suggest the import of designing
group and individual interview discussions raised      better valuation survey instruments. The results seem
wetland beauty comports favorably with the find-       to suggest the value of researchers using groups to
ings of De Jong and Schellens (1998) concerning       learn about the array of services that matter to bene-
focus group and individual interview data. The       ficiaries before using individual interviews to validate
mangrove ecosystem focus groups did lead researchers    such findings. Likewise, it seems important to use
176                      MICHAEL D. KAPLOWITZ

                               groups and individual interviews raised chivita collec-
sequential qualitative methods to evaluate how best
                               tion and lagoon fishing in discussions. Conversely,
to communicate and increase information exchange
                               the ecosystem services that only occupy a minor or
concerning ecosystem services in value elicitation
                               cursory place in the communities’ appreciation of
instruments.
                               wetland services do not differ significantly in their
Researchers’ perceptions and beneficiaries’          frequency of discussion in focus groups and individual
understanding                        interviews.
                                 However, the frequency that several ecosystem
The literature is full of lists of use and nonuse services  services raised in focus groups and individual inter-
that in some but not all cases can be associated with    views did differ significantly. The extractive ecosystem
mangrove ecosystems (e.g., Barbier, 1994; Barbier et     services that differed significantly may be thought
al., 1997; Janssen and Padilla, 1996; Spaninks and van    of as sub-components of the more general mangrove
Beukering, 1997). These mangrove services include      ecosystem “fishing” service. The difference in these
on-site fisheries, fuelwood collection, timber harvests,   frequencies may be a function of the difference in the
off-site fishery support, aquaculture, carbon sequestra-   dynamics of a focus group discussion and a one-on-
tion, growing of medicinal plants, biodiversity, recrea-   one depth-interview. For example, shrimp collection
tion, transportation, meat production, flood control,     (mentioned in 75% of focus groups and 16% of indi-
storm protection, option values, existence values, and    vidual interviews) happens to be an occasional and
bequest values. A daunting set of services to have to    contentious phenomenon in the lagoon. The recent
include in a particular valuation effort. However, the    construction of a duck habitat restoration dike by
findings show that by using qualitative methods, bene-    Ducks Unlimited and activities of the Mexican Navy,
ficiaries can help researchers narrow the set ecological   according to participants, have resulted in drastic
services to those most relevant for study.          curtailment of the once annual or biannual inundation
  In Chelém, the focus groups and individual        of shrimp in the lagoon. The data show that it is 16
interviews left no doubt that lagoon fishing (espe-      times more likely that shrimp collection be raised in
cially for “chivita”, crab, and shrimp) is of utmost     focus groups than individual interviews. The lower
importance to local people. A few nonconsumptive       frequency that individual interviews raised the topic
uses and possible nonuse values were articulated       of shrimp collection may well reflect the decreased
by respondents (nongame species and the beauty of      role of shrimp collection in beneficiaries’ use of
the ecosystem) while the relative insignificance of      the mangrove lagoon. The topic’s high frequency of
ecosystem storm protection services and wood collec-     discussion in focus groups may reflect a collective
tion was also made apparent. Furthermore, the small     need or desire of individuals to process or air feelings
role that mangrove wood and wood collection plays      associated with the loss of this service.
in the lives of local beneficiaries in Chelém Lagoon       Similarly, a statistically significant divergence
contrasts with the findings of Kovacs (1999). Together,    between focus group and individual interview data was
the use of focus groups and individual interviews      observed in salt extraction data. At one time, indi-
allowed the researcher to identify those service most    viduals in the region could construct salt ponds, flood
relevant to local beneficiaries and to further investiga-   them with seawater, allow the water to evaporate, and
tion.                            then collect and sell crystallized sea salt. However, the
                               area’s lucrative salt mining business has been defunct
Significance of differences
                               for years. The change followed the flooding and ipso
                               facto enlarging of Chelém Lagoon that resulting when
The most frequent and least frequent ecosystem
                               the Mexican government dredged and constructed a
services raised using the two methods were not statis-
                               safe harbor and naval station in the lagoon in the late
tically different across methods. This seems to imply
                               1960s and early 1970s (Paré and Fraga, 1994). Like
that, regardless of method, participants recall and
                               the shrimp collection data, individual interviews raised
articulate common wetland ecosystem uses equally
                               salt extraction as an ecosystem service significantly
at the extremes of usage or importance in focus
                               less often than focus groups. It is about 19 times more
groups and individual interviews. For example, the
                               likely that a focus group raise salt extraction than an
collection of chivita (Melongena melongena) from
                               individual interview raise that same topic. People’s
the muddy bottom of Chelém Lagoon has become
                               discussion of the lagoon in groups seemed to trigger
the predominant subsistence strategy for the regions’
communities.6 Chivita collection has replaced more      discussion of the loss of ecosystem services, like salt
                               extraction.
