Data for: Bi-modal response strategy in Daphnia to ambush predation risk

Predation, a well-established major selective force, can manifest through either consumptive or non-consumptive effects. The relative impact of such effects is predicted to vary depending on the predator hunting strategy. In aquatic systems, great attention has been paid to coursing predators, such as fish, and their effects on the behaviour of zooplankton. However, less information is available regarding more ambush-style predators. To remedy this paucity, we utilised a 3D-tracking platform to record groups of Daphnia magna under control or predation risk conditions from the ambush, invertebrate predator Erythromma najas. This design allowed us to test if there are anti-predator responses in multiple metrics of swimming behaviours. Furthermore, we recorded actual predation events allowing the distinction between predator exuded info-chemicals and those produced during the predation event. We demonstrate that predation risk was greatest for those that swam at 85% of the available depth and averaged 8.1 mm/s. Examining each individuals swimming behaviour separately shows that predation risk did not exert an impact on any of the prey response metrics. Interestingly, however, we show that Daphnia conform to one of two strategies whilst under predation risk; either swim fast high up in the water column or swim slowly close to the bottom. Hence, this dichotomous behaviour is driven by strategies combining speed and depth in different constellations. As a result, our study demonstrates that Daphnia can detect the presence of a predator, but not immediately the act of predation. In a broader context, our findings highlight the importance of considering both the spatial and temporal dimensions of predation events in order correctly detect anti-predator responses.

The data here is the raw positions of each individual tracked for the above study.

Methods
The information on collection of data is found in the associated manuscript and the only processing in is the concatenation of individual files into a single dataset, included here.