INTRODUCTION
There is growing interest in the reproductive connectivity of marine organism populations associated with the design of Marine Protected Area networks. Accordingly the role of man-made structures in facilitating colonization of hard substrata, and subsequent dispersal of reproductive propagules or juvenile stages, is receiving worthwhile attention (e.g. Bacchiocchi & Airoldi, Reference Bacchiocchi and Airoldi2003; Moschella et al., Reference Moschella, Abbiati, Ĺberg, Airoldi, Anderson, Bacchiocchi, Bulleri, Dinesen, Frost, Gacia, Granhag, Jonsson, Satta, Sundelof, Thompson and Hawkins2005; Hawkins et al., Reference Hawkins, Moore, Burrows, Poloczanska, Mieszkowska, Herbert, Jenkins, Thompson, Genner and Southward2008, Reference Hawkins, Sugden, Mieszkowska, Moore, Poloczanska, Leaper, Herbert, Genner, Moschella, Thompson, Jenkins, Southward and Burrows2009). However, it may be that as many rare species have poor, or poorly understood, dispersal traits, due either to a short pelagic larval phase or direct development, the importance of artificial substrata in their persistence and movement may be underestimated.
The dogwhelk Nucella lapillus (L.) is a predatory gastropod inhabiting intertidal and subtidal hard structures along North Atlantic shores. The species is carnivorous, feeding preferentially on sedentary prey (e.g. barnacles, by prizing plates apart; and mussels by boring through the shell) (Moore, Reference Moore1938; Morgan, Reference Morgan1972; Crothers, Reference Crothers1985), but can adapt to other prey types (Largen, Reference Largen1967; Crothers, Reference Crothers1985). Having no recognized pelagic stage, upon emergence from capsules fully formed juvenile N. lapillus (crawlaways), can migrate down shore where feeding has been observed on the annelid worm Spirorbis borealis (Moore, Reference Moore1936) or small barnacles (Crothers, Reference Crothers1985). Juvenile N. lapillus have also been observed at the same tidal level as their egg cases sheltering in empty barnacle carapaces (Feare, Reference Feare1970). It may be considered that direct development may restrict dispersive abilities, but rafting (Bryan et al., Reference Bryan, Gibbs, Hummerstone and Burt1986; Martel & Chia, Reference Martel and Chia1991), passive transport (Colson & Hughes, Reference Colson and Hughes2004) or a small amount of air trapped in the shell (Adachi & Wada, Reference Adachi and Wada1999), have been suggested as colonization facilitators in direct developing gastropods. In addition shells of small gastropods (Littorina littorea) have been noted to roll in the surf, in a similar way to saltation in sediments (Amos et al., Reference Amos, Sutherland, Cloutier and Patterson2000).
Because of tributyltin (TBT) use as an antifoulant, many UK N. lapillus populations declined or became absent from shores, particularly close to ports (e.g. Spence et al., Reference Spence, Bryan, Gibbs, Masters, Morris and Hawkins1990). Dogwhelks are highly sensitive to TBT (Evans & Nicholson, Reference Evans and Nicholson2000) with levels as low as 1 ngl−1 inducing imposex (imposition of male sex organs on the female) (Matthiessen & Gibbs, Reference Matthiessen and Gibbs1998) thus the organism proved to be a sensitive and valuable indicator for marine-TBT levels. In female N. lapillus, imposex advances to irreversible sterility and death occurs as vas deferens growth blocks the anterior oviduct (Gibbs & Bryan, Reference Gibbs and Bryan1986).
