Content uploaded by Leonard Manda
Author content
All content in this area was uploaded by Leonard Manda on Aug 19, 2023
Content may be subject to copyright.
Available via license: CC BY
Content may be subject to copyright.
Co-management brings hope
for effective biodiversity
conservation and socio-
economic development in Vwaza
Marsh Wildlife Reserve in Malawi
Leonard Manda
1,2
*, Kolawole
´Valère Salako
3
, Andrew Kataya
4
,
Sèdjro A. T. Affossogbe
3
, Dalo Njera
5
, William O. Mgoola
4
,
Achille Ephrem Assogbadjo
6
and Brice Sinsin
6
1
Biological Sciences Department, Mzuzu University, Mzuzu, Malawi,
2
African Centre of Excellence for
Neglected and Underutilized Biodiversity (ACENUB), Mzuzu University, Mzuzu, Malawi,
3
Laboratoire
de Biomathe
´matiques et d’Estimations Forestières (LABEF), Universite
´d’Abomey-Calavi, Cotonou,
Benin,
4
Department of National Parks and Wildlife, Lilongwe, Malawi,
5
Forestry Department, Mzuzu
University, Mzuzu, Malawi,
6
Laboratoire d’e
´cologie applique
´e (LEA), Universit´e d’Abomey-Calavi,
Cotonou, Benin
Co-management has been widely promoted in protected area management on
the premise that it may simultaneously enhance biodiversity conservation
outcomes and improve livelihoods of the park-border communities. However,
the success of this management approach remains a growing debate raising the
question of its effectiveness. To contribute to this debate, we used local
community perceptions and secondary ecological data to assess the extent to
which co-management has effectively contributed to biodiversity conservation
and socio-economic development outcomes in the Vwaza Marsh Wildlife
Reserve. Face-to-face individual interviews using a semi-structured
questionnaire were used to collect data on the perceptions of co-
management from 160 purposively selected heads of households. A desk
study was used to collect data on trends in animal populations, animal
mortality, and prohibited activities including incidences of poaching for the
past 30 years (pre-and post-introduction of co-management). Results showed
that local communities have positive perceptions of the conservation work in the
Vwaza Marsh Wildlife Reserve. Further, there was an improved people-park
relationship and a recovery of animal populations in the reserve after the
introduction of co-management. These findings point to the success of co-
management in the area. However, misunderstandings over revenue sharing
were still a thorny issue, somehow creating mistrust between parties. We
concluded that while it may still be early to achieve more demonstrable
conservation outcomes, co-management appears to bring hope for effective
Frontiers in Conservation Science frontiersin.org01
OPEN ACCESS
EDITED BY
Rebecca Kariuki,
Nelson Mandela African Institution of
Science and Technology, Tanzania
REVIEWED BY
Matthew Grainger,
Norwegian Institute for Nature Research
(NINA), Norway
Inza Kone
´,
Swiss Centre for Scientific Research,
Co
ˆte d’Ivoire
Jean Huge
´,
Open University of the Netherlands,
Netherlands
*CORRESPONDENCE
Leonard Manda
lmanda8@gmail.com;
manda.l@mzuni.ac.mw
RECEIVED 14 December 2022
ACCEPTED 06 June 2023
PUBLISHED 16 August 2023
CITATION
Manda L, Salako KV, Kataya A,
Affossogbe SAT, Njera D, Mgoola WO,
Assogbadjo AE and Sinsin B (2023)
Co-management brings hope for effective
biodiversity conservation and socio-
economic development in Vwaza Marsh
Wildlife Reserve in Malawi.
Front. Conserv. Sci. 4:1124142.
doi: 10.3389/fcosc.2023.1124142
COPYRIGHT
© 2023 Manda, Salako, Kataya, Affossogbe,
Njera, Mgoola, Assogbadjo and Sinsin. This is
an open-access article distributed under the
terms of the Creative Commons Attribution
License (CC BY). The use, distribution or
reproduction in other forums is permitted,
provided the original author(s) and the
copyright owner(s) are credited and that
the original publication in this journal is
cited, in accordance with accepted
academic practice. No use, distribution or
reproduction is permitted which does not
comply with these terms.
TYPE Original Research
PUBLISHED 16 August 2023
DOI 10.3389/fcosc.2023.1124142
biodiversity conservation and socio-economic development in the Vwaza Marsh
Wildlife Reserve. Participatory evaluation of co-management involving key
stakeholders is recommended in the Vwaza Marsh Wildlife Reserve based on
the findings of this study and lessons learnt over the years.
KEYWORDS
community perceptions, collaborative management, protected area management,
community-based natural resources management, rural livelihood
Introduction
Biodiversity conservation in protected areas (PAs) is evolving
from coercive to more people-centred approaches. Since the 1980s,
there has been a growing recognition of the critical role of local
communities in the management of PAs, with many studies
emphasising the need for more inclusive and people-oriented
approaches (Wells and McShane, 2004;Dawson et al., 2021;
Gatiso et al., 2022;Ma et al., 2022). This has been a reaction to
the failures of exclusionary conservation approaches to achieve
conservation, economic, and development goals of PAs (Berkes,
2004;Watson et al., 2014). Indeed, PA managers, especially in
developing countries where socio-economic and cultural factors
predispose some people to over-rely on the immediate environment
to meet their daily livelihood needs (Shackleton and de Vos, 2022;
Zhang et al., 2022), are continuously faced with the challenge of
balancing multiple goals of PAs (McShane et al., 2011;Allendorf
and Yang, 2013;Gidebo, 2022;Yousefpour et al., 2022).
The management challenges faced are in part emanating from
the historical context of biodiversity conservation in PAs. For a long
time, PA management has mostly been the sole responsibility of
national governments, modelled on Yellowstone National Park, the
first PA established in 1872 (Andrade and Rhodes, 2012) and based
on values that have less regard for local people and their livelihoods
(Andrade and Rhodes, 2012;Sarkki et al., 2015;Umar, 2018). The
exclusion further meant the prohibition of local communities from
accessing wildlife resources that have been supporting their very
lives since time immemorial (Berkes, 2004;Berkes, 2009;Sarkki
et al., 2015). Consequently, antagonism and conflicts between local
communities and government agencies have not been uncommon.
The consequences have been widespread rebellion, hostility, rule-
breaking, and increased anthropogenic pressure in PAs such as
poaching and deforestation (Anthony and Szabo, 2011;Spracklen
et al., 2015;Gray et al., 2018;Geldmann et al., 2019;Matseketsa
et al., 2019). This may essentially reduce the quality and quantity of
ecosystem services provided by PAs to people, ultimately leading to
the disruption of local livelihoods. These challenges, coupled with
many mixed reports on the success of PAs (Leverington et al., 2010;
Muhumuza and Balkwill, 2013;Watson et al., 2014;Oldekop et al.,
2016;Singh et al., 2019;Robson et al., 2022), have continually put
into question the effectiveness of PAs as conservation and
development cornerstones (Anthony and Szabo, 2011).
Collaborative management or Co-management has been
defined as the sharing of power and responsibility between
government or state and local resource users (Berkes, 2009). In
this governance and management system of natural resources, each
party has roles and obligations, and is accountable to each other. It
has been widely promoted on the premise that it may enhance
biodiversity conservation outcomes while also serving the interests
of social and economic development (Berkes, 2004;Berkes, 2009). It
is believed that biodiversity conservation efforts would become
more successful if communities that carry the burden of living
with wildlife were involved in the conservation work and benefited
from it (Berkes, 2004;Anthony, 2007). In this respect, Sub-Saharan
Africa has been one of the experimental regions for co-management
(Zulu, 2013;Umar, 2018). However, since its introduction, its
effectiveness remains debatable and mixed results are reported in
the literature on its success across different ecosystems (Fabricius
and Collins, 2007;Campbell et al., 2013;Oldekop et al., 2016;
Kabeer et al., 2018;Rahman, 2022). Moreover, both access and
restrictions to wild resources seem to escalate extraction of
prohibited resources, varying with socio-economic and cultural
factors (Shova and Hubacek, 2011;Muhumuza and Balkwill,
2013). This suggests the need for context-specific evaluation of
co-management as a conservation approach.
