Current predictions show that on the Atlantic coast of Europe there will be “very likely” more frequent heat waves, longer in duration and greater in intensity, as well as increases in the intensity and frequency of extreme precipitations that will modify coastal salinity (Christensen et al., 2007; Trenberth, 2012; IPCC, 2014; Jacob et al., 2014; Hoegh-Guldberg et al., 2018; Viceto et al., 2019; Carvalho et al., 2021; but see Cardoso Pereira et al., 2020; Lorenzo and Álvarez, 2020). In fact, from 1982 to 2010 almost half of the world’s coastlines had already experienced a significant decrease in the frequency of extremely cold days and more than a third had faced an increase in extremely hot days (Lima and Wethey, 2012).

These acute events are important in shaping communities and their effects can be widespread and long lasting provoking important biogeographical changes (e.g., Daufresne et al., 2007; Mulholland et al., 2009; Byrnes et al., 2011; Wethey et al., 2011; Frölicher and Laufkötter, 2018; Shanks et al., 2020). Temperature affects primarily physiology and metabolism of ectotherms resulting in changes in energy allocation among metabolism, maintenance, growth, reproduction, and protective responses such as heat shock proteins or antioxidant pathways, which require energy (Pörtner and Farrell, 2008; Eymann et al., 2020). Likewise, changes in salinity produce energetic demands of osmoregulation with allocation of energy in growth and cellular maintenance at the expense of reproduction (Brett, 1979).

The reproductive cycle of bivalves is controlled by the interaction between endogenous and environmental factors (Normand et al., 2008). Among those environmental factors, temperature is a forcing variable for the gametogenic cycle in bivalves (e.g., Sastry, 1966; Xie and Burnell, 1994; Albentosa et al., 2007) defining both the starting point and the rate of gonadal development, whereas food influences the duration of the reproductive cycle (Lubet, 1959; Ruíz et al., 1992) and the percentage of mature individuals (Delgado and Pérez-Camacho, 2007). It has been demonstrated that under increased stresses, such as those predicted from climate change scenarios, bivalves may allocate energy away from growth and reproduction toward costly physiological defenses in order to survive (Petes et al., 2008). And the potential for fecundity loss is one of the most likely consequences of that scenario. For instance, elevated temperature reduces gamete development in Mytilus galloprovincialis (Múgica et al., 2015) and fertilization success in Tridacna maxima (Armstrong et al., 2020); abrupt changes in salinity affect spawning of Ruditapes philippinarum (Dang et al., 2010). Thus, the very energetically costly reproductive processes (Gosling, 2003) may be impaired or suppressed (Wingfield and Sapolsky, 2003) with enormous negative consequences for the structure and dynamics of populations of coastal organisms. Failures in recruitment of juveniles will prevent replenishment of shellfish beds (Beukema and Dekker, 2005; Petes et al., 2007; Talmage and Gobler, 2011), and, ultimately, the decline or local extinction of populations (Parmesan, 2006; Berg et al., 2010).

The impacts of climate change on the structure and functioning of coastal marine ecosystems, especially estuaries, are a major focus of concern (Ringwood and Keppler, 2002; Harley et al., 2006; Grilo et al., 2011) even more when the affected species support important fisheries (La Peyre et al., 2009; Möller et al., 2009; Parada et al., 2012; Peteiro et al., 2018; Domínguez et al., 2020, 2021; Woodin et al., 2020).

Small-scale fisheries (SSFs, Small-Scale and Spatially Structured Fisheries targeting sedentary resources with artisanal gears, sensu Orensanz et al., 2005) are of enormous importance to both harvest and employment (Mills et al., 2011). This is particularly true in Galicia (NW of Spain) where fisheries in the intertidal and shallow subtidal of the native clams Ruditapes decussatus (Linnaeus, 1758) (grooved carpet shell) and Venerupis corrugata (Gmelin, 1791) (pullet carpet shell), the introduced species R. philippinarum (Adams and Reeve, 1850) (Manila clam), and the cockle Cerastoderma edule (Linnaeus, 1758) contribute an annual value of ∼74 million € and involve ∼7,100 fishers.¹

These four bivalves are dioecious broadcast spawners without sexual dimorphism, each displaying a different reproductive cycle. In Galicia (Spain) the reproductive cycle of R. decussatus starts in January with the initiation of gametogenesis although the proliferation of gametes occurs during spring, followed by spawning in summer (June–August) (Rodríguez-Moscoso, 2000; Ojea et al., 2004). This species has a great capacity for gonadal regeneration thus allowing multiple spawnings per season (Matias et al., 2013). Gametogenic reabsorption after spawning seems to be unimportant (Rodríguez-Moscoso, 2000) with only some degenerate oocytes (cytolysis) and no major hemocyte infiltration. Resting phase happens in November–December with the presence of reserve tissue formed by interfollicular muscle and vesicular cells (polygonal cells with an eccentric nucleus) inside the follicles, which disappear as the gonadal development advances (Delgado and Pérez-Camacho, 2007).

