Fig 1.
Overview of snail species used in the experiment.
Snail species used in the experiment with their characteristic body shapes. Prey snail species used in the experiment with A. helena (a) include: Pond snails (Lymnaea sp.; b), Trumpet snails (Melanoides tuberculata; c), Quilted Melania snails (Tarebia granifera; d), and Ramshorn snails (Planorbella sp.; e). Top-down view of a typical experimental compartment (f) containing the predator and the four types of prey (in equal relative abundances). Note that panel f is on a different scale from the other panels, roughly 2:1. Photos credit: Oksana Gonchar (University of Leicester, UK).
Fig 2.
Total prey consumption overall and per treatment per predator.
Total consumption of each prey species by A. helena (n = 97) over all treatments (a) and total consumption of prey snails at each treatment over the 14-day experimental period (b). All prey types were regularly consumed, however different numbers were consumed (dark grey: consumption; light grey: abundance). In (b) prey species are respectively, Ramshorn snails, Malaysian Trumpet snails, Pond snails, and Quilted melania snails. Box-plots with interquartile ranges and median, whiskers are up to the most extreme value within 1.5 times the interquartile range from the interquartile range. Individual points represent individual predators, outliers are indicated by triangles.
Fig 3.
Prey selection in each treatment.
Mean α (±95% CI) of each prey type in each treatment (a-e) based on feeding without depletion. The prey abundance in a particular treatment is indicated in the label of each panel (altogether 8 snails were used in each treatment). In the case where the confidence interval overlaps with the dashed line (expected feeding under random prey selection) there is no selective feeding on a prey species.
Table 1.
Comparison of consumption of prey types within treatments.
Each row shows the P-value for the null hypothesis that there is no difference in consumption between the prey species. Results come from GLMM’s testing the likelihood of individual prey snails within treatments being consumed, based on species. Multiple comparisons were Tukey corrected within treatment. Prey species are Ramshorn snails (R), Malaysian Trumpet snails (T), Pond snails (P), and Quilted melania (Q) snails.
Fig 4.
Weekly prey preference during the four week treatment.
Preference of A. helena for each of the prey species in the four week feeding trial (prey composition: 1R:2T:1P:4Q). Mean values for all snails with ±95% CI. CI’s overlapping with the dashed line (expected consumption under random feeding per prey type) indicated no selection.
Fig 5.
Individual preference of A. helena predators within the population measured in all feeding treatments.
The left column (a-d) shows the relative abundance of the most consumed prey plotted against the abundance of the same type of prey in the environment. The solid black line shows the fitted model. Note that the fitted model is the same in all four subplots. The right column (e-h) presents the relative abundance of most consumed type of prey plotted against the total numbers of prey eaten. Again, the black solid line indicates the fitted model. In all graphs, each point corresponds to an individual A. helena predator. Individual data points are slightly jittered horizontally.
Fig 6.
Histogram of Wi-values with corresponding P-value distribution for A. helena in each treatment.
Petraitis Wi-values (left column) indicate the level of individual specialisation compared the the population diet. Higher values indicate a diet more similar to that of the population as a whole. The population mean deviation is indicated by the vertical dashed line. Histogram of P-values (right column), significant values (<0.05; left of the dashed vertical line) indicate a diet breadth narrow than that of the population, i.e. less variation in prey species consumed.