help radfit()

Dear all,

I am working with radfit() function in vegan with a df which contains a big
amount of zeros. When is not zero, I have numbers with comma, which
represent proportions. I am using the family Gamma. I got negative values
for some models. Below an example of the output:

      Null  Preemption   Lognormal        Zipf  Mandelbrot
37  694.8004  -413.17581  -565.23622  492.945299  -482.93299
38  415.2763  -340.93024  -363.86510  469.338760  -379.28155
39 1221.0254  -310.94620 -1490.60959 1131.261212  -342.56346
40 1448.9174  -328.61790  -794.12383  761.840479  -439.40301
41 1007.8350  -376.34689  -493.21855  455.072273  -486.60574

Someone can explain if is a problema to get this negative values?

Thanks for your help,

Manoeli