Animal Movement - Science topic

Hi, I have the same issue! Did you ever figure it out?

Hi Wren. How many (approximately) rays are you going to be looking at? It is all fine and well to try to distinguish between rays without markers at numbers, say, less than 4 -5, but once you get beyond that number it becomes a nightmare! I would strongly suggest that you consider a marking strategy, not only to be able to produce rigorous science, but also for the sake of your sanity! Having 'been there' in terms of identifying individual elasmobranchs on sight, I would highly recommend making your life easier and keeping your data unambiguous by using markers. I hope this helps! Best of luck.

I used R package GLATOS to build Random Walk simulated tracks constrained into a defined polygon

See applcation in our paper:

Hi Colin,

Sorry for the slow reply -- I've been thinking a lot about your question. I study cow behavior and have seen others use area explored in the event of variable GPS time-lags. I think you'd find a lot in the literature about large ungulates that deals with this problem. That said, I think it would still be applicable to your cobra example in that you are tracking multiple snakes for multiple days throughout the breeding season (and perhaps across years, too?); I suspect that a daily MCP calculation for four fix locations per day between 0600-1800 would exaggerate the area explored, but would also offer a consistent metric between each snake and season. This way you could average the MCPs per season to see their general area explored. I think you could make a solid argument that a larger area explored would be facilitated by moving more. There is always the case that the animal is moving a lot in a small area and this would be the hardest thing to rule out, but I can't think of another metric that would account for the time-lagged dateset.

There is a easy way to do this in PYTHON.

The function called "resample" and "groupby" are most used function for time series analysis.

I second David's suggestion. Just to add: If you have really large data sets (TB) and or you want to run really complex simulations which may be parallelized, you might consider using a computer cluster. Universities often have such clusters, and there are possibilities to by access to commercial clusters. It saves you own investment and administration.

Check

1. www.pangolinsg.org;

2. A conservation assessment of Smutsia temminckii

3. Ethnozoological Survey of Traditional Uses of Temminck’s Ground Pangolin (Smutsia temminckii) in South Africa

4. IUCN Pangolin Specialist working group

After some additional searching, I've found the "speedfilter" function in the new (April 2019) "trip" package to work well. More info here:

Tongue on cheek, some might say the invasive animals are not what (or rather who) we think they are. Edward O. Wilson's (2017) most impassioned book to date argues we need to dedicate fully half the surface of the Earth to nature. That said, responses to this quite commonsensical query might be enriched by a look at the effects of conservation corridors.

Wilson, E. O. (2017). Half-Earth: Our planet's fight for life. Liveright.

Janelle Goeke

We offer a secure cloud-based app that allows you to automate your crab behavioral research.

It can be accessed anywhere in the world by a computer, smartphone, or tablet without downloading anything.

Wouldn't it be amazing to do your behavioral analysis in front of the TV watching Netflix...

You don't need anything fancy in order to get started, typically you can set up your smartphone, and let it record, while you're in a separate room. There is no fear of an interruptance caused by a researcher, etc.

Although we are not an open-source program. Our prices are very competitive when compared to Noldus' EthoVision. You can even share your data with your collaborators.

Please feel free to start your free trial at www.behaviorcloud.com or email me at chloe@behaviorcloud.com for more details.

Hi, it appears that no one has really answered your question, so I am going to take a stab at it and hopefully it helps. If you are using Esri products in the Geoprocessing tools there are the Spatial Analyst and Tracking Analyst extensions that are really helpful. You can build a model, through modeling building and test how it works. They do require some level of knowledge with ArcGIS and GIS. However, there are online tutorials for ArcGIS extensions, and YouTube videos, for free. I have seen in other forums people suggesting Hawths’ tool for spatial ecology, but I am not familiar with that software at all that I am aware of.

or if you read enough which most college students do check out the choices in YouTube videos

https://www.google.com/search?rlz=1C1AKJH_enUS704US704&ei=Ff6aXMHjIIGK5wK30KeIAQ&q=Tracking+analysis+site%3Ayoutube.com&oq=Tracking+analysis+site%3Ayoutube.com&gs_l=psy-ab.3...114159.118694..120182...0.0..0.66.931.15......0....1..gws-wiz.......0i71.BRqbCIRygBg

Also: a book is out on explaining Modeling suitability, movement and interaction

I hope this helps, Good luck with your project.

Indrikis Krams We have developed an affordable option compared to Noldus, Anymaze, Cleversys, and allows you to access from any device as well as share with colleagues and collaborators. We also provide great technical support, training, and consulting services as well. Please check us out at www.behaviorcloud.com

Thank you for your help! I will take a look at this package.