Ruditapes philippinarum, native to the Pacific, was introduced for culture in Europe in the 1970s because of its high adaptability (Latrouite and Claude, 1976; Pérez-Camacho and Cuña, 1985; Drummond et al., 2006). Nowadays it is one of the most commercially exploited molluscs in the world. It has a long reproductive period and its gonadal development is quite asynchronous within the gonad of each individual and between individuals. It is in sexual resting from October until January, when the gametogenesis starts, reaching advanced gametogenesis in March. In April the gametes are mature and a period of successive spawning interspersed with gonad recovery lasts from April to September (Rodríguez-Moscoso et al., 1996; Delgado and Pérez-Camacho, 2007).

In Galicia V. corrugata shows a long reproductive cycle (Villalba et al., 1993) with ripe and spawning stages observed throughout the year, although with larger number of ripe follicles between February and June (Cerviño-Otero et al., 2007). During autumn the gonad develops asynchronously, showing simultaneously in the same gonad mature and early gametogenic follicles. Gametogenesis occurs between February and March, reaching maturity in March although spawning starts in May continuing until July. In August and September a new maturation period begins with sporadic spawning in autumn. Sexual rest is very occasional since as soon as residual gametes appear, vesicle cells are formed and new germinal lines develop (Cerviño-Otero, 2011).

The onset of gametogenesis in the cockle C. edule in Galicia takes place at the end of the summer (September to October), progresses throughout the winter, and reaches maturation in the spring. The first spawning occurs in April and May and, after gonad restoration, another spawning episode takes place in May and June. During July and August, most of the population shows signs of gonad resorption, although a less-intensive spawning event can occur at the end of summer/beginning of autumn (Martínez-Castro and Vázquez, 2012).

The physiological responses to temperature and salinity fluctuations of these four bivalves have been recently investigated by our research team in mesocosm experiments in which field conditions were simulated (Macho et al., 2016; Peteiro et al., 2018; Domínguez et al., 2020). Early recruits of C. edule showed a lethal limit at salinity of ∼15 and a marked reduction of activity (i.e., continuous valve closure and inhibition of respiration and ammonium excretion rates) at salinities between 15 and 30 (Peteiro et al., 2018). After 6 days of exposure to different salinity fluctuations, adults of C. edule, R. decussatus, R. philippinarum, and V. corrugata showed a dramatic reduction of pumping activity, scope for growth (SFG) and burrowing activity at the lowest salinities (i.e., 5, 10, and 15), but responses under higher salinity (20, 25, and 30) varied among species and across seasonal periods (Domínguez et al., 2020). While C. edule was the most affected species in autumn showing no fast recovery after the end of the stress episode, R. decussatus was more resistant in all seasons (Domínguez et al., 2020).

Similarly, temperature stress experiment performed for these species, agree to identify acute responses to short heatwaves (Domínguez et al., 2021). After 2 days of exposure to sediment heatwave during low tide, C. edule and V. corrugata suffered significant mortalities and also a dramatic decrease in SFG as well as reduction in burrowing activity compared to R. decussatus and R. philipinnarum (Domínguez et al., 2021).

While the effects of acute environmental stress on the SFG and behavior (see above), or early-life history stages (e.g., Talmage and Gobler, 2011) in bivalves have been documented, the effect on the reproductive performance in the context of climate stressors has not yet been investigated (but see Xu et al., 2016) even though the reproductive response of clams and cockles under future climate change scenarios is crucial for evaluating population dynamics (Morgan et al., 2013; Xu et al., 2016).