By the 1980s N. lapillus became extinct in areas of high shipping and maritime activity, and, particularly along the UK south coast, many surviving populations were depleted, suffering with high imposex levels (Spence et al., Reference Spence, Bryan, Gibbs, Masters, Morris and Hawkins1990; Bryan & Gibbs, Reference Bryan, Gibbs, Newman and McIntosh1991). In 1985 France was the first nation to place restrictions on TBT use and in 1987 the UK also banned the antifoulant on vessels <25 m length overall. Many other nations put in place similar bans (Champ, Reference Champ2000), however, its use on larger ships continued (Bates & Benson, Reference Bates and Benson1993). A global ban was finally ratified in September 2008 (Sonak et al., Reference Sonak, Pangam, Giriyan and Hawaldar2009), but there are some difficulties with the legal issues (Gipperth, Reference Gipperth2009; Kotrikla, Reference Kotrikla2009). Following progressive TBT restrictions, UK sites where N. lapillus was absent began to be recolonized (Evans et al., Reference Evans, Leksono and McKinnell1995; Harding et al., Reference Harding, Rodgers, Davies and Moore1997; Bray & Herbert, Reference Bray and Herbert1998; Crothers, Reference Crothers1998, Reference Crothers2003; Birchenough et al., Reference Birchenough, Evans, Moss and Welch2002; Colson & Hughes, Reference Colson and Hughes2004). Some populations recovered to pre-TBT densities in just ten years (Colson & Hughes, Reference Colson and Hughes2004) with increasing egg production and changes in population structure documented (Morton, Reference Morton2009). However, possibly due to residual and potentially continuing contamination (Kotrikla, Reference Kotrikla2009; Oliveira et al., Reference Oliveira, Richardson, Sousa, Takahashi, Tanabe and Barroso2009), the recovery of dogwhelk populations has proved to be slow in some areas (e.g. Smith et al., Reference Smith, Thain and Barry2006).
One such area is the Solent region on the UK south coast: a centre for international shipping, and a focus for maritime recreation from Selsey (Sussex) to Poole Harbour (Dorset) and on the Isle of Wight. Much of the coastline is soft sediment, sand, gravel or soft chalk, with hard substrata suitable for N. lapillus and their prey few and far between. Isolated dogwhelk populations are believed to have existed historically on the mainland coast at Netley on Southampton Water (Langston et al., Reference Langston, Bryan, Burt and Pope1994) and between Hurst Spit and Calshot (Moore, Reference Moore1936), but they were not found in surveys in 1986–1989 (Spence et al., Reference Spence, Bryan, Gibbs, Masters, Morris and Hawkins1990) and 1996–2002 (Bray, Reference Bray2005). It seems that N. lapillus has been absent from much of the mainland coast of central southern England since the 1980s, potentially due to TBT pollution and unsuitable habitat. In the 1970s, large populations were present on the northern Solent coast of the Isle of Wight, mainly on piers and breakwaters. By the mid-1980s, with TBT now extensively used on all kinds of marine vessels, these populations had mostly become extinct, although those on the relatively uncontaminated rocky shores on the island's southern coast remained, albeit significantly affected by imposex (Herbert, Reference Herbert1988; Bray & Herbert, 1998; Herbert et al., Reference Herbert, Bray, Hawkins, Collins and Ansell2000).
Of the 354 km of developed North Solent coastline approximately 80% is protected from erosion with a variety of man-made structures (Williams et al., Reference Williams, Bray, Lloyd Jones, Steyl, Hudson and Nicholls2009). Over the last two decades sea defences have changed, with combinations of rip-rap, rock armour, concrete breakwaters and rock groynes, which have often replaced wooden structures (Dong, Reference Dong2004). These have created intermittent artificial ‘rocky shores’ which may be rapidly colonized by ‘fouling organisms’ (e.g. Bacchiocchi & Airoldi, Reference Bacchiocchi and Airoldi2003; Chapman & Bulleri, Reference Chapman and Bulleri2003). Yet, as N. lapillus has no recognized pelagic stage, it could be expected that the colonization or recolonization of these structures, would be slow, even as TBT concentrations potentially diminish (Smith et al., Reference Smith, Thain and Barry2006). Moreover, the Solent region has a complex hydrography and tidal regime, that together with fast offshore currents around headlands immediately east and west of the region may create barriers to colonizers from existing populations (Herbert et al., Reference Herbert, Southward, Sheader and Hawkins2007, Reference Herbert, Southward, Sheader and Hawkins2009).
This study was prompted by initial casual observation of dogwhelks on rock groynes constructed at Highcliffe (Christchurch Bay), where N. lapillus were not previously documented. Subsequently the aims were to monitor and record the recovery of N. lapillus populations, against TBT decline, on the south coast, and in particular the colonization of man-made substrates across the Selsey–Poole coastline. As they may prove to be source sites for mainland population recovery, data from recovering populations on the Isle of Wight, on both natural and man-made substrate were also included to give a complete regional assessment.