Regards

You are most welcome.

@ Sang-Il Oh : In the computer vision field, object(s) tracking can be applied to solve your problem.

Yes, definitely. By triangulating (or tessellating) the frames in a video and applying shape detection, analysis and classification techniques, it is possible to organized video frames into clusters containing objects (e.g., ground beetles) with similar behavior. For more about this, see

Foundations of Computer Vision

Dear Jacob, make sure to have look at the recent papers by Justin Calabrese and coworkers. They use time-continuous, not time discrete ones, to fit tracking data, which allows them to better take autocorrelation into account. They also developed a powerful R script to actually use this new approach. I am not an expert on all this, but it seems to me that this is the best you can find in movement ecology. Cheers, Volker

the 'amt' package in R will do just that, see this preprint:

Hi Renee,

For apply the method that I am talking about, you need to have camera-trap data. This method have been applied in rodents, and I have applied it in ungulates, with nice results.

See "Bias in estimating animal travel distance: the effect of sampling frequency, Rowcliffe et al 2012" if you want more details.

I agree with you Ronny. We have to accept some dependence (literal meaning) between animal location at the time of a previous fix, where it is now, and where it can potentially go before the next fix. But I understand mathematical people focus on something different, statistical independence of points as a basic assumption of kernel density estimation. Nevertheless, in many previous animal tracking studies it was acceptable to use kernel methods with long intervals between fixes (reducing the issue of auto-correlation), and could okay for your fixes 3 times per night with 2 h intervals. Newer methods like aKDE from Fleming et al and Brownian Bridge movement models I think are more important for people who get high frequency fixes. Might be interesting to try both with your owl data. It sounds a very interesting study, best wishes for your analysis.

Dear Jared,

I second everything Julien said. If you know R or can afford the time to learn from the vignettes of the packages, use R (but indeed it would be critical to know what you want to do to direct you more precisely). If you don't know R, just have a few HR to compute, and you are fine with location-based kernel HR estimate (which are generally ok for a set of independent locations), you may want to try this:

which we made at the lab for user-friendly simple HR estimation (Credits go to C. Bernard). It's an online app, it runs R adehabitatHR package in the background, will allow you to play with the smoothing parameter, and you can select the %UD you want to get. You can export a shapefile of your UD.

Best,

simon

Hi Julia,

I have been using e-obs collars on Eurasian lynx in a montane area in Turkey. To my experience if you have severeal high hill tops ascendant to your study area it is very easy to download the data when your animals are in the inactive phase of the day. I could download data many times over more than 5kms away and once even from 16kms from a mountain slope to a hill top directly looking to each other!

I agree with Miha that if the animal is moving during the download, connection breaks very often. So, I always try around 11am to 3pm when the animals are generally inactive and resting on slopes.

As far as I know e-obs is very successful on bird tracking as well.

Cheers,

Deniz

SkillSpector can also be usefull. 

If you want species specific information try: http://www.fishbase.org

It's a good starting point.

Dear Dr. Ashneel Ajay 

Thank you for your suggestion and send the article. 

Thank you

Hey Mariettte,

I just refer to my question and add two additional links.

They USE the UTM

All the best

Leon

If a passive drop-off system would meet your needs, you might look at the attached publication.

Dear Bernardo,

You should always worry about residuals. If you are using GLMs (generalized linear models), then you need not assume iind (independently identically normally distributed) residuals. However, depending on the link function, you must assume some other residual distribution function. Looking at the residuals after fitting the model allows you to check how well the residual structure fits the assumption you made.

Some other comments: even in GLM, the parameters of the assumed distribution of the residuals must be the same across all predicted values. Sometimes (often in biology), this is not the case: the larger the predictor values, the larger the variance in the predicted values. If you have several predicted values for a predictor value, then you can estimate the variance at each predictor variable and perhaps scale accordingly.

The other issue has to do with the relation between sum of squared residuals and AIC (or AICc, depending on how many data points you have). I cannot write the general formula here (no formula editor available), but the general formula for the likelihood, not assuming normality, is available in Burnham and Anderson. Perhaps someone has programmed is in R (which I do not use). You will have to derive it, perhaps, depending on your link function. The warning: do not use the derived term (-n ln(RSS/n) + ln(2 pi) +1) in the formula for AIC (AICc), because this formula is valid for iind residuals.

Maybe some of this needs to be tailored to your modeling random walks.

