The effects of climate change on reproduction and recruitment success of the acorn barnacle semibalanus balanoides at its southermost european distribution limit (Galicia, Spain)

Resumo

The intertidal zone is a model system for examining the effects of climate change both because of the rapidity of its response, and because of the rich historical record (Wethey and Woodin 2008). An intertidal rocky shore organism widely used for parameterization of mechanistic species distribution surveys is the acorn barnacle Semibalanus balanoides (Linnaeus, 1767) (Crustacea, Cirripedia). S. balanoides provides an excellent study organism for being a common, widely-distributed member of boreo-artic communities whose populations are easily manipulated in the field and for which there is a wide literature on historical distributions (Fischer-Piette 1936, Southward and Crisp 1954, Connell 1961, Southward et al. 1995, Hawkins et al. 2003), and physiology (Barnes 1963, Crisp and Patel 1969). Moreover, it has been long used as a model system for being sessile in their adult phase and easily tracked as they grow in the rocky intertidal (Southward et al. 1995, Hawkins et al. 2003, Blythe 2008), as well as for being sensitive to changes in temperature (Hutchins 1947). S. balanoides is an obligate cross-fertilizing hermaphrodite with internal fertilization (Crisp 1954), whose breeding occurs synchronously once a year (Barnes 1963). Gonads begin to develop during spring, although they are not conspicuous until summer when the penis develops and enlarges, reaching its maximum size just before the onset of fertilization in the late fall (Moore 1935, Barnes 1958, Crisp 1964). In the UK populations, copulation takes place from November to early December (Crisp 1964). After copulation the penis degenerates and grows again for the following reproductive season (Barnes and Stone 1972). Penis and gonad development in the population is highly synchronous, and their development has been related to low temperatures, as the ovary and penis do not develop at temperatures over 15°C (e.g. Barnes 1972). Fecundity of the species varies with age and location (Barnes 1989) and it has also been related to temperature variations (Barnes and Barnes 1976). Fertilized embryos are incubated in two egg sacs in the mantle cavity over winter, and nauplii larvae are released from the barnacle in the spring in synchronization with the spring algal bloom (Barnes 1956, Crisp 1964). The larvae then pass through six nauplius stages, before metamorphosing into the lecithotrophic cyprid stage, which will locate a suitable place to settle (Rainbow 1984). Settlement and recruitment of S. balanoides have been strongly positively correlated with cold winters (Southward 1991, Jenkins et al. 2000, Drévès 2001, Abernot-Le Gac et al. 2018). In North America, S. balanoides has been recorded in the Pacific coast from the Bering Sea (Alaska) as far south as British Columbia, Canada (Carroll and Wethey 1990), and in the Atlantic coast from the Labrador Sea as far south as Cape Hatteras, USA (Wells et al. 1960, Crickenberger and Wethey 2018a). In Europe, S. balanoides is found from Greenland and the Svalbard Islands to the middle of the French coast on the Bay of Biscay (Crisp and Fischer-Piette 1959, Feyling-Hanssen 1953, Høpner Petersen 1966, Stubbings 1975). However, surprisingly, there is a southern isolated disjunct population in Galicia, in the NW Iberian coast (Fischer-Piette and Prenant 1956, Fischer-Piette 1963, Macho 2006, Macho et al. 2010, Wethey et al. 2011b, Herrera et al. 2019). Many studies have related the southern range limit regression of the species in the Atlantic coast with warming temperatures both in USA (Jones et al. 2012, Crickenberger and Wethey 2018b) and Europe (Southward 1991, Wethey and Woodin 2008, Wethey et al. 2011b, Mieszkowska et al. 2014). In Europe, S. balanoides has been contracting northward over the past century and has suffered reduced abundance during the long term warming (Wethey et al. 2011b). Abundance and distribution records suggest a distribution range reduction in the NW Iberian population. The observations made by Fischer-Piette and Prenant (1957), Ardré et al. (1958), Fischer-Piette (1963), Fischer-Piette and Seoane-Camba (1963), Barnes and Barnes (1966) and Wethey and Woodin (2008), suggest that S. balanoides in NW Iberia was more abundant and more widely distributed along the region in the past, going south as far as Viana do Castelo (Portugal) during the 50s, but disappearing from Portugal in the last survey done in 2006 (Wethey and Woodin 2008), and as far as Ortigueira (A Coruña) in Northern Galicia. Wethey and Woodin (2008) suggested that the species has contracted from reproductive populations throughout the NW Iberian Peninsula in the 1950s and 1960s (Ardré et al. 1958, Fischer-Piette and Seoane-Camba 1963, Barnes and Barnes 1966) to a single reproductive population in the Ría de Arousa (Galicia, Spain), which represents a compression of the reproductive range of the species of 50 km from the south to 250 km from the north since the 1960s-1990s (Wethey and Woodin 2008). Wethey and Woodin (2008) results were in agreement with their Winter Cold Limitation of Reproduction Hypothesis, which states that changes in the southern limit of S. balanoides are caused by intolerance of winter body temperatures above 10°C, leading to reproductive failure. Therefore, they suggested that contraction in distribution of S. balanoides in NW Iberia has been caused by reproductive failure at temperatures above 10-12°C, which is the threshold proposed as crucial for fertilization success based on northern European populations (Barnes 1963, Crisp and Patel 1969, Crickenberger and Wethey 2018b). Although such low seawater temperatures are rarely met in Galicia, S. balanoides presence in the Galician rias might be mainly explained by the upwelling system in the region (Fraga 1981), which is a major driver of ecological processes due to the fine-scale temperature distribution. Galician coast is colder than both Portugal towards the south and Cantabrian coast of Spain towards the northeast (Alvarez et al. 2005b). It has been suggested that populations near the species distribution limit are likely to be more sensitive to climate change (Alcock 2003, Svensson et al. 2005). Because of that, regions where the range limits of the species are found close together are relevant places for the assessment of climate change effects (Southward et al. 1995). Given that the southernmost European distribution limit of S. balanoides in Galicia meets both assumptions, there is a particular interest in understanding the mechanisms that govern its distribution in the area and the effect that climate change can produce on their populations, as a model to understand the effect of global change on population dynamics of intertidal communities. So, the overall aims of this PhD dissertation is to determine the mechanisms governing the effects of climate change on reproduction, settlement and recruitment of S. balanoides, as a model organism of rocky intertidal communities, and its biogeographic boundaries in the northwest of the Iberian Peninsula. According to this, the main working hypothesis is that reproductive cycle, fertilization, settlement and recruitment of S. balanoides in Galicia will be higher in colder locations and years, in agreement with the Winter Cold Limitation of Reproduction Hypothesis (Wethey and Woodin 2008). We used a combination of laboratory experiments and field observations in the Rías Baixas to test the hypothesis that high temperatures inhibit the S. balanoides penis development, becoming a limiting factor for fertilization and therefore, for reproductive success at its southernmost distribution limit (Chapter 2). We used four experimental temperature treatments in the laboratory experiment (14°C, 17°C, 20°C and 23°C). In the field, we selected two locations in the Ría de Arousa (A1 and A2), where temperatures were expected to be lower, and also two locations in the Ría de Vigo (V1 and V2), where temperatures were expected to be warmer. Results from the laboratory and the field were used to estimate fertilization probabilities at different temperatures and population densities. High temperatures inhibit penis development of S. balanoides as determined in the laboratory experiments, and therefore it diminishes the fertilization probabilities of the species, as confirmed in the field. Penis development was completed in barnacles held at temperatures of 17°C or lower. The mean length of the fully developed penises (relaxed) was far greater at 14 and 17°C than at temperatures of 20 and 23°C. Barnacles maintained at 14 and 17°C showed a number of annulations corresponding to field observations in the UK of fully developed penises immediately before copulation (Barnes and Stone 1972, Barnes 1992), while the animals from warmer temperature treatments (20 and 23°C) in the laboratory and from the Ría de Arousa showed a lower number of annulations. These results suggest that the S. balanoides population in the Ría de Arousa does not show a fully developed penis, in terms of length and number of annulations, as observed on other populations in the