This study was aimed at evaluating the extent to which co-
management has effectively contributed to biodiversity
conservation and socio-economic development outcomes in the
Vwaza Marsh Wildlife Reserve (VMWR) in Malawi. According to
Pressey et al. (2021), conservation impact of a PA is the sum of
avoided biodiversity loss and promoted recovery relative to
outcomes without protection. We therefore examined the
effectiveness of co-management on biodiversity conservation by
quantifying changes in the abundance and mortality of animal
species, and illegal activities (Sarkar et al., 2022), On the other hand,
effectiveness on the socio-economic development outcomes was
investigated by documenting the goods and services that local
communities obtain from the reserve, and other perceived
benefits including development projects attributable to co-
management. The VMWR became one of the pilot sites for co-
management in the early 2000s through a donor-funded project
(Zulu, 2012) concomitant with the global changes in biodiversity
conservation approaches. The Department of National Parks and
Wildlife (DNPW) entered into a governance and management
Manda et al. 10.3389/fcosc.2023.1124142
Frontiers in Conservation Science frontiersin.org02
working relationship with local communities bordering the VNWR
to improve the conservation goals and human well-being in the
area. However, as with other PAs (Muhumuza and Balkwill, 2013;
Oldekop et al., 2016), limited evidence of the success of co-
management since its introduction in Malawi partly undermines
the upscaling of this management approach. Several previous
studies have often used perceptions and attitudes of local
communities bordering PAs as a proxy for evaluating the
effectiveness of co-management (e.g., Bennett, 2016;Allendorf,
2020;Katswera et al., 2022;Rampheri et al., 2022)bywayof
focusing on the park-people relationships and benefits accrued to
local communities.
It is widely recognised that attitudes and perceptions of the
park-border community towards conservation work can provide
insights into their behaviour, compliance with conservation
regulations, willingness to participate, and acceptability of
conservation management (Bennett, 2016;Katswera et al., 2022).
As such, understanding these perceptions is crucial for developing
successful, long-term conservation and management strategies
(Ellwanger et al., 2015). Perceptions can also be used as a starting
point for developing the park-people relationship which is crucial
for local community commitment to effectively participate in
conservation work (Vodouheet al., 2010). For PAs where there
has been a park management shift from a strictly state-controlled
approach to a governance and management agreement with the
bordering local communities such as the VMWR, local community
perceptions may also help identify conservation conflicts that may
inform conservation decision-making (Janssens et al., 2022). More
importantly, perceptions may also aid in assessing the performance
of conservation work and eventually help in the development of an
adaptive co-management approach that could lead to improved
conservation and development outcomes (Bennett, 2016;Abukari
and Mwalyosi, 2020). It has further been suggested that quantifying
perceptions and attitudes towards ecosystem conservation of local
communities can help assess the success of current conservation
goals and actions and inform future practices (Xu et al., 2022).
Positive perceptions in this regard are indicative of a successful
conservation approach (Bennett, 2016;Allendorf, 2020).
Already, community perceptions have been used to facilitate
proper conservation in PAs (see Liu et al., 2010;Allendorf et al.,
2012). However, rarely has a single study looked at both ecological
and social data collected inside and outside PAs (but see Sarkar
et al., 2022), thus undermining more conclusive evidence of what is
working, what is not, and why (Pendleton et al., 2018). Moreover,
although elements of co-management have been examined
elsewhere in Malawi (Gordon, 2017), a systematic assessment of
the perceptions and ecological conservation outcomes before or
after the change in the management approach in the VMWR is
opaque in the literature. Such information would be important to
guide management decisions for the improvement of co-
management in the VMWR. It has been suggested that local
people’s knowledge can have a valid relevance in PA management
and could assist in developing effective conservation strategies, as
well as in the overall improvement of local socio-ecological systems
(Afriyie et al., 2021).
Previous studies in the VMWR that tackled elements of co-
management and/or local community resource use include those
that focused on spatial patterns of resource use and prohibited
activities (McShane, 1990), resource use by communities based on
permit records (O’Sullivan, 2019), and patterns of bushmeat
hunting and consumption (van Velden et al., 2020). In this study,
we used local community knowledge, perceptions, and secondary
ecological data (animal population estimates and animal mortality)
to assess the extent to which co-management has contributed to the
effectiveness of biodiversity conservation and development
outcomes in the VMWR. We made the following assumptions:
(1) that co-management had provided local communities with
increased access to resources important for improved livelihood
in the VMWR (Muhumuza and Balkwill, 2013;Woodhouse et al.,
2022); (2) that co-management had brought a positive perception of
biodiversity conservation among local communities in the VMWR
(Woodhouse et al., 2022); (3) that co-management had reduced
extraction of prohibited resources in the VMWR (Woodhouse et al.,
2022); (4) that co-management had provided local communities
with a problem-solving platform to improve their quality of life
(Berkes, 2009); and (5) that choices, knowledge, and perceptions of
local communities of biodiversity conservation and sustainable use
were influenced by their socio-demographic factors (age, gender,
level of formal education, distance from the reserve boundary, and
membership to a conservation institution) (Shibia, 2010;
Muhumuza and Balkwill, 2013).
Materials and methods
Study site and context
Vwaza Marsh Wildlife Reserve (VMWR) is located at 33° 28’E
and 11° 00’S, covering an area of about 986 km
2
. The reserve is
found in Rumphi District and lies on the watershed between Lake
Malawi and the Luangwa valley, with the western part of its
boundary forming the Malawi-Zambia border (Mgoola and
Msiska, 2017)(Figure 1). The VMWR has an annual mean
temperature range of 16 - 28°C; altitude varying from 1000 to
1660 m a.s.l, and a mean annual rainfall range of 700 - 1100 mm,
often falling from November to April. Temperatures are lowest
during the early dry season (May to August) and highest in the late
dry season (September to November) (Mgoola and Msiska, 2017).
The major characteristic features of the reserve including
microhabitat types, and animal and plant communities are
reported elsewhere (Engel et al., 2013;Mgoola and Msiska, 2017;
Sichinga, 2019). The reserve contains the most extensive wetlands
in the plateau area of Malawi and is bordered by a community that
practices mixed farming of crops and livestock (McShane, 1990;
Mgoola and Msiska, 2017). The human population density has
grown from a mean density of about 28 persons/km
2
in 1990
(McShane, 1990) to about 63 persons/km
2
in 2020 (van Velden
et al., 2020). The park-border community is mainly composed of
the Tumbuka, Phoka, and Ngoni ethnic tribes, and are traditionally
hunting tribes (van Velden et al., 2020). As with many other areas in
Manda et al. 10.3389/fcosc.2023.1124142
Frontiers in Conservation Science frontiersin.org03
Malawi, poverty and food insecurity are some of the major
challenges. According to the World Bank (Caruso and Sosa,
2022), nearly 73.5% of the population in Malawi lives on US$
1.90 per day and about 2.3 million people (approx. 12% of the
country’s population) are perpetually food insecure.
The historical context of the VMWR is well elaborated by
McShane (1990). Suffice it to say that efforts to protect the area
began in 1941 with the current size and status established in 1977
characterised by the eviction of local communities who had settled
in the reserve since the Early Iron Age (McShane, 1990). It is further
reported that Vwaza attracted several settlers owing to the species
richness of large-sized mammals including elephants. As such,
Vwaza became a route of trade, especially in ivory connecting yet
another elephant-rich area of eastern Zambia to Lake Malawi
through to the Indian coast (McShane, 1990).
Elements of co-management in the VMWR began in the early
1990s following the Wildlands and Human Needs Project
championed by the World Wildlife Fund (WWF) (McShane,
1990). However, as was the case with many countries in the
region, the lack of relevant supporting policy slowed down
progress (Zulu, 2013). The management approach was revisited
in 1996 following Malawi’s attainment of multiparty democracy in
1994 and the adoption of a new constitution that emphasises the
devolution of power to the people. The decentralisation process that
started thereafter (Malawi Government, 1998)providedmore
thrust for local community involvement in decision-making
processes including decisions regarding the management of
natural resources.
