Does climatic warming explain why an introduced barnacle finally takes over after a lag of more than 50 years?
- First Online:
- Cite this article as:
- Witte, S., Buschbaum, C., van Beusekom, J.E.E. et al. Biol Invasions (2010) 12: 3579. doi:10.1007/s10530-010-9752-5
- 263 Views
Invading alien species may have to await appropriate conditions before developing from a rare addition to the recipient community to a dominance over native species. Such a retarded invasion seems to have happened with the antipodean cirripede crustacean Austrominiusmodestus Darwin, formerly known as Elminius modestus, at its northern range in Europe due to climatic change. This barnacle was introduced to southern Britain almost seven decades ago, and from there spread north and south. At the island of Sylt in the North Sea, the first A. modestus were observed already in 1955 but this alien remained rare until recently, when in summer of 2007 it had overtaken the native barnacles Semibalanus balanoides and Balanus crenatus in abundance. At the sedimentary shores of Sylt, mollusc shells provide the main substrate for barnacles and highest abundances were attained on mixed oyster and mussel beds just above low tide level. A. modestus ranged from the upper intertidal down to the subtidal fringe. Its realized spatial niche was wider than that of the two natives. We suggest that at its current northern range in Europe a long series of mild winters and several warm summers in a row has led to an exponential population growth in A. modestus.
KeywordsAlien speciesAustrominius modestusInvasionClimate changeNorth Sea
Translocation of species across natural barriers by human carriers has become a process of increasing global importance in aquatic coastal ecosystems (Carlton 1985; Grosholz 2002). The success of alien invaders is primarily a function of (1) propagule pressure such as at ports or open aquaculture facilities for coastal aquatic species (2) superiority by means of attributes evolved in another evolutionary context (3) release from co-evolved enemies (4) synergisms with other alien invaders (5) availability of resources in recipient ecosystems, and (6) environmental change which may stress natives but benefits pre-adapted aliens (i.e., Alpert 2006; Barney and Whitlow 2008; Sax et al. 2007; Westphal et al. 2008). The last mentioned process may turn an alien species previously considered to be a harmless addition to the recipient biota into an invader of high competitive ability. The recent global warming may entail such shifts in dominance when aliens originate from a warmer environment than their recipient regions (Dukes and Mooney 1999; Occhipinti-Ambrogi 2007; Reise et al. 2008). Thus, in cold temperate waters the biota may already comprise several aliens suspected to be ‘sleepers’. Most sources of introduced species, ports and aquaculture facilities, are located in warmer regions. Introduced long ago, sleepers may be awakened by recent climatic warming, and then potentially may run for dominance. Such time lags can be found in all phases of the invasion process of alien species (Crooks and Soulé 1999; Crooks 2005). An end of a time lag is often caused by changes in abiotic conditions of the receiving environment but may also be triggered by other factors like changes in the invader (e.g. genetics) or relaxation of biotic pressures. However, the awakening of ecological sleepers after decades due to global warming is in contrast with the common pattern of introduced aliens which after a short initial phase enter a boom and bust development, followed by a long adjustment phase with more moderate densities (Reise et al. 2006).
The barnacle Austrominius modestus Darwin (1854), formerly known as Elminiusmodestus (O’Riordan et al. 2009; Buckeridge and Newman 2010), originating from southern Australia and New Zealand, arrived in British waters some time prior to 1943, presumably by navy vessels or flying boats (Bishop 1947; Crisp 1958). Since then it has extended its range into the North Sea and the Atlantic shores, probably by a combination of larval drift and hitchhiking on boats or as larvae in ballast water (Kühl 1954; Crisp 1958; Barnes and Barnes 1968). The species is also known from the western Mediterranean Sea since 1990 (Galil 2009). In 1955 the first five individuals of A. modestus were recorded at the shore of the island of Sylt in the North Sea (Kühl 1963), which is the site of this study. In subsequent years, A. modestus has been found regularly in low numbers, became relatively common from 1993 to 1995, then was almost absent after the severe winter 1995/96, and since then gradually increased but remained rare relative to the two native barnacles, the intertidal Semibalanus balanoides (L.) and the subtidal Balanus crenatus (Bruguière), which are overlapping around low tide level (Luther 1977, 1987; pers. obs.). Other Balanomorpha occasionally encountered around Sylt are Balanus balanus (L.) on a subtidal boulder groin; this species also occurs together with B. perforatus and B. improvisus (Darwin) on drift material washed ashore; occasionally the latter has been found attached to mussels, and Verruca stroemia (Müller) to Pacific oysters imported from Ireland (Büttger et al. 2008; Reise 1998, Reise unpubl. data).