MATERIALS AND METHODS
The Selsey–Poole coastline, including the Isle of Wight, was surveyed to identify suitable substrates (natural and man-made structures constructed of natural stone or concrete, in moderately exposed or exposed locations) that could be expected to support N. lapillus. The distribution of rocky structures was recorded by use of Ordnance Survey maps, data and aerial photographs.
Seven sites between Sandbanks and Calshot were surveyed in 2006. These were re-examined in 2008 in addition to the coast from east of Southampton Water to Selsey, giving a total of eleven mainland sites assessed in detail in 2008. Five sites on the Isle of Wight were surveyed in 2007. These results are compared with historic (1995 onwards) data where available (Bray, Reference Bray2005).
Where N. lapillus were identified on mainland sites, individuals were counted over three 10-minute timed site searches conducted at low spring tides; Isle of Wight sites were searched in a single 30 minute period, so overall effort was the same. In the mainland surveys shell size was measured to the nearest 0.1 mm. Thirty-five adults were then randomly selected to undergo imposex assessment using methods as defined by Gibbs & Bryan (Reference Gibbs and Bryan1986) and Gibbs et al. (Reference Gibbs, Bryan, Pascoe and Burt1987) in which the vas deferens sequence (VDS) and relative penis size index (RPSI) are calculated as measures of TBT impact. The VDS in females is assessed through stages 0 (no VDS) to 6 (oviduct blocked with aborted egg capsules; stage 5 and over are sterile) and the RPSI calculated as the ‘bulk’ (length3) of the average female penis length as a percentage of the average male penis length. Imposex was not assessed where low numbers were recorded (where less than 50 individuals were found, although recovering populations on the Isle of Wight were treated conservatively). VDS data have often been presented as the female population mean value (e.g. Evans et al., Reference Evans, Nicholson, Browning, Hardman, Seligman and Smith1998, Reference Evans, Kerrigan and Palmer2000). However, as the VDS is assessed through categorical values, they are presented here as the range and median to give a representation of a population's central tendency VDS.
In terms of substrate and food resource, sites were included which appeared suitable for N. lapillus, but had none present: for example Hayling and Southsea were unoccupied, but had extensive suitable substrate; Sowley was occupied in 2006, but not in 2008. Food resources were not surveyed in detail in heavily industrialized estuaries such as Southampton Water and Portsmouth Harbour where no N. lapillus appeared to be present, apart from Netley (Southampton Water) which was monitored as part of another study (Bray, Reference Bray2005) (Figure 1; Table 1). Using stratified random 0.25 m2 quadrats, percentage cover of the main N. lapillus prey resources, common mussels (Mytilus edulis) and barnacles (Semibalanoides balanoides and Elminius modestus), was assessed in the mid to lower eulittoral zone (Lewis, Reference Lewis1964; Stephenson & Stephenson, Reference Stephenson and Stephenson1949, Reference Stephenson and Stephenson1972) as were algae, at each occupied site surveyed in 2008.
Population of Nucella lapillus and imposex impact at sites in the Solent, Southern UK. (A) Site locations: a, Sandbanks; b, Hengistbury Head; c, Highcliffe; d, Barton-on-Sea; e, Hurst Spit; f, Sowley; g, Calshot; h, Netley; i, Lee-on-Solent; j, Southsea; k, Hayling Island; l, Selsey Bill; m, Yarmouth; n, Hanover Point; o, Ventnor; p, Bembridge; (B–D) status of populations at each site: (B) 1995–2002; (C) 2006; (D) 2007–2008 (with charts indicating the prevalence of imposex and population size).
Site characteristics and dogwhelk population and imposex data for mainland coast Sandbanks (Dorset) to Selsey Bill (Sussex). Survey data for Lee-on Solent, Southsea, Hayling Island and Selsey Bill were only gathered in 2008. No individuals were found at Netley in either year. Abundance shows total number found in three 10 minute searches.
RSPI, relative penis size index; VDS, vas deferens sequence; N/A, not applicable.
