Two-species model and spatial dependence

[sunita]

I am looking at interaction between two primate species in southern India. I have 8 spatial replicates of 500m each from 286 cells. The sampling was not done on trails but the segments were continuous. I ran the spatial dependence model and compared it with the standard model. The AIC values indicated that the spatial dependence model was better for both species and teh theta0 and theta1 estimates were not equal indicating spatial dependence.
My objective is to look at species interaction, and to run the two-species model. However, currently there is no way to incorporate the spatial dependence while modeling species interaction. How does one incorporate the same?
Can anyone give me some pointers on how to deal with this? Is it possible to run the same in R?
Thanks in advance.
Sunita

[darryl]

Hi Sunita,
As far as I'm aware there's no code publicly available for the model you're proposing (two species with correlated repeat surveys). You may be able to roughly approximate it by defining a survey-specific covariate that =1 if species seen in previous survey and =0 otherwise. It's not the same as the 'spatial dependence' model, but may be better than nothing.

When you say you have 8 spatial replicates, does that mean you have 8 rows of data for each species?

Cheers
Darryl

[sunita]

Thanks Darryl.
We have sampled 286 cells (covering an area of about 650 sq km) - so have 286 rows of data for each of the species. Within each cell we surveyed 8 replicates (total of 4km walked per cell). In a few cells, however, we were unable to survey all 8 replicates.
Sunita