INTRODUCTION
The horse-mussel, Modiolus modiolus (Linnaeus, 1758) occurs singularly, in clumps, or as high-density, species-rich biogenic reefs in temperate coastal regions around the world (Rees et al., Reference Rees, Sanderson, Mackie and Holt2008; Sanderson et al., Reference Sanderson, Holt, Ramsay, Perrins, McMath and Rees2008; Rees, Reference Rees2009). Modiolus modiolus reefs of conservation importance can be identified by their extent, percentage cover of live M. modiolus, and associated communities (Morris, Reference Morris2015). The decline of M. modiolus reefs in the North East Atlantic marks them as protected marine conservation features (Rees, Reference Rees2009; Cook et al., Reference Cook, Fariñas-Franco, Gell, Holt, Holt, Lindenbaum, Porter, Seed, Skates, Stringell and Sanderson2013). They are therefore identified under several conservation drivers: The Marine (Scotland) Act 2010; the Habitats and Species Directive (Council Directive 92/43/EEC); the OSPAR Convention; and the Marine Strategy Framework Directive (MSFD; Council Directive 2008/56/EC).
The extent and density of M. modiolus reefs have previously been used to determine reef condition (Rees, Reference Rees2009; Moore et al., Reference Moore, Harries and Trigg2012; Fariñas-Franco et al., Reference Fariñas-Franco, Sanderson and Roberts2014). However, these measures indicate little about their future prospects. Instead, parameters such as size-frequency distribution, morphometries and growth rates may be more appropriate for understanding long-term reef status. An understanding of reef-based differences will also support development of ecologically relevant climate envelope models and aid in illuminating potential impacts (e.g. altered ecosystem function due to shifts in body size) of a changing climate (Peck et al., Reference Peck, Clark, Morley, Massey and Rossetti2009; Somero, Reference Somero2010; Sandford & Kelly, Reference Sanford and Kelly2011).
Growth of M. modiolus is usually rapid in the first 4–6 years in order to reach an adult shell length (~35–40 mm) (Anwar et al., Reference Anwar, Richardson and Seed1990). At this size predation pressure is typically reduced (Comely, Reference Comely1978; Holt et al., Reference Holt, Rees, Hawkins and Seed1998) and animals can then spend more of their energy resources on reproduction as opposed to growth. Modiolus modiolus are reported to be sexually mature between 4–8 years old (Wiborg, Reference Wiborg1946; Rowell, Reference Rowell1967; Jasim & Brand, Reference Jasim and Brand1989). Temperature influences growth rates in bivalves, with slower growth rates characteristic of higher latitudes and greater depths (MacDonald & Thompson, Reference MacDonald and Thompson1985; Sato, Reference Sato1994). Near-bed seawater temperature data for the UK indicates a difference of up to 3–4°C between northern and southern latitudes, and consequently one would expect temperature driven variation in growth rates across M. modiolus reef habitats (see Gormley et al., Reference Gormley, Porter, Bell, Hull and Sanderson2013; Seidov et al., Reference Seidov, Baranova, Biddle, Boyer, Johnson, Mishonov, Paver and Zweng2013). It has also been suggested that gonad development is slower in bivalves in colder temperatures, as such conditions may lead to late maturity and subsequently allow northern populations to attain a larger size before energy is focused on reproduction (Sato, Reference Sato1994). The availability of food is another factor that may affect growth rates, suggesting slower growth rates would be observed in deep-water populations, where food is more limited (MacDonald & Thompson, Reference MacDonald and Thompson1985). Additionally, intertidal M. modiolus populations have demonstrated slower growth rates than populations in fast flowing currents, e.g. on oilrigs in the North Sea (Anwar et al., Reference Anwar, Richardson and Seed1990), where increased flow rates reduce the energy required to filter feed (Wildish & Peer, Reference Wildish and Peer1983); however, differences in food availability and feeding windows between intertidal and offshore sites may also play a role (Lesser et al., Reference Lesser, Witman and Sebnens1994).