Subsequently, in 2000, the DNPW entered into a governance and
management agreement with the local communities living within a
5 km distance from the gazetted boundary of the VMWR. Among the
items included in the agreement are: (i) the Resource Use Programme
(RUP), whereby members of the communities bordering the VMWR
are issued with permits allowing them to harvest prescribed resources
from the reserve within the 5km distance from the reserve border into
the park; (ii) a revenue sharing scheme, whereby communities receive a
fraction (currently at 25%) of the total revenue accrued from park entry
and concession fees for community development projects. In return,
communities are expected to perform several voluntary activities meant
to foster the conservation of biodiversity such as surveillance, ensuring
the protection of the wire fence, and reporting any suspected prohibited
activities. For governance purposes, Natural Resources Committees
(NRCs) were established at the village level. These are grouped into
zones that operate at the Traditional Authority level. The Nyika-Vwaza
Association (NVA) was subsequently established in 2003 as an
umbrella body to coordinate the affairs of the communities in their
respective zones and NRCs. The NVA encompasses two major and
closely located PAs in northern Malawi, Nyika National Park and
VMWR.Bythetimeofthisstudy,therewere60NRCsin12zones
against an initial 13 NRCs at the start of the programme in 2000.
FIGURE 1
Map of the Vwaza Marsh Wildlife Reserve showing locations of study villages.
Manda et al. 10.3389/fcosc.2023.1124142
Frontiers in Conservation Science frontiersin.org04
According to the Extension Officer at the VMWR, Veronica Mhango
(per. comm., April 2022), it was the communities themselves that
demanded the formation of new NRCs over the years after realising the
importance of being involved in the management and conservation of
natural resources under the co-management approach.
Sampling and data collection
We used a multi-stage sampling approach where we selected
five of the 12 zones and 20 of the 60 NRCs. A somewhat similar
approach was used in the area by van Velden et al. (2020). The
major differences from van Velden et al. (2020) are that we selected
heads of households within the 5km distance from the reserve
boundary that had lived in the area from the inception of co-
management (not less than 25 years) since the purpose of our study
was to assess the impact of the change in the conservation approach;
while van Velden et al. (2020) used individuals above 18 years old
from 231 households. Secondly, we only selected zones with
proportionally reported greater activities in the reserve (McShane,
1990;O’Sullivan, 2019). The zones considered were Bowe, Kazuni,
Lake Kazuni, Lusani, and Thunduwike. Finally, from the pool of
purposively selected key informants that met our criteria, we
randomly selected both men and women, and members and non-
members of the NRCs drawn from different villages.
We used face-to-face individual interviews to collect data. A
semi-structured questionnaire (composed of closed-and open-
ended questions) was used for the interview survey (Young et al.,
2018;Rampheri et al., 2022). According to Young et al. (2018),
semi-structured questionnaires are preferable as they take care of
the shortcomings associated with structured (or close-ended) and
unstructured (or open-ended) interviews. Indeed, closed-ended
questions allow for the comparison of responses between
respondents and for conducting quantitative analyses; while
open-ended questions provide respondents an opportunity to
freely express themselves and provide detailed responses
(Rampheri et al., 2022). Data were collected between January and
April 2022. But preliminary data that gave the impetus to this study
was collected in 2010 under the auspices of the Rufford Small
Grants for Nature Conservation.
We collected data on the socio-demographic attributes of
respondents. These included gender, ethnic tribe, age, residence
period, level of formal education, estimated distance from the
reserve boundary, and membership status to NRC (member or
non-member). To determine the prescribed resources that local
communities obtain from the VNMR, respondents were asked to
state the resources they directly obtain from the reserve under the
RUP in the co-management arrangement. To examine the
perceptions of local communities towards conservation, we first
assessed (i) the perceived importance and/or benefit of the co-
management approach compared to the previous approach; (ii) the
willingness of non-members to join NRCs including their
motivating reasons as an indicator of local community willingness
to participate in conservation work; and (iii) perceived importance
of conserving biodiversity in the VMWR about ecosystem services.
We also investigated knowledge of their roles in the co-
management arrangement. Finally, based on their experiences
and/or lessons learnt from the co-management arrangement to
date, respondents were asked to suggest actions for improvement on
their part as communities and on the part of the DNPW.
We used secondary ecological data from annual reports at the
DNPW as a proxy to examine the effectiveness of ecological
conservation outcomes. Ecological data was on trends in animal
populations (1985 –2021 for 17 species), animal mortality (1988 –
2021 for 30 species), and prohibited activities (1993-2021). The
DNPW collects animal population data through annual surveys
using aerial transect counts followed by ground truthing using
ground transects. This is mostly done during the dry season when
visibility is good. Data collected includes animal species abundance
which provides trends and population estimates. Animal mortality
data is collected daily and reported every month during routine
patrols by the park staff. Some of the parameters collected include
species name, status of carcasses (including, cause of mortality if
possible), and GPS coordinates. Staff have cyber tracker gadgets
which enable them collect GPS coordinates, track their routes and
estimate coverage for monthly patrols. The collected data is
aggregated annually. As for the prohibited activities, the DNPW
in VMWR, like in other PAs in Malawi uses several parameters to
assess and categorise these activities. These include dead animals,
gunshots from poachers, snares encountered in the reserve, tree
cutting, and signs of illegal camps encountered in the reserve. Data
on prohibited activities is collected during routine daily and
monthly patrols, when dealing with problem animal control, and
when conducting annual ground animal surveys.
Approval to conduct this research was sought at different stages.
The study was conducted in conformity with the requirements of
Mzuzu University Research and Ethics Committee (MZUNIREC).
Approval was also granted by the Director of the DNPW.
Thereafter, verbal approval was sought from the leadership of the
Nyika Vwaza Association, and the Resource Use Committees and
local village heads in Vwaza. Moreover, participation by local
communities was also based on their willingness to participate
following a verbal free, prior informed consent.
Data analysis
Descriptive statistics were used for data analyses to describe
trends. Specifically, bar charts were used to illustrate the prescribed
resources (%) obtained from the studied communities in the
VMWR and the perceived importance of conserving biodiversity
in the VMWR. Pie charts were used to show local community
perceptions of the importance of co-management in the VMWR,
Further, various diversity indices including the Simpson (1-D),
Shannon (H) and Evenness (e^H/S) were computed to depict the
diversity and patterns of use of wild resources/NTFPs among the
five zones. For this purpose, we treated a particular wild resource/
NTFP (e.g., fruits, medicine, mushroom) as a “species”such that the
citation of a resource was considered its presence while non-citation
was its absence. Diversity indices have been used to show patterns of
Manda et al. 10.3389/fcosc.2023.1124142
Frontiers in Conservation Science frontiersin.org05
use and knowledge of non-timber forest products in several studies,
with evenness indicating the distribution of the resource (e.g., Ladio
and Lozada, 2004). For instance, higher evenness (values closer to 1)
indicates similar distribution of the resource in the area.
The population of animal species as estimated by the VMWR
was recorded as individuals per species per year of estimation and a
line graph was plotted to show the abundance trend for each species
(Western et al., 2009). The same was done for animal mortality. For
prohibited activities, the number of dead animals encountered per
year and the number of activities and/or events encountered were
also plotted in line graphs to show variations in trends. We built a
Sankey diagram of the suggestions made by communities using the
SankeyMATIC software (https://sankeymatic.com/build/) to show
their relative importance. Sankey diagrams are one of the visual
tools. They are used for illustrating categories of information and
how their flow relates to each other with proportional arrow
magnitudes (https://www.sankey-diagrams.com/). They have wide
applications in different disciplines including natural resource
management (e.g., Brodie et al., 2022). A correlation analysis was
also done to assess which of the suggestions were correlated. The
rationale for this was to identify convergent and non-convergent
suggestions which may reflect synergies and trade-offs. All
descriptive statistics were calculated in R software (R
Development Core Team, 2021).