Recent studies have shown that intertidal barnacles respond in range and abundance to current climatic warming (i.e., Mieszkowska et al. 2006; Southward et al. 2005; Wethey and Woodin 2008), and it has been suggested that increased abundances of A. modestus in Ireland (Lawson et al. 2004) and on the rocky island of Helgoland not far from Sylt (Franke and Gutow 2004; Reichert and Buchholz 2006) may have been facilitated by warming. A. modestus is a small, short-living, highly fecund barnacle with a physiological optimum above 20°C (Harms 1999) and suffers high mortality during cold winters (Crisp 1964; pers. obs.). At Sylt, the last severe winter with strong effects on the littoral fauna occurred in 1995/96 (Armonies et al. 2001) and since then temperatures have been increasing (MacKenzie and Schiedek 2007; Martens et al. 2008). We therefore conducted an assessment of the distribution and abundance of barnacles around Sylt. Except for coastal defence structures, shores are composed of sediments. Biogenic substrates such as mollusc shells constitute the main substrate for barnacles. We hypothesize that A. modestus has improved its position relative to the natives S. balanoides and B. crenatus, which used to dominate the higher and lower shore, respectively.
Materials and methods
Study area and experimental sites
Six sites were sampled on the sheltered eastern side and one site at the exposed western side of the island Sylt (Fig. 1). At the sheltered site, four transects from high tide line (+0.9 m above mean water level) to low tide line (−0.9 m below mean water level) were surveyed to assess barnacle distribution and abundance. Adjacent to each transect, we investigated barnacle occurrence at a subtidal site (2–6 m below mean low tide level; S1-S4 in Fig. 1). In addition to the transects, we quantified barnacle density at two sheltered groins (G1 and G2 in Fig. 1) at the eastern side and at one exposed groin (G3 in Fig. 1) at the western side of Sylt.
Each transect crossed sediments with very little potential substrate as well as a mixed bed of native blue mussels Mytilus edulis L. and introduced Pacific oysters Crassostrea gigas (Thunberg) with plenty of substrate for barnacles. In the area, Pacific oysters were regularly introduced for aquaculture purposes since the mid of the 1980s, and subsequently a spread by oyster larvae to native mussel beds was observed (Reise 1998). Since 2005, most mussel beds have turned to oyster dominated reefs due to comparatively high water temperatures in summer enabling high recruitment success of C. gigas (Diederich et al. 2005; Kochmann et al. 2008).
Depending on barnacle densities, differently sized quadrates of 0.25 × 0.25 m (smallest size) to 10 × 10 m (largest size) were used for sampling, placed randomly at stations along each transect. At each transect station 3–7 replicate samples were surveyed. Within each quadrat barnacles were separated by species and counted. Of the three barnacle species encountered in this survey (S. balanoides, B. crenatus and A. modestus), only those with a basal diameter >2 mm were considered.
Stations along transects were subjectively chosen to include all tidal zones and habitat types. The three groins for coastal defence which were surveyed (Fig. 1), were made out of wooden pilings set up in double rows. Due to high barnacle density, a small frame of 2.5 × 2.5 cm was used for counting, with 36 replicates at each of the three groins. All surveys were carried out in May and September 2007, respectively.
Long-term data on abundance of A. modestus attached to mussel and oyster beds in the List tidal basin were available from the literature (Albrecht and Reise 1994; Görlitz 2005), unpublished surveys of the authors, and from a blue mussel bed monitoring program conducted annually since 1998 in the northern Wadden Sea (Nehls and Büttger 2006).
To reveal the pattern of recruitment across the tidal gradient, cleaned oysters (maximal shell lengths 120–150 mm) were offered to barnacle cypris larvae in Königshafen, a small embayment within List tidal basin (Fig. 1). Cylindrical cages (13 cm high, 25 cm in diameter) were made out of 1-cm² plastic meshes and were stabilized with a plastic ring above and below. In July 2007 a total of 20 cages with 3 oysters each were exposed at four tidal levels: upper (0.3 m above mean water level), mean (±0.0 m mean water level which is equivalent to chart datum) and lower intertidal (0.6–0.8 m below mean water) and sub tidal fringe (1.0 m below mean water level). At each tidal height five cages were set up in a row with distances of 1 m between them. Each cage was anchored by 3 iron rods and pushed 5 cm deep in the sediment to prevent predators intruding from underneath. At regular intervals, cages were checked for entangled drift algae that were removed when necessary. Three months later in October 2007 the number of barnacles that had settled on the oysters was counted.