If I can help any further, please contact me. Hermann

You can also try AnyMaze (expensive). There is also IDTracker, which runs on Matlab 2014 or higher, which is free. You need to download and compile it. 

Dear Gareth,

As far as I know, there are not many things published on that topic in Central Asia, and of course only in Russian. 

you have that book chapter in English, a bit more recent than the 1970 one: Bibikov, D.I. (1982) ‘‘Wolf Ecology and Management in the USSR’’, in F.H. Harrington and P.C. Paquet (eds) Wolves of the World. Perspectives of Behavior, Ecology and Conservation, pp. 120–33. Park Ridge, NJ: Noyes.

I got a copy of Vyrypaev & Vorobjev book on wolves in Kyrgyzstan:

Vyrypajev VA, Vorobjev GG. 1983. Volk v Kirgizii. Frunze: Ilim. 94 p.

However I would be cautious in using it since I detected some mistakes in the tables they give inside.

I published two papers about human-wolves relationships in Kyrgyzstan (they are available on RG, and you may find something in the references...), as well as a PhD in French. If you have any question about that specific topic, you are welcome to contact me!

Nicolas

Hi Miquel,

one solution is to synchronize time between your GPS and reader, and to have the GPS recording continuously the tracking path. When you export the data of your reader and GPS, you merge the two data files (GPS track and detections) by date/time. Maybe not the best solution, but it works fine! :)

See the description in the following paper: https://www.researchgate.net/publication/274704028_Linking_lab_activity_with_growth_and_movement_in_the_wild_explaining_pace-of-life_in_a_trout_stream

Dear Sam, thanks so much for your reply. I will try to find out more about it. Best, Stefano

I have a paper coming out with Lyons in Behavioral Research Methods describing how to use the Microsoft Kinect gaming device. You should observe the animal, then you need to writ the scoring program, so it is not off the shelf. The Kinect costs several hundred dollars. 

Another alternative is the Geospatial Modelling Environment from SpatialEcology.com. The download is free. Look for the command "movement.pathmetrics"

Thank you, I've sent an email :)

You could measure the stride length for each step cycle - 3 consecutive imprints (or sets of manual + pedal imprints in case of quadruped animals) = 1 step cycle, 11 imprints = 9 step cycles - and then do an averaging over all step cycles (mean stride length).

If the trackway pattern shows considerable variation (e.g. curves) you should rather calculate distinct velocities for different parts of the trackway and discuss that.

I guess you recover the devices. Then consider exploring the I-GotU,  a small, cheap (U$S 50) and good GPS originally designed for trekkers. You can made a cover to protect it find the way to deploy them in a small mammal. Our group used them with coromorants and worked well.

hello Nikolaos

I used Kinovea (freeware) to retrieve the target coordinates in 2D on both camera.

Then I concatenate two 2D data with a DLT (direct linear transformation) to obtain 3D with matlab, labview or openCV (freeware) script

Abdel- Aziz Y. I, Karara H.M., 1971. Direct Linear Transformation from Comparator Coordinates into ObjectSpace Coordinates.ASP Symposium on Close- Range Photogrammtery, Falls Church, VA, pp.1-18.

enjoy

Patrick

Institut of Mouvement SciencesMarseille, France

Hi Anne,

Did you ever find a solution?

I'm only just reading this now, but I think the function you are looking for is actually gdltraj()

Sure.

There are many studies that used stable isotope of C and N isotopes for this.

Best and Schell (1996) with stable isotopes of carbon and nitrogen indicated the seasonal movements for southern right whale Eubalaena australis.

Ontogenetic migration of maturing and mature male of sperm whale Physeter macrocephalus to high latitude by segregating from natal groups in low latitudes, which was apparent but not conclusive for long, was decoded by Mendes et al. (2007) through stable isotope ratios of carbon and nitrogen from dentine collagen.

There are several other studies also, especially of Turtles, that underscored their migratory dependencies.

Migratory dichotomy, practiced by the sea turtles to understand their behaviour, ecology and demography. Zbinden et al. (2011) though the aid of stable nitrogen isotope discriminated the two foraging regions area preferred by the loggerhead seas turtle Caretta caretta in the Mediterranean region to understand the relative importance of geographically separated foraging regions and associated phenotypic variations accruing from this discrimination.

Similar, but ontogenetic dietary based oceanic–to–neritic migration has been confirmed by McClellan et al (2010) for juvenile loggerhead sea turtles, which has important bearing for the loggerhead individual survivorship, stage duration and time of maturity.

1. Best, P. B., & Schell, D. M. (1996). Stable isotopes in southern right whale (Eubalaena australis) baleen as indicators of seasonal movements, feeding and growth. Marine Biology, 124(4), 483-494.