Finally, we tested whether (and how) the socio-demographic
attributes of the respondents (gender, age, level of formal education,
distance from the reserve boundary, and membership status
(member vs non-member of NRC) influence community choices,
knowledge, and perceptions about conservation work in the
VMWR. The data consisted of two dependent binary variables for
reasons for participation in the community management (benefits/
incentives, and relationships), and three dependent binary variables
regarding knowledge of the roles of NRC Community members
(removing litter, community sensitisation, and removing snares or
traps). Five socio-demographic independent variables, including
gender (binary: male versus female), age (categorical with three
levels: young, adult, and old persons), level of formal education
(categorical with four levels: none, primary, secondary, tertiary),
distance from the reserve boundary (categorical with two levels:
short, and long), and membership status (two levels: member vs
non-member of NRC) were considered to assess their effects on the
dependent variables. One commonly used approach is to apply a
binary logistic/probit model on each binary dependent variable and
considering the five independent variables. However, this approach
is not efficient mainly because we have multiple dependent variables
which may be correlated. For instance, the response of an informant
concerning benefits or incentives could be linked to his/her
response concerning relationships. The multivariate probit (MVP)
model is a generalisation of the probit model used to estimate
several correlated binary outcomes jointly (Campbell et al., 2013;
Mittal and Mehar, 2016). The MVP accounts for correlated
responses when estimating the effect of predictors on dependent
variables and is more efficient than applying several univariate
logistic/probit models. The model was implemented in STATA
(version 15.1, STATA Corporation, College Station, TX, USA) with
the function mvprobit program. This function uses a simulated
maximum likelihood estimation approach to provide robust
parameter estimation (Cappellari and Jenkins, 2003).
Results
Socio-demographic characteristics of
the respondents
A total of 160 individuals (46.25% women: 53.75% men), mostly
belonging to the Tumbuka ethnic tribe (77.50%), with a few Ngoni
(11.88%) and two other tribes (Lambya and Chewa) participated in
this study. The majority of the respondents (58.13%) were under 40
years old and most had lived in the area for nearly 30 years
(56.26%), with farming as their main occupation (96.25%).
Slightly over half of the respondents (55.00%) had attained formal
education to the primary school level, and most of them (76.25%)
were residing within a 3km distance from the reserve boundary.
While 50% of the respondents belonged to various NRCs, only
42.5% of these had been members for up to five years (Table 1).
Resources legally obtained from the VMWR
by local communities
The respondents cited 11 prescribed wild resources (non-timber
forest products) that they obtained from the VMWR under the
RUP. Among the most cited resources were medicine (93%),
mushrooms (91%), and edible fruits (69%), while reeds (11%) and
sisal (1%) were among the least cited (Figure 2).
Among the five zones, three diversity indices [Simpson (1-D),
Shannon (H′), and Evenness (e^H′/S)consistentlyflagged
Thunduwike Zone (S=9;N= 234) as a zone where local
communities obtained comparably high diversity of wild
resources (non-timber forest products including fruits, medicine,
mushrooms) followed by Lake Kazuni Zone (S= 10; N=
173) (Table 2).
Perceptions of local communities of
conservation
Local communities’perceptions of the
importance of co-management in the VMWR
Compared to the era before co-management was introduced,
nearly all the respondents (99%; n = 160) expressed positive
perceptions towards co-management stating that things had
somehow improved following its introduction in the VMWR,
with only 1% indicating the opposite. Principally, respondents
indicated that co-management had: (1) improved the relationship
between communities and the park staff (99% of the respondents);
(2) provided local communities with legal access to prescribed
resources to complement their daily needs (98% of the
respondents); and (3) brought tangible benefits to the area
Manda et al. 10.3389/fcosc.2023.1124142
Frontiers in Conservation Science frontiersin.org06
including community development projects and household
livelihood interventions. Some of the community development
projects initiated in various areas included school blocks, bridges,
and boreholes. At the household level, interventions cited included
small livestock pass-on programmes and carbon payment schemes.
Willingness and motivating factors for
community participation in conservation work
About 83.75% (67 out of 80) of the non-members of the NRCs
that participated in the study expressed willingness to join and
participate in conservation work in the area. The respondents cited
TABLE 1 Socio-demographic characteristics of the respondents in the VMWR by Zone.
Parameter Zone
Bowe Kazuni Lake Kazuni Lusani Thunduwike Total
Gender
Female 14(43.75) 13(40.63) 18(50.26) 9(28.13) 20(62.50) 74(46.25)
Male 18(56.25) 19(59.38) 14(43.75) 23(71.88) 12(37.50) 86(53.75)
Tribe
Tumbuka 25(78.13) 24(75.00) 26(81.25) 25(78.13) 24(75.00) 124(77.50)
Ngoni 1(3.13) 7(21.88) 4(1250). 3(9.38) 4(12.50) 19(11.88)
Others 6(18.75) 1(3.13) 2(6.25) 4(12.50) 4(12.50) 17(10.63)
Age
Young: age≤40 years 19(59.38) 16(50.00) 15(46.88) 23(71.88) 20(762.50) 93(58.13)
Adult: 40 <age ≤60 years 12(37.50) 14(43.75) 15(46.88) 8(25.00) 8(25.00) 57(35.63)
Old: age> 60 years 1(3.13) 2(6.25) 2(6.25) 1(3.13) 4(12.50) 10(6.25)
Residential period
≤30 years 16(50.00) 12(37.50) 23(71.88) 16(50.00) 23(71.88) 90(56.25)
>30 years 16(50.00) 20(62.50) 9(28.13) 16(50.00) 9(28.13) 70(43.75)
Main occupation
Farmers 28(87.50) 30(93.75) 32(100.00) 32(100.00) 32(100.00) 154(96.25)
Others 4(12.50) 2(6.25) 0(0.00) 0(0.00) 0(0.00) 6(3.75)
Level of formal education
None 1(3.13) 1(3.13) 3(9.38) 0(0.00) 2(6.25) 7(4.38)
Primary 17(53.13) 23(78.88) 19(59.38) 13(40.63) 16(50.00) 88(55.00)
Secondary 12(37.50) 8(25.00) 10(31.25) 19(59.38) 13(40.63) 62(38.75)
Tertiary 2(6.25) 0(0.00) 0(0.00) 0(0.00) 1(3.13) 3(1.88)
Distance from reserve boundary
0 - 3 Km 18(56.25) 20(62.50) 2990.63) 31(96.88) 24(75.00) 122(76.25)
3.1 - 5 Km 14(43.75) 12(37.50) 3(9.38) 1(3.13) 8(25.00) 38(23.75)
Membership period to NRCs
Non-member 16(50.00) 16(50.00) 16(50.00) 16(50.00) 16(50.00) 80(50.00)
Member
1-5 years 14(43.75) 13(40.63) 13(40.63) 13(40.63) 15(46.88) 68(42.50)
6-10 years 0(0.00) 0(0.00) 0(0.00) 0(0.00) 0(0.00) 0(0.00)
11-15 years 1(3.13) 0(0.00) 3(9.38) 2(6.25) 1(3.13) 7(4.38)
16-20 years 0(0.00) 1(3.13) 0(0.00) 1(3.13) 0(0.00) 2(1.25)
>20 years 1(3.13) 2(6.25) 0(0.00) 0(0.00) 0(0.00) 3(1.88)
Values in parentheses are percentages.
Manda et al. 10.3389/fcosc.2023.1124142
Frontiers in Conservation Science frontiersin.org07
three motivating factors, namely (1) benefits and/or incentives that
members of the NRCs get (100%); (2) the urge to take part in
management decision-making for the Reserve (98.51%); and (3) the
feeling to further help build the relationship between communities
and the park staff (16.42%). Some of the benefits and/or incentives
mentioned included temporary employment or piece works in the
reserve, tours to various PAs, livestock pass-on programmes, and
first-hand access to information regarding community development
programmes in the area.