middle of its geographic range. Results suggested that penis length in the Ría de Vigo population is shorter than in the Ría de Arousa. Fertilization probability was higher in the Ría de Arousa (up to approximately 95%), where temperatures are 1-2°C lower than in the Ría de Vigo, where fertilization probabilities do not exceed 44%. The higher fertilization probability in A1 location in the Ría de Arousa was explained by the population density which is higher than it is in the A2 location. Fertilization in the warmer ria (Ría de Vigo) will be negatively affected not only by low population densities, but also by shorter penis lengths caused by exposure to higher temperatures. Even in the Ría de Arousa, if temperatures continue rising in a warming scenario, the reproductive success can be diminished up to approximately 42%. The reproductive cycle of S. balanoides at its southernmost European distribution limit in Galicia (NW Iberian Peninsula) was studied in two locations in the Ría de Arousa (A1 and A2) with different temperatures during a four-year time series (2012-2016) (Chapter 3). The results allowed us to affirm that the species is capable of reproducing in Galicia, even though the sea temperature conditions proposed for northern European populations as necessary for the gonad maturation (less than 15ºC) (Rognstad and Hilbish 2014) and fertilization (less than 10ºC) (Barnes 1963) are not met. In addition, reproductive output was significantly greater in the colder location (A2) than in the warmer one (A1), indicating that even slightly differences in temperature between two locations separated by less than 10 km, as in our study, can result in a significant difference in reproductive output. Our results indicated that S. balanoides reproductive output in both locations was higher when upwelling was more favorable, as during the year 2014, when low downwelling was recorded, and therefore cold water was retained longer inside of the rias, suggesting that upwelling regime of the Rías Baixas is probably the main factor driving S. balanoides presence in this southernmost limit. Despite of being at its southernmost distribution limit, S. balanoides population in Ría de Arousa shows fecundity values comparable to those at its middle distribution range (Barnes and Barnes 1968). The only report in the literature of the fecundity of the Galician population was made by Barnes and Barnes (1968), who estimated 7160 embryos for an individual of 1.5 mg body dry weight in the Ría de Muros, 25 km northward the Ría de Arousa, where few individuals can be found nowadays, we found a similar number of embryos for an individual of 1.5 mg body dry weight in A1 and a significantly higher number in A2. In laboratory experiments, Rognstad and Hilbish (2014) found considerably less reproductive output between barnacles held at 13°C and lower temperatures (7 and 10°C) during the embryo development period; however, although 13°lies in the range observed in our study area, fecundity values in Galicia are similar to those of the lowest temperature. The cyclic pattern observed in S. balanoides in Galicia is in agreement with those described elsewhere (Bousfield 1954, Høpner Petersen 1966, Barnes 1989). Fluctuations in the weight of ovary/ovisac and somatic tissue were related to seasonal availability of food. The onset of breeding for S. balanoides in the Ría de Arousa is similar to that described in northernmost European latitudes as in southwest of England, during late November-early December (Crisp 1959). Our results were in accordance to the reported fertilization date by Barnes and Barnes (1976) who indicated it occurred around November, 22 in the Rías Baixas. Crisp (1959) pointed out that the onset of fertilization results from a series of changes beginning some months earlier, which are associated with a fall in metabolism rate with decreasing temperatures and influenced by environmental factors; our results indicate that a higher production of embryos is almost 100% explained by a combination of environmental conditions (upwelling, food availability, suspended sediments and salinity) exactly three months previous to the fertilization event, supporting the idea that environmental factors that determine fertilization act over a large period of time. As described elsewhere (Barnes et al. 1963, Crisp 1964, Barnes 1989), in Galicia, larvae are released coinciding with the spring algal bloom. Significant correlation between suspended sediment and hatching period was found, while Chl-a concentration did not have a significant effect even though it helped explain the variance. However, it