conventional lagoon fishing and collection of crab as
                                 Apparently, focus group data can leave researchers
the most important ecosystem service. Therefore, it
                               with an impression about the significance of a resource
is no surprise that more than 90% of both the focus
                                                       177
                     IDENTIFYING ECOSYSTEM SERVICES

service that substantially differs from the impres-    cates that the notion of a nonconsumptive or nonuse
sion left by individual interview data. The differences  ecosystem service may be difficult for individuals to
observed in the frequencies of the discussions of crab   conceptualize and associate with an ecosystem without
collection, salt extraction, and shrimp fishing activ-   the benefit of a dynamic exchange of information (e.g.,
ities illustrate that specific components of inclusive   informational priming in a survey instrument). The
use values (e.g., Lagoon Fishing) are more likely     use of multiple qualitative methods would seem valu-
to be raised in focus groups rather than individual    able to researchers charged with the task of designing
interviews.                        a study or instrument addressing beneficiaries’ stated
                              preferences or values for nonconsumptive use services
Better understanding from qualitative methods       and nonuse services associated with complex ecosys-
                              tems.
There was not obvious difference in the frequency that
groups and individuals raise chivita collection or the
broader discussion topic of lagoon fishing. It seems    Conclusion
that virtually every family in the two communities,
at one time or another has adopted chivita collection   This study demonstrates that use of multiple quali-
as part of their subsistence survival strategy. Further-  tative methods can help researchers develop a
more, it is common for almost everyone in the area to   more complete understanding of beneficiaries’ natural
refer to himself or herself as a “pescador” (fisherman).  resource values. Reliance upon one qualitative method,
This despite the fact that many of these individuals    focus groups or individual interviews, would have
provide for themselves and their families by working    provided researchers with a less than complete under-
in nearby factories or doing construction work. Not    standing of beneficiaries’ uses, perceptions, and values
only do individuals perceive themselves as fisher-     associated with their shared mangrove ecosystem. This
people, it was learned throughout the groups and inter-  study shows the two qualitative research methods to be
views that respondents include chivita collection, crab  complementary, not substitute, methods for learning
and shrimp collection together with line and net fishing  about ecosystem services.
for other species when speaking about lagoon fishing.      The study demonstrates the value of using indi-
What makes this especially important, is that local    vidual interviews in addition to focus groups in the
researchers from nearby Mérida working on coastal     economic valuation study design process. Such inter-
zone management in the region were surprised to learn   views can be used to evaluate the effectiveness of
of the extent to which the respondents relied upon     the instrument at communicating complex information
chivita collection. It was their belief that chivita was  as well as to judge respondents’ ability to meaning-
a minor component of residents’ subsistence strategy    fully undertake the requisite valuation tasks. Incorpo-
and that near-shore fishing in the gulf was the predom-   rating both focus groups and individual interviews into
inant occupation in the area.               the conceptualization and design phase of valuation
  The researcher learned that unfortunately as one    studies seems capable of shaping ecosystem valuation
respondent put it,                     research so that it is more concerned with “what people
                              value.”
 We used to make a living fishing in the sea . . . Now
                                The incorporation of statistical examination of
 you can’t make a profit more than 2–3 months
                              focus group and individual interview data on
 from fishing in the sea . . . The same problem is also
                              ecosystem services illustrates that the two methods
 happening in the estuary, it used to be that you
                              generate different ecosystem service data. The find-
 could take all the crab you wanted. Now only the
                              ings suggest that focus group ecosystem service data
 small ones are around . . . While some try to work
                              reflect differences that may be attributable to dynamic
 elsewhere, people sustain their families with chivita
                              processing of information. This finding is in line with
 from the wetland (Transcript 18).