Perceived importance of conserving biodiversity
in the VMWR
Other than direct utilitarian benefits obtained from the VMWR,
respondents were asked to mention the importance of conserving
biodiversity in the reserve. In this regard, the respondents
demonstrated having adequate knowledge of the various ecosystem
services that biodiversity in the VMWR brings to the area. These
included climate regulation (100% of the respondents), carbon
sequestration (99.38% of the respondents), and pollination (99.38%)
(n=160) (Figure 3).
Key roles of local communities in the co-
management arrangement in the VMWR
Both members and non-members of the NRCs demonstrated
knowledge of local communities’roles in the conservation of
biodiversity in VMWR under the co-management arrangement.
Some of the key roles cited included patrolling and reporting illegal
activities (87% of the respondents) and removing litter, especially
plastics from along the wire fence (51% of the respondents)
(n=160) (Figure 4).
Trends in animal populations, animal
mortality, and prohibited activities in
the VMWR
Animal population trends
The results showed that animal species populations were
relatively much higher by 1985 but they almost crashed by 1996
(Figure 5). The only remarkable exception in this regard was the
elephant (Loxodonta africana) population. For instance, the
population of buffaloes (Syncerus caffer) decreased from a record
900 (1985) to only less than 200 in 1996 and since then an average
of 151.33 (median = 147) buffaloes have been recorded per year. On
the other hand, no individual eland (Taurotragus oryx) had been
spotted in the reserve after 1996 from an estimated population of 75
animals in 1985; while zebra (Equus quagga) appears to have only
re-emerged in 2007 since 1985, and average of 7.86 individuals had
since been recorded in the surveys between 2007 and 2021
(Figure 5A). The population of roan antelope (Hippotragus
equinus) decreased from 700 (1985) to barely less than 100
animals by 1996 (Figure 5B). Since then, its population has
remained roughly below 200 (mean = 159; median = 162).
Similarly, there were approximately 1000 common duikers
(Sylvicapra grimmia) and 700 bushbucks (Tragelaphus scriptus)in
1985 but the population sizes of these two species had reduced
drastically to <100 and 10 by 1996 for the common duiker and
bushbuck, respectively (Figure 5C) and have continued to fluctuate
below 100 animals to date. A similar trend was observed for
warthog (Phacochoerus africanus) and grysbok (Raphicerus
meanotis). There were about 1500 warthogs and 1000 grysbok in
the mid-1980s, but the populations of these two species drastically
TABLE 2 Diversity of NTFPs in the studied five zones in the VMWR based on three diversity measures [Simpson (1-D), Shannon (H′), and Evenness (e^H′/S).
Metric Bowe Kazuni Lake Kazuni Lusani Thunduwike
Taxa (S)8109109
Individuals (N) 175 153 228 173 234
Simpson (1-D) 0.853 0.867 0.877 0.852 0.885
Shannon (H′) 1.944 2.101 2.103 2.001 2.172
Evenness (e^H′/S) 0.874 0.817 0.910 0.740 0.976
FIGURE 2
Prescribed resources obtained from the VMWR by the studied
communities.
Manda et al. 10.3389/fcosc.2023.1124142
Frontiers in Conservation Science frontiersin.org08
reduced to <100 animals by 1996 and have remained almost the
same to date (Figure 5D).
Animal mortality
Generally, the results showed an increasing trend in animal
mortality over the period (Supplementary Figure 1).Thiswas
particularly evident for large body-sized mammals such as
elephant, buffalo, and hippopotamus (Supplementary Figure 1A),
large-antelopes e.g., roan antelope and kudu, and medium-sized
antelopes e.g., bushbuck and impala (Supplementary Figures 1B,C),
and for warthog (Supplementary Figure 1D). As expected, the trend
was less conspicuous for the non-edible species in the area such as
the mongoose (Supplementary Figure 1E) and carnivorous animals
like leopards (Supplementary Figure 1F).
Prohibited activities
The results showed an increasing trend in the number of
animals found dead in the reserve attributable largely due to
poaching, with an annual average of 2.27 (median = 1.5)
elephants and 21.82 (median=18.5) other species over the period
of 1993-2021 (Supplementary Figure 2A). Further, the number of
snares found in the reserve (annual mean=378.64; median=151.5)
and cut trees (annual mean=234.73; median=222.5) also showed an
increasing trend over the same period (Supplementary Figure 2B).
Suggestions for improvement in the
co-management approach
As resource users, local communities suggested areas or actions
they thought would enhance conservation and sustainable use of
biodiversity in the VMWR under the co-management approach
based on the lessons learnt over the years. On their part, most
respondents suggested tree planting on communal land (75%) and
livestock farming (72.50%). On the part of the DNPW, the
communities suggested undertakings such as the provision of
social amenities (65.63%), provision of small livestock such as
goats (62.50%), and timely sharing of revenue accrued from the
reserve (26.88%) (Figure 6).
When these suggestions were correlated, the results showed a
somewhat moderate negative correlation between livestock farming
and stopping poaching (r≥-0.45), and a moderate positive
correlation between tree cutting and stopping poaching (r≥
0.37) (Table 3).
Socio-demographic factors influencing
choices, knowledge, and perceptions of
local communities on biodiversity
conservation in the VMWR
The Multivariate probit (MVP) model showed that choices that
local communities make about participation in conservation work
significantly vary depending on their socio-demographic attributes
(P = 0.05, n = 160)(Supplementary Table 1). The model showed
FIGURE 3
Perceived importance of conserving biodiversity in the VMWR.
FIGURE 4
Key roles of local communities in the co-management arrangement
in the VMWR.
Manda et al. 10.3389/fcosc.2023.1124142
Frontiers in Conservation Science frontiersin.org09
that age of individuals was the most influencing factor, with the
probability of the old (age>60 years) citing benefits and/or
incentives being significantly lower than that of the adults (40
<age ≤60 years) citing this reason as a motivating factor (P = 0.004;
S.E. = 0.471; coef. = -1.364). Gender, level of formal education,
distance from the reserve boundary, and membership status did not
affect this reason. However, female respondents frequently cited the
need to help build the relationship between park staff and the local
community more than did male respondents (P = 0.005; S.E. =
0.325; coef. = -0.907). The level of formal education, distance from
the reserve boundary, and membership status did not affect this
aspect. Further, there was a significant negative correlation between
the two reasons (benefits and/or incentives and relationship with
the park staff) (rho21; P = 0.002; S.E. = 0.162; coef. = -0.496)
(Supplementary Table 1A).
Significant variations were observed among local communities
regarding knowledge of their roles in the co-management
arrangement (P = 0.0001; n = 160) largely influenced by gender
and membership status of the respondents. It was found that the
probability of male respondents citing removing litter was lower
than female respondents (P = 0.040; S.E. = 0.215; coef. = 0.441).
Whereas, the probability of non-members of NRCs citing removing
litter was higher than members (P = 0.002; S.E. = 0.218; coef. =
0.669). However, non-members cited community sensitisation less
frequently than members of the NRCs (P = 0.011; S.E. = 0.217; coef.
= 0.551). Further, male respondents cited removing snares more
frequently than female respondents (P = 0.003; S.E. = 0,226; coef. =
0.668), while non-members cited this role less frequently than did
members of the NRCs (P = 0.002; S.E. = 0.227; coef. = 0.716). The
model further showed that there were significant negative
correlations between removing litter and community sensitisation
(rho21; P = 0.003; S.E. = 0.116; coef. = -0.351), removing litter and
removing snares (rho31; P = 0.026; S.E. = 0.122; coef. = -0.272), and
community sensitisation and removing snares (rho32; P = 0.008;
S.E. = 0.127; coef. = -0.336)(Supplementary Table 1B).
Discussion
This study assessed the extent to which co-management has
effectively contributed to biodiversity conservation and socio-
economic development outcomes in the Vwaza Marsh Wildlife
Reserve based on perceptions of local communities bordering the
reserve. It also utilised secondary ecological data on trends in the
animal population, animal mortality, and prohibited activities.
The results are discussed in the subsequent thematic subsections,
and implications for conservation are suggested under each theme.
Local communities’access to resources in
the VMWR
Based on the responses of the studied communities, it is evident
that co-management has improved access by local communities to
NTFPs in the VMWR thereby enabling them to meet their daily needs
(assumption one). This is contrary to the previous management
A
B
DC
FIGURE 5
(A–D) Estimated population trends for 17 animal species in the VMWR from 1985-2021.
