[R-sig-eco] Model for zero-inflated species abundance data, allowing for spatial autocorrelation?

[Peter Solymos]

Kay and Alexandre,

INLA approach might be fine, but given the data you described, I would
rather think about what might cause the zero inflation (90% zeros) and
the spatial autocorrelation and pick a model accordingly. For example
if the zero inflation might be caused by low abundance related to low
habitat suitability, underlying occupancy/range pattern, or detection
error, or some combination of these. Depending on the cause of the
zeros, the treatment of autocorrelation might change. If you have all
nonzero counts clumped together, it might be explained by a spatial
covariate effect. If not, 10% nonzero count in the range of 1-6 won't
necessarily allow a meaningful estimation of spatial autocorrelation
(of course depending on sample size). On the other hand, if you have
nonzero counts scattered in the landscape, estimating autocorrelation
again will be difficult as it roughly decreases at a rho^distance
rate.

Detection error is a whole other can of worms, but using single visit
(without temporal replication) to sites actually allow to correct for
detection error as long as covariates are available (see svabu
function in detect package and references on svabu help page, this
implements ZIP model with detection error and without spatial
autocorrelation).

Cheers,

Peter

Péter Sólymos
Alberta Biodiversity Monitoring Institute
and Boreal Avian Modelling project
Department of Biological Sciences
CW 405, Biological Sciences Bldg
University of Alberta
Edmonton, Alberta, T6G 2E9, Canada
Phone: 780.492.8534
Fax: 780.492.7635
email <- paste("solymos", "ualberta.ca", sep = "@")
http://www.abmi.ca
http://www.borealbirds.ca
http://sites.google.com/site/psolymos



On Tue, Feb 14, 2012 at 5:25 PM, Kay Cichini <kay.cichini at gmail.com> wrote:
> Alexandre, many thanks, there seems to be plenty of useful content on the
> linked site!
>
> Am 14.02.2012 14:56 schrieb "Alexandre Villers" <villers.alexandre at gmail.com
>>:
>>
>> I answered by mistake on the R SIG GEO.... sorry for cross posting.
>>
>> ---------------------------------------------------------------------
>>
>> Hi,
>>
>>
>> If you are not afraid to switch to Bayesian methods (but fast ones !),
> the INLA package could do that for you. ( www.r-inla.org )
>> Cheers
>>
>>
>> Alex
>>
>> On 14/02/2012 15:36, Kay Cichini wrote:
>>>
>>> Hi all,
>>>
>>> I'd very much appreciate pointers at methodology/R-packages/functions
>>> suited for modelling zero inflated species abundance data (counts) that
>>> most likely will show spatial autocorrelation. Species abundances range
>>> between 0 and 6, the proportion of zeroes is about 90%. I have about 6
>>> nominal covariables (partly heavily unbalanced).
>>>
>>> Yours,
>>> Kay
>>>
>>>        [[alternative HTML version deleted]]
>>>
>>> _______________________________________________
>>> R-sig-ecology mailing list
>>> R-sig-ecology at r-project.org
>>> https://stat.ethz.ch/mailman/listinfo/r-sig-ecology
>>>
>>
>> _______________________________________________
>> R-sig-ecology mailing list
>> R-sig-ecology at r-project.org
>> https://stat.ethz.ch/mailman/listinfo/r-sig-ecology
>
>        [[alternative HTML version deleted]]
>
> _______________________________________________
> R-sig-ecology mailing list
> R-sig-ecology at r-project.org
> https://stat.ethz.ch/mailman/listinfo/r-sig-ecology
>