The status of Clibanarius erythropus after a recent range expansion to Great Britain, with the highest latitude recording of a gravid individual

In 2016, the range of the hermit crab Clibanarius erythropus expanded to South West Britain for the second time. C. erythropus primarily lives in the Mediterranean and the Atlantic coast of Europe from the Bay of Biscay to Morocco. The species has now been recorded on both the north and south coast of the South West peninsula of the UK from Newtrain Bay, on the north coast of Cornwall, to Wembury, on the south coast of Devon. It is unknown if the crab’s reappearance in the UK has been caused by a one-off colonisation event or by a continued influx of larvae. The population in the UK is made up of individual within a narrow size bracket, indicating a single colonisation event took place, and that the population is an ageing one. However, we also report the highest latitude recording of a gravid individual for the species. A lack of gravid individuals was suggested to be why the species was unable to sustain its presence in the UK following a previous colonisation in 1960. This discovery hints that rising water temperatures may allow C. erythropus and other warm-water species to expand and sustain themselves in the UK. We also found crossover in shell utilisation between C. erythropus and the native hermit crab Pagurus bernhardus, suggesting that competition might occur between the two species.


Background
Clibanarius erythropus is a warm water hermit crab whose distribution extends from Morocco to the coast of France and includes the Mediterranean (Tricarico et al. 2009). In the 1950s, observations suggested that the range of C. erythropus was expanding northwards beyond the Bay of Biscay around Brittany with the highest latitude record being at Roscoff in 1955 (Southward and Southward 1977).
The species made even further progression northwards when it was discovered in the UK in the winter of 1959/60 (Carlisle and Tregenza 1961). However, at the majority of sites, the species' appearance was short-lived. In 1967, the 'Torrey Canyon' oil spill and subsequent toxic cleaning chemicals were observed to have directly killed most, if not all, the population (Southward and Southward 1977). After 1967 records of C. erythropus occurred at only two known sites, Wembury and Marazion (Southward and Southward 1977). These populations seemed unsustainable and disappeared sometime around the 1980s (Hawkins et al. 2017;Southward and Southward 1988).
It has been predicted that C. erythropus would return to the southwest of the UK as the water temperatures rose (Hawkins et al. 2008;Southward et al. 1995) and, in March 2016, C. erythropus was rediscovered in Cornwall at Castle beach (50°8′52″N, 5°3′19″W). This recording was followed by sightings at multiple locations along both the north and south coast.
The reappearance of C. erythropus is unlikely to have been caused by direct human translocation. When translocation occurs, sightings typically start at a single site and then spread outwards (Farnham and Morrell 1983). As recordings of C. erythropus appeared simultaneously across the coast, it is more likely that its planktonic larvae drifted to the UK from further south.
We describe the current known distribution of C. erythropus in the UK, as well as data on the abundance of C. erythropus at five locations in Cornwall. To explore the age structure of C. erythropus in the UK, we recorded the size range of the individuals. Hermit crabs cannot be aged directly and the shell species a hermit crab is occupying alters the growth rate (Bertness 1981); therefore, we use size as an approximation of age. If the UK population of C. erythropus is capable of reproducing or is the result of a continued influx of pelagic larvae, then the age structure of the UK population should resemble those at other sites in Europe. C. erythropus, in Europe, have cephalothorax shield lengths ranging from 1.3 mm to 7.4 mm, with a high ratio of individuals having shield lengths smaller than 3.0 mm (Cuesta et al. 2016). Gherardi and Benvenuto, (2001) found the mean C. erythropus cephalothorax shield length in Europe was 1.99-2.10 mm, although this does vary between sites . Conversely, if there has been a single incursion of pelagic larvae and no selfrecruitment has occurred, individuals in the UK should be restricted to a narrow size range. Long term monitoring of the UK rocky shore has shown that the range of many intertidal species can alter in response to a changing climate (Hawkins et al. 2008;Mieszkowska et al. 2006).
We suggest that the reappearance of C. erythropus on UK coasts could be an important indicator of how the ranges of other warm-water species could expand to the UK over time as climate changes.
We also explored the shell species that C. erythropus uses in the UK, in order to make predictions as to how the species could interact with the native UK fauna; mainly the UK's only abundant intertidal hermit crab Pagurus bernhardus, through shell competition. From the 1960s to the 1980s, C. erythropus predominantly utilised Nucella lapillus shells (Southward and Southward 1988) even though this is not the most abundant intertidal gastropod (Wilson-Brodie et al. 2017).
Rockpools were selected by moving parallel to the water along the mid tide mark, and all rockpools greater than two square metres were searched. Each rockpool was searched for 10 min per two square metres of the water's surface area. The species of shell occupied by any C. erythropus found was recorded according to Crothers (2001). In order to measure crab size, a camera (Olympus Tough TG-4) was placed 40 cm above the ground, pointing vertically downwards at a ruler, which had divisions of one millimetre, on a flat substrate. Each crab was placed adjacent to the ruler with the aperture of their shell facing vertically upwards. This induced the hermit crab to partly emerge from their shell and as this occurred, photographs were taken at 4x optical zoom. After each individual was measured, they were placed back into the rockpool. If an individual did not fully emerge within 10 min, they were placed back into the rockpool and another sought.
Photographs were analysed using Image J software (Schneider et al. 2012). The different anatomical structures were measured using the ruler in each photograph as a guide. The cephalothorax shield width (CSW), the cephalothorax shield length (CSL), and the total length of the cephalothorax (TCL) were recorded for each animal.

