[R-sig-Geo] Estimation of utilization distribution and home range within a polygon

[Tom Philippi]

Manuel--
I don't think that there is one single "right" method for estimating
homeranges or spatial utilization contours with spatial boundaries.

Before deciding on the computations, I would encourage you to carefully
think about and define how movement & spatial utilization react to the
polygon boundaries.  Are the boundaries the equivalent of a sheer wall,
where individuals use space normally right up to the boundary, or is there
a gradient of repulsion extending some (short) distance from the boundary
(e.g., too shallow water) so use drops off a bit before the boundary, or is
the boundary attractive or reflecting, where potential movements across the
boundary become repeated head-banging against the boundary (e.g., desert
tortoises and other herps hitting drift fences)?

At the micro-scale, your pdf from each fish position is something like a
2-d Epanechnikov or normal bump of volume 1/N.  If that bump spans (or even
gets near to) your boundary, you probably want the volume to remain 1/N.
The options may be to clip at the boundary & renormalize by a scalar to
retain 1/N, or shift to an anisotropic kernel elongated along the
boundary.  Just doing the full  kernelling and then clipping the homerange
reduces the weight given to that observation near the boundary relative to
the other observations, biasing your estimated utilization away from the
edges.  For places like the Everglades where water depth, nutrients, and
vegetation are very different near the edges of wetland compartments, that
bias would be seriously misleading.

 Perhaps none of this matters for the spatial scales of your fish, boundary
effects in your wetlands, and the frequency of position locations of
individual fish (with 1000 observations per fish, fractionally
downweighting a couple of observations has little effect).

My suggestion can be thought of as an expansion of Mike's: "You can just
clip after the fact, but as you surely know that defeats the
purpose if the boundary has meaning to the intensity of behaviour" above to
cases beyond just clipping.  Mike is both more succinct and faster at
responding than I am.

Tom 2

On Mon, Aug 15, 2016 at 8:38 AM, Michael Sumner <mdsumner at gmail.com> wrote:

> It sounds like maybe your input constraint is not in a form that kernelUD
> likes, perhaps there is room to read about the error message and explore
> the details?
>
> You otherwise will need more specialist methods to model within a complex
> boundary.
>
> The soapfilm tools in mgcv is one example:
> http://www.fromthebottomoftheheap.net/2016/03/27/soap-film-smoothers/
>
> Some of the trajectory/ tracking packages probably have stuff, review the
> packages in the list here:
>
> https://cloud.r-project.org/web/views/SpatioTemporal.html
>
> There are few other related packages on CRAN from recent times not on that
> list, and many there have had some significant development so worth
> exploring.
>
> You can just clip after the fact, but as you surely know that defeats the
> purpose if the boundary has meaning to the intensity of behaviour.
>
> (tripEstimation does facilitate this generally,  but I'd say that's much
> too specialist and buried too deep to be helpful).
>
> Cheers, Mike.
>
> On Tue, 16 Aug 2016 at 01:08 Manuel Spínola <mspinola10 at gmail.com> wrote:
>
> > Dear list members,
> >
> > Is there a way to estimate utilization distribution and home range area
> > within a polygon.  I am working with spatial location on a fish in a
> > wetland but the utilization distribution and the home range cannot go
> > outside the water boundary.
> >
> > I tried using a spatial polygon as spatial line and specify this in the
> > boundary argument of the kernelUD function of the adehabitatHR package
> but
> > I got an error:
> >
> > > ud = kernelUD(dfs_in[,2], h="href", grid = 100, boundary = riol)
> >
> > Error in .boundaryk(SpatialPoints(x, proj4string =
> > CRS(as.character(pfs1))),  :
> >   non convenient boundary: turning angles > pi/2
> >
> > "riol" is the spatial polygon coerced to spatial line
> >
> > Thank you very much in advance
> >
> > Manuel
> >
> > --
> > *Manuel Spínola, Ph.D.*
> > Instituto Internacional en Conservación y Manejo de Vida Silvestre
> > Universidad Nacional
> > Apartado 1350-3000
> > Heredia
> > COSTA RICA
> > mspinola at una.cr <mspinola at una.ac.cr>
> > mspinola10 at gmail.com
> > Teléfono: (506) 8706 - 4662
> > Personal website: Lobito de río <
> > https://sites.google.com/site/lobitoderio/>
> > Institutional website: ICOMVIS <http://www.icomvis.una.ac.cr/>
> >
> >         [[alternative HTML version deleted]]
> >
> > _______________________________________________
> > R-sig-Geo mailing list
> > R-sig-Geo at r-project.org
> > https://stat.ethz.ch/mailman/listinfo/r-sig-geo
>
> --
> Dr. Michael Sumner
> Software and Database Engineer
> Australian Antarctic Division
> 203 Channel Highway
> Kingston Tasmania 7050 Australia
>
>         [[alternative HTML version deleted]]
>
> _______________________________________________
> R-sig-Geo mailing list
> R-sig-Geo at r-project.org
> https://stat.ethz.ch/mailman/listinfo/r-sig-geo
>

	[[alternative HTML version deleted]]