Utilising Marxan Planning Platform for Informing Conservation Areas in Rwanda

Executive Summary

This study aims to evaluate the current protected areas in Rwanda and propose an alternative, more comprehensive conservation planning solution that focuses on focal species and the critical ecoregions from which they rely on. Utilising the Marxan Planning Platform, modifications were made to the species penalty factor (SPF) and to the target adjustment parameters based on the conservation needs of each study species according to their IUCN Red List status and the degree of spatial overlap these species had within each ecoregion within our study. Critically endangered species, including the black rhino, African forest elephant, and hill’s horseshoe bat, were designated a relatively higher SPF and target value to prioritize their inclusion in the conservation planning solution output. In contrast, endangered and vulnerable species received adjusted SPFs and target adjustments accordingly. The adjusted parameters yielded a more effective conservation solution, with an increase in protection coverage across all focal species and ecoregions compared to the default scenario. These findings support the recommendation to establish buffer zones around the protected areas to mitigate human-wildlife conflict, specifically addressing issues such as illegal poaching, deforestation activities, and agricultural expansion. Stakeholder engagement through educational outreach and employment opportunities have also been proposed in an effort to foster buy-in amongst the local community and support local livelihoods.

Introduction

The issue of defining effective biodiversity conservation practices in developing countries has sparked considerable debate in recent years (Rutagarama & Martin 2006). Given these challenges, assessing diverse conservation strategies is essential to evaluate the efficacy of Rwanda’s current national protected area system and explore potential conservation solutions. The significance in doing so is underscored in its ability to inform and hasten environmental decision-making processes.

Driven by variations in topography and climate, human population distribution across East Africa is notably imbalanced with greater than 300-500 persons/km2 in some locations (Li et al. 2021). Specifically, Rwanda had the highest population density in Africa as of 2006 (Rutagarama & Martin 2006). With a rapidly growing population, this region experiences excess pressure on natural land resources, leading to habitat degradation, biodiversity loss, excessive deforestation, and food security challenges (Li et al. 2021). The rapid increase in human population alongside the issues concerning food security in Rwanda specifically, have likely contributed to extensive poaching activities known to be reported in this region (Lunstrum and Givá 2020, Li et al. 2021, Moore et al. 2021, Dushimana 2022 ). The detrimental impacts of poaching activities can impact populations of all species, including non-target species, within a protected area (Moore et al. 2021). Furthermore, the landscape has undergone widespread habitat conversion to cropland due to a strong dependence on agriculture for sustenance (Karamage et al. 2017, Li et al. 2021). As these anthropogenic pressures intensify, the impact on natural ecosystems, as well as the proliferation of native species and habitats, becomes increasingly urgent.

In Rwanda, certain species hold a valuable role in the maintenance of biodiversity and ecosystem health. The chimpanzee (Pan troglodytes), is considered endangered by the IUCN Red List and faces repercussions from habitat loss and poaching (Humle et al. 2016). The black rhinoceros (Diceros bicornis), is subject to intense illegal poaching pressures, with approximately 95% of rhino horn supplied to illegal markets in Southeast Asia originating from Africa (Emslie et al. 2019). As a result, this species is classified as critically endangered on the IUCN Red List (Emslie 2020). The critically endangered African forest elephant (Loxodonta cyclotis) has suffered a likely irreversible 80% population decline over the past three generations (93 years) (Thouless et al. 2016, Gobush et al. 2021). Their decline is driven by poaching for ivory, habitat degradation and fragmentation, and agricultural development, which exacerbates human-wildlife conflict (Ngama et al. 2016). The critically endangered hill’s horseshoe bat (Rhinolophus hilli) and vulnerable leopard (Panthera pardus) are similarly affected by habitat degradation and loss, which compromises their persistence (Fahr et al. 2002). This is specifically pertinent to R. clivosus, which has an estimated range of only 8 km2, and is found solely in Nyungwe National Park, where they suffer a decline in habitat quality and population abundance (Webala et al. 2021).

These key species inhabit several critical ecoregions, including the Victoria Basin Forest Savanna Mosaic, Ruwenzori Virunga Montane Moorlands, Central Zambezian Miombo Woodlands, and Albertine Rift Montane Forests. The Albertine Forest features dense vegetation that not only offers habitat for the black rhino, which thrives in thick bush, but also provides nesting areas for hill’s horseshoe bats within its dense canopies. Furthermore, the high insect population within these forests serves as a primary food source for the hill’s horseshoe bat, while the elephants benefit from the fruit produced by the trees and may ring-bark them to access nutrients within the bark. Additionally, these elephants rely on the muddy areas created by rainfall which averages between 1,200 and 2,200 mm per annum in this location. The mud is used for both hydration and as a natural method for parasite removal and sun protection. Due to the prevalence of poaching in Rwanda, the dense cover of forests offer a higher chance of survival for these species. Conservation strategies that account for the biogeographic context of these ecoregions are crucial for safeguarding a comprehensive range of representative areas and the species that inhabit them, particularly those that depend on expansive terrain or are particularly vulnerable to human-induced threats (Olson et al. 2001). Therefore, integrating these ecoregions into conservation planning assessments is vital, as it complements and supports the preservation of species that rely on these habitats. This study aims to evaluate conservation strategies for the protection of key species and their habitats within Rwanda’s critical ecoregions and to explore potential expansions of existing protected areas.