is unclear what the effect of rising temperatures due to climate change might be on the embryo development process since laboratory experiments with S. balanoides from Galicia have shown that embryo development rate can be accelerated by high temperatures (Herrera unpubl. data), causing a potential mismatch between larvae release and phytoplankton bloom, as Crickenberger and Wethey (2018b) suggested for the southern range limit in Galicia by using biogeographic models. Such a mismatch would have potentially detrimental consequences for this southernmost population, as could be observed on the settlement and recruitment pattern study in Chapter 4. Although temperatures in Galicia are higher than they are in the middle of the S. balanoides distribution range, the reproductive output of the species seems comparable to that found further north, suggesting the adaptation of the species to local environmental conditions and probably the existence of natural selection on individuals with better reproductive performance. Settlement and recruitment are the keystone for persistence of populations. In sessile marine species like barnacles, recruitment is influenced by the oceanographic factors at which planktonic larvae are subjected. Larvae can be transported over wide geographic regions, far away from their origin location, therefore, both local and regional scales need to be considered. We studied settlement and recruitment patterns of S. balanoides in the NW of the Iberian Peninsula, its southernmost European distribution limit, using a combination of regional and local scale surveys (Chapter 4). We tested several alternative hypotheses regarding the mechanisms that may control the geographic pattern, and temporal variations in intensity of recruitment. At regional scale, we assessed abundance and distribution of adults and recruits along the NW Iberian coast, in 30 locations, from the Ría de Foz (Galicia) to Porto (Portugal), in two consecutive years (2015 and 2016). At local scale, we studied the settlement patterns of S. balanoides in the Ría de Arousa (locations A1 and A2) and Ría de Vigo (locations V1 and V2) in the same years. Results from local and regional scale observations were analyzed in relation to environmental factors. Adult densities were very variable along the region; the highest densities were recorded in the Rías Baixas, where there were also shorter nearest neighbor distances (NNDs) among individuals. Thus, only 6 out of 30 sampled locations along the NW Iberian coast were likely larval sources since they had a mean NND less than the 3 cm, the maximum penis length (Herrera et al. 2019). This leads us to conclude that the Rías Baixas, in center of the local geographic range, is the larval source, contributing to regional-scale recruitment and to the maintenance of less dense sink populations in regions A (to the north) and C (to the south), within the dispersal range of larvae (ca. 50 km; Crisp 1958, Southward 1967, Crickenberger et al. 2017, Crickenberger and Wethey 2018b). Consistent with this conclusion, low to null recruitment was found outside the Rías Baixas in localities with no adults or where those adults were so sparse that mating is not possible. Within the center of the geographic range (Rías Baixas), year to year variation in recruitment was consistent with temperature variations. We found higher recruitment rates at regional scale in the year 2015 which was colder during January-February when embryonic development occurs, consistent with the idea that low SSTs are required for high recruitment success (Southward 1991, Drévès 2001, Wethey et al. 2011b, Abernot-Le Gac et al. 2018). Thus, temperature conditions seem to have been determinant not only to the reproductive success but also to the recruitment success, similar to the temperature-dependent relationship proposed by other authors (Southward 1967, 1991, Drévès 2001, Wethey et al. 2011b, Rognstad et al. 2014, Rognstad and Hilbish 2014, Abernot-Le Gac et al. 2018, Crickenberger and Wethey 2018a, b). Local scale assessment of settlement patterns, showed that settlement intensity in both years was highly variable in time and space in both rias, which is consistent with the variability described for this species at northern latitudes (e.g. Hawkins and Hartnoll 1982, Kendall et al. 1985, Wethey 1985, Jenkins et al. 2000). Settlement in the warmer ria (Ría de Vigo) was lower than in the colder one (Ría de Arousa). Nonetheless, the observed settlement rates (mean of 0.02 settlers cm-2) were three orders of magnitude less than settlement rates (mean of 80 settlers cm-2) recorded