In the PGLS script, I used normalized percent detections as response variables, scaled median_startTime, logged and scaled median_peak_freq, and included univariate models with single predictors and multi-variate models with different pair combinations of predictors. In this, I don't get a significant value for median_startTime in any of the models, but there is a marginal significance when trophic niche is used as a single predictor.
There are only minor text edits in the other scripts.
Hi Vijay,
In the PGLS script, I used normalized percent detections as response variables, scaled median_startTime, logged and scaled median_peak_freq, and included univariate models with single predictors and multi-variate models with different pair combinations of predictors. In this, I don't get a significant value for median_startTime in any of the models, but there is a marginal significance when trophic niche is used as a single predictor.
There are only minor text edits in the other scripts.
Excited to hear your thoughts!
Thank you very much, Pavithra