Scientific method: making predictions knight used natural differences among ponds in the university of florida’s ordway-swisher biological station to create a unique field experiment. she selected 8 ponds that were similar in size and depth. four of the ponds contained fish, but the other four contained no fish. a diagram showing the experimental setup. four ponds with fish. four ponds without fish. when fish are present in ponds, they consume the aquatic larvae of dragonflies. adult dragonflies are voracious predators of other flying insects, including various flies, bees, and butterflies. flies, bees, and butterflies pollinate a shrub called st. john’s wort that grows around all the ponds. knight predicted that the presence or absence of fish would cause variation in the abundance of dragonflies, which might influence the abundance of pollinators, thereby affecting the reproductive success of plants that depend on those pollinators, such as st. john’s wort. her prediction took into account previous research that had demonstrated the type of interactions that occur between the organisms in her study system. drag the pink labels to indicate which type of interaction occurs between each pair of species. based on those interactions, drag the blue labels to indicate the population changes knight predicted for each species in the two different treatments. labels can be used once, more than once, or not at all. fish: a. predation, dragonflies: b predation, insects: c. mutualism dragonfly population: d. goes down, insect population: e. goes up, st. john's wort population: f. goes up, dragonfly population: g. goes up, insect population: h. goes down, st. john's wort population: i. goes down