Thus, the aim of the present study was to experimentally evaluate the potential effect of short episodes of low salinity as well as sediment heatwaves on the reproductive cycle of these bivalves. Both low salinity and temperature stress treatments consisted of four ramps that followed similar profiles to those experienced by bivalves in the field during tidal cycles. Consequently, a low salinity stress experiment was repeated in autumn, winter and spring when episodes of heavy rains occur in the Atlantic coast of Europe while a heatwave experiment was performed during summer. Considering the available physiological data from previous studies, it is reasonable to hypothesize that each of these four species may respond differently to increasing environmental stress, potentially impairing or suppressing reproductive success. Also, we expect a more acute effect during gametogenesis and spawning than during the resting period of the gametogenic cycle.

Four experiments were performed in a mesocosm system at Estación de Ciencias Mariñas de Toralla (ECIMAT)² of the Universidade de Vigo (Galicia, NW Spain). Low salinity stress experiments were independently run in December 2015, March and May 2016; the heatwave stress experiment was done in July 2016.

During the four experiments, a total of 1,605 histological slides out of 2,112 were finally used since parasitized gonads or those with very little gonadal tissue were not considered: 438 R. decussatus (170 undifferentiated, 142 females, 126 males), 395 R. philippinarum (122 undifferentiated, 156 females, 117 males), 394 V. corrugata (3 undifferentiated, 167 females, 224 males), and 378 C. edule (36 undifferentiated, 173 females, 169 males). The mean lengths (±SD) of these bivalves were 43.39 ± 1.99 mm for R. decussatus; 42.04 ± 1.37 mm for R. philippinarum; 39.27 ± 2.28 mm for V. corrugata, and 30.65 ± 1.37 mm for C. edule.

No mortality was recorded during the salinity stress experiments. During the heatwave experiments, mortality of R. decussatus was less than 4% in all treatments. R. philippinarum only suffered mortality of 2% at temperatures <37°C but 30% at 37°C treatment. Mortality of V. corrugata was 12% at temperatures <37°C and 40% at 37°C treatment. The greatest mortality was experienced by C. edule with more than 75% at 37°C and 12% at temperatures <37°C.

Successful reproduction is a key factor for the survival of species. Reproduction itself implies physiological stress (Domouhtsidou and Dimitriadis, 2001; Garmendia et al., 2010) so any additional exogenous stress during the reproductive period will be dramatic since, to ensure survival, animals have to allocate a large amount of energy to support maintenance costs (Wingfield and Sapolsky, 2003), thus limiting the energy available for growth, storage, and reproduction (Petes et al., 2008; Kooijman, 2010; Stumpp et al., 2012). The effects of stress on reproduction are very complex, varying with time of the year, food supply or height on the shore in the case of intertidal organisms (Petes et al., 2007, 2008). Our results support those statements because both low salinity and heatwave stress did have important effects on reproduction of the four studied bivalves. Furthermore, the response was species-specific as well as varied with the time of the year, therefore with the annual gametogenic cycle, being more acute in March and May, at the peak of the reproductive season.

During sexual resting or early stages of gametogenesis, in the December experiment, a delay in the gametogenic development was observed in the clams R. decussatus, R. philippinarum, and Venerupis corrugata at salinities lower than 15 and at salinities lower than 10 in the cockle Cerastoderma edule. This delay lasted three days after the end of the stress only for R. decussatus. Similarly, delayed gametogenesis has been reported in bivalves subjected to thermal (Fearman and Moltschaniwskyj, 2010) and pollution stress (Gagné et al., 2002; Gauthier-Clerc et al., 2002). It is also known that seawater acidification delays reproductive processes of R. decussatus (Range et al., 2011). Surprisingly SFG of this species did not change with the lowest salinity (Domínguez et al., 2020).

Variations of salinity between 5 to 20 and 10 to 25 in March and May, during the peak of the reproductive season, triggered massive spawning followed by gonadal recovery in C. edule, a similar response to that produced by acidification in R. philippinarum during gonadal maturation (Xu et al., 2016). But early spawning of C. edule under stress could lead to a mismatch between the presence of their planktotrophic larvae and the phytoplankton bloom that serves as a food supply, causing potentially starved larvae (Philippart et al., 2003) and thus reducing recruitment success (Beukema et al., 2009).

Another observed response in these experiments was the change in the morphology of the oocytes of R. decussatus in March and May and of V. corrugata and C. edule in May. Oocytes are quite vulnerable to changes in salinity (Gallo et al., 2020) and changes in the morphology could be attributed to changes in the osmotic pressure. Structural changes in oocytes and breakdown processes related to osmotic stress are common in invertebrates (Ram et al., 2004; Simmonds and Barber, 2016). Marked drops in salinity should be followed by an influx of water, which may harm an oocyte unless it can tolerate such changes in volume or, alternatively, can osmoregulate and, thus, prevent marked changes in volume. Whether or not oocytes of these bivalves can osmoregulate, is a question that remains unanswered.

