There is a long history (at least since 1995) in estimating animals in a mark-recapture framework from remote camera data. Here are a few titles:

1) A critique of density estimation from camera‐trap data

2) The use of camera traps for estimating jaguar Panthera onca abundance and density using capture/recapture analysis

3) A hierarchical model for estimating density in camera‐trap studies

4) Evaluating the potential biases in carnivore capture–recapture studies associated with the use of lure and varying density estimation techniques using photographic-sampling data......

Thanks for this Brian. Very helpful. I am not actually recapturing the hedgehogs but the principle is the same. I am recording every instance (within each 15 min period) that a particular individual presents itself in front of the camera. Analysis, to date, has shown that some animals do not appear for weeks but will then make a solitary appearance and vanish again (from the cameras). Obviously, the fact that they do not appear on the cameras does not mean that they are not in the vicinity as the positioning of the cameras (although hopefully optimum) does not capture images from all angles and all areas of the study location.

If you don't know program Mark should be worth to take a look:

http://warnercnr.colostate.edu/~gwhite/mark/mark.htm

Also has an associated forum:

http://www.phidot.org/forum/index.php .

Thanks Mario. I'll have a look at these.

Are you able to identify individuals?

With the exception of a very small number, yes.

The quick answer is: Yes.

The longer answer requires reading (and I recommend):

Rowcliffe, J. M., Field, J., Turvey, S. T. and Carbone, C. (2008), Estimating animal density using camera traps without the need for individual recognition. Journal of Applied Ecology, 45: 1228–1236. doi: 10.1111/j.1365-2664.2008.01473.x

Be careful not to confuse abundance data with density indications.

Thanks very much Scott. That will be very useful. The majority of the hedgehogs that I am monitoring are marked but the markings on some individuals fade with time (even though every effort is made to remark them) and babies are born each year to add to the numbers. These juveniles are unmarked so I think that the publication you mention will help a great deal.

This book might be helpfull: http://www.springer.com/life+sciences/animal+sciences/book/978-4-431-99494-7

Dear Toni,

I hope that you can find some useful informations in this handbook.

I have not read it yet, but its title and contents seem to be in connection with your question.

http://snowleopardconservancy.org/pdf/screen111705.pdf

Dear Toni,

recently spatially explicitly capture-recapture (SECR) models have been developed. I strongly recommend to use the package secr in R. In addition, long capture history (I mean captures belonging to several years) have been used to estimate a lot of useful biological parameters (check Karanth 2006 Ecology).

Finally, there is a recent paper (unfortunately I don't remember the authors name) in which secr are combined with marked and unmarked populations (check something for cougar and Florida).

Good luck for the work

Thanks very much Stefano for taking the time to look out the Karanth publication. I will look at it and also track down the other you mention. With my hedgehogs I don't do anything invasive such as attaching anything to them as I want them to live a totally normal life without impedances. This is why I have gone down the route of using emulsion marks which have turned out to be very successful. With one albino that I released last year, and which I am monitoring, I obviously used black paint. This still stands out as clear as a bell. This particular animal has proved many doubters wrong in terms of survivability. He disappeared from the cameras for 6 months then appeared again on April 9 this year and is on the cameras every night without fail.

No more remote cameras but remote sensing can give some approximation about big mammals, not those which are living deep inside earth or in water bodies like ocean.

Is this using GPS proximity sensors and similar tracking devices?

We have recently used remote camera trapping to estimate the population size of leopards in Namibia. The results where very satisfactory and we are planning to expand the study to estimate the population of the cheetah. We are also using the remote camera trapping to estimate the population of Hartmann's Mountain Zebra at a larger scale. We have piloted the camera trapping of zebra and has compared the results with those of the aerial survey. The outcome was amazing - a very close/similar estimate of the zebra population. So it can be done!

Hi Kenneth, thanks very much for this very useful piece of information. I have been using 3 remote cameras since April 2011 and have found them to be good but haven't been sure how reliable the data are for assessing population density. I released two rehabilitated albino hedgehogs to the area I am monitoring in York, UK, just over a year ago. The male vanished from the cameras on October 9 last year then reappeared on April 7 this year - and every night since. His sister has not appeared on the footage since November last year. A chance conversation yesterday, with someone who lives at the edge of the special monitored area, revealed that the female albino is still roaming the area ( as two albinos have been seen at the same time) despite not having appeared once on any of the cameras. This is great news but very surprising to me that she keeps to such a narrow route and has somehow managed to avoid the cameras. This throws into doubt somewhat the reliability of using the cameras as a tool for assessing population density of hedgehogs. Do you have a publication documenting your recent work with leopards? I would be very interested to read it.