RESULTS
On the mainland coast between Sandbanks and Selsey Bill, 95 individual or grouped structures, and sub-optimal natural substrates enhanced by anthropogenic debris such as concrete blocks were recorded as having potential to support N. lapillus. Most are within 5 km of each other, but there is an approximately 13 km gap between Sandbanks and Hengistbury Head and only a single outfall pipe between Hurst Spit and Langley (2.5 km south-west of Calshot) (Figure 1). Nucella lapillus populations were recorded at eight mainland sites across the study area (in 2006 a single dogwhelk only was recorded at a ninth site, Sowley); seven of the eight populations were on man-made substrates. No dogwhelks were recorded in Southampton Water or Portsmouth Harbour despite suitable habitat in docksides, breakwaters and sea walls, and none were found at Hayling or Southsea despite extensive suitable substrates and relatively ‘clean’ conditions.
In 2008 the largest population recorded was on rock groynes at Barton-on-Sea (total 307 specimens). Total N. lapillus counts in 2008 were generally slightly lower than those of 2006. (Figure 1; Table 1), however total counts from Highcliffe and Calshot increased in 2008.
Where N. lapillus were recorded, prey assemblages varied from site to site. Hayling Island had the highest percentage barnacle cover (49.7 ± 13.4%), but no N. lapillus. The sites with the lowest mean barnacle cover were Sandbanks with 1.0% (±0.6 SE); and Selsey (3.0 ± 0.4% barnacle cover) where N. lapillus were noted feeding upon the common periwinkle, Littorina littorea (see Crothers, Reference Crothers1985); the only other observed potential prey type present at Sandbanks was limpets on which N. lapillus will also forage (Crothers, Reference Crothers1985). An independent sample t-test comparison for percentage barnacle cover at Sowley and Selsey showed they were not significantly different, but whilst Selsey supported 276 N. lapillus, none were recorded at Sowley. Although Southsea did not support N. lapillus, mean percentage covers of each prey taxon were similar to Hengistbury Head (which supported a dogwhelk population); independent sample t-tests showed that the prey assemblages were not significantly different.
In 2006, all mainland sites (Figure 1) showed TBT impacts except Hurst Spit. For 2008, of the seven mainland sites supporting N. lapillus sufficient for imposexing, only Highcliffe was unaffected (RPSI, 0, VDS, 0), a reduction from the 2006 VDS range of 0–4 (Table 1).
In 2006 and 2008 the Sandbanks VDS range was 0–4. The 2008 Hengistbury Head, Barton-on-Sea and Hurst Spit populations showed little TBT impact. Only one female from each exhibited imposex (Figure 2; Table 1), with the one at Barton-on-Sea at VDS stage 3. Nevertheless, both Barton-on-Sea and Hurst Spit populations had an increased VDS range over 2006 results (Table 1). The remaining mainland locations with populations present (Table 1), Calshot and Selsey, had VDS ranges of 0–3, but, as with all sites, except Sandbanks in 2006 (Table 1), in both sample years the median VDS value was 0, highlighting the dominance of unaffected females. However, in 2008 the similar (>40%) percentage of affected females at the sites near commercial ports (Sandbanks and Calshot) highlights the on-going imposex impact at such locations (Table 1). In addition, despite suitable habitat, the other commercial traffic-affected location, Southsea, did not support any individuals, though this may equally be attributable to N. lapillus not yet having achieved successful colonization because of TBT impacts.
The frequency of each vas deferens sequence stage for mainland Solent sites in 2006 and 2008. No imposex data were collected from Calshot and Selsey in 2006. No specimens exhibited stage 5 or stage 6 criteria.
Mean male penis lengths varied from 1.7 mm (Calshot) to 3.1 mm (Barton-on-Sea) in 2008. Sandbanks had the greatest mean female penis length (0.34 mm) reaching 13.6% of the mean male penis length (2.52 mm). This is reflected in the RPSI of 0.25%; all other sites had RPSIs below 0.09% (Table 1).
Assessment on the Isle of Wight in 2007 showed population recovery on Solent shores (Bembridge, Ventnor and Hanover Point) and man-made structures at Cowes and at Yarmouth. At Bembridge and Hanover Point there were reduced imposex levels compared to 1997 values (Table 2). The other sites were not assessed for imposex in 1997 due to low abundance; Cowes was not assessed in either 1997 or 2007.
Site characteristics, dogwhelk population and imposex data for Isle of Wight for 1997 and 2007. Abundance data show number found in 30 minutes.
RSPI, relative penis size index; VDS, vas deferens sequence; N/A, not applicable.