To investigate differences in recruitment of A. modestus between blue mussels and Pacific oysters as substrate, similar cages were set up at low tide level from July to September 2007 with cleaned M. edulis and C. gigas (lengths 75–90 mm). Six cages enclosed 3 oysters and 6 others had 3 mussels each. Cages were arranged in pairs within a 10 × 10 m plot.
Abundances are given as arithmethic means with standard deviation (SD). For statistical treatment, analysis of variance (ANOVA) was used. Significant differences were analysed using Tukey’s honest significant difference (HSD). Homoscedasticity of variances of dependent variables were tested by using Levene’s Test. Individual tests on homogeinity of variances are not presented as these have little bearing on our overall conclusion on the dominance of A. modestus in the tidal zone. Heterogeneity of variances was reduced by log-transformation. For some analyses, however, transformation did not remove heterogeneity. In these cases untransformed data were used (Underwood 1997). Interpretations of analyses for which variances were heterogeneous were made cautiously, noting the robustness of analyses where sufficient samples are used. Effects were considered to be statistically significant, if P was <0.05.
Barnacle occurrence and distribution
Mean abundances (MA: individuals >2 mm diameter m−2) and coefficients of variance (CV) of barnacles, Austrominius modestus, Semibalanus balanoides and Balanus crenatus, averaged over all transect abundance data in early and late summer 2007
Settlement of A. modestus on cleaned and caged oysters exposed from July to October 2007 was rather variable and significantly different between tidal zones (ANOVA, P < 0.007). Means (SD) per oyster in the upper, mid and lower intertidal were 14.2 (15.8), 22.9 (32.8) and 9.9 (8.7), respectively. At the shallow subtidal site settlement was significantly lower with only 1.0 (0.9) per oyster (Tukey’s HSD, P < 0.004), reflecting the low share of A. modestus on dredged material. No significant difference was found between mussels and oysters as substrate for settling A. modestus, with 2.2 (1.5) and 1.4 (1.0) individuals cm−2, respectively.
Recent climate change is affecting species distributions, abundances and phenologies across ecological systems (Walther et al. 2002; Parmesan and Yohe 2003). This is apparent in various biota of the North Sea (i.e., Beare et al. 2004; Beaugrand et al. 2003; Kirby et al. 2007; Perry et al. 2005; Pörtner and Knust 2007; Wiltshire and Manly 2004). We here suggest that in northern temperate coastal waters, alien species may spearhead biotic responses to climatic warming and that on top of an increasing tide of invaders their share in the recipient biota will be further enhanced by climate change. The barnacle Austrominius modestus originating from the SW Pacific may exemplify this trend and be added to three other alien invaders which also took benefit from recent warming at the shores of the island of Sylt.
At the sheltered upper shore, the cord grass Spartina anglica, introduced into the Wadden Sea already in the 1920 s, recently began to spread again, suggested to be triggered by an earlier warming in spring (Loebl et al. 2006). When Pacific oysters Crassostrea gigas were introduced for farming in the 1980 s, it was assumed that they could not recruit at the prevailing low temperatures but recent warming in the summer months greatly facilitated their recruitment (Diederich et al. 2005). C. gigas has advanced to become the dominant reef builder in the lower tidal zone since the early 2000 s (Nehls et al. 2006). Another introduced suspension feeder, the American slipper limpet Crepidula fornicata, present around Sylt since the 1930 s, has been limited by the occasional severe winters (Thieltges et al. 2004). These severe winters have become rare and the limpet population became as abundant as never before on shallow subtidal bottoms (Thieltges et al. 2009). With A. modestus we may now have a fourth alien species likely to have taken benefit from the recent trend in warming long after its initial introduction because Harms and Anger (1989) already showed that A. modestus is strongly negatively affected by severe winters in the area.
Since the first record in 1955, A. modestus has never been abundant and the two native barnacles S. balanoides and B. crenatus dominated in the intertidal and subtidal zones, respectively (Luther 1987, Buschbaum 2002a). After the severe winter 1995/96, local records suggest an exponential increase in population density. With our survey in 2007, it appears that we have just hit the year when A. modestus has become numerically dominant over the others in the intertidal zone. Revisiting some intertidal sites early in 2009 revealed an almost complete takeover of A. modestus on mussels and oysters with almost no S. balanoides left anymore (Reise, in prep.).
