Fig. 3
From: Seasonal trophic controls drive population variability in a foundational marine copepod
Boosted regression tree results for Wilkinson Basin (a) and Jordan Basin (b). Bars give each predator’s relative predictive influence (%), calculated as the normalized frequency at which each predator was selected as improving model fit. Results show spring-summer predator averages being used to predict C. finmarchicus (Calfin) abundance in spring (light blue/green) and fall (dark blue/green). Here, the Spring response variable indicates bottom-up control (Spring Calfin → Predators) while the Fall response variable indicates top-down control (Predators → Fall Calfin).