                              others’ research that shows increased interaction and
                              exchange of information improves respondents’ under-
  The individual interview data and the focus
                              standing of complex ideas (Schwarz, 1997; Schwarz
group data about beneficiaries’ uses and perceptions
                              and Sudman, 1995; Sudman et al., 1996). The system-
of Chelém Lagoon services appear to be compli-
                              atic statistical analysis of individual interview and
mentary. While both methods revealed information
                              focus group data can provide an empirical basis for
about ecosystem services, the relative weight that
                              better understanding of ecosystem services and their
each of the services received differed by method. For
                              value to respondents.
example, ecosystem beauty was raised in every focus
group. However, only 11 percent of individual inter-
viewees raised ecosystem beauty. This contrast indi-
178                         MICHAEL D. KAPLOWITZ

Notes                               Bryman, A. (1988). Quantity and Quality in Social Research.
                                   London: Unwin Hyman.
                                  Carson, R. T. (1998). “Valuation of tropical rainforests: Philo-
1. Contingent valuation (CV) studies elicit economic values
                                   sophical and practical issues in the use of contingent valu-
  for environmental amenities and natural resources using
                                   ation,” Ecological Economics 24: 15–29.
  carefully designed and administered surveys. CV studies
                                  Carson, R. T. and R. C. Mitchell (1993). “The issue of scope
  are one type of stated-preference approach researchers use
                                   in contingent valuation,” American Journal of Agricultural
  to reveal how individuals value environmnetal and natural
                                   Economics 75: 1263–1267.
  resources.
                                  Carson, R. T., W. M. Hanemann, R. J. Kopp, A. Krosnick, R.
2. Focus groups are carefully planned discussions designed
                                   C. Mitchell, S. Presser, P. A. Ruud, and V. K. Smith (1994).
  to learn about subjects’ perceptions on a defined area
                                   Prospective Interim Lost Use Value Due to DDT and PCB
  of interest in a permissive, nonthreatening environment.
                                   Contamination in the Southern California Bight. La Jolla,
  They are conducted by a skilled moderator who follows
                                   California: Natural Resource Damage Assessment, Inc.
  a discussion guide and involve as few to as many as 12
                                  Chilton, S., T. Burton, M. Jones, and G. Loomes (1998).
  informants.
                                   “A qualitative examination of preference reversals,” The 1st
3. Individual interviews (also called unstructured, explor-
                                   World Congress of Environmental and Resource Economists,
  atory, intensive, in-depth, and depth interviews) are guided
                                   June 25–27. Venice, Italy.
  conversations whose goal is to elicit from interviewees
                                  Chilton, S. M. and W. G. Hutchinson (1999). “Exploring diver-
  (also called informants) rich, detailed materials that can
                                   gence between respondent and researcher definitions of the
  be used in qualitative analysis. The interviewer used the
                                   goods in contingent valuation studies,” Journal of Agricul-
  same discussion guide as used in focus groups to guide the
                                   tural Economics 50: 1–16.
  one-on-one conversations.
                                  Clark, J. R. (1996). Coastal Zone Management Handbook. Boca
4. Dr. David Morgan is a highly regarded and widely
                                   Raton, Florida: CRC Press.
  published focus group researcher. He is a Professor in the
                                  De Jong, M. and P. J. Schellens (1998). “Focus groups
  Institute on Aging and the Department of Urban Studies and
                                   or individual interview? A comparison of text evaluation
  Planning at Portland State University. His works include
                                   approaches,” Technical Communication 45: 77–88.
  such classics as Focus Groups as Qualitative Research
                                  Farnsworth, E. J. and A. M. Ellison (1997). “The global
  (1988) and The Focus Group Kit (1998).
                                   conservation status of mangroves,” Ambio 26: 328–334.
5. Dr. Robert Weiss is Director of the Work and Family
                                  Fredricks, M. and S. I. Miller (1997). “Some brief notes on
  Research Unit and Professor at the University of Massachu-
                                   the ‘unfinished business’ of qualitative inquiry,” Quality and
  setts. Weiss is renown as a qualitative researcher and the
                                   Quantity 31: 1–13.
  author of Learning from Strangers: The Art and Method of
                                  Freeman, A. M. (1993). The Measurement of Environmental
  Qualitative Interview Studies (1994).
                                   and Resource Values. Washington, DC: Resources for the
6. Chivita (Melongena melongena) is a small mollusk found
                                   Future.
  in the mud flats on estuaries. It is also known as a West
                                  Gray, J. H. and I. L. Denstein (1998). “Integrating quantitative
  Indian Crown Conch.
                                   and qualitative analysis using latent and manifest variables,”
                                   Quality and Quantity 32: 419–431.