Methods

This study investigates Rwanda, a biodiverse country in East Africa with four national parks: Akagera National Park, Nyungwe National Park, Gishwati-Mukura National Park, and the Volcanoes National Park. Considering Rwanda’s ecological significance, an assessment of its national protected area system may serve to inform environmental decision making processes. The Marxan Planning Platform (MaPP) was therefore utilised as a tool for modeling conservation solutions with regard to variables such as cost, current protected areas, and the conservation status of various species.

An existing planning unit grid supplied by TNC (2024) was imported into MaPP in a zipped Shapefile format as the foundational layer. This grid, represented by hexagonal units of 20 km2, was utilised as a foundation for which to base the subsequent conservation scenarios. A map of the existing protected areas (IUCN categories II and IV) in Rwanda were then uploaded as a layer into MaPP, adjusting the conservation area threshold to 75% as suggested in TNC (2024). This threshold value signals when a planning unit is considered protected. The default settings for this project were implemented by TNC (2024); however, these parameters have been modified to better suit the aim of our study. To account for conservation costs, we applied the Human Modification Index (HMI), which ranges from 0 (low modification) to 1 (highly modified). While the default setting for the cost layer illustrated an “equal area” uniform cost across the study site, the HMI was better suited to inform our cost estimate layer considering its comprehensive spatial assessment impact of anthropogenic stressors, thereby serving as a proxy for ecosystem condition.

Additional layers included features representing the spatial distributions of key species: black rhinoceros black rhinoceros, chimpanzee, elephant, hill’s horseshoe bat, and leopard. Habitat features were also included, encompassing the Victoria Basin Forest Savanna Mosaic, Ruwenzori Virunga Montane Moorlands, Central Zambezian Miombo Woodlands, and Albertine Rift Montane Forests. By default, the feature parameters were set at a 30% target, meaning the conservation scenario will aim to protect 30% of each species’ range. In the present study, this target was adjusted to scale protection levels to the conservation status of each species: 30% for the vulnerable leopard (Panthera pardus), 40% for the endangered chimpanzee (Pan troglodytes), and 50% for critically endangered black rhinoceros (Diceros bicornis), African forest elephant (Loxodonta cyclotis), and hill’s horseshoe bat (Rhinolophus hilli). Although the datasets provided by TNC (2024) did not specify the elephant species, both the African savanna (Loxodonta africana) and forest elephant are known to occur in Rwanda (Loxodonta cyclotis) are present in Rwanda, with distinct IUCN conservation statuses: endangered and critically endangered, respectively. The literature indicates a strong association with conservation efforts in Rwanda, particularly in the Nyungwe National Park (Pluymptre et al. 2002, Niyonsaba and Ndokoye 2023). Therefore, to adopt a conservative approach, we used the African forest elephant’s status to determine conservation targets and parameter modifications.

Ecoregion target parameters were adjusted based on habitat overlap with key species and the conservation status of thereof. For instance, the Albertine Rift Montane Forests serve as habitat for all focal species in our study, and was therefore et at a 40% target, while the Central Zambezian Miombo Woodlands – despite inhabiting black rhino, elephant, and leopard, was set at 20% due to limited overlap within Rwanda’s borders. Notably, the target for the Ruwenzori Virunga Montane Moorlands was set at 30%. Although this ecoregion supports the same species as the Zambezian Miombo Woodlands, a larger portion of its habitat lies within Rwanda’s borders. Lastly, the Victoria Basin Forest savanna mosaic target was set at 30% because it inhabits the critically endangered elephant and black rhino and vulnerable leopard. Although it is also home to the endangered chimpanzee, this falls outside of Rwanda borders. In light of this, it is comparatively similar to the Ruwenzori Virunga Montane Moorlands in that they are home to the same species within Rwanda borders.

