Ben Augustine

Nailed it

🤔🧐 Some of the better fits. Used dirichlet-multinomial model for telemetry counts in cells, within cell location informs avail sigma. Random RSF slopes for individuals. Need better covariates or spatial random effects, but seems pretty cool. Daily locs, probably need to thin more.

Not sure who this Ben is, but they freak me out weekly--what did i agree to, and who is asking me questions!?

Probably conceptually similar to step selection function except decisions being made from the activity center. Both assume functional form for availability. Of course, this idea here is ignoring correlated observations due to animal movement

Definitely a resource selection probability function.

I guess ability to estimate spatial scale of availability requires a functional form to be specified, so no free lunch

I'm sure someone has done this before, but I need to look around. I think this is Lele's weighted distribution but using a parametric assumption for the availability distribution instead of locations, and allowing one to estimate the spatial scale of avail instead of assuming when making unused locs

I'm biting the bullet and confronting it with some GPS data. Will report back later when the fit sucks.

Here, movement is less costly along the road. Or say it is a ridge. Could model anisotropic HRs where the angle of elongation isn't the same for all individuals, say ridge lines, valley bottoms. Can do this with current LCP models, but they introduce unwanted individual heterogeneity in space use/p

Nothing new here, but inference about landscape connectivity and resource selection could differ a lot depending on how availability is defined. Here, decay is halfnormal, hazard, annular-normal (territorial sps?)

"Your rigorous survey design and statistical methods may say density is low, but last week I was out in a helicopter and saw tons of them"

Row number for road was one off before. So they don't hang out in the road if it is costly to move into it unless there are good resources there to offset the movement cost.

Simulation of home range placement and within home range resource selection with connectivity constraints for availability.

Assuming the road is narrow relative to grid res and animals can spend time in road cells if they contain good habitat, just are less likely to cross them.

I've had someone else say they didn't think it mattered if they didn't account for the dependence between parameters when updating them in MCMC. It absolutely does! In this case, a lot.

Correction to ‘An evaluation of multi‐species occupancy models with correlated species occurrences’ Click on the article title to read more.

Multivariate Probit occupancy model explored recently by Dorazio et al. works much better than initially presented.

besjournals.onlinelibrary.wiley.com/doi/10.1111/...

ah. read that as it took 2 days to fit once

wild it would be so close even if the code was the same

checks out 😆

As search effort accuracy decreases, individual heterogeneity in detection probability is introduced

Exploring impact of search effort resolution for Royle-Young model. How accurate does it need to be? In this scenario, probably 1sigma or less. 2sigma only -4% bias, not bad. 4sigma -8%

compensatory heterogeneity, detection proportional to space use, etc.

e.g., if an animal won't go there, we disallow activity centers from being located there. But the observation model should also reflect that an animal won't go there--it must spend more of it's time in the areas it does go to and detection there should go up relative to if it could go anywhere.

I think both of these are assumptions made about how the process model should behave, but I am thinking about the consequences of these assumptions for the observation model, which typically is the product of within home range space use and detection given space use (combined into det function)

Here, an animal can't cross the fence. So all the time that theoretically could spent on the other side of the fence needs to be allocated back to the side it lives on.

Has anyone ever pointed this out in the literature before?

One thing regular SCR does not account for is if you have a hard boundary like hab mask, detection probability should generally go up inside the hab mask if the detection function extends beyond it. Ignoring this can introduce false individual heterogeneity in p. Proper RSF model is one solution.

This got me wondering who came up with the analogy. Appears to be Bernoulli

en.wikipedia.org/wiki/Urn_pro...

We are recruiting several PhD students. If you are interested in conservation conflicts, population ecology, quantitative ecology, or biodiversity, check these projects out (links in the thread, feel free to reach out):

a man wearing 3d glasses and a hard hat says " like a boss " ALT: a man wearing 3d glasses and a hard hat says " like a boss "

Passing those posterior predictive checks

😂

Plot is only for 10 species, but decent ESS for species correlations in 3 minutes.

Recently came across this Dorazio et al. paper showing that parameter expansion can make probit multispecies occupancy models much more efficient to sample.

besjournals.onlinelibrary.wiley.com/doi/full/10....

I wanted to learn a new trick so I made a nimble version.

github.com/benaug/Probi...