Figure 1.
Conceptual model of prey responses to multiple predators.
The figure depicts a prey responding with different traits to two predator types that are functionally equivalent in their initial step of prey acquisition (i.e. identical feeding mode C) and functionally inverse in their further feeding modes (i.e. opposing feeding modes A and B). At any combination of density of predators the prey expresses the non-specific defense trait, c, at the same level simply because it perceives cues for predation risk from any of the predators that share feeding mode C. As the risk from the predator with feeding mode A increases, the prey expresses a specific defense trait, a, that optimally protects against that single predator. For instance, assume that the concentration of cues in the form of kairomones or alarm cues from conspecifics being consumed by a certain predator type increases, then the prey develops certain phenotypic defense traits to avoid consumption by that specific predator. The same holds for the trait protecting against the predator with feeding mode B: the prey expresses a specific defense phenotype, b, when perceiving increased predation risk from that single predator. When the prey perceives the risk from both predator types as equal, then theory predicts it should express an intermediate phenotype, a-b. The intermediate phenotype describes a defense that lies between the extreme ends of the phenotypes expressed in response to only a single predator i.e. a phenotype that provides equally bad or good protection against either predator type.
Figure 2.
Zebra mussels' phenotype expression changes with predation risk.
Anti-predator responses of zebra mussels from the four experimental treatments: either a single crayfish (crayfish symbol), a single roach (fish symbol), both predators combined (both symbols next to each other) or no predator at all (minus symbol). Different lower case a and b letters denote significant differences for p<0.05 after post-hoc tests. Error bars denote 95% confidence intervals A: shell shape expressed as a principal component PC1 explaining 77% of the total variation in the contour shape of all analyzed mussels (the contours on the y-axis depict ±2SD of the mean of PC1). B: shell strength as resistance to crushing in the form of the residuals from the correlation between mussel size and strength. C: growth measured in mm increase in the three shell dimensions length, height, and width compared to the start of experiment. D: Attachment strength as inferred from the number of byssus threads with which individual zebra mussels were attached to the surface.
Figure 3.
Shell strength and growth are correlated in zebra mussels.
A: experimental data as mean growth (measured in mm increase in the 3 shell dimensions length, height, and width) in relation to the corresponding shell strength means from the 28 experimental tanks with 16 mussels each. B: field data as individual growth in the first year post settlement in relation to shell strength of 556 mussels. Dashed lines denote 95% confidence intervals for the fit.
Figure 4.
Shell shape and shell strength are correlated in zebra mussels.
A: experimental data from 448 mussels. B: field data from 756 mussels. The contours on the y-axis depict ±2SD of the mean of PC1. Dashed lines denote 95% confidence intervals for the fit. Note that we here present individual values for the experimental mussels since we obtained both a shape and strength measure for each mussel separately.
Figure 5.
Shell shape and growth are correlated in zebra mussels.
A: experimental data as mean growth (measured in mm increase in the 3 shell dimensions length, height, and width) in relation to the corresponding shell shape as means from the 28 experimental tanks with 16 mussels each. B: field data as individual growth in the first year post settlement in relation to individual shape of 556 mussels. The contours on the y-axis depict ±2SD of the mean of PC1. Dashed lines denote 95% confidence intervals for the fit.
Figure 6.
Field data reveal high phenotypic diversity in zebra mussels.
Mean (±SE) shell shape (triangles = PC1, squares = PC2) and shell strength (filled circles) of 84 mussels from each of 9 sampling sites. The contours on both y-axes depict ±2SD of the mean of PC1 and PC2.
Figure 7.
Predators of zebra mussels differ locally in their density.
Density of roach (white bars, fish symbol) and crayfish (black bars, crayfish symbol) at the sampled sites expressed as catch per unit effort (four Nordic gill-nets in the case of roach and 20 baited traps in the case of crayfish). Note that crayfish from site 4 were lost after catch.
Figure 8.
Population density and substrate influence shell strength in zebra mussels.
Relationship between shell strength of zebra mussels and the two major shell strength determining factors (as identified by the PLS model): A intra-specific zebra mussel density and B percentage of bottom covered by stones. The r of the regressions of shell strength with the zebra mussel density are 0.8 (p<0.001) and with the percentage of lake bottom covered by stones 0.76 (p = 0.017).