UK distribution
The first confirmed recording of C. erythropus in the UK since 1985 (Southward and Southward 1988) was at Castle beach in Falmouth (50°8′52″N, 5°3′19″W), on the 12th March 2016. Adrian Rowlands, who was taking part in a Shoresearch citizen science survey run by Cornwall Wildlife Trust, found a single individual that was identified as C. erythropus by Matt Slater.
The survey in 2017 recorded a total of 29 C. erythropus, at Porthleven (16 individuals), Prisk Cove (12 individuals), and Bream Cove (one individual). Across all sites, we attained 13 measurements of CSW, 11 measurement of CSL, and seven measurements of TCL (Table 1). Sixteen individuals did not emerge far enough to take any of their measurements in the allocated time.

Size structure
Mean CSL was 2.47 mm (range 1.42-3.15 mm), mean CSW was 2.81 mm (range 1.73-3.63 mm), and mean TCL was 5.95 mm (range 5.16-7.65 mm). Both the largest and the smallest individuals were found at Porthleven where the majority of measurements were taken ( Table 1). The mean CSL was significantly larger than the CSL of individuals from Gherardi and Benvenuto (2001) (one-sample t-test; t 10 = 2.85, p < 0.05; t 10 = 3.70, p < 0.01, Fig. 2).

Shell utilisation
C. erythropus was found utilising four different species of gastropod shell, across all study sites. Nucella lapillus was occupied by 83% of C. erythropus (24 individuals), Littorina littorea was occupied by 6.8% (two individuals), Tritia reticulata was occupied by 6.8% (two individuals), and Buccinum undatum was occupied by 3.4% (one individual). The number of C. erythropus occupying each species of gastropod shell differed significantly (X 3 = 51.69, p < 0.001). Elongated, opposed to globose, shells were utilised by 93% of individuals.
Individuals utilising N. lapillus shells had a mean CSL of 2.49 mm (range 2.18-2.86). The smallest individual (1.47 mm) was found utilising a B. undatum shell and the largest (3.15 mm) a L. littorea shell. One individual occupying a T. reticulata shell had a CSL of 2.76 mm (Fig. 2).

Gravid individual
On October 8th 2018, at Gyllyngvase West, a preliminary study that involved extracting three individuals from their shells was conducted. One individual was gravid with a mass of 300+ eggs gathered round its abdomen. Each egg had a visible eye pigment which was irregularly rounded, indicating they are at a late stage of development (Turra and Leite 2007). The gravid individual was found occupying a N. lapillus shell and had a CSL of 2.4 mm (Fig. 3). The other two individuals were a male with a CSL of 5.4 mm and a non-gravid female with a CSL of 3.3 mm both occupying a L. littorea shell. Fig. 2 The size structure (cephalothorax shield length) of Clibanarius erythropus across different species of Gastropod shell from all study sites. Dashed line marks the mean cephalothorax shield length of clustered individuals studied in Europe by . Photographs adapted from those of H. Zell Fig. 3 Photographs of the gravid individual discovered at Gyllyngvase beach (50°8′ 43″N, 5°3′ 48″ W) on the 8th October 2018. A shows the individual fully extracted from its shell with the egg mass gathered around its abdomen using its pleopods. B shows the individual partly extracted with the egg mass held inside the cavity of the N. lapillus shell