Physical conditions also have a strong influence on shell morphology with bivalves generally exhibiting high plasticity in response to local environmental conditions (Seed, Reference Seed1968; O'Connor, Reference O'Connor2010). For example, bivalves with narrow shaped shells reduce the effects of drag and the risk of dislodgement, and are typically found in areas with high flow rates (Seed, Reference Seed1968, Reference Seed, Rhoads and Lutz1980; Steffani & Branch, Reference Steffani and Branch2003). Reef density has also been seen to influence the morphology of bivalves, with narrower shells found in denser populations. This may be a response to food availability, disease prevalence and/or physical interference (Alunno-Bruscia et al., Reference Alunno-Bruscia, Bourget and Frechette2001; Lauzon-Guay et al., Reference Lauzon-Guay, Hamilton and Barbeau2005; Caill-Milly et al., Reference Caill-Milly, Bru, Mahé, Borie and D'Amico2012). Depth (Etter & Rex, Reference Etter and Rex1990) and substrate type (Seed, Reference Seed, Rhoads and Lutz1980) have also been cited as influencing morphology in bivalves. A recent study on the morphology of M. modiolus shells found significant differences in shell shape between sites within an enclosed loch, probably due to differences in current speed and sediment type between sites (Fariñas-Franco et al., Reference Fariñas-Franco, Sanderson and Roberts2014).
Recruitment of juveniles influences population structure, and effective recruitment relies on successful spawning and settlement of larvae. The spawning season of M. modiolus is poorly understood and tends to vary greatly between reefs, though is probably linked to temperature. Settlement of larvae at certain reefs is also highly variable and sporadic between years (Wiborg, Reference Wiborg1946; Seed & Brown, Reference Seed and Brown1977; Brown, Reference Brown1984). Comely (Reference Comely1978) suggested settlement differences between reefs were due to larvae originating from outside the reef area; hence recruitment would be affected by hydrodynamic conditions, particularly as M. modiolus larvae can remain within the water column for ~1 month (Schweinitz & Lutz, Reference Schweinitz and Lutz1976; Roberts et al., Reference Roberts, Allcock, Fariñas Franco, Gorman, Maggs, Mahon, Smyth, Strain and Wilson2011). In addition, a population may act as a larval sink or source, resulting in recruitment differences between reefs (Lipcius et al., Reference Lipcius, Eggleston, Schreiber, Seitz, Shen, Sisson, Stockhausen and Wang2008). Variation in mussel density on a reef may also influence recruitment. Adult aggregations provide refuge, and consequently can increase post-settlement survival by protecting juvenile M. modiolus from predation (Comely, Reference Comely1978; Holt et al., Reference Holt, Rees, Hawkins and Seed1998).
Population structure, density or shell morphology may influence the biodiversity of a given reef, either directly through the abundance and diversity of associated communities (Gutiérrez et al., Reference Gutiérrez, Jones, Strayer and Iribarne2003; O'Connor & Crowe, Reference O'Connor and Crowe2007; Ragnarsson & Burgos, Reference Ragnarsson and Burgos2012) or indirectly through the reef's capability for habitat modification (Allen & Vaughn, Reference Allen and Vaughn2011). It is important for such factors to be taken into consideration in Marine Protected Area (MPA) monitoring programmes for successful conservation management of these biodiversity hotspots.
The aim of the present study was to compare population structure, shell morphology and growth rates of M. modiolus reef populations from three sites situated across the extent of the UK distribution for the habitat including the Lleyn Peninsula in Wales (southern range), off Port Appin in Western Scotland (mid-range) and in Scapa Flow in the Orkney Isles, Scotland (northern range). The main hypothesis of the work was that sites would vary significantly in these aspects with observed differences related to differences in latitude (as a proxy of temperature), current flow and mussel density. Differences in these population parameters are likely to have important implications for the conservation management of M. modiolus populations in a changing climate.
