Technical Comments

Comment on "Divergent Induced Responses to an Invasive Predator in Marine Mussel Populations"

See allHide authors and affiliations

Science  06 Apr 2007:
Vol. 316, Issue 5821, pp. 53
DOI: 10.1126/science.1135099


Freeman and Byers (Reports, 11 August 2006, p. 831) presented evidence for the rapid evolution of antipredator defenses in the mussel Mytilus edulis. However, their analysis is confounded by three issues. Samples from some sites are likely to have included a second species, M. trossulus; their manipulation of chemical cues does not preclude other interpretations; and they failed to establish an adaptive significance to shell thickening.

Freeman and Byers (1) recently reported that blue mussels (Mytilus edulis) from southern and northern New England differ in the degree to which they express inducible shell thickening when exposed to waterborne cues from two crab predators. Although mussels from both regions increased shell thickness after exposure to the green crab, Carcinus maenas, only southern mussels exhibited a significant increase in shell thickness upon exposure to the Asian shore crab Hemigrapsus sanguineus. Because H. sanguineus has recently become established on the U.S. Atlantic coast but is, as yet, absent from northern New England, Freeman and Byers concluded that differential shell thickening in blue mussels was due to the rapid evolution of antipredator defenses in southern New England mussels. We believe that the experiments conducted by Freeman and Byers did not fully account for the presence of two morphologically similar mussel species in northern New England. Their conclusions are further weakened by an incomplete manipulation of chemical cues, and their data fall short of clearly demonstrating an adaptive significance to shell thickening.

Freeman and Byers's conclusion of intraspecific change rests on the premise that they are comparing an inducible response in a single species, M. edulis. However, the morphologically indistinguishable sibling species M. trossulus is sympatric with M. edulis in northern New England. We strongly suspect that the presence of this second species of mussel in their northern samples confounds their experimental design. Freeman and Byers classified only two sites in eastern Maine (Lubec and Cutler) as sympatric on the basis of early results published by Rawson et al. (2). More extensive sampling of intertidal sites has shown that the zone of sympatry extends farther southwest than originally thought; mussel populations in both eastern and central Maine are a mosaic, with some populations composed predominantly of M. edulis whereas nearby populations may predominantly contain M. trossulus (Fig. 1). Repeated sampling has also shown that the frequency of M. trossulus at some locations changes substantially from year to year. Although such data were not available to Freeman and Byers at the time of their study, they failed to recognize that species boundaries and hybrid zones in marine taxa with dispersive larvae are likely to be highly dynamic. In addition, while Freeman and Byers sampled juvenile mussels from subtidal locations, the data presented in (2) come from samples of intertidal adult mussels. We have found substantial frequencies of M. trossulus settling at subtidal sites where the adults in neighboring intertidal populations are predominantly or exclusively M. edulis. We do not currently know whether selection is reducing the frequency of M. trossulus during ontogeny or whether species composition differs consistently between subtidal and intertidal sites. However, we consider it inevitable that Freeman and Byers's northern mussels included an appreciable frequency of M. trossulus. If we are correct, then the lack of a response to Hemigrapsus in northern mussels could be entirely or partially due to interspecific differences in the response to predators.

Fig. 1.

Mussel populations in central and eastern Maine are a mosaic of M. edulis and M. trossulus. Substantial frequencies of M. trossulus alleles (black) at the diagnostic marker Glu5′; can be found as far west as the Penobscot Bay region. Sites 12 to 24 (underlined) are those that have been sampled subsequent to Rawson et al. (2). Between 30 and 48 adult mussels (20- to 50-mm shell length) were assayed at each of these sites. Nearly identical results were obtained for a second diagnostic marker, ITS.

Freeman and Byers (1) state that the rapid evolution of novel predator recognition is driven by natural selection acting on cue specificity or thresholds in southern New England mussels. Although this is a reasonable hypothesis, the authors did not collect the data needed to support this conclusion. Their experimental procedure did not examine mussel responses to different levels of potential waterborne crab cues (i.e., no thresholds were determined), nor did the authors present any biochemical evidence to suggest that the cues associated with C. maenas and H. sanguineus effluent are different. Because shell thickening by mollusks can be enhanced when waterborne predator cues are accompanied by cues from damaged conspecifics (3, 4), a more ecologically relevant test of predator recognition would have included treatments exposing mussels to the complex cues created by a predator feeding on conspecifics. Indeed, an alternative hypothesis to explain the data is that southern mussels may have learned to associate cues of damaged conspecifics with the scent of Hemigrapsus, whereas northern mussels have not. If so, this is an example of a short-term behavioral response, not rapid evolution.

The evolutionary interpretation offered by Freeman and Byers is further weakened by their failure to establish an adaptive significance to the thicker shells induced by Hemigrapsus. Bivalve shells are composed of inorganic compounds in an organic matrix, so a thicker shell is not necessarily a stronger shell. In the absence of a test to evaluate the resistance of shells to crushing, which is a standard feature of shell-thickening studies in other mollusks (3, 4), it is not clear whether the thicker shells developed by southern mussels actually confer any increased resistance to predation by Hemigrapsus. Consequently, we view Freeman and Byers's results as evidence of region-specific variation in shell thickening, the basis of which deserves further exploration, but believe that the adaptive interpretation proposed by these authors is not warranted.

References and Notes

View Abstract

Navigate This Article