The species penalty factor is a setting that adjusts the emphasis for the model to achieve the specific target for key species. This is critical for ensuring key species will achieve their target when balancing costs. The default setting by TNC (2024) was 1; however, this has been adjusted to adhere to the conservation needs of each species according to their status in the IUCN Red List and their degree of overlap with the ecoregions in our study. The critically endangered black rhino was given an SPF of 5, whereas an SPF of 2 was granted to elephants, and an SPF of 1.5 was given to the hill’s horseshoe bat. The endangered species (chimpanzee) and vulnerable species (leopard) were given an SPF of 1. In regard to the ecoregions, the SPF parameters reflect the target values in that the Albertine Rift Montane Forest was given an SPF of 4.5, the Ruwenzori Virunga Montane Moorlands was given 2, the Central Zambezian Miombo Woodlands was given a 0.5, and the Victoria Basin forest savanna mosaic was given a 2. These alterations in SPF allow for the prioritization among species based on conservation status and among ecoregions according to habitat overlap with the study species and spatial extent within Rwanda borders.

To minimize fragmentation and encourage solution clustering, boundary length modifier (BLM) adjustments were calibrated from 0.001 to 10, following TNC (2024) guidelines. Upon exploring the results and investigating the relationship between the BLM value and cost, a BLM of ~1.5 was selected for optimal clumping. Marxan ran 10 times using the default advanced settings, with exception for the conservation feature missing proportion, which was set to 0.95. After running the scenario, the best solution was chosen from the solutions table and the report was downloaded. The target achievement function was then utilised as a visual aid to evaluate conservation targets for each feature.

Results

The adjusted parameters in my conservation planning model was an improvement relative to the default settings presented in workshop 1 (Fig. 1, 2, 3). Key species of interest and their habitats received increased attention in my model, which aligns well with the modified target and SPF values. In the Albertine Rift Montane Forest, my solution achieved a 41% coverage, surpassing the target value of 40% and the default scenario which reached only 30% coverage. This increase is necessary in order to aid in the protection of focal species such as the critically endangered hill’s horseshoe bat. Similarly, my adjusted scenario provided a slight increase for the Ruwenzori Montane Moorlands, with a 47% coverage, in contrast to 31% in the default parameters outlined in workshop 1. Adjustments to the settings also increased protection levels in the Central Zambezian Miombo Woodlands and the Victoria Basin Forest Savanna Mosaic, achieving 79% and 50% coverage, respectively. These values are significantly above the default scenario outcomes of 52% and 39%. This is an essential takeaway because it allows for a more comprehensive approach for habitat conservation across a more representative range of ecoregions and the species in which they inhabit.

The modified parameters had notable impacts on species-specific conservation targets, particularly for those classified as critically endangered. The black rhino reached its target of 50% coverage in my model, whereas the default scenario provided only 35% protected area coverage. Protection for the elephant reached 86%, which is a modest increase from the 83% protection offered from the default scenario. My adjustments also enhanced outcomes for species that had previously missed targets, such as the leopard which received only 23% protection in the default scenario. This provides a significant comparison with the 92% protection offered by the model created in this study. The chimpanzee and hill’s horseshoe bat coverage showed a slight variation from the default scenario, achieving a 66% and 92% coverage in the new model respectively, compared to the 65% and 93% protection offered in the default scenario.

Recommendations

Based on these findings, it is recommended that conservation efforts prioritize the Albertine Rift Montane Forest due to its role in providing habitat for all focal species within this study, including the hill’s horseshoe bat, which suffers from a highly confined range within this ecoregion. Expanding protected areas within this ecoregion and the Ruwenzori Montane Moorlands could enhance habitat security and connectivity, ensuring that key species such as the critically endangered black rhino and elephant can maintain viable ranges and reside in a quality habitat that has the capacity to support their ecological roles, movement parameters, and dietary requirements. Moreover, establishing and preserving biodiversity corridors between these proposed protected areas could facilitate gene flow and migration, particularly for wide-ranging species such as the leopard and elephant.

To address and mitigate human-wildlife conflict, the land use around these protected areas should adhere to a buffer layer of protection. This ‘buffer’ should have constraints on agricultural expansion and increased patrol to aid in the reduction of illegal poaching and deforestation (logging) activities. Engaging with a diverse array of stakeholders and providing jobs to the local community would aid in creating buy-in and potentially increase the efficacy of the proposed protected areas. This would require a dedicated educational outreach program to ensure the involvement of the general public surrounding these areas. The monetary benefits surrounding the tourism industry would provide an economic incentive for the local community members to support and adhere to the legal requirements of these protected areas.

References                                                            

Dushimana, J. C. (2022). Assessment of factors influencing poaching and impact on conservation of Nyungwe National Park: A case study of Cyato and Rangiro sectors (Doctoral dissertation, University of Rwanda (College of science and Technology)).

