Logistic regression problem

Hi there,
I'm doing some comparisons between characteristics of used and unused nest trees of birds. I have 155 nests and 155 randomly chosen trees without nests. I've collected 6 independent variable that I think should influence the use of the trees by birds. When I run the logistic model with all 6 variables I get 2 that are signifcant (i.e., p<0.05); however,the beta coefficients are extremly small (i.e., SPSS cuts them off and prints out 0.000), plus the confidence intervals overlap 0. How can these be significant variable according to the p-value, yet other indices suggest they have a very limited (or non-existent) effect?
Thanks for the help.