behaviour comparisons at two sites


I have the frequencies of different behaviors of individuals for two sites. I have the individual frequencies for each nest observed i.e. so many minutes brooding and then minutes off the nest and responding to a disturbance. As each observation is of the same length then they could be looked at as frequencies or I could even make it into percentages as needed. I have also calculated the mean for each behavior for each site.

Now I need to know if the frequency of behavior is different for each of the sites. Would it be better to use each individual observation frequencies for each category in turn and use a t-test, or should I use means and do a Chi or G-test?

Also, I am going to then split the results for each age group (i.e 1 week, two week, three week and four weeks after incubation began). Should I yet again compare each behaviour separately or can I use something like a 2-way ANOVA to test for an overall difference in the frequency of behaviors between age and site? I am sorry if I am completely wrong in this but I am relatively new to statistics and when I think I have it, I take a break from stats and when I get back to it it feels new again.

Thanks. I may have more questions in the future (I have already asked some), so please bear with me.