MATERIALS AND METHODS
Mussels were collected by clearing ~1 m2 plots on M. modiolus ‘reefs’ (cf. Morris, Reference Morris2015) using scuba. Four plots were cleared in Scapa Flow (58°53′0.446N 03°11′0.255W; 23 m below chart datum (BCD)), three plots were cleared off Port Appin (56°33′0.029N 05°25′0.468W; 21 m BCD) and four plots were cleared from north of the Lleyn Peninsula (52°56′0.516 N 04°38′0.070W; 30 m BCD); (Figure 1). These sites were chosen as representative of different reef types found in the UK with regards mussel density, geographic location (i.e. latitude) and flow regime.
Study sites: Scapa Flow (1), Port Appin (2) and North Lleyn (3) were sampled in the present study. The Humber Estuary (4), Irish Sea (5), Isle of Man (6), Strangford Lough (7), Isle of Mull (8) and Ling Bank (9) were sampled in the Anwar et al. (Reference Anwar, Richardson and Seed1990) study.
The North Lleyn reef is currently the largest known M. modiolus reef in the UK, with an extent of ~349 ha (Lindenbaum et al., Reference Lindenbaum, Bennell, Rees, McClean, Cook, Wheeler and Sanderson2008). The reef lies within and forms a feature of the Pen Lŷn a'r Sarnau SAC and is believed to be the most southerly reef in the UK (Lindenbaum et al., Reference Lindenbaum, Bennell, Rees, McClean, Cook, Wheeler and Sanderson2008). The reef has high densities of M. modiolus with raised reef structures present across the reef's extent (Lindenbaum et al., Reference Lindenbaum, Bennell, Rees, McClean, Cook, Wheeler and Sanderson2008; Sanderson et al., Reference Sanderson, Holt, Ramsay, Perrins, McMath and Rees2008). The Scapa Flow reef is the mostly northerly reef in this study and is less dense than the North Lleyn reef and comprised of three main areas with a combined extent of 42 ha (Sanderson et al., Reference Sanderson, Hirst, Fariñas Franco, Grieve, Mair, Porter and Stirling2014). The Port Appin reef is approximately mid latitude between the other two reefs, and is the smallest of the three reefs with an extent of ~2 ha. The reef is continuous within this area with high densities of M. modiolus but lacks the raised structures found on the North Lleyn reef (Moore et al., Reference Moore, Harries and Trigg2012).
A current meter (MIDAS ECM, Valeport Ltd, Devon, UK) was placed at each site for 6 days during a neap cycle, recording current speed, direction and water depth every 5 s. The densities of the M. modiolus reefs were estimated using 0.25 m2 photo quadrats using the method and counting rules outlined in Cook et al. (Reference Cook, Fariñas-Franco, Gell, Holt, Holt, Lindenbaum, Porter, Seed, Skates, Stringell and Sanderson2013). Nine and five photo quadrats were taken in 2014 from Scapa Flow and Port Appin reefs, respectively. Photo quadrats could not be taken in 2014 on the North Lleyn reef, so seven photo quadrats taken in 2009 (Cook et al., Reference Cook, Fariñas-Franco, Gell, Holt, Holt, Lindenbaum, Porter, Seed, Skates, Stringell and Sanderson2013), were analysed instead.
Length frequency and morphometric data
The maximum shell-lengths of all M. modiolus were measured to the nearest 0.1 mm using digital vernier calipers, before being returned to the reef. In addition, width and height of the first 50 mussels were recorded for morphometric analysis. Debris from the cleared plots was washed through a 1 mm sieve in order to include juveniles in the analysis. This was fixed in seawater buffered 10% formaldehyde solution, and subsequently sorted by hand to find juveniles. Juveniles were measured using the same method as described above. Additional length, width and height measurements of M. modiolus were collated from historical surveys (2010–2015) at the same sites.