Emslie, R. 2020. Diceros bicornis. The IUCN Red List of Threatened Species 2020: e.T6557A152728945. https://dx.doi.org/10.2305/IUCN.UK.2020-1.RLTS.T6557A152728945.en

Emslie, R., Milliken, T., Talukdar, B., Burgess, G., Adcock, K., Balfour, D. and Knight. M.H. 2019. African and Asian Rhinoceroses – Status, Conservation and Trade: A report from the IUCN Species Survival Commission (IUCN/SSC) African and Asian Rhino Specialist Groups and TRAFFIC to the CITES Secretariat pursuant to Resolution Conf. 9.14 (Rev. CoP17). CoP18 Doc. 83.1 Annex 2. CITES Secretariat, Geneva, Switzerland. http://www.rhinoresourcecenter.com/pdf_files/156/1560170144.pdf.

Fahr, J., Vierhaus, H., Hutterer, R. and Kock, D. 2002. A revision of the Rhinolophus maclaudi species group with the description of a new species from West Africa (Chiroptera: Rhinolophidae). Myotis 40: 95-126.

Gobush, K.S., Edwards, C.T.T, Maisels, F., Wittemyer, G., Balfour, D. & Taylor, R.D. 2021. Loxodonta cyclotis (errata version published in 2021). The IUCN Red List of Threatened Species 2021: e.T181007989A204404464. https://dx.doi.org/10.2305/IUCN.UK.2021-1.RLTS.T181007989A204404464.en. Accessed on 10 October 2024.

Humle, T., Maisels, F., Oates, J.F., Plumptre, A. & Williamson, E.A. 2016. Pan troglodytes (errata version published in 2018). The IUCN Red List of Threatened Species 2016: e.T15933A129038584. http://dx.doi.org/10.2305/IUCN.UK.2016-2.RLTS.T15933A17964454.en             

Karamage, F.; Zhang, C.; Fang, X.; Liu, T.; Ndayisaba, F.; Nahayo, L.; Kayiranga, A.; Nsengiyumva, J.B. Modeling Rainfall-Runoff Response to Land Use and Land Cover Change in Rwanda (1990–2016). Water 2017, 9, 147.

Kennedy, C. M., Oakleaf, J. R., Theobald, D. M., Baruch‐Mordo, S., & Kiesecker, J. (2019). Managing the middle: A shift in conservation priorities based on the global human modification gradient. Global Change Biology, 25(3), 811–826. https://doi.org/10.1111/gcb.14549

Li, C., Yang, M., Li, Z., & Wang, B. (2021). How will Rwandan land use/land cover change under high population pressure and changing climate?. Applied sciences, 11(12), 5376.

Lunstrum, E., & Givá, N. (2020). What drives commercial poaching? From poverty to economic inequality. Biological Conservation, 245, 108505.

Moore, J. F., Uzabaho, E., Musana, A., Uwingeli, P., Hines, J. E., & Nichols, J. D. (2021). What is the effect of poaching activity on wildlife species?. Ecological Applications, 31(7), e02397.

Ngama, S., Korte, L., Bindelle, J., Vermeulen, C. and Poulsen, J.R. 2016. How bees deter elephants: beehive trials with forest elephants (Loxodonta africana cyclotis) in Gabon. PLoS One 11(5): e0155690. DOI: 10.1371/journal.pone.0155690.

Olson, D. M., Dinerstein, E., Wikramanayake, E. D., Burgess, N. D., Powell, G. V., Underwood, E. C., ... & Kassem, K. R. (2001). Terrestrial Ecoregions of the World: A New Map of Life on Earth: A new global map of terrestrial ecoregions provides an innovative tool for conserving biodiversity. BioScience, 51(11), 933-938.

Plumptre, A. J., Masozera, Fashing, M., P.J., Mcneilage, A, Ewango, C., Kaplin, B.A., Liengorla, I. (2002). ‘Biodiversity surveys of Nyungwe forest reserve in SW Rwanda’ WCS working paper No. 19.

Thouless, C., Dublin, H. T., Blanc, J., Skinner, D. P., Daniel, T. E., Taylor, R., ... & Bouché, P. (2016). African elephant status report 2016. An update from the African Elephant Database, (60), 10.

TNC. 2024. The Marxan Planning Platform: Tutorial (Rwanda). Global Science, The Nature Conservancy. (TNC), Arlington, Virginia, USA. https://marxanplanning.org

Webala, P., Flanders, J., Frick, W.F. & Fahr, J. 2021. Rhinolophus hilli. The IUCN Red List of Threatened Species 2021: e.T44781A203829053. https://dx.doi.org/10.2305/IUCN.UK.2021- 3.RLTS.T44781A203829053.en