Shell lengths varied considerably, ranging from 12.2 mm (Hurst Spit 2008) to 45.4 mm (Highcliffe 2006). Hurst Spit (2006) had the highest proportion of small individuals (7.8% up to 14 mm). Lee-on-Solent's population (2008 only) consisted of only very large individuals ranging from 38.7 to 44.4 mm, although, only five specimens were found at this site. Two sample Kolmogorov–Smirnov Z-tests (KSz) (Siegel, Reference Siegel1956) were applied to the 2006 and 2008 data to determine whether neighbouring populations had similar size–frequencies, that might indicate the presence of metapopulations augmented by dispersers. Barton-on-Sea and Hurst Spit had similar populations in 2006 (6.3 km apart), and Hengistbury Head was similar to both Hurst Spit and Calshot in 2008; (12.1 km and 21.1 km apart respectively). All other neighbouring sites had significant differences (2006, KSz all others P < 0.01; 2008, all others P < 0.05). Analysis of 2006 and 2008 data demonstrated a significant negative logarithmic correlation between abundance and mean shell length (r = 0.73, N = 14, df = 13, P < 0.05; y = –3.6678 ln(x) + 48.124) (Figure 3).
Relationship between abundance and mean shell length at each, mainland Solent site in 2006 and 2008. Plot omits data (one individual) from Sowley in 2006.
DISCUSSION
The results show an unexpectedly rapid spread of dogwhelks across the Solent into areas which have either never supported known populations, or where they have been extinct for around thirty years. This increase in distribution has taken place as restriction on the use of TBT has tightened, but appears to have been facilitated by the widening occurrence of man-made hard substrates. Where suitable natural habitat was present, such as at Bembridge on the Isle of Wight, engineered structures on the shore may have acted as stepping stones across less favourable habitats thus accelerating re-colonization and regional recovery of populations. Distances between occupied sites varied from as little as 1.3 km (Highcliffe to Barton-on-Sea) to 35.6 km (Lee-on-Solent to Selsey Bill). This investigation confirmed that man-made rocky structures have potential to provide a valuable intertidal habitat for recovering N. lapillus populations. However, they are not essential for populations to survive, since N. lapillus were abundant at Selsey Bill, where there was little or no man-made or natural rocky substrate, but areas of gravel, large cobbles and concrete debris stabilized by timber groynes. Nucella lapillus aggregations were present on shingle surrounded single boulders and a highly TBT impacted population was recorded at Selsey as long ago as 1990 (Spence et al., Reference Spence, Bryan, Gibbs, Masters, Morris and Hawkins1990) with evidence of high TBT levels experimentally recorded using imposex indices by Harding et al. (Reference Harding, Bailey and Davies1992).
Whilst six recently established N. lapillus populations have been identified, with the mainland Solent dogwhelk distribution apparently having increased significantly, their persistence may be unstable because of the high proportion of juveniles (indicated by the smaller shells). The large population at Barton-on-Sea declined slightly from 2006 to 2008, while numbers increased at Highcliffe and Selsey; the notable abundance decline at Hurst Spit may be attributed to factors such as wave exposure and predation pressure influencing N. lapillus abundance (Miller et al., Reference Miller, Fernandes and Read1999) as well as bad weather hindering searches, rather than TBT pollution. ‘Abundance’ in this study actually refers more to the rate of individuals found (animals per three 10 minute replicate searches), which reflects both the population density and the nature of the area being searched, so is subject to some degree of error; deep crevices take more time to investigate than flat surfaces. However, abundance as recorded appeared to have a significant relationship with shell size; smaller shells were found at sites newly colonized or recovering.