Manda et al. 10.3389/fcosc.2023.1124142
Frontiers in Conservation Science frontiersin.org10
approach where the park-border communities were denied access to
valuable resources in the reserve. Increasing benefits to local
communities through RUPs is one of the preconditions for co-
management and has been widely reported (e.g., Muhumuza and
Balkwill, 2013), although some studies have reported otherwise. For
instance, despite the introduction of co-management in Blouberg
Nature Reserve in South Africa, the majority of the households
interviewed (87%) (n = 290) reported not to have received benefits
from the reserve (Rampheri et al., 2022). The reported discrepancies in
the literature on benefits from PAs, according to Zhang et al. (2022),
may in part point to differences in the cognition of benefits by
communities with varying daily needs, influenced by different socio-
economic, political, and cultural factors. This highlights the need for
carefully assessing basic resources that are valuable to the park-border
communities, since such resources may be crucial for co-management
programme support (Shackleton and Shackleton, 2004).
Like in many developing countries, increasing cultural beliefs in
medicines from natural products and limited access to conventional
medicine (Drury, 2020), and perpetual food insecurity challenges
make NTFPs an important source of health care service and a food
security safety valve in southern Africa (Shackleton and de Vos,
2022). It is not surprising, therefore, to find medicine and food
(including mushrooms, edible caterpillars, and edible fruits) being
some of the important resources cited by local communities in the
VMWR. Medicine and food are two of the important resources
upon which livelihoods of many rural people rely worldwide
(Shackleton and de Vos, 2022).
Further, the results showed slight spatial variations in the
diversity of resources accessed by the park-border community,
with communities in the Thunduwike Zone using a somewhat
relatively higher diversity of resources (Table 2). This may reflect
differences in resource use among zones with a more intense use
and dependence on the VMWR in the eastern side of the reserve
where the Thunduwike zone is located or it may perhaps indicate a
relatively higher diversity of resources found in this area. Increased
human pressure on biodiversity on this side of the reserve has been
reported in previous reports (O’Sullivan, 2019). These results may
be used to guide where more focus on intense monitoring of the
RUP should be done in collaboration with the NRC members.
Although access to basic resources by the bordering communities
is not in itself a panacea for effective conservation and has also been
blamed for conflicting ecological principles of conservation (Shova
and Hubacek, 2011;Muhumuza and Balkwill, 2013), many other
studies have suggested that local communities are likely to value and
support conservation work if such initiatives address their basic needs
(Shackleton and Shackleton, 2004;Gordon, 2017;Woodhouse et al.,
2022). This is especially true in developing countries like Malawi,
given the heavy dependency on forest and wildlife resources of the
park-border communities in these countries to meet their livelihood
needs (Muhumuza and Balkwill, 2013;Shackleton and de Vos, 2022).
FIGURE 6
Sankey diagram showing suggested areas for improvement in co-management based on the perceptions of local communities in the VMWR;
Responsible actors (left) and suggested actions (right). Numbers represent frequency of citation.
Manda et al. 10.3389/fcosc.2023.1124142
Frontiers in Conservation Science frontiersin.org11
TABLE 3 Correlation matrix of the suggested areas for improvement in co-management in the VMWR.
Stop
poaching
Stop
cutting
trees
Livestock
farming
Tree
planting
Raise
awareness
Diversify
crop
farming
Promote
natural
regeneration
Provide
social
amenities
Provide
livestock
Timely
revenue
sharing
Provide
farm
inputs
Create
employment
opportunities
Provide
business
loans
Provide
tree-planting
pots
Stiffen laws
against
poaching
Extend the
fish harvesting
period
Stop
poaching 1.00
Stop cutting
trees 0.37 1.00
Livestock
farming -0.45 -0.36 1.00
Tree
planting -0.38 -0.16 0.03 1.00
Raise
awareness 0.08 -0.11 -0.18 -0.30 1.00
Diversify
crop farming -0.02 -0.11 0.06 -0.25 -0.05 1.00
Promote
natural
regeneration -0.21 -0.13 0.04 0.02 -0.03 0.05 1.00
Provide
social
amenities 0.02 -0.20 0.06 0.16 -0.12 0.04 0.01 1.00
Provide
livestock -0.10 -0.09 0.07 -0.03 0.09 0.05 -0.01 -0.21 1.00
Timely
revenue
sharing 0.04 0.23 -0.07 0.02 -0.03 -0.06 0.05 -0.15 -0.05 1.00
Provide farm
inputs -0.05 -0.02 0.00 -0.10 0.04 0.05 -0.02 -0.19 0.36 0.01 1.00
Create
employment
opportunities 0.27 0.14 -0.19 -0.13 0.20 -0.06 -0.15 -0.19 -0.23 -0.01 -0.07 1.00
Provide
business
loans -0.03 0.01 0.08 0.06 -0.03 -0.02 -0.06 -0.31 -0.03 0.11 0.04 -0.08 1.00
Provide tree-
planting pots 0.00 0.09 0.05 -0.01 -0.10 -0.01 -0.11 -0.08 0.16 -0.19 -0.04 -0.13 -0.11 1.00
(Continued)
Manda et al. 10.3389/fcosc.2023.1124142
Frontiers in Conservation Science frontiersin.org12
Perceptions of local communities
on conservation
Unlike other studies (e.g., Rampheri and Dube, 2021;Rampheri
et al., 2022) that largely found negative community perceptions of
nature conservation in PAs thus pointing towards failure of the co-
management approach, the findings of the current study have
shown that co-management had created a positive attitude
towards conservation among the local communities bordering the
VMWR (assumption two). Owing to this, and as noted in this study,
even respondents that were not members of the NRCs were willing
to become members of the NRCs so as to fully participate in the
conservation work. The positive perceptions expressed by
communities in the VMWR may relate to their increased access
to valuable resources, involvement in biodiversity conservation, and
social development programmes initiated in the area through the
co-management arrangement. Similar results were found in
Pendjari National Park in Benin (Vodouheet al., 2010).
According to Vodouheet al. (2010), the positive behaviour of
local communities towards conservation of biodiversity within the
Pendjari National Park was highly correlated with the effective
involvement of local communities in the management strategy that
involved more effectively local communities. Further, participants’
perceptions of biodiversity conservation were strongly related to
locally perceived benefits (Vodouheet al., 2010).
Recently, a global summary of local residents’attitudes towards
PAs (Allendorf, 2020) has shown that communities are more likely
to be more positive towards less strict PAs than strict ones. Bennett
et al. (2019) found that perceptions of good governance and social
impacts were stronger predictors of increasing support for
conservation work among small-scale fishermen in marine PAs
from six countries in the Mediterranean Sea. Besides, it is widely
considered that people’s positive perceptions of protected areas are
not only a key indicator of PA conservation success (Allendorf,
2020) but they also ultimately ensure the support of local
constituents thus enabling the long-term success of conservation
work (Bennett, 2016). Therefore, the positive perceptions of
conservation work as found in the VMWR point to a somewhat
successful co-management in the area. Importantly, the perceptions
provide an opportunity to achieve greater and more effective long-
term support for conservation initiatives in the VMWR.
The finding that local communities were able to associate the
biodiversity of the VMWR with various benefits they get from the
reserve including regulating benefits (e.g., pollination), material
benefits (e.g., water source), and non-material benefits (e.g.,
tourism) (Figure 3) is a demonstration of their appreciation of
nature’s contribution to people. This appreciation, according to
(Allendorf and Yang, 2013), is important for local communities to
recognise a common ground between their livelihoods and a PA.
