Stephen, You'd be one of the lucky ones if your data actually fit a Poisson. You'll probably find some level of overdispersion and need to handle that with, e.g., a negative binomial or gamma distribution. Either way, for some methods to handle those zeros in a flexible way, check out the excellent vignette in the package 'pscl'. You can find it by Googling "regression models count data R". It will be the first result; a PDF. The package 'zicounts' handles zero-inflated binomial, poisson, and negative binomial, but it seems a bit less flexible and less actively maintained than 'pscl'.
Also, I wonder about the question you're asking... the factors appear to be all about location. Are you modeling any processes, or just looking for pattern? If it's all about pattern, which seems to fit your description of what you're after in variance components, you're really just interested in getting a good model, not "testing." I've always found Ben Bolker's "flow chart" for models helpful - pg. 397 (http://www.zoo.ufl.edu/bolker/emdbook/). Dave Hewitt VIMS, Gloucester Point, VA At 03:38 PM 1/12/2008 -0400, you wrote: >Hello Ecolog - I was wondering if anyone had any advice on the following >problem. > >I have a data set that is infested by a plague of zeros that is causing me >to violate all assumptions of classic parametric testing. These are true >zeros in that the organisms in question did not occur in my randomly sampled >quadrats. They are not "missing data" > >I have a fully nested Hierarchical design >My response variable is density obtained from quadrat counts. >my explanatory variables are as follows > >Region (3 levels-fixed) >Location(Region) (4 levels - random >Site(Location(Region)) (4 levels - random) > >My plan was to analyze the data with a nested anova and then proceed to >calculate variance components to allow me to parse out the variance that >could be attributed to each spatial scale in my design. Since it is known >that violations of assumptions severely distort variance components in >random factors, i would really like to clean up my data set to meet the >assumptions but as of yet i have found no acceptable remedial measure. > >Has anyone else run into this problem when analysing abundance data. I am >aware of conditional models, but i have no practical experience with them >and i am not even sure how to proceed with analysis in that case. I have >been using the R program to tackle this problem and i have also found no >advice on the r-help mailing list. > >Thanks for any help that can be provided > >Stephen Cole >Marine Ecology Lab >Saint Francis Xavier University