Growth rate data
Thirty mussels of varying lengths were selected from each site across an even size range. Acetate peels were used to age the mussels as described in Richardson et al. (Reference Richardson, Crisp and Runham1979) and Anwar et al. (Reference Anwar, Richardson and Seed1990). One valve from each mussel was cut longitudinally, along the umbone-rim axis, using a circular saw. To prevent breakages, shells <60 mm in length were set in clear polyester casting resin before being cut. Once cut, the shell half containing the umbone was sanded, polished and then etched in 1% hydrochloric acid. The edge of the shell was dipped in acetone before being placed on an acetate sheet. After drying, the sheet was pulled from the shell, leaving a ‘peel’ of the inner nacreous layer. All peels were aged via use of a dissection microscope with each of the dark bands in the middle nacreous layer representing a year of growth (Figure 2).
Acetate peel of a 16-year-old M. modiolus specimen. White arrows indicate some of the dark winter bands in the younger half of the shell, laid down each year within the middle nacreous layer.
Analyses
All statistical analyses were completed using the statistical software R (release 3.1.1, 2014). One-way ANOVAs with follow-up pairwise comparisons were used to test for differences in flow rates, collected with the current meter, between the sites. Length frequency measurements were converted to percentage size frequency and plotted for each sampled site along with approximate ages based on the growth curves. One-way ANOVAs, with follow-up pairwise comparisons were then used to test the differences in the mean length of the mussels between sites.
Differences in morphometric ratios between shell measurements were compared between sites. The ratios of height-length were used as a measure of shell elongation, width-height as a measure of shell inflation and width-length as a measure of shell obesity (Zieritz & Aldridge, Reference Zieritz and Aldridge2009; Fariñas Franco et al., Reference Fariñas-Franco, Sanderson and Roberts2014). Due to the limited number of mussels <50 mm found on some reefs, all mussels <50 mm were removed from the morphometric analysis. These morphometric ratios were not normally distributed; therefore, Kruskal–Wallis tests were used to test for statistically significant differences in the shell shape ratios between sites. Additionally, approximations of the shell shapes of M. modiolus at the same age from North Lleyn, Port Appin and Scapa Flow reefs were created according to calculated ratios and growth rates.
Values for maximum length (L ∞), growth rate (K) and tθ were identified to produce growth curves based on von Bertalanffy's growth equation for each of the populations. Starting parameters were estimated using VbStarts in the FSA package within R (Fish R, Reference Fish2014). These starting values were then used to calculate the coefficients using the non-linear least-squared regression within R. L ∞ and K values were plotted with the best-fit von Bertalanffy growth model superimposed. These values, along with L ∞ and K values from six other populations reported in Anwar et al. (Reference Anwar, Richardson and Seed1990) (Figure 1), were used to test for correlations with maximum tidal flow and latitude. After assumptions of normality and equal variance were confirmed, a Pearson's Product Moment Correlation test was used to compare these L ∞ and K values against predicted maximum flow rates (BERR, 2008) and latitude of the nine sites. The current meter data was not used in this analysis, as benthic current flow data was not available for the additional six sites. The relationship between latitude and flow was tested separately using Spearman's Rank Correlation.
RESULTS
The mean current speed at the North Lleyn site was 0.244 m s−1, significantly faster than at Scapa Flow (0.029 m s−1, P < 0.001) and Port Appin (0.106 m s−1, P < 0.001). The current at Port Appin was also significantly faster than at Scapa Flow (P < 0.001). The North Lleyn reef had 24.6 M. modiolus per m2 and 21% coverage, the Port Appin reef had 16 per m2 and 16% coverage, and the Scapa Flow reef had 11.1 per m2 and 11% coverage.
Length frequency distributions for each of the sites were bimodal (Figure 3), most pronounced in the Port Appin and North Lleyn populations. The mussels sampled at Scapa Flow were significantly larger (mean 88.75 mm) than those at Port Appin (mean 50.1 mm; P < 0.001) and North Lleyn (mean 56.04 mm; P < 0.001).
Percentage size frequency (mm) of M. modiolus from Scapa Flow (N = 118), Port Appin (n = 218) and North Lleyn (n = 113). Numbers above bars indicate approximate age of size classes using the growth rates for each site.