Increased temperatures during summer months may enhance the breeding success in A. modestus which is a cross-fertilising hermaphrodite with multiple broods throughout the warmer season, particularly with water temperatures above 20°C (Harms 1999, O’Riordan and Muphy 2000). Substantial settlement of juveniles takes place from June to October, and newly metamorphosed A. modestus grow rapidly and can reach maturity in about 8 weeks, while longevity is <2 years (Harms and Anger 1983). By contrast, the native S. balanoides is a boreal cold water species, with peak settlement already between April and May. It grows slowly and may get several years old. B. crenatus is also a boreal cold water species with settlement of larvae mainly from April to May (Luther 1987). This may explain the reverse dynamics of barnacle abundances observed in the course of our survey in summer 2007. Abundance of A. modestus increased, while numbers of S. balanoides and B. crenatus decreased.
Abundances of A. modestus and S. balanoides were highest in the low intertidal zone where mixed mussel and oyster beds occurred whereas B. crenatus was more abundant in the subtidal zone. A. modestus and S. balanoides are both capable of oxygen uptake from air and are tolerant to moderate desiccation stress, whereas the subtidal B. crenatus is not able to survive long periods of emergence (Foster 1971). In contrast to S. balanoides, A. modestus has the ability to settle in the upper subtidal as well (Harms and Anger 1983, Watson et al. 2005). Of the three barnacle species, A. modestus has the widest vertical range in the littoral zone, and this is reflected by the lowest coefficient of variation over all transect abundance data in this study. On wooden groins, A. modestus occurred higher up than S. balanoides. This agrees with observations at other shores (Foster 1971; Lawson et al. 2004). Further, we found no evidence of substrate specifity nor for a preferred surrounding by either oysters or mussels, and settlement occurred at the sheltered as well as the exposed shore. A. modestus seems to have a wider habitat niche than the native barnacles at Sylt. Overall, A. modestus is bound to warmer waters and there behaves more opportunistically than S. balanoides and B. crenatus (for review see Rainbow 1984). We know of no other environmental trend (i.e., no declining salinity, van Beusekom et al. 2008; no increasing food supply, van Beusekom et al. 2009) nor space for settlement) than temperature which could explain the observed boom in abundance of A. modestus in 2007. Differences in total density of all barnacle species at the study sites are typical for the area and may be caused by variable recruitment success and predation pressure.
We observed A. modestus frequently settling upon S. balanoides. Nevertheless, we do not assume effective competition for space as it has been suggested for coasts in southern Britain (Crisp 1958) because there is usually sufficient empty space still available on groins as well as on dead shells and living epibenthic molluscs at the sedimentary shores around Sylt. Barnacle diversity at the shore of Sylt is low compared to more southern locations as for example Lough Hyne on the south coast of Ireland. There three more species, Chthamalus montagui, C. stellatus and Verruca stroemia, are common (Lawson et al. 2004; Watson et al. 2005). Inside this Lough, A. modestus was still absent in the 1950s, was recorded at low numbers in the 1980 s, and finally attained dominance in the summer of 2001. Thus, even with five other species being abundant, no evidence for diversity resistance was found, nor for displacement of natives, and total barnacle cover has increased over the decades.
At Sylt we lack quantitative data on barnacles from earlier times but a very strong variability in cirripede larvae and settlement has been noted (Buschbaum 2002b). This, together with extreme patchiness of biogenic substrate for settlement, may preclude competitive displacement at the scale of a shore. S. balanoides is more likely to decline in range and abundance for climatic reasons (Wethey and Woodin 2008). With the alien A. modestus there is a potent substitute to fill the gap in case S. balanoides fades away because of warming.
However, it cannot be excluded that A. modestus may pre-empt space before Chthamalus spp.—now common at more southern shores of the North Sea—could expand their range into the eastern North Sea in the wake of climatic warming. Barnacles are easily shipped around the world or are transported with farmed oysters (Bishop 1951; Kerckhof and Cattrijsse 2001; Wolff and Reise 2002), and several alien barnacles seem to be ante portas to become established once the currently too cold waters warm up further in the North Sea (Kerckhof and Cattrijsse 2001; Kerckhof et al. 2007; Wiegemann 2008). The availability of artificial substrates, floating or fixed, may ease their immigration. Thus, there are good prospects that the low-diversity balanid fauna in the eastern North Sea will become invaded by more barnacles in addition to A. modestus.
Regardless whether aliens will have negative or positive effects on native barnacles or other biota in the recipient ecosystem, the addition contributes to the irreversible transformation of the inherited and co-evolved coastal biota into a more fortuitous and globalized assemblage where anything goes. Presumably this emerging assemblage will be better prepared for the changes to come than a pristine biota. However, the prize is a loss in regional identity and integrity.
We thank Elisabeth Herre for substantial help with the figures. Patrick Polte, Nils Volkenborn, Alfred Resch and Kay von Böhlen were essential to set-up the ring-experiment. Heike Büttger and Georg Nehls provided long-term data on barnacle abundance attached to beds of mussels and oysters.