top of page


Short pieces on research and other things

In time, below you'll find a collection of easy to digest articles relating to my research outputs and other interesting news in wildlife ecology and conservation.  For now, you'll find a short article on the use of tourist photographs for surveying wildlife densities in tourism areas.


Photo credit: Gary Whyte @ Unsplash

The third chapter of my PhD was recently published. This work looked at leopard scent marking behaviours and how they might be impacted by environmental features, like roads. An overview:

Scent marking is a type of communication where individuals leave chemical signals that contain information in the environment. These marks (and where they are placed in the environment) can contain information on things like who owns a territory, the sexual status of the signaler, and the fitness of individuals. For species where individuals occur in low densities or cover large areas, and so who are unlikely to meet others of the same species, scent marks can help individuals coordinate movements and maintain social structure. For example, scent marks can help individuals find each other for mating, maintain boundaries with territorial neighbours, and keep track of members within their social group.

On the other hand, scent marks can be costly to produce. There are energetic costs associated with producing scent marks and with travelling around landscapes to place marks in the right areas. There are also costs if marks are encountered by species who use them to cause signalers harm or steal their resources (e.g. food). As a result, species can use different strategies to reduce signaling costs and increase the chances of scent marks being encountered by the right individuals. One strategy, for example, might involve changing where scent marks are placed in the environment. For many species, relatively little is known of how scent marks are distributed across the landscape and how this might be impacted by the environment.

I used data collected from following leopards and GPS radio collars to look at leopard scent marking behaviours in the Okavango Delta. Leopards are territorial. So, I looked at how leopard scent marking behaviours change across areas of their territories (central v boundary areas) and across the pathways they travel along (natural v roads).

In a nutshell, I found that leopards within our study area visited scent marking sites on roads in boundary areas of their territory more quickly than those elsewhere, perhaps to keep boundary scent marks fresh to deter intruders. I also found evidence that roads function as key locations for chemical information. Leopards scent-marked over four times as frequently and investigated over three times as frequently when travelling on roads than when travelling along natural routes. This means that these human-made features now play an important role in leopard communication in this landscape.

Understanding how animals communicate and how human changes to the landscape can impact this is key in ecology. It can help us to document and so reduce the negative impacts that we might be having on species. More than that, it can also pave the way for novel approaches for reducing human-wildlife conflict - like the bio-boundary project, run by my collaborators at Botswana Predator Conservation. .

To read my full paper, head on over to:

Or send me a message to get an author-copy.

Photo by Geran de Klerk on Unsplash

The second of my PhD chapters is now available. In a nutshell:

Encounters between wildlife are important. They can, for example, impact an animal’s ability to find suitable mates, whether they lose a kill to a dominant predator, and even the transmission of diseases. And so encounters can impact the density of wildlife within an area. But for large carnivores, we know relatively little about the factors influencing whether two individuals from different species are likely to meet. This is because large carnivores tend to occur in low densities, they cover wide areas of land, and they typically occur over landscapes that are difficult for researchers to navigate across. As a consequence, directly observing encounters between individuals is challenging. Even modern technologies, like animal-worn GPS sensors, haven’t traditionally been able to collect data on encounters because of hardware and software limitations. In recently published work, however, alongside collaborators from the BPCT and RVC, I used custom-built, high-resolution GPS radio collars to look at likely encounters between leopards, lions, African wild dogs, and cheetahs. From 2012 to 2018, we recorded 115 leopard-competitor encounters at our study site in northern Botswana. We found that encounters initiated by smaller competitors tended to occur in habitats with poor visibility, suggesting that encounters in these cases are a result of imperfect decision making. We also found that encounters peaked during periods of shared high-activity overlap between species and often varied with moonlight availability.

These results provide insights into how the environment can shape encounters within communities of coexisting competitors and how changes in activity patterns (caused by other things) could impact competition dynamics within species. Looking at these insights alongside the actual fitness consequences of encounters could help us to understand how changing environments might impact the viability of species who share their habitats with competitors.

Or send me a message to get an author-copy.

Image credit: K Rafiq & M. Claase

Below is an edited and slightly longer version of an article published in 'The Conversation'


Animal populations have declined on average by 60% since 1970, and it's predicted that around a million species are at risk of extinction . As more of the Earth’s biodiversity disappears and the human population grows, protected landscapes that are set aside to conserve biodiversity are increasingly important. Yet most are significantly underfunded, and many of Africa’s most treasured areas operate in funding deficits of hundreds of millions (USD).