Shell morphometric ratios of inflation, elongation and obesity were found to be significantly different between sites. Mussels from the North Lleyn reef were significantly less elongated than mussels from the Port Appin Reef or Scapa Flow reef (respective Kruskal–Wallis χ2 = 3.614 and 2.111; P < 0.001 and P < 0.05). Mussels from the North Lleyn reef were also significantly less inflated (respective Kruskal–Wallis for inflation χ2 = 5.729 and 4.618; P < 0.001), and less obese (respective Kruskal–Wallis for inflation χ2 = 7.733 and 6.754; P < 0.001) than mussels from the Port Appin reef or Scapa Flow reef (Figure 4). Overall, this gave the mussels from the North Lleyn reef a more streamlined profile, having 1% less height and 4.6% less width at a given length compared with the other reefs, and 8.4% less height at a given width compared with the other reefs. The approximate shell shapes of M. modiolus, at a similar age, from the three sites are illustrated in Figure 5.
The relationship between length and height (A), length and width (B), and height and width (C) of M. modiolus from Scapa Flow (N = 110), Port Appin (N = 96) and North Lleyn (N = 95).
An approximation of the shell shapes of M. modiolus of the same age from North Lleyn, Port Appin and Scapa Flow.
Von Bertalanffy growth curves revealed that M. modiolus from the Scapa Flow population had a higher L ∞ (159.8 mm) and a lower K value (0.04) compared with the other sites in this study (Figure 6; Table 1). North Lleyn M. modiolus had a lower L ∞ value (L ∞ 110.9 mm), and a lower K value (0.059) compared with the Port Appin population (L ∞ 122.82, K 0.061).
Von Bertalanffy growth curves of the three M. modiolus populations analysed in this study: Scapa Flow, Port Appin and North Lleyn.
L ∞ (mm) and K constant of the von Bertalanffy equation LI (t) = L ∞ (1-e-K (t-tθ)), from the three M. modiolus populations studied in this paper as well as the six populations studied by Anwar et al. (Reference Anwar, Richardson and Seed1990). Depths at which the samples were taken are included, as well as approximate flow rates (BERR, 2008).
The K and L ∞ values obtained from the Anwar et al. (Reference Anwar, Richardson and Seed1990) study are listed in Table 1 with predicted max flow rates and depths. A significant positive relationship was found between latitude and L ∞ (r = 0.74, t = 2.85, P < 0.05; Figure 6) and a significant negative relationship was found between flow and L ∞ (r = −0.82, t = −3.4836, P < 0.05; Figure 7). There was no significant relationship between K and any environmental variable. There was also no significant correlation between flow and latitude (T = 10, P = 0.39).
Relationship between the latitude and the L ∞ (mm) (A), and the chart surface flow and L ∞ (mm) (B), of the respective M. modiolus growth curves from each of the sites.
DISCUSSION
The present study found that M. modiolus from different reefs exhibited different demographic profiles, morphologies and growth rates, that varied with tidal flow and latitude. Growth rates and maximum achievable sizes increased with current flow and latitude. Both factors may also directly contribute to the observed variations in juvenile abundance and shell morphology, but indirect impacts via changes to energy budgets should also be considered (Sokolova et al., Reference Sokolova, Frederich, Bagwe, Lannig and Sukhotin2012).