                                  Hamilton, L. and S. Snedaker (1984). Handbook for Mangrove
                                   Area Management. Hawaii: East/West Center.
References
                                  Hamilton, L., J. Dixon, and G. Owen Miller (1989).
                                   “Mangroves forests: An undervalued resource of the land and
Arrow, K., R. Solow, E. Leamer, P. Portney, R. Rader, and H.
                                   sea,” in E. M. Borgese, N. Ginsburg, and J. R. Morgan (eds.),
 Schuman (1993). “Report of the NOAA panel on contingent
                                   Ocean Yearbook 8. Chicago: University of Chicago Press.
 valuation,” Federal Register 58: 4601–4614.
                                  Hirsch, D. and A. Mauser (1992). The Economic Values of
Aylward, B. and E. B. Barbier (1992). “Valuing environmental
                                   Mangroves: Two Case Studies – Mida Creek and Funzi Bay,
 functions in developing countries,” Biodiversity and Conser-
                                   August–December. Amsterdam: University of Amsterdam.
 vation 1: 34–50.
                                  Hutchinson, W. G., S. M. Chilton, and J. Davis (1995).
Bann, C. (1997). An Economic Analysis of Alternative
                                   “Measuring non-use value of environmental goods using
 Mangrove Management Strategies in Koh Kong Province,
                                   the contingent valuation method: Problems of information
 Cambodia, November. Singapore: Economy and Environ-
                                   and cognition and the application of cognitive questionnaire
 ment Program for South East Asia.
                                   design methods,” Journal of Agricultural Economics 46:
Barbier, E. B. (1994). “Valuing environmental functions: Trop-
                                   97–112.
 ical wetlands,” Land Economics 70: 155–173.
                                  Instituto Nacional de Estadística Geografía e Informática
Barbier, E. B., M. Acreman, and D. Knowler (1997). Economic
                                   (INEGI) (1992). Yucatán-resultados definitivos: Datos por
 Valuation of Wetlands: A Guide for Policy Makers and Plan-
                                   AGEB urbana. XI censo general de poblacion vivienad, 1990.
 ners. Cambridge, UK: Ramsar Convention Bureau, Depart-
                                   Aguascalientes, Mexico: Instituto Nacional de Estadística,
 ment of Environmental Economics and Management, Univer-
                                   Geografía e Informática.
 sity of York.
                                  Janssen, R. and J. E. Padilla (1996). Valuation and Evaluation of
Bennet, E. L. and C. J. Reynolds (1993). “The value of a
                                   Management Alternatives for the Pagbilao Mangrove Forest,
 mangrove area in Sarawak,” Biodiversity and Conservation
                                   October. Amsterdam: Institute for Environmental Studies.
 2: 359–375.
                                                                179
                        IDENTIFYING ECOSYSTEM SERVICES

Kaplowitz, M. D. (1998). “Research intentions and respondent     Perrings, C. (1995). “The economic value of biodiversity”
 perceptions: Identifying a nexus in valuing mangrove ecosys-     in Heywood, V.H. (ed.), Global Biodiversity Assessment.
 tems,” Fifth Biennial Meeting of the International Society for    Cambridge: Cambridge University Press.
 Ecological Economics, November 15–19. Santiago, Chile.       Ruitenbeek, H. J. (1992). Mangrove Management: An
Kaplowitz, M. D. (1999). “Conflicting agendas in the Yucatán,”     Economic Analysis of Management Options with a Focus on
 International Review of Comparative Public Policy 11: 141–      Bintuni Bay, Irian Jaya. Jakarta and Halifax: Environmental
 156.                                 Management Development in Indonesia Project (EMDI) and
Kaplowitz, M. D. and J. P. Hoehn (1998). “A total value hypoth-    Dalhousie University.
 esis for ecosystem valuation in a cross-cultural setting,” First  Schkade, D. A. and J. W. Payne (1994). “How people respond to
 World Congress of Environmental and Resource Economists,       contingent valuation questions: A verbal protocol analysis of
 June 25–27. Venice, Italy.                      willingness to pay for an environmental regulation,” Journal
Kitzinger, J. (1994a). “The methodology of focus groups:        of Environmental Economics and Management 26: 88–109.
 The importance of interaction between research participants,”   Schwarz, N. (1997). “Cognition, communication, and survey
 Sociology of Health and Illness 16: 103–121.             measurement,” in R. J. Kopp, W. Pommerehne, and N.