The problems facing wildlife are complex, but information, data, can help steer the way, and evidence-based conservation decisions are recognised as a key component of effective conservation. Data on population numbers and trends can help divert limited resources to the right places at the right times, and ultimately, it can help increase the number of species that can be saved. Yet our world is changing far more rapidly than we can understand it. For many species and areas, information is sporadic or absent.

In unfenced wilderness, scientists rarely have an inventory on the exact numbers of species in an area at a particular time. Instead they make inferences on the numbers or densities of wildlife using one of many different survey approaches, including camera traps, track surveys, and drones. These methods can estimate how much and what kind of wildlife is present, but large amounts of effort (time and money) is often needed to generate reliable density estimates.

Camera traps are remote cameras, activated by movement, that can collect vast quantities of data by taking photographs of passing animals. However, camera trap surveys can cost tens of thousands of dollars to run and once in the wild, cameras are at the mercy of curious wildlife. Track surveys rely on specialist trackers, who aren’t always available; and drones, whilst promising, have restricted access to many tourism areas in Africa. All of this makes wildlife monitoring difficult to carry out over large areas and repeat, limiting our access to the information needed for fast and effective conservation.

Citizen science on Safari

Tourism is one of the fastest growing industries in the world. In sub-Saharan Africa alone it is thought that 42 million people travelled to the region in 2018 . Many visit for the unique wildlife and unknowingly collect valuable conservation data with their phones and cameras: wildlife photographs. Photographs on social media are already being used to help target illegal wildlife trade and have been suggested as an aid to sustainable tourism.

Yet tourists and their guides are still an often overlooked source of information for monitoring wildlife. Every day, businesses and governments harvest user generated data for their own purposes, but conservation, until recently, has largely passed this opportunity by. Could your holidays snaps help monitor endangered wildlife?

Male lion looking back at the photographer. Image credit: K.Rafiq

In a recent study, we tested exactly this. Partnering with a tour operator in Botswana, we approached all guests passing through a safari lodge over three months in the Okavango Delta, and we asked them if they were interested in contributing their photographs towards conservation. For those interested, we provided them with a small GPS logger - the type commonly used for tracking pet cats - so that we could see where the images were being taken.

Images were later collected, processed, and passed through computer models to estimate the densities of five large African carnivore species: lions, spotted hyaenas, leopards, African wild dogs, and cheetahs. To validate the results, we compared the densities to those from three of the most popular carnivore survey approaches in Africa: camera trapping, track surveys, and call-in stations (where sounds are played through a loudspeaker to draw-in and count the wildlife of interest).

The data from the tourist photographs provided similar estimates of wildlife densities as the other survey approaches and was much cheaper to collect and process (providing savings of up to ~$840 per survey season). Plus, while the costs of the tourist photograph method (mainly around researcher time spent processing the data) were similar to the track surveys, we typically got more information from the images than the tracks. Exactly who was sighted? What condition were they in? How were they behaving? This is all information that could be teased from the images but remained hidden from tracks. It was also the only method to detect cheetahs in the area – though so few were sighted that their total density couldn’t be confirmed.

Over the survey period, cheetahs were only detected by the tourist photograph method. Image credit: K. Rafiq.

Thousands of wildlife photographs are taken every day, and the study showed that we can use statistical models to cut through the noise and get valuable, robust data for conservation. But, there are challenges. Mainly, relying on researchers to visit tourist groups to coordinate photograph collection is far from sustainable. Luckily, that’s where wildlife tour operators could come in.

Tour operators could help collect tourist images to share with researchers. If you were to combine this coordinated data collection effort with recent advances in Artificial Intelligence for image processing, well, you could have a simple, scalable, low-cost framework for monitoring charismatic wildlife within tourism areas. As with most things, it’s not just the idea, it’s the execution.

The silver bullet for wildlife monitoring?

Are tourist photographs the silver bullet for wildlife monitoring? No. How effective tourist photographs can be will depend on the area and the species to be monitored, but there is little doubt that they are a valuable resource we need to harness. The method is best suited for monitoring large species that live in areas often visited by tourists – species that tend to have high economic and ecological value. Yet while this approach perhaps isn’t as well suited to smaller species, it can still indirectly support their conservation by helping protect the landscapes they live in.

The line between true wilderness and landscapes modified by humans is becoming increasingly blurred, and more people are visiting wildlife in their natural habitats. This isn’t always a good thing, but maybe conservationists can use these travels to their advantage and help conserve some of the most iconic species on our planet.

Silhouette of an African wild dog, made up of a mosaic of wildlife images. Main image credit: M. Claase. Collage create using Andrea Mosiac

The peer-reviewed paper on this research can be found at

bottom of page