Differences in juvenile abundance are tentatively interpreted here because of the low number of sites studied. Nevertheless, the increased abundance of juveniles at North Lleyn and Port Appin (Figure 3) may be due to the higher current flows observed (Comely, Reference Comely1978; Brown, Reference Brown1984). Such conditions may lead to increased connectivity via improved larval supply from other reefs in the Irish Sea (Anwar et al., Reference Anwar, Richardson and Seed1990; Rees et al., Reference Rees, Sanderson, Mackie and Holt2008; Gormley et al., Reference Gormley, Mackenzie, Robins, Coscia, Cassidy, James, Hull, Piertney, Sanderson and Porter2015b) and Loch Linnhe respectively (Rees, Reference Rees2009; Moore et al., Reference Moore, Harries and Trigg2012). Conversely, the decreased abundance of juveniles in Scapa Flow may be caused by limited connectivity, due to reduced tidal flow and a lack of known neighbouring reefs (Rees, Reference Rees2009; Gormley et al., Reference Gormley, Hull, Porter, Bell and Sanderson2015a). Furthermore, while the Scapa Flow population had the largest L ∞, it also had a lower mussel density compared with the Port Appin and North Lleyn reefs. Although gamete production is correlated with body size, population density has been shown to be equally important to fertilization success and reproductive output (Levitan, Reference Levitan1991), and consequently could partially account for lowered abundance of juveniles in the northern population. The higher structural complexity found on reefs with higher densities might also afford more shelter to juveniles from predation (Comely, Reference Comely1978; Holt et al., Reference Holt, Rees, Hawkins and Seed1998). An increase in reef complexity has been shown to increase post-settlement survival in other bivalves (Gutierrez et al., Reference Gutiérrez, Jones, Strayer and Iribarne2003; Nestlerode et al., Reference Nestlerode, Luckenbach and O'Beirn2007).
The high mussel density and narrow-shaped shells of mussels from North Lleyn, and low density and globular-shaped shells from Scapa Flow and Port Appin, support the view that crowding leads to narrower shells (Alunno-Bruscia et al., Reference Alunno-Bruscia, Bourget and Frechette2001; Lauzon-Guay et al., Reference Lauzon-Guay, Hamilton and Barbeau2005; Caill-Milly et al., Reference Caill-Milly, Bru, Mahé, Borie and D'Amico2012). However, the lower density at Port Appin does not support this hypothesis. Moreover, given that flow rates were significantly higher at North Lleyn than at Port Appin or Scapa Flow, flow rate, rather than density, may be the stronger driver of shell morphology. The narrow-shaped shells from North Lleyn are possibly a morphological adaptation to allow a better hold within the sediment and reduce the risk of dislodgement in fast flowing currents (Seed, Reference Seed1968, Reference Seed, Rhoads and Lutz1980; Steffani & Branch, Reference Steffani and Branch2003; Fariñas-Franco et al., Reference Fariñas-Franco, Sanderson and Roberts2014).
A significant positive relationship was observed between latitude and L ∞ when including data from the Anwar et al. (Reference Anwar, Richardson and Seed1990) study. Bergmann's rule, that a species will demonstrate a larger body size at higher latitudes (i.e. lower temperatures), may explain this relationship (Berke et al., Reference Berke, Jablonski, Krug, Roy and Tomasovych2013). Oxygen concentration generally limits the size of ectotherm species as surface area to volume ratios decrease with increased body size, thereby reducing gas exchange capability. The increased solubility of oxygen at lower seawater temperatures however, coupled with lowered metabolic demands under such conditions, reduces these size constraints in animals at higher latitudes (Chapelle & Peck, Reference Chapelle and Peck2004; Makarieva et al., Reference Makarieva, Gorshkov and Li2005; Moran & Woods, Reference Moran and Woods2012). However, Berke et al. (Reference Berke, Jablonski, Krug, Roy and Tomasovych2013) highlight that there is enormous diversity in size-latitude relationships for marine bivalves. Body size is influenced by a complex interaction of physiological, ecological and evolutionary drivers that affect growth rates, food availability, reproductive output, predation pressure, longevity and various other factors. Under increased thermal stress (e.g. at the edge of a biogeographic range), strain is placed on internal physiological systems. The related costs of energy acquisition, conversion and conservation, impact energy budgets with negative consequences for growth and reproduction (Sokolova et al., Reference Sokolova, Frederich, Bagwe, Lannig and Sukhotin2012). There are numerous studies that demonstrate temperature-induced changes to various aspects of energetics including growth, metabolism, reproductive output and condition index in ectotherm species (Pörtner, Reference Pörtner2002, Reference Pörtner2012; Lesser & Kruse, Reference Lesser and Kruse2004; Hofmann & Todgham, Reference Hofmann and Todgham2010; Sokolova, Reference Sokolova2013). Additionally, given that shell formation and repair are energetically expensive (Palmer, Reference Palmer1992), increasing temperature may have negative repercussions for shell growth. Elevated temperature may also negatively affect shell biomineralization, particularly when coupled with limited food availability (Mackenzie et al., Reference Mackenzie, Bell, Birchenough, Culloty, Sanderson, Whiteley and Malham2013; Thomsen et al., Reference Thomsen, Casties, Pansch, Körtzinger and Melzner2013). Additionally, variation in depth between sites, particularly with regard to those deeper sites examined by Anwar et al. (Reference Anwar, Richardson and Seed1990), may influence the growth of mussels at the same latitude but at different depths, especially as previous research has shown that even small changes in temperature can influence bivalve growth (Almada-Villela et al., Reference Almada-Villela, Davenport and Gruffydd1982). Likewise, there are additional abiotic (e.g. photoperiod) and biotic (e.g. food availability) factors which vary with latitude and also influence growth that future work could consider (Strömgren, Reference Strömgren1976; Brodte et al., Reference Brodte, Knust and Pörtner2006).