These results are consistent with the recent findings in the PAs of
the Sundarbans mangroves of Bangladesh, where co-management
has built an awareness in favour of biodiversity conservation and
the efficient use of natural resources (Rahman, 2022). Such an
awareness of the people’s already-existing perceptions, according to
Allendorf and Yang (2013) and McShane et al. (2011), may provide
a conducive environment to initiate a discussion of win-win
TABLE 3 Continued
Stop
poaching
Stop
cutting
trees
Livestock
farming
Tree
planting
Raise
awareness
Diversify
crop
farming
Promote
natural
regeneration
Provide
social
amenities
Provide
livestock
Timely
revenue
sharing
Provide
farm
inputs
Create
employment
opportunities
Provide
business
loans
Provide
tree-planting
pots
Stiffen laws
against
poaching
Extend the
fish harvesting
period
Stiffen laws
against
poaching -0.08 -0.04 0.04 -0.03 -0.01 -0.06 0.10 -0.10 -0.13 -0.04 -0.08 -0.13 -0.11 -0.10 1.00
Extend the
fish
harvesting
period 0.10 0.05 0.03 0.10 -0.05 -0.02 -0.06 0.05 -0.23 0.13 -0.10 0.03 0.05 -0.06 -0.06 1.00
Red, negative correlation; Blue, positive correlation.
Manda et al. 10.3389/fcosc.2023.1124142
Frontiers in Conservation Science frontiersin.org13
scenarios regarding conservation work. These findings further
provide hope for a successful conservation arrangement in the
VMWR, given the shreds of evidence that conservation efforts are
likely to succeed if local communities become aware of the values of
conservation in form of various ecosystem services accrued to them
(Allendorf and Yang, 2013;Muhumuza and Balkwill, 2013).
Co-management and extraction of
prohibited resources
The animal population crash and the increase in animal
mortality in the VMWR around the mid-1990s seem to have
coincided with the political wind of change that took place in
Malawi and the political environment that ensued after 1994 in the
nascent years of multiparty democracy. Two factors could explain
these observations. With the wind of democratic rights and
freedoms, either some pockets within the local communities
thought this was their time to claim what was forcefully taken
away from them following their eviction from the park, or the
DNPW suddenly became overwhelmed as the previous
management approach of coercion could no longer be used. The
upsurge in lawlessness in the reserve during this transition might
have led to either killing of animals or migration of animals to other
areas such as Nyika National Park and neighbouring Zambia.
Although it has been suggested that there is generally no uniform
relationship between democracy and the state of the environment
(Arvin and Lew, 2011), evidence abound in the literature of
increased prohibited activities and biodiversity loss linked to
democracy, especially in countries with low economies and young
democracies (Buitenzorgy and Mol, 2011;Ryden et al., 2020). For
instance, Buitenzorgy and Mol (2011) suggest that countries in
democratic transition experience the highest deforestation rates
compared to non-democracies and mature democracies. Besides,
indicators of increasing levels of prohibited activities such as
poaching and growing requests for land and accessibility to the
VMWR were reported prior to 1994 (McShane, 1990). Although
this study did not ask the respondents about illegal activities
undertaken in the area and associated threats to biodiversity in
the reserve, the mentioning of removal of snares (35%) as one of the
key roles of the communities (Figure 4) and the suggestion to stop
poaching by nearly 34% of the respondents (Figure 6) as one of the
ways to improve co-management in the area signify indirect
admission of the seriousness of poaching in the reserve by the
communities themselves. While several factors such as climate
change and loss of suitable habitat may negatively impact animal
populations, poaching is regarded the major factor associated with
declining wildlife populations in Africa (Mutti et al., 2023). It may
be suggested, therefore, that poaching was more serious prior to the
establishment of co-management, especially in the transition period
into multiparty democracy, which might have led to the animal
population crashes in the VMWR (Figure 5).
According to Allendorf et al. (2012), higher levels of community
participation are generally related to higher levels of compliance.
However, our study found increasing trends in prohibited activities
even after the introduction of co-management. At this stage, those
that are actually involved in prohibited activities are not known: is it
members of the NRCs or non-members or both, or other people
outside the border zone? Nonetheless, the finding contradicts our
expectation that co-management would reduce or completely avoid
the occurrence of prohibited activities in the reserve (assumption
three). Interestingly and encouragingly, despite the increasing trend
in the reported prohibited activities, most species seemed to have
recovered from the crash they suffered in 1996 after 2000 including
the record re-emergence of the zebra, coinciding with the
establishment of co-management (Figure 5). It maybe suggested
therefore that the observed trends in prohibited activities probably
reflect a continuation of an existing trend that was otherwise less
reported prior to the introduction of co-management. It may also
imply that probably increased efforts of removing traps and snares
are paying dividends.
Our study did not assess reasons behind the increasing trends in
prohibited activities in the VMWR. However, van Velden et al.
(2020) have suggested that hunting for income, a preference for the
taste of wild meat and added diversity in the diet are key drivers of
bushmeat consumption in Malawi. In their exploratory study on the
prevalence of hunting and consumption of wild meat in four PAs in
Malawi including the VMWR, van Velden et al. (2020) found that
nearly 39% and 4 -19% of the population (n=1562) consumed wild
meat and engaged in hunting, respectively, with consumption being
more prevalent in poorer households. This may mean that
poaching, probably driven by poverty and culture, is still an issue
in the VMWR despite the introduction of co-management. Indeed,
poverty (Knapp, 2012;Matseketsa et al., 2022) and a culture of
hunting for bushmeat (Tuu et al., 2008;Viollaz et al., 2022) have
been reported as key drivers of non-compliance in PAs. According
to Muhumuza and Balkwill (2013), such factors may make some
members of communities unable to appreciate incentives obtained
from PAs but get involved in prohibited activities like poaching. For
example, it has long been suggested that poor people may be forced
to overexploit wildlife resources even if they are faced with various
sanctions including arrests, fines, and imprisonment (McShane,
1990;Knapp, 2012).
On the other hand, while some people may be aware of the
prohibited activities but offenses such as poaching may not be
viewed as deviant (Viollaz et al., 2022) since bushmeat has all along
been consumed as part of their tradition and culture (Tuu et al.,
2008). Similar cultural and traditional practices have been raised to
justify claims for access to resources in other ecosystems including
aquatic ecosystems (Williams, 2021). Thus, it may be suggested that
some members of the local communities in the VMWR might have
taken advantage of the RUP to undertake prohibited activities like
snaring in the reserve. Involvement of resource use committees in
prohibited activities and practices of corruption have been reported
in Bangladesh (Rahman, 2022) and Kibale National Park in Uganda
(Solomon et al., 2012). Moreover, local community accessibility to
reserves has also been implicated in driving prohibited activities in
other PAs such as the Bwindi Impenetrable National Park in
Uganda (Bitariho et al., 2022) and Bardia National Park in Nepal
(Shova and Hubacek, 2011).
The increasing trend in snaring in the VMWR is worrisome
(Supplementary Figure 2B), as is globally the case (Watson et al.,
Manda et al. 10.3389/fcosc.2023.1124142
Frontiers in Conservation Science frontiersin.org14
2013;Gray et al., 2018). According to Gray et al. (2018), snaring is
one of the major drivers of defaunation since snares are cheaply
made, easy to set, difficult to detect, and indiscriminate. As such,
while removal of snares was reported in the VMWR, Gray et al.
(2018) argue that removal alone is largely ineffective as snares will
continuously be replaced in the absence of proactive search, arrest,
and prosecution of snare-setters, along with incentives not to hunt.
Several suggestions aimed at curbing prohibited activities such as
snaring have been made elsewhere including compensating local
communities to forfeit overexploitation of natural resources
(Amadu et al., 2021), criminalising the possession of snares (Gray
et al., 2018), and provision of alternative livelihood interventions
(Gray et al., 2018;Brittain et al., 2022;Willis et al., 2022). For
instance, park-border communities in northern Ghana have
demonstrated willingness to accept an average annual amount of
GH¢3346.26 (US$ 339.36) and GH¢1487.67 (US$ 150.87) per
household as compensation to forfeit the exploitation of market
and non-market forest resources, respectively (Amadu et al., 2021).
Unfortunately, while compensation may help curb snaring and
other prohibited activities, implementation of this strategy in many
PAs including the VMWR may not be sustainable, considering that
most PAs in Africa have a limited financial resource base for PA
management (Muhumuza and Balkwill, 2013;Baghai et al., 2018).