Kitzinger, J. (1994b). “Focus groups: Method or madness,” in      Schwarz (eds.), Determining the Value of Non-Marketed
 Boulton, M. (ed.), Challenge and Innovation: Methodolog-       Goods. Boston: Kluwer-Nijhoff.
 ical Advances in Social Research on HIV/AIDS. New York:      Schwarz, N. and S. Sudman (1995). Answering Questions:
 Taylor and Francis.                         Methodology for Determining Cognitive and Communicative
Kovacs, J. M. (1999). “Assessing mangrove uses at the local      Processes in Survey Research. San Francisco: Jossey-Bass
 scale,” Landscape and Urban Planning 43: 201–208.          Publishers.
Krippendorff, K. (1980). Content Analysis: An Introduction      Spaninks, F. and P. van Beukering (1997). Economic Valuation
 to Its Methodology. Beverly Hills, California: Sage Public-     of Mangrove Ecosystems: Potential and Limitations, July.
 ations.                               Amsterdam: Institute for Environmental Studies.
Krueger, R. A. (1994). Focus Groups: A Practical Guide for      Strauss, A. and J. Corbin (1990). Basic of Qualitative Research:
 Applied Research, 2nd edn. Thousand Oaks, California: Sage      Grounded Theory Procedures and Techniques. Newbury
 Publications.                            Park, California: Sage Publications.
Kunstadter, P., E. C. F. Bird, and S. Sabhasri, (1985). Man     Sudman, S., N. B. Bradburn, and N. Schwarz (1996). Thinking
 in the Mangroves: The Socio-Economic Situation of Human       About Answers: The Application of Cognitive Processes to
 Settlements in Mangrove Forests. Tokyo: United Nations        Survey Methodology. San Francisco: Jossey-Bass Publishers.
 University.                            Swallow, S., M. Spencer, C. Miller, P. Paton, R. Deegen, L.
Mandondo, A. (1997). “Trees and spaces as emotion and         Whinstanley, and J. Shogren (1998). “Methods and applica-
 norm laden components of local ecosystems in Nyamaropa        tions for ecosystem valuation: A collage,” in B. Kanninen
 communal land, Nyanga District, Zimbabwe,” Agriculture        (ed.), Proceedings of the first Workshop in the Environmental
 and Human Values 14: 353–372.                    Policy and Economics Workshop Series. Washington, DC:
Maxwell, J. A. (1996). Qualitative Research Design: An Inter-     Office of Research and Development and Office of Policy, US
 active Approach. Thousand Oaks, California: Sage Publica-      Environmental Protection Agency.
 tions.                               Talawar, S. and R. E. Rhoades (1998). “Scientific and local
Mitchell, R. C. and R. T. Carson (1989). Using Surveys to       classification and management of soils,” Agriculture and
 Value Public Goods: The Contingent Valuation Method.         Human Values 15: 3–14.
 Washington, DC: Resources for the Future.             Weber, R. P. (1990). “Basic content analysis,” in M. S. Lewis-
Mitsch, W. J. and J. G. Gosselink (1993). Wetlands, 2nd edn.      Beck (ed.), Quantitative Applications in the Social Sciences,
 New York: Van Nostrand Reinhold.                   2nd edn. Newbury Park, California: Sage.
Morgan, D. (1997). Focus Groups as Qualitative Research 2nd     Weiss, R. S. (1994). Learning from Strangers: The Art and
 edn. Thousand Oaks, California: Sage Publications.          Method of Qualitative Interview Studies. New York: The Free
Morgan, D. L. (1988). Focus Groups as Qualitative Research.      Press.
 Newbury Park, California: Sage Publications.
Morgan, D. L. (1996). “Focus groups,” in J. Hagan and K. S.     Address for correspondence: Michael D. Kaplowitz, J.D.,
 Cook (eds.), Annual Review of Sociology. Palo Alto: Annual     Ph.D., Michigan State University, 311A Natural Resources,
 Reviews.                              East Lansing, MI 48824, USA
Morgan, D. L., R. A. Krueger, A. U. Scannell, and J. A. King     Phone: +1-517-355-0101; Fax: +1-517-353-8994;
 (1998). Focus Group Kit. Thousand Oaks, California: SAGE      E-mail: kaplowit@msu.edu
 Publications.
Paré, L. and J. Fraga (1994). La costa de Yucatán: Desarrollo
 y vulnerabilidad ambiental. Mexico: Universidad Nacional
 Autónoma de México. Instituto de Investigaciones Sociales.