in S. balanoides optimum distribution areas in northern Europe (Connell 1961) and North America (Wethey 1985, Minchinton and Scheibling 1991, Bertness et al. 1996). Food supply may also influence larval recruitment success, especially when there could be a mismatch between phytoplankton abundance and larval release as it happens in S. balanoides (e.g. Barnes 1956). Our results regarding the difference in timing between sufficient chlorophyll-a concentration and Nauplius Larval Development (NLD) suggested that food supply may have an important role on the differences found between 2015 and 2016. The higher settlement and recruitment rates of 2015 coincide with the coupling between NLD and a sufficient food supply period (which was not the case in 2016), necessary for a successful larval development (Barnes 1956). The timing of settlement suggested that larvae are transported far away from their origin location. In 2015, settlement period in both locations (A1 and A2) of the colder ria started at the same time when larvae release and NLD occurred in those same locations, therefore it is impossible that larvae that settled at that precise moment originated in those same locations, since once that nauplius larvae were released they required at least 13 days to develop into the settling form of cyprid, given that SST during that period was 12°C (relationship from Harms 1984). Larval release can vary up to a month among subpopulations at broad scale (Kendall et al. 1985), so it is likely that such larvae came from other source location where hatching took place sooner, probably due to higher local temperatures accelerating embryo development (Kendall et al. 1985). In 2016, larvae release in locations A1 and A2 occurred previous to the beginning of settlement which makes plausible that larvae settling two weeks after were produced in those same locations. The results were generally consistent with the ocean currents hypothesis, which proposes that ocean currents have a dominant influence on larval recruitment at local and regional scales (Pineda et al. 2007, Dubois et al. 2007, Broitman et al. 2008, Rognstad et al. 2014, 2018, Crickenberger and Wethey 2018b). Ocean current flow paths on weekly scales were in directions that connect sites where adult reproduction is favored to sites with low adult densities where we observed recruitment, suggesting that the regional oceanic current system plays an important role on large-scale variability in recruitment of larvae and, therefore, on the recolonization of the species when winter temperature conditions are favorable. Transport of larvae to the shore has been related to wind direction (Bennell 1981, Hawkins and Hartnoll 1982, Kendall et al. 1985, Bertness et al. 1996), since high settlement rates are expected from onshore winds. We did not find evidence of a dominant effect of wind on settlement variation since less than 1% of the variance in settlement was explained by wind direction. Variations in the intensity of upwelling can control the concentration of larvae close to the coast, since upwelling should move larvae offshore and downwelling should bring them close to shore (Roughgarden et al. 1988, Shkedy and Roughgarden 1997, Connolly et al. 2001, Iles et al. 2011). Our results did not support this hypothesis, as we did not find a significant relationship between the upwelling index during settlement period and settlement rates in any year. Other oceanographic processes that have not been considered in the present study, like tidal bores, internal waves or local scale hydrodynamic differences might be responsible for onshore transport of larvae (Pineda 2000, Shanks et al. 2010). The present Ph.D. dissertation contributes to the understanding of the main driving mechanisms influencing S. balanoides population dynamics at its southernmost European distribution limit in Galicia. The obtained results have allowed us to conclude that temperature is the main factor determining the population performance of S. balanoides in Galicia, which is derived from the upwelling regime in the region. S. balanoides distribution in Galicia is mainly explained under the Winter Cold Limitation of Reproduction Hypothesis, since most of the limitations for reproduction and recruitment success derive from winter temperature conditions. The temperature threshold for S. balanoides reproductive success in Galicia is 14-17°C, rather than 10-12°C as proposed for further north populations. The Rías Baixas are likely acting like a refuge for S. balanoides at its southernmost European distribution limit, and most likely for other boreal species.