2. Mendes, S., Newton, J., Reid, R. J., Zuur, A. F., & Pierce, G. J. (2007). Stable carbon and nitrogen isotope ratio profiling of sperm whale teeth reveals ontogenetic movements and trophic ecology. Oecologia, 151(4), 605-615.

3. Zbinden, J. A., Bearhop, S., Bradshaw, P., Gill, B., Margaritoulis, D., Newton, J., & Godley, B. J. (2011). Migratory dichotomy and associated phenotypic variation in marine turtles revealed by satellite tracking and stable isotope analysis. Marine Ecology Progress Series, 421, 291-302.

4. McClellan, C.M., Braun–McNeil, J. Avens, L., Wallace, B.P. and Read, A.J. 2010. Stable isotopes confirm a foraging dichotomy in juvenile loggerhead sea turtles. J Experiment Mar Biol Ecol, 387, 44–51.

Also follow K.A. Hobson work related with tropic linkages

Dear Luis,

Unfortunately, there is no easy way to measure bite force.

An accurate form to measure bite force is using piezo force transducers mounted on a something that the animals can bite. Someone from the biophysics or engineering department should be able to build that for you.

However, that build will only you give the maximum bite force. You mentioned that you want to measure the bite force in different prey, and that is not an easy task. Perhaps you can mount the force transducers in a animal-shaped cast or something like that.

For more information I suggest you this paper:

Meijden, A.; Langer, F; Boistel, R; Vagovic, P; Heethoff, M (2012) Functional morphology and bite performance of raptorial chelicerae of camel spiders. Journal of Experimental Biology 215, 3411-3418.

I hope this helps.

Best wishes and good luck.

Are you tracking videos? I have used this tracking software ^ too. If you have a pc and not a mac, the programme "Kinovea" (http://www.kinovea.org/) does an even better job. Although I like "Tracker" for its simplicity (and mac compatibility), I used it to analyse locomotion (specifically velocity) in Northern Quolls.

Hi. You can use the R package "scrbook" to simulate same dataset under different densities, trap configuration and maybe home range from target species. This is the link to download the package: https://sites.google.com/site/spatialcapturerecapture/home

In R, try:

library(scrbook)

?scrbook # Look for SCR0bayes

I also used DayGlo a few years ago but the operator would not sell me the powder. I learned from another person that they will send you "samples" upon request. That's free! I received enough to try out a few colors and actually do a complete project. If you're in the experimental phase of learning what color is best, they may send you samples if you simply ask. From what they told me, it's just the powder before the paint is made. So this may be the same with any company that makes or uses fluorescent paint. If you find other means, please keep us updated.

A plant that could move ever very slowly would seem to have many advantages so I am wondering why this has never evolved and if it would be possible to evolve such plants.

Problem solved :) Simon Benhamou (who developed the method) provides a software package which does the job, and he has been very generous with suggestions, and offers of assistance should I need it:

You may wish to pick up the thread in the literature for DEMP, Density Equalizing Map Projections. These can be thought of as "rubber sheeting" a surface so that the density of points on the surface is then of uniform density. You may also wish to read a bit about Cartograms, of which DEMP is a special case. It also is worth noting that most currently used spatial analysis methods can adjust for the underlying density of the thing being measured using heterogeneous Poisson models.

Choosing openCV for your project seems to be a good choice for your setting. If you have already an experienced programmer doing the coding you are lucky. Nonetheless the overall costs of doing the customization for a single project are likely to exceed the 5 Tsd $ pricetag for the programmer, if you need to hire someone in Germany (50-100 hrs work). If you don't want to get to expensive with the software you might also think about continuos manual tracking using an video overlay or timelapse images for movement control. Here is one approach based on the ethovision software Brendele J., Annas E., Weirich C., Hoy S. (2009): Results of computer supported measurements of distances moved by VideoMotionTracker. Proceedings of the 41. Internationale Tagung Angewandte Ethologie, Freiburg, Germany, but the same effort can be coded with reasonable workload without using ethovison for the backgroundlayer and tracking.

Please keep us updated, how openCV managed the job.

Test if your beetles walk in the glass of the petri dish, that is, all the flour is above them. In this case, you could use a regular video tracking from bellow, not above the flour.

Hi, yes should be 45°. I agree with you. Position the mirror slightly above the lizard with the camera positioned laterally slightly above the height of the animal. Now you can record the lateral and the dorsal aspect with one camera simultaneously... I would try to avoid oblique views. Cheers