Discussion
After a gap of 30 years, records of C. erythropus now frequently occur on both the north and south coast of Cornwall. Our measurements of CSL, the most commonly used measurement of size for C. erythropus, suggests that the range of sizes found in the UK population is reduced compared to those in the Mediterranean and Iberian Peninsula . No individual was larger than 3.5 mm (Fig. 2), indicating that individuals in Cornwall have not been able to reach the species' maximum size. This conforms with the theory that C. erythropus has only recently reappeared in the South West and was caused by a single incursion of the species pelagic larvae to the UK. The conclusion that C. erythropus arrived in a single incursion event and is an ageing population is based on the measurement of only 11 individuals and should be treated with caution. If this interpretation is correct, and no further colonisation events occur, we should see a gradual increase in the ratio of larger individuals as the population ages. The growth rate of C. erythropus is unknown, meaning we cannot age the individuals in order to estimate when the larvae first settled in the UK. However, being part of the Clibanarius genus, C. erythropus is notably different from the other species of hermit crab found UK. This, combined with the active marine biological recording community in Cornwall, means the species is unlikely to have gone unnoticed for long. Thus the influx of larvae probably occurred in 2015 or early 2016.
The number of C. erythropus found, during this study, varied between survey sites (Table 1). This may have been because of an actual difference in abundance between sites or it could be because the species tends to congregate into small areas, possibly to increase shell exchange . As such, the stochasticity of finding clusters of individuals may have caused the high numbers of C. erythropus found at Prisk Cove compared to Gyllngvase beach and Bream Cove. At Porthleven, while only 16 individuals were measured and recorded due to constraints of the tide cycle, a high number of C. erythropus inhabited nearly all other rockpools at the site.
The majority of C. erythropus were found in N. lapillus (83%), with other shells being utilised at a much lower percentage, e.g. L. littorea (6.8%). This is similar to the percentages (72% N. lapillus and 11% L. littorea) used 2 years after the first colonisation, in 1961 (Southward and Southward 1977). In 1976 all individuals were found in N. lapillus (Southward and Southward 1988). This study further supports the notion that N. lapillus is the gastropod species most commonly occupied by C. erythropus, in the South West of the UK. However, in comparison to our results in 2017, research conducted in autumn 2016 showed a higher percentage use of Littorea spp (70%), while N. lapillus was occupied by only 22% (Antony 2017). This suggests that within the first year of colonising, C. erythropus occupied Littorea spp most frequently, but as individuals age, they began occupying N. lapillus.
It was suggested that the disappearance of C. erythropus in the 1980s was partly caused by the reduction in N. lapillus due to Tributyltin pollution (Southward and Southward 1988). The banning of Tributylin and the subsequent increase in N. lapillus (Birchenough et al. 2002) may have been a factor in allowing C. erythropus to return to the UK, but as C. erythropus did not reappear until 2016 and in an apparent single incursion this is unlikely to be the only factor involved.
The presence of C. erythropus in the UK may affect P. bernhardus the most common intertidal hermit crab in the UK. Records of P. bernhardus become rarer further south than Brittany (GBIF.org 2019); this means the expansion of C. erythropus from the Bay of Biscay to the UK has caused overlap in the ranges of the two species. While behavioural and dietary differences discussed in Southward & Southward (1977) may allow the species to coexist into the future, interspecific shell competition could occur. In the UK, C. erythropus is predominantly utilising N. lapillus shells, which are elongated and similar to those occupied in its traditional range (Tricarico et al. 2009). However, C. erythropus prefers globose shells , which allow them to reach larger sizes (Cuesta et al. 2016) and have increased locomotion (Benvenuto et al. 2003). It is possible that in the UK P. bernhardus is outcompeting C. erythropus for the preferred globose shells, such as L. littorea. However, C. erythropus, in the UK, have not reached the largest size class of the species and may currently be unable to utilise the larger globose L. littorea shells. As we predict the population of C. erythropus will grow into these large size classes in the future, competition between C. erythropus and P. bernhardus, for the larger and globose shell species, may become fiercer.
Gravid individuals of C. erythropus were not observed during the previous appearance in the UK and the sea surface temperature of the South West was thought to be too low for reproduction (Southward and Southward 1977). Within the Mediterranean and Atlantic range of C. erythropus, sea surface temperature extremes range from~10 to 25°C. Females are gravid during the warm summer months (Harms 1992), and on the Iberian coast, C. erythropus can reproduce where the average sea surface temperature in summer is 16.8°C . Harms (1992) concluded that larvae can develop successfully above temperatures somewhere between 15 and 18°C. At 15°C, the larvae developed through all 4 zoea stages but did not develop into the final megalopa stage, which can move into an empty shell and moult into the adult form. At 18°C complete larval development was accomplished.
The gravid individual found during this study was discovered on the 8th of October 2018 when the sea surface water temperature around the south coast of Cornwall was 13.3°C (Channel Coastal Observatory, 2019). As water temperatures fall rapidly during this time of year, and given the results of Harms (1992), it is unlikely that these eggs would have produced larvae capable of surviving through to the adult stage. However, sea water temperatures around the southwest of the UK did exceed 15°C, and briefly 18°C, over the summer of 2018, and the maximum temperature of UK rockpools can reach 24°C (Hopkin et al. 2006). As such, if C. erythropus were able to produce larvae earlier in the year a full reproductive cycle may be possible.
The return of C. erythropus to the UK may serve as in interesting proxy for other warm water species that have been expanding their range northwards in recent years (Keith et al. 2011;Mieszkowska et al. 2007;Mieszkowska et al. 2006), for example the Marble crab, Pachygrapsus marmoratus (Ingle and Clark 2008). The size structure of individuals suggests that the recent colonisation of C. erythropus to the UK was a single event and recruitment remains rare. Thus, ocean currents and temperatures may not be conducive to regular incursions by pelagic larvae from species native to European waters. Monitoring the distribution of C. erythropus across the UK, as well as observing the current populations for signs of reproduction or recruitment, could further our understanding of how quickly C. erythropus and other warmwater species can gain a larger foothold in the southwest of the UK as the climate changes.

Conclusion
The UK population of C. erythropus appears to be an ageing one, resulting from larvae founding populations in just 1 year. However, the discovery of a gravid individual suggests the potential reproduction of C. erythropus and that the species could extend its presence in the UK indefinitely. It took 25 years for C. erythropus to disappear from the UK after the previous colonisation event (Southward and Southward 1988). Therefore, even if no self-recruitment occurs C. erythropus will likely remain a noteworthy addition to the UK's intertidal fauna for another two decades.