The continued use of TBT antifouling paint on large vessels and the restricted water and sediment movement within ports prevents complete recovery from TBT poisoning (Smith et al., Reference Smith, Thain and Barry2006). TBT has been observed to persist in sediments and could potentially provide a ‘reservoir’ of contamination for decades (Maguire, Reference Maguire2000). Benthic sediments are the primary means of TBT bioaccumulation in deposit feeders, particularly bivalves (Langston et al., Reference Langston, Bryan, Burt and Gibbs1990). Contaminated sediments could continue to be a TBT pollution source well into this century (Langston & Pope, Reference Langston and Pope1995). Desorption of TBT from sediments increases concentrations in overlying water to quantities which exceed the Environmental Quality Standard of 2 ng l−1 (Langston & Pope, Reference Langston and Pope1995). Wave simulation experiments indicate that TBT paint particles move upwards to the sediment surface where it is most available (Eggleton & Thomas, Reference Eggleton and Thomas2004) and simulated removal of the top sediment layer releases TBT back into the water column, leading to a higher incidence of imposex (Svavarsson et al., Reference Svavarsson, Granmo, Ekelund and Szpunar2001). In addition, not only does the disturbance of contaminated sediment cause remobilization of TBT at dredge sites, levels at disposal sites have been implicated in imposex impacts in marine protected areas (Boersma & Parrish, Reference Boersma and Parrish1999; Terlizzi et al., Reference Terlizzi, Delos, Garaventa, Faimali and Geraci2004).
Ongoing dredging and disposal within the region may threaten complete recovery of N. lapillus populations. Measurements of TBT taken from East Brambles Buoy (2.5 km offshore of Calshot, at the entrance to Southampton Water) from February 2000 to February 2007, gave a mean water concentration of 2.465 ng l−1 (unpublished data, D. Lowthion, Environment Agency, 2007), enough to cause imposex in dogwhelks (Gibbs et al., Reference Gibbs, Bryan, Pascoe and Burt1987) and Calshot had 45% of females with imposex in 2008 (Table 1). Thomas et al. (Reference Thomas, Blake and Waldock2000, Reference Thomas, Fileman, Readman and Waldock2001) indicated that TBT would be a factor in non-target organism impacts in the region for some time. This may continue to be reflected by the intensity of maritime activity around the Port of Southampton as well as sediment disturbance due to regular large scale dredging.
It is unclear why N. lapillus has not yet colonized Southsea, close to Portsmouth Harbour. Assessment of the potential food resources for dogwhelks indicate that the site is capable of supporting a population, yet the proximity of a port may be a key factor; its absence may be attributed to TBT levels as the site is within 1 km of Portsmouth Naval base and a commercial ferry route. Alternatively, it may be that recruitment is limited due to local hydrodynamics, and lack of nearby source populations; Lee-on Solent, itself with a small, isolated population, is 8.3 km away with extensive docks and dredged channels in between.
For mainland sites there was an overall decrease in RPSI between 2006 and 2008, except for at Barton-on-Sea and Hurst Spit, where although values increased (Table 1) they were still very low. The continued evidence of TBT impact in the Barton-on-Sea/Hengistbury Head/Hurst Spit area, which is distant from the larger regional harbours, may be due to persistent low level contamination or a local hotspot (perhaps TBT in sediments), or the irreversibility of imposex (Bryan et al., 1986). The lifespan of N. lapillus is typically around six years (Feare, Reference Feare1970; Crothers, Reference Crothers1985) meaning that individuals alive when TBT concentrations were higher may still be present. In 2008 all sites had low RPSI-negligible in comparison with other south coast sites in the 1990s (Spence et al., Reference Spence, Bryan, Gibbs, Masters, Morris and Hawkins1990; Evans et al., Reference Evans, Leksono and McKinnell1995, Reference Evans, Evans and Leksono1996; Huet et al., Reference Huet, Paulet and Le Pennec1996), though they were apparently not completely TBT-free; with the exception of Highcliffe where the increased population has colonized man-made groynes.
Similarities in shell lengths between sites several kilometres apart could simply be attributed to wave exposure (comparable at these locations) rather than an indication of population connectivity, as it has been shown to influence phenotypic expression in N. lapillus (Staiger, Reference Staiger1957; Kitching et al., Reference Kitching, Muntz and Ebling1966). However, medium distance movements of 10–100 km have proven fairly common for N. lapillus, assuming populations showing genetic similarity have been founded or augmented by one or another (Colson & Hughes, Reference Colson and Hughes2004), though the means of travel are poorly understood. Hydrodynamic forces may carry individuals for distances of up to 1 km during storms (Bray, Reference Bray2005) and N. lapillus individuals were found rolling between rock groynes at both Hengistbury Head and Highcliffe. In addition birds may drop individuals (Crothers, Reference Crothers1985), but dislodgement and passive transport is more likely.