A significant negative relationship between flow rate and the maximum theoretical length (L ∞) was also observed in M. modiolus when including data from Anwar et al. (Reference Anwar, Richardson and Seed1990). Such findings could be an indication of the energetic demands of byssal thread production under high flow conditions, reducing the risk of dislodgement, and reducing scope for growth (Comely, Reference Comely1978; Okamura, Reference Okamura1986; Anwar et al., Reference Anwar, Richardson and Seed1990; Fariñas-Franco et al., Reference Fariñas-Franco, Sanderson and Roberts2014). The findings could also possibly highlight an upper limit to flow conditions that are conducive to efficient food uptake, as current velocity determines the flux of material available for feeding (Lesser et al., Reference Lesser, Witman and Sebnens1994).
An increase in global CO2 concentration has caused sea temperatures to rise, especially in the latter part of the last century and, under current climate change emission scenarios, is predicted to continue to increase (3–5°C by 2100) (IPCC, 2014). For M. modiolus reefs at the limit of the species’ thermal tolerance (e.g. the North Lleyn reef), warming is likely to be a contributory factor to potential decline. Mean bottom temperature at this southern aspect of the distribution was 11°C in 2009 (Gormley et al., Reference Gormley, Porter, Bell, Hull and Sanderson2013) and regional summer seawater temperatures as high as 17–18°C have been reported (CEFAS Coastal Temperature Network). Warming is likely to cause range shifts in species and habitats as species align their distributions to match their physiological tolerances (Doney et al., Reference Doney, Ruckelshaus, Duffy, Barry, Chan, English, Galindo, Grebmeier, Hollowed, Knowlton, Polovina, Rabalais, Sydeman and Talley2012). Consequently, M. modiolus reefs may respond to future climate change by shifting distribution further northward. Climate-population models have indicated that changes in population demographics, abundance and size may accompany such climate-driven range shifts (Hare et al., Reference Hare, Alexander, Fogarty, Williams and Scott2010). Results here provide baseline values against which changes in such factors may be monitored thereby aiding detection of potential climate change impacts. Further, where populations are genetically connected (Gormley et al., Reference Gormley, Hull, Porter, Bell and Sanderson2015a, Reference Gormley, Mackenzie, Robins, Coscia, Cassidy, James, Hull, Piertney, Sanderson and Porterb) and thus may have similar adaptive capacity, consideration of the effects of temperature in more southern regions may give some indication of effects to be expected in more northerly populations. However, given both the longevity of M. modiolus and current acceleration in rates of climate change, many reefs will have limited opportunity for adaptation.
