Dear all, I try to fit MANYGLM models to my species community data. I am working with bird data and want to see if forest restoration lead to changes in bird communities. I have 4 treatments and 10 stand per treatment. I work with presence/absence data. The 40 plots I use differ in size (between 3-22 ha), therefore I want to correct for plot-size in the model. My first question is if it is possible to include a covariate (plot-size) in a MANYGLM model. Until now I treated MANYGLM just like a normal GLM and expected that the model correct for plot-size but I am not sure how it works. Model: < manyglm(birdmva~Area+Treatment, data=x, family="binomial")
Secondly, I also want to know which species contribute to the differences in community structure. But I am a little confused about the univariate test! First (if possible in MANYGLM), what to do with the covariate. Secondly, how do you determine the order of the variables in the model, if you change order in the global MANYGLM model, univariate outcome chance significantly. Both multivariate test and at species level. < anova.manyglm(model, test="LR") Example with spider dataset: Univariate test model 1 Analysis of Deviance Table Model: manyglm(formula = spiddat ~ moss + soil.dry, data=x, family = "neg.binomial") Multivariate test: Res.Df Df.diff Dev Pr(>Dev) (Intercept) 27 moss 26 1 71.86 0.001 *** soil.dry 25 1 104.51 0.001 *** Univariate test model 2 Model: manyglm(formula = spiddat ~ soil.dry + moss, data=x, family = "neg.binomial") Multivariate test: Res.Df Df.diff Dev Pr(>Dev) (Intercept) 27 Soil.dry 26 1 147.30 0.001 *** Moss 25 1 29.07 0.068 . Moss is significant in the first model but not in the second! This is the first time I work with MANYGLM, thus any assistance would help. Thanks in advance. Martijn Versluijs __________________________________________________________________________________ PhD-Student Department of Wildlife, Fish, and Environmental Studies (Vilt, fisk och milj�) Faculty of Forest Sciences (Skogsvetenskapliga fakulteten) Swedish University of Agricultural Sciences (SLU) S-901 83 Ume�, Sweden [[alternative HTML version deleted]]
_______________________________________________ R-sig-ecology mailing list R-sig-ecology@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-ecology