I have paired data for bird nest sites and corresponding data for a random point linked to the nest site. The random site is dependent on the nest site due to pacing x meters at y angle from the nest. To compare vegetation measures at the two locations, I am trying to find the best test. I am unsure if the linkage between the sites disqualifies a t-test. The data are paired because the random site is dependent on the nest site, but it seems a paired t-test is the same subject under two circumstances. A 2 sample t-test doesn't seem right because the samples have a connection to each other. Is Anova the best way, since there would be an effect of the nest site on the random point? Thank you for any guidance.