Under the OSPAR Convention for the Protection of the Marine Environment of the North East Atlantic 1992, M. modiolus reefs are listed as Priority Marine Habitats (determined as ‘threatened and/or declining species and habitats’) (OSPAR, 2009). The maintenance of such habitats is therefore key to the achievement of ‘Good Environmental Status’ under the European Union (EU) Marine Strategy Framework Directive (OSPAR, 2012). The findings of the present study have important implications when considering how Marine Protected Areas (MPAs) for these reefs are managed, particularly in a changing climate (Gormley et al., Reference Gormley, Porter, Bell, Hull and Sanderson2013, Reference Gormley, Hull, Porter, Bell and Sanderson2015a). Although the underlying causes of variation in demographics between sites would require further investigation, strong reproduction and post-settlement survival (collectively referred to as recruitment and measurable as juvenile abundance) are nevertheless desirable characteristics for reefs within an MPA. From a cost-benefit perspective, MPAs with stronger recruitment are more likely to achieve conservation objectives. In some cases, maintaining strong recruitment might be a site-specific management consideration (self-recruiting reefs) or it might require the management of larval supply from outside the MPA (cf. Gormley et al., Reference Gormley, Mackenzie, Robins, Coscia, Cassidy, James, Hull, Piertney, Sanderson and Porter2015b), thus requiring a network approach. At present, no MPA management plans or conservation objectives give detailed consideration of how recruitment should be managed. For a habitat type that has declined (Rees, Reference Rees2009), restoration may be necessary to maintain the shellfish reefs within the MPA (cf. Fariñas-Franco et al., Reference Fariñas-Franco, Sanderson and Roberts2014) or to create stepping-stones in planned climate migration (Gormley et al., Reference Gormley, Hull, Porter, Bell and Sanderson2015a, Reference Gormley, Mackenzie, Robins, Coscia, Cassidy, James, Hull, Piertney, Sanderson and Porterb). In such cases, Fariñas-Franco et al. (Reference Fariñas-Franco, Sanderson and Roberts2014) considered that M. modiolus ecophenotypes might need to be matched to MPAs if donor populations were to be used in the rehabilitation of declining populations in that MPA. For example, a globular shell shape may not fare well if translocated to an area of greater flow (Hiscock et al., Reference Hiscock, Tyler-Walters and Jones2002) and morphological adaptation is unlikely to be rapid in these slow growing, long-lived species (Fariñas-Franco et al., Reference Fariñas-Franco, Sanderson and Roberts2014). The present study provides widespread evidence of significantly different ecophenotypes linked to flow and latitude and therefore further emphasizes these conclusions.
Significant morphological and demographic variation between locations also gives cause for consideration that, since body size is often linked to fecundity, and population size is clearly linked to total reproductive output (and probably recruitment) some populations may be a higher priority for conservation management than others. Whether it is preferable to direct conservation management effort towards high densities of smaller individuals such as off the North Lleyn (mean length 56.04 mm, 30.8 ind. m2) or reefs with low densities of bigger animals such as in Scapa Flow (mean length of 88.7 mm and 5.80 ind. m2) would require a more detailed analysis of total reproductive outputs and vectors.
This study presents differences between M. modiolus populations in terms of their demographics, morphology and growth. Overall the study highlights that conservation management needs to carefully consider the demographics and morphology of protected M. modiolus populations in the prioritization of management effort, in assessments of the future prospects of protected areas, and the consideration of restoration. These implications are particularly relevant to planning for a changing climate.
ACKNOWLEDGEMENTS
Thank you to Dan Harries, Hamish Mair, Joanne Porter, Rebecca Grieve, Ashley Cordingley and Robert Harbour for their support during sample and data collection. José Fariñas-Franco gave advice on producing acetate peels; Kate Gormley provided some of the historical sample metadata; and Rebecca Grieve provided quadrat photographs of the reefs at Scapa Flow for density estimates. Hamish Mair also provided feedback on early drafts of the manuscript, for which we are grateful. Special thanks also go to Rohan Holt and the marine monitoring team at Natural Resource Wales for facilitating sample collection from the reef at North Lleyn.
FINANCIAL SUPPORT
This work received funding from the MASTS pooling initiative (The Marine Alliance for Science and Technology for Scotland) and their support is gratefully acknowledged. MASTS is funded by the Scottish Funding Council (grant number HR09011) and contributing institutions.
