Fig 1.
Experimental tank setup consisting of a tank with dimensions of 30 x 30 x 30 cm, two high-speed cameras in a fixed stereo rig, LED lights for illumination and a ventilator to prevent condensation on the glass.
Fig 2.
Manual tracking analysis with Argus of left and right camera recording; tracking points: eye of the larvae (red), snout-tip of the stickleback (blue), tracking period: first hunt over 700 frames (5 s). Here the whitefish larva escaped. The setup partially mimics the pelagic situation of Lake Constance.
Fig 3.
Resulting data points of the 3D analysis were smoothed in Python 3. Depicted are different stages of a failed predation (stickleback: blue, whitefish: red). Tracking started 3 s before capture or escape (400 frames).
Fig 4.
To identify changes in swimming speed in successful and unsuccessful whitefish hunts, the first breakpoint was selected using segmented linear regression analysis of the three second period before the biting event or attempt.
Fig 5.
Predator and prey tracks before and after an unsuccessful hunt and before a successful hunt.
The biting attempt of the predator was at time = 0 s. Tracks were divided into two prey fish categories (whitefish, and roach & perch). The speed in cm/s and distance in cm was derived from smoothed 3D tracks of prey and predators. Tracks were averaged over each species and outcome category. The grey-coloured period of predation describes the hunt of the stickleback.
Table 1.
Performance data of predator and prey fish during the hunt.
Fig 6.
Maximum and average speed of prey.
Maximum and average speed of prey (cm/s) in failed hunts divided by species and size class.
Fig 7.
Average turning angle of the predator.
Average turning angle of the predator during the failed hunts of whitefish over the four size classes of prey.
Table 2.
Variables predicting the successful outcome of the predation from a predator’s perspective.
Fig 8.
Output of the multivariate logistic regression depicted the likelihood of the predator to capture the prey in relation to A) the average acceleration of the prey and B) the average speed of the prey, N = 75 predation events, including all prey species and outcomes.
Fig 9.
Projected predator and prey relationships.
Summary of the observed relationships between predator (stickleback) and prey (whitefish) behaviour at different prey length classes: A) Hunting success of sticklebacks decreases with prey size, which was shown to be correlated to B) an increase in complexity of predator avoidance behaviours. A and B are based on previously published data from the same experiment [33]. The current study indicates that: C) predators more often adjust their swimming direction when following larger whitefish larvae; D) larger whitefish are increasingly adept at detecting the start of the hunt as measured by an increase in swimming speed; B+D could explain why predators increasingly need to adjust their track C) which in turn might decrease hunting performance (A).