Shore macrofauna can travel hundreds of kilometres by rafting (as adults or eggs) on detached clumps of floating seaweed (Ingólfson, Reference Ingólfson1995). To determine whether this is the case for N. lapillus, placing surface drifters and performing genetic examination could highlight whether gene flow follows the direction of the currents (Muhlin et al., Reference Muhlin, Engel, Stressel, Weatherbee and Brawley2008). The long-standing, relatively unaffected, intertidal populations on the south coast of the Isle of Wight (Herbert Reference Herbert1988; Bray & Herbert, Reference Bray and Herbert1998) may have been a source of founding individuals on the mainland. Populations on the north coast of the Isle of Wight had become extinct by the mid-1980s (Herbert, Reference Herbert1988; Bray & Herbert, Reference Bray and Herbert1998), yet have recovered since recolonizing individuals were observed in 1995 at East Cowes.
It is more likely that populations on the mainland are being replenished and/or augmented by dispersers from subtidal populations that have survived the worst of the contamination. Crothers (Reference Crothers1998, Reference Crothers2003) postulated that a recolonized population of N. lapillus in Watermouth Cove in the Bristol Channel was repopulated by a subtidal population, taking 13 years to travel 30 miles. We noted a subtidal population on an outfall pipe at Hengistbury Head that extends into Christchurch Bay where there are shallow rocky ledges. Thus it may be that the Hurst Spit population was founded by subtidal colonizers, possibly explaining between-site similarity in cumulative shell lengths. Whether dispersers can form a viable population that persists in time depends on, amongst other things, demographic, environmental and genetic stochasticity (Shaffer, Reference Shaffer1981), prey availability and also, more practically, engineering activity maintaining or changing sea defence structures. Some populations will not be successful because the habitat is unsuitable for them. This is most likely the case for Sowley where prevalence of prey was limited and substrates marginal, thus colonizers (as seen in 2006) cannot establish. The recovery of N. lapillus could be partly attributed to selection for individuals that have the ability to breed successfully, despite high concentrations of TBT (Birchenough et al., Reference Birchenough, Evans, Moss and Welch2002; Huet et al., Reference Huet, Paulet and Le Pennec1996, Reference Huet, Le Goïc and Gibbs2008). The recessive and uncommon male genital defect Dumpton Syndrome (DS) results in reduced penis size in males, and also in females affected by TBT (Gibbs, Reference Gibbs1993). The condition was previously undetected, but TBT impacts highlighted it in some isolated populations (Gibbs, Reference Gibbs2005). Populations exhibiting DS have survived in areas of high TBT contamination, as DS-affected females avert sterilization by imposex (Huet et al., Reference Huet, Paulet and Le Pennec1996). Males taken from Calshot had smaller penis sizes than other sites, a symptom of DS. However, as DS occurs in isolated populations and this work suggests possible connectivity between recolonizing groups, this may be because of general underdevelopment.
This study shows that following UK restrictions on the use of TBT, N. lapillus populations in the southern UK region have recovered and spread into locations where they may not have been previously present. The evidence presented here leads us to hypothesize that this recovery may have been accelerated through the construction of new sea defences, as the animals have been found on these structures at locations where they were previously unrecorded. These structures provide a new habitat, offering intertidal ‘stepping stones’ to sites where dogwhelks may not have been previously present or were eliminated by TBT pollution some 30–40 years ago. Some locations remain uncolonized, possibly due to residual TBT, or unsuitability and isolation; sites near to ports still show indications of TBT impact expressed as imposex, and no colonization has taken place within industrialized estuaries although extensive suitable substrate is present. The means of long-range dispersal of N. lapillus in the Solent remains unclear, but what has been demonstrated is that man-made substrates can provide useful habitat and facilitate more rapid recovery of N. lapillus as TBT pollution declines. For species with a short pelagic larval stage or with direct development, the population connectivity between patches of harder substrata may therefore be much greater than previously thought, even within a hydrodynamically complex region such as the Solent and central south coast of England.
ACKNOWLEDGEMENTS
The authors would like to thank John Crothers for his comments on an earlier version of the manuscript; Jenny Mallinson (National Oceanography Centre); and several volunteers who assisted with field data collection and David Lowthion (Team Leader, Marine Team SE, Environment Agency) who provided data on TBT. Three anonymous referees provided helpful comments which considerably improved this paper. This research received no specific grant from any funding agency, commercial or not-for-profit sectors.