I recommend you to use SCR because you get explicit estimates of animal density instead of population size. You can use "classical" SCR if you can identify all animals; or SCR for both marked and unmarked populations if you can't recognize all animals. This paper is very interesting: SPATIALLY-EXPLICIT MODELS FOR INFERENCE ABOUT DENSITY IN UNMARKED POPULATIONS (http://arxiv.org/pdf/1112.3250.pdf)

Thanks very much indeed for this information Jose.

This is an excellent paper Jose and will be extremely useful in helping me to assess hedgehog density in the area I am monitoring. I particularly like the fact that the model suggested in the paper makes allowances for not always being able to identify individual animals due to similarity of markings or, in the case of my animals, fading of the original mark (made using paint emulsion on the spines which does not harm the hedgehog).

Hi, I used remote video camera traps and SECR to estimate density of wild living and unhabituated chimpanzees, gorillas and elephants, paper just published (link = http://onlinelibrary.wiley.com/doi/10.1002/ece3.670/full). You will also find reference to quite a few other studies which have used SECR in the paper. I think remote cam traps and SECR are a really effective way to get density estimates and it's no problem if you don't capture / identify the whole population, the SECR package (in R) accounts for this. Another useful paper is attached. Hope its helpful.

Dear mam,

First of all for the camera trapping study, you need to have adequate number of camera traps. By using three cameras in a fixed location you can not estimate the population density. Because in the camera trapping sampling the observer (camera trap) is stationary and animals are in movement. Thus here the capture of the animals depend on the movement behaviour, range use, daily movement of animals in addition to that camera flash (since you are using a IR flash camera this problem may not be there) but by curiosity if you go and check the camera every day then your presence will have influence on animals avoidance of camera.

1. First know about the actual area used by the animal, its ranging behaviour, where they go for forage, usual movement path,etc.

2. You can still estimate number of animal by constantly moving camera from one place to other

3. Then divide the total area into smaller sections (perferrably into rectangular grids, then select the best possible site where you will get 100% sure about capture of animals (eg the movement path of the animal)

4. Then place your three cameras in there selected adjacent girds in rows or columns for 15 days for minimum upto 60 days maximum. Then move the camera to next adjacent grids by this you can cover large area for example for 3 to 6 months.

5. You need to keep record of when you are setting up the camera when your are removing, atleast once in 3 to four days you need to check the camera while going to check camera dont walk straight to camera walk from behind.

6. The data recording, protocols where already been referred to you. you can follow that.

7. if you can identify individuals it is well and good if not also you can estimate number based on number of captures.

Thanks and goodluck

Thanks very much for your response. Very helpful indeed. The animals that I am monitoring only appear between dusk and dawn and I remove the camera cards during daylight hours which means that they are not disturbed by my presence. I am restricted regarding where I can place the 3 cameras due to the location of trees for attaching them. Fortunately I can identify most of the animals (European hedgehogs) by virtue of marks that I have made using a small amount of emulsion on their spines. Each mark is unique to each individual hedgehog. To date I have been able to establish the extent to which hedgehogs travel in the area, when they are active and the length of time that they survive. Interaction between animals, mostly considered to be a solitary species, has also been captured on the cameras. Thanks again for all your suggestions.

Hi Toni. In the new book by Royle et al (Spatial Capture-Recapture) you can find interesting pieces of code in R + BUGS to deal some common problems in camera trapping, as miss-identification (blurry photos), and others. You can also find codes for SECR heterogeneity, time and behavior, and interesting applications for Bayesian p-value. You can browse the book at this link: http://www.barnesandnoble.com/sample/read/9780124059399

Hello， everyone, i am trying to study the population density of giant panda using remote cameras, i feel it is difficult to distinguish different individuals, because most pictures of pandas are similar, although we set up over 50 cameras and got hundreds pictures of giant panda.

Hi Jindong. Do you know the home range of the giant panda? If you have some data about that and your grid of camera trap is adecuate (see the conditions in the attached paper) you can use your data and a code for density calculation for unmarked population. (http://arxiv.org/pdf/1112.3250.pdf). I'm working now in density estimation with unmarked otter.

For a gentle introduction to SECR models the following paper is also very useful: Borchers, D. L. A non-technical overview of spatially explicit capture-recapture models Journal of Ornithology, 2012, 152, 435-444