On the other hand, Gray et al. (2018) have suggested legislative
reform in Southeast Asia that criminalises the possession of snares
including materials used for their construction, inside and in the
border zone of PAs, as well as consistent enforcement of such
legislation as a way of curbing poaching. This, according to Gray
et al. (2018), ought to be combined with longer-term awareness-
raising activities aimed at changing cultural attitudes and behaviours
related to the consumption of wildlife products. However,
considering that law enforcement based on arrests and fines alone
may not be effective against poaching (Knapp, 2012;Moreto and
Charlton, 2021), proactive snaring prevention through use of
informal guardianship as advanced by Viollaz et al. (2022) could
become a useful strategy in the VMWR. This also bodes well with the
proposition of a community problem-solving policing model where
local communities are actively involved in anti-poaching (Moreto
and Charlton, 2021). Mapping of prohibited activities such as
poaching could therefore serve as a starting point to guide park
managers in the VMWR in the formulation of targeted management
strategies in this regard (Degbelo et al., 2022). The work initiated in
VMWR by McShane (1990) on the spatial mapping of prohibited
activities may provide a valuable baseline in this respect.
Alternative livelihood interventions implemented at the
household level and supported by awareness raising that address
both food (animal protein) and income have also been suggested as
possible strategies (Brittain et al., 2022;Willis et al., 2022). For
instance, a recent study in Dja Faunal Reserve in Cameroon on
alternative livelihood interventions (Brittain et al., 2022) has shown
that alternative projects that offer both food and income-generating
activities could reduce household rates of hunting and consumption
of wild meat. Therefore, for successful alternative livelihood projects
in the VMWR, an understanding of the characteristics of alternative
projects preferred by the communities using a scenario-based
approach (Brittain et al., 2022) may be considered before the
finalisation and implementation of project designs.
Given that communities around the VMWR are primarily
farmers, and that prohibited activities in the area tend to increase
with declining human food security (van Velden et al., 2020),
agricultural-related interventions suggested by the communities
themselves (Figure 6) would probably improve the economic
activities in the area. These may help reduce reliance on wildlife-
based economies. This is also important considering that local
community involvement in the current community development
projects seems to be failing to reduce the hunting and consumption
of wild meat (van Velden et al., 2020). However, given the growing
human population size around the VMWR and associated soil
degradation (McShane, 1990;van Velden et al., 2020), indigenous
multipurpose tree species with the potential to increase agricultural
production and selected in a participatory manner with farmers
(Leakey et al., 2022) could play a critical role in this regard.
Crucially, participatory revision of the access permit, especially
the need for park staff to accompany communities during resource
collection activities as recommended in Majete Wildlife Reserve
(Gordon, 2017) ought to be considered in the VMWR.
Co-management as a problem-solving
platform for improved human well-being
It has long been suggested that co-management is not a fixed
state but a continuous problem-solving process evolving over time
(Berkes, 2009). It would, therefore, seem that the initial phases of
co-management have provided local communities in the VMWR
with a platform to rethink their problems and learn from their past
experiences with co-management. This has enabled them to suggest
various ways of improving not only their quality of life (assumption
four) but also biodiversity conservation (Figure 6). These
suggestions may be crucial for developing an adaptive co-
management in VMWR that could increase both conservation
and development goals. Moreover, adapting natural resource
management based on the feedback from resource users,
according to Meijaard et al. (2021), could lead to positive
outcomes for both the environment and the well-being of people.
However, it is also recognised that benefits from co-management
implementation may take a while, given that this is a new
phenomenon (Pailler et al., 2015), thus underscoring the need for
a continued search for ways to improve the co-management
approach based on the socio-economic, cultural and political
contexts. Therefore, the suggestions made by the local
communities in VMWR may further be explored, and the
scenario-based approach (Brittain et al., 2022) could become a
useful tool in this respect. Such an exploration could help come up
with feasible and targeted interventions relevant to the local
communities in the VMWR (Allendorf et al., 2012).
Manda et al. 10.3389/fcosc.2023.1124142
Frontiers in Conservation Science frontiersin.org15
Age, gender, and membership to a natural
resource institution of local communities
influence their decisions, knowledge, and
perceptions in the VMWR
Our results have shown that age, gender, and membership to the
NRC influence conservation decisions, knowledge, and perceptions of
local communities in the VMWR, partly confirming our last
assumption (assumption five). This is consistent with many other
studies (Muhumuza and Balkwill, 2013)thatshowtheinfluence of
these three factors on conservation work. Specifically, our results
show variations in factors driving decisions for participation in
conservation work based on age. That adults (40 <age ≤60 years)
cited benefits more than old individuals (age>60 years) as a
motivating factor may mean that adults are perhaps driven by
material benefits more than other benefits of biodiversity as a
source of livelihood to fend for their families. This suggests that
adults more than the old individuals in the VMWR are likely to
engage in prohibited activities, given that older individuals are more
likely to engage with nature and avoid environmental harm in their
behaviours (Wiernik et al., 2013).
The results have further revealed the underlying gender
disparities prevalent in the area of natural resources management.
It would appear gender norms and practices have shaped the
knowledge and perceptions of community members in the VMWR,
with different roles ascribed to a different gender. For instance,
women appear to get more involved in removing litter as opposed
to men who cited removing snares more than women. Traditionally,
hunting for bushmeat is done by men. As such, men are better placed
to identify traps and snares. Similar observations regarding
differences in community roles based on gender have been reported
in forest-fringe communities in Ghana (Asumang-Yeboah et al.,
2022). Further, like in many countries in Africa where
belongingness to an organisation increases knowledge of
communities about biodiversity conservation (Muhumuza and
Balkwill, 2013), members of NRCs in the current study seemed to
be more exposed to information about the roles of communities in
the co-management. These results generally highlight the importance
of paying attention to community differentiation and attributes such
as age, gender, and membership to local institutions, as inattention to
such factors may limit the potential effectiveness of co-management
(Muhumuza and Balkwill, 2013;Viollaz et al., 2022).
Conclusions
We used a combination of perceptions of local communities and
ecological data (trends in animal population, mortality, and prohibited
activities) to assess the extent to which co-management has effectively
contributed to biodiversity conservation and socio-economic
development outcomes in the Vwaza Marsh Wildlife Reserve
(VMWR). This study has shown that populations of most animal
species are showing recovery signs from the crash suffered during the
period prior to the establishment of co-management. It has also shown
that co-management has created positive perceptions of local
communities of conservation and socio-economic development
work in the area, thus partly indicating the success of the
management approach. The positive perceptions are attributable to
improved benefits from the reserve and improved park-people
relationships. Furthermore, the study has shown that co-
management can provide an opportunity for an actively engaged
community to make suggestions that aim at improving not only their
quality of life but also ecological goals. Considering the resource
constraints (financial and human) in the VMWR, as is common with
PAs in the global south (Watson et al., 2014), and given that this is a
new management approach in complex and context-specific
relationships (Allendorf et al., 2012;Pailler et al., 2015), we conclude
that while it might be too early to achieve both conservation and
development goals, our findings provide hope for an adaptive and
effective co-management in the VMWR. We recommend a study on
the spatial distribution of traps and snares in relation to animal
occupancy, resource use zone and core zone to help channel resources
for monitoring. So too, a study to quantify sensitive behaviour related
to prohibited activities including poaching using specialised
questioning techniques such as the Randomized Response
Techniques (RRTs) (Ibbett et al., 2023) would help establish the
prevelaence of such rule-breaking behaviours among community
members in the Vwaza. Further, future conservation activities in the
area should take into consideration age, gender, and membership to
natural resource committees (NRCs) of the local communities as these
appeared to significantly influence local communities’decisions,
knowledge, and perceptions of conservation work. Furthermore,
misunderstandings over revenue sharing should be addressed
transparently, as this might be a cause of conflict for the
management of the VMWR. Finally, participatory evaluation of co-
management in the VMWR, taking into consideration the findings of
this study and lessons learnt over the years, is also recommended.
Data availability statement
The raw data supporting the conclusions of this article will be