* Based on the study of Bijl, H & Heltai, M. (2022). A narrative review on the use of camera traps and machine learning in wildlife research. Columella - Journal of Agricultural and Environmental Sciences, 9(2): 47-69. DOI: https://doi.org/10.18380/SZIE.COLUM.2022.9.2.47

A mighty roaring red stag in autumn, a swiftly moving fox, or even the cute newborn wild boar piglets in spring.

Every hunter or game manager who uses a trail camera on their hunting ground knows the excitement when the camera captures something.

However, hunters and game managers are not the only ones using these handy tools.

So for what kind of applications can camera traps be useful?

Trail cameras for wildlife science

Although trail camera manufacturers mainly serve the needs of hunters, over the last few decades it has become an increasingly used scientific tool in wildlife research.

Camera traps have many benefits and can be used in a variety of situations.

What, where and how many?

Trail cameras basically provide information where humans cannot be present. 

Therefore, it is a great way to monitor animal populations remotely and provide real-time observations. This becomes especially handy when animals are rare, endangered, or elusive in nature. 

Although there are many census methods (both direct and indirect), trail cameras can be an additional tool to gain information on the occupancy of species and their distribution in an area.

It is also a common way to estimate population size and has been used for a long time for a variety of species, such as white-tailed deer, bighorn sheep, wild boars, bears, red foxes, and different wild cat species.

Wildlife health

Camera traps can also be used as a part of large programs monitoring disease by observing the contact of animals that are susceptible to disease transmission. 

This can mean direct or indirect individual contact, like nuzzling, faecal-oral contact, and site visitation. This type of contact (or disease transmission) can be within species (intraspecific) but also between different species (interspecific).

For example, camera traps can take moment-of-contact pictures between wild deer and farmed deer. As chronic wasting disease is quickly spreading in North America, this method can detect possible ways of transmission.

Plus, it is also easy to monitor the vaccine intake of animals by means of bait consumption.

Protect the eggs!

Another way to use camera traps in wildlife research is to identify nest predators.

Normally, predation signs like hair, scats, or tracks near the nests are used. However, to collect this evidence, humans need to approach the nest which can disrupt the nesting pattern or scare away predators. 

This obviously conflicts with the very aim of the examination - which is to detect and identify predators.

Therefore, remote cameras are an efficient tool to provide information on the type of predator and the time the predation happened without directly interfering.

This method is mainly used for bird’s nests, but nests of reptiles can be studied just as much with this method.

Sleeping or awake?

Radiotags on animals give a good indication of their daily or seasonal activities. 

However, this method is often invasive as the animals need to be handled beforehand which is stressful for the animal. Additionally, it is an expensive and labour-intensive method which does not work on elusive or shy species.

Therefore, camera traps are a non-invasive method that can be used to investigate animal activity patterns because of the time stamps on the images.

Another new, creative way is mounting the cameras on the animal itself. This has the same idea as the radiotags but with the added imagery.

This was done in Alaska on brown bears where they used neck-mounted cameras to determine the killing rates of moose and caribou. With this method, they actually found higher kill rates than previous estimates with other methods.

Additionally, it gives insight into the diet of the species, which brings me to the next point.

Feeding habits

There are multiple methods for analysing the diet of wildlife. These include (manual) direct observation and the analysis of scats/faeces (see the previous article on microscopic analysis of herbivore faeces here), stomachs, and prey remains.

Camera trapping for diet analysis is an alternative for observations and is mainly used for nesting raptors. 

Although no humans are needed for this method, a lot of food items remain unidentified to the species level and usually underestimate smaller-sized prey. Therefore, this needs to be taken into account when using this method for this purpose.

Wildlife corridors

With the increase in infrastructure and urban areas, wildlife habitat is more and more fragmented. Unfortunately, this often results in wildlife-vehicle collisions when animals move from one area to another.

One of the solutions to tackle this is building wildlife crossings to provide safe movement for the animals.

For ensuring that the corridors are properly functioning, monitoring them with remote cameras is a great way to assess which species use the crossings and how often.

Downsides of camera trapping

Although camera traps can be used in a variety of ways, there are some disadvantages. 

These negative aspects include, among others, that remote cameras are very susceptible to their environment. Not only can the vegetation hinder the view, but they are also sensitive to theft and vandalism. 

It would probably take some trial and error to place the camera in the right spot.

That being said, they are dependent on where we place them and how and where they are being used. Otherwise, we just end up with little to no data.

Talking about data; one of the major downsides of camera trapping is processing its data.

A large number of images and the false positives and negatives all need to be examined manually and filtered before we end up with a useful dataset.

However, as technology advances, there are tools that could aid in this process.

One of these emerging tools includes artificial intelligence. You can read about the use of machine learning in camera trap data processing here.