Fences are among conservation’s most controversial interventions.
To some, they are essential for conserving wildlife, minimising encroachment, and preventing the type of conflict that happens when humans come into contact with wildlife.
To others, fences represent exclusion. They break the landscape up into pieces, prevent wild animals from moving freely over long distances and create tensions between protected areas and neighbouring communities.
These debates show that fences are more than just infrastructure – they represent a decision about the purpose of a protected area and the type of protected area a society wants to create.
Read more: Africa’s savannah elephants: small ‘fortress’ parks aren’t the answer – they need room to roam
We are a group of researchers who study how protected areas are governed and managed. We’ve been involved in research on protected area governance and management within the broader environment for many years. In our most recent study, we asked: does fencing shape how a park’s characteristics and functioning differ from the surrounding landscape?
To find out, we analysed how land is used and covered inside and outside the boundaries of 60 national parks across 17 countries south of the Equator. We looked at satellite images and compared parks that were fully fenced, partially fenced, and unfenced between 2020 and 2024.
We found that fully fenced parks were most different to the world just outside. Cropland and human settlements, roads and other developed areas were generally much more common outside the parks than inside. Natural cover, like trees and shrubs, was higher inside fully fenced parks such as the Addo Elephant National Park in South Africa.
In unfenced parks, the change across the boundary was often less abrupt. Parks merged more gradually into their surrounding landscapes, such as in the case of the Chobe National Park in Botswana.
Partially fenced parks sit between these two extremes. Some of their edges are tightly controlled while others remain more open, as seen in places such as the Kasungu National Park in Malawi.
Read more: Can ‘climate corridors’ help species adapt to warming world?
Our research shows that fences can form a rigid boundary or allow a park to be partly open. Not using fencing at all connects parks to the surrounding environment.
The choice affects the livelihoods of people living near the park, human-wildlife conflict, tourism and the biodiversity of the area.
The full fence effect: trade-offs between protection and connectivity
In parks facing heavy pressures from farming, settlement, poaching or repeated incursions, a hard boundary can be a vital management tool.
But that same hard edge can also come at a cost.
Our findings suggest that fencing can also increase human-induced pressure on the boundary. Human activity may become concentrated there, meaning that people, livestock, infrastructure and wildlife pressures are pushed up against the same narrow edge instead of being spread more gradually across the wider landscape.
This can lead to greater tension where wildlife, people and land-use pressures meet. For example, in South Africa’s Kruger National Park, fencing does not eliminate all boundary conflicts. In some areas, communities live right next to the fence. When animals break through the fence, they destroy crops, kill livestock and damage property – there’s immediate human-wildlife conflict.
A fence can help prevent encroachment but may also fragment landscapes, restrict animal movement, and deepen the separation between parks and the people and ecosystems around them. Socially, this can reinforce the feeling that the park is isolated from nearby communities and their livelihoods.
For species that rely on movement, like elephants and wild dogs, this trade-off might give them protection inside the park. However, it can also limit their movement between seasonal resources and breeding sites, and stop their natural migrations.
For local communities, fences can help reduce certain risks, but they can also restrict access and strengthen the idea of separation.
Fence decisions are not only ideological. They are also practical. Large or potentially dangerous animals, such as elephants, buffalo or lions, need fencing to ensure the safety of both visitors and wildlife. Constructing and maintaining fences over vast areas can be costly. The choice of setting up a fence or not is influenced by a mix of ecological needs, safety concerns and budgets.
Neither enclosed nor open: the hybrid park approach
One of the most interesting findings from our research concerns partial fencing.
When challenges such as human-wildlife conflicts and potential fencing costs need to be considered, the partially fenced or hybrid park is an option.
In partially fenced parks, we still found clear inside versus outside differences, but without the sharp, rigid boundaries often seen in fully fenced parks.
Read more: Wetlands in South Africa’s Addo elephant park are in danger: what’s being done to protect them
Partially fenced parks point to a more nuanced way of thinking about fencing where the choice is not only between fortress conservation and an open system.
An example is the western boundary of the Singita Grumeti concession in the vicinity of the Serengeti National Park in Tanzania. Human-wildlife conflicts are prevalent here. A partial fence here reflects ecological needs, financial realities and surrounding land uses.
How parks without fences work
Unfenced parks work quite differently. In our study, they showed the weakest contrast between inside and outside. The boundary often appeared more like a gentle transition than a sharp edge. Natural vegetation frequently extended beyond the official park boundary and farming and other human land uses often blended into the park edge. Examples include Kafue National Park and South Luangwa National Park in Zambia.
This approach offers both advantages and challenges. Without fences, parks feel less like isolated islands. But’s more difficult to prevent human activities from gradually encroaching on protected land.
What needs to happen next
First, it’s clear that areas can be protected in law, but operate very differently depending on how their boundaries are defined and managed.
Second, fencing decisions should be guided by clear goals that are appropriate for the place, its animals and its communities. If the priority is to prevent land conversion and keep a sharp conservation core, full fencing might be appropriate. If the goal is to promote more ecological connectivity across a mixed-use landscape, then partial fencing or no fencing might be preferable.
Read more: As human population grows, people and wildlife will share more living spaces around the world
For the general public, this is important because protected areas are connected to society. How their boundaries are managed affects biodiversity, tourism, land use, rural livelihoods, ecological connectivity and human-wildlife conflicts. Fencing is, therefore, more than just a technical choice for park managers: it influences how conservation is implemented on the ground.
Claudine Alberts (Roos) has not received any funding related to this paper. She has, however, received funding for waste-related conservation research from the CSIR/DSTI Waste Roadmap grant.
Dirk Cilliers, Francois Pieter Retief, Jurie Moolman, Reece C Alberts, and Ruhan Verster do not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and have disclosed no relevant affiliations beyond their academic appointment.
This article was originally published on The Conversation. Read the original article.
