146 out of 221 microsites at a moderately exposed shore in central California showed a positive correlation between force and wave height; at some of these microsites force increased nonlinearly toward a statistically defined limit, making it difficult to classify a shore in regard to its exposure.
One might also expect that when higher water velocities are encountered, mussels will produce more byssal threads to increase strength of attachment. However, this is not the case, as shown by Moeser et al. (2006), who reported that in flume experiments mussels (M. edulis) significantly reduced thread production at water velocities above 15 cm s−1 (but see contrary findings in Zardi et al. 2007b). Similar findings were reported in a subsequent study on four mussel species, M. trossulus, M. galloprovincialis, M. californianus and Modiolus modiolus by Carrington et al. (2008). For all four species, velocities above 20 cm s−1 visibly hindered the mussel’s ability to extend its foot beyond the margin of the shell, a posture that must be held for several minutes in order to mold and attach a new thread to the substrate. Even in the exposed shore species, M. californianus, velocities above 30 cm s−1 precluded thread formation. Altogether, this study established 50 cm s−1 as a reasonable threshold limiting byssal thread production in solitary mussels. The authors found that flow was greatly ameliorated within mussel aggregations, ranging from 0.1 to 10% of free‐stream velocity, thus explaining why mussels can persist on shores with water flows in excess of their physiological limits.
Wave action also has a controlling influence on mussel bed communities by causing dislodgement through lift and drag, especially when mussel beds are dense and firmly packed, as on the majority of wave‐exposed shores (Figure 3.1). When dislodgement occurs, new space is created for colonisation. The risk of dislodgement in M. edulis increases with flow speed and mussel size and decreases with mussel tenacity or attachment strength; the latter varies twofold during the year, but this cycle is not aligned precisely with seasonal patterns of wave velocity (Carrington 2002). In a subsequent study, mussels (M. galloprovincialis) on rocky shores were found to allocate resources to reduce risk of dislodgement (smaller, thicker shell, stronger byssal threads) instead of promoting growth and reproduction (Babarro & Carrington 2013).
Denny (1995 and references therein) outlined a method for using information regarding the long‐term ‘waviness’ of the ocean to predict the rate of disturbance for individual organisms. This approach potentially provides a mechanism for ‘predicting the weather’ at arbitrary sites on the shore and can thus serve as a mechanistic link between wave climate and the rate of physical disturbance of individuals in wave‐swept populations. The rate of disturbance can then be used to predict the practical limits to a species' distribution. Several studies have used this mechanistic model to predict frequency and severity of mussel dislodgement (Hunt & Scheibling 2001; Schneider et al. 2005; Carrington et al. 2009; Cole & Denny 2014). To illustrate, using inputs of wave height and mussel attachment strength on wave‐exposed shores on Rhode Island, United States (Bass Rock and Black Point), Carrington et al. (2009) found that when averaged over three years, the predictions of the mechanistic model matched observed patterns, with minimal dislodgment from November to July and peak dislodgment in the hurricane season between August and October (Figures 3.7 and 3.8). Overall, the mechanistic model accounted for over 80% of the variation in mean monthly wave dislodgment during their study (linear regression, r 2 = 0.85, P < 0.001, N = 12, observed = 0.08 + 0.574 × predicted).
Mussel aggregations provide a habitat for a diverse assemblage of organisms, some of which attach directly to the mussel shells (epibionts). Three‐dimensional epibionts (e.g. kelp, barnacles and slipper limpets) increase the height of the mussel and thus the probability it will be dislodged, as compared to epibionts that form thin layers (e.g. encrusting bryozoans, colonial tunicates and sheet‐like sponges). The increased chance of dislodgement of mussels covered with 3D epibionts is due to increased hydrodynamic forces exerted on the mussel, specifically faster flow velocities that lead to higher drag‐induced loading (Garner & Litvaitis 2017 and references therein; O’Connor et al. 2006). When mussels were artificially fouled, attachment strength increased by 7%, or by 17% during winter (Garner & Litvaitis 2013). Attachment strength and position within mussel beds were measured in two mussel species, M. edulis and M. galloprovincialis (Schneider et al. 2005). Results indicated that M. edulis moved more frequently and more quickly to the exterior of simulated mussel beds than did M. galloprovincialis. Together with measurements of attachment strength in the field, Schneider et al. (2005) used a wave force model to examine the probability of dislodgement for each species under a range of water velocities. By preferentially moving to the exterior of beds, the results suggested that M. edulis experiences higher dislodgement rates due to exposure to large hydrodynamic forces than does M. galloprovincialis. Because of lower attachment strengths, M. edulis is also predicted to have higher mortality rates than M. galloprovincialis in interior portions of the bed. Thus, the authors concluded that differential movement behaviour may contribute to the differential genotype‐specific mortality rates observed in the Mytilus spp. hybrid zone in southwest England (see Chapter 9) and is an example of behaviour potentially modifying rates of exogenous selection in an intertidal hybrid zone. However, a genetic survey of 20 sites on Irish coasts has indicated no apparent advantage for either M. galloprovincialis or M. edulis genotypes at exposed shore locations (Gosling et al. 2008).
Figure 3.7 Photographs of a quadrat from Black Point, Rhode Island Sound, United States, taken on (top) October 3 and (bottom) October 21, 2001. Mussels (M. edulis) were dislodged by waves between the two censuses, as evidenced by the large gap of bare rock in the latter image. Scale bar 5.0 cm.
Source: From Carrington et al. (2009). Reproduced with permission from John Wiley and Sons.
Figure 3.8 Summary of mussel dislodgment in Rhode Island, 2001–2003. Bars are means (± SE) of monthly values of predicted (black) and observed (gray) mortality.
Source: From Carrington et al. (2009). Reproduced with permission from John Wiley and Sons.
Substrate
In wave‐exposed areas, Mytilus requires a hard and stable substratum such as rocks or large boulders on which to form beds, while in sheltered areas infaunal beds may occur on gravel or even in quite sandy areas. Mussel larvae settle on a wide variety of substrates (e.g. rocks and ridged, hard surfaces, filamentous macroalgae, hydroids, eelgrass, byssus of conspecific adults, artificial substrates such as polypropylene fibrous ropes or artificial seaweed; see details on the mechanism of byssus attachment to substrates and larval settlement cues in Chapters 2 and 5, respectively). The capability of mussels to attach byssal threads to the substratum and to form dense aggregations has permitted them to colonise both hard‐ and soft‐bottom habitats, where they attach to one another because little suitable attachment substratum is available (Aguilera et al. 2017). Specifically, shells of recently dead mussels, for example, are common in mussel beds, yet they do not offer the same hold as the shells of living conspecifics. However, shells (empty, alive or fragmented)