Microgeographical variation in shell strength in the flat periwinkles Littorina obtusata and Littorina mariae

The strength of molluscan shells has been shown to vary in adaptive ways in a number of species and one of the main factors thought to be involved is shell-crushing by predators. A recent study found that the sibling species of flat periwinkle Littorina obtusata and Littorina mariae showed significant differences in the rates at which shell strength increased with shell length in specimens which had been collected from the same location, where the species were sympatric. This paper describes differences between the shells of the two species from a number of localities around Milford Haven in Dyfed, Wales, and local geographical variation in the shells. Littorina mariae, which is normally found at lower tidal levels than L. obtusata, matures at a smaller shell length. Both species reinforce the shell as they grow since shell strength, determined as the maximum force applied by a hydraulic tensile testing machine before the shell cracked, is strongly positively allometric; it increases at a rate close to the cube of shell length whilst isometric growth would result in strength increasing in proportion to the square of shell length. Because L. mariae matures earlier and reinforces the shell at a smaller size, the mature shell of L. mariae is substantially stronger on average than that of a similar sized but immature L. obtusata. At maturity the shell strengths of the two species are not very different despite the substantial difference in mean shell length. Strength varies significantly from shore to shore, and with the level of the shore from which the animals were collected. Strength increases down the shore in both species. Shell strength decreases with exposure to wave action in L. mariae but increases with exposure in L. obtusata; there is also substantial shore-to-shore variation which is not explained by exposure. Path analysis was used to explore the relationship between shell strength and other measured shell parameters (mass, length, height, thickness). The best predictor of shell strength in both species is a parameter which is heavily positively loaded on LN (shell mass) and strongly offset by negative loadings on LN (shell length) and LN (shell height). This is logical because for a given shell length a heavier shell will be thicker and stronger, whilst for a given shell mass a bigger shell will be thinner and therefore weaker. Such differential variation of shell mass and shell length explains most of the geographical variation observed in shell strength; shells are stronger in snails collected from one place than from another because, for the same shell length they are heavier or, to put it the other way, because at the same shell mass, they are smaller.