Resorption of gametes with a severe haemocytic infiltration and formation of vesicle cells was the common response of R. philippinarum and V. corrugata at the lowest salinities in March and May. Despite having the same effect on reproduction, only SFG of V. corrugata drastically diminished by low salinities (Domínguez et al., 2020). Even without stress, some species like R. philippinarum present a considerable number of individuals in gonad resorption during the phase of sexual resting and throughout the process of gonadal maturation that evidences a high capacity for gametogenic regeneration and ability to recover the energy invested in the production (Delgado and Pérez-Camacho, 2007). However, stress also causes resorption. To avoid osmotic stress bivalves frequently close their valves, thereby losing between 1 and 24% of the energy acquired (Gosling, 2015). Specifically, these clams do it at salinity 15 and below (Carregosa et al., 2014; Verdelhos et al., 2015; Domínguez et al., 2020) and the cockle at salinities lower than 10, reducing their SFG (Verdelhos et al., 2015; Domínguez et al., 2020) mainly due to a reduction in the feeding activity (Domínguez et al., 2020). Consequently, their reproductive response to stress, both during early gametogenesis with a delay in the gonadal development and during advanced gametogenesis with resorption of gametes, could be attributed to an energetic limitation due to a cessation of feeding since continuous and adequate energy supply is required for gonad ripening (Beninger and Lucas, 1984; Gosling, 2003; Drummond et al., 2006). For instance, mussels experiencing food shortage are not able to maintain ripe gametes in the follicles (Bayne and Thompson, 1970) and seawater acidification forces them to reabsorb the gametes as an energy saving (Bibby et al., 2008).

Our experimental but realistic atmospheric heat waves also triggered gonadal resorption in V. corrugata, R. decussatus, and C. edule which was almost total at the highest temperature. This response was expected for V. corrugata and C. edule, since after two days of exposure to the same heatwaves, both species suffered a dramatic decrease in their SFG as well as reduction in burrowing activity. On the contrary no effect was expected for R. decussatus since they live burrowed at 13 cm depth within the sediment and depth in sediment buffers the temperature actually experienced (e.g., Johnson, 1965; Harrison and Phizaclea, 1987; Domínguez et al., 2021). The morphology of their oocytes was also modified although such modification suggests other mechanisms rather than osmotic pressure. Several studies found decreasing quality of oocytes in vertebrates and invertebrates with increasing temperature, affecting mainly processes during maturation (Múgica et al., 2015; Gallo et al., 2020), with different underlying mechanisms, e.g., oxidative stress, DNA methylation (Gallo et al., 2020). Contrastingly, R. philippinarum was not affected by temperature treatments other than a tendency toward a larger proportion of gonads in resorption at the highest temperature. Mostly, higher temperature of sea water results in initiation and acceleration of reproductive processes (Lubet, 1959), hence, raising temperature is a common practice in aquaculture to bring bivalves into spawning condition (e.g., Sastry, 1966; Helm and Bourne, 2004; Delgado and Pérez-Camacho, 2007). The most common response of bivalves to thermal stress is the complete release of gametes (Schreck et al., 2001; Petes et al., 2007, 2008) in order to reallocate the energy from reproduction toward defense and repair mechanisms (Múgica et al., 2015) although these spawned gametes can be of poorer quality (Martínez et al., 2000; Ojea et al., 2008). Even more, gonads can remain spawned-out, not able to restore the gonad after spawning, probably due to a lack of energy to regenerate gametes (Petes et al., 2008) caused by drastic reduction of SFG (Domínguez et al., 2021). Alternatively elevated temperatures can restrain reproduction (Heasman et al., 1996; Jeffs et al., 2002), or slow down gametogenesis by the increased demands of metabolism that limits energy for reproduction as described for Mytilus galloprovincialis (Fearman and Moltschaniwskyj, 2010).

Energy balance is fundamental in environmental stress adaptation and tolerance in marine organisms (Pörtner and Farrell, 2008; Sokolova et al., 2012). As discussed, the gonadal cycle of these species may be affected by changes in SFG (Domínguez et al., 2020, 2021). Gametogenesis is an energy demanding process that results in changes in vulnerability, reducing allocation of energy to grow and burrow (Joaquim et al., 2011). In a previous study we calculated that during high tide in a Galician fishing bed, seawater salinity was ≤15 during 9.5% of the time in autumn and during 21 and 23% of the time in winter and spring, respectively (Domínguez et al., 2020). This yields to a loss of >20% of the original SFG value of these four studied species due to exposures to salinities ≤15, that can be even worse in spring when SFG is already low even without stress. Since the probability of being in energy deficit is very high during spring, that could contribute to a slowdown of reproduction in all species but R. philippinarum in March, and in May with dramatic, not only ecological, but also economic consequences. Bivalve fisheries already experience high spatial and temporal variability in catches that are related to high mortality due to strong fluctuations in environmental conditions such as temperature and salinity (Juanes et al., 2012; Parada et al., 2012; Morgan et al., 2013; Aranguren et al., 2014). This fact is aggravated when reductions in reproductive output could lead to a recruitment failure into adult populations (Shanks et al., 2020) especially given the extremely high mortality (>95%) during pelagic larval development and benthic recruitment (e.g., Rumrill, 1990; Gosselin and Qian, 1997). For short-lived species as these bivalves if low salinity episodes in winter and spring are followed by a heat wave in summer, reproduction may be compromised, and the impact magnified if this situation repeats during consecutive years. Long term consequences in the recruitment of these commercially important species are of crucial importance for the sustainability of the fishery.

Negative impacts of the introduced R. philippinarum on other native bivalves, mainly R. decussatus, such as decrease of densities and changes in the distribution, have been claimed as the reason for the decrease of the native clam populations along the European coast (Pranovi et al., 2006; ICES, 2008). However, interspecific competition with other bivalves (Lee, 1996; Byers, 2005), specifically with R. decussatus (Juanes et al., 2012; Bidegain and Juanes, 2013), for space or resource has not yet been clearly demonstrated. According to these last authors, low salinity events, but not competition, modulate the distribution of R. philippinarum and R. decussatus, with R. decussatus being in more freshwater influenced areas since floods may have a stronger effect on the mortality of R. phillippinarum. However, our results point toward a reproductive failure of R. decussatus over time since under high temperature and low salinity stress, reproductive success of R. philippinarum is higher than R. decussatus, mainly during heat waves in summer when R. decussatus is spawning. This could lead to a failure in the recruitment that, sustained over time, and together with overfishing (Massapina and Arrobas, 1991; Maia et al., 2006) could be the reason for the decrease of the populations of this clam. Consequently, the introduced Manila clam with faster maturation, greater number of spawning events, longer and greater reproductive activity, faster growth (Laruelle et al., 1994; Delgado and Pérez-Camacho, 2007; Moura et al., 2017, 2018) and greater tolerance to salinity and temperature stress (Domínguez et al., 2020, 2021) contribute to the observed larger abundance of the introduced clam. Furthermore, R. philippinarum, but not R. decussatus, is typically seeded in shellfish beds (e.g., Royo et al., 2002; Melià and Gatto, 2005; Mantovani et al., 2006) artificially increasing its abundance. In fact, R. philippinarum has become a natural population and one of the most commercially exploited bivalves along the Atlantic and Mediterranean coast of Europe (FAO, 2020) due to the drastic declines of the native carpet shell clam. Currently in Galicia catches of R. philippinarum are 10 times higher than those of the native R. decussatus and five times higher than 20 years ago³ (Figure 9). If current predictions for the Atlantic coast of Europe, i.e., more frequent heat waves, longer in duration and greater in intensity, as well as increases in the intensity and frequency of extreme precipitation, hold true, R. philippinarum may replace the native R. decussatus in Galicia as has already happened in Arcachon Bay in France and the lagoon of Venice in Italy (Marin et al., 2003; Caill-Milly et al., 2006). The ecological and socioeconomic consequences need to be analyzed to develop adequate management and conservation strategies. This new information is useful in order to promote government intervention to manage these resources sustainably in a context in which the frequency and intensity of extreme events will increase.

The datasets presented in this article are not readily available because dataset are still used in ongoing research. Requests to access the datasets should be directed to EV, eotero@uvigo.es.

EV analyzed the histological slides. EV and CO did the statistical analysis. EV and CO wrote the manuscript which was edited and discussed by DW, SW, and LP